06_境外及港澳台调查报告_part01.md

06_境外及港澳台调查报告 · 整合分片 part01

本片含 189 份报告。

FD06029 ｜ 2024_美国_NIOSH_face202402.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2024 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Page i

Career Engine Company Officer Dies in Abandoned, Single-Family Structure Fire – North Carolina Executive Summary On March 29, 2024, a 53-year-old career firefighter with the rank of captain (E3 captain) died from injuries sustained in an abandoned, single-family, structure fire after becoming lost within the structure on March 26, 2024. E3 captain was the company officer of the first due engine company making an interior fire attack. The single-family residential structure was an approximately 1,400 square feet, one story, wood frame construction. The structure was deemed abandoned by the local housing office and was boarded up prior to the incident. At approximately 0000 hours, the local communications center received a 9- 1-1 call reporting a fire in the structure from a bystander. At 0002 hours, the local communications center dispatched a box alarm assignment to the address for a reported structure fire. Engine 3 (E3), Battalion Chief 7 (BC7), Ladder 1 (L1), Engine 1 (E1), and Quint 2 (Q2) were dispatched. At 0005 hours, the communications center advised the responding fire companies that the local police department was on-scene advising the “backside of house is fully engulfed” and that it was “believed everyone is out of the house.” E3 arrived on-scene at 0006 hours and reported a working fire. The E3 firefighter stretched a 200’ 1¾ - inch attack line to the front door on side Alpha to prepare for an interior fire attack. L1 arrived on-scene at 0007 hours and split their crew of four. The L1 captain and a firefighter reported to side Alpha to meet with the crew from E3 and forced open the front door. E3 captain and firefighter along with the L1 captain and firefighter made entry into the structure and observed moderate, lazy, and brown smoke with no significant heat. No immediate fire conditions were encountered as the crew from L1 proceeded to conduct a left-hand search while the E3 crew advanced into the structure. BC7 arrived on-scene and assumed incident command (IC). IC repositioned for a better view of the active fire on side Delta at the side Charlie corner. REPORT F2024-02 • March 2025 Side Alpha post fire event (Courtesy of Fire Marshal Office)

Page ii Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 IC requested a water supply be established to E3 from E1 upon their arrival from a nearby hydrant, requested electrical utility services from the communications center, and requested on-scene fire companies switch to the tactical radio channel. The IC assigned Q2 as the rapid intervention crew (RIC) upon arrival and established an offensive strategy which was announced via the radio system on the tactical radio channel. At approximately 0009 hours, E1 was establishing a water supply and informed the IC that a 1¾- inch backup attack line had been stretched to side Alpha at the side Delta corner. The nozzle firefighter from E3 continued to advance into the structure approximately eight to ten feet before conditions began to change with zero visibility. L1 continued a left-hand search entering a bedroom while E3 captain reported he was going back to the entry door to “get more hose.” Both the L1 interior crew and E3 firefighter began to observe changing fire conditions in the interior with decreased visibility and increased temperatures. The firefighter from E3 began to flow water from the 1 ¾ - inch attack line towards side Charlie (rear) of the structure from the inside with no observed changes in fire conditions. At 0012 hours, the company officer from E1 informed IC that heavy fire was venting out the side Alpha door. IC acknowledged the radio message and observed a second room on side Delta flashover through the exterior windows of the structure. IC announced a change in strategy to a defensive mode and requested a personnel accountability report (PAR). IC was informed that L1 and E3 still had firefighters inside the structure when IC specifically requested that they withdraw from the structure. At 0014:57 hours, the apparatus operator from Q2 informed IC that crews were outside the structure but that E3 captain was unaccounted for. RIC informed the IC that E3 captain was not located on the exterior of the structure and RIC was deployed to the interior at 0017:24 hours. Approximately two minutes after the RIC deployment to the interior, the RIC could hear an audible personal alert safety system (PASS) device towards the rear of the structure (side Charlie). RIC located the E3 captain in the rear of the kitchen on side Charlie at the Bravo corner at approximately 0021 hours. E3 captain was removed by being drug to the front door on side Alpha. E3 captain was awake and alert upon removal from the structure with personal protective equipment (PPE) and self-contained breathing apparatus (SCBA) still intact. He was airlifted from the scene to a local burn center where he succumbed to his injuries three days later. Contributing Factors • Vacant and abandoned building marking system • Crew integrity • Two-way portable radio communications • Thermal Imaging (TI) device usage • Fire dynamics in relation to flow path • Strategy and tactics for transitional fire attack • Safety and survival training for self-rescue.

Page iii Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Key Recommendations Fire departments should: • Consider working with local officials to develop a vacant/abandoned building management and marking system. • Ensure that crew integrity is properly maintained by visual, direct, or verbal (voice or radio) contact when operating in an immediately dangerous to life and health (IDLH) atmosphere. • Develop policies and ensure that all personnel are trained on the features and use of portable radio operations, particularly as they relate to building muscle memory in daily response habits. • Ensure personnel understand the capabilities, use, and limitations of thermal imaging (TI) devices and utilize them during interior structural firefighting operations. • Develop or revise policies, strategies, and tactics to include modern research of fire dynamics and the movement of air and fire within a structure. • Develop policies, strategies, and tactics related to the exterior extinguishment or resetting of the seat of fire, slowing both vertical and/or horizontal fire spread prior to interior firefighting operations. • Develop training initiatives that focus on the knowledge, skill, and ability of firefighter safety/survival self-rescue training to include emergency egress and low-profile clearance.

The National Institute for Occupational Safety and Health (NIOSH) initiated the Fire Fighter Fatality Investigation and Prevention Program to examine deaths of firefighters in the line of duty so that fire departments, firefighters, fire service organizations, safety experts and researchers could learn from these incidents. The primary goal of these investigations is for NIOSH to make recommendations to prevent similar occurrences. These NIOSH investigations are intended to reduce or prevent future firefighter deaths and are completely separate from the rulemaking, enforcement and inspection activities of any other federal or state agency. Under its program, NIOSH investigators interview persons with knowledge of the incident and review available records to develop a description of the conditions and circumstances leading to the deaths in order to provide a context for the agency’s recommendations. The NIOSH summary of these conditions and circumstances in its reports is not intended as a legal statement of facts. This summary, as well as the conclusions and recommendations made by NIOSH, should not be used for the purpose of litigation or the adjudication of any claim. For further information, visit the program website at www.cdc.gov/niosh/firefighters/fffipp/ or call toll free 1-800-CDC-INFO (1-800-232-4636).

Career Engine Company Officer Dies in Abandoned, Single- Family Structure Fire – North Carolina

Page 1 REPORT F2024-02 • March 2025 Introduction On March 26, 2024, a 53-year-old career firefighter, holding the rank of captain, with 19 years in the department died while engaged in an interior fire attack at a single-family, one-story, wood frame structure fire. On April 1, 2024, the U.S. Fire Administration (USFA) notified the National Institute for Occupational Safety and Health (NIOSH) of this incident. On April 28 through May 2, 2024, two investigators representing the NIOSH Fire Fighter Fatality Investigation and Prevention Program (FFFIPP) traveled to North Carolina to investigate this incident. The NIOSH investigators met with the fire chief and staff, training division, inspections and investigations bureau, local city officials, and the 9-1-1 communications center. NIOSH also conducted interviews with fire officers and firefighters that responded to the box alarm incident. The NIOSH investigators inspected and photographed the deceased’s PPE and the SCBA. NIOSH investigators also reviewed the training records of specific personnel involved in the incident and reviewed the department’s standard operating procedures (SOPs) and professional development model. Fire Department The deceased firefighter (E3 captain) was part of a career fire department that has four stations with 50 uniformed personnel. This fire department serves a population of approximately 19,500 within a 20 square mile area and is led by a fire chief who is assisted by a deputy fire chief. The department has three battalion chiefs who oversee each of the three shifts. Each battalion chief also performs ancillary duties at the discretion of the fire chief such as training and safety. The fire chief and deputy chief respond off duty to incidents within the city that require additional support services. The four fire stations are geographically spread across the city limits and are staffed 24-hours a day. The fire department staffs two engine companies, one mid-mount ladder company, and a 75ft quint company used internally as a truck company. The two engine companies and quint are staffed daily with a minimum of three personnel while the minimum staffing for the ladder company is four personnel. The fire department participates in both automatic and mutual aid responses with neighboring departments throughout the county. Since this incident, the fire department has increased minimum staffing to four personnel on each apparatus daily after being awarded a Staffing for Adequate Fire and Emergency Response (SAFER) grant. The fire department does not provide transporting emergency medical services (EMS) but does respond to cardiopulmonary resuscitation (CPR) and lift assist incidents.

Page 2 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Training, Education, and Professional Development The E3 captain had extensive experience volunteering with other local fire departments around the area. He served as a fire instructor at the local community college and was actively involved in training and education within the region. The E3 captain had received International Fire Services Accreditation Congress (IFSAC) certifications in various courses to include Firefighter I and II, Instructor I and II, Inspector I, Fire Officer I and II, Driver Operator pump and aerial, and other courses. The minimum hiring qualifications for a position within the fire department include national certification in Firefighter I and II as well as Hazardous Material Operations Level. Qualifications for firefighters are largely left to the authority having jurisdiction (AHJ) with certifiable courses available through the Office of the State Fire Marshal. The fire department utilizes third-party software for continuing education and professional development, conducts weekly training, and offers training quarterly, specifically with mutual-aid partners and company officer development. The fire department is recognized by the state as a certified rescue department and conducts an annual hazardous materials drill. The department has a professional development plan to encourage the professional growth, training, and education of personnel looking to advance through the department rank structure. Apparatus, Staffing, and Communications Emergency calls are received into a local communications center which dispatches all emergency services for the county. The fire department utilizes an 800-megahertz (MHz) radio system for communications and response which is a part of a statewide public safety telecommunications system. The communications center is staffed with six dispatchers who are trained to a national standard and receive approximately 275,000 calls for service per year. No mobile data terminals (MDT’s) are utilized within the department for joint information sharing between units in the field and the communications center. Since this incident though, new technology has been acquired to provide a live stream from a 9-1-1 caller to dispatchers and responding personnel. A third-party notification system is used within the department to assist with dispatch location, hydrant locations, and information. Training is being adopted to incorporate the communication center dispatchers into fire-based scenarios and training. At the time of incident, each individual firefighter in the department was assigned a personal portable radio to be utilized while on-duty. Tables 1 and 2 below outline the responding apparatus to the box alarm including off duty chief staff to provide incident support.

Table 1 Box alarm assignment

Resource Designation Staffing level Engine 3 3 Ladder 1 4

Page 3 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Quint 2 4 Engine 1 3 Battalion Chief 7 1

Table 2 Off duty responding chief staff Resource Designation Staffing level Chief 6 (Deputy Chief) 1 Chief 1 (Fire Chief) 1

Personal Protective Equipment Upon arrival at the scene and during the incident, E3 captain was wearing his department issued structural firefighting turnout gear including coat, pants, gloves, Nomex hood, helmet, and boots that met the current requirement of NFPA 1971, Standard on Protective Ensembles for Structural Fire Fighting and Proximity Fire Fighting (NFPA 1971, 2018). The SCBA with an integrated personal alert safety system (PASS) device was approved by NIOSH, certified as meeting the 2013 edition of NFPA 1981, Standard on Open-Circuit Self-Contained Breathing Apparatus (SCBA) for Emergency Services, by a third-party laboratory (NFPA 1981, 2013). E3 captain was also wearing a fire department t-shirt and carrying a portable radio with a corded remote speaker microphone (RSM). The NIOSH investigators examined and photographed the PPE at the local Office of the Fire Marshal. Upon visual inspection of the PPE, extensive thermal degradation was noted to the turnout pants and coat, both on the outer shell and moisture barriers. This damage was more specifically to the left side, indicating that the E3 captain was laying prone with his right-side shielded during flashover. The Nomex hood, helmet, SCBA, and boots also suffered extensive thermal degradation to the left side. The SCBA facepiece had severe staining and discoloration with no visibility through the facepiece lens. However, it was intact and operational. The SCBA regulator and supply hose had deformities due to melting. However, both were still intact. Although, the PPE suffered extensive thermal degradation, exposure to extreme temperatures exceeding the PPE’s capabilities is not considered to be a contributing factor in this incident. Weather and Road Conditions The weather on March 25, 2024, at approximately 2354 hours, was reported as cloudy. This was approximately 8 minutes before the box alarm dispatch. The temperature was 48°F with winds east at 5 mph. The humidity was 50% with no precipitation noted. Weather conditions at the time of incident are not considered to be a contributing factor in the incident (Weather Underground, 2024).

Page 4 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Structure This single-family, non-sprinkled, residential structure was situated on a 0.61-acre lot located in an urban environment. Fire hydrant access to the property was available and utilized during the incident. The local government zoning office listed the structure as a 1,436 square foot, single-story residential home. The structure had electricity service from the local electrical grid system through the weather head of the structure. However, the local power company had blanked the electrical meter which prevented any electricity from entering the structure.

Photo: 1 Aerial view of structure with sides of the structure labeled A (Alpha), B (Bravo), C (Charlie), and D (Delta) (Courtesy of Google Earth) This structure had four-bedrooms and two-bathrooms. It was built in 1930 and was wood frame construction with a vinyl siding exterior. It contained 2x4 interior stud walls that were covered with both common drywall as well as lath and plaster interior finishings. The sloping roof structure was constructed of wood roof rafters covered in a metal roof exterior finishing. For this report, side Bravo is orientated to the north followed by side Charlie to the east, side Delta to the south, and side Alpha to the west (see Photo 1). NIOSH investigators were able to gain access to the structure to take photographs and measurements.

Page 5 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02

Diagram 1: Structure diagram (Courtesy of NIOSH)

Page 6 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02

Diagram 2: First floor diagram and approximate location E3 captain was found. Note: Diagram not to scale. (Courtesy of NIOSH)

Page 7 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Timeline The following timeline is a summary of events that occurred as the incident evolved. Not all incident events are included in this timeline. The times are approximate and were obtained by examining the dispatch records, audio recordings, witness statements, and other available information. The timeline is not intended, nor should it be used, as a formal record of events. Time (hours) Fireground Response, Operations, and Details 0000 • 9-1-1 call received by communications center indicating a house on fire • Local police discover the house fire and request the fire department for a “fully engulfed” house 0002 • E3, E1, L1, Q2, and B7 dispatched to structure fire 0004 • All fire companies enroute 0005 • Communications center advises responding fire companies that police are on-scene. Police advised backside of structure is “fully engulfed,” everyone believed to be out, and several calls received 0006 • E3 arrived on-scene, pulling past the address on side Alpha and advised “working fire” • Visible fire observed on side Delta at the side Charlie corner 0007 • L1 arrived on-scene on side Alpha in front of the structure • L1 split the crew of four into two teams (interior and exterior) • BC7 arrived on-scene, assumed IC, announced an offensive strategy, and requested all companies switch to the tactical radio channel • BC7 repositioned vehicle to have a view of side Delta but doesn’t have a clear line of sight to side Alpha. Incident command post (ICP) established inside the BC7 vehicle • IC informed E1 that they need a supply line to E3

0008 • E1 arrived on-scene and stretched approximately 200’ of 5-inch supply line from E3 to the hydrant just down and across the street from the scene • IC assigned Q2 as the RIC upon arrival 0009 • E1 to IC – Backup line stretched and charged to side Alpha • E3 and L1 made entry into the structure from the front door on side Alpha

Page 8 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Time (hours) Fireground Response, Operations, and Details 0010 • Outside vent firefighter informed IC that live wires were down at the side Bravo/Charlie corner • Q2 arrived on-scene • IC requested electric company response from communications center 0011 • Q2 informed IC that RIC was established • RIC was established at the side Alpha/Bravo corner • E1 company officer requested a 360 degree walk around from the Q2 company officer 0012 • E1 company officer informed IC that heavy fire was venting out the side Alpha door • IC acknowledged and agreed to a defensive strategy • E3 apparatus operator informed IC that a water supply was established 0012:31 • IC announced via the radio a change to a defensive strategy • IC observed a second bedroom on side Delta flashover through the windows 0013:01 • IC inquired via the radio if all on-scene companies were out of the structure • IC’s visibility is limited to side Delta with a partial view of side Alpha 0013:10 • IC was informed that L1 and E3 are still interior 0013:20 • IC ordered L1 and E3 via the radio to exit the structure • Fire and smoke conditions began to rapidly change, coming out of the door on side Alpha 0013:40 • E3 apparatus operator confirmed with IC that he ordered an evacuation • E3 apparatus operator began to sound the air horn on the apparatus to signal evacuation • E1 apparatus operator deployed a 200’ 2 ½ - inch line on side Delta and requested it be charged 0014:50 • IC asked if all companies were out of the structure

Page 9 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Time (hours) Fireground Response, Operations, and Details

• Visible fire was venting out the front door on side Alpha 0014:57 • Q2 apparatus operator advised command that E3’s company officer was still inside • Visible fire venting from the front door on side Alpha was being attacked by a 1 ¾-inch attack line • The interior 1 ¾ -inch attack line was removed from the structure with the nozzle firefighter during the evacuation 0015:30 • L1 company officer called for E3 captain on the radio with no response 0015:39 • L1 company officer asked over the radio if anyone had eyes on the E3 captain 0016:07 • IC ordered the 1 ¾-inch back-up attack line to the front porch on side Alpha • IC left the ICP to gain a better view of side Alpha 0017:08 • RIC advised IC that the E3 captain still had not been located on the exterior of the structure • RIC worked along the exterior of side Bravo, utilizing a TI device and looking into windows trying to locate E3 captain 0017:24 • IC deployed the RIC to the interior from side Alpha • RIC deployed and made entry on side Alpha with a 1 ¾-inch attack line • E1 company officer joined the RIC crew upon entry 0018:03 • IC notified the communications center of a Mayday on the main radio channel • Communications center rebroadcasted the Mayday declaration on the main radio channel 0018:59 • IC requested the next three closest departments respond to the scene 0019:16 • Chief 6 on-scene

Page 10 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02

Time (hours) Fireground Response, Operations, and Details 0019:36 • E1 company officer advised IC they could hear a PASS device in the rear of the structure • RIC encountered several rooms of fire and were searching as they continued firefighting towards side Charlie 0020:51 • RIC confirmed with the IC that they were still working towards side Charlie and could still hear a PASS device • RIC continued to encounter fire throughout the structure while utilizing their 1 ¾-inch attack line 0021:31 • RIC to IC – “I got him, I got him, I got him” • RIC confirmed the E3 captain was breathing and SCBA was functioning. RIC silenced the PASS device on E3 captain’s SCBA 0022:19 • RIC announced they had E3 captain on side Bravo at the windows and needed a ladder • RIC was working in zero visibility and attempted to utilize the Rapid Intervention Universal Air Connection (RIC-UAC) J’P-8from the RIC kit 0023:44 • RIC informed the IC to have EMS ready at the door of side Alpha • RIC changed extrication plans when they heard the low air alarm activate on E3 captain’s SCBA 0024:28 • L1 company officer confirmed with IC that all companies were outside the structure • E3 captain was awake and talking after the doffing of his SCBA 0024:38 • IC informed the communications center that the missing firefighter had been removed from the structure 0046:08 • E3 captain transported from the scene to the landing zone 0058:53 • Medical helicopter departed the landing zone to transport the E3 captain to a local trauma center

Page 11 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Investigation On March 26, 2024, at 0000 hours, the local communications center received a 9-1-1 call reporting a house on fire. The 9-1-1 caller told the communications dispatcher upon questioning that “nobody is inside” and that the “back of the house” was on fire. Simultaneously, a patrol officer from the local police department radioed the communications center to start the fire department for a “fully engulfed” structure fire but was later corrected to the fire being in the rear (side Charlie) of the structure. The communications center dispatcher instructed the 9-1-1 caller to follow standard prearrival instructions to include not entering the structure and to safely keep bystanders free of the area. The communications center entered the information into the computer aided dispatch (CAD) system and initiated the box alarm assignment for the location. Box alarm fire companies were dispatched at 0002 hours consisting of E3, E1, L1, Q2, and BC7. Assigned fire companies were enroute by 0004 hours (see Table 3). At 0005 hours, the communications center advised the responding fire companies that the local police department was on- scene, and advising the rear of the structure was “fully engulfed” and with everyone believed to be out of the structure. During the fire department response, the local police department worked to gather information from bystanders on the scene and began to control traffic within the area while leaving access for arriving units. Table 3 On-scene resource assignments Apparatus Assignment BC7 Incident Command E3 Interior Fire Attack – 200’ 1 ¾-inch attack line L1 Interior search and rescue / outside vent crew E1 Backup attack line side Alpha – 200’ 1 ¾-inch attack line Q2 RIC – side alpha/bravo corner

E3 (apparatus of the deceased) arrived first on-scene at 0006 hours. E3 pulled past the location to leave access for the ladder company and provided an on-scene report of “working fire.” No further information or size-up information was provided by the E3 captain to incoming fire companies. The firefighter from E3 stretched a 200’ 1 ¾ -inch attack line to the front door on side Alpha as L1 arrived at 0007 hours. L1 split their four-person crew into two working groups for both interior and exterior operations. The company officer and a firefighter from L1 reported to the front door on side Alpha and utilized traditional forcible entry techniques to force open and control the door while the apparatus operator and firefighter (outside crew) from L1 checked utilities, assisted with the large diameter

Page 12 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 supply line, and prepared L1 for aerial operations. Before making an interior offensive fire attack through side Alpha, E3 captain returned to E3 to retrieve the TI device from the apparatus. At 0007 hours, BC7 arrived on-scene, assumed IC, announced an offensive strategy via the radio, and requested all on-scene personnel switch to the tactical channel per department policy. BC7 had to reposition his department vehicle to better observe the fire conditions on side Delta but had limited visibility to side Alpha. The ICP was established inside BC7’s vehicle. At 0008 hours, IC requested that E1 establish a water supply from a nearby hydrant. E1 stretched approximately 200’of 5-inch supply line from E3 to secure the water source. IC assigned Q2 as the RIC and set up at the side Alpha/Delta corner. The E3 crew (company officer and firefighter) and L1 crew (company officer and firefighter) made entry into the structure from side Alpha at approximately 0009 hours and observed brown smoke with limited visibility throughout the structure. No visible fire was observed upon entry as L1 proceeded to conduct a left-hand search while E3 pushed forward into the structure. E1 advised the IC that a backup 1 ¾ -inch attack line had been stretched and charged to side Alpha. L1 exterior crew advised IC that live wires were down at the service drop to the structure at the side Bravo/Charlie corner. IC acknowledged this message and requested utility services to the scene.

Diagram 3: Apparatus, attack, supply line placement. Note: Diagram not to scale. (Courtesy of NIOSH)

Page 13 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 At approximately 0011 hours, the company officer from E1 requested the RIC (i.e., Q2) conduct a 360- degree walkaround of the structure which was acknowledged. E3’s captain and firefighter continued to advance into the structure a few feet, with both periodically stopping to check for accountability. The E3 firefighter began to experience changes in both smoke and temperatures with visible fire being seen inside the structure. He began to flow water towards the rear of the structure when he heard E3 captain yell that he was going to get more hose. At 0012 hours, the company officer from E1 advised IC that heavy fire was venting out the side Alpha door. IC acknowledged the message and observed a middle room in the structure on side Delta flashover through the windows. The IC announced a change in strategy on the radio to a defensive operation. This message was acknowledged by the apparatus operator of E3 who began to sound the airhorns to indicate to interior crews that they needed to withdraw from the structure. A 200’ 2 ½”-inch smoothbore attack line was then stretched to side Delta. At 0013:01, IC inquired if all companies had exited the structure when he was informed that L1 and E3 interior crews were still inside the structure. IC called both L1 and E3 captains on the radio and instructed them to evacuate the house. However, this message was never acknowledged by E3 captain. As noted in interviews conducted by NIOSH, the fire conditions at this point began to rapidly deteriorate both on the interior and exterior. At 0014:50 hours, the IC asked again if all companies had been able to evacuate the structure. The apparatus operator of Q2 advised IC that E3 captain was still inside the structure. L1 company officer and firefighter, along with the nozzle firefighter from E3, exited the structure on side Alpha. Personnel from E1 were using the 1 ¾- inch backup attack line to knock back the fire that was coming out the side Alpha door. The initial 1 ¾ - inch attack line was removed out of the structure by the nozzle firefighter upon his exit. The company officer from L1 called for E3 captain directly on the radio at 0015:30 hours with no answer. L1 company officer then asked all on-scene companies if anyone had E3 captain in their sight. At 0016:07 hours, the IC ordered that the backup line be placed into service at side Alpha and the IC left the ICP to gain a better view of the operations on side Alpha. RIC advised IC that they had scanned the windows alongside Bravo and the exterior of the structure but were unable to account for the E3 captain. IC deployed the RIC including the company officer from E1 to the interior of the structure. RIC deployed with the 1 ¾-inch attack line that was obtained from the E3 firefighter who had exited the structure. IC notified the communications center on the main fire channel of a Mayday deployment at 0018:03 hours and requested the next closest three fire companies to respond to the scene. E1 company officer (now attached to the RIC) advised at 0019:36 hours that a PASS device could be heard towards the rear (side Charlie) of the structure from the interior. RIC encountered several rooms of active fire as they continued to advance down the center hallway towards side Charlie. RIC provided a status report at 0020 hours to the IC that they were still working their way to side Charlie. At 0021:31 hours RIC transmitted to IC, “I got him, I got him, I got him.” RIC was able to confirm the E3 captain was breathing and awake with his SCBA appearing to be functioning properly. RIC silenced the PASS device on the SCBA and advised they were near some windows on side Bravo. They requested a ladder to the window. RIC was working in zero visibility and attempted to use the

Page 14 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 RIC/UAC in conjunction with the RIC kit. At approximately 0023 hours, RIC drug the E3 captain out of the structure through the hallway and requested that EMS be ready on side Alpha. L1 informed IC that all personnel including E3 captain were out of the structure on side Alpha at approximately 0024 hours. IC informed the communications center at 0024:38 hours that the missing firefighter had been located and removed from the structure. E3 captain was awake and talking upon being removed from the structure. He was treated by on-scene EMS personnel before being transported to a nearby landing zone at 0046:08 hours. He was air lifted to a trauma center at approximately 0058 hours. Unfortunately, E3 captain later succumbed to the injuries sustained during the incident three days later. Fire Origin and Cause At the time of this report being published, the Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) field office and investigating partners noted that the origin, cause, and classification were still under investigation. Cause of Death The Office of the Medical Examiner listed the cause of death as complications from thermal cutaneous injuries, partial and full thickness burns, anasarca, and cerebral edema. Contributing Factors Occupational injuries and fatalities are often the result of one or more contributing factors or key events in a larger sequence of events that ultimately result in the injury or fatality. NIOSH investigators identified the following items as key contributing factors in this incident that ultimately led to the fatality: • Vacant and abandoned building marking system • Crew integrity • Two-way portable radio communications • TI device usage • Fire dynamics in relation to flow path • Strategy and tactics for transitional fire attack • Safety and survival training for self-rescue. Recommendations Fire departments should: Recommendation #1: Consider working with local officials to develop a vacant/abandoned building management and marking system. Discussion: The local housing authority knew this structure was an abandoned property. The property was boarded up on the exterior to prevent unauthorized access. However, some of the windows had

Page 15 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 been broken over the years. The structure had experienced a fire event years before this incident but was structurally intact with intact dry wall and lath and plaster interior finishings. The home was not legally occupied at the time of incident. It contained minimal debris and living furnishings, and it was not serviced by the local electrical utility service. The city government and fire department did not have a collaborative approach to managing and marking abandoned structures. The AHJ is responsible for training and certifying inspection and code enforcement personnel to ensure that the appropriate fire and building safety codes are utilized and enforced. Based upon jurisdiction, fire and life safety inspection programs may be in the fire, building, or code enforcement departments. In some fire departments, this responsibility is performed by firefighters, fire prevention divisions, or fire officers. When the authority to perform fire inspections and code enforcement is not assigned to the fire department, the fire department should develop a working relationship with the department that has this authority [IFSTA, 2016]. NFPA 1201, Standard for Providing Fire and Emergency Services to the Public, Section 6.3 states, “the fire and emergency service organization (FESO) shall seek to establish a good working relationship with the agency or authority responsible for enforcing the building code so that the review of the design, construction, alteration, or demolition of buildings and structures can be monitored to identify fire protection concerns.” This includes ensuring the investigation of fires and other emergencies to assist with developing an effective hazard and risk prevention program [NFPA 1201, 2020]. This also includes inspections, permits, code violations, fire protection maintenance and service, and any other vital information. When fire departments and fire leadership maintain a familiarity with their AHJ’s local ordinances, this enhances their ability to have input in renovations and use group changes for occupancy types within their jurisdiction. NFPA 1030, Standard for Professional Qualifications for Fire Prevention Program Positions, Chapter 7 provides the job requirements for the position of Fire Inspector. It states that the duties of this position involve conducting fire and life safety inspections, to include enforcement actions, and analyses of new and existing structures and properties for construction, occupancy, fire protection, and exposures. This includes identifying occupancy classifications, evaluating fire protection systems and equipment, and recognizing hazards and deficiencies according to applicable policies, codes, and standards adopted and maintained by the AHJ [NFPA 1030, 2024]. A hazardous building management program may reduce the risks posed by these buildings and ideally composed of three phases to include determination, remediation, and marking. Phase 1 – Determination: Determination consists of identification, evaluation, designation, and communication. • Identification: Personnel physically locate and evaluate the potential hazardous buildings. A common challenge encountered is the ability to legally access the property, especially in the case of private and residential properties where the fire department would not normally have the legal authority to access the property outside of an emergency incident. It is critical that anyone attempting to access a property fully understands their legal standing to avoid possible

Page 16 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 criminal charges and hostile confrontations with property owners. Fire department personnel should work with the local AHJ to legally gain access to the properties as warranted. • Evaluation: Personnel responsible for a building management program must have the subject matter expertise to evaluate the condition of the buildings. Beyond determining if a building is vacant or abandoned, they need to have the necessary expertise to determine if it has structural integrity defects or conditions that could warrant a status of being formally condemned. Fire department personnel must exercise caution and take all necessary preventative measures to ensure their safety during the evaluation process. Hazardous buildings can present several hazards outside of the structure itself, to include the presence of unauthorized people, concealed traps, and hazardous materials. If there are questions as to the structural integrity of a building, the fire department should utilize outside subject matter expertise as needed to ensure that they can safely enter and evaluate the building. • Designation: Personnel determine the status of the building upon completion of the evaluation. The fire department can determine the status of the building for the purposes of their risk/benefit analysis during an incident. However, the fire department may not have the legal authority to formally condemn a building unless it poses an imminent risk to public safety. This is another part of the process where it is critical to work with the local AHJ if there is a need to have a building formally condemned so that it can be demolished and removed through the appropriate means. • Communication: After the status of the building has been determined, it should be effectively communicated to all parties involved (property owner, building dept, fire marshal, fire department, dispatch center, etc.).

Phase 2 – Remediation: Remediation includes demolition, marking, and securing.

• Demolition: For a derelict or condemned building, the best approach through risk management principles is to remove the hazard by having the building demolished. If a building remains, it may still be inhabited by unauthorized occupants, and it may still catch fire, forcing a fire department response. Most fire departments lack the necessary legal standing or resources needed to carry out the demolition of a building, so it is critical for them to form strategic partnerships with local agencies who can assist. • Marking: Proper marking benefits the responding fire department in recognizing the building status. There are several references that can be used for guidance when it comes to developing and implementing a building marking system.

As an example, the International Association of Arson Investigators and United States Fire Administration Abandoned Building Project utilizes a simple system consisting of a box with a diagonal line to indicate the condition of the building. The abandoned building project toolbox can be found at http://www.interfire.org/features/AbandonedBuildingProjectToolBox.asp.

Page 17 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02

Figure 1. High risk building marking system (Courtesy of United States Fire Administration) The 2021 International Fire Code (IFC) Appendix J Building Information Sign (International Fire Code, 2020) and the NFPA 1 Fire Code, Annex C - Fire Fighter Safety Building Marking System (Zevotek, 2022) utilize a more elaborate and detailed marking system based on a Maltese Cross that goes beyond just indicating the status of the building.

Figure 2. Updated and detailed high risk building marking system (Courtesy of NFPA) Although marking buildings can have a benefit, there can be localized challenges to implementing this action based on unique circumstances. Some fire departments have run into challenges when trying to implement these programs due to concerns over the negative stigma attached to these markings when they are concentrated in a given area or neighborhood. Therefore, it may be useful to form strategic partnerships with local community partners to explain the lifesaving benefits of implementing a marking system.

• Securing: If unauthorized individuals can be prevented from entering the building, it may eliminate the possible ignition source of a fire and prevent the fire department from having to enter for search & rescue operations. Securing the building may include both physical barriers

Page 18 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 to deny access and partnering with local law enforcement or private security agencies to identify and prevent unauthorized access. Securing a building can have pros when it comes to preventing unauthorized access; however, it can also have cons if the fire department needs to enter the building. As such, the fire department needs to be aware of how to access and remove any security features established.

Phase 3 – Emergency Response: Emergency response includes recognition and risk management.

• Recognition: Through various means available (dispatch information, CAD notes, building markings, and/or effective situational awareness), responding personnel must have the ability to proactively recognize the status of a hazardous building as soon as possible. A multilayered system with repetitive information is best if a failure occurs with one of the layers. For example, if a building marker ends up missing, but the information is relayed at dispatch from a CAD note, the responders will still be aware of the hazardous building’s status. • Risk Management: The key decision makers (1st arriving officer, IC, safety officer) need to incorporate the status of the hazardous building into their risk/benefit analysis and IAP. This includes recognizing both the “status” of potential occupants and the building (NIOSH, 2024).

Recommendation #2: Ensure that crew integrity is properly maintained by visual, direct, and/or verbal (voice or radio) contact when operating in an IDLH atmosphere. Discussion: In this incident, E3 captain chose an offensive strategy that was reinforced and transmitted via the radio by BC7 upon arrival and establishing IC. The crew (captain and firefighter) from E3 proceeded to stretch an interior attack line and made entry with L1 (captain and firefighter) through the front door on side Alpha. During the initial stretch of the 1 ¾-inch attack line, the E3 captain told the nozzle firefighter that he was returning to the front door “to get more hose.” As fire conditions changed, the nozzle firefighter of E3 was unable to establish voice communications with E3 captain. During the investigation, NIOSH was unable to find any witnesses who observed E3 captain making it back to the front door.

Additionally, upon the arrival of BC7 and repositioning the BC7 vehicle for a better view of side Delta, the IC was unaware that either E3 or L1 made entry into the structure. There were no radio communications from either L1 or E3 about making entry into the structure. At 0012 hours when E1 informed the IC of changing fire conditions venting outside side Alpha through the door, the IC was unaware that E3 and L1 crews were interior. The IC specifically ordered both L1 and E3 company officers via the radio to exit the structure. The withdrawal order was never acknowledged via the radio by E3 captain. It was later determined through investigation that the portable radio being carried by E3 captain was never turned on during this incident nor did it ever register in the portable radio log from the communications system (see recommendation #3). Crew integrity is essential to fireground accountability. Fire departments should have personnel accountability policies and procedures in place to track crews and individuals. A department’s accountability system should have the capabilities to track and identify who is operating in the hazard zone, their assignment, and location. The collection of personal accountability tags or passports placed

Page 19 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 on an apparatus does not assist in maintaining awareness to the fireground location of crews or individuals. NFPA 1550 Standard for Emergency Responder Health and Safety, Section 17.7 provides guidance on accountability for all resources, including personnel (NFPA 1550, 2024): • Specific means (Passport tags, software, etc.) to identify and track responders entering hazardous areas • Accounting for personnel who arrive on-scene by other means besides apparatus • Providing a process to rapidly account for all responders on-scene. Recommendation #3: Develop policies and ensure that all personnel are trained on the features and use of portable radio operations particularly, as they relate to building muscle memory in daily response habits. Discussion: In this incident, the portable radio issued to E3 captain was never turned on nor did it ever register on the radio log during the incident, preventing him from hearing any of the fire scene audio about rapidly changing fire conditions, change in strategy from offensive to defensive, and the request from the IC to evacuate the structure. There were no radio transmissions made from E3 captain after giving an initial on-scene radio report made from the mobile radio inside the cab of E3. Two-way portable radio communications are a critical component of the firefighter’s complete PPE. Personnel should build muscle memory and consistency in incorporating their portable two-way radio device into the appropriate donning of assigned PPE and turning on the portable radio once on-scene in accordance with department policy. In a position paper written by the International Association of Fire Chiefs (IAFC), they state, “Information from outside can prevent crews from advancing into dangerous conditions inside” (IAFC, 2009). Two-way portable radio communications enhance general situational awareness for first responders, confirm and communicate strategy and tactics, and most importantly, allow for firefighters to transmit distress signals and communications such as declaring a Mayday or an urgent message. If not already in place, fire department policies should outline response communications and training should be developed so that personnel understand the features and capabilities of two-way radio communications. NFPA 1802 Standard on Two-Way, Portable RF Voice Communications Devices for use by Emergency Services Personnel in the Hazard Zone provides examples of features that are incorporated into two-way portable radios that may be helpful to personnel to be aware of and trained on to include (NFPA 1802, 2021): • Power/volume knob • Selector knob • EAB (emergency alert button) • Time-out timer • Out-of-range Detection.

Page 20 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Recommendation #4: Ensure personnel understand the capabilities, use, and limitations of thermal imaging (TI) devices and utilize them during interior structural firefighting operations. Discussion: In this incident, E3 captain returned to the engine company after the initial attack line was stretched to side Alpha to retrieve the TI device assigned to E3. During the initial fire attack, it is unknown as to whether he ever used the TI device. Following the Mayday and removal of E3 captain to side Alpha, the TI device was discovered just inside the front door on side Alpha. The TI device was not found with the E3 captain, on his PPE, or in the room where he was found. The Western Fire Chiefs Association noted that TI devices are helpful in many situations while improving both speed and effectiveness (Western Fire Chiefs Association, 2024). While the fire service provides many opportunities for TI device usage such as hazardous materials, search and rescue, fire attack, etc., one of the primary components to TI use is to maintain broad situational awareness to the conditions inside zero visibility environments. The ability to detect changes in fire conditions, flow paths, and egress can all be applicable reasons to use TI during interior firefighting. Personnel should be trained to specific characteristics, features, modes, and overall capabilities of a department’s specific TI device. Fire departments should use caution when selecting TI devices that are incorporated into other PPE components like SCBA and helmets to ensure the technology meets NFPA 1801 Standard on Thermal Imagers for the Fire Service if being utilized as the primary TI device. Personnel should be mindful to incorporate the use of TI where available and build the muscle memory needed when TI devices are incorporated into the daily wearing and use of PPE ensembles. Recommendation #5: Develop or revise policies, strategies, and tactics to include modern research of fire dynamics including the movement of air and fire within a structure. Discussion: In this incident, E3 approached the scene from the north traveling south and was able to observe sides Bravo, Alpha, and Delta upon arrival. Visible and well-developed fire was observed on side Delta at the Charlie corner. E3 provided an on-scene report of “working fire.” The E3 firefighter stretched a 1 ¾ - inch attack line to side Alpha as the E3 company officer chose an interior fire attack. Conditions on side Alpha didn’t reveal anything significant to note with only light smoke around the porch area. The crew from E3 met with the L1 crew to utilize traditional forcible entry techniques to gain access to the interior on side Alpha. The front door on side Alpha remained open as both E3 and L1 made entry into the structure for an interior fire attack. E3 and L1 interior crews had been inside the structure for approximately four minutes when the E1 company officer relayed the message to the IC that fire was venting out the front door on side Alpha. Within this four-minute time frame the fire had been drawn from the side Delta/Charlie corner through the structure and out the front door. Upon investigation, the abandoned structure was largely intact with limited combustible furnishings. A flow path is “the movement of heat and smoke from the higher pressure within the fire area towards the lower pressure areas accessible via doors, window openings, and roof structures” (NFPA Glossary of Terms, 2021). Flow paths consist of an inlet and an exhaust with the direction of travel being determined by pressure. Heat and smoke in a high-pressure area will travel to an area of lower

Page 21 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 pressure. It is possible to have multiple flow paths within a structure dependent upon the size of the building, openings, closures such as fire doors, and overall structure design. Personnel working in the flow path (between the seat of the fire and the exhaust) operate in a significantly risky environment. Coordinated ventilation tactics should be utilized to redirect the flow path from interior operations. The International Fire Service Training Association notes that firefighters advancing hose lines or ventilating windows create newly formed flow paths between the seat of the fire and ventilation points to the exterior (IFSTA, 2018).

Figure 3. IFSTA Essentials of Firefighting 7th Edition Figure 4.47 (Courtesy of IFSTA) When identifying the flow path within a structure, personnel should be mindful to wind conditions. Wind can create pressure changes within a structure or building which could spread smoke and fire into non impacted parts of the structure or building. Interior firefighting personnel should be mindful to these changes in conditions that place the interior firefighting personnel in the exhaust portion of the flow path. Sudden changes in temperature and smoke are factors that should alert interior personnel of imminent changes to conditions. Water should be applied to the fire as quickly as possible utilizing the reach of an effective fire stream that is delivering an appropriate volume of water that will overcome the heat release rate. In some cases, it may be appropriate to introduce water from the exterior of a structure through a window/door with an attack line or begin the fire attack from the exterior with a coordinated blitz attack flowing 300 gallons per minute or greater before committing personnel to the interior. Fire departments should cool interior spaces from the safest and closest location possible, especially in vent limited spaces.

Recommendation #6: Develop policies, strategies, and tactics related to the exterior extinguishment or resetting of the seat of fire, slowing both vertical and/or horizontal fire spread prior to interior firefighting operations.

Page 22 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Discussion: In this incident, the fire began inside the abandoned structure and was immediately discovered and reported to the 9-1-1 communications center. A local law enforcement officer on patrol also quickly discovered the structure fire and requested the fire department. Upon arrival of E3, a well- developed fire was burning on side Delta at the Charlie corner that had vented to the exterior of the structure. Both the 9-1-1 caller and local law enforcement officer provided information that the structure was believed to be empty of any occupants and this information was relayed to responding fire companies. Fires venting from the exterior with both vertical and horizontal spread can present a hazard to arriving fire department personnel. First arriving company or chief officers on-scene should be mindful to the many factors involving fire dynamics when considering the appropriate strategy for the incident. Paraphrased from NFPA 1700 Guide for Structural Firefighting, Chapter 9 (NFPA 1700, 2021), these include: • Smoke and fire conditions – Read the smoke based on volume, velocity, color, and density • Fuel load – Increased fuel loads may increase higher heat release rates • Openings – Windows, doors, and structural openings that can provide a pathway for fire extension • Flow path – Air and gases move from higher pressure to lower pressure via doors and windows within a structure • Weather Conditions – Wind should be evaluated to recognize the hazards of a wind-driven fire • Accessibility of the structure – Apparatus placement, water supply, and access for fire attack and ground ladders • Fire progression – 360-degree walkaround predictions made to fire spread while considering fire dynamics and outcomes of fire attack and ventilation • Fire Control Positioning – Location of the seat of the fire, survivability profiling, and access.

Consideration should be given to a transitional attack when the first arriving company or chief officers on-scene and observe active and well-developed fire accessible from or involving the exterior of a structure that is spreading vertically, horizontally, or both with survivable space within the structure. Company or chief officers should include the type and occupancy status of a structure and a risk/benefit analysis when making tactical decisions. A transitional attack is “the application of a fire stream from the exterior of a structure to improve conditions prior to interior fire control” (NFPA Glossary of Terms, 2021). Recommendation #7: Develop training initiatives that focus on the knowledge, skill, and ability of firefighter safety/survival self-rescue training to include emergency egress and low-profile clearance. Discussion: In this incident, E3 captain made an interior fire attack through the front door on side Alpha. After the nozzle firefighter advanced a short distance into the structure, E3 captain informed the nozzle firefighter that he “was going to get more hose.” E3 captain was never observed returning to the front door on side Alpha nor was any hose advanced into the structure. E3 captain had become lost and disoriented within the structure and was found in the rear (side Charlie at Bravo) corner by the RIC

Page 23 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 after a Mayday had been declared by L1 on behalf of E3 captain. It is unknown what actions E3 captain performed to self-rescue. However, two smaller windows were present in the laundry room where he was found by RIC. It can be concluded from interviews and photographs that E3 captain assumed a prone position on the floor and experienced most of the thermal assault to his left side. Interior structural firefighting personnel should be trained to both recognize and avoid, when possible, the following hazards as listed in the International Fire Service Training Association (IFSTA) Essentials of Fire Fighting (IFSTA, 2024): • Air emergencies • Lost/disoriented • Entanglement • Rapid fire development • Collapse/trapped.

Further, the following skills should be incorporated into a firefighter’s self-rescue training program. This list is not intended to be all inclusive but serve as general guidance. • Declaring Maydays – Location, Unit, Name, Air, Resources, and PASS device activation • Reduced profile maneuvers • Window egress techniques • Emergent ladder egress/escape • Emergency bail out systems • Disentanglement from debris and wires • Wall breaches.

The International Association of Fire Fighters (IAFF) (FGS) cadre notes “Fire fighters must be trained to perform potentially life-saving actions if they become lost, disoriented, injured, low on air, or trapped” (IAFF, 2024). Funded in part by a grant from the U.S. Department of Homeland Security and the Assistance to Firefighters (FIRE Act), the IAFF offers both online and on-site training related directly to firefighter survival training and tactics. A 2023 article posted to FireRescue1 provides some realistic training examples and drills that can assist fire departments in training for declaring Maydays, orientation, props, low air emergencies, and trapped firefighter simulations. Andrew Beck, firefighter/EMT and author of this FireRescue1 article notes that readers should “Define your parameters, practice the procedures, and ensure it’s realistic. You will prepare your folks to succeed when it’s truly life or death” (Beck, 2023). References Beck, A. [2023]. Practice makes automatic: Simple mayday training to instill good habits. Retrieved September 2024, from FireRescue1. IAFC [2009]. Position Statements. Retrieved August 2024, from International Association of Fire Chiefs.

Page 24 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 IAFF [2024]. Fire-Ground-Survival. Retrieved 2024, from International Association of Fire Fighters. IFSTA [2016]. Fire Inspection and Code Enforcement. 8th ed. Stillwater, OK: International Fire Service Training Association. IFSTA [2018]. Fire Dynamics. In O. S. University, Essentials of Firefighting 7th Edition (p. 162). Stillwater, OK: International Fire Service Training Association. IFSTA [2024]. Firefighter Survival During Interior Operations. In Essentials of Fire Fighting (Eighth ed., p. 456). Stillwater, OK: International Fire Service Training Association. International Fire Code [2020]. Appendix J Building Information Sign. Retrieved June 2024, from International Fire Code. NFPA [2013]. NFPA 1981, Standard on Open-Circuit Self-Contained Breathing Apparatus (SCBA) for Emergency Services. Quincy, MA: National Fire Protection Association. NFPA [2018]. NFPA 1971, Standard on Protective Ensembles for Structural Fire Fighting and Proximity Fire Fighting. Quincy, MA: National Fire Protection Association. NFPA [2020]. NFPA 1201, Standard for providing fire and emergency services to the public. Quincy, MA: National Fire Protection Association. NFPA [2021]. NFPA 1700, Guide for Structural Fire Fighting Chapter 9. Strategic Considerations. Quincy, MA: National Fire Protection Association. NFPA [2021]. NFPA 1802, Standard on Two-Way, Portable RF Voice Communications Devices for use by Emergency Services Personnel in the Hazard Zone. Quincy, MA: National Fire Protection Association. NFPA [2024]. NFPA 1030, Standard for professional qualifications for fire prevention program positions, Quincy, MA: National Fire Protection Association. NFPA [2024]. NFPA 1550, Standard for Emergency Responder Health and Safety Section 17.7. Quincy, MA: National Fire Protection Association. NFPA Glossary of Terms [2021]. Quincy, MA: National Fire Protection Association. Weather Underground [2024]. Greensboro, NC Weather History. Retrieved July 2024, from wunderground.com. Western Fire Chiefs Association [2024, July 9]. Thermal Imaging Camera (TIC) in Firefighting. Retrieved September 4, 2024, from Western Fire Chiefs Association. Zevotek, R. [2022, July 15]. Firefighter Safety Building Marking Systems. Retrieved June 2024, from National Fire Protection Association.

Page 25 Career Engine Company Officer dies in Abandoned, Single-Family Structure Fire – North Carolina Report # F2024-02 Investigator Information This incident was investigated by Patrick R. Montague (former), Investigator and Louis (Rick) Lago (former), Investigator, both with the Fire Fighter Fatality Investigation and Prevention Program, Surveillance and Field Investigations Branch, Division of Safety Research, NIOSH located in Morgantown, WV. The report was written by Patrick R. Montague (former), retired Battalion Chief, and credentialed Chief Fire Officer. An expert technical review was provided by Timothy S. Cowan, Deputy Fire Chief, Dewitt Fire District in Central New York. A technical review was also provided by the National Fire Protection Association, Emergency Response and Responder Safety Division. Disclaimer The information in this report is based upon dispatch records, audio recordings, witness statements, and other information that was made available to the National Institute for Occupational Safety and Health (NIOSH). Information gathered from witnesses may be affected by recall bias. The facts, contributing factors, and recommendations contained in this report are based on the totality of the information gathered during the investigation process. This report was prepared after the event occurred, includes information from appropriate subject matter experts, and is not intended to place blame on those involved in the incident. Mention of any company or product does not constitute endorsement by NIOSH, Centers for Disease Control and Prevention (CDC). In addition, citations to websites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or products. Furthermore, NIOSH is not responsible for the content of these websites. All web addresses referenced in this document were accessible as of the publication date.

FD06030 ｜ 2024_美国_NIOSH_face202403.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2024 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Page i Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Executive Summary On May 30, 2024, a 33-year-old career firefighter (E31FF1), was seriously injured during flashover while conducting search operations in a one-story residential structure. A second firefighter (E31FF2), partnered with E31FF1, also suffered injuries during the event. At 21:02 hours, multiple 9-1-1 calls were received reporting a structure fire. Several of the calls reported occupants in the structure. The following units were dispatched: Photo 1: Side Alpha of the structure post-fire. (Courtesy of the fire department) Engine 31 (E31), Engine 74 (E74), Engine 73 (E73), Engine 71 (E71), Truck 71 (T71), and Battalion Chief 7 (BC7). Rescue 31 (R31) was also added. At 21:07 hours, E31 arrived on-scene. The E31 officer gave a quick size-up and assumed Command. E31’s crew, E31FF1 and E31FF2, were assigned fire suppression. As they were preparing to enter the structure, multiple civilians indicated a child with disabilities was trapped in the home. They decided to make a quick entry without a hoseline from the Side Alpha entry door. R31 arrived on-scene and was assigned to proceed with establishing a backup hoseline. Command reported E31’s change of assignment confirming they were not in possession of a hoseline when they entered the structure. Command then ordered R31 to perform fire suppression; however, there was a delay in R31 initiating a fire attack. The occupant approached R31’s crew frantically telling them her child was in the structure. As an E31 crew entered the structure, they traveled down a center wall and made a right turn to a hallway. As they proceeded down the hallway, they encountered a bedroom on the right side (Side Alpha). E31FF2 entered the first bedroom on Side Alpha while E31FF1 proceeded to look in the second bedroom located on Side Alpha/Delta corner. Conditions were rapidly changing and E31FF1 informed E31FF2 that “we need to get out of here.” During this time, an R31 crew retrieved and put a hoseline in service outside the structure, but there was no water pressure. The preconnected hoseline was kinked at the REPORT F2024-03 • April 2025

Page ii Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 discharge located on the right side of the pumper. As E31’s crew started to exit the structure they traveled down the hallway toward Side Bravo. A flashover occurred with fire traveling throughout the entire interior open space consisting of the living room, dining area, and kitchen. When E31FF2 got to the end of the hallway he turned left, leading him to the Side Alpha entry door exiting the structure. However, E31FF1 became disoriented and traveled across the living room toward the Side Bravo wall. Becoming more confused and disoriented he reversed direction, returning down the hallway, until he got to the end. He reentered the bedroom at the Side Alpha/Delta Corner which he was searching previously and exited out of the bedroom window. Once on the exterior of the building, he was treated by advanced life support (ALS) personnel and was transported to a hospital burn center. The fire was extinguished, and no occupants were inside the structure. E31FF1 sustained third degree burns covering 45% of his body. Contributing Factors • Size-up and risk assessment • Risk-benefit analysis • Fire behavior/fire dynamics and tactics • Freelancing/fireground operations procedures • Incident Safety Officer (ISO) • On-scene radio communications • Uncoordinated ventilation by occupants/flow path. Key Recommendations Fire departments should ensure: • Initial and ongoing size-ups and risk assessments are conducted throughout the incident. • Initial arriving firefighters to a fire with a report of trapped occupant(s) consider the risk- benefit analysis of commencing a search versus committing to fire suppression operations, if staffing is limited. • Firefighters and fire officers are trained in fire dynamics, recognition of flow paths, and interior fire attack techniques utilizing door control. • Fireground operations procedures are developed, trained on, and followed by fire officers and firefighters. • Response plans include a dedicated and trained ISO. • Firefighters protect new and older radios and cables by wearing radio straps under their turnout coat, with the radio extended below the bottom of the coat and the antenna tilted away to protect the cable from damage. • Fire prevention efforts educate the community to help change knowledge, attitudes, and behaviors that can reduce risks, injuries, and fires within the community.

Page iii Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03

The National Institute for Occupational Safety and Health (NIOSH) initiated the Fire Fighter Fatality Investigation and Prevention Program to examine deaths of fire fighters in the line of duty so that fire departments, fire fighters, fire service organizations, safety experts and researchers could learn from these incidents. The primary goal of these investigations is for NIOSH to make recommendations to prevent similar occurrences. These NIOSH investigations are intended to reduce or prevent future firefighter deaths and are completely separate from the rulemaking, enforcement, and inspection activities of any other federal or state agency. Under its program, NIOSH investigators interview persons with knowledge of the incident and review available records to develop a description of the conditions and circumstances leading to the deaths in order to provide a context for the agency’s recommendations. The NIOSH summary of these conditions and circumstances in its reports is not intended as a legal statement of facts. This summary, as well as the conclusions and recommendations made by NIOSH, should not be used for the purpose of litigation or the adjudication of any claim. For further information, visit the program at www.cdc.gov/niosh/firefighters/fffipp/ or call 1-800-CDC-INFO (1-800-232-4636).

Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon

Page 1 REPORT F2024-03 • April 2025 Introduction On May 30, 2024, a 33-year-old firefighter (E31FF1) was seriously injured during a rapid-fire progression while searching for a reported trapped child in a residential structure fire. His injuries consisted of burns to 45% of his body. E31FF1’s partner was a 28-year-old firefighter (E31FF2) and was injured with minor burns to his upper body. On June 4, 2024, the chief of the fire department notified the National Institute for Occupational Safety and Health (NIOSH) of this serious injury (SI) incident and requested an investigation. On July 14-19, 2024, two investigators from the NIOSH Fire Fighter Fatality Investigation and Prevention Program (FFFIPP) along with a representative from the International Association of Firefighters (IAFF) traveled to Oregon to investigate this SI incident. The investigators conducted interviews with command officers, fire officers, firefighters, and other emergency personnel who were on-scene during the response. Also, the investigators inspected the personal protective equipment (PPE) used by the firefighters and reviewed fire department standard operating guidelines (SOGs), training, dispatch, and other records. Fire Department The career fire department in this SI incident has a 60 square mile jurisdiction, serves a population of 110,000 residents, responds to an average of 22,000 calls annually, and provides emergency medical services (EMS). The fire department is comprised of 114 uniformed personnel and seven staffed stations. There are three shifts (A, B, and C) which operate on a 24-hours on, 72-hours off, 48-hours on, 72-hours off schedule. Each shift is led by a battalion chief and includes station captains, lieutenants, and firefighters. The fire department is led by a fire chief and assistant chief.

Divisions overseen by the assistant chief include: • Emergency operations division: Responsible for emergency medical calls, fire suppression services, and specialized responses including confined space, high angle rope rescue, structural collapse, water rescue, hazardous materials response, and wildland firefighting. • Training and safety division: Responsible for providing the department’s minimum training standards for all levels of certifications. • Life safety division: Responsible for applying fires codes to new construction, ensuring appropriate fire access, water supply, safety features, including fire alarm systems, and sprinklers systems are code compliant. The fire marshal leads the division, which handles all fire investigations, determining cause for known arson fires, significant fire loss, and fire fatalities.

Page 2 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 Training, Education, and Professional Development The fire department maintains a training division that provides initial training and continuing skills development for all firefighters, fire officers, and EMS personnel. Newly hired firefighters go through a 13-week academy at the training center and follow the curriculum in NFPA 1001 Firefighter Task book, which was adopted by Oregon Department of Public Safety. Upon completion, new firefighters are paired with a senior firefighter and spend the next four shifts at a fire station. After four shifts with a senior firefighter, the new firefighters are deemed fit to be the lone firefighter in the back of an apparatus. After the training academy, firefighters receive a minimum of 60 hours of fire training every two years. All firefighters in the department are trained at the minimum level of Emergency Medical Technician (EMT) Basic through state certification. All engine companies are ALS units, which means each has a firefighter/paramedic on board.

E31FF1 had three years of total fire service with the department. He maintained training in and held numerous certifications as Fire Fighter I, Fire Fighter II, Fire Apparatus Driver/Operator, Hazmat Awareness, and EMT provider. Apparatus, Staffing, and Communications At 21:03 hours, the following units were dispatched for Box 3118:

Apparatus Staffing Arrival On-Scene Engine 31 (E31) 4 21:07 Rescue 31 (R31) added 2 21:07 Battalion Chief 7 (BC7) 1 21:08 Engine 71 (E71) 3 21:08 Engine 74 (E74) 3 21:09 Truck 71 (T71) 4 21:09 Engine 73 (E73) 3 21:10

The flashover in this incident occurred at 21:08 hours, prior to the arrival of five of the dispatched units. Building Construction The one-story residential structure was a type V, wood-frame constructed building (see Photos 2 and 3). It was 936 sq ft and built in 1956. There was a 992 sq ft attached garage on the Bravo/Charlie corner of the building. In 2010, the garage was converted into a fully functional residence. There was no access from the main home to the converted garage. The fire building contained three bedrooms and one full bathroom. The living room, dining room, and kitchen were designed with an open floor plan. The interior finishes consisted of gypsum board walls and vinyl laminated flooring.

Page 3 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03

Personal Protective Equipment At the time of the SI incident, the seriously injured firefighter (E31FF1) was wearing full structural firefighting turnout gear and a NIOSH Approved® SCBA. The investigators examined and photographed the PPE (see Photos 4a, 4b, and 4c). Investigators noted that the portable radio worn by E31FF1 was worn in a radio pocket of the turnout coat with the remote speaker microphone (RSM) cable positioned around the firefighter’s collar. This positioning first, directly exposed the cable to fire conditions and second, caused the radio to unintentionally transmit, which rendered the channel useless for fireground communications. Extensive thermal damage was noted to both the portable radio and RSM. Photo 2: Side Alpha of the structure pre-fire. (Courtesy of the fire department) Photo 3: Side Charlie of the structure pre-fire. (Courtesy of the fire department)

Page 4 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03

Weather Conditions On May 30, 2024, at 20:53 hours, the weather condition was fair, winds out of the Northwest at 9 mph with no reported wind gust. The temperature was 60°F, the dew point was 40°F, the humidity was 55%, and there was no precipitation in the last 24 hours leading up to the incident [Weather Underground 2024]. Investigation At 21:02 hours, the first of several 9-1-1 calls were received by dispatch reporting a one-story residential structure on fire. The initial condition was described by the caller as smoke coming from the structure. Additional calls were received with many of the callers unsure of the exact address and indicating they saw a person in the front door. At 21:03, hours, Box 3118 was dispatched for a structure fire. The dispatcher, unable to determine the exact address, shared two possible locations. E31, E74, E73, E71, T71 and BC7 were dispatched with Operations 4 as the designated radio channel. At 21:05, R31 requested to be added to the assignment. At 21:06 hours, dispatch provided an update regarding the correct location and the report of possible people inside the structure. Medic 346 was also added to the assignment. Prior to the arrival of the responding units, occupants attempted to perform ventilation to air out the building which inadvertently facilitated a significant change in fire behavior. Photos 4a, b, and c: E31FF1’s structural turnout gear (4a), gloves (4b), and portable radio showing extent of damage (4c). (Courtesy of NIOSH)

Page 5 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 E31 arrived on-scene at 21:07 hours, reporting a one-story wood-frame house with smoke showing from Side Alpha. The E31 officer established Command and began to operate from the inside of E31’s cab. Command ordered E31’s crew, consisting of E31FF1 (seriously injured firefighter) and E31FF2 (injured firefighter), to perform fire suppression. Additionally, Command assigned R31 to establish a backup hoseline. As E31’s crew exited the unit, one civilian told them there was a child with disabilities in the bedroom on Side Alpha. At the same time, two other civilians notified Command that there were people inside the house. As the incident continued, the civilians began to scream in both English and Spanish about the trapped occupants which added stress to the firefighters operating on-scene. Based upon these reports, E31FF1 and E31FF2 donned their SCBA facepieces at the front door of Side Alpha. The two firefighters went into “rescue mode” to enter and rapidly search the structure without the protection of a hoseline. As E31FF1 and E31FF2 prepared to enter the structure (see Photo 5), R31 arrived on-scene. After hearing reports of a trapped child, a R31 firefighter (R31FF1) went to Side Charlie without notifying Command. Once he arrived on Side Charlie, a bystander stated there was a child in the bedroom that was hearing impaired. R31FF1 entered the structure on Side Charlie searching for the child. R31FF2 proceeded to E31 as E31’s driver/operator (E31DO) pulled a 1¾ hoseline from the officer’s side of the crosslay. The line was severely kinked due to failure of the webbing that secures the hose bed during travel releasing from the retention ring. As E31FF1 and E31FF2 entered the structure, Command realized the firefighters did not have a hoseline. Command made an announcement on the radio stating the E31 crew entered the structure without a handline and requested R31 to perform fire suppression on Side Alpha. At the same time, R31FF1 observed worsening conditions, exited Side Charlie, and returned to Side Alpha. E31FF1 and E31FF2 proceed through the front door and made their way along the righthand wall of the open living room to reach a hallway (see Diagram 1). They made a right turn down the hallway. E31FF2 entered a bedroom on the right side of the hallway (Side Alpha) to conduct a search. While he was searching the bedroom, E31FF1 remained in the hallway but proceeded further down to look in the next bedroom (Alpha/Delta corner). He observed worsening conditions inside with a rapid increase in heat as firefighters on the outside observed the smoke coming from the Side Alpha door change from gray to black with a glow visible below the smoke layer. E31FF1 returned to the first bedroom and told E31FF2 that they need to exit the structure. E31FF1 led the way with E31FF2 following after exiting the bedroom. At the same time, E31DO charged the handline pulled from the crosslay. R31FF2 advised that there was no pressure in the line. E31FF1 followed the righthand wall of the hallway, entered the middle of the living room, and became disoriented. E31FF2 followed the lefthand wall (keeping one shoulder on the wall) of the hallway, Photo 5: E31FF1 and E31FF2 preparing to enter the structure. (Courtesy of the fire department)

Page 6 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 making the left when entering the living room, and exited through the Side Alpha door in a standing position. E31FF1 reversed direction and proceeded down the hallway towards the second bedroom that he previously attempted to search (Side Alpha/Delta corner). BC7 arrived on-scene and assumed Command. He directed E31’s officer to join his crew for fire suppression. E31FF2 proceeded to Side Bravo to observe conditions where he was asked whether E31FF1 was also out of the structure. Having observed another firefighter on Side Alpha, he indicated that E31FF1 was clear. E71 arrived on-scene and began to secure a water supply to the hydrant and assist with fire suppression. The handline R31FF2 operated on Side Alpha still had no pressure. He was joined with E31FF2 and an E71 firefighter. The E31DO increased discharge pressure to fix the issue as T71 arrived on-scene and was assigned primary search. They proceeded with flaking out the hose with little results. It was determined the hose was kinked at the preconnected hoseline at the discharge located on the officer’s side of E31. Engine 74 arrived on-scene and was tasked with providing a 360 size-up. An unintelligible radio transmission was heard from E31FF1 with an open mic that lasts 55 seconds. Dispatch confirmed to Command that an E31 radio had an open mic which disabled all fireground communications. E74 firefighters proceeded with their size-up, moving counterclockwise (from Side Delta to Side Charlie). At the same time, E31FF1 exited the structure through the bedroom 2 window of Side Alpha T71 who had been advised there was a trapped child in the bedroom of Side Alpha proceeded to the window of bedroom 2 and cleared it with a hook with the intent of entering for a search. Taking the window caused the smoke layer to transition to flaming combustion. Based upon conditions observed, they decided to enter the structure from Side Charlie instead. Dispatch again relayed that an E31 radio had an open mic. E31FFI proceeded to the Side Alpha/Bravo corner and was met by the E31 officer who was checking on him based on his visual condition. At 21:10 hours, E31’s officer radioed Command that he had a firefighter who needed immediate medical attention. E73 arrived on-scene, taking the medical assignment, as Command requested a second alarm. They rushed to E31FF1 with ALS equipment. Engine 74 firefighters pulled a booster line from E31 and applied water to E31FF1 to cool down his turnout gear and SCBA. With E73 firefighters, they began to remove his PPE and cut his turnout pants off. He was placed on oxygen by a non-rebreather mask while waiting for Medic 340 to arrives on- scene. At 21:13 hours, Medic arrived and E31FF1 was loaded and transported to a hospital burn center for his injuries. The fire was soon declared under control with additional searches not finding the reported trapped child.

Page 7 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03

Fire Cause and Origin According to the fire department’s life safety division, the fire was started when a burning candle ignited plastic shower curtains in the bathroom. Prior to the arrival of the responding units, occupants attempted to perform ventilation to air out the building which inadvertently facilitated a significant change in fire behavior which led to the flashover. Contributing Factors Occupational injuries and fatalities are often the result of one or more contributing factors or key events in a larger sequence of events that ultimately result in serious injuries or fatalities. NIOSH investigators identified the following items as key contributing factors in this SI incident that ultimately led to the SI:

• Size-up and risk assessment • Risk-benefit analysis • Fire behavior/fire dynamics and tactics • Freelancing/fireground operations procedures • ISO • On-scene radio communications • Uncoordinated ventilation by occupants/flow path Diagram 1: Floor plan, view from Side Alpha. (Courtesy of the fire department)

Page 8 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 Cause of Serious Injury According to the fire department, E31FF1 sustained third degree burns covering 45% of his body due to the flashover. Recommendations Fire departments should ensure:

Recommendation #1: Initial and ongoing size-ups and risk assessments are conducted throughout the incident. Discussion: Dispatch advised responding units to this SI incident there was a possibility of occupants inside the structure. Once the units arrived, an occupant indicated a child with disabilities was trapped in the structure. Additional bystanders approached fire personnel reporting trapped occupants. There was a rapid entry into the structure without the protection of a hoseline. The speed of the decision prevented a full scene-size up and risk assessment to be completed prior to entry. Once C7 arrived, he gave an order to E74 to conduct a 360; however, the flashover had already occurred, and no updated fire conditions were transmitted as a result of the 360.

Continuous communication supports effective risk assessments. It also allows the incident commander and all personnel operating at an incident to be aware of changing conditions and adjust to avoid hazards or mitigate risks. Performing the 360° is an important component of the scene size-up and can be used in the risk assessment. The International Association of Fire Chiefs’ Rules of Engagement for Structural Firefighting recommends that the first rule for ICs is to rapidly conduct or obtain a 360° situational size- up of the incident. Many incidents contain obstacles to prevent the viewing of all sides of a structure. This included the attached garage and fence that obstructed the view to Side Charlie. When the 360° reconnaissance is achieved, it provides the incident commander and personnel knowledge of the building layout, construction, access/egress points, fire location and direction of spread, and obstacles or hazards. It allows the incident commander and personnel to know how much air is entering the structure from existing openings and what options are available for future ventilation [NIOSH 2017].

Recommendation #2: Initial arriving firefighters to a fire with a report of trapped occupant(s) consider the risk-benefit analysis of commencing a search versus committing to fire suppression operations, if staffing is limited. Discussion: Ongoing research from the Fire Safety Research Institute (FSRI), a part of UL Research Institutes, shows the importance of getting effective water on the fire as quickly as possible from a position of advantage. If staffing and resources permit, both search and suppression operations should occur simultaneously and without delay. Should the initial arriving unit be limited in staffing, strategy and tactics may need to incorporate sequential versus simultaneous operations until more personnel are on the scene and available. As such, a decision will need to be made as to whether the personnel and resources will commence a search for the reported occupant or engage in fire suppression [Weinschenk et al 2022; Regan et al. 2020; Zevotek 2018].

Page 9 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 Prior to making a decision to commence a search, possibly without the protection of a hoseline, several things need considered:

• The current fire location and progression in the structure • The reported location of the occupant(s) • The estimated arrival of the next unit capable of beginning suppression operations • The ability to isolate and compartmentalize spaces during the search

If the first arriving fire officer decides to commence a search without the protection of a hoseline, in lieu of suppression operations, the following actions should occur:

• Communicate this decision via fireground radio including the need for immediate suppression operations upon the arrival of the next unit capable of doing so • Upon entering the structure, attempts should be made to enact door control at the entry point to limit the available fresh air and subsequent fire growth • As the search crews move throughout the structure, all compartmentalizable spaces should be isolated if possible • If the fire room(s) are located during the search, attempts should be made to isolate the fire areas from the remainder of the searchable space • Pressurized water can(s) should be brought with the search crew to provide protection as needed

Recommendation #3: Firefighters and fire officers are trained in fire dynamics, recognition of flow paths, and interior fire attack techniques utilizing door control. Discussion: In 2009, the fire academy removed training on fire behavior and fire dynamics from its curriculum to accommodate a shortage of instructors. Upon observing the location of the fire, E31FF1 did not close the bathroom door to isolate it. R31FF1 entered the Side Charlie of the structure and other firefighters cleared additional windows which created additional flow paths. Research has been conducted by NIST, FSRI, and the International Society of Fire Service Instructors on modern fire dynamics involving heat release rates, new building construction techniques, and the composition of modern home furnishings. In years past, these home furnishings were largely composed of natural products such as wood and cotton. Today, many of these home furnishings are now manufactured using petroleum-based synthetic materials that generate high heat release rates as well as toxic and flammable smoke. Consequently, common, everyday structure fires can create an oxygen depleted environment with fuel rich smoke that can sustain combustion given the appropriate mixture of air and heat. A flow path is defined by NFPA as “the movement of heat and smoke from the higher pressure within the fire area towards the lower pressure areas accessible via doors, window openings, and roof structures” [NFPA 1410 2020]. Flow paths consist of an inlet and an exhaust with the direction of travel being determined by pressure. Heat and smoke in a high-pressure area will travel to an area of lower pressure. It is possible to have multiple flow paths within a structure dependent upon the size of the building, openings, closures such as fire doors, and overall structure design. Personnel working in the

Page 10 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 flow path (between the seat of the fire and the exhaust) are operating in a significantly risky environment. Coordinated ventilation tactics should be utilized to redirect the flow path from interior operations. The International Fire Service Training Association (IFSTA) notes in the 7th Edition of the Essentials of Firefighting that, “when firefighters advance a hoseline or ventilate windows to make entry into a building, they establish new flow paths between the fire compartment (in this case the location) and exterior vents of the building” (see Figure 1) [IFSTA 2018].

Figure 1. Movement of Smoke. (Courtesy of IFSTA)

When identifying the flow path within a structure, personnel should be mindful to wind conditions. Wind can create pressure changes within a structure or building which could spread both smoke and fire into non impacted parts of the structure or building. Interior firefighting personnel should be mindful to these changes in conditions that place the interior firefighting personnel in the exhaust portion of the flow path. Sudden changes in temperature and smoke are factors that should alert interior personnel of imminent changes to conditions. Understanding that modern fuel loads inside of residential structures creates flammable gasses that will sustain combustion interior fire attacks should utilize thermal imaging devices. Water should be applied to the fire as quickly as possible utilizing the reach of an effective fire stream that is delivering an appropriate volume of water to overcome the heat release rate. In some cases, it may be appropriate to introduce water from the exterior of a structure through a window/door with an attack line or begin the fire attack from the exterior with a coordinated blitz attack flowing 300 gallons per minute or greater before committing personnel to the interior. Fire departments should be cooling interior spaces from the safest and closest location possible, especially in vent limited spaces. Maintaining door control is critical during interior operations. Opening a door changes the ventilation profile of the structure and may have an adverse effect on fire behavior [IFSTA 2013]. Controlling the door limits the fresh air available to the fire and reduces the flow path. Door control is a critical

Page 11 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 component of coordinated ventilation and fire attack. Once water is applied to the fire, it is no longer ventilation limited. This allows for more rapid and safer ventilation [Gray and Norwood 2024]. The use of a pressurized water extinguisher to reduce the amount of fire to gain control of the fire room door can be effective. This would isolate the fire. Even if the door was partially burned through, it would limit fire growth and spread and provide time for primary search.

Fire departments can develop training from available resources or requisition existing training. Fire behavior/fire dynamics trainings are available through FSRI, fire academies, and private fire training organizations. Fire departments can develop SOGs using the information and response techniques learned from these trainings.

Recommendation #4: Fireground operations procedures are developed, trained on, and followed by fire officers and firefighters. Discussion: During this SI incident, multiple firefighters engaged in freelancing and employed a number of different tactics and strategies based upon individual decisions. This began with the two firefighters going into “rescue mode” to enter and rapidly search the structure without the protection of a hoseline. NIOSH investigators found that “rescue mode” was a tactic that was not defined in any procedure or policy for fireground operations. After realizing he was in danger, E31FF1 did not activate his PASS device or employ Mayday/firefighter survival techniques. Fire departments should develop fireground operations procedures that are clear and understood by all firefighters. Established procedures create better structure and accountability of actions during an incident. These procedures should include the following areas relative to this incident: Transfer of Command When command is transferred, it is the responsibility of the IC to communicate critical information to the person taking over. This includes the current accountability, assignments, and actions of on-scene personnel; current incident conditions, incident conditions, safety concerns, and benchmarks [NIOSH 2024]. This includes any other critical information that may have been missed while other units are enroute to the incident. Communications Effective fireground radio communication is an important tool to ensure fireground command and control as well as helping to enhance fire fighter safety and health. The radio system must be dependable, consistent, and functional to ensure that effective communications are maintained, especially during emergency incidents. Key components for effective fireground communications include requirements for transmitting strategic modes of operation and situational reports; personnel accountability reports, emergency traffic including Mayday; and receipt of critical transmissions [NFPA 1550 2024].

Page 12 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 Mayday/Firefighter Survival Firefighters should be trained in how to call a Mayday when in danger. Any delay in calling a Mayday reduces the chance of survival and increases the risk to other firefighters trying to rescue the “downed” firefighter. Firefighters should have 100% confidence in their competency to declare a Mayday and follow the GRAB LIVES Mayday procedures. Fire departments should ensure that any personnel who enter a hazard zone have been trained and are proficient in Mayday competencies [IAFF 2010a; 2010b]. Training on Fireground Operations To ensure the proficiency and competency of fire department members, fire departments should conduct annual training and skills evaluations to verify firefighters understand the established fireground operations procedures. NFPA 1550, Standard for Emergency Responder Health and Safety, requires a fire department to establish and maintain a training, education, and professional development program with the goal of preventing occupational deaths, injuries, and illnesses. This ensures member competencies are maintained to execute all responsibilities effectively, efficiently, and safely [NFPA 1550 2024]. NFPA 1410, Standard on Training for Emergency Scene Operations, defines basic evolutions, which are adaptable to local conditions and serves as a method for the evaluation of minimum acceptable job performance during training for fire suppression and rescue activities [NFPA 1410 2020]. Proficiency training for fireground operations and emergency incidents should occur annually. This training should include scene size-up, situational awareness, use of an incident management system, personnel accountability system, strategy and tactics, search and rescue, hoseline operations, ladder operations, ventilation, thermal imaging cameras, fireground communications, use of rapid intervention teams, and Mayday operations [NIOSH 2024]. Recommendation #5: Response plans include a dedicated and trained ISO. Discussion: During this SI incident, the first IC operated from the cab of E31. Additional firefighters on- scene were tasked with fire suppression and search. A number of firefighters began tasks without notifying the incident commander of their actions or where they would be operating. At no time was a dedicated ISO established to supervise operations, fire behavior, and conduct accountability. An ISO should be trained to NFPA 1550 which defines the requirements for the IC, including establishing a fixed command post, personnel accountability, the use of staff aides and RICs, and the appointment of an ISO and assistant safety officer(s) (as needed). The standard addresses the expectations and authority of the ISO. Expectations and authority include determining hazardous incident conditions, advising the incident commander to modify control zones or tactics to address corresponding hazards, communicate fire behavior and forecast growth, and estimate building/structural collapse hazards. This also includes the authority to stop or suspend incident operations based on imminent threats posed to firefighter safety [NFPA 1550 2024]. The ISO should be separate from the IC, operations, or accountability positions so they can focus on their responsibilities and the primary objective of continually assessing any and all on-scene hazards to firefighter life and safety. A personnel accountability system is a system that readily identifies both the location and function of all members operating at an incident scene [NFPA 1550 2024]. This system is implemented during an

Page 13 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 incident to collect and maintain the status and location of the resources working in, or potentially working in, an immediately dangerous to life and health IDLH environment. All members operating at an incident are responsible for understanding and participating in this system. The IC is responsible for but may delegate certain responsibilities to another person such as the ISO. An integral part of the accountability system is to make sure that the firefighters who are assigned and operating in the hazard zone are accounted for throughout the entire incident. A properly initiated and enforced personnel accountability system can enhance firefighter safety and survival [NIOSH 2024]. A functional personnel accountability system should have the ability to identify:

• Members operating in the hazard zone (who) • Where members are in the hazard zone (where) • Conditions in the hazard zone (conditions) • Actions used in the hazard zone (actions) • Paths of access and egress in and out of the hazard zone (exits) • RICs and their assignments

Different methods and tools are available for resource accountability, including: • Tactical worksheets • Command boards • Apparatus riding lists • Company responding boards • Electronic bar-coding systems • Accountability tags or keys

Recommendation #6: Firefighters protect new and older radios and cables by wearing radio straps under their turnout coat, with the radio extended below the bottom of the coat, and with the antenna tilted away to protect the cable from damage. Discussion: In this incident, E31FF1 was able to self-rescue out of a window but sustained significant burn injuries. The portable radio he had was placed in a radio pocket of his turnout coat with the RSM cable positioned around his collar. This positioning directly exposed the cable to fire conditions, eventually causing the radio to unintentionally transmit, rendering the channel useless for fireground communications. Extensive thermal damage was noted to both the portable radio and RSM.

Portable radios serve as an essential safety device allowing each firefighter to immediately report, or be notified of, changing conditions and emergencies that can affect their survival. As such, it is critical firefighters understand the conditions in which these devices may fail and how to protect them. Between 2004 and 2014, NIST conducted studies to determine the effect of high temperatures (160°C/320°F) on portable radios. Some of these studies demonstrated that direct exposure to high temperatures can melt the RSM and cable, resulting in loss of functionality for interior firefighters and disabled broader fireground communications due to unintentional transmissions. In these studies, failures occurred most often when portable radios were worn on the exterior of structural firefighting turnout gear with the RSM cable directly exposed to high temperatures (which was also the case in this current SI incident) [NIST 2006]. NIST findings from user-centered interviews noted that firefighters reported

Page 14 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 being aware that the RSM cable is known to melt in high temperature environments, with a number of firefighter fatalities related to failure of radios and subsequent loss of communications [NIST 2018]. These studies led to the development of NFPA 1802. The standard includes a requirement for portable radios and their accessories to withstand temperatures up to 500°F for five minutes [NFPA 1802 2021].

Radios meeting the NFPA 1802 standard were first certified in 2022. It is likely that older radios are still in use and may not offer the same heat tolerance as newer radios designed to meet the standard. Firefighters may employ additional protection or other tactics for both new and older radios and cables with PPE.

Based on other studies, the NFPA 1802 standard, and this SI incident, proper placement of portable radios entails wearing the radio on straps, under the turnout coat, and with the radio extended below the bottom of the coat with the antenna tilted away to protect the RSM cable from melting (see Photo 6). This placement of the portable radio allows the RSM cable to be protected from direct exposure. Additionally, this placement significantly reduces the possibility of an entanglement hazard involving the RSM cable and improves signal strength from the radio antenna. A 2013 report by the Fairfax County Fire & Rescue Department includes diagrams for proper wearing of a radio strap and portable radio for interior firefighting operations. Fire departments may consider equipping firefighters with radio straps and updating standard operating procedures or guidelines to include proper use during interior firefighting operations. Fire academies and fire instructors may opt to include radio strap use in both entry level and advanced firefighter training.

Recommendation #7: Fire prevention efforts educate the community to help change knowledge, attitudes, and behaviors that can reduce risks, injuries, and fires within the community. Discussion: In this incident, occupants attempted to perform ventilation to air out the building which inadvertently facilitated a significant change in fire behavior which led to the flashover.

Fire and life safety education is a critical element of community-risk reduction. This effort works to change knowledge, attitudes, and behaviors among community residents to reduce risks, injuries, and fires [IFSTA 2015a]. Fire departments act on this effort by training firefighters and other personnel as Fire and Life Safety Educators (FLSEs). FLSEs coordinate and deliver educational programs that teach people about a particular hazard and how to reduce or prevent the risks [NFPA 1030 2024]. Photo 6: Firefighter wearing portable radio on a strap under coat. (Courtesy of the National Volunteer Fire Council)

Page 15 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 Fire and life safety education initiatives can teach community members about fire and injury prevention, including what actions they can take during an emergency. Example topics of these programs include: • The importance of smoke alarms to reduce deaths in structure fires • The use of child safety seats to lessen injuries of children involved in vehicle accidents • The need for carbon monoxide detectors and proper use of gas-powered appliances • Closing bedroom doors while sleeping or closing doors upon egressing during a fire incident Specific to this incident, an applicable topic could be teaching community members to close doors during an evacuation to limit oxygen ingress and to report fires as soon as possible to get firefighters dispatched immediately. Fire Safety Research Institute’s ‘Close Your Door’ encourages those both trapped in a room during a fire as well as those who can safely leave a home to close as many doors as possible to limit the growth and spread of the fire. The webpage, https://fsri.org/programs/close-before- you-doze offers a series of free resources, including videos that fire departments can use to educate the public. These types of initiatives can be provided by fire departments in the form of scheduled lectures at community centers or fire stations, school visits by firefighters for fire prevention week, or through posting informational graphics on social media [IFSTA 2015a]. References Fairfax County Fire & Rescue Department [2013]. Portable radio placement in the IDLH. Fairfax, VA: Fairfax County Fire & Rescue Department, Communications Division. Gray S and Norwood PJ [2024]. Modern firefighting tactics in residential structures: Ventilation, coordination, and control. Fire Engineering. IAFF [2010a]. IAFF Fire Ground Survival Training Program. Washington, DC: International Association of Fire Fighters. IAFF [2010b]. Self-survival procedure. Washington, DC: International Association of Fire Fighters. IFSTA [2013]. Essentials of fire fighting. 6th ed. Stillwater, OK: International Fire Service Training Association. IFSTA [2015a]. Fire and life safety educator. 3rd ed. Stillwater, OK: International Fire Service Training Association. IFSTA [2018]. Essentials of fire fighting. 7th ed. Stillwater, OK: International Fire Service Training Association. NFPA [2020]. NFPA 1410, Standard on training for emergency scene operations. Quincy, MA: National Fire Protection Association. NFPA [2021]. NFPA 1802, Standard on two-way, portable RF voice communications devices for use by emergency services personnel in the hazard zone. Quincy, MA: National Fire Protection Association.

Page 16 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 NFPA [2024]. NFPA 1550, Standard for emergency responder health and safety. Quincy, MA: National Fire Protection Association. NFPA [2024]. NFPA 1030, Standard for professional qualifications for fire prevention program positions, Quincy, MA: National Fire Protection Association. NIOSH [2017]. NIOSH fire fighter fatality investigation helps incident commander choose strategy and tactics to control flow path and protect fire fighters. By Bowyer M., Loflin M., Merinar T., Miles S., Moore P., Wertman S., Orr B., and Webb S. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication 2017–200. NIOSH [2024]. Volunteer firefighter killed after becoming trapped at an assisted living facility fire and two firefighters injured – New York. By Loflin M and Attwood W. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2021-10. NIST [2006]. NIST Technical note 1477: Testing of portable radios in a fire fighting environment. By Davis WD, Donnelly MK, and Selepak MJ. Gaithersburg, MD: U.S. Department of Commerce, National Institute of Standards and Technology. NIST [2018]. Voices of first responders -identifying public safety communications problems: Findings from user-centered interviews phase 1, volume 1. By Choong YY, Dawkins S, Furman S, Greene KK, Prettyman SS, and Theofanos MF. Gaithersburg, MD: U.S. Department of Commerce, National Institute of Standards and Technology. Regan J, Bryant J, and Weinschenk C [2020]. Analysis of the coordination of suppression and ventilation in single-family homes. Columbia, MD: UL Fire Safety Research Institute. SKCFTC [2023]. South King County Fire Training Consortium, Command Procedures: Structure Fires. Kent, WA. South King County Fire Training Consortium. March 2023. Weather Underground [2024]. Weather History. The Weather Underground. Date accessed: November 14, 2024. Weinschenk C and Stakes K [2022]. Analysis of search and rescue tactics in single-story single-family homes part III: Tactical considerations. Columbia, MD: UL Fire Safety Research Institute. Zevotek R, Stakes K, and Willi J [2018]. Impact of fire attack utilizing interior and exterior streams on firefighter safety and occupant survival: Full-scale experiments. Columbia, MD: UL Firefighter Safety Research Institute. Investigator Information This incident was investigated by Louis (Rick) Lago (former), Investigator, and Patrick R. Montague (former), Investigator, with the Fire Fighter Fatality Investigation and Prevention Program, Surveillance

Page 17 Career Firefighter Seriously Injured During Search Operations in a Residential Structure Fire – Oregon Report # F2024-03 and Field Investigations Branch, Division of Safety Research, NIOSH located in Morgantown, WV. Jeffrey Seaton, an expert in firefighter survival accompanied the team as a representative of the IAFF.

This investigation report was authored by Louis (Rick) Lago (former) and Dr. Wesley R. Attwood, Investigator and Program Advisor, with the Fire Fighter Fatality Investigation and Prevention Program, NIOSH. The report was also authored by Keith Stakes, Principal Engineer, Fire Safety Research Institute, part of the UL Research Institutes.

Dan Madrzykowski from the Fire Safety Research Institute, part of the UL Research Institutes, provided an expert review of the investigation report. A subject matter expert review was provided by Brian Chalfant, Staff Fire Instructor, Allegheny County Emergency Services – Fire Academy, Pennsylvania. The NFPA Emergency Response & Responder Safety Division provided a technical review. Additional Information Underwriters Laboratories (UL) The Fire Safety Research Institute (FSRI), part of the UL Research Institutes, continues to work with fire departments and fire service organizations to conduct research on fire dynamics, fire safety issues, and fire ground operations. Access to reports from completed studies and information from on-going studies can be found at https://fsri.org. Access to free online training on evidence-based firefighting (more than 30 course modules in all) can be found at https://training.fsri.org. Disclaimer The information in this report is based upon dispatch records, audio recordings, witness statements, and other information that was made available to the National Institute for Occupational Safety and Health (NIOSH). Information gathered from witnesses may be affected by recall bias. The facts, contributing factors, and recommendations contained in this report are based on the totality of the information gathered during the investigation process. This report was prepared after the event occurred, includes information from appropriate subject matter experts, and is not intended to place blame on those involved in the incident. Mention of any company or product does not constitute endorsement by NIOSH, Centers for Disease Control and Prevention (CDC). In addition, citations to websites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or products. Furthermore, NIOSH is not responsible for the content of these websites. All web addresses referenced in this document were accessible as of the publication date. NIOSH Approved is a certification mark of the U.S. Department of Health and Human Services (HHS) registered in the United States and several international jurisdictions.

FD06031 ｜ 2024_美国_NIOSH_face202404.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2024 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Page i

27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Executive Summary On August 1, 2024, a 27-year-old male career firefighter (FF1) with no known medical conditions participated in a mutual aid response to a residential structure fire. FF1 was inside the building with an active fire for approximately 4 minutes before exiting the front door. Shortly after exiting, the firefighter appeared to have a seizure and collapse to the ground. Nearby firefighters who witnessed the seizure and collapse rushed to assist FF1. FF1’s gear and self-contained breathing apparatus (SCBA) were quickly removed, and an initial medical assessment found FF1 to be pulseless and apneic. Cardiopulmonary resuscitation (CPR) was initiated. The initial firefighters tending to FF1 were all basic emergency medical technicians (EMT-B). The first due engine, which was the initial interior fire attack crew, was advanced life support (ALS) equipped and had a firefighter paramedic (FFP) assigned to the crew but no ambulance for patient transport was immediately available. The FFP initiated and continued advanced cardiac care of FF1 until arrival of the ALS ambulance. While en-route, the cyanide antidote hydroxocobalamin was administered. FF1 regained a pulse enroute to the hospital but lost it as he was being taken into the emergency department (ED). The ED staff worked FF1’s resuscitation for approximately 15 minutes before he was pronounced dead. The Medical Examiners report listed FF1’s cause of death as “unspecified cardiac arrest sustained while fighting a fire”. Of note, FF1 had a syncopal event approximately 2 months prior but the cause was not identified, and no further medical evaluation was done. At the approximate time of arrival at the residential fire, the outdoor temperature was 96℉ with 44% humidity resulting in a heat index of 103.4℉. However, it is unknown if either the syncopal event or heat stroke was a factor in this death. Key Recommendations NIOSH offers the following recommendations to reduce the risk of sudden cardiac arrest among firefighters at this and other fire departments across the country. • Key Recommendation #1: Fire departments should consider implementing annual medical examinations as per National Fire Protection Association (NFPA) 1582 recommendations. • Key Recommendation #2: Fire departments should encourage members to reduce and if possible, eliminate intake of/exposure to known risk factors for adverse cardiovascular events and rhabdomyolysis. F2024-04 • February 2025

Page ii 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 • Key Recommendation #3: Fire departments should increase awareness of conditions that increase risk of heat-related illness (HRI) and implement heat stress management policies to reduce those risks. • Key Recommendation #4: Ensure emergency medical service (EMS) with transport capability is on-scene at any fire or special operation location in the event emergency medical care and transport is needed. • Key Recommendation #5: Fire officers should ensure crew integrity when entering, operating in, and exiting from an environment immediately dangerous to life or health (IDLH). • Key Recommendation #6: Fire departments should provide physicians tasked with pre- employment and return to work medical evaluations, descriptions of the essential tasks and duties and not rely on the cadet/incumbent to relay this information. • Key Recommendation #7: Fire departments should ensure that a physician signs off on pre- employment and return to work evaluations.

The National Institute for Occupational Safety and Health (NIOSH) initiated the Fire Fighter Fatality Investigation and Prevention Program to examine deaths of fire fighters in the line of duty so that fire departments, fire fighters, fire service organizations, safety experts and researchers could learn from these incidents. The primary goal of these investigations is for NIOSH to make recommendations to prevent similar occurrences. These NIOSH investigations are intended to reduce or prevent future fire fighter deaths and are completely separate from the rulemaking, enforcement and inspection activities of any other federal or state agency. Under its program, NIOSH investigators interview persons with knowledge of the incident and review available records to develop a description of the conditions and circumstances leading to the deaths in order to provide a context for the agency’s recommendations. The NIOSH summary of these conditions and circumstances in its reports is not intended as a legal statement of facts. This summary, as well as the conclusions and recommendations made by NIOSH, should not be used for the purpose of litigation or the adjudication of any claim. For further information, visit the program Web site at www.cdc.gov/niosh/firefighters/fffipp/about.html or call toll free 1-800-CDC- INFO (1-800-232-4636).

27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas

Page 1 F2024-04 • February 2025 Introduction On August 1, 2024, a 27-year-old male career firefighter (FF1) with no known medical conditions participated in a mutual aid response to a residential structure fire. FF1 belonged to Fire Department A (FD-A) which has multiple stations. FF1 was moved from his normally assigned FD-A station 1 to FD- A station 2 for that day due to staffing alignment. This was the first time FF1 worked from FD-A station 2. The structural fire was located in Fire Department B (FD-B)’s service area and FF1’s crew responded as part of a mutual aid agreement. The incident command for the fire was FD-B. FF1 was inside the building with an active fire for approximately 4 minutes before exiting through the front door. Shortly after exiting, the firefighter appeared to have a seizure and collapsed to the ground. During the investigation, the National Institute for Occupational Safety and Health (NIOSH) investigators interviewed the following people: • Chiefs of fire departments involved in the incident • Fire department staff and emergency medical service (EMS) personnel involved in the incident • County Medical Director • Fire Marshal • Medical Examiner The NIOSH investigator reviewed the following documents: • Standard operating procedures from FD-A and FD-B • Mutual aid agreements between involved departments • Narratives and incident reports from all departments involved • Incident location photos and schematics • Radio communication audio files from incident • Cyanokit training, assignment, and use protocols • FF1 Training Records

Page 2 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 • FF1 helmet cam during August 1, 2024, residential fire response • Medical examiner’s report • Pre-hospital care reports and data from LifePak used by EMS staff in FF1’s resuscitation • Emergency Department (ED) records from 5/16/24 and 8/1/24 visits • Pre-employment medical evaluation • Incident/injury reports involving FF1 during his employment with FD-A Fire Departments FD-A is a career municipal fire department with approximately 500 members. FD-A has 22 stations, covering 168 square miles, and serving a population of over 490,000. FD-A responded to this incident under an automatic aid agreement. There were 3 fire departments involved in this incident. FD-B is a career department serving the county jurisdiction around the municipality of FD-A. The structure fire occurred in FD-B’s jurisdiction. FD-B had multiple apparatus respond to this incident. The Battalion Chief that assumed Incident Command (IC) is a member of FD-B. Fire Department C (FD-C) is a small city department within the same county as FD-A and FD-B. FD-C is a career department that also responded to this incident under an automatic aid agreement. Engine 1 (E1) from FD-C was the first engine company to arrive on-scene. E1 was ALS capable with equipment and a staffed firefighter/paramedic. The initial primary Incident Safety Officer was a Captain from FD-B. The interior Safety Officer was from FD-C. Membership and Training FD-A hiring process components consists of a two-part aptitude/interview process: • Phase 1 consists of a physical abilities evaluation in a group setting where candidates are individually scored and may qualify for phase 2 • Phase 2 is an individually tested and scored, multi-station obstacle course, modeled after the Candidate Physical Abilities Test (CPAT), and a panel interview In addition, applicants must submit to a background investigation and undergo a physical examination and lab screening. The physical examination is National Fire Protection Association (NFPA) 1582 compliant (Standard on Comprehensive Occupational Medical Program for Fire Departments) and is conducted by a city appointed physician. Once accepted, the recruit training academy is 22 weeks in length. Eight weeks are dedicated to Emergency Medical Technician training and certification. Fourteen weeks are dedicated to basic training in firefighting, special rescue, and hazardous materials. Recruits participate in ride-outs with operations companies during the academy.

Page 3 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 Preplacement/Periodic/Return to Work Medical Evaluations At the time of FF1’s hire in 2019, FD-A required the pre-employment medical evaluation to include a physical exam, pulmonary function testing, blood tests (standard chemistry panel and complete blood count), urine drug screen, urinalysis, chest x-ray, lumbosacral spine x-ray, audiogram, functional capacity evaluation (FCE), cardiac treadmill stress test and completion of the Occupational Safety and Health Administration (OSHA) Respirator Clearance form. At the time of FF1’s hire through the period of his employment, there were no department requirements for annual medical examinations. The majority of FF1’s pre-employment evaluation was completed on February 19, 2019. The physical exam, blood tests, pulmonary function testing, x-rays, and cardiac treadmill test were all unremarkable. His urinalysis was positive for ketones but was otherwise normal. He was noted to have “good exercise tolerance” on his treadmill test and his functional capacity evaluation was reported in 2 components; The physical classification of the worker was designated as “very heavy”, and the physical demand classification of job was designated as “medium heavy”. These classifications were defined according to the table below (Table 1). The final determination of the FCE dated February 22, 2019, was that FF1 was “able to meet functional requirements of job based on tests completed”.

Light Light/Medium Medium Medium/Heavy Heavy Very Heavy Occasional (1%–33%) 20 lbs. 35 lbs. 50 lbs. 75 lbs. 90 lbs.

100 lbs.

Frequent (34%–66%) 10 lbs. 15 lbs. 25 lbs. 35 lbs. 50 lbs.

50 lbs.

Constant (67%–100%) Negligible 5 lbs. 10 lbs. 15 lbs. 20 lbs.

20 lbs.

It was noted on the Respirator Clearance form under job description that “a respirator job description was not provided by the Company and this recommendation is based solely upon history/physical information gained today”. The FD’s policy manual states that the “description of essential job tasks is included in position classifications associated with specific titles, or ranks, for each commissioned Fire Department position” It is unclear why these were not provided to the physician conducting FF1’s pre- placement medical evaluation. The final determination was that FF1 was cleared with “no restrictions of respirator use”. FF1 had 2 illness/injury incidents during his tenure at FD-A. The first incident occurred on July 15, 2020, during participation in a structure fire response when FF1 tripped over a hoseline and fell to the ground. During his fall, his bunker pants were pushed up above his knees, exposing his skin resulting in “slight burns to both shins” when he contacted the heated floor. He was treated at the local ED and discharged to return to work without restrictions. A return-to-work form dated the same day of the incident was completed confirming he was released back to full duty without restriction. The required work release form included a list of essential tasks and functional capabilities required for those tasks. The second incident was dated November 18, 2020, and was described as an inhalational exposure to COVID-19 occurring while engaged in CPR despite wearing proper personal protective equipment (PPE). The form does not list what PPE was worn, only that all personnel on 2 engine crews and EMS

Page 4 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 crew were exposed/involved in this incident and “all engine 20 members were wearing proper PPE”. No treatment was documented. The form indicated FF1 was not admitted to a hospital and was not seen in the Occupational Medicine Facility. It is unknown if FF1 was subsequently tested for COVID-19 to determine if an infection resulted from that possible exposure. The form asked if work restrictions were prescribed by a clinician. The response to this question was “no” but the form was not signed by any physician, so it is unclear if a physician reviewed this incident report or if there was a reviewing physician and they did not sign the report. Wellness/Fitness Programs FD-A has no formal on-going wellness and fitness programs. The only mandated physical occurs at pre-employment. The department’s physical fitness training policy allows for one hour of physical training per 24-hour shift. Exercise equipment is furnished by the department. Investigation On August 1, 2024, at 07:00 hours, FF1 began his tour of duty on his regularly assigned shift and apparatus at FD-A station 1 on Engine 10 (E10). Soon after arrival at the firehouse, FF1 was reassigned to station 2 as Engine 2 (E2) driver in another part of the city. During the morning at station 2, FF1 checked out his equipment and apparatus, performed physical training which included an upper body weight workout of shoulder press and lateral pull-downs, and took two EMS calls on E2. After completing the second EMS call, FF1 went to lunch then picked up drinks and snacks at a store. During interviews, firefighters from both E10 and E2 stated that FF1 was his typical self, having no complaints and doing nothing out of the ordinary. The E2 crew was less familiar with FF1’s routines, but there was nothing that raised any suspicion to any physical or medical concerns. FF1 was in good physical condition, did regular cardiovascular and strength training workouts, and he was an avid hunter. To understand the brevity of the events in the early stages of this incident, a basic timeline follows. The timeline is derived from 4 sources: the dispatch log of the incident, FF1’s self-contained breathing apparatus (SCBA) data log, helmet camera footage from FF1, and eyewitness interviews. There is an approximate 30–60 second margin of error in times associated with the SCBA data log. This includes entry and exit from the structure. The SCBA internal timekeeper was off 4–5 minutes, commonly referred to as time drift. Time drift is not uncommon with SCBA’s internal timekeeping mechanisms and is usually rectified at the time of regular equipment service and software updates. Time Action 07:00 Shift start 14:56 Dispatch to structure fire 15:02 Battalion Chief (BC1) on -scene 15:03 First engine company (Engine 1 from Fire Department C) on-scene 15:06 Engine 2 marks on-scene (FF1’s engine company) 15:11 FF1 goes on air, breathing from his SCBA

Page 5 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 15:12 FF1 makes entry into the structure 15:13 Primary search marked “all clear” 15:16 SCBA regulator disengaged. It is reasonable to conclude that he exits the structure at this time. 15:22 Multiple eyewitness accounts of FF1’s seizure activity and collapse 15:22 Care is initiated immediately following witnessed collapse of FF1 At 14:56, E2 was dispatched to a residence fire as part of a mutual aid response. BC1 arrived first and E2 would be the second engine on scene. BC1 reported smoke and flames coming from the residence. E2 marked on scene at 15:06 and was assigned to pull a 2nd hoseline for fire attack. The E2 crew pulls and sets up a line from the first due engine, Engine 1 (E1). E1 was assigned as the initial attack crew, and they already had their primary attack line inside the structure. The second hoseline was charged and the E2 crew made entry into the residence. FF1 helmet cam video shows FF1 pulling at the exterior roofline when his crew enters the structure, which allows another crew to enter the structure ahead of FF1. This provided an approximate 1-minute delay of FF1 meeting up with his crew inside the residence. Upon entering the structure, FF1 catches up with his crew quickly and assists with moving small piles of burning material with his hook to be extinguished by his crew’s hoseline. At approximately 15:14, FF1 shuts off his helmet camera. In the next 1–2 minutes, FF1 tells one of his engine crew members that he must exit the structure. The firefighter crew member asks FF1 if he’s ok and FF1 responds with a “yes”. FF1 exits the structure at approximately 15:16. During this same period an Incident Safety Officer reports to the IC that a small amount of fire needed some attention under the front porch. The IC tasked E2 crew with this assignment. Interviews with witnesses believe that FF1, being already outside the structure, heard the redirected assignment for E2 and made his way to this part of the porch in front of the structure. Firefighters at different locations on the fireground reported something did not appear right as FF1 came down the 3–4 steps off the wooden porch and made his way to the area underneath the porch that had some fire. These firefighters could not specify what about FF1 caught their attention as he exited the structure, only that he did. FF1 then began to do some work with his hook at a spot fire under the porch. At 15:22, FF1 was witnessed to have collapsed from standing position and then had seizure activity at which time a medical emergency was called. Nearby crew members responded to FF1 and by the time they arrived at his position, the seizure activity had stopped, and he was noted to be in cardiac arrest. On-scene FF/EMTs responders were basic life support (BLS) certified so they removed his gear and started chest compressions immediately while the Firefighter Paramedic (FFP) working the interior was directed to FF1. At this time a call was placed for an ALS transport unit. Resuscitation followed standard BLS protocol with placement of an I-Gel Airway®. A Lucas® device was placed by the FFP to assist with compressions at 15:43. Although FF1 displayed asystole for the majority of the resuscitation during which multiple doses of epinephrine were administered, the Lifepak-15 did advise and delivered one shock of 360J at 15:31 as per pre-hospital records (this differed from the timestamp on the data output of the device which noted this time to be 15:27). Medical command advised administration of 2 doses of sodium bicarbonate (this offsets metabolic acidosis that occurs with prolonged cardiac arrest). At 15:50, the cyanide antidote, hydroxocobalamin, was administered followed by one dose of calcium chloride. At 16:01, return of spontaneous

Page 6 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 circulation was noted but this was transient. Additional doses of epinephrine were administered prior to arrival at ED at 16:08. The Lucas® device was removed at 16:10 for a pulse check which did not find one, and asystole was confirmed on the cardiac monitor. Compressions were resumed and additional doses of epinephrine, calcium chloride, and sodium bicarbonate were administered. FF1 showed ventricular fibrillation at 16:15 and a shock of 300J was provided but this degraded back to asystole by the time compressions were stopped for the next pulse check at 16:19. Additional medications were given but he remained in asystole. Time of death was noted as 16:25. Medical Findings On one of his pre-placement medical history forms completed in 2019, FF1 stated he had already been in the fire service for 2 years. As this was a hearing history form, there were questions focused on medical conditions involving hearing/vestibular issues which were all negative but none of the provided forms showed a question that asked for any/all past medical history. The Fire Marshal covering FD-A confirmed that FF1 did not have a primary care provider. This was also noted in the May 2024 ED visit records. ED records from August 1, 2024, admission note a past medical history of chickenpox, hay fever, and allergic rhinitis with no known allergies. The autopsy was conducted on August 2, 2024, the day after FF1 died, by the Chief Medical Examiner (CME) of the district in which the fatality occurred. The cause of death was determined to be “unspecified cardiac arrest while fighting a fire”. The manner of death was listed as accident. The CME noted evidence of the medical interventions engaged as part of the resuscitation efforts as well as contusions and a tongue hematoma likely related to his seizure. FF1’s height was 70 inches and weight was 185 lbs. There were no structural or vascular abnormalities in the brain. The heart weight was within normal parameters at 440 grams. There were no structural abnormalities of the heart, and the coronary arteries only showed mild atherosclerosis. No evidence of ischemia or infarction were noted. Histological examination of sections from the brain, heart, liver, and kidneys were all normal. Two slides of lung tissue showed congestion, pulmonary edema, and interalveolar (inside the small end airways where gas exchange occurs) hemorrhage. Intracardiac blood was collected and tested negative for ethanol, carbon monoxide, acetone, and a panel of 17 common drugs of abuse including prescription medications and illicit substances. During the Fire Marshal’s inventory of FF1’s personal effects, he found a receipt from a local supermarket dated August 1, 2024, and time stamped at 14:38, approximately fifteen minutes before FF1 responded to the residential fire at 14:56. FF1 purchased several caffeinated drinks and snack foods. These purchases caused the Fire Marshal to be concerned regarding FF1’s use of caffeinated products. These concerns were relayed to the medical examiner who decided to collect a femoral blood sample to test for a caffeine level. This sample resulted a caffeine level of 4.6 micrograms per milliliter (µg/ml). As per the laboratory that conducted the analysis, caffeine levels in the range 3–15 µg/ml are reflective of “usual stimulant levels” in adults [NMS 2022]. Discussion Operations/Tactics This incident exhibits many of the best elements of automatic agreements. The closest apparatus and personnel respond to an emergency incident regardless of the “patch” on the sleeve and the “shield” on

Page 7 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 the apparatus door. Within the first few minutes, the IC from FD-B worked with companies from multiple agencies to get an aggressive fire attack in motion, and the primary search was cleared within 10 minutes of the first engine’s arrival. NIOSH investigators found the initial operations to be adequate. The IC and his assignments effectively mitigated the fire. However, good practice and NFPA consensus standards dictates that no firefighter should be entering or leaving the fire structure alone [NFPA 2024]. FF1’s helmet camera footage showed he entered the structure following firefighters that were not part of the E2 crew, and quickly caught up with the E2 crew. FF1 spent approximately 4 minutes inside and probably was never more than about 10 feet inside the front door before notifying a member of his crew he was going to exit the structure. His E2 officer was unaware that FF1was exiting the structure until after incident command reassigned the crew to the side A porch for a small fire under the porch. FF1 was not with his crew when he disengaged his SCBA at 15:16 through his collapse at 15:22. If his collapse had not been seen by other nearby responders who began to render aid, there could have been a significant delay in others being aware he was experiencing a medical emergency and initiating medical care. Even when all tactics are accomplishing the objectives of an incident action plan (IAP), it is important to ensure that basic safety tenets are followed. On scene EMS and related transport service are issues that must be addressed in all emergency response plans. In this incident an operational ALS engine company was part of the initial response increasing FF1’s chances of survival. As per NFPA 1550, Standard for Emergency Responder Health and Safety states in Paragraphs 10.6.9 – 10.6.11 and A.10.6.9 that basic life support or advanced life support (preferred) with transportation capabilities should be available on scene when members are operating at a structure fire or performing special operations, and for medical monitoring at HAZMAT incidents, and considered based on risk management at other emergency operations [NFPA 2024]. If resources are challenged, efforts should be made to assign EMS transporting units once an incident is confirmed to be a working incident by first arriving crews so that any responding personnel who experience a medical emergency can be immediately transported to a higher level of care. NIOSH investigators identified delays in obtaining EMS transport as this resource was not assigned as part of the original working incident and it was just by happenstance that an ALS level provider was on-scene as part of the responding staff to provide care until an appropriate transport unit arrived. In these jurisdictions, agencies place paramedics and ALS equipment on apparatus but those apparatus are not part of formal EMS response. In this incident, E1 was ALS staffed and equipped giving FF1 the best chance of survival. Many similar responses could occur in other parts of these jurisdictions and there would be no ALS staffed and equipped apparatus. Knowing where those apparatus are and their availability for response would be helpful when transporting units are unavailable. Fire departments should consider creating a notation on their dispatch records to denote whether staff assigned to apparatus have ALS provider capacity with the associated equipment. Medical FF1’s cause of death was “unspecified cardiac arrest sustained while fighting a fire”. FF1 did not have any known prior medical history. NIOSH investigators cannot make any determinations beyond what the medical examiner stated as to what caused the seizure and cardiac arrest that occurred to FF1 on August 1, 2024. However, there are some aspects to FF1’s fatality that may be helpful to discuss.

Page 8 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 Syncope The Fire Marshal noted that FF1’s neighbors had witnessed him have a syncopal episode sometime around April-May 2024, but he was unable to confirm ED evaluation. Records obtained by NIOSH investigators from an ED near FF1’s home showed that he was seen in that ED on May 16, 2024. FD-A’s records show that FF1 had worked May 15, 2024, 07:00 to May 16, 2024, 07:00 and during that shift, he responded to 3 calls with the last call dispatching him from the station at 17:00 for an “unauthorized burning”. ED records show that FF1 stated that he started mowing his lawn around 18:00 on May 16, 2024, and had completed 3 passes with his push mower when he suddenly lost consciousness and woke up on the ground. Neighbors witnessed the collapse and reported that he had aroused by the time they reached him. FF1 stated he felt his speech was initially slurred when he came to on the ground. The neighbors contacted FF1’s family and he was transported to the ED in their private vehicle. FF1 reported that he “did not drink very much, had not really had much to eat today” and felt that he “woke up after a minute”. His job as a firefighter was noted on the chart along with a denial of any past medical history. He denied any recent illness, fever, chills, headache, chest pain or pressure, or abdominal pain. Vital signs on 18:58 arrival to ED were oral temperature of 98.5℉, heart rate 98 beats per minute (bpm), respiratory rate 18 breaths per minute, blood pressure 142/79 millimeters of mercury (mmHg) and oxygen saturation 96% on room air. His physical examination was unremarkable, and no trauma related to his fall from standing height was noted. Head CT was negative. Orthostatic syncope (also called postural hypotension) is tested by obtaining blood pressures when the patient is supine (lying flat on their back) and then immediately after being asked to stand. A positive test for orthostatic syncope is defined as a drop in systolic blood pressure by at least 20 mmHg and a drop in diastolic blood pressure of at least 10 mmHg within 3 minutes of standing. In FF1’s case, his orthostatic blood pressures did not meet both of these conditions going from 142/79 mmHg while supine to 127/74 mmHg while standing but he reported feeling “shaky” on standing. An electrocardiogram (EKG) showed a normal sinus rhythm of 93 bpm with no ST segment depression/elevation indicative of cardiac ischemia/infarction. It is important to note that the aforementioned maneuver and blood pressure testing to check for orthostatic syncope has a sensitivity for hypovolemia/dehydration of only 70% [Probst 2022]. Blood tests ordered by the ED physician included a complete blood count which did not show evidence of anemia and a marker for cardiac injury was negative. Chemistry panels showed a normal blood glucose level, liver function, and a ratio of blood urea nitrogen (BUN) to creatinine of 6. Dehydration may be evidenced by a BUN: creatinine ratio > 20. ED notes state that FF1 felt better after 2 liters of intravenous fluids. Repeat vital signs prior to discharge: blood pressure 134/83 mmHg, heart rate 88 bpm, respiratory rate 12 breaths/min and oxygen saturation 97% on room air. Hypovolemia / dehydration is often accompanied by a heart rate > 100 bpm and/or low blood pressure which normalizes once the fluid/volume deficit is corrected but FF1 did not have those on arrival to the ED and his vital signs did not change significantly after receiving IV fluids. Although FF1 stated he felt better after receiving IV fluids, which would be subjective evidence of dehydration/hypovolemia, his chemistry panel did not reflect this, he did not meet criteria for orthostatic syncope, and there were no significant changes in his vital signs between ED arrival and after receiving the IV fluids as would be expected after correcting a volume deficit. Therefore, his

Page 9 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 discharge diagnosis of “vasovagal syncope or dehydration” included instructions to stay hydrated, and “pursue further outpatient evaluation with the primary care physician or the other designated or consulting physician as outlined in the discharge instructions”. Vasovagal syncope (VVS) is purely a clinical diagnosis (i.e., there is no diagnostic test for it). VVS is considered a type of reflex or neurally mediated syncope and it is a separate clinical entity from orthostatic syncope. In VVS, the patient usually experiences symptoms immediately preceding the LOC such as nausea, lightheadedness, sweating, warmth, or chills, and blurred or darkening vision. Triggers for vasovagal syncope may include fear/anxiety, prolonged standing, viewing a traumatic scene, etc. [Probst 2022]. However, it was noted FF1 did not have a primary care physician and there were no specialist referrals on the discharge form. FD-A confirmed that FF1 did not have a primary care provider at the time of his death. A first syncopal episode should have further evaluation to determine the cause especially when the initial evaluation does not clearly identify the cause as in this case [Probst 2022; Shen et al. 2017]. Syncope refers to a transient loss of consciousness (LOC) caused by inadequate blood flow to the brain. The LOC associated with a syncopal episode is characterized by “rapid onset, short duration, and spontaneous complete recovery”. Syncope can be a presenting symptom of many conditions ranging from life-threatening to benign. Therefore, it is important for someone experiencing a first syncopal episode to have a thorough medical evaluation. An initial evaluation of a syncopal episode often occurs in an ED that conducts an initial assessment for immediate life-threatening conditions including but not limited to acute myocardial infarction, aortic dissection, arrythmia (ventricular tachycardia/fibrillation, high degree atrioventricular blocks, etc.), hypertrophic cardiomyopathy, pulmonary embolism, intracranial bleed, etc. Other causes of syncope are related to vascular tone such as vasovagal/neurocardiogenic disorder, orthostatic hypotension (body cannot maintain perfusion to brain when going from supine to standing position), disorders of the carotid sinus which contain baroreceptors to monitor mean arterial blood pressure (MAP) and can send signals to the heart and brain to adjust vessel tone, heart rate, cardiac stroke volume, etc. to compensate for drops in the MAP. Syncope can even be caused by “bearing down” when coughing or during defecation. Once immediate life-threatening causes are ruled out, the patient can then be provided outpatient referrals to various specialists (i.e., cardiology, neurology, etc.) to determine the cause of the syncopal episode which will then guide treatment options. Other conditions may mimic syncope such as transient low blood sugar, hypoxia due to symptomatic anemia, and certain types of seizures [Furlan et al. 2024; Grossman and Badireddy 2023; Martone et al. 2024]. As per FD-A’s Leave Policy dated August 2019, “An employee who is unable to report for duty due to illness or injury shall notify their regularly assigned supervisor or a supervisor where the employee is assigned for the affected duty shift…..It is the employee’s responsibility to keep their supervisor informed of progress of the illness or injury on a monthly basis and an estimated return to duty date. When 48 or more consecutive hours of sick leave are used, the employee must provide a Commissioned Fire Personnel Return to Work Form, prior to or upon returning to work. This form must be signed by a health care provider and include information indicating the employee is able to perform assigned duties”. In a later section of that same document under the “Return to Work” section, it states that “a doctor’s release is required when returning to duty”. FD-A records show FF1 returned to work for his next scheduled shift on May 18, 2024, 0700 to May 19, 2024, 0700. Since his syncopal episode and subsequent ED evaluation occurred during non-duty hours and he was able to resume

Page 10 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 work as scheduled, he was not required by FD-A policy to inform his supervisor or undergo a Return To Work evaluation. FD-A does not require annual medical evaluations of its members as per recommendations in NFPA Standard 1582. In this case, lack of an annual NFPA medical examination may represent a missed opportunity for FF1 to report the syncopal episode to a physician who could have coordinated specialist evaluations to determine if he had an underlying medical condition that could have impacted his ability to perform his job especially since FF1 did not have his own primary care physician to fulfill this role. Recurrent syncope is listed as a cardiovascular condition that could impact the safe performance of essential firefighter job tasks in Section 9.7 of this standard. In Section 7.5 of the standard, it states that the annual medical examination should include a medical history questionnaire to “provide baseline information with which to compare future medical concerns” and this questionnaire should include “changes in health status and known occupational exposures since the previous annual evaluation” so this may have been a way for the syncopal event to have been captured in his FD medical record and sent for further evaluation [NFPA 2022]. Sudden Cardiac Death Sudden cardiac death (SCD) has been a recognized entity in the U.S. Fire Service for over a decade. Although cardiovascular disease (CVD) events of all types, including heart attack, are the leading cause of non-traumatic death in firefighters, SCD is thought to be mainly due to arrhythmias which usually occur in the setting of pre-existing cardiac conditions involving structural abnormalities such as an enlarged heart (cardiomegaly), atherosclerotic cardiovascular disease (ASCVD), thickening of the heart muscle (hypertrophy), etc. [Smith et al. 2013]. Many schools have instituted pre-participation screening of potential athletes for undiagnosed hypertrophic cardiomyopathies (HCMs), thickening of the portion of the heart muscle that separates the ventricles that can bulge out resulting in blood flow obstruction during physical activity and/or place an individual at higher risk for potentially fatal arrythmias leading to SCD. HCMs are the most common cause of SCD among pediatric/ young adult athletes accounting for about 36% of cases [Horeczko 2022]. Endurance and strength training in athletes can induce remodeling of the heart’s structure and conduction system making an evaluation by a cardiologist a prudent action prior to approving participation in sports. The American Heart Association (AHA) has developed a standardized screening tool for this purpose [Lammlein et al. 2018]. HCM was noted as the most common cause of death in US athletes overall under 35 years of age while atherosclerotic coronary artery disease is the most common cause of SCD in athletes over 35 years of age [Maron et al. 2007]. FF1’s coronary arteries were clear and there were no structural abnormalities found in the heart. The weight of the heart at 400 grams was within the normal range for adult males [Sens and Hughes 2021]. While HCM, heart valve defects, etc. are associated with anatomical abnormalities, arrythmias may result in SCD without leaving macroscopic or histological evidence to be found at autopsy. Some arrythmias are intermittent and would only be found during prolonged Holter monitoring if not captured on a snapshot EKG [Bragg et al. 2024]. There are a class of cardiac arrythmias involving defects in the protein that forms calcium channels because of genetic mutations. Calcium, sodium, and potassium ion channels in the myocardial cell membranes are the base components of the cardiac conduction system. Molecular pumps use these channels to create the voltage gradients that form the electrical signals sent to the various structures of the heart resulting in coordinated contraction for

Page 11 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 efficient pumping. Calcium channel defects include Brugada Syndrome as well as Long and Short QT Syndromes and of these, Long QT Syndrome (LTQS) is the most common. Studies have identified ten different genetic mutations causing the defect in the protein that forms the calcium ion channel in LQTS [Grant 2009; Jouriles 2022]. FF1 had an exercise stress test via treadmill done as part of his preplacement medical evaluation in 2019 and an EKG done during his ED visit for syncope in 2024 and neither noted abnormalities consistent with these or other types of cardiac conduction disorders. One of the challenges in deaths due to SCD, is that there is often no evidence of the cause to be found during autopsy as in this case. This was reflected in a large study using autopsy data from the U.S. Department of Defense’s Cardiovascular Death Registry which analyzed 902 nontraumatic sudden deaths of individuals over 18 years of age. Although 79% of these deaths were determined to be due to cardiac causes, no cause could be identified in 21% of these cases [Eckart et al. 2011]. There have been reports of possible increased risk of SCD following COVID-19 infection; particularly in those who develop long COVID-19 [Karimi et al. 2024; Stimart and Hipkins 2024]. However, since FF1 did not have any documented testing confirming COVID-19 infection following his suspected exposure in November 2019, no further association can be made. Concerns were also raised if COVID-19 vaccination could have played a role in this fatality. There was no documentation provided that FF1 received a COVID-19 vaccination. However, there is no evidence that COVID-19 vaccinations have resulted in sudden cardiac death [Liko and Cieslak 2024]. Review of data provided through the Vaccine Adverse Event Reporting System (VAERS) did show an association between COVID-19 vaccination and occurrence of myocarditis, an inflammation of the heart muscle, and/or pericarditis, an inflammation of the fluid filled sac around the heart, in children and young adults. These conditions are usually accompanied by severe symptoms of chest pain, heart palpitations, and/or shortness of breath and elevations of blood markers for inflammation (increased white blood cell count, erythrocyte sedimentation rate, etc.) and myocardial damage (troponin I, etc.) [CDC 2023]. A large, retrospective study of data from 4 health plans covering over 100 million U.S. adults found that myocarditis and/or pericarditis following mRNA COVID-19 vaccination was rare with only 411 documented cases among 15 million people receiving approximately 27 million doses of vaccine [Wong et al. 2022]. There were no reports of FF1 having symptoms consistent with myocarditis or pericarditis. His EKG and the blood test during his May 16, 2024, ER visit did not show any evidence of myocardial damage. Heat Stroke/Rhabdomyolysis Historical weather data showed that on the date and approximate time of his syncopal episode, the outdoor temperature was 75°F with 60% humidity resulting in a heat index of 75°F. On August 1, 2024, at 1500, the approximate time of arrival at the residential fire, the outdoor temperature was 96°F with 44% humidity resulting in a heat index of 103.4°F. It was noted that the cab of the apparatus FF1 was driving during his August 1, 2024, shift was not air conditioned and that crewmembers had reported FF1 stating that he felt hot inside the cab. Exertional heat stroke can occur even in seemingly “cool” environments such as a 25-year-old firefighter who died of this after 30 minutes of engaging in an indoor maze course whose thermostat was set to 66°F while in full turnout gear and SCBA [NIOSH 2018]. It is vital for fire departments not to underestimate the contribution of exertional heat to the net heat load especially while engaged in firefighter essential tasks in a hot environment as in this case.

Page 12 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 Rhabdomyolysis is the constellation of signs and symptoms that result when muscle tissue is damaged. This damage can be caused by a variety of insults such as direct mechanical disruption, overheating the muscle, overexertion, direct toxicity from medications, decreased blood flow to muscles due to medical conditions such as sickle cell disease or use of vasoconstrictive substances such as caffeine, etc. [CDC 2024; Cervellin et al. 2017]. Both heat stroke and rhabdomyolysis may result in seizures and death but since there is no documentation of any temperature taken in the ED and no creatine kinase (CK) levels, the blood marker for rhabdomyolysis, it is not possible to determine if either condition played a part in this death. NIOSH has created information sheets for structural firefighters regarding their risk for rhabdomyolysis: What Structural Fire Fighters Need to Know about Rhabdomyolysis along with a wallet card that structural firefighters can show to healthcare providers when they present for treatment of possible signs and symptoms of rhabdomyolysis. There is a companion document for healthcare providers of structural firefighters reminding them of their increased risk for rhabdomyolysis as a result of their job titled, “Rhabdomyolysis in Structural Fire Fighters: A Patient Population at Risk”. General guidance regarding heat stress evaluation and mitigation along with information on identification of heat-related illness (HRI) including heat-related rhabdomyolysis, is provided in the most recent edition of the NIOSH heat stress criteria document [NIOSH 2016]. Structural firefighters may be exposed to multiple different sources of heat contributing to their overall heat load. Aside from ambient environmental conditions that can be exacerbated due to relative humidity reflected by the local heat index, other heat sources may include heat generated by the fire they are fighting, and metabolic heat generated by their muscles due to the exertion involved in performing their essential job tasks. The turnout gear worn as part of their PPE, can worsen firefighter heat exposure due to increased exertion induced by the extra weight that the firefighter must carry resulting in the generation of more metabolic heat. In addition, the heavy materials used in turnout gear can trap heat by blocking the body’s usual methods of dissipating heat through evaporative heat loss from sweating and direct heat transfer from the body to the environment from dilated blood vessels underneath the skin. NIOSH has conducted Health Hazard Evaluations regarding heat exposure and rhabdomyolysis in structural firefighters during training [NIOSH 2015, 2019]. It was also noted that the bay in FD-A in which the exercise equipment is located was not air conditioned. Fitness centers should be located in areas where HVAC systems can offset the heat and/or humidity (from sweat and exhalation) generated by those utilizing the workout area in order to reduce the possible adverse health effects of heat exposure during physical fitness training. Studies have also shown that the indoor air conditions of fitness centers have impacts on the effectiveness of both aerobic and anaerobic training [Huang et al. 2021]. The American College of Sports Medicine recommends an air temperature 65°F–68°F (18.3℃–20℃) and a relative humidity of 40%–60% for areas in which aerobic exercise, strength training, or Pilates workouts are occurring [ACSM 2012]. Caffeine Toxicity Caffeine is a stimulant that impacts the cardiovascular system by increasing heart rate directly and increasing blood pressure through its vasoconstrictive properties, both of which place additional strain on the heart which can increase risk for an acute cardiac event especially in the setting of increased physical exertion. Vasoconstriction also results in decreased blood flow to the supplied tissues thereby

Page 13 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 increasing the risk for rhabdomyolysis. The products noted by the Fire Marshal on the receipt found in FF1’s personal effects included Mio water enhancer, Celsius® Energy Drink, and Liquid I.V. The Mio product line is made by Kraft Heinz food company and comes in 3 different types: Original, Hydrate, and Energy. The original and hydrate lines are caffeine free while Mio Energy contains 90 mg caffeine per serving and each 1.62 fluid ounce (fl oz) bottle of Mio Energy contains 12 servings for a total caffeine content of 1.08 grams per bottle. Celsius® Energy Drink is produced by a company of the same name, and it comes in 3 different lines: Celsius® original (12 oz cans), Celsius® Essentials (16 oz cans), and Celsius® On-the-Go (0.17–0.20 oz powder packets). These products contain between 200–300 milligrams (mg) caffeine per can/packet. The Liquid IV line has 5 products: Hydration Multiplier®, Hydration Multiplier® sugar-free, Immune Support, Gut Health, and Energy Multiplier. Both versions of the Hydration Multiplier® along with the Immune Support, and Gut Health products do not contain any caffeine. The Energy Multiplier contains 100 mg caffeine per serving. The Fire Marshal’s notes did not specify which specific product of each of these product lines were purchased. Caffeine is a widely used stimulant and it is estimated that 85% of the US population drinks at least one caffeinated beverage per day. Moderate caffeine intake of less than 400 mg/day is considered safe. Caffeine content of common items is 30–60 mg per 12 oz. of soda, 50 mg in 8 oz of black tea, and 80 mg per 8 oz coffee. Caffeine is quickly absorbed through the gastrointestinal tract and crosses the blood-brain barrier to directly affect the central nervous system to cause increased alertness and decreased fatigue. It impacts the autonomic nervous system by increasing heart rate, respiratory rate, and blood pressure. Blood pressure is increased via constriction of the smooth muscle in the walls of blood vessels. Excessive intake can lead to caffeine toxicity resulting in heart attack, cardiac arrest, tachycardia, hypertension, and rhabdomyolysis along with seizures, hallucination, anxiety, nausea/vomiting and abdominal pain. Case reports of deaths due to caffeine toxicity have reported postmortem blood caffeine levels 160–400 µg/ml [Banerjee et al. 2014; Brower and Swatek 2024]. NMS labs who conducted the postmortem caffeine test on FF1 reported that levels > 50 µg/ml are considered toxic. Therefore, FF1’s postmortem caffeine level of 4.6 µg/ml would not be considered high enough to have played a significant role in his death. However, first responders should monitor their caffeine intake due to its potential to exacerbate cardiac demands in a job that is already strenuous. In addition, due to its vasoconstrictive properties, caffeine is a known risk factor for rhabdomyolysis and increases the risk for HRI by opposing the body’s dilation of the small blood vessels under the skin for heat dissipation. It is good practice for firefighters to reduce exposures to known risk factors for rhabdomyolysis and HRI due to the risks for the conditions inherent in their job. Cyanide Toxicity Cyanide is a known product of combustion for many items found in a residential location. Inhalation of cyanide byproducts in fire and smoke inhalation victims can be a cause of cardiac arrest. Therefore, the actions of the EMS crew in administering the cyanide antidote, hydroxocobalamin, were reasonable. Cyanokit®, the brand name that hydroxocobalamin is marketed under and how it is referred to in the FD-B protocol, was approved by the U.S. Food and Drug Administration in 2006. The manufacturer states that its use is indicated for “known or suspected cyanide poisoning. If clinical suspicion of cyanide poisoning is high, Cyanokit® should be administered without delay” [Serb Pharmaceuticals

Page 14 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 2024]. One of the challenges of cyanide toxicity is that blood tests for cyanide often take hours to result. Since cyanide acts by disrupting cellular processes, hypoxia and metabolic acidosis quickly evolve and the latter can be detected by rapid tests for lactate and an anion gap [Huzar et al. 2013]. However, disruption of cellular functions that occur during a cardiac arrest can themselves result in elevated lactate levels and metabolic acidosis, so these conditions are not specific to cyanide toxicity. Although FF1’s postmortem blood sample was not tested for cyanide, it was tested for carbon monoxide, another combustion byproduct that would lend evidence to smoke inhalation, and that was not detected. One study measured combustion byproduct gases in fire smoke including carbon monoxide, hydrogen cyanide, hydrogen sulfide, ammonia, nitrogen dioxide, and sulfur dioxide over a period of 16 months within the Chicago Fire Department found that of all the 6 gases measured, carbon monoxide had the highest maximum exposure and the highest total concentration [Fabian et al. 2014]. Carbon monoxide and cyanide levels in the blood of smoke inhalation victims have also been shown to correlate [Baud et al. 1991]. FF1 was wearing his SCBA when he entered the residence and had disengaged it when he exited the structure. Since he was not breathing supplied air at the time of his collapse, cardiac arrest from cyanide exposure from smoke inhalation was possible. However, it would be unlikely to have a fatality due to smoke inhalation injury without evidence of any exposure to carbon monoxide. Regardless, due to the fast progression of cyanide toxicity and its lethality, presumptively treating a cardiac arrest at a fire scene with the cyanide antidote is appropriate and in line with current FD-B protocol but due to the lack of evidence of CO exposure, it is unlikely that cyanide toxicity played a role in this fatality. Contributing Factors Occupational injuries and fatalities are often the result of one or more contributing factors or key events in a larger sequence of events that ultimately result in the injury or fatality. Since the cause of death could not be definitively determined in this case, there is no evidence that the following recommendations would have prevented this fatality, but they are being provided here as a reminder of good practice. Recommendations Recommendation #1: Fire departments should consider implementing annual medical examinations as per NFPA 1582 recommendations. Discussion: In this case and many others, firefighters may not have their own primary care physician and the provider of the annual NFPA medical evaluation may be the only physician they see regularly. Section 9.1.2 of this standard states that “The occupational medical evaluation shall include a medical history, examination, and any medical tests required to assess medical conditions that can affect an individual's ability to safely and effectively perform the essential job tasks.” In this case, if an annual medical examination had been conducted between May 16, 2024, and August 1, 2024, it is unknown if FF1 would have disclosed his syncopal event and if so, if a further evaluation of it coordinated by the physician conducting the exam would have prevented his death but it does represent a potential missed opportunity.

Page 15 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 Recommendation #2: Fire departments should encourage members to reduce and if possible, eliminate intake of/exposure to known risk factors for adverse cardiovascular events and rhabdomyolysis. Discussion: Firefighters have multiple risk factors for adverse cardiac events and rhabdomyolysis due to various inherent aspects of the essential job tasks including intense, physical exertion while in a hot environment. It would be prudent to avoid or at least reduce intake/exposure to additional known risk factors for both cardiac events and rhabdomyolysis such as caffeine use, poor nutrition leading to formation of cholesterol plaques inside blood vessels decreasing blood flow, etc. Fire departments should also encourage seasonal influenza and COVID-19 vaccinations as both these infections have been associated with increased risk for rhabdomyolysis [Singh and Scheld 1996; Singh et al. 2020]. Ensuring that fitness centers inside fire stations and apparatus cabs are air conditioned, will help reduce risk of HRI, heat-related rhabdomyolysis, and adverse cardiac events while encouraging improved physical fitness. Recommendation #3: Fire departments should increase awareness of conditions that increase risk of heat-related illness (HRI) and implement heat stress management policies to reduce those risks. Discussion: Heat stress is the net exposure experienced which can be from multiple sources such as ambient environmental conditions, active heat source such as a fire, and heat generated by muscles through physical exertion. Heat exposure in structural firefighters can be exacerbated by wearing turnout gear that can trap heat due to disruption of the body’s usual routes of heat dispersion of evaporative heat loss from sweating or direct conductive heat transfer to the environment. Heat strain is the body’s response to heat stress and consists of a spectrum of clinical conditions from heat cramps to the most severe and potentially life-threatening heat stroke. Although many heat sources are unavoidable for structural firefighters, efforts should be made to mitigate those that can be reduced. In this case, on the day FF1 died, the outdoor temperature was 96°F with 44% humidity resulting in a heat index of 103.4°F. The cab of the apparatus FF1 was driving that shift was not air conditioned and crewmembers had reported FF1 stating that he felt hot inside the cab. Fire departments should consider purchasing apparatus with air-conditioned cabs to reduce the heat stress firefighters inside that cab experience due to environmental conditions but also to act as a cooling location if a firefighter should develop heat-related illness on scene before an EMS unit is present. Similarly, fitness centers located in fire stations should also be air conditioned to reduce the risk of HRI in those using those facilities. Lastly, this fatality occurred shortly after the incident response began, otherwise with an event of extended duration, a rehab location would be protocol to provide cooling and rehydration along with a crew rotation as a heat stress mitigation approach. Recommendation #4: Ensure EMS with transport capability is on-scene at any fire or special operation location in the event emergency medical care and transport is needed. Discussion: Based on resource availability in this jurisdiction, EMS transporting units were not automatically sent to structure fires on the initial dispatch nor at the report of a working incident by initial on-scene firefighting personnel. Normal operations in this jurisdiction give EMS transport unit requesting authority to the IC. In this incident, the IC requested a transporting unit upon witnessing FF1 collapse. Given normal jurisdictional procedures, this operation appropriately requested an EMS transport unit.

Page 16 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 Strain on EMS system resources can affect many aspects of emergency response, and that challenge has no doubt impacted agencies across the United States. It is imperative to find ways to get needed resources to an emergency incident. In this case, ALS capability was part of the first due engine company and the efforts of the firefighter paramedic, along with on-scene firefighter EMT-Basics, afforded FF1 the best opportunity for survival. In this incident, the response delay of the transporting unit did not delay advanced care. However, at fire scenes where a FFP is not on a responding apparatus, there would be a delay in initiation of ALS care and transport if an ALS EMS unit is not dispatched as soon as a working fire is identified. The NFPA 1550, Standard for Emergency Responder Health and Safety states in Paragraphs 10.6.9 – 10.6.11 and A.10.6.9 that basic life support or advanced life support (preferred) with transportation capabilities should be available on scene when members are operating at a structure fire or performing special operations, and for medical monitoring at HAZMAT incidents, and considered based on risk management at other emergency operations. [NFPA 2024]. Recommendation #5: Fire officers should ensure crew integrity when entering, operating in, and exiting from an environment immediately dangerous to life or health (IDLH). Discussion: In this incident, the crew was assigned to pull a 2nd 1-3/4-inch hoseline for interior attack. Crew members reported that they went in together. Once inside, it was reported that the crew was split into two teams, with FF1 and the nozzleman in the living room to hit spot fires. Helmet cam video from FF1 provides a few minutes of conditions and actions beginning outside readying to enter the residence and continues inside the residence until the helmet cam is turned off. The video shows: The fire officer entered and was followed by the nozzleman and back up. However, before FF1 enters, several firefighters from other companies were in line at the side A door working their way in. These firefighters pushed the hoseline while FF1 pulled at hot spots along the roofline. FF1 then entered and made his way into the residence before meeting up with his crew. The crew worked on dousing a debris fire from roofing materials collapsed into the residence. In the video, the crew goes out of view as FF1 remains in the same area of the residence with other firefighters passing by. The crew appears again in the video before the video is turned off. The video does not continue to when FF1 tells one of the crew that he’s leaving the structure. FF1 reportedly exits on his own. The fire officer was unaware . that FF1 was exiting the structure until after IC reassigned the crew to the side A porch for a small fire under the porch. NFPA 1550, Standard for Emergency Responder Health and Safety states in Paragraph 10.5.6 that company officers shall maintain an ongoing awareness of the location and condition of all company members. Paragraph 10.5.7 states that, where assigned as a company, members shall be responsible to remain under the supervision of their assigned company officer [NFPA 2024]. The International Association of Fire Chiefs’ Safety, Health, and Survival Section has redefined the Rules of Engagement for Structural Fire Fighting. One of the objectives is to ensure that firefighters always enter a burning building as a team of two or more members and no firefighter is allowed to be alone at any time while entering, operating in, or exiting a building. A critical element for firefighter survival is crew integrity. Crew integrity means firefighters stay together as a team of two or more. They must enter a structure together and remain together at all times while in the interior, and all members come out together.

Page 17 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 Crew integrity starts with the company officer ensuring that all members of the company understand their riding assignment, have the proper personal protective equipment, and have the proper tools and equipment. Upon arrival at the incident, the company is given a task to perform by the incident commander. The company officer communicates to the members of the company what their assignment is and how they will accomplish their assignment. Just as the members of a company enter a hazardous environment together, they should leave together to ensure that crew integrity is maintained. If one member has to leave, the whole company leaves together [IAFC 2012]. It is the responsibility of every firefighter and company officer to stay in communication or contact with crew members at all times by visual observation, voice, or touch while operating in the hazard zone. All firefighters should maintain the unity of command by operating under the direction of the incident commander, division/group supervisor, or their company officer at all times. The ultimate responsibility for crew integrity and ensuring no members get separated or lost rests with the company officer. If crew integrity is lost, firefighters are placed at increased risk. A Mayday should be called if any member cannot be accounted for as described above. Company officers should give an accountability report upon exiting the hazard zone to either Command or their assigned division or group supervisor. Recommendation #6: Fire departments should provide physicians tasked with pre-employment and return to work medical evaluations, descriptions of the essential tasks and duties and not rely on the cadet/incumbent to relay this information. Discussion: NFPA 1582 Section 4.1.5 states “The fire department shall provide the fire department physician with a fire service overview, current job descriptions, and the essential job tasks required for all fire department positions and ranks [NFPA 2022].” On FF1’s preplacement Respirator Physical – Physician’s Recommendations form, there are 2 options under the “Job Description” section, and they are: • Yes, a specific respirator job description was provided by the Company. This recommendation is based upon this description and history/physical information gained today. • No, a respirator job description was not provided by the Company and this recommendation is based solely upon history/physical information gained today. The nurse who signed off on that form on February 19, 2019, selected the “No” option. Fire Departments should consider including in their pre-employment medical evaluation form packets a document containing a description of their essential tasks to ensure all providers involved in these evaluations have access to that information and can affirm their awareness of these tasks upon signing off on their respective components of the candidate’s medical condition. Recommendation #7: Fire departments should ensure that a physician signs off on pre-employment and return to work evaluations. Discussion: NFPA 1582 states medical evaluations should be conducted by a Fire Department Physician as per Section 7.2.1 “The fire department, the fire department physician, and member organizations where they exist shall be responsible to convey the purposes and importance of the annual occupational medical evaluation to members and to the AHJ”. Section 3.3.6 defines a Fire Department Physician as “A licensed doctor of medicine or osteopathy who has been designated by the

Page 18 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 fire department to provide professional expertise in the areas of occupational safety and health as they relate to emergency services”. Section 4.2.2. states “When medical evaluations are conducted by a physician or medical provider other than the fire department physician, the evaluation shall be reviewed and approved by the fire department physician.” In FF1’s case, his pre-employment clearance was signed off by an advanced practice registered nurse without clear evidence of physician oversight or documentation of review by the fire department physician. References American College of Sports Medicine [2012]. ACSM’s Health/Fitness Facility Standards and Guidelines. Fifth Edition. Champaign, Illinois: Human Kinetics. Banerjee P, Ali Z, Levine B, Fowler DR [2014]. Fatal caffeine intoxication: a series of eight cases from 1999 to 2009. J Forensic Sci 59(3):865–868. Baud FJ, Barriot P, Toffis V, Riou B, Vicaut E, Lecarpentier Y, Bourdon R, Astier A, Bismuth C [1991]. Elevated blood cyanide concentrations in victims of smoke inhalation. N Engl J Med 325(25):1761– 1766. Bragg S, Brown B, DeCastro AO [2024]. Arrhythmias and Sudden Cardiac Death. Prim Care 51(1):143–154. Brower JO, Swatek JL [2024]. Beyond the buzz: the fatal consequences of caffeine overconsumption. J Anal Toxicol 48(8):535–540. Centers for Disease Control and Prevention [2023]. Clinical considerations: myocarditis after COVID-

19 vaccines | CDC. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. Centers for Disease Control and Prevention [2024]. Rhabdomyolysis. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. Cervellin G, Comelli I, Benatti M, Sanchis-Gomar F, Bassi A, Lippi G [2017]. Non-traumatic rhabdomyolysis: background, laboratory features, and acute clinical management. Clin Biochem 50(12):656–662. Eckart RE, Shry EA, Burke AP, McNear JA, Appel DA, Castillo-Rojas LM, Avedissian L, Pearse LA, Potter RN, Tremaine L, Gentlesk PJ, Huffer L, Reich SS, Stevenson WG [2011]. Department of Defense Cardiovascular Death Registry Group. Sudden death in young adults: an autopsy-based series of a population undergoing active surveillance. J Am Coll Cardiol 58(12):1254–1261. Fabian TZ, Borgerson JL, Gandhi PD, Baxter CS, Ross CS, Lockey JE, Dalton JM [2014]. Characterization of firefighter smoke exposure. Fire Technol 50:993–1019. Furlan L, Jacobitti Esposito G, Gianni F, Solbiati M, Mancusi C, Costantino G [2024]. Syncope in the emergency department: a practical approach. J Clin Med 13(11):3231 Grant AO [2009]. Cardiac ion channels. Circ Arrhythm Electrophysiol 2(1):185–194. Grossman SA, Badireddy M [2024]. Syncope. In: StatPearls [Internet]. Treasure Island, FL: StatPearls Publishing. Available from: https://www.ncbi.nlm.nih.gov/books/NBK442006/.

Page 19 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 Horeczko T [2022]. Chapter 165: Pediatric Cardiac Disorders. In Rosen’s Emergency Medicine 10th Edition. Walls R, Hockberger R, Gausche-Hill M, Erickson TB, and Wilcox SR Editors. Published by Elsevier. Huang C, Que J, Liu Q, Zhang Y [2021]. On the gym air temperature supporting exercise and comfort. Build Environ 206:108313. Huzar TF, George T, Cross JM [2013]. Carbon monoxide and cyanide toxicity: etiology, pathophysiology and treatment in inhalation injury. Expert Rev Respir Med 7(2):159–170. IAFC [2012]. Rules of engagement for structural firefighting: increasing firefighter survival. Draft 10 developed by the Safety, Health and Survival Section International Association of Fire Chiefs. Jouriles NJ [2022]. Chapter 68: Pericardial and myocardial disease. In Rosen’s Emergency Medicine 10th Edition. Walls R, Hockberger R, Gausche-Hill M, Erickson TB, and Wilcox SR Editors. Published by Elsevier. Pgs. 1112–1126. Karimi M, Faal Hamedanchi N, Ansari K, Nahavandi R, Mazdak M, Javaherchian F, Koochaki P, Asadi Anar M, Shirforoush Sattari M, Mohamaditabar M [2024]. Rhabdomyolysis secondary to COVID-19 infection and vaccination: a review of literature. Front Med (Lausanne). Lammlein KP, Stoddard JM, O'Connor FG [2018]. Preparticipation screening of young athletes: identifying cardiovascular disease. Prim Care 45(1):95–107. Liko J, Cieslak PR [2024]. Assessment of risk for sudden cardiac death among adolescents and young adults after receipt of COVID-19 vaccine - Oregon, June 2021-December 2022. MMWR 73(14):317– 320. Maron BJ, Thompson PD, Ackerman MJ, Balady G, Berger S, Cohen D, Dimeff R, Douglas PS, Glover DW, Hutter AM Jr, Krauss MD, Maron MS, Mitten MJ, Roberts WO, Puffer JC [2007]. American Heart Association Council on Nutrition, Physical Activity, and Metabolism. Recommendations and considerations related to preparticipation screening for cardiovascular abnormalities in competitive athletes: 2007 update: a scientific statement from the American Heart Association Council on Nutrition, Physical Activity, and Metabolism: endorsed by the American College of Cardiology Foundation. Circulation 115(12):1643–1655. Martone AM, Parrini I, Ciciarello F, Galluzzo V, Cacciatore S, Massaro C, Giordano R, Giani T, Landi G, Gulizia MM, Colivicchi F, Gabrielli D, Oliva F, Zuccalà G [2024]. Recent advances and future directions in syncope management: a comprehensive narrative review. J Clin Med 13(3):727. NFPA [2022]. NFPA 1582. Standard on Comprehensive Occupational Medical Program for Fire Departments. Quincy, MA: National Fire Protection Association. NFPA [2024]. NFPA 1550. Standard for Emergency Responder Health and Safety. Quincy, MA: National Fire Protection Association. NIOSH [2015]. Evaluation of heat stress, heat strain, and rhabdomyolysis during structural firefighter training. By Eisenberg J, Methner M, Dowell CH, Mueller C. Cincinnati, OH: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, Health Hazard Evaluation Report 2012-0039-3242.

Page 20 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 NIOSH [2016]. NIOSH criteria for a recommended standard: occupational exposure to heat and hot environments. By Jacklitsch B, Williams WJ, Musolin K, Coca A, Kim J-H, Turner N. Cincinnati, OH: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication No. 2016-106. NIOSH [2018]. Cadet dies from hyperthermia and exertional heat stroke during indoor SCBA maze training—Texas. Cincinnati, OH: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2016-05. NIOSH [2019]. Evaluation of rhabdomyolysis and heat stroke in structural firefighter cadets. By Eisenberg J, Li JF, Feldmann KD. Cincinnati, OH: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, Health Hazard Evaluation Report 2012-0039-3242. NMS Labs [2022]. Caffeine, Blood-0930B. Horsham, PA: NMS Labs. Probst M [2022]. Chapter 11: Syncope. In Rosen’s Emergency Medicine 10th ed. Walls R, Hockberger R, Gausche-Hill M, Erickson TB, Wilcox SR eds. Published by Elsevier.

Serb Pharaceuticals [2024]. CYANOKIT® (hydroxocobalamin) | Cyanide Poisoning Treatment. Sens MA, Hughes R [2021]. Diagnostic pathology: forensic autopsy, 1st ed. Internal Examination Cardiovascular System. Philadelphia, PA: Elsevier. Shen WK, Sheldon RS, Benditt DG, Cohen MI, Forman DE, Goldberger ZD, Grubb BP, Hamdan MH, Krahn AD, Link MS, Olshansky B, Raj SR, Sandhu RK, Sorajja D, Sun BC, Yancy CW [2017]. 2017 ACC/AHA/HRS Guideline for the Evaluation and Management of Patients With Syncope: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. Circulation 136(5):e25–e59. Singh B, Kaur P, Reid RR [2020]. Case reports: rhabdomyolysis associated with COVID-19. Am Fam Physician 102(11):645–648. Singh U, Shceld WM [1996]. Infectious etiologies of rhabdomyolysis: three case reports and review. Clin Infect Dis 22(4):642–649. Smith DL, Barr DA, Kales SN [2013]. Extreme sacrifice: sudden cardiac death in the US Fire Service. Extrem Physiol Med 2(1):6. Stimart HL, Hipkins B [2024]. The negative effects of long COVID-19 on cardiovascular health and implications for the presurgical examination. J Osteopath Med. Online ahead of print. Wong HL, Hu M, Zhou CK, Lloyd PC, Amend KL, Beachler DC, Secora A, McMahill-Walraven CN, Lu Y, Wu Y, Ogilvie RP, Reich C, Djibo DA, Wan Z, Seeger JD, Akhtar S, Jiao Y, Chillarige Y, Do R, Hornberger J, Obidi J, Forshee R, Shoaibi A, Anderson SA [2022]. Risk of myocarditis and pericarditis after the COVID-19 mRNA vaccination in the USA: a cohort study in claims databases. Lancet 399(10342):2191–2199.

Page 21 27-Year-old Firefighter has Seizure then Dies of Cardiac Arrest while Responding to Residential Fire – Kansas Report # F2024-04 Investigator Information This incident was investigated jointly by the NIOSH Fire Fighter Fatality Investigation and Prevention Program (FFFIPP) Medical Team based in the Division of Field Studies and Engineering in Cincinnati, Ohio and the FFFIPP Trauma Team based in the Division of Safety Research in Morgantown, West Virginia. This investigation was conducted and this report was co-authored by Judith Eisenberg MD, MS, and Mr. Rob Saunders of the FFFIPP Medical Team and Tammy Schaeffer of the FFFIPP Trauma Team. Dr. Eisenberg is a board certified Emergency Medicine physician and Rob Saunders is a former NIOSH Technical Information Specialist. Mr. Saunders retired after 31 years with the Pike Township Fire Department, Indianapolis, Indiana. In addition to having served as a firefighter, paramedic, heavy rescue technician, and rescue diver, he has held the positions of company officer, Division Chief of Emergency Medical Services, Deputy Chief of Operations, and Fire Chief. Tammy Schaeffer is a Safety & Occupational Health Specialist. In addition, Mrs. Schaeffer served as a Pennsylvania Paramedic for 20 years and as a U.S. Army Combat Medic for 8 years, with deployment for Operation Iraqi Freedom in 2004 Disclaimer Mention of any company or product does not constitute endorsement by the National Institute for Occupational Safety and Health (NIOSH). In addition, citations to Web sites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or products. Furthermore, NIOSH is not responsible for the content of these Web sites.

FD06032 ｜ 2024_美国_NIOSH_face202405.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2024 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Page i

Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Executive Summary On September 4, 2024, a 35-year-old career battalion chief (Battalion 1) died while operating at a tractor trailer fire. He was the command officer at the incident and was killed while investigating the type of cargo that was held within the refrigerated trailer involved in the fire. At approximately 0921 hours, the local public safety answering point (PSAP) received a call for a vehicle fire. While the incident was being dispatched to the responding units, the call type was updated to a tractor trailer fire with black smoke and multiple callers. Battalion 1 and Engine 11 were dispatched to the incident. Upon hearing the update from the dispatch center, Battalion 1 requested an additional engine company and that station 1 be re-paged for a tanker response. At approximately 0928 hours, Battalion 1 arrived on-scene and positioned past the involved tractor trailer, blocking traffic. Battalion 1 established command and sized the incident up as a tractor trailer fire with heavy fire involvement at the rear of the tractor, including tires and diesel fuel, with unknown cargo and occupant status. Battalion 1 provided a further update to all responding units that there was off- gassing at the rear of the tractor with an unknown source. He ordered Engine 11 to stop short, deploy an attack line from the front bumper, and utilize foam. Engine 11 stopped approximately 100’ short of the involved tractor trailer and deployed the 100’ 1 ¾-inch front bumper attack line. The Chief of the department (Chief 1101), who arrived just before Engine 11, assisted with deployment of the attack line. Engine 11’s officer extended the line with a 50’ section of attack line that was stored in the bumper well. At approximately 0931 hours, Battalion 1 updated dispatch that they had water on the fire and walked away from the engine crew. At approximately 0932 hours, while investigating at the rear of the trailer, Battalion 1 updated all units on-scene that the cargo was still unknown, but the rear of the trailer indicated it was refrigerated. At approximately 0933 hours, Battalion 1 attempted to open the rear trailer doors. The trailer Photo 1: Fire conditions on arrival of Battalion 1. (Photo courtesy of the fire department) REPORT F2024-05 • March 2025

Page ii Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 exploded and Chief 1101 called to Battalion 1 on the radio stating they were okay. Battalion 1 did not acknowledge the message. At approximately 0934 hours, Chief 1101 updated dispatch that the trailer had exploded, was fully involved in fire, and he was going to check on Battalion 1. At approximately 0935 hours, Chief 1101 updated dispatch that Battalion 1 was down and requested air medical. At approximately 0937 hours, Chief 1101 assumed command of the incident and confirmed Battalion 1 was deceased on- scene. Contributing Factors • Standard operating procedures/guidelines (SOPs/SOGs) • Incident command • Risk/benefit analysis • Staffing • Hazard recognition, including control zone establishment with unknown hazard. Key Recommendations • Fire service organizations should develop and maintain SOPs/SOGs with industry best practices • Incident Commanders (ICs) should establish an Incident Management System (IMS) at all incidents • ICs should integrate risk assessment as a risk management practice into their fireground strategy and tactic SOGs/SOPs • Fire service organizations should follow the appropriate standards for organization and deployment of fire suppression operations, emergency medical operations, and special operations • ICs should establish control zones on incidents where immediately dangerous to life and health (IDLH) environments exist • Fire service training should provide insight and education towards commonly unrecognized pressure vessels in the form of intermodal freight trailers and mixed cargo trailers.

For further information, visit the program website at www.cdc.gov/niosh/firefighters/fffipp/ or call toll free 1-800-CDC-INFO (1-800-232-4636). To access the report slides that summarize this incident and the recommendations, visit www.cdc.gov/niosh/firefighters/programs/pdfs/face202405RS.pdf. The National Institute for Occupational Safety and Health (NIOSH) initiated the Fire Fighter Fatality Investigation and Prevention Program to examine deaths of fire fighters in the line of duty so that fire departments, fire fighters, fire service organizations, safety experts and researchers could learn from these incidents. The primary goal of these investigations is for NIOSH to make recommendations to prevent similar occurrences. These NIOSH investigations are intended to reduce or prevent future fire fighter deaths and are completely separate from the rulemaking, enforcement and inspection activities of any other federal or state agency. Under its program, NIOSH investigators interview persons with knowledge of the incident and review available records to develop a description of the conditions and circumstances leading to the deaths in order to provide a context for the agency’s recommendations. The NIOSH summary of these conditions and circumstances in its reports is not intended as a legal statement of facts. This summary, as well as the conclusions and recommendations made by NIOSH, should not be used for the purpose of litigation or the adjudication of any claim.

Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia

Page 1 REPORT F2024-05 • March 2025 Introduction On September 4, 2024, a 35-year-old battalion chief died while operating at a tractor trailer fire. On September 5, 2024, the U.S. Fire Administration notified the National Institute for Occupational Safety and Health (NIOSH) of this incident. From September 25 through October 2, 2024, two NIOSH Fire Fighter Fatality Investigation and Prevention Program (FFFIPP) investigators conducted interviews virtually with fire officers and firefighters who responded to the incident. Photos taken at the scene, computer aided dispatch (CAD) notes, radio traffic audio, helmet camera, body camera, and drive camera footage obtained from the fire department were reviewed. NIOSH investigators also reviewed the training records of personnel involved in the incident and reviewed the department’s SOPs and professional development model. Due to the virtual investigation, on September 16, NIOSH’s FFFIPP contacted the Fire Safety Research Institute (FSRI), part of the UL Research Institutes, to request assistance with fire analysis of this incident. Facts and the timeline surrounding the incident were discussed with the NIOSH team and gaps in the information were identified. Subsequently, on September 18, 2024, FSRI sent a team member to meet with the Chief of the Department to examine the remains of the trailer at a storage site in Georgia. The goal of the examination was to better understand the construction and fuel load of the refrigerated trailer itself. A sample of rigid foam insulation from the trailer was provided for testing to identify the foam type, and its potential as a fuel source for the smoke explosion. Fire Department Battalion 1 was a member of a combination fire department that has five stations with 31 full-time, 20 part-time, and 3 volunteer personnel that serve a population of approximately 25,000 within a 406 square mile area. The fire department is managed by a career fire chief employed by the county. The combination fire department is responsible for providing service to the entire county and has automatic aid agreements with the eight other volunteer departments located throughout the county. The area served by the department contains a myriad of occupancies including resorts, an airport, single and multiple family dwellings, light commercial, warehouses, and heavy industrial complexes. The fire department is responsible for structural fire, aircraft rescue firefighting services, hazardous material response, technical rescue, and emergency medical services (EMS). All department members, except for the fire chief, work a 24-hours on duty, 48-hours off duty rotating schedule. The minimum staffing for the department is 9 personnel including an on-duty battalion chief,

Page 2 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 with full staffing of 11 personnel. The on-duty battalion chief oversees the staffing of five engine companies, one hazardous materials unit, and one brush unit operating out of five separate fire stations. Two personnel are assigned to each of the five stations. Fire and EMS are separate divisions within the department. Training and Experience Battalion 1 had 17 years of firefighting experience with a background in both the career and volunteer fire service. The minimum qualifications for the battalion chief position within the department include national certification in Firefighter I and II, Apparatus Operator-Pumper, Fire Officer 1, 2, and 3, Fire Instructor 1 and 2, Incident Safety Officer, and State Emergency Medical Technician Intermediate (EMT-I) or Advanced (AEMT), or Paramedic.

Battalion 1 had various professional certifications including Firefighter I and II, Hazardous Materials Awareness and Operations, Apparatus Operator - Pumper, Fire Officer 1, and Fire Instructor 1. He was promoted to the rank of battalion chief with an agreement that all necessary qualifications would be completed before the end of July 2026. This exemption was granted due to the department changing their professional development program for minimum standards within each rank. The department training records also indicated he had completed continuing education training in a variety of topics, such as pipeline emergencies, strategies and tactics of fire control, and building construction. Apparatus, Staffing, and Communications The minimum daily staffing of the combination fire department is comprised of four engine companies (minimum of two personnel each) and a battalion chief. The full, daily staffing of the combination fire department is comprised of five engine companies (minimum of two personnel each) and a battalion chief. The engine company personnel are responsible for staffing of and response with other apparatus housed in their station based upon call type. Emergency calls are received into a county PSAP which utilizes a Motorola P25 800-Megahertz (MHz) encrypted tri-county simulcast trunking system and a CAD system. All dispatchers are state certified and required to complete a minimum of 10 hours of continuing education annually. The PSAP personnel in this incident averaged over 25 hours of continuing education and is currently in the process of obtaining Emergency Medical Dispatching certification. Each dispatcher handles the call and incident through completion, regardless of type of call. The various CAD event codes may prompt the call taker to ask additional and specific questions to provide to dispatched companies. This PSAP has a minimum staffing of two operators per shift with typical staffing of three or four operators per shift. The county system averages 40,000 calls yearly including emergency and non-emergency incidents for the sheriff, local, and state police departments, a combination fire department, eight volunteer fire departments, and three EMS departments. The PSAP also receives calls for public works, city workers, three animal control offices, city water, city gas, Department of Natural Resources (DNR), and railroads. This incident was dispatched using the event code for a vehicle fire and the CAD generated a specific, predetermined response plan to dispatch the appropriate units. See Table 1 for the companies dispatched on the original call and Table 2 for the fire unit status and staffing at the time of the explosion.

Page 3 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Table 1: Initial Incident Dispatch Resource Designation Staffing Level Engine 11 2 Personnel Battalion 1 1 Personnel

Table 2: Fire Unit Status and Staffing at Time of Explosion Resource Designation Staffing Level Status Engine 11 2 Personnel On Scene Battalion 1 1 Personnel On Scene Chief 1101 1 Personnel On Scene Chief 8 1 Personnel On Scene Engine 12 2 Personnel Enroute Tanker 1 0 Personnel Dispatched, not enroute Personal Protective Equipment Upon arrival at the scene, Battalion 1 exited his command vehicle in his daily work uniform (i.e., uniform boots, long pants, sweatshirt, and ball cap). He was wearing a radio strap and holder that held his portable radio and remote speaker microphone. As the incident progressed, Battalion 1 also donned a high visibility safety vest as well as his leather structural firefighting helmet. The helmet met current Occupation Safety and Health Administration (OSHA) standards. Environmental Conditions The weather on September 4, 2024, at 0853 hours was fair with no precipitation. The temperature was 69°F with winds out of the north northeast at 10 miles per hour. The humidity was 63% [Weather Underground 2024]. It should be noted that in Diagram 2 and other images contained within this report, the wind at the location of the fire was out of the east-southeast direction. This may be due to obstructions or topography that are different from the recording weather station.

Page 4 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Involved Vehicle The vehicle involved in this incident was a tractor trailer style commercial vehicle (see Diagram 1 and Photo 2). The trailer portion of the vehicle consisted of a container chassis with a refrigerated intermodal container secured to the chassis with twistlocks on all four corners of the container. The refrigerated intermodal container had a portable generator, or genset, affixed to the bulkhead of the container. The genset in this incident ran on diesel fuel and included a self- contained diesel fuel tank on the lower portion of the genset device. Based on the information provided by the UL Research Institutes team, a visual inspection of the diesel tank did not present any evidence of bulging or rupture. The container was loaded with frozen chicken. The 2024 Emergency Response Guidebook (ERG) identifies this type of trailer as a type 111 intermodal freight container. These containers are for general freight that carry bulk or non-bulk packages. They may transport hazardous materials/dangerous goods in small packages (e.g., bags, boxes, drums) or intermediate bulk containers (IBCs), sometimes referred to as “totes”. These

Photo 2. Type 111 Intermodal Freight Container secured to trailer chassis with genset. (Photo courtesy of UL FSRI) Diagram 1: Type 111 road trailers. (Courtesy of 2024 Emergency Response Guidebook)

Page 5 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 containers typically have the doors for entering the compartment at the rear opposite the bulkhead [PHMSA 2024]. A summary of FSRI’s trailer construction materials analysis findings can be found in Appendix One. Timeline The following timeline is a summary of events that occurred as the incident evolved. Not all incident events are included in this timeline. The times are approximate and were obtained by examining the dispatch records, audio recordings, witness statements, and other available information. All times are approximate and rounded to the closest minute. The timeline is not intended, nor should it be used, as a formal record of events. Time (hours) Fireground Operations, Response, and Details 0921 • PSAP case created and unit recommendations for vehicle fire response initiated. 0923 • Battalion 1 and Engine 11 dispatched. o PSAP updated responding units that they have received multiple calls of a tractor trailer box truck with black smoke coming from it. 0924 • Battalion 1 and Engine 11 enroute. 0927 • Battalion 1 requests additional engine company and Station 1 be re-paged for a Tanker. 0928 • Battalion 1 on-scene, established command. o “Heavy fire involvement at the rear of the tractor, unknown occupant status, unknown cargo”

0929 • Battalion 1 provides update to all responding units. o “Some type of off-gassing at the rear of the tractor, unknown source” 0930 • Battalion 1 orders Engine 11 to stop short and deploy the bumper line and begin using foam. 0931 • Engine 11 has water on the fire. 0932 • Battalion 1 updates all units on-scene stating, “unknown what the cargo is, it is a refrigerated trailer according to the rear.” 0933 • Trailer explodes.

Page 6 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Investigation On September 4, 2024, at approximately 0921 hours, the local PSAP received a call for a vehicle fire. The case was generated in the CAD system and the call taker gathered information. While the incident was being dispatched, the call type was updated to a tractor trailer fire with black smoke reported by multiple callers. The companies dispatched on the original call included Battalion 1 and Engine 11, see Table 1. Upon hearing the update from the dispatch center, Chief 1101 called Battalion 1 by mobile phone to suggest the upgrade of the assignment as well as to let Battalion 1 know he was enroute. Battalion 1 said he would be able to see what they were dealing with momentarily and they hung up. Battalion 1 could see a significant smoke column and requested an additional engine company and that Station 1 be re-paged for a tanker response, see Table 2. Chief 8 self-dispatched to the scene to assist. Time (hours) Fireground Operations, Response, and Details • Chief 1101 calls command, “everyone is 10-4 over here.” • Chief 1101 calls Battalion 1, “Are you okay?” 0934 • Chief 1101 updates dispatch, “large explosion, trailer now fully involved, the entire trailer blew apart.” • Chief 1101 going to check on Battalion 1. 0935 • Chief 1101 updates all units, “cannot find Battalion 1.” • Chief 1101 updates, “Battalion 1 is down. Needs helicopter.” 0937 • Chief 1101 assumes command. o Chief 1101 confirmed that Battalion 1 is deceased on scene. o Chief 1101 updates still working on fire attack, contents of trailer still unknown. 0938 • Chief 1101 confirmed contents of the trailer as frozen chicken. 0958 • Chief 1101 transfers command to Engine 13 o Still fighting fire, Personnel accountability report (PAR) okay, 6 personnel working. 1028 • Fire is knocked down. Hitting hot spots. 1528 • Command terminated. • Scene turned over to Sheriff’s Office. • Roadway still closed.

Page 7 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 At approximately 0928 hours, Battalion 1 arrived on-scene and positioned his vehicle past the involved tractor trailer on the North side, blocking the south bound lanes to close the road in that direction. Battalion 1 established command of the incident and sized the incident up as a tractor trailer fire with heavy fire involvement at the rear of the tractor including the tires and diesel fuel tanks. The fire from the rear of the tractor provided direct flame exposure to the underside of the front portion of the trailer. The fire extended up between the rear of the tractor’s cab and the front of the trailer. The cargo and occupant status were unknown upon his arrival and relayed to the dispatch center. Battalion 1 provided an additional update to all responding units that there was off-gassing and fire involvement at the rear of the tractor from an unknown source. He ordered Engine 11 to stop short, South of the incident, deploy an attack line from the front bumper, and utilize foam. Engine 11 stopped approximately 100’ short of the involved tractor trailer and deployed the 100’ 1¾- inch front bumper attack line. Chief 1101, who arrived just before Engine 11, assisted with deployment of the attack line. Engine 11’s officer extended the line with a 50’ section of attack line that was stored in the bumper well. At approximately 0931 hours, Battalion 1 updated dispatch that they had water on the fire and walked away from the engine crew. At approximately 0932 hours, while investigating at the rear of the trailer, Battalion 1 updated all units on-scene that the cargo was still unknown but the rear of the trailer indicated it was refrigerated. At approximately 0933 hours, as Battalion 1 attempted to open the rear trailer doors, the trailer exploded Diagram 2: Approximate locations of all on scene apparatus and attack line at the time of explosion. (Diagram prepared by NIOSH)

Page 8 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 and Chief 1101 called to Battalion 1 on the radio, stating they were okay. Battalion 1 did not acknowledge the message. At approximately 0934 hours, Chief 1101 updated dispatch that the trailer had exploded, was fully involved in fire, and he was going to check on Battalion 1. At approximately 0935 hours, Chief 1101 updated dispatch that Battalion 1 was down and requested a medical helicopter. At approximately 0937 hours, Chief 1101 assumed command of the incident and confirmed Battalion 1 was deceased on-scene. Chief 1101 updated that crews were still working on fire attack, and contents of the trailer were still unknown. At approximately 0938 hours Chief 1101 confirmed with the driver that the cargo was frozen chicken.

Photo 3 and Photo 4. View of fire behavior on passenger side of vehicle (top photo) and view from the fire attack team as Battalion 1 declared water on the fire (bottom photo). (Photos courtesy of fire department)

Page 9 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 At approximately 0958 hours, Chief 1101 transferred command to Engine 13, crews were still fighting fire, PAR was okay, and they had six personnel working. At approximately 1028 hours, command reported the fire was knocked down with crews still extinguishing hot spots. At approximately 1528 hours, command was terminated, the scene was turned over to the Sheriff’s Office, and the roadway would remain closed. Cause of Death According to the State Bureau of Investigation, Department of Forensic Sciences report, an autopsy was performed by the Office of the Medical Examiner. The manner of death was classified as an Photo 5. View of driver side of involved vehicle following extinguishment. (Photo courtesy of fire department) Photo 6. View of passenger side of involved vehicle following extinguishment. (Photo courtesy of fire department)

Page 10 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 accident. It was noted Battalion 1 was in close proximity to a vehicular fire and explosion. The cause of death was listed as blunt force, blast, and thermal injuries of the head and torso. Contributing Factors Occupational injuries and fatalities are often the result of one or more key contributing factors in a larger sequence of events. NIOSH investigators identified the following items as key contributing factors that likely led to this fatality: • SOPs/SOGs • Incident Command • Risk/benefit analysis • Staffing • Hazard recognition, including control zone establishment with unknown hazard. Recommendations Recommendation #1: Fire Service Organizations should develop and maintain SOPs/SOGs with industry best practices. Discussion: SOPs/SOGs establish the basic framework for conducting emergency operations safely and effectively. Some procedures are specifically directed toward safety; others support safe operations by establishing a system of predictable and consistent operations. Coordination, consistency, and standard approaches are all important and valuable components of safe emergency operations [USFA 2018]. In this incident, the IC was the first to arrive on the scene. The IC provided a scene size up report and established command. He positioned his apparatus to provide a safe working area for incoming crews and to establish a Traffic Incident Management Area (TIMA) [NFPA 1091 2024]. The IC then made the effort to perform his 360-degree size up as dictated by the department SOGs. But, due to the smoke drift, only about 270-degrees of the incident was viewed immediately upon his arrival. Once Battalion 1 exited his vehicle he completed his 360-degree size-up, provided his incoming units with updates including something off-gassing at the rear of the tractor, the level of fire involvement, and their initial set of orders on arrival. Following the initiation of fire attack by the first arriving engine and additional personnel who arrived in non-suppression apparatus, the IC continued to operate within the hot zone of the incident in an investigation mode. The purpose of the IC operating within the hot zone performing investigation was to determine the cargo being transported by the motor carrier. NFPA 1550 section 6.2.2, Standard for Emergency Responder Health and Safety, states that fire departments should prepare and maintain written policies and SOPs that document the organization structure, membership, roles and responsibilities, expected functions, emergency operations, and training requirements. This includes:

Page 11 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 • the types of standard evolutions that are expected to be performed • the evolutions that must be performed simultaneously or in sequence for different types of situations • the minimum number of members required to perform each function • the number and types of apparatus and personnel that will be dispatched to different types of incidents in accordance with NFPA 1710 or NFPA 1720. The standard also states these SOPs should specify the procedures that will be employed to initiate and manage operations at the scene of an emergency incident and post-incident control and mitigation of emergency scene contaminants [NFPA 1550 2024]. The department in this incident provided copies of their SOGs for the NIOSH investigators to review. The SOGs contain the requirements for personnel operating on vehicle fires, specifically commercial vehicle fires. The SOG covering incidents where fire is showing requires personnel to be in full protective clothing, including self-contained breathing apparatus. If the vehicle involved is a tractor trailer, the SOG states that command should be in the investigation mode to determine if any hazardous materials are involved. If there are hazardous materials, the type and amount of material should be determined, and command should refer to the appropriate hazardous material SOG. If there are no hazardous materials or if the vehicle involved is a bus, then a search for any victims should be made. Fire service organizations with similar SOGs could benefit from adding language about conducting this type of investigation from outside of the hot zone, preferably uphill and upwind whenever possible. Additionally, ensuring the incident response meets the requirements set forth by the applicable NFPA 1710 or NFPA 1720 standard allows for additional personnel to respond and perform necessary tasks, reducing task oversaturation by individual personnel. The frequency at which SOGs should be reviewed and revised is also specified within NFPA 1550 section 6.2.4. The standards articulate that fire departments should evaluate current trends and research to determine if policies and procedures are appropriate at least annually or following a near miss or catastrophic event [NFPA 1550 2024]. The guidelines under which the department was operating on this incident were effective as of August 2022 and were due for review in August 2023 based upon the recognized best practice set forth in the standard. The department confirmed that they perform SOG reviews and training on a rolling calendar basis and intend to continue doing so. Fire service organizations could benefit from revisiting SOPs/SOGs in their system at least annually as well as evaluating SOPs/SOGs for appropriate risk management strategies. As written, SOPs/SOGs support safe operations and avoid unnecessary risk while allowing for deviation when necessary. Recommendation #2: ICs should establish an IMS on all incidents. Discussion: ICs are responsible for everything that occurs on the fireground. This includes being able to decipher all the information that is presented to them upon their arrival at the incident, make sound decisions, develop an Incident Action Plan (IAP), and employ an IMS on all incidents. Personnel

Page 12 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 cannot control the specific emergency they will face when they arrive to an incident. However, a well- developed command system and training program provides a solid base for personnel to refer to while operating [Pieper 2020]. NFPA 1550 Chapter 21 Command Safety states that the first arriving responder from an emergency services organization has responsibility for the incident, should assume the role of IC for the incident, and maintain the following responsibilities [NFPA 1550 2024]: • Conduct an initial and ongoing situational assessment of the incident • Establish an effective communications plan • Develop incident objectives from the situational assessment and form applicable strategy and tactics • Develop an IAP for the management of the incident • Review, evaluate, and revise the strategy and tactics based on the needs of the incident and deploy resources as needed • Provide for the continuity, transfer, or termination of command • Maintain an effective supervisory span of control. Industry best practices recommend that the IC establish an incident command post (ICP) and remain in the immediate area for the duration of the incident. It is up to the department and IC to determine the most appropriate command post for their incident. Some command post examples include the front lawn or Alpha/address side of a structure, inside the cab of the apparatus with the best vantage point, or the IC’s vehicle. NFPA 1550 standard 21.11 states how the IC should choose and set up their ICP in the initial stages of an incident. Specifically, the standard states how to provide safety for the command staff of an incident based on whether it is a single commander or a team. It would benefit the IC to select an ICP that includes the following considerations [OSFC 2003, NFPA 1550 2024]: • Sound – If the ICP is quiet, there is less of a chance for the IC to miss radio transmissions. • Location – Although ICs want to be able to see as much as they can, it is important to provide a buffer from staging crews and distractions. The ICP should be located in the cold zone of an incident. • Stationary – Moving ICs may find it more difficult to maintain focus as there are more opportunities for them to be distracted. • Better radio – Most apparatus and command vehicles have a more powerful radio than the portable radio issued to personnel. • Protection from the elements – A warm or cool and dry ICP will allow for the IC to concentrate on the incident.

Page 13 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 • Place to write and record – It is imperative that an IC can write or record information related to operations, whether it be crew assignments, which units are in staging, mayday information, or even something as simple as a phone number. • Better intel equipment – An IC wants to know anything and everything about the incident, whether it is hazardous materials, occupancy data, target hazard information, etc. There are times when conditions outside the ICs control prevent the first arriving officer from taking a command role and they must engage in operations on the fireground. This could be an instance of less- than-ideal staffing, an imminent rescue, or a situation that requires the IC to take an investigation style stance. Investigation mode means the IC must determine what conditions are for the incoming crews. This could include determining involvement of hazardous materials, fire conditions, occupancy status, etc. Referring to the previous recommendation, fire department policies should acknowledge this reality while also setting the expectation that ICs should be performing command functions and not operational functions. Exceptions to this are infrequent and not the rule of day-to-day operations. If the IC must perform in an investigative role, it is imperative that they transfer command to the next arriving qualified officer or remove themselves from the command role as soon as possible [Pieper 2020]. The SOGs provided by the fire department in this incident for vehicle fires specifically direct the IC to be in an investigation mode and not in a command position. This forces the IC to be engaged in some level of operations inside the hot zone versus being in a controlled and safer environment with a good vantage point of ongoing operations. The Incident Command SOG for the involved department also does not specify the methodology the command officers should follow on arrival other than establishing command, giving the size-up report, mode of operation, and delegating assignments. Specifying the need for a relatively stationary command setup and the necessity of employing an accountability system and IAP for the incident is important. The Incident Command Structure (ICS) is a standardized approach to the command, control, and coordination of on-scene incident management that provides a common hierarchy within which personnel from multiple organizations can be effective. ICS specifies an organizational structure for incident management that integrates and coordinates a combination of procedures, personnel, equipment, facilities, and communications. Deploying the ICS on every incident can help hone and maintain skills needed to coordinate efforts effectively, is applicable across disciplines, and enables incident managers from different organizations to work together [FEMA 2017]. Recommendation #3: ICs should integrate risk assessment as a risk management practice into their fireground strategy and tactic SOGs/SOPs. Discussion: In this incident, the IC followed the department SOGs that were specifically developed for a tractor trailer fire. He established command, provided a size-up, gave the initial company orders, and requested additional resources. Then, based on the direction of the SOGs, the IC entered the hot zone multiple times to investigate what the tractor trailer was hauling. He stated over the radio that the trailer contained unknown content on more than one occasion. He stated the trailer was refrigerated.

Page 14 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Applying different risk management techniques in coordination with following the SOGs may have limited the number of personnel operating in the hot zone of the incident. The concepts of operational risk management have always been important considerations for emergency response organizations [USFA 2018]. Some risks are unavoidable and are accepted as part of the duty of a firefighter or line officer. The nature of the activities of emergency responders exposes them to a high level of inherent risk. Conditions or circumstances can exist in which the proper response is to not take actions that could place the lives of responders in danger [USFA 2018]. Fire service personnel can reduce the risk of injury or death while operating on the fireground if they are able to identify the nature of the threat, determine the risk potential, and incorporate risk management practices into the operational plan. A high level of risk management is involved in directing emergency operations and in regulating the responders’ exposure to emerging incident hazards. Emergency incident risk management expands the standard approach to directing and conducting emergency operations by incorporating risk evaluation and the assessment of optional approaches to the problems at hand. NFPA 1550 standard 18.3.2.1 states that the IC should integrate risk management practices in the regular functions of incident command and at emergency operations the IC should evaluate the risk to members operating at the scene [NFPA 1550 2024]. This means the IC should determine the life safety profile of the incident and apply the most appropriate level of risk to first responders consistent with the following principles shown in Chapter 10.4 of NFPA 1550 [2024]: • Activities that present a significant risk to the safety of members should be limited to situations where there is a potential to save endangered lives. • Activities that are routinely employed to protect property should be recognized as inherent risks to the safety of members, and actions should be taken to reduce or avoid these risks. • No risk to the safety of members should be acceptable when there is no possibility to save lives or property. • In situations where the risk to fire department members is excessive, activities should be limited to defensive operations. There are times in which a tactical command style can be beneficial. Chief Curt Isakson explains that the command style known as “tactical command” is out of a command vehicle, equipped with all personal protective equipment, and involved in a more up close and personal size-up which assists in not only managing the incident but also in getting personally involved in the mitigation of the incident. He also expresses his perspective of the importance of choosing the command style or mode to fit the incident and that there is not always a one size fits all approach. The command location and style must fit the incident to provide the best outcome for life-safety of both civilians and firefighters, while still placing property as a second priority [Isakson 2021]. In this incident, the contents of the trailer were unknown other than it was refrigerated. The bulk of the fire was located between the rear of the tractor and the trailer bulkhead. Bystanders told the IC that the

Page 15 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 driver was out of the vehicle and had fled the scene. The department SOG for tactical priorities listed the hierarchy of the incident as follows:

1. Life safety
2. Incident stabilization
3. Property conservation.

Knowing the driver of the tractor trailer departed the vehicle partially eliminates the life safety concern. Without knowing if it was a driver team or a single driver, the priority cannot totally be ignored without contacting the driver who fled the scene. Battalion 1 gave orders to the first arriving engine company to stop short of the incident and to stretch an attack line for fire attack. This began the process of addressing the priority of incident stabilization. Risk assessment, as part of an overall risk management process, is ongoing throughout the entirety of an incident starting at the arrival of the first emergency responder until the incident is deemed over and command is terminated. All actions and tactics should have any new or changed information applied to them to ensure personnel safety and operational effectiveness. The role of the IC can be very overwhelming and difficult when the incident is complex and multifaceted. The IC has overall authority for the management of the incident and for some it can be a daunting task. The ICS provides a framework that allows it to be adapted and scaled based upon the incident [NFPA 1550 2024]. Recommendation #4: Fire service organizations should follow the appropriate standards for organization and deployment of fire suppression operations, emergency medical operations, and special operations. Discussion: The department involved in this incident became a combination fire department in 2019. NFPA 1720 section 3.3.16.1 defines a combination fire department as one having emergency service personnel comprising less than 85 percent majority of either volunteer or career membership. NFPA 1720 section 3.3.16.2 defines a volunteer fire department as a fire department having volunteer emergency service personnel comprising 85 percent or greater of its department membership [NFPA 1720 2020]. Although it is not clear which standard they were following, the department in this incident should have been utilizing the information in NFPA 1710 due to the percentage of volunteer membership being approximately 5.6 percent of the total membership. Chapter 5 of NFPA 1710 provides minimum standards for staffing of various fire suppression companies. The department involved in this incident is responsible for primarily staffing engine companies. Engine companies are fire companies whose primary functions are to pump and deliver water as well as perform basic firefighting and search and rescue. Further, these companies should be staffed with a minimum of four on-duty members. However, the standard recommends that these companies be staffed with a minimum of five on-duty members in first-due response zones that experience a high number of incidents, geographical restrictions, geographical isolation, or urban areas, as identified by the authority having jurisdiction [NFPA 1710 2020].

Page 16 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 If the department were operating under the NFPA 1720 standard as a combination department, the following information should be utilized with the implementation of more volunteer membership. Based on the metrics provided by the department, the county could be considered a rural and remote area. Specifically, there is less than 50 people per square mile and a travel distance from the second due engine is greater than eight miles [USCB 2020]. The data preceding this shows that the staffing should be six personnel within 14 minutes of dispatch at least 80% of the time. Table 3 displays the staffing and time standards for volunteer and combination fire departments as specified in NFPA 1720. Table 3: Staffing and Response Time (Adapted from NFPA 1720) Demand Zone Demographics Minimum Staff to Respond Response Time (Minutes) Meets Objective (%) Urban area

1,000 people/mi2

(2.6 km2) 15 9 90 Suburban area 500 – 1,000 people/mi2 (2.6 km2) 10 10 80 Rural area people/mi2 (2.6 km2) 6 14 80 Remote area Travel distance ≥ 8 mi (12.87 km) 4 Directly dependent on travel distance 90 Special risks Determined by AHJ Determined by AHJ based on risk Determined by AHJ 90

Page 17 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Recommendation #5: ICs should establish control zones on incidents where IDLH environments exist. Discussion: NFPA 1550, chapter 18.3.1.4, states the IC should have the responsibilities and duties of all unassigned incident command system positions. In organizations where personnel are not plentiful and arriving early at incidents, this creates a situation in which task oversaturation for the IC can happen quickly. NFPA 1550 chapter 5 states the Incident Safety Officer (ISO) role is a position within the ICS at an incident or planned event that is charged with ensuring action is taken to mitigate any immediate life safety threats. Within the scope of the ISO position is the necessity to determine hazardous incident conditions and advise the IC to establish or modify control zones at any given incident. These control zones must be communicated to on-scene personnel and entry into the hazardous area must be controlled by the ISO as specified in chapter 5 of NFPA 1550. These actions require the ISO to possess comprehensive knowledge of hazardous conditions, operations, departmental SOPs/SOGs, control zones protocols, and the IMS. An ISO also needs to be able to evaluate the effect of proximity for incident hazards so that risk to members will be limited to emergency responders assigned tasks to mitigate the incident (i.e., in the hot zone) [NFPA 1550 2024]. In this incident, the IC applied various techniques that fell under responsibilities of an ISO. He ordered Engine 11 to stop short of the affected vehicle, placing it outside of the hazard zone. He applied the process outlined in the department SOGs and attempted to locate the driver as well as determine if the cargo was considered a hazardous material. The IC also communicated all other hazards noticed during his ongoing investigation of the incident including that there was something off-gassing at the rear of the tractor and that the trailer was refrigerated. The International Fire Service Training Association (IFSTA), Essentials of Fire Fighting 8th Edition: Firefighter 1, states that control zones can have a variety of labels depending on operations being conducted within those zones. The most common terms used to describe these areas are hot, warm, and cold zones. The hot zone is the area closest to the hazard such as a fire or hazardous material. This is where trained personnel work to mitigate the problem. Only personnel directly tasked in mitigating the problem are allowed to enter this zone, which limits crowds and confusion at the most critical area of the scene. Diagram 3: Incident Scene Control Zones 1. (Courtesy of International Fire Service Training Association)

Page 18 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 These personnel wear protective clothing and equipment appropriate for the incident’s hazards [IFSTA 2024]. The warm zone is immediately outside the hot zone and serves as the transition area. Personnel in the warm zone directly support personnel working in the hot zone by providing emergency lighting, fire protection, or other support functions. These personnel wear full protective clothing and equipment and are ready to enter the hot zone if needed [IFSTA 2024]. The cold zone surrounds the warm and hot zones. The cold zone may include the ICP and locations for the rapid intervention crews, public information officer, rehabilitation area, and staging areas for personnel and portable equipment. The outer boundary of the cold zone serves as the control line for the general public (i.e., crowd-control line). Personnel working in the cold zone should wear protective clothing and equipment appropriate for the tasks they must perform and the hazards they may encounter [IFSTA 2024]. Photo 7 and Photo 8. Driver side of trailer just prior to explosion (left photo) and change of smoke behavior and expansion of container as explosion is occurring (right photo). (Photos courtesy of fire department)

Page 19 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Recommendation #6: Fire Service training should provide insight and education around commonly unrecognized pressure vessels in the form of intermodal freight trailers and mixed cargo trailers. Discussion: The 2024 edition of the ERG indicates there are several types of road trailers in use in the United States and includes a road trailer identification chart for emergency responders. Several of these trailers call out either design pressures or maximum allowable working pressures (MAWP). The road trailer involved in this incident was a 111 intermodal freight container. The ERG lists distinctive characteristics and uses for this type of trailer [PHMSA 2024] Including: • For general freight that carry bulk or non-bulk packages • May transport hazardous materials/dangerous goods in small packages (e.g., bags, boxes, drums) or IBCs, sometimes referred to as “totes” • Cargo door(s) in the rear. What is not provided through the ERG is the MAWP for several types of trailers including the 111 intermodal freight container. The other trailer types identified with no noted MAWP include the 117 compressed gas/tube trailer, 134 dry bulk cargo trailer, and 137 vacuum tanker. The fire service is not taught to recognize type 111 trailers as a pressure vessel. When impinged by fire, these trailers can become pressurized. The same fire dynamics principles that apply to structure fires, apply here as well. As the temperatures inside of a closed compartment increase, the pressure inside the compartment also increases. Without a sufficient pressure relief vent(s), the pressure continues to build. The Type 111 trailers, when configured for refrigerated applications, have void spaces between the inner and outer floor, walls, and roof that are filled with combustible foam plastic insulation. This means that even an empty trailer has a significant fuel load. Additionally on this unit, there was an integrated refrigeration unit that consisted of pressurized cooling lines and a compressor system much like in a typical refrigerated trailer. The various containers integrated into the refrigeration system could suffer a catastrophic failure as the refrigerant becomes heated when exposed to fire. This could eventually lead to a boiling liquid expanding vapor explosion (BLEVE). In this incident, the refrigeration overpressure appeared to ventilate and maintain safer pressures. This incident is not believed to be the result of a BLEVE. In the Principles of Heavy Vehicle Extrication by IFSTA [2023], personnel are taught the five-step process for heavy vehicle extrication which includes the following steps:

1. Identify vehicle and cargo weight hazards (identify the load)
2. Stop the crush
3. Lower the smaller vehicle
4. Lift the load
5. Separate and extricate.

In this incident, the vehicle was upright on its wheels. The first applicable step of the process upon arrival is identifying the load. Part of identifying the load is contacting the driver of the vehicle who

Page 20 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 should know what they are hauling. In this incident, Battalion 1 was unable to locate the driver of the vehicle because they left the immediate area. The driver was located later into the incident by Chief 1101 and the cargo information was relayed after the explosion occurred. The trailer was not placarded to indicate any hazardous materials and Battalion 1 approached the rear of the trailer to further investigate. He eventually tried to open the rear cargo doors of the trailer at the same time the trailer suffered a catastrophic failure. The IFSTA Principles of Heavy Vehicle Extrication states that if a cargo area of a tractor trailer or box truck is involved in fire, consider calling for additional suppression resources as these fires require extensive overhaul operations [IFSTA 2023]. The fire service at large is aware of the potential surrounding a catastrophic failure of tank trailers or fuel tanks such as a BLEVE or a burn through that contributes to significant fire growth and potential. However, the same care is not taught for type 111 trailer responses, which may support more complacent responses. A smoke explosion, like this case, can happen in any fire in a compartment. This could be a structure fire, or something different such as the trailer in this instance. The key indicators as the compartment is filled with products of combustion and oxygen is consumed or limited all it takes is the introduction of oxygen to the mixture to support a violent reaction. The IFSTA Principles of Heavy Vehicle Extrication teaches the explosion hazards of undeployed air bags, energy- absorbing bumper struts, support struts (such as in hoods, trunks, and hatchbacks), suspension struts, shock absorbers, fuel containers, seatbelt pretensioners, and vehicle tires. The fire service should be taught that even though explosion hazards are less prevalent than fire hazards, there is a risk associated with explosion hazards at heavy vehicle incidents [IFSTA 2023]. The dangers of the unrecognized pressure vessel can be deadly, as seen in this incident. One pressure vessel that is present on every vehicle fire are the tires of the vehicle. Fire service personnel are educated of the potential dangers when these rupture but is not emphasized to the level it could be. Commercial vehicle tires contain significantly more pressure than passenger car tires before they are on fire. Tires rupturing without being impinged in fire have incredible force potential and fires amplify the force potential. Another commonly overlooked or unrecognized pressure vessel are the various air bags in commercial vehicles. These are in the suspension, the cab supports, and even the seats inside the passenger compartments. Fire service personnel must be educated to identify these other less known pressure vessels including the cargo compartments of the vehicle, especially when they are the refrigerated style. Pressure can build in these containers unbeknownst to fire service personnel operating in their vicinity with potentially deadly outcomes. A summary of FSRI’s fire analysis findings can be found in Appendix Two. References FEMA [2017]. National incident management system. Washington, D.C.: U.S. Department of Homeland Security, Federal Emergency Management Agency.

IFSTA [2023]. Principles of heavy vehicle extrication. Stillwater, OK: International Fire Service Training Association.

Page 21 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Isakson C [2021]. Command & control success: Command post location. Retrieved October 2024, from Firehouse: https://www.firehouse.com/leadership/incident-command/article/21243635/command- control-success-command-post-location

ISFSTA [2024]. Essentials of fire fighting. 8th ed. Stillwater, OK: International Fire Service Training Association.

NFPA [2020]. NFPA 1710, Standard for the organization and deployment of fire suppression operations, emergency medical operations, and special operations to the public by career fire departments. Quincy, MA: National Fire Protection Association.

NFPA [2020]. NFPA 1720, Standard for the organization and deployment of fire suppression operations, emergency medical operations, and special operations to the public by volunteer fire departments. Quincy, MA: National Fire Protection Association.

NFPA [2024]. NFPA 1061, Standard for traffic incident management personnel professional qualifications. 12. Quincy, MA: National Fire Protection Association.

NFPA [2024]. NFPA 1550, Standard for emergency responder health and safety. Quincy, MA: National Fire Protection Association.

OSFC [2023]. Officers command: 8 functions of command. Lewistown, PA: Pennsylvania Office of the State Fire Commissioner.

PHMSA [2024]. Emergency Response Guide. Washington, D.C.: U.S. Department of Transportation Pipeline and Hazardous Materials Safety Administration.

Pieper S [2020]. 4 essential elements for fire incident command policies and procedures. Lexipol.

USCB [2020]. QuickFacts Greene County, Georgia. Washington, D.C.: U.S. Department of Commerce, U.S. Census Bureau.

USFA [2018]. Risk management practices. Washington, D.C.: U.S. Department of Homeland Security, Federal Emergency Management Agency, U. S. Fire Administration. Retrieved October 2024, from

Weather Underground [2024]. Augusta, GA weather history. Weather Underground. Investigator Information This incident was investigated by Louis (Rick) Lago (former), Investigator, and Stephen J. Ringer, Investigator, both with the Fire Fighter Fatality Investigation and Prevention Program, Surveillance and Field Investigations Branch, Division of Safety Research, NIOSH located in Morgantown, WV. The report was written by Stephen J. Ringer. Dan Madrzykowski and Keith Stakes from the Fire Safety Research Institute, part of the UL Research Institutes, provided an expert review of the

Page 22 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 expert review of the investigation report. The NFPA Emergency Response & Responder Safety Division also provided a technical review. An expert technical review was provided by Thomas Cook, CPSE credentialed Chief Fire Officer and Chief Training Officer. Disclaimer The information in this report is based upon dispatch records, audio recordings, video recordings, witness statements, and other information that was made available to the National Institute for Occupational Safety and Health (NIOSH). Information gathered from witnesses may be affected by recall bias. The facts, contributing factors, and recommendations contained in this report are based on the totality of the information gathered during the investigation process. This report was prepared after the event occurred, includes information from appropriate subject matter experts, and is not intended to place blame on those involved in the incident. Mention of any company or product does not constitute endorsement by NIOSH, Centers for Disease Control and Prevention (CDC). In addition, citations to websites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or products. Furthermore, NIOSH is not responsible for the content of these websites. All web addresses reference in this document were accessible as of the publication date.

Page 23 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Appendix One Trailer Construction Materials Findings (Courtesy of UL FSRI) At the time of the examination, the heavily damaged tractor trailer had been moved to a storage area and was basically in 5 pieces (areas). The remains of refrigerated trailer (intermodal shipping container), the trailer frame that the intermodal container sat on, remains of the tractor, the diesel generator that powered the refrigeration system, and cargo that is in 3 roll off dumpsters. The intermodal portion of the refrigerated trailer had been reduced to the floor, front wall, and some sections of the sides, Figure 1. The floor, walls and roof of the refrigerated trailer had void spaces between the exterior metal of the trailer and the interior metal lining of the trailer. By design, the void spaces were filled with a blown in foam that sets up as a rigid foam plastic. This provides the insulation for the refrigerated trailer. This foam could be a significant source of fuel. It was noted that foam insulation had been consumed completely or showed significant mass loss in areas of the trailer that had been exposed to the exterior fire sources.

Photo 9 and Photo 10. Photographs of the remains of the refrigerated trailer. Rear view on the left and front view on the right. Most of the foam insulation had been burned out of the front portion of the trailer. (Photos courtesy of UL FSRI)

Page 24 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05

Photo 11 and Photo 12. Some of the remaining foam insulation. Partially charred foam from a wall panel (left photo) and partially charred foam in the floor section of the trailer (right photo). (Photos courtesy of UL FSRI) Photo 13 and Photo 14. The foam sample from the trailer flooring section that was provided for testing. Bottom of the sample (left photo) and top of the sample (right photo). (Photos courtesy of UL FSRI)

Page 25 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Foam Analysis The foam samples used for testing were cut out of the foam from the truck, so that no char or residue were part of the test samples. A pyrolysis gas chromatography with mass spectrometry (Py/GC-MS) analysis of the foam was conducted. This type of analysis is used to evaluate the chemical composition of pyrolyzates from a wide range of organic materials, providing information to identify the test material. Organic materials such as foam plastics that can be pyrolyzed can be analyzed by GC-MS. The mass spectrum is used to identify the compounds in the sample and compare them against a library of known materials, to identify the best match for the pyrolyzed sample. Samples of the foam were prepared as a powder to ensure the sample was homogeneous for Py/GC- MS and microscale combustion calorimetry experiments. The temperature range for the pyrolysis of the sample was 300-900 °C with a temperature ramp of 5 °C/s. Pyrolyzates were directed into the GC- MS column and identified by the MS. Peaks consistent with polyisocyanurate foam were observed in the chromatogram. The polyisocyanurate foam was also subjected to a series of tests to determine its potential heat release rate capability. Powdered foam samples were tested with the microscale combustion calorimeter (MCC) in general accordance with ASTM D 7309-21. The samples were heated from room temperature to 700°C at a heating rate of 1°C/s. The peak specific heat release rate in the MCC was observed at 371 °C and was measured as 293 W/g and the specific heat of combustion was 25.4 MJ/kg. Foam samples, approximately 100 mm square and 25 mm thick, were tested in the cone calorimeter in general accordance with ASTM E 1354-23. Cone samples were tested at radiant exposures of 25 kW/m2, 50 kW/m2, and 75 kW/m2. Two of the cone calorimeter results for the 75 kW/m2 are presented here, the averaged peak heat release rate of 697 kW/m2 and the averaged heat of combustion of 20.2 MJ/kg. The higher heat flux was chosen to represent the heat flux closest to flame impingent from the outside of the truck. The measure protocols and the full dataset created can be viewed at https://materials.fsri.org/materialdetail/polyisocyanurate-spray-foam Compared to materials like wood, https://materials.fsri.org/materialdetail/wood-stud, under similar test conditions, the foam insulation has an average peak heat release rate twice as high as wood, and an average heat of combustion, 30% higher than wood. The fire service take away, even if the refrigerated trailer were empty, it still has a large fuel load, with the potential for significant fire or explosion hazard. The material analysis of the foam shows that its pyrolyzates and combustion products could have been a source of fuel for the smoke explosion that occurred within the trailer.

Page 26 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05

Summary In addition to fuel sources such as tires and diesel fuel, the fire service needs to recognize other fuels that are used in the construction of the tractors and trailers, including fiberglass and foam insulation. Furthermore, based on the fire analysis of the incident there are several considerations for the fire service regarding refrigerated trailer fires due to the insulating foam, the trailer contains a significant amount of fuel (hazard), even if it is empty. Photo 15 and Photo 16. One of the foam samples prepared for testing in the cone calorimeter (left photo) and amount of flame generated by the foam sample during a cone calorimeter test (right photo). (Photos courtesy of UL FSRI)

Page 27 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05 Appendix Two Fire Analysis Findings (Courtesy of UL FSRI) Still images taken from a video of the incident show notable changes to smoke and fire coming from the trailer. The smoke and flames prior to suppression seem to cycle, build, stop, and start again, see Photos 17 through 19. After initial knockdown, the smoke from the trailer again seems to cycle or pulse until the explosion, Figure 8. After the explosion the ventilation limited/fuel rich environment is the truck resulted in a vent point ignition, followed by an increased amount of fire, Figures 9 and 10. The images captured throughout the incident captured pulsating fire behavior indicative of the onset of a smoke explosion.

Photo 17 and Photo 18. The photograph on the left shows the fire at approximately 0928, the fire has been burning for almost eight minutes. The images are approximately two seconds apart. The photograph on the right shows a noticeable amount of dense black smoke from the underside of the trailer. It seems to have enough pressure to flow against the direction of the wind, then it stops. (Photos courtesy of UL FSRI)

Page 28 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05

Photo 19 and Photo 20. This set of images was recorded between 0929 and 0930. The interval between the image on the left and the image on the right is about five seconds. These images show flames jetting upwind from a localized point under the trailer. This may be an indication of pressurized fire gases exiting an opening in the trailer. (Photos courtesy of UL FSRI) Photo 21 and Photo 22. Between 0930 and 0931, just prior to water on the fire, the image on the left shows that the flame jetting had stopped and then started again 35 seconds later. Seven seconds later the flame jet stopped again. (Photos courtesy of UL FSRI)

Page 29 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05

Photo 23, Photo 24, and Photo 25. These images are one second apart between 0932 and 0933. Dense black smoke flow from under the trailer once again is pushing out. This was seconds before the explosion. (Photos courtesy of UL FSRI) Photo 26 and Photo 27. These images provide a sense of the amount of energy released by the explosion in terms of a blast, projectiles, and thermal radiation. The image on the right shows the latched door that was blown off the trailer. (Photos courtesy of UL FSRI)

Page 30 Career Battalion Chief Dies from Explosion while Fighting Tractor Trailer Fire – Georgia Report # F2024-05

Summary The fire analysis showed it is important to recognize that ventilation limited fire conditions can occur in compartments on wheels, such as this refrigerated trailer, just as they occur in compartments in fixed structures. Fires burning inefficiently in a ventilation limited compartment can generate significant amounts of gaseous fuel. When a portion of the compartment contains an explosive mixture of oxygen and gaseous fuel, all that is needed to generate a smoke explosion is a source of heat. Based on the fire analysis of the incident there are several considerations for the fire service regarding potential fire behavior of refrigerated trailers: 1) Compartment fire dynamics apply to compartments that are part of a tractor trailer such as: the closed refrigerator trailer, tanks, and tires. 2) Watch for pulsating fire behavior indicative of the potential for a smoke explosion and maintain a safe distance.

Photo 28 and Photo 29. After the explosion, the inside of the trailer was still a fuel rich environment providing the conditions for a “vent point” ignition, as shown on the left. As additional air was entrained into the smoke the area of flaming combustion increased. The source of the fuel for the fire is mainly the trailer and not the cargo. (Photos courtesy of UL FSRI)

FD06033 ｜ 2024_美国_NIOSH_face202406.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2024 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Page i

Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Executive Summary On November 6, 2024, a 42-year- old career firefighter died, and a 38-year-old firefighter was injured while performing rapid intervention team (RIT) tasks at a vacant warehouse fire. The two firefighters were assigned to the RIT team on their arrival at the scene. At approximately 22:50 hours, the office of emergency communications (OEC) received the first call reporting a three-story warehouse on fire with an approximate location. Several additional calls reporting a warehouse on fire were received which gave a more specific address. The OEC dispatched the first alarm assignment for a warehouse fire at approximately 22:52 hours which included District Chief 20 (DC20), District Chief 8 (DC08), Engine 18 (E18), Engine 20 (E20), Engine 40 (E40), Engine 23 (E23), Tower 18 (T18), Ladder 20 (L20), and Safety Officer 30 (SF30). At approximately 22:57 hours, E18 updated responding units of a working fire. E18 arrived on-scene at approximately 22:58 hours to a “large warehouse fire showing from the roof, heavy fire” and assigned their side of the building as the Alpha side of the structure. E20 approached from the cross street and pulled past the fire building leaving room for L20. DC20 arrived and positioned across the street from the Alpha/Delta corner in a parking lot and established himself as the incident commander (IC). He reported “heavy fire from a one-story warehouse.” E18 declared a defensive fire. The OEC broadcasted an emergency tone and rebroadcasted the defensive fire message across the channel. At approximately 23:00 hours, L20A provided updated observations to IC, “Command, TIC (thermal imaging camera) reading through open doors is about 1,000 degrees at the roof, heavy smoke, fire, and F2024-06 March 2025 Photo 1: Delta side of structure post-collapse. (Courtesy of the fire department)

Page ii Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Report # F2024-06 some of the roof is starting to collapse.” IC acknowledged the report and requested both aerial devices be set up for elevated master stream operations. IC then updated all units that accountability was in place. E40 arrived and was assigned water supply for L20. E23 arrived and IC assigned them as rapid intervention team (RIT). At approximately 23:02 hours, L20A provided a conditions, actions, needs (CAN) report stating, “Collapse of the roof, heavy fire, exposure on the Charlie/Delta side going to have to get inside, already starting to smoke up. Trying to get water on it now. L20 in position to hit it with the pipe, no further needs.” IC acknowledged the update and stated over the channel, “District 20 to all companies, do not go interior.” E23B (deceased firefighter) and E23C (injured firefighter) gathered the RIT equipment from the IC post and staged it on the Alpha/Delta corner of the fire building. IC broadcasted over the channel, “Command to all companies on Delta division, let’s stay clear of that wall, it looks like it is going to collapse outwards.” E23’s crew, as part of their RIT duties, went to the Delta side man door and began attempting to force the burglar bars and doors open to “soften” the structure. At approximately 23:05 hours, IC requested a second alarm be dispatched. As this was called, E20 updated IC there was a collapse on the Delta side. Less than a minute later, E23A declared a Mayday. IC acknowledged the Mayday. OEC also notified IC of the Mayday and inquired if IC would like the third alarm assignment. IC requested the third alarm. L20 updated IC that one person had been removed, and IC requested a count on how many personnel were missing. L20 responded possibly one more and they were going to attempt a PAR. A moment later L20 updated IC there were two firefighters missing, they were actively working to get to them and requested operations go to a different radio channel. At approximately 23:09 hours, R42 arrived and were assigned to the Delta side collapse for extrication of the trapped firefighters. IC updated OEC that the second firefighter is being extricated. At approximately 23:13 hours, L20 updated IC that all victims were out, and personnel were clearing the collapse zone. IC requested all personnel evacuate the area of the Delta side and that OEC cancel the third alarm. L20 exited the collapse zone with the second firefighter, and they needed advanced life support (ALS) to meet them at the front of L20. OEC notified IC that L20A was showing an alarm and asked for confirmation that all personnel were accounted for. IC confirmed all personnel were accounted for and requested all personnel stay away from the building and out of the collapse zone. E23C (injured firefighter) was transported to the hospital by ambulance 23. E23B (deceased firefighter) was transported to the hospital by medic 18 and was later pronounced deceased at the hospital. Contributing Factors • Pre-incident planning • Standard operating procedures/guidelines (SOPs/SOGs) • Strategy and tactics • Risk versus benefit analysis • Safety officer presence • Change of occupancy • Vacant/dangerous building inspections

Page iii Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Report # F2024-06 Key Recommendations Fire service organizations should: • Develop pre-incident plans for structures within their first due response area, specifically dangerous or vacant structures. • Develop and maintain SOPs/SOGs. • Ensure ICs match their tactical strategy with the incident at hand. • Train all personnel to employ risk management principles when operating on the incident scene. • Strategically deploy safety officers. Additionally, governing municipalities (federal, state, regional/county, and local) should: • Complete change of occupancy inspections per applicable fire and life safety codes. • Communicate across departmental boundaries.

For further information, visit the program at www.cdc.gov/niosh/firefighters/fffipp/ or call 1-800-CDC-INFO (1-800-232-4636). The National Institute for Occupational Safety and Health (NIOSH) initiated the Fire Fighter Fatality Investigation and Prevention Program to examine deaths of firefighters in the line of duty so that fire departments, firefighters, fire service organizations, safety experts and researchers could learn from these incidents. The primary goal of these investigations is for NIOSH to make recommendations to prevent similar occurrences. These NIOSH investigations are intended to reduce or prevent future firefighter deaths and are completely separate from the rulemaking, enforcement, and inspection activities of any other federal or state agency. Under its program, NIOSH investigators interview persons with knowledge of the incident and review available records to develop a description of the conditions and circumstances leading to the deaths in order to provide a context for the agency’s recommendations. The NIOSH summary of these conditions and circumstances in its reports is not intended as a legal statement of facts. This summary, as well as the conclusions and recommendations made by NIOSH, should not be used for the purpose of litigation or the adjudication of any claim.

Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas

Page 1 F2024-06 March 2025 Introduction On November 6, 2024, a 42-year-old firefighter died, and a 38-year-old firefighter was injured while engaged in RIT tasks at a vacant warehouse fire. On November 8, 2024, the fire department notified the National Institute for Occupational Safety and Health (NIOSH) of this incident and requested an investigation. From November 19 – November 24, 2024, two investigators representing the NIOSH Fire Fighter Fatality Investigation and Prevention Program (FFFIPP) conducted interviews with fire officers and firefighters who responded to the incident. Site visits, photos taken at the scene, computer aided dispatch (CAD) notes, radio traffic audio, and surveillance camera footage were obtained from the fire department and reviewed. NIOSH investigators also reviewed the training records of specific personnel involved in the incident and reviewed the department’s SOGs and professional development model. Fire Department The career fire department in this incident consists of approximately 4,000 uniformed personnel and approximately 100 civilian personnel. Approximately 3,800 of those personnel are assigned to the emergency response division that provides fire protection and emergency medical services (EMS). The department’s emergency response division works a 72-day duty cycle. Members work a total of 480 hours that are made up of 18 regular 24-hour shifts and two 24-hour debit day shifts. The emergency response division consists of four shifts following this schedule. The shifts, not including debit shifts, typically appear on a calendar as follows: 24 hours on duty, 24 hours off duty, 24 hours on duty, 120 hours off duty. There are 93 fire stations located throughout the city that serve a population of approximately 2.3 million people across 637 square miles. The fire department divides the city into quadrants: northeast, northwest, southwest, and southeast. These are further divided into 21 districts. The fire department is certified as an Insurance Services Office class 1 department and is accredited through the Commission on Fire Accreditation International as of 2021. It provides aircraft rescue firefighting (ARFF) for two airports within the city. The fire department participates in a county wide mutual aid agreement for fire protection and suppression that was drafted in June 2017 and automatically renews every year. This agreement is across all departments within the county whether they are career, combination, or volunteer and it includes more than 71 different entities. The fire department responds to over 395,000 calls for service annually. The emergency response division is made up of two deputy chiefs, three safety officers, three cascade units, one command van, one rehab unit, 21 district chiefs, 88 engine companies, 33 ladder companies, five tower

Page 2

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas companies, 57 ambulances with an additional five during peak times, 41 medic units, five squads, seven ambulance supervisors, two senior ambulance supervisors, one EMS district chief, three rescue companies, four hazardous materials companies, and 16 ARFF units. The minimum daily staffing for the department is 849 personnel across all companies. The deputy chiefs, safety officers, cascade units, command van, ambulance supervisors, senior ambulance supervisors, EMS district chief and rehab unit are each staffed with a single person. The district chiefs, ambulances, medic units, and squads are each staffed with two personnel. Engine companies, ladder companies, tower companies, rescue companies, and hazardous materials companies are each staffed with four personnel. Ambulance units are classified as basic life support (BLS) while squads and medic units are classified as ALS. The fire department provides EMS service up to and including ALS-transport. Training and Experience The fire department hires through a civil service process. The minimum requirements to apply are that the applicant must be between 18 and no more than 36 years old when they take the oath of office, have completed 15 accredited college hours or two years of full-time active-duty military service with an honorable discharge, and have two or less moving violations in the last 36 months. Upon accepting a conditional job offer, the applicant must attend the academy and complete between eight and ten months of daily rigorous physical training and classroom instruction. Upon graduation, applicants have certifications in the Texas Commission on Fire Protection fire basic, Texas Department of State Health Services emergency medical technician (EMT) basic, and the National Registry of Emergency Medical Technicians. Once the candidate successfully graduates from the academy, they are assigned to their fire station. Once assigned to their fire station and shift, all personnel complete a minimum of 24 hours of on-the-job training and continuing education per month. Personnel within the department have a variety of career pathways they may choose to pursue once they have reached the rank of engineer operator (EO). Pursuing a career pathway involves a process of either promotion or acceptance into one of the “closed shops” within the fire department. Examples of the career pathways personnel may pursue include the OEC, inspection division, or arson division. The deceased firefighter (E23B) had eight and a half years of experience with the department. The deceased firefighter (E23B) participated in a variety of monthly company level trainings. Fire department training records show participation in training such as hazardous materials, RIT skills, building construction, multi- company drills, equipment and apparatus familiarization, EMS continuing education, area familiarization, and tactical decision making on offensive and defensive fires. Per department records, E23B firefighter routinely logged 30 hours of training per month. The injured firefighter (E23C) had approximately six years of experience with the department. The injured firefighter (E23C) participated in a variety of monthly company level trainings. Fire department training records show participation in training such as Mayday training, building construction, operational readiness,

Page 3

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas EMS continuing education, and truck company tactics. Per department records, E23C firefighter routinely logged between 24 and 30 hours of training per month. Apparatus, Personnel, and Accountability Equipment The minimum staffing for the career fire department in this incident is two personnel assigned to a district chief vehicle (district chief and incident command technician (ICT)) and four personnel assigned to engines, truck, and rescue companies. The city often must “brown out” companies due to staffing issues while maintaining a sufficient standard of cover for fire protection and EMS services. Staffing vacancies are filled with personnel working their debit day, picking up optional overtime, or being mandated to work overtime. Personnel are assigned to a specific position and apparatus by the staffing office. OEC calls are received by civilian call takers. The OEC dispatchers for the fire side of the center are sworn personnel at the rank of communications captain or above with a minimum of five years within the department. The OEC dispatchers utilize a digital trunking 700-megahertz encrypted system for all radio communications. The fire department in this incident utilized a pre-established run card assignment based on information provided by the civilian call taker. A first alarm assignment for a warehouse fire included two district chief units, four engines, two truck companies, a medic unit, and a safety officer. The three rescue companies spread throughout the city were recently added to working fires across the city. However, it was noted by the fire department that this incident type was missed in the change. The captain assigned to R42 self- dispatched their unit to this fire upon hearing E18 call the working fire. Tables 1 and 2 below provide staffing levels and location when the collapse took place. Table 1: First Alarm units at the time of collapse Resource Designation Staffing Level (#) Status and Location District Chief 20 (DC20) 2 On-scene with Command District Chief 8 (DC08) 2 On-scene with Alpha side operations Engine 18 (E18) 4 On-scene, side Alpha Engine 20 (E20) 4 On-scene, side Delta Engine 40 (E40) 4 On-scene, Alpha/Delta corner at hydrant Engine 23 (E23) 4 On-scene, RIT, side Delta Tower 18 (T18) 4 On-scene, side Alpha Ladder 20 (L20) 4 On-scene, side Delta Safety 30 (SF30) 1 Arrived after removal of second firefighter from collapse Rescue 42 (R42) 4 Self-dispatched to the incident due to recent changes in run card assignments. Arrived after

Page 4

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas the collapse, assigned to extrication of trapped firefighter Medic 40 (M40) 2 Cancelled by OEC when M18 became available and was closer to the incident. Medic 18 (M18) 3 (2 assigned, 1 intern) On-scene, requested to be added to the incident as they were available and nearby

Table 2: Second Alarm units at the time of collapse Resource Designation Staffing Level (#) Status and Location District Chief 26 (DC26) 2 Enroute. Arrived after second firefighter was extricated District Chief 19 (DC19) 2 Enroute. Arrived after second firefighter was extricated Engine 17 (E17) 4 Arrived as second firefighter was extricated Engine 25 (E25) 4 Arrived as second firefighter was extricated Engine 42 (E42) 4 Enroute. Arrived after second firefighter was extricated Engine 27 (E27) 4 Enroute. Arrived after second firefighter was extricated Ladder 7 (L7) 4 Enroute. Arrived after second firefighter was extricated Ladder 26 (L26) 4 Enroute. Arrived after second firefighter was extricated Safety 24 (SF24) 1 Enroute. Arrived after second firefighter was extricated

The fire department also has specific callsigns for each position on an apparatus. Table 3 provides the relationship between the position on the unit and the callsign. For example, L20A is the senior captain on Ladder 20. Table 3: Callsign by riding position Apparatus Type Riding Position Callsign District Chief District Chief Incident Command Technician DC#A DC#B Engine Captain Firefighter 1 Firefighter 2 E#A E#B E#C

Page 5

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Engineer Operator E#D Tower/Ladder Senior Captain Firefighter 1 Firefighter 2 Engineer Operator T#A or L#A T#B or L#B T#C or L#C T#D or L#D Rescue Captain Firefighter 1 Firefighter 2 Engineer Operator R#A R#B R#C R#D Medic Engineer Operator Firefighter 1 Firefighter 2 (or intern) M#A M#B M#C Ambulance Engineer Operator Firefighter 1 A#A A#B Safety Officer Safety Officer Safety#

Accountability Equipment The fire department utilizes an electronic accountability system that integrates the individual firefighter’s self-contained breathing apparatus (SCBA) and portable radio and transmits information to a computer terminal in the command vehicle. The accountability, or SDI terminal, is operated by the ICT that responds with the district chief. This system provides ICs with real-time accountability of on-scene firefighters. This includes digital time stamps, fireground priority reminders, and other critical command tools for managing firefighter Mayday situations. Through multiple audio and visual alerts, SDI bridges communication gaps commonly encountered on the fireground to help ICs organize and manage incidents. The system also facilitates comprehensive handling of Mayday messages, captures radio transmissions, and identifies firefighters in distress.

Page 6

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Personal Protective Equipment The deceased and injured firefighters were wearing their department issued personal protective equipment (PPE) ensemble. This ensemble included structural firefighting turnout gear that met the requirements of NFPA 1971, Standard on Protective Ensembles for Structural Fire Fighting and Proximity Fire Fighting [NFPA 1971 2018]. Both firefighters were wearing a NIOSH Approved® SCBA with an integrated personal alert safety system (PASS) device that were certified to the 2013 edition of NFPA 1981, Standard on Open-Circuit Self-Contained Breathing Apparatus (SCBA) for Emergency Services [NFPA 1981 2013]. Both firefighters in this incident were wearing the appropriate facepiece but were not connected to and breathing from their SCBA at the time of the collapse. NIOSH investigators examined and photographed the PPE ensemble and SCBA of each firefighter at the arson evidence office. The SCBA of E23B had signs of crush damage to the high- pressure lines near the PASS device (See Photos 2 and 3). No evidence was identified to suggest that the SCBA units contributed to the fatality or injury. Environmental Conditions The weather on November 6, 2024, at 22:50 hours was mostly cloudy. The temperature was 76°F with winds from the east-northeast at 5 mph. The Photo 2 (top) & 3 (bottom): Show damage to the high-pressure hose near the PASS device of the deceased firefighter’s SCBA. (Courtesy of NIOSH)

Page 7

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas humidity was 91% with no precipitation. At 23:50 hours, the humidity increased to 94% with no precipitation with winds from the east at 6 mph [Weather Underground 2024]. Building Construction The county central appraisal district records for the property states the vacant warehouse was 13,160 sq. ft. and built in 1955 of average quality. This total square footage included the 1,680 sq. ft. of space that was designated for two story offices. The building also had a partial partition wall between the two overhead doors on the Alpha side, closest to the Delta side. This wall did not go from floor to ceiling or from wall to wall. The building had no cooling system but contained unit heaters. The wall construction was primarily concrete block with a brick veneer on what appeared to be the office portion of the building (See Photos 1, 4, 5, and 6). The brick veneer post collapse appears to have a gap or separation between the veneer and the block wall in the post collapse images, if this space existed pre-collapse, it increases the risk of collapse. One of the six overhead style doors on the loading dock (Alpha) side of the building had a canopy roof over it. The office (Delta) side of the building had a small porch with steps on either side to a double door storefront style entrance with burglar bars over the doors. This entrance also had a canopy over it with several slots in the brick veneer toward the Alpha side of the building where it appeared that an original, larger canopy had once Photo 5: Photo showing interior details of the building and roof construction. (Courtesy of the fire department) Photo 4: Detail of Delta side wall showing the brick veneer over block wall, slots for previous canopy roof tie ins, roof surface covering. (Courtesy of the fire department)

Page 8

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas been installed. The Delta side canopy roof was constructed of dimensional lumber with a wooden veneer and topped with asphalt roofing material. The Bravo side of the building also had one overhead style door. The building had no fire protection system [HCAD 2024]. The roof construction was made of steel bar joists landed on wide flange steel beams which were supported by brick columns or smaller wide flange steel columns throughout the warehouse. The bar joists were topped with steel decking and roll style asphalt roofing material. This should be noted during departmental pre-incident planning, especially when there may be a lack of additional supports to maintain structural integrity. If the bar joists fail early into an incident, the roof and attached walls of the structure will separate.

Timeline The following timeline is a summary of events that occurred as the incident evolved. Not all incident events are included in this timeline. The times are approximate and were obtained by examining the dispatch records, audio recordings, witness statements, and other available information. All times are approximate and rounded to the closest minute. Each bullet occurred within the minute time block in the closest approximated order. The timeline is not intended, nor should it be used, as a formal record of events. Photo 6: View of the structure from the Alpha/Delta corner post collapse and during overhaul operations. (Courtesy of the fire department)

Page 9

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Time (Hours) Fireground Operations, Response, and Details 22:50 • First call received by PSAP for a 3-story warehouse on fire. o Additional callers begin calling and giving a different location than the initial call. 22:52 • First alarm companies dispatched to a warehouse on fire at the address given by the initial call. 22:54 • First alarm assigned companies enroute to the scene. 22:56 • SF30 enroute. • Responding units given Northeast TAC 4 (NETAC 4) channel assignment for operations. 22:57 • E18 “We’ve got a column, working fire.” o OEC copies and rebroadcasts “Working fire.” o E18 repeats this message again within the minute. 22:58 • E18 arrives on scene, “Large warehouse, fire showing from roof, heavy fire, Alpha side is 66th Street.” o OEC copies message. • E20 arrives on-scene. • L20 on-scene and reports “Working fire, mobile command.” • DC20 on-scene, “Assuming command, heavy fire from a 1-story warehouse.” 22:59 • E18 declares a Defensive Fire. • OEC asks E18 to repeat. • E18 repeats, “Defensive Fire.” • OEC transmits the emergency tone and rebroadcasts, “All units, this is a Defensive Fire.” • IC orders L20 to set up the ladder pipe. • T18 also reports they are setting their ladder pipe up and opening the garage doors on the Alpha side. 23:00 • IC copies T18. • L20 – “Command, TIC reading through open doors is about 1,000 degrees at the roof, heavy smoke, fire, and some of the roof is starting to collapse. o IC acknowledges the report, requests ladder pipes be set up on Ladder 20 and Tower 18. • IC – “SDI in place.” • E40 – “On location, connecting to a plug and supplying L20.”

Page 10

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Time (Hours) Fireground Operations, Response, and Details • E23 – “On location.” o IC – “E23, I’m going to need you to be RIT.” 23:01 • T18 – “Command, we’re going to have E18 supply T18. We’ve got a plug right behind T18. o IC copies the message. • Medic 18 requests assignment to the incident in place of Medic 40. • IC – “E40, supply L20, rear bumper.” • E20 – “Command, D side Exposure, pulling 1.75” to protect it, we don’t have a supply.” o IC copied the message. • E23 crew gets RIT equipment from DC20’s vehicle. 23:02 • E18 establishes their water supply. • E40 acknowledges their assignment, asks if they are RIT after water supply. o IC – “E23 is RIT.” • L20 – “We’re set up for water.” • L20A – “Ready to give CAN report.” 23:03 • IC – “L20, Go ahead.” o L20A – “Collapse of the roof, heavy fire, exposure on C/D side going to have to get inside, already starting to smoke up. Trying to get water on it right now. L20 in position to hit it with the pipe, no further needs.” o IC – “Received, District 20 to all companies, do not go interior.” • E20 requests personnel to assist with dragging supply line to a nearby hydrant. o IC assigns this task to Medic 18. • E23 crew stages RIT equipment at Alpha/Delta corner of building. 23:04 • IC requests status of water supply. • E18C – “Water’s coming.” • IC updates OEC that water supply is established to both ladder pipes. • IC – “Command to all companies on Delta Division, let’s stay clear of that wall, it looks like it’s going to collapse outwards.” • District 8 on location. o IC assigns District 8 side Alpha. • E23B and E23C walk up to Delta side porch to “soften” the structure. 23:05 • IC requests 2nd alarm dispatched. o OEC copies request for 2nd alarm.

Page 11

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Time (Hours) Fireground Operations, Response, and Details • E20 – “Command, we got a collapse on the Delta side.” o IC – “That’s received. Command to all companies, stay away from Delta Division.” • E23B’s radio changes to the wrong channel. E23B’s SCBA pressure is noted as having dropped approximately 650psi since arrival on-scene. 23:06 • E23A declares Mayday. o IC – “We have a positive Mayday.” o OEC – “Command you had a mayday, possibly L20D.” o IC – “Received, yes, positive Mayday, we’re working on getting him out right now.” o OEC – “Command, I just sent your 2nd alarm, do you want the 3rd alarm for your Mayday?” o IC – “Yes, give me that 3rd alarm.” • E23B’s SCBA shows another approximately 600psi drop. SCBA also showing in full alarm. 23:07 • ICT sends an electronic PAR check over the radio to E23B. E23B’s SCBA pressure drops another approximately 1200psi. • L20 – Updates IC that one person was removed. o IC copies, inquires about total missing personnel. o L20 states they think one more, going to get a PAR. • E1 arrives. o IC requests they report to IC Post. • E40 – Updates they have one more person trapped under debris. 23:08 • E23B’s SCBA pressure drops approximately 450psi. SDI showed the remaining pressure at 0psi. • L20 advises IC there are two firefighters missing, they are actively working, requests operations go to a different channel and the rescue stay on this channel. o IC copies. 23:09 • Rescue 42 arrives. o IC assigns Rescue 42 to the Delta side for extrication of trapped firefighter. 23:10 • IC inquires if there is another firefighter missing. 23:11 • Tower 18 requests water. • OEC calls IC with 10-minute incident clock and advises L20A is showing in alarm.

Page 12

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Time (Hours) Fireground Operations, Response, and Details o IC requests L20 PAR report, other personnel answer he is OK, working the rescue. • Alpha division requests additional medic units from IC. • E23B’s PASS alarm shows cleared in electronic accountability system. 23:12 • IC requests two medic units added to the incident. • Tower 18 updates they have water flowing on Alpha side of building. • IC advised the second firefighter is being extricated. 23:13 • L20 updates all victims are out, all personnel clearing the collapse zone. • IC requests all personnel evacuate the area of Delta side. • IC requests OEC cancel the 3rd alarm. • Engine 25 and Engine 17 arrive. • L20 updates they are exiting the collapse zone with the 2nd victim. They need ALS, automated external defibrillator (AED) protocol is in effect. 23:14 • IC asks all units to clear Delta side. • OEC cancels 3rd alarm. • L20 requests to ALS meet them at front of L20. • IC orders Medic 17 to Delta side. 23:15 • DC 26 and DC 46 arrive. • L20 requests ALS a third time. • IC orders E18 and T18 to stay back from the wall on the Alpha Division, L20 is about to start flowing in that area. • SF30 arrives. 23:16 • Engine 42 arrives. • T18 reports bulk of the fire is knocked on the Alpha side. Asks if IC wants the stream redirected. • OEC notifies IC that L20A is showing in alarm. Asks for confirmation that all personnel are accounted for. 23:17 • IC confirms all personnel are accounted for. • IC requests all personnel stay away from the building. • E27, Medic 40, L26 arrive on-scene. 23:18 • L20 updates all wounded are out and everyone is outside the collapse zone. • OEC updates IC they are at the 20-minute incident clock. 23:44 • IC declares the incident under control.

Page 13

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas

Investigation On November 6, 2024, at approximately 22:50 hours, a 9-1-1 call was received by the OEC for a three-story warehouse on fire with an approximate location. As the first call was being handled by civilian call takers, additional calls were being answered for the incident of a warehouse on fire near the location provided by the initial call. As the call taker began entering information into the CAD terminal, the first alarm suggested units populated. The OEC personnel dispatched the suggested units to the area the first caller provided the call taker. The dispatched units included DC20, DC08, E18, E20, E40, E23, T18, L20, and SF30. Within four minutes of the initial call, all first alarm units were enroute to the incident location. OEC assigned this incident NETAC4 for operations. At approximately 22:57 hours, E18 updated “We’ve got a column, working fire.” OEC copied the update and rebroadcasted the working fire. At approximately 22:58 hours, E18 arrived on-scene and sized the incident up as, “Large warehouse, fire showing from the roof, heavy fire, Alpha side is 66th street.” OEC copied the message. E20 approached from the cross street (i.e., Delta side) within seconds of E18’s size up. L20 arrived and positioned behind E20 on the Alpha/Delta corner (See Photo 7) and reported, “Working fire, mobile command.” Seconds later DC20 arrived and updated, “Assuming command, heavy fire from a one-story warehouse.” Time (Hours) Fireground Operations, Response, and Details 23:53 • E23B transported by Medic 18 to a nearby hospital. • E23C transported by Ambulance 23 to a nearby hospital.

Page 14

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas At approximately 22:59 hours, E18 broadcasted this incident as a defensive fire. OEC asked E18 to repeat their last message. E18 repeated, “Defensive fire.” OEC transmitted an emergency tone and rebroadcasted “All units, this is a defensive fire.” IC ordered L20 to set up their ladder pipe. T18 reported setting up their ladder pipe up and opening garage doors on the Alpha side. IC copied the report from T18. At approximately 23:00 hours, L20A radioed IC, “Command, TIC reading through open doors is about 1,000 degrees at the roof, heavy smoke, fire, and some of the roof is starting to collapse.” IC acknowledged the report and requested the ladder pipes on T18 and L20 be set up. IC updated all units that SDI was in place. E40 arrived and advised IC, “On location, connecting to a plug and supplying L20.” E23 arrived and IC assigned them RIT responsibilities. At approximately 23:01 hours, T18 updated, “Command, we’re going to have E18 supply T18. We’ve got a plug right behind T18.” IC acknowledged the message. Medic 18 requested to be added to the incident instead of Medic 40. IC ordered E40 to supply L20 through the rear bumper. E20 updated, “Command, D side exposure, pulling 1¾ to protect it, we don’t have a supply.” IC acknowledged the update. E23 crew went to DC20’s vehicle and retrieved the RIT equipment for their assignment. E18 established their water supply. E40 acknowledged their assignment and asked if they were to be RIT after establishing a water supply. IC advised E23 is RIT. L20 advised they were setup for water. L20A called IC and was ready to give a CAN report. Photo 7: Arrival conditions of L20 at the Alpha/Delta corner. (Courtesy of the fire department)

Page 15

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas At approximately 23:02 hours, IC was ready for the CAN report. L20A stated, “Collapse of the roof, heavy fire, exposure on C/D side going to have to get inside, already starting to smoke up. Trying to get water on it right now. L20 in position to hit it with the pipe, no further needs.” IC responded, “Received, District 20 to all companies, do not go interior.” E20 requests personnel to assist with dragging a supply line to a nearby hydrant. IC assigned the crew of Medic 18 to assist. E23 staged the RIT equipment on the ground near the Alpha/Delta corner of the building. At approximately 23:04 hours, IC requested the status of the water supply to which E18C responded, “Water’s coming.” IC updated OEC that a water supply was established to both ladder pipes. DC08 arrived as IC updated all companies on scene, “Command to all companies on Delta Division, let’s stay clear of that wall, it looks like it’s going to collapse outwards.” IC assigned DC08 to side Alpha. E23B and E23C walked up onto the Delta side porch to “soften” the building as part of the RIT assignment. At approximately 23:05 hours, IC requested the 2nd alarm be dispatched. E20 called IC, “Command we got a collapse on the Delta side” (See Diagram 1). IC responded, “That’s received. IC to all companies, stay away from Delta Division.” The ICT noticed that E23B’s radio suddenly changed to the wrong channel for the incident and that their SCBA pressure dropped approximately 650 psi since their arrival on scene. At approximately 23:06 hours, E23A declared a Mayday. IC acknowledged the mayday. OEC called IC, “Command you had a Mayday, possibly L20D.” IC confirmed the message, “we’re working on getting him out right now.” OEC confirmed to IC the 2nd alarm was dispatched and asked if IC wanted the 3rd alarm for the Mayday. IC asked for the 3rd alarm. E23B’s SCBA showed approximately 600psi more drop as well as the PASS showing it was in full alarm. At approximately 23:07 hours, L20 updated IC that one firefighter had been removed. The ICT sent an electronic PAR check over the radio to E23B. E23B’s SCBA pressure dropped an additional 1,200 psi. IC asked how many personnel were missing. L20 thought one more and went to get a PAR. E1 arrived and was ordered to report to the IC post. E40 updated that they had one more person trapped under debris. A moment

Page 16

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas later L20 advised IC there were two firefighters missing and actively working to get them. L20 requested operations go to a different radio channel so the rescue could have this one. E23B’s SCBA pressure dropped an additional 450 psi and the electronic accountability system in use by the department showed there was 0psi remaining in the SCBA. Diagram 1: Scene layout from above at time of collapse. (Prepared by NIOSH)

Page 17

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas At approximately 23:09 hours, R42 arrived and assigned to the Delta side for extrication of the trapped firefighter. IC asked if there was another firefighter missing. At approximately 23:11 hours, T18 requested water. OEC updated IC they were at the 10-minute incident clock and advised L20A was showing in alarm. IC requested a PAR for L20, other personnel respond he was ok and working the rescue. T18 updated they had water flowing on Alpha side of the building. IC updated the second firefighter was being extricated. E23B’s SCBA PASS alarm showed cleared in the electronic accountability system as well. At approximately 23:13 hours, L20 updated all victims and personnel clearing the collapse zone were out. IC requested all personnel evacuate the area of the Delta side and that OEC cancel the 3rd alarm. L20 exited the collapse zone with the second victim, and they needed ALS and AED protocols in effect. IC again asked all units to clear from the Delta side. L20 requested ALS meet them at the front of L20. IC ordered Medic 17 to Delta side. At approximately 23:15 hours, L20 requested ALS a third time. IC ordered E18 and T18 to stay back from the wall on the Alpha division, L20 to be flowing water in that area. SF30 arrived on location. At approximately 23:16 hours, T18 reported the bulk of the fire was knocked down on the Alpha side and asked if IC wanted the stream redirected. OEC notified IC that L20A was showing in alarm and asked for confirmation that all personnel were accounted for. IC confirmed that all personnel were accounted for and requested all personnel stay away from the building. At approximately 23:18 hours, L20 updated that all wounded personnel were out, and everyone was outside of the collapse zone. EOC updated IC they were at the 20-minute incident clock. E23C was transported to the hospital by Ambulance 23 and treated for minor injuries. E23B was transported to the hospital by medic 18 and was later pronounced deceased at the hospital. IC declared the fire under control at 23:44 and the scene was turned over to the arson division. Fire Origin and Cause The cause of the fire was incendiary/arson. Cause of Death The cause of death listed in the county medical examiner’s report is mechanical asphyxia with the manner of death being listed as homicide. Contributing Factors Occupational injuries and fatalities are often the result of one or more contributing factors or key events in a larger sequence of events that ultimately result in injuries or fatalities. NIOSH investigators identified the following items as key contributing factors in this incident: • Pre-incident planning • SOPs/SOGs • Strategy and tactics

Page 18

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas • Risk versus benefit analysis • Safety officer presence • Change of occupancy • Vacant/dangerous building inspections Recommendations Fire service organizations should: Recommendation #1: Develop pre-incident plans for structures within their first due response area, specifically dangerous or vacant structures. The fire department in this incident has a section dedicated to the authority having jurisdiction’s (AHJ) approved fire code compliance for several adopted International Code Council (ICC) fire codes: • 2021 International Fire Code (IFC) with specific amendments for application within the city. • 2021 International Building Code • 2021 International Existing Building Code (IEBC) • 2021 International Residential Code • 2021 International Swimming Pool and Spa Code • 2021 International Energy Conservation Code • 2021 Uniform Mechanical Code • 2021 Uniform Plumbing Code • 2023 National Electric Code with all applicable amendments The city handles vacant and dangerous building inspections and the associated responsibilities. The city provides a hotline for residents and visitors to report concerns. Tips received from the hotline are transferred to the appropriate entity within the city for follow up and any needed actions. Residential building concerns fall within the department of neighborhoods, inspection, and public services. Commercial building concerns fall within the responsibility of public works inspection and investigation division. These entities are entirely separate from the fire department’s fire marshal office. In this incident, the fire building was not occupied by an active business. The property had employed an individual to occupy the building as a form of security but there was no formal security in place. The individual was permitted by the building owner to live within the confines of the warehouse while it was in this unoccupied or vacant state. Chapter 3 of NFPA 1660, Standard for Emergency, Continuity, and Crisis Management: Preparedness, Response, and Recovery, defines a vacant building as a building that is currently unoccupied or unused and for which there is intention to reoccupy and reuse in the future [NFPA 1660 2024]. Section 311.1 General of the 2021 IFC specifies that temporarily unoccupied buildings, structures, premises, or portions thereof, including tenant spaces, shall be safeguarded and maintained in

Page 19

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas accordance with sections 311.1.1 through 311.6Section 311.2.1 Security of the 2021 IFC states that exterior and interior openings that are open to other tenants or unauthorized persons shall be boarded, locked, blocked, or otherwise protected to prevent entry by unauthorized individuals [International Code Council 2021]. City code of ordinances, Article 9, Division 5 for dangerous buildings lists several ways the city classifies a dangerous building including: a vacant building, regardless of its structural condition, that has been unsecured for more than seven consecutive/inconsecutive days in any 30-day period [Municode 2024]. The vacant property in this incident was known to the city public works department and had recently undergone several rounds of exterior inspections. The structure had been occupied as recently as one month prior to this incident. Although the inspections had multiple failures, none of them qualified the building as dangerous; however, it was noted that all the inspections conducted through this process were only from the exterior of the building. Interior inspection of vacant properties can only be conducted using an administrative warrant unless personnel are invited in by the property owner. Several of the line personnel interviewed by NIOSH investigators felt more communication regarding building inspections would be beneficial. It was noted during the field visit that the department recently began performing pre-incident plans for structures by emergency response personnel within the city. This was based on a risk assessment point total that fell within the range of High (36-42 points) or Special (43-49 points), requiring a pre-fire plan be performed. Appendix A shows the building’s risk assessment and classifications provided by the department as well as their assigned point values for each. Chapter 17 of NFPA 1660 states that pre-incident plans will assist personnel in effectively managing incidents and events for the protection of occupants, responding personnel, property, and the environment [NFPA 1660 2024]. Because many fire departments are unable to develop a pre-incident plan for all the structures within their jurisdiction, departments may opt to prioritize plans for structures that have 1.) elevated or unusual fire hazards, and 2.) life safety considerations [NIOSH 2024]. The information gathered and included in the pre-incident plan can help ensure the appropriate tactics are employed in the most effective way. Chapter 17 of NFPA 1660 (2024) indicates that the pre-incident plan should be a cooperative effort among the plan developer, facility management and operations staff, and responding personnel. Further, the plan should be coordinated with the incident management system (IMS). A key component to the pre-incident planning process is the provision for training and education for portions that involve unique or unusual operations. Pre-incident plans should also be available to the IC while operating at an incident. When developing pre-incident plans, items to consider include [NFPA 1660 2024]: • Potential life safety hazard, including emergency responder safety • Structure size and operations complexity • Economic impact • Importance to the community

Page 20

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas • Location and seasonal variations • Presence of hazardous materials • Susceptibility to natural disasters When developing a pre-incident plan the developer should [NFPA 1660 2024]: • Visit the property to become familiar with its layout, contents, construction, and protection features. • Use a standardized pre incident plan document which is utilized throughout the AHJ response area. Chapter 18 of NFPA 1660 states that pre-incident plan physical elements and site considerations should be classified into five groups [NFPA 1660 2024]:

1. Construction
2. Building management systems and utilities
3. External site conditions
4. Internal and external security
5. Fences or other barriers

Section 18.2 of NFPA 1660 discusses the construction of the building and what should be included in the pre-incident plan such as the size of the building in its entirety (i.e., the overall height, number of stories, square footage, and exact or approximate year of construction. Within the construction section of the pre- incident plan, including data on various points of construction may be helpful such as [NFPA 1660 2024]: • Wall construction, insulation, and integrity • Roof construction and integrity • Floor construction and integrity • Pertinent building features • Floor plan with room identifier, occupancy, and use of each room • Location, types, and construction of access features • Areas where fire, products of combustion, or other contaminants could spread due to a lack of structural barriers • Atriums • Storage arrangements • Fire command center location, access, and fire rating of the area fire walls Performing a pre-incident plan and recording the data allows the personnel responding to that structure to get inside and around the building before there is an emergency incident. This allows the personnel a basic familiarity with the location prior to the introduction added stressors during an emergency response. The 2021 IEBC allows the city to perform an inspection whenever a change of occupancy occurs within a structure. A change of occupancy is defined as [ICC 2021]:

Page 21

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas • any change in the occupancy classification of a building or structure • any change in the purpose of, or a change in the level of activity within, a building or structure • change of use This code provides an opportunity for the city to perform an inspection and pre-incident plan anytime a building changes owners or tenants to ensure awareness of hazards on the premises. While performing change of use and change of occupancy inspections, AHJ’s and fire departments could help make buildings safer by ensuring compliance to codes and standards. It is imperative to the safety of personnel that the fire department and the AHJ are operating from the same edition of codes so there are not discrepancies in enforcement. The fire department in this incident is currently in process of formally implementing pre- incident plans in emergency response. Recommendation #2: Develop and maintain SOPs/SOGs. The fire department in this incident provided their SOGs to NIOSH investigators for review. The department has well-defined guidelines that provide a reference framework for personnel to use when they respond to incidents. The fire department also provides tracking on each document for when they are reviewed or revised with a date. In the fire service, SOPs/SOGs exist to give personnel a basis to make split second decisions effectively and safely. If constructed well, operating personnel may not need to deviate from the written guidelines or policies. SOPs/SOGs allow fire departments to approach operational risk management in a standard manner that will include a full range of safety considerations [U.S. Fire Administration 2018]. Incorporating SOPs/SOGs in all facets of operations and training allows personnel to execute safety considerations as part of the IAP. NFPA 1550, Standard for Emergency Responder Health and Safety, Chapter 6 outlines that fire departments should prepare and maintain written policies and SOPs that document the organization structure, membership, roles and responsibilities, expected functions, emergency operations, and training requirements including the procedures that will be employed to initiate and manage operations at the scene of an emergency incident. NFPA 1550 also states that fire departments evaluate current trends and research annually or following a near miss or catastrophic event to determine if policies and procedures are appropriate [NFPA 1550 2024]. Department guidelines had many segments that had not been reviewed in two or more years with many not having revisions in four or more years prior to the date of this incident. Recommendation #3: Ensure ICs match their tactical strategy with the incident at hand. In this incident, the fire department was dispatched to a working fire at a vacant warehouse. Moments after arrival, E18 declared a defensive strategy for operations. Each portion of the building type for this incident has its own set of tactical considerations to consider. Chapter 12 of NFPA 1700 provides specific tactical

Page 22

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas considerations for each of these building types. Section 12.13 is specific to abandoned and vacant structures, while section 12.15 is specific for warehouses. Abandoned or vacant structures are those no longer in use and could be in an unknown condition or compromised state (e.g., weakened structural components, missing or holes in the floors, and/or structural deficiencies). When operating at these types of structures [NFPA 1700 2021]: • Consider exterior fire control prior to entry. • Anticipate early collapse. • Anticipate unpredictable or increased fire activity or impediments to normal firefighting operations based on gutted, deteriorated, and modified interiors. • Consider occupancy by squatters and transients. Warehouse and storage fires may be complex incidents that can expose responders to varying challenges and hazards. The complexity of these incidents is a function of several factors and can require significant resources to mitigate. Fires in these types of occupancies can involve large open areas and high fuel loads. Key risk factors for warehouse fires include [NFPA 1700 2021]: • Construction features including construction type, total building size, details of fire-rated enclosures, and the presence of large open fire areas • Types and hazard level of material stored • Storage configurations such as height and type (e.g., rack storage, floor storage) • Presence, type, and suitability of fire protection and detection systems. • Available methods to facilitate ventilation such as roof vents, smoke control, and exhaust systems • Availability and adequacy of water supply • Equipment and machines that involve material handling Section 12.15.2 of NFPA 1700 states that preplanning of warehouse and storage occupancies is a critical aspect of enabling an effective fire response. Section 12.15.5 of NFPA 1700 discusses how high fuel loads and long structural spans can facilitate structural collapse, suggesting that structural conditions be continually evaluated. A common denominator in both vacant fire building and warehouse fire building tactics is an increased risk of structural collapse. For large fires where the fire has vented through the roof, defensive operations should be initiated. Specifically, personnel should be evacuated from within the interior while roof areas and appropriate collapse zones are established and enforced. Aerial water streams and ground-level monitors can be used to help control the spread of fire and protect against exposures [NFPA 1700 2021]. The construction of a building can also help the IC identify an appropriate tactical strategy. The fire building in this incident was Type II construction consisting of approximately 13,160 square feet with 1,680 sq. ft. of the building designated as office space. The building contained no means of automatic fire suppression. The

Page 23

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas roof was constructed of lightweight trusses topped with metal decking and asphalt roofing material. There was also no noted fire detection system. This combination allowed the fire to free burn until a passerby called to report the fire. The building also had burglar bars on the windows and a storefront style door on the Delta side. Type II construction is split into two categories, Type II-A (protected) and Type II-B (unprotected). Type II- A requires that structural components have one-hour fire resistance, meaning the building will perform similarly to Type I but with a lower requirement for fire resistance. Type II-B allows structural components to remain unprotected, meaning major structural components have no fire resistance. A Type II-B building cannot be expected to provide structural stability under fire conditions and the failure of unprotected steel members must be anticipated. Chapter 2 of the Essentials of Fire Fighting 8th Edition Firefighter 2 discusses that when Type II construction is exposed to temperatures above 1,000°F, unprotected steel will expand and twist, pushing out walls [IFSTA 2016]. RIT operations can be separated into a few stages depending upon the scale of the incident and tactical strategy employed. The collapse of the Delta side wall occurred while the personnel assigned to E23 were in process of working through what can be described as stage one. In the article The Fire Scene: The Three Stages of RIT Operations, stage one is described as the prevention stage (Salka Jr., 2024), meaning the RIT team members assemble and place RIT tools in a staging area. Oftentimes ladders are positioned around the structure for rapid escape if conditions dictate the need. Another prevention duty of the RIT team is “softening the building” or the removal of window bars or other barriers that could prevent a firefighter retreating out through a window or door. These actions all take time to complete, and the RIT team operations may benefit from having these actions performed before the RIT team is deployed in a rescue situation [Salka Jr. 2024]. Recommendation #4: Train all personnel to employ risk management principles when operating on the incident scene. There were decisions made early into this incident that did not employ appropriate risk management principles. E18, within moments of arrival, declared this is a defensive fire over NETAC4, which was acknowledged by IC and rebroadcast by OEC with alert tones. L20A arrived within seconds of E18, performed a 360 of the structure and provided the findings to IC as well as a CAN report for effective containment of the fire. The CAN report informed the IC that the temperatures at the ceiling of the building were 1,000°F and portions of the roof collapsed. Since the early 1990s, the fire service has operated under the mantra of “Risk a lot to save a lot, risk a little to save a little” [Avsec 2023]. Performing an offensive fire attack from inside a structure increases the inherent risk to the personnel performing the attack, whereas a defensive posture may provide an added level of safety by keeping personnel outside the structure. Fire departments should provide personnel with a standardized risk management approach that could be employed at any incident [NIOSH 2020]. Personnel

Page 24

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas operating at an incident should be cognizant of the safety of themselves and their crew to maintain maximum effectiveness. Chapter 10 Incident Management of NFPA 1550 includes a section on risk management during emergency operations and while it specifically mentions the IC to include risk management in the IAP, the same could be applied to all personnel. Risk management principles should be utilized based on the following [NFPA 1550 2024]:

1. Activities that present a significant risk to personnel safety should be limited to situations where

there is a potential to save endangered lives.

1. Activities that are routinely employed to protect property are recognized as inherent risks to the

safety of personnel, and actions are taken to reduce or avoid these risks.

1. No risk to the safety of personnel is acceptable when there is no possibility to save lives or property.
2. Where the risk to fire department personnel is excessive, activities are limited to defensive

operations. Initiative 3 of the 16 firefighter life safety initiatives from the Everyone Goes Home project suggests that personnel focus greater attention on integrating risk management principles with incident management at all levels, including strategic, tactical, and planning responsibilities. This integration allows for appropriate risk management at all levels of operations from the newest to most experienced firefighter. Initiative 4, empowerment, states that all firefighters be empowered to stop unsafe practices [Everyone Goes Home 2025]. Annex J, Risk Management Plan Factors, of NFPA 1550 provides information on developing a risk management plan and includes a sample that fire departments could implement as a framework and then fine tune for their situation. The steps outlined in Annex J, per NFPA 1550 [2024] are:

1. Risk identification: Looking at every aspect of the fire department’s operation and listing potential

problems

1. Risk evaluation: Evaluating each risk identified in step one and determining the likelihood and

potential severity (e.g., expense) of the occurrence.

1. Establishing priorities for action: Evaluating the frequency and severity of risks guides how

departments establish priorities. A risk evaluated to be a low probability, but high severity often ranks as a high-priority item [NFPA 1550 2024]. Risks that have a lower likelihood can often be placed lower on the priority list.

1. Risk control: Once risks are identified and evaluated, a control for each should be implemented and

documented with one of two primary methods – 1.) Eliminate or avoid the risk or activity that presents the risk, or 2.) Develop steps to control the level of the risk.

1. Other control methods: Other suggested methods of controlling risk include developing,

implementing, and enforcing a safety program, developing SOPs/SOGs disseminating training, and completing inspections.

Page 25

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Recommendation #5: Strategically deploy safety officers. The fire department in this incident had three personnel on each shift assigned to the position of safety officer for the entirety of the city. That equates to one safety officer for every 31 firehouses on average. Fire departments may find added benefits, such as response time, if they increase the number of personnel trained to perform the duties of a safety officer. With the safety officer position operating as a separate but incorporated part of the IC staff who looks for hazards and works with personnel, they may provide a separate, individual focus on the safety aspects of the incident. In other words, they are not assigned operational tasks but are solely focused on the safety of the personnel on scene and changing conditions separate from the IC and their plan. Mayday declarations within career fire departments oftentimes occur within 20 to 35 minutes of the initial on scene time according to the 2021 Project Mayday report. Of 12,227 declared Maydays, approximately 6,352 or 51.95% were within this window of on scene time. In this incident, the Mayday occurred within seven minutes of the arrival of the first unit, putting this incident within the lowest frequency of occurrences. Only 0.5% of Maydays included within the sample set occurred in less than 10 minutes of the first arriving unit [Abbott 2021]. It is also known that a high percentage of Mayday declarations occur when only one chief officer is on scene, as high as 90% in 2019 [Marsar 2019]. This incident occurred approximately one minute after the arrival of the second district chief and ten minutes before the arrival of the assigned safety officer. The incident safety officer (ISO) is responsible for identifying activities that pose an imminent threat to firefighter safety and subsequently stop, alter, or suspend the identified activity. When an ISO identifies unsafe conditions, operations, or hazards that do not pose an imminent threat to firefighters, they should take appropriate action through the IC to mitigate or eliminate the unsafe condition. Upon arrival to the scene, the ISO should conduct a quick, face-to-face with the IC if possible, to cover the scene layout, established perimeters or control zones, initial hazards, and the strategies and tactics employed in the IAP by the IC. ISOs should perform their own, independent 360 of the scene focused on personnel safety rather than for tactical reasons [Eskwitt 2018]. Fire departments may benefit if the deployment of safety officers occurs as part of the initial assignment to any incident where personnel could be assigned tasks within an immediately dangerous to life and health (IDLH) atmosphere such as a fire, hazardous materials incident, or technical rescue. Across the fire service it is common for the first arriving officer to establish command as the IC and subsequently pass command to the first arriving chief officer for the remainder of the incident. However, there are many times in which the second arriving command officer is tasked to operations or a division of the incident. Fire departments could see significant benefit to the safety of their personnel by assigning the second arriving command officer as the ISO and altering the assignments for high-risk incident types to include three command officers on the initial assignment. This would allow for the first arriving officer to assume command, the second arriving officer to assume the ISO role, and the third officer to be tasked according to the IAP based on their arrival.

Page 26

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas The third officer could be assigned what has been traditionally assigned to the second arriving officer. This change would allow for personnel safety to be elevated on the hierarchy of the IAP.

Additionally, governing municipalities (federal, state, regional/county, and local) should:

Recommendation #6: Complete change of occupancy inspections per applicable fire and life safety codes. The building was vacant as the previous occupant relocated operations. The lack of a building occupant falls under the adopted codes as being classified as a “dangerous building with unlawful existence.” The public works department performed building inspections at this location prior to the incident but could only do so from the exterior as the building was secured. The AHJ could benefit from adding an additional process to notify various departments, including the fire department, when buildings in their respective territories are added to the dangerous or vacant building list. When the AHJ is notified of a change in occupancy, they could enact a joint inspection between fire department personnel as well as other municipal departments to allow for faster and clearer hazard communication across traditional departmental borders. Because tenants utilize structures for different purposes, when tenants of commercial buildings change, so can the fire load and hazards inherent to the building. It could go from dry good storage, hazardous material storage, to medical supplies, or just about anything else the building owner chooses to allow. NFPA 1, Fire Code, Chapter 1, Section 1.3, states that this code applies to new and existing conditions and more specifically in repairs, renovations, alterations, reconstruction, change of occupancy, and additions to buildings and requires conformity. Further, NFPA 1 states plans for a new occupancy be submitted to the AHJ prior to the change of occupancy of any existing building. NFPA 1 also requires that plans be submitted for review by the AHJ prior to altering the means of egress or fire protection systems of any existing building [NFPA 1 2024]. The issuance of an occupancy permit following a request for a change of occupancy within a structure should be dependent on satisfactorily following all applicable codes and variances adopted and enforced by the AHJ. NFPA 1, chapter 4, states that in any building or structure, a change from one occupancy classification to another shall be permitted only where the space conforms with code and variance requirements that apply to new construction for the proposed new use, except the following [NFPA 1 2024]:

1. Where, in the opinion of the AHJ, the proposed occupancy or change in use is not more hazardous

than the existing use, based on life safety and fire risk, the AHJ shall be permitted to approve such a change of occupancy provided compliance with the requirements of the adopted code for buildings of like occupancy or use are specifically incorporated to safeguard the life, health, and welfare of persons.

Page 27

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas

1. Change of tenants or ownership shall not be construed to be a change of occupancy classification

where the nature of use and assigned occupancy classification remain the same. However, the ability to make these requirements more stringent lie with the AHJ. Specifically, the AHJ can require any change of occupancy request to be approved prior to the action taking place and enforce notification of any change of occupancy occurring. The 2021 IFC defines a change of occupancy as any time the code requires a greater degree of safety, accessibility, structural strength, fire protection, means of egress, ventilation, or sanitation than exists in the current building or structure such as a change in occupancy classification, or the purpose of or level of activity within a building [ICC 2021]. To ensure compliance prior to normal business operations, the AHJ could adopt a process that includes a site inspection for code compliance and incorporates a pre-fire plan as part of the process. Incorporating a pre- incident planning component into this process could allow responding personnel to be updated on any changes with the contents, occupied hours, and any fire protection changes to the structure. This process could also allow the responding personnel a tactical and safety advantage prior to arriving on scene by having at least a basic understanding of what they may be responding to. Recommendation #7: Communicate across departmental boundaries. Cross-departmental communication can ensure the disbursement of information and improve personnel safety within local government. This includes communication to other parts of the local government as well as the residents and visitors of the municipality. Communications is of the utmost importance for local governments and the complexity of disbursement continues to increase as society becomes more technology driven. Local government provides essential services, enforces regulations, and makes decisions that impact the lives of those residing, working, and visiting their communities every day and the information related to these items need to be disseminated to everyone within the community regardless of their abilities or backgrounds [Pond 2023]. Communication across departments can also support community engagement and legal obligations. While it is imperative that local government entities ensure communication from the outward facing perspective, it is just as important to foster communication internally among departments. The local government where this incident occurred had multiple entities that had various aspects of code compliance and building complaints within the municipality. None of these entities had a process for communicating with other entities that may be involved. Creating a procedure for cross-departmental communication could help identify potential hazards in various occupancies. The building in this incident was originally reported as a potentially dangerous building but only one segment of the local government was aware of that.

Page 28

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Another form of cross-department communication that agencies may find useful are visuals. As an example, marking the building on the exterior could provide a visual indicator for arriving responders that the building may need to be approached with different tactics. This could provide the IC with a visual indicator and reminder to set up operational perimeters and collapse zones earlier into the incident, potentially from the first arriving company onward. This type of system, sometimes referred to as a High-Risk Building Management Program (HRBMP) addresses the needs associated with hazard communication for these types of structures [NIOSH 2022]. An HRBMP consists of three phases, a determination, remediation, and emergency response. The remediation phase consists of two parts, demolition and marking. For buildings that fall within an HRBMP, remediation can involve clear communication and signage. The International Association of Arson Investigators (IAAI) and the United States Fire Administration (USFA) have developed an abandoned building project that depicts specific signage to be affixed to these types of buildings as seen in Diagram 2. However, this is just one example. It is likely that local government agencies have their own programs and systems for communicating such information. References Abbott, D. E. (2021). Project Mayday. Retrieved January 2025, from http://www.firefightermayday.org/: http://www.firefightermayday.org/public/res/pdf/PM_COIC_Part_1.pdf Avsec, R. (2023, March 8th). "Risk a little to save a lot": A new look at an old mantra. Retrieved January 2025, from FireRescue1: https://www.firerescue1.com/fire-products/personal-protective-equipment- ppe/articles/risk-a-little-to-save-a-lot-a-new-look-at-an-old-mantra-glmhuy2ZKFxVw62c/ Eskwitt, S. (2018, September 12). First 5 Minutes on the Fireground: Critical Actions for Incident Safety Officers. Retrieved January 2025, from Lexipol: https://www.lexipol.com/resources/blog/first-5- minutes-on-the-fireground-critical-actions-for-incident-safety-officers/ Everyone Goes Home. (2025). 3. Risk Management. Retrieved January 2025, from https://everyonegoeshome.com: https://everyonegoeshome.com/16-initiatives/3-risk-management/ Everyone Goes Home. (2025). 4. Empowerment. Retrieved January 2025, from https://everyonegoeshome.com: https://everyonegoeshome.com/16-initiatives/4-empowerment/ Diagram 2: IAAI/USFA high-risk building marking system (Courtesy of USFA)

Page 29

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Harris Central Appraisal District. (2024). Real Property Records. Retrieved December 2024, from hcad.org: https://hcad.org/property-search/real-property/real-property-search-by-address International Code Council. (2021). 2021 International Existing Building Code. Country Club Hills, IL, United States of America. Retrieved January 2025 International Code Council. (2021). 2021 International Fire Code (IFC). Country Club Hills, IL, United States of America. Retrieved December 2024 International Fire Service Training Association. (2016). Building Construction Related to the Fire Service (4th ed.). (O. S. University, Ed.) Stillwater, Oklahoma, United States of America: Fire Protection Publications. Retrieved January 2025 International Fire Service Training Association. (2024). Essentials of Fire Fighting Firefighter 2 (8th ed.). (O. S. University, Ed.) Stillwater, Oklahoma, United States of America: Fire Protection Publications. Retrieved January 2025 Marsar, S. (2019, June 1). Your First Mayday Shouldn't be the Real Thing! Retrieved January 2025, from Fire Engineering: https://www.fireengineering.com/firefighting/your-first-mayday-shouldnt-be-the- real-thing/ Municode. (2024, October 3). Code of Ordinances City of Houston, Texas. Retrieved December 2024, from https://library.municode.com/tx/houston/codes/code_of_ordinances?nodeId=COOR_CH10BUNEPR _ARTIXBUST_DIV5DABU Muni-Link. (2024, December 5). Local Gov Communcation Strategies. Retrieved January 2025, from muni- link.com: https://muni-link.com/blog/local-government-communication-strategies/ NFPA 1. (2024). Fire Code. Quincy, MA, United States of America: National Fire Protection Association. Retrieved January 2025 NFPA 1550. (2024). Standard for Emergency Responder Health and Safety. Quincy, MA, United States of America: National Fire Protection Association. Retrieved October 2024 NFPA 1660. (2024). Standard for Emergency, Continuity, and Crisis Management: Preparedness, Response, and Recovery. Quincy, MA, United States of America: National Fire Protection Association. Retrieved December 2024 NFPA 1700. (2021). Guide for Structural Fire Fighting. Quincy, MA, United States of America: National Fire Protection Association. Retrieved January 2025

Page 30

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas NFPA 1971. (2018). Standard on Protective Ensembles for Structural Fire Fighting and Proximity Fire Fighting. Quincy, MA, United States of America: National Fire Protection Association. Retrieved December 2024 NFPA 1981. (2013). Standard on Open-Circuit Self-Contained Breathing Apparatus (SCBA) for Emergency Services. Quincy, MA, United States of America: National Fire Protection Association. Retrieved December 2024 NIOSH. (2020). F2020-10: Career captain and career firefighter die after running out of air during a search in a public library - California. by Miles S., Bowyer M., Funke J. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health. Retrieved January 2025 NIOSH. (2022). F2020-04: Career firefighter killed in a structural collapse while conducting fire attack and search in a derelict single-family residence - Missouri. By Richardson M., Miles S. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health. NIOSH. (2024). F2021-10: Volunteer firefighter killed after becoming trapped at an assisted living facility fire and two firefighters injured - New York. by Loflin M., Attwood W.,. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health. Retrieved December 2024, from https://www.cdc.gov/niosh/firefighters/programs/pdfs/face202110.pdf Pond, M. (2023, October 9). Breaking Barriers: Communications Access in Local Goverment. Retrieved January 2025, from www.lexipol.com: https://www.lexipol.com/resources/blog/breaking-barriers- communications-access-in-local-government/ Salka Jr., J. J. (2024, June 14). The Fire Scene: The Three Stages of RIT Operations. Retrieved January 2025, from Firehouse: https://www.firehouse.com/operations-training/article/21221071/the-fire- scene-the-three-stages-of-rit-operations U. S. Fire Administration. (2018). Risk Management Practices. Retrieved January 2025, from www.usfa.fema.gov: https://www.usfa.fema.gov/downloads/pdf/publications/risk_management_practices.pdf Weather Underground. (2024, November 6). Houston, TX Weather Hustory. Retrieved December 2024, from Weather Underground: https://www.wunderground.com/history/daily/us/tx/houston/KHOU/date/2024-11-6

Page 31

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Investigator Information This incident was investigated by Louis (Rick) Lago, Investigator, and Stephen J. Ringer, Investigator, both with the Fire Fighter Fatality Investigation and Prevention Program, Surveillance and Field Investigations Branch, Division of Safety Research, NIOSH located in Morgantown, WV. The report was written by Stephen J. Ringer and Dr. Wesley R. Attwood, Investigator and Program Advisor. Dan Madrzykowski and Keith Stakes from the Fire Safety Research Institute, part of the UL Research Institutes, provided a technical review of the investigation report. A subject matter expert review was provided by Michael Italia, Zoning Officer/Building Code Official, Barry Isett & Associates. The NFPA Emergency Response & Responder Safety Division also provided a technical review. Disclaimer The information in this report is based upon dispatch records, audio recordings, video recordings, witness statements, and other information that was made available to the National Institute for Occupational Safety and Health (NIOSH). Information gathered from witnesses may be affected by recall bias. The facts, contributing factors, and recommendations contained in this report are based on the totality of the information gathered during the investigation process. This report was prepared after the event occurred, includes information from appropriate subject matter experts, and is not intended to place blame on those involved in the incident. Mention of any company or product does not constitute endorsement by NIOSH, Centers for Disease Control and Prevention (CDC). In addition, citations to websites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or products. Furthermore, NIOSH is not responsible for the content of these websites. All web addresses reference in this document were accessible as of the publication date. NIOSH Approved is a certification mark of the U.S. Department of Health and Human Services (HHS) registered in the United States and several international jurisdictions.

Page 32

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas Appendix A: Building Risk Assessments and Classifications (Courtesy of the fire department) A. Life Hazards (4 – 15 points) a. Occupancy Type i. Light (1 point) - Light hazard occupancies are defined as occupancies or portions of other occupancies where the quantity and/or combustibility of contents is low and fires with relatively low rates of heat release are expected. Examples: Animal Shelters Hotels Offices Churches Institutional Residential Homes Clubs Kennels Restaurants Educational Museums Theaters and Auditoriums Hospitals Nursing Homes Used attics Libraries

ii. Ordinary (3 points) - Ordinary hazard occupancies are defined as occupancies or portions of other occupancies where the quantity and combustibility of contents are low to high and fires with relatively moderate rates of heat release are expected. Examples: Agricultural Facilities Glass Products Manufacturing Post Office Processing Automobile Parking Horse Stables Printing and Publishing Bakeries Laundries Racetrack Stables Canneries Leather Goods Manufacturing Repair Garages Cereal Mills Local Fairgrounds Resin Application Area Chemical Plants (ordinary) Machine Shops Textile Manufacturing Confectionary Products Metal Working Tire Manufacturing Distilleries Mercantile/Retail Tobacco Product Manufacturing Dry Cleaners Paper and Pulp Mills Wood Machining Electronic Plants Paper Processing Plants Wood Product Assembly Exterior Loading Docks Piers and Wharves

Feed Mills

Page 33

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas iii. Extra Hazard (5 points) - Extra hazard occupancies are defined as occupancies or portions of other occupancies where the quantity and combustibility of contents are very high, flammable, or combustible liquids are present, dust, lint, or other materials are present, introducing the probability of rapidly developing fires with high rates of heat release. Examples: Asphalt Saturating Plywood and Particle Board Manufacturing Combustible Hydraulic Fluid Areas Printing (inks, flashpoints, 100° F) Die Casting Rubber Facilities Flammable Liquids Spraying Sawmills Flow Coating Solvent Cleaning Manufactured/Modular Home Assemblies Textile Picking, Opening, Blending Metal Extruding Upholstering with Plastic Foams Plastics Processing Varnish and Paint Dipping b. Number of Occupants - Number of occupants that are typically in the building. Buildings that fluctuate with occupants have its maximum occupancy limit, designed occupancy, or the estimated amount for a normal event listed (i.e., Church – use the typical Sunday Service). i. 0 to 10

1 point ii. 11 to 50

2 points iii. >50

3 points c. Occupant Mobility – are the occupants able to egress the building without any assistance? i. Ambulatory (yes)

1 point ii. Non-ambulatory (no)

3 points d. Building Code Classification - The Building Code classifies occupancies in specified categories. The Standard of Cover manual recommends that various types of classifications be grouped together into low, moderate, high, and special risks to ensure appropriate deployment of fire suppression resources. See Building Code Classification with Examples for definition of each classification. i. Group (M)-Mercantile

1 point ii. Groups (E)-Educational, (R-1 to R-4)-Residential

2 points iii. Groups (A)-Assembly, (B)-Business, and (S)-Storage 3 points iv. Groups (I)-Institutional and (F)-Factory

4 points B. Building Type (7 to 25 points) a. Type I (1 point) - Type I Fire Resistive construction has structural components that are non- combustible materials, usually steel or concrete, that offers a fire-resistance rating that ensures the efficiency of the inherent built-in fire protection. These types of construction are most common in Groups B and I building code classifications.

Page 34

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas b. Type II (1 point) - Type II Non-Combustible construction may not afford any fire-resistive rating because it is generally designed with exposed structural elements. This type of construction is most common in Low-rise and Group R building code classifications. c. Type III (2 points) - Type III Ordinary construction is usually divided into protected and unprotected subtypes. Exterior walls are usually masonry and built with brick or block walls and have a wooden roof, wall and floor assembly which is not protected against fire. These buildings are frequently found in "warehouse" districts of older cities. This type of construction is most common in Group B and R building code classifications. d. Type IV (2 points) - Type IV Heavy Timber construction have large cross-sectional areas with all elements (columns, pillars, arches, floors, and roofs) being heavy wood design with a minimum nominal dimension of 8 inches or greater. Exterior walls are usually made of noncombustible masonry walls and usually found in older factories and mills. e. Type V (3 points) - Type V Wood Frame construction has mainly wood components for exterior and interior walls are made many other materials permitted by international and local building codes. The three most common subtypes relating to framing and foundations include Post and Beam, balloon frame, and platform frame. This type of construction is most common in Group R building code classifications. C. Square Footage - This is the square footage per floor of the building. This shall not be reduced by firewalls or separations. The County Property Appraiser files should be used if the square footage is not able to be determined. a. 0 to 2,500 sq. ft.

1 point b. 2,501 to 10,000 sq. ft.

2 points c. >10,000 sq. ft.

3 points D. Number of Stories - This is the total number of stories for the building. If a building contains a basement, this should be added as a story. Buildings with attics that can be inhabited should also be added as an additional story. a. 1 to 3

1 point b. 4 to 6

3 points c. >7

5 points E. Fire Protection - Identify if the building is protected by fire protection equipment. If the building is not completely covered, they should be considered as having none. a. Sprinkler System and Fire Alarm Monitored

1 point b. Fire Alarm or Sprinkler System Not Monitored

2 points c. None

3 points F. Fire Flow a. 0 to 1,000 gpm

1 point b. 1,000 to 3,000 gpm

2 points c. 3,001 to 4,000 gpm

3 points

Page 35

Report # F2024-06 Career firefighter dies and another injured while performing rapid intervention team tasks at vacant warehouse fire – Texas d. >4,000 gpm

5 points G. Age of Building a. 0 to 10 years

1 point b. 11 to 21 years

2 points c. >21 years

3 points H. Access to Building - The risk to the occupancy increases if access to the property is difficult. This delays operations to mitigate the incident. Examples of difficult access are parts of building not accessible to apparatus and security gates. a. No Access Problems

1 point b. Moderate Access (Security Gate)

2 points c. Significant Access Problems (50 percent inaccessible)

3 points I. Community Value (3 – 9 points) a. Historic Value - Buildings that if they exhibit a catastrophic event would have a historic impact on the community. Examples would include historic buildings, churches museums. i. No

1 point ii. Yes

3 points b. Social Value - Buildings that if they exhibit a catastrophic event would have a social impact on the community. Examples would include City Hall, Libraries, Police Stations, Recreation Centers, schools, and churches. i. No

1 point ii. Yes

3 points c. Economic Value - Buildings that if they exhibit a catastrophic event would have an economic impact on the community. Examples would include large industrial properties such as Big Box stores, and large grocery stores. i. No

1 point ii. Yes

3 points Officers should add up all points associated with each category to determine a Final Risk Assessment Score. Do to the extremely high number of buildings in each response area, a goal should be set at performing a Pre-Fire Plan on buildings with High and Special risk values first. District Chiefs and Company Officers should consider performing a Pre-Fire Plan on buildings with Low and Moderate risk as time permits. Final Risk Assessment Score

Low 14 to 22 points Moderate 23 to 35 points High 36 to 42 points Special 43 to 49 points

FD06034 ｜ 2025_美国_NIOSH_face202502.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2025 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Page i

One Firefighter Dies During Unprotected Search in Nine- Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Executive Summary On January 6, 2025, 25-year-old firefighter died after becoming disoriented and trapped while conducting an unprotected search in a nine-apartment residential building fire. At 19:44 hours, the Fire Communications Center (FCC) dispatched Engine 2, Engine 4, Engine 9, Engine 29, Ladder 2, and Battalion 1 for a building fire. A caller noted that apartment #9 was involved. After receiving multiple calls, the FCC upgraded the incident to a second alarm at 19:45 hours. The second alarm dispatched Engine 1, Engine 22, Ladder 29, and Rescue 1. At 19:47 hours, Engine 2 arrived on-scene and provided a size up of a two-story apartment building, lazy smoke visible, but no flames. They pulled a pre-connect hoseline for fire attack while Engine 29 arrived and established a water supply. At 19:48 hours, Ladder 2 and Engine 9 arrived on-scene. At 19:50 hours, Ladder 29 and Battalion 1 arrived on-scene. Battalion 1 requested a conditions, actions, needs (CAN) report. Engine 2A responded that they had water on the fire from the exterior Side Charlie/Delta and the fire was located on the first division. Engine 9 entered and was performing searches on the second division. Engine 2 made entry from Side Charlie/Delta for fire control and primary search. Ladder 2 entered the building from the front door located on Side Alpha to perform searches on the first division. Engine 9 reported to Battalion 1 that they did not have a hoseline and that there was no fire on the second division. At 19:54 hours, Battalion 1 assigned Engine 4 to stage on Side Alpha with their rapid intervention team (RIT) gear and requested that Ladder 29 perform a 360 size-up of the building and secure utilities. At 19:55 hours, Engine 9A called a Mayday noting they were cut off with fire in the stairwell on the second division. Ladder 29A reported they had flames lapping up on the Side Delta toward the second division. Battalion 1 ordered multiple units to pull additional hoselines and assist with rescue of the Engine 9 firefighters. At 19:59 Photo 1: The apartment building, post-fire, Side Alpha (Courtesy of NIOSH) REPORT 2025-02 • April 2025

Page ii One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 hours, Engine 9A reported that they were on the second division landing on Side Alpha. He further reported that his facepiece was melting, they had two firefighters, two firefighters were missing, and they needed water. At 20:01 hours, Engine 9A reported to Battalion 1 they had two firefighters out of the building with Rescue 1 but were still missing two firefighters. Multiple units reported that deteriorating fire conditions were delaying entry to the building. Engine 9A advised Battalion 1 that the missing firefighters were Engine 9C and Engine 9D (deceased firefighter). One minute later, Engine 9C transmitted that he was outside the building. At 20:21 hours, Engine 4A transmitted that they found the downed firefighter on the Bravo/Charlie corner on the first division and requested assistance. At 20:24 hours, Ladder 29A transmitted priority traffic that if all personnel were accounted for, all units should go to a defensive strategy as the second division was now fully involved. At 20:25 hours, Battalion 1 advised the FCC that the victim had been extricated and moved to EMS. Engine 9D was transported to a local hospital where he was pronounced deceased. Contributing Factors • Unprotected interior search • Hoseline selection and deployment • Door control/flow path • Mayday operations • Change of occupancy. Key Recommendations Fire departments should: • Ensure that interior search crews’ means of egress are protected by a staffed hoseline. • Use the appropriate hoseline based on conditions present. • Train on recognition of flow paths and interior fire attack techniques utilizing door control. • Employ proactive RIT operations for access and egress.

Governing municipalities (federal, state, regional/county, and local) should: • Ensure that the delegated zoning or building code enforcement authority shares change of occupancy information with the fire department.

The National Institute for Occupational Safety and Health (NIOSH) initiated the Fire Fighter Fatality Investigation and Prevention Program to examine deaths of fire fighters in the line of duty so that fire departments, fire fighters, fire service organizations, safety experts and researchers could learn from these incidents. The primary goal of these investigations is for NIOSH to make recommendations to prevent similar occurrences. These NIOSH investigations are intended to reduce or prevent future firefighter deaths and are completely separate from the rulemaking, enforcement, and inspection activities of any other federal or state agency. Under its program, NIOSH investigators interview persons with knowledge of the incident and review available records to develop a description of the conditions and circumstances leading to the deaths in order to provide a context for the agency’s recommendations. The NIOSH summary of these conditions and circumstances in its reports is not intended as a legal statement of facts. This summary, as well as the conclusions and recommendations made by NIOSH, should not be used for the purpose of litigation or the adjudication of any claim. For further information, visit the program at www.cdc.gov/niosh/firefighters/fffipp/ or call 1-800-CDC-INFO (1-800-232-4636).

One Firefighter Dies During Unprotected Search in Nine- Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii

Page 1 REPORT F2025-02 • April 2025 Introduction On January 6, 2025, 25-year-old firefighter died after becoming disoriented and trapped while conducting an unprotected search in a nine-apartment residential building fire. On January 27, 2025, the U.S. Fire Administration (USFA) notified the National Institute for Occupational Safety and Health (NIOSH) of this incident. On February 2 – 9, 2025, an investigator representing the NIOSH Fire Fighter Fatality Investigation and Prevention Program (FFFIPP), along with a representative from the International Association of Fire Fighters (IAFF), traveled to Hawaii to investigate this incident. The investigators conducted interviews with command officers, fire officers, firefighters, fire investigators, and other emergency personnel who were on-scene at the incident. NIOSH investigators inspected the personal protective equipment (PPE) used by the firefighter and inspected the apartment building. The investigators reviewed fire department standard operating guidelines (SOGs), training records, dispatch records, witness statements, investigation documents, and the medical examiner’s report. Fire Department This career fire department has a 597 square mile jurisdiction, serves a population of one million residents, annually responds to an average of 60,000 calls, and provides emergency medical services (EMS) at the Basic Life Support (BLS) non-transport level. The fire department maintains 43 stations and an aircraft station.

The fire department is comprised of 1,120 uniformed personnel across five battalions. Fifteen suppression battalion chiefs and two administrative battalion chiefs manage the five battalions that are divided into three platoons each. Daily staffing is 333 firefighters. Staffing for engine and aerial companies is four to five personnel which includes a fire captain, an apparatus operator, and two to three firefighters. Hazmat and rescue companies are staffed with five personnel, including a fire captain, an apparatus operator, and three hazmat technicians or rescue technicians. The fire department’s leadership is comprised of a fire chief, deputy fire chief, and four assistant chiefs who manage the four divisions.

Assistant chiefs oversee these divisions: • Fire Operations manages incident response. • Administrative Services performs budget, human resources, occupational safety and health, inventory management, and apparatus maintenance and replacement.

Page 2 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 • Support Services manages the training and research bureau and fire prevention bureau which conduct activities such as firefighter recruit training, incumbent firefighter training and education, certification, fire code enforcement, fire investigation, and public education. • Planning and Development facilitates the accreditation process, maintains the department’s strategic plan, grant funding programs, FCC, and radio operations. Training, Education, and Professional Development The fire department maintains a training and research bureau that provides initial firefighter recruit training, incumbent firefighter training, and continuing education for all firefighters, fire officers, and EMS personnel. The initial training for recruits consists of eight months of academy lecture and hands- on skills development. Recruits graduate the academy with International Fire Services Accreditation Congress (IFSAC) professional certification as Fire Fighter I, Fire Fighter II, Fire Apparatus Driver/Operator, and Hazmat Awareness and Operations. Recruits also graduate with a City Operator certification and as a nationally registered emergency medical technician (EMT).

Engine 9D (deceased firefighter) had 1.5 years of fire service with the department. He maintained training in and held numerous certifications as Fire Fighter I, Fire Fighter II, Fire Apparatus Driver/Operator, Hazmat Awareness and Operations, and EMT provider.

During this incident, the first four arriving apparatus on-scene had acting company officers (Engine 2, Engine 29, Ladder 2, and Engine 9). All were trained to the professional development set by the department, including a national hazard zone incident command training and certification program. Apparatus and Communications The fire department maintains 63 frontline apparatus, including 43 engines, 16 aerials, and four tankers. The aerials include a total of eight two-piece combination apparatus (equipped with a 100’ mid-mount aerial ladder), six quint apparatus (equipped with a 75’ rear mount aerial), and two tower apparatus (equipped with an 85’ rear mount platform). The fire department has two hazmat companies, two heavy rescue companies, and three helicopters.

Table 1. Units dispatched to the building fire between 19:44 – 19:45 hours

First Alarm Units Dispatched Second Alarm Units Dispatched Apparatus Arrival On-Scene Apparatus Arrival On-Scene Engine 2 19:47 Rescue 1 19:49 Engine 29 19:47 Ladder 29 19:50 Ladder 2 19:48 Engine 1 19:56 Engine 9 19:48 Engine 22 19:57 Engine 4 19:49

Battalion 1 19:50

Page 3 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 The primary Public Safety Answering Point (PSAP) dispatches for the jurisdiction’s city services. These include law enforcement, fire, EMS, and mass transit. The FCC has four platoons with six dispatchers each. All fire dispatchers are also certified firefighters. Radio communications are conducted on dedicated tactical channels that are assigned to different areas of the jurisdiction. Box assignments include four engines, a ladder, and a battalion chief. The FCC can upgrade the incident based on what is reported. Second alarm adds two engines and a ladder. Third alarm adds two engines, a ladder, and an additional battalion chief. The portable radio designation/identifier for each firefighter is based upon assignment in each apparatus. This includes “A” for officer, “B” for driver/operator, and “C” or “D” for firefighters. Building Construction and Occupancy Status This single-family residence, built in 1906, was located at the front of the property near the street. Official department of planning and permitting records indicate that the building was relocated to the rear of the property in 1954 (see Photo 2), and an additional residence was built on the same property against the street. The two- story building involved in the incident was 32 ft x 62 ft, with 1,770 square feet on the first division, and 550 square feet on the second division. After the building was relocated, it was converted from a single-family residence into a nine-unit multi-family residence, with six units on the first division and three units on the second division. The building had a post and pier foundation, 2” x 6” floor joists, and 1” x 12” floor sheathing (see Photo 3). The walls were a combination of 2” x 4” studs with 1” x 12” interior sheathing and 1” exterior shiplap siding, and single wall construction. Single wall construction is unique to Hawaii and utilizes 1” x 12” boards oriented horizontally on one side, and 1” x 12” boards oriented vertically on the opposite side to create walls without any vertical structural members or studs (see Photo 4). The roof consisted of conventionally framed 2” x 4” rafters with 1” x 4” skip sheathing, which had been covered over by plywood at some point. The building had been Photo 2: The apartment building (red square), post- fire, view from above on Side Alpha. (Courtesy of the fire department) Photo 3: Post and pier foundation. (Courtesy of NIOSH)

Page 4 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 remodelled during its life cycle, and as a result, there was a combination of 1” shiplap and Canec fiberboard interior wall finishes. Canec is a fiberboard building material made from sugar cane bagasse, which is the residual fiber that remains after juice has been extracted from the sugar cane. Canec fiberboard was used extensively as an interior wall finish until 1964. In this incident, the Canec interior wall coverings were heavily charred and likely contributed to flame spread within the structure (see Photo 5).

Throughout the building, there were several doors that were permanently shut to separate the individual dwelling spaces. These doors did not provide sufficient fire separation and even contributed to fire spread between units (see Photo 6). Occupancy Status In 1955, after the building had been relocated to the rear of the property, official building records show that the building’s occupancy classification was listed as “Hotel & Apartments”. This is the first recorded documentation indicating that the building was no longer a single-family residence. In 1974, a Certificate of Occupancy was granted for the building to contain 10 individual apartments. These occupancy changes were never shared with the fire prevention bureau, and as a result, it was never included in their inspection program or identified as a multi-family occupancy. The apartment layout within the building is shown below in Diagrams 1 and 2.

Photo 4, 5, and 6: Single wall construction (left) Canec fiberboard interior wall finishes (middle), and door (right). (Courtesy of NIOSH)

Page 5 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02

Diagram 1: First division of the apartment building with windows shown in blue. (Courtesy of the Bureau of Alcohol, Tobacco, Firearms and Explosives)

Page 6 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02

Incident Timeline The following timeline is a summary of events that occurred as the incident evolved shortly after 19:44 hours on Monday, January 6, 2025. Not all incident events are included in this timeline. The times are to Diagram 2: Second division of the apartment building with windows shown in blue. (Courtesy of the Bureau of Alcohol, Tobacco, Firearms and Explosives)

Page 7 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 the minute and were taken from the fire departments’ National Fire Incident Reporting System (NFIRS) fire reports, dispatch log, on-scene accountability documentation, and interview notes. Time Fireground Operations, Response, and Details January 6, 2025

19:44 Hours • FCC dispatched Engine 2, Engine 4, Engine 9, Engine 29, Ladder 2, and Battalion 1 for a building fire with the caller noting apartment #9. 19:45 Hours • FCC upgraded the incident to a second alarm and dispatched Engine 1, Engine 22, Ladder 29, and Rescue 1. 19:47 Hours • Engine 2 arrived on-scene and provided a size up of a two-story apartment building, smoke visible, but no flames. • Engine 29 radioed that they would establish a water supply. 19:48 Hours • Ladder 2 and Engine 9 arrived on-scene. 19:49 – 19:50 Hours • Engine 4, Ladder 29, and Battalion 1 arrived on-scene. 19:51 Hours • Engine 2A reported having water on the fire from the exterior and were trying to go to the interior. The fire location was reported to be on the first division on Side Charlie/Delta. 19:53 Hours • Engine 2A made entry for a primary search. • Engine 9A was completing a primary search on the second division without a hoseline – no fire identified. 19:54 Hours • Engine 4 staged on Side Alpha with their RIT gear. • Ladder 29 performed a 360 size-up to report and secure utilities to the building. 19:55 Hours • Engine 9A called a Mayday noting they were cut off on the second division with fire in the stairwell. • Ladder 29A reported flames lapping up on Side Delta toward the second division.

Page 8 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 Time Fireground Operations, Response, and Details 19:57 Hours • Engine 9A requested a hoseline while reporting to Battalion 1 that they were on the second division Side Alpha, above the stairwell, overcome with flames. 19:58 Hours • FCC advised all units on the fireground that a Mayday has been declared and requested all companies to maintain radio discipline. 19:59 Hours • Engine 9A reported being on the second division landing but unsure what side of the building. His facepiece was melting, they had PAR of two, were missing two firefighters, and they needed water. 20:01 Hours • Engine 9A reported to Battalion 1 that two firefighters were out with Rescue 1 but were still missing two firefighters. • Engine 22A, not yet entered, reported being on Side Bravo and hearing screams for help in the back corner of the building on Sides Bravo/Charlie. 20:02 – 20:03 Hours • Engine 9A advised Battalion 1 that the missing firefighters were Engine 9C and Engine 9D (deceased firefighter). • Engine 9C said they were outside the building but Engine 9D was still unaccounted for. 20:04 Hours • Battalion 1 requested the FCC send two EMS units to the scene. 20:06 Hours • Ladder 29A had a ladder on Side Alpha, two ladders on Side Bravo, and were unable to ladder Sides Charlie and Delta due to the fire. 20:07 Hours • Engine 29A was on the Charlie/Delta corner with heavy fire involvement and partial roof collapse. 20:09 Hours • Engine 22A had three hoselines conducting fire attack from the exterior on Sides Bravo and Charlie. They reported no longer hearing screams for help and did not have control of the fire yet. 20:17 Hours • Engine 22 advised Battalion 1 that they had water on the fire on the second division of the Sides Bravo and Charlie but were unable to control the fire from their position.

Page 9 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 Personal Protective Equipment At the time of the incident, Engine 9D was wearing full structural firefighting turnout gear and a NIOSH Approved® MSA Model G1, 45-minute, 4500 psi SCBA. The investigators inspected and photographed the PPE. Engine 9D had a personal two-way radio that was confirmed to be in the “on” position and on the correct tactical channel. During the incident, Engine 9D made two inaudible radio transmissions. He manually activated his PASS device approximately seven minutes after turning on his SCBA. There was no evidence to suggest that the turnout gear or SCBA contributed to the fatality. Weather Conditions At 18:53 hours on January 6, 2025, the outdoor temperature was 77°F, dewpoint was 70°F, the wind was out of the southwest at 8 mph, there had been no precipitation in the last 24 hours, and conditions were fair [Weather Underground 2025]. Investigation At 19:44 hours, the FCC dispatched Engine 2, Engine 4, Engine 9, Engine 29, Ladder 2, and Battalion 1 for a building fire, with the caller noting apartment #9. After receiving multiple calls, the FCC upgraded the incident to a second alarm at 19:45 hours. The second alarm dispatched Engine 1, Engine 22, Ladder 29, and Rescue 1. The FCC advised all incoming units to operate on the B2 tactical channel 3 for the incident.

Time Fireground Operations, Response, and Details 20:21 Hours • Engine 4A transmitted priority traffic that they found the downed firefighter on the Bravo/Charlie corner and requested assistance. 20:24 Hours • Ladder 29A transmitted via priority traffic that if all personnel were accounted for, all units should go to a defensive strategy as the second division was now fully involved. • Engine 4A came down with the victim through Side Alpha and needed medical assistance immediately. • Battalion 1 ordered all units not assisting with the victim to completely exit and evacuate the structure. • The FCC transmitted emergency tones, informed all personnel that strategy is changed from offensive to defensive, all personnel exit the building and give PARs upon exiting. 20:25 Hours • Battalion 1 advised the FCC that the victim was extricated and being transferred to EMS. Engine 9D was then transported to a local hospital and pronounced deceased.

Page 10 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 At 19:47 hours, Engine 2 arrived on-scene and provided a size up of a two-story apartment building, lazy smoke visible, but no flames. Engine 2 pulled a pre-connect, 200 feet of 1 3/4”, for fire attack and requested the next arriving engine establish a water supply. Upon arrival, Engine 29 connected to the hydrant with 100 feet of 4” supply line and relayed water to Engine 2. Engine 29’s crew observed grey smoke as they approached the building. At 19:48 hours, Ladder 2 and Engine 9 arrived on-scene. Engine 2A requested their hoseline be charged and requested another company to bring forcible entry tools. Engine 2 firefighters observed smoke and fire coming through a window on Side Delta (apartment #9 kitchen). At 19:49 hours, Engine 4 arrived on-scene.

At 19:50 hours, Ladder 29 and Battalion 1 arrived on-scene. Battalion 1 staged 500 feet down the road from the incident and performed command from his vehicle using a tactical board and aide. Engine 2A again requested their hoseline be charged. Engine 29D made the same request for a charged hoseline but was on the wrong tactical radio channel. Engine 2A made an additional, new request to have their hoseline pressure increased. Engine 2B boosted the pressure to 120psi. Engine 2B previously secured the water supply before sending pressure to the hoseline. At 19:51 hours, Battalion 1 requested confirmation on incident status, indicating he had Engine 2 with a pre-connect stretched into the building; Engine 29 providing a water supply; and Engine 9, Ladder 2, Ladder 29, and Engine 4 on-scene. After Battalion 1requested a CAN report, Engine 2A responded that that they had water on the fire from the exterior (Charlie/Delta corner) and were going to try to go interior. Battalion 1 requested the location of the fire and asked if another company was assisting. Engine 2A responded that the fire was located on the first division. At 19:52 hours, Battalion 1 assigned Engine 29 to stage on Side Alpha as a 2 ½” pre-connect hoseline was stretched to Side Alpha.

At 19:53 hours, Engine 2A told Battalion 1 that they were making entry for a primary search on Sides Charlie/Delta of the first division. Rescue 1A then said that Ladder 2 firefighters were making entry on Side Alpha of the first division. They observed smoke coming from the sill on the Alpha/Delta corner prior to making entry. Battalion 1 requested that Ladder 2 complete a primary search on the first division. Engine 9A told Battalion 1 that they were conducting a primary search on the second division. Battalion 1 asked Engine 9A if they had a hoseline, to which Engine 9A said no and that there was no fire on the second division. Engine 9, however, had a set of irons and a thermal imaging camera with them. Upon entering the second division, they had limited visibility with light colored smoke but no heat.

At 19:54 hours, Battalion 1 assigned Engine 4 to stage on Side Alpha with their RIT gear. Battalion 1 requested that Ladder 29 perform a 360 size-up to report and secure utilities to the building. Engine 29A indicated they had an electric service line above them on the Alpha/Delta corner with flames blowing out on Side Delta. As Ladder 2 completed a search of apartment #3 (Alpha/Bravo corner), they observed light smoke and a flash coming from down the hallway (towards Side Charlie). As Ladder 2 firefighters completed searches of apartments #1 and #3, a firefighter kicked open the door of apartment #2. Upon entering between two to three feet into apartment #2, heat and flames were observed spreading towards the firefighter from Side Charlie of the room. Within approximately 10 seconds, the room flashed, and the firefighter immediately exited apartment #2. Ladder 2 was forced to evacuate out Side Alpha. This fire extended out of apartment #2, into the hallway, and eventually up the stairs to the second division.

Page 11 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 At 19:55 hours, Engine 9A called a Mayday noting they were cut off on the second division with fire in the stairwell. He also engaged his radio’s emergency alert button. Battalion 1 acknowledged the transmission and Ladder 29A reported there were flames lapping up on Side Delta toward the second division. They deployed another pre-connect hoseline, 200 feet of 1 ¾”, off Engine 2. The FCC confirmed to Battalion 1 that they had a Mayday from Engine 9A. Battalion 1 requested the FCC to sound the emergency tones. Engine 9 firefighters were forced to the floor due to rapid heat and zero visibility.

At 19:57 hours, the FCC transmitted the emergency tones and Battalion 1 advised all companies operating at the incident that they had a confirmed Mayday. Engine 9A reported to Battalion 1 that they were on the second division, above the stairwell, overcome with flames. They requested a hoseline. Battalion 1 requested Engine 1 to stretch a line to the second division and assist with getting the Engine 9 firefighters out. Engine 22A then advised that they were on Side Bravo with a hoseline set up. Battalion 1 requested they stretch that hoseline to the second division and get Engine 9 out. Engine 22A ordered their firefighters to make entry. At 19:58 hours, the FCC advised all units on the fireground that a Mayday has been declared and requested all companies to maintain radio discipline. Ladder 2A advised Battalion 1 that they had a ground ladder and asked if he would like them to ladder the building.

At 19:59 hours, Battalion 1 requested a CAN report from Engine 9. Engine 9A responded that they were on the second division landing but unsure what side of the building. He further reported that his facepiece was melting, they had PAR of two, were missing two firefighters, and they need water. Battalion 1 then requested a CAN from Engine 22A who responded that they were making entry. Engine 22B advised that their apparatus was positioned at the Alpha/Bravo corner with only one hoseline deployed. Battalion 1 requested Engine 4 to stretch another hoseline off Engine 22 to get into the second division to assist with the Mayday. Engine 4 deployed 200 feet of 1 ¾”. Two Engine 9 firefighters, Engine 9A and 9B, found a door at the end of the hallway (apartment #8). Engine 9B used strength and an axe to breach through the door unaware of where it went, but thinking the environment had to be better on the other side. Engine 9A was now alone and found the hole in the door by feel. Both firefighters exited a window leading to a small roof. They then climbed down a 24 ft ladder thrown by Rescue 1.

At 20:01 hours, Engine 9A reported to Battalion 1 they have two firefighters out of the building with Rescue 1 but are still missing two firefighters. Battalion 1 acknowledged. Engine 22A reported that they were on Side Bravo and had not made entry yet. They heard screams for help in the back corner of the building. Battalion 1 requested a status report from Ladder 29 to soften the building to assist with the rescue operation. Ladder 29A reported that they have one ground ladder in place and two firefighters from Engine 1. Battalion 1 requested that all sides of the building be softened. At 20:02 hours, Engine 9A advised Battalion 1 that the missing firefighters were Engine 9C and Engine 9D (deceased firefighter). One minute later, an inaudible radio transmission was received from Engine 9C. Battalion 1 responded asking Engine 9C to locate Engine 9D, turn on their personal alert safety system (PASS) devices and a flashlight, and find a window to call for help. Engine 9C responded that he was outside the building but Engine 9D was still unaccounted for. Battalion 1 then attempted to call Engine 9D by name on the radio and requested they respond.

Page 12 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 At 20:04 hours, the FCC requested confirmation that the only missing firefighter at the incident was Engine 9D. Battalion 1 confirmed to the FCC and then relayed the message to Engine 22 and Engine 4. Engine 22A confirmed they are looking for Engine 9D but that they were still knocking down the fire and hadn’t made entry. Battalion 1 then requested the FCC send two EMS units to the scene. The FCC confirmed that one EMS unit was on-scene, another was being dispatched, and asked Battalion 1 if he wanted a third alarm. Battalion 1 declined the third alarm but requested two racks of SCBA air cylinders and a fire investigator be sent to the scene. He also advised the FCC that an electric company representative on-scene.

At 20:06 hours, Battalion 1 requested a CAN report from Engine 4. Engine 4A responded that they were on Side Bravo stretching a hoseline to get to the second division to make entry. Ladder 29A then advised that they have a ladder on Side Alpha, two ladders on Side Bravo, and were unable to ladder Sides Charlie and Delta due to the fire. At 20:07 hours, Engine 29A reported that they switched out with Engine 2 personnel. They stretched a 1 ¾” bumper line from Engine 2. They reported they are on the Charlie/Delta corner with heavy fire involvement and partial roof collapse. Ladder 2A then reported they cleared the first three rooms on the first division and requested to exit the building to swap out air cylinders. Battalion 1 acknowledged, and Engine 2A then reported that they were exiting to swap out air cylinders.

At 20:09 hours, Engine 22A requested Engine 22B to increase water pressure to their hoseline. Battalion 1 then requested they provide a CAN report. Engine 22A advised that they have three hoselines conducting fire attack from the exterior on Sides Bravo and Charlie. They reported they no longer heard screams for help and did not have control of the fire yet. Battalion 1 acknowledged. At 20:11 hours, Battalion 1 assigned Ladder 2 to stage on Side Alpha and requested a CAN report from Engine 4 who reported they were still trying to get up to the second division. At 20:13 hours, Engine 29A then reported that they were at the Charlie/Delta entrance looking in but could not see Engine 9D and needed water. Battalion 1 asked if they were interior or looking from exterior to which Engine 29A responded that they were exterior firefighting the fire. Battalion 1 requested they continue into the building to search the entire room. At 20:15 hours, Battalion 1 requested that Engine 2 stage on Side Alpha and that Engine 1 back up Engine 29 on the second division. He requested that they back them up on their hoseline or take another with them. Engine 29A then responded with a correction that they were on the first division of the Charlie/Delta corner and making their way to the second division.

At 20:17 hours, Engine 22 advised Battalion 1 that they had water on the fire on the second division of the Sides Bravo and Charlie but were unable to control the fire from their position. Battalion 1 then attempted to raise Engine 9D on the radio again. Ladder 29A reported that cleared from the second division of the Delta corner while Battalion 1 requested a CAN report from Engine 4. At 20:20 hours, Engine 29A reported the roof was fallen through with conditions changing to heavy smoke. Engine 4A advised Battalion 1 that they were still on the first division and completed a search of three rooms on Side Bravo and one on Side Delta. They reported they were making their way to the second division and still had fire on the first division Side Charlie. The FCC advised Battalion 1 that Engine 29 had a second Mayday. Battalion 1 requested a CAN report from Engine 29. Engine 29A requested to cancel their

Page 13 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 Mayday stating it was emergency traffic about the roof collapsing and second division fire penetration. Engine 29A confirmed they were not in an emergency situation.

At 20:21 hours, Engine 4A transmitted priority traffic that they found the downed firefighter on the first division, Bravo/Charlie corner and requested assistance. Engine 4A noted that Engine 9D’s PASS was active, his foot was stuck on something, and he did not have a low air alarm. Battalion 1 acknowledged while Engine 29 reported they were going to swap out air cylinders and recycle. Ladder 29A reported they were conducting fire attack from the Charlie/Delta corner and had partial roof collapse on the second division. Battalion 1 requested Ladder 29A assistance with the rescue. Ladder 29A declined as they were conducting fire attack from the Side Delta exposure. At 20:22 hours, Engine 2A reported they were on the Charlie/Delta corner making entry to assist with the rescue. Battalion 1 requested confirmation of the unit that found the downed firefighter. The FCC then responded that they believed it was Engine 4 that had Engine 9D.

At 20:24 hours, Ladder 29A transmitted priority traffic that if all personnel are accounted for, all units should go to a defensive strategy as the second division was now fully involved. Battalion 1 confirmed and requested a CAN report from Engine 4 along with their location. Engine 4A responded that they were coming out with the victim through Side Alpha and needed medical assistance immediately. Battalion 1 then ordered all units not assisting with the victim to completely exit and evacuate the structure. The FCC then sounded the emergency tones and advised all units on-scene that per Battalion 1, they were to change strategies from offensive to defensive. The FCC requested all units exit the building and conduct PAR checks. At 20:25 hours, Battalion 1 advised the FCC that the victim had been extricated and they were getting him to EMS. Engine 9D was transported to a local hospital where he was pronounced deceased. Fire Origin and Cause The fire department, in collaboration with the Bureau of Alcohol, Tobacco, Firearms and Explosives National Response Team, completed an on-scene origin and cause fire investigation. The fire was classified as accidental due to unattended cooking. The fire originated in the kitchen of apartment #9, located at the rear of the structure. Contributing Factors Occupational injuries and fatalities are often the result of one or more contributing factors or key events in a larger sequence of events that ultimately result in injuries or fatalities. NIOSH investigators identified the following items as key contributing factors in this incident that ultimately led to the fatality: • Unprotected interior search • Hoseline selection and deployment • Door control/flow path • Mayday operations • Change of occupancy

Page 14 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 Cause of Death According to the medical examiner report, Engine 9D’s cause of death was thermal injuries and inhalation of combustible products. Recommendations Fire departments should:

Recommendation #1: Ensure that interior search crews’ means of egress are protected by a staffed hoseline. Discussion: In this incident, Engine 9 made entry and conducted interior searches on the second division. When they entered, they had clear conditions, lights on within the building, and no indication of fire. They did not take a hoseline with them to control rapid fire growth or provide protection. Their means of egress was cut off by rapidly developing fire from the first division.

When conducting a search or fire attack above a fire, it is important for firefighters to recognize the potential danger of being trapped by the fire. This danger is influenced by the construction of the building. In buildings such as the one in this incident, vertical fire spread is more rapid. Flames may spread vertically and trap firefighters operating above the fire by traveling up the interior stairs, within concealed spaces, through windows, or up the combustible exterior siding [NIOSH 2012].

Fire departments should ensure that a charged hoseline is either advanced with the interior search crew or is operated by another firefighter to provide protection while entering an immediately dangerous to life and health environment. The only exception for entering a structure without an operational hoseline is when an occupant can be seen or heard within a few feet of the structure entry or the fire area. Additionally, deployed backup lines should be operational and dedicated to protect the interior crews and their means of egress [NIOSH 2012]. A backup hoseline performs three critical functions [IFSTA 2013]:

• Protecting the primary interior attack team from extreme fire behavior • Protecting the means of egress for interior team(s) • Providing additional fire suppression capacity in the event fire volume increases

Recommendation #2: Use the appropriate hoseline based upon current conditions. Discussion: At this incident, the fire department had several 1 ¾” hoselines and one 2 ½” hoseline deployed around the structure for fire suppression. As the fire grew in intensity, units notified the incident commander that they had water on the fire, but it was not enough to gain control of the fire. Use of deck cannons from on-scene apparatus were not effective due to the limited access and obstructions to the fire building. This delay in fire suppression hampered the ability of RIT and others involved in the rescue operation to enter the structure.

Page 15 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 Firefighters need to provide the appropriate amount of water in the initial attack to put out the fire and stop potential fire growth. Fire departments generally use an 1 ¾” hoseline as the initial preconnect for residential structure fires. These lines are operated to provide 160 gallons per minute. These hoselines can be effective for suppression when the fire is confined to a single room and is in either the incipient or growth stages. However, modern furnishings such as televisions and mattresses are increasingly being manufactured with synthetic materials with high levels of hydrocarbons. These furnishings burn at greater speeds and intensity, facilitating rapid fire growth and extension [DeStefano 2016].

During an incident, firefighters should select and deploy the appropriate hoseline based upon fire conditions. While monitoring fire suppression activities, the incident commander, incident safety officer, or operations should evaluate whether the hose size is appropriate for the current fire conditions and other factors including [IFSTA 2015; IFSTA 2013]:

• Fire load and material involved • Location of fire • Size of building and fire area • Potential fire spread

Selecting the correct size of hoseline is critical. Inability to provide a sufficient volume of water for the size of the fire may delay extinguishment and expose firefighters to danger from rapid fire development.

Recommendation #3: Train on recognition of flow paths and interior fire attack techniques utilizing door control. Discussion: At this incident, Ladder 2 firefighters were completing searches in apartments #1 and #3 as a firefighter kicked open the door of apartment #2. Heavy fire immediately exited apartment #2 and created a flow path into the hallway, forcing Ladder 2 to evacuate out Side Alpha.

A flow path is “the movement of heat and smoke from the higher pressure within the fire area towards the lower pressure areas accessible via doors, window openings, and roof structures” [NFPA Glossary of Terms 2021]. Flow paths consist of an inlet and an exhaust with the direction of travel being determined by pressure. Heat and smoke in a high-pressure area will travel to an area of lower pressure. It is possible to have multiple flow paths within a structure dependent upon the size of the building, openings, closures such as fire doors, and overall structure design. Personnel working in the flow path (between the seat of the fire and the exhaust) operate in a significantly risky environment.

Maintaining door control is a critical component of coordinated ventilation and fire attack. Opening a door changes the ventilation profile of the structure and may have an adverse effect on fire behavior [IFSTA 2013]. Controlling the door limits the fresh air available to the fire and reduces the flow path. Once water is applied to the fire, it is no longer ventilation limited. This allows for more rapid and safer ventilation [Gray and Norwood 2024].

Page 16 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 Recommendation #4: Employ proactive RIT operations for access and egress. Discussion: This fire department utilized rotating units for RIT until the arrival of a heavy rescue company. During this incident, Engine 4 was assigned to stage on Side Alpha with their RIT equipment. After the Mayday was called, rescue efforts were ordered with multiple units being assigned to assist with extricating Engine 9. These units were unable to make entry into the structure through the existing access and egress points due to the fire conditions and delay in effective fire suppression. Effective RIT operations are dependent on proactive efforts. Upon arrival, the RIT officer, accompanied by one member of the RIT, will get a report from the incident commander and then should perform an incident scene survey while the remaining RIT members assemble the RIT equipment. During the 360- degree survey, the RIT officer and members should look for ways in and out of the structure, including window configurations, fire escapes, and construction features. The RIT officer should note the feasibility for placement of ground ladders for rescue or escape purposes. The RIT officer has a responsibility to set up and secure a suitable secondary egress for interior crews [Rowett 2018; Toledo Fire & Rescue Department 2012]. After these tasks are completed, the RIT equipment is put in place and the RIT officer informs the incident commander that a 360-degree survey is complete and the RIT is ready to intervene, if necessary. The entire RIT should stay in an area immediately accessible to the building for rapid deployment and maintain radio contact with the incident commander. The RIT officer should brief all RIT members with the results of the incident scene survey [Toledo Fire & Rescue Department 2012]. The RIT officer and members will coordinate with the incident commander to formulate rescue plan contingencies and to monitor radio and fireground conditions. RIT protection is not a passive assignment. This is a process of ongoing information gathering and diligent scene monitoring until the unit is released by the incident commander [NIOSH 2024; NFPA 1407 2020].

Laddering:

The RIT should work to place a ladder to at least one window on each side of the building on every division that firefighters are operating (i.e., laddering). Ground ladders improve the ability to perform vent-enter-search operations to rescue a firefighter as well as the ability of the Mayday firefighter to self- rescue. Once the RIT has determined the need for an egress ladder, it is necessary to remove the window glass, but only after conferring with the incident commander that the removal of the window will not affect firefighting operations. Placing the egress ladder(s) at the window(s) should be announced over the radio by the RIT officer. [NIOSH 2024; Rowett 2018; Toledo Fire & Rescue Department 2012].

Softening:

After identifying existing ways in and out of the structure, the RIT should soften the building by forcing doors and removing devices such as security bars from windows. This proactive tactic ensures the RIT can quickly enter the building in the event of a firefighter emergency and improves the ability of self- rescue for firefighters operating in the interior. The goal of softening is to remove any obstructed building access, but it should be performed in a way that does not introduce an unintended flow path [NIOSH 2024; Rowett 2018].

Page 17 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 Breaching:

In situations where the existing ways in and out of the structure, such as windows and doors, are not accessible due to fire conditions, the RIT should identify areas to enter by breaching walls. The initial survey of the structure includes identification of the construction features of the structure. Along with RIT equipment, the RIT should be able to identify areas to enter the building by breaching walls. The wall breach maneuver is generally taught as a firefighter survival technique to escape deteriorating conditions and the building. These maneuvers can also be employed to enter the structure to affect a rescue during a firefighter emergency [NIOSH 2024; Kiurski 2011]. All necessary actions, to include wall breaches and window conversions, should be executed to gain access to the interior of a structure.

Governing municipalities (federal, state, regional/county, and local) should:

Recommendation #5: Ensure that the delegated zoning or building code enforcement authority shares change of occupancy information with the fire department. Discussion: In 1955, official building records show that the building’s occupancy classification was listed as “Hotel & Apartments.” This is the first recorded documentation indicating that the building was no longer a single-family residence. In 1974, a Certificate of Occupancy was granted for the building to contain 10 individual apartments. These occupancy changes were never shared with the fire prevention bureau, and as a result, it was never included in their inspection program or identified as a multi-family occupancy. During the incident, the fire department believed they were responding to a single-family dwelling. The fire department in this incident maintains a fire prevention bureau that enforces fire codes for the jurisdiction, administers plans reviews, conducts fire inspections, and provides a fire and life safety education program for residents. The bureau utilizes the Hawaii State Fire Code for the jurisdiction which is an adoption of NFPA 1 Fire Code with modifications. The AHJ in this incident enforces zoning, subdivision, building, and housing ordinances through a department of planning and permitting. Through ordinance, this department enforces the 2018 International Building Code (IBC). NFPA 1201, Standard for Providing Fire and Emergency Services to the Public, Section 6.3 recommends that fire departments establish a good working relationship with the agency or authority responsible for enforcing the building code so that the review of the design, construction, alteration, or demolition of buildings and structures can be monitored to identify fire protection concerns. This includes sharing vital information between the entities from inspections, permits, code violations, fire protection maintenance and service, and investigation of fires or other emergencies [NFPA 1201 2020; IFSTA 2016]. This also includes relevant information from change of occupancy permits and approvals. When a change of occupancy is completed, the application of different code requirements may be required [IBC 2018]. Additionally, this can change factors such as the number of occupants being housed, storage of hazardous materials, or other unique hazards that can require significant change in

Page 18 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 fire suppression and rescue tactics and strategy [IFSTA 2016]. This information is critical for fire department pre-incident planning. NFPA 1620 [2020] recommends that pre-incident plans document occupancy type and use including considerations for occupant load, accountability, means of egress, and assistance for people with self-evacuation limitations. When added to computer-aided dispatch, this information can assist incident commanders with formulating an incident action plan [NIOSH 2019]. References DeStefano M [2016]. Big fire, big water; little fire, little fire. Fire Engineering. Gray S and Norwood PJ [2024]. Modern firefighting tactics in residential structures: Ventilation, coordination, and control. Fire Engineering. IBC [2018]. International Building Code. Washington, D.C.: International Code Council. IFSTA [2013]. Essentials of fire fighting. 6th ed. Stillwater, OK: International Fire Service Training Association. IFSTA [2015]. Fire and emergency services safety officer. 2nd ed. Stillwater, OK: International Fire Service Training Association. IFSTA [2016]. Fire inspection and code enforcement. 8th ed. Stillwater, OK: International Fire Service Training Association. Kiurski T [2011]. Training days: Wall breach maneuver. Fire Engineering. NFPA [2019]. NFPA 1730, Standard on organization and deployment of fire prevention inspection and code enforcement, plan review, investigation, and public education operations. Quincy, MA: National Fire Protection Association. NFPA [2020]. NFPA 1201, Standard for providing fire and emergency services to the public. Quincy, MA: National Fire Protection Association. NFPA [2020]. NFPA 1407, Standard for training fire service rapid intervention crews. Quincy, MA: National Fire Protection Association. NFPA [2020]. NFPA 1620, Standard for pre-incident planning. Quincy, MA: National Fire Protection Association. NFPA [2021]. Glossary of terms. Quincy, MA: National Fire Protection Association. NIOSH [2012]. Volunteer fire fighter caught in a rapid fire event during unprotected search, dies after facepiece lens melts – Maryland. By Tarley J, Miles, S, Loflin M, and Merinar T. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2011-02.

Page 19 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 NIOSH [2019]. Career firefighter dies after becoming disoriented in a three-story apartment building – Massachusetts. By Loflin M, Merinar T, and Kline K. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2018-18. NIOSH [2024]. Volunteer firefighter killed after becoming trapped at an assisted living facility fire and two firefighters injured – New York. By Loflin M and Attwood W. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2021-10. Rowett A [2008]. The proactive RIT. Firehouse. Toledo Fire & Rescue Department [2012]. Rapid intervention team (RIT) standard operating procedure C82. Toledo, OH: Toledo Fire & Rescue Department. Weather Underground [2025]. Weather history. The Weather Underground. Date accessed: February 14, 2025. Investigator Information This incident was investigated by Patrick R. Montague, Investigator, with the Fire Fighter Fatality Investigation and Prevention Program, Surveillance and Field Investigations Branch, Division of Safety Research, NIOSH located in Morgantown, WV and James Johnson, Fire Ground Survival Master Instructor and Codes and Standards Representative, International Association of Fire Fighters. This investigation report was authored by Dr. Wesley R. Attwood, Investigator and Program Advisor, with the Fire Fighter Fatality Investigation and Prevention Program, Surveillance and Field Investigations Branch, Division of Safety Research, NIOSH and James Johnson. A subject matter expert review was provided by Karl Fippinger, Vice President, Fire and Disaster Mitigation, International Code Council. Additional Information International Code Council (ICC) The International Code Council is the leading global source of model codes and standards and building safety solutions that include product evaluation, accreditation, technology, training, and certification. The Code Council's codes, standards, and solutions are used to ensure safe, affordable, and sustainable communities and buildings worldwide. Access to codes and standards can be found at https://www.iccsafe.org/products-and-services/codes-standards/. Disclaimer The information in this report is based upon dispatch records, audio recordings, witness statements, and other information that was made available to the National Institute for Occupational Safety and Health (NIOSH). Information gathered from witnesses may be affected by recall bias. The facts, contributing factors, and recommendations contained in this report are based on the totality of the information gathered during the investigation process. This report was prepared after the event occurred, includes

Page 20 One Firefighter Dies During Unprotected Search in Nine-Apartment Building that was Originally Constructed to be a Single-Family House – Hawaii Report # F2025-02 information from appropriate subject matter experts, and is not intended to place blame on those involved in the incident. Mention of any company or product does not constitute endorsement by NIOSH, Centers for Disease Control and Prevention (CDC). In addition, citations to websites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or products. Furthermore, NIOSH is not responsible for the content of these websites. All web addresses referenced in this document were accessible as of the publication date. NIOSH Approved is a certification mark of the U.S. Department of Health and Human Services (HHS) registered in the United States and several international jurisdictions.

FD06035 ｜ 2025_美国_NIOSH_face202503.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2025 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Page i

Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Executive Summary On February 12, 2025, a career firefighter died after being trapped by the collapse of two vacant commercial buildings. At 21:39 hours, Squad 21 (SQ21), Quint 21- 2 (QT212), Quint 21-3 (QT213), Tower 21 (TW21), Engine 21-4 (EN214), and Car 21 were dispatched for reports of a building fire. The fire ambulance (7121) arrived on-scene and provided a size-up confirming a working fire with heavy smoke and visible flames from 1 Main to 3 Main. Once the incident commander (IC) arrived, he requested a 2nd alarm, additional apparatus, fire investigators, safety officer, and utilities be dispatched to the scene. The off-duty chief officers went to the scene when they heard the 2nd alarm requested. On-scene crews began fire suppression operations deploying hoselines around the buildings and evacuating residents from 5 Main while two other firefighters began opening 1 and 3 Main to flow water. At 21:48 hours, IC announced a defensive strategy for operations and ordered all units to exit the buildings. At 21:51 hours, IC reported a major collapse on Side Charlie on talk group 3 and dispatch channels. One firefighter (SQ21 firefighter) was trapped by this collapse under the fully intact facade which spanned from the sidewalk to a street parked car on Side Alpha. For about 19 minutes, multiple crews attempted to extricate the SQ21 firefighter from under the facade. After 20 minutes, two firefighters lifted the facade using spreaders and a combination tool. This provided enough room for crews to pull the SQ21 firefighter out towards the building. The SQ21 firefighter was placed on a stretcher and loaded into 7121 while a fire investigator performed CPR. Fire ambulance 7121 went to the local hospital where the SQ21 firefighter was later pronounced deceased. Photo 1: Side Alpha showing extrication operation, post collapse. (Courtesy of the fire department) REPORT F2025-03 • July 2025

Page ii Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Contributing Factors • Strategy and tactics • Wind-driven fire/structural collapse • Divisions/groups and incident safety officers (ISOs) • Rapid intervention team/crew • Personnel accountability system • Vacant buildings and arson Key Recommendations Fire departments should: • Establish and enforce collapse zones when a defensive strategy begins. • Educate fire officers and firefighters in building performance under fire conditions and the potential for structural collapse. • Establish divisions/groups with a supervisor or use multiple safety officers for fireground management and risk assessment at geographically complex scenes, as early in the incident as possible. • Ensure a rapid intervention team/crew is dedicated, assigned, and in place during structural firefighting operations to immediately respond to a firefighter emergency. • Use a personnel accountability system to identify the location and function of all operating personnel.

Governing municipalities (federal, state, regional/county, and local) should: • Ensure the delegated zoning or code enforcement authority, law enforcement agency, and fire department collaborate to proactively remediate vacant and abandoned buildings to reduce arson.

The National Institute for Occupational Safety and Health (NIOSH) initiated the Fire Fighter Fatality Investigation and Prevention Program to examine deaths of fire fighters in the line of duty so that fire departments, fire fighters, fire service organizations, safety experts and researchers could learn from these incidents. The primary goal of these investigations is for NIOSH to make recommendations to prevent similar occurrences. These NIOSH investigations are intended to reduce or prevent future firefighter deaths and are completely separate from the rulemaking, enforcement, and inspection activities of any other federal or state agency. Under its program, NIOSH investigators interview persons with knowledge of the incident and review available records to develop a description of the conditions and circumstances leading to the deaths in order to provide a context for the agency’s recommendations. The NIOSH summary of these conditions and circumstances in its reports is not intended as a legal statement of facts. This summary, as well as the conclusions and recommendations made by NIOSH, should not be used for the purpose of litigation or the adjudication of any claim. For further information, visit the program at www.cdc.gov/niosh/firefighters/fffipp/ or call 1-800-CDC-INFO (1-800-232-4636).

Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York

Page 1 REPORT F2025-03 • July 2025 Introduction On February 12, 2025, a career firefighter (SQ21 firefighter) died after being trapped by the collapse of two vacant commercial buildings. On February 13, 2025, the U.S. Fire Administration (USFA) notified the National Institute for Occupational Safety and Health (NIOSH) of this incident. On June 8 - 13, 2025, an investigator representing the NIOSH Fire Fighter Fatality Investigation and Prevention Program (FFFIPP) traveled to New York to investigate this incident. The NIOSH investigator conducted interviews with command officers, fire officers, firefighters, fire investigators, and other emergency personnel who were on-scene at the time of the incident. Also, the NIOSH investigator interviewed representatives of the zoning and code enforcement departments, law enforcement agency, and the county communications center. The investigator reviewed fire department standard operating guidelines (SOGs), training records, dispatch records, witness statements, and state agency investigation documents. Fire Department The career fire department in this incident has an 11.5 square mile jurisdiction, serves a population of 47,000 residents, annually responds to an average of 11,000 calls, and provides emergency medical services (EMS). The fire department has 117 uniformed personnel and maintains four shifts: A, B, C, and D. Each shift is staffed by a minimum of 21 personnel across five fire stations. Shifts operate on a 24-hour schedule from 08:00 to 08:00 the following day. Each shift is assigned a shift commander (assistant chiefs). The fire department’s leadership contains a fire chief, fire marshal, deputy chief, three assistant chiefs, a fire training instructor, eight captains, and 20 lieutenants. The fire department has five divisions: • Fire Suppression manages incident response. • EMS provides basic life support first response services and advanced life support transport at the paramedic level. • Special operations work land and water rescue as well as vehicle extrication and hazardous materials response. • Training runs a fire academy that provides initial, advanced, and routine training and certification programs. • Fire prevention leads activities such as fire code enforcement/inspection, fire investigation, and public education.

Page 2 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Training, Education, and Professional Development The fire department maintains a training division that provides initial training and continual skills development for all firefighters, fire officers, and EMS personnel. The training division curriculum goes beyond the New York State required curriculum for career firefighters. Initial training for recruits consists of 18 weeks of daily academy training for a total of 720 hours. Recruits graduate with Pro Board® certifications for NFPA 1010 Fire Fighter I, Fire Fighter II as well as NFPA 470 Hazardous Materials/Weapons of Mass Destruction at the Operations level.

The IC (Car 21) had 25 years of fire service experience at the fire department. He held numerous certifications such as Pro Board® NFPA 1041 Fire Instructor I as well as New York State Basic Firefighter, Officer Level I, Instructor Level I, Municipal Fire Instructor, and EMT-Basic.

SQ21 firefighter had 11 years of fire service with the department. He held numerous certifications such as Pro Board® NFPA 1001 Fire Fighter I and Fire Fighter II as well as New York State paramedic. Apparatus, Staffing, and Communications At 21:39 hours, the following units were dispatched for a building fire (see Table 1):

Table 1. Units dispatched and arrival time Apparatus Staffing Arrival On-Scene Quint 21-2 (QT212) 5 21:42 Car 21 (IC) 1 21:42 Tower 21 (TW21) 4 21:42 Squad 21 (SQ21) 3 21:42 Quint 21-3 (QT213) 3 21:43 Engine 21-4 (EN214) 3 21:46

At the time of the incident, the county Public Safety Answering Point (PSAP) dispatched for seven EMS agencies, 33 fire departments, and eight law enforcement agencies. The PSAP was staffed by a supervisor and nine dispatchers per shift. The county used a UHF digitally trunked radio system with four designated dispatch talk groups (channels) for fire and EMS, one for each of the county’s four “battalions.” The fire department in this incident was dispatched on one talk group with four other fire departments that were assigned battalion 2. Tone alerting for all fire and EMS was handled on a separate paging talk group. Per PSAP protocols, all fireground communications were conducted on an operations talk group that was assigned by the PSAP. Like all talk groups on the system, the operations talk group was repeated and recorded. The PSAP had a primary dispatcher monitoring the dispatch talk group for every working fire. The operations talk group was monitored, when possible, in case of emergencies such as Maydays and “urgents.” Each fire officer and firefighter was given one portable radio with scan function.

All portable radios in the county had a programmed emergency alert button (EAB) to use when a firefighter was in immediate danger. The county’s protocol for EAB noted that upon activation, the portable radio

Page 3 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 would automatically switch to the emergency county talk group channel and result in an open mic for the firefighter to say their name, emergency, and location. Building Construction This incident involved two commercial buildings that were the end of a row next to a river, 1 and 3 Main (see Photo 2). Official zoning department records indicated that 3 Main was built around 1885 and was about 25 ft x 75 ft, with 1,875 square feet on the first floor, and another 1,875 square feet in the basement. It was wood frame construction with clapboard siding on Side Bravo and Side Charlie. It had a flat roof and a small parapet wall. The roof and floor joists were pocketed into the adjacent brick building, 5 Main (see Photo 3). The original post-in-ground foundation was replaced in 1949 with steel beams and concrete piers. The building originally was used for the commercial manufacturing of carriages. Over the years, it was remodeled and the storefront had two entrances on Side Alpha. There was also a stone facade, supported by a wood frame, that was added before 1992. In the 1940’s, three billboards were installed on the roof which spanned the length of the building. The billboards were present at the time of the incident.

Historical records indicated that 1 Main was built in 1938 as a wood frame addition to 3 Main to serve as a shoeshine shop. There was a shared doorway between both buildings. It was 24 ft x 21 ft, with 504 square feet on the first floor. 1 Main also had a small basement with steel beams and concrete piers as a foundation. It was remodeled with its storefront also having two entrances on Side Alpha. Both buildings had a sidewalk on Side Alpha and on-street parking. At the time of the incident, both 1 and 3 Main had most of their doors and storefront windows on Side Alpha boarded up with plywood.

Photo 2. The row of buildings in the incident in order from left to right: 1, 3, 5, 7, 9, and 11 Main. Red box shows 1 and 3 Main. (Courtesy of the fire department)

Page 4 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Occupancy Status At the time of the incident, both 1 and 3 Main were registered as vacant buildings. The authority having jurisdiction (AHJ) maintains a vacant property registration program. Property owners annually register both residential and commercial properties as vacant with the AHJ if they are unoccupied for more than four months or in foreclosure. This program allows the AHJ to track vacant properties but assigns responsibility to the property owner to ensure the buildings are secured and maintained per New York State code. The program requires an exterior visual inspection of each vacant property by a code enforcement representative annually to ensure these responsibilities are met. When an inspection finds that a vacant building is unsecured, a certified letter is sent to the property owner that instructs them to secure the property within seven days. A code enforcement representative re-inspects the building after seven days to verify the building is secured. If it is not, a work order is entered for the AHJ’s parks department to board up and secure the building.

Inspections of 1 and 3 Main

Visual exterior inspections of both 1 and 3 Main were conducted in 2022 by code enforcement. 1 and 3 Main received violation notices for trash accumulation, other property maintenance issues, and the Photo 3. The pockets that were built into 5 Main that supported roof and floor joists of 3 Main. The image was taken post-fire. (Courtesy of the fire department)

Page 5 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 buildings being unsecured (see Photo 4). Per New York State code, the fire prevention division performed fire and life safety inspections once every three years for the type of commercial occupancy at 1 and 3 Main.

The last inspection for 1 Main was conducted in 2022 and results indicated that the building was unoccupied with structural issues observed, such as holes in the floor. Fire prevention division records showed that the building was last occupied in 2011 when it was used as a clothing retail store.

A joint inspection of 3 Main was performed on February 8, 2023, by code enforcement and fire prevention representatives when it was found to be unsecured. The roof joists and stairway of 3 Main were observed to be pulling out of the pockets of the brick wall of 5 Main with some of the joists being unsupported. Fire prevention division records showed that the building was last occupied in 2008 as a barber shop. Pre-Incident Hazard Identification The fire prevention division informed fire department personnel about the structural issue documented in 2023 by placing a “do not enter” electronic hazard notice for both 1 and 3 Main in the emergency response dispatch software. This software displayed a hazard marker on the map for the buildings to notify responding units of the hazard. The notes for this hazard notice indicated “unstable structure/floor.” Additionally, the fire prevention division communicated this hazard to shift personnel for awareness. At the time of the incident, the responding units were aware of the documented structural issues observed in 2023, informing firefighters to not enter either building for fire suppression operations. Incident Timeline The following timeline is a summary of events that occurred as the incident evolved on February 12,

1. Not all incident events are included in this timeline. The times are to the minute and were taken

from the fire departments’ National Fire Incident Reporting System (NFIRS) fire reports, dispatch log, and interview notes. Photo 4. Side Charlie of 3 Main in 2022. (Courtesy of the fire department)

Page 6 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Time Fireground Operations, Response, and Details February 12, 2025

21:37 Hours • PSAP received a 9-1-1 call for a fire in an abandoned building. 21:39 Hours • PSAP transmitted a full alarm assignment for a building fire. SQ21, QT212, QT213, TW21, EN214, and Car 21 were dispatched with operations assigned to talk group 3. 21:40 Hours • 7121 was in the area during dispatch. Upon hearing the PSAP, 7121 stopped at scene and provided a size-up confirming a working fire with heavy smoke and visible flames from 1 Main with extension to 3 Main. They requested QT212 respond to Side Charlie on arrival. 21:42 Hours • QT212, Car 21 (IC), TW21, and SQ21 arrived on-scene. • QT212 reported heavy smoke in the area, and believed the fire had possibly extended farther than 3 Main. QT212 officer went to Side Charlie and reported heavy fire in the basement and first floor of 1 Main and that he believed both buildings were vacant. • Car 21 established himself as IC and parked at the end of the row opposite of the bridge. He requested a 2nd alarm be transmitted and for spare apparatus to be staffed. He also requested fire investigators, a safety officer (SO21), and utility companies to be dispatched to the scene. • The off-duty assistant chiefs (CH21A, CH21B, and CH21D) went to the scene when they heard the 2nd alarm requested. • TW21 parked at the end of the bridge in anticipation of a defensive operation. • SQ21 pulled past TW21 and staged in the road in front of 3 and 5 Main. • SQ21 officer and TW21 officer had a face-to-face with the IC who ordered no entry into the buildings. He requested their crews to begin opening the doors and windows of the buildings for suppression. • SQ21 officer, backed up by SQ21 firefighter, prepared to flow water into 3 Main using a hoseline from SQ21. 21:43 Hours • The PSAP dispatched fire departments 26, 32, 35, and 59 to standby in their stations. 21:44 Hours • QT213 arrived on-scene and IC requested they deploy a hoseline to the Side Charlie/Side Delta corner for exposure protection. 21:45 Hours • 7121 officer returned to his apparatus and radioed the PSAP that he was proceeding to the local hospital with a patient they picked up prior to the incident. 21:47 Hours • IC requested the PSAP dispatch a mutual aid truck and engine to the scene as well as a 3rd alarm to staff spare apparatus. The PSAP dispatched Truck 59 and Engine 26-1.

Page 7 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Time Fireground Operations, Response, and Details • EN214 arrived on-scene and prepared to establish a water supply from a hydrant to QT213. 21:48 Hours • IC announced a defensive strategy for operations and ordered all units to exit the buildings. He requested all apparatus driver/operators on-scene to sound their air horns to reinforce the change to a defensive strategy. He made the announcement first on talk group 3 then the dispatch channel. • QT212 officer immediately responded reporting that all occupants were evacuated from an adjacent building in the row. • The PSAP dispatched the 3rd alarm units. 21:51 Hours • IC reported a major collapse on Side Charlie on both talk group 3 and dispatch channels. • SQ21 firefighter was trapped under the fully intact collapsed facade on Side Alpha which spanned the sidewalk to a street parked car. • The Side Charlie collapse was confirmed by QT212 officer and was followed by an announcement from IC that a parapet had collapsed on Side Alpha. • Immediately following the collapse, IC requested a personnel accountability report (PAR) starting with QT212. 21:52 Hours • IC reported to the PSAP that there was a firefighter down on Side Alpha. 21:53 Hours • A Mayday was called with the PSAP sounding the emergency tones. • A QT212 firefighter at the collapsed facade activated their EAB. • IC requested all manpower to Side Alpha. • Multiple on-scene units were assigned to the extrication operation. 21:55 Hours • IC called an “urgent” on talk group 3 requesting all available manpower to Side Alpha to assist with the downed firefighter. 21:56 Hours • PSAP requested confirmation that either Chief 21 (CH21) or the IC received an additional Mayday (EAB activation) from QT212. • IC assigned CH21B to resolve the additional Mayday. 21:58 Hours • 7121 arrived back on-scene from the local hospital and was requested by IC to report to Side Alpha to assist with the downed firefighter. • A firefighter radioed on talk group 3 that they needed a line on the fire on Side Alpha. • IC and a QT212 firefighter grabbed hoselines and TW21 directed its master streams to the fire of 3 Main to protect the extrication operations. 21:59 – 22:02 Hours • CH21B advised the PSAP that the Mayday was for the firefighter that was trapped by the collapse that they are working to extricate. • IC requested PAR from QT212 who responded that they were working on the down firefighter. • PSAP sounded emergency tones stating there was a second Mayday from QT212 (EAB activation).

Page 8 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Personal Protective Equipment At the time of the incident, SQ21 firefighter was wearing full structural firefighting turnout gear and a NIOSH Approved® self-contained breathing apparatus (SCBA). No evidence was identified to suggest that the structural firefighting turnout gear or SCBA contributed to the fatality. Weather Conditions At 21:39 hours on February 12, 2025, the outdoor temperature was 25°F, dewpoint was 16°F, the wind was out of the southeast at 16 mph, and conditions included light snow [Weather Underground 2025]. Time Fireground Operations, Response, and Details • IC again requested PAR from QT212 who responded they were working on the down firefighter. 22:03 Hours • IC requested a status from QT213 on Side Charlie who responded that the fire was somewhat knocked down and they were getting ready to open 5 Main to look for fire extension. 22:12 Hours • IC notified the PSAP that QT212 had PAR, and their Mayday was for the member of SQ21 that they were working to extricate. • The facade was lifted by the SQ21 driver/operator using spreaders and a QT212 firefighter using a combination tool. This provided room for crews to pull SQ21 firefighter out towards the building. 22:13 Hours • CH21A notified IC that SQ21 firefighter was fully extricated from under the facade. The PSAP acknowledged that extrication was complete. • SQ21 firefighter was placed on a stretcher and loaded into 7121 while fire investigator 21-2 (FI212) performed CPR and 7121 crew provided advanced life support care. • 7121 departed the scene to the local hospital where SQ21 firefighter was later pronounced deceased. 22:15 – 22:18 Hours • IC requested a progress report from TW21 who responded that there was still heavy fire because of the wind. QT213 reported that they were holding the fire on Side Charlie. • IC initiated a PAR check for all on-scene units with all units confirming PAR. February 13, 2025

00:43 Hours • The fire was declared under control with on-scene units working on hot spots. 16:59 Hours • All units cleared the scene.

Page 9 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Investigation At approximately 21:37 hours, the county PSAP received a 9-1-1 call from an occupant of a nearby residence reporting a fire in an abandoned building. At 21:39 hours, the PSAP transmitted a full alarm assignment for the reported building fire around the intersection at Main and Front. SQ21, QT212, QT213, TW21, EN214, and Car 21 were dispatched and assigned talk group 3. While units were enroute, crews in both SQ21 and TW21 discussed their knowledge of the deteriorated conditions of 1 and 3 Main. Officers of each apparatus informed their crews that no one enter the buildings.

At the time of dispatch, 7121 was transporting a patient to the local hospital and just entered the bridge headed towards the front of 1 Main. 7121 parked on the side of the road after clearing the bridge to investigate and provided a scene size-up. The 7121 officer walked up to 1 Main and observed smoke coming out of the boarded up windows on Side Alpha. The smoke billowed out of the boarded windows as if the building was charged with smoke. He walked onto the bridge and observed flames breaking through Side Bravo of 1 Main and heavy smoke coming from Side Charlie of 3 Main. The 7121 officer provided a detailed size-up of his observations to the PSAP that confirmed the working fire in 1 Main extended into 3 Main. He requested QT212 respond to Side Charlie on arrival. As he returned to Side Alpha, dark grey smoke from 1 Main became so heavy and banked down that he could no longer see his apparatus parked across the street (see Photo 5). The severe smoke conditions were caused by strong winds coming off the river. These winds blew into both buildings, advancing the situation to a wind- driven fire.

Photo 5. Example of smoke condition on Side Alpha. (Courtesy of the fire department)

Page 10 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 At 21:42 hours, QT212, Car 21 (IC), TW21, and SQ21 arrived on-scene. QT212 parked at the intersection at the end of the row opposite the bridge and reported heavy smoke in the area, believing the fire had possibly extended farther than 3 Main. Their crews received reports of elderly occupants that needed assistance to evacuate the buildings next to 3 Main. The crews broke into two teams. The QT212 officer went to Side Charlie and assessed conditions. He reported heavy fire in the basement and first floor of 1 Main (see Photo 6). He also reported that he believed both buildings were vacant and requested hoselines to Side Charlie. The other QT212 crew assisted police officers with evacuating the occupants of the other buildings in the row.

Car 21 established himself as the IC and parked at the end of the row opposite the bridge. He requested a 2nd alarm be transmitted and for spare apparatus to be staffed. He also requested fire investigators, SO21, and utility companies to be dispatched to the scene. He walked along the row to the bridge and assessed conditions on Side Alpha of each building. When he was at the bridge, he noted that the buildings were hard to see because of the amount of smoke present. The off- duty assistant chiefs went to the scene when they heard the 2nd alarm requested.

Seeing the heavy smoke, the TW21 officer had the driver/operator park at the end of the bridge in anticipation of a defensive operation. SQ21 pulled past TW21 and staged in the road in front of 3 and 5 Main. The SQ21 officer and the TW21 officer had a face-to-face with IC who ordered no entry into the buildings. He requested their crews begin opening the doors and windows of the buildings for fire suppression. TW21 crews were unable to remove the plywood of 1 Main with hand tools. A TW21 firefighter retrieved a chainsaw and began cutting holes in the plywood. The SQ21 officer observed police officers trying to open 3 Main. He started removing the plywood on the windows and doors with hand tools while the SQ21 driver/operator and the SQ21 firefighter stretched a 1 ¾-inch hoseline around the parked cars to the front of 3 Main. The TW21 firefighter cut a hole in the plywood the SQ21 officer was working on. The SQ21 officer, backed up by the SQ21 firefighter, began flowing water into 3 Main. The SQ21 officer did not observe any active fire in 3 Main but believed the fire was in the basement. He continued to flow water, periodically stopping to evaluate conditions.

The PSAP dispatched fire departments 26, 32, 35, and 59 to standby in their stations. IC radioed EN214 and asked if they were on-scene yet. When QT213 arrived on-scene, IC requested they deploy a hoseline Photo 6. Conditions on Side Charlie upon arrival of QT212. (Courtesy of the fire department)

Page 11 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 to the Side Charlie/Side Delta corner for exposure protection. QT213 crews deployed 300 feet of 1 ¾- inch hoseline to Side Charlie of 3 Main which was fully involved. The QT213 officer had crews deploy an additional 2 ½-inch hoseline after seeing the volume of fire. The QT212 officer reported on talk group 3 that his crew was evacuating occupants from 5 Main out through the exterior stairs of Side Charlie. EN214 radioed IC stating they were almost on-scene and asked if he wanted them to establish a water supply. IC responded requesting their location and gave orders for EN214 to establish a water supply to QT213. The 7121 officer returned to his apparatus and radioed the PSAP that he was proceeding to the local hospital with their patient.

IC requested the PSAP to dispatch a mutual aid truck and engine to the scene and a 3rd alarm. He requested the QT212 driver/operator establish a water supply to TW21. As QT213 performed fire suppression on Side Charlie of 3 Main, EN214 arrived on-scene and prepared to establish a water supply from a hydrant. While at the same time, a QT212 firefighter deployed 300 feet of 1 ¾-inch hoseline to Side Alpha from QT212.

At 21:48 hours, IC announced a defensive strategy for operations and ordered all units to exit the buildings. He requested all apparatus driver/operators on-scene to sound their air horns to reinforce the change to a defensive strategy. He made the defensive strategy announcement first on talk group 3 then on the dispatch channel. The QT212 officer immediately responded reporting that all occupants had been evacuated from an adjacent building in the row. The PSAP dispatched the 3rd alarm units. Hearing the order from IC and the QT212 officer’s response, the SQ21 officer and other firefighters on Side Alpha assumed the order was for interior crews that were in adjacent buildings evacuating occupants. IC requested TW21 crews set up the aerial ladder as the SQ21 driver/operator began to deploy a 2 ½-inch hoseline.

IC requested QT213 and EN214 to deploy a blitz gun to Side Charlie to protect exposures. EN214 secured the water supply to QT213 while firefighters from QT212 and QT213 added EN214’s blitz gun to the deployed 2 ½-inch hoseline. CH21A and SO21 arrived on-scene. IC radioed on the dispatch channel asking if any chief officers were on-scene. CH21A responded that he was on-scene and IC ordered him to Side Charlie to manage the exposure protection operations. SO21 had a face-to-face with IC who requested he also stage on Side Charlie. The QT212 driver/operator completed the water supply to TW21 who began operating two master streams from the bucket into the fire of 1 Main. Fire ambulance 7121 was requested to return to the scene as they left the local hospital. At this time, the SQ21 officer told the SQ21 firefighter that their suppression operations weren’t working and they needed to regroup. The SQ21 firefighter began to walk away from the front of 3 Main towards Side Delta. The SQ21 officer then dropped the 1 ¾-inch hoseline and followed.

At 21:51 hours, IC reported a major collapse on Side Charlie on both talk group 3 and dispatch channels. 3 Main experienced a progressive collapse that started when the roof failed and caused the billboards to drop (see Photo 7 and 8). This caused the Side Bravo wall to fall outward and push the walls of 1 Main. This also caused the facade on Side Alpha to collapse outwards onto the sidewalk toward the street parked cars. The SQ21 firefighter was trapped under the fully intact facade which spanned the sidewalk to a street parked car. His head to his diaphragm were covered by building materials. His waist, one arm, and legs were exposed through a window frame. The SQ21 officer was hit in the head by the facade,

Page 12 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 knocking off his helmet. The collapse caused zero visibility conditions on Side Alpha; some smoke conditions occasionally improved because of shifting winds. The Side Charlie collapse was confirmed by the QT212 officer and was followed by an announcement from IC that a parapet had collapsed on Side Alpha. Immediately following the collapse, IC requested a PAR starting with QT212. IC did not witness the facade collapse onto the SQ21 firefighters because of the severe smoke conditions.

CH21D arrived on-scene. A TW21 firefighter and the SQ21 driver/operator notified IC that the SQ21 firefighter was trapped under the facade. IC reported to the PSAP that there was a firefighter down on Side Alpha. The PSAP requested confirmation from IC that there was a firefighter down on Side Alpha. This transmission was followed by a Mayday with the PSAP sounding the emergency tones. The PSAP attempted to contact the Mayday firefighter twice with no response. Since the firefighter calling the Mayday had their transmissions being stepped on, a QT212 firefighter at the collapsed facade activated their EAB. IC then requested all manpower to Side Alpha and QT212 crews from Side Charlie made their way to Side Alpha through a building in the row. CH21A was assigned to lead the extrication operation. As EN211 arrived on-scene they reported to IC, who requested they assist with extricating the SQ21 firefighter. Along with crews from QT212 and SQ21, they began the operation by attempting to lift the facade, but it was too heavy. Crews also attempted to pull SQ21 firefighter by his lower body which was exposed through the window frame. This was also unsuccessful.

Photo 7 and 8. Progressive collapse of 3 Main showing the billboard drop and pushing out the Side Bravo wall. (Courtesy of the fire department)

Page 13 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 7121 went to the scene from the local hospital as FI212 arrived on-scene. IC requested that EMS bring a stretcher to Side Alpha. A QT213 firefighter reported to IC that the “single-story” building only had part of the Side Alpha and Side Bravo walls left standing. This was followed by IC calling an “urgent” on talk group 3 and requesting all available manpower to Side Alpha to assist with the downed firefighter. The SQ21 officer responded requesting a saw to Side Alpha for crews working the extrication to cut the facade frame apart (see Photo 9).

The PSAP requested confirmation that either CH21 or the IC received an additional Mayday (EAB activation) from QT212. CH21 responded noting that he was not on-scene and could not confirm. IC assigned CH21B to resolve the additional Mayday. He began to search for QT212 officer to verify they had PAR. When 7121 arrived on-scene, IC requested they report to Side Alpha to assist with the downed firefighter. Upon acknowledging their arrival report, the PSAP requested that 7121 confirm with IC that they were aware of the Mayday from QT212. IC responded by requesting a PAR from QT212 as the 7121 officer staged a stretcher by the collapsed facade on the street side of the parked cars. A firefighter radioed on talk group 3 that they needed a line on the fire on Side Alpha (see Photo 10). Following the collapse, the wind began to blow direct heavy smoke, heat, and fire over the edge of the sidewalk towards the facade and extrication operations. IC and a QT212 firefighter grabbed hoselines and TW21 directed its master streams to the fire of 3 Main to protect the extrication operations. The EN214 driver/operator established a water supply to SQ21 which was running low on tank water. Crews switched out operating the hoselines to protect the extrication operation from smoke, heat, and flames.

Photo 9. The collapsed facade on Side Alpha. (Courtesy of the fire department)

Page 14 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03

CH21B advised the PSAP that the Mayday was for the firefighter trapped by the facade, who they were working to extricate. The PSAP acknowledged while IC requested PAR from QT212. QT212 responded that they were working on the downed firefighter. A firefighter requested water on the fire on Side Alpha. This transmission was followed by emergency tones from the PSAP about a second Mayday from QT212 (EAB activation). The PSAP requested confirmation from IC that they were aware of the second Mayday. CH21B confirmed as fire investigator 21-3 (FI213) arrived on-scene. Crews began to use chainsaws to cut the wood frame of the facade. One chainsaw kept stalling from the amount of smoke while the other would not cut the material of the wall. Again, IC requested PAR from QT212 who responded they were working on the downed firefighter. IC requested confirmation that QT212 knew where all its firefighters were. IC requested a status from QT213 on Side Charlie who responded that the fire was somewhat knocked down and they were getting ready to open 5 Main to look for fire extension. IC requested TW21 stop their master streams as they were working against the extrication operations.

EN219 and SQ219 arrived on-scene and were assigned to assist with the extrication operations by staging tools. IC notified the PSAP that QT212 had PAR, and their Mayday was for the member of SQ21 that they were working to extricate. The SQ21 driver/operator used spreaders and a QT212 firefighter used a combination tool to lift the facade together. This provided enough room for crews to pull the SQ21 firefighter out towards the building. At 22:13 hours, CH21A notified IC that the SQ21 firefighter was fully extricated from under the facade. The PSAP acknowledged that extrication was complete. The SQ21 firefighter was placed on a stretcher and loaded into 7121 while FI212 performed CPR and the 7121 crew provided advanced life support care. IC requested the PSAP to notify the local Photo 10. View of fire that became wind-driven from bridge with master streams and hoseline attempting to protect the extrication operation. (Courtesy of the fire department)

Page 15 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 hospital of the SQ21 firefighter being transported to them. 7121 went to the local hospital where the SQ21 firefighter was later pronounced deceased.

IC requested a progress report from TW21 who responded that there was still heavy fire because of the wind. QT213 reported that they were holding the fire on Side Charlie. IC then initiated a PAR check for all on-scene units with all confirming PAR. At this time, IC was transferred to CH21D, and on-scene units were rotated off. Multiple firefighters were treated for smoke inhalation because they worked the extrication operation without an SCBA.

The next day, the fire was declared under control with on-scene units still working on hot spots. All units cleared the scene that afternoon. Three firefighters were evaluated at a local hospital for smoke inhalation, an overexertion injury, and head injuries.

The fire department created a peer support program three years ago. After the incident, the fire department activated their internal peer support team for those who responded to this incident. The availability of the peer support team and its mental health resources were communicated to the department personnel via email, text message, word-of-mouth, and through the supervisory chain. The resources were also available to spouses and children of responding personnel. Fire Origin and Cause The State cause and origin investigation determined that the fire started in 1 Main and was classified as arson. The fire was believed to have been started by an individual experiencing homelessness. Cause of Death According to the county coroner report, the cause of death of the SQ21 firefighter was traumatic asphyxiation. Contributing Factors Occupational injuries and fatalities are often the result of one or more contributing factors or key events in a larger sequence of events that ultimately result in injuries or fatalities. The NIOSH investigator identified the following items as key contributing factors in this incident that ultimately led to the fatality: • Strategy and tactics • Wind-driven fire/structural collapse • Divisions/groups and ISOs • Rapid intervention team/crew • Personnel accountability system • Vacant buildings and arson

Page 16 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Recommendations Recommendation #1: Fire departments should establish and enforce collapse zones when a defensive strategy begins. In this incident, the IC announced a defensive strategy for operations and ordered all units to exit the buildings three minutes before the collapse. He requested all apparatus driver/operators on-scene sound their air horns to reinforce the change to a defensive strategy (i.e., exterior fire control). The QT212 officer immediately responded, reporting that all occupants had been evacuated from an adjacent building in the row. Hearing the order from IC and the QT212 officer’s response, the SQ21 officer and other firefighters interviewed by NIOSH assumed the order was only for interior crews in adjacent buildings who were evacuating occupants. A collapse zone was not established or enforced by the IC, company officers, or ISO prior to the collapse.

In a defensive operation, the IC should make a reasonable determination of what the fire may consume and where it can be stopped. The IC should acknowledge that if the fire has significantly strengthened, some or all of the building may be lost. This determination should be based on the fire department’s capability to extinguish the fire without placing firefighters in unnecessary risk. Conducting a thorough scene size-up and deploying effective tactics on what can be saved and where the fire can be stopped minimizes losses [NIOSH 2025a]. NFPA 1700 states that abandoned or vacant structures are often in an unknown state of condition or compromise, which could result in weakened structural components, holes in floors, and structural deficiencies. When controlling fires in these structures, a defensive strategy should be considered to avoid entry, and early collapse should be anticipated. Structural deficiencies in these buildings may result in unpredictable and increased fire activity [NFPA 1700 2021]. A decision for a defensive strategy is based on the incident’s hazards outweighing the ability to safely operate inside the structure. Recognizing and defining collapse signals and risk through exclusion and control zones are fundamental to collapse zone management principles [NIOSH 2014]. Clear announcement of a defensive strategy is crucial as it helps all responders understand the IC’s goal (control the spread from the outside) [NIOSH 2025a]. When the structure is visibly unstable, a collapse zone equal to one and a half times the height of the building should be established at minimum [NFPA 1550 2024]. This perimeter size keeps firefighters and other personnel out of imminent danger in the event of a collapse. When a collapse zone is established, the IC should communicate a “no re-entry” strategy until otherwise directed [NIOSH 2008]. The IC, with the help of the ISO(s), should define, establish, and enforce collapse areas. Once in place, the IC must clearly communicate where the collapse area(s) is/are to all firefighters operating on the fireground. The IC, division/group supervisors, and company officers should ensure firefighters do not enter the collapse areas. Preventing deviance from the strategy requires communication with everyone on the fireground, including the dispatcher and other responding units [NIOSH 2025b]. Recommendation #2: Fire departments should educate fire officers and firefighters in building performance under fire conditions and the potential for structural collapse. At this incident, the majority of responding personnel knew that the condition of the incident buildings were vacant and deteriorated. Some crews decided while enroute that no one would enter them for

Page 17 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 interior operations. Recognizable characteristics of a wind-driven fire were present based on smoke conditions and fire behavior. Due to their single-story structure, many of the firefighters interviewed by NIOSH expressed they did not predict the buildings to fail so quickly.

Understanding a building’s design and structural anatomy, construction methods and materials, and vulnerabilities under fireground conditions has been linked to safer firefighting operations and firefighter survivability [NIOSH 2013]. Knowledge of building construction is critical to help firefighters recognize the potential for structural collapse. During the growth stage, the fire consumes combustible structural members. During the decay stage and post-suppression activities, the structure is further weakened due to the poor state of remaining structural members and the buildup of water. The contents of a building, such as furniture or machinery, also contribute to the potential for structural collapse by [IFSTA 2015; NIOSH 2025b]: • Adding fuel load into the structure and generating higher temperatures that weaken structural components. • Adding weight to the weakened structural members. • Retaining water, which increases weight of contents and applies more stress on the structural members. The progressive collapse of 3 Main resulted in the stone facade on Side Alpha falling straight out as a unit. Exterior masonry and veneer walls are not load-bearing. They are attached with ties and supports to the structure [IFSTA 2013]. When the load-bearing walls of 3 Main failed due to the collapse, the weight of and imposed load from the stone on the facade pulled it straight out onto the sidewalk. It is important for firefighters to be aware of how masonry or stone veneer construction influence the potential for structural collapse. The term “wind-driven” fire is used to describe a fire in which the wind has the potential to, or is already causing, a dramatic and sudden increase in fire, heat, and smoke conditions. When responding to a reported structure fire, an overriding consideration concerning size-up must be wind conditions and their potential effect on the fire. The key to successfully operating at wind-driven fires depends on recognizing the wind-impacted fire conditions that may change a seemingly routine fire into a “blowtorching” fire [NIOSH 2016]. At this incident, the poor structural condition of the buildings, including multiple openings paired with the wind from the Side Bravo/Side Charlie corner facilitated rapid fire growth. This led to a quicker structural failure of the buildings. Fire officers and firefighters should be trained to consider building performance under fire conditions, such as wind-driven fires, to better recognize the potential for structural collapse and adjust tactics as applicable. Recommendation #3: Fire departments should establish divisions/groups with a supervisor or use multiple safety officers for fireground management and risk assessment at geographically complex scenes, as early in the incident as possible. This incident involved a geographically complex scene with a row of buildings near a river. The fire department in this incident used the training division instructor (SO21) as the designated ISO. The ISO

Page 18 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 only responds when requested by the IC. At the time of this incident, SO21 was off-duty and responded from home. He arrived on-scene two minutes before the collapse. As the incident escalates and additional alarms are requested, the span of control must be expanded to maintain effective command and coordination of the incident. This tactical-level management can be supported by establishing divisions or groups to direct operations in specific geographic areas or manage incident functions [IFSTA 2015]. Within a division/group, firefighters advise their supervisor of work progress and provide accountability for crew members engaging in task level activities. The IC should assign divisions/groups to a supervisor early. This is especially important when firefighters are operating from tactical positions that the IC has little or no direct control over (e.g., out of sight). All requests for additional resources or assistance within a division/group are directed to the supervisor who is responsible for communicating with the IC. Supervisors can provide ongoing conditions, actions, needs (CAN) reports to the IC of all four sides and the interior of an incident which may influence tactics and strategy [SKCFTC 2023]. Division/group supervisors can also assist in providing PARs when requested by the IC, ISO, or operations. Tactical-level management at geographically complex scenes can also be accomplished by using multiple safety officers. The ISO provides a fire department with a higher level of expertise to perform the necessary incident scene functions and assist the IC with fireground safety. An ISO can perform initial and ongoing size-ups throughout the incident. Expectations and authority for the ISO include determining hazardous incident conditions, advising the IC to modify control zones or tactics to address corresponding hazards, communicating fire behavior and forecasting growth, and estimating building/structural collapse hazards. The ISO also has the authority to stop or suspend incident operations based on imminent threats to firefighter safety [NFPA 1550 2024]. The ISO should be separate from the IC, operations, or accountability positions so they can focus on their responsibilities and the primary objective of continually assessing all on-scene hazards to firefighter life and safety [NIOSH 2025b].

Larger fire departments can consider having one or more full-time dedicated ISOs who are on duty and can routinely respond to working fires (e.g., full-time shift ISOs). This can include increasing staffing to fulfill this role. In smaller departments, every officer should have the ability to function as the ISO when assigned by the IC [NIOSH 2025c]. Whether the ISO position is predesignated or filled by a fire officer, ISOs should be trained in how to assist the IC and other officers in fireground operations. While a fire department may use an appointed officer as an ISO, they may be delayed in recognizing a hazardous situation or operation. This delay can be mitigated by training all individuals who may be appointed as an ISO, to ensure they have clear understanding of responsibilities and expectations if deemed necessary [Dodson 2021; Sullivan 2012; NIOSH 2013]. Specific to this incident, the IC might have considered the use of multiple or assistant ISOs for fireground management and risk assessment at such a geographically complex scene. Incidents at large geographical areas such as commercial buildings or scenes with access issues that hamper the ability to do a 360 (e.g., riverside, row of buildings) often require expansion of the ISO role. Consequently, assigning assistant ISOs into branches, divisions, or groups for specific areas of responsibility (e.g. Side Charlie) might be necessary [NIOSH 2016].

Page 19 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Recommendation #4: Fire departments should ensure a rapid intervention team/crew is dedicated, assigned, and in place during structural firefighting operations to immediately respond to a firefighter emergency. The fire department in this incident used the 7121 crew as their assigned rapid intervention team (RIT) for working fires. The 7121 crew would collect the RIT equipment stored in Car 21 and stage it at the scene. At this incident, 7121 was the first arriving unit. On arrival, the IC observed 7121 on-scene and believed those crews were there for RIT assignment. However, 7121 departed the scene for the local hospital with the patient they had from a previous call for service. No RIT was present on-scene at the time of the collapse or extrication operation. Effective RIT operations are dependent on being proactive. On arrival, the RIT officer with one member of the RIT, receives a report from the IC. This report informs the RIT of on-scene details before they perform an incident scene survey to complete their evaluation. The remaining RIT members assemble the RIT equipment. After these tasks are completed and the RIT equipment is in place, the RIT officer informs the IC that a scene survey is complete and the RIT is ready, if needed. The entire RIT should stay immediately accessible for rapid deployment and maintain radio contact with the IC. The RIT officer should brief all RIT members with results of the incident scene survey [Toledo Fire & Rescue Department 2012]. The RIT officer and members coordinate with the IC to form rescue plan contingencies and to monitor radio and fireground conditions. RIT protection is an active assignment. This is an ongoing process of comprehensive information gathering and diligent scene monitoring until the unit is released by the IC [NIOSH 2024; NIOSH 2025d; NFPA 1407 2020]. RIT personnel should be qualified with operations level training of structural collapse search and rescue, as established in NFPA 2500. This includes being able to perform tasks such as conducting hasty primary and secondary search operations (low and high coverage) to locate victims trapped on, inside, and beneath collapse debris; accessing victims trapped inside and beneath collapse debris; and performing extrication operations involving packaging, treating, and removing victims trapped within and beneath collapse debris [IFSTA 2019; NFPA 2500 2022]. In terms of this incident, the RIT equipment was not staged and ready for use during the extrication operation, this includes appropriate tools based upon conditions present. RIT equipment not only includes lifesaving resources such as spare SCBA cylinders within RIT kits, but also forcible entry, cutting, and striking tools. This may include battery-powered chainsaws that are unaffected by the smoke conditions present on Side Alpha when crews attempted to cut the facade [IFSTA 2019]. Recommendation #5: Fire departments should use a personnel accountability system to readily identify the location and function of all operating personnel. At this incident, the IC was not able to maintain accountability of on-scene personnel because of smoke conditions on Side Alpha and the geographically complex scene. Additionally, the multiple Maydays and PAR attempts for QT212 could not be resolved in a timely manner. A personnel accountability system is a system that identifies the location and function of all members operating at an incident scene [NFPA 1550 2024]. This system is activated during an incident to collect

Page 20 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 and maintain the status and location of personnel that may be working in an immediately dangerous to life and health environment. All personnel operating at an incident are responsible for understanding and participating in this system. The IC is responsible for all personnel but may delegate certain responsibilities to another person such as the ISO. An integral part of the accountability system is to make sure that the firefighters who are assigned and operating in the hazard zone are accounted for throughout the entire incident. A properly initiated and enforced personnel accountability system can improve firefighter safety and survival [NIOSH 2024]. A functional personnel accountability system can identify [NIOSH 2025e]: • Members operating in the hazard zone • Where members are in the hazard zone • Conditions in the hazard zone • Actions used in the hazard zone • Paths of access and egress in and out (i.e., exits) of the hazard zone • Rapid intervention teams/crews and their assignments

Different methods and tools are available for resource accountability, including [NIOSH 2025e]: • Tactical worksheets • Command boards • Apparatus riding lists • Company responding boards • Electronic bar-coding systems • Accountability tags or keys

Completion of personnel accountability can be assigned to a staff aide (e.g. incident command technician) by the IC. Functions of the staff aide include maintaining the tactical worksheet; maintaining personnel accountability of all members operating at the incident (resource status and situation status); monitoring radio communications on the dispatch and operations channels; control information flow; and accessing reference material and pre-incident plans. Some fire departments use firefighters as staff aides and other fire departments use fire officers to serve as a staff aide for a command officer. This position can be appointed, or fire departments can consider increasing staffing to fulfill this position [NIOSH 2016; NIOSH 2025d].

Recommendation #6: Governing municipalities (federal, state, regional/county, and local) should ensure the delegated zoning or code enforcement authority, law enforcement agency, and fire department collaborate to proactively remediate vacant and abandoned buildings to reduce arson. In this incident, the AHJ maintained a vacant property registration program that provided contact information and generated fees to cover municipal costs associated with these properties. This included a vacant property officer who interacted with owners and municipal departments, emphasizing compliance more than enforcement. At the time of the incident, there were more than 350 vacant properties registered with the AHJ. The fire department in this incident generally responds to 10 or more fires annually in vacant and abandoned buildings.

Page 21 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Many fire departments have vacant/abandoned/derelict (high-risk) structures in their jurisdictions. These types of buildings are of increasing concern to the life and safety of firefighters. These structures lack structural integrity and may contain unknown hazards [USFA 2023]. The best way to prevent vacant building fires is to reduce or eliminate vacant (i.e., high-risk) buildings. However, most fire departments lack the necessary legal standing or resources to carry out the demolition of a building. So, it is critical for them to form strategic partnerships with zoning, code enforcement, and law enforcement agencies of their AHJ to develop and implement strategies for timely remediation of these buildings [NIOSH 2025a].

After vacant and abandoned buildings have been identified, the fire department should work with the AHJ to remediate them by securing all entrances or demolition. Preventing unauthorized individuals from entering a building can reduce the risk of a fire igniting and arson. Securing the building might include physical barriers to deny access [NIOSH 2025a]. The effectiveness of physical barriers should be evaluated during building inspection. Once secured, buildings must be patrolled with regular frequency to make sure they remain secure. High visibility of law enforcement patrols can also be an effective control in deterring unauthorized access [USFA 2023].

The National Vacant Properties Campaign describes several strategies to address the problem of vacant properties and provides examples of how these strategies have been used. Strategies include [National Vacant Properties Campaign 2024]:

• Land banks for property seized from nonpayment of taxes • Rehabilitation programs for owner-occupied housing and home repair programs • Information systems to capture data about individual properties and neighborhoods that allow developing problems to be identified, tracked, and addressed • Property maintenance codes related to occupied housing that reduce the likelihood a property falls into serious disrepair and abandonment • Nuisance abatement authority that allows municipalities to address threats to the general public, typically, through administrative hearings rather than courts • Receivership, which is legal action that places property in dispute under the control of an independent person (e.g., court-appointed neutral party) Post-Incident Fire Department Prevention Actions After this incident, the fire department implemented changes to its incident response and fireground operations. These changes were based on the department’s critique of the incident on February 12, 2025.

• Collapse hazards in CAD data: Vacant and derelict buildings that have response hazards identified from inspections and crew observations are now recorded in the fire department’s CAD data to ensure that arriving units have additional information on incident buildings prior to arrival. • Building Owner Remediation:

Page 22 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 When the fire prevention division identifies fire and life safety code violations at commercial buildings, the fire marshal’s office issues violation notices for building owners to appear in court if the code violations are not remedied. This includes vacant buildings. References Dodson D [2021]. Fire department incident safety officer, 3rd Ed. Burlington, MA: Jones & Bartlett Learning. IFSTA [2013]. Essentials of fire fighting. 6th ed. Stillwater, OK: International Fire Service Training Association.

IFSTA [2015]. Fire and emergency services safety officer. 2nd ed. Stillwater, OK: International Fire Service Training Association.

IFSTA [2019]. Rapid intervention teams. 2nd ed. Stillwater, OK: International Fire Service Training Association.

National Vacant Properties Campaign [n.d.]. Strategies and technical tools. Date accessed: June 13, 2025.

NFPA [2020]. NFPA 1407, Standard for training fire service rapid intervention crews. Quincy, MA: National Fire Protection Association. NFPA [2021]. NFPA 1700, Guide for structural fire fighting. Quincy, MA: National Fire Protection Association.

NFPA [2022]. NFPA 2500, Standard for operations and training for technical search and rescue incidents and life safety rope and equipment for emergency services. Quincy, MA: National Fire Protection Association. NFPA [2024]. NFPA 1550, Standard for emergency responder health and safety. Quincy, MA: National Fire Protection Association. NIOSH [2008]. Career fire fighter dies when trapped by collapsed canopy during a two alarm attached garage fire – Pennsylvania. By Koedam R, Merinar T, and Kovacevic L. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2007-08. NIOSH [2013]. Career lieutenant and fire fighter killed and two fire fighters injured by wall collapse at a large commercial structure fire - Pennsylvania. By Miles S., Bowyer M., Merinar T., Tarley J., and Loflin M. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2012-13.

Page 23 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 NIOSH [2014]. Preventing deaths and injuries to fire fighters by establishing collapse zones at structural fires. Cincinnati, OH: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health (NIOSH), Publication No. 2014-120. NIOSH [2016]. Lieutenant and fire fighter die and 13 fire fighters injured in a wind-driven fire in a brownstone – Massachusetts. By Bowyer ME, Loflin ME, and Miles ST. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2014-09. NIOSH [2024]. Volunteer firefighter killed after becoming trapped at an assisted living facility fire and two firefighters injured – New York. By Loflin M and Attwood W. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2021-10. NIOSH [2025a]. Three firefighters die and another seriously injured during the collapse of an abandoned, derelict, 3-story row house fire – Maryland. By Funke JR, Attwood WR, Haas EJ, Miles ST, Bowyer M, and Loflin ME. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2022-01. NIOSH [2025b]. A military firefighter killed and two firefighters injured by a wall collapse at a barn fire – West Virginia. By Loflin ME, Attwood WR, and Ringer SJ. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2021-09. NIOSH [2025c]. Career candidate firefighter found unresponsive at a residential structure fire and dies 5 days later–Illinois. By Loflin ME and Saunders R. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2021-20. NIOSH [2025d]. Firefighter dies and lieutenant seriously injured at a multi-family residential structure fire – Connecticut. By Loflin ME, Attwood WR, and Funke JR. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2021-11. NIOSH [2025e]. Lieutenant dies from a floor collapse in residential structure fire with unpermitted renovations – Illinois. By Loflin ME, Attwood WR, and Austin E. Morgantown, WV: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, F2022-09. SKCFTC [2023]. Command procedures: Structure fires. Kent, WA: South King County Fire Training Consortium. March 2023.

Sullivan J [2012]. ISO’s role when transitioning from an offensive to a defensive attack. FireRescue, April: 86–89.

Page 24 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 Toledo Fire & Rescue Department [2012]. Rapid intervention team (RIT) standard operating procedure C82. Toledo, OH: Toledo Fire & Rescue Department. USFA [2023]. Reducing arson at vacant and abandoned buildings. Emmittsburg, MD: U.S. Fire Administration.

Weather Underground [2025]. Weather history. The Weather Underground. Date accessed: June 6, 2025. Investigator Information This investigation report was authored by Dr. Wesley R. Attwood, Senior Investigator, with the Fire Fighter Fatality Investigation and Prevention Program, Surveillance and Field Investigations Branch, Division of Safety Research, NIOSH. James Johnson, Fire Ground Survival Master Instructor and Codes and Standards Representative, and Jeffrey Seaton, an expert in firefighter survival, International Association of Fire Fighters, provided a subject matter expert review of the investigation report. Additional Information Preventing Deaths & Injuries to Firefighters by Establishing Collapse Zones at Structure Fires Firefighters are at significant risk for injury or death due to structural collapse during firefighting operations. Structural collapse often occurs without warning. NIOSH recommends that the IC establish defensive operations and collapse zones when there is potential for a structural collapse during fire- fighting operations. Access more information at: https://www.cdc.gov/niosh/docs/wp-solutions/2014- 120/.

Partnering to Reduce the Risk to Firefighters Responding to High-risk Buildings This NIOSH Science Blog discusses strategic partnerships fire departments and authorities having jurisdiction can develop to create solutions to large community safety issues such as abandoned or condemned structures.

NFPA 1700, Guide for Structural Fire Fighting (2021 edition) NFPA 1700, Guide for Structural Fire Fighting, 2021 edition, is the first NFPA document connecting fire dynamics research and its application to strategy, tactics, and best practices for firefighters in controlling fires within a structure. Disclaimer The information in this report is based upon dispatch records, audio recordings, witness statements, and other information that was made available to the National Institute for Occupational Safety and Health (NIOSH). Information gathered from witnesses may be affected by recall bias. The facts, contributing factors, and recommendations contained in this report are based on the totality of the information gathered during the investigation process. This report was prepared after the event occurred, includes information from appropriate subject matter experts, and is not intended to place blame on those involved in the incident. Mention of any company or product does not constitute endorsement by NIOSH, Centers for Disease Control and Prevention (CDC). In addition, citations to websites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or

Page 25 Career Firefighter Dies after being Trapped by Collapse of Two Vacant Commercial Buildings – New York Report # F2025-03 products. Furthermore, NIOSH is not responsible for the content of these websites. All web addresses referenced in this document were accessible as of the publication date. NIOSH Approved is a certification mark of the U.S. Department of Health and Human Services (HHS) registered in the United States and several international jurisdictions.

FD06036 ｜ 2026_美国国家消防协会_曼哈顿单楼梯公寓火灾调查报告（英文）.md

分类：06_境外及港澳台调查报告 ｜ 年份：2026 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：md

一、基本情况

2026年5月4日，纽约曼哈顿一栋仅设单一疏散楼梯的公寓楼发生致命火灾。初步报道指出潜在致因包括单一疏散楼梯、敞开的门户以及已知的合规缺陷。

二、起火原因

后续信息确认：火灾由被人随意丢弃在首层门厅纸箱上的香烟引燃，烟气与火焰沿楼梯间向上蔓延；至少2名遇难者被发现于靠近屋顶的楼梯间顶部。

三、责任与致灾因素（待全面调查）

• 早期信息有限，尚不清楚建筑内安装了哪些消防系统及其是否按预期动作；
• 需弄清烟气如何进入楼梯间、人员为何在楼梯间内被烟气困住；
• 不能仅以“单一楼梯”草率归因——多层建筑（含多楼梯）在烟气/火焰进入楼梯间时同样发生过重大伤亡（如2022年布朗克斯17人遇难火灾）。

四、整改与教训

NFPA强调：致命火灾后立即下结论存在风险，应待完整调查厘清所有致因；无论建筑形式，工作联动的联网烟感、可靠自闭式防火门、自动喷淋的防护价值已被充分研究证明。对单楼梯高层提案的放宽，应在开放、共识性的规范制定流程中评估，而非以单一既有建筑火灾误导立法。

FD06037 ｜ 2004_英国_fire-in-the-engine-room-on-the-suction-dredger-arco-avon-with-loss-of-1-life__ArcoAvon_Com.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2004 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A HSE Research Report 222

HSE Health & Safety Executive Tests for the ignition and flame spread of clothing fabrics subjected to angle grinder sparks Results from ten generic fabrics and variants Prepared by Building Research Establishment Ltd for the Health and Safety Executive 2004 RESEARCH REPORT 222

HSE Health & Safety Executive Tests for the ignition and flame spread of clothing fabrics subjected to angle grinder sparks Results from ten generic fabrics and variants Peter Fardell, C.Chem. MRSC. MSc. and Carol Rock, BSc. Centre for Fire Performance and Suppression Fire Division Building Research Establishment Ltd Garston Watford Herts WD25 9XX Test methods for measuring the times to ignition and rates of flame spread of fabrics subjected to sparks from a hand angle grinder have been developed. Such tests are needed by the HSE as there is a regular incidence of burn injuries caused through workwear fabrics igniting and causing burns to the wearer through the use of angle grinders. Separate apparatus was used for measuring time-to-ignition and flame spread and development details have been previously reported. Ten generic fabric fabrics have been tested (sixteen, including variants) with the prototype apparatus, including synthetic, natural, synthetic/natural mixtures, Flame Retarded (FR), laundered FR and fabrics contaminated with grease and paint. Many fabrics could ignite after less than a minute with one “FR” fabric igniting after ten seconds. Another FR fabric, checked for correct FR content was resistant to ignition and flame spread.. Examples of non-ignition were observed with both natural and synthetic fabrics but a different mechanism was apparent in each case. Natural fabrics tended to form a protective carbonaceous char whereas synthetic fabrics tended to melt and form a hole around the stream of sparks, preventing or reducing the likelihood of ignition. Flame-spread rates were measured horizontally and at 25O to the horizontal. Most fabrics showed a well defined rate of flame spread which was usually significantly faster in the inclined position. Overall rates between 9mm/min and 891 mm/min were measured although in some cases flaming ceased before the whole length of sample had been consumed. Two full scale tests using a manikin clothed in FR (Proban) cotton and a non FR cotton “boiler suits” confirmed the small scale test results and illustrated the value of FR fabrics in resisting ignition and flame spread from hand grinder sparks. The non FR boiler suit ignited after twenty seconds and burnt to completion. The FR boiler suit did not ignite after sustained continuous application of a stream of sparks over a five minute period. This report and the work it describes were funded by the Health and Safety Executive (HSE). Its contents, including any opinions and/or conclusions expressed, are those of the authors alone and do not necessarily reflect HSE policy. HSE BOOKS

© Crown copyright 2004 First published 2004 ISBN 0 7176 2841 8 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means (electronic, mechanical, photocopying, recording or otherwise) without the prior written permission of the copyright owner. Applications for reproduction should be made in writing to: Licensing Division, Her Majesty's Stationery Office, St Clements House, 2-16 Colegate, Norwich NR3 1BQ or by e-mail to hmsolicensing@cabinet-office.x.gsi.gov.uk ii

ACKNOWLEDGEMENTS The authors would like to thank Miss Leah Amendt, Miss Sheila Warbus and John Cormican for providing enthusiastic and skilful practical help with all stages of this project. iii

iv

CONTENTS EXECUTIVE SUMMARY 1 1 INTRODUCTION 3 2 DESCRIPTION OF THE PROJECT 5 2.1 General Background 5 2.2 Fabrics Used 5 2.3 Replacement for fabric 2. 7 2.4 Contaminated samples 7 2.5 Effect of changing the sample holder for the ignition studies. 8 2.6 Effect of FR additive to the cotton fabric. 9 2.7 Testing of the ten generic fabrics and variants 9 2.8 Full scale dressed manikin tests 9 3 FINDINGS AND DISCUSSION 11 3.1 Summary of ignition and flame spread results 11 3.2 Natural and synthetic fabrics 13 3.3 FR treated fabrics and effects of laundering 13 3.4 Fabrics contaminated with grease or paint. 13 3.5 Manikin fire test 14 3.6 Detailed test results 15 APPENDIX 1 - Test data obtained from the “time-to-ignition” study of the ten generic fabrics and variants. APPENDIX 2 – Test data obtained from the “flame-spread rates” study of the ten generic fabrics and variants. APPENDIX 3 – Ancillary measurements 4 CONCLUSIONS AND RECOMMENDATIONS 17 REFERENCES 35 v

vi

EXECUTIVE SUMMARY Apparatus and procedures have been developed to measure the times to ignition and flame-spread rates of clothing fabrics subjected to angle grinder sparks. The development work has been separately reported in previous Milestone deliverables. This final report describes results obtained using the prototype apparatus and procedures on ten generic fabric samples and variants, chosen by the HSE. The time-to-ignition apparatus utilised a proprietary, commonly available, 115mm disc angle grinder, retained in a proprietary stand. The work piece used was a hollow square section bar of mild steel, 50mm x 50mm x 7mm wall thickness, which enabled the apparatus to generate a steady and repeatable stream of sparks over periods in excess of 5 minutes. For measurement of the time-to-ignition, the tested fabric was held in a vertical wire mesh holder which allowed the stream of grinder sparks to impinge at the base of the sample in a “U” shaped vertical channel. The sparks were allowed to impinge for various fixed time intervals after which the sample was inspected for self-propagating flames. Each time interval was repeated between 3 and 5 times. When every trial over a particular time period produced self-propagating flaming, this was recorded as the time-to-ignition. For measurement of the flame spread rate procedure, it was not possible to utilise the same apparatus used for time-to-ignition. This was due to uncertainties in measuring the flame front and because the flame spread following ignition could be very variable depending on subtle folds developing and sections of the sample falling at random during flaming. It was therefore agreed to utilise a separate apparatus for the flame spread rate studies. This apparatus was based on an existing standard used to test the horizontal flame spread of fabrics used in the interior of vehicles in the USA. Each flame spread rate determination was repeated and the mean value used as the quoted flame spread rate. Finally, two tests were carried out using a full size manikin dressed in an FR and then a non FR boiler suit,using the standard stream of angle grinder sparks used in the main study. In summary the main findings were: · Natural and synthetic fabrics were not particularly distinguishable by their ignition and flame spread performance – although the mechanisms were different in each case. Synthetic fabrics tended to melt, natural fabrics tended to char. · A proprietary FR fabric ignited readily. · A laundered FR cotton fabric had a reduced time to (partial) ignition compared with an unlaundered sample but remained resistant to flame spread. · Grease contamination showed marginally reduced ignition times and a tendency to increased flame-spread rates over the uncontaminated samples. · Paint contamination increased the ignition time and reduced the rate of flame spread with a cotton sample. With a contaminated polyester sample, time-to-ignition was reduced and flame- spread rates were marginally increased over the uncontaminated sample. 1

· The FR boiler suit resisted ignition and failed to ignite after 5 minutes of continuous application of a stream of sparks. The non-FR boiler suit ignited after approximately 20 seconds and continued to burn to completion. A number of ancillary studies were also carried out to investigate particular observations. These were: · Validation trials for a substitute fabric which had been discontinued during the study. (The substitute fabric was deemed equivalent). · Investigation of the time-to-ignition changes occurring following a change of sample holder and operator. (The changes were ascribed to the change of sample holder and were not found to be significantly operator dependant). · Obtaining elemental analysis of a substitute FR fabric following the ready ignition observed with an original FR fabric. (This “pedigree” fabric was shown by elemental analysis to contain the expected level of FR treatment and exhibited the expected ignition characteristics). It is recommended that the apparatus and methodology developed in this study be adopted for use by the HSE as a basis for testing the ignition and rate of flame spread performance of work wear fabrics when subjected to sparks from an angle grinder. It is further recommended that some aspects of the study (e.g. fabric combinations, contaminated samples and development of performance criteria) should be carried out in future work. 2

1 INTRODUCTION This study was undertaken for the HSE because of concerns over the incidence of injuries sustained in the workplace from fires caused by angle grinder sparks impinging on clothing. A simple but realistic test was required which could measure times to ignition and flame-spread rates from a range of fabrics which might be used in work wear. Statistics gathered by the UK’s Health and Safety Executive (HSE) as part of their “FIREX” database [1] have shown a continuing incidence of burn injuries caused by hand operated angle grinder sparks impinging on clothing. There have been at least ten such incidents over the last ten years including three serious injuries. The Fire and Risk Sciences (FRS) division of the Building Research Establishment was invited to design, develop and establish standard test apparatus and procedures for measuring the time-to- ignition and rates of flame spread on clothing fabrics subjected to constant stream of angle grinder sparks. The intention is to use the methods to determine the ignition and flame spread characteristics of fabrics which may be encountered in the workplace, both as normal and workwear clothing. There was no requirement to set criteria for hazard classification although this would form a logical next step. This is the final milestone report and describes a study of the ignition and flame spread characteristics of ten generic work wear fabrics and variants, (chosen by the HSE following the development stages) subjected to angle grinder sparks. The methods used for measuring times to ignition and flame spread have been developed in previous work under this contract and have been separately reported [2,3]. Progress reports, visits by and to the HSE project officer and other communications have been made at regular intervals throughout the study. 3

4

2 DESCRIPTION OF THE PROJECT 2.1 General Background Previous work on this project has been fully described in reports of meetings with the HSE project officer and two Milestone Reports [2,3] but in summary, the initial work involved setting up a proprietary hand angle grinder and stand and developing a technique for generating a standard continuous shower of sparks from a suitable work piece. This was followed by the development of a suitable fabric sample holder and technique for measuring time-to-ignition of tested fabrics. For flame spread measurements, an existing small-scale horizontal flame spread apparatus [4] was modified and used. It had been hoped that the ignition and flame spread parameters would be measured on the same apparatus and during the same test. However it proved very difficult and subjective to measure flame- spread rates following ignition of a vertical sample with grinder sparks. The flame front was difficult to follow and although possibly the worst case for flame spread, (vertical orientation) the variations in flaming behaviour (e.g. through folds developing in a random manner and fabric dropping from the sample holder) it was decided to separate the testing for ignition and flame spread behaviour. This gave the opportunity to test flame spread in a much more controlled manner (using a modified existing standard test protocol [4]) and to investigate the effect of sample orientation on flame-spread rates. A wide range of fabric types including pure natural, pure synthetic and natural/synthetic mixtures were studied in the development stages of the project. 2.2 Fabrics Used Table 1 summarises the fabrics and their variants used over the entire study both during the development work (as previously reported [2,3]) and as reported here. The fabrics tested for this report, using the prototype apparatus and procedures, are shown in bold type in table 1. This includes the ten generic fabrics, fabric combinations, fabrics contaminated with grease or paint, and laundered and unlaundered FR cotton fabrics. All fabrics were conditioned for a minimum of 48 hours to Standard conditions (50%RH +/- 5% and 23OC +/- 2OC) prior to testing. . 5

Table 1 – Summary of fabric fabrics used over the entire study (for this report in bold). Fabric Number Description Mass per unit area (kg/m2) Remarks 1 35% polyester/65% cotton 0.092 2 100% cotton close weave 0.156 Replaced by 15. 3 100% cotton FR treated (interliner) 0.140 4 100% polyester (“Charmeuse” satin) 0.091 5 100% cotton (open weave) 0.174 6 81% acrylic/19% polyester 0.212 7 59% cotton/41% modacrylic 0.402 8 55% polyester/45% viscose FR treated 0.281 9 100% wool (suiting “dogtooth”) 0.182 10 100% cotton FR treated (corduroy) 0.599 11 Stretch denim (cotton) 0.377 12 Polyester fabric and wadding (Quilting) 0.219 13 100% polyester (purple) 0.104 14 100% polyester over 100% wool N/A Two layers of fabric 15 Cotton close weave (replacement for 2, blue/white) 0.161 Checked for similarity with 2 16 “Proban” treated cotton - unlaundered N/A CHNP (41.1%, 6.3%, 2.1%, 2.3%) 17 “Proban” treated cotton - laundered N/A To BS EN ISO 10528:1995 18

100% cotton with grease contaminant

Fabric 15 + grease 19

100% cotton with paint contaminant

Fabric 15 + paint 20

100% polyester with grease contaminant

Fabric 13 + grease 21

100% polyester with paint contaminant

Fabric 13 + paint 22

Boiler suit (cotton, non FR)

23

Boiler suit (cotton FR)

6

1. 3

Replacement for fabric 2. Following the use of this fabric (a “close weave” cotton) for ignition and flame spread study, more fabric was needed for the “contamination” studies (see 2.4). Unfortunately it was discovered from the supplier that this was a discontinued line. Although there was an option for this fabric to be specially made by the manufacturer it was decided that due to the expense and time period required, a similar fabric would be selected which was likely to be in stock for an extended period. Fabric 15 was therefore obtained and used for the rest of the “contamination” study in place of fabric 2, following checks on the similarity of the new fabric to the old in terms of time-to-ignition and flame spread (horizontal sample only). 2.4 Contaminated samples Two fabric types – one 100% synthetic (fabric 13) and one 100% natural (fabric 15 – the replacement for fabric 2) were chosen for study where the fabric had been contaminated with a petroleum-based workshop grease and an organic solvent paint. The contaminated fabrics were designated 18,19,20 and 21 (see table 1). The whole area of fabric was coated in each case for the ignition and flame spread tests, with an allowance made for the different sized specimens required for each test. With the grease contaminant, application was attempted by smearing a weighed portion of grease on a spatula over the whole area of a fabric both for the time-to-ignition and flame spread samples. It was possible to obtain a reasonably uniform coating with this technique after several trials. The samples were tested within 2 hours of application to avoid any significant evaporation of any volatiles. With the paint contaminant, application was attempted by simply brushing on the paint to the fabric surface as uniformly as possible, and allowing the paint to dry over a 24 hour period. The “loading” of paint was measured by weighing the fabric before and after application of the paint but it proved difficult to make the loadings identical for the two fabric types, and a fairly wide variation of loading had to be accepted. Table 2 summarises the fabrics chosen, the contaminants used and the contaminant loading achieved. 7

Table 2: A summary of the contaminated tests fabrics tested Fabric Description Contaminant Approximate mass applied. Ignition Flame Spread 18 100% Cotton close weave “Castrol Spherol L-EP2” petroleum based grease. 2g over an area 300mm x 100 mm 1g over an area 300mm x 35mm 20 100% polyester “Castrol Spherol L-EP2” petroleum based grease. 2g over an area 300mm x 100 mm 1g over an area 300mm x 35mm 19 100% Cotton close weave “Dulux Trade High Gloss” organic solvent based paint. 3.1g over an area of 300mm x 100mm 1.4g over an area of 300mm x 35mm 21 100% polyester “Dulux Trade High Gloss” organic solvent based paint. 1.6g over an area of 300mm x 100mm 0.68g over an area of 300mm x 35 mm 2.5 Effect of changing the sample holder for the ignition studies. During the development stages of this work, a simple variable “handy angle” sample holder was used for the measurement of times to ignition of fabric fabrics subjected to angle grinder sparks. [2]. This enabled various configurations of fabric to be presented to the stream of grinder sparks, to achieve a configuration which gave the most consistent and measurable ignition results. However following the establishment of the optimum dimensions for the sample holder it was decided to use a more accurately constructed, fixed dimensions, open wire mesh sample holder for the final series of tests on the ten generic fabrics and variants. However early trials with this holder, with a different operator, showed significantly shorter times to ignition for the same fabric. A short series of tests was therefore performed on two of the fabrics (9 and 12) which showed the most variability in order to establish whether the effect was due to the change of sample holder or to the change of operator. 8

1. 6

Effect of FR additive to the cotton fabric. One of the ten generic fabrics chosen for study in this phase of the work was an FR cotton, (fabric 3). It was intended to use this fabric to determine the effects of a standard laundry cycle [5] on the flammability of an FR fabric but this particular fabric surprisingly proved to ignite readily from the grinder sparks (but showed little flame spread). It was therefore decided to include a “pedigree” FR cotton fabric treated with “Proban” (fabric16) in the programme with elemental analysis to determine the level of FR present, and check this level with the usually accepted level for an FR cotton fabric [6]. Accordingly a sample of a “Proban” treated cotton was purchased and an elemental analysis (C, H, N, and P) obtained. 2.7 Testing of the ten generic fabrics and variants The testing apparatus and procedures used for the work carried out for this report was as previously described in detail [2,3]. In summary, for measurement of the time-to-ignition, the tested fabric was held in a vertical wire mesh holder which allowed the stream of grinder sparks to impinge at the base of the sample in a “U” shaped vertical channel. The sparks were allowed to impinge for various fixed time intervals after which the sample was inspected for self-propagating flames. Each time interval was repeated between 3 and 5 times. When every trial over a particular time period produced self- propagating flaming, this was recorded as the time-to-ignition. Appendix 1 contains the detailed results of the time-to-ignition study. For measurement of the flame spread rate, it was not possible to utilise the same apparatus used for time-to-ignition. This was due to uncertainties in measuring the flame front and because the flame spread following ignition could be very variable depending on subtle folds developing and sections of the sample falling at random during flaming. It was therefore agreed to utilise a separate apparatus for the flame spread rate studies. This apparatus was based on an existing standard used to test the horizontal flame spread of fabrics used in the interior of vehicles in the USA. Each flame spread rate determination was repeated and the mean value used as the quoted flame spread rate. Appendix 2 contains the detailed results of the flame spread rate study. 2.8 Full scale dressed manikin tests Two cotton “boiler suits” were obtained from a recognised supplier [7], one treated with “Proban” flame retardant, the other untreated. These boiler suits were used to clothe a manikin with mineral wool inserts to simulate the torso and arms and legs. Each dressed manikin was subjected to a stream of grinder sparks below the waistline and observations made of ignition and flame spread. 9

10

3 FINDINGS AND DISCUSSION 3.1 Summary of ignition and flame spread results Table 3 summarises the ignition and flame spread data for the ten chosen fabrics and variants. For time-to-ignition, the results were obtained using the wire frame sample holder thus repeating several of the tests carried out in the development stages which used a “handy angle” sample holder. Three to five successful ignitions (i.e. flame propagating away from the influence of the grinder sparks) were used as the criteria for ignition over a given time interval. In some cases differences in times to ignition were observed, depending on the orientation (i.e. weft lengthwise or width wise). These are indicted in table 3 where the time-to-ignition was altered sufficiently to place the result in another time period. For the flame spread rate trials, only the additional fabrics required for test following the development stages were tested. Flame-spread rates for the other fabrics are reported here as obtained during the development stages. 11

Table 3 – Summary of time-to-ignition and flame spread results for the ten generic fabrics and variants. Fabric number Fabric description Time-to-ignition (Seconds). 1 Mean rate of flame spread (horizontal) (mm/min) Mean rate of flame spread (25Oincline) (mm/min) 1 35% polyester, 65% cotton 10 176 889 2 100% cotton (close weave) 20 131 891 3 100% cotton “FR” (interliner) 10 0 450 4 100% polyester Not measurable (melted and formed hole). 0 85 (extinguished after 15 seconds) 9 100% wool (suiting) 45 0 0 10 100% cotton “FR” (corduroy) Did not ignite. 0 143 (extinguished after 46 seconds) 11 Denim (cotton) stretch fabric 20 (lengthwise) 30 (widthwise) 84 407 12 Polyester fabric quilt over wadding 10 80 391 13 100% polyester Not measurable (melted and formed hole). 0 248 (extinguished after 20 seconds) 14 100% polyester over 100% wool (two layers) 30 0 0 16 100% cotton with “Proban” FR treatment – unlaundered sample. 30 (only small flamelets observed for 2 seconds). 0 0 17 100% cotton with “Proban” FR treatment – laundered sample. 10 (only small flamelets observed for 2 seconds). 0 0 18 100% cotton (fabric 15) with grease contaminant. 15 140 728 19 100% cotton (fabric 15) with paint contaminant. Ignited in only one trial at 60s on non painted area. 91 496 20 100% polyester (fabric 13) with grease contaminant. Ignited in only one trial at 40s. 254 (extinguished after 10 secs) 9.4 (extinguished after 12 seconds) 21 100% polyester (fabric 13) with paint contaminant. 30 170 587 12

1. 2

Natural and synthetic fabrics Time-to-ignition results did not particularly distinguish between natural and synthetic fabrics. Flame- spread rates were generally significantly faster for the 25O inclined sample, over the horizontal sample. In some cases no flame spread was observed with a horizontal orientation but was observed in the 25O inclined orientation. With the synthetic fabrics a hole was often punched through the fabric preventing further impingement of sparks onto the sample, which in some cases produced a “no-ignition” result. Flame- spread rates on these samples were however often measurable, particularly in the 25O inclined orientation. With natural fabrics a char tended to form which appeared to inhibit ignition. Wool was particularly resistant. A polyester covering over wool (fabric 14) ignited after 30 seconds but a flame spread rate could not be determined in either of the two orientations attempted for this double-layered fabric. Fabrics consisting of synthetic/natural fibre mixes were not particularly distinguishable from their “pure” counterparts. With fabric combinations (i.e. two layers) there was some evidence of intermediate behaviour from that measured with the individual components, but this was somewhat tenuous and needs further investigation. 3.3 FR treated fabrics and effects of laundering One proprietary FR fabric (fabric 3) ignited readily (after 10 seconds). This required the inclusion of a “pedigree” fabric, analysed for FR content (fabric 16). The phosphorus content measured (2.3%) is consistent with the recognised correct loading of “Proban” cotton fabric in this application [6]. The unlaundered “Proban” treated cotton fabric (fabric 16) ignited after 30 seconds but produced small flamelets which extinguished after a few seconds. After being subjected to a standard laundry cycle, the FR “pedigree” fabric (fabric 17) ignited after 10 seconds (but with small, barely propagating flames). No flame spread was observed with either the laundered or unlaundered samples. 3.4 Fabrics contaminated with grease or paint. Grease contamination on a cotton fabric (fabric 18) reduced the time-to-ignition from 20 seconds on the uncontaminated sample to 15 seconds. Flame spread rate was slightly increased in a horizontal orientation and slightly decreased in the 25 degree inclined orientation compared to the uncontaminated sample. (However, note that a substitute fabric for fabric 2 had to be used for this “contamination” trial). Paint contamination on a cotton fabric (fabric 19) increased the time-to-ignition from 20 seconds on the uncontaminated sample to over 60 seconds. Flame-spread rates were significantly reduced in both orientations, over the uncontaminated sample. (However, note that a substitute fabric for fabric 2 – fabric 15 - had to be used for these “contamination” trials). 13

Grease contamination on a polyester fabric (fabric 20) did produce a marginal ignition at 40 seconds. The uncontaminated sample (fabric 13), melted and developed a hole, preventing further impact from sparks. A flame-spread rate for the contaminated sample was observed in the horizontal orientation with (atypically) a much lower rate in the 25O orientation, but in both tests the flame extinguished well before complete combustion. No spread was observed with the uncontaminated sample when horizontal but the 25O orientation uncontaminated sample did produce a measurable flame spread but again flaming extinguished well before the sample had been consumed. Paint contamination on a polyester fabric (fabric 21), produced a reliable ignition after 30 seconds. The uncontaminated sample (fabric 13) melted and developed a hole, preventing further impact from sparks. Flame-spread rates of the contaminated sample were measured with both sample orientations with complete combustion of the sample in each case. The flame spread rate was much faster with the 25O orientation. These variable results must be considered to be due in part to the difficulties experienced in applying a uniform and consistent level of contamination on the fabrics and to the variable flammability nature of the base materials, particularly the polyester fabric. Further work is needed on this aspect to confirm (or refute) the findings from this limited study. 3.5 Manikin fire test The FR treated boiler suit resisted ignition and did not produce any signs of self-propagating flames after exposing the fabric to a sustained stream of sparks over a five minute period. The fabric glowed, smouldered and fumed during application of the sparks but this ceased shortly after the stream of sparks was stopped. The non FR boiler suit ignited readily after approximately 20 seconds when the stream of sparks was stopped. It continued to burn to completion. Figures 1 and 2 illustrate the behaviour of the FR and non FR boiler suit in the manikin tests: Figure 1. Grinder sparks being applied to manikin (FR “Proban” cotton boiler suit). No ignition occurred after five minutes of sustained application of the sparks 14

Figure 2. Grinder sparks have been applied to a manikin (non-FR boiler suit). The result after approximately 45 seconds (ignition occurred at approximately 20 seconds). 3.6 Detailed test results Appendices 1 and 2 provide details and results of the all trials carried out to obtain these data for ignition times and flame-spread rates respectively. Appendix 3 provides details and results of the ancillary measures carried out during this project i.e. · Trials for a substitute for fabric 2. · Establishing the degree of operator dependence for the ignition study · Elemental analysis of a “Proban” flame retarded cotton fabric. 15

16

4 CONCLUSIONS AND RECOMMENDATIONS 1. Following extensive development in earlier reported studies, [2,3] an apparatus for measuring the time-to-ignition and an apparatus for measuring rate of flame spread has been developed for work wear fabrics exposed to angle grinder sparks. 2. Ten generic fabrics and variants were chosen by the HSE to use in these tests with the two sets of apparatus. The ten fabrics and variants included pure synthetic fabrics, pure natural fabrics, synthetic/natural fibre mixtures, fabric combinations, FR fabrics (in one case laundered and unlaundered) and a natural and synthetic fabric both contaminated with paint and grease. In summary, main conclusions were: · The range of synthetic and natural fabrics tested showed ignition times and flame- spread rates covering a wide range but there was no particular distinction between values obtained for the two types. · Synthetic fabrics tended to melt and natural fabrics tended to form a surface char, both effects having a strong influence on performance in the ignition and flame spread tests. · Natural/synthetic fibre mixes were not particularly distinguishable from their “pure” counterparts. · Combinations (i.e. two layers) of synthetic and natural fabrics showed some evidence of intermediate performance between the two components. · A cotton fabric marked as “FR interliner” ignited readily. · A cotton fabric purchased as “FR” and checked for acceptable FR content by elemental analysis did exhibit very limited ignition and flame spread behaviour, both before and after being subjected to a single standard laundry cycle.[5]. · Synthetic and natural fabrics contaminated (separately) with grease and paint did not show a particularly degraded performance – in some cases a marginal improvement was observed. However these results must be considered as provisional (and subject to further confirmation) due to problems with ensuring a uniform application of contaminant to the samples and the variability in behaviour of the base fabrics. · Tests on a full sized manikin clothed in separate tests with an FR and non FR “boiler suit” demonstrated the value of FR treated materials when subjected to sparks from an angle grinder. The FR boiler suit did not ignite after 5 minutes of sustained spark impingement whereas the non FR boiler suit ignited after approximately 20 seconds and burnt to completion. 3. It is recommended that the test apparatus and methodologies as developed in this study should be used as the basis for a Standard test method for determining the times to ignition and rates of flame spread of fabrics likely to be encountered in the workplace, when subjected to a stream of sparks from an angle grinder. 4. Further development work is recommended, particularly in the areas of fabric combinations, contaminated fabrics and to develop performance criteria. 17

18

APPENDIX 1 - Test data obtained from the “time-to-ignition” study of the ten generic fabrics and variants. (Note – “Width” and “Length” terms relate to the direction of the warp threads of the fabric in the sample holder. Fabric: 1. 35%polyester/65%cotton. Orientation: Width (Length) Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 N (Y) N (Y) Y (Y) Y (Y) Y (Y) 10 Y Y Y Y Y

15

-

-

20

-

-

30

-

-

40

-

-

Other

-

-

Fabric: 2. Cotton close weave. (note – replaced by fabric 15 for “contamination” studies). Orientation: Width (Length) Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks

5

-

-

10 N N

N

- 15 N (N) N (N)

N (N)

- 20 Y (Y) Y (Y)

Y (Y)

-

30

-

-

40

-

-

Other

-

-

19

Fabric: 3. 100% cotton FR treated (interliner). Orientation: Width (Length) Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 N (N) N (N) N (N) N (N) N (N) Slight scorching 10 Y (Y) Y (Y) Y (Y) Y (Y) Y (Y) Surprising for an FR product

15

-

-

20 (Y) (Y) (Y) (Y) (Y) Surprising for an FR product

30

-

-

40

-

-

Other

-

-

Fabric: 4. 100% polyester. Orientation: Width Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 N N

N

- No sign of ignition/fumes 10 N N

N

Slight melting 15 N N N N N Slight melting 20 N N N -

Melted, punched hole. 30 - - - - - 40 N N N -

Melted, punched hole Other 120 N N N - - Melted, punched hole 20 Fabric: 9. 100% wool (suiting). Orientation: Width (Length) Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 N

N

-

No sign of scorching 10 N

N

-

No sign of scorching 15 N N

N

- Slight scorching 20 N (N) N (N)

N (N)

- More scorching, light fuming 30 N (N) N (N)

N (N)

- Scorching and moderate fuming 40 Y (N) N (Y) N (Y) Y (N)

• (Y)

Transient flames for approx. three seconds. 45 Y (Y) Y (Y)

Y (Y)

- Fabric: 10. 100% cotton FR treated (corduroy). Orientation: Length Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 N

N

-

No signs of fumes/scorching 10 N

N

-

No signs of fumes/scorching 15 N

N

-

No signs of fumes/scorching 20 N N

N

Some scorching, hole developed 30 - - - - - 40 N N N - - Some scorching, punched hole Other 120 N N N N N Some scorching, punched hole. Expected from FR? 21 Fabric: 11. Stretch denim (cotton). Orientation : Width (Length) Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 N

N

-

No sign of fumes or scorching 10 N(N) N(N)

N(N)

- Slight scorching 15 Y(N) Y(Y) Y(Y) N(Y) N Heavy scorching before flame 20 N(Y) Y(Y) Y(Y) N(Y) Y(Y) Heavy scorching before flame 30 Y Y Y Y Y Fabric: 12. Polyester fabric and wadding. Width (Length) Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 N (N) N (N)

N (N)

- Some scorching 10 Y (Y) Y (N)

Y (N)

-

15

-

-

20

-

-

30

-

-

40

-

-

Other

-

-

22

Fabric: 13. 100% polyester. Orientation: Width (Length) Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 N

N

-

Small holes from sparks 10 N

N

-

Small holes from sparks 15 N

N

-

Scorching and small hole 20 N (N) N (N)

N (N)

- Scorching and larger hole 30 N

N

-

Scorching and larger hole 40 N (N) N (N)

N (N)

- Scorching and larger hole Other (120) N N

N

Scorching large hole punched. Sparks not contacting fabric. Fabric: 14. 100% polyester over 100% wool (two fabrics). Orientation: Width (Length) Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 - - - - - 10 - - - - - 15 - - - - - 20 N (N) N (N) N (N) - - Scorching, melted hole. 30 Y (N) Y (Y) Y (Y) (N) (Y) 40 - - - - - Other - - - - - 23 Fabric: 16. “Proban” treated cotton. Unlaundered Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks

5

-

-

10 N N

N

Slight scorching 15 N N N - - 20 N Y Y Y - Fuming and charring, flaming for 2 seconds only. 30 Y Y Y Y Y Fuming, charring, flaming for 2 seconds only. 40 Y Y Y Y Y Fuming, scorching, flaming for 2 seconds only. Other - - - - - Fabric: 17. “Proban” treated cotton. Laundered Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks 5 N N N - - 10 Y Y Y - - Flaming stopped after 2 seconds 15 Y Y Y - - Flaming stopped after 2 seconds 20 - - - - - 30 - - - - - 40 - - - - - Other - - - - - 24 Fabric: 18. 100% cotton (fabric 15) with grease contaminant. Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks

5

-

-

10 N N

N

- 15 Y Y

Y

-

20

-

-

30

-

-

40

-

-

Other

-

-

Fabric: 19. 100% cotton (fabric 15) with paint contaminant. Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks

5

-

-

10

-

-

15

-

-

20 N N

N

-

30

-

-

40 N N

N

- Other 60 N N

Y

- Flaming only on non painted area. 25

Fabric: 20. 100% polyester (fabric 13) with grease contaminant. Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks

5

-

-

10

-

-

15

-

-

20 N N

N

- Scorching and hole punched through

30

-

-

40 N N

Y

- Scorching, hole punched flaming in one test. Other (60) N

N

-

Fabric: 21. 100% polyester with paint contaminant. Time sparks applied (seconds) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Remarks

5

-

-

10

-

-

15 N N

N

- Slight hole and light fumes 20 N Y Y Y Y Hole but continued flaming 30 Y Y Y Y Y Holed but continued flaming.

40

-

-

Other

-

-

26

APPENDIX 2 – Test data obtained from the “flame-spread rates” study of the ten generic fabrics and variants. Fabric Time in seconds to reach indicated flame spread distance in mm and (inches) – Reference Horizontal Sample. Number (Table 1) and repeat number 0 (0) 25.4 (1) 50.8 (2) 76.2 (3) 101.6 (4) 127.0 (5) 152.4 (6) 177.8 (7) 203.2 (8) 228.6 (9) 254.0 (10)# 1.1 16 28 31 40 52 59 67 77 84 95 108 1.2 13 20 29 36 44 52 60 69 77 85 95 2.1 20 32 43 54 68 79 91 101 110 121 134 2.2 17 30 38 51 64 75 84 98 108 120 136 3.1 0 No flame spread 3.2 0 4.1 0 4.2 0 9.1 0 No flame spread 9.2 0 10.1 0 10.2 0 11.1 44 67 88 108 128 145 163 183 199 217 235 11.2 43 64 83 101 119 136 152 168 186 202 218 12.1 23 45 67 92 111 135 150 170 183 201 222 12.2 25 44 64 85 105 122 142 160 175 192 210 13.1 2 No further flame spread 13.2 3 14.1 No flame spread 14.2 16.1 0 No flame spread from ignition source 16.2 0 27

1. 1

0 No flame spread from ignition source 17.2 0 18.1 25 35 47 56 66 76 85 95 105 117 131 18.2 25 36 49 61 72 85 93 102 115 125 137 19.1 25 42 60 75 90 107 125 141 157 178 197 19.2 26 42 61 77 95 110 126 141 158 174 192 20.1 7 13 20.2 Very slight propagation from source (few mm) before extinguishing 21.1 6 16 27 43 52 60 67 75 86 94 98 21.2 6 12 20 27 34 48 58 70 82 89 94 28

Fabric Reference Number (Table 1) and repeat number Time in seconds to reach indicated flame spread distance in mm and (inches) – Angled Sample (25O) 0 (0)* 25.4 (1) 50.8 (2) 76.2 (3) 101.6 (4) 127.0 (5) 152.4 (6) 177.8 (7) 203.2 (8) 228.6 (9) 254.0 (10)# 1.1 4 7 9 11 13 24 1.2 4 19 2.1 3 17 2.2 5 27 3.1 4 6 36 3.2 4 9 40 4.1 6 No further flame spread 4.2 6 15 No further flame spread 9.1 0 No flame spread 9.2 0 10.1 12 28 37 46 No further flame spread 10.2 10 20 No further flame spread 11.1 12 18 23 27 31 34 36 40 43 47 50 11.2 13 19 24 28 32 35 39 41 45 48 50 12.1 10 13 18 20 23 26

30

-

48 12.2 10 15 20 22

25

-

-

50 13.1 12 20 No further flame spread 13.2 15 20 14.1 No flame spread 14.2 16.1 13 No further flame spread 16.2 12 17.1 13 No further flame spread 17.2 13 18.1

10

- 15

16

-

26

- 30 29

1. 2

7 10 13

15

-

22

- 29 19.1

8

-

15

20

30

- 36 19.2 10 14 17 20 24

26

35

40

44 20.1 Very slight flame spread before extinguishing 20.2 6 10 No further flame spread 21.1 3

7

10

15

20

25 28 21.2

3

-

12

14

21

23

30 30

APPENDIX 3 – Ancillary measurements During the study a number of ancillary measurements were made to investigate various observations connected with the procedures. For completeness these are presented here. Trials for a substitute for fabric 2 The new fabric (15) was required as the original (fabric 2) was discontinued before the contamination trials (which required fabric 2), could take place. Table A3.1 shows the comparative data obtained for these two fabrics. Table A3.1 - Comparative data for an original and replacement cotton fabric Fabric Mass/unit area (kg/m2) Time-to-ignition (seconds) Mean horizontal flame spread (mm/min) Mean horizontal flame spread (repeat) (mm/min) 2 0.156 20 134 128 15 0.161 20 98.7 101 Although fabric 15 showed some difference in flame spread rate, the difference was considered small enough (in the context of the range of flame-spread rates observed throughout the whole study) to enable the fabric to be used as a substitute for the discontinued fabric 2 for the “contamination” trials in this study. Establishing operator dependence for the ignition study The development work for subjecting fabric samples to a consistent shower of sparks from an angle grinder, initially utilised a sample holder made from “handy angle”. This was to enable a variety of configurations of fabric to be readily trialled to ensure the final design gave the most repeatable and reliable results. On completion of the development work, a final design was constructed from an open mesh wire sheet to the final required dimensions. However, early trials with the new sample holder, and with a different test operator showed significant differences in times-to-ignition; i.e. giving shorter times with the new sample holder. In order to check whether this was due to the change of sample holder or operator, two fabrics were chosen and trialled by both operators with both types of sample holder. 31

Table A3.2 below summarises the results: Table A3.2: Results from test using different operators and the two different fabric sample holders Sample Operator Fabric Run 1 Run 2 Run 3 holder (time/ignition) (time/ignition) (time/ignition) Handy angle JC 9 60sec/no 60sec/no 60sec/no Open wire JC 9 30sec/no 40s/yes 40sec/no 45s/yes 40sec/no 45s/yes Handy angle JC 12 5sec/no 10sec/no 5sec/no 10sec/yes 20sec/yes 10sec/yes Open wire JC 12 5sec/no 5sec/no 5sec/no 10sec/yes 10sec/yes 10sec/yes Handy angle PJF 9 60sec/no 60sec/no 60sec/no Open wire PJF 9 30sec/no 40 sec/no 40sec/no 45sec/yes 45sec/yes 45sec/yes Handy angle PJF 12 5se/no 5sec/no 5sec/no 10sec/yes 10sec/yes 10sec/yes Open wire PJF 12 5sec/no 5sec/no 5sec/no 10sec/yes 10sec/yes 10sec/yes As shown, the differences appear to be due to the different sample holder rather than the different operator, although the differences with sample 12 between handy angle and open wire holder were small. The differences are tentatively ascribed to the larger “heat sink” properties of the handy angle holder, over the open wire mesh holder, resulting in somewhat longer ignition times (where indeed ignition did actually occur) with the handy angle holder. Elemental analysis of a flame retarded fabric Due to the surprising result with one FR fabric (3) which ignited in 10 seconds (although it possessed a low flame spread) it was considered desirable to obtain a “Proban” treated cotton fabric (fabric 16) and subject this to elemental analysis to ensure the loading of FR treatment appeared consistent with correct “Proban” treatment. 32

Table A3.3 gives the result of a C, H, N and P analysis fabric 16 (“Proban” treated cotton). Table A3.3 % carbon expressed as C 41.13 41.12 % hydrogen expressed as H 6.29 6.36 % nitrogen expressed as N 2.12 2.02 % phosphorus expressed as P 2.33 2.28 Through discussions with a major supplier manufacturer of “Proban” treated fabrics an elemental phosphorus content of 2.3% to 3% was considered a satisfactory “loading” in this application [6]. 33

34

REFERENCES 1. “FIREX” – HSE Technology Division’s Database. 2. “Hand grinder sparks: Development of a test for ignition and flame spread on work wear. An apparatus and procedure to compare times to ignition.” June 2002. Fardell P.J. BRE Client Report 206550. 3. “Hand grinder sparks: Development of a test for ignition and flame spread on work wear. An apparatus and procedure to compare flame-spread rates. August 2002. Fardell P.J. BRE Client Report 206551. 4. Title 49 – Transportation #571.302 Federal Motor Vehicle Safety Standard (FMVSS) “Standard 302: Flammability of interior fabrics” US Federal Register Vol. 36.N0 232. 5. “Textiles. Commercial laundering procedure for textile fabrics prior to flammability testing. (reduced procedure)”BS EN ISO 10528 :1995. 6. Phosphorus levels in “Proban” treated textiles. Northenden Textiles Ltd., Rengate House, 221 Palatine Road, Didsbury, Manchester M20 2EE. Mr. Tony Purcell (Private Communication). 7. Portwest Clothing Ltd. Market Street, Goldthorpe, Rotherham S. Yorkshire S63 9HB 35

Printed and published by the Health and Safety Executive C30 1/98 Printed and published by the Health and Safety Executive C1.10 05/04

ISBN 0-7176-2841-8 RR 222 78071 7 62841 4 £10.00 9

Annex B MSN 1731 (M+F)

1 MERCHANT SHIPPING NOTICE MSN 1731 (M+F) Summary • This Notice gives notice of new regulations governing the provision of personal protective equip- ment, the Merchant Shipping and Fishing Vessels (Personal Protective Equipment) Regulations

1. These Regulations supersede the Merchant Shipping (Protective Clothing and Equipment)

Regulations 1985 (S.I. No. 1664), and come into force on 25 October 1999 • Annex 1 gives the design standards for personal protective equipment in use on board ships, for specified work activities and situations, in order to comply with regulation 5(2)(a) of the new Regulations. Introduction The Merchant Shipping and Fishing Vessels Personal Protective Equipment Regulations 1999 This Merchant Shipping Notice is an integral part of the Merchant Shipping and Fishing Vessels (Personal Protective Equipment) Regulations 1999. Notice to employers of crew, masters, safety officers and safety representatives. This Notice supersedes Merchant Shipping Notices No. M1195 and M1358. 1. The Regulations require employers to ensure that personal protective equipment (PPE) is provided for their workers who are engaged in, or at risk from, a hazardous work activity on board a United Kingdom ship1. 2. The Regulations are subject to the general rule that use of PPE is always a last resort, where risks cannot be avoided or reduced to a safe level by means of collective protection, or safe systems of work. 3. PPE must be provided free of charge to the workers, except that, where use of the equipment is not exclusive to the work place, workers may be required to contribute towards the cost. 1 “worker” includes trainees and apprentices, but does not include persons who are training in a sail training vessel. 4. Where, traditionally, workers provide their own PPE, the employer remains responsible for ensuring that workers are equipped with appropriate PPE, and that they use it when engaged in work of the types outlined in Annex 1. 5. The equipment issued must be “suitable”, which is defined as : (a) in relation to any work process described in [this] Merchant Shipping Notice MSN 1731 (M+F), of the kind and to the standard specified [in that Merchant Shipping Notice], in relation to that work process; (b) appropriate for the risks to which he worker is exposed and to the task which he is performing, without itself leading to any increased risk;

2 (c) correctly fitting the worker, or capable of being adjusted to fit; (d) taking into account ergonomic requirements and the worker’s state of health; and (e) compatible with any other equipment the worker has to use at the same time, so that it continues to be effective against the risk” 6. In addition, the employer must ensure that the PPE supplied is easily accessible, and properly stored and maintained, and that where appropriate, instructions are available to the workers who are required to carry out any maintenance. The equipment must be regularly inspected, in accordance with the manufacturers instructions, and its operation checked. Respiratory protective equipment must be always be checked before and after use. 7. The employer must ensure, so far as is reasonably practicable, that PPE issued under the regulations is used as instructed - eg that workers do not use it for a purpose for which it is not designed, and that it is put on and worn correctly. 8. Workers must receive adequate and appropriate training so that they are aware of the risks against which the PPE is designed to protect them, and of when and how to use it and look after it correctly. This may include demonstrations of the wearing of PPE, where appropriate. 9. Workers are required to wear and use the PPE which has been issued to them when appropriate, and to comply with any training and instruction provided. Standards of design and manufacture 10. The specifications for PPE are set out in the Annex. The list covers the PPE most commonly used on ships, but is not exclusive. 11. The letters “EN” stand for “European Norm”. Where no “EN” standard is available, a BS standard is quoted. The letters “BS” refer to a British Standard. The standards are those to which the clothing and equipment should comply and the date which appears will be the date on which the latest revision of the relevant Standard was published, including all amendments at the date of this Merchant Shipping Notice. 12. Any reference to an EN or BS standard contained in the annex means that standard or an alternative Standard which provides, in use, equivalent levels of safety, suitability and fitness for purpose. 13. The standards of equipment given in this Merchant Shipping Notice do not apply to life saving appliances or other equipment which is subject to the Merchant Shipping (Marine Equipment) Regulations. (S.I. 1999/1957). 14. Publications mentioned in Annex 1 are available from: “BS” and “EN” Specifications: The British Standards Institution 389 Chiswick High Road London W4 4AL The Code of Practice, “Noise Levels in Ships” is available from The Stationery Office. MSPP3 (Seafarers Health and Safety) The Maritime and Coastguard Agency Bay 2/1 Spring Place 105 Commercial Road Southampton SO15 1EG Tel: 01703 329390 Fax: 01703 329251 August 1999 © Crown copyright 1999 [MC 122/6/053] An executive agency of the Department of the Environment, Transport and the Regions

3 ANNEX 1 STANDARDS OF PERSONAL PROTECTIVE EQUIPMENT Note: all protective clothing should conform to EN 340 : 1993 - ‘Protective clothing. General requirements.’ Work activity Protective clothing and equipment to be provided Full title of Standard 1 Any process or activity involving a reasonably foreseeable risk to the head from falling or moving objects. Head protection EN 397 : 1995 Specification for industrial safety helmets. 2. When working in areas where the circumstances involve a reasonably foresee- able risk to the head from bruising or abrasion. Scalp protection to EN 812 : 1997 Industrial bump caps. 3. When entering or working in a space or working with machinery or equipment where the noise level exceeds 85dB(A). Hearing protection complying with section 10 and appendix 3 of the Code of Practice for Noise Levels in Ships published by the Department of Transport (1990): EN 352-1 : 1993 EN 352- 2 : 1993 EN 352- 3 : 1996 EN 458 : 1994 Ear muffs. Ear plugs. Ear muffs attached to an industrial safety helmet. Hearing protectors. Recommen- dations for selection, use, care and maintenance. 4. Welding and gas cutting. Eye and face protection to EN 175 : 1997 Personal protection. Equipment for eye and face protection during welding and allied processes. EN 166 : 1995 Personal eye protection. Specifications. EN 379 : 1994 Specification for filters with switchable or dual luminous transmittance for personal eye protectors used in welding and similar operations. EN 169 : 1992 Specification for filters for personal eye protection equipment used in welding and similar operations.

4 Work activity Protective clothing and equipment to be provided Full title of Standard Body protection to EN 470-1 : 1995 Additional protection may be required in some situations (eg for particularly intense welding/cutting operations) Protective clothing for use in welding and allied processes. General requirements. Electric arc welding (in addition to above) Safety footwear to BS 7193 Specification for lined lightweight rubber overshoes and overboots 5. Any work activity in which there is a reasonably foreseeable risk of injury to the eye from particles, frag- ments or injurious substances. Eye protection to EN 166 As above. 6. a) Any work activity involving working in an atmosphere which is likely to be hazard- ous to health. Protection against nuisance dust mist, particles and dust of low toxicity. Note - The following items should be selected and maintained according to BS 4275 : 1997 Disposable dust respirators conforming to EN 149 : 1991 General purpose dust respirators conform- ing as appropriate to one of the following: EN 136 : 1998 EN 140 : 1998 EN 141 : 1990 EN 143 : 1990 EN 371 : 1992 EN 372 : 1992 EN 1827: 1999 Guide to implementing an effective respiratory protective device programme. Specification for filtering half- masks to protect against particles. Respiratory protective devices: Full face masks. Respiratory protective devices: Half masks and quarter masks Respiratory protective equipment: Gas filters and combined. Specification for particle filters used in respiratory protective equipment. Specification for AX gas filters and combined filters against low boiling organic compounds used in respiratory protective equipment. Specification for SX gas filters and combined filters against specific named compounds used in respiratory protective equipment. Half masks without inhalation valves, with separate filters to protect against gases or gases and particles or particles only.

5 Work activity Protective clothing and equipment to be provided Full title of Standard b) Protection against toxic dusts and gases of low toxicity. Respirators conforming as appropriate to one of the following: BS 7355 (EN 136) BS 7356 (EN 140) EN 141 EN 143 EN 371 EN 372 EN 405 : 1992 Note: particulate filters may be incorporated for some applications. As above. As above. As above. As above. As above. As above. Valved filtering half masks for gases or gases and particles. EN 1827: 1999 Half masks without inhalation valves, with separate filters to protect against gases or gases and particles or particles only. c) Protection against toxic dust. Powered dust respirators, powered dust hoods and blouses conforming as appropriate to one of the following: EN 136 EN 143 EN 12942: 1998 EN 12941: 1998 As above (note: this only applies to the mask). As above. Specification for power assisted particle filtering devices incorporating full face masks, half masks or quarter masks. Respiratory protective devices. Specification for powered particle filtering devices incorporating helmets or hoods. d) Protection against highly toxic atmospheres; or where there is oxygen deficiency; or as an alternative to the items above, where suitable. Breathing apparatus conforming to: EN 1146 : 1997 (for self-rescue only) “Escape sets” Respiratory protective devices for self rescue. Self contained open- circuit compressed air breathing apparatus incorporating a hood (compressed air apparatus with hood). Requirements, testing, marking.

6 Work activity Protective clothing and equipment to be provided Full title of Standard EN 137 : 1993 Specification for respiratory protective devices: self contained open-circuit compressed air breathing apparatus. EN 138 : 1994 Respiratory protective devices. Fresh air hose breathing apparatus for use with full face mask, half mask or mouthpiece assembly. EN 139 : 1994 Respiratory protective devices. Compressed air line breathing apparatus for use with a full face mask, half mask or mouthpiece assembly. Requirements, testing, marking. EN 269 : 1994 Respiratory protective devices. Powered fresh air hose breathing apparatus incorporating a hood. EN 270 : 1994 Respiratory protective devices. Compressed air line breathing apparatus incorporating a hood. Requirements, testing, marking. EN 271 : 1995 Respiratory protective devices: Compressed air line or powered fresh air hose breathing apparatus incorporating a hood for use in abrasive blasting operations. EN 402 : 1993 Respiratory protective devices for escape. Self contained open-circuit compressed air breathing apparatus with full face mask or mouthpiece assembly. 7. Any process or activity involving working in an area where there is a reasonably foreseeable risk of injury from substances which are corro- sive or likely to be absorbed through the skin. Protective overalls, gloves or head gear, whichever is appropriate: EN 340 : 1993 EN 465 : 1995 Protective clothing: General requirements. Protective clothing. Protection against liquid chemicals. Performance requirements for chemical protective clothing with spray-tight connections between different parts of the clothing (Type 4 equivalent).

7 Work activity Protective clothing and equipment to be provided Full title of Standard EN 466 : 1995 EN 467 : 1995 Protective clothing. Protection against liquid chemicals. Perfor- mance requirements for chemical protective clothing with liquid-tight connections between different parts of the clothing (Type 3 equivalent). Protective clothing. Protection against liquid chemicals. Performance requirements for garments providing protection to parts of the body. 8. Any process or activity involving a reasonably fore- seeable risk of injury to the hands unless the use of hand protection would increase the risk. Hand protection conforming as appropriate to : EN 374 Protective gloves against chemicals and micro-organisms. EN 374-1 : 1994 Terminology and performance requirements. EN 374- 2 : 1994 Determination of resistance to penetration. EN 374- 3 : 1994 Determination of resistance to permeation by chemicals. EN 388 : 1994 Protective gloves against mechanical risks. EN 407 : 1994 Protective gloves against thermal risks EN 420 : 1994 General requirements for gloves. EN 511 : 1994 Protective gloves against cold. 9. Any process or activity involving particular risk of injury to the feet. Foot protection conforming to : EN 345 ; or Safety footwear for professional use. EN 346, whichever is appropriate: Protective footwear for professional use. EN 345- 1 : 1992 Specification. EN 345- 2 : 1996 EN 346- 1 : 1992 Additional specifications. Specification. EN 346-2 : 1996 Additional specifications. EN 347-1 : 1992 Occupational footwear for professional use. EN 347-2 : 1996 Additional specifications.

8 Work activity Protective clothing and equipment to be provided Full title of Standard

1. Work aloft or in any other

area where there is a reason- ably foreseeable risk of falling a distance of more than 2 metres. Safety belt or harness and associated lanyard conforming to the following: EN 353-1 : 1992 EN 353- 2 : 1992 EN 354 : 1992 EN 355 : 1992 EN 360 : 1992 EN 361 : 1992 EN 362 : 1992 EN 363 : 1992

• or where the use of portable ladders is

necessary, such ladders to be used in accordance with Chapter 15 of the Code of Safe Working Practices for Merchant Seamen. Specification for guided type fall arresters on a rigid anchorage line. Specification for guided type fall arresters on a flexible anchorage line. Personal protective equipment against falls from a height. Lanyards. Personal protective equipment against falls from a height. Energy absorbers. Personal protective equipment against falls from a height. Retractable fall arrangements. Personal protective equipment against falls from a height. Full body harnesses. Personal protective equipment against falls from a height. Connectors. Personal protective equipment against falls from a height. Fall arrest systems.

1. Any work carried out from an

overside position or in an ex- posed position where there is reasonably foreseeable risk of falling or being washed over- board or any work carried out in or from a ship’s boat. A lifebuoy with sufficient line attached ready for immediate use and either a Maritime and Coastguard Agency approved lifejacket or a lifejacket conforming as appropriate to one of the following, taking into account the area of operation: EN 394 : 1994 EN 396 : 1993 Life jackets and personal buoyancy aids. Additional items. Life jackets and personal buoyancy aids. Life jacket 150.

9 Work activity Protective clothing and equipment to be provided Full title of Standard EN 399 : 1993 Partially inherent lifejackets must have at least 89 N of inherent buoyancy; and with the inflatable sections relying on automatic inflation. Life jackets and personal buoyancy aids. Life jacket 275.

1. Any work activity where it is

necessary to carry out repair or maintenance work on or near exposed live electrical equipment and there is a reasonably foreseeable risk of injury. Rubber gloves conforming to: BS 697 : 1986 EN 60903 : 1992 Protective sleeves conforming to EN 60984 : 1993 An insulating mat (except where specially insulated flooring is installed) conforming to BS 921 : 1976 Rubber soled footwear (no standard necessary). Note - gloves, sleeves and mats should protect against the appropriate voltage. Specification for rubber gloves for electrical purposes [4 classes of gloves rated at 650v and above] Gloves and mitts of insulating material for live working. Sleeves of insulating material for live working. Specification. Rubber mats for electrical purposes.

1. Any work activity involving

a reasonably foreseeable risk of injury from vehicle move- ment eg during ro-ro operations. Suitable high-visibility garment conforming to EN 471 : 1994 Specification for high-visibility warning clothing.

1. Any work process involving

exposure to heat EN 531 : 1995 Protective clothing for industrial workers exposed to heat (excluding fire-fighters’ and welders’ clothing).

1. Work in engine rooms or any

area where there is a risk of fire. Overalls made of fabric of low flammability - eg

• natural fibre, high cotton content;
• non-flammable clothing

as appropriate Cotton or cotton and polyester clothing with flame-retardant finishes are available to protect against sparks and flame.

FD06038 ｜ 2010_英国_MAIB_man-overboard-from-twin-rig-trawler-aquarius-with-loss-of-1-life__Aquarius_CombinedAnnexes.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2010 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A Annang Nuertey (deceased) contract of employment

Annex B MGN 430 (F) Checks on Crew Certification and Drills

Maritime and Coastguard Agency Logo

MARINE GUIDANCE NOTE

MGN 430 (F)

FISHING VESSELS: Checks on Crew Certification and Drills

Notice to all Owners, Operators, Managing Agents, Skippers and Crew This Note should be read in conjunction with MGN 411 - Training and Certification Requirements for the Crew of Fishing Vessels and their applicability to Small Commercial Vessels and Large Yachts, MSN 1770, The Code of Safe Working Practices for the Construction and Use of 15m Length Overall to less than 24m Registered Length Vessels, the Fishing Vessels (EC Directive on Harmonised Safety Regime) Regulations 1999, No.2998 as amended by SI 2003/1112 and the Fishing Vessels (Safety Provisions) Rules 1975, as amended.

PLEASE NOTE:- Where this document provides guidance on the law it should not be regarded as definitive. The way the law applies to any particular case can vary according to circumstances - for example, from vessel to vessel and you should consider seeking independent legal advice if you are unsure of your own legal position.

Summary. This Note provides guidance on the checks that MCA surveyors will make regarding crew certification and drills. Surveyors will:

• ensure that the correct Certificates of Competency are held and Safety Training courses have been undertaken by skippers and crew; • check that written health and safety policies are in place and completed risk assessments have been carried out. • witness emergency drills as part of the Renewal and Intermediate surveys on the vessel or at any other time as deemed necessary by the MCA. • confirm that emergency drills (Fire, Collision/Grounding, Man Overboard, Abandon ship, Anchoring) are practiced monthly and when a new crew member joins the vessel. • if practicable, and when there is no evidence that drills have been conducted and it is considered the crew are not trained for an emergency, ask vessels to proceed from the harbour to a safe anchorage to undertake anchoring drills. This increases the validity of the drill and provides a more challenging, realistic environment.

The Note sets out various scenarios for different types of emergency drills, which the MCA surveyors may ask to see, sets out what the crew need to demonstrate and the key issues (Cont over.)

• 1 -

a surveyor will be looking at. The Note sets out the guidance on possible requirements of skippers and crew during emergencies which can be demonstrated during drills and musters. It should be noted that this guidance is generic and it is the owner, skipper and crew who should decide what steps are necessary and who should undertake them.

Fishing vessels – Crew certification and emergency drills

Introduction 1.1. The Marine Accident Investigation Branch (MAIB) report entitled “Analysis of UK Fishing Vessel Safety 1992 to 2006” recommends that the MCA ensures that the current mandatory training requirements for fishermen are strictly applied. Other accident reports have highlighted the lack of emergency procedures and the need to carry out drills which are a statutory requirement. Owners, skippers and crew should note that it is the flag state’s responsibility (i.e. the MCA for UK flagged vessels) to take such measures as it may deem necessary to ensure that crews are adequately trained in their duties in the event of emergencies.

Actions by Surveyors 2.1 Surveyors who undertake surveys or inspections on fishing vessels will check the following documents in particular before endorsing or renewing United Kingdom Fishing Vessel Certificates or International Fishing Vessel Certificates.

• Certificates of Competency for the Deck and Engineer Officers. Details of specific requirements are contained within the Fishing Vessels (Certification of Deck Officers and Engineer Officers) Regulations 1984 SI No. 1984/1115.

• Certificates of Equivalent Competency for Deck and Engineer Officers. Details of specific requirements are set out in the Fishing Vessels (Certification of Deck Officers and Engineer Officers) Regulations 1984 SI No. 1984/1115 and MSN 1825 (F) Certificates of Equivalent Competency: Fishing Vessels.

• Certificates confirming that fishermen not holding Certificates of Competency have attended the basic safety training courses listed in the Fishing Vessels (Safety Training) Regulations 1989 SI No. 1989/126.

1. 2

Further guidance on the requirements set out in 2.1 above can be found in MGN 411.

1. 3

Surveyors will also check that when the crew is greater in number than five, there is in place a written health and safety policy and that risk assessments have been completed for the various operations undertaken on board. Both these requirements are contained within SI No. 1997/2962 - Merchant Shipping and Fishing Vessels (Health and Safety at Work) Regulations 1997. Without a written policy and assessments, operators will have no evidence or defence that they have actually met their obligations.

1. 4

Attending surveyors will also witness the crew undertaking suitable emergency drills. Guidance for surveyors, skippers and crew specific to fishing vessels is annexed to this notice. If practicable, and when there is no evidence that drills have been conducted and it is considered the crew are not trained for an emergency, vessels will be asked to proceed from the harbour to a safe anchorage to undertake anchoring drills. This increases the validity of the

• 2 -

drill and provides a more challenging, realistic environment. The attending survey will take into account the weather conditions and other safety aspects when deciding if the vessel will be asked to proceed from harbour for a drill.

1. 5

In addition, attending surveyors will satisfy themselves that the emergency drills as required by either the Fishing Vessels (EC Directive on Harmonised Safety Regime) Regulations 1999 SI No. 1999/2998 as amended, the Fishing Vessel (Safety Provisions) Rules 1975 SI No. 1975/330 as amended or by MSN1770(F) have been carried out and correctly documented.

1. 6

Providing that all of the above are found to be satisfactory and in accordance with the relevant legislation for vessels of 15m Length overall and above, the vessel’s certificate may be renewed/endorsed. If however any of the above is found to be unsatisfactory or not proven then it will be noted on the Report of Inspection/Survey as a deficiency and the vessel’s certificate will not be renewed or endorsed until drills are completed satisfactorily.

Action by Owners

1. 7

In respect of fishing vessels of less than 15m Length Overall, from 1 July 2010 the MCA now issues a Small Fishing Vessel Certificate. If your vessel was inspected on or after this date you will be required to present a copy of this Certificate to the Register of Shipping and Seamen (RSS) as proof of compliance with the Small Fishing Vessel Code MSN 1813 when registering or re-registering your vessel. Vessels inspected before 1 July 2010 will need to present the Report of Inspection (MSF 1606) which would have been provided to the owner at the time.

More Information

Fishing Vessels Safety Unit Vessel Policy Branch Maritime and Coastguard Agency, Bay 2/30 Spring Place, 105 Commercial Road, Southampton, SO15 1EG.

Tel :

+44 (0) 23 8032 9154. Fax :

+44 (0) 23 8032 9104. e-mail:

fishing @mcga.gov.uk.

General Inquiries: infoline@mcga.gov.uk .

MCA Website Address: www.mcga.gov.uk .

File Ref:

MS88/1/426

Published:

January 2011

Please note that all addresses and

telephone numbers are correct at time of publishing.

© Crown Copyright 2011.

Safer Lives, Safer Ships, Cleaner Seas.

When printed by the MCA the material used contains a minimum 75% post-consumer waste paper

• 3 -

Annex to MGN 430 - Fishing vessels – Crew certification and emergency drills

The investigations by the MAIB include incidents where crews have had to abandon ship in extremely poor weather conditions. One vessel grounded on a lee shore with engine trouble, another caught fire, and another took on a great deal of water whilst preparing to fish in heavy weather.

It is mandatory that emergency drills be carried out monthly on all United Kingdom fishing vessels over 15 metres Length Overall and a record of these drills should be entered into the vessel’s log book. An entry should be made if one of these drills has been missed along with a reason why. An emergency drill should also be carried out when a new crew member joins the vessel.

The following is guidance for musters and drills witnessed during inspections and surveys.

The emergency drill could take the form of:-

• Engine Room, Accommodation or Factory Deck Fire; • Collision/ Grounding; • Man Overboard; • Abandon ship; and • Emergency Anchoring

The drill should refresh basic safety training and add an element of reality of working as part of a team onboard their own vessel. In witnessing a safe and effective drill, it is important that as many of the regular crew are present as possible.

Drills cannot replace the written risk assessment but are a vital part of the necessary control measures within that risk assessment.

It is essential that all crew members undertake drills and play their part in the safe day to day running of the vessel.

• 4 -

Anchor Drill

Purpose

If practicable, and taking into account the considerations in paragraph 2.4 of this MSN, the vessel will proceed from the harbour or dock to a safe anchorage position where an anchoring drill will be carried out.

The crew will demonstrate:

• that the vessel can safely deploy and recover an anchor within a reasonable time to simulate an emergency anchoring scenario; and

• the deployment of the anchor, with the crew dressed suitably including gloves and correctly donned lifejackets.

Scenario

The vessels steering gear has failed in the fairway into a busy harbour. Bring the vessel to anchor and show the correct day signal.

Key Issues

The key issues are:

• crew dressed in Personal Protective Equipment (PPE);

• anchor handling and methodology safe;

• anchor cable marked to show length of cable deployed;

• crew aware of length of cable remaining onboard;

• communication between wheelhouse and anchor party acceptable; and

• anchor deployed safely and operation acceptable.

• 5 -

Muster Stations Drill

Purpose

The crew will demonstrate:

• awareness of their personal muster station and that they can don a lifejacket quickly and correctly;

• that those allocated to the lifeboat or liferaft deck can prepare a liferaft for manual launching; and

• that those responsible for other duties undertake them quickly and competently.

Scenario

The vessel will simulate working conditions as if the vessel were fishing with crew asleep in cabins. The crew alarm will be activated and the crew will go to muster stations dressed appropriately to abandon ship, with warm clothing and lifejackets correctly donned.

Crew “asleep” in the cabin will exit the cabin using the emergency escapes. The “duty watch” will close all doors and vents on way to the muster station.

Key Issues

The key issues are:

• crew alarm activated and suitable;

• crew to correct muster stations;

• crew suitably dressed in warm clothing and lifejackets;

• liferafts made ready for deployment;

• duties as posted on Muster List completed; and

• stowage of lifejackets, risks understood and acceptable.

The crew will be accounted for and a report made to the skipper.

Liferafts will be “made ready” to deploy manually but will not be deployed.

Lifejacket Storage Position

Careful thought needs to go into deciding where to store the lifejackets, they need to be stored where there is airflow and where they will not get damp. Access to the lifejackets should be the main concern of the crew, they will be using emergency exits and access to inside the vessels should not be allowed. Can the crew therefore access the lifejackets easily?

Lifejacket donning instructions should relate to the lifejacket types carried on the vessel. More than one type of lifejacket means displaying more than one type of lifejacket donning instructions which need to be displayed where the lifejackets are stored and in crew areas.

• 6 -

Misuse of the Emergency Alarm

The emergency alarm should be distinctly different from the call system used to call the crew out on deck to haul the gear.

• 7 -

Fire Drill

Purpose

To demonstrate:

• knowledge of ship board equipment and vents;

• donning and use of breathing equipment;

• practical use of hoses and nozzles;

• safety considerations; and

• use of fixed fire suppression gas systems.

During this fire drill the main fire fighting hoses will be deployed and demonstrated. Once the surveyor is satisfied that the condition of the hoses and nozzles are to standard then the vessel will simulate “dead ship” and the emergency lights, fire hoses and equipment will be tested.

Scenario

Fire reported in area of high fire risk e.g. engine room, galley or spaces with electrical heaters. Prepare to fight the fire and prevent the fire from spreading to other areas of the vessel.

Key Issues

The key issues are

• knowledge of shipboard fire fighting equipment;

• all vents closed and fans stopped;

• fire mains and hoses proved;

• emergency pumps and hoses proved;

• breathing equipment and safety procedures acceptable;

• use of gas suppression systems understood and appreciated; and

• limitations of fire fighting equipment carried appreciated.

• 8 -

Man Overboard Drill

Purpose

To demonstrate

• that the crew can conduct an effective search of the vessel;

• the crew are able to quickly launch the rescue boat to recover a person from the water; and

• the rescue boat crew are aware of the effects of hydrostatic squeeze and how it will affect a casualty suffering from hypothermia.

Scenario

The crew will undertake a “man overboard” drill and will launch and man the rescue boat. The rescue boat will be readied and swung out to a side specified by the attending surveyor and launched into the water in a safe and controlled manner.

A member of the crew is believed to fallen overboard and has not been seen for some time. The attending surveyor will indicate when this crew man was last seen and if the crewman was seen to have fallen overboard.

Key Issues

The key issues are

• rescue boat stores to correct scale (as applicable) and serviceable;

• rescue boat launched in safe manner, crew suitably dressed;

• rescue boat and davit in serviceable condition; and

• search undertaken of vessel for missing crew member.

On completion of this drill the rescue boat will be recovered to the vessel and readied for immediate use.

Man Overboard Recovery System

There are several “man overboard” recovery systems in use on board fishing vessels at this time, Jason’s Cradle and Markus Net to name just two. These systems are permitted to replace rescue boats only when an exemption has been applied for and granted.

Crews should be well trained in the use of these systems and appreciate the limitations of the use of these recovery systems in poor weather conditions as well as fine.

Key Issues

The key issues are:

• system inspected and serviced;

• crew well trained in the use of the system carried;

• 9 -

• system deployed correctly;

• first aid requirements anticipated; and

• system re-stowed and readied for immediate use.

• 10 -

Annex 2 – Guidance on Drills for different types of emergencies

Man Overboard Guidance

The following provides guidance on the actions of the vessel’s skipper and crew. These actions are generic and may vary from vessel to vessel and it is recommended that you identify the requirements particular to your vessel through practice drills.

General Requirements of All:

• Sound Crew Alarm.

• Crew to muster stations with warm clothing and lifejackets donned correctly.

• Was the crewman seen to have fallen overboard? Yes / No.

• If yes, throw lifebuoy with smoke / light float attached to mark position.

• If no, note time and position and consider using smoke / light float anyway to mark a datum position. This will give a visual marker to searching vessels and aircraft as an indication of the tide and surface water movement in the search area from a given time.

Skipper:

• Press MOB function on Navigational aid, if fitted and crew alarm.

• Inform ships mate with all details.

• Send DSC alert and commence voice transmission on VHF, MF or HF as appropriate.

• Inform Coastguard of any updated information and the description of missing person.

• Haul fishing gear if fishing.

• If not fishing and navigation allows commence Williamson turn.

• Pass any additional information to the Coastguard and any other vessels assisting.

• Keep all search units up to date by sending situation reports regularly.

Mate:

• Crew muster report to skipper, crewman missing? Yes / No.

• Collect details of missing crewman and pass to skipper as soon as possible, time and place last seen, clothing type and colours, age and state of health.

• Ensure crew are dressed appropriately.

• Post lookouts forward, wing of the bridge port and starboard also on a high point aft.

• Search vessel for missing crewman and prepare to launch rescue boat if

• 11 -

carried.

• Organise man overboard recovery system.

• Organise dry clothing and first aid equipment, prepare to treat for hypothermia.

• 12 -

Fire Fighting Guidance

The following provides guidance on the actions of the vessel’s skipper and crew. These actions are generic and may vary from vessel to vessel and it is recommended that you identify the requirements particular to your vessel through practice drills.

General Requirements of All:

• Sound Crew Alarm.

• Crew to muster stations with warm clothing and lifejackets donned correctly.

Skipper:

• Account for all crew, and send DSC alert and follow with voice transmission on VHF, MF or HF.

• Consider manual launching of liferafts.

• In receipt of all available information consider fighting the fire.

• Monitor vessel stability with engineer regarding water used for fire fighting.

• Consider emergency fuel stops, before activation, consider how power loss affects the crew’s fire fighting capability. Loss of main fire hoses and lighting.

• Prior to use of CO2 gas or other fixed fire fighting medium, order compartments to be cleared and checked, all crew evacuated from effected area.

• Ensure CO2 or other fixed fire fighting medium, firing point is evacuated in case gas leaks from bottles and valves.

• On receipt of the regular situation reports which indicate that the fire is now out of control, consider abandoning the vessel to the liferafts, (proceed to abandon ship).

Mate:

• Crew muster report to skipper.

• If liferafts deployed ensure access for boarding and safety of the raft from heat and smoke.

• Take charge of fire fighting parties.

• Close all ventilation flaps and doors to engine room, accommodation, galley and factory deck.

• Delegate crew member to don fireman’s outfit and breathing gear if carried.

• If crew numbers allow delegate first aid party and a safe, warm and dry compartment for any injured crew to rest and receive treatment.

• Give regular situation reports to skipper.

• 13 -

Engineer:

• Stop all fans in and around affected and adjacent compartments.

• Ascertain extent and severity of the fire and report to the skipper; assist the mate where possible with fire fighting parties.

• Check water ingress into compartments from fire hoses and fire fighting appliances.

• Prime emergency fire hoses and pumps as these may be required if emergency fuel stops activated.

• 14 -

Taking Water or Sinking Guidance

The following provides guidance on the actions of the vessel’s skipper and crew. These actions are generic and may vary from vessel to vessel and it is recommended that you identify the requirements particular to your vessel through practice drills.

General Requirements of All:

• Sound Crew Alarm.

• Crew to muster stations with warm clothing and lifejackets donned correctly.

Skipper:

• Send DSC alert on appropriate frequency.

• Follow with voice transmission on appropriate frequency.

• If call for assistance is answered request portable pumps.

• Consider if pumps are going to cope.

• Consider evacuation of non essential crew members.

• Consider ‘abandon ship’ (go to ‘abandon ship’).

Mate:

• Muster details to skipper.

• On the instructions from the skipper deploy the liferafts and rescue boat if carried.

• Secure liferafts in safe area on vessel and provide safe means of boarding.

• Keep skipper informed on progress.

Engineer:

• Check compartments for water ingress.

• Monitor bilge alarms.

• Maintain bilge pumps.

• Keep skipper informed of progress of pumps and water levels.

• Advise skipper on stability state of the vessel.

• Advise skipper on possible future of water levels and speed of water ingress.

• Advise skipper if additional pump would make a difference.

• Close all oil and fuel vents to try to prevent pollution.

• 15 -

Collision or Grounding Guidance

The following provides guidance on the actions of the vessel’s skipper and crew. These actions are generic and may vary from vessel to vessel and it is recommended that you identify the requirements particular to your vessel through practice drills.

General Requirements of All:

• Sound Crew Alarm.

• Crew to muster stations with warm clothing and lifejackets donned correctly.

Skipper:

• Send DSC Alert VHF, MF or HF.

• Transmit voice message on appropriate radio frequency.

• Check with mate on status of crew.

• Check with engineer on status of hull integrity.

• Check with other vessel if collision, on status of hull integrity and injuries to crew.

• Decide if evacuation to other ship is necessary or their crew to your vessel.

• If aground consider deploying stern anchors.

• Try to establish depth of water around the vessel and in which direction lies deeper water.

• Consider deployment of liferafts / rescue boat.

• Consider ‘abandon ship’ to liferafts (go to ‘abandon ship’).

Mate:

• Report muster details to skipper.

• Check for injured or trapped crewmen.

• Investigate watertight integrity of your vessel.

• Delegate work party to try to stop ingress of water.

• Keep skipper informed of progress.

Engineer:

• Start bilge pumps and monitor bilge alarms.

• Try to establish what damage has been sustained.

• Assist mate with designated work party to stop any ingress of water.

• Instigate frequent reports to skipper.

• 16 -

Abandon Ship Guidance

The following provides guidance on the actions of the vessel’s skipper and crew. These actions are generic and may vary from vessel to vessel and it is recommended that you identify the requirements particular to your vessel through practice drills.

General Requirements of All:

• Sound Crew Alarm.

• Crew to muster stations with warm clothing and lifejackets donned correctly.

Skipper:

• Issue order to ‘abandon ship’ by verbal order to mate and / or over ship’s intercom system.

• Crew go to liferaft embarkation points, not returning to cabins to get personal items.

• To ensure all crew are accounted for and have embarked liferafts.

• DSC alert sent and voice transmission to include the fact that crew are abandoning the vessel to liferafts, position, number of crew and any crew with injuries.

• Deploy EPIRB and SART if carried.

• Take charge of liferaft number 1, collect remaining VHF portable radio, log book and remaining equipment in grab bags, consider additional water for injured crewmen.

Mate:

• Account for all crew members and report to skipper.

• Supervise deployment of liferafts, pay due care to weather conditions for boarding.

• Supervise deployment of rescue boat if carried.

• Supervise embarkation of liferafts and account for all crew members boarding rafts as indicated on muster list.

• Tether liferafts (and the rescue boat if carried) together.

• Take command of liferaft number 2 taking additional grab bag, first aid kit, portable VHF radio and flares.

Abandoning the Vessel

If possible survival craft should be boarded dry, but if it is necessary to abandon a vessel by jumping directly into the water, the following procedure should be followed.

Hold lifejacket, block off nose and mouth, keep feet together, check below, avoid obstructions,

• 17 -

jump feet first, look ahead parallel to horizon, attempt to land near and slightly ahead of survival craft. Do not jump into boats or on top of liferafts, the contents could cause you serious injury.

In the Water

A survivor in the water should swim away from a sinking vessel as quickly as possible since when it founders wreckage and debris may surface with great force along with oil.

Get out of the water as soon as possible and remember swimming increases the rate of heat loss, try to avoid any likelihood of hypothermia.

• 18 -

Annex 3 – Muster Lists explained

Muster List

Muster lists are to inform the seaman where to go and what to do in the event of an alarm and one of the first tasks of a new crew member is to find out where he is expected to muster and his duties in the event of an alarm sounding.

The lists should contain the information required to muster the crew together at any given moment. Use of bunk or cabin numbers will enable the crew to establish their muster point and save you time on making new lists when crew leave or join the vessel. Don’t forget to allow for additional crew members or other persons carried. Keep muster points to a minimum and in areas easily monitored from the wheel house.

Lists should also clearly state initial duties and responsibilities in the case of an emergency. Crew mustered at the liferafts deck should prepare the liferafts for manual launching when the command has been given. The Chief Engineer will close up engine room and ensure all vents and fans throughout the vessel are shut down.

The pages at the end of this notice contain muster lists for vessels with small or large crew numbers.

Muster lists are in three stages which are:

Stage 1: Muster Station

At the sounding of any alarm the crew will dress in warm clothing suitable to abandon ship and a lifejacket (with light) donned appropriately and proceed to allocated muster stations and remain there until released. Crewmen turning up with T-shirts will not last long in a liferaft in the winter and even summer nights can be cold enough to induce hypothermia.

En-route to their muster stations all crew will close and secure doors, ventilators and undertake their emergency task before arrival at their muster station. Crew allocated a muster station at a liferaft or lifeboat stowage position will prepare the craft for manual deployment, not deploying until ordered to do so by the skipper or ship’s officer.

Consideration should be given both to the boarding of the liferafts when deployed and the positioning of boarding ladders may be beneficial.

The actions for Stage 2 will vary depending on the nature of the emergency. Therefore two options for Stage 2 are set out below.

Stage 2: Fire Fighting

Once all the crew have been accounted for and liferafts / lifeboats deployed and readied for use the crew guided by the mate will undertake fire fighting duties as detailed and practiced. At least two crew members should be allocated to breathing apparatus and the safety line signals should be known by all. The skipper will be busy enough transmitting safety or distress traffic on the vessel’s radio equipment. Ensure all vents are closed including those which may be within the winch or other deck housing; all vent fans should be stopped.

A further muster of the crew should be undertaken before firing of the CO2, or other fire smothering gas, to ensure that no persons are in the space about to be filled with the gas. Evacuation of the immediate area of the gas bottles should also be considered due to possible leaks from the valves. Gas bottles should be checked to confirm that the gas from all bottles has been deployed, a “tell tale” sign would be “frosting” of the bottom of each bottle.

• 19 -

Stage 2: Man Overboard

Lookouts should be posted and communications agreed; the first aid kit, recovery system and blankets should be available and deployed. A course should be plotted, reversed and Pan Pan or Mayday DSC alert and voice broadcast sent. The crewman who is suspected to have fallen overboard should be identified and the crew questioned as to last sighting of this person.

The vessel should be searched to ensure that the crewman is not still on board.

Stage 3: Abandon Ship

The order is to be given by the skipper or master and relayed to the crew via the mate or ship’s officers. Where possible announcements by tannoy should be avoided as it is likely to induce panic. Information should be given on the muster list as to which liferaft or lifeboat that each crew member is assigned.

At the order given by the skipper the crew will prepare to abandon ship, no crew member will be given access to the inside of the vessel but will proceed to their designated liferaft or lifeboat. All lifeboats or liferafts should be launched if possible and manned, not forgetting the EPIRB, additional first aid kit, additional distress flares, portable VHF radios and the SARTs if carried. Some crew members will have been allocated special tasks, deploy EPIRB or take additional flares, first aid kit or SART.

The liferafts should be secured together where possible and remain in the vicinity of the vessel or the last known position which was broadcast with the distress message or DSC information. Consideration should be given to a “grab bag” prepared with “goodies” such as bottled water and tinned high energy chocolate or sugar sweets. Remember survival course training regarding safety and hypothermia.

• 20 -

The following pages are examples of muster lists. You should prepare your own muster lists based on your own vessel layout, equipment and number of crew.

Muster List (Larger Vessels) Name of Fishing Vessel: Crew Member Stage 1 Muster at muster station with warm clothing and lifejacket securely fastened Stage 2 Form work parties to save the ship and attend to casualties Stage 3 Abandon ship on verbal order of the skipper

Muster Point Emergency Parties Abandon Ship Skipper Muster point 1. Wheel House Don lifejacket. Inform Coastguard by radio. Maintain communications with Coastguards. Issue Verbal command Broadcast Mayday. Deploy EPIRB and portable VHF. Ensure all crew has embarked to liferafts. Take charge of liferaft No 1. Mate Muster point 1. Wheel House Don lifejacket Account for all crew.

Take charge of emergency parties. Broadcast to crew, abandon ship to liferafts. Deploy first aid kit, flares, SART and take charge of liferaft No 2. Bunk 3 Engineer Muster point 2. Boat Deck Don lifejacket. Ready liferafts for deployment. Under mates direction undertake first aid duties. Deploy liferaft to lee side of vessel, secure painter. Abandon ship to liferaft 1. Bunk 4 Muster point 2. Boat Deck Don lifejacket. Ready liferafts for deployment. Under mates direction.

Deploy liferaft to lee side of vessel, secure painter. Abandon ship to liferaft 2. Bunk 5 Muster point 3. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferaft 1. Bunk 6 Muster point 3. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferaft 2. Bunk 7 Muster point 3. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferaft 1. Bunk 8 Muster point 3. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferaft 2. Bunk 9 Muster point 3. Fish Room Hatch Don lifejacket await further orders Under mates direction. Abandon ship to liferaft 1. Bunk 10 Muster point 3. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferaft 2. Bunk 11 Muster point 3. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferaft 1. Bunk 12 Muster point 3. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferaft 2. Additional Crew Muster point 2. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferafts as directed by the mate.

21 -

22 - Muster List (Smaller Vessels) Name of Fishing Vessel: Crew Member Stage 1 Muster at muster station with warm clothing and lifejacket securely fastened Stage 2 Form work parties to save the ship and attend to casualties Stage 3 Abandon ship on verbal order of the skipper Muster Point Emergency Parties Abandon Ship Skipper Muster point 1. Wheel House Don lifejacket. Inform Coastguard by radio. Maintain communications with Coastguards. Issue Verbal command Broadcast Mayday. Deploy EPIRB and portable VHF. Ensure all crew has embarked to liferafts. Take charge of liferaft No 1. Mate Muster point 1. Wheel House Don lifejacket account for all crew. Take charge of emergency parties. Broadcast to crew, abandon ship to liferafts. Deploy first aid kit, flares, SART and take charge of liferaft No 2. Bunk 3 Engineer Muster point 2. Boat Deck Don lifejacket. Ready liferafts for deployment. Under mates direction undertake first aid duties. Deploy liferaft to lee side of vessel, secure painter. Abandon ship to liferaft 1. Bunk 4 Muster point 2. Boat Deck Don lifejacket. Ready liferafts for deployment. Under mates direction. Deploy liferaft to lee side of vessel, secure painter. Abandon ship to liferaft 2. Bunk 5 Muster point 3. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferaft 1. Bunk 6 Muster point 3. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferaft 2. Additional Crew Muster point 2. Fish Room Hatch Don lifejacket await further orders. Under mates direction. Abandon ship to liferafts as directed by the mate.

Annex C Annual self-certification declaration form

                                                      

Annex D Guard Vessel Inspection and Specification Document

APPENDIX B GUARD VESSEL INSPECTION AND SPECIFICATION DOCUMENT Vessel Name: Reg No: Client/Project: Audit Time & Date: Inspector: Port of Inspection: No. Yes No Remarks 1 Vessel Spec. (see attached) completed 2 FFSC Guard Vessel Operational Procedures Manual on board 3 Guard Vessel Contract and Project Papers on board 4 Company Policy on Substance Abuse - (Flyer on display) 5 Vessel Certification Issue Date Expiry Date Remarks Load Line Exemption Certificate of Registry Fishing Vessel Certificate Radio Licence Certificate Stability Book Guardship Stability Risk Assessment Folder Last reviewed: 6 Crew Certification Name Rank D.O.B. Marine Cert. Marine Cert. First Aid Fire Fighting S.Awareness Sea Survival Medically Fit Name Rank D.O.B. Marine Cert. Marine Cert. First Aid Fire Fighting S.Awareness Sea Survival Medically Fit Name Rank D.O.B. Marine Cert. Marine Cert. First Aid Fire Fighting S.Awareness Sea Survival Medically Fit Revision: 2.0 Page 1 of 6 Jan 2013

APPENDIX B GUARD VESSEL INSPECTION AND SPECIFICATION DOCUMENT 7 Safety Items Quantity Expiry/Service Date Remarks Liferafts (Make)* HRUs for Liferafts (Makes)* Life Buoys and Smoke Marker EPIRB HRU for EPIRB (Make)* Pyrotechnics (plus Solas Chart No.1) Line Throwers Immersion / Survival Suits* Lifejackets Fire Extinguishers Fire Blanket in Galley Fire Pump and Hoses Bilge Pump Bilge Alarm(s) Tested: Fixed Fire Fighting System Type: Compass (Date last corrected - Devation Card) Medical Kit (Pharmacy Receipt Date + Class) * Category: Navigation Lights and Sound Signals Checked/tested: Waterproof Torch VHF radio (fixed and hand held) Safety Drills (date last undertaken)

• - Servicing Receipt Required

8 Miscellaneous Yes No Remarks Publications/Charts (appropriate electronic and paper charts) Almanac KIS UKCS (current update) FIshSAFE (current update) Binoculars Camera Search Light (hand held - properly charged) MARPOL Regulations (knowledge of) 9 General Condition of Vessel Remarks/Observations/Recommendations Hull Decks Engine Room/Machinery Space/Steering Compartment Galley Accommodation 10 Questions for Fishing Vessels undertaking Guard Duty When did the vessel last have a MCA inspection and what were the findings? Have there been any recent MCA spot checks – what were the findings? What action was taken on the non conformities? Answer: When was the last critical maintenance performed on critical equipment such as main engines/generators? Answer: When was the vessel last slipped/major maintenance carried out? Answer: Lights Expiry Date. Revision: 2.0 Page 2 of 6 Jan 2013

APPENDIX B GUARD VESSEL INSPECTION AND SPECIFICATION DOCUMENT 10 Questions for Fishing Vessels undertaking Guard Duty (continued) Has the vessel a closed shelter deck? If so what arrangements for removal of entrapped water are there? Answer: What is required to maintain watertight integrity of the vessel? Are all escape hatches, watertight doors, deadlights, scuppers etc in good working condition and properly tested? Answer: What are the competencies onboard? Please insert appropriate number(s) in boxes below. Is there a dedicated engineer? Answer: Class 1 Fishing Class 2 Fishing Eng Officer 5 Day Bridge Watch Keeping Eng Watch Navigation/other Watch Keeping Dedicated Engineer - answer yes or no. What is the safety regime onboard? Does the crew wear Personal Floatation Devices (PFD) and appropriate PPE at all times on deck? Answer: What is the Vessel’s weather operating criteria for remaining on guard station? Answer: Has Skipper noted the 'Guardship Stability - Notice to Skipper' information contained within the Vessel's Guardship Stability Booklet and can he confirm that vessel will be loaded as per the noted conditions? If there is any alteration to the noted conditions, the Skipper must confirm that he is aware of the effect that any devation will have to the Vessel's stability. Answer: What is the watch keeping regime offshore? In particular – who mans the bridge, what hours are worked and what competencies are taking a bridge watch? Answer: What accidents/incidents/breakdowns have occurred in last 12 months and if any and if appropriate, what corrective / future preventative action was taken? Answer: Regular Skipper/ Mate or Engineer on board? Please answer yes/no in appropriate boxes below. Answer: Regular Skipper Regular Mate Regular Engineer Are any of the crew going on the guard trip on prescribed medication? If so, (a) Is the Crew member / Skipper aware of the effect that this medication may have on Watch Keeping / other Guard Vessel related duties and, (b) Has the individual an adequate supply of this medication for the scheduled duration of the guard trip? Answer(s): Revision: 2.0 Page 3 of 6 Jan 2013

Vessel Specification: Vessel Details: Name Fishing Vessel Number Date Built Date of any Major Modifications Hull Length Breadth Depth Load Line Exemption Certificate Valid until UKFV Certificate Valid until Tonnage (Gross & Net) Main Engine Type and hp/kw Auxiliary Engine Type and hp/kw Cruising Speed Fuel Oil Capacity Freshwater Capacity Lube Oil Capacity Hydraulic Oil Capacity Maximum Number of days on Location Call Sign Mobile No. Telex No./Email Satellite Phone No. Skipper Owner Accommodation (no. of persons & berthing arrangements) Electronics Navigation (Type) Communications (Type) Radar (1) - Main Radio (1) - Radar (2) - Main Radio (2) - Plotter (1) - VHF (1) - Plotter (2) - VHF (2) - DGPS - Mobile Tel - Echo Sounder - Sat Tel - FishSAFE - Telex/Email - AIS - Navtex - Additional Comments: GUARD VESSEL INSPECTION AND SPECIFICATION DOCUMENT APPENDIX B Revision: 2.0 Page 4 of 6 Jan 2013

Photographs: APPENDIX B GUARD VESSEL INSPECTION AND SPECIFICATION DOCUMENT Revision: 2.0 Page 5 of 6 Jan 2013

Vessel Name: Reg No: Client Project: Audit Time and Date: Inspector: Port of Inspection: No. Close Out Category Date Close Out Category Definitions P = Prior to sailing 1 = Within 1 month window 3 = Within 3 month window I declare that the answers provided in relation to the Questions contained in Section 10 of the Guard Vessel Inspection and Specification document are full and accurate and hereby accept the above findings from today's Guard Vessel Inspection. Project Specific Brief provided by SFF Services Representative - Yes/No (delete as applicable). Skipper's Signature: Skipper's Name: Inspector's Name: Inspector's Name: Checked by - Signature: Checked by - Name: Findings Close Out Action APPENDIX B GUARD VESSEL INSPECTION AND SPECIFICATION DOCUMENT Revision: 2.0 Page 6 of 6 Jan 2013

Annex E Aquarius documented man overboard recovery procedure

Annex F Aquarius log book - Muster and training records

Annex G Risk Assessment - General Working on Deck

Annex H Aquarius LOLER record of inspections

Annex I MAIB Safety Flyer

SAFETY FLYER TO THE FISHING INDUSTRY Aquarius: Fatal man overboard on 17 August 2015 Narrative In the early hours of the morning on 17 August 2015, a fisherman on board the 20.8m twin rig trawler Aquarius (Figure 1) was thrown overboard violently when a rope stopper parted. The vessel had just cleared the port of Aberdeen, Scotland, and its crew were attempting to re-mark the trawl warps when the accident happened. The skipper reacted quickly to stop the vessel, and the crew threw lifebuoys to the casualty. However, the recovery attempt was unsuccessful and, about 10 minutes later, the casualty sank out of view. The skipper then raised the alarm but, despite a search involving numerous vessels and a helicopter, the casualty’s body was not recovered. In order to mark the steel wire warp, the crew had streamed it over the stern. The stopper was used to take the strain of the trailing warp so that the crew could lower its inboard section on to the deck (Figure 2). The MAIB investigation established that: • The stopper parted under tension because a man-made fibre rope had been used instead of a chain, and a proper stopper hitch had not been applied. • The casualty was thrown overboard because he had positioned himself within the bight of the slackened trawl warp. Figure 1 Figure 2

• The method used to mark the trawl warp was unnecessarily hazardous and the outcome was entirely foreseeable. • The crew were unable to recover the casualty back on board because neither they, nor their vessel had been adequately prepared to deal with such emergency situations. ◦◦ The vessel did not carry a dedicated means of recovering a person from the water. ◦◦ The crew did not conduct periodic emergency drills. • The casualty sank and his body was not recovered because he was not wearing a lifejacket or other type of personal flotation device while working on the vessel’s open deck. • Aquarius had been poorly operated; its safety equipment was not properly maintained, hazardous activities had not been risk assessed and its crew worked prolonged hours and did not receive the minimum amount of rest or leave required by UK legislation. Safety Lessons The importance of developing and promoting a strong safety culture within an inherently hazardous industry and ensuring that a vessel and its crew are properly prepared to deal with emergency situations is repeatedly highlighted by MAIB in fishing vessel accident reports. Neither of these safety critical factors was evident on board Aquarius; the key lessons to take from this accident are: • Fishing vessel owners and skippers must make every effort to promote a strong safety culture on board their vessels; hazardous activities should be properly planned and safe systems of work developed. This is particularly important when carrying out non-standard or one-off tasks. • Crew should always wear lifejackets or personal flotation devices when working on deck, especially when working on fishing gear that is to be towed, shot or recovered. • The dangers of standing in the bight of a rope, or over a slacked warp should be well understood by all seafarers and, in accordance with basic seamanship good practice, must be avoided at all times. • Whenever a person falls or is thrown overboard, the alarm should be raised immediately, preferably by use of the vessel’s VHF radio digital selective calling function. This will ensure the speediest response from the emergency services and other vessels operating in the area. • The importance of conducting regular manoverboard recovery cannot be emphasised enough; emergency drills will help ensure that the vessel’s safety equipment remains fully operational, is suitable for its intended use and that the crew are familiar with its use. • Tiredness is a major hazard to all fishermen, and in order to reduce the risk of crew fatigue skippers must ensure that they monitor and record hours of rest. This flyer and the MAIB’s investigation report are posted on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch Email: maib@dft.gsi.gov.uk First Floor, Spring Place 105 Commercial Road Tel: 023 8039 5500 Southampton, SO15 1GH

FD06039 ｜ 2011_英国_MAIB_gas-explosions-on-general-cargo-ship-nortrader-with-1-person-injured__Nortrader_CombinedAn.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2011 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A IMSBC Code - schedule entry for aluminium smelting/remelting by-products

PRECAUTIONS Prior to loading this cargo, a certificate shall be provided by the manufacturer or shipper stating that, after manufacture, the material was stored under cover, but exposed to the weather in the particle size to be shipped, for not less than 3 days prior to shipment. Whilst the ship is alongside and the hatches of the cargo spaces containing this cargo are closed, the mechanical ventilation shall be operated continuously as weather permits. During handling of this cargo, “NO SMOKING” signs shall be posted on decks and in areas adjacent to cargo spaces and no naked lights shall be permitted in these areas. At least two self-contained breathing apparatus, in addition to those required by SOLAS regulation II-2/10.10, shall be provided on board. Bulkheads between the cargo spaces and the engine- room shall be gastight. Inadvertent pumping through machinery spaces shall be avoided. VENTILATION Continuous mechanical ventilation shall be conducted during the voyage for the cargo spaces carrying this cargo. If maintaining ventilation endangers the ship or the cargo, it may be interrupted unless there is a risk of explosion or other danger due to interruption of the ventilation. In any case mechanical ventilation shall be maintained for a reasonable period prior to discharge. Ventilation shall be arranged such that any escaping gases are minimized from reaching living quarters on or under the deck. CARRIAGE For quantitative measurements of hydrogen, ammonia and acetylene, suitable detectors for each gas or combination of gases shall be on board while this cargo is carried. The detectors shall be of certified safe type for use in explosive atmosphere. The concentrations of these gases in the cargo spaces carrying this cargo shall be measured regularly, during voyage, and the results of the measurements shall be recorded and kept on board. DISCHARGE No special requirements. CLEAN-UP Water shall not be used for cleaning of the cargo space which has contained this cargo, because of danger of gas.

EMERGENCY PROCEDURES SPECIAL EMERGENCY EQUIPMENT TO BE CARRIED Nil EMERGENCY PROCEDURES Nil EMERGENCY ACTION IN THE EVENT OF FIRE Batten down and use CO2 if available. Do not use water. If this proves ineffective, endeavour to stop fire from spreading and head for the nearest suitable port. MEDICAL FIRST AID Refer to the Medical First Aid Guide (MFAG), as amended.

Annex B IMSBC Code - 16 point cargo information list and cargo declaration form

Section 4 Assessment of acceptability of consignments for safe shipment 4.1 Identification and classification 4.1.1 Each solid bulk cargo in this Code has been assigned a Bulk Cargo Shipping Name (BCSN). When a solid bulk cargo is carried by sea it shall be identified in the transport documentation by the BCSN. The BCSN shall be supplemented with the United Nations (UN) number when the cargo is dangerous goods. 4.1.2 If waste cargoes are being transported for disposal, or for processing for disposal, the name of the cargoes shall be preceded by the word “WASTE”. 4.1.3 Correct identification of a solid bulk cargo facilitates identification of the conditions necessary to safely carry the cargo and the emergency procedures, if applicable. 4.1.4 Solid bulk cargoes shall be classified, where appropriate, in accordance with the UN Manual of Tests and Criteria, part III. The various properties of a solid bulk cargo required by this Code shall be determined, as appropriate to that cargo, in accordance with the test procedures approved by a competent authority in the country of origin, when such test procedures exist. In the absence of such test procedures, those properties of a solid bulk cargo shall be determined, as appropriate to that cargo, in accordance with the test procedures prescribed in appendix 2 to this Code. 4.2 Provision of information 4.2.1 The shipper shall provide the master or his representative with appropriate information on the cargo sufficiently in advance of loading to enable the precautions which may be necessary for proper stowage and safe carriage of the cargo to be put into effect. 4.2.2 Cargo information shall be confirmed in writing and by appropriate shipping documents prior to loading. The cargo information shall include: .1 the BCSN when the cargo is listed in this Code. Secondary names may be used in addition to the BCSN; .2 the cargo group (A and B, A, B or C); .3 the IMO Class of the cargo, if applicable; .4 the UN number preceded by letters UN for the cargo, if applicable; .5 the total quantity of the cargo offered; .6 the stowage factor; .7 the need for trimming and the trimming procedures, as necessary;

.8 the likelihood of shifting, including angle of repose, if applicable; .9 additional information in the form of a certificate on the moisture content of the cargo and its transportable moisture limit in the case of a concentrate or other cargo which may liquefy; .10 likelihood of formation of a wet base (see subsection 7.2.3 of this Code); .11 toxic or flammable gases which may be generated by cargo, if applicable; .12 flammability, toxicity, corrosiveness and propensity to oxygen depletion of the cargo, if applicable; .13 self-heating properties of the cargo, and the need for trimming, if applicable; .14 properties on emission of flammable gases in contact with water, if applicable; .15 radioactive properties, if applicable; and .16 any other information required by national authorities. 4.2.3 Information provided by the shipper shall be accompanied by a declaration. An example of a cargo declaration form is set out in the next page. Another form may be used for cargo declaration. As an aid to paper documentation, Electronic Data Processing (EDP) or Electronic Data Interchange (EDI) techniques may be used.

FORM FOR CARGO INFORMATION for Solid Bulk Cargoes BCSN Shipper Transport document number Consignee Carrier Name/means of transport Port/place of departure Port/place of destination Instructions or other matters General description of the cargo (Type of material/particle size) Gross mass (kg/tonnes) Specifications of bulk cargo, if applicable: Stowage factor: Angle of repose, if applicable: Trimming procedures: Chemical properties if potential hazard*:

• e.g., Class & UN No. or “MHB”

Transportable Moisture Limit Group of the cargo Group A and B* Group A* Group B Group C * For cargoes which may liquefy (Group A and Group A and B cargoes) Moisture content at shipment Relevant special properties of the cargo (e.g., highly soluble in water) Additional certificate(s)* Certificate of moisture content and transportable moisture limit Weathering certificate Exemption certificate Other (specify)

• If required

DECLARATION I hereby declare that the consignment is fully and accurately described and that the given test results and other specifications are correct to the best of my knowledge and belief and can be considered as representative for the cargo to be loaded. Name/status, company/organization of signatory Place and date Signature on behalf of shipper

Annex C Carrying solid bulk cargoes safely - booklet

Carrying solid bulk cargoes safely Guidance for crews on the International Maritime Solid Bulk Cargoes Code In conjunction with

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 2 Contents Introduction

1. General requirements for carrying solid bulk cargoes

Accepting cargoes for shipment Loading

1. The Code’s three cargo groups

Group A cargoes – cargoes which may liquefy Group B cargoes – cargoes with chemical hazards Group C cargoes – cargoes which are neither liable to

liquefy nor possess chemical hazards

1. Other hazards associated with carrying solid bulk cargoes
2. Checklist and flowchart for accepting and loading solid

bulk cargoes Appendix 1 IMO regulations and guidance relating to the transport of solid bulk cargoes Appendix 2 The Code section by section 3 4 4 6 7 8 10 14 15 16 19 21

Introduction Carrying solid bulk cargoes involves serious risks, which must be managed carefully to safeguard the crew and the ship. These risks include reduced ship stability, and even capsizing, due to cargo liquefaction; fire or explosion due to chemical hazards; and damage to ship structures due to poor loading procedures. The main legislation governing safe carriage of solid bulk cargoes is the International Maritime Solid Bulk Cargoes (IMSBC) Code, which became mandatory on January 1, 2011, under the SOLAS Convention . This pocket guide will help you understand the IMSBC Code’s key requirements and give you greater confidence in managing the risks of carrying solid bulk cargoes and achieving compliance with SOLAS. It outlines the precautions you should take before accepting cargoes for shipment and the procedures you should follow for safe loading and carriage, and details the primary hazards associated with the different types of solid bulk cargo. It also contains a quick reference checklist and flowchart summarising the steps you need to follow. Note: This guide is only an introduction to the IMSBC Code; you should always consult the full Code to check the requirements for each cargo you are carrying. It does not cover carriage of grain in bulk; this is covered by the International Grain Code. Look for “Consult the Code” headings within this publication, which refer you to specific sections of the Code for more information. Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 3 Note: This guide is only an introduction to the IMSBC Code; you should always consult the full Code to check the requirements for each cargo you are carrying. It does not cover carriage of grain in bulk; this is covered by the International Grain Code. Look for “Consult the Code” headings within this publication, which refer you to specific sections of the Code for more information. Introduction Carrying solid bulk cargoes involves serious risks, which must be managed carefully to safeguard the crew and the ship. These risks include reduced ship stability (and even capsizing) due to cargo liquefaction; fire or explosion due to chemical hazards; and damage to ship structures due to poor loading procedures. The main legislation governing safe carriage of solid bulk cargoes is the International Maritime Solid Bulk Cargoes (IMSBC) Code, which became mandatory on January 1, 2011, under the SOLAS Convention1. This pocket guide will help you understand the IMSBC Code’s key requirements and give you greater confidence in managing the risks associated with the carriage of solid bulk cargoes. It outlines the precautions you should take before accepting cargoes for shipment and the procedures you should follow for safe loading and carriage, and details the primary hazards associated with the different types of solid bulk cargo. It also contains a quick reference checklist and flowchart summarising the steps you need to follow. 1For a list of other IMO legislation and guidance relating to bulk cargoes, see Appendix 1.

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 4

1. General requirements for

carrying solid bulk cargoes No matter what solid bulk cargo you are carrying, the same general requirements apply for accepting them for shipment and loading them. Section 2 of this pocket guide covers the typical requirements for different types of cargo. Accepting cargoes for shipment Information required from the shipper Before you can accept a cargo for shipment, the shipper must provide the Master with valid, up-to-date information about the cargo’s physical and chemical properties. The exact information and documentation they must provide is listed in the Code under ‘Assessment of acceptability of consignments for safe shipment; Provision of Information’, and includes the correct Bulk Cargo Shipping Name2 and a declaration that the cargo information is correct.

Checking the cargo schedule Individual cargoes are listed in ‘schedules’ which are contained in Appendix 1 of the Code. These describe each cargo’s properties and detail the requirements for handling, stowing and carrying it safely. You must always consult the relevant schedule in the Code to find out what hazards the cargo presents. 2This is the cargo’s official name used in the Code. Cargo being worked Consult the Code – see Section 4 Consult the Code – see Appendix 1

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 5 Accepting cargoes not listed in the IMSBC Code The list of individual cargoes contained in the Code is not exhaustive. If a cargo not listed in the Code is presented for shipment, the shipper and the appropriate competent authorities3 must follow this process:

1. Before loading, the shipper must provide details of the characteristics and

properties of the cargo to the competent authority of the port of loading.

1. Based on this information the competent authority of the port of loading

will assess the acceptability of the cargo for shipment.

– If the assessment defines the cargo as Group A or B4, the competent

authorities3 will set the preliminary suitable conditions for carriage.

– If the cargo is Group C4 then carriage can be authorised by the port of

loading and the competent authorities of the unloading port and flag

state will be informed of the authorisation.

1. In both cases, the competent authority of the port of loading will give the

Master a certificate stating the characteristics of the cargo and the required

conditions for carriage and handling. The competent authority of the port of

loading will also provide the same information to the IMO.

Exemptions Under section 1.5 of the Code, a competent authority (or authorities) can grant an exemption which allows ships to carry a cargo outside the requirements specified in its schedule, provided that equivalent provisions have been put in place. Agreement of all three competent authorities is required to ship a cargo under an exemption. Acceptance of an exemption by authorities not party to it is discretionary: i.e., if the loading port authority issues an exemption, the unloading port and flag state authorities can choose to accept it or reject it. An exemption can be valid for up to five years and does not necessarily lead to the creation of a new or revised schedule. 3The competent authorities of the port of loading, the port of unloading and the flag state. 4The Code categorises cargo into three Groups – A, B and C. See page 7 for their definitions. Consult the Code – see Section 1.3 Consult the Code – see Section 1.5

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 6 Loading Inspecting and preparing cargo spaces In general, before loading a cargo you must inspect and prepare the cargo spaces, checking that: – bilge wells and strainer plates are prepared to facilitate drainage and prevent cargo entering the bilge system – bilge lines, sounding pipes and other service lines are in good order – cargo space fittings are protected from damage, and – measures are in place to minimise dust entering living quarters or other

interior spaces, or coming into contact with moving parts of deck machinery and external navigational aids. Distribution and stability You must also make sure that cargoes are properly distributed throughout the ship’s holds to provide adequate stability and ensure that the ship’s structure is never overstressed. Information can be found in the ship’s stability information booklet or you can use loading calculators if they are available. The Master will need to calculate the stability for the anticipated worst conditions during the voyage as well as for departure and demonstrate that the stability is adequate.

Loading Plan Before loading or unloading, the Master and the terminal representative must agree a Loading Plan to ensure that the permissible forces and moments on the ship are not exceeded. What this Plan should include is detailed in the Code of Practice for the Safe Loading and Unloading of Bulk Carriers (the BLU Code) and in the checklist on page 16.

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 7

1. The Code’s three cargo groups

The IMSBC Code categorises cargoes into three groups: Group A – cargoes which may liquefy if shipped at a moisture content exceeding their Transportable Moisture Limit (TML)5. Group B – cargoes which possess a chemical hazard which could give rise to a dangerous situation on a ship. Group C – cargoes which are neither liable to liquefy (Group A) nor possess chemical hazards (Group B). Cargoes in this group can still be hazardous. You can find the Group for a particular cargo in its schedule. *GOSCjCBLGAICJMPC'K?ECAMSPRCQWMD+2"

5The TML is the maximum moisture content considered safe for carriage. See page 9.

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 8 Group A cargoes (cargoes which may liquefy) What is liquefaction and how does it affect cargo? Liquefaction means that a cargo becomes fluid (liquefies). On ships, this happens when the cargo is compacted by the ship’s motion. Cargoes which are prone to liquefaction contain a certain quantity of moisture and small particles, although they may look relatively dry and granular when loaded. Liquefaction can lead to cargo shift6 and even to the capsize and total loss of the ship, and can occur even when cargoes are cohesive and trimmed level. Examples of Group A cargoes Mineral concentrates Mineral concentrates are refined ores in which valuable components have been enriched by eliminating most waste materials. They include copper concentrate, iron concentrate, lead concentrate, nickel concentrate and zinc concentrate. Nickel ore7 There are several types of nickel ore which vary in colour, particle size and moisture content. Some may contain clay-like ores. Coal Coal (bituminous and anthracite) is a natural, solid, combustible material consisting of amorphous carbon and hydrocarbons. It is best known as a Group B cargo due to its flammable and self- heating properties, but it can also be classed as Group A because it can liquefy if predominantly fine (i.e., if 75% is made up of particles less than 5mm in size). In these cases, it is classed as both Group A and B. 6Cargo shift can be divided into two types: sliding failure or liquefaction consequence. Trimming the cargo can prevent sliding failure. 7Nickel ore is not to be confused with nickel concentrate. Coal being loaded. Coal is most AMKKMLJW?%PMSN A?PEM @SR can also be classed as both A and B. Consult the Code – see Section 7

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 9

Hui Long ?@SJIA?PPGCPUFGAF Q?LIGL2FCA?SQCU?Q @CJGCTCBRM@CJGOSCD?ARGML

NMQQG@JWBSCRMRFC2+* being exceeded. Typical requirements for accepting and loading Group A cargoes Information required from the shipper To control the risks of liquefaction, Group A cargoes are tested to determine their Transportable Moisture Limit (TML) and their actual moisture content before they can be shipped. The TML is the maximum moisture content considered safe for carriage. The actual moisture content of the cargo must be below the TML. The information supplied by the shipper to the master must include a signed certificate stating the TML, and a signed certificate or declaration of the cargo’s actual moisture content. What you need to do Follow these steps when carrying Group A cargoes to reduce the risk of liquefaction: – Make sure the shipper has supplied the required information, including the TML and the actual moisture content. – Only accept the cargo if the actual moisture content is less than its TML. – Carry out visual monitoring during loading. If there are any indications of high moisture content, stop loading and seek further advice. – Consider trimming the cargo to reduce the likelihood of cargo shift. – Take measures to prevent water or other liquids entering the cargo space during loading and throughout the voyage. Practical guidance on managing the risks of liquefaction is also available from the UK P&I Club at www.ukpandi.com/loss-prevention Consult the Code – see Sections 7 and 8

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 10 Group B cargoes (cargoes with chemical hazards) Group B cargoes are classified in two ways within the Code: ‘Dangerous goods in solid form in bulk’ (under the International Maritime Dangerous Goods (IMDG) Code; and ‘Materials hazardous only in bulk’ (MHB). You will find this information in the “characteristics” section of the cargo’s schedule. Cargoes classified as dangerous goods in solid form in bulk will also have a ‘UN’ number in the Bulk Cargoes Shipping Name.

Dangerous goods in solid form in bulk In the Code these cargoes are classed as follows: Class 4.1: Flammable solids Class 4.2: Substances liable to spontaneous combustion Class 4.3: Substances which, in contact with water, emit flammable gases Class 5.1: Oxidizing substances Class 6.1: Toxic substances Class 7: Radioactive materials Class 8: Corrosive substances Class 9: Miscellaneous dangerous substances and articles. Materials hazardous only in bulk (MHB) MHB cargoes are materials which possess chemical hazards when transported in bulk that do not meet the criteria for inclusion in the IMDG classes above. They present significant risks when carried in bulk and require special precautions. They are described as follows: Combustible solids: materials which are readily combustible or easily ignitable Self-heating solids: materials that self-heat Solids that evolve into flammable gas when wet: materials that emit flammable gases when in contact with water Solids that evolve toxic gas when wet: materials that emit toxic gases when in contact with water Toxic solids: materials that are acutely toxic to humans if inhaled or brought into contact with skin Corrosive solids: materials that are corrosive to skin, eyes, metals or respiratory sensitisers. Consult the Code – see Section 9

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 11 Examples of Group B cargoes and the risks they present The major risks associated with Group B cargoes are fire and explosion, release of toxic gas and corrosion.

Coal Coal may create flammable atmospheres, heat spontaneously, deplete oxygen concentration and corrode metal structures. Some types of coal can produce carbon monoxide or methane. Direct reduced iron (DRI) DRI may react with water and air to produce hydrogen and heat. The heat produced may cause ignition. Oxygen in enclosed spaces may also be depleted. Metal sulphide concentrates Some sulphide concentrates are prone to oxidation and may have a tendency to self-heat, leading to oxygen depletion and emission of toxic fumes. Some metal sulphide concentrates may present corrosion problems. Organic materials Ammonium nitrate-based fertilisers Ammonium nitrate-based fertilisers support combustion. If heated, contaminated or closely confined, they can explode or decompose to release toxic fumes and gases. Wood products transported in bulk8 Wood products transported in bulk are listed in a new schedule to the Code: Wood Products – General. They include logs, pulpwood, roundwood, saw logs and timber. These cargoes may cause oxygen depletion and increase carbon dioxide in the cargo space and adjacent spaces.

2FCB?K?ECA?SQCB@W?"0'CVNJMQGML 'K?ECAMSPRCQWMD SPEMWLCQ 8These are wood products loaded and discharged by methods such as elevators and grabs. They are distinct from wood products listed in other schedules.

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 12 Typical requirements for accepting and loading Group B cargoes Information required on board ship for dangerous goods in solid form in bulk To carry dangerous goods in solid form in bulk, your ship must have a Document of Compliance for the Carriage of Dangerous Goods, supplied by the ship’s flag or classification society. The Master must have a special list, manifest or stowage plan identifying the cargo’s location, and there must be instructions on board for emergency response. Segregation Because of their potential hazards, many Group B cargoes are incompatible and must be segregated. When segregating cargoes, you should take into account any secondary risks they present.

!M?JMLjPCGL?A?PEMFMJB Consult the Code – see Section 9.3

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 13 Specific risk mitigation measures The cargo’s schedule and the information provided by the shipper will detail the precautions you must take when carrying Group B cargoes. The following are some of the common risk mitigation measures you will employ. Fire and explosion Depending on the cargo, precautions against fire and explosion may include ventilating or inerting the cargo holds and the enclosed spaces adjacent to the holds: in some cases the ventilators will need to be explosion proof. The atmosphere in the cargo holds and the enclosed spaces adjacent to the holds may also need to be monitored with an appropriate gas detector. Toxic gas Toxic gas risks will be mitigated using natural or forced mechanical ventilation. The choice of ventilation will depend on the type of cargo and the properties of the gas (i.e., whether it creates an explosive atmosphere). You may need to monitor the cargo hold atmosphere. Corrosion Corrosion can be caused by some Group B cargoes and their residues. A coating or barrier may need to be applied to the cargo space structures before loading. Before loading and unloading corrosive cargoes, make sure the cargo space is clean and dry. "0'UFGAFF?QQCJDFC?RCB'K?ECAMSPRCQWMD SPEMWLCQ

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 14 Group C cargoes (cargoes which are neither liable to liquefy nor possess chemical hazards) Although Group C cargoes do not present the dangers associated with Group A and B cargoes, they can still carry risks. Examples of Group C cargoes, their risks and mitigation measures Iron ore and high density cargoes These cargoes can be extremely dense and can overstress the tanktop. Make sure that their weight is evenly distributed during loading and during the voyage so that the tanktop is not overstressed, and also consider trimming the cargo. Loading rates of iron ore are normally very high and you should also consider the ship’s ballasting operations and loading sequences. Sand and fine particle materials Fine particle materials can be abrasive. Silica dust is easily inhaled and can result in respiratory disease. You should take appropriate precautions to protect machinery and accommodation spaces from the dust of sand and fine particle cargoes, and to prevent the cargo from getting into the bilge wells. People who may be exposed to cargo dust should wear goggles or other equivalent dust eye-protection, dust filter masks and protective clothing. Cement Cement may shift when aerated during loading. Dust can also be produced from this cargo. Follow the precautions for sand and fine particle materials (above).

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 15

1. Other hazards associated with carrying

solid bulk cargoes Entering enclosed spaces Always follow the appropriate procedures before entering an enclosed space, taking into account IMO Resolution A.1050(27) – Revised Recommendations for Entering Enclosed Spaces Aboard Ships. Note: after a cargo space or tank has been tested and generally found to be safe for entry, small areas may exist where oxygen is deficient or toxic fumes are still present. Pesticides The risks of using pesticides include the accumulation of gas in spaces adjacent to the cargo holds being treated or, if it is essential to ventilate the cargo holds, accumulation of fumigant in accommodation and working areas. Fumigants such as Phosphine and Methyl Bromide are poisonous to humans and if they are not handled correctly, they can also represent a fire risk. They should only be used by specialists and not by the ship’s crew. Carry out any fumigation in line with the IMO Circular, MSC.1/Circ.1264 – Recommendations on the Safe Use of Pesticides in Ships Applicable to the Fumigation of Cargo Holds. This is contained in the supplement to the Code. The ship should carry gas-detection equipment, adequate respiratory protective equipment, a copy of the latest version of the Medical First Aid Guide for Use in Accidents Involving Dangerous Goods (MFAG), and appropriate medicines and medical equipment.

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 16

1. Checklist and flowchart for accepting

and loading solid bulk cargoes The checklist on the next two pages and the loading flowchart on page 18 bring together all the key procedures you must follow when accepting and loading any solid bulk cargo for shipment. Always consult the Code itself to ensure that all mandatory measures and specific advice have been implemented. Download a large version of the flowchart at www.lr.org/imsbc Checklist Before accepting and loading a cargo, ask yourself the following questions:

1. Has the shipper delivered the cargo information and documentation listed in

the Code under ‘Assessment of acceptability of consignments for safe shipment; Provision of Information’, including the correct Bulk Cargo Shipping Name (BCSN), and provided a declaration that the cargo information is correct?

1. Have you consulted the relevant IMSBC Code schedule to find out the cargo’s

general and specific hazards?

1. If the cargo is listed as Group A, have you followed all procedures relating

to the safe carriage of cargoes which may liquefy?

1. If the schedule indicates the cargo is Group B has the shipper provided

a statement that the chemical characteristics of the cargo are, to the best of his knowledge, those present at the time of loading?

a) If the cargo is classified as ‘dangerous goods in solid form in bulk’ does the

Master have a special list, manifest or stowage plan identifying its location,

are there instructions on board for emergency response, and does the ship

have a Document of Compliance for the Carriage of Dangerous Goods?

(Continued on the next page)

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 17 Checklist (continued)

1. Whatever the cargo Group, have you taken the recommended precautions

to remove or minimise the cargo’s hazard, including:

a) preparing recommended safety equipment and procedures?

b) activating any cargo monitoring equipment ready for loading?

1. Have the Master and Terminal Representative agreed a Loading Plan to ensure

that the permissible forces and moments on the ship are not exceeded during loading or unloading? This should include the sequence, quantity and rate of loading or unloading, the number of pours and the de-ballasting or ballasting capability of the ship.

1. Are there instructions to suspend the loading or unloading operation if

the ship’s limits are exceeded, or are likely to be exceeded if the operation continues?

1. Are you monitoring the cargo loading or unloading procedure, is the ship’s

draught being monitored and recorded in the logbook, and have any significant deviations been corrected?

• Before sailing on the loaded passage, have you considered other factors, such

as the ingress of water, which could affect the cargo during the passage? Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 18

Shipper has delivered all required cargo information? IMSBC Code consulted regarding cargo and associated hazards? Master and crew have the necessary ship data and information to prepare ship for loading? Cargo spaces inspected and prepared for cargo loading and stowage? Is the cargo categorised as ‘dangerous goods in solid form in bulk’ by SOLAS regulation VII/7? Does the Master have a special list/manifest/stowage plan identifying its location, are there instructions on board for emergency response, and does the ship have a ‘Document of Compliance for the Carriage of Dangerous Goods’? Has the Loading Plan been agreed with the Master and terminal representative? Are there instructions to suspend the loading/unloading operation if the ship’s limits are exceeded, or are likely to be exceeded if continued? YES YES LOAD (monitoring the cargo loading operation) A B C Not listed in IMSBC Code? NO TML Certificate? Moisture content (MC) declaration? MC lower than TML? Cargo spaces free from liquids? Visual monitoring conducted? Trimming considered? Stowage/segregation acceptable? Atmospheric conditions suitable and monitoring tools operational? Applicable dust precautions in place? Ventilation meets operational and IMSBC requirements? Competent authority has provided the Master with certificate stating carriage requirements (if Group A or B, agreed by authority of port of unloading and flag)? IMSBC Cargo Schedule consulted? Applicable precautions taken? DO NOT LOAD Cargo Group? NO NO NO NO NO NO NO NO Loading flowchart – follow the steps to see if it is safe to load your cargo

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 19 Appendix 1 – IMO regulations and guidance relating to the transport of solid bulk cargoes Mandatory regulations International Convention for the Safety of Life at Sea (SOLAS) – Chapter VI: Carriage of cargoes – Chapter VII: Carriage of dangerous goods – Chapter XII: Additional safety measures for bulk carriers International Maritime Solid Bulk Cargoes (IMSBC) Code – facilitates the safe stowage and shipment of solid bulk cargoes by providing information on the risks associated with their shipment, and the procedures to be adopted for carriage. International Code for the Safe Carriage of Grain in Bulk (International Grain Code) – contains regulations and guidance for ships engaged in the carriage of grain in bulk. Code of Safe Practice for Ships Carrying Timber Deck Cargoes – provides stowage, securing and other operational safety measures designed to ensure the safe transport of timber cargoes, primarily stowed on deck. MARPOL Annex V – Regulations for the Prevention of Pollution by Garbage from Ships – includes requirements for discharge of dry bulk cargo residues and cargo hold wash-water containing cargo residues. Guidance Code of Practice for the Safe Loading and Unloading of Bulk Carriers (BLU Code) – assists those responsible for the safe loading or unloading of bulk carriers to carry out their functions and to promote the safety of bulk carriers. Circulars and Resolutions – MSC.1/Circ.1160 – Manual on Loading and Unloading of Solid Bulk Cargoes for Terminal Representatives. – MSC. 1/Circ.1357 – Additional Considerations for the Safe Loading of Bulk Carriers

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 20 Circulars and Resolutions (continued) – MSC. 1/Circ.908 – Uniform Method of Measurement of the Density of Bulk Cargoes – MSC. 1/Circ.1395 – Lists of Solid Bulk Cargoes for which a Fixed Gas Fire-extinguishing System may be Exempted or for which a Fixed Gas Fire- extinguishing System is Ineffective – Resolution A.1050(27) – Revised Recommendations for Entering Enclosed Spaces Aboard Ships – MSC. 1/Circ.1264 – Recommendations on the Safe Use of Pesticides in Ships Applicable to the Fumigation of Cargo Holds – BC. 1/Circ.69 – Contact Names and Addresses of the Offices of Designated National Competent Authorities Responsible for the Safe Carriage of Grain and Solid Bulk Cargoes.

Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 21 Appendix 2 – the Code section by section Section 1 – General provisions – including application, definitions, related SOLAS regulations Section 2 – General loading, carriage and unloading precautions – cargo

distribution and loading/unloading procedures Section 3 – Safety of personnel and ship - poisoning, corrosive and asphyxiation hazards, dust hazards, flammability and ventilation Section 4 – Assessment of acceptability of consignments for safe shipment – identification of cargoes, and the tests and documentation required for their safe carriage Section 5 – Trimming procedures Section 6 – Methods of determining angle of repose Section 7 – Cargoes that may liquefy – the dangers of liquefaction, conditions under which liquefaction may take place and precautions to prevent it Section 8 – Test procedures for cargoes that may liquefy Section 9 – Materials possessing chemical hazards – classification of hazards, stowage and segregation requirements Section 10 – Carriage of solid bulk wastes – their classification, segregation and stowage Section 11 – Security provisions Section 12 – Stowage factor conversion tables Section 13 – References to related information and recommendations Appendices Appendix 1 – Individual schedules of solid bulk cargoes Appendix 2 – Laboratory test procedures, associated apparatus and standards Appendix 3 – Properties of solid bulk cargoes Appendix 4 – Index of solid bulk cargoes Supplement – this contains supporting documents including the BLU Code and IMO circulars (see Appendix 1)

Lloyd’s Register is an independent risk management organisation that works to improve its clients’ quality, safety, environmental and business performance throughout the world. Our expertise and activities cover shipping, railways, other land-based industries and oil and gas. Working from around 245 locations we employ some 7,600 personnel. We operate independently of any government or other body. Lloyd’s Register is a trading name of Lloyd’s Register Group Limited and its subsidiaries. For further details please see www.lr.org/entities To order additional copies of this pocket guide, please visit the Lloyd’s Register webstore: www.webstore.lr.org About the UK P&I Club One of the world’s largest P&I mutuals, the UK P&I Club currently insures around 200 million gross tons of owned and chartered ships in 60 countries. The Club publishes loss prevention material through a wide range of media on topics such as hazardous cargo in containers, human error, personal injury and maritime security. The full range of Club activities can be viewed at www.ukpandi.com About Intercargo Intercargo exists in order to present a unified voice for shipowners and operators in the dry bulk sector. Intercargo’s primary objective is to link industry stakeholders in a commitment to a safe, efficient and environmentally friendly dry cargo maritime industry. For further details see www.intercargo.org The full range of Club activities can be viewed at www.ukpandi.com

Annex D Material Safety Data Sheet for U-IBA

MATERIAL SAFETY DATA SHEET According to EU Directive nr. 1907/2006 and nr 1272/2008 en 453/2010 Untreated Incinerator bottom ash

Page: 1 of 6 Produced by: IndusTox Consult (fj)

Version: 1.0 Edition: 23-02-2012

• IDENTIFICATION OF THE MATERIAL

1. 1. Identification of the compound

Product name: Untreated incinerator bottom ash Chemical name: A mixture of particles consisting of minerals like calcium carbonate, calcium sulphate, iron oxide en aluminium silicates together with scrab and metal particles like aluminium, lead, zinc and copper Commercial name: Raw IBA CAS nr: Not applicable 1.2. Usage of the material Raw IBA is the untreated IBA which is produced as a residue by incineration of household and commercial waste. Raw bottom ash contains big pieces of bottom ash together with scrap and non ferrous metals mainly aluminium. Also there is a minor piece of unburned material in it. The product is heterogeneous. De raw bottom ash is upgraded by treatment to produce IBA which is a building material which is used in the road construction industry. The treatment consists of separating the scrap, non-ferrous particles and unburned pieces form the mineral pieces which are minimized to a maximum grain size of 40 mm.

1. 3. Identification of the producer

Producer: Address: Place: Telephone: Fax: E-mail: Web site:

1. 4. Telephone numer in case of Emergency

Producer: Telephone (24 hours a day): National contact: Telephone:

• IDENTIFICATION OF THE RISKS

1. 1. Classification of the material

Classification according to EU 1272/2008 No risk classification. This product contains respirable crystalline silica as an impurity. The concentration of respirable quartz is less than 1%. Therefore it is non- hazardous. The Industrial Mineral Producers (IMA) pose that silicoses is the critical effect of inhalation of respirable crystalline silica and that lung cancer is a secondary effect. This product is classified according to the criteria of the Industrial Mineral Producers (IMA). The product contains Pb as an impurity. Lead compounds are harmful to the reproduction. The concentration is less than 0,3 % and therefore the product is not classified as harmful to the reproduction. MATERIAL SAFETY DATA SHEET According to EU Directive nr. 1907/2006 and nr 1272/2008 en 453/2010 Untreated Incinerator bottom ash

Page: 2 of 6 Produced by: IndusTox Consult (fj)

Version: 1.0 Edition: 23-02-2012

EG-classification/EWC Non-hazardous waste with code 19.01.12 according to the European Waste catalogue: bottom ash and slag which are not in the category of bottom ash containing hazardous compounds”.

1. 2. Labelling

No risk labelling

1. 3. Other risks

None

• SAMENSTELLING / INFORMATIE OVER DE BESTANDDELEN

Name Percentage (v/v) Risks of the constituents Raw AEC-bodemas 100%

Pb-compound 1 - 2 g/kg H360, H332, H302, H373, H410 Zn-compound 3 - 4 g/kg H302, H314, H410 Sum of 6 PCB’s < 175 µg/kg H373, H410 Dioxins (as TEQ) 0,02- 0,05 µg/kg H373, H350 Total-PAH < 30 mg/kg H350, H340, H360, H317, H410 Respirable crystalline silica, under which quartz Max 0,01 % (v/v) H350, H373

• First Aid Arrangements

Rae IBA is rich of chalk and has a pH of 9 - 12 and therefore it is an alkaline product. The following points are addressed to this property 4.1. Ogen Effect: First aid: Possible irritating for the eyes. Red eyes and tears.

Flush with water and when necessary consult a doctor.

1. 2. Inhalation

Effect:

First aid: Possible light irritation of the air way, but not to expect.

Fresh air; when necessary consult a doctor.

1. 3. Swallowing

Effect:

First aid: Inflamed feeling in the stomach.

Flush mouth, drink water, consult the Urgent Help of the nearest hospital when the complaint last longer than one hour.

MATERIAL SAFETY DATA SHEET According to EU Directive nr. 1907/2006 and nr 1272/2008 en 453/2010 Untreated Incinerator bottom ash

Page: 3 of 6 Produced by: IndusTox Consult (fj)

Version: 1.0 Edition: 23-02-2012

1. 4. Skin

Effect:

First aid: Possibly light irritating, red skin.

Wash skin and hands with soap before eating, drinking, toileting and smoking, take of polluted clothes and wash regularly. When skin eruption occurs ask a doctor

• FIRE PREVENTION

Fire prevention by: Suitable quenching aids The material is inflammable. No dangerous situation can be expected in case of fire.

All quenching aids are permitted

• MEASUREMENTS WHEN DUST IS RELEASED ACCIDENATALY

1. 1. Personal precaution

Direct contact with the skin is not advisable. Wear complete covering working clothes, use working gloves and wear safety goggles and breath protection FFP3.

1. 2. Environment precaution

Keep windows and doors of building and vehicles closed when possible. Avoid contact of raw bottom ash with the sewer, surface water or direct contact with the soil.

1. 3. Cleaning methods

Avoid flying away of dust and spray after enduring dryness. Spoiled product can be discharged according to the European Waste catalogue (code 19.01.12). Don’t flush the spoiled material into the sewer or into the surface water.

• HANDLING AND STORAGE

1. 1. Safe handling

Dust category S5: Hardly sensitive to dust when > 15% of moisture. Can be categorized to higher class when the moisture content is < 15% Prevent forming of dust when handled. Use protection of airways when ventilation is not possible. Don’t eat, drink and smoke on the working place. Wash hands after using. Take of personal protection and working clothes when an eating place is entered.

MATERIAL SAFETY DATA SHEET According to EU Directive nr. 1907/2006 and nr 1272/2008 en 453/2010 Untreated Incinerator bottom ash

Page: 4 of 6 Produced by: IndusTox Consult (fj)

Version: 1.0 Edition: 23-02-2012

1. 2. Conditions for save storage

Raw bottom ash has puzzolanic properties whereby the granulate will form a bound mass. When stored in a closed facility ventilate.

• PRECAUTIONS TO CONTROL EXPOSURE / PERSONAL PROTECTION

1. 1. Limits for exposure

Technical arrangements: When handling IBA take measurement to prevent dust. Exposure limits Target limit for inhaling: 3,0 mg/m3 als TGG-8 uur.

Personal protection means Inhalation Working when dust: dusk mask with FFP3-filter. Hands Working with contact: Working gloves. Eye Working when dust: Safety goggles. Skin and body In dirt surroundings: Complete covering working clothes

• FYSICALK AND CHEMICAL PROPERTIES

1. 1. General information

Fysical form: Solid compound both lumps and dust. Can contain big pieces of metals Smell: Typical light cement smell Colour: Dark grey till black 9.2. Important health, environmental and safety information pH: Between 9 en 12 Density: When deposited ca. 1550 kg/m3 , Densed ca. 1800 kg/m3 Damp density: Not applicable Melting point: Not applicable Flame pint: Inflammable Boiling point: Not applicable Explosive: Non-explosive Solubility in water: Insoluble. A part of the metal compounds and salts can leach.

• STABILITY AND REACTIVITY

1. 1. Stability product

Avoid contact with groundwater because compounds which are harmful to the environment can be leached. When stored for a long time the temperature can rise very little because of cementious reactions..

MATERIAL SAFETY DATA SHEET According to EU Directive nr. 1907/2006 and nr 1272/2008 en 453/2010 Untreated Incinerator bottom ash

Page: 5 of 6 Produced by: IndusTox Consult (fj)

Version: 1.0 Edition: 23-02-2012

1. 2. Materials to avoid

Not known and not to expect

1. 3. Hazardous decomposing compounds

When contacted with water (H2O) possible formation of hydrogen gas (H2-gas). Under reducing circumstanced ammonia can be formed (NH3). Accidentally fosfine (PH3) can escape.

1. 4 Reactivity

Not known and not to expect

• TOXICOLOGAL INFORMATION

1. 1. Acute toxicity

Lethal dose (LD)50 oral, rat: Not known Direct consequences: When dust enters the eyes it can irritate. When swallow it can irritate the stomach and intestinal tube

1. 2. Chronically toxicity

Long term consequences: Not known.

ECOLOGICAL INFORMATION Ecotoxicity: Ecotoxicologal information is not known. It is to be expected that because of pH adjustment there can be effect in aquatic systems

INSTRUCTIONS FOR REMOVING 13.1. Waste/Unused products

Take spoiled material back to the pile or bring it to a certified handler according to the EWC.

1. 2. Packing

No special demands

• INFORMATION ABOUT TRANSPORT

UN-number and class: None ADR/ADNR/RID: Not applicable

General: No special precautions needed according to the transport demands for non hazardous goods. Avoid dust formation and spoiling during loading.

MATERIAL SAFETY DATA SHEET According to EU Directive nr. 1907/2006 and nr 1272/2008 en 453/2010 Untreated Incinerator bottom ash

Page: 6 of 6 Produced by: IndusTox Consult (fj)

Version: 1.0 Edition: 23-02-2012

Cover when transported.

OBLIGATORY LEGAL INFORMATION EG classification waste materials: Non-hazardous waste with code 19.01.12 according to the European Waste catalogue: bottom ash and slag which are not in the category of bottom ash containing hazardous compounds”. EG-hazardous classification EG-hazardous classification is not applicable because bottom ash is a waste. According to the classification of the Industrial Mineral Producers Association (IMA) this material is to be classified as non-hazardous (see also section 16)

OTHER INFORMATION 16.1. Other information IMA classification rules quartz

The Industrial Minerals Producers (IMA) have determined that respirable crystalline silica (=RCS) is best to classified as STOT RE Category 1 because of the danger of silicoses. STOT means ‘Specific Target Organ Systemic Toxicant’. RE means “Repeated Exposure”. Based on scientific evidence it is determined that is general necessary to inhale over a longer period repeated significant amounts of RCS on the working spot before possible health effect can occur. Because of the classification all products that contain RCS, despite the fact this is an impurity, additive or constituent has to be classified as:

• STOT RE 1, when de RCS concentration ≥10%;
• STOT RE 2, when de RCS concentration is between 1 and 10%.

Annex E UN Test N.5, extract from the Manual of Tests and Criteria

Annex F Marchwood Scientific Services flammable gas release test reports

MVV Devonport 117/563 2 February 2017

INTRODUCTION

Marchwood Scientific Services Ltd (MSSL) has been commissioned to perform an assessment of gas evolution for a provided sample of Incinerator Bottom ash (IBA).

Background- MSSL received a representative sample of IBA from MVV Devonport referenced MVV_IBA_250117_52_AWr for analysis. The analysis requirement was to assess the material for flammability/gas evolution in accordance with the latest United Nations Recommendations on the Transport of Dangerous Goods (ADR Regulations) In accordance with the ADR Regulations a substance would be listed as Class 4.3 Hazardous when a substance, which, when in contact with water emits a flammable gas. The regulations also state that Packing Group 111 shall be assigned to any substance which reacts slowly with water at ambient temperature such that the measured rate of evolution of flammable gas is greater than 1 litre per kilogram of substance per hour and which does not meet the criteria of Packing Groups 1 and 11. The regulations include the Manual of Tests and Criteria Part 111 Section 33.4 and specify a test that must be undertaken to assess a substance as above. The provided sample of IBA was tested by MSSL using methods in accordance with those specified, TEST RESULTS-

The results of the laboratory analysis of the provided samples are provided below. Note that for all of the tests undertaken the material was ground to µm particle size. This was not only required in order to obtain a sample of the correct specification for testing but also the regulations require this if initial inspection of a material indicates the presence of finer particles. This would indicate that the material may breakdown in transit.

Test specified in Section 33.4.1.4.3.2-

In accordance with this test a small quantity of substance ( diameter) was placed in a trough of demineralised water at 20° C

RESULT- NO GAS OBSERVED

Test specified in Section 33.4.1.4.3.3-

In accordance with this test a small portion of the ground ( diameter) material was placed on a filter paper (as support) and floated on the surface of an evaporating dish containing distilled water at 20° C.

RESULT- NO GAS OBSERVED

Test specified in Section 33.4.1.4.3.4-

In accordance with this test a portion of the ground ( diameter) material was placed in a pile measuring 20mm high and 30mm diameter and a hollow made in the top of the pile. A few drops of water were then added to the hollow.

RESULT- NO GAS OBSERVED

MVV Devonport 117/563 3 February 2017

Test specified in Section 33.4.1.4.3.5-

This method specifies that a portion of the ground ( µm diameter) material is placed in a conical flask and then water applied to the solid using a separating funnel. The rate/quantity of gas evolved, if any, is then measured over a prolonged period and the evolution rate calculated and compared with the limit value of 1 litre of gas evolved per kilogram of substance per hour.

Results- The method does not specify the quantity of water that should be applied to the solid material therefore some experimentation was undertaken. In order to accurately measure the rate of gas evolved a Buchner conical flask was used with side-arm. The side arm of the conical was then connected to a Tipping Bucket gas flow meter designed to measure very low gas flows. The Tipping Bucket gas flow meter is employed as part of the approved apparatus used as part of the Biomethane Potential (BMP) and Dynamic Respiration (DR4) tests. The separating funnel was placed onto the Buchner flask and the water added. The separating funnel was then removed and the Buchner flask sealed. The gas evolved was then measured. Applying a few drops of water resulted in no measurable gas flow. The gas was monitored over a period of 5 hours to check there was no delayed reaction. The test was then repeated using increasingly large quantities of water and the gas evolution measured. The results are provided in the table below-

Quantity of Solid used (grams) Quantity of water added (mls) Quantity of Solid used (kilograms) Gas evolved (litres/kilogram per hour) # Limit value (litres/kilogram per hour) Test Pass/Fail 19.2 3.0 0.019

1 Pass 18.6 10.0 0.019

1 Pass 18.4 20.0 0.018

1 Pass

this value is the maximum gas evolution rate per hour. Visibly some very small short-lived

emissions of gas were observed within the first minute of each of the tests but these were below the measuring capabilities of the test apparatus.

CONCLUSION The measured gas evolution rate for this material was below the 1 litre/Kg/Hour limit value.

Annex G Report by The Hydrogen Hazards Unit, London Southbank University

Asse

EXPL SCHO LOND 103 B LOND SE1 0

http://

Rep ssment of e OSION & F OL OF EN ON SOUTH OROUGH ON AA www.lsbu.a port to th the likel explosion FIRE RES NGINEERI H BANK ROAD ac.uk/esbe e Marine y hydroge (13th Ja 11t EARCH G NG UNIVERSIT /efrg/inde

Acciden

en concen nuary, Pl

th April 20 ROUP TY x.shtml t Investig tration i ymouth a 017 ation Bra nvolved i anchorage

anch n the Nor e) trader

Assessment of the likely hydrogen concentration involved in the Nortrader explosion (13th January, Plymouth anchorage)

From details of the accident investigation supplied it is clear that the explosion was initiated in the fo’c’s’le and propagated back into the main cargo hold causing further ignition of a second much larger volume of hydrogen.

In this case we can safely assume that the flammable gas mixture has not detonated (see Note A) and the explosion can be classed as a vented deflagration. In simple terms the explosion is likely to have proceeded as follows.

Prior to the explosion the cargo hold was nominally sealed. Following the propagation of the explosion back into the cargo hold the pressure in the hold will have begun to rise as hydrogen/air is combusted. The flame will accelerate (depending on the hydrogen concentration and turbulence generated by flow over obstacles) as it propagates down the length of the hold. After a period of time (in this case estimated to be of the order of 0.15s) the pressure generated has caused the cleats to fail and hatches to lift. Depending on the rate of combustion and hence rate of pressure rise the lifting of the hatches will limit the ultimate pressure experience by the hold/hull to some degree i.e. they act as explosion relief vents. In the case of hydrogen, vents that are both very large and that can open rapidly (low inertia) are usually required to prevent an enclosure experiencing a significant overpressure above that required to open the vent.

The primary aim of this report is the estimation of the likely concentration of hydrogen present in the hold at the time of ignition.

Minimum possible quantity of hydrogen The absolute minimum quantity of hydrogen involved in the explosion i.e. that needed to lift the weight of the hatches and break the cleats can be readily estimated by equating the hatch weight combined with the breaking force of the cleats to a force exerted on the hatches by an internal pressure in the cargo hold.

From the data supplied, the mass of all hatches (8x14 Tonne + 4 Tonne) equates to 116 Tonne and multiplying by gravity (9.8 m/s2) a weight of 1138 kN. For the cleats the total break force works out to be ((56 x 8.75 Tonne + 10 x 9 Tonne) x 9.8 = 5690 kN. This then gives a combined force required to lift the hatches and break the cleats (F) of 6828 kN.

The internal pressure in the hold required to produce this amount of lift can then be estimated as F divided by the total surface area of the hatches (S) of 556 m2. This then gives a pressure (Pv) of 12kPa (1.78 psi, 0.123 bar, 0.121 atm).

For explosions inside closed volumes the pressure generated can normally be equated to that determined theoretically by assuming constant volume adiabatic combustion. This pressure is most accurately determined using a thermochemical equilibrium solver such as STANJAN1 or GASEQ2. The results of such calculations for hydrogen/air at an initial pressure of 1 atm and 298K are given in Figure 1.

It can be seen from figure 1, that the concentration required to produce 0.121 atm is theoretically very small and works out to be only 0.3%(v/v). Estimating the ullage volume of the hold (Note B) as 1900 m3 this gives a volume of hydrogen of 6m3 .

However, as can also be seen from figure 1, this value is well below the Lower Flammable Limit of 4% (v/v). For an explosion to generate close to 0.121 atm at 0.3% hydrogen (v/v) the 0.3%(v/v) would need to be confined to a localised region (e.g. all in a thin layer at the top of the hold or in a localised plume close to the fo’c’s’le) at over 4% (v/v). For the situation described in this incident it seems highly unlikely that there would be significant stratification of hydrogen in the hold for the following reasons:

1 STANJAN is a computer program which calculates chemical equilibrium using the method of element potentials. It was developed by Professor W. C. Reynolds, Department of Mechanical Engineering, Stanford University, Stanford, CA 94305 2 GASEQ is a thermochemical equilibrium solver for combustion problems developed by Chris Morley. http://www.gaseq.co.uk/

 From information provided by the MAIB it seems likely that the aluminium in the ash and hence the hydrogen production was distributed fairly evenly across the surface of the ash  The generation and hence release of hydrogen to the cargo ullage was likely slow and chronic it would therefore be unlikely to have a large plume that would extend to ceiling level before mixing.  The relatively high temperature of the waste bed would induce convection currents in the cargo hold and aid mixing.

On the assumption that the hydrogen is well mixed above the cargo the concentration in the cargo ullage would have to have been at least at the lower flammable limit of 4% for ignition to occur.

However, if well mixed, the ullage concentration would almost certainly have been significantly above 4%(v/v) (concentration as volume per volume percentage). Whereas with most fuels there is negligible difference between the concentration at which a mixture becomes flammable and when it becomes explosive, this is not true for hydrogen. For most fuels the upwards flammability limit corresponds to an adiabatic flame temperature of around 1000°C and the concentration is typically around 50% of the stoichiometric3 concentration. In the case of hydrogen, however, the theoretical adiabatic flame temperature of a 4% (v/v) mixture is only 350°C and the concentration is 14% of stoichiometric. The reason for this is related to the high diffusivity of hydrogen in comparison to oxygen. Preferential diffusion occurs and the flame becomes unstable and breaks into cells. The hydrogen concentration at the convex flame cells increases and hence the localised flame temperatures, in excess of the calculated adiabatic flame temperature, are generated at the flame fronts. Near LFL (i.e.4%(v/v)) mixtures of hydrogen/air are therefore less “explosive” than near LFL mixtures of most other fuels (smaller temperature/pressure rises, slower flame velocities and incomplete combustion). Between 4 & 9 % the overpressure experienced is some fraction of that predicted by thermodynamic calculation, as shown by the green shaded region on figure 1, (probably almost zero for most quiescent vessels) and depends on the geometry of

3 The minimum concentration required to react with all the oxygen present. In the case of hydrogen/air 29.5%(v/v) .

the vessel, the point of ignition, initial turbulence of the mixture, and obstructions in the vessel which could cause turbulence. Flammability limits are often determined observing flame propagation in a tube. The 4%(v/v) limit is for upwards flame propagation i.e. ignition source at the bottom of a vertical tube . If ignition is at the top downwards propagation the length of the tube is not observed until 9%(v/v). If the tube is mounted horizontally sideways propagation is achieved around 6%(v/v). Additionally, mixtures at the upwards LFL of 4% require a flame or a high energy spark for ignition (15kV & 10J is typical for test apparatus).

Returning to the Nortrader incident, it was reported that there was an ignition of a volume of hydrogen/air mixture, starting inside a switch box in the fo’c’s’le, which then propagated back into the cargo hold (probably via an insecure cargo light access plate). For this to have occurred it seems likely that the hydrogen concentration in the cargo hold (and the fo’c’s’le) would have been at least 6%(v/v). At less than 6%(v/v) it’s unlikely the ignition in the foc’s’le would have propagated back into the cargo hold. For the foc’s’le to contain 6%(v) hydrogen the mixture vented into there from the hold would have had to have been at least that concentration. Additionally given the shape of the cargo hold, mixtures at much less than 6%(v/v), would quite likely not have generated sufficient overpressure to lift the hatches.

In summary it is concluded that the hold was most likely at a concentration of at least 6%(v/v) 114 m3 of hydrogen assuming an ullage volume of 1900 m3.

Maximum possible quantity of hydrogen. In order to try to place an upper bound on the hydrogen concentration in the hold, calculations have been carried out to gauge the likely effectiveness of the lifting of the hatches in reducing the overpressure experience by the vessel structure. It is clear that there was relatively little damage to the vessel. Regarding explosions occurring in industry in general from a comparison of damage to vulnerable refinery parts, by Stephens as cited in [1], significant damage might be expected at 0.3 bar and serious damage at 0.5 to 1 bar. It is also likely from the photos following the explosion that the hatchcovers were not lifted significantly since they are not

significantly displaced from their original positions (some returned to their original positions but were clearly lifted to an extent as their cleats had broken) and so the vent area that was realised during the explosion would have been a relatively small fraction of the total hatch area.

One of the most important parameters that is required for explosion relief vent sizing is the rate of pressure rise within the vessel. It is found that the maximum rate of pressure rise in a vessel of volume (V) is characterised by the “cubic law” given below:

G 3 / 1 max K constant ) / (   V dt dP

KG is known as the explosion index (deflagration index) and normally quoted for hydrogen as 550 bar.m.s-1. KG is a measure of the violence of an explosion. A value of 550 bar.m.s-1 is a worst case value applicable to hydrogen/air mixtures at around 30%(v/v). Hydrogen has one of the highest KG values and is therefore difficult to vent. However, at lower hydrogen concentrations, flame speeds and rate of pressure rise reduce considerably. The KG value at lower concentrations KGc can be estimated from the eqn.

max S S K K c G Gc  

Where Smax is the maximum fundamental burning velocity (Smax= 3 m/s for hydrogen/ air), Sc is the fundamental burning velocity at a specific concentration. At 10% hydrogen in air (burning velocity 0.1 m/s) this suggests a KG of only 18 bar m s-1. Measured KG for low concentrations of hydrogen are sparse but data is presented for H2/O2 mixtures at an initial pressure of 5 bar absolute and 20°C in Reference 2. At 10%(v/v) the KG value was observed to be only 10 bar m s-1, reducing to only 0.8 bar m s-1 at 6%(v/v). This would suggest that the above equation is possibly slightly pessimistic.

NFPA68 [3] provides a number of venting correlations intended for the design of relief vents on process plant. The equation, from NFPA68, below applies to enclosure capable of withstanding more than 0.1 bar.

      3 2 5722 0 5817 0 10 1.0 1754 .0 0567 .0 log 1265 .0 V P P P K A stat red red G            

Where

A

geometrical vent area (m2)

KG

Explosion index bar m s-1

Pstat

Static activation pressure (bar)

Pred

Reduced explosion pressure (bar)

V

enclosure volume (m3)

For long enclosures an expression to calculate additional vent area to be added to that calculated from the above equation is given by.

750 2 2              D L K A A G v

Unfortunately this only applies up to a length to diameter ratio of 5. Taking the length to diameter as 5 in the case of the cargo hold will potentially underestimate the required area.

The above equations can be applied to the Nortrader explosion to gauge what scale of vent opening would have been needed to be realised in order to keep the overpressure experience by the vessel hull etc ( i.e. Pred ) to a level commensurate with the observed damage, for varying initial hydrogen concentrations (by varying KG). Assuming that Pred = 0.3 bar and taking Pstat as 0.123 bar the vent area can be determined for different KG values and hence hydrogen concentrations. The free volume (ullage) in the cargo hold was estimated as 1900 m3 (Note B). These are summarised in the table below.

H2%(v/v) KG (bar m s-1) Pred (bar) Vent Area, m2 6 1 0.3 * 10 18 0.3 40

15 70 0.3 103 30 550 0.3 694 6 1 0.5 * 10 18 0.5 30 15 70 0.5 77 30 550 0.5 515 6 1 1 * 10 18 1 20 15 70 1 51 30 550 1 344

• Correlation goes negative for KG values less than 3 suggesting little vent area

would be required.

It can be seen from the above table that above 10% Hydrogen the required vent area increases substantially. Additionally the vent inertia of the hatch panels is considerably in excess of that recommended for explosion relief vents (i.e. 200 kg/m2 as opposed to 20 kg/m2). Bjerketvedt [1] et al state that, in their experience vent panels of 30 to 50 kg/m2 are too heavy to have a significant effect on peak pressure. The higher the hydrogen concentration and rate of pressure rise the greater the influence of the vent intertia.

It is clear that during the incident the hatch panels did move and vent the explosion to some degree, but it seems unlikely that they could have been opened sufficiently within the timescales required to effectively vent anything other than a weak explosion. Had the panels been lifted sufficiently to realise a vent area of 15m2 that would equate to lifting all the panels vertically to a height of around 0.1m.

Overall, while it is difficult to be certain, the maximum hydrogen concentration in the cargo hold was unlikely to have been much in excess of around 9%(v/v) hydrogen where combustion would become fast and complete. This would correlate with a flame temperature of around 1000°C.

Conclusions In summary, considering the evidence available, the likely hydrogen concentration in the cargo hold at the time of the explosion was in the range of 6 to 9 %(v/v). This equates to a volume of between 114 and 171 m3. The damage could have been caused by a smaller total amount of hydrogen confined to a localised volume at higher concentration, but this seems less likely given the distributed and slow/steady nature of the hydrogen release. The upper bound on the concentration is based on the limited damage observed and the displacement of the hatches and their likely poor effectiveness at venting an explosion (limiting overpressure) at higher concentrations.

References

D. Bjerketvedt et al. “Gas Explosion Handbook”, Journal of Hazardous Materials 52 (1997) 1-150

Schroeder V., Holtappels K., Explosion characteristics of hydrogen-air and hydrogen-oxygen mixtures at elevated pressures, International Conference on Hydrogen Safety, September 8-10 2005, Congress Place, Pisa, Italy

NFPA 68, “Guide for venting of deflagrations”, National Fire Protection Association, Quincy, MA, 2002 Note A The distinction is important in assessing the likely consequences of an explosion. During a deflagration the flame propagation is subsonic and the pressure produced by an explosion inside an enclosure can be approximated to a static pressure, whereas a detonation exerts localised shock loadings. In the case here, detonation can be excluded on the basis that primarily a) likely hydrogen concentration was too low to sustain a detonation and also b) likely insufficient run – up distance for a DDT (Deflagration to Detonation Transition) considering that the ullage is relatively uncongested and c) the nature of damage is inconsistent with a detonation.

Note B The volume was calculated by taking the hold dimensions as 61.5 x 9 x 6.16 = 3409 m3 and then subtracting the volume of the cargo (1505 m3 ) giving an ullage volume in the hold of 1900 m3. The cargo volume was estimated assuming a load of 2333 tonne and a density of 1550 kg/m3. .

Figure 1 The AICC overpressure produced for a given hydrogen/air mixture

Annex H MAIB safety flyer

SAFETY FLYER TO THE SHIPPING INDUSTRY Explosion of gas released from cargo of unprocessed incinerator bottom ash on Nortrader, 13 January 2017 Narrative At 1447 on 13 January 2017, the dry cargo vessel Nortrader suffered two explosions in quick succession. The vessel was anchored in Plymouth Sound and had loaded unprocessed incinerator bottom ash (U-IBA) the day before at Plymouth. The first explosion was in the forecastle store and the second in the cargo hold. The force of the explosion in the cargo hold dislodged all the heavy steel hatch covers and all but one of the holding down cleats were broken. Many of the hatch covers and sections of the hold coaming were distorted. The chief engineer was in the forecastle store at the time, testing the emergency fire pump. He suffered second degree burns and had to be airlifted to a nearby hospital. His recovery was prolonged and he was only declared fit for work again 4 months later. The vessel was out of service until 20 April. Nortrader at Victoria Wharf, Plymouth

The MAIB investigation established the following: •• Nortrader’s master had received an email from the vessel’s charterer that the intended cargo of U-IBA was a ‘non dangerous’ cargo. •• Hydrogen gas released from the cargo seeped into the forecastle store through a slack cargo light access cover. The hold was unventilated at the time. •• All the natural ventilation outlets in the forward compartment, including the forecastle store, were closed in preparation for imminent heavy weather. This allowed hydrogen to accumulate in the forecastle store. •• The starter panel for the emergency fire pump had a broken fixing lug at its door and was therefore not sealed. •• The first explosion was likely to have been triggered by the arcing between the contact terminals of the emergency fire pump starter panel when it was switched off. •• There had been 34 similar shipments of U-IBA from Plymouth to the Netherlands. Despite U-IBA not being listed in the International Maritime Solid Bulk Cargoes (IMSBC) Code, no steps had been taken to seek approval from the competent authorities for its carriage by sea. Safety lessons •• The only approved means of ensuring safe carriage of solid bulk cargoes is through the application of the IMSBC Code. The schedules included in Appendix 1 of the Code clearly classify cargoes and identify the necessary requirements for safe carriage. Where no schedule is listed for a specific cargo, the required classification procedure is stated at Section 4 of the Code, with further guidance provided by MSC Circ.1453, Marine Guidance Note (MGN) 512 and the guidance document Carrying solid bulk cargoes safely published by Lloyd's Register, UK P&I Club and INTERCARGO. •• The application of the IMSBC Code has been mandatory since 2011. It replaced the Code of Safe Practice for Solid Bulk Cargoes (BC Code), which was for guidance only. •• The IMSBC Code requires the shipper to provide the vessel’s master with a comprehensive list of information about the intended cargo. This flyer and the MAIB’s investigation report are posted on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch Email: maib@dft.gsi.gov.uk First Floor, Spring Place 105 Commercial Road Tel: 023 8039 5500 Southampton, SO15 1GH

FD06040 ｜ 2012_英国_carbon-monoxide-poisoning-on-board-the-sports-cruiser-emma-louise-with-loss-of-2-lives__20.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2012 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H ACCIDENT REPORT ACCIDENT REPORT VERY SERIOUS MARINE CASUALTY REPORT NO 1/2023 APRIL 2023 1 Carbon monoxide poisoning on board the sports cruiser Emma Louise resulting in two fatalities at Port Hamble Marina, River Hamble, England on 11 January 2022 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an such investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2023 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: +44 (0)23 8039 5500 SUMMARY On the morning of 12 January 2022, two men were found unconscious on board the sports cruiser, Emma Louise, which was berthed in Port Hamble Marina, on the River Hamble, England. It was later established that both men had died because of carbon monoxide poisoning. The two men had boarded the boat the previous evening and had left the engine running while they remained in the boat’s covered cockpit area. The MAIB investigation concluded that the boat’s exhaust gas, which contained carbon monoxide, had likely been funnelled into the cockpit by an inflatable towable ski ring that was suspended from the transom of Emma Louise. The levels of carbon monoxide would have increased in the cockpit, rendering both men unconscious and causing their deaths. There was no means of alerting the men to the danger because a carbon monoxide alarm had not been fitted to the boat. In November 2022, the MAIB participated in raising awareness of hazards to boat users during Carbon Monoxide Awareness Week and, in conjunction with the publication of this report, has issued a flyer to summarise this accident and highlight the key safety lessons to the boating community. Emma LouiseThe sports cruiser Emma Louise moored starboard side to a finger pontoon.

2 FACTUAL INFORMATION Narrative At about 1700 on 10 January 2022, the owner of Emma Louise and his brother-in-law arrived at Port Hamble Marina, having travelled from their homes in Slough, England. They boarded Emma Louise, which was moored to a pontoon, and prepared to spend two nights on board. Shortly afterwards they went ashore to eat at a nearby restaurant and, at about 2000, returned to Emma Louise, where they slept on board. On the morning of 11 January, the owner went home to attend to work matters while his brother-in-law visited Hamble. During the late afternoon they met up outside the restaurant in Port Hamble Marina and boarded Emma Louise at approximately 1700. It was drizzling with light, occasionally gentle, northerly wind, the air temperature was 10˚C and there was no swell; there were squally rain showers later in the evening. The men started the boat’s engine and switched on a portable electric radiator, the cockpit lights, navigation lights and the entertainment radio. They then sat in the cockpit with its canopy closed and each drank a beer. At 1727, the owner called his wife and told her that he and his brother-in-law planned to stay on board Emma Louise that evening and have a meal and some wine. At about 1000 on 12 January, the owner’s wife attempted to call her husband but he did not answer. She became concerned and called the Port Hamble Marina office. As a result of her call, two of the marina’s staff walked to the pontoon. They found Emma Louise with its engine idling. They could see the two occupants through the closed cockpit canopy window and both of them were unresponsive when the marina staff called out to them. The marina staff opened the canopy and turned the boat’s engine off. They then checked the two men and called the emergency services, having been unable to find a pulse on either of them. At about 1030, ambulance staff attended and were followed shortly afterwards by officers from both the police and the fire and rescue service. At around 1050, both men were pronounced deceased. The casualties Emma Louise’s owner, William Traynor, was 44 years old, 1.87m tall and weighed 96kg. He bought the boat in May 2020 and later named it Emma Louise, initially keeping it on the River Thames. In May 2021, he relocated Emma Louise to Port Hamble Marina, and he and his family frequently used it on the Solent. The owner’s brother-in-law, Martin Steventon, was 39 years old,1.76m tall and weighed 71kg. He had little boating experience but enjoyed spending time on board Emma Louise with his brother-in-law. On 19 January, postmortems identified the primary cause of death as Carbon monoxide (CO) toxicity to both men. The levels of blood carboxyhaemoglobin (COHb)1 saturation were recorded as 75% for the owner and 72% for his brother-in-law. The postmortems indicated blood alcohol levels of 0.5 and 1.5 times the drink driving limit in England for the owner and his brother-in-law respectively; the postmortem report stated that the alcohol levels could possibly be attributed to putrefaction of the bodies. Emma Louise description Emma Louise was a 5.5m glass reinforced plastic Fletcher 18 GTS sports cruiser, built in 1996. The boat had a cockpit area with a cushioned transverse bench seat to aft, below which the engine was fitted into a separate compartment (Figure 1). The cockpit had two pedestal mounted seats; the helm and boat controls were mounted in a console in front of the starboard seat. There was a low wraparound 1 Carboxyhaemoglobin describes the combination of CO and haemoglobin that forms in red blood cells when CO is inhaled or produced in normal metabolism.

3 Figure 1: Emma Louise plan (a) and side elevation (b) For illustrative purposes only: not to scale Console Battery compartment Denatured alcohol stove Bilge blower Ventilation grille Ventilation grille Seat Bench seat Engine Engine Transom PORT STARBOARD Covered cockpit Enclosed cabin Portable electric cool box (under deck hatch) Portable electric radiator Mattress Seat Cabin entrance Windscreen a Closed forward hatch AFT FORWARD b Flagpole socket Canvas canopy Waterline exhaust Transom Covered cockpit Enclosed cabin windscreen to the front of the console. A single burner denatured alcohol stove was stowed behind a hatch in the port side inner bulwark of the cockpit. A portable electric cool box was stowed below a hatch in the cockpit floor. The cockpit area was covered by a canvas canopy, which had clear plastic windows and was supported by a tubular frame and fastened to the hull and windscreen by means of snap poppers. The canopy included panels that provided access and ventilation into the cockpit when they were open; these were secured with heavy duty zips and twist connectors when closed. There were small gaps where the canopy fitted around a railing mounted on the boat’s transom. The canopy was otherwise well-fitted, with no large air gaps where it met the boat’s hull and windscreen, and the canvas material was in good condition with no rips. Forward of the cockpit area there was a small enclosed cabin with a mattress. The cabin ceiling was fitted with a hatch that opened onto the foredeck. Emma Louise was equipped with a four cylinder 3.0 litre inboard petrol engine that developed 135 horsepower through a steerable outdrive. The engine had a wet exhaust system, which vented the exhaust gases and cooling water through a hull fitting in the centre of the boat’s transom at the waterline.

4 An electric fan, known as a bilge blower, ventilated the engine compartment through a grille in the port side of the hull’s topsides. There was a separate battery compartment adjacent to the engine compartment, with a ventilation grille to the starboard side of the hull’s topsides. The owner had the stern drive rebuilt and the engine replaced in August 2021. The boat’s electrical power was supplied by 12 volt direct current (VDC) batteries, which were charged by an alternator via a belt drive from the engine. The batteries operated the engine starter motor, navigation lights, cockpit lights, entertainment radio, bilge blower and portable electric cool box. A shore power cable supplied 240 volt alternating current (VAC) to a double domestic socket, into which an 800 watt portable electric radiator was plugged. Emma Louise and crew as found condition The marina staff found Emma Louise moored starboard side to a finger pontoon with its bow pointing to the north. The boat’s engine was running at idle out of gear with the bilge blower turned off. The canvas canopy was zipped closed and fastened down and the cabin ceiling hatch was secured shut. The denatured alcohol stove was stowed away. The entertainment radio, cockpit lights and navigating lights were turned on and the portable electric cool box, which contained some food and beer, was plugged in. An inflatable towable ski ring (known as a ringo) was suspended from a flagpole socket at the transom of Emma Louise (Figure 2). The ringo was protected by a removeable cover. The 240 VAC shore supply cable was plugged into a socket on the pontoon; the portable electric radiator was switched on and was warm. The owner of Emma Louise was lying on his back on the deck, parallel to the aft bench seat, and his brother-in-law was slumped in the seat that was behind the helm, with his face against the electric radiator. Both men were fully clothed. They had each consumed half of their bottle of beer; the meal and red wine were untouched. There had been no activity on either of their mobile phones since 1800 on 11 January 2022. Figure 2: Emma Louise's transom with ringo hanging from flagpole socketLooking down onto the ringo in its removable cover. It is clearly hooked onto the flagpole socket and rests against a folded-up engine mounting board as it hangs over the motor cruiser’s transom. Boat inspection tests Emma Louise was relocated to Southampton, England, where the MAIB carried out a series of tests in its as found condition; 160kg of dispersed weight was added to simulate the effect of the two men on the boat’s draught and trim. There were light northerly breezes during the tests, with air temperatures of approximately 6˚C. The initial series of tests used a forward-looking infrared (FLIR)2, Optical Gas Imaging (OGI) camera to film the passage and direction of the engine’s exhaust gas. Gas detectors3 were also placed on the boat in the following locations: • above the waterline exhaust on the transom outside the canopy; • on top of the vertical cushion of the aft bench seat; 2 The FLIR camera was supplied and operated by FLIR Systems Ltd. FLIR OGI cameras can be tuned to the wavelength of specific gases such as CO and CO2 and make these invisible gases appear as dark smoke when filmed. 3 Honeywell BW™ Flex series portable multiple Gas Detector, capable of measuring up to 2000ppm of CO; ppm represents the number of units of a substance present in a million parts of another substance. Ringo under removable cover Flagpole socket

5 • on the console; • in the engine compartment. A domestic CO detector4 was also placed on the console. The test results from the idling engine showed that: • 2000 parts per million (ppm) of CO was detected in the atmosphere just above the waterline exhaust; • 2000 ppm of CO and 19.5% of oxygen (O2)5 was detected in the atmosphere in the covered cockpit above the bench seat; • The exhaust gas tended to funnel upwards from the waterline between the ringo and the boat’s transom (Figure 3a) and then seeped through small gaps between the boat’s canopy and transom railing (Figures 3b and 3c); • Exhaust gas was rising on convection currents above the portable electric radiator; • of CO was detected in the engine compartment; • The domestic CO detector on the console alarmed within 10 minutes of starting the test; • Light grey engine exhaust gas was emitted just below the waterline. The second series of tests carried out in the boat’s as found condition used CO gas test detectors6 with a greater CO measuring range than used for the initial tests. These were positioned just above the waterline engine exhaust and within the cockpit, with the following results: • ≤7000ppm of CO was measured in the atmosphere just above the waterline exhaust; • The CO concentration in the covered cockpit varied, with an average of 984ppm during a 120-minute test, reaching 1400ppm within 35 minutes and a peak value of 2000ppm after 100 minutes. A further test was carried out with the covered ringo removed from the transom (Figure 4) using the same CO detectors. The CO levels in the cockpit reached 200ppm after about 14 minutes before settling at around 300ppm after 30 minutes. Exhaust gas and boat ventilation Common sources of gas emissions found on pleasure boats include internal combustion engine and portable generator exhausts, cookers and cabin heaters. Typical exhaust emissions from a well-serviced petrol engine at idle can represent 50,000ppm (5%) CO, with the rest of the exhaust gas comprised mostly of carbon dioxide (CO2). Engines running at less than optimal conditions due either to wear or poorly adjusted fuel settings tend to emit both black smoke rich in unburnt fuel and a higher than normal CO content. Insufficient ventilation in boats can lead to the accumulation of dangerous CO levels where exhaust gases had entered cabins and cockpits. Carbon monoxide poisoning CO is a highly poisonous gas that is 3% lighter than air. It is a colourless, odourless and tasteless product of incomplete combustion. When CO is inhaled COHb forms, which readily replaces oxygen in the blood. Common symptoms of low exposure to CO poisoning include headache, nausea, stomach pain, difficulty breathing, tiredness and confusion. The effects become more severe as the CO concentration and duration of exposure increase (Table 1) and can be expedited where the atmosphere is also deficient in O2 and rich in CO2. The effects of a COHb level of 60 to 70% in blood include unconsciousness, convulsions, respiratory failure and death. 4 CO detector conformed to British Standard EN 50291-1:2010+A1:2012 Electrical apparatus for the detection of carbon monoxide in domestic premises. 5 Normal O2 content in air is 21.5%. 6 Dräger X-am® 5000 multi gas detectors with data-logging functionality capable of measuring up to 10,000ppm CO.

6 Figure 3: Representation of the likely passage of exhaust gas from waterline exhaust (a) into covered cockpit (b and c) based on the FLIR OGI camera Image A shows the sports cruiser’s transom angled to show the railings behind the ringo as it hangs from the flagpole socket. Three arrows between the transom and the ringo indicate how gas from the waterline exhaust tended to funnel up behind the ringo towards the small gaps between transom railings and the canvas canopy. Image B is an exterior close-up highlighting gaps in the sport cruiser’s canvas canopy where it is fitted around two bars to mount railings to the transom. Arrows illustrate how the exhaust gas rose from behind the top of the ringo toward these gaps and seeped into the otherwise enclosed cockpit. Image C shows the gaps in the canopy from inside the cockpit. The railing bars sit behind an aft bench seat and arrows illustrate the path of the gas through gaps into the enclosed area. Exhaust gas enters cockpit through small gaps Ringo a b c Exhaust gas rises from waterline exhaust and funnels upwards behind ringo Small gap between canopy and transom railing Figure 4: Emma Louise's transom with ringo removed Stern view of the sports cruiser Emma Louise moored starboard side to a finger pontoon. The cockpit canopy is closed and the ringo has been removed to reveal the transom. Grab rails run either side of central transom steps and a flagpole socket attached near to the top of starboard side rail is indicated. Another arrow points to the central location of the wet exhaust below the swim platform and just above the waterline. Flagpole socket Waterline exhaust

7 Table 1: Effects of carbon monoxide exposure (NORSOK Standard Z-013) CO (concentration) Physiological effects 1500 ppm Headache after 15 minutes, collapse after 30 minutes, death after 1 hour 2000 ppm Headache after 10 minutes, collapse after 20 minutes, death after 45 minutes 3000 ppm Maximum "safe" exposure for 5 minutes, danger of collapse in 10 minutes 6000 ppm Headache and dizziness in 1 to 2 minutes, danger of death in 10 to 15 minutes 12800 ppm Immediate effect, unconscious after 2 to 3 breaths, danger of death in 1 to 3 minutes Carbon monoxide detectors CO detectors are relatively inexpensive and are both easy to install and test. They are designed to sense CO levels that have accumulated over a period of time or suddenly increased, which would then trigger an audible and visual alarm. CO detectors for domestic use in the UK are required to meet BS EN 50291- 1, while detectors that comply with the more stringent BS EN 50291-2 standard are best suited for boats. Both types are required to activate an alarm when predetermined levels of CO are detected (Table 2). Table 2: British Standard CO detector alarm concentration and activation levels CO (concentration) Alarm must not activate before Alarm must activate before 30 ppm 120 minutes Not applicable 50 ppm 60 minutes 90 minutes 100 ppm 10 minutes 40 minutes 300 ppm Not applicable 3 minutes Carbon monoxide legislation and information At the time of the accident CO detectors were not required to be fitted on board recreational craft operating in UK coastal waters. The Royal Yachting Association (RYA) provided guidance on the potential sources and danger of CO on its website, recommending that CO alarms were fitted on board boats and regularly tested by the owners. Pleasure craft operating on the majority of the UK’s inland waterways were expected to undergo a Boat Safety Scheme (BSS)7 inspection every 4 years. The BSS inspection included an assessment of the boat’s ventilation and, since April 2019, had required that a CO alarm was fitted on board. The BSS publication Carbon Monoxide Safety On Boats was produced in collaboration with the Council of Gas Detection and Environmental Monitoring (CoGDEM) and was updated in May 2019. The publication was available on the BSS and CoGDEM websites and highlighted the hazard presented by CO, including a warning for mariners that: on a moving boat and even when moored, cockpit awnings can act as a funnel to draw engine gas inside the boat. [sic] 7 The purpose of the BSS is to help minimise the risk of boat fires, explosions or pollution harming users of inland waterways.

8 On the purchase of CO detectors, it advised boat owners to: … choose ones that meet the international standard BS EN 50291-2; these are best suited for boats. Some river and coastal marinas have voluntarily adopted the BSS inspection scheme for resident boats but this was not a requirement at Port Hamble Marina. In 2015, the UK All-Party Parliamentary Carbon Monoxide Group published its Carbon Monoxide: From Awareness to Action report that highlighted many CO safety challenges, including the difficulties of raising and relying on CO awareness to ensure individuals are adequately protected from the dangers. Similar accidents Over the past 10 years the MAIB has investigated five accidents involving CO poisoning on pleasure vessels that (including this one) resulted in the tragic loss of nine lives. In April 2014, two recreational boat users died of CO poisoning while sleeping on board the motor cruiser Arniston when gas leaked from a portable generator exhaust (MAIB Report 2/20158). The occupants had not been alerted to the presence of CO because the alarms fitted on board did not work. The MAIB made a recommendation to multiple organisations, including the BSS, MCA, RYA and CoGDEM to coordinate efforts to raise awareness in the pleasure boat community about the dangers of CO and the importance of fitting CO alarms. In June 2016, CO poisoning resulted in two fatalities on board the motor cruiser Love for Lydia, which was licenced to operate on an inland waterway (MAIB Report 9/20179). The petrol driven inboard engine had been left running while alongside, probably to charge batteries, causing the exhaust from the engine to fill the cockpit canopy and spread into the cabin. No CO alarm was fitted on board and the occupants were overcome by the gas. The MAIB made several recommendations, including to the MCA to continue to build on current initiatives by coordinating relevant organisations to focus efforts on raising the awareness of the dangers of CO and the importance of fitting carbon monoxide alarms. It also made a recommendation to the BSS regarding the compulsory installation of CO alarms in craft participating in the BSS scheme. In November 2016, the owner of pleasure boat Vasquez (MAIB Report 18/201710) was overcome and died from CO poisoning when exhaust gas transferred from the engine compartment into the boat’s enclosed canopy and cabin. The petrol engine’s exhaust system, which had been modified during the boat’s life, had leaked into the engine compartment. No CO alarm was fitted on board. In December 2019, two men died of CO poisoning from a diesel cabin heater exhaust that was leaking into the cabin on board the motor cruiser Diversion (MAIB Report 4/202111). The cabin heater’s exhaust was not designed for marine use, its connection was not gas tight and the heater had neither been serviced nor checked by a professional heater installer. The boat’s owner and his friend were preparing to go to sleep when they were overcome by the CO; they were not alerted to the danger because no CO alarm was fitted on board. In 2020, the Office for National Statistics recorded 21 fatalities in England and Wales attributed to accidental CO poisoning in land-based scenarios. 8 https://www.gov.uk/maib-reports/carbon-monoxide-poisoning-on-motor-cruiser-arniston-with-loss-of-2-lives 9 https://www.gov.uk/maib-reports/carbon-monoxide-poisoning-on-board-the-motor-cruiser-love-for-lydia-with-the-loss-of-2- lives 10 https://www.gov.uk/maib-reports/carbon-monoxide-poisoning-on-motor-cruiser-vasquez-with-loss-of-1-life 11 https://www.gov.uk/maib-reports/carbon-monoxide-poisoning-on-motor-cruiser-diversion-with-loss-of-2-lives

9 ANALYSIS The owner of Emma Louise and his brother-in-law died from CO poisoning due to a build-up of exhaust gas within the boat’s covered cockpit while it was moored alongside. The CO was emitted as part of the exhaust gases released from the boat’s running engine; it is likely that the gases seeped under the canopy at the transom into the cockpit. The CO was able to accumulate and quickly reach fatal levels because the canopy was well-fitted with no large air gaps, which resulted in a lack of ventilation. The return to the boat When the two men returned to Emma Louise on the night of the accident the weather conditions were squally. Once they were on board and had turned on the cockpit lights they secured the canopy and switched on the portable electric radiator, probably, to keep warm and dry. There was no evidence that they planned to take the boat out on the river that evening so it is probable that the navigation lights were turned on in error. As the radiator was plugged into the domestic socket supplied by the shore power cable it is likely that the engine was started to maintain power to the cockpit lights, portable cool electric box and entertainment radio, which was supplied by the boat’s batteries. The source and path of carbon monoxide The alcohol stove had not been used on the evening of the accident and has been discounted as a potential source of CO. MAIB testing found no evidence of an exhaust leak inside the engine compartment that might have transferred into the cockpit. The tests confirmed that the engine exhaust exited freely into the atmosphere at the waterline (Figure 3a), indicating that the boat’s draught and the lack of swell did not set up a back pressure on the exhaust that might have affected engine performance. The amount of CO in the atmosphere at the exhaust during the tests (≤7000ppm) and the lack of black smoke in the exhaust gas emitted did not provide any indication that the engine, which had been replaced in the year before the accident, was defective. Although atmospheric conditions on the evening of the accident were not identical to the MAIB post- accident tests, it is likely that the passage of the exhaust gas into the boat would have been similar. When the engine was started it is probable that the warm exhaust gas rose and was funnelled between the ringo and the boat’s transom up to the lower edge of the canopy, where it then seeped into the cockpit through the small gaps around the transom railing (Figure 5). The CO rich and O2 deficient atmosphere that was detected above the aft bench seat during the tests showed the potential for the incoming gas to degrade the atmosphere in the cockpit. Although CO is only slightly lighter than air, it is possible that convection currents rising from the warm portable electric radiator caused the CO’ within the exhaust gas to mix with the air and distribute within the cockpit, which was fully enclosed by the canopy. The canopy was intact, fully closed up and well-fitted to the hull and windscreen, while the forward hatch was closed and the engine bilge blower was not running. Consequently, it is likely that only a small amount of the polluted atmosphere would have been refreshed with clean air. The collapse and poisoning The air quality in the cockpit’s atmosphere is likely to have deteriorated relatively quickly. The as found position of the two men suggested that the owner of Emma Louise was sitting on the aft bench seat and his brother-in-law was in the seat behind the helm at the time of the accident. Neither man was significantly under the influence of alcohol, so it is unlikely that their alcohol intake had a major effect on their physiology. It is possible that the air at their head height was sufficiently saturated with CO and CO2 and was slightly deficient in O2; this would have likely caused both men to fall unconscious at a similar time. The absence of mobile phone activity after 1800 and the fact that the meal and wine were untouched suggest that they fell unconscious not long after they arrived on board Emma Louise. The levels of accumulated CO identified during the MAIB tests combined with CO2 and air that was slightly deficient in O2 showed that similar exposure to the two men was sufficient for them to remain unconscious and for their blood COHb to rise above 70%, thereby causing their deaths later that evening.

10 Figure 5: Representation of possible path of exhaust gas within Emma Louise’s covered cockpit area For illustrative purposes only: not to scale Exhaust gas enters cockpit and circulates on convection currents above portable electric radiator in front of starboard chair Exhaust gas enters cockpit through small gaps between canopy and transom railing Exhaust gas rises from waterline exhaust and funnels upwards behind ringo Transom Covered cockpit Enclosed cabin AFT FORWARD Ringo Closed forward hatch Although the MAIB tests demonstrated that some exhaust gas did transfer into the cockpit when the engine was running with the ringo removed, it is likely that the two men’s exposure to CO would have been at a much lower level that probably would not have resulted in their deaths. CO is always present within the exhaust of an engine that runs on fossil fuel; however, the risk of CO poisoning can be reduced if a boat’s engine is not run while the vessel is stationary and canopies and cabins are well ventilated. Use of carbon monoxide detectors The MAIB tests showed that if either a domestic or marine approved CO detector had been installed in the cockpit of Emma Louise it would inevitably have alarmed within 10 minutes of the boat’s engine being started. This would have alerted the men to the presence of CO and provided them with an opportunity to stop the engine and ventilate the cockpit. An operational CO detector could have prevented the needless loss of lives in this case and also the tragic accidents on board the pleasure craft Diversion, Vasquez, Love for Lydia and Arniston. It is essential that CO alarms are fitted in any areas where carbon monoxide could accumulate and pose a risk to health, including boats’ cockpits and accommodation areas. Preference should be given to selecting detectors marked as meeting BS EN 50291-2, which are intended for use in a marine environment. It is critical to fit the CO alarm properly, following the manufacturer’s installation instructions, and to regularly test it. Although there is no requirement for pleasure boat owners operating from most coastal marinas to install a CO detector, the value of such an inexpensive and easy to fit safety device on all pleasure craft cannot be underestimated. It also makes good sense to adhere to the BSS inspection regime, which includes a requirement for CO alarms following the MAIB’s recommendation to the BSS as part of the Love for Lydia investigation. Carbon monoxide awareness The high number of MAIB investigations into CO-related fatalities involving pleasure craft over the last 10 years, coupled with the number of land-based fatalities due to CO poisoning in 2020, demonstrate the very real threat the gas poses. However, despite the range of good quality information produced by the RYA, CoGDEM and BSS, it appears that awareness of the danger of CO and how to prevent it causing harm remains low within the pleasure boat community.

11 Although MAIB recommendations made as part of the Arniston and Love for Lydia investigations to tackle this lack of awareness were accepted by the MCA and other stakeholders, the tragic loss of two lives in this accident demonstrates that there is further work to be done by industry stakeholders to raise CO awareness among users of coastal pleasure craft. CONCLUSIONS • The two men died on board Emma Louise as a result of CO poisoning due to a build-up of exhaust gas within the boat’s covered cockpit area where they were seated. • It is likely that the exhaust gas was funnelled into the covered cockpit of Emma Louise by an inflatable towable ski ring that was suspended at the transom. • The cockpit canopy on board Emma Louise was intact and well fitted and the forward hatch was secured. Therefore, there was little ventilation to enable either the CO to disperse or fresh air to enter the boat. • It is possible that convection currents rising from the warm portable electric radiator caused the CO within the exhaust gas to distribute within the cockpit. • The two men had no means of being alerted to the rising CO level before being rendered unconscious and then succumbing, as there was no CO alarm on board Emma Louise. • The increasing number of fatalities caused by CO poisoning on board pleasure craft indicates that there is further work to be done by industry stakeholders to raise CO awareness among pleasure craft users. ACTION TAKEN MAIB actions • In November 2022, the MAIB participated in the UK government’s Carbon Monoxide Awareness Week, released a statement from the Chief Inspector of Marine Accidents12 and produced an information video13 in support of the campaign. • The MAIB has issued a safety flyer highlighting the hazards presented by carbon monoxide and distributed this to marina operators, pleasure boat trade organisations and the yachting media in conjunction with the publication of this report. RECOMMENDATIONS Given the recommendations issued by the MAIB as a result of its Love for Lydia investigation, and the subsequent actions taken during this investigation, no further recommendations are made in this report. 12 https://www.gov.uk/government/news/carbon-monoxide-awareness-week-2022 13 https://www.youtube.com/watch?v=jNatvhBrlqo&ab_channel=MarineAccidentInvestigationBranch

12 VESSEL PARTICULARS Vessel’s name Emma Louise Flag UK Type Fletcher 18 GTS sports cruiser Registered owner Privately owned Manager(s) Not applicable Year of build 1996 Construction Glass reinforced plastic Length overall 5.5m Gross tonnage Not applicable Minimum safe manning Not applicable Authorised cargo Not applicable VOYAGE PARTICULARS Port of departure Not applicable Port of arrival Not applicable Type of voyage Moored Cargo information Not applicable Manning 2 MARINE CASUALTY INFORMATION Date and time 11 January 2022 in the early evening Type of marine casualty or incident Very Serious Marine Casualty Location of incident Port Hamble Marina, Hampshire Place on board Cockpit Injuries/fatalities 2 fatalities Damage/environmental impact None Ship operation Moored Voyage segment In port External & internal environment Light northerly wind; drizzle; 10°C. Persons on board 2

FD06041 ｜ 2012_英国_carbon-monoxide-poisoning-on-motor-cruiser-vasquez-with-loss-of-1-life__MAIBInvReport-18_2.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2012 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H VERY SERIOUS MARINE CASUALTY REPORT NO 18/2017 AUGUST 2017 1 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an such investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2017 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gsi.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 ACCIDENT REPORT Fatal CO poisoning on board motor cruiser Vasquez while moored at Cardiff Yacht Club 12 November 2016 SUMMARY At approximately 1200 (UTC1) on 12 November 2016, the owner of the 7.75m motor cruiser Vasquez fell unconscious after being overcome by carbon monoxide (CO) that had emitted from his boat’s inboard petrol engine. Although rescuers came to his aid and conducted cardio-pulmonary resuscitation, it was not possible to save his life. Two of the rescuers had to be treated for the effects of CO poisoning following the accident. The CO was found to have originated from the rubber bellows of the wet-exhaust system of the engine, which was not only leaking gas but also water. The boat’s engine had not been regularly serviced, and evidence revealed that the exhaust system of the engine had been modified during the boat’s life. Notwithstanding the initiatives already taken by various stakeholders from the recreational boating sector, there remains a significant task to raise boat owners’ awareness about the dangers of CO. Due to the odourless nature of CO, the fitting of a detector/alarm remains the only effective warning that the poisonous gas is present. Given the recommendations issued following the MAIB’s recent Love for Lydia investigation, no further recommendations are made. 1 UTC - Universal Co-ordinated Time. Figure 1: Vasquez

2 FACTUAL INFORMATION Narrative During the morning of 12 November 2016, Ray Milton travelled to Cardiff Yacht Club to check on his motor cruiser Vasquez (Figure 1), which was moored in the marina. Ray regularly visited the club to socialise with other club members, usually on a weekly basis, but on this occasion had not visited for a couple of weeks. He had arranged to meet a friend at the club that morning. At approximately 1120, Ray’s friend walked down to where Vasquez was moored stern-to. He stepped onto the starboard bathing platform and leaned into the cockpit canopy, which was unzipped on the starboard side. Ray was sitting at the helm position and the aft floor area had been lifted to give access to the engine bay. There was water in the bilge and Ray was running the electric bilge pump to clear it. There was also water in the cabin, which was unusual. Once the water level was below the starter motor of the engine, Ray wanted to see if the engine would start. He turned the ignition key and the engine started successfully. The friend suggested putting the engine in gear ahead while still moored so that the boat’s bow would lift slightly and allow the water in the cabin to run aft into the engine bay bilge, where it could be pumped out easily. Ray followed this suggestion and water was seen to be running aft and being successfully pumped out by the electric bilge pump. About 10-15 minutes after arriving, the friend left and he asked Ray to take the engine out of gear so that he could step back onto the pontoon. He expected to catch up with Ray later in the clubhouse. After the friend had stepped onto the pontoon, Ray re-engaged the engine and at some stage replaced the cockpit flooring above the engine bay. At 1210, as he walked to the clubhouse, the friend called Ray on his mobile phone; there was no reply. He assumed Ray was busy and would ring back. He entered the clubhouse and ordered a drink at the bar. Approximately 30 minutes later, he tried calling Ray again, but still there was no reply. At 1242, he called another club member whom he knew was on a boat near Vasquez and asked him to check on Ray. When the other club member reached Vasquez he found the engine still running ahead, but there was no sign of Ray. With the assistance of another club member, who was an off-duty firefighter and happened to be nearby, he managed to pull the stern lines in sufficiently so that they could step onto the aft bathing platforms. Initially there was no sign of Ray, but they then saw a foot next to the helm. Emergency response The firefighter undid the port aft zip of the cockpit canopy and entered the cockpit. He saw Ray lying head-down in a pool of water in the cabin, with his feet stretched aft as if he had fallen forward. The firefighter turned off the engine ignition and the other club member raised the alarm by firstly calling the clubhouse for assistance and then, at 1248, calling for an ambulance. Ray was rolled onto his back and cardio-pulmonary resuscitation (CPR) was commenced. Shortly afterwards a further club member arrived from the clubhouse and climbed on board to assist the firefighter. They took it in turns to administer chest compressions and rescue breaths. After 10-15 minutes, the firefighter felt dizzy and developed a headache. The club member conducting CPR noticed the colour draining from the firefighter’s face and asked him if he was okay. The firefighter said he was fine and continued with CPR. As the firefighter’s condition appeared to worsen the other club member conducting CPR asked for the firefighter to be helped out of the boat, which was duly carried out by one of the small crowd of people now gathered on the pontoon.

3 The paramedics arrived shortly afterwards and were initially directed to attend to the firefighter. After examining him, the paramedics quickly removed the cockpit cover and then took over the CPR on Ray. With CPR continuing, Ray was lifted onto the aft seating area and Vasquez was towed round to another pontoon to allow better access ashore. Ray was then transferred to hospital by ambulance. The two club members who had administered CPR were given oxygen and then taken to hospital for further tests and treatment. The firefighter had a carboxyhemoglobin level of 20% when first assessed, while the other club member had a reading of 9.5%; both indicating CO poisoning. Ray did not recover consciousness and was declared deceased at hospital. The postmortem examination recorded that the cause of death was CO poisoning, noting a carboxyhemoglobin level of 51.8%. Vessel owner The owner of Vasquez, Ray Milton, was 72 years old and was a retired lorry driver. He had owned various motor boats throughout his life and had been a member at Cardiff Yacht Club for approximately 20 years. He bought Vasquez from a fellow club member 5-6 years prior to the accident. In common with many recreational boat owners, he neither held formal boating qualifications nor was he required to do so. Vessel description Vasquez was a Draco 2400ST motor cruiser constructed in 1985/86 in Norway. The design was certified by Det Norske Veritas in compliance with Nordic Requirements, a Norwegian boat construction standard. It was certified to carry up to 1400kg including a maximum of 10 persons. The vessel had a forepeak cabin with seating area, dedicated sleeping berth and toilet compartment. The cockpit, which could be enclosed with a canopy, contained the helm position (Figure 2) as well as a small galley and further seating. The engine compartment was situated under the floor of the cockpit at the aft end. The boat was driven by an inboard V6 Volvo Penta AQ225F petrol engine that was naturally aspirated (Figure 3). The vessel was fitted with an OMC stern drive (Figure 4), which provided the boat’s propulsion. It was not possible to determine if the stern drive was original or had been fitted later in the boat’s life. The engine had a wet-exhaust system, whereby cooling water, which was sucked in via ports in the stern drive, was combined with the engine exhaust gas and expelled from the stern drive through the hub of the propeller. Figure 2: Cockpit with view forward into the cabin

4 Situated in the engine compartment were the boat’s batteries as well as three bilge pumps. One bilge pump, which was positioned under the engine, was electrically driven and was operated by holding down a switch at the helm position, while the other two were hand pumps. There was a drain hole at the base of the engine compartment forward bulkhead that allowed water to flow freely through the bilge. Although a switch on the helm console was labelled ‘blower’, no blower was found in the engine compartment. Whether it had been fitted in the past and removed or had never been fitted is unknown. The blower, if fitted, would have normally been used to clear the engine bay of any combustible fumes prior to starting the engine. Figure 3: Volvo Penta AQ225F petrol engine Figure 4: OMC stern drive

5 Use of Vasquez Since buying Vasquez the owner had not used it frequently for trips, although he had visited the boat to check it and to start the engine periodically while at its mooring. During his ownership one major piece of work, costing over £1000, was undertaken on the engine by a local marine professional. However, it is believed that no regular servicing had been undertaken during the last 5-6 years. The owner used the boat very little in the 2-3 years prior to the accident and had discussed selling the boat with friends. However, in the months before the accident, following a trip on a friend’s boat to Portishead, the owner had taken a renewed interest in Vasquez and, a couple of months prior to the accident, had taken a trip around the bay. During the trip, the engine temperature had run very high, so after returning to the yacht club the boat was lifted out of the water. The cooling water inlets on the stern drive were found blocked. The stern drive was cleaned off and re-antifouled, and the oil in the stern drive was changed. Some 4-5 weeks before the accident Vasquez was taken on another test run and the over-heating problem had not re-occurred. However, the engine had not run smoothly at high revolutions and it was clear that it would need further attention before the boat could be used reliably outside Cardiff Bay. Engine testing and inspection Following the fatal accident, the MAIB conducted a reconstruction. The boat’s engine was started with the cockpit floor in place and the canopy erected, but with the aft starboard zip open. Atmosphere monitors and a domestic CO detector were positioned inside the canopy and viewed through the clear plastic of the canopy. After 40 minutes with the engine running, a CO reading of 250ppm2 was recorded on the atmosphere monitor at the helm position. The engine was then switched off, and over the next 5 minutes the CO level increased to 550ppm, before starting to slowly decrease. The local Volvo agent was then contracted to conduct an inspection of the engine and to establish the source of the CO leak. With the engine stopped and the boat afloat, a small water leak was apparent emanating from the rubber bellows on the wet- exhaust system. With the engine running, a considerable amount of exhaust fumes and water were emitted from the port and starboard bellows (Figure 5). 2 ppm - parts per million: HSE workplace exposure limits for CO - long term exposure limit 30ppm (8hrs), short term exposure limit 200ppm (15 minutes). Figure 5: Exhaust leak from rubber bellows

6 On closer examination, it was apparent that the exhaust system was not in accordance with manufacturer’s drawings and had been modified during the life of Vasquez. The following were observed: •• The exhaust manifolds were not original (Figure 6). •• The exhaust risers were not original (Figure 7). •• The cow horn castings had been cut down (Figure 7). •• The exhaust elbow had been connected on both sides of the engine with two unequal rubber bellows and a stepped aluminium tube retained by stainless steel hose clips (Figure 7). Although it could not be confirmed, the lack of corrosion on the bolts attaching the risers to the exhaust manifold suggested that the exhaust system had been included in the work carried out in the last 3-4 years. Figure 6: Drawing of original exhaust manifolds

7 Carbon monoxide hazard and legislation CO is produced by incomplete combustion. Potential sources include propulsion engines, generators, paraffin and oil lamps, log burners and barbecues. CO is a highly poisonous gas that weighs the same as air. It cannot be seen, smelled, tasted or felt. When breathed in, CO readily replaces oxygen in the human bloodstream by forming carboxyhemoglobin. This prevents oxygen from supplying the heart, brain and other vital organs. Common symptoms of low exposure to CO poisoning include headache, feeling sick, stomach pain, difficulty breathing, or feeling tired or confused. In greater concentrations a person will feel drowsy and lethargic, and may have difficulty walking and moving around. CO poisoning is not a problem limited to recreational craft and can also be an issue in the home. CO detectors/alarms of varying types are widely available and will provide an early warning of the presence of the poisonous gas. Many detectors have an audible and visual alarm, which operate when a certain Figure 7: Exhaust riser and rubber bellows Hose clips Rubber bellows Stepped aluminium tube Cow horn Exhaust riser

8 level of exposure to CO has been reached. However, they are dependent on having a power supply, often in the form of a small battery. Legislation has been introduced requiring CO detectors/alarms to be fitted: since 2010 in some new dwellings; since 2011 in new caravans; and since 2015 in rented properties. CO detectors/alarms, although widely recommended by organisations such as the Gas Safe Charity, CoGDEM3 and Boat Safety Scheme (BSS)4, are not required to be fitted under any UK regulations on board recreational craft. The Recreational Craft Directive (RCD), introduced in 1998, also has no provision for CO monitors, but does require inboard engines to be fitted in an enclosure separated from living quarters and installed to minimise the hazard from toxic fumes. However, these requirements would not have applied to Vasquez given that its construction predated the introduction of the RCD. There is no specific legislation that applies to the maintenance of recreational craft that operate in coastal regions. However, those operating on inland waterways have to satisfy the requirements of the BSS. Similar accidents In April 2014 two fatalities occurred on the motor cruiser Arniston5. The owner had fitted a portable generator in the boat’s engine compartment and rigged an exhaust system to discharge combustion gases overboard. The exhaust system failed during use while the vessel was moored, resulting in the exhaust fumes filling the motor cruiser’s cabin and leading to the death of two of the occupants. They had not been alerted to the presence of CO because the alarms fitted on board were not working. In June 2016 a further two fatalities resulted from CO poisoning on the motor cruiser Love for Lydia6. The petrol-driven inboard engine had been left running while alongside, probably to charge batteries. The exhaust from the engine filled the cockpit canopy and spread into the cabin. There was no CO detector/alarm fitted on board and the occupants were overcome by the fumes. The MAIB made recommendations to: the Maritime and Coastguard Agency regarding the co-ordination of raising CO awareness of the leisure boating community; British Marine to seek clarification on carbon monoxide requirements within the RCD; and the BSS regarding the compulsory installation of CO monitors in craft participating in the BSS scheme. ANALYSIS Accident scenario Ray Milton died from CO poisoning because his boat’s engine exhaust system was leaking the odourless gas into the cockpit and cabin area of his boat. With no CO detector/alarm fitted to his boat, he was unaware of the gas’s increasing presence. Over a period of time, with little ventilation to disperse the gas, a sufficient concentration of CO built up, such that he was overcome by the gas and collapsed in the cabin. The first responder on scene was also unaware of the presence of CO, and commenced CPR on Ray in the cabin not realising the danger he was placing himself in or that he would be breathing further CO into Ray’s lungs during CPR. The MAIB’s reconstruction demonstrated that, although the first responder had switched off the engine, the CO levels increased significantly in the following few minutes, making the situation worse. This was probably because, with the engine running, it was drawing in some of the atmosphere from the cockpit. With the engine switched off this effect was removed. 3 CoGDEM - The Council of Gas Detection and Environmental Monitoring. 4 Boat Safety Scheme - This scheme’s purpose is to help minimise the risk of boat fires, explosions, or pollution harming users of inland waterways. 5 www.gov.uk/maib-reports/carbon-monoxide-poisoning-on-motor-cruiser-arniston-with-loss-of-2-lives 6 www.gov.uk/maib-reports/carbon-monoxide-poisoning-on-board-the-motor-cruiser-love-for-lydia-with-the-loss-of-2-lives

9 CO awareness Department of Health statistics from 2013 show that around 40 people in the UK are killed by accidental CO poisoning each year. At least a further 4,000 people are treated in hospital for non-fatal CO poisoning. The number of further individuals who might have suffered from CO poisoning and not visited hospital as they have assumed some other illness is responsible (given the common symptoms) is impossible to quantify. However, it is apparent that awareness of the dangers of CO is not widespread in the UK as a whole, and this equally applies to those involved with recreational craft. In 2015, the All-Party Parliamentary Carbon Monoxide Group published a report, entitled ‘Carbon Monoxide: From Awareness to Action’, which highlighted many issues regarding CO safety, including the difficulties of raising CO awareness. Within the context of recreational boaters, several organisations, notably the BSS, CoGDEM and the Royal Yachting Association (RYA), have been proactive in raising awareness of the dangers posed by CO, and the BSS ‘Trusted Messenger’ initiative, which started in April 2017, has the potential to significantly improve CO safety on inland waterways. However, a significant task remains to raise CO awareness among the users of coastal recreational craft. It is welcome news that co-ordination of the task to raise awareness is underway following the MAIB’s Love for Lydia report’s recommendation. CO detectors/alarms and ventilation For those recreational craft not registered for use on inland waterways, such as Vasquez and numerous others operating around the UK coast, there is no requirement to fit a CO detector/alarm. CO is often regarded as a product of poor combustion and, therefore, poorly maintained equipment. However, as demonstrated in the Love for Lydia accident, a properly maintained engine can still emit significant quantities of CO from its exhaust. Simply mooring near a motor boat with its engine running could expose the crew to CO. A suitable CO detector/alarm is the only effective way to be warned of the presence of the gas. Once warned of the presence of the gas, steps can be taken to reduce its concentration by removing the source and by increasing ventilation. Unfortunately, when trying to keep warm while inside a recreational craft, the natural tendency is to try to limit draughts. This will only help with the build up of CO and is further justification for fitting a CO detector/alarm. There is a common misconception that CO detectors/alarms used in the marine environment are prone to false alarms. This is rarely the case. If the alarm sounds it will be because CO is present. The alarm must sound at relatively low concentrations of CO to provide enough warning for effective remedial action to be taken. CO detectors/alarms intended for recreational craft are tested to a rigorous standard (EN 50291-2), which should reassure boat owners that when a CO alarm sounds, suitable action must be taken. Recreational craft maintenance Although new recreational craft have to meet the requirements of the RCD, once purchased there is no requirement for ongoing maintenance of a private leisure vessel. In addition, for private recreational craft operating around the UK coast, there is no requirement to regularly have the safety of the vessel checked. There is, therefore, a significant onus on a boat owner to ensure their boat is maintained properly and that it remains safe to operate.

10 Although the full maintenance history of Vasquez’s engine could not be determined, it was apparent that the exhaust system had been modified and two rubber bellows added to the wet-exhaust pipework. Rubber bellows on a wet-exhaust system are vital in maintaining the watertight integrity of a boat and it is advisable to inspect them regularly and replace them every 2-3 years, as the rubber can harden and deteriorate. In this case, in addition to the rubber bellows being older than 3 years, the trial trip during which the engine overheated probably damaged the rubber. This ultimately led to a water/exhaust fumes leak, which caused Vasquez’s cabin to slowly flood and CO gas to enter the engine compartment while the engine was running. The RYA offers courses and provides guidance on various aspects of boating to educate boat owners on how to stay safe when afloat. It is important, however, to appreciate the extent of one’s knowledge when it comes to self-maintenance of a boat, particularly with regard to its engine, and when professional advice should be sought. Regular engine servicing in accordance with the engine manufacturer’s guidelines should ensure an engine remains reliable and safe to use. CONCLUSIONS •• Ray Milton died from CO poisoning because, without a CO detector/alarm being fitted to his boat, he was unaware that CO from his boat’s engine exhaust was emitting into the cockpit and cabin area. •• Given the recent fatal accidents from CO poisoning on recreational craft, there remains a significant task for the various stakeholders of the industry to assist in raising CO awareness. •• Given CO’s odourless nature, the fitting of a CO detector/alarm remains the only effective warning against being poisoned by the gas. •• CO detectors/alarms intended for recreational craft are tested to a rigorous standard, which should reassure boat owners that when a CO alarm sounds, suitable action needs to be taken immediately. •• It is important to seek professional advice and regularly service a boat’s engine, in accordance with the manufacturer’s guidelines, to ensure it remains reliable and safe to use.

11 ACTION TAKEN MAIB actions The Marine Accident Investigation Branch has: •• Issued a Safety Bulletin7, in August 2016, following the two fatalities on the motor cruiser Love for Lydia, highlighting the dangers of CO poisoning and the need for detectors. RECOMMENDATIONS Given the recommendations issued following the MAIB’s recent Love for Lydia investigation, no further recommendations are made. 7 www.gov.uk/maib-reports/safety-warning-after-carbon-monoxide-poisoning-on-motor-cruiser-results-in-loss-of-2-lives

12 SHIP PARTICULARS Vessel’s name Vasquez Flag Not applicable Classification society Not applicable IMO number/fishing numbers Not applicable Type Draco 2400ST motor cruiser Registered owner Privately owned Manager(s) Not applicable Construction Glass reinforced plastic Year of build 1985/86 Length overall 7.75m Gross tonnage Not applicable Minimum safe manning Not applicable Authorised cargo Not applicable VOYAGE PARTICULARS Port of departure Cardiff Port of arrival Cardiff Type of voyage Moored Cargo information Not applicable Manning 1 MARINE CASUALTY INFORMATION Date and time 12 November 2016, approximately 1200 Type of marine casualty or incident Very Serious Marine Casualty Location of incident Cardiff Yacht Club marina Place on board Cockpit Injuries/fatalities 1 fatality Damage/environmental impact None Ship operation Moored Voyage segment In port External & internal environment No breeze, dry day Persons on board 1

FD06042 ｜ 2012_英国_catastrophic-engine-failure-and-fire-on-board-ro-ro-passenger-ferry-wight-sky__MAIBInvRepo.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2012 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H SERIOUS MARINE CASUALTY REPORT NO 14/2018 JULY 2018 1 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an such investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2018 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 ACCIDENT REPORT Catastrophic engine failure, resulting in a fire and serious injuries to the engineer on board Wight Sky, off Yarmouth 12 September 2017 SUMMARY At 2133 on 12 September 2017, while approaching Yarmouth, Isle of Wight, the ro-ro passenger ferry Wight Sky (Figure 1) suffered a catastrophic failure of one of its main propulsion engines, followed by a fire. The fire was brought under control in less than 2 minutes, but the vessel’s engineer, who had been standing near the engine, suffered serious burn injuries to his hands and face. Although he was discharged from hospital 7 days later, he was subsequently diagnosed with post- traumatic stress disorder. An examination conducted by the vessel’s engine maker, Volvo Penta UK, concluded that the most probable trigger for the failure was debris in the engine’s oil channels following rebuild. The MAIB investigation established the following: •• The engine had been completely rebuilt and failed after 5½ hours of operation. •• The vessel’s soft patches 1 had not been removed, necessitating the engine to be lowered piecemeal into the engine room. 1 Soft patches: steel plates bolted down and sealed flush with the vehicle deck, that can be removed to allow large pieces of ship’s machinery or equipment to be removed/inserted. Figure 1: Wight Sky

2 •• Debris could have entered the engine’s oil channels during the 3 days that the partially assembled engine had been exposed to the elements. •• Analysis of oil samples from the engine indicated that accelerated wear had commenced before the engine failure. •• The power supply to the essential services switchboard, which distributed power to critical equipment including the fixed fire-extinguishing system, was lost 27 minutes after the accident. Actions have been taken by the vessel’s owner, Wightlink Limited, and by Volvo Penta UK, to address some of the issues identified in this report. A recommendation has been made to Volvo Penta UK, to consider offering wear particle detection technology for marine engines that cannot be easily serviced on board. FACTUAL INFORMATION Narrative On 12 September 2017, the ro-ro passenger ferry Wight Sky departed Lymington at 2100 2 for the 40-minute passage to Yarmouth. There were 40 persons on board. Main propulsion engine 2 (ME2) had been put back into service following a complete rebuild by Volvo Penta authorised service engineers RK Marine Limited (RKM). The crew of the morning shift had tested ME2 off load and had verified that all alarms and shutdowns were functioning correctly. During the afternoon shift, the vessel’s engineer tested the engine on load, and on departure from Lymington that evening all four main engines were sharing the sea load. At around 2130, the engineer went down to the forward engine room from the machinery monitoring room (MMR), located on the main vehicle deck immediately above, to note the machinery running parameters for the logbook. He saw that the high temperature cooling water header tank for ME2 required topping up and, while he was filling the tank with coolant, he heard an unusually loud sound from the engine. The mechanic was attending to main propulsion engine 1 (ME1) located in the same space (Figure 2a). At 2133:51, as the engineer rushed towards ME2’s emergency shutdown button, there was a loud explosion from its crankcase (Figure 2b). The engineer, who was wearing a full sleeved cotton boiler suit, was momentarily engulfed in a ball of fire and experienced intense pain. He ran out of the engine room and, assisted by the mechanic, went to the 2 All times are UTC. Mechanic Engineer Figure 2a: CCTV footage before engine explosion (CCTV time was 13 minutes slower than alarm system time) ME1 ME2

3 MMR. They called the bridge using the talk back system and apprised the master of the situation. At the time, the master was making his approach to the ferry terminal, which was approximately 300m away. The bosun was at the wheel. The master abandoned berthing, and at 2135 released the Hi-fog 3 fire suppression system for the forward engine room. The master observed on the closed-circuit television (CCTV) screen that the fire appeared to be extinguished less than a minute after the Hi-fog had been activated. On the master’s order, the bosun made an announcement to inform the passengers of the incident and to reassure them that the situation was under control. The master then called the emergency services and asked for ambulance and fire services to attend the vessel at the berth. Wight Sky berthed at 2142 and all passengers and vehicles were disembarked in an orderly fashion. Meanwhile the ship’s crew attempted to make the engineer comfortable. Shortly after berthing, an ambulance with paramedics arrived on board Wight Sky and, after assessing the engineer, took him to the hospital. The fire brigade arrived at approximately 2200. When firefighters entered the engine room a few minutes later, the Hi-fog system had stopped. On inspecting the engine room, the firefighters found a small fire still burning inside ME2’s crankcase. They extinguished this with a portable dry powder extinguisher. On the afternoon of 13 September, Wight Sky was towed to Portsmouth for repairs. The damaged engine was replaced with a new one and the vessel was returned to service on 26 October. The engineer suffered burns to his hands and face. Although he was discharged from hospital after 7 days, he was subsequently diagnosed with post-traumatic stress disorder. He had not returned to work at the time of publishing this report. Sequence of events and damage observed The following sequence of events was captured by the vessel’s Kongsberg alarm recording and CCTV systems: •• 2133:08 ME2 lube oil pressure shutdown •• 2133:10 ME2 crankcase pressure; coolant level; common shutdown •• 2133:51 Explosion (CCTV footage) •• 2134:25 Fire alarm •• 2134:59 Hi-fog activated •• 2135:58 Fire apparently extinguished (CCTV footage) •• 2202:29 Hi-fog pump unit fault 3 Hi-fog: Marioff Hi-fog System, a water mist-based fire suppression system. Figure 2b: CCTV footage showing engine explosion

4 Most of the visible damage was observed in and around cylinder unit 4 of ME2. The big end bearing was distorted radially (Figure 3a), white metal had melted onto the bearing pin on the crankshaft (Figure 3b) and two of the four big end bearing bolts had sheared at the joining face between the top and bottom halves of the bearing keep. The outboard side of the crankcase was punctured (Figure 3c), creating a large opening through which several components, including the gudgeon pin, connecting rod and fragments of the piston and cylinder liner had been ejected. There was fire damage on and around the engine (Figure 3d). Company and crew Wightlink Limited (Wightlink) operated eight vessels on scheduled routes between the Isle of Wight and the south coast of England. Three sister vessels, Wight Sky, Wight Light and Wight Sun operated on the Lymington-Yarmouth route. The other vessels in the fleet, including two high speed vessels, operated on routes from Portsmouth to the Isle of Wight. On the day of the accident Wight Sky’s crew comprised the master, mate, bosun, three deck ratings, an engineer and a mechanic. The master and engineer held STCW 4 certificates of competency appropriate to their ranks. The master had 7 years’ experience and the engineer 26 years’ experience in their respective roles on board Wightlink ferries. Layout of machinery space and engine maintenance Wight Sky had a Voith-Schneider cycloidal propulsion unit at each end of the vessel. Each propeller shaft was belt-driven by a pair of Volvo Penta D16C-A MH, six-cylinder engines, with an output of 485kW per engine at 1800rpm

1. ME1 and ME2 were located in the forward engine room and ME3 and ME4 in the

aft engine room. The Hi-fog fire-extinguishing system with its tanks, pumps and release switches was located in the space between the two engine rooms (Figure 4). There was an agreement between the vessel’s classification society, Lloyd’s Register, Wightlink and their service agents, RKM 6 that the Volvo Penta engines on all three sister vessels would be completely stripped down and examined every 16000 to 20000 hours. RKM had carried out 16 engine rebuilds on the three sister vessels, including that of ME2 of Wight Sky. The assembly of the engine’s block complete with liners, pistons, connecting rods, bearings and crankshaft, but without the cylinder head or sump pan was referred to as the ‘short block’. In all 16 instances, the short blocks were built in RKM’s workshop and were delivered to the vessels’ engine rooms via the emergency escape hatch. The water and oil channels, internal components and machined surfaces were protected by a loose plastic sheet during transit and while waiting to be lowered into the engine room. Openings to engine room Each engine room had a 0.69m by 0.69m emergency escape hatch on the starboard side leading to the vehicle deck above. The aperture of the hatch was restricted by a handrail and ladder. The short block was the largest engine assembly that could be moved through the hatch opening. All of Wight Sky’s four main propulsion engines had a ‘soft patch’ in the deckhead directly above them that could be removed to permit the removal and replacement of complete engines (Figure 5). Each opening was 3.85m by 1.95m and was closed with a steel plate bolted down and sealed flush with the vehicle deck. At the deckhead level in the engine room, the soft patch was insulated and covered over with cladding plates held in place by pop rivets. Opening the soft patch involved the removal of several items associated with the engine, including a large section of exhaust pipe, crankcase breather, engine lifting beam and several brackets supporting the pipes. 4 STCW - International Convention on Standards of Training, Certification and Watchkeeping for Seafarers 1978, as amended. 5 rpm: revolutions per minute. 6 The company carried out all the major servicing of Volvo Penta engines in the Wightlink fleet.

5 Figure 3: Damage to main engine 2 3a - Distorted big end bearing cylinder unit 4 3b - Bearing pin of cylinder unit 4 3c - hole in engine crankcase 3d

6 Figure 4: General arrangement of machinery spaces Watertight doors Hi-fog system Forward machinery space Aft machinery space Main engine 1 Main engine 2 Clutch Coupling Belt drive Voith Schneider propulsion Output shaft Belts

7 0.69m 0.69m 3.85m 1.95m Figure 5: Soft patch and escape hatch

8 Wightlink and RKM had agreed to move the engine in and out of the engine room using the emergency escape hatches. In October 2017, the replacement ME2 was lowered into the engine room through the soft patch opening. It took external contractors 8 hours to remove the soft patch and 16 hours to refit it. Previous engine failure on Wight Sky In July 2017, ME2 developed a coolant loss problem. The ship’s crew and RKM were unable to diagnose the cause of this. Several cooled components, including the turbocharger and associated exhaust piping, were replaced, but the problem persisted. On 4 August, the cylinder head was replaced by RKM, but on testing the engine thick white smoke was observed at the engine exhaust. On 11 August, the turbocharger was replaced, when the engine was run up one of the two exhaust valves on cylinder unit 5 sheared, resulting in the lower half becoming embedded in the piston. The failure was not investigated but was attributed by RKM to incorrectly adjusted valve tappet clearances. Engine rebuild Following the failure of ME2 on 11 August, RKM, with the agreement of Wightlink, decided to complete a full rebuild of the engine. The work commenced on August 18 at RKM’s workshop. The engine components used for the rebuild were a mixture of new and used components from other D16 engines in the fleet. The crankshaft was polished and crack tested at a specialist workshop subcontracted by RKM. The crankshaft oil channels were blown through with compressed air before assembly. The engine block was pressure washed, dried and a rust inhibitor applied to its internal surfaces. The short bock was assembled using the polished crankshaft and new pistons, cylinder liners and bearing shells. The short block and the remaining components of the engine were transported to Lymington ferry terminal on Friday, 25 August. There is conflicting evidence as to whether it was delivered on board and left on deck or remained on the quay. On 29 August, the Tuesday following the August bank holiday, the short block, cylinder head and sump pan were lowered separately into the forward engine room via the emergency escape hatch by Wightlink staff. Between 29 August and 4 September, two diesel fitters from RKM completed the engine rebuild in Wight Sky’s engine room. At the time, they were unable to assemble the coupling connecting the engine to the output shaft as the thread of one of the coupling’s bolt holes was damaged. On 4 September, RKM tested the engine uncoupled from the output shaft and left the vessel with instructions to the vessel’s engineer to load test the engine once it had been connected to the output shaft. Engine lubrication system The lubrication system (Figure 6) consisted of a crankshaft-driven oil pump, which delivered the sump oil through two full flow filters of 40-micron mesh size and a third, part-flow filter of 10-micron mesh size. The total quantity of lubricating oil was 55 litres. An oil channel along the inboard side of the engine block conveyed the lubricating oil to the main bearings. From the main bearing the oil was fed to the big end bearing immediately forward of it through a drilled channel. Piston cooling and lubrication oil was carried through a separate channel on the outboard side of the block.

9 Examination of the engine Prior to the explosion, ME2 had been run for 5½ hours since its rebuild. On 19 September, the engine cylinder block, sump pan, clutch and external pipes of ME2 were removed by RKM in the presence of MAIB inspectors. The engine short block, with all the remaining components, was packed and delivered to temporary storage on 4 October and to Volvo Penta’s UK head office at Warwick on 9 November. On 15 November, specialist engineers from Volvo Penta, Sweden, examined the engine components in the presence of representatives of the MAIB, RKM and Wightlink. The tightening torque of the bolts on bearing keeps of all the undamaged main and big end bearings was checked and confirmed to be correct. The crankshaft journal dimensions were also checked to establish that they were true. Signs of overheating at main bearing 5 at the top end of its journal nearest to big end bearing 4 were observed. Upon dismantling main bearing 5, its shells were found to have turned (Figure 7). Both the bearing shells and journal had suffered severe damage, but there was no remaining evidence of debris embedded in the bearing shells. The investigation report by Volvo Penta concluded that, when the shells for main bearing 5 turned, they blocked the lubricating oil channel on the main journal leading from the journal to cylinder unit 4’s crank pin. This caused the big end bearing to seize on the crank pin, leading to the failure of the bearing bolts. The rotating crankshaft then hit the unrestrained connecting rod, which impacted against the side of the crankcase, puncturing it and breaking the piston cooling oil gallery. The report concluded: We believe the breakdown was occurred due to remaining debris in the engine oil channels and the main bearing 5 after the overhaul. And also considering that the breakdown occurred only 5 hrs. after the refurbishment. [sic] Part-flow filters of 10-micron mesh size Full-flow filters of 40-micron mesh size Lubricating oil to main journal crank pin and gudgeon pin Oil pump suction Oil in sump Piston cooling oil nozzle Figure 6: Engine lubricating oil system Forward filter Crankshaft driven oil pump Centre filter

10 Connecting rod Bearing shell Big end bearing Gudgeon pin Piston Crankpin Drilled oil passage Main journal Turned bearing shell Figure 7: Schematic of crankshaft assembly

11 Tests carried out The forward and centre filter (Figure 6), along with oil samples from forward filter housing and the engine sump were analysed independently for their physical and chemical characteristics and accumulated wear particles. The report on the analysis of particles in the filters stated: Both of the oil filters tested contained high quantities of particulate matter...Some particles exhibited evidence of localised heating. Some of the particles observed were rather large (over 1mm in size). A few critical parameters are compared in Table 1. Oil from sump Oil from forward filter Iron (ppm7) 17 8 Copper (ppm) 27 27 Lead (ppm) 18 13 PQ Index8 8 217 Water (ppm) 2472 279 Flash point - indication of fuel dilution (°C) 135

200

Table 1: Comparison of test results of engine lubricating oil samples To understand the 7mode of failure of the big end bearing bolts of cylinder unit 4, they were also independently tested. 8The test report concluded that all bolts failed by ductile failure (either in tension or shear). The report further observed that the two bolts on one side of the bearing keep had failed in tension, and the two on the other side in shear. Essential services switchboard The forward and aft engine rooms had their own main switchboards. A common essential services (ES) switchboard was connected to each main switchboard by a dedicated circuit breaker. The ES switchboard supplied electrical power to critical equipment such as the water-mist fire suppression system 9 (Hi-fog), watertight doors, rescue boat davit, main hydraulic power pack and emergency lighting. The ES circuit breakers had an automatic and manual mode and were intended to be left in automatic mode so that irrespective of which main switchboard was in use, the critical equipment was not deprived of electrical power. On the day of the accident, the forward switchboard was in use. It shut down just after 2200 after the electrical power supply automatically switched over to the aft switchboard. As a result of an earlier oversight, the ES circuit breaker for the aft switchboard had been left in the manual mode, so the ES switchboard was left without power. This caused the loss of all essential services, including the Hi-fog system. 7 ppm: parts per million 8 PQ Index: A relative measurement of the total ferrous metal content of oil detected by magnetic field. 9 The Hi-fog system had a reserve tank of water pressurised by nitrogen, sufficient to provide up to 60 seconds of coverage for the protected space.

12 Previous accidents and failures In February 2018, D16 main propulsion engine on Wight Light shut down on low oil pressure. The analysis of the sump oil from the engine indicated a large amount of copper, lead and iron wear particles. Subsequent strip down examination of the engine by RKM, found that the main bearing shells of cylinder unit 2 had turned. On 7 September 2017, the 15.87m crew transfer vessel, Windcat 8, was on passage to Grimsby, UK, from the Lynn Wind Farm in the North Sea with two crew and eight windfarm technicians on board. Shortly after setting off, the vessel’s port engine, which was the same make and model as that on Wight Sky, suffered catastrophic damage and caught fire. The MAIB investigation 10 identified that one of the engine’s connecting rod big end shell bearings had failed, resulting in the connecting rod assembly releasing and penetrating the engine crankcase. In November 2015, Wight Sun suffered a failure of its D16 main propulsion engine. The cause of failure was established to be turned shells of a main bearing. ANALYSIS Immediate consequences of the accident Wight Sky was approaching the ferry terminal at Yarmouth when ME2 exploded, causing a fire in the engine room and injuring the engineer. The master was faced with a very difficult situation and, in co-ordination with the bosun, executed a series of quick actions that ensured the safety of the vessel, passengers and crew. The engineer was extremely fortunate that his injuries were not life threatening. The shrapnel ejected from the engine during the explosion thankfully missed him and, despite being engulfed in fire, his cotton boiler suit protected most of his body from the heat. The consequences of this accident could have been far worse. Accident mechanism All the bolts from undamaged bearings had been torqued to the recommended levels, and metallurgical tests on the failed big end bearing bolts established that there was no evidence of fatigue. The engine ran for 5½ hours before the breakdown; had the engine been incorrectly assembled it is likely that the failure would have occurred soon after it had been started. Volvo Penta’s finding, that debris in the oil channels was the likely cause of the bearing shells turning, was not supported by direct evidence, such as embedded debris in the shells. However, debris of significant hardness and size, if lodged between the main bearing and crankshaft journal, has the potential to disrupt the hydrodynamic lubrication, resulting in metal to metal contact. This would be rapidly followed by the bearing shells turning, blocking the oil supply to the big end bearing, which then resulted in its failure. Soft patches The soft patches were not used to move the engines in or out of the engine rooms due to the disruption their removal would cause. Therefore, RKM planned its work around the use of the emergency hatch. This required the engines to be partially disassembled and rebuilt in the engine room, transporting the majority of the components in the short block. 10 MAIB Report 1/2018, catastrophic engine failure, resulting in a fire on board Windcat 8 off the Lincolnshire coast.

13 The source of any debris that might have entered the oil channels cannot be known for certain. However, ME2’s short block had been exposed to the elements for 3 days with only a loose plastic sheet for protection, and debris could have entered the oil channels during this time. Rebuilding the entire engine under a clean and controlled environment and transferring it directly into the engine room through the soft patch would have significantly reduced the likelihood of debris ingress during transit and while awaiting fitting. Analysis of test results The test results of the sump oil sample revealed high levels of water and fuel contamination that were not present in the oil sample from the filter. Since the mechanically-driven lubricating oil pump would have stopped with the engine, the filter oil sample was representative of the engine oil’s condition before the explosion. The contamination of the sump oil was a result of the damage and subsequent fire suppression. The presence of a high number of particles larger than 10 microns as indicated by a high PQ index (Table 1), combined with a low count of small iron particles in the oil sample from the forward filter, suggests a sudden and rapidly progressing wear-down. The unusually large wear particles in the filters, some of which were larger than 1mm, also implies that accelerated wear had started before developing into a catastrophic failure. This wear-down probably started before the main bearing shells turned, since the total starvation of oil supply would have resulted in nearly instant seizing of the big end bearing, giving insufficient time for the wear particles to find their way to the filter. Therefore, the wear particles exhibiting heat damage were most likely to have been from main bearing 5, generated before the bearing shells turned. It is evident from the alarm logs that the lubricating pressure shutdown was activated a few seconds before the explosion. By that time, the damage had already been done. As the engine was not fitted with a particle detector or other means of detecting rapidly progressing wear, there was no possibility of receiving an early warning before the engine failed. The previous accident on Windcat 8 and other failures on sister vessels of Wight Sky could also have potentially been prevented had a particle detector system been fitted. Loss of power to essential services When the power supply for Wight Sky switched from the forward to aft switchboard, the ES switchboard lost all power because the dedicated ES circuit breaker had been left in the manual mode. As a result, the Hi-fog system, along with the watertight doors and several other pieces of critical equipment, lost electrical power. The Hi-fog system was very effective in extinguishing the fire within a minute of having been activated. However, had the changeover to the aft switchboard taken place immediately after the accident, the Hi-fog system would have provided further protection for only 60 seconds, increasing the potential for re-ignition and growth of the fire. There were several fuelled vehicles parked on the deck directly above the forward machinery space, so the consequences of the loss of electrical power to the ES switchboard could have been very serious. To avoid the loss of power to critical equipment, the importance of the ES switchboard circuit breakers needs to be recognised and suitable controls established to ensure these are not routinely left in the manual mode.

14 CONCLUSIONS •• The consequences of this accident could have been far worse; the engineer was fortunate that his injuries were not life threatening, and the quick actions of the bridge team ensured the safety of the vessel, passengers and crew. •• Volvo Penta’s investigation into the engine failure concluded that: ○ ○Debris in the engine’s oil channels was most probably responsible for initiating the failure. ○ ○When the shells of main bearing 5 turned, the lubrication supply to big end bearing 4 was blocked, resulting in destructive failure. •• Rebuilding the engine in a clean and controlled environment and transferring it complete into the engine room would have reduced the likelihood of debris ingress. •• The engine was not fitted with a wear detector, so there was no means of receiving a warning before the engine failed. •• The essential services switchboard aft circuit breaker had been left in manual mode, resulting in the loss of power to critical equipment, including the fixed fire-fighting system. ACTION TAKEN Wightlink Limited has: •• Agreed with RKM that all engines will be fully assembled and load-tested prior to delivery on board. •• Put in place signage to ensure that the circuit breakers for the essential services switchboard are correctly set at all times. Volvo Penta UK has: •• Written to all Volvo Penta dealerships in the UK and Ireland to provide guidance on good practice for repair and rebuild and to ensure that: ○ ○Volvo Penta UK is informed of any major engine failure resulting in injury, fire or flood. ○ ○Where appropriate, soft patches are removed to allow removal and reinstallation of complete engines. ○ ○Engine assembly is completed in a clean environment to prevent debris being built into an engine. ○ ○Following rebuilds, engines are load-tested on a dynamometer and certificates issued confirming the required performance. ○ ○Records of component measurements are kept to confirm that they are within tolerance and fit for reuse.

15 RECOMMENDATIONS Volvo Group (UK) Ltd is recommended to: 2018/120 Consider offering wear particle detection technology for Volvo Penta marine engines that cannot be easily serviced on board. Safety recommendations shall in no case create a presumption of blame or liability

16 SHIP PARTICULARS Vessel’s name Wight Sky Flag United Kingdom Classification society Lloyd’s Register IMO number/fishing numbers 9446984 Type Ro-ro passenger ferry Registered owner Wightlink Limited Manager(s) Wightlink Limited Year of build 2008 Construction Steel Length overall 62.40m Gross tonnage 2546 Minimum safe manning 10 Authorised cargo Passengers, private and commercial vehicles VOYAGE PARTICULARS Port of departure Lymington Port of arrival Yarmouth (Isle of Wight) Type of voyage Internal waters Cargo information 19 motor vehicles, 30 passengers Manning 10 MARINE CASUALTY INFORMATION Date and time 12 September 2017, 2133 Type of marine casualty or incident Serious Marine Casualty Location of incident 50°42’.41 N, 1°30’.03 W (300m from Yarmouth ferry terminal) Place on board Forward engine room Injuries Severe burn injuries to engineer Damage/environmental impact Main propulsion engine 2 damaged beyond repair Ship operation In service Voyage segment Arrival External & internal environment Moderate sea, light breeze Persons on board 40

FD06043 ｜ 2012_英国_catastrophic-engine-failure-and-subsequent-fire-on-board-the-site-investigation-vessel-kom.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2012 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

SERIOUS MARINE CASUALTY REPORT NO 10/2026 MAY 2026 1 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an such investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2026 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: +44 (0)23 8039 5500 ACCIDENT REPORT ACCIDENT REPORT Catastrophic engine failure and subsequent fire on board the site investigation vessel Kommandor Susan in the Firth of Forth, Scotland on 25 January 2025 SUMMARY At 1305 on 25 January 2025, the site investigation vessel Kommandor Susan suffered a catastrophic failure of a diesel generator engine while conducting sea trials in the Firth of Forth, Scotland. The failure resulted in an engine room fire and a complete power blackout. The crew responded promptly, extinguished the fire and mustered safely. The vessel lost propulsion and began drifting eastwards. Attempts to deploy anchors were unsuccessful because the anchor winches required electrical power to operate. Kommandor Susan was eventually returned to Leith harbour with limited propulsion restored. The investigation found that the engine failure was caused by premature wear of bearings fitted during a major overhaul in 2019. These components were not approved by the engine original equipment manufacturer and exhibited weaker material bonding than genuine parts. The extended service intervals applied to the engines were valid only for original equipment manufacturer components, making the use of substitute parts a critical factor in the failure. The vessel’s former owner had minimal oversight of the overhaul process and assumed that original equipment manufacturer parts were used. This lack of verification meant that the presence of substitute components remained undiscovered and was not communicated to subsequent owners. Additionally, the vessel’s anchoring procedure did not account for the risk of power loss, leaving the anchors inoperable during the emergency. The vessel’s owner, Gardline Shipping Limited, has since rebuilt the affected engine, and overhauled all other generator engines on board Kommandor Susan, with original equipment manufacturer parts. The Chief Inspector of Marine Accidents has written to the vessel’s previous owner, Hays Ships Limited, outlining the importance of a structured supervision system that provides clear accountability measures and the need to adopt real-time progress monitoring for all critical maintenance activities. Kommandor Susan Image courtesy of fabianv (ShipSpotting.com)

2 FACTUAL INFORMATION Narrative On the morning of 25 January 2025, the crew of Kommandor Susan prepared the vessel for sea trials from the Port of Leith, Scotland. At 1020, the chief engineer (C/E) started diesel generator (DG) 1 and DG4 to warm the engines. By 1142, a pilot had boarded, and the bridge team had completed their pre-departure checks. At 1220, Kommandor Susan entered the lock to exit Leith harbour, clearing the lock at 1252 before proceeding into the main channel of the Firth of Forth. At 1305, the C/E heard a change in engine tone from one of the running DGs, followed immediately by a loud bang and the activation of the fire alarm. On opening the engine control room (ECR) door, the C/E observed that the engine room was filled with smoke and oil mist. One minute later, the C/E informed the bridge that there was a fire in the engine room. In response to the incident, the second engineer (2/E) stopped both DGs from the ECR switchboard, resulting in a blackout and loss of propulsion. The emergency generator started automatically and restored power to essential systems, including emergency lighting, but not to propulsion. The master sounded the general alarm from the bridge, prompting the crew to muster stations. Once everyone was accounted for, the chief officer (C/O) prepared a fire team to enter the engine room. Equipped with a dry powder extinguisher, the fire team located and extinguished several small pockets of burning oil in the bilges on both sides of DG1 before exiting the space. With no propulsion available, Kommandor Susan began drifting eastwards. The master instructed the deck crew to drop anchor to arrest the drift, but both anchor winches were clutched into their respective electric motors and the clutches could not be released due to the lack of power. The pilot advised the master that support had been requested and, at 1416, the tug Kittiwake was made fast to Kommandor Susan’s port quarter. Shortly after, the C/E advised that DG4 could be safely restarted. Limited propulsion was restored once DG4 was operational, which allowed Kommandor Susan to manoeuvre, and anchor winch clutches could also be released at this time. Kommandor Susan berthed alongside at 1806, having returned to Leith harbour assisted by two tugs. Environmental conditions On 25 January 2025, the weather in the Firth of Forth was mostly dry with a south-westerly force 5 to force 6 wind. The visibility was good and the ebbing tide was running at around 2 knots to the east. Kommandor Susan Kommandor Susan was built in 1999 as an 83m offshore supply vessel. In 2018, the vessel was sold to Hays Ships Limited (Hays Ships). In 2022, Kommandor Susan was sold to Gardline Shipping Limited (Gardline). In 2019, significant works had been undertaken to convert the vessel to work in the specialist site investigations sector. Site investigation vessels performed offshore geotechnical and geophysical surveys.

3 Kommandor Susan’s propulsion system comprised four Caterpillar 3516B-TA diesel generators (Figure 1), each with a rated output of 1,901 kilowatts (kW). This system was used to power two 2,200kW azimuth propulsion units, an 883kW drop-down azimuth thruster and an 800kW tunnel bow thruster. Kommandor Susan’s crew were all appropriately qualified to serve on a UK registered vessel. Regular fire drills were completed on board to prepare the crew to deal with emergency situations. This included the use of portable extinguishers, fire hoses, and breathing apparatus. The emergency response teams also practised raising the alarm, isolating fuel and ventilation systems, and deploying fixed firefighting installations such as carbon dioxide or water mist systems. Communication protocols were tested to maintain coordination between the engine room, bridge, and emergency response teams. Drills also emphasised evacuation routes, muster points, and casualty handling. Diesel generators The standard service intervals for Caterpillar 3516B-TA engines were based on load, operating environment and fuel quality. For marine applications using Ultra Low Sulphur Diesel, Caterpillar’s recommended service intervals were 5,000 hours for a top-end service and 10,000 hours for a major overhaul. The top-end service included cylinder head removal and inspection. The major overhaul included the removal of components such as pistons, liners and bearings as well as the replacement of all bearings, pistons and piston rings. Connecting rods could also be replaced at this time based on their condition. The service interval could be extended with Caterpillar’s agreement following the first major overhaul. Several factors were considered in Caterpillar’s decision to approve an increase, including total fuel usage, oil analysis, historic power loading and the condition of the parts removed during the major overhaul. Caterpillar had extended the service intervals on Kommandor Susan’s 3516B-TA engines to 24,000 hours, which was not an uncommon decision for similarly loaded engines if genuine Caterpillar components were used in subsequent overhauls. Major overhaul service intervals could be extended up to 40,000 hours. At the time of the accident DG1 had operated for 65,700 hours in total, of which 13,466 running hours were since the 2019 overhaul. The previous four oil analysis reports for DG1 did not show any significant amounts of wear particles. The most recent of these was based on an oil sample taken on 6 January 2025, some 19 days and 127 running hours before the failure. All four of Kommandor Susan’s DGs underwent major overhauls during Hays Ships’ ownership: DG1 and DG3’s overhauls were completed in January 2019; DG2 and DG4’s overhauls were Looking down onto four large pieces of machinery in the vessel’s engine room. In the foreground is DG1, which is fire damaged and has a red and white hazard tape cordon at one end. Figure 1: Kommandor Susan's four diesel generators DG1

4 completed in November 2020. All four overhauls were carried out by Holderness Ship Repairers Limited (Holderness)1, which sourced many of the major components required for the engine overhauls from G&T Engineering Limited2. Hays Ships’ technical management was made aware that some of the component parts for the engine overhauls would be sourced from the USA, but Holderness did not provide specific supplier details or certification. During the 2019 major overhaul of the engines, Hays Ships’ technical management simultaneously oversaw Kommandor Susan’s conversion with several major engineering projects demanding their attention. Post-accident inspection of DG1 It was observed, post-failure, that DG1’s cylinder 5 and cylinder 6 pistons had fractured and released their connecting rods from the piston gudgeon pins and the big ends had detached from the crankshaft. The loose connecting rods had punched through their respective sides of the crankcase (Figure 2). The cylinder 5 and cylinder 6 pistons remained in their cylinder liners. The remains of the small end and big end bearings were found in the crankcase; all were found in a high state of degradation and were heat affected. Other detached parts included crankshaft counterweights and lower sections of the pistons, and both gudgeon pins were found in the crankcase. 1 Holderness Ship Repairers Limited ceased trading in November 2024. 2 G&T Engineering Limited ceased trading in February 2023. Figure 2: Damage to DG1 View of scorched and fire-damaged machinery, highlighting DG1’s crankcase on the right-hand side. Looking towards a hole DG1’s crankcase, exposing the fire-damaged internal machinery.Several detached component parts. Damaged parts DG1 crankcase Images courtesy of MAIB and Kommandor Susan 2/E

5 The failed DG1 was removed from Kommandor Susan and transported to a Caterpillar authorised agent, Pon Power, in Papendrecht, the Netherlands. Pon Power technicians stripped the engine into its component parts for detailed inspection. The technicians found that many of the major components of the rotating assembly were not original equipment manufacturer (OEM) parts. The non-OEM parts identified in DG1 were manufactured by Costex Tractor Parts (CTP). The USA-based CTP stated on its website that it provided New Replacement Parts for Caterpillar. CTP was not listed as an authorised dealer on Caterpillar’s official dealer locator. The DG1 components identified as CTP parts included pistons, connecting rods, small end and big end bearings, and main bearings. The Pon Power technicians compared a CTP big end bearing found in DG1 and a genuine OEM big end bearing and identified that the white metal surface of the OEM big end bearing had an aluminium backing layer, while the white metal surface of the CTP big end bearing had a copper backing layer (Figure 3). Pon Power produced a report on its findings that stated the most likely order of events to be: • Run out of connecting rod bearing five • Bearing hammered thin • Connecting rod cap number five breaks at a normal stress raiser • Connecting rod number five damages counterweight five and bottom of liner five and six • Counterweight five comes loose and falls into the oil pan • Piston number 6 gets stuck due to the damaged liner • Connecting rod bolts break by overload and connecting rod cap number six comes off. Pon Power’s report concluded that: With the facts gathered the most likely root cause is failure of connecting rod bearing number five. The excessive wear on the other connecting rod bearings show that the most likely root cause for the bearing failure is weak bonding of the copper layer. Evidence of weak material bonding was found in several bearings (Figure 4). Figure 3: Comparison between genuine OEM (left) and CTP bottom end bearings (right) Image courtesy of Pon PowerTwo curved pieces of metal placed side by side. There is a copper tint to the edge of the bearing on the right-hand side.

6 The report made two recommendations: • Make use of original Caterpillar parts as service interval schedules are only applicable for these parts. Non-Caterpillar parts used in the engine tent to have a shorter service life. [sic] • Check the use of non-Caterpillar bearings in the other engines on the vessel. If so, plan a early change of the main and connecting rod bearings. [sic] Parts approval process The approval process for a ship’s engine parts involved several steps to ensure compliance with safety and performance standards. Classification societies required manufacturers to submit detailed documentation, including their manufacturing process descriptions, quality control measures, and test results. Typically, there were four steps:

1. Application submission – manufacturers to provide technical specifications, factory layout

drawings, and quality control procedures.

1. Testing and inspection – classification societies to conduct inspections and tests to verify

compliance with industry standards.

1. Certification and approval – issued if the parts met the required standards.
2. Regular renewal and compliance checks – to ensure continued regulatory compliance.

The investigation found no evidence that CTP manufactured parts had been class approved by any marine classification societies. Three separate bearings, each overlaid with an illustrated circular outline highlighting exposed copper beneath its white metal surface. Image courtesy of Pon Power Figure 4: Evidence of weak material bonding on several bottom end bearings​

7 REGULATION AND GUIDANCE Emissions of nitrogen oxides (NOx) from marine diesel engines were regulated under the International Convention for the Prevention of Pollution from Ships (MARPOL Convention)3 Annex VI4, Regulation 13, which applied to all engines over 130kW. Compliance was achieved through certification under the NOx Technical Code5 2008 and required an Engine International Air Pollution Prevention (EIAPP) certificate and ongoing in-service verification. This verification was usually in the form of a NOx file that recorded all NOx-emitting components and documented post-maintenance and post-overhaul updates. The NOx limits were defined within a set of tiered emission standards based on the ship’s construction date and engine speed, with the earliest compliance applicable to vessels constructed on or after 1 January 2000. Kommandor Susan was built in 1999, so was not required to comply with the MARPOL Convention Annex VI, Regulation 13 on NOx limits. This meant that no NOx file existed, which would have tracked compliant components and spare parts. Previous accidents At 2003 on 16 April 2018, the Lithuania registered roll-on/roll-off (ro-ro) cargo vessel Finlandia Seaways suffered a catastrophic main engine failure that caused serious structural damage to the engine and a fire in the engine room (MAIB report 2/20216). The vessel’s third engineer suffered serious smoke-related lung, kidney and eye injuries during their escape. The investigation identified that non-OEM maintenance techniques introduced latent defects, which led to engine failure. At 2013 on 19 September 2021, a fire broke out in the auxiliary engine room on the Finland registered ro-ro cargo ship Finnmaster during departure from Hull, England (MAIB report 13/20257). Finnmaster lost power, but the fire was later extinguished and the vessel safely re-berthed with the assistance of tugs. The auxiliary engine room was significantly damaged, but there were no injuries. The flexible hose installed in the fuel system during a modification did not meet the required standard, was fitted in an inappropriate position, and had not been subject to approval or oversight by the responsible classification society. ANALYSIS Overview The catastrophic engine failure fire on board Kommandor Susan occurred when DG1 suffered a mechanical breakdown, leading to a fire in the engine room and a complete blackout. Emergency response procedures were followed, all crew were accounted for, and the fire was successfully extinguished. However, the vessel lost propulsion and drifted until limited propulsion was restored and tug assistance arrived. Kommandor Susan eventually returned to Leith harbour. 3 The MARPOL Convention was adopted on 2 November 1973 at the International Maritime Organization. 4 Prevention of Air Pollution from Ships (entered into force 19 May 2005). 5 Technical Code on Control of Emission of Nitrogen Oxides from Marine Diesel Engines. 6 MAIB report 2/2021: Finlandia Seaways 7 MAIB report 13/2025: Finnmaster

8 Engineering oversight Hays Ships had overseen major engine overhauls during its ownership of Kommandor Susan. The company expected that Holderness would use genuine Caterpillar parts for these overhauls based on a long-standing relationship and mutual trust. Coupled with competing priorities during the vessel’s conversion project, this resulted in Hays Ships providing minimal oversight to Holderness and no verification of authenticity for the parts used, enabling the installation of non- OEM components that later contributed to the engine failure. Similar to the lack of oversight highlighted in both the Finlandia Seaways and Finnmaster cases, the lack of structured supervision and verification during Kommandor Susan’s main engine overhaul process meant that Hays Ships was unaware that non-OEM parts had been used. This oversight created a latent risk that was not communicated to Gardline when it later purchased the vessel. Since no NOx file was required for Kommandor Susan and Hays Ships was not directly supervising the work, there was no control of the suitability of parts used in the vessel’s engine overhauls. The failure to implement robust contractor assurance measures allowed critical safety vulnerabilities to persist undetected, ultimately leading to DG1’s catastrophic failure. Bearing failure Post-failure inspection revealed that non-OEM bearings manufactured by CTP were fitted during the overhaul. These bearings had a copper backing layer instead of the aluminium layer found in genuine Caterpillar bearings; weaker bonding of these layers reduced durability. The extended service interval agreed by Caterpillar was approved on the basis that genuine parts were used. The CTP parts were designed for a standard service interval. Use of the CTP parts beyond their design life led the connecting rod bearings to fail, initiating a chain of mechanical failures that culminated in the engine failure and subsequent fire. The extended service intervals were inappropriate for these substitute components, accelerating the risk of failure. Anchoring The crew attempted to deploy the anchors to arrest the vessel’s drift when propulsion was lost. However, both anchors were clutched into their electric motors and could not be released without power. Industry good practice is to have anchors cleared away and ready to drop when navigating in pilotage conditions, and any procedural oversight posed a significant safety hazard during emergencies involving total power loss. The inability to anchor increased Kommandor Susan’s vulnerability to environmental conditions, grounding and collision risks until tug assistance arrived. CONCLUSIONS • The catastrophic failure of Kommandor Susan’s DG1 engine and resultant engine room fire was caused by the premature wear of non-OEM components, fitted without Hays Ships’ knowledge during the engine’s last major overhaul. • The agreed extended service intervals were inappropriate when non-OEM components were in use, accelerating the risk of failure. • Hays Ships exercised limited oversight of the contractors performing the engine overhauls. The company was therefore unaware that non-OEM parts had been used, and unable to adequately determine the risk of failure or inform the vessel’s new owner that non-OEM parts were fitted in the engines.

9 • The failure to implement robust contractor assurance measures allowed critical safety vulnerabilities to persist undetected, ultimately leading to the catastrophic engine failure. • The inability to anchor increased Kommandor Susan’s vulnerability to environmental conditions, grounding and collision risks until tug assistance arrived. Kommandor Susan’s onboard procedure for emergency anchoring did not identify the risk of having the winch drum permanently clutched in. ACTIONS TAKEN MAIB actions The Chief Inspector of Marine Accidents has written to Hays Ships Limited, outlining the importance of robust oversight during critical maintenance activities and the need to adopt a system of structured supervision, ensuring clear accountability and real-time progress monitoring. Actions taken by other organisations Gardline Shipping Limited has: • Rebuilt DG1 with genuine spares, including a new engine block; • Completed major overhauls of DG2, DG3 and DG4 using genuine Caterpillar parts; • Changed the onboard anchoring procedure, highlighting that clutches must remain disengaged when the winches are not in use to ensure they remain ready for use in the event of power failure. RECOMMENDATIONS In view of the actions already taken, no recommendations have been made.

10 SHIP PARTICULARS Vessel’s name Kommandor Susan Flag UK Classification society Det Norske Veritas IMO number/fishing numbers 9177844 Type Site investigation vessel Registered owner Gardline Shipping Limited Manager(s) Gardline Shipping Limited Year of build 1999 Construction Steel Length overall 83.7m Registered length 75.36m Gross tonnage 3,388 Minimum safe manning 10 Authorised cargo Not applicable VOYAGE PARTICULARS Port of departure Leith, Scotland Port of arrival Leith, Scotland Type of voyage Sea trial Cargo information Not applicable Manning 14 MARINE CASUALTY INFORMATION Date and time 25 January 2025 at 1305 Type of marine casualty or incident Serious Marine Casualty Location of incident The Firth of Forth, Scotland Place on board Engine room Injuries/fatalities None Damage/environmental impact Substantial damage to diesel generator Ship operation Sea trials Voyage segment Departure External & internal environment Wind south-westerly force 5 to force 6; mostly dry; good visibility Persons on board 14

FD06044 ｜ 2012_英国_catastrophic-engine-failure-resulting-in-a-fire-on-crew-transfer-vessel-windcat-8__MAIBInv.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2012 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H SERIOUS MARINE CASUALTY REPORT NO 1/2018 FEBRUARY 2018 1 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an such investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2018 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gsi.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 ACCIDENT REPORT Catastrophic engine failure, resulting in a fire on board Windcat 8 off the Lincolnshire coast 7 September 2017 SUMMARY On 7 September 2017, the 15.87m crew transfer vessel, Windcat 8, was on passage to Grimsby, UK, from the Lynn Wind Farm in the North Sea with two crew and eight windfarm technicians on board. Shortly after setting off, the vessel’s port engine suffered catastrophic damage and caught fire. The passengers were quickly transferred on to Windcat 31 and the fire was contained within the port engine space and soon extinguished. Windcat 8’s port engine was badly damaged and the vessel was towed to Grimsby by Windcat 30. There was no pollution and no injuries. The MAIB investigation identified that the catastrophic damage to Windcat 8’s port main engine was caused by the failure of a piston connecting rod big end shell bearing, which resulted in the connecting rod assembly releasing and penetrating through the engine crankcase. The investigation also identified that: •• A high oil temperature alarm prior to the engine failure warranted a more cautious approach to the use of the engine until a deeper technical investigation could be conducted. •• The fire resulted from the ignition of oil vapour released from the damaged crankcase and did not spread beyond the port engine. •• The fixed fire extinguishing system was rendered ineffective because not all the port engine space vents were closed prior to its release. •• The safety of the passengers was given a high priority. A recommendation has been made to Windcat 8’s owner, Windcat Workboats BV Ltd, which is aimed at improving its crews’ reactions to engine alarms and engine space fires. Windcat 8 Image courtesy of Windcat Workboats BV Ltd

2 NARRATIVE At approximately 16201 on 7 September 2017, the crew transfer vessel2 Windcat 8, owned and operated by Windcat Workboats BV Ltd (Windcat Workboats), departed wind turbine LN09 on the Lynn Wind Farm (Figure 1) to return eight technicians to Grimsby, UK. The vessel’s crew comprised a master and a deckhand. It was daylight with a gentle breeze and slight seas. As Windcat 8 cleared the windfarm, the master increased the vessel’s speed to 21 knots (approximately 1840rpm). At 1638, a high engine oil temperature alarm activated on the port engine indication panel on the forward console in the main cabin (Figure 2). The master immediately set both engines to idle and told the deckhand to go to the port engine space to investigate. The deckhand checked the port engine’s cooling system and oil level but he did not identify anything unusual. He returned to the cabin, closing the port engine space deck hatch behind him (Figure 3), and reported his findings to the master. As the high oil temperature alarm on the indication panel had now reset, the master decided to resume passage. At 1640, he increased the vessel’s speed back to 21 knots and monitored the port engine temperature, which remained normal. 1 All times are UTC(+1) 2 Crew transfer vessels are used to provide support services to offshore windfarms Figure 1: Extract of Admiralty Chart 1190-0 showing accident location Reproduced from Admiralty Chart 1190-0 by permission of HMSO and the UK Hydrographic Office Grimsby Windcat 8's track 1620 1646 (Fire) Lynn Wind Farm Inner Dowsing Wind Farm

3 Figure 3: Aft deck Exhaust vent Supply vent Natural vent Port engine space deck hatch Figure 2: Deck plan and lay out Fire extinguishing system activation panel Engine space access Forward console Natural vent Fuel shut off

4 At 1646, there was a loud bang in the port engine space, the port engine stopped, and the port engine space fire alarm activated. The master saw from the closed-circuit television (CCTV) monitor that the port engine was alight (Figure 4a)3. In response, he stopped the starboard engine, shouted that there was a fire, instructed the deckhand to close down the port engine space and activate its fire extinguishing system. He also stopped the ventilation fans servicing the port engine space and told the eight technicians to prepare for evacuation. The deckhand went onto the aft deck and closed the flap on the port engine space supply vent. He then activated the port engine space’s fire extinguishing system using controls located on the cabin’s aft bulkhead, behind the last row of seats (Figure 2). By 1648, the port engine space was filled with the extinguishing medium, which obscured the fire on the CCTV monitor (Figure 4b) and a white gaseous cloud billowed from the port engine space’s natural air vent on the aft deck (Figure 4c) and into the cabin through its open door. The technicians left the cabin and mustered on the foredeck. At 1649, the master broadcast a “Mayday” using the fixed very high frequency radio on channel 16. Windcat 30 and Windcat 31, also owned and operated by Windcat Workboats, were nearby and closed to assist. At approximately 1650, the deckhand supervised the transfer of the technicians from Windcat 8 to Windcat 31. He then returned to the aft deck, where plumes of dark smoke were now coming from the port engine space’s natural vent and entering the cabin through its open door. He leaned through the smoke and operated the remote fuel shut- off for the port engine sited next to the vent (Figure 4d). The master then sealed the natural and exhaust vents. 3 The times shown on the CCTV recordings were 7 minutes slow Figure 4a: Initial fire Figure 4b: Port engine space immediately following the operation of the fixed fire extinguishing system Figure 4c: Fire extinguishing powder billowing from the natural air vent Figure 4d: Deck hand shutting off fuel supply

5 By 1653, the smoke had cleared from the port engine space and flames were observed burning along the top of the port engine (Figure 5a). Four minutes later the port engine was again obscured by smoke. Meanwhile, the master and the deckhand rigged a fire hose on the aft deck in readiness for boundary cooling. The remote fuel shut-off for the starboard engine was also operated. At 1700, the fire appeared to be out and Windcat 8 was taken in tow by Windcat 30 for passage to Grimsby. The master referred to the vessel’s emergency checklists and later re-started the starboard engine to expedite the transit. The vessels arrived at Grimsby Fish Dock at 2100 and were met by local fire and rescue services personnel, who confirmed that the fire in Windcat’s port engine space had been extinguished. Post-incident examination Examinations of Windcat 8’s port engine by Windcat Workboats’ engineers and MAIB inspectors identified that the lower part of the piston, its connecting rod and the piston gudgeon pin from its No.6 cylinder had broken through the crankcase. As a result, the engine-mounted fuel pump had sheared off and the fuel pump casing was fractured. Plastic shielding on the top of the engine had been completely consumed by the fire and the plastic coolant water header tank mounted on the top of the engine, along with conduit and wiring on the side of the engine, had also burned (Figure 5b). Secondary heat damage was limited to a fire detection head and a warning light immediately above the engine. Windcat 8’s port engine, a Volvo Penta D16C-A MH, was removed from the vessel and sent to Haisma Scheeps & Industriemotoren BV (Haisma Scheeps) in the Netherlands for thorough examination. The examination was completed on 21 September, and identified that the big end shell bearing on the No.6 piston connecting rod had been destroyed by heat (Figures 6 and 7). The level of the damage prevented analysis of the bearing to determine the precise cause of its failure. Figure 5a: Fire on engine top Figure 5b: Fire damage to port engine Figure 6: Heat damage to No.6 piston connecting rod Inset: Remains of shell bearing

6 The thorough examination also identified that the camshaft, main bearings and the other connecting rod bearings were in pristine condition and that none of the pistons had seized. In addition, the lubricating oil system appeared to have functioned as expected. Components damaged by the heat generated by the failure of No.6 connecting rod big end bearing included the No.6 crankpin journal (plastic deformation) (Figure 8) and No.6 and No.7 main bearings. Physical damage included: •• The failure of the No.6 connecting rod bearing cap bolts. •• The separation of No.6 piston crown and skirt. The piston crown had struck the open exhaust valve, shattering the ceramic exhaust valve cam follower and damaging the running surface of the exhaust cam lobe. •• A hole in the engine crankcase in way of No.6 piston (Figure 9). Operation Windcat 8, a 15.87m aluminium catamaran, was one of 39 crew transfer vessels operated by Windcat Workboats, mainly in the European offshore wind power generation industry. The vessel was certified under the Brown Code ‘A code of practice for the safety of small workboats and pilot boats under 24 metres’ and was designed and built specifically for transporting up to 12 technicians and their equipment. The catamaran’s square bow enabled the vessel to ‘push on’ to the wind turbines and create a stable platform for the technicians to safely disembark and embark. The vessel typically operated for 12 hours each day, of which 3 hours were spent on passage to and from the wind turbines and the remaining 9 hours were spent on site. It was usual for the vessel to remain ‘pushing on’ to the wind turbine on which the technicians were working. Figure 7: Schematic of crankshaft assembly Piston Gudgeon pin Shell bearing Big end bearing Connecting rod No.6 crankpin journal

7 Propulsion Windcat 8’s propulsion system consisted of twin shaft fixed pitch propellers powered by two Volvo Penta engines via reversing gearboxes. The engines were rated at 551kW at 1900rpm and were sited in separate spaces (port and starboard). Access to the spaces was via hatches on the aft deck (Figures 2 and 3). Remote fuel shut-off mechanisms were sited on the aft deck and were painted red. The port engine was installed in Windcat 8 in July 2015 as part of an engine service exchange contract with Haisma Scheeps, a Volvo Penta accredited service agent. Before installation, the engine had completed 10844 running hours and had been completely stripped down and rebuilt at Haisma Scheeps’ workshops using new components supplied by Volvo Penta. The engine’s oil capacity was 55 litres and its normal operating temperature was 90°C. The high oil temperature alarm was set to activate at 125°C with an intended delay of 30 seconds. Other engine alarms included high exhaust and coolant temperatures and low oil pressure. The engine was not fitted with a protection device to automatically shut it down in the event of defined maximum operating parameters, which could include high oil temperature, being exceeded. From its installation on board Windcat 8, the engine completed 2275 running hours during which the engine was maintained in accordance with Windcat Workboats’ planned maintenance programme, and there had been no reported problems with its operation. In addition, neither the results of the analyses of regular oil sampling, nor the weekly records of the engine’s pressures and temperatures taken when the vessel was either at ‘full speed’ or ‘pushing on’ indicated the presence of any defects or abnormalities. However, the records showed that when ‘pushing on’ at 1300rpm (50% engine speed), the turbocharger exhaust temperatures were higher than when the vessel was on passage and the engines were at their maximum speed of 1900rpm. The engine’s idle speed was approximately 800rpm. Figure 8: Plastic deformation of crankpin Figure 9: Hole in engine crankcase Exit path of connecting rod assembly

8 Engine space ventilation Windcat 8’s engine spaces each had a forced air supply vent, a forced air exhaust vent and a natural air vent. The fans for the supply and exhaust vents were operated by switches on the cabin’s forward console. The supply vent was closed by the movement of a lever that operated an internal flap, and the exhaust and natural vents were closed by sliding plates over their external grills. The operating lever for the supply vent, and the sliding plates, were painted red. The sliding plates were also marked ‘Fire Flap’ (Figure 2). The integrity of the ventilation flaps and sliding plates was tested annually. It was also checked during fire drills. Fixed fire extinguishing system In common with all Windcat Workboats’ vessels, the engine spaces on Windcat 8 were protected by a manually-operated Stat-X fixed fire extinguishing system, supplied and fitted by Marine Fire Safety Limited. The system incorporated a combination of smoke and heat detectors linked to control boxes on the main cabin’s aft bulkhead that also housed the activation mechanism and an audible alarm. The system used a fine particulate agent, potassium carbonate, in aerosol form to extinguish the fire by acting on the chain reaction of the combustion process. To fully suppress flames, the potassium carbonate particles had to be maintained at sufficient density. The system’s operating manual recommended that the engine space doors and hatches remained closed for at least 10 minutes after system activation. Instructions posted adjacent to the Stat-X system’s control boxes advised that the affected space should not be re-entered for at least 15 minutes. Emergency procedures There was no guidance or instructions on board Windcat 8 regarding action to be taken on the activation of engine alarms. Emergency checklists kept on the vessel’s forward cabin console included actions to be taken in the event of abandonment, collision, grounding, man overboard and fire. Regarding an engine space fire, the checklist’s actions included: Shutdown of engines, ventilation Pull fuel shut offs and fit fire flaps and close louvres Activate the engine room smothering system if appropriate Instructions posted adjacent to the Stat-X fire suppression control units stated: In the event of the engine room alarm sounding Switch off engines and machinery Evacuate all personnel from engine room Close all engine room ventilation Carry out fire procedures Crew Windcat 8’s master held an STCW4 accredited Maritime and Coastguard Agency Master’s certificate for coded vessels less than 200 gross tonnes. He had worked for Windcat Workboats for 2 years and had successfully completed a seagoing performance assessment. He had also completed familiarisation training on board Windcat 8 on 5 September 2017 before taking over as master the following day. The deckhand held an STCW navigational watch rating certificate. He had worked for Windcat Workboats for 8 months and had successfully completed a performance assessment at sea. He had also completed familiarisation training on board Windcat 8 when he joined the vessel on 24 May 2017. 4 STCW - International Convention on Standards of Training, Certification and Watchkeeping for Seafarers 1978, as amended.

9 Drills Windcat Workboats required the crews on board its vessels to follow a drills schedule that detailed the types of drill to be conducted and their frequency. It also prescribed the locations of the fire drills with engine room fire drills being scheduled for April and September. Windcat Workboats required the drills to be recorded in each vessel’s deck logbooks on completion. The company also required crews to forward photographs taken during the drills to its UK office. During 2017, the master had completed 16 emergency drills, including fire drills, on board other Windcat vessels and a manoverboard drill on board Windcat 8 during the afternoon of 7 September. The deckhand had completed 23 emergency drills in 2017 including five fire drills. The fire drills did not include any engine space fire drills on board Windcat 8. ANALYSIS Failure mechanism The initial inspection of Windcat 8’s port engine in Grimsby and the later thorough examination of the engine in the Netherlands, showed that the locus of the catastrophic failure was the big end bearing associated with the No.6 piston connecting rod. The condition of the bearing apparently quickly deteriorated soon after Windcat 8 departed from the windfarm. Such a deterioration would have restricted or stopped the flow of oil that was necessary to lubricate the bearing’s moving surfaces. At an engine speed of approximately 1840rpm, the resulting friction between the bearing and the crankpin journal would have generated high temperatures. Initially, the heat would have been conducted through the surrounding oil, but it is evident from the activation of the oil high temperature alarm that the oil temperature eventually exceeded the capacity of the oil cooler and reached 125°C. Following the oil high temperature alarm, the master reduced the port engine’s speed to approximately 800rpm. As a result, the heat generated from the friction between the bearing and the crankpin journal would have reduced. That the high temperature alarm automatically reset, indicates that the oil cooler again lowered the oil temperature to below 125°C. However, as soon as the engine’s speed was again increased to approximately 1840rpm, the big end bearing inevitably would have deteriorated further, heating the surfaces of the crankpin journal and the connecting rod bearing sufficiently to cause plastic deformation of the metal surfaces (Figure 8). Such heat would have also reduced the tensile strength of the bolts holding the connecting rod bearing cap in place. As a result, the bolts failed and the No.6 connecting rod assembly was released at high speed. That the oil high temperature alarm did not activate for a second time was possibly due to the time it would have taken the engine’s 55 litres of oil to again reach 125°C, together with the 30-second delay set. Engine loading It was not possible to pinpoint the cause of the bearing’s failure due to the extensive heat and mechanical damage sustained (Figure 6). However, the engine had been rebuilt in 2015 by an accredited Volvo Penta service agent using new components supplied by Volvo Penta. It had also run for 2270 hours over a period of 26 months, during which its performance had been routinely monitored and found to be within expected parameters. In addition, the thorough examination in the Netherlands did not find other defective parts or defects in the cooling and lubrication systems. Therefore, it is highly unlikely that the failure was the result of incorrect component assembly or a lack of oil. It is more likely that the failure was due to a weakness in the thin-walled shell bearing itself.

10 The likelihood of a relative weakness in the big end bearing leading to its failure was possibly increased by Windcat 8 ‘pushing on’ to wind turbine towers for several hours each day. When ‘pushing on’ the recorded turbocharger exhaust temperatures indicate that the loading on the port engine was higher at 1300rpm than it was at 1900rpm with the vessel on passage. The higher loading when ‘pushing on’ would have resulted in more fuel being injected and an increase in the force acting on the pistons, bearings and crankpin journals. In such circumstances, any weakness or defect in the bearing’s assembly would have been more vulnerable to failure than usual. Engine alarms Automatic engine shutdown arrangements are a technical means of preventing engine damage, but they are not required to be fitted to vessels certified under the Brown Code. However, no guidance or instructions were provided on board Windcat 8 that specified the actions to be taken in response to the differing engine alarms. Therefore, the vessel’s master had to rely on his experience and knowledge to decide on appropriate responses. On this occasion, when the port engine high oil temperature alarm activated, the master immediately reduced the engine’s speed to idle and instructed the deckhand to investigate. As the engine’s cooling and lubrication systems appeared to be working normally and the alarm had reset, the master then resumed passage at 21 knots. At the time, this action was logical and well-intended. However, in hindsight, that the alarm remained unexplained and the engine was not protected by an automatic shutdown arrangement, a more cautious approach in the use of the engine would have been appropriate before it was again put under load. Emergency response The transfer of the eight windfarm technicians from Windcat 8 to Windcat 31 within 5 minutes of the fire in the port engine space being detected indicated that the master gave high priority to their safety. However, the closing down and isolation of the port engine space lacked co-ordination. Although the ventilation fans were stopped, all the ventilation flaps were closed, the Stat-X system was released and the fuel shut-offs were activated, the sequence in which the actions were taken rendered the release of the Stat-X fixed fire extinguishing system ineffective. When the No.6 piston connecting rod assembly penetrated the engine crankcase, oil vapour and intense heat would have released simultaneously, and it is evident that the oil vapour ignited (Figure 4a). Within seconds, the deckhand had activated the Stat-X system in the port engine space and its effect was immediate (Figure 4b). However, it was also short-lived because much of the potassium carbonate quickly escaped through the space’s natural and exhaust vents that remained open (Figure 4c). Consequently, the density of the potassium carbonate was not maintained for sufficient time for it to act on the combustion process, and the fire continued to burn on the engine’s side and top, consuming residual fuel from the fractured oil pump, wiring looms, plastic conduit and shielding, and the plastic coolant header tank (Figures 5a and 5b). The fire extinguished at about 1700, largely because there was little else in its vicinity to burn and the air supply into the port engine space had been restricted by the eventual closure of all the ventilation flaps. Although all the vents were closed within about 2 minutes, this action was almost certainly prompted by the smoke being emitted and entering the cabin, rather than a recognition of the need to maintain the density of the Stat-X aerosol within the engine space. Windcat 8’s master and deckhand had worked on board Windcat Workboat vessels fitted with the Stat-X system for 2 years and 8 months respectively and both had participated in fire drills. In addition, the release procedure for the Stat-X system was straightforward and was posted adjacent to the system’s control boxes and included in the onboard checklists. The vent flaps were also clearly marked. Therefore, notwithstanding the increased stress levels experienced by the master and the deckhand during the emergency, the sequence of their actions indicates that a greater focus on the Stat-X system release procedure is warranted during onboard training.

11 CONCLUSIONS •• The damage to the port engine was caused by the failure of a piston connecting rod big end shell bearing, which led to the connecting rod assembly releasing and penetrating through the engine crankcase. •• A high oil temperature alarm prior to the engine failure warranted a more cautious approach to the use of the engine until a deeper technical investigation could be conducted. •• The fire resulted from the ignition of oil vapour released from the damaged crankcase and did not spread beyond the port engine. •• The fixed fire-fighting system was rendered ineffective because not all the port engine space vents were closed prior to its release. •• The safety of the embarked technicians was given a high priority. ACTION TAKEN Windcat Workboats BV Ltd has: •• Instructed crews on board vessels fitted with Volvo Penta D16 engines to stop the affected engine when a high oil temperature alarm activates, and not to restart the engine until the reason for the alarm has been determined. •• Identified lessons to be learned from the accident and emergency response that will be highlighted in future crew training. •• Programmed additional fire drills on board its vessels. •• Ensured that signage on board its vessels reminds crew to close vent systems before activating engine space fire suppression systems. RECOMMENDATIONS Windcat Workboats BV Ltd is recommended to: 2018/101 Provide training and guidance on the action to be taken for all critical propulsion alarms, and to introduce measures to ensure its crews understand the principles of the installed fixed fire-fighting systems and are fully familiar with procedures for their operation. Safety recommendations shall in no case create a presumption of blame or liability

12 SHIP PARTICULARS Vessel’s name Windcat 8 Flag United Kingdom Classification society Not applicable IMO number/fishing numbers 913358 Type Crew transfer vessel Registered owner Windcat Workboats BV Ltd Manager(s) Windcat Workboats BV Ltd Year of build 2007 Construction Aluminium Length overall 15.87m Registered length 15.87m Gross tonnage 26.20 Minimum safe manning 2 Authorised cargo 12 passengers VOYAGE PARTICULARS Port of departure Grimsby Port of arrival Grimsby Type of voyage Coastal Cargo information 8 passengers Manning 2 MARINE CASUALTY INFORMATION Date and time 7 September 2017 16:39 Type of marine casualty or incident Serious Marine Casualty Location of incident 53° 16.5N 000° 23.1E 4.5nm north-north-west of the Inner Dowsing Wind Farm Place on board Port engine space Injuries/fatalities None Damage/environmental impact Significant damage to port engine Ship operation On passage Voyage segment Mid-water External & internal environment Daylight; Wind force 3; Sea state 3 Persons on board 10

FD06045 ｜ 2012_英国_fire-on-board-fishing-trawler-ardent-ii-resulting-in-extensive-damage__MAIBInvReport12_201.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2012 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H VERY SERIOUS MARINE CASUALTY REPORT NO 12/2017 JUNE 2017 1 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an such investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2017 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gsi.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 ACCIDENT REPORT Fire on board fishing vessel Ardent II (INS 127) while alongside in Port Henry Basin, Peterhead 16 August 2016 SUMMARY At approximately 0500 on 16 August 20161, while Ardent II was alongside in Peterhead, an electrical fire started in the crew mess room. A short while later one of the crew got out of bed to go to the toilet, smelt and saw smoke, and raised the alarm. All three crew who were on board managed to escape onto the quay before the fire took hold. The fire service tackled the blaze, finally extinguishing it the following day. The vessel was extensively damaged and was later declared a constructive total loss. The MAIB investigation identified that: •• An electrical fault or failure of a multi-socket adapter in the crew mess room was the most likely cause of the fire. •• There was no smoke detector fitted in the crew mess room and it was fortunate that one of the crew awoke and discovered the fire at an early stage. This alerted the crew to the fire and enabled them to evacuate the vessel without delay. •• The fact the quay offered the crew a safe haven is likely to have influenced their decision to evacuate the vessel immediately, rather than to attempt to contain and extinguish the fire. 1 All times are UTC+1. Ardent II Courtesy of www.shipspotting.com

2 The MAIB has issued a flyer to the fishing industry highlighting the safety issues identified in this report, and endorsing the practice of regular visual inspection and Portable Appliance Testing (PAT) of electrical equipment as an effective means for reducing the risk of electrical fires. More explicit instruction on which spaces on board similar fishing vessels require a means of fire detection is intended to become mandatory in 2017. No recommendations are made in this report. FACTUAL INFORMATION Narrative On 11 August 2016, the trawler Ardent II returned to Peterhead from a fishing trip. The catch was landed, shore power was connected, and the vessel’s machinery was shut down. Four crew went home, leaving three crew, who were Filipino and lived on board. On 15 August, several contractors worked on the vessel conducting weld repairs and servicing the vessel’s fire-fighting appliances. The vessel’s engineer arrived during the morning to oversee the contractors, conduct routine checks, change the compressor’s oil and complete a deep clean of the engine room. The work was in preparation for a pre-guardship duty2 inspection which was scheduled to be carried out later in the week. The contractors completed their work and stood a fire watch over the weld repairs for a further 30 minutes before departing. At approximately 1800, the three crew cooked a meal using an electric rice cooker in the crew mess room and a microwave oven in the galley. During the evening, the engineer briefly ran the port auxiliary engine to enable him to pump out the fish room bilge. At 2200, two crew went to bed. The third crew man went to bed after 2345 having briefly spoken to the engineer, who told him that he would lock up the vessel when he left. At 0230 on 16 August, the engineer finished his work and departed the vessel via the wheelhouse door, noting nothing unusual as he passed through the crew mess room. He locked the wheelhouse door with a padlock as he left. At about 0515, one of the crew exited the accommodation and entered the crew mess room on his way to the toilet/washroom. As he did so, he smelt what he considered to be burning plastic and saw black smoke, but there were no obvious signs of flames or heat. He shouted to alert the other two crew, who then exited the accommodation into the crew mess room, unfastened the watertight door, and passed through the doorway into the aft net drum space and then onto the quay. At the same time, the crewman who had raised the alarm entered the wheelhouse, opened the wheelhouse door window, unlocked the padlock using a key from his pocket, opened the door, and passed onto the aft upper deck and then onto the quay. The three crew then alerted the crew of a nearby fishing boat and they called the fire service at 0537. Ten minutes after the crew had escaped from Ardent II, flames were seen emitting from the watertight doorway between the crew mess room and the aft net drum space. At 0546, the first fire appliance arrived on scene and another two arrived within the next 20 minutes. At 0611, the fire incident commander reported that the vessel was well alight and was moored by the bow only, as the stern lines had burnt through, and he requested a further three fire appliances be sent to the scene. The fire crews tackled the blaze from the quay as they were unable to board the vessel at that time. By 0620, with the vessel’s crew accounted for, arrangements were made to push the vessel back alongside. 2 Guardship duty is when a vessel is contracted to guard an offshore installation such as a pipeline or oil rig.

3 By 1209, the fire on Ardent II was believed to be extinguished. However, flames were later seen in the accommodation and further fire-fighting was necessary (Figure 1). The fire service continued damping down until the following afternoon, at which time there were no further hot spots or signs of smoke. Although still afloat, Ardent II was extensively damaged by the fire and was later declared a constructive total loss. Crew Ardent II’s skipper held a Class 2 Fishing Vessel Certificate of Competency, which he had obtained 18 years prior to the accident. All the crew had either completed the necessary sea survival, first-aid, fire-fighting and safety awareness courses in the UK or, in the case of the Filipino crew, possessed recognised equivalent qualifications. All three Filipino crew had completed at least two contracts on UK fishing vessels, the contracts varying between 8 and 10 months in length. They had also spent a minimum of 6 months on Ardent II, and lived on board the vessel. All the crew were smokers apart from the skipper. Smoking was permitted in the crew mess room, but not in the accommodation. Vessel layout The central living area of Ardent II, situated on the weather deck, housed the galley, crew mess room, the toilet/washroom and a skipper’s cabin (Figure 2). The galley was equipped with an electric cooker, fridge and freezer, microwave oven and kettle. A portable electric hob was available in the galley for use when the vessel was connected to shore power as the electric cooker drew sufficient current to trip the shore connection when other electrical appliances were also in use. The crew mess room provided seating at two tables. It was equipped with a television and a satellite box, located forward of the exhaust trunking, that were usually left on stand-by. There was also a freezer on the starboard side forward of the engine room access hatch. The freezer was plugged into a multi-socket adapter located adjacent to it. Other electrical appliances, including a rice cooker and the crew’s mobile phones, were plugged into the adapter when required. No such appliances were plugged into the adapter at the time of the accident. The accommodation incorporated seven bunks and was fitted with an escape hatch to the aft net drum space on the port side. Normal access to the accommodation was via a vertical ladder from the crew mess room. A separate vertical ladder from the crew mess room gave access to the engine room. The wheelhouse, which incorporated a fire detection and alarm system panel, was accessed from the crew mess room via a stairway with a door fitted at the top. An aft door provided access from the wheelhouse to the aft upper deck. Figure 1: Ardent II on fire on 16 August 2016

4 Wash room Galley Skipper’s cabin Crew mess room Accommodation access hatch Engine room access hatch Accommodation escape hatch Engine room escape trunk Stairs up to wheelhouse Freezer Tables Exhaust trunking Figure 2: Ardent II layout

5 Fire protection, alarms and fire-fighting appliances Fire protection fitted to Ardent II was limited to the galley boundary. The galley was separated from the crew mess room by a fire door, but this was normally secured open, its auto-closure having been disconnected previously (Figure 3). A fire detection and alarm system was fitted to the vessel. The system was connected to smoke and heat detectors in the engine room, a heat detector in the galley and a smoke detector in the accommodation. The fire alarm, which sounded in both the wheelhouse and accommodation, was last tested in May 2016. Eight portable fire extinguishers were positioned throughout the vessel. These included 9-litre water extinguishers in the accommodation and crew mess room, and a 4.5kg dry powder extinguisher in the galley. The engine room was fitted with a CO2 fixed fire-extinguishing system, which could be activated from the aft net drum space. Two electrically driven pumps in the engine room could also be used as fire pumps, which fed fire hoses located on the main deck. A manual fire pump was located inside the main deck shelter. Maintenance, survey and inspection Ardent II was partly-owned by Grampian Sea Fishing Ltd, who provided general vessel management services. This included organising contractors for repairs and servicing, and arranging vessel inspections and surveys. Auto-closure mechanism Figure 3: Galley fire door with auto-closure disconnected

6 The vessel was last surveyed by the Maritime and Coastguard Agency (MCA) on 2 March 2016. On the basis of this survey, an International Fishing Vessel Certificate and a Load Line Exemption Certificate were issued, valid until 28 May 2017 and 2 March 2017 respectively. On 11 May 2016, a pre-guardship duty inspection was conducted by the Scottish Fishermen’s Federation. All identified deficiencies were closed out by 19 May 2016. Regulations and guidance Ardent II was built in 1986 in Campbeltown and was 24.36m registered length. When constructed, the vessel was certified against the requirements of The Fishing Vessels (Safety Provisions) Rules 1975 (1975 Rules), which include specific requirements for vessels under 24.4m registered length. Under the 1975 Rules, neither an automatic fire detection and alarm system nor a fixed fire-extinguishing system for machinery spaces was required for steel vessels of less than 24.4m registered length. However, Ardent II was fitted with a fixed fire-extinguishing system and a fire detection and alarm system for the engine room, which exempted the vessel from the requirement for the divisions separating the engine room from the accommodation, service spaces and control stations to be insulated to the ‘A60 standard’3. In 2005, the MCA published Marine Guidance Note (MGN) 291 (F) - Fire Detection and Alarm Systems on Fishing Vessels, which specified that fishing vessels over 24m registered length should have a fire detection and alarm system protecting the accommodation, service spaces and control stations. In 2008, the fire detection and alarm system on Ardent II was extended by connecting a heat detector in the galley and a smoke detector in the accommodation. As a result of a harmonisation of standards, Ardent II was later required to meet regulations for vessels over 24m registered length, resulting in an International Fishing Vessel Certificate being issued. In 2014, the MCA published MGN 501 (F), announcing a voluntary code of practice for fishing vessels of 24m registered length and over. In respect of fire detection and alarms systems, the Code requires such a system to be: ‘…installed and so arranged as to detect the presence of fire in all accommodation spaces and service spaces except those which afford no substantial fire risk, such as void spaces and sanitary spaces.’ The Code is intended to become mandatory in 2017, replacing, among other legislation, the 1975 Rules. Regulation 14 of The Merchant Shipping and Fishing Vessels (Provision and Use of Work Equipment) Regulations 2006 states: ‘The employer shall ensure that all ship’s electrical equipment and installations are so constructed, installed, operated and maintained that the ship and all workers are protected against electrical hazards.’ Similar accidents In 2008 a fire broke out in the galley/mess area of the fishing vessel Vision II4, while alongside in Fraserburgh. The three crew, who were living on board, all died as a result of the fire. Contributing factors to this accident were the disabling of the fire detection and alarm system and an inability of the galley fire door to self-close. 3 So insulated where necessary with suitable non-combustible materials such that, if the division is exposed to the standard fire test, the average temperature of the unexposed side of the division will rise not more than 139ºC above the initial temperature nor will the temperature at any one point, including any joint, rise more than 180ºC above the initial temperature within 60 minutes. 4 MAIB Report No 8/2009.

7 In 2010, as a result of this and other accidents, the MCA published MGN 413 (F) - Voluntary Code of Practice for Employment of Non European Economic Area (EAA) Fishing Crew, and MGN 425 (M+F) – Assessment of Risks for those sleeping on “Dead Ships”. MGN 413 (F) highlighted the voluntary code that had been produced by the SFF to highlight the social and practical responsibility of employing non- EEA fishing crew. The following extracts are relevant to this accident: •• ‘Do the alarms sound loud and clear in the accommodation and sleeping cabins? Consider fitting supplementary smoke detectors. •• Are fire doors self-closing or kept closed? •• Have heaters and cooking appliances been checked and are they safe to use in the accommodation? •• Are all electric appliances safe, with correct wiring, fuses (e.g. radios, TVs, toasters, kettles, phone chargers etc)?’ The following extracts of MGN 425 (M+F) are also relevant to this accident: •• (3.1) It is recommended that all vessels should have fire (smoke) and bilge alarms fitted with at least two independent power sources. The alarm should be so fitted that it can be heard within the accommodation in areas where crew may be sleeping. •• (3.3) It is strongly recommended that the alarm systems be capable of detecting, not only smoke and high bilge levels but noxious and flammable gases. •• (4.2.6) Are fire doors self-closing or kept closed (Note: unapproved holdbacks should be removed only electromagnetic holdbacks linked to the fire detection system are acceptable)? •• (7.1.10) Are all electric appliances safe, with correct wiring, fuses (e.g. radios, TV’s, toasters, kettles, phone chargers etc.)? ANALYSIS Source of the fire A detailed examination of the fire damage was undertaken to try to establish the source of the fire. The fire patterns and other evidence indicated the likely source of the fire was in the crew mess room. This ruled out it being a potential consequence of welding work carried out the day before the accident. Although the galley area was heavily damaged, the extent to which items had survived in this compartment also eliminated the possibility of the fire having originated there. Given that the crew were smokers and were permitted to smoke in the crew mess room, a discarded cigarette was considered as a possible source of ignition. However, based on the physical post- accident evidence of the scene and the time at which the fire was discovered, a discarded cigarette was considered to be an unlikely source. On excavation of the fire scene (Figure 4), it was apparent that the seat of the fire was at deck level in the gap between the freezer and the seating in the crew mess room (Figure 5). An aluminium plinth edging in this area had melted, there was heavy charring of the wood structure, and remnants of the multi-socket adapter (Figure 6), with the freezer plugged in, were found in the debris. A section of the fridge compressor electric cable was found intact, suggesting that the freezer itself was not the source of the fire. In light of this evidence, it is concluded that an electrical fault or failure of the multi-socket adapter was the most likely cause of the fire.

8 Figure 4: Fire damage to crew mess room Figure 5: Location of source of fire Engine room hatch Freezer Table supports Freezer side

9 Fire detection and alarm system It was fortunate that one of the crew awoke and discovered the fire shortly after it had started. In the absence of a smoke detector in the crew mess room, this alerted the crew to the fire and triggered them to evacuate the vessel without delay. The nearest sensor to the probable source of the fire was a heat detector in the galley, which would not have activated until later, or possibly not at all if the fire door to the galley had been shut. Ardent II’s fire detection and alarm system was initially restricted to the engine room as a means of exempting the engine room from being fitted with ‘A60 standard’ divisions. Efforts were subsequently made to upgrade the vessel’s fire detection and alarm capability but the description in the regulations of areas to be covered allowed the crew mess room to remain excluded. A more explicit instruction on which spaces on board similar fishing vessels require a means of fire detection to be fitted is included in MGN 501 (F), which is intended to become mandatory in 2017 and should address this omission. Action on discovery of the fire The fire quickly spread throughout the vessel. By using available portable fire extinguishers, and closing doors and other openings on discovery of the fire, the crew might have been able to restrict its spread, or even extinguish it, before they evacuated onto the quay. Such actions might then have enabled the fire service to start tackling the fire at an earlier stage in its development and so limit the resulting damage. However, the fact that the quay offered the crew a safe haven is likely to have influenced their decision to evacuate the vessel immediately, rather than to attempt to contain and extinguish the fire. Figure 6: Remnants of multi-socket adapter Remnants of multi-socket adapter Aluminium plinth edge

10 Electrical safety on board While good practice suggests that electrical appliances should be switched off and unplugged when not in use, some appliances, such as freezers and fridges, require a constant power supply. Ashore, the Electricity at Work Regulations 1989 place a responsibility on employers that any electrical equipment that has the potential to cause injury is maintained in a safe condition. The Regulations do not specify what needs to be done, by whom, or how frequently. However, Portable Appliance Testing (PAT) is a common method used by employers to discharge this responsibility. The Merchant Shipping and Fishing Vessels (Provision and Use of Work Equipment) Regulations 2006 place a similar responsibility on employers in relation to electrical equipment on fishing vessels. Regular visual inspection of electrical equipment to check for bare wires, that appropriate fuses are in place, and for signs of burning, together with regular PAT, would provide an effective means for reducing the risk of electrical fires. Living on board fishing vessels The MAIB investigation of the fire on board Vision II highlighted a number of concerns about crew living on board fishing vessels while in port. MGN 413 (F) and MGN 425 (M+F) summarise many of these issues and provide guidance on ensuring crew safety. The practice of crew locking themselves inside a vessel is an understandable measure from a security aspect and, as demonstrated in this accident, is unlikely to hinder their escape provided that they are immediately alerted of the need to do so and have the necessary capacity. However, in other circumstances, the situation could present significant problems for the fire service in both rescuing crew and fighting the fire. The issue of access to a fishing vessel in an emergency is something that owners and operators should take fully into account as part of their risk assessment before allowing crew to live on board in port. CONCLUSIONS •• An electrical fault or failure of a multi-socket adapter in the crew mess room was the most likely cause of the fire. •• There was no smoke detector fitted in the crew mess room and it was fortunate that one of the crew awoke and discovered the fire at an early stage. It enabled the crew to be alerted to the fire and triggered them to evacuate the vessel without delay. •• The fact that the quay offered the crew a safe haven is likely to have influenced their decision to evacuate the vessel immediately, rather than to attempt to contain and extinguish the fire. •• MGN 501 (F), which contains guidance on fire protection, is intended to be made mandatory in 2017. This will mean that the crew mess rooms of fishing vessels like Ardent II are fitted with a means of fire detection in the future. •• Regular visual inspection of electrical equipment to check for bare wires, that appropriate fuses are in place, and for signs of burning, together with regular PAT, would provide an effective means for reducing the risk of electrical fires. •• Although not a contributory factor of this accident, access to a fishing vessel in an emergency is something that owners and operators should take fully into account as part of their risk assessment before allowing crew to live on board in port.

11 ACTION TAKEN MAIB actions The Marine Accident Investigation Branch has: Issued a flyer to the fishing industry highlighting the safety issues identified in this report, and endorsing the practice of regular visual inspection and PAT of electrical equipment as an effective means for reducing the risk of electrical fires. RECOMMENDATIONS In view of current regulation and guidance, and that the voluntary code of practice for fishing vessels of 24m registered length and over is intended to become mandatory in 2017, no recommendations are made in this report. Safety recommendations shall in no case create a presumption of blame or liability

12 SHIP PARTICULARS Vessel’s name Ardent II Flag UK Classification society Not applicable IMO number/fishing numbers INS127 Type Fishing trawler Registered owner Private Manager(s) Grampian Sea Fishing Ltd Construction Steel Length overall 26.70m Registered length 24.36m Gross tonnage 251 Minimum safe manning Not applicable Authorised cargo Fish VOYAGE PARTICULARS Port of departure Peterhead Port of arrival Peterhead Type of voyage Not applicable Cargo information None Manning 7 MARINE CASUALTY INFORMATION Date and time 0500 16/8/16 (UTC+1) Type of marine casualty or incident Very Serious Marine Casualty Location of incident Peterhead Place on board Crew mess room Injuries/fatalities None Damage/environmental impact Constructive Total Loss Ship operation Alongside/moored Voyage segment Alongside/moored External & internal environment Dark, light breeze Persons on board 3

FD06046 ｜ 2015_英国_fire-in-the-engine-room-on-the-suction-dredger-arco-avon-with-loss-of-1-life__MAIBInvRepor.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2015 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H ACCIDENT REPORT VERY SERIOUS MARINE CASUALTY REPORT NO 17/2016 SEPTEMBER 2016 Report on the investigation of a fire in the engine room on the suction dredger Arco Avon 12 miles off the coast of Great Yarmouth resulting in one fatality on 18 August 2015

Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2016 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch Spring Place 105 Commercial Road Southampton Email: maib@dft.gsi.gov.uk United Kingdom Telephone: +44 (0) 23 8039 5500 SO15 1GH Fax: +44 (0) 23 8023 2459 Press enquiries during office hours: 01932 440015 Press enquiries out of hours: 020 7944 4292

CONTENTS GLOSSARY OF ABBREVIATIONS AND ACRONYMS SYNOPSIS 1 SECTION 1 - FACTUAL INFORMATION 2 1.1 Particulars of Arco Avon and accident 2 1.2 Narrative 3 1.2.1 Pre-accident events 3 1.2.2 The accident 5 1.2.3 Post-accident events 9 1.3 Environmental conditions 9 1.4 Vessel, equipment and manning 9 1.4.1 Vessel background 9 1.4.2 Machinery 10 1.4.3 Fuel system 11 1.4.4 Engine room CO2 fixed fire-extinguishing system 11 1.4.5 Manning 15 1.5 Safety management, survey and inspection 16 1.6 International Maritime Organization initiatives 19 1.7 Accident site examination 19 1.8 Abrasive wheels 20 1.9 Personal protective equipment 25 1.10 Other technical investigations 26 1.11 Medical care 30 1.12 Previous / similar accidents 31 1.13 Voyage data recording 31 SECTION 2 - ANALYSIS 32 2.1 Aim 32 2.2 Overview 32 2.3 Third engineer’s actions 32 2.4 Lone watchkeeping 33 2.5 Source of ignition 34 2.6 Abrasive wheels and hot work 35 2.7 Personal protective equipment 35 2.8 Emergency response 36 2.9 Medical treatment 37 2.10 Maintenance and survey 37 2.11 Safety management system 38 SECTION 3 - CONCLUSIONS 39 3.1 Safety issues relating to the accident that have been addressed or resulted in recommendations 39 3.2 Other safety issues relating to the accident 40 3.3 Other safety issues that have been addressed or resulted in recommendations 40 3.4 Other safety issues 40 SECTION 4 - ACTIONS TAKEN 41 SECTION 5 - RECOMMENDATIONS 42

FIGURES Figure 1

Chart showing location of accident Figure 2

Port engine room vent Figure 3a

Engine room main deck access Figure 3b

Engine room tank top Figure 4

Fuel system Figure 5

CO2 cylinder room Figure 6

CO2 system operating control cabinet Figure 7a

CO2 cylinder valve Figure 7b

CO2 manifold valve Figure 8a

Tools and angle grinder as found Figure 8b

Accident scene with tools and cable reel Figure 9

Pipe showing leak and partially detached hoop bracket Figure 10

Defective fuel pipe Figure 11a - Hoop bracket Figure 11b - Hoop bracket wear Figure 12a - Evidence of additional fretting Figure 12b - Evidence of pipe interference Figure 13

Pipe support stanchion Figure 14a - CAT scan of angle grinder Figure 14b - CAT scan of angle grinder switch Figure 15a - Spray test Figure 15b - Spray test Figure 16

Coverall ignition and flame spread test ANNEXES Annex A

HSE Research Report 222 Annex B

MSN 1731 (M+F)

GLOSSARY OF ABBREVIATIONS AND ACRONYMS AB

Able Seaman ANSI

American National Standards Institute AVR

Automatic Voltage Regulator BA

Breathing apparatus BST

British Summer Time CAT Scan

Computerized axial tomography scanner cm

centimetre CO2

Carbon dioxide COSWP

Code of Safe Working Practices for Merchant Seamen consolidated edition 20101 DoC

Document of Compliance FR

Fire retardant FRS

Fire and Rescue Service HAML

Hanson Aggregates Marine Limited HSE

Health and Safety Executive IMO

International Maritime Organization ISM

International Safety Management (Code) kg

kilogramme kW

kilowatt “Mayday”

A word used internationally as a distress signal in voice radio communication meaning the prefix to “Grave and Imminent Danger to a Person, Ship, Aircraft or Other Vehicle Requiring Immediate Assistance” MCA

Maritime and Coastguard Agency MCR

Machinery control room mg

milligramme MGN

Marine Guidance Note MGO

Marine gas oil mm

millimetre MSN

Merchant Shipping Notice 1 Replaced with Code of Safe Working Practices for Merchant Seafarers 2015 edition in September 2015.

NCN

Non-conformance note NFRS

Norfolk Fire and Rescue Service OOW

Officer of the watch PMS

Planned maintenance system PPE

Personal protective equipment PPE 99

The Merchant Shipping and Fishing Vessels (Personal Protective Equipment) Regulations 1999 PTO

Power take-off PTW

Permit to Work RNLI

Royal National Lifeboat Institution s

second SCMG

Ship Captain’s Medical Guide SMC

Safety Management Certificate SMS

Safety management system STCW

International Convention on Standards of Training, Certification and Watchkeeping for Seafarers 1978, as amended UMS

Unmanned machinery space UTC

Universal co-ordinated time V

Volt VDR

Voyage Data Recorder VHF

Very high frequency TIMES: all times used in this report are BST (UTC +1) unless otherwise stated

1 SYNOPSIS At 2312 on 18 August 2015, a fire broke out in the engine room of the dredger Arco Avon while the vessel was loading a sand cargo approximately 12 miles off Great Yarmouth, UK. Arco Avon’s third engineer, Anthony Jones, was badly burned in the fire and later died as a result of his injuries. As a consequence of the fire, which was extinguished following activation of the CO2 smothering system, the vessel lost all power and remained at anchor with its dredging equipment partially deployed for 9 days until it could be taken under tow and delivered to a repair facility. The MAIB investigation concluded that: • The third engineer was attempting to repair a failed fuel pipe when fuel, under pressure in the pipe, ignited. • No one else on board was aware of a failed fuel pipe or that the third engineer had apparently decided to repair it. • The third engineer was using a portable angle grinder to access the repair site, and had underestimated the risk of doing so. • Possible contributing factors to the third engineer’s underestimation of risk were that he had neither carried out a formal risk assessment nor sought a permit to work prior to commencing the repair. The Maritime and Coastguard Agency has taken actions to review and update: • The Code of Safe Working Practices for Merchant Seafarers to address the hot work hazard of sparks generated by the use of fixed and portable angle grinders. • Merchant Shipping Notice 1870 (M+F) to introduce an appropriate standard for overalls for work in engine rooms or other area where there is a risk of fire. • The Ship Captain’s Medical Guide to provide clear guidance on the appropriate medical treatment for serious burns. The Maritime and Coastguard Agency has been recommended to more widely promulgate the contents of IMO Circular MSC.1/Circ.1321 i.e. that owners and operators conduct periodic inspections of low pressure fuel system components. Bureau Veritas has been recommended to advise its surveyors of the contents of the circular. Hanson Aggregates Marine Limited has been recommended to review and, as appropriate, amend its safety management system to ensure, inter alia, that manning levels, watchkeeping duties and communication procedures provide for safe engine room operations at all times.

2 SECTION 1 - FACTUAL INFORMATION 1.1 PARTICULARS OF ARCO AVON AND ACCIDENT SHIP PARTICULARS Vessel’s name Arco Avon Flag United Kingdom Classification society Bureau Veritas IMO number/fishing numbers 8508383 Type Suction dredger Registered owner Hanson Aggregates Marine Limited Manager(s) Hanson Aggregates Marine Limited Construction Steel Year of build 1986 Length overall 98.6m Registered length 92.74m Gross tonnage 3474 Minimum safe manning 8 Authorised cargo Dredge spoil VOYAGE PARTICULARS Port of departure Dagenham Port of arrival Dagenham (planned) Type of voyage Commercial Cargo information Sand Manning 10 MARINE CASUALTY INFORMATION Date and time 18 August 2015, 2312 Type of marine casualty or incident Very Serious Marine Casualty Location of incident Approximately 12 miles off Great Yarmouth, UK Place on board Engine room Injuries/fatalities 1 fatality Damage/environmental impact Fire damage to engine room power cabling and control systems. Smoke damage throughout engine room. Ship operation Loading (dredging) Voyage segment Mid-water External & internal environment Wind: F4, sea state: moderate Persons on board 10

3 1.2 NARRATIVE 1.2.1 Pre-accident events At approximately 1200 on 18 August 2015, Arco Avon left Dagenham and proceeded along the east coast of the UK to its designated loading area. Loading involved dredging within a defined geographical box, designated ‘240’ by The Crown Estate2, located approximately 12 nautical miles off Great Yarmouth. The dredging operation usually took about 5 hours to complete. At 2000, approximately 1 hour before arrival at the loading area, the officer of the watch (OOW) called the third engineer to prepare the machinery for loading. The third engineer then proceeded to the engine room and changed the alarm and monitoring system from its remote, unmanned configuration, to local control for manned machinery space operations. He then started No.2 generator and transferred the vessel’s electrical load from the shaft alternator on to No.2 generator and prepared the dredging pump which would be powered by the shaft alternator. The normal arrangement for power distribution during loading operations would have been for No.3 generator to provide power for the vessel’s services with the shaft alternator supplying power for the dredging pump. Due to a defect on No.3 generator’s voltage regulator this machine was unavailable for use. Loading commenced at approximately 2124, after which there was no further communication between the bridge team and the third engineer prior to the accident. 2 The Crown Estate is a diverse portfolio of UK buildings, shoreline, seabed, forestry, agriculture and common land that generates revenue for the government each year. Arco Avon

4 Figure 1: Chart showing location of accident Reproduced from Admiralty Chart BA 1543 by permission of the Controller of HMSO and the UK Hydrographic Office.

5 1.2.2 The accident Loading proceeded as planned until 2312, when a fire detector activated. The fire detection panel on the bridge indicated that the fire was within the engine room. The alarm was reset on the fire detection panel, which was situated on the bridge, but it immediately reactivated, sounding throughout the accommodation on each occasion. The alarm was then reset again but it activated for a third time, alerting the rest of the crew, some of whom had already responded to the initial alarm and had proceeded to the bridge. The remaining crew then mustered rapidly on the bridge, which was the designated emergency station, with the exception of the third engineer, who was on watch alone in the engine room, and the chief officer, who was asleep following his period of duty. The master looked aft as the crew were mustering and noted smoke coming from engine room vents (Figure 2). Having established that there was a fire in the engine room, the master immediately made two telephone calls to the engine room, both of which were unanswered. Following a discussion with the master, the chief engineer was about to leave for the engine room to investigate further when the third engineer telephoned the bridge. He told the chief engineer that there was a fire in the vicinity of No.2 generator and that he had been hurt. The chief engineer told him to evacuate the engine room. The call had been made from the machinery control room (MCR), which meant that the engineer had to re-enter the engine room and exit through the fire as there was no direct escape route from the MCR. Figure 2: Port engine room vent Soot residues caused by smoke prior to the vents being closed

6 The master ordered the designated fire team to dress in fire suits and breathing apparatus (BA) in preparation to fight an engine room fire. The team consisted of the second engineer and an able seaman (AB). The chief engineer left the bridge and went aft to assume the role of on-scene commander. He was accompanied by the bosun, whose role was to shut the engine room ventilation flaps. At approximately 2314 the master began to recover the dredging equipment as it was essential to raise the dredge pipe off the seabed while the vessel still had sufficient electrical power so as to prevent significant damage. Approximately 1 minute later, once the pipe was clear of the seabed, he stopped the dredging pump. These actions were carried out concurrently with fire-fighting preparations. As the chief engineer and the bosun reached the aft end of the main deck, they met the third engineer, who had just exited the engine room through the doorway to the main deck (Figure 3a). His clothes appeared to have been burned off, and he was showing obvious signs of physical trauma, but he was able to walk unaided. He spoke briefly to the chief engineer, repeating the information relayed in his earlier telephone call. The chief engineer told the bosun to take the third engineer to the crew mess room and to summon first-aid assistance; he then relayed this information to the master via a portable very high frequency (VHF) radio. At approximately 2322, as the bosun and third engineer were crossing the main deck, the vessel’s main power supply failed, leaving the deck in darkness. After a few seconds, power was briefly restored but then failed again. Several seconds later, the emergency generator started, restoring emergency lighting. The bosun and the third engineer were then able to continue to the mess room, where the third engineer was left in the care of the cook. The bosun then returned to the aft end of the main deck to continue closing the machinery space ventilation flaps. The chief engineer had remained at the aft end of the main deck and opened the engine room access door. He was met by dense smoke and heat, and concluded that the fire team would be unable to enter the engine room. When the main electrical power failed at 2322, the voyage data recorder (VDR) ceased recording and power for recovery of the dredging equipment was lost. The second engineer dressed in a fire suit and BA, which was located on the bridge, while the fire team AB went to the forward machinery access void space to collect an additional fire suit and BA. The AB then took his equipment to the crew mess room, where the cook and another AB helped him to dress. The plan was for the fire team to proceed along one of the two tank top void spaces and enter the engine room through the door at the aft end of the void (Figure 3b). The chief officer had assisted the second engineer to dress and had left the bridge with him to take control of the fire team. However, the second engineer went ahead and proceeded alone along the starboard void space. The chief officer began preparations to rig a fire-fighting hose, for use by the fire team, from a hydrant at the forward end of the starboard void space. On arrival at the engine room entrance the second engineer opened the access door and quickly realised that the prevailing heat and smoke would prevent the fire team from entering the space. Aware that he was detached from the team and without means of communication, he closed the engine room door and began to retreat back along the starboard void. On reaching the forward end of the void, he met the fire team AB. The second engineer explained the situation to the AB,

7 Figure 3a: Engine room main deck access Main deck engine room access door Location of CO2 system control Figure 3b: Engine room tank top Fire scene Port void engine room access door Starboard void engine room access door

8 following which the two firefighters reported their findings to the chief officer. The chief officer then contacted the master using VHF radio and passed the information that the fire team were unable to enter the engine room. At 2325, the master broadcast a “Mayday”. He followed this up at 2336 with a request for evacuation of an injured crew member. Humber Coastguard responded by tasking a rescue helicopter and a Royal National Lifeboat Institution (RNLI) all-weather lifeboat to stand by. The chief engineer reported by VHF radio to the master and recommended that the engine room CO2 fixed fire-extinguishing system be activated to extinguish the fire. The master required a personnel muster to be completed before he would give approval to activate the CO2 system. Completion of the muster was delayed as a result of some confusion regarding the location of the second engineer. Once approval had been given, the chief engineer operated the CO2 system from its remote operating position in the main deck engine room access lobby. The chief engineer believed that the CO2 system had functioned correctly; however he went to the CO2 cylinder room to check. He found that approximately 50% of the cylinders had failed to operate, which he then manually activated. At 2338 the master reported to the Coastguard that CO2 flooding of the engine room had been carried out. Following operation of the CO2 system, the crew focused their efforts on containment of the fire through engine room boundary cooling and temperature monitoring. The only cooling required was on the main deck, directly above the seat of the fire. Water for boundary cooling was supplied by the vessel’s diesel-driven emergency fire pump that was located in the forward machinery space. During this period, the port anchor was deployed to ensure that the vessel remained in a navigationally safe position. The third engineer remained in the crew mess room accompanied by the cook, who had been designated the first-aider. Although visited by several members of the crew, including the master, his medical care was left solely to the cook and this was limited to the provision of liquid refreshment in the form of fruit squash. At 2352, the master reported to the Coastguard that the fire was under control, the engine room had been shut down and that the surrounding decks were cooling. The master reiterated concern over the serious condition of the third engineer and, at 2356, the Coastguard gave an estimated time of arrival of 15 minutes for an RNLI lifeboat and 35 minutes for a rescue helicopter. At 0013 on 19 August, two crew from the Gorleston RNLI lifeboat, one of whom was a paramedic, boarded Arco Avon. The paramedics assessed the third engineer’s condition and commenced medical treatment, which included administration of medical oxygen and pain relief medication. Following the arrival on scene of rescue helicopter R912, discussions between the helicopter pilot and the master identified the stern of the vessel as the most appropriate area for winching operations. The third engineer was placed in a stretcher and, with some difficulty, he was manoeuvred along the deck to the agreed winching position. At 0137, R912 departed for the Norfolk and Norwich University Hospital with the third engineer embarked. However, despite further medical treatment, the third engineer died of his injuries later that day.

9 As a result of discussion between the Coastguard and Norfolk Fire and Rescue Service (NFRS), consideration was given to airlifting a fire-fighting team out to the vessel. However, at 0250, the master reported to the Coastguard that the fire had been extinguished and that the situation on board was under control. Following this, the Coastguard released all search and rescue units from the scene and the decision was made that the NFRS team would board the vessel as it arrived in port. The vessel’s owner, Hanson Aggregates Marine Limited (HAML) arranged for a tug, Sea Beaver, to stand by Arco Avon in case further assistance was required. Arco Avon remained at anchor overnight with all of its remaining crew on board. 1.2.3 Post-accident events It was initially intended that Arco Avon would be towed into Great Yarmouth to allow an assessment of the damage and a forensic examination of the fire scene. However, owing to the vessel’s increased draught as a result of its partially deployed dredging equipment, there was insufficient depth of water for it to enter Great Yarmouth. The dredging equipment could not be recovered due to a lack of electrical power and possible damage to the vessel’s main switchboard, which was located in a compartment within the engine room and could not therefore be accessed. Despite an examination by a salvage team who had been appointed by the owners, an alternative method of recovering the equipment safely could not be identified. On the evening of 19 August, some of Arco Avon’s crew departed the vessel, leaving the master, chief engineer, second engineer and chief officer on board with a tug standing by in attendance. The crew noted that water was entering the engine room bilges. It was concluded that a seawater-fed cooler in the vicinity of the seat of the fire was the most likely cause of sea water ingress. During the dredging operation the main sea water suction intakes had been changed over from the main engine room to the forward machinery space to reduce contamination caused by the dredging process. This meant that the crew were able to enter the forward machinery space and close the sea water isolation valves to stem the leak. A salvage pump was provided by the tug to allow removal of the flood water. On the evening of 20 August the master, chief engineer, second engineer and chief officer were replaced by a relief crew. In view of a forecast of deteriorating weather and high winds for the area, a salvage operation was initiated by HAML and, on 22 August, the vessel was towed to a sheltered anchorage. Arco Avon was taken to a repair/layby berth in Hull on 27 August. 1.3 ENVIRONMENTAL CONDITIONS At the time of the accident, weather conditions were good, with a moderate sea state and force 4 winds. 1.4 VESSEL, EQUIPMENT AND MANNING 1.4.1 Vessel background Arco Avon was a suction dredger operated by HAML, the UK marine division of the Heidelberg Cement Group. The vessel was constructed in 1986 and was the first of four sister vessels built by Appledore Shipbuilders Limited.

10 The vessel was built to the requirements of The Merchant Shipping (Cargo Ship Safety Construction and Survey) Regulations 1984 and The Merchant Shipping (Fire Protection) Regulations 1984. Compliance with the regulations was verified under survey by Bureau Veritas (BV). Arco Avon was classed by BV for dredging up to a distance of 15 miles offshore. The vessel’s recent operations had involved loading sand and gravel off the coast of Great Yarmouth and discharging the cargo at Dagenham. 1.4.2 Machinery Propulsion Arco Avon’s propulsion comprised a Mirrlees Blackstone KMR6 Mk3, 4 stroke, single acting, 6 cylinder in line diesel engine driving a controllable pitch propeller through a reduction gearbox. At the free end of the engine, was a power take-off (PTO) driving a shaft alternator (known on board as No.1 generator) through a flexible coupling. The shaft alternator supplied 1,800kW at 660V. Auxiliary generators Arco Avon was equipped with two auxiliary generators: • A Mirrlees ESL 8 Mk II 8 cylinder engine rated at 1065kW (known on board as No.2 generator) that generated electricity at 660V. • A Caterpillar 3408 DI-TA 8 cylinder engine (designated the harbour generator on the vessel’s plans, although known on board as No.3 generator), that generated electricity at 415V. Generator configuration Arco Avon’s electrical supply and distribution system could be configured in a number of different ways to suit particular operations. When in transit, all electrical power was provided by the shaft alternator through a step down transformer to 415V to supply the vessel’s services. When loading, the shaft alternator provided power directly to the dredging pump, which required 660V, with all other services supplied by No.3 generator. When discharging, No.2 generator provided all required electrical power. At the time of the accident, Arco Avon’s No.3 generator was out of service due to a fault with its automatic voltage regulator (AVR) that limited the supply voltage to about 380V. This prevented it from being synchronised with the other generators. Although spare parts to rectify the defect had been delivered before the vessel’s departure from Dagenham on 18 August, these were found to be incorrect and the fault was not repaired. The chief engineer conducted an impact assessment and concluded that the vessel could continue to operate. The defect meant that loading operations would be carried out using the shaft alternator to supply power for the dredge pump and No.2 generator to supply the rest of the electrical load. However, the lack of redundancy meant that if any further electrical faults developed, dredging operations would have to be curtailed. The master concurred with the assessment but the operational

11 programme was amended to carry a sand load to Dagenham instead of a mixed load of sand and gravel to Belgium. This was to allow service engineers to meet the vessel at Dagenham and repair the defective AVR. 1.4.3 Fuel system Arco Avon’s main engine and auxiliary generators ran on marine gas oil (MGO), commonly referred to as diesel. Fuel was stored in two wing bunker tanks, treated directly through a centrifugal purifier, and transferred to two daily service tanks (Figure 4). Fuel for the main engine was pumped from the service tanks via booster pumps to a main engine fuel supply rail, from which high pressure injection pumps took their suction. Surplus fuel was returned to the suction side of the booster pumps via a pressure regulating valve designed to maintain the return side of the system at a minimum pressure of approximately 1.8 bar. The auxiliary generator engines’ fuel supplies were gravity-fed directly from the service tanks with returns directed back to these tanks. The fuel system pipework was manufactured from steel and complied with ANSI3 schedule 40, which classified it as a 1½ inch (38mm) pipe, with an outside diameter of 1.90 inches (48mm) and a wall thickness of 0.15 inch (4mm). The pipe ran beneath the engine room floor plates on the outboard side of No.2 generator. It was supported along its length by steel hoop brackets, clamping it against support stanchions. 1.4.4 Engine room CO2 fixed fire-extinguishing system The engine room fixed fire-extinguishing arrangements were provided by a Kidde CO2 smothering system. The system comprised 23 x 45kg CO2 cylinders arranged for simultaneous release that was designed to achieve a ≥40% by volume concentration of CO2 in the space. The CO2 cylinders were located in a dedicated storage room on the starboard side at the aft end of the main deck (Figure 5). The amount of CO2 needed to reduce the oxygen level to a point at which fuels are prevented from burning varies depending upon the properties of individual fuels. To ensure that smothering systems are effective, the concentration of CO2 required is relatively high (40% in a marine environment) and this therefore results in an atmosphere which cannot sustain human life. The CO2 system was activated from an operating cabinet located in the access lobby to the engine room at the aft end of the main deck (Figure 6). The cabinet door was kept locked with a key available in a ‘break glass’ box on the front of the cabinet. Opening the cabinet automatically sounded an alarm and stopped the engine room ventilation fans. Inside the cabinet were two pilot cylinders, each with its own isolation valve, and two main system valves. Valve A was operated to open the discharge manifold to the engine room distribution pipework. Valve B was then opened to allow gas pressure from the pilot cylinders to operate a piston in the discharge valve fitted to each of the main cylinders (Figure 7a) which, in turn, allowed gas from the main cylinders into the discharge manifold. If the remote operation failed, the discharge manifold valve could be opened locally in the CO2 cylinder room (Figure 7b) and the main CO2 cylinders could be opened manually. 3 American National Standards Institute.

12 Figure 4: Fuel system Main engine fuel supply rail and high pressure injection pumps Pressure regulating valve Return line Daily service tanks Main engine booster pumps Bunker storage tanks

13 Figure 5: CO2 cylinder room Figure 6: CO2 system operating control cabinet Isolation valves Main system valves Pilot cylinders

14 Figure 7a: CO2 cylinder valve Manual operating valve Main cylinder discharge valve Figure 7b: CO2 manifold valve Discharge manifold valve

15 1.4.5 Manning Arco Avon had been issued with a safe manning document requiring a minimum of eight crew, including a chief engineer, a second engineer and one additional watchkeeping engineer. There was no requirement for any engine room ratings. The document was endorsed for Unmanned Machinery Space (UMS) operations stating: ‘a lone watchkeeper must not attend the machinery spaces unless a suitable watch alarm is fitted and operational or an alternative system is in place.’4 At the time of the accident, there were 10 crew on board, with the engineering department’s manpower at the minimum level required by the safe manning document. Arco Avon was manned with a crew of UK nationals. During transits to the loading area, they operated with a deck officer and lookout on the bridge and the engine room in the UMS mode. When dredging, the bridge manning was supplemented by the master, who controlled the dredging operation, and the engine room reverted to a manned condition with a single engineer officer as watchkeeper. All crew members were employed through Hanson Ship Management Limited, a crewing agency dedicated to providing manpower for HAML vessels. Master The master was an experienced seafarer and had worked within the dredging industry for more than 20 years. He held an STCW II/2 qualification and had been master on Arco Avon since 2011. Chief engineer The chief engineer had started his career in the Royal Navy and, after serving approximately 8 years, left to join the Merchant Navy. His civilian seafaring was predominantly within the dredging industry, where he had fulfilled the roles of both chief engineer and technical superintendent. He held an STCW III/2 qualification and had been chief engineer of Arco Avon since 2013. Third engineer Anthony (Tony) Jones had served for 26 years in the Royal Navy. On leaving in 1994, he had an interest in a motorcycle repair business and then relocated from Devon to Kent, where he worked in the ship repair industry before joining the Merchant Navy. He spent 2 years working on ferries before moving to the dredging industry. He joined Hanson Ship Management Limited in 2013 and was appointed to Arco Avon as third engineer. He held an STCW III/1 qualification. Tony was well regarded on board Arco Avon and within the company. He was noted for being extremely ordered, and for identifying issues that needed attention and completing tasks as soon as practicable. Tony died as a consequence of severe burns he received in the fire. Postmortem examination results gave the cause of death as multiple organ failure. Subsequent toxicology results indicated that he had consumed no alcohol. 4 Section 15.9.1 of the Code of Safe Working Practices for Merchant Seamen (COSWP) 2010 edition, which was the current version at the time of the accident, states that seafarers should only enter or remain in an unmanned machinery space alone to carry out a specific task that they may be expected to complete in a comparatively short time. Furthermore, before entering, leaving and at regular intervals while within the space, they must report by telephone or other means provided to the duty deck officer. Similar guidance is now provided in the Code of Safe Working Practices for Merchant Seafarers 2015 edition.

16 1.5 SAFETY MANAGEMENT, SURVEY AND INSPECTION Safety Management System (SMS) HAML operated a fleet of eight dredgers (seven in service and one in long-term lay-up). Seven vessels were managed under a Document of Compliance (DoC) issued to the company by the MCA on 15 November 2010 and individual Safety Management Certificates (SMC) issued to each ship. The DoC had been subject to annual verification by the issuing authority, the last audit having been carried out on 11 February 2015. Arco Avon’s SMC was issued on 27 May 2015. The attending MCA auditor noted that: • There was a comprehensive system in place for risk assessments and permits to work. • The maintenance of the deck and engine room departments was of a high standard. • Engine room maintenance and cleanliness were of a high standard. • The SMS contained a list of potential shipboard emergencies, which formed part of the vessel’s regular emergency drill programme. • During the fire drill the ability and knowledge of the two BA wearers were particularly good. The auditor issued two non-conformance notes (NCN), one related to crew familiarisation with SMS updates and the other to evidence of annual document review. Permits to work Within the master’s standing orders, which applied to all crew, there was reference to the permit to work (PTW) system and risk assessments. The following statement appeared under the heading of risk assessments: ‘The vessel operates its own risk assessment procedure. The ships staff are to generate a risk assessment for all tasks carried out onboard prior to any job being undertaken. All personnel onboard should be involved in risk assessment. All risk assessments should be used in conjunction with the company permit to work system. All risk Assessments should be reviewed every 12 months.’ [sic] No evidence was found to indicate that a documented risk assessment had been conducted or a PTW had been raised in respect of the attempted fuel pipe repair that is concluded to have resulted in the fire. Lone working The SMS contained instruction relating to lone working in the engine room: ‘Engine Room Procedures AE09 Lone Working in the Engine Room

17 Ref – CoSWP 15.9 – Unmanned Machinery Space This procedure applies to both Engineer Officer of the Watch and any other member of the crew entering the engine room at any time.

1. The term ‘Engine Room’ applies to all machinery spaces
2. The use of machinery spaces as a means of access to other areas of the ship

should be avoided

1. Whenever anyone enters a machinery space, whether alone or with a co-worker,

the bridge Officer of the Watch must be informed

1. Whenever anyone enters a machinery space they should be equipped with a

hand held radio which is in contact with the bridge OOW

1. It is accepted that there may be areas of the engine room where there is no, or

poor, radio reception. These areas should be identified and additional safeguards put in place, eg inform the bridge OOW and increase frequency of telephone contact

1. Whenever a lone worker or watchkeeper is in the engine room, the Patrol (or

Watchkeeping) Alarm must be activated

1. If the Patrol Alarm is out of service then an alternative robust procedure must

be put in place, eg regular phone or radio contact at intervals of no more than 15 minutes or increase the manning of the watch

1. Whenever work is carried out in the engine room or machinery spaces, especially

by a lone worker, particular attention should be paid to safe working practices as laid down in the CoSWP, in particular (but not exclusively) with regard to – • Manual handling • Machinery guarding • Personal Protective Equipment – especially hearing protection • The handling of dangerous substances • The use of hand tools • The use of portable electrical equipment’ Loading operations The chief engineer’s standing orders also formed part of the SMS and contained a specific section relating to loading operations: ‘LOADING OF SHIP

1. The Bridge will notify the Duty Engineer before loading is due to commence.
2. De-ballasting can commence on instruction from the Bridge OOW.
3. Ensure sufficient power is available and systems ready for loading.

18

1. Change over SW suctions from Aft to Forward. The Forward SW suction only is

to be used whilst dredging.

1. Before starting check IHC power pack and Dredge pump L.O. tank levels. Check

Auto-greaser level. Ensure flushing pumps are running.

1. During load the Duty Engineer will progress routine duties and planned

maintenance as required.

1. Duty Engineer is to inform the Bridge of his whereabouts should he need to leave

the machinery space.

1. On completion of loading the Bridge will normal inform the Engine Room that

loading is completed.

1. SW suction can now be changed from Forward to Aft.
2. The bridge will inform the Engine Room de-watering may commence.
3. Inform the Chief Engineer of any problems immediately.’ [sic]

Maintenance Coupled with the SMS was a computer-based planned maintenance system (PMS). The PMS was used for both routine tasks based on a formal planned maintenance strategy, and for the recording of ad hoc tasks generated manually to record breakdowns and repairs. Classification society surveys Arco Avon’s certificate of class was issued by BV on 10 November 2011. A range of class surveys had been carried out by BV in February 2015. These included annual, continuous and occasional machinery surveys. The following is an extract from Bureau Veritas Rules for the Classification of Ships: ‘Chapter 3 - ‘Scope of Surveys’ (All Ships) 3.5 Systems in machinery spaces 3.5.2 The fuel oil, lubricating oil, hydraulic oil, thermal oil, and feed and cooling water systems, together with pressure filters, heaters and coolers used for essential services, are to be opened up and examined or tested, as considered necessary by the Surveyor. Safety devices for the foregoing items are to be examined.’ In respect of the class surveys conducted on board Arco Avon, there had been no concerns raised relating to the low pressure fuel pipework or the fuel system in general.

19 1.6 INTERNATIONAL MARITIME ORGANIZATION INITIATIVES IMO circular MSC.1/Circ.1321 – Guidelines for Measures to Prevent Fires in Engine-Rooms and Cargo Pump-Rooms was issued on 11 June 2009. The circular contained annexed guidelines detailing the prevention measures. Part III Chapter 1 of the Annex concerns the control of flammable oils within engine rooms. It covers issues with design and installation, maintenance, system pressure pulses and vibration. The following is an extract: ‘1.5 Inspection and maintenance 1.5.1 The ship safety management system should contain procedures to identify vibration, fatigue, defects, poor components and poor fitting of the fuel system and ensure that proper attention to protecting hot surfaces is maintained. Means, such as check lists should be prepared to ensure that all procedures are followed at major overhauls and that all components, supports, restraints, etc., are refitted on completion of such work. The installed system should be routinely inspected for: .1 verification of the adequacy of its supports and the condition of its fittings; .2 evidence of fatigue stresses to welded or brazed pipes and connections; .3 assessment of the level of vibration present; and .4 condition of the lagging or shielding of hot surfaces. 1.5.2 Components of the fuel system should be comprehensively examined, particularly threaded connections, at each dismantling. 1.5.3 Injection pump holding-down bolts should be proved tight by testing with a torque spanner at frequent intervals (not to exceed 3 months). 1.5.4 The supports and retaining devices of the low pressure fuel system should be checked at regular intervals (not to exceed 6 months), to be proved tight and to provide adequate restraint. The lining of such devices should be examined for wear and renewed if they provide insufficient support.’ The circular invited Member Governments to bring the annexed guidelines to the attention of ship designers, owners, operators, shipbuilders and other parties concerned. This had not been formally actioned in the UK. 1.7 ACCIDENT SITE EXAMINATION An examination of the accident site was conducted by MAIB inspectors from 28 August 2015. There was evidence of heat and flame damage on the outboard side of No.2 generator which extended around the area local to the fire scene. The contents of a mezzanine deck used for storage directly above the fire scene were burned, and plastic fittings and containers had melted with debris from the deck falling down onto the accident site. There was severe smoke damage throughout the engine room.

20 Smoke patterns and heat damage generally radiated away from the deck adjacent to the bedplate of No.2 generator. Electrical power cables and control wiring in the vicinity of No.2 generator had suffered extensively from heat damage. Floor plates adjacent to No.2 generator were missing, giving access to a defective fuel pipe located beneath the floor plates. An angle grinder, a spanner, and a ratchet with an attached socket were located on the floor plates adjacent to the access opening (Figure 8a). A cable reel to provide power to the angle grinder was plugged into a nearby power socket (Figure 8b). A number of hose clips and pieces of rubber jointing material were found on the workshop bench along with spare angle grinder discs and, in the vicinity, an operators’ handbook. The hoop bracket supporting the fuel pipe at the position of the leak was hanging below the pipe with one retaining nut still in place (Figures 9 and 10). The other end of the bracket was sheared off and the remaining nut was seized onto the bracket. The bracket had significant material wastage and was twisted along its length (Figures 11a and 11b). Along the length of the defective fuel pipe there was evidence of fretting in way of the support brackets and from contact with other pipes where flanged pipe joints rested against the fuel line (Figures 12a and 12b). The pipe supporting stanchion in way of the hanging hoop bracket displayed evidence of a cut, the width of which matched the thickness of the angle grinder disc (Figure 13). 1.8 ABRASIVE WHEELS Safe use of, and the associated hazards relating to abrasive wheels is referred to both in COSWP and by the Health and Safety Executive (HSE) in its publication HSG17 (third edition published 2000) ‘Safety in the use of abrasive wheels’. COSWP section 20.4 gives guidance on the use of portable power-operated tools but does not specifically refer to angle grinders. Section 20.6 is titled abrasive wheels and contains information relating to fixed bench grinders and their abrasive wheels but makes no reference to the use of portable grinders. HSG17 is a comprehensive document which recognises that portable and hand-held grinding machines form the largest number of abrasive wheels in use and details checks, tests and standards relating to the equipment. Neither the guidance in COSWP nor that within HSG17 makes any reference to the hazards created by the sparks generated during grinder operation. Although not referred to in its guidance (HSG17), in 2004 the HSE produced a research report (HSE Research Report 2225), which studied the ignition and flame spread of clothing fabrics subjected to angle grinder sparks (Annex A). The findings of this project are discussed at Section 1.10. COSWP chapter 24 identifies that risk assessments and appropriate control measures are required to protect personnel affected by hot work. It states that welding and flame-cutting activities in places other than the workshop should generally be the subject of a PTW. While acknowledging welding and flame-cutting as hot work activities, the chapter makes no mention of cutting and grinding using abrasive wheels. 5 Research Report 222 – Tests for the ignition of flame spread of clothing fabrics subjected to angle grinder sparks – prepared by Building Research Establishment Ltd for the HSE 2004.

21 Figure 8a: Tools and angle grinder as found Debris from mezzanine deck Angle grinder Socket and ratchet Spanner Cable reel Tools Fire site No.2 generator Figure 8b: Accident scene with tools and cable reel

22 Figure 9: Pipe showing leak and partially detached hoop bracket Figure 10: Defective fuel pipe Pipe support stanchion Hoop bracket Hole in pipe Defective pipe Hole in pipe Evidence of fretting

23 Figure 11a: Hoop bracket Figure 11b: Hoop bracket wear Fractured end Wear from bracket location in support stanchion

24 Figure 12a: Evidence of additional fretting Figure 12b: Evidence of pipe interference Evidence of fretting Evidence of fretting

25 The Code of Safe Working Practices for Merchant Seafarers 2015 edition provides similar guidance to that referred to above in respect of COSWP consolidated edition 2010, which it replaced in September 2015. One of the risk assessments (Engine Room Work Shop Activities) in Arco Avon’s SMS, noted the need to keep flammable material away from the bench grinder in the workshop. However, use of the bench grinder was listed as a non-hot work activity and the instruction was not extended to the use of portable grinders. The operators’ handbook for the portable angle grinder found at the accident site stated, in two separate sections, that the tool should not be operated near flammable materials as sparks could ignite them. While Arco Avon’s crew generally acknowledged that use of an angle grinder constituted a hot work activity, a PTW was often not implemented, particularly in respect of work carried out within the machinery spaces. 1.9 PERSONAL PROTECTIVE EQUIPMENT The Merchant Shipping and Fishing Vessels (Personal Protective Equipment) Regulations 1999 (PPE 99) require ship owners and employers to ensure that personal protective equipment (PPE) is provided for seafarers engaged in, or at risk Figure 13: Pipe support stanchion Hoop bracket Angle grinder witness mark

26 from hazardous work activities on board UK registered ships. MSN 1731 (M+F)6 (Annex B) details standards of PPE and states that work in engine rooms or any area where there is a risk of fire requires overalls made of fabric of low flammability. To enhance safety, HAML policy had been for all crew to wear high-visibility coveralls when on the outside deck of their vessels and, with time, this practice had extended to the engine room. During an onboard safety committee meeting on 5 May 2014, crew members on a sister vessel had raised concerns that the high-visibility coveralls, which were only supplied as a polyester/cotton garment, may not afford the level of low flammability required under PPE 99. Subsequently, HAML revised its policy and issued a Fleet Directive (E356) dated 18 December 2014, detailing the PPE that each employee was entitled to have, and the process for ordering and obtaining it. Engineers were allowed three white cotton coveralls and one high-visibility polyester/cotton coverall. The revised PPE policy also stated that engineers should wear a high-visibility vest over their cotton coveralls when on the outside deck. 1.10 OTHER TECHNICAL INVESTIGATIONS Angle grinder The angle grinder found on the engine room floor plates adjacent to the failed fuel pipe was examined using a computerized axial tomography scanner (CAT scan) to determine its operating condition (Figure 14a). The scan identified that the operating switch contacts were open, indicating that the grinder had been switched off (Figure 14b). Fuel pipe testing Flow and spray tests on the defective fuel pipe were conducted by the MAIB in conjunction with Burgoynes, Consulting Scientists and Engineers, who were acting on behalf of HAML. Tests were carried out using water at a range of pressures from 1 through to 5 bar, with 2 bar being deemed most representative of the likely pressure in the pipe at the time of the accident. The tests showed a powerful jet of fluid spraying from the hole in the defective pipe. The flow and pressure were sufficient for the fluid to atomise as it hit the ground beneath the test rig (Figure 15a). When a screwdriver was introduced into the jet to simulate the position of the hoop bracket, the degree of atomisation increased significantly (Figure 15b). PPE flammability testing Flammability testing of coveralls available to the third engineer (67% polyester 33% cotton and 100% cotton) along with a control sample from a new 67% polyester 33% cotton coverall were carried out at a UK test facility. Limited flame spread testing according to the provisions of EN ISO 15025:2002 of three fabric samples was carried out: Sample 1 - 100% cotton – used but laundered. Sample 2 - 67% polyester 33% cotton – used but laundered. Sample 3 - 67% polyester 33% cotton – new and unused. 6 Replaced with MSN 1870 (M+F) in January 2016. The details relating to engine room PPE (overalls) remain unchanged.

27 Figure 14a: CAT scan of angle grinder Figure 14b: CAT scan of angle grinder switch Open contacts

28 Figure 15a: Spray test Figure 15b: Spray test Defective pipe Spray pattern representative of impact with hoop bracket

29 The tests showed that each of the samples was made from a fabric that supported combustion after removal of the ignition source. These results demonstrated that neither of the coverall types available to the third engineer would have given any flame or heat protection. Following the initial tests, a further sample of each garment was soaked in diesel to determine the likely effects of ignited diesel on each fabric type. Adding diesel significantly increased the burning rate of all the samples. None of the coveralls from which the samples were taken were labelled to indicate that they were designed or supplied to give heat or flame protection. The following standards are recognised as applicable to PPE tested against EN ISO 15025:2002 to meet limited flame spread claims: EN ISO 14116:2015: Protective clothing. Protection against flame. Limited flame spread materials, material assemblies and clothing. EN ISO 11611:2015: Protective clothing for use in welding and allied processes. EN ISO 11612:2015: Protective clothing. Clothing to protect against heat and flame. HSE Research Report 222 (Annex A) tested a number of fabrics for ignition and flame spread when they were subjected to angle grinder sparks. The report studied natural and synthetic fabrics and both untreated and fire retardant (FR) garments. The report’s conclusions include the following: • ‘The range of synthetic and natural fibres tested showed ignition times and flame-spread rates covering a wide range but there was no particular distinction between values obtained for the two types. • Synthetic fibres tended to melt and natural fibres tended to form a surface char, both effects having a strong influence on performance in the ignition and flame spread tests. • Natural/synthetic fibre mixes were not particularly distinguishable from their “pure” counterparts. • A cotton fabric marked as “FR interliner” ignited readily. • A cotton fabric purchased as “FR” and checked for acceptable FR content by elemental analysis did exhibit very limited ignition and flame spread behaviour, both before and after being subject to a single standard laundry cycle. • Tests on a full sized manikin clothed in separate tests with an FR and non FR “coverall” demonstrated the value of FR treated materials when subjected to sparks from an angle grinder. The FR coverall did not ignite after 5 minutes of sustained spark impingement whereas the non FR coverall ignited after approximately 20 seconds and burnt to completion.’ (Figure 16) Courtesy of HSE Figure 16: Coverall ignition and flame spread test after approximately 45 seconds (ignition occurred at approximately 20 seconds)

30 1.11 MEDICAL CARE Requirements for the provision of onboard medical care are defined in The Merchant Shipping (Maritime Labour Convention) (Minimum Requirements for Seafarers etc.) Regulations 2014, which are promulgated through MGN 482 (M) Maritime Labour Convention, 2006: Medical Care. Under these regulations, ship owners are required to provide seafarers with medical care on board ship. This is achieved through a combination of the provision of medical stores and training of personnel. The Merchant Shipping Act 1995 Part III Chapter 53 states that where a ship does not carry a doctor, the master is responsible for ensuring that any necessary medical attention is given by him or under his supervision by a duly appointed person. On Arco Avon, the master’s standing orders stated that the OOW was to administer first-aid treatment and that the master was to be informed. The master was also to be informed immediately of any major injury occurrences, and was to be consulted if there were any doubts concerning treatment. The Ship Captain’s Medical Guide (SCMG) is published by the MCA for use on ships not carrying a doctor. The guide gives the following advice in respect of burns: ‘Burns and scalds, clothing on fire ■ by far the best way to put out a fire on a person is to use a dry powder fire extinguisher at once; ■ if a dry powder extinguisher is not available, then lay the person down and smother the flames by wrapping him in any available material (not made of man-made fibre), or throw buckets of water over him, or use a hose; ■ make sure all smouldering clothing is extinguished. NOTE: The powder from a fire extinguisher will not cause much, if any, eye damage. Most people shut their eyes tightly if sprayed with powder. Any powder which gets in the eye should be washed out immediately after the fire has been extinguished and while cooling is being undertaken. Heat burns and scalds ■ all heat burns should be cooled as quickly as possible with running cold water (sea or fresh) for at least ten minutes, or by immersing in cold water and keeping the injured part in motion; cooling of extensive burns (>15%) should be avoided as hypothermia will result; ■ if it is not possible to cool the burn on the spot, the casualty should be taken to where cooling can be carried out; ■ try to remove clothing gently but do not tear off any which adheres to the skin; ■ then cover the burned areas with a dry, non-fluffy, dressing which is larger than the burns and bandage in place; ■ further treatment as in Chapter 4.

31 Chapter 4 ■ burns of over 18% of the body surface in adults or 10% in children or older persons … Until removed to hospital put the patient to bed and seek to restore the fluid balance by encouraging the patient to drink as much as possible. Put rehydration powder into the drink according to the instructions; (if not available dissolve 1 teaspoonful of salt in 1/2 litre of water). If vomiting occurs and persists, fluid per rectum may be necessary. Relieve pain and start antibiotic treatment. Remove rings, jewellery or constricting items of clothing. Anxiety may be relieved by giving diazepam 5 mg, repeated every 4 hours. Cling film makes a good temporary dressing for large burns.’ The SCMG was not referred to on board Arco Avon following the accident and prior to the third engineer’s evacuation by helicopter. 1.12 PREVIOUS / SIMILAR ACCIDENTS The MAIB and other authorities have investigated a number of fires and incidents relating to failures in low pressure fuel systems. These have largely been the result of flange/joint failures rather than pipe failures, but have resulted in similar atomised fuel leaks. Norsea On 2 September 2002, the passenger/ro-ro cargo ferry Norsea suffered an engine room fire while on passage in the North Sea. A low pressure fuel pipe had fretted against its clamp and housing, causing a small hole in the pipe’s wall. Following its own investigation of the accident, the management company introduced a monthly procedure for checking the security of fuel pipes, clamps and mounting bolts, and inspection of pipework for chafing. It also fitted additional clamps to the pipework. At the time of the accident, Norsea’s engine room was manned by a lone engineer. The MAIB investigation report No.16/2003 states: ‘It might be argued that machinery spaces which are manned are likely to provide an enhanced level of vessel safety. However, when attended by a lone watchkeeper, who is protected only by a watch alarm designed for short inspection visits, any improvement in vessel safety might be at the price of placing that watchkeeper at an elevated level of risk’. 1.13 VOYAGE DATA RECORDING The information from the VDR was downloaded on 21 August 2015 while the vessel was still at anchor off Great Yarmouth. The information had been correctly saved by the crew. However, contrary to VDR performance standards and for reasons that have not been established, the unit did not continue to record using its own uninterruptible power supply when the vessel lost its main power supply.

32 SECTION 2 - ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 OVERVIEW Arco Avon suffered an engine room fire that resulted in the death of the third engineer. Physical evidence from the fire scene indicates that the third engineer was attempting to remove a securing bracket so he could make a temporary repair to a leaking low pressure fuel return pipe on the main engine fuel system. The main engine remained running throughout the repair attempt. During the process his overalls became soaked in fuel and, when he used a portable angle grinder to remove the bracket, high energy sparks from the grinding disc ignited the atomised fuel from the leak and his diesel soaked clothing. This resulted in the fire in the engine room. He was able to exit the engine room and reach medical assistance, but not before he had sustained burns that ultimately proved fatal. It is not known why the third engineer attempted such a hazardous repair on his own and without informing the bridge watchkeeper or the chief engineer of either the problem or his intentions. However, the onboard routines for lone working in the engine room specifically allowed maintenance to be carried out during the watch. While there was a documented requirement to complete a PTW this was often waived for work in the machinery spaces. Further, reports of the third engineer’s character indicate that once he had started the repair task he would likely have persevered to complete it on his own despite it becoming increasingly challenging. 2.3 THIRD ENGINEER’S ACTIONS It is likely that the third engineer became aware of a fuel leak through the smell of diesel in the engine room as there was no evidence of any fuel system or bilge alarms on the engine room data logger. Access covers in the engine room floor plates had been lifted, which suggests that he conducted a search of the area to find a diesel leak. It is apparent that he found the defective fuel pipe beneath the floor plates adjacent to No.2 generator. Tools and equipment had been systematically prepared to carry out a repair from which it is concluded that he had decided to take autonomous action to repair the failed fuel pipe. Although it was both required (chief engineer’s standing orders and SMS Engine Room Procedure AE09 Lone Working in the Engine Room) and expected as good practice, on this occasion the third engineer informed neither the chief engineer nor the bridge OOW of a problem. The master’s standing orders also stated a requirement for risk assessments and permits to work to be completed. By acting alone, without any other members of the crew being aware of his intentions, the third engineer compromised both his own and the rest of the crew’s safety.

33 Had the third engineer informed the chief engineer and/or bridge OOW of the situation and his intended action, or had there been another watchkeeper in the engine room, it is probable that a discussion and some form of dynamic risk assessment would have been conducted. His omission in not informing the chief engineer or bridge OOW is likely to have been influenced by the onboard culture of routine lone working and absence of regular and frequent communication. The apparent initial repair attempt may be rationalised through the third engineer’s experience and positive attitude to prompt remedial action. At first glance, it could have appeared to be a relatively quick and straightforward task to release the hoop bracket and to slide a hose clip along the pipe to cover the hole. However, to access the hoop bracket retaining nuts, the third engineer would have needed to stand in the bilge adjacent to the pipe run. He would have been in very close proximity to the leak and would almost certainly have been soaked in fuel from the spray created by the leak. Notwithstanding this, once the extent of fretting had been uncovered, the realisation that a hose clip and rubber jointing would not suffice should have been apparent and ought to have prompted a review of his intentions. His apparent decision to use an angle grinder to cut the pipe support stanchion is therefore more difficult to rationalise as once the securing bracket had been freed it would have been relatively easy to pull the pipe away from the stanchion in order to fit a hose clip. Despite the lack of direct guidance in respect of angle grinders and hot work, given his experience, the third engineer ought to have recognised the extreme risk of using any tool capable of generating sparks in close vicinity to atomising fuel, causing him to review his intended actions. During the fuel pipe testing following the accident, using a screwdriver to simulate the position of the hoop bracket indicated that the diesel spray leaving the hole in the pipe and impacting on the defective hoop bracket and surrounding structure would have atomised to a degree likely to have resulted in a readily ignitable fuel source. The examination of the angle grinder showed that the operating switch contacts were open. This finding, along with the positioning of the grinder, suggest that it is likely to have been placed in its resting position as a deliberate act by the third engineer. The witness mark on the pipe support bracket matched the width of the angle grinder disc, indicating that the tool had been used as part of the third engineer’s apparent repair attempt. In reviewing the task, the third engineer would have been aware that isolating the fuel system would have involved stopping the main engine which, in turn, would have interrupted the loading programme. Once he had reached this critical point in the event, it might have been his professional pride and confidence in his ability to successfully complete the repair that drove him to carry on with the task. 2.4 LONE WATCHKEEPING Arco Avon was certified for UMS operations, but at the time of the accident the engine room was in a manned condition with a single engineer in attendance. SMS Engine Room Procedure AE09 Lone Working in the Engine Room mandated the use of the UMS patrol alarm whenever a lone worker or watchkeeper was in the engine room. These instructions were compliant with the vessel’s safe manning document and with the guidance available in COSWP (section 15.9.1).

34 The requirement to use the patrol alarm was acknowledged by the vessel’s crew. However, it was deemed an impractical tool to cover the periods when the engine room was manned by a lone watchkeeper. The use of telephone or radio to keep the bridge OOW informed of the watchkeeper’s location was a more normal procedure. This system was recognised within the SMS but only for periods when the patrol alarm was out of service. Given the routine, rather than exceptional, practice of not using the patrol alarm, the requirement to communicate with the bridge OOW at not more than 15-minute intervals had been allowed to lapse, as evidenced by the lack of communication on this occasion from 2124, nearly 2 hours prior to the accident. As noted during the Norsea investigation (section 1.12), the practice of operating machinery spaces with a lone watchkeeper not only risks a loss of contact in a potentially hazardous environment, but it also encourages an individual to act autonomously. With no local co-worker to challenge an individual’s actions, the risks of commencing and subsequently compounding hazardous activities are exacerbated. In the case of Arco Avon, there was a specific requirement, mandated through the chief engineer’s standing orders, that during loading operations the duty engineer was to progress routine duties and conduct planned maintenance while on watch. This instruction effectively condoned lone working and wasnot consistent with the guidance contained in section 15.9.1 of COSWP or with similar guidance now provided in the Code of Safe Working Practices for Merchant Seafarers 2015 edition. The patrol alarm was designed specifically for unmanned operations. It allowed the on-call engineer to investigate alarms or conduct machinery space inspections while maintaining a form of contact. However, it is not a suitable system to allow a watchkeeper to perform maintenance tasks during their period of duty. 2.5 SOURCE OF IGNITION The fire scene investigation revealed extensive heat and smoke damage in the vicinity of the failed fuel pipe. Two potential sources of ignition were considered: the running No.2 generator, which was adjacent to the defective fuel pipe, and sparks generated through use of the portable angle grinder. The exhaust pipe on No.2 generator was at the top of the engine and ran along its outboard side. This was the component on the generator most likely to have been a source of sufficient heat to ignite a diesel spray. However, the flame spread pattern on the side of the generator indicated that the fire had started at the base of the engine and had spread up its side towards the exhaust system. Furthermore, the fuel pipe testing following the accident indicated that the diesel spray from the defective pipe would have been directed away from the generator and into the engine room bilge. Therefore it was concluded that the generator was unlikely to have provided the ignition source. The location of the angle grinder and the witness mark on the fuel pipe support stanchion provided strong evidence that the angle grinder had been used during the third engineer’s apparent repair attempt. The fuel pipe testing demonstrated that an atomising spray of fuel would have resulted as the fuel impinged on the hanging hoop bracket and adjacent engine room structure. This, coupled with the findings of

35 HSE’s Research Report 222 (Annex A), led to the conclusion that the high energy sparks from the portable angle grinder were the probable source of ignition for the fire. 2.6 ABRASIVE WHEELS AND HOT WORK The use of angle grinders for both cutting and grinding is commonplace on board ships. They are exceptionally versatile tools, well suited to general repair work both within and outside of an engine room. It is widely accepted that the sparks generated by abrasive wheels during grinding and cutting operations contain sufficient energy to ignite flammable substances. Therefore, both grinding and cutting processes constitute hot work and should be treated as such. Furthermore, there are a number of documented fires in shore-based industries that have been caused as a result of using angle grinders. Guidance relating to hot work and use of abrasive wheels on board UK registered ships is readily available in COSWP. However, in its current form, the Code could be misinterpreted as not recognising grinding as a hot work activity. COSWP defines hot work as welding and flame cutting, but makes no specific mention of cutting and grinding using abrasive wheels. Conversely, the section relating to abrasive wheels concentrates on the hazards associated with disc and wheel failures, but does not identify flammability hazards related to the generation of sparks. The evidence from this and other accidents, coupled with research findings, clearly indicate that the sparks generated by using fixed and portable angle grinders produce a hot work hazard. However this is not currently acknowledged in marine industry guidance. 2.7 PERSONAL PROTECTIVE EQUIPMENT During the evidence gathering phase of the investigation it was noted that, despite the change in company policy relating to the provision of cotton coveralls, the majority of engineers on HAML vessels continued to wear polyester/cotton high-visibility coveralls. MSN 1731 (M+F) (Annex B) stated that for Work in engine rooms or any area where there is a risk of fire the PPE provided should be Overalls made of fabric of low flammability. The MSN gave the examples of natural fibre, high cotton content and non-flammable clothing. However, it did not quote a standard to indicate suitability and fitness for purpose. The limited flame spread test carried out on sample garments during this investigation and the HSE study into ignition and flame spread of fabrics subject to angle grinder sparks demonstrated that cotton coveralls, on their own, do not offer suitable protection against ignition and flame spread. The fact that coveralls of ‘high cotton content’ were listed in MSN 1731 (M+F) as suitable for use in engine rooms as a protection against fire is a significant concern. This might have led to an unfounded confidence that a cotton garment would provide a recognised level of fire protection. Notwithstanding the findings in relation to the properties of cotton coveralls, even fabric with an FR finish is likely to ignite and continue to burn after being soaked with diesel. The presence of atomised fuel coupled with the wicking effect of a diesel- saturated coverall resulted in an extremely flammable garment that could be readily ignited by the introduction of sparks generated from the use of an angle grinder.

36 2.8 EMERGENCY RESPONSE Following activation of the fire alarm, Arco Avon’s crew mustered quickly and the decision-making process was generally well considered and appropriate. However, the uncoordinated approach to the engine room by the members of the fire team led to the second engineer becoming separated from the fire team AB. The lack of co-ordination was the direct result of the two firefighters dressing in widely separated locations and no re-muster and briefing from the fire team controller (chief officer) before deployment. A further concern was that both the chief engineer and the second engineer opened different access doors to the engine room while alone and without any fire-fighting medium for protection. Had the two doors been opened simultaneously, a through draught could have introduced additional oxygen, causing the fire to burn more intensely. The second engineer was new to the vessel and his lack of familiarity with its layout might explain why he became separated from the fire team AB. However, neither he nor the chief engineer should have attempted to open an engine room access door until the fire team were assembled with the appropriate equipment and had been fully briefed on their required actions. Re-entry of a closed compartment containing a confirmed or suspected fire is an extremely hazardous activity that requires comprehensive planning and a co-ordinated approach. Current entry protocols for a machinery space fire are for the BA team to be ‘on air’ and equipped with a charged fire hose, and for them to initially crack open the door approximately 10cm for 10s to ascertain the conditions within the compartment before proceeding. In his role of on-scene commander, the chief engineer should have taken control at the scene and directed the fire team to attempt the investigation following a thorough assessment of boundary conditions. Once the decision to use the CO2 fixed fire-extinguishing system had been made it was important to activate the system rapidly. The sooner the system could be operated, the more effective it was likely to be. That the second engineer had become separated from the fire team AB delayed a full headcount from being achieved, which consequently delayed the master’s approval to activate the system. Having operated the CO2 system from the control cabinet, the chief engineer’s decision to confirm the successful activation of the system from the cylinder room, was entirely appropriate. On entering the room, he discovered that 50% of the cylinders had failed to operate, and so manually opened the remaining cylinders to complete the system discharge. The reason why many of the cylinders had not operated remotely could not be established. Once the CO2 system had been operated, checking the cylinder room should have been a controlled operation with the room being treated as an enclosed space. There was a significant risk that CO2 would have escaped into the cylinder room and thus depleted the oxygen content of the atmosphere. Unless the atmosphere inside a CO2 cylinder room can be confirmed as safe, BA must be used when entering the room or the initial emergency could be compounded with additional casualties. Following operation of the CO2 system, containment through boundary cooling using the emergency fire pump and temperature monitoring was competently implemented and maintained.

37 2.9 MEDICAL TREATMENT The first-aid treatment given to the third engineer immediately following his evacuation from the engine room was rudimentary. It was not until an RNLI paramedic arrived that care, commensurate with the extent of his injuries, was instigated. Although Arco Avon’s master’s standing orders required the OOW to administer first-aid treatment, care of the third engineer was left to the cook. While the cook had undertaken elementary first-aid training, it would not have prepared him to deal with a casualty in such a serious condition. MGN 482 (M) Maritime Labour Convention, 2006: Medical Care reiterates that The Merchant Shipping (Training and Certification) Regulations 1997 require all deck officers to be trained in medical first-aid or medical care. Furthermore, as confirmed in Arco Avon’s master’s standing orders, the Merchant Shipping Act 1995 gives the master responsibility for ensuring that medical treatment is given either by him or under his supervision. All of the crew were shocked by the extent of the injuries to the third engineer, and it is concluded that this contributed significantly to the low level of medical care administered by them. The published guidance available on board was not referred to as treatment was left entirely to the cook, who did not have ready access to the SCMG. Given the current SCMG guidance that cooling of extensive burns (>15%) should be avoided as hypothermia will result, reference to it might not have prompted any cooling of the third engineer’s burns. Without cooling, the extent and consequent effect of the burns would have continued to increase for a significant time. Independent medical advice suggests that cooling is effective for up to 3 hours but most effective in the first 30 minutes. While cooling of extensive burns may risk inducing hypothermia, this could be managed by selectively cooling individual areas while maintaining body temperature through the use of clean dry blankets or clothing and, where possible, operating in a heated compartment. In this case, the severity of the injuries was such that any action by the crew was unlikely to have saved the third engineer. 2.10 MAINTENANCE AND SURVEY Inspection of the fuel system pipework following the accident revealed that the fuel leak had been the result of fretting caused by relative movement between the pipe and one of the hoop brackets. The fretting had resulted in a deep groove around approximately 50% of the pipe’s circumference and a 2mm diameter hole at the bottom of the pipe. The condition of the defective pipe was consistent with other pipes located beneath the engine room floorplates. HAML’s PMS recorded both planned and breakdown maintenance. However, there were no specific tasks within the system for inspection of the low pressure fuel system. The vessel was classed with BV and had been subject to surveys every year since build. While BV’s guidance to surveyors included notes relating to the fuel system, no defects or anomalies had been noted in respect of Arco Avon. The IMO had recognised that inspection of the fuel system was an important element of fire prevention. IMO circular MSC.1/Circ. 1321, dated 11 June 2009, was issued to provide guidelines for measures to prevent fires in engine rooms and cargo pump

38 rooms. This circular, the contents of which had not been promulgated by the MCA to the shipping industry, recommends a 6-monthly inspection of all low pressure fuel system components to be included in the vessel’s SMS. The majority of the pipe runs forming the main engine low pressure fuel system were located beneath the floor plates in Arco Avon’s engine room and were not readily accessible for inspection or survey. The layout of the pipework allowed flanges from one pipe to come into contact with another, resulting in wear due to fretting. More damaging were the system’s steel hoop brackets used to clamp the pipes against their support posts. These brackets were not fitted with any form of packing between the pipe and bracket. Vibration associated with the normal operation of the vessel and possible vibration induced by pressure pulses within the system allowed the brackets to loosen. The combination of vibration and material loss through corrosion resulted in loosening of the brackets in service, with some failing completely. The loose brackets then allowed relative movement between the materials, resulting in fretting of the pipework. On examination following the accident, the condition of the low pressure fuel system pipes was generally found to be poor and not commensurate with the overall material condition of the engine room. The lack of a requirement in HAML’s PMS for fuel system pipework inspections to form part of the vessel’s routine maintenance was compounded by the discretionary nature of BV’s survey of fuel system pipework, and this allowed defective pipework below the engine room floor plates to go undetected. 2.11 SAFETY MANAGEMENT SYSTEM The auditor who issued Arco Avon’s SMC on 27 May 2015 noted that there was a comprehensive system in place for risk assessments and PTWs. However, the SMS did not contain any risk assessments relating to hot work procedures involving grinding or cutting activities using abrasive wheels. While there were extensive records of completed PTWs, none of them referred to a risk assessment, either generic or task specific, contrary to the master’s standing orders that a risk assessment was to be generated prior to any job being carried out. While HAML’s SMS provided a tool to support its vessels in operating in a safe and consistent manner, operational inconsistencies on board Arco Avon were not detected by the vessel’s shore management team. There appears to have been no challenge during internal audits in relation to particular anomalies between direct instruction (standing orders) and safe systems of work (e.g. regularity and frequency of communications with engine room staff, and the use and coupling of risk assessments and PTWs). This suggests that while HAML operated an SMS that was compliant with the ISM Code, the practical application of the SMS was deficient in some areas.

39 SECTION 3 - CONCLUSIONS 3.1 SAFETY ISSUES RELATING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

Although required by Arco Avon’s chief engineer’s standing orders and elsewhere in the vessel’s SMS, the third engineer informed neither the chief engineer nor the bridge OOW of a fuel leak and his apparent intended action to repair it. His reason for not doing so is likely to have been influenced by the onboard culture of routine lone working and absence of regular and frequent communication. [2.3]

The UMS patrol alarm was deemed on board Arco Avon to be an impractical tool to cover the periods when the engine room was manned by a lone watchkeeper. [2.4]

Given the routine, rather than exceptional, practice of not using the patrol alarm, the vessel’s SMS requirement to communicate with the bridge OOW at not more than 15-minute intervals had been allowed to lapse. [2.4]

Arco Avon’s chief engineer’s standing orders requiring the duty engineer to progress routine duties and conduct planned maintenance while on watch, effectively condoned lone working and was not consistent with the guidance provided in section 15.9.1 of COSWP or with similar guidance now provided in the Code of Safe Working Practices for Merchant Seafarers 2015 edition. [2.4]

From the results of fuel pipe testing following the accident and the findings of HSE Research Report 222, it is concluded that high energy sparks from the portable angle grinder were the probable source of ignition for the fire. [2.5]

The fact that sparks generated by using fixed and portable angle grinders produce a hot work hazard is not currently acknowledged in marine industry guidance. [2.6]

The presence of atomised fuel coupled with the wicking effect of a diesel-saturated coverall, regardless of the fact that the third engineer was not wearing a coverall of 100% cotton, resulted in an extremely flammable garment that could be readily ignited. [2.7]

A combination of vibration and fuel pipe support bracket material loss through corrosion had resulted in loosening of the brackets in service and consequent fretting of the pipework. The fretting had resulted in a hole in a low pressure fuel pipe located below the engine room floor plates. [2.10]

The contents of IMO circular MSC.1/Circ.1321, which recommends a 6-monthly inspection of all low pressure fuel system components to be included in the vessel’s SMS, had not been promulgated by the MCA to the shipping industry. [2.10]

The lack of a requirement in HAML’s PMS for fuel system pipework inspections to form part of the vessel’s routine maintenance was compounded by the discretionary nature of BV’s survey of fuel system pipework, and allowed defective pipework below the engine room floor plates to go undetected. [2.10]

While HAML’s SMS provided a tool to support vessels in operating in a safe and consistent manner, operational inconsistencies on board Arco Avon were not detected by the vessel’s shore management team. [2.11]

40 3.2 OTHER SAFETY ISSUES RELATING TO THE ACCIDENT7

The third engineer’s apparent initial repair attempt may be rationalised through his experience and positive attitude to prompt remedial action. However, his apparent decision to then use an angle grinder to cut the pipe support stanchion is more difficult to rationalise. [2.3]

The third engineer would have been aware that isolating the fuel system would have involved stopping the main engine which, in turn, would have interrupted the loading programme. It might have been his professional pride and confidence in his ability to successfully complete the repair that drove him to carry on with the task. [2.3] 3.3 OTHER SAFETY ISSUES THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

The fact that coveralls of ‘high cotton content’ were listed in MSN 1731 (M+F) might have led to an unfounded confidence that a cotton garment would provide a recognised level of fire protection. [2.7]

An uncoordinated approach to the engine room led to the fire team being separated and was a direct result of the two firefighters dressing in widely separated locations with no re-muster and briefing from the fire team controller before deployment. [2.8]

The chief engineer and second engineer opened different access doors to the engine room while alone and without any fire-fighting medium for protection. Had the two doors been opened simultaneously, a through draught could have introduced additional oxygen causing the fire to burn more intensely. [2.8]

Following activation of the CO2 fixed fire-extinguishing system, the chief engineer entered the CO2 cylinder room without wearing BA and without first checking and confirming the atmosphere inside the room as safe. [2.8]

Contrary to Arco Avon’s master’s standing orders, medical care of the third engineer was left to the cook, who had not received the level of training to prepare him to deal with a casualty in such a serious condition. However, the severity of injuries was such that any action taken by the crew was unlikely to have saved the third engineer. [2.9]

Current SCMG guidance that cooling of extensive burns (>15%) should be avoided as hypothermia will result, is not consistent with independent medical advice received by the MAIB following the accident. [2.9] 3.4 OTHER SAFETY ISSUES7

The practice of operating machinery spaces with a lone watchkeeper not only risks a loss of contact in a potentially hazardous environment, but it also encourages an individual to act autonomously. With no local co-worker to challenge an individual’s actions, the risks of commencing and subsequently compounding hazardous activities are exacerbated. [2.4] 7 These safety issues identify lessons to be learned. They do not merit a safety recommendation based on this investigation alone. However, they may be used for analysing trends in marine accidents or in support of a future safety recommendation

41 SECTION 4 - ACTIONS TAKEN The Maritime and Coastguard Agency has taken actions to review and update: • The Code of Safe Working Practices for Merchant Seafarers to address the hot work hazard of sparks generated by the use of fixed and portable angle grinders. • Merchant Shipping Notice 1870 (M+F) to introduce an appropriate standard for overalls for work in engine rooms or any area where there is a risk of fire. • The Ship Captain’s Medical Guide to provide clear guidance on the appropriate medical treatment for serious burns. Hanson Aggregates Marine Limited has issued: • Fleet Directive E363 detailing amendments to personal protective equipment requirements on its vessels. • Fleet Directive M33 requiring fleet-wide inspection of fuel systems. • Fleet Directive M34 requiring fleet-wide inspection of high temperature surface insulation and spray shields.

42 SECTION 5 - RECOMMENDATIONS The Maritime and Coastguard Agency is recommended to: 2016/136 Review International Maritime Organization circular MSC.1/Circ.1321 – Guidelines for measures to prevent fires in engine-rooms and cargo pump-rooms, and, as appropriate, promulgate its contents to the shipping industry. Hanson Aggregates Marine Limited is recommended to: 2016/137 Review and, as appropriate, amend its safety management system to ensure: • Its planned maintenance system includes a 6-monthly inspection of all low pressure fuel system components, as recommended in IMO circular MSC.1/Circ.1321. • Manning levels, watchkeeping duties and communication procedures provide for safe engine room operations at all times. • Portable angle grinding is included and addressed as a hot work activity. • Fire-fighting training is enhanced to address and correct re-entry techniques. • A procedure is included for entering the CO2 cylinder room following system activation. • Improved oversight by shore management to identify operational inconsistencies, particularly with regard to the use of risk assessments and permits to work. • The master’s role and responsibilities, when dealing with medical casualties, is reinforced. Bureau Veritas is recommended to: 2016/138 Advise its surveyors of the contents of IMO circular MSC.1/Circ.1321 – Guidelines for measures to prevent fires in engine-rooms and cargo pump-rooms. Safety recommendations shall in no case create a presumption of blame or liability

Marine Accident Report

FD06047 ｜ 2016_英国_flooding-and-sinking-of-the-wooden-potter-majestic__MAIBInvReport_16_2016.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2016 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H VERY SERIOUS MARINE CASUALTY REPORT NO 16/2016 JULY 2016 1 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of such an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2016 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gsi.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 ACCIDENT REPORT Foundering of fishing vessel Majestic (LK678) 5 nautical miles off Yell, Shetland 21 January 2016 SUMMARY At 14431 on 21 January 2016 the 16m wooden potter Majestic (LK678) sank close to a gas pipeline off the Point of Fethaland, Shetland. The vessel’s sinking resulted from a flood in the engine room that had started sometime between 0900 and 1000. The skipper and crewman abandoned into a liferaft shortly after they became aware of the ingress of water and were rescued by a nearby fishing vessel. No efforts were made to stem the flood and the vessel foundered approximately 4 hours after it was abandoned. There were no injuries and no pollution. The pipeline’s owner, British Petroleum, later arranged for sandbags to be placed on the seabed between it and the wreck to prevent damage to the pipeline. 1 All times are UTC. Majestic Image courtesy of Mike Pennington

2 The cause of the engine room flood is unknown but it was probably caused by a failure within a seawater system or sea valve. The investigation identified that: • The engine room’s bilge alarms sounded in the wheelhouse and were not heard by the skipper and the crew who were working on the deck. Consequently, the engine room had been flooding for up to 1 hour before the crew were alerted. • The rate of the flooding would have been slowed by the operation of the electric bilge pump and possibly by stopping the main engine. • The effectiveness of the crew’s response to the flood was probably reduced because they did not routinely conduct emergency drills. • The prompt abandonment and notification to the coastguard and vessels in the area reduced the risk to the lives of the crew. This accident indicates that relevant guidance and the lessons from similar accidents were not being sufficiently heeded. In the accident, Majestic’s wheelhouse was left unattended for a prolonged period and the crew were not practised in reacting to a flood on board. In addition, they did not wear personal flotation devices while working on deck and did not don lifejackets when abandoning the vessel. A recommendation made to Majestic’s owners is aimed at improving the safe operation of any vessels they may own or operate in the future. FACTUAL INFORMATION Narrative At 0530 on 21 January 2016, the 16m wooden potter Majestic (LK678) sailed from Toft, Shetland, and headed for fishing grounds approximately 5nm to the north-east of Point of Fethaland (Figure 1). On board were the vessel’s skipper and a crewman; both men were wearing oilskin trousers and jackets. The weather was fine and clear and the visibility was good. There was a moderate south-easterly breeze and the sea was slight with a 1m swell. Majestic reached the fishing grounds at about 0630. Over the next 3½ hours, the skipper and the crewman worked three fleets of creels. This involved hauling, emptying, re-baiting, re-shooting the creels and processing the catch of brown crab. The hauler and its associated hydraulics were operating, which made it very noisy on the deck. During this period, other than when manoeuvring the vessel between the fleets, the wheelhouse was left unattended. Each fleet of creels took approximately 1 hour to process. At about 1000, the skipper went to the wheelhouse and heard the engine room’s bilge alarms sounding. He opened a forward wheelhouse window and shouted to the crewman on the deck below to check the engine room for flooding. Accordingly, the crewman lifted the engine room escape hatch (Figure 2). The skipper leant out of the wheelhouse window and he and the crewman saw that the engine room was flooded. The water level was above the deck plates, half way up the side of the engine casing. The skipper and the crewman immediately went to the wheelhouse roof and launched the liferaft. The skipper then returned to the wheelhouse and, at 1010, he pressed the very high frequency (VHF) digital selective calling (DSC) distress alert. He also broadcast a “Mayday” via VHF radio, channel 16. The DSC distress message was received by the Shetland Coastguard Operation Centre (CGOC) and the nearby fishing vessels Fairway II (LK270) and Kestrel (LK268). The coastguard watch officer was

3 Figure 1: Yell sound fishing grounds (inset: Orka gas pipeline and position of sinking Reproduced from Admiralty Chart BA 1239 by permission of the Controller of HMSO and the UK Hydrographic Office. Fishing grounds Point of Fethaland Toft Position flood discovered Gas pipeline Final position of Majestic

4 Liferaft stowage Engine compartment escape hatch Work bench and hauler Vivier tank flush deck opening Liferaft stowage Work bench and hauler Cabin Engine compartment Engine compartment escape hatch Vivier tank For’d hold Figure 2: Vessel layout

5 unable to acknowledge the message due to a technical fault but he quickly established contact with Majestic’s skipper on VHF radio, channel 16. Fairway II and Kestrel were 4nm and 3.4nm respectively from Majestic; Fairway II immediately started to close to assist and Kestrel’s fishing gear was hauled. Majestic’s skipper informed the coastguard that the vessel was sinking due to a flood in the engine room. He also advised that it was too dangerous to go into the compartment and that he intended to abandon into a liferaft. The coastguard watch officer advised the skipper to take with him a hand-held VHF radio and the vessel’s Emergency Position Indicating Radio Beacon (EPIRB). The watch officer also tasked rescue helicopter R900, which was on a training exercise in the area. The Aith and Lerwick all-weather lifeboats (AWL) were not tasked due to their distance from Majestic, an assessment that the vessel was sinking quickly and the relatively close proximity of R900, Fairway II and Kestrel. On leaving the wheelhouse, Majestic’s skipper took a hand-held VHF radio and his mobile phone, but he did not collect the EPIRB. At about 1017, the skipper and the crewman boarded the liferaft; neither wore a lifejacket. The men cut the liferaft painter and then paddled the liferaft clear of Majestic. Majestic’s engine was running but it was not in gear. At about 1022, R900 arrived on scene soon followed by Fairway II. At 1033, Majestic’s skipper and crewman were recovered from the liferaft onto the fishing vessel (Figure 3). Majestic remained upright and on an even keel (Figure 4) and over the next 4½ hours the vessel drifted towards the Orka Vow gas pipeline to the north (Figure 1 inset); Fairway II and Kestrel were in attendance. Shetland CGOC informed British Petroleum (BP), the pipeline’s owners, of the situation and, at 1333, the coastguard also discussed the possibility of taking Majestic in tow with Fairway II’s skipper. However, Fairway II’s skipper considered it too dangerous to put anyone on board. At about 1440, Majestic heeled to port and trimmed by the stern. It sank 3 minutes later in a depth of 100m. The vessel’s EPIRB released and activated and was recovered by Kestrel’s crew. Post-foundering actions A 500 metre temporary exclusion zone was established over the site of Majestic’s wreck and BP deployed a guard vessel to the area. A remotely operated vehicle (ROV) survey identified that Majestic had settled on the seabed 2 metres from the gas pipeline; the vessel was heeling towards the pipeline. As a precaution, the ROV was used to place large sandbags between Majestic and the pipeline to prevent the wreck from toppling (Figure 5). Crew Majestic’s skipper (aged 34) and the crewman (aged 35) were brothers and had been fishermen all their working lives. They had co-owned Majestic since 2012 and crewed the vessel themselves. Other crewmen had occasionally been employed to assist but recruitment and retention of fishermen was difficult due to competition from the oil and gas industry. The skipper held a Seafish2 under 16.5m skipper certificate3 and the crewman held a fishing vessel class 1 Certificate of Competency (CoC) as well as a fishing vessel engineer class 2 CoC. Both men had completed the mandatory Seafish training courses4. Neither the skipper nor the crewman usually wore lifejackets when working on deck. They also did not routinely conduct emergency drills. 2 Seafish – the Sea Fish Industry Authority works across all sectors of the UK seafood industry to promote good quality and sustainable seafood, and to improve the safety and standards of training for fishermen. 3 The under 16.5m skippers’ certificate is a voluntary qualification administered by Seafish aimed at increasing the navigational and engineering knowledge of skippers and watchkeepers on small fishing vessels. 4 New entry fishermen must complete basic safety courses in sea survival, elementary first-aid; fire-fighting and health and safety. Fishermen with two years’ experience must also complete a 1-day mandatory safety awareness course run by Seafish.

6 Figure 3: Majestic’s crew recovered by Fairway II Figure 4: Majestic at 1036

7 Construction Majestic was built in 1977 to a traditional wooden design with oak frames and iroko planking. The vessel complied with The Fishing Vessels (Safety Provisions) Rules 1975 and comprised three main internal compartments: fish hold (that contained a vivier5 tank), engine room and accommodation (Figure 2). The vessel’s bulkheads were not required to be watertight but the bulkhead between the fish hold and the engine room was required to be of solid and substantial construction. The bulkhead between the engine room and the accommodation, and the engine room deck head were steel and met the B15 fire rating standard6. 5 The vivier tank was a holding tank with a capacity of approximately 12t that was used to keep shell fish alive. Pumps maintained a continuous supply of fresh sea water to the tank, the level of which was controlled by an overflow system that discharged excess water overboard. 6 A bulkhead or deck head constructed of non-combustible material with an insulation value that ensures the average temperature of the unexposed side to a fire will not rise more than 140°C above the original temperature, nor will the temperature at any one point, including any joint, rise more than 225°C above the original temperature within 15 minutes. Figure 5: Sandbags positioned against the wreck

8 Seawater inlets and discharges overboard The table below summarises Majestic’s seawater inlets and discharges and the status of the associated machinery at the time of the flood: Engine Room Inlets / Discharges Inlet System Discharge Running 50mm (S) Main Engine cooling water 50mm (P) Yes Gearbox cooling water (supplied from engine cooling water) 40mm (P) Yes 40mm (P) Vivier Tank circulating pump Unknown Yes 50mm (P) 13m3 /hr Deck Wash / bilge pump 50mm (S)

Yes

11m3 /hr Electric Engine / Room bilge pump 30mm (P)

No

Hand operated Engine Room bilge pump 60mm (S) No

Other discharges

11m3 /hr Electric Fish Hold bilge pump

No

6m3 /hr Electric Accommodation bilge

No

Hand operated Fish Hold bilge pump No (P) = Port side, (S) = Starboard side Bilge pumps and alarms Majestic was fitted with four power driven-bilge pumps: two in the engine room, one in the fish hold and one in the accommodation. Hand-operated bilge pumps were also fitted in the engine room and the fish hold and a portable engine-driven salvage and fire pump was carried on board. The engine room bilge pumps comprised one electric pump and one engine-driven pump with rated capacities of 11m3 and 13m3 per hour respectively. The bilge pumps in the fish hold and accommodation were also electric-powered with rated capacities of 11m3 and 6m3 per hour respectively. The on/off switches for the electric bilge pumps were located on a panel in the wheelhouse. Two bilge water level detectors were fitted in Majestic’s engine room; a ‘high’ level and a ‘low’ level. Bilge water level detectors were also fitted in the accommodation and the fish hold. Both the visual and aural alarms associated with the water level detectors were located in the wheelhouse. The engine room bilges were routinely pumped three or four times each day to remove accumulated water. The crew had not identified why this water had been accumulating. Surveys and maintenance The Maritime and Coastguard Agency (MCA) surveyed Majestic in 2012 after the vessel had grounded. The attending surveyor noted that repair was required to planking on the vessel’s sides and internal stiffeners. Other than that the vessel’s stability book was not on board at the time of the survey, no further deficiencies were identified. The MCA also completed a United Kingdom Fishing Vessel Certificate renewal survey on board Majestic on 17 August 2013 in Macduff shipyard, Scotland. During the survey, all sea valves were removed and were found to be in a satisfactory condition, as was the vessel’s non-ferrous seawater pipework. However, survey of the hull identified that extensive caulking was required in way of the forefoot.

9 The surveyor noted that routine drills were not recorded on board. A fire drill and an abandon ship drill were conducted during the survey. Lifejackets were worn for the abandon ship drill; Majestic carried six abandon ship lifejackets and four inflatable personal flotation devices (PFD). In April 2014, Majestic was taken out of the water for routine maintenance. The work carried out included: • Caulking and painting of the hull • Servicing of machinery • Replacement of the vivier tank. ANALYSIS Flooding mechanism It is evident that the flood into Majestic’s engine room started between 0900 and 1000 when the wheelhouse was unattended. That only the bilge alarms for the engine room were sounding when the skipper returned to the wheelhouse, indicates that, by then, the floodwater had not spread to a significant level in the fish hold or the accommodation. The source of the floodwater in Majestic’s engine room is unknown. Water ingress into the engine room bilges was a regular occurrence that required the bilges to be pumped frequently. However, this was probably due to degraded caulking in the vessel’s iroko planking. It is difficult to keep vessels made of iroko watertight as the wood is very hard and the caulking between planks quickly loses integrity. It is calculated that Majestic would have required to take on board 100m3 of water in order to lose stability. Therefore, as the vessel survived for between approximately 4 hours 40 minutes and 5 hours 40 minutes after the flooding started, the flooding rate must have been between 17.5 and 21m3/hr. This rate of water ingress was significantly higher than usual. Therefore, it was almost certainly due to a material failure of pipework or a seal on one of the engine room’s seawater systems connected to seawater inlets or discharge overboard valves, rather than the loss of caulking. Safety-critical alarms MGN 313(F)7 states that the wheelhouse must not be left unattended at any time, primarily to ensure that a proper lookout is maintained and that a vessel is navigated safely. However, the equipment installed in fishing vessels’ wheelhouses is not usually limited to navigation equipment and propulsion and steering control systems. Safety-critical alarms, such as fire and bilge-level alarms, are also fitted. Therefore, if a wheelhouse is left unattended, the risk of these alarms going unnoticed is significantly increased. On board Majestic, the purpose of the automatically activated bilge alarms was to provide the crew with an early notification of a flood. However, both the skipper and the crewman were working on deck and did not hear the bilge alarms over the noise of the deck hydraulics and machinery. As a result, the flood went unnoticed for up to 1 hour. Had the skipper and the crewman been alerted as soon as the bilge alarms sounded, the likelihood of them being able to enter the engine room to turn on the motor- driven bilge pump, and to identify and resolve the source of the water ingress would have increased significantly. Despite the guidance in MGN 313 (F) and numerous investigation reports, it is still common practice for fishing vessel wheelhouses to be left unattended. In the last year, the MAIB has published two reports into the loss of two vessels primarily because their wheelhouses were unmanned: the collision between 7 Keeping a Safe Navigational Watch on Fishing Vessels.

10 Good Intent (SY79) and Silver Dee (B310);8 and the fire and foundering of Karinya (FR699)9. In the latter case, the fire alarm in the wheelhouse could not be heard by the skipper, who was on deck operating winch controls. The desire to minimise manpower costs and the reported difficulties in recruiting fishermen often result in minimum crewing that makes it likely that the wheelhouses on board many fishing vessels will be left unattended during certain phases of the fishing process. In view of the need to maintain a safe navigational watch, such a situation cannot be condoned. However, as leaving wheelhouses unattended for short periods is likely to persist, it would be beneficial for safety-critical alarms to be audible throughout a vessel, and not only in the wheelhouse. In this respect, a two stage alarm similar to a navigation watch alarm, in which a general alarm sounds when an initial alarm has not been acknowledged within a defined time period, could be of benefit. Crew response The skipper’s quick decision to abandon ship after seeing the water level in the engine room, the immediate launch of the liferaft and the prompt activation of the DSC backed up by a “Mayday” broadcast, ensured the risk to the lives of the skipper and the crewman was reduced considerably. As a consequence, they were able to abandon Majestic in a controlled and safe manner and external assistance was provided very quickly. However, notwithstanding the skipper’s understandable reluctance to enter the engine room to investigate the source of the flood or to start the motor-driven bilge pump, several actions were not taken that potentially could have delayed the need to abandon and might have prevented the vessel’s loss. Notably, the electric bilge pump was not switched on and the vessel’s main engine was not stopped. With a capacity of 11m3/hr, the operation of the electric bilge pump would have reduced the flooding rate by about 50%. In addition, if, as is possible, the water ingress was caused by a failure connected with the engine’s cooling system, the ingress rate would have probably decreased further as soon as the engine was stopped. Onboard drills A factor that almost certainly reduced the skipper and crewman’s ability and willingness to tackle the flood in the engine room was that they did not routinely carry out emergency drills on board. MSN 1770 (F)10 requires fishing vessel crews to carry out at least one emergency drill each month. Drills are used to rehearse anticipated emergency scenarios, including flooding, and are intended to provide training, reduce confusion, and verify the adequacy of emergency response actions and equipment. The scope of flooding drills that can be conducted on vessels such as Majestic, that have few crew, is limited. Nevertheless, ‘table-top’ discussions covering the actions to take and the equipment available to deal with floods in different compartments would have assisted the skipper and the crewman with the development of ship-specific plans and procedures. Consequently, the importance of operating the electric-driven bilge pump and stopping the engine as soon as a flood in the engine room was discovered, might have been identified. Wearing of lifejackets That the skipper and the crewman did not wear PFDs when working on deck and did not don lifejackets when abandoning is of concern. A more rapid foundering, stumbling when entering the liferaft, or difficulties when deploying the liferaft could easily have resulted in one or both of the men ending up in the sea. 8 Collision between the fishing vessels Silver DeeV (B310) and Good Intent (SY79) resulting in the foundering of Silver Dee, Irish Sea, 29 July 2015, MAIB report 4/2016. 9 Fire and foundering of fishing vessel Karinya (FR 699), Moray Firth 4 October 2015, MAIB report 7/2016. 10 The Fishing Vessels Code of Safe Working Practice for the Construction and Use of 15 metre length overall (LOA) to less than 24 metre registered length (L) Fishing Vessels.

11 Despite the advantages of wearing a PFD when working on the open deck at sea, many fishermen still do not wear them. In 2000 the MAIB made its first recommendation about the compulsory wearing of lifejackets by fishermen working on deck. In the intervening years, there has been a succession of discussions, education programmes and research projects that have had very limited success. Since 2012, European funding has also been used by the various fishing industry associations to provide PFDs to thousands of fishermen free of charge. Despite these initiatives, the culture of the fishing industry has been slow to change, and fishermen continue to drown who might otherwise have lived had they been wearing a PFD when they entered the water. CONCLUSIONS • The flooding was probably caused by a failure within a seawater system inside the engine room. • The engine room had been flooding for up to 1 hour before the crew were alerted. • The engine room’s bilge alarms sounded in the wheelhouse and were not heard by the skipper and the crew who were working on the deck. • The prompt abandonment and notification to the coastguard and vessels in the area reduced the risk to the lives of the crew. • The rate of the flooding would have been slowed by the operation of the electric bilge pump and possibly by stopping the main engine. • The effectiveness of the crew’s response to the flood was probably reduced by not routinely conducting emergency drills. ACTION TAKEN MAIB actions The MAIB has issued a flyer to the fishing industry highlighting the safety issues identified in this report. RECOMMENDATIONS The owners of Majestic are recommended to: 2016/135 Take steps to ensure on any vessel they own or skipper in the future that: • The dangers of leaving the wheelhouse unattended are fully assessed so that practical measures can be adopted to mitigate such risks; • Emergency drills are conducted in accordance with MSN 1770 (F) and; • PFDs are worn by all crew when working on the open deck at sea. Safety recommendations shall in no case create a presumption of blame or liability

12 SHIP PARTICULARS Vessel’s name Majestic Flag United Kingdom Classification society Not applicable IMO number/fishing numbers LK678 Type Potter Registered owner Privately owned Manager(s) Not applicable Year of build 1977 Construction Wood Length overall 16.0 Registered length 15.98 Gross tonnage 51 Minimum safe manning Not applicable Authorised cargo Not applicable VOYAGE PARTICULARS Port of departure Toft Port of arrival Not applicable Type of voyage Coastal Cargo information Not applicable Manning 2 MARINE CASUALTY INFORMATION Date and time 21 January 2016 at 1443 Type of marine casualty or incident Very Serious Marine Casualty Location of incident 60°45’54N, 001°17’01W - 5nm NW of Yell, Shetland Place on board Ship / engine room Injuries/fatalities None Damage/environmental impact Total loss, no significant pollution Ship operation Fishing Voyage segment Mid water External & internal environment Moderate breeze, slight seas with a 1m well, good visibility and fine weather. Persons on board 2

FD06048 ｜ 2016_英国_man-overboard-from-twin-rig-trawler-aquarius-with-loss-of-1-life__MAIBInvReport18_2016.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2016 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H ACCIDENT REPORT VERY SERIOUS MARINE CASUALTY REPORT NO 18/2016 OCTOBER 2016 Report on the investigation of a fatal man overboard from the fishing vessel Aquarius (BF 89) east of Aberdeen on 17 August 2015

Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2016 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch Spring Place 105 Commercial Road Southampton Email: maib@dft.gsi.gov.uk United Kingdom Telephone: +44 (0) 23 8039 5500 SO15 1GH Fax: +44 (0) 23 8023 2459 Press enquiries during office hours: 01932 440015 Press enquiries out of hours: 020 7944 4292

CONTENTS GLOSSARY OF ABBREVIATIONS, ACRONYMS AND TERMS SYNOPSIS 1 SECTION 1 - FACTUAL INFORMATION 2 1.1 Particulars of Aquarius and accident 2 1.2 Narrative 3 1.3 Environmental conditions 9 1.4 Crew 9 1.5 Aquarius 10 1.5.1 General 10 1.5.2 Fishing gear 11 1.6 Crew working patterns 11 1.7 Marking the trawl warps 11 1.8 The rope stopper 13 1.9 The Code of Safe Working Practice for the Construction and Use of 15 to 24 Metre Fishing Vessels 13 1.10 Surveys and inspection requirements 16 1.10.1 Certification process for 15 to 24 metre fishing vessels 16 1.10.2 Annual self-certification 16 1.10.3 Offshore industry guard ship inspections 17 1.11 Survey and inspection performance record 17 1.12 Post-accident inspection 18 1.13 Management of deficiencies 18 1.14 Life-saving appliances 19 1.15 Vessel Safety Management 20 1.15.1 General duties 20 1.15.2 Safety management system 20 1.15.3 Emergency procedures and drills 21 1.15.4 Risk assessments 23 1.15.5 Work equipment inspections 23 1.15.6 Personal protective equipment and working lifejackets 24 1.15.7 Stability 25 1.16 Safe working practices 25 1.16.1 General safety guidance 25 1.16.2 Guidance provided by the International Maritime Organization 25 1.16.3 The Fishermen’s Safety Guide 26 1.16.4 Marine Guidance Notes 26 1.16.5 The Code of Safe Working Practices for Merchant Seamen 26 1.17 Hours of work and rest requirements 27 1.18 Previous incidents on Aquarius and similar accidents 28 1.18.1 Aquarius 28 1.18.2 UK fishing vessel fatal man overboard accidents 29 1.18.3 Annual self-certification 30 SECTION 2 - ANALYSIS 31 2.1 Aim 31 2.2 Overview 31 2.3 The failure of the rope stopper 31

1. 4 Working practices on board Aquarius

32 2.4.1 Marking the trawl 32 2.4.2 Standards of general seamanship 33 2.4.3 Leadership and teamwork 34 2.5 Man overboard 34 2.6 Emergency response 34 2.7 Emergency preparedness 35 2.8 The use of personal flotation devices on deck 36 2.9 Effects of tiredness and fatigue 37 2.10 Hours of work and rest 37 2.11 Vessel safety management 38 2.11.1 Vessel’s safety management system 38 2.11.2 Risk assessment 38 2.11.3 Safety culture and its effect on behaviours 39 2.12 Certification process and equipment maintenance 39 2.12.1 Vessel owner’s annual self-certification process 39 2.12.2 Regulatory oversight and the management of deficiencies 40 SECTION 3 - CONCLUSIONS 41 3.1 Safety issues directly contributing to the accident that have been addressed or resulted in recomendations 41 3.2 Safety issues not directly contributing to the accident that have been addressed or resulted in recommendations 41 SECTION 4 - ACTIONS TAKEN 43 4.1 MAIB Actions 43 4.2 Actions taken by other organisations 43 SECTION 5 – RECOMMENDATIONS 44

FIGURES Figure 1

Demersal trawl net arrangement Figure 2a

Stern view of Aquarius Figure 2b

Port trawl warp Figure 3a

Reconstruction of rope stopper applied Figure 3b

Reconstruction of rope stopper trapped in block after warp slackened off Figure 4

Plan view of deck with warp pulled inboard Figure 5

Reconstruction of marking the warp Figure 6

Positions of the deck crew when the stopper failed Figure 7

Track of Aquarius on departing Aberdeen Figure 8

Wheelhouse electronic displays Figure 9

Reconstruction of application of stopper hitch from cleat on port side to the warp Figure 10

Reconstruction of stopper trapped in the gantry block as warp is veered Figure 11

Endless chain available for use on Aquarius Figure 12

Chain stopper applied ANNEXES Annex A

Annang Nuerty (deceased) contract of employment Annex B

MGN 430 (F) – Checks on Crew Certification and Drills Annex C

Annual self-certification declaration form Annex D

Guard Vessel Inspection and Specification Document Annex E

Aquarius documented manoverboard recovery procedure Annex F

Aquarius logbook – muster and training records Annex G

Risk Assessment – General Working on Deck Annex H

Aquarius LOLER record of inspections Annex I

MAIB Safety Flyer

GLOSSARY OF ABBREVIATIONS, ACRONYMS AND TERMS °C

degrees Celsius CoSWP

Code of Safe Working Practices for Merchant Seamen DSC

Digital Selective Calling EPIRB

Emergency Position Indicating Radio Beacon FISG

Fishing Industry Safety Group GMDSS

Global Maritime Distress and Safety System HF

High Frequency IMO

International Maritime Organization ISO

International Organization for Standardization kg

kilogram kt(s)

knot(s) (nautical mile(s) per hour) kW

kilowatt L

Length LOA

Length Overall LOLER

The Merchant Shipping & Fishing Vessel (Lifting Operations and Lifting Equipment) Regulations 2006 LSA

Life-Saving Appliance m

metre MCA

Maritime and Coastguard Agency MF

Medium Frequency MGN

Marine Guidance Note mm

millimetre MOB

Man Overboard MSN

Merchant Shipping Notice N

Newton PFD

Personal Flotation Device PPE

Personal Protective Equipment PUWER

The Merchant Shipping & Fishing Vessel (Provision and Use of Work Equipment) Regulations 2006 RNLI

Royal National Lifeboat Institution

SAR

Search and Rescue

Seafish

Sea Fish Industry Authority SFF

Scottish Fishermen’s Federation SIAS

Ship Inspection And Survey SOLAS

International Convention for the Safety of Life at Sea 1974 t

tonne VHF

Very High Frequency

VTS

Vessel Traffic Services

UKFVC

United Kingdom Fishing Vessel Certificate UTC

Universal Co-ordinated Time Terms The Working Time - The Fishing Industry Code of Practice on Working Standards Code

Time Standards 15-24m FV Code

Code of Safe Working Practice for the Construction and Use of 15 metre length overall (LOA) to less than 24 metre registered length (L) Fishing Vessels TIMES: all times used in this report are UTC +1 unless otherwise stated

Image courtesy of Sean Boyce, MarineTraffic.com Aquarius

1 SYNOPSIS In the early hours of the morning on 17 August 2015, a fisherman on board the 20.8m twin rig trawler Aquarius was struck and thrown overboard violently by a steel wire trawl warp when a rope stopper parted. The vessel had just cleared the port of Aberdeen, Scotland, and its crew were attempting to re-mark the trawl warps when the accident happened. The skipper reacted quickly to stop the vessel, and the crew threw lifebuoys to the casualty. However, the recovery attempt was unsuccessful and the casualty sank out of view approximately 10 minutes later. Despite a search involving numerous vessels and a helicopter, the casualty’s body was not recovered. In order to mark the steel wire warp, the crew had streamed it over the stern. The stopper was used to take the strain of the trailing warp so that the crew could lower its inboard section on to the deck. The MAIB investigation established that: • The stopper parted under tension because a man-made fibre rope had been used instead of a chain, and because the way it had been applied deviated from well- established good practice. • The casualty was thrown overboard because he had positioned himself within the bight of the slackened trawl warp. • The crew were unable to recover the casualty back on board because neither they, nor their vessel had been adequately prepared to deal with such emergency situations. • The casualty’s body was not recovered because he was not wearing a lifejacket or other type of personal flotation device while working on the open deck. The Maritime and Coastguard Agency had surveyed and inspected Aquarius on numerous occasions during the previous 9 years. It had identified Aquarius as a poorly run vessel and issued it with 137 deficiencies; many of these related to safety management and were of a repetitive nature. The underlying factors that contributed to this accident included: a total lack of proactive safety management; a poor level of onboard safety culture; and the crew suffering from tiredness and fatigue. Recommendations have been made to the owners of Aquarius, the vessel’s manning agency and the Maritime and Coastguard Agency. These are aimed at improving the levels of safety management and emergency preparedness on board Aquarius; the working conditions and hours of rest for non-UK nationals on board UK flagged fishing vessels; and the capability of the electronic systems used by the Maritime and Coastguard Agency to manage and monitor deficiencies and poor performing fishing vessels.

2 SECTION 1 - FACTUAL INFORMATION 1.1 PARTICULARS OF AQUARIUS AND ACCIDENT SHIP PARTICULARS Vessel’s name Aquarius Flag UK Classification society Not applicable Fishing numbers BF 89 Type Stern trawler Registered owner MB Aquarius Ltd Manager(s) n/a Construction Steel Year of build 1994 Length overall 20.80m Gross tonnage 189t Main engine power 600kW Minimum safe manning Not applicable Authorised cargo Fish VOYAGE PARTICULARS Port of departure Aberdeen Intended port of arrival Aberdeen Type of voyage Fishing Manning 6 MARINE CASUALTY INFORMATION Date and time 17 August 2015 at about 0140 Type of marine casualty or incident Very Serious Marine Casualty Location of incident 2 miles east of Aberdeen harbour Place on board Aft main deck Injuries/fatalities One fatality Damage/environmental impact None Ship operation Single rig trawling for squid External & internal environment Westerly wind at 10kts; calm to slight sea; visibility good (night); sea temperature 13.8°C. Persons on board 6

3 1.2 NARRATIVE Shortly after midnight on Monday 17 August 2015, the UK registered fishing vessel Aquarius entered Aberdeen harbour, Scotland, and berthed on the fish quay. Once secure alongside, the vessel’s crew landed their catch of 47 boxes of squid. On completion, the skipper moved Aquarius to an adjacent quay to create space for another fishing vessel to come alongside. Before departure from Aberdeen, the skipper gathered his five-man crew together and informed them that he needed to re-mark the vessel’s steel wire trawl warps (Figure 1). He explained that this would be done once the boat was clear of the harbour, and that he wanted the warps to be marked at 125 fathoms1. At about 0100, the crew let the mooring ropes go and the skipper manoeuvred Aquarius away from the quayside. As the boat was manoeuvred out of the harbour the crew secured the trawl doors to the stern gantry and disconnected the port side trawl warp (Figures 2a and 2b). Once Aquarius was clear of the harbour entrance, the crew veered the port trawl warp winch drum, and lowered the warp over the stern and into the sea. The crew stopped the winch at the warp’s 100 fathom marker and waited for the skipper to come aft and measure the additional 25 fathoms. The skipper was busy navigating the vessel clear of harbour traffic, and directed Annang Nuertey, one of the vessel’s Ghanaian deck crew, to carry on and complete the job. Under Annang’s instruction, the warp was veered a further 25 fathoms and marked temporarily with adhesive tape. One of the crewmen then used a length of synthetic fibre rope to apply a stopper2 to the tensioned steel wire warp (Figures 3a and 3b). The winch was again veered and the inboard section of the warp became slack as the strain transferred to the rope stopper. With the weight of the streamed warp being taken by the rope stopper, the crew pulled the slackened section of warp inboard and laid it on the deck (Figure 4). Annang knelt on the deck outboard of the slackened warp and began to open its wire strands with a marlin spike; his intention was to insert a fibre rope marker (Figure 5). Another crew member knelt on the opposite side of the warp to assist in holding it securely. The three remaining crew members stood inboard of the warp towards the stern of the boat (Figure 6). At the same time, the skipper was gradually altering the vessel’s course to head south towards his intended fishing grounds, steaming at a speed over the ground of 5 to 6kts (Figure 7). At about 0140, the slackened warp suddenly snapped tight with a loud bang, striking Annang and catapulting him backwards over the boat’s port side and into the water. On hearing the loud bang the skipper put the engine to stop and ran out of the wheelhouse towards the main fishing deck aft. On his way, the skipper met a crew member running in the opposite direction to fetch a lifebuoy. The crew member was shouting “Annang overboard, Annang overboard”. The skipper passed a powerful torch to him with the instruction to shine it on Annang in the water. The skipper then returned to the wheelhouse and briefly put the boat’s engine astern. 1 A fathom is a unit of length equal to 6 feet (1.8m), chiefly used in reference to depth of water. 2 A stopper is a device that is used to temporarily take the weight off a rope that is under strain.

4 Image courtesy of Seafish Figure 1: Demersal trawl net arrangment

5 Figure 2a: Stern view of Aquarius Trawl doors Figure 2b: Port trawl warp Trawl warp

6 Figure 3a: Reconstruction of rope stopper applied Figure 3b: Reconstruction of rope stopper trapped in block after warp slackened off

7 Escape Up Down Escape Hatch (flush) Hatch (flush) × Figure 4: Plan view of deck with warp pulled inboard Figure 5: Reconstruction of marking the warp Trawl winches Stopper Trawl doors Net winches Trawl warp

8 Figure 6: Positions of the deck crew when the stopper failed Figure 7: Track of Aquarius on departing Aberdeen Reproduced from Admiralty Chart BA 1446 by permission of the Controller of HMSO and the UK Hydrographic Office. Position at time of accident Aquarius departure track Escape Up Down Escape Hatch (flush) Hatch (flush) × Trawl winches Stopper Trawl doors Net winches Key Annang (deceased) Filipino crewman Other crew

9 One of the crew threw a lifebuoy into the water, but it landed too far away from Annang for him to seize. At about 0143, the skipper manoeuvred Aquarius to place Annang directly astern, close to the vessel and within the area of sea illuminated by the vessel’s floodlights. Another lifebuoy was thrown, and this landed about 3 to 4m away from Annang, but he was unable to swim to it. The crew shouted encouragement to Annang as he treaded water; Annang made noises but did not respond coherently. Shortly afterwards, the deck crew saw Annang sink below the surface. He did not reappear. At 0155, the skipper called Aberdeen Vessel Traffic Services (VTS) on his mobile phone and reported that one of his crew had gone overboard. Aberdeen VTS then alerted the Aberdeen coastguard, which in turn transmitted a “Mayday” relay broadcast on very high frequency (VHF) radio channel 16. The Royal National Lifeboat Institution’s (RNLI) all-weather Severn Class lifeboat Bon Accord, and an inshore D Class lifeboat, were launched from Aberdeen at

1. The search and rescue (SAR) helicopter Bond 1 was also scrambled. The

coastguard co-ordinated a systematic search of the area for Annang, with numerous fishing vessels and oil support ships in the vicinity tasked to assist. At about 0400, the two lifebuoys thrown from Aquarius were recovered, along with one yellow Wellington boot of the type worn by Annang. Further searching proved unsuccessful, and the SAR effort was called off later that morning. Annang’s body was not found. 1.3 ENVIRONMENTAL CONDITIONS The accident occurred during the hours of darkness; the visibility was good and a light Westerly breeze was blowing at 10kts. The sea state was calm to slight and the sea temperature was recorded at 13.8°C. The charted depth of water where the accident occurred was about 30m and the surface of the seabed was shown on the chart as sand and gravel. 1.4 CREW Aquarius was certified to carry a crew of five, but was being operated with a crew of six at the time of the accident. The skipper of Aquarius was a 46 year old UK national and was also part owner of the vessel. He was a career fisherman and had qualified as Deck Officer (Fishing Vessel) Class 2 in 1999. He also held an Engineer Officer (Fishing Vessel) Class 2 qualification. Four of his five crew members were Ghanaian nationals; the fifth was Filipino. The Ghanaian crew were all supplied by the crewing agency PG Manning Ltd and employed on a seaman’s contract of employment (Annex A). The duration of the contract between the vessel and the crew members was 15 months (+/- 3 months on mutual consent). They lived on board Aquarius for the duration of their contract.

10 The crew members were all career fishermen and had completed the mandatory training3 required to work on board a UK registered fishing vessel. In addition to their basic safety training, all crew had completed the Sea Fish Industry Authority (Seafish) safety awareness for experienced fishermen course, the syllabus of which included accident prevention and risk assessment. None of the crew, with the exception of the skipper, had any navigation, radio communications or watchkeeping qualifications. Annang Nuertey was 47 years old. He had previously been employed in various fisheries around the world and had worked on several Scottish boats prior to joining Aquarius in January 2015. His command of the English language was above the local Ghanaian average. The other three Ghanaian crewmen were aged between 37 and 46 years old. Two had been on board Aquarius for over 6 months, the third had joined the vessel 3 weeks before the accident. Although the least experienced on board Aquarius, the third Ghanaian crewman had completed two previous contracts on UK registered fishing vessels, and had known Annang in Ghana. The Filipino crewman was 45 years old and had been working on Aquarius for 5 years. His command of the English language was good. The skipper had not nominated a deck crew leader, but typically passed his instructions to Annang or the Filipino crewman as they had the most fishing experience and the best command of the English language. 1.5 AQUARIUS 1.5.1 General Aquarius was a steel hulled stern trawler built in Buckie, Scotland, in 1994 and was originally named Crystal River. It was registered in Banff, Scotland, with the fishing vessel registration number BF 89, and its length overall (LOA) was 20.8m. The vessel was owned by MB Aquarius Ltd, a company which had four shareholders, one of whom was the skipper at the time of the accident. United Fish Selling Ltd, based in Buckie, provided agency services to Aquarius and several other local fishing vessels. Its services included the procurement of spare parts, quota management, liaison with authorities and crewing management. Aquarius carried a digital selective calling (DSC)4 enabled ICOM GM651 VHF radio set and a Tron 30s MkII emergency position indicating radio beacon (EPIRB). 3 New entry fishermen must complete basic safety courses in sea survival; first-aid; fire-fighting and fire prevention; and health and safety. Fishermen with 2 years’ experience must also complete a 1 day safety awareness and risk assessment course approved by Seafish, as required by the Maritime and Coastguard Agency. 4 A DSC-enabled VHF radio allows the operator to transmit a substantial amount of critical information, including the vessel’s position, to the coastguard and nearby vessels in an emergency by the pressing of a button and without the need for voice communication.

11 1.5.2 Fishing gear Aquarius operated mainly as a twin rig demersal trawler and fished principally for white fish such as cod, monkfish and haddock. Aquarius had reached its annual quota for white fish the week before the accident, and the skipper had re-rigged the vessel to catch squid. The rig consisted of a single net, two sets of sweeps, bridles and warps, and a pair of trawl doors (Figure 1). The length of the bridles/sweeps was shortened from 80 fathoms to 40 fathoms to fish in shallower waters, and the cod-end bag changed to a smaller mesh size. The trawl warps were 22mm diameter, right-hand ordinary lay, galvanised steel wire ropes. Each steel wire warp was spooled on to its own separate winch, with one warp connected to each trawl door. The net was hauled and shot from its own independent drum winch. 1.6 CREW WORKING PATTERNS Whenever possible, Aquarius was operated 24 hours a day, 7 days a week. Other than landing its catch, the vessel spent time in port only if it was under repair, weather-bound or unable to fish due to lack of available quota. The skipper shared a 10 day on and 10 day off working routine with a contracted skipper who had been working on Aquarius for a number of years. While the crew were on contract, there was no provision for them to take leave. In the months leading up to the accident, Aquarius had been fishing for white fish around the coast of Scotland. A typical working period involved 1 day steaming to the fishing grounds, 5 or more days fishing, and 1 day steaming back to port to land the catch. The fishing routine was a continuous cycle of shooting the nets, towing the gear for about 5 hours, hauling in the nets and processing the catch. The crew got broken rest periods while steaming to and from the fishing grounds. While fishing, the crew typically got 2 to 3 hours rest every 6 hours after processing the fish and before hauling in the next catch. When the skipper was resting, the deckhands would take turns to man the wheelhouse for periods of 2 hours. When the skipper began to fish for squid, the crew’s work routine changed. As squid was fished only during daylight hours and spoils quickly once caught, the working day was split between steaming to fishing grounds, repeatedly shooting/hauling nets, processing the catch, and steaming back to port to land the day’s catch. During the 24-hour period prior to the accident, the crew got 11 hours rest. This rest was split into multiple periods, most being about 1½ hours in duration, and none more than 4 hours in length. 1.7 MARKING THE TRAWL WARPS When fishing, the mouth of the net is kept open by trawl doors. To ensure that the fishing gear is towed at its optimum efficiency it is necessary to know how far apart the trawl doors are. Electronic trawl monitoring systems are often fitted on vessels to determine and monitor door spread, but nonetheless many fishermen still prefer to use traditional measurement and calculation methods, particularly on board smaller fishing vessels.

12 Aquarius was fitted with an electronic Rapp Marine PTS Pentagon winch control system and a Simrad ITI acoustic trawl monitoring system. Both systems provided computer display readouts in the wheelhouse (Figure 8). The skipper had limited confidence in his vessel’s electronic systems and preferred to physically mark his warps to give a visual reference of how much wire had been payed out. The skipper used strands of man-made fibre rope to mark his warps. To do this, the crew used marlin spikes to open up the steel wire warps and wove the markers between the warps’ steel wire strands. Due to the effects of warp stretch and splice repairs, the skipper typically re-measured and, where necessary, re-marked his warps every 2 months. This periodic process was usually carried out in port, where the crew could haul the warps along the quay to measure and re-mark them. During the 2-week period prior to the accident, the warps were measured and re-marked in a similar manner to that on the day of the accident, on three separate occasions with the vessel at sea. On the last occasion, 2 days before the accident, an additional mark was inserted into the warp at 100 fathoms in preparation to fish for squid. During one of the evolutions, the Filipino crewman raised safety concerns over the way in which the task was being conducted. However, his warnings were ignored and he was aggressively rebuked by his Ghanaian crew mates. Figure 8: Wheelhouse electronic displays ITI SIMRAD unchanged from twin rig configuration PTS Pentagon showing single rig configuration

13 1.8 THE ROPE STOPPER The rope used by the crew to stopper the warp was a three stranded, 14mm diameter PolysteelTM man-made fibre rope. Its tensile strength5 was 3.8t. The length of rope used for the task was cut from a new coil a few days earlier, and had been used as a stopper on two previous occasions. The manner in which the stopper was applied on those occasions was similar to that on the day of the accident. Following the accident, remnants of the rope stopper were found on the gantry cleat. These were removed by the crew because they wanted to use the cleat to moor the vessel alongside. The remnants of the rope were subsequently disposed of overboard prior to the vessel’s arrival back into Aberdeen. During the MAIB inspectors’ initial inspection of the accident site, the deck crew were asked to demonstrate how the stopper had been applied. It was observed that the crewman initially secured one end of the stopper rope to a cleat on the port side of the stern gantry (Figure 9). He then looped the other end of the rope over the warp and around the cleat three times before tying it off around the warp with a series of half hitches. When the warp winch was veered, it was noted that the stopper hitch moved aft and entered the gantry block (Figure 10). The steel wire warp was then seen to slip a short distance through the hitch before the stopper tightened and took the strain. 1.9 THE CODE OF SAFE WORKING PRACTICE FOR THE CONSTRUCTION AND USE OF 15 TO 24 METRE FISHING VESSELS Aquarius was required to comply with the Code of Safe Working Practice for the Construction and Use of 15 metre length overall (LOA) to less than 24 metre registered length (L) Fishing Vessels (15-24m FV Code)6. The aim of the Code was to set standards of safety and protection for all staff on board fishing vessels. It set minimum standards for construction, machinery, equipment and also stability. To comply with the 15-24m FV Code, the vessel owners were responsible for ensuring that the vessel: • Was built, equipped, surveyed, certified and maintained and operated in accordance with the relevant provisions of the Code. • Was subjected to annual self-certification inspections. • Continued to comply with the requirements of the Code in service. • Was operated by appropriately qualified and certificated crew who had completed mandatory training courses. • Was not operated as a fishing vessel without a valid UK fishing vessel certificate being in force. The full text of the 15-24m FV Code was set out by the Maritime and Coastguard Agency (MCA) in its Merchant Shipping Notice (MSN) 1770 (F) 5 The tensile strength is the load at which a new rope, tested under laboratory conditions, can be expected to break. 6 The Fishing Vessels (Safety of 15-24 Metre Vessels) Regulations 2002 gave statutory force to the 15-24m FV Code.

14 Figure 9: Reconstruction of application of the stopper hitch from cleat on port side to the warp

15 Figure 10: Reconstruction of stopper trapped in the gantry block as warp is veered

16 1.10 SURVEYS AND INSPECTION REQUIREMENTS 1.10.1 Certification process for 15 to 24 metre fishing vessels The MCA issues UK fishing vessel certificates that routinely are valid for 5 years. Before a certificate is issued or renewed, fishing vessels are surveyed by appointed surveyors to verify compliance with the 15-24m FV Code and applicable legislation. In order to verify ongoing compliance with the Code, an intermediate inspection by an appointed surveyor must be carried out between the 2nd and 3rd anniversary of the certificate issue/renewal date. Certificates may be issued for periods of less than 5 years if the surveyor considers that the standard of the vessel requires an enhanced level of scrutiny. Fishing vessel owners are required to present their vessels for survey within the prescribed time frames or prior to undertaking any major repairs or modifications. Aquarius’s fishing vessel certificate had been renewed on 29 January 2013 and was valid until 18 December 2017. For the certificate to remain valid an intermediate inspection was required between 28 December 2014 and 28 December 2015. Although not overdue, this intermediate inspection had not been carried out prior to the accident. The MCA’s Marine Guidance Note (MGN) 430 (F) Checks on Crew Certification and Drills (Annex B) provided information to fishing vessel owners and skippers on the checks its surveyors will make regarding crew certification and the conduct of emergency drills during vessel surveys and inspections. The guidance explained that surveyors will: • ensure that the correct certificates of competency are held and safety training courses have been undertaken by skippers and crew; • check that written health and safety policies are in place and completed risk assessments have been carried out; • witness emergency drills as part of the renewal and intermediate surveys on the vessel or at any other time as deemed necessary by the MCA; • confirm that emergency drills (fire, collision/grounding, man overboard, abandon ship, anchoring) are practiced monthly and when a new crew member joins the vessel; • if practicable, and when there is no evidence that drills have been conducted and it is considered the crew are not trained for an emergency, ask vessels to proceed from the harbour to a safe anchorage to undertake anchoring drills. This increases the validity of the drill and provides a more challenging, realistic environment. 1.10.2 Annual self-certification In addition to the survey and inspection requirements detailed in paragraph 1.10.1, Aquarius’s owners or their delegated representative were required to check the vessel annually to ensure that: i. all firefighting appliances, lifesaving appliances and safety equipment that are carried on board the vessel have been suitably maintained and are within date;

17 ii. the radio equipment is functioning correctly; iii. the shipborne navigational equipment, nautical publications and lights, shapes and sound signal appliances, that are required for compliance with the Collision Regulations, are carried on board and are functioning correctly; iv. the risk assessment remains appropriate to the vessel’s fishing method and mode of operation; v. no known alteration, damage or deterioration to the vessel or its equipment has occurred in service that would affect the vessel’s compliance with the requirements of the Code or the vessel’s stability; vi. weathertight doors and hatches are functioning correctly; and vii. crew training and certification are valid. On completion of each annual check, owners were required to sign declarations confirming compliance with the mandated standards, and to retain them on board for subsequent inspections. Self-declaration forms (Annex C) were provided for use by fishing vessel owners in Annex 2 of MSN 1770 (F). The skipper of Aquarius could not produce any self-declaration forms and there was no evidence that the annual verification checks had been carried out. This fact had been identified during the MCA’s post-accident inspection of the vessel, and recorded as a safety-critical deficiency in the inspection report. MCA surveyors had also identified this omission during several previous surveys and inspections. 1.10.3 Offshore industry guard ship inspections Guard ships are employed to patrol and protect valuable offshore assets in the North Sea. In particular, they are used to keep fishing vessels and, where applicable, other marine traffic clear of vulnerable offshore assets. In order to generate additional income, fishing vessel owners could apply for a guard ship contract. Fishing vessels used for guard ship duties were required to hold a Load Line Exemption Certificate, and meet the safety specifications set out by SFF Services Ltd in its Guard Vessel Inspection and Specification Document (Annex D). In order to achieve this, participating fishing vessels were required to be inspected annually. Aquarius was presented annually for guard ship inspections between 2011 and

1. The vessel was never contracted to conduct guard ship duties.
2. 11

SURVEY AND INSPECTION PERFORMANCE RECORD In the 9 years prior to the accident, Aquarius was surveyed and/or inspected by MCA surveyors on nine separate occasions. These were conducted for the maintenance of Aquarius’s fishing vessel certificate and verification of compliance with the guard ship safety specification. The MCA surveyors recorded 137 deficiencies during these nine surveys and inspections.

18 Of the deficiencies recorded, 65% were classified as safety-critical and needed to be rectified before the vessel’s departure from the port of survey. During this period, the MCA’s visiting surveyors held frequent discussions with the skipper concerning the repeated nature of many of the deficiencies identified, the lack of risk assessments and safety drills. On 21 March 2013, the MCA conducted a targeted inspection that focused on crew certification and competency following concerns raised by the Scottish Fishery Protection Agency when it boarded the vessel at sea on 15 March 2013. 1.12 POST-ACCIDENT INSPECTION On 18 August 2015, a targeted post-accident general inspection of Aquarius and its safety equipment was carried out in Aberdeen by an MCA surveyor. During the surveyor’s initial sampling process he identified so many safety-critical deficiencies that he suspended the inspection and detained the vessel. At the skipper’s request, the surveyor returned to the vessel on 21 August to conduct a release from detention inspection. This was required in order that the vessel be allowed to proceed to the port of MacDuff for repairs. During that inspection, the surveyor recorded 23 deficiencies, two of which needed to be rectified before the vessel proceeded to sea, and 18 required action to be taken at the next port. On 26 August 2015, an MCA surveyor carried out an intermediate fishing vessel certificate inspection in MacDuff. Following that inspection, Aquarius was put to sea without rectifying the safety-critical deficiencies identified by the MCA surveyor. The vessel was subsequently detained on its return to the harbour at the end of August. 1.13 MANAGEMENT OF DEFICIENCIES On completion of each vessel survey or inspection the vessel details and the deficiencies identified were recorded by the MCA surveyor on a paper report form7; a copy of the report was passed to the vessel owner/skipper. The content of the inspection report form was later entered into the MCA’s electronic Ship Inspection And Survey (SIAS) database. When outstanding deficiencies were rectified, vessel owners were required to inform their local MCA marine office. The marine office staff would then update the information held within the SIAS database. For deficiencies identified on board Aquarius, the vessel’s agent (United Fish Selling), assumed the responsibility of reporting to the regulator that they had been rectified. The SIAS database could be interrogated to provide reports of outstanding deficiencies for individual vessels, but there was no facility within the database to alert surveyors when a deficiency had passed its required action date. The surveyors in the Aberdeen marine office typically monitored the status of deficiencies by reviewing the paper files of individual vessels. An attempt was made within the Aberdeen marine office to proactively identify fishing vessels with a poor safety record. The aim of this project was to concentrate the surveyor resources on a few targeted vessels to improve their safety management. The project had identified Aquarius as one of the top 10 vessels to be targeted in the region. 7 MCA MSF 1602 and 1603 forms - Report of Inspection and/or Survey of United Kingdom/ Dependent Territory/ Foreign Vessel.

19 1.14 LIFE-SAVING APPLIANCES The requirements for Aquarius’s life-saving appliances (LSA) were set out in Chapter 7 of the 15-24m FV Code. As a minimum, the vessel was required to carry: • At least two liferafts8. • A lifejacket for each person on board, plus an additional two lifejackets. • At least two lifebuoys, one of which should be provided with a self-igniting light and self-activating smoke signal, and the other provided with a buoyant line of at least 18m in length. • A means of recovering a person from the water. • A line-throwing appliance. • Six red rocket parachute flares, two buoyant orange smoke signals and four red hand flares. • One hand-held VHF radio. • One float-free9 satellite EPIRB. Liferafts should be readily available for safe and rapid use in an emergency. They should also be stowed in such a manner that enables them to float free from their stowage, inflate and break free from the vessel in the event of its sinking. During the MCA’s post-accident inspection on 18 August 2015, the port liferaft on board Aquarius was found to be rigged incorrectly and would not have deployed automatically as required. For a crew of six, Aquarius should have carried eight emergency use lifejackets, but the MCA surveyor found only five on board during his post-accident inspection, two of which did not have the required lights. There were three lifebuoys on board, one of which was fitted with a light and smoke signal, but they did not have the required standard of retroreflective tape and markings. The lifebuoy fitted with the self-igniting light and self-activating smoke signal was not easily accessible because a pilot ladder had been stowed and lashed down in front of it. This lifebuoy was not used during the manoverboard rescue attempt. Aquarius did not carry a dedicated means of recovering a person from the water. The MCA’s record of particulars for Aquarius, dated 29 January 2013, listed Patching and Line as the vessel’s manoverboard recovery equipment. The skipper and his crew were unable to explain what was meant by Patching and Line. Aquarius was equipped with the distress flares required by the 15-24m FV Code, however the inspection found that the service life had expired in July 2014 (13 months before the accident); the boat’s medical kit was also out of date. The vessel’s Global Maritime Distress and Safety System (GMDSS) logbook had not been maintained as required. The EPIRB’s hydrostatic release unit expired in September 2014, and its battery life expired 4 months before the accident. There were also no records to indicate that the EPIRB had been tested at monthly intervals. 8 The liferafts had to meet the requirements of SOLAS 1974 as amended. 9 Float-free in relation to LSA means that the appliance is automatically released from a sinking vessel and is ready for use.

20 1.15 VESSEL SAFETY MANAGEMENT 1.15.1 General duties In accordance with Regulation 5 of the Merchant Shipping and Fishing Vessels (Health and Safety at Work) Regulations 1997, an employer10 has a general duty to: ‘ensure the health and safety of workers11 and other persons so far as is reasonably practicable’. In order to fulfil their general duties, the owners of Aquarius were required to endeavour to: • Avoid or minimise risks • Evaluate unavoidable risks and take actions to minimise them, and • Adopt safe work patterns and procedures. In practice, the day-to-day management of the shipboard operations and safety was under the control of the skipper. 1.15.2 Safety management system Seafish provided safety management guidance for the UK fishing industry on its website, and reminded fishing vessel owners that: Vessels are required to be operated in a safe manner, it is the responsibility of the owners/skippers of fishing vessels to ensure that they are compliant with the The Merchant Shipping and Fishing Vessels (Health and Safety at Work) regulations 1997. In order to help fishing vessel owners and skippers manage their vessels safely, and to comply with their regulatory obligations, Seafish created a structured safety management folder (safety folder). The safety folder was aimed at helping skippers/ owners to produce a management system for their vessel and contained examples of risk assessments for different types of fishing operations, as well as a vessel safety policy template, and vessel safety checklists. The safety folder was available in hard copy and was supplied to fishermen attending the Seafish safety awareness course. Alternatively, it could be completed and maintained online12. The safety folder included checklists, templates and forms for: • Vessel safety policy statement • Emergency procedures and drills • Crew details and induction training • Muster plan 10 Employer means a person by whom a worker is employed under a contract of employment. 11 Worker means any person employed by an employer under a contract of employment, including trainees or apprentices. 12 www.safetyfolder.co.uk

21 • Risk assessments • Records of equipment inspections • Stability. A paper copy of the Seafish safety folder was carried on board Aquarius. The safety folder’s initial entries were made in late 2012 and early 2013. These entries were coincident with the fishing vessel certificate renewal survey conducted on 29 January 2013. The vessel owners’ safety policy statement included their intent to comply with all relevant health and safety legislation. It also included a commitment to review the vessel’s equipment registers, safety procedures and risk assessments every 12 months, or following a significant event. The folder and the vessel’s risk assessments had not been reviewed or updated since their initial compilation, and there was no other safety management system in use on board. 1.15.3 Emergency procedures and drills Section 8.1.2.1 of the 15-24m FV Code stated that: The skipper should ensure that the crew are trained in the use of all lifesaving and fire appliances and equipment with which the vessel is provided and should ensure that all members of the crew know where the equipment is stowed. Such training should be carried out in drills, held in port or at sea, at intervals of not more than one month. For a vessel with a crew of five or more, a muster list is required containing clear instructions for each crew member to follow in the event of an emergency. There was no muster list posted on board Aquarius. The emergency procedures listed in the Seafish safety folder included man overboard; location and recovery. The MCA provided generic guidance on the conduct of emergency drills on board fishing vessels for surveyors, skippers and crew in an annex to MGN 430 (F).13 The guidance for manoverboard included: Crew general: • Sound crew alarm • Crew to muster stations with warm clothing and lifejackets donned correctly • Was the crewman seen to have fallen overboard? Yes / No • If yes, throw lifebuoy with smoke / light float attached to mark position • If no, note time and position and consider using smoke / light float anyway to mark a datum position. This will give a visual marker to searching vessels and aircraft as an indication of the tide and surface water movement in the search area from a given time. 13 MGN 430 (F) – Checks on Crew Certification and Drills (see Annex D)

22 Skipper: • Press MOB function on navigational aid, if fitted and crew alarm • Inform ships mate with all details • Send DSC alert and commence voice transmission on VHF, MF or HF as appropriate • Inform Coastguard of any updated information and the description of missing person • Haul fishing gear if fishing • If not fishing and navigation allows commence Williamson turn • Pass any additional information to the Coastguard and any other vessels assisting • Keep all search units up to date by sending situation reports regularly. Mate: • Crew muster report to skipper, crewman missing? Yes / No • Collect details of missing crewman and pass to skipper as soon as possible, time and place last seen, clothing type and colours, age and state of health • Ensure crew are dressed appropriately • Post lookouts forward, wing of the bridge port and starboard also on a high point aft • Search vessel for missing crewman and prepare to launch rescue boat if carried • Organise man overboard recovery system • Organise dry clothing and first aid equipment, prepare to treat for hypothermia. Guidance regarding the use of the manoverboard recovery systems emphasized that crews should be well trained in their use, and appreciate the limitations of their use in poor weather conditions as well as fine weather. Aquarius’s manoverboard recovery procedure (Annex E), as stated in its safety folder, was: Our procedure for a man overboard is to keep a visual on the man while the others prepare a ring and line. The best location for retrieving a man is on the stern where it’s the lowest point and we have 2 winches to haul a person on board very easily. [sic]

23 The vessel’s safety folder also contained a log sheet to record the conduct of emergency drills at monthly intervals. These log sheets had not been used and there were no records of any emergency drills having been conducted within the vessel’s logbook. The last recorded safety drill was carried out on board Aquarius during an MCA inspection in January 2013 (Annex F). The Ghanaian crew members’ details had not been entered into the relevant pages in the safety folder, and there was no record of them having received their induction training. They had not carried out any manoverboard drills, or undertaken any structured emergency training since joining the vessel. 1.15.4 Risk assessments The Merchant Shipping and Fishing Vessels (Health and Safety at Work) Regulations 1997 required that fishing vessel owners risk assess their vessels’ working operations, and review the assessments at regular intervals. MSN 1770 (F) incorporated this requirement and stated that: ‘A health and safety risk assessment should be used to satisfy the obligation of providing information to crew members of the measures taken for their own protection.’ The risk assessments contained within the safety folder on board Aquarius included one for general working on deck (Annex G). The hazards identified for work activities on deck included handling fishing gear, falling overboard and sudden capsize or loss of vessel. The consequences of falling overboard and sudden capsize or loss of vessel were drowning and death, and the control measures prescribed in the risk assessment were: • When working with the risk of M/O wear safety harness [sic] • Lifejackets worn when working on deck [sic] The crew were not familiar with the vessel’s documented risk assessments, and there was no risk assessment or safety procedure for the task of marking the warps. 1.15.5 Work equipment inspections The Merchant Shipping & Fishing Vessel (Provision and Use of Work Equipment) Regulations 2006 (PUWER), and the Merchant Shipping & Fishing Vessel (Lifting Operations and Lifting Equipment) Regulations 2006 (LOLER) apply to all UK registered fishing vessels. All tools, machinery, and equipment used at work are covered by the requirements of PUWER. This includes everything from a trawl winch to a gutting knife. PUWER requires that work equipment is suitable for the work to be carried out, or is properly adapted for that purpose, and may be used by workers without impairment to their health or safety.

24 Work equipment that is also lifting equipment will additionally be subject to the requirements of LOLER. Both regulations require that equipment be identified, maintained and inspected. Equipment used for lifting under LOLER regulations is subject to mandatory inspection routines and examination. The LOLER block accounting system contained in the Seafish safety management folder on Aquarius had not been updated for some considerable time (Annex H) and, along with the lack of crane wire certification, had been noted as deficiencies in previous MCA surveys. 1.15.6 Personal protective equipment and working lifejackets The Merchant Shipping and Fishing Vessels Personal Protective Equipment Regulations 1999 require employers to provide personal protective equipment (PPE) for their workers when they are engaged in, or at risk from, a hazardous work activity on board a UK registered vessel. Regulation 10 requires that: “The employer shall take all reasonably practicable steps to ensure that any personal protective equipment provided to workers under regulation 6(1) is used as instructed.” In circumstances where there is a foreseeable risk of crew falling overboard, the recognised PPE includes a safety harness and lanyard, a personal flotation device14 (PFD), and possibly also the use of a thermally-insulated immersion suit as appropriate. Aquarius’s owners had recently provided eight Parmaris 150N PFDs for use by the crew while working on deck. This was the direct result of lessons learned following a recent fatal man overboard incident on board another local fishing vessel, Beryl (see paragraph 1.18.2). These PFDs incorporated automatic and manual inflation, and complied with international safety standards15. Each member of the deck crew was given one of the PFDs and had received instruction on its use. The crew regularly wore their PFDs on deck during fishing operations, but on occasions the skipper had observed them working on deck without them. The skipper did not routinely enforce the requirement for the crew to wear PFDs when they were working on deck. 14 PFDs are divided into the following two main classes: those that provide face up in-water support to the user regardless of physical conditions (lifejackets); and those that require the user to make swimming and other postural movements to position their face out of the water (buoyancy aids). The selection of the appropriate PFD is dependent on the task undertaken and the environment in which it is conducted. PFDs that do not require intervention, such as auto-inflation lifejackets, are suited to activities where persons are likely to enter the water unexpectedly. 15 ISO 12402-3:2006 – Personal flotation devices – Part 3: Lifejackets, performance level 150 – Safety requirements.

25 1.15.7 Stability Aquarius’s stability booklet was originally approved by the MCA on 26 January 1999. On 29 January 2013 Aquarius underwent a lightship16 stability check. During this process it was found that an additional 2.5t of fishing gear was carried on board. This caused the vessel to fail certain stability safety criteria. Section 3.1.3.3 of the 15-24m FV Code stated: The carriage of unnecessary spare gear, stores and parts, the accumulation of debris and the cumulative effects of minor modifications over time can adversely affect the vessel’s lightship weight and centre of gravity. Attention should be made to limiting these effects if lightship growth and the possibility of adverse effects on the vessel’s stability are to be avoided. To make the vessel safe, the skipper was required to remove the additional 2.5t of fishing gear immediately. This was confirmed in a declaration signed and dated on 22 December 2014, by the vessel’s agent, as having being completed. During the MCA’s post-accident survey an additional 2.5t of fishing gear was again found on board. 1.16 SAFE WORKING PRACTICES 1.16.1 General safety guidance Chapter 6 of the 15-24m FV Code contained generic safety guidelines for the protection of the crew, but did not contain detailed guidance on safe working practices and general seamanship. In order to help inform fishing vessel owners and skippers of best practice, methods of mitigating hazards and reducing risks, the International Maritime Organization (IMO), the MCA and fishing industry bodies provided a variety of guidance and information. 1.16.2 Guidance provided by the International Maritime Organization The IMO provided international guidance on safety and health practice for fishermen, and safety and health requirements for the construction and equipment of fishing vessels in its Code of Safety for Fishermen and Fishing Vessels 2005. Although the IMO’s Code of Safety did not specifically discuss the task being attempted on board Aquarius when the stopper failed and Annang was thrown overboard, it did give the following warnings: • During the handling of mooring lines or other wires or ropes, care should be taken not to stand in the bights • Crew should pay special attention to ropes connected to the net such as bridles, false headlines, etc., when shooting. They should never stand in the bights of such ropes 16 Lightship is the weight of the vessel with no fuel, cargo, water, fishing gear or stores

26 1.16.3 The Fishermen’s Safety Guide The MCA’s Fishermen’s Safety Guide, originally written in 2008 and updated in May 2015, was endorsed by the UK Fishing Industry Safety Group (FISG). The guide provided a broad range of advice in respect of fishing safety, and its aim was to provide guidance and promote safety awareness to all. Section 3 of the guide contained the following warning: Do not stand on a slack warp laid on the deck; if the ‘stopper’ chain slips, it may suddenly become tight, throwing you up and perhaps overboard. The Fishermen’s Safety Guide is freely available to download from the internet: Aquarius did not have a copy on board. 1.16.4 Marine Guidance Notes The MCA’s MGN 415 (F), Fishing Vessels: The Hazards Associated with Trawling, Including Beam Trawling and Scallop Dredging was published in 2010. The guidance provided did not directly relate to the task being conducted at the time of the accident although it did provide warnings, similar to those given by the IMO, about the danger of standing in the bights of ropes. 1.16.5 The Code of Safe Working Practices for Merchant Seamen The MCA produced the Code of Safe Working Practices for Merchant Seamen (CoSWP)17 primarily for the use of merchant seamen on United Kingdom registered vessels. Although not required to be carried on board fishing vessels, CoSWP is an internationally recognised standard of good seamanship practices and is freely available to download from the internet. Chapters 21.17 and 26.3.18 of CoSWP provided guidance on the use of stoppers. Although much of the guidance referred specifically to mooring ropes, the same principles apply to stoppering any rope or wire under tension. The guidance stated that: • Natural fibre rope should be stoppered with natural fibre. • Man-made fibre rope should be stoppered with man-made fibre stopper (but not polyamide). • The ‘West Country’ method (double and reverse stoppering) is preferable for ropes. • Wire moorings should be stoppered with chain, using two half-hitches in the form of a cow hitch, suitably spaced with the tail backed up against the lay of wire, to ensure that the chain neither jams nor opens up the lay of the wire. This information was repeated in MGN 308 (M+F) Mooring, Towing or Hauling Equipment on all vessels – safe installation and safe operation. 17 The MCA issued a revised version of CoSWP on 4 September 2015 and renamed it The Code of Safe Working Practices for Merchant Seafarers. This revised version of CoSWP contained similar guidance.

27 An endless chain (Figure 11) for stoppering the trawl warps was carried on board Aquarius and had been used by the skipper and his crew on numerous occasions previously. 1.17 HOURS OF WORK AND REST REQUIREMENTS The Working Time: Sea Fishermen Regulations 2004 set out the working time limits and the amount of rest and annual leave to which workers on board UK registered fishing vessels were entitled. The MCA provided guidance on the application of the working time regulations in MSN 1786 (F)18. The regulations specified that: • A worker’s working time shall not exceed 48 hours per 7-day period averaged out over a year. • A worker is entitled to adequate rest, and the total hours of rest should not be less than 10 hours in any 24-hour period and 77 hours in any 7-day period. • A worker is entitled to paid annual leave of at least 4 weeks. • A skipper can require a worker to work any hours in an emergency. For objective or technical reasons, or for reasons having to do with the organisation of the work, the working time standards on board certain types of fishing vessels might be unachievable. In such cases, while the standards remain as the benchmark, exceptions to the limits may be allowed provided that the general 18 MSN 1786 (F): Application of the Fishing Vessels (Working Time: Sea-fishermen) Regulations 2004 Figure 11: Endless chain available for use on Aquarius

28 principles of the health and safety of the workers are respected. Such exceptions should take account of more frequent or longer leave periods or the granting of compensatory leave. The Fishing Industry Code of Practice on Working Time Standards (Working Time Standards Code), which was contained as an annex to MSN 1786 (F), was a Government-approved exception that applied to certain types of fishing vessels. This exception applied to Aquarius as a white fish trawler provided that compensatory rest was given to the crew to offset those occasions when the standards set out in Clause 7 of the Working Time Standards Code were not met. In order to comply with the working time regulations, employers were required to keep records that adequately demonstrated that employed sea-fishermen were receiving the minimum rest to which they were entitled. The regulations also required these records be retained for a period of 2 years and be made available for inspection. No records of rest were kept on board Aquarius. The contract of employment (Annex A) drafted by PG Manning Ltd for the Ghanaian crew of Aquarius, was between the vessel and the crew member. It made no mention of the applicable working time regulations, maximum working hours, minimum rest periods, compensatory arrangements or the requirement for paid annual leave. 1.18 PREVIOUS INCIDENTS ON AQUARIUS AND SIMILAR ACCIDENTS 1.18.1 Aquarius The MAIB accident information database contains records of seven incidents involving Aquarius. These include: • July 2013, and again in February 2015, Aquarius was towed back to port following an engine failure. • January 2010, Aquarius suffered a control system failure while crossing Lerwick harbour, and was towed alongside by the local lifeboat. • October 2009, the vessel’s wheelhouse was flooded when the windows were smashed by heavy seas. • January 2008, the vessel suffered partial flooding in the engine room as a result of a leaking stern gland. This was due to poor maintenance after the gland had been repacked but not tightened. The vessel was towed into port at Scrabster. • December 2005, Aquarius was towed into port by the lifeboat after a fuel tank valve failed, causing the vessel to run out of fuel. • August 2005 Aquarius collided with another fishing vessel, Fertile II, while pair trawling; Fertile II sank. The Chief Inspector of Marine Accidents wrote to Aquarius’s skipper recommending that he improve the standard of lookout, communications and maintenance procedures on board his vessel.

29 In addition to the reported accidents recorded on the MAIB database, the following incidents occurred: • October 2014, the skipper was fined for ‘intentionally obstructing a sea fishery officer’. • August 2014, the vessel’s stand-by skipper was fined for illegally landing fish and falsifying records. • March 2013, the owners of Aquarius were issued with an improvement notice for their failure to have a qualified GMDSS operator on board. • January 2013, the skipper was issued an official warning for operating Aquarius without a valid UK fishing vessel certificate. 1.18.2 UK fishing vessel fatal man overboard accidents From the period 1992 up to and including this accident, there have been 117 fatal man overboard accidents on UK registered fishing vessels. On 9 July 2015 the skipper of the twin rig stern trawler Enterprise fell overboard through one of the fishing net shooting ports when the vessel lurched in heavy seas. Although all of the other crew were wearing PFDs, he was not. It was estimated that he was in the water for 30 to 40 minutes before being recovered on board. Despite resuscitation efforts by the crew he did not survive. On 10 February 2015 a crewman on board the 28m twin rig trawler Beryl was carried overboard by the vessel’s port trawl net. The crewman was carried through the port shooting door and into the sea. He was conscious and managed to hold onto the net. The crewman was wearing a PFD and it inflated when he entered the water. Beryl’s crew spent almost 50 minutes trying to recover him back on board, but they were unsuccessful. The crewman was eventually recovered by a rescue craft launched from an offshore support vessel. He was transferred to a rescue helicopter and flown to hospital, but he did not survive. On 29 January 2012 a crewman was swept overboard from the twin rig trawler Zenith while standing on top of the aft bulwark rail during hauling operations. The vessel was quickly manoeuvred to bring the casualty alongside, but the crew were unable to recover him back on board and he slipped from their reach and was lost. He was not wearing a PFD. On 11 November 2009 a Filipino crewman was dragged into the sea from the deck of the fishing vessel Osprey by a net running over the side. Despite his crew mates’ attempts to recover him using a line and a lifebuoy, the crewman disappeared from the surface after about 12 minutes and was never recovered. He was not wearing a PFD. On 13 August 2008 a crewman was swept overboard from the twin rig trawler New Dawn by a towing chain while shooting gear. The skipper jumped into the sea to help the crewman but quickly began to succumb to the effects of the cold water. After some difficulty the skipper was recovered. Unfortunately, the crewman, who was not wearing a PFD, was not recovered.

30 On 3 September 2007 a crewman on board the stern trawler Apollo died after being dragged overboard by a trawl net during hauling. The vessel was quickly manoeuvred alongside the casualty, who grabbed hold of a life-ring thrown to him by his crew mates, and was pulled alongside. Once alongside he began to lose consciousness and, despite his crew mates’ efforts, was unable to be revived when he was recovered to the deck. The casualty, who had not been wearing a PFD, was in the water for approximately 15 minutes before his final recovery to the deck. The Chief Inspector of Marine Accidents wrote to the Chief Executive of the MCA suggesting that he consider extending an initiative for MCA’s surveyors to assist fishermen in their assessment of the risks encountered during fishing operations. 1.18.3 Annual self-certification On 30 March 2014 the owner/skipper of the 10m scallop dredger Ronan Orla was fatally injured when he became entangled on the warping drum of the vessel’s winch. One of the findings from the MAIB investigation was that self-certification, required to confirm the status of safety equipment, had not been carried out. A recommendation was made to the MCA regarding this within the report 12/2015: 2015/129 Amend Marine Guidance Note 502(F) The Code of Practice for the Safety of Small Fishing Vessels to require owners of under 24m fishing vessels to submit copies of their annual self-certification declarations to the regulator. The MCA rejected this recommendation.

31 SECTION 2 - ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 OVERVIEW Ghanaian crewman, Annang Nuertey, was thrown overboard from the UK registered stern trawler Aquarius because he was kneeling within a bight of a trawl wire that suddenly came under tension when the stopper taking the weight of the wire failed. The crew were unable to recover him because the vessel was not effectively equipped to recover a person from the water, its emergency procedures lacked detail, and the crew had not practised manoverboard recovery. 2.3 THE FAILURE OF THE ROPE STOPPER The crew of Aquarius were attempting to mark the vessel’s port trawl warp while underway at sea when the accident happened. In order to do this, the warp was streamed over the vessel’s stern and towed along the seabed. The application of the stopper allowed the crew to slacken the inboard section of the warp and lay it onto the deck. The length of rope used for the stopper was cut from a coil of 3 stranded, 14mm diameter PolysteelTM rope, and had been used to conduct a similar evolution on at least two previous occasions. Remnants of the failed rope were found on board after the accident, but were thrown overboard prior to the vessel arriving back in port. As the remnants of the stopper had been discarded prior to the vessel returning to port, the condition of the rope used to create the stopper and its failure mode could not be verified. However, taking into consideration the sudden nature of the stopper failure, the method used to apply it, and that the remnants were found attached to the gantry cleat after the failure, it was apparent that the rope must have parted. Given this, the investigation considered three main scenarios that could have caused the stopper rope to part. These were: • Tensile overload due to the strain acting on the stopper exceeding the minimum breaking load of the PolysteelTM rope. • Tensile failure due to loss of residual strength as a result of previous rope damage. • Cutting of the man-made fibre rope by the steel wire warp following the failure of the stopper hitch. Aquarius was making 5 to 6kts over the ground and the skipper was executing a slow turn to starboard when the stopper failed. About 210m of the 22mm diameter steel wire warp had been streamed over the stern, over 100m of which would have been dragging along the seabed. Calculations indicate that, had the stopper rope been in good condition, it should have had sufficient tensile strength to hold the weight of the streamed warp. However, despite the surface of the seabed being sand and gravel, the risk of the trailing warp becoming snagged would have been ever present. Regardless of the condition of the rope used to create the stopper, had the warp snagged on the seabed the sudden increase in tensile load would have been more than sufficient to cause the stopper to part.

32 When the crew demonstrated how they had applied the stopper (Figure 9) it was evident that the method used was not in accordance with the guidance for good practice set out in seamanship manuals and CoSWP. It was also noted that the stopper hitch entered the gantry block and the warp slipped a short distance through it before the hitch tightened and the stopper took the strain. PolysteelTM rope is often used on board fishing vessels as it has good abrasion resistance and UV protection characteristics. It is about 40% stronger than equivalent diameter polypropylene and polyethylene ropes. Despite the rope only having been used to stopper the vessel’s warps on a couple of previous occasions, it is highly likely that it would have suffered significant abrasion damage each time the steel wire warp slipped over it. This abrasion damage would have decreased the tensile strength of the rope, and therefore increased the risk of it parting under tension. The manner in which the stopper was applied would also have presented a high risk of the stopper hitch failing. If the steel wire warp had suddenly started to slip through the stopper hitch, it would very quickly have cut through the man-made fibre rope. However, it is most likely that the stopper failed because the rope used to create it parted under tensile overload. It is also likely that the rope’s residual strength had been significantly reduced due to previous abrasion damage. Whatever the failure mode, it is without doubt that the method used to stopper the warp contributed significantly to this accident. 2.4 WORKING PRACTICES ON BOARD AQUARIUS 2.4.1 Marking the trawl The marking of Aquarius’s trawl warps, by opening up the steel wire strands and inserting lengths of man-made fibre rope, was an activity that the skipper and his crew carried out on a regular basis. The task was typically undertaken on the quayside when the vessel was in port. The reason for marking the warps was to allow the skipper to place the net on or near the sea bed, and also by varying the amount of wire veered to maintain the mouth of his net at its optimum width to catch fish. The electronic winch control and acoustic trawl monitoring systems fitted to Aquarius should have removed the need to mark the warps manually. However, the skipper had little confidence in the technology and it was apparent that the trawl monitoring system had not been set up for single net trawling (Figure 8). Having decided to rely on warp marks, it would have been much safer to have undertaken the task on the quayside in Aberdeen before sailing, where facilities were available by prior arrangement. If it was considered essential to conduct the task at sea, steps could easily have been taken to make the evolution much safer. For example, the warp could have been streamed once the vessel was clear of the harbour and marked while the vessel was drifting. The crew had recent experience of marking the warps in a similar manner at sea with the vessel making way. The successful completion of the task on several previous occasions is likely to have influenced some of the crew’s perception of risk, and led to them taking a complacent approach to their own safety. Nevertheless, the dangers associated with the task were obvious but had not been risk assessed, and there were no safe systems of work in place to follow. The activity undertaken was unnecessarily hazardous, and with proper planning the accident could easily have been avoided.

33 2.4.2 Standards of general seamanship The standard of seamanship executed in the build-up to this accident fell well short of recognised good practice. The guidance and warnings contained in publications such as the Fishermen’s Safety Guide, IMO’s Code of Safety for Fishermen and Fishing Vessels and the MCA’s MGNs and CoSWP appeared to have been ignored. Of note: • A man-made fibre rope was used to stopper the steel wire rope. • The way the stopper was applied increased the risk of slippage and failure. • The casualty positioned himself in the bight of a rope. It is a well-established seamanship practice to use chains or specially designed equipment when stoppering steel wire ropes. This should have been well understood by Aquarius’s crew, and man-made fibre rope should not have been considered suitable for use when stoppering the vessel’s steel wire warps. The hitch used to create the stopper, and the manner in which the stopper was applied, were not in accordance with the guidance provided in CoSWP or seamanship manuals (Figure 12). The stopper should have been applied in such a way that it could take the strain before it entered the gantry block. The effect of crushing the stopper rope against the steel groove of the gantry block wheel would have caused further damage to the man-made fibre rope. Standing in the bight of a rope is a hazard that all seafarers are taught to avoid during basic training. Furthermore, the subsequent dangers are constantly emphasised in seamanship manuals, safety posters and safety booklets. It is therefore difficult to understand why a seafarer as experienced as Annang would position himself in the bight of Aquarius’s slackened warp. This is especially so given the knowledge that an ad hoc stopper arrangement was being used to take the strain. It is equally difficult to understand why his fellow crew mates did not intervene and tell him to reposition himself. The space available on the deck, in which to conduct the work, was probably the dominant factor that led Annang to position himself in the danger zone. Figure 12: Chain stopper applied Secured end Free end

34 2.4.3 Leadership and teamwork The skipper did not formally appoint one of his crew to act as his first mate or to be the vessel’s senior deckhand. He often oversaw deck activities himself and typically relayed his orders through Annang or the Filipino crewman. Being the oldest and most experienced fisherman, the Ghanaian crew members looked to Annang for leadership, but also accepted direction from their Filipino crew mate. It was apparent that there had been a predominantly good working relationship among the deck crew, using English as the common language. However, this had recently been undermined when, a few days prior to the accident, the Ghanaian crew members reacted aggressively towards the Filipino crewman following his protests about their use of a fibre rope stopper, in a similar unsafe manner. Culturally, this would have brought shame upon the Filipino19. The lack of an appointed leader, the alienation of the Filipino crewman, and Annang’s status among his Ghanaian crew mates almost certainly had an influence on the way the tasks were conducted on deck. Similarly, these factors probably resulted in an unwillingness among Annang’s crew mates to challenge unsafe acts. This might explain why none of the other crew members intervened when Annang stood in the bight of the warp, and a fibre rope was used to stopper it. 2.5 MAN OVERBOARD In preparation for the task, Annang had positioned himself in the bight of the slackened warp. He was kneeling on the deck and leaning over the warp, with a marlin spike in one hand and the wire rope in the other, when the stopper parted. A second crewman was positioned on the other side of the warp and was helping to hold it steady (Figure 6). When the stopper parted, the slackened warp snapped tight with considerable force. Given Annang’s position within the bight of the warp, it is likely that it struck him across his chest and under his armpits as it snapped tight. The energy with which the crewman was lifted off the deck and catapulted overboard would almost certainly have severely stunned and disorientated him, and was probably sufficient to cause internal injuries. Nevertheless, Annang was alive when he entered the water and was able to keep himself afloat for several minutes without the aid of a PFD. 2.6 EMERGENCY RESPONSE When the crew raised the alarm the skipper’s initial reaction was to focus on monitoring Annang’s position in the water, and manoeuvring Aquarius astern towards him. The crew threw lifebuoys towards their crew mate in the water and shouted encouragement, but despite these efforts Annang soon succumbed and sank below the surface of the water. About 15 minutes after the casualty entered the water, and having realised that the attempts to rescue him had failed, the skipper used his mobile phone to alert Aberdeen VTS. 19 https://geert-hofstede.com/philippines.html

35 The initial steps taken by the skipper and his crew were broadly in line with those listed in the vessel’s manoverboard recovery procedure; however, they fell well short of recognised good practice. Many of the key initial responses listed by the MCA in MGN 430 (F) were not carried out. In particular: • A “Mayday” call was not transmitted. • The crew did not use the lifebuoy that had been rigged with smoke and light signals to mark the position of the casualty. • The skipper did not activate the manoverboard function on the vessel’s navigation aids. Alternative options for recovery of a person in the water were discussed within section 2.6 of the recent MAIB report 26/2015, into the fatal manoverboard from Beryl (BF 440). This included the consideration of deployment of a liferaft to provide either a temporary place of safety, or as a platform to aid recovery. When the crew initially raised the alarm the skipper was under the misconception that he and his crew would be able to recover Annang quickly, and without the need for external support. This assessment was wrong, and the emergency response was ineffective. 2.7 EMERGENCY PREPAREDNESS In order to minimize the consequences of marine accidents and incidents, fishermen and their vessels need to be prepared to deal with a variety of emergency situations. Fishermen are prepared through the delivery of training and the provision of guidance and procedures. Fishing vessels are prepared through design, the provision of LSA and other safety equipment. To ensure that training has been effective, emergency procedures are fully understood and safety equipment is suitable for its intended use, fishing vessel crews should conduct regular and realistic emergency response drills. Aquarius did not carry all the LSA required for a fishing vessel of its size, and the majority of the safety equipment it did carry was not being properly maintained. Of particular note: the vessel did not carry a dedicated means of recovering a person from the water, or a sufficient number of lifejackets for all on board; one of the liferafts had not been properly rigged; access to one of the lifebuoys had been obstructed; and the distress flares and first-aid supplies were out of date. The lifebuoy that was obstructed was the one that had been rigged with smoke and light signals, and it was not used during the rescue attempt. The vessel’s documentation listed patching and line as the equipment to be used to recover a person from the water, but the vessel’s manoverboard recovery procedure prescribed the use of a ring and line. No one on board knew what the term patching and line meant, and the limitations of relying on lifebuoys with lifelines to recover a person from the water became apparent as the emergency situation developed. The crew on board Aquarius were all experienced fishermen; they had completed the UK’s mandated sea survival and safety awareness courses, and should have had at least a basic understanding of what to do when Annang was thrown overboard. However, the vessel’s manoverboard procedure contained little or no detail on the fundamental steps to take in such an emergency situation, and indicated that it would be easy to recover a person from the water over the

36 vessel’s stern. As the crew did not conduct regular emergency drills, the vessel’s manoverboard procedure was not properly challenged, and the crew were denied the opportunity to practise and hone their skills. The MAIB has investigated numerous similar fishing vessel accidents (paragraph 1.18.2) and has repeatedly highlighted the dangers of falling overboard while working on deck, and the difficulty of recovering a person from the water. The MCA, fishing industry federations, and UK maritime safety organisations have also made great efforts to highlight these issues. However, it was clear that the skipper and owners had seriously underestimated the dangers associated with such an event. The response to the crewman falling overboard should have been instinctive to all on board; in particular, the alarm should have been raised immediately by activating the DSC function on the vessel’s VHF radio. It was clear that Aquarius and its crew had not been properly prepared to deal with the situation. Had the vessel been adequately equipped, its safety equipment properly maintained, and its crew regularly drilled, Annang’s chances of survival would have been increased considerably. 2.8 THE USE OF PERSONAL FLOTATION DEVICES ON DECK The risk assessments contained in Aquarius’s safety folder identified falling overboard as a hazard to persons working on deck, and recognised that the consequences of such an event could be fatal. In order to reduce the likelihood of a man overboard drowning, the control measures listed in the risk assessment for working on deck included the wearing of PFDs. In order to implement this control measure, the owners of Aquarius had provided each crewman with a 150N automatic inflation PFD. However, the crew did not wear their PFDs when they attempted to mark the trawl warps; the skipper did not have a PFD. When a person enters cold sea water after falling or being thrown overboard they often die swiftly due to the effects of cold water shock, or over a period of time due to the intake of water in turbulent seas. It is difficult to estimate with any accuracy the likely survival times for people immersed in water as there are many uncertainties. The temperature of the sea water, and the physiological attributes and health of the person immersed are key factors that can influence survival times. An uninjured person may be able to survive for several hours immersed in sea temperatures of 13.8°C.20 However, without the buoyant support of a PFD, a person’s survival time is often measured in minutes and seconds, rather than hours. The extent of injuries suffered, and the effect of cold water immersion when Annang was thrown overboard, are unknown. However, he was alive and treading water for several minutes before he sank out of sight. As the weather at the time of the accident was benign, it is almost certain that Annang would have been recovered had he been wearing his PFD. This was particularly so given the proximity of other vessels and shore-based rescue craft. The wearing of a PFD not only increases the chances of survival for a person who unexpectedly enters the water, but it also increases the likelihood of them being recovered back on board21. The procurement of the PFDs and their distribution to 20 Golden. F. Tipton M (2002). Essentials of Sea Survival pp129-139. 21 In the case of a person lost at sea, it can take up to 7 years under current UK legislation for them to be declared dead. This can cause years of financial and administrative problems for the family.

37 the crew on board Aquarius is to be applauded; however the failure to ensure that they wore them at all times when working was a significant factor in their inability to recover Annang once he fell into the water. 2.9 EFFECTS OF TIREDNESS AND FATIGUE When working, Aquarius was operated 24 hours a day, 7 days a week, by two skippers who shared a 10-day on / 10-day off work routine. The deck crew were all foreign nationals and were employed on 15-month contracts through manning agencies. The crew lived on board for the duration of their contracts and worked continuously without leave breaks. During the 24-hour period prior to the accident, the crew worked a total of 13 hours, but their rest periods were broken, with none lasting longer than 4 hours. It was apparent that Aquarius and its crew were being worked extremely hard for prolonged periods of time; therefore the risk of crew fatigue was high. This risk might have been heightened by the recent change to the vessel’s operating routine, as the switch to day running to fish for squid would have had an impact on their sleeping patterns. Sleep deprivation has a cumulative and negative effect on an individual’s health, and after several sleepless nights the mental effects become more serious. Under such circumstances, an individual’s ability to concentrate and make sound decisions will be impaired and, as a result, the risk of injury and accidents increases. Even if the crew were not feeling the chronic effects associated with long working hours and continuous periods of broken rest, tiredness and a desire to complete the job quickly so that they could sleep ahead of another busy day, almost certainly influenced their approach to the task of marking the warps. 2.10 HOURS OF WORK AND REST MSN 1786 (F) provided clear guidance on the application of the UK’s working time regulations for sea-fishermen employed on UK registered fishing vessels. The purpose of the regulations was to ensure that fishermen receive adequate rest and not less than a set minimum period of annual leave. In order to assess compliance, employers were required to keep records of each individual crew member’s hours of rest. The aim of this requirement was to ensure employers had the information necessary to monitor their crews’ work and rest patterns, and take steps to minimise the risk to their health and safety arising from fatigue. The difficulties in applying strict working time standards across a complex and diverse fishing industry were acknowledged in MSN 1786 (F), and allowance was made in the regulations for the application of exceptions to the mandated limits. The fishermen’s Working Time Standards Code provided a set of government-approved generic exceptions for certain classes of fishing vessel; Aquarius was subject to these exceptions. Nevertheless, the core requirements set out in the Working Time Standards Code were similar to those in the regulations. In circumstances where crew are expected to exceed the mandated working time limits, employers were required to ensure they were given more frequent and longer leave periods, or received compensatory rest.

38 The crew members on board Aquarius were not afforded the minimum standards required by UK legislation; of particular note: • Their hours of rest were not recorded • They continually worked in excess of 48 hours per week • They did not receive 4 weeks’ annual leave or compensatory leave; and • They did not get a 6-hour rest period during the 24 hours leading up to the accident. The Ghanaian crew members’ contracts of employment, issued by a UK manning agency, made no reference to the fishermen’s working time regulations, and did not include maximum hours of work, or annual and compensatory leave entitlements. The vessel’s British skippers took regular 10-day breaks, but the foreign nationals were expected to work and live on board Aquarius without any opportunity to take periods of compensatory leave. 2.11 VESSEL SAFETY MANAGEMENT 2.11.1 Vessel’s safety management system In accordance with the Merchant Shipping and Fishing Vessels (Health and Safety at Work) Regulations 1997, the owners and skippers of Aquarius had a general duty to ensure, as far as was reasonably practicable, the health and safety of their crew. In order to do this, they were required to evaluate unavoidable risks and take actions to minimise them. This included the adoption of safe working patterns and procedures. These requirements, along with a set of prescriptive safety standards, were repeated in the 15-24m Fishing Vessel Code. The owners and skippers of Aquarius used a paper copy of the Seafish safety folder to document the vessel’s risk assessments, procedures and equipment registers. However, it had been compiled prior to its fishing vessel certificate renewal survey in January 2013, and had not been amended or reviewed since. When reviewed, the safety folder was found to contain a minimal amount of vessel-specific information, its documented procedures lacked detail, and it was apparent that it had been hastily compiled in order to maintain the vessel’s UK fishing vessel certification. Aquarius’s poor safety inspection record, and the material condition of the vessel’s safety equipment at the time of the accident indicated that the skipper and owners had adopted a reactive approach to safety management, and took action to meet mandated safety standards only when prompted by deficiencies identified by MCA surveyors. In effect, the owners and skipper were using the MCA as a superintendence and safety management service. 2.11.2 Risk assessment The lack of risk assessments had been raised as deficiencies by MCA surveyors in four out of five inspections that took place on board between January 2011 and January 2013, and had been discussed with the skipper at the time. When the vessel’s safety folder was compiled, the owners relied solely on the generic risk assessments produced by Seafish.

39 Relevant to this accident, the task of marking the trawl was conducted on a regular basis, but a documented risk assessment had not been produced. The method used to mark the trawl warps at sea was unnecessarily hazardous, and the failure of the stopper and subsequent loss of a crewman overboard were entirely foreseeable. This specific risk was documented in The Fishermen’s Safety Guide (paragraph 1.16.3), which contained the following warning: Do not stand on a slack warp laid on the deck; if the ‘stopper’ chain slips, it may suddenly become tight, throwing you up and perhaps overboard. Had the task of re-marking the trawl warps at sea been subject to a risk assessment process, the difficulties of working with a lack of space, the danger zone within the bight of the warp, the application of an appropriate stopper, and the need to wear a PFD could have been identified. This might have prompted the skipper to avoid the risk and do the job before leaving port, or put plans in place to do the task in a much safer manner at sea. 2.11.3 Safety culture and its effect on behaviours The safety culture on board a vessel is the product of individual and group values, attitudes, perceptions, competencies, and patterns of behaviour that determine the crew’s commitment to health and safety management. The prevailing level of safety culture achieved on board a vessel will have a significant effect on the way tasks are done on board, and on what the crew do when no one is watching. In order to achieve a strong safety culture on board a fishing vessel, the skipper must demonstrate his commitment to safety management, and vigorously promote safe systems of work. If a skipper has a natural, unconscious bias for production over safety, or a tendency to focus on the short-term and on being highly reactive, a poor level of safety culture will prevail. Symptoms of poor cultural factors can include: • Widespread, routine procedural violations • Failure to comply with the vessel’s safety management system; and • Management decisions that appear to consistently prioritise the catching of fish over safety. On Aquarius, all three of these indicators had been present for a prolonged period of time, and it was evident that an extremely low level of safety culture existed on board. Safety management should not be considered an unnecessary burden; it should be a natural function of good vessel management. Boats with a strong safety culture are those that are well run, well maintained and employ well-trained crew. As a result, they typically have fewer deficiencies identified, fewer accidents and less resultant downtime. 2.12 CERTIFICATION PROCESS AND EQUIPMENT MAINTENANCE 2.12.1 Vessel owner’s annual self-certification process The owners of Aquarius were responsible for ensuring that the vessel was equipped, surveyed, certified, maintained and operated in accordance with the relevant provisions of the 15-24m Fishing Vessel Code. To assure themselves that the vessel continued to comply with the Code throughout its service, the owners or their

40 appointed representative were required to check the vessel annually. On completion of each annual check the owners should sign a declaration confirming compliance; this should be held on board the vessel for subsequent inspection. The annual self-certification process is fundamental to the UK’s approach to the certification of non-SOLAS (under 24m LOA) fishing vessels. However, no self- declaration forms were found on board Aquarius during its post-accident inspection. This shortcoming is not uncommon and, as highlighted in previous reports, MAIB inspectors often discover that annual self-certifications have either not been completed, or are not available for inspection. Following the Ronan Orla accident (paragraph 1.18.3), the MAIB recommended that the MCA requires vessel owners to submit copies of their self-declaration forms to the MCA for monitoring scrutiny. This recommendation was rejected by the MCA because it did not align with the agency’s 10 year strategy, which was to encourage fishermen to take more responsibility for their own safety. Notwithstanding this aspiration, at the time of this investigation it is evident that the annual inspection and self-certification process was not achieving its designed aim. 2.12.2 Regulatory oversight and the management of deficiencies In the 9 years prior to the accident, Aquarius was surveyed and/or inspected by MCA surveyors on nine separate occasions, the last being a targeted inspection on 21 March 2013. In total, 137 deficiencies were raised during this time, of which 65% were classed as safety-critical. Many of the deficiencies the MCA raised related to the management of safety, with similar problems being highlighted on a regular basis. The local MCA marine office had identified Aquarius as a problem vessel, and its surveyors had spent time explaining to the skipper the nature of the shortcomings they were finding. Despite these efforts, the guidance offered was not heeded and the longstanding safety management shortfalls were again evident during the MCA’s post-accident surveys. Deficiencies identified during surveys were recorded in the MCA’s electronic SIAS system. The closure of non-critical deficiencies relied largely on the honesty and integrity of vessel owners’ reports that rectification action had been completed. In the case of Aquarius, it was apparent that some of the deficiencies raised by the MCA were not properly addressed, but had been reported by the vessel’s agent as being rectified. The SIAS database could not automatically identify poor performers or alert surveyors when outstanding deficiencies had reached their time-bound limit. In order for a marine office to monitor a particular vessel, or fleet of vessels, a surveyor had to manually interrogate the SIAS database and/or review the paper files. Notwithstanding the MCA’s objective of encouraging fishermen to take responsibility for their own safety, it must have a robust system for following up on deficiencies raised during survey and inspection if it is to deal effectively with consistently sub-standard vessels.

41 SECTION 3 - CONCLUSIONS 3.1 SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMENDATIONS

The casualty was thrown overboard when a rope stopper that had been applied to a steel wire warp suddenly failed. [2.2]

It is likely that the rope used to create the stopper parted under tensile overload; the material condition of the rope used to create the stopper, and the method used to apply it were significant contributing factors. [2.3]

The casualty was thrown overboard because he had positioned himself within the bight of the slackened warp. [2.5]

The actions taken by the skipper and his crew to recover the casualty were not in line with the guidance provided by the MCA for a manoverboard situation, and fell well short of standard responses taught on training courses. Of particular note, the alarm was not raised until after the crewman had sunk below the surface of the water. [2.6]

Aquarius and its crew were not properly prepared to deal with the emergency situation. The vessel was not effectively equipped to recover a person from the sea, its safety equipment was not properly maintained, its emergency procedures lacked detail, and its crew had not practised manoverboard recovery. [2.7] 3.2 SAFETY ISSUES NOT DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

Marking a trawl warp by streaming it over the stern of the vessel while underway was unnecessarily hazardous. With proper planning, a safer way of achieving the task could have been identified and the accident could easily have been avoided. [2.4.1]

The crew were all experienced fishermen, but the standard of basic seamanship demonstrated by them was extremely low. The material used to create the stopper and the way it was applied increased the risk of its failure. [2.4.2]

The casualty sank and his body was not recovered because he was not wearing a lifejacket or other type of PFD when he entered the water. [2.8]

Aquarius’s crew were being worked extremely hard for prolonged periods of time. Tiredness and a desire to rest ahead of a busy day’s fishing almost certainly influenced the crew’s approach to the task of marking the trawl warps. [2.9]

Aquarius’s British skipper’s took regular 10-day breaks, but the foreign nationals employed as crew were expected to live and work on board the vessel without any opportunity to take periods of compensatory leave. [2.10]

The skipper and owners of Aquarius adopted a reactive approach to safety management, and took action to meet mandated safety standards only when prompted by deficiencies identified by MCA surveyors. [2.11.1]

42

The accident could have been avoided if the risks of undertaking this task had been properly assessed, and the danger zone formed by the bight of the warp had been identified. [2.11.2]

The skipper and owners of Aquarius consistently prioritised the catching of fish over the safety of the vessel and its crew. This resulted in the promotion of a poor safety culture. [2.11.3]

The skipper and owners had not completed annual self-assessment declarations. Previous investigations have identified that this regulatory shortcoming is endemic across the UK’s under 24m fishing vessel fleet. [2.12.1]

The MCA’s database does not facilitate the identification and targeting of higher risk fishing vessels of under 24m for inspections. Nor does it alert surveyors to outstanding deficiencies to ensure their timely closure. [2.12.2]

43 SECTION 4 - ACTIONS TAKEN 4.1 MAIB ACTIONS The Marine Accident Investigation Branch has:

Published a Safety Flyer (Annex I) to disseminate the main lessons from this accident to the fishing industry 4.2 ACTIONS TAKEN BY OTHER ORGANISATIONS MB Aquarius Ltd has:

Purchased a manoverboard recovery cage and undertaken in-harbour drills and discussions with crew on its deployment and use.

Commenced recording of hours of work and rest for the vessel’s staff.

Amended the work pattern of Aquarius to 10-days on and 5-days off routine.

Changed the work contracts for the deckhands to facilitate back-to-back rota.

Appointed a mate and senior deck hand.

Implemented the use of the SFF safety management system to help improve safety management on board.

Reviewed and updated its risk assessments and recorded such in the Seafish Safety Folder.

Implemented mandatory wearing of PFDs when working on deck. The Maritime and Coastguard Agency has:

Begun to transfer its electronic files, including survey reports and deficiency records to the Single Vessel Database section of the Consolidated European Reporting System. On completion its SIAS database will be discontinued.

44 SECTION 5 – RECOMMENDATIONS The Maritime and Coastguard Agency is recommended to: 2016/139 Develop the capability within its new Consolidated European Reporting System/Single Vessel Database to automate the management of inspection and survey deficiency records so that consistently sub-standard vessels can be quickly identified and targeted, and marine offices are alerted if deficiencies are not rectified within stipulated time frames. 2016/140 Review its monitoring and enforcement of “The Working Time: Sea Fishermen Regulations 2004” to ensure that fishermen, and in particular foreign fishermen living on board their vessels, are achieving the statutory levels of rest and annual leave. The owners of Aquarius are recommended to: 2016/141 Conduct a thorough review of the vessel’s safety management system and take robust actions to improve the safety culture on board Aquarius and any other vessels they might own or operate. Particular attention should be given to ensuring compliance with all appropriate health and safety regulations, the 15-24m FV Code and the hours of work and rest regulations. 2016/142 Ensure that Aquarius and its crew are properly prepared to deal with emergency situations through the conduct of regular and realistic emergency drills. PG Manning Ltd is recommended to: 2016/143 Amend its fishermen’s contracts of employment to include reference to The Working Time: Sea fishermen Regulations 2004 and the employees’ hours of rest and leave entitlements. Safety recommendations shall in no case create a presumption of blame or liability

Marine Accident Report

FD06049 ｜ 2017_英国_gas-explosions-on-general-cargo-ship-nortrader-with-1-person-injured__MAIBInvReport26_2017.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2017 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H ACCIDENT REPORT SERIOUS MARINE CASUALTY REPORT NO 26/2017 DECEMBER 2017 Report on the investigation of the explosion of gas released from a cargo of unprocessed incinerator bottom ash on Nortrader at anchorage in Plymouth Sound on 13 January 2017

Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2017 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch First Floor, Spring Place 105 Commercial Road Southampton Email: maib@dft.gsi.gov.uk United Kingdom Telephone: +44 (0) 23 8039 5500 SO15 1GH Fax: +44 (0) 23 8023 2459 Press enquiries during office hours: 01932 440015 Press enquiries out of hours: 020 7944 4292

CONTENTS GLOSSARY OF ABBREVIATIONS AND ACRONYMS SYNOPSIS 1 SECTION 1 - FACTUAL INFORMATION 2 1.1 Particulars of Nortrader and accident 2 1.2 Narrative 4 1.2.1 Instructions to load 4 1.2.2 Cargo loading and anchorage 4 1.2.3 Explosions in the cargo hold 6 1.2.4 Post-accident response 7 1.2.5 Damage 8 1.2.6 Post-accident actions 8 1.3 Vessel 11 1.3.1 Cargo hold 11 1.3.2 Hatch covers 12 1.3.3 Forecastle store 12 1.4 Crew 12 1.5 Unprocessed incinerator bottom ash 12 1.5.1 Production and storage 12 1.5.2 Sea transportation 14 1.6 Organisations involved in transportation of U-IBA 14 1.6.1 Vessel owners 14 1.6.2 Disponent owner of the vessel 14 1.6.3 Shipper of the cargo 16 1.7 Regulations and guidance for the carriage of bulk cargo 16 1.7.1 SOLAS Convention 16 1.7.2 International Maritime Solid Bulk Cargoes Code 16 1.7.3 Further guidance on the application of the International Maritime Solid

Bulk Cargoes Code 18 1.7.4 National regulations 18 1.8 Competency examination syllabus for senior deck officers 18 1.9 Vessel safety management system 20 1.10 Vessels certified to carry dangerous cargo 20 1.11 Regulations and guidance for waste management 20 1.11.1 Waste classification 20 1.11.2 Testing protocol for waste classification 21 1.11.3 Environment Agency regulations 21 1.11.4 Material safety data sheet for U-IBA 22 1.12 Tests 22 1.12.1 United Nations Manual of Tests and Criteria 22 1.12.2 Gas generation potential and homogeneity of U-IBA 23 1.12.3 MVV Environment Devonport tests 24 1.13 Burgoynes & Partners investigation 26 1.14 Estimation of potential hydrogen production 26 1.14.1 Production of hydrogen 26 1.14.2 Study to estimate hydrogen volume 26 1.15 Previous/similar accidents 27 1.15.1 Accidents at sea 27 1.15.2 Accidents ashore 27

SECTION 2 - ANALYSIS 28 2.1 Aim 28 2.2 The explosion 28 2.2.1 Conditions for the explosion 28 2.2.2 Connection between cargo hold and forecastle store 28 2.2.3 Ignition leading to explosions 29 2.3 Decision to load 29 2.3.1 Precedent 29 2.3.2 Master 29 2.3.3 Shipper 29 2.3.4 UK regulations 30 2.4 Uniqueness of cargo or vessel 30 2.4.1 Cargo 30 2.4.2 Vessel 30 2.4.3 Conclusion 31 2.5 Classification of U-IBA 31 2.5.1 Carriage of U-IBA by sea 31 2.5.2 Classification of U-IBA ashore 31 2.6 Adequacy of United Nations Test N.5 31 2.6.1 Non-replicability of test results 31 2.6.2 Rate of release of flammable gas 32 SECTION 3 - CONCLUSIONS 33 3.1 Safety issues directly contributing to the accident that have been addressed or resulted in recommendations 33 3.2 Safety issues not directly contributing to the accident that have been addressed or resulted in recommendations 33 SECTION 4 - ACTION TAKEN 34 4.1 MAIB actions 34 4.2 Actions taken by other organisations 34 SECTION 5 - RECOMMENDATIONS 35

FIGURES Figure 1

Nortrader at Victoria Wharf, Plymouth Figure 2

Location of the accident Figure 3

Forecastle deck and entrance to forecastle store

Figure 4

Emergency fire pump starter panel (inset: reverse of door, showing broken door lug) Figure 5

The chief engineer’s burnt parka coat Figure 6

Damage caused by the explosion Figure 7

Evidence of short-lived heat damage

Figure 8

Cargo hold lamp access (insets: views from forecastle store with maintenance hatch closed and removed) Figure 9a

General arrangement of the vessel Figure 9b

Hatchcovers undergoing temporary repairs Figure 10

U-IBA from Nortrader after the explosion Figure 11

Quick guide to IMSBC Code Figure 12

Comparison of U-IBA sample from Nortrader with national averages (extracted from the WRc report) TABLES Table 1

Comparison of test results of U-IBA loaded on Nortrader

ANNEXES Annex A

IMSBC Code - schedule entry for aluminium smelting/remelting by- products Annex B

IMSBC Code - 16 point cargo information list and cargo declaration form Annex C

Carrying solid bulk cargoes safely - booklet Annex D

Material Safety Data Sheet for U-IBA Annex E

UN Test N.5, extract from the UN Manual of Tests and Criteria

Annex F

Marchwood Scientific Services flammable gas release test reports Annex G

Report by The Hydrogen Hazards Unit, London Southbank University

Annex H

MAIB safety flyer

GLOSSARY OF ABBREVIATIONS AND ACRONYMS BC Code

Code of Safe Practice for Solid Bulk Cargoes BCSN

Bulk Cargo Shipping Name DOC

Document of compliance EA

Environment Agency ESA

Environment Services Association EU

European Union EWC

European Waste Catalogue H&V

Hudig & Veder BV H2

Chemical formula of hydrogen gas H2O

Chemical formula of water HHU

Hydrogen Hazards Unit HP

Hazard property IBA

Incinerator bottom ash IMDG Code

International Maritime Dangerous Goods Code IMO

International Maritime Organization IMSBC Code

International Maritime Solid Bulk Cargoes Code INTERCARGO - International Association of Dry Cargo Ship Owners l/kg/hour

litres per kilogramme per hour LoW

List of Wastes MARPOL

[MARPOL 73/78] International Convention for the Prevention of Pollution from Ships, 1973 as modified by the Protocol of 1978 MBL

Minimum breaking load MCA

Maritime and Coastguard Agency MGN

Marine Guidance Note MIN

Marine Information Note MSDS

Material Safety Data Sheet MVVD

MVV Environment Devonport Limited NH3

Chemical formula of ammonia

NLS

The Environment Agency’s National Laboratory Services PH3

Chemical formula of phosphene SMS

Safety Management System SOLAS

International Convention for the Safety of Life at Sea 1974, as amended SR

Standard Rules

STCW

International Convention on Standards of Training, Certification and Watchkeeping for Seafarers 1978, as amended (STCW Convention) t

tonnes TC

Time charter U-IBA

Unprocessed incinerator bottom ash UN

United Nations UN Test N.5

United Nations Test method for substances which in contact with water emit flammable gases VHF

Very high frequency

WM3

Guidance on the classification and assessment of waste, Technical Guidance WM3 TIMES: all times used in this report are UTC unless otherwise stated.

1 SYNOPSIS At 1447 on 13 January 2017, the Antigua & Barbuda registered dry cargo vessel Nortrader suffered two explosions in quick succession while anchored in Plymouth Sound, England. The vessel was loaded with a cargo of 2333 tonnes of unprocessed incinerator bottom ash. The first explosion was in the forecastle store and the second in the cargo hold. The chief engineer, who was inside the forecastle store at the time, suffered second degree burns and was airlifted to a nearby hospital. He was repatriated to Ukraine after 12 days and was declared fit for duty 4 months later. The vessel suffered extensive damage and was out of service until 20 April. The MAIB investigation established that the explosions were caused by the ignition of hydrogen gas released from the cargo. Prior to this accident there had been 34 similar shipments of incinerator bottom ash from Plymouth to the Netherlands and, despite it not being listed in the International Maritime Solid Bulk Cargoes Code, no steps had been taken to seek approval from the competent authorities for its carriage. The investigation also found that the testing protocols in place for assessing if the waste was capable of producing flammable gases were inappropriate and inadequate. The MAIB has published a safety flyer to disseminate the lessons from this accident and improve awareness of the International Maritime Solid Bulk Cargoes Code. The MAIB has also provided the Maritime and Coastguard Agency with the results of laboratory tests and technical research conducted during the investigation. The Maritime and Coastguard Agency has set up tripartite agreements between the UK, the Netherlands and several other administrations for the safe carriage of incinerator bottom ash and proposed its inclusion in the International Maritime Solid Bulk Cargoes Code. The Maritime and Coastguard Agency has been recommended to update the Merchant Shipping (Carriage of Cargoes) Regulations to refer to the International Maritime Solid Bulk Cargoes Code and to work with the Environment Agency to ensure that test protocols for the classification of cargoes are fit for purpose. Recommendations have also been made to Nortrader’s owners to review their safety management system to reflect the requirements of the International Maritime Solid Bulk Cargoes Code.

2 SECTION 1 - FACTUAL INFORMATION 1.1 PARTICULARS OF NORTRADER AND ACCIDENT SHIP PARTICULARS Vessel’s name Nortrader Flag Antigua & Barbuda Classification society Bureau Veritas IMO number 9557393 Type Dry cargo Registered owner NTO Shipping GmbH & Co.KG Manager(s) NTO Shipping GmbH & Co.KG Construction Steel Year of build 2012 Length overall 87.91m Gross tonnage 1934 tonnes Minimum safe manning 6 Authorised cargo General cargo VOYAGE PARTICULARS Port of departure Plymouth, England Port of arrival Beverwijk, the Netherlands (intended) Type of voyage Short international Cargo information 2333t of unprocessed incinerator bottom ash Manning 7 MARINE CASUALTY INFORMATION Date and time 13 January, 1447 Type of marine casualty or incident Serious Marine Casualty Location of incident Plymouth Sound, England Place on board Cargo hold Injuries 1 Damage/environmental impact Material damage to ship, no environmental impact Ship operation Normal service Voyage segment Anchored External environment Calm seas, light breeze, good visibility, occasional rain Persons on board 7

3 Nortrader

4 1.2 NARRATIVE 1.2.1 Instructions to load On 10 January 2017, the master of Nortrader received instructions by email to load a full cargo of unprocessed incinerator bottom ash (U-IBA) at Plymouth, England for Beverwijk in the Netherlands. The email was sent from Hudig & Veder Chartering BV (H&V), a company based in the Netherlands. It specified that the cargo was non dangerous and stated: …cargo also include some foreign materials which is no problem. Cargo can be loaded/discharged in rain. [sic] On receipt of these instructions, the master and chief officer referred to the International Maritime Solid Bulk Cargoes Code (IMSBC Code), but did not find an entry for U-IBA. The master decided to load as instructed, and did not receive any further information about the cargo. 1.2.2 Cargo loading and anchorage Nortrader berthed at Victoria Wharf, Plymouth, at 0655 on 12 January. Loading commenced at 0815, and by 1600 that day 2333t of U-IBA had been loaded in heavy and persistent rain (Figure 1). After completion of the final draught survey, the master checked the weather forecast and decided to anchor off Plymouth and wait for the imminent bad weather to pass. At 1900, the vessel left the berth and anchored 30 minutes later in Plymouth Sound (Figure 2). Figure 1: Nortrader at Victoria Wharf, Plymouth

5 Figure 2: Location of the accident Vessel anchoring position Reproduced from Admiralty Chart BA 1900-0 by permission of the Controller of HMSO and the UK Hydrographic Office

6 1.2.3 Explosions in the cargo hold The following morning, on the bridge, the chief officer instructed the bosun to wash the deck and superstructure. The chief engineer, also present on the bridge at the time, requested that the bosun use the electric emergency fire pump for the task in order to test it. The bosun started the emergency fire pump from the bridge, but a short while later he reported to the chief engineer that it was not delivering any water. He then changed over to the main fire pump and the crew completed washing by 1200. At approximately 1440, the chief engineer went to the forecastle store (Figure 3) where the emergency fire pump main starter panel was located. He opened the door to enter the store and used the hook on its outside to secure the door in the open position. He then went down the ladder into the store and started the emergency fire pump. From the sound of its operation he suspected that the pump was running dry and so he returned to the forecastle and opened a fire hydrant, to confirm that no water was being pumped. He then returned to the store and stopped the emergency fire pump using the stop button on the starter panel (Figure 4). As soon as he stopped the pump, there were two loud explosions in quick succession. The fire detection system registered this event at 1447. The first explosion threw the chief engineer onto the emergency fire pump starter panel. The second explosion pushed him violently backwards and he fell on the deck under the starter panel. Figure 3: Forecastle deck and entrance to forecastle store Entrance to forecastle store Cargo hold vent fan suction

7 1.2.4 Post-accident response At the time of the accident, the third officer was on watch. On hearing the explosions, the master and the rest of the crew immediately came to the bridge. They saw the chief engineer staggering along the starboard side of the main deck between the cargo holds and shipside protective railing. The bosun met him halfway between the forecastle and the accommodation, and escorted him to the mess room. The synthetic fibre-filled parka1 coat that the chief engineer had been wearing had melted onto his body and head. He was conscious and complained of being in extreme pain. Once in the mess room, the chief officer cut the chief engineer’s coat off him (Figure 5) and attempted to make him comfortable. The master called the coastguard using very high frequency (VHF) radio to inform them about the explosions and asked for immediate medical assistance for the chief engineer. By 1531, rescue helicopter R924 was in position above the vessel. A paramedic was landed on board Nortrader and, following a quick assessment, the chief engineer was lifted into the helicopter and evacuated to Derriford Hospital in Plymouth from where he was subsequently transferred to the Acute Burns Clinic 1 Parka coat: a coat used in cold weather usually stuffed with a synthetic fibre or down. Figure 4: Emergency fire pump starter panel (inset: reverse of door, showing broken door lug)

8 at Southmead Hospital, Bristol. He had suffered first degree burns to his face and second degree burns to his body, both hands and lower extremities. On 25 January, he was discharged and repatriated to Ukraine, where his treatment was continued for 3 weeks. He was declared fit for duty in May 2017. 1.2.5 Damage The explosions dislodged and distorted all nine of the vessel’s steel hatch covers, breaking all but one of the 66 hatch cleats in the process (Figure 6). Some of the hatch covers dropped into the cargo hold and were found resting on top of the cargo. The cargo hold coaming bar was also damaged in several locations and the cargo ventilation trunking tower flaps were deformed. There was no fire damage, although there was evidence of intense heat within the forecastle store. A parka coat, like that worn by the chief engineer, which had been hung at the entrance to the forecastle store, had been severely melted and several laminated sheets and plastic document wallets were heat damaged (Figure 7). 1.2.6 Post-accident actions A Maritime and Coastguard Agency (MCA) surveyor inspected Nortrader while still at anchor and gave the master permission to return to Plymouth. On 14 January, Nortrader berthed at Victoria Wharf, and over the next 3 days discharged the entire cargo. The cargo was returned to the storage shed at Victoria Wharf. The MCA prohibited further sea transportation of U-IBA from the UK until the procedures described in the IMSBC Code for cargoes not listed in it had been completed. Figure 5: The chief engineer’s burnt parka coat

9 Figure 6: Damage caused by the explosion Dislodged hatch covers Distorted hatch covers Broken cleat

10 Figure 7: Evidence of short-lived heat damage

11 Nortrader’s Classification Society, Bureau Veritas, carried out a damage assessment survey, which concluded that the vessel had not sustained any structural damage, and it was granted a dispensation to sail to Rotterdam for repair. Five of the nine hatch covers were replaced along with all of the hatch cleats and the cargo hold coaming bar. The cargo ventilation trunking tower flaps were repaired. The vessel returned to service on 20 April. 1.3 VESSEL 1.3.1 Cargo hold Nortrader had a single cargo hold of 3409m3 capacity. It was possible to sub-divide the hold with two movable bulkheads. At the time of the accident, the bulkheads were not in use and were stored at the aft end of the hold. There were two forced draught fans for ventilation. These were not in use and their inlets had been closed and dogged down after loading, in preparation for poor weather. There were two cargo lamps at the forward end of the hold. The access to the starboard lamp was from within the forecastle store. The access plate for this was found loosely fitted using two of ten bolts, and the sealing gasket was missing (Figure 8). The previous cargo carried by the vessel had been cattle feed. After this cargo was discharged, the holds had been washed clean and ventilated. Figure 8: Cargo hold lamp access (insets: views from forecastle store with maintenence hatch closed and removed)

12 1.3.2 Hatch covers The hold had eight large hatch covers each weighing 14t, and a smaller central one weighing 5t. These hatch covers were made tight against the coaming bars with 66 hatch cleats that had minimum breaking loads (MBL) of 8.75t (athwartships cleats) and 9.0t (longitudinal cleats). There were also 40 steel wedges (3.5t MBL) used to interlock the tops of adjacent hatch covers (Figure 9a and b). 1.3.3 Forecastle store The forecastle store was located in a compartment forward of the hold. It was accessed via a door on the forecastle deck and down a short stairway. The starter panels for various equipment, including the emergency fire pump and cargo hold ventilation fans, were located in the forecastle store. The emergency fire pump and bow thruster were located at the bottom of the compartment and were accessed by a vertical ladder from the store. At the time of the accident, the emergency fire pump starter panel door was not properly closed as one of its two fastening lugs was broken. The forecastle store was used to store consumables for deck maintenance. The rope and paint stores were located forward of the forecastle store and both stores were open to each other. The store was naturally ventilated by gooseneck vent pipes to the forecastle above. On the day of the accident, all these vents had been closed in preparation for heavy weather. 1.4 CREW Nortrader’s crew comprised the master, chief officer, third officer, chief engineer and three ratings from the Russian Federation and Ukraine. Their contracts were 4 months long, and none of them had any previous experience on ships carrying U-IBA. The master held an STCW2 II/2 (unlimited) certificate of competency gained in 1991 and had completed six contracts as master on Nortrader. He was 59 years old. The chief engineer was 31 years old and held an STCW III/2 (unlimited) chief engineer’s certificate of competency. He had completed four contracts on Nortrader and had last joined the vessel in September 2016. 1.5 UNPROCESSED INCINERATOR BOTTOM ASH 1.5.1 Production and storage U-IBA is the ash accumulated at the bottom of the furnace when waste material is incinerated. The cargo of U-IBA loaded onto Nortrader was generated at the incinerator plant owned and operated by MVV Environment Devonport Limited (MVVD) located in Devonport, Plymouth. MVVD started operating in June 2015 and handled approximately 250,000t of non-recyclable waste in a year, which included 80,000t of commercial waste. Waste from the Plymouth and Torbay areas and Devon County Council was incinerated at the plant and the heat energy produced was recovered in a combined heat and power plant. The waste was not sorted prior to incineration. 2 International Convention on Standards of Training, Certification and Watchkeeping for Seafarers 1978, as amended (STCW Convention).

13 Figure 9a: General arrangement of the vessel Emergency fire pump Position of chief engineer just before the explosion Figure 9b: Hatch covers undergoing temporary repairs 5t centre section Steel locking wedges Coaming bar

14 At the time of the accident, there were 32 operational municipal waste incineration plants in the UK. All of these except for MVVD processed their incinerator bottom ash (IBA) either on site, or off-site within the UK. Transportation to off-site processing plants was carried out using open top or tarpaulin covered freight lorries. Other than during initial start-up, the combustion in the furnace was sustained entirely by the waste, and the furnace temperature ranged from 800°C to 1100°C. On leaving the furnace the IBA was quenched in water before being stored in an open bunker adjacent to the incinerator. The cargo discharged from Nortrader was typical of the U-IBA from the MVVD facility and included items such as unburnt paper, domestic gas cylinders and wheels from motor vehicles (Figure 10). Approximately 200t of U-IBA per day was transferred by freight lorries to the storage shed at Victoria Wharf in Plymouth. The shed had a maximum capacity of 4000t and was naturally ventilated by its open front. The U-IBA was stored in a large pile in the shed where old and new batches were not segregated. 1.5.2 Sea transportation Since MVVD started operating in June 2015, there had been 34 shipments of U-IBA from Plymouth to Beverwijk on 26 vessels similar in tonnage to Nortrader. The description of the cargo for all shipments had been identical to that provided during the engagement of Nortrader. More than a third of these vessels had the document of compliance (DOC) required to carry dangerous goods, although this was not a requirement of the charter party3 agreement for shipment. The first shipment of U-IBA employed the vessel Kine, which had a DOC for dangerous cargoes. On that occasion, H&V acted as a broker between the ship owner and the shipper of the cargo, Rock Solid BV. The material safety data sheet (MSDS) for U-IBA was provided to Kine. After the first shipment, except when masters of vessels specifically requested for it, the MSDS was not sent to the vessels. 1.6 ORGANISATIONS INVOLVED IN TRANSPORTATION OF U-IBA 1.6.1 Vessel owners Nortrader’s registered owner was NTO Shipping GmbH based in Germany. NTO Shipping GmbH owned and operated 18 vessels, and was part of the EMS-Fehn Group, which offered services including logistics, forwarding, port operation and crewing. 1.6.2 Disponent owner of the vessel From 15 December 2016 Nortrader was on time charter to Hudig & Veder (H&V), a shipping company based in Rotterdam, making it the disponent4 owner of the vessel. H&V offered a variety of services including chartering, cargo inspection, stevedoring and port agency. It also owned nine ships similar in construction and size to Nortrader. 3 A charter party is the contract between the owner of a vessel and the charterer for the use of a vessel. 4 The disponent owner is a person or company that assumes the legal responsibilities of the registered owner of the vessel during the term of a time charter or bareboat charter.

15 Figure 10: U-IBA from Nortrader after the explosion Unburnt paper

16 On 10 January 2017, H&V chartered Nortrader to the shipper, Rock Solid BV, for a single voyage to load U-IBA at Plymouth on 12 January for discharge at Beverwijk. The charter party agreement described the cargo as follows: a full and complete cargo avi-slakken (bodem as) stw dw in bulk, untreated incinerator bottom ash in bulk, cargo non dangerous, non-imo classed. [sic] H&V acted as brokers for the first shipment of U-IBA from Plymouth. For the following 33 shipments they either used their own vessels or engaged vessels on short-term charters. 1.6.3 Shipper of the cargo The shipper was Rock Solid BV, a company based in the Netherlands that specialised in processing U-IBA to recover mineral residues. The remainder of the ash was prepared for use in the road building and construction industries, landfill and the production of fertilisers. MVVD and Rock Solid BV had agreed a contract that committed MVVD to supply Rock Solid BV with 68000t of U-IBA annually. Rock Solid BV took ownership of the U-IBA at the point of production and was responsible for transporting it from MMVD to its processing facility at Alkmaar in the Netherlands. The Bill of Lading5 for the cargo loaded on Nortrader showed MVVD as the shipper, but for all practical purposes Rock Solid BV was the shipper of the cargo. 1.7 REGULATIONS AND GUIDANCE FOR THE CARRIAGE OF BULK CARGOES 1.7.1 SOLAS Convention The International Convention for the Safety of Life at Sea 1974, as amended (SOLAS) Chapter VI ‘Carriage of Cargoes’, parts A and B contains the mandatory provisions concerning the carriage of bulk cargoes. It requires the governments involved in the transportation chain to provide information regarding the cargo, including the precautions required for its safe carriage. SOLAS refers to the IMSBC Code for further information about the cargo and requirements for its safe carriage. 1.7.2 International Maritime Solid Bulk Cargoes Code The IMSBC Code entered into force on 1 January 2011 under the provisions of the SOLAS Convention. Appendix 1 of the Code, Individual schedules of solid bulk cargoes, listed all solid cargoes permitted to be shipped in bulk and provided a schedule of requirements for the safe handling and carriage for each. The Code also grouped cargoes according to the hazards associated with their carriage6. Those in groups A or B were also entered into the International Maritime Dangerous Goods (IMDG) Code. An example of a schedule for a cargo that can develop small quantities of hydrogen and ammonia, aluminium smelting/remelting by-products, processed, is at Annex A. 5 A bill of lading is a legal document between the shipper and the carrier providing details of the quantity, type and destination of the cargo, and also serves as a receipt of shipment. 6 Group A consisted of cargoes that may liquefy if shipped at a moisture content in excess of their transportable moisture limit. Group B consisted of cargoes that posed a chemical hazard which could give rise to a dangerous situation on a ship. Group C consisted of cargoes that did not belong to Groups A or B.

17 Section 4.2.2 of the IMSBC Code contained a list of information about the intended solid bulk cargo that the shipper was to provide to the master prior to loading. Of the 16 items on this list, item 1 was the bulk cargo shipping name (BCSN) if listed in the IMSBC Code, item 2 was the cargo group (A, B, A and B, or C) and item 14 was properties on emission of flammable gases in contact with water, if applicable. The MSDS was not part of the required information. In addition to this list of information, the shipper was required to provide the master with a completed and signed cargo declaration form (Annex B). Solid bulk cargoes not listed in Appendix 1 of the IMSBC Code were not permitted to be loaded until the competent authority at the loading port had authorised the intended cargo. In the UK, the competent authority was the MCA and the procedure to be followed was stated in Marine Guidance Note (MGN) 512, Solid Bulk Cargoes

• Guidelines for the submission of information and completion of the format for the

properties of cargoes not listed in the International Maritime Solid Bulk Cargoes (IMSBC) Code and their conditions of carriage, issued in June 2014. This note was specifically aimed at all ship owners, ship operators, terminal operators, port authorities, classification societies, agents, charterers, shippers, consignors, masters and all other parties involved in the transport of solid bulk cargoes by sea. At section 1.1 it stated: Shippers of solid bulk cargoes intended for carriage and which are not listed in appendix 1 of the International Maritime Solid Bulk Cargoes (IMSBC) Code, are required, prior to loading, to provide the Competent Authority of the port of loading with the characteristics and properties of the cargo in accordance with Section 4 of the Code. The International Maritime Organization’s (IMO) Marine Safety Committee (MSC) Circular 1453, Guidelines for the submission of information and completion of the format for the properties of cargoes not listed in the International Maritime Solid Bulk Cargoes (IMSBC) Code and their conditions of carriage provided a questionnaire to aid the gathering of information about an intended cargo. The application of these guidelines became mandatory on 1 January 2015. Under Hazardous properties, the questionnaire contained the following question: Does the cargo react with water causing toxic or flammable gases to be released? Which gases? How toxic or flammable are the gases? What is the rate of evolution? If the proposed cargo was assessed to belong to group A or B of the IMSBC Code, the competent authorities at the ports of loading and discharge, along with the Flag State of the intended vessel were required to agree conditions for the carriage of that cargo; this was known as a tripartite agreement. If the proposed cargo was assessed to belong to Group C in the IMSBC Code, the competent authority was able to permit the carriage without involving other parties. Between the introduction of the IMSBC Code in 2011 and this accident, the MCA had not received any request or enquiry concerning the carriage of cargoes not included in the IMSBC Code.

18 1.7.3 Further guidance on the application of the International Maritime Solid Bulk Cargoes Code In addition to the guidance provided by MGN 512 and IMO MSC Circ.1453, Lloyd’s Register, UK P&I Club and the International Association of Dry Cargo Ship Owners (INTERCARGO) published Carrying solid bulk cargoes safely, included at Annex C. This straightforward guide on the application of the IMSBC Code stated in its introduction: Before you can accept a cargo for shipment, the shipper must provide the Master with valid, up-to-date information about the cargo’s physical and chemical properties. The exact information and documentation they must provide is listed in the Code under ‘Assessment of acceptability of consignments for safe shipment; Provision of Information’, and includes the correct Bulk Cargo Shipping Name and a declaration that the cargo information is correct. A loading flowchart contained in this guide provided clear guidance to the seafarer on the steps required to be taken before agreeing to load solid bulk cargoes (Figure 11). 1.7.4 National regulations The UK’s Merchant Shipping (Carriage of Cargoes) Regulations 1999 (Statutory Instrument 336) stated: The owner and master of every ship to which these Regulations apply, other than a ship engaged in the carriage of grain, shall ensure that appropriate documentation, relevant to the cargo and its stowage and securing, which should specify in particular the precautions necessary for the safe carriage of that cargo by sea, is carried on board… The regulation referred to the Code of Safe Practice for Solid Bulk Cargoes (BC Code). The BC Code was superseded by the IMSBC Code in January 2011 but the regulation had not been updated to reflect this. In contrast with the IMSBC Code, the BC Code provided non-mandatory guidance and did not require masters to take action before loading a non-listed cargo. The MCA published Marine Information Note (MIN) 349, Solid Bulk Cargoes – Adoption of the International Maritime Solid Bulk Cargoes (IMSBC) Code in 2009 advising the industry, including ship owners, managers, masters and officers that the application of the IMSBC Code would become mandatory through SOLAS Convention (as amended) from 1 January 2011. This MIN expired on 31 May 2010. 1.8 COMPETENCY EXAMINATION SYLLABUS FOR SENIOR DECK OFFICERS The STCW syllabus for chief officers and masters required, under the heading Cargo handling and stowage at the management level, that candidates demonstrate the: Ability to establish procedures for safe cargo handling in accordance with the provisions of the relevant instruments such as IMDG Code, IMSBC Code, MARPOL7 73/78 Annexes III and V and other relevant information. 7 MARPOL: International Convention for the Prevention of Pollution from Ships

19 Carrying solid bulk cargoes safely © Lloyd’s Register/UK P&I Club/Intercargo, 2013 Page 18

Shipper has delivered all required cargo information? IMSBC Code consulted regarding cargo and associated hazards? Master and crew have the necessary ship data and information to prepare ship for loading? Cargo spaces inspected and prepared for cargo loading and stowage? Is the cargo categorised as ‘dangerous goods in solid form in bulk’ by SOLAS regulation VII/7? Does the Master have a special list/manifest/stowage plan identifying its location, are there instructions on board for emergency response, and does the ship have a ‘Document of Compliance for the Carriage of Dangerous Goods’? Has the Loading Plan been agreed with the Master and terminal representative? Are there instructions to suspend the loading/unloading operation if the ship’s limits are exceeded, or are likely to be exceeded if continued? YES YES LOAD (monitoring the cargo loading operation) A B C Not listed in IMSBC Code? NO TML Certificate? Moisture content (MC) declaration? MC lower than TML? Cargo spaces free from liquids? Visual monitoring conducted? Trimming considered? Stowage/segregation acceptable? Atmospheric conditions suitable and monitoring tools operational? Applicable dust precautions in place? Ventilation meets operational and IMSBC requirements? Competent authority has provided the Master with certificate stating carriage requirements (if Group A or B, agreed by authority of port of unloading and flag)? IMSBC Cargo Schedule consulted? Applicable precautions taken? DO NOT LOAD Cargo Group? NO NO NO NO NO NO NO NO Loading flowchart – follow the steps to see if it is safe to load your cargo Figure 11: Quick guide to IMSBC Code

20 1.9 VESSEL SAFETY MANAGEMENT SYSTEM The vessel’s safety management system (SMS) contained specific and detailed procedures concerning cargo operations under a section titled Procedure Cargo Operations. It stated that the master, with the assistance of the chief officer, was responsible for assisting the owner in meeting his obligation to receive, stow, carry safely and deliver the cargo. It further stated: In doing so, the reputation of the Owners and the Master will assure that the vessel is regarded as a reliable carrier … Such a good reputation has a positive financial effect both in the freight business and with insurance companies. Under the sub-section titled Cargo Acceptance, further instructions to the master were available to ensure the cargo was as specified, undamaged and the quantity as agreed. In the final section Dangerous Goods the SMS stated: The IMDG-Code and / or the IMSBC-Code (newest editions) are mandatory for the carriage of dangerous goods. There was no guidance in the SMS concerning the carriage of a cargo not included in the IMSBC Code. The SMS stated that the working language on board was English. 1.10 VESSELS CERTIFIED TO CARRY DANGEROUS CARGO SOLAS Chapter II-2 Regulation 19.4 listed the special requirements for vessels carrying dangerous goods. These included specific requirements for fire-fighting and detection systems, safety of electrical equipment (where permitted by the Administration to be fitted inside enclosed cargo spaces), personnel protection, bilge pumping and ventilation systems. Additionally, cable penetrations of decks and bulkheads were required to be sealed to prevent the passage of gas or vapour. The requirements for ventilation stated: Adequate power ventilation shall be provided in enclosed cargo spaces. The arrangements shall be such as to provide for at least six air changes… , and for removal of vapours from the upper or lower parts of the cargo spaces, as appropriate. 1.11 REGULATIONS AND GUIDANCE FOR WASTE MANAGEMENT 1.11.1 Waste classification European Union (EU) Directive 2008/98/EC on waste, generally known as the ‘Waste Framework Directive’, lays down the principles of waste management to prevent health risks to human beings and harm to the environment. The European Waste Catalogue (EWC) is a hierarchical list of waste descriptions. It is divided into 20 main chapters, each of which has a two-digit code between 01 and 20. IBA was coded as 19.01.12 and was not appended with an asterisk mark, indicating that it was not considered hazardous8. 8 Pursuant to Article 1(4) first indent of Directive 91/689/EEC.

21 The UK incorporated the guidance provided by the Waste Framework Directive into Guidance on the classification and assessment of waste, Technical Guidance WM3 (WM3). This document was jointly published by the Environment Agency (EA) in England and its equivalents in Scotland, Wales and Northern Ireland. WM3 provided step by step instructions for classifying waste as required by the List of Wastes (LoW) Regulations 2005, which itself was a transposition of the EWC into UK law. (The LoW Regulations 2005 were revoked in July 2015 and domestic legislation has since referred to the Waste Framework Directive.) In the LoW, U-IBA was classified as ‘mirror hazardous’, which implied that it could be either hazardous or non-hazardous depending on the results of regular testing. In the UK, incinerator operators who were members of the Environmental Services Association9 (ESA) undertook to test the U-IBA they produced on a monthly basis. 1.11.2 Testing protocol for waste classification To determine the classification of waste material, WM3 categorised 15 hazard properties (HP) numbered HP1 to HP15. Of these, HP3 defined the flammability of the waste product together with its ability to emit flammable gases in dangerous quantities when in contact with water. The threshold for categorising a product as dangerous was its ability to generate more than 1 litre of gas per hour per kg (l/kg/ hour). In 2012, the ESA developed a test protocol for identifying the hazards associated with U-IBA10, which was endorsed by the EA. Following the guidance provided in WM3, this protocol discounted six hazard properties including HP3. One of the sub categories of the discounted HP3 was HP3A: substances and preparations which in contact with water or damp air, evolve highly flammable gases in dangerous quantities. The ESA protocol justified the discounting of HP3 by stating the following: IBA has been through an incineration process and therefore compounds present in the ash are not flammable. The ESA protocol also recommended that full characterisation and assessment of the 15 hazard properties be undertaken annually; MVVD had not carried out such annual tests. MVVD was not a member of the ESA and contracted out all its U-IBA testing to Marchwood Scientific Services, also not a member of the ESA. However, as the ESA protocol was the only available guidance for U-IBA testing, it was universally followed by members and non-members of ESA, including Marchwood Scientific Services. Before this accident, none of the waste incineration facilities in the UK tested U-IBA for HP3. 1.11.3 Environment Agency regulations The EA’s Standard Rules SR2012 No13 on the treatment of IBA stated: hydrogen gas is released from the IBA during the ageing process as aluminium reacts with calcium hydroxide and water to form aluminium hydroxide. 9 The Environmental Services Association is a trade association for the UK’s waste management industry. 10 Assessment of hazard classification of UK IBA – Report for the January-June 2011 IBA dataset.

22 SR2012 No13 also stated that if the ash was stored in a building or under cover, adequate ventilation should be provided to safely disperse any gas generated. Furthermore, it stated that where there was the likelihood of a flammable or explosive atmosphere being formed, such areas should be assessed in accordance with the Dangerous Substances and Explosive Atmosphere Regulations (2002). 1.11.4 Material safety data sheet for U-IBA The purpose of an MSDS is to provide information regarding the safe handling and use of a material. They provided useful information to ships’ crews regarding the risks associated with contact with a cargo and, as such, should be carried on board. However, MSDS are not required by the IMSBC Code as they had no place in assessing the suitability of a cargo for safe carriage. The MSDS for U-IBA (Annex D) was supplied by Rock Solid BV to H&V in October

1. Another MSDS for processed incinerator bottom ash aggregate was also

supplied in 2016 to H&V by Rock Solid BV, and this was erroneously included in the voyage charter party for the carriage of U-IBA by Nortrader. On the fifth page of both of these 6-page documents, the MSDS stated under the heading ‘Hazardous decomposing compounds’: When in contact with water (H2O) possible formation of hydrogen gas (H2-gas). The MSDS for U-IBA further stated: Under reducing circumstanced ammonia can be formed (NH3). Accidentally fosfine (PH3) can escape [sic] The MSDS also stated that the EWC waste catalogue classified U-IBA as non-hazardous. The spaces in the MSDS labelled ‘Identification of the producer’ and ‘Telephone number in case of Emergency’ were left blank. 1.12 TESTS 1.12.1 United Nations Manual of Tests and Criteria The United Nations publication Recommendations on the transport of dangerous goods, Manual of tests and criteria, commonly known as the UN Manual of Tests and Criteria, contained guidance for competent authorities to identify and classify goods as dangerous for transportation. This publication did not apply directly to bulk transport of dangerous goods in bulk carriers but the IMSBC Code refers to the UN Manual of Tests and Criteria, part III, for the classification of such cargoes. The competent authority had the discretion to vary the tests or require additional tests if they judged deviation from the prescribed tests to be necessary. UN Test N.5, Test method for substances which in contact with water emit flammable gases, described in section 33.4.1.4 (Annex E) of the UN Manual of Tests and Criteria, was designed to determine if a substance should be categorised as dangerous with respect to its ability to generate flammable gases when in contact with water. The test involved grinding the substance to a homogeneous mixture and measuring the gases released when small quantities (heaps of diameter 2mm and 20 mm) were mixed with distilled water. These results were then extrapolated to provide an estimate of gas released in litres per kg per hour (l/kg/hr). If this exceeded 1 l/kg/hour, the substance was classified as dangerous.

23 Due to the heterogeneous nature of U-IBA and the presence of large metallic objects, it was very difficult to ensure that samples created for testing were similar in composition to the bulk U-IBA from which they were taken. Despite following a detailed sampling protocol11 developed by the ESA, this investigation has identified significant variability across samples. 1.12.2 Gas generation potential and homogeneity of U-IBA Following the accident, on 17 January, MAIB inspectors collected 4 samples, 20kg to 25kg each, from the discharged cargo. These samples were placed in sealed containers and stored at MVVD. The MAIB commissioned WRc (an environmental consultancy, recognised as experts in waste characterisation and classification) to organise the testing of these samples and interpret the results. WRc was tasked with answering two questions:

1. Was the U-IBA loaded on Nortrader similar in physical and chemical

characteristics to that produced by other incinerator plants in the UK and the MVVD plant from the time of its commissioning?

1. What gases are released from the U-IBA samples of Nortrader and how much

of each gas was the sample capable of releasing? On 8 February WRc collected the sealed samples taken by the MAIB from MVVD along with a fresh sample from the MVVD incineration plant. Two methods were used to test these samples for the release of flammable gases:

1. UN Manual of Tests and Criteria, Test N.5;
2. An in-house test developed by the EA’s National Laboratory Services (NLS).

The main difference between these two test methods was the quantity of water added to the samples. For the UN Test N.5, the liquid-to-solid ratio was 1.5:1 and for the NLS test it was 6:1. The sample preparation method employed by WRc was to separate the ferrous and non-ferrous metal items, grind the remaining sample to less than 4mm particle size, subsample and grind again to less than 250 microns. For flammability tests, the separated metal was then ground to less than 1mm particle size and returned to the sub-samples proportionately. The reported results can be summarised as follows: ● ●UN Test N.5: No gas was produced by the three sub-samples taken from the MAIB sample. One of the three sub-samples from the WRc sample produced hydrogen at the rate of 0.43 l/kg/hour, stopping after 6 hours. ● ●NLS in-house flammability test: All three sub-samples of the MAIB sample produced hydrogen at an average rate of 0.41 l/kg/hour and all three sub- samples of the WRc sample produced hydrogen at an average rate of 0.77 l/ kg/hour. 11 Four crane grab samples are taken from the ash pit. Five samples of 10kg each are taken from each grab producing a sub-sample of 200kg of ash. This sample is then mixed, coned and quartered (a method to reduce systemic bias). Two opposite quarters are then selected for tests.

24 The MAIB samples were also tested to quantify metals, and several other parameters including alkalinity and moisture content. The report stated that, assuming all the non-ferrous metal in the U-IBA sample provided by MAIB was metallic aluminium, the total theoretical hydrogen production potential was 31 l/kg. The WRc report concluded as follows: ● ●In general the U-IBA collected from the ship on the 17th of January 2017 exhibited similar chemical characteristics to UK wide U-IBA samples analysed by WRc. ● ●The concentration of major elements and metals with the potential to form flammable gases were within the expected range for UK U-IBA from municipal waste facilities based on WRc’s in-house data set. ● ●Neither the MVVD U-IBA sample collected from the ship post the explosion, or the sample of fresh U-IBA collected directly from the Devonport facility were hazardous by HP3-A (Highly flammable – water reactive waste). ● ●Although neither U-IBA samples tested were hazardous by HP3; when exposed to water a low rate of H2 generation was observed during both the UN test N.5 and the in-house NLS test method (0.43 and 0.77 l kg-1 h-1 respectively). Considering the rate of H2 production, scenario specific risk assessments should be conducted regarding U-IBA transport and storage where confined conditions are used. There could be a potential risk of explosive atmosphere formation in certain conditions when the quantity of U-IBA is large relative to the effective air space. 1.12.3 MVV Environment Devonport tests Every month, in accordance with EA requirements, MVVD had two samples of U-IBA tested using the ESA protocol. Figure 12 shows a comparison between the results of these tests and WRC’s in-house data made up of the test results from over 21 similar incinerator facilities in the UK. A comparison between the test results of the routine test carried out by MVVD and the sample collected by the MAIB after the accident is shown in Table 1. The total metal content in the MAIB sample was estimated as 10.6%, of which the non-ferrous content was 2.6%. Parameter Routine test carried out by MVV after production Test carried out post- accident by a different lab Aluminium12 [mg/kg] 19,248 26,246 Moisture [% weight] 14.5 12.4 pH value or measure of alkalinity 12.6 11.8 Table 1: Comparison of test results of U-IBA loaded on Nortrader 12 The aluminium content reported in these tests is that found in the clinker and does not contribute towards the production of hydrogen.

25 Figure 12: Comparison of U-IBA sample from Nortrader with national averages (extracted from the WRc report) Key ANC: Acid neutralising capacity TC: Total carbon TOC: Total organic carbon LOI: Loss on ignition

26 Following the accident, at the request of Rock Solid BV, MVVD sent 10 samples of U-IBA to be tested for the release of flammable gases. These tests, carried out in accordance with UN test N.5, showed negligible gas release for all 10 samples. Marchwood Scientific Services’ report is included at Annex F. 1.13 BURGOYNES & PARTNERS INVESTIGATION Burgoynes & Partners were appointed by Nortrader’s hull and machinery insurers to complete an independent investigation of the explosion. Their report stated that the explosion was most likely to have been the result of hydrogen released from the cargo. Burgoynes tested for gas within small holes dug into a heap of U-IBA after it had been discharged ashore from Nortrader and measured 1.4% hydrogen and 0.2% carbon monoxide. The test was repeated the next day, but neither gas was registered on the gas meter. 1.14 ESTIMATION OF POTENTIAL HYDROGEN PRODUCTION 1.14.1 Production of hydrogen Hydrogen is a colourless, odourless gas that is 14 times lighter than air. It is highly volatile and burns easily in air at concentrations between 4% and 75%. Compared to other flammable gases, hydrogen has the lowest minimum ignition energy, which is a measure of the electrical energy contained in a spark across two conductors sufficient to cause ignition of the flammable mixture13. The predominant source of hydrogen generation in U-IBA is the reaction of elemental aluminium with water in alkaline conditions. Other metals such as iron, nickel and copper can also produce hydrogen, but their contribution is negligible when compared to aluminium. 1.14.2 Study to estimate hydrogen volume The MAIB commissioned the Hydrogen Hazards Unit (HHU) of London South Bank University to carry out a study to determine the approximate concentration of hydrogen gas required to generate the explosive forces experienced on board Nortrader. The MAIB provided HHU with relevant information, including the drawings and dimensions of the cargo hold and forecastle store, weights and material of hatch covers, breaking strengths of cleats and wedges used to secure the hatch covers, and photographs of the damage. The results of the tests carried out by WRc, and other U-IBA test results made available to the MAIB, were not supplied to HHU. The HHU report (Annex G) stated that the flame generated during the first explosion would have propagated into the cargo hold and that the pressure in the hold would have risen as the hydrogen air mixture within it combusted. It was estimated that after approximately 0.15 second, the hatch covers would have lifted, allowing the pressure built up in the hold to vent to atmosphere. The temperature of the hydrogen flame front was estimated to have been approximately 1000°C. 13 LECTURE. Sources of hydrogen ignition and prevention measures (Compiled by S.Tretsiakova-McNally; reviewed by D.Makarov).

27 The report concluded: In summary, considering the evidence available, the likely concentration in the cargo hold at the time of the explosion was in the rage of 6 to 9% (v/v). This equates to a volume of between 114 and 171 m3 [of hydrogen gas]. 1.15 PREVIOUS/SIMILAR ACCIDENTS 1.15.1 Accidents at sea In January 2008, two crew members died of asphyxiation in the forward store room of the dry cargo vessel Sava Lake (MAIB report 15/2008). The oxygen level had been depleted by the cargo of steel turnings, a cargo the vessel was not authorised to carry. The atmosphere in the cargo hold was inadvertently equalised with that of the forward store due to modifications carried out on the ventilation trunking. 1.15.2 Accidents ashore In 2006, following an explosion in a disused oil well that was used to store U-IBA, the Swedish Thermal Engineer Research Institute carried out a study to estimate the hydrogen generation potential of U-IBA.14 In 2009, the Health and Safety Executive in the UK reported severe injuries to a worker who was using an angle grinder to cut steel reinforcement in a concrete structure which utilised U-IBA. Several other incidents of cracks developing in concrete and overlay asphalt on roads, due to the production of hydrogen by the U-IBA used, have also been reported by organisations, including the UK’s Health and Safety Executive. 14 Maria Arm and Johanna Lindeberg, 2006: Gas generation in incinerator ash.

28 SECTION 2 - ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 THE EXPLOSION 2.2.1 Conditions for the explosion Laboratory tests have established that hydrogen gas was the only flammable gas released by the U-IBA carried on board Nortrader and that a sufficient quantity of hydrogen could have been released from the cargo to form an explosive atmosphere. Several underlying conditions combined to cause the explosion: ● ●The presence of 2.6% non-ferrous metals including aluminium, the alkalinity of the cargo, and an increased moisture content due to heavy rain during loading led to the generation of hydrogen. ● ●The 19 hours spent at anchor allowed the generated hydrogen to accumulate in the unventilated cargo hold. ● ●The unsecured and unsealed cargo hold lamp access cover in the forecastle store allowed hydrogen gas to enter the forecastle store. ● ●The closed vents and access door of the forecastle store allowed the hydrogen to accumulate in the forecastle store. ● ●The broken lug of the emergency fire pump starter panel door allowed the accumulated hydrogen to enter the space in the circuit breaker. 2.2.2 Connection between cargo hold and forecastle store The access plate for the starboard cargo hold lamp fitting had been left loosely secured by two bolts with no gasket. It is likely that the bulb in this fitting was prone to regular failure, due to the vibration generated in the area during the bow thruster operation, resulting in a need for frequent access. However, leaving the access plate slack and unsealed effectively connected the cargo hold with the forecastle store, allowing the atmosphere of both spaces to equalise. Had Nortrader been certified to carry dangerous goods, the access to the cargo lamps might have been better designed and more strictly controlled. As on Sava Lake (see 1.15.1), the gas tight integrity of the cargo hold of Nortrader was breached. On Nortrader, this allowed the lighter than air hydrogen generated by the cargo to rise and enter the forecastle store. Hydrogen being an odourless, colourless gas, the chief engineer would not have been alerted to its presence when he entered the space.

29 2.2.3 Ignition leading to explosions The propensity of hydrogen to form an explosive mixture with air across a broad range of concentrations, and the ability of a weak ignition source to ignite this mixture, resulted in the first explosion. The ignition source for the hydrogen air mixture was most likely to have been electrical arcing between the contactors in the switch gear in the emergency fire pump starter panel when the chief engineer stopped the pump. The initial explosion in the forecastle store would have led to a rapid increase in pressure in the store. This pressure was responsible for throwing the chief engineer against the starter panels before it was released through the open weathertight door. The flame front from the initial explosion propagated through the unsealed cargo lamp access cover and ignited the hydrogen air mixture in the hold. The pressure created in the hold by the resulting explosion was sufficient to break all the hatch cleats and dislodge the hatch covers. There were no subsequent explosions since the hold was then open to atmosphere, stopping any further accumulation of hydrogen. HHU estimated the flame front temperature to have been 1000°C. However, the passage of the flame front would have been very quick as the accumulated hydrogen burned. The chief engineer appeared to have suffered the full brunt of the first explosion, causing the synthetic fibre-filled coat he was wearing to melt onto his body. 2.3 DECISION TO LOAD 2.3.1 Precedent There had been 34 shipments of U-IBA from Plymouth to Beverwijk since the incinerator plant started operating in 2015. However, U-IBA was not listed in the IMSBC Code and no steps had been taken to seek approval from the competent authorities for its carriage, as required by the Code. Had the explosion on Nortrader not taken place, these unauthorised and dangerous shipments would probably have continued without change. 2.3.2 Master The master had the ultimate responsibility for accepting the cargo and could have challenged the instructions he received from H&V on the basis that the cargo was not listed in the IMSBC Code. However, Nortrader’s SMS incorrectly stated that the IMSBC Code was only to be used when carrying dangerous cargoes. This was the way that the IMSCB Code’s predecessor, the BC Code, had been used. The charter party stated that U-IBA was non dangerous and non-IMO classed (implying that it was not a recognised dangerous cargo under the IMDG Code). Therefore, it is not surprising that the master, along with the masters of the vessels carrying the previous 34 shipments, followed the loading instructions from H&V.

30 2.3.3 Shipper All cargoes of U-IBA from MVVD had been arranged by Rock Solid BV through H&V. Since U-IBA was not listed in the IMSBC Code, Rock Solid BV (the shipper of the cargo) should have provided the MCA (the competent authority at the port of loading) with the required information regarding the intended cargo, prior to the first carriage of this cargo in June 2015. However, Rock Solid BV was unaware of this requirement as it had no shipping expertise, and expected H&V to provide marine advice as required, despite there being no written agreement between them to this effect. H&V restricted its action to identifying and chartering ‘suitable’ vessels, without referring to the IMSBC Code. Its selection was based on the limited information contained in the MSDS for U-IBA provided by Rock Solid BV in October 2013 (Annex D). While this was inappropriate use of the MSDS, it did state that U-IBA could release hydrogen and other dangerous gases when in contact with water. In contrast, the charter party for Nortrader inaccurately stated that U-IBA was non dangerous. The charter party also stated that U-IBA was non-IMO classed although it would have been more accurate to state that U-IBA had not been given a BCSN since it had never been assessed, as required by the IMSBC Code. 2.3.4 UK regulations The UK Carriage of Cargoes Regulations had not been updated, and referred to the guidance provided by the superseded BC Code and not the mandatory IMSBC Code. However, this factor did not influence the skipper’s decision to load U-IBA or contribute in any way to the accident on Nortrader. Furthermore, it is considered unlikely that past shipments were carried out with the intent of exploiting any potential loophole in the UK regulations. 2.4 UNIQUENESS OF CARGO OR VESSEL 2.4.1 Cargo There had been 34 shipments of U-IBA from MVVD since June 2015, without any untoward incidents. Yet the cargo loaded on Nortrader caused a major explosion. The human consequences of this accident could have been severe, and it is extremely fortunate that the chief engineer survived the explosion. The cargo loaded on Nortrader comprised U-IBA produced at MVVD over a period of 15 consecutive days. The tests carried out by WRc on samples collected after the explosion established that the chemical and physical characteristics of the cargo were consistent with those of the U-IBA from other facilities in the UK. 2.4.2 Vessel More than a third of the vessels employed to carry the previous shipments were certified to carry dangerous cargoes. Breaches between cargo hold and adjoining spaces were therefore less likely to have existed and electrical equipment, where fitted in the cargo space, would have been intrinsically safe, thus reducing the potential for ignition sources. It is also possible that some vessels operated natural

31 or forced exhaust ventilation systems for the cargo holds while carrying U-IBA, thus protecting themselves by avoiding the potential for the accumulation of hydrogen. However, Nortrader’s design and construction, including that of its cargo hold and ventilation systems, were similar to many of the other vessels that had been engaged to carry U-IBA from MVVD. 2.4.3 Conclusion Neither the U-IBA cargo loaded onto Nortrader nor the vessel itself were unique in any way that might have contributed to this accident. 2.5 CLASSIFICATION OF U-IBA 2.5.1 Carriage of U-IBA by sea U-IBA had not been assessed or classified for carriage in bulk by sea. The only approved means of ensuring safe carriage of solid bulk cargoes is through the application of the IMSBC Code. The schedules included in Appendix 1 of the Code clearly classify cargoes and identify the necessary requirements for safe carriage. Where no schedule is listed for a specific cargo, the required classification procedure is stated at Section 4 of the Code with further guidance provided by MSC Circ.1453, MGN 512 and the guidance document Carrying solid bulk cargoes safely (Annex C). It is likely that this accident would not have happened had the IMSBC Code requirements been followed, as the dangers would have been identified and procedures for safe carriage developed and implemented prior to loading. 2.5.2 Classification of U-IBA ashore The UK’s LoW regulation, revoked in July 2015, categorised U-IBA as mirror hazardous. Monthly testing was voluntarily undertaken by the majority of incinerator facilities in the UK, based on a protocol developed by the ESA and endorsed by the EA. This protocol represented a significant step forward in the understanding of the classification and properties of IBA and was a good example of an industry led protocol. However, the protocol discounted the potential for the generation of flammable gases on the erroneous assumption that the incineration process would have rendered the U-IBA non-flammable, without considering the possibility of the ash releasing flammable gases when wet. Ashore, the release of hydrogen gas from moist U-IBA was a recognised phenomenon. The EA’s SR2012 No13 identified that wet U-IBA would release hydrogen and stated that ventilation was required to disperse the hydrogen where U-IBA was stored under cover. In view of this, the EA’s endorsement of the ESA test protocol that did not include HP3 was inappropriate. 2.6 ADEQUACY OF UNITED NATIONS TEST N.5 2.6.1 Non-replicability of test results Flammable gas release tests conducted by Marchwood Scientific Services on 10 fresh samples, and on samples collected by the MAIB after the accident, did not produce any hydrogen when subject to UN Test N.5. However, the fresh sample

32 collected by WRc from MVVD tested positive for hydrogen release. The NLS in-house method test of the MAIB sample proved positive, although a smaller quantity of hydrogen was released compared to the fresh WRc sample during this test. Sample preparation, inter-lab variability, the variability in the two test methodologies, and the aging of the sample may account for some of the difference in results. The UN Test N.5, like the NLS in-house method test, assumed that the sub samples used for testing were representative of the cargo. However, visual observation of U-IBA and its test results demonstrate that it was highly heterogeneous in nature and, despite the best efforts to reconstitute the test samples as per their original composition, it is difficult if not impossible to achieve reliable and repeatable results. 2.6.2 Rate of release of flammable gas It is evident that the release of hydrogen was localised and perhaps limited to pockets. If the results of the UN Test N.5 on the WRc sample are extrapolated, the total hydrogen produced in 6 hours (the duration after which the gas production was observed to have ceased) would amount to 6019m3. The strength of explosion of such a quantity of hydrogen would have resulted in far worse damage to the vessel. The estimation by HHU of London South Bank University, that the quantity of hydrogen present in the hold was unlikely to have exceeded 171m3, was considered to be more realistic. At the time of the explosion, the hold had been sealed for approximately 19 hours. It cannot be known whether the cargo was still producing hydrogen, but subsequent testing showed that some was still being produced in one sample. The production of 171m3 of hydrogen from 2333t of cargo equates to a minimum theoretical rate of release of approximately 0.004l/kg/h. While this is a rough approximation, it is clear that the 1l/kg/h threshold for categorising a cargo as hazardous (UN Test N.5) is inappropriate for bulk cargoes, and that the competent authorities’ discretion to require additional tests should be exercised. For purposes of transportation in bulk by sea, the release of any quantity of flammable gas should be considered dangerous. The MCA has recognised this and has categorised U-IBA as a cargo posing a chemical hazard that could give rise to a dangerous situation on a ship.

33 SECTION 3 - CONCLUSIONS 3.1 SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

Hydrogen gas was the only flammable gas released by the U-IBA carried on board Nortrader, and a sufficient quantity of hydrogen could have been released from the cargo to form an explosive atmosphere. [2.2.1]

The gas tight integrity of the cargo hold was breached, allowing the hydrogen generated by the cargo to enter the forecastle store. [2.2.2]

The ignition source for the hydrogen air mixture was most likely to have been electrical arcing between the contactors in the switch gear in the emergency fire pump starter panel. [2.2.3]

There had been 34 shipments of U-IBA from Plymouth to Beverwijk despite U-IBA not being listed in the IMSBC Code, during which no steps had been taken to seek approval from the competent authorities for its carriage. [2.3.1]

Nortrader’s SMS incorrectly stated that the IMSBC Code was only to be used when carrying dangerous cargoes. [2.3.2]

The shipper, Rock Solid BV, was unaware of the requirements of the IMSBC Code. [2.3.3]

Hudig & Veder restricted its action to identifying and chartering ‘suitable’ vessels, without referring to the IMSBC Code. [2.3.3]

Had the IMSBC Code requirements been followed it is likely this accident would not have happened as the dangers would have been identified and procedures for safe carriage developed and implemented prior to loading. [2.5.1]

For purposes of transportation in bulk by sea, the release of any quantity of flammable gas should be considered dangerous. [2.6.2] 3.2 SAFETY ISSUES NOT DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

The UK Carriage of Cargoes Regulations had not been updated and referred to the guidance provided by the superseded BC Code and not the mandatory IMSBC Code. [2.3.4]

The test protocol adopted by the ESA and endorsed by the EA erroneously discounted the potential for the generation of flammable gases. [2.5.2]

The Environment Agency’s endorsement of the ESA test protocol that did not include HP3 was inappropriate. [2.5.2]

Due to its heterogeneous nature, it is difficult if not impossible to achieve reliable and repeatable results using UN Test N.5 on U-IBA. [2.6.1]

The 1l/kg/h limit suggested by (UN Test N.5) is inappropriate for bulk cargoes. [2.6.2]

34 SECTION 4 - ACTION TAKEN 4.1 MAIB ACTIONS The Marine Accident Investigation Branch has: ● ●Published a safety flyer intended to promulgate the lessons learned from this accident with particular emphasis on the application of the IMSBC Code. ● ●Worked with several trade bodies including the International Group of Protection & Indemnity Clubs, International Chamber of Shipping, International Association of Dry Cargo Ship Owners, and Institute of Chartered Shipbrokers to help disseminate the safety flyer. ● ●Provided the MCA with the results of laboratory tests and technical research conducted during this investigation to support its submissions to the IMO. 4.2 ACTIONS TAKEN BY OTHER ORGANISATIONS The Maritime and Coastguard Agency has: ● ●Set up a tripartite agreement between the UK, the Netherlands and several other Administrations for the safe carriage of U-IBA. ● ●Proposed two papers to the International Maritime Organization: to include U-IBA in the schedule of authorised cargoes in the IMSBC Code; and to raise the issue of the inadequacy of UN test N.5 for non-homogeneous bulk cargoes. MVV Environment Devonport Limited has: ● ●Installed hydrogen gas monitors and a hydrogen warning system in the U-IBA storage bunker. ● ●Revised the risk assessment and procedures for U-IBA storage and movement from the site. The Environment Services Association has: ● ●Published a notice on its website about incinerator bottom ash, which includes the hazard of hydrogen release. ● ●Commissioned further tests on un-ground U-IBA samples for the generation of hydrogen gas.

35 SECTION 5 - RECOMMENDATIONS The Maritime and Coastguard Agency and the Environment Agency are recommended to: 2017/153 Work collaboratively to identify reliable methods and protocols for testing non-homogeneous solid bulk cargoes for the property of evolving flammable gases when wet. The Maritime and Coastguard Agency is recommended to: 2017/154 Update The Merchant Shipping (Carriage of Cargoes) Regulations 1999 with appropriate references to the IMSBC Code. Hudig & Veder BV is recommended to: 2017/155 Review its operating procedures to ensure that the requirement to apply the provisions of the IMSBC Code to all bulk cargoes is clear. NTO Shipping GmbH & Co.KG is recommended to: 2017/156 Review its safety management system to ensure that the requirement to apply the provisions of the IMSBC Code to all bulk cargoes is clear. Safety recommendations shall in no case create a presumption of blame or liability

Marine Accident Report

FD06050 ｜ 2017_英国_man-overboard-from-ocean-rowing-boat-toby-wallace-with-loss-of-1-life__MAIBInvReport2_2017.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2017 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H ACCIDENT REPORT VERY SERIOUS MARINE CASUALTY REPORT NO 2/2017 FEBRUARY 2017 Report on the investigation of the fatal man overboard from the ocean rowing boat Toby Wallace North Atlantic Ocean 14 February 2016

Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2017 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch Spring Place 105 Commercial Road Southampton Email: maib@dft.gsi.gov.uk United Kingdom Telephone: +44 (0) 23 8039 5500 SO15 1GH Fax: +44 (0) 23 8023 2459 Press enquiries during office hours: 01932 440015 Press enquiries out of hours: 020 7944 4292

CONTENTS GLOSSARY OF ABBREVIATIONS AND ACRONYMS SYNOPSIS 1 SECTION 1 - FACTUAL INFORMATION 2 1.1 Particulars of Toby Wallace and accident 2 1.2 Narrative 3 1.2.1 Events leading to the man overboard 3 1.2.2 Crew actions 3 1.3 Hindcast 8 1.4 Crew 8 1.4.1 Skipper 8 1.4.2 First mate 9 1.4.3 Pay per place crew 9 1.4.4 Preparation 9 1.4.5 Watch routine 10 1.4.6 Previous man overboard 10 1.5 Boat design and construction 10 1.6 Onboard equipment 11 1.6.1 General 11 1.6.2 Lifejackets, tethers and personal locator beacons 11 1.6.3 Ankle leashes 11 1.7 Oceanus Rowing Ltd 11 1.8 Ocean rowing 12 1.9 Atlantic Challenge race rules 12 1.10 Local requirements 13 1.11 Pleasure vessel regulations 13 1.11.1 Definition 13 1.11.2 Applicable regulations 14 1.12 Governing body 14 1.13 Previous accidents 15 1.13.1 General 15 1.13.2 Fire Ant 15 SECTION 2 - ANALYSIS 16 2.1 Aim 16 2.2 Loss overboard 16 2.3 Boat design 16 2.4 Carriage and use of safety equipment 17 2.4.1 Requirements 17 2.4.2 Ankle leash 17 2.4.3 Lifejackets and tethers 17 2.4.4 Personal locator beacon 18 2.4.5 Man overboard 18 2.5 Preparation and training 19 2.6 Emergency response 19 2.7 Governance 19 2.7.1 Commercial operation 19 2.7.2 Best practice 20

SECTION 3 - CONCLUSIONS 22 3.1 Safety issues directly contributing to the accident that have been addressed or resulted in recommendations 22 3.2 Other safety issues not directly contributing to the accident 22 SECTION 4 - ACTIONS TAKEN 23 4.1 Actions taken by other organisations 23 SECTION 5 - RECOMMENDATIONS 24

FIGURES Figure 1

Toby Wallace – position on 14 February 2016 Figure 2

Rowing positions Figure 3

Ankle leash and foot straps Figure 4

Toby Wallace before abandonment Figure 5

Toby Wallace’s crew climbing on board Sea Pearl Figure 6

Toby Wallace – abandoned and drifting

Figure 7

An ocean rowing boat fitted with guardrails

Figure 8

An ocean rowing boat fitted with guard wires Figure 9

Use of a safety harness, tethers and jackstay

GLOSSARY OF ABBREVIATIONS AND ACRONYMS

AIS

Automatic Identification System EPIRB

Emergency Position-Indicating Radio Beacon GPS

Global Positioning System kt

knot m

metre MGN

Marine Guidance Note MHz

megahertz mm

millimetre MRCC

Maritime Rescue Co-ordination Centre nm

nautical mile PLB

Personal Locator Beacon RYA

Royal Yachting Association SAR

Search and Rescue UTC

Universal Time Co-ordinated VHF

Very High Frequency TIMES: all times used in this report are UTC (-1) unless otherwise stated

1 SYNOPSIS On 14 February 2016, the 10.8m ocean rowing boat Toby Wallace was on passage in the North Atlantic Ocean, between the Canary Islands and Barbados. Shortly after 2300, a rower was swept overboard by a large wave and he soon became separated from the boat. The seven rowers remaining on board were unable to stop the boat, and they eventually lost sight of him. The following day, French and Portuguese fixed wing aircraft and the Bahamas registered cargo ship Sea Pearl searched the area, but the missing rower was not found. Toby Wallace’s crew were recovered on board Sea Pearl, leaving the abandoned rowing boat to drift. Sea Pearl arrived in Macapa, Brazil, 6 days later. The investigation identified that: ● ●The rower had not been adequately tethered to the boat. ● ●The boat’s low gunnels1 afforded little protection against a person falling overboard. ● ●The rower became separated from the boat because the leash he was wearing detached from his ankle and the boat could not be stopped in the prevailing conditions. ● ●The rower’s likelihood of being found was reduced because he was not wearing a lifejacket or a personal locator beacon. ● ●The boat was operating on a commercial basis but it was not certified as a commercial vessel. This was the second occasion in 2 days in which rowers in an ocean rowing boat operated by Oceanus Rowing Ltd had been rescued in the North Atlantic Ocean by a passing merchant vessel. Earlier on 14 February 2016, the four crew of the ocean rowing boat Fire Ant were rescued by the Liberia registered bulk carrier Rio Grita after the boat was damaged in rough seas. Ocean rowing is a niche sport that has grown in popularity, but the boats used and their operation, is largely unregulated. They are also vulnerable when on ocean passage, where assistance is rarely immediately at hand. Although elements of best practice for the sport have been established over the years, these have not yet been formalised or applied consistently. Recommendations have been made to British Rowing, the governing body for ocean rowing in England, and to the Maritime and Coastguard Agency that are intended to improve the governance of UK registered ocean rowing boats and ensure minimum safety standards on board those that are commercially operated. A recommendation made to Oceanus Rowing Ltd is aimed at improving the safety of its boats on ocean crossings in the future. 1 Gunnel/gunwale; upper edge of the boat’s sides.

2 SECTION 1 - FACTUAL INFORMATION 1.1 PARTICULARS OF TOBY WALLACE AND ACCIDENT SHIP PARTICULARS Vessel’s name Toby Wallace Flag United Kingdom Classification society Not applicable IMO number/fishing numbers Not applicable Type Ocean Rowing Boat Registered owner Privately owned Manager(s) Oceanus Rowing Ltd Construction Carbon fibre – epoxy resin Year of build 2011 Length overall 10.8m Registered length Not applicable Gross tonnage Not applicable Minimum safe manning Not applicable Authorised cargo Not applicable VOYAGE PARTICULARS Port of departure Puerto De Mogan, Gran Canaria Port of arrival Barbados Type of voyage International Cargo information Not applicable Manning 8 MARINE CASUALTY INFORMATION Date and time 14 February 2016 at 2301 UTC (-1) Type of marine casualty or incident Very Serious Marine Casualty Location of incident 18° 28.89N 38° 59.19W Place on board Open rowing deck Injuries/fatalities 1 fatality Damage/environmental impact None Ship operation On passage Voyage segment Mid water External & internal environment Wind east-north-east at Beaufort Force 6; sea state 5 (3m significant wave height); sea temperature 22°C; weather rain Persons on board 8

3 1.2 NARRATIVE 1.2.1 Events leading to the man overboard On 28 January 2016, the 10.8m ocean rowing boat Toby Wallace departed from Puerto De Mogan, Gran Canaria for passage to Barbados, West Indies (Figure 1). On board were eight rowers who aimed to complete the voyage in less than 31 days. Toby Wallace’s crew were divided equally into two watches. The skipper was in charge of one watch and the first mate was in charge of the other. The rowers in each watch were allocated one of four rowing positions (Figure 2). The stroke2 sat furthest aft (position 1). When on watch, the skipper sat immediately behind the stroke (position 2) and the mate sat in the bow seat (position 4). By 14 February, Toby Wallace had made good progress in steady force 4-5 east-north-east trade winds. The boat was making good between 3kts and 4.5kts and was approximately half-way across the North Atlantic Ocean, 1500nm from land. At 2300, Toby Wallace was heading downwind and down sea and was making good a course of 253° at 3.9kts3.The ‘on-watch’ crew were seated with their feet secured to footplates with straps. Each of the rowers had a leash attached around their right ankle (Figure 3), the other end of which was secured to a jackstay4 that ran along the port side of the deck. As the bow oarsman (position 4) took a planned 3 minute break from rowing, the skipper saw a large wave approaching off the port quarter and shouted “big wave”5 to warn the crew. The wave broke over Toby Wallace and swept all of the rowers, apart from the skipper, off their seats. The ‘stroke’, Mike Johnson, landed on the starboard side of the deck from where he then rolled over the boat’s starboard gunnel into the sea. Mike’s ankle leash detached and he was quickly separated from Toby Wallace as the rowing boat continued to move ahead through the water. The ankle leash was recovered inboard. Mike was wearing oilskins, training shoes and a head torch fitted with a red lamp. 1.2.2 Crew actions The skipper immediately shouted “man overboard” and attempted to steer Toby Wallace to port in order to position the boat across the wind and sea. He also told the rowers to ‘hold water’6. Meanwhile, the off-watch crew in the aft cabin input a man overboard mark in the global positioning system (GPS) plotter and broadcast a “Mayday”7 via very high frequency (VHF) radio. 2 The stroke rower sets the stroke rate and rhythm for the rest of the crew to follow. 3 Toby Wallace’s position and course, and speed over the ground were periodically transmitted via satellite. 4 In this context ‘Jackstay’ means a strong line, usually of flat webbing or a wire stay running fore and aft along the sides or a deck of a boat to which a safety harness can be attached. 5 ‘Big wave’ was the recognised verbal warning on board to alert the crew that the boat was about to be swamped. 6 ‘Hold water’ – a command to stop a rowing boat by squaring the oar blades in the water. 7 “Mayday” – the international distress message (spoken).

4 Reproduced from Admiralty Chart BA 4015-0-1 by permission of the Controller of HMSO and the UK Hydrographic Office Figure 1: Toby Wallace – position on 14 February Toby Wallace 2300 on 14 February 2016 Fire Ant 0800 on 13 February 2016 Cape Verde Ponta Delgada Dakar Puerto de Mogan

5 Figure 2: Rowing positions 3 4 Forward cabin Aft cabin Figure 3: Ankle leash and foot straps Ankle leash Jackstay Foot strap Ankle strap Watch leader 2 Stroke 1

6 Toby Wallace was turned across the sea but the crew were unable to row the boat into the wind and they could see that the distance to the red light from Mike’s head torch was increasing. To try and stop the boat, a drogue8 and a parachute anchor9 were deployed over the port and starboard sides respectively. The parachute anchor soon became detached, so the crew attached various items including the liferaft valise and lifejackets to a line and threw them over the side. Water ballast was also added to some compartments to make the boat heavier. At approximately 2322, the skipper activated the emergency position-indicating radio beacon (EPIRB)10. The EPIRB’s distress signal was received by the Maritime Rescue Co-ordination Centre (MRCC) Falmouth, UK, which confirmed the boat’s location with the duty officer at Oceanus Rowing Ltd, Toby Wallace’s operator. The duty officer also provided MRCC Falmouth with the boat skipper’s satellite telephone contact details. At 0002 the following day, MRCC Falmouth contacted Toby Wallace’s skipper via satellite telephone. He advised that the wind was 35kts and that he could not manoeuvre Toby Wallace towards the man overboard. He also advised that the boat’s AIS receiver was showing no other vessels in the area. Toby Wallace’s crew last saw the red light from Mike’s head torch at about 0022. The boat’s crew released rocket flares at regular intervals and tied a light on the end of an upright oar to act as a beacon. By 0130, the crew were in a state of shock. Waves continued to break over Toby Wallace’s deck and two of the rowers started to suffer from hypothermia. At 0300, Toby Wallace was still drifting at a speed of 1.6kts to the west. Eventually, all seven remaining crew took shelter in the forward and aft cabins. As Toby Wallace was in the Portuguese search and rescue (SAR) region, MRCC Falmouth passed responsibility for co-ordination of the incident response to MRCC Ponta Delgada, Azores (Figure 1). MRCC Ponta Delgada arranged for a French Falcon aircraft from Dakar, Senegal, and a Portuguese P3C Orion aircraft from Portugal to search for the missing rower. The Bahamas registered cargo ship Sea Pearl, which was about 113nm to the east of Toby Wallace, was also diverted. The Falcon aircraft departed from Dakar at 0550 and, having refuelled in Cape Verde, it was on scene between 1100 and 1230. Sea Pearl arrived at about 1230 and the P3C Orion joined the search from about 1430 until 1530, but Mike Johnson was not found. At 2000, Toby Wallace’s crew were recovered on board Sea Pearl (Figures 4 and 5), leaving Toby Wallace to drift (Figure 6). Sea Pearl arrived in Macapa, Brazil, 6 days later. The last transmission received from Toby Wallace’s AIS was on 9 May, when the boat was approximately 500 miles north-east of the British Virgin Islands. 8 A drogue is a conical shaped bag that is towed astern in order to reduce the speed of advance. 9 A parachute anchor is typically made of woven material that is shaped like a round parachute and is intended to be deployed over the bow on a bridle arrangement in order to stop a boat head into wind. 10 EPIRBs alert search and rescue services by transmitting a coded message on the 406MHz distress frequency. Some EPIRBs also have built-in GPS that provide positional information. The EPIRB carried on board Toby Wallace was not fitted with a GPS.

7 Figure 4: Toby Wallace before abandonment Figure 5: Toby Wallace’s crew climbing on board Sea Pearl

8 1.3 HINDCAST A Met Office hindcast report for Toby Wallace’s position at the time of the man overboard concluded: All available evidence suggests that Toby Wallace was impacted by a steady east or east-northeast trade wind of force 5 or 6 during the 14th and 15th February 2016. Significant wave height was of approximately 3 metres with a zero upcrossing period of 6 or 7 seconds11. An 11 second swell of 1.0-1.5 metres was aligned with the direction of wind wave, but crossed by a southeast going swell of approximately 1.0 metres with a peak period of 15 seconds. In addition, this swell train was opposed by a weak northwest going current modelled 0.3 to 0.4 knots. These ‘crossing seas’ would have made for steeper than normal wave scenario. Maximum individual wave heights, crest to trough would have been double significant wave height within a 6-hour period; 6.6 metres on the 14th and 6.2 metres on the 15th. [sic] 1.4 CREW 1.4.1 Skipper Toby Wallace’s skipper was the boat’s registered owner. He was 43 years old and a UK national who had completed nine ocean crossings, four of which had been on board Toby Wallace. The skipper held a Royal Yachting Association (RYA) 11 MAIB note. Zero upcrossing period is a method of measuring wave periodicity. Figure 6: Toby Wallace – abandoned and drifting

9 powerboat qualification, and had completed RYA training courses in sea survival, VHF radio and Ocean Yachtmaster theory. He received a payment/salary for skippering Toby Wallace on its Atlantic crossing. Toby Wallace’s skipper had been very actively involved in ocean rowing since 1997 and Toby Wallace was the last of 64 ocean rowing boats that he had designed and built. In 2003, he established ‘Woodvale Events Ltd’ and purchased the commercial rights to operate the ‘Atlantic Rowing Race’. Between 2004 and 2011, through Woodvale Events Ltd and then Woodvale Challenge Ltd, the skipper operated six ocean rowing races, developed safety rules for ocean racing and negotiated and agreed minimum boat equipment and crew qualification requirements with the Tenerife Port Authority12. In 2007, the skipper organised the first ‘pay per place’ ocean rowing expedition. 1.4.2 First mate The first mate was an experienced ocean rower, having completed four previous Atlantic rowing events. He was 68 years old and had completed RYA training courses in sea survival, first-aid, VHF radio and Ocean Yachtmaster theory. The first mate was a late addition to the crew due to the withdrawal of the intended person for this role. He met Toby Wallace’s paying crew in Puerto De Mogan prior to the start of the row. 1.4.3 Pay per place crew The remainder of Toby Wallace’s crew comprised UK, French and Zimbabwean nationals aged between 20 and 68. One of the crew was female. None held any maritime qualifications or had completed any recognised maritime training courses but all had sporting and/adventure activity backgrounds. Each was charged £15000 towards the expenses incurred in preparing, equipping and transporting Toby Wallace from the UK to Gran Canaria and from Barbados back to the UK. One of the crew had previously had surgery for a percutaneous transluminal coronary angioplasty13 as well as having a full knee arthroplasty14. He was initially incapacitated through seasickness but was then unable to row due to problems with his knees. The deceased, Mike Johnson, was a 21 year-old Zimbabwean and UK national who had been involved in rowing since 2007 and had competed in a number of junior world rowing championships. His other sporting interests included cross-country running and cycling. 1.4.4 Preparation Between 22 and 23 August 2015, Toby Wallace’s skipper and four of the paying crew had rowed together on board the ocean rowing boat Britannia, which was similar to Toby Wallace, in Colwyn Bay, Wales. All eight members of the crew rowed together on several occasions and practised a man overboard recovery, using a bucket to represent a person in the water, in calm conditions in Puerto De Mogan before setting off for Barbados. 12 The Port Authority of Santa Cruz de Tenerife also manages the ports of Santa Cruz La Palma, Los Cristianos, San Sebastian de La Gomera and La Estaca. See also paragraph 1.10. 13 Surgical treatment to open a narrowed or blocked artery using a wire mesh tube called a “stent”. 14 Replacement of knee joint with an artificial joint.

10 1.4.5 Watch routine After departing from Gran Canaria on 28 January 2016, Toby Wallace’s crew worked a rotating watch routine with each day divided in to 11 periods of between 1 and 4 hours. Watch changes were achieved by the on-going rowers relieving the off-going rowers one at a time. The crew were only able to move about the boat by crawling on their hands and knees. When not on-watch, the stroke and the watch leader rested in the aft cabin and the rowers allocated the forward positions (positions 3 and 4) rested in the forward cabin. Prior to the accident, Mike Johnson had assisted the skipper in the repair of the boat’s water maker when off-watch during 14 February and had not rested for about 8 hours at the time he was lost overboard. 1.4.6 Previous man overboard During the first 2 weeks of the crossing, two of Toby Wallace’s paying crew fell overboard on separate occasions while moving on the open deck in daylight. Both crewmen managed to hold onto the boat after falling into the water and were recovered back on board without injury. 1.5 BOAT DESIGN AND CONSTRUCTION Toby Wallace was built in 2011 by its skipper. The boat was initially named Titan, and was entered on the UK small ships register in 2012 as a pleasure vessel. It was renamed Toby Wallace in February 2014 and had successfully completed four Atlantic crossings. Toby Wallace was made from carbon fibre/epoxy resin and was 10.8m in length with a breadth of 1.8m and a draught of 0.45m. The boat was based on an ocean rowing boat design that made it self-righting in the event of capsize. It was also manoeuvrable when steering downwind and it could be steered across the sea without difficulty. However, it is reported that Toby Wallace could not be turned or steered into wind in wind speeds above 18kts. Factors influencing the boat’s design included the presumed prevalence of capsize as the cause of previous rowing fatalities and the skipper’s personal experience of capsize both as a rower and as an ocean race organiser. Toby Wallace’s open deck was self-draining and had four rowing positions, each fitted with an aft-facing sliding seat and a fixed footplate with foot straps (Figure 3). Oars were mounted into rowlocks on the gunnels. In common with the other 63 ocean rowing boats built by Woodvale Events Ltd and Woodvale Challenge Ltd, the height of the gunnels above the deck was approximately 300mm in order to keep the oars relatively level when rowing. A rudder on the aft transom was controlled via an electronic pilot control system or by lanyards attached to the tiller. Two dagger boards were fitted towards the bow to aid directional stability and manoeuvrability. Electrical power for navigation aids, radios and a water maker was provided by solar panels and a wind generator.

11 1.6 ONBOARD EQUIPMENT 1.6.1 General The equipment carried on board Toby Wallace, including lifesaving and safety equipment, was based on detailed lists developed by the boat’s owner/skipper when employed by Woodvale Challenge Ltd (see paragraphs 1.7 and 1.8). 1.6.2 Lifejackets, tethers and personal locator beacons Eight manual inflation lifejackets, each fitted with a gas bottle and a lifting becket, were carried on board Toby Wallace, four being located in each of the forward and aft cabins. Four tethers that could be attached to the lifejackets were also carried. Neither the lifejackets nor the tethers were routinely used by the crew. The vessel was equipped with an EPIRB, and one personal locator beacon (PLB)15 was carried on board in an emergency grab bag. 1.6.3 Ankle leashes The ankle leashes were provided on board Toby Wallace (Figure 3) to prevent the rowers from becoming separated from the boat in the event of capsize. The leashes were approximately 3m in length, which was sufficient to allow the wearer to keep clear of the boat as it self-righted. The leashes were manufactured from plastic coated cord with straps fitted with Velcro16-type hook and loop fastenings at each end. The ankle strap was wider than the strap secured to the jackstay line on the port side of the deck. The jackstay line was made from 10mm Dynema rope (breaking strength 4650kg). The rowers wore the leashes during the hours of darkness but they were not routinely worn in daylight hours. During the watch change at 2200 on 14 February 2016, the skipper carried out a leash check soon after his watch were seated. Each rower confirmed that their leash was attached to their ankle and to the line running along the deck. 1.7 OCEANUS ROWING LTD Oceanus Rowing Ltd (formerly Woodvale Challenge Ltd) was established in 2015 to provide individuals with the opportunity to participate in ocean rowing. The company was owned by an experienced ocean rower who was the wife of Toby Wallace’s skipper. The boats operated by the company were registered as privately owned, but they were also listed among its financial assets. During ocean crossings, the company employed a duty officer who was responsible for monitoring the boats’ progress, acting as a point of contact and managing social media and publicity. The company’s website included: Oceanus Rowing provides individuals from a wide variety of ages, abilities and backgrounds with the chance to row an ocean. 15 A PLB is a waterproof and buoyant personal safety device designed to enable SAR services to quickly locate an individual in an emergency. When activated, a PLB transmits a coded message, which includes positional information on the 406MHz distress frequency. 16 Velcro is a registered trademark.

12 We have a fleet of specifically designed ocean rowing boats to offer a unique experience for prospective ocean rowers on-board one of our incredible boats. Rowers can apply for places and pay a single fee on a Pay-Per-Place basis, which covers all your costs to do an ocean row (excludes flights & port accommodations, which we can help with discounted rates). It is an ideal solution for adventurous people who may be unable to devote the time or be able to run their own ocean rowing campaign – it is a cost effective, fun and exciting way to do an ocean row with an outstanding, knowledgeable and experienced professional skipper to give you a one off experience! So join us today – be involved at the cutting edge of ocean rowing with the added incentive of having an adventure of a lifetime rowing an ocean! Ocean rowing is a dangerous sport…and that he/she undertakes the challenge at his/her own risk. When assessing applicants’ suitability to complete an ocean crossing, Oceanus Rowing Ltd took into account their sporting background, sense of adventure, previous rowing experience and their ability to work in a team in confined conditions over a prolonged period of time. 1.8 OCEAN ROWING Ocean rowing is a growing endurance sport that combines physical and mental stamina with technical expertise. The sport has flourished since the first modern races in the 1990’s, although it attracts fewer participants than other physical challenges such as polar expeditions. To date, fewer than 650 successful ocean crossings have been recorded17. Prospective ocean rowers are able to participate in the sport by purchasing a boat, assembling a crew and deciding when and where to travel from and to. Alternatively, they are able to pay for a place on board a boat operated by an ocean rowing company, or they can participate in an organised event or race. The first trans-Atlantic rowing race was held in 1997. Since then, races have been held on a regular basis. Between 2003 and 2012, the rights to organise an annual sponsored race across the Atlantic Ocean were owned by Woodvale Challenge Ltd. Since then, the race has been organised by Atlantic Campaigns SL, based in Spain and has been named the ‘Atlantic Challenge’. 1.9 ATLANTIC CHALLENGE RACE RULES The Atlantic Challenge race starts from La Gomera, Tenerife, and finishes in Antigua, West Indies. The rules that have been adopted by Atlantic Campaigns SL include, inter alia, crew qualification and training requirements and requirements for boat survey, boat registration, boat stability data, and equipment. The current rules were based on those initially developed by Woodvale Challenge Ltd between 2003 and 2011. 17 Data on ocean rowing events and records is kept by the Ocean Rowing Society, which was established in 1983.

13 With regard to qualifications and training, the rules require that each competitor: ● ●Shall have attended the Ocean Rowing Course. ● ●Shall hold a valid Short Range Radio Certificate (VHF). ● ●Shall hold a valid RYA Basic Sea Survival course completion certificate. ● ●Shall hold a valid First Aid at Sea certificate. Each team must also have: Rowed their boat for a minimum of 72 hours, with at least 12 hours of the 72-hour period conducted prior to arrival in La Gomera and at least 12 hours of the 72-hour period conducted during the hours of darkness. During this time all equipment shall be tested. With regard to life-saving equipment, when on deck each rower is required to wear a PLB and either a safety harness and line or a combination lifejacket with a harness and safety line. A support yacht is provided during the Atlantic Challenge. The primary purpose of the support yacht, which carries limited supplies, is to reduce the need to call for third-party assistance. 1.10 LOCAL REQUIREMENTS In light of the experience gained from the numerous rowers getting into difficulty shortly after starting to cross the Atlantic Ocean, the Port Authority of Santa Cruz do Tenerife introduced a number of conditions for departure. These requirements were negotiated with Woodvale Challenge Ltd and included that boats be registered with a national administration, skippers to have completed an RYA Ocean Yachtmaster (theory) course and sea survival training, and that boats be suitably equipped for ocean crossings. Similar requirements did not apply in other port areas in the Canary Islands such as Puerto De Mogan. 1.11 PLEASURE VESSEL REGULATIONS 1.11.1 Definition The Merchant Shipping (Vessels in Commercial Use for Sport or Pleasure) 1998 (SI 1998/2771), as amended, defines a pleasure vessel as: (a) any vessel which at the time it is being used is: (i) (aa) in the case of a vessel wholly owned by an individual or individuals, used only for the sport or pleasure of the owner or the immediate family or friends of the owner; or

14 (bb) in the case of a vessel owned by a body corporate, used only for sport or pleasure and on which the persons on board are employees or officers of the body corporate, or their immediate family or friends; and (ii) on a voyage or excursion which is one for which the owner does not receive money for or in connection with operating the vessel or carrying any person, other than as a contribution to the direct expenses of the operation of the vessel incurred during the voyage or excursion; or (b) any vessel wholly owned by or on behalf of a members’ club formed for the purpose of sport or pleasure which, at the time it is being used, is used only for the sport or pleasure of members of that club or their immediate family, and for the use of which any charges levied are paid into club funds and applied for the general use of the club; and (c) in the case of any vessel referred to in paragraphs (a) or (b) above no other payments are made by or on behalf of users of the vessel, other than by the owner. 1.11.2 Applicable regulations Marine Guidance Note (MGN) 538 (M) - Pleasure Vessels – Guidance and Applicable UK Regulations, provides an overview of the Merchant Shipping legislation that is applicable to pleasure vessels. Paragraph 3.2 of the MGN states: Pleasure vessels less than 13.7 metres in length are not covered by any statutory requirements for the carriage of Firefighting Equipment or Life-Saving Appliances, however, the owner has a duty of care under health and safety legislation where people are employed on-board (see Section 6.4). Owners and skippers should refer to the RYA and British Marine for guidance of any safety equipment for their vessels. 1.12 GOVERNING BODY With regard to the use of commercial vessels for sport or pleasure, MGN 280 (M) - Small Vessels in Commercial Use for Sport or Pleasure, Workboats and Pilot Boats – Alternative Construction Standards includes: 3.4.1. The Code deals with the safety of the vessel and its occupants but not sport or pleasure activities involving special safety requirements. 3.4.2. The objectives for sport have been set out by Government. The principle of self-determination for sports bodies has been encouraged to the extent that when it has been necessary to impose some form of control on such bodies

• such as safety or environmental matters - the policy has usually been to

encourage the bodies to adopt voluntary codes or procedures which would have the same effect as regulation. 3.4.3. In 1990, the Minister for Sport commissioned a review into safety in water sports. The review concluded that the current system of self-regulation developed by the governing bodies of sport is sufficient to meet their responsibility for the safety of sports participants…

15 Ocean rowing boats are not specifically covered by MGN 280 (M), or any of the extant codes of practice for small craft. British Rowing is the national governing body (England) for rowing and has over 30 000 members. It promotes the safety of the sport and provides advice to rowing clubs on planning events and on risk management. To date, British Rowing has had little involvement with ocean rowing. 1.13 PREVIOUS ACCIDENTS 1.13.1 General Since 1990, there have been 204 unsuccessful attempts to row across an ocean that necessitated crew rescue. Records indicated that up to 1999, five fatalities had occurred during ocean rowing crossing attempts. There has been one fatality between 1999 and 2015. There is insufficient evidence available to accurately determine the circumstances of these fatalities. 1.13.2 Fire Ant During the morning of 13 February 2016, Oceanus Rowing Ltd’s ocean rowing boat Fire Ant capsized several times when 400nm north-west of Cape Verde (Figure 1). During the capsizes, the all-female crew of four rowers lost their oars and the boat’s steering system and the GPS were damaged. Fire Ant’s skipper, who was the owner of Oceanus Rowing Ltd, alerted the company’s duty officer via satellite telephone. In turn, the duty officer contacted the coastguard operations centre, Falmouth, and requested assistance. After Fire Ant’s position had been determined by the manual release of the boat’s EPIRB, the coastguard broadcast a “Mayday” to all ships in the area via Enhanced Group Call via Inmarsat C, an element of the Global Maritime Distress and Safety System. The Liberia registered bulk carrier, Rio Grita, responded to the “Mayday” broadcast and diverted towards Fire Ant. Sixteen hours later, the ocean rowing boat’s crew were recovered on board Rio Grita, which arrived in Port Alfred, Canada on 22 February 2016.

16 SECTION 2 - ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 LOSS OVERBOARD Mike Johnson was swept from his seat by a large wave that broke over and swamped Toby Wallace. He then rolled over the boat’s low gunnel because he was not adequately tethered. The straps on the footplate by his seat were a rowing aid and the ankle leash was being worn to prevent him separating from the boat in the event of capsize. Neither the foot straps nor the leash, which was 3m in length, were intended to prevent him from falling overboard. After falling overboard, Mike’s ankle leash detached and he quickly became separated from Toby Wallace. The crew were unable to stop the boat from drifting away from him and, although their actions to raise the alarm were timely, it took 11 hours for SAR units to reach Toby Wallace’s remote position in the mid-Atlantic. The sea water temperature was 22°C and Mike was possibly conscious when the red light from his head torch was last seen. That Mike was not wearing a lifejacket or a PLB reduced considerably the likelihood of him surviving and being found. 2.3 BOAT DESIGN Many experienced ocean rowers consider the greatest risk to ocean rowing boats and their crews is capsize. Consequently, like many other ocean rowing boats, Toby Wallace was designed to self-right following capsize. However, it is apparent from the circumstances of this accident, that the risk of a rower falling or being swept overboard from Toby Wallace was also significant. Toby Wallace’s gunnels were only 300mm above the deck in order to keep the rowers, the rowlocks and the sea water surface on a similar plane. The gunnels offered virtually no protection to the crew from the risk of falling overboard. Although the risk of falling overboard was mitigated to some degree by the crew crawling when moving on the deck, it would have been an ever-present concern when operating in rough seas. Toby Wallace’s design also increased the difficulty of recovering a man overboard. In the force 5 to force 6 winds experienced on 14 February, the boat could not be turned into the wind. Further, despite the use of the rudder, holding water, the deployment of a drogue, a parachute anchor and other items, and ballasting to make the boat heavier it could not be stopped in the water. As a result, over the near 1.5 hours that Mike Johnson’s head torch remained visible, Toby Wallace drifted further away from him, and the boat’s crew were unable to attempt a recovery.

17 2.4 CARRIAGE AND USE OF SAFETY EQUIPMENT 2.4.1 Requirements Toby Wallace had been registered by its owner as a pleasure vessel as defined by The Merchant Shipping (Vessels in Commercial Use for Sport or Pleasure) Regulations 1998. As such, and as the boat was also less than 13.7m, it was not required to carry lifesaving equipment. The lifesaving and safety equipment carried on board Toby Wallace was based on the equipment lists developed by Woodvale Challenge Ltd and the skipper’s previous experiences of ocean rowing, rather than on any regulatory requirements. 2.4.2 Ankle leash The use of ankle leashes on board Toby Wallace was a practical measure that was primarily intended to keep a rower attached to the boat following capsize. Ankle leashes are frequently used as the ankle is a suitable location for a long tether and the likelihood of chaffing is reduced. Ankle leashes also do not interfere with body movement while rowing. However, the type of ankle leashes used on board Toby Wallace were the type widely used by surfers and paddle boarders. They were not manufactured or tested to meet any national or international safety standards and the velcro-type fastenings would have been prone to degrade over time through use, immersion and ultra-violet light. In addition, although the breaking strength of the jackstay running along the deck to which they were attached was over 4.5t, the breaking strength of its fixing points was not known. Therefore, both the leash and the jackstay had potential weaknesses that impacted on their reliability. At the start of the watch at 2200 on 14 February 2016, Mike Johnson had confirmed that his ankle leash was attached to his ankle and the jackstay. Therefore, unless he removed his ankle strap later in the watch, his rapid separation from Toby Wallace after falling overboard indicates that the Velcro-type ankle strap on his leash had come undone or slipped off his leg. When Mike fell into the water, Toby Wallace was still making approximately 3kts of headway. Consequently, it is possible that sufficient force was generated to separate the Velcro-type fastenings, which might have degraded over time in the marine environment or might not have been securely fastened from the outset. 2.4.3 Lifejackets and tethers Toby Wallace was equipped with eight manual inflation lifejackets and four tethers. The vessel was not carrying any form of safety harnesses, and so the tethers could only be used in combination with the lifejackets. Mike was wearing oilskins at the time of the accident, but neither a lifejacket nor tether. Had he been doing so, the tether would have kept him beside Toby Wallace and he would have been able to inflate the lifejacket if he needed buoyancy before the crew helped him back on board. The prolonged wearing of lifejackets while ocean rowing is commonly considered to be impractical. The wet and salty environment can lead to clothing and lifejackets chafing the rowers’ skin as they row, resulting in open sores. Further, there is a view

18 that lifejackets can impede the action of rowing. As a consequence, when conditions allow, some rowers prefer to row naked. That Toby Wallace’s crew did not routinely wear lifejackets when rowing was not unusual among ocean rowers. As the tethers on board could be used effectively only if they were attached to the lifejackets, they too remained unused. Conditions at the time of the accident were sufficiently poor that the crew on deck had elected to wear oilskins. That the circumstances had not also prompted them to wear their lifejackets and connect their tethers indicates they had not considered there was a risk of falling or being washed overboard. 2.4.4 Personal locator beacon Unlike Atlantic Challenge rowing, where each rower is required to wear a PLB when on deck, Toby Wallace’s crew were not required to do so, nor were there sufficient PLB’s on board for each member of the crew. In sea water at a temperature of 22ºC, Mike Johnson could have been expected to survive between 2 and 40 hours18 following his immersion shortly after 2300 on 14 February. Given that the SAR aircraft were on scene the following day from 1100 and Sea Pearl from 1230, had Mike been wearing a PLB, it would potentially have enabled MRCC Ponta Delgada to locate and track him in the water and to direct the aircraft and Sea Pearl towards his position. 2.4.5 Man overboard That two of Toby Wallace’s rowers had fallen overboard a few days after the boat had set off from Puerto De Mogan highlights that the risk of falling overboard from such a low-sided boat with a narrow beam was significant. On those occasions, the rowers managed to hold onto the boat and were recovered without injury. In the rough sea conditions experienced when Mike Johnson was swept overboard, he was unable to hold onto Toby Wallace. Toby Wallace would not steer into strong winds. For reasons of practicality, its crew also did not wear lifejackets which prevented the use of the tethers that were carried on board. Therefore, the successful recovery of Mike Johnson depended on him remaining connected to the boat and the boat being stopped in the water. It is of concern that neither were achieved due to the detachment of Mike’s ankle leash and the parachute anchor. Although Toby Wallace had conducted a manoverboard drill in Puerto De Mogan, the safety equipment carried on board and the lack of instructions and procedures for its use, indicate that the risk of a rower falling overboard had not been properly considered. The increased likelihood of a person falling over the side due to the boat’s low gunnels, the difficulties likely to be encountered in recovering a person from the water due to its manoeuvring characteristics, and the foreseeable flaws in the reliability of the ankle leashes were all significant factors that warranted a far more robust assessment of the safety equipment required and how and when it was to be used. 18 United States Search and Rescue Task Force – Cold water survival.

19 2.5 PREPARATION AND TRAINING Toby Wallace’s skipper and first mate were experienced ocean rowers, but the remainder of the crew were not. Therefore, it was unlikely that the short time some of the crew rowed together over the weekend in August 2015, and all of the crew rowed together shortly before leaving Gran Canaria, was sufficient to adequately prepare them for a 31-day ocean crossing. The man overboard recovery in the relatively sheltered waters of Puerto De Mogan would not have prepared them to cope with a man overboard at night in seas with 6m waves. Furthermore, in view of the nature of the intended voyage, it is astonishing that sea survival training was not a requirement for all crew and that the crew did not have to meet a medical standard. 2.6 EMERGENCY RESPONSE The capsize and abandonment of Fire Ant on 13 February 2016 and the loss of a rower overboard and abandonment of Toby Wallace 2 days later highlight the vulnerability of ocean rowing boats. In both cases, Oceanus Rowing Ltd’s duty officer in the UK was pivotal to the safety of the boats’ crews. By alerting the coastguard to the difficulties encountered by Fire Ant and then quickly confirming the position of Toby Wallace, and providing contact details for the boat’s skipper, the duty officer ensured that SAR action was initiated promptly. However, the accidents involving Toby Wallace and Fire Ant also demonstrate that small boats that cross the oceans independently rely heavily on passing merchant vessels and distant SAR services for assistance in the event of an emergency. Unlike participants in organised events, such as the Atlantic Challenge, which have a support vessel, independent boats frequently have little or no assistance at hand. It took Sea Pearl about 12 hours to reach Toby Wallace, and it took Rio Grita 16 hours to reach Fire Ant. As prompt assistance is unlikely to be available in mid ocean, the need to ensure that all ocean rowing boats and their crews are appropriately prepared and equipped is compelling. 2.7 GOVERNANCE 2.7.1 Commercial operation It is clear from the fees paid by the ‘pay for place’ rowers, the skipper’s remuneration, and that Toby Wallace was listed as one of Oceanus Rowing Ltd’s company assets that the vessel was being commercially operated. However, Toby Wallace could not comply with the requirements of MGN 280, and there were no alternative standards set by the national governing body that could be applied instead. It is arguable that, without appropriate certification, Toby Wallace should not have operated commercially, but there were no avenues by which commercial certification could be achieved. Nonetheless, the registration of Toby Wallace as a pleasure vessel in the skipper’s name was inappropriate as it gave a misleading impression of the vessel’s status, and of the standards of practice that should have been applied to its activities. Although ocean rowing is an extreme and potentially dangerous sport, commercial operators have a duty of care towards rowers who are paying to participate. Relevant considerations to be taken into account when discharging this duty of care could include: the suitability of the vessel; provision of individual and collective safety

20 equipment; and that the rowers themselves are trained, mentally prepared and medically fit to undertake the crossing. The significant variance in the approaches adopted by Atlantic Campaigns SL (see section 1.9) and by Oceanus Rowing Ltd indicate a need to develop a set of codes or procedures for ocean rowing to guide those with a duty of care towards fee paying customers. 2.7.2 Best practice Only one ocean rower is recorded as having died during an ocean crossing between 1999 and 2015 (paragraph 1.13.1), despite the high number of failed attempts. This is likely to be due to improved vessel design, particularly with regard to the ability of the craft to self-right following capsize. There is also much evidence that some operators are attempting to improve safety by the fitting of safety features such as guardrails (Figure 7), guard wires (Figure 8), and by the use of safety harnesses and tethers (Figure 9). However, other operators are concerned that measures taken to reduce the risk of a person being lost overboard could increase the risk of entrapment following capsize. These disparate approaches can result in a single type of hazard receiving prominence at the expense of measures to mitigate other risks. There is therefore a need for the ocean rowing community to coalesce for the purposes of sharing best practice and developing safety standards for their sport. If deemed appropriate, these safety standards could be taken forward by the national governing body (see section 1.12) and proposed to the marine regulators for use as the basis for commercial certification. Figure 7: An ocean rowing boat fitted with guardrails Image courtesy of www.alanlau.co.uk

21 Figure 8: An ocean rowing boat fitted with guard wires Figure 9: Use of a safety harness, tethers and jackstay Jackstay Safety tether Safety harness

22 SECTION 3 - CONCLUSIONS 3.1 SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

The rowers on board Toby Wallace were not suitably tethered. The ankle leashes were intended to keep rowers close to the boat in the event of capsize and the foot straps were a rowing aid. [2.2]

Toby Wallace’s gunnels were only 300mm above the deck and provided virtually no protection to the crew from the risk of falling overboard. [2.3]

Toby Wallace’s design increased the difficulty of recovering a man overboard. The boat could not be turned into the strong wind or stopped and, as a result, its crew were unable to even attempt a recovery. [2.3]

The ankle leashes used on board Toby Wallace were not manufactured or tested to meet any national or international safety standards, and were therefore not sufficiently reliable for use as safety equipment. [2.4.2]

Although lifejackets carried on board Toby Wallace were not used by the crew, this was common practice among ocean rowers. [2.4.3]

Only one personal locator beacon was carried on board. [2.4.4]

The risk of a person falling overboard from Toby Wallace had not been properly assessed. Consequently, appropriate mitigation measures had not been put in place. [2.4.5]

The crew’s preparation and training were insufficient to deal with a 31-day ocean passage and the recovery of a person overboard at night and in rough seas. [2.5]

Toby Wallace was operating as a commercial vessel but was not certified to do so. The need to address the governance of commercially operated ocean rowing boats is compelling. [2.7.1]

Ocean rowing has developed best practice that has improved safety, but inconsistencies in its application indicate that it could be better promulgated. [2.7.2] 3.2 OTHER SAFETY ISSUES NOT DIRECTLY CONTRIBUTING TO THE ACCIDENT19

The accidents involving Toby Wallace and Fire Ant demonstrate the dependency of independent ocean rowing boats on passing merchant vessels and distant SAR services for assistance. [2.6] 19 These safety issues identify lessons to be learned. They do not merit a safety recommendation based on this investigation alone. However, they may be used for analysing trends in marine accidents or in support of a future safety recommendation

23 SECTION 4 - ACTIONS TAKEN 4.1 ACTIONS TAKEN BY OTHER ORGANISATIONS Atlantic Campaigns SL has: ● ●Forwarded a proposed ‘Ocean Rowing Code of Practice’ to British Rowing for review, adoption and promulgation.

24 SECTION 5 - RECOMMENDATIONS British Rowing and the Maritime and Coastguard Agency are recommended to: 2017/101 Work together in order to assess the feasibility of developing means by which commercially operated ocean rowing boats can demonstrate equivalent safety standards to those required of other small vessels in commercial use for sport or pleasure. British Rowing is recommended to: 2017/102 Liaise with stakeholders to develop and promulgate a best practice guide or a code of practice for ocean rowing, taking into account, inter alia: ● ●Boat design, construction and stability ● ●Minimum training requirements ● ●Minimum equipment requirements ● ●Onboard procedures ● ●Shore-based and seaborne support. Oceanus Rowing Limited is recommended to: 2017/103 Review its risk assessments for the conduct of future ocean crossings and take measures as necessary to ensure its crews are appropriately fit, trained and competent, and the necessary equipment, training and procedures are in place to reduce the risk of rowers coming to harm to as low as reasonably practicable. Safety recommendations shall in no case create a presumption of blame or liability

Marine Accident Report

FD06051 ｜ 2017_英国_safety-warning-after-explosion-on-container-vessel-manhattan-bridge-results-in-1-fatality.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2017 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Extracts from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 Regulation 5: “The sole objective of a safety investigation into an accident under these Regulations shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of such an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” Regulation 16(1): “The Chief Inspector may at any time make recommendations as to how future accidents may be prevented.” NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2017 See http://www.nationalarchives. gov.uk/doc/open-government- licence for details. All bulletins can be found on our website: https://www.gov.uk/maib For all enquiries: Email: maib@dft.gsi.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 Press Enquiries: 01932 440015 Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000 M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H SAFETY BULLETIN SB1/2017 June 2017 1 Auxiliary boiler explosion on board the container ship Manhattan Bridge at Felixstowe container terminal, England resulting in one fatality and one serious injury on 19 January 2017 Figure 1: Manhattan Bridge Image courtesy of Ron van de Velde

2 MAIB SAFETY BULLETIN 1/2017 This document, containing safety lessons, has been produced for marine safety purposes only, on the basis of information available to date. The Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 provide for the Chief Inspector of Marine Accidents to make recommendations at any time during the course of an investigation if, in his opinion, it is necessary or desirable to do so. The Marine Accident Investigation Branch is assisting the Japan Transport Safety Board (JTSB) in carrying out an investigation into the auxiliary boiler explosion on board Manhattan Bridge, resulting in one fatality and one serious injury. The JTSB will publish a full report on completion of the investigation. Steve Clinch Chief Inspector of Marine Accidents NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall not be admissible in any judicial proceedings whose purpose, or one of whose purposes, is to apportion liability or blame. This bulletin is also available on our website: www.gov.uk/maib Press Enquiries: 01932 440015; Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000 Steve Clinch's signature

3 BACKGROUND At about 2304 on 19 January 2017, an auxiliary boiler furnace explosion occurred in the engine room on board the Japan registered container ship Manhattan Bridge (Figure 1) as it was berthing alongside a container terminal in Felixstowe, England. Manhattan Bridge’s second engineer and an engine room oiler were investigating a boiler flame failure alarm at the time and were caught by the blast. The oiler suffered severe injuries and died soon after the explosion. The second engineer suffered burn injuries to his face and right arm, which required a skin graft. The Japan Transport Safety Board (JTSB) is conducting a full investigation into the causes and circumstances of the accident and, in accordance with the IMO Casualty Investigation Code, will publish its findings in due course. The UK Marine Accident Investigation Branch (MAIB) deployed inspectors to Felixstowe to conduct an initial accident site investigation. Its findings have prompted the MAIB to publish this safety bulletin, which is designed to raise awareness of a safety issue that might be linked to the initial boiler flame failures. EVENTS LEADING UP TO THE ACCIDENT Manhattan Bridge had been operating in the North Sea Sulphur Emission Control Area (SECA) for several days prior to the accident. In order to comply with international emissions control standards1, the auxiliary boiler fuel supply had been switched from heavy fuel oil (HFO) to marine gas oil (MGO). The MGO was loaded at Rotterdam in November 2016 and was declared as meeting the quality standards set out in ISO 8217:2005 – Petroleum products – Fuels (class F) - Specifications of marine fuels. In the hours leading up to the accident, the auxiliary boiler had cut out several times due to flame or ignition failures, and on each occasion, the fault was investigated and the boiler reset by the second engineer. The boiler explosion occurred while the second engineer and the oiler were trying to restart the boiler burner unit following a flame failure cut-out. The force of the explosion blew open the boiler burner unit door (Figure 2) and propelled the burner’s air diffuser into the engine room (Figure 3). The oiler was standing directly in front of the burner unit and the second engineer was close by. INITIAL FINDINGS Following the incident, examination of the boiler fuel system by the burner unit manufacturer identified the build-up of waxy deposits in the supply filter, sufficient to restrict the fuel flow (Figure 4). Samples of the MGO being burnt at the time of the accident were taken by the MAIB and sent to a laboratory for analysis. The samples were tested in accordance with specifications set out in the latest ISO 8217:2017 standard, which included Cloud Point (CP), Cold Filter Plugging Point (CFPP) and Pour Point (PP) tests. The tests found that the fuel had a CFPP of 14°C and a PP of less than -9°C, requiring a minimum fuel operating temperature of 15°C. The ambient air temperature at Felixstowe on 19 January 2017 was about 4°C, low enough for wax formation. The CP of the fuel could not be obtained because the test samples had a dark appearance, which was attributed to the mixing of residual HFO deposits with the MGO in the system pipework. 1 On 1 January 2015, the sulphur emissions limits within the designated maritime SECAs were reduced from 1.0% to 0.1% by mass, which has resulted in an increased demand for MGO bunkers across the shipping sector.

4 Figure 2: Boiler and burner unit Figure 3 Burner unit Boiler fuel supply pipework Failed burner hinge Air diffuser

5 SAFETY ISSUE Industry reports indicate an increased incidence of boiler and marine diesel engine performance problems in colder waters following the implementation of the more stringent sulphur emissions limit. This has been attributed to the increased paraffin content found in some low sulphur distillate fuels (MGOs) and the subsequent formation of waxy deposits or crystals as the fuel temperature falls. Restricted fuel flow due to wax deposits in filters and pipework can cause intermittent and incomplete combustion to the point of flame failure. The paraffin content of MGOs varies globally due to the regional composition of crude oil and variation in refinery processes. There are three measurable stages in the waxing process for distillate fuels; these are CP, CFPP and PP. The first discernible stage, CP, is defined as the temperature at which wax crystals start to visibly form in the fuel and a transparent fuel becomes cloudy or hazy. The CFPP is the lowest temperature where the fuel of a set volume, drawn, by vacuum, through a standardised filter (45 micron) within a specified time (60 sec) still continues to flow. The PP is the lowest temperature at which the fuel will continue to flow when cooled. The PP does not provide any indication of the temperature at which filtration issues may occur. Prior to March 2017, the ISO 8217 standard, often used by the shipping industry as the baseline specification when ordering and testing fuel oil bunkers, focused on PP and did not include test specifications for CP or CFPP. As a consequence, the MGO loaded on board Manhattan Bridge in Rotterdam was not subject to CP or CFPP testing. Figure 4 Waxy deposits in fuel line filter

6 SAFETY LESSONS It is essential that vessel operators carefully consider anticipated ambient air and sea temperatures that will be experienced during the voyage when purchasing low sulphur MGO bunkers. Such information should be used to identify the required cold flow characteristics of the fuel being supplied using CP and CFPP as key metrics. When this is impractical, it is important to establish the CP and CFPP of the fuels carried on board through sample testing. When operating in cold climates, the risk of waxy residue developing in the vessel’s fuel lines can be controlled by: ● ●Closely monitoring the visual appearance of low sulphur MGO bunkers for signs of wax precipitation. ● ●Conducting regular fuel filter inspections and close monitoring of fuel system pressures. ● ●Maintaining the temperature of the low sulphur MGO in the vessel’s tanks and pipework above the CP and CFPP temperatures to avoid the possibility of filter blocking. The addition of cold-flow improver chemicals to the low sulphur MGO in the vessel’s storage tanks should only be considered as a last resort under the strict guidance of an additive supplier. More detailed guidance2 on the cold flow properties of MGOs can be found on the International Council on Combustion Engines (CIMAC) website: www.cimac.com. Issued June 2017 2 01/2015 CIMAC Guideline: Cold flow properties of marine fuel oils. 06/2015 CIMAC Position Paper: New 0.1% sulphur marine (ECA) fuels.

FD06052 ｜ 2017_英国_sinking-of-vivier-creel-boat-louisa-with-loss-of-3-lives__MAIBInvReport17_2017.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2017 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H ACCIDENT REPORT VERY SERIOUS MARINE CASUALTY REPORT NO 17/2017 JULY 2017 Report on the investigation of the foundering of the fishing vessel Louisa SY30 while at anchor off the Isle of Mingulay in the Outer Hebrides on 9 April 2016 resulting in three fatalities

Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2017 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch Spring Place 105 Commercial Road Southampton Email: maib@dft.gsi.gov.uk United Kingdom Telephone: +44 (0) 23 8039 5500 SO15 1GH Fax: +44 (0) 23 8023 2459 Press enquiries during office hours: 01932 440015 Press enquiries out of hours: 020 7944 4292

CONTENTS GLOSSARY OF ABBREVIATIONS AND ACRONYMS SYNOPSIS 1 SECTION 1 - FACTUAL INFORMATION 2 1.1 Particulars of Louisa and accident 2 1.2 Background 3 1.3 Narrative 3 1.3.1 The accident 3 1.3.2 Emergency response 4 1.4 Crew 8 1.5 Vessel and operation 9 1.5.1 Vessel description 9 1.5.2 Vivier system 10 1.5.3 Engine room 11 1.5.4 Bilge system 11 1.5.5 Hold and fore peak 12 1.5.6 Accommodation 12 1.6 Wreck recovery 12 1.7 Vessel examination 13 1.7.1 Vessel status 13 1.7.2 Hull and structure integrity 14 1.7.3 Pipe systems 14 1.7.4 Refloating 16 1.8 Modelling of foundering scenarios 16 1.8.1 General 16 1.8.2 Engine room flooding 16 1.8.3 Hold flooding 17 1.8.4 Hold flooding in combination with water on deck 17 1.9 Lifesaving appliances 17 1.9.1 Lifesaving appliances carried on board Louisa 17 1.9.2 Liferaft 17 1.9.3 Post-accident inspections of the HRU and liferaft 18 1.9.4 Testing and examination of the CO2 inflation cylinder 18 1.9.5 Lifejackets – regulatory requirements 19 1.9.6 Approval process for novel lifesaving appliances 22 1.9.7 Approval of the Cosalt Premier (32kg or more) solid-filled lifejacket 22 1.9.8 MAIB lifejacket trials and subsequent testing 24 1.9.9 Emergency position indicating radio beacon 25 1.10 Search and rescue 25 1.10.1 EPIRB description 25 1.10.2 Cospas-Sarsat system description 27 1.10.3 Terminology 28 1.10.4 UK EPIRB response 28 1.10.5 HMCG operational manning 29 1.11 Effects of cold water immersion 30 1.12 Urgency of response 31 1.13 Previous/similar accidents 31

SECTION 2 - ANALYSIS 35 2.1 Aim 35 2.2 Overview 35 2.3 Vessel status 35 2.4 Fatigue 36 2.5 Vessel foundering 36 2.6 Abandoning the vessel 37 2.7 Cold water survival 38 2.8 Lifesaving appliances 38 2.8.1 The liferaft - onboard management 38 2.8.2 The liferaft - servicing 39 2.8.3 Lifejackets 39 2.8.4 Inflatable PFDs 40 2.8.5 LSA maintenance 40 2.9 Search and rescue 41 2.9.1 Summary of key times and events 41 2.9.2 UKMCC/HMCG interface 41 2.9.3 Speed of response 42 2.9.4 Response to Louisa’s EPIRB alert 43 2.9.5 Incident location 43 2.9.6 Network activity 44 SECTION 3 - CONCLUSIONS 45 3.1 Safety issues directly contributing to the accident that have been addressed or resulted in recommendations 45 3.2 Other safety issues 46 SECTION 4 - ACTION TAKEN 47 4.1 MAIB actions 47 4.2 Actions taken by other organisations 47 SECTION 5 - RECOMMENDATIONS 48 FIGURES Figure 1

Louisa – general arrangement Figure 2

Accident location and positional information Figure 3

Louisa – working deck Figure 4

Louisa – transporting to River Clyde Figure 5

Lousia – engine controls as found Figure 6

System drawing – as found Figure 7

Liferaft at Barra police station Figure 8

CO2 cylinder as found

Figure 9

Recalled CO2 cylinder of a similar type to that of Louisa’s cylinder Figure 10

Lifejackets at Barra police station Figure 11

MAIB lifejacket trials Figure 12

Louisa – EPIRB Figure 13

Cospas-Sarsat system diagram Figure 14

Hold/fore peak drain TABLES Table 1

Loss condition estimated loading Table 2

Summary of key times and events ANNEXES Annex A

Materials Technology Ltd – CO2 cylinder report

Annex B

Marchwood Scientific Services – water analysis report Annex C

Extract from A.689(17) and MSC.81(70) – Recommended lifejacket water performance tests Annex D

Extracts from MSC.200(80) – Amended recommended lifejacket water performance tests Annex E

Extract from MSC.378(93) – Further amendments to recommended lifejacket water performance tests Annex F

Extract from MSC.378(93) – Amended design and construction requirements for an RTD Annex G

HMCG’s standard operating procedure for satellite distress beacon alerts Annex H

MAIB Flyer to the Fishing Industry Annex I

Cosalt Premier (32kg or more) lifejacket – testing and type approval certificates

GLOSSARY OF ABBREVIATIONS AND ACRONYMS AIS

Automatic Identification System ALB

All Weather Lifeboat ARCC

Aeronautical Rescue Co-ordination Centre BOSIET

Basic Offshore Safety Induction and Emergency Training (course) CERS

Consolidated European Reporting System CGOC

Coastguard Operations Centre cm

centimetre CO2

Carbon Dioxide DLA

Deputy Launch Authority EC

European Community EFF

European Fisheries Fund EMSA

European Maritime Safety Agency EPIRB

Emergency Position Indicating Radio Beacon GEOLUT

Geostationary Local User Terminal GEOSAR

Geostationary Search and Rescue (satellite) GNSS

Global Navigation Satellite System HMCG

HM Coastguard hr

hour HRU

Hydrostatic Release Unit IAMSAR

International Aeronautical and Maritime Search and Rescue (Manual) IHAC

(Department for Transport) In House Analytical Consultancy (team) IMDatE

Integrated Maritime Data Environment IMO

International Maritime Organization kg

kilogram kt

knot kVA

kilo-volt-ampere kW

kilowatt

LEOSAR

Low Earth Orbiting Search and Rescue (satellite) LOA

length overall LSA

Lifesaving appliances LUT

Local User Terminal m

metre MCA

Maritime and Coastguard Agency MCC

Mission Control Centre MGN

Marine Guidance Note MHz

Megahertz mm

millimetre MOO

Maritime Operations Officer MRCC

Maritime Rescue Co-ordination Centre MSN

Merchant Shipping Notice N

Newton NMOC

National Maritime Operations Centre nm

nautical mile ORC

Offshore Racing Council PDF

Protected Data Field PFD

Personal Flotation Device RAF

Royal Air Force RNLI

Royal National Lifeboat Institution rpm

revolutions per minute RTD

Reference Test Device s

second SAR

Search and Rescue SFF

Scottish Fishermen’s Federation Sitrep

Situation report SMOO

Senior Maritime Operations Officer

SOLAS

International Convention for the Safety of Life at Sea 1974, as amended t

tonne UKMCC

United Kingdom Mission Control Centre UTC

Universal Co-ordinated Time VHF

Very High Frequency VMS

Vessel Monitoring System TIMES: all times used in this report are UTC unless otherwise stated

1 SYNOPSIS Early on 9 April 2016, the fishing vessel Louisa foundered, with the loss of three lives, while anchored close to the shore in Mingulay Bay in the Outer Hebrides. The skipper and crew, who had been working long hours before anchoring late the previous evening, had woken suddenly as the vessel was sinking rapidly by the bow. They were able to escape to the aft deck, activate the emergency position indicating radio beacon (EPIRB), and to don lifejackets. However, they were unable to inflate the liferaft as they abandoned the vessel. At 0232, an alert from Louisa’s EPIRB was detected by a geostationary search and rescue satellite and forwarded through the Cospas-Sarsat system to the United Kingdom Mission Control Centre (UKMCC). The UKMCC relayed the alert to HM Coastguard, but confusion over terminology resulted in delays before search and rescue units were sent to the scene. When Barra lifeboat arrived in Mingulay Bay, the crew were able to assist one crewman, who had swum to the shore and climbed onto rocks. They located the uninflated liferaft and beside it found the skipper and one crewman unresponsive and face down in the water, despite wearing approved abandonment lifejackets. The other crewman had attempted to swim to the shore and was found, also face down and unresponsive, still wearing his lifejacket, close to the beach. The skipper’s body was lost as the lifeboat crew attempted to recover him and remains missing. The MAIB investigation included salvaging the wreck to determine the cause of flooding, inspection and testing of the liferaft, lifejacket trials and testing, and a review of the search and rescue response. The Maritime and Coastguard Agency has since taken action to enhance its guidance in respect of liferaft servicing requirements. The circumstances of this accident, and subsequent trials and testing undertaken, have raised concerns about the effectiveness of the lifejackets worn by Louisa’s skipper and crew. Recommendations have been made to Louisa’s owners regarding vessel maintenance, safety equipment servicing and risk assessments, and to the liferaft servicing company and its sub-contractor in respect of work processes. The Maritime and Coastguard Agency has been recommended to: urgently conduct research designed to confirm or otherwise the suitability of lifejacket water performance test requirements, and to bring any shortcomings identified to the attention of the International Maritime Organization; and to update and enhance its response to satellite distress beacon alerts.

2 SECTION 1 - FACTUAL INFORMATION 1.1 PARTICULARS OF LOUISA AND ACCIDENT SHIP PARTICULARS Vessel’s name Louisa Flag United Kingdom Classification society Not applicable IMO number/fishing numbers SY30 Type Vivier creel boat Registered owners Privately owned Manager(s) Not applicable Construction Steel Year of build 2008 Length overall 14.95m Registered length 14.15m Gross tonnage 32 Minimum safe manning Not applicable Authorised cargo Not applicable VOYAGE PARTICULARS Port of departure Uig, Isle of Skye Port of arrival Uig, Isle of Skye (intended) Type of voyage Commercial Cargo information Shellfish Manning 4 MARINE CASUALTY INFORMATION Date and time 09/04/2016 0232 UTC Type of marine casualty or incident Very Serious Marine Casualty Location of incident Isle of Mingulay Place on board Hold Injuries/fatalities 3 fatalities Damage/environmental impact Constructive total loss/none Ship operation At anchor Voyage segment Anchored External & internal environment Night, visibility good, wind light airs variable direction, sea state slight. Sea temperature 10ºC, air temperature 8ºC. Persons on board 4

3 1.2 BACKGROUND Louisa generally operated around the Western Isles fishing for lobster, crab and prawns. Each trip was usually 6 days’ duration, departing on a Monday afternoon and landing the catch on the following Sunday, during which time the catch was kept alive in a vivier tank. The intensity of fishing varied depending upon quantity of catch, weather conditions and the skipper’s preference. The relief skipper normally aimed to recover and shoot each string of creels twice during the trip, whereas the regular skipper aimed to recover and shoot each string three times per trip. The intention on this trip was for each string to be recovered and relaid three times, which could amount to hauling about 1,000 creels per day. Catch was normally landed at the port of Uig on the Isle of Skye, with a weekly call into Stockinsh on the Isle of Lewis to allow the owners to carry out any maintenance required. The crew generally completed 3 weeks on board, followed by a week of leave, but the regular skipper could spend several months on the vessel without a break. There was a pool of seven crew (including the relief skipper) with either the skipper plus three, or occasionally four, crew on board during each fishing trip. 1.3 NARRATIVE 1.3.1 The accident On Sunday 3 April 2016, Louisa berthed in Uig to discharge catch and conduct a crew change during which the relief skipper left and Paul Alliston, the vessel’s regular skipper, took over. Another crew member left at short notice, and sailing was delayed while a relief crewman travelled from the mainland to join the vessel. Photograph courtesy of Trawlerphotos.co.uk Louisa

4 Louisa sailed late evening on Monday 4 April, and motored for about 7 hours to reach the first string of creels. Fishing then continued through Tuesday, Wednesday and Thursday, with the live catch being loaded into the vivier tank. The vessel berthed at North Bay on the Isle of Barra to load an additional 20 boxes of fish offal as bait. Having loaded the bait, the vessel returned to its fishing grounds. Some of the boxes of bait were stored in the port bait locker, a refrigerated compartment in the hold. The remainder of the boxes were stowed in various locations on the main deck with a number placed directly aft of the hold hatch coaming. Fishing continued into Friday evening when Louisa’s skipper recognised that the crew, who had been working for approximately 20 hours per day, were too tired to continue safely. He decided to anchor in Mingulay Bay (a recognised anchorage off the uninhabited Isle of Mingulay), to enable the crew and himself to rest. At approximately 2230, he anchored Louisa about 200m from the shore. He and the crew then set about tidying the working deck and hosing down using the deck wash pump, while a meal was prepared. When the meal was ready the skipper and crew went to the mess room to eat it, after which they all went to bed in the accommodation. The crew woke suddenly in the early hours of Saturday morning, 9 April, to find the vessel significantly down by the head and apparently foundering bow first. Wearing only light clothing they quickly exited the accommodation via the wheelhouse and onto the aft deck, by which time the vessel’s bow was fully submerged and water was encroaching on the wheelhouse windows (Figure 1). The skipper told the crew to launch the liferaft and activate the EPIRB while he collected lifejackets from a locker located in the wheelhouse. One of the crew released the liferaft canister from its cradle and unsuccessfully attempted to inflate the liferaft by pulling on the painter. The canister was then removed from the cradle and its securing straps cut to remove the uninflated liferaft. Further attempts were then made to operate the liferaft’s CO2 inflation cylinder, all of which failed. By this time, the crew were standing on the aft bulkhead of the wheelhouse with the vessel almost vertical and foundering rapidly. The crew then worked together to place some plastic buoys inside the liferaft to give it buoyancy, and then pushed it into the water. By now, the skipper and crew had all donned lifejackets. Two of the crew jumped onto the liferaft, which initially supported their weight but, as the others jumped on, it began to sink. Martin Johnstone immediately decided to attempt to swim to the shore. The skipper and remaining crew, conscious of their sea survival training, decided to remain with the partially floating liferaft. After several minutes it became apparent that the liferaft was drifting further out to sea. After discussing their options, one of the remaining crew also decided to attempt the swim to shore while the skipper and Christopher Morrison chose to remain with the liferaft. Louisa foundered and sat on the seabed in approximately 10m of water. 1.3.2 Emergency response At 0232 UTC, an alert from Louisa’s EPIRB was detected by geostationary search and rescue (GEOSAR) satellite MSG3, which automatically forwarded it to the French geostationary local user terminal (GEOLUT). The information then

5 Figure 1: Louisa - general arrangement Bulkhead Fore peak Hold hatch

6 automatically passed to the United Kingdom Mission Control Centre (UKMCC). The information included a position for the beacon of 56º 45.00’N, 007º 30.00’W (Figure 2). By 0239, UKMCC had initiated a new incident on the HM Coastguard (HMCG) live information system (Vision 4) and had informed Coastguard Operations Centre (CGOC) Falmouth via fax and telephone call that the alert had been received. The alert was initially logged as an unresolved 406 alert. At 0244 UKMCC entered the position for the EPIRB into Vision 4. The EPIRB’s details were then entered and, by 0250, UKMCC had passed co-ordination of the incident response to CGOC Falmouth. At 0251, CGOC Falmouth made CGOC Stornoway aware of the alert and requested it attempt to contact Louisa by all available means. Despite numerous calls to the registered emergency numbers associated with the EPIRB and very high frequency (VHF) radio calls to the vessel, CGOC Stornoway was unable to establish contact with Louisa or its owners. At 0241, an alert from Louisa’s EPIRB was detected by GEOSAR satellite MSG2 and passed to the Greek GEOLUT. This information, which was again automatically passed to UKMCC, included an updated position for the beacon of 56º 48.73’N, 007º 37.40’W. It is reported that UKMCC received this information at 0251. However, the refined position did not appear on Vision 4 until 0312 when it became visible to CGOC Falmouth. At 0306, the National Maritime Operations Centre (NMOC) noted on Vision 4 that it was aware of the EPIRB alert and was monitoring the narrative. During a telephone conversation between CGOC Falmouth and UKMCC, UKMCC staff expressed uncertainty as to the accuracy of the updated position and that a low earth orbiting search and rescue (LEOSAR) satellite pass would be able to confirm the position. With the next LEOSAR satellite pass not scheduled until 0410, CGOC Falmouth staff pressed for confirmation that the EPIRB alert included a Global Navigation Satellite System (GNSS) position. After further discussion, UKMCC staff confirmed that the updated position of Louisa’s EPIRB was accurate to within 120m. At 0316, CGOC Falmouth passed responsibility for co-ordinating the Search and Rescue (SAR) response to the incident to CGOC Stornoway, and confirmed that the updated position for Louisa’s EPIRB was accurate to 120m. At 0322, CGOC Stornoway staff paged Barra all weather lifeboat (ALB) deputy launch authority (DLA), who telephoned them to obtain further information about the incident. He then authorised an immediate launch of Barra ALB and requested that CGOC Stornoway page the lifeboat’s crew. The lifeboat crew were paged at 0333 and the ALB launched at 0340. At 0327, CGOC Stornoway staff contacted the Aeronautical Rescue Co-ordination Centre (ARCC) and requested the launch of a SAR helicopter. ARCC staff requested Stornoway CGOC page the crew for aircraft R948. At 0344, CGOC Stornoway recorded in Vision 4 that it had paged the aircraft crew and that a fault with the communication equipment had caused a delay in activating the pagers.

7 Figure 2: Accident location and positional information Reproduced from Admiralty Chart BA 1796-0 by permission of the Controller of HMSO and the UK Hydrographic Office

8 At 0356, CGOC Humber staff contacted CGOC Stornoway to advise that a position taken from the European Maritime Safety Agency (EMSA) IMdatE1 system placed Louisa in Mingulay Bay at 0140 that morning. Barra ALB proceeded to Mingulay Bay and arrived on scene at 0413. The crew immediately set off white parachute flares to illuminate the area. There was no sign of Louisa, which by then had foundered in 10m of water. The ALB’s crew saw and headed towards a white flashing light, which was attached to Louisa’s uninflated liferaft. At 0424, near to the liferaft, the ALB crew found the skipper and Chistopher Morrison, both unresponsive and face down in the water. While recovering the casualties, the skipper’s body slipped from his lifejacket and was lost. Christopher was recovered from the water but was later declared deceased caused by drowning. After a further 30 minutes of searching, the lifeboat’s crew saw another light to the north of the bay. This light was attached to the lifejacket of a surviving crewman who had swum ashore and was clinging to the rocks. The ALB despatched its tender and landed a crew member on the rocks to provide warm clothing to the survivor. The ALB then continued searching and used its tender to recover Martin Johnstone, who was unresponsive and face down in the water, approximately 50m from the shore. Martin was later declared deceased caused by drowning At 0425, R948 reported that it would be on scene at 0510. By 0539, R948 had recovered the surviving crewman from the rocks. The survivor confirmed that there had been four persons on Louisa, all of whom were now accounted for. The skipper’s body remains missing. 1.4 CREW The skipper, Paul Alliston, was 42 years old. He had been associated with Louisa since build, initially as crew, and had taken over as skipper approximately 6 years before the accident. He held a Seafish2 Under 16.5m Skipper Certificate (unrestricted), which included both bridge and engine room watchkeeping certificates and Intermediate Stability Awareness. He had also completed the mandatory basic safety training courses. Martin Johnstone was 29 years old. As well as working as a fisherman he had served in the UK military. Of the mandatory basic safety training courses, he had completed Basic Fire-fighting and Prevention, Basic First-aid, and Safety Awareness and Risk Assessment courses, but not the required Basic Sea Survival course. He had undertaken the Basic Offshore Safety Induction and Emergency Training (BOSIET) course when looking for work in the offshore industry, but this was not recognised as an equivalent to the Basic Sea Survival course. Christopher Morrison was 27 years old. An experienced fisherman and a regular member of Louisa’s crew, he had completed all the mandatory basic safety training courses. 1 The Integrated Maritime Data Environment (IMDatE) is a technical framework that collects and combines data from EMSA’s maritime applications and other external sources to provide a comprehensive and configurable service to users. It supports the relay of data from Vessel Monitoring Systems (VMS) and Satellite AIS between maritime applications. 2 Seafish is an executive non-departmental public body, sponsored by the Department for the Environment, Food and Rural Affairs.

9 The fourth and only surviving crew member was 27 years old. He was an experienced crab fisherman and regular member of Louisa’s crew. He had completed the mandatory basic safety training courses and was the only member of the crew to have completed the 1-day Basic Health and Safety course.3 Louisa’s skipper and crew were share fishermen, which meant that they were not employed under a contract of service and, instead, received a share of the vessel’s profits as remuneration. 1.5 VESSEL AND OPERATION 1.5.1 Vessel description Louisa was built for static gear creel fishing and was designed with a vivier system to keep the catch alive (Figure 3). As required under Merchant Shipping Notice (MSN) 1813 (F), The Fishing Vessels Code of Practice for the Safety of Small Fishing Vessels, Louisa was built in accordance with the Seafish Construction Standards for new fishing vessels less than 15m length overall. Prior to registration, the vessel was subject to a Maritime and Coastguard Agency (MCA) inspection of safety equipment and crew qualifications. This inspection was required to be repeated every 5 years, with the vessel’s owners completing annual self-certification of compliance in the intervening years. As recommended by MSN 1813 (F), a freeboard and stability information booklet was produced for Louisa by the vessel’s designer following a lightship inclining experiment. The booklet contained a section entitled ‘Working Instruction’ in which the following was stated: ‘The accumulation of free water in the hold must be avoided as this will cause loss of stability due to its transverse movement. Non-watertight pound divisions will not effectively prevent this. Check bilge levels frequently and pump out as necessary, the functioning of the hold bilge alarm should also be verified at the start of each trip.’ [sic] Louisa’s normal operating pattern was broadly consistent with the voyage cycle assumption in the freeboard and stability information booklet, which assumed a ½ day voyage between departure port and fishing grounds, 5 days on the fishing grounds and a further ½ day return voyage from the fishing grounds to arrival port. 3 Following the discovery that many experienced fishermen who joined the UK fishing industry after 1 January 2005 had not completed the mandatory Basic Health and Safety course, the MCA decided that: 

‘Fishermen who joined a UK fishing vessel for the first time after 1 January 2005 and before to 1 June 2014 and did not undertake the Basic Health and Safety course do not have to complete this course provided that they;

a) have completed the Safety Awareness and Risk Assessment course and

b) can demonstrate the date of joining a UK vessel prior to 2014 to the satisfaction of an MCA surveyor.

Fishermen who joined a UK fishing vessel for the first time after 1 June 2014 and did not undertake the UK Basic Health and Safety course must complete this course within a time specified by an MCA surveyor. This applies regardless of whether the fisherman holds the Safety Awareness and Risk Assessment Course.’

10 The booklet required that the vivier tank remained full with the water level pressed up the access trunk to the overflows. It stated: ‘If the water level dropped, a large free surface would develop and the vessel’s stability would be dangerously reduced.’ 1.5.2 Vivier system The purpose of a vivier system is to keep the vessel’s catch alive once removed from its natural environment, as the quality of shellfish degrades rapidly after death. To achieve this, circulation pumps maintain a steady flow of clean, aerated sea water throughout the system. On Louisa, the vivier system consisted of a tank with a capacity of 22t located in the centre of the vessel. The tank boundaries comprised the forward bulkhead of the engine room, the aft bulkhead of the forward hold and the vessel’s sides up to the level of the bait locker decks. A trunking led from the top of the vivier tank to finish flush with the main deck. Within the trunking were port and starboard overflow pipes. In the bottom of the vivier tank was a grid of perforated pipes through which the sea water was pumped by one of two dedicated pumps. Both pumps were non-self- priming and had a capacity of 125t/hr at 1,500rpm. The port pump was powered by an electric motor and the starboard by a hydraulic motor. The electric pump was rarely used as it generated excessive noise and vibration. The starboard pump was driven via a power take off from either the main engine (when at sea) or the Figure 3: Louisa - working deck Catch chute Hold hatch cover Vivier tank hatch Pound boards Discard chute

11 starboard auxiliary engine (when in port or at anchor). The deck wash pump could also be configured to feed the vivier tank, at a reduced throughput of approximately 35t/hr. 1.5.3 Engine room The engine room was located aft of amidships and forward of the accommodation. It was accessed via a door from the accommodation. The main engine produced 230kW at 2,000rpm and turned the vessel’s propeller through a gearbox, which incorporated a slipping clutch. This arrangement could be used in conjunction with the throttle to control ahead and astern propulsion. At the forward end of the engine, driven from the crankshaft, were power take offs to provide hydraulic power for the steering gear and the vivier system’s main pump. The engine room contained two auxiliary engines to provide electrical power for the vessel. The starboard auxiliary provided 45kVA at 1,500rpm and had a power take off to drive the hydraulic vivier system pump when the main engine was shut down. A smaller air-cooled engine had been retrofitted to Louisa by the owners to provide a level of redundancy. However, this was out of commission at the time of the accident and was planned to be replaced in May 2016. Also in the engine room was a general service pump (known on board as the deck wash pump), which could be configured to pump the bilges, supply the deck wash hose or supply the vivier system. The pump was powered by an electric motor and had a capacity of 35t/hr at 2,900rpm. 1.5.4 Bilge system As required by MSN 1813 (F), Louisa was equipped with a bilge pump and a bilge level alarm, with sensors (float switches) in the hold and engine room to provide warning to the crew when working inside or outside the wheelhouse. The general service pump could pump water from the engine room and/or the hold via fixed pipelines running to dedicated bilge wells in each of these compartments. In addition to the general service pump, the owners had fitted stand-alone electric bilge pumps in both the engine room and hold bilge wells. The bilge alarm float switches in the hold and engine room bilge wells were connected to an audible/visual alarm panel in the wheelhouse. Following a previous pipe failure, which had resulted in the engine room flooding, the owners had fitted a remote alarm sounder in the crew cabin to alert the skipper and crew that the bilge alarm had activated in the event the wheelhouse was unmanned. During the investigation, the alarm sounder in the crew cabin was found to have been disabled. It was reported that small quantities of melt water from the refrigerated bait storage lockers routinely collected in the hold bilge well, which caused the hold bilge alarm to sound when the vessel rolled. The alarm sounder in the crew cabin had regularly disturbed the skipper and crew, and consequently they had disabled it.

12 1.5.5 Hold and fore peak The hold was located forward of the vivier tank. Incorporated within the hold were two refrigerated bait storage lockers: one on the port side and one on the starboard side outboard of the vivier tank trunk. The lockers could each hold approximately 2t of bait and were accessed via insulated, but not watertight, doors in the hold. The hold was used for general storage of ropes and other spares and also included a small domestic chest freezer. The fore peak was located forward of the hold and included the bulbous bow. Although separate compartments, the fore peak and hold were interconnected by way of a small (30mm diameter) drainage hole located in the bilge. 1.5.6 Accommodation The accommodation comprised a galley/mess room and a single berth skipper’s cabin at main deck level, and a four-berth cabin for the crew on the deck below the galley and aft of the engine room. The designated skipper’s cabin on the main deck was only used for storage; the skipper and crew all slept in the lower four-berth cabin. The vessel’s hydraulic steering gear was located in the four-berth cabin, which also contained a fire detector/alarm and a bilge alarm sounder (that had been disabled, see 1.5.4). 1.6 WRECK RECOVERY To ascertain the cause of Louisa foundering, the wreck was salvaged and placed ashore in a secure facility to allow a thorough examination. A pre-salvage dive survey found no evidence of damage or breaches to the hull; that all visible doors and hatches were open; and that those freeing ports that could be seen were clear. The divers observed through the wheelhouse windows that the main engine controls were set with the propeller engaged and set to run astern at apparently low speed. After large amounts of loose rope had been cleared from around the vessel, lifting slings were placed beneath the vessel’s hull and it was lifted onto a crane barge. The wreck was then secured to the barge and transported to a facility on the River Clyde, where it was prepared for inspection (Figure 4). Extensive cleaning of the vessel was required to remove the remnants of the catch (approximately 4t of dead shellfish) and other detritus to make it safe for access.

13 1.7 VESSEL EXAMINATION 1.7.1 Vessel status The initial focus of the examination centred on checking and recording the positions of all system valves and machinery controls. As identified during the dive survey, the main engine controls were found set to run at apparently low speed with the propeller engaged astern (Figure 5). The vivier system valves and power supply were configured to operate the system using the main hydraulic pump driven from the main engine power take off (Figure 6). The status of the deck wash pump could not be ascertained because the switching arrangement provided no means for confirming whether or not the pump was being electrically powered at the time of the accident. However, the lock stop had not been activated, indicating that the pump was probably running. All machinery space and accommodation doors were found open. The hold hatch cover was found closed but not secured, and some lengths of rope were between the hatch cover and its coaming. It was reported that, prior to the accident, the hatch cover was open and secured against pound boards located forward of the hatch coaming, and that a number of bait boxes were stacked aft of the hatch coaming. It was not possible to check this information as the pound boards had become detached during the foundering and the bait boxes had moved. Figure 4: Louisa - transporting to River Clyde

14 1.7.2 Hull and structure integrity There was no evidence of external damage to Louisa’s hull and structure. To test its integrity, the vivier system’s inlet and outlet valves were closed and the tank was filled to the level of the overflow pipes in the trunking. Tank boundaries and pipelines were examined for signs of leakage. No leakage was observed and the water level remained constant. The vivier tank was then emptied and the hold filled. Hold boundaries (which included the engine room) and pipe runs were examined, but again no leakage was observed. It was noted at this stage that the fore peak filled at the same time as the hold, and this subsequently led to the discovery of the drainage hole between the two compartments. The vivier overflow pipes and the discard chute ran through sections of the hold. The overflow pipes and discard chute were blanked off at their hull outlets and filled with water, but again no leakage into the hold was observed. 1.7.3 Pipe systems To test Louisa’s piping, connection boxes were welded over the sea water inlets on the vessel’s hull and the systems were tested under pressure using a portable pump. The systems tested included: ● ●Bilge system (engine room and hold) ● ●Engine and generator cooling system Figure 5: Louisa - engine controls as found Ahead Astern Throttle/gearbox control Winch control (unused) Slipping clutch

15 ● ●Refrigeration condenser ● ●Vivier system ● ●Deck wash line ● ●Redundant prawn wash line. All systems were confirmed as being intact, with no leakage noted. Sea suction Discharge Bilge suction Hold bilge Prawn wash line shut shut shut open open open open shut shut shut Sea suction open shut open Sea suction shut open Deck wash pump Hydraulic Viver Pump Engine Room Vivier tank Hold N/R ER Bilge Deck wash line (open) Seized Vivier Discharge Manifold N/R shut Electric Vivier pump shut Deck wash pump suction manifold Overboard discharge Overboard discharge Valve Pump Figure 6: System drawing (valves as found)

16 1.7.4 Refloating Once all testing had been completed, Louisa was craned back into the water and the vivier tank refilled to replicate as far as possible the vessel’s condition at the time of the accident. The vessel was left floating for approximately 4 hours, remaining in a stable condition with no ingress of water, before being lifted ashore once more. 1.8 MODELLING OF FOUNDERING SCENARIOS 1.8.1 General A Wolfson HST computer model was used to explore the likely flooding scenarios that could have led to Louisa’s foundering. The model included the hull, tanks and watertight bulkheads. The fore peak was assumed to be part of the hold as they were connected via a 30mm diameter pipe. The forward section of sheltered deck space was also modelled to allow flooding on deck to be simulated. Hatch openings and freeing ports were included as warning points in the model to establish when they became submerged. The loading of the vessel at the time of loss was estimated from information derived from interviews and from the vessel’s freeboard and stability information booklet. Table 1 shows the loading condition assumed at the time of loss. Item Quantity (t) Notes Fuel tanks (port and starboard) 2.7 in each 85% full, not cross connected Daily service fuel tank 0.227 Full Vivier tank 22.87 Full (any catch would displace water so no net effect) Fresh water tank 1.68 80% full Bait and boxes 2.0 As per freeboard and stability booklet Crew (4) 0.3 Stores 0.5 As per freeboard and stability booklet Table 1: Loss condition estimated loading 1.8.2 Engine room flooding Given the most likely sources of internal flooding were within Louisa’s engine room, flooding of that space was modelled first. With the engine room free to flood, the model remained afloat with a small stability margin and with a stern trim.

17 1.8.3 Hold flooding With 35t of flood water in the hold, which equated to 60% of the compartment’s volume, the model became heavily trimmed by the bow with negligible stability remaining. There was still positive freeboard forward and no downflooding points were submerged. 1.8.4 Hold flooding in combination with water on deck The effect of water on deck was explored. However, for water to collect on the main deck, the vessel required a significant bow trim (of 0.5m between the draught marks) to prevent the forward-most freeing ports from shedding water overboard. Louisa normally operated with a stern trim. It was estimated that to achieve such a bow trim required 15t of water in the hold. A further 5t of water, equating to a depth of 0.5m, was then required on the main deck for the model’s stability to become marginal. Consideration was given to whether the propeller driving slowly astern while Louisa was at anchor would induce sufficient bow trim to allow water to collect on the main deck. Approximate calculations indicated a bow trim of 0.07m might result from the propeller driving astern with the main engine idling. Considering that a trim change of over 1m was required to cause water to collect forward, the probability of this effect was considered to be negligible. 1.9 LIFESAVING APPLIANCES 1.9.1 Lifesaving appliances carried on board Louisa Louisa was equipped with the following lifesaving appliances: ● ●1 Oceanic 8, Lifeguard marine liferaft ● ●5 Cosalt Premier (32kg or more) solid-filled lifejackets ● ●1 McMurdo Smartfind 406MHz EPIRB ● ●5 Mullion Compact 150N personal flotation devices (PFDs). The Mullion PFDs had been issued free of charge by the Scottish Fishermen’s Federation (SFF) under the Personal Flotation Device (PFD) initiative supported by the European Fisheries Fund (EFF). The PFDs were designed for use by fishermen in the working environment and aimed to combat the ongoing loss of life through crew falling overboard during normal fishing operations. The PFDs supplied to the crew of Louisa had remained in their original packaging, were stored in the skipper’s cabin, and were out of date for their required annual servicing. 1.9.2 Liferaft MSN 1813 (F) required Louisa’s owners to ensure that safety equipment carried on board the vessel was suitably maintained and serviced in accordance with manufacturers’ instructions. Louisa was required to carry a liferaft, and the Code

18 recommended it should either be float free or fitted with a Hydrostatic Release Unit (HRU). Marine Guidance Note (MGN) 499 (M+F)4, required the liferaft to be serviced at least annually at a service station accredited by the manufacturer. Louisa’s Oceanic 8, Lifeguard marine liferaft had been manufactured in 1990 to Offshore Racing Council (ORC) standards5. The liferaft was located on the wheelhouse roof and secured in a cradle, incorporating an HRU. The HRU was designed to release the liferaft from its cradle if the unit was submerged to a depth of between 2m and 4m. Louisa’s owners had hired the liferaft from Premium Liferaft Services on a 3-year contract that had commenced on 23 April 2014. The hire contract allowed for the liferaft to be replaced annually to meet the servicing requirements. The last recorded service of the liferaft was on 7 March 2014, just prior to the beginning of the contract. Furthermore, at the time of the accident, Premium Liferaft Services’ accreditation from the manufacturer to service this brand of liferaft had lapsed despite the company’s efforts to secure continuation training for its technicians. 1.9.3 Post-accident inspections of the HRU and liferaft During examination of the recovered vessel, MAIB inspectors noted that the HRU had an expiry date of March 2015, and it had not been rigged correctly. The liferaft’s painter, which should have been secured via a weak link (designed to part once the liferaft had inflated) had been tied directly to the cradle. In the event, these shortfalls were immaterial because the crew had deployed the liferaft manually before the vessel foundered; however, it had failed to inflate. Following the accident, the uninflated liferaft was recovered to Barra police station, where MAIB investigators conducted an initial inspection (Figure 7). Subsequent inspections of the liferaft were carried out by an independent liferaft servicing company and by the current manufacturing representatives of Lifeguard products. The findings of these examinations were that: ● ●The liferaft inflated successfully when subjected to a gas inflation test. ● ●The liferaft’s condition indicated that it was suitable for use and that its condition was commensurate with its age. 1.9.4 Testing and examination of the CO2 inflation cylinder Inspection of the CO2 inflation cylinder noted that it was devoid of gas but contained approximately 1.5 litres of water. Examination of the CO2 cylinder at an independent test facility (Annex A), including dye penetrant testing, revealed surface cracks in the valve attachment threads at the neck of the cylinder. Laboratory analysis of the water (Annex B) highlighted a high saline content indicating that this was likely to have been seawater. 4 MGN 499 (M+F) Life Saving Appliances: Inflatable Liferafts, Marine Evacuation Systems, Inflatable Lifejackets and Hydrostatic Release Units – Servicing Requirements, has now been replaced by MGN 553 (M+F) for non- SOLAS inflatable LSA. 5 ORC liferafts were introduced, following the 1979 Fastnet Race, by the Offshore Racing Council (ORC) for yachts racing under its rules. The Sydney Hobart Race in 1998 brought the durability of these liferafts into question and led to a revision of the standard. ORC liferafts are being phased out by the MCA in accordance with MGN 553 (M+F).

19 Examination of the valve assembly showed that the operating mechanism had been correctly activated. Further examination indicated that the valve assembly had not been tightened into the cylinder to the manufacturer’s recommended torque setting. In November 2011, Premium Liferaft Services had sub-contracted refurbishment of the cylinder to Thameside Fire Protection Company Limited. The cylinder was one of a batch of nine cylinders that had been sent for 5-yearly refurbishment and refilling. The process included cleaning, inspection, pressure testing, painting and refilling of the cylinders. Prior to filling, the shoulder of the cylinder was date-stamped to indicate that pressure testing had taken place. A label was adhered to the cylinder showing its tare weight6, gas charge (CO2 and N2) and full weight, which was 8.40kg (Figure 8). When the cylinder was weighed following the accident, its tare weight was 8.40kg. It was therefore concluded that, following refurbishment in 2011, the cylinder had not been refilled prior to its return to Premium Liferaft Services. Following this discovery, MAIB inspectors arranged for the remaining eight cylinders from the batch of nine to be recalled and weighed. All were found to have been correctly filled and labelled. One of the recalled cylinders of a similar type to that of Louisa’s cylinder is shown in Figure 9. 1.9.5 Lifejackets – regulatory requirements MSN 1813 (F) required small fishing vessels to carry an abandonment lifejacket for each person on board. The lifejackets were required either to be of the solid-filled type, or should have had automatic gas inflation and at least 150N buoyancy. SOLAS requires that the in-water performance of a lifejacket be evaluated in compliance with the recommendations of the International Maritime Organization (IMO). The IMO’s recommendations on testing lifesaving appliances are set out in 6 Tare weight = the weight of the cylinder plus valve assembly (some companies also include label weight) measured before adding the gas charge. The gas charge is a mixture of carbon dioxide (CO2) and Nitrogen (N2) which is added by weight. Figure 7: Liferaft at Barra police station CO2 cylinder Activation head

20 resolution A.689(17) and, for lifesaving appliances installed on board on or after 1 July 1999, in resolution MSC.81(70) (Revised recommendation on testing of life-saving appliances). Recommendations for lifejacket water performance tests, set out in both A.689(17) and MSC.81(70), are reproduced in Annex C. Figure 8: CO2 cylinder as found TARE WEIGHT 5.550 kg CARBON DIOXIDE 2.780 kg NITROGEN 0.070 kg FULL WEIGHT 8.400 kg Figure 9: Recalled CO2 cylinder of a similar type to that of Louisa’s cylinder TARE WEIGHT 8.360 kg CARBON DIOXIDE 2.780 kg NITROGEN 0.070 kg FULL WEIGHT 11.210 kg

21 The water performance tests were intended to determine the lifejacket’s ability to assist a helpless person or one in an exhausted or unconscious state, and to show that the lifejacket did not unduly restrict movement. The test regime required able-bodied test subjects who were all good swimmers and collectively met prescribed selection criteria in terms of gender, height and weight. Recognizing a need to introduce more precise provisions, the IMO adopted amendments to the revised recommendation on testing of lifesaving appliances on 13 May 2005 in resolution MSC.200(80). The amended recommendations for lifejacket water performance tests, set out in MSC.200(80), are reproduced in Annex D. Further amendments were introduced by resolution MSC.378(93) on 22 May 2014, reproduced in Annex E. MSC.200(80) introduced a requirement for the in-water performance of a lifejacket to be evaluated by comparing its performance with that of a Reference Test Device (RTD)7. The design and construction requirements for an RTD were amended by resolutions MSC.226(82) and MSC.378(93), reproduced in Annex F. The test standards are not absolute and incorporate a risk-based approach, particularly in respect of the water performance righting tests. In both A.689(17) and MSC.200(80), donning tests require that test subjects carry out initial tests wearing normal clothing and then a further test wearing heavy-weather clothing. However, water performance tests are carried out in fresh water under still conditions, with the test subjects wearing only swimming costumes. Resolution A.689(17) section 2.9.8 states: ‘When evaluating the results of a test in accordance with 2.9.5, 2.9.7 and 2.9.8, the Administration may, in exceptional circumstances, disregard the results of a test on a subject if the results show a very slight deviation from the specified criteria, provided the Administration is satisfied that the deviation can be attributed to the unusual size and stature characteristics of the test subject and the results of tests on other subjects, chosen in accordance with 2.9.2, show satisfactory performance of the lifejacket.’ MSC.200(80) defines the requirements for test subjects (Annex D Table 2.1) and requires a minimum of 12 test subjects. Each subject is required to complete the righting test six times using the lifejacket and six times using the RTD. The period of time taken for the test subject’s mouth to come clear of the water is measured to the nearest 1/10th second, and the highest and lowest times recorded against each device are discarded. MSC.200(80) goes on to state the following relating to the water performance test results: ‘2.8.7 After the water tests described in 2.8.5 and.6 above: .1 Turning time: The average turn time for all subjects in the candidate lifejacket should not exceed the average time in the RTD, and the number of ‘no-turns’, if any, should not exceed the number in the RTD; .2 Freeboard: The average freeboard of all the subjects should not be less than the average for the RTD; 7 An RTD is a lifejacket manufactured to a specification as detailed in Annex F.

22 .3 Torso angles: The average of all subjects’ torso angles should not be less that the average for the RTD minus 5º; .4 Faceplane (head) angles: The average of all subjects’ faceplane angles should be not less than the average for the RTD minus 5º; .5 Lifejacket light location: The position of the lifejacket light should permit it to be visible over as great a segment of the upper hemisphere as is practicable.’ MSC.378(93) was developed following an IMO sub-committee investigation into inconsistencies in the performance of prototype lifejackets against the RTD. The revised criteria relaxed the performance criteria against the RTD, allowing a lifejacket to turn in 1s longer than the average time of the RTD and a reduction in the required freeboard height, and faceplane and torso angles. 1.9.6 Approval process for novel lifesaving appliances Before giving approval to a novel lifesaving appliance or arrangement, SOLAS requires an Administration to ensure that such appliances provide safety standards at least equivalent to the requirements of SOLAS and the Life Saving Appliance (LSA) Code and that they have been evaluated and tested based on the guidelines developed by the IMO in resolution A.520(13). The following is an extract: ‘3.1.2 Individual buoyancy equipment should: .8 in calm fresh water, be capable of lifting the mouth of a completely relaxed person wearing normal clothing at least 120 mm clear of the water; .9 in calm fresh water, be capable of turning a completely relaxed person wearing normal clothing from any position in the water to one where the mouth is clear of the water within 5 s…’ 1.9.7 Approval of the Cosalt Premier (32kg or more) solid-filled lifejacket Louisa was equipped with Cosalt Premier (32kg or more) solid-filled lifejackets, which the skipper and crew all donned before abandoning the vessel (Figure 10). These lifejackets were marked ‘LLOYDS REGISTER APPROVED, AUTHORISED BY MCA’8 and were of a type commonly used on passenger ships, other commercial vessels and leisure craft. The lifejacket type was tested against A.689(17) by the Department of Transport’s Newcastle Marine Office in April 1992 (Annex I). The water performance tests included a requirement for a minimum of six test subjects to complete three gentle swimming strokes, then to simulate a state of utter exhaustion with their heads down and lungs partially filled. The time from the last stroke until the mouth became clear was then noted. The time was not to exceed 5s. The test was then repeated after the subject had exhaled and, again, the time was not to exceed 5s. 8 Lloyd’s Register was authorised by the MCA to certify and approve lifejackets against IMO requirements. The MCA was not involved in testing or authorising individual products.

23 During the first test, four of the eight test subjects exceeded 5s (albeit by up to only 0.5s) and, after exhalation, two of the eight test subjects exceeded 5s (by less than 1s). These tests were then repeated by the Newcastle Marine Office in May 1992 (Annex I) using the same eight test subjects, at which time all the test subjects’ airways were clear of the water within 5s during both tests. There is no indication in either of these test reports of any modifications or improvements having been carried out to the lifejacket. Following its initial approval, the lifejacket type underwent a number of modifications relating to material specifications, each of which resulted in partial retesting of the lifejackets. The version of the lifejacket type on board Louisa was issued a ‘Certificate of Type Approval’ (included at Annex I) by Lloyd’s Register on 6 May 2004 valid until 5 May

1. The certificate and the accompanying ‘EC Type Examination (Module B)

Certificate’ did not reference any further water performance tests and referred only to material and buoyancy tests carried out by AMTEC Laboratory and Fleetwood Testing Laboratory between 1998 and 1999 (Annex I). The lifejacket type ceased production in 2010 and was replaced by the Cosalt Premier 2010, which was designed to meet the requirements of MSC.200(80). Figure 10: Lifejackets at Barra police station

24 1.9.8 MAIB lifejacket trials and subsequent testing In view of the RNLI reports relating to the discovery and recovery of the casualties, the MAIB undertook trials of one of the recovered Cosalt Premier (32kg or more) lifejackets and a similar and widely available solid-filled approved lifejacket, which had been produced by another manufacturer in 2014 and tested against the requirements of MSC.200(80). Also tested was a PFD of the same make and design as those available but unused on Louisa. The trials used only two subjects and only loosely resembled the test protocols contained in MSC.200(80). However, they closely matched the requirements of A.689(17). Both test subjects were approximately 176cm tall, one weighing 66kg and the other 104kg. Each subject trialled both of the solid-filled lifejackets. During the righting tests neither lifejacket was able to turn either test subject, leaving them floating face down in the water until they were forced to abort the test to take breath (in excess of 20s) (Figure 11). The inflatable PFD righted a face down test subject within 2s on each occasion. Following liaison between the MAIB and the MCA, the Cosalt Premier (32kg or more) solid-filled lifejacket used in the MAIB trials was transferred to the MCA for further testing. These tests were carried out by an approved testing organisation and included the testing of an additional Cosalt Premier (32kg or more) solid-filled lifejacket. The test protocols used were in accordance with MSC.81 (70). Both lifejackets failed to meet the required test standard. The MCA also commissioned Figure 11: MAIB lifejacket trials

25 the testing of a similar, widely available solid-filled approved lifejacket against the requirements of MSC.200 (80). This lifejacket also failed to meet the required test standard. 1.9.9 Emergency position indicating radio beacon MSN 1813(F) recommends that small fishing vessels are equipped with an EPIRB. Louisa’s McMurdo Smartfind 406MHz EPIRB, which was fitted in a float-free housing on the wheelhouse roof, was designed to provide both alerting and location functions. Louisa’s EPIRB was manually activated by the crew and functioned correctly despite the battery having an expiry date of March 2016 and, therefore, being overdue for replacement (Figure 12). 1.10 SEARCH AND RESCUE 1.10.1 EPIRB description An EPIRB is a distress alerting device that, once activated, transmits a 406MHz digitally encoded radio message at a power of 5 Watts for a ‘burst’ duration of approximately 0.5s. Transmissions are repeated once every 50s until the beacon’s battery is exhausted. EPIRB transmissions contain information relating to the country in which the beacon is registered and the identification of the vessel in distress. EPIRBs can be an alert only type or GNSS encoded to provide a location function. Positional information for alert only (non GNSS) EPIRBs is obtained through Doppler processing9 by the local user terminals (LUTs) and requires a minimum of 3 data points from the EPIRB transmissions. GNSS enabled EPIRBs also transmit their position. The standard location protocol for a GNSS enabled EPIRB refers to coarse and refined position accuracy. For UK SAR purposes, a coarse position is assumed to have a maximum error of 15nm assuming that the GNSS position is accurate. A refined position is accurate to within 120m. The EPIRB’s transmission is arranged in two protected data fields: the first protected data field (PDF-1) contains the beacon’s identification information and its ‘coarse’ position; the second protected data field (PDF-2) contains a correction to the beacon’s coarse position to produce a ‘refined’ position. PDF-1 contains the maximum level of protection from signal interference and is received in the initial transmission. 9 Prior to the introduction of GNSS enabled beacons, the Cospas-Sarsat system made location of the beacons possible by Doppler processing. LUTs detecting non-GEOSAR satellites interpret the Doppler frequency shift heard by LEOSAR satellites as they pass over a beacon transmitting at a fixed frequency. The interpretation determines both bearing and range. The bearing and range are measured from the rate of change of the heard frequency, which varies both according to the path of the satellite in space and the rotation of the Earth. This triangulates the position of the beacon. A faster change in the Doppler indicates that the beacon is closer to the satellites orbit. If the beacon is moving toward or away from the satellite track due to the Earth’s rotation, it is on one side or other of the satellite’s path. Doppler shift is zero at the closest point of approach between the beacon and the orbit.

26 Battery expiry date MAR 2016 Figure 12: Louisa EPIRB

27 1.10.2 Cospas-Sarsat system description The Cospas-Sarsat system consists of LEOSAR satellites at an altitude of 850km and GEOSAR satellites at between 19,000 and 24,000km altitude. Both types of satellite detect transmissions from EPIRBs, Emergency Locator Beacons, and Personal Locator Beacons, and transmit data received from these to LUTs. LUTs process the data to identify and locate a beacon and its corresponding mission control centre (MCC), to whom the data is forwarded. The MCC then provides the distress alert information to the appropriate SAR service (Figure 13). The Cospas-Sarsat system was initially designed to work with LEOSAR satellites which provide both independent location of beacons by Doppler processing and GNSS reception of encoded beacons. Subsequently, the alerting function was incorporated into GEOSAR satellites to improve system coverage both for the Figure 13: Cospas-Sarsat system diagram

28 alerting function for all beacons and the location capability for GNSS encoded beacons. Expansion of the system to include GEOSAR satellites has improved system coverage for encoded location and enhanced initial alert timeliness. 1.10.3 Terminology In addition to coarse and refined positions, the following terminology relevant to this accident is used by operators dealing with Cospas-Sarsat beacon detections: The GEOLUT passes the alert to the beacon’s host MCC as soon as it has a confirmed message (valid message with complete PDF-1 field). It will also transmit unconfirmed messages if confirmation has not been achieved within 5 minutes. The GEOLUT will also pass updated information to the MCC as the encoded position in the beacon’s transmitted alert message changes (refined position). An EPIRB alert is categorised in terms of its type and location. The alert type is undetermined until it is confirmed as being real or false. Where a satellite is unable to detect the EPIRB’s location, the alert is detect only. This can be a non GNSS EPIRB detected by a GEOSAR or by a LEOSAR with insufficient data points. An unresolved location means that a LEOSAR satellite has detected the EPIRB’s location using Doppler processing, producing two possible positions: one true and one false. A resolved location means that a second satellite pass has validated one of the possible positions as being the EPIRB’s true position. 1.10.4 UK EPIRB response In the UK, the response to a UK EPIRB was managed by two separate organisations: UKMCC and HMCG. UKMCC was manned by RAF personnel and based in Scotland at RAF Kinloss. Distress alerts are initially received by UKMCC through the Cospas-Sarsat system. Alerts from EPIRBs (and all other alerts over water) are then passed to HMCG, who co-ordinate the SAR response. HMCG’s point of contact for EPIRB alerts has traditionally been CGOC Falmouth, but this is now shared with the National Maritime Operations Centre (NMOC). HMCG’s standard operating procedure for satellite distress beacon alerts is reproduced in Annex G. On receipt of a beacon alert, UKMCC first checks that it is a new alert and not additional data relating to a previous or ongoing beacon activation. When a new EPIRB alert is received, UKMCC passes the information directly to HMCG by telephone and fax, and by initiating a new incident on the Vision 4 system. UKMCC and HMCG check the EPIRB registration on the Consolidated European Reporting System (CERS) database for information relating to the vessel’s name, its operation and emergency contact details. Both HMCG and UKMCC can update the live feed on the Vision 4 system.

29 Once the registration details have been recovered, HMCG attempts to contact the vessel and/or its owners through the emergency contact details listed in the CERS database. Attempts are also made to confirm the vessel’s normal area of operation. Once the EPIRB’s location has been confirmed, co-ordination of the incident is passed to the relevant CGOC. Louisa’s EPIRB initial alert type was categorised by UKMCC as undetermined and its location as unresolved. Subsequently, UKMCC referred to the EPIRB’s updated location as a refined position. The position descriptors of coarse and refined, associated with GNSS enabled EPIRBs, do not feature in HMCG’s standard operating procedure for satellite distress beacon alerts. 1.10.5 HMCG operational manning HMCG has a network of 10 area CGOCs and one NMOC. The NMOC is both an operation centre in its own right and the co-ordinator of operational activities across the network. Although the CGOCs have the ability to function autonomously, the system is designed to operate as a network with each CGOC receiving support from within the network to meet its operational needs. A duty controller monitors activity levels across the network and matches the available resource to meet operational demands. The role of the duty controller expands to monitor the progress of ongoing operations and to assist, or to direct other operators to assist, as the situation develops. The duty controller is nominally based at NMOC, but could operate from any of the CGOCs. Whenever practicable, SAR activity is co-ordinated by the CGOC within whose geographical area the incident has occurred. During the CG response to the foundering of Louisa there were 25 officers manning the network. Four officers (including a trainee controller) were based at the NMOC. The CGOCs had a maximum of three officers on watch, though CGOC Aberdeen and CGOC London, each had only one officer on watch. The network had been the subject of a study by the Department for Transport’s In House Analytical Consultancy (IHAC) team, which determined that the minimum manning across the network should not drop below a total of 21 suitably qualified personnel. The network had only one duty controller covering the shift, who was based in CGOC Belfast and was on a break at the time of the alert and did not rejoin the network until 0427. His role was being covered by a trainee at NMOC who had been employed by HMCG for only 3 months. At the start of the night shift on 8 April, the duty controller had asked the senior maritime operations officer (SMOO) at CGOC Stornoway to provide SMOO cover for nine additional areas. There was a total of 10 SMOOs on duty across the network, but none located in the NMOC. CGOC Falmouth had two officers available. The SMOO was away from her post taking a scheduled meal break during the incident. CGOC Stornoway had two officers on watch, a SMOO and a trainee Maritime Operations Officer (MOO). Once CGOC Falmouth had received a refined position for Louisa’s EPIRB, co-ordination of the incident was passed to CGOC Stornoway. There, the SMOO and the trainee MOO attempted to deal with a wide range of tasks. These included requesting SAR support (ALB and aircraft), briefing the RNLI and ARCC, paging and tasking the ALB crew, and paging the aircraft crew. Additional workload was caused

30 by a number of system-related deficiencies. These included a fault with the UKMCC fax machine, which had been defective for a week, and an inability to automatically transfer EPIRB data into Vision 4, a condition that had been known for 18 months and which consequently required manual transfer of data that took several minutes. These deficiencies were compounded by faults with the Vision 4 system, which resulted in terminals freezing or shutting down. Additional communications problems included an inability of the ARCC to access the VHF radio paging system for helicopter crews, and known communications blackspots in the Barra area. 1.11 EFFECTS OF COLD WATER IMMERSION The human body’s typical reaction to immersion in cold water (under 15°C) is normally considered in four stages:

Cold water shock Cold water shock takes place within the first 30s to 2 minutes and is generally associated with a gasp reflex as the body comes into contact with the cold water, along with hyperventilation and a dramatic increase in heart rate and blood pressure. If the head goes underwater during this stage, the inability to hold one’s breath will often lead to water entering the lungs in sufficient quantities to cause death. The increased heart rate and blood pressure can result in cardiac arrest, especially if the casualty has an existing cardiovascular condition. Panic can cause the hyperventilation to continue even after the initial physiological effects have subsided.

Cold incapacitation Cold incapacitation usually occurs within 2-15 minutes of entering the water. The blood vessels are constricted as the body tries to preserve heat and protect the vital organs. This results in the blood flow to the extremities being restricted, causing cooling and consequent deterioration in the functioning of muscles and nerve ends. Useful movement is lost in the hands and feet, progressively leading to the incapacitation of arms and legs. Unless an effective lifejacket is worn, death by drowning occurs as a result of impaired swimming.

Hypothermia Hypothermia occurs when the human body’s core temperature drops below 35°C (it is normally about 37°C). Depending on circumstances, this can occur after 30 minutes. Symptoms of moderate hypothermia, when the body’s temperature is between 28-32°C, include inattention, confusion, difficulty moving and loss of coordination. Loss of consciousness is associated with severe hypothermia, when the body’s temperature is below 28°C. The body’s core temperature can continue to drop even after the casualty has been recovered from the water if the re-warming efforts are not effective.

31

Circum-rescue collapse Circum-rescue collapse can occur just before, during or after rescue due to a variety of mechanisms that result in unconsciousness or death. Collapse just before rescue may occur when a casualty relaxes mentally resulting, among other things, in a sudden drop of stress hormones, possibly leading to drop in blood pressure. 1.12 URGENCY OF RESPONSE The International Aeronautical and Maritime Search and Rescue (IAMSAR)10 manual states that in water at 10ºC, the 50% survival time for a normally-clothed individual is estimated to be about 2 hours, with a recommended search time of 12 hours. The manual recognises that there are many factors that can influence survival times, including the use of survival suits, lifejackets and liferafts, and the adverse effects of wind-chill, breaking waves and strong currents. It also notes that an individual’s level of fitness and body fat percentage can further influence survivability. 1.13 PREVIOUS/SIMILAR ACCIDENTS Aquila - 20 July 2009 Aquila was a 13.4m long scallop dredger, which capsized east of the Isle of Muck, off the west coast of Scotland, with the loss of three crew. The MAIB report11 concluded among other things that the speed of capsize resulted in the vessel’s liferaft and EPIRB becoming trapped in the superstructure. Therefore, although they had released from their stowage cradles, neither of these important safety items were able to operate as intended. The Maritime Rescue Co-ordination Centre (MRCC) Clyde, which initially received the distress call, immediately tasked a SAR helicopter. However, when it was realised that the accident had occurred outside MRCC Clyde’s area of operation, MRCC Clyde transferred co-ordination responsibility to MRCC Stornoway, in whose area the accident had occurred, and stood the helicopter down. The decision to transfer co-ordination of the incident and stand the helicopter down caused a delay of 23 minutes in a rescue helicopter reaching the accident site. The MCA committed to make arrangements to make ARCC the single tasking authority for SAR helicopters with effect from 1 April 2010 and to commence a review of: ● ●The exchange of information between MRCCs and ARCC when aircraft are tasked to an incident. ● ●The handover of live incidents between MRCCs. ● ●The selection of appropriate SAR assets regardless of MRCC boundaries. ● ●VHF radio procedures for managing SAR assets. 10 The IAMSAR manual is in three volumes and its purpose is to assist States in meeting their own SAR needs and their international obligations under SOLAS. 11 MAIB Report No 5/2010.

32 ● ●The VHF radio coverage in the area of the accident. In view of these commitments the MAIB did not make further recommendations. Ocean Way – 2 November 2014 The 17.07m fishing vessel Ocean Way (FR 349), a trawler with a crew of five, capsized and sank 100nm off the north-east coast of England. The MAIB report12 concluded that: although accurate positional information for the vessel became available, this was not initially recognised by the SAR authority due to its cognitive bias towards information that placed the vessel further inshore; it is important that procedures are followed to avoid essential information being overlooked when staff are engaged in demanding situations; the opportunity to task a rescue helicopter at an earlier stage of the incident was missed and; the efficient conduct of SAR missions would be more effective if all UK MRCCs were equipped with computer systems capable of gaining access to essential information from all available databases. The report also observed that the availability of contemporaneous Vessel Monitoring System (VMS) information to MRCCs would prove of great value in identifying, locating and assisting fishing vessels in distress. Had the vessel’s EPIRB been fitted with an integral GNSS it is probable that a helicopter would have been tasked about 50 minutes sooner. As a result of this incident, the MCA reported it had: ● ●Included a requirement that EPIRBs be equipped with GNSS in proposed new Codes of Practice for the safety of fishing vessels. ● ●Ensured that the computer equipment in all MRCCs was capable of accessing vessels’ positional information in a timely manner. ● ●Updated staff training provision to ensure appropriate consideration was given to the use of all resources when responding to requests for assistance. ● ●Reviewed its staff training in operational procedures to ensure the correct handover of incident co-ordination is followed in the future. The MAIB recommended13 that the MCA: ● ●Take action to ensure that the EPIRBs required to be carried on UK registered fishing vessels are equipped with integral GNSS receivers. Annie T - 4 October 2015 A crewman from the 9.15m creel fishing vessel Annie T (CY1) was carried overboard by the fishing gear when his foot became caught in a bight of rope while he was moving an end weight on the working deck. The skipper recovered the crewman back on board about 10 minutes later but was unable to resuscitate him. 12 MAIB Report No 23/2015. 13 Recommendation 2015/154.

33 The MAIB report14 concluded that: The accident occurred because the crewman was in an unsafe position and became entangled in a rope while shooting creels. Neither the skipper nor the crewman wore lifejackets during fishing although three new, unused, constant wear lifejackets were found on board the vessel following the accident. Neither the skipper nor crewman appreciated the effects of cold water shock and, as a consequence, had not taken precautions to mitigate these effects. The MCA was recommended15 to: Prioritise the introduction of legislation that will require the compulsory wearing of personal flotation devices on the working decks of all fishing vessels while at sea. Apollo – 18 April 2016 At 0630 on 18 April 2016, a crewman fell overboard from the stern of the 23.95m fishing vessel Apollo (INS 179) while the nets were being hauled. The weather was rough at the time of the accident and the crewman was not wearing a personal flotation device (PFD). An extensive search was carried out by coastguard helicopters, military aircraft and Apollo’s crew. The crewman’s body was found nearly 4 months later. The MAIB report16 found that following a previous fatal man overboard in 2007 the crew had been required to wear lifejackets. However, the practice had fallen into disuse and lifejackets had not been worn for several years despite constant wear lifejackets being available on board. It also identified that fishermen generally lack awareness of the debilitating effects of cold water immersion, specifically cold shock and cold incapacitation. The report considered similar manoverboard cases that have been reported to the MAIB since 2007. In the majority of these cases the person in the water was initially responsive and able to help themselves before they rapidly succumbed to the incapacitating effects of cold water. Harvester - 28 April 2016 The potter Harvester (M999) grounded on rocks in Abereiddy Bay, North Pembrokeshire, and sank a short time later. There was no indication of any crew on board at the time of the grounding. A large-scale SAR operation commenced and the body of a crew member was recovered from the water 3nm from where Harvester had foundered. He was wearing neither a lifejacket nor other buoyancy aid. The second crew member was not found despite an extensive search. The MAIB report17 concluded that neither crew member was wearing a PFD, significantly lowering their chances of survival. 14 MAIB Report No 21/2016. 15 Recommendation 2016/146. 16 MAIB Report No 23/2016. 17 MAIB Report No 22/2016.

34 The report also noted that a personal locator beacon is a very useful additional means of raising the alarm particularly, as in this case, if no one is left on board and the only other means of raising the alarm remains on the vessel. Harvester’s automatic identification system (AIS) unit was switched off at the time of the accident. An historical track of the vessel’s movements would have been particularly valuable to those involved in the initial search and rescue operation.

35 SECTION 2 - ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 OVERVIEW Louisa sank by the bow and foundered while at anchor in Mingulay Bay. Following recovery and investigation of the wreck, it is concluded that the cause of the foundering was flooding of the hold. The skipper and crew had all gone to bed in the accommodation, and the hold bilge alarm failed to wake them because the alarm sounder in the crew cabin had previously been disabled. The events leading up to the flooding and the circumstances in which it occurred could not be conclusively determined, but water probably entered the hold from the deck wash hose. When abandoning the vessel, the crew attempted to launch the liferaft. However, the liferaft failed to inflate because its CO2 cylinder was empty, which resulted in the skipper and crew having to enter the water. Although they had all donned lifejackets, the skipper and two crew became unresponsive through cold water immersion and were later found face down in the water by the rescue services. Before abandoning Louisa, the crew activated the vessel’s EPIRB. Although UKMCC had received an initial alert by 0239, the first request for a SAR asset was not made by HMCG until 0322. Had the rescue services arrived on scene earlier, it is possible that there would have been more survivors. 2.3 VESSEL STATUS At the time of the accident, Louisa was at anchor approximately 200m from the shore in Mingulay Bay with the skipper and crew asleep in the accommodation. Post-accident examination of the vessel indicates at the time of the accident that: ● ●The main engine was running at apparently slow speed with the propeller engaged astern, and the vivier system hydraulic pump was being driven from the main engine power take off. ● ●All machinery space and accommodation doors were left open, and the hold hatch cover was secured in the open position. ● ●The deck wash pump had probably been left running (see section 2.5). While the skipper had been known, on occasion, to leave the engine running astern for short periods while moored alongside, such as when loading bait, it would be highly unusual to leave the engine driving the vessel astern when the vessel was at anchor, especially when the wheelhouse was unmanned. More normal would have been for the main engine to be shut down and the hydraulic vivier pump driven from the auxiliary engine. However, it is possible that although the main engine was left running to power the vivier pump, the propeller was only engaged astern by the

36 skipper in an attempt to overcome the rapidly increasing bow trim. Also, while not always the case, it would have been a routine task to shut down the general service (deck wash) pump before the crew took their rest. 2.4 FATIGUE Louisa’s crew had been fishing continually for 4 days, working approximately 20 hours per day. During that period they had not had any dedicated rest or meal breaks, they had mainly sustained themselves by eating snacks, and had slept when possible during the relatively short transits between each string of creels. Besides being short of sleep, the crew were physically exhausted by the work required to shoot and recover the strings of creels and to process the catch. Ultimately, the skipper had anchored the vessel because he considered both he and the crew were too tired to continue fishing safely. While the skipper’s decision to halt operations to enable the crew and himself to rest was wise, by then the crew were fatigued. The evidence indicates that they interrupted their work to eat a meal, and afterwards went straight to bed. The consequence of this was that the vessel’s machinery was not shut down / reconfigured appropriately, which created the conditions for the accident to occur. Further, because the bilge alarm sounder in the crew cabin had been disabled, the skipper and crew did not receive any early warning that the hold was flooding. Instead of managing work routines to prevent fatigue and so ensure adequate levels of safety were maintained, the skipper worked the crew and himself to a state of tiredness such that the safety of the vessel and its crew was compromised. The work ethic that resulted in this behaviour can be linked to the share remuneration scheme, which encouraged the skipper and crew to maximise the catch whenever suitable conditions provided. 2.5 VESSEL FOUNDERING There was no evidence that Louisa had suffered external damage, all hull and tank boundaries were intact, and seawater piping systems retained their integrity throughout post-accident system testing. Modelling of foundering scenarios showed that 35t of water in the hold would have heavily trimmed the vessel by the bow, with negligible stability remaining. Flooding the engine room would have caused the vessel to sit deeper in the water with a stern trim but would not have resulted in foundering. Therefore, flooding of the hold was considered to have been the probable cause of the foundering. The drainage hole between the hold and the fore peak effectively rendered these spaces as one compartment, exacerbating the effects of floodwater in the hold which, in turn, would have accelerated the foundering (Figure 14). Following extensive trials and testing, and in the absence of any other plausible explanation, it is concluded that water probably entered the hold from the deck wash hose. When the crew went for their meal, the general service (deck wash) pump was probably running, the hold hatch cover was open and a number of bait boxes were stacked aft of the hatch coaming. It is probable that the hose, which could flail around the deck if not firmly secured, became wedged between the bait boxes, and

37 this allowed water to spray against the open hatch cover and fall into the hold. As the general service (deck wash) pump had a rated capacity of 35t/hr, the water in the hold could have accumulated to a critical point, in the worst case, within an hour. 2.6 ABANDONING THE VESSEL The skipper and crew woke as the vessel acquired a significant bow down trim. Once roused, they all exited the accommodation through the wheelhouse, but on reaching the aft deck they were forced to stand on the aft bulkhead of the wheelhouse due to the attitude of the vessel. There was no evidence to suggest that abandon ship drills had been routinely practised on Louisa but the skipper’s decision to return to the wheelhouse to collect lifejackets and then instruct the crew to operate the EPIRB and prepare the liferaft, was prudent and appropriate. Although cutting the liferaft canister straps and removing the liferaft from the canister was not in accordance with the manufacturer’s operating instructions, it was an appropriate action in the circumstances and should not have prevented the liferaft from being inflated had the CO2 cylinder been charged. The necessary speed of the abandonment gave no opportunity for the skipper and crew to don warm clothes. Their survival in water of 10ºC therefore relied heavily on them being able to board the liferaft and / or a rapid rescue. Figure 14: Hold/fore peak drain Drain hole Permanent ballast

38 2.7 COLD WATER SURVIVAL Without an inflated liferaft, the effect of cold water immersion was a significant hazard to the crew as they abandoned Louisa. That the skipper and his crew were able to discuss survival options once they were in the water indicates that they had overcome the immediate effects of cold water shock. That they were wearing lifejackets when they entered the water is likely to have been a major contributory factor to their surviving the initial shock by assisting them in keeping their airways clear of the water as they got their breathing under control. Thereafter, the crew would have suffered from cold incapacitation as their hands and legs cooled and they progressively lost dexterity and co-ordination. This would have reduced their ability to conduct survival tasks, and increased their reliance on the ability of the lifejackets they were wearing to maintain their airways clear of the water. As they became increasingly hypothermic, they would eventually have lost consciousness, but could have survived had their airways remained clear of the water until they were rescued. Postmortem examination results in respect of Christopher Morrison and Martin Johnstone indicate drowning as the cause of death. It is therefore uncertain to what extent they had succumbed to the effects of hypothermia at the time of death. By the time the ALB found the first unresponsive casualties, they had been in the water for approximately 1 hour and 52 minutes. The 50% survival rate for normally clothed individuals in water of 10ºC is predicted to be 2 hours. It is possible, therefore, that more of the crew would have survived had their lifejackets been more effective in keeping their mouths clear of the water and had rescue services arrived sooner. 2.8 LIFESAVING APPLIANCES 2.8.1 The liferaft - onboard management Louisa’s liferaft was on hire to the owners from Premium Liferaft Services. The hire agreement included annual replacement to comply with the requirement under MGN 499 (M+F) for annual servicing. However, the liferaft had not been changed for 2 years and was out of date for service. It was secured to the vessel with an HRU, which was also found to be out of date. When the crew attempted to operate the liferaft, it failed to inflate owing to the CO2 cylinder being empty. Although past its expiry date, the HRU was found to have operated and to have cut the liferaft canister securing strap. The shortcomings identified above indicate that insufficient attention was being paid on board Louisa to the fitting and servicing of the vessel’s liferaft. Had the liferaft been serviced/changed annually, as required, it is very likely that its replacement would have functioned correctly when it was needed on 9 April 2016.

39 2.8.2 The liferaft - servicing Refurbishment of the CO2 cylinder fitted to Louisa’s liferaft had been sub-contracted by Premium Liferaft Services to Thameside Fire Protection Company Limited in November 2011. Post-accident examination of the cylinder found that it had not been filled post-refurbishment, which had not been identified when its filled weight was written on the cylinder; there were surface cracks in the valve attachment thread at the neck of the cylinder; and, the valve assembly had not been tightened into the neck to the manufacturer’s recommended torque setting. The investigation found that Thameside Fire Protection Company Limited did not have access to any manufacturer’s information relating to maintenance of the cylinder or valve assembly, and relied on general requirements for cylinder testing under the transportation of pressurised gas cylinders BS EN 1968:2002. They were subject to third party audit in respect of their fire protection business but not in respect of liferaft cylinder testing. Premium Liferaft Services had not carried out any audit of Thameside Fire Protection Company Limited before sub-contracting work to the company. The sub-contractor’s omission to refill Louisa’s liferaft CO2 cylinder was almost certainly an isolated incident. However, it was allowed to occur owing to deficiencies in the work process. If there had been a formal work specification detailing each step of the process, including a requirement to record the cylinder weight before and after servicing, this deficiency ought to have been detected. The work processes and recording of cylinder maintenance were not well defined, with no detailed work specification issued by Premium Liferaft Services on despatch of the cylinders. The guidance in MGN 499 (M+F) relating to liferaft servicing also lacked definition in this regard. 2.8.3 Lifejackets The circumstances of this accident and the subsequent trials and testing of the Cosalt Premier (32kg or more) solid-filled lifejackets have raised concerns about the effectiveness of the lifejackets worn by Louisa’s skipper and crew. The second set of tests carried out by the Department of Transport in May 1992 on the Cosalt Premier (32kg or more) solid-filled lifejackets met the requirements specified by the IMO (see section 1.9.7). If these tests had been conducted with the test subjects wearing normal clothing, and similar results had been achieved, it would have also met the corresponding requirements for novel individual buoyancy equipment. However, the conditions applied to the water performance tests both in 1992 and under current requirements are restricted (test subjects wearing only swimming costumes and in still, fresh water). Against the testing requirements for the Cosalt Premier (32kg or more) solid-filled lifejacket (A.689(17) and MSC.81(70)), an Administration has, in particular circumstances, discretion to disregard a very slight deviation from the specified criteria. Against the current testing requirement (MSC.200(80)), there is a possibility of a lifejacket meeting the test requirements having recorded a number of ‘no-turns’. Therefore, the IMO testing requirements acknowledge that an approved lifejacket may, under some circumstances, be incapable of satisfactorily turning a face down, unresponsive casualty, and keeping their airway clear of the water. This highlights

40 the absence of any factor of safety to take account of the lifejacket being used in operational conditions i.e. with an unspecified range of subjects wearing clothing other than swimming costumes in other than calm conditions. Prototype lifejackets that initially fail approval tests could be retested using the same or different test subjects (within the defined criteria) and subsequently pass the required tests without modification, as was the case with the Cosalt Premier (32kg or more) solid-filled lifejacket in 1992. Conversely, lifejackets that pass the required tests could subsequently fail the tests without modification, as demonstrated by the tests commissioned by the MCA following this accident (see section 1.9.8). The variability of outcome evidenced by testing of the Cosalt Premier (32kg or more) solid-filled lifejackets raises doubts about the robustness of the testing regime. Further amendments to the recommended lifejacket water performance tests, introduced on 22 May 2014 (MSC.378(93)), notably relax the turning time and freeboard requirements against amended design and construction requirements for an RTD, reducing pre-existing safety factors. Paul Alliston, Christopher Morrison and Martin Johnstone had all undertaken sea survival training during which they were taught how to don lifejackets correctly. That all three were found unresponsive and face down in the water indicates that their lifejackets were incapable of turning them onto their backs and keeping their airways clear. Assuming that they had donned and secured their lifejackets in accordance with their training and the manufacturer’s instructions, this finding supports that of the MAIB lifejacket trials and the MCA commissioned testing, and questions the suitability of both historical and extant lifejacket water performance test protocols. 2.8.4 Inflatable PFDs The MAIB lifejacket trials found that the inflatable PFD, of the same make and design as those available but unused on Louisa, righted a face down test subject within 2s on each occasion. The PFDs issued to Louisa were intended for use during normal fishing operations rather than vessel abandonment. However, they had remained in their original packaging, showed no signs of use and were out of date for servicing. This was not unexpected as, despite campaigns to encourage their use and as highlighted in the Annie T and Harvester MAIB reports, many UK fishermen have not yet accepted the rationale for routinely wearing these valuable safety devices. 2.8.5 LSA maintenance The LSA on Louisa was commensurate with carriage requirements but significant items were out of date for test or replacement. The liferaft should have been serviced annually but was overdue for service by a year, the HRU was found to have functioned satisfactorily but was overdue for replacement, and the EPIRB’s battery was also overdue for replacement. While not part of the formal LSA outfit, the PFDs were also out of date for service. Even where items of safety equipment have been hired, the responsibility to maintain and service them to ensure they remain in date and fit for purpose rests with the vessel’s owners. Notwithstanding this, Premium Liferaft Services operated a system for prompting hirers when payment was due based on invoice dates.

41 However, the invoice dates did not align with service dates. With minor adjustments and a focus on servicing requirements, the system could have provided a potentially useful prompt to Louisa’s owners that the liferaft was due for replacement under their existing contract. 2.9 SEARCH AND RESCUE 2.9.1 Summary of key times and events Time (UTC) Event 0232 Satellite detects EPIRB alert. 0239 New incident input to Vision 4. UKMCC advises CGOC of EPIRB alert. 0241 Second satellite detects EPIRB alert. 0244 UKMCC enters coarse EPIRB position into Vision 4. 0250 UKMCC passes co-ordination to CGOC Falmouth. 0251 CGOC Falmouth requests CGOC Stornoway to contact vessel. 0251 UKMCC receives refined EPIRB position. 0306 NMOC notes on Vision 4 that it is aware of the EPIRB alert and that it is monitoring narrative. 0312 UKMCC enters refined EPIRB position into Vision 4. 0316 CGOC Falmouth confirms with UKMCC that refined position is accurate to 120m and passes co-ordination to CGOC Stornoway. 0322 CGOC Stornoway pages Barra ALB DLA. 0327 CGOC Stornoway contacts ARCC and requests SAR helicopter. 0333 CGOC Stornoway pages Barra ALB crew. 0340 Barra ALB launches. 0344 CGOC Stornoway pages R948 aircraft crew. 0356 CGOC Humber advises CGOC Stornoway that IMdatE system placed Louisa in Mingulay Bay at 0140. 0413 Barra ALB arrives on scene in Mingulay Bay. 0424 Barra ALB locates two unresponsive casualties. 0539 R948 SAR helicopter reports having recovered survivor. Table 2: Summary of key times and events 2.9.2 UKMCC/HMCG interface At the outset of the emergency response, there was confusion between UKMCC and CGOC Falmouth with respect to the application of the descriptors used to define the alert. On this occasion resolved, unresolved, refined and unrefined, appeared in the coastguard log, which differed from the standard terminology used by Cospas-Sarsat and UKMCC (see 1.10.3). There was also a lack of understanding of the operation of the Cospas-Sarsat system relating to encoded and alert only beacons.

42 The coastguard operator recognised an unresolved alert but was unsure of the term unrefined which was used by the UKMCC operator to describe the location in place of the standard term coarse position. In turn, the UKMCC operator was unsure how the embedded global positioning satellite (GNSS) function of the EPIRB worked and, although aware that the coarse position had accuracy of 15nm associated with it, was unable to satisfactorily explain this to Falmouth. The reliance on a LEOSAR with a pass time of 0410 was unreasonable given the environmental conditions and predicted survival times. Even when the second, refined position was received the UKMCC operator was still concerned about a lack of data points, which had no relevance in relation to a position generated by a GNSS embedded EPIRB. The inconsistent terminology for EPIRB positions used by UKMCC and coastguard, coupled with insufficient knowledge of the EPIRB and satellite system, caused confusion and resulted in a delay in prosecuting SAR action. 2.9.3 Speed of response On receipt of Louisa’s EPIRB alert, UKMCC checked that it was a new alert, and then informed CGOC Falmouth and initiated a new incident on Vision 4. UKMCC and CGOC Falmouth then checked the EPIRB’s registration details, and UKMCC categorised the alert as an unresolved 406 alert with a location of 56 45.00N 007 30.00E. UKMCC passed co-ordination to CGOC Falmouth, who then requested CGOC Stornoway to initiate local information gathering. CGOC Falmouth implemented HMCG’s standard operating procedure for satellite distress beacon alerts based on receipt of a detect only or unresolved alert, meaning that the EPIRB’s position will only be confirmed following local information gathering and/or through the unresolved location becoming resolved following a second satellite pass. In either event, co-ordination remained with CGOC Falmouth until the EPIRB’s position was confirmed. Although included as an option in the procedure, in view of the uncertain timescale required to confirm an EPIRB’s position following a detect only or unresolved alert, and the potential inaccuracy of an unresolved location, no consideration is normally given to tasking SAR assets until such time as the position is confirmed. The procedure states: ‘An EPIRB activation is always a DISTRESS ALERT’, implying that SAR tasking should be initiated as soon as the EPIRB’s position is confirmed. The introduction of GNSS enabled EPIRBs provides an opportunity to significantly improve SAR speed of response. A coarse position is subject to an assumed maximum error of 15nm, which is sufficiently accurate to identify the relevant CGOC and to start tasking SAR assets in readiness to deploy once the refined position, accurate to within 120m, is received. Local information gathering is still important to attempt to confirm whether or not an alert is false, prior to receipt of the refined position and a decision to deploy SAR assets. However, it is no longer a priority for the purpose of confirming the EPIRB’s position. The benefit of GNSS enabled EPIRBs with respect to SAR response time was highlighted in the Ocean Way MAIB report.

43 HMCG’s standard operating procedure for satellite distress beacons currently makes no specific reference to the handling of GNSS enabled EPIRB alerts. 2.9.4 Response to Louisa’s EPIRB alert Louisa’s EPIRB alert provided a coarse position that was 5.54nm from its actual position and within CGOC Stornoway’s area of operations. CGOC Falmouth was advised of the EPIRB’s coarse position at 0239. A refined EPIRB position was received by UKMCC at 0251, but it was not until 0316 that CGOC Falmouth passed co-ordination to CGOC Stornoway, and not until 0322 that CGOC Stornoway started SAR asset tasking. Contributing to this delay might have been that CGOC Falmouth was not made aware of the refined position until UKMCC entered the information into Vision 4 at 0312. Although contrary to HMCG’s standard operating procedure for satellite distress beacon alerts, if CGOC Falmouth had passed co-ordination to CGOC Stornoway at 0239, CGOC Stornoway could have started to task SAR assets in readiness to deploy once the refined position had been received. In the event, Barra ALB launched at 0340, incurring an unnecessary delay to the SAR response of 49 minutes. If SAR asset tasking had started once the refined position had been received by UKMCC at 0251, a delay of 31 minutes could have been avoided. However, in the event, this position was not apparent to CGOC Falmouth until it was entered into Vision 4 at 0312. 2.9.5 Incident location The EPIRB’s coarse position was categorised as unresolved, prompting CGOC Falmouth to request CGOC Stornoway to initiate local information gathering in accordance with HMCG’s standard operating procedure. Although not a confirmed position, the EPIRB’s coarse position was actually 5.54nm south-east of Mingulay Bay and in an area that a Stornoway registered fishing vessel could reasonably be expected to be operating. In the absence of an AIS transmission from Louisa, no personal locator beacons being worn by the skipper and crew, and no success in calling the EPIRB’s registered emergency contacts, CGOC Stornoway was reliant on establishing a confirmed position by other means. The benefits of an historical track of a vessel’s movements, and of wearing personal locator beacons, to those involved in SAR, were highlighted in the Harvester MAIB report. Vessel monitoring system (VMS) positional data for Louisa was available to HMCG and could have been used to assist in establishing the vessel’s location. Louisa’s last positon had been transmitted at 0140 and put the vessel in Mingulay Bay. VMS positional data was accessed by neither CGOC Falmouth, CGOC Stornoway nor the NMOC, and it was not until 0356 that CGOC Humber highlighted this information. The Ocean Way MAIB report highlighted the benefits of CGOCs having computer systems capable of gaining access to essential information from all available databases. In response, the MCA gave assurance that the computer equipment in all CGOCs would be capable of accessing vessel positional data in a timely manner. HMCG’s omission to use all available tools suggests insufficient training and manpower resource for its CGOCs to act autonomously.

44 2.9.6 Network activity Although each CGOC can act autonomously, the current coastguard model is of a network of individual nodes which combine to provide support as required. This is overseen by a network controller who manages routine workload to match network resources and provides supervisory cover in the event of an operational incident. The network duty controller was on a break from 0300 to 0430, not returning to his post until after the SAR assets were on scene and the casualties had been located. During this period, the network was being monitored by a trainee who had been employed by HMCG for only 3 months. CGOC Stornoway’s duty SMOO and MOO, who was still undergoing training, attempted to deal with a wide range of tasks. Additional workload was caused by a number of system-related deficiencies and communication problems. A lack of network interaction and supervision meant that the network was ineffective in supporting the SAR operation to the extent of eliminating confusion and ensuring a timely response to the EPIRB alert. The Aquila and Ocean Way MAIB reports highlighted the need for effective network communications and co-ordination between CGOCs and the need to follow effective procedures to ensure important information is not overlooked in demanding situations.

45 SECTION 3 - CONCLUSIONS 3.1 SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

While all on board went to bed in the accommodation, the configuration of Louisa’s machinery was inconsistent with best practice and demonstrated an underestimation of the risks associated with flooding and foundering. [2.3]

The bilge alarm sounder in the crew cabin had previously been disabled and prevented early notification of hold flooding to the skipper and crew. [2.4]

Instead of managing work routines to prevent fatigue and so ensure adequate levels of safety were maintained, the skipper drove the crew and himself to a state of tiredness such that the safety of the vessel and its crew was compromised. [2.4]

Modelling of possible foundering scenarios identified flooding of the hold as the probable cause of Louisa’s foundering. The probable means by which the water entered the hold was from the deck wash hose. [2.5]

The deck wash pump had probably been left running, and it is likely that the hose became wedged between bait boxes stacked aft of the hold hatch coaming, allowing water to spray against the open hold hatch cover. [2.5]

The necessary speed of vessel abandonment gave no opportunity for the skipper and crew to don warm clothes. Their survival in water of 10ºC therefore relied more heavily on entry to the vessel’s liferaft and/or a rapid rescue. [2.6]

It is possible that there would have been more survivors had the skipper and crew’s lifejackets been more effective in prolonging their survival and/or had the rescue services arrived sooner. [2.7]

The liferaft servicing company’s sub-contractor omitted to refill Louisa’s liferaft CO2 cylinder owing to deficiencies in the work process. The work processes and recording of cylinder maintenance were not well defined and no detailed work specification was issued by the liferaft servicing company on despatch of the cylinders. [2.8.2]

The guidance in MGN 499 (M+F) relating to liferaft servicing lacked definition with regard to a liferaft servicing company sub-contracting work in respect of maintaining and refilling liferaft CO2 cylinders. [2.8.2]

IMO testing requirements acknowledge that an approved lifejacket may, under some circumstances, be incapable of satisfactorily turning a face down, unresponsive casualty and keeping their airway clear of the water. This highlights an absence of any factor of safety to take account of the lifejacket being used in operational conditions i.e. with an unspecified range of subjects wearing clothing other than swimming costumes in other than calm conditions. [2.8.3]

46

Assuming that they had donned and secured their lifejackets in accordance with the manufacturer’s instructions, the fact that the skipper and two crew were found face down in the water supports the findings of MAIB trials and the MCA commissioned testing following the accident and questions the suitability of lifejacket water performance test requirements. [2.8.3]

Significant items of safety equipment on board Louisa were out of date for test or replacement. [2.8.5]

The liferaft servicing company operated a system for prompting hirers when payment was due based on invoice dates. With minor adjustments and a focus on servicing requirements, the system could have provided a potentially useful prompt to Louisa’s owners that the liferaft was due for replacement under their existing contract. [2.8.5]

Inconsistent terminology for EPIRB positions used by UKMCC and CGOC Falmouth, coupled with insufficient knowledge of the Cospas-Sarsat system, caused confusion and resulted in an unnecessary delay in prosecuting SAR action. [2.9.2]

HMCG’s standard operating procedure for satellite distress beacons currently makes no specific reference to the handling of GNSS enabled EPIRB alerts. [2.9.3]

HMCG’s omission to use all available tools suggests insufficient training and manpower resource for its CGOCs to act autonomously. [2.9.5]

A lack of network interaction and supervision meant that the network was ineffective in supporting the SAR operation to the extent of eliminating confusion and ensuring a timely response to the EPIRB alert. [2.9.6] 3.2 OTHER SAFETY ISSUES18

The work ethic that resulted in the skipper driving himself and the crew to a level of tiredness that compromised their safety, can be linked to the share remuneration scheme, which encouraged them to grasp every opportunity to continue fishing and maximise the catch whenever suitable conditions prevailed. [2.4]

The PFDs issued to Louisa, which were intended for use during normal fishing operations rather than vessel abandonment, had remained in their original packaging. Despite campaigns to encourage their use, many UK fishermen have not yet accepted the rationale for routinely wearing these valuable safety devices. [2.8.4] 18 These safety issues identify lessons to be learned. They do not merit a safety recommendation based on this investigation alone. However, they may be used for analysing trends in marine accidents or in support of a future safety recommendation

47 SECTION 4 - ACTION TAKEN 4.1 MAIB ACTIONS The MAIB has: Issued a Flyer to the Fishing Industry (Annex H). 4.2 ACTIONS TAKEN BY OTHER ORGANISATIONS The Maritime and Coastguard Agency has:

Issued MGN 553 (M+F) – Life-Saving Appliances – Inflatable Non-SOLAS Liferafts, Lifejackets, Marine Evacuation Systems, Danbuoys and Lifebuoys – Technical Standards and Servicing Requirements, which includes guidance for liferaft servicing companies on sub-contracting work in respect of servicing liferaft CO2 cylinders.

Conducted testing of the Cosalt Premier (32kg or more) solid-filled lifejacket used in the MAIB trials, an additional Cosalt Premier (32kg or more) solid-filled lifejacket and a similar, widely available solid-filled approved lifejacket. All three lifejackets failed to meet required test standards (see section 1.9.8).

48 SECTION 5 - RECOMMENDATIONS The Maritime and Coastguard Agency is recommended to: 2017/130 Urgently conduct research to confirm or otherwise the effectiveness of SOLAS lifejacket water performance test requirements to ensure approved lifejackets will satisfactorily turn a face down, unconscious person onto their back with sufficient orientation and buoyancy to maintain their airway clear of the water. Any shortcomings in the water performance test requirements that may be identified should be brought to the attention of the International Maritime Organization for action. 2017/131 Update and enhance its response to satellite distress beacon alerts, particularly with regard to GNSS enabled EPIRBs, in respect of: ● ●HMCG’s standard operating procedure. ● ●Staff training, in terms of both Cospas-Sarsat system knowledge and HMCG’s operational requirements including the definition of standard terminology in relation to beacon alerts. ● ●Network functionality, reliability, supporting interactivity and resource, in terms of both manpower and equipment. Premium Liferaft Services is recommended to: 2017/132 Update its liferaft servicing procedures to ensure: ● ●Any anomalies in the recorded CO2 cylinder weight can be readily identified. ● ●Definitive work specifications are issued to sub-contractors. ● ●Selected sub-contractors are suitably qualified to undertake the specified work. ● ●Introduce a formal process to advise hirers when their liferafts are due for service. ● ●Compliance with the content of MGN 533 (M+F). Thameside Fire Protection Company Limited is recommended to: 2017/133 Introduce liferaft CO2 cylinder servicing procedures to ensure: ● ●Any anomalies in the recorded CO2 cylinder weight can be readily identified. ● ●Sufficient documentation is held to facilitate servicing a CO2 cylinder in accordance with the liferaft servicing company’s work specification and the particular liferaft manufacturer’s instructions.

49 Louisa’s owners are recommended to: 2017/134 With respect to any fishing vessel they may own in the future, ensure that the vessel remains compliant with the relevant mandatory Code of Practice by: ● ●Developing a planned maintenance system to ensure the vessel is maintained and its safety equipment serviced in accordance with statutory requirements and manufacturers’ instructions. ● ●Conducting formal risk assessments appropriate to the vessel’s anticipated range of activities. Safety recommendations shall in no case create a presumption of blame or liability

Marine Accident Report

FD06053 ｜ 2018_英国_MAIB_entry-to-enclosed-space-on-fishing-vessel-sunbeam-with-loss-of-1-life__Sunbeam_CombinedAnn.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2018 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A MGN 309(F) Fishing Vessels —The Dangers of Enclosed Spaces

1 - MARINE GUIDANCE NOTE

MGN 309 (F)

Fishing Vessels – The Dangers of Enclosed Spaces

Notice to all Owners, Builders, Employers, Skippers and Crews of Fishing Vessels.

Summary

This note warns of the unseen risks from the build-up of gases in “enclosed” spaces including refrigerated Salt Water (RSW) tanks and provides guidance on the avoidance of such risks and of the safe entry into enclosed spaces.

Key Points

• Be aware that unless absolutely necessary, avoid enclosed spaces whilst at sea • Be aware of what constitutes a dangerous enclosed space • Be aware of the procedures for preventing toxic gases in RSW tanks

If entry to enclosed space is necessary

• Be aware of the procedures for prevention of risk • Be aware of how to test the atmosphere • Be aware of the preparations required before entering • Be aware of how to work while in an enclosed space • Be aware how to leave

• Introduction

1. 1 Whilst on board a United Kingdom fishing vessel operating off the west coast of Africa

three crew members tragically lost their lives.

1. 2 The Marine Accident Investigation Branch (MAIB) report identified the release of

dangerous gases from decaying fish, fish waste or offal as the cause of the accident. The source of the gases was a Refrigerated Salt Water (RSW) tank.

1. 3 The tank had not been cleaned or chilled after use and a residue of fish and sea water

remained. Hydrogen, cyanide, hydrogen sulphide and carbon dioxide gas accumulated over a period of several days; the effect being accelerated in tropical temperatures.

1. 4 Gases were released into an occupied working space and inhaled when a side door to the

RSW tank was opened, with fatal consequences to three of the crew and injuries to others.

• 2 -

1. 5 In another case a vessel’s engineer lost his life as a result of high carbon monoxide (CO)

poisoning. A portable, petrol-engined pump was being used with the pump exhausting directly into the engine room. 1.6 These incidents further highlight:

.1 that all spaces receiving little or no ventilation, including RSW tanks, should be treated as potentially dangerous;

.2 the need for proper training in the use of plant and equipment; and

.3 the importance of completing a thorough safety risk assessment and of informing the crew of the measures taken for their safety and health (see MGN20).

• WHAT IS MEANT BY A DANGEROUS “ENCLOSED SPACE”?

1. 1 A dangerous enclosed space:

.1 One that is poorly ventilated or sealed with the oxygen reduced to low levels; or

.2 where toxic and/or flammable gases have built up to dangerous levels.

1. 2 Many forms of chemical reaction can cause low oxygen levels or dangerous gases to build

up, for example the decay of waste material or the exhaust from machinery, one common source is corrosion or rusting, which can significantly reduce the oxygen content in a space. All enclosed spaces therefore need to be treated with caution before opening or entering.

1. 3 Examples of such spaces on fishing vessels are:

.1 RSW tanks; .2 Fuel tanks; .3 Fish holds containing decaying fish, fish waste or offal; .4 Fresh water tanks; .5 Sea or fresh water ballast tanks; .6 Void spaces; and .7 Stores containing chemicals.

• OPENING OR ENTERING AN ENCLOSED SPACE

1. 1 Unless it is absolutely necessary, DO NOT open or enter enclosed spaces whilst at sea. It

is better to wait until the vessel is in port. Then “Call in the specialists” to certify that the air in the space is safe.

1. 2 If entry is necessary, the following procedures are recommended prior to opening or

entering, together with others that may be identified as a result of risk assessment:-

• PREVENTION OF RISK

1. 1 Ensure tanks are empty and not under pressure. Be patient and make sure that tanks are

completely empty.

1. 2 Ensure filling pipes to tanks are isolated, that all valves are shut and notices are put up to

prevent tanks from being filled by someone else.

1. 3 Ventilate spaces THOROUGHLY. Mechanical or natural means of ventilation may be

used, with more time given for natural ventilation.

• 3 -
• TESTING THE ATMOSPHERE

1. 1 A person competent in the use of the equipment should test the atmosphere (air) inside a

space.

1. 2 When testing the air for the first time, breathing apparatus should be worn.

1. 3 An oxygen meter should be used, with a steady reading of 20% before entry is

considered.

1. 4 A combustible gas indicator (“explosimeter”) should be used to test for combustible

gases where these are suspected, especially where this may be a build up of hydrogen (e.g. in fuel tanks).

1. 5 Toxic gases should also be tested for where any risk is suspected, the test equipment

being specific to the gas (e.g. for unclean RSW tanks, test for hydrogen sulphide and hydrogen cyanide).

REMEMBER

IF TESTING APPARATUS IS NOT ON BOARD AND DOUBT EXISTS ABOUT THE SAFETY OF THE SPACE, DO NOT ENTER!

• PREPARING TO ENTER

1. 1 The following should be carried out before entry.

1. 2 Ensure the space is well lit, including the entrance.

1. 3 Remove sources of ignition from clothing, such as matches or lighters.

1. 4 Position appropriate rescue equipment at the entrance, such as breathing apparatus,

(intrinsically safe) torches, life lines and hoists.

1. 5 Limit the number of persons entering the space.

1. 6 Place an attendant at the entrance, who should remain there until everyone has left the

space.

1. 7 Agree and test a suitable means of communication.

• DURING ENTRY TO A SPACE

1. 1 Whilst inside the enclosed space, the following procedures should be observed.

1. 2 Regular communication should be maintained.

1. 3 Continue ventilating the space.

1. 4 Every so often, re-test the atmosphere in the tank and if any doubt exists, stop the

operation.

1. 5 Should an emergency occur, the attendant at the entrance should raise the alarm and

should NOT ENTER the space himself until help has arrived.

1. 6 Breathing apparatus MUST be worn during any rescue from a space. Once making

contact with a casualty, he must be removed to safety and given first aid as necessary.

• 4 -
• LEAVING A SPACE

1. 1 After completion of the operation, everyone should leave the space and the entrance

should be resealed (unless further entry is required).

• RSW TANKS

1. 1 MAIB commissioned research into the toxic gases produced by fish as they decay. The

study revealed that:-

.1 Temperature influences the level of toxic gases produced by decaying fish in sea water;

.2 At 45°C, a “half and half” mix of rotting fish and sea water produced dangerous levels of hydrogen cyanide, hydrogen sulphide and carbon dioxide after only 24 hours;

.3 At 35°C, similar results to 45°C were obtained;

.4 At 20°C, a “half and half mix” of rotting fish and sea water produced dangerous levels of hydrogen cyanide, hydrogen sulphide and carbon dioxide after just 64 hours;

.5 At 5°C, only traces of the three gases were measured after 10 days.

1. 2 The research identifies the importance of the following procedure for the RSW tanks,

particularly when operating in warmer waters:-

.1 Immediately clean all un-chilled RSW tanks after use, removing all residues.

.2 If cleaning has not occurred immediately and a mixture of fish and sea water has been left for more than a few hours, flush the tank through (using the appropriate sea water pump) and fully ventilate the tank.

.3 Ensure adequate ventilation exists in spaces adjacent to RSW tanks.

.4 On no account open or enter RSW tanks known to contain sea water/decaying fish unless the full procedures for entry into an enclosed space have been carried out ( see paragraphs 4-8 above).

• FURTHER ADVICE

1. 1 Crew of vessels that operate at sea for longer periods, particularly on those of 24 metres

and over in registered length may have need to enter enclosed spaces whilst at sea. Further guidance on enclosed spaces may be found elsewhere, such as the Code of Safe Working Practices for Merchant Seamen.

• 5 -

Further Information

Further information on the contents of this Notice can be obtained from:

Fishing Vessel Safety Branch Bay 2/05 Maritime and Coastguard Agency Spring Place 105 Commercial Road Southampton SO15 1EG

Tel :

+44 (0) 23 8032 9163 Fax :

+44 (0) 23 8032 9447 e-mail:

General Inquiries: 24 Hour Infoline infoline@mcga.gov.uk 0870 600 6505

MCA Website Address: www.mcga.gov.uk

File Ref:

MS 088/001/436

Published:

January 2006

© Crown Copyright 2006

Safer Lives, Safer Ships, Cleaner Seas

Printed on material containing minimum 75% post-consumer waste

Annex B Extract of Sunbeam’s Safety Folder including risk assessment Form G12 – Maintenance Work

SUNBEAM SAFETY FOLDER SUNBEAM - FR487 SAFETY FOLDER DOCUMENT CONTROL Contents Preface Authorisation For Issue ) Distribution List Amendment Record Comments Sheet Abbreviations SECTION 1: FISHING VESSEL POLICY STATEMENT ................................................... 1.1 SECTION 2: EMERGENCY PROCEDURES .................................................................... 2.1 SECTION 3: CREW LIST AND STATEMENT ................................................................... 3.1 SECTION 4: SAFETY EQUIPMENT CHECKLIST............................................................ 4.1 SECTION 5: RISK ASSESSMENT.................................................................................... 5.1 Appendix 5.1 Standa Appendix 5.2 Standa~ Appendix 5.3 Standar Appendix 5.4 Standar Appendix 5.5 Standar isk As orm Appendix 5.6 Assessment Forms (Spare forms) Appendix 5.7 Action Lists (Spare forms} Appendix 5.8 Crewmember (Spare forms} Appendix 5.9 Reported Faults (Spare forms} Appendix 5.1 o Substance Abuse Policy ing/seining ing and Dredging FR487 AMENDMENT No. 00 DA TE: August 2000 REVISION 00 PAGE: i of vii

SUNBEAM FR487 SAFETY FOLDER PREFACE This document has been prepared for the SUNBEAM - FR487, and has been compiled following an initial meeting/training session where safety details regarding the vessel were confirmed. Whilst we are confident that this SAFETY FOLDER contains the relevant details required to comply with regulations, it remains the responsibility of the Person Responsible for the Health and Safety (See Section 1 ), to ensure that the information is correct and accurate for your vessel. Please ensure that the required signatories are obtained to complete your SAFETY FOLDER. Regulations The Merchant Shipping and Fishing Vessels (Health and Safety at Work) Regulations 1997 which came into force 31st March j19981fe'tJU el Prepare a written state as often. as may be ap Ith and safety and revise this n five workers}. Carry out risk assessments to identify the risks of health and safety of workers in the normal course of their activities or duties (Applies on all vessels on which any worker is employed). This document has been produced to comply with the Merchant Shipping and Fishing (Health and Safety at Work) Regulations 1997, and to the guidance given in MCA Marine Guidance Note MGN20 (based on the Safety Folder supplied by Seafish Industry Authority). Controlled Document The distribution of this doc will only be accepted for documents issued by the d Review and. Update ot be divided. Responsibility r revisions} of the original . This manual will be subject to regular review and update, when all copy holders will have the opportunity to express opinions and suggest improvements. However, the document control system allows for continuous update of this document. As such, any user may at any time identify an error or suggest an improvement by using the Comments Sheet. Copy Holder's Responsibilities It is the responsibility of the registered copy holder to maintain the accuracy of this document by ensuring that all updates are promptly incorporated and acknowledged. We recommend that you the contents with crewme updated during your onboa rd your vessel, and discuss re for fire can be tested and Your decision to compile this SAFETY FOLDER is a initial step towards a new SAFETY CULTURE, and as this develops the contents of your SAFETY FOLDER will reflect your commitment to the safety practices aboard your vessel. Thank you for selecting Cabern Ltd., to assist you to compile your SAFETY FOLDER. Your SAFETY CULTURE is a continuing process, and Cabern Ltd., will be pleased. to offer our services for future developments ( ). AMENDMENT No. 00 DATE: August 2000 REVISION 00 PAGE: ii of vii

Sunbeam FR487 - Risk Assessment Review~itt9 This risk assessment and health and safety folder in connection with this vessel has been approved and checked by the following persons: Name Position Date Signed Crewmember Statement The safety equipment and procedures have been explained to me and I have been informed of the risk assessments that have been made, and will comply fully with all requirements for health and safety in connection with this vessel. Print Name Date Signed

) SUNBEAM SAFETY FOLDER AUTHORISATION FOR ISSUE Name:

Address:

Telephone:

AMENDMENT No. 00 DATE: August 2000 FR487 REVISION 00 PAGE: iii of vii

SUNBEAM FR487 SAFETY FOLDER DISTRIBUTION LIST This document is a controlled document and is identified by a unique number. Distribution is to Copy Holders as listed below. The Copy Holder of the numbered document is responsible for ensuring that any changes which may be issued are incorporated. Copy No.1 of the document is the Master Copy and is held by the Vessel Skipper. NOTE: Copy No. 1 will provide the latest information available, other controlled copy holders will be provided with the r·ei:~~ve nts. COPY NUMBER 1 AMENDMENT No. 00 DA TE: August 2000 Vessel Skipper or RESPONSIBLE PERSON REVISION 00 PAGE: iv of vii

SUNBEAM SAFETY FOLDER AMENDMENT NO. AMENDMENT No. 00 DATE: August 2000 DATE AMENDMENT RECORD PAGES AFFECTED FR487 SIGNATURE REVISION 00 PAGE: v of vii

) SUNBEAM FR487 SAFETY FOLDER SECTION 1 FISHIN POLICY The following section contains the Safety Policy, Responsible Person information, and vessel details. AMENDMENT No. 00 DATE: August 2000 REVISION 00 PAGE: 1.1 of 3

SUNBEAM FR487 SAFETY FOLDER FISHING VESSEL SAFETY POLICY STATEMENT This Safety Policy Statement sets out how we intend to operate this vessel in compliance with the Merchant Shipping and Fishing Vessels (Health and Safety at Work) Regulations 1997 and other health and safety regulations, to minimise the risk of accidents and ill health. Included in this statement are safety equipment, emergency measures, and risk assessments for activities and areas of the vessel. These will be reviewed every 12 months ( or sooner if significant changes have been made). Personal and Protective Equipment (PPE), information, . training, and the operating procedures necessary for the safety of the vessel and crew will be provided as required by the regulations. Name: Address: Telephone:

Signed: Person responsible for Health and Safety (if different from above) : Address: AMENDMENT No. 00 DATE: August 2000 REVISION 00 PAGE: 1.2 of 3

SUNBEAM FR487 SAFETY FOLDER ·S TION 5 The foll win i the Risk Assessment sheets for this vessel. Spare sheets, Action Lists, and Risk Assessment information. AMENDMENT No. 00 DATE: August 2000 REVISION 00 PAGE: 5.1 of 5

Vessel Name: SUNBEAM Standard Risk Assessment Form Activity or area Maintenance Work (Cont.) Other Assessor Name Possible hazards Electrical tools and equipment Liquid propane gas (LPG) Electrical work Enclosed spaces may have an unsafe atmosphere Possible conseauences Fatal electric shock/injuries mage, chemical , Respiratory lems 'lr.ttinn Suffocaf explosio Electric -u--•·• ... - , ~T)t},; I Deaths I I I Vessel Registration: FR487 F/P s F/PxS 1 3 3 1 I 3 I 3 1 I 3 I 3 1 I 3 ~.,, 1 3 3 Assessor Signature Reference No : G012 MAINTENANCE WORK Control measures necessary with respect to your vessel All tools must be in a safe condition, with effective controls, guards and other safety features correctly in place. Extension cables must be in good condition and carefully rigged to avoid damage or persons tripping over them. Safety circuit breakers should be used to oive protection. Correct PPE to be used. When using chemicals ensure that the safety information supplied with the chemical is studied and all precautions followed. Correct PPE to be used. Ensure that only trained experienced people are permitted to maintain refrigeration systems. Correct PPE to be used. Ensure that all persons are aware of the possible dangers. Not applicable to this vessel. Ensure that electrical fittings are in good order and that fuses or circuit breakers are correctly rated. Only competent persons should be allowed to work on electrical systems. spaces rn.ust beJreate~ \Vi.t~ caution. ,,. " , · · · · H'ations such as painting II persons involved wear nt time after painting for allowing free access. spaces, suCM*as taffffs that have held fuel or oil, do not enter until checks have been made to ensure that it is free of explosive gases and has a safe atmosphere.

1. Do not enter any space that has been sealed without first

making checks that it is safe. Ensure that crewmembers are aware of the dangers and the rocedures to follow. Date Assessed:/.( ........ /, .... ... ;:: ..

Frequency/Probability (F/P) (How likely that harm may occur) 1 Very unlikely 2 Unlikely 3 Likely Severity (S) (How harmful) 1 Slightly harmful 2 Harmful 3 Very harmful Electrical T'.ool~',ndtE•,~ipm,~rat .. : All toolsmust be;!n~ .. s.,fe c~;tfl!P,i'!ith~ffectiv~~ebntrols, guards and other safety features · correctlyi;;,fri;,1,plac~ ~~pstt,111,1~,blesJ1tnust be in good condition and carefulf,Y rigged,,Jo avof9 dl!lmf~e qt P,.rtomJ1,';fripping over them. Safety circuit breakers should·be usetf;to give1l)rotsctltitf Chemicals If chemicals are being used for cleaning or other treatl'li!.§ the safety information supplied with the chemical is · followed. .1,Yy,/.J,;~'t;:;,1•2~r,,i, Refrigerant gases. Refrigerant gases should be treated with caution as a leakage may displace the oxygen from the atmosphere rendering persons unconscious. In liquid form, they will bum the skin. Only trained experienced people should be permitted to maintain refrigeration systems. Liquid propane gas (LPG) known as 1Calor Gas' Leakage of LPG is liable to result in an explosion and fire. Being heavier than air, the gas will collect in the bottom of the vessel and hence may be undetected. Aside from the danger of explosion, the gas may also suffocate. Great care is needed with LPG equipment and portable cylinders should be stored in a well ventilated area. Ensure that all persons are aware of the dangers. · Electrical work Only competent persons should be permitted to carry out electrical maintenance work. High power systems can give fatal electric shocks and 12 or 24 V systems can cause fires if badly maintained. Risk Factors (F/P x S)

1

No action is needed

2

Can be tolerated, but make sure that it does not become worse 3/4 - Take action but subject to it being reasonable and sensibly possible

6

Must be attended to, you must reduce the risk

9

Cannot be accepted and work/activity must not continue Enclosed spaces All enclosed spaces must be treated with caution. I. Ensure adequate ventilation when operations such as painting are taking place and make sure that all persons involved wear a suitable respirator. Allow sufficient time after painting for thorough ventilation of fumes before allowing free access. enclosed spaces, such as tanks that have held fuel or oil, do not enter I checks have been made to ensure that it is free of explosive gases and s a safe atmosphere. Ill. Do not enter any space that has been sealed without first making checks that it is safe. Ensure that er permit to work rocedures to follow. A -,.

Annex C MAIB Safety Bulletin 4/2018

Extracts from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 Regulation 5: “The sole objective of a safety investigation into an accident under these Regulations shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of such an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” Regulation 16(1): “The Chief Inspector may at any time make recommendations as to how future accidents may be prevented.” NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2018 See http://www.nationalarchives. gov.uk/doc/open-government- licence for details. All bulletins can be found on our website: https://www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 Press Enquiries: 01932 440015 Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000 MAR I N E ACCID E NT I N V EST IGATI ON BRAN CH SAFETY BULLETIN SB4/2018 October 2018 Working in refrigerated salt water tanks Fatal enclosed space accident on board the fishing vessel Sunbeam (FR487) at Fraserburgh, United Kingdom on 14 August 2018

MAIB SAFETY BULLETIN 4/2018 This document, containing safety lessons, has been produced for marine safety purposes only, based on information available to date. The Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 provide for the Chief Inspector of Marine Accidents to make recommendations at any time during the course of an investigation if, in his opinion, it is necessary or desirable to do so. The Marine Accident Investigation Branch is carrying out an investigation into a fatal enclosed space accident on board the fishing vessel Sunbeam on 14 August 2018. The MAIB will publish a full report on completion of the investigation. Andrew Moll Chief Inspector of Marine Accidents NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall not be admissible in any judicial proceedings whose purpose, or one of whose purposes, is to apportion liability or blame. This bulletin is also available on our website: www.gov.uk/maib Press Enquiries: 01932 440015; Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000

1 BACKGROUND Sunbeam (Figure 1) was a 56m UK registered pelagic trawler. Its home port was Fraserburgh, Scotland, and it was typically manned by a crew of eleven. In the weeks prior to the accident, it had been fishing for herring in the North Sea and landing its catch in Lerwick, Shetland. The vessel had nine refrigerated salt water (RSW) tanks for storing its catch. On 10 August 2018, Sunbeam arrived at Fraserburgh. It had caught and landed its seasonal quota of herring and was being prepared for a planned refit period. During the refit the vessel’s owner intended to replace Sunbeam’s refrigeration plant. INITIAL FINDINGS At about 0900 on 14 August, Sunbeam’s crew arrived at the vessel’s berth ready to begin work. The vessel’s refrigeration plant had been shut down after landing the final catch at Lerwick, and its RSW tanks had been pumped out and tank lids opened in preparation for deep cleaning. At some time between 1200 and 1350, Sunbeam’s second engineer entered the aft centre RSW tank (Figure 2) and collapsed. At about 1350, the second engineer was seen lying unconscious at the aft end of the tank by a crewmate, who immediately raised the alarm. Three of the vessel’s crew entered the tank and tried to resuscitate the second engineer but they soon became dizzy, confused and short of breath. One of the crew managed to climb out of the tank unaided, the other two crewmen and the second engineer were recovered onto the open deck by two crewmen wearing breathing apparatus. The two crewmen made a full recovery, but the second engineer could not be resuscitated and died. It is unclear when and why the second engineer entered the tank. However, evidence indicated that his intention was to sweep the residual seawater that had settled at the aft end of the tank forward in to the tank’s bilge well. No safety procedures for entering or working in RSW tanks had been completed before he entered the tank. Figure 1: FV Sunbeam

2 Tests of the atmosphere in the tank following the accident showed that the level of oxygen at the bottom was less than 6% (normal level should be 20.9%). Further tests of both the tank atmosphere and residual water samples showed the presence of Freon R22, the refrigerant gas used in the RSW tank’s refrigeration plant. The MAIB’s initial investigation identified that the refrigeration plant sea water evaporators had suffered several tube failures resulting in a number of repairs (Figure 3). It is likely that the refrigerant leaked through one or more failed tubes into the seawater system, and was released into the RSW tank. Freon R22 is four times heavier than air so it will displace oxygen at the bottom of an enclosed space, such as an RSW tank. It is a toxic, tasteless and mostly odourless gas. If it is deeply inhaled, it can cut off vital oxygen to blood cells and lungs. SAFETY LESSONS The RSW tanks on board Sunbeam were, by design, enclosed spaces that did not have a fixed means of positive ventilation. Such spaces can become dangerously hazardous to life. The atmosphere in the tanks can become oxygen deficient through the effects of corrosion, or toxic through the decomposition of sludge or fish, or, as in this case, the accidental release of gas. Other hazards, such as flooding and heat exhaustion can also be a threat to life. Figure 2: Aft centre RSW tank

3 It is the responsibility of vessel owners/operators to ensure that suitable measures are taken to safeguard the crew. All work activities should be subject to risk assessment and safe systems of work should be put in place. Working in enclosed spaces is particularly hazardous, and procedures for entering and working in them should be robust and understood. Similarly, rescue plans need to be put in place and fully understood and should be practised. Widely recognized safety controls for working in enclosed spaces include:

●Atmosphere testing.

●Provision of positive ventilation.

●Safety sentry at entry point.

●Breathing apparatus available for rescue team.

●Safety harness and means of recovering an unconscious person. It is also the responsibility of crew members to behave in a safe manner. This is particularly important when working alone. Tubes plugged by previous repairs Air bubble Figure 3: Starboard evaporator tube leak

4 This was a tragic accident, which nearly resulted in multiple fatalities. The crew did not appreciate the levels of risk they were taking, even after the second engineer had collapsed. The Maritime and Coastguard Agency provides further guidance in its Marine Guidance Note MGN 309 (F) Fishing vessels: the dangers of enclosed spaces, and the Fishermen’s Safety Guide. The findings from an investigation into a similar accident on board the pelagic trawler Oileάn An Óir, in Ireland in 2015, which resulted in two fatalities, also highlight the potential dangers of RSW tanks.1 RECOMMENDATION Sunbeam’s owners are recommended to: S2018/129 Conduct risk assessments specifically for entering and working in RSW tanks and provide safe operating procedures for its crew to follow and appropriate levels of safety equipment to use. Safety recommendations shall in no case create a presumption of blame or liability 1 The Department of Transport, Tourism and Sport in Ireland issued Marine Notice No.43 of 2016.

Annex D MAIB Safety Flyer to the Fishing Industry

SAFETY FLYER TO THE FISHING INDUSTRY Fatal enclosed space accident, on board the pelagic trawler Sunbeam on 14 August 2018 Narrative During the afternoon of 14 August 2018, a second engineer working on board the pelagic trawler Sunbeam was found collapsed inside a refrigerated salt water (RSW) tank. Sunbeam was in Fraserburgh and the crew were preparing for a refit; the evidence suggested that the second engineer had entered the RSW tank to sweep away residual water. When the second engineer was found, other members of the crew rushed to his aid, including three others who entered the tank, one of whom also collapsed. The crew members in the tank were rescued by Sunbeam’s mates, who were wearing breathing apparatus; only three of the four crew who entered the tank survived. Safety Lessons

Sunbeam’s RSW tanks were enclosed spaces because they had limited openings, no ventilation and were not intended for continual worker occupation. The accident happened because the second engineer entered the tank without any of the safety precautions normally associated with such a hazard being in place, specifically: the atmosphere was not monitored, there was no plan for the work or a rescue, and the second engineer was working alone without means of communication. Sunbeam

Entering Sunbeam’s RSW tanks without safety precautions having been implemented had become normalised by the crew as this had been completed without consequence over many years of the vessel’s operations. However, on this occasion the atmosphere could not support life as refrigerant gas (Freon in this case) had leaked into the space through failed evaporator tubes in one of the vessel’s refrigeration plants.

While a leak of refrigerant gas might have been less foreseeable than hazards such as oxygen depletion from corrosion or hydrogen sulphide from rotting fish, it simply serves to underpin the critical need for atmosphere monitoring in enclosed spaces.

All work activities should be subject to risk assessment and safe systems of work. Working in enclosed spaces is particularly hazardous, and procedures for entering and working in them should be robust and understood. Similarly, rescue plans need to be put in place and rehearsed. This flyer and the MAIB’s investigation report are posted on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch Email: maib@dft.gov.uk First Floor, Spring Place Tel: 023 8039 5500 105 Commercial Road Southampton SO15 1GH Publication date: December 2020 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an such investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This safety flyer is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2020 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned.

FD06054 ｜ 2018_英国_MAIB_two-catastrophic-engine-failures-one-resulting-in-a-fire-on-board-ro-ro-passenger-ferry-wight-sky__WightSky-Annexes.pdf.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2018 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A Wightlink Ltd - Lymington Harbour Safe Operating Procedure

Annex B Cylinder Liner Cavitation Erosion On Diesel Engines

Page 1 of 2

This information is provided from the best available sources. However, AERA does not assume responsibility for data accuracy or consequences of its application. Members and others are not authorized to reproduce or distribute this material in any form, or issue it to their branches, divisions, or subsidiaries, etc. at a different location, without written permission.

© AERA Engine Builders Association. All rights reserved. Cylinder Liner Cavitation Erosion On Diesel Engines

The AERA Technical Committee offers the following information on cylinder liner cavitation erosion on diesel engines. This information should be considered anytime maintenance is being performed on a diesel engine.

The cavitation erosion that is found in diesel engines on the exterior side of the cylinder liners (wet) has been a theme of investigation by engine and corresponding component manufacturers, even though no definite manner of eliminating the problem has been found. See Figure 1.

Figure 1. Severely Eroded liner

The degree and size of erosion or decay as well as its form and distribution on the affected zones can vary from engine to engine and inclusively from cylinder to cylinder within the same engine.

The affected areas form vertical strips or patches sometimes in alignment with the thrust face of the piston within the cylinder, or also formed immediately over the top sealing ring of the liner. The erosion can penetrate the wall of the cylinder and permit the flow of antifreeze/coolant to the oil or vice-versa.

The cavitation erosion is caused by excess in harmonic vibrations of the engine and in some cases by loose fitting liners that result in a fast formation and implosion of small vapor bubbles within the coolant which attack the cylinders’ liner wall. The vibration in combination with collapsing bubbles produces an

Page 2 of 2

This information is provided from the best available sources. However, AERA does not assume responsibility for data accuracy or consequences of its application. Members and others are not authorized to reproduce or distribute this material in any form, or issue it to their branches, divisions, or subsidiaries, etc. at a different location, without written permission.

© AERA Engine Builders Association. All rights reserved. erosive effect over the surface of the liner’s exterior face. This occurs as the piston moves up and down within the cylinder causing vibration especially over the area of thrust in the cylinder liner.

We can say that the wall of the sleeve, which comes in contact with the coolant, quickly moves inwards and outwards striking these vapor formations. During this process, tiny bubbles are formed that implode or collapse violently producing shock waves against the liner’s wall. The results of these implosions at the impact area of the liner’s wall have been calculated to reach over 10,000º F with pressures of over 10,000 psi.

A specific material, which would prevent cavitation erosion within reasonable cost, has not yet been found. Nevertheless, some coatings can be applied to delay cavitation erosion until it is time for major engine repairs.

In order to reduce or prevent cavitation, manufacturer recommendations should be followed.

In many cases, cavitation can be avoided by reducing harmonic vibrations. For example, by making sure the injection complies with the manufacturer’s specifications, the engine’s speed is governed according to the manufacturer’s data and that the control functions. Also, do not underestimate the importance of correct clearances between liner and cylinder blocks; incorrect liner fit is a serious contributor to liner vibration resulting in cavitation erosion.

Engine manufacturers have specific additives for the coolant called SCA (supplementary cooling additive). These additives will form a protective coating over the cylinder liner, which is exposed to the coolant, therefore reducing cavitation damage. With time, additive concentration reduces in the system, consequently making it necessary to follow system maintenance recommendations in order to keep the adequate levels. The following points will be supplied:

• pH control to avoid corrosion
• Water hardness control to avoid mineral deposit formation
• Cavitation protection

In all cases, it is required that all manufacturer’s recommendations regarding coolants, base additives, additives, coolant filters as well as maintenance schedules should be strictly adhered to or followed at all times.

Annex C Event 1: number 6 connecting rod debris analysis

1. UK0-0870

Exponent is an international consultancy with offices located in UK, Germany, Switzerland, USA, China and Hong Kong SAR 26th February, 2019 , Volvo Penta UK Subject: Volvo Penta – Wightlink Ferry Exponent Project No.: 1808759.UK0-0870 Dear , At your request, Exponent International Limited (Exponent) has carried out an analyses of the material located within the lubricating oil channel of connecting rod no.6, during the stripdown of a Volvo Penta D16 engine, serial number 1016027578. This stripdown was carried out in Warwick on the 29th January, 2019, by Volvo Penta staff in the presecence of representatives from the Marine Accident Investigation Branch, Lloyd’s Register, Wightlink and RK Marine as part of ongoing investigations into various engine failures which have occured onboard Wightlink ferries since August 2017. At an inspection on the 19th February, 2019 at an Exponent Labortory within The Natural History Museum, London, a connecting rod from the subject engine was visually inspected and samples collected from the lubrication channel by Dr. , an Exponent Materials Engineer. The lubrication channel openings and particles after collection are shown in FIGURE 1, FIGURE 2 and FIGURE 3, respectively. These particles were cleaned in an organic solvent (acetone) and prepared for scanning electron microscopy (SEM) and qualitative chemical analysis via energy dispersive x-ray spectroscopy (EDS). No analysis of the connecting rod material was performed.

Volvo Penta UK 26th February, 2019 Page 2

1. UK0-0870

SEM is an analysis technique whereby high magnification images of a sample can be created by interrogating the surface with an electron beam. When the electron beam from an SEM interrogates a material surface, x-rays are released by the sample.1 These x-rays have a characteristic energy that depends on the element being interrogated. As a result, it is possible to determine what elements are present in a material through EDS.

In certain cases, it may be possible to estimate the relative amounts of each element present through a semi-quantitative analysis, but the sample must be specially prepared (i.e., flat and polished) to successfully perform this analysis;1 the size and morphology of the particles precluded such semi-quantitative analysis, thus the presence of particular elements in the particles is noted without specific information regarding the concentration of each element in the particles. Additionally, EDS does not provide information regarding the electronic state of the elements (i.e., what elements are bonded together), so specific information regarding the composition of the particles is not provided. Two particles, both approximately 1 mm3, were collected from the subject connecting rod and arbitrarily named “Particle 1” and “Particle 2.” Both particles were examined via SEM/EDS. Representative SEM images of the Particles 1 and 2 are shown in FIGURE 4 and FIGURE 5, respectively. The contrast in FIGURE 4 and FIGURE 5 represents compositional differences: brighter areas represent heavier elements (e.g., iron, Fe) while darker areas represent lighter elements (e.g., carbon, C). Note, the contrast is not uniform across all of the images, thus a particular shade of grey does not represent the same element image-to-image. Smaller, micron-scale particles were observed decorating the larger 1 mm3 particles. The larger 1 mm3 particles appear to consist mainly of carbon (C), while the smaller particles consist of a mix of aluminium (FIGURE 6), iron (FIGURE 7), nickel (FIGURE 7), copper (FIGURE 7) and lead (FIGURE 8). Traces of other elements, including but not limited to, phosphorus (P), calcium (Ca), sulphur (S) and silicon (Si) were also detected in various small particles decorating Particles 1 and 2.

1 ASM Handbook Volume 10 Materials Characterization, ASM International, Metals Park, OH USA, 1986, pp. 518 – 535.

Volvo Penta UK 26th February, 2019 Page 3

1. UK0-0870

FIGURE 1. Representative photograph of the subject connecting rod.

FIGURE 2. Representative photographs of the lubrication channel openings in the subject connecting rod. Particles were visible in both openings.

Volvo Penta UK 26th February, 2019 Page 4

1. UK0-0870

FIGURE 3. Representative photograph of the two particles collected from the subject connecting rod.

FIGURE 4. Representative SEM image of Particle 1.

Volvo Penta UK 26th February, 2019 Page 5

1. UK0-0870

FIGURE 5. Representative SEM image of Particle 2.

Volvo Penta UK 26th February, 2019 Page 6

1. UK0-0870

FIGURE 6. Representative EDS spectrum of a small particle decorating Particle 1. Aluminium (Al) was detected in this smaller particle.

Volvo Penta UK 26th February, 2019 Page 7

1. UK0-0870

FIGURE 7. Representative EDS spectrum of a small particle decorating Particle 1. Iron (Fe), nickel (Ni) and copper (Cu) were detected in this smaller particle.

Volvo Penta UK 26th February, 2019 Page 8

1. UK0-0870

FIGURE 8. Representative EDS spectrum of a small particle decorating Particle 2. Lead (Pb) was detected in this smaller particle.

Volvo Penta UK 26th February, 2019 Page 9 1808759.UK0-0870 To summarise, during the analyses of No.6 connecting rod lubricating oil channel taken from a Volvo Penta D16 engine, serial number 1016027578, carbon particles were detected surrounded by wear metal material. If we can be of further assistance in this matter, pelase do not hesistate to contact us. Your sincerely, Exponent International.

FD06055 ｜ 2018_英国_capsize-and-sinking-of-fishing-vessel-solstice-with-loss-of-1-life_2018-20-_Solstice_A.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2018 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A The Code of Practice for the Safety of Small Fishing Vessels mandatory equipment for decked vessels of less than 10m registered length

21 ANNEX 1.4 CODE OF PRACTICE FOR THE SAFETY OF SMALL FISHING VESSELS: CHECK LIST OF REQUIREMENTS: SEE ADDITIONAL GUIDANCE IN ANNEX 2 Equipment need not be MCA approved provided it is fit for its intended purpose. DECKED Vessels of less than 10 metres (L) “Decked vessels” means a vessel with a continuous watertight weather deck that extends from stem to stern and has positive freeboard throughout, in any condition of loading the vessel. Item Remarks/compliance Expiry/Service Date Liferaft(s) (for vessels of 7 metres (L) and over) - sufficient capacity for all persons on board vessel and appropriate for area of operation – See Annex 2 for guidance Lifejackets - 1 per person 2 Lifebuoys (1 with 18 metre buoyant line attached) or 1 Lifebuoy (fitted with 18 metre buoyancy line) +1 Buoyant Rescue Quoit 3 Parachute Flares 2 Hand-held Flares 1 Smoke Signal (buoyant or hand held) 1 Multi-purpose Fire Extinguisher (fire rating 5A/34B) 1 Multi-purpose Fire Extinguisher for oil fires (fire rating 13A/113B) Gas Detector 1 Fire Blanket (light duty) in galley or cooking area (if applicable) Fire Detectors 1 Fire Pump + Hose and 1 Fire Bucket and lanyard 1 Satellite EPIRB or Personal Locator Beacon(s) – 1 per person VHF Radio – DSC fixed or hand held Bilge Pump Bilge Level Alarm Approved Navigation Lights & Sound Signals Anchor and cable/warp Compass Waterproof Torch Medical Kit in accordance with The Merchant Shipping and Fishing Vessels (Medical Stores) Regulations 1995 No.1802 or any superseding regulations Radar Reflector CO Alarms for every enclosed space that has a fired cooking or heating appliance or where engine exhausts penetrate the wheelhouse or crew space Note: The checklist represents the minimum safety equipment requirements and owners should consider carrying additional safety equipment. It is recommended that if you carry the Satellite EPIRB, you also carry Personal Locator Beacons for each member of the crew, and if you carry Personal Locator Beacons, you also carry a Satellite EPIRB. The liferaft, which is mandatory for vessel of 7 metres (L) and over and strongly recommended for vessels under 7 metres (L), should be fitted in accordance with the manufacturer’s instructions. Coastguard Operations Centres maintain a listening watch only on VHF Channel 16. The primary means of distress and urgency alerting should be via VHF DSC. Substantial modifications or alterations affecting the vessel’s dimensions, structure or stability, the removal or repositioning of equipment, changes in the vessel’s mode of fishing and/or its gear or the fitting of additional equipment shall be investigated, prior to making any changes, to ensure that the vessel ensure that the vessel is stable for its intended purpose and/or will continue to comply with the stability requirements of this Code. In addition such modifications or alterations shall only be carried out after consultation and with the approval of the MCA.

Annex B Coastguard OMS Vessel overdue guidance

Vessel - Overdue Incidents involving reports of vessels missing or overdue The report may be made because the vessel failed to make a scheduled call or because of its non arrival at a port/harbour Overdue leisure vessels will often be reported by a concerned friend or relative Additional considerations need to be made for commercial vessels reported overdue Related Links SafeTrx EMSA Maritime Applications Portal SAD - Datum Area SAD - Datum Line Information Gathering Who - first informant details Who -vessel • Name, callsign and MMSI number • Type, size and description of vessel, CG66 • Radio • Mobile Phone / Communications devices • satcomms • EPIRB • SART • Other tracking device What - Activity or passage details • Departure point • Departure time • Car registration number • Activity • ETA Try to contact the vessel by any means possible as soon as practical Persons On Board • How many • · Profiles (e.g. age; sex, physical state etc.) • Experience, Familiarity with the area ls/might skipper be using SafeTrx trip monitoring or similar App? 0 Ask if there are any other potential sources of information, e.g. Clubs, Website 0 Look for latest entries on their Facebook / social media pages

Ask the first informant to call if they have any further information, or if the vessel re.turns Declare Phase and Follow Phase Actions Mission Conduct O could info from IMDatE be useful (e.g. FVMS) Vessel • Type and size • Capability and equipment • Condition (i.e. seaworthiness) • What means of communication are there on board mobile phone, VHF radio etc. • LSE Persons on board • How many • Crew composition • Profiles age, sex, health etc. • Experience Vessel Activity If fishing vessel A.overdue FVs often due to distress situation A.check FVMS via IMDatE When was vessel due in port or to market Is owner/agent aware or concerned - if 1st informant, probably something wrong Does FV have: • EPIRB • AIS • MAFF/DEFRA tracker (VMS - check IMDatE) Do crew have PLBs If making passage • Last known position • Type and size of vessel • Did the vessel depart • Is the ETA reasonable given environmental/ tidal conditions or could the vessel still be under pa·ssage Weather over passage Tidal conditions

• • Could the vessel have arrived or taken port en-route If not making passage • What activity was being undertaken • Location or area of activity • Radius of capability How far could the vessel get before running out of fuel or needing to return A1f the vessel departed from within another Ops Centre's area of responsibility, contact and liaise with the other Ops Centre A1f vessel using, or possibly using, SafeTrx app, contact the NMOC Search Plan 0 1s a search plan required • Alert and Distress Phase - Request a RIPA enquiry using a SPOC to try to obtain location information on any mobile phones/devices Consider • · Datum Area • Datum Line Environment • Poor weather may have caused the vessel to take shelter • Poor visibility may have caused the vessel to become lost or delayed • Strong head winds or flat calm (for sailing vessels) may cause delay • Tide -direction and rate over period A Forecast conditions and possible haza'rds along or near route must be considered when determining emergency phase and level of response 0 Contact On Call Duty Personnel - see Duty Rosters for contact details • Duty Controller • Duty Press Officer • Duty CPSO • Lloyds • DEFRA/SFPA • Local Fisheries Officer • MAIS • Send EU_SITREP to CERS • MCA Duty Surveyor (if applicable) • FCO (if applicable) • MCA Enforcement Branch (if applicable) Taskin Facilities

• • • • • Phases • Distress • Alert • Uncertainity Mission Conduct • • • • • Welfare • Welfare • Blood Borne Viruses • TRiM Prdcedure Conclusion • • Revision: 2.0 - Published:

Annex C Safety Flyer to the Fishing Industry

SAFETY FLYER TO THE FISHING INDUSTRY Fishing vessel Solstice, capsize and sinking resulting in the loss of one life, 26 September 2017 Narrative At 1938 on 26 September 2017, the 9.9m fishing vessel Solstice (Figure 1) capsized in calm weather conditions about 7 miles south of Plymouth. The skipper and crewman were rescued from the vessel’s upturned hull (Figure 2) about 5½ hours later, but the vessel’s owner was trapped and drowned in the wheelhouse. Solstice later sank. The scallop dredger had recently been modified to operate as a stern trawler and its owner, skipper and crewman were in the process of hauling their second catch of the day on board when the capsize occurred. The net’s cod-end was full of fish, moss and sand, and started to move uncontrollably along the transom as the vessel rolled in the light swell. The capsize was rapid, and the crew did not have time to raise the alarm before they entered the water. As the vessel was not equipped with an emergency position indicating radio beacon (EPIRB) and the crew did not carry personal locator beacons (PLBs), they were wholly reliant on family and friends realizing they were overdue and alerting the coastguard. Figure 1: Solstice Figure 2: Upturned hull floating on surface (photograph taken the following morning)

Safety lessons • Solstice capsized because it did not have sufficient transverse stability to safely lift the contents of its net on board over the stern. This was primarily because the weight in the net was excessive and the height of the lifting point at the stern was high (Figure 3). • The dangers of modifying small fishing vessels and changing fishing methods are well known and the consequences on stability are often fatal. However, the vessel had no stability data and a safe method of fishing had not been developed. • It is important to have a clear understanding of your vessel’s stability, but the impact that fishing methods have on stability cannot be emphasized enough. Talk to your local fishing vessel surveyors and seek expert advice if necessary before commencing any structural modifications, but especially modifications to change or alter the method of fishing. • It was obvious that the weight in the net was excessive, but the crew remained determined to lift the catch on board. No catch, no matter how valuable, is worth a life. Stop and seriously consider your options and, ultimately, be prepared to let it go. • The survivors were extremely fortunate that Solstice’s upturned hull remained afloat as they were not wearing lifejackets when they entered the water, and it took over 5 hours for them to be found and rescued. Personal flotation devices should always be worn when working on deck and emergency use lifejackets should be readily available. • It was also fortunate that the survivors were found. Their friends and family did not know where they had been fishing and, because Solstice was not equipped with an automatic identification system (AIS) transceiver, the coastguard could not identify the vessel’s last recorded position. The carriage of an EPIRB and/or PLBs would have resulted in an immediate and location- focused emergency response. The carriage of an AIS transceiver in this case would have been of invaluable benefit. This safety flyer and the MAIB’s investigation report are on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch Email: maib@dft.gov.uk First Floor, Spring Place Tel: 023 8039 5500 105 Commercial Road Southampton SO15 1GH Publication date: 6 December 2018 Figure 3: Solstice scallop dredging four weeks before the accident Lifting point

FD06056 ｜ 2018_英国_engine-failure-and-subsequent-fire-on-ro-ro-cargo-vessel-finlandia-seaways-with-1-person-i.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2018 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

ACCIDENT REPORT ACCIDENT REPORT Report on the investigation of the engine failure and fire on board the ro-ro cargo vessel Finlandia Seaways resulting in injury to one crewman 11 miles east of Lowestoft 16 April 2018 SERIOUS MARINE CASUALTY REPORT NO 2/2021 FEBRUARY 2021

The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2021 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch First Floor, Spring Place 105 Commercial Road Southampton SO15 1GH United Kingdom Email: maib@dft.gov.uk Telephone: +44 (0) 23 8039 5500 Press enquiries during office hours: 01932 440015 Press enquiries out of hours: 020 7944 4292

CONTENTS GLOSSARY OF ABBREVIATIONS AND ACRONYMS SYNOPSIS 1 SECTION 1 - FACTUAL INFORMATION 2 1.1 Particulars of Finlandia Seaways and accident 2 1.2 Narrative 3 1.2.1 Passage from Zeebrugge 3 1.2.2 The engine failure and subsequent fire 3 1.2.3 Emergency response 4 1.2.4 Engine room fire damage 12 1.3 Crew 12 1.4 Environmental conditions 13 1.5 Finlandia Seaways 13 1.6 Main engine 13 1.6.1 Overview 13 1.6.2 Engine design and key components 14 1.6.3 Engine condition monitoring and protection systems 16 1.6.4 Post-fire inspection 17 1.6.5 Engine room alarms 21 1.6.6 Planned maintenance 21 1.6.7 Outsourcing of main engine work 22 1.7 Diesel Service Group UAB 22 1.7.1 Overview 22 1.7.2 Certification 22 1.8 Connecting rod piston pin bush renewal 23 1.8.1 MAN guidance and procedure 23 1.8.2 Diesel Service Group guidance and procedure 24 1.9 Engine manufacturer’s post-failure engine inspection and guidance 27 1.9.1 Engine inspection and failure mode analysis 27 1.9.2 Botnia Seaways 31 1.9.3 Engine design modifications and updates 33 1.10 Metallurgical analysis 33 1.11 External oversight 34 1.12 Emergency response 36 1.12.1 Carbon dioxide fixed fire-extinguishing system 36 1.12.2 Means of escape 37 1.12.3 Emergency escape breathing devices 39 1.12.4 Voyage data recorder recovery 40 1.13 Previous/similar accidents 40 1.13.1 Thetis D 40 1.13.2 Power station, Spain 41 SECTION 2 - ANALYSIS 42 2.1 Aim 42 2.2 Accident overview 42 2.3 Engine failure 42 2.3.1 Sequence of events 42 2.3.2 Identification of the component failure 43 2.3.3 Failure mode 43

1. 4 Fracture mechanism

44 2.4.1 General 44 2.4.2 Fatigue fracture 44 2.4.3 Notch effect and toughness 44 2.4.4 Localised heating 46 2.4.5 Inspection and analysis 47 2.5 Maintenance management 47 2.5.1 Overview 47 2.5.2 Connecting rod small end overhaul process 48 2.5.3 Rotation of connecting rod small ends 48 2.5.4 Diesel Service Group quality control 49 2.5.5 DFDS quality control 50 2.5.6 Classification Society oversight 50 2.6 Emergency response 51 2.6.1 Overview 51 2.6.2 Carbon dioxide bottle identification 51 2.6.3 Effective numbers and distribution of EEBDs 52 2.6.4 VDR 52 SECTION 3 - CONCLUSIONS 53 3.1 Safety issues directly contributing to the accident that have been addressed or resulted in recommendations 53 3.2 Safety issues not directly contributing to the accident that have been addressed or resulted in recommendations 54 3.3 Other safety issues not directly contributing to the accident 54 SECTION 4 - ACTION TAKEN 55 4.1 MAIB actions 55 4.2 Actions taken by other organisations 55 SECTION 5 - RECOMMENDATIONS 57 FIGURES Figure 1: Finlandia Seaways Figure 2: Engine room purifier room Figure 3: Purifier room door Figure 4: Access to ECR, CCR from vehicle deck Figure 5: Control station for the engine room fuel systems’ quick-closing valves Figure 6: Purifier room and engine room tween deck (shaded blue) with 3/E’s escape route Figure 7: Elevation showing the primary escape route from the purifier room to the main vehicle deck and the secondary escape route taken by the third engineer

Figure 8: Escape route vertical ladder and safety chains Figure 9: Upper vehicle deck Figure 10: CO2 remote release station Figure 11: CO2 bottle room Figure 12: Fire damage to upper engine room (looking forward) Figure 13: MAN 48/60 engine (engine shown 14V48/60B) Figure 14: Connecting rod assembly Figure 15: B5 small end with piston pin Figure 16: Crankcase oil mist detector Figure 17: Port side main engine Figure 18: Starboard crankcase damage at B5 Figure 19: Section of A5 connecting rod (after removal from engine room) Figure 20: B5 components on starboard side deck plates Figure 21: A5 cylinder viewed from engine sump Figure 22: B5 liner (looking up to inlet/exhaust valves from engine sump) Figure 23: Diesel Service Group workshop and milling machine (inset) Figure 24: Diesel Service Group: small end piston pin bearing bush renewal procedure Figure 25a: A5 small end upper half Figure 25b: A5 small end fatigue fracture Figure 25c: A5 looking forward and fatigue fracture Figure 26: MAN – Finlandia Seaways’ small ends investigation Figure 27: Botnia Seaways small end heat damage (circled) Figure 28: Carbon dioxide bottle operating valve Figure 29: Distribution of EEBDs in or near to the engine room and escape routes Figure 30: Stress/strain curve Figure 31: Crack propagation Figure 32: Plastic protractor under polarising light showing stress concentration

TABLES Table 1: Key alarms relating to the engine failure and engine room fire Table 2: MAN’s planned maintenance schedule Table 3: A5 small end overhauls and inspections Table 4: Finlandia Seaways’ main engine connecting rods, including small end damage Table 5: Lloyd’s Register Component Survey List ANNEXES Annex A MAN inspection report Annex B MAN customer information no. 344/2015 Annex C The Test House (Cambridge) Ltd metallurgical report Annex D Lloyd’s Register Examination of Surveyable Machinery Items by Chief Engineers

GLOSSARY OF ABBREVIATIONS AND ACRONYMS

APT

Annual Performance Test

bar

metric unit of pressure: 1bar is equal to 100,000 pascals

BSU

Bundesstelle für Seeunfalluntersuchung CCR

Cargo control room C/E

Chief engineer

C/O

Chief officer CO2

Carbon dioxide CPP

Controllable pitch propeller

CSM

Continuous Survey Machinery

DFDS

DFDS Seaways AB-Lithuania

DSG

Diesel Service Group UAB ECR

Engine control room

EEBD

Emergency escape breathing device Hz

Hertz IMO

International Maritime Organization kg

kilogramme kW

kilowatt

LMSA

Lithuanian Maritime Safety Administration LR

Lloyd’s Register

LTSA

Lithuanian Transport Safety Administration

MAN

MAN Diesel and Turbo SE renamed MAN Energy SE on 26/6/2018

MAN PrimeServ

Brand name for MAN’s aftersales and maintenance service division mbar

millibar

MCA

Maritime and Coastguard Agency

MGN

Marine Guidance Note mm

millimetre nm

nautical miles

OMD

Oil mist detector

OOW

Officer of the watch Piston pin bush

Connecting rod small end piston pin bearing bush

PMS

Planned Maintenance System

RINA

Registro Italiano Navale

RMRS

Russian Maritime Register of Shipping rpm

revolutions per minute

SOLAS

International Convention for the Safety of Life at Sea 1974, as amended

TAIID

Lithuanian Transport Accident and Incident Investigation Division

TTH

The Test House (Cambridge) Ltd

UPS

Uninterruptible power supply

UTC

Universal Co-ordinated Time V

volt

VDR

Voyage data recorder VHF

Very high frequency

2/E

Second engineer 3/E

Third engineer TIMES: all times used in this report are UTC unless otherwise stated

1 SYNOPSIS At 2003 on 16 April 2018, the Lithuanian registered ro-ro cargo vessel Finlandia Seaways suffered a catastrophic main engine failure that caused serious structural damage to the engine and a fire in the engine room. The vessel’s third engineer, who was on duty in the engine room at the time, suffered serious smoke-related lung, kidney and eye injuries during his escape. Finlandia Seaways was 11nm east of Lowestoft on a regular voyage from Zeebrugge to Rosyth when one of the main engine’s connecting rods broke. Several of the engine’s major internal rotating components were thrown through the side of the crankcase into the engine room, and a short but intense fire occurred. Within 20 minutes the crew had conducted a muster, sealed the engine room, activated its carbon dioxide fixed fire-fighting system and extinguished the fire. The third engineer was recovered by coastguard helicopter to Norfolk and Norwich hospital for medical care, and made a successful recovery. The MAIB’s technical investigation was carried out with support of the Lithuanian Transport Accident and Incident Investigation Division and the engine manufacturer, MAN Diesel and Turbo SE. The investigation identified that a fracture of the connecting rod small end had led to the sudden failure of the main engine. The investigation also found that the method used to replace the connecting rod small end piston pin bearing bushes by the vessel manager’s maintenance support contractor had introduced stress raisers that signficantly increased the likelihood of crack initiation and fatigue failure. Other factors that contributed to the engine failure included: standards of maintenance management; lack of appreciation of the importance of following the engine manufacturer’s instructions for the removal and refitting of the piston pin bearing bushes; and external oversight of the engine maintenance process. With regard to the emergency response, although the carbon dioxide fire-fighting system was activated successfully, the third engineer was fortunate to have survived given that there were no emergency escape breathing devices on his escape route. In common with other accidents in which carbon dioxide has been released following a fire, the inability to confirm which gas bottles had discharged hampered re-entry to the engine room. In addition to this, the voyage data recorder did not record the incident due to the uninterruptible power supply failing. On 27 July 2018, the MAIB and the Lithuanian Transport Accident and Incident Investigation Division wrote to MAN Diesel and Turbo SE and the vessel’s classification society recommending that they provide technical advice to Finlandia Seaways’ operator to reduce the likelihood of a similar accident occurring in the future. In response, Finlandia Seaways’ sister vessel was withdrawn from service and its engine connecting rods removed and replaced. In addition to many actions taken by stakeholders as a result of this accident, further recommendations aimed at addressing the safety issues raised in this report have been made to the vessel operators, DFDS Seaways AB-Lithuania and its engine maintenance support contractor, Diesel Service Group.

2 SECTION 1 - FACTUAL INFORMATION 1.1 PARTICULARS OF FINLANDIA SEAWAYS AND ACCIDENT SHIP PARTICULARS Vessel’s name Finlandia Seaways Flag Lithuania Classification society Lloyd’s Register IMO number/fishing numbers 9198721 Type Ro-ro cargo vessel Registered owner DFDS Seaways AB-Lithuania Manager(s) DFDS Seaways AB-Lithuania Construction Steel Year of build 2000 Length overall 162.58m Gross tonnage 11530 Minimum safe manning 13 Authorised cargo Freight vehicles VOYAGE PARTICULARS Port of departure Zeebrugge Port of arrival Rosyth Type of voyage International Cargo information A mixture of heavy goods vehicles, containers, cars, campervans and caravans Manning 19 Passengers 4 MARINE CASUALTY INFORMATION Date and time 16 April 2018 at 2003 (2203 ship’s time) Type of marine casualty or incident Serious Marine Casualty Location of incident 11 miles east of Lowestoft Place on board Engine room Injuries/fatalities One injury Damage/environmental impact Main engine structural damage; fire damage in engine room Ship operation On passage Voyage segment Mid-water External & internal environment South-westerly wind, slight sea Persons on board 23

3 1.2 NARRATIVE 1.2.1 Passage from Zeebrugge At 1445 on 16 April 2018, Finlandia Seaways’ (Figure 1) main engine was started in preparation for leaving port. On duty in the engine room were the chief engineer (C/E), second engineer (2/E) and an electrician. The vessel’s relief C/E was also in the engine room. At 1510, engine control was passed to the bridge and, at 1545, the vessel departed Zeebrugge on passage to Rosyth, Scotland. At 1600, the on watch 2/E was relieved by the third engineer (3/E) and the former left the engine room. At 1610, the vessel cleared Belgium’s 12nm coastal limit and the engineers began to change over the engine fuel supply from low sulphur marine diesel oil to heavy fuel oil. This was completed at 1642 and, by 1800, the C/E, relief C/E and electrician had all left the engine room. The on watch 3/E monitored machinery operating parameters from the engine control room (ECR) and undertook hourly inspection rounds of the engine room. Figure 1: Finlandia Seaways The ro-ro cargo vessel Finlandia Seaways Image courtesy of www.shipspotting.com 1.2.2 The engine failure and subsequent fire At about 2000, the 3/E left the ECR and commenced his hourly rounds of the engine room and machinery spaces. Shortly after entering the purifier room (Figure 2), he heard loud metallic knocking sounds coming from the engine room. At the same time, powerful vibrations were being felt on the bridge by the officer of the watch (OOW), who immediately checked the propeller pitch angle and main engine speed gauges. The OOW saw that both gauges were fluctuating and immediately reduced the propeller pitch from 60% to 25%. He then saw that the fuel consumption meter was also fluctuating and giving abnormal readings, while various engine warning lights were illuminated.

4 Figure 2: Engine room purifier roomThe purifier room on board Finlandia Seaways The 3/E opened the purifier room door, looked out into the engine room and saw smoke rising from the port forward end of the main engine. The noise and vibration levels rapidly increased and, sensing danger, the 3/E ran back into the purifier room and crouched behind one of the purifiers. Almost immediately, at 2003, there was a loud bang and flames flashed past the open purifier room door (Figure 3) and the space filled with thick black smoke. The ship blacked out1, but within seconds the emergency generator cut-in and the emergency lighting came on. Despite this, the smoke in the engine room had reduced the 3/E’s visibility to almost nothing. 1.2.3 Emergency response The master, who had been alerted by the strong ship vibrations, arrived on the bridge as the blackout occurred. The fire alarm system had been activated and the bridge alarm panel indicated that all zones within the engine room were affected. The master immediately took command and put the vessel’s controllable pitch propeller (CPP) to zero pitch. He then instructed the OOW to sound the general alarm and make an intercom broadcast stating that it was a genuine emergency and that the crew should go to their muster stations. At about the same time, the chief officer (C/O) arrived on the bridge and reported that he had seen smoke emitting from the funnel. 1 The sudden and total failure of the ship’s main electrical power supply.

5 Figure 3: Purifier room doorOpened door to the purifier room with some heat damage visible Like the master, many of the crew members had been alerted by the ship vibrations and had already begun to react. The C/E had noticed that the main engine’s No.A5 cylinder exhaust gas temperature deviation alarm had activated on his cabin alarm panel. He initially thought it could be related to a fuel system problem. As he left his cabin for the engine room, the blackout occurred. The C/E continued aft to the port side of the main vehicle deck and opened the engine room access door. He was met by significant smoke and some heat. He shouted out for the 3/E but heard no reply and was forced back onto the vehicle deck. He then tried to enter the ECR via the cargo control room (CCR) (Figure 4) and was again faced with smoke and heat. There was no sign of the 3/E. The C/E called the bridge from the steering gear space and informed the master of the fire in the engine room. He then operated the engine room fuel systems’ quick-closing valves (Figure 5) and fuel pump emergency stops from the control box adjacent to the ECR. Heat damage to door

6 Figure 4: Access to ECR, CCR from vehicle deck Ship schematic showing the access routes from the vehicle deck to the engine and cargo control rooms Access from ECR to engine room Engine control room Cargo control room

7 Figure 5: Control station for the engine room fuel systems’ quick-closing valves Opened control box containing the engine room fuel systems’ quick-closing valves In the purifier room, the 3/E realised that his nearest escape route would take him past the main engine at cylinder head level and decided to use the secondary escape route at the aft end of the engine room (Figure 6). He took a deep breath and then left the purifier room. On entering the dense black smoke, the 3/E ducked underneath the main engine exhaust gas trunking and ran aft past the two auxiliary generators. The escape route (Figure 7) led him to the first of three vertical ladders (Figure 8). When he climbed the first ladder, he struggled to get past the ladder platform guard rail safety chains and fell back down the ladder three times, losing his torch in the process, before making good his escape. The vessel’s four passengers were taken to their assembly point on the upper vehicle deck and an initial crew muster was attempted at the port lifeboat. Some of the crew involved in the fire-fighting effort reported in via hand-held radios. The engine room fire party, which had arrived on the upper vehicle deck and was running out fire hoses, reported the 3/E missing. A second muster took place on the upper vehicle deck, where the 2/E and motorman had already begun to shut the engine room fire flaps. At 2013, the 3/E escaped from the engine room through the funnel casing weathertight door onto the upper vehicle deck (Figure 9). He then collapsed onto the deck, gasping for breath and struggling to see. The 2/E saw the 3/E collapse and went to help him. The 2/E radioed the bridge and informed the master that the 3/E required medical assistance. At 2015, on instruction from the master, the OOW contacted Humber Coastguard and requested a helicopter.

8 Figure 6: Purifier room and engine room tween deck (shaded blue) with 3/E’s escape route Ship schematic showi ng third engineer's se condary escape route thro ugh the purifier roo m and engine room

9 Figure 7: Elevation showing the primary escape route from the purifier room to the main vehicle deck and the secondary escape route taken by the third engineer

(Green: primary escape route | Blue: secondary escape route) Ship schematic showing the elevations for the primary and secondary routes FWD AFT Mezzanine deck Upper vehicle deck

10 Figure 8: Escape route vertical ladder and safety chains Vertical ladder leading to a platform with guard rails and safety chains Safety chain Ladder Platform Guard rails Figure 9: Upper vehicle deck The upper vehicle deck on Finlandia Seaways with the funnel casing weathertight door opened Funnel casing weathertight door

11 With confirmation that the 3/E had been found, all crew and passengers accounted for, the fire flaps closed and the engine room fuel supply isolated, the C/E advised the master to release the engine room’s carbon dioxide (CO2) fixed fire-extinguishing system. The master then called the 2/E and ordered him to operate the CO2 system for the engine room. The 2/E and the second officer went to the CO2 remote release station (Figure 10) on the forecastle deck and, at about 2020, released the CO2 into the engine room. The 2/E then went down to the main deck and entered the CO2 bottle room (Figure 11). He found that the control lever for the engine room had operated correctly and that the CO2 pipework had frosted. He also noted that the main CO2 supply line pressure gauge was indicating about 20bar pressure, and a short time later he saw it reduce to zero. Figure 10: CO2 remote release stationCO2 remote release station on the forecastle deck Figure 11: CO2 bottle roomCO2 bottle room on the main deck At 2027, Humber Coastguard tasked rescue helicopter R912 based at Humberside to evacuate the injured 3/E. Two warships were also in the vicinity and were tasked to close on Finlandia Seaways and stand-by if required. While the medical team continued to provide medical support to the 3/E, other crew members were tasked to monitor engine room boundary temperatures. This was initially carried out by using their hands and, latterly, with fridge thermometers. Temperatures were measured and logged regularly and appeared to reduce over time. The master instructed the bosun to drop the port anchor; one shackle ran out before it jammed. The starboard anchor was then released and it, too, initially jammed. However, with assistance from the bosun it ran out and held the ship with four shackles.

12 At 2203, after diverting to refuel, R912 arrived at Finlandia Seaways. At 2247, the 3/E was recovered into the helicopter and taken to Norfolk & Norwich hospital, where he received medical care for smoke-related lung, kidney and eye injuries. At 2337, Finlandia Seaways’ master advised Humber Coastguard that the fire situation had been stabilised. Finlandia Seaways was initially towed to the Hawke anchorage on the River Humber. A further tow was arranged, with the vessel arriving in Immingham on 20 April 2018, where it was boarded by the Humberside Fire & Rescue Service to confirm that the engine room fire had been extinguished. 1.2.4 Engine room fire damage Fire damage was sustained primarily at the tween deck level in the engine room, with severe heat damage to the upper part of the main engine and engine room deckhead (Figure 12). Most of the heat damage was sustained around the main engine cylinder heads, and to auxiliary equipment on the periphery at the forward end of the engine room. Further aft, towards the two auxiliary generators, smoke damage was the major contributor. On the bottom plates, the main area of heat damage was at the port forward engine room access door and the adjoining stairwell. Impact damage to the access door caused by the components ejected from the main engine had enabled heat and smoke to enter the stairwell, which had then funnelled up the next two decks to the ECR. Figure 12: Fire damage to upper engine room (looking forward) Severe heat damage to the upper part of the main engine and engine room deckhead (looking forward) 1.3 CREW Finlandia Seaways was crewed by 19 Lithuanian nationals, one of whom had just joined the vessel in Zeebrugge as the relief C/E. Crew members normally worked 4 weeks on, 4 weeks off; or 6 weeks on, 6 weeks off. Port side mezzanine walkway Main engine Deck head Cylinder head A1

13 Although Finlandia Seaways had an unattended machinery space2 notation, the vessel’s engine room was normally manned at all times with the 2/E and 3/E undertaking 4 and 6-hour watches in rotation3. The C/E, electrical engineer and motorman were day workers and typically worked from 0800 to 1800. 1.4 ENVIRONMENTAL CONDITIONS At the time of the accident the wind was southerly, force 4-5, the sea was slight with fair weather and the visibility was good. 1.5 FINLANDIA SEAWAYS Finlandia Seaways was a Lithuanian registered roll-on roll-off cargo vessel. It was owned by DFDS Seaways AB-Lithuania (DFDS) and operated by DFDS A/S. It was built in 2000 and purchased, along with its sister vessel Botnia Seaways, by DFDS in 2009. When first acquired, Finlandia Seaways and Botnia Seaways were named Tor Finlandia and Tor Botnia. The vessels were renamed in 2012. Finlandia Seaways had provided a regular service on the Zeebrugge to Rosyth route since 2009. It had been classed by Lloyd’s Register (LR) since new and, at the time of the accident, its certificate of class was valid until 9 August 2020. DFDS A/S is a Danish shipping and logistics company. At the time of the accident it operated 28 vessels on passenger and freight services on 24 routes across northern Europe and within the Mediterranean Sea. DFDS Lithuania’s management and technical teams were based in Klaipeda, Lithuania. 1.6 MAIN ENGINE 1.6.1 Overview The vessel’s main engine was a 12,600kW MAN Diesel and Turbo SE (MAN) 12V 48/60 single-acting medium speed four-stroke diesel engine (Figure 13). It had two banks of six cylinders arranged in a “V” configuration with two turbochargers, one per bank. The ‘48/60’ referred to the bore of the cylinders and the stroke of the pistons in centimetres. The engine operated at a fixed speed of 500rpm, with load changes controlled by fuel input. It was connected via a flexible coupling to a single reduction gearbox, which drove a single CPP at an operating speed of 155rpm. The engine also drove a 400V, 1,500kW, 50Hz electrical alternator. At the time of the accident, the engine had a total of 110,662 running hours. The MAN 12V 48/60 engine was commonly used in both ship propulsion and shoreside power generation systems. 2 Unattended machinery spaces are required to have levels of automation that allow them to be periodically unattended. Typically, outside normal daily working hours. If there is a malfunction in any machinery, or any operating parameters are exceeded, an alarm will be sounded in the duty engineer’s cabin and other communal accommodation areas. It is then the duty engineer’s responsibility to investigate. 3 Engine room watch routine – 2/E: 0000-0400 and 1200-1800; 3/E: 0400-0800 and 1800-0000.

14 Figure 13: MAN 48/60 engine (engine shown 14V48/60B) Catalogue image of Finlandia Seaways's main engine model 1.6.2 Engine design and key components The engine design comprised a single-piece cast iron crankcase. The cylinder liners were separated from each other and were mounted in the engine frame, and the crankshaft was underslung. The crankcase was fitted with crankcase explosion relief valve doors, two per crank throw, i.e. one on each side. The crankcase relief valve operating pressure was 50mbar. The forged crankshaft had two counterweights per crankpin journal to compensate for the oscillating masses. The counterweights were held in place by anti-fatigue bolts. All the major bolting arrangements within the crankcase were hydraulically tensioned. The cylinders were numbered 1 to 6 from aft to forward and labelled “A” for the port side bank and “B” for the starboard side bank. Each cylinder bank had a camshaft, which operated four valves per cylinder head and the fuel injection pumps. A cylinder and its associated valves, piston and con rod was referred to as a ‘unit’, i.e unit A1, unit B1 etc. The piston connecting rods were cast in one piece. During manufacture, the connecting rod shank was cut below the small end eye, flanged and bolted together. This was to facilitate piston removal without the need to remove the connecting rod big end bearing. The shank’s flanged connection was the weakest part of the connecting rod assembly (Figure 14). The connecting rod small ends were stamped with the manufacturer’s unique identification code numbers (e.g. COC 1234) and LR identification numbers for survey purposes. A one-piece piston pin bearing bush (piston pin bush4) was 4 Connecting rod piston pin bearing bushes are also often referred to as small end bushes. Camshaft casing Crankcase doors Crankcase Oil tray Cylinder block Cylinder heads A1

15 installed into the eye of the small end (Figure 15). The surface finish of the inner face of the small end eye was prepared during manufacture by shot-peening5 to reduce the potential for fatigue fractures developing under high operational loads. The masses of key rotating and oscillating components within the engine were: piston – 337kg; piston pin – 102kg; connecting rod – 655kg; and crankshaft counterweight – 525kg. 5 Shot peening is a surface cold worked process of bombarding the material with small high velocity spherical metal particles (shot), which introduces residual compressive stress. Figure 14: Connecting rod assembly Drawing of a connecting rod assembly

16 Figure 15: B5 small end with piston pinSmall end with bearing bush and piston pin highlighted 1.6.3 Engine condition monitoring and protection systems The engine was equipped with the following crankcase monitoring systems:

● main bearing temperature monitoring system,

● splash oil monitoring system, and

● oil mist detector (OMD). The crankshaft bearing temperature monitoring system used temperature sensors that were fitted in the underslung crankshaft main bearings. The sensors transmitted electronic signals to the engine safety system. The splash oil monitoring system measured and compared the oil temperature dripping from the connecting rod bearings. Both temperature monitoring systems could trigger an alarm and automatically shut down the engine. The OMD was manufactured by Schaller Automation and was fitted on the starboard side of the engine crankcase (Figure 16). The OMD measured and compared the oil vapour concentrations present in each crankcase compartment and provided an audible and visual alarm when anomalies were identified. The OMD could also trigger an engine shutdown if high levels of mist developed in one or more compartments. Bearing bush Piston pin

17 Figure 16: Crankcase oil mist detectorOil Mist Detector fitted on the starboard side of the engine crankcase High oil mist concentrations are created in a crankcase by the presence of hot spots, which cause oil to evaporate and then form a mist when they start to condense. The ignition of the oil mist can lead to violent crankcase explosions, the consequence of which can be severe, including death and serious injury to crew, and extensive engine damage. 1.6.4 Post-fire inspection Finlandia Seaways’ main engine suffered substantial structural and component damage during its failure. The engine crankcase structure was ruptured on the port and starboard sides of units five and six, with crankcase explosion relief valve doors and large sections of the crankcase casting being propelled several metres across the floorplates (Figures 17 and 18). A5’s connecting rod (Figure 19), the lower half of its piston and its crankshaft counterweight had been ejected through the starboard side of the crankcase. The connecting rod had separated at its small end eye, shank flange and big end bolted connections. The big end bearing cap was found at the forward end of the engine room deck plates on the port side. B5’s connecting rod and its piston had also been ejected from the crankcase and were found on the starboard deck plates (Figure 20), the connecting rod assembly having separated at its bolted shank flange. The broken upper half of A5’s small end, and B5’s big end lower half, were found in the engine room starboard bilge.

18 Figure 17: Port side main enginePort side main engine showing broken crankcase units and the post-accident position of the crankcase door Broken crankcase units 5 and 6 Crankcase door Figure 18: Starboard crankcase damage at B5 Starboard side of crankcase units five and six showing damage and debris in the sump Broken crankcase Crankshaft big end bearing pin A5 and B5 Debris in sump

19 Figure 19: Section of A5 connecting rod (after removal from engine room) Connecting rod separated at its small end eye, shank flange and big end bolted connections Figure 20: B5 components on starboard side deck plates Connecting rod and piston on starboard side deck plates The upper part of A5’s piston and its piston pin6 remained in the upper section of the cylinder liner (Figure 21); the lower section of the cylinder liner had shattered and dropped into the sump. Although badly damaged at its lower end, B5 cylinder liner did not suffer the same level of damage as A5 (Figure 22). Impact damage had occurred to the crankshaft at number five crankpin journal and the crankpin webs. The camshaft covers for both banks had become detached and the A-bank camshaft had also been partially displaced. 6 Piston pins are also widely referred to as gudgeon pins. B5 piston B5 conrod small end

20 Figure 21: A5 cylinder viewed from engine sump Upper part of A5’s piston and its piston pin shown in situ in the upper section of the cylinder liner Piston Piston pin Upper section of liner Figure 22: B5 liner (looking up to inlet/exhaust valves from engine sump) Broken B5 cylinder liner, looking up to the cylinder head and four valves Broken liner Cylinder head and valves

21 1.6.5 Engine room alarms During the engine failure and ship blackout many alarms activated in the engine control room and on the bridge. A summary of key alarms is listed in Table 1 below. Table 1: Key alarms relating to the engine failure and engine room fire Ship’s time (UTC+2) Alarm Alarm State 2203:21 Main engine crankcase oil mist High 2203:22 Main bearing 6 high temperature Sensor (Fault) 2203:22 Main engine splash oil monitor Fail 2203:24 Lube oil pump stand-by Start 2203:25 Main engine slowdown 2203:28 Main bearing 5 high temperature Sensor (Fault) 2203:28 Main bearing 7 high temperature Sensor (Fault) 2203:32 Main bearing 1 high temperature Sensor (Fault) 2203:32 Main bearing 2 high temperature Sensor (Fault) 2203:32 Main bearing 3 high temperature Sensor (Fault) 2203:32 Main bearing 4 high temperature Sensor (Fault) 2203:35 Fire alarm 2203:43 Cylinder A5 exhaust gas out temp Low 1.6.6 Planned maintenance The main engine planned maintenance work on board Finlandia Seaways was conducted in accordance with the schedules contained in the vessel’s electronic planned maintenance system (PMS). The PMS schedules were based on MAN’s engine running hours maintenance schedule (Table 2). Table 2: MAN’s planned maintenance schedule Description Hours Check one big end bearing 6,000 Remove, clean/check one piston 6,000 Check all pistons 12,000 Remove one piston pin 12,000 Disassemble one piston 12,000 Replace all big bearing shells and small end bearing bushes 36,000

22 Day to day defect maintenance and minor planned maintenance work on the main engine were carried out by the vessel’s engineers. This included the replacement of fuel injectors and fuel pump servicing. In addition, the engineers conducted routine inspections of the engine’s crankcase, camshafts, bearings, liners, bolt security and lubricating oil supply lines. They also took regular fuel and lubricating oil samples and sent them ashore for analysis. Major maintenance work, such as piston removal, inspection and overhaul was usually carried out by shore contractors during maintenance layover periods at Rosyth. The attendance of the shore contractors was facilitated by the vessel’s technical superintendents. The maintenance requirements for the main engine connecting rods and pistons were set out in the same PMS schedules, under the headings No.1 Unit A; No.1 Unit B, etc. To assist with piston overhauls and enable swifter maintenance period turnarounds for Finlandia Seaways and Botnia Seaways, DFDS held six spare small ends. To help monitor small end running hours, the vessels’ C/Es recorded their running hours, based on their COC number, on a separate spreadsheet. 1.6.7 Outsourcing of main engine work Since taking operational control of Finlandia Seaways and Botnia Seaways in 2009, DFDS had outsourced most of the vessels’ main engine repair work to the ship repair company Diesel Service Group UAB (DSG). Initially, DSG was tasked to work alongside MAN service engineers but, over time it was contracted to carry out increasing amounts of the routine main engine work alone. MAN was used only to do complex tasks that required specialist tools and specific expertise. Recent work items undertaken on board Finlandia Seaways by MAN included replacing a camshaft in 2016, replacing the Vulcan shaft coupling and overhauling the engines’ oil and freshwater pumps. DFDS had held discussions with DSG to ascertain its capability to undertake specific tasks. In this respect, DFDS considered that DSG had the necessary tools and skills to undertake replacement of the connecting rods’ piston pin bushes. No written agreement for this was recorded. The last visit by DSG (concurrent with an engine inspection by ship’s engineers) had taken place on 3 April 2018, with engine running hours at 110,365. During the visit, DSG replaced the engine thrust bearing and carried out crankshaft deflection measurements. 1.7 DIESEL SERVICE GROUP UAB 1.7.1 Overview DSG was established in 1995 and was based in Klaipeda, Lithuania. It specialised in the repair of marine diesel engines and other engine room and deck equipment. The company had provided a service for a wide range of ships. Many of its technical staff had previously worked in the Western Ship repair yard, Klaipeda. 1.7.2 Certification DSG was certified by the Lithuanian Maritime Safety Administration (LMSA)7 to perform repairs, checks and testing of ships’ main and auxiliary mechanisms, devices and related systems. DSG’s certification was renewed on 22 January 2016, and it was audited by LMSA on 28 February 2017; no non-conformities were raised. 7 The LMSA was renamed the Lithuanian Transport Safety Administration (LTSA) on 1 December 2017.

23 DSG was also registered with the Russian Maritime Register of Shipping (RMRS), which had issued a ‘Certificate of Firm Conformity’ in November 2015. The certificate was valid for 5 years subject to annual confirmation and stated that DSG was certified for, among other things: Code 22014000 ‘Construction, conversion, modernization and repair of items of technical supervision products’, include:

1. Repair of ship’s engines up to 8500kW. [sic]

DSG was also accredited by the Italian classification society Registro Italiano Navale (RINA) with a ‘Workshop Approval Statement’, covering the design, diagnostics, maintenance and repair of ships’ systems and equipment. The certificate was valid until 18 March 2019. The approved workshop status was based on evidence provided, including qualification levels and experience, and a review of the workshop procedures and instructions. The statement did not specify any engine size or power limitation. DSG was not required to have a quality management system but was in the process of applying for ISO 9001:20008 and gaining third party certification. In this respect, a quality management system was under development, including a quality policy, project planning, division of responsibilities and quality objectives. To perform the maintenance on various engines, it had a range of marine engine manufacturers’ instructions, including MAN and S.E.M.T. Pielstick engines. DSG was not an accredited service provider for MAN. 1.8 CONNECTING ROD PISTON PIN BUSH RENEWAL 1.8.1 MAN guidance and procedure The wear limit between the piston pin and its bearing bush was 0.46mm. From experience, MAN expected the bearing pin bush to be replaced after 36,000 running hours. MAN’s piston pin bush removal procedure required the use of a milling machine to accurately cut a groove along the internal face of the bush. The cut needed to be a sufficient depth to leave less than 0.5mm of material on its outer edge without cutting into the surface of the connecting rod small end. The compression force acting on the bush due to its interference fit9 in the eye of the small end, caused the remaining thin layer of material to collapse inward, allowing the bush to be removed. New bushes were required to be cooled and shrunk using liquid nitrogen before they could be inserted in the small ends. The MAN 48/60 engine maintenance manual stated: Important! If required, the piston pin bush should be replaced in a Service Workshop, as the bush is to be cooled down before being installed. 8 ISO 9001:2000 specifies requirements for a quality management system where an organisation needs to demonstrate its ability to consistently provide a product that meets the customer and applicable regulatory requirements. 9 An interference fit (press fit and shrink fit) is a frictional connection for rotating components. Joint pressure in the friction surface between the two components is necessary for the torque to be effectively transmitted; this pressure is generated by the deformation of the two components, e.g. a shaft and its bearing.

24 MAN PrimeServ, the brand name for MAN’s service division, offered aftersales and through-life maintenance support for all MAN engines and systems. It provided a worldwide, round-the-clock service and had a global network of service workshops and test facilities. It also offered a variety of training courses for its customers’ marine engineers and other technical staff. MAN PrimeServ had two service stations worldwide that it considered equipped and able to conduct the bush renewal process; one of them was based at MAN’s headquarters in Hamburg, Germany. 1.8.2 Diesel Service Group guidance and procedure DSG carried out work on a range of vessels and engine types, including five operated by DFDS. On average, DSG had replaced two to three piston pin bushes per year for DFDS, of which about 24 were fitted to the MAN 12V 48/60 engines on board Finlandia Seaways and Botnia Seaways. The removal and installation of the MAN connecting rod piston pin bushes was carried out by DSG at its workshops in Klaipeda (Figure 23). DSG’s documented piston pin bush removing and installing procedure for the MAN 12V 48/60 engine (Figure 24) was similar to that followed by MAN. To remove the piston pin bush in accordance with the procedure, one side of the bush had to be bored away using a milling machine before pressing it out. The new bush needed to be cooled using liquid nitrogen to – 195ºC and measured before being inserted into the small end eye; a specialist tool was required to be used when aligning the bush. DSG employed one person to oversee the work and produce the technical repair document, and only one technician was permitted to use the milling machine. DSG was unable to confirm whether anyone else, including possible subcontractors, had undertaken this work. Figure 23: Diesel Service Group workshop and milling machine (inset) Diesel Service Group workshop showing position of the milling machine The milling machine

25 Figure 24: Diesel Service Group: small end piston pin bearing bush renewal procedure DSG’s documented piston pin bush removing and installing procedure for the MAN 12V 48/60 engin e

26 DSG carried out similar piston pin bush replacements on S.E.M.T Pielstick engines. The S.E.M.T Pielstick manufacturer’s instructions specified that the connecting rod small end bearing bush could be removed by making two diametrically opposed saw cuts in the bush. In February 2017, DSG carried out a 12,000-hour overhaul of Finlandia Seaways unit A5. The total engine running hours was 103,007 and the remarks recorded in the vessel’s PMS included:

●Piston with connecting rod small end:

○Dismantling, disassembling, cleaning, measurements, disassembling of piston, replacement of piston head (used), assembly of piston with new studs,

○Replacement of conrod small end, assembling of piston and conrod small end with new small end bush, installing with new piston rings in accordance with engine manual.

○Installing with new piston rings. The work was carried out in Rosyth, and the small end was replaced with one of DFDS’s spare small ends (COC 61584), which had been overhauled in November

1. A summary of A5 small end overhauls based on engine running hours only is

set out in Table 3. Table 3: A5 small end overhauls and inspections Date Engine operating hours Remarks 20/12/2002 15,885 Piston pin checked ok 03/03/2005 30,631 Pin and bush inspected 25/06/2007 46,279 Piston pin checked 02/02/2010 60,009 Piston pin bush checked, measured the clearance 03/10/2012 75,058 Fitted connecting rod with new piston pin bush 07/10/2014 87,637 Piston pin bush replaced 21/02/2017 103,007 Replacement of con rod small end with new small end bush [COC 61584] The MAIB and Lithuanian Transport Accident and Incident Investigation Division (TAIID) inspectors visited DSG on 16 July 2018. DSG was unable to identify when, or by whom, the A5 piston pin bush had been replaced. A subsequent inspection by LTSA on 4 September found that DSG was unable to provide evidence to confirm that liquid nitrogen had been used to shrink the piston pin bush during the overhaul of the small ends. As DSG was not accredited by MAN, it did not have access to MAN’s piston pin bush removal and installation tools and procedures.

27 1.9 ENGINE MANUFACTURER’S POST-FAILURE ENGINE INSPECTION AND GUIDANCE 1.9.1 Engine inspection and failure mode analysis MAN attended Finlandia Seaways after the accident while it was berthed alongside in Immingham. The engine debris had been left untouched to allow their site examination. Independently, but like the MAIB’s initial examination, the focus of interest became the A5 connecting rod. The upper half of the broken small end was found in the debris field on the starboard side of the engine, and the severely damaged piston pin bush was found nearby. Inspection of the failed surfaces of the small end identified a potential fracture face on one side of the small end eye (Figures 25a, 25b and 25c). The remnants of the failed A5 small end, the B5 small end and various other components, including the OMD, were taken ashore by MAN for further analysis. MAN examined the failed A5 connecting rod small end on 22 June 2018 at its laboratory in Augsburg, Germany. MAIB inspectors and a DFDS representative were in attendance during the examination process. Work carried out included:

● microscopic examination of the small end fracture faces,

● metallurgical composition testing, and

● notch toughness testing10. The MAN investigation of A5 connecting rod small end (COC 61584), identified that:

● The small end was originally fitted to Botnia Seaways at build in 1999 and was the latest design version.

● It had never been serviced by MAN.

● The fracture might have occurred over a period of weeks or months.

● Due to the damage on the fracture surface, it was not possible to identify the locus of the fracture.

● The chemical composition did not fully correspond with the material specification. The ductile properties of the material were slightly below the required specification.

● A microsection showed a quenched and tempered microstructure but with an aggregation of non-metallic inclusions. The examination of B5 (COC 113544) found no abnormalities and MAN’s investigation established that it was delivered in 2007. In addition, the investigation report (Annex A) concluded that:

● The OMD alarm was the first indicator of a problem, quickly followed by physical destruction of internal engine components. 10 A standardized high strain-rate test, which determines the amount of energy absorbed by a material during fracture.

28

● The OMD did not have the ability to record alarms or faults to assist the investigation.

● The A5 small end was probably the first major component to exit the crankcase. For engines of the type 48/60, MAN also identified that over half of its engines in service had accumulated over 100,000 running hours, with two operating at 180,000 running hours without similar small end failures. Figure 25a: A5 small end upper half Fracture surface on the upper half of the A5 small end Figure 25b: A5 small end fatigue fracture Photos and diagram to illustrate fatigue fracture on the A5 small end Bearing bush A5 Fatigue facture Crack initial side Fracture surface Base images courtesy of MAN Energy Solutions

29 Figure 25c: A5 looking forward and fatigue fractureDrawing of the A5 connecting rod, looking forward, showing position of the fatigue fracture Fatigue fracture Connecting rod A5 Base image courtesy of MAN Energy Solutions

30 At MAN’s request, the remaining 10 connecting rod small ends from Finlandia Seaways were sent to MAN for inspection. After the piston pin bushes were removed, seven were found to have cuts, or notches, at the edges and on the internal faces of the small end eyes (Figure 26 and Table 4). The cuts were symptomatic of the use of a disc cutter or angle grinder. Small ends: A3; B1; B2 and B4 were subsequently identified as having the same respective overhaul and installation dates. It is probable that the overhauls were carried out at the DSG workshop in Klaipeda before installation while the vessel was alongside in Rosyth. Small end (COC No) Date installed in engine Date overhauled by DSG New piston pin bush fitted Cuts or notches found in small end eye Engine running hours at installation Small end running hours at time of engine failure11 A1 (62085) 29/12/2016 unknown Yes Yes 102,091 88,765 A2 (61868) 06/06/2017 unknown Yes Yes 105,058 89,233 A3 (62009) 15/08/2017 15/08/2017 No Yes 106,399 91,344 A4 (62110) 06/06/2017 unknown Yes Yes 105,058 88,946 A5 (61584) 21/02/2017 11/2016 Yes Unknown 103,007 90,467 A6 (113538) 18/04/2017 unknown Yes No 104,096 88,677 B1 (61993) 02/01/2017 02/01/2017 No Yes 102,140 89,455 B2 (61944) 18/04/2017 18/04/2017 No No 104,096 88,466 B3 (61948) 07/11/2016 07/11/2016 No No 101,146 91,237 B4 (61857) 21/12/2016 21/12/2016 No Yes 102,091 90,389 B5 (113544) 29/12/2016 unknown Yes No 102,091 87,455 B6 (61996) 21/02/2017 unknown Yes Yes 103,007 89,048 Table 4: Finlandia Seaways’ main engine connecting rods, including small end damage 11 Engine running hours: 110,662

31 Figure 26: MAN – Finlandia Seaways’ small ends investigation Pa ge f rom MAN's investig ation show ing ph o tographs of the seven small ends wit h cuts, or not ches, at the edge s and on the internal faces of the smal l end eyes 1.9.2 Botnia Seaways Following the initial damage assessments and MAN findings, on 27 July 2018 MAIB and TAIID wrote jointly to MAN and LR raising concerns about the risk of similar catastrophic engine failures on board Botnia Seaways and other vessels operating MAN engines overhauled by DSG. In the letters, the engine manufacturer and

32 classification society were recommended to provide technical advice on actions to be taken by vessel operators at risk. DFDS subsequently withdrew Botnia Seaways from service. Between 6 and 25 August 2018, while out of service in Marseille, France, all of Botnia Seaways’ main engine connecting rods were replaced. During the initial inspection of the small ends, the attending LR surveyor identified heat marks on the inner edge of one of the small ends (Figure 27). MAN specialists examined the marks and concluded that they were not caused during engine operation. MAN’s assessment was that the localised heat marks had been caused by the application of intensive heat from something like a gas welding torch. MAN suggested that Figure 27: Botnia Seaways small end heat damage (circled)Heat marks on the inner edge of one of the small ends from Botnia Seaways' main engine connecting rods

33 this had been to assist with the piston pin bush removal/installation process. MAN further concluded that the intense local heating would have produced an annealing12 effect of the metal, resulting in structural changes affecting the material strength. MAN’s subsequent investigation of the 12 small ends returned for inspection identified:

● Notches/cuts on three small ends.

● Localised heat marks on three small ends.

● Damaged shank bolts on three connecting rods.

● Various levels of other spurious damage and wear. 1.9.3 Engine design modifications and updates MAN regularly provided targeted customer service information to advise customers on maintenance updates for specific engine types. Since February 2000, MAN had distributed 94 customer information documents for the 48/60 engine, including all the variants. In 2015, MAN circulated customer information notice No.344/2015 (PCI 344/2015) (Annex B), which advised that the connecting rod be renewed after 100,000 running hours. This was primarily based on increased ovality identified in the big end bearing housings. On receipt of the notice, DFDS checked the ovality of a selection of its connecting rod big ends during routine maintenance and, finding no ovality, kept them in service. Although DFDS had received them, the MAN customer information notices issued after Finlandia Seaways’ build were not available on board. 1.10 METALLURGICAL ANALYSIS As the initiation point of the A5 small end fracture was not identified during the examinations conducted by MAN, the MAIB commissioned The Test House (Cambridge) Ltd (TTH) to undertake further metallurgical analysis work. The remains of Finlandia Seaways’ A5 small end, and B5 and B6 small ends were delivered to TTH in October 2018. TTH’s initial inspection of A5 identified that: The fracture surface showed a flat/featureless surface, clear evidence of beach marks, post fracture mechanical damage on the fracture surface and bearing contact surface and a secondary crack in the oilway groove perpendicular to the fatigue fracture. The conclusions contained in TTH’s report (Annex C) included: The engine had failed in a catastrophic manner as a result of the A5 connecting rod small end failing due to fatigue. The fracture had initiated at the bearing contact surface and propagated through the body. 12 Annealing is a heat treatment that alters the physical and sometimes chemical properties of a material to increase its ductility and reduce its hardness, making it more workable.

34 Due to post fracture mechanical damage it was difficult to identify the point of initiation. The intact A6 small end had gouge marks on the edges of the bearing bush contact surface. The TTH report stated that: Although the intact small end [A6] showed no evidence of consequential features due to the gouge marks, such practises should be avoided at all costs as they can conclude to points of stress raisers. 1.11 EXTERNAL OVERSIGHT Classification societies conduct periodic reviews of critical machinery components to ensure they are in an appropriate condition; normally 20% of Class items are inspected per year over the mandated 5-year survey cycle. In accordance with LR rules and regulations, Finlandia Seaways was operated on a Continuous Survey Machinery (CSM) cycle. Consequently, its C/Es were authorised to carry out examinations of selected machinery while the ship was at sea or in port when LR was not represented. The process was explained by LR in its document Examination of Surveyable Machinery Items by Chief Engineers, effective 1 March 2015 (Annex D). The C/Es’ records of examination were audited annually by an LR surveyor and confirmatory surveys carried out. LR provided a template statement (Annex D) for completion by the C/E with a requirement that a copy be provided to the surveyor and a second copy remain on board. Applicable machinery items that could be examined by a C/E included:

● main engine connecting rods,

● piston pins, and

● piston pin bushes. The connecting rod was a certified item and had an LR stamp on it, but its piston pin bush was considered a consumable item. The replacement of a small end piston pin bush was considered routine maintenance on the part of the ship owner, and LR did not expect to be called in by an owner to witness routine maintenance between its mandated surveys. LR did expect to be notified if a certified component, such as the connecting rod, had been deemed beyond economical repair and was replaced, had a defect that was repaired, or if the component was modified. Cylinder A5 had previously been overhauled on 7 October 2014; the C/E’s statement of examination was dated 29 December 2014. The examination was then credited under the CSM cycle by an LR surveyor on 31 January 2015 (Table 5).

35 Table 5: Lloyd’s Register Component Survey List Conrod Description Date last credited by LR Type of credit13 A1 Cylinder, cover, bearings, piston, conrod, valves and gears January 2015 Complete A2 Cylinder, cover, bearings, piston, conrod, valves and gears January 2015 Complete A3 Cylinder, cover, bearings, piston, conrod, valves and gears January 2015 Complete A4 Cylinder, cover, bearings, piston, conrod, valves and gears January 2015 Complete A5 Cylinder, cover, bearings, piston, con-rod, valves and gears January 2015 Complete A6 Cylinder, cover, bearings, piston, conrod, valves and gears September 2017 Confirmatory B1 Cylinder, cover, bearings, piston, conrod, valves and gears January 2015 Complete B2 Cylinder, cover, bearings, piston, conrod, valves and gears January 2015 Complete B3 Cylinder, cover, bearings, piston, conrod, valves and gears January 2015 Complete B4 Cylinder, cover, bearings, piston, conrod, valves and gears September 2017 Confirmatory B5 Cylinder, cover, bearings, piston, conrod, valves and gears September 2017 Confirmatory B6 Cylinder, cover, bearings, piston, conrod, valves and gears September 2017 Confirmatory A subsequent C/E’s statement was completed on 4 September 2017 and included various engine parts examined since February 2017. These included cylinder B6 on 21 February 2017: No.6 STBD bank cylinder, cover, piston, cylinder liner, connecting rod & bearings, valves & gears were inspected during overhaul of cyl No6B. All parts cleaned, measured and found in good sound condition. Replaced piston rings, all o-rings and gaskets. No6B cyl was tested in work after assembling. No remarks. [sic] There was no reference to the replacement of the small end bush that had also occurred. 13 Items credited with ‘Complete’ indicate that the surveyor conducted a full survey of the item. ‘Confirmatory’ indicates that the surveyor was provided documentary evidence from the chief engineer regarding the condition of the equipment.

36 DFDS only had the C/E’s statement of examinations for cylinders A1, A2 and A5 that were overhauled during 2014, and LR interim certificates for July 2012 and December 2017. The local Klaipeda LR surveyor did not routinely retain his copies of the survey results as there was no obligation to do so. During their examinations, LR surveyors were advised to take pictures of the general condition of compartments and also of defects/observations. For specific engine components, photographs would be deemed necessary only if defect observations were made. As no defect observations had been made, no photographic evidence of surveys was available. 1.12 EMERGENCY RESPONSE 1.12.1 Carbon dioxide fixed fire-extinguishing system Finlandia Seaways’ CO2 fixed fire-extinguishing system comprised 221 x 53kg gas bottles, with the capacity to protect the engine room, upper cargo hold, lower cargo hold and paint store. Each protected space required a specified number of bottles to ensure that the correct percentage of CO2 was injected to smother a fire. Thirty-one bottles were required for the engine room, and they were released via the remote release station on the forecastle deck. The system could also be operated from the CO2 bottle room. Confirmation that CO2 had been released into the protected space was achieved by entering the CO2 bottle room and checking the appropriate control valve had activated and the manifold gas pressure gauge had dropped to zero. Frosting of the bottles and manifold pipes provided a further visual indication that gas had been released. Information to enable the crew to easily identify which spaces individual CO2 bottles protected was not provided in the bottle room, and the gas bottle operating valves did not provide a visible means of identifying whether they were open or closed (Figure 28). Following the activation of the CO2 system, it was unclear how many bottles had been released. Consequently, a specialist service engineer was required to attend the vessel and confirm that all 31 CO2 bottles had been released into the engine room. Marine Guidance Note (MGN) 389 (M+F) Operating Instructions and Signage for Fixed Gas Fire-Extinguishing Systems was published in 2009 by the Maritime and Coastguard Agency (MCA) following a series of fire investigations conducted by the MAIB, which identified issues relating to the quality and effectiveness of operating instructions and signage. The MGN quotes Chapter 5, Section 2.1.3.3 of the International Maritime Organization’s (IMO) Fire Safety Systems Code, which states: The means of control of any fixed gas fire-extinguishing system shall be readily accessible, simple to operate and shall be grouped together in as few locations as possible at positions not likely to be cut off by a fire in a protected space. At each location there shall be clear instructions relating to the operation of the system having regard to the safety of personnel. MGN 389 (M+F) recommends the use of colour coded controls for different discharge zones, and emphasises the importance of ensuring gas has been correctly released for a protected zone immediately after activation. For systems such as that fitted on board Finlandia Seaways, the MCA advised crew to look for frosting around the bottom of the discharged gas bottles.

37 Figure 28: Carbon dioxide bottle operating valveGas bottle operating valve showing no visible means of identifying whether they were open or closed 1.12.2 Means of escape Under the International Convention for the Safety of Life at Sea (SOLAS) Amendments (Chapter II-2, Part D, Regulation 13 – Means of escape), ships of more than 1000gt shall have two means of escape from each machinery space of category A14. Each escape shall be widely separated from each other and lead to the open deck. One of the escapes shall have a continuous fire shelter from the lower part of the space to a safe position outside. Finlandia Seaways met the means of escape requirements, with the continuous fire protection route being the primary escape route via the steel self-closing door at the port-forward bottom engine deck, up the enclosed stairwell via the control room and to the main deck (Figure 29). 14 Spaces which contain internal combustion machinery used for main propulsion.

38 Figure 29: Distribution of EEBDs in or near to the engine room and escape routes Plan diag ram shows distribu tion of EEBD s in or near to engi ne room and escape routes

39 The secondary means of escape, as used by the 3/E, via the funnel casing aft of the generators, was not enclosed. This route could be accessed from the bottom engine room deck, via a steel ladder near the propeller shaft on the port side, and up through a hinged steel hatch grating at generator deck level. 1.12.3 Emergency escape breathing devices Emergency escape breathing devices (EEBDs) are part of ships’ safety equipment, and their primary purpose is to enable crew to escape from a smoke-filled compartment. They should provide a minimum air supply of 10 minutes and are mandatory under SOLAS for ships where a safety equipment certificate applies. SOLAS 2000 Amendments (Chapter II-2, Part D, Regulation 13) introduced the requirement for emergency escape breathing devices (EEBDs). Section 4.3 Emergency escape breathing devices states: 4.3.1 On all ships, within the machinery spaces, emergency escape breathing devices shall be situated ready for use at easily visible places, which can be reached quickly and easily at any time in the event of fire. The location of emergency escape breathing devices shall take into account the layout of the machinery space and the number of persons normally working in the spaces. IMO MSC/Circ.849 (8 June 1998) – Guidelines for the performance, location, use and care of emergency escape breathing devices in section 4.6 states: Unless personnel are individually carrying EEBDs, consideration should be given for placing such devices along the escape routes within the machinery spaces or at the foot of each escape ladder within the space. IMO MSC/Circ.1081, Unified Interpretation of the Revised SOLAS Chapter II-2, 13 June 2003, relating to EEBDs, includes: Regulation 13.4.3 Emergency escape breathing devices (EEBD)

1. This interpretation applies to machinery spaces where crew are normally

employed or may be present on a routine basis.

1. In machinery spaces for category A containing internal combustion

machinery used for main propulsion, EEBDs should be positioned as follows: .1 one (1) EEBD in the engine control room, if located within the machinery space; .2 one (1) EEBD in workshop areas. If there is, however, a direct access to an escape way from the workshop, an EEBD is not required; and .3 one (1) EEBD on each deck or platform level near the escape ladder constituting the second means of escape from the machinery space (the other means being an enclosed escape trunk or watertight door at the lower level of the space). Alternatively, a different number or location may be determined by the Administration taking into consideration the layout and dimensions or the normal manning of the space.

40 The unified interpretation applies to all ships constructed on or after 1 July 2003. Finlandia Seaways had three EEBDs located in the engine room machinery spaces: one in the ECR, one in the workshop, and one on the engine room bottom deck at the aft end adjacent to the propeller shaft (Figure 29). A further six EEBDs were provided in the forward accommodation block (one on each deck) and one on the main deck forward equipment room. The LR Record of Approved Cargo Ship Safety Equipment (dated 31/12/12) identified 10 EEBDs, with two located in the engine room and eight in the accommodation. The engine room escape routes and locations of the EEBDs were clearly marked on Finlandia Seaways’ fire control and safety plan, which had been approved by LR on behalf of the vessel’s Administration. The nearest EEBD to the 3/E when he began his escape from the engine room tween deck, was one deck below, at the aft end adjacent to the propeller shaft. 1.12.4 Voyage data recorder recovery Finlandia Seaways was equipped with a Kelvin Hughes MDP-A5 voyage data recorder (VDR). It was fitted in 2010 and was supported by an uninterruptible power supply (UPS) battery to enable it to maintain recording capability when ship’s power was lost. The UPS batteries were last replaced in January 2015. Shortly after the fire, the company and master were requested to save the data recorded on the VDR. The MAIB recovered the VDR data as part of its investigation and discovered nothing had been recorded after the blackout. On 11 September 2017 a Kelvin Hughes representative had conducted an annual performance test (APT). The APT report identified that the UPS battery had an expiry date of 1 March 2019 and, during testing, sustained power to the VDR for more than 1 hour 55 minutes after the main power supply was removed. The report was returned to Kelvin Hughes, vetted and approved. 1.13 PREVIOUS/SIMILAR ACCIDENTS 1.13.1 Thetis D On 26 October 2015, the container ship Thetis D suffered a major main engine failure and subsequent engine room fire while on passage. Its MAN 58/64 CD engine suffered serious damage to its number 7 piston and connecting rod big end bearing. It also suffered structural damage to the crankcase adjacent to the bottom end bearing. The investigation report15 published by the German government’s marine transport accident investigation body, Bundesstelle für Seeunfalluntersuchung (BSU), identified that the bottom end bearing had been inspected by the vessel’s C/E 7 months earlier. The findings of the examination of the failed components by MAN and an independent engineer identified that after the inspection of the bottom end bearing, the conrod bolts had not been tightened correctly. The loss of bolt tension eventually resulted in bottom end bearing damage, piston seizure and ejection of engine components through the crankcase. 15 BSU report 431/15 – Serious engine damage followed by fire on board the cargo vessel THETIS D on 26 October 2015 in the Kiel Bight.

41 BSU recommended that Thetis D’s owners take the necessary steps to ensure qualified engine room personnel prepare, carry out and record maintenance and servicing works on key components properly. They also recommended that the classification society, at its next International Safety Management audit, focus on the procedure for maintaining and servicing key components to ensure it exists and is implemented. 1.13.2 Power station, Spain A Spanish power station using MAN 48/60 engines suffered a catastrophic engine failure during operation. MAN’s investigation identified the failure of a connecting rod small end to be the cause. Its investigation found that the technicians operating the engine had carried out a small end overhaul on the failed connecting rod. During the overhaul the technicians used a disc cutter to remove the piston pin bushes from the small ends. The disc cutter had also cut into the eye of the small end. MAN concluded that the disc cutter damage had introduced stress raisers, which eventually led to fatigue fracture and small end failure.

42 SECTION 2 - ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 ACCIDENT OVERVIEW Finlandia Seaways’ main engine suffered a catastrophic failure that resulted in the violent ejection of heavy internal components from the crankcase into the engine room. The subsequent release of hot oil vapours caused a fire to break out and the engine room to become engulfed in thick black smoke. The 3/E was in the engine room at the time and he was extremely fortunate to have escaped and survived. In this section of the report, the engine failure mode will be identified and the factors that contributed to it will be analysed. The emergency response by the vessel’s crew, and the levels of emergency preparedness for the casualty event, will also be discussed. 2.3 ENGINE FAILURE 2.3.1 Sequence of events The initial indications that there was a problem with Finlandia Seaways’ main engine were loud metallic knocking noises heard by the on watch 3/E in the purifier room, and powerful vibrations felt throughout the ship by the OOW, master, C/E and the rest of the crew. The first engine alarm to activate was the main engine crankcase oil mist – high alarm, which was immediately followed by number six main bearing high temperature and splash oil monitor failure alarms. Within 2 seconds of the first alarm the engine’s standby lubricating oil pump started, and the automatic engine slow down sequence began. Ten seconds later, the engine room fire alarm was activated. OMDs are designed to identify and alert engineers to small changes in crankcase oil mist levels and therefore protect an engine against the potential consequences of hot spots created over a period of time due to events such as bearing wear or piston ring blow past. In cases where oil mist is ignited, the sudden increase in crankcase pressure should be released via the crankcase explosion relief valve doors. Although the initial alarm indicated that an oil mist had been generated in the crankcase, there was no evidence to suggest that a crankcase explosion had occurred prior to the catastrophic engine failure. The witness accounts and machinery alarm sequence clearly indicate that the accident was the consequence of a sudden major engine component failure rather than as a result of gradual wear and/or lubrication starvation. The loud mechanical knocking noises, automatic starting of the standby lubricating oil pump, the sequential failure of internal alarm sensors and the fact that the crankcase explosion relief valve doors did not open prior to the ejection of rotating components into the engine room, support this conclusion.

43 2.3.2 Identification of the component failure The structural damage to the engine crankcase was restricted to units five and six, with internal components from A5 and B5 cylinders being released into the engine room. It was therefore apparent that the initial component failure must have occurred in either A5 or B5 cylinder. The accident site damage inspection conducted by MAIB and MAN engineers was focused on identifying the likely locus of the failure and eliminating the components that had suffered secondary damage due to contact with other moving parts or sudden loss of lubrication. Both A5 and B5 connecting rods had failed and been thrown into the engine room. B5’s piston had also been thrown out of the crankcase, but the upper half of A5 piston was found to be seized in its damaged cylinder liner, with its piston pin still in place. It was therefore evident that the A5 connecting rod had failed at its small end eye. Either the piston had seized within the liner to such an extent that the connecting rod small end eye had sheered under tension, or the small end itself had suffered a metallurgical failure. The types of friction marks and liner damage typically associated with piston seizures due to misalignment, lack of liner lubrication or ring failure were not found when the recovered pieces of the broken A5 cylinder liner and A5 piston skirt were inspected. Furthermore, had a piston seizure caused the catastrophic failure, the connecting rod’s flanged shank connection should have failed first as it was the weakest part of the connecting rod assembly. Because of this, the failure of the A5 connecting rod small end was identified as the most likely cause of the catastrophic engine failure. When the small end failed, the A5 connecting rod would have begun flailing around inside the crankcase. At 500rpm, damage within A5 cylinder unit developed rapidly, tearing away alarm cabling and causing B5 cylinder components to also fail. Within seconds, the heavy rotating components were released and, with great energy, thrown through the crankcase into the engine room. 2.3.3 Failure mode The A5 connecting rod’s small end had completed just over 90,000 running hours. This was 10,000 hours fewer than the recommended limit set by MAN for the connecting rod, 20,000 less than the engine operating hours and up to 80,000 fewer than some of those accumulated on some similar MAN engines. This suggests that the failure had most likely been the result of either damage caused during the overhaul of the small end, an assembly error during the overhaul process or during its installation, or a metallurgical defect introduced during the initial manufacturing process. MAN’s metallurgical analysis of A5’s small end identified the presence of non-metallic inclusions within the material’s microstructure, and that the ductility of the metal was marginally below that required in its material specification. Although both factors would have an adverse effect on the material strength and durability of the small ends, MAN determined that the deviations were minor and within the acceptable safety margin limits set out in published guidelines. Both the MAN and TTH examination reports dismissed a manufacturing defect as being a likely cause of the failure.

44 When the upper half of the A5 connecting rod small end was found, a brief inspection of its failure faces was sufficient to identify the development of a fatigue fracture on one side of the small end eye (Figure 25a and 25b). Following closer examination of the failure faces, both MAN and TTH concluded that the mode of failure was a fatigue fracture. 2.4 FRACTURE MECHANISM 2.4.1 General Fatigue fracture is one of the primary failure modes in structural metal components and is often referred to as metal fatigue. Fatigue fracture results from cyclic stresses that are below the ultimate tensile strength, or even the yield strength of the material (Figure 30). As the original design strengths are not exceeded and the only warning sign of an impending fracture is a crack, which is often hard to see, fatigue damage is particularly dangerous. 2.4.2 Fatigue fracture Fatigue failure occurs in three stages: crack initiation, crack growth (slow propagation through the metal) and fracture (rapid failure of the remaining intact metal) (Figure 31). Fatigue cracks almost always start at the surface of a material. This is because:

● The most highly stressed fibres are usually located at the surface (bending fatigue). And

● The intergranular flaws, which precipitate tension failure are more frequently found at the surface. Crack initiation points, crack growth and propagation rates can be established by analysing the macroscopic beach marks that are usually created on a failure face prior to final fracture. TTH analysed the beach marks found on the small end’s failure face, and determined that the fracture had initiated at the bearing bush contact surface and had propagated outward. It also identified a secondary crack in the oilway groove perpendicular to the fatigue fracture. Since fatigue cracks generally initiate at a surface, the surface condition of the component being loaded will influence its fatigue life. Surface roughness is important because it is directly related to the level and number of stress concentrations on the surface. The higher the stress concentration the more likely a crack is to develop. Smooth surfaces increase the time before crack development. Notches, scratches and indentations can act as stress raisers and therefore decrease the time before crack development. The bush mating surface of the internal face of the small end eyes had been shot-peened to increase crack resistance. However, the introduction of notches, caused during the bush removal process, introduced stress raisers into the small end, increased the likelihood of fatigue crack initiation and therefore fatigue failure. 2.4.3 Notch effect and toughness Toughness is the ability of a metal to deform plastically and to absorb energy before fracture. There are several variables that have a major influence on the toughness of a material. These variables include: strain rate, temperature and notch effect.

rack growth and f racture

46 The third variable, notch effect, involves the distribution of stress. A material might display good toughness when the applied stress is uniaxial. However, when a multiaxial stress state is produced due to the presence of a notch, the material might not withstand the simultaneous elastic and plastic deformation in the various directions. A notch causes a disruption to the stress flow and introduces localised stress concentrations. The change of section concentrates stress most strongly where the transition is sharp (Figure 32). The introduction of a notch, or cut, in a material will affect the fracture development. The shape of the notch will also have an effect on how the stress propagates through the material. Fatigue loading is cumulative, and the fatigue strength decreases more rapidly over the cycling lifetime for notched samples. Figure 32: Plastic protractor under polarising light showing stress concentration Stress concentration at an inside corner of a plastic protractor 2.4.4 Localised heating Heat marks were found on several of the small ends removed from Botnia Seaways, all of which were condemned and had to be replaced. It was apparent that the heat marks were the result of localised heating applied by a gas welding torch or similar device. Temperature can also influence the fatigue life of materials. As the material’s temperature increases, so the fatigue life often decreases. A cyclic temperature change encourages failure by thermal fatigue; when the material heats in a non-uniform way, some parts of the structure expand more than others. This

47 non-uniform expansion introduces a stress within the material, and when the structure cools and contracts, stresses of the opposite nature are then imposed. As a consequence of the thermally-induced stresses and strains, fatigue fracture can eventually occur. Large masses of steel of heavy section will cool more slowly than small components of thin section. If not uniformly heated and quenched to produce a harder material overall, the effect of gradual cooling will introduce stresses between the hard and brittle surface and the softer layers below. These stresses can eventually initiate crack development. Uneven heating and subsequent cooling can also exacerbate crack formation. 2.4.5 Inspection and analysis MAN’s inspection and testing of the upper half of Finlandia Seaways’ failed A5 connecting rod small end was unable to identify the fatigue crack initiator due to heavy secondary damage of the part. However, MAN’s initial hypothesis on the failure, based on the visual inspection of the crack surface and the age of the small end, was that the relative movement between the bearing bush and the small end during the small end’s estimated 90,000 running hours exceeded the allowable value. This could have been caused by:

● plastic deformation of the small end (ovality), or

● heavily worn bearing bush, or

● maintenance irregularities during the bearing bush replacement. TTH’s examination assessed that the fatigue crack had started in the region of the oil groove, with a secondary crack within the groove itself, although post-failure damage prevented an accurate identification of the start point. The later inspection of Finlandia Seaways’ remaining small ends, and those of Botnia Seaways, confirmed that poor overhaul practices, in respect of notch and heat damage, were the most probable cause of the A5 small end failure. 2.5 MAINTENANCE MANAGEMENT 2.5.1 Overview The routine maintenance and major component overhauls for Finlandia Seaways’ and Botnia Seaways’ MAN 12V 48/60 main engines was undertaken in accordance with the schedules contained in the vessels’ electronic PMS. Most of the maintenance schedules were based on the accumulated running hours set by MAN. Minor routine maintenance was carried out by ship’s crew, but major overhauls were contracted out by DFDS. This included the inspection and, where necessary, replacement of the connecting rod small end piston pin bushes. To facilitate quick maintenance turnaround times, DFDS held a stock of six spare overhauled small ends for use on either vessel. Finlandia Seaways’ main engine maintenance and overhauls had been carried out in accordance with the schedules set out in the PMS, and were recorded electronically by its engineers. However, the post-accident examinations and laboratory tests carried out during this investigation, clearly showed that MAN’s guidance and

48 recommended procedures for overhauling the connecting rod small ends had not been followed, and the small ends had been damaged during the piston pin bush removal and fitting process. The extent of the damage was clearly visible to the naked eye but was either not identified during the engine/component overhauls or was ignored. 2.5.2 Connecting rod small end overhaul process MAN had a detailed set of procedures for the removal and installation of the 48/60 engine connecting rod piston pin bush, and its maintenance manual warned that they should be replaced in a Service Workshop because they had to be cooled before being installed. MAN’s aftersales service division, PrimeServ, offered around- the-clock support from its global network of service centres, which included two workshops specifically equipped to remove and fit the piston pin bushes. DFDS’s technical management was aware of MAN’s requirements and contracted DSG to carry out most of its engine overhauls. DSG was not a MAN-accredited service agent and did not have access to all of MAN’s maintenance guides and overhaul procedures. However, DSG’s documented piston pin bush removal and installation procedures were similar to MAN’s and included the use of a milling machine to cut a groove in the bush during the removal process, as well as liquid nitrogen and use of a special alignment tool during the fitting process. DSG had also designated a competent fitter to undertake the specific task of removing the piston pin bushes at its workshops in Klaipeda. Despite this, seven of the intact 11 small ends removed from Finlandia Seaways’ engine were found to have notch damage on the internal surfaces of their small end eyes. Three of Botnia Seaways’ small ends had similar damage and three had evidence of localised heat damage. Although the technical investigations did not identify when the damage had been done, it was clearly apparent that it had been caused by using a disc cutter and a gas torch during the overhauls undertaken by DSG’s service engineers. DSG’s failure to follow its own procedures might have been influenced by the fact that its engineers had experience of working on smaller engines where mechanical saws and presses were often used to remove and fit piston pin bushes. It is also possible that some of the overhauls were carried out on board the ships or in local workshops that did not have the specialist tooling required or a supply of liquid nitrogen. Regardless, the extent of the damage caused during the piston pin bush overhauls was visually apparent and, therefore, it was obvious that DSG’s service engineers did not appreciate the potential consequences of the stress raisers they had introduced. 2.5.3 Rotation of connecting rod small ends The two major components that comprised the split shank connecting rod – the big end and small end – each had their own COC identification numbers. As the two parts were not specifically matched to each other, it enabled DFDS to replace the small ends with spares, and therefore reduce the time taken to conduct connecting rod overhauls. Consequently, there was a disparity between engine running hours (110,662 hours at failure) and individual small end running hours, of roughly 20,000 hours, with the A5 small end having accumulated just over 90,000 hours in total. This was further complicated because the six spare small ends had been rotated between Finlandia Seaways and Botnia Seaways.

49 The rotation of the small ends could be considered efficient in terms of keeping vessel downtime to a minimum, but it required an operating hours’ log for both the connecting rod big end and the small end. Although the onboard records were maintained, the DSG maintenance documents gave the COC numbers for only some of the small ends, having misunderstood which identification number to use. This introduced the possibility for errors to occur. MAN’s customer information notice PCI 344/2015 required the connecting rods to be replaced after 100,000 running hours. MAN viewed the connecting rod as a whole unit, however PCI 344/2015 referred only to the risk of bottom end ovality as the reason for replacement. Although DFDS understood that PCI 344/2015 related to the whole connecting rod, in referring to only the bottom end as the reason for replacement, MAN potentially introduced confusion over whether replacement included the small end. Regardless, DFDS did not replace the connecting rod big ends after they reached 100,000 hours; instead, it measured a selection of big ends for ovality. Finding no ovality, DFDS decided to continue operating the connecting rods. This decision was made without consulting either MAN or LR. MAN had not considered the possibility of a vessel operator rotating small ends and keeping a supply of spares. It might make sense to occasionally replace a damaged small end with a new or overhauled spare. However, PCI 344/2015 referred to the connecting rod in its entirety, regardless of its separate components, and the manufacturer’s instructions should have been followed. Nevertheless, the ability to separate the small end from the shank, which included the big end, introduced the possibility of them being seen as separate components and maintained independently of each other. 2.5.4 Diesel Service Group quality control DSG was an established ship repair company and was certified and accredited to carry out repairs to ships’ auxiliary systems and main propulsion engines by various state authorities and maritime organisations. It had worked on DFDS MAN engines for almost 10 years and had replaced at least 24 piston pin bushes on board Finlandia Seaways and Botnia Seaways. It had also worked alongside PrimeServ technicians and was in the process of gaining ISO9001-2000 accreditation for its quality management system. DSG employed one person to oversee the work and produce its technical repair documents. However, documentation giving dates when each small end was overhauled, and by whom, was not available, and identification of every small end from DSG records was unclear as DSG had not always used the small end COC numbers for identification purposes. The visits to DSG by the MAIB, TAIID and LMSA included requests for documentation to identify when and how small end bush replacement was carried out. DSG was unable to provide this documentation. It was also apparent from the MAIB’s and TAIID’s visit to DSG that the damage to the small ends was viewed as inconsequential when compared with the mass of metal that made up the small end. As such, no comments or other references to the damage were identified in the available documentation.

50 DSG’s Certificate of Firm Conformity for the construction, conversion, modernization and repair of items of technical supervision products, issued by RMRS, included an engine power limitation of 8,500kW. This limitation implied that the Russian maritime administration did not consider DSG competent or equipped to undertake work on more powerful engines, such as MAN’s 12,600kW 48/60 engine. DSG attributed the power rating discrepancy to an administrative error, but was unable to substantiate this. However, the RINA ’Working Approved Statement’, which included the repair of ships’ main engines, did not specify a power limit. Given DSG’s experience and the number of the ship repairer’s other external accreditations, it is possible that the power rating limit set on its RMRS certificate was an administrative error. However, this should have been noticed and rectified by DSG’s quality control manager. Similarly, the poor record keeping identified during the MAIB, TAIID and LMSA visits to DSG’s workshops in Klaipeda, and the standards of work its service engineers were allowed to carry out over a prolonged period of time, were strong indicators of significant weaknesses in the company’s quality management system. 2.5.5 DFDS quality control DSG’s work on board Finlandia and Botnia Seaways was overseen by the vessels’ engineers and a DFDS technical superintendent. It was unclear what processes the vessels’ engineers and the technical superintendent followed when the overhauled small ends were returned to the vessels. There were no written inspection or quality control procedures, and it is unknown if any of the vessels’ engineers thoroughly examined the small ends before they were installed in the engine. It was apparent from MAN’s photographs (Figure 26) of the seven damaged small ends, that the cut marks at the periphery of the small end eyes were visible to the naked eye even with the bush installed. Consequently, either no pre-installation inspection took place, or there was no concern about the cut marks. The latter was probably the most likely as, when the damage was discussed with DFDS’s technical management team, a similar view to DSG was expressed. Any damage identified to major engine components, no matter how inconsequential it might appear, should be properly assessed, and any actions taken, or decisions made should be recorded in the planned maintenance system. The extent of the damage found to the connecting rods fitted to Finlandia Seaways and Botnia Seaways was significant and should have been highlighted to MAN and LR. 2.5.6 Classification Society oversight Finlandia Seaways’ main engine connecting rods were class survey items, but LR was unaware of the damage caused to the small ends during overhaul, or the fact that the connecting rod big ends had accumulated more running hours than recommended by the engine manufacturer. Finlandia Seaways was being managed on a CSM cycle, and its C/Es were authorised by LR to carry out examinations and report on the condition of selected machinery items. LR would then carry out a confirmatory survey with a review of the records of examination by the C/E and a general examination of items to be credited. It was apparent from the C/Es’ statements of examined machinery items that the replacement of piston pin bushes, or the use of the spare small ends, was not being reported. Furthermore, the small end damage had not been photographed or noted

51 in the vessel’s planned maintenance records. Consequently, once the small ends had been installed in the engine, the opportunity for an LR surveyor to identify a safety issue was lost. During machinery examinations, it is the vessel operator’s responsibility to present satisfactory maintenance and monitoring records and valid supporting information, such as original equipment manufacturers’ maintenance recommendations and service updates, in order for the class surveyor to complete the survey. Otherwise, as in this case, machinery items could have been credited which, according to the manufacturer, might already exceed the recommended service life. 2.6 EMERGENCY RESPONSE 2.6.1 Overview The vessel’s machinery alarm log (Table 1) showed how rapidly events developed before the engine shut down and the fire started. It was apparent that the crew had little opportunity to conduct an emergency shutdown of the engine to prevent further damage. After the catastrophic engine failure, the subsequent fire, caused by the ejection of hot oil, was intense but short-lived, and fire damage was localised. The fire was limited, in part, by the lack of readily available fuel sources, but also by the prompt actions of the crew, including closing the fire flaps on deck, shutting off the fuel quick-closing valves, stopping the fuel pumps and ultimately the injection of CO2 into the engine room. 2.6.2 Carbon dioxide bottle identification It was evident to the vessel’s crew, when they entered the CO2 bottle room, that the fixed fire extinguishing system had been activated and that CO2 had been released into the engine room. This was because the engine room injection valve was open, the discharge pipe and some bottles had frosted, and the manifold pressure gauge was seen to drop to zero. However, it was unclear how many bottles had been released. This was because the 221 gas bottles in the bottle room had not been clearly marked to show which protected spaces they served. It was several days after Finlandia Seaways arrived at Immingham that a specialist in fire protection systems confirmed that all 31 CO2 bottles assigned to the engine room had been released. Although DFDS had operated Finlandia Seaways for many years, and some of its crew had served on board many times, information to identify the 31 bottles assigned to the engine room was not readily available on board. It is apparent, both from this case, and from previous fire investigations as identified in MGN 389, that inadequate signage can potentially compound an already serious situation. If it is not possible to identify which, or how many bottles have been released, it leaves the crew in a predicament. On ships where several compartments are protected, each with a specified number of allocated bottles, crew need to be confident that a sufficient amount of gas has been released. MGN 389 states the measures that should be applied when checking the bottles that have been released, including frosting around the base of the bottles. However, in a large multiple bottle bank system it may not be feasible to inspect the base of bottles at the rear of a bank where only the bottle valves are visible. Having comprehensive

52 guidance on board to enable identification of which bottles relate to a specific compartment will assist with checks after releasing CO2, and reduce the likelihood of potentially fatal inadvertent releases of gas during post-accident activities. 2.6.3 Effective numbers and distribution of EEBDs There were three EEBDs in Finlandia Seaways’ engine room machinery spaces: one in the ECR, one in the workshop and one on the bottom plates close to the secondary escape route, meeting the requirements of MSC/Circ.849. As Finlandia Seaways was constructed before 1 July 2003, MSC/Circ.1081 was not applied, but, if it had been applicable, it would have required one EEBD to be positioned on each deck or platform level near the escape ladders, constituting the second means of escape. The escape route used by the 3/E to evacuate the engine room was the designated secondary means of escape, but only one EEBD was located at the lowest point of the secondary escape route, below the tween deck. The 3/E was trapped by the fire at the aft end of the engine room tween deck and had no choice but to use the secondary escape route. Fortunately, he was able to locate the escape route ladder in dense black smoke, and despite falling off the first vertical ladder several times, he was able to climb out of the engine room onto the open deck without an EEBD. EEBDs are designed to provide a minimum 10-minute air supply to allow personnel trapped by smoke or other toxic atmospheres to escape. To enable their use in an emergency, the number and distribution of EEBDs within a compartment needs to be considered carefully. If the 3/E had been able to don an EEBD during his escape, he would not have experienced the problems or suffered the injuries he did. It is unfortunate that the unified interpretation in MSC/Circ.1081 was not retrospectively applied to all ships given that installing extra EEBDs would not be an overly onerous requirement for vessel operators. On Finlandia Seaways, anyone trapped by smoke on the tween deck and unable to reach the engine workshop would need to climb down to the bottom plates to get an EEBD before using one of the two escape routes. Climbing down to retrieve an EEBD before making an escape is not particularly intuitive in an emergency. It is apparent from this accident that more consideration could be given to the number and location of EEBDs in Finlandia Seaways’ engine room. As a minimum, an additional EEBD at the secondary escape ladder on the tween deck level would be a sensible addition. 2.6.4 VDR VDRs provide crucial evidence to aid safety investigations. The VDR on board Finlandia Seaways stopped recording as soon as the ship blacked out. This denied investigators information that could have supported various elements of the investigation, including propulsion and communications data. Initial indications suggested that the UPS battery had failed. Although the VDR system had passed its yearly APT inspection in November 2017, the battery was incapable of providing power 5 months later. Recognising that older or incorrect batteries may not provide the longevity required, in May 2018 Kelvin Hughes issued a technical advice sheet, reducing the battery replacement period to 3 years from the previously stipulated 4 years. The supply of a battery maintenance kit also became mandatory from 1 January 2019.

53 SECTION 3 - CONCLUSIONS 3.1 SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

The fire on board Finlandia Seaways was the consequence of a sudden major engine component failure, which led to the ejection of heavy engine parts from the crankcase and release of hot oil vapours into the engine room. [2.3.1]

The catastrophic failure of Finlandia Seaways’ main engine was caused by a fracture of the A5 connecting rod small end. The A5 small end was found to have suffered a fatigue fracture. [2.3.2, 2.3.3]

The introduction of notches, probably caused during the bush removal process, introduced stress raisers into the small end, increased the likelihood of fatigue crack initiation and therefore fatigue failure. Practical testing has shown that a notch in a loaded component will reduce its lifetime-to-failure. [2.4.2, 2.4.3]

MAN’s inspection of Finlandia Seaways’ remaining small ends, and those of Botnia Seaways, confirmed that poor overhaul practices were the most probable cause of the A5 small end failure. [2.4.4, 2.4.5]

This investigation found that MAN’s guidance and recommended procedures for overhauling the connecting rod small ends had not been followed and the small ends had been damaged during the piston pin bush removal and fitting process. [2.5.1]

Seven of the eleven intact small ends removed from Finlandia Seaways’ engine were found to have notch damage on the mating surface with the bush. These were clearly the result of the use of a disc cutter to remove the bushes. Additionally, heat marks identified on Botnia Seaways’ small ends, and the lack of evidence provided by DSG on the use of liquid nitrogen, indicates that both heating and using a disc cutter on the bush insitu was common practice. [2.5.2] 7. The conrods were not overhauled in accordance with DSG’s written procedures, and maintenance was not properly recorded. [2.5.4]

There was a lack of appreciation by DSG or DFDS of the potential consequences of the stress raisers introduced during overhaul of the connecting rod small ends, and the damage was viewed as inconsequential when compared with the mass of metal that made up the small end. [2.5.4, 2.5.5]

Finlandia Seaways’ main engine connecting rods were class survey items, but LR was not informed of the damage caused to the small ends during overhaul, or that connecting rods had accumulated more running hours than recommended by the engine manufacturer. [2.5.6]

54 3.2 SAFETY ISSUES NOT DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

Localised heating was found on several small ends from Botnia Seaways. If not uniformly heated, the large mass of steel of the small end could introduce stresses, which could eventually initiate crack development. [2.4.4, 2.5.2]

On receipt of the customer information notice PCI 344/2015, DFDS decided to continue operating the connecting rods without consulting either MAN or LR. Had a discussion taken place, it might have led to the replacement of the entire connecting rods, identified the differential in operating hours of the small ends and the fact that small ends were being rotated between the two vessels. [2.5.3]

Regarding the overhaul of small ends, DFDS had no written inspection or quality control procedures, but given the notch marks on the small ends were visible to the naked eye, it is evident any examination by the vessel’s engineers was insufficiently thorough. [2.5.5]

It is apparent from the C/E’s statements of examined machinery items that small end bearing bush renewal, or the replacement of the small end with a spare, was not being reported to LR. [2.5.6]

This case and previous fire investigations, as identified in MGN 389, have highlighted that having comprehensive on board guidance, to enable identification of which CO2 bottles relate to a specific compartment, will assist with checks after releasing CO2. [2.6.2]

The 3/E was very fortunate to have survived the fire given the lack of an EEBD placed along the secondary means of escape. If he had been able to don an EEBD, he would not have suffered the stress and injury that resulted from his escape from the engine room. [2.6.3] 7. It is unfortunate that the unified interpretation in MSC/Circ.1081 was not retrospectively applied to all ships given that installing a few extra EEBDs would not be an overly onerous requirement for vessel operators and it would improve the opportunity to successfully escape from machinery spaces in an emergency. [2.6.3]

The VDR UPS battery failure denied data that could have supported various elements of the investigation, including propulsion data and communications. [2.6.4] 3.3 OTHER SAFETY ISSUES NOT DIRECTLY CONTRIBUTING TO THE ACCIDENT16

The customer information notice PCI 344/2015, in referring to only the bottom end as the reason to replace the connecting rod and the ability to separate the 48/60 engine connecting rod shank from the small end, potentially introduced confusion over whether maintenance of each component could be carried out separately. [2.5.3] 16 These safety issues identify lessons to be learned. They do not merit a safety recommendation based on this investigation alone. However, they may be used for analysing trends in marine accidents or in support of a future safety recommendation.

55 SECTION 4 - ACTION TAKEN 4.1 MAIB ACTIONS The MAIB and Lithuanian Transport Accident and Incident Investigation Division have: Issued a combined recommendation17 on 27 July 2018 to MAN Energy Solutions and to Lloyd’s Register to:

● Provide technical advice to DFDS on the actions the company should take to minimise the risk of a similar catastrophic engine failure on board Botnia Seaways, and any other vessel operators whose MAN engines might have been subjected to similar maintenance practices. 4.2 ACTIONS TAKEN BY OTHER ORGANISATIONS DFDS Seaways AB-Lithuania has:

● Inspected and renewed all 12 connecting rods on board Botnia Seaways.

● Carried out main engine overhauls on board five of the company’s vessels with reference to the MAIB/TAIID recommendation and manufacturers’ service letters.

● Increased the number of staff in their preventative maintenance system support team.

● Issued a fleet safety bulletin that instructs crews on board their vessels:

○to be familiar with CO2 systems, that they have clear operating instructions and the number of cylinders for each compartment is understood;

○to assess safe means of escape from the engine room and the location and number of EEBDs with respect to MSC/Circ.1081.

● Reviewed its database of contractors to ensure that those who carry out main engine repair and maintenance work on board its vessels are suitably qualified.

● Continued to review its company PMS on board vessels it operates to ensure that vessel equipment manufacturers’ maintenance, repair and replacement guidance is followed.

● Complied with the Kelvin Hughes technical advice sheet regarding VDR battery replacement. MAN Energy Solutions has:

● Investigated the cause of the A5 small end failure on board Finlandia Seaways.

● Analysed the remaining 11 small ends retrieved from Finlandia Seaways’ main engine. 17 Recommendation 2018/121

56

● Examined the small ends retrieved from Botnia Seaways’ main engine.

● Liaised with LR to provide appropriate guidance to DFDS in respect of the MAIB/ TAIID recommendation. Diesel Service Group has:

● Provided assurances that only Finlandia Seaways and Botnia Seaways, using the MAN 12V 48/60 main engine, were at risk of small end damage during bearing overhaul. Lloyd’s Register has:

● Liaised with MAN Energy Solutions to provide appropriate guidance to DFDS in respect of the MAIB/TAIID recommendation.

● Attended Botnia Seaways to inspect the small ends bearings removed from the main engine.

● Assessed whether other vessels, classed by LR and maintained by DSG, were at risk.

● Updated “The Examination of Surveyable Machinery Items by Chief Engineers” to include:

○the scope of service overhauls specified by the original equipment manufacturer;

○that the chief engineer’s maintenance and monitoring records should be made available;

○if the surveyor is not satisfied with the chief engineer’s statement, records or operating condition, they may request items be opened for inspection.

● Updated its instructions to surveyors relating to CSM surveys. Lloyd’s Register (Klaipeda) has:

● Undertaken to discuss the issue with technical superintendents of shipowners, operators and managers and/or ships’ C/Es when diesel engines repair/ maintenance works are undertaken by contractors, with respect to engine manufacturers’ recommendations. Kelvin Hughes has:

● On 14 May 2018, issued a VDR technical advice sheet for the supply of a battery maintenance kit to address regulatory requirement changes where old or incorrect type batteries had been used that failed to perform in the event of an incident. It advised that, on installation, the battery should be given a 3-year expiry date. The policy became mandatory from 1 January 2019.

57 SECTION 5 - RECOMMENDATIONS DFDS Seaways AB-Lithuania is recommended to: 2021/102 Review and improve how its chief engineers conduct class-related equipment examinations as part of the Continuous Survey Machinery cycle to ensure that examinations are conducted thoroughly and reported accurately. Diesel Service Group (Klaipeda) is recommended to: 2021/103 Fully apply equipment manufacturers’ maintenance and repair guidance and procedures. 2021/104 Review and, as necessary, amend its record keeping in order to generate a full and auditable record of the maintenance carried out by its staff. 2021/105 Review and update staff training to ensure familiarity with engineering methods appropriate for the various repair and overhaul tasks, backed up with a suitable quality assurance process to ensure standards are maintained. Safety recommendations shall in no case create a presumption of blame or liability

Marine Accident Report

FD06057 ｜ 2018_英国_unintentional-release-of-carbon-dioxide-from-fixed-fire-extinguishing-systems-on-ro-ro-vessels-eddystone-and-red-eagle_2018-16-_Eddystone_and_Red_Eagle.pdf.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2018 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H ACCIDENT REPORT MARINE INCIDENT REPORT NO 16/2018 SEPTEMBER 2018 Report on the investigation of unintentional release of carbon dioxide Eddystone in the Red Sea on 8 June 2016 and Red Eagle in Southampton Water on 17 July 2017

Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2018 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch First Floor, Spring Place 105 Commercial Road Southampton Email: maib@dft.gov.uk United Kingdom Telephone: +44 (0) 23 8039 5500 SO15 1GH Fax: +44 (0) 23 8023 2459 Press enquiries during office hours: 01932 440015 Press enquiries out of hours: 020 7944 4292

CONTENTS GLOSSARY OF ABBREVIATIONS AND ACRONYMS GLOSSARY OF TECHNICAL TERMS SYNOPSIS 1 SECTION 1 - BACKGROUND 2 1.1 Fixed fire-extinguishing systems 2 1.2 Carbon dioxide (CO2) as a fire-extinguishing medium 2 1.2.1 Properties of CO2 2 1.2.2 Effect of CO2 on human physiology 2 SECTION 2 - FACTUAL INFORMATION: EDDYSTONE 4 2.1 Particulars of Eddystone and incident 4 2.2 Narrative 5 2.2.1 Incident 5 2.2.2 System reinstatement 7 2.3 Company, vessel and crew 8 2.3.1 Company and vessel 8 2.3.2 Crew and other personnel 8 2.4 Fire-extinguishing system description 8 2.4.1 Configuration 8 2.4.2 Protection 8 2.4.3 Cylinder valves 10 2.5 Inspection and maintenance 11 SECTION 3 - FACTUAL INFORMATION: RED EAGLE 12 3.1 Particulars of Red Eagle and incident 12 3.2 Narrative 13 3.2.1 Incident 13 3.2.2 System reinstatement 13 3.3 Company, vessel and crew 15 3.3.1 Company 15 3.3.2 Vessel and crew 15 3.4 Fire-extinguishing system description 15 3.4.1 Configuration 15 3.4.2 Protection 17 3.4.3 Cylinder valves 17 3.5 Inspection and maintenance 19 3.6 Tests carried out 19 3.6.1 Cylinder valves’ leak test 19 3.6.2 Pressure switch test 20 SECTION 4 - COMMON TOPICS 22 4.1 Safer alternatives to CO2 22 4.2 Regulations and guidance 22 4.2.1 International requirements 22 4.2.2 Cylinder valves 22

1. 2.3

System testing 23 4.2.4 Land-based installations: requirements for CO2 cylinders and cylinder

valves 24 4.2.5 Approval of service suppliers 25 4.3 Previous accidents 25 4.3.1 Published reports 25 4.3.2 Cases involving sister vessels to Eddystone 26 4.3.3 Cases from Mariners’ Alerting and Reporting Scheme 27 4.3.4 United States Coast Guard reports 27 4.3.5 Research on CO2 related accidents 27 SECTION 5 - ANALYSIS 29 5.1 Aim 29 5.2 Overview 29 5.3 Mechanism of CO2 release 29 5.3.1 Eddystone 29 5.3.2 Red Eagle 30 5.4 Cylinder valve leakage 30 5.4.1 Eddystone 30 5.4.2 Red Eagle 31 5.4.3 Maintenance 31 5.5 Use of ship’s service air 31 5.6 Protection Systems 31 5.6.1 Vent valves 31 5.6.2 Leakage alarms 32 5.7 Ambiguity of guidance 32 5.7.1 Hydrostatic testing of high pressure cylinders 32 5.7.2 Cylinder valve testing, inspection and maintenance 33 5.7.3 Summary 34 5.8 Service suppliers 34 SECTION 6 - CONCLUSIONS 36 6.1 Safety issues directly contributing to the accident that have been addressed or resulted in recommendations 36 SECTION 7 - ACTION TAKEN 37 7.1 Actions taken by other organisations 37 SECTION 8 - RECOMMENDATIONS 38

FIGURES

Figure 1

CO2 cylinder pressure and temperature relationship

Figure 2

Eddystone’s CO2 room: bank A

Figure 3

Schmöle cylinder valve

Figure 4

Eddystone’s CO2 room: bank B

Figure 5

Bulging in outer skin of rubber hoses

Figure 6

Eddystone system drawing (inset: non-return valve in situ)

Figure 7

Non-return valve (A) and check valve (B)

Figure 8

Dense white cloud in CO2 room (inset: cylinder valve actuation head)

Figure 9

Red Eagle system drawing showing pressure switch

Figure 10

Trigger line

Figure 11

DAB D10963 1/2” CO2 cylinder valves with D10964 manual/pneumatic actuator with detent Figure 12a - Leaking DAB valve 97056 Figure 12b - Evidence of corrosion Figure 12c - Evidence of cleaning with wire wheel Figure 12d - Brass particles embedded in seating face TABLES

Table 1

Effect of CO2 on human physiology ANNEXES

Annex A

DAB Inspection Report – Red Eagle – 31st July 2017

Annex B

IMO circular - MSC.1/Circ.1318

Annex C

Fire Industry Association guidance note

Annex D

UR Z17 - Procedural Requirements for Service Suppliers

Annex E

MCA Safety Bulletin No. 12

GLOSSARY OF ABBREVIATIONS AND ACRONYMS

ADR

European Agreement Concerning the International Carriage of Dangerous Goods by Road

Barber

Barber Brothers Limited

BS

British Standard

CO2

Carbon dioxide

DAB

DAB Fire Engineering Company Ltd

EN

Normes Européennes (European Standard)

DNV GL

Det Norske Veritas – Germanisher Lloyd

EPA

Environmental Protection Agency

FIA

Fire Industry Association

FSS Code

International Code for Fire Safety Systems

GFT

Griffin Fire & Training Limited

HSE

Health and Safety Executive

IACS

International Association of Classification Societies

IMO

International Maritime Organization

ISO

International Organization for Standardization

kg

kilogram

LR

Lloyd’s Register

MARS

Mariners’ Alerting and Reporting Scheme

MCA

Maritime and Coastguard Agency

MOD

UK’s Ministry of Defence

MSC

Maritime Safety Committee

MSIS

Merchant Ships Instructions to Surveyors

Nm

Newton metres Ocean Safety - Ocean Safety Limited

RESMAR

RESMAR Limited

Ro-ro

Roll on, roll off

SOLAS

International Convention for the Safety of Life at Sea 1974, as amended

UR

Unified Requirement

USCG

United States Coast Guard GLOSSARY OF TECHNICAL TERMS

CO2 cylinder

Container for storing CO2 under pressure

Cylinder valve

The valve fitted on a cylinder, which is opened either manually or through the action of gas when CO2 is required for extinguishing fire

Bank

The set of CO2 cylinders designated for release into a space

Busting disc

A diaphragm that is designed to rupture at a pre-determined pressure

Check valve

A valve that restricts the flow of the medium to one direction. For the purpose of this report, check valves are not spring-loaded

Master cylinder

A cylinder that is activated first, causing the CO2 gas to enter the trigger line to discharge the remaining designated cylinders Non-return valve - A valve that restricts the flow of the medium in one direction. For the purpose of this report, non-return valves are spring-loaded

Pilot cylinder

A cylinder containing gas (usually CO2 or nitrogen) and stored in an enclosure separate and remote from the main CO2 cylinders

Pilot line

The piping system that leads the gas from the pilot cylinder(s) to the CO2 cylinder bank

Remote release

The means to discharge CO2 into a compartment on fire that is located away from the room containing the CO2 cylinders

Trigger line

The piping system that leads CO2 from the master cylinder(s) to activate the remaining cylinders in the bank TIMES: all times used in this report are ship’s time (UK summer time UTC+1 for Red Eagle and UTC+3 for Eddystone)

Eddystone Red Eagle

1 SYNOPSIS On 8 June 2016, the roll on, roll off cargo vessel Eddystone experienced an unintentional release of carbon dioxide from its fixed fire-extinguishing system while on passage in the Southern Red Sea. A similar incident took place on 17 July 2017 on board the roll on, roll off passenger ferry Red Eagle while on passage from the Isle of Wight to Southampton. In both cases, the engine room distribution valve for the carbon dioxide gas remained closed and gas leaked out into the compartment where the carbon dioxide cylinders were stored. Fortunately, no one was harmed in either of these incidents. However, a report issued by the UK Health and Safety Executive stated that the unintended release of carbon dioxide from fire-extinguishing systems caused 72 deaths and 145 injuries, mainly in the marine industry, between 1975 and 2000. In both incidents, the release of carbon dioxide was found to be the result of malfunctioning cylinder valves. The MAIB investigation revealed that maintenance of the fire-extinguishing systems had been inadequate, and that the available guidance for the marine industry on the maintenance and inspection of carbon dioxide fixed fire-extinguishing systems was insufficient. The Maritime and Coastguard Agency has issued a safety bulletin to the operators of all UK registered vessels regarding the appropriate maintenance of carbon dioxide fixed fire- extinguishing systems. This bulletin clarifies the UK administration’s position with respect to the application of the international requirements for the testing of gas cylinders. AW Ship Management (Eddystone’s manager) and Red Funnel (Red Eagle’s owner/ manager) have replaced the components that contributed to these incidents on the sister vessels in their fleets. Recommendations have been made to: • The Maritime and Coastguard Agency, to seek clarification from the International Maritime Organization of the maximum permitted periodicity between hydrostatic testing of individual high pressure cylinders (MSC.1/Circ.1318); and, to ensure that all safety devices fitted to carbon dioxide fixed fire-extinguishing systems are maintained and surveyed appropriately. • Det Norske Veritas – Germanischer Lloyd and Lloyd’s Register, to raise with the International Association of Classification Societies the issue of the quality of service provided by approved service suppliers in the maintenance of carbon dioxide fixed fire-extinguishing systems. • The owners of Red Eagle, to review the design of the carbon dioxide fire-extinguishing systems fitted to their vessels where the leakage of a single cylinder valve causes the entire system to discharge.

2 SECTION 1 - BACKGROUND 1.1 FIXED FIRE-EXTINGUISHING SYSTEMS The International Convention for the Safety of Life at Sea 1974, as amended (SOLAS) required machinery spaces of category A1 to be protected by a permanently installed fixed fire-extinguishing system that uses one of the following as the extinguishing medium: a. Gas b. High expansion foam c. Water-spray. Carbon dioxide (CO2) is an inexpensive, widely available, effective fire-extinguishing medium and is used for most marine applications where gas is the chosen medium. 1.2 CARBON DIOXIDE (CO2) AS A FIRE-EXTINGUISHING MEDIUM 1.2.1 Properties of CO2 CO2 is odourless, colourless and exists in gaseous form at atmospheric pressures and temperatures. CO2 has a high rate of expansion, enabling large areas to be flooded quickly. It is non-flammable and does not support combustion. When released into a confined space, a concentration of 20% or more CO2 is sufficient to displace the lighter oxygen molecules, smothering any fire. However, it provides almost no cooling effect and there is a risk of re-ignition if the space is subsequently vented before it has had sufficient time to cool. Above the critical temperature of 31.1°C, CO2 cannot be liquified by the application of pressure. A typical fire-extinguishing system cylinder at 20°C will have a vapour pressure of approximately 57 bar; this will rise to 74 bar at the critical temperature. Further increase in temperature will lead to increased pressure within the cylinder (Figure 1). CO2 will permeate through rubber at approximately five times the rate of air2. 1.2.2 Effect of CO2 on human physiology The main drawback of CO2 as a fire-extinguishing medium is that the amount required to suppress a fire is higher than the amount required to cause harm to human beings. Above a concentration of 10% it causes unconsciousness, followed by death. At concentrations of 17% and above, survival time is less than a minute (Table 1). 1 Machinery spaces of category A are those spaces and trunks to such spaces that contain either: 1. internal combustion machinery used for main propulsion; 2. internal combustion machinery used for purposes other than main propulsion where such machinery has in the aggregate a total power output of not less than 375kW; or 3. any oil-fired boiler or oil fuel unit, or any oil-fired equipment other than boilers, such as inert gas generators, incinerators, etc. 2 J.D. Edward & S.F. Pickering, 1920, Permeability of rubber to gases.

3 3 Shortness of breath Image courtesy of https://www.epa.gov/snap/carbon-dioxide-fire-suppressant-examining-risks Table 1: Effect of CO2 on human physiology Figure 1: CO2 cylinder pressure and temperature relationship

4 SECTION 2 - FACTUAL INFORMATION: EDDYSTONE 2.1 PARTICULARS OF EDDYSTONE AND INCIDENT SHIP PARTICULARS Vessel’s name Eddystone Flag UK Classification society Lloyd’s Register IMO number 9234070 Type Ro-ro cargo Registered owner Foreland Shipping Limited Manager(s) AW Ship Management Limited Construction Steel Year of build 2002 Length overall 193.0m Gross tonnage 23235 Authorised cargo Military equipment VOYAGE PARTICULARS Port of departure Jeddah, Saudi Arabia Port of arrival Duqm, Oman Type of voyage International Cargo information Military equipment Manning 21 MARINE CASUALTY INFORMATION Date and time 8 June 2016 at 1625 Type of marine casualty or incident Marine incident Location of incident 13 28.4 N; 042 37.4 E, Southern Red Sea Place on board CO2 room Injuries/fatalities None Damage/environmental impact None Ship operation Normal service Voyage segment Transit Internal and external environment Ambient air temperature 32°C, calm sea Persons on board 30

5 2.2 NARRATIVE 2.2.1 Incident At 1625 on 8 June 2016, while the roll on, roll off (ro-ro) cargo vessel Eddystone was in the Southern Red Sea on passage to Duqm, Oman, the fixed fire-extinguishing system’s CO2 release alarm sounded. The engine room had been changed over to unmanned machinery space watchkeeping mode but the second engineer, who was on duty that day, was in the engine room at the time, attending to a purifier alarm. He telephoned the chief engineer to apprise him of the situation and remained in the engine control room, which was separated from the engine room. The main engine, generators and engine room vent fans all continued to operate normally. The chief engineer and the electrotechnical officer inspected the remote CO2 release station and saw that it was undisturbed. They then went to the door of the CO2 room, where they heard a loud hissing noise from within. The chief engineer informed the master, who sent the chief officer to assist them. The chief engineer and chief officer donned breathing apparatus and entered the upper level of the CO2 room, which contained the 65 CO2 cylinders that comprised bank A4 of the vessel’s fixed fire-extinguishing system (Figure 2). A white cloud was seen forming at the far corner of the compartment and all the cylinder valves (Figure 3) were in the open position. They then went down to the lower level, where the reserve set of 65 CO2 cylinders comprising bank B (Figure 4) were located. All of bank B’s cylinder valves were in the closed position. 4 Eddystone’s CO2 cylinders were stored as two independent banks, A and B. Figure 2: Eddystone's CO2 room: bank A Safety pin Trigger line To bank B Rubber CO2 outlet hose CO2 manifold

6 Figure 3: Schmöle cylinder valve Over centre linkage Set screw Valve spindle Valve lid Over centre linkage Set screw Bursting disc Actuator plunger Pilot/ trigger line To manifold Valve in closed position Valve in open position

7 To prevent the release of CO2 from bank B, the chief engineer disconnected the pilot activation lines to both banks of cylinders and left bank B in the manual activation mode, allowing it to be activated from within the CO2 room if required. CO2 gas from bank A continued to leak from a few failed cylinder outlet rubber hoses over the next 4 days, during which time most of the rubber hoses in bank A developed bulges in their outer skin (Figure 5). 2.2.2 System reinstatement On 19 June, when Eddystone called at Bahrain, after Duqm, a CO2 fire-extinguishing system specialist attended the vessel. All four pilot cylinders and the cylinders of bank B were weighed to confirm they contained the full charge of gas. After carrying out visual checks on all the components, the pilot cylinders were reconnected and the system was put back in service with bank A isolated. The vessel then continued its planned voyage to Marchwood, UK. When Eddystone reached Marchwood, RESMAR Limited (RESMAR), the company contracted by the managers of Eddystone to carry out the annual inspection and maintenance of the CO2 fire-extinguishing systems on the vessels in their fleet, attended the vessel with its subcontractor, Barber Brothers Limited (Barber). All the cylinders from bank A were transported to Barber’s facility in Birmingham, refilled and returned to the vessel. Between 28 July and 2 August RESMAR replaced eight check valves and one non-return valve that had been identified to be leaking, and renewed all the rubber hoses in both banks. Figure 4: Eddystone's CO2 room: bank B Engine room distribution valve Figure 5: Bulging in outer skin of rubber hoses Bulging outer skin

8 2.3 COMPANY, VESSEL AND CREW 2.3.1 Company and vessel Eddystone was one of six sister vessels built for the transport of military cargoes and vehicles. Eddystone and three other sister vessels were owned by Foreland Shipping Limited and managed on their behalf by AW Ship Management Limited. All the vessels were on charter to the UK’s Ministry of Defence (MOD) and plied regularly between Marchwood, UK, and ports in the Middle East. 2.3.2 Crew and other personnel At the time of the incident, there were 21 crew members and nine military personnel on board Eddystone. All were UK nationals. The manning on board exceeded the MCA’s minimum safe manning requirements. 2.4 FIRE-EXTINGUISHING SYSTEM DESCRIPTION 2.4.1 Configuration The MOD required a fully redundant fixed fire-extinguishing system for the engine room, so all six sister vessels had been fitted with an additional bank of CO2 cylinders. Eddystone’s system had been supplied by Tyco Engineering Services, and comprised two banks, A and B, each of 65 cylinders including two master cylinders. There were also two remote release stations, each containing two pilot cylinders and capable of operating either bank (Figure 6). In operation, the gas from the pilot cylinders was routed through two independent lines: one for opening the engine room distribution valve and the other for activating the master cylinders of the chosen bank. The CO2 from the master cylinders released into the manifold through a spring-loaded non-return valve at the end of the cylinder rubber outlet hose. At the cylinder end of the non-return valve, a T-branch was fitted through which the gas was led, via the trigger line, to the release mechanism to trigger the remaining cylinders in the bank. Every tenth cylinder in each bank had a similar arrangement to maintain the trigger line pressure. The gas from each of the remaining cylinders was released to the manifold through a check valve, which consisted of a ball valve maintaining the seal by the weight of the ball. Check valves were not spring-loaded (Figure 7). 2.4.2 Protection To prevent an accidental release of CO2, the pilot lines were fitted with vent valves set to remain open normally and to close if the pressure exceeded 1.5 bar. In the event of leakage through the pilot cylinders’ valves, the vent valves would prevent gradual pressure build-up in the pilot lines that might otherwise lead to the cylinder valves being triggered. The manifold was fitted with a pressure switch designed to trigger an alarm if minor leaks from the cylinder valves caused the pressure in the manifold to rise. This pressure switch was not shown in the system drawing. During a visit to the vessel by MAIB inspectors on 24 August 2017, the chief engineer removed the pressure switch, which had a stated range of 0 to 10 bar, and tested it; it was found to be set at 12 bar. To protect the main CO2 manifold from extreme overpressure, the manifold was also fitted with a pressure relief valve set to open at 120 bar.

9 Figure 6: Eddystone system drawing (inset: non-return valve in situ) Vent valve Check valve Non-return valve Key Pilot line Main CO2 manifold Trigger line Bank A Bank B Relief valve

10 2.4.3 Cylinder valves The system’s main CO2 cylinders were all fitted with Schmöle5 K85-50 cylinder valves (Figure 3). The internal mechanism of these valves consisted of a valve spindle and valve lid sealing against the seat. The seat sealing surface was made of polyether ether ketone or PEEK, a synthetic material with high resistance to thermal, chemical and aqueous degradation. The valve was kept closed by an over-centre linkage tensioned by a set screw acting on the top of the valve spindle. To ensure that the valves remained closed irrespective of the ambient temperature and the associated variation of cylinder pressure, the manufacturer required this set screw to be tightened to between 10Nm and 11Nm at installation. These valves also provided a bursting disc diaphragm to protect the cylinder against extreme overpressure. 5 Now owned by VTI Ventil Technik GmbH. A B Figure 7: Non-return valve (A) and check valve (B)

11 The Schmöle valves were designed to be operated either manually, using the operating lever, or remotely, via the application of a minimum pneumatic pressure of 20 bar through the pilot or trigger line to the valves activation plunger mechanism. The tightening torque of the valve body to the cylinder was required to be between 160Nm and 180Nm. The manufacturer’s instruction manual stated that the valves did not have any design-related restrictions regarding their service life, and were fit for 100 releases. 2.5 INSPECTION AND MAINTENANCE Before the incident, between 2 and 5 October 2015, RESMAR carried out the annual inspection and maintenance of Eddystone’s fixed fire-extinguishing system. In December 2015, it also replaced eight cylinders with pressure tested cylinders in compliance with the generally understood requirement to hydrostatically pressure test 10% of cylinders every 10 years. During the annual inspection and maintenance, the ship’s service air, supplied directly from the main air receiver through a reduction valve set to approximately 7 bar, was used to blow through the system and pressure test the main manifold. The cylinder valves were visually inspected, and all gas cylinders were checked with liquid level detectors to ensure they contained their full charge of CO2. When MAIB inspectors visited Eddystone in August 2017, RESMAR technicians working on board reported that the service air pressure was insufficient to trigger the leakage alarm pressure switch. Neither Barber nor RESMAR were able to provide any documentary evidence to confirm that the valve set screws had been set to the required torque at any time before or after the incident.

12 SECTION 3 - FACTUAL INFORMATION: RED EAGLE 3.1 PARTICULARS OF RED EAGLE AND INCIDENT SHIP PARTICULARS Vessel’s name Red Eagle Flag UK Classification society Not applicable IMO number 9117337 Type Ro-ro cargo ferry Registered owner Southampton Isle of Wight & South of England Royal Mail Steam Packet Company Limited (known as Red Funnel) Manager(s) Southampton Isle of Wight & South of England Royal Mail Steam Packet Company Limited Construction Steel Year of build 1996 Length overall 93.22m Gross tonnage 4075 Authorised cargo Passengers and vehicles VOYAGE PARTICULARS Port of departure East Cowes Port of arrival Southampton Type of voyage Internal waters Cargo information 57 passengers, 34 vehicles Manning 8 crew members MARINE CASUALTY INFORMATION Date and time 17 July 2017 at approximately 0435 Type of marine casualty or incident Marine incident Location of incident Solent (off Bramble Banks) Place on board CO2 room Injuries/fatalities None Damage/environmental impact None Ship operation Normal service Voyage segment Transit Internal and external environment CO2 room temperature 18°C, calm sea Persons on board 65

13 3.2 NARRATIVE 3.2.1 Incident At 0420 on 17 July 2017, the ro-ro ferry Red Eagle departed East Cowes, Isle of Wight, bound for Southampton. Approximately 15 minutes after departing, the CO2 release alarm sounded in the engine room. The chief engineer and mechanic, who were in the engine control room at the time, ran out of the space, exiting onto the vehicle deck immediately above. The CO2 room and its remote release station were located on opposite sides of this deck. The chief engineer inspected the remote CO2 release station and saw that it was undisturbed. He then went to the door of the CO2 room, where he heard a loud hissing sound from within. He opened the door to the room slightly, and through a very small opening saw a dense white cloud inside (Figure 8). The chief engineer telephoned the bridge and discussed the situation with the master. They concluded that no emergency response was required other than ensuring that no one entered either the CO2 room or the engine room. The master then telephoned Griffin Fire & Training Limited (GFT), who were contracted to maintain the CO2 fire-extinguishing systems on the company’s ro-ro ferries, and asked them to attend the vessel as soon as possible. He also informed the vessel’s owners, who in turn notified the MCA. Red Eagle berthed at 0515 and, after disembarking the passengers and their vehicles, moved to a layby berth nearby. The vessel was taken out of service pending the restoration of the CO2 system. A GFT technician arrived on board at 0630 and, having donned breathing apparatus, went inside the CO2 room to investigate. He could hear several of the cylinder valves rattling due to the back pressure in the outlet manifold, and CO2 still leaking into the room from some of the cylinder outlet hose to manifold connections (Figure 8). To prevent further discharge of gas, the technician removed all the cylinder valve actuation heads (Figure 8 inset). An MCA surveyor arrived on board later that morning. At the request of the surveyor, the GFT technician proved that neither the remote nor local release mechanism for the CO2 system had been activated to cause the cylinders to discharge. 3.2.2 System reinstatement All 26 of the system’s CO2 main cylinders and the pilot cylinder were sent to Ocean Safety Limited (Ocean Safety), subcontractors to GFT. Ocean Safety established that the two master cylinders and the pilot cylinder contained their full charge; six cylinders had completely discharged and the remaining cylinders were partially discharged. They were unable to establish which cylinder valve had leaked, causing the system to discharge. Ocean Safety emptied and pressure tested all the CO2 cylinders, examining and refurbishing the cylinder valves as required. It was during the process of refilling the cylinders that the MCA asked the vessel’s owners to replace all the cylinder valves as the valve that had caused the leakage could not be identified. Ocean Safety then ordered 26 new cylinder valves from DAB Fire Engineering Ltd (DAB) in York. The filled cylinders were emptied and refilled with the new valves fitted, and the vessel returned to normal service on 21 July after the system had been reinstated.

14 The final report by GFT following its reinstatement of the CO2 system stated: A new 10-370 bar latching pressure switch was fitted to the manifold. This was tested and witnessed by the MCA representative. The MAIB subsequently pointed out that the purpose of this pressure switch was to raise an alarm in the event of the manifold becoming pressurised to 1 bar as a result of leakage past cylinder valves. The new pressure switch was not in accordance with the system drawing and could not raise the alarm before the system was Figure 8: Dense white cloud in CO2 room (inset: cylinder valve actuation head) Frosted union of check valve

15 triggered. GFT responded that they did not have the system drawings for Red Eagle. GFT later reported that they were unable to find a pressure switch of the right specification and was directed by the MAIB to report the matter to the MCA. GFT was subsequently informed by the CO2 system manufacturer that the original pressure switch had been unavailable since 2013 and that the pressure switch GFT had fitted was correct. 3.3 COMPANY, VESSEL AND CREW 3.3.1 Company Red Eagle was owned and operated by The Southampton Isle of Wight and South of England Royal Mail Steam Packet Company Limited, commonly known as Red Funnel. Red Eagle was one of three Raptor class sister vessels employed on a regular passenger, vehicle and freight transportation service between the Isle of Wight and Southampton. Red Funnel also operated a foot passenger only service between the Isle of Wight and Southampton with four high speed catamarans. 3.3.2 Vessel and crew Between 2003 and 2005, the three Raptor class vessels in the Red Funnel fleet were lengthened at Remontowa shipyard in Gdansk, Poland. Red Eagle was the last vessel to be extended. Due to the increased volume of the engine room resulting from the extension work, the fixed fire-extinguishing system needed an additional six CO2 cylinders to be added to the existing 20. Following this work, the main CO2 manifold up to the engine room distribution valve was pressure tested to 122 bar as required by Lloyd’s Register (LR), the classification society that witnessed and surveyed the work. The vessel’s crew comprising the master, mate, engineer, mechanic and four deck ratings satisfied the MCA’s safe manning requirements. 3.4 FIRE-EXTINGUISHING SYSTEM DESCRIPTION 3.4.1 Configuration Red Eagle’s fixed fire-extinguishing system was supplied by Wormald Engineering (later part of Tyco Engineering Services) and comprised 26 cylinders, including two master cylinders. The remote release station, which contained the pilot CO2 cylinder, was located on vehicle deck E, the deck above the engine room. In operation, the pilot gas was routed through two independent lines: one for opening the engine room distribution valve and the other for activating the master cylinders. The CO2 from the master cylinders released into the manifold through check valves at the end of the cylinder outlet hoses (Figure 9). From the end of the manifold, gas was led via the trigger line to the release mechanism of each cylinder valve in turn, causing them to open and release into the manifold through a check valve (Figure 10). The trigger line was not separated from the manifold as in the system on Eddystone.

16 Figure 9: Red Eagle system drawing showing pressure switch Check valve Key

17 3.4.2 Protection No part of the system was fitted with a vent valve to protect from gradual pressure build-up caused by leaking cylinder valves. The manifold was fitted with a widely available Danfoss KPS 33 pressure switch, designed to trigger the CO2 release alarm in the event of minor leaks from the cylinder valves as well as act as a CO2 discharge alarm. This pressure switch had a stated range of 0 to 3.5 bar and was to be set to break at 1 bar rising (Figure 9). To protect the main CO2 manifold from extreme overpressure, the manifold was also fitted with a bursting disc designed to rupture at 190 bar. 3.4.3 Cylinder valves The system’s main CO2 cylinders were all fitted with DAB D10963 1/2” CO2 cylinder valves (Figure 11), each of which had a D10964 manual/pneumatic actuator with detent fitted on top. When activated by pilot gas pressure or the manual release mechanism, the servo piston was pushed down, causing the firing pin to push the pilot seat down, opening the pilot valve. This allowed cylinder pressure to enter the top of the valve seat (made of a synthetic material called Nylon 6), pushing it Figure 10: Trigger line Trigger line CO2 manifold Check valve

18 Figure 11: DAB D10963 1/2" CO2 cylinder valves with D10964 manual/pneumatic actuator with detent Balance chamber Bursting disc Pilot seat Servo piston CO2 cylinder pressure Firing pin Valve seat Valve Valve Actuator head

19 further down until the detent on the actuator locked, keeping the valve open. These valves also provided a bursting disc diaphragm to protect the cylinder from extreme overpressure. A small hole had been drilled into the side of the servo piston, allowing the escape of any gas that might have become trapped within it. A restrictor wire inserted into this hole served to reduce the cross-sectional area of the hole to the required aperture. In DAB valves manufactured before 2000, this restrictor wire was made of copper. With the exception of the valve seat, all other internal components and the body of the valves were made of brass and stainless steel. The actuators fitted to the top of the valves were designed to be operated either manually, using the operating lever, or remotely, via the application of pneumatic pressure through the pilot or trigger line to the actuator. The instructions for the DAB actuator stated that a minimum pneumatic pressure of 31 bar was required to open it, when the cylinder pressure was between 45 and 50 bar. During the tests carried out at DAB, it was observed that an activation pressure of 4.1 bar was sufficient to activate the valve with a cylinder pressure of 50 bar (simulated with nitrogen gas). Although the valves were designed to be serviced, the manufacturer’s instruction manual stated: ‘DAB do not recommend that CO2 valves be serviced or refurbished in the field’. 3.5 INSPECTION AND MAINTENANCE On 9 June 2016, GFT replaced three cylinders with pressure tested cylinders in compliance with the generally understood requirement to hydrostatically pressure test 10% of cylinders every 10 years. On 19 December 2016, GFT completed the annual inspection and maintenance of Red Eagle’s fixed fire-extinguishing system. Following a visual inspection of the cylinder valves, which did not raise any concern, all the cylinder valve actuators were removed and proved to be working by pressurising the main manifold with dry compressed air. The system’s pressure switch activated when the manifold was pressurised. The pressure switch set point was not checked at the time. Prior to the incident on Red Eagle, Ocean Safety believed that DAB had ceased trading, and so refurbished cylinder valves on the three Raptor class vessels with spare parts held at its workshop. No records were maintained of these refurbishments. 3.6 TESTS CARRIED OUT 3.6.1 Cylinder valves’ leak test Following the accidental release of CO2, DAB tested all 26 cylinder valves from Red Eagle. The MAIB witnessed these tests. Nitrogen at 50 bar was applied to the cylinder side of the valves and the outlets were monitored for leakage using soap solution. Of the 26 valves tested, two (serial numbers 37017 and 97056) were found to be leaking.

20 The leakage from valve 97056 was very small. The bubble formed in the soap solution during testing took a few seconds to burst (Figure 12a). On dismantling this valve, evidence of corrosion was apparent (Figure 12b) and the DAB report of the tests (Annex A) stated: When valve 97056 was disassembled, it was noted that the outlet side of the valve was heavily discoloured internally and rust had formed in the outlet. As the valve is manufactured from brass and stainless steel it was surmised that the rust had come from the discharge hose fitting which are normally manufactured from zinc plated steel. [sic] The report also noted that the bonnet assembly showed signs of pitting associated with being cleaned with a wire wheel, or similar. The valve had obviously been disassembled previously during its service life (Figure 12c). With regard to the source of the leak, the seating area of the valve seat was found to have small particles, subsequently identified as brass, embedded in it (Figure 12d). The DAB report stated: From experience, this brass swarf was the most likely cause of the leak. The leak however, was so slight that it would have taken many months if not longer for the leak to build up pressure inside the manifold and discharge the system. The report concluded: Once a pressure of 60 PSI6 is achieved in the manifold the first of the valves would start to open causing a cascade effect in the rest of the system opening the remaining 25 valves. The leakage from valve 37017 was severe and, when the valve was dismantled, a piece of the copper restrictor wire from the servo piston was found to have lodged open the pilot cone of the valve seat. The DAB report stated that this could only have occurred when the system was activated. DAB confirmed that, based on the serial numbers of the cylinder valves, 20 had been supplied before the year 2000 and six additional valves were supplied in 2006 when the vessel was extended. 3.6.2 Pressure switch test The Danfoss KPS 33 pressure switch fitted to the CO2 manifold was examined by a local company specialising in pressure switch calibration. It found that: ● ●The switch was set to open at 0 bar rising ● ●Upon application of pressure up to 4 bar, the switch did not operate ● ●The lower diaphragm appeared to be contorted ● ●The upper plate that the mechanism attached to was bent upwards. 6 60 PSI = 4.14 bar.

21 Figure 12a: Leaking DAB valve 97056 Figure 12c: Evidence of cleaning with wire wheel Figure 12d: Brass particles embedded in seating face Brass particles embedded in seating face Figure 12b: Evidence of corrosion Corrosion

22 SECTION 4 - COMMON TOPICS 4.1 SAFER ALTERNATIVES TO CO2 Fixed fire-extinguishing systems that use CO2 pose a significant risk to human life if used within a confined space. Safer, alternative systems have therefore been developed, and shore-based fixed fire-extinguishing systems are increasingly using: Argogen (IG55), Argonite (IG55), Inergen (IG541), FM200 and NOVEC 1230. The document entitled ‘Fire protection standard for UK power networks operational sites’ was published in January 2016 by UK Power Networks. It set out to provide guidance on the broad fire safety legislative responsibilities and current fire safety standards as well as maintaining best practice from previous documentation concerning fire protection of UK power networks’ operational sites. It stated: Where gaseous fixed fire extinguishing systems are installed it is preferred that Carbon Dioxide is not used. 4.2 REGULATIONS AND GUIDANCE 4.2.1 International requirements SOLAS refers to the International Code for Fire Safety Systems (FSS Code) for the technical requirements of fixed gas fire-extinguishing systems. The FSS Code requires all gas-based fixed fire-extinguishing systems to be fitted with an audible and visual warning of the release of gas into spaces where personnel normally work or have access. For machinery spaces protected by CO2 systems, the FSS Code requires that 85% of the CO2 shall be discharged within 2 minutes, to cover 35%7 of the gross volume of the largest machinery space protected, including the engine casing. Red Eagle was categorised as a Class IV passenger ship engaged only on voyages in inland waters. Class IV vessels were required to comply with Statutory Instrument (SI) 1998 No.1011, The Merchant Shipping (Fire Protection: Small Ships) Regulations

1. The requirements for fixed gas systems were contained in MSN 1666, The

Merchant Shipping (Fire Protection) Regulations 1998, which formed an integral part of SI 1998 No.1011. The requirements set out in MSN 1666 did not include maintenance and testing routines as the MCA deferred to the International Maritime Organization’s circular MSC.1/Circ 13188 (Annex B), published in June 2009, for guidance on this subject. 4.2.2 Cylinder valves MSC.1/Circ.1318 aimed to provide the minimum recommended level of maintenance and inspections for fixed carbon dioxide fire-extinguishing systems on all ships, and are intended to demonstrate that the system is kept in good working order as specified in SOLAS regulation II-2/14.2.1.2. It listed the requirements and provided checklists for use by service suppliers to maintain a record of their inspections and maintenance. MSC.1/Circ.1318 stated that maintenance and inspections should be carried out in accordance with the manufacturer’s instructions, and listed monthly 7 or 40% of the gross volume excluding casings as defined in section 2.2.1.3 of FSS Code, whichever is larger. 8 MSC.1/Circ.1318,‘Guidelines for the maintenance and inspections of fixed carbon dioxide and fire- extinguishing systems’.

23 and annual checks that should be conducted. It further stated that biennially for passenger ships and at intermediate, periodic or renewal surveys of all other ship types, the following maintenance, inter alia, should be carried out: ● ●The hydrostatic test date of all storage containers should be checked. ● ●High pressure cylinders should be subjected to periodical tests at intervals not exceeding 10 years. At the 10-year inspection, at least 10% of the total number provided should be subjected to an internal inspection and hydrostatic test** The asterisks in the quote above refer to a footnote to the International Organization for Standardization (ISO) Standard, Gas cylinders – Seamless steel gas cylinders – Periodic inspection and testing, ISO 6406. The standard states that if a valve is to be reintroduced on to a cylinder, it shall be inspected and maintained to ensure optimal performance. A recommended maintenance plan is then outlined in an annex to the standard, and includes the following: Maintenance of the valve should include general cleaning, together with replacement of elastomers and worn or damaged components, packing and safety devices, where necessary. However, ISO 6406 refers to EN 14189 Transportable gas cylinders. Inspection and maintenance of cylinder valves at time of periodic inspection of gas cylinders, now replaced by BS EN ISO 22434:2011, Transportable gas cylinders – Inspection and maintenance of cylinder valves. This standard provides further guidance on external examination with the cylinder valve both in situ and removed from the cylinder. It states that if external examination reveals no defects or anomalies, the valves may be reintroduced into service. The MCA’s Merchant Ships Instructions to Surveyors (MSIS) stated: If the fixed CO2 cylinders are found to be in a very good condition at the 10 year interval, then the MCA will accept the hydrostatic test requirements specified in MSC/Circular 1318 paragraph 6.1.2, provided that all cylinders are tested within a 20 year maximum period. 4.2.3 System testing MSC.1/Circ.1318 under the heading Minimum recommended maintenance states: the discharge piping and nozzles should be tested to verify that they are not blocked. The test should be performed by isolating the discharge piping from the system and flowing dry air or nitrogen from test cylinders or suitable means through the piping. It also states under Example Service Charts: Manifold tested for leakage, by applying dry working air. Det Norske Veritas – Germanischer Lloyd (DNV GL) Class Guidelines CG-0058 re-iterate these requirements under Maintenance of safety equipment, which state that CO2 lines and nozzles should be blown through with dry air or nitrogen to confirm that they are not blocked.

24 Land-based guidance also requires the use of nitrogen or a suitable alternative in British Standard BS 5306-4 Fire-extinguishing installations and equipment on premises – Part 4: Specification for carbon dioxide systems which provides the following guidance: After assembly, the system shall be thoroughly blown through with nitrogen or suitable alternative … 4.2.4 Land-based installations: requirements for CO2 cylinders and cylinder valves In November 2015 the United Kingdom Fire Industry Association (FIA) published Guidance on the periodic testing of transportable gas containers used in fire- extinguishing systems (Annex C). This guidance gave recommendations on how fire-extinguishing system containers should be treated. Within the UK, the design, manufacture, inspection and test, and transport of dangerous goods, including transportable pressure equipment, is governed by the Carriage of Dangerous Goods and Use of Transportable Pressure Equipment Regulations. These regulations implement the European Transportable Pressure Equipment Directive and the European Agreement Concerning the International Carriage of Dangerous Goods by Road (ADR). FIA’s guidance refers to instruction P200 of ADR, which specifically requires CO2 cylinders to be pressure tested at 10-yearly intervals to 190-250 bar. The United States National Fire Protection Agency imposes the requirement for hydrostatic testing of all cylinders which are discharged after 5 years of the last test and imposes a maximum limit of 12 years between hydrostatic tests. On cylinder valves, the FIA guidance states: For safety reasons container [cylinder] valves should not be reused after removal from containers unless the following conditions are met: ● ●The valve has been refurbished in accordance with manufacturer’s recommendations ● ●The connection thread to the container and discharge hose are inspected to ensure they are within tolerance and undamaged. The guidance notes that, where cylinder valves have a taper thread, they are often found to be unfit for reuse after a single fitment and removal due to the distortion of the brass thread when tightened against the steel cylinder. ISO 6183, Fire protection equipment -- Carbon dioxide extinguishing systems for use on premises -- Design and installation, states that CO2 is no longer recommended for use in normally occupied spaces. Similarly, the National Fire Protection Agency of the United States of America recommends that CO2 should not be used in new total flooding systems for spaces that are normally occupied.

25 4.2.5 Approval of service suppliers The International Association of Classification Societies (IACS) provides guidance for the approval of service suppliers in Unified Requirement (UR) Z17, Procedural Requirements for Service Suppliers (Annex D). UR Z17 defines a service supplier as: A person or company, … provides services for a ship or a mobile offshore drilling unit such as measurements, tests or maintenance of safety systems and equipment, the results of which are used by surveyors in making decisions affecting classification or statutory certification and services. UR Z17 defines a subcontractor as: A Person or Company providing services to a Manufacturer or approved/ recognized service supplier, with a formal contract defining the assumption of the obligations of the service supplier. UR Z17 has detailed requirements on the training, personnel and quality systems of the service supplier, which are also applicable to their subcontractors (except those only providing equipment). It also requires the classification society that is providing the approval to audit the service supplier every 5 years. Firms engaged in servicing fixed fire-extinguishing systems are not required to be approved by the system’s manufacturer, but need to demonstrate an understanding of the principles involved with gas, foam, deluge, sprinkler and water mist systems, as relevant for the approval being sought. Access to system drawings, manufacturers’ instruction manuals and bulletins is required for the service supplier and nominated subcontractors. The service supplier approval certificates (for the inspection and maintenance of portable fire extinguishers, and gas and foam type fixed fire-extinguishing systems) were issued to RESMAR by Lloyd’s Register (LR), and to GFT by DNV GL. Eddystone had been serviced by GFT from September 2010 to October 2013 and by RESMAR from 2014 onwards. GFT had been Red Eagle’s service supplier since the vessel entered service. 4.3 PREVIOUS ACCIDENTS 4.3.1 Published reports There are several published reports concerning the release of CO2 from fixed fire- extinguishing systems. In 2004, the release of 5 tonnes of CO2 on board the Hong Kong registered containership YM People9 killed the vessel’s master, chief engineer, chief officer and third engineer. All the deceased were in the CO2 room at the time and were attempting to vent the manifold into which CO2 had inadvertently released while the chief engineer was preparing the system for inspection and maintenance. 9 Improper release of carbon dioxide on board M.V.”YM People” on 27.9.2004 causing four fatalities (https://www.mardep.gov.hk/en/publication/pdf/mai040927.pdf).

26 In 2010, the offshore support vessel Marsol Pride10 experienced a discharge of CO2 into the manifold due to a leaking Schmöle cylinder valve. Subsequently, the main engine room distribution valve developed a leak, which resulted in the engine room being flooded with CO2. Fortunately, there were no casualties. In 2011, a shore-based service engineer was seriously injured on board the tug SD Nimble11 when six cylinders of CO2 were accidentally discharged shortly after the tug had slipped from its berth. The engineer had been testing components of the vessel’s fixed CO2 fire-extinguishing system, in the CO2 room, without first disconnecting the system’s activation pilot line. The engineer lost consciousness, but was airlifted to a local hospital and recovered following a long period of recuperation and therapy. 4.3.2 Cases involving sister vessels to Eddystone Accidents on board two of Eddystone’s sister vessels were reported to the MAIB. These reports did not result in investigations by the MAIB, but the circumstances of the accidents as reported are outlined below: Sister vessel 1 of Eddystone In June 2017, technicians working on the CO2 fire-extinguishing system decided to pressure test the manifold after they noticed a corroded section on it. When they applied service air pressure to the manifold, all 65 cylinders in bank B of the system discharged into the manifold and leaked out through various unions between the cylinder outlet hoses and the manifold. There were no injuries. The subsequent investigation carried out by the chief engineer found that on 124 of 130 cylinders, the set screws on cylinder valves were not torqued to 11nm as required by the instruction manual for Schmöle valves. The service supplier, based in Barcelona, also carried out an investigation and reported as follows: CAUSES OF THE ACCIDENTAL RELEASE

1. One of the Check valves from Master Cylinder was not working properly.
2. The pilot loops from manifold check valves were not disconnected before apply

compressed air.

1. The bolts for calibrating the CO2 valves of many cylinders were untightened (less

than 11 Nm). Because that even the valves were secured by safety pin some of them were partially released. [sic] 10 Report 10-203: Marsol Pride, uncontrolled release of fire-extinguishing gas into engine room, Tui oil and gas field, 27 May 2010 (https://taic.org.nz/sites/default/files/inquiry/documents/10-203%20Final%20Published.pdf). 11 Report on the investigation of the accidental discharge of carbon dioxide on board SD Nimble, resulting in serious injury to a shore-based service engineer at Her Majesty’s Naval Base Faslane 23 August 2011 (https://www.gov.uk/maib-reports/accidental-discharge-of-carbon-dioxide-when-testing-the-fixed-co2-fire- extinguishing-system-on-tug-sd-nimble-at-hm-naval-base-faslane-scotland-with-1-person-injured).

27 Sister vessel 2 of Eddystone In January 2013, while the chief engineer was putting the CO2 system back into service at the end of the dry docking period, CO2 discharged from 14 cylinders whose safety pins had already been removed. The engine room distribution valve, which was locked shut, prevented the release of CO2 into the engine room. The cause of leakage was subsequently determined to be leakage from a Schmöle cylinder valve. 4.3.3 Cases from Mariners’ Alerting and Reporting Scheme The Mariners’ Alerting and Reporting Scheme (MARS) was a confidential reporting system run by The Nautical Institute to allow full reporting of accidents and near misses on board vessels, without fear of identification or litigation. Anonymous MARS reports based on these accidents and incidents were held in a publicly- accessible database as a free service to the industry. MARS 201311 reported on an inadvertent release of CO2 from 11 cylinders into the manifold, which subsequently leaked into the CO2 room of a vessel. The subsequent investigation established that a cylinder valve set screw had not been tightened to the correct torque, resulting in the valve opening when the ambient temperature in the CO2 room reached 50°C. MARS 201120, reported the failure of a CO2 extinguishing system to release during an engine room fire. This was found to have been due to several deficiencies in the maintenance of the system. These included leaking connections, failed distribution valves, failure of pilot cylinders to release, and water accumulation and corrosion inside the pipelines. The system had been inspected and serviced by an authorised service supplier shortly before the incident. 4.3.4 United States Coast Guard reports In July 2017, the USCG published Safety Alert 07-17, titled CO2 Hazards are Nothing New. But we’d like to remind you of what not to do! The alert reported a near fatal accident due to the unintentional release of CO2 while coastguard officers were inspecting a vessel. Technicians working on the system were responsible for releasing the CO2. 4.3.5 Research on CO2 related accidents In a paper12 exploring the risks of CO2 as a fire-suppressant, 146 deaths were found to be associated with the use of CO2 in the period 1940 to 1998. The data excluded deaths caused by fires, and concluded: These deaths point to a need for additional safety measures when using CO2. A large proportion (68%) of the post-1975 incidents was marine-related. Examination of the causes of the incidents indicated that only a limited number of crewmembers had the training and authority to activate the systems. Those crewmembers without training may not have had a true appreciation of the dangers that surround exposure to high CO2. Consequently, additional safety measures might be warranted for marine applications. 12 S.R. Skaggs at Halon Options Technical Working Conference (May 1998): Examining the risks of CO2 as a fire suppressant.

28 This data was further corroborated in a separate study13 by the US Environmental Protection Agency (EPA), which concluded: Examination of the accident records shows that a disproportionately large number of accidents involving carbon dioxide have occurred on marine vessels. And: Maritime regulations (46 CFR Part 76.1514 and SOLAS) do not provide detailed requirements to ensure safety of personnel. These maritime regulations can be contrasted with the NFPA15 standard that has more specific suggestions to protect personnel against the adverse effect of carbon dioxide. The report went on to recommend: Improvement of maritime regulations would at least provide specific requirements that would presumably help reduce the accidental exposures that occur in marine applications. The UK’s Health & Safety Executive (HSE) draws upon the data from the US EPA study in a major paper16 and states: From 1975 to 2000, a total of 51 incident records were located that reported a total of 72 deaths and 145 injuries resulting from accidents involving the discharge of CO2 from fire extinguishing systems. The review indicates that the majority of reported incidents occurred during maintenance on or around the CO2 fire protection system itself. 13 U.S. Environmental Protection Agency, February 2000: Carbon Dioxide as a Fire Suppressant: Examining the Risks. 14 U.S. domestic regulation. 15 The National Fire Protection Association (NFPA) is a U.S. trade association. 16 Peter Harper HSE: Assessment of the major hazard potential of carbon dioxide (CO2).

29 SECTION 5 - ANALYSIS 5.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 5.2 OVERVIEW The concentration of CO2 required to be released into a space to extinguish a fire is more than double that required to kill a human being within a minute. Therefore, any accidental release of CO2 can, and often does, have fatal consequences. Ashore, the use of CO2 is discouraged in spaces where people may be present. The data collated by the EPA on fatal accidents and injuries from 1975 to 2000 most likely influenced this move against the use of CO2. Several safer alternatives have been developed, and the move away from using CO2 as an extinguishing agent has improved the safety of land-based systems. The marine industry’s reluctance to adopt a similar approach is likely to cost more lives in the future. The two incidents investigated for this report were both caused by leakage of CO2 through cylinder valves. Unlike the Marsol Pride incident, in both of these cases the distribution valves held, protecting ship’s staff from exposure to CO2, reducing the potential for injury or loss of life. It is likely that in both cases the leakage of the cylinder valves, and the failure of the alarms designed to warn of such leakage, was attributable to the work of approved service suppliers operating within the ambiguous and limited international requirements regarding these systems and their maintenance. 5.3 MECHANISM OF CO2 RELEASE 5.3.1 Eddystone Immediately after the incident on board Eddystone, the chief engineer observed that all 65 cylinder valves in bank A had opened, and that the manual and remote releases remained undisturbed. In order for this to have occurred, one or more cylinder valves must have leaked, pressurising the manifold. This led to the pressurisation of the trigger line, through the non-return valve that was subsequently found to be leaking, and the discharge of all the cylinders in bank A. It was not possible to identify which cylinder valve had initiated the discharge, and there were no available records for the torque settings of the Schmöle valve set screws for either bank of cylinders from before or after the incident. The design of Schmöle valves was such that if their set screws were inadequately torqued, the likelihood of the valve leaking would rise significantly with increased cylinder pressure. At the time of the unintended release, Eddystone had just entered the Southern Red Sea, and the ambient temperature of the CO2 room had risen above CO2’s critical temperature. The fill density of the cylinders was not recorded but, in any event, given the minimum permissible fill density, this temperature rise would have caused the pressure within the CO2 cylinders to rise rapidly to over 76

30 bar at 32°C (Figure 1). Therefore, it is considered almost certain that the leakage on Eddystone was the result of one or more incorrectly torqued Schmöle valve set screws. There is a history of similar incidents involving Schmöle type cylinder valves leaking as a result of incorrect torques having been applied to their set screws at installation or service. These include the MARS 201311 report and the two incidents on Eddystone’s sister vessels, where one was found to have had almost all of its Schmöle valve set screws incorrectly torqued. 5.3.2 Red Eagle Following the incident on Red Eagle, it was established that all but the two master cylinders had been released, despite the manual and remote releases remaining undisturbed. Red Eagle’s fixed fire-extinguishing system was designed such that pressurisation of the manifold, normally from the master cylinders, would lead to the release of all the remaining cylinders through a direct connection between the manifold to the trigger line (Figure 10). The unintended release was caused by one, or more, of the cylinders (other than a master cylinder) leaking into the manifold until the activation pressure was reached in the trigger line, when the remaining cylinders started being activated. Both master cylinders remained undischarged because their trigger lines were not connected to the main manifold. During tests carried out at DAB’s facilities it was established that the actuators would start opening cylinder valves at a trigger pressure of 4.14 bar. Cylinder valves 97056 and 37017 were also identified to be leaking during these tests. However, the leakage from valve 37017 was considered to have resulted from the incident. The localised frosting of the union between the check valve and the cylinder outlet hose visible in Figure 8, taken immediately after the incident, was a result of cooling caused by the rapid expansion of CO2 as it escaped. This indicated that this was a likely source of leakage from the system following the unintended release. Therefore, assuming the check valves were in good working order and that the manifold was completely gas tight at low pressure, it is possible that the slow leak found in valve 97056 led to a gradual build-up of pressure. However, this would have taken several weeks to reach the 4.14 bar needed to trigger the discharge. It is also possible that the valve that leaked, causing the system to discharge, had subsequently re-seated and held pressure during tests conducted at DAB. 5.4 CYLINDER VALVE LEAKAGE 5.4.1 Eddystone Schmöle valves, such as those fitted on board Eddystone, were designed to operate 100 times before they needed to be replaced, and they could not be refurbished. However, it was critical that the correct torque was applied to their set screws at installation, and checked during the annual inspection and maintenance. In the absence of any record of the set screw torques having been set or checked, it is likely that this was not done during the system’s last inspection 8 months before the incident. The setting and checking of the set screws’ torque are specific to Schmöle valves, so do not appear in any of the general guidance or requirements for annual inspection and maintenance. It should, however, form part of the vessel-specific worklist developed by the service supplier.

31 5.4.2 Red Eagle The CO2 cylinders fitted on Red Eagle were equipped with DAB valves. There was no written guidance concerning the service life of these valves, although after the incident DAB stated that they should be replaced every 10 years. The valves were designed to be serviced, but the manufacturer’s maintenance instructions stated that they were not to be refurbished by the end-user, except in an emergency. The leaking valve that led to the unintended release of CO2 on board Red Eagle could not be identified with certainty. Therefore, the specific cause of the leakage cannot be known. However, the valves showed evidence of refurbishment by service suppliers, including the use of a wire wheel or similar to clean the internal components. This is contrary to the manufacturer’s instructions and led to the entrapment of brass particles on the sealing surface of one valve, causing it to leak. 5.4.3 Maintenance Given the potentially fatal consequences of unintentional discharge of CO2 from fixed fire-extinguishing systems, the current ‘fit and forget’ approach to cylinder valves is unsafe. Regular inspection and maintenance in line with the manufacturer’s instructions is of paramount importance in ensuring the safety of these systems. 5.5 USE OF SHIP’S SERVICE AIR Ship’s service air is typically taken directly from the air receiver and, as a result, contains both moisture and oil. Unlike control air, service air is not dried or filtered. Therefore, applying service air to CO2 systems introduces moisture, leading not only to internal corrosion of pipework and fittings, but also to the risk of blockages caused by internal icing when CO2 is released. It is for these reasons that MSC.1/Circ.1318 requires the use of nitrogen or dry air. The annual inspection and maintenance of Eddystone’s CO2 system was carried out in October 2015. When the system was being reinstated following the incident in July 2016, eight check valves and one non-return valve were found to be leaking. It is likely that the practice of using ship’s service air contributed to the premature failure of these fittings. 5.6 PROTECTION SYSTEMS 5.6.1 Vent valves Fixed fire-extinguishing system designers were aware of the potential for cylinder valves to leak during their service life. To mitigate the risk of unintentional system discharge posed by such leakage, many systems were fitted with vent valves. These were designed to prevent pressure build-up by venting CO2 leakage to atmosphere, typically up to a pressure of about 1.5 bar. To ensure normal operation when activated, these valves would close when the system was subjected to operational pressure. However, vent valves were not mandatory for CO2 fixed fire-extinguishing systems. Eddystone’s system was fitted with vent valves in its pilot lines. However, the leakage occurred into the manifold and the trigger line, and did not affect the pilot lines.

32 5.6.2 Leakage alarms The leakage alarms fitted to both vessels activated only once the cylinders released CO2 into the manifolds. The alarm set points were outside the range of the respective pressure switches, and it was fortunate that they triggered the alarm. Had the distribution valves been open (or failed as a result of the shock loading), allowing CO2 to release into the engine rooms, it is possible that no alarm would have sounded since the manifold pressure would have been significantly lower. Leakage alarms were not required by national or international regulations. Therefore, they were not mentioned in the Example Service Chart for high pressure CO2 systems included in MSC.1/Circ.1318 (Annex B) as it only included checks on components required by the FSS Code. As a result, their purpose and function were not understood by the service suppliers or surveyors involved, and they had been subjected to inappropriate tests. RESMAR technicians recognised that the 7 bar pressure available from the ship’s service air on Eddystone was insufficient to trigger the alarm. Had they understood the purpose of the alarm, it would have been clear that it should have been set to trigger at the lowest pressure possible. The pressure switch for the alarm fitted to Red Eagle was found to have been adjusted to trigger at 0 bar as it was probably damaged by the application of compressed air at a pressure above its maximum rated pressure. It is likely that at some time in the past, the setting of the pressure switch was adjusted until the alarm was triggered during annual pressure testing of the manifold. Without reference to the system drawing, the range, set point and purpose of the alarm were misunderstood. This was further demonstrated by the fitting of a switch with a range of 10-370 bar when re-instating the system following the CO2 release. That this switch was accepted by the MCA surveyor and Johnson Controls International demonstrates the need for improved guidance on the maintenance and operation of these systems. This is especially true in the case of legacy systems, when the rationale behind the design features of the original system becomes lost or is misunderstood following a change of company ownership. 5.7 AMBIGUITY OF GUIDANCE 5.7.1 Hydrostatic testing of high pressure cylinders MSC.1/Circ.1318 section 6 provides recommended minimum maintenance for fixed CO2 fire-extinguishing systems. With respect to hydrostatic testing, sub-section 6.1.2 initially states that the hydrostatic test date of all storage containers should be checked. It is not immediately clear what the purpose of checking the dates is, although the text that immediately follows it: High pressure cylinders should be subjected to periodical tests at intervals not exceeding 10 years, could imply that hydrostatic testing should not exceed a 10-year periodicity. However, the subsequent sentence, further reinforced by point 19 of Example Service Charts in the Appendix, states that at the 10-year inspection, at least 10% of the total number provided should be subjected to an internal inspection and hydrostatic test. This can result in the interpretation that the same 10% of cylinders may be hydrostatically tested every 10 years, with the result that 90% of a vessel’s cylinders could remain untested throughout its service life.

33 Many administrations have chosen to interpret the above requirement for testing periodicity as meaning that only 10% of cylinders need to be hydrostatically tested every 10 years. In the UK, this interpretation is negated by the MCA’s guidance to surveyors that stipulates that all CO2 cylinders should be tested within 20 years. In this context, it is noteworthy that the maintenance requirements for land-based systems make it mandatory that all CO2 cylinders be hydrostatically tested within a 10-year period (12 years in the USA). The divergence in approach between land and marine organisations to the safety of CO2 systems is hard to reconcile, especially as a malfunctioning CO2 installation on an ocean-going vessel has the potential to leave the vessel without a fixed fire-extinguishing system until it arrives at a port where the system can be reinstated. Given that the marine industry, unlike shore-based industries, is not making the transition to safer alternatives to CO2, there is an urgent need for the International Maritime Organization (IMO) to clarify MSC.1/Circ.1318 section 6 to remove any ambiguity about the requirements for hydrostatic testing of high pressure cylinders. 5.7.2 Cylinder valve testing, inspection and maintenance MSC.1/Circ.1318 provides no guidance on the testing and maintenance of cylinder valves, beyond recommending an annual visual inspection and stating that any maintenance should be carried out according to the system manufacturer’s instructions. It does refer, however, in a footnote, to ISO 6406, which in turn refers to EN 14189 (now superseded by BS EN ISO 22434:2011). ISO 6406 provides some guidance on valve maintenance, recommending general cleaning, together with replacement of elastomers and worn or damaged components, without specifying the need to refer to manufacturer’s instructions. BS EN ISO 22434:2011 permits valves to be re-used without maintenance if an external inspection reveals no defects or anomalies. In practice, due to the ambiguity in MSC.1/Circ.1318 outlined in 5.7.1 above, the only test carried out on cylinder valves is a leak test when the cylinders are refilled following hydrostatic testing. It is therefore possible for 90% of a vessel’s CO2 cylinder valves to remain untested for the life of the vessel. Since it is not possible to establish by visual inspection that the set screw on a Schmöle type valve is correctly torqued, 90% of these could also remain incorrectly set for the life of a vessel, or until they leak. In the incident involving Red Eagle, it is possible that CO2 leaked through a seal contaminated with brass particles that had resulted from the service supplier using a wire wheel or similar while refurbishing the DAB valve. This refurbishment was not in accordance with manufacturer’s instructions (see 5.4.2), but was likely a consequence of the service supplier believing DAB was no longer trading. Finally, FIA guidance (see 4.2.4) recommends that cylinder valves with tapered threads, the most common types used in the marine industry, are replaced when removed after a single construction due to likelihood that the brass thread will have distorted upon tightening into a steel cylinder. This guidance is not replicated in MSC.1/Circ.1318.

34 From the above, it can be concluded that ambiguity exists over the appropriate servicing standards to be applied to cylinder valves, and that the generic guidance available to the industry varies in quality. The ambiguity is further compounded by the uncertain requirements for hydrostatic testing of CO2 cylinders as described in the previous section. 5.7.3 Summary Given the hazard that CO2 poses to human health, there is a need for the IMO to clarify MSC.1/Circ.1318 section 6 to remove any ambiguity about the requirements for hydrostatic testing of high pressure cylinders. [5.7.1] It can be concluded that ambiguity exists over the appropriate servicing standards to be applied to cylinder valves, and that the generic guidance available to the industry varies in quality. [5.7.2] 5.8 SERVICE SUPPLIERS CO2 based fixed fire-extinguishing systems are entirely reliant on the ability of ships’ staff to release the required quantity of CO2 into the appropriate space in the event of an emergency. Flag administrations, classification societies, ships’ owners, operators and crew all rely on approved service suppliers to ensure that these systems are in a continuous state of readiness, by means of regular maintenance and testing. Service suppliers are often called upon to attend unfamiliar vessels when they call into port. It is accepted that working on an unfamiliar system within the constraints of a vessel’s busy schedule in port could be challenging. However, both Eddystone and Red Eagle were vessels known to their respective service supplier. The investigations into the incidents described in this report found significant deficiencies in the inspection and maintenance provided by service suppliers. These included: ● ●Refurbishing of cylinder valves contrary to the manufacturer’s guidance. ● ●Not maintaining records when cylinder valves were refurbished. ● ●Blowing through or pressurising Eddystone's system with ship’s service air instead of dry air or nitrogen. ● ●Working on systems without adequate documentation such as system drawings. ● ●Fitting an inappropriate pressure switch. ● ●Setting pressure switches outside their operating range. ● ●Not testing pressure switches. ● ●Not maintaining control and oversight over subcontracted third parties.

35 Many past incidents have occurred during maintenance, and as a result of poor maintenance of system components. The incidents on SD Nimble and sister vessel 1 of Eddystone; the cases reported in MARS 201311; and the USCG report citing several serious deficiencies observed immediately after a system service, demonstrate that such deficiencies are widespread. This is further supported by the findings of the UK HSE and the US EPA. The Procedural Requirements for Service Suppliers, UR Z17 (Annex D), describes requirements placed by the classification societies on those approved to maintain safety systems and equipment. However, it is apparent from the findings of this report, that the level of service given by some approved service suppliers regularly fails to maintain the safety of CO2 based fixed fire-extinguishing systems on board ships.

36 SECTION 6 - CONCLUSIONS 6.1 SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

It is almost certain that the CO2 fire-extinguishing system on Eddystone discharged because of one or more incorrectly torqued Schmöle cylinder valve set screws. [5.3.1] 2. The unintended release of CO2 on Red Eagle was caused by one or more cylinder valves leaking into the manifold, causing the system to discharge. [5.3.2] 3. Red Eagle’s fixed fire-extinguishing system was designed such that pressurisation of the manifold would lead to the release of all the remaining cylinders. [5.3.2] 4. It is likely that the torque on Eddystone’s Schmöle valve set screws had not been checked during the system’s last inspection 8 months earlier. [5.4.1] 5. Red Eagle’s cylinder valves showed evidence of refurbishment by service suppliers, which led to the entrapment of brass particles on the sealing surface of one valve, causing it to leak. [5.4.2] 6. Regular inspection and maintenance of CO2 cylinder valves in line with the manufacturer’s instructions is of paramount importance in ensuring the safety of these systems. [5.4.3] 7. The practice of using ships’ service air to blow through CO2 systems is likely to have contributed to the premature failure of fittings on these systems. [5.5] 8. The purpose and function of leakage alarm switches was not understood by the service suppliers or surveyors involved, resulting in them being incorrectly set and subjected to inappropriate tests. [5.6.2] 9. Given the hazard that CO2 poses to human health, there is a need for the IMO to clarify MSC.1/Circ.1318 section 6 to remove any ambiguity about the requirements for hydrostatic testing of high pressure cylinders. [5.7] 10. It can be concluded that ambiguity exists over the appropriate servicing standards to be applied to cylinder valves, and that the generic guidance available to the industry varies in quality. [5.7.2]

Flag administrations, classification societies, ship’s owners, operators and crew all rely on approved service suppliers to ensure that CO2 fixed fire-extinguishing systems are in a continuous state of readiness, by means of regular maintenance and testing. [5.8] 12. The level of service given by approved service suppliers regularly fails to maintain the safety of CO2 based fixed fire-extinguishing systems on board ships. [5.8]

37 SECTION 7 - ACTION TAKEN 7.1 ACTIONS TAKEN BY OTHER ORGANISATIONS The Maritime and Coastguard Agency has: Issued a safety bulletin to the operators of all UK registered vessels clarifying the ambiguity of MSC.1/Circ.1318 regarding the hydrostatic testing periodicity for CO2 cylinders and the appropriate maintenance of CO2 fixed fire-extinguishing systems and their components (Annex E). Red Funnel has: Replaced the CO2 cylinder valves on all the Raptor class vessels in its fleet. AW Ship Management Limited has: Replaced the non-return and check valves on all the sister vessels of Eddystone.

38 SECTION 8 - RECOMMENDATIONS The Maritime and Coastguard Agency is recommended to: 2018/123 Ensure that surveys check that all safety devices fitted to CO2 fixed fire- extinguishing systems are appropriately maintained and surveyed. 2018/124 Seek clarification from the IMO of the maximum permitted periodicity between hydrostatic tests of individual high pressure cylinders, as detailed in MSC.1/ Circ.1318 ANNEX B 6.1.2. Det Norske Veritas – Germanischer Lloyd and Lloyd’s Register are recommended to: 2018/125 Propose to the International Association of Classification Societies that an investigation be carried out into the application of Procedural Requirements for Service Suppliers, UR Z17. This should take into consideration the finding of this report, that the level of service provided by approved service suppliers regularly fails to maintain the safety of CO2 based fixed fire-extinguishing systems on ships. Red Funnel is recommended to: 2018/126 Review the design of the CO2 fixed fire-extinguishing systems fitted to its vessels where the leakage of a single cylinder valve causes the entire system to discharge. Safety recommendations shall in no case create a presumption of blame or liability

Marine Accident Report

FD06058 ｜ 2018_英国_unintentional-release-of-carbon-dioxide-from-fixed-fire-extinguishing-systems-on-ro-ro-vessels-eddystone-and-red-eagle_2018-16-_Eddystone_and_Red_Eagle_Annexes.pdf.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2018 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A DAB Inspection Report — Red Eagle — 31st July 2017

Inspection Report – Red Eagle – 31st July 2017 Background DAB Fire Engineering Ltd were originally contacted on Tuesday the 18th of July by of the Maritime and Coastguard Agency and informed of an accidental CO2 discharge aboard the Red Eagle Passenger ferry. The CO2 system was fitted with DAB CO2 Cylinder Valves (D10963/D) and DAB Manual Pneumatic Actuators (D10964/B). It should be noted that DAB Fire Engineering Ltd is not the original supplier of the valves and actuators - These were previously supplied by DAB Engineering Company Limited - but as the new manufacturer of DAB products we agreed to help with the incident. DAB Fire were subsequently contacted by of the Marine Accident Investigation Branch and asked if we could help with the inspection and testing of the valves. 26 valves were sent to DAB for inspection and all inspection and testing was witnessed by . Initial Inspection All valves were unpacked and the serial numbers logged as below. A quick visual inspection showed that there were 4 separate dates of valves. The serial numbers of the oldest valves (37000 range) were commensurate with the age of the vessel. (1996) The 39938 valve would be approximately circa 1997/8. These valves are before the date of the electronic records so were unable to date the valves exactly. The valve marked 73719 was new enough to be included in our serial number log and showed as being supplied to Tyco Engineering Services on 04th September 2000. The valves marked 97050, 97056, 97057, 97054 and 97059 were part of a batch supplied to Tyco Marine Services on 2nd November 2004. Externally all valves looked to be in good condition considering their age, with the exception of valve no 97056 which showed discolouration and evidence of rust inside the outlet port. Leak Testing All valves were subjected to a standard seat leakage test which consists of pressurising the valve to 40 Bar through the syphon tube thread and applying a soapy solution to the outlet port to produce a film. Any leakage would be demonstrated by a bubble forming and bursting. All valves passed the leakage test except for valves no 37017 and 97056. Valve 37017 had a reasonably major leakage and was retested with a flow meter on the outlet to try and measure the leak rate. The leak was over the range of our meter and far in excess of 5 litres per minute. Valve 97056 had a very minor leak that was far below the range of our flow meter. The bubble forming on the outlet took approx 20 seconds to burst.

Internal inspection. The leaking valves were disassembled and inspected internally. Valve 37017 Valve 37017 looked to be in reasonable condition internally, but it was noted that the restrictor wire had dislodged from the servo piston and had been caught between the main seat and the pilot check. This would’ve caused the leak. Despite the leak, it was determined that this leak was not the cause of the system activation as the leak would’ve been obvious when the cylinder was filled and the broken wire had most likely been sitting on top of the main seat for some time and lodged between the pilot check and the seat when the system was activated. Valve 97056 When valve 97056 was disassembled, it was noted that the outlet side of the valve was heavily discoloured internally and rust had formed in the outlet. As the valve is manufactured from brass and stainless steel it was surmised that the rust had come from the discharge hose fitting which are normally manufactured from zinc plated steel.

Upon further inspection, it was also noted that the bonnet assembly showed signs of pitting associated with being cleaned with a wire wheel, or similar. The valve had obviously been disassembled previously during its service life.

The main seat was inspected using an eye loupe and small flecks of brass swarf were visible in the main seating area. From experience, this brass swarf was the most likely cause of the leak. The leak however, was so slight that it would have taken many months if not longer for the leak to build up pressure inside the manifold and discharge the system.

Activation pressure test. A new CO2 valve was fitted to our test rig and a pressure of 50 Bar applied to the inlet. A standard D10964 Manual / Pneumatic actuator was fitted to the valve, and pressure increased to the activation port until the valve began to fire. At approximately 60PSI the valve Pilot seat began to open slightly.

Conclusion Without gaining access to the rest of the CO2 suppression system it is difficult to conclude if the leaking valves were the cause of the accidental initiation of the system. If the complete manifold and pilot system aboard the Red Eagle was 100% sealed then it would be possible for the very minor leak in valve no 97056 to slowly build up over time and provide enough pressure to slowly activate the rest of the system. Once a pressure of 60 PSI is achieved in the manifold the first of the valves would start to open causing a cascade effect in the rest of the system opening the remaining 25 valves. My recommendation would be for a bleed valve to be fitted to the pilot system to prevent accidental build up of pressure in the future. DAB has always unofficially suggested a life expectancy of 10 years for all of its products, though the life span may be extended if the products are well maintained. It was also noted that the system also contained a DAB Pressure Operated Direction valve on the end of the manifold. If this valve is of the same age as the rest of the system it should also be replaced as it is past its service life and is not up to current testing regulations.

DAB Fire Engineering Ltd. 16th August 2017

Annex ‘A’ CO2 Bottle Valve Serial No Comments 1 37017 No Leak 2 37039 No Leak 3 37012 No Leak 4 37009 No Leak 5 37032 No Leak 6 37049 No Leak 7 37001 Major leak on outlet in excess of 5 litres per minute 8 39938 No Leak 9 37037 No Leak 10 37020 No Leak 11 37045 No Leak 12 37052 No Leak 13 37033 No Leak 14 37034 No Leak 15 37042 No Leak 16 97050 No Leak 17 97056 Very minor leak on outlet. 18 97057 No Leak 19 97054 No Leak 20 73719 No Leak 21 97059 No Leak 22 37031 No Leak 23 37048 No Leak 24 37038 No Leak 25 37053 No Leak 26 37028 No Leak

Annex B IMO Circular - MSC.1/Circ.1318

INTERNATIONAL MARITIME ORGANIZATION 4 ALBERT EMBANKMENT LONDON SE1 7SR E Telephone: 020 7735 7611 Fax: 020 7587 3210 IMO Ref: T4/4.01 MSC.1/Circ.1318 11 June 2009 GUIDELINES FOR THE MAINTENANCE AND INSPECTIONS OF FIXED CARBON DIOXIDE FIRE-EXTINGUISHING SYSTEMS 1 The Committee, at its eighty-sixth session (27 May to 5 June 2009), having considered the proposal by the Sub-Committee on Fire Protection, at its fifty-third session, approved Guidelines for the maintenance and inspections of fixed carbon dioxide fire-extinguishing systems, as set out in the annex. 2 Member Governments are invited to apply the annexed Guidelines when inspecting fixed carbon dioxide fire-extinguishing systems on board all ships and bring them to the attention of ship designers, shipowners, equipment manufacturers, and other parties concerned.

I:\CIRC\MSC\01\1318.doc

MSC.1/Circ.1318 ANNEX GUIDELINES FOR THE MAINTENANCE AND INSPECTIONS OF FIXED CARBON DIOXIDE FIRE-EXTINGUISHING SYSTEMS 1 General These Guidelines provide the minimum recommended level of maintenance and inspections for fixed carbon dioxide fire-extinguishing systems on all ships, and are intended to demonstrate that the system is kept in good working order as specified in SOLAS regulation II-2/14.2.1.2. These Guidelines are intended to supplement the fire-extinguishing system manufacturer’s approved maintenance instructions. Certain maintenance procedures and inspections may be performed by competent crewmembers, while others should be performed by persons specially trained in the maintenance of such systems. The onboard maintenance plan should indicate which parts of the recommended inspections and maintenance should be completed by trained personnel. 2 Safety Whenever carbon dioxide fire-extinguishing systems are subjected to inspection or maintenance, strict safety precautions should be followed to prevent the possibility that individuals performing or witnessing the activities are placed at risk. Prior to performing any work, a safety plan should be developed to account for all personnel and establish an effective communications system between the inspection personnel and the on-duty crew. Measures to avoid accidental discharges such as locking or removing the operating arms from directional valves, or shutting and locking the system block valve should be taken as the initial procedure for the protection of personnel performing any maintenance or inspections. All personnel should be notified of the impending activities before work is begun. 3 Maintenance and inspection plan Fixed carbon dioxide fire-extinguishing systems should be kept in good working order and readily available for immediate use. Maintenance and inspections should be carried out in accordance with the ship’s maintenance plan having due regard to ensuring the reliability of the system. The onboard maintenance plan should be included in the ship’s safety management system and should be based on the system manufacturer’s recommendations including: .1 maintenance and inspection procedures and instructions; .2 required schedules for periodic maintenance and inspections; .3 listing of recommended spare parts; and .4 records of inspections and maintenance, including corrective actions taken to maintain the system in operable condition. I:\CIRC\MSC\01\1318.doc

1. 2

In addition, on low pressure systems the inspections should verify that:

MSC.1/Circ.1318 ANNEX Page 2 4 Monthly inspections 4.1 At least every 30 days a general visual inspection should be made of the overall system condition for obvious signs of damage, and should include verification that: .1 all stop valves are in the closed position; .2 all releasing controls are in the proper position and readily accessible for immediate use; .3 all discharge piping and pneumatic tubing is intact and has not been damaged; .4 all high pressure cylinders are in place and properly secured; and .5 the alarm devices are in place and do not appear damaged. .1 the pressure gauge is reading in the normal range; .2 the liquid level indicator is reading within the proper level; .3 the manually operated storage tank main service valve is secured in the open position; and .4 the vapour supply line valve is secured in the open position. 5 Annual

inspections The following minimum level of maintenance and inspections should be carried out in accordance with the system manufacturer’s instructions and safety precautions: .1 the boundaries of the protected space should be visually inspected to confirm that no modifications have been made to the enclosure that have created uncloseable openings that would render the system ineffective; .2 all storage containers should be visually inspected for any signs of damage, rust or loose mounting hardware. Cylinders that are leaking, corroded, dented or bulging should be hydrostatically retested or replaced; .3 system piping should be visually inspected to check for damage, loose supports and corrosion. Nozzles should be inspected to ensure they have not been obstructed by the storage of spare parts or a new installation of structure or machinery; .4 the manifold should be inspected to verify that all flexible discharge hoses and fittings are properly tightened; and I:\CIRC\MSC\01\1318.doc

MSC.1/Circ.1318 ANNEX Page 3 .5 all entrance doors to the protected space should close properly and should have warning signs, which indicate that the space is protected by a fixed carbon dioxide system and that personnel should evacuate immediately if the alarms sound. All remote releasing controls should be checked for clear operating instructions and indication as to the space served. 6 Minimum

recommended maintenance 6.1 At least biennially (intervals of 2 years ± 3 months) in passenger ships or at each intermediate, periodical or renewal survey* in cargo ships, the following maintenance should be carried out (to assist in carrying out the recommended maintenance, examples of service charts are set out in the appendix): .1 all high pressure cylinders and pilot cylinders should be weighed or have their contents verified by other reliable means to confirm that the available charge in each is above 90% of the nominal charge. Cylinders containing less than 90% of the nominal charge should be refilled. The liquid level of low pressure storage tanks should be checked to verify that the required amount of carbon dioxide to protect the largest hazard is available; .2 the hydrostatic test date of all storage containers should be checked. High pressure cylinders should be subjected to periodical tests at intervals not exceeding 10 years. At the 10-year inspection, at least 10% of the total number provided should be subjected to an internal inspection and hydrostatic test*. If one or more cylinders fail, a total of 50% of the onboard cylinders should be tested. If further cylinders fail, all cylinders should be tested. Flexible hoses should be replaced at the intervals recommended by the manufacturer and not exceeding every 10 years; and .3 the discharge piping and nozzles should be tested to verify that they are not blocked. The test should be performed by isolating the discharge piping from the system and flowing dry air or nitrogen from test cylinders or suitable means through the piping. 6.2 At least biennially (intervals of 2 years ± 3 months) in passenger ships or at each renewal survey in cargo ships, the following maintenance should be carried out by service technicians/specialists trained to standards accepted by the Administration: .1 where possible, all activating heads should be removed from the cylinder valves and tested for correct functioning by applying full working pressure through the pilot lines. In cases where this is not possible, pilot lines should be disconnected from the cylinder valves and blanked off or connected together and tested with full working pressure from the release station and checked for leakage. In both cases this should be carried out from one or more release stations when installed. * Refer to Survey guidelines under the Harmonized System of Survey and Certification, 2007 (resolution A.997(25)). ** Refer to standard ISO 6406 – Periodic inspection and testing of seamless steel gas cylinders. I:\CIRC\MSC\01\1318.doc

MSC.1/Circ.1318 ANNEX Page 4 If manual pull cables operate the remote release controls, they should be checked to verify the cables and corner pulleys are in good condition and freely move and do not require an excessive amount of travel to activate the system; .2 all cable components should be cleaned and adjusted as necessary, and the cable connectors should be properly tightened. If the remote release controls are operated by pneumatic pressure, the tubing should be checked for leakage, and the proper charge of the remote releasing station pilot gas cylinders should be verified. All controls and warning devices should function normally, and the time delay, if fitted should prevent the discharge of gas for the required time period; and .3 after completion of the work, the system should be returned to service. All releasing controls should be verified in the proper position and connected to the correct control valves. All pressure switch interlocks should be reset and returned to service. All stop valves should be in the closed position. I:\CIRC\MSC\01\1318.doc

MSC.1/Circ.1318 ANNEX Page 5 APPENDIX EXAMPLE SERVICE CHARTS HIGH PRESSURE CO2 SYSTEM Date: Name of ship/unit: IMO No.: Technical description No. Text Value 1 Manufacturer 2 Number of main cylinders 3 Main cylinders capacity (each) 4 Number of pilot cylinders 5 Pilot cylinder capacity (each) 6 Number of distribution lines 7 Oldest cylinder pressure test date 8 Protected space(s) 9 Date flexible hoses fitted/renewed Description of inspection/Tests No. Description Carried out Not carried out Not applicable Comment 1 Release controls and distribution valves secured to prevent accidental discharge 2 Contents in main cylinders checked by weighing 3 Contents in main cylinders checked by liquid level indicator 4 Contents of pilot cylinders checked 5 All cylinder valves visually inspected 6 All cylinder clamps and connections checked for tightness 7 Manifold visually inspected 8 Manifold tested for leakage, by applying dry working air 9 Main valve and distribution valves visually inspected 10 Main valve and distribution valves tested for operation 11 Time delay devices tested for correct setting* 12 Remote release system visually inspected 13 Remote release system tested 14 Servo tubing/pilot lines pressure tested at maximum working pressure and checked for leakages and blockage 15 Manual pull cables, pulleys, gang releases tested, serviced and tightened/adjusted as necessary 16 Release stations visually inspected 17 Warning alarms (audible/visual) tested 18 Fan stop tested* 19 10% of cylinders and pilot cylinder/s pressure tested every 10 years 20 Distribution lines and nozzles blown through, by applying dry working air 21 All doors, hinges and locks inspected* 22 All instruction and warning signs on installation inspected 23 All flexible hoses renewed and check valves in manifold visually inspected every 10 years 24 Release controls and distribution valves reconnected and system put back in service 25 Inspection date tags attached * If fitted as part of the CO2 system. I:\CIRC\MSC\01\1318.doc

MSC.1/Circ.1318 ANNEX Page 6 LOW PRESSURE CO2 SYSTEM Date: Name of ship/unit: IMO No.: Technical description No. Text Value 1 Manufacturer 2 No. of tanks 3 Tanks capacity (tonnes) 4 Number of pilot cylinders 5 Pilot cylinder capacity (each) 6 Number of distribution lines 7 Protected space(s) Description of inspection/Tests No. Description Carried out Not carried out Not applicable Comment 1 Tank main service valve closed and secured to prevent accidental discharge 2 Distribution valves verified closed 3 Check correct function of level indicator 4 Contents of CO2 tank checked by tank level indicator 5 Contents of CO2 tank checked by riser tube reading 6 Contents of CO2 tank checked by level control valve 7 Supports of tank inspected 8 Insulation on tank inspected 9 Safety valves of tank inspected 10 Safety valves of tank tested 11 Contents of pilot cylinders checked 12 Start/stop function of cooling compressors tested 13 All connected electrical alarms and indicators tested 14 Main manifold valve inspected 15 Main manifold valve tested 16 Distribution valves inspected 17 Distribution valves tested 18 Release stations inspected 19 Total flooding release mechanism inspected 20 Total flooding release mechanism tested 21 Time delay devices tested for correct setting* 22 Warning alarms tested 23 Fan stop tested* 24 Distribution lines and nozzles inspected 25 Distribution lines and nozzles tested 26 Distribution lines and nozzles blown through 27 All doors, hinges and locks inspected* 28 All instruction plates inspected 29 Tank main service valve reopened and secured open 30 System put back in service 31 Inspection date tags attached * If fitted as part of the CO2 system. I:\CIRC\MSC\01\1318.doc

Annex C Fire Industry Association guidance note

Guidance Note Guidance on the Periodic Testing of Transportable Gas Containers used in Fire Extinguishing Systems

Guidance on the Periodic Testing of Transportable Gas Containers • Version 2 • November 2015 • Tel: +44 (0)20 3166 5002 • www.fia.uk.com 2 of 6 Guidance on the Periodic Testing of Transportable Gas Containers used in Fire Extinguishing Systems BACKGROUND ........................................................................................................................................... 3 PERIODIC TESTING ................................................................................................................................. 4 REUSE OF CONTAINER VALVES ........................................................................................................ 5 REFERENCES ............................................................................................................................................... 6

Guidance on the Periodic Testing of Transportable Gas Containers • Version 2 • November 2015 • Tel: +44 (0)20 3166 5002 • www.fia.uk.com 3 of 6 BACKGROUND There is currently some confusion on the date at which extinguishing agent containers are required to be submitted for periodic testing under the Pressure Equipment Directive, the Transportable Pressure Equipment Directive and The Carriage of Dangerous Goods Regulations, as stated in the relevant Standards. This FIA guidance gives recommendations on how fire extinguishing system containers should be treated.

Guidance on the Periodic Testing of Transportable Gas Containers • Version 2 • November 2015 • Tel: +44 (0)20 3166 5002 • www.fia.uk.com 4 of 6 PERIODIC TESTING BS EN 1968:2002, applicable to seamless steel gas containers, BS EN 1802:2002, applicable to seamless aluminium alloy gas containers, BS EN 1803:2002, applicable to welded carbon steel gas containers, all state, in clause 3, that: • In the case of containers used for emergency purposes eg fire extinguishers, breathing apparatus, it is the responsibility of the person in possession (owner or user) to submit it for periodic inspection within the interval specified in Annex B. • In Annex B, the interval for containers used for CO2, Nitrogen and Argon and gas mixtures used as extinguishing agents is 10 years. In The Safe Use of Gas Containers, a guidance document issued by the Health and Safety Executive, under the heading ‘How to reduce risks’, sub heading ‘periodic examination’, makes references to Instruction P200 in ADR, as required by the The Carriage of Dangerous Goods Regulations. Instruction P200 also gives the periodic inspection intervals for CO2, Nitrogen and Argon containers as 10 years. Therefore, all fixed extinguishing agent containers (eg HFC-227ea (FM-200), FK 5-1-12 (Novec 1230), HFC-23 (FE-13), HFC-125 (FE-25), IG 541 (INERGEN), IG 01(Argon), IG 55 (ARGONITE), IG 100 (Nitrogen) and CO2 etc) must be inspected, re-tested and certified accordingly prior to the 10th anniversary of their manufacture or most recent test. Note 1: Containers found to be in service beyond the 10th anniversary of their manufacture or most recent test should be removed from service. Note 2: The previous dispensation, under BS 5430 Part 1 and Part 2, for fire protection containers to have retesting deferred to 20 years, no longer applies.

Guidance on the Periodic Testing of Transportable Gas Containers • Version 2 • November 2015 • Tel: +44 (0)20 3166 5002 • www.fia.uk.com 5 of 6 REUSE OF CONTAINER VALVES For safety reasons container valves should not be reused after removal from containers unless the following conditions are met: • The valve has been refurbished in accordance with manufacturer’s recommendations • The connection thread to the container and discharge hose are inspected to ensure they are within tolerance and undamaged Note: Where container valves use a taper thread form to marry to the container these are often found to be outside of the tolerances after a single fitment and removal.

Guidance on the Periodic Testing of Transportable Gas Containers • Version 2 • November 2015 • Tel: +44 (0)20 3166 5002 • www.fia.uk.com 6 of 6 Tudor House, Kingsway Business Park, Oldfield Road, Hampton, Middlesex TW12 2HD Tel: +44 (0)20 3166 5002 • www.fia.uk.com DISCLAIMER The information set out in this document is believed to be correct in the light of information currently available but it is not guaranteed and neither the Fire Industry Association nor its officers can accept any responsibility in respect of the contents or any events arising from use of the information contained within this document. REFERENCES • Safe Use of Gas Containers , Health And Safety Executive http://www.hse.gov.uk/press/2002/e02100.htm • The Carriage of Dangerous Goods and Use of Transportable Pressure Equipment Regulations 2009 (CDG 2009) Update Regulations http://www.hse.gov.uk/cdg/manual/adrcarriage.htm • ADR 2015, UNECE http://www.unece.org/trans/danger/publi/adr/adr2015/15ContentsE.html

Annex D UR Z17 - Procedural Requirements for Service Suppliers

Z17

Page 1 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) Procedural Requirements for Service Suppliers

CONTENTS

General

Objective

Definitions

Application

Procedure for Approval and Certification

Certification

Information Regarding Alterations to the Certified Service Operating System

Cancellation of Approval

Existing Approvals

Annex 1 Special Requirements for Various Categories of Service Suppliers

Revision notes:

Rev.4 added in Annex, Section 10, 11 & 12, with reference in 3.1.2.

Rev.5 clarified applicability to thickness measurement companies in 3.1.1 and Annex 1,

1. 1.

Rev.6 is to be uniformly implemented by IACS Societies and Associates from 1 January 2008.

Rev.7: Section 13 is added to Annex 1. This section applies to requests for recognition of test laboratories received on or after 1 January 2008.

Rev.8: Reference to IACS Recommendations 101 and 102 added.

Rev.9: Includes procedures for approval of test laboratories against res. MSC.288(87)

and applies to requests for recognition of approval of testing laboratories received on or

after 1 July 2013. However, deletion of the reference to PR 34 in Rev.9 applies from

1 July 2012.

Rev.10: Full document review carried out - Rev.10 is to be uniformly implemented by IACS Societies from 1 January 2016.

Rev.11: Full document review in order to verify the compliance with R.O. Code, IMO Res. MSC 349(92), carried out - Rev.11 is to be uniformly implemented by IACS Societies from 1 July 2016.

Rev.12: Section 15 of Annex 1 revised - Rev.12 is to be uniformly implemented by IACS Societies from 1 January 2018. Z17 (1997) (Rev.1 June 1999) (Rev.2 Nov 1999) (Rev.3 July 2002) (Rev.4 July 2003) (Rev.5 Feb 2004) (Rev.6 June 2007) (Rev.7 Nov 2007) (Rev.8 July 2008) (Rev.9 June 2012) (Corr.1 Aug 2012) (Rev.10 Jan 2015) (Rev.11 June 2015) (Rev.12 Nov 2016)

Z17

Page 2 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 1. General

1. 1

To approve firms providing services, such as measurements, tests or maintenance of safety systems and equipment, the Society is to apply procedures in this unified requirement and relevant Annex 1.

Objective 2.1 The objective of this procedure is to set minimum requirements for approval and certification of service suppliers and is applicable to both initial and renewal audits.

Definitions

Manufacturer: A company that manufactures equipment required to be periodically serviced and/or maintained.

Service Supplier (A Service Supplier or category of Service Supplier may be referred to here after simply as ‘supplier’): A person or company, not employed by an IACS Member, who at the request of an equipment manufacturer, shipyard, vessel's owner or other client acts in connection with inspection work and provides services for a ship or a mobile offshore drilling unit such as measurements, tests or maintenance of safety systems and equipment, the results of which are used by surveyors in making decisions affecting classification or statutory certification and services.

Agent: A Person or Company authorised to act for or to represent a Manufacturer or

approved/recognized service supplier.

Subsidiary: A Company partly or wholly owned by a Manufacturer or approved/recognized service supplier.

Subcontractor: A Person or Company providing services to a Manufacturer or approved/recognized service supplier, with a formal contract defining the assumption of the obligations of the service supplier.

Z17

Page 3 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 4. Application

1. 1

This procedure applies to the approval of the following categories of service suppliers:

1. 1.1 Statutory services

• Firms engaged in servicing inflatable liferafts, inflatable lifejackets, hydrostatic release

units, inflatable rescue boats, marine evacuation systems

• Firms engaged in inspections and testing of radio communication equipment

• Firms engaged in inspections and maintenance of self contained breathing apparatus

• Firms engaged in annual performance testing of Voyage Data Recorders (VDR) and

simplified Voyage Data Recorders (S-VDR)

• Firms engaged in sound pressure level measurements of public address and general

alarm systems on board ships

• Firms engaged in inspections of low location lighting systems using photo luminescent materials and evacuation guidance systems used as an alternative to low-location lighting systems

• Firms engaged in the servicing and maintenance of lifeboats, launching appliances, on-

load release gear and davit-launched liferaft automatic release hooks.

• Firms engaged in inspection, performance testing and maintenance of Automatic

Identification Systems (AIS)

1. 1.2 Classification and/or Statutory services:

• Firms engaged in thickness measurements on ships except

(1) non-ESP ships less than 500 gross tonnage and

(2) all fishing vessels.

• Firms carrying out in-water survey of ships and mobile offshore units

• Firms engaged in inspections and maintenance of fire extinguishing equipment and systems

• Firms engaged in tightness testing of closing appliances such as hatches, doors etc.

with ultrasonic equipment

• Firms engaged in measurements of noise level on board ships

• Firms engaged in examination of Ro-Ro ship’s bow, stern, side and inner doors

• Firms engaged in testing of coating systems in accordance with IMO Resolution

MSC.215(82), as amended, and IACS UI SC223 and/or MSC.288(87), as amended.

• Firms engaged in tightness testing of primary and secondary barriers of gas carriers

with membrane cargo containment systems for vessels in service

Z17

Page 4 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 4.1.3 Where a Society accepts work of a third party (eg., service supplier) approved by itself, the Society shall verify the performance of such services. For statutory service, the flag State may increase the scope of verification to be applied to these services. The process shall be defined within the Society’s quality management system. For the purpose of accountability to the flag State, the work performed by the third party (eg., service supplier) constitutes the work of the Society and shall be subject to the requirements incumbent upon the Society under the RO Code IMO MSC.349 (92) and MEPC.237(65).

1. 2

Where the results of the following service providers are used by a Surveyor of a Society in making decisions affecting classification services then that service provider must be approved and verified by that Society.

• Firms engaged in thickness measurements on ships except

(1) non-ESP ships less than 500 gross tonnage and

(2) all fishing vessels.

• Firms carrying out in-water survey of ships and mobile offshore units

• Firms engaged in tightness testing of closing appliances such as hatches, doors etc.

with ultrasonic equipment

1. 3

Where such services are used by Surveyors in making decisions affecting statutory certification and service, the firms are subject to approval and verification by the Society where the Society is so authorised by the relevant flag Administration (i.e. the flag of the ship on which the servicing is to be done or the service equipment is to be used). For such services the Society may accept approvals done by:

i. the flag Administration itself,

ii. duly authorized organizations acting on behalf of the flag Administration, or

iii. other organizations those are acceptable to the flag Administration (e.g. other

governments, etc.).

1. 4

Use of the approved service suppliers is not mandatory for the following services, unless instructed otherwise by the flag Administration with respect to statutory certification

• Firms engaged in inspections of low location lighting systems using photo luminescent materials and evacuation guidance systems used as an alternative to low-location lighting systems

• Firms engaged in sound pressure level measurements of public address and general

alarm systems on board ships

• Firms engaged in measurements of noise level onboard ships

• Firms engaged in testing of coating systems in accordance with IMO Resolution

MSC.215(82) as amended and IACS UI SC223 and/or MSC.288(87) as amended

• Firms engaged in examination of Ro-Ro ships bow, stern, side and inner doors

Z17

Page 5 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 4.5 Detailed requirements specific to the various categories of suppliers are given in Annex

1. National and/or international requirements may give additional requirements. References

to such national and/or international requirements are given in Annex 1.

Z17

Page 6 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 5. Procedure for Approval and Certification

1. 1

Submission of documents

1. 1.1 The following documents are to be submitted to the Society for review. General

requirements concerning suppliers are given in 5.2, and specific requirements as relevant, in Annex 1.

• Outline of company, e.g. organisation and management structure, including subsidiaries

to be included in the approval/certification

• List of nominated agents, subsidiaries and subcontractors

• Experience of the company in the specific service area

• For categories of Service Suppliers that require authorization from manufacturers,

manufacturer’s documentary evidence that the Service Supplier has been authorized or

licensed to service the particular makes and models of equipment for which approval is

sought shall be provided.

• List of operators/technicians/inspectors documenting training and experience within the

relevant service area, and qualifications according to recognised national, international

or industry standards, as relevant

• Description of equipment used for the particular service for which approval is sought

• A guide for operators of such equipment

• Training programmes for operators/technicians/inspectors

• Check lists and record formats for recording results of the services referred to in

Annex 1

• Quality Manual and/or documented procedures covering requirements in 5.5

• Documented procedures for communication with the crew prior to commencing work, so

that it is safe to decommission the equipment being maintained, and to provide a safe

system of work in place

• Evidence of approval/acceptance by other bodies, if any

• Information on the other activities which may present a conflict of interest

• Record of customer claims and of corrective actions requested by certification bodies

1. 2

General requirements:

1. 2.1 Extent of Approval – The supplier shall demonstrate, as required by 5.2.2 – 5.2.11, that

it has the competence and control needed to perform the services for which approval is sought.

1. 2.2 Training of personnel – The supplier is responsible for the qualification and training of

its personnel to a recognised national, international or industry standard as applicable. Where such standards do not exist, the supplier is to define standards for the training and qualification of its personnel relevant to the functions each is authorised to perform. The

Z17

Page 7 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) personnel shall also have adequate experience and be familiar with the operation of any necessary equipment. Operators/technicians/inspectors shall have had a minimum of one year tutored on-the-job training. Where it is not possible to perform internal training, a program of external training may be considered as acceptable.

1. 2.3 Supervision – The supplier shall provide supervision for all services provided. The

responsible supervisor shall have had a minimum of two years of experience as an operator/technician/inspector within the activity for which the supplier is approved. For a supplier consisting of one person, that person shall meet the requirements of a supervisor.

1. 2.4 Personnel records – The supplier shall keep records of the approved

operators/technicians/inspectors. The record shall contain information on age, formal education, training and experience for the services for which they are approved.

1. 2.5 Equipment and facilities – The supplier shall have the necessary equipment and

facilities for the service to be supplied. A record of the equipment used shall be kept and available. The record shall contain information on maintenance and results of calibration and verifications. The Society shall assess and record the validity of previous measuring results when the equipment is found not to conform to requirements. The Society shall take appropriate action on the equipment affected.

1. 2.6 Control of data:

When computers are used for the acquisition, processing, recording, reporting, storage, measurement assessment and monitoring of data, the ability of computer software to satisfy the intended application shall be documented and confirmed by the service supplier. This shall be undertaken prior to initial use and reconfirmed as necessary. Note: Commercial off-the-shelf software (e.g. wordprocessing, database and statistical programmes) in general use within their designed application range may be considered to be sufficiently validated and do not require any subsequent confirmation.

1. 2.7 Where several servicing stations are owned by a given company, each station is to be

assessed and approved except as specified in 5.5.3

1. 2.8 Procedures – The supplier shall have documented work procedures covering all

services supplied.

1. 2.9 Subcontractors – The supplier shall give information of agreements and arrangements if

any parts of the services provided are subcontracted. Particular emphasis shall be given to quality management by the supplier in following-up such subcontracts. Subcontractors providing anything other than equipment shall also meet the requirements of sections 5.2 and 5.5.

1. 2.10 Verification – The supplier shall verify that the services provided are carried out in

accordance with approved procedures.

1. 2.11 Reporting – The report shall be prepared in a form acceptable to the Society. The

report should detail the results of inspections, measurements, tests, maintenance and/or repairs carried out. Special guidelines may be given in Annex 1. The report shall include a copy of the Certificate of Approval.

1. 2.12 Documented procedures and instructions should be available for the recording of

damages and defects found during inspection, servicing and repair work. This documentation is to be made available upon request.

Z17

Page 8 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 5.3 Auditing of the Supplier – Upon reviewing the submitted documents with satisfactory result, the supplier is audited in order to ascertain that the supplier is duly organised and managed in accordance with the submitted documents, and that it is considered capable of conducting the services for which approval/certification is sought.

1. 4

Certification is conditional on a practical demonstration of the performance of the specific service as well as satisfactory reporting being carried out. At renewal audits, evidence of performance, verified by class surveyor, since the previous audit is sufficient to satisfy this requirement.

1. 5

Quality System

1. 5.1 The supplier shall have a documented system covering at least the following:

• code of conduct for the relevant activity

• maintenance and calibration of equipment

• training programmes for operators/technicians/inspectors

• supervision and verification to ensure compliance with operational procedures

• recording and reporting of information

• quality management of subsidiaries, agents and subcontractors

• job preparation

• periodic review of work process procedures, complaints, corrective actions, and

issuance, maintenance and control of documents

1. 5.2 A documented Quality system complying with the most current version of ISO 9000

series and including the above items, would be considered acceptable.

1. 5.3 If a manufacturer of equipment (and/or its service supplier) applies to a Society for

inclusion of its nominated agents and/or subsidiaries in the approval, then it must have implemented a quality system certified in accordance with the most current version of ISO 9000 series. The quality system must contain effective controls of the manufacturer’s (and/or service supplier’s) agents and/or subsidiaries. The nominated agents/subsidiaries must also have in place an equally effective quality system complying with the most current version of ISO 9000 series. Such approvals shall be based upon an evaluation of the quality system implemented by the parent company against the most current version of ISO 9000 series. The Society may require follow-up audits on such agents or subsidiaries against the most current version of ISO 9000 series to confirm adherence to this quality system.

1. 6

Service Suppliers Relations with the Equipment Manufacturer

1. 6.1 A company which works as a service station for manufacturer(s) of equipment (and as a

service supplier in this field), shall be assessed by the manufacturer(s) and nominated as their agent. The manufacturer shall ensure that appropriate instruction manuals, material etc. are available for the agent as well as proper training of the agent’s technicians. Such suppliers shall be approved either on a case by case basis, or in accordance with 5.5.3.

Z17

Page 9 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 6. Certification

1. 1

Upon satisfactory completion of both the audit of the supplier and the demonstration test, as applicable, the Society may issue a Certificate of Approval stating that the supplier’s service operation system has been found to be satisfactory and that the results of services performed in accordance with that system may be accepted and utilised by the Society’s Surveyors in making decisions affecting classification or statutory certification, as relevant. The Certificate shall clearly state the type and scope of services and any limitations or restrictions imposed including type of equipment and/or names of Manufacturers of equipment where this is a limiting restraint. The supplier may also be included in the Society’s record of approved service suppliers.

1. 2

Renewal or endorsement of the Certificate is to be made at intervals not exceeding five (5) years by verification through audits that approved conditions are maintained or, where applicable, on expiry of the supplier’s approval received from an equipment Manufacturer, whichever comes first. In the latter case, the Society is to be informed in due course by the Service Supplier. Individual Societies may require renewal or endorsement of the Certificate at intervals shorter than five (5) years and may require intermediate audits. For firms engaged in thickness measurements, renewal/endorsement of the Certificate is to be made at intervals not exceeding 3 years by verification that original conditions are maintained.

Z17

Page 10 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 7. Information Regarding Alterations to the Certified Service Operating System

1. 1

When any alteration to the certified service operating system of the supplier is made, such alteration is to be immediately informed to the Society. Re-audit may be required when deemed necessary by the Society.

Cancellation of Approval 8.1 The Society reserves the right to cancel the approval and to inform the IACS Members accordingly (For Firms engaged in thickness measurements refer to PR23).

1. 2

Approval may be cancelled in the following cases:

1. 2.1 Where the service was improperly carried out or the results were improperly reported.

1. 2.2 Where a Surveyor finds deficiencies in the approval service operating system of the

supplier and appropriate corrective action is not taken.

1. 2.3 Where alterations have been made to the Company’s Quality System relevant to the

service supplier certificates, without written notification to the Society.

1. 2.4 Where the intermediate audit, if requested as per 6.2, has not been carried out.

1. 2.5 Where wilful acts or omissions are ascertained.

1. 2.6 Where any deliberate misrepresentation has been made by the Service Supplier.

1. 3

A supplier whose approval was cancelled, may apply for re-approval provided it has corrected the non-conformities which resulted in cancellation, and the Society is able to confirm it has effectively implemented the corrective action.

1. 4

Expiration or cancellation of the Supplier’s parent company approval automatically invalidates approval of all agents and subsidiaries if these are certified according to 5.5.3.

Existing Approvals Approvals for the categories of service suppliers granted before the date of implementation of UR Z17 by a society may remain valid as stated in the respective certificates for a period up to but not exceeding 3 years. Renewals of such certificates must be carried out in accordance with UR Z17.

Z17

Page 11 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) ANNEX 1

Special Requirements for Various Categories of Service Suppliers

• Firms engaged in thickness measurements on ships

1. 1

Extent of engagement – Thickness measurement of structural material of ships except

(1) non-ESP ships less than 500 gross tonnage and

(2) all fishing vessels.

1. 2

Supervisor – The responsible supervisor shall be qualified according to a recognised national or international industrial NDT standard (e.g. EN 473 level II as amended or ISO 9712 level II as amended).

1. 3

Operators – The operators carrying out the measurements shall be certified to a recognised national or international industrial standard (e.g. EN 473 level I as amended or ISO 9712 level I as amended) and shall have adequate knowledge of ship structures sufficient to elect a representative position for each measurement.

1. 4

Equipment – On coated surfaces, instruments using pulsed echo technique (either with oscilloscope or digital instruments using multiple echoes, single crystal technique) are required. Single echo instruments may be used on uncoated surfaces, which have been cleaned and ground.

1. 5

Procedures – Documented work procedures are at least to contain information on inspection preparation, selection and identification of test locations, surface preparation, protective coating preservation, calibration checks, and report preparation and content.

1. 6

Reporting – The report shall be based on the guidelines given in UR Z7, UR Z7.1, UR Z7.2, Z10.1, Z10.2, Z10.3, Z10.4 and Z10.5, as relevant.

1. 7

Verification – The supplier must have the Surveyor’s verification of each separate job, documented in the report by the attending Surveyor(s) signature.

Z17

Page 12 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in tightness testing of closing appliances such as hatches, doors

etc. with ultrasonic equipment

1. 1

Extent of engagement – Ultrasonic tightness testing of closing appliances such as hatches, doors etc.

1. 2

Operators – The operator is to have the following qualifications:

• Have knowledge of different closing appliances such as hatches, doors etc. including

their design, functioning and sealing features

• Have experience with the operation and maintenance of different closing appliances

such as hatches, doors etc.

• Be able to document theoretical and practical training onboard in using the ultrasonic

equipment specified

1. 3

It shall be demonstrated to the Surveyor that the ultrasonic equipment is fit for the purpose of detecting leakages in closing appliances.

1. 4

Procedures – The supplier shall have documented work procedures which shall include the manual for the ultrasonic equipment specified, its adjustment, its maintenance, its operation and approval criteria.

Z17

Page 13 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms carrying out in-water survey of ships and mobile offshore units

1. 1

Extent of engagement – In-water survey of ships and mobile offshore units by diver or Remote Operated Vehicle.

1. 2

Training of personnel – The supplier is responsible for the qualification of its divers and the diving equipment utilised when carrying out inspection. Knowledge of the following shall be documented:

• Ship’s underwater structure and appendages, tail shaft, propeller, rudder and its

bearings, etc.

• Non-destructive testing in accordance with a recognised national or international

industrial NDT standard. This requirement only applies if an in-water survey company

performs non-destructive testing

• Bearing clearance measurements on rudders and tail shaft

• Under-water video monitoring with TV-monitors on deck, as well as still picture work

• Operation of under-water communication system

• Special equipment and tools e.g. hull cleaners, grinders, cutters, etc.

1. 3

A plan for training of personnel in the reporting system, minimum Rule requirements for relevant ship types, ship’s underwater structure, measuring of bearing clearances, the recognition of corrosion damage, buckling and deteriorated coatings, etc. shall be included.

1. 4

Supervisor – The supervisor shall be qualified according to the supplier’s general requirements and shall have a minimum of two years’ experience as a diver carrying out inspection.

1. 5

Divers carrying out inspection – The diver carrying out the inspection shall have had at least one year’s experience as an assistant diver carrying out inspections (including participation in a minimum of 10 different assignments).

1. 6

Equipment – The following shall be available:

• Closed circuit colour television with sufficient illumination equipment

• Two-way communication between diver and surface staff

• Video recording device connected to the closed circuit television

• Still photography camera

• Equipment for carrying out thickness gauging, non-destructive testing and

measurements, e.g. clearances, indents, etc., as relevant to the work to be performed

• Equipment for cleaning of the hull

• Remote Operated Vehicle, if applicable

Z17

Page 14 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 3.7 Procedures and guidelines – The supplier shall have documented operational procedures and guidelines for how to carry out the inspection and how to handle the equipment. These shall include:

• Two-way communication between diver and surface

• Video recording and closed circuit television operation

• Guidance of the diver along the hull to provide complete coverage of the parts to be

inspected

• Guidance for the operation and maintenance of the Remote Operated Vehicle, if

applicable

1. 8

Verification – The supplier must have the Surveyor’s verification of each separate job, documented in the report by the attending Surveyor(s) signature.

Z17

Page 15 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in inspection and maintenance of fire extinguishing equipment and

systems

1. 1

Extent of engagement – inspections and maintenance of fire-extinguishing equipment and systems such as fixed fire extinguishing systems, portable fire extinguishers and fire detection and alarm systems.

1. 2

Extent of Approval

• Service Suppliers are to have professional knowledge of fire theory, fire-fighting and fire-extinguishing appliances sufficient to carry out the maintenance and/or inspections, and to make the necessary evaluations of the condition of the equipment

• In demonstrating professional knowledge, Service Suppliers are to have an

understanding of the various types of fires and the extinguishing media to be used on

them

• For fixed fire-extinguishing systems, Service Suppliers are to demonstrate an

understanding of the principles involved with gas, foam, deluge, sprinkler and water-

mist systems, as relevant for the approval being sought

1. 3

Procedures

• Service Suppliers are to have documented procedures and instructions on how to carry

out the servicing of the equipment and/or system. These are to either contain or make

reference to the Manufacturer’s servicing manuals, servicing bulletins, instructions and

training manuals, as appropriate, and to international requirements

• Additionally they are to make reference to any requirements (e.g. what markings should

be appended to the equipment/system)

1. 4

Reference Documents - The Service Supplier is to have access to the following

documents:

• Manufacturer’s servicing manuals, servicing bulletins, instructions and training manuals,

as appropriate

• Type Approval certificates showing any conditions that may be appropriate during the

servicing and/or maintenance of fire-extinguishing equipment and systems

• SOLAS, MSC.1/Circular.1318 (Guidelines for the Maintenance and Inspections of Fixed

Carbon Dioxide Fire-Extinguishing Systems), International Code for Fire Safety

Systems (FSS Code), ISO 6406 (Periodic inspection and testing of seamless steel gas

cylinders), and any documentation specified in the authorisation or license from the

equipment manufacturer

• MSC/Circ.670 (Guidelines for the Performance and Testing Criteria and Surveys of

High Expansion Foam Concentrates for fixed Fire-Extinguishing Systems)

• MSC/Circ.798 (Guidelines for the Performance and Testing Criteria and Surveys of

Medium Expansion Foam Concentrates for fixed Fire-Extinguishing Systems)

• MSC/Circ.799 ((Guidelines for the Performance and Testing Criteria and Surveys of

Expansion Foam Concentrates for fixed Fire-Extinguishing Systems of Chemical

Tankers)

Z17

Page 16 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) • MSC.1/Circ.1312 (Revised Guidelines for the Performance and Testing Criteria and

Surveys of Foam Concentrates for fixed Fire-Extinguishing Systems as corrected by

MSC/Circ.1312/Corr.1)

• MSC.1/Circ.1432 (Revised Guidelines for the maintenance and Inspection of Fire

Protection Systems and Appliances)

• IMO Res. A. 951(23) – Improved guidelines for marine portable fire extinguishers

• MSC.1/Circ.1370 – Guidelines for the design, construction and testing of fixed

hydrocarbon gas detection systems

• Guidelines adopted by IMO for fire extinguishing equipment and systems specifically

intended for service by service suppliers

1. 5

Equipment and Facilities

1. 5.1 General Requirements

• If Service Suppliers undertake shore-based inspecting and maintenance, they should

maintain and implement procedures for workshop cleanliness, ventilation and

arrangement, with due cognisance of the spares and extinguishing media being stored,

to ensure safe and effective working procedures

• Service Suppliers undertaking inspecting and maintenance of equipment and systems

onboard are to provide the appropriate facilities to either complete the work onboard or

remove the necessary items to their workshops

1. 5.2 Equipment

Sufficient and appropriate spares and tools are to be available as applicable, which should include:

• Various scales to weigh items

• Means to hydrostatically pressure test components/systems/storage bottles

• Liquid/gas, flow meters, as appropriate

• Pressure gauges or manometers

• In the cases of foam concentrates and portable fire-extinguishers, chemical analysis

equipment and a testing bay, respectively; and

• Specific equipment/spares as may be specified by Manufacturer

• Level measuring equipment for bottles

• Recharging facilities for pressurized bottles, extinguishers and cartridges

Z17

Page 17 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in servicing inflatable liferafts, inflatable lifejackets, hydrostatic

release units, inflatable rescue boats, marine evacuation systems

1. 1

Extent of engagement

• Servicing of inflatable liferafts, inflatable lifejackets, hydrostatic release units and/or

inflatable rescue boats

• Servicing of marine evacuation systems

1. 2

Equipment and facilities – IMO Res. A.761(18) as amended by MSC.55(66) gives recommendations on conditions for the approval of servicing stations for inflatable liferafts which shall be observed as relevant. Where inflatable liferafts are subject to extended service intervals, MSC.1/Circ.1328 should also be followed.

1. 3

Procedures and instructions – The supplier shall have documented procedures and instructions for how to carry out service of equipment. Where inflatable liferafts are subject to extended service intervals in accordance with the requirements of SOLAS Regulation III/20.8.3, MSC.1/Circ.1328 should be followed in addition to Resolution A.761(18) as amended by MSC.55(66).

1. 4

The supplier shall provide evidence that it has been authorised or licensed to service the particular makes and models of equipment for which approval is sought by the equipment’s manufacturer.

1. 5

Reference Documents - The Service Supplier is to have access to the following documents:

• IMO - Resolution A.761(18) - Recommendation on Conditions for the Approval of

Servicing Stations for Inflatable Liferafts - (adopted on 4 November 1993), amended by

Resolution MSC.55(66)

• IMO - Resolution MSC.55(66)

• IMO – MSC.1/Circ.1328 – Guidelines for the Approval of Inflatable Liferafts Subject to

Extended Service Intervals Not Exceeding 30 Months

• Manufacturer’s servicing manuals, servicing bulletins, instructions and training manuals,

as appropriate

• Type Approval certificates, showing any conditions that may be appropriate during the

servicing and/or maintenance of inflatable liferafts, inflatable rescue boats, inflatable

lifejackets, and hydrostatic release units

• LSA code/Chap.IV, 1995 SOLAS Conference Resolution 4 regarding marine evacuation

systems

Z17

Page 18 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in inspections and testing of radio communication equipment

1. 1

Extent of engagement

• Surveys, inspection, testing, and/or measurement of radio equipment aboard ships or

mobile offshore units for compliance with SOLAS regulations

• Annual testing of 406 MHz satellite EPIRBs for compliance with SOLAS Regulation

IV/15.9

• The principles of this section also apply to Service Suppliers involved in inspection,

performance testing and maintenance of Automatic Identification Systems (AIS).The

Service Supplier is to be familiar with the equipment with which it will be involved, such

as being a service agent for the equipment manufacturer

1. 2

Reference documents – The supplier shall have access to the following documents:

• SOLAS 1974 as amended

• IMO Res. A.789(19): Specification on the survey and certification functions of

recognised organisations acting on behalf of the administration

• MSC/Circ.1040/Rev.1 – Guidelines on Annual Testing of 406 MHZ Satellite EPIRBs

• MSC.1/Circ.1252 – Guidelines on Annual Testing of the Automatic Identification System

(AIS)

• SN/Circ.227, SN/Circ.227/Corr.1 and 245 – Guidelines for the Installation of a

Shipborne Automatic Identification System (AIS) and amendments thereto

• ITU Radio Regulations

• IMO Performance Standards for the equipment for which the Service Supplier is

approved

• Flag State Administration requirements

• Relevant parts, if any, of the Society’s Rules and Guidelines

1. 3

Supervisor – The supervisor shall have a minimum two years education from a technical school, experience as inspector, and should preferably hold a General Operator’s Certificate (GOC) or a GMDSS Radioelectronic Certificate (REC), recognised by the ITU, to operate or test radio transmitters. He should be aware of any local conditions for radio signal propagation, of regional radio stations and their facilities, and of the GMDSS infrastructure.

1. 4

Radio inspector – The inspector carrying out the inspection shall have passed the internal training of the supplier in Radiotelephony, GMDSS, and initial and renewal surveys, as applicable. The inspector shall also have at least one year’s technical school training or as alternative hold evidence that he followed a technical course approved by the relevant Administration, at least one year’s experience as an assistant radio inspector and should preferably hold an appropriate National Radio Operators Certificate, recognised by the ITU, such as a GMDSS General Operator’s Certificate (GOC) or a GMDSS Radioelectronic Certificate (REC). He should be aware of any local conditions for radio signal propagation, of regional radio stations and their facilities, and of the GMDSS infrastructure.

Z17

Page 19 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 6.5 Equipment and facilities

1. 5.1 The supplier shall have the major and auxiliary equipment required for correctly

performing the inspection. A record of the equipment used shall be kept. The record shall contain information on manufacturer and type of equipment, and a log of maintenance and calibrations.

1. 5.2 A standard which is relevant to the radio equipment to be tested shall be available for

the equipment and shall be cited in the inspection report.

1. 5.3 For equipment employing software in conjunction with the testing/examination, this

software shall be fully described and verified.

1. 6

Minimum required instruments:

• Equipment for measuring frequency, voltage, current and resistance

• Equipment for measuring output and reflect effect on VHF and MF/HF

• Equipment for measuring modulation on MF/HF and VHF (AM, FM, PM)

• Acid tester for checking specific gravity of lead batteries

• Tester for checking of correct output from Free-Float Satellite EPIRB

• Equipment for testing the performance of Automatic Identification Systems (AIS)

1. 7

Procedures and instructions – The supplier shall have documented procedures and instructions for how to carry out testing and examination of radio equipment. Procedures and instructions for operating each item of the testing/inspection equipment shall also be kept and be available at all times.

Z17

Page 20 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in inspections and maintenance of self contained breathing

apparatus

1. 1

Extent of Engagement - inspections and maintenance of self-contained breathing apparatus, Emergency Escape Breathing Devices (EEBD)

1. 2

Extent of Approval

• The supplier shall document and demonstrate that it has knowledge of the equipment

and systems sufficient to carry out the inspections and testing of self-contained

breathing apparatus to identify standards and to make the necessary evaluation of the

condition of the equipment

• In demonstrating professional knowledge, Service Suppliers are to have an

understanding of the operational requirements involved with self-contained breathing

apparatus and how these are to be maintained

• Additionally, Service Suppliers are to demonstrate the necessary safety requirements

applicable to such equipment

1. 3

Procedures

• Service Suppliers are to have documented procedures and instructions on how to carry

out the servicing of the equipment and/or system. These are to either contain or make

reference to the Manufacturer’s servicing manuals, servicing bulletins, instructions and

training manuals, as appropriate

• Additionally they are to make reference to any requirements (e.g. what markings should

be appended to the equipment/system) and how they should be applied

1. 4

Reference Documents - The Service Supplier is to have access to the following documents:

• Manufacturers’ servicing manuals, servicing bulletins, instructions and training manuals,

as appropriate

• Type Approval certificates showing any conditions which may be appropriate during the

servicing and/or maintenance of self-contained breathing apparatus

1. 5

Equipment and Facilities

1. 5.1 General Requirements

• If Service Suppliers undertake shore-based inspecting and maintenance, they should

maintain and implement procedures for workshop cleanliness, ventilation and

arrangement, with due cognisance of the spares and pressurised bottles being stored,

to ensure safe and effective working procedures

• Service Suppliers undertaking inspecting and maintenance of equipment and systems

onboard are to provide the appropriate facilities to either complete the work onboard or

remove the necessary items to their workshops

Z17

Page 21 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 7.5.2 Equipment

• Sufficient and appropriate spares and tools are to be available for repair, maintenance

and servicing of self-contained breathing apparatus in accordance with the

requirements of the Manufacturers

• These are to include, as required by the self-contained breathing apparatus equipment

and/or systems:

• Various scales to weigh items

• Means to hydrostatically pressure test components/systems/storage bottles

• Flow meters; and

• Pressure gauges or manometers

• Equipment for checking air quality

• Recharging facilities for breathing apparatus

Z17

Page 22 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in examination of Ro-Ro ships bow, stern, side and inner doors

1. 1

Extent of engagement - inspection of securing and locking devices, hydraulic operating system, electric control system for the hydraulics, electric indicator systems, and supporting, securing and locking devices and tightness testing.

1. 2

The supplier is to be certified to the most current version of ISO 9000 series.

1. 3

Supervision - In addition to 5.2.3, the requirement to have had a minimum of two years experience as operator/technician/inspector within the activity, a Supervisor is to have a minimum two years related education from a technical school.

1. 4

Training of personnel - Operators carrying out Non-Destructive Examination (NDE) are to be qualified to a recognised National or International Standard for the methods used.

1. 5

Reference documents - The supplier shall have access to the following reference documents:

• IMO - International Convention on the Safety of Life at Sea (SOLAS) 74/78, as

amended

• ISO 9002:1994 - Quality systems - Model for quality assurance in production,

installation and servicing

• UR Z24 - Survey Requirements for Shell and Inner Doors of Ro-Ro ships, or its

equivalent, by the relevant class society

1. 6

Required Equipment

1. 6.1 For Inspection of Supporting Securing and Locking Devices, Hinges and Bearings:

• Equipment for measuring clearances (i.e. feeler gauges, vernier calipers, micrometers).

• Non-destructive examination (i.e. dye penetrant, magnetic particle inspection)

1. 6.2 For Tightness Testing:

• Ultrasonic leak detector or equivalent

1. 6.3 For Inspection of Hydraulic Operating System:

• Pressure gauges

• Particle counter for analysing the quality of hydraulic fluid

1. 6.4 For Inspection of Electric Control System and Indication System:

• Digital multi-meter

• Earth fault detector

1. 7

Procedures and Instructions

1. 7.1 The supplier shall have access to drawings and documents, including the Operating

and Inspection Manual.

Z17

Page 23 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 8.7.2 The supplier shall have access to the service history of the doors.

1. 7.3 The supplier should use, complete and sign a checklist which has been found

acceptable by the classification society.

Z17

Page 24 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in annual performance testing of Voyage Data Recorders (VDR) and

simplified Voyage Data Recorders (S-VDR)

1. 1

Extent of engagement – Testing and servicing of Voyage Data Recorders (VDR) and Simplified Voyage Data Recorders (S-VDR) in accordance with SOLAS Chapter V Regulation 18.8 and IMO - MSC.1/Circular.1222 - Guidelines on Annual Testing of Voyage Data Recorders (VDR) and Simplified Voyage Data Recorders (S-VDR), as applicable.

1. 2

Extent of Approval

1. 2.1 The supplier shall provide evidence that he has been authorised or licensed by the

equipment’s manufacturer to service the particular makes and models of equipment for which approval is sought.

1. 2.2 Where the Service Supplier is also the Manufacturer of the Voyage Data Recorder

(VDR) or Simplified Voyage Data Recorder (S-VDR) and has elected to apply IMO - MSC.1/Circular.1222 - Guidelines on Annual Testing of Voyage Data Recorders (VDR) and Simplified Voyage Data Recorders (S-VDR) in its entirety for the purpose of acting as a Service Supplier engaged in annual performance testing, the following is to apply:

• The Manufacturer is responsible for appointing Manufacturer’s Authorised Service

Stations to carry out annual performance testing

• The Manufacturer is required to be an Approved Service Supplier and is to satisfy the

requirements for Service Suppliers engaged in annual performance testing of Voyage

Data Recorders (VDR) and Simplified Voyage Data Recorders (S-VDR), as applicable

• The Manufacturer’s Authorised Service Station is not required to be an Approved

Service Supplier

• The Manufacturer is to demonstrate that IMO - MSC.1/Circular.1222 - Guidelines on

Annual Testing of Voyage Data Recorders (VDR) and Simplified Voyage Data

Recorders (S-VDR) is applied in its entirety

1. 3

Procedures

1. 3.1 The Service Supplier shall have documented procedures and instructions.

1. 3.2 Where the Service Supplier is also the Manufacturer of the Voyage Data Recorder

(VDR) or Simplified Voyage Data Recorder (S-VDR) and has selected to apply IMO - MSC.1/Circular.1222 - Guidelines on Annual Testing of Voyage Data Recorders (VDR) and Simplified Voyage Data Recorders (S-VDR) in its entirety for the purpose of acting as a Service Supplier engaged in annual performance testing, the following is to apply:

• The Manufacturer is to have documented procedures for the assessment and

authorisation of Manufacturer’s Authorised Service Stations who carry out annual

performance testing

• The Manufacturer is to have documented procedures for the review of Manufacturer’s

Authorised Service Stations annual performance test reports, analysis of the Voyage

Data Recorder (VDR) and Simplified Voyage Data Recorder (S-VDR) 12 hour log and

the issue of annual performance test certificates to the Owner/Operator

Z17

Page 25 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) • The Manufacturer is to maintain a list of Manufacturer’s Authorised Service Stations

that can be accessed (by any available means, e.g. via a nominated contact point or

from the Manufacturer’s website) upon request

1. 4

Reference Documents

1. 4.1 The Service Supplier is to have access to the following documents:

• IMO - International Convention on the Safety of Life at Sea (SOLAS), 74/78, Ch V, Reg

1. 8. – Approval, surveys and performance standards of navigational systems and

equipment and voyage data recorder

• IMO - MSC.1/Circular.1222 - Guidelines on Annual Testing of Voyage Data Recorders

(VDR) and Simplified Voyage Data Recorders (S-VDR) - (11 December 2006)

• IMO - Resolution A.861(20) (adopted on 27 November 1997) as amended by IMO

Resolution MSC.214(81) and revised by IMO Resolution MSC.333(90) - Performance

Standards for Shipborne Voyage Data Recorders (VDRs)

• IMO - Resolution MSC.163(78) - Performance Standards for Shipborne Simplified

Voyage Data Recorders (S-VDRs) - (adopted on 17 May 2004), as amended by IMO

Resolution 214(81)

1. 4.2 The Service Supplier is to have access to applicable industry performance standards,

e.g.:

• IEC 61996 - Maritime navigation and radiocommunication equipment and systems -

Shipborne voyage data recorder (VDR)

• IEC 61996–2 - Maritime navigation and radio communication equipment and systems –

Shipborne voyage data recorder (VDR) – Part 2: Simplified voyage data recorded (S-

VDR) – Performance requirements, method of testing and required test results

1. 4.3 The Service Supplier is also to have access to any documentation specified in the

authorisation or license from the equipment manufacturer.

1. 5

Equipment and Facilities - In addition, the Service Supplier shall have equipment as specified in the authorisation or license from the equipment Manufacturer.

1. 6

Reporting - Test Report

1. 6.1 The Service Supplier shall issue a certificate of compliance as specified in the

International Convention on Safety of Life at Sea (SOLAS 1974), as amended, Ch V, Reg 18.8.

1. 6.2 Annual Performance Test of VDR and S-VDR should be recorded in the form of the

model test report given in the Appendix to MSC.1/Circular.1222, signed and stamped by the Service Supplier and attached to the annual performance test certificate.

1. 6.3 Where the Service Supplier is also the Manufacturer of the Voyage Data Recorder

(VDR) or Simplified Voyage Data Recorder (S-VDR) and has selected to apply IMO - MSC.1/Circular.1222 - Guidelines on Annual Testing of Voyage Data Recorders (VDR) and Simplified Voyage Data Recorders (S-VDR) in its entirety for the purpose of acting as a Service Supplier engaged in annual performance testing, the Manufacturer is to make arrangements for the following:

Z17

Page 26 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) • Review of the Manufacturer’s Authorised Service Station annual performance test

report

• Analysis of the recorder’s 12 hour log

• Checking of the master record/database for the recorder

1. 6.4 Issue of the annual performance test certificate to the Owner/Operator within 45 days of

completion of the annual performance test.

Z17

Page 27 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in inspections of low location lighting systems using photo

luminescent materials and evacuation guidance systems used as an alternative to low- location lighting systems

1. 1 Extent of engagement – Luminance measurements on board ships of low location

lighting systems using photo luminescent materials.

1. 2 Operators – The operator is to have the following qualifications:

• Have adequate knowledge of the applicable international requirements (namely SOLAS

reg. II-2/13.3.2.5, IMO Res. A.752(18) - Guidelines for the Evaluation, Testing and

Application of Low-Location Lighting on Passenger Ships, ISO 15370-2010, FSSS

Code Chapter 11)

• Be able to document theoretical and practical training onboard in using equipment

specified

1. 3 Equipment – The measuring instrument shall incorporate a fast-response photometer

head with CIE (International Commission on Illumination) photopic correction and have a measurement range of at least 10-4 cd/m2 to 10 cd/m2.

1. 4 Procedures – Documented work procedures are at least to contain information on

inspection preparation, selection and identification of test locations.

1. 5 Reporting – The report shall conform to Annex C of ISO 15370-2010.

1. 6 Verification – The supplier must have the Surveyor’s verification of each separate job,

documented in the report by the attending Surveyor’s signature.

1. 7 Reference Documents

The Service Supplier is to have access to the following documents:

• IMO - International Convention on the Safety of Life at Sea (SOLAS), 74/78 Ch II-2, Pt

D, Reg 13.3.2.5 – Marking of escape routes

• IMO – Fire Safety Systems (FSS Code), Ch 11 – Low-location lighting systems

• IMO - Resolution A.752(18) - Guidelines for the Evaluation, Testing and Application of

Low-Location Lighting on Passenger Ships - (adopted on 4 November 1993)

• ISO 15370:2010 - Ships and marine technology - Low-location lighting on passenger

ships – Arrangement

• MSC/Circ.1168 – Interim guidelines for the testing, approval and maintenance of

evacuation guidance systems used as an alternative to low-location lighting systems

Z17

Page 28 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in sound pressure level measurements of public address and

general alarm systems on board ships

1. 1 Extent of engagement – Sound pressure level measurements of public address and

general alarm systems on board ships.

1. 2 Operators – The operator is to have the following qualifications:

• Have adequate knowledge of the applicable international requirements (SOLAS Reg.

III/4 and III/6, LSA CODE Chapter VII/7.2, IMO Code on alarms and indicators, 1995)

• Be able to document theoretical and practical training onboard in using equipment

specified

1. 3 Equipment – The measuring instrument shall be an integrating sound level meter with

frequency analyser capabilities complying with IEC (International Electrotechnical Commission) 60651 and IEC 61672, type 1 precision class with, at least an A-weighting frequency response curve and 1/3 octave and 1 octave band filters, complying to IEC 61260, as appropriate for the measurements to be carried out. In addition microphones shall be of the random incidence type, complying with IEC 60651.

1. 4 Procedures – Documented work procedures are at least to contain information on

inspection preparation, calibration, selection and identification of test locations.

1. 5 Reporting – The report shall describe, as a minimum, the environmental conditions of

the tests and, for each test location, the ambient noise level or the speech interference level, as appropriate for the measurements to be carried out. The report shall conform to any other specific requirement of the Society.

1. 6 Verification – The supplier must have the Surveyor’s verification of each separate job,

documented in the report by his signature.

1. 7 Reference documents

The Service Supplier is to have access to the following documents:

• SOLAS 74/78, Ch III, Pt A, Reg 4 – Evaluation, testing and approval of life-saving

appliances and arrangements

• SOLAS 74/78, Ch III, Pt B, Reg 6 – Communications

• International Life-Saving Appliance (LSA) Code, Ch VII, Reg 7.2 – General alarm and

public address system

• IMO - Code on Alarms and Indicators, 1995 as amended

• IEC 60651 (2001-10) - Sound level meters

• IEC 61672 - Electroacoustics - Sound level meters

• IEC 61260 - Electroacoustics - Octave-band and fractional-octave-band filters

Z17

Page 29 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in testing of coating systems in accordance with IMO Resolution

MSC.215(82) as amended and IACS UI SC223 and/or MSC.288(87) as amended

1. 1 Laboratories

1. 1.1 Extent of Engagement - Testing of coatings systems according to IMO Resolution

MSC.215(82), as corrected by IMO MSC.1/Circ.1381 and amended by IMO Resolution 341(91) and IACS UI SC223 and/or MSC.288(87), as corrected by IMO MSC.1/Circ.1381 and amended by IMO Resolution 341(91).

1. 1.2 The laboratory is to provide to the Society the following information:

• A detailed list of the Laboratory test equipment for the coating approval according to the

IMO Resolution MSC.215(82) as amended and/or MSC.288(87) as amended.

• A detailed list of reference documents comprising a minimum those referred to in IMO

Resolution MSC.215(82) as amended and/or MSC.288(87) as amended for the coating

approval.

• Details of test panel preparation, procedure of test panel identification, coating

application, test procedures and a sample test report.

• Details of exposure method and site for weathering primed test panels.

• A sample daily or weekly log/form for recording test conditions and observations

including unforeseen interruption of the exposure cycle with corrective actions.

• Details of any sub-contracting agreements (if applicable).

• Comparison test report with an approved coating system or laboratory if available.

1. 1.3 Reporting – Reference is made to the following IACS Recommendations:

• Rec. 101: IACS Model Report for IMO Resolution MSC.215(82) Annex 1 “Test

Procedures for Coating Qualification”

• Rec. 102: IACS Model Report for IMO Resolution MSC.215(82) Annex 1 “Test

Procedures for Coating Qualification”, Section 1.7 – Crossover Test

1. 1.4 Audit of the test laboratory is to be based on this procedure and the standards listed in

the IMO Resolution MSC.215(82) as amended and/or MSC.288(87) as amended for the coating approval.

Z17

Page 30 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in the servicing and maintenance of lifeboats, launching

appliances, on-load release gear and davit-launched liferaft automatic release hooks

1. 1 Extent of engagement – Servicing and maintenance of lifeboats, launching appliances,

on-load release gear and davit-launched liferaft automatic release hooks.

1. 2 Extent of Approval

1. 2.1 The contents of this procedure apply equally to manufacturers when they are acting as

Service Suppliers.

1. 2.2 Any Service Supplier engaged in the thorough examination, operational testing, repair

and overhaul of lifeboats, launching appliances, on-load release gear and davit-launched liferaft automatic release hooks carried out in accordance with SOLAS regulation III/20 should be qualified in these operations for each make and type of equipment for which they provide the service, and provide manufacturers documentary evidence that they have been so authorized or they are certified in accordance with an established system for training and authorization in accordance with MSC.1/Circ.1277, as amended.

1. 2.3 In cases where an equipment manufacturer is no longer in business or no longer

provides technical support, Service Suppliers may be authorised for the equipment on the basis of prior authorization for the equipment and/or long term experience and demonstrated expertise as an authorized service provider.

1. 3 Qualifications and Training of Personnel

Service Suppliers should be trained and qualified in the operations for which they are authorised, for each make and type of equipment for which they provide the service. Such training and qualification should include, as a minimum:

1. 3.1 Employment and documentation of personnel certified in accordance with a

recognized national, international or industry standard as applicable, or an equipment manufacturer’s established certification program. In either case, the certification program should be based on the guidelines in the appendix for each make and type of equipment for which service is to be provided.

1. 3.2 The education and training for initial certification of personnel should be documented

and address, as a minimum:

• Causes of lifeboat accidents

• Relevant rules and regulations, including International Conventions

• Design and construction of lifeboats, including on-load release gear and launching

appliances

• Education and practical training in the procedures specified in annex 1 of

MSC.1/Circ.1206/Rev.1 for which certification is sought

• Detailed procedures for thorough examination, operational testing, repair and overhaul

of lifeboats, launching appliances and on-load release gear, as applicable; and

• Procedures for issuing a report of service and statement of fitness for purpose based on

MSC.1/Circ.1206/Rev.1 (annex 1, paragraph 15)

Z17

Page 31 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 13.3.3 The education and training for the personnel should include practical technical training on actual inspection and maintenance using the equipment (lifeboats, launching appliances and/or on-load release gear) for which the personnel are to be certified. The technical training should include disassembly, reassembly, correct operation and adjustment of the equipment. Classroom training should be supplemented by field experience in the operations for which certification is sought, under the supervision of an experienced senior certified person.

1. 3.4 At the time of initial certification and at each renewal of certification, the service

supplier shall provide documentation to verify personnel’s satisfactory completion of a competency assessment using the equipment for which the personnel are certified.

1. 3.5 The Service Supplier shall require refresher training as appropriate to renew the

certification.

1. 4 Reference Documents - The Service Supplier is to have access to the following

documents:

• IMO - MSC.1/Circ.1206/Rev.1, as amended, Measures to Prevent Accidents with

Lifeboats

• IMO – MSC.1/Circ.1277, as amended, Interim Recommendation on Conditions for

Authorization of Service Providers for Lifeboats, Launching Appliances and On-Load

Release Gear

• IMO – Resolution A.689(17), recommendation on testing of life-saving appliances and,

for life-saving appliances installed on board on or after 1 July 1999, Resolution

MSC.81(70), revised recommendation on testing of life-saving appliances

• For servicing and repair work involving disassembly or adjustment of on-load release

mechanisms, availability of the equipment manufacturer’s specifications and

instructions

• Type Approval certificate showing any conditions that may be appropriate during the

servicing and/or maintenance of lifeboats, launching appliances and on-load release

gear

1. 5 Equipment and Facilities - The Service Supplier is to have access to the following:

• Sufficient tools, and in particular any specialized tools specified in the equipment

manufacturer’s instructions, including portable tools as needed for work to be carried

out on board ship

• Access to sufficient materials, spare parts and accessories as specified by the

equipment manufacturer for repairing lifeboats, launching appliances and on-load

release gear, as applicable

• For servicing and repair work involving disassembly or adjustment of on-load release

mechanisms, availability of genuine replacement parts as specified or supplied by the

equipment manufacturer

1. 6 Reporting - The report should conform to the requirements of MSC.1/Circ.1206/Rev.1

(annex 1, paragraph 15). When repairs, thorough examinations and annual servicing are completed, a statement confirming that the lifeboat arrangements remain fit for purpose should be promptly issued by the Service Supplier.

Z17

Page 32 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in measurements of Noise level Onboard Ships

1. 1 Extent of engagement

Sound pressure level measurements onboard Ship.

1. 2 Supervisor

The supervisor shall have a minimum of 2 years of experience as an operator in sound pressure level measurements.

1. 3 Operators

The operator is to have the following qualifications:

• Knowledge in the field of noise, sound measurements and handling of measurement

equipment

• Adequate knowledge of the applicable international requirements (SOLAS Regulation

II-1/3-12, as amended, and IMO Code on noise levels onboard Ships, as amended,)

• At least 1 year’s experience, including participation in a minimum of 5 measurement

campaigns as an assistant operator

• Training concerning the procedures specified in IMO Code on Noise Levels onboard

Ships

• Be able to document theoretical and practical training onboard in using a sound level

meter

1. 4 Equipment

1. 4.1 Sound level meters

Measurement of sound pressure levels shall be carried out using precision integrating sound level meters. Such meters shall be manufactured to IEC 61672-1(2002-05)1, as amended, type/class1 standard as applicable, or to an equivalent standard acceptable to the Administration2.

1 Recommendation for sound level meters. 2 Sound level meters class/type 1 manufactured according to IEC 651/IEC 804 may be

used until 1 July 2016.

1. 4.2 Octave filter set

When used alone, or in conjunction with a sound level meter, as appropriate, an octave filter set shall conform to IEC 61260 (1995)3, as amended, or an equivalent standard acceptable to the Administration.

3 Octave-band and fractional-octave-band filters

1. 4.3 Sound Calibrator

Sound calibrators shall comply with the standard IEC 60942 (2003-01), as amended, and shall be approved by the manufacturer of the sound level meter used.

Z17

Page 33 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 14.4.4 Calibration

Sound Calibrator and sound level meter shall be verified at least every two years by a national Standard laboratory or a competent laboratory accredited according to ISO 17025 (2005), as amended. A record with a complete description of the equipment used shall be kept, including a calibration log.

1. 4.5 Microphone wind screen

A microphone wind screen shall be used when taking readings outside, e.g. on navigating bridge wings or on deck, and below deck where there is any substantial air movement. The wind screen should not affect the measurement level of similar sounds by more than 0.5 dB(A) in "no wind" conditions.

1. 5 Procedures and instructions

1. 5.1 The supplier shall have documented procedures and instructions to carry out service

of the equipment.

Documented work procedures are at least to contain information on inspection preparation, selection and identification of sound level measurement locations, calibration checks and report preparation.

1. 5.2 The supplier shall have access to the following documents:

SOLAS 1988, as amended (Reg.II-1/3-12)

Resolution A.468(XII) and IMO Resolution MSC.337(91) code on noise levels on board

ships

Resolution A.343(IX) Recommendation on methods of measuring noise levels at listening posts

The Society’s Rules and Guidelines 14.6 Reporting

A noise inspection report shall be made for each ship. The report shall comprise information on the noise levels in the various spaces on board. The report shall show the reading at each specified measuring point. The points shall be marked on a general arrangement plan, or on accommodation drawings attached to the report, or shall otherwise be identified. The format for noise inspection reports is set out in appendix 1 of IMO Code on Noise Levels onboard Ships and may conform to any other specific requirement of the society (refer to IMO circular MSC.337(91)).

1. 7 Verification

The supplier must have the Surveyor’s verification of each separate job, documented in the report by his signature.

Z17

Page 34 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont)

1. Firms engaged in tightness testing of primary and secondary barriers of gas

carriers with membrane cargo containment systems for vessels in service

1. 1 Extent of engagement

Firms carrying out the following:

• Global Vacuum Testing of Primary and Secondary Barriers

• Acoustic Emission (AE) Testing

• Thermographic Testing

1. 2 Requirements for firms engaged in global testing of primary and secondary barriers

1. 2.1 Testing Procedures – Testing is to be carried out in accordance with cargo

containment system designer’s procedures as approved by the society.

1. 2.2 Authorization - The supplier is to be authorized by the system designer to carry out the

testing.

1. 2.3 Equipment – Equipment is to be maintained and calibrated in accordance with

recognized national or international industrial standards.

1. 2.4 Reporting – The report is to contain the following:

• Date of testing

• Identity of test personnel

• Vacuum decay data for each tank

• Summary of test results

1. 3 Requirements for firms engaged in acoustic emission (AE) testing

1. 3.1 Testing procedures – The supplier is to have documented procedures based upon

recognized national or international industrial standards to perform ultrasonic leak test using AE sensors for the secondary barrier of membrane cargo containment systems. The procedures are to include details of personnel responsibilities and qualification, instrumentation, test preparation, test method, signal processing, evaluation and reporting.

Note: The differential pressure during testing should not exceed the containment system designer’s limitations.

1. 3.2 Supervisor – The responsible supervisor shall be certified to a recognized national or

international industrial standard (e.g. Level II, ISO-9712 as amended or SNT-TC-1A as amended) and have one year experience at Level II.

1. 3.3 Operators – The operators carrying out the acoustic emission (AE) testing shall be

certified to a recognized national or international industrial standard (e.g. Level I, ISO-9712 as amended or SNT-TC-1A as amended) and shall have adequate knowledge of ship structures sufficient to determine sensor placement.

Z17

Page 35 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) 15.3.4 Equipment – Equipment is to be maintained and calibrated in accordance with recognized national or international industrial standards or equipment manufacturer’s recommendations.

1. 3.5 Evaluation of acoustic emission (AE) testing – Must be carried out by the supervisor or

individuals certified to a recognized national or international industrial standard (e.g. Level II, ISO-9712 as amended or SNT-TC-1A as amended) and have one year experience at Level II.

1. 3.6 Reporting – The report is to contain the following:

• Date of testing

• Supervisor and operator(s) certifications

• Description of time and pressure of each cycle of test

• List and sketch detailing location of possible defects

1. 4 Requirements for firms engaged in thermographic testing

1. 4.1 Testing Procedures – Testing is to be carried out in accordance with the cargo

containment system designer’s procedures as approved by the society.

1. 4.2 Authorization - The supplier is to be authorized by the system designer to carry out the

testing.

1. 4.3 Supervisor – The responsible supervisor shall be certified to a recognised national or

international industrial standard (e.g. Level II, ISO-9712 as amended or SNT-TC-1A as amended) with additional certification in infrared/thermal testing. Certification by the supplier is not allowed and must be obtained through an independent certification body. SNT-TC-1A certified personnel must provide evidence that training on Level II or above has been administered by an independent training body centrally certified to ASNT or a comparable nationally recognized certification scheme.

1. 4.4 Operators – The operators carrying out the imaging shall be certified to a recognized

national or international industrial standard (e.g. Level I, ISO-9712 as amended or SNT-TC- 1A as amended) with additional certification in infrared/thermal testing and shall have adequate knowledge of ship structures sufficient to determine position for each identified image, and of the containment system to understand the basis of the testing. Certification by the supplier is not allowed and must be obtained through an independent certification body. SNT-TC-1A certified personnel must provide evidence that training on Level I or above has been administered by an independent training body centrally certified to ASNT or a comparable nationally recognized certification scheme.

1. 4.5 Equipment – Thermal cameras and sensors are to be in accordance with the system

designer’s procedures with regards to sensitivity, accuracy and resolution.

Equipment are to be in accordance with recognized standard (IEC, etc.) with regards their safety characteristics for the use in hazardous areas (in gas explosive atmosphere), maintained and calibrated in accordance with the maker’s recommendations.

1. 4.6 Evaluation of thermographic images – Must be carried out by the supervisor or

individuals certified to a recognized national or international industrial standard (e.g. Level II, ISO-9712 as amended or SNT-TC-1A as amended) with additional certification in

Z17

Page 36 of 36 IACS Req. 1997/Rev.12 2016 Z17 (cont) infrared/thermal testing. Certification by the supplier is not allowed and must be obtained through an independent certification body. SNT-TC-1A certified personnel must provide evidence that training on Level II or above has been administered by an independent training body centrally certified to ASNT or a comparable nationally recognized certification scheme.

1. 4.7 Reporting – The report is to contain the following:

• Date of testing

• Supervisor and operator(s) certifications

• Differential pressures of all phases

• List and sketch detailing location of thermal indications

• Thermographic images of all phases of testing for thermal indications

• Evaluation of thermal images indicating possible leaks

End of Document

Annex E MCA Safety Bulletin No. 12

Page 1 of 2 SAFETY BULLETIN NO. 12

Date of issue: September 2018

SAFETY BULLETIN 12 - Accidental CO2 releases onboard 2 UK Merchant Vessels The Issue The Maritime & Coastguard Agency is aware of two serious and potentially near fatal accidental CO2 releases on UK ships in the last two years. In both cases the CO2 leaked from the manifold into the CO2 room. In both cases the remote release valves were untouched and the CO2 alarms operated, alerting the crew and averting fatalities. However, these incidents follow a concerning pattern of similar incidents. The MCA would like to remind operators that CO2 is highly asphyxiating, a 9% concentration causes unconsciousness within minutes and 17% causing death within just a couple of minutes. CO2 is also both colourless and odourless. Key findings of a recent MAIB investigation: • Many systems are designed such that a single leaking valve can discharge the entire system. • Lack of clarity on life/service intervals and maintenance requirements of cylinder valves. • There is perceived over reliance on shore-based contractors who may have poor knowledge of the specific system fitted onboard. • CO2 leaked from the systems and was not contained by the pipework and manifold.

Image - Common arrangement of pilot bottles connected to system with flexible pipes Some key areas that should be highlighted which will reduce the risk of reoccurrence of similar incidents; • When bottles are required to be refilled, it is important that valves should be either serviced or replaced at least in line with manufacturers recommendations. With

Page 2 of 2 regards maintenance of the valves the MCA and MAIB would like to draw the attention of service agents and ships operators to BS EN ISO 22434:2011 - Transportable gas cylinders - Inspection and maintenance of cylinder valves which provides relevant guidance on the issue of maintenance where manufacturers are silent on the issue. • MCA interpretation of MSC.1/Circ.1318 requires that 10% of high pressure CO2 cylinders are hydrostatically tested at their 10 year anniversary. Furthermore, in line with BS EN 1968-2002 – All remaining cylinders must be hydrostatically tested by the 20 year anniversary. • Flexible pipework must be replaced at intervals specified by the manufacturer or at the 10 year anniversary whichever is sooner as per MSC.1/Circ.1318. • The IMO FSS Code, chapter 5.2.1.1.3 requires that crew should be checking quantities of fire extinguishing medium. Given the numbers of bottles involved, methods such as weighing of cylinders are highly impractical. There are now commonly available simple methods such as ultrasonic liquid level gauges which facilitate easy in situ level testing which operators should consider in order that their crews can readily and safely check the levels of CO2 thus enabling early detection of a potential problem. The UK would consider this appropriate as per the IMO FSS Code Chapter 5.2.1.1.3 and this could be built into the planned maintenance system. • The MCA would remind operators of the benefit of marking the cylinders and checking CO2 levels at least annually. • The MCA would also recommend that operators test and ensure the correct operation of any pressure switches and alarms within their systems.

Further information

For further information please contact marinetechnology@mcga.gov.uk or phone on +44 2038 172000

FD06059 ｜ 2019_英国_explosion-and-fire-on-chemical-tanker-stolt-groenland__InterimReport201912-StoltGroenland.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2019 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H INTERIM REPORT SERIOUS MARINE CASUALTY December 2019 1 Interim report on the investigation of the explosion and fire on board the chemical tanker Stolt Groenland on 28 September 2019, Ulsan, Republic of Korea The information contained in this interim report is based on investigations to date. Readers are cautioned that there is the possibility new evidence may become available that might alter the circumstances as depicted in this report. Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of a safety investigation into an accident under these Regulations shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of such an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2019 See http://www. nationalarchives.gov.uk/doc/ open-government-licence for details. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459

INTERIM REPORT As the full investigation report will not be published within 12 months of the accident date, this interim report is published, pursuant to Regulation 14(2)(b) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012. Stolt Groenland Image courtesy of https://www.fleetmon.com NARRATIVE On 17 August 2019, Stolt Groenland, a Cayman Islands registered chemical/ products tanker operated by Stolt Tankers B.V. sailed from Houston, USA for passage to Kobe, Japan via the Panama Canal. The tanker was carrying 20 different chemical cargoes in 37 of its 39 cargo tanks. After discharging four cargo tanks off Kobe on 23 September and six cargo tanks at the Odfjell Terminal in Ulsan, Republic of Korea, between 26 and 27 September, Stolt Groenland moored alongside the Yeompo Quay in Ulsan. Two further cargo tanks were then discharged via a ship to ship transfer with Stolt Voyager. On completion, Stolt Voyager moored ahead of Stolt Groenland. At about 0600 on 28 September, the Singapore registered chemical/ products tanker Bow Dalian secured alongside Stolt Groenland’s port side. The purging of Bow Dalian’s cargo tanks with nitrogen supplied from shore- side vehicles was then commenced in readiness for a ship-to-ship cargo transfer with Stolt Groenland.

2 At 1043, vapour started to release from the pressure vacuum valve for Stolt Groenland’s number 9 starboard (9S) cargo tank, which contained styrene monomer. About 2 minutes later, a high- level alarm indicated that the level in 9S cargo tank had reached 95%, soon followed by a high-high-level alarm indicating that the level had increased to 98%. By now, Stolt Groenland’s on-watch deck officer and chief officer had made their way to the cargo control room and they saw from the cargo monitoring system that the pressure inside 9S cargo tank was rapidly rising. Suddenly, at 1050, two explosions were seen and heard in rapid succession in way of the tanker’s cargo manifold (Figure 1). The resulting fireball passed very close to a road bridge above the quay (Figure 2). Figure 1: Tank rupture and ignition of released vapour Vapour release from PV valve (9S cargo tank) 9S cargo tank Figure 2: Fireball viewed from the Ulsan bridge

3 Stolt Groenland’s and Bow Dalian’s crews immediately operated and directed foam monitors towards the respective cargo manifolds. Bow Dalian’s cargo manifold drench system was also activated. The fire on board Stolt Groenland was very intense and large plumes of thick black smoke were emitted. Stolt Groenland’s crew evacuated using the free-fall lifeboat at the stern while Bow Dalian’s crew were taken off by Korean Coastguard boats that had arrived at the scene. DAMAGE AND INJURIES The fire was fought from the shore and from the water and lasted until early the following morning. The damage in way of Stolt Groenland’s cargo manifold and accommodation block was extensive (Figure 3). The following persons were injured: Stolt Groenland’s gangway watchkeeper, who was blown over the guardrails by the force of the explosion; a deck rating on board Bow Dalian; and, a number of shore workers, including firefighters and police during the emergency response. STYRENE MONOMER Styrene monomer is used to make plastics, paints and synthetic rubber. It is a colourless to yellowish oily liquid with an aromatic odour and is highly flammable. It has a flash point of 32°C (90°F). It is also toxic and reactive and is generally stable at ambient temperature. Polymerization1 of the monomer is initiated by heat or contact with peroxides. The polymerization process is exothermic and the monomer’s temperature may rise to the point where the reaction becomes very rapid and self-sustaining (runaway polymerization). Normally, temperatures above 65°C (149°F) are needed to initiate runaway polymerizations. To reduce the possibility of polymerization in storage and during transportation, Tert- Butylcatechol (TBC) inhibitor is added to styrene monomer, typically at a concentration of between 10-15ppm, but higher levels might be needed depending on storage conditions and expected voyage duration. The depletion of TBC over time and its effectiveness as an inhibitor are affected by temperature, oxygen levels and water. INVESTIGATION On 28 September, the Maritime Authority of the Cayman Islands requested the MAIB investigate this accident in accordance with the International Maritime Organization Casualty Investigation Code2. During the on-site investigation in Ulsan, the MAIB was assisted by the Korean Maritime Safety Tribunal. In addition, the Singapore Transport Safety Investigation Bureau supported MAIB’s evidence collection on board Bow Dalian, along with the vessel’s owners. Access to Stolt Groenland was controlled by the Korean Coastguard and restricted by the toxicity of the atmosphere on board. MAIB inspectors eventually boarded the tanker on 8 October 2019 with the co-operation and assistance of Stolt Tankers B.V., the vessel’s operator. A further inspection was carried out on 12 November 2019. Stolt Tankers B.V. has provided a copy of the data saved on the tanker’s voyage data recorder capsule. 1 Polymerization is a chemical reaction in which two or more molecules combine to form larger molecules that contain repeating structural units. However, if a large amount of monomer is involved, and if the reaction is strongly exothermic, the monomers may combine too quickly. Consequently, heat and pressure increase. 2 IMO Resolution MSC 255(84)

4 9C cargo tank area Approximate area of tank split (at deck level) 9S cargo tank area Figure 3: Stolt Groenland starboard cargo manifold

5 INITIAL FINDINGS The explosions on board Stolt Groenland were probably caused by the rupture of the deck above 9S cargo tank, followed immediately by the ignition of the styrene monomer vapour that was then released. The rupture was due to over-pressurisation and the likely sources of the ignition were static electricity, sparks or elevated steel deck plate temperatures resulting from the tank rupture. A large hole was found in way of the tank’s common bulkhead with number 9 centre (9C) cargo tank (Figure 4) and its hatch cover had been blown off. No cargo operations or deck maintenance were in progress at the time. VDR data showed that the temperature of the styrene monomer had reached 100ºC at the time of the explosion. Such an elevated temperature indicates that the cargo was polymerizing. Figure 4: 9S tank rupture Polymerized styrene monomer

6 Stolt Groenland had loaded 5,250 tonnes of styrene monomer at the LBC terminal in Houston, U.S.A. between 7 and 8 August. The cargo was distributed between three stainless steel cargo tanks: numbers 9S, 6P and 6C. Prior to loading, the tanks had been washed and inspected and wall wash tests had been conducted; no problems were recorded. The tanks were not purged with nitrogen before loading and none were adjacent to heated cargoes in accordance with cargo loading instructions. The concentration of TBC inhibitor in the styrene monomer in the shore tank was 11.3ppm but this was increased by the addition of 3 US gallons of liquid TBC to each of the destination tanks before loading was commenced. The target concentration of the TBC was 17ppm. The loading was overseen by an American Cargo Assurers cargo surveyor acting on behalf of the supplier of the styrene monomer, Ineos Styrolutions, and the purchaser, Samsung C&T Corporation. The certificate of inhibitor, issued by the cargo surveyor at the time of loading, stated that the TBC should remain effective for between 60 and 90 days. It also stated that the inhibitor was oxygen dependent and that the ideal temperature for the styrene monomer was between 60ºF and 85ºF (15.5ºC – 29.4ºC). Stolt Groenland did not carry additional inhibitor. ACTIONS TAKEN Stolt Tankers B.V. engaged salvors to assist in making Stolt Groenland safe. The discharge of the remaining chemical cargo from the undamaged tanks was completed by Stolt personnel on 29 October 2019. The Ulsan Regional Office of Oceans and Fisheries has prohibited the transport and unloading of some hazardous cargoes, including styrene monomer, at several terminals under the Ulsan bridge. OTHER RECENT INCIDENT On 20 November 2019, the MAIB was informed of an incident on board the Cayman Islands registered chemical/products tanker Stolt Focus in which the temperature of a styrene monomer cargo had steadily increased. Efforts to stabilize the temperature, including the injection of additional inhibitor, were unsuccessful. Following consultation with chemists ashore, the crew distributed the cargo between four cargo tanks and mixed it with sea water, which stabilized the styrene monomer. THE MAIB INVESTIGATION WAY AHEAD The MAIB’s investigation will focus on, among other things:

● the factors contributing to the initiation of the polymerization of the styrene monomer,

● cargo monitoring, and

● the emergency response. The MAIB will continue to work closely with all the stakeholders involved in order to progress the investigation as swiftly as possible and ensure appropriate action can be taken in order to

7 further reduce the likelihood of similar accidents in the future. Once the MAIB’s investigation is complete, a report will be drafted and distributed to key stakeholders for a 30-day consultation period prior to it being published. In the interim, chemical tanker owners/operators are reminded to:

● Adhere to the carriage and storage instructions details on the safety data sheet, the certificate of inhibitor, and those provided by the charterer.

● Witness the addition of inhibitor into each cargo tank.

● Closely monitor cargo temperature for unexplained increases. They should also ensure that their crews are familiar with the action to take in the event of the styrene monomer self-heating/polymerizing. REQUEST FOR INFORMATION The MAIB is seeking any information from ship owners, ship and terminal operators, or individuals regarding any accidents or ‘near-misses’ involving the carriage of styrene monomer on board ships, including any actions subsequently taken. The MAIB’s email address is maib@dft.gov.uk and further contact details are at www.gov.uk/maib.

8 VESSEL DETAILS Vessel’s name Stolt Groenland Flag Cayman Islands Classification society Det Norske Veritas Germanischer Lloyd (DNVGL) IMO number 9414072 Type Chemical/Products Tanker Registered owner Stolt Tankers B.V. Manager Stolt Tankers B.V. Year of build 2009 Construction Steel Length overall 182.72 Gross tonnage 25881 Authorised cargo Chemicals/Products VOYAGE PARTICULARS Port of departure Ulsan, Republic of Korea Port of arrival Ulsan, Republic of Korea Type of voyage International Cargo information 31178 tonnes of chemicals Manning 25 MARINE CASUALTY INFORMATION Date and time 28 September 2019 at 1050 (UTC+9) Type of marine casualty or incident Serious Marine Casualty Location of incident Ulsan, Republic of Korea Place on board Cargo tanks/Deck Injuries/fatalities 1 (Stolt Groenland), 1 (Bow Dalian). 15 shore workers/officials were also reported to have been injured. Damage/environmental impact Toxic fumes Ship operation Moored alongside Voyage segment Moored alongside External & internal environment Daylight. Air temperature: 24ºC. Wind: light airs Persons on board 25

FD06060 ｜ 2019_英国_safety-warning-about-drivers-remaining-in-vehicle-cabs-while-ferries-are-at-sea_2019-_SB.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2019 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Extracts from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 Regulation 5: “The sole objective of a safety investigation into an accident under these Regulations shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of such an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” Regulation 16(1): “The Chief Inspector may at any time make recommendations as to how future accidents may be prevented.” NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2019 See http://www.nationalarchives. gov.uk/doc/open-government- licence for details. All bulletins can be found on our website: https://www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 Press Enquiries: 01932 440015 Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000 M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H SAFETY BULLETIN SB1/2019 March 2019 Dangers posed to freight vehicle drivers by remaining in their vehicle cabs while on board ro-ro ferries at sea

MAIB SAFETY BULLETIN 1/2019 This document, containing safety lessons, has been produced for marine safety purposes only, based on information available to date. The Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 provide for the Chief Inspector of Marine Accidents to make recommendations at any time during the course of an investigation if, in his opinion, it is necessary or desirable to do so. The Marine Accident Investigation Branch is carrying out an investigation into the cargo shift of freight vehicles on board the ro-ro passenger ferry European Causeway while on passage from Larne to Cairnryan in Scotland. The MAIB will publish a full report on completion of the investigation. Andrew Moll Chief Inspector of Marine Accidents NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall not be admissible in any judicial proceedings whose purpose, or one of whose purposes, is to apportion liability or blame. This bulletin is also available on our website: www.gov.uk/maib Press Enquiries: 01932 440015; Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000

1 BACKGROUND The MAIB is investigating the shift and toppling over of freight vehicles (Figures 1, 2, 3 and 4) on board the P&O ro-ro passenger ferry European Causeway during heavy weather while on passage from Larne, Northern Ireland, to Cairnryan, Scotland. In accordance with international regulations and industry best practice, P&O does not permit passengers to remain on the ro-ro decks when at sea. INITIAL FINDINGS On 18 December 2018, European Causeway was on passage from Larne to Cairnryan when it encountered very strong winds and very rough seas, which caused the ship to roll heavily. As a result, 9 of the 40 freight vehicles on board toppled over, with several vehicles sustaining damage. The MAIB investigation has found that at least six drivers had remained in their freight vehicle cabs during the crossing despite being instructed by the ship’s crew to vacate the ro-ro deck after they had parked their vehicles. Figure 1: Trucks on upper deck aft Figure 2: Truck toppled over Figure 3: Truck toppled over on main deck

2 Four drivers were found in the freight vehicles that had toppled over, with one remaining trapped until he could be freed by the emergency services that were waiting in Cairnryan. Fortunately, nobody was hurt during the accident. The investigation has uncovered that the problem of drivers remaining in their vehicle cabs on the ro-ro deck while ferries are on passage is not unique to this route or to P&O. SAFETY LESSONS A ferry’s ro-ro deck is a hazardous and potentially life-threatening environment. While a ferry is at sea, the ro-ro decks should be occupied by only trained professional seafarers who are required to undertake safety and security patrols. Drivers who remain on the vehicle deck of ro-ro ferries pose a danger to themselves, and can cause a delay to the emergency response, particularly in the event of a fire. Any delay to the activation of fire suppression systems on the vehicle deck due to the need to undertake a muster and headcount of all persons on board could have catastrophic consequences to the whole vessel, its passengers and the environment. Furthermore, drivers who have remained in their vehicle cabs could be in danger of asphyxiation by the fire, or as a result of the fire suppression systems that may be released by ship’s staff. ACTIONS TAKEN P&O Ferries has contacted ferry operators in the United Kingdom who it considers may be affected by the issue of drivers remaining in vehicle cabs on ro-ro decks. Its aim is to encourage operators to contribute to a discussion forum to collectively eliminate this problem. All companies operating ferries to the United Kingdom are strongly encouraged to engage positively with this safety initiative, to work across the industry to develop a cohesive and co- operative approach to resolve this urgent safety issue. RECOMMENDATION The Road Haulage Association Ltd is recommended to: S2019/106 Distribute this Safety Bulletin to its members and encourage them to take robust action to improve and assure driver safety by helping ferry operators eliminate the issue of drivers remaining in the cabs of freight vehicles on ro-ro decks. Safety recommendations shall in no case create a presumption of blame or liability Figure 4: Damaged truck and mini bus

FD06061 ｜ 2019_英国_two-catastrophic-engine-failures-one-resulting-in-a-fire-on-board-ro-ro-passenger-ferry-wight-sky__2019-InterimReport-WightSky.pdf.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2019 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H SERIOUS MARINE CASUALTY May 2019 1 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of a safety investigation into an accident under these Regulations shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of such an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2019 See http://www. nationalarchives.gov.uk/doc/ open-government-licence for details. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 As the full investigation report will not be published within 12 months of the accident date, this interim report is published, pursuant to Regulation 14(2)(b) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012. INTERIM REPORT Interim report on the investigation into two catastrophic engine failures on board the ro-ro passenger ferry Wight Sky on 26 August 2018 and 14 December 2018 BACKGROUND The UK registered roll-on roll-off passenger ferry Wight Sky (Figure 1) was one of three W-Class sister vessels operated between Lymington, Hampshire, and Yarmouth, Isle of Wight, by Wightlink Ltd since 2008. It was propelled and steered by two five-blade Voith Schneider units, one forward and one aft. Depending on the power required, each propulsion unit could be driven by one or two Volvo Penta D16 six-cylinder diesel engines. Two main engines (ME), ME1 and ME2, were located in the forward engine room, with ME3 and ME4 in the aft engine room. On 12 September 2017, Wight Sky’s ME2 suffered a catastrophic failure during the vessel’s approach to Yarmouth, which resulted in an engine room fire and caused serious injuries to an engineering officer. The engine had recently been overhauled ashore by an independent local Volvo Penta dealer, following which it was rebuilt in the engine room on board the vessel. It had been running for less than 6 hours when it failed. Volvo Penta’s investigation concluded that the lubrication supply to a big- end bearing had probably been blocked by debris that had been allowed to enter the engine’s oil channels during the rebuild. The information contained in this interim report is based on investigations to date. Readers are cautioned that there is the possibility new evidence may become available that might alter the circumstances as depicted in this report. Figure 1: Wight Sky Photo by user:geni, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=47612090

2 The accident was investigated by the MAIB and the investigation report 1 concluded that ‘Rebuilding the engine in a clean and controlled environment and transferring it complete into the engine room would have reduced the likelihood of debris ingress.’ Subsequently fully assembled engines were shipped in/out of the W-Class via the designated route. NARRATIVE On 26 August 2018, Wight Sky’s replacement ME2 catastrophically failed (Figure 2) as the vessel was preparing to enter the Lymington river. One of the engine’s connecting rods (Figure 3) and various other internal engine components were ejected through the engine block. The engine lubricating oil ignited, causing a fireball to engulf the engine (Figure 4). The engine room was unmanned at the time. The bridge team were immediately alerted to the situation by the activation of alarms on the engine alarm monitor and could see flames and smoke in the forward engine room via CCTV, and they swiftly activated the fixed fire-fighting water-mist system. This action quickly contained, and then extinguished the fire. Following discussions with the coastguard, the vessel’s master returned Wight Sky to Yarmouth and disembarked the passengers. There were no injuries. The engine that failed had accumulated 2241 running hours since its installation. It was a new build engine that had been delivered from Volvo Penta’s factory in Sweden and installed by the local Volvo Penta dealer. Initially, the failure appeared to be identical to the previous one and the MAIB began a new investigation. 1 MAIB report No 14/2018 - Catastrophic engine failure, resulting in a fire and serious injuries to the engineer on board Wight Sky, off Yarmouth, 12 September 2017. Ignition of engine oil after crankcase breached Figure 2: Catastrophic failure of Wight Sky's replacement ME2 on 26 August 2018 (various pieces of engine debris circled)

3 Figure 3: Failed connecting rod Figure 4: Engine engulfed by flame from ignition of hot engine oil

4 On 14 December 2018, Wight Sky’s ME4 suffered a catastrophic failure as the vessel was closing on the berth at Lymington. The engine was also new, having run for only 380 hours. Again, a connecting rod bottom-end failed, and various internal engine components were ejected through the engine block. There was no fire, but the fixed fire-fighting system was operated immediately as a precaution. There were no injuries and the vessel berthed safely. INVESTIGATION The MAIB is working closely with Wightlink Ltd, Volvo Penta, the local Volvo Penta dealer, Lloyd’s Register (Wight Sky’s Classification Society) and the Maritime and Coastguard Agency to investigate the two latest engine failures. Due to the apparent similarities between the failure of ME4 and the previous failures of ME2, the MAIB investigation findings will be set out in a combined report. The technical investigation team identified two other recent non-catastrophic engine failures that occurred on board Wight Sky’s sister vessel, Wight Light, and included them in the scope of the investigation. One of the Wight Light engines suffered a crankshaft main bearing failure on 2 February 2018, the other a melted piston and partial seizure on 3 August 2018. The engines were sent to Volvo Penta’s facility in Gothenburg for dismantling and examination (Figure 5) to determine the cause of failure. The initial examination of Wight Sky’s ME2 that failed on 26 August 2018 did not identify any obvious cause of failure, and it was brought back to the UK for metallurgical examination. In addition to seeking to identify the direct causes and mechanisms of the engine failures, the investigation has also focused on other factors that might have contributed to the series of failures. These include, inter alia, vessel and system operating parameters; propulsion system alignment and design; and maintenance management. The investigation team has also re-examined the Wight Sky engine that failed on 12 September 2017 and reviewed the investigation of an identical engine that suffered a similar failure on board a windfarm vessel in September 2017  2. 2 Report No 1/2018 - Catastrophic engine failure resulting in a fire on crew transfer vessel Windcat 8, 7 September 2017. Figure 5: Removed engine after delivery to Volvo Penta's factory in Gothenburg

5 INITIAL FINDINGS The investigation into the series of Volvo Penta D-16 engine failures experienced on board Wightlink’s Yarmouth to Lymington ferries, and Wight Sky in particular, is technically complex and is ongoing. The initial findings support the conclusions made in MAIB report No.14/2018 with respect to the first Wight Sky ME2 engine failure, and have identified direct causes for two of the other four engine failures: • Wight Sky’s newly built ME4 suffered a catastrophic failure after 380 running hours because of an assembly error during build in the Volvo Penta factory. • Wight Light’s ME1 seized because of an assembly error during overhaul conducted by the local Volvo Penta dealer. Although direct causes have not been established for all five engines, the investigation has identified several underlying factors that might have contributed to this unusually high incidence of failures. These include the way the vessels and their engines are driven during entry to and departure from port; the set- up of the engines’ auxiliary systems; the management of maintenance on board and ashore; quality control during engine assembly and overhaul. ACTIONS TAKEN Following the failure of Wight Sky's main engine (ME4) in December 2018, Wightlink immediately withdrew its W-class vessels from service. After discussions between Wightlink, the MCA, Lloyd’s Register, and Volvo Penta, a mitigation plan was put in place to enable the vessels to return to service. The initial mitigation measures included weather and engine load constraints, enhanced engine monitoring and shutdown procedures, and restrictions on personnel entering the machinery spaces when the main engines were running. Further precautionary steps have also been taken as the technical investigation has progressed, including: shorter periods between routine fuel injector overhauls; modifications to the engine alarm system; replacement of flexible engine mounts; reduction of oil replacement and oil sampling hours; and testing of engine coolant. Following the discovery of the assembly error Volvo Penta has strengthened its standard operation process to prevent this from occurring again. Volvo Penta confirmed the same error had not been repeated on engines manufactured around the same time. THE MAIB INVESTIGATION WAY AHEAD The MAIB will continue to work closely with all the stakeholders involved in order to progress the investigation as swiftly as possible and ensure appropriate action can be taken in order to further reduce the likelihood of similar accidents in the future. Once the MAIB’s investigation is complete, a report will be drafted and distributed to key stakeholders for a 30-day consultation period prior to it being published.

6 SHIP PARTICULARS Vessel’s name Wight Sky Flag United Kingdom Classification society Lloyd’s Register IMO number/fishing numbers 9446984 Type Ro-ro passenger ferry Registered owner Wightlink Limited Manager Wightlink Limited Year of build 2008 Construction Steel Length overall 62.40m Gross tonnage 2546 tonnes Authorised cargo Passengers, private and commercial vehicles VOYAGE PARTICULARS Port of departure Yarmouth, Isle of Wight Yarmouth, Isle of Wight Port of arrival Lymington, Hampshire Lymington, Hampshire Type of voyage Internal waters Internal waters Cargo information 117 passengers, 29 cars, 1 lorry, 3 vans 43 passengers, 17 vehicles Manning 10 9 MARINE CASUALTY INFORMATION Date and time 26 August 2018, 1835 14 December 2018, 0655 Type of marine casualty or incident Serious Marine Casualty Serious Marine Casualty Location of incident 50° 44’ 41N 1˚ 30’ 42W 50m from Lymington ferry berth Place on board Forward engine room Aft engine room Injuries/fatalities None None Damage/environmental impact No.2 main engine damaged beyond repair No.4 main engine damaged beyond repair Ship operation Commercial voyage Commercial voyage Voyage segment Mid-water Arrival External & internal environment Near gale; slight sea; moderate visibility Gentle breeze; calm sea; good visibility Persons on board 127 52

FD06062 ｜ 2020_英国_collision-between-2-fire-and-rescue-service-boats-with-loss-of-1-life__2020-17-FRSRIBs.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2020 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

ACCIDENT REPORT ACCIDENT REPORT Report on the investigation of the collision between two fire and rescue service boats resulting in one fatality River Cleddau, Milford Haven, Wales 17 September 2019 VERY SERIOUS MARINE CASUALTY REPORT NO 17/2020 NOVEMBER 2020

The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be in­ admissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2020 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch First Floor, Spring Place 105 Commercial Road Southampton United Kingdom SO15 1GH Email: maib@dft.gov.uk Telephone: +44 (0) 23 8039 5500 Fax: +44 (0) 23 8023 2459 Press enquiries during office hours: 01932 440015 Press enquiries out of hours: 020 7944 4292

CONTENTS GLOSSARY OF ABBREVIATIONS AND ACRONYMS SYNOPSIS 1 SECTION 1 - FACTUAL INFORMATION 2 1.1 Particulars of N207 and Rescue 1 and accident 2 1.2 Background 3 1.3 Environmental conditions 4 1.4 Narrative 4 1.5 Emergency response 9 1.6 Boats’ crews 9 1.6.1 N207 9 1.6.2 Rescue 1 10 1.7 Mid and West Wales Fire and Rescue Service 10 1.7.1 Overview 10 1.7.2 N207 10 1.7.3 Rescue 1 11 1.7.4 Boat procedures 13 1.7.5 Risk assessments 14 1.7.6 Control of the activity 14 1.7.7 Training and familiarisation 15 1.7.8 Audit, inspection and peer review 17 1.8 Regulation and oversight of FRS boat operations 18 1.8.1 Overview 18 1.8.2 National Fire Chief’s Council 19 1.8.3 Regulations for boat operations 19 1.8.4 Other boat operating codes 19 1.9 Port Authority 20 1.10 Previous/similar accidents 20 1.10.1 Collision between two RIBs Osprey and Osprey II (MAIB Report

No 10/2017) 20 1.10.2 Two Cardiff Bay Yacht Club RIBs (MAIB Report No 19/2011) 21 SECTION 2 - ANALYSIS 22 2.1 Aim 22 2.2 Overview 22 2.3 The collision 22 2.3.1 Rescue 1 22 2.3.2 N207 23 2.4 Lookout 23 2.5 Planning and control of the training and familiarisation evolutions 23 2.6 Qualifications, crewing and training 24 2.6.1 Overview 24 2.6.2 Training 24 2.6.3 Management oversight of training and certification 25

1. 7 Procedures

25 2.7.1 Personal protective equipment 25 2.7.2 Suitability of procedures and risk assessments 25 2.8 Boat operating codes and guidance 26 SECTION 3 - CONCLUSIONS 28 3.1 Safety issues directly contributing to the accident that have been addressed or resulted in recommendations 28 3.2 Safety issues not directly contributing to the accident that have been addressed or resulted in recommendations 29 SECTION 4 - ACTION TAKEN 30 4.1 Actions taken by other organisations 30 SECTION 5 - RECOMMENDATIONS 31

FIGURES Figure 1: N207 (Zodiac Milpro FC470) Figure 2: Rescue 1 (Delta 6.4 RIB) Figure 3: Milford Haven and River Cleddau Figure 4: Location of crew in N207 Figure 5: Tracks of boats before collision Figure 6: Manoeuvring leading to the collision Figure 7: Console arrangement of Rescue 1 Figure 8: Windscreen of Rescue 1 Figure 9: Training matrix Figure 10: Location of Caldey Island ANNEXES Annex A: Extracts from Standard Operating Procedure 2.08 - Flooding Annex B: Extracts from Standard Operating Procedure 2.10 - Rescues; Working Near, On, Or In Water Annex C: Extracts from Appliance, Equipment, Technical Instruction for: Fire Service Boats 5.17 Annex D: Health and Safety Procedural Guidance Document 5.19 - Risk Assessment Annex E: Milford Haven Port Authority - Leisure Guide

GLOSSARY OF ABBREVIATIONS AND ACRONYMS

AET

Appliance, Equipment, Technical instruction for fire service boats AWOA

All Wales Operational Assurance

DEFRA

Department for Environment Food and Rural Affairs FRCO

Flood Rescue Concept of Operations FRS

Fire and Rescue Service GMDSS

Global Maritime Distress and Safety System hp

horsepower kts

knots (nautical miles per hour) m

metres MCA

Maritime and Coastguard Agency MGN

Marine guidance note MHPA

Milford Haven Port Authority MWWFRS

Mid and West Wales Fire and Rescue Service NFCC

National Fire Chief’s Council RBO

Rescue Boat Operator

RIB

Rigid Inflatable Boat

RYA

Royal Yachting Association SRT

Swift Water and Flood Rescue Technician WRBO

Water Rescue Boat Operator VHF

Very High Frequency TIMES: all times used in this report are UTC + 1 unless otherwise stated.

1 SYNOPSIS At about 1125 on 17 September 2019, two fire and rescue service boats - a 4.7m Zodiac Milpro inflatable boat and a 6.4m Delta rigid inflatable boat - were in collision while undertaking boat training and familiarisation on the River Cleddau near Milford Haven, Wales. The collision resulted in one of the firefighters who was taking part in the training on board the Zodiac, being struck by the Delta RIB and sustaining fatal injuries. The collision occurred because both boats were operating at speed and carrying out un-coordinated manoeuvres in the same stretch of the river. The Delta RIB’s helmsman attempted to pass between the Zodiac and the shoreline as his boat exited a large turn made at speed. The Zodiac’s helmsman, initially unaware of the Delta RIB approaching, turned his boat sharply to port as part of his training manoeuvres, which resulted in the boats heading towards each other. The subsequent action taken by both boats to avoid a collision was unsuccessful. The MAIB investigation found that the accident could have been avoided had the training activities been properly planned. Furthermore:

●No-one was in overall charge of the training and familiarisation activities, so they were not properly managed, briefed or communicated between the crews of both boats. ● The operation of the boats did not adhere to the requirements of the local standard operating procedures or risk assessments. ●

●The generic risk assessment for boat operations was inadequate.

●The Fire and Rescue Services in the United Kingdom did not operate boats to a common standard or code of practice when not employed on flood rescue duties. Mid and West Wales Fire and Rescue Service has undertaken its own internal investigation; has implemented a peer review by two other fire and rescue service authorities for various activities, including boat operations; and has reviewed and updated its standard operating procedures and risk assessments for fire service boats. Mid and West Wales Fire and Rescue Service has been recommended to review the crewing and staff qualification requirements for its boats, implement measures to maintain crew competency as boat helmsman, and improve its auditing process to include its boat operations. The National Fire Chief’s Council has been recommended to adopt a nationwide standard of boat operations for all craft capable of being operated at sea by any fire and rescue service across the United Kingdom.

2 SECTION 1 - FACTUAL INFORMATION 1.1 PARTICULARS OF N207 AND RESCUE 1 AND ACCIDENT SHIP PARTICULARS Vessel’s name N207 Rescue 1 Type Zodiac Milpro FC470 Delta 6.4 Rigid Inflatable Boat Registered owner Mid and West Wales Fire and Rescue Service Mid and West Wales Fire and Rescue Service Manager(s) Milford Haven Fire and Rescue Service Milford Haven Fire and Rescue Service Construction Duratane™ fabric Glass Reinforced Plastic hull, Hypalon tubes Year of build 2017 2001 Length overall 4.7m 6.4m VOYAGE PARTICULARS Port of departure Gelliswick Gelliswick Port of arrival Neyland Neyland Type of voyage Training Familiarisation Cargo information n/a n/a Crewing 3 2 MARINE CASUALTY INFORMATION Date and time 17 September 2019; 1125 Type of marine casualty or incident Very Serious Marine Casualty Location of incident River Cleddau Place on board Forward starboard bow Forward bow Injuries/fatalities One fatality, one minor injury None Damage/environmental impact Superficial damage to boat None Ship operation n/a n/a Voyage segment n/a n/a External & internal environment Lt airs, smooth water, good visibility Persons on board 3 2

3 1.2 BACKGROUND The Mid and West Wales Fire and Rescue Service (MWWFRS) was required by Welsh Assembly legislation to provide a flood rescue service within its operational area. Its Milford Haven fire station had two rescue boats: a 4.7m Zodiac Milpro inflatable boat, callsign N 207 (N207) (Figure 1), which was used principally for flood rescue; and a 6.4m Delta rigid inflatable boat (RIB), callsign Rescue 1 (Rescue 1) (Figure 2), which was used primarily to assist firefighters in the event of an emergency at the local oil and gas terminal jetties. On the evening of 16 September 2019, a hazardous material spillage de-contamination exercise, planned for the following day at the Milford Haven fire station, was cancelled. The 10 firefighters scheduled to undertake the exercise discussed alternative training options and agreed to assemble at the station and take the boats out for boat handling practice and flood rescue training. This decision was made because two of the firefighters were due to travel to North Wales 2 days later to complete Water Rescue Boat Operator (WRBO) training on the Menai Strait using boats identical to N207. Figure 1: N207 (Zodiac Milpro FC470) The 4.7m Zodiac Milpro inflatable boat, call sign N 207 Figure 2: Rescue 1 (Delta 6.4 RIB) The 6.4m Delta rigid inflatable boat (RIB), callsign Rescue 1

4 1.3 ENVIRONMENTAL CONDITIONS The wind was variable in direction and had a strength of force 1 to 2. It was a sunny day with very few clouds. The temperature was about 19°C and the visibility was excellent. High water at Neyland was at 0828 with a height of 6.72m, and low water was at 1430 with a height of 1.10m. 1.4 NARRATIVE At about 0800 on 17 September 2019, the 10 firefighters assembled at Milford Haven fire station and were split into two teams of five. One team was chosen to go out in the boats during the morning while the other manned the fire station. The plan was to swap the teams at lunchtime. At about 0830, following a short discussion, the morning boat team was split into two groups. One of the firefighters was also a WRBO instructor (hereafter referred to as the trainer) and he and the two WRBO trainees were assigned to N207. The two remaining firefighters, each of whom held the rank of crew manager, were assigned to Rescue 1. N207 was made ready at the fire station, then towed down to a launch site at Gelliswick (Figure 3). When they arrived, the three firefighters donned their dry-suits, boots and Palm Rescue 800 buoyancy aids. The boat was equipped with Nitro XT water safety helmets, but they were not worn by the crew. At 0938, the firefighters launched and boarded N207. One of the firefighters secured the engine kill-cord to his wrist and took the helm; the trainer positioned himself in the middle of the boat and the other firefighter sat in the bow. In the meantime, the towing vehicle and trailer returned to the fire station to collect Rescue 1. At about 0945, N207 left Gelliswick and was steered eastwards through the Milford Haven waterway and toward the Cleddau bridge. While en route the trainer pointed out different types of buoyage to the firefighters, discussed tides, and talked about his plans for the morning. At the bridge, the trainer asked the firefighters to practise some swift water boat manoeuvres. However, the flow of the ebbing tide at the bridge was insufficient to practise some intended rope and engine work, so they moved the boat a short distance upstream to an old fish farm structure on the south shore where the tide was ebbing with more strength. The firefighters then swapped places and the second firefighter took over the helm. At 1050, Rescue 1 was launched with the two crew managers on board. Each was wearing a dry-suit, boots and Crewsaver Offshore 290N inflatable lifejacket. Rescue 1 was equipped with Geko open face marine safety helmets, but they were not worn by either crew member. The helmsman positioned himself on the port side of the boat behind the steering wheel, and secured the engine kill-cord to his lifejacket. The other crew manager, who was seated on the starboard side next to the helmsman, assumed the role of crewman and used a hand-held VHF radio to inform the local coastguard that Rescue 1 was on the water. He also explained that they would be engaged in training activities in the eastern part of Milford Haven waterway and in the River Cleddau. Rescue 1’s helmsman steered the RIB eastwards, under the Cleddau bridge, and met up with N207 at the disused fish farm structure. During a shouted discussion between the two boat crews it was agreed that each would make their way independently upstream, practising operating the boats at high speed, and rendezvous again at Mill Bay on the south shore (Figure 3).

5 Figure 3: Milford Haven and River Cleddau Google Earth satellite image of Milford Haven and River Cleddau area Image courtesy of Google Earth

6 As there were no speed restrictions in that part of the river, both helmsmen proceeded ahead at, or close to, their boat’s full speed. Rescue 1, which was travelling at a speed of about 28kts, arrived at Mill Bay ahead of N207, which was being driven at its full speed of 16kts. By about 1115, both boats had arrived at Mill Bay and were rafted together, drifting about 20m from the south shore and pointing upstream. As they drifted, the crews discussed the local topography, the nature of the riverbed and local depths of water. During the conversation they agreed to proceed further upstream to the area of Lawrenny and rendezvous again there. Note: The accounts of the witnesses with regard to the events after the boats left Mill Bay are not wholly consistent, and this section of the narrative is based upon the investigator's understanding of the most likely sequence of events. Shortly after 1120, the two boats parted, with Rescue 1 proceeding ahead and accelerating to about half full speed. Meanwhile, N207 proceeded upstream at variable speeds, with its helmsman sitting on the starboard side of the boat practising manoeuvres that involved large course alterations. N207’s trainer was sitting in the middle of the boat with his back against the port side tube, facing to starboard. The other firefighter, Josh Gardener, was sitting in the bow on the port side (Figure 4). Once ahead and clear of N207, Rescue 1’s helmsman commenced a large full-circle turn of the boat to port (Figure 5a). As the boat neared the end of its final turn to port travelling at about 15kts, the helmsman saw N207 ahead of him in the distance, slightly to port, and noted that it now appeared to be headed downstream towards Mill Bay (Figure 5b). He decided to pass between N207 and the shoreline. As Rescue 1 came out of its turn, the helmsman’s attention was briefly taken away from N207 and was split between operating the steering in front of him, checking the engine gauges, and controlling the throttles. Figure 4: Location of crew in N207 For illustrative purposes only: not to scale

7 Figure 5: Tracks of boats before collision Illustati on showing Rescue 1 ah ead and clear of N207 and commencing a full-circle turn to port Illustrat ion to show N207 ahead of Rescue 1 in the distance, turning to port and heading downstream Rescue 1 N207 5a 5b Rescue 1 N207 For illustrative purposes only: not to scale

8 After turning N207 to port and now headed downriver, the helmsman saw Rescue 1 closing rapidly in front of him. In an attempt to avoid crossing ahead of it, he pulled the outboard motor tiller sharply towards him to turn the boat further to port. The trainer saw that Rescue 1 was fast approaching, and realised that they were on a collision course. He shouted a warning to the helmsman and pointed at Rescue 1. When Rescue 1’s helmsman looked ahead, he saw N207 close in front of him and attempted to avoid a collision by increasing speed and turning full helm to starboard (Figure 6). At about 1125, despite the attempts of both helmsmen, the two boats collided. Figure 6: Manoeuvring leading to the collision Illustati on showing tracks of t he boats before the collision with an inset showing the angle of impact 6a: Collision For illustrative purposes only: not to scale Tracks of the boats before the collision

9 Rescue 1’s bow struck N207 in the middle of the boat at an angle of about 45° and rode over its buoyancy tube toward its bow (Figure 6a). Rescue 1’s helmsman pulled his kill-cord, which stopped the boat’s engines. The force of the impact threw N207’s helmsman forward, which pulled out his kill-cord and stopped the boat’s engine. The firefighter in the bow of N207 was knocked or thrown into the water. He was struck on the head during the collision and as the boats separated was seen floating motionless on the water. 1.5 EMERGENCY RESPONSE N207’s trainer immediately jumped into the water to attempt to rescue the casualty, but it quickly became apparent to him, from the severity of the injuries, that the firefighter was dead. Rescue 1 was manoeuvred alongside the casualty in the water, and his body was recovered onto the boat using the onboard casualty recovery equipment. The trainer climbed back into N207 and brought it alongside Rescue 1. He then transferred into the bigger boat with N207’s remaining crewman to assist. At 1127, the crew manager who had been acting as crew on Rescue 1 made a telephone call to the fire station watch manager and told him what had happened. At 1129, he called the emergency services and reported the accident. While the phone calls were being made, N207 was placed under the casualty recovery equipment and secured alongside Rescue 1. The deceased’s body was then lowered into the smaller boat, placed in a body bag, and covered over. Rescue 1 then motored slowly downstream toward Neyland yacht club. At 1145, while on passage downstream, the crewman on board Rescue 1 received a call from the local ambulance service to verify that attempts at resuscitation could not be carried out by the firefighters, and to inform them that they would be met on arrival. At 1155 Rescue 1 arrived at Neyland yacht club pontoon, where they were met by members of the police, ambulance service and fire and rescue service (FRS). The casualty was formally declared deceased by a paramedic from the ambulance service. All crew of the two boats were breathalysed by the police; all recorded a zero-alcohol result. A postmortem external examination carried out by a pathologist a few days later confirmed that the casualty had died as a result of a severe head injury. 1.6 BOATS’ CREWS 1.6.1 N207 Firefighter Josh Gardener had joined the FRS on 10 September 2018. Prior to this, he had worked at sea on small craft operating in offshore windfarms. He had completed the Royal Yachting Association (RYA) powerboat level 2 training in June 2019.

10 The trainer held the rank of firefighter and had worked in the FRS since 1996. He held a RYA advanced powerboat operator certificate that had been commercially endorsed, as well as RYA Day Skipper for Sail and Power Craft, RYA radar operator certificates and GMDSS1 radio operator certificate of competency. He was also qualified to instruct both the RYA powerboat level 2 scheme and the FRS WRBO course, and would periodically act as an instructor at the fire service RYA approved training schools. Several of the trainer's qualifications had been obtained by the trainer at his own expense, including his commercial endorsement. The other crew member of N207 also held the rank of firefighter, and had joined the full-time FRS in January 2019. He had previously worked as a retained part-time firefighter since 2013. He had also completed the RYA powerboat level 2 training in June 2019. 1.6.2 Rescue 1 Both crew on board Rescue 1 were very experienced firefighters and both held the rank of crew manager within the fire service. As designated boat operatives, they had both gained the RYA powerboat level 2 qualification and had completed the WRBO training at Menai. Neither held any other marine qualifications. 1.7 MID AND WEST WALES FIRE AND RESCUE SERVICE 1.7.1 Overview MWWFRS has a legal responsibility under the Fire and Rescue Services Act 2004 to make provision for fire-fighting purposes to protect life and property from fire, and to attend road traffic collisions and other emergencies. In addition, the Fire and Rescue Services (Emergencies) (Wales) (Amendment) Order 2017 introduced a duty requiring MWWFRS to make provision to respond to flooding and inland water emergencies. The Department for Environment, Food and Rural Affairs (DEFRA) produced a Flood Rescue Concept of Operations (FRCO), which set out the processes for managing and maintaining flood rescue capability. DEFRA also controlled the process for the national co-ordination of flood rescue assets in England and Wales. Six Zodiac Milpro FC470 boats were operated by MWWFRS and were identified within the national asset register. Their operation and local procedures followed the DEFRA FRCO requirements. 1.7.2 N207 N207 was one of 16 Zodiac Milpro FC470 inflatable boats purchased by North Wales FRS in October 2017 using a grant from the Welsh Government for flood rescue service. It was one of six boats subsequently transferred to MWWFRS in March 2018 as an equipment upgrade to replace older boats that had come to the end of their service life. The six boats were deployed around the MWWFRS region, with N207 being allocated to Milford Haven. All of the boats were fitted with a long-shaft Mariner 30hp 4-stroke outboard engine with throttle-controlled gearbox, a propeller guard, and were supplied on a trailer. This model of boat was also fitted with speed tubes on the hull underwater, which provided lift and lateral stabilisation, 1 GMDSS – Global Maritime Distress and Safety System

11 thereby reducing drag and improving both stability and manoeuvring capability. This type of boat was referred to as a Swift Water and Flood Rescue Technician (SRT) boat. 1.7.3 Rescue 1 Rescue 1 was a Delta 6.4m RIB that came into service with MWWFRS in 2001 to replace a boat that had come to the end of its service life. It was fitted with twin 90hp two-stroke engines, which were not fitted with propeller guards, and was stored on a trailer at the fire station. The engine throttles and gauges were to the right of the steering wheel in the middle of the console in front of the crew (Figure 7). A large windscreen was mounted on top of the console, which was surrounded by a tubular framework (Figure 8). When standing astride the boat’s jockey seat, the helmsman’s view was not obstructed by the windscreen and framework. However, when standing, the helmsman partially obstructed the seated crewman’s field of view to port. Figure 7: Console arrangement of Rescue 1Rescue 1's console Wheel Throttle Engine gauges

12 Figure 8: Windscreen of Rescue 1Rescue 1's console Windscreen Tubular framework The boat originally had a dual function of providing an operational platform for an FRS dive team, and supporting firefighters working on the oil and gas terminal jetties in the Milford Haven waterway. It could also be used as a transport platform to support an emergency response to local offshore islands. To enable it to operate in areas outside of Milford Haven waterway, Rescue 1 was initially coded to the Maritime and Coastguard Agency (MCA) Merchant Shipping (Small Workboats and Pilot Boats) Regulations 1998. However, shortly after the RIB came into service, the dive team was disbanded. As a result, Rescue 1’s certification of compliance to the code was not renewed at the end of its validity period. MWWFRS completed an analysis to establish the need to maintain Rescue 1 as an operational asset subsequent to the disbanding of the dive team and, at that time, it was decided to keep the RIB in service.

13 1.7.4 Boat procedures The MWWFRS documented management system was available to staff via its ‘@Work’ intranet portal. Within this there were two procedures related to the operation of N207 and Rescue 1, each of which had an associated risk assessment. These were Standard Operating Procedure (SOP):

● SOP 2.8 - Flooding (Extracts at Annex A)

● SOP 2.10 - Rescues; Working Near, On, Or in Water sic. SOP 2.8 was mainly concerned with inland water operations and only made reference to Rescue 1 as an asset: Milford Haven also has a larger boat (Rigid Inflatable Boat) that may be suitable for Large Flooding Incidents. SOP 2.10 stated that: This Standard Operating Procedure (SOP) considers hazards, risks and control measures for fire & rescue service personnel attending incidents involving rescues; working near, on, or in water. It incorporates the Generic Risk Assessment outcomes of MWWFRS Risk Assessment RA(SOP)2.10. [sic] SOP 2.10 made no reference to Rescue 1. The risk assessments for each of these SOPs did not include any reference to the various hazards of working in a boat, or using it as a platform for operations. In addition to the SOPs, MWWFRS produced an ‘Appliance, Equipment, Technical Instruction for Fire Service Boats’ (AET) number 5.17 entitled ‘Appliance, Equipment, Technical Instruction for: Fire Service Boats’ (Extracts at Annex C). This document stated: The Service has a duty to make a suitable and sufficient assessment of the health and safety risks to employees and to provide them with comprehensive and relevant information on the risks identified. The document examines the hazards, risks and control measures relating to fire service personnel using the Fire Service Boats. [sic] Section 2.1 of the document defined the qualifications required by MWWFRS to be held by the crew: SRT Boats SRT boats are crewed by personnel trained and competent to RYA (Royal Yachting Association) Level 2 as a minimum. If operating in swift moving water then the crew must be trained and competent to Rescue Boat Operator (RBO) level as a minimum.

14 RIB The only RIB within the service is based at Milford Haven and the crews will hold the following additional qualifications RYA advanced powerboat certificate (for all helms) Radar course GMDSS (Global Maritime Distress Safety System). [sic] While the trainer in N207 held the qualifications required by MWWFRS within AET 5.17 to operate Rescue 1, neither crew manager in Rescue 1 had the stipulated qualifications required to operate the RIB. 1.7.5 Risk assessments All senior station staff within MWWFRS had undertaken formal training in the development of risk assessments, and station staff were very experienced in undertaking dynamic risk assessments. MWWFRS produced a ‘Health and Safety Procedural Guidance Document 5.19 – Risk Assessment’ (Annex D). This document provided guidance for those carrying out risk assessments, and Appendix B of the document specifically focused on training risk assessment. There were no generic risk assessments for undertaking boat training available via the @Work portal. However, several local risk assessments were made available during the investigation. These were task specific risk assessments for boat training exercises that had been conducted between 2005 and 2015, but none of these had been reviewed and none contained reference to the hazard of boat collision. A risk assessment created in July 2018 entitled ‘Water Rescue Boat Operators Training at Milford Haven Waterway’, specifically covered the majority of activities being undertaken by N207 and Rescue 1 on the day of the accident. However, this document was not available to Milford Haven fire station staff via the @Work portal. An internal review of MWWFRS risk assessments undertaken in August 2018 identified that, of the 119 risk assessments listed in the training section, 88 items (74%) were overdue for review. 1.7.6 Control of the activity Prior to undertaking station-based or off-station training or familiarisation, MWWFRS expected that the person in-charge of the activity would ensure that

● a risk assessment would be carried out for the activity, and

● a safety brief would be undertaken for the activity. At the general morning briefing in Milford Haven fire station on the day of the accident, nobody was identified as being in charge of the water-based activities for that morning’s session. Although an informal dynamic risk assessment was undertaken prior to the boats departing the fire station and launching, no formal

15 documented risk assessment was carried out, no reference was made to SOPs 2.08 and 2.10, or AET 5.17, and no formal safety briefing was carried out that included both boat crews. 1.7.7 Training and familiarisation In common with other FRS authorities around the UK, MWWFRS used pdrPro, a commercially available planner-based maintenance of competence system to record their training activities. This electronic recording system was maintained up-to-date by the individual firefighter and monitored by their line manager. In addition, MWWFRS followed the DEFRA FRCO recommendation for skill sustainability training that: ‘personnel spend a minimum of one day per annum at a suitable swift water training venue, under the supervision and tutorship of appropriately competent trainers’ It was MWWFRS’s policy to undertake an internal annual verification of staff boat operations’ competence using qualified instructors from its own RYA approved and accredited training school. At Milford Haven, the last time that this verification had been undertaken was on 17 July 2017. Milford Haven fire station had an extensive daily training plan for all firefighters to maintain competence in all aspects of their work and core skills (Figure 9). This also included training and familiarisation with N207 and Rescue 1. This boat-specific work was intended to be a combination of time at the fire station and on-water training and familiarisation. The daily plan was necessarily superseded by operational emergency call-outs, but was intended to be sufficiently frequent that MWWFRS training requirements and parameters set within the pdrPro system were met. Firefighters participating in on-water training were not available to respond immediately to emergency call-outs, unlike those undergoing training on land. To enable on-water training and familiarisation with the boats, off duty staff were required to be called in, or cover provided by another station, potentially increasing response time. This made the ability to conduct on-water training and boat familiarisation difficult to achieve, unless the boats were called out for emergency response. In the 12 months prior to the accident, N207 had been called out for emergency response seven times and had undertaken one day of training/familiarisation. In that same 12-month period, Rescue 1 had not been used operationally, but had undertaken 5 days of training/familiarisation. Most of the training/familiarisation had taken place within the Milford Haven waterway, however Rescue 1 had undertaken a training exercise on 21 June 2019 at Caldey Island (Figure 10), and boat familiarisation in August 2019 that had taken it beyond the Milford Haven waterway, and to sea.

16 Figure 9: Training matrixMilford Haven fire station's extensive daily training plan

17 Figure 10: Location of Caldey Island Map location of Caldy Island Caldey Island Image courtesy of Google Earth As part of their WRBO training, firefighters were expected to be able to work with other SRT craft, but not RIBs such as Rescue 1, and it focused on flood rescue scenarios. In the 18 months prior to the accident, N207 and Rescue 1 had been on only one joint training and familiarisation activity, on 10 July 2018, which had taken place in Milford Haven waterway. All full-time firefighters at Milford Haven fire station were required to have completed the RYA powerboat level 2 certificate of competency and the WRBO certificate. There was no evidence found that MWWFRS had provided any additional training since 2004 for firefighters to obtain the RYA advanced powerboat certificate, radar qualifications or GMDSS radio operator required by MWWFRS to operate Rescue 1. This meant that as firefighters had left the service at Milford Haven, the number of firefighters qualified to the MWWFRS required standard to operate Rescue 1, was diminishing. The lack of a sufficient number of firefighters qualified to this higher standard had been raised with MWWFRS regional management several times, most recently by local management at Milford Haven in June 2019. 1.7.8 Audit, inspection and peer review MWWFRS undertook an internal audit of Milford Haven fire station on 9 January 2018 and a station inspection on 29 November 2018. The internal audit used a standard checklist that had not been adapted to include the equipment and training required at the stations operating boats. Omissions included the operation of flood rescue craft and the unique status and requirements at Milford Haven for operating Rescue 1. Consequently, there was no evidence that the operational status of N207, Rescue 1 or the associated procedures and risk assessments were considered during the internal audit.

18 The station inspection in November 2018 included an exercise of SRT techniques undertaken in the exercise yard at Milford Haven fire station. In the report it was noted that the demonstration ‘highlighted the skills of the crews’ and a comment that ‘Overall an excellent standard from a Training and Delivery perspective’. The inspection report included the following areas: 6.1 Operational Risk Management, and 6.2 People and Organisational Development & Training Delivery The lack of practical on-water training, lack of suitably qualified rescue boat crew, and the lapsed certification and operating standards for Rescue 1 were not identified as non-conformities in the internal audit report. During 2017 and 2018, peer review assessment teams, drawn from other UK based FRS regions, visited MWWFRS as part of the All Wales Operational Assurance (AWOA) process. The final report was circulated to the Chief Fire Officer in May 2018 and stated that: ‘On introduction of the statutory requirement to respond to water related incidents a small number of stations were enhanced in their water response capabilities. This additional training was absorbed without difficulty into the T&D2 Department, however it was apparent that should any further significant enhancement in equipment introduction and associated personnel training over and above of current levels would have a significant impact on the department with respect to the number of available trainers/Instructors for initial acquisition training and ongoing assessment and reaccreditation. Also this would have a significant impact on the T&D training delivery annual plan.’ [sic] and ‘Appropriate levels of initial acquisition, maintenance of competence and re accreditation via the RYA and WRBO are well established in line with the national requalification standards and timeframes.’ [sic] The peer review took into account the flood rescue capability and it referred to national standards required under the DEFRA FRCO. It did not, however, take into account the operational standards of training and certification required for Rescue 1 as a unique asset at Milford Haven. 1.8 REGULATION AND OVERSIGHT OF FRS BOAT OPERATIONS 1.8.1 Overview Although the majority of boats operated by various FRSs in the United Kingdom fall into the SRT category, and comply with the standards required by the DEFRA FRCO, several authorities also operate larger RIBs, similar to Rescue 1, and some authorities operate specialist fire-fighting boats in port areas. As fire service boats are not operated for pleasure, for legislative purposes they are classified by the MCA as commercial craft. 2 T&D – Training and delivery

19 1.8.2 National Fire Chief’s Council The National Fire Chief’s Council (NFCC) represents the whole of UK FRS with the aim of improving national co-ordination, reducing duplication of work and supporting local delivery of service. The chair of the NFCC is a single point of contact for the UK government, fire professionals and local authorities. The programme of work for the NFCC is coordinated by various committees who engage with subject matter experts. The NFCC’s national operational guidance documents were produced by its committees and were promulgated via its website3. Guidance for dealing with fires on ships and flood water rescue operations was available on this website. The national operational guidance did not provide information about certification of boats or categorisation of waters around the UK. 1.8.3 Regulations for boat operations The operation of small commercial craft around the UK is regulated either by the MCA or by the local authority, depending on whether the craft is operating ‘at sea’ or in ‘categorised waters’. Categorised waters include near coast, estuaries, rivers and other inland waters that are graded from A (most sheltered) to D (least sheltered), as defined in Merchant Shipping Notice (MSN) 1827 (M) Amendment 2. Small commercial vessels of less than 24m in length, that carry no more than 12 persons on board and proceed to sea, are required to comply with The Merchant Shipping (Small Workboats and Pilot Boats) Regulations 1998 (as amended), which is applied by the Workboat Code Edition 2, or The Merchant Shipping (Vessels in Commercial Use for Sport or Pleasure) Regulations 1998 as applied by MGN 280. Small commercial craft operating in categorised waters are regulated by the local authority responsible for those waters. The local authority sets its own requirements for commercial boat operations, which could include requiring the craft to comply with the Workboat Code Edition 2 or MGN 280. Milford Haven waterway within a line drawn from South Hook to Thorn Point was classified as Category C waters, so MWWFRS boats operating in this area were required to comply with regulation set by the local port authority. 1.8.4 Other boat operating codes The MCA allows police boats and rescue boats (such as those operated by the Royal National Lifeboat Institution) to operate under either the MGN 518(M) Police Boat Codes, Amendment 3, or the MGN 466(M) Rescue Boat Code, as applicable. These codes were developed between the MCA and the relevant organisations in recognition of the specific status, equipment and training required for these craft to operate effectively at sea, and the codes are accepted by local authorities as meeting their requirements to operate in categorised waters. In the absence of a specific code for FRS boats, MWWFRS boats operating at sea were required to comply with the Workboat Code. When operating in Milford Haven waterway and the River Cleddau, MWWFRS boats were required to comply with any regulations set by the local port authority. 3 www.ukfrs.com

20 1.9 PORT AUTHORITY Milford Haven Port Authority (MHPA) is the statutory harbour authority for Milford Haven waterway and the tidal region of the River Cleddau, including the section of the river where the accident occurred. MHPA is responsible for the navigation, safety and traffic regulation of all vessels navigating within this area. This is achieved through the use of byelaws, General Directions, Special Directions and Notices to Mariners, and included the requirement that all commercial and leisure users within the MHPA area adhere to the International Regulations for Preventing Collisions at Sea 1972 (as amended). Rule 5 of these regulations states: Every vessel shall at all times maintain a proper look out by sight and hearing as well as by all available means appropriate to the prevailing circumstances and conditions so as to make a full appraisal of the situation and of the risk of collision. MHPA had designated the area upriver from the Cleddau bridge to the vicinity of Lawrenny as a water-ski area and it was not subject to speed restrictions. Various river areas, including upriver past Lawrenny, were designated as ‘Dead Slow Minimum Wake’ zones within the MHPA Leisure Guide (Annex E). These minimum wake zones were routinely patrolled by an MHPA water ranger to ensure compliance. As the local authority responsible for ensuring standards of operations, MHPA generally adopted the MCA coding and certification requirements for commercial vessels operating in its area. Once the coding for Rescue 1 had lapsed, an agreement had been put in place between MHPA and MWWFRS to continue to operate Rescue 1 within the MHPA area without the requirement for it to be coded. This agreement had never been extended to cover the SRT boats such as N207, and no agreement was in place with any other port authority within the MWWFRS area to allow them to operate without adhering to local commercial craft coding and certification requirements. 1.10 PREVIOUS/SIMILAR ACCIDENTS 1.10.1 Collision between two RIBs Osprey and Osprey II (MAIB Report No 10/20174) Two passenger carrying RIBs, Osprey and Osprey II, collided in the Firth of Forth. A passenger who was sitting on an inflatable tube of Osprey II was crushed between Osprey’s bow and Osprey II’s helm console, resulting in her sustaining serious injuries. While proceeding in parallel at a speed of around 6 knots (kts), the skipper of each RIB increased speed and commenced a power turn away from each other with the intention of passing each other in the course of completing a round turn. However, as the RIBs turned towards each other it became apparent to both skippers that the RIBs were in danger of colliding. Although they both acted quickly to reduce the speed of their respective vessels and so lessen the impact, they were unable to prevent the collision. The manoeuvre had previously been carried out successfully on several occasions but it had not been formally risk assessed, and no thought had been given regarding what to do if a collision situation developed. 4 https://www.gov.uk/maib-reports/collision-between-rigid-inflatable-boats-osprey-and-osprey-ii-resulting-in- serious-injuries-to-1-passenger

21 1.10.2 Two Cardiff Bay Yacht Club RIBs (MAIB Report No 19/20115) Two club RIBs belonging to Cardiff Bay Yacht Club collided at night while transporting a number of children across Cardiff Bay. The RIBs were proceeding at about 20kts in the dark, and carried no navigation lights or torches. One RIB had one seat and six occupants; the other, two seats and seven occupants. Those without designated seats were sitting on the inflatable tubes. As a result of the accident, three children, who were sitting on the inflatable tubes, were thrown into the water. One of the children sustained a traumatic brain injury, two others required extended medical treatment and several of the children from both boats suffered bruising and soreness following the collision. 5 https://www.gov.uk/maib-reports/collision-between-2-rigid-inflatable-boats-while-transporting-a-number-of- children-across-cardiff-bay-wales-with-3-people-injured

22 SECTION 2 - ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 OVERVIEW The FRS boats N207 and Rescue 1 collided because they were carrying out uncoordinated manoeuvres, at speed, in the same stretch of water and neither helmsman realised the risk until it was too late. Firefighter Josh Gardener died because he was struck on the head by Rescue 1 before or after being knocked out of N207 into the water. This section of the report will analyse the circumstances that led to the collision, the reasons why the boats were being operated in the same stretch of water, and the underlying factors that might have contributed to the accident and resultant loss of life. 2.3 THE COLLISION Once the two boats had left the vicinity of Mill Bay, they were headed for the same agreed location near Lawrenny, with Rescue 1 ahead and clear of N207 until its helmsman decided to execute a large full-circle turn to port. 2.3.1 Rescue 1 It was unclear why Rescue 1’s helmsman decided to undertake the full-circle turn given its proximity to N207, rather than proceeding upstream to the agreed rendezvous location. This manoeuvre resulted in the need for Rescue 1 to subsequently overtake or pass close to N207, which was conducting a series of turns under the supervision of the trainer. The turn to port took Rescue 1 downstream of N207, so it approached the slower boat from behind. When the helmsman saw N207 heading downstream toward Mill Bay as he executed the final stages of the turn, he thought that it was going to cross clear ahead of him and did not expect N207 to suddenly turn into his path. The crewman was not keeping an effective lookout and keeping the helmsman informed of the location of N207, so as Rescue 1 came out of its turn and the helmsman looked up, he was surprised to see that collision with N207 was imminent. The emotional and behavioural response to being surprised might have led to delayed collision avoidance action and disrupted problem solving. His reaction time to implement his decision - to accelerate and turn sharply to starboard - would have been affected by his competence and familiarity with Rescue 1 and its controls. The qualification to RYA powerboat level 2 should have been sufficient to drive the boat, however the helmsman’s lack of practical experience would have diminished his level of competence. Given the speed Rescue 1 was travelling, it would have been prudent to have planned a greater passing distance from the slower boat, whose exact intentions were unknown.

23 2.3.2 N207 The three crew of N207 were aware that Rescue 1 had accelerated away from Mill Bay and away from their boat. However, their collective attention then returned to the boat handling training and practice for the two firefighters. The positions of the trainer in the middle and the firefighter forward in the boat meant that they could both see forward, to starboard and astern, but neither had an easy view to port without physically moving position. The helmsman steering the boat was seated on the starboard side tube aft and had a view to port, but was focused on practice manoeuvres. As a result, during Rescue 1’s turn to port, it went out of the immediate view of the crew of N207. Because none of N207’s crew were keeping a proper lookout, the helmsman was completely unaware of Rescue 1’s position when he initially turned to port and headed downstream. After seeing the position and direction of travel of Rescue 1, the helmsman made the decision to turn further to port to avoid crossing ahead of the approaching RIB. However, because of the closing speed between the two boats, there was insufficient time for the helmsman to bring his boat further to port out of the way of Rescue 1. 2.4 LOOKOUT The lack of effective lookout and awareness of each boat’s position led to the collision situation arising. On N207 the crew were focused on their training, part of which should have been to nominate a member of the crew to keep a good lookout. With only two boats on the water in that locality, the lookout should have been able to provide the helmsman with information on Rescue 1’s position, course and speed. With that information the helmsman would not have turned into the path of Rescue 1. On board Rescue 1, the crewman had not been tasked with keeping a lookout and was not paying attention to the activity of N207. Had he done so, the risk of collision might have been identified earlier, brought to the attention of the helmsman, and the collision avoided. In the RYA publication, ‘Start Powerboating’, written to accompany the powerboat level 2 course, it states: When intending to travel at higher speeds, you must keep a really good all-round look-out. Check it is clear before you turn and be aware of other water users... In close proximity and with the relative speeds involved, there was little time to make decisions and take evasive action, especially given the level of competence and experience of most of the crew involved. 2.5 PLANNING AND CONTROL OF THE TRAINING AND FAMILIARISATION EVOLUTIONS MWWFRS procedures for any training activity required that a full briefing was undertaken prior to commencement. This briefing would be expected to include reference to any specific procedures related to the activity, any guidance and risk assessments, identification of the person in-charge and a full discussion of the intended activity. Because the decision to undertake the training/familiarisation

24 was agreed on the evening prior to the accident, there was little time to plan it thoroughly. The firefighters had assembled in the morning and held an informal discussion about the activity for the day, but because there was no formal briefing, the MWWFRS’s pre-activity planning requirements were not met. Additionally, although the trainer was identified as being in charge of the training undertaken on N207, neither of the crew of Rescue 1 were nominated as being in charge of that boat, and no individual had been identified as being in overall charge of the activity on-water that morning. Consequently, the morning’s activities were not formally planned, and the role of each boat was not discussed. With the two firefighters planned to undertake the training course in Menai the following day, the purpose and structure of the training and activities being undertaken by N207 on the day of the accident were clear, and under the control of a qualified trainer. What was unclear was the purpose and intention of the two crew managers crewing Rescue 1. Unlike N207, the crew of Rescue 1 had no plan of activities and there was no agreement to support, or be part of, N207’s training. Had formal planning and briefing taken place, steps might have been taken to avoid both boats operating at speed in the same part of the river. 2.6 QUALIFICATIONS, CREWING AND TRAINING 2.6.1 Overview The firefighters’ qualification to RYA powerboat level 2 provided them with training that was sufficient to operate boats similar to N207 and Rescue 1 as leisure users. However, since the firefighters were at work, they were obliged to comply with the higher qualification standards set by MWWFRS for operating its boats. MWWFRS’s procedures did not distinguish between the crewing requirements to be met when the boats were on operations, and when staff were undergoing training or accruing boat familiarisation time in controlled conditions. These requirements included both the numbers of crew per boat, and the level of qualifications they held. For Milford Haven fire station, which operated two boats, it was difficult to undertake any afloat training or boat familiarisation activities while complying with the operational qualification and crewing requirements. This difficulty was evidenced by the few times that either boat had been used solely for training or familiarisation in the 12 months prior to the accident at Milford Haven. 2.6.2 Training The Health and Safety Executive defines competence as being: the combination of training, skills, experience and knowledge that a person has and their ability to apply them to perform a task safely. The records of training within the MWWFRS pdrPro software recorded only that firefighters had been training with the boats, but did not record how much time any individual spent in charge or helming. The firefighters had initially been trained to a suitable standard, but the systems were not in place to maintain or build their levels of competence. There was a lack of opportunity to undertake training, to practise and hone boat skills, gain experience

25 and demonstrate knowledge. Further, the skill sustainability training that had been undertaken was focused on the SRT boats and WRBO requirements. It did not include the unique challenges of operating Rescue 1, which was the sole asset of its type within the MWWFRS area. Finally, the assurance mechanisms for checking crew competence were not effective as the MWWFRS annual verification of continued competence had not been undertaken, and was 14 months overdue. At the time of the accident, it is possible that some of Milford Haven’s firefighters were suffering from a gradual erosion of their levels of practical boat handling competence, and a general reduction in their level of safety awareness for operating their boats; in particular, Rescue 1. 2.6.3 Management oversight of training and certification MWWFRS required all firefighters crewing boats to hold the RYA powerboat level 2 qualification as a minimum, but these requirements were enhanced for the helmsman operating Rescue 1 as the boat’s operational role required it to operate offshore day and night. Had effective planning and briefing been carried out on the morning of the accident, with reference to the extant procedures, the senior officers might have recognised that there were insufficient qualified staff allocated to conduct the planned training with N207 and Rescue 1 simultaneously, and that neither of the crew managers were qualified to helm Rescue 1. The opportunities for MWWFRS management to recognise that there was a shortfall in training and familiarisation for firefighters at Milford Haven were missed several times during audits, inspections and reviews. It is apparent that once Rescue 1 had ceased to be required as a support craft for the disbanded dive team, its operational need and function were diminished. Operational documentation made scant reference to it, and the higher standard of training required for it to operate offshore and be commercially coded was no longer provided. 2.7 PROCEDURES 2.7.1 Personal protective equipment The firefighters in both boats were provided with protective equipment. These were detailed in MWWFRS procedures and as mitigation measures within risk assessments. All firefighters were wearing dry-suits with fleece undersuits, a buoyancy aid or lifejacket, and quick release heavy boots. However, on the morning of the accident, the decision was made to not wear protective headgear on either boat. The reason why this decision was taken is unclear, but some firefighters reported the helmets to be uncomfortable to wear and to hamper communications. In addition, it was a warm and sunny day. The headgear provided on N207 was of a sufficient standard to provide some impact protection. Nevertheless, given the force of the impact and the extent of the injuries suffered, it is unlikely that the wearing of the head protection provided would have prevented the firefighter’s death. By not wearing helmets, none of the firefighters on either boat complied with the personal protective equipment requirements contained within section 2.2 of AET 5.17 (Annex C).

26 2.7.2 Suitability of procedures and risk assessments MWWFRS SOPs 2.08 Flooding and 2.10 Rescues: Working Near, On, or In Water, focused on flood rescue and the boats used for that purpose. Neither SOP covered the use of Rescue 1 to respond to incidents in Milford Haven or offshore islands. Furthermore, neither of the risk assessments associated with these procedures made any reference to risks associated with operating boats. They provided very general reference to existing control measures that offered little guidance to the reader. Any additional control measures required to mitigate the risk were limited to a generic ‘Regular assessment required’ statement. Milford Haven fire station did not have local SOPs for routine boat handling familiarisation or for water rescue training. Local risk assessments had been created, but these had been for specific activities. AET 5.17 contained more useful information and requirements, but had not been updated for some time. The information concerning the SRT boats within it was incorrect, referring to boats that had been replaced by the Zodiac Milpro 470. There were other weaknesses in the procedures because by not differentiating between familiarisation, instruction and operational call-out crewing requirements, staff were provided with inconsistent messages. AET 5.17 contained an illustration of operations with an SRT boat manned by only one person, and promotional photographs of FRS staff undertaking WRBO training in Menai showed boats with only two FRS crew, contrary to the MWWFRS requirement that all boats were to be operated by three crew. While standards for emergency operation will dictate stringent requirements, the ability to train and conduct familiarisation under the same procedures was difficult, and at local level this led to the divergence from MWWFRS SOPs. 2.8 BOAT OPERATING CODES AND GUIDANCE As part of the investigation, several UK FRSs were contacted to obtain information on the operating standards adopted for their boats. For SRT boats the DEFRA FRCO formed the basis of standards across the FRSs. However, because there was no common approach across the UK to any adopted standard for larger craft, there was a variety of responses. Some boats were certificated to the workboat code, some were not certificated but operated to the standards of the rescue boat code, and some craft were not operating to any code. It was also apparent that various port authorities made assumptions about the operating standards of the FRS craft in their area of responsibility and did not verify to what standard the boats were being operated. Rescue 1 was well maintained and fully equipped for its intended tasks. However, any FRS boat being operated at sea outside categorised waters, including boats such as N207, is required to adhere to a standard agreed with the MCA. In addition, when operating on its categorised waters, most port authorities require commercial craft to be operated to standards of certification that meet the MCA requirements for coded craft. As the SRT boats frequently operate outside the DEFRA FRCO and within port areas, they must meet these requirements. It was apparent that there was no awareness by MWWFRS that local authority requirements to operate all

27 of its boats to an approved standard should apply. For Rescue 1, the lapse in the workboat code certification meant that mobilisation to offshore islands, such as the training exercise at Caldey Island earlier in 2019, and the training in August 2019 did not comply with the MCA certification requirements. Appropriate certification of its boats would provide a greater level of assurance regionally and nationally that FRS boats are being operated safely by trained and experienced staff, and that the boats are properly maintained and equipped. To assist the various FRS authorities around the UK, guidance should be provided by the NFCC. Their operations committee for marine operations and water rescue should consult with the MCA and stakeholder authority representatives for categorised waters, regarding a common approach to the application of an appropriate standard.

28 SECTION 3 - CONCLUSIONS 3.1 SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

The firefighter on board N207 was fatally injured after being struck on the head by Rescue 1 when the two boats collided. [2.2]

N207 and Rescue 1 collided because they were operating at speed in close proximity when N207 turned into the path of Rescue 1. [2.2]

It is unclear why Rescue 1’s helmsman decided to undertake a circular turn to port in the vicinity of N207, but given the speed the RIB was travelling it would have been prudent to have planned a greater passing distance from the slower boat, whose exact intentions were unknown. [2.3.1]

The crew of N207 were focused on their boat handling training, and the helmsman was unaware of Rescue 1’s position when he turned sharply into its path. Due to the closing speed of the two vessels, his subsequent actions to avoid collision were not effective. [2.3.2]

Neither crew was keeping an effective lookout, and so lacked awareness of the two boats’ relative positions and movements. [2.4]

Mid and West Wales Fire and Rescue Service’s pre-activity planning requirements were not met and its SOPs were not followed, with the consequence that: no individual had responsibility for the overall activity; no-one was nominated to be in charge of Rescue 1, which had no clear task; Rescue 1 had insufficient crew, neither of whom held the qualifications required to be the RIB’s helmsman; and, no steps were taken to prevent both boats operating in the same stretch of water in an uncoordinated manner. [2.5, 2.6.3]

It was difficult for Milford Haven fire station, which operated two boats, to comply with Mid and West Wales Fire and Rescue Service crewing requirements, which reduced the amount of on-water training and familiarisation conducted in the 12 months prior to the accident. [2.6.1]

At the time of the accident it is possible that Milford Haven’s firefighters were suffering from a gradual erosion of the levels of practical boat handling competence, and a general reduction in their levels of safety awareness for operating their boats; in particular, Rescue 1. [2.6.2]

Although the firefighters crewing the boats were not wearing the head protection required by AET 5.17, it is unlikely that head protection would have been sufficient to save the firefighter’s life when he was struck by Rescue 1. [2.7.1]

Mid and West Wales Fire and Rescue Service’s procedures for boat operations had not been updated for some time, contained misleading information, and made insufficient reference to Rescue 1 as a pre-determined operational asset. [2.7.2]

29

While the standards for emergency operation will dictate stringent requirements, the ability to train and conduct familiarisation under the same procedures was difficult, and at local level this led to the divergence from Mid and West Wales Fire and Rescue Service’s SOPs. [2.7.2]

The investigation has found that some port authorities made assumptions that FRS craft operating in their areas complied with an approved standard. However, it was apparent that Mid and West Wales Fire and Rescue Service had overlooked the requirement to operate all of its boats to an approved standard, or that within a local authority or harbour area locally required standards could apply. [2.8] 3.2 SAFETY ISSUES NOT DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

The recording of training and monitoring of competence was not robust enough to identify when a firefighter was helming a boat, and therefore ensure that sufficient experience was logged. [2.6.2]

The chance to recognise that shortfalls in training, experience, qualifications and boat certification by Mid and West Wales Fire and Rescue Service was missed during internal audits and inspections and through external peer review. [2.6.3]

Mid and West Wales Fire and Rescue Service standard operating procedures did not contain information relevant to the operation of the Zodiac Milpro 470 or the Delta 6.4m RIB. Furthermore, it did not distinguish between requirements for using the boats when staff were undertaking familiarisation, instruction or were on emergency call-out. [2.7.2]

The Fire and Rescue Service was not fully aware of the certification requirements for commercially operated boats operating in categorised and non-categorised waters, and hence Rescue 1 had been operated at sea without the necessary levels of certification. [2.8]

Standards for Fire and Rescue Service boats operating in categorised waters and at sea had been overlooked nationally. [2.8]

30 SECTION 4 - ACTION TAKEN 4.1 ACTIONS TAKEN BY OTHER ORGANISATIONS Mid and West Wales Fire and Rescue Service has:

● Appointed an independent peer review to be conducted of water-based activities by another regional Fire and Rescue Service.

● Appointed its own investigation team and commenced an internal investigation.

● Reviewed and revised its procedures 2.1, 2.08, 2.10, and AET 5.17.

● Initiated a full review of water-based requirements.

● Completed an assessment of the future provision of Rescue 1 and taken the decision in April 2020 to remove it from service.

● Reviewed the risk assessments within Mid and West Wales Fire and Rescue Service stations for special appliance activities.

● Secured funding to provide staff with Institution of Occupational Safety and Health (IOSH) and National Examination Board in Occupational Safety and Health (NEBOSH) training.

● Engaged Kent Fire and Rescue Service to undertake peer review of boat operations.

● Set up a group to discuss introducing additional on-watch boat trainers and to make recommendations to senior management.

● Commenced a risk management plan for incidents at offshore islands, including consideration of commercial contract for the transportation of firefighters if needed.

● Engaged with regulatory authorities concerning operational standards for Fire and Rescue Service boats.

● Upgraded the standard of helmets used as Personal Protective Equipment on its SRT boats from Nitro XT to Manta SAR helmets.

● Met with stakeholders in the Milford Haven waterway area to clarify the provision of a safety boat during incidents.

● Engaged with a local tug operator with the aim of establishing a memorandum of understanding for the provision of emergency response to incidents in the Waterway.

● Commenced a gap analysis to encompass the historical evolution of business planning process over the previous 20 years.

31 SECTION 5 - RECOMMENDATIONS The Mid and West Wales Fire and Rescue Service is recommended to: 2020/130 Undertake a review of the crewing and staff qualification requirements for boats within MWWFRS to determine appropriate levels for familiarisation, training and emergency operations status and include the requirement within revised procedures and guidance. 2020/131 Introduce a method of recording time spent as helmsman of its boats and implement a minimum number of hours required as helmsman to maintain competency. 2020/132 Include reference to its boats within the internal audit and inspection report process. The National Fire Chief’s Council is recommended to: 2020/133 Consult with the Maritime and Coastguard Agency and the UK Harbour Master’s Association to introduce a standard code for the operation of all fire and rescue service craft when in categorised or non-categorised waters. Safety recommendations shall in no case create a presumption of blame or liability

Annex A Extracts from Standard Operating Procedure 2.08 - Flooding

Standard Operating Procedure for: 2.8 Mid & West Wales Fire & Rescue Service SOP - 2.8 Standard Operating Procedure for: Flooding This Standard Operating Procedure (SOP) considers hazards, risks and control measures for fire & rescue service personnel attending incidents involving flooding. It incorporates the Generic Risk Assessment outcomes of MWWFRS Risk Assessment RA(SOP)2.8. AIDE MEMOIRE AIDE MEMOIRE 1 SCOPE 1.1 Fire & rescue response to flooding 1.2 Statutory Duties 2 SPECIFIC HAZARDS & RISKS 2.1 Flowing water 2.2 Primary hazards associated with flooding 2.3 Large blocked culverts 2.4 Flash floods 2.5 Working conditions 2.6 Water borne diseases 2.7 Water temperature 2.8 Specialist plant & equipment 2.9 Manual handling 3 KEY CONTROL MEASURES 3.1 Pre-planning 3.2 MWWFRS response to flooding 3.3 MWWFRS appliances & equipment

1. 4 Personal protective equipment (PPE)
2. 5 Safe systems of work in floodwater
3. 6 Training
4. 7 Flood & weather warnings
5. 8 Closing Highways and Leaving the Scene

4 OPERATIONAL PROCEDURES 4.1 Prior to mobilisation 4.2 En-route 4.3 Incident plan 4.4 Equipment & PPE for flooding incidents 4.5 The correct use of MAWWFRS flood response assets 4.6 High volume pumps (HVPS) 4.7 Sandbags 4.8 Urban search and rescue (USAR) 4.9 Requesting helicopter assistance 4.10 Low risk flooding 4.11 Body recovery 4.12 Incident command unit 4.13 Multi-agency command 4.14 Personal hygiene 4.15 Crew welfare & reliefs 4.16 Activation of vehicle airbags at flooding incidents

5 ADDITIONAL CONSIDERATIONS

1. 1 EAW flood warning codes

6 REFERENCES 6.1 Sources of Information 6.2 Related Documents 6.3 Questions & Answers

• SCOPE

1. 1

Fire & rescue service response to flooding Flooding affects every part of a community’s normal daily existence, from shelter in domestic properties, to services such as fuel & power supplies, transport networks and communications such as telephone/computer lines. Flooding can also lead to the contamination of food and water supplies. The Fire & Rescue Service (FRS) does not have a statutory responsibility to attend flooding incidents, however Mid and West Wales Fire & Rescue Service (MWWFRS) will respond to calls for assistance in connection with flooding, including rescues from floodwater and pumping out. This SOP should be read in conjunction with SOP 2.1 – Rescues from Ice & Unstable Ground, and SOP 2.10-Rescues: Working Near, On, or In Water.

1. 2

Statutory Duties Historically the three Welsh Fire and Rescue Authorities (FRAs) respond to flooding and water rescue incidents using their existing general powers to respond to ‘other eventualities’ that cause or are likely to cause death, injury or illness, as described in Section 11 of the Fire and Rescue Services Act 2004 (‘2004 Act’). There was, however, no statutory requirement for them to do so. The Fire and Rescue Services (Emergencies) (Wales) (Amendment) Order 2017 (“the Order”) introduced a duty requiring the three Welsh FRAs to make provision to respond to flooding and inland water emergencies. Rescue and protection in case of flooding and inland water emergencies (1) A fire and rescue authority must make provision in its area, to the extent that it considers it reasonable to do so, for the purpose of – a. rescuing people, or protecting them from serious harm in the event of an emergency involving flooding; and b. rescuing people in the event of an emergency involving inland water.

(2) The duty within paragraph (1) applies to an emergency falling within section 58(a) of the Fire and Rescue Services Act 2004. (3) In this article – “flooding” means any case where land not normally covered by water becomes covered by water; “inland water” means – a. rivers, streams and canals; and b. lakes, reservoirs and water-filled quarries.” c. [back to top]

• KEY CONTROL MEASURES

1. 1

Pre-planning Pre-planning within MWWFRS for an emergency response to flooding combines training and equipment provision for crews, with spate-condition procedures. Where widespread flooding is likely to affect one or more areas across MWWFRS, Service Control may take necessary measures to deal with the larger amount of calls coming in from the public. Mapping flood plain overlays for the region are considered from the perspective of both Business Continuity Management (i.e. the likelihood of fire stations being affected by flooding) and for emergency response/assistance to members of the public. Pre-planning also includes an advance weather warning notification system for operational personnel. Integrated (multi-agency) major flood response The operational or physical response to a flood disaster will initially involve the emergency services as they will receive the first calls for assistance from the public. In the early stages, the priority will be the rescue of casualties from life threatening situations, removing victims from the affected area and mitigation of damage to property. In major flooding the extent of the area affected and the demands on the emergency services can over extend their resources, and it may take some

time to identify priorities and establish a robust and efficient command structure. In such cases, a coordinated major flood response will potentially involve Local Authorities, the Fire & Rescue Service, Police, Ambulance, Maritime Coastguard Agency, Natural Resources Wales, and military personnel, together with volunteer organisations such as the Women’s Royal Voluntary Service, the British Red Cross and the Salvation Army. Detailed plans for dealing with major flooding will in general be prepared and maintained by a Local Authority Emergency Planning Officer. These plans will include: • The initial emergency response; • Rescue operations; • Provision of specialist equipment; • Reception areas and rest centres; • Establishing police casualty bureaus; • Provision of information through the media; • VIP visits; • The recovery phase; • Finance. 1 Pump or Major pump (Nearest resource Mobilised)

1. 2

Flood response assets of MAWWFRS The following team types have been established in the MAWWFRS region to deal with large scale flooding incidents. These teams have the capabilities as outlined in the DEFRA Flood Rescue National Enhancement Project – Flood Rescue Concept of Operations. Type E teams – Bank based All front line appliances can provide this response. The teams can provide the following functions at flooding incidents: • Pumping Operations. • Bank Based Rescue. • Shallow Water Wading (less than height of fire boot) in low energy water. Type D teams – Wading These stations have the capability to perform wading duties (low energy water and less than waste height) in order to perform rescues, provide humanitarian services and assist with pumping operations during major flooding incidents. These teams are located in the following stations: • Newtown • Llanfair Caereinion

• Talgarth • Lampeter • Cardigan • Fishguard • Pembroke Dock • Ponturdulais • Morriston • Neath • Glynneath • Haverfordwest • Machynlleth • Pontardawe • Llanelli • Ammanford • Llandysul • Llandovery • Llandrindod Wells • Haverfordwest • Fishguard • Pembroke Dock

The High Volume Pump crew based in Ammanford will also have the capability. Type C– Swift water rescue technician Swift water rescue technician (SRT) teams will have the additional capabilities to allow personnel to perform contact rescues by means of tethered swimming. They will also have the equipment to perform technical rescues using tethered boats. The teams are currently based in the following stations: • Carmarthen • Hay on Wye • Aberystwyth • Swansea Central • Welshpool Service boat stations The Service has powered safety boats (inflatable) that are suitable for deployment during flooding incidents, and are based in the following locations: • Swansea Central. • Carmarthen. • Aberystwyth. • Hay on Wye. • Milford Haven

• Welshpool Milford Haven also has a larger boat (Rigid Inflatable Boat) that may be suitable for Large Flooding Incidents.

1. 3

MWWFRS appliances/equipment Appliances/equipment available within MWWFRS which can be deployed at flooding incidents include: • Major and portable pumps; • High Volume Pumps (HVPs); • Inflatable walkways; • Safety boats; • Personal flotation devices (lifejackets and buoyancy aids); • Dry suits; • Hose inflation kits; • Throw lines.

1. 4

Personal protective equipment (PPE) Depending upon the Level of Response, PPE for use at flooding incidents includes: • Wet weather clothing. • Lifejackets -all appliances. • Dry suits - boats, swift water rescue (Type C) and wading personnel (Type D). • Specific safety headwear and gloves. • Powered boat operators PPE.

• REFERENCES

1. 1

Sources of Information • Generic Risk Assessment 2:8 Rescues – Involving Flooding •

1. 2

Related Documents • SOP 2.1 – Rescues from Ice & Unstable Ground • SOP 4.5 - Use of Helicopters at Incidents • SOP 2.10 – Rescues In or near Water Environment Agency Wales – Flood Warning Plans for local authorities

ORIGINAL AUTHOR STATION MANAGER GRIFFITHS 613 ORIGINAL ISSUE SEPTEMBER 2010 REVIEWED BY STATION MANAGER C THOMAS 612 LAST REVIEWED NOVEMEBER 2016 AUTHORISED BY GROUP MANAGER MCLINDEN CURRENT VERSION VERSION 1 [back to top]

Annex B Extracts from Standard Operating Procedure 2.10 - Rescues; Working Near, On, Or In Water

Standard Operating Procedure for:: 2.10 Mid & West Wales Fire & Rescue Service SOP - 2.10 Standard Operating Procedure for: Rescues; Working Near, On, Or In Water This Standard Operating Procedure (SOP) considers hazards, risks and control measures for fire & rescue service personnel attending incidents involving rescues; working near, on, or in water. It incorporates the Generic Risk Assessment outcomes of MWWFRS Risk Assessment RA(SOP)2.10. AIDE MEMOIRE AIDE MEMOIRE 1 SCOPE 1.1 Working in, on or adjacent to water 1.2 Statutory Duties 2 SPECIFIC HAZARDS & RISKS 2.1 Types of water hazard 2.2 Effects of cold water 2.3 Hazard areas 2.4 Weirs/stoppers 2.5 Flooding 2.6 Water borne diseases 2.7 Blue green algae 2.8 Pollution 2.9 Working conditions 2.10 Debris 2.11 Poor Rescue Attempts

3 KEY CONTROL MEASURES 3.1 Pre-planning 3.2 Rescue assets of MAWWFRS 3.3 PDA 3.4 PPE 3.5 Additional equipment carried on appliances 3.6 Methods of rescue 3.7 Specialist rescue teams 3.8 Training 3.9 Specialist Advisors 3.10 Analytical Risk Assessment 3.11 Closing Highways and Leaving the Scene

4 OPERATIONAL PROCEDURES 4.1 Mobilisation to incidents in, on or near water 4.2 Low risk water hazards 4.3 Hazard zones 4.4 PPE 4.5 Specialist Water Rescue Team - Options 4.6 Rescues from moving water 4.7 Sub-surface rescue/recovery 4.8 Body recovery 4.9 Assistance from the RNLI 4.10 Requesting helicopter assistance 4.11 Personal hygiene & Welfare 4.12 Communications 4.13 Search operations

5 ADDITIONAL CONSIDERATIONS

1. 1 Characteristics of flowing water

6 REFERENCES 6.1 Sources of Information 6.2 Related documents 6.3 Questions & Answers

• SCOPE

1. 1

Working in, on or adjacent to water There is an increasing public expectation that Fire and Rescue Services (FRSs) will attend water-related incidents to save life or render humanitarian services. Mid and West Wales Fire and Rescue Service (MWWFRS) responds to such incidents in accordance with Section 11 of the Fire & Rescue Services Act 2004. MWWFRS often attends incidents that require personnel to work in, on or adjacent to water. The conditions at such incidents can vary from still ponds & canals to fast flowing water, weirs and areas of mud and slurry. Climatic extremes can add further to the complexity of these incidents and significantly change the approach required (e.g. extreme cold resulting in water courses forming ice and presenting a new set of hazards to crews). Further potential variables include geographic factors such as remote locations, steep banking and restricted access. This SOP should be read in conjunction with SOP 2.1 – Rescues from Ice & Unstable Ground, and SOP 2.8 – Flooding.

1. 2

Statutory Duties Historically the three Welsh Fire and Rescue Authorities (FRAs) respond to flooding and water rescue incidents using their existing general powers to respond to ‘other eventualities’ that cause or are likely to cause death, injury or illness, as described in Section 11 of the Fire and Rescue Services Act 2004 (‘2004 Act’). There was, however, no statutory requirement for them to do so. The Fire and Rescue Services (Emergencies) (Wales) (Amendment) Order 2017 (“the Order”) introduced a duty requiring the three Welsh FRAs to make provision to respond to flooding and inland water emergencies.

Rescue and protection in case of flooding and inland water emergencies (1) A fire and rescue authority must make provision in its area, to the extent that it considers it reasonable to do so, for the purpose of – a. rescuing people, or protecting them from serious harm in the event of an emergency involving flooding; and b. rescuing people in the event of an emergency involving inland water. (2) The duty within paragraph (1) applies to an emergency falling within section 58(a) of the Fire and Rescue Services Act 2004. (3) In this article – “flooding” means any case where land not normally covered by water becomes covered by water; “inland water” means – a. rivers, streams and canals; and b. lakes, reservoirs and water-filled quarries.”

[back to top]

• KEY CONTROL MEASURES

1. 1 Pre-planning (Ops 45)

Station Managers must make an effective assessment and operational plan for all areas of open water in their station ground. Where applicable recording of the plan may be made using SOP 1.4 Operational Intelligence Gathering. Consideration must be made in the following areas • The geospatial reference (12 figure reference); • The name of the water hazard, including alternative/local names; • A description of the type of water hazard (e.g. static – quarry); • An approximation of size/area/width/depth; • The pre-determined attendance (PDA); • Other services/agencies that may deploy (e.g. Mountain Rescue, requested via police, RNLI); • Access points/RVPs for appliances (four wheel and non four wheel drive access); • Access points for personnel; • Access points for deploying throw lines, inflatable hose, inflatable walkways; • Access/launch points for rescue boats;

• Access points for damming/pumping (where appropriate). For rivers and/or larger water hazards, the geospatial reference may relate to access points for appliances or launch points for FRS boats.

1. 2 Rescue assets of MAWWFRS

The following team types have been established in the MAWWFRS region to deal with large scale flooding incidents. These teams have the capabilities as outlined in the DEFRA Flood Rescue National Enhancement Project – Flood Rescue Concept of Operations and may also be utilised for water rescue incidents: Type E teams – Bank based All front line appliances can provide this response. The teams can provide the following functions at incidents: • Pumping operations. • Bank based rescue (shout, reach and throw). • Shallow water wading (less than height of fire boot) in static or slow moving water. Crews are required to wear eye protection when working on embankments near undergrowth and vegetation. Type D teams – wading These stations have the capability to perform wading duties (low energy water and less than waste height) in order to perform rescues, provide humanitarian services and assist with pumping operations during major flooding incidents. These teams are located in the following stations: • Newtown • Llanfair Caereinion • Talgarth • Cardigan • Lampeter • Fishguard • Pembroke Dock • Pontarddulais • Morriston • Neath • Glynneath • Pontardawe • Llanelli • Llandysul • Llandovery • Llandrindod Wells

The High Volume Pump crew based in Ammanford will also have the capability, however due to their role as HVP operators they will not be declared as a national asset Non powered boats will be strategically based at certain wading stations and may be deployed for evacuation purpose as part of wading rescue operations. Note – Wading teams are not on the PDA for water rescue incidents, however an OiC may call upon them to perform limited wading duties at an incident involving low energy water (shallow, still or very slow moving). Type C– Swift water rescue technician Swift water rescue technician (SRT) teams will have the additional capabilities to allow personnel to perform contact rescues by means of tethered swimming. They will also have the equipment to perform technical rescues using tethered boats. The teams are currently based in the following stations: • Carmarthen. • Hay on Wye. • Aberystwyth • Swansea Central. • Milford Haven • Welshpool All type C stations will have powered safety boats (inflatable) that are suitable for deployment during water rescue incidents. It is the intention of the Service to train all type C stations up to type B standard. Type B is Rescue Boat Operators. Type B teams will be qualified to operate rescue crafts in swift water environments, and using rescue crafts as a platform for swift water rescues, and tethered boat operations. A minimum of three trained boat handlers will be required to crew a rescue craft when committed for water rescue operations.

1. 4 Personal protective equipment (PPE)

Depending upon the Level of Response, PPE for use when working in, on or adjacent to water includes: • Lifejackets/PFD. • Dry suits. • Specific safety headwear. • Powered boat operators/swift water rescue technician PPE. • Crews wearing fire helmets MUST undo chin strap and slacken head band within helmet

• Eye Protection (Visor or Glasses)

• REFERENCES

1. 1

Sources of Information • Fire & Rescue Services Act 2004 • Generic Risk Assessment 2.1 – Ice & Unstable Ground • Generic Risk Assessment 2.8 – Flooding •

1. 2

Related Documents • SOP 2.1 – Rescues from Ice & Unstable Ground • SOP 2.8 – Flooding

•

ORIGINAL AUTHOR STATION MANAGER GRIFFITHS 613 ORIGINAL ISSUE SEPTEMBER 2010 REVIEWED BY Operational risk management (858) (612) LAST REVIEWED 29 March 2018 AUTHORISED BY GROUP MANAGER (248) CURRENT VERSION VERSION 4 [back to top]

Annex C Extracts from Appliance, Equipment, Technical Instruction for: Fire Service Boats 5.17

Appliance, Equipment, Technical Instruction for Fire Service Boats Mid & West Wales Fire & Rescue Service AET - 5.17 Appliance, Equipment, Technical Instruction for: Fire Service Boats The Service has a duty to make a suitable and sufficient assessment of the health and safety risks to employees and to provide them with comprehensive and relevant information on the risks identified. The document examines the hazards, risks and control measures relating to fire service personnel using the Fire Service Boats. 1 SCOPE 1.1 Introduction 2 SPECIFIC HAZARDS & RISKS 2.1 General 2.2 PPE 2.3 Getting to Work 3 PROCEDURAL GUIDANCE 3.1 General Description 3.2 Rigid Inflatable Boat 3.3 Avon ERB 400 3.4 Standard Boat Inventory 3.5 Work Boats 3.6 Mobilising & Crewing 3.7 Standard Testing and Maintenance 3.8 Flushing of Engine (after use) 4 Aide Memoir

1. 1 Safety
2. 2 General Use & Information
3. 3 Inventory
4. 4 Standard Testing & Maintenance

5 SOURCES OF INFORMATION/ APPENDICIES 5.1 Points of Reference

• SCOPE

1. 1

Introduction Within Mid & West Wales Fire & Rescue Service’s geographical area, there are several main arterial waterways, as well as costal areas. MAWWFRS has three types of boat available for incident response on these waterways • Powered Inflatables (SRT boats) • Powered Rigid Inflatable Boat (RIB) • Unpowered workboats This Appliance/Equipment Technical Instruction outlines the technical information, significant hazards and risks associated with operating Fire Service Boats, and lays down the procedural guidance to be followed when using them. [back to top]

• SPECIFIC HAZARDS & RISKS

1. 1

GENERAL SRT Boats SRT boats are crewed by personnel trained and competent to RYA (Royal Yachting Association) Level 2 as a minimum. If operating in swift moving water then the crew must be trained and competent to Rescue Boat Operator (RBO) level as a minimum RIB The only RIB within the service is based at Milford Haven and the crews will hold the following additional qualifications RYA advanced powerboat certificate (for all helms) Radar course GMDSS (Global Maritime Distress Safety System). Workboats Unpowered workboats are available to enable the movement of casualties and members of the public across areas of standing or slow moving water by wading teams They may also be used by SRT teams in line with their training and safe working practices

1. 2

P.P.E. All boat operators / handlers will wear Service supplied PPE SRT boats SRT boat operators will wear their SRT PPE when crewing the boat. RIB All crew will wear specialist helmets and the Crewsaver

Crewfit Life Jacket Twin275 (see AET 6.12). Work Boats All personnel handling workboats will wear wading PPE as a minimum. Any crew members working within 3 metres of the water’s edge must wear appropriate PPE (see SOP 2.8 and 2.10 for details)

1. 3

GETTING TO WORK All powered boats must have a minimum crew of three; all qualified to RYA level 2 standard as a minimum All boat operators are to work within the safe systems of work as laid down in SOP 2.8 Flooding and 2.10 Working on or near water.

[back to top]

• PROCEDURAL GUIDANCE

1. 1 GENERAL DESCRIPTION

The Boats used in the service are for effecting rescues from inland and specific costal waterways, they are also used for providing safety for crews working on or near water, therefore providing a safe system of work. There are three types of boat currently in use within MAWWFRS:- • Delta 650X(based in Milford Haven) • Avon ERB400 (Based with each SRT team) • Avon M9 workboat (Non mobile within each command) Training Boats and Operational Spares based at Earlswood Training Centre:- AVON W 465 Heavy Duty Workboat x3 • 5 chambers and keel all inflatable • Length 4.65 metre • Width 2.12 metre • Payload 10 persons or 1000kg • Aluminium deck Powered by outboards 30 HP Mariner 2 Stroke Engine

1. 2 RIGID INFLATABLE BOATS (RIBs)

A Rigid Inflatable Boat (RIB) is a light-weight but high performance and high capacity boat, constructed with a solid, shaped hull and flexible cylindrical shaped, air inflated tubes to assist in its buoyancy. • The design is stable and seaworthy, and the inflatable collar means that the buoyancy is not lost if a large quantity of water is shipped aboard. • RIBs are commonly between 4 to 7 metres (13 to 24 ft) long, and are propelled by one or more outboard motors. Generally the power of the motors is in the range of 5 to 500 horse power (hp). Their shallow draft allows for high manoeuvrability, speed and relative immunity to damage in low speed collisions. 3.2.1 Specification The Rib is kept inflated and strapped onto its transportation trailer, with the outboard engines fitted and associated equipment stored inside the boat. RIB at Milford Haven

Boat Type

Delta Rib

Length

1. 5 mtr

Engine type -

2 x Mariner 2 stroke (90Hp each) Maximum capacity

Min crew 3 persons (not max capacity)

Serial number

GB-DPS16473E101

Trailer type - Hallmark super rollercoaster 8 Additional information Night vision capability Offshore category C DSC/VHF/GPS Radar Thermal Imaging Salvage & pumping (firefighting) capability

1. 3 Avon ERB 400 (SRT Boat)

The Avon Professional Emergency Rescue Boats can be used for variable depth residential flood work, river flooding including fast flowing water operations, dam and sluice rescues where accurate shore line positioning is necessary, the craft has also been proven as an effective tool in ice and quick sand rescues. An integral aluminium slatted deck is covered with tough Avon fabric to ensure 'trip free' search and rescue missions. Load securing points are fitted to the deck to allow fixing of specialised equipment in the craft. The transom can be fitted with an outboard engine. The large bearing surface of the inflatable tubes and floor spread the rescue load over a wide base, making the craft ideal for rescue operations over thin ice or soft mud.

The ERB has been designed to be capable of being stored deflated and inflated using a compressed air cylinder when the incident scene has been reached. However the Avon ERB 400’s used by MAWWFRS boats are kept inflated and are transported to the incident on a trailer. If the pressure of the buoyancy tube drops crews can use a BA cylinder to bring the tube back up to pressure. 3.3.1 SPECIFICATION Capacity • Maximum number of persons : 8 • Maximum load (persons, motor & gear) : 900 kg • Floorboard usable area : 2.26m2 • Buoyancy tube volume : 1.334m3 Dimensions • Overall length : 4.0m • Inside length: 2.8m • Overall width : 1.9m • Inside width : 0.88m • Weight empty : 91 kg • Buoyancy tube diameter : 51 cm • Folded dimensions : 137 x 72 x 45 cm Safety / Inflation • Number of airtight compartments in main buoyancy tube : 4 • Inflation valve : 4 plus keel • Inflatable keel : 1 • High pressure air (Cylinder) inflation points : 2 • Overpressure valves : 4 plus keel • Total airtight chambers: 4 plus keel

Engine Power • Engine: shaft length : (20”) long shaft • Useful HP range : 10-20 HP • Maximum HP: 25HP (unguarded propeller outboard) or 30 HP (guarded propeller) • Maximum weight of engine : 75 kg The ERB has been equipped with a 25hp Mariner marathon engine which has been fitted with a propeller guard. SPECIFICATIONS MARATHON 25 Horsepower 30 hp Dry weight 48 kg Revolutions per minute (RPM) 5000-5850 Displacement 429 cc Cylinder layout In-line Number of cylinders 2 Starting Manual Alternator system power 76 watt Trim positions 6 Induction system Loop charged (1carb) Gear shift Forward (F) Neutral (N) Gear ratio 1.92:1 Steering Tiller: M, ML Recommended boat transom type Short 435 mm - Fuel tank type Standard Fuel type Gasoline Closed cooling N/A Engine technology 2 strokes The ERB 400 Boats are located at: • Aberystwyth • Welshpool

• Swansea Central • Carmarthen • Milford Haven • Hay on Wye

1. 4 STANDARD BOAT INVENTORY (RIB and SRT boats)

Boat including paddles Outboard engine Galvanised Trailer Lifejackets x 4 Waterproof lamp & charger x 1 Waterproof bag for radio x 1 First Aid kit x 1 Engine tool kit & propeller x 1 Anchor, chain & line (30 mtr) x 1 6m Painter (Bow line) x 1 Throw line in bag x 2 Body bag x 1 Plastic survival bag x 1 ERB Inventories may vary slightly from station to station, therefore this inventory is to be taken as a standard requirement.

1. 5 M9 Workboats

The work boats have been supplied for crews to be able to transport people across stretches of slow, low energy water. The craft is designed to be exceptionally stable even when tested in white water conditions. The craft is lightweight and can be stowed in a bag. Its stowed weight is 20kg which can be carried by one man in line with the service manual handling standing order. The bag has several handles which can accommodate a multi person lift should it be deemed necessary in line with a dynamic risk assessment. The M9 workboats will be stored deflated and can be inflated at the incident using the supplies regulator and a BA cylinder. The Cylinder is connected to the regulator and the cylinder opened, the pressure relief valve will operate to indicate that the required pressure has been reached. When inflating the regulator and cylinder will become cold and may cause frosting. Personnel inflating the work boat must wear appropriate PPE (Gloves and eye protection) 3.5.1 Specification Length Beam Tube Weight Overall Inside Overall Inside 2.74m 2.01m 1.48m 0.70m 0.38m 20kg Stowed Displacement Capacity Dry Swamped Weight Person 89 x 46cm 1100kg 640kg 400kg 4

1. 5.2 Transportation

Both the RIBs and the SRT boats are transported via a trailer and a dedicated vehicle assigned to the specific boat station. The workboats are non-mobile items of equipment and will be taken to an incident by command staff on request. There are several options for mobilising the workboat using both fire appliances and support vehicles. Below are some suggestions on how the work boats may be transported to an incident. Remove GTS

Stow in tool Locker

Stow in rear cab In response car (Skoda Yeti shown)

The suggestions above are not a definitive guide on how the work boats may be mobilised and commands may make other arrangements for their transport to an incident.

1. 6

MOBILISING & CREWING Milford Haven RIB The boat will be mobilised with a minimum crew of three operators qualified to the levels detailed in section 2.0 ERB 400 These boats will be mobilised as part of an SRT response as detailed in SOP 2.10 Working on or near water Workboats These boats are part of a command’s non mobile equipment and will be available on request through control.

1. 7

STANDARD TESTING & MAINTENANCE All the equipment within the standard inventory is to be inspected and maintained by Fire Service personnel in accordance with manufacturers instructions and this Appliance, Equipment &Technical document & Fire Service Manual Vol’1: Test Frequency On Acceptance After Use After Repair Weekly Quarterly 6 Monthly Annually On OIC’s Discretion On completion, Standard Test records must be completed accordingly. Due to the large numbers of various types of equipment within the inventory, standard testing should be staggered to the following frequencies: Weekly • Mainscheme Radio on the designated vehicle • Hand-held radios (portable mainscheme radio if applicable) • Water-proof lighting • Keel and sponson (ensure that each section is inflated) • Life jackets (visual) Monthly • Engine (engine tool kit & propeller)

If the engine has not been used for training or at an incident within the past 30 days then the engine should be subjected to an after use flush test as listed in section 3.8 of this AET Quarterly • Torches • Extinguishers • Trailer unit • Life jackets • All Lines • Anchor, chain & line • 1st Aid Kit Other items of equipment which may be present over & above minimum requirement: • Ratchet strap • Boat Sling • Strops & ‘D’ Shackles • Winch 6 Monthly (By insurance engineer) • Boat Sling • Strops & ‘D’ Shackles

Annually • Lifejackets 3.7.1 When carrying out a visual standard test on this equipment, things to look out for are: Boat & Trailer • The boats main structure is in good condition and that the keel & sponson are adequately inflated. • The trailer unit is in good condition, all fixings are secure, all straps & ratchets operate correctly and are in good condition and that the wheels are road-worthy. • It is also important to ensure that the engine is thoroughly flushed with clean water after use. Ancillary equipment • Ensure that the equipment is clean and dry whilst in storage. • Check that the body or structure of the item is not damaged, and that any nuts, bolts and/or fixtures are tight. • Check that metallic items of equipment are free from rust, if necessary use cloth/wire brush to remove rust and smear with oil or petroleum jelly. • Ensure that any moving parts are in good order and that the device works efficiently. The item should be lightly oiled or greased if necessary. • If the condition of a particular item of equipment is unsatisfactory, then an order for a replacement is to be completed using the ops 50 procedure where necessary. On completion, Standard Test records must be completed accordingly.

1. 8 FLUSHING OF ENGINE WITH CLEAN WATER (after use)

After the boat has been used, it will be necessary for the engine to be flushed with clean water, in order to remove any silt, grit or other fine debris which may have been induced into the engine during operation. Therefore on return to station, the procedure for flushing the engine (outboard motor) is:

• Locate a drum behind the boat and fill with clean water, the advised container is to be a 45 gallon drum with the top removed. • Crew is then to lower the engine into the container, taking the utmost care when carrying out this procedure. • When the engine is lowered and the propeller is fully submerged, the Crew are to ensure that container and engine are both secure, only then can the engine be turned on. THE ENGINE IS TO REMAIN IN NEUTRAL AT ALL TIMES. • The engine is then left to idle for approx 5 minutes, enabling enough time to flush the engine thoroughly, at the same time ensuring that a fine jet of water in continually expelled from the water coolant drain. • Once flushed, turn off engine, and crew is to lift the engine clear of the drum. FULL PPE INCLUDING GLOVES AND SAFETY GLASSES ARE TO BE WORN AT ALL TIMES

[back to top]

• Aide Memoir

1. 1

SAFETY All boats within Mid & West Wales Fire & Rescue Service are only to be operated by trained personnel (Minimum RYA Level 2). FULL PPE IS TO BE WORN i.e: All operators are to wear Dry-suits, Helmets etc

1. 2

GENERAL USE & INFORMATION The Boats used in the service are for effecting rescues and providing safety for crews working on or near water, at inland and specific costal waterways. There are three types of boat currently in use within MAWWFRS:- • Delta 650X(based in Milford Haven) • Avon ERB400 (Based with each SRT team) • Avon M9 workboat (Non mobile within each command)

The Delta 650X RIB located at: • Milford Haven The Inflatable Boats are located at: • Aberystwyth • Milford Haven • Carmarthen • Llanelli • Welshpool • Hay on Wye • Swansea Central The Avon W465 training boat/operational spare located at: • Earlswood Training Centre

1. 3

INVENTORIES Inventory found within this document is the minimum requirement for an operational boat, however may vary slightly from station to station.

1. 4

STANDARD TESTING & MAINTENANCE Test Frequency On Acceptance After Use After Repair Weekly Quarterly 6 Monthly Annually On OIC’s Discretion On completion, Standard Test records must be completed accordingly.

[back to top]

• SOURCES OF INFORMATION (& APPENDICIES if necessary)

<> 5.1 POINTS OF REFERENCE Manufacturers information DEFRA guidance SOP 2.8 Flooding SOP 2.10 Working on or near water

[back to top]

ORIGINAL AUTHOR EAT 636 ORIGINAL ISSUE October 2007 REVIEWED BY

LAST REVIEWED

July 2015 AUTHORISED BY Head of Tech Svc's GM KW 124 CURRENT VERSION Version 2 REVIEW PERIOD

Annex D Health and Safety Procedural Guidance Document 5.19 - Risk Assessment

1. 19-Risk-Assessment-July-2018July 2018

1 MID & WEST WALES FIRE & RESCUE SERVICE HEALTH & SAFETY PROCEDURAL GUIDANCE DOCUMENT

1. 19 – Risk Assessment

US – Indicates Updated Section (US) - (for initial circulation / publishing)

Contents

1. 0

Risk Assessments 2.0 Review of Procedure Appendix A – Risk Assessment Guidance Appendix B – Training Risk Assessment Appendix C – How to complete a risk assessment

Document Information

Security Classification N/A Document Owner Corporate Risk Department Document Author Head of Corporate Risk Date of Issue July 2018 Version Number 01 Date of Last Review July 2018 Date Last Amended July 2018 Date of Next Review July 2021 Approval Details Director of Operational Support and Improvement Equality Impact Assessment Reference N/A

1. 19-Risk-Assessment-July-2018July 2018

2

1. 0

Risk Assessment

1. 1

Risk assessments can be carried out by any member of staff within a Department or Station.

1. 2

Advice and guidance on undertaking risk assessments can be found within this procedure. Additional information on lone workers can be found within Procedure CR 5.21 Lone Working.

1. 3

This guidance is for use by all staff to carry out risk assessments across the broad spectrum of Service activities within their own sphere of competence.

1. 4

A list of those members of staff who have successfully passed either the NEBOSH or IOSH exam can be obtained from the People and Development Department. The holders of these qualifications are able to assist any member of staff with risk assessments. When approached, a qualified member of staff should respond positively and should bear in mind that it is their knowledge of the risk assessment procedures that is required and not an in-depth knowledge of the task or job to be assessed. This will be provided by the individual requesting assistance. Therefore, any risk assessment being undertaken should involve both the worker and where requested the NEBOSH or IOSH holder.

1. 5

To enable uniformity of standards to be monitored and to assess the effectiveness of the risk assessment programme, copies of all risk assessments are to be forwarded to the Corporate Risk Dept, Service Headquarters, for uploading to the Service @work site. A copy should also be retained within the Station /Department. Heads of Area Response /Department Heads will have the responsibility of implementing any remedial action that may be required. (This will normally be as a result of a stage two assessment).

1. 6

Department Heads will indicate on the form what action has been taken to resolve any problems before passing the assessments onto the Corporate Risk Dept.

1. 0

Review of Procedure

This procedure shall be reviewed following:

i. receipt of new information; ii when introducing change to or new working procedures; iii introduction of new agreements; iv in any event after 36 months; v at any other time when the current procedure is deemed to be invalid.

1. 19-Risk-Assessment-July-2018July 2018

3

Appendix A

Risk Assessment Guidance

This guidance is intended to help personnel to carry out Risk Assessments by following the procedure detailed below.

An Assessment is nothing more than a careful examination of what in your work, can harm people/property. This assessment will help to decide whether sufficient precautions have been taken; or if more needs to be done to prevent harm.

The overall aim is to make sure that no one is injured, becomes ill and that damage does not occur to equipment or property.

The important things that need to be decided are whether the hazards that have been identified are significant and whether they have been covered by adequate control measures so that the resulting risk is small.

e.g. Exposure to some chemicals presents a serious risk to health. However, by wearing Chemical Protection Suits, and Positive Pressure Breathing Apparatus, this Risk becomes negligible.

Risk Assessment Methodology

The process of Risk Assessment is carried out in two stages.

Stage 1

Identifies all significant Hazards that are present in the Workplace and takes account of:-

• who will be harmed; • the chance of that harm occurring (likely occurrence); • how much harm will occur (likely severity).

Hazards which present Risks that cannot easily be controlled require further attention, and the Stage II Assessment should be undertaken.

Stage 2

Identifies:

• the Hazards that require more attention; • the extra controls that are required to reduce the Risk to an acceptable level; • who will provide the extra controls; • when the controls are to be implemented.

The Risk Assessment should be completed by Personnel who are competent in the subject that is being assessed by their knowledge, experience or training.

1. 19-Risk-Assessment-July-2018July 2018

4

Stage 1 Form (Guidance For Completion)

Does any task expose personnel or other persons to Hazardous Substances or Hazardous Environments that may affect their health?

N.B. The “Hazard Type – Guidance to Stage 1 Assessment” and “Manual Handling Assessment Stage 1” can be used in association with the guidance below.

Activity/Workplace

In order to identify the Hazards present, all the activities or Workplaces must be listed. If a procedure is being assessed, all the events of that procedure should be listed in sequence.

Hazard

The next stage is to identify the Hazards associated with each of the “Areas of Assessment” as applied to each Event, Activity or Workplace.

The Areas of Assessment are:-

Actions what tasks are being undertaken that present hazards;

Environment what elements of the working environment act on the tasks being undertaken that may cause hazards e.g. heat, humidity, cold, slippery surfaces, poor lighting, rough terrain etc.;

Equipment does any equipment being used present hazards to the operations or other persons? e.g. light weight, pump hydraulic rescue equipment, disc cutters;

Occupational Health does any task expose personnel or other persons to hazardous substances or hazardous environments that may affect their health?

e.g. • biological agents;

• noise;

• smoke;

• radiation;

• fumes;

• lubricants;

• sewers;

• flammable liquids etc.

Litigation Are there any acts or omissions possibly performed by personnel that would be deemed as negligent towards other personnel, other persons, equipment, property or the environment.

1. 19-Risk-Assessment-July-2018July 2018

5

Persons at Risk Once all the significant hazards have been established, the persons that will be at risk from the hazards must be identified e.g. operational personnel, contractors, members of the public etc. This should also include the numbers of persons involved.

Existing Controls Any existing control measure relating to each Hazard should be identified and recorded. These will be taken into account when determining the Risk Rating.

Risk Rating The final part of the Stage 1 Assessment process will be to calculate the Risk Rating for each identified Hazard. This is done by estimating the likelihood of harm occurring and the Likely Severity should an accident happen.

Any existing control measure relating to each Hazard should be identified and recorded. These will be taken into account when determining the Risk Rating.

As mentioned earlier the existing control measures should be taken into consideration at this point, and the result of the calculation applied to the Risk Rating Interpretation.

Any risk that requires further action should be subject to a Stage 2 Assessment.

On Completion of the STAGE 1 Assessment Form

• one copy is to be forwarded to the Head of Area Response / Head of Department; • one copy is to be forwarded to the Corporate Risk Department; • one copy is to be kept in the Station / Department.

Stage 2 Form (HSRA 8) Guidance for Completion

Review of the assessment should be carried out on a regular basis.

Activity/Hazard The Hazard in question should be identified.

Controls Required Further control measures that are required to reduce the Risk to an acceptable level should be recorded. If the Risk Rating is very high and further control measures cannot reduce it, then it would be unacceptable to continue, and the activity should be halted.

Action by and Target Date Personnel should be identified who are going to put the controls in place and a date set for when they are to be introduced.

1. 19-Risk-Assessment-July-2018July 2018

6

Monitor and Review of Assessment It is the results of this stage ii assessment that will be recorded, monitored and reviewed periodically

The Monitoring process should ensure that the Safety Controls put in place stay in place and remain effective. This should be reflected in the ‘Controls Required’ column on the Stage II Assessment form (HSRA 8).

The review, and if necessary a modification of the Assessment should be carried out on a regular basis.

Unless:

• the nature of the work changes;

• the appreciation of Hazards and Risk increases;

• there is reason to suspect that the Assessment is no longer valid;

• accidents / near misses occur which were previously unforeseeable.

On completion of the STAGE 2 Assessment Form

• One copy is to be forwarded to the Head of Area Response / Head of Department; • One copy is to be forwarded to the Corporate Risk Department; • One copy is to be kept in the folder or electronically at the location.

1. 19-Risk-Assessment-July-2018July 2018

7

Appendix B

Training Risk Assessment

Training for Realism

Risk that is introduced into training must be a planned exposure with a definite purpose in mind.

• The standard drill, as set out in the Fire Service Training manual is the cornerstone of all practical training.

• The procedures and safety precautions written into the drills ensure that the safety of personnel taking part is paramount. This results in little exposure to risk whilst performing those drills.

• As skills develop, personnel should progress to more advanced types of training exercises.

• Ideally, the first time firefighters are exposed to a risk should not be at an operational incident where they would rely solely on someone else’s experience to deal with that risk.

• Personnel must, therefore, be exposed to risk whilst training and this is recognised by the Health and Safety Executive, for further guidance please see the following link.

• “If, in order to discharge his Section 2 duty, the employer has to expose the firefighter to an element of risk on the Training Ground, then the Health and Safety at Work Act does not prevent it, provided that all is done by the way of supervision and the provision of safeguards to ensure that overall, the arrangements are a safe as can sensibly be”.

Also;

“The exposure of firefighters to the risk of the training ground must be a planned exposure with a definite purpose in mind and one for which the way ahead has been prepared in previous training”.

• In order to follow this concept, drills and exercises must be formulated that introduce elements of realism. These can be conducted on station drill grounds and also at off station sites/premises.

• A drill yard at a Fire Station or Training Centre is normally a safe area until Fire Service activities are performed on it. Hazards are introduced into these training areas which have the potential for serious harm and only the controls of the risks that have been created by the hazards prevent injuries and damage from occurring.

1. 19-Risk-Assessment-July-2018July 2018

8

On Station

As more Risk has to be introduced into training, then the control of that risk must be increased:-

• As more risk has to be introduced within training scenarios, as suggested by the Health and Safety Executive, then the controls of these risks must be increased.

• In order to assist Officers-in-Charge of drills/exercises to identify hazards and, therefore, determine the extent that crews are being exposed to risks, a guide has been produced that will aid the assessment of those risks and once identified controls of the risks can be formulated.

• The guide provided is for general purpose risk assessment use but can easily be adapted for any situation, including training scenarios. The assessment is carried out in two stages. Refer to the Risk Assessment Procedure Appendix ‘A’.

Planning and Organising Arrangements of the Safety Controls

• The controls that have been identified must be planned and organised in such a way that they remain in place throughout the training period and that also includes the making up of equipment following the completion of the exercise.

Monitoring of the Safety Controls

• The Officer in Charge of the exercise must observe the safety controls and note if they were effective during the training period. This includes actions taken against recommendations of a safety brief. Nominated safety officers may also carry out this function.

Review of the Safety Controls

Any observations relating to the success or not of the safety controls must be raised as part of the debrief. Any significant observations that have any bearing on Service or Fire Service activities must be reported via the Operational Learning System (OLS).

The ORM Department will sift through any reported learnings for actioning by the Head of ORM Department. An email will then be circulated to the relevant Department to take appropriate action.

In the event that a learning outcome is highlighted relating to risk critical activities, the Head of ORM will forward an email to the relevant Department.

All actions undertaken are then fedback into the OLS system for audit purposes.

Observations would include:

• any injuries, Near Miss Events occurring;

1. 19-Risk-Assessment-July-2018July 2018

9 • observations regarding safety controls; • any comments that will improve safety in future training periods.

Off Station Exercise Sites

Risk Assessment is carried out in two stages

• Opportunities to hold Training Exercises at off station locations occur from time to time and full advantage should be made of them.

• The procedure for assessments will be similar as for station drills, but more reliance will be placed on the Hazard Type Guidance to Stage I because the site/premises being assessed will be relatively unfamiliar.

Safety Representatives

• Persons who have been nominated as Safety Representatives under the Safety Representatives and Safety Committee Regulations 1977 have legal rights to inspect the workplace and this includes exercise sites.

• Every effort should be made to liaise with the relevant health, safety and welfare representative in order than if they so request, they can inspect the site and details of the exercise can be viewed by them.

Risk Assessment Proforma

The following risk assessment forms will be available for Officers in Charge of exercises to carry out their assessments.

• Manual Handling Assessment – Stage I (HSMHRA Initial) Blank Manual Handling Risk Assessment Forms (HSMHRA 5) • Blank Stage I Risk Assessment Forms (HSRA 7) • Blank Stage II Risk Assessment Forms (HSRA 8)

1. 19-Risk-Assessment-July-2018July 2018

10

Appendix C

How to complete a Risk Assessment Identify the activity e.g. Removing 45mm delivery hose from locker. Carrying box of photo copying paper, hose training YFF Identify all the hazards associated with the risk Slip trip fall, Manual Handling, Crush impact amputation injury, fall form height. Identify who may be at risk e.g. Service personnel, other emergency service personnel, members of the public, environment, animals.

Identify existing control measures e.g. Training, instruction, supervision, SOP’s, AET’s Mechanical aids (trolley, Pull out shelves, ladder gantry, Standard test PPE. Using the Ranking of Hazard by Risk Guidance Taking all control measures into consideration decide upon the likely occurrence of an event happening Identify Risk Enter Likely Occurrence and Likely Severity values below using the equation below:-

LO x LS = RR Risk Rating Interpretation Using the guidance chart consider your actions

Review date. Identify all the hazards associated with the risk Slip trip fall, Manual Handling, Crush impact amputation injury, fall form height. Identify Risk Enter Likely Occurrence and Likely Severity values below using the equation below:-

LO x LS = RR Risk Rating Interpretation Using the guidance chart consider your actions

Review date.

1. 19-Risk-Assessment-July-2018July 2018

11

Appendix D

Annex E Milford Haven Port Authority - Leisure Guide

CAUTION: These areas have a high incidence of close quarters situations between craft of less than 20 metres in length, and vessels which can only safely navigate within the fairway. Recreational craft should navigate with extreme caution in these areas and should monitor VHF channel 12 at all times when within the Haven. Users are advised that Port Vessels and Police Marine Unit vessels patrol near Terminals to maintain BL27. NOTE: Tugwash may be experienced at the entrance and in the west channel whilst escorting tankers. 5 BLASTS (DANGER) You must make clear your intentions and indicate that you are getting out of the way 2 BLASTS 1 BLAST Long blasts every two minutes ship operating in fog 3 BLASTS 2019 Leisure User Guide Please see over for more information about the Waterway. Dead Slow Minimum Wake These areas are designated for navigational safety and importance for wildlife. Bays, creeks and tributaries provide vital feeding and breeding areas. At low water most of these areas may dry out or become very shallow revealing submerged natural features. The restriction on speed and wake reduces disturbance and helps protect wildlife from swamping or being panicked by fast moving vessels, and the environment from damaging effects of wake and vessel contacts. Operating at slow speed in these areas also reduces the possibility of serious injury or damage and also helps to reduce conflicts between users and ensures safe navigation for all. Jet Ski/PWC Aquabatics Area Personal Watercraft (PWC) aquabatics are ONLY permitted in the Pwllchrochan Flats PWC area. PWCs are permitted elsewhere in the Haven if they navigate considerately with due care and attention and in compliance with Harbour Bye-laws and zoning. Water-skiing Areas In the interest of safety, vessel owners should be mindful of their wake in these areas and transit with caution. Potential Renewable Test Bed Sites Sensitive Habitat Zone - Voluntary No Anchoring National Park Castle Hotel or Public House Yacht Club Launch Site Fuel Public Pontoon (Seasonal) Privately owned/operated Pontoon (Seasonal) Marker Buoys Moorings (Commercial) Day Visitor Moorings These moorings are for the use of vessels up to a max length of 40ft and max weight of 10 tonnes during daylight only. The skipper must ensure sufficient depth of water and swinging area is available for the vessel to be secured. Full conditions of use can be found at www.mhpa.co.uk/marine-leisure-services/ Defibrillator (accessible from water) Defibrillators can be found at: Hazelbeach: public toilet wall at the top of the pontoon Burton: public toilet wall in the car park Neyland Marina: top of the gangway entrance Milford Marina: Pier Head Dale: the boathouse Lawrenny Arms: behind the bar Black Tar: top of slipway. All Pilot Vessels and the Water Ranger RHIB carry defibrillators (contact details overleaf.) Leading Lights in the Haven Haven Approach: West Blockhouse (WBH) Watwick Beacon (WB) 022(T) West Channel: Great Castle Head (GCH) Little Castle Head (LCH) 040(T) Main Channel: Popton and Bullwell beacons (PP) (BW) in line 095(T) 094(T) Milford Shelf: Beacons above Newton Noyes (NN) 080(T) 087(T) Dale

• Approximate Ferry Times

St. Ann’s inwards at 1145 hours Depart Ferry Terminal at 1445 hours St. Ann’s inwards at 2345 hours Depart Ferry Terminal at 0245 hours NOT TO BE USED FOR NAVIGATION Carew and Cresswell Rivers Navigable only with great care and local knowledge. Pembroke River Navigable only with great care and local knowledge St Ishmaels Sandy Haven Gelliswick Angle Milford Haven Llanstadwell Neyland Pembroke Dock Pembroke Burton Llangwm Hook Landshipping Lawrenny Cresswell Quay Carew Upton Little Milford Haverfordwest Castle Pill Neyland Marina Toll Bridge Rudder’s Boatyard Cresswell River Garron Pill Sprinkle Pill Eastern Cleddau Western Cleddau Beggar’s Reach Carew River Benton Castle Millin Pill Picton Point Picton Castle Slebech Blackpool Mill Pembroke Ferry Cosheston Pill East Llanion Marine Hobbs Point Wear Spit Beacon Ferry Port* Westfield Pill Nature Reserve Milford Marina Milford Haven Port Control Thorn Island Stack Rock RNLI Lifeboat Station Tanker Terminal Keep Clear Tanker Terminal Keep Clear Tanker Terminal Keep Clear East Channel West Channel St Ann’s Head Lighthouse Lighthouse (WBH) (WB) (GCH) (LCH) (PP) (BW) (NN) Pwllchrochan Flats (shallow at low tide) Dale Anchorage Area used by sail training centres for teaching beginners. Proceed with caution. Mid Channel Rocks Valero Refinery Puma Energy Ltd. Pembroke Power Station South Hook LNG Valero Pembrokeshire Oil Terminal & Dragon LNG www.mhpa.co.uk f PortofMilfordHaven t @mh_port Horn Signals Area used by sail training centres for teaching beginners. Proceed with caution. List of Admiralty Charts 1478, 2878, 3273, 3274, 3275

Leisure Charts SC5620 folio | Imray Y26 and 2600 ATLAS refer Pembroke Barrage To request barrage opening, ring County Hall 01437 764551 Operating times 07:00-19:00 weekdays 09:00-12:00 Saturdays Please request 72hrs in advance Please respect this environment: take your litter home with you and follow the Pembrokeshire Marine Code. Together we’ll protect the amazing plants and wildlife found along the Waterway. To find out how you can protect marine wildlife go to www.pembrokeshiremarinecode.org.uk or download the Pembrokeshire Marine Code App. •

It is recommended that you: Inform Tell others where you are going and when you will be back. Tides and Weather Check the tides and forecast to ensure conditions are and will remain suitable for your vessel and activity. SOS Device Carry a marine VHF radio with Ch16 and capable of listening to Ch12 (harbour working channel) as well as any other means of calling for help. Wear a LifejacketYou never know when you might need it and it could just save your life. Engine and Equipment checks Ensure you have appropriate working equipment and spares aboard and that all moorings or anchors being used are capable of withstanding extreme weather conditions. Training Ensure your vessel is handled in a competent manner and avoid taking unnecessary risks. RNLI safety information is available by visiting www.rnli.org.uk The Water Ranger The Water Ranger patrols the Waterway in a grey RHIB with Patrol on the sides from St Ann’s Head at the entrance to the Haven, upriver to Haverfordwest on the Western Cleddau, and Blackpool Mill on the Eastern Cleddau. The main aim of this service is to provide information, advice and assistance to all Waterway users, police the zoning system and enforce the appropriate Bye-laws. www.mhpa.co.uk/water-ranger If you wish to contact the Water Ranger please call 01646 696100. Further information for recreation users can be found by visiting www.mhpa.co.uk/marine-leisure-services Waterway patrol vessels Port pilot vessels operate throughout the Waterway 24/7 and can be recognised by their green hulls and white superstructure. All patrol vessels keep radio watch on channel 12. Sharing the Estuary Safely You are sharing the Haven with wildlife, commercial activity and a whole range of other leisure activities. To promote the safety of navigation and reduce conflict between users, the Waterway is zoned for high speed activity. The zones also ensure that the sensitive areas of the Haven remain tranquil and can be used for quiet enjoyment. The Milford Haven Port Authority Bye-laws (see page 14) control navigation and should be followed at all times. Conservation & Wildlife The Haven is a special place, protected for its tranquil, scenic landscape and varied wildlife. The upper and lower reaches form part of the Pembrokeshire Coast National Park because of their unspoilt character. For more information on exploring the National Park visit: www.pembrokeshirecoast.org.uk In spring and summer the peaceful bays and rivers form breeding grounds for many birds, whilst in winter up to 25,000 birds such as shelduck, teal, widgeon, curlew and redshank come here to feed. Many parts of the foreshore are Sites of Special Scientific Interest and the whole Haven forms part of the Pembrokeshire Marine Special Area of Conservation because of its international importance for marine wildlife above and below the water. For more information visit www.pembrokeshiremarinesac.org.uk Human activities in the Haven can damage and disturb wildlife and its habitats. Please think and act in an environmentally friendly way by following the zoning, Bye-laws and Pembrokeshire Marine Code. Following the Marine Code will help you to enjoy the best of the coast without disturbing sensitive wildlife. www.pembrokeshiremarinecode.org.uk Recreation Activities A whole range of water-based recreation activities take place on the Waterway, including sailing, power boating, water-skiing, personal watercraft use (jet-skiing), windsurfing, canoeing, fishing, diving and bathing. Activities take place both on an informal basis and through organised events. Navigation Notes Commercial Vessel Speed Large commercial vessels may, at times, make passage at higher speeds than expected to maintain proper control and steerage. Users should monitor the approach of vessels and remain clear when required. Approach for Milford Docks Leading lights 348° fixed blue in line with fixed blue on white background attached to wall below the Lord Nelson Hotel. Cleddau Bridge There may be confused water by bridge piers approximately 1 hour before and after High Water springs. Lifeboats A Tamar class lifeboat and an inflatable ‘D’ class inshore lifeboat are based at Angle. People in danger and needing rescue services should contact Milford Coastguard - 999 - ask for Coastguard, or call on VHF channel 16 using accepted emergency procedures. Public Pontoons (Mar-Oct) A maximum stay of 12hrs is permitted with no return within 24hrs. The ‘T-Head’ must remain clear at all times for embarking /disembarking. Fishing Restrictions Welsh Government legislation and the Bass Order 1990 restrict fishing between 30th April and 1st November in areas of the Waterway. Fishing methods, minimum sizes and catch quantities are also controlled throughout. For information call 0300 025 3500. Fishing for salmon, trout, freshwater species and eels within the Waterway requires a valid UK rod licence. Local Weather Information Weather forecasts are broadcast by Milford Haven Coastguard on channels 16 and 84. Times: Inshore weather 0750, repeated three hourly, and 1950, repeated three hourly. Strong wind and gale warnings are broadcast as soon as they have been received and on subsequent three hour forecasts. ∏ ✔ ® Tidal Information Differences at High Water Approximate times to be applied to Milford Haven High Water predictions Dover +5 hrs Tenby -14 mins Fishguard +1 hr 3 mins Swansea +5 mins Neyland +2 mins Ilfracombe -30 mins Haverfordwest +10 mins Padstow -50 mins Tidal Streams Across the entrance to Milford Haven the streams run approximately at right angles to the line of approach; well within the entrance they run nearly parallel to the channel. The tidal streams run approximately as follows: Position Interval from High Direction Water Milford Haven 1 mile outside entrance +0455 East -0125 West Maximum spring rate in each direction is 2.25 knots Inside the entrance -0555 Ingoing +0030 Outgoing Maximum spring ingoing rate is 1.5 knots and outgoing rate is 1.75 knots It should be noted that heavy swell may be experienced until inside St Ann’s Head in adverse conditions as far as Angle Light-bouy. Tidal streams within the Haven run in the direction of the deepwater channel as far as Wear Point (51°42’N, 04°59’W). In the reach above the Wear Point the ingoing stream is deflected to the north side of the channel by Carr Spit (51°42’N, 04°57’.8W). In Pembroke Reach the ingoing stream runs principally in the channel north of Dockyard Bank, attaining a spring rate of 2.0 knots; it is weak and irregular in the channel south of this bank, in which it attains a spring rate of 0.5 knot. The outgoing stream runs strongly in the channel south of the Dockyard Bank, attaining a spring rate of 2 knots. It is weak on the south side of the channel, north of the bank; in the north part of the latter channel an eddy runs strongly during the outgoing stream. Safety FOG - recommended procedure for small craft in visibility of 0.5 mile or less

1. All craft with VHF radio maintain a listening watch on channel 12
2. Provided you are aware of your position and you intend staying

out of the fairway, monitor the commercial activity reported on channel 12

1. If you are uncertain of your position, it may be prudent to

anchor clear of the shipping channel until visibility improves Sound signals used by vessels in the area (Bye-laws No 9) Signal Meaning Four short blasts in A power-driven vessel underway rapid succession is about to turn around (altering course more than 135°) Followed after a short interval by One short blast Vessel turning with head to starboard or Two short blasts Vessel turning with head to port Three short blasts Vessel operating to stern One long blast, two Power-driven vessel or vessel not short blasts under command unable to give way to a sailing vessel Five short blasts A vessel is doubtful if you are taking sufficient avoiding action Æ f 7 ! 1 ∫ Been involved in an incident on the Waterway? Report to Milford Haven Port Control on 01646 696136/7 or complete an incident report form, found on the Port's website: goo.gl/yqhWwz Further Information Key Contacts VHF Channel STD number Milford Haven Port Control 12 01646 696137 Water Ranger 12 & 16 01646 696100 Milford Pierhead 14 01646 696310 Castlemartin Firing Range 16 01646 662367 Emergency Numbers HM Coastguard 999 or 01646 690909 Police (inc. Marine Unit) 999 or 0845 330 2000 Customs 0800 595000 Withybush Hospital, Haverfordwest 01437 764545 Natural Resources Wales (Incident Hotline) 0300 065 3000 Berths/Boatyards/Chandlery Dale Sailing 01646 603110 East Llanion Marine 01646 686866 Lawrenny Quay 01646 651212 Milford Marina and Boatyard 01646 696312/3 Neyland Marina 01646 601601 Quadra Marine 01646 278270 Rudders Boatyard 01646 600288 Windjammer Marine 01646 699070 Yacht Clubs Dale 01646 636362 Neyland 01646 600267 Pembrokeshire 01646 692799 Pembroke Haven 01646 684403 Other Tourist Information Centre, Haverfordwest 01437 775244 Pembrokeshire Coast National Park 01646 624800 Pembrokeshire County Council Beaches Dept. 01646 602105 Crime Stoppers 0800 555111

Marine Accident Report

FD06063 ｜ 2020_英国_grounding-and-recovery-of-container-feeder-vessel-thea-ii-and-tug-svitzer-josephine__2020.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2020 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

ACCIDENT REPORT ACCIDENT REPORT Report on the investigation of the grounding and recovery of the container feeder vessel Thea II and the tug Svitzer Josephine in the approaches to the Humber Estuary on 15 December 2018 SERIOUS MARINE CASUALTY REPORT NO 15/2020 AUGUST 2020

The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be in­ admissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2020 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch First Floor, Spring Place 105 Commercial Road Southampton United Kingdom SO15 1GH Email: maib@dft.gov.uk Telephone: +44 (0) 23 8039 5500 Fax: +44 (0) 23 8023 2459 Press enquiries during office hours: 01932 440015 Press enquiries out of hours: 020 7944 4292

CONTENTS GLOSSARY OF ABBREVIATIONS AND ACRONYMS SYNOPSIS 1 SECTION 1 - FACTUAL INFORMATION 2 1.1 Particulars of Thea II and Svitzer Josephine 2 1.2 Narrative 3 1.2.1 The grounding 3 1.2.2 Actions post-grounding 6 1.2.3 Recovery of Svitzer Josephine and Thea II 8 1.3 Damage sustained 11 1.3.1 Thea II 11 1.3.2 Tug Svitzer Josephine 11 1.3.3 Tug Yeoman 11 1.4 Environmental conditions 12 1.5 Thea II 13 1.5.1 Crew 13 1.5.2 Vessel background 13 1.5.3 Thea II’s managers 13 1.5.4 Anchor arrangement and guidance 14 1.5.5 Failure of main engine 15 1.6 Svitzer Josephine 15 1.6.1 Crew 15 1.6.2 Tug background 15 1.7 Yeoman 17 1.7.1 Crew 17 1.7.2 Vessel background 17 1.8 Humber Port and Vessel Traffic Service 18 1.8.1 Overview 18 1.8.2 Risk assessment 18 1.8.3 Accident statistics 18 1.8.4 Environmentally sensitive sites 19 1.8.5 Humber Estuary Serious Marine Emergency Plan 19 1.8.6 Vessel traffic services 19 1.8.7 Immingham Oil Terminal duty fire tug 22 1.9 Guidance on anchoring 22 1.9.1 Length of anchor cable deployed 22 1.9.2 Use of a second anchor 23 1.10 The role of the Secretary of State’s Representative 23 1.11 The role of the duty counter-pollution and salvage officer 25 1.12 Towage and salvage contracts 25 1.13 Similar accident 26 SECTION 2 - ANALYSIS 27 2.1 Aim 27 2.2 Overview 27 2.3 Thea II – grounding 27 2.3.1 Main engine failure 27 2.3.2 Thea II dragging anchor 27 2.4 Svitzer Josephine – grounding 28 2.4.1 Master’s appreciation of risk 28 2.4.2 Vessel Traffic Services Humber’s advice 29

1. 5 Salvage of Thea II

30 2.5.1 Emergency response by ABP Humber 30 2.5.2 Emergency response by Thea II’s managers 31 2.5.3 Emergency response by the SOSREP and the DCPSO 32 2.6 The tow to the anchorage 32 SECTION 3 - CONCLUSIONS 34 3.1 Safety issues directly contributing to the accident that have been addressed or resulted in recommendations 34 3.2 Other safety issues not directly contributing to the accident 35 SECTION 4 - ACTIONS TAKEN 36 SECTION 5 - RECOMMENDATIONS 37 FIGURES Figure 1: Thea II, post-grounding, alongside in Hull Figure 2: Svitzer Josephine Figure 3: Overview of the accident track from 1501 on 15 December until 0115 on 16 December Figure 4: Thea II and Svitzer Josephine viewed from the search and rescue helicopter, aground on the Binks Shoal Figure 5: Yeoman (Hong Kong class) tug Figure 6: Thea II’s passage, under tow, from the Binks shoal to the deep water anchorage Figure 7: Yeoman’s propellers fouled by Svitzer Josephine’s towline Figure 8: Thea II’s D’Hone, high holding power anchor Figure 9: Svitzer Josephine’s underwater arrangements Figure 10: Marine protected areas at Spurn Head Figure 11: VTS display showing up to date S57 chart data Figure 12: Scope of deployed anchor Figure 13: Use of the second anchor

GLOSSARY OF ABBREVIATIONS AND ACRONYMS ABP

Associated British Ports

AHM

Assistant Harbourmaster

AIS

Automatic Identification System ALB

All-weather lifeboat

ASD

Azimuth stern drive BA

British Admiralty

BIMCO

Baltic and International Maritime Council

BRM

Bridge Resource Management C/E

Chief engineer

DCPSO

Duty counter-pollution and salvage officer DPA

Designated Person Ashore

ECP

Emergency check-off plans

ECS

Electronic chart system

GPS

Global positioning system GTGP

• Guide to Good Practice

HESMEP

Humber Estuary Serious Marine Emergency Plan IACS

International Association of Classification Societies IALA

International Association of Lighthouse Authorities

INS

Information Service IOT

Immingham oil terminal kg

kilogram kt

knot kW

kilowatt LOF

Lloyd’s Open Form m

metre

MCA

Maritime and Coastguard Agency

MGN

Marine Guidance Note mm

millimetre

MRC

Marine Response Centre

NAS

Navigation Advisory Service NCP

National Contingency Plan

nm

nautical mile

OOW

Officer of the watch

PMSC

Port Marine Safety Code RNLI

Royal National Lifeboat Institution

SAR

Search and rescue

SMCP

Standard maritime communications phrases

SMS

Safety management system

SOSREP

Secretary of State’s Representative STCW

• International Convention on Standards of Training, Certification and

Watchkeeping for Seafarers 1978, as amended (STCW Convention) t

tonne

TOS

Traffic Organisation Service VHF

Very high frequency radio

VSP

Voith Schneider propeller

VTS

Vessel Traffic Service

VTSO

Vessel Traffic Service operator TIMES: all times used in this report are UTC unless otherwise stated.

1 SYNOPSIS At approximately 1500 on 15 December 2018, the main engine of the Cyprus registered container feeder vessel, Thea II, failed while approaching the Humber pilot station. With no propulsion the master anchored the vessel in gale force weather conditions. While awaiting the arrival of tugs, the weather deteriorated further and Thea II dragged its anchor towards the Binks shoal. At 1739, before Thea II could be towed clear, the container vessel and one of the tugs attending, Svitzer Josephine, ran aground. Search and rescue assets were deployed, but no personnel were evacuated from either vessel. As the tide rose, both vessels refloated. Just before 2200, Svitzer Josephine was able to get underway and return to harbour. However, Thea II’s tow was delayed while the vessel’s managers attempted to agree contractual terms with another tug operator. At about 0100 the following day, its managers signed Lloyd’s Open Form. The vessel was then towed to the deep water anchorage before being taken into harbour on 16 December 2018. There were no injuries and no pollution as a result of this accident. The investigation found that Thea II had lost propulsion due to the failure of its main engine governor, which could not be repaired, and the master had no choice but to anchor the vessel and await the arrival of tugs. However, as the weather deteriorated and the vessel began to drag towards the Binks shoal, the master did not deploy all his remaining chain or his vessel’s second anchor. Had he done so, it would have slowed or arrested the vessel’s rate of drift and allowed more time for tugs to secure lines and tow the vessel clear. The investigation also found that the tug Svitzer Josephine grounded because the master, focused on passing the towline to the crew of Thea II, lost positional awareness and did not appreciate the rate of drift of both vessels towards the Binks shoal. Thea II refloated on the rising tide. However, despite the best efforts of the harbourmaster, the Maritime and Coastguard Agency and local tugs, in difficult weather conditions, Thea II’s managers only agreed salvage terms 2½ hours later, under the threat of formal direction from the Secretary of State’s Representative. This delay almost led to Thea II re-grounding, risking damage to an environmentally sensitive area. As a result of this accident, action has been taken by Associated British Ports Humber and Svitzer Marine Ltd, to review their risk assessments and operating procedures. The Secretary of State’s Representative has continued an industry wide engagement programme to promote a greater understanding of his role. A recommendation has been made to Thea II’s managers, TS-Shipping, to review their emergency response organisation and procedures.

2 SECTION 1 - FACTUAL INFORMATION 1.1 PARTICULARS OF THEA II AND SVITZER JOSEPHINE SHIP PARTICULARS Vessel’s name Thea II Svitzer Josephine Flag Cyprus United Kingdom Classification society DNV GL Lloyd’s Register IMO number/fishing numbers 9107394 8919219 Type Container feeder vessel Tug Registered owner MS Thea II Schiffahrts GmbH & Co. KG Svitzer Marine Ltd Manager(s) TS-Shipping GmbH & Co. KG. Svitzer Marine Ltd Construction Steel Steel Year of build 1995 1991 Length overall 94.09m 28.13m Gross tonnage 2899 364 Minimum safe manning 6 3 Authorised cargo Containers None VOYAGE PARTICULARS Port of departure Amsterdam, Netherlands Immingham, United Kingdom Port of arrival Hull, United Kingdom Immingham, United Kingdom Type of voyage International Coastal Cargo information 85 containers Not applicable Manning 11 3 MARINE CASUALTY INFORMATION Date and time 15 December 2018, at 1739 Type of marine casualty or incident Serious Marine Casualty Location of incident Approaches to the Humber Estuary Injuries/fatalities None None Damage/environmental impact Minor damage to ship’s bottom, steering gear and deck vents Water damage to engine room ventilation system and accommodation spaces Ship operation At anchor Underway Voyage segment Arrival Mid-water External & internal environment The wind was south-easterly force 9-10, accompanied by a 4-5m sea Persons on board 11 3

3 1.2 NARRATIVE 1.2.1 The grounding On the evening of 14 December 2018, Thea II departed Amsterdam, Netherlands, bound for Hull, United Kingdom. The vessel was loaded with 85 containers, 46 of which were carried on deck (Figure 1).The container feeder vessel Thea II alongside after the accident Figure 1: Thea II, post-grounding, alongside in Hull Thea II’s passage across the North Sea was uneventful. However, by the following afternoon on 15 December, as Thea II approached the Humber, the weather deteriorated and the wind increased to south-easterly force 8, accompanied by a 2-3m sea. At 1501 an operator at vessel traffic service (VTS) Humber observed that Thea II’s speed had reduced to zero and enquired using Very High Frequency (VHF) radio if there was a problem. Thea II’s master responded that the main engine had failed, the ship was now 'not under command’ and that it required tug assistance. The VTS operator (VTSO) advised the master to anchor in his present position while tugs were sought. Three minutes later, Thea II’s master reported that the vessel had anchored using its port anchor with 5 shackles of cable in the water. The VTSO subsequently contacted Humber coastguard and notified them of the incident. The VTSO then tried to raise the crew of Svitzer Josephine (Figure 2), Immingham Oil Terminal’s (IOT) duty fire tug, on VHF radio, to enquire if they were able to assist Thea II. The call went unheard as the tug’s bridge was not manned. The VTSO then contacted the IOT duty officer and requested that he activate the duty fire tug alarm. The alarm was heard on board the tug, and Svitzer Josephine’s master immediately contacted VTS Humber. The VTSO informed him that the container feeder vessel Thea II had broken down, was anchored on the north side of the channel and needed urgent tug assistance. Svitzer Josephine’s master agreed to assist. He then asked the VTSO to notify the Svitzer office of their tasking so that they could arrange a replacement fire tug.

4 Figure 2: Svitzer Josephine The tug Svitzer Josephine Image courtesy of balticshipping.com Shortly after Thea II had anchored, the port suspended pilotage due to the deteriorating weather conditions. The VTSO observed Thea II moving very slowly to the north and, concerned that it might be dragging its anchor, called the vessel on VHF radio: “At the moment you are moving at about 0.5kts to the north, caution shallow water. You may need to consider dropping more cable. Do not drop another anchor, but you may need to drop more cable.” Thea II’s master confirmed that he understood the instructions and the vessel’s drift appeared to stop. At 1507, having briefed his crew, Svitzer Josephine’s master got the tug underway and commenced the 14nm passage downriver towards Thea II. During the passage the master quickly realised that a second tug would be required to safely tow the container vessel. He contacted the Svitzer office, which tasked Svitzer Castle to assist. At 1546, shortly after sunset1, Svitzer Castle’s master reported to VTS Humber that it was on passage to join Svitzer Josephine to tow Thea II into port. A harbour pilot was also placed on stand-by to board the disabled vessel if weather conditions improved. At 1630, Thea II’s master first emailed the vessel’s managers informing them that Thea II’s engine had stopped and the vessel was anchored by the side of the channel. At 1633, just before low water Spurn Head2, the VTSO observed Thea II drifting north-west at 1.7kts (knots) (Figure 3). He called the vessel and enquired if Thea II was dragging its anchor. The master confirmed that this was the case and reported that he was deploying a further shackle of anchor cable. However, this did not stop the vessel’s drift. 1 Sunset was at 1540. 2 Low water Spurn Head was at 1637.

5 Figure 3: Overview of the accident track from 1501 on 15 December until 0115 on 16 December Chart showing an overview of the accident Reproduced from Admiralty Chart 0104-0 by permission of HMSO and the UK Hydrographic Office 1501 - Thea II loses propulsion 2337 - Harbourmaster issues special direction 1739/44 - Svitzer Josephine and Thea II ground 0115 - Irishman and Yeoman tow Thea II to deep water anchorage 0054 - Thea II agrees LOF with SMS Towage 1505 - Thea II anchors 1629 - Thea II drags anchor 1707 - Svitzer Josephine leaves navigational channel 2207 - Svitzer Josephine refloats and returns to Immingham under its own power 1726 - Svitzer Josephine commences passing tow to Thea II 2330 - Irishman and Yeoman on scene, ready to pass lines Time Wind (dirn/ force Tidal force (dirn/kts)

6 Shortly afterwards, Svitzer Josephine rounded Spurn Head, followed 20 minutes later by Svitzer Castle. Svitzer Josephine’s master noted that the weather had deteriorated, and the wind was now south-east force 9-10, accompanied by a 4-5m sea. At 1707, as the two tugs left the main channel, Svitzer Josephine’s master and the VTSO discussed the situation. The tug’s master confirmed its draught as 5.4m and expressed his concern that, given the weather conditions, he was unsure if his tug would be able to secure a tow to Thea II and move it into deeper water. The VTSO acknowledged the tug master’s concern and emphasised that Thea II was in need of urgent assistance. He finished the exchange by stating: “Yeah that’s fine and having worked at Spurn for some time I’ll keep an eye on you and if I think you’re getting too close to the bank I’ll let you know. I don’t have a chart but from low water experience [Thea II] should be OK at the moment.” At the time, Thea II, also with a draught of 5.4m, was 7 cables from grounding on the Binks shoal. Svitzer Josephine’s master then contacted Thea II’s crew. He established that the container vessel’s port anchor was deployed and directed Svitzer Castle to stand off while he made his approach to pass the towline. It took 15 minutes for Thea II’s crew to make the towline fast. At 1739, as Svitzer Josephine’s master tried to tow Thea II, he found his tug was aground. He reported this to VTS Humber, who in turn contacted Thea II’s master, who responded that his vessel was also aground. 1.2.2 Actions post-grounding At 1751, concerned that there might be a need to evacuate personnel from both vessels, Humber coastguard requested the launch of the Spurn Head Royal National Lifeboat Institution (RNLI) all-weather lifeboat (ALB). Shortly after the lifeboat was underway, Svitzer Josephine’s engine room fire alarm sounded, and the master ordered both engines to be stopped. With no propulsion, the tug pounding the seabed and waves breaking over the vessel, the master became concerned for the crew’s safety and ordered them to don immersion suits and lifejackets. He then contacted Humber coastguard to request evacuation. The coastguard acknowledged his call for assistance and activated a search and rescue (SAR) helicopter to attend. At 1758, the Humber harbourmaster activated the Humber Estuary Serious Marine Emergency Plan (HESMEP) and established the Marine Response Centre (MRC) at the port’s Grimsby office. At 1824, the Maritime and Coastguard Agency (MCA) and the Secretary of State’s Representative (SOSREP) were informed of the incident. At 1900 the SAR helicopter arrived on scene but, having assessed the situation, the pilot declared the weather conditions out of limits for conducting a winch transfer (Figure 4). The ALB offered to evacuate non-essential personnel from Thea II and Svitzer Josephine, but as the watertight integrity of both vessels was still intact their masters declined.

7 Figure 4: Thea II and Svitzer Josephine viewed from the search and rescue helicopter, aground on the Binks ShoalVideo still from footage taken by rescue helicopter showing the vessels aground Image courtesy of Maritime and Coastguard Agency The harbourmaster arrived at the Grimsby port offices shortly after 1900. Conscious of the opportunity provided by the rising tide and forecast moderation in the weather, he contacted another of the port’s tug operators, SMS Towage Ltd, to establish if they had vessels that could assist. As a result of these discussions it was agreed that SMS’s shallow draught (3.5m) Hong Kong class tugs, Yeoman and Irishman, would attend once they had completed their tasking at Hull. At approximately 1920 the two tugs commenced their passage downriver towards Thea II. Thea II’s master had been in contact with the vessel’s managers by email since 1630, updating them on the tugs’ arrival. The chief engineer (C/E) was also in email correspondence regarding the engine defect. However, there was no mention of dragging anchor until 1948, when the master informed the managers that the tugs had been unsuccessful and that Thea II had dragged anchor and was now aground. The vessel’s Designated Person Ashore (DPA) responded by asking the master to provide the completed emergency check-off plans (ECP) for main engine failure/ grounding. In a further email, at 2015, the managers advised the master “do not ask the captain of the tugs for anything and do not sign anything”. The ECPs were completed by Thea II’s officer of the watch (OOW). Shortly after 2030 Svitzer Josephine’s master began to re-evaluate his situation. He sensed the tug’s pounding had become less intense and assessed that the tug was beginning to refloat. Aware that the tide was rising, and that the engine room fire alarm had stopped, he decided to send the chief engineer below to restart the tug’s main engines.

8 1.2.3 Recovery of Svitzer Josephine and Thea II At 2139 Thea II’s master received an email from the vessel’s managers informing him that insurance brokers were in the process of sourcing appropriate tugs to assist. At 2145, the tugs Yeoman and Irishman (Figure 5) left the main channel to approach Thea II. Shortly afterwards, Yeoman reported to VTS that the weather conditions were unsuitable to pass towlines. Figure 5: Yeoman (Hong Kong class) tug Yeoman - a Hong Kong class tug Courtesy of transportsofdelight By about 2200 Svitzer Josephine had refloated and was led into deep water by the ALB. As he manoeuvred his tug clear, the master payed out the vessel’s 130m towline until its weak link broke and the line detached from the winch. This left the tug’s towline attached to Thea II. Once clear, Svitzer Josephine returned to Immingham under its own power, escorted by Svitzer Valiant. Just before high water at 2235, VTS noted Thea II had started to drift towards the Spurn peninsula, so contacted the vessel to enquire if it was going to drop a second anchor. Thea II’s master responded, saying that he did not intend to do so as they were drifting into deep water on the other side of the Binks Shoal. At 2245, after discussion between the harbourmaster, tug skippers and the ALB coxswain, it was decided that the ALB would guide Yeoman and Irishman’s masters to the bow and stern of Thea II. The ALB led Irishman to the east of the Binks shoal to Thea II’s stern and then led Yeoman to the vessel’s bow. During Yeoman’s approach to Thea II the tug was set on to the container vessel’s anchor cable, causing some damage to its hull. By 2330, both SMS tugs were close enough to pass lines to the container vessel. Concurrently, the Humber harbourmaster updated the SOSREP on his intentions to secure the tugs to Thea II and tow it clear. At 2300, Thea II’s master emailed the vessel’s managers, stating his concern over their delay in authorising the tow. Thirty minutes later the vessel’s managers responded, informing him that they were ‘working on it’. The master continued to express his concern to them, approximately every 20 minutes, over the delay in granting him permission to take the tugs’ lines.

9 At 2337, concerned that the tide was falling and that Thea II’s crew had not taken lines from the tugs, the harbourmaster issued a special direction on VHF radio to Thea II’s master, directing him to take tugs’ lines. This direction was acknowledged by Thea II’s master, but he did not issue instructions to his crew to take lines from the tugs. Throughout the incident, Humber coastguard had kept the MCA Duty Counter- Pollution and Salvage Officer (DCPSO) updated regarding the situation. The DCPSO, in turn, kept the SOSREP apprised of the situation. The DCPSO also commissioned a model of the environmental risk posed by the grounded vessels and liaised with Thea II’s managers and insurers. The vessel’s managers and insurers were initially confident that they would be able to contract tugs to tow their vessel clear. However, at 2342, concerned that Thea II had yet to take lines from the attending tugs, Humber coastguard contacted the DCPSO, and the Humber harbourmaster was also in contact with the SOSREP. Several telephone conversations involving the DCPSO and Thea II’s managers followed. During these discussions it became clear that negotiations between SMS Towage, the tugs’ owners, and Thea II’s managers/insurers, had stalled. Thea II’s managers/insurers wanted to employ the tugs under a fixed price towage contract, while the tug owners, SMS Towage, given the extreme weather conditions and risk to their vessels, were only willing to salvage the container vessel under Lloyd’s Open Form (LOF). At 0007 on 16 December, VTS informed Thea II on VHF: “[Thea II] your position is not good. You are in a very bad position at this time, if you don’t accept the assistance from the tugs.”

10 Figure 6: Thea II’s passage, under tow, from the Binks shoal to the deep water anchorage Chart showing passage from Binks Shoal to deep water anchorage Reproduced from Admiralty Chart 0107 by permission of HMSO and the UK Hydrographic Office 0110 - Yeoman and Irishman commence tow of Thea II. Led into safe water by lifeboat 0215 - Yeoman port propeller fouled 0520 - Pullman arrives and takes over tow from Yeoman 0640 - Thea II unexpectedly deploys anchor 0911 - Thea II anchored in deep-water anchorage

11 arranged for a replacement tug, Pullman, to get underway. The harbourmaster advised the tugs to remain clear of the windfarm exclusion zone and hold their current position until the replacement tug arrived. At 0520, the replacement tug, Pullman, arrived on station and, once Yeoman had been released, took over as Thea II’s bow tug. The disabled Yeoman was left to drift safely to the north. At 0640, Thea II’s master sighted anchored vessels nearby and, thinking that his vessel had arrived at the anchorage, dropped anchor. After discussion with the tugs and VTS, Thea II quickly recovered its anchor and the tow to the deep water anchorage continued. At 0911, Thea II anchored in the deep water anchorage. Once released, Irishman proceeded north, recovered the disabled tug Yeoman and towed it back to port. Pullman remained on station until Thea II was safely towed into harbour the following day. 1.3 DAMAGE SUSTAINED 1.3.1 Thea II Once Thea II’s cargo had been discharged at Hull, the vessel was towed to Germany for repair. After the vessel had been dry docked it was found that: ● The rudder and steering gear were damaged, but operational. ● The vessel’s hull was intact, but the hull bottom paintwork was damaged. ● There was minor damage to the propeller blades. ● Two quarter deck ventilation pipes had been broken by the tug’s towline. Thea II was out of service for 6 weeks for repair. The vessel resumed its charter in February 2019. 1.3.2 Tug Svitzer Josephine Post-accident in-water survey of Svitzer Josephine revealed that the void space containing the starboard engine room vent fan had flooded. This had shorted the fan motor and probably triggered the fire alarm. The starboard weather deck door had also vibrated open during the grounding, causing water damage to the tug’s accommodation and recreational spaces. Initially it was thought that the tug’s hull and Voith Schneider propellers (VSP) were undamaged. However, during the dry-docking of the vessel in June 2019, extensive cracking was identified along the skeg and adjoining keel plating, which had not been detected during the in-water survey. 1.3.3 Tug Yeoman Yeoman’s hull was damaged when, during its approach to Thea II, it was pushed onto the container vessel’s anchor cable. Later during the tow, the tug lost the use of both engines. Yeoman was towing Thea II bow to bow, and its master initially

12 assumed that its propellers had been fouled by fishing gear. However, when Yeoman was in dry dock it was discovered that the port and starboard azimuth units had been fouled by a towline (Figure 7). The following damage was also discovered: ● The internal engine room mountings were out of tolerance. ● The starboard bilge keel and starboard side of the hull were damaged. The vessel was in dry dock for 10 weeks and returned to service in February 2019. Post-accident inspection of the rope recovered from Yeoman’s azimuth units suggests that it was an 80mm, 3-strand ‘Strongline’ polyester, Langhorst towing hawser, attached to a 15m yellow/orange towing pennant. This was similar to Svitzer Josephine’s towline, which had been discarded by Thea II’s crew. Figure 7: Yeoman’s propellers fouled by Svitzer Josephine’s towline Towline caught up in azimuth stern drives Image courtesy of SMS Towage Ltd 1.4 ENVIRONMENTAL CONDITIONS During the afternoon of 15 December, weather conditions deteriorated as forecast. At 1500, when Thea II first anchored, the wind was south-easterly force 8 accompanied by a 2-3m sea. By 1630, when Thea II began to drag its anchor, the wind had increased to storm force 10, with a 4-5m sea. However, by 2300, just after high water, conditions had begun to moderate; the wind had veered to south-south- east gale force 8 and the sea reduced to 2m. During the following morning, while Thea II was under tow, conditions improved further. The wind veered to the south and then south-west and reduced from force 6 at 0100 to force 4 at 0900. The tidal streams in the approaches to the Humber Estuary near the Binks shoal flowed south-south-west on the flood and north-north-east on the ebb tide, at rates of up to 2kts. Initially the tidal stream set north-north-east until 1800, south-south- west until midnight, before setting north-north-east again (Figure 3). Buoyant towing pennant Non-buoyant towline

13 Tides in the Humber were affected by weather conditions and, as a result, their height differed from those predicted. Normally, these variations were between 20 and 30cm. However, on 15 December, with storm force winds from the south-south- east, Spurn Head tide gauge showed the actual height of tide was between 0.9 and 1; 1m less than predicted. The seabed had a charted depth of between 0 and 10m, and the bottom composition was described as sand and shells; judged to be moderate holding ground. 1.5 THEA II 1.5.1 Crew Thea II had a crew of 11 made up of five Russian/Ukrainian officers and six Filipino crew. The 62 year old Russian master had worked at sea for over 40 years and held an STCW II/2 certificate of competency. He had completed four, 4-month contracts as master of Thea II. The master did not keep watches in harbour, but at sea kept the 0800-1200 and 2000-0000 watches. On the day of the accident, the master had been on watch as normal in the morning and had returned to the bridge at 1430 to oversee the vessel’s entry into the Humber. He then remained on the bridge throughout the incident, only returning to his cabin to rest once the vessel was towed clear of the Binks shoal. The 62 year old Ukrainian C/E had worked at sea for over 40 years and held an STCW III/2 certificate of competency. He had served as C/E of Thea II for over 13 years. The C/E, assisted by a Filipino oiler, worked during the day, adjusting his hours as required so that he could keep watch in the engine room during pilotage. 1.5.2 Vessel background Thea II had been managed by TS-Shipping GmbH & Co. KG. from build. Since March 2018 the vessel had been on time charter to Samskip, employed on a regular route between the Netherlands and the United Kingdom. Thea II’s primary means of navigation was by paper charts, supported by an electronic chart system (ECS) as a navigational aid. The vessel was also fitted with two navigational radars: Automatic Identification System (AIS) and two global positioning systems (GPS). As Thea II was less than 3000gt the vessel was not fitted with a voyage data recorder. 1.5.3 Thea II’s managers Thea II’s managers were based in Haren, Germany. At the time of the accident their office was closed for the weekend and the vessel’s DPA and safety auditor coordinated the company’s response remotely from their homes. During the accident, the DPA and safety auditor made attempts to contact Thea II’s master by mobile phone, but with no success. The master, instead, preferred all communications with the managers to be via email. After being initially informed by the master, the managers did not monitor Thea II’s position using AIS as they were under the misconception that the ship was safely anchored, awaiting tug assistance.

14 1.5.4 Anchor arrangement and guidance Thea II’s anchor arrangements comprised a forecastle windlass and two 1998kg, high holding power, D’Hone anchors (Figure 8). The port anchor was attached to 12 shackles and the starboard anchor to 9 shackles of 40mm hardened steel cable. Figure 8: Thea II’s D’Hone, high holding power anchor D'Hone high holding power anchor Image courtesy of TS-Shipping GmbH and Co. KG Thea II’s safety management system (SMS) guidance on anchoring included: ●That the OOW needs to consider the condition of the holding ground, weather and tidal conditions as well as traffic flow. ●To detect dragging, the vessel’s position should be checked frequently preferably by different methods (visual bearings, radar bearing and distance). ●If the weather deteriorates or there is a risk of dragging, inform the master and make engines ready for immediate manoeuvre. If necessary, consider lowering a second anchor. The vessel’s SMS did not provide guidance on the scope of anchor cable to be deployed. Fluke Shank

15 Thea II’s paper chart showed the vessel’s position had been fixed at 1500, 1638 and

1. It also showed the depth of water on anchoring, including the height of tide

(2.3m), was approximately 12m. 1.5.5 Failure of main engine Thea II’s main engine was a MAN B&W, 2950kW, diesel engine connected via a gearbox to a controllable pitch propeller and shaft generator. The shaft generator was used while entering and leaving harbour to operate the 250kW bow thruster. The vessel was also fitted with two generators, both of which were operating at the time of the accident. Shortly before 1500 on 15 December, the C/E had been working in the engine room with the oiler when he heard the engine speed governor alarm sound. The engine revolutions slowly reduced until the shaft stopped. The C/E immediately went to investigate. He checked the fuel system but could find nothing wrong. He then inspected the governor control unit. There, he found one of the fuses had tripped, which he reset. The C/E then attempted to restart the main engine, without success. He repeated this on three or four occasions. At each attempt, the engine briefly ran before shutting down. He then spoke to the OOW on the bridge, who informed him that the ship was going to anchor. The C/E continued to investigate the main engine defect, liaising by email with the vessel’s managers. By about 1600 he had identified the problem as a defective circuit board within the electronic governor control system. A spare circuit board was not carried on board. The C/E advised the master that specialist technical support from the manufacturer would be required to repair the defect. Until then, the vessel’s main engine would be out of action. 1.6 SVITZER JOSEPHINE 1.6.1 Crew Svitzer Josephine carried a crew of three comprising the master, an OOW (deck) and the C/E, all of whom had joined the vessel at 0830 on 15 December 2018. At sea, the master conned and navigated the tug. The OOW assisted the master when required, and during towing operations worked on deck handling the towline. The C/E oversaw the safe operation of the tug’s machinery and, when towing, operated the towing winch from the wheelhouse. The 54 year old master had worked at sea for over 20 years. He had joined Svitzer tugs as an able seaman before gaining his STCW II/3 tug master’s certificate of competency in 2016. He had been working as relief master on various tugs since March 2016. Svitzer Josephine’s minimum safe manning document permitted the tug to reduce from six to three crew when operating within Humber port limits. 1.6.2 Tug background The UK registered Svitzer Josephine was owned by Svitzer Marine Ltd. It was a twin VSP tractor tug and had a bollard pull of 53t. The VSPs were protected from damage by debris or grounding by means of protective steel structures known as ‘stools’ (Figure 9). The tug was designed to tow over its stern and was fitted with a

16 towing winch that held a 130m long, 80mm diameter towline. The towline was not buoyant, however the 15m rope pennant, secured to the eye of the towline, was. Svitzer Josephine, in common with the other Svitzer tugs involved, was based at Immingham and supported shipping movements throughout the Humber Estuary. The tug’s normal area of operations confined the vessel to operating within the Humber Estuary’s navigation channels. The SMS required the master to maintain a minimum under keel clearance of 1m. Svitzer Josephine’s primary means of navigation was using paper charts. These were kept up to date by the crew and stored on the bridge within a cardboard tube for use as required. The tug was also fitted with a radar, an ECS, an echo sounder, and a GPS. During the emergency response Svitzer Josephine’s master navigated the tug using its ECS. Svitzer Josephine’s echo sounder transducer was fitted in front of its VSPs. The echo sounder’s minimum depth alarm was set at 1m. However, the position of its transducer, like many tugs, meant that it was prone to false alarm due to VSP wash and, on the day of the accident, the echo sounder’s audible warning alarm had been silenced. Figure 9: Svitzer Josephine’s underwater arrangements Skeg Towing winch Protective steel plates Position of echo sounder transducer Voith Schneider propeller

17 The company’s SMS guidance for vessels fitted with an ECS stated:

●Correctly updated paper charts must be carried covering the entire operational area.

●Paper charts must be used as primary means of navigation. The charts must be prepared & readily available (not in a drawer).

●All electronic charts used by the unofficial electronic chart system, must be kept updated in accordance with the recommended update frequency from the manufacturer/supplier of the charts and publications. On the day of the accident, the tug’s paper charts did not include BA chart 104, ‘Approaches to the Humber Traffic Separation Scheme’. This chart encompassed the area where Thea II anchored and both vessels subsequently grounded. The SMS guidance on fixing the tug’s position stated that it:

●Must be done at regular intervals.

●Must be done utilizing either the radar, visual landmarks, navigational marks or a combination.

●Must be recorded in the logbook or the chart utilized.

●If applicable the echo sounder read out must be compared with the charted depths in the electronic chart system when position fixing.

●Svitzer do not accept position fixing solely by GPS.

●Svitzer do not accept past track as a means of position fixing. 1.7 YEOMAN 1.7.1 Crew Yeoman was operated by a crew of three: a master, deckhand and an engineer. While towing, the master conned the tug, the deckhand remained on the bridge and operated the winch, while the engineer worked on deck passing and returning the towline. Yeoman was certified under The Small Commercial Vessel and Pilot Boat Code of Practice (MGN 280). The 66 year old master held a commercially endorsed Royal Yachting Association, Yachtmaster Offshore (Power Driven Craft) certificate of competency and had completed a voluntary towing endorsement. 1.7.2 Vessel background The UK registered Yeoman was operated by SMS Towage. It was 24m in length, had a 40t bollard pull and was propelled by a twin azimuth stern drive (ASD). Known locally as a ‘Hong Kong’ class tug because of its country of build, Yeoman’s towing winch was fitted to its forecastle and it was designed to primarily tow over the bow. Irishman, which was also involved in the salvage operation, was a sister vessel to Yeoman.

18 1.8 HUMBER PORT AND VESSEL TRAFFIC SERVICE 1.8.1 Overview Associated British Ports (ABP) was the Statutory and Competent Harbour Authority for the Humber Estuary, responsible for ensuring the safety of marine operations within its area. The accident took place within the port’s limits. ABP’s authority for the Humber Estuary was underpinned by legislation, which permitted the Humber harbourmaster to issue general and special directions. 1.8.2 Risk assessment ABP Humber conformed to the UK Port Marine Safety Code (PMSC) and the associated Guide to Good Practice (GTGP). As part of the code the port was required to ensure that: All marine risks are assessed and are eliminated or reduced as low as reasonably practicable. Of the port’s 43 generic risk assessments, two were relevant to this accident: ‘Equipment failure (vessel)’; and ‘Grounding: Sea to Cleeness’.

● Equipment failure (vessel). The worst credible scenario was based on equipment failure leading to loss of control of a vessel resulting in serious injuries; moderate damage to the vessel/infrastructure, with no measurable environmental impact; and, minor local media attention. Overall, vessel equipment failure was assessed as a ‘low’ risk. The most significant control measures included: the requirement for ships to declare defects prior to arrival/departure, the use of bridge resource management (BRM), harbour directions, port state inspection and pilotage.

● Grounding: Cleeness to Sea. The worst credible scenario was based on a vessel aground that could not be refloated on the same tide, which resulted in serious damage to the vessel and environment and the potential for serious injuries and serious impact on port business. Overall, grounding was assessed as a ‘medium’ risk. The most significant control measures included: aids to navigation; AIS coverage; BRM; stakeholder communication; accurate hydrographic data; passage planning; pilotage; tug availability; and VTS. 1.8.3 Accident statistics In 2018 the port recorded 28 minor incidents during 33,000 vessel moves, where vessels had suffered main engine failure while operating within the Humber port area. This equated to approximately one incident every other week.

19 In the last 20 years the port recorded four groundings in the approaches to the Humber Estuary (‘Cleeness to sea’). In each case, the vessel had been aground for only a short period, before getting underway without tug assistance. 1.8.4 Environmentally sensitive sites The Humber port area contained several environmentally sensitive sites. Thea II and Svitzer Josephine grounded in a Marine Conservation Zone3 and Special Protection Area4, which were adjacent to a Special Area of Conservation and Ramsar site5 (Figure 10). The environmental modelling, initiated by the DCPSO, showed that marine diesel or lube oil released from Thea II or Svitzer Josephine could have been deposited on to the Spurn peninsula before being swept into the North Sea. 1.8.5 Humber Estuary Serious Marine Emergency Plan The purpose of HESMEP was to provide a framework for the management of serious marine emergencies within, and close to, the Humber port area. The plan gave the harbourmaster the lead for coordinating the emergency response to incidents located within the port boundary, while the coastguard led the emergency response for those outside it. The HESMEP was exercised annually. The last exercise involving the port, the coastguard and other emergency services had been conducted a few weeks before the accident. 1.8.6 Vessel traffic services VTS Humber was located at Grimsby. It was operated as a 24-hour service and was manned at all times by three personnel: a VTS supervisor (known as the assistant harbourmaster (AHM)), supported by two VTSOs. VTS’s role was to monitor and regulate shipping within the Humber port area. The AHM and operators used a bespoke computer system, which combined data from several radar sites, vessel AIS and environmental information to produce a comprehensive picture of maritime activity within the Humber and its approaches. Charts for the VTS area were available in both electronic and paper form. The electronic chart data could be displayed on the VTSO’s computer system in simple, intermediate and detailed views (Figure 11). In most cases operators configured their display to use the uncluttered, simple view, so that they could monitor vessel traffic without distraction. 3 A Marine Conservation Zone is a type of marine nature reserve in UK waters. They are established under the Marine and Coastal Access Act (2009) and are areas designated with the aim to protect nationally important, rare or threatened habitats and species. 4 A Special Protection Area is a designation under the European Union Directive on the Conservation of Wild Birds. Under the Directive, Member States of the European Union have a duty to safeguard the habitats of migratory birds and certain particularly threatened birds. 5 A Ramsar site is a wetland site designated to be of international importance under the Ramsar Convention. The Convention on Wetlands, known as the Ramsar Convention, is an intergovernmental environmental treaty established in 1971 by UNESCO, which came into force in 1975.

20 Figure 10: Marine protected areas at Spurn Head Chart showing the prot ected areas around Spurn Head Image courtesy of Natural England

21 Figure 11: VTS display showing up to date S57 chart data Simple view on VTS display Simple view Intermediate view on VTS display Intermediate view Detailed view on VTS display Detailed view Governed by the International Association of Lighthouse Authority (IALA) regulations, VTS Humber provided vessels with the following services:

● Information Service (INS). A service that provided relevant information to vessels within the VTS area. This included vessel movements/ intentions, weather or hydrographic information, and information concerning manoeuvring limitations of vessels.

● Traffic Organisation Service (TOS). A service that coordinated vessel traffic within the VTS area to prevent dangerous situations developing.

22 VTS Humber did not provide:

● Navigation Advisory Service (NAS). This service would have provided essential and timely navigation information. It would have included the provision of navigational advice or instruction. However, the IALA is clear that VTS provision of NAS did not relieve masters of their responsibility for the safety of their vessels. To ensure VTS information and direction were understood by foreign crew, for whom English might not have been their first language, the operators were taught to use IMO Standard Maritime Communication Phrases (SMCP). A key element of SMCP was the use of specific message markers, such as the following: ‘instruction’, ‘advice’, ‘warning’, ‘information’, ‘question’, ‘answer,’ ‘request’ and ‘intention’ to ensure the significance of the message that followed was understood. All VTS Humber operators on watch at the time of the accident had completed approved IALA VTS courses appropriate for their roles. 1.8.7 Immingham Oil Terminal duty fire tug ABP Humber had contracted Svitzer Marine Ltd to provide a duty fire tug ready and equipped to respond to emergencies at either Immingham or South Killingholme Oil Jetties. The duty tug was at 10 minutes’ notice to get underway and could be contacted by VHF or emergency alarm. This contract did not commit Svitzer Marine to provide emergency towage to incidents elsewhere within the port. 1.9 GUIDANCE ON ANCHORING 1.9.1 Length of anchor cable deployed In order for the maximum holding power of an anchor to be achieved, the anchor flukes must be buried in the seabed and the shank must be pulled horizontally. If the anchor’s shank is lifted by only 5° from the seabed, then the anchor’s holding power would be reduced by 25% (Figure 12). To ensure that this does not happen the crew must deploy sufficient cable. Industry guidelines on the length of anchor cable to be deployed state that:

● the length of cable in metres must be between 6 and 10 times the depth of water; or,

● the length of cable in shackles must be greater than: Using a charted depth of 12m these guidelines indicate that Thea II’s master should have deployed between 2.6 and 5.2 shackles of cable. When Thea II initially anchored, 5 shackles of cable were deployed, which was increased to 6 shackles once the vessel started to drag its anchor. These guidelines recommend that in bad weather additional anchor cable be deployed. Furthermore, in strong winds, high-sided vessels like Thea II, with containers stowed on deck, can be prone to yaw either side of their anchor cable. At the extremity of each yaw, the vessel would snatch at the cable, placing it under

23 Figure 12: Scope of deployed anchor considerable load and increasing the risk of the anchor breaking free and dragging. Thea II was yawing approximately 35° either side of its cable immediately prior to dragging its anchor. 1.9.2 Use of a second anchor A method to reduce the likelihood of dragging anchor in bad weather is to deploy a second anchor (Figure 13) 6, by either:

● Dropping a second anchor ‘underfoot’. Employing this method, the second anchor is deployed with sufficient cable to just touch the seabed. The drag provided by the second anchor ‘underfoot’ slows the vessel’s rate of yaw, reducing the likelihood of the other anchor breaking free and dragging. Or

● Deploy a second anchor with a similar length of cable to the first. However, this method is suitable only when the vessel is expected to be at anchor for a short period of time. If the vessel swings, due to wind or tide, its cables will become fouled and can be difficult to untangle when the anchors need to be recovered. For this reason, many mariners are reluctant to deploy a second anchor in this manner. 1.10 THE ROLE OF THE SECRETARY OF STATE’S REPRESENTATIVE The SOSREP’s role, as described in the UK National Contingency Plan (NCP), is to represent the Secretary of State for Transport by removing/reducing the risk to safety, the environment and property, arising from a maritime accident. In support of the NCP, the SOSREP regularly engages directly with UK port authorities, as well as harbour and towage industry associations. When possible, SOSREP is also involved in exercises undertaken by ports and the UK coastguard where pollution and salvage responses are tested. 6 Admiralty Manual of Navigation volume 1. Sand and shells An anchor provides max holding power when its flukes are fully embedded, and its shank lies on the sea bed. If the shank is lifted by 5º the holding power will be reduced by 25%.

24 Figure 13: Use of the second anchor 35 35 1 1 2 2 3 3 4 Dropping second anchor underfoot Dropping a second anchor 1st anchor (maximum length of cable) Preferred ‘let go’ position for 2nd anchor (underfoot) 1st anchor 2nd anchor ‘Let go’ position for 1st anchor Alternative ‘let go’ position for 2nd anchor (underfoot) ‘Figure-of-eight’ yawing pattern reduced by second anchor being dropped off under foot ‘Let go’ position for 2nd anchor

25 The UK NCP states that, when an accident takes place within a harbour area, the SOSREP will monitor the situation and tacitly approve the harbour authority’s response to actions. However, the SOSREP can intervene and use his powers in support of the harbourmaster if required. The SOSREP’s powers, defined in the Merchant Shipping Act 1995, include being able to issue the following:

● An initial caution. This is normally issued by the coastguard or the DCPSO where the actions of the master/owners have been unsuccessful, advising them that, under UK law, the SOSREP has the power to order the vessel to take a tug or accept tug assistance. The caution warns the master/owner that if they fail to make adequate arrangements for assistance, the SOSREP will formally direct action and recover the costs from them.

● Formal direction. The SOSREP can issue a formal direction ordering the vessel to accept tug assistance. If the vessel fails to do so, then not only will action be taken to recover costs from the owners, but also the owners/ master will have committed a criminal offence under UK law and be subject to prosecution. The SOSREP’s initial caution and formal direction would be issued by VHF radio, followed up with an email. There is no requirement to issue an initial caution prior to issuing a formal direction; similarly, an initial caution can be issued without the need to escalate it to a formal direction. In this case, while no initial caution had been issued, the DCPSO had informed the vessel’s managers and insurers that if Thea II did not take lines from the tugs by 0100 on 16 December, the SOSREP would issue a formal direction for the vessel to do so. 1.11 THE ROLE OF THE DUTY COUNTER-POLLUTION AND SALVAGE OFFICER The MCA’s counter-pollution and salvage branch has specific responsibility for counter-pollution preparedness and response at sea, as well as management of the UK’s stockpile of equipment and dispersant. Members of the team keep week-long duties as the DCPSO, coordinating the national response to marine incidents in UK waters that have the potential to result in pollution or salvage. As part of their role they have the authority to liaise with the SOSREP and other stakeholders as required. 1.12 TOWAGE AND SALVAGE CONTRACTS Thea II’s managers/insurers had two contractual options for negotiating the towage of their vessel to a place of safety: a contract tow or LOF. Negotiation of a contract tow is often based on Baltic and International Maritime Council (BIMCO) standard towage contracts. These allow the towage contract to be drawn up, using a proforma. The payment is based on a daily charter fee, paid for by the vessel’s managers/insurers on completion of the tow.

26 In contrast, LOF, the standard contract for marine salvage, does not specify the sum paid to salvors. Instead, based on the principle of ‘no cure no pay’, it requires salvors to use their ‘best endeavours’ to recover the ship to a place of safety in order to receive payment. LOF can be agreed over VHF and allows the salvage to start quickly without the salvors, insurers and owners having to negotiate contractual terms. Where salvage is successful, the renumeration awarded under LOF is intended to incentivise salvors to respond to stricken vessels. However, while the scale of the reward is likely to be considerably larger than that received for a contract tow, it is often paid many months after the event following arbitration. 1.13 SIMILAR ACCIDENT On 18 December 2018, the Russia registered bulk carrier Kuzma Minin grounded after dragging its anchor in Falmouth Bay, England. The vessel was in ballast and was successfully refloated on the next high water. There was minor damage to the vessel and there were no injuries or pollution. The MAIB investigation (Report 11/2019) identified that the vessel dragged anchor in wind speeds of over 50kts. The vessel’s movement was quickly detected by the crew and the master attempted to get the vessel underway. However, the anchor was fouled by a chain and, while the crew attempted to clear this, the vessel was blown ashore. Following the accident, the port strengthened existing measures to check visiting vessels have protection and indemnity insurance, and to improve the safety of vessels at anchor in Falmouth Bay. A recommendation was made to the vessel’s owner aimed at ensuring its vessels are appropriately resourced and meet the requirements of international conventions.

27 SECTION 2 - ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 OVERVIEW While approaching the Humber pilot station, the container feeder vessel Thea II suffered an engine failure that could not be immediately repaired and which meant that the vessel had to anchor in gale force conditions. The master had no choice but to ask for tug assistance to enter port. While at anchor the winds increased to storm force, leading to Thea II dragging its anchor as insufficient chain and no second anchor were deployed. Subsequently, while trying to establish a tow, both Svitzer Josephine and Thea II grounded on the Binks shoal. The tug grounded because its master, focused on securing a towline, lost positional awareness and did not fully appreciate the drift rate of the two vessels. The delays in Thea II securing a tow after refloating were due to the managers’ lack of awareness of the vessel’s situation. Despite numerous emailed requests by Thea II’s master, and a special direction issued by the Humber harbourmaster instructing him to secure tug towlines, it was only the threat of intervention by SOSREP that persuaded the managers to take action to prevent their vessel re-grounding in an environmentally sensitive area. 2.3 THEA II – GROUNDING 2.3.1 Main engine failure Thea II, which was 23 years old, was a well-maintained vessel with a reputation for being mechanically reliable. Its C/E, with over 13 years’ experience on board, was able to quickly identify the engine defect. This revealed that Thea II had lost propulsion due to the failure of its main engine governor circuit board. Without a replacement component, the engine could not be repaired. Thea II did not carry a spare circuit board, as such a repair would normally be carried out by service engineers. Consequently the master had no option but to anchor his vessel and summon assistance. 2.3.2 Thea II dragging anchor Thea II maintained its position for almost 90 minutes, lying to its port anchor with 5 shackles of cable in the water. The master believed that his vessel would be at anchor for only a short period awaiting the arrival of tugs; the amount of cable he had deployed reflected this. However, as the wind increased to storm force 10, the vessel started to yaw, and shortly afterwards began to drag its anchor. The master’s response was to deploy a further shackle of cable.

28 This was unsuccessful in stopping the anchor dragging, and Thea II continued to drift towards the Binks shoal at 1.7kts. The weather conditions were now extreme and, while they exceeded the International Association of Classification Societies’ (IACS) design threshold for the vessel’s anchor,7 the master still had several options to arrest or slow Thea II’s drift towards the shore. The first option was to deploy all remaining cable on the port anchor. Thereafter the master could also, as the vessel’s SMS advised, have deployed the starboard anchor, either ‘underfoot’, or as a second anchor. Had the master taken any of these actions they might have stopped, or at least reduced Thea II’s rate of drift towards the Binks shoal. Had the vessel’s rate of drift been reduced by 0.5kt, it would have allowed the tugs a further 25 minutes to secure lines and tow Thea II to safety. The master was reluctant to drop a second anchor because, like many mariners, he was concerned that the two anchor cables would become entangled. In addition, at 1505 the VTSO had advised Thea II not to drop a second anchor because it might have hampered or prevented the vessel from securing tugs. Therefore, while it is unsurprising that Thea II dragged its anchor in such severe weather conditions, the master could have taken further action to slow or arrest the vessel’s drift. Moreover, the additional time gained might have been sufficient for the tugs to attach their lines and prevent the vessel from grounding. 2.4 SVITZER JOSEPHINE – GROUNDING 2.4.1 Master’s appreciation of risk Svitzer Josephine’s master left Immingham expecting to tow a disabled ship, anchored by the side of the main channel, into harbour. Yet, while on passage downriver the task changed to intercepting and towing to safety a disabled vessel that was dragging its anchor towards shallow water. As the tug rounded Spurn Head, the master began to appreciate the severity of the weather and the challenge associated with passing a towline to the disabled vessel. His VHF conversations with VTS indicate that he was unsure as to whether to proceed. However, with the AHM’s assurance, he decided to attempt to secure a towline, as his vessel and Svitzer Castle were the only assets available to prevent Thea II from grounding. The task faced by Svitzer Josephine’s master was unusual. Routinely employed in harbour towage, the master was more accustomed to operating in the familiar waters of the Humber Estuary, under the direction of a ship’s master or harbour pilot. However, on this occasion Svitzer Josephine had left its normal area of operation and was proceeding independently, making a challenging night time approach, in storm force conditions, to assist a disabled vessel. As the master turned the tug to make the approach to Thea II’s starboard shoulder, his crew of three were fully occupied. The OOW was on deck passing the towline to the container vessel and the engineer was operating the towing winch. The master was manoeuvring the tug, monitoring the activity on the upper deck of both vessels, while occasionally glancing at the tug’s navigational equipment (radar, ECS and echo sounder). It was therefore unsurprising that he lost positional awareness and had not seen that both vessels would shortly ground on the Binks shoal. 7 IACS required anchor equipment to moor the vessel up to 135m in length in winds up to 25m/s (49.6kts) and currents up to 2.5m/s (4.9kts).

29 The guidance on the conduct of a tug’s navigation in Svitzer’s SMS was detailed but did not match day to day practice. Svitzer Josephine was fitted with an ECS, and the crew were therefore obliged to use paper charts as the primary means of navigation. However, chart work was not a routine activity on board the tug, and the paper charts therefore remained in the storage tube. Instead, Svitzer Josephine’s master relied on his knowledge and experience to keep the vessel navigationally safe. Ordinarily, a further safeguard was that, when tugs were towing larger, deeper vessels, the pilot or master of those vessels would help keep the tugs navigationally safe by ensuring that the vessel they were piloting did not ground. Even if BA chart 104 had been on board, given that all three of the crew were fully occupied the use of paper charts to plot and monitor the tug’s position was impracticable. The echo sounder alarm, prone to false alarm, had been silenced and, while it would have indicated the approaching shoal, neither the master nor the OOW were visually monitoring it. Svitzer Josephine’s master did not fully appreciate the level of risk he was exposing his vessel to when he took the tug out of the main channel to attend Thea II. He did not have a passage plan to keep his vessel safe from grounding or an effective method of monitoring his position. The tug grounded because the master became task focused on passing the towline to Thea II, lost positional awareness and did not appreciate the rate of drift of either the casualty vessel or his own tug towards the Binks shoal. 2.4.2 Vessel Traffic Services Humber’s advice Until the harbourmaster took over at 1900, VTS Humber led the port’s response to this accident. The VTSOs were quick to spot that Thea II had lost propulsion, and identified when the vessel started to drag its anchor. However, as the disabled container vessel began to drag its anchor at 1.7kts towards the Binks shoal, their advice became less helpful. In particular, when Svitzer Josephine’s master weighed up the risk of leaving the fairway to provide assistance to Thea II, the VTSO was keen that he should assist the casualty. He even offered to “keep an eye” on the tug, letting it know if it “got too close to the bank”. In effect, VTS staff went beyond their remit of providing an information and traffic service, tacitly offering the tug navigation advice, albeit caveated by the statement that they did not have a chart. In the event, VTS staff provided no warning to Svitzer Josephine before the tug grounded. ABP Humber’s new VTS facility at Grimsby was state of the art, and combined meteorological, tide, radar and AIS information, all of which could be displayed alongside the latest chart data. The AHM and his VTS team had the tools necessary to estimate the time of Thea II’s grounding. Had they done so, and had they shared this information with Svitzer Josephine, it might have prompted the tug master to move his vessel to deeper water and prevent his vessel grounding. While it might not have been appropriate for the VTSOs to provide the tug with direct navigational advice, they could have used the SMCP message markers ‘caution’ or ‘warning’ to alert both vessels to the approaching shoal. While responsibility for the safety of the tug rested with its master, timely advice from VTS Humber might have prompted him to take action to prevent his vessel grounding.

30 2.5 SALVAGE OF THEA II 2.5.1 Emergency response by ABP Humber Once Thea II and Svitzer Josephine were aground, ABP initiated HESMEP. The harbourmaster, who had been briefed on the accident by phone, arrived at the MRC at about 1900 and took over control of the port’s emergency response. His initial concern was for the safety of those on board the two grounded vessels. The RNLI ALB and SAR helicopter were deployed. However, by 1930, having established that neither master wished to evacuate his crew, the harbourmaster’s priority had shifted to minimising risk to the environment. To do this he needed to expedite the salvage of both vessels before they grounded again as the tide fell. As the tide rose Svitzer Josephine refloated and returned to harbour under its own power. However, Thea II remained aground. The harbourmaster’s challenge was that, while he held the legal responsibility for managing emergencies within the port, he had no direct control of any salvage assets, such as tugs, to recover Thea II. Strictly speaking ABP’s IOT fire tug agreement only contracted Svitzer tugs to support emergencies at Immingham and South Killingholme terminals. While the statistics of machinery breakdowns in ABP Humber’s area would suggest this lack of salvage assets would be a concern, in practice these breakdowns had always been resolved, without external assistance to the vessel’s crew. With a Svitzer tug aground, the harbourmaster commenced discussions with one of the other tug operators, SMS Towage Ltd, to establish whether they had any assets that could assist. This led to their shallow draught Hong Kong class tugs being deployed to assist the casualty. He also requested that the coxswain of the RNLI ALB use his vessel’s echo sounder to survey the water around Thea II so that the tugs could be safely guided into position when they arrived at the scene. Concurrently, he kept the SOSREP and other stakeholders apprised of developments, as best practice would dictate. The harbourmaster’s plan was put into place. By 2330 Thea II was afloat, with two tugs close enough to pass lines and tow the vessel clear. However, with the tide falling there was now an urgent need to tow the casualty vessel into safe water. The only option available to the harbourmaster was to use his statutory powers to issue a special direction ordering Thea II’s master to take the tugs’ lines. Even then, Thea II’s master refused to do so until he had received his manager’s approval. By transmitting the special direction to the master solely by VHF radio, its legal significance might not have been fully appreciated. Had it been backed up by email to Thea II, its agent, managers and insurers might all have better understood the status and importance of the harbourmaster’s special direction. Furthermore, the master, whose first language was not English, and German managers who, by this stage, were controlling the decision whether to take tug’s lines, would have both received the same message. Given the assets available, ABP Humber’s emergency response to the grounding of both vessels was well managed. The harbourmaster correctly assessed that if Thea II was not towed clear during this tidal cycle, there would have been a significant risk of the vessel grounding on the Spurn peninsula, with the increased likelihood of environmental damage to several environmentally sensitive sites. The harbourmaster used his power of special direction appropriately and, when

31 unsuccessful, engaged with the SOSREP to ensure the timely removal of the casualty. In future, to guarantee that harbourmasters’ special directions are fully understood by all parties, the port should consider reiterating significant VHF communications by email. This would help ensure that key decision makers, who might not be on board the vessel, understand the significance of the direction given and the gravity of the situation. 2.5.2 Emergency response by Thea II’s managers The initial emails from Thea II’s master to the vessel’s managers, insurers and agent advised them that the crew were unable to repair the engine defect and that the vessel was at anchor awaiting tugs to tow it into harbour. In these initial emails, the master did not inform the company of the severe weather conditions, or that the vessel was dragging its anchor. In return, the company advised the master that: “Towage is the worst option always! …all officers have to know. If towage is requested, then you have to ask insurance broker always first!!!” Shortly after receiving the advice above, the master informed the company that Thea II was aground. Having been unable to speak to the master by phone, the DPA’s appreciation of the vessel’s situation was limited by the master’s brief emails. As a result, believing there was plenty of time, his advice to the master was not to ask for assistance from tugs, nor sign LOF, while the insurance brokers sourced suitable towage assets to recover the vessel. As high water approached, the master’s emails became more urgent. He repeatedly asked Thea II’s managers for their intentions to tow his vessel to safety. Once Thea II was afloat, the frequency of the master’s requests for advice increased and the tone of his emails became more anxious. These email exchanges indicate that the managers still did not fully appreciate their vessel’s predicament. They were unaware of the exact location of the grounded Thea II, the severity of the weather, and the limited tidal window available to recover their vessel. They were also unaware that Thea II was aground in an environmentally sensitive area. In part, this was due to the limited detail contained in the master’s concise emails and a lack of verbal communication directly with the master. However, it was also as a result of Thea II’s managers’ focus on the commercial aspects of the tow, at the expense of them seeking a detailed understanding of their vessel’s situation. It was difficult to gain such an understanding while operating from home, without access to a full set of navigational publications that would have been available in their office. A further issue that delayed Thea II’s salvage was the managers’ instruction to the master not to engage tugs, nor sign LOF, without their permission. This approach undermined the master’s authority. When the harbourmaster issued his special direction, Thea II’s master, aware of his vessel’s precarious situation, wanted to take tugs’ lines but did not feel empowered to do so. He was very conscious that the tide was falling and, unless his vessel was under tow by approximately 0130, it was likely that Thea II would re-ground. Furthermore, the only tugs close enough to assist were Yeoman and Irishman, already at the scene. However, despite the master’s repeated requests, the managers took a further 1 hour 24 minutes before they granted approval for Thea II to take tugs’ lines; a delay that meant the ship was towed clear of the bank with an estimated 20 minutes to spare before it would have re-grounded.

32 For Thea II’s managers, this accident probably presented them with one of the most demanding emergency situations that they had faced. Unable to speak to the master by phone, they had to gather as much information as possible about their vessel’s situation from the master’s brief emails, but they did not seek out other information sources such as AIS or local information. Their priority was agreeing the best possible terms with tugs and salvors without fully appreciating the full situation faced by the crew of Thea II. They had to balance the safety of the crew, marine environment and the ship and cargo. To successfully do so required a well- established and exercised emergency response organisation that had as much information as possible at its disposal. In this case, their organisation, operating remotely, did not grasp the seriousness of the situation, the need for urgent action, or the potential environmental risk posed by their vessel. 2.5.3 Emergency response by the SOSREP and the DCPSO Humber coastguard kept the DCPSO informed of Thea II’s situation throughout the incident. When it became clear that, despite direction Thea II had not taken tugs’ lines, the DCPSO began to actively lobby the vessel’s owners/insurers to persuade them to take urgent action to safeguard their vessel. These discussions culminated with the DCPSO informing Thea II’s insurers that if the vessel had not taken tugs’ lines by 0100 on 16 December, the SOSREP would issue a ‘formal direction’ for the vessel to do so. The DCPSO’s approach followed the intent of the UK NCP. This required the harbourmaster to take the lead for managing accidents within their port area, keeping the SOSREP informed so that he could tacitly approve the actions being taken. However, when the harbourmaster’s special direction failed to elicit the required response, the DCPSO’s threat that the SOSREP would intervene had the desired effect, causing the insurers to agree that the container vessel should be salvaged under LOF. The DCPSO’s threat of action by SOSREP worked and ultimately ensured that Thea II was towed clear into safe water. Importantly, this accident and the grounding of Kuzma Minin (see 1.13.2), demonstrate the importance of SOSREP’s regular engagement with the ports’ industry and other stakeholders to promote his role in counter-pollution and salvage operations around the United Kingdom coast. 2.6 THE TOW TO THE ANCHORAGE At approximately 2200, Svitzer Josephine jettisoned its tow and got underway. This left Thea II with 130m of heavy, 80mm towline attached to its forecastle, which needed to be cleared before Yeoman’s line could be made fast. Having examined the 43m of towline recovered from Yeoman’s propellers, it appears that Thea II’s crew attempted to recover Svitzer Josephine’s hawser. However, having brought the first 43m on board, the crew made the decision to cut and jettison the remaining 87m. Subsequently, as conditions moderated once Thea II was clear of the Binks, the crew made the decision to dispose of the remaining 43m of towline.

33 This remnant of the towline was probably thrown overboard just before 0200. While the 80mm hawser sank, the buoyant 15m messenger did not, and it was drawn into Yeoman’s propellers. This caused the tug to lose the use of both its ASDs. Fortunately, Irishman was able to control the tow until the tug Pullman arrived on scene and took over the tow from the disabled Yeoman. Emergencies require rapid decisions to be taken, the consequences of which are often not fully understood at the time. In this case, Svitzer Josephine’s towline being jettisoned by Thea II’s crew disabled Yeoman and almost undermined the integrity of the towing operation to salvage the vessel. Had Thea II’s crew made a proactive decision to recover the towline on board, instead of simply jettisoning it, the disabling of Yeoman could have been avoided.

34 SECTION 3 - CONCLUSIONS 3.1 SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS

Thea II lost propulsion due to the failure of its main engine governor central processing unit. [2.3.1]

Thea II’s crew could not repair the main engine. As a result, Thea II’s master had no choice but to anchor and request tug assistance. [2.3.1]

It was unsurprising that Thea II dragged its anchor in the severe weather conditions. However, had Thea II’s master deployed a second anchor and/or increased the scope of the anchor already deployed, it would have slowed, and might have arrested the vessel’s rate of drift towards the Binks shoal, allowing more time for a tug to secure a towline. [2.3.2]

While VTS Humber had earlier advised Thea II not to deploy a second anchor, it was the master’s concern - that a second anchor would become entangled with the first - that resulted in his decision not to deploy the starboard anchor. [2.3.2]

Svitzer Josephine’s master did not fully appreciate the level of risk he was exposing his vessel to when he took the tug out of the main channel to attend Thea II. [2.4.1]

Svitzer Josephine grounded because the master, focused on passing the towline to Thea II, lost navigational positional awareness and did not appreciate the rate of drift of both vessels towards the Binks shoal. [2.4.1]

The assistant harbourmaster and his VTS team had the tools necessary to estimate the time of Thea II grounding. Had they done so, this would have allowed the time available to secure Svitzer Josephine to be calculated, enabling a more informed decision or additional precautions to be taken. [2.4.2]

While responsibility for the safety of the tug rested with its master, timely advice from VTS Humber might have prompted him to take action to prevent Svitzer Josephine grounding. [2.4.2]

The harbourmaster correctly assessed that if not towed clear during this tidal cycle, it is likely Thea II would have been blown on to the Spurn peninsula, risking potential damage to several environmentally sensitive sites. [2.5.1]

The harbourmaster used his powers of special directions appropriately and, when Thea II did not take the tugs’ lines, successfully engaged the Secretary of State’s Representative to ensure Thea II was towed into safe water. [2.5.1]

Transmitting the harbourmaster’s special direction by email, as well as by radio, to Thea II’s master, agent and managers would have ensured that all of the key decision makers understood the significance of the direction, the gravity of the situation and the need to take urgent action. [2.5.1]

The vessel’s managers’ instruction to Thea II’s master, not to take tugs’ lines without their prior approval, undermined his authority to act in the best interest of his ship and crew. [2.5.2]

35

It is estimated that Thea II was only 20 minutes from re-grounding on a falling tide when the vessel’s managers agreed that the master could accept a tow under Lloyd’s Open Form. [2.5.2]

Thea II’s managers, operating remotely from home, did not fully appreciate the seriousness of the situation, the need for urgent action, or the environmental risk posed by their vessel. [2.5.2]

The duty counter-pollution and salvage officer’s threat of action by the Secretary of State’s Representative worked, and ultimately ensured that Thea II was towed clear into safe water. [2.5.3] 3.2 OTHER SAFETY ISSUES NOT DIRECTLY CONTRIBUTING TO THE ACCIDENT8

The remnant of the towline discarded by Thea II’s crew, subsequently fouled the tug Yeoman’s propellers, disabling it. Fortunately, Irishman was able to control the tow until the tug Pullman replaced Yeoman. [2.6] 8 These safety issues identify lessons to be learned. They do not merit a safety recommendation based on this investigation alone. However, they may be used for analysing trends in marine accidents or in support of a future safety recommendation

36 SECTION 4 - ACTIONS TAKEN ABP Humber has:

● Conducted an internal investigation into this accident.

● Reviewed relevant risk assessments.

● Ensured local tugs are carrying relevant charts.

● Reviewed:

○Fire tug arrangements; and,

○The Memorandum of Understanding between ABP and the Humber coastguard. Svitzer Marine Ltd has:

● Conducted an internal investigation into this accident and delivered several recommendations to:

○Review the use of echo sounder and electronic chart systems.

○Improve the governance of, and preparation for, ‘special service’ and ‘non- routine’ towing operations.

○And it has reiterated the requirement to produce passage plans when operating outside the tug’s normal area of operations. TS-Shipping GmbH & Co. KG. has:

● Reviewed the safety critical engine spares carried on Thea II with the engine manufacturer.

37 SECTION 5 - RECOMMENDATIONS TS-Shipping GmbH & Co. KG. is recommended to: 2020/128 Review the company’s emergency response organisation and procedures with the aim of improving decision making and the clarity of advice provided to its vessels. Safety recommendations shall in no case create a presumption of blame or liability

Marine Accident Report

FD06064 ｜ 2020_英国_man-overboard-from-dredger-cherry-sand-with-loss-of-1-life__2020-9-CherrySand-Annexes.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2020 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A Extracts from the Merchant Shipping Notice 1886 (M+F) Maritime Labour Convention, 2006 Work in Fishing Convention, 2007 (ILO No. 188) Medical Examination System: Appointment of Approved Doctors and Medical and Eyesight Standards

Relationship with the Merchant Shipping (Maritime Labour Convention) (Medical Certification) Regulations 2010, as amended  The MLC 2010 Regulations need to be read in conjunction with this Merchant Shipping Notice, as it sets out some of the key details which supplement the legal obligations in those Regulations. Failure to comply with those obligations may be a criminal offence under the Regulations.  In order to assist with this Paragraph 8 contains an index which clearly sets out and/or identifies the provisions in this Notice which relate to the relevant provision in the Regulations. Relationship with the Merchant Shipping (Work in Fishing Convention) (Medical Certification) Regulations 2018  The FV 2018 Regulations need to be read in conjunction with this Merchant Shipping Notice, as it sets out some of the key details which supplement the legal obligations in those Regulations. Failure to comply with those obligations may be a criminal offence under the Regulations.  In order to assist with this Paragraph 9 contains an index which clearly sets out and/or identifies the provisions in this Notice which relate to the relevant provision in the Regulations.

1. Introduction
2. 1 The Merchant Shipping (Maritime Labour Convention) (Medical Certification) Regulations

2010, (referred to in this Notice as “the MLC 2010 Regulations”) and the Merchant Shipping (Work in Fishing Convention) (Medical Certification) Regulations 2018 (“the FV 2018 Regulations”) implement the medical certification provisions of the Maritime Labour Convention 2006 and the ILO Work in Fishing Convention (No. 188) for the UK. 1.2 They also implement, for merchant seafarers, the requirements of clause 13 of the Social Partners’ Agreement set out in the Annex to Council Directive 1999/63/EC of 21 June 1999 (the Seafarers’ Working Time Directive)1. 1.3 The provisions set out in this notice also comply with Regulation I/9 and Standard A.I/9 of the IMO’s International Convention on Standards of Training, Certification and Watchkeeping (STCW), to which the UK is a signatory and the International Convention on Standards of Training, Certification and Watchkeeping for Fishing Vessel Personnel (STCW(F), which the UK has not yet ratified. Terminology 1.4 In this notice, 1.4.1 the fisherman or seafarer being examined or in respect of whom a decision is made about their fitness to work at sea or to fish is referred to as “the candidate”; 1.4.2 “Ship” includes “fishing vessel”. 1.4.3 “Master” includes skipper. 1.4.4 “go to sea”, “sea service” etc includes work on a fishing vessel regardless of area of operation. 1 As amended by Council Directive 2009/13/EC. The requirements of the remaining clauses of that Agreement, relating to hours of work, are implemented separately in the Merchant Shipping (Maritime Labour Convention) (Hours of Work) Regulations 2018.

1. 4 Eyesight Standards
2. 4.1 The Approved Doctor must also ensure that the candidate meets the visual acuity

and colour vision Standards. The UK’s Standards for visual acuity comply with the international standards laid down in the IMO Convention on the Standards of Training, Certification and Watchkeeping (STCW 78 as amended). Candidates will be tested for each eye separately and must meet both the higher standard required with visual aids (where needed), and a minimum standard without visual aids, in order to ensure that in an emergency they have adequate vision to carry out emergency procedures safely. For colour vision, candidates are screened using Ishihara plates. It is essential that candidates applying for Certificates of Competency as deck or dual career (merchant/fishing) officers have full colour vision. 3.4.2 A deck applicant who fails the Ishihara plate test may arrange for their colour vision to be re-tested at an MCA approved CAD test centre (listed at Annex B). The CAD test centre will charge a fee for the CAD test, which is payable directly to the test centre. Failure in this test will mean that a medical certificate may only be issued with a restriction precluding lookout duties at night. Trainee Deck officers however will be made permanently unfit. Other deck candidates applying for their first MCA Certificate of Competency (CoC), Able Seafarer Deck Certificate of Proficiency or Rating Certificate who have failed the Ishihara plate test will only be considered if they have subsequently passed the MCA Holmes Wright B (HWB) lantern test or CAD test and hence carry no restriction on their seafarer medical certificates precluding lookout duties.  Existing deck seafarers who have failed the Ishihara plate test but passed the MCA HWB lantern test and have evidence of this will still be considered to meet the colour vision standards and do not need to undertake a CAD test.  Existing deck seafarers who have failed the Ishihara plate test and passed the HWB lantern test but do not have evidence of this will need to take the MCA CAD test. If they do not take the test, or fail it, they may continue working in their current roles. They will be given an ENG 1 noting that they do not meet the colour vision standards with an appropriate restriction.  Existing deck seafarers who wish to qualify for an MCA CoC for the first time, or a higher CoC, will have to meet the colour vision standards (i.e. Ishihara plates or MCA CAD test passed, or provide evidence that they have previously passed the MCA HWB lantern test). Their seafarer medical certificate must not have any restriction on lookout duties.

Table - SUMMARY OF STANDARDS REQUIRED Category of Seafarer Basic Visual Acuity Standard (unaided) Higher Visual Acuity Standard (aided if necessary) Near Vision (both eyes together aided or unaided) Colour Vision Visual Field Deck or dual career Better eye 6/60 Other eye 6/60 Better eye 6/6 Other eye 6/12 N8 Ishihara or CAD No pathological field defect Engineer/ Radio 6/60 6/18 6/18 N8 Ishihara or Farnsworth D15 or City University Sufficient to undertake duties efficiently Others Sufficient to undertake duties efficiently Those who become monocular in service with no evidence of progressive eye disease in the remaining eye Deck

6/60

6/6

N8 Ishihara or CAD No pathological field defect Eng/Radio

6/60

6/9

N8 Ishihara or Farnsworth D15 or City University Sufficient to undertake duties efficiently Others Sufficient to undertake duties efficiently There should be a sufficient period of adaptation after becoming monocular to enable stairs to be descended rapidly and safely. Notes 1. No diplopia, congenital night blindness, retinitis pigmentosa or any other serious or progressive eye disease is permitted. 2. If bifocal glasses are worn there should be a period of adaptation first because of the risk of falls. 3. Where glasses or contact lenses are needed to meet the vision standard, a spare pair (distance and near vision if necessary) should be carried. 4. Aids to colour vision e.g. red-tinted x-chroma, chromas lenses and chromagen lenses are not permitted. 5. Seafarers who suffer pathological field defects should have a field of vision at least 120o in the horizontal measured by the Goldman perimeter using the iii/4 setting (or equivalent perimetry). In addition there should be no significant defect in the binocular field which encroaches within 20o of fixation above or below the meridian. Homonymous or bitemporal defects which come close to fixation whether hemianopic or quadrantopic are not accepted.

MSN 1886(M+F) Appendix 2 to Annex A Table - GUIDANCE ON ASSESSMENT OF MINIMUM ENTRY-LEVEL AND IN-SERVICE PHYSICAL ABILITIES FOR SEAFARERS Shipboard task, function event or condition 3 Related physical ability A medical examiner should be satisfied that the candidate4 Routine movement around vessel: – on moving deck – between levels – between compartments Maintain balance and move with agility Climb up and down vertical ladders and stairways Step over coamings (e.g. Load Line Convention requires coamings to be600 mm high) Open and close watertight doors Has no disturbance in sense of balance Does not have any impairment or disease that prevents relevant movements and physical activities Is, without assistance 5, able to: – climb vertical ladders and stairways – step over high sills – manipulate door closing systems Note 1 applies to this row Routine tasks on board: – use of hand tools – movement of ship’s stores – overhead work – valve operation – standing a four–hour watch – working in confined spaces – responding to alarms, warnings and instructions – verbal communication Strength, dexterity and stamina to manipulate mechanical devices Lift, pull and carry a load (e.g. 18 kg) Reach upwards Stand, walk and remain alert for an extended period Work in constricted spaces and move through restricted openings (e.g. SOLAS requires minimum openings in cargo spaces and emergency escapes to have the minimum dimensions of 600 mm × 600 mm – SOLAS regulation 3.6.5.1) Visually distinguish objects, shapes and signals Hear warnings and instructions Give a clear spoken description Does not have a defined impairment or diagnosed medical condition that reduces ability to perform routine duties essential to the safe operation of the vessel Has ability to: – work with arms raised – stand and walk for an extended period – enter confined space – fulfil eyesight standards (table A–I/9) – fulfil hearing standards set by competent authority or take account of international guidelines – hold normal conversation Note 1 applies to this row Emergency duties 6 on board: – Escape – Firefighting – Evacuation Note 2 applies to this row Don a lifejacket or immersion suit Escape from smoke-filled spaces Take part in firefighting duties, including use of breathing apparatus Take part in vessel evacuation procedures Does not have a defined impairment or diagnosed medical condition that reduces ability to perform emergency duties essential to the safe operation of the vessel Has ability to: – don lifejacket or immersion suit – crawl – feel for differences in temperature – handle firefighting equipment – wear breathing apparatus (where required as part of duties) Notes to the table:

1. Rows 1 and 2 describe: (a) ordinary shipboard tasks, functions, events and conditions; (b) the corresponding

physical abilities which may be considered necessary for the safety of a seafarer, other crew members and the ship; and (c) high-level criteria for use by medical practitioners assessing medical fitness, bearing in mind the different duties of seafarers and the nature of shipboard work for which they will be employed.

• Row 3 describes: (a) ordinary shipboard tasks, functions, events and conditions; (b) the corresponding physical

abilities which should be considered necessary for the safety of a seafarer, other crew members and the ship; and (c) high-level criteria for use by medical practitioners assessing medical fitness, bearing in mind the different duties of seafarers and the nature of shipboard work for which they will be employed.

• This table is not intended to address all possible shipboard conditions or potentially disqualifying medical

conditions. Parties should specify physical abilities applicable to the category of seafarers (such as “Deck officer” and “Engine rating”). The special circumstances of individuals and for those who have specialized or limited duties should receive due consideration.

• If in doubt, the medical practitioner should quantify the degree or severity of any relevant impairment by means

of objective tests, whenever appropriate tests are available, or by referring the candidate for further assessment.

• The term “assistance” means the use of another person to accomplish the task.
• The term “emergency duties” is used to cover all standard emergency response situations such as abandon

ship or firefighting as well as the procedures to be followed by each seafarer to secure personal survival. Annex B UK Dredging risk assessment UKD/002

Annex C Cherry Sand’s Safe System of Work No.5

Annex D UK Dredging New Joiner Induction

FD06065 ｜ 2020_英国_safety-warning-about-carbon-monoxide-poisoning-after-the-loss-of-2-lives-on-the-motor-crui.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2020 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H SAFETY BULLETIN SAFETY BULLETIN SB2/2020 OCTOBER 2020 Carbon monoxide poisoning on board the motor cruiser Diversion on the River Ouse, in York, resulting in two fatalities, on 4 December 2019 Extracts from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 Regulation 5: “The sole objective of a safety investigation into an accident under these Regulations shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of such an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” Regulation 16(1): “The Chief Inspector may at any time make recommendations as to how future accidents may be prevented.” Press Enquiries: 01932 440015 Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000 NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2020 See http://www. nationalarchives.gov.uk/doc/ open-government-licence for details. All bulletins can be found on our website: https://www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459

MAIB SAFETY BULLETIN 2/2020 This document, containing safety lessons, has been produced for marine safety purposes only, on the basis of information available to date. The Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 provides for the Chief Inspector of Marine Accidents to make recommendations or to issue safety lessons at any time during the course of an investigation if, in his opinion, it is necessary or desirable to do so. The Marine Accident Investigation Branch is carrying out an investigation into into the fatal carbon monoxide poisoning on board the motor cruiser Diversion. The MAIB will publish a full report on completion of the investigation. Andrew Moll Chief Inspector of Marine Accidents NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall not be admissible in any judicial proceedings whose purpose, or one of whose purposes, is to apportion liability or blame. This bulletin is also available on our website: www.gov.uk/maib Press Enquiries: 01932 440015 Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000

1 BACKGROUND On 4 December 2019, the owner of the 9.14m motor cruiser Diversion, and a friend, were killed by carbon monoxide poisoning after they returned from an evening out and prepared to sleep on board whilst the boat was moored on the River Ouse in York. The motor cruiser had been built by the owner over a period of 11 years and was completed in 2001. Since then, the owner had carried out several modifications to the boat, including the installation of a second-hand diesel-fuelled cabin heater in 2018. It was a cold evening and the cabin heater had been left running while the two casualties were eating and drinking with friends in the city centre. INITIAL FINDINGS The cabin heater (Figure 1) had been installed in a cabinet beneath the boat’s control panel and steering position. It drew combustion air from inside the boat and diesel fuel from one of the boat’s fuel tanks. The fuel/air mixture was burnt in the heater and the exhaust gas was discharged to outside, via a silencer, through an adjacent hull fitting. Circulating air was drawn from inside the cabinet, warmed by the heater and discharged into the cabin, via a nozzle mounted at deck level next to the steering position (Figure 2). The initial investigation found that the exhaust silencer was not approved for marine use, it was not gas tight and its connection to the exhaust pipework was loose. Consequently, a small but steady flow of carbon monoxide was leaking into the cabinet, which was then drawn into the heater’s air inlet and blown into the cabin space. As the carbon monoxide enriched air was drawn into the heater’s combustion chamber, and recirculated through the cabin, the cycle repeated itself and the levels of carbon monoxide gradually increased to lethal levels. The initial installation of the diesel heater was not checked by a professional engineer and it had never been serviced. The heater was one of four potential sources of carbon monoxide on board the motor cruiser, which was not fitted with a carbon monoxide alarm. The other potential sources were the boat’s engine, a portable petrol generator and a methylated spirit fuelled cooker. The MAIB has investigated fatal accidents1 where all three were identified as the source of the deadly toxic gas. It is a mandatory requirement under the Canal and Rivers Trust’s and the Environmental Agency’s Boat Safety Scheme (BSS) to fit carbon monoxide alarms on board all boats sailing on most inland waterways. Diversion was last surveyed, and its BSS Certificate re-issued in February 2019; the requirement to fit an alarm became mandatory 6 weeks later on 1 April 2019. 1 MAIB investigations: Eschol (report no 24/2014); Arniston (report no 2/2015); Love for Lydia (report no 9/2017) and Vasquez (report no 18/2017).

2 Figure 1: Heater installation with exhaust insulation removed Non-marine silencer leaking CO into air inlet Flexible exhaust pipe Combustion air inlet

3 Figure 2: Heater position Air outlet Heater under console

4 SAFETY LESSONS

● Diesel-fuelled cabin heaters are frequently installed in boats, trucks, caravans and motor homes. They are readily available to purchase both new and second-hand and may not necessarily be installed by a professional installation engineer. However, it is essential that they are installed in accordance with the manufacturer’s instructions and all components, used in a marine environment, should be approved for marine use. They should also be inspected by a suitably qualified engineer after installation and be regularly serviced and tested for leaks.

● Carbon monoxide is a by-product of combustion appliances fuelled by oils, solid fuel or gas. It has no smell, no taste, is colourless and is extremely difficult for human senses to detect. Therefore, it is essential that carbon monoxide alarms are fitted in areas where carbon monoxide could accumulate and pose a risk to health (such as the accommodation areas of boats). When selecting a carbon monoxide alarm, preference should be given to those marked as meeting safety standard EN 50291- 2:2010, which are intended for use in a marine environment. It is essential to fit alarms following the manufacturer’s guidance, to test them routinely using the test button and not to ignore them.

● Carbon monoxide is a silent killer. Its symptoms can be similar to colds, flu, hangovers or even Covid-19; headaches, dizziness, nausea, vomiting, tiredness, confusion, stomach pain and shortness of breath. If carbon monoxide poisoning is suspected, stop the source, get to the open air and seek medical attention.

FD06066 ｜ 2020_英国_safety-warning-after-inadvertent-activation-of-condensed-aerosol-fire-extinguishing-system.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2020 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H SAFETY BULLETIN SB1/2020 MARCH 2020 Inadvertent discharge of a FirePro condensed aerosol fire extinguishing system during its installation on board the fishing vessel Resurgam (PZ 1001) on 15 November 2019 resulting in one fatality Extracts from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 Regulation 5: “The sole objective of a safety investigation into an accident under these Regulations shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of such an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” Regulation 16(1): “The Chief Inspector may at any time make recommendations as to how future accidents may be prevented.” Press Enquiries: 01932 440015 Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000 NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2020 See http://www. nationalarchives.gov.uk/doc/ open-government-licence for details. All bulletins can be found on our website: https://www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: 023 8039 5500 Fax: 023 8023 2459 SAFETY BULLETIN Image courtesy of www.marinetraffic.com Resurgam

MAIB SAFETY BULLETIN 1/2020 This document, containing safety lessons, has been produced for marine safety purposes only, on the basis of information available to date. The Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 provides for the Chief Inspector of Marine Accidents to make recommendations or to issue safety lessons at any time during the course of an investigation if, in his opinion, it is necessary or desirable to do so. The Marine Accident Investigation Branch is carrying out an investigation into the fatality of a shore- based engineering apprentice who was working in the engine room of the fishing vessel Resurgam in Newlyn on 15 November 2019. The MAIB will publish a full report on completion of the investigation. Andrew Moll Chief Inspector of Marine Accidents NOTE This bulletin is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall not be admissible in any judicial proceedings whose purpose, or one of whose purposes, is to apportion liability or blame. This bulletin is also available on our website: www.gov.uk/maib Press Enquiries: 01932 440015 Out of hours: 020 7944 4292 Public Enquiries: 0300 330 3000

1 BACKGROUND On 15 November 2019, the UK registered fishing vessel Resurgam was in Newlyn, England undergoing maintenance. An engineer and an apprentice from the owner’s shore-based support team were working on the main engine in the engine room. Also working in the engine room were two contractors installing a new FirePro condensed aerosol fire extinguishing system. During the installation and without warning, the fire extinguishing system partially and inadvertently discharged, filling the engine room with a dense cloud of aerosol fire suppressing particles (Figure 1). Both installation contractors and the company’s engineer managed to evacuate, but the apprentice collapsed in the engine room. He was later recovered by the local fire and rescue service but was found not breathing and could not be resuscitated. Figure 1: Typical discharge of a condensed aerosol fire suppressant (not at time of accident) Aerosol generator INITIAL FINDING The exact causes and circumstances of this accident are still being investigated and the findings will be published by the MAIB in a full investigation report. However, during the inadvertent discharge, it is evident that the apprentice inhaled a high concentration of the suppressant particles and this significantly contributed to the fatality.

2 FirePro’s Installation and User Manual and its product’s material safety data sheets had recognised the inadvertent discharge of the system, particularly during installation and maintenance, as a hazard. However, the loss of life was not identified as a potential outcome; therefore, the risk associated with inhaling or ingesting a large volume of the suppressant particles was not fully appreciated or protected against. SAFETY LESSONS Vessel owners, operators and those contracted to install FirePro and other similar condensed aerosol fire extinguishing systems should be fully aware of the potential risk to life from exposure to the aerosol particles. Safety precautions should be put in place to ensure that personnel are not exposed to this hazard:

● Prior to intentional discharge of a condensed aerosol system, there should be visible and audible alarms to alert personnel. Checks should also be made to ensure the protected compartment has been evacuated before the system is activated.

● When condensed aerosol fire extinguishing systems are being installed or maintained the system should be fully isolated to guard against inadvertent activation, non-essential personnel should be clear of the area and an enclosed space rescue plan should be in place. RECOMMENDATION FirePro is recommended to: S2020/114 Issue a safety alert to the owner/operators of vessels fitted with its systems and its network of marine installation/maintenance engineers highlighting the circumstances of this accident and advising them of appropriate measures to take to reduce the risk of exposure to fire suppressant particles. Safety recommendations shall in no case create a presumption of blame or liability

FD06067 ｜ 2020_英国_subsea-explosion-resulting-in-damage-to-crab-potting-vessel-galwad-y-mor-and-injuries-to-c.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2020 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

ACCIDENT REPORT ACCIDENT REPORT Report on the investigation of a subsea explosion resulting in crew injuries and vessel damage to the crab potting vessel Galwad-Y-Mor (BRD 116) 22 nautical miles north of Cromer, England on 15 December 2020 SERIOUS MARINE CASUALTY REPORT NO 1/2022 JANUARY 2022

The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. Front Cover image courtesy of Andrew Oliver © Crown copyright, 2022 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All MAIB publications can be found on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch First Floor, Spring Place 105 Commercial Road Southampton SO15 1GH United Kingdom Email: maib@dft.gov.uk Telephone: +44 (0)23 8039 5500 Press enquiries during office hours: +44 (0)1932 440015 Press enquiries out of hours: +44 (0)300 7777878

CONTENTS SYNOPSIS 1 SECTION 1 – FACTUAL INFORMATION 2 1.1 Particulars of Galwad-Y-Mor and accident 2 1.2 Narrative 3 1.3 Environmental conditions 5 1.4 Galwad-Y-Mor 6 1.4.1 Vessel description 6 1.4.2 Crew 6 1.4.3 Vessel operation 9 1.4.4 Vessel stability 9 1.5 Crew injuries and vessel damage 9 1.5.1 Crew injuries 9 1.5.2 Vessel damage 9 1.6 Examination of rope and metal fragment 13 1.6.1 Explosive trace examination 13 1.6.2 Metallurgical examination of metal fragment 13 1.7 Ordnance 18 1.7.1 Unexploded ordnance found at sea 18 1.7.2 Injuries to people on vessels subjected to subsurface exploding ordnance 18 1.7.3 Effect to vessels subjected to subsurface exploding ordnance 19 1.7.4 Records of at sea unexploded ordnance 19 1.7.5 Potential unexploded ordnance near the Galwad-Y-Mor accident location 20 1.7.6 Unexploded ordnance guidance available to mariners 20 1.8 Previous/similar accidents 23 SECTION 2 – ANALYSIS 24 2.1 Aim 24 2.2 The accident 24 2.2.1 The ordnance 24 2.2.2 The explosion 24 2.2.3 Crew injuries 24 2.2.4 Vessel damage 25 2.2.5 Anticipation of the danger from unexploded ordnance 25 2.2.6 UXO location information available to fishermen 26 2.3 Post-accident 26 2.3.1 Emergency response 26 2.3.2 Vessel survivability 26 SECTION 3 – CONCLUSIONS 27 3.1 Other safety issues directly contributing to the accident 27 SECTION 4 – ACTION TAKEN 28 4.1 MAIB actions 28 SECTION 5 – RECOMMENDATIONS 28

FIGURES Figure 1: Location of the accident Figure 2: Galwad-Y-Mor settled low in the water Figure 3: Galwad-Y-Mor general arrangement plan Figure 4: Galwad-Y-Mor hull and keel detail Figure 5: Galwad-Y-Mor crab potting arrangement Figure 6: Position of crew and injuries sustained at time of the accident Figure 7: Galwad-Y-Mor structural damage Figure 8: Galwad-Y-Mor machinery damage Figure 9: Galwad-Y-Mor interior damage Figure 10: Steel fragment detail Figure 11: SC250 high explosive bomb Figure 12: Possible UXO in Dudgeon Wind Farm survey area Figure 13: SC250 high explosive bomb found and detonated in Dudgeon Wind Farm survey area TABLES Table 1: Results of explosive trace examination Table 2: Elemental composition of the metal fragment

GLOSSARY OF ABBREVIATIONS AND ACRONYMS ALB

all-weather lifeboat DNT

dinitrotoluene EH

exploratory hunting EOD

explosive ordnance disposal FRB

fast rescue boat FTB

fast transfer boat GP

general purpose HE

high explosive kg

kilogram kW

kilowatt m

metre mm

millimetre MGN

Marine Guidance Note MOD

Ministry of Defence MSN

Merchant Shipping Notice MVC

microvoid coalescence ng

nanogram nm²

square nautical mile

OREI

Offshore Renewable Energy Installations

PFD

personal flotation device RDX

Research Department eXplosive RN

Royal Navy RNLI

Royal National Lifeboat Institution RoW

Receiver of Wreck t

tonne

TNT

trinitrotoluene UKAS

United Kingdom Accreditation Service

UKHO

UK Hydrographic Office UTC

universal time coordinated UXO

unexploded ordnance VHF

very high frequency WWI

The First World War WWII

The Second World War TIMES: all times used in this report are UTC unless otherwise stated. Galwad-Y-Mor Image courtesy of Andrew Oliver

1 SYNOPSIS On 15 December 2020, the 14.95m crab potting vessel, Galwad-Y-Mor, disturbed a piece of unexploded ordnance on the seabed while recovering crab pots in the North Sea, approximately 22 nautical miles off Cromer, England. The ordnance detonated and the ensuing explosion threw Galwad-Y-Mor up from the surface of the sea, causing significant crew injuries and damage to the vessel. The crew were rescued and evacuated to local hospitals and Galwad-Y-Mor was later towed to Grimsby. The MAIB investigation found that: ●The ordnance was an air-dropped bomb that had remained intact on the seabed since The Second World War. ●The bomb detonated on the seabed and the shock wave and gas bubble from the explosion hit Galwad-Y-Mor. ●The position of most seabed unexploded ordnance is unknown and Galwad-Y-Mor’s crew could not have anticipated the fouling of a bomb in the crab potting string. ●Galwad-Y-Mor’s crew training, experience, length of service together and emergency preparedness improved their survival chances. ●Galwad-Y-Mor’s hull was well constructed and able to withstand the force of the nearby seabed explosion. Based on this accident’s circumstances, no action has been taken by external stakeholders and no recommendations made. The aim of this report is to highlight the dangers that still exist with unexploded ordnance in the seas around the UK, and the actions to take should fisherman encounter any. In this case, the skipper and crew could not have foreseen the explosion and their level of preparedness to deal with such an emergency saved lives.

2 SECTION 1 – FACTUAL INFORMATION 1.1 PARTICULARS OF GALWAD-Y-MOR AND ACCIDENT SHIP PARTICULARS Vessel’s name Galwad-Y-Mor Flag United Kingdom IMO number/fishing numbers BRD 116 Type Crab potting vessel Registered owner The Galwad-Y-Mor Shellfish Company Manager(s) Not applicable Construction 2007 Year of build Steel Length overall 14.95m Registered length 12.90m Gross tonnage 63.23 VOYAGE PARTICULARS Port of departure Grimsby Port of arrival Grimsby Type of voyage Commercial Cargo information Shellfish Manning 7 MARINE CASUALTY INFORMATION Date and time 15 December 2020 at 1122 Type of marine casualty or accident Serious Marine Casualty Location of accident 53°18.59’N  001°15.46’E Place on board Hull and all compartments Injuries/fatalities Significant injuries to crew members Damage/environmental impact Extensive deformation to hull plating, engine room flooded and severe shock damage in all internal compartments Ship operation Fishing, recovering pots External & internal environment Wind, south-westerly force 3-4, sea state slight/moderate, visibility good Persons on board 7

3 1.2 NARRATIVE At about 2000 on 13 December 2020, the crab potting vessel, Galwad-Y-Mor left Grimsby with seven crew on board and spent the following day shooting and recovering strings of crab pots in the North Sea. Galwad-Y-Mor made passage overnight to the next crabbing ground; the skipper and five of the crew took rest while the night watchman navigated the vessel. On the morning of 15 December, Galwad-Y-Mor was operating in fishing grounds east of The Wash (Figure 1). The vessel was loaded with approximately 3 tonnes (t) of catch in the seawater-filled vivier tank1, 7t of bait, and approximately 7000 litres of fuel. At about 1110, the crew started to haul in a string of 100 crab pots from 30m water depth; the string’s main line was about 1.2 miles long. The skipper was seated in the wheelhouse pilot chair, five crew were on the main deck, and the night watchman was in bed in the crew cabin. After approximately 15 crab pots had been hauled on board, the crew signalled to the skipper that there was a lot of tension to the main line. The skipper increased the vessel’s engine speed in an attempt to free the potting string from the seabed. At 1122, there was an explosion external to Galwad-Y-Mor and three loud bangs were heard by the crew on the main deck. The vessel was thrown about. Propulsion and electrical power immediately failed. The main deck was deluged with seawater and one crew member’s personal flotation device (PFD) automatically inflated. The skipper had hit his head and was dazed; four of the crew were severely injured but all remained conscious. The wheelhouse equipment was seriously damaged, water flooded onto the main deck, and into the engine room, and the vessel settled low in the water (Figure 2). The skipper roused the night watchman and made a distress call with a handheld very high frequency (VHF) radio on channel 16. Unsure that the distress call had been received, he contacted fishing vessel Ingenuity’s skipper via mobile phone messaging service and requested that a distress message be relayed to the coastguard. Galwad-Y-Mor’s skipper launched the liferaft with the help of a crew member and ordered the crew to prepare to abandon ship. At 1148, the coastguard tasked a search and rescue helicopter to fly to the accident site. At 1158, a Royal National Lifeboat Institution (RNLI) all-weather lifeboat (ALB), based at Cromer, was also tasked to Galwad-Y-Mor by the coastguard. The captain of wind farm offshore support vessel Esvagt Njord, positioned approximately four miles to the south of the accident site, heard the VHF distress call and dispatched his vessel’s fast rescue boat (FRB) to help Galwad-Y-Mor’s crew. Galwad-Y-Mor’s skipper assisted three crew members to board the FRB once it had arrived, and then cut away the potting string from Galwad-Y-Mor’s hauler. After hearing reports of the severity of crew injuries, Esvagt Njord’s captain then dispatched his vessel’s fast transfer boat (FTB) to the accident site. Once it had arrived alongside, Galwad-Y-Mor’s skipper and the remaining crew members boarded the FTB and both boats returned to Esvagt Njord. By 1159, they were hoisted on board. Esvagt Njord’s crew and the team of wind farm contractors on board, which included a paramedic, administered first aid to the injured fishermen. 1 A vivier tank is used for storing live shellfish.

4 Figure 1: Location of the accident Reproduced from Admiralty Chart 0002 by permission of HMSO and the UK Hydrographic Office Accident location Fishing grounds Reproduced from Admiralty Chart 2182A by permission of HMSO and the UK Hydrographic Office

5 Figure 2: Galwad-Y-Mor settled low in the water Image courtesy of Andrew Oliver At 1350, two of Galwad-Y-Mor’s crew were airlifted from Esvagt Njord by the coastguard helicopter and taken to hospital. At 1430, four crew were taken ashore to Cromer in the ALB and then to hospital by ambulance. At 1550, the last injured crew member was taken to hospital by the coastguard helicopter. Galwad-Y-Mor’s owner arranged for the tug GPS Avenger to salvage Galwad-Y-Mor. At about 2130, Galwad-Y-Mor was taken under tow to Grimsby, and was lifted out of the water into a local ship repair facility the following morning. On 27 December 2020, Galwad-Y-Mor’s owner boarded Ingenuity and, on arrival at the accident location, recovered the remainder of the potting string. A distorted metal fragment was found in one of the crab pots; the MAIB sent the fragment and a section of rope from the crab pot string to explosive analysis specialists and metallurgists for forensic examination. 1.3 ENVIRONMENTAL CONDITIONS At the time of the accident, the wind was southerly force 3 to 4, visibility was good and the sea state was slight to moderate.

6 1.4 GALWAD-Y-MOR 1.4.1 Vessel description Galwad-Y-Mor (Figure 3) was designed by Macduff Ship Design Ltd and built in 2007 by MMS shipyard in Hull. The vessel had a 6mm steel-plated hull, set on frames 450mm apart, and a box keel (Figure 4). Galwad-Y-Mor had a shelter deck with a wheelhouse and deck crane. The main deck had aft accommodation, including a galley, crew mess and shower. At its forward end was a working area with a hydraulic hauling winch and catch handling table and a store that was also used as a working mess room. There was a ‘tween deck clearance of approximately 2.3m between the main deck and shelter deck. The lower deck consisted of a six berth cabin, separate skipper’s cabin, four fuel tanks and a steering compartment to aft. The engine room was positioned at midships and there was a vivier tank, bait hold and freshwater tank to forward. The engine room had a 265kW main engine that provided power for propulsion. There was also a 100kW main generator, harbour generator and main electrical switchboard fitted in the engine room. The wheelhouse was equipped with three VHF radios, one with Digital Selective Calling2 functionality for sending distress messages and two handheld radios. The navigation equipment consisted of two radars, a chart plotter, an echo sounder and two fishing computers. The wheelhouse floor to deckhead clearance was approximately 2.5m. Interior linings and wheelhouse equipment were fixed to wood battens bonded to the steel structure. 1.4.2 Crew Galwad-Y-Mor was crewed by a skipper, night watchman and five deckhands. The crew had worked together for several years and held the mandatory qualifications to work on a fishing vessel of this size. The crew had undergone safety familiarisation and periodic training on board Galwad-Y-Mor. Safety routines, such as man overboard, abandon ship and fire drills were carried out each month in accordance with the requirements of Merchant Shipping Notice (MSN) 1871 Amendment 1 (F) Code of Practice for the Safety of Small Fishing Vessels of less than 15m Length Overall. At the time of the accident, all crew on the main deck were wearing automatic inflation PFDs and the two crew working around the hauling table were wearing hard hats. The main deck watertight doors to the accommodation and engine room and fish hold hatch cover were all closed. This was standard practice when Galwad-Y-Mor was at sea. 2 Digital Selective Calling (DSC) is a method of quickly sending designated pre-programmed digital messages that automatically include a vessel’s identification and location.

7 Figure 3: Galwad-Y-Mor general arrangement plan Image courtesy of Macduff Ship Design

8 Figure 4: Galwad-Y-Mor hull and keel detail Hull steel plating 6mm thick Frame spacing 450mm Image courtesy of Macduff Ship Design

9 1.4.3 Vessel operation Galwad-Y-Mor was owned and operated by the Galwad-Y-Mor Shellfish Company that also owned Ingenuity, a second potting vessel. Galwad-Y-Mor typically caught crab and lobster in the North Sea. Both vessels had regularly caught crab in the accident’s vicinity and it was common for their pots and lines to catch on objects such as boulders on the seabed. The skipper navigated Galwad-Y-Mor during daytime passages and potting operations. The night watchman usually navigated the vessel to the next crabbing ground overnight. At the start of potting operations, the skipper manoeuvred the vessel towards a pot string marker buoy while steering by hand. The crew retrieved the marker buoy and used the hydraulic hauling winch, which had a 2t lifting capacity, to heave the string of pots onto the catch handling table. The skipper switched the steering to autopilot and used the throttle to manually control the vessel’s speed. Two deckhands worked either side of the catch handling table, emptying each pot in turn and dropping the crabs into the vivier tank. The other three deckhands rebaited the pots and stacked them on the main deck in preparation for the next shoot (Figure 5). Each string consisted of 60 to 100 crab pots. Each pot weighed approximately 30kg when full; the recovery, rebaiting and shooting of a string of pots took about an hour. Galwad-Y-Mor’s skipper recorded the track of each crab pot string in the fishing computer. It was typical to spend 7 days at sea, catch around 14t of crab and use about 1t of bait per day. 1.4.4 Vessel stability Galwad-Y-Mor‘s owner was supplied with a stability booklet on the vessel’s delivery and Galwad-Y-Mor had not been significantly altered since its build. To minimise free surface effect3 the vessel’s load conditions were based on a full vivier tank, which both the owner and skipper were aware of. The vessel’s intact stability passed the various MSN 1871 Amendment 1 (F) load condition requirements. Section 3.37 of the MSN required damage stability to be calculated for multihulled fishing vessels, which was not applicable to Galwad-Y-Mor. 1.5 CREW INJURIES AND VESSEL DAMAGE 1.5.1 Crew injuries Five of the seven crew experienced significant injuries in the accident, some of which were life-changing (Figure 6). Most of the injured crew required operations and extended stays in hospital, during which one crew member contracted COVID-19. 1.5.2 Vessel damage There was significant structural damage to Galwad-Y-Mor. Hull plating was pushed in by up to 40mm between frames and the starboard side hull plating was holed. The transverse bulkheads were buckled, and the port side bilge keel was twisted (Figure 7). 3 Free surface effect occurs when a ship’s tank or compartment is partially filled and any rolling motion causes the liquid to move and reduce ship stability.

10 Figure 5: Galwad-Y-Mor crab potting arrangement Shooting hatch Pots baited and stacked Normal lead of potting string Crabs into Vivier tank Hauling table Image courtesy of Macduff Ship Design

11 Figure 6: Position of crew and injuries sustained at time of the accident Skipper: 1.85m tall, 100kg. Seated in wheelhouse. 3 broken vertebrae, sternum and orbital bone. Knee injuries. Deckhand 1: 1.87m tall, 100kg. Fractured skull, pelvis, arm, ribs and toes. Ruptured bladder Deckhand 2: 1.87m tall, 89kg. Broken leg, damaged knees and elbow. Punctured lung Deckhand 3: 1.8m tall, 76kg. Head injuries: plate to cranium, loss of sight in one eye Deckhand 4: 1.76m tall, 97kg. Minor injuries to shoulder, ankles, heels, neck and back Deckhand 5: 1.83m tall, 120kg. Cuts and bruises Night watchman: 1.8m tall, 76kg. Prone in upper bunk. Broken leg, hip and shoulder Image courtesy of Macduff Ship Design

12 Figure 7: Galwad-Y-Mor structural damage Pushed in hull plating Port side bilge keel twisted Hull plating holed Buckled transverse bulkhead

13 The main engine, gearbox, generator, switchboard, hydraulic power pack and miscellaneous small pumps sheared from their rigid mountings. The main seawater inlet valve was sheared at its inlet flange and the engine room was flooded to the top of the compartment. The rudder actuator was displaced from the rudder stock (Figure 8). The wheelhouse equipment and linings were displaced from their mountings. The outfitting of the galley, crew mess, toilet and cabin was significantly damaged (Figure 9). There was a great deal of disruption to external deck areas with numerous instances of scattered equipment, damage to mast-mounted antennae, displaced light fittings, and broken pipe clamps. 1.6 EXAMINATION OF ROPE AND METAL FRAGMENT 1.6.1 Explosive trace examination The recovered metal fragment and crab pot rope were tested for traces of a range of explosives in line with United Kingdom Accreditation Service (UKAS)4 standard operating procedures and methods. Table 1 shows the test results: Explosive Quantity of explosive recovered from sample/ ng5 Metal Fragment Rope Dinitrotoluene (DNT) Indicated 0 Picrate Indicated 0 Research Department Explosive (RDX) 0.2 0 R-salt6

0 Trinitrotoluene (TNT) 1600 9 Table 1: Results of explosive trace examination Trinitrotoluene (TNT) was identified on both the metal fragment and the rope. TNT is used widely across the world in many explosive applications and was commonly used in First World War (WWI) and Second World War (WWII) ordnance. The quantity of TNT detected on the metal fragment was consistent with the item having been in direct contact with the explosive or an item heavily contaminated with it. The quantity of TNT detected on the rope was also consistent with it being in contact with the explosive. The small amounts of some of the other explosives found on the metal fragment were too low to provide the basis for conclusive analysis. 1.6.2 Metallurgical examination of metal fragment The metal fragment (Figure 10) was magnetic and measured 94mm in length, 129mm in width and had a 7.41mm average thickness. Visual inspection showed that the metal fragment had deformed into a curved shape. Rust was visible on 4 UKAS is the national accreditation body appointed by UK government to assess organisations that provide certification, testing and calibration services. 5 Nanogram one thousand millionth of a gram. 6 R-salt is a derivative of Research Department eXplosive.

14 Figure 8: Galwad-Y-Mor machinery damage Switchboard sheared from mountings Sheared main seawater inlet valve Displaced rudder actuator Engine sheared from rigid mountings

15 Figure 9: Galwad-Y-Mor interior damage Wheelhouse Crew toilet Crew mess

16 Figure 10: Steel fragment detail Image courtesy of the 1710 Naval Air Squadron Face A Fracture surface 1 Fracture surface 4 Fracture surface 2 Face B Fracture surface 3 Figure 11: SC250 high explosive bomb 'Bottling' section of 250kg HE bomb Images courtesy of the Defence Science and Technology Laboratory

17 both sides and observed as more progressed along the concave face (Face B), indicative of longer exposure to a corrosive environment or an accelerated reaction to corrosion. The metal fragment’s four fracture surfaces (Figure 10) were examined with scanning electron microscopy and showed features consistent with ductile microvoid coalescence (MVC)7. These features are indicative of ductile overload, which is the failure mode that occurs when a material is loaded to beyond its ultimate tensile strength. Optical profilometry showed machine tool marks on the metal fragment’s convex face (Face A) (Figure 10) consistent with a turning tool8 machining process. The metal fragment was examined by scanning electron semi-quantitative Energy Dispersive X-ray Analysis9, which established the metal’s elemental composition (Table 2). Elements found Weight as a % of the fragment Silicon 0.5-0.6 Vanadium 0.4-0.5 Phosphorus Not applicable Chromium 1.4-1.9 Manganese 0.6-0.8 Carbon 0.42 Sulphur 0.02 Nickel Not applicable Iron 96.597.8 Other elements

Table 2: Elemental composition of the metal fragment The metal fragment’s composition was consistent with a medium carbon steel that is obsolete and historic in nature. The metal fragment’s properties were compared with that of historic ordnance and it was found that:

The material thickness was greater than that commonly used in torpedoes and depth charges.

The material thickness was greater than that used in WWI and WWII British and German sea mines. Also, the elemental composition of the fragment was remarkably different from a WWII German mine casing, which tests found was made from iron with 0.5% manganese and no other alloying elements.

The material thickness and machining method was comparable with that used on the ‘bottling section’ of a general purpose (GP) high explosive (HE) air-dropped bomb (Figure 11). The fragment’s elemental composition was 7 MVC is a process where voids form and join in a material resulting in a failure. 8 Machining process used to manufacture round components. 9 This is a widely applied elemental microanalysis method capable of identifying and quantifying all elements in the periodic table except Hydrogen, Helium, and Lithium.

18 comparable with materials used in German SC10 250 GP air-dropped HE bombs, manufactured between 1940 and 1942. SC bombs were the most deployed air-dropped WWII German bomb and weighed 1800kg, 1000kg, 250kg or 50kg respectively. The SC250 bomb was 1173mm long, 385mm in diameter, and contained approximately 123kg of explosive that consisted of TNT and ammonium nitrate. Historic bombing records indicated that approximately 10% of HE bombs dropped during WWII did not detonate11. 1.7 ORDNANCE 1.7.1 Unexploded ordnance found at sea In a paper titled Low order deflagration for UXO disposal for the commercial sphere, submitted to a UK parliamentary sub-committee in September 2020, it was estimated that there are approximately 500,000 (100,000t) pieces of unexploded ordnance (UXO) in the seas around the UK. This UXO may originate from German or allied air-dropped bombs jettisoned to sea after aborted bombing missions, German air-dropped bombs used to target allied shipping, depth charges, torpedoes, sea mines. Since WWII, the Royal Navy (RN) also disposed of expired ordnance at sea in approved designated areas. UXO contains an explosive charge that can produce unstable degradation products over time. If an immersed UXO casing remains intact, the explosive and any degradation products are not washed away by seawater. When UXO is disturbed, the degradation products can detonate and cause a secondary detonation of the main explosive charge. An unstable fuse can also cause intact UXO to detonate if disturbed. At sea, WWI and WWII UXO is highly volatile and explosive ordnance disposal (EOD) specialists tend to dispose of it by controlled explosion rather than attempt to defuse. 1.7.2 Injuries to people on vessels subjected to subsurface exploding ordnance Injuries sustained due to a subsurface ordnance explosion are categorised as follows: ● Primary injuries are blast type injuries caused by proximity to an explosion and include shock wave damage to internal organs such as lungs and intestines. ● Secondary injuries are penetrative injuries caused by interaction of the body with fragments from the explosion, such as shrapnel. ● Tertiary injuries are blunt force trauma injuries caused by the body being rapidly thrown about by the blast. ● Quaternary blast injuries are thermal burn injuries due to exposure to the explosive blast. 10 Sprengbombe cylindrish German to English translation: high explosive cylindrical bomb. 11 The Imperial War Museum: https://www.bbc.com/future/article/20150922-these-nazi-bombs-are-more- dangerous-now-than-ever-before.

19 1.7.3 Effect to vessels subjected to subsurface exploding ordnance The damage experienced by a vessel from an underwater explosion is dependent on its proximity to the ordnance at the time of the explosion. In the case of a vessel striking a piece of ordnance such as a contact mine, the ensuing explosion causes direct damage to the hull in the form of a hole, and secondary damage to equipment from fragmentation of the mine casing and the vessel’s structure. Damage caused by contact with ordnance usually includes flooding and the sinking of the vessel. A more distant underwater explosion of a piece of ordnance creates a gas bubble that rises to the surface, collapses, and is filled with seawater moving at high velocity, which creates a water plume jettisoned into the air. If the collapsing gas bubble interacts with a vessel’s hull, the energy from the in-rush of seawater can hole the hull or even break it apart. Ordnance detonated on the seabed creates a shock wave that tends to exert hydraulic pressure to any nearby vessels on the surface, causing them to be thrown about. Damage may include pushing in of hull plating and displacement of equipment. In large explosions, nearby vessels can be thrown upwards out of the water, causing the unsupported hull to hog and then abruptly sag, which can result in a broken keel. It is common for people to hear more than one bang when a vessel encounters a single underwater explosion. This is due to reverberation of sound caused by the pulse bubble and shock wave. Blast effect modelling of a seabed ordnance explosion at a 30m water depth showed that the onset of surface vessel damage would require a minimum explosive charge of 121.5kg TNT equivalent12 and would result in a surface gas bubble of 2.5 to 3.0m in diameter. The Ministry of Defence (MOD) recommend that, to safely avoid damage when clearing UXO, non-military vessels should keep a distance of at least 530m during a controlled underwater explosion of ordnance this size. 1.7.4 Records of at sea unexploded ordnance The Receiver of Wreck (RoW) is a government body that deals with cases of voluntary salvage wreck material across the UK. The RoW keeps records of reported UXO findings and activities such as EOD clearance in UK waters. In the five years prior to this accident, the RoW listed 49 instances of ordnance found at sea, of which 31 related to UXO found in fishing nets. The rest occurred during dredging, diving, subsea surveying and mine hunting. The UK Hydrographic Office (UKHO) is the MOD agency responsible for providing hydrographic and marine geospatial data to mariners and maritime organisations across the world. The UKHO held some records of British and German WWII minefields, collated from RN reference books, historic admiralty charts and copies of original German records stored in the RN Historic Branch archive. The UKHO considered the currency of this information to be questionable due to inaccuracy of positional data at the time of mine laying, the effect of the passage of time on submerged mines and the lack of detail of the effectiveness of mine clearance activities. Known seabed ordnance dumps are marked on published admiralty 12 This is the method of quantifying the energy released in explosions.

20 charts. Because the location of other seabed UXO such as air-dropped bombs, depth charges and torpedoes are unknown, they are not marked on UKHO charts or other charting systems used by the fishing industry. Developers of Offshore Renewable Energy Installations (OREI) often consult with commercial organisations that hold digitalised records of WWII minefields when developing plans for new OREI. It is commonplace to carry out extensive seabed UXO surveys prior to installation of windfarms, subsea turbines and subsea power or telecom cables. If UXO is detected, then further surveys may be carried out to determine if they can be left in position, moved, or detonated. A UXO survey report is then submitted to the marine licensing authority responsible for the geographical OREI area, such as The Crown Estate. UXO survey information can be downloaded free of charge from The Crown Estate’s Marine Data Exchange website13. A pre-installation seabed survey of the Dudgeon Offshore Wind Farm site, which was 1.5nm to the east of the Galwad-Y-Mor accident, covered a 90 square nautical mile survey area (Figure 12). It identified numerous potential UXO targets, of which 244 were classed as high risk to the wind farm’s installation and operation. EOD specialists cleared several UXO, including German SC bombs, using controlled explosion methods and post detonation fragment collection (Figure 13). 1.7.5 Potential unexploded ordnance near the Galwad-Y-Mor accident location The Galwad-Y-Mor accident occurred beneath WWII flight paths used by both allied and German air forces, which jettisoned numerous bombs into the sea. The location was also close to coastal shipping routes and known British and German historic WWII minefields. Although the minefields were swept by the RN during WWII and in post-war years, whether this activity was completely effective is unknown. None of the RoW recorded UXO findings were within a 10nm radius of the Galwad-Y-Mor accident, and the nearest charted ordnance dump lay more than 100nm to the east. In February 2021, the RN carried out an exploratory hunting (EH) survey of a 1nm² area centred on a datum at the first reported position of the Galwad-Y-Mor accident, which was approximately 3.5nm north-west of the explosion. As a result of the EH survey, the RN concluded that this location was not historically used to dump munitions. 1.7.6 Unexploded ordnance guidance available to mariners The Maritime and Coastguard Agency’s Marine Guidance Note (MGN) 323 (M+F) Explosives Picked Up At Sea, provided guidance to owners, masters, skippers and crew of trawlers, fishing vessels, dredgers and other vessels involved in seabed operations. The MGN advised that explosives found at sea can be volatile and should not be landed on deck or tampered with; suspected ordnance should be reported to the coastguard. It also stated that fishing vessels suspecting they have an item of ordnance in their trawls should reduce vessel speed, transit to a safe area, and lower their trawl to the seabed in a depth of less than 30m if possible, while notifying the coastguard of their position. 13 www.marinedataexchange.co.uk.

21 Figure 12: Possible UXO in Dudgeon Wind Farm survey area Reproduced from Admiralty Chart 2182A by permission of HMSO and the UK Hydrographic Office Position of accident Dudgeon Wind Farm survey area UXO survey ST13504 Possible UXO targets Low risk Moderate risk High risk Dudgeon wind turbines Dudgeon cables Image courtesy of Equinor UK Ltd

22 Figure 13: SC250 high explosive bomb found and detonated in Dudgeon Wind Farm survey area 250kg HE bomb found on seabed 250kg HE bomb fragment post clearance detonation Image courtesy of Equinor UK Ltd

23 The Crown Estate’s Guidance Note Dealing with munitions in marine sediments, provided guidance for dealing with potential UXO recovered from the seabed by vessels dredging for aggregates, undertaking navigational dredging, and discharging cargos. This document detailed procedures for identifying potential UXO and the notification of retained EOD specialists, police or coastguard dependent on the vessel’s location. 1.8 PREVIOUS/SIMILAR ACCIDENTS In December 2020, a Marine Scotland research vessel operating in the Firth of Clyde landed a 350kg explosive charge German WWII mine on board. The crew alerted the coastguard and were evacuated from the vessel by the RNLI. An RN EOD team lowered the mine to the seabed and carried out a controlled explosion. In February 2017, a dredger recovered what was believed to be a German SC250 bomb in Portsmouth Harbour, England. The bomb was towed out to sea and blown up by an RN EOD team. In November 2016, fishing vessel Star of Annan (OB50) picked up what was thought to be a chemical bomb while dredging for scallops in the Irish Sea. The bomb exploded and caused a small fire on the deck of the fishing vessel and chemical injuries to crew members.

24 SECTION 2 – ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 THE ACCIDENT 2.2.1 The ordnance Although the external explosion to Galwad-Y-Mor occurred in the vicinity of historic WWII offensive and defensive minefields, the material thickness and elemental composition of the recovered metal fragment shows that it was unlikely it originated from a German or British sea mine. The elemental composition, thickness, machine tool marks and the exposure to TNT of the recovered metal fragment suggest with a high degree of certainty that it originated from an unexploded WWII German HE SC250 air-dropped bomb. The metal fragment’s deformation and ductile overload indicates that it was exposed to an explosive event. The lack of overall corrosion to the recovered fragment suggests that the bomb was buried below the seabed for several years. The light corrosion to one surface of the metal fragment indicates that the bomb was intact and not flooded, therefore the explosive charge had not been washed away by seawater. The lack of a major hole to Galwad-Y-Mor’s hull indicates that the vessel did not directly contact the bomb. It is likely that direct exposure to the full detonation of ordnance, containing 123kg explosive charge, would have blown the vessel apart. Galwad-Y-Mor’s damage was commensurate with the explosive charge detonating below the vessel at a water depth of 30m, therefore it is likely that the bomb fully detonated on the seabed. 2.2.2 The explosion The string of pots the crew were attempting to recover at the time of the explosion most likely became fouled on an SC250 bomb. It is possible that the bomb was dragged along the seabed and caught on an obstacle, such as a boulder, causing the potting string to tighten on the hauler. As Galwad-Y-Mor’s skipper increased vessel power, the disturbance of the bomb caused either an unstable fuse to activate or explosive charge degradation products to detonate and this resulted in the bomb exploding on the seabed. 2.2.3 Crew injuries Most of the injuries experienced by the crew were of a blunt force trauma nature and did not include penetrating wounds, burns or major internal damage (Figure 6). These were consistent with tertiary injuries most likely caused by the acceleration forces the crew experienced when Galwad-Y-Mor was thrown about. The head injuries to deckhands 1 and 3, who were not wearing safety helmets, suggests they struck the deckhead at the time of the accident. The skipper was seated when the

25 bomb exploded and struck his head on the wheelhouse deckhead, which caused injuries to his eye socket and face. His back and knee injuries were probably incurred when he landed back down on the deck. Although the physical injuries were significant to five of the seven crew, they were fortunate not to be killed. The crew member working on the starboard side of the main deck could have been ejected overboard. Although wearing a working PFD, his survivability might have been threatened by the impeded ability of the injured crew to retrieve him. 2.2.4 Vessel damage The potting string quickly tightening at the hauler suggested that the bomb was directly below the vessel on the seabed prior to the explosion. The hull plating deformation, buckled bulkheads, and machinery and equipment displacement were commensurate with Galwad-Y-Mor interacting with the shock wave created by the explosion. It is likely that the single hole to the hull plating was caused by the plating bending in way of the adjacent structure and fracturing as a result. It is therefore concluded that Galwad-Y-Mor was close to the explosion and well below the MOD’s recommended 530m safe stand-off distance for non-military vessels to be unaffected by a seabed explosion of this size. Nevertheless, Galwad-Y-Mor’s thick shell plating presented good resistance to the shock wave generated by the explosion, and the heavy section box keel was able to resist bending forces set up by the pulse bubble underneath the vessel. The main seawater inlet valve was sheared either by the effect of the seawater system being pressured by the shock wave, or by the displacement of the machinery it was connected to. The sheared seawater inlet valve allowed water to enter and flood the engine room. Although the transverse bulkheads had buckled, they remained intact and prevented the seawater transferring to other compartments. Galwad-Y-Mor’s machinery mountings and internal fit out were damaged because they were not designed to withstand the forces created by the action of the vessel being thrown about. 2.2.5 Anticipation of the danger from unexploded ordnance Both Galwad-Y-Mor and Ingenuity had fished for crabs in the vicinity of the accident for many years and had not experienced any previous UXO interactions. The seabed positions of air-dropped UXO are unmarked on UKHO charts and are largely unknown, apart from sporadic at-sea findings recorded by the RoW and offshore installation licensing authorities. Consequently, the crew could not have known of the impending danger from the fouling of a bomb and its ensuing explosion. If the bomb had not exploded and been hauled on board, the information contained within MGN 323 (M+F) would have helped the crew deal with the bomb and notify the appropriate authority.

26 2.2.6 UXO location information available to fishermen Apart from known ammunition dumping grounds that are marked on navigation charts, there is little information readily available to fishermen concerning the presence of UXO on the seabed. No single authority holds data concerning the location of UXO found on the seabed. Information is held by the RoW, the UKHO, The Crown Estate, and commercial owners or installers of offshore infrastructure such as wind farms, pipelines and cables. It is known that some areas around the UK were subjected to more wartime activity than others. For example, the Strait of Dover, the River Thames and estuary, and the areas around naval ports. While it may be possible to infer from this that there is a greater potential risk of encountering wartime UXO, many of these areas have been fished since 1945 without apparent incident. Although the requirement to report UXO encounters to the RoW and Coastguard exists within MGN 323, anecdotal evidence suggests that there is significant under-reporting. 2.3 POST-ACCIDENT 2.3.1 Emergency response The crew’s actions after the explosion were both prompt and effective; five of the seven crew were significantly injured and yet they were able to send a distress message, launch a liferaft, and evacuate to Esvagt Njord’s rescue boats. It is likely that Galwad-Y-Mor crew’s formal training, emergency drills, and combined length of onboard service prepared them to deal with this unforeseen emergency scenario and enabled them to take swift action that, undoubtedly, saved lives. The emergency response and rescue that Esvagt Njord provided was also prompt and effective. The onboard triage of the casualties was aided by the presence of a trained paramedic and increased the survivability of Galwad-Y-Mor’s injured crew members. 2.3.2 Vessel survivability Galwad-Y-Mor remained afloat despite having one flooded compartment that caused the vessel to settle low in the water close to its maximum draught of 3.6m at the stern. The flooding of the engine room increased the vessel’s draught and reduced the freeboard to such an extent that the main deck was at sea level, which allowed seawater onto the deck. As all main deck watertight doors and hatches were closed, and the transverse main deck bulkheads remained watertight, the seawater was prevented from entering other compartments, and the stability of the vessel was not compromised further. Despite severe damage and flooding, Galwad-Y-Mor had good reserve buoyancy. However, it was fortunate that the weather conditions were favourable, which reduced the risk of the vessel capsizing while under tow to a place of safety.

27 SECTION 3 – CONCLUSIONS 3.1 OTHER SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT

Galwad-Y-Mor was severely damaged and its crew injured because a German high explosive bomb, dropped during WWII, exploded on the seabed beneath it. It is highly likely that the bomb was a WWII German HE SC250 air-dropped bomb. [2.2.1]

Even after many years of submersion, UXO can be highly volatile, become unstable, and explode unexpectedly. [2.2.1]

Galwad-Y Mor’s potting gear disturbed the bomb and it subsequently exploded. [2.2.2]

The bomb’s explosion released a gas bubble and pressure wave that threw Galwad- Y-Mor about on the surface, causing significant injuries to crew and damage to the vessel. The effect of the explosion could have resulted in multiple fatalities. [2.2.1, 2.2.3]

Galwad-Y-Mor’s machinery, on board equipment and internal outfitting were considerably damaged, but its hull was strong enough to withstand the forces generated by a nearby seabed explosion, which went far beyond the hull’s design requirement. [2.2.4]

There were no records or information available to fishermen that could have alerted Galwad-Y-Mor’s skipper to the risk of snagging unexploded ordnance in the area where he had laid his pots. [2.2.6]

Although Galwad-Y-Mor’s crew could not have anticipated the fouling of a bomb in the potting string, their training, experience, length of service together and emergency preparedness improved their chances of survival. [2.3.1]

Although Galwad-Y-Mor flooded during the accident, the vessel’s reserve buoyancy and calm weather conditions meant the vessel did not capsize. [2.3.2]

28 SECTION 4 – ACTION TAKEN 4.1 MAIB ACTIONS The MAIB has:

● On 21 December 2020, published a preliminary assessment14 into the Galwad-Y- Mor accident. SECTION 5 – RECOMMENDATIONS In view of the actions already taken, no recommendations have been made. 14 https://www.gov.uk/government/publications/preliminary-assessment-galwad-y-mor/explosion-resulting-in- damage-to-fishing-vessel-galwad-y-mor.

Marine Accident Report

FD06068 ｜ 2021_英国_carbon-monoxide-poisoning-on-motor-cruiser-diversion-with-loss-of-2-lives__2021-03-Diversi.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2021 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

M A R I N E A C C I D E N T I N V E S T I G AT I O N B R A N C H ACCIDENT REPORT ACCIDENT REPORT VERY SERIOUS MARINE CASUALTY REPORT NO 4/2021 APRIL 2021 Carbon monoxide poisoning on board motor cruiser Diversion, resulting in two fatalities at the Museum Gardens quay, River Ouse, York on 4 December 2019 1 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an such investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2021 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned. All reports can be found on our website: www.gov.uk/maib For all enquiries: Email: maib@dft.gov.uk Tel: 023 8039 5500 SUMMARY At about 2000 on 4 December 2019, the bodies of two men were discovered in the forward cabin of the 9.14m motor cruiser Diversion, which was moored to a quay on the River Ouse (Figure 1) in the centre of York, England. The bodies were those of the vessel owner and his friend who had spent the previous evening in the city centre socialising with former work colleagues and were spending the night on board. Both men had died as a result of carbon monoxide poisoning. The MAIB investigation concluded that: • The carbon monoxide had leaked into the cabin from the boat’s diesel-fuelled cabin heater exhaust. • The cabin heater’s exhaust silencer was not designed for marine use. Its connection to the exhaust pipe system was not gas tight, the installation had not been checked by a professional heater installer, and it had not been serviced. • The cabin ventilation system did not meet the requirements of the Boat Safety Scheme and this might have increased the rate at which the carbon monoxide accumulated in the boat’s cabin space. • The owner and his friend were not alerted to the danger because a carbon monoxide alarm had not been fitted. Figure 1: Diversion moored in York

2 As a result of the investigation’s initial findings, the MAIB issued safety bulletin number SB2/2020, highlighting the importance of having diesel-fuelled cabin heater installations checked and regularly serviced, and also the requirement to have a carbon monoxide alarm fitted. The safety bulletin has been referenced in Boat Safety Scheme publications and has been featured in the 2020 Carbon Monoxide awareness week. FACTUAL INFORMATION Narrative At about 1200 on 3 December 2019, the owner and two of his friends made their way onto the motor cruiser Diversion at its mooring in York Marina. Once on board, they readied the boat for a 4 mile trip upriver to York city centre, where they had arranged to attend a pre-Christmas social gathering with former work colleagues. They left the marina at about 1230 for York, where the owner and one of his friends planned to sleep on board Diversion overnight. During the trip the owner used the stove in the main cabin to make coffee while one of his friends steered the boat. It was a cold, clear day so the owner also turned on the boat’s diesel-fuelled cabin heater. At about 1330, they arrived in the city centre and moored Diversion on the northern bank of the river, adjacent to Museum Gardens. At about 1400, the three men went ashore, and spent the afternoon drinking and socialising with their former colleagues in several city centre pubs. One of the owner’s two friends had to return home in the early evening and left them at about 1930 to catch a train. The owner and his remaining friend had a meal with some of the group then returned to Diversion at about 2200, and prepared to go to bed. At about midday on the following day, the families of the owner and his friend became concerned that they had not heard from either man. They made several telephone calls to try to locate them, but without success, and later in the afternoon they reported their concerns to the police. A police officer was dispatched to the quayside and confirmed that the boat was still moored there, and that its accommodation was in darkness and no one appeared to be on board. The officer was then instructed to continue on patrol. With no further contact from the two men, the families contacted the police once again to express their increasing concern. At about 2000, another police officer boarded Diversion and found the main cabin door unlocked, with all lights switched off. The police officer entered the boat and found the bodies of the owner and his friend in the forward cabin bunk space; one was lying on the floor, the other was seated on the forward cabin entrance steps, and it was apparent that they had prepared themselves to go to bed. The police officer immediately called for backup and, at 2045, the Fire and Rescue Service arrived on the boat. They tested the atmosphere in Diversion’s accommodation and declared it safe for entry. Paramedics entered the boat and at 2115 the owner and his friend were declared deceased. Diversion was towed back to York Marina on 7 December 2019, where it was secured by the police. On 9 December post-mortems were carried out, which found that the primary cause of death for both men was carbon monoxide (CO) poisoning; both also had caffeine and alcohol in their system. The accident was not reported to the MAIB until June 2020; the investigation started shortly afterwards. The casualties Diversion’s owner, Ronnie Holmes, was 60-years-old and had worked as a power station engineer for over 30 years. He retired in 2015 and spent his time carrying out home improvements, fixing motorcycles and boating. Before retirement he had enjoyed scuba diving with friends.

3 Stephen Jewitt had retired around the same time as Ronnie, was 63-years-old and had worked as an engineer at the power station, where he met and became friends with Ronnie. Stephen also owned a boat which he had used for diving trips with Ronnie. Vessel description Diversion was a 9.14m steel hulled boat (Figure 2) that had been built by its owner based on plans purchased from boat designer Bruce Roberts. The owner had started the construction of Diversion in 1990 and built it over a period of 11 years on land behind his home. The boat was launched at York Marina in 2001, where it was based for the next 19 years. The owner used Diversion predominantly on the River Ouse, and frequently entertained family and friends on board. During the build, the owner had made several alterations to the design, which included the installation of a second engine and changes to the cabin arrangements. The forward cabin was fitted with twin bunk beds and was wood lined, but it had no thermal insulation to the hull. The steering wheel, engine throttles and engine control console were located on the starboard side of the main cabin, and there was a 4kW1 twin burner methylated spirit stove on the port side. In 2018, the owner fitted a second-hand diesel- fuelled cabin heater that he had purchased from an auction website. The owner had also purchased a portable petrol generator and coupled it up to a 24V battery charger in the engine space. Two spare back-up batteries had been placed in the engine bay by the owner before they left York Marina. Diversion was licensed by the Canal and River Trust to operate on the River Ouse and was examined under the Boat Safety Scheme (BSS) as a mandatory requirement of this licence. 1 kW – Kilowatt Figure 2: Diversion general arrangement Toilet/washroom Main cabin Forward cabin Steering console Engine bay

4 Boat Safety Scheme The BSS is a public safety initiative, jointly owned by the Canal and River Trust and the Environment Agency. The purpose of the BSS is to help minimise the risk of boat fires, explosions, or pollution from harming third party users of the waterways. Approximately 60,000 BSS boat examinations are carried out by BSS accredited examiners every 4 years. Their checks are visual and do not involve disassembly of equipment. The BSS states on its website that its examinations are not a full condition survey and that owners should ensure boats and appliances are competently installed and serviced according to manufacturer’s instructions. On 1 April 2019, BSS had amended its requirements to include a working CO alarm2. The BSS extensively publicised this requirement on its website and in boating publications but found that about 10% of the boats presented for examination in the year April 2019 to April 2020 did not have a CO alarm fitted. Diversion did not have a CO alarm fitted at the time of the accident or when it was last presented for a BSS examination in February 2019. The owner was not present at the BSS examination and was not advised of the imminent requirement to install a CO monitor in the examination report. Documentary evidence indicates that a CO alarm had been fitted previously, but the reasons for its subsequent removal prior to the survey in February 2019 are unknown. On board investigation and tests MAIB did not become aware of the accident until June 2020, consequently testing of equipment onboard was carried out under warm summer conditions. It is possible, therefore, that the performance of the cabin ventilation during the wintery conditions at the time of the accident differed from those found during testing. Four potential sources of CO were found on board Diversion when the boat was inspected. These were the main engines, portable generator, cabin stove and cabin heater; all were subject to onboard examination and testing. The boat’s engines and portable generator were not running when the police first boarded the boat. The evidence indicated that they were not running at the time of the accident. When they were tested, no exhaust gas leaks were found, and no CO was measured within the boat’s accommodation. The methylated spirit fuelled stove was found turned off and both burner reservoirs contained fuel. When the stove was lit, the CO emissions measured in the main cabin initially exceeded 20ppm3 but dropped to approximately 10ppm once the burners were warm. When the emergency services arrived on board the boat, the diesel-fuelled heater was found to be powered up, but it had switched itself into the standby mode and was not operating. When tested on board, the heater was observed to emit approximately 20ppm of CO from its external exhaust, and CO was detected in the cabin spaces after the heater had been run for extended periods. CO levels in excess of 50ppm were measured in the main cabin more than 16 hours after the heater had been shut down. The cabin heater Diversion’s cabin heater was an Eberspächer 1.8kW D1LC air heater. It was installed in a cabinet in the main cabin beneath the steering console and was operated from a panel mounted on the top of the console. Circulating air was drawn from inside the cabinet, warmed by the heater, and discharged to the boat’s main cabin via a nozzle mounted at deck level next to the steering position (Figure 3). The cabin temperature was set by means of an adjustable control knob, and the heater started and stopped thermostatically. The heater also drew its combustion air from inside the cabinet, and a pump supplied diesel fuel from the boat’s port side fuel tank. The fuel/air mixture was burned in the heater, 2 alarm meeting safety standard EN 50291-2:2010, 3 ppm – parts per million

5 Figure 3: Installation of Diversion’s cabin heater Heater circulating air outlet Heater under console

6 and the exhaust gases containing CO and carbon dioxide (CO2) were discharged outside via a silencer connected to a through-hull fitting on the starboard side. The exhaust system was fully insulated with heat proof fibreglass and covered by foil tape. The manufacturer’s installation instructions, available via the internet, provided guidance regarding the requirements of the heater’s fuel system, exhaust, combustion air, electrical connection and circulating air. It is unknown if the owner read the instructions before he installed the heater and there was no record to indicate the installation was checked by an approved Eberspächer agent, or of the heater being serviced since it was fitted. The heater’s exhaust lagging was removed to enable closer inspection of the system, which comprised of a Webasto exhaust silencer, connected to the heater and boat’s side via flexible pipes. The Eberspächer heater exhaust port had a 24mm outside diameter stub pipe, but the Webasto silencer inlet had a 22mm outside diameter stub pipe. This meant that the two sizes were incompatible, and the flexible exhaust that connected the heater to the silencer was loose, and leaked exhaust gas. In addition, the silencer had spot welded seams and a condensate drain hole, which also leaked exhaust gas when the heater ran (Figure 4). Figure 4: Gas leak path from condensate drain and spot welded seam The most well-known brands of diesel-fuelled heaters are Webasto and Eberspächer. These are commonly installed into lorries, earth moving machinery, caravans and boats. Diversion’s Webasto silencer was not designed to be fitted in marine installations because it was not gas tight and it was incompatible with the exhaust piping. As a result, the leaking exhaust gas was drawn into the heater’s circulating air inlet and then blown into the main cabin (Figure 5). The heater would often be left running while the owners were ashore when using Diversion on cold days. On one occasion in October 2019, the owner and his wife experienced nausea and sickness after sleeping on board with the cabin heater running. Exhaust silencer as found Exhaust silencer removed Exhaust gas leakage path from loose inlet pipe connection Gas leak path from condensate drain and riveted seam

7 Figure 5: Leakage path of exhaust gas through cabin heater Examination and testing of the cabin heater The heater and associated equipment were removed from the boat and taken to an Eberspächer service agent for extensive examination and testing. The heater’s service panel cover was found to have an unbroken seal across it, confirming that it had not been serviced since installation. The heater’s internal glow plug was in good condition and there were no soot deposits found. The heater was found to start and stop thermostatically as per the manufacturer’s design, and the low supply voltage cut-out switched the heater to standby mode at 21V. Exhaust gas readings were close to design specifications with no trace of CO found at the circulating air outlet. The heater was then retested with its exhaust silencer positioned adjacent to the circulating air intake, comparable to the ‘as found’ arrangement on board Diversion. Leaking exhaust gas from the exhaust silencer’s seams, condensate drain, and loose inlet connection was sucked into the heater’s circulating air inlet and approximately 8ppm CO was then measured at the circulating air outlet. It was noted that an onboard heater service would include a check of the glow plug, exhaust gas analysis, visual assessment of the installation while the heater was running, and a test of the circulating air for any trace of combustion gas such as CO2 or CO. Air/CO mix blown into cabin Exhaust leaks ingested into heater circulation air inlet Air/CO mix ingested into heater combustion inlet pipe Air/CO from cabin returns to heater circulation air inlet

8 Cabin ventilation Boat cabins are often fitted with ventilation to provide fresh air for boat users and combustion air for fossil fuel burning cookers, water heaters, wood burners and diesel-fuelled air heaters. Insufficient air supply leads to incomplete combustion within such appliances and a discharge of unburnt fuel and higher than normal levels of CO from their exhaust systems. Diversion’s upper-level ventilation consisted of four mushroom type vents mounted in the roof of the accommodation. The two forward vents provided ventilation to the forward cabin space. These vents were permanently open, had no closing mechanism, and were found to be clear of obstruction. The two aft vents were mounted in the roof of the main cabin and had dampers fitted into them that were found in the closed position. Low level ventilation was provided via a louvre type grille mounted in the bottom of the cabin door. Holes had been drilled through the door behind the grille, and approximately half of these were found to be blocked with dust. This reduced the effective capability of the ventilator grille (Figure 6). The BSS used a standard formula4 for calculating minimum ventilation requirements when they examined Diversion. This formula used values based on the thermal output rating of fossil fuel burning appliances installed, and the number of persons for which the accommodation space was designed. There were some errors found in the application of the formula during the 2019 BSS examination of Diversion: an incorrect low thermal rating of the stove was used, and the cabin heater was not included. Additionally, there was a miscalculation of the available ventilation on board because the closable mushroom vents were included, in error. 4 derived from BS 5482—3 Annex G (Code of Practice for the installation of solid fuel heating and cooking appliances – Part 3: Installations in boats, yachts and other boats), Figure 6: Wheelhouse dour louvre vent Holes blocked Holes behind louvre grille

9 As part of the investigation, the BSS formula was used to re-assess Diversion’s ventilation requirement, and to re-calculate the actual available ventilation on board. It was found that the available ventilation was approximately 60% of the minimum required (Table 1). Table 1: Comparison of ventilation requirement and ventilation availability BSS calculated minimum ventilation requirement Feb 2019 BSS calculated available ventilation Feb 2019 MAIB calculated minimum ventilation requirement MAIB calculated available ventilation 9200 mm 18469 mm2 12192 mm2 7304 mm2

Carbon monoxide poisoning CO is a highly poisonous gas that weighs nearly the same as air and cannot be seen, smelled, tasted or felt by humans. CO readily replaces oxygen in the human bloodstream by forming carboxyhaemoglobin, which prevents oxygen from being supplied to the heart, brain and other vital organs. The blood carboxyhaemoglobin saturation level found during the post-mortem examinations of Diversion’s owner and his friend were 58.7% and 61.8% respectively; blood carboxyhaemoglobin saturation more than 50% is lethal to humans. Common symptoms of low exposure to CO poisoning include headache, nausea, stomach pain, difficulty breathing, or feeling tired or confused. In greater concentrations a person will feel drowsy and lethargic and may have difficulty moving around. The symptoms of carbon monoxide exposure continue to intensify as the concentration or time frame increases. An exposure at 800ppm can kill the person within a few hours, at 1500ppm it can cause death within an hour, at 3000ppm within half an hour and at 6000ppm death can occur within 15 minutes. Furthermore, exposure at 12,000ppm can result in instant death for the person. Similar accidents In November 2016, the owner of pleasure boat Vasquez5 was overcome and died as a result of CO poisoning, which had leaked from the petrol engine’s exhaust system. The exhaust system had been extensively modified and the engine had not been regularly serviced. No CO alarm was fitted on board. In June 2016, two fatalities resulted from CO poisoning on the motor cruiser Love for Lydia6 . The petrol- driven inboard engine had been left running while alongside, probably to charge batteries. The exhaust from the engine filled the cockpit canopy and spread into the cabin. There was no CO alarm fitted on board and the occupants were overcome by the fumes. The MAIB made recommendations to the Maritime and Coastguard Agency regarding the co-ordination of raising CO awareness of the leisure boating community; to British Marine to seek clarification on CO requirements within the recreational craft directive; and to the BSS regarding the compulsory installation of CO alarms in craft participating in the BSS scheme. In April 2014, two recreational boat users died of CO poisoning while sleeping on board the motor cruiser Arniston7 . The owner had fitted a portable generator in the boat’s engine compartment and rigged an exhaust system to discharge combustion gases overboard. The exhaust system failed, and fumes filled the motor cruiser’s cabin. The occupants had not been alerted to the presence of CO because the alarms fitted on board did not work. In January 2014, two fishermen died from CO poisoning on board the scallop dredger Eshcol8 . A butane fuelled cooker grill, which leaked CO, had been left on to warm the boat’s cabin as the crew slept. There was no CO alarm fitted and it is believed the cooker had never been serviced. 5 MAIB report 18/2017 6 MAIB report 9/2017 7 MAIB report 2/2015 8 MAIB report 14/2014

10 ANALYSIS Overview The two men on board Diversion died from CO poisoning due to a build-up of exhaust gas within the cabin spaces. The gas leaked from the cabin’s air heater exhaust silencer and a loose exhaust pipe connection. The nausea experienced by Diversion’s owner and his wife 2 months earlier was probably caused by CO poisoning from the cabin heater exhaust leaks. The accumulation of carbon monoxide Because it was a cold day and Diversion’s hull was uninsulated, the boat’s cabin heater was turned on shortly after leaving York Marina. It is likely that the heater was left running when the friends went ashore so that the accommodation would be warm on their return later that evening. When the heater was running, small amounts of the leaking exhaust gas were drawn into the heater’s circulating air intake and then discharged into the main cabin. Over a period of hours, the air in the main and forward cabins became increasingly polluted with CO and CO2. As the heater continued to operate, its combustion air drawn from inside the boat also became increasingly saturated with exhaust gases, and as a result the quality of combustion within the heater reduced. This added to the increased amount of CO in the exhaust gas and further polluted the air inside the boat. The evidence of caffeine found during the post-mortem examinations of both men suggests it is possible that when they returned to Diversion and went inside, they had a coffee before going to bed. Once inside the boat, they would have started to breathe in the CO enriched atmosphere, and it is likely that the effects of alcohol consumption made them unaware of the early signs of CO poisoning. As the men prepared to go to sleep in the forward cabin it was apparent that they were both overcome by CO, rendered quickly unconscious, and later died. The cabin heater would have continued to operate until Diversion’s battery voltage dropped to approximately 21V. The heater then automatically shut down to standby mode, which stopped additional build-up of CO inside the boat. The cabin heater The undamaged heater glow plug indicated that the heater had not been used frequently. The tests showed that the heater unit had no internal exhaust gas leaks and was in good working order. It is not known whether the Webasto exhaust silencer was an original fit to a previous heater, was supplied with the replacement Eberspächer heater unit, or was purchased separately. However, the silencer was not approved for marine use and was designed to be fitted externally. The gas leaks from its body, the condensate drain hole, and from the loose pipe connection, were obscured by the exhaust insulation and were probably not noticed by the owner when he installed the heater. An Eberspächer service agent appraisal of the heater installation, which is recommended by BSS, would probably have highlighted the issues with both of these items. Also, a service agent’s circulating air test from the running heater would have shown the presence of CO, and both of these areas of concern would have been pointed out to Diversion’s owner for rectification. Cabin ventilation Without removing the louvre grille from the door, it was impossible to tell that the drilled holes behind it were mostly blocked. While this did not cause a problem when the door was opened frequently, allowing fresh air into the main cabin, when it was closed for long periods the air flow through the boat was impeded.

11 Because the calculation used at the time of the BSS examination of the boat was not accurate, the correct level of ventilation required was mis-identified, and the levels of available ventilation were overstated. The situation was made worse, because the ventilation holes in the main cabin door were partially blocked. Onboard testing in summer conditions demonstrated that CO tended to remain in the boat’s atmosphere for extended periods. The impeded air flow through the cabin vents would have increased the likelihood and rate that CO accumulated within the cabin space. Carbon monoxide alarms Because a BSS examination is required to be carried out only every 4 years, it is feasible that some boat owners may not comply with the CO alarm installation requirement that came into force in April 2019 until their boat is next examined. This could be as late as 2023; which was the case for Diversion. It is surprising that the imminent requirement to have a CO monitor installed to Diversion was not highlighted during the last BSS examination. Taking into account the statistical 10% noncompliance rate at the time of being presented for examination, this could mean that more than 4000 BSS examined boats may not currently have a CO alarm fitted, representing a potential CO poisoning risk to the owners. On board Diversion, because there was no CO alarm fitted, neither the owner nor his friend were alerted to the presence of dangerous levels of CO within the cabin. Since CO in the atmosphere is undetectable by humans until the effects of poisoning start to become apparent, the installation of an alarm may be the final defence available to protect individuals. When a CO monitor goes into alarm, it should never be ignored, or considered to be defective. CONCLUSIONS • Diversion’s owner and his friend died after being poisoned by CO that had leaked from the boat’s cabin heater exhaust system and accumulated in the accommodation spaces. • Diversion’s owner had fitted the cabin heater using some non-marine components which allowed the exhaust to leak. • The CO leaks had not been identified because the diesel-fuelled heater installation had not been checked by an approved installation engineer, nor had it been serviced. • CO accumulated in the forward cabin to a lethal level and the cabin ventilation system, which was partially restricted, did not allow the CO to clear. • The casualties were not alerted to the danger because a CO alarm had not been fitted. ACTIONS TAKEN MAIB actions The MAIB issued Safety Bulletin SB2/20209 on 15 October 2020. Actions taken by other organisations The BSS has extensively published references to Safety Bulletin SB2/2020 and it has been featured in 2020 CO awareness week, which took place from 23 to 29 November. RECOMMENDATION As a result of the actions taken, no recommendations have been made. 9 MAIB safety bulletin SB2/2020

12 VESSEL PARTICULARS Boat’s name Diversion Type Motor cruiser Registered owner Privately owned Year of build 1990-2001 Construction Steel Length overall 9.14m Gross tonnage 7 Tonnes (approximately) VOYAGE PARTICULAS Port of departure York Marina, Naburn Port of arrival York City Type of voyage Pleasure MARINE CASUALTY INFORMATION Date and time Overnight 3 to 4 December 2019 Type of marine casualty or incident Very Serious Marine Casualty Location of incident Moored at Museum Gardens quay, York Place on board Forward cabin Injuries/fatalities 2 fatalities Damage/environmental impact None External environment 2ºC Wind southerly 5mph Persons on board 2

FD06069 ｜ 2021_英国_cargo-tank-explosion-and-fire-on-chemical-tanker-stolt-groenland__2021-09-StoltGroenland-R.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2021 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

ACCIDENT REPORT ACCIDENT REPORT INVESTIGATED BY THE MAIB ON BEHALF OF THE CAYMAN ISLANDS GOVERNMENT Red Ensign Group logo Cayman Maritime logo Report on the investigation of the cargo tank explosion and fire on board the chemical tanker Stolt Groenland Ulsan, Republic of Korea 28 September 2019 SERIOUS MARINE CASUALTY REPORT NO 9/2021 JULY 2021

This investigation was carried out by the UK Marine Accident Investigation Branch (MAIB) on behalf of the Cayman Islands Government in accordance with the Memorandum of Understanding between the MAIB and the Red Ensign Group Category 1 registries of Isle of Man, Cayman Islands, Bermuda and Gibraltar. Extract from The Cayman Islands Merchant Shipping (Marine Casualty Reporting and Investigation) Regulations, 2018 – Regulation 4: “The sole objective of a marine safety investigation into an accident under these Regulations shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It is not the purpose of a marine safety investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This report is not written with litigation in mind and, pursuant to Regulation 22 of The Cayman Islands Merchant Shipping (Marine Casualty Reporting and Investigation) Regulations, 2018, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes, is to attribute or apportion liability or blame. All MAIB publications can be found on our website: www.gov.uk/maib Front cover image courtesy of Lappino (shipspotting.com) For all enquiries: Marine Accident Investigation Branch First Floor, Spring Place 105 Commercial Road Southampton SO15 1GH United Kingdom Email: maib@dft.gov.uk Telephone: +44 (0)23 8039 5500 Press enquiries during office hours: +44 (0)1932 440015 Press enquiries out of hours: +44 (0)300 7777878 Maritime Authority of the Cayman Islands Breezy Castle Unit 1 125 Owen Roberts Drive P.O. Box 2256 Grand Cayman KY1-1107 www.cishipping.com/contact-us © Crown copyright, 2021 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned.

CONTENTS GLOSSARY OF ABBREVIATIONS AND ACRONYMS SYNOPSIS 1 SECTION 1 – FACTUAL INFORMATION 2 1.1 Particulars of Stolt Groenland and accident 2 1.2 Narrative 3 1.2.1 Passage and cargo discharge 3 1.2.2 Explosions 7 1.2.3 Emergency response 7 1.3 Environmental conditions 10 1.4 On-site investigation 10 1.5 Damage 10 1.6 Onboard records 11 1.7 Styrene monomer 12 1.7.1 Description 12 1.7.2 Polymerisation 16 1.7.3 Polymerisation inhibitors 16 1.8 Cargo system 18 1.9 Loading in Houston 18 1.9.1 Planning and documentation 18 1.9.2 The cargo 20 1.9.3 Styrene and inhibitor mixing 21 1.9.4 Styrene sampling 21 1.10 Cargo temperature changes during the voyage 23 1.11 Crew 25 1.11.1 General 25 1.11.2 Master 25 1.11.3 Chief officer (Ulsan) 25 1.11.4 Chief officer (loading and voyage) 25 1.11.5 Voyage routines 25 1.12 Cargo management ashore 26 1.13 Onboard cargo procedures 27 1.13.1 Planning and stowage 27 1.13.2 Cargo monitoring 27 1.14 IBC Code 27 1.15 Shipping industry guidance 29 1.15.1 Tanker Safety Guide 29 1.15.2 Chemical Distribution Institute Guidance 29 1.16 Storage and handling guidance 30 1.16.1 Styrene Monomer: Safe Handling Guide (2018) 30 1.16.2 Safe handling sheets 31 1.16.3 Safety data sheets 32 1.17 Recent inspections, audits and vetting 32 1.18 Yeompo Quay 32 1.19 Styrene monomer accidents and incidents 33 1.19.1 Stolt Focus 33 1.19.2 Other maritime accidents 34 1.19.3 Shore-based accidents 34 1.20 Styrene sample and overall analysis results 35

SECTION 2 – ANALYSIS 37 2.1 Aim 37 2.2 Explosion and fire mechanisms 37 2.3 Initiation of polymerisation 37 2.4 Heat transfer 38 2.5 Polymerisation inhibitor mixing 39 2.6 Polymerisation inhibitor depletion 39 2.7 Cargo stowage 40 2.8 Temperature monitoring 42 2.9 Action in the event of polymerisation 43 2.10 Fire-fighting 44 2.11 Stolt Focus incident 44 SECTION 3 – CONCLUSIONS 46 3.1 Safety issues directly contributing to the accident that have been addressed or resulted in recommendations 46 3.2 Safety issues not directly contributing to the accident that have been addressed or resulted in recommendations 47 3.3 Other safety issues not directly contributing to the accident 47 SECTION 4 – ACTION TAKEN 48 4.1 MAIB actions 48 4.2 Actions taken by other organisations 48 SECTION 5 – RECOMMENDATIONS 49 FIGURES Figure 1: Ofdjell terminal and Yeompo Quay, Ulsan Figure 2: Cargoes remaining following discharge in Ulsan Figure 3: CCTV still showing vapour release from 9S p/v valve Figure 4: CCTV still showing tank rupture and vapour release Figure 5: CCTV still showing vapour ignition Figure 6: Fireball at Ulsan bridge Figure 7: Bow Dallian main deck with burned styrene deposits Figure 8: Styrene residue sample Figure 9: Overalls from Bow Dallian crewman Figure 10: 9S heating control valves

Figure 11: Aerial view of main deck and location of 9S Figure 12: 9S tank rupture Figure 13: 9S access hatch Figure 14: Midships deckhouse Figure 15: Accommodation interior Figure 16: Accommodation front Figure 17: Cargo temperatures recorded on the VDR Figure 18: Certificate of Inhibitor

Final sample analysis results

9S Styrene Monomer Temperature 20-28 September TABLES

Table 2: Temperatures of cargoes discharged in Kobe and Ulsan

Table 4: TBC depletion in styrene (if stored under air)

Table 8: Calculated TBC depletion using figures provided in the Plastics Europe Styrene Monomer: Safe Handling Guide (2018)

GLOSSARY OF ABBREVIATIONS AND ACRONYMS

2/O

Second officer

3/O

Third officer

ACA

American Cargo Assurance

AIS

Automatic identification system

CCR

Cargo control room

CDI

Chemical Distribution Institute

C/O

Chief officer

ERT

Emergency Response Team

HMD

Hexamethylenediamine

IBC Code

International Code for the Construction and Equipment of Ships Carrying Dangerous Chemicals in Bulk

IMO

International Maritime Organization

Ineos

Ineos Styrolution America LLC

INTERTANKO

International Association of Independent Tanker Owners

iStow

iStow Tanker cargo stowage software

KMST

Korea Maritime Safety Tribunal

LPD

LPD Lab Services Limited

m

metre

m³

cubic metres

mb

millibar

MARPOL

International Convention for the Prevention of Pollution from Ships

MCO

Marine Compliance Officer

OS

Ordinary seaman

OTK

Odfjell terminal in Ulsan, Republic of Korea

ppm

parts per million

PTZ

Phenothiazine

p/v

Pressure/vacuum

SDS

Safety Data Sheet

SHC

Specific Heat Capacity

SMS

Safety Management System

SOLAS

International Convention for the Safety of Life at Sea 1974, as amended

ST B503

Shore Tank 503

STCW

International Convention on Standards of Training, Certification and Watchkeeping for Seafarers 1978, as amended (STCW Convention)

Stolt

Stolt Tankers B.V.

t

tonne

TBC

4-tert-Butylcatechol

TEA

Triethanolamine

TSG

Tanker Safety Guide

UTC

Universal co-ordinated time

VCR

Voyage cargo report

VDR

Voyage data recorder

Voranol

Glycerol, propoxylated and ethoxylated

YMP3

Yeompo Quay, berth 3 TIMES: all times used in this report are UTC + 9 unless otherwise stated.

Stolt Groenland Permission requested via email. Image courtesy of Hannes van RijnThe chemical tanker Stolt Groenland at sea

1 SYNOPSIS On 28 September 2019, a cargo tank containing styrene monomer on board the Cayman Islands registered chemical tanker Stolt Groenland ruptured causing an explosion and fire. The tanker was moored alongside a general cargo berth in Ulsan, Republic of Korea and the Singapore registered chemical tanker Bow Dalian was moored outboard. The ignition of the styrene monomer vapour resulted in a fireball, which reached the road bridge above. Both vessels were damaged, and two crew suffered minor injuries. Fifteen emergency responders were injured during the fire-fighting, which lasted for over 6 hours. The rupture of the styrene monomer tank resulted from a runaway polymerisation that was initiated by elevated temperatures caused by heat transfer from other chemical cargoes. The elevated temperatures caused the inhibitor, added to prevent the chemical’s polymerisation during the voyage, to deplete more rapidly than expected. Athough the styrene monomer had not been stowed directly adjacent to heated cargo, the potential for heat transfer through intermediate tanks was not fully appreciated or assessed. Critical temperature limits had been reached before the vessel berthed under the road bridge in Ulsan. The tanker’s crew did not monitor the temperature of the styrene monomer during the voyage, and therefore were not aware of the increasingly dangerous situaution. A similar dangerous styrene monomer polymerisation incident had occurred a couple of weeks ealier on board another Stolt Tankers B.V. ship, Stolt Focus. The heat generated by the polymerisation process was noticed before the critical runaway temperature was reached. The styrene momomer cargoes on board both tankers was loaded at a similar time from the same tank in Houston and were exposed to similar environmental conditions. The incident on board Stolt Focus was not reported to the ship’s Flag State or other masters in the Stolt Tankers B.V. fleet. Following the accident, the Ministry of Oceans and Fisheries, Republic of Korea, prohibited ship-to-ship transfer operations for dangerous cargo on general cargo berths in Ulsan. Stolt Tankers B.V. took immediate action to ensure that the temperatures of all cargoes carried on board its ships were monitored and reported to its shore management. It also took steps to enhance crew awareness on the hazards of inhibited and heat sensitive cargoes. The company is developing technological and administrative initiatives to assist with the safe stowage and monitoring of heat sensitive cargoes. A recommendation has been made to Stolt Tankers B.V. aimed at ensuring the wider marine chemical sector benefits from the lessons learned from the Stolt Focus incident and research initiatives that were carried out as a result of this accident. Recommendations have also been made to the Cayman Island Shipping Registry, the Chemical Distribution Institute and Plastics Europe (Styrene Producers Association). These are intended to assist in ensuring that the guidance provided in certificates of inhibitor and styrene monomer handling guides is consistent and achievable given the limitations of equipment and testing facilities on board ships.

2 SECTION 1 – FACTUAL INFORMATION 1.1 PARTICULARS OF STOLT GROENLAND AND ACCIDENT SHIP PARTICULARS Vessel’s name Stolt Groenland Flag Cayman Islands Classification society Det Norske Veritas Germanischer Lloyd IMO number/fishing numbers 9414072 Type Chemical/products tanker Registered owner Stolt Tankers B.V. Manager(s) Stolt Tankers B.V. Construction Steel Year of build 2009 Length overall 182.72m Beam 32.24m Gross tonnage 25881 Deadweight 43478 Minimum safe manning 17 Authorised cargo Chemicals and oil products in bulk VOYAGE PARTICULARS Port of departure Kobe, Japan Port of arrival Ulsan, Republic of Korea Type of voyage International Cargo information Chemicals in bulk 27117 tonnes Manning 25 MARINE CASUALTY INFORMATION Date and time 28 September 2019 at 1050 Type of marine casualty or incident Serious Marine Casualty Location of incident Ulsan, Republic of Korea Place on board Cargo tanks and deck Injuries/fatalities 2 crew suffered minor injuries (1 on board Stolt Groenland and 1 on board Bow Dalian). 15 shore workers/officials were also reported to have been injured. The severity of their injuries is not known Damage/environmental impact Toxic vapour released to atmosphere Ship operation Moored alongside Voyage segment Moored alongside External & internal environment Daylight. Air temperature: 24ºC Wind: Light airs. Humidity 93% Persons on board 26

3 1.2 NARRATIVE 1.2.1 Passage and cargo discharge Between 2 and 17 August 2019, Stolt Groenland, a Cayman Islands registered chemical tanker operated by Stolt Tankers B.V. (Stolt), loaded 20 different chemical cargoes from terminals in Texas, USA. Cargo was carried in 37 of the tanker’s 39 cargo tanks for delivery in various ports in the Far East (Table 1). Tanks 3 Port (3P), 7 Starboard (7S), 8P, 9P, 10P, and 11P contained heated cargoes. The temperatures of the cargoes in 3P and 7S were required to be maintained at between 40ºC and 45ºC, and the temperatures of the cargoes in 8P, 9P, 10P, and 11P (Hexamethylenediamine (HMD)), the last cargo to be loaded, had to be maintained at between 45°C and 50°C. The temperature of the HMD at loading was 61°C. Table 1: Cargo stowage on departure from Houston on 17 August Detai ls of the car go stowage showing locations of h eated cargo and styrene monomer

4 The cargoes also included 5245t of styrene monomer, loaded on 7 and 8 August, which was divided among cargo tanks 6P (1789t), 6 Centre (6C) (1671t), and 9S (1785t), to which a polymerisation inhibitor had been added (see Section 1.7.2). The tanker sailed for passage to Japan on 17 August. Between 22 and 24 September, Stolt Groenland’s cargo tanks 5S, 8C (part cargo), 8S and 11C were discharged to barges while the vessel was at anchor off Kobe, Japan. The vessel then sailed to Ulsan, Republic of South Korea, where six tanks containing adiponitrile were discharged at the Odfjell terminal (OTK) on 26 and 27 September. During discharge on 26 September, the tanker’s chief officer (C/O) handed over to his relief, who had arrived shortly after the vessel moored alongside. The off-going C/O departed the vessel at about 1500. During the afternoon of 27 September 2019, Stolt Groenland shifted from the OTK to Yeompo Quay, berth 3 (YMP3) underneath the Ulsan Bridge (Figure 1). The tanker moored starboard side to the quay and during the evening, a cargo of Voranol was discharged from cargo tanks 10C and 11S to Stolt Voyager via ship-to-ship transfer. On completion, Stolt Voyager moored ahead of Stolt Groenland. The temperatures of the cargoes discharged in Kobe and at OTK recorded on the ullage reports are shown in Table 2. The attending cargo surveyor in Ulsan noticed that the temperature of the adiponitrile in 9C (48.8ºC) was elevated, but as it remained within the charterer’s specified maximum temperature of 50ºC, he did not discuss this with the C/O and no action was taken. The distribution of the cargoes remaining on board Stolt Groenland following the discharges in Kobe and OTK is shown at Figure 2. Table 2: Temperatures of cargoes discharged in Kobe and Ulsan Date Tank Cargo Place Discharge Temperature 22-23 Sep 5S Toluene Kobe, Japan 33.2°C 22-23 Sep 8C (part cargo) Triethanolamine (TEA) Kobe, Japan 43.6°C 22-23 Sep 8S ALFOL Kobe, Japan 39.1°C 22-23 Sep 11C MEA Kobe, Japan 48.2°C 26-27 Sep 1S ADIPONITRILE OTK, Ulsan 27.2°C 26-27 Sep 2P ADIPONITRILE OTK, Ulsan 30.0°C 26-27 Sep 3C ADIPONITRILE OTK, Ulsan 29.8°C 26-27 Sep 5C ADIPONITRILE OTK, Ulsan 30.5°C 26-27 Sep 6S ADIPONITRILE OTK, Ulsan 31.9°C 26-27 Sep 9C ADIPONITRILE OTK, Ulsan 48.8°C 27 Sep 10C VORANOL YMP3, Ulsan 46.0°C 27 Sep 11S VORANOL YMP3, Ulsan 38.4°C

5 Figure 1: Ofdjell terminal and Yeompo Quay, Ulsan Satellite image courtesy of Google Maps Satellite image shows Ofdjell terminal and Yeompo Quay, Ulsan Ofdjell terminal Yeompo Quay

6 Figure 2: Cargoes remaining following discharge in Ulsan Heated (3P, 7S, 8P, 9P, 10P and 11P) Empty (1S, 2P, 3C, 5C, 5S, 6S, 8S, 9C, 10C, 11C, 11S, 12P and 12S) Styrene monomer (6P, 6C and 9S) Other cargo (the tanks that are left) Plan il lust ration in dic ates position of heated, empty, styrene monomer and o ther cargo rem aining on Sto lt Groenland FORWARD PORT STARBOARD AFT Cofferdam Cofferdam

7 1.2.2 Explosions At about 0600 on 28 September, the Singapore registered chemical tanker Bow Dalian was secured alongside Stolt Groenland’s port side. The purging of Bow Dalian’s cargo tanks with nitrogen supplied from shoreside vehicles was then started in readiness for a ship-to-ship cargo transfer with Stolt Groenland. At 1043, vapour started to release from the pressure/vacuum (p/v) relief valve for Stolt Groenland’s 9S cargo tank (Figure 3). About 2 minutes later, the tank’s high-level alarm (indicating that the level in 9S cargo tank had reached 95%) activated the warning siren and strobe light on the bridge front. The warning indicators were noticed by the ordinary seaman (OS) on gangway watch, and he immediately notified the third officer (3/O), the on-watch deck officer, via hand-held radio. The 3/O was working in the master’s office and went to the unmanned cargo control room (CCR) to investigate. The OS’s radio transmission was also heard by the C/O, and he joined the 3/O in the CCR. Shortly afterwards, the high-high-level siren on the bridge front sounded, indicating that the level in 9S cargo tank had increased to 98%. The alarm was again reported by the OS. The C/O and 3/O saw from the cargo monitoring system that the pressure inside 9S cargo tank was 1340mb, and observed it rise quickly to 2000mb. At 1050, two explosions occurred in rapid succession in way of the vessel’s cargo manifold (Figure 4). The second explosion ignited the released styrene monomer vapour (Figure 5) and the resulting fireball passed very close to the road bridge above the quay (Figure 6). It also ignited vapours released from the 9S and 9C p/v valves. 1.2.3 Emergency response Stolt Groenland On board Stolt Groenland, the force of the second explosion blew the gangway watchman over the starboard side guardrails on the main deck. The OS initially held on to one of the rails, but he was soon forced to let go due to the heat from the fire. He fell into the water between the quay and the vessel, but he was able to climb over a fender and on to the quay. The C/O activated the fire alarm and the tanker’s crew mustered at their fire stations. The C/O also activated the deck foam monitor system and directed the port monitor toward the cargo manifold. The 3/O did likewise with the starboard monitor. However, the master soon ordered everyone to the lifeboat station on the boat deck due to the intensity of the fire and the thick black smoke. The crew used the freefall lifeboat to abandon the tanker. Bow Dalian Following the explosions, Bow Dalian’s C/O activated the vessel’s emergency alarm and the deck foam fire extinguishing system. He and the master also directed the monitors toward Stolt Groenland’s main deck and cargo manifolds. The crew then abandoned ship via a rope ladder on the vessel’s port side on to Korea Coast Guard vessels, which had arrived to assist.

8 Figure 3: CCTV still showing vapour release from 9S p/v valve Vapour starting to release Figure 4: CCTV still showing tank rupture and vapour release Vapour explosion from ruptured tank Figure 5: CCTV still showing vapour ignition Ignition of the released styrene monomer vapour Vapour release Vapour explosion from tank 9S

9 Figure 6: Fireball at Ulsan bridge Image courtesy of @803_Gorani (Twitter) Fireball very close to the road bridge above the quay Stolt Stolt was informed of the explosions and fire on board Stolt Groenland by a marine compliance officer (MCO) who was on board Stolt Voyager, and immediately activated its Emergency Response Team (ERT) in the Netherlands. The ERT communicated with its MCO via telephone and ‘WhatsApp messenger’ and established that Stolt Groenland’s crew were safe. At 1528, the ERT were informed of a further explosion toward the tanker’s accommodation. However, at 1612, it received a report that the fire had dampened down and that no flames were visible. Shortly afterwards, the fire was reported to have been extinguished, although there was still smoke in the accommodation. Stolt’s protection and indemnity insurers, Gard, appointed the Netherlands based salvors Ardent, to assist Stolt’s emergency response team. It also e-mailed its fleet to check styrene monomer cargoes that were being carried. This was later backed up by fleet Flash Report no.2019/08 highlighting the precautions required when styrene monomer is carried. Local emergency services Local fire brigade shore appliances arrived at YMP3 at 1101 and started to fight the fire. Korea Coast Guard fire-fighting tugs and launches assisted and towed Bow Dalian to a berth clear of Yeompo. A total of 726 emergency responders participated in the firefighting and rescue operations, with 117 fire-fighting appliances being deployed to the scene. A total of two ships’ crew and 15 emergency responders suffered injuries and were taken to hospital for observation or treatment.

10 1.3 ENVIRONMENTAL CONDITIONS During the cargo loading in Texas and during passage in the Caribbean, the average air and sea temperatures experienced by Stolt Groenland were about 30°C, peaking at 37°C air temperature during the day. These averages reduced to about 25ºC during the passage across the Pacific Ocean. In Kobe, the sea temperature was 26ºC and the maximum and minimum air temperatures were 30ºC and 28ºC. In Ulsan, the sea temperature was 24ºC with maximum and minimum air temperatures of 25ºC and 19ºC. 1.4 ON-SITE INVESTIGATION On 28 September, the Maritime Authority of the Cayman Islands requested the MAIB investigate this accident in accordance with the International Maritime Organization (IMO) Casualty Investigation Code (IMO Resolution MSC 255(84)). During the on-site investigation in Ulsan, the MAIB was assisted by the Korea Maritime Safety Tribunal (KMST). Access to Stolt Groenland was controlled by the Korea Coast Guard and restricted by the toxicity of the atmosphere on board. On 3 October, with the co-operation of Bow Dalian’s owners and the Transport Safety Investigation Bureau of Singapore, MAIB inspectors and KMST staff visited Bow Dalian in Ulsan. The tanker’s weather decks and accommodation block were spattered with burned residues (Figure 7). A sample of residue (Figure 8), and the overalls (Figure 9) from a crewman who had been on deck during the period of vapour release from 9S’s p/v valve, were removed for analysis (see Section 1.20). On 7 October, the analysis of samples of styrene monomer taken from Stolt Groenland’s cargo tanks 6C and 6P indicated that the concentrations of the inhibitor 4-tert-Butylcatechol (TBC) were 8 parts per million (ppm) and 7ppm respectively. The following day, MAIB inspectors were allowed access to the vessel. However, the presence of hazardous chemicals and fire damage limited the inspection to the main deck primarily. A further inspection on 12 November was co-ordinated by the Korea Coast Guard and the KMST. These inspections, and inspections conducted by or on behalf of Stolt, confirmed that the heating coils in 9C and 9S were closed and blanked (Figure 10), and that vapours released from the p/v valves for 9S and 9C had ignited. 1.5 DAMAGE A large hole was found in Stolt Groenland’s main deck in way of cargo tank 9S and its common bulkhead with cargo tank 9C (Figures 11 and 12). The access hatch cover to 9S had also been blown off (Figure 13). The manifold area was fire damaged to varying degrees, and the midships deckhouse was burned out (Figure 14). The tanker’s accommodation block sustained extensive internal fire damage (Figure 15). The deck between the technical room and the CCR had collapsed, and both spaces were burned out. Heat and smoke had also penetrated the bridge and damaged much of its equipment. Other than broken windows, many of which were broken by firefighters during the emergency response, external damage to the front of the accommodation block was minimal (Figure 16).

11 1.6 ONBOARD RECORDS Cargo-related records and voyage related records such as ‘noon reports’ were available on ‘Veslink’, a networked system that was used on board Stolt Groenland to store data and to communicate with interested parties ashore. In addition to mandated information, Stolt Groenland’s voyage data recorder (VDR) recorded cargo tank levels, volumes, temperatures, and liquid densities. All cargo-related records (digital and paper) that were held in the CCR and the cargo samples stored in the midships deckhouse were destroyed in the fire. However, the cargo temperature data from 0001 (UTC) on 30 August to 0150 (UTC) on 28 September (Figure 17), the time range of the data that was available, was recovered from the VDR. Figure 7: Bow Dallian main deck with burned styrene depositsMain deck spattered with burned residues Figure 8: Styrene residue sample Close-up of styrene residue sample to scale at approximately 90cm across Figure 9: Overalls from Bow Dallian crewmanBlackened overalls of crewman on deck during period of vapour release

12 1.7 STYRENE MONOMER 1.7.1 Description Styrene monomer (also known as ethenylbenzene, phenylethylene, phenylethene, vinylbenzene, or cinnamene) is an aromatic hydrocarbon and a building block of the plastics industry. It is commonly used in the manufacture of a variety of plastic, rubber and polystyrene products. Over 30 million tonnes of styrene are produced annually, much of it transported by sea to plastics production plants. Styrene monomer is a colourless, transparent liquid under ambient conditions and has a distinctly sweet odour. It is a volatile and flammable substance with a flashpoint of 32ºC. Its boiling point is 145ºC and its auto ignition temperature is 490ºC. The harmful effects of styrene monomer include severe irritation to eyes and mucous membranes, as well as gastrointestinal effects. Chronic exposure to styrene monomer leads to central nervous system dysfunction, such as headache, fatigue, weakness, depression, hearing loss and nerve damage. Figure 10: 9S heating control valves9S heating control valves in which coils were closed and blanked

13 Figure 11: Aerial view of main deck and location of 9S Stolt Groenland’s main deck showing cargo tanks 9S and 9C and the approximate area of tank split 9C cargo tank area Approximate area of tank split (at deck level) 9S cargo tank area

14 Figure 12: 9S tank rupturePolymerised styrene in ruptured tank Figure 13: 9S access hatchAccess cover to 9S blown off Polymerised styrene

15 Figure 14: Midships deckhouse Midships deckhouse burned out Figure 15: Accommodation interiorExtensive internal fire damage to accommodation interior

16 Figure 16: Accommodation frontMinimal damage to accommodation exterior 1.7.2 Polymerisation Polymerisation is a chemical reaction, or process in which a monomer or a mixture of monomers is converted into a polymer such as polystyrene. Styrene polymerises slowly at normal ambient temperatures but very rapidly at elevated temperatures. It can be accelerated by heat, the lack of dissolved oxygen, the lack of a polymerisation inhibitor, and when contaminated by oxidising agents and most halides. The polymerisation process is exothermic and, if the resulting heat is not removed, the bulk styrene temperature may rise to a level at which polymerisation is self-sustaining and very rapid. This is referred to as ‘runaway polymerisation’ and will usually be initiated by temperatures above 65°C. During a runaway polymerisation, the cargo will expand causing pressure to increase to the point that vapour is released from tank vents or p/v valves. In some cases, the resulting build-up of pressure is sufficient to rupture the tank. 1.7.3 Polymerisation inhibitors To prevent polymerisation during storage and transportation, an inhibitor must be added. TBC, the most commonly used polymerisation inhibitor, is a solid but is often mixed with methanol to produce a liquid before adding to styrene monomer. In shore storage facilities, TBC is added to styrene monomer through dosing systems. For the marine transportation of styrene monomer in bulk, the TBC is added to the bottom of a cargo tank before loading. TBC should be stored in non-reactive, light resistant containers at ambient temperature. It is not known to have a limited shelf life.

17 Figure 17: Cargo temperatures recorded on the VDR Line gra ph s hows tan k te mper ature s b et wee n 1 6 and 27 S ept emb er (U TC) , hea tin g coi l v al ves op en ed and r ise of s tyr ene m ono mer t emp era tu re ove r 3 d ays 30.0 40.0 50.0 60.0 70.0 Stolt Groenland cargo tank temperatures Temperature ºC 9S styrene monomer temp 65° Heating applied Kobe, Japan Ulsan OTK YMP 30 August September 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Heated HMD cargo in tanks 8P, 9P, 10P and 11P Heated cargo in 7S 6P-16 6C-17 6S-18 7P-19 7C-20 7S-21 8P-22 8C-23 8S-24 9P-25 9C-26 9S-27 10P-28 10C-29 10S-30 11P-31 11C-32 11C-32 9S - styrene monomer >1°C temp rise each day

18 The TBC level must be maintained within a specified concentration range, typically between 10-15ppm, to prevent polymer formation. The presence of dissolved oxygen is required for the inhibitor to function properly and it is the inhibitor that controls oxygen depletion. Where ambient temperatures are likely to exceed the flash point of the cargo, it might be desirable to use a nitrogen blanket to minimize the risk of fire or explosion. However, the atmosphere above styrene monomer should contain 3% to 8% (volume) of oxygen to maintain inhibitor effectiveness and therefore avoid polymerisation. Whether stored under air or a nitrogen blanket, the inhibitor concentration is depleted more rapidly at elevated temperatures. Consequently, styrene monomer is usually kept below 30°C when transported and below 20°C in storage tanks ashore. 1.8 CARGO SYSTEM Stolt Groenland was built in 2009 and was one of six sister vessels. The tanker had 39 integral cargo tanks of varying capacities, which were separated into three fore to aft sections by two transverse cofferdams. The tanks in the fore and aft sections (tanks 1-3 and 12-13) were constructed of zinc coated mild steel; the mid-section (tanks 4-11) was of solid duplex stainless steel construction. The tanks’ bulkheads were corrugated. All stainless steel tanks were fitted with heating coils, capable of maintaining a temperature of up to 85°C. The heating medium was water, and the system was controlled manually via valves on deck. Each cargo tank was independent and had its own pump and pipeline arrangement. The tanks were protected against over and under pressures by individual p/v relief valves, which were situated at two vent stacks on the main deck. The lifting pressure of the p/v valves was 206mb. Tank liquid levels were measured by a closed radar system, and tank pressures and temperatures by internal sensors. The temperature sensors were located at low and middle levels in each tank. Data from the radar and tank sensors was monitored via an Ariston computerised cargo monitoring system in the CCR (with a slave display on the bridge). The data was also fed to the vessel’s CargoMax stability computer. The information displayed by the Ariston system on the cargo monitor screens was determined by user preference. The system could be configured to show individual tanks or groups of tanks. The user was also able to set cargo level, temperature and pressure alarms for each tank. Stolt Groenland’s crew could not test or monitor for TBC concentration, oxygen or polymer levels in the styrene monomer cargoes. These values could only be determined in shore laboratories. 1.9 LOADING IN HOUSTON 1.9.1 Planning and documentation The cargo stowage plan for the loading between 2 and 17 August was first sent to Stolt Groenland by the tanker’s shore-based ship operator (see Section 1.11) on 19 July. Inconsistencies between the voyage cargo report (VCR) and the vessel’s iStow Tanker (iStow) cargo stowage computer, were raised by the C/O. These were subsequently addressed by the ship operator, and the plan was re-issued. Following further iterations of the plan, during which the intended stowage of the styrene monomer cargo was amended from cargo tanks 6P, 6C, and 10S, to 6P, 6C and 9S,

19 a final iStow file and VCR was issued by the ship operator on 1 August. The plan was checked and agreed by Stolt Groenland’s C/O and approved by the tanker’s master. The VCR listed the nominated cargoes for the forthcoming voyage. It also cross- referenced the bill of lading number with the chemical name, quantity, ports of loading and discharge, cargo-handling requirements, and safety references. It included: 430 STYRENE MONOMER Generic specs: HANDLING UPDATED 10 JUL 18 FOR SHELL REVISION 9 DATED 01 APR 18 HEAT ADJ NO HEATED CARGO TO BE STOWED ADJACENT. IMO Ships carrying this product must be provided a Certificate of Protection (Inhibitor Certifcate) from the shipper in order to comply with the requirements of 15.13.3. 1) Name and amount of additive/inhibitor 2) If additive is Oxygen dependent 3) Date additive was added to cargo and duration of effectiveness 4) Temperature limitations affecting effective lifetime of additive 5) Action that needs to be taken should voyage exceed effective lifetime of additive or a statement concerning the Oxygen level required if the exclusion of air method is being used to stop oxidization as per 15.13.4. If an Inhibitor Certificate is to be issued it MUST state if the Inhibitor is Oxygen Dependent, or not, and if it is Oxygen Dependent it MUST state the Oxygen level which will allow the inhibitor to be effective. If you do not receive this documentation please alert your Ship Operator as soon as possible. IBC Code 15.19.6 High Level alarm in cargo tank IBC Code 16.6.1 NO ADJACENT HEAT as this cargo can potentially go into a state of POLYMERIZATION, DECOMPOSITION. THERMAL INSTABILITY OR EVOLUTION OF GAS IBC Code 16.6.2 HEATING COILS IN TANK(S) loading this cargo MUST BE BLANKED OFF IMO Name STYRENE MONOMER INCOMPAT Polymerization catalysts. Strong oxidizing agents. The Shell document referred to in the VCR was their Cargo Handling Sheet for styrene monomer, which was intended for use by vessels chartered by Shell Chemicals. It included: Loading Temperature Range: 13 – 23 °C / 55 – 73 °F Transit Temperature Range: Ambient Discharge Temperature Range: Ambient

20 Maximum Heating Coil Temperature: Blanked Off Adjacent Maximum Cargo Temperature: 35°C Note 1: This product is heat sensitive, self-reactive and inhibited. Note 2: Shell Chemical SM is generally inhibited with Para-tertiary Butyl Catechol (p-TBC), typically at 10-20 ppm or more, depending on duration of voyage. Where required to add additional inhibitor to a loaded tank of styrene, this should be done using closed equipment. If the equipment is not available and there is a requirement to add inhibitor, the local Chemical MTA should be consulted. DAILY LOG: During the voyage the vessel shall maintain a daily log of the following and, upon request, send the log to the responsible Shell Chemicals Charterer: 1. cargo (Styrene) temperature 2. adjacent cargo temperature 3. air and sea water temperature If during the voyage any of the following is observed, the responsible Shell Chemicals Charterer shall be notified immediately: ●1 °C rise of cargo (styrene) temp per day, over 3 consecutive days ●> 2 °C rise of cargo (styrene) temperature within any 24 hours ●Cargo temperature at any time raises above > 30°C ●O2 content in cargo (styrene) tank ≤4% by volume, when inerted After completion of cargo (Styrene) discharge the vessel shall provide a copy of the daily log of temperature/pressure/O2 content to the responsible Shell Chemicals Charterer. Due to risk of polymerization, Styrene should not be carried in tanks serviced by a cargo pump room. 1.9.2 The cargo The styrene monomer carried on board Stolt Groenland was supplied by Ineos Styrolution America LLC to the LBC Terminal in Houston, USA. The cargo was pumped from shore tank B503 (ST B503) to the tanker during 7 and 8 August 2019 and was designated as cargo OBL430. The charterer on the Bill of Lading was Samsung C&T Corporation. Prior to loading, the cargo tanks 6P, 6C and 9S had passed wall wash tests and visual inspections conducted by a cargo surveyor from American Cargo Assurance (ACA), and a preloading acceptance certificate was issued.

21 During the afternoon of 7 August, the ACA cargo surveyor and Stolt Groenland’s C/O completed a preloading checklist and exchanged cargo documentation which included a safety data sheet (SDS) for styrene monomer (stabilised) issued by Ineos Styrolution. Shortly afterwards, the surveyor carried out a final visual check of 6P, 6C, and 9S and then poured 3 US gallons (11.4 litres) of liquid inhibitor TBC into each of the tanks through their respective cleaning hatches. The addition of the inhibitor was witnessed by a terminal representative. No inhibitor was provided to the tanker’s crew to carry on board during the voyage. The loading of the styrene monomer started at 1735 (UTC-5) on 7 August from one shore cargo hose. The hose was connected to the tanker’s manifold via a splitter spool piece connected to tanks 6P and 6C. A cross-connection on the manifold also enabled the simultaneous loading of cargo tank 9S. The temperature of the styrene monomer stored in ST B503 was 13ºC. Between 0146 and 0340 on 8 August, pumping was stopped while awaiting the results of the shoreside testing of styrene samples from the cargo tanks. The analysis did not identify any issues, and loading was completed by 1520. The observed temperatures of the styrene monomer in Stolt Groenland’s cargo tanks shortly after the loading was completed ranged between 16.7 and 17.2ºC. At 1550, the ACA surveyor sealed the access and cleaning hatches for cargo tanks 6P, 6C and 9S. He then issued a Certificate of Inhibitor (Figure 18), which was signed by the attending terminal representative and Stolt Groenland’s C/O. This stated that the ideal temperature for the cargo was 60-85ºF (15.5-29.4ºC) and that if these conditions were exceeded, the cargo should be monitored for the inhibitor level and polymerisation, adding additional inhibitor as needed. 1.9.3 Styrene and inhibitor mixing The TBC liquid added to the tanks before cargo loading was thick and viscous, and its mixing with the styrene monomer relied on the agitation achieved during the loading process and the sloshing caused by vessel movement on passage. The partial agitation caused by the flow of the styrene monomer cargo as it loads into the tank was not sufficient to adequately mix the TBC with the styrene cargo. For this reason, TBC concentration tests, done both during and immediately after loading were prone to inaccuracy, and were therefore not undertaken. Instead, calculations were conducted and the notation ‘adjusted to XX ppm’ was used in the shipment analysis reports. This process was standard industry practice for styrene bulk transportation by sea. 1.9.4 Styrene sampling During the loading of the styrene monomer, the ACA surveyor took 1 US pint (0.47 litre) samples of the cargo on opening ST B503 from the dock-line, the cargo manifold and the 6P, 6C, and 9S pump stacks. The surveyor also took samples from each tank at the first foot1 and the final levels. The surveyor passed the samples to the C/O and these were then stored in the midships deck housing. The surveyor also sent a sample from ST B503, and cargo tank samples to an Ineos laboratory ashore for testing. The samples from 6P, 6C and 9S were combined for both the first foot and final levels, and were tested as composite samples. 1 The ‘first-foot’, or about 30cm, of loaded cargo was sampled to check if the cargo pipeline system was clean.

22 Figure 18: Certificate of InhibitorCertificate of inhibitor issued by ACA surveyor which was signed by attending terminal representative and Stolt Groenland’s C/O

23 The Styrene Shipment Analysis report concerning the composite final sample (Figure 19) stated that the TBC concentration was adjusted to 17ppm and the Certificate of Inhibitor stated that the ‘Days Effective’ was 60-90 days. The concentration level was a calculation based on the TBC concentration in ST B503 (11.3ppm) and the amount of inhibitor that was added into the tanks prior to loading. 1.10 CARGO TEMPERATURE CHANGES DURING THE VOYAGE Stolt Groenland’s styrene monomer cargo was due to be discharged in An-Ping, Taiwan about 56 days after it was loaded in Houston. The VDR provided hourly cargo temperature records for the 30-day period prior to the explosion (Figure 17). At 1200 (UTC) on 30 August the recorded temperatures of the styrene monomer in 6P, 6C and 9S were 31.5, 31.4 and 37.2°C respectively. The temperatures of the HMD in 9P and the adiponitrile in 9C were 51.5 and 43.9°C respectively. The temperature variations recorded in cargo tanks 8P, 9P, 10P and 11P showed that the HMD had been routinely heated to about 55°C and then allowed to drop to about 45°C. This was achieved by opening and closing the heating coil valves; it typically took about 2 days to heat the HMD and 5 days to for its temp to drop. This weekly heating cycle resulted in temperature variations of about 2.5°C in 9C and 0.5°C in 9S. On 18 September, the heating coil valves for 8P, 9P, 10P and 11P were opened and the temperature was increased to about 55°C, where it was maintained ahead of the planned cargo discharge programme. At 1200 (UTC) on 18 September, the temperatures of the cargoes in 8S, 9S, 10S and 11S were 37, 37.7, 37.8 and 36.1°C respectively. The subsequent daily 1200 (UTC) temperatures of the starboard tanks are shown in Table 3. Table 3: Midday temperatures for tanks 8S, 9S, 10S and 11S between 19 and 27 September 2019 Dates 8S midday temperature (°C) 9S midday temperature (°C) 10S midday temperature (°C) 11S midday temperature (°C) 19 Sept 37.3 37.6 37.8 36.2 20 Sept 37.7 37.9 37.7 36.2 21 Sept 37.9 38.4 37.9 36.2 22 Sept 38.5 39.4 38.1 36.5 23 Sept 39.4 41.8 38.6 36.7 24 Sept 40.1 44.9 39.3 36.8 25 Sept 46.0 48.9 40.1 36.9 26 Sept 43.4 54.5 40.7 37.4 27 Sept 44.9 66.7 41.6 38.7 The temperature of the styrene monomer in 9S rose by 1°C on three consecutive days between 21 and 24 September and rose by 2.4°C between 22 and 23 September. It reached 65°C at about 1000 (UTC) on 27 September and 100.8°C immediately before the explosion.

24 Figure 19: Final sample analysis results

25 1.11 CREW 1.11.1 General Stolt Groenland’s crew comprised 10 officers and 15 ratings. The officers were Russian nationals and the ratings Filipino nationals. All crew held the appropriate STCW2 certification, including the chemical tanker basic training endorsement. The length of contract for the officers was 4 months. Two of the ratings were pumpmen, whose duties primarily concerned cargo operations. The master, senior officers and deck officers also had completed the STCW chemical tanker advanced training (see Section 1.15.2) within the last 5 years in accordance with Regulations V/1-1 and V/1 – 2. They viewed styrene monomer as a comparatively benign cargo when inhibited, and none had previously experienced problems with its carriage. 1.11.2 Master Stolt Groenland’s master had worked for Stolt since 1999, serving on a variety of ships. He had been a C/O for 5½ years and was promoted to master in 2009. Since 2016, he had been Stolt Groenland’s regular master, and prior to that he had served on its sister ships. The master last joined Stolt Groenland on 14 August 2019. 1.11.3 Chief officer (Ulsan) The C/O that joined the vessel in Ulsan started his career with Stolt in 2008 and had remained with the company throughout. He was promoted to second officer (2/O) in 2011, and to C/O in 2016. It was his fifth contract on Stolt Groenland, and he had also served on the vessel’s sister ships. 1.11.4 Chief officer (loading and voyage) The C/O that oversaw the cargo loading operation and completed the passage to Ulsan started his sea career as cadet in 2006 and worked on board oil tankers before joining Stolt as a 3/O in 2011. He was promoted to 2/O in 2012 and to C/O in

1. The C/O had served as C/O on Stolt Groenland three times, usually with the

same master and crew. He last joined the tanker in June 2019. 1.11.5 Voyage routines On passage, Stolt Groenland’s crew carried out watchkeeping duties and routine maintenance. The C/O worked days, supervising the deck crew and the pumpmen as well as completing administrative tasks and inspections. The C/O was responsible for monitoring the cargo and ensuring that the heated cargoes remained within the temperature ranges specified on the VCR. The adjustment of the heating controls to keep the heated cargoes within these parameters was delegated to one of the pumpmen. The second pumpman assisted with monitoring and maintaining pressures in the cargo tanks that had been protected with a nitrogen blanket. 2 International Convention on Standards of Training, Certification and Watchkeeping for Seafarers 1978, as amended (STCW Convention)

26 1.12 CARGO MANAGEMENT ASHORE Stolt was the world’s largest operator of chemical tankers, with a fleet of over 100 vessels. Cargo operations were co-ordinated from its main operations centres in Rotterdam, Houston and Singapore, and each vessel had a nominated ship operator who was responsible for cargo allocation and cargo loading and discharge programmes. The ship operator was the master’s shore-based point of contact for cargo-related issues. Stolt Groenland’s ship operator was based in Houston and was typically responsible for between four and five tankers at any one time. The ship operator had 23 years’ experience in the shipping industry and had worked for Stolt as a ship operator for 4 years, having initially completed 6 months’ on the job training under the supervision of an experienced manager. The ship operator used the iSTOW software, supported by Stolt’s ‘Chemscan’ database, to plan cargo stows. The database contained details of each chemical cargo, including temperature and nitrogen requirements, and enabled the iSTOW programme to identify inconsistencies in stowage plans related to aspects such as heated and cooled cargoes, heat in adjacent tanks, and cargo quality requirements. As a result, the iSTOW software informed the ship operator of the mix of cargoes a ship was able to carry. The user guide for iSTOW described the system as: …a computerized system that assists ship operators in the cargo stowing process. The system provides a “click, drag, and drop” feature for assigning of cargoes from a list to a cargo tank. STOW makes no attempt to stow the ship for the ship operator. Instead, it is a tool to assist the ship operator in the stow process and to alert them when any rules or regulations are violated. The system records stow changes made by the ship operator, sends and prints stow plans (including versions for last three cargoes and tank condition). Once the Stow Plan has been completed, the ship operator can ask the system to perform checks of the current Stow Plan. These checks use a combination of information from the Stolt CHEMSCAN Database and IMOS3. For example, if one of our customers has requested that we carry their cargo in a stainless-steel tank and that the adjacent cargoes are not heated above a certain temperature, the system will detect if there is a potential problem and alert the ship operator. Other checks are US Coast Guard compatibility, commingling, and last cargo restrictions. Additional checks are scheduled to be developed in compliance with changes in rules and regulations. On completion of verifying an intended cargo stowage plan in iSTOW, the ship operator forwarded the iSTOW file and the VCR to the vessel to check the viability of the stowage plan regarding factors including cargo compatibility, stability and trim. Stolt Groenland’s ship operator was aware of the potential for styrene monomer to polymerise and understood that the chemical should not be stowed adjacent to a heated tank. The ship operator assessed that the stowage of styrene monomer in 9S and the stowage of heated cargoes in 9P and 7S (i.e. separated by 9C and 8S) was acceptable. 3 Ship voyage management software.

27 1.13 ONBOARD CARGO PROCEDURES 1.13.1 Planning and stowage Stolt Groenland’s safety management system (SMS) included a cargo and ballast operations manual that provided guidance on the various phases of cargo operations. The standard operating procedures section of the manual included: Ship Operator stowage proposals as described in Cargo Orders Communication procedure are presented as proposals and must be closely studied taking into consideration the port and load berth rotation, draft, list, trim, loading sequence, etc. While operations try to give the Master maximum flexibility in making alterations, the Master and Chief Officer will often find that because of heating requirements, coating compatibility, tank sizes, pumps and cargo specifications, last cargo and customer requirements, there is often a limited choice of stowage. …It is the ship’s responsibility to make a detailed check of proper cargo stowage based on the proposal received from the Ship Operator. .. The final stow is always subject to the Master’s approval. No issues with the stowage of the styrene monomer cargo carried on board Stolt Groenland were raised by the tanker’s master or C/O. 1.13.2 Cargo monitoring The cargo and ballast operations manual also included: Care for the Customers cargo does not cease after loading, it continues throughout the voyage to the discharge port and during the discharge. Some cargoes may require additional attention and control during the voyage in respect to their single or combined concerns to: safety, quality, handling, cleaning, costumer’s demand etc. Ongoing cargo care is the most important aspect of the chemical tanker trade and the Chief Officer has the primary responsibility for conducting and recording these tasks… Voyage instructions required that the temperatures of cargoes requiring heating or cooling be recorded in a cargo temperature log at least once a day and for the log to be available for scrutiny by the receivers of the cargo at discharge ports. The instructions also required the temperatures of non-heated and non-cooled cargoes to be monitored for signs of abnormal activity such as polymerisation. There was no specific requirement for the temperatures of non-heated and non-cooled cargoes to be recorded. 1.14 IBC CODE SOLAS4 Chapter VII and MARPOL5 Annex II require chemical tankers built after 1 July 1986 to comply with the International Code for the construction and equipment of ships carrying dangerous chemicals in bulk and index of dangerous chemicals 4 International Convention for the Safety of Life at Sea, 1974, as amended. 5 International Convention for the Prevention of Pollution from Ships, 1973, as amended.

28 carried in bulk (IBC Code). In addition to prescribing design and construction standards, The IBC Code tabulates the minimum requirements for over 750 dangerous chemicals. References in the IBC Code to cargo temperature, inhibited cargoes, and cargoes not to be exposed to excessive heat, include: Chapter 7 – Cargo temperature control 7.1.5 Means shall be provided for measuring the cargo temperature. 7.1.5.4 When overheating or overcooling could result in a dangerous condition, an alarm system which monitors the cargo temperature shall be provided. (see also Operational Requirements in 16.6.) Chapter 15 – Special requirements 15.13 Cargoes protected by additives 15.13.3 Care shall be taken to ensure that these cargoes are sufficiently protected to prevent deleterious chemical change at all times during the voyage. Ships carrying such cargoes shall be provided with a certificate of protection from the manufacturer, and kept during the voyage, specifying:

1. the name and amount of additive present;
2. whether the additive is oxygen dependent;
3. date additive was put in the product and the duration of

effectiveness;

1. any temperature limitations qualifying the additives effective

lifetime; and

1. the action to be taken should the length of the voyage

exceed the effective lifetime of the additives. Chapter 16 – Operational requirements 16.6 Cargoes not to be exposed to excessive heat 16.6.1 Where the possibility exists of a dangerous reaction of a cargo, such as polymerization, decomposition, thermal instability or evolution of gas, resulting from local overheating of the cargo in either the tank or associated pipelines, such cargo shall be loaded and carried adequately segregated from other products whose temperature is sufficiently high to initiate a reaction of cargo (see 7.1.5.4). 16.6.2 Heating coils in tanks carrying this product shall be blanked off or secured by equivalent means.

29 1.15 SHIPPING INDUSTRY GUIDANCE 1.15.1 Tanker Safety Guide The Tanker Safety Guide (2014) (TSG), was published by the International Chamber of Shipping to promulgate good practice. With reference to styrene monomer, the guide stated that polymerisation was very often initiated by elevated temperatures, which could reduce the effectiveness of the inhibitor or reduce the inhibitor’s effective life. It also stated: These inhibitors are designed to be effective for a set period of time at a specified temperature. It is therefore essential that the timed effectiveness of the inhibitor is sufficient for the voyage and includes a good safety margin. Since elevated temperature can reduce the effectiveness of the inhibitor, or reduce its effective life, it is essential that heat sources are kept away from these cargoes and that the temperature is closely monitored on at least a daily basis, or more frequently if recommended by the cargo manufacturer. An increase in cargo temperature that is not related to ambient weather conditions or adjacent cargo temperatures may be an early indication that a polymerisation process has started. In such instances, the cargo manufacturers should be contacted immediately to advise appropriate counter measures which may include the addition of more inhibitor or the cooling of adjacent structures. Should the increase in temperature be rapid then the decision to jettison cargo may be the only option in order to avoid serious structural damage to the cargo tank and the ship. Regarding cargo stowage and handling, the TSG included: ●That heated cargoes will be stowed so as to be compatible with cargoes in adjacent tanks; ●That heated cargoes will be stowed so as not to be adjacent to heat sensitive cargoes or when a heat source could lead to a dangerous reaction; ●Some cargoes require an inhibitor to ensure that they remain chemically stable during transit. Such cargoes should not be stowed adjacent to heated cargoes; ●Some cargoes are liable to self react under certain conditions…The temperature of cargoes that may self-react should be closely monitored. Unexpected changes are an early indicator of a possible self-reaction. Should the temperature rise be in excess of what is expected, taking into account the ambient conditions and the temperature of adjacent cargoes, then this should be treated as an emergency and handled accordingly… 1.15.2 Chemical Distribution Institute Guidance The Chemical Tanker Operations for the STCW Advanced Training Course – A Practical Guide to Chemical Tanker Operations (2018) was produced by the Chemical Distribution Institute (CDI) to provide an aide-mémoire for crews and a study guide for advanced training in chemical operations.

30 Regarding the polymerisation and inhibited cargoes, among other things, the guide stated: Heat can initiate polymerization in reactive monomers, producing yet more heat in an exothermic reaction. Inhibited self reactive cargoes: should not be loaded adjacent to heated cargoes and heating systems will require isolating. 1.16 STORAGE AND HANDLING GUIDANCE 1.16.1 Styrene Monomer: Safe Handling Guide (2018) The Plastics Europe Styrene Monomer: Safe Handling Guide (2018)6 (the Guide) was produced by the Styrene Producers Association, members of which included Ineos Styrolution Europe GmbH and Shell Chemicals Europe BV. Ineos Styrolution provided copies of the safe handling guide to all its customers. The Guide was not referenced in Stolt’s VCR, nor was it passed to Stolt Groenland. The Guide referred to transportation of styrene monomer by road, rail, barge and sea. It included: Heated adjacent cargoes should be avoided, even if separated by cofferdams. Styrene should not be loaded into cargo tanks adjacent or corner-to-corner to a cargo having a temperature of 30°C (86°F) or higher even if separated by a cofferdam. Heating styrene to above this temperature will reduce shelf life and increase the risk of polymerisation in the cargo tank. For styrene storage ashore, Appendix 2 of the Guide included the following checks for polymerisation: ●Polymer content (< 10 ppm, depending on product spec.) ●Temperature (<< 2-3°C/day). If the temperature rises 1°C/day, it is advised to be alert and keep monitoring the temperature actively. Re-circulation could stop the temperature rise. A 2-3°C/day temperature increase is a typical indication of the onset of a runaway polymerization. The temperature needs to be monitored continuously. ●TBC levels (target >10 ppm wt). At temperatures below 15°C in the tank/ container, weekly sampling should be sufficient; above 25°C daily sampling is recommended. Normal TBC levels are between 10 and 15 ppm (for some applications higher concentrations are required). Below 10 ppm TBC polymer levels can slowly increase; below 4 ppm the TBC is not effective and accelerated polymerization will occur. ●Oxygen levels (3-8 volume % in the vapour phase). 6 Plastics Europe is a Pan-European trade association representing plastics manufacturers, of which the Styrene Producers Association was a sector group.

31 It explained that in certain circumstances it is possible for a runaway reaction to occur even if there is TBC and oxygen present in the styrene. These included: the presence of contaminants; the presence of rust in the tank; high (local) temperatures (>40°C); and non-homogeneous distribution of the TBC and oxygen in the tank contents. The guide also explained that TBC is not active long enough at high temperatures, since the reaction rate and therefore the depletion rate becomes too high. It included indicative depletion rates for TBC at varying temperatures (Table 4). Table 4: TBC depletion in styrene (if stored under air) Temperature Depletion Days/1ppm TBC Content TBC Content Shelf Life °C [Initial ppm] [end] Assured days 25 11 15 10 55 30 7 15 10 35 40 1.5 15 10 7 1.16.2 Safe handling sheets In addition to the Shell Chemicals Cargo Handling Sheet for styrene monomer that was referenced in Stolt Groenland’s VCR (see Section 1.9.1), similar guidance was provided by other styrene producers. Requirements for TBC concentrations and styrene temperatures contained in these guides are summarised in Table 5. Table 5: Comparison of manufacturers’ requirements for styrene monomer Producer Recommended TBC for storage Storage or carriage temperature Maximum temperature Remarks Americas Styrenics 10-15ppm Ambient None quoted Cooling recommended for storage where temperatures exceed 24ºC. Vessels containing styrene should be protected from external sources of heat. Shell Europe 10-20ppm Ambient 30ºC Not adjacent to heat. Adjacent maximum temp 35ºC. Lyondellbasell 10-15ppm Ambient 30ºC Styrene monomer should not be stored adjacent to tanks or compartments that are heated. Chevron Phillips 10ppm 24ºC 30ºC Stowed with respect to compatibility, heat adjacency and other requirements.

32 1.16.3 Safety data sheets SDSs are issued by chemical manufacturers in a harmonised format to provide generic safety information on products. SDSs are not specific to marine transport. The transport information provided in the SDS for styrene monomer (stabilised) issued by Ineos Styrolution did not include guidance regarding the bulk transport of the product by sea. 1.17 RECENT INSPECTIONS, AUDITS AND VETTING On 22 May 2019, Stolt Groenland was audited by a Stolt superintendent. The internal audit report noted that the SMS was well implemented, the senior officers were diligent, and that the tanker was very well maintained. Two non-conformities and 11 observations were recorded. The non-conformities concerned the absence of records of atmospheric checks when tank cleaning, and the recording of working hours. The observations were related to minor errors and omissions in documentation. On 6 July 2019, Stolt Groenland was inspected by a Port State Control officer in Long Beach, USA. No deficiencies were identified. On 6 August 2019, Stolt Groenland was vetted by a CDI inspector in Point Comfort, USA. One of the questions included in the CDI questionnaire was: Are officers aware of the documentation and handling requirements for cargoes and inhibitors, and if the cargo carried is required to be inhibited, is the required information available? No deficiencies were recorded, and the vetting report noted: Due to the various annex II cargoes carried on the vessel, current and previous cargo operations plans were checked and found in compliance with IBC and company procedures. And Observed the cargo handling and monitoring equipment on deck and in the CCR during cargo operations and found in good condition overall. 1.18 YEOMPO QUAY The port of Ulsan was administrated by the authority of the Ministry of Oceans and Fisheries. The berths were managed and operated by three different entities; the Ministry, Ulsan Port Authority and private companies. In the case of Yeompo Quay, berths 1 and 2 were operated and managed by Ulsan Port Authority while berth 3, where the accident occurred, was operated by the Ministry. Several berths and jetties in Ulsan harbour were specified for handling dangerous cargoes, and they were listed in the Ministry’s publication Safety Navigation Guide for Oil Tankers (Ulsan Port). The guide did not list Yeompo Quay as an authorised berth for handling dangerous cargo.

33 Stolt Groenland’s berth at Yeompo Quay had been arranged by the ship’s local agent who was given permission by the port authority to use it for ship-to-ship transfer operations. 1.19 STYRENE MONOMER ACCIDENTS AND INCIDENTS 1.19.1 Stolt Focus During the investigation into the causes and circumstances of the explosion on board Stolt Groenland, MAIB was made aware of a styrene monomer polymerisation incident that occurred on board Stolt Focus a couple of weeks earlier. Styrene loading Between 3 and 4 August 2019, the Cayman Islands registered chemical tanker, Stolt Focus, also operated by Stolt, loaded 4900t of styrene monomer into nine stainless steel cargo tanks at the LBC Terminal, Houston. The styrene was produced by Ineos Styrolutions and the product was loaded from ST B503. The temperature of the styrene monomer in the shore tank was 13.33°C and the TBC concentration was 11.7ppm. The observed temperatures the styrene shortly after it was loaded on board were between 16.6 and 21.5°C. The certificate of inhibitor issued by the attending ACA surveyor stated that 14 US gallons of liquid TBC had been divided between the cargo tanks and that the target TBC concentration was 21ppm. The effective days (60 – 90) and inhibitor statement shown on the certificate were identical to the certificate issued to Stolt Groenland (Figure 18). The certificate was signed by the surveyor and the tanker’s C/O. Passage Stolt Focus departed Houston on 12 August 2019 and arrived in Kobe on 13 September. At about this time elevated temperatures were discovered in some of the tanks containing styrene monomer and actions were taken to cool the cargo and contain the situation. The incident was not reported to the Flag State at the time. On 20 November 2019, Stolt informed the Cayman Islands Registry that Stolt Focus had experienced a problem with the styrene monomer cargo. An email to the Registry included: Stolt Focus observed a steady rise in temperature of the Styrene Monomer onboard, all our efforts to stabilize temperature rise failed, after consultation with our chemists it was decided to mix the Styrene Monomer with sea water to stop the reaction which was successful. As a result, we now have 4 tanks with this mixture which is now unpumpable and therefore we need to reduce viscosity by diluting with Benzine. SOLAS is not so clear on dilution of cargo’s and we would like your guidance to ensure we do correctly and meet all relevant regulations. Our initial plan was to carry out alongside in Malaysia which is now a no go as we are unable to get permission from Malaysian authorities.

34 Stolt proposed to dilute the styrene monomer with Benzine in international waters. The intended ratio facilitated the addition and mixing of the ‘short stop’ inhibitor7 phenothiazine (PTZ), which would stabilise the styrene to enable its incineration ashore. The Cayman Islands Registry approved Stolt’s proposal subject to several conditions concerning: risk assessments, the readiness and functioning of equipment, enhanced monitoring, and the notification of transfer and disposal intentions to coastal states. Request for further information Due to the co-incidence of the heating of styrene monomer cargoes on board Stolt Focus and Stolt Groenland (same supplier, same port of loading, same shore tank, similar timeframe, similar trading route), the MAIB requested Stolt to provide further information on the circumstances leading to the remedial action taken on board Stolt Focus. Stolt refused to do so, as it did not consider the heating and subsequent disposal of the styrene monomer cargo to be a marine accident or marine incident. A note on an emergency response team log sheet recorded during the Stolt Groenland accident stated ‘Same LOADING TERMINAL as S/FOCUS’. 1.19.2 Other maritime accidents In December 2019, the MAIB issued an interim report on the explosion and fire on board Stolt Groenland in which ship owners, ship and terminal operators, and individuals were requested to forward information on accidents or ‘near-misses’ involving the carriage of styrene monomer on board ships. Only one response was received. This concerned one of four tanks of styrene monomer loaded in Houston that was rejected after a high concentration of polymer was identified at the discharge port in China. The temperature of the styrene monomer reached 37°C but was cooled by running water on the deck. In addition, inhibitor was provided by the discharge terminal, and was added in the presence of a P & I surveyor. The polymerisation of the styrene was checked, and the cargo was eventually brought back within specification by a specialist chemical contractor. It was suspected that the polymerisation was caused by oxygen depletion in the styrene monomer. The vent stack from the styrene tank was next to a vent stack from a Hexene cargo that was nitrogen padded. As the nitrogen in the Hexene tank had to be topped-up by bottle through the vent stack, it was possible that the crew used the styrene vent stack in error and inadvertently introduced nitrogen into the styrene tank. Stolt did not report any previous accidents or incidents involving styrene monomer. 1.19.3 Shore-based accidents In the absence of a global chemicals accident database, the accidents resulting from runaway polymerisation of styrene monomer shown in Table 6, have been collated from several sources. 7 A ‘short stop’ is a free radical scavenger that can be added to a polymerising system to temporarily halt the polymerisation. When the short stop is consumed the polymerisation will continue.

35 Table 6: Land-based runaway styrene monomer polymerisation accidents since 1990 Date Location Injuries Fatalities 7 May 2020 Visakhapatnam, India

300

11 30 June 2005 Mesa, USA 0 1 8 April 2004 Jiangsu, China 0 1 2 April 2003 Addyston, USA 0 1 27 March 2000 Pasadena, USA 71 1 6 October 1999 Chiayi, Taiwan 1 0 23 June 1999 Pasadena, USA 21 2 27 June 1998 Channahon, USA 1 0 24 December 1998 Kanagawa, Japan 0 0 21 January 1998 Kaohsiung, Taiwan 4 0 26 January 1996 Chiayi, Taiwan 1 0 5 July 1994 Kaohsiung, Taiwan 0 1 The identified causes of the accidents varied, but insufficient temperature monitoring was a common contributor. 1.20 STYRENE SAMPLE AND OVERALL ANALYSIS RESULTS The sample of residue (Figure 8) and the overalls (Figure 9) recovered from Bow Dalian during the onsite investigation (see Section 1.4) were sent to LPD Lab Services Limited (LPD) for analysis. The laboratory used Fourier Transform Infrared Spectrometry (FTIR)8 and Pyrolysis gas chromatography–mass spectrometry (pyGC-MS)9 analysis to try to identify potential indications of causative factors in the explosion. The LPD report stated that the depletion of TBC is affected by heat, water and air, with heat being the most important. LPD explained that additional inhibitor should be added when inhibitor levels drop below 10 ppm, and that normal levels are 10-15 ppm, but some customers require up to 60 ppm. The laboratory analysis report found that: ●The plastic blob was confirmed as polystyrene or a polystyrene based material. ●There were some indications of the occurrence of dimers and other products of styrene, that would be indicative of an uncontrolled polymerisation process having occurred. 8 FTIR is a sensitive technique particularly for identifying organic chemicals in a whole range of situations including solid, liquid and gas samples. FTIR can also be used to characterise some inorganic compounds. 9 pyGC-MS is a method of chemical analysis in which the sample is heated to decomposition to produce smaller molecules that are separated by gas chromatography and detected using mass spectrometry.

36 ●There is no TBC detectable in the polymer blob material. This is expected and is further indication of uncontrolled polymerisation as it is expected TBC is consumed prior to the ‘runaway’ polymerisation process. ●Styrene was identified on the overalls. This was observed in all areas but appeared more intense in areas of contamination. ●In high contamination areas (contaminated overalls) TBC was identified. ●If it is assumed that the styrene contamination of the overalls took place at the same time of the polymerization reaction and explosion was observed, then it can be stated that inhibitor was present, and detectable. ●Whilst quantification was not undertaken in this investigation, the amount of TBC in the overalls is likely to be significantly above the detection level of trace contaminants (typically between 1-10ppm). Therefore, it is unlikely that the absence of TBC was key in the explosion. LPD found no other clear indication of any initiator compounds associated with the chemical initiation of polymerisation. Based on the results of its analysis, LPD concluded that: At this point in the investigation it is thought that rather than being associated with a lack of TBC, which was clearly present from being found on the overalls, the runaway styrene polymerisation reaction is likely to be associated with insufficient dissolved oxygen during storage, perhaps from poor venting or mixing. So, the TBC was unable to control the rate of polymerisation. The heat removal would then be the driving factor in the polymerisation reaction promoting reaction acceleration or thermal runaway.

37 SECTION 2 – ANALYSIS 2.1 AIM The purpose of the analysis is to determine the contributory causes and circumstances of the accident as a basis for making recommendations to prevent similar accidents occurring in the future. 2.2 EXPLOSION AND FIRE MECHANISMS It was evident, from the exponential temperature gradient for 9S recorded on the VDR (Figure 17), that the explosion and fire on board Stolt Groenland resulted from the runaway polymerisation of the styrene monomer cargo. The sudden release of large quantities of vapour from the 9S p/v valve (Figure 3), the activation of the high-level and high-high-level alarms on the bridge front in quick succession, and the rapid increase in the pressure within 9S that was seen by the C/O and 3/O in the CCR, support this. The runaway polymerisation resulted in the rapid volumetric expansion of the styrene monomer, which increased the pressure within 9S at a rate that exceeded the venting capacity of the tank’s p/v valve. As a result, the tank ruptured through the main deck and the common bulkhead with the empty cargo tank 9C (Figures 11 and 12). The styrene monomer vapour ignited almost immediately following the rupture. As the temperature of the cargo in 9S was well below the auto-ignition point for styrene monomer when the explosion occurred, ignition of the vapour must have been caused by an external source such as sparks created by static electricity or metal to metal contact during tank rupture. 2.3 INITIATION OF POLYMERISATION Styrene monomer polymerisation is initiated by heat, the lack of dissolved oxygen, the lack of an inhibitor, or contact with free-radical initiators such as peroxides. To mitigate against these potential initiators, the styrene monomer carried on board Stolt Groenland had been: ● Loaded under air, which significantly reduced the likelihood of oxygen depletion. ● Inhibited with a quantity of TBC that was intended to remain effective for a minimum of 60 days given the cargo volume and the ambient temperatures anticipated during the voyage. ● Stored in stainless steel tanks that had been wall wash tested and visually inspected, which reduced the risk of contamination. ● Sample tested on completion of loading (Figure 19) and found to be within specification with the peroxide concentration of 6ppm being well below the allowable maximum of 50ppm.

38 ● Stowed in tanks that were not adjacent to tanks10 that contained cargoes that were required to be heated, and for which the heating coils had been blanked. However, despite these precautions, runaway polymerisation occurred. The styrene monomer carried in 9S had been heated to levels well above the ambient temperatures experienced during the voyage. The VDR data (Figure 17) showed that the temperature of the styrene monomer stored in 9S exceeded 37°C during the 29-day period between 30 August and 28 September. The styrene monomer loaded in Houston had been maintained at 13°C in a shore tank; by the end the loading process on 8 August, its temperature had increased by about 4°C. Given that the ambient temperature in Houston was about 30°C and cargoes loaded into the tanks adjacent to 9S all exceeded 30°C, it is likely that the temperature of the styrene monomer would have reached or been higher than the maximum temperature of 29.4°C (85°F) specified on the certificate of inhibitor (Figure 18) when the vessel departed Houston on 17 August, and therefore during the whole voyage. Although unlikely to have been the single cause of the runaway polymerisation event, the elevated temperatures significantly increased the rate of TBC depletion and therefore the risk of polymerisation. It was therefore apparent that the elevated temperature of the cargo in 9S was the key factor in this accident. 2.4 HEAT TRANSFER As the air and sea temperatures in the Houston area, and during the tanker’s passage to and through the Panama Canal, were about 30°C, it was inevitable that the temperature of the styrene monomer and the other ‘ambient’ cargoes would rise accordingly. Any further increase of cargo temperatures required additional heating. The only source for the heating of Stolt Groenland’s ‘ambient’ cargoes was the heated cargoes, particularly the HMD in 8P to 11P, which was 61°C when loaded. The VDR (Figure 17) showed that the HMD temperature was being maintained between 45°C and 55°C by the periodic use of the vessel’s cargo heating system. It also showed that, due to the effects of heat transfer, the temperatures of the cargoes in the adjacent centre tanks (8C to 11C) had settled at between 40ºC to 44ºC, and the corresponding starboard side tanks (8S to 11S) between 36ºC and 38ºC. That the HMD was loaded at such a high temperature and later heated above its specified maximum carriage temperature of 50°C, might not have affected its quality, but it would have exacerbated the levels of heat transfer. The rate and degree of the heat transfer from the HMD cargo to the styrene monomer in 9S would have been influenced by the relatively low specific heat capacity (SHC) of the adiponitrile in 9C and that of the styrene itself (Table 7). SHC is the amount of heat energy required to raise the temperature of a substance per unit of mass. A substance with a low SHC will heat up (and cool down) more readily than a substance with a high SHC. In this case, the adiponitrile and styrene would have readily heated, and avenues for heat loss through adjacent tanks, the empty ballast tanks below, and the sun-exposed deck above were limited. 10 It should be noted that, contrary to guidance contained in the Plastics Europe’s Styrene Monomer: Safe Handling Guide, 6C (which contained styrene monomer) was corner to corner with 7S (which contained a heated cargo).

39 Table 7: The specific heat capacity of various materials at 20ºC Material Specific heat capacity (J/kgK) Water 4200 Glycerol 2400 Acrylonitrile 2090 Styrene 1730 Sodium (liquid) 1230 Adiponitrile 1190 Copper 0.386 2.5 POLYMERISATION INHIBITOR MIXING The TBC concentration was calculated and not measured. It was then poured into the open tanks and the styrene monomer loaded on top of it with no means of mechanical mixing. Poor distribution of the TBC and oxygen in the styrene monomer can lead to low concentrations at certain zones in the tank, which could lead to runaway zones in the bulk contents. The methods of adding and mixing the inhibitor on board Stolt Groenland were rudimentary and relied heavily on vessel and cargo movement during the voyage. Although the styrene monomer was loaded and its inhibitor added 10 days before the vessel departed Houston, these were tried and tested industry practices and, given the levels of TBC found in 6P and 6C after the accident, it is less likely to have been a factor. 2.6 POLYMERISATION INHIBITOR DEPLETION The target TBC concentration on loading was 17ppm. This was based on the amount of TBC required to ensure it remained effective for the duration of the voyage i.e. above 10ppm on arrival. The certificate of inhibitor issued on behalf of American Cargo Assurance indicated that the TBC should have remained effective for a period of 60 to 90 days at a recommended temperature range between 15.6ºC and 29.4ºC (60ºF and 85ºF). The duration between loading the styrene monomer on board Stolt Groenland in Houston and its planned discharge in An-Ping, Taiwan, was 56 days. Using the indicative depletion rates provided by Plastics Europe in its Styrene Monomer: Safe Handling Guide (Table 4): ● At a temperature of 25ºC, the TBC would have depleted at a rate of approximately 1ppm every 11 days. At that rate, it would have taken 77 days for TBC levels to drop from 17ppm to 10ppm (Table 8). ● At a temperature of 30ºC, the TBC would have depleted at a rate of 1ppm every 7 days and therefore could have dropped to 10ppm within 49 days. This was less than the duration the cargo was planned to be carried on board. ● At 37ºC, the TBC depletion rate would have been significantly higher.

40 Table 8: Calculated TBC depletion using figures provided in the Plastics Europe Styrene Monomer: Safe Handling Guide (2018) Ambient Temperature ºC Depletion rate (Days/1ppm) Effective TBC content (17ppm – 10ppm) Assured number of effective days 25ºC 11 7 77 30°C 7 7 49 40ºC 1.5 7 10.5 Tests carried out on the samples taken from 6P and 6C after the accident found the TBC content in each tank to be 8ppm and 7ppm respectively. As the temperature of the styrene monomer in 9s was 5 to 6°C higher than that in 6P and 6C, the depletion rate of its TBC would have been significantly higher and therefore, its concentration level on arrival Ulsan much lower. TBC inhibitor is ineffective below 4ppm and at 40°C this level could be reached within 21 days. The laboratory tests conducted by LPD (Section 1.20) identified traces of TBC on the overalls worn by a crewmember who was working on Bow Dalian’s deck when vapour from 9S cargo tank was venting from its p/v valve. This suggests that TBC was still present in the styrene monomer at the time of the explosion. However, given the prolonged elevated temperature of the cargo in 9S, it is highly likely that the TBC had become ineffective at some point during the voyage and probably before Stolt Groenland arrived at Ulsan. The elevated temperature of the styrene monomer in 9S and the reduced level of TBC, would have increased the likelihood of dissolved oxygen depletion in the styrene monomer and, consequently, increased the rate of polymerisation. As polymerisation is an exothermic reaction, the styrene monomer temperature continued to rise, and on 27 September it reached 65°C (Figure 20), the threshold for runaway polymerisation – the point at which the reaction could no longer be arrested. 2.7 CARGO STOWAGE The IBC Code requires cargoes that present a risk of a dangerous reaction due to localised overheating heating, to be loaded and carried adequately segregated from other products whose temperature is sufficiently high to initiate a reaction of cargo. Reference to this requirement was made in Stolt Groenland’s VCR, which also emphasised that ‘heated cargoes will be stowed so as not to be adjacent to heat sensitive cargoes or when a heat source could lead to a dangerous reaction’. Shell’s cargo handling sheet for styrene monomer stated that the maximum temperature of the cargo in adjacent tanks should not exceed 35°C. Similar guidance was provided by other industry bodies. The Styrene Monomer: Safe Handling Guide went further and stated that ‘styrene should not be loaded into cargo tanks adjacent or corner-to- corner to a cargo having a temperature of 30°C (86°F) or higher even if separated by a cofferdam’.

41 Figure 20: 9S Styrene Monomer Temperature 20-28 September Lin e gra ph sh ows ta nk te mpe ra tur es be twe en 16 an d 2 7 Sep tem be r ( UTC ), he ati ng co il va lve s o pe ned an d ris e o f sty ren e mon ome r tem per at ure ov er 3 day s 30.0 40.0 50.0 60.0 Temperature ºC 70.0 Tank temperatures 16 to 27 September (UTC) 6P-16 6C-17 6S-18 7P-19 7C-20 7S-21 8P-22 8C-23 8S-24 9P-25 9C-26 9S-27 10P-28 10C-29 10S-30 11P-31 11C-32 11C-32 9S – styrene monomer >1°C temp rise each day 65ºC Heating coil valves opened to 7S, 8P, 9P, 10 and 11P 16 17 18 19 20 21 22 23 24 25 26 27 September Kobe, Japan Ulsan OTK YMP

42 Some industry publications focussed on the need to physically segregate heated cargo from heat sensitive cargo rather than specifying maximum temperature limits for adjacent tanks. This approach did not take into consideration the risk of heat transfer from tank to tank across a cargo stow. The HMD in tanks 8P to 11P was clearly heated cargo and its temperature was maintained using the vessel’s cargo heating system. The cargoes in tanks 8C to 11C were not considered to be heated cargo; however, because of heat transfer, they were being carried at temperatures well in excess of the limits set out by Shell and Plastics Europe. It was entirely foreseeable, and could not have been unexpected, that the styrene monomer cargo in 9S would be heated by the cargo in 9C, which itself been had been heated by the HMD in 9P. The likelihood of elevated temperatures due to heat transfer was acknowledged in the TSG, which stated that ‘an increase in cargo temperature that is not related to ambient weather conditions or adjacent cargo temperatures may be an early indication that a polymerisation process has started’. The polymerisation of the styrene monomer on board Stolt Groenland indicates that the precaution of not stowing heat sensitive chemicals immediately next to heated cargoes is not always sufficient in meeting the ‘adequate’ segregation requirement of the IBC Code (see Section 1.14). The likelihood and extent of heat transfer from heated cargoes to heat sensitive cargoes via intermediate tanks must also be accurately determined. The stowage of the styrene monomer on board Stolt Groenland was planned by an experienced ship operator using the iStow programme, and appeared to comply with the applicable regulatory and onboard requirements and to align with industry guidance. The tanker’s C/O then scrutinised the proposed plan before it was approved by the master. However, although the maximum temperature of the styrene monomer cargoes, 85°F or 29.4°C, was stated on the certificate of inhibitor, no calculations were undertaken during the planning of the stowage to ensure that heating from adjacent tanks did not lead to this being exceeded. Instead, the ship operator, and the tanker’s master and C/O relied on their experience. This stemmed from several factors, including: ● The calculation of heat transfer was complex due to the differing SHCs, tank construction, tank levels, stowage durations, and other variables. Consequently, it was outside usual practice. ● The complexity of the heat transfer calculation was such that it was outside the scope of the iStow software. ● The master and chief officer’s knowledge of heat transfer was limited to the syllabus of the STCW training. They had also not previously encountered any difficulties when transporting styrene monomer. The instructions and guidance that were available in the IBC Code, TSG and the CDI guide were clear that heat sensitive and heated cargoes should not be stowed in adjacent tanks. However, little or no emphasis was placed on industry guidance, the potential for heat transfer through intermediate tanks, or how such transfer was predicted. Instead, reliance was placed on temperature monitoring.

43 2.8 TEMPERATURE MONITORING Stolt Groenland’s SMS required the temperatures of all non-heated cargoes, which included the styrene monomer in 9S, and the adiponitrile in 9C, to be monitored daily. To assist in meeting these requirements and to comply with the IBC Code, remote temperature measurement and temperature alarms were available via the Ariston cargo monitoring system. That Stolt Groenland’s crew were unaware of the elevated temperatures and polymerisation of the styrene monomer in 9S, until warned by the activation of the high-level alarm on the bridge front, indicates that the temperatures of the non-heated cargoes were not monitored at all, and that no temperature alarms were set. The crew also either did not notice, or did not recognise the significance of the elevated temperatures of the cargoes discharged in Kobe and Ulsan (Table 2), notably 8C (43.6°C), 8S (39.1°C), 11C (48.2°C), 9C (48.8°C), 10C (46.0°C) and 11S (38.4°C). In Ulsan, such oversight might have been due to competing demands encountered by the off-going and incoming C/Os when handing over during cargo operations. The absence of temperature monitoring of the non-heated cargoes stemmed from the crew’s view that styrene monomer was a benign cargo when inhibited, and that no previous problems or difficulties with its carriage had been experienced. Consequently, whereas the temperature of the heated cargoes was monitored and maintained, primarily for reasons of quality, the styrene monomer and the other non-heated cargoes that were transported at ‘ambient’ were less of a concern, even if heat sensitive. Had the temperatures of the styrene monomer been closely monitored and the maximum temperature of 85°F (29.4°C) stated on the certificate of inhibitor been adhered to, this accident would have been avoided. Concerns should have been raised and the charterer alerted well before Stolt Groenland entered the ports in Kobe and Ulsan, and maybe even before it left Houston. The charterer should have been informed once the temperature of the styrene monomer reached 30°C; this threshold was probably crossed before the vessel left Houston. The next opportunity to alert the charter and take urgent action was missed when the cargo temperature in 9S increased by 1°C on three consecutive days and 3°C in 24 hours (Figure 20). These two events occurred on 24 September while the vessel was in Kobe and 4 days before the explosion. The runaway temperature of 65°C was reached at about 1000 on 27 September, shortly before Stolt Groenland moved to the Yeompo quay. 2.9 ACTION IN THE EVENT OF POLYMERISATION The arrest of the polymerisation of styrene cargo in four of Stolt Focus’s cargo tanks was achieved ultimately by mixing the styrene with sea water (see Section 1.19.1). This action was successful, but it raises concerns over the options available to tanker crews encountering elevated styrene monomer temperatures. The certificate of inhibitor for the styrene monomer issued to Stolt Groenland on completion of loading (Figure 18) stated that if the maximum temperatures were exceeded then the cargo should be monitored for the inhibitor level and polymerisation, and additional inhibitor added as required. However; ● Stolt Goenland’s crew had no means to test for TBC concentration.

44 ● No means were available on board the tanker to test for polymers. ● The charterers did not provide additional TBC for the voyage. ● There were no means available to mix additional TBC with the styrene monomer. As a result, even if Stolt Groenland’s crew had identified the elevated temperatures in 9S. 6P and 6C, the actions stated on the certificate of inhibitor were not viable. In most cases, beyond monitoring tank and adjacent tank temperatures, and monitoring oxygen levels where a nitrogen blanket is used, the ability of tanker crews to prevent or arrest polymerisation of styrene monomer is limited to reducing tank temperatures either by cooling the tank exterior, or by flooding the tank with water. Therefore, it is important that the remedial actions to be taken, as listed on the certificate of inhibitor and associated guidance, are achievable and focused on sea transportation rather than shore-based processes. For reasons of expertise and practicability, this will probably be limited to seeking advice from cargo manufacturers or chemists ashore. 2.10 FIRE-FIGHTING The immediate reactions of Stolt Groenland’s and Bow Dalian’s crews to supress the fire with fixed foam appliance, and to abandon their respective vessels, were positive and assisted in ensuring their own safety. The fire on board Stolt Groenland burned for 5½ hours, during which time it spread from the manifold area to the accommodation block. The path of the fire spread was not clear during the on-site inspection, as the accommodation front was relatively free from significant damage. It is possible that the spread into the accommodation resulted from flammable styrene vapour entering from either the gas detection or cargo tank heating systems. The actions of the local emergency services were also timely and considerable given that Yeompo Quay was a general cargo berth and, unlike chemical and oil product terminals, there were no enhanced fire-fighting arrangements or emergency response procedures in place. Given the berth’s proximity to Ulsan and the Ulsan Bridge (Figures 1 and 6), and the toxic smoke that was emitted from the explosion and fire, the potential consequences to the surrounding environment were severe. 2.11 STOLT FOCUS INCIDENT The polymerisation incident that occurred on board Stolt Focus a couple of weeks earlier (Section 1.19.1) was of particular interest to the investigation as the two vessels had both loaded styrene monomer cargoes produced by the same supplier, in the same port, from the same shore tank during a similar timeframe. Furthermore, the voyage routes and therefore environmental conditions were similar. A key difference was that the target concentration for the TBC on board Stolt Focus was significantly higher (21ppm rather than 17ppm). This might have been due to the planned carriage time, however, the calculated effective days documented on the certificate of inhibitor was the same for both vessels (60 – 90 days). The cargo

45 stowage arrangements, the cause and magnitude of the elevated temperatures, and the circumstances that led to the elevated temperatures being identified on board Stolt Focus are unknown and therefore comparisons cannot be made. It was disappointing that Stolt would not provide any details of the Stolt Focus incident, particularly as the company had apparently not experienced any other similar accidents or near misses while transporting styrene monomer. It was also surprising that the company did not consider the occurrence to be a marine incident11 as defined in the IMO Casualty Investigation Code12. It is evident from the accident on board Stolt Groenland and the number of major styrene monomer explosions that have occurred on land, that the incident on board Stolt Focus, if not corrected would have endangered the safety of the ship, its occupants, and any other persons nearby. Therefore, in accordance with the Flag State regulations13, it should have been reported. In not reporting or informing other ships’ masters of the Stolt Focus marine incident, an opportunity to avoid the explosion in Ulsan was lost. In not sharing the circumstances of the incident with this investigation, an opportunity for the wider industry to learn more lessons from the near miss has been lost. Regardless of any commercial sensitivities, it would be beneficial to future safety if Stolt would inform the chemical tanker industry of any additional lessons it has learnt from its styrene monomer polymerisation occurrences. 11 A marine incident means an event, or sequence of events, other than a marine casualty, which has occurred directly in connection with the operations of a ship that endangered, or, if not corrected, would endanger the safety of the ship, its occupants or any other person or the environment. 12 Code of International Standards and Recommended Practices for a Safety Investigation into a Marine Casualty or Marine Incident. 13 Cayman Islands’ The Merchant Shipping Law (2016 Revision) – The Merchant Shipping (Marine Casualty Reporting and Investigation) Regulations, 2018.

46 SECTION 3 – CONCLUSIONS 3.1 SAFETY ISSUES DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS 1. The explosion and fire on board Stolt Groenland resulted from the runaway polymerisation of the styrene monomer cargo in cargo tank 9S. [2.2] 2. The polymerisation of the styrene monomer in 9S was initiated by the cargo’s elevated temperature for much of the voyage, which reduced the effectiveness of the TBC inhibitor. [2.3, 2.6] 3. The elevated temperature of the styrene monomer in 9S resulted from the transfer of heat from the HMD cargo in the port side tanks, via the adiponitrile cargo in 9C. [2.4] 4. The precaution of not stowing the styrene monomer next to the heated HMD cargo was not sufficient in meeting the adequate segregation requirement of the IBC Code. [2.7] 5. The probability of heat being transferred from the HMD cargo tanks to the styrene monomer cargo was not fully considered during the planning and approval of the cargo stowage. [2.7] 6. Calculations to predict heat transfer during cargo stowage planning were not conducted because they were complex and outside the capabilities of the ship operator and the tanker’s crew. They were also outside the scope of the cargo stowage software. [2.7] 7. Instructions and guidance were clear that inhibited cargoes should not be stowed adjacent to heated cargoes, but the likelihood of heat transfer through adjacent or intermediate cargo tanks was not covered in detail. [2.7] 8. Despite being a requirement in Stolt Groenland’s SMS, the temperature of the styrene monomer was not monitored, and the temperature alarms available on the cargo monitoring system were not set. The crew also either did not notice, or did not recognise the significance of, the elevated temperatures of the cargoes discharged in Kobe and Ulsan. [2.8] 9. The absence of temperature monitoring of the styrene monomer was influenced by the crew’s view that it was a benign cargo when inhibited, and that no previous problems or difficulties with its carriage had been experienced. [2.8]

47 3.2 SAFETY ISSUES NOT DIRECTLY CONTRIBUTING TO THE ACCIDENT THAT HAVE BEEN ADDRESSED OR RESULTED IN RECOMMENDATIONS 1. The actions to be taken on encountering elevated temperatures in the styrene monomer cargo on board Stolt Groenland, which were stated on the certificate of inhibitor, were not viable. The crew had no means of testing for TBC concentration or polymers, and no additional TBC was carried. [2.9] 2. Given the proximity of Yeomp Quay to Ulsan and the Ulsan Bridge, and the toxic smoke that was emitted from the explosion and fire on board, the potential consequences to the surrounding environment were severe. [2.10] 3.3 OTHER SAFETY ISSUES NOT DIRECTLY CONTRIBUTING TO THE ACCIDENT 1. In not reporting or alerting other tanker masters to a similar polymerisation incident that occurred 2 weeks earlier, an opportunity was lost to prevent this accident. [2.11]

48 SECTION 4 – ACTION TAKEN 4.1 MAIB ACTIONS The MAIB has issued an interim report that alerted the chemical tanker industry to the circumstances of the accident and requested information about previous similar accidents or incidents. 4.2 ACTIONS TAKEN BY OTHER ORGANISATIONS Stolt Tankers B.V. has:

● Issued fleet notices highlighting issues with inhibited cargo carriage.

● Introduced a requirement for ships to report cargo temperatures to the ship operator ashore three times per week.

● Revised procedures for the setting of cargo-related alarms.

● Enhanced procedures for the checking of cargo stowages that contain both heated and inhibited cargoes.

● Commenced a concerted campaign through its marine compliance officers to raise crew awareness of inhibited cargoes.

● Started to investigate the possibility of upgrading its cargo stowage software to calculate and predict heat transfer within a stowage containing both heated and heat-sensitive cargoes.

● Started to work with manufacturers and industry bodies to ensure that cargo- handling instructions and protocols for inhibited cargoes are practical and achievable.

● Investigated the use of polymer testing equipment on board ships. Unfortunately it was not considered a viable prospect with current technology. The Ministry of Oceans and Fisheries, Republic of Korea has prohibited ship-to-ship transfer operations for dangerous cargo on general cargo berths in Ulsan. The International Chamber of Shipping has amended the Tanker Safety Guide (Chemicals), published January 2021, section 6.3.4 Stowage planning, to make it clear that:

● The stowage of heated and inhibited cargoes can result in a dynamic situation in which the degree of heat transfer may be complex and difficult to predict.

● One tank separation between heated and heat sensitive cargoes might not be sufficient. The Chemical Distribution Institute has instigated the production of a best practice paper regarding the dosing of ship cargo tanks which will address, among other things, inhibitor quantity validation.

49 SECTION 5 – RECOMMENDATIONS Cayman Island Shipping Registry, through the UK as the Member Government for the Red Ensign Group to the International Maritime Organization, is recommended to: 2021/117 Propose to the IMO a revision to Section 15.13 of the IBC Code to:

●Include in the certificate of protection the actions to be taken in the event of a cargo falling outside of the manufacturer’s specified oxygen and temperature limits, and that

●Any actions should be realistic, taking account of the limitations on board ships regarding the monitoring, adding, and mixing of inhibitor during the voyage. International Chamber of Shipping is recommended to: 2021/118 Promulgate this report to its members. INTERTANKO is recommended to: 2021/119 Promulgate this report to its members. Chemical Distribution Institute is recommended to: 2021/120 Amend its publication ‘Chemical Tanker Operations for the STCW Advanced Training Course – A Practical Guide to Chemical Tanker Operations’ to make it clear that:

●The stowage of heated and inhibited cargoes can result in a dynamic situation in which the degree of heat transfer may be complex and difficult to predict.

●One tank separation between heated and heat sensitive cargoes might not be sufficient.

●Promulgate this report to its members. Plastics Europe (Styrene Producers Association) is recommended to: 2021/121 Work with its members to incorporate the lessons learned from this accident in its Styrene Monomer: Safe Handling Guide. Stolt Tankers B.V. is recommended to: 2021/122 Share with INTERTANKO the circumstances and lessons learned from the Stolt Focus incident and the results of its research into improved stowage software, to enable prediction of heat transfer and cargo behaviour. Safety recommendations shall in no case create a presumption of blame or liability.

Marine Accident Report

FD06070 ｜ 2021_英国_crush-incident-on-scallop-dredger-olivia-jean-with-loss-of-1-life__2021-05-OliviaJean-Anne.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2021 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A Crew induction proforma

OLIVIA JEAN TN 35 www.safetyfolder.co.uk Crew Induction form The location and use of safety equipment and fire equipment: Liferaft Location Life jackets & lights & whistles Lifebuoy and light Pyrotechnics EPIRB First aid equipment Portable fire equipment Fixed fire equipment & detection Deck / fire pump & hose/s Vent closures / fuel shut offs Fire buckets with lanyards Wear and use of PFD`s. The routine operation, procedure or general knowledge of: Windlass / anchoring Preparing vessel for sea Watch keeping responsibilities LPG system Berthing mooring lines Fishing equipment & safety Helm / engine controls & pilot Electrical & emergency electrical Communication equipment Refuelling Bilge & tank pumping Hatches and closures Garbage / oil / sewage disposal Record keeping Incident reporting Navigational equipment The response in the event of an emergency: General alarm signal Medical / injury Fire on board / E/R alarms Escape routes Person overboard Abandon vessel / remain with vessel Flooding / grounding / bilge alarm Fixed fire equipment & detection Deck / fire pump & hose/s Vent closures / fuel shut offs Metal buckets with lanyards Acknowledgement of instruction provided by the master / owner and received by the crew for those items initialled above Name of Crew Signature & Date Name of Owner / Skipper Signature & Date Name of Emergency Contact for crew Phone 12 / 35

Annex B Olivia Jean’s risk assessments

Annex C MAIB Safety Flyer to the Fishing Industry

SAFETY FLYER TO THE FISHING INDUSTRY Fatal accident to a crewman on board the scallop dredger Olivia Jean (TN35), north-east of Aberdeen, Scotland on 28 June 2019 Narrative At about 2200 on 28 June 2019, an Indonesian crewman on board the scallop dredger Olivia Jean was fatally injured after being struck on the head by one of the vessel’s scallop dredge towing bars while working on deck. The crewman had replaced two worn dredges on the towing bar and stood clear as the skipper used the winches and derrick to lift and realign the gear against the vessel’s tipping door. Unfortunately, one of the towing bar’s securing chains had not been released and the dredge gear became snagged. Although the skipper shouted instructions to the crewman to remain clear as he attempted to free the gear, the crewman stepped between the snagged bar and the accommodation superstructure just as the snagged bar released and swung inboard. The crewman suffered crush injuries to his head and was airlifted from the vessel and taken to hospital for emergency treatment. He died 12 days later. The dredge gear winches and the two crew members working on deck were being controlled by the skipper from the wheelhouse. The location of the accident was out of the skipper’s line of site and he was reliant on a CCTV screen that was positioned behind him to monitor the area. The deck crew’s level of English comprehension was poor, and they did not speak a common language. Olivia Jean

In August 2020, a British crewman on the Olivia Jean was struck by a towing bar during a dredge gear shooting operation. He moved out of the designated safe zone before being given the clearance to leave, stepped into an unsafe area and was struck by the towing bar when it unexpectedly moved. The injured fisherman was taken to hospital for emergency treatment. Safety lessons

1. The shooting, hauling, and moving of dredge gear are extremely hazardous activities that need

to be tightly controlled. No matter how experienced the crew on deck might be, they need to be properly supervised and their safety closely monitored.

1. The crew on deck may not be visible to the winch or machinery operator, so clear, explicit and

easily understood communications are vital to the avoidance of dangerous situations from developing. This is particularly the case where the crew do not share the same first language.

1. Learning lessons from previous accidents can prevent injuries and save lives. In this case, the

lessons learned and corrective action implemented did not prevent a near identical accident from occurring to the crew on board in August 2020. It was very fortunate that it did not result in another fatality. This flyer and the MAIB’s investigation report are posted on our website: www.gov.uk/maib For all enquiries: Marine Accident Investigation Branch First Floor, Spring Place 105 Commercial Road Southampton SO15 1GH Email: maib@dft.gov.uk Tel: +44 (0)23 8039 5500 Publication date: May 2021 Extract from The United Kingdom Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 – Regulation 5: “The sole objective of the investigation of an accident under the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents through the ascertainment of its causes and circumstances. It shall not be the purpose of an such investigation to determine liability nor, except so far as is necessary to achieve its objective, to apportion blame.” NOTE This safety flyer is not written with litigation in mind and, pursuant to Regulation 14(14) of the Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to attribute or apportion liability or blame. © Crown copyright, 2021 You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned.

FD06071 ｜ 2022_英国_person-overboard-from-motor-cruiser-diamond-emblem-1-with-loss-of-1-life__2022-5-DiamondEm.pdf

分类：06_境外及港澳台调查报告 ｜ 年份：2022 ｜ 地域： ｜ 严重程度： ｜ 场所： ｜ 起火原因： ｜ 来源格式：pdf

Annex A Funnell 42 Sundeck Ferry Emblem Fleet Declaration of Conformity

Declaration of Conformity

Manufacturer Ferry Marina Ferry Road Horning Norfolk NR12 8PS

Description of the Craft Brand Name:

Funnell 42 Sundeck Type:

Fully decked motorboat Type of Hull:

GRP low wash hull Design Category:

D Propulsion type:

Inboard diesel engine Installed engine power:

32kw Maximum engine power: 45kw Length:

1. 8m

Beam:

1. 66m

Craft Identification Number:

This declaration of conformity is issued under the sole responsibility of the manufacturer. I declare on behalf of the craft manufacturer that the craft mentioned above complies with all applicable essential requirements in the way specified.

Name: ____________________________ Position: ____________________

Signature: ___________________________________

Documentation and Standards used to comply with essential safety requirements 2.1 CIN

EN ISO 8666:2002 2.2 Builders plate

EN ISO 10087:2006 2.3 Protection from falling overboard

Empirical test data 2.4 Visibility from steering position

Technical manual pg 1and Appendix K 2.5 Owners manual

BS EN ISO 10240:2004 3.1 Structure

Empirical data 3.2 Stability and freeboard

Trial data following MCA code of practice 3.3 Buoyancy and flotation

Trial data 3.4 Openings

BS EN ISO 12216:2002 and CE marked doors and windows 3.5 Flooding

Technical manual Appendix B 3.6 Maximum recommended load

Technical manual pg 4 3.7 Liferaft stowage

Technical manual pg 5 3.8 Escape

Technical manual pg 5 and Appendix B 3.9 Anchoring, mooring and towing

BS EN ISO 1584:2003 4.0 Handling Characteristics

Trial data in Appendix C 5.1 Inboard engine, ventilation and exposed parts Technical manual pg 5 & 6 5.2 Fuel system and fuel tanks

BS EN ISO 10088:2001 and Appendix F 5.3 Electrical systems

BS EN ISO 10133:2001 5.4 Steering systems

Technical manual pg 6 5.5 Gas

Technical manual pg 7 and Appendix A 5.6 Fire protection

Technical manual pg 7 and Appendix B 5.7 Navigation lights

Technical manual pg 8 and Appendix D 5.8 Discharge prevention

Technical manual pg 8

Annex B Funnell 42 Sundeck Ferry Emblem Fleet Technical Manual

Funnell 42 Sundeck Ferry Emblem Fleet

Technical Manual Page 1 of 9

INDEX

Boat Design Category 2.1 CIN 2.2 Builders Plate 2.3 Protection from Falling Overboard 2.4 Visibility From Main Steering Position 2.5 Owners Manual 3.1 Structure 3.2 Stability and Freeboard 3.4 Openings in Hull, Deck and Superstructure 3.5 Flooding 3.6 Recommended Load 3.7 Life-raft Stowage 3.8 Escape 3.9 Anchoring, Mooring, Towing 4 Handling Characteristics 5.1 Engine, Engine Ventilation, Exposed parts 5.2 Fuel System 5.3 Electrical System 5.4 Steering System 5.5 Gas System 5.6 Fire Protection 5.7 Navigation Lights 5.8 Discharge Prevention 5.9 Toilet System 5.10 Water System Appendix A Gas system Appendix B Openings, downflooding heights, escape, fire protection Appendix C Handling Appendix D Electrical diagram Appendix E Steering system Appendix F Fuel System Appendix G Toilet system and water system Appendix H Standards used Appendix I

Parts used Appendix J Wiring numbers Appendix K Visibility

Funnell 42 Sundeck Ferry Emblem Fleet

Technical Manual Page 2 of 9

1 Boat Design Category

The vessel is to be Category D as per RCD and Category A for MCA stability requirements. Requirements for Category D as follows - This craft is designed to operate in winds of up to Beaufort force 4 and the associated wave heights (occasional maximum waves of 0.5m). such conditions may be encountered in sheltered inland waters and in coastal waters in fine weather.

1. 1

Craft Identification Number

The CIN number will be made up as such for each boat. GB-FYM 36

G

01

C 9 09

Manufacturers code / Length / Type / Number of Build / Month / Year / Model year

The CIN number will be sandblasted to an area close to the transom on the starboard aft corner. It will also be placed in the inside of the boat behind the rudder position.

1. 2

Builders Plate

The builders plate will be an engraved plate attached to the lower part of the lower helm position. It states the recommended load, maximum numbers of persons, design category, CE logo, Ferry Marina.

1. 3

Protection from Falling Overboard

There are handrails located all round the deck area, also a rail around the screen area on the front deck. There is a hand rail on the aft steps up to the upper helm. These rails are all made from 316 stainless steel with a diameter of 1”, they are screwed in place onto the superstructure. There is a high toe rail all round the side decks to help prevent slipping. As a means of re-boarding there is a set of rails permanently attached on the transom with a grab bar above the deck to assist re-boarding.

Funnell 42 Sundeck Ferry Emblem Fleet

Technical Manual Page 3 of 9

1. 4

Visibility from the Main Steering Position

The main steering position is from the upper helm. This position gives clear all round vision in all directions, however, there are large blind spots at the waterline level as shown in Appendix K. These blind spots will only hinder visibility at waterline level not where other larger vessels are present, as shown on the second diagram, these are easily visible from the seated position. Small canoes etc will not be seen if in blind spots for up to 5m away from the standing helm position. Movement up to 1m away from the helm position will reduce this blind spot down to 2.8m astern and 2.5m on the port side. Approx height for the canoe measurement is given at 60cm above the water. The lower helm does not give any visibility astern or to starboard aft sections. There is a warning notice in place for this at the helm position.

1. 5

Owners Manual

See attached manual.

1. 1

Structure

The hull and superstructure have been manufactured by Waveney River Centre Limited or Hardy Marine. The hull and superstructure are bonded together using two layers of GRP and flow coat overtop to ensure water tightness. The floor bearers are constructed from 3” wood of various lengths and depths, all floor bearers have been glassed over. The entire bilge area has been flow coated to ensure water tightness. The bulkheads are made from ¾” pre-laminated ply wood. The bulkheads have been bonded to the superstructure and floor using freefix along the joints. The deck supports are constructed from 1” ply and are glassed into place. The roof throughout the aft cabin is mainly a false roof, this is constructed by forming a batten lattice across the void and using soft lined panels attached to the battens. Some of the saloon roof is also constructed in a similar way. The rest of the saloon roof is loft lined panels, these are attached to spacing battens onto the inner GRP. The design was completed by John Moxham – Marine Architect, all calculations with regards to scantlings etc. were made by him prior to handing over completed drawings.

Funnell 42 Sundeck Ferry Emblem Fleet

Technical Manual Page 4 of 9

The Rubbing strips are made from 2 ½” rubber D section which is bolted in place.

1. 2

Stability and Freeboard

The test carried out for stability and freeboard allowed for ten people at 85kg each (average). The test was carried out at this level to allow for a change of layout to accommodate more people, or to allow for a private vessel taking friends on board. The vessel was tested with all people to the port side and starboard side, with the driver staying at the helm position. Sharp right and left turns were completed at different speed intervals to test stability.

1. 4

Openings in the Deck, Hull and Superstructure.

See Appendix B for locations

1. 5

Flooding

The roof vents are of a UFO type which prevent ingress of water. The hull vents are over 600mm above the waterline. The through hull fittings all remain over 250mm above the NLWL. There is one bilge pump fitted capable of pumping 1000 gallons per hour. Openings shown in Appendix B

1. 6

Recommended Load

Fuel

149kg Water

528kg Waste

419kg Persons 8 600kg (75kg each) Luggage

320kg (40kg each)

Total 2016kg

Total stated on the builders plate for people and luggage 920kg.

Funnell 42 Sundeck Ferry Emblem Fleet

Technical Manual Page 5 of 9

1. 7

Life raft Stowage

A stowage point is available on the upper helm area, a label is affixed here for reference.

1. 8

Escape

Escape points are located in Appendix B.

1. 9

Anchoring, Mooring and Towing

There are six stainless steel cleats located around the deck area. These are bolted in place with a wooden pad underneath and washers to prevent the bolts being ripped out of the deck. These cleats are suitable for use to tie off mooring ropes and tying off any mud weights. The fwd and aft cleats are suitably strong enough for emergency towing. The boat is not designed to tow another vessel for any long distances. The boat can be towed either along side a tug boat or using the front cleats. There is a recess for the mud weight on the front deck, the mud weight is not suitable for use on the aft deck, this is stated in the owners manual.

4 Handling Characteristics

See Appendix C for data on the handling characteristics.

1. 1

Engine, Engine Ventilation and System

The engine is located under the aft centre section of the saloon seating. It is contained over an engine tray which has be flow coated to ensure no ingress of oil into the GRP.

Funnell 42 Sundeck Ferry Emblem Fleet

Technical Manual Page 6 of 9

The engine bearers are constructed from 1 ½” laminated ply, which have been glassed into place. The engine is mounted onto the beds using coach bolts and rubber mountings. Ventilation for the engine comes under the floorboards from large hull vents both sides and is also fitted with an extractor fan, linked to the ignition system.

1. 2

Fuel System

The 186 ltr fuel tank is located in the mid double cabin under the deck area. The tank is screwed into place using the welded tags on the corners of the tank. An electronic fuel gauge is located in the dash. The filler pipe and vent pipe are ISO 7840 type pipe. The vent pipe is located above the level of the filler. The vent is fitted with a flame trap. The deck filler is clearly marked for use with diesel with a red engraved label. The fuel feed and return for the engine are piped to the filter in seamless copper. The pipes onto the engine are piped using ISO7840 type fuel flex. The filter is all metal and conforms to the fire resistance specification. The manual fuel shut off is located next to the filter. There is also an electrical fuel shut off located next to the manual shut off, this is operated from a feed to the ignition. The fuel is automatically shut off when the ignition is turned off and switched on when the ignition is on again. The fuel line to the Eberspacher heater is also piped in seamless copper. The fuel shut off for the heater is located next to the heater. All fuel lines are clipped at recommended intervals to remove any vibration. The diagram for the fuel system is located in Appendix F.

1. 3

Electrical System

A complete diagram of the electrical system can be found in Appendix D.

1. 4

Steering System

The steering system is operated using hydraulic pipes. The diagram for the pipe locations is located in Appendix E. The steering system can be operated from the upper and lower helm positions.