I. Summary of Fire Incidents in 2005

8 FOREWORD

We are pleased to issue the revised edition of an English booklet entitled "Korean Fire Data" which offers an outline of Korean fire statistics during the year 2005.

The first edition was published in 1997 in accordance with the "The agreement of Cooperation between the NFPA (National Fire Protection Association) and the KFPA (Korean Fire Protection Association)." It was probably the first Korean fire related statistics in English which paved the way for disseminating of information throughout the world including USA. Since then, the revised editions have been issued annually.

Take a look at the Korean fire data in the year 2005.

Thirty two thousand and three hundred and forty (32,340) fires were attended by the public fire services, a decrease of 1.2% from the previous year. The death rose by 4.3% from 484 in 2004 to 505 in 2005 and injuries rose by 0.9% from 1,820 in 2004 to 1,837 in 2005. Property loss has accounted for 171.37 billion Won (approximately US $180 million) in 2005, 16.9% increased from 146.63 billion Won (approximately US $154 million) in 2004.

With regard to the property loss in 2005, it still showed increase rate of

1. 9%, average 6.2% increase per annum for the past ten year period. In addition, the rate of fire occurrence in Korea has increased by an average of 2.3% per annum for the past ten years.

The above fire statics shows that the fire safety in Korea still has to attain more solid establishment in various areas, when compared to the advanced countries such as USA, Japan and Western European nations. Therefore Korea should make sincere efforts to increase investment in the fire infrastructure and to enhance the people’s safety awareness in cooperation with the government and the private sectors.

We hope this booklet serves not only to deliver the current Korean fire data but to reduce the worldwide fire disaster.

CONTENTS

❏ 2005 Fire Report ··························································· 1

Ⅰ. Summary of Fire Incidents in 2005 ································· 3

Ⅱ. Fires for the 10-Year Period (1996-2005) ······················· 7

❏ Fire Investigation Reports ··········································· 27

❍ Laboratory Fires ·································································· 29

❍ Explosion Cases ·································································· 41

❍ Analysis of Plant Fire ······················································· 49

SOURCE : Fire Statistics Yearbook 2005

by the National Emergency Management Agency of

the Republic of Korea

2005 FIRE REPORT

I. Summary of Fire Incidents in 2005

1. Fire Loss Summary

2. 2005 Fires by Causes

3. 2005 Fires by Property Types

4. 2005 Fires by Months

5. 2005 Fires by Major Cities and Provinces

Thirty two thousand and three hundred and forty (32,340) fires were attended by public fire services, a decrease of 1.2% from the year before.

Death toll rose by 4.3% from 484 in 2004 to 505 in 2005 and injuries rose by 0.9% from 1,820 in 2004 to 1,837 in 2005.

Property loss accounts for ¥171.37 billion in 2005, 16.9% increased from ¥146.63 billion in 2004.

The leading causes of 2005 fires were electrical distribution involved, arson, smoking, ember and children playing etc. and they account for 62.5% of the total fires in 2005. Electrical fires involved ranked first, occupied about 30.9% of all fires in 2005.

Twenty-six percent (26.0%, 8,411) of the total 32,340 fires occurred in homes, including houses and apartments ranked first, vehicles fires ranked second or 18.2% (5,884), factory fires ranked third or 10.4% (3,358).

6. Catastrophic Fires for the Last 5 Year Period (2001 - 2005)

• A catastrophic fire is defined as following according to the "Fire Investigation & Report Regulation".

• Death toll : Not less than five people

• Injuries : Not less than ten people

• Property loss : Not less than two billion Won(¥2 billion)

Catastrophic fires for the last 5 years marked an average 82.7 % increase rate per year. But the 5 catastrophic fires in 2005 represented a 400 % increase from 2004.

The 6 fire deaths in 2005 showed an decrease of 85.4 % from 2001, but a increase of 20.0 % from 2004, with a 76.9 % increase per year for the last 5 year.

The 74 fire injuries in 2005 showed a 7.5% decrease from 2001

In 2005, property damage caused by fires amounted to ¥25,078 million, which marked a increase of 2,485.4 % and 417,866.7 % respectively from 2001 and 2004, with a 86,189.0 % increase per year for the last 5 year.

II. Fires for the 10-Year Period (1996-2005)

1. Number of Fires for 1996-2005

〔图〕插图

• TF : Total Fires, S : Structures, VH : Vehicles, Ships, Aircraft, etc

*IR:Increase Rate

The number of fires for the last 10-year period has shown an average 2.3 % increase rate per year since 1996. In 2005, total 32,340 fires occurred to be up by 12.8 % from 1996. Fires in Structures marked an average 2.3 % increase rate per year, and resulted in a 14.1 % rise from 23,106 fires in 1996 to 26,370 fires in 2005. Vehicle (vehicles, ships, aircraft) fires have risen with an average 3.1 % rate per year to mark a 7.4 % increase rate in 2005 compared to 1996.

1. Fire Casualties for 1996-2005

〔图〕插图

• TF : Total Fires, * IR : Increase Rate

○ Trend in fire deaths

• Deaths in 2005 : increased average 0.7 % per annum for the last 10-year period

• Numbers in 2005 compared to 2004 : 4.3% increase

• Numbers in 2005 compared to 1996 : 14.3 % decrease

○ Trend in fire injuries

• Injuries in 2005 : increased average 1.4 % per annum for the last 10-year period

• Numbers rate in 2005 compared to 2004 : 0.9 % increase

• Numbers rate in 2005 compared to 1996 : 12.4% increase

1. Property Loss for 1996-2005

〔图〕插图

• PL : Property Loss, * IR : Increase Rate

○ Trend in property loss

• Increase trend for the last 10-year period : increased average 6.2 % per annum

• Numbers in 2005 compared to 2004 : 16.9% increase

• Numbers in 2005 compared to 1996 : 51.5% increase

○ Property loss order by places

• 1st : factories ¥39,274 mil. (22.9 % of total ¥171,374 mil.)

• 2nd : residential ¥21,371 mil. (12.5 % of total ¥171,374 mil.)

• 3rd : vehicles ¥19,643 mil. (11.5 % of total ¥171,374 mil.)

○ Property loss order by causes

• 1st : electrical fires ¥44,577 mil. (26.0 % of total ¥171,374 mil.)

• 2nd : arson ¥11,009 mil. (6.4 % of total ¥171,374 mil.)

• 3rd : embers ¥7,389 mil. (4.3 % of total ¥171,374 mil.)

1. Fires by Cause for the 5-Year Period 2001-2005

4-1. Electrical Fires

〔图〕插图 O Fire trend

• 5-year : decreased average 3.2 % per annum

• Numbers in 2005 compared to 2004 : 4.4% decrease

• Numbers in 2005 compared to 2001 : 18.8 % decrease

O Fire casualties

• 5-year trend : deaths and injuries decreased average 1.6 % and 3.5% per annum respectively

• Numbers in 2005 compared to 2004 : deaths up by 46.7 % and injuries down by 1.5 %

• Numbers in 2005 compared to 2001 : deaths and injuries down by 42.9 % and 15.0 % respectively

○ Property loss

• 5-year trend : decreased average 4.9 % per annum

• Numbers in 2005 compared to 2004 : 6.8 % decrease

• Numbers in 2005 compared to 2001 : 34.3 % decrease

4-2. Smoking materials

〔图〕插图

○ Fire trend

• 5-year : decreased average 4.8 % per annum

• Numbers in 2005 compared to 2004 : 8.2% decrease

• Numbers in 2005 compared to 2001 : 26.0% decrease

○ Fire casualties

• 5-year trend : deaths and injuries increased average 12.5% and 2.8% per annum respectively

• Numbers in 2005 compared to 2004 : deaths up by 76.5 % and injuries down by 13.2 %

• Numbers in 2005 compared to 2001 : deaths up by 15.4% and injuries down by 19.3%

○ Property loss

• 5-year trend : decreased average 4.7% per annum

• Numbers in 2005 compared to 2004 : 19.3 % decrease

• Numbers in 2005 compared to 2001 : 36.3 % decrease

4-3. Arson

〔图〕插图

○ Fire trend

• 5-year : increased average 5.5% per annum

• Numbers in 2005 compared to 2004 : 1.1 % increase

• Numbers in 2005 compared to 2001 : 22.8 % increase

○ Fire casualties

• 5-year trend : deaths and injuries increased average 32.7% and 8.7% per annum respectively

• Numbers in 2005 compared to 2004 : deaths and injuries down by 6.3 % and 9.9 % respectively

• Numbers in 2005 compared to 2001 : deaths and injuries up by 12.5% and 3.6% respectively

○ Property loss

• 5-year trend : increased average 13.4% per annum

• Numbers in 2005 compared to 2004 : 1.4% increase

• Numbers in 2005 compared to 2001 : 31.7% increase

4-4. Children Playing

○ Fire trend

• 5-year : decreased average 7.1 % per annum

• Numbers in 2005 compared to 2004 : 19.1 % decrease

• Numbers in 2005 compared to 2001 : 18.3% decrease

○ Fire casualties

• 5-year trend : deaths increased average 2.5% per annum and injuries decreased average 12.7% per annum

• Numbers in 2005 compared to 2004 : deaths and injuries down by 60.0 % and 37.9 % respectively

• Numbers in 2005 compared to 2001 : deaths and injuries down by 69.2 % and 30.8 % respectively

○ Property loss

• 5-year trend : decreased average 5.8% per annum

• Numbers in 2005 compared to 2004 : 32.1 % decrease

• Numbers in 2005 compared to 2001 : 36.7% decrease

4-5. Embers

〔图〕插图

O Fire trend

• 5-year : increased average 3.1 % per annum

• Numbers in 2005 compared to 2004 : 0.2 % decrease

• Numbers in 2005 compared to 2001 : 0.2 % decrease

O Fire casualties

• 5-year trend : deaths and injuries increased average 90.5 % and 26.3 % per annum respectively

• Numbers in 2005 compared to 2004 : deaths down by 12.5 % and injuries up by 11.5 %

• Numbers in 2005 compared to 2001 : deaths and injuries down by 50.0 % and 10.8 % respectively

○ Property loss

• 5-year trend : increased average 8.1 % per annum

• Numbers in 2005 compared to 2004 : 15.2 % decrease

• Numbers in 2005 compared to 2001 : 20.6% decrease

4-6. Oils

〔图〕插图

O Fire trend

• 5-year : decreased average 1.7 % per annum

• Numbers in 2005 compared to 2004 : 5.4 % decrease

• Numbers in 2005 compared to 2001 : 15.6% decrease

O Fire casualties

• 5-year trend : injuries increased average 11.6% per annum

• Numbers in 2005 compared to 2004 : deaths and injuries up by 200.0 % and 45.7 % respectively

• Numbers in 2005 compared to 2001 : deaths and injuries up by 80.0 % and 29.7 % respectively

○ Property loss

• 5-year trend : increased average 8.2 % per annum

• Numbers in 2005 compared to 2004 : 36.5 % increase

• Numbers in 2005 compared to 2001 : 17.3 % increase

4-7. Gases

〔图〕插图

O Fire trend

• 5-year : decreased average 17.8 % per annum

• Numbers in 2005 compared to 2004 : 15.9 % decrease

• Numbers in 2005 compared to 2001 : 60.3 % decrease

O Fire casualties

• 5-year trend : deaths and injuries decreased average 16.2 % and 13.1 % per annum respectively

• Numbers in 2005 compared to 2004 : deaths down by 47.8 % and injuries up by 0.9 %

• Numbers in 2005 compared to 2001 : deaths and injuries down by 45.5 % and 42.6 % respectively

○ Property loss

• 5-year trend : decreased average 12.0 % per annum

• Numbers in 2005 compared to 2004 : 18.8 % decrease

• Numbers in 2005 compared to 2001 : 45.6% decrease

1. Fires by Property Type for the 5-Year Period 2001-2005

5-1. Residential (Houses & Apartments)

〔图〕14.0FC(in hundreds)12.010.0Fires(in thousands)8.06.0FiresFire Casualties(FC....PL(in ten4.0-·★--Property Loss(PLthousand ○ Fire trend

• 5-year : decreased average 2.7% per annum

• Numbers in 2005 compared to 2004 : 3.2% decrease

• Numbers in 2005 compared to 2001 : 16.0% decrease

○ Fire casualties

• 5-year trend : deaths increased average 0.2% per annum and injuries decreased average 0.6% per annum

• Numbers in 2005 compared to 2004 : deaths and injuries down by 2.7% and 3.3% respectively

• Numbers in 2005 compared to 2001 : deaths and injuries down by 14.0 % and 11.2 % respectively

○ Property loss

• 5-year trend : decreased average 1.1 % per annum

• Numbers in 2005 compared to 2004 : 3.7% decrease

• Numbers in 2005 compared to 2001 : 19.1% decrease

5-2. Factories

○ Fire trend

• 5-year : decreased average 3.2% per annum

• Numbers in 2005 compared to 2004 : 2.7% decrease

• Numbers in 2005 compared to 2001 : 14.2 % decrease

○ Fire casualties

• 5-year trend : deaths increased average 3.7% per annum and injuries decreased average 5.7% per annum

• Numbers in 2005 compared to 2004 : deaths up by 44.4 % and injuries down by 17.4 %

• Numbers in 2005 compared to 2001 : deaths up by 52.9 % and injuries down by 7.6 %

○ Property loss

• 5-year trend : increased average 1.8% per annum

• Numbers in 2005 compared to 2004 : 11.9% increase

• Numbers in 2005 compared to 2001 : 4.7% decrease

5-3. Vehicles

〔图〕插图

O Fire trend

• 5-year : increased average 0.1 % per annum

• Numbers in 2005 compared to 2004 : 2.1 % decrease

• Numbers in 2005 compared to 2001 : 1.5% decrease

O Fire casualties

• 5-year trend : deaths and injuries increased average 60.0 % and 27.6 % per annum respectively

• Numbers in 2005 compared to 2004: injuries down by 23.1 %

• Numbers in 2005 compared to 2001 : deaths and injuries up by 76.2 % and 12.4 % respectively

○ Property loss

• 5-year trend : increased average 2.9 % per annum

• Numbers in 2005 compared to 2004 : 12.0 % decrease

• Numbers in 2005 compared to 2001 : 8.1 % decrease

5-4. Restaurants

○ Fire trend

• 5-year : increased average 1.8% per annum

• Numbers in 2005 compared to 2004 : 1.4% decrease

• Numbers in 2005 compared to 2001 : 2.8% decrease

○ Fire casualties

• 5-year trend : deaths and injuries increased average 4.6 % and 4.4 % per annum respectively

• Numbers in 2005 compared to 2004 : deaths and injuries up by 8.3% and 27.0% respectively

• Numbers in 2005 compared to 2001 : deaths and injuries up by 8.3 % and 42.3 % respectively

○ Property loss

• 5-year trend : decreased average 0.04% per annum

• Numbers in 2005 compared to 2004 : 13.9% decrease

• Numbers in 2005 compared to 2001 : 31.6% decrease

5-5. Shops

〔图〕插图 ○ Fire trend

• 5-year : decreased average 0.9% per annum

• Numbers in 2005 compared to 2004 : 1.9% decrease

• Numbers in 2005 compared to 2001 : 1.1% decrease

○ Fire casualties

• 5-year trend : deaths and injuries increased average 24.1 % and 0.3 % per annum respectively

• Numbers in 2005 compared to 2004 : deaths and injuries up by 200.0 % and 4.3 % respectively

• Numbers in 2005 compared to 2001 : deaths down by 25.0 % and injuries up by 28.0 %

○ Property loss

• 5-year trend : increased average 0.3 % per annum

• Numbers in 2005 compared to 2004 : 11.5% decrease

• Numbers in 2005 compared to 2001 : 10.6% decrease

5-6. Workplace

〔图〕插图 O Fire trend

• 5-year : increased average 3.5 % per annum

• Numbers in 2005 compared to 2004 : 25.9 % increase

• Numbers in 2005 compared to 2001 : 0.5 % decrease

O Fire casualties

• 5-year trend : deaths decreased average 3.8 % per annum and injuries increased average 28.3 % per annum

• Numbers in 2005 compared to 2004 : deaths down by 40.0 % and injuries up by 126.9 %

• Numbers in 2005 compared to 2001 : deaths and injuries down by 57.1 % and 24.4 % respectively

○ Property loss

• 5-year trend : increased average 11.0 % per annum

• Numbers in 2005 compared to 2004 : 25.7% increase

• Numbers in 2005 compared to 2001 : 39.9 % decrease

5-7. Warehouses

〔图〕插图

O Fire trend

• 5-year : decreased average 3.0 % per annum

• Numbers in 2005 compared to 2004 : 3.3 % increase

• Numbers in 2005 compared to 2001 : 3.8 % decrease

O Fire casualties

• 5-year trend : injuries increased average 14.1 % per annum

• Numbers in 2005 compared to 2004 : deaths and injuries up by 50.0 % and 4.5 % respectively

• Numbers in 2005 compared to 2001 : injuries up by 27.8 %

○ Property loss

• 5-year trend : increased average 10.1 % per annum

• Numbers in 2005 compared to 2004 : 14.1 % increase

• Numbers in 2005 compared to 2001 : 36.6% increase

5-8. Hotels & Motels

〔图〕插图

O Fire trend

• 5-year : decreased average 1.1 % per annum

• Numbers in 2005 compared to 2004 : 2.0 % decrease

• Numbers in 2005 compared to 2001 : 9.5 % decrease

O Fire casualties

• 5-year trend : deaths and injuries increased average 14.0 % and 6.2 % per annum respectively

• Numbers in 2005 compared to 2004 : deaths up by 23.1% and injuries down by 20.7 %

• Numbers in 2005 compared to 2001 : injuries down by 9.9%

○ Property loss

• 5-year trend : increased average 5.3 % per annum

• Numbers in 2005 compared to 2004 : 38.4 % increase

• Numbers in 2005 compared to 2001 : 4.1 % decrease

5-9. Schools

○ Fire trend

• 5-year : decreased average 0.8 % per annum

• Numbers in 2005 compared to 2004 : 18.7% decrease

• Numbers in 2005 compared to 2001 : 16.2% decrease

○ Fire casualties

• 5-year trend : injuries increased average 1.4% per annum

• Numbers in 2005 compared to 2004 : deaths and injuries down by 100.0 % and 77.8 % respectively

• Numbers in 2005 compared to 2001 : injuries down by 75.0 %

○ Property loss

• 5-year trend : decreased average 5.5 % per annum

• Numbers in 2005 compared to 2004 : 46.9 % decrease

• Numbers in 2005 compared to 2001 : 20.0% decrease

FIRE INVESTIGATION REPORT

Laboratory Fire

Laboratory Fires

I. Overview

Large lab fires are rare as labs handle rather small amount of hazardous materials but they sometimes see deadly fire incidents as occurred in a lab of Nuclear Engineering Department at Seoul National University in 1999. In particular, chemical labs are highly likely to face potential risks as they deal with many kinds of hazardous chemical substances. In many cases a failure of safety management has led to a failure of first response, causing devastating damage.

Labs are used for education purpose at elementary, middle, high schools and universities and for research purpose at institutes, industries and graduate schools.

According to the report on school fires in recent 5 years issued by National Emergency Management Agency, fires in school are not at highly worrisome levels (see Table 1 and Table 2), on the other hand the report on analysis of fires at special-purpose buildings in schools conducted by Korean Fire Protection Association shows that fires at these buildings in schools in recent 5 years reach 6.5 cases × 10$^{2}$/year in frequency, which is unusually high figures.

Table 1. School Fires and Fire Loss

Table 2. School Fires by Cause

II. Fire Cases

Small fires in laboratories often tend to be underestimated as they are not recognized well from outside. Recent fires in labs at universities, institutes and industries are described in Table 3. February. 2007, both chemical waste leakage and fire occurred in the same university.

Table 3. Lab Fire Description

1. Fire Case at KIST (Korea Institute of Science and Technology)

1.1 General

● Date: September 19, 2006 at 18:50~22:33

● Damaged Property: Younji building at KIST

● Location: First floor of L-3 research building at Seoul KIST

● Cause: Unknown

● Fire Loss

(1) Human loss: None

(2) Property Loss : 147,4110,000 won (Immovables : 52,000,000 won, Movables : 95,000,000 won)

1.2 Overview of the building

(1) General

The building (approximately 4,881m $^{2}$ of total floor area) was R.C structure and had one underground level and 3 above ground occupiable floors. It had a 1-story high temporary structure made of Sandwich Styrofoam panels at the center of the building and each floor was not separated by fire protection system. At the center of the first floor was the temporary structure and on one side were labs of concrete structure. The second and third floors were open area with labs which were on the same side of the building as the first floor labs.

(2) Separation

The Building has a stairway at the center. The stairway services all 3 floors. The temporary structure at the center of the first floor was not connected to the labs. The labs on each floors had wooden doors.

(3) Structure

The area of ignition was Fuel Cell Research Center on the first floor. First floor section was covered Sandwich Styrofoam panels. Lab 1 and Lab 2 were divided. Lab 1 had an office in front of the door which was separated from the lab space by cabinet and desk.

〔图〕插图

Fuel Cell lab 1

Fuel Cell lab 2

Material Lab

× : Proof of Short Circuit — : Electric Wire

Fig. 1. Electric System Diagram of Area of Ignition

〔图〕插图

Photo 1. Damage of Lab 1

〔图〕插图

Photo 2. Damage of Lab 2

1.3 The Fire

(1) Ignition Area

The person in charge of the Fuel Cell lab 1 had left the lab where the ignition appear to have started about 20 minutes before the fire was discovered, which was 18:31. A Russian researcher in the lab 2 also had gone out at about 18:37 before the fire was recognized. The fire broke out at around 18:50 when there was no one in any of the labs.

There are a few reasons to regard the Fuel Cell lab 1 as the area of ignition.

First, the pattern of fire spread told that it occurred in the lab 1. The fire spread through the ceiling spaces and into other labs.

Second, according to the witness, Mr.Whang, when he went to the lab 1 where he thought the fire occurred after the fire alarm set off, he saw smoke and flame inside the lab. He immediately reported and took initial action with a fire extinguisher stored in the building. Mr.Park, the researcher working in the Material lab(Next to Fuel cell lab 2) on that day in the same building told that he saw black smoke coming out of the ventilation duct on the ceiling in his lab and he realized that it was coming from the Fuel Cell lab 1. He and a building maintenance employee escaped when they failed to suppress the fire.

Third, the proof of a short circuit was discovered in Fuel Cell lab 1 where initial witnesses stated they saw the fire.

• 35 -

(2) Ignition Source

The office in the Fuel Cell lab 1 where the fire originated had desks, a chair, a bookrack and a cabinet. Inside the lab were research equipment, desks and switchboard. Potential sources of ignition included gas, arson, cigarette, chemicals and lingering embers. A Short circuit in origin as electricity related source remained suspected, but it was hard to prove exactly how it was involved as the lab was severely damaged.

(3) Fire Spread

The fire originated at point A in the Fuel Cell Lab 1, spread through the ceiling spaces and into other places. It is believed that the fire was recognized when smoke and fumes from the fire extended to the Material Lab through ventilation duct.

1. Recommendations for Laboratory Fire Safety

1. 1 Recommended Lab Structure

(1) Laboratory structure should comply with Table 4.

(2) Laboratory should be separated from non-laboratory by fire proof construction that can endure at least one hour from non-laboratory regardless of fire-resistance rating described in Table 4

(4) In places where water supply can cause large fire and loss of human life, appropriate

automatic fire suppression system should be placed in stead of sprinkler system.

(5) Cross points of floors, ceilings and walls with piping, wiring, cables, wires, air duct,

air pressure tube, duct and other similar devices should be built in fire proof manner.

(6) Any openings on the floor that have the potential for hazardous materials to leak through should be sealed safely and a bump should be made around the openings to avoid the leakage through the openings.

(7) Fire separations that need an hour fire resistance rating should have doors with fire-resistant material that can endure at least half an hour. Fire separations that need two hour fire resistance rating should have fire resistant-treated doors that can last at

• 37 -

(5) High Temperature and High Pressure facility

Laboratory emergency plans should contain the followings.

(1) Alarm activation

(2) Evacuation procedures and building re-entry information

(3) A shut-off procedure of equipments or available emergency measure

(4) Fire suppression activities

(5) Non fire danger

(6) Information for emergency response team

FIRE INVESTIGATION REPORT

Explosion Cases

• 41 -

Explosion Cases

Ⅰ. Forward

Unlike fire, Explosion gives no time for people to evacuate as the incident happen in a short time interval. It is important to prevent leakage of hazardous materials primarily as Inflammable gas and steam involved in explosions can detonate easily by low ignition energy.

If leak of Inflammable gas and vapor already started, response work of removing and diluting hazardous materials would be widely used and explosion proof devices would be also used to suppress source of ignition like spark caused by electricity and friction. Physical explosion like vapor explosion simply caused by high pressure has low rate of pressure rise compared to chemical explosion and the possible damage can be considerably reduced through pressure relief.

While some explosions occur when safety measures or basic requirements are not followed, others happen just by a lack of information about characteristics of chemicals and processing methods.

In this report, cases of hydrogen gas explosion, vapor explosion and oil mist explosion are introduced to see characteristics of the explosions.

Ⅱ. Danger of Explosion

When explosion happens, the air tries to balance with pressure. The sudden increase in noise and pressure by explosion cause great destruction and damage. And the pieces from destroyed containers can smash into nearby facilities and buildings, creating additional damage. Furthermore, if the materials involved in the explosion are inflammable, fire will be followed, causing more damage.

1. The Blast Pressure Wave Effect

The air that moves by explosion pressure produces blast pressure wave which is divided into two phases (the Positive-Pressure Phase and Negative-Pressure Phase) by the direction of force at the point of explosion.

In the positive phase, expanding gas rushes far away from the point of detonation. The positive phase is normally powerful than the negative phase and covers most of the damage caused by pressure. In the negative phase, the rapidly expanding gas rolls outward from the point of detonation, pushing away surrounding air and forming relatively low pressure at the center of the point of detonation. When the positive phase disappears the air comes back to the point of detonation to make the low pressure balanced.

〔图〕P50PressureAmbient, Po0PositivephaseNegative phaseDuration, t5Duration, toTime after explosion\n

Fig 1. Changes in Explosion Pressure by time

2. Fragmentation Effect

Fragments can produce great damage and casualties within a larger area from the point of detonation. They also can destroy electric power line, fuel gas & combustible fuel pipe and storage containers, affecting the magnitude of fire and causing additional blast.

3. Thermal Effect

Flame explosion releases flame gas and great energy that heat the surrounding air. The energy can ignite fire on nearby combustible materials and people can also get a burn.

• 43 -

The fire after the blast increases property and human losses and complicate investigation by interrupting the determination of which one occurred first, the blast or fire.

Thermal Effect caused by inflammable vapor in BLEVE (Boiling Liquid Expanding Vapor Explosion) consists of fireball and firebrand. Fireball which is produced at the instant of detonation or after the detonation release radiant heat in a short time. Firebrand scatters fragments at high temperature or on fire during detonation. All of these effects can play a big roll in causing fires in wider areas far away from the point of detonation.

1. Seismic Effect

Quake that is generated when blast pressure wave is produced or big structures drop on the ground can be transmitted through the ground and create further damage on structures and underground facilities, pipelines, tanks or cables.

Ⅲ. Incident Cases

1. Hydrogen Gas Leak Explosion

1. 1 General

• Date : September 20, 2004

• Location : Ulsan

• Human loss : 2 deaths

1. 2 The Incident

When the hydrogen compressor was run in front of the panel of the pressure-type explosion-proof structure during Polypropylene process, hydrogen entered into a electric panel through nitrogen purge pipe situated at the inlet side of the compressor and it caused electric spark, contributing to explosion. Two employees at work in front of the panel were killed by being hit by the cover of panel at the scene.

5 days before the incident occurred, hydrogen compressor (a) was switched by compressor (b) for repair. For Compressor (a), as planed, a blind was inserted in flare

stack pipe connected to cylinder block to change 2 interior valves and purge pipe connected to cylinder block was blocked for backflow prevention. And the purge was done by pressuring with nitrogen (7kg/cm $^{2}$) through purge line on the inlet side of the compressor. Considering of the situation, there were possibilities that all 3 shutoff valves of the purge pipe were left open or a valve that leaked was remained open.

1.3 Problems and Measures

(1) When it comes to changed factors, workers should comply with some procedures including maintaining changed factors, danger assessment, document upgrade, filling out a form of safety work procedure, employee training, mechanic checks prior to the operation.

(2) A blind on the nitrogen pipe for purge at the inlet side of a compressor needs to be equipped at the back of shutoff valve in order to make it to be installed in the middle of operation. It is also required to be separated from nitrogen pipe for keeping the positive pressure in a control panel.

(3) Once any equipment has been repaired, regular checks and assessment should be provided to check various issues, such as openness/closure of valves and compressors, problems with attachments related to any instruments and periodic leakage, prior to start-up of that equipment.

〔图〕插图

Photo 1. The exploded Electric Control Panel

• 45 -

1. Vapor Explosion

1. 1 General

• Date : March, 2005

• Location : Cheongwon, Chungbuk Province

• Human loss : 2 deaths and 2 injuries

1. 2 The Incident

The explosion occurred in a flue/smoke tube boiler which was used to provide steam for recycling waste paper using firebrand in an incinerator. The blast scattered boiler plates and smoke tubes up to 100 meters away and claimed 4 lives including workers nearby.

After the cleaning was done, the boiler was operated for 2 hours without opening steam emission valve. The low water level alarm was not activated, which was stopped when water was drained and water was not refilled. The temperature of the incinerator kept going up to 500°C (932°F) and steam failed to be released to steam header or outside as steam emission valve of the boiler was closed. As the pressure inside boiler went up, safety valve set to run at 8kg/cm² was activated, emitting only small amount of steam. It was when the boiler operator opened silencer to release the steam faster. When the switch of water supply pump was on, the operator opened air emission valve of the water supply pump to emit the air. Water supply pump started to provide water automatically. 20 minutes later heated boiler exploded.

1. 3. Problems and Measures

(1) A work procedure for stopping and restarting a boiler and principles of emergency response wasn't followed and a work procedure form wasn't written.

(2) The boiler was supposed to have regular checks within validity period but it didn't carried out.

(3) Pressure relief device wasn't tested by the pressure gauge approved by the national calibration organization at least once a year.

(4) The system that completely control heat flow overheating a boiler was inappropriate.

The boiler, a waste heat boiler, at the scene couldn't effectively close off the heat flow

• 46 -

due to the chimney's stack effect even though F.D Fan and I.D fan were stopped. When a boiler has low water level and overpressure it is needed to make the heat flow bypass in order to prevent the firebrand from being provided.

1. Oil Mist Explosion

1. 1 General

•Date : Sep, 2004

•Location : Gyeongju in Gyeongbuk Province

•Human loss : 1 death

1. 2 The incident

When an oil tank of waste boiler was being cut with an oxygen cutter in a ferrous scrap yard, oil mist remained in the tank exploded. A worker with third degree burn from the blast was taken to a hospital but died. The worker either didn't check whether there was inflammable material remained, but he just kept doing his job ignoring the danger.

1. 3 Problems and Measures

Even when kerosene with relatively high flash point is dealt with, there is always possibility of ignitable vapor remained, which means workers must make sure that air ventilation is done to be within the explosion limit range and water is sprayed on the cutting part to reduce the temperature.

FIRE INVESTIGATION REPORT

Analysis of Plant Fire

Analysis of Plant Fire

I. Overview

Fires at plants and workshops are in third place in frequency after fires at houses/apartments and vehicles, and account for 10% of total fires in Korea.

Property loss, however, represents 30% of total financial loss and there would be even greater financial loss through business interruption. For the past 30 years all 10 big fires happened in plants and it is ever more important to prevent fires and minimize the loss.

By cause, fires caused by electricity and embers account for nearly half of the total fires and cigarettes and arsons for 10%.

Table 1. Fire Loss at Plants and Workshops in recent 5 years.

Table 2. Fires by Location in 2005

Table 3. Fires at Plants and Workshops by Month in 2005

II.Fire Cases

1. General

● Location: A plant in Gyeongbuk Province

● Date: March, 2005 at 23:25

● Ignition Area : Inside oven

● Property Loss : 7 million Won

● Human Loss: None

● Cause of Fire : Problems with electric cooling fan

2. The Outline of Plant Work

It was to cast Liquid silicon in stator used for vehicle, to transport the stator to a oven and solidify the silicon at 250°C (482°F).

The oven on fire was not a precision work equipment but played a big role in production process by making the silicon cast into the stator solid.

3. Ignition and Suppression

The fire broke out in the dry part of the oven where 39 stators (3×13) were stored. A factory worker discovered that smoke came out of the oven and asked for help. He opened the cover of the oven and suppressed the fire with extinguisher. However he failed to extinguish the fire immediately.

The cover was not opened with ease as it worked by electric power. He had to open side covers of the oven and spray the dry chemical extinguisher into the gaps of the sides.

Large amount of powder was found in the damaged cooling fan later.

• 51 -

Ⅲ. Case Analysis

The oven worked 24 hours and 365 days. As there was no extra equipment that replaces the existing one, it was impossible to clean and repair the oven on a regular basis. Inside oven were piled dusts and lubricant mist. Electrical equipments and heated atmosphere gave the place a good condition to ignite.

1. Source of Ignition

• Electrical coil : It was inside of the metal pole in the dry part.

• Heated air : Air temperature of the dry part was around 250°C (482°F).

• Electric system :The Cooling fan was stored in the Cooling part.

• Friction heat : There was possibility that friction in rotation part of cooling fan caused heat.

• Spark : There was possibility that some metal pieces crashed into the fan and caused spark.

1. Ignited material

•Dusts : The oven had a Suction device that can emit air. It might have caused the inflow of dusts.

•Oil Mist : Mist from lubricant and silicon might have been generated due to high temperature of the dry part.

•Silicon : The flash point of silicon for stator is 115°C (239°F).

•Lubricant : The Flash point of lubricant used for conveyer belt is 240°C (464°F).

•Electric wire coating material : Heat resistant PVC wire was used.

•Palette : Plastic was used.

1. Incident Scenario

(1) Ignition by defective connection with electric power system of cooling fan

•Possibility

Dusts and oil mist on power connection part were cooled and solidified, making a short

• 52 -

circuit, igniting and burning the coating materials of electric wires.

•Analysis of Evidence

The coating materials of the electric wires which were close to the cooling fan was

severely damaged.

The damage in the oven covered horizontally from central part where cooling part was

to air outlet, and upward was only air outlet damage.

(2) Ignition by mechanical energy of the cooling fan

•Possibility

Some metal pieces crashed into the cooling fan and caused spark. There was another speculation that a problem with motor axis of the cooling fan generated friction heat, but investigation found low possibility on that speculation.

•Analysis of Evidence

The investigation of the collected cooling fans (11 of 12 were identified) showed no evidence of damage.

(3) Natural ignition by heated air in the dry part

•Possibility

Hot wire tubes were in the dry part. There was possibility of ignition by silicon [flash

point : 115°C (239°F)], lubricant [flash point : 240°C (464°)] and oil mist as the dry

part was run at around 250°C (482°F).

Considering of the damage pattern in the oven, the possibility that the fire originated

in the upper part and extended to the lower part was low.

•Analysis of evidence

As the inside of the oven was where airflow was made compulsorily, it is less possible that oil mist was within the flammability limit range and as 39 Stators were in the dry part on that day, which was unusual condition.

• 53 -

Ⅳ. Problems and measures

1. Cause of the Incident

A lack of evidence (the fire scene was cleared out before the investigation) and inconsistent eyewitness testimonies were obstacles to an accurate investigation. However the cause was determined by some identified evidences and analysis of the fire behavior.

(1) It is determined that the fire originated due to the electrical problem of cooling fan, extended to the coating material of electric wires and dusts, and spread toward vents.

(2) It is also possible that ignition started in the heated atmosphere with great oil mist in

the dry part. In that case the ignited material would have dropped on the lower part

(cooling part), contributing to a second fire, but no damage found there.

(3) Other causes were determined to indicate no possibility of ignition.

1. Safety Measures

The following safety Measures for an oven can be provided.

•Sources of ignition : Dusts should be removed and lubricant with high flash point should be used.

•Wire connection part : Dusts and carbon accumulation should be removed.

•Location of control panel : Control panel should be separated form main board.

•Metal pieces : An equipment that keeps any metal pieces away should be installed.

•Fire spread prevention : Indoor fan and vent fan should be closed down in emergency.

•Control system : It should be set to ring the alarm when the temperature surpass 10% of set temperature and to avoid products from moving to next steps when the temperature surpass 20% of that.

•Automatic fire suppression system : Gas-type fire suppression should be installed.

•Communication : Effective system for cooperation from the employees at the fire scene is needed.

•Documentation : Documents about fire safety such as a list of dangerous facilities and reports on fires should be written.

〔图〕插图

Photo 1. Damage in the vent and duct : The connection part was melted.

〔图〕に、彼の\n

Photo 2. Piled dusts inside of the oven

This is provided under the Agreement of Cooperation between the KFPA and the NFPA. Please feel free to contact Risk Research & Development Department of KFPA if you have any inquires. (E-Mail kfpa@kfpa.or.kr, Tel 82-2-780-8111, Fax 82-2-783-4094)

2006 KOREAN FIRE DATA

Published in March 2007

Edited by Risk Research & Development Department Published by Korean Fire Protection Association

35-4 Yoido-Dong, Yongdungpo-Ku

Seoul 150-885, KOREA

TEL 82-2-780-8111

FAX 82-2-783-4094

E-Mail

kfpa@kfpa.or.kr// http://www.kfpa.or.kr Registered No. 13-794 (1997.8.14)

Printed by LeeWhan (Tel 82-2-764-1116)

ISBN 978-89-88874-65-3 93530