韩国_KFPA_火灾统计年报_화재조사보고서

8

2004

KOREAN FIRE DATA

This is provided under the Agreement of Cooperation between the Korean Fire Protection Association and the National Fire Protection Association

1. 3

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 2003.

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 and paved for disseminating the information to the world including USA. Since then, the revised editions have been issued annually.

Take a look at the Korean fire data in the year 2003.

Thirty one thousand and three hundred and seventy two (31,372) fires were attended by the public fire services, a decrease of 4.8% from the previous year. The death rose by 51.5% from 491 in 2002 to 744 in 2003 and injuries rose by 19.8% from 1,744 in 2002 to 2,089 in 2003. Property loss has accounted for 151.59 billion Won (approximately US $151 million) in 2003, 5.7% increased from 143.44 billion Won (approximately US $143 million) in 2002.

With regard to the property loss in 2003, it still showed increase rate of 5.6%, average 18.0% increase per annum for the past ten year period. In particular Korean non-life insurance companies say that the actual property losses may be estimated to several times as much as the above official figures. In addition, the rate of fire occurrence in Korea has increased by an average of 5.6% 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

$\blacksquare$ 2003 Fire Report $\ldots$ 1

1. Summary of Fire Incidents in 2003 $\ldots$ 3

II. Fires for the 10-Year Period (94-03) 7

Fire Investigation Reports 27

$\blacktriangleright$ Fire in Lodging Houses 29

► Explosion in Chemical Plant $\ldots$ 37

► Gas Fire in a House $\ldots$ 45

2003 FIRE REPORT

SOURCE : "2003 Fire Statistics Report

by the Ministry of Government Administration and Home Affairs of the Republic of Korea

I. Summary of Fire Incidents in 2003

1. Fire Loss Summary

2. '03 Fires by Causes

3. '03 Fires by Property Types

4. '03 Fires by Months

5. '03 Fires by Major Cities and Provinces

Thirty one thousand and three hundred and seventy two(31,372) fires were attended by public fire services, a decrease of 4.8% from the year before.

Death toll rose by 51.5% from 491 in 2002 to 744 in 2003 and injuries rose by 19.8% from 1,744 in 2002 to 2,089 in 2003. (mainly because of increase in Daegu Subway fires in 2003)

Property loss accounts for ¥151.59 billion in 2003, 5.7% increased from ¥143.44 billion in 2002.

The leading causes of 2003 fires were electrical distribution involved, smoking, arson, ember and children playing etc. and they account for 65.5% of the total fires in 2002. Electrical distribution fires involved ranked first, occupied about 34% of all fires in 2003.

Twenty-seven percent(27%, 8,474) of the total 31,372 fires occurred in homes, including houses and apartments ranked first, vehicles fires ranked second or 19.3%(6,049), factory fires ranked third or 10.9%(3,416).

6. Catastrophic Fires for the Last 5 Year Period('99 - '03)

• 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 2.4% increase rate per year. But the 10 catastrophic fires in 2003 represented a 100% increase from 2002.

The 230 fire deaths in 2003 showed a increase of 191.1% and 422.7% respectively from 1999 and 2002, with a 98.8% increase per year for the last 5 year.

The 202 fire injuries in 2003 showed a 119.6% increase from 1999, and a increase of 1,246.7% from 2002, with a 222.1% increase per year for the last 5 year.

The Daegu Subway Fire(the death toll 192, the injured 148) cause a sudden rise of casualties in 2003

In 2003, property damage caused by fires amounted to ¥15,163 million, which marked an decrease of 3.9% from 1999, but an increase of 13,085.2% from 2002, with a 2,646.4% increase per year for the last 5 year.

II. Fires for the 10-Year Period ('94-'03)

1. Number of Fires for '94-'03

〔图〕插图

• According as an alteration of classificatory criterion by the source, it is to change as follows

• 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 5.6% increase rate per year since 1994. In 2003, total 31,372 fires occurred to be up by 42.3% from 1994. Fires in Structures marked an average 5.4% increase rate per year, and resulted in a 39.5% rise from 18,95 fires in 1994 to 25,235 fires in 2003. Vehicle(vehicles, ships, aircraft) fires have risen with an average 6.9% rate per year to mark a 55.4% increase rate in 2003 compared to 1994.

2. Fire Casualties for '94-'03

〔图〕插图

• TF: Total Fires, *IR: Increase Rate

▶ Increase trend in fire deaths

• Deaths in 2003 : increased average 3.7% per annum for the last 10-year period

• Numbers in 2003 compared to 2002 : 51.5% increase

• Numbers in 2003 compared to 1994 : 34.1% increase

▶ Increase trend in fire injuries

• Injuries in 2003 : increased average 6.1% per annum for the last 10-year period

• Numbers rate in 2003 compared to 2002 : 19.8% increase

• Numbers rate in 2003 compared to 1994 : 57.8% increase

1. Property Loss for '94-'03

〔图〕插图

• PL : Property Loss, * IR : Increase Rate

▶ Increase trend in property loss

• Increase trend for the last 10-year period : increased average 18.0% per annum

• Numbers in 2003 compared to 2002 : 5.7% increase

• Numbers in 2003 compared to 1994 : 14.3% increase

▶ Property loss order by places

• 1st : factories ¥48,007mil.(31.7% of total ¥151,590mil.)

• 2nd : residential ¥22,671mil.(15.0% of total ¥151,590mil.)

• 3rd : vehicles ¥17,212mil.(11.4% of total ¥151,590mil.)

▶ Property loss order by causes

• 1st : electrical fires W49,898mil.(32.9% of total W151,590mil.)

• 2nd : arson ¥15,151mil.(10.0% of total ¥151,590mil.)

• 3rd : embers ¥12,036mil.(7.9% of total ¥151,590mil.)

4. Fires by Cause for the 5-Year Period '99-'03

4-1. Electrical Fires

〔图〕插图 ▶ Fire trend

• 5-year : decreased average 0.3% per annum

• Numbers in 2003 compared to 2002 : 4.7% decrease

• Numbers in 2003 compared to 1999 : 4.8% decrease

▶ Fire casualties

• 5-year trend : deaths and injuries decreased average 0.7% and 0.1% per annum respectively

• Numbers in 2003 compared to 2002 : deaths and injuries downed by 12.6% and 2.7% respectively

• Numbers in 2003 compared to 1999 : deaths and injuries up by 28.8% and 21.3% respectively

▶ Property loss

• 5-year trend : decreased average 2.1% per annum

• Numbers in 2003 compared to 2002 : 13.9% decrease

• Numbers in 2003 compared to 1999 : 8.7% decrease

4-2. Smoking materials

〔图〕插图

$\blacktriangleright$ Fire trend

• 5-year : decreased average 2.5% per annum

• Numbers in 2003 compared to 2002 : 13.8% decrease

• Numbers in 2003 compared to 1999 : 22.1% decrease

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths and injuries increased average 29.6% and 13.1% per annum respectively

• Numbers in 2003 compared to 2002 : deaths and injuries up by 50.0% and 35.1% respectively

• Numbers in 2003 compared to 1999 : deaths and injuries up by 35.0% and 27.1% respectively

$\blacktriangleright$ Property loss

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

• Numbers in 2003 compared to 2002 : 16.6% decrease

• Numbers in 2003 compared to 1999 : 29.4% decrease

4-3. Arson

〔图〕16.0Fires14.0...··Fire Casualties(FC-.·-Property Loss(PL)12.010.0PL(in thousands8.06.0FC(in hundreds4.0Fires(in thousand $\blacktriangleright$ Fire trend

• 5-year : increased average 1.8% per annum

• Numbers in 2003 compared to 2002 : 15.9% increase

• Numbers in 2003 compared to 1999 : 32.4% increase

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths and injuries increased average 48.3% and 147.0% per annum respectively

• Numbers in 2003 compared to 2002 : deaths and injuries up by 243.8% and 75.7% respectively

• Numbers in 2003 compared to 1999 : deaths and injuries up by 264.6% and 103.7% respectively

※ The Daegu Subway Fire(the death toll 192, the injured 148) caused a sudden rise of casualties in 2003

$\blacktriangleright$ Property loss

• 5-year trend : increased average 14.9% per annum

• Numbers in 2003 compared to 2002 : 85.1% increase

• Numbers in 2003 compared to 1999 : 92.3% increase

4-4. Children Playing

〔图〕插图

$\blacktriangleright$ Fire trend

• 5-year : decreased average 7.6% per annum

• Numbers in 2003 compared to 2002 : 7.3% increase

• Numbers in 2003 compared to 1999 : 30.6% decrease

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths and injuries increased average 100.8% and 34.7% per annum respectively

• Numbers in 2003 compared to 2002 : injuries up by 8.8%

• Numbers in 2003 compared to 1999 : deaths and injuries downed by 86.2% and 65.4% respectively

$\blacktriangleright$ Property loss

• 5-year trend : decreased average 10.7% per annum

• Numbers in 2003 compared to 2002 : 20.2% decrease

• Numbers in 2003 compared to 1999 : 34.5% decrease

4-5. Embers

〔图〕插图 $\blacktriangleright$ Fire trend

• 5-year : increased average 4.9% per annum

• Numbers in 2003 compared to 2002 : 8.4% decrease

• Numbers in 2003 compared to 1999 : 7.9% increase

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths and injuries increased average 91.7% and 19.2% per annum respectively

• Numbers in 2003 compared to 2002 : deaths and injuries up by 550.0% 42.9% respectively

• Numbers in 2003 compared to 1999 : deaths and injuries up by 188.9% and 9.6% respectively

$\blacktriangleright$ Property loss

• 5-year trend : increased average 25.1% per annum

• Numbers in 2003 compared to 2002 : 65.8% increase

• Numbers in 2003 compared to 1999 : 56.5% increase

4-6. Oils

〔图〕插图

$\blacktriangleright$ Fire trend

• 5-year : decreased average 4.9% per annum

• Numbers in 2003 compared to 2002 : 0.8% increase

• Numbers in 2003 compared to 1999 : 3.5% decrease

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths decreased average 42.5% per annum and injuries increased average 3.5% per annum

• Numbers in 2003 compared to 2002 : no change in injury numbers

• Numbers in 2003 compared to 1999 : injuries downed by 9.1%

$\blacktriangleright$ Property loss

• 5-year trend : increased average 4.0% per annum

• Numbers in 2003 compared to 2002 : 4.5% decrease

• Numbers in 2003 compared to 1999 : 38.5% increase

4-7. Gases

〔图〕插图

$\blacktriangleright$ Fire trend

• 5-year : decreased average 11.5% per annum

• Numbers in 2003 compared to 2002 : 16.2% decrease

• Numbers in 2003 compared to 1999 : 44.4% decrease

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths increased average 15.4% per annum and injuries decreased average 8.7% per annum

• Numbers in 2003 compared to 2002 : deaths and injuries up by 50.0% and 22.4% respectively

• Numbers in 2003 compared to 1999 : deaths up by 8.0% and injuries downed by 35.0%

$\blacktriangleright$ Property loss

• 5-year trend : decreased average 16.6% per annum

• Numbers in 2003 compared to 2002 : 10.5% increase

• Numbers in 2003 compared to 1999 : 14.8% decrease

5. Fires by Property Type for the 5-Year Period '99-'03

5-1. Residential (Houses & Apartments)

〔图〕插图 $\blacktriangleright$ Fire trend

• 5-year : decreased average 2.86% per annum

• Numbers in 2003 compared to 2002 : 4.5% decrease

• Numbers in 2003 compared to 1999 : 14.7% decrease

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths and injuries increased average 5.6% and 3.0% per annum respectively

• Numbers in 2003 compared to 2002 : deaths and injuries up by 27.3% and 12.7% respectively

• Numbers in 2003 compared to 1999 : deaths and injuries up by 9.9% and 7.7% respectively

$\blacktriangleright$ Property loss

• 5-year trend : decreased average 3.4% per annum

• Numbers in 2003 compared to 2002 : 17.8% decrease

• Numbers in 2003 compared to 1999 : 23.7% decrease

5-2. Factories

〔图〕插图

$\blacktriangleright$ Fire trend

• 5-year : increased average 0.3% per annum

• Numbers in 2003 compared to 2002 : 3.5% decrease

• Numbers in 2003 compared to 1999 : 6.6% decrease

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths and injuries increased average 13.1% and 1.9% per annum respectively

• Numbers in 2003 compared to 2002 : deaths and injuries up by 4.2% and 14.4% respectively

• Numbers in 2003 compared to 1999 : deaths and injuries up by 31.6% and 1.8% respectively

$\blacktriangleright$ Property loss

• 5-year trend : increased average 4.6% per annum

• Numbers in 2003 compared to 2002 : 34.2% increase

• Numbers in 2003 compared to 1999 : 8.0% decrease

5-3. Vehicles

〔图〕插图

$\blacktriangleright$ Fire trend

• 5-year : increased average 2.4% per annum

• Numbers in 2003 compared to 2002 : 4.4% increase

• Numbers in 2003 compared to 1999 : 10.2% increase

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths and injuries increased average 77.4% and 45.3% per annum respectively

• Numbers in 2003 compared to 2002 : deaths and injuries up by 356.7% and 227.3% respectively

• Numbers in 2003 compared to 1999 : deaths and injuries up by 417.0% and 285.7% respectively

$\blacktriangleright$ Property loss

• 5-year trend : increased average 9.5% per annum

• Numbers in 2003 compared to 2002 : 33.4% increase

• Numbers in 2003 compared to 1999 : 64.5% increase

〔图〕插图

5-5. Shops

▶ Fire trend

• 5-year : decreased average 0.5% per annum

• Numbers in 2003 compared to 2002 : 12.9% increase

• Numbers in 2003 compared to 1999 : 2.1% decrease

▶ Fire casualties

• 5-year trend : deaths decreased average 11.4% per annum and injuries increased average 3.8% per annum

• Numbers in 2003 compared to 2002 : deaths downed by 72.2% and injuries up by 32.8%

• Numbers in 2003 compared to 1999 : deaths and injuries downed by 73.7% and 22.1% respectively

▶ Property loss

• 5-year trend : decreased average 4.1% per annum

• Numbers in 2003 compared to 2002 : 23.7% decrease

• Numbers in 2003 compared to 1999 : 19.9% decrease

5-6. Workplaces

〔图〕插图

$\blacktriangleright$ Fire trend

• 5-year : decreased average 2.4% per annum

• Numbers in 2003 compared to 2002 : 14.6% decrease

• Numbers in 2003 compared to 1999 : 16.3% decrease

$\blacktriangleright$ Fire casualties

• 5-year trend : deaths increased average 50.8% per annum and injuries decreased average 6.9% per annum

• Numbers in 2003 compared to 2002 : deaths and injuries downed by 50.0% and 55.3% respectively

• Numbers in 2003 compared to 1999 : injuries downed by 54.3%

$\blacktriangleright$ Property loss

• 5-year trend : increased average 14.4% per annum

• Numbers in 2003 compared to 2002 : 89.6% increase

• Numbers in 2003 compared to 1999 : 52.3% increase

5-7. Warehouses

〔图〕插图

$\blacktriangleright$ Fire trend

• 5-year : decreased average 2.1% per annum

• Numbers in 2003 compared to 2002 : 18.0% decrease

• Numbers in 2003 compared to 1999 : 22.0% decrease

$\blacktriangleright$ Fire casualties

• 5-year trend : injuries decreased average 6.8% per annum

• Numbers in 2003 compared to 2002 : deaths and injuries downed by 50.0% and 48.0% respectively

• Numbers in 2003 compared to 1999 : deaths and injuries downed by 50.0% and 27.8% respectively

$\blacktriangleright$ Property loss

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

• Numbers in 2003 compared to 2002 : 33.1% decrease

• Numbers in 2003 compared to 1999 : 16.4% decrease

5-8. Hotels & Motels

〔图〕插图

▶ Fire trend

• 5-year : increased average 3.5% per annum

• Numbers in 2003 compared to 2002 : 10.5% increase

• Numbers in 2003 compared to 1999 : 9.7% decrease

▶ Fire casualties

• 5-year trend : deaths and injuries increased average 11.2% and 19.7% per annum respectively

• Numbers in 2003 compared to 2002 : deaths downed by 10.7% and injuries up by 29.3%

• Numbers in 2003 compared to 1999 : injuries up by 26.0%

▶ Property loss

• 5-year trend : decreased average 3.8% per annum

• Numbers in 2003 compared to 2002 : 19.5% decrease

• Numbers in 2003 compared to 1999 : 35.9% decrease

5-9. Schools

〔图〕插图

$\blacktriangleright$ Fire trend

• 5-year : increased average 2.1% per annum

• Numbers in 2003 compared to 2002 : 1.2% decrease

• Numbers in 2003 compared to 1999 : 10.5% increase

$\blacktriangleright$ Fire casualties

• 5-year trend : injuries increased average 75.2% per annum

• Numbers in 2003 compared to 2002 : injuries up by 60.0%

• Numbers in 2003 compared to 1999 : injuries up by 1100.0%

$\blacktriangleright$ Property loss

• 5-year trend : decreased average 2.2% per annum

• Numbers in 2003 compared to 2002 : 0.1% increase

• Numbers in 2003 compared to 1999 : 13.9% increase

FIRE INVESTIGATION REPORT

Fire in lodging Houses

Fire in Lodging Houses

I. Current Status of Fire

As shown in Table 1, fire in lodging houses has increased by 3.5% in average for 5 years, the number of death increased by 11.2%, the number of injury increased by 19.7%. However, it is shown that the damage of property has reduced by 3.8% in average.

Table 1. Status of fire in lodging houses for last 5 years

(unit : person, million won)

II. Cause of Fire

As shown in Table 2, for the number of fires in hotels and inns by cause, the fire caused by electricity occupies 34%, the fire by smoking and arson shows 30% respectively. It is problematic that it is hard to prevent the fire by the negligence of guests' wrong doing such as unattended smoking or the intentional fire.

Table 2. Number of fire by cause

(2003)

Ⅲ. Accident Cases

□ Hotel A

1. General

Location : Pyeongtaek-si, Gyeonggi-do

· Date : 01:33, Jan. 11, 2005

· Origin : Rm. 305

· Property loss : 48,000 US dollar

· Casualties : 3 death, 6 injury

· Cause : flammable liquid

2. Overview

A hotel is a 4 story building with 1,876m $^{2}$ of total floor area, where the ground floor and 2nd floor are used as bathhouses, 3rd and 4th floors as guest rooms.

At the time of accident, fire fighters from 3 fire stations arrived at the hotel, but the fire has already reached at its peak and it was difficult to reduce the damage. Because of this fire, 3 guests including one man and two women who was sleeping in their room were suffocated by smoke to death and 210m $^{2}$ of 3rth floor was destroyed. A man and a woman of in their 50s were found in Rm. 305 and 307 on the 3rd floor and another woman found dead in the corridor. Eyewitnesses stated that flames shot up from the third floor after 3 times of "popping" sound were made.

Police investigators discovered the traces of flammable liquid were found in Rm. 305 of the 3rd floor.

〔图〕YTM\n

Photo 1. Firing telegraph pole by radiation heat

〔图〕插图

Photo2. Destroyed guest room

〔图〕O\n

Photo3. Traces of the fire spreading through windows

□ Hotel B

1. General

· Location : Yangsan-si, Gyeongsangnam-do

· Date : 00:35, Apr. 28, 2005

· Origin : Rm. 307

· Property loss : 34,000 US dollar

· Casualties : 9 injury

· Cause : Overheated gas burner

2. Overview

B Hotel is a building of 5 floors with 1,589 m $^{2}$ of total floor area, where the ground floor was used for parking lot and liquor store, the second floor for Karaoke, 3rd~5th floors for guest rooms.

It was investigated that the fire was caused by the radiation heat from overheated iron pot and moved to the floor cover and the surrounding combustible materials while a long-term guest was sleeping with the iron pot put on the portable gas burner for meal.

The fire spread over the surrounding combined with several processes as follows : Firstly, the fire spread to the corridor because the man in RM 307 could not close the door due to the strong flame during the evacuation after he had failed to extinguish the fire himself.

Secondly, the fire spread to the flammable materials including mattresses, chairs and drink boxes left in the fire exit and the stairway next to Rm. 307.

Thirdly, the fire expanded rapidly to the forth and fifth floors through the stairway because the door to the third floor stairway remained open.

〔图〕插图

Photo5. Floor of origin in Rm. 307

〔图〕-\n

Photo 6. Damaged passage of the 3rd floor

3. Problems and measures

Since lodging houses have closed spaces, they may become late to perceive the fire. Especially if a fire breaks out when there are guests sleeping, the loss of life may be huge because they become late to react and respond physically to the fire. Since it is hard to take preconditions for the accident due to smoking or arson, it is required to reduce the damages after the fire, apart from the preventive activities.

Meanwhile, people must wait calmly for the rescue while covering the crack in the door with wet towels if they can not escape the fire.

A. Fire facilities

Most lodging houses are equipped with alarm that operates by smoke or heat and large-scale buildings such as hotels are equipped with automatic extinguishing facilities such as sprinkler systems, but the damages often increased because the fire extinguishing facilities do not function. Accordingly, it is important to manage fire extinguishing facilities for rapid escape and early extinguishment.

B. Combustion preventing facilities

The fire protection division by floor or area is essential to reduce the damages by fire. And the stairway division is important because it prevents not only vertical

spread of flames or smoke but facilitates the escape.

Recently, the facilities are equipped for the prevention of fire expansion at some level, but the open places are often left while the occupancy of building is changed or the building is repaired. Accordingly, it is necessary to attach door checks to the fire protection door and take thorough actions for fittings and cooling/heating ducts where the piping and wires are installed because the division of them is insufficient as they are not peculiar places.

C. Escape facilities

If a fire breaks out and the one-direction fire exit is closed, people must be able to escape through another route. However, the fire exit is often closed for the convenience of management, which results in the damages of lives.

Accordingly, it is required to ensure 2 or more fire exits. Also, the exit stairway must be connected directly to the exit floor and has a structure that people can escape safely from a fire. Make sure not to use general elevators because they have no safety facilities against fire.

The escape route and guidance sign must be installed on guest rooms and corridors for emergency.

D. Combustible materials

Unlike offices, lodging houses have basically combustible materials including bedding, so they have conditions for the expansion of fire. Accordingly, it is necessary to limit the combustible materials as much as possible and not to leave them on the stair landing.

FIRE INVESTIGATION REPORT

Explosion in Chemical Plant

Explosion in Chemical Plant

1. General

· Location : Ulsan-si

· Date : 08:05, Sep. 20, 2004

· Origin : Polypropylene process

· Property loss : Process damage

· Casualties : 2 death

· Cause : Electric spark

2. Overview

Workers were operating the compressor in the panel of pressure explosion-proof structure for hydrogen compressor in the polypropylene process. Hydrogen flow into the electrical panel through the nitrogen purge line installed on the inlet of compressor and the explosion was caused by the electrical spark of the inner panel. Two workers working in front of the panel were hit by the cover of panel and died on the spot.

3. Materials related to Accident

$\blacksquare$ Characteristics of hydrogen gas

• CAS NO. : 1333 - 74 - 0

• NFPA Class : Fire = 4

• Combustion equation : $2H_2 + O_2 \rightarrow 2H_2O$

• LEL : 4.0%(Vol)

• Minimum Oxygen Concentration, MOC : 1/2 x 4% = 2%(Vol)

$\blacksquare$ Panel

• Outer dimensions: 1,000 x 1,000 x 600 (mm)

• Design pressure : $500mmH_2O$

• Used materials : STS304

• Set pressure of overpressure protection equipment : $250mmH_2O$

• Pressure alarm set value within the electrical panel

·Overpressure alarm: 300mmH2O

· Low pressure alarm : 30mmH₂O

• Panel cover of Pressure explosion-proof structure

· Dimensions: 3 x 1,000 x 1,000 (mm)

· Weight : About 60kg(including attachments)

$\blacksquare$ Nitrogen supply equipment

• Main supply pressure and pipe size : $7\text{kg/cm}^2$ x $1/2$ "

• Set pressure of 1st pressure regulator(RV1) : 3kg/cm$^{2}$

• Set pressure of 2nd pressure regulator(RV3) : 100mmH2O(for normal operation)

· Minimum nitrogen flow at normal operation : $14N \ell / min$

• Set pressure of 2nd pressure regulator(RV2) : 320mmH₂O(for internal replacement)

• Minimum flow at internal replacement : $380N \ell / min$

4. Occurrence of Fire

Five days before the accident, hydrogen compressor(A) was replaced by hydrogen compressor(B) for repair works. To replace 2 internal valves of the hydrogen compressor(A) under the maintenance plan, the blind was inserted into the flare line connected to the compressor cylinder block, the purge line connected to the cylinder block was shut off for the countercurrent prevention, and purge was executed by pressurizing the nitrogen (7kg/cm$^{2}$) through the purge line on the compressor inlet. Accordingly, we could not completely rule out the possibility that the shut-off valves(in 3 places) on the purge line were left open after purge or one leaked valve was shut off. At the day of accident, then preparation work was done to replace hydrogen compressor(A) with hydrogen compressor(B) for the

change of polypropylene products.

Since the relevant internal panel must maintain positive pressure to operate hydrogen compressor(A), nitrogen was supplied by opening all valves on the nitrogen line for positive pressure. It seemed that the leaked hydrogen started to flow into the electrical panel together with nitrogen on the existing nitrogen line at the time and the concentration of hydrogen within the panel was increased over time.

It seems that hydrogen flowed through the purge line on the inlet of compressor reached to LEL (4 vol%) or higher and formed the explosion atmosphere. It was analyzed that the concentration of hydrogen in the nitrogen supply piping was 5.5% about 7 hours after the accident.

Investigations on the shapes of damaged containers, pumps and piping in the site, found that the reactor among the equipment remained crashed by external impact but had no trace of internal explosion. It was reversed in the site with intermediate vessel and product vessel level gage attached.

5. Causes of Explosion

(1) Basis of Estimation

It is possible that the explosion has occurred by operating electrical switches in the condition that hydrogen flowed backward to the nitrogen line installed within the panel for positive pressure by the leakage of check valve on the purge line connected to the inlet of compressor and formed the explosion surroundings within the panel. The results of examining the matters about combustibles, ignition sources and oxygen flow by item are as follows:

① Explosives

It is likely that hydrogen(with pressure of 34~35kg/cm$^{2}$, 99.9%) thought to be cause of the explosion flowed backward to the nitrogen line for positive pressure within the panel of pressure explosion-proof structure through the purge line on the inlet of compressor due to the leakage of 2 check valves, the leakage of 1 shut-off valve and opening or uncertain closing of 2 shut-off valves and stayed within the panel to form the explosive concentration.

• It was found that hydrogen flowed in as the concentration of hydrogen was shown to be 5.4% and 2.84% respectively as a result of analyzing the components on the nitrogen purge line passed through 2 check valves and 2 shut-off valves at 15:00 and 17:20 after the accident. It is estimated that the concentration of hydrogen analyzed at 17:20 was lowered by the opening of valve in the first sampling.

② Ignition source

Since the electrical panel had a pressure explosion-proof structure and the electrical equipment within the panel was not explosion-proof, sparks could be caused by operating switches that have contacts within the panel.

③ Oxygen

It is possible that the valve of nitrogen line for positive pressure of electrical panel at the neighboring point was shut off to shorten the purge time when fulfilling nitrogen purge through the purge line on the inlet of compressor to replace 2 valves on the inlet of hydrogen compressor on September 15th, at this time nitrogen that remained within the panel leaked through spare nozzle of panel or the opening of inlet on the cable and oxygen in the air flowed in the internal panel instead, and such condition had been maintained until the morning on September 20th when hydrogen compressor(A) was operated again.

(2) Examination of site evidences

① The panel cover (weight of about 60kg) was flown away and left the trace of severe impact on the equipment at the point of 3.7m at a level.

② Comparing the trace of impact with the original location of panel, it is about 1.4° below at the vertical direction (about 6 cm below from the point of 3.7 m) and the panel was flown almost horizontally. Also the direction of flight was almost not changed by the feldspar attached to open the cover on the outside of panel (leaned about just 3° to the installed point of feldspar) with the feldspar broken.

③ Among 6 walls of electrical panel, the bottom side supported robustly by the section steel had slight bending, while on the rest of them the excessive

separation was caused and the tag welding part of internal bracket was mostly broken by instantaneous overpressure.

④ All of 12 volts tightening the cover and even the external feldspar were destroyed simultaneously.

⑤ Considering that the volts supporting both of steel plates on the lower panel were sheared and destroyed by the overpressure at the explosion, it seemed that the explosion was caused within the operating panel. Also, considering that the overpressure protection equipment remained completely damaged(broken), it supported that the rapid overpressure was caused.

6. Problems and Measures

(1) Thorough management

For the work that comes under the management of changing element, it is necessary to perform thorough the management of changing element, risk evaluation, data update, preparation of safe operation manual, worker training and per-operation check.

(2) Supplementing nitrogen piping and operating panel

It is necessary to install the blind of nitrogen piping for purge at the inlet of compressor on the rear of shut-off valve so that the blind can be installed during operation with separated from the nitrogen piping for positive pressure maintenance of operating panel. Also, it is strongly recommended to supply instrument air instead of nitrogen line.

• Install the shut-off valve on the purge nitrogen piping of $H_2$ Compressor cylinder block so that nitrogen for positive pressure of operating panel will be always supplied ;

• Install breeder valve on the top operating panel(set pressure: 200mmH₂O) so that nitrogen and dusts can be discharged in the work of internal panel replacement.

(3) Thorough pre-work check

It is necessary to check thoroughly the opening or closing of compressor and

various valves, defects and regular leakage of accessories related to various instruments before operation after repair or maintenance works done.

(4) Installation of recorder

It is necessary to install a recorder that will record the operation of equipment since it serves for the main control room to identify the situations before accident and the causes of accident after accident.

〔图〕插图

Photo 1. Exploded electrical panel

〔图〕Panel Open.2\n

Photo 2. Panel cover damaged by the impact after explosion(flew away 3.7m and bent by the clash with other equipment.)

FIRE INVESTIGATION REPORT

Gas Fire in a House

Gas Fire in a House

1. General

· Location : Pyeongtaek-si, Gyeonggi-do

· Date : 22:30, Feb. 28, 2004.

· Origin : House

· Property loss : Partial damage of kitchen by fire

· Casualties : 1 death

· Cause : Gas leaks

2. Overview

The flame spread over the surroundings as soon as the victim unknowingly turned on the gas stove while LPG(Liquified Petroleum Gas) was in the air, which caused serious burning to the victim who died later in the hospital during treatment.

The traces of fire were discovered all over the inside of kitchen located within a remodeled small room, and there was no trace of explosion(refer to photo 1 and 2).

〔图〕插图

Photo 1. Trace of fire on the ceiling in the room

〔图〕插图

Photo 2. Wall paper burned behind the gas stove

3. Occurrence of Fire

Explosion was not occurred because the gas did not leak enough to cause it. Later it was found to be a flash fire caused by flammable gas.

The results of investigation on the fire scene are as follows :

• There were traces in the kitchen that foodstuffs in the pan boiled over and were scorched.

• Gas stove operated normally and there was no gas leakage.

• However, the safety extinguishing device at the right fuel injection did not work.

• The intermediate valve was not installed and the whole piping was made of PVC hose.

• PVC hose was connected to the kitchen through the lattice window of the door.

• Gas leakage was not found on the connection of gas facilities and the burner itself. (refer to Photo 3)

At the time of accident, the victim wore only pants and was easily exposed to the heat of fire, who was diagnosed as 85% of generalized burn. (refer to Photo 4)

〔图〕插图

Photo 3. Burners in the gas stove

〔图〕插图

Photo 4. Pants of victim

4. Causes of Explosion

Fire pattern of accident site was found to be the fire by gas leakage, the safety extinguishing device of right burner in the gas stove did not work normally and the safety valve was found to have been transformed as a result of disassembling appraisal. (refer to Photo 5 and 6).

Accordingly, it is considered that the fire started as soon as the victim operated the ignition coke while he was not aware that the valve was open due to inexperienced operation of ignition coke when he used the gas stove. Electronic valve unit of safety regulator and its working principles are shown in Fig. 1.

〔图〕插图

Photo 4. The front of safety valve

(The upper valve was bent and did not work normally)

〔图〕插图

Photo 5. The side of safety valve

〔图〕Electric valve unitPushrodElectric magnetic bar Safety ValveThermocouplePushPilot flameto pilota piece of irongas\n

Fig 1. Safety valve structure of electronic valve unit

5. Problems and Measures

(1) Thorough management

The accident was not prevented because the safety device of gas burner did not work properly. Accordingly, it is necessary to replace timely old or failed gas burner. Since the safety check and publicity about gas facilities are insufficient in small towns or isolated places unlike large cities, the measures about this are required.

(2) Use of metal pipe

For the piping exceeding 3m, it is required to use metal pipe instead of vinyl hose and install intermediate valve between the piping and gas burner.

〔图〕outside streetstorehousebarngateroomkitchendoorrestroomkitchen garden\n

Fig 2. Plane figure of house

〔图〕350cmThe door was covered with sootThere was no damage to windowor glasses, but there were trace236cmof melting of bug-p

Fig 3. Plot plan of inside of kitchen

This is provided under the Agreement of Cooperation between the KFPA and the NFPA. Please feel free to contact Risk Management Department of KFPA if you have any inquires.

(E-Mail kfpa@kfpa.or.kr, Tel 82-2-780-8111, Fax 82-2-783-4094)

2004 KOREAN FIRE DATA

Published in March 2005

Edited by Risk Management 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 89-88874-45-5 93530