Skip to content
From Fire Safety Academy Wiki
Contents
  1. 1Definition and thresholds
  2. 2Share of fire deaths
  3. 3Long-term trends
  4. 4Recurring factors
  5. 5Notable fires
  6. 6Codes, priorities and societal risk
  7. 7See also
  8. 8Notes
  9. 9References

A catastrophic multiple-death fire is a single fire that kills several people at once. In the statistics of the US National Fire Protection Association (NFPA), a catastrophic fire kills at least five people in a residential structure or at least three in another fire.[1] Such fires draw most of the public attention given to fire, and many of them have led to major changes in building and fire codes.[2][3] Yet in the United States from 1991 to 2005 they almost always accounted for less than 10 percent of structure fire deaths, the exception being 2001.[1] Big multiple-death fires occur in places where many people gather, such as large office complexes, hotels and nightclubs, whereas most fire deaths occur one at a time in homes.[4]

Well-known examples are the 1980 MGM Grand Hotel fire in Las Vegas (85 deaths), the 1986 DuPont Plaza Hotel fire in Puerto Rico (98) and the 2003 Station nightclub fire (100).[5][6][7] Recurring factors are inadequate or blocked exits, overcrowding, combustible finishes, late warning and the absence of sprinklers.[8][9] A 2005 NFPA report quoted by Huggins (2008) found no record of a fire killing more than two people in a completely sprinklered building with a properly operating system. It excepted explosions, flash fires and deaths of fire brigade members or employees during fire suppression.[10]

Definition and thresholds

Hirschler (2008a) calls the NFPA definition "relatively inclusive": a catastrophic fire is a residential structure fire in which five or more people die, or any other fire in which three or more people die.[1] The threshold has changed over time. In the period 1961–1980, the threshold for a multiple-death fire was three deaths in any property; by 2008 it was five deaths in residential properties.[8] NFPA's incident record keeping is best for fires in which 10 or more people died, and Hall and Cote (2008) use these as the principal measure of the effect of codes and standards.[11] As a rule, only incidents with 50 or more deaths are identified by name in NFPA statistical work.[11]

NFPA's data on these fires come from its fire experience survey and its Fire Incident Data Organization (FIDO).[1] Ahrens and Stewart (2008) reported confidence that the data were complete because NFPA made every effort, including surveys of state fire marshals' offices, to track down multiple-death fires.[12] In contrast with property loss data, no comparable omissions occur among multiple-death fires.[4] In risk analysis, "catastrophic" is also used as a severity class. A table from NFPA 551, reproduced by Hall and Watts (2008), defines a catastrophic fire as one that will produce death, multiple deaths or injuries, or a disastrous impact on operations.[13]

Share of fire deaths

Catastrophic fires make up a small part of US structure fire deaths in fire statistics. In US structure fires of 1991–2005, they caused between 98 and 320 deaths a year, or 3.2 to 8.0 percent of all structure fire deaths, in every year except 2001.[1] In 2001 the share rose to 51.9 percent, because the attacks of 11 September 2001 caused 2,451 of that year's 5,671 structure fire deaths.[1] In those years other than 2001, more than 90 percent of US structure fire deaths occurred in fires below NFPA's catastrophic-fire threshold. Fires with at least 25 deaths are extremely rare: Hirschler (2008a) reports five such incidents in 1991–2005, accounting for 5 percent of structure fire deaths, most of those deaths in the 2001 attacks. His table names six separate fire locations, listing the World Trade Center and Pentagon on one 11 September line.[1][14]

Bar chart of the yearly share of US structure fire deaths in catastrophic fires, 1991 to 2005, with 2001 off the scale
Share of US civilian structure fire deaths that occurred in catastrophic fires, 1991–2005; 2001 includes the attacks of 11 September.

Big multiple-death fires occur in high-occupancy places such as large office complexes, hotels or nightclubs. Most fire deaths, by contrast, occur in low-occupancy places such as dwellings and individual apartments, and occur as single-fatality fires.[4] The same holds for vehicles: big multiple-death fires may involve airliners or large ships, while most vehicle fire deaths occur in cars and trucks.[4] Hall (2008) reported that homes and private vehicles accounted for more than 90 percent of US fire deaths.[15] The overall US fire death rate in 2004 was 13 civilian fire deaths per million population.[16]

The pattern is not limited to the United States. Hirschler (2008a) notes that in the United States and other countries the vast majority of fire deaths occur in private homes and vehicles, where the public feels safest and fire safety requirements are minimal.[14] Van Rossum (2027) likewise observes that most fatal fire victims die at home.[17] Per hour of stay, the chance of dying in a fire is much greater in a residential building than in an assembly building. The Dutch detection requirements are nevertheless stricter for assembly buildings than for residential ones.[18]

Long-term trends

The deadliest fires have become less frequent. From 1900 through 1954, the United States had 44 fires and explosions with death tolls of 100 or more. From 1955 to the time of writing there were only five. They were a Kentucky restaurant fire in 1977, the Oklahoma City bombing of 1995, a Florida in-flight fire in 1996, the World Trade Center attack of 2001 and the Station nightclub fire of 2003.[19] Hall and Cote (2008) describe this as a decline from nearly one fire of that severity per year to less than one per decade.[19]

Stacked bar chart of the death tolls of the three deadliest US fires and explosions in each decade from 1900 to 2005
Deaths in the three deadliest US fires and explosions of each decade, 1900–2005.

This decline can be expressed as an exceedance frequency, the frequency of events with at least a given number of deaths. Such frequencies are the basis of the F-N curves used for societal risk, which plot the cumulative frequency F of events against the number of fatalities N.[20] As an illustration, a simple estimate from a historical record is the number of events divided by the length of the period:

F(N)=nNT

Symbol Description Unit
F(N) Frequency of events with at least N deaths per year
nN Number of events with at least N deaths in the period –
T Length of the period year

As a worked example, for N = 100 the 44 events of 1900–1954 give F = 44/55 = 0.80 per year. Using the roughly 50-year period from 1955 through the mid-2000s gives roughly 5/50 = 0.1 per year, about one event per decade.[19] For US structure fires of 1991–2005 and N = 25, counting the six fires named in Hirschler's table separately gives F = 6/15 = 0.40 per year.[1] The two estimates cover different scopes: the first includes all fires and explosions, the second only structure fires.

The record is dominated by a few types of setting. Most of the 101 US fires and explosions that killed 50 or more people from 1900 to 1999 were mining incidents, 59 in all.[21] Seven of the 11 deadliest single-building fires and explosions in US history occurred in places of assembly.[8] In almost every decade, the deadliest fires include fires that are not building fires, from ship fires early in the twentieth century to airliner fires at its end.[22] For fires this severe, only decade-by-decade comparisons can separate real trends from random variation.[4]

Recurring factors

The same factors appear in many places of assembly. More than a quarter of the 746 occupants of the Rhythm Club in Natchez, Mississippi, died in 1940. At the Cocoanut Grove nightclub in Boston in 1942, nearly half of the more than 1,000 occupants died. Overcrowding, inadequate exits and highly combustible decorations were cited in both fires, with roughly 200 victims trapped behind one revolving door at the Cocoanut Grove.[8] The 1977 Beverly Hills Supper Club fire in Kentucky killed 165 people among 2,400 to 2,800 occupants, two to three times the facility's capacity.[8] A separate pattern is arson at the normal exit: in the 1990 Happy Land social club fire in New York City, only six of the 93 occupants survived.[23]

Diagram linking recurring factors of multiple-death fires to the fires in which they were cited and to code responses
Recurring factors in US multiple-death fires, examples and code responses.

Fast fire spread over combustible surfaces is another common element. In New Zealand, the 2013 commentary on the Building Code fire clauses cites the Cocoanut Grove, MGM Grand and Station Night Club fires as showing how quickly fire can spread on combustible surface material. Because the code cannot control the contents of a building, its Clause C3.4 places controls on the surface finishes in certain areas instead.[24] Late warning is a third element. Purser (2002) identifies a failure in early detection and effective warning as a common feature of many major fire disasters. In most of the cases he lists, evacuation began late and accurate, authoritative instructions were not given.[9]

Hall and Cote (2008) list the problems with detection and alarm most often cited in major fires with large loss of life. They are the absence of needed systems, misapplication of equipment, lack of maintenance and improper response by occupants to the alarm.[25] Facilities for vulnerable people have had their own series of such fires. From 1970 to 1973, eight fires in facilities caring for the aged each killed at least 10 people, 112 in total. From 1979 to 1984, residents of board and care homes had a risk of dying in a multiple-death fire five times that of residents of other residential properties.[5]

Code compliance makes a large difference. In places of assembly, Hall and Cote (2008) find that the post-Cocoanut Grove code changes worked: large-loss-of-life fires have not occurred where code compliance was observed, except in rapid-onset situations such as explosions.[23] In hotels and motels, NFPA statistics for 1988–1997 showed sprinklers cutting the chance of dying in a fire by 91 percent.[8] Hall and Cote (2008) conclude that people who die in fires do so in the kinds of fires that codes do not reach. Otherwise they die in properties that codes do not reach, because the codes are not adopted or not enforced.[23] Locked and blocked exits, for example, contributed to the 25 deaths in a North Carolina chicken processing plant fire in 1991.[26]

Notable fires

The MGM Grand Hotel fire in Las Vegas on 21 November 1980 killed 85 people.[5] It was discovered at about 7:10 a.m. in an unoccupied restaurant on the casino level, reached flashover and spread through the casino. Heat and smoke then travelled through seismic joints, elevator shafts and stairways to the 21 residential floors, and guests were not alerted by the public address or alarm system.[27] According to one account, 619 people were taken to hospitals, and about 300 were evacuated by helicopter. It was the second most serious hotel fire in US history, after the Winecoff Hotel fire of 1946 with 119 deaths.[28]

The fire ran counter to the industry view that fire problems belonged to older or poorer facilities. It inspired an industry response that combined widespread code compliance with fire safety provisions that often went beyond the codes.[5][8] In 1980, sprinklers were reported present in only one of every nine hotel or motel fires; the latest data cited by Hall and Cote (2008) showed them in 40 percent.[8]

The DuPont Plaza Hotel fire in San Juan, Puerto Rico, in December 1986 killed 98 people in the hotel casino. The fire was set by arsonists in a ballroom across a wide stair from the casino. Hotel staff tried to put it out, but casino occupants were not warned.[6]

Schematic plan of the DuPont Plaza casino with the fire in the ballroom, blocked main doors and an inward-opening rear door
Schematic of the 1986 DuPont Plaza Hotel casino fire: flashover blocked the main entrance, leaving one rear door that opened inward.

Within an estimated 12 minutes the fire flashed over and blocked the main doors, leaving a single rear door that opened inward. Fleeing victims piled up against it, and virtually all the victims were found there.[6] With other hotel fires, it led to the US Hotel and Motel Fire Safety Act of 1990. Fruin (2008) states that the law required federal employees travelling on official business to stay in accommodations with hardwired smoke alarms and, above three storeys, sprinklers.[6]

The Station nightclub fire in West Warwick, Rhode Island, in February 2003 killed 100 people and injured more than 180. Pyrotechnics used by a band ignited soundproofing behind the stage, and a fire official stated that the club was fully involved within 3 minutes.[7] Most occupants headed for the front entrance and jammed it, although three other exits were usable; their exit signs were obscured by smoke.[7] Hall and Cote (2008) note that the fire involved materials that promoted rapid spread, an unusual heat source in indoor fireworks, and no sprinklers.[8] As a result, the 2006 edition of the NFPA Life Safety Code required automatic sprinklers in all new nightclubs with an occupant load of 50 or more and in existing ones with 100 or more.[23]

Codes, priorities and societal risk

Catastrophic fires have shaped fire codes for more than a century. Four building fires of 1903–1911, including the Iroquois Theatre and the Triangle Shirtwaist Factory, were largely responsible for the appointment in 1913 of the NFPA Committee on Safety to Life. A series of multiple-death fires in the 1940s led to major changes in the Building Exits Code, as NFPA 101 was then known.[3] A 1958 school fire in Chicago, which killed 95 people, 92 of them children, drew a quicker and more sweeping response than any other single US fire. In the following four decades, no US school fire killed 10 or more people.[11]

Hirschler (2008a) observes that this reactive pattern has limits. Code changes after a single tragedy often do not fully take into account the probability of such a fire, and they tend to address a narrow issue in a prescriptive way.[1] Defining fire safety priorities mainly by reacting to major incidents is, he adds, not unique to the United States.[14] Newsworthiness is not the same as severity: the 1963 Golden Age Nursing Home fire killed 63 people but drew little attention, because it occurred the day after the assassination of President Kennedy.[4]

The choice of measure shapes priorities. Hall (2008) notes that fire deaths can be counted as deaths, fatal fires, multiple-death fires or deaths in multiple-death fires, and that counting fatal fires reduces the weight of multiple-death fires.[29] In Hall's hypothetical example, the MGM Grand fire caused more deaths than all other Las Vegas fires of many years combined. Counting deaths alone would have sent a Las Vegas fire educator's resources to preventing another such fire. In his view, a single event of historic severity can justify expanded attention, but not exclusive attention.[29] The public often seems far more upset by one fire that kills five people than by five fires that kill one person each.[29]

Risk analysis treats this concern as societal risk. The Australian Building Codes Board describes societal risk as the risk from hazards that can affect large numbers of people at nearly the same time. It is the relationship between frequency and the number of people suffering a given level of harm.[30] Its 2020 handbook calls for consideration of both individual and societal risk. It says that a performance solution should offer at least the safety level of a reference building meeting the deemed-to-satisfy provisions.[30]

In one example in that handbook, an apartment building with a single stair and longer travel distances is compared with a two-stair reference. A badly chosen reference would hide a doubling of the population at risk and of fire starts, with a significant effect on societal risk.[31] A table from NFPA 551 reproduced by Hall and Watts (2008) lists quantitative fire risk methods that include plots of frequency against number of fatalities and the determination of societal risk to the entire population.[32]

See also

Notes

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Hirschler (2008a), p. 6-36.
  2. ↑ Hall and Cote (2008), p. 3-8.
  3. ↑ 3.0 3.1 Hall and Cote (2008), p. 3-22.
  4. ↑ 4.0 4.1 4.2 4.3 4.4 4.5 Ahrens et al. (2008), p. 3-87.
  5. ↑ 5.0 5.1 5.2 5.3 Hall and Cote (2008), p. 3-24.
  6. ↑ 6.0 6.1 6.2 6.3 Fruin (2008), p. 4-97.
  7. ↑ 7.0 7.1 7.2 Fruin (2008), p. 4-98.
  8. ↑ 8.0 8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 Hall and Cote (2008), pp. 3-25–3-26.
  9. ↑ 9.0 9.1 Purser (2002), p. 2-142.
  10. ↑ Huggins (2008), p. 16-32.
  11. ↑ 11.0 11.1 11.2 Hall and Cote (2008), p. 3-23.
  12. ↑ Ahrens and Stewart (2008), p. 3-57.
  13. ↑ Hall and Watts (2008), p. 3-141.
  14. ↑ 14.0 14.1 14.2 Hirschler (2008a), p. 6-37.
  15. ↑ Hall (2008), p. 5-107.
  16. ↑ Ahrens et al. (2008), p. 3-80.
  17. ↑ Van Rossum (2027), §7.
  18. ↑ Van Rossum (2027), §1.2.
  19. ↑ 19.0 19.1 19.2 Hall and Cote (2008), p. 3-5.
  20. ↑ Barry (2002), p. 5-210.
  21. ↑ Hall and Cote (2008), p. 3-27.
  22. ↑ Ahrens et al. (2008), p. 3-84.
  23. ↑ 23.0 23.1 23.2 23.3 Hall and Cote (2008), p. 3-26.
  24. ↑ Ministry of Business, Innovation and Employment (2013), p. 19.
  25. ↑ Hall and Cote (2008), p. 3-18.
  26. ↑ Demers (2008a), p. 20-204.
  27. ↑ Bryan (2008), p. 4-20.
  28. ↑ Bryan (2008), p. 4-21.
  29. ↑ 29.0 29.1 29.2 Hall (2008), p. 5-106.
  30. ↑ 30.0 30.1 Australian Building Codes Board (2020b), p. 22.
  31. ↑ Australian Building Codes Board (2020b), pp. 46–47.
  32. ↑ Hall and Watts (2008), p. 3-140.

References

  • Ahrens, M., Frazier, P., & Kelch, G. (2008). Use of fire incident data and statistics. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-61–3-92). National Fire Protection Association.
  • Ahrens, M., & Stewart, S. (2008). Fire data collection and databases. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-43–3-59). National Fire Protection Association.
  • Australian Building Codes Board. (2020b). Handbook: Fire Safety Verification Method (Print version 1.1).
  • Barry, T. F. (2002). Quantitative risk assessment in chemical process industries. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 5-176–5-213). National Fire Protection Association; Society of Fire Protection Engineers.
  • Bryan, J. L. (2008). Human behavior and fire. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 4-3–4-47). National Fire Protection Association.
  • Demers, D. P. (2008a). Industrial occupancies. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 20-203–20-210). National Fire Protection Association.
  • Fruin, J. J. (2008). Techniques of crowd management. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 4-93–4-102). National Fire Protection Association.
  • Hall, J. R., Jr. (2008). Using data for public education planning and decision making. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 5-103–5-118). National Fire Protection Association.
  • Hall, J. R., Jr., & Cote, A. E. (2008). An overview of the fire problem and fire protection. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-3–3-30). National Fire Protection Association.
  • Hall, J. R., Jr., & Watts, J. M., Jr. (2008). Fire risk analysis. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-135–3-143). National Fire Protection Association.
  • Hirschler, M. M. (2008a). Concepts and protocols of fire testing. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 6-35–6-60). National Fire Protection Association.
  • Huggins, R. (2008). Automatic sprinkler systems. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 16-29–16-53). National Fire Protection Association.
  • Ministry of Business, Innovation and Employment. (2013). Commentary for Building Code clauses C1–C6 and Verification Method C/VM2.
  • Purser, D. A. (2002). Toxicity assessment of combustion products. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 2-83–2-171). National Fire Protection Association; Society of Fire Protection Engineers.
  • Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.