1. The analysis in this chapter is restricted to commercial airliners. The FAA regulates small commuter airlines and private aircraft separately. [BACK]

2. A few major airlines have safety departments. Several of these airlines conduct occasional safety tests and participate in the development of some nongovernment safety standards. But most airlines do not do any of these things. The explanation of these differences is an important question for future research. [BACK]

3. Federal Aviation Administration, Federal Aviation Regulation (F.A.R.) 121.309(c). See, generally, 14 C.F.R., part 121. [BACK]

4. FAA, "Hand Fire Extinguishers in Transport Category Airplanes and Rotocraft," Advisory Circular 20-42, September 1, 1965. [BACK]

5. National Transportation Safety Board, Recommendation A-73-067 (closed on May 3, 1979, following "acceptable action" by the FAA). [BACK]

6. FAA response (May 2, 1974) to NTSB recommendation A-73-067. [BACK]

7. FAA, Advisory Circular 20-42A, July 29, 1980. [BACK]

8. FAA, General Notice 8000.212, November 29, 1980. [BACK]

9. Letter from W. B. Durlin, vice president, engineering and quality control, Frontier Airlines, to Sheldon King, FAA air carrier maintenance inspector, Aurora, Colo., December 16, 1983. Obtained through Freedom of Information request for correspondence concerning compliance with FAA General Notice 8000.212. [BACK]

10. Boeing apparently concluded that Halon 1211 worked better than water extinguishers in all simulated fire tests but a large newspaper fire. The company did not officially release the results of its tests, but an unofficial account appears in L. M. Krasner, Factory Mutual Research Corporation, Study of Hand-Held Fire Extinguishers Aboard Civil Aviation Aircraft, DOT/ FAA/Ct-82/42 (Springfield, Va.: National Technical Information Service, 1982), 20. [BACK]

11. Hal Quinn, "A Burning Nightmare at 31,000 Feet," MacLean's, June 13, 1983, p. 19. [BACK]

12. The NSTB is an independent agency charged with investigating transportation accidents and recommending regulatory changes. Its reports are often highly influential. NFPA, in cooperation with the U.S. Fire Administration and the National Bureau of Standards, investigates major fires of all sorts for the purpose of improving private standards. [BACK]

13. See, generally, Thomas J. Klem, National Fire Protection Association, "Investigation Report: Air Canada DC-9 Aircraft Fire, Greater Cincinnati Airport, June 2, 1983" (photocopy, n.d.). [BACK]

14. National Transportation Safety Board, Recommendations A-83-71 (thermal discharge extinguishers for lavatories) and A-83-72 (Halon extinguishant), forwarded to J. Lynn Helms, administrator, Federal Aviation Administration, October 31, 1983. [BACK]

15. The three committees were the House Subcommittee on Transportation, Aviation, and Materials; the House Subcommittee on Investigations and Oversight of the Public Works and Transportation Committee; and the Senate Labor and Human Resources Committee. [BACK]

16. U.S. House, Committee on Public Works and Transportation, Legislation to Improve Airline Safety, Hearings, July 26, August 1-2, 1984 (98-59) (Washington, D.C.: Government Printing Office, 1984), 161. [BACK]

17. See, for example, H.R. 3264, 98th Cong., 2d sess. (requiring smoke detectors); H.R. 3793, 98th Cong., 2d sess. (requiring a "comprehensive, manually operated fire extinguishing system"); H.R. 5518, 98th Cong., 2d sess. (requiring high-buoyancy life vests). [BACK]

18. U.S. House, Committee on Public Works and Transportation, Legislation to Improve Airline Safety, 2 (Congressman Mineta). [BACK]

19. James Ott, "Smoke Detectors Start Conflict," Aviation Week and Space Technology, September 24, 1984. [BACK]

20. The first digit in the number represents the number of carbon atoms in the compound molecule; the second digit, the number of flourine atoms; the third digit, the number of chlorine atoms; and the fourth digit, the number of bromine atoms. [BACK]

21. To qualify for a Class A rating from UL, it is necessary to meet minimum performance criteria on an excelsior fire, a wood crib fire, and a wood panel fire. The excelsior fire test, for example, involves six pounds of fuel distributed over a 2'10" x 5'8" test area. The wood crib test involves extinguishment of a crib composed of fifty wood members arranged in ten layers of five members each. Specific details for test construction, arrangement, ignition, and fire attack strategy are contained in UL 711. [BACK]

22. Charles Perrow disagrees, arguing that "we simply do a lot of driving and very little flying, giving us the impression that the risk of the latter is much smaller. In terms of exposure (fatalities per hour) they may be very similar" (Normal Accidents: Living with High-Risk Technologies [New York: Basic Books, 1984], 126). While Perrow is correct that comparing absolute numbers of fatalities may be misleading, he is wrong in attempting to remedy the problem by comparing the number of fatalities per hour. That measure would make sense only if airplanes and automobiles traveled equally fast. An hour of "exposure" to airplane travel is not commensurate in any meaningful sense with an hour of car travel. An hour of airplane travel should be compared with the "exposure" necessary to accomplish the same distance in a car, since that is the choice faced by an individual. Under this measure, commercial airplane travel would have to be considered at least five times safer than car travel. [BACK]

23. Robert J. Johansen, "Using Statistics in Aviation Underwriting," Best's Review (Prop/Casualty), February 1982, 42. [BACK]

24. U.S. House, Committee on Public Works and Transportation, Legislation to Improve Airline Safety, 58. [BACK]

25. See, for example, Executive Order 12291, February 17, 1981. [BACK]

26. For an explanation of the $350,000 value placed on a statistical traveler's life, see FAA, Office of Aviation Policy, Economic Values for Evaluation of Federal Aviation Administration Investment and Regulatory Programs (Springfield, Va.: National Technical Information Service, 1981), 12-27. [BACK]

27. Federal Register 49 (May 17, 1984): 21010 21015 (notice of proposed rulemaking). [BACK]

28. A phenomenon is characterized by a Poisson distribution if the average number of events during a given time period can be estimated from past experience. Further, if the time period is subdivided into much smaller intervals, the probability that an event will occur in such a small interval is extremely low and remains constant for every small interval, and the probability that two or more events can occur within such a small interval is so low that it can be assigned a probability of zero. Finally, the occurrence of an event within any small interval is independent of the occurrence of events in other small intervals and is also independent of when that small interval occurs within the larger time period. [BACK]

29. Thomas Hopkins, "Economics of Safety" (Paper presented to the Flight Safety Foundation Conference on Cabin Safety, Washington, D.C., December 13, 1984), 154. [BACK]

30. The error has been corrected in subsequent FAA analyses. [BACK]

31. The Federal Aviation Regulation on Training States that each crew member must "operate" an extinguisher in training exercises during initial training and every twenty-four months thereafter. According to a report commissioned by the FAA: "Actual practice among the carriers varies broadly from brief discharge of agent on a makeshift fire to passing empty extinguishers around with individuals operating the discharge assembly" (Krasner, "Study of Hand-held Fire Extinguishers Aboard Civil Aviation Aircraft," 44). [BACK]

32. H. M. Hollingsworth, U.S. Coast Guard Research and Development Center, Obstructured Fire Tests on Hand-held Fire Extinguishers for Recreational Boats, CG-D-133-75 (Springfield, Va.: National Technical Information Service, 1975). [BACK]

33. The FAA employees do not both officially "represent" the FAA. NFPA, concerned about the "balance" of its committees, resisted the idea of having two FAA representatives on one committee. Through a compromise, an employee of the Airports Division participates as an individual, not as a representative of the FAA. He attends not so much to influence NFPA standards as to gain technical information and contacts helpful in his job at the FAA. [BACK]

34. One representative was from the FAA's Research Facility. His primary interest was in the arcane dispute (described later) about various Halon extinguishing agents. The other was, as noted, from the Airports Division. He considered NFPA 408 entirely outside his area. "I am only concerned with fire extinguishers if they are inside an airport facility," he explained. [BACK]

35. There are seventeen other NFPA aviation safety standards covering such items as aircraft hangars, airport terminals, loading walkways, and aircraft engine-testing facilities. [BACK]

36. National Fire Protection Association, Aircraft Hand Fire Extinguishers, NFPA 408-1956 (Boston: NFPA, 1956), sec. 110 (scope), p. 6 (the various versions of this document are hereafter referred to as NFPA 408, followed by the year of publication). [BACK]

37. The same engineer describes an intensive battle in the mid 1970s to get Boeing to consider supplying aircraft with Halon fire extinguishers. "For years Boeing was married to dry chemical," so it was not until the airline presented its own test data on the effectiveness of Halon that they made headway with Boeing. [BACK]

38. An amendment to the 1970 version provided an exception to the general rule that extinguishers shall be carbon monoxide or water: "Extinguishers utilizing halgonated extinguishing agents classified by the Underwriter's Laboratories, Inc., as failing in UL Toxicity Group 5 or 6 may be substituted for the carbon dioxide extinguishers if it can be shown that there is sufficient free-air volume within the aircraft cabin space to avoid producing serious irritating effects on the occupants" (NFPA 408-1973, sec. 312, p. 5). [BACK]

39. Sale of publications (primarily standards) accounted for almost $14 million in 1984, close to two-thirds of NFPA's income (NFPA Balance Sheet, Fire Journal, May 1985, 82). [BACK]

40. NFPA 408-1980, sec. 3-1.3, p. 7. [BACK]

41. Ibid., sec. A-2-3-4.1, p. 10. [BACK]

42. For example, as discussed in chapter 4, if NFPA allows only metal chimneys passing the 2100° F fire test, UL will have to change its standard accordingly. [BACK]

44. NFPA 408-1980, sec. A-5-3.3. [BACK]

45. National Fire Protection Association, Technical Committee Documentation: 1984 Annual Meeting (Quincy, Mass.: NFPA, 1984), Comments 408-8-(2-3.3.1 [New]), 408-9-(2-3.6), and 408-10-(table 3-1.1), p. 5. [BACK]

46. See, generally, Yair Aharoni, The No-Risk Society (Chatham, N.J.: Chatham House, 1981). [BACK]

47. "Statement by Congressman Norman Y. Mineta," in Proceedings of Cabin Safety Conference and Workshop: December 11-14, 1984, DOT/FAA/ ASF100-85/01 (Springfield, Va.: National Technical Information Service, 1985), 9. [BACK]