Logic dictates that to best prepare for future hazardous-materials transportation accidents, we should analyze past incidents. But since thousands of such accidents occur each year, analysis can be difficult. There are, however, relatively few multiple-fatality transportation accidents due to hazardous-materials releases. The U.S. Department of Transportation (DOT) has maintained records on such incidents in interstate commerce since 1970.

The DOTs Research and Special Programs Administration (RSPA) has 240 reports, covering 369 deaths, in its files tor the period from 1970 to 1990 The following, based on RSPA records, are 20 major, multiple-fatality’ transportation-related hazardous-materials accidents for the period covering 1970 to 1990. involving 120 deaths (the incidents have been assigned letters to aid in analysis).


  • A. June 4, 1971: six killed in Waco, Georgia, in an explosives fire.
  • B. November 5, 1971: three killed in Chicago, Illinois, by gasoline fire during unloading into a storage tank.
  • C. November 3, 1973: three killed at Logan Airport in a Pan Am plane crash caused by leaking nitric acid cargo.
  • D. January 11, 1974: four killed in West St. Paul, Minnesota, by propane fire during its unloading into a storage tank.
  • E. July 19, 1974: seven killed in Decatur, Illinois, in a liquefied petroleum gas (I.PG) railroad tank car derailment and fire.
  • F. August 2, 1974: three killed on Wyoming Highway 789 in a gasoline cargo tank fire.
  • G. February 26, 1975: three killed in a Stanton, California, fire that occurred while gasoline was being pumped out of an underground storage tank.
  • H. April 29, 1975: 15 killed in Eagle Pass, Texas, by an LPG cargo tank failure and fire.
  • I. May 11, 1976: five killed in Houston, Texas, by an anhydrous ammonia cargo tank failure.
  • J. July 9, 1976: three killed in Tazewell, Virginia, by a train/gasoline cargo tank fire.
  • K. December 15, 1976: three killed in Marland, Oklahoma, by a train/crude oil cargo tank fire.
  • L February 14, 1978: eight killed in Chicago, Illinois, by a corrosive liquid storage tank unloading and mixing accident.
  • M. February 22, 1978: 15 killed in Waveriy, Tennessee, when an LPG railroad tank car derailed, with subsequent
  • failure and fire.
  • N. February 26, 1978: eight killed in Youngstown, Florida, when a railroad tank car carrying chlorine derailed, with subsequent tank failure.
  • O. August 7, 1980: four killed in Oasis, California, in a JP4 cargo tank fire.
  • P. September 15, 1981: seven killed in Huntsville, Alabama, by a train/gasoline cargo tank fire.
  • Q. November 14, 1981: eight killed west of Canon City, Colorado, in a gasoline cargo tank fire.
  • R. June 30, 1986: three killed in Grand Rapids, Michigan, in a gasoline cargo tank fire.
  • S. December 23, 1988: eight killed in Memphis, Tennessee, in an LPG cargo tank failure and fire.
  • T, December 27, 1989: four killed in Claremont, California, in a gasoline cargo tank fire.

Note: The infamous 1973 Kingman, Arizona, and 1988Kansas City, Missouri, incidents, both of which resulted in firefighter deaths, technically were out of the transportation system and thus not included in the RSPA fatal accident reports.


Since we are attempting to prepare ourselves for future emergencies, we need to determine the skills that were necessary to handle the incidents listed above. As part of this process, we can eliminate from consideration those incidents in which involvement of the hazardous material did not in itself require action by public safety agency responders. A clearcut example is the Logan Airport Pan Am crash caused by a nitric acid leak. Even though the leak disabled the airplane crew and caused the crash, a nitric acid leak was not a consideration for the fire department—the plane crash was.

If hazardous-material involvement is not a consideration for responders, we will eliminate the incident from our analysis. In incidents A, B, C, and G, for instance, the hazardous material that caused the incident was not a problem for the emergency responders who subsequently came to the scene. This leaves 354 deaths to be analyzed.


Next, we will categorize the incidents according to the common skills or knowledge required to handle them satisfactorily. For example, JP4 can be handled the same way as gasoline, for all practical purposes. Thus, incident O can be grouped with incident P, since there was no real difference in the abilities required of the emergency responders facing the two incidents. Incident K could be included in the same category with incidents Q and P, since again, relatively similar techniques would be used to handle a crude oil cargo tank fire and a gasoline cargo tank fire. The categories could be grouped as follows:

Category 1: Incidents F, J, K, O, P, Q, R, and T involved flammable/combustible liquids transported in highway cargo tanks. As noted, a relatively similar approach is called for whether the flammable/combustible liquid was gasoline, JP4. or crude oil (35 deaths).

Category 2: Incidents D, E, H, M, and S involved LPG transported in either highway cargo tanks or railroad tank cars (62 deaths).

Category 3•’ Incident I involved anhydrous ammonia transported in a highwaycargo tank (five deaths).

Category 4: Incident I. involved a corrosive liquid transported in a highway cargo tank. It was an unusual incident, involving a mistake by either employees at the plant… where the acid was delivered or the delivery truck driver. The product was unloaded into the wrong storage tank, reacted with an incompatible chemical, and caused toxic vapors that killed eight plant employees (eight deaths).

Category 5: Incident N involved chlo* rine transported in a railroad tank car (eight deaths).

We now have categorized the worst multiple-fatality accidents that occurred over the past 20 years in hazardousmaterials transportation. Certain patterns appear, such as the involvement of highway cargo tanks or railroad tank cars in all five of the categories. Also, flammable/ combustible liquids and flammable gases ’ in cargo tanks and railroad tank cars account for all but four of the 20 incidents.

It is tempting to draw some immediate conclusions. However, although these 20 incidents represent the worst multiplefatality transportation accidents from 1970 to 1990, they account for only 120 of the 369 deaths. We should not attempt to draw significant conclusions based on only one-third of the total fatalities. Ana-s lyzing the other two-thirds, or 67 percent, of the finalities reveals that most can be grouped into the same general categories used for the major incidents.

Of the remaining 249 deaths, 179 were caused by flammable/combustible liquids—gasoline, jet fuel, diesel fuel, and others. Of the 70 deaths now unaccount_ ed for, 29 were caused by I.PG or another flammable gas, leaving 41 fatalities, nine of which were caused by anhydrous ammonia releases. Of the 32 deaths left, four ‘ were due to explosives. The remaining 28 , fatalities were caused by other hazard classes and chemicals.

Adding the nonmajor incident figures to the previously developed category figures, we account for 341 deaths, or more than 92 percent of all fatalities caused by hazardous-materials releases. This gives a fairly complete picture of what causes hazardous-materials transportation fatalities.

Category 1. Flammable/combustible liquids, almost exclusively in highway cargo tanks, 214 deaths.

Category 2. LPG, almost exclusively in highway cargo tanks and railroad tank cars, 78 deaths.

Category 3Anhydrous ammonia, almost exclusively in highway cargo tanks, 14 deaths.

Category 4. Corrosive liquid delivered into the wrong storage tank, eight deaths.

Category 5. Chlorine in a railroad tank car, eight deaths.

‘ITie fatality data show several significant facts:

  • Eighty-two percent (292 of 354) of the fatalities involved either flammable/ combustible liquids or flammable gases.
  • Of the 68 deaths not accounted for by flammable/combustible liquids or flammable gases, 22, or about one-third of them, involved either anhydrous ammonia or chlorine.
  • Kighty-nine percent (316 of 354) of
  • the deaths almost exclusively involved highway cargo tanks or railroad tank cars.
  • Less than eight percent of the fatalities involved chemicals other than explosives, flammable/combustible liquids and gases, anhydrous ammonia, chlorine, or the fumes from improperly mixing one chemical with another in a storage tank.

The single incident in Category 4, in which an industrial accident resulted inmultiple fatalities, is difficult to address. This accident was caused by the improper addition of a chemical, delivered to a facility, to a tank that already contained another chemical. The two materials were not compatible and quickly produced toxic vapors. The emergency responders had to deal with the casualties and not directly with the chemicals involved. Other incidents involving the accidental mixing of incompatible chemicals in storage tanks have taken place, but the records indicate they are relatively uncommon.


Our initial objective was to use past multiple-fatality transportation accident data to best determine what knowledge and skills we had to master to prepare ourselves for future incidents. The almost inescapable conclusion reached from the facts stated above is that the overwhelming majority of our hazardous-materials training should concentrate on the field behavior of flammable/combustible liquids and flammable gases, the field behavior of anhydrous ammonia and chlorine, and the accident performance of highwaycargo tanks and railroad tank cars.

Instruction involving the field behavior of flammable/combustible liquids and gases, anhydrous ammonia, and chlorine is considerably different than teaching just the technical chemical characteristics of these materials. Most experienced responders agree that being able to state from memory the chemical formula of a substance usually isn’t very relevant when you’re up to your ankles in it. What you really need to understand is what might happen next, based on what has occurred in previous accidents.

We also concluded that we had to know about highway cargo tanks and railroad tank cars. Since these containers are involved in the majority of fatalities involving hazardous-materials transportation, it seems apparent on its face that we need to know more about them. Most emergency responders are not able to answer the following, sometimes critical, questions: Will a burning gasoline cargo tank explode? Can you stop leaks from gasoline cargo tank dome covers? What are the functions of the various valves and fittings on compressed gas cargo tanks? How can you stop or reduce leaks from valves or fittings on railroad tank cars? How can product be removed from highway cargo tanks and railroad tank cars?

Our study of hazardous-materials transportation accident fatalities over the past 20 years has led to some interesting conclusions. It appears that the traditional instruction emergency responders receive overemphasizes chemistry and underemphasizes the things we really need to know. What is needed now is a change in priorities. We must begin teaching the knowledge and skills critical to our ability to mitigate future accidents.

Most programs in hazardous-materials instruction are top-heavy in chemistry. Such an emphasis may be necessary for handling fixed-facility incidents. However, our analysis of 369 transportation deaths from 1970 to 1990 clearly shows that we need to redirect our training. We need to teach a new set of skills and knowledge that will better prepare us to respond to transportation accidents in the next 20 years.

No posts to display