Archives of Personal Papers ex libris Ludwig Benner, Jr.
line   - - - - - -Last updated on Thursday, November 25, 2004
[Return to Home Page ]   [ Investigation Research Roundtable ]   [ Contact Host via r lbjr05 at ]

Click here to order reproduction permissions for public use beyond fair use ( Item R070) for this Paper


National Fire Protection Association 81st Annual Meeting
Washington, DC
May 16-20, 1977
By Ludwig Benner, Jr.
Hazardous Materials Specialist National Transportation Safety Board

The views expressed are those of the author and do
not necessarily reflect the views of the National Transportation Safety Board.



This paper examines the expectations of firemen in hazardous materials transportation emergencies, and suggests that until firemen know how to predict the behavior of hazardous materials in emergencies, it seems unreasonable to expect them to be able to favorably change or influence that behavior. Some dangerous assumptions on which present training is based are described.


The purpose of this paper is to discuss the continuing vulnerability of firemen in hazardous materials emergencies, probe the reasons this vulnerability continues, and indicate what firemen might do to reduce their vulnerability.


In its October 20, 1976, letter to the Secretary of Transportation, the National Transportation Safety Board mentioned a continuing series of incidents in which firemen lost their lives or were seriously injured while attempting to cope with hazardous materials transportation emergencies. In the seven years I have been with the Safety Board, it appears that the rate of casualties among firemen responding to hazardous materials emergencies has not declined appreciably.

In those accidents which the Safety Board has investigated, there seems to be a common thread. The responding firemen involved were doing what they thought was expected of them when they were injured. These expectations were examined in the Safety Board’s report of a railroad accident at Houston, Texas, and in other subsequent reports. For one type of emergency, concerted efforts to change these expectations seem to be producing favorable results. However, the efforts involve only one of the many, many types and groups of hazardous materials likely to be encountered in emergencies.

The conclusion is apparent. Until the other hazardous materials are treated in the same way as the example I will discuss shortly, firemen will continue to be vulnerable to injury and death in future hazardous materials emergencies. This, then, is the problem I want to discuss:

what can be done to reduce this continuing vulnerability of firemen in

future hazardous materials emergencies???


Examination of the case histories of liquefied gas tank car emergencies is instructive. Beginning in 1970, with the disastrous railroad accident in Crescent City, Illinois, special attention was given to the threat posed by DOT type 112A and 114A involved in railroad derailments and fires. The photographs and motion pictures of the Crescent City disaster received widespread distribution, and were used to show what can happen when these cars are involved in railroad accidents. In 1971, 40 firemen were casualties as they fought the fires involving vinyl chloride which had spilled in a tank car derailment in Houston, Texas. In its report of its investigation of this accident, the Safety Board explored the decision making process of the firemen during that emergency. This resulted in a Safety Board recommendation to improve the guidelines being offered firemen to glide their actions during such emergencies. In April, 1974, the National Fire Protection Association published and gave wide distribution to revised recommendations for handling emergencies involving liquefied petroleum gas (LPG) in these types of tank cars when they were in the wrecks. NFPA followed this advisory with a revised film on the handling of LP gas fires involving tank cars in accidents, in its widely distributed film “BLEVE.” These revised instructions shifted the focus of the firemen’s attention from the inherent characteristics of the LP gas and specific instructions for “cooling the tanks” to a description of the behavior of the tank cars in fires and a message that withdrawal could be the best response. This important shift in the approaches is a milestone. The film conveys to firemen how they should expect these hazardous materials containers with flammable liquefied compressed gases to behave in an emergency, and it identifies ways they can determine whether or not this behavior might occur.

Now, if firemen can predict what events are going to happen, they can take measures to favorably influence that course of events and outcome in specific emergencies. And therein lies the key to understanding the continuing vulnerability of firemen in these kinds of emergencies. UNTIL FIREMEN KNOW HOW TO PREDICT THE BEHAVIOR OF HAZARDOUS MATERIALS INVOLVED IN AN EMERGENCY, IT SEEMS UNREASONABLE TO EXPECT THEM TO BE ABLE TO FAVORABLY CHANGE OR INFLUENCE THAT BEHAVIOR!

I would like to try to further clarify your understanding of this basic problem by introducing systems analysis terms of reference for tour discussion. My own experience indicates that systems analysis is a very useful tool for gaining an understanding of complex phenomena.

Let’s begin with the general systems model shown below. We can describe the handling of a hazardous material emergency in terms of input, operation, output and feedback. The model is illustrated below:


To clarify these terms, compare the fire department with a computer. For a computer, the inputs govern the outputs. The old expression “ garbage-in, garbage-out” is a familiar and appropriate one here. Inputs to a computer include both the raw data and the program of instructions for its operation. By analogy, the inputs at the scene of a hazardous materials emergency response include the raw data at the scene of a specific .incident, plus the “programming” of the emergency response personnel for their operations during the emergency. Deficient on-scene data or deficient firefighter “programming" can both produce deficient outputs— that is, increased harm. The processing of the input data in accordance with the input programs by a computer involves a series of logical electronic "decisions" as the data proceeds through the computer toward the intended output. In much the same way, firemen process data and make decisions during the course of emergency events to produce the emergency response output, that is, the emergency outcome. In the operation of a computer, the output is generally some document that meets a clearly specified form and quality. For hazardous materials emergency responses, (he specified form and quality of outputs are poorly defined. The definition of desired emergency response outputs needs attention desperately. Without dwelling on this problem, I suggest you think of fire department outputs in terms of the reductions in casualties or losses achieved by the response, as compared with what would occur if no response were mounted by the firemen.

Sometimes computers are designed to adjust their operations based on “feedback” generated while the computer is operating. This “self—adjustment” during the computer operation can be compared with the adjustment of emergency response operations during an emergency.

Now, in this context, let’s examine the status of the input data, programs, operations, outputs and feedback for hazardous materials emergency response by firemen. Except for LPG tank car wrecks, I can find almost no instructions to firemen specifying either the data they should look for in emergencies, or instructing them in how to predict the behavior of the material involved. The instructions are generally limited to signs or placards, markings or labels, or people who have information about the hazardous materials. These instructions do not speak to the identification of energy transfers, transfer mechanisms, stressed elements of the systems involved, or other indicators of the probable behavior of the hazardous materials during the emergency This, in my view, demonstrates deficiencies in the "programming" of the fire services for such emergencies. It appears that this traditional “attack and extinguish” programming of firemen, in combination with their experience in “ size-up” of fire emergencies is assumed to satisfy the programming needs for firemen. The continuing casualties demonstrate that this assumption simplify is not justified. Experience also indicates that programming firemen to refer to manuals during emergencies is ineffective for several reasons, including the lack of availability when needed, problems in interpreting the generalized instructions in specific cases, and the uncertainty which can arise when instructions are vague or contradictory.

If you think I am alleging that present instructions inadequately “program” firemen for hazardous materials emergencies, you have received the heart of my message.

There are literally hundreds of thousands of people engaged in the production, processing, handling and transportation of hazardous materials. All these people expect the goods to move without incident, but if there is an incident, they all expect the firemen to handle the ensuing emergency. Most stand ready to help the firemen if the firemen ask for help. Note that the initiative for asking for help rests with the firemen. This places a further burden for diagnosing the emergency on the firemen. Yet, have any of you seen any instructions that indicate when and in what circumstances firemen should ask for expert help? As you look over the training material available to firemen, put yourself in the position of being called on to interpret these instructions under the stress of an emergency. I think you will soon appreciate the problem I am trying to define here, and understand why it exists.

Carrying this analysis one step further leads us to what I believe to be the source of the firemen’s continuing vulnerability in hazardous materials emergencies. The example of the LPG tank cars demonstrates what can be achieved when the behavior of hazardous materials is identified and widely disseminated to the fire service. I believe we will see a substantial reduction or even elimination of firefighter fatalities in his type of accident in the future, as the understanding conveyed by the AAR instruction and “BLEVE” film spreads. This, in turn, suggests that if firemen could be instructed in the way all hazardous materials are “programmed” to behave in emergencies, it would go a long way toward enabling them to produce safer emergency outcomes in the future.

However, today the understanding of these programmed behaviors in transportation emergencies is almost nonexistent. The principal reason these programs are not widely available to persons developing training courses or dealing with emergencies is that almost no one has made any special effort to record the behavior of hazardous materials shipments during emergencies, nor to record the successful or unsuccessful emergency actions taken by firemen to influence this behavior favorably.

The conditions under which the transportation of a hazardous material is authorized by regulatory agencies, and performed by shippers and carriers largely determine the behaviors programmed into hazmat shipments in emergencies. At present, there is no requirement that this “programmed” emergency behavior be identified or documented prior to moving shipments in the transportation systems. The Safety Board has, on several occasions, called for the preparation of safety analyses and the recording of accident data to identify this “programmed” emergency behavior. It is unclear how the programmed behaviors were determined or considered in the preparation of the existing guidelines for emergency response actions. Additionally, in the absence of documentation, new experience gained in accidents can not be used to regularly refine our current understanding of these “programs.” Thus, it appears that firemen will continue to remain vulnerable in these types of emergencies in the foreseeable future.


Frankly, public safety officials are faced with a difficult dilemma. They can not refuse to respond to hazmat emergencies. Yet, improvement of the “ programmed responses “ is not presently practical because of the present state of knowledge about the “programmed behavior” of hazmat shipments in emergencies. In these circumstances, what should be done?

First, the fire service should use every means at its disposal at the local, State and Federal levels to bring about documentation of the emergency behavior presently “programmed” into hazmat shipments under today’s regulations. The Safety Board, in proposing safety analysis and hazmat emergency information system refinements, has suggested directions fort these activities. However these developments will take time before they begin to produce soundly programmed emergency responses for the local firefighters. In the meantime, emergencies will continue to occur. What might be done to improve their outcomes until a better understanding of “programmed” behavior“ is achieved and disseminated?

One approach for rapid resolution of this problem was proposed in April 1977 by the Safety Board. Specifically, the Safety Board asked that the development of a “hazmat emergency squad” be studied, and a plan developed” to help firemen diagnose hazmat emergencies, devise responses which would” produce better outcomes, and monitor the progress as the emergency progresses toward a conclusion.

The proposal contemplated the early notification of the presence of one or more hazmat to the local public safety officials by the transportation operator as soon as an incident or spill was detected, and the subsequent linking of the public safety officials to experts with knowledge of “programmed hazmat behavior” through voice communications. The availability of this real time, personal emergency assistance during the critical early diagnostic stages might be one way of overcoming the local officials’ inability to predict the “programmed hazmat behavior.“ This approach would facilitate training of personnel in the over 29,000 fire departments who might on any given day at any moment be called upon to cope with a hazmat emergency without adequate “programming” for handling it.

In closing, I would” like to highlight some of the most important assumptions that the fire service should be aware of during any response to a hazmat transportation emergency today. These include:

  • every local fire department is expected to be able to handle any hazmat transportation emergency at any time;
  • every fire officer of every fire company is expected to know when he should call for help in a specific emergency;
  • every fire officer of every fire company is expected to be able to interpret the written instructions for handling hazmat emergencies which might be available to him; and
  • better training will resolve any problems that exist.
I suggest you weigh these assumptions critically as you formulate your own future actions in this field.

Historical note

This paper was presented to stimulate the individuals at greatest risk to reexamine the underlying philosophy, system of assumptions and procedures addressed to hazardous materials emergency response guidance and training. It was presented at a time when the training needs for firefighters were becoming evident, because of the challenges then facing responders.

LB December 26, 1999

Return to top: