A National Transportation Safety Board (NTSB) special investigation, undertaken in response to observed hazardous materials emergency response problems, revealed a lack of practical, currently available information about hazardous materials behavior in transportation accidents. Researchers, mathematical modelers, emergency response personnel others voiced a need for data to support predictions about how materials could be expected to behave in accidents. The NTSB hazardous material accident Spill Map program was developed to fill this need. NTSB Spill Maps now report such behavior in a tentatively standardized format. The maps feature a time-sequenced display of dispersion patterns and ranges, weather at the time(s) reported, injury/atality exposure locations, and a synopsis of the accident scenario. Although these maps are currently used by the Safety Board to report behaviors of hazardous materials released in transportation accidents, applications can be seen for a much broader range of releases. The Safety Board’s development of a reporting standard for hazardous materials releases supports an accelerating shift in the basic philosophy for controlling hazardous materials transportation safety risks. In the past, the basic safety objective was to contain hazardous products safely under “conditions normally incident to transportation.” however, accidents and failures of containment systems frequently resulted in emergencies that had to be handled by local emergency response personnel. Resulting casualties among response personnel and other personnel enlarged the Safety Board’s concern to include not only accident prevention objectives, but also casualty reduction during accidents. One way to reduce these casualties was to improve the decisionmaking process, so emergency response personnel could logically think their way through an emergency to achieve the safest possible outcome. This process requires predictive estimates of the likely behavior of hazardous materials present.

At a public hearing held by the NTSB in April 1978, emergency response personnel reported problems encountered in developing pre-emergency plans, and making tactical decisions during emergencies involving hazardous materials. many voiced the belief that existing technical manuals, emergency action guides, and other information sources do not provide adequate information needed to identify how a hazardous material might behave in a specific emergency. The Safety Board reports of accidents resulting in the tragic deaths of at least 50 firefighters since 1968 clearly demonstrate the need for improved methods of handling hazardous materials emergencies. Emergency services cannot be expected to adequately plan for and make life or death decisions at hazardous materials emergencies when information which will help them identify how a given material will probably act in a specific emergency situation is not available.

An analysis of the hearing testimony, previous accidents, and subsequent interviews with emergency response personnel identified the need to determine how a hazardous material will behave before tactical decisions towards the safest outcome can be made. That is necessary to assess exposures risk and safe operational distances. Five aspects of the hazardous materials’ potential behavior must be understood before effective pre-emergency plans can be formulated, or tactical on-scene decisions needs can be resolved:

• where is the hazardous material or container likely to go if released during the emergency?
• why is the hazardous material likely to go there?
• when is the hazardous material likely to go there?
• how will the hazardous material get there?
• what harm will occur when the hazardous material gets there?

Already available information sources were surveyed to determine the quality and quantity of information provided with respect to these points. Accident reports, the most common source of information on past hazardous materials accidents, contained data which helped determine probable accident cause, yet lacked information regarding hazardous materials behavior and casualty relationships. Research information such as mathematical modeling provides one basis for forecasts of hazardous material behavior; however data for validation of predictions is limited. Emergency action guides, which usually provide lists of various known properties and characteristics, can be linked to hazardous materials behavior only in general terms. As a result, instructions are brief, general, and vague. No presently available information source provided adequate information required for the emergency response decision process.

During the investigation it was noted that some mapping was already employed in accident reports and some of the more comprehensive manuals to display dispersion boundaries, casualty location, or damages. However, no general standard to record the events was established. A variety of methods, scales, map bases, and information content were being used.

NTSB APPROACH TO MAPS

The NTSB approach towards displaying hazardous material behavior information was suggested by data obtained in an investigation of a tank truck accident in 1976 involving release of anhydrous ammonia. Infrared photographs, U.S. Geological Survey (U.S.G.S.) maps, and events-sequence analysis were used to record the accident events. Analysis of data recorded by the Safety Board, the Houston Fire Department, and the Houston Air Pollution Control Board established relationships between product dispersion, environmental conditions, and injuries which could be mapped. To record these relationships, timing of events, timing of the dispersion pattern, a timing of exposures to the hazardous materials had to be shown. A graphic display on a standardized map base was selected to present this information. The relationships could be displayed on timesequenced maps, constructed to illustrate hazardous materials dispersion patterns for the progressive phases during the emergency, and showing weather data and injury/fatality information.

To fulfill users’ needs, the NTSB established some basic criteria for map base choice, information content, and layout. The selection of the most desirable map base was governed by several considerations. A map base with a uniform scale, available nationwide, at low cost to all users, was considered essential. A consistent map scale would enable comparison of hazardous material effects, regardless of individual accident locations. The map base should include transportation corridors and topographic features. Display of exposures in the potential path of released hazardous materials would be useful. Descriptive information about the accident should be included in a consistent form, as should weather conditions and fatality location information. A uniform means of representing hazardous materials dispersion effects had to be formulated.

A short synopsis summarizing the sequence of events, and a summary showing the locations and degree of injury resulting from hazardous material exposure was also needed, to support the behavior prediction effort.

The U.S. Geological Survey 7-1/2 minute series, 1:24,000 scale, topographic map series was chosen as the standard mapping base. Scales 1:12,000 or 1:6,000 scale were also provided, to permit reporting of small release patterns. These maps are available from U.S.G.S. for most parts of the country. Mapping of other areas is to be completed soon; in such areas, a 15 minute series, 1:62,500 scale map is available and may be enlarged to the 1:24,000 scale or greater. U.S.G.S. topographic maps include relief features, waterways, towns, and buildings. Roads, highways, railroads, pipelines, airports, and docks are shown. Thus, an accident location on the transportation corridor can be displayed as accurately as data permit. The inclusion of buildings on the U.S.G.S. maps also helps indicate potential exposures in the path of a hazardous material. The U.S.G.S. maps are readily available for $1.25 each. They are presently used for a variety of related applications, including local civil defense and planning, so that many users already may be familiar with the map scale and symbols.

Using the U.S.G.S. topographic map as a base, the behavior of the hazardous material is displayed at various times during the emergency. Time sequencing is important for emergency response personnel, because needed response actions change with time. Selecting the times at which maps will be prepared is related to the changes in the hazardous material behavior pattern. Maps try to show the growth of the hazardous material dispersion pattern, changes in area affected due to environmental conditions such as wind direction, changes in number of people affected, or other changes in the behavior which affect emergency response actions.

Both surface and airborne effects are displayed. These displays are prepared from observed data, supplemented by estimates suggested by witness’ comments. Estimated surface effects appear as shaded areas. Solid lines enclosing surface effects represent observed patterns. An estimated boundary of airborne effects to represented by a dotted line; and observed boundary is represented by a dash line. A key on each map identifies these representations.

Weather conditions at the times mapped are displayed in a block that indicates wind direction and speed, sky condition, present weather, visibility, and air temperature. (“Present weather” is the form of precipitation present.) Data in this block may change with time on each map in the sequence.

An x,y, coordinate grid system is used to pinpoint the location of fatal exposures. Scales are numbered along the horizontal (x-axis) and vertical (y-axis) edges of the map. Using x,y coordinates, approximate locations of hazardous materials exposures leading to death are tabulated on the map.

Descriptive data about the accident is provided in a box containing: type of accident, date and time of accident, time of event displayed, time after initial hazardous material release, product released, container type, quantity of product released, map base, and map scale. This information is shown on each map in the seauence.

USES OF NTSB SPILL MAPS

These maps present observation of hazardous material behavior in actual accidents. From such information, many applications are foreseen. Model verification, release comparisons, and accident simulations are some of the expected uses. The mapping exercise may be expanded from recording releases due to transportation accidents to mapping other releases, such as at storage or manufacturing facilities. Maps eventually may be referenced on-scene during hazardous materials emergencies, providing a foundation for critical decisions.

For each application, precaution must be taken to accommodate the variables of weather, product, amount released, population density, and topography in different situations. Locations, environmental conditions, terrain features, release mechanics, etc. are unique for each release. The dispersion of a released product in a mountainous location, influenced by specific weather conditions, may be compared only in general terms to a similar release in an area with little topographic relief. Accident simulations in areas with like characteristics may also be accomplished when the variables are considered.

Accident comparison, based on spill size and injurious effects as related to commodity and transportation mode, is helping to distinguish the more dangerous or lethal shipments in accidents involving hazardous materials. Using the standard scale, Spill Maps will facilitate such comparisons. They provide estimates of fatal exposure areas and other exposure ranges which should help emergency response personnel to appraise potential lethality to guide their response decisions.

Information from a past accident may be used to simulate that accident on a local community map. Simulation of mapped accidents in a local community can provide insights about the nature and magnitude of local problems which might arise in information gathering, emergency assessment, obtaining access to the site, logistics, evacuation, treatment of victims, resource limitations, communication and command needs, and other decision dilemmas or data needs. Participants in simulations can obtain experience in resolving actual problems without risking exposure to actual on-scene dangers. Planners in local communities may use simulations to identify “target hazard” areas where emergency service tactics, command actions, coordination, and possible courses of action could be practiced.

The variables of topography, weather, released product quantity, etc. can be used to test models which make predictions such as downwind dispersion patterns. Such predictions may be tested against actual occurrences using information provided by the hazardous materials accident Spill Maps. Through such verification, the confidence level in mathematical models could increase. Subsequently, emer9ency action guides which rely heavily on models, could be upgraded.

Maps depicting hazardous material releases may also be used as a basis in establishing or amending regulations. For example, the Department of Housing and Urban Development is interested in maps in relation to siting under its housing programs. Significant differences between commodities behavior when spilled will be evident on maps. For various materials, the timing of the dispersion pattern formed and the harm that results has been noted to differ significantly; slower dispersion rates seem to produce lower losses in comparable situations. Such comparisons may lead to a rationale for changes in regulatory. requirements for high loss-producing commodities.

In the future, hazardous material accident Spill Maps are expected to aid on-scene response activities. Stored maps and associated data from past accidents at a one-call information center could be recalled to help local emergency service personnel assess the dangers during an actual emergency. Emergency response coordinators would be able to call the center and obtain information about accidents with similar circumstances. Local variables could be interpreted and probable hazardous materials behavior then could be predicted. Recall and display capabilities within emergency response time frames seem to be technically feasible if libraries of the maps were to be developed in computer-compatible form.

The basic map elements needed to fulfill the range of applications are (1) time sequencing, (2) standard scale, and (3) an adequate number of maps. The first two elements are available on the NTSB hazardous materials accident Spill Maps. The third, however, can only be accomplished with the help of others. Standardized hazardous materials map preparation is presently performed solely by the NTSB, to our knowledge. Involvement of outside organizations would multiply the effectiveness and utility of the maps.

WHAT’S NEXT

Other organizations have expressed interest in this mapping project. Between 1500 and 2000 parties on the Safety Board’s present mailing list receive copies of each map. In addition to interested users of the maps, a number of individuals have requested additional information for preparing the maps. Written procedures for preparing maps have been drafted. Presently the Board is soliciting comments as to how these maps might be improved to meet particular needs. In addition, attempts are .being made to involve other organizations in mapping hazardous material releases.

Any significant release of a hazardous product in transportation, workplace, industry, etc. could be recorded on U.S.G.S. maps to provide standardized Spill Maps. The current goal to increase the hazardous materials data base is attainable with the cooperation of all who share an interest in understanding their accident behavior. Maps could be provided by those interested workplace releases of hazardous materials. They could be prepared by those responsible for recording environmental insults. Maps could be constructed by those who respond to incidents where hazardous materials are released, an interest which already has been expressed. Each new map of a hazardous material release will add to the degree of certainty in predicting the behavior in future releases.

As more sources become increasingly involved in reporting and using hazardous materials accident Spill Maps, it eventually will be necessary to establish a clearinghouse for map publication and distribution to users. The Safety Board is not equipped to handle such a voluminous project, which will likely include maps for non-transportation uses. Public or private agencies may have to share this task, as usage becomes more widespread.

The idea of hazardous materials accidents Spill Maps is being accepted with enthusiasm by several users, As other users become familiar with this new data source and possible applications, we are seeing more and more excitement about the project, and are getting more and more comments and suggestions about them. This is the most important step for successful use of the maps. Because a successful program depends so greatly on out-side participation, comments from all those who might see a possible application of these Spill Maps for their own purposes are genuinely welcomed by the Safety Board. Any comments and suggestions to make these maps more useful for specific applications should be forwarded to the Safety Board, Washington, D.C. 20594, Attention TE-40. The current standardized way of presenting the data is only tentative until these comments have been reviewed and worked into a final map system, The final standardized map system is expected to be proposed by the Safety Board by the end of 1980. We hope your inputs will be avail-able to help shape the future of this program.

Historical notes:

This paper was presented by Ms Rote to the American Institute of Chemical Engineers Loss Prevention Symposium, June 8-12, 1980 in Philadelphia Pa. It presented a new initiative we undertook to help refine predictions and validation of hazardous materials releases. An earlier presentation to the industry has proded shippers of hazmats to explore the prediction of hazmat release effects. The MAPS never gained momentum after I left the NTSB apparently because the demand for validation of predictions was not pursued at the Board.