The analyses of specific reports using the STEP method, is presented and discussed in this section. The events sets and comments show the accident process and suggest problems and corrective action options that will be further discussed and summarized in the next section.

2.2. Accidents analyzed.

Accident reports from the following accidents were used in the analysis.

>

Case	Location	Date	Type
1.	Checotah, OK	4 August 1985	Fire & Explosions
2.	Benson, AZ	24 May 1973	Fire & Explosions
3.	Waco, GA	4 June 1971	Fire & Explosion
4.	Roseville, CA	28 April 1973	Fire & Explosions
5.	Ajo AZ	25 June 1976	Fire
6.	Bonnieville, KY	10 May 1985	Fire
7.	W. Frankford, IL	18 Dec 1973	Fire
8.	Lincoln, AL	20 May 1987	Fire & Explosion
9.	Keystone, WV	5 April 1979	Fire & Explosion	>
10.	Ishpeming, MI	26 August 1973	Fire

Each accident was flow charted on a worksheet to the extent permitted by the information reported. The worksheets are found in Appendix 2.

2.3. Highway Accidents Analyzed

Each event set from the worksheet for each case is analyzed in this section. The numbers following each event block specify the row and column where the event is located on the worksheet in Appendix 2.

2.3.1. Checotah, OK, 4 August 1985 Fire & Explosion

This accident involved the rear-ending of a passenger automobile by a tractor-semi-trailer rig transporting 10 bombs, followed by a fire and explosions. The discussion of the events sets in this accident focuses on the highway vehicle transporting Class A explosives which was involved in the collision and explosions, and those events. A second vehicle, not part of that process, could be analyzed to determine what role second drivers might play in accidents of this type, if sufficient data are reported.

2.3.1.1. Collision events

Tractor strikes vehicle [2-1.4]

24 gal fuel escape from auto gas tank [8-1.4]

fire erupted. [4-1.6]

Because of the frequency with which this event set can be observed in day-to-day traffic operations, this set must be considered a probable event set when transporting hazardous materials by highway, and therefore should be included in any general model of the explosives A mishap process. Changes that might be introduced to control this event set are beyond the scope of this study because such action lies outside the OHMR.

2.3.1.2. Initial fire events

Data from which the origin of the fire could be described was not reported with any degree of certainty in any of the cases. The data that were reported included the following events set:

fire erupted [4-1.6]

tractor trailer stopped over fire [2-2.0]

fire engulfed front of tractor and rear of auto [4-2.6]

In the 8 highway accidents studied, this event set was reported in two large loss cases (1,3) and may have occurred in a third (8) In cases 1 and 3, the vehicles came to rest upright in the fires. From the reports, it is unclear whether the vehicles might have been brought to a stop in a different location, changing the relationship within this event set For a general model, this set should be included.

2.3.1.3. Driver actions after accident began.

Data from all eight highway vehicle cases demonstrates that the driver of the truck transporting the explosives is likely to survive the accident relatively uninjured, and capable of exercising emergency response functions. The specific event set follows:

Truck stopped over fire [2-2]

? [1-2.8]

Driver retrieved fire extinguisher from tractor [2-3/6]

While the driver actions between the time the truck stopped and the time he retrieved the fire extinguisher are unreported, it is logical to conclude that he was physically able to function sufficiently well to do this.

2.3.1.4. Driver fire fighting actions.

The driver apparently did what he could to extinguish the fire, but did not produce the desired results.

Driver tried to extinguish fire on both vehicles [ 1-4.2]

Driver exhausted fire extinguisher without getting fire out [2-5.2]

The report describes but does not explain why this events set occurred. The events set suggests that the driver tried to do what he was probably programmed to do - attempt to put the fire out with the tools provided him. The outcome suggests changes might merit consideration. First, the tools that are provided drivers might be reviewed to ascertain what other types of tools might be made available to surviving drivers to put to work on controlling fires. Secondly, review of what drivers are programmed to do, and the decision making process they are programmed to execute, now and with possible new tools, might suggest possible new fire control options as such fires are better understood. The absence of data after the extensive incident experience to date suggests that development of improved investigator training and reporting requirements are needed to gain this understanding.

2.3.1.5. Driver rescue efforts

The driver then directed his energies to the rescue of persons involved in the crash.

Driver exhausted fire extinguisher without getting fire out [2-5.2]

? [2-6.0]

Driver helped pull passenger out of auto [2-6.2]

It is not clear what, if anything, the driver did between the time he exhausted the extinguisher and the time he started to help pull the passenger out of the auto. What is significant is that the driver had time to do all this and still escape the later explosion. This adds weight to the suggestion that the actions drivers are programmed to take in such circumstances merit review to determine the most effective task sequencing and how to perform the tasks.

2.3.1.6. Police response

A local police officer was dispatched to the collision scene after a local citizen phoned a report of the accident to the police offices.

? [7-3.6]

Police dispatcher sent police officer to scene [7-4.0]

Police officer reported fuel tank truck was involved to police dispatcher [6-7.0]

Police dispatcher notified Fire Dept., State Police, Ambulance [7-7.6]

This event set resulted in the initial problem definition for the emergency response that followed. At least three response units entered the hazardous area with a mistaken understanding of the danger they faced. The problem was later redefined, but the initial information made available in emergency response communication channels was erroneous. While the outcome in this case was acceptable, the increased risk such erroneous information poses is apparent. The report does not provide additional details about causal relationships among these events. Since it was fairly recent and the participants survived without incident, further investigation of this events set might prove fruitful in terms of size-up training needs that might be suggested for police officers.

The event set also relates to the driver task analysis discussed above. The actions by drivers to provide risk information to the emergency response personnel who need it should be included in the driver post-collision task analysis. This may involve some diagnostic training for drivers.

2.3.1.7. Fire behavior after initiation.

The next events set considers the behavior of the fire.

fire engulfed front of tractor and rear of auto [4-2.6]

Flames rose high above the wreckage [4-7.0]

(Fire heated cargo)[7-8.2]

Despite the drivers attempts to extinguish the fire, it continued to burn after the collision. While "fire" is used as the actor here, the actual actors were probably the gasoline and components of the vehicles, as well as grass and perhaps other ground cover. The fire burned about 15 minutes before the cargo reacted. The report does not explicitly state that the fire heated the cargo, but the event had to occur to produce the subsequent events and is therefore included to illustrate how inferred events can be incorporated into the worksheets.

The main point of this events set is that in a collision of this type the fuel load available is sufficient to sustain a fire for at least 15 minutes without adding any additional fuel from external sources. This provides guidance for the prediction of time during which explosive cargo is likely to be at risk to heat of fires in such accidents. The size of the fire provides an indication of the heat flux to which cargo would be exposed.

2.3.1.8. Driver traffic control actions

The driver took time, after trying to extinguish the fire and help the auto passenger, to try to warn persons in the threatened area and persons who might enter the threatened area about the risk. This is a form of risk disclosure, in real time, and should be considered in the post collision task analyses for the drivers of explosive vehicles.

Driver told auto occupants he was transporting explosives [2-8.0]

Driver returned to top of hill to warn people in approaching vehicles [2-8.2]

The point raised by these actions is that the drive constitutes a risk disclosure resource, and the need for an assessment of the effectiveness of the risk disclosure by this resource on the classes of persons addressed. Persons with whom the driver has a lingering contact may require a different message than persons with whom the driver has only a transient contact, such as the driver of an oncoming vehicle, or emergency response personnel. The report does not state what the driver was trained to do in these circumstances, suggesting that "common sense" (adaptive responses based on subjective judgments under stress) may be relied on for these critical communications, rather than pre-planned and trained (habituated) responses.

2.3.1.9. Firefighter actions

The firefighters were responding to a tank truck fire, because of earlier communications. In the circumstances, it is understandable why they approached the burning wreckage as they did.

Firefighters approached site on entrance ramp from Old 69 [5-9.0]

Fire chief saw Class A explosives placard on back of semi-trailer [5-9.8]

Firefighter verified information on placard [5-11.0]

Fire Chief ordered firefighters to withdraw [5-12.2]

Firefighters safely withdrew fire trucks to safer location [5-13.4]

The report suggests that the first explosion occurred before the last event in the set, and the photograph of the explosion reinforces this timing. While ambiguous, the report data suggests that the placards were observed before the first explosion of the cargo, but the withdrawal occurred after the first explosion. Of special interest in this events set is the implied awareness of the information about the cargo available to them, and their use of their observations on site to change their response plan. This represents successful use of one of the regulatory safeguards.

The timing of the events in the set (probably less than ten minutes from arrival to withdrawal) suggests that the actions of surviving drivers could reduce the risks to the emergency responders.

2.3.1.10. Driver-firefighter interactions

The next events set provides a glimpse of driver interactions with firefighters.

Driver spoke with firefighter [2-11.0]

? [ 5-11.2]

Driver walked west to Dep. Sheriff's car[2-13.2] [5-13.4]

The report does not state what the firefighter did with the driver's information about the cargo, but it does indicate that the information did not reach the Fire Chief. This suggests a need for training of firefighters at all levels to be on the lookout for vehicle drivers, and possibly for drivers to have some kind of authoritative credential to display to emergency response personnel to establish their credibility and the value of their information ion mishaps. This provides another aspect of the driver task design review addressed above.

2.3.1.11. Cargo behavior in fire

The report presents some information about the behavior of the cargo in the fire. The timing of the events in this set is particularly import, for its implications for controlling the size of the loss, and possible intervention options.

~ 03:45 Part of cargo near front of trailer exploded [4-12.6]

? [4-14.6] [5-16.6]

~ 04:02 More of cargo exploded [4-16.4]

? [ 4-18.4] [5-16.6]

~ 04:22 Third cargo explosion created large crater in roadway [4-19.2]

The heat from the fire changed the state of the cargo until the first "explosion" occurred. about 15 minutes after the collision The nature of this explosion is not clear from the report, but it was probably a pressure rupture of one of the bombs, based on the relatively small distance affected when this happened; the firefighters' vehicle on the roadway was able to withdraw apparently under its own power.

The behavior of the bombs between the explosions (the "?" in the events set) was apparently not observed or not reported. Nor were the effects on the fire reported, although a photograph of one of the explosions indicates that the fire was spread by the explosions.

The set suggests that some limited time is available to take ameliorative actions to reduce the size of the loss in accidents involving this type of cargo. While the risks of intervention escalate rapidly from the onset of the fire, they might be controllable and acceptable if, in future accidents, high-value exposures are threatened. Analysis of this behavior -- either observed or hypothesized -- using disciplined logic techniques such as fault trees, failure mode and effects analyses in combination with Energy trace and Barrier analyses or STEP techniques might disclose some way to intervene. The potential consequences and the relative probability of the events set (also observed in cases 2, and 4) indicate such analyses merit funding.

2.3.1.12. Post-explosion firefighter actions.

After the first relatively modest explosion firefighters remained concerned about firefighting, but withdrew to a safer location and applied water. The report does not indicate whether this was a serious fire suppression attempt or cargo-cooling effort or what the purpose of the water spray was.

~ 03:45 Firefighters safely withdrew fire trucks to safer location [5-13..6]

Firefighters sprayed water at fire from area below bridge on 69 [5-17.2

2 firefighters fell off fire truck due to explosion [5-19.8

Two firefighters were subsequently knocked off their vehicle because of the decision to spray water at the fire from the location below the bridge on highway 69, and presumably are among the 8 injured emergency response personnel. No scale is shown on the map or indicated from the photographs, but it appears their apparatus may have been positioned at a distance further from the fire than the hose spray could reach. If true, the expected effects of the water on the fire would not have been realized. This set, plus the 8 other emergency response injuries reported, suggest that the time available for ameliorative actions was used for traditional tactics, with the result that two firefighters were exposed to subsequent explosion effects with slim prospects of reducing the explosion risks and changing the outcome of the fire in this case. That reinforces the need described in 2.2.1.12 for review of the amelioration tasks that might be introduced between the time the fire begins and the time the first explosion occurs.

2.3.1.13. Post-explosion cargo behavior.

The third explosion about 52 minutes after the fire began was the most violent, but its effects are of interest because of what happened.

04:22 Third cargo explosion created large crater in roadway [4-19.4]

04:22+? Intact bomb landed on berm an est. 20 feet from crater[4-21.6]

04:22 + ? Intact casing with 80-90% of explosive burned out comes to rest 150' from crater[4-21.6].

04:22 + 3-1/2' Filled bomb casing section comes to rest about 150' from crater[4-21.6]

In addition to the intact bomb which did not explode although it was subjected to intense forces of the explosion, another 3-1/2 -foot-long rear bomb section filled with explosive filler was found around 150 feet southwest of the crater. Another undamaged bomb casing with about 80-90 percent of the explosive burned out was found around 150 feet south of the crater. A third large section of casing was also found about 150 feet away from the crater. The distribution of these remnants suggests that the complete filled bomb may have been throw free of the fire during the first or second "explosion" while the two large remnants found 150 feet away were probably blown there by the third explosion. The other seven bombs apparently detonated, and broke up into smaller fragments.

This set suggests several points. First, the intact bomb near the crater survived a violent explosion, whether if was blown free of the trailer during the first, second or third explosion. If the survival mechanism could be identified, it might suggest some options for reducing the mass explosion risk, such as the third explosion, that exists in such accidents. A second point is that the explosion dispersed parts of the cargo with a residual explosive potential posing some non-zero risk until this part of the cargo was safely removed from the accident scene. The third point is that the removal of the dispersed parts of the cargo required special expertise for Class A explosives

2.3.1.14. Explosion consequences

In addition to the dispersion of parts of the cargo, the explosion produced harmful effects on nearby exposures.

~ 04:22 Third cargo explosion created large crater in roadway[4-19.4]

Fire continued to burn (5:08 hrs) [4-20.4]

Over-pressure + flying debris damaged nearby structures[5-19.8]

The harmful work performed on structures by the Class A explosives was not detailed on the worksheet since harm to structures from these kinds of explosions is well known and documented. Additionally, changes to structures along transportation corridors to protect them from such harm are, a priori, not feasible because of the massive economic costs involved and the low risk to a specific structure posed by this kind of accident. The more cost-effective approach would be to make the explosions less harmful, rather than hardening the exposures, because of the smaller number of units that would be affected.

A second point related to the explosion consequences was the sudden escalation of the scope of the harmful consequences resulting from the explosion. Rather than dealing with a point source problem, the on-scene personnel were now required to act on a wider fire, and structural damages. This is an attribute of Class A explosions that differs from the risks posed by most other classes of hazardous materials.

2.3.1.15. Shipper response actions

The dispersion of the cargo during the third explosion, and the nature of the clean-up of the portions of the cargo that did not explode required expertise not found in most local response organizations. Recognizing this the on-scene emergency officials contacted the shipper for help.

03:45 + ? Driver walked to Deputy Sheriff's car [1-13.4]

04:22+?Highway Patrol called Army 61st Ord. Det. Explosive Ordnance Disposal [9-22.2][

04:22+? 61st recommended evacuation of 4-mi radius, do nothing until two hours after fire out, etc.[10-22.2]

08:32 61st team arrived on site to clean up, provide ordnance disposal support to community[1026.0]

The basis for the blast advice is not reported.

One point raised by this set is the feasibility of establishing a 4-mile radius evacuation zone in the time available before the third explosion - 52 minutes. The likelihood that any community could evacuate a roughly 36 square mile area in 52 minutes is small, indicating that more realistic alternative advice should be developed by the shipper for such emergencies.

A second point is that the arrival of on-site help followed the notification by about 4 hours. This seems to be a reasonable response time in such circumstances. During this 4 hour period, the local officials were confronted with the need to cope with the emergency. Meanwhile the driver was at the site, and available. These circumstances further demonstrate that review of the tasks during an accident for which drivers should be trained and prepared merits renewed attention for Class A explosive shipments.

08:32 61st team arrived on site to clean up, provide ordnance disposal support to community[1026.0]

61st team concluded cleanup operations on Aug 7 [10-28.2]

The arrival of the EOD team, and subsequently supporting National Guard and Pine Bluff Arsenal personnel, on site to assist with the clean up of the residual Class A explosive, and the duration of their stay -- almost 3 days -- suggest two points. The clean up of the debris from this type of cargo can be complicated and may require substantial specialized expertise. The cleanup expertise required appears to be a decision best left to experts furnished by the shipper. The second point is that such clean-up can extend the period during which a community can be disrupted by Class A explosive accidents for several days. -- almost 3 at Checotah. This suggests that procedures for cleanup should be reviewed to determine how the duration of the cleanup effort might be reduced.

2.3.1.16. Evacuation

The report is silent about the process that led the Mayor to order the evacuation of Checotah after the damage had been done by the explosion, so to speak.

04:22+? 61st recommended evacuation of 4-mi radius, do nothing until two hours after fire out, etc.[10-22.2]

? [9-22.2] [7-23.2]

~ 06:00 Mayor ordered evacuation of Checotah

Given the advice from the 61st, damages to structures, the continuing fires at the scene, and uncertainty about remaining explosives from the cargo, the evacuation order by the Mayor seems logical. However, the order did not come until over 1-1/2 hours after the third explosion. The time required to execute the order and complete the evacuation was not reported.

This events set suggest questions about the evacuation as an emergency response option in such accidents. Given the nature and timing of the risks, sheltering would seem a far superior option compared to evacuation, all other aspects remaining the same. Alternatively, reduction of the area at risk by changing the cargo behavior in an accident would be the preferable strategy to evacuation or shelter, which requires action by a large number of individuals, for reducing explosion risks.

2.3.1.17. Accident reporting

The information in this report was among the most extensive reported about this type of accident and was very useful for that reason.

2.3.2. Waco, GA 4 June 1971 Fire & Explosion

This accident involved a head-on collision of a passenger automobile and a tractor-semi-trailer rig transporting about 25000 pounds of commercial explosives, followed by a fire and explosions resulting in 5 fatalities, 38 injuries and several million dollars damages. The discussion of the events sets in this accident focuses on the highway vehicle transporting Class A explosives which was involved in the collision and explosions, and those events. As stated previously, this kind of accident where fuel is released during the collision dynamics occurs with sufficient frequency that it must be part of the risk assumed in highway transportation of hazardous materials.

2.3.2.1. Collision events

As in Case 1, a highway vehicle collision initiated a fire that involved the freight vehicle carrying Class A explosives..

Automobile crossed over center line of 2-lane highway [5-1.4]

Tractor smashed automobile[2-1.6]

? [4-2.2]

Fire engulfed tractor windshield + left side of tractor cab [4-2.4]

The is a different event set from that observed in Case 1, but with a similar result -- fire involving the vehicle carrying Class A explosives.. Again, changes that might be introduced to control this event set are beyond the scope of this study because such action lies outside the OHMR.

2.3.2.2. Initial fire events

The initial fire events occurred during the collision dynamics, in a manner similar to Case 1.

Fire engulfed tractor windshield + left side of tractor cab [4-2.4]

Tractor/trailer came to rest in fire, blocking road [2-4.0]

? [1-4.0]

Fire burned alongside entire left side of trailer [4-4.6]

The comments applicable to Case 1 are applicable here. It is not clear from the report whether vehicle damage would have precluded the driver from traveling further down the road out of the fuel and flames, to prevent the fire from burning along the trailer. Driving a vehicle away from a collision when there is fire is an action that contradicts most drivers' intuitive response to a collision, but viewed from the perspective of changing the overall risk of the accident, removing the vehicle from the fire source would reduce the greater risk of explosion.

2.3.2.3. Driver actions after accident began.

The driver of the vehicle carrying the Class A explosives was able to function after this collision.

Tractor/trailer came to rest in fire, blocking road [2-4.0]

Driver decided fire was uncontrollable with 2 5# 10BC extinguishers in vehicle

Driver exited via left tractor window

This events set discloses that the driver considered using the fire extinguishers but rejected their use while still in the cab of the tractor. The report does not state whether the driver left the extinguishers in the cab of the truck, although it implies that he did, nor does it describe why the driver made this decision, but from the description of the post-crash fire, the driver's actions appear logical. Subsequent events reinforce the point that a review of the driver's tasks merits consideration as a result of this accident, also. It is not clear whether the additional time the driver might have provided the firefighters would have been enough to change the course of events in this Case. However, in the absence of guidelines to use for training drivers and for helping them set priorities in such situations, the decision making of the drivers should not be disputed.

2.3.2.4. Driver fire fighting actions.

In this Case, the truck driver made several observations about the fire but did not change his mind about fighting the fire. Instead, the driver elected to try to warn approaching traffic away from the fire.

Driver smelled strong odor of gasoline[1-6.8]

Driver saw billowing black smoke around rear tractor wheels [1-7.8]

Driver ran to west to eastbound truck driver stopped in traffic [1-9.0]

The events set indicates that the driver was sufficiently concerned about the fire to gather data that might influence his decision about fighting the fire, since he was apparently aware of the consequences of the fire. However, once he had left the truck cab which was at least partially engulfed in flames, his only source of firefighting capability was no longer accessible to him. In reviewing driver tasks, the fire suppression tools available to the driver of a vehicle should be considered along with their potential use by the driver.

2.3.2.5. Driver rescue efforts

The truck driver did not attempt to rescue the automobile occupant, according to the report.

2.3.2.6. Police response

The community first notified of the accident had a combination police and fire department. Since the accident was in another jurisdiction, the police did not respond. The county sheriff was en route when the explosion occurred. The state highway patrol

Citizen called Bremen police/fire dept to report truck on fire on Rt 78 in Waco[8-8.0]

2 firefighters left immediately for fire in fire truck[6-8.2]

Resident notified Bremen police/fire dispatcher that burning truck was loaded with dynamite[8-11.4]

Police departed for scene immediately[7-11.6]

This events set suggests difficulties in communicating information among the units involved in the response. The report does not describe what happened, but it appears the new information available to the dispatcher could not be communicated directly to the firefighters. The events suggest that the process by which communication of the information about the Class A Explosives travels from a truck driver to the emergency response communications center to the field units with a need to know would be useful during the review of driver tasks after collisions.

2.3.2.7. Fire behavior after initiation.

The fire which broke out during the collision continued to burn and grow in intensity. The driver's report of b lack smoke around the rear tractor wheels suggests a tire fire.

Fire burned alongside entire left side of trailer [4-6.0]

Driver saw billowing black smoke around rear tractor wheels

Fire flames reached top of tractor and trailer [4-11.6]

Fire reached middle of semi-trailer [4-16.4]

bout the time flames reached the middle of the semi-trailer, the en-masse explosion occurred. The time between the collision and the en-masse explosion of the cargo was estimated to be about 10-15 minutes. From the actions of the driver, it is likely the elapsed time was more nearly 15 minutes, rather than 10 minutes.

(continued on next page.) 2.3.2.8. Driver traffic control actions

The driver was able to communicate with other truck drivers on both sides of his stopped vehicle, despite the fire that was burning.

Driver told eastbound truck driver dynamite was on burning truck[1-9.6]

Driver told eastbound truck driver to stop eastbound traffic [1-10.6]

Driver ran to east[1-11.4]

Driver shouted same message to westbound truck driver who had stopped [1-12.4]

Driver waved arms, shouted to people to keep away from burning dynamite truck[1-13.4]

Driver went to top of hill west of scene[1-14.4]

This events set provides an indication of the workload that a driver can perform after a collision and before a Class A explosive cargo detonates. In the time available, the driver went from one side of the wreckage to another, and still had time to move a survivable distance from the vehicle before the cargo exploded. The supports the suggested reexamination of driver tasks, mentioned in Case 1.

2.3.2.9. Firefighter actions

The firefighters responded quickly with a minimum crew, arriving on the scene about 10 minutes after the collision, and a few minutes before the large explosion, but apparently after the initial rapid fire detonations had occurred.

2 firefighters left immediately for the fire in Waco [6-8.6]

Firefighters arrive on scene[6-14.2]

Firefighters began to fight blaze[6-15.2]

? [6-16.2]

Firefighters stop hosing fire[6-17.2]

2.3.2.10. Driver-firefighter interactions

No reported interaction occurred between the driver and firefighters.

2.3.2.11. Cargo behavior in fire

The cargo apparently reacted in two phases, involving the blasting cap cargo in the dromedary and then the dynamite and other explosives in the semi-trailer..

Series of rapid-fire detonations occurred

(thought to be blasting caps)[3-13.2]

Cargo detonated (en masse) [3-17.8]

Cargo detonation kills 5, injures 33 people, produces $1,000,000 damages [3-19.0]

2.3.2.12. Post-explosion firefighter actions.

Two firefighters were killed in the explosion. Other firefighting actions were not reported.

2.3.2.13. Post-explosion cargo behavior.

No mention is made in the report of unexploded cargo surviving the explosion

2.3.2.14. Explosion consequences

The detonation of the cargo in the semi-trailer killed 5, injured 33 and produced over $1000000 of property damages. The number of casualties and the amount of the loss (in 1973 dollars) provides a measure of the potential consequences of this type of accident for risk assessment purposes.

2.3.2.15. Shipper response actions

No mention is made in the report of a shipper response to the accident.

2.3.2.16. Evacuation

Except for the actions of the driver, no evacuation of exposed persons was made or attempted by the emergency response personnel in the roughly 15 minutes before the explosion.

2.3.2.17. Accident reporting

The information in this report was among the most extensive reported about this type of accident and was very useful for that reason.

2.3.3. Ajo AZ, 25 June 1976 Fire

This Case involved a single vehicle accident and fire which apparently consumed all the cargo, including about 10,000 pounds of Class A explosives. The data about the accident was flow charted but the reports provide so little information that a detailed discussion of every element of the accident scenario, similar to the discussion in Cases 1 and 3, is omitted from this section.

2.3.3.1. Collision events

In this accident, the collision dynamics involved striking a concrete abutment, and the truck overturning

Truck ran off left side of roadway

Truck hit concrete abutment

Truck overturned

The report does not specify the speed involved, nor does it provide any indication of driver capabilities after the accident, so it is not possible to offer a judgment about the impact forces to which the Class A explosive cargo was exposed. However, is seems reasonable to assume that as the collision dynamics occurred, the load in the truck was jostled, and shifted significantly before the fire.

These events could be interpreted to suggest that the risks of explosions being triggered by severe collision dynamics are low.

2.3.3.2. Cargo behavior in fire

Little is known about the cargo behavior in the fire.

Truck burned

Fire destroyed hazmat

This events set is so abstract that only the most general points can be intimated. One point is that after a vehicle upsets or turns over, the likelihood of fire resulting in a detonation of the cargo may be reduced.

2.2.1.17. Accident reporting

This is an example of a lost opportunity to learn from an accident because of the limitations on the reporting requirements, and the quality of the information furnished voluntarily. Information about the collision events, initial fire events, driver actions after accident began, driver fire fighting actions, driver rescue efforts, police response, fire behavior after initiation or initiation, driver traffic control actions, firefighter actions, driver-firefighter interactions, cargo behavior in fire, post-explosion firefighter or emergency response actions, post-explosion cargo behavior, explosion consequences, shipper response actions, and evacuation are not adequate for analysis.

2.3.4. Bonnieville, KY 10 May 1985 Fire

This Case involved a vehicle accident and fire during which a small shipment of 638 pounds of Class A explosives cargo burned without detonating. The available data about the accident was flow charted but the report provides so little information that a discussion of the accident scenario can not be provided .

2.3.5. W. Frankford, IL 18 Dec 1973 Fire

This Case involved a single-vehicle fire during which the Class A explosives cargo burned without detonating. The available data about the accident was flow charted but the reports provide so little information that a detailed discussion of every element of the accident scenario, similar to the discussion in Cases 1 and 3, can not be provided in this section.

2.3.5.1. Initial fire events

The fire originated in a tire. The report does not indicate how soon after the rupture the driver detected the fire and stopped the vehicle, or how the fire began. Therefore no conclusions can be drawn from this events set.

. Tire ruptured [1-1.6]

? [2-1.8]

Driver maintained vehicle movement[2-2.0]

Tire ignited[1-2.6]

Tire fires are not uncommon in tractor-semi-trailer highway operations, and must be considered likely to occur when assessing transportation risks.

2.3.5.2. Driver fire fighting actions.

The driver eventually detected the fire, and tried to extinguish it with the fire extinguishers available on the truck. The report is ambiguous about whether the tire was burning or the trailer was burning at this stage of the process, and what the driver tried to extinguish.

Driver tried to extinguish fire with two 5# fire extinguishers [2-3.4]

? [5-3.6]

Fire continued to burn [5-4.6]

This events set indicates that either the two fire extinguishers were ineffective against the fire at the stage it was detected by the driver, or that the driver did not use the extinguishers effectively on whatever was burning. Without more detail, no conclusions can be drawn about what happened in this specific incident. It can be concluded that the extinguishers did not provide the driver with the extinguishing capacity he needed in this case, because the fire continued to burn.

When the driver's fire extinguishing efforts were unsuccessful, the driver initiated other action.

Driver unhooked trailer [2-5.6]

Driver pulled tractor away [2-6.2]

Driver opened doors on trailer [2-7.2]

This events set suggests that the driver was attempting to remove the tractor from the threat of the trailer fire, and possibly from the threat of an explosion. The report provides no details about this decision, or the decision to open the doors on the trailer, or the timing or effects of those decisions and actions on the outcome.

2.3.5.3. Cargo behavior in fire

Little information about the cargo behavior in the fire is provided in the report, except that the fire consumed the entire cargo of explosives. The cargo consisted of eight hundred 50# bags of oxidizer, accompanied by 4 packages of Class A explosives and 2 packages of Class C explosives.

Fire continued to burn [5-4.6]

Fire consumed entire cargo of explosives [5-8.0]

This events set suggests that some mechanism was present by which the Class A explosives were protected against explosion during the fire to prevent an explosion. The report provides no details from which further analysis can be made.

2.3.5.4. Accident reporting

The outcome of this accident suggests that some useful lesson might be learned by understanding the progression of events in this case more thoroughly. Unfortunately this accident occurred over 10 years ago, so the likelihood of learning much more about it is low.

2.3.6. Lincoln, AL 20 May 1987 Fire & Explosion

This case involved the collision of a truck carrying 42,095 pounds of Class A explosive ammunition for cannon, a fire and explosion of the cargo. The available data about the accident was flow charted but the report provides little information about the accident.

2.3.6.1. Initial fire events

This case involved two vehicles traveling in the same direction.

( Began to pass tractor-semi-trailer in Right lane) [3-2.0]

Right front struck left rear of Veh #2[3-3.2]

? burst into flames[3-4.4]

The events in this set indicate that the initial fire began while the vehicles were still in motion. The report does not provide enough detail to determine the fuel load, timing or other aspects of the accident needed for analysis.

2.3.6.2. Driver actions after accident began.

When the collision and fire occurred, the driver brought the truck to a stop.

? burst into flames [3-4.4]

Stopped truck 100-150' forward in median[2-5.6]

This events set suggests that the driver's reaction was to bring the truck to a stop in the shortest possible distance. In the absence of data about the fire, this may or may not have been the most desirable action by the driver.

2.3.6.3. Cargo behavior in fire

The Class A explosive cargo reportedly exploded during the fire.

? (tractor) burst into flames [3.4.6]

Struck Trailer and cargo (dump truck) also burned [6-8.6]

Cargo Exploded [4-8.4]

The cargo on the struck trailer, a dump truck, also burned. This suggests that the fire was extensive, either before or after the explosion.

In the absence of addition data, the significance of these events can not be assessed, except that the explosion again demonstrates the possible consequence of transporting Class A explosives by highway

2.3.6.4. Accident reporting

Adequate information about the accident with respect to collision events, initial fire events, driver actions after accident began, driver fire fighting actions, driver rescue efforts, police response, fire behavior after initiation or initiation, driver traffic control actions, firefighter actions, driver-firefighter interactions, cargo behavior in fire, post-explosion firefighter or emergency response actions, post-explosion cargo behavior, explosion consequences, shipper response actions, and evacuation is not available to permit further analyses of the report submitted to satisfy OHMR reporting requirements.

2.3.7. Keystone, WV 5 April 1979 Fire & Explosion

This case involved a single vehicle which caught fire while being driven on a public highway, and subsequently exploded.

2.3.7.1. Initial fire events

The initial fire events were not reported.

?

? began to burn at rear of truck bed

This events set reflects the uncertainty about the initial fire events in this case. It is unlikely that a tire fire was involved because of the terms used to describe where the fire started. Without more data about what the driver observed and a complete description of the entire cargo, the origin of the fire can not be suggested. However, the information provided suggests that the cargo itself can not be ruled out as a possible source.

2.3.7.2. Driver actions after accident began.

The report does not specify what happened to draw the driver's attention to the fire in the truck.

Driver noticed fire in truck bed

Driver pulled truck over to side of road

The events set suggests that the fire had a sufficient start to enable the driver to see it while the vehicle was moving. No basis is available for estimating the relative timing between the initiation of the fire and its detection.

2.3.7.3. Driver fire fighting actions.

The driver may have used fire extinguishers on board the truck to try to extinguish the fire, although the report is ambiguous on what exactly the driver did.

Driver tried to extinguish fire

Driver realized fire was out of control

This events set indicates that whatever the driver did, the action was less than adequate to stop the fire. Significantly, fire suppression was the first reaction by the driver, as it has been in several other Cases. In the absence of data describing what the driver did, or the time the driver spent fighting the fire, the influence of the decision to perform this task on the change in the risks in this Case can not be assessed. Since injuries did occur, the time spent fighting the fire may have reduced the time spent warning bystanders.

2.3.7.4. Driver traffic control actions

Driver realized fire was out of control

Driver began warning people in surrounding area

2.3.7.5. Cargo behavior in fire

The cargo exploded. The report indicates 16 minor injuries, which presumably resulted from the explosion.

Truck (Cargo) exploded (detonated)

Explosion resulted in minor injury to 16 persons

This events set indicates that the driver's attempts to warn people in surrounding area were not completely successful. The report does not describe what actions the warned people took, or who was injured, but the end result of the explosion was the injuries. The case again illustrates the need to review driver tasks following the onset of a fire in transportation.

2.3.7.6. Accident reporting

Information to analyze other aspects of this Case is not reported, including initial fire events, driver actions after accident began, driver fire fighting actions, driver rescue efforts, police response, fire behavior after initiation or initiation, driver traffic control actions, firefighter actions, driver-firefighter interactions, cargo behavior in fire, post-explosion firefighter or emergency response actions, post-explosion cargo behavior, explosion consequences, shipper response actions, and evacuation.

2.3.8. Ishpeming, MI 26 August 1973 Fire

This Case involved a fire which started while a truck was being driver along a public highway at night, and burned up the truck and 1500 pounds of Class A explosives in wooden boxes.

2.3.8.1. Initial fire events

The origin of the fire is unknown. The report contained speculation about the origin which is not used here.

2.3.8.2. Driver actions after accident began.

In this incident, the driver pulled the truck to the side of the road and, presumably, stopped.

Driver smelled smoke [2-2.0]

Driver pulled truck over to side of road [2-3.0]

Driver saw smoke at back of truck[2-4.0]

The events set suggests the driver reacted conventionally to the detection of smoke.

2.3.8.3. Driver fire fighting actions.

The driver's observation of smoke led him to try to find and extinguish the fire.

Driver opened tailgate to put out fire [2-5.0]

Driver saw no open flame [2-5.6]

Driver looked under and around truck [2-6.2]

Driver saw truck burst into open flame inside box 2-7.0]

Driver concluded no hope of putting fire out [2-7.8]

This events set indicates the driver was faced with a difficult diagnostic challenge that he was probably unprepared to meet. In retrospect, it appears that the nature of the fire in this incident was such that it was not readily detectable, and by the time the driver completed his diagnostic procedure, the fire had advanced to the point where it could no longer be controlled.

2.3.8.4. Fire behavior after initiation.

The fire was not visible until after the driver opened the tail gate to put out the fire.

Fire began to burn at unknown location [5-1.8]

Driver opened tailgate to put out fire

? burst into flame [5-7.0]

This events set suggests that the fire may have been starved for oxygen before the driver made air available to it by opening the tailgate, and further that degassing of flammable vapors was occurring as the fire burned, wherever it was burning. The "?" is used to indicate that the material which burst into flames is unknown. Further inquiry might permit logical reasoning to deduce scenarios which allowed the fire to intensify before flames became visible.

2.3.8.5. Driver traffic control actions

When the fire became visible, the driver interpreted the observations to predict that an explosion was possible, and acted on that diagnosis.

Driver saw possibility of an explosion [2-8.6]

Driver stopped all cars 2000' on each side [2-9.4]

The events set suggests that the driver was available for traffic control tasks and performed them. Since the cargo did not explode, it is unclear what effect the driver's action had on the outcome.

2.3.8.6. Cargo behavior in fire

About 1500 pounds of Class A explosive was in the fire. Since the fire events are unknown, it is unclear when the explosive started burning.

(Cargo began to burn) [4-7.6]

? [4-10.6]

Cargo stopped burning [4-11.0][

The events set shows a "?" between the beginning and end of the cargo aspects of the fire, because cargo stopped burning does not necessarily follow cargo bean to burn when dealing with explosives. If more were known about the load configuration and materials involved, deductive logic might be used to determine why the load didn't detonate during the fire.

2.3.8.7. Accident reporting

Insufficient data is reported about initial fire events, driver actions after accident began, driver fire fighting actions, driver rescue efforts, police response, fire behavior after initiation or initiation, driver traffic control actions, firefighter actions, driver-firefighter interactions, cargo behavior in fire, post-explosion firefighter or emergency response actions, post-explosion cargo behavior, explosion consequences, shipper response actions, and evacuation to complete worksheet.

Go to Section C: Railroad Accident Analyses
Return to Section A