BY VYTO BABRAUSKAS
The April 17, 2013, ammonium nitrate (AN) explosion at the West Fertilizer Company in West, Texas, was one of the more infamous accidents involving numerous line-of-duty deaths to firefighters. Fire Engineering covered the news and subsequent developments extensively throughout 2014 and 2015. (See “Explosion in West, Texas: Lessons Learned from Multiple Deaths,” Fire Engineering, March 2015, 137-145. http://bit.ly/1gn7jxv) (photos 1, 2).
One upshot of that disaster was that the National Fire Protection Association (NFPA) declared that NFPA 400, Hazardous Materials Code, the NFPA’s primary means of addressing AN accidents, will be revised to incorporate in its provisions what was learned about that disaster. The 2016 edition (available toward the end of 2015) has now been finalized, so it is important to assess if it will do enough for the safety of firefighters and local communities.
|(1) The West, Texas, AN explosion (Photos 1 and 2 courtesy of the Texas State Fire Marshal’s Office.)|
First, it is useful to take a broader perspective on AN disasters and look at their history. AN is a chemical (NH4NO3) that has the ability to explode.1 Its explosive potential has been recognized for more than 200 years. But unlike some materials with this trait, its explosions are not easily initiated. In engineering parlance, it is an insensitive explosive. It is also an explosive with a moderately low energy yield if it does explode-between 35 and 50 percent that of TNT. This refers to AN by itself. AN combined with fuel oil-ANFO-has a much higher energy yield,2 but the discussion here will be solely on fertilizer-grade AN and not ANFO. Neither of these traits is sufficient to prevent a recurrence of disasters. Farmers using AN as a fertilizer need large quantities. Facilities providing the fertilizer must, in turn, be set up to supply this demand, and large stockpiles have to be anticipated. Thus, even though AN is relatively insensitive and has a low explosive yield, large and devastating explosions are possible.
AN explosions are not just some unexpected rarity; they have been distressingly recurrent. Subsequent to the West disaster, authors writing in an Italian fire safety magazine3 published an article “Ammonium Nitrate: A Century’s Worth of Explosions.” But there have been many more accidents in the United States than in Italy. The safety issues with AN are quite different within the manufacturing operations of a factory, compared to the issues in storage and transportation of the product. Here, we will only look at explosions that did not involve manufacturing operations. It is also useful to exclude ANFO explosions and explosions where other explosive materials near the AN were involved. A review of the literature indicates that around 60 accidents have been described over the past 100 years where a significant fire involving AN occurred in a storage or transportation environment. Table 1 lists 18 of these accidents that resulted in explosions, not just fires.
Uncontrolled Fire: Prerequisite for Disaster
After most of these accidents, certainly those that resulted in deaths, investigating bodies try to determine why it happened. Typically, the investigators try to determine what was unique about the case. This approach is unsound, unsafe, and uninstructive. The consequences are that lessons are not learned and the disasters keep recurring. If examined at a fine enough level of detail, no disaster will ever be found to be a duplicate of a previous incident; thus, the circumstances will always be “new and novel.” The presumed motivation of this thinking is that if the circumstances are always “new and novel,” then no one can be blamed for not anticipating them. But with that mind-set, no lessons are learned. The way to avoid repeating the mistakes of the past is to use a different thinking process.
|(2) The aftermath. (Photos 1 and 2 courtesy of the Texas State Fire Marshal’s Office.)|
One must not ask, “What was unique about this latest accident?” but “What has been the common factor in all the past disasters?” In the case of AN fertilizer disasters, the answer is clear: All of the accidents involved an uncontrollable fire. It does not matter if the AN was stored in a warehouse, onboard a ship, in a railcar, or on a truck-none of these facilities ever blew up unless there was an uncontrollable fire. “Uncontrollable” thus refers not to the size of the fire or the pile but to the circumstances of the accident. Rapid fire detection, a well-equipped fire department, an adequate water supply, and a host of other fireground issues will determine whether a fire is controllable. A large fire can be controlled in some circumstances, but a modest-sized one may defeat the forces at hand in another scenario.
Categorizing the incident types according to how the material was stored-e.g., in a warehouse or onboard a ship-is counterproductive and has impeded learning the lessons from the incidents. The AN fertilizer in all these incidents, regardless of the storage/transport modes involved, is identical: It was in granular form; quiescent; and originally at some reasonable ambient temperature; in most of the cases, no extraneous chemicals played a role unless they caused an initial fire. Only by disregarding the misleading distinctions of the mode of storage can the big picture emerge.
West Disaster Precedent
An uncanny precedent to the West disaster occurred on October 2, 2003, in the small village of Saint-Romain-en-Jarez, in France.4 A farmer stored in a barn three to five tons of AN fertilizer in “big bags,” which is a common way of delivering the fertilizer in France. The fire evidently started in an electrical fixture in the barn and quickly caused a major fire, which was fed by numerous crates and other stored combustibles.
The arriving fire department had difficulties establishing a reliable water supply, thus it sensibly focused on protecting the exposures and had to let the main fire continue burning. Abruptly, as the fire was already decreasing because of fuel consumption, a hissing sound was heard and a massive explosion occurred (photo 3). A sizable part of the village was devastated, and some 29 people were injured; fortunately, there were no fatalities. Among the injured were 18 firefighters, nine of them seriously, plus three police officers.
|(3) Results of an AN explosion in a farmer’s barn in Saint-Romain-en-Jarez, France. (Photo courtesy of the French Ministry for Sustainable Development.)|
Debris from the blast ignited spot fires up to a half mile away. In the West, Texas, disaster, an electrical fault is thought to be one of three potential causes, the fire department also experienced water supply difficulties, and the explosion also came without warning. The major difference between the two incidents is that the French explosion involved one-tenth the quantity of AN at West, Texas. Although the French incident is exceptionally instructive, perhaps it was disregarded because of the lack of fatalities.
Although 100 percent of the explosions in Table 1 were preceded by an uncontrolled fire, a fire does not have to result in an explosion of AN. Case histories indicate that, historically, there is about a 30-percent probability that a fire involving AN will lead to an explosion. Furthermore, the likelihood that a fire involving AN will lead to an explosion causing fatalities is only 15 percent. But it is little comfort to the relatives of the victims or to the community that the general probability is 85 percent that an AN incident will conclude with nobody killed. Given the fact that AN incidents are relatively rare to begin with, many of those responsible for safety justify inaction because of the low probability of a catastrophe. But that should not be the policy when the consequences of fire safety inadequacies can be disastrous.
Preventing Uncontrollable Fires
If it is understood that an AN disaster will not happen without an uncontrollable fire, then the necessary strategy is clear: prevent uncontrollable fires. The focus here will be solely on AN in buildings, since protection measures for transportation are different. To achieve adequate safety does not require elaborate schemes. Observing a few pivotal fire safety principles would suffice:
- A more safe product formulation than straight-AN fertilizer.
- Noncombustible building facilities.
- Noncombustible AN containers.
- Allow no places where molten AN can pool.
- A monitored fire alarm system.
- An automatic fire sprinkler system fed by a reliable water supply.
- Mandatory, not optional, safety provisions.
Most of these provisions are obvious to safety professionals, but a couple need explanation. It was already learned in the 1920s that there are safer alternative products for providing nitrogen to crops. Not counting products that are physically very different (e.g., liquid fertilizers), it is possible to combine AN with some other substances to achieve a product much less likely to detonate. Many countries in Europe have for decades now used calcium ammonium nitrate (CAN) as a safer alternative.5 Ammonium sulfate nitrate (ASN) is also used in some countries. In addition, there have been a number of safer alternatives patented but not yet in widespread use. The AN manufacturers in the United States; however, have resisted making any changes to the status quo, and CAN has been available only as an import from Europe.
Analysis of AN explosions indicates that the melting of the material plays a crucial role in the initiation of detonations. A thin layer of molten material is unlikely to detonate; a pool of substantive depth may. Thus, it is important to avoid creating places, such as elevator pits, where AN could accumulate in depth if it melts and flows.
Who Can Prevent AN Fires?
In principle, a number of entities could prevent uncontrollable AN fires, such as the facility owner, but in practice, none has been doing so. But this would be an unreasonable expectation because owners are not required to do so; AN distribution facilities are commonly small businesses that only have agricultural expertise and none in industrial loss prevention; and viable guidance has not been available for them even if they wanted to make such efforts.
The federal government has three agencies with a mission in this area: the Occupational Safety and Health Administration, the Environmental Protection Agency, and the Department of Homeland Security (DHS). Yet, the federal Government Accountability Office documented that although the DHS has been operating a successful program against diversion of AN into the hands of those who might want to improvise explosive devices, none of the agencies has had any effective programs for fire or explosion safety at the storage facilities.6
State or local authorities might also ensure safety, but this has little basis in fact. Fertilizer is an agricultural commodity; thus, the primary warehousing will necessarily be in rural areas. The West, Texas, fertilizer building was not in any municipality; it was just outside the West city limits. Typically, such rural areas may have no building department and depend on volunteer fire services. Inspection of facilities for chemical hazards is not within anybody’s job description or skill set in such rural areas.
This leaves the producers of AN. Unlike fertilizer distribution facility owners or rural local officials, AN producers are major chemical manufacturing firms. In many parts of the chemical industry, manufacturers have voluntarily espoused the principles of “product stewardship.” One of its tenets is that a chemical manufacturer is responsible for a molecule from the time it is first produced to the time that it ceases to exist. Under such a principle, a manufacturer will not sell hazardous chemicals to a buyer without first personally ascertaining that the buyer can handle them safely. Typically, this means that the manufacturer sends a representative to inspect the safety of the buyer’s facility before authorizing the first sale of product. Companies that wish to belong to the American Chemistry Council, the National Association of Chemical Distributors, or the Society of Chemical Manufacturers and Affiliates must agree to abide by product stewardship principles.
Unfortunately, U.S. manufacturers of AN fertilizer have not accepted this. Although they process anti-diversion paperwork to the satisfaction of the DHS, they consider that their responsibility for the safety of the product stops when it leaves the factory. Regarding obligations to their buyers- e.g., companies such as West Fertilizer Co.-they consider their responsibilities discharged by publishing a material safety data sheet (MSDS) and posting it on their Web sites. The MSDS of the U.S. manufacturers, in turn, have been distressingly brief and lacking; the main safety instruction in them is to refer to NFPA 400, Hazardous Materials Code-in other words, “Let the NFPA take care of it.”
NFPA 400, Hazardous Materials Code
Relying on the NFPA would never be appropriate, simply because the NFPA is not an organization that inspects facilities, but inspection is crucial if meaningful compliance is expected. Nonetheless, it is important to examine the adequacy of NFPA 400. The 2013 edition (issued June 2012) was in effect at the time of the West, Texas, explosion. The U.S. Chemical Safety Board (CSB) evaluated the 2013 edition and concluded7 that many of its safety provisions “appear to be confusing or contradictory, even to code experts, and are in need of comprehensive review in light of the West disaster.” They also pointed out that the Code failed to prohibit wooden buildings or wooden AN storage bins and failed to require a sprinkler system. Worse yet, “grandfathering” provisions meant that only new buildings, not existing ones, would need to comply with various safety provisions.
But the situation was even worse than that, since the provisions of NFPA 400-2013 did not strictly mandate any of the crucial safety requirements considered above. In addition to the technical shortcomings, safety provisions will not ensure anybody’s welfare if they are not mandatory. Here, the situation with NFPA 400-2013 was also deeply problematic. Apart from being so obtuse as to be “confusing to code experts,” along with outright “grandfathering” away most of the crucial safety aspects of building construction, the document left much of the determination of what can or cannot be done to the authority having jurisdiction (AHJ). In the case of the West Fertilizer Co., there was no such entity. The town of West had no building code and no ability to inspect industrial premises in a meaningful way. But the facility was outside the West city limits and would not have been subject to its inspections even if the city had had a building department. A reasonable interpretation of such a code, in the absence of an AHJ, is that anything that an AHJ would potentially be able to allow is automatically allowed.
In response to the CSB’s urging of code revisions, the NFPA has just approved NFPA 400-2016. A few of the glaring omissions in the 2013 version have been remedied (Table 2). Notably, a monitored alarm system and an automatic fire sprinkler system are now specified, as is clear wording that prohibits AN pooling places such as elevator pits. Combustible building construction and wooden bins are also prohibited for new construction, but this is likely to have a negligible safety impact. Use of AN as a fertilizer has been decreasing for many years, and most of the storage/distribution facilities are many decades old. Because usage is declining, there is little construction of new facilities, so to improve fire safety, the focus would have to be on existing operations, not new construction.
But the main reason a new NFPA code edition is not likely to do anything to save firefighter lives is the location of these facilities. NFPA 400 gives great latitude to the AHJ to make specific decisions for any particular facility. The majority of the facilities are in sparsely populated rural areas that either do not have an AHJ or cannot afford a building department with useful expertise in industrial explosion hazards. Thus, irrespective of what the NFPA might publish, technically sophisticated AHJs are not likely to appear where they are not present today.
Will the Emergency Response Guidebook Help?
Finally, what about the place where the rubber meets the road, the Emergency Response Guidebook (ERG)? The firefighters at West had copies of the ERG in their trucks, and they knew about the book and its purpose. For AN fires, the current ERG (2012 ed.) specifies: “Small fire: Use water ….” “Large fire: flood fire area with water from a distance ….” NFPA 400-2016 introduces some additional concepts that may possibly be incorporated into a future edition of the ERG, including the requirements for an emergency action plan and for sirens to alert the population and a suggestion to evacuate up to a one-mile radius of the incident site.
But it is dubious that any of these provisions would be sufficient to prevent uncontrollable fires from developing into disasters. It is easy to say that nobody will get hurt if everybody is far away from the seat of an explosion, but an evacuation is difficult to implement. Many individuals are likely to be unable to evacuate themselves and, in the West disaster, a retirement home was one of the nearby properties; one resident who died was 96 years old. Although it might be the safest option for firefighters and emergency medical service personnel to pick up and run fast, this is hardly consistent with their sense of responsibility or the community’s expectations.
A successful evacuation of civilians, sirens notwithstanding, is not likely to happen without fire/rescue personnel going door-to-door. In the West disaster, about 12 minutes elapsed between the arrival of the first engine and the explosion. It is not reasonable to expect that a door-to-door tour, plus active assistance to civilians in need, could be accomplished anywhere close to that time.
Unfortunately, responses to these types of incidents often focus just on actions that took place after the 911 call was made. By then, it may be too late to prevent a disaster. In some circumstances, no amount of training or prefire preparation will suffice to make an incident of this type tolerable, even if it were to happen in the jurisdiction of a well-trained, large city fire department. Even if firefighters pull back from firefighting and focus solely on evacuation, no fire department can safely take all the needed actions if a major explosion is going to occur while personnel are still setting up incident command. Thus, facilities must be designed in such a way that they do not expose firefighters to this kind of danger. As with NFPA 400, it is not reasonable to expect that any revision of the ERG will provide safety-by the time the ERG is needed in such an incident, the battle has been lost.
1. Chaptal, Mia, Elements of Chemistry (third ed.), vol.1. GG and J Robinson, London (1800). 237.
2. Babrauskas, V. Ignition Handbook, Fire Science Publishers/Society of Fire Protection Engineers, Issaquah WA (2003).
3. Pasturenzi, C, L Gigante, and P Cardillo. “Nitrato d’ammonio: Un secolo di esplosioni.” La Revista dei Combustibili 67:2, 19-28 (2013).
4. Aria No. 25669, “Incendie dans un hangar agricole et explosion d’engrais,” 2 Octobre 2003, Saint Romain en Jarez (Loire), France.” Ministè;re chargé du développement durable, Paris, France. An English translation of the report has also been published. Unfortunately, the English translation mis-translates “ammonitrates” (ammonium nitrate fertilizer) as “ammonium nitrate solution.” In fact, “ammonitrates” denotes the granular form of AN, and solutions of AN are not sold for farm use in France.
5. Médard, L A, Accidental Explosions, 2 vols., Ellis Horwood, Chichester, England (1989).
6. “Chemical Safety: Actions Needed to Improve Federal Oversight of Facilities with Ammonium Nitrate.” (GAO-14-274), U.S .Government Accountability Office, Washington (2014). http://www.gao.gov/products/GAO-14-274.
7. “Preliminary Findings of the U.S. Chemical Safety Board from Its Investigation of the West Fertilizer Explosion and Fire,” U.S. Chemical Safety Board, Washington (April 2014). http://www.csb.gov/assets/1/19/west_preliminary_findings.pdf.
VYTO BABRAUSKAS, Ph.D., earned degrees in physics and structural engineering and a Ph.D. in fire safety. As a researcher at the National Institute of Science and Technology, he developed devices to measure the heat release rate of products and developed a computer program for modeling the development of room fires. He founded a consulting firm in 1993 and provides fire safety science expertise to fire investigation and litigation.
What Can Firefighters Do?
Clearly, firefighter lives will remain in danger if mass civilian evacuations are required. But this will only be necessary if AN is allowed to be stored in combustible facilities in which there is danger of uncontrolled fire developing. Such hazards would no longer exist if the AN storage facilities are rebuilt properly as noncombustible structures with appropriate fire safety features. For example, no incompatible storage-i.e., any combustible materials (wood, paper, plastics, hydrocarbon liquids, and so forth). AN is an oxidizer and presents a grave danger if combustible materials come in contact with it. This is already part of the NFPA’s safety prescriptions. The existence of NFPA 400-2016, by itself, will do nothing to advance this objective. And any “solution” that does not involve mandatory and thorough upgrading of unsafe buildings will not achieve positive results.
But the picture need not be bleak. AN would not be stored in unsafe facilities if the AN manufacturers stopped selling to customers with such facilities. So, the most viable strategy for the fire service is to persuade the manufacturers that they must meet the requirements of Product Stewardship and agree to sell AN to a customer only if the product will be stored in a fire-safe storage building. Such a change can happen overnight if the industry chooses to comply.
Fire Engineering Archives