Friday, 19 April 2013

Likely sequence of events during West Fertilizer explosion, Texas

UPDATE: I have been quoted in Chemistry & Industry 77, May 2013

The cause for the recent incident at the West Fertilizer site in Texas is under investigation and remains unknown, but many parallels can be drawn from previous similar events involving large quantities of inorganic fertilizers.

Aftermath of the mass explosion following a fire in West, Texas, April 17, 2013.
Photo by REUTERS, Mike Stone.
It is the ammonium nitrate (AN) that poses the best known fire and explosion hazards in fertilizer storage sites, especially of the NPK fertilizer type (nitrogen, phosphorous and potassium). Some media outlets are speculating about exploding anhydrous ammonia tanks, but that is a very rare event not ever observed before. Unfortunately, mass fires and explosions in AN warehouse are not uncommon events (average worldwide frequency is about one every three years). One example, in 2001 the AN warehouse of a fertilizer plant in Toulouse, France, exploded resulting in 30 people dead and >2000 injured. The blast wave shattered windows up to 3 km away [source: wikipedia].

I am confident the Fire Service was aware that situation was very difficult and probably had a special emergency plan to deal with this particular site. They attended the fire to comply with their duty in the face of extreme danger. Their main priority would be to control the fire so it does not grow to the critical size when an explosion of AN could be triggered. The science behind a mass explosion following a fire in AN plants is still in bare bones, we know so little, and cannot be predicted. So imagine how difficult it is to deal with the emergency.

The source of the hazard is the exothermic decomposition of AN which begins around 200-230 ◦C. It has been suggested that it follows two reaction paths (the second is more exothermic):

NH4NO3→ N2O + 2 H2O
4 NH4NO3→ 3 N2 +2 NO2 +8 H2O

(a) Unreacted NPK fertilizer granules and (b) cross section showing partially reacted sample with 4 phases visible. Photos from Hadden and Rein 2007.

The fire could have been initiated by self-sustaining decomposition (SSD). This is the phenomenon in which the temperature of a bed of AN-fertilizer rises due to spontaneous heat generation until thermal runaway leads to a fire. The flames would have had then spread to other flammable materials in the plant, like supplies, fuel, packaging, offices or vehicles. SSD of fertilizers is promoted by chemical compounds present in NPK and also the accidental contamination with organic materials. It can start at around 100 ◦C, which is a significantly lower temperatures than that required for pure AN decomposition.

A likely sequence of events is that an accidental heat source (e.g. hot work, hot surface, small fire) starts a SSD reaction in a bed on AN-fertilizer which slowly grows and leads to the fire that the Fire Service were battling. At some point, the flames grow faster than expected and rapidly heat very large quantities of AN, which leads to detonation (=explosion and blast caused by the very rapid decomposition of AN inside an enclosure).

A large detonation wave like this one devastates life and structures over a wide area around the point of origin. Moreover, the burning fertilizer becomes airborne with the explosion and lands further away igniting subsequent fires, as seen in the aftermath of this explosion.

Most of the information I used is from our 2007 paper:
- R Hadden, G Rein, Small-scale experiments ofself-sustaining decomposition of NPK fertilizer and application to events aboard the Ostedijk in 2007, Journal of Hazardous Materials 186, pp 731–737, 2011. doi:10.1016/j.jhazmat.2010.11.047.