Friday, April 19, 2013

That devastating explosion at the Texas fertiliser plant - was it nitrous oxide assisted?

 Update: 11th May 2013:   Criminal probe opened into Texas fertilizer plant explosion
(This Independent article initially invited comments - to which I responded - but the article later appeared stripped of comments, presumably for legal reasons).

This is just a holding post. I'll insert the comments I've been posting on the Guardian and Independent later. They will show an evolution in my thinking, informed (or misinformed|) by the Indy quoting from an US inspector's report that the site DID have ammonium nitrate as well as anhydrous ammonia.

As the search and rescue operation continued, agents from the federal Bureau of Alcohol, Tobacco, Firearms and Explosives scoured the scene for clues to the cause of the blast, which officials said had destroyed about 50 homes. A recent report submitted to the Texas Department of State Health Services suggested the facility contained a stockpile of up to 270 tonnes of ammonium nitrate, and 100,000lb of liquid ammonia.
(ed: normally, when you link to a report in a newspaper, that report stays the same. The UK's Independent is a notable exception.  Articles are not only edited but can be completely rewritten - while still appearing under the same URL. Let the internet surfer - or in this case blogger - beware)

Here's the way my thoughts are going. First, one does not expect liquid ammonia (NH3) to behave like rocket fuel when tanks are ruptured by fire. Escape of ammonia gas yes. Fireball, no. Surely there would be well known precedents if that were the case, with regulations siting the tanks far from population centres. Secondly, one does on the other hand have precedents too numerous to mention of ammonium nitrate producing devastating explosions that have flattened entire communities and killed scores, sometimes hundreds. But there is scarcely any mention of ammonium nitrate as a causative agent in connection with West-comma- Texas near Waco.
(ed: the above sentence was true at the time of writing; we have since learned that some 270 tonnes (yes!) was stored on that site, in breach of National Homeland Security regs).

Here's my hunch. There was BOTH liquid ammonia AND ammonium nitrate on site, and the heat of the initial fire (wooden pallets?) was sufficient to melt the ammonium nitrate and then rupture the ammonia storage tanks.

When molten ammonium nitrate decomposes it produces nitrous oxide, N2O, a gas with similar oxidising properties to oxygen, able to support combustion ( even relights a glowing splint as I recall, the "standard" test for oxygen).

Ammonia gas is reluctant to burn in air, unless introduced into an existing flame, and raised to its air/ammonia ignition temperature of close on 1000 degrees C.

But suppose there had been a mixture of ammonia gas and nitrous oxide, the latter behaving like oxygen. Might that have not been the reason for ammonia behaving like rocket fuel?

NH4NO3 ->  N2O  + 2H2O
ammonium nitrate -> nitrous oxide and steam

2N2O  -> 2N2 + O2
nitrous oxide -> nitrogen + oxygen

4NH3 + 3O2  ->  2N2 + 6H2O
ammonia + oxygen -> nitrogen + steam

Result - in gas phase - BANG! 

Overall word equation:

ammonium nitrate(solid)  + ammonia (gas) -> nitrogen(gas) and steam  (highly exothermic reaction, both products gaseous, to produce a vast fireball and powerful pressure wave)

Further reading: Firefighting and Anhydrous Ammonia

Sample comment from the Guardian:

@Daveinireland - All we know for certain is that liquified, pressurized anhydrous ammonia was on site.
Theoretically it could have produced the fireball if the ignition temperature of air/ammonia gas mixtures of close on 1000 degrees C had been reached. But if the accident investigators know their explosives chemistry, they will be looking for traces of solid ammonium nitrate and/or for any evidence that the solid was either shipped in, or made on site by reacting ammonia and nitric acid.
Methinks we would have had more precedents in the past for what happened at West, Texas if anhydrous liquid ammonia really were given to behaving like rocket fuel when there's a nearby fire.

Sample comment from the Independent

on Hundreds believed injured in Texas fertiliser plant blast 8 hours ago
The short science lesson would have been useful, had it been correct.
Ammonia is a gas at ordinary temperature and pressure. It does not form explosive mixtures with air, but will combust to nitrogen and steam if introduced into an existing flame, the latter being maintained. In other words, ammonia and air alone will not produce a conflagration and certainly not an explosion. Were that the case, there would have been hundreds of explosions from leaking refrigeration plants.
It can be liquified under pressure alone (normal temp and pressure). The only significance of that -33 degrees C is that any leak of liquid ammonia would result in the liquid boiling at the temperature, and quickly changing to highly toxic gas.
As for the fireball, it's hard to see how that could be due to liquid ammonia, even if the tanks ruptured, although the gas might have assisted an existing fire, as mentioned above. What is far more probable is that solid ammonium nitrate was also on site, and that as we know can be a very dangerous and unpredictable substance if exposed to heat alone. Leaving aside specialized detonation to produce instant decomposition with powerful shock wave ("fertilizer bomb") ammonium nitrate produces vast amounts of gas when it thermally decomposes to nitrous oxide and steam, both gaseous, and the only products. One imagines that it was a runaway decomposition that produced the vast fireball that was for all intents and purposes an explosion, even if the chemical purists prefer to describe it as a rapid decomposition rather than explosion.
It's bad enough that liquid ammonia was stored so close to a community - but if reports that ammonium nitrate was also there are correct, then words simply fail one. Have they not read their history books? Many, probably most, of the worst most devastating chemical disasters in history in which hundreds have lost their lives have involved harmless-looking, salt-like ammonium nitrate.
NH4NO3 (solid) -> N2O (gas) + 2H2O (steam)
That is the chemical equation for a potential time bomb - especially if fire breaks out in the vicinity. All that reaction needs is the right (wrong) kind of kick start.

newsjunkie aka sciencebod

Guardian (again)


"Officials at first suggested the explosion was caused by the anhydrous ammonia igniting, but it was revealed Thursday through Texas state records that the plant also possessed 270 tons of ammonium nitrate, a much more volatile, dry solid, at the end of 2012. Records also suggested that in 2006, the Environmental Protection Agency (EPA) issued a $2,300 fine for deficiencies in the plant's risk management plan."

If there had been even a modest amount of ammonium nitrate as well as anhydrous ammonia on site, and the two were pre-heated close to one another, there is an explanation for that massive fireball.
Ammonium nitrate decomposes to nitrous oxide, N2O, an oxidant that behaves almost like pure oxygen (relights a glowing splint). Ammonia, reluctant to burn in air except at high ignition temperatures (close on a 1000 degrees C) can burn in pure oxygen, so probably in nitrous oxide too, maybe at much lower temperatures than 1000 degrees.

Result: liquid ammonia becomes effectively rocket fuel if/when there is ammonium nitrate + fire nearby.

Postscript: have just this minute discovered a paper that backs up my hypothesis: nitrous oxide supports the combustion and flame propagation of ammonia, the main product being nitrogen gas.

Here's the title and authors of the 1964 paper:

Combustion of ammonia supported by oxygen, nitrous oxide or nitric oxide: Laminar flame propagation at low pressures in binary mixtures

  • Department of Physical Chemistry, University of Leeds, Leeds 2 UK


Postscript: Virgin Galactic's SS2 uses a solid rubber compound as fuel, which is burned in liquid NITROUS OXIDE.  Never undersestimate the oxidising power of nitrous oxide.