NOTE: I received this really detailed analysis of at least
one potential explanation of why the Athens, TX ammonium nitrate facility did
not explode Thursday like the one did in West, TX a year ago. It came to me in
an email from Jim Overman who has graciously allowed me to post it here.
Patrick:
I really think there needs to be a lot of technical work
done on the response of AN to fire. I have gotten a lot of questions from
friends about one had a devastating explosion and the other just burned.
You can include the incident in Bryan, TX a few years ago as well.
I'm sure that research would reveal countless others.
Certainly, there are many factors involved in what happened
in each of these circumstances. I still think a significant factor is the
volume to surface area ratio. As I have stated several times, the rate of
reaction is dependent on concentration of the reactants (which in the case of
the gas phase of any reaction is definitely related to pressure) and the
temperature.
As you know, the rate of reaction doubles with every 10
degree (K) increase in temperature. (I recognize this is only a rough
guideline). Also, the rate of energy production of a reaction is directly
related to the rate of the reaction (we are discussing exothermic reactions).
This is one reason we have runaway reactions. We utilize this
intentionally with commercial explosives.
If the rate is very slow, we can use such common terms as
corrosion in reference to the reaction. If the reaction is more vigorous,
we use terms like "fire". It really gets exciting when the rate
is high enough to use terms like deflagration and if the reaction moves fast
enough through the medium to exceed the speed of sound, when refer to it as a
detonation.
Essentially, in oxidation reactions like those between iron
and oxygen, the only difference between rusting and burning is the rate of
reaction. We all know that steel wool will burn in pure oxygen.
This is an example of concentration of the reactants influencing the
rate.
In general, we use water effectively on Class A fires
because the evaporation of the water removes heat from the reactants (e.g.,
wood and oxygen) faster than it is being produced. (It is also pertinent
to this discussion that the water vapor produced reduces the concentration of
the other reactants.) Now, consider fires in bulk materials. If we
can't get water to the hot reactants fast enough, the reaction continues to
produce heat faster than the evaporating water removes it and the fire will
continue to burn until there are not enough reactants (concentration).
Now, look closely at pictures of the AN storage in Athens
[PJC – For example
here
and
here].
The piles are not especially deep and the bins are are more like rooms an
not very high. That means that the ration of volume to surface area is
low. In other words, the total surface area of a given bulk is higher in
a shallow pile with higher length and width. This is the relevant heat
transfer equation:
Fourier's Law express
conductive heat transfer as
(1)
where:
q = heat transfer (W, J/s,
Btu/s)
A = heat transfer area (m2,
ft2)
dT = temperature difference
across the material (K or oC, oF)
s = material thickness (m, ft)
Note that the only variables that really count are A and s.
As A gets smaller and s gets larger, q gets smaller. In other
words, physics works and less heat gets removed from the bulk and the reaction
rate goes up. As the temperature of the burning material goes up, dT
actually increases so sometimes the change in dT will balance the other factors
and we will have steady state burning. This explains why "cooling
the pile with hose streams sometimes works. However, if the increase in
dT does not balance things out, we have a recipe for disaster.
There are many other variables involved, but unless
contamination is a factor, I suspect this may be the most significant.
After all is said and done, standards that do not address this issue will
fall short when fires actually occur.
Jim Overman
Commentary
Jim makes some very important points about the heat transfer
effect in an ammonium nitrate fire. As he mentions in his closing paragraph
contamination of the ammonium nitrate may also play an important role. This was
noted in the Chemical Safety Board’s
interim
report on the West explosion. Another factor that could come into play is
building collapse as that can have a direct pressure effect, reduce heat transfer
and impede the flow of water to the ammonium nitrate stack.
At this point it does not look like the Chemical Safety
Board will be doing an investigation of this accident. There was no loss of
life and there was only limited (on a grand scale) property damage involved. As
a single incident goes this just isn’t important enough to tie up the CSB’s
limited resources. As we are finding more and more of these under-regulated
ammonium nitrate storage facilities across the country, perhaps it would be
important for the CSB to get involved in the investigation to help prevent
another West Fertilizer type explosion. That would almost certainly take a
congressional request, something unlikely to come from the Texas congressional
delegation.