There seem to be any number of ideas about am. nitrate.
The stuff is slightly hygroscopic. It's a useful (and fairly cheap) nitrogen source so it gts used as a fertiliser.
Unfortunately, as a powder, it gets damp then forms into solid lumps that are difficult to handle.
So someone came up with the idea of coating prilled AN with wax. The wax stops the grains sticking together.
Even more unfortunately this was found to accidentally generate a fairly good explosive (albeit, one that's quite hard to detonate). Further refinements of this gave rise to the low density prill based ANFO used as an explosive.
Since the decomposition of AN is exothermic, it's easy to see how the reaction could get out of control.
In particyular, the reaction is catalysed by impurities so, as the AN decomposes and leaves the tube (as H2O/N2O) any impurities left behind get more amd more concentrated. This would increase the effective reaction rate.
That raises the temperature.
I guess you can all see the punchline now.