A competent study (i.e., including x-ray diffraction to look at the specific species of iron, iron oxide, aluminum & aliminum oxide, and anaerobic ignition) will tell if there really is any thermite in the samples.
I am curious why Ferrar says that standard office fires will only get up to about "half of 1100°C".
The folks at Manchester Fire Engineering disagree:
The number he is quoting is true for an external fire (see BSEN1991 "external fire" - purple line - in graph above), but not for one contained as in the rubble pile.
And standard fire curves:
show 1100°C being reached after about 3 hours of burning. How many weeks did those fires burn underground? 10 weeks? 15?
This data shows a consistent pattern of the temp of large steel I-beams reaching about 1050°C after 2 hours,
and still rising...
Finally, all of the curves show the higher the fuel loads, the higher the max temperatures. Using Bazant's compaction ratio of 0.2 gives, as a first order estimate, a fuel area density in the rubble pile 5x greater than in the precollapse building.
I seriously doubt that even the highest fuel load scenario in the tests reported above had the equivalent of five floors worth of office contents jammed into one floor.
