Now this is something that can probably be tested or at least discussed using mathematics and known values. If someone can come up with the explosive power of JP4 when ignitiing 10,000 gallons (and of course not all of it would ignite at once, and some of it might be still be bulk liquid and some spread out in a fuel-air mixture). Anybody up to some calculating?
Back of the envelope:
A fuel-air mix has an RF of about 5, so 10,000 gallons (at about 6lbs/gallon) is an explosive force of 150 tons of TNT (0.15 kilotons), assuming it was all aerosol and ignited simultaneously. Obviously, this was not the case.
The impact caused a rapid combustion, rather than explosion (the damage was heat based, much reduced blast wave compared to explosives). Assuming a contained liquid (RF value ~.8), It's equivalent to 4 tons of TNT (0.004 kilotons), which is really not that large (I've seen larger explosions live).
Seems to me (eyeballing, and recall this is back-of-the-envelope calculation) that the explosion witnessed falls somewhere between these, towards the low end. This would make sense if part of the fuel was dispersed when ignited. The disperesed fuel provided the initial fireball, the remaining fuel (no longer contained) was not explosive, but simply burned.
As pointed out, the fuel tanks are located in the wings and central area of the aircraft, not in the front portion. The tanks would have had to rupture upon entry and then an ignition source would have had to ignite the fuel. Meanwhile at least some of the fuel is still travelling at an appreciable fraction of the 500MPH speed (that old man Newton, again). Given that the building was only about 208 Ft across (IIRC, please correct me if I am wrong) the fuel/wreckage, ignition sources (engine heat, for example) would cross that space in:
500MPH = 733.5 Feet per Second (FPS)
400MPH = 586.8 FPS
300MPH = 440.1 FPS
So it seems plausible that the fuel, exploding forward as the tanks ruptured, would be able to explode out of the building in the manner we saw. But I am willing to listen to your conclusions if the above is not satisfactory.
If it helps, a fuel explosion typically has expansion speeds on the order of 15,000 fps (IIRC, someone may want to check this). I know it's about 24,000 fps for C-4 and other cutting charges, 21,000 fps for Ammonium nitrate and other cratering charges, and I believe fuels are from 12,000 to 18000 fps. Then, consider that the plane has to travel partway intot he building (for the fuel tanks to impact), and also much have an ignition source (heat/spark/electricity, which souldn't be much of an issue but still takes time). The plane itself is about 160 feet long, and about 60 feet from nose to wing. So, we have a travel distance of about 100 feet from wing impact, until the plane is completely inside the building. More precisely, the wings and fuel tanks are only 20 to 30 feet long (front to back), so that's really what we have to work with. How much would the fireball expand before being inside the building? So, taking our 733 fps speed, and looking at a 30 ft distance, we have about .041 seconds for expansion, which gives us a distance of 615 feet. Note that this assumes a gasloine bomb that detonates on impact. In reality, the container would be broken open before ignition, which would limit this expansion further. Not to mention there would be a slight delay between impact and ignition, perhaps as much as .01 seconds, which would knock another 25% off this distance.
Not this is all back-of-the-envelope calcualtion, I'd have to check sources to get anything more precise.
