BAFFLED INVESTIGATORS still have no convincing explanation for the spectacular failure of WTC7, the first major modern steel framed building to collapse catastrophically solely as a result of fire.
In their report, the investigators admit that their best hypothesis has 'only a low probability of occurrence', and that much more 'research, investigation and analyses are needed'.
Unlike the failure of the twin towers, which showered debris in all directions as their external structure peeled away, WTC7 appeared to implode, with the facade coming straight down and creating only a small debris field.
Videotape evidence has convinced investigators that the most likely location for the initial failure was on the fifth floor, near the eastern end of the building.
This was in an area of massive transfer trusses and columns extending over seven storeys and transmitting loads from the conventional upper storeys into foundations located irregularly in and around a pre-existing major electricity substation below.
Most of these floors were given over to switchgear and transformers, with the only major fire load coming from the fuel storage for the emergency generators located at various levels within the building.
The problem for the investigators was constructing a scenario in which enough of this fuel could get into the area of the fifth floor where the failure probably occurred. Even then, it had to burn long enough and hot enough to cause the steel to yield.
Fires originally broke out on WTC7 on at least six floors from floor six upwards immediately after the collapse of the second twin tower.
The building's sprinkler system appears to have failed to cope and, with the building successfully evacuated, the emergency services made no attempt to fight the fires.
As the fire spread, observers noted that the smoke plume from the upper floors was dark, while that from the lower floors was distinctively white.
The ASCE/FEMA report makes no attempt to clarify this particular mystery, focusing instead on a two story mechanical equipment room on the fifth and sixth floors.
Two major trusses ran through this space. Outside the double doors was a generator set and fuel supply lines connected to diesel tanks at ground level. Suppose, says the report, that these fuel lines were partially ruptured during the initial impact, which also cut the mains power?
The diesel generators would automatically have cut in, supplied by a fuel pump with a capacity of 280litres/minute, powered by an electric supply that continued until collapse. Even if the generators eventually choked on the fumes, the pressure differential caused by the partial rupture would have kept the pump running.
Investigators found a maximum of 45,000 litres of fuel was missing from basement tanks. At 280litres/minute, this would have lasted less than three hours - but if only 1,100litres/minute was actually escaping - this could have continued for seven hours.
Such a flow would certainly have contained enough energy to cause the trusses to fail. But between the fuel line and the trusses stood a double door in a masonry wall, so the investigators have to assume that either the door was left open or that it was never intended as a fire door.
As well as calling for more analyses of the WTC7 collapse itself, the report recommends more study into the 'adequacy of current code design provisions for members whose failure could result in large scale collapse'. This includes levels of fire resistance.