I was under the impression that the graphite never burned at Windscale, only the metal fuel. What graphite was lost likely simply evaporated from the heat of the fire but did not really contribute much to sustaining it. If it had, then I think the fire would have kept raging when they turned the cooling fan off. A lot of people assume that because graphite is pure carbon, it burns like some kind of high grade coal. In reality, graphite is a very strange material and is extraordinarily resistant to fire even though its constituent, carbon, should thermodynamically favor combustion. Its incredible stability is why it has been used as electrodes in open air arc lamps. The only instance I have heard of graphite actually burning is at Chernobyl reactor number 4, where it was subjected to uncommonly intense external heat from recently-melted nuclear fuel. Even there, there is debate as to how much graphite actually burned. In any case, graphite's flammability or lack thereof has never actually caused any nuclear accidents and it shouldn't really be all that much more dangerous than water moderated reactors. Containment's a big deal too. If Chernobyl had had TMI's outer containment, it would have been minor and if TMI had had Chernobyl's outer containment, it would have been a lot more like Chernobyl.