Why is it appropriate to compare the worst case for various methods of power generation? I think it's more appropriate to weigh the potential failure cases of each by the probability of them occurring, and I suspect if we did that, nuclear wouldn't be the most harmful method of power generation.