mheslep said:
First, that is fallacious, as I think you'll see on rereading, begging the question: "It is extremely difficult to get an accurate risk assessment of an event that is exceedingly rare/unlikely." and then go on to state it is a "fact that it is exceedingly rare/unlikely".
Well, visibly it didn't happen yet. Maybe it never will. In order to get an idea, an order of estimate, of some class of events which are so elusive that no a priori mathematical model can tell you, you need some statistics. Visibly we're not talking about a probability of more than 1 in 10 years, right ? There have now been enough "slices of 10 years" where it wasn't strictly speaking impossible to have a nuclear terrorist bomb, and we didn't have one. So claiming that we will have one every 3 years is not going to be correct. But is it one every 20 years, or one every 200 years ? Or one every 2000 years ? We'd need at least 2 or 3 events to get even an idea of that. Even one event would already exclude the "every 2000 years".
Because it makes a big difference. If it is "one every 200 years, or one every 2000 years", then it really is something one shouldn't worry too much about. Cities do get leveled "regularly" on that time scale, be it by war events, natural catastrophes, accidents, whatever. So if we are on that time scale, the size of the event doesn't even matter much. It's in the noise of what happens in any case.
If it is every 10 or 20 years, then things are different. Then they start to be an observable risk - still much smaller than most daily risks like driving a car, but it becomes something one should keep an eye on.
That doesn't mean one should be careless. All measures that do not entail large economic costs and can make it more difficult to have such a terrorist attack, are to be taken of course. That's a matter of balancing the costs and the benefits. So strict accounting of spend fuel, security at the reprocessing facilities, inspections, independent control mechanisms etc... are necessary.
That said, there's no reason that a terrorist bomb that explodes, say, in Europe, must come from material that came from a factory in Europe. (fill in "US", or whatever place).
But taking drastic measures, like banning nuclear power, or banning fuel reprocessing (and hence dividing the potential energy use of nuclear fuel by 100) for a potentiality of which there has never even been a single manifestation, sounds somewhat crazy to me. It might be like the guy who puts red poles around his house, against attacks of the giraffes. When his neighbor points out that there aren't any giraffes in the country, even less groups of giraffes that attack houses, he says that one should thank him, it's thanks to his red poles. They're more efficient than he thought!
It might be that a nuclear terrorist bomb is in fact impossible, simply because of the structure of terrorism. Maybe it is not their thing. Maybe the work and organization needed to make a nuclear weapon is incompatible with the way most organized terrorist organizations work. Maybe it doesn't fit in their politics. In that case, the effective probability of having a nuclear terrorist attack is in fact zero.
As someone pointed out, a much easier strike by a terrorist group would be blowing up a dam. It has never happened either. Happily, nobody stopped building dams in the beginning of the 20th century because of this potentiality.
