Incorrect. The area of dispersion is actually quite limited. Widespread dispersion requires release of significant amounts of stored energy. Nuclear explosions produce fallout because of this. The same occurred at Chornobil, wherein the internal energy contained in the fuel drove the material out into the environment.
Additionally, widespread disperal usually occurs when there is significant vaporization of the reacting mass. The explosive materials must reach very high temperatures so its constituent atoms have tremendously high kinetic energy. They then move about the environment, losing energy and eventually plate out or condense onto other materials, far removed from the explosion site.
An event driven by an external force simply doesn't match up on these terms. Sure, you can push materials around somewhat, but in terms of blasting things far and wide, to the tune of some of these widely speculative "doomsday" scenarios involving "dirty nukes" or whatever the kooks are calling them these days, its simply not in the cards, based on the laws of mother nature.
Before you blow me off as not knowing what I am talking about, I did some of the modeling and some experiments involving release of materials from nuclear material transport accidents back when DOE was doing some of the initial transport cask design and validation. I was also a partner on a consulting job for a utility group that was looking into risk management that involved modeling of release terms and dispersion from power plant accidents. If you have any kind of reasonably accurate meteorological model that incorporates realistic transport mechanisms, it predicts quite limited dispersion.
1) Modeling. There are lots of supercomputers constantly "modeling" the weather. That doesn't mean anyone can tell you if it's going to rain tomorrow... More seriously, some really smart people "modeled" the effects of Iraq igniting the Kuwaiti oil fields. Remember Sagan did the talk show circuit doing the chicken little thing. When the oil fields were ignited, the end effect was little more than a heck of a lot of work for the engineering people to put them out. Modeling is fun, but models aren't reality and computer models of complex things have a pretty iffy public record. I'm not saying you're wrong, I'm just gagging on your appeal to authority of running simulations.
>Unless you postulate some kind of isotopic separation process occurring, you're not going to be able to assemble an explosively-critical mass in any kind of meltdown scenario.
2) But if you have a meltdown scenario, almost by definition you are going to have, to use every day language, really, really hot stuff coming into contact with stuff that is -- comparatively -- really, really cold. As the hot stuff causes the "cold" stuff to get volatile, vaporise, the sudden pressure changes might not be an "explosion" per se, but they possibly could have similar effects to an explosion, and disperse bad stuff over a wide area.
-- KotS
I went through a Nuclear, Biological & Chemical school in the Army, worked with nuke systems, and also read the ordnance (blowing stuff up) manuals for fun. I can confirm everything you're saying, although to be honest, if I hadn't also been reading this in the library some of this would be going over my head. The Army simplifies everything so much.