Nuclear meltdown: race to save reactors in Japan

Kevin,
I don’t think that this is likely to be “Chernobyl on steroids.” First of all, Chernobyl was a steam explosion, followed by a graphite fire on a reactor running at 100% or close to it when the accident happened. This generated a great deal of heat: enough to produce the convection necessary to provide enough lift to the particles to get them up high enough where they could travel significant distances.
There does not seem to be enough of a source of heat at Fukushima to generate enough lift to come even close to Chernobyl. This is assuming of course, that Units #1, #2 and #3 remain sub-critical and that the not-so-spent fuel in the “spent fuel” pool at Unit #4 also remains sub-critical. If these assumptions hold, then Fukushima will probably release only or almost only fission by-products and release of significant amounts of actinides, especially plutonium, may be avoided. One can only hope.
You may be right about the source of hydrogen in Units #1 and #3 being primarily from inside the reactor. The pool may have produced some hydrogen, but probably not enough to create the size of explosions that happened. I read something interesting about Unit #2 though. One thought is that, since it was probably the only one whose core was completely uncovered, it was also probably the only one that melted. Zircalloy is supposed to be an excellent conductor, so, even if the rods are partially submerged in water, that may be enough to avoid melting, although probably not enough to avoid damage to the rods. If one or more of the rods did melt, however, the molten metal would be so hot when it dropped into the water at the bottom of the reactor that some of the water would instantly flash to steam, perhaps with enough force to over pressurize the reactor vessel and cause loss of integrity.
My heart goes out to everyone at risk in Japan and one can only hope that a major catastrophe is averted.
I have built large industrial facilities in earthquake zones, and, in principle, I don’t see any reason why a reprocessing facility or a nuclear plant cannot or should not be built there. One just needs to be think differently. It also helps to have some experience or acquire some experience with what has happened structurally elsewhere during strong earthquakes. For example, water or other liquid storage facilities belong at or below ground, period. Different types of foundations need to be considered. For example, the Latin American tower (http://en.wikipedia.org/wiki/Torre_Latinoamericana) in Mexico City was built in the fifties with an innovative foundation and has performed extremely well in earthquakes. There is no intrinsic reason why these cannot be built correctly. Hopefully, Units #5 and #6 survive intact as close examination of those structures and how they fared in this 9.0 earthquake would be of invaluable, if not, irreplaceable value to provide further knowledge that can be used to build truly earthquake-proof plants. You can model a structure all you want, but there is simply no substitute for real data – as is so evident in climate science, for example.
P.S. I don’t think it matters where a fault line is. We cannot hope to map them all, because some of them only show up when a new earthquake hits. The “fault line” criterion does not inform the design and construction of a facility. Instead, one should simply focus on what the maximum accelerations are likely to be and use that information to design the structure. Then the only issue when an earthquake hits would be whether the structure was subjected to accelerations that exceeded the design parameters.