George Monbiot had an interesting take on this. If Fukishima is obsolete/inferior design/construction, and a Richter 9 EQ is as bad as it gets, this is the worst case scenario for natural disaster at a nuke. If this is as bad as it is likely to get, vulnerabiity to natural disaster is no reason not to support nuke development. In fact the opposite is true: it has a superior safety record vis a vis natural forces as compared with any other energy source.
Monbiot's argument, not mine, but not easy to refute, if you consider that Chernobyl and TMI were both pilot error events.
