On Rick Katz's A, B or C:

I won't letter ...

1. Older plants were designed with the expectation (perhaps with a wink and a nod) that they would see 40 years of active life; the damaged Fukushima plants were that generation. So if the last previous large tsunami was over 1,000 years earlier -- and records were from a different part of the coast -- then do you really build that into your scenarios? Ditto an M9.0 earthquake, when there were none of that size in Japan's recent history (though in fact the shock itself did not cause particular problems).

Building a house in a 100-year floodplain is clearly a bad idea, since there's a high chance of a flood within the 50-year time horizon that builders use in the US, but how about a 500 year floodplain? That's older than the northern European settlement. I think it's not at all unusual to find whole towns that are in "potential"
floodplains for which there is no written "historical"
record of a flood.

2. The disaster was because several backup systems failed, if it was only the tsunami then there wouldn't have been a problem. It was also that the plant was knocked off the grid, and not just because of a downed line, but multiple landslides and downed lines that meant they needed 11 days to reconnect. I think that was also outside the engineering scenarios. Hindsight is tricky, but disaster analysis means that the next disaster will be due to a different string (not
individual) series of unanticipated events. We don't make the same mistakes, but neither are we prescient.
[But when it comes to coal mining, we do see skimping on safety, repeatedly -- there are too many mines and it is too easy to delay and (if necessary) pay off inspectors. Yet as incompetent as the TEPCO management may be, short of making different design decisions in the mid-1960s, it doesn't look as though a greater degree of competence could have made a difference:
there was simply no way to cool the reactors, given the chain of events, and very limited ability to prevent the hydrogen explosions.]

3. I would curious as to the damage assessments to the newer plants at Fukushima. They happened not to be operating....so this is purely curiosity. My vague recollection is that they had different backup systems.
We should remember too that the Onagawa plant was hit harder both in terms of the initial shock and the subsequent tsunami but did not experience problems. If I read this correctly, each new plant was better, but older plants (especially much older plants) near the end of their operating lives were not retrofitted.
Again, if a plant is soon to be decommissioned, how much retrofitting do you want to do, particularly if it leads to shutdowns in the interim?

Now this is rehashing some ground, but I think it is a useful reminder given Rick's note on the psychology of decision-making under uncertainty (and particularly in the face of long odds), and on the "simple"
cost/benefit and other tradeoffs that underlie engineering decisions, since budgets are always finite.

mike smitka=