The Tohoku Earthquake and the future of nuclear power

Both the Mercury-News and the LA Times had articles today on whether the problems of Japanese nuclear plants after the Great Tohoku Quake of 2011 (now upgraded to a 9.0 on the Richter scale) could happen in California. This of course is part of the renewed stirring of the nuclear controversy in the light of this tragic quake.

The Merc story was disappointing. After noting that two plants provide 4.7 GW (12%) of California’s electricity, it quoted PG&E (operator of Diablo Canyon) and Southern California Edison (operator of San Onofre) saying predictable things and opponents saying predictable things. It also (as is want around here) gave undue credence to fringe critics rather than actual experts.

Perhaps the most interesting thing in both articles is the discussion of the seismic fault under Diablo Canyon that was discovered after the plant was built. In particular, the Merc quoted the hometown state senator, Sam Blakeslee, who has a Ph.D. from UCSB and various published papers on seismic issues. It quoted Dr. Blakeslee as asking experts to study fully the safety implications of the new fault.

On the other hand, the LA Times quoted actual independent experts (i.e. university scientists) saying that a 9.0 is not very likely to happen near either plant, with “low 7s” being the largest quake expected at either plant. Big tsunamis won’t happen here either because we don’t have offshore subduction zones.

However, as an engineer, I found troubling two questions that were not directly addressed.

The Great Tohoku Quakeis larger than any in Japan’s recorded history. This reminds me of Katrina, which was a large hurricane than was anticipated — after the fact, the Army Corps of Engineers called it a “400 year” storm, but the city had planned for a “100 year” storm.

The largest earthquake in California’s history is the Fort Tejon quake of 1857 (magnitude 7.9). However, the largest earthquake in US history was the 1964 Alaska quake (9.2) which also caused a tidal wave in California, killing 11.

When it comes to record high and low temperatures, record wind or rain, record earthquake magnitude — these are always a first. If every city plans for a 100 year storm, some will not see that 100 year storm over a 100 years, and others will see the 200 or 400 or 1000 year storm. (And, of course, many cities are over 100 years old.)

This suggests to me that planning for earthquakes — at least when there’s a high safety implication — should be planning for more extreme events. If NorCal doesn’t have a 200 year quake in the 21st century, perhaps SoCal will.

Estimating the size of the largest quake is not a policy — or political question — but a scientific question. Of course, science is so politicized nowadays (particularly due to the impact of groupthink on access to funding) that a purely scientific evaluation may be impossible.

Finally, the biggest lesson of TEPCO plant in Fukushuima — the one that will cause power engineers to rip their plans and start over — is that what happens outside the dome is as important as what happens inside the dome. The engineering failure was not nuclear or structural, but in systems design.

The 40-year-old Japanese reactor containment vessel did its job, holding up to the largest earthquake ever. However, the emergency cooling plans depended on the availability of power from outside the plant, and the infrastructure did not survive the quake well enough to provide power for the cooling pumps. Apparently the backup diesel generators worked, but did not survive the tsunami.

The LA Times article talks about gravity fed emergency cooling reservoirs located onsite at Diablo Canyon and San Onofre. Will those reservoirs (and their piping) survive a direct 8.0 earthquake? Will that onetime supply of water be enough to keep the reactor cooled if there is no electric power for 7 days? 14 days?

Again, these are engineering problems, not political problems — Californians should hope that PG&E and SCE will share their revised emergency plans after the lessons of Fukushima have been fully studied.