The US and Japan take notably different approaches.
A short article in the American Geophysical Union’s journal EOS looks at how Japan is responding to the revelations that are coming out of Fukushima, and it compares that response to the regulatory situation in the US. The most obvious response has been the shutdown and safety review that took all of Japan’s reactors offline; for most of them, the reviews are still in progress.
There was a legislative response as well. The agency responsible for the plants, which has been accused of suffering from regulatory capture, was reorganized. In addition, new safety rules were put in place and are helping govern the restart. The EOS paper looks at the experience with the Tsuruga Nuclear Power Plant, which would be the oldest operational facility in Japan if given the green light to restart.
Tsuruga is situated on the coast in a steep valley that faces a narrow bay. A high ridge separates it from the Sea of Japan, which is the opposite coast from the faults most likely to generate a large tsunami. That’s the good news. The bad news is that the valley and bay are formed by a fault that extends north-south for at least 10 km (its full extent is unmapped). The foundations of one of the reactors are less than 200 m from the main fault, and a smaller side fault extends directly underneath reactor 2.
When Tsuruga was built in the 1970s, these faults were considered inactive. But further research has shown that the main fault (Urasoko) has shifted a number of times over the past 120,000 years; current maps of Japanese faults call Urasoko either “active” or “possibly active.” That bit of knowledge has triggered an extensive safety review.
To understand the risk, workers cut trenches across both faults on the site and used changes in the sediment to date earthquakes. (For example, the surface sediment deposits should be uninterrupted until the layers deposited just before the last major earthquake.) These trenches showed that the smaller faults closest to the rectors had been inactive for at least 120,000 years and were not triggered when the Urasoko fault shifted. The same work built a careful chronology of events at the main Urasoko fault, which can then be incorporated into the safety assessment.
The authors compare that assessment with how things have worked in the US and appear to suggest that a simple determination that a fault is “active” or “inactive” probably isn’t the best way to go about things. They compare that process to the one under which California’s Diablo Canyon power plant was licensed, which required an ongoing assessment of seismic risks. Plant operators have thus established a Long Term Seismic Program that updates safety considerations based on current research and models. In an example given in the article, the risk analysis data was updated significantly in the wake of 1989’s Loma Prieta earthquake near Santa Cruz, CA.
Read more at What to do with a nuclear plant on a fault line
Related article: 断層上にある原発、日米のリスク評価の違い via Wired