Robert J. Geller: A Seismologist’s View of Nuclear Safety Issues in Japan — Part II


Robert J. Geller, Professor of Geophysics in the Graduate School of Science of the University of Tokyo

Idea Gallery is a recurring editorial series on Forum on Energy in which guest authors provide their perspective on issues affecting nuclear energy. Today’s guest author is Robert J. Geller, Professor of Geophysics (seismology) in the Graduate School of Science of the University of Tokyo.

>>Read Part I of the guest perspective piece from Dr. Geller.


In this, the second part of my article, I discuss the road forward, post-Fukushima, for nuclear power in Japan. I write both as a seismologist (here’s my publication list) and as a permanent resident of Japan who has been a faculty member at the University of Tokyo since 1984, who is fluent in Japanese, and who has made his own life here. What I write here is inseparably informed by all of these diverse perspectives.  

At the time of the 2011 Tohoku earthquake Japan’s governing party was the Democratic Party of Japan (DPJ), and Naoto Kan was the prime minister. He was succeeded as prime minister by another DPJ member, Yoshihiko Noda, in September 2011. After the general election of December 2012, the Liberal Democratic Party (LDP) and their coalition partners took power, and Shinzo Abe became prime minister. The fact that Japan has had three prime ministers and that there have been even more frequent changes of other cabinet ministers in the 29 months since the earthquake has unquestionably impeded progress in resolving issues related to nuclear safety.  

Of these three prime ministers, Kan advocated phasing out nuclear power. Noda was ambivalent, but under his premiership two reactors at Kansai Electric Power Company’s Ohi nuclear power station (NPS) were restarted (one of these reactors was recently temporarily shut down for scheduled maintenance, and the second will be shut down temporarily for scheduled maintenance later this month — Read the story in Japanese). Abe is pro-nuclear, but during the first nine months of his premiership no visible progress has been made toward restarting additional reactors. As discussed below, this delay is due in part to teething problems of Japan’s new Nuclear Regulation Authority (NRA).  

Since the Fukushima accident, Japan has managed to get by without using nuclear power, with the exception of the two reactors at Ohi. However, maintaining Japan’s electricity supply without nuclear power has been expensive because of the cost of fuel imports, has required full use of many relatively old fossil fuel plants, and may not be sustainable on a long-term basis. Thus economic and energy security considerations suggest that restarting as many as possible of Japan’s 54 nuclear power reactors is desirable. On the other hand, in the wake of the Fukushima accident, it obviously is necessary to address safety issues, especially those involving natural hazards such as earthquakes and tsunamis, before reactors are restarted.

Earthquake and tsunami hazards

An NPS (or indeed, any other structure) must be designed to withstand some specified levels of hazard (earthquake shaking, tsunami height, and so on), called the design basis. Until someone decides on particular numbers for the design basis, the NPS designers can’t decide on the amounts of steel, concrete, etc., to be used in the NPS.

Good engineering is conservative, so typically some safety margin over the bare minimum is incorporated. Nevertheless, if a natural disaster exceeds the design basis by a large margin the possibility of failure of the structure exists. The possibility of failure also exists for events below the design basis if there are flaws in design or construction. Thus good engineering practice for NPSs (“defense in depth”) includes tiers of precautionary measures so that even if one set of defenses fails, an accident will still be prevented if possible, or at least its consequences will be minimized.

In an engineer’s ideal world, he or she would be able to consult a qualified seismologist; obtain reliable and accurate information on the “maximum earthquake” and “maximum tsunami” expected at a particular NPS site; and then ensure safety by designing the NPS to withstand those hazards. Unfortunately, that’s not how the earth works.

About 100 years ago, when the systematic study of earthquakes began, it was widely assumed, mostly due to lack of data, that earthquakes in each region recurred more or less cyclically. However, seismology has progressed in the past 100 years, and we now know, thanks in large part to rigorous statistical studies by Drs. Yan Kagan and David Jackson of UCLA, that earthquake activity is not a simple cyclical process.

Research is ongoing, but, as a result of work done at UCLA and many other institutions, we now have the following picture of earthquake activity. The size of earthquakes is quantified by the magnitude, originally defined by Charles Richter, but now measured using a formula devised by Dr. Hiroo Kanamori of Caltech. Each increase of 1.0 in magnitude means the energy released by the earthquake goes up by a factor of 30. For small earthquakes the number of earthquakes in a given space-time box decreases on average by a factor of ten when the magnitude increases by 1.0. However, as the magnitude increases above some “corner magnitude,” which probably varies from region to region, but is probably typically about M8.0, the number of earthquakes starts to fall off much more rapidly, so there are far fewer than 10 percent as many M9 earthquakes as M8 earthquakes. The largest earthquake recorded to date in the 100-year-or-so history of instrumental seismology is the M9.5 earthquake of 1960 off the coast of Chile, but the UCLA researchers recently estimated that there might be a magnitude 10 earthquake off the coast of Tohoku once every ten thousand years or so.

In summary, above M8.0 or so we reach a point beyond which the probability of larger earthquakes decreases exponentially as the magnitude increases further, but there is no such thing as a sharply defined “maximum earthquake” or a “maximum tsunami.” That’s the way the earth works, and we have to deal with the earth as it is, not as we’d like it to be.

Regaining Public Trust

As discussed in Part I, before the Fukushima Daiichi NPS accident, the nuclear industry and government just told the public that NPS operation was “safe.” Depending on one’s point of view this was either a lie or an oversimplification—but it certainly wasn’t the whole, undiluted truth.

The first step toward regaining public trust should be a public statement by the prime minister acknowledging that nuclear power is not “safe.” He should clearly state that everything in life has some risks, and that his government will abandon the “safe/unsafe” dichotomy in favor of a policy of objective and quantitative evaluation of risks.

Before Fukushima, it appears that the government and nuclear industry believed that the public was unable to accept the idea of living with risk. But, I think that if the concepts are properly explained, the public will accept an honest discussion of risks. I also think that after Fukushima the public will no longer accept oversimplified assurances of perfect safety.

The division of responsibility between the government and the nuclear regulators should be recalibrated. The government must of necessity rely on expert advice in establishing safety regulations for NPSs, but it should be made clear that the regulator’s job is to evaluate risks, and that the final decision on what risks are acceptable is the responsibility of the government. Based on expert advice, the prime minister should honestly and accurately summarize the risks entailed in restarting nuclear power stations in Japan, and explain why (if the government so decides) he views the risks to be acceptable.

The government and the nuclear industry should publicly provide detailed information on risks and countermeasures on their websites, at visitor centers and in other easily accessible locations. If the industry followed the prime minister by publicly acknowledging risks and discussing them, then the pre-Fukushima talking heads could be sidelined and replaced by actual engineers, both from Japan and from overseas, many of whom, if given the chance, are capable of clearly and comprehensibly informing the public. Open discussion of risks would allow open preparation for various possible accident scenarios, which could contribute greatly to mitigating the damage in the event of an accident.

Fukushima—Managing the Aftermath

Since March 2011 it seems that news from Fukushima has repeatedly been delayed and has often only been released piecemeal. An early example was the failure of the government to issue timely reports on the direction of windborne radioactivity, and the most recent example is the belated reports of large-scale leakage of highly radioactive water, some of which is apparently leaking into the sea. In my opinion, public trust in the government and Tokyo Electric Power Co. (TEPCO), the operator of the Fukushima Daiichi NPS, would be enhanced in the long run by fully and promptly announcing all news, even if the news is bad.

TEPCO’s known financial assets appear to be dwarfed by reasonable estimates of its liabilities—including site cleanup and damage compensation—and it appears unable to survive as a normally solvent private company. In view of the importance of electric power to the economy, there appears to be no practical alternative to nationalizing TEPCO, and eventually refloating it after the injection of public funds to cope with the direct and indirect costs of the accident. However, for the moment, in the absence of government action, TEPCO is staggering on as a kind of zombie company. As a result of TEPCO’s strained finances, site cleanup efforts have been slow, and final settlement of damages to people who will be forced to permanently relocate and/or who lost their businesses or livelihoods has been indefinitely delayed.

Officials of local and prefectural governments in regions that contain NPS sites, as well as residents of such localities, are undoubtedly watching events at Fukushima with great interest. Even people who are favorably disposed to nuclear power in general may be concerned about what might befall them in the event of a nuclear accident in their backyard. Failure by the national government to compensate residents of Fukushima for the damages they incurred due to the accident will dampen the willingness of other localities to agree to restarting nuclear reactors.

Nuclear Regulation

After Fukushima it was generally agreed that the previous regulators had been insufficiently independent of the nuclear industry, and the new NRA was created. The start-up of the NRA has not been altogether smooth. The NRA must avoid being overly influenced by the nuclear industry, but communication with the industry is also essential, and the NRA may not yet have found the best balance. The NRA should probably also be more proactive in learning from the experience of regulatory agencies in other countries.

Great attention is being paid by the NRA to the issue of faults near nuclear power stations (a fault is simply a discontinuous surface in the earth where slip has occurred, usually due to a past earthquake). The NRA has apparently decided that if a fault near a nuclear power station has slipped in the past 400,000 years, this is grounds for rejecting an application for permission to restart the NPS. The NRA might benefit from the example of the Diablo Canyon NPS in the United States, which was allowed to continue to operate after quantitative evaluation by the regulators of the risk posed by a nearby fault.

It appears to me that many NPSs on the Sea of Japan may not have sufficiently strong tsunami defenses, and that this issue should be studied on a priority basis. The NRA, government, and plant operators do not appear to be paying sufficient attention to this vital issue.

Fukushima is on Japan’s Pacific Coast, but there is also a significant tsunami risk on the coast of the Sea of Japan, where many of Japan’s NPSs are located. During the past 30 years large tsunamis struck the coast of Akita in 1983 and Okushiri Island, off Hokkaido, in 1993. It’s unlikely that these are the largest tsunamis to have hit the Sea of Japan coastline in the last, say, 10,000 years, so it seems prudent to prepare for considerably larger tsunamis. It’s also hard to believe that Okushiri Island and Akita are peculiarly dangerous, and that other sites on the Sea of Japan coast have much lower tsunami risk. In fact, it appears to be reasonably well established that a large tsunami struck Masuda, in Shimane prefecture, in the year 1026. Prudence thus calls for any nuclear power station anywhere on the Sea of Japan coast to be able to withstand a tsunami considerably larger than that which hit Okushiri Island in 1993.

Without minimizing faulting risk, it appears to me that the NRA is placing too much weight on the risk of relatively small earthquakes that might have a low average rate of occurrence (say once every 400,000 years) while under-emphasizing the risk due to tsunamis that might occur 100 times more frequently (say, once every 4,000 years), and, furthermore, which might be much more damaging to an NPS when they do occur.

Living with Risk

Up to this point I’ve only discussed nuclear risks. But we shouldn’t forget that fossil fuel electric power generation also has big risks. Many power plants are built on landfill, which might liquefy in the event of a large earthquake. In the event of a major earthquake and/or tsunami, the contents of oil tanks might well spill into Tokyo Bay or similar coastal bays, with massive fires and environmental damage. And a war in the Middle East could easily disrupt supplies of crude oil, which would impact electrical power availability within a short time. So we shouldn’t focus only on the risks of nuclear energy while ignoring the existence of other risks.

In closing, let me emphasize the gravity of the issues that are at stake. It behooves pro-nuclear advocates to understand and acknowledge that there is no way to ensure “absolute safety”—the best that can be done is to minimize risks. And the consequences of nuclear accidents are all too apparent—just look at Fukushima. On the other hand, while I’m in favor of further development of renewable energy sources, it appears to me that the numbers for immediately going completely non-nuclear just don’t add up.

Let’s accept that we’re faced with finding the least bad solution to Japan’s energy problem—there are no truly good solutions. What should we do?

Rather than simply letting each utility company apply for permits to restart reactors of its own choosing, the NRA and government should proactively rank plants from safest to riskiest, and only license restarting the minimum number of reactors needed to ensure a stable energy supply.

And finally, when considering the risks posed by various natural hazards, we shouldn’t ever forget that “it’s the tsunami, stupid.”


Part I of Dr. Geller’s Idea Gallery series may be found here