The four reactors at Fukushima Daiichi damaged in the March 11 earthquake and tsunami belong to the Tokyo Electric Power Co. (TEPCO). Not surprisingly, TEPCO’s financial situation has become perilous. In announcing its annual results, the company recorded a direct loss of more than ¥1 trillion ($12 billion). This does not include several significant additional costs. Several of TEPCO’s other nuclear reactors remain down in part due to the regulatory and political reaction to the accident. The company will need to purchase replacement power both for the four lost reactors, and for the add-on shut down reactors. Most importantly, the ¥1 trillion does not include the liability for the various types of damage caused by the accident. Estimates for the size of this liability range from 2-11 times the ¥1 trillion loss already booked. Following the accident, TEPCO’s stock quickly lost ¾ of its market value, or more than ¥2.7 trillion ($32 billion). It is widely believed that the liabilities from the accident make TEPCO insolvent, but for a government decision to support the company financially.
In a nice (but paywalled) article in Energy Risk, Stephen Maloney, discusses the lessons from the TEPCO disaster for other nuclear companies and for the industry as a whole. Among those lessons, perhaps the most important one is that
The Fukushima incident clearly demonstrates that a reactor accident can be the single largest financial risk facing the chief risk officer (CRO) of a nuclear generating company, far outpacing market, credit, and operational risks, combined.
The US industry confronted this fact after the Three Mile Island accident, and implemented important changes to industry standards. In particular, the Institute of Nuclear Power Operations (INPO) was created as a best practices organization which has significant disciplinary power tied to the industry liability insurance system. Undoubtedly, how well the US system is operating is open to debate, but it is certainly instructive to examine that history.
Maloney also advertises Nassim Taleb’s assertion that we systematically underestimate the likelihood of disasters like the Fukushima accident because we ignore the estimation errors in our predictions of the likelihood of key events–“the future has thicker tails than the past.” Therefore, probabilistic risk assessments (PRAs) which guide nuclear safety regulations systematically underestimate the size of tail events that have huge consequences. Making that assertion is easy, and probably wise. Figuring out what to do about it is a more challenging task.