The Fukushima nuclear accident demonstrates that the economic costs of a severe nuclear accident can be considerable. The current cost estimates for the Fukushima accident include
• Support for accident evacuees. TEPCO (2014) estimated that, as of January 15, 2014, its compensation payments to the evacuees and businesses affected by radiological releases from the Fukushima Daiichi plant would be more than ¥5 trillion ($50 billion).1
• Offsite decontamination. Japan’s National Institute of Advanced Industrial Science and Technology estimates that decontamination in Fukushima Prefecture will cost ¥2.5-5.1 trillion (~$25-51 billion) (Mainichi, 2013).
• Onsite decommissioning. TEPCO (2014) estimates its site cleanup costs at Fukushima Daiichi at ¥2 trillion (~$20 billion).
• Replacing power from idled nuclear plants. The undamaged units at the Fukushima Daiichi plant (Units 5 and 6) will never operate again (TEPCO, 2014). In addition, all of Japan’s other nuclear power plants have been idled for about 3 years as a result of the accident. Japan’s utilities have paid an estimated ¥7.3 trillion ($73 billion) for fuel in fiscal year 2012, double the amount in FY 2010, in large part because of the need to buy liquefied natural gas to replace the power from the shutdown nuclear
1 Additional detail provided in Asahi Shimbun (2013).
• Other costs. The Institut de Radioprotection et de Sûreté Nucléaire (IRSN) estimated the cost of a Fukushima-scale accident in France (Pascucci-Cahen and Momal, 20123). The estimate included about €166 (~$215 billion) in costs for image or reputation losses, including loss of food exports, reductions in other exports, and loss of tourism. The estimated total loss of tourism in Japan is about ¥1 trillion ($10 billion) (World Travel and Tourism Council, 2011); this loss is attributable to the earthquake and tsunami as well as the nuclear accident.
The total cost of the Fukushima Daiichi accident could therefore exceed ¥20 trillion (~$200 billion).
It is instructive to compare these costs to the estimates developed by the USNRC staff for a hypothetical accident at the Peach Bottom nuclear plant in Pennsylvania. These cost estimates were used in the staff’s backfit analysis for filtered vents (Borchardt, 2012a, Enclosure 5C, Table 7, Case 2):
• A collective population dose to workers and the public out to 50 miles (accounting for reductions due to evacuation) of 0.53 million rem, which, valued at $2,000/rem, translated into damage of about $1 billion;
• A loss of the use of offsite land and property due to radioactive contamination of $1.9 billion;
• A loss of onsite value of $3.2 billion (Borchardt, 2012a, Enclosure 5C, Table 8). This includes the loss of use of an average of 1.75 nuclear power reactors at a BWR plant site (Borchardt, 2012a, Enclosure 1, p. 15, Table 1).
The total estimated costs for the hypothetical accident at the Peach Bottom plant are therefore about $6 billion.
The USNRC staff estimated that most of the offsite damage ($2.5 billion) and $1.2 billion of the onsite damage could be prevented by the installation of filtered vents. After multiplying the savings of $3.7 billion by a probability of 2 × 10−5 accidents per reactor year (i.e., one accident every 50,000 years on average) and by 17.6 years (the assumed remaining 25 years of reactor life discounted by 3 percent per year), the savings per
2 In March 2012, the Japan Atomic Industrial Forum estimated the extra cost at $40 billion/year; see http://www.world-nuclear.org/info/Country-Profiles/Countries-G-N/Japan/#.UeHlrFOzLEh.
3 The technical analysis behind this IRSN analysis was published in 2013. An English translation (“Methodology Used in IRSN Nuclear Accident Cost Estimates in France,” IRSN, PRP-CRI/SESUC/2014) was made available to the committee prior to its publication.
The cost estimates for the accident at the Fukushima Daiichi plant (~$200 billion) are about 33 times higher than the USNRC cost estimate for a hypothetical accident at the Peach Bottom plant (~$6 billion). The primary reasons for these differences are the following:
1. The relatively low USNRC estimate of costs associated with the calculated contamination of 354 km2 (140 square miles) above 15 curies/km2, which is approximately equal to the threshold that has been used for long-term evacuation in Japan (which affected about 625 km2 of land).4
2. The USNRC assumed that the operation of other nuclear power plants would not be affected, unlike the situation in Japan where virtually all nuclear power plants have been shut down.
Differences between accident costs in Japan and the United States can be expected—as can differences in accident costs for different sites in the United States. Nevertheless, the large differences noted above serve to illustrate that cost estimates—and associated backfit rule decisions—are sensitive to the assumptions made in developing those estimates.
The point of this appendix is not to critique the USNRC’s analysis—the committee did not perform an in-depth review of this analysis because it is outside the statement of task for the study. The committee offers this example to demonstrate that severe accidents such as occurred at the Fukushima Daiichi plant can have large costs and other consequences that are not considered in USNRC backfit analyses. These include national economic disruption, anxiety and depression within affected populations, and deterioration of social institutions arising from a loss of trust in governmental organizations.
The USNRC is launching a multiyear process for updating its regulatory guidance for backfit analyses. One focus of the update is to improve calculations of the economic consequences of a reactor accident, taking into account lessons learned from the accident at Fukushima Daiichi. The USNRC is also reevaluating how qualitative factors are used in the backfit analysis process.
4 IRSN (2012a, Fig. 6-24). See also Table 6-11, which shows the areas contaminated above this level outside the 20-km radius around the plant as 320 km2.