You are here

New studies: How safe is nuclear power?

Nuclear Monitor Issue: 

There are two broad methods of assessing the risks of nuclear power reactors. The nuclear industry calculates the probabilities of accident scenarios but these 'probabilistic risk assessments' are flawed and consistently underestimate the true risks, as discussed in Nuclear Monitor #803.1

The second method of assessing reactor risks is to analyze the historical record. One such study, by Thomas Rose and Trevor Sweeting, has recently been published in the Bulletin of the Atomic Scientists. Rose and Sweeting analyze all past core-melt accidents and estimate a failure rate of 1 per 3704 reactor–years.2

The authors state:

"By our calculations, the overall probability of a core-melt accident in the next decade, in a world with 443 reactors, is almost 70%. (Because of statistical uncertainty, however, the probability could range from about 28% to roughly 95%.) The United States, with 104 reactors, has about a 50% probability of experiencing one core-melt accident within the next 25 years."

The authors also analyzed the role that learning from past accidents can play over time, using a much larger database of accidents and not just core-melt accidents, and conclude that few or no learning effects are in evidence. In their words, their statistical analysis finds "a probability for a (minor or major) accident in a nuclear power plant of about 1 in 1000 reactor years and shows no evidence of a learning effect."

Their findings come with caveats. Information is hard to come by, partly because the International Atomic Energy Agency does not publish a full list of International Nuclear Event Scale-rated events. Choices necessarily made by scholars tackling these issues greatly affect the conclusions. For example Rose and Sweeting exclude core-melt accidents in research reactors, they exclude the Windscale / UK 1957 fire on the grounds that it involved a military reactor, and they count Fukushima as one core-melt accident instead of three.

Rose and Sweeting conclude with a parting shot at the IAEA for its indefensible refusal to release data it has at its disposal:

"In conclusion, the number of core-melt accidents that can be expected over time in nuclear power stations is larger than previously expected. To assess the risk of similar events occurring in the future, it is necessary to determine whether nuclear power operators learn from their experiences. Our work shows that it is possible to investigate such learning effects through statistical analysis. Until the IAEA makes the relevant data available, however, the full story of accident probability and learning effects will remain untold."

Scientists for Global Responsibility

A somewhat similar analysis by Spencer Wheatley, Benjamin Sovacool and Didier Sornette has been published by Scientists for Global Responsibility.3 The authors compiled a dataset of 184 events from 1950 to 2014 that resulted in losses of US$20 million (€18m) or more (inflation-adjusted). One of their conclusions is more positive than Rose and Sweeting: they find that the frequency of accidents dropped substantially after Three Mile Island (TMI) and Chernobyl, and has remained relatively constant since.

That is no reasons for complacency as the authors go on to explain:

"This is good news, but not an adequate improvement: the post-TMI distribution is so heavy tailed that the expected severity is mathematically infinite. This is reflected by the fact that the severity of Fukushima is larger than the sum of all remaining events. This point cannot be emphasized enough, as it implies that, if one wants to reduce the total risk level, one needs to effectively exclude the possibility of the most extreme events. Put simply, we need to move to a situation where major nuclear accidents are virtually impossible."

On the basis of their analysis the authors estimate that:

  • one event per year causing damage in excess of US$20 million should be expected.
  • there is at least a 50% probability of a Chernobyl-type event (causing about US$32 billion (€28.7b) in damage costs) happening in the next 30-60 years.
  • there is at least a 50% probability of a Fukushima-type event (US$170 billion, €153b) happening in the next 65-150 years.

They state that while their estimates are highly uncertain, they are much larger than what industry estimates would suggest.

U.S. safety regime flawed

"I am confident that the legacy of Fukushima Daiichi will be a sharper focus on nuclear safety everywhere," said IAEA Director General Yukiya Amano in a March 10 media release. "There is widespread recognition that everything humanly possible must be done to ensure that no such accident ever happens again."4

But the reality doesn't match the rhetoric and the situation in the U.S. provides one example. The Union of Concerned Scientists (UCS) has released a report on the failure of the U.S. nuclear power industry to adequately respond to safety flaws in the five years since Fukushima, as well as the failures of the Nuclear Regulatory Commission (NRC).5

After Fukushima, the NRC set up a task force to analyze what happened at Fukushima and assess how to make U.S. reactors safer. In July 2011, the task force offered a dozen recommendations to help safeguard U.S. nuclear plants in the event of a Fukushima-scale accident. Unfortunately, the NRC has since rejected or significantly weakened many of those recommendations and has yet to fully implement the reforms it did adopt. The UCS report also finds that the NRC abdicated its responsibility as the nation's nuclear watchdog by allowing the industry to routinely rely on voluntary guidelines, which are, by their very nature, unenforceable.

Among many other problems, the NRC decided to continue to allow plant owners to develop their own voluntary plans for managing a core-melt accident, rejecting a task force recommendation to require them to do so. If plans are voluntary, the NRC has no authority to review them or issue citations when they are deficient.

"Once again, the NRC is ignoring a key lesson of the Fukushima accident: Emergency plans are not worth the paper they are printed on unless they are rigorously developed, maintained, periodically tested, and subject to NRC inspection and enforcement," said Edwin Lyman from the UCS. "When it comes to many critical safety measures, the NRC is allowing the industry to regulate itself."

The UCS recommends a revised regulatory framework; expedition of transfer of spent fuel to dry casks; increased emergency planning zone sizes (beyond the current 10-mile radius); increased NRC oversight of operator guidelines instead of voluntary guidelines that are not subject to NRC enforcement; and validation of FLEX strategies that aim to make emergency equipment readily available to reactors during extreme events.


1. 'Nuclear accidents and risk assessments', 7 May 2015, Nuclear Monitor #803,

2. Thomas Rose & Trevor Sweeting, March 2016, 'How safe is nuclear power? A statistical study suggests less than expected', Bulletin of the Atomic Scientists, Volume 72, Issue 2,

3. Spencer Wheatley, Benjamin Sovacool and Didier Sornette, March 2016, 'Statistically assessing of the risks of commercial nuclear energy', Scientists for Global Responsibility Newsletter no.44,

4. Nicole Jawerth, 10 March 2016, 'Five Years After Fukushima: Making Nuclear Power Safer',

5. Union of Concerned Scientists, March 2016, 'Preventing an American Fukushima: Limited Progress Five Years after Japan's Nuclear Power Plant Disaster',

See also: Elliott Negin, 10 March 2016, 'Five Years After Fukushima, U.S. Nuclear Safety Upgrades Lagging',

UCS, March 2016, 'Preventing an American Fukushima (2016)',

Edwin Lyman et al., 2014, 'Fukushima: The Story of a Nuclear Disaster',