OECD/NEA: Comparing Nuclear Accident Risks with Those from Other Energy Sources

[As it is relevant to current radiation hysteria, I bumped the time-stamp on this Sept, 2010 post to bring it to the top — Ed.]

The Nuclear Energy Agency of the OECD published in 2010 Comparing Nuclear Accident Risks with Those from Other Energy Sources (PDF). The report is aimed at politicians (“policy makers”) with the goal of educating them on how incredibly safe nuclear power is compared to alternative energy sources. On the damages side, the report only examines severe accidents for all non-nuclear sources, comparing these to modeled latent fatalities from nuclear operations (because there are no real-world deaths in OECD experience). I’ll cover the full life cycle comparative health effects in a separate post – to include e.g., long term health damage from fossil fuel burning.

This OECD/NEA report is the most complete and up-to-date (that I know of) that addresses the anti-nuclear fear-mongering of the “China Syndrome”. Unfortunately, there is no two-minute “elevator pitch” that explains accurately why nuclear power is so safe — it is a complex subject.

From the Executive Summary:

Many countries are reconsidering the role of nuclear energy in their energy mix, as a means to alleviate the concerns over climate change, security of energy supply and the price and price volatility of fossil fuels. However, nuclear energy remains a contentious technology in some political circles and in the minds of many members of the public.

One of the issues that causes concern is that of the safety of nuclear power plants. However a rational choice of energy sources should involve an even handed comparison of the risks presented by the various energy chains available. There is little real value in rejecting one source if that which replaces it presents even greater hazards. The purpose of this document is to provide energy policy makers with quality data and information that will enable an understanding of how accident risks are managed in nuclear plants and also provide a rational analysis of the relative risks presented by the various major energy chains used for the production of electricity.

The report starts by considering a major component of the design philosophy adopted in nuclear reactors, explaining the concept of defence in depth. Defence in depth is implemented through the combination of consecutive and independent levels of protection that would all have to fail before harmful effects could be caused to people or to the environment. If one level of protection or barrier were to fail, the subsequent level or barrier is still available to provide protection.

Next, the report discusses the important issue of the safety culture of operating organisations in maintaining a low level of risk. The quality of an operator’s safety culture cannot be measured directly. The international community has developed a number of indicators which are tracked and compared to allow a judgement of the performance trends in nuclear power plants. The report presents data for the indicators of unplanned automatic trip rate, worker collective and worker individual radiation exposure. The data shows that there have been very positive trends in all of these indicators over the last two decades in all the regions of the world and in all types of reactors.

The risk associated with the operation of a nuclear plant is that radioactivity is released to the environment, resulting in exposure by and health effects to the population. Since significant releases of activity are extremely rare, reliance on statistics of events is not possible. The report uses the analytical technique of probabilistic safety assessment (PSA) by which potential accidents, their probabilities of occurrence and their consequences can be assessed. It is common to look at the outcomes in terms of the theoretical probabilities of core damage (an accident in which the fuel cladding is ruptured, for example by overheating and melting) and the more severe events in which significant radioactivity breaches the primary circuit and the secondary containment and is released to the environment. These two measures are termed the theoretical core damage frequency (CDF) and the theoretical large release frequency (LRF). While these are not actual statistics on accident rates, they serve to illustrate the trends.

The report looks at the “as originally designed” CDFs and LRFs over the evolution of reactor designs from Generation I to Generation II and on to Generation III/III+. It shows that, over this evolution, there has been a very significant reduction in both CDF and LRF. While this clearly indicates that modern designs are extremely safe, it is important to recognise that earlier designs have also been back-fitted with safety improvements, often evaluated using the techniques of PSA. If the world turns to nuclear energy in large measure to alleviate the energy issues it confronts, it can be expected that this evolution in CDF and LRF reduction will continue and it is desirable that it does so.

The report then looks at real accident data from full energy chains, using an impressive collection of data assembled by the Paul Scherrer Institute (PSI) in Switzerland. Using this severe accident data (events that have resulted in 5 or more prompt fatalities that have actually occurred from 1969 onwards) it compares the outcomes with the theoretical accident outcomes from PSA analysis (since there are no real nuclear accident data from OECD countries and only one data point from non-OECD countries). This shows that, contrary to the expectation of many people, nuclear power generation presents a very low risk in comparison to the use of fossil fuels.

Latent fatality rates for modern nuclear plants can only be assessed using PSA. PSA studies available for the Mühleberg Swiss nuclear power plant show there is about a 1 in 1 million-year probability of an accident causing more than 2 000 latent fatalities. For OECD countries, frequency-consequence curves show that the risk of a nuclear accident with more than 100 latent fatalities is a factor of ten or more lower than the risk of an accident with 100 immediate fatalities from coal, oil, natural gas or hydro energy chains, and almost a factor of one thousand lower than the risk from LPG.

I also recommend this short summary of the study findings on Chernobyl, which saves me writing it. Excerpt (in the following I would add the qualifier “thoroughly discredited LNT hypothesis”):

Considering the long-lasting health concerns in areas affected by Chernobyl, the OECD quoted a range of 9000-33,000 eventual deaths from effects of Chernobyl over the next 70 years. These figures come from reports by the European Commission, World Health Organisation, International Atomic Energy Agency as well as Russian authorities, and they depend on the land area considered and how the effects of low radiation dose are understood.   

The report heavily qualifies these figures by pointing out that they are based on the contentious “linear dose response relationship with no threshold” (LNT) hypothesis, and that if the same logic is applied to the background radiation normally experienced by all of us, those figures would be insignificant. “For the 70 years over which the above fatality figures were calculated for the accident, the collective dose from natural background would be 910,000,000 person-Sieverts (assuming a constant population), some 1500 times larger, therefore theoretically causing 1500 times as many fatalities (some 50 million) due to exposure to natural background radiation. However there is no way to definitely confirm these figures for Chernobyl, since death rates from all cancers are very much higher.”

The report also says that “extrapolating these (Chernobyl) nuclear risks to current OECD countries is not appropriate because OECD plants use other, safer technologies that are operated under a stricter regime than was in force in Ukraine at the time of the Chernobyl accident.”