Assumptions and accidents

If we want to continue with nuclear energy, we should think the risks of nuclear power through to their logical conclusion.

Satellite image of Fukushima Daiichi nuclear plant after the recent earthquake (Photo: Reuters/Digital Globe/Handout)
The German acronym ‘GAU’ stands for ‘größter anzunehmender Unfall’, i.e. the greatest accident that has to be assumed. Since, however, there are also accidents that are even worse, the Germans have also coined the term ‘Super-GAU’, which refers to an accident that goes beyond our assumptions – to an accident such as the one at the Fukushima nuclear power plant taking place in Japan today. The two terms reveal human hubris and the way that our assumptions stop us pursuing risk calculations through to their logical and necessary conclusion.

Though ‘GAU’ stands for the worst accident that has to be assumed, it would be more appropriate to talk in terms of the worst accident that has been assumed. It refers to the type of accident that power companies and governments think of as highly unlikely but just about feasible, weighing also economic and political considerations. Power plants have thus to be designed in a way that they withstand such an accident – hence the English equivalent term ‘design-basis accident’. The Japanese reactors, for instance, were designed to cope with an earthquake of magnitude 8.2, a tsunami and the subsequent failure of the cooling system. With adequate provision, things should not get that far, but in the event that the ‘GAU’-scenario comes to pass, there are backup systems such as the diesel generators at Fukushima on standby for emergency cooling.

So when the ‘GAU’ fails to materialise, this is not because it couldn’t but because the security precautions were effective in preventing it.

This experience of security is deceptive, however. It lulls power companies and governments into the belief that no accident in excess of the assumptions of a ‘design-basis accident’ can occur. But every ’Super-GAU’ or ‘beyond design-basis accident’ – e.g. Chernobyl, Challenger and, most recently, the Macondo oil platform in the Gulf of Mexico – shows once again painfully that accidents may well go beyond our assumptions. In Japan, the reactors were not designed to cope with a magnitude 9.0 tremor. They were designed to cope with a failure of the cooling system but not with one of the emergency cooling system, because it seemed inconceivable.

What is today’s non-assumed ‘Super-GAU’ is tomorrow’s to-be-assumed ‘GAU’. It is when this scenario has actually occurred that we at last have to concede its feasibility and that denial of its unlikely possibility ceases to be an option. Subsequently, power companies and governments factor it into their risk calculation and take counteractive measures, defining thus a new ‘design-basis accident’. That’s why we believed that there could be no second Chernobyl. Indeed, we were surprised at the time by the inconceivable, but we have been mindful of it ever since and have ratcheted up our safety precautions accordingly. That is why Lithuania and Bulgaria had to take their seemingly unsafe Soviet-type nuclear power plants off the grid upon accession to the EU. We, power companies and governments considered it impossible that a meltdown à la Chernobyl might ever happen in a – more modern – Western reactor.

Regarding other unlikely possibilities, however, we remain in denial. Because as rational and progressive human beings we cling to the view that we understand what we control and know what can happen … until the day comes when our assumptions are shown to be wrong.

The real question about nuclear energy was therefore never whether it could be controlled or made controllable through further technological progress, because there can never be a guarantee. The probability of another ‘Super-GAU’ is very low, but always above zero. Instead, the question should have been put in the following terms: Are we willing to take the non-excludable residual risk? There is nothing uncommon in taking a risk; we all do it in the daily traffic. A residual risk remains in spite of mandatory seat belts and helmets, and sometimes it is these safety precautions that let us take a higher risk when we think that wearing a helmet we can allow ourselves to drive faster. Nor is it light-headed to take a risk – as long as we enter into it willing to bear the consequences. As a society we increase motorised traffic despite the accident risk because we deem the advantages of a higher mobility more important than the unavoidable road casualties.

But it counts as pure negligence, if we take a risk on the assumption that the emergency case won’t happen. If we fool ourselves about the risks and don’t think the thinkable. Are we as a society, are power companies and governments ready to bear the consequences of running nuclear power plants? The very real consequences of a very unlikely but possible ‘beyond design-basis accident’? Have we taken the risk calculation process through to its logical conclusion or do we want to continue believing that another ‘Super-GAU’ is impossible?

It is better than nothing when governments in the EU and elsewhere want to re-examine all the reactors – if there is a second, third or fourth layer of emergency measures – and increase the safety precautions as appropriate. In doing so, however, they are again and vainly trying to find an answer to the question whether nuclear power could be made controllable. After Chernobyl, we ratchet up safety precautions aimed at excluding a second Chernobyl. After Fukushima, we can try to exclude a second Fukushima. But even then another ‘Super-GAU’ remains thinkable.

The real question that governments and we as a society have to respond to is whether we really are willing to bear the consequences of a nuclear ‘Super-GAU’ such as the one the Japanese are now struggling to cope with. In other words, we have to answer the question, do the advantages of nuclear power outweigh its disadvantages? Or do we believe these consequences go beyond what we are ready to bear? In the latter case we have to abandon nuclear energy before the unlikely possible event destroys not only our assumptions.

 

Who is Marcel Viëtor?

Marcel Viëtor (vietor@dgap.org) is Resident Fellow for Energy and Climate Policy at the Alfred von Oppenheim Center for European Policy Studies at the German Council on Foreign Relations (DGAP) and Associate Fellow at the stiftung neue verantwortung (snv), a Berlin-based think tank on responsible leadership.

The fallout of Fukushima

The nuclear disaster in Japan has re-ignited the debate on the desirability of nuclear energy. For a thoughtful discussion of this complex issue, we highly recommend an essay written in 2009 by Warner ten Kate and Jacques de Jong of the Clingendael International Energy Programme (CIEP), Some Policy Challenges of a Global Nuclear Renaissance.

A more specific, also highly interesting look at the failings of Japanese nuclear energy policy is given by Jan-Hein Christoffels of CIEP in this article: Earthquake Alarm - The Kashiwazaki nuclear incident and the consequences for Japan's nuclear policy.

Philippe Dumas, Manager of the European Geothermal Energy Council, pointed out to us that, for geological reasons that are all too obvious, Japan has large geothermal resources, but – unlike for example Iceland – it has never chosen to develop those. Instead, says Dumas, it chose to rely on nuclear power. See this 2010 Country Update of the World Geothermal Congress on Japan’s geothermal resources.