Nuclear reactor safety in the post-Fukushima world

Since the Fukushima-Daiichi nuclear power plant disaster in 2011, there have been calls for increasing safety levels in nuclear plants worldwide. The incident in Japan was the fourth significant accident in the 55-year history of nuclear reactor operation.

The first occurred in 1957 at the Windscale reactor in the UK. Two decades later, in 1979, a reactor at Three Mile Island in the USA was severely damaged, though radioactive material releases were slight. The third incident is well-known: the 1986 disaster at Chernobyl, Ukraine, where the destruction of a reactor by steam explosion, fire and core disruption had significant health and environmental consequences, mainly due to fission product release and dispersion, as well as a human death toll at the site itself.

But throughout the nuclear age, countries and international bodies have, of course, been developing a variety of approaches and systems to promote safety and minimize the risk of accidents. In the UK and US, for example, nuclear reactor safety has, since the latter part of the twentieth century, been based on a comprehensive risk assessment approach in which experts, at the design stage, identify what could go wrong—based on a detailed knowledge of the components, materials, energy flows and core neutronics.

The experts then identify the various paths that an accident might follow, taking into account the plant design and its safety features such as emergency shutdown control rods, emergency core cooling and pressure vessel strength. They also calculate probabilities for each scenario (how often in ten thousand years a given accident might happen). For each hypothetical accident, they calculate the likely release of radioactive materials and model what the consequences would be for human life, health and the environment. Once all the above is complete, the so-called ‘envelope of risks’ for the planned reactor can be examined. If any part of this represents a risk above that which is judged to be tolerable, improvements can be made.

A landmark event in the UK was the Public Enquiry into the construction of a pressurised water reactor at the existing nuclear site at Sizewell. This enquiry revealed that neither the operators nor the Nuclear Installations Inspectorate had a numerical basis for comparing the tolerability of risks from nuclear reactors to such things as deaths from earthquakes, aircraft crashes and lightning strikes. The Sizewell enquiry was adjourned until these things were developed and approved. The result was two seminal documents: ‘The Tolerability of Risk from Nuclear Power Stations HSE, London 1988 (revised 1992)’; and ‘Safety Assessment Principles for Nuclear Facilities (revised as 2006 Edition, Revision 1, HSE, Bootle, UK).

The US has also engaged in studying these issues. The latest document is the US Nuclear Regulatory Commission’s State-of-the-Art Reactor Consequence Analyses, which analyzed the potential consequences of severe accidents at the Surry Power Station, Virginia, and the Peach Bottom Atomic Power Station in Pennsylvania as real examples of nuclear reactors. The project combined up-to-date information about the plants’ layout and operations with local population data and emergency preparedness plans. This information was then analyzed using computer codes that incorporate decades of research into reactor accidents. More