Nuclear reactor safety goals and assessment principles

Although regulatory requirements for the safety of nuclear reactors are still largely deterministic, two national regulatory authorities have recently introduced the use of quantitative evaluations of risks based upon the probabilities of accidents. These developments are contained in the safety assessment principles promulgated in 1979 and used in licensing by the UK Nuclear Installations Inspectorate, and the safety goals adopted for trial use by the US Nuclear Regulatory Commission early in 1983. Both approaches incorporate forms of risk limitations and optimisation, and therefore appear to be adaptions of the radiation protection principles which the International Commission on Radiological Protection applies to exposures from controlled sources of radiation. However, in the regulatory authorities' approaches: there are no upper limits of risk specified; and optimisation appears to be required only if specified levels of individual risk are exceeded. Thus, in effect, the regulatory authorities have introduced a de minimis concept for risks of accidental exposure. The UK and the US approaches to risk assessment are discussed and problems associated with each of them are identified. Ways are suggested in which the use of probabilistic risk assessment could be extended