|
Abstract
|
The paper offers an overview of the current regulatory background within the United States, emphasising nuclear regulation, related to the establishment of quantitative safety goals as a way to respond to the key issue of 'How safe is safe enough?' The approach described in the paper, based on use of probabilistic risk assessments techniques is centred around definition of quantitative safety regions. Regions are termed: design basis, safety margin, the design capability, and safety research.
The design basis region is bounded by the frequency of 10−4 per reactor year and consequences of no identifiable public injury. The frequency of 10−4 per reactor year, is associated with the total projected lifetime of a commercial US nuclear power programme. Events which have a 50% chance of happening are included in the design basis region. In the safety margin region, which extends below the design basis region, protection is provided against events whose probability of not happening during the expected course of the US nuclear power programme is within the region of 50 to 90%. Setting the lower mean frequency to this region of 10−5 per reactor year is equivalent to offering 90% assurance that an accident of given severity will not happen. A rare event with a mean frequency below 10−5 can be predicted to occur. However, accidents predicted to have a probability of less than 10−6 are 99% certain not to happen at all, and are thus not anticipated to affect public health and safety. Safety goals associated with individual risk to a maximum exposed member of the public, general societal risk, and public property risk are proposed in the paper. The derivation and rationale behind the above stated goals are briefly discussed. |