Risk-informed decision making: Application in nuclear power plant design & operation

Generally, a risk-informed analysis will involve a relatively broad spectrum of disciplines and realistic engineering analysis as required in NRC guidance. Traditionally, engineering analysis in commercial nuclear power applications has lacked detailed uncertainty analysis and quantification. Instead, "worst case" or credible extremes are evaluated separately. Where uncertainty has been evaluated [16, 17], it is not clearly related to risk or scenario importance. Quantification of risk and uncertainty is a basic need in a risk-informed regulatory paradigm. The risk-informed pilot project developed and used a new integrative approach, which could be more suitable in the risk-informed regulatory setting. More specifically, in this project (a) underlying physical and chemical models for the post-LOCA PRA basic events, associated with the concerns raised in GSI-191, were developed outside the PRA and (b) uncertainties were propagated in the physical models in order to have their results ready to be incorporated into PRA. The probabilities of the basic events (and their associated uncertainties), as they were obtained from the models of the underlying physical and chemical phenomena, were then incorporated into PRA. The uncertainties in the probabilities of basic events were also propagated through the PRA scenarios. The risk-informed pilot project is unique in some aspects such as it : (1) advances the PRA methodologies and applications by the integration of underlying physical and chemical phenomena, (2) presents the state-of-the art integration of probabilistic methods (e.g., uncertainty analysis, statistical analysis, risk analysis) and deterministic techniques (e.g., experimental testing, computer-aided design, and computational fluid dynamics), and (3) arose from collaborative work by teams of experts from industry and academia. The risk-informed project started in academia, but without private sector industry support, practical implementation would not have been possible. The utility and, particularly, its staff members assigned to solve the problem, maintained close relationships with the academic faculty, making it possible to react quickly to challenging research needs and to search for scientific and practical solutions. It is absolutely essential to understand how to incorporate practical experience from industrial practitioners in a project like the pilot project in order to produce a truly implementable program that can be an on-going operating experience. The progress of the research and its implementation have also significantly benefitted from the discussions and feedback received from regulatory representatives. The methodology used in the pilot project can be applied for other risk-informed projects related to nuclear power plants. Some of the authors are currently proposing to apply a similar methodology to fire PRA [18] and seismic PRA [19]. Another important area is the incorporation of underlying safety culture and organizational factors into risk-informed regulation. One of the authors has worked in this area of research [20] and, is advancing the benefits for applications to the nuclear power industry.