Implication of PSA uncertainties on risk-informed decision making

Abstract

Nuclear power plants risk-informed policy is introduced in order to improve safety decision making and regulatory efficiency. The corresponding regulatory guides define the acceptable risk measures and their changes resulting from the modifications in the licensed design of the nuclear power plant. The risk measures used in the acceptance guidelines are the core damage frequency and large early release frequency. The risk measures and their corresponding changes are assessed by the Probabilistic Safety Assessment. The uncertainties of Probabilistic Safety Assessment should be appropriately addressed in the context of the decision making, considering their implication on the obtained results. The Probabilistic Safety Assessment uncertainties include epistemic uncertainties resulting from parameter, model, and completeness uncertainties. The paper presents the obtained results from the uncertainty analysis of the Probabilistic Safety Assessment of the reference nuclear power plant and their implication on risk-informed decision making. The paper focuses particularly on parameter and model uncertainties. The analysed modification is extension of the test interval of the emergency diesel generators. The core damage frequency is the used risk measure in the analysis. The need for the appropriate consideration of the uncertainties in the Probabilistic Safety Assessment in order to adequately support the risk-informed decision making is identified. The deficiency of usage of percentile measures is identified and acknowledged. The need for the adaptation of the risk-informed decision-making principles considering new nuclear power plants is recognized.