A simplified quantification method for seismic risk assessment of nuclear fuel cycle facilities using Clark approximation

Nuclear fuel cycle facilities are important elements for supporting the efficient use of energy resources by establishing a nuclear fuel cycle. However, given that the risks of these facilities are lower than those of nuclear power plants, it is considered reasonable to apply simplified assessment methods when evaluating seismic risks. In this study, a simplified quantification method is proposed for seismic risk assessment at such facilities. Traditional simplified methods have streamlined the assessment process by selecting only representative components, often neglecting others. In contrast, the proposed method simplifies the required computational processes while considering all components by applying Clark approximation. Clark approximation is a mathematical method for approximating the maximum of two normal distributions as a new normal distribution. The proposed method was validated by comparing its seismic probabilistic risk assessment with those performed using Monte Carlo simulations and traditional simplified methods. Results showed that although the proposed method overestimated the high confidence of low probability of failure by a relative difference of 0.15 compared with that of the Monte Carlo method under completely independent condition, the overall plant-level fragility curve was generally within the range of the 5% and 95% confidence fragility curves. The proposed method accounted for the impact of correlated failure, which is critical in seismic risk assessments. Thus, this method enabled the seismic risk assessment of nuclear fuel cycle facilities in a simplified manner without compromising accuracy, potentially contributing to examining risk mitigation measures and developing risk-informed safety regulations for these facilities.