Development of a BDD-based methodology to estimate importance measures in seismic probabilistic risk assessment

Probabilistic risk assessment (PRA) evaluates the core damage frequency (CDF) as the plant risk by considering failure events such as failure of structures, systems, and components (SSCs) and failure of operator's actions. To calculate the CDF, minimal cut-sets (MCSs) related to core damage are obtained from the PRA model developed with event trees and fault trees. Using these MCSs, quantification is performed and importance measures are estimated to confirm the impact of the failure events considered in the MCSs on the CDF. However, in the case of seismic PRA, since probabilities of seismic-induced failure events are relatively higher than that of random failure events and can lead to an overestimation of the CDF, the quantification of the CDF should be performed by converting the MCSs into a binary decision diagram (BDD) form and this form should also be considered in the importance measure for seismic event. In this study, we review the importance measures calculated using MCSs, which is mainly used in PRA for internal events, and the method of calculating these measures considering the BDD form is developed and applied to confirm the effect on failure events in seismic PRA.