Probabilistic risk assessment for interdependent critical infrastructures: A scenario-driven dynamic stochastic model

Critical infrastructures (CIs) are becoming increasingly important in social public services; however, various CI risk events emerge constantly with potential damages. Probabilistic risk assessment (PRA) is a quantitative measurement of risk occurrence probability that is used to support risk profile judgment and weakness identification. However, the inherent features of risk factor multiplicity, CI interdependency, and dynamic stochasticity render PRA for CIs more challenging. The purpose of this study is to investigate a PRA model for CIs with a comprehensive consideration of the inherent features and an integrated utilization of multisource data. A multidimensional PRA scenario analysis is first conducted from the perspectives of scenario elements, scenario evolution, and scenario effect. Subsequently, to support PRA for interdependent CIs, a scenario-driven dynamic stochastic model is developed based on a three-stage solution with accurate feature quantification, and effective utilization of objective factual records and subjective expert judgment. Furthermore, the applicability and effectiveness of the proposed model are demonstrated through a case study. It is indicated that the PRA results obtained using the proposed model are beneficial for decision makers to clarify the overall risk profile and determine the risk-alert periods, high-frequency risk factors, and high-risk CIs to support CI risk prevention.