Multi-peril loss estimation of industrial facilities

In a simplistic representation, we can define system risk as Risk = Hazard X Vulnerability, where hazard represents the probabilistic occurrence rates of catastrophic events, while vulnerability represents the probabilistic failure rate of the system subject to the hazard. An industrial facility, which is a complex system of multiple components, is prone to multiple hazards during its lifetime. Traditionally, total risk to the facility is computed by summing risk from individual hazards, which is adequate when hazards occur independently. However, when the multiple hazards show spatio-temporal correlation, estimating the total risk as merely a summation of individual hazard risks results in an inaccurate risk estimate. Moreover, the vulnerability of components of an industrial facility varies from one hazard to another. In this paper, we present a methodology for risk assessment of industrial facilities that accounts for the combined occurrence of multiple hazards and the variable vulnerability of components to each hazard. This method is based on simulating a time-dependent stochastic catalog of hazards, developed considering the interactions between hazards, focusing on earthquake and tsunami events. The lack of empirical data for developing hazard and vulnerability models imparts severe uncertainty to several components of the multi-hazard risk assessment methodology. We demonstrate the methodology on a coal fired power plant that is exposed to earthquake and tsunami hazard.