Assessing nuclear power plant safety and recovery from earthquakes using a system-of-systems approach

We adopt a 'system-of-systems' framework of analysis, previously presented by the authors, to include the interdependent infrastructures which support a critical plant in the study of its safety with respect to the occurrence of an earthquake. We extend the framework to consider the recovery of the system of systems in which the plant is embedded. As a test system, we consider the impacts produced on a nuclear power plant (the critical plant) embedded in the connected power and water distribution, and transportation networks which support its operation. The Seismic Probabilistic Risk Assessment of such system of systems is carried out by Hierarchical modeling and Monte Carlo simulation. First, we perform a top-down analysis through a hierarchical model to identify the elements that at each level have most influence in restoring safety, adopting the criticality importance measure as a quantitative indicator. Then, we evaluate by Monte Carlo simulation the probability that the nuclear power plant enters in an unsafe state and the time needed to recover its safety. The results obtained allow the identification of those elements most critical for the safety and recovery of the nuclear power plant; this is relevant for determining improvements of their structural/functional responses and supporting the decision-making process on safety critical-issues. On the test system considered, under the given assumptions, the components of the external and internal water systems (i.e., pumps and pool) turn out to be the most critical for the safety and recovery of the plant. - Highlights: • We adopt a system-of-system framework to analyze the safety of a critical plant exposed to risk from external events, considering also the interdependent infrastructures that support the plant. • We develop a hierarchical modeling framework to represent the system of systems, accounting also for its recovery. • Monte Carlo simulation is used for the quantitative evaluation of the model, with respect to the safety and recovery of the system