On Compound Hazards in Cost-Effective Recovery Management

The failure of an infrastructure component can compromise the overall performance of a system. Approaches to quantifying component vulnerability typically focus on single-hazard scenarios, neglecting, for example, the effect of consecutive hazards on the system, which could increase the vulnerability of the system over time. This can lead to inaccurate risk or resilience assessments and consequently to inadequate decision-making. In this work, we apply a framework for modeling the evolution of component vulnerabilities in response to compound hazards using state-dependent fragility curves. The approach employs a Markovian state transition model in which a hazard event can alter the state of system components. Using this approach, we illustrate how the interaction between hazards can affect the degradation of components. We show that the compounding effect of two hazardous events depends on both the time lag between the two and the recovery capacity of the affected components. We illustrate this relationship by analyzing hazard time series with different correlation strength, and by defining a minimal recovery rate required to maintain system performance below a prescribed failure threshold. Finally, we incorporate a cost analysis by linking recovery and replacement processes to cost functions, allowing the computation of these costs over time. This allows the identification of appropriate strategies when hazards are correlated versus independent.

Keywords: Natural hazards, compound hazards, fragility curves, Markov chain, infrastructure, recovery function, cost-benefit analysis.