Multi-hazard life-cycle consequence analysis of deteriorating engineering systems

Abstract

Probabilistic life-cycle consequence (LCCon) analysis (e.g., assessment of repair costs, downtime, or casualties over an asset’s service life) can enable optimal life-cycle management of critical assets under uncertainties. This can lead to effective risk-informed decision-making for future disaster management (i.e., risk mitigation and/or resilience-enhancing strategies/policies). Nevertheless, despite recent advances in understanding, modeling, and quantifying multiple-hazard (or multi-hazard) interactions, most available LCCon analytical formulations fail to accurately compute the exacerbated consequences, which may be due to incomplete or absent repair actions between different interacting hazard events. This paper introduces a discrete-time, discrete-state Markovian framework for efficient multi-hazard LCCon analysis of deteriorating engineering systems (e.g., buildings, infrastructure components) that appropriately accounts for complex interactions between natural-hazard events and their effects on a system’s performance. The Markovian assumption is used to model the probability of a system being in any performance level (i.e., limit state) after multiple hazard events inducing either instantaneous and/or gradual deterioration and after potential repair actions through implementing stochastic (transition) matrices. The LCCon estimates are then obtained by combining the performance limit state distributions with suitable system-level consequence models in a computationally efficient manner. The proposed framework is illustrated for two case studies subject to earthquake and flood events, as well as environment-induced corrosion during their service life. The first is a reinforced concrete building and the second is a transportation road network with a reinforced concrete bridge.