A Markovian approach to infrastructure life‐cycle analysis: Modeling the interplay of hazard effects and recovery

Abstract

AbstractProbabilistic life‐cycle cost (LCC) analysis facilitates optimal management of infrastructure systems under uncertainty. Despite recent advancements, accurate and efficient estimation of the LCC of systems facing risk of exposure to multiple stochastic occurrences of hazards has remained a significant challenge. This paper proposes a Markovian LCC framework that substantially reduces the often very high computational cost of probabilistic LCC analysis especially for large systems. The presented methodology also enhances the accuracy of LCC estimates in multiple fronts. It captures damage accumulations due to incomplete or no repair actions over multiple stochastic hazard exposures in a system's lifetime. Moreover, it tracks the state of infrastructure vulnerability during repair processes and incorporates uncertainties in downtimes associated with damage states of the system. Finally, it estimates the entirity of loss induced by a hazard and avoids the truncation error that exists in annual loss estimates for scenarios where repair downtimes are greater than one year. With these new capabilities, the proposed LCC framework is applied to a bridge in a seismic zone to estimate the LCC and analyze complex interactions of stochastic hazards and their effects with recovery efforts. Results indicate that neglecting effects of damage accumulation and partial improvements in the course of repair can lead to 20% underestimation and 14% overestimation of the expected hazard‐induced LCC, respectively.