Probabilistic dynamic resilience quantification for infrastructure systems in multi-hazard environments

Abstract

Resilience has been evolving as a key criterion for infrastructure systems as it ensures the system's dynamic performance pre-, during, and post-hazard disruptions. However, estimating these performances is challenging due to system and operation complexities, and the probabilistic dynamic nature of infrastructure system. Moreover, infrastructure systems are usually exposed to multi-hazard environments, with their own probabilistic behavior, leading to additional complexity in terms of estimating the system response and, subsequently, the overall system resilience. As such, this study develops a probabilistic resilience-centric system dynamics modeling approach to quantify infrastructure dynamic resilience based on a holistic representation of infrastructure systems under multi-hazard scenarios, whereby the probabilistic natures of both the hazards and system are incorporated. Unlike the traditional resilience quantification approaches that represent system resilience by a single value calculated after the system's full recovery, the developed model focuses on tracking the temporal evolution of system resilience along the entire period of system performance deterioration and recovery. A real-world hydropower dam, as an example for infrastructure systems, in British Columbia, Canada is used as a demonstration application to show model utility in developing resilience-guided assessment plans for infrastructure systems. Overall, the developed approach empowers the decision-makers with insights into critical operational periods, the required time to reach specified resilience targets, and the efficiency of risk mitigation measures in real-time.