Resilience and risk-based restoration strategies for critical infrastructure under uncertain disaster scenarios

Authors：Kong, Jingjing; Zhang, Chao^*; Simonovic, Slobodan P.

Source：Sustainable Cities and Society, 2023, 92: 104510.

DOI：10.1016/j.scs.2023.104510

• Critical infrastructure
• System resilience
• Resource deployment and allocation
• Stochastic optimization
• Risk management

Improving the resilience of critical infrastructure (CI) systems to natural disasters is essential for society's longterm development. Considering the uncertainty in disaster scenarios, this study first proposes a two-stage riskneutral stochastic model to develop restoration resource deployment and allocation strategies with the objective of maximizing the expected value of resilience (EVR) of a CI system. Secondly, a risk metric is applied to assess the risk on the value of resilience imposed by the uncertainty. And a multi-objective optimization model for risk management is proposed, in which minimizing the value of the risk metric is formulated as a secondary objective to the EVR. The trade-offs between the EVR and the risk objective are assessed via the epsilon-constraint method. Finally, the utility of proposed models is illustrated using the electric power system of the Greater Toronto Area (GTA). The results show that: (i) there is a significant advantage in using risk-neutral strategies to improve system resilience over strategies without considering the characteristics of disaster scenarios; (ii) the value of the risk metric can be significantly reduced at the expense of decreasing EVR; and (iii) if the use cost of the resources is considered, there are optimal amounts of available resources for risk-neutral and risk-averse CI system operators. The models can assist CI system operators in selecting effective emergency response strategies for disasters with uncertain scenarios.

• Institution
  6