A dynamic Bayesian network approach to characterize multi-hazard risks and resilience in interconnected critical infrastructures

A new paradigm for risk assessment has emerged, recognizing the escalating frequency and severity of disasters associated with natural hazards. Conventional risk assessments often fail to capture the dynamic and interconnected nature of disruptions within infrastructure systems during failure scenarios. This study introduces a Dynamic Bayesian Network (DBN) framework, designed to assess risk in interconnected infrastructure systems under complex hazard scenarios. The framework addresses the limitations of static models by dynamically capturing the progression of disruptions during failure and the restoration process during recovery. Using a case study in Saint Lucia, a Caribbean Island susceptible to natural hazards, this study examines the complex network of critical infrastructure. The DBN framework explores various failure scenarios, highlighting the cascading effects across infrastructure sectors, and captures the probabilistic hazard conditions and functional dynamics during disruption and restoration processes. Results from the case study illuminate the heightened vulnerability of the international airport and tourism sectors, emphasizing the interdependencies and propagation of failures within the infrastructure system. By investigating failure scenarios, the DBN approach characterizes the complex interactions between infrastructure systems, providing valuable insights into how multi-hazard events affect interconnected networks. These findings underscore the critical need for dynamic, real-time risk assessments that consider both short-term disruptions and long-term recovery processes. The study highlights the urgency of embracing dynamic risk assessment methodologies and offers a foundation for developing adaptive, multi-hazard risk assessment strategies to enhance the resilience of critical infrastructure networks.