A probabilistic framework to simulate the dynamics of post-earthquake rescue operations

Post-earthquake rescue studies often overlook critical factors influencing rescue operations, such as injury aggravation, transportation disruptions, and debris removal. This study presents a probabilistic framework to evaluate rescue operations in terms of time and cost. It integrates probabilistic models of earthquake occurrence, buildings’ response, debris generation, transportation network disruptions, and injury aggravation into an Agent-Based Simulation (ABS) model to replicate rescue and debris removal processes. A Developed Minimum Spanning Tree (DMST) algorithm is incorporated to dynamically update patient transfer times based on real-time transportation network conditions. Key contributions include: (1) unifying hazard, damage, casualty, debris, and transportation models to simulate post-earthquake rescue operations; (2) assessing costs associated with injury aggravation and fatalities due to delayed rescues; (3) quantifying the impact of road blockages on transportation network functionality; and (4) optimizing rescue efforts by dynamically identifying optimal routes using the DMST algorithm. The framework is tested on the INSURER City virtual platform, yielding probabilistic distributions for buildings’ damage, injuries, fatalities, costs, rescue response times, and transportation network performance. Sensitivity analysis reveals that while a random debris removal team with a rescue crew achieves a cost reduction of 1.55 %, the most significant benefits come from targeted debris removal combined with rescue crews, resulting in a cost reduction of 2.33 %. It also indicates that increasing the number of rescue teams is more effective in reducing fatalities and costs than modifying debris removal strategies. This framework provides valuable applications in assessing the post-disaster serviceability of transportation networks, enhancing rescue operations, reducing response times, and minimizing casualties.