Enhancing Radiological Emergency Response Through Agent-Based Model Case 1: Effectiveness of Staged Evacuation

Abstract

In the event of radiological emergencies, such as nuclear accidents, the speedy but well-ordering evacuation of residents is imperative to ensure their health and safety. Staged evacuation has been studied as a method to reduce evacuation time. However, a method is needed to confirm the effectiveness of staged evacuation and to optimize the strategy. This study evaluates the performance of staged evacuation with more realistic factors which are implementable in a simulation platform. The evacuation simulation was conducted using PRISM (Platform for Radiological Emergency Integrated Simulation Model), developed by Kyung Hee University and utilizing an Agent-Based Model. A city was modeled using GIS data for road networks, buildings, and administrative districts. The population distribution was made non-uniform based on building layout data, and evacuation areas were divided based on administrative districts for demonstrating purpose. The simulation varied the order of evacuation groups and the evacuation start time interval, comparing evacuation times and average speeds to evaluate each strategy. The results reached a few insights: (1) Certain staged evacuation strategies are effective in reducing evacuation times within the PAZ. (2) A staged evacuation strategy that prioritizes areas closest to the nuclear plant resulted in longer evacuation times compared to other evacuation sequences. (3) There is no significant relationship between evacuation time and evacuation speed. In other words, even if the evacuation times of strategies are similar, the evacuation speeds may differ. Since evacuation speed indicates road congestion, a strategy with a slightly longer evacuation time but a faster evacuation speed may be chosen depending on the objective.