Numerical Simulation Model for Large-Scale Emergency Evacuation of Subway Passengers

Abstract At present, passenger density has significantly increased with the expansion of high-density subway systems. The risk of large-scale passenger emergency evacuation has also significantly increased. According to the literature review, research on the correlation between emergency stopping positions on trains and evacuation efficiency is insufficient. Additionally, the present literature lacks studies on the coupling relationship between fire-induced smoke hazards and evacuation in subways. These problems may lead to inaccurate assessment of casualties and efficiencies of subway passenger evacuation. Therefore, this study resolves these limitations by systematically investigating the key variables of train stopping positions, number of evacuees, shoulder widths, movement speeds, and crowd control protocols. Crucially, it integrates the coupled effects of fire smoke dispersion on evacuation dynamics. This study uses Fire Dynamic Simulator and Pathfinder simulation software to simulate the development of subway fires and the passenger evacuation process. It quantitatively assesses safety efficiency during fires and presents a new method for evaluating the efficiency of safe evacuation. Results indicate that passenger evacuation efficiency on the platform is higher than in the tunnel. Under defined conditions, the platform-level evacuation rate exhibits a critical inflection point at 3,296 evacuees. Evacuation efficiency declines rapidly once this passenger threshold is exceeded. The platform occupancy directly determines the overall success rate of subway evacuation. This inflection point methodology will be helpful for planning evacuation strategies in emergencies.