Computational modeling reveals double-sided evacuation maximizes survival during subway tunnel fires, highlighting the need for low-level smoke extraction alongside sprinklers.
Key Points
To develop and validate an integrated model that couples fire progression with human egress to determine pedestrian survivability thresholds in subway tunnels.
Coupled Large-Eddy Simulation computational fluid dynamics with agent-based pedestrian trajectory modeling for 2,160 evacuees in a 3D tunnel featuring a 2% longitudinal gradient.
Simulated 9.5 MW and 12 MW fire heat release rates across three evacuation strategies: Single-Sided Evacuation (SSE), Double-Sided Evacuation (DSE), and Sprinkler-Assisted Single-Sided Evacuation (SSE-S).
Continuously tracked lethal thresholds using Fractional Effective Dose (FED) indices, 60°C temperature boundaries, and 500 ppm carbon monoxide dispersion fronts.
The tunnel gradient produced an intense stack effect driving temperatures beyond 1200°C, causing SSE to create fatal crowd bottlenecks with a peak Fractional Effective Dose of 15.33 compared to a survivable peak FED of 0.52 under DSE.
Sprinklers in the SSE-S scenario successfully suppressed the fire within 105 seconds but caused thermal cooling that collapsed smoke buoyancy, precipitating toxic gas stratification directly into the pedestrian breathing zone.
Cite This Study
Kırlangıçoğlu et al. (2026) studied this question.