Simulation-Based Performance Analysis of Ventilation and Fire Protection System in Road Tunnel Fires

Abstract This paper evaluates the performance of ventilation and fire-protection system in a 200 metre road tunnel using PyroSim/FDS. Three design fires were modelled — heavy goods vehicle (100 MW), bus/truck (25 MW) and passenger car vehicle (10 MW)— under two configurations: natural ventilation (baseline) and an integrated system combining longitudinal mechanical ventilation with sprinklers. Under baseline conditions, simulations show rapid smoke back-layering, rising thermal loads and loss of tenability within minutes. The integrated configuration preserved upstream tenability and directed smoke downstream, moderating—but not entirely eliminating—severe local heating; in the integrated runs the fire growth was constrained, with HRR behaviour capped at values comparable to those observed under the natural-ventilation cases. The discussion interprets thermal fields, smoke and CO distribution, visibility and sensor/sprinkler response, and draws implications for evacuation routing, control strategy for jet fans, and performance-based design. Simulation fidelity was corroborated using an Artificial Neural Network (ANN) regression (R = 0.97–0.99), supporting the reliability of the numerical findings. The results underline the limitations of natural ventilation alone and the practical benefits of coordinated ventilation–suppression strategies for tunnel fire safety. Keywords: Tunnel fire, Ventilation, Sprinkler suppression, PyroSim, Back-layering