Probabilistic analysis of fire casualties in an urban traffic tunnel considering the uncertainty of vehicle queue

Fire accidents could lead to severe casualties, especially in urban traffic tunnels, due to the concentration of users and limited space. However, it is challenging to evaluate the fire casualties in tunnel fires due to the complicated development of fire environment and the uncertainty of personnel distribution, which is influenced by vehicle queue. In this paper, a prediction model on evaluating probability of fire casualties in urban traffic tunnels is proposed, considering the evolution of the thermal and carbon monoxide distribution, as well as the uncertainty of vehicle queue. Firstly, the general framework of the model is introduced and the procedure for generating the probability curve is outlined. Subsequently, each module of the model, including the evolution of the fire scenario, simulation of evacuation process, evaluation of fire casualty and fitting of probability curve is elaborated in details. Then, this model is applied to a three-lane traffic tunnel in Shanghai to investigate the influence of heat release rate, fire source locations and ventilation systems on the probability curves of fire casualty expectation. The results show that there is a certain correlation in the impact of heat release rate and fire source location on fire casualty expectation. Besides, the ventilation condition has a significant impact on fire casualty expectation. The expected values of fire casualty expectation under longitudinal ventilation system could sometimes be only one-tenth of those under the natural ventilation condition. This study makes some contributions to evaluating the personnel loss in tunnel fire and lays a foundation for assessing and improving resilience of tunnel.