Passenger safety is one of the main goals of railway tunnels design, being the evacuation of a burning train one of the worst scenarios, where any delay to start the evacuation is crucial for passenger’s survival. Previous research works have separately studied the evacuation of railway tunnels due to fires and the effect of tunnel slope on gas and smoke spread, but none of them has addressed both factors together. In this work, we developed a quantitative approach to assess the time delay to start the evacuation, depending on the tunnel slope. The methodology is based on statistical analysis of simulation results. The proposed model, based on linear multiple regression with an R-square value close to 90%, explains the number of fatalities as a function of the time delay to start the evacuation and the tunnel slope. The statistical model used in this study predicts more than one fatality for each second of delay in starting the evacuation. Moreover, tenable conditions for safe evacuation in case of fire cannot be easily guaranteed in inclined tunnels with more than 1 km length and natural ventilation.

旅客安全是铁路隧道设计的主要目标之一，这是最坏的情况之一，即疏散燃烧中的火车，疏散疏散对于旅客的生存至关重要。先前的研究工作已经分别研究了火灾引起的铁路隧道疏散以及隧道边坡对瓦斯和烟气扩散的影响，但没有一个能够同时解决这两个因素。在这项工作中，我们开发了一种定量方法来评估开始疏散的时间延迟，具体取决于隧道的坡度。该方法基于对模拟结果的统计分析。所提出的模型基于R值接近90％的线性多元回归，解释了死亡人数与开始疏散和隧道坡度的时间延迟的关系。在这项研究中使用的统计模型预测，疏散开始的每一秒延迟都会导致不止一个人死亡。此外，在长度超过1公里且自然通风的倾斜隧道中，难以轻易保证有可靠的疏散条件，以免发生火灾。