Optimization of Emergency Egress Doors in Road Tunnels

As a general approach, the required maximum spacing between the emergency exits in road tunnels is provided not to exceed 300 m, as stated in NFPA 502. It is also stated that there is not considered to be any minimum exit separation, while most typical exit separations are between 30 m - 200 m and appropriate exit separation distances can only be determined by engineering analysis of emergency egress requirements.

This study focuses on the incident tunnel and aims to analyze the effects of spacing and clear widths of egress doors on evacuation time. To accomplish this, generic tunnel geometries with various lengths and varying number of lanes are modeled to accommodate different number of vehicles to be present along the tunnel. Egress pathways leading directly to an exit are also modeled for the tunnels with various clear widths exceeding 1.12 m as per NFPA 502. Four types of vehicles are modeled with varying numbers in the tunnels and their initial locations are provided to demonstrate a highly-congested (almost stopped) condition with different spacing between each other. The number of motorists is calculated based on the total number of vehicles and additionally increased to provide varying numbers for each scenario. Hence, four variables, i.e., exit doors spacing, exit door clear width, egress pathway clear width, total number of motorists, are considered to be varying for each scenario.

Emergency evacuation simulations are then performed using Pathfinder, to obtain the total evacuation time, i.e. RSET (Required Safe Egress Time), for the incident tunnel. Finally, a supervised machine learning algorithm is developed to evaluate the effects of each variable on evacuation time, and the possibility of optimizing the egress doors.