Numerical Simulation of Air Pollutant Distribution in Urban Tunnels

Abstract

Numerical simulation of air pollution dispersion inside tunnels is a suitable method for studying air pollutants’ spatial distribution and for evaluating the tunnel’s ventilation efficiency. In the present work, Fluent 6.2, a computational fluid dynamic software, has been used for full-scale numerical simulation of air flows and carbon monoxide (CO) concentrations inside the Resalat Tunnel of Tehran. Fans and vehicles are simulated with source of momentum and porous jump boundary conditions, respectively. Also, source of air pollutants, i.e., vehicle emissions, is simulated as uniform area source in the tunnel floor. Modeling results for concentrations of CO are validated by measurement data in 24 points adjacent to the fans inside the tunnel. Calibration of model indicated that the moving wall porous jump method for simulation of vehicle geometry and their effects, momentum source for fans modeling, standard k − ϵ scheme for turbulence modeling, and hexahedral mesh type are proper choices for the developed model. The results show a good correlation (R = 0.9) between modeling and measurement data. Five different scenarios (namely 1–vehicles to be stopped, 2–fans off, 3–two vertical ventilation ducts considered in the tunnel, 4–vehicles with Euro-IV emissions standard instead of Euro-II, and 5–blowing power of fans increased to twice the present) are examined for CO concentrations inside the tunnel. The numerical simulations for these scenarios are modified using a relation between measurement data and modeling results. Following the modification, results show that at the last measurement point (near the end of the tunnel), concentrations of CO is 59, 77, 9, 23, and 14 ppm for the five mentioned scenarios, respectively, and it is 32 ppm for normal condition. Hence, appropriate measures may be undertaken by the city authorities for air quality improvements in urban tunnels.