Real-Time Emission Factors of Motor Vehicle Air Pollutants Based on Tunnel Experiments

Jiang Weisi; Kong Shaofei; Yan Yingying; Wu Jian; Zheng Huang

doi:10.3799/dqkx.2025.170

Abstract：Dynamic emission factors of motor vehicle air pollutants are a key parameter limiting the improvement of real-time emission inventory accuracy.In this study,a tunnel in a metropolitan area was selected as an observation site.Real-time monitoring of the concentrations of seven air pollutants(CO,NO,NO2,NOx,SO2,BC,and PM2.5)and traffic flow was conducted using online monitoring instruments and surveillance cameras.Vehicle type and speed were obtained by applying the YOLOv8l deep learning object detection model in combination with the SORT tracking algorithm.Based on the emission intensity ratio method,the average emission factors for the entire fleet and for different vehicle categories were estimated to be(1 064.9±479.8),(496.5±209.3),(55.5±30.4),(578.6±267.6),(6.3±2.2),(3.3±1.5),and(37.7±19.2)mg/(km∙veh)for CO,NO,NO2,NOx,SO2,BC,and PM2.5,respectively.The overall fleet emission factors were(634.7±477.2),(266.0±142.9),(26.4±13.5),(302.3±159.5),(3.5±1.9),(2.0±1.1),and(19.8±12.3)g/km for these pollutants,respectively.During the observation period,the average daily traffic volume on weekends was 88.6％of that on weekdays.Except for PM2.5,weekday emission factors for all pollutants were 1.0-1.48 times higher than those on weekends.Hourly analysis showed that the proportion of diesel vehicles during the early morning was 1.6 times that of other periods,with pollutant emission peaks 2.0-3.5 times the daily average.Fleet emissions exhibited a bimodal diurnal pattern,peaking during morning(07:00-09:00)and evening(17:00-19:00)rush hours at 1.8-3.3 times the daily average.The findings provide essential data and scientific support for constructing high-resolution dynamic motor vehicle emission inventories and implementing refined control strategies for vehicular pollutant emissions.