A field study of CO2 and particulate matter characteristics during the transition season in the subway system in Tianjin, China

The daily passenger flow and ride times for subway systems are increasing in many large cities. Understanding the characteristics of carbon dioxide (CO2) and particulate matter (PM) in subway stations and train cars is essential in reducing passenger exposure to these pollutants. This study conducted field sampling on two subway lines in Tianjin, China (an old line, L1, and a new line, L6), during the transition season for a period of three weeks. During the first week, the heating, ventilation, and air conditioning (HVAC) systems in subway cars were turned off, while during the following two weeks, the systems were turned on. Measurements were conducted in subway cars, in stations (6 typical above-ground and underground stations), and in the ambient environment simultaneously. The CO2 concentrations in the cars were linearly related to passenger density (p < 0.05), yet independent of the HVAC system. When the HVAC system was off, the CO2 concentrations during the morning and evening peak periods were 2016 ± 567 ppm and 1534 ± 498 ppm, respectively, with the maximum concentrations observed at 3849 ppm and 3282 ppm. When the CO2 concentration was higher than 1500 ppm (Chinese air-quality standard), the exposure time were about 37.6 and 26.5 min, respectively, in the morning and evening peak periods. After the HVAC system was turned on, the CO2 concentrations in the cars dropped by 21–27%. Meanwhile, PM1, PM2.5 and PM10 concentrations in subway cars were 11.1 ± 7.6 μg/m3, 23.1 ± 6.5 μg/m3 and 68.8 ± 15.5 μg/m3, respectively. The ratios of PM concentration in carriage/in station (I/O) for PM1, PM2.5 and PM10 were 0.57, 0.53, and 0.42, respectively. Operating the HVAC systems in subway cars reduced the I/O ratios by 28.2–49.2%. The full-screen doors used on line L6 performed better than the half-screen doors used on line L1 in preventing the diffusion of particles from the tunnel to platforms by the piston effect. Based on the sampling results, possible air control strategies for CO2 and particle pollutants were also evaluated.