Survey of air exchange rates and evaluation of airborne infection risk of COVID-19 on commuter trains

Shinohara, Naohide; Sakaguchi, Jun; Kim, Hoon; Kagi, Naoki; Tatsu, Koichi; Mano, Hiroyuki; Iwasaki, Yuichi; Naito, Wataru

To identify potential countermeasures for coronavirus disease (COVID-19), we determined the air exchange rates in stationary and moving train cars under various conditions in July, August, and December 2020 in Japan. When the doors were closed, the air exchange rates in both stationary and moving trains increased with increasing area of window-opening (0.23–0.78 /h at 0 m(2), windows closed to 2.1–10 /h at 2.86 m(2), fully open). The air exchange rates were one order of magnitude higher when doors were open than when closed. With doors closed, the air exchange rates were higher when the centralized air conditioning (AC) and crossflow fan systems (fan) were on than when off. The air exchange rates in moving trains increased as train speed increased, from 10/h at 20 km/h to 42/h at 57 km/h. Air exchange rates did not differ significantly between empty cars and those filled with 230 mannequins representing commuters. The air exchange rates were lower during aboveground operation than during underground. Assuming that 30–300 passengers travel in a train car for 7–60 min and that the community infection rate is 0.0050%–0.30%, we estimated that commuters’ infection risk on trains was reduced by 91%–94% when all 12 windows were opened (to a height of 10 cm) and the AC/fan was on compared with that when windows were closed and the AC/fan was off.