Refinement of vehicle emission factor estimation in tunnel experiments using CFD and field measurement.

Tunnel experiments are widely used to determine fleet-average vehicle emission factors (EFs) under real-world conditions. However, limited sampling locations often lead to inaccuracies due to spatial heterogeneity in wind and pollutant dispersion. This study integrates CFD modeling with tunnel measurements to quantify and correct such errors. Simulations reveal that EFs derived from field measurements alone can be underestimated by up to 40 %, particularly under low traffic density and weak natural wind conditions. The estimated correction schemes for EFs were evaluated and validated using practical case studies. The results showed that the average EFs of NOx, CO, and PM2.5 for the tunnel fleet were underestimated by approximately 24.50, 161.66, and 3.33 mg km-1·veh-1, respectively. Although the corrected EFs for diesel vehicles remain significantly higher than those for petrol vehicles, their relative contribution to the fleet-average EFs is notably reduced. The CFD analysis also highlights that external atmospheric conditions strongly influence internal tunnel flows, especially within 200 m of the entrance, offering valuable guidance for tunnel siting and sampling strategies. This study enhances the reliability of tunnel-derived EFs and supports the standardization of tunnel experiment protocols, contributing to more accurate emission inventories and targeted air quality management policies.