Estimating regional CO2 and NOx emissions from road transport using real-world data-based emission factors in Korea

The average-speed emission model (Speed-based model), a widely used and simple method of calculating road vehicle emissions, offers easy accessibility by expressing emissions as a function of average speed. However, there are limitations in expressing emissions generated through complex mechanisms simply as a function of speed. Real-world driving tests using a portable emission measurement system can incorporate the impact of vehicle driving load on emissions. In this study, we analyzed real-world emissions data from 94 light-duty vehicles and developed time-based emission factors depending on vehicle speed and vehicle-specific power (VSP). We also propose a speed-VSP based model to estimate regional CO2 and NOx emissions by combining time-based emission factors and vehicle operating times. The speed-based model and Speed-VSP based model exhibit a 44% difference in NOx emissions and a 29% difference in CO2 emission. In a comparison of the two models against RDE test results, the speed-VSP based model achieved high accuracy in predicting NOx and CO2 emissions with a lower root mean square error (RMSE). Specifically, for NOx emissions predictions, the speed-VSP based model achieved an RMSE of 122-270 mg/km, while the speed-based model showed a much higher RMSE of 435-476 mg/km. For CO2 emissions predictions, the speed-VSP based model achieved an RMSE of 34-56 mg/km, while the speed-based model showed a much higher RMSE of 36-72 mg/km. The results of this study present an opportunity to reassess and improve conventional method of measuring and evaluating emissions from road transport.