There is growing evidence that vehicle emissions under real-world driving conditions can be significantly higher than certified emissions values as tested in the laboratory. As part of China’s adoption of more stringent China 6 emissions standard, China is adopting a more representative cycle, the Worldwide Harmonized Light Vehicles Test Procedure (WLTP), adding real-driving emissions (RDE) testing requirements, and introducing a comprehensive in-use compliance program.
For this study, two modern China 5 gasoline cars were tested in the laboratory on a chassis dynamometer and on the road by using a Portable Emissions Measurement System (PEMS). Laboratory chassis dynamometer tests included standard China 5 type-approval tests (NEDC, 25℃, and cold start) and tests under various cycles and conditions to reflect real-world complexity, such as low/high ambient temperature, cold/hot start, air conditioning operation, and on the WLTP. The RDE test routes, equipment, and ambient conditions all met the provisions in the China 6 RDE regulation.
Vehicle A, a 1.6 L small sedan with port fuel injection (PFI), and Vehicle B, a 2.4 L multi-purpose vehicle with gasoline direct injection (GDI), both passed the China 5 type-approval tests over the NEDC. However, nitrogen oxides (NOX) emissions of Vehicle A on the NEDC after a hot start and on the standard WLTP with a cold start increased significantly and exceeded the China 5 limit. In addition, the NOX results of Vehicle A in RDE tests were on average 1.6 times the China 5 laboratory limit. NOX emissions from Vehicle B were even lower than the China 6b limit under all laboratory and RDE tests. Our in-depth investigation indicates that the high NOX of Vehicle A is most likely attributable to a poor and lenient design of the fuel injection control strategy.
Both vehicles show significantly higher emissions of carbon monoxide (CO) over the more dynamic WLTP tests. This is probably because the engines were running at rich air-to-fuel ratios during the higher engine loads imposed by harder accelerations in the WLTP. In RDE tests, CO emissions in some cases exceeded the China 5 limit by 2.8 times. The results provide sound arguments that CO emissions from gasoline cars are not properly controlled under real-world driving conditions and need particular attention from the regulators.
Particle number (PN) emissions from Vehicle A, a conventional PFI car, managed to stay at low levels in all circumstances, while Vehicle B, a GDI car, had an order of magnitude higher PN emissions in laboratory and RDE tests compared with Vehicle A. Our results highlight the importance of better controlling particle emissions from GDI cars.