A Euro 4 Light-Duty Diesel vehicle equipped with a diesel particulate filter (DPF) was circulated to 9 labs where repetitions of the current regulatory New European Drive Cycle (NEDC) were conducted. Regulated gaseous and improved (with cyclone, filter temperature 47 ± 5• C, constant filter face velocity, high precision balance at all labs) particulate mass (PM) measurements were also conducted. A reference particle number (PN) measurement system measuring non-volatile particles was circulated along with the test vehicle. Labs also tested their own PN systems built to comply with the reference system's performance specifications. The mean PN emissions level of the vehicle was below 1 × 10 11 particles/km. The intra-lab variability (repeatability) was ∼40% and the inter-lab variation was ∼25%. The study showed that the new PN method had similar variability to other gaseous pollutants such as carbon monoxide and hydrocarbons and better than the PM (intra-lab variability ∼55% and inter-lab ∼35%). Even with the improved PM method the emissions of the vehicle were similar to the background level (∼0.4 mg/km) and the method was subject to volatile artifact. The PN method showed greater sensitivity than the PM method as it could distinguish the DPF fill state or different preconditioning states of the vehicle. However, the PN emission level of the vehicle estimated by the reference system were on average 15% higher than any given lab's own system, indicating that the procedures and calibration designed for the standardization of performance should be precisely defined and followed. This work has been conducted in the framework of the PMP project (run under the auspices of the UNECE-GRPE). The authors would like to thank all the laboratories and companies that participated. In addition, Matter Engineering AG for providing the dilution system of the reference system, TSI Incorporated for providing the particle number counter of the reference system, and Dekati Ltd and Horiba for providing particle number systems at the labs that did not have their own. AECC receive thanks for providing the reference vehicle, Concawe for providing the fuel and the lubricant for the reference vehicle and AEA Technology Environment for the calibrations of the reference system. The authors would also like to thank Dr. Nikolaos Stilianakis for his insights on the statistical issues.Address correspondence to Jon Andersson, Ricardo UK, Chemistry Department, Shoreham Technical Centre, BN43 5FG, Shoreham-bySea, U.K. E-mail: Jon.Andersson@ricardo.com
INTRODUCTIONThe established method to measure particle emissions for type approval tests is gravimetric analysis of filter samples, taken from a full exhaust flow dilution tunnel. However, for low-emission vehicles, which are already present in the market, concerns have been raised about the suitability of the method. For example Chase et al. (2004) showed that a major part of the collected mass consists of volatiles (volatile artifact). Tests at various laboratories have shown high variability for th...