Comparability of particle emission measurements between vehicle testing laboratories: a long way to go

The characterization of automotive exhaust aerosol today involves determination not only of the mass emission rate but additional particle properties such as number and surface concentration and size distribution. However, despite the fact that several measurement campaigns are conducted world-wide utilizing various sampling systems and measurement devices, there is no consistent information on the uncertainty range for such measurements. This paper presents the within laboratory and between laboratories variability expected for measurements of exhaust particle properties, processing the data of several instruments and driving cycles collected in a specifically designed interlaboratory study. Results show that within laboratory variability is in the range of 20% of the mean value, excluding outliers. However, between laboratories variability can regularly reach or exceed 50% when small deviations from the sampling protocol occur, bringing significant implications to the comparability of the results. However, continuation of the study demonstrated that it is possible to decrease the between labs uncertainty range below 20% when laboratories get more experienced in the strict application of the measurement protocol. The study concludes that the comparability of exhaust aerosol measurements can only be verified with enforcement and strict application of common sampling, measurement and calibration protocols.