Abstract

Reference measurements of particulate matter (PM) in ambient air are performed by sampling a known volume of air through a filter for 24 hours. By measuring the mass difference of the filter before and after sampling the concentration of particulate matter may be determined. Procedures for the performance of these measurements are described in European Standards; additional requirements for Dutch monitoring networks are laid down in a Netherlands Technical Agreement. In order to investigate whether application of this procedure - when applied by different networks - leads to comparable results, an inter-laboratory comparison has been performed. Seven monitoring networks from Belgium and The Netherlands have measured PM10 simultaneously from April to July 2012 at a monitoring site of RIVM. A total of 9 reference samplers has been used. After validation by each of the participants the measurement results have been sent to the pilot laboratory (RIVM) for evaluation. The evaluation has been performed using the international standards ISO 13528 and ISO 5725 part 2, resulting in performance indicators for each network but also for the reference method as such. The mean PM10 concentration level found during the comparison is 16,1 µg/m3, the range of concentrations is between 7,2 and 41,6 µg/m3. These rather low levels may be explained from prevailing typical Dutch weather conditions. The evaluations performed show that the results for all networks are generally in good agreement. On average, 96 per cent of all results have so-called z-scores, determined using ISO 13528 by applying a fit-for-purpose standard deviation of 8,3 per cent, that are good. Application of ISO 5725-2 statistics indicate one set of results to be a straggler for means and one to be a straggler for standard deviation. Both fin dings imply that all networks have an appropriate level of quality control for the performance of the measurements. When evaluating the method as a whole, the resulting reproducibility relative standard deviation is 6,8 per cent resulting in an expanded measurement uncertainty of 2,5 µg/m3. This is within the uncertainty calculated in prEN 12341:2012 by adding contributions derived from method performance requirements, showing that it is possible overall to meet these requirements. Lastly, a number of permutations of the implementation of the reference method have been compared. It is noted, however, that the results of these comparisons are to be treated with care because of the low number of samples within single permutations.