Proficiency testing schemes in Europe have shown that there was a need to improve the quality of the measurements to assess exposure to harmful chemicals in air at the workplace. Efforts were required to provide calibrants and CRMs for quality control, but also to improve the performance in sampling..
A collaborative European project has been initiated under the BCR programme to study the performance of personal samplers most widely used for health related sampling of workplace air. In view of the implementation of European Standard EN 482, which outlines minimum performance criteria for personal samplers, bias and precision of methods applied by participating laboratories were assessed by means of intercomparisons.
The primary objective was to study and quantify errors of the overall measurement procedure hereby giving priority to the identification of errors associated to the sampling step. This evaluation would contribute to the improvement of methods and to the promotion of good laboratory practice, and would accelerate the elimation of methods leading to substantial measurement errors. Both pumped and diffusive sampling methods had to be included as well as GC and HPLC analysis after solvent as well as thermal desorption.
The design of the experiments allowed to verify compliance of method-laboratory combinations to EN 482. For most aromatic and chlorinated hydrocarbons, only 10 % of results did not fulfil the overall criterion of < 30% uncertainty. For other organic groups and analytes this was substantially different. The most extreme example was acrolein where from 50 to 80 % of the method-laboratory combinations did not comply to the uncertainty criterion of 30 %.
The project has gained much information on sampling and analysis parameters like breakthrough volumes, uptake rates, desorption efficiencies, sampling volumes, active versus diffusive sampling and others.
It was demonstrated and published that the most widely used method for aldehydes needs to be revised thoroughly. Variables in sampling conditions which may lead to errors were clearly identified and have resulted for polar compounds to guidelines for the selection of sorbents and desorption liquids.
Sampling exercises were organised for : a) aromatic hydrocarbons and chlorinated hydro-carbons, b) esters, ketones and ethers; c) aldehydes. The experimental work took place in a purpose built facility where the project partners could sample simultaneously from a precisely controlled synthetic workplace atmosphere. The concentration levels in the sampling manifold were typically set at a level of 0.5 - 2 times the Limit Value. The duration of each sampling experiment, the choice and the type of analytes, the water vapour content and the interfering compounds were modified in order to determine their influence on the overall uncertainty of the methods used. Participants were able to test any personal sampling system they wished.