MEL-MESL: Intercomparison Exercise IAEA-417
WORLD-WIDE INTERCOMPARISON ON THE DETERMINATION OF CHLORINATED PESTICIDES, PCBs AND PETROLEUM HYDROCARBONS IN A SEDIMENT SAMPLE IAEA-417
Introduction
The assessment of environmental contamination requires accurate analytical data on pollutant concentrations in the environmental components. For this purpose the analytical methods used by the laboratories need to be validated and tested. This is commonly achieved through the analyses of certified reference materials and participation in interlaboratory analytical comparison exercises. Interlaboratory comparison are also essential to check the accuracy of the analytical results and, furthermore, are a means to stimulate for better analytical performance.
The Marine Environmental Studies Laboratory (MESL) of IAEA-MEL has conducted intercomparison exercises on trace organic compounds for over twenty years as part of its contribution to IAEA's Analytical Quality Control Service, UNEP's Regional Seas Programme, and, occasionally, in association with the Intergovernmental Oceanographic Commission (of UNESCO) GIPME (Global Investigation of the Pollution in the Marine Environment) programme. These exercises are organized on a continuous basis, once per year, and the participation of the laboratories is cost free.
Organization of the intercomparison exercise
Scope of the intercomparison
The scope of the exercise was to compare the determination of organochlorine compounds, petroleum hydrocarbons and sterols in a marine sediment sample simultaneously analysed in worldwide laboratories. In the organization of this exercise, as well as in similar ones organized by the IAEA-MEL, the recommendations of internationally harmonized protocols were followed.
Preparation of the sample
A large sediment sample was collected from the lagoon of Venezia, Italy, for use as an intercomparison material. This sediment was deep-frozen, freeze-dried, ground and sieved through a 250 µm stainless steel sieve. The sediment fraction of particle size less than 250 µm was further homogenized by mixing in a stainless steel rotating drum for two weeks. Then, aliquots of about 40 g were packaged into glass bottles with aluminium screw caps and sealed with Teflon tape.
The homogeneity of the material for organochlorine compounds, petroleum hydrocarbons and sterols was assessed by determining the concentration of some compounds (chlorinated pesticides, petroleum hydrocarbons and sterols) in 10 sample aliquots taken randomly in the bulk of the powder. A one way variance analysis of the results indicated that the material could be considered as homogeneous.
Sample dispatch and data return
In January 2001, 206 bottles of sediment labeled IAEA-417 were distributed to laboratories world-wide along with a sample description and general instructions to analysts. Results were received until the end of December 2001, date of closure of the exercise.
Results of analyses were reported by 97 laboratories of which, 36 provided data for both chlorinated and petroleum hydrocarbons, 42 provided results for organochlorine compounds only and 19 laboratories provided data on petroleum hydrocarbons only. Only 4 laboratories analyzed the sample for sterols.
Statistical evaluation of the results
Results reported by the participants were extensively compiled in a report distributed to all participants and available upon request from the IAEA-MEL1. In the report, a compilation of analytical methods used, from pretreatment of the sample to gas chromatographic conditions, was also included. Each laboratory is represented by a confidential code number, so that the anonymity of each result is preserved.
For the statistical evaluation of the data, the laboratory results were organized in data sets, and the range of results, the arithmetic mean of all individual results and the standard deviation (n-1) were computed for each compounds. Outliers were then removed using the Box-and-Whisker plot of Statgraphics 5.0, and the range, mean and standard deviation recalculated. The overall non-parametric median and the corresponding 95 % confidence intervals of the median were computed for each compound. Results obtained by individual laboratories can thus be easily compared with the median values and confidence intervals.
Results and discussion
Organochlorine compounds
In total, 78 laboratories reported data for organochlorine compounds, of which 53 also reported data for PCBs. Figure 1 shows the results for a selection of organochlorine pesticides.
Petroleum hydrocarbons
In total, 55 laboratories provided data for petroleum hydrocarbons.
Conclusion
The overall results of this intercomparison exercise indicate that there are still widespread difficulties with the determination of many persistent organic pollutants. The most probable cause is because many laboratories worldwide did not adopt Quality Assurance/Quality Control (QA/QC) procedures as yet. Nevertheless in some regions, specific programmes have been successful with the training of analysts and promoting adoption of QA/QC. This is the case for the Mediterranean region, where 48 laboratories from 14 countries reported results for the sample IAEA-417. This high participation reflects the increasing attention to QA/QC in this region, which has been backed by the UNEP Mediterranean Action Plan.
Intercomparison exercises represent only one component of data quality assurance and provide only sporadic indicators of data quality. The most valuable approach to quality assurance is through the regular analysis of reference material and by plotting the resulting data on quality control charts. This provides a continuous feed-back to the analyst and is an essential tool for monitoring the day-to-day quality of data and assuring accurate and comparable results in future exercises.
As a further outcome of this worldwide intercomparison, reference concentration values could be assigned to the sediment sample. This sample is suitable to check the accuracy of analytical measurements when marine sediments from medium polluted area are analyzed for petroleum hydrocarbons and chlorinated pesticides. Furthermore, the sediment sample IAEA-417 will extend the range of concentrations and matrices available for quality control purposes.
Reference
Villeneuve J.-P., de Mora S.J., Cattini C., Worldwide and regional intercomparison for the determination of organochlorine compounds and petroleum hydrocarbons in sediment sample IAEA-417, International Atomic Energy Agency, Vienna, Report IAEA/AL/131 (2002).
Figure 1 Concentration values reported by individual laboratories (symbols) and the assigned reference value (solid line) and the 95 % confidence interval (dotted lines) for selected chlorinated pesticides in IAEA-417 sample. UNEP's Regional Seas Programmes are identified with different symbols. Outlier data have been removed from graphical presentation.
