Fifteen years of proficiency testing of total petrol hydrocarbon determination in soil: a story of success

Abstract

The total petrol hydrocarbon (TPH) content in soil is determined by gas chromatographic separation and flame ionisation detection according to ISO 16703 in routine laboratories for about 20 years. The development of the interlaboratory variability observed with this analytical procedure over 15 years in a proficiency testing scheme conducted annually with more than 170 participants is evaluated in detail. A significant improvement of the reproducibility standard deviation among participants is observed over the years and attributed to an increasing familiarity with the procedure. Nevertheless, the determination of TPH in the environmentally relevant mass fraction range between 500 mg/kg and 10 000 mg/kg in soils or sediments is far from reaching the reproducibility standard deviations predicted by the Horwitz curve. It is seen that laboratories with sporadic participation tend to report higher bias, while a core group of laboratories participating on a regular basis arrived at reproducibility standard deviations below 20 %. Results from a given laboratory obtained on two different samples tend to be highly correlated in the same PT round indicating a sound repeatability. Expectedly, the within-laboratory correlation between results from consecutive rounds was considerably lower. However, results from consecutive rounds with a temporal distance of 1, 2 or 3 years revealed largely similar correlations which suggests that the within-laboratory reproducibility adjusts to a constant level at least after a period of 1 year.