Estimation of uncertainty in the sampling and analysis of polychlorinated biphenyls and polycyclic aromatic hydrocarbons from contaminated soil in Brighton, UK

Abstract

The heterogeneity of environmental samples is increasingly recognised, yet rarely examined in organic contamination investigations. In this study soil samples from an ex-landfill site in Brighton, UK were analysed for polychlorinated biphenyl (PCB) and polycyclic aromatic hydrocarbon (PAH) contamination by using a balanced sampling protocol. The analytical technique of gas chromatography–mass spectrometry was found to be fit for purpose by the use of duplicate samples and the statistical analysis of variances, as well as of certified reference materials. The sampling uncertainty was found to significantly overweigh the analytical uncertainty, by a factor of 3 and 6 for PCBs and PAHs, respectively. The soil samples showed a general trend of PCB concentration that was under the recommended target level of 20ng/g dry weight. It is possible that one site alongside the main road may exceed the 20ng/g target level, after taking into consideration the overall measurement uncertainty (70.8%). The PAH contamination was more severe, with seven sites potentially exceeding the effect-range medium concentrations. The soil samples with relatively high PCB and PAH concentrations were all taken from the grass verge, which also had the highest soil organic carbon content. The measurement uncertainty which was largely due to sampling can be reduced by sampling at a high resolution spacing of 17m, which is recommended in future field investigations of soil organic contamination.