Method Dependency in the Measurement of BTEX Levels in Contaminated Soil

Abstract

Environmental consultants and analytical laboratories are increasingly realizing the importance of analytical method selection and validation for the reliable measurement of soil contamination. Aromatic hydrocarbons (BTEX) is one class of environmentally significant soil contamination for which testing is required by regulatory authorities in most industrialized countries. Unfortunately, in most guidelines there is a lack of direction for the selection of a testing method for BTEX and, in practice, any one of a variety of methods may be employed. The fact that these may not all yield comparable results is a source of much frustration in the industry and there is a paucity of published research into this problem. In a number of carefully controlled experiments, the method dependency of measured BTEX levels in contaminated soil samples has been investigated. Three commonly employed methods, namely, (methanol extraction) purge-and-trap/gas chromatography with mass selective detection (P&T/GCMSD), headspace/GCMSD and dichloromethane (DCM) extraction/gas chromatography with flame ionization detection (GCFID) have been compared in the analysis of 109 gasoline-contaminated soil samples collected from station sites in Melbourne, Australia. Measurable BTEX concentrations were recorded in 92% of the samples using P&T/GCMSD, 59% using DCM/GCFID, and 40% using headspace/GCMSD. Correspondingly, the magnitudes of the recovered concentrations were significantly higher by P&T/GCMSD than by DCM/GCFID, which in turn were significantly higher than the magnitudes determined by headspace/GCMSD. These trends are evident for both clay and sandy soils. These studies clearly demonstrate that, for the three commonly employed methods described, measured BTEX levels are extraction and analytical method dependent in at least two different soil types.