Assessment of soil pollution based on total petroleum hydrocarbons and individual oil substances

Different oil products like gasoline, diesel or heavy oils can cause soil contamination. The assessment of soils exposed to oil products can be conducted through the comparison between a measured concentration and an intervention value (IV). Several national policies include the IV based on the so called total petroleum hydrocarbons (TPH) measure. However, the TPH assessment does not indicate the individual substances that may produce contamination. The soil quality assessment can be improved by including common hazardous compounds as polycyclic aromatic hydrocarbons (PAHs) and aromatic volatile hydrocarbons like benzene, toluene, ethylbenzene and xylenes (BTEX). This study, focused on 62 samples collected from different sites throughout The Netherlands, evaluates TPH, PAH and BTEX concentrations in soils. Several indices of pollution are defined for the assessment of individual variables (TPH, PAH, B, T, E, and X) and multivariables (MV, BTEX), allowing us to group the pollutants and simplify the methodology. TPH and PAH concentrations above the IV are mainly found in medium and heavy oil products such as diesel and heavy oil. On the other hand, unacceptable BTEX concentrations are reached in soils contaminated with gasoline and kerosene. The TPH assessment suggests the need for further action to include lighter products. The application of multivariable indices allows us to include these products in the soil quality assessment without changing the IV for TPH. This work provides useful information about the soil quality assessment methodology of oil products in soils, focussing the analysis into the substances that mainly cause the risk. •Management of soils affected by oil products requires appropriate soil assessment.•Soil contamination by oil products can be assessed through TPH concentration.•Evaluation of PAH and BTEX concentrations improves soil quality assessment.•Multivariable analysis simplifies soil assessment by reducing the number of variables.