Dielectric Response of a Variable Saturated Soil Contaminated by Non-Aqueous Phase Liquids (NAPLs)

In recent years, several studies have been conducted both in saturated and unsaturated soils to detect non-aqueous phase liquid (NAPL) hydrocarbon contamination in soils and groundwater by means of the time domain reflectometry (TDR) technique. This technique is widely used for measuring the dielectric permittivity and bulk electrical conductivity of multiphase systems. Only accurate knowledge of the dielectric response of soil matrix- water-NAPL (saturated condition) or soil matrix-air-water-NAPL (unsaturated condition) systems can allow the volumetric NAPL content (θNAPL) to be determined in the soil. This paper investigates the influence of NAPL contamination (corn oil, a non-volatile and non-toxic NAPL, was used) on TDR measurement in a volcanic soil, relating dielectric permittivity of the multiphase soil system to volumetric fluid content θf (i.e. water+NAPL). The soil samples were oven dried at 105 °C and passed through a 2 mm sieve. Known quantities of soil, water and oil were mixed and repacked into plastic cylinders (15 cm high and 9.5 cm in diameter); 40 different combinations of water and oil were tested, with θNAPL varying from 0.05 to 0.40 by 0.05 cm3/cm3 increments. A volumetric mixing model with three (soil matrix-water-NAPL) or four (soil matrix-air-water-NAPL) phases permitted conversion from a dielectric permittivity domain into a θf domain. The results show that, the amount of contaminant in soil can be inferred if the total volume of pore fluid θf and the dielectric permittivity of the contaminated soil are known. Further work will be built on this initial study, concentrating on: i) enhancing the model linkage and validating it with new laboratory results; ii) validating the developed TDR interpretation tool with field results.