Vacuum extraction of a nonaqueous phase residual in a heterogeneous vadose zone

Abstract

The fate of an infiltrating nonaqueous phase liquid (NAPL) in a heterogeneous vadose zone was evaluated under laboratory conditions. Vapor extraction was then used to remove the contaminant, and the rate of removal was compared with the predictions of simple mathematical models. The laboratory experiments were conducted in idealized soil systems in which a fine-grained, low permeability lens insert was located in a coarse-grained sand layer. The infiltration study indicated that an essentially saturated layer of NAPL was trapped at the base of a fine-grained lens. A model of this process showed that capillary forces were responsible for the observed retention. Entrapment of NAPL within the fine-grained lens was expected to reduce the effectiveness of any in-situ extractive technology such as soil vacuum extraction (SVE) for the removal of the NAPL. Simple models of the mass transfer from the lens-entrapped NAPL to an extracting vapor in the coarse sand layer were developed and applied to laboratory experiments in which toluene, either pure or in a mixture, was injected within an idealized low permeability lens. Mass transfer resistances in the extracting vapor-phase controlled the release of pure toluene from the fine lens at low air flowrate (Reynolds number, Re ≤ 0.5) while intra-lens diffusion resistances were dominant at higher air flow rate (Re ≥ 3). Liquid phase resistances were always important in the extraction of toluene from a NAPL mixture. The models were found to accurately predict the removal rate of the toluene from the low permeability lens under each of the experimental conditions.