The measurement of the diffusion rates of volatile organic compounds through laboratory-prepared and intact soils.

Abstract

A device has been built, refined, and utilized to measure the fluxes of volatile organic compounds through laboratory-prepared soil cores and intact cores. The effective diffusion coefficients for VOCs in soils can be determined from steady state fluxes using Fick's First Law. Sorption characteristics can be estimated using kinetic information. The two compounds studied were trichloroethylene and benzene; the soil media tested were sand and glass beads at moisture contents ranging from 0 to 80% of water saturation. The temperatures tested were 10 to 25$\sp\circ$C. Effective diffusion coefficients for each compound in soils with moisture contents of 20 to 50% closely matched the diffusion coefficients calculated with the Millington-Quirk equation. At higher moisture contents (e.g., 60-80%) the measured coefficients were 32.1 to 93.0% lower than the calculated coefficients, indicating the lack of applicability of the Millington-Quirk equation at high moisture contents. Completely dry media showed possible mass-transfer limitations as the measured coefficients were 3 to 19% lower than the calculated ones. Flux curves were modeled using the EPA's One Dimensional Finite Difference Vadose Zone Leaching Model (VLEACH). Using the experimentally determined soil porosity ($\theta$), air-filled porosity ($\rm\theta\sb{g}$), and moisture content, the model was matched to the non-steady-state flux. With this fitting, the model was able to accurately predict the steady-state flux for that experiment. It was then shown that an experiment need not run to steady-state to estimate the final steady-state flux. This held for a variety of soil moisture contents and a variety of temperatures. Intact cores from a potentially contaminated site were examined but the results were unusable because the cores were irregular, had pulled away from the cylinder walls, and had large cracks, all indicative of soil moisture loss or other damage. The device nonetheless reached steady-state, indicating that for more well defined soil cores, the device would probably work as well as it did for laboratory-prepared cores. Improvements in methods and materials are suggested that would lead to more accurate results and thus better agreement between experimental and calculated results. Future possibilities for research with this device are the analysis of more than one compound at a time, having several devices doing analyses simultaneously, and obtaining a soil's characteristics on the basis of the non-steady-state and steady-state behavior of a well-characterized compound as it diffuses through that soil.