ACCOUNTING FOR VERTICAL CONCENTRATION AVERAGING IN MONITORING WELLS WHEN USING THE DOMENICO EQUATION AND IMPLICATIONS FOR DETERMINING BIODEGRADATION RATES IN INVERSE MODELING

Domenico (1987) presented an advection-dispersion model that included first order decay based on an analytical model derived by Domenico and Robbins (1985). Domenico's model has become widely used in evaluations of risk associated with ground water contamination from gasoline spills. Robbins (1989) further modified the Domenico and Robbins equation to examine the influence of using purged and partially penetrating monitoring wells on contaminant detection, mapping and modeling. A major finding of that study was that vertical concentration averaging that takes place during purging can result in a mapped plume that differs substantially from the actual three-dimensional contaminant distribution. Hence, contaminant transport parameters, including dispersion coefficient values and retardation factors, derived from inverse modeling using models that do not account for concentration averaging can be highly inaccurate and result in poor transport predictions. In this study, a version of the Domenico model is presented that takes into account vertical concentration averaging. The model is derived by integration of the ERF terms in the Domenico model that account for vertical spreading. Results of sensitivity analyses suggest that not accounting for vertical concetration averaging when using the original Domenico model may result in greatly over estimating plume bioattenuation. Furthermore, estimating biodegradation rate constants using ratios of degrading to recalcitrant contaminants as a means to account for unknown transport parameters, such as dispersion, may be problematic if the vertical concentration distributions of the contaminants differ. The revised Domenico model provides a means to improve risk assessments using monitoring well data. However, if highly accurate plume evaluations and risk assessments are desired, one must fully account for the three-dimensional distribution of contamination. This requires using data derived from multilevel samplers as opposed to averaged concentration information obtained from typical monitoring wells.