抄録
PCE and TCE saturation distributions are generated with stochastic models to incorporate spatially varying aquifer properties. The influence of aquifer property correlation on fluid migration and entrapment is explored in cross-sectional 2D numerical two-phase saturated flow models extracted from 3D geostatistical realizations generated from well-published aquifer data in a nonuniform sandy aquifer. An effort to model a mixture of DNAPL is exercised by employing five compounds DNAPL; a 6:1:1:1:1 by volume ratio of PCE, toluene, 2-chlorotoluene, 1-bromohexane, and 1-bromoheptane. Comparisons of PCE, TCE, and DNAPL mixture suggest that the choices involving chemical waste compositions greatly influenced the saturation and distribution of DNAPL (i.e., pathways and organic spreading). The organic wastes released into the system can escape to the other 2D cross sections as the waste mixtures are more mobile compared to pure phase. Coupled application of stochastic model of aquifer spatial variability as well as chemical waste compositions can significantly influence predicted DNAPL infiltration depth, entrapment, and recovery. The resultant distribution profiles of DNAPL mass within the source zone also have implications for DNAPL recovery and subsequent downstream mass fluxes in remediation operations.
