Explaining "anomalous" solute transport at the Macrodispersion Experiment (MADE) site from a geological perspective

Abstract:

Stochastic realizations of the spatial distribution of five lithofacies are used to simulate the tritium plume observed during second large-scale tracer experiment at the Macrodispersion Experiment (MADE) site. Lithofacies are identified from previously collected aquifer samples on the basis of relative percentages of gravel, sand and fines, and characteristic grain diameters. Geostatistical simulations of lithofacies assemblage are initially generated with the transition probability approach based on a Markov Chain model calibrated against borehole lithological data. Hydraulic conductivity (K) values, which are estimated with an empirical expression based on grain-size data, are then mapped according to the simulated distribution of lithofacies. With this approach, K fields in the numerical simulations are directly linked to the spatial distribution of the lithofacies. Since the model of physical heterogeneity takes into account sharp vertical K contrasts at the decimeter scale as well as the lateral connectivity of highly conductive sediments, we are able to accurately match the experimental data with a relatively simple transport model based on the advection-dispersion equation (ADE). Transport simulations show that one well-interconnected lithofacies, with a significantly higher K and accounting for 12% of the total volume, is responsible for the anomalous transport behavior indicated by the asymmetric shape of the plume and by variations of the dispersion rate in both space and time. Our analysis provides a lithological basis to the hypothesis that transport at MADE site is controlled by a network of high permeable sediments embedded in a less permeable matrix. It also explains the calibrated value of the ratio of mobile to total porosities used in previous models based on the dual-domain mass transfer approach. This study stresses the importance of geologically plausible conceptualizations of the subsurface for making accurate predictions on the fate of contaminants in highly heterogeneous aquifers. These conceptualizations may be developed through integration of raw geological data with expert knowledge, interpretation and geostatistical methods.