DETERMINATION OF EFFECTIVE HYDROLOGIC PARAMETERS USING HIGH RESOLUTION EXPERIMENTAL STRATIGRAPHY

A geostatistical analysis is conducted to obtain the statistical parameters of natural log hydraulic conductivity using a high resolution image of sediment architecture from the Experimental Earthscape Facility, University of Minnesota. A parallel finite element groundwater flow model is developed using Metis for mesh partitioning and Aztec for parallel iterative solution; solute transport is solved via random walk particle tracking, also parallelized. A finite element grid is generated using Lagrit to represent the conductivity map with 424,217 nodes and 845,208 elements. A geological framework model is developed based on stratigraphic analysis; numerical, analytical and stochastic methods are used to estimate the effective conductivity (K*) and macro-dispersivity (A) in flow/transport simulations. Using up-scaling, the “scale-effect” in K* is also addressed. We conclude that stochastic model prediction of K* becomes more accurate when a local statistical homogeneity is identified. K* exhibits “scale effect” due to heterogeneity; asymptotic values may be reached after data support exceeds the correlation length. This implies that K* for a hydrostratigraphic layer may only be obtained based on large-scale tests that incorporate flow effects on the order of the correlation length. Simple averaging based on core-scale measurements can lead to severe errors. For chosen boundary conditions, the hydraulic head of the framework model based on K* computed with periodic boundary conditions gives the best approximation to the reference heads computed with all conductivity information. Numerical tracer tests in the heterogeneous model indicate that A depends on the initial plume size/shape and the magnitude of local dispersion. Within a basin with multi-scale conductivity heterogeneity, the groundwater velocity is statistically non-stationary, accordingly, macro-dispersivity experiences fluctuations indicating alternate plume expansion and contraction during its migration.