T191 Geologic Controls on Chemical Migration in Fractured and Carbonate Aquifers I

Wednesday, 31 October 2007
8:00 AM - 12:00 PM

MODELING OF HYDRAULIC RESPONSE AND TRACER MIGRATION IN A FRACTURED-CARBONATE AQUIFER

Paper number: T191-9
Program: Technical Sessions
Time: 10:05 AM – 10:20 AM

Authors

Abstract

Fractured-carbonate aquifers provide important, but vulnerable, sources of groundwater. The dual-porosity nature of these aquifers-- high-permeability fractures transmit the majority of the water while the lower-permeability matrix blocks provide the storage capacity -- make them exceedingly susceptible to contamination, challenging to characterize, and difficult to model. Management of these vulnerable aquifers sometimes requires the use of numerical models for various purposes including wellhead protection, design of monitoring systems, and remediation of contaminated sites. Existing numerical codes range from traditional porous-medium models to discrete fractured network models. A “hybrid” modeling approach, in which selected high-permeability fractures are simulated as discrete features in a porous-medium model, has also been used by several researchers. Currently, there is limited guidance as to which modeling approaches are appropriate for predicting chemical migration in fractured-carbonate aquifers.

Detailed hydrogeologic characterization of a small study site (40 x 25 m) located in a dolomite quarry in northeastern Wisconsin and the results of several tracer experiments at the same site provide a comprehensive data set that we have used to evaluate various approaches for numerical modeling of groundwater flow in fractured carbonates. A regional analytical element model was used to determine the boundary conditions for the site-specific model of the study area. A traditional porous-medium model, MODFLOW was successfully used to simulate the results of an “open hole” pumping test. While the traditional porous-medium approach was adequate to simulate the hydraulic response of the aquifer, it proved inadequate to predict transport of a conservative tracer even when hydraulically-significant bedding-plane fractures were incorporated into the model. We are currently simulating the results of one of the natural-gradient tests using a fracture-network approach (FracMan).