EVALUATION OF GROUNDWATER FLOW AND TRANSPORT CHARACTERISTICS OF THE MATRIX OF LIMESTONE

Limestone aquifers serve as the primary water supply for roughly 1/4 of the US population. Unfortunately, these aquifers have been widely contaminated with chlorinated solvents and other organic chemicals. While the rapid spread of these contaminants primarily occurs through fractures within the limestone, the vast majority of the volume of the limestone is composed of the rock matrix. The matrix is highly porous (normally 5 to 30%), but is also relatively low in permeability (K ~10-3 to 10-7cm/s). The distribution of organic matter and clay within the matrix, which are also highly variable within the matrix, have been seen to retard the movement of these contaminants. In addition, the process of drilling to obtain cores often compromises the ability to measure the concentration of these contaminants within the cores due to volatilization. Thus, at many sites the fate and transport properties of these chemicals in the dissolved and non-aqueous phases within the matrix of limestone is not well established. This situation induces a high degree of uncertainty in the predictions of the amount and distribution of contamination, which inhibits the ability to optimize remediation strategies.

As a case study a series of unfractured cores from the chlorinated solvent contaminated limestone aquifer at Camp Stanley Army Base located in San Antonio, Texas, were examined for effective porosity, water saturated hydraulic conductivity, fraction organic carbon, and clay content. Preliminary results show that the effective porosity varied between 11 and 22% and water saturated hydraulic conductivity varied between 10-5 to 10-6cm/s. Secondary experiments using the core holder were performed to evaluate and compare the flow and transport properties of common organic contaminants and groundwater tracers. The results of the tests were analyzed with the flow and contaminant transport models MODFLOW and MT3D. In the model simulations a range of primary and secondary porosity and linear and non-linear sorption were examined. These simulations showed that a dual porosity model most accurately reflects the system. After thirty cores from the site were characterized the distribution of those properties was plotted and analyzed using geostatistics.