South-Central Section–40th Annual Meeting (6–7 March 2006)

Paper No. 4
Presentation Time: 8:20 AM-12:00 PM

CHARACTERIZATION OF ROUGHNESS IN FRACTURES USING COMPUTED TOMOGRAPHY


ROBERTSON, Wendy Marie, Geological Sciences, University of Texas at Austin, Department of Geological Sciences, 1 University Station C1100, Austin, TX 78712-0254, wendyr@mail.utexas.edu

Characterization of fluid flow through fractures is an important aspect of hydrogeology because up to 80% of the world's aquifers have significant fracture-dominated flow. One important aspect of flow in fractures is how the asperities along fracture walls affect the fluid flow to complicate the classic cubic law model. Studies of fluid flow through fracture systems could become more precise, effective, and cost efficient if a model could be made of these affects by using a relatively small rock sample. In this study, 3 granite and 2 tuff samples, sizes approximately ten centimeters wide by fifteen centimeters long and five to seven centimeters thick were analyzed for asperities and aperture changes along their fracture faces by using three dimensional analyses of X-ray computed tomography (CT) scans. The surface analysis was placed into a computerized flow modeling system to predict the behavior of fluid through the fracture. The rock samples were subjected to flow tests using a unique apparatus designed to measure flow volumes and head differences through fracture samples. Flow tests were conducted with varying head differences and flow volumes and boundaries for laminar flow through the samples were determined. The computer model compared flow test data with simulations run in MODFLOW to determine the accuracy of a predictive flow study upon a small rock sample. Initial tests indicate that when the size of asperities in a fracture is within two orders of magnitude of aperture, it greatly alters the behavior of fluid flow in the fracture from the ideal parallel plate model to a much more complex model involving channels of flow with varying velocities. In future tests, the goal is to correlate patterns of roughness and aperture variations to channels with a known flow rates so that flow in a fracture can be estimated precisely with values for roughness coefficient and aperture.