Seismic Anisotropy Dependence on Fluids, Fractures, and Stress: Physical Modeling with Bakken and Barnett Shale Field Cases

A major goal of geophysical research is to understand and predict the seismic response of fluids, fractures, and stress in the subsurface. In this dissertation, we explored different forward modeling and field techniques with the goal of quantifying seismic anisotropy and its relationship with subsurface states. This dissertation includes two series of physical modeling experiments as well as analyses of wide-azimuth 3D data from the Williston Basin in North Dakota and the Fort Worth Basin in Texas. In the first set of lab experiments, we explored fluid substitution effects in a synthetic anisotropic medium. We observed the effects of different fluids on wide-azimuth P-wave NMO ellipses from our synthetic composite rock sample. We find that fluid substitution from air to water can increase inherent anisotropies by as much as 50%. We also observe changes in P-wave NMO ellipses as a function of different fluid saturants in the synthetic sample. In the second set of experiments, we varied uniaxial normal stress and measured transit-time and its associated effects on a layered synthetic orthorhombic medium. The experiment was designed to measure the dynamic elastic properties of sedimentary reservoir rocks deposited in layers under stress. Results show a general increase in all measured velocities with stress ranging from 4% to 10%. We also observed anisotropic behavior a priori to both orthorhombic and VTI symmetries in different principal axes of the synthetic sample as uniaxial stress changes. 3D wide-azimuth data from the Ross Field of the Williston Basin in North Dakota is fully processed using conventional techniques. We presented a velocity-based workflow for inverting for the direction and intensity of preferred orientations within the subsurface. We demonstrate a potential for using wide-azimuth P-wave seismic data as a tool for subsurface characterization in shale reservoirs. Lastly, using fully processed wide-azimuth 3D dataset and wells from the Fort Worth Basin in Texas, we presented a workflow that integrated RMO analysis and azimuthally sectored inversions to generate a broad overview of subsurface orientation in the Barnett shale play.Earth and Atmospheric Sciences, Department o