Using Seismic Velocity Anisotropy to Predict Fractures: A Calibrated Case Study from the Green River Basin of Wyoming
Shanley, Keith W.1, J. Chris Besler2, William A. Miller3, John F.
1 The Discovery Group, Inc, Denver, CO
2 Stone Energy LLC, Denver, CO
3 Miller Consulting Services, Littleton, CO
4 Halliburton Energy Services, Denver, CO
Economic gas production from low-permeability reservoirs is generally thought to require the presence of natural fractures. Detection and mapping of fracture trends is difficult, however, it has been suggested that velocity anisotropy within both P-wave and converted-wave seismic data could be used to characterize these features. Fundamental to this process is the implicit assumption that in relatively undeformed strata velocity anisotropy primarily reflects variations in fracture density and orientation. Although the theoretical basis is well understood, there remain few calibrated data-sets that investigate the relationship between anisotropy and fractures.
A wide-azimuth 3D survey designed to facilitate anisotropy measurements was acquired across an area of 105 mi2 (269 km2) in the southern Green River Basin in southwest Wyoming. Maps over a range of depths show substantial vertical and lateral variation in the magnitude of velocity anisotropy and the orientation of the fast and slow velocity vectors suggesting strong variations in fracture density and orientation. Wells drilled within the survey area sampled a wide range of velocity anisotropy conditions. Several wells were logged with formation imaging tools so that fracture sets and stress orientations could be independently detected, measured, and their orientations quantified for purposes of calibrating the seismic data volume. Borehole fracture measurements were compared with corresponding seismic anisotropy data across specified intervals. Borehole data are remarkably uniform in terms of stress orientation, fracture orientation, and fracture density, in sharp contrast to the highly variable anisotropy data. A similar study conducted in a more structurally complex portion of the Basin yielded similar results. We conclude that the use of velocity anisotropy to infer the presence of fracture trends remains elusive and additional research is required to fully understand the controls on velocity anisotropy.