Fracture Characterization Using Multicomponent Elastic Waves

Fractures impact reservoir management, drilling, and well completion. Therefore, we characterize fractures systems to understand subsurface stress and flow fields. I studied four topics on fracture characterization for both applied and earthquake seismology: (1) the effect of fractured reservoir layer thickness on reflected P wave amplitudes; (2) the effect of fractured reservoir layer thickness on shear-wave splitting; (3) the effect of normal and tangential fracture compliances on anisotropy and shear wave splitting; and (4) the effect of incidence angle and subducting slab dipping angle on the transmitted seismic wavefield. I considered a 3-layer model in which the middle layer is anisotropic. I applied full-wave modeling based on plane wave expansion in anisotropic media. The modeling includes all reflections and mode conversions. In topic 1, I studied how layer thickness affects the reflected P wave amplitude variation versus azimuth caused by an incident P plane wave in an HTI fractured layer. I found that to accurately extract fracture orientation and density using azimuthal P-wave traveltime as a function of azimuth, the thickness of the fractured reservoir layer cannot be less than the P wavelength. In topic 2, I found that the splitting time for reflected shear waves is related to the thickness of the fractured layer. To observe splitting from seismic data, the layer thickness needs to be ~5 times the S wavelength. In topic 3, I used linear-slip boundary conditions to represent fractures. I found that shear wave splitting is only affected by the tangential compliance, Z_T, because the Thomsen’s parameter, γ, only relates to tangential compliance, Z_T, and not the normal compliance, Z_N. From my modeling, I find it difficult to observe shear wave splitting from synthetic seismic data if the Z_N/Z_T is low (<0.4) even if the fracture thickness is thick enough. In topic 4, I extracted polarization, ϕ, of the fast S wave and the time delay, dt, between the fast and slow S waves measured from seismic records to characterize the possible anisotropy in subducted slabs. I found that ϕ and dt are related to the anisotropic property, the slab dipping and incidence angles.