Fracture prediction based on attenuative anisotropy theory and its application to a shale gas reservoir

Shale gas reservoirs require artificial hydraulic fracturing to achieve commercial gas production. Continental Jurassic shale reservoirs have several thin interlayers (e.g. sandstone and carbonate), which lead to pseudo-anisotropy of the seismic amplitude and velocity, and traditional technology for predicting anisotropic fractures has difficulty in identifying a true target. Based on the traditional theory of azimuthal anisotropy and the equivalent medium theory, in this paper, the approximate formula for the reflection coefficient of attenuating anisotropy and the equation for the azimuthal attenuating elastic impedance are derived. With the objective function established using Bayesian and group sparse inversion methods, parameters such as the fluid terms and fracture characteristics are obtained by solving the elastic matrix, and a highly accurate prediction of the 3D fracture density is obtained. The validity of this method is confirmed via a test on a single well and application to actual seismic data for a shale gas reservoir.