Application of anisotropic frequency-dependent inversion to shale reservoirs in the Zhaotong Demonstration Area, Sichuan Basin

Abstract The accurate characterization of fracture density and gas saturation is crucial for successful shale gas reservoirs developments. We present an integrated approach that combines rock physics modeling with anisotropic frequency-dependent inversion to characterize shale reservoirs in the Zhaotong Demonstration Area, Sichuan Basin. First, based on drilling, logging, and geological data, a shale model was constructed using self-consistent approximation (SCA), differential equivalent medium (DEM), and Chapman fracture models. The model was used to study the effects of fracture density and gas saturation on the dispersion and attenuation of P-wave velocity, S-wave velocity, and the anisotropic parameters ε and δ. Based on this analysis, a rock physics interpretation template was constructed to provide a theoretical foundation for the anisotropic frequency-dependent inversion. Parameter sensitivity analysis indicated that P-wave and S-wave velocity dispersions are sensitive to fracture density, whereas the dispersions of anisotropy parameters ε and δ are influenced by gas saturation. Time-frequency analysis of the near-, mid-, and far-angle gathers using inverse spectral decomposition based on the generalized S-transform was performed to obtain data volumes at different frequencies. By establishing an objective function based on L1 norm regularization using the Rüger approximation formula, the spectral projection gradient for L1 minimization (SPGL1) was employed to solve this objective function, thereby obtaining the dispersion properties. Frequency-dependent AVO inversion was subsequently applied using the derived R ger approximation equation to extract dispersion attributes. In addition, the fracture density and gas saturation were predicted using a rock physics template and validated against well-log data to prove the effectiveness of the proposed method.