Inversion of Shale Reservoir Elastic Properties Using Deep-Guided Waves in Vertical Transversely Isotropic Elastic Media: Methodology and Sensitivity Analysis

Abstract

The elastic properties of shale reservoirs are crucial for reservoir geomechanical modeling, fracture imaging, and fracturing scheme design. Traditional surface acquisition-based seismic inversion methods have low spatial resolution and cannot accurately delineate the characteristic properties of thinner reservoirs. The compressional (P-) and shear (S-) waves excited by a perforation shot or microearthquake propagate in the deep waveguide and are reflected on the top and bottom interfaces, constructively interfering to form the deep-guided wave. The deep-guided wave has high-frequency content and notable dispersive features. The dispersion represents kinematic information and can be used to image low-velocity structures. This article presents a gradient-based method for deep-guided wave dispersion inversion in anisotropic elastic media. We derive the analytical expressions of the sensitivity functions of dispersion curves with respect to density, P- and S-wave velocities, and Thomsen anisotropic parameters. The sensitivity functions can serve as gradients in gradient-based optimization methods for guided wave dispersion inversion. Compared with stochastic methods, the proposed gradient-based guided wave dispersion inversion method improves computational efficiency and provides a sensitivity analysis of dispersion curves with respect to different anisotropic elastic parameters. The latter is important for estimating the confidence and uncertainty of the inversion results. Numerical tests demonstrate that the proposed method can successfully reconstruct the target S-wave velocity profile in low-velocity layers. The proposed methodology provides a viable path toward real-time inversion and analysis of guided wave dispersion in unconventional reservoir monitoring.