Five-Dimensional Seismic Anisotropic Inversion Method for Reservoirs with Tilted Fractures

Fractured reservoirs are widespread in unconventional hydrocarbon plays, and the geometry and orientation of fracture networks directly control fluid flow and productivity. To enhance prediction accuracy in reservoirs containing a single set of inclined fractures, high-precision anisotropic inversion methods are urgently needed. This study targets complex reservoirs dominated by one set of inclined fractures and develops a five-dimensional (5D) seismic anisotropic inversion workflow to improve the accuracy and stability of fracture-related parameter estimation. First, we employ the Schoenberg linear-slip model to construct an equivalent tilted transverse isotropic (TTI) medium for single inclined fractures and derive its approximate stiffness matrix under weak-anisotropy assumptions. Next, using the Bond transformation, we formulate a linearized azimuthal reflection-coefficient equation for TTI media that explicitly relates P-wave modulus, shear modulus, density, fracture normal/tangential compliance, and fracture dip angle. Building on this theoretical foundation, we design a multi-stage, time–frequency inversion scheme within a Bayesian framework, integrating parameter-correlation constraints and low-frequency regularization to bolster robustness against noise and nonuniqueness. We apply this workflow to well-constrained, optimized 5D seismic gathers and validate it via synthetic tests and field data. Results show that our method accurately recovers the anisotropic parameters of inclined fractures, with inversion outputs in excellent agreement with drilling and logging measurements. The proposed approach significantly improves prediction accuracy and reliability in complex fractured reservoirs, offering a solid theoretical and technical basis for future exploration and development.