Zero-order Fourier coefficient inversion of azimuthal seismic data for fluid detection in fractured reservoirs

Fractured reservoirs play a critical role in unconventional hydrocarbon development. However, the complex spatial distribution of underground fractures induces strong reservoir heterogeneity, posing significant challenges to fluid prediction and hindering efficient hydrocarbon development. To address this, we propose a fluid prediction method based on zero-order Fourier coefficient inversion (ZFCI) of azimuthal seismic data, which integrates spatial coordinates, azimuth, and offset and provides more comprehensive information than conventional poststack seismic data for fractured reservoir characterization. Initially, we develop a horizontally transverse isotropic-oriented rock-physics workflow characterizing vertically aligned fractures within fluid-saturated reservoirs, formulating the anisotropic Gassmann fluid term (AGF) as the key indicator. Subsequently, through Fourier series expansion with respect to azimuth, we derive a zero-order term of the P-wave reflection coefficient containing the AGF, thereby enabling robust fluid prediction that is independent of prior fracture symmetry axis azimuth. Finally, a Bayesian inversion framework is constructed, integrating Cauchy prior constraints and low-frequency model regularization to achieve stable and accurate fluid indicator inversion. Single-well tests demonstrate that the established rock physics model achieves high accuracy in estimating S-wave velocity and fracture weakness parameters, with the fluid indicator effectively delineating reservoir hydrocarbon. Critically, the theoretical model comparison demonstrates that the proposed ZFCI method, despite requiring no prior fracture symmetry axis azimuth, achieves fluid indicator accuracy that matches conventional azimuth-dependent inversion when supplied with the correct prior azimuth. Field data applications further confirm the method’s robustness and consistency with well-log interpretation. This work provides a novel and practical approach for high-precision fluid prediction in frontier fractured reservoirs.