A fluid identification method for carbonate reservoirs based on anisotropic media theory and Bayesian EM algorithm

Reservoir fluid identification plays a crucial role in the exploration and development of carbonate reservoirs, including hydraulic fracturing and sweet spot prediction. Carbonate reservoirs with aligned fractures typically exhibit anisotropy, and such medium complexity further increases the difficulty of fluid identification. Starting from anisotropy theory, this paper parameterizes carbonate reservoirs as transversely isotropic media containing a set of vertical fractures with a horizontal symmetry axis. Based on this, an anisotropic reflection coefficient equation is derived, including the fluid indicator, quasi-fracture tangential weakness, and quasi-fracture normal weakness. Within the Bayesian theoretical framework, the expectation-maximization (EM) algorithm is integrated to construct a fluid identification method for carbonate reservoirs. Inversion results of synthetic and field seismic data demonstrate small errors compared with log data, which corroborates the accuracy and reliability of the proposed method for carbonate reservoir fluid detection.