Rock-Physics-Constrained Prestack Inversion Method Based on Fuzzy c-Means Clustering Strategy for Direct Prediction of Different Types of Effective Reservoirs

The prestack seismic inversion technique serves as a critical method for obtaining elastic parameters of subsurface sediments, as it further guides the lithologic prediction of reservoirs. As exploration targets become more complex, identifying sandstones with different porosity and permeability characteristics (i.e., different types of effective reservoirs) based on accurate predictions of reservoir lithology distribution has become a new research focus. However, the fact that different types of effective reservoirs are distributed in sandstones and tend to have different sensitive elastic parameters with respect to lithology prevents traditional direct inversion strategies from making direct predictions of targets. To address this limitation, we propose a rock-physics-constrained prestack inversion method integrating fuzzy c-means (FCM) clustering strategy to achieve simultaneous inversion of sensitive elastic parameters and direct identification of effective reservoir types. First, sensitive elastic parameters for lithology discrimination and effective reservoir characterization were determined through rock-physics analysis. The results revealed that the P-wave to S-wave velocity ratio (Vp/Vs) effectively discriminates lithology, while P-wave impedance (Ip) can indicate different types of effective reservoirs. Second, we derived approximate reflection coefficient equations for Vp/Vs, Ip, and density characterization using a generalized elastic impedance equation, which reduces conversion errors of elastic parameters and enhances computational efficiency. Finally, accurate estimation of elastic parameters coupled with direct reservoir characterization was achieved through the developed rock-physics-constrained prestack inversion method incorporating FCM clustering strategy. Both synthetic model validation and field case studies demonstrate the method’s effectiveness, providing a feasible solution for predicting different types of effective reservoirs.