Seismic fluid identification is the main goal of current prestack seismic inversion. Various kinds of fluid indicators are used for fluid detection in industry today. However, the existing methods cannot always provide reliable fluid prediction owing to the insensitivity to fluid response and the lack of converted wave constraints. The equivalent fluid bulk modulus is an effective fluid factor based on matrix-fluid decoupling, which can provide persuasive evidence for fluid detection. Combining poroelasticity theory and matrix-fluid decoupling theory, we have deduced a new PS-wave linear amplitude versus offset approximation equation that provides estimations of equivalent fluid bulk modulus, rigidity, porosity, and density. Then, the joint inversion of PP- and PS-waves based on matrix-fluid decoupling was executed in a Bayesian framework with constraints from rock physics and well-log data obtaining elastic parameter estimation of high precision directly. We tested the new method on a synthetic example and field multicomponent data, and the results indicated that the estimated fluid factor matched with well-data interpretation and geology information because of adding converted wave information and avoiding indirect inversion error. This demonstrated that the new method can enhance the quality of fluid detection and provide reliable geophysical evidence for reservoir characterization.
Fluid identification in tight sandstone reservoirs based on a new rock physics model