| The quantitative evaluation of time-lapse seismic data remains a challenge due to poor match between the model predictions and the actual seismic data.; Velocity anisotropy is one important reason for the mismatch. I compile experimental velocity-anisotropy data from cores to explore the empirical relationships between anisotropy parameters and general well-log information. Then, I develop a method to estimate Thomsen's anisotropy parameters &egr; and gamma using a regression of the data in the crossplot domain between velocity and porosity. I present an application result of the method to demonstrate the significance of the correction.; Next, using the corrected velocity, I present a method of impedance decomposition into Vp, Vs, and rho using three elastic impedances derived from the seismic inversion of angle stacks. In general, the maximum stack angle of seismic data is limited to be less than 30°, which is not wide enough to obtain the stable calculation result. I discuss the effect of noise on the analysis as the most important reason that decomposition is difficult. I show an innovative method incorporating rock-physics bounds as constraints for the analysis. I apply it to an actual dataset from an offshore oil field; I demonstrate the result of analysis for sand-body detection.; Based on the estimated Vp, V s, rho and shale volume from the elastic impedances, I develop a workflow to determine the saturation of formation-water, oil and gas from seismic data. First, I consider the pressure effect and the saturation scale of fluids for time-lapse seismic analysis. Second, I demonstrate a deterministic approach to computing the fluid saturation to evaluate time-lapse seismic data. In this approach, I derive the physical properties of the water-saturated sandstone reservoir. Then, by comparing the in-situ-fluid-saturated properties with the 100% formation-water-saturated reservoir properties, I determine the bulk modulus and the density of the fluid phase in the reservoir. Solving three equations simultaneously (bulk modulus, density and the total saturation), I compute the saturation of each fluid. In the result of the case study, I emphasize the validity of the workflow for quantitatively detecting the fluid displacement and for delineating a remaining oil accumulation. |
