Nonlinear AVO Inversion Theory And Method For Shale Reservoir

The prestack AVO/AVA inversion technique is one of the most commonly used seismic prediction techniques for shale reservoir exploration and development.Based on the AVA inversion results,the physical parameters,oil-bearing properties,TOC content etc of shale reservoir can be further predicted and estimated.Enhancing the accuracy of AVA inversion results can provide more reliable seismic information for the prediction of the sweet spots of shale reservoir and the fine description of reservoir in the middle and late stages of development.Therefore,this dissertation carries out studies on prestack AVO/AVA inversion of shale reservoir in aspect of improving the accuracy of inversion results.Considering the unique characteristics of shale reservoirs,several new strategies and methods of prestack nonlinear AVO/AVA inversion are proposed here.The conventional elastic parameters dominated by P-and S-wave velocity and density are importanted data supported for shale reservoir exploration and development.These parameters are usually obtained from pre-stack AVO inversion based on the approximate formulas of the Zoeppritz equations.However,the assumptions and low calculation accuracy of the approximation formulas greatly limit the estimated accuracy of these parameters.These limitations can be handled by using exact Zoeppritz equations.Therefore,a direct inversion of conventional elastic parameters based on the exact Zoeppritz equations is carried out.The inversion objective function is then constructed by utilizing Bayesian theory.Meanwhile,the statistical correlations of the elastic parameters and prior distribution characteristics are introduced as priori information.In addition,the modified Cauchy distribution was introduced in order to overcome the problem that the sparsity of priori constraints too strong which cause the weak reflection is suppressed so seriously that it’s not easy to recovery the weak reflection information.The Gauss-Newton method and the combination method of the generalized linear inversion(GLI)and iterative reweighted least squares(IRLS)algorithm are used to quickly solve the nonlinear objective function.In contrast,the second method is more convenient for inversion stability analysis.Both synthetic data and field data examples demonstrate the feasibility and effectiveness of the proposed method.Young’s modulus and Poisson’s ratio are important rock mechanics parameters directly related to the exploration and development of shale reservoirs.However,the cumulative errors caused by indirect calculation and the low calculation accuracy of the approximation formulas greatly limit the acquisition accuracy of these parameters.In this paper,the Zoeppritz equations is converted from the classical form to a new form containing the chosen elastic constants of Young’s modulus,Poisson’s ratio and density by using the rock physics theory.Then,the direct inversion of the Young’s modulus and Poisson’s ratio based on these equations is achieved,which effectively improves the estimated accuracy of these two parameters.Shale reservoirs generally exhibit strong anisotropy characteristics and can be approximately equivalent to VTI media when their structures are relatively simple.Therefore,the nonlinear AVO/AVA inversion method of VTI media is studied.Aiming at the problem that the approximate reflection coefficients formulas of VTI media have low calculation accuracy and the exact reflection coefficients equations of VTI media have very complicated forms and strong nonlinearity,the new reduced approximation equations are derived under the assumption of weak anisotropy.Then the objective function is constructed based on the new reduced approximation equations.Meanwhile,the blockiness constraint term is introduced to enhance layer boundaries.Finally,the combination method of the GLI and IRLS algorithm is used to solve the objective function.This method effectively improves the estimated accuracy of VTI media parameters.The inter-salt shale oil reservoirs have obvious interbed features,which can be effectively identified by the high-frequency information contained in the seismic data.Therefore,the frequency-division joint inversion and the above-mentioned nonlinear AVO/AVA inversion method are combined to achieve high resolution prestack joint inversion for salt shale oil reservoirs.