Seismic Inversion Method For Prediction Of Shale Reservoir Sweet Spots

More and more attention has been paid to shale gas exploration and development,and the prediction of sweet spots is of great significance for shale gas exploration.The prestack AVO inversion can be used to obtain the shale brittleness index which is the key index for the prediction of sweet spots.Because of the lack of the PS data,the traditional prestack AVO inversion only uses the PP data,which also leads to the error of the S-wave velocity inversion.Joint AVO inversion of PP and PS data can provide high precision inversion results,especially the S-wave velocity.Three-term AVO inversion has the unstable problem because of the density term being not sensitive to the amplitude.The correlation information of the AVO parameters can be used as prior information to stabilize the inversion.The Bayesian theory is a statistical method,which can be used as the framework of AVO inversion,besides,the statistical correlation of AVO parameters will also bring errors to the inversion.The trivariate Cauchy prior probability distribution ensures that the inversion is stable and of high resolution because of including the correlation information of the AVO parameters.In this paper,a joint PP and PS prestack inversion based on a trivariate Cauchy prior probability distribution is proposed.The joint PP and PS inversion makes the inversion result more precise and noise resistant.The trivariate Cauchy distribution ensures that the inversion is stable and of high resolution.Numerical examples indicate that the inversion results of P-velocity,S-velocity and density are more stable,and more accurate than that of using PP data only.The shale model test shows that the method can well predict sweet spots from random noise contaminated data.