塔中奥陶系碳酸盐岩缝洞型储层贝叶斯叠前反演预测研究

为了对塔中地区顺托果勒南区块SN1井区碳酸盐岩缝洞型储层进行精细预测，开展了基于三维叠前地震数据的贝叶斯反演技术应用研究，重点论述了测井曲线重构、岩石物理分析、敏感弹性参数验证、叠前反演角道集优选等关键技术环节。利用神经网络技术建立测井响应与地层参数之间的关系模型，重构出垮塌井段的纵、横波速度及密度测井曲线;通过建立缝洞型储层的地震地质模型，利用正、反演技术同时验证了贝叶斯叠前反演在SN1井区的适用性及敏感弹性参数识别储层的有效性。最后将叠后曲率属性与叠前反演的敏感弹性参数属性进行融合，对碳酸盐岩缝洞型储层进行精细刻画，预测出SN1井区两条相交断裂带的交会区域为最有利勘探区带。

In order to precisely predict the fractured-vuggy carbonate reservoir characteristics of SN1 well area in Tazhong area,the application research of Bayesian inversion technique based on 3D prestack seismic data is conducted.This paper is focusing on well logs reconstruction,rock physical analysis,verification of sensitive elastic parameters and angle-gathers optimization of prestack inversion.By using neural network technique,we established the relation models between log response and reservoir parameters,with which we reconstructed the well logs of S-wave and P-wave velocities and density.Making use of forward and inversion techniques for seismic-geological models of fractured-vuggy carbonate reservoir,we proved the capability of prestack Bayesian inversion and the effectiveness of reservoir recognition by sensitive elastic parameters in SN1 well area.At the end of this paper,by the integration of poststack curvature attribute and sensitive elastic parameters from the prestack Bayesian inversion we carried out fine characterization of fracture-vuggy reservoirs of SN1 well area and predicted that the intersection area of the two fault zones is the most favorable exploration area.