【摘要】 四川盆地焦石坝地区下志留统龙马溪组下部沉积了一套较厚的海相暗色泥页岩,由于其含气的优质泥页岩储层速度、密度变化小,波阻抗响应特征不明显,利用常规波阻抗反演技术进行优质泥页岩的定量预测存在困难。为此,提出了利用拟声波反演进行优质泥页岩定量预测的技术,即基于优质泥页岩具有高TOC值的这一特征,首先利用小波多尺度分解技术提取声波测井曲线中的低频信息,再利用信息融合技术将TOC曲线与原始声波曲线重构成拟声波曲线,结合该区地震数据和稀疏脉冲波阻抗反演技术实现了优质泥页岩的定量预测。预测结果表明:该区龙马溪组下部的优质泥页岩整体发育,主体部位优质泥页岩厚度大于35m,由东北向西南优质泥页岩厚度增加。拟声波反演预测结果与后续实钻结果基本一致,预测误差小于1m,同时龙马溪组优质泥页岩的纵横向展布特征也得到了有效刻画。更多还原

【Abstract】 Gas shale in the Lower Silurian Longmaxi Fm in the Jiaoshiba area of the Sichuan Basin has insignificant wave impedance response characteristics due to small variations of velocity and density,thus it is difficult to perform quantitative prediction of highquality shale reservoirs with the conventional impedance inversion technology.In view of this,pseudo-acoustic impedance inversion was proposed.Due to the high TOCof high-quality shales,the multi-scale wavelet decomposition technology was first used to extract low-frequency information in the sonic logs.Then,the information fusion technology was used to rebuild pseudo-acoustic logs based on TOClogs and the original acoustic logs.The pseudo-acoustic logs were used to quantitatively predict the high-quality shale reservoirs in combination with the seismic data and sparse spike impedance inversion.The results show that high-quality shales are widely developed in the Lower Longmaxi Fm in the study area.It is over 35 mthick in the core area but is relatively thin in the areas with well-developed faults.As a whole,its thickness increases from northeast to southwest.The prediction results were basically verified by follow-up drilling results.The error of prediction is less than 1 m.In addition,the vertical and lateral distribution of high-quality shales was also well defined by using this technology.更多还原