龙门山前侏罗系沙溪庙组致密砂岩储层叠前地震预测

龙门山前侏罗系沙溪庙组致密砂岩横向变化快、非均质性强、孔隙结构复杂，储层与围岩阻抗差异小，岩石物理规律不清，且地震叠前道集资料品质差、测井横波资料匮乏，优质储层预测困难。针对以上问题，通过叠前反褶积、阿尔法滤波、剔除拟合振幅补偿优化处理了叠前CRP道集，采用改进的Xu-White模型预测了横波，并通过贝叶斯稀疏脉冲反演获得了叠前弹性参数VP/VS，预测了砂体边界，最后在砂体边界控制下，利用蒙特卡罗—贝叶斯分类方法反演了孔隙度，预测了致密砂岩有利储层发育区。研究表明：贝叶斯稀疏脉冲反演简化了稀疏脉冲反演过程，求解更稳定，砂岩边界控制下的孔隙度反演排除了泥岩影响，提高了孔隙度预测精度，该技术在龙门山前侏罗系沙溪庙组致密砂岩有利储层预测中取得了较好的效果，具有一定的推广应用价值。

Prestack seismic prediction of tight sandstones for Jurassic Shaximiao Formation in front of Longmenshan

The tight sandstones of Jurassic Shaximiao Formation had the features of rapid lateral changes，strong heterogeneities and complex pore structure in front of Longmenshan.Rock physics characteristics were similar between tight sandstones and mudstones.In addition，the quality of prestack data was poor and the shear wave logging was scarce.This made reservoir prediction difficult.The CRP gather was optimized by a series of corresponding methods，including Decon，Alpha Filter and AMP Compensation.Shear wave was predicted with modified Xu-White model.After that，the prestack parameter of VP/VS was acquired by Bayesian sparse pulse inversion.And sandstone boundary was predicted using VP/VS.Finally，reservoir porosity was calculated by Monte Carlo-Bayesian classification under the control of sandstone boundary，and the effective reservoirs were determined.Research indicated that Bayesian sparse pulse inversion simplified the sparse pulse inversion process and its solution was more stable.Porosity inversion with the control of sandstone boundary avoided the mudstone affect and improved the accuracy of porosity prediction.The new method had achieved excellent results in the effective reservoir prediction of tight sandstone for Jurassic Shaximiao Formation in front of Longmenshan and had significant implications for late exploration.