【摘要】 常规多元地震属性储层定量预测技术是一种快捷有效的储层预测方法,然而对于覆盖面积较大、沉积环境较复杂的地区,该法的预测精度就会明显降低。为此,本文提出了基于沉积特征分区域的多元地震属性储层定量预测方法,首先利用地震波形分类技术将研究区按沉积特征划分为不同区域,然后按区域分别统计多元地震属性与储层信息的关系并进行线性拟合,最终将各区域拟合结果综合为全区储层信息预测结果。在松辽盆地北部大庆长垣以东徐家围子地区登娄库组二段储层预测中,利用104口井砂岩数据及优选的7种地震属性,采用逐步线性回归法进行全区拟合,其相关系数仅为0.48;采用基于沉积特征分区域的多元地震属性储层定量预测方法进行全区拟合,相关系数达0.83,且预测结果符合地质认识及沉积规律。更多还原

【Abstract】 Conventional multiple seismic attributes quantitative reservoir prediction method are fast and effective.However the accuracy of this method would decrease if the study area is large and depositional environment is complex.To this end,the paper proposes a quantitative reservoir prediction method by multiple-seismic-attributes based on dividing the study area to different regions by depositional characteristics.Firstly,the study area is divided into different regions according to their depositional characteristics by seismic waveform classification,and then linear fit is progressed between multiple seismic attributes and reservoir information in different regions,finally,fitting results of different regions are synthesized as the predicting results of reservoir information in the whole area.The reservoir prediction of Denglouku formation in Xujiaweizi area,east of Daqing placanticline in the north of Songliao Basin,was made.The sandstone data of 104 wells and 7 optimized seismic attributes were used,the fitting degree was only 0.48 by the gradual linear regression method,which was increased to 0.83 by the multiple seismic attributes quantitative reservoir prediction method based on dividing regions according to sedimentary characteristics,at the same time the predicted results agreed with the sedimentary laws and geological knowledge.更多还原