分频线性回归技术在储集层厚度预测中的应用  

Application of Spectrum Decomposition Linear Regression to Thickness Prediction of Reservoir

作　　者：陈香朋[1] 吕小伟[2] 张营革[1] 

机构地区：[1]中国石化胜利油田有限责任公司物探研究院沾车室,山东东营257000 [2]中国石化胜利油田有限责任公司新疆勘探开发中心地质工艺研究所,山东东营257000

出　　处：《新疆石油地质》2012年第1期98-101,共4页Xinjiang Petroleum Geology

摘　　要：应用地震波分频属性能较好地刻画储集层厚度在平面上的分布规律,但在实际应用中,单一的分频属性存在多解性,不能较好地预测储集层的厚度。在理论模型分析的基础上,利用多个单频属性进行线性回归,可预测储集层厚度。该方法在大王庄地区沙河街组二段储集层厚度预测中取得了良好的效果,预测结果与钻探结果具有较好的一致性,表明该方法在储集层厚度预测中具有很大的应用潜力。Seismic spectrum decomposition attribute can be applied to well sculpture the lateral distribution of reservoir thickness,but the single spectrum attribute can't be used to well predict the reservoir thickness due to its multiple solutions.Based on theoretical model analysis,several single-spectrum attributes are regressed followed by being used to predict reservoir thickness.This method is applied to the reservoir thickness prediction of No.2 member of Shahejie formation in Dawangzhuang area,and the result shows that the predicted result is accordant with actual result by drilling.

关 键 词：分频属性 线性回归 储集层厚度 预测 

分 类 号：P631.445.91[天文地球—地质矿产勘探]