多总体逐步判别分析法在复杂油水层识别中的应用  

The application of multigroup stepwise discriminant analysis for complex oil,water and dry layers identification

作　　者：王雅春[1,2] 田春阳[3] 张振伟[1] 金铁文[4] 杨宏松[1] 

机构地区：[1]大庆石油学院地球科学学院,黑龙江大庆163318 [2]中国石油大学(北京)资源与信息学院,北京102249 [3]大庆油田有限责任公司第二采油厂,黑龙江大庆163414 [4]辽河油田分公司生产运行处,辽宁盘锦124010

出　　处：《大庆石油学院学报》2010年第2期26-32,共7页Journal of Daqing Petroleum Institute

基　　金：黑龙江省教育厅科学技术研究项目(11531z13)

摘　　要：综合敖南油田敖21区块葡萄花油层录井、测井和试油资料,考虑与油水层识别有关的10个参数(冲洗带电阻率、地层真电阻率、声波时差、自然电位、自然伽马、泥质含量、有效孔隙度、含油饱和度、深侧向电阻率、密度),应用逐步判别分析的方法,优选地层真电阻率、自然电位、自然伽马和深侧向电阻率等4个作用比较显著的参数,建立敖21区块葡萄花油层的油层、油水同层、水层和干层的判别函数模型.通过回判率检验和F分布检验表明,建立的判别模型效果显著,油水同层和干层判别的符合率达到100%,油层和水层判别的准确率达到80%以上.Mud logging,well logging and formation testing data are used to set up equations for oil,water and dry layers identification of Putaohua reservoir in Ao 21 block of Aonan oilfield.Here,10 variables such as flushed zone formation resistivity,true formation resistivity,acoustic,spontaneous potential,natural gamma ray,clay content,effective porosity,oil saturation,deep lateral resistivity,density are considered,4 variables such as true formation resistivity,spontaneous potential,natural gamma ray and deep lateral resistivity are selected into the functions.The oil,oil-water,water and dry layers identification in the studied area with discriminant functions is proved to be very effective with a precision of more than 80%.

关 键 词：交会图法 判别分析法 油水层识别 敖21区块 葡萄花油层 

分 类 号：P631.84[天文地球—地质矿产勘探]