基于经验模态分解和能量熵判别的火成岩岩性识别方法——以春风油田石炭系火成岩储层为例  

Identification of igneous reservoir lithology based on empirical mode decomposition and energy entropy classification: A case study of Carboniferous igneous reservoir in Chunfeng oilfield

作　　者：韩玉娇 袁超[1] 范宜仁[2] 葛新民[2] 范卓颖 杨文超 Han Yujiao;Yuan Chao;Fan Yiren;Ge Xinmin;Fan Zhuoying;Yang Wenchao(Department of Well Logging and Remote Sensing Technology,PetroChina Research Institute of Petroleum Exploration and Development, Beijing 100080,China;School of Geosciences,China University of Petroleum(East China),Qingdao,Shandong 266580,China;College of Geosciences,China University of Petroleum(Beijing),Beijing 102249,China;Shengli Well Logging Co.Ltd.,SINOPEC,Dongying,Shandong 257096,China)

机构地区：[1]中国石油勘探开发研究院测井与遥感所,北京100080 [2]中国石油大学(华东)地球科学与技术学院,山东青岛266580 [3]中国石油大学(北京)地球科学学院,北京102249 [4]中国石化胜利石油工程有限公司测井公司,山东东营257096

出　　处：《石油与天然气地质》2018年第4期759-765,共7页Oil & Gas Geology

基　　金：中国石油天然气集团公司科学研究与技术开发项目(2015A-3601)

摘　　要：火成岩储层具有"喷发模式多样化、岩性成因多元化、矿物组合多变化"的特征,岩性精细识别难度大,严重制约了储层参数的计算及后续油气开发。以准噶尔盆地春风油田石炭系储层为例,结合岩心和薄片等分析测试资料将储层发育的岩性分为5类:玄武岩、玄武质安山岩、安山岩、凝灰岩和火山角砾岩。在明确不同岩性测井响应特征的基础上,采用"逐级剥离"的思想,利用交会图法识别出了凝灰岩和火山角砾岩。针对较难识别的火山熔岩,引入经验模态分解算法将常规测井资料转化为多个频带的本征模态函数集,得到了各熔岩不同测井参数经验模态函数的能量熵,应用判别算法实现了火成岩岩性的精细识别。区块应用结果表明:该方法岩性识别整体符合率高达93.7%,有效提高了岩性识别精度。Igneous reservoirs are characterized by diversity of eruption patterns and lithology geneses,and variation in mineralogy assemblage.It is really difficult to identify lithology in details,which in turn will greatly hinder the correct calculation of reservoir parameters,as well as subsequent hydrocarbon development strategies.Therefore,the study took a case study of the carboniferous igneous reservoirs in Chunfeng oilfield,Junggar Basin.The lithological categories of the reservoirs are basalt,basaltic andesite,andesite,tuff and volcanic breccias.A combination of core,thin section and other tests’data were used in the study.Then the logging response characteristics of different lithologies were clarified.The tuff and volcanic breccia were identified with the cross plot technique and“progressive stripping”concept.For the hard-to-identify volcanic lava,we used the empirical mode decomposition algorithm to convert the conventional logging data into multiple band sets of intrinsic mode functions,and attained the energy entropy of empirical mode function of various logging parameters for lavas.Then the precise identification of igneous lithologies was realized by using the discriminant algorithm.In applying the method to the block,we found that the overall matching rate of the proposed method is 93.7%,which has greatly improved the accuracy of lithological identification.

关 键 词：经验模态分解 能量熵 岩性识别 火成岩 春风油田 

分 类 号：TE122.2[石油与天然气工程—油气勘探]