基于常规测井识别微生物碳酸盐岩岩相新方法及应用  

New method for quantitative identification of carbonate lithofacies based on conventional logging curves and its application

作　　者：李昌 贾俊[3] 沈安江 王亮 梁正中[5] 李振林[6] LI Chang;JIA Jun;SHEN An-jiang;WANG Liang;LIANG Zheng-zhong;LI Zhen-li(PetroChina Hangzhou Research Institute of Geology,Hangzhou 310023,China;PetroChina Key Laboratory of Carbonate Reservoirs,Hangzhou 310023,China;Mianyang Teachers'Colloge,Mianyang 621000,China;Chengdu University of Technology,Chengdu 610500,China;Yulin University,Yulin 719000,China;CNPC Logging,Xi'an 771000,China)

机构地区：[1]中国石油杭州地质研究院,杭州310023 [2]中国石油天然气集团公司碳酸盐岩储层重点实验室,杭州310023 [3]绵阳师范学院,绵阳621000 [4]成都理工大学,成都610500 [5]榆林学院,榆林719000 [6]中国石油集团测井有限公司测井应用研究院,西安771000

出　　处：《地球物理学进展》2020年第5期1792-1802,共11页Progress in Geophysics

基　　金：中国石油科技重大专项03课题“深层油气储层形成机理与分布规律”(2018A-0103);国家专项“寒武系-中新元古界碳酸盐岩规模储层形成与分布研究”(2016ZX05004002);国家专项“四川盆地大型碳酸盐岩气田开发示范工程”(2016ZX05052);陕西省自然科学基础研究计划(2017JM4017)联合资助.

摘　　要：岩相是沉积储层研究的重要基础,由于取心成本高,而录井资料不能精细描述岩石结构组分,利用测井资料识别岩相成为主要手段.然而由于碳酸盐岩强非均质性,导致测井识别基于岩石结构组分分类的岩相难度较大,特别微生物发育使得情况更加复杂,电成像测井虽然可以识别微生物构造,然而电成像资料一般较少,目前利用常规测井仍是主要手段,传统交会图版法和机器学习方法都是基于测井数值的绝对差异划分不同岩相,而对较小的相对差异并不敏感,另外机器学习方法依赖于学习样本的数量和质量,取心资料少的情况,难以发挥其优越性.为此,本文提出了一种基于常规测井曲线色彩图版的岩相识别新方法,其优势是分析测井曲线的相对差异,将经验认识融入到岩相识别中,能够有效解决岩心资料少影响识别效果的难题.根据岩-电关系分析,获取经验认识,优选敏感测井曲线,将优选测井曲线之间相对关系转化为不同颜色的组合,与岩心标定,根据经验认识调整颜色,最终建立不同岩相的典型色彩识别图版,从而实现微生物碳酸盐岩岩相测井识别.以四川盆地MX地区震旦系灯影组微生物碳酸盐岩地层为例,实际应用并取心资料验证表明:灯影组分为4类岩相,其识别符合率能够达到75%以上,该方法的应用对MX地区沉积微相精细研究工作起到了推动作用,同时也丰富了现有碳酸盐岩岩相测井识别方法.Lithofacies is an important basis for the study of sedimentary reservoirs.However,because of the strong heterogeneity of carbonate rocks,it is difficult to identify the lithofacies based on the classification of rock fabric components.And due to the high cost of coring,using logging data to identify lithofacies become the main technical means.For microbial carbonate rocks,the logging characteristics of lithofacies are more complex because of the more strong heterogeneity and the microorganisms development.Although electrical image logging can identify microbial structural features,such as laminations and laminations,but the data is limited,so it is still the main method to identify microbial carbonate lithofacies with conventional logging data.The traditional cross-plots methods and machine learning methods are based on the absolute difference of logging data to divide the different lithofacies,but not sensitive for the smaller relative difference.In addition,the recognition effect of machine learning methods depends heavily on the quantity and quality of learning samples.Therefore,a colormap method is proposed with conventional logging curve in this paper,which is specially for the case of small number of cores.Analysis of the relationship between lithofacies and logging characteristics can help to choose the more sensitive logging curve.The relative difference between the selected logging curves is transformed into colormap.By core calibrating,the typical colormap for the different lithofacies is established and used to identify lithofacies of microbial carbonate.A case study of microbial carbonate of Sinian Dengying formation in MX area Sichuan Basin,four dominant dolomites representing major depositional settings were classified,including algae dolomite,grainstone,mudstone and argillaceous mudstone.The coring data verification shows the recognition coincidence rate more than 75%.The application of this method promotes the fine research of sedimentary microfacies in MX area.It also enriches existing carbonate

关 键 词：碳酸盐岩结构组分 常规测井 色彩模式 岩相识别 

分 类 号：P631[天文地球—地质矿产勘探]