基于主控因素分析的A87区块S油层复杂油水层识别方法  

Identification method of complex oil-water layers in S reservoir in A87 block based on analysis of main controlling factors

作　　者：宋延杰[1,2] 姜靖 唐晓敏 王超[3] 乔莎莎 Song Yanjie;Jiang Jing;Tang Xiaomin;Wang Chao;Qiao Shasha(College of Geo-science,Northeast Petroleum University,Daqing 163318,China;Accumulation and Development of Unconventional Oil and Gas,State Key Laboratory Cultivation Base Jointly-constructed by the Province and the Ministry of Science and Technology,Daqing 163318,China;Geological Research Institute of the 9th Oil Production Plant,Daqing Oilfield Co.,Ltd.,Daqing 163853,China)

机构地区：[1]东北石油大学地球科学学院,黑龙江大庆163318 [2]非常规油气成藏与开发省部共建国家重点实验室培育基地,黑龙江大庆163318 [3]大庆油田有限责任公司第九采油厂地质研究所,黑龙江大庆163853

出　　处：《能源与环保》2023年第4期154-165,共12页CHINA ENERGY AND ENVIRONMENTAL PROTECTION

基　　金：黑龙江省自然科学基金项目(D2015012);黑龙江省省属本科高校基本科研业务费项目(HBHZX202003)。

摘　　要：A87区块S油层储层物性差异大、含泥含钙较重,油水层识别难度大。利用岩心分析和测井及试油数据,研究泥质含量、钙质含量和孔渗参数与电阻率之间的关系,并结合典型的低阻油层、高阻水层以及不同孔渗油气层的测井响应特征和岩性物性特征,得出泥质、钙质、物性是影响研究区电阻率测井响应的主要因素。依据电阻率幅值特征与侵入特征的响应机理以及典型油水层的测井响应特征,得出区分水层与油层或油水同层最佳参数为R ILD/R LLD、(R ILM-R ILD)/R ILD,区分油层与油水同层的最佳参数是CNL、R LLD×Φ^(2)/1000,建立了油水层识别图版。对深侧向电阻率进行泥质和钙质影响校正,建立了泥钙校正后的油水层识别图版。利用泥钙校正的敏感测井响应与参数,建立了油水层识别决策树模型和支持向量机模型。与校正前的敏感参数建立的油水层识别图版相比,泥钙校正后的敏感参数建立的油水层识别图版、决策树模型和支持向量机模型精度和解释符合率均有一定提高。应用研究区12口投产井进行验证,表明3种方法均能较好地判别研究区油水层,且决策树模型识别研究区油水层效果最好。The physical properties of reservoir S in A87 block vary greatly,with heavy mud and calcium content,making it difficult to identify oil and water layers.This article uses core analysis,logging,and oil testing data to study the relationship between mud content,calcium content,porosity and permeability parameters,and resistivity.Based on the logging response characteristics and lithological properties of typical low resistivity oil layers,high resistivity water layers,and different porosity and permeability oil and gas layers,it is concluded that mud,calcium,and physical properties are the main factors affecting the resistivity logging response in the study area.Based on the response mechanism of resistivity amplitude characteristics and invasion characteristics,as well as the logging response characteristics of typical oil and water layers,the optimal parameters for distinguishing between water layers and oil layers or oil-water layers are R ILD/R LLD and(R ILM-R ILD)/R ILD,and the best parameters to distinguish oil layers from oil-water layers are CNL and R LLD×Φ^(2)/1000,and then,the interpretation chart for oil-water layers is established.The deep lateral resistivity is corrected for influence of mud and calcium,and the corrected interpretation chart for oil-water layers is established.The decision tree model and support vector machine model are established to interpret oil-water layers with corrected sensitive log responses and parameters.Compared with the oil-water layer recognition chart established by the sensitive parameters before correction,the accuracy and interpretation accuracy of the oil-water layer recognition chart,decision tree model,and support vector machine model established by the sensitive parameters after mud calcium correction have been improved to a certain extent.The application of 12 production wells in the research area is verified,and it is found that all three methods can effectively identify the oil-water layers in the research area,and the decision tree model has the best effect i

关 键 词：复杂油水层 主控因素 泥质和钙质校正 图版法 决策树模型 支持向量机模型 

分 类 号：P631[天文地球—地质矿产勘探]