徐鹏宇,周怀来,赵霞,周捷,刘俊平,陶柏丞
XU Pengyu,ZHOU Huailai,ZHAO Xia,ZHOU Jie,LIU Junping,TAO Bocheng

• 1:成都理工大学能源学院
• 2:成都理工大学地球物理学院
• 3:成都理工大学油气藏地质及开发工程国家重点实验室

摘要(Abstract)：

由于受沉积环境、物性、含气水饱和度等因素的影响，非均质碳酸盐岩常规测井储层评价的储层参数识别精度不高。针对这一问题，根据研究区实际情况将储层类型分为孔隙型、裂缝-孔隙型、孔洞型及裂缝-孔洞型4类，再利用斜率关联度选取滑动窗口内对储层类型敏感的测井响应曲线作为极限学习机的训练目标，通过对相邻滑动窗口的叠加来识别目标层的储层类型，然后分别建立各储层类型的孔隙度解释模型。通过计算修正后流动单元指数，应用离散岩石聚类技术划分孔渗等级，建立了渗透率解释模型。结果表明：极限学习机能保证分类精度在95%以上，单次训练耗时0.01 s；采用4种孔隙度解释模型的ELM算法决定系数平均值为0.838，优于PSO-SVM算法；结合流动单元指数和离散岩石聚类技术将孔隙空间类型划分为11小类，能有效地对复杂孔隙空间进行表征。该方法得出的测井综合评价结果可为复杂碳酸盐岩储层气藏的合理高效开发提供必要的支撑。
Conventional logging evaluation of heterogeneous carbonate reservoirs has not high identification accuracy of reservoir parameters due to the influence of sedimentary environment,reservoir property and gas and water saturation. To solve this problem,according to actual situation of studied area,the reservoirs are classified into porous type,fractured-porous type,vuggy type and fractured-vuggy type,and slope coefficient correlation of is used to select logging response curves which are sensitive to reservoir type in sliding window as training parameters of extreme learning machine(ELM). Reservoir types of target layer are identified by superposition of adjacent sliding windows,and then porosity interpretation models of are established. Based on calculation of modified flow unit index,porosity and permeability grades are divided by discrete rock clustering technique,and permeability interpretation model is established. The research shows that extreme learning machine can ensure classification accuracy more than 95%,with single training took 0.01s. Average determination coefficient of ELM algorithm using 4 porosity interpretation models is 0.838,which is better than PSO-SVM algorithm. Combined with flow unit index and discrete rock clustering technique,pore space types are divided into 11 categories,which can effectively characterize complex pore space. Comprehensive logging evaluation results obtained by this method provides necessary support for rational and efficient development of complex carbonate gas reservoirs.

关键词(KeyWords)： 碳酸盐岩;斜率关联度;极限学习机;滑动窗口;流动单元指数;离散岩石聚类
carbonate rock;slope coefficient correlation;extreme learning machine;sliding window;flow unit index;discrete rock clustering

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金项目“致密储层裂缝系统诱发地震异常的机理及其与储层产能的关系”（41874143）

作者(Author): 徐鹏宇,周怀来,赵霞,周捷,刘俊平,陶柏丞
XU Pengyu,ZHOU Huailai,ZHAO Xia,ZHOU Jie,LIU Junping,TAO Bocheng

DOI: 10.19597/j.issn.1000-3754.202203051

参考文献(References)：

• [1]杨生超，邱隆伟，刘魁元，等.渤海湾盆地济阳坳陷邵家洼陷沙四段湖相碳酸盐岩储层特征与分类评价[J].天然气地球科学，2016,27(4):628-638.YANG Shengchao,QIU Longwei,LIU Kuiyuan,et al. Characteristics and classification and evaluation of lacustrine carbonate reservoir of the fourth member of Shahejie Formation of Shaojia Sag, Jiyang Depression, Bohai Bay Basin[J]. Natural Gas Geoscience,2016,27(4):628-638.
• [2]代鹏，丁文龙，曹自成，等.塔里木盆地塔中南坡地区奥陶系优质海相碳酸盐岩储层测井评价[J].物探与化探，2016,40(2):243-249.DAI Peng, DING Wenlong, CAO Zicheng, et al. Logging evaluation of high quality Ordovician marine carbonate reservoir on the south slope of Central Uplift Belt,Tarim Basin[J]. Geophysical and Geochemical Exploration, 2016, 40(2):243-249.
• [3]郝雪峰，尹丽娟，林璐.济阳坳陷油藏类型及属性分布有序性[J].油气地质与采收率，2016,23(1):8-13.HAO Xuefeng, YIN Lijuan, LIN Lu. Ordered distribution of reservoir type and its attributes in Jiyang depression[J].Petroleum Geology and Recovery Efficiency, 2016, 23(1):8-13.
• [4]任杰.碳酸盐岩裂缝性储层常规测井评价方法[J].岩性油气藏，2020,32(6):129-137.REN Jie. Conventional logging evaluation method for carbonate fractured reservoir[J].Lithologic Reservoirs,2020,32(6):129-137.
• [5]黄长兵，黄泽贵，王小飞，等.普光气田飞仙关组三段微生物碳酸盐岩储层特征及主控因素[J].大庆石油地质与开发，2021,40(1):17-25.HUANG Changbing, HUANG Zegui, WANG Xiaofei, et al.Characteristics and main controlling factors of the microbial carbonate reservoirs in the 3rd member of Feixianguan Formation in Puguang Gas Field[J]. Petroleum Geology&Oilfield Development in Daqing,2021,40(1):17-25.
• [6]司马立强，疏壮志.碳酸盐岩储层测井评价方法及应用[M].北京：石油工业出版社，2009.SIMA Liqiang, SHU Zhuangzhi. Logging evaluation method and application of carbonate reservoir[M]. Beijing:Petroleum Industry Press,2009.
• [7]宋修章，吕正祥，章顺利，等.川西坳陷新场构造上沙溪庙组储层下限分析[J].成都理工大学学报（自然科学版），2016,43(4):431-437.SONG Xiuzhang,LüZhengxiang,ZHANG Shunli,et al. Study on the lower limit parameters of the Shangshaximiao Formation reservoir in Xinchang structure,West Sichuan depression,China[J]. Journal of Chengdu University of Technology(Science&Technology Edition),2016,43(4):431-437.
• [8] LI Y, KANG Z J, XUE Z J, et al. Theories and practices of carbonate reservoirs development in China[J]. Petroleum Exploration and Development,2018,45(4):712-722.
• [9]赵军，刘彦斌，王菲菲，等.碳酸盐岩缝洞型储层类型识别与分类预测[J].中国岩溶，2018,37(4):584-591.ZHAO Jun, LIU Yanbin, WANG Feifei, et al. Identification and classification prediction of fractured-vuggy reservoir type in carbonate rocks[J]. Carsologica Sinica, 2018, 37(4):584-591.
• [10]刘冬冬，杨东旭，张子亚，等.基于常规测井和成像测井的致密储层裂缝识别方法：以准噶尔盆地吉木萨尔凹陷芦草沟组为例[J].岩性油气藏，2019,31(3):76-85.LIU Dongdong, YANG Dongxu, ZHANG Ziya, et al. Fracture identification for tight reservoirs by conventional and imaging logging:A case study of Permian Lucaogou Formation in Jimsar Sag, Junggar Basin[J]. Lithologic Reservoirs, 2019, 31(3):76-85.
• [11]张锐锋，李拥军，黄显华，等.巴彦河套盆地变质岩储层测井综合评价方法[J].大庆石油地质与开发，2021,40(6):133-143.ZHANG Ruifeng, LI Yongjun, HUANG Xianhua, et al.Comprehensive logging evaluation method for the metamorphic reservoir in Bayanhetao Basin[J]. Petroleum Geology&Oilfield Development in Daqing,2021,40(6):133-143.
• [12]郭建宏，张占松，张超谟，等.基于灰色系统与测井方法的煤层气含量预测及应用[J].物探与化探，2020,44(5):1190-1200.GUO Jianhong, ZHANG Zhansong, ZHANG Chaomo, et al.Prediction and application of coalbed methane content based on gray system and logging method[J]. Geophysical and Geochemical Exploration,2020,44(5):1190-1200.
• [13]郭建宏，张占松，张超谟，等.用地球物理测井资料预测煤层气含量：基于斜率关联度-随机森林方法的工作案例[J].物探与化探，2021,45(1):18-28.GUO Jianhong, ZHANG Zhansong, ZHANG Chaomo, et al.The exploration of predicting CBM content by geophysical logging data:A case study based on slope correlation random forest method[J]. Geophysical&Geochemical Exploration, 2021,45(1):18-28.
• [14]程翊珊，李治平，许龙飞，等.预测油层无机积垢的BP神经网络方法[J].大庆石油地质与开发，2021,40(3):84-93.CHENG Yishan, LI Zhiping, XU Longfei, et al. BP neural network method for predicting the inorganic scaling in the reservoir[J]. Petroleum Geology&Oilfield Development in Daqing,2021,40(3):84-93.
• [15]孙予舒，黄芸，梁婷，等.基于XGBoost算法的复杂碳酸盐岩岩性测井识别[J].岩性油气藏，2020,32(4):98-106.SUN Yushu,HUANG Yun,LIANG Ting,et al. Identification of complex carbonate lithology by logging based on XGBoost algorithm[J]. Lithologic Reservoirs,2020,32(4):98-106.
• [16]何健，武刚，聂文亮，等.基于近似支持向量机的裂缝分类方法[J].岩性油气藏，2020,32(2):115-121.HE Jian,WU Gang,NIE Wenliang,et al. Fracture classification method based on proximal support vector machine[J].Lithologic Reservoirs,2020,32(2):115-121.
• [17]陈芊澍，文晓涛，何健，等.基于极限学习机的裂缝带预测[J].石油物探，2021,60(1):149-156.CHEN Qianshu,WEN Xiaotao,HE Jian,et al. Prediction of a fracture zone using an extreme learning machine[J].Geophysical Prospecting for Petroleum, 2021, 60(1):149-156.
• [18]徐睿，梁循，齐金山，等.极限学习机前沿进展与趋势[J].计算机学报，2019,42(7):1640-1670.XU Rui,LIANG Xun,QI Jinshan,et al. Advances and trends in extreme learning machine[J]. Chinese Journal of Computers,2019,42(7):1640-1670.
• [19]肖新平，谢录臣，黄定荣.灰色关联度计算的改进及其应用[J].数理统计与管理，1995,14(5):27-30.XIAO Xinping, XIE Luchen, HUANG Dingrong. A modified computation method of grey correlation degree and its application[J]. Mathematical Statistics and Management,1995,14(5):27-30.
• [20]马保国，成国庆.一种相似性关联度公式[J].系统工程理论与实践，2000(7):69-71.MA Baoguo,CHEN Guoqing. A formula of similarity correlation degree[J]. Theory and Practice of System Engineering, 2000(7):69-71.
• [21]曹明霞.灰色关联度模型正负性问题的研究及其改进[J].系统工程与电子技术，2008,30(6):1086-1088.CAO Mingxia. Study and improvement on properties of positive and negative of grey incidence degree model[J]. Systems Engineering and Electronics,2008,30(6):1086-1088.
• [22]张绍良，张国良.灰色关联度计算方法比较及其存在问题分析[J].系统工程，1996,14(3):45-49.ZHANG Shaoliang,ZHANG Guoliang. Comparison between computation models of grey interconnet degree and analysis on their shortages[J]. Systems Engineering,1996,14(3):45-49.
• [23] DENG W Y, ZHENG Q H, LIAN S G, et al.Ordinal extreme learning machine[J]. Neurocomputing, 2010, 74(1-3):447-456.
• [24]朱林奇，张冲，周雪晴，等.融合深度置信网络与核极限学习机算法的核磁共振测井储层渗透率预测方法[J].计算机应用，2017,37(10):3034-3038.ZHU Linqi, ZHANG Chong, ZHOU Xueqing, et al. Nuclear magnetic resonance logging reservoir permeability prediction method based on deep belief network and kernel extreme learning machine algorithm[J]. Journal of Computer Applications,2017,37(10):3034-3038.
• [25]雍世和，张超谟.测井数据处理与综合解释[M].北京：中国石油大学出版社，2006.YONG Shihe, ZHANG Chaomo. Logging data processing and comprehensive interpretation[M]. Beijing:China University of Petroleum Press,2006.
• [26] ZHANG Z, ZHANG H, LI J, et al. Permeability and porosity prediction using logging data in a heterogeneous dolomite reservoir:An integrated approach[J]. Journal of Natural Gas Science and Engineering,2021,86:103743.
• [27]罗贞耀.用侧向资料计算裂缝张开度的初步研究[J].地球物理测井，1990,14(2):83-92.LUO Zhenyao. A preliminary study on calculating fracture opening using lateral data[J]. Geophysical Logging, 1990, 14(2):83-92.
• [28]袁伟，何胜林，焦祥燕，等.北部湾盆地低渗—特低渗含砾砂岩储层渗透率评价方法[J].大庆石油地质与开发，2020,39(5):127-134.YUAN Wei, HE Shenglin, JIAO Xiangyan, et al. Evaluating method for the permeabilities of low permeability and ultra-low permeability pebbly sandstone reservoirs in Beibuwan Basin[J]. Petroleum Geology&Oilfield Development in Daqing,2020,39(5):127-134.
• [29]袁迎中，严文德，戚志林，等.基于页岩气多重运移机制的二项式视渗透率公式[J].特种油气藏，2021,28(3):76-80.YUAN Yingzhong, YAN Wende, QI Zhilin, et al. Binomial apparent permeability formula based on multiple migration mechanism of shale gas[J]. Special Oil&Gas Reservoirs,2021,28(3):76-80.
• [30] MIRZAEI-PAIAMAN A,SABOORIAN-JOOYBARI H, POURAFSHARY P. Improved method to identify hydraulic flow units for reservoir characterization[J]. Energy Technology, 2015, 3(7):726-733.