油气勘探

核磁共振横向弛豫时间谱高斯混合聚类及应用

展开
  • 1.中国石油大学(华东)地球科学与技术学院,山东青岛266580;
    2.中国石油天然气集团有限公司测井重点实验室,西安710077;
    3.海洋国家实验室海洋矿产资源评价与探测功能实验室,山东青岛266071;
    4.中国地质调查局油气资源调查中心,北京100083;
    5.中国石油集团测井有限公司,西安710077;
    6.中国石油勘探开发研究院,北京100083;
    7.中国石油青海油田公司,甘肃敦煌736202;
    8.中海石油(中国)有限公司湛江分公司,广东湛江524057;
    9.中国石化石油工程技术研究院,北京100101
葛新民(1985-),男,江西于都人,博士,中国石油大学(华东)副教授,主要从事测井岩石物理、核磁共振测井和复杂油气层测井评价方法等研究。地址:山东省青岛市黄岛区长江西路66号,中国石油大学(华东)工科楼C534室,邮政编码:266580。E-mail: gexinmin2002@163.com

收稿日期: 2021-08-30

  网络出版日期: 2022-03-16

基金资助

国家自然科学基金(42174142); 国家科技重大专项(2017ZX05039-002); 中国石油天然气集团有限公司测井重点实验室运行基金(2021DQ20210107-11); 中央高校基本科研业务费专项(19CX02006A); 中国石油天然气集团有限公司重大科技项目(ZD2019-183-006)

An unsupervised clustering method for nuclear magnetic resonance transverse relaxation spectrums based on the Gaussian mixture model and its application

Expand
  • 1. School of Geosciences, China University of Petroleum, Qingdao 266580, China;
    2. CNPC Key Well Logging Laboratory, Xi’an 710077, China;
    3. Laboratory for Marine Mineral Resources, Qingdao National Laboratory for Marine Science and Technology, Qingdao 266071, China;
    4. Oil and Gas Survey Center of China Geological Survey, Beijing 100083, China;
    5. China Petroleum Logging Co. Ltd., Xi’an 710077, China;
    6. PetroChina Research Institute of Petroleum Exploration & Development, Beijing 100083, China;
    7. PetroChina Qinghai Oilfield Company, Dunhuang 736202, China;
    8. Zhanjiang Branch of CNOOC Ltd., Zhanjiang 524057, China;
    9. Sinopec Research Institute of Petroleum Engineering, Beijing 100101, China

Received date: 2021-08-30

  Online published: 2022-03-16

摘要

为使核磁共振测井横向弛豫时间(T2)谱的定量表征结果更为直观地反映储集层类型和孔隙结构,提出基于高斯混合模型(GMM)的T2谱无监督聚类和孔隙结构定量识别方法。首先对T2谱数据进行主成分降维,减弱数据间的相关性;其次采用高斯混合模型概率密度函数对降维数据进行拟合,结合期望值最大化算法和赤池信息准则变化率得到模型参数和最佳聚类群集;最后分析不同聚类群集的T2谱特征、孔隙结构类型等,并与T2几何平均值、T2算术平均值等进行对比,通过数值模拟和核磁共振测井资料验证算法有效性。研究表明,基于GMM方法的聚类结果与T2谱形态、T2谱、孔隙结构、油气产能等具有很好的对应性,为孔隙结构定量识别、储集层级别划分和产能评价等提供新的手段。

本文引用格式

葛新民, 薛宗安, 周军, 胡法龙, 李江涛, 张恒荣, 王烁龙, 牛深园, 赵吉儿 . 核磁共振横向弛豫时间谱高斯混合聚类及应用[J]. 石油勘探与开发, 2022 , 49(2) : 296 -305 . DOI: 10.11698/PED.2022.02.08

Abstract

To make the quantitative results of nuclear magnetic resonance (NMR) transversal relaxation (T2) spectrums reflect is proposed the type and pore structure of reservoir more directly, an unsupervised clustering method was developed to obtain the quantitative pore structure information from the NMR T2 spectrums based on the Gaussian mixture model (GMM). We conducted the principal component analysis on T2 spectrums in order to reduce the dimension data and the dependence of the original variables. The dimension-reduced data was fitted using the GMM probability density function, and the model parameters and optimal clustering numbers were obtained according to the expectation-maximization algorithm and the change of the Akaike information criterion. Finally, the T2 spectrum features and pore structure types of different clustering groups were analyzed and compared with T2 geometric mean and T2 arithmetic mean. The effectiveness of the algorithm has been verified by numerical simulation and field NMR logging data. The research shows that the clustering results based on GMM method have good correlations with the shape and distribution of the T2 spectrum, pore structure, and petroleum productivity, providing a new means for quantitative identification of pore structure, reservoir grading, and oil and gas productivity evaluation.

参考文献

[1] 肖立志. 核磁共振成像测井与岩石核磁共振及其应用[M]. 北京: 科学出版社, 1998.
XIAO Lizhi. Nuclear magnetic resonance imaging logging and rock nuclear magnetic resonance and its application[M]. Beijing: Science Press, 1998.
[2] 肖立志, 谢然红, 廖广志. 中国复杂油气藏核磁共振测井理论与方法[M]. 北京: 科学出版社, 2012.
XIAO Lizhi, XIE Ranhong, LIAO Guangzhi. Theory and method of NMR logging for Chinese complex oil and gas reservoirs[M]. Beijing: Science Press, 2012.
[3] DUNN K J, BERGMAN D J, LATORRACA G A. Nuclear magnetic resonance petrophysical and logging applications[M]. New York: Elsevier, 2002.
[4] 胡法龙, 周灿灿, 李潮流, 等. 基于弛豫-扩散的二维核磁共振流体识别方法[J]. 石油勘探与开发, 2012, 39(5): 552-558.
HU Falong, ZHOU Cancan, LI Chaoliu, et al. Fluid identification method based on 2D diffusion-relaxation nuclear magnetic resonance (NMR)[J]. Petroleum Exploration and Development, 2012, 39(5): 552-558.
[5] MUTINA A, BACHMAN N, RENDON L. Fast multidimensional NMR logging provides advanced fluid characterization at a step change in logging speed[R]. London, UK: SPWLA 59th Annual Logging Symposium, 2018.
[6] SUN B Q. In situ fluid typing and quantification with 1D and 2D NMR logging[J]. Magnetic Resonance Imaging, 2007, 25(4): 521-524.
[7] KAUSIK R, JIANG T M, VENKATARAMANAN L, et al. Reservoir producibility index (RPI) based on 2D T1-T2 NMR logs[R]. Texas, USA: SPWLA 60th Annual Logging Symposium, 2019.
[8] ZHANG H, WANG Y, FANG T, et al. First application of new generation NMR T1-T2 logging and interpretation in unconventional reservoirs in China[R]. SPE 202261-MS, 2020.
[9] 范宜仁, 刘建宇, 葛新民, 等. 基于核磁共振双截止值的致密砂岩渗透率评价新方法[J]. 地球物理学报, 2018, 61(4): 1628-1638.
FAN Yiren, LIU Jianyu, GE Xinmin, et al. Permeability evaluation of tight sandstone based on dual T2 cutoff values measured by NMR[J]. Chinese Journal of Geophysics, 2018, 61(4): 1628-1638.
[10] 李鹏举, 谷宇峰. 核磁共振T2谱转换伪毛管压力曲线的矩阵方法[J]. 天然气地球科学, 2015, 26(4): 700-705.
LI Pengju, GU Yufeng. Matrix method of transforming NMR T2 spectrum to pseudo capillary pressure curve[J]. Natural Gas Geoscience, 2015, 26(4): 700-705.
[11] 闫建平, 温丹妮, 李尊芝, 等. 基于核磁共振测井的低渗透砂岩孔隙结构定量评价方法: 以东营凹陷南斜坡沙四段为例[J]. 地球物理学报, 2016, 59(4): 1543-1552.
YAN Jianping, WEN Danni, LI Zunzhi, et al. The quantitative evaluation method of low permeable sandstone pore structure based on nuclear magnetic resonance (NMR) logging[J]. Chinese Journal of Geophysics, 2016, 59(4): 1543-1552.
[12] GE X M, FAN Y R, ZHU X J, et al. Determination of nuclear magnetic resonance T2 cutoff value based on multifractal theory: An application in sandstone with complex pore structure[J]. Geophysics, 2015, 80(1): 11-21.
[13] 丁娱娇, 柴细元, 邵维志, 等. 基于核磁共振T2谱集中度的低孔隙度低渗透率储层饱和度参数研究[J]. 测井技术, 2017, 41(4): 405-411.
DING Yujiao, CHAI Xiyuan, SHAO Weizhi, et al. Key parameters of water saturation based on concentration of T2 spectrum distribution[J]. Well Logging Technology, 2017, 41(4): 405-411.
[14] 白松涛, 程道解, 万金彬, 等. 砂岩岩石核磁共振T2谱定量表征[J]. 石油学报, 2016, 37(3): 382-391.
BAI Songtao, CHENG Daojie, WAN Jinbin, et al. Quantitative characterization of sandstone NMR T2 spectrum[J]. Acta Petrolei Sinica, 2016, 37(3): 382-391.
[15] 张超谟, 陈振标, 张占松, 等. 基于核磁共振T2谱分布的储层岩石孔隙分形结构研究[J]. 石油天然气学报, 2007, 29(4): 80-86.
ZHANG Chaomo, CHEN Zhenbiao, ZHANG Zhansong, et al. Fractal characteristics of reservoir rock pore structure based on NMR T2 distribution[J]. Journal of Oil and Gas Technology, 2007, 29(4): 80-86.
[16] GE X M, FAN Y R, CAO Y C, et al. Reservoir pore structure classification technology of carbonate rock based on NMR T2 spectrum decomposition[J]. Applied Magnetic Resonance, 2014, 45(2): 155-167.
[17] 钟吉彬, 阎荣辉, 张海涛, 等. 核磁共振横向弛豫时间谱分解法识别流体性质[J]. 石油勘探与开发, 2020, 47(4): 691-702.
ZHONG Jibin, YAN Ronghui, ZHANG Haitao, et al. A decomposition method of nuclear magnetic resonance T2 spectrum for identifying fluid properties[J]. Petroleum Exploration and Development, 2020, 47(4): 691-702.
[18] BAUER K, KULENKAMPFF J, HENNINGES J, et al. Lithological control on gas hydrate saturation as revealed by signal classification of NMR logging data[J]. Journal of Geophysical Research. Solid Earth, 2015, 120(9): 6001-6017.
[19] ANAND V. Novel methodology for accurate resolution of fluid signatures from multi-dimensional NMR well-logging measurements[J]. Journal of Magnetic Resonance, 2017, 276: 60-68.
[20] ROJAS P A, RINCÓN M, NETTO P, et al. Application of machine learning tool to separate overlapping fluid components on NMR T2 distributions: Case studies from laboratory displacement experiment and well logs[R]. SPE 197684-MS, 2019.
[21] XU R, DENG T Q, JIANG J J, et al. Integration of NMR and conventional logs for vuggy facies classification in the Arbuckle Formation: A machine learning approach[J]. SPE Reservoir Evaluation & Engineering, 2020, 23(3): 917-929.
[22] 孟昆, 王胜建, 薛宗安, 等. 利用核磁共振资料定量评价页岩孔隙结构[J]. 波谱学杂志, 2021, 38(2): 215-226.
MENG Kun, WANG Shengjian, XUE Zong’an, et al. Quantitative evaluation of shale pore structure using nuclear magnetic resonance data[J]. Chinese Journal of Magnetic Resonance, 2021, 38(2): 215-226.
[23] 闫伟林, 张兆谦, 陈龙川, 等. 基于核磁共振技术的古龙页岩含油饱和度评价新方法[J]. 大庆石油地质与开发, 2021, 40(5): 78-86.
YAN Weilin, ZHANG Zhaoqian, CHEN Longchuan, et al. New evaluating method of oil saturation in Gulong shale based on NMR technique[J]. Petroleum Geology & Oilfield Development in Daqing, 2021, 40(5): 78-86.
[24] 陈惠, 冯春珍, 赵建鹏, 等. 基于分形与核磁共振测井的储层孔隙结构表征与分类[J]. 测井技术, 2021, 45(1): 50-55.
CHEN Hui, FENG Chunzhen, ZHAO Jianpeng, et al. Pore structure characterization and classification based on fractal theory and nuclear magnetic resonance logging[J]. Well Logging Technology, 2021, 45(1): 50-55.
[25] 高新, 靳国旺, 熊新, 等. 融合差异图与高斯混合模型相结合的SAR图像变化检测[J]. 测绘科学技术学报, 2020, 37(1): 68-73.
GAO Xin, JIN Guowang, XIONG Xin, et al. Change detection in synthetic aperture radar images based on image fusion and Gaussian mixture model[J]. Journal of Geomatics Science and Technology, 2020, 37(1): 68-73.
[26] 戚贵玲, 何冰冰, 张榆锋, 等. 基于高斯混合模型聚类的B超图像颈动脉内膜和中膜厚度检测[J]. 生物医学工程学杂志, 2020, 37(6): 1080-1088.
QI Guiling, HE Bingbing, ZHANG Yufeng, et al. Detection of carotid intima and media thicknesses based on ultrasound B-mode images clustered with Gaussian mixture model[J]. Journal of Biomedical Engineering, 2020, 37(6): 1080-1088.
[27] 张美霞, 李丽, 杨秀, 等. 基于高斯混合模型聚类和多维尺度分析的负荷分类方法[J]. 电网技术, 2020, 44(11): 4283-4293.
ZHANG Meixia, LI Li, YANG Xiu, et al. A load classification method based on Gaussian mixture model clustering and multi-dimensional scaling analysis[J]. Power System Technology, 2020, 44(11): 4283-4293.
[28] 兰志刚, 靳卫卫, 朱明亮, 等. 基于高斯混合模型的海冰图像非监督聚类分割研究[J]. 海洋科学, 2011, 35(11): 97-100.
LAN Zhigang, JIN Weiwei, ZHU Mingliang, et al. Sea ice image segmentation with unsupervised clustering based on the Gaussian mixture model[J]. Marine Sciences, 2011, 35(11): 97-100.
[29] 张宝一, 陆浩, 杨莉, 等. 顾及梯度的高斯混合模型在三维属性场空间聚类中的应用[J]. 地质找矿论丛, 2019, 34(3): 460-470.
ZHANG Baoyi, LU Hao, YANG Li, et al. Application of the gradient-based Gaussian mixture model to spatial clustering of three-dimensional attribute field[J]. Contributions to Geology and Mineral Resources Research, 2019, 34(3): 460-470.
[30] 王亚利, 都伟冰, 王双亭. 高斯混合模型自动阈值法遥感冰川信息提取[J]. 遥感学报, 2021, 25(7): 1434-1444.
WANG Yali, DU Weibing, WANG Shuangting. Extracting glacier information from remote sensing imageries by automatic threshold method of Gaussian mixture model[J]. Journal of Remote Sensing, 2021, 25(7): 1434-1444.
[31] YANG M S, LAI C Y, LIN C Y. A robust EM clustering algorithm for Gaussian mixture models[J]. Pattern Recognition, 2012, 45(11): 3950-3961.
[32] NOWAKOWSKA E, KORONACKI J, LIPOVETSKY S. Clusterability assessment for Gaussian mixture models[J]. Applied Mathematics and Computation, 2015, 256: 591-601.
[33] 王一妹, 刘辉, 宋鹏, 等. 基于高斯混合模型聚类的风电场短期功率预测方法[J]. 电力系统自动化, 2021, 45(7): 37-43.
WANG Yimei, LIU Hui, SONG Peng, et al. Short-term power forecasting method of wind farm based on Gaussian mixture model clustering[J]. Automation of Electric Power Systems, 2021, 45(7): 37-43.
[34] 柴秀俊, 王宏伟, 王林, 等. 基于高斯混合聚类的切换系统的辨识[J]. 控制理论与应用, 2021, 38(5): 634-640.
CHAI Xiujun, WANG Hongwei, WANG Lin, et al. Identification of switched systems based on Gaussian mixture clustering[J]. Control Theory & Applications, 2021, 38(5): 634-640.
[35] FRIEL N, MCKEONE J P, OATES C J, et al. Investigation of the widely applicable Bayesian information criterion[J]. Statistics and Computing, 2017, 27(3): 833-844.
[36] 张发才, 李喜旺, 樊国旗. 基于高斯混合聚类的风电出力场景划分[J]. 计算机系统应用, 2021, 30(1): 146-153.
ZHANG Facai, LI Xiwang, FAN Guoqi. Wind power output scene division based on Gaussian hybrid clustering[J]. Computer Systems & Applications, 2021, 30(1): 146-153.
[37] GE X M, WANG H, FAN Y R, et al. Joint inversion of T1-T2 spectrum combining the iterative truncated singular value decomposition and the parallel particle swarm optimization algorithms[J]. Computer Physics Communications, 2016, 198: 59-70.
[38] 周海涛, 王志刚, 刘昌明. 基于主成分分析和高斯混合模型的耐火材料损伤信号分类[J]. 武汉科技大学学报, 2014, 37(4): 269-272.
ZHOU Haitao, WANG Zhigang, LIU Changming. Classification of refractory damage signals based on the principal component analysis and Gaussian mixture model[J]. Journal of Wuhan University of Science and Technology, 2014, 37(4): 269-272.
[39] 申波, 毛志强, 樊海涛, 等. 基于主成分分析技术计算蚀变地层孔隙度的新方法[J]. 测井技术, 2012, 36(2): 130-134.
SHEN Bo, MAO Zhiqiang, FAN Haitao, et al. A new porosity calculation method based on principal component analytical technology for altered formation[J]. Well Logging Technology, 2012, 36(2): 130-134.
[40] 钟仪华, 李榕, 张志银, 等. 基于主成分分析的离散过程神经网络水淹层动态预测方法[J]. 测井技术, 2010, 34(5): 432-436.
ZHONG Yihua, LI Rong, ZHANG Zhiyin, et al. Dynamic recognition method for water-flooded layer with discrete process neural network based on the principal component analysis[J]. Well Logging Technology, 2010, 34(5): 432-436.
文章导航

/