Please wait a minute...
E-mail Alert Rss
 
物探与化探  2021, Vol. 45 Issue (1): 18-28    DOI: 10.11720/wtyht.2021.1508
  地质调查·资源勘查 本期目录 | 过刊浏览 | 高级检索 |
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
郭建宏1,2(), 张占松1,2(), 张超谟1,2, 周雪晴1,2, 肖航1,2, 秦瑞宝3, 余杰3
1.长江大学 地球物理与石油资源学院,湖北 武汉 430100
2.长江大学 油气资源与勘探技术教育部重点实验室,湖北 武汉 430100
3.中海油研究总院,北京 100027
The exploration of predicting CBM content by geophysical logging data: A case study based on slope correlation random forest method
GUO Jian-Hong1,2(), ZHANG Zhan-Song1,2(), ZHANG Chao-Mo1,2, ZHOU Xue-Qing1,2, XIAO Hang1,2, QIN Rui-Bao3, YU Jie3
1. College of Physics and Petroleum Resources, Yangtze University, Wuhan 430100, China
2. Key Laboratory of Exploration Technologies for Oil and Gas Resources, Ministry of Education, Yangtze University, Wuhan 430100, China
3. CNOOC Research Institute, Beijing 100027, China
全文: PDF(3196 KB)   HTML
输出: BibTeX | EndNote (RIS)      
摘要 

煤层气含量是煤层勘探开发研究的重点参数之一,由于煤层气含量受多因素影响,能有效预测其含量至关重要。本文将斜率关联度法与随机森林算法相结合,以地球物理测井资料为基础进行煤层气含量预测。首先利用改进的斜率关联度法,计算得到对煤层气含量敏感的测井曲线,再利用交叉验证法探究合适的随机森林决策树个数,并结合选出的超参数利用随机森林算法预测煤层气含量。以沁水煤田柿庄北区3号层为例,对该区块进行评价预测,并将预测结果与多元回归模型拟合结果进行对比,同时对本文方法模型的泛化性进行研究分析。结果表明,应用斜率关联度法对测井曲线与煤层气含量进行分析计算能准确有效地找到可用于煤层气含量预测的测井曲线;用随机森林算法训练得到的模型预测非夹矸段煤岩的煤层气含量准确,计算结果可信度高,在夹矸段预测能力较弱,总体对煤层气勘探开发有指导意义,具有实际应用价值。
服务
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章
郭建宏
张占松
张超谟
周雪晴
肖航
秦瑞宝
余杰
关键词 ： 煤层气含量,  斜率关联度法,  测井曲线,  随机森林,  地球物理测井资料    
Abstract：

Coalbed methane content is one of the key parameters in coal seam exploration and development research. Due to the influence of many factors on coalbed methane content, it is very important to predict coalbed methane content effectively. In this paper, slope correlation degree method and random forest algorithm are combined to predict coalbed methane content based on geophysical logging data. Firstly, the improved slope correlation degree method is used to obtain the favorable geophysical logging curves for CBM content prediction, and then the cross validation method is used to explore the appropriate number of random forest decision trees, and the random forest algorithm is used to predict the coalbed methane content for the logging curve sequence with positive correlation. With the No.3 seam in Shizhuang north area of Qinshui coalfield as an example, the block was evaluated and predicted with the results compared with the results of multiple regression model, and the anti-interference ability of the model was studied and analyzed. The results show that the application of slope correlation method to analyzing and calculating the geophysical logging curve and coalbed methane content can accurately and effectively find the logging curve that can be used to predict the content of coalbed methane, the model trained by random forest algorithm is accurate in predicting the content of coalbed methane in the non-gangue section, and the calculation result has high reliability, but the prediction ability is weak in the gangue section. The results obtained by the authors are of guiding significance to the exploration and development of coalbed methane and have practical application value.
Key words： coalbed methane content    slope correlation method    logging curve    random forest    geophysical logging data
收稿日期: 2020-11-08      修回日期: 2020-11-19      出版日期: 2021-02-20
ZTFLH:  P631  
基金资助:国家科技重大专项(2016ZX05060001-012)
通讯作者: 张占松
作者简介: 郭建宏(1997-),男,山东招远人,主要研究方向为测井方法与解释、煤层气测井智能评价。Email:87942024@qq.com
引用本文:   
郭建宏, 张占松, 张超谟, 周雪晴, 肖航, 秦瑞宝, 余杰. 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 Jian-Hong, ZHANG Zhan-Song, ZHANG Chao-Mo, ZHOU Xue-Qing, XIAO Hang, QIN Rui-Bao, YU Jie. The exploration of predicting CBM content by geophysical logging data: A case study based on slope correlation random forest method. Geophysical and Geochemical Exploration, 2021, 45(1): 18-28.
链接本文:  
https://www.wutanyuhuatan.com/CN/10.11720/wtyht.2021.1508      或      https://www.wutanyuhuatan.com/CN/Y2021/V45/I1/18
Fig.1  煤层气含量与测井参数间的关系
参数 测试气量/
(m3·t-1)		 自然伽马/
API		 自然电位/
mV		 补偿密度/
(g·cm-3)		 声波时差/
(μs·m-1)		 补偿中子/
(V·V-1)		 深电阻率/
(Ω·m)		 浅电阻率/
(Ω·m)
范围 5.91~26.07 30.4~109.4 26~134 1.19~1.89 384~489 0.41~0.56 165~33766 470~18355
平均值 18.83 55.3 82 1.39 419 0.49 4845 4428
Table 1  3号煤层测井响应范围
样号 测试气量/
(m3·t-1)		 深度曲线/
m		 自然伽马/
API		 自然电位/
mV		 补偿密度/
(g·cm-3)		 声波时差/
(μs·m-1)		 补偿中子/
(V·V-1)		 深电阻率/
(Ω·m)		 浅电阻率/
(Ω·m)
A1-1 17.32 972.94 52.96 134.34 1.21 395.9 0.47 11946 9024
A1-2 18.93 975.28 59.63 41.22 1.27 418.5 0.44 5121 3892
A1-3 15.90 976.00 45.51 42.10 1.20 415.1 0.46 5224 4563
? ?
A9-2 7.81 956.55 77.90 25.93 1.38 411.0 0.44 1514 88
Table 2  3号煤层斜率关联度计算样本
γ(x0,x1)
深度曲线		 γ(x0,x2)
自然伽马		 γ(x0,x3)
自然电位		 γ(x0,x4)
声波时差		 γ(x0,x5)
补偿密度		 γ(x0,x6)
补偿中子		 γ(x0,x7)
深电阻率		 γ(x0,x8)
浅电阻率
关联度 0.134 0.056 -0.049 0.163 0.183 0.168 0.196 -0.076
关联序 5 6 7 4 2 3 1 8
Table 3  3号煤层斜率关联度计算结果
Fig.2  斜率关联度计算前后随机森林袋外误差结果
Fig.3  交叉验证探究决策树范围结果
Fig.4  决策树个数为500时袋外误差
Fig.5  斜率关联度—随机森林预测煤层气含量结果
测试气量/(m3·t-1) 预测气量/(m3·t-1) 绝对误差/(m3·t-1) 相对误差/%
A1-2 18.93 16.67 2.26 0.12
A7-5 23.53 18.93 4.60 0.20
A7-3 18.83 19.90 1.07 0.06
A7-2 18.91 19.19 0.28 0.01
A2-6 19.07 17.66 1.41 0.07
A8-5 22.44 20.79 1.65 0.07
A6-2 15.14 17.77 2.63 0.17
A4-2 17.76 19.08 1.32 0.07
A3-1 15.96 20.59 4.63 0.29
A4-3 18.77 18.10 0.67 0.04
平均值 2.05 11.10
Table 4  3号煤层测试集预测结果
Fig.6  A7井3号煤层气含量预测成果
Fig.7  A3井3号煤层响应与实验值分析
[1] 赵庆波. 中国煤层气地质特征及勘探新领域[J]. 天然气工业, 2004,24(5):4-7.
[1] Zhao Q B. Geological features of the coalbed methane in China and its new exploration domains[J]. Natural Gas Industry, 2004,24(5):4-7.
[2] 孟召平, 田永东, 雷旸. 煤层含气量预测的BP神经网络模型与应用[J]. 中国矿业大学学报, 2008(4):28-33.
[2] Meng Z P, Tian Y D, Lei Y. Prediction models of coal bed gas content based on BP neural networks and its applications[J]. Journal of China University of Mining & Technology, 2008(4):28-33.
[3] 连承波, 赵永军, 李汉林, 等. 煤层含气量的主控因素及定量预测[J]. 煤炭学报, 2005,30(6):726-729.
[3] Lian C B, Zhao Y J, Li H L, et al. Main controlling factors analysis and prediction of coal bed gas content[J]. Journal of China Coal Society, 2005,30(6):726-729.
[4] 娄剑青. 影响煤层气井产量的因素分析[J]. 天然气工业, 2004,24(4):62-64.
[4] Lou J Q. Factors of influencing production of coal bed gas wells[J]. Natrual Gas Industry, 2004,24(4):62-64.
[5] 高波, 马玉贞, 陶明信, 等. 煤层气富集高产的主控因素[J]. 沉积学报, 2003,21(2):345-349.
[5] Gao B, Ma Y Z, Tao M X, et al. Main controlling factors analysis of enrichment condition of coalbed methane[J]. Acta Sedimentologica Sinica, 2003,21(2):345-349.
[6] 吴永平, 李仲东, 王允诚. 煤层气储层异常压力的成因机理及受控因素[J]. 煤炭学报, 2006,31(4):475-479.
[6] Wu Y P, Li Z D, Wang Y C. The formation mechanisms of abnormal pressure and factor in control of the coal bed gas in Qinshui Basin[J]. Journal of China Coal Society, 2006,31(4):475-479.
[7] 叶建平, 武强, 王子和. 水文地质条件对煤层气赋存的控制作用[J]. 煤炭学报, 2001,26(5):63-67.
[7] Ye J P, Wu Q, Wang Z H. Controlled characteristics of hydrogeological conditions on the coalbed methane migration and accumulation[J]. Journal of China Coal Society, 2001,26(5):63-67.
[8] 陈跃, 汤达祯, 许浩, 等. 基于测井信息的韩城地区煤体结构的分布规律[J]. 煤炭学报, 2013,38(8):1435-1442.
[8] Chen Y, Tang D Z, Xu H, et al. The distribution of coal structure in Hancheng based on well logging data[J]. Journal of China Coal Society, 2013,38(8):1435-1442.
[9] 李贵红, 张鸿, 崔永君, 等. 基于多元逐步回归分析的煤储层含气量预测模型——以沁水盆地为例[J]. 煤田地质与勘探, 2005,33(2):22-25.
[9] Li G H, Zhang H, Cui Y J, et al. A predictive model of gas content in coal reservoirs based on multiple stepwise regression analysis: A case study from Qinshui Basin[J]. Coal Geology & Exploration, 2005,33(2):22-25.
[10] Kim A G. Estimating methane content of bituminous coal beds from adsorption data[R]. United States Department of the Interior, Report of Investigations-Bureau of Mines 8245, 1977: 1-11.
[11] Ahmed U, Johnston D, Colson L. An advanced and integrated approach to coal formation evaluation[C]//SPE22736, 1991, 755-770.
[12] Hawkins J M, Schraufnagel R A, Olszewsk A J. Estimating coal bed gas content and sorption isotherm using well log data[C]// SPE24905, 1992: 491-501.
[13] 邵先杰, 孙玉波, 孙景民, 等. 煤岩参数测井解释方法——以韩城矿区为例[J]. 石油勘探与开发, 2013,40(5):559-565.
[13] Shao X J, Sun Y B, Sun J M, et al. Logging interpretation of coal petrologic parameters: A case study of Hancheng mining area[J]. Petroleum Exploration and Development, 2013,40(5):559-565.
[14] 董红, 侯俊胜, 李能根, 等. 煤层煤质和含气量的测井评价方法及其应用[J]. 物探与化探, 2001,25(2):138-143.
[14] Dong H, Hou J S, Li N G, et al. The logging evaluation method for coal quality and methane[J]. Geophysical and Geochemical Exploration, 2001,25(2):138-143.
[15] 田敏, 赵永军, 颛孙鹏程. 灰色系统理论在煤层气含量预测中的应用[J]. 煤田地质与勘探, 2008,36(2):24-27.
[15] Tian M, Zhao Y J, Zhuansun P C. Application of grey system theroy in prediction of coalbed methane content[J]. Coal Geology & Exploration, 2008,36(2):24-27.
[16] 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.
[16] Guo J H, Zhang Z S, Zhang C M, et al. Prediction and application of coalbed methane content based on grey system and logging method[J]. Geophysical & Geochemical Exploration, 2020,44(5):1190-1200.
[17] 梁亚林, 原文涛. 测井预测煤层气含量及分布规律——以山西省沁水煤田为例[J]. 物探与化探, 2018,42(6):1144-1149.
[17] Liang Y L, Yuan W T. The prediction of the content and distribution of coalbed gas: A case study in the Qinshui coalfield based on logging[J]. Geophysical and Geochemical Exploration, 2018,42(6):1144-1149.
[18] 黄兆辉, 邹长春, 杨玉卿, 等. 沁水盆地南部TS地区煤层气储层测井评价方法[J]. 现代地质, 2012,26(6):1275-1282.
[18] Huang Z H, Zou C C, Yang Y Q, et al. Coal bed methane reservoir evaluation from wireline logs in TS District, southern Qinshui Basin[J]. Geoscience, 2012,26(6):1275-1282.
[19] 金泽亮, 薛海飞, 高海滨, 等. 煤层气储层测井评价技术及应用[J]. 煤田地质与勘探, 2013,41(2):42-45.
[19] Jin Z L, Xue H F, Gao H B, et al. Technology for evaluation of CBM reservoir logging and its application[J]. Coal Geology & Exploration, 2013,41(2):42-45.
[20] 潘和平, 黄智辉. 煤层含气量测井解释方法探讨[J]. 煤田地质与勘探, 1998,26(2):58-60.
[20] Pan H P, Huang Z H. Discussion on the interpretation method of coalbed methane content[J]. Coal Geology & Exploration, 1998,26(2):58-60.
[21] 吴东平, 吴春萍, 岳晓燕. 煤层气测井评价的神经网络技术[J]. 天然气勘探与开发, 2001,24(1):31-34.
[21] Wu D P, Wu C P, Yue X Y. Neural network of coal bed gas logging evaluation[J]. Natural Gas Exploration & Development, 2001,24(1):31-34.
[22] 连承波, 赵永军, 李汉林, 等. 基于支持向量机回归的煤层含气量预测[J]. 西安科技大学学报, 2008,28(4):707-709.
[22] Lian C B, Zhao Y J, Li H L, et al. Prediction of coal bed gas content based on support vector machine regression[J]. Journal Center of Xi’an University of Science and Technology, 2008,28(4):707-709.
[23] Breiman L. Bagging predictors[J]. Machine Learning, 1996,24(2):123-140.
[24] 冯明刚, 严伟, 葛新民, 等. 利用随机森林回归算法预测总有机碳含量[J]. 矿物岩石地球化学通报, 2018,37(3):475-481.
[24] Feng M G, Yan W, Ge X M, et al. Predicting total organic carbon content by random forest regression algorithm[J]. Bulletin of Mineralogy,Petrology and Geochemistry, 2018,37(3):475-481.
[25] 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.
[25] Xiao X P, Xie L C, Huang D R. A modified computation method of grey correlation degree and its application[J]. Journal of Applied Statistics and Management, 1995,14(5):27-30.
[26] 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.
[26] Ma B G, Cheng G Q. A formula of similarity correlation degree[J]. Systems Engineering-Theory & Practice, 2000(7):69-71.
[27] 李明凉. 灰色关联度新判别准则及其计算公式[J]. 系统工程, 1998,16(1):68-70.
[27] Li M L. A new descriminant byelaw for grey interconnet degree and its calculation formulas[J]. Systems Engineering, 1998,16(1):68-70.
[28] 张绍良, 张国良. 灰色关联度计算方法比较及存在问题分析[J]. 系统工程, 1996,14(3):45-49.
[28] Zhang S L, Zhang G L. Comparison between computation modles of grey interconnet degree and analysis on their shortages[J]. Systems Engineering, 1996,14(3):45-49.
[29] Breiman L, Cutler A. Random forests[J]. Machine Learning, 2001,45(1):5-32.
doi: 10.1023/A:1010933404324
[30] Ho T K. The random subspace method for constructing decision forests[J]. IEEE Transactions on Pattern Analysisand Machine Intelligence, 1998,20(8):832-844.
[31] 贾承造, 郑民, 张永峰. 中国非常规油气资源与勘探开发前景[J]. 石油勘探与开发, 2012,39(2):129-136.
[31] Jia C Z, Zheng M, Zhang Y F. Unconventional hydrocarbon resources in China and the prospect of exploration and development[J]. Petroleum Exploration and Development, 2012,39(2):129-136.
[32] 雍世和, 张超馍. 测井数据处理与综合解释[M]. 东营: 中国石油大学出版社, 2007: 134-139.
[32] Yong S H, Zhang C M. Logging data processing and comprehensive interpretation[M]. Dongying: China University of Petroleum Press, 2007: 134-139.
[1] 陈超群, 戴慧敏, 冯雨林, 杨泽, 杨佳佳. 基于Sentinel-2A的孙吴地区土壤有机质反演研究[J]. 物探与化探, 2022, 46(5): 1141-1148.
[2] 赵理芳, 李爱勇, 王导丽, 张明鹏, 周锡明. 基于电阻率测井曲线的大地电磁测深标定[J]. 物探与化探, 2022, 46(3): 737-742.
[3] 郭建宏, 杜婷, 张占松, 肖航, 秦瑞宝, 余杰, 王灿. 基于支持向量机与地球物理测井资料的煤体结构识别方法[J]. 物探与化探, 2021, 45(3): 768-777.
[4] 刘伟男, 张超谟, 朱林奇, 胡松, 孔政, 邓瑞. 页岩气水平井TOC测井评价新方法[J]. 物探与化探, 2021, 45(2): 423-431.
[5] 彭刘亚, 解惠婷, 冯伟栋. 基于随机森林算法的砂土液化预测方法[J]. 物探与化探, 2020, 44(6): 1429-1434.
[6] 郭建宏, 张占松, 张超谟, 陈芷若, 张鹏浩, 汤潇, 秦瑞宝, 余杰. 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.
[7] 李江, 智敏, 朱书阶. 岩浆岩地震波阻抗反演与厚度预测[J]. 物探与化探, 2020, 44(5): 1233-1238.
[8] 张瑞, 陈刚, 潘保芝, 蒋必辞, 杨雪, 刘丹. 基于细菌觅食优化广义回归神经网络的煤层气含量预测[J]. 物探与化探, 2016, 40(2): 327-332.
[9] 赵莹. 小波分析在松辽盆地北部高分辨率层序地层学中的应用[J]. 物探与化探, 2013, 37(2): 310-313.
[10] 杨孛, 李瑞, 杨滔. 测井曲线构建技术在BMM地区的应用[J]. 物探与化探, 2013, 37(2): 333-337.
[11] 戴方尧, 崔若飞, 陈同俊. 多参数岩性反演在煤田地震勘探中的应用[J]. 物探与化探, 2013, 37(1): 104-107.
[12] 王明健, 国景星, 袭著纲.
惠民凹陷唐庄地区沙河街组层序地层和沉积相[J]. 物探与化探, 2010, 34(1): 49-53.
[13] 侯伯刚, 刘文岭, 罗娜. 地震反演中测井数据的预处理[J]. 物探与化探, 2009, 33(3): 331-336.
[14] 王艳梅, 邓霜岭, 黄元清, 陈传发, 薛勇. 含煤层系测井曲线特征及其沉积环境[J]. 物探与化探, 2009, 33(3): 290-293.
[15] 冯宝华. 火山沉积岩及其矿床研究中伽马测井曲线解释的误区[J]. 物探与化探, 2008, 32(2): 171-174.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
京ICP备05055290号-3
版权所有 © 2021《物探与化探》编辑部
通讯地址：北京市学院路29号航遥中心 邮编：100083
电话：010-62060192；62060193 E-mail：whtbjb@sina.com