基于支持向量机(SVM)的不同咸化程度烃源岩总有机碳含量预测模型

初勇志; 刘成林; 太万雪; 阳宏

doi:10.11781/sysydz202204739

基于支持向量机(SVM)的不同咸化程度烃源岩总有机碳含量预测模型

doi: 10.11781/sysydz202204739

初勇志^{1, 2,},
刘成林^{1, 2, ,},
太万雪^{1, 2},
阳宏^{1, 2}

1.
中国石油大学地球科学学院, 北京 102249
2.
中国石油大学(北京) 地球科学学院, 北京 102249

基金项目:

国家自然科学基金面上项目“咸化湖盆条件下盐类对地层超压的作用机制研究” 41872127

详细信息

作者简介:
初勇志(2000—), 男, 从事油气成藏和资源评价研究。E-mail: chuyongzhi18@163.com

通讯作者:
刘成林(1970—), 男, 博士, 教授, 博士生导师, 从事油气资源评价与非常规油气地质研究。E-mail: liucl@cup.edu.cn

中图分类号: TE122.1
计量
- 文章访问数: 297
- HTML全文浏览量: 61
- PDF下载量: 41
- 被引次数: 0
出版历程
- 收稿日期: 2021-12-12
- 修回日期: 2022-07-01
- 刊出日期: 2022-07-28

Prediction model of TOC contents in source rocks with different salinity degrees based on Support Vector Machine (SVM)

CHU Yongzhi^{1, 2
,},
LIU Chenglin^{1, 2
, ,},
TAI Wanxue^{1, 2},
YANG Hong^{1, 2}

1.
State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum (Beijing), Beijing 102249, China
2.
College of Geosciences, China University of Petroleum (Beijing), Beijing 102249, China

摘要

摘要: 总有机碳含量(TOC)作为评价烃源岩有机质丰度的重要参数，其精确预测对油气勘探开发具有重要意义。目前总有机碳含量预测以ΔlogR方法、多元回归分析等基于统计分析的方法为主，存在泛化能力弱、主观性强等问题。机器学习方法的引入，可有效解决这类非稳定性、非线性、高复杂性的问题，但当下的研究仍停留在方法的比较与选取层面，没有对优良模型进行深入分析并检验其适用性。采用应用效果更好的支持向量机模型进行总有机碳含量预测，选取渤海湾盆地渤中凹陷古近系东营组淡水湖相和柴达木盆地西部狮子沟地区古近系咸化湖相烃源岩作为研究对象，对模型的效果进行检验与对比。通过相关性和XGBoost特征重要性分析，选定声波时差(DT)、体积密度(DEN)、自然电位(SP)、自然伽马(GR)、深度等作为输入层，以总有机碳含量作为输出层，确立SVM烃源岩总有机碳含量预测模型。研究结果表明，模型在应用至差异较大的沉积环境时具有很强的泛化能力以适应不同地区的地质特征；测井曲线对于烃源岩有机质丰度的敏感性由于沉积环境存在差异而有所区别。该模型在渤海湾盆地淡水湖相区域的应用中相关性更高，误差更小。
- 总有机碳含量预测 /
- 测井资料 /
- 支持向量机 /
- 渤海湾盆地 /
- 柴达木盆地
Abstract: The total organic carbon (TOC) content is an important parameter for the evaluation of abundance of organic matter in source rocks, and its predicting accuracy is of great significance to oil and gas exploration and development. At present, TOC prediction is mainly based on statistical analysis methods such as ΔlogR method and multiple regression analysis, problems such as weak generalization ability and strong subjectivity exist. The introduction of machine learning methods can effectively solve these problems of instability, nonlinearity, and high complexity. However, current research remains at the level of method comparison and selection with no indepth analysis of good models and their applicability. In this paper, a Support Vector Machine (SVM) model with better application effects was used to predict TOC contents of source rocks with different salinity degrees. As source rocks of freshwater lacustrine facies, the Paleogene Dongying Formation in the Bozhong Sag of Bohai Bay Basin and Paleogene source rocks in the Shizigou area of the western Qaidam Basin as saline lacustrine facies source rocks were selected to test and compare the effectiveness of the model. Through correlation analysis and XGBoost feature importance analysis, the logging sonic differential time (DT), volume density (DEN), spontaneous potential (SP), Gamma ray (GR) and depth were selected as the input layer, while the TOC was used as the output layer to establish a TOC prediction model based on SVM. Results show a strong generalization ability when applied to different sedimentary environments. It can adapt to the geological characteristics of different regions. The sensitivity of logging curves to the abundance of organic matter in source rocks varies in different sedimentary environments, which makes the model more relevant when applying to the fresh water lacustrine facies area in the Bohai Bay Basin.
- TOC prediction /
- logging data /
- Support Vector Machine /
- Bohai Bay Basin /
- Qaidam Basin

HTML全文

图 1 渤海湾盆地渤中凹陷构造位置及新生界地层划分

据文献[11]，有修改。

Figure 1. Tectonic location and stratigraphic division of Cenozoic, Bozhong Sag, Bohai Bay Basin

下载: 全尺寸图片幻灯片

图 2 柴达木盆地狮子沟背斜构造位置及新生界地层分布

据文献[21], 有修改。

Figure 2. Distribution of Cenozioc in Shizigou anticline and its tectonic location, Qaidam Basin

下载: 全尺寸图片幻灯片

图 3 渤海湾盆地渤中凹陷和柴达木盆地狮子沟地区实测总有机碳含量与测井参数相关图

Figure 3. Measured TOC contents and logging parameters of samples from Bozhong Sag of Bohai Bay Basin and Shizigou area of Qaidam Basin

下载: 全尺寸图片幻灯片

图 4 渤海湾盆地渤中凹陷和柴达木盆地狮子沟地区总有机碳含量实测和预测数据相关性

Figure 4. Correlations of measured and predicted TOC contents of samples from Bozhong Sag of Bohai Bay Basin and Shizigou area of Qaidam Basin

下载: 全尺寸图片幻灯片

图 5 渤海湾盆地渤中凹陷烃源岩总有机碳含量预测模型BZ13-2-5井单井对比

Figure 5. Single well comparison of source rock TOC prediction model in Bozhong Sag, Bohai Bay Basin

下载: 全尺寸图片幻灯片

表 1 渤海湾盆地渤中凹陷烃源岩样品不同核函数预测效果对比

Table 1. Comparison of forecasting effects of different core functions applied on source rock samples from Bozhong Sag, Bohai Bay Basin

预测效果线性核函数多项式核函数高斯核函数

R² 0.95 0.53 0.92

MSE 0.04 0.41 0.06

MAE 0.17 0.46 0.21

下载: 导出CSV

参考文献(34)

[1]	SCHMOKER J W. Determination of organic-matter content of Appalachi and Evonian shales from Gamma-ray logs[J]. AAPG Bulletin, 1981, 65(7): 1285-1298.
[2]	秦建强, 付德亮, 钱亚芳, 等. 烃源岩有机质丰度预测的地球物理研究进展[J]. 石油物探, 2018, 57(6): 803-812. doi: 10.3969/j.issn.1000-1441.2018.06.002 QIN Jianqiang, FU Deliang, QIAN Yafang, et al. Progress of geophy-sical methods for the evaluation of TOC of source rock[J]. Geophysical Prospecting for Petroleum, 2018, 57(6): 803-812. doi: 10.3969/j.issn.1000-1441.2018.06.002
[3]	区舫, 杨辉廷, 黄晓兵. X气田飞仙关组礁滩相储层测井解释方法应用[J]. 油气藏评价与开发, 2021, 11(5): 744-752. doi: 10.13809/j.cnki.cn32-1825/te.2021.05.012 OU Fang, YANG Huiting, HUANG Xiaobing. Application of logging interpretation method for reef shoal reservoir in Feixianguan Formation of X Gas Field[J]. Reservoir Evaluation and Development, 2021, 11(5): 744-752. doi: 10.13809/j.cnki.cn32-1825/te.2021.05.012
[4]	PASSEY Q R, CREANEY S, KULLA J B, et al. A practical model for organic richness from porosity and resistivity logs[J]. AAPG Bulletin, 1990, 74(5): 1777-1794.
[5]	MENDELZON J D, TOKSOZ M N. Source rock characterization using multivariate analysis of log data[C]//SPWLA 26th Annual Logging Symposium. Texas: Society of Petrophysicists & Well-Log Analysts, 1985: 1-21.
[6]	赵兴齐, 陈践发, 郭望, 等. BP神经网络在西湖凹陷烃源岩评价中的应用[J]. 测井技术, 2013, 37(5): 567-571. doi: 10.3969/j.issn.1004-1338.2013.05.022 ZHAO Xingqi, CHEN Jianfa, GUO Wang, et al. The application of BP neural network to the source rocks evaluation in Xihu Sag[J]. Well Logging Technology, 2013, 37(5): 567-571. doi: 10.3969/j.issn.1004-1338.2013.05.022
[7]	王惠君, 赵桂萍, 李良, 等. 基于卷积神经网络(CNN)的泥质烃源岩TOC预测模型: 以鄂尔多斯盆地杭锦旗地区为例[J]. 中国科学院大学学报, 2020, 37(1): 103-112. https://www.cnki.com.cn/Article/CJFDTOTAL-ZKYB202001011.htm WANG Huijun, ZHAO Guiping, LI Liang, et al. TOC prediction model for muddy source rocks based on convolutional neural network (CNN): a case study of the Hangjinqi area of the Ordos Basin[J]. Journal of University of Chinese Academy of Sciences, 2020, 37(1): 103-112. https://www.cnki.com.cn/Article/CJFDTOTAL-ZKYB202001011.htm
[8]	石创, 朱俊章, 龙祖烈, 等. 基于概率神经网络的烃源岩TOC预测: 以珠江口盆地陆丰南区为例[J]. 断块油气田, 2019, 26(5): 561-565. https://www.cnki.com.cn/Article/CJFDTOTAL-DKYT201905005.htm SHI Chuang, ZHU Junzhang, LONG Zulie, et al. Prediction of total organic carbon in source rocks by probabilistic neural network: a case study of southern Lufeng area in Pearl River Mouth Basin[J]. Fault-Block Oil and Gas Field, 2019, 26(5): 561-565. https://www.cnki.com.cn/Article/CJFDTOTAL-DKYT201905005.htm
[9]	张成龙, 陶士振, 白斌, 等. 基于支持向量机模型的烃源岩有机碳含量预测: 以鄂尔多斯盆地为例[J]. 天然气地球科学, 2019, 30(5): 761-768. https://www.cnki.com.cn/Article/CJFDTOTAL-TDKX201905016.htm ZHANG Chenglong, TAO Shizhen, BAI Bin, et al. Source rock TOC content prediction based on the support vector machine model: an application in Ordos Basin[J]. Natural Gas Geoscience, 2019, 30(5): 761-768. https://www.cnki.com.cn/Article/CJFDTOTAL-TDKX201905016.htm
[10]	李欣, 李建忠, 杨涛, 等. 渤海湾盆地油气勘探现状与勘探方向[J]. 新疆石油地质, 2013, 34(2): 140-144. https://www.cnki.com.cn/Article/CJFDTOTAL-XJSD201302004.htm LIN Xin, LI Jianzhong, YANG Tao, et al. Oil-gas exploration status and future targets in Bohai Bay Basin[J]. Xinjiang Petroleum Geology, 2013, 34(2): 140-144. https://www.cnki.com.cn/Article/CJFDTOTAL-XJSD201302004.htm
[11]	夏庆龙, 徐长贵. 渤海海域复杂断裂带地质认识创新与油气重大发现[J]. 石油学报, 2016, 27(S1): 22-33. https://www.cnki.com.cn/Article/CJFDTOTAL-SYXB2016S1003.htm XIA Qinglong, XU Changgui. New geological understandings and major hydrocarbon discoveries in the complex fault zone of Bohai Sea[J]. Acta Petrolei Sinica, 2016, 27(S1): 22-33. https://www.cnki.com.cn/Article/CJFDTOTAL-SYXB2016S1003.htm
[12]	孙永河, 漆家福, 吕延防, 等. 渤中坳陷断裂构造特征及其对油气的控制[J]. 石油学报, 2008, 29(5): 669-675. https://www.cnki.com.cn/Article/CJFDTOTAL-SYXB200805008.htm SUN Yonghe, QI Jiafu, LV Yanfang, et al. Characteristics of fault structure and its control to hydrocarbon in Bozhong Depression[J]. Acta Petrolei Sinica, 2008, 29(5): 669-675. https://www.cnki.com.cn/Article/CJFDTOTAL-SYXB200805008.htm
[13]	HAO Fang, ZHOU Xinhuai, ZHU Yangming, et al. Charging of the Neogene Penglai 19-3 field, Bohai Bay Basin, China: oil accumulation in a young trap in an active fault zone[J]. AAPG Bulletin, 2009, 93(2): 155-179.
[14]	郭玉新. 渤中凹陷埕岛东坡东三段沉积物重力流类型及沉积模式[J]. 油气地质与采收率, 2021, 28(3): 14-24. https://www.cnki.com.cn/Article/CJFDTOTAL-YQCS202103003.htm GUO Yuxin. Types and sedimentary models of sediment gravity flows of Ed³ member in east slope of Chengdao area, Bozhong Sag[J]. Petroleum Geology and Recovery Efficiency, 2021, 28(3): 14-24. https://www.cnki.com.cn/Article/CJFDTOTAL-YQCS202103003.htm
[15]	牛成民, 王飞龙, 何将启, 等. 渤海海域渤中19-6潜山气藏成藏要素匹配及成藏模式[J]. 石油实验地质, 2021, 43(2): 259-267. doi: 10.11781/sysydz202102259 NIU Chengmin, WANG Feilong, HE Jiangqi, et al. Accumulation factor matching and model of Bozhong 19-6 buried hill gas reservoir, Bohai Sea area[J]. Petroleum Geology & Experiment, 2021, 43(2): 259-267. doi: 10.11781/sysydz202102259
[16]	林会喜, 熊伟, 王勇, 等. 济阳坳陷埕岛潜山油气成藏特征[J]. 油气地质与采收率, 2021, 28(1): 1-9. https://www.cnki.com.cn/Article/CJFDTOTAL-YQCS202101002.htm LIN Huixi, XIONG Wei, WANG Yong, et al. Hydrocarbon accumulation in Chengdao buried hill of Jiyang Depression[J]. Petroleum Geology and Recovery Efficiency, 2021, 28(1): 1-9. https://www.cnki.com.cn/Article/CJFDTOTAL-YQCS202101002.htm
[17]	谢玉洪. 渤海湾盆地渤中凹陷太古界潜山气藏BZ19-6的气源条件与成藏模式[J]. 石油实验地质, 2020, 42(5): 858-866. doi: 10.11781/sysydz202005858 XIE Yuhong. Gas resources and accumulation model of BZ19-6 Archean buried-hill large-scale gas reservoir in Bozhong Sag, Bohai Bay Basin[J]. Petroleum Geology & Experiment, 2020, 42(5): 858-866. doi: 10.11781/sysydz202005858
[18]	庞雄奇, 郭永华, 姜福杰, 等. 渤海海域优质烃源岩及其分布预测[J]. 石油与天然气地质, 2009, 30(4): 393-397. https://www.cnki.com.cn/Article/CJFDTOTAL-SYYT200904005.htm PANG Xiongqi, GUO Yonghua, JIANG Fujie, et al. High-quality source rocks and their distribution prediction in the Bohai Sea waters[J]. Oil & Gas Geology, 2009, 30(4): 393-397. https://www.cnki.com.cn/Article/CJFDTOTAL-SYYT200904005.htm
[19]	江涛, 李慧勇, 胡安文, 等. 渤中西洼东三段烃源岩特征与油气成藏模式[J]. 特种油气藏, 2017, 24(6): 12-17. https://www.cnki.com.cn/Article/CJFDTOTAL-TZCZ201706003.htm JIANG Tao, LI Huiyong, HU Anwen, et al. Source rock characterization and hydrocarbon accumulation of Dong3 Member in west Bozhong Sag[J]. Special Oil & Gas Reservoirs, 2017, 24(6): 12-17. https://www.cnki.com.cn/Article/CJFDTOTAL-TZCZ201706003.htm
[20]	王翔宇. 渤海湾盆地渤中凹陷渐新统东营组三段烃源岩预测及评价[D]. 荆州: 长江大学, 2019. WANG Xiangyu. Prediction and evaluation of the source rocks of the third member of the Oligocene Dongying Formation in the Bozhong Sag, Bohai Bay Basin[D]. Jingzhou: Yangtze University, 2019.
[21]	刘溪溪, 岳鑫, 袁文虎, 等. 柴达木盆地西部狮子沟背斜构造区深部卤水水化学特征及演化分析[J]. 盐湖研究, 2019, 27(1): 73-81. https://www.cnki.com.cn/Article/CJFDTOTAL-YHYJ201901010.htm LIU Xixi, YUE Xin, YUAN Wenhu, et al. Hydrochemical characte-ristics and evolutionary process of deep brines from Shizigou anticline structure in Qaidam Basin, China[J]. Journal of Salt Lake Research, 2019, 27(1): 73-81. https://www.cnki.com.cn/Article/CJFDTOTAL-YHYJ201901010.htm
[22]	魏学斌, 沙威, 沈晓双, 等. 柴达木盆地油气勘探历程与启示[J]. 新疆石油地质, 2021, 42(3): 302-311. https://www.cnki.com.cn/Article/CJFDTOTAL-XJSD202103007.htm WEI Xuebin, SHA Wei, SHEN Xiaoshuang, et al. Petroleum exploration history and enlightenment in Qaidam Basin[J]. Xinjiang Petroleum Geology, 2021, 42(3): 302-311. https://www.cnki.com.cn/Article/CJFDTOTAL-XJSD202103007.htm
[23]	舒豫川, 胡广, 庞谦, 等. 柴达木盆地咸湖相烃源岩特征: 以英西地区下干柴沟组上段为例[J]. 断块油气田, 2021, (2): 179-186. https://www.cnki.com.cn/Article/CJFDTOTAL-DKYT202102008.htm SHU Yuchuan, HU Guang, PANG Qian, et al. Characteristics of source rocks of salt lake facies in Qaidam Basin: taking upper member of Xiaganchaigou Formation in Yingxi region as an example[J]. Fault-Block Oil and Gas Field, 2021, (2): 179-186. https://www.cnki.com.cn/Article/CJFDTOTAL-DKYT202102008.htm
[24]	王琳霖, 于冬冬, 浮昀, 等. 柴达木盆地西部构造演化与差异变形特征及对油田水分布的控制[J]. 石油实验地质, 2020, 42(2): 186-192. doi: 10.11781/sysydz202002186 WANG Linlin, YU Dongdong, FU Yun, et al. Tectonic evolution and differential deformation controls on oilfield water distribution in western Qaidam Basin[J]. Petroleum Geology & Experiment, 2020, 42(2): 186-192. doi: 10.11781/sysydz202002186
[25]	沈亚, 李洪革, 管俊亚, 等. 柴西地区古近系—新近系含油凹陷构造特征与勘探领域[J]. 石油地球物理勘探, 2012, 47(S1): 111-117. https://www.cnki.com.cn/Article/CJFDTOTAL-SYDQ2012S1021.htm SHEN Ya, LI Hongge, GUAN Junya, et al. Structure features and exploration potential of the Paleogene-Neogene depressions in western Qaidam Basin[J]. Oil Geophysical Prospecting, 2012, 47(S1): 111-117. https://www.cnki.com.cn/Article/CJFDTOTAL-SYDQ2012S1021.htm
[26]	付锁堂, 马达德, 陈琰, 等. 柴达木盆地油气勘探新进展[J]. 石油学报, 2016, 37(S1): 1-10. https://www.cnki.com.cn/Article/CJFDTOTAL-SYXB2016S1001.htm FU Suotang, MA Dade, CHEN Yan, et al. New advance of petroleum and gas exploration in Qaidam Basin[J]. Acta Petrolei Sinica, 2016, 37(S1): 1-10. https://www.cnki.com.cn/Article/CJFDTOTAL-SYXB2016S1001.htm
[27]	中华人民共和国国家质量监督检验检疫总局. GB/T 19145-2003: 沉积岩中总有机碳的测定[S]. 北京: 中国标准出版社, 2003. General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China. GB/T 191452003: Determination of total organic carbon in sedimentary rock[S]. Beijing: China Standard Press, 2003.
[28]	SONDERGELD C H, NEWSHAM K E, COMISKY J T, et al. Petrophysical considerations in evaluating and producing shale gas resources[C]//SPE Unconventional Gas Conference. Pittsburgh, Pennsylvania, USA: Society of Petroleum Engineers, 2010.
[29]	VAPNIK V N. Estimation of dependences based on empirical data[M]. New York: Springer-Verlag, 1982.
[30]	VAPNIK V N. The nature of statistical learning theory[M]. New York: Springer-Verlag, 1995.
[31]	杨斌, 匡立春, 孙中春, 等. 一种用于测井油气层综合识别的支持向量机方法[J]. 测井技术, 2005, 29(6): 511-514. https://www.cnki.com.cn/Article/CJFDTOTAL-CJJS200506011.htm YANG Bin, KUANG Lichun, SUN Zhongchun, et al. On support vector machines method to identify oil & gas zone with logging and mudlog information[J]. Well Logging Technology, 2005, 29(6): 511-514. https://www.cnki.com.cn/Article/CJFDTOTAL-CJJS200506011.htm
[32]	李新虎. 基于不同测井曲线参数集的支持向量机岩性识别对比[J]. 煤田地质与勘探, 2007, 35(3): 72-76. https://www.cnki.com.cn/Article/CJFDTOTAL-MDKT200703021.htm LI Xinhu. Lithology identification methods contrast based on support vector machines at different well logging parameter[J]. Coal Geology & Exploration, 2007, 35(3): 72-76. https://www.cnki.com.cn/Article/CJFDTOTAL-MDKT200703021.htm
[33]	牟丹, 王祝文, 黄玉龙, 等. 基于SVM测井数据的火山岩岩性识别: 以辽河盆地东部坳陷为例[J]. 地球物理学报, 2015, 58(5): 1785-1793. https://www.cnki.com.cn/Article/CJFDTOTAL-DQWX201505028.htm MOU Dan, WANG Zhuwen, HUANG Yulong, et al. Lithological identification of volcanic rocks from SVM well logging data: case study in the eastern depression of Liaohe Basin[J]. Chinese Journal of Geophysics, 2015, 58(5): 1785-1793. https://www.cnki.com.cn/Article/CJFDTOTAL-DQWX201505028.htm
[34]	太万雪, 刘成林, 田继先, 等. 柴达木盆地西部古近系咸化湖盆烃源岩总有机碳含量预测[J]. 特种油气藏, 2021, 28(1): 74-80. https://www.cnki.com.cn/Article/CJFDTOTAL-TZCZ202101010.htm TAI Wanxue, LIU Chenglin, TIAN Jixian, et al. Prediction of total organic carbon content of source rocks in Paleogene salinized lake basin in western Qaidam Basin[J]. Special Oil & Gas Reservoirs, 2021, 28(1): 74-80. https://www.cnki.com.cn/Article/CJFDTOTAL-TZCZ202101010.htm