基于特征点和改进卷积神经网络的压裂水平井裂缝参数高效快速反演

doi:10.7623/syxb202509008

石油学报 ›› 2025, Vol. 46 ›› Issue (9): 1751-1763.DOI: 10.7623/syxb202509008

• 油田开发 • 上一篇

基于特征点和改进卷积神经网络的压裂水平井裂缝参数高效快速反演

陈志明^1,2, 朱海锋^1,2, 赵辉³, 董鹏^1,2, 廖新维^1,2, Yu Wei⁴

1. 中国石油大学(北京)石油工程学院北京 102249;
2. 中国石油大学(北京)人工智能学院北京 102249;
3. 长江大学石油工程学院湖北武汉 430100;
4. 美国得克萨斯大学奥斯汀分校得克萨斯州奥斯汀 78712

收稿日期:2024-05-23 修回日期:2025-05-09 发布日期:2025-10-11
通讯作者: 陈志明,男,1989年12月生,2018年获中国石油大学(北京)博士学位,现为中国石油大学(北京)教授,主要从事非常规油气藏试井反演理论与技术方法研究。
作者简介:陈志明,男,1989年12月生,2018年获中国石油大学(北京)博士学位,现为中国石油大学(北京)教授,主要从事非常规油气藏试井反演理论与技术方法研究。Email:zhimingchn@cup.edu.cn
基金资助:
国家自然科学基金项目(No.52074322,No.52274046)和北京市自然科学基金项目(3232027)资助。

Efficient and rapid inversion of fracture parameters of fracturing horizontal wells based on feature points and improved CNN

Chen Zhiming^1,2, Zhu Haifeng^1,2, Zhao Hui³, Dong Peng^1,2, Liao Xinwei^1,2, Yu Wei⁴

1. College of Petroleum Engineering, China University of Petroleum, Beijing 102249, China;
2. College of Artificial Intelligence, China University of Petroleum, Beijing 102249, China;
3. College of Petroleum Engineering, Yangtze University, Hubei Wuhan 430100, China;
4. University of Texas at Austin, Austin Texas 78712, USA

Received:2024-05-23 Revised:2025-05-09 Published:2025-10-11

摘要/Abstract

摘要： 压裂水平井能大幅提高低渗及特低渗油气藏的最终采收率,具有重要应用前景,而其裂缝参数反演是压裂评价和产能预测的前提。试井分析方法是裂缝参数反演的重要手段之一,但目前的方法以人工反演为主,其主观性和多解性强,为裂缝参数反演结果带来了极大不确定性。为解决人工反演耗时长、受试井工作者主观因素影响强、不同试井工作者水平差异大等问题,利用特征点法及改进卷积神经网络,基于压裂水平井油藏的半解析解形成一套智能参数反演方法。对压力数据进行归一化处理,先通过训练好的卷积网络反演改造区渗透率、无因次导流能力、裂缝半长及裂缝间距,再通过特征点法得到无因次井筒储集系数及表皮系数,验证结果表明无因次井筒储集系数、表皮系数、改造区渗透率、无因次导流能力、裂缝半长和缝间距的反演值与真实值的相对误差分别为5.00%、6.01%、7.96%、6.10%、4.70%、0.25%,平均相对误差为5.00%,反演曲线与真实曲线符合程度良好,验证了该方法的可靠性。对实际现场的两次压力恢复数据进行智能反演,并与人工反演进行对比,结果显示智能反演的无因次井筒储集系数下降39.37%、表皮系数增加15%、改造区渗透率下降26.53%、无因次导流能力下降19.45%、裂缝半长下降12.82%、有效井长下降30.29%,符合现场实际情况,从而验证了该方法的实用性。

关键词: 人工智能, 压裂水平井, 卷积神经网络, 特征点法, 试井解释, 参数反演

Abstract: Hydraulically fractured horizontal wells significantly enhance the ultimate recovery factor of low-permeability and ultra-low-permeability reservoirs, demonstrating critical application prospects in the development of such reservoirs. The inversion of fracture parameters serves as a prerequisite for fracturing evaluation and productivity prediction. The well test analysis method is one of the important methods for fracture parameter inversion, but the current approach mainly relies on manual inversion, which is highly subjective and suffers from non-uniqueness, introducing great uncertainties into inversion results. To address issues such as long inversion time, strong influence by subjective factors from well test workers, and significant disparities in professional proficiency, an intelligent parameter inversion method is developed based on the semi-analytical solution of fractured horizontal reservoirs, integrating the Feature Point Method and improved Convolutional Neural Network (CNN). The pressure data were normalized, and retrofit zone permeability, dimensionless conductivity, fracture half-length and fracture spacing were first inverted by the trained convolutional network, and then the dimensionless well reservoir coefficient and epidermal coefficient were obtained by the feature point method. The verification results show that the relative errors between the inverted and the true values of the the dimensionless wellbore storage coefficient, skin factor, stimulated zone permeability, dimensionless conductivity, fracture half-length, and fracture spacing are 5.0 %, 6.01 %, 7.96 %, 6.1 %, 4.7 % and 0.25 %, respectively, with an average relative error of 5.00 % . The inversion curve exhibits excellent agreement with the true curve, confirming the method's reliability. Field applications demonstrate that for two sets of pressure buildup data, the intelligent inversion reduces the inversion time by 39.37 % compared to manual methods. The skin factor increases by 15 %, stimulated zone permeability decreases by 26.53 %, passive diversion capacity drops by 19.45 %, fracture half-length shortens by 12.82 %, and effective wellbore length decreases by 30.29 %, which aligns with field realities, verifying the method’s practical applicability.

Key words: artificial intelligence, fractured horizontal well, convolutional neural networks, feature point method, well test explanation, parameter inversion

中图分类号:

TE353

陈志明, 朱海锋, 赵辉, 董鹏, 廖新维, Yu Wei. 基于特征点和改进卷积神经网络的压裂水平井裂缝参数高效快速反演[J]. 石油学报, 2025, 46(9): 1751-1763.

Chen Zhiming, Zhu Haifeng, Zhao Hui, Dong Peng, Liao Xinwei, Yu Wei. Efficient and rapid inversion of fracture parameters of fracturing horizontal wells based on feature points and improved CNN[J]. Acta Petrolei Sinica, 2025, 46(9): 1751-1763.

参考文献

[1] 胡素云,李建忠,王铜山,等. 中国石油油气资源潜力分析与勘探选区思考[J].石油实验地质,2020,42(5):813-823.
HU Suyun,LI Jianzhong,WANG Tongshan,et al.CNPC oil and gas resource potential and exploration target selection[J].Petroleum Geology & Experiment,2020,42(5):813-823.
[2] 罗红文,李海涛,李颖,等.低渗透气藏压裂水平井产出剖面与裂缝参数反演解释[J].石油学报,2021,42(7):936-947.
LUO Hongwen,LI Haitao,LI Ying,et al.Inversion and interpretation of production profile and fracture parameters of fractured horizontal wells in low-permeability gas reservoirs[J].Acta Petrolei Sinica,2021,42(7):936-947.
[3] 计秉玉,张文彪,何应付,等.油藏地质建模与数值模拟一体化内涵及发展趋势[J].石油学报,2024,45(7):1152-1162.
JI Bingyu,ZHANG Wenbiao,HE Yingfu,et al.Connotation and development trends of integration between geological reservoir modeling and numerical reservoir simulation[J].Acta Petrolei Sinica,2024,45(7):1152-1162.
[4] 刘合,任义丽,李欣,等.岩心智能识别技术内涵与展望[J].石油学报,2024,45(8):1296-1308.
LIU He,REN Yili,LI Xin,et al.Connotation and prospect of intelligent recognition technology for cores[J].Acta Petrolei Sinica,2024,45(8):1296-1308.
[5] 杨德彬,何新明,张恒,等.塔河油田主体区奥陶系表层岩溶带智能识别及缝洞发育规律[J].石油学报,2024,45(2):374-389.
YANG Debin,HE Xinming,ZHANG Heng,et al.Intelligent identification of Ordovician epikarst zones and development laws of fractures and vugs in the main area of Tahe oilfield[J].Acta Petrolei Sinica,2024,45(2):374-389.
[6] 殷启帅,张来斌,钟弘成,等.基于神经网络优化算法的系泊系统智能设计[J].石油学报,2025,46(5):967-976.
YIN Qishuai,ZHANG Laibin,ZHONG Hongcheng,et al.Intelligent analysis and design of tensioned mooring system based on backpropagation neural network and genetic algorithm[J].Acta Petrolei Sinica,2025,46(5):967-976.
[7] 王卫星,成绥民,李汝勇,等.人工智能在试井解释中的应用研究[J].西南石油学院学报,1991,13(2):31-37.
WANG Weixing,CHENG Suimin,LI Ruyong,et al.Use of artificial intelligence in well test interpretation[J].Journal of Southwest Petroleum Institute,1991,13(2):31-37.
[8] 成绥民,李论,孙志林,等.神经网络专家系统在试井解释中的应用[J].西安石油学院学报,1997,12(2):32-33.
CHENG Suimin,LI Lun,SUN Zhilin,et al.Application of neural networks and expert system in the interpretation of well test[J].Journal of Xi’an Petroleum Institute,1997,12(2):32-33.
[9] 胡泽,赵必荣.水平井试井解释自动参数识别新方法[J].天然气工业,1997,17(2):71-74.
HU Ze,ZHAO Birong.A new method for discrimination of automatic parameters in horizontal well test analysis[J].Natural Gas Industry,1997,17(2):71-74.
[10] 邓远忠,陈钦雷.试井解释图版拟合分析的神经网络方法[J].石油勘探与开发,2000,27(1):64-66.
DENG Yuanzhong,CHEN Qinlei.Using artificial neural network to realize type curve match analysis of well test data[J].Petroleum Exploration and Development,2000,27(1):64-66.
[11] AL-KAABI A A U,LEE W J.Using artificial neural nets to identify the well-test interpretation model[J].SPE Formation Evaluation,1993,8(3):233-240.
[12] VAFERI B,ESLAMLOUEYAN R,AYATOLLAHI S.Automatic recognition of oil reservoir models from well testing data by using multi -layer perceptron networks[J].Journal of Petroleum Science and Engineering,2011,77(3/4):254-262.
[13] ADIBIFARD M,TABATABAEI-NEJAD S A R,KHODAPANAH E.Artificial Neural Network (ANN)to estimate reservoir parameters in Naturally Fractured Reservoirs using well test data[J].Journal of Petroleum Science and Engineering,2014,122:585-594.
[14] GHAFFARIAN N,ESLAMLOUEYAN R,VAFERI B.Model identification for gas condensate reservoirs by using ANN method based on well test data[J].Journal of Petroleum Science and Engineering,2014,123:20-29.
[15] ALMARAGHI A M,EL-BANBI A H.Automatic reservoir model identification using artificial neural networks in pressure transient analysis[R].SPE-175850,2015.
[16] 李祯,郭奇,卜亚辉,等.基于深度学习的饱和度场样本库建立及预测[J].石油学报,2024,45(4):698-707.
LI Zhen,GUO Qi,BU Yahui,et al.Establishment and prediction of sample pool of saturation field based on deep learning[J].Acta Petrolei Sinica,2024,45(4):698-707.
[17] ALQAHTANI N,ARMSTRONG R T,MOSTAGHIMI P.Deep learning convolutional neural networks to predict porous media properties[R].SPE191906,2018.
[18] 赵春兰,屈瑶,王兵,等.一种基于2D-CNN深度学习的钻井事故等级预测新方法[J].天然气工业,2022,42(12):95-105.
ZHAO Chunlan,QU Yao,WANG Bing,et al.A new method for predicting drilling accident level based on 2D-CNN deep learning[J].Natural Gas Industry,2022,42(12):95-105.
[19] CHU Hongyang,LIAO Xinwei,DONG Peng,et al.An automatic classification method of well testing plot based on convolutional neural network (CNN)[J].Energies,2019,12(15):2846.
[20] LIU Xuliang,LI Daolun,YANG Jinghai,et al.Automatic well test interpretation based on convolutional neural network for infinite reservoir[J].Journal of Petroleum Science and Engineering,2020,195:107618.
[21] LI Daolun,LIU Xuliang,ZHA Wenshu,et al.Automatic well test interpretation based on convolutional neural network for a radial composite reservoir[J].Petroleum Exploration and Development,2020,47(3):623-631.
[22] DONG Peng,CHEN Zhiming,LIAO Xinwei,et al.Application of deep learning on well-test interpretation for identifying pressure behavior and characterizing reservoirs[J].Journal of Petroleum Science and Engineering,2022,208:109264.
[23] 齐占奎,张新鹏,刘旭亮,等.一种基于一维卷积神经网络的试井模型智能识别方法[J].油气井测试,2024,33(2):72-78.
QI Zhankui,ZHANG Xinpeng,LIU Xuliang,et al.An intelligent method for identifying well testing models based on one-dimensional convolutional neural network[J].Well Testing,2024,33(2):72-78.
[24] 王欢.低渗透油藏体积压裂井动态反演[D].北京:中国石油大学(北京),2015.
WANG Huan.Dynamic inversion of volumetric fracturing wells in low permeability reservoirs[D].Beijing:China University of Petroleum,2015.
[25] 梁光跃,廖新维,赵晓亮.特征点拟合法在非牛顿流试井解释中的应用[J].西南石油大学学报(自然科学版),2011,33(6):94-100.
LIANG Guangyue,LIAO Xinwei,ZHAO Xiaoliang.Characteristic points matching method applied in non-newtonian reservoir well testing[J].Journal of Southwest Petroleum University(Science & Technology Edition),2011,33(6):94-100.
[26] 廖新维,陈晓明,赵晓亮,等.低渗油藏体积压裂井压力特征分析[J].科技导报,2016,34(7):117-122.
LIAO Xinwei,CHEN Xiaoming,ZHAO Xiaoliang,et al.Pressure transient analysis of volume fracturing well in low permeability oil reservoir[J].Science & Technology Review,2016,34(7):117-122.
[27] ZHU Haifeng,CHEN Zhiming,ZHAO Xin,et al.Intelligent parameter inversions for activated fault based on well testing constraints and transfer learning[J].SPE Journal,2025,30(6):3896-3916.
[28] HAN Jiapeng,CHEN Zhiming,ZHOU Biao.Transient flow analysis of multistage fractured horizontal wells with varied natural fractures connection relationships[J].SPE Journal,2025,30(4):1984-2001.
[29] FU Zhuojia,CHEN Wen,YANG Haitian.Boundary particle method for Laplace transformed time fractional diffusion equations[J].Journal of Computational Physics,2013,235:52-66.
[30] 同登科,陈钦雷.关于Laplace数值反演Stehfest方法的一点注记[J].石油学报,2001,22(6):91-92.
TONG Dengke,CHEN Qinlei.A note on the Stehfest method of Laplace's numerical inversion[J].Acta Petrolei Sinica,2001,22(6):91-92.
[31] KINGMA D P,BA J.Adam:a method for stochastic optimization[C]//Proceedings of the 3rd International Conference on Learning Representations.San Diego:ICLR,2015.
[32] 陈志明,YU Wei,石璐铭,等.压裂水平井的多模式裂缝网络试井模型及参数评价——以吉木萨尔页岩油为例[J].石油学报,2023,44(10):1706-1726.
CHEN Zhiming,YU Wei,SHI Luming,et al.Well test model and parameter evaluation of multi-mode fracture network in fractured horizontal well:a case study of Jimsar shale oil[J].Acta Petrolei Sinica,2023,44(10):1706-1726.
[33] 雷群,翁定为,熊生春,等.中国石油页岩油储集层改造技术进展及发展方向[J].石油勘探与开发,2021,48(5):1035-1042.
LEI Qun,WENG Dingwei,XIONG Shengchun,et al.Progress and development directions of shale oil reservoir stimulation technology of China National Petroleum Corporation[J].Petroleum Exploration and Development,2021,48(5):1035-1042.
[34] 齐洪岩,王振林,张艳宁,等.吉木萨尔凹陷芦草沟组页岩油藏甜点分类[J].新疆石油地质,2025,46(2):127-135.
QI Hongyan,WANG Zhenlin,ZHANG Yanning,et al.Classification of sweet spots in shale oil reservoir of Lucaogou formation in Jimsar sag,Jurggar Basin[J].Xinjiang Petroleum Geology,2025,46(2):127-135.

基于特征点和改进卷积神经网络的压裂水平井裂缝参数高效快速反演

Efficient and rapid inversion of fracture parameters of fracturing horizontal wells based on feature points and improved CNN

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	胡连博, 陈志明, 廖新维, Yu Wei, Sepehrnoori Kamy. 基于"压—焖—采"多数据协同分析的页岩油井裂缝参数反演与产能预测[J]. 石油学报, 2025, 46(8): 1550-1567.
[2]	刘合, 任义丽, 李欣, 朱如凯, 胡延旭, 刘茜, 苏乾潇, 吴健平, 李彬. 岩心智能识别技术内涵与展望[J]. 石油学报, 2024, 45(8): 1296-1308.
[3]	计秉玉, 张文彪, 何应付, 段太忠, 刘合. 油藏地质建模与数值模拟一体化内涵及发展趋势[J]. 石油学报, 2024, 45(7): 1152-1162.
[4]	孙歧峰, 李克昊, 段友祥, 张依旻, 宫法明. 基于卷积神经网络与特征聚类的荧光薄片分析方法[J]. 石油学报, 2024, 45(3): 548-558.
[5]	赵娅, 管玉, 贾迪. 基于卷积神经网络与视觉变换器的微观剩余油分类方法[J]. 石油学报, 2024, 45(12): 1818-1832.
[6]	刘合, 李艳春, 贾德利, 王素玲, 乔美霞, 屈如意, 温鹏云, 任智慧. 人工智能在注水开发方案精细化调整中的应用现状及展望[J]. 石油学报, 2023, 44(9): 1574-1586.
[7]	刘喜武, 王心宇, 刘宇巍, 张金强, 刘炯, 刘倩. 中国陆相页岩油地震勘探技术现状及发展方向[J]. 石油学报, 2023, 44(12): 2270-2285.
[8]	陈志明, Yu Wei, 石璐铭, 胡连博, 廖新维. 压裂水平井的多模式裂缝网络试井模型及参数评价——以吉木萨尔页岩油为例[J]. 石油学报, 2023, 44(10): 1706-1726.
[9]	刘文岭, 韩大匡. 数字孪生油气藏:智慧油气田建设的新方向[J]. 石油学报, 2022, 43(10): 1450-1461.
[10]	孙金声, 刘凡, 程荣超, 冯杰, 郝惠军, 王韧, 白英睿, 刘钦政. 机器学习在防漏堵漏中研究进展与展望[J]. 石油学报, 2022, 43(1): 91-100.
[11]	罗红文, 李海涛, 李颖, 汪凤, 向雨行, 蒋贝贝, 于皓. 低渗透气藏压裂水平井产出剖面与裂缝参数反演解释[J]. 石油学报, 2021, 42(7): 936-947.
[12]	李宁, 徐彬森, 武宏亮, 冯周, 李雨生, 王克文, 刘鹏. 人工智能在测井地层评价中的应用现状及前景[J]. 石油学报, 2021, 42(4): 508-522.
[13]	王敏生, 光新军. 智能钻井技术现状与发展方向[J]. 石油学报, 2020, 41(4): 505-512.
[14]	姜瑞忠, 何吉祥, 姜宇, 范海军. 页岩气藏压裂水平井Blasingame产量递减分析方法建立与应用[J]. 石油学报, 2019, 40(12): 1503-1510.
[15]	欧阳伟平, 孙贺东, 张冕. 考虑应力敏感的致密气多级压裂水平井试井分析[J]. 石油学报, 2018, 39(5): 570-577.