物理信息神经网络驱动的井筒温度场求解方法

doi:10.7623/syxb202502009

石油学报 ›› 2025, Vol. 46 ›› Issue (2): 413-425.DOI: 10.7623/syxb202502009

物理信息神经网络驱动的井筒温度场求解方法

刘雪琪¹, 王志远¹, 魏强², 王敏³, 孙小辉⁴, 王雪瑞⁴, 张剑波¹, 尹邦堂¹, 孙宝江¹

1. 中国石油大学(华东)石油工程学院山东青岛 266580;
2. 中国石油塔里木油田公司轮南采油气管理区新疆库尔勒 841000;
3. 中国石油塔里木油田公司克拉采油气管理区新疆库尔勒 841000;
4. 中国石油大学(华东)计算机科学与技术学院山东青岛 266580

收稿日期:2024-04-09 修回日期:2024-10-24 出版日期:2025-03-13 发布日期:2025-03-13
通讯作者: 王志远,男,1981年12月生,2009年获中国石油大学(华东)油气井工程专业博士学位,现为中国石油大学(华东)教授、博士生导师,主要从事油气井多相流动理论及应用等方面的研究工作。Email:wangzy1209@126.com
作者简介:刘雪琪,男,1998年11月生,2021年获中国石油大学(华东)海洋油气工程专业学士学位,现为中国石油大学(华东)石油与天然气工程专业博士研究生,主要从事油气井多相流动理论及应用、井筒传热、天然气水合物开采等方面的研究。Email:lxq6snow7@163.com
基金资助:
国家自然科学基金重点项目"超深气井生产管柱泄漏精准识别与压力管控"(No.52434002)、国家自然科学基金基础科学中心项目"超深特深层油气钻采流动调控"(No.52288101)、国家重点研发计划项目"压井处置智能决策支持与控制系统"(2023YFC3009204)、国家自然科学基金联合基金项目"深水控压钻完井地层-井筒多场耦合机理与压力调控"(No.U21B2069)和山东省重大科技创新工程项目"深水复杂钻井多相流动模拟关键技术与监测装备"(2022CXGC020407)资助。

A method for solving wellbore temperature field driven by physical information neural network

Liu Xueqi¹, Wang Zhiyuan¹, Wei Qiang², Wang Min³, Sun Xiaohui⁴, Wang Xuerui⁴, Zhang Jianbo¹, Yin Bangtang¹, Sun Baojiang¹

1. School of Petroleum Engineering, China University of Petroleum, Shandong Qingdao 266580, China;
2. Lunnan Oil and Gas Production Management Area, PetroChina Tarim Oilfield Company, Xinjiang Korla 841000, China;
3. Kela Oil and Gas Production Management Area, PetroChina Tarim Oilfield Company, Xinjiang Korla 841000, China;
4. College of Computer Science and Technology, China University of Petroleum, Shandong Qingdao 266580, China

Received:2024-04-09 Revised:2024-10-24 Online:2025-03-13 Published:2025-03-13

摘要/Abstract

摘要： 深水、深层油气钻探过程中对井筒温度场计算的实时性要求高,高精度、高效率的井筒温度场求解方法是精确计算流体物性、精细保障井筒流动安全的关键。将井筒温度场模型以损失函数形式嵌入神经网络,利用自适应权重和自适应学习率的优化方法提高训练效率, 建立了物理信息神经网络驱动的井筒温度场求解方法,分析了钻井和气井测试期间井筒温度的瞬态变化。研究结果表明:钻井期间,与有限差分算法相比,钻杆温度和环空温度的平均误差分别为0.847 % 和0.725 % ,井底温度和井口温度的平均误差分别为0.162 % 和1.047 % ,计算效率提高约150倍;与现场实测数据对比,物理信息神经网络驱动的预测解与有限差分数值解的平均误差分别为2.16 % 和2.27 % ,规避偏微分方程的截断误差有助于提高模型精度;气井测试2 d内,天然气水合物生成风险的推断时间为0.728 1 s,该方法可应用于水合物生成区域的快速预测。提出的求解方法在保证计算精度的同时,可大幅度提高计算速度。

关键词: 钻井, 气井测试, 井筒温度, 物理信息神经网络, 快速预测

Abstract: In the drilling process of deepwater and deep oil and gas, there is a high demand for real-time calculation of the wellbore temperature field. Therefore, a high-precision and high-efficiency wellbore temperature field solution method is the key to accurately calculate fluid properties and precisely guarantee the safety of wellbore flow. In this study, a wellbore temperature field model is embedded into the neural network in the form of loss function, and the optimization method of self-adaptive weight and self-adaptive learning rate is used to improve the training efficiency. Further, the paper establishes a method for solving the wellbore temperature field driven by physical information neural network, and analyzes the transient changes in wellbore temperature during drilling and gas well testing. The results show that during drilling, the average errors of drill pipe temperature and annular temperature are 0.847 % and 0.725 % , respectively, and those of bottom hole temperature and wellhead temperature are 0.162 % and 1.047 % , respectively, from which it can be seen that the computational efficiency is improved by about 150 times when compared with the finite difference algorithm. Compared with the field measurments, the average errors of the predicted solution driven by the physical information neutral network and the finite difference numerical solution are 2.16 % and 2.27 % , respectively, and the model accuracy can be improved by avoiding the truncation errors in partial differential equations. During the gas well testing for two days, the inferred time for the risk of natural gas hydrate formation is 0.728 1 s, and this method can be applied to quickly predict the hydrate formation areas. In conclusion, the proposed solution method can not only ensure the calculation accuracy, but also significantly improve the computational speed.

Key words: drilling, gas well testing, wellbore temperature, physical information neural network, quick prediction

中图分类号:

TE21

刘雪琪, 王志远, 魏强, 王敏, 孙小辉, 王雪瑞, 张剑波, 尹邦堂, 孙宝江. 物理信息神经网络驱动的井筒温度场求解方法[J]. 石油学报, 2025, 46(2): 413-425.

Liu Xueqi, Wang Zhiyuan, Wei Qiang, Wang Min, Sun Xiaohui, Wang Xuerui, Zhang Jianbo, Yin Bangtang, Sun Baojiang. A method for solving wellbore temperature field driven by physical information neural network[J]. Acta Petrolei Sinica, 2025, 46(2): 413-425.

参考文献

[1] 杨进,李磊,宋宇,等.中国海洋油气钻井技术发展现状及展望[J].石油学报,2023,44(12):2308-2318.
YANG Jin,LI Lei,SONG Yu,et al.Current status and prospects of offshore oil and gas drilling technology development in China[J].Acta Petrolei Sinica,2023,44(12):2308-2318.
[2] 贾承造.中国石油工业上游前景与未来理论技术五大挑战[J].石油学报,2024,45(1):1-14.
JIA Chengzao.Prospects and five future theoretical and technical challenges of the upstream petroleum industry in China[J].Acta Petrolei Sinica,2024,45(1):1-14.
[3] 刘岩生,张佳伟,黄洪春.中国深层—超深层钻完井关键技术及发展方向[J].石油学报,2024,45(1):312-324.
LIU Yansheng,ZHANG Jiawei,HUANG Hongchun.Key technologies and development direction for deep and ultra-deep drilling and completion in China[J].Acta Petrolei Sinica,2024,45(1): 312-324.
[4] 李军,杨宏伟,张辉,等.深水油气钻采井筒压力预测及其控制研究进展[J].中国科学基金,2021,35(6):973-983.
LI Jun,YANG Hongwei,ZHANG Hui,et al.Progress of basic research on wellbore pressure control in deepwater oil and gas drilling and production[J].Bulletin of National Natural Science Foundation of China,2021,35(6):973-983.
[5] 徐宝昌,尤香凝,孟卓然,等.基于回压激励响应的实时地层压力反演方法[J].石油学报,2024,45(6):1009-1018.
XU Baochang,YOU Xiangning,MENG Zhuoran,et al.Real-time formation pressure inversion method based on back-pressure excitation response[J].Acta Petrolei Sinica,2024,45(6):1009-1018.
[6] 汪海阁,黄洪春,毕文欣,等.深井超深井油气钻井技术进展与展望[J].天然气工业,2021,41(8):163-177.
WANG Haige,HUANG Hongchun,BI Wenxin,et al.Deep and ultra-deep oil/gas well drilling technologies:progress and prospect[J].Natural Gas Industry,2021,41(8):163-177.
[7] 夏顺雷,李军,柳贡慧,等.基于无迹卡尔曼滤波的井筒压力实时校正模型[J].石油机械,2022,50(9):10-18.
XIA Shunlei,LI Jun,LIU Gonghui,et al.Real-time correction model of wellbore pressure based on unscented Kalman filter[J].China Petroleum Machinery,2022,50(9):10-18.
[8] CAO Lihu,SUN Jinsheng,ZHANG Bo,et al.Sensitivity analysis of the temperature profile changing law in the production string of a high-pressure high-temperature gas well considering the coupling relation among the gas flow friction,gas properties,temperature,and pressure[J].Frontiers in Physics,2022,10:1050229.
[9] 刘劲歌.深水油气井开采期间环空压力预测及释放机理研究[D].北京:中国石油大学(北京),2019.
LIU Jinge.Research on the annular pressure buildup and mitigation mechanism in deepwater wells[D].Beijing:China University of Petroleum,2019.
[10] 匡立春,刘合,任义丽,等.人工智能在石油勘探开发领域的应用现状与发展趋势[J].石油勘探与开发,2021,48(1):1-11.
KUANG Lichun,LIU He,REN Yili,et al.Application and development trend of artificial intelligence in petroleum exploration and development[J].Petroleum Exploration & Development,2021,48(1):1-11.
[11] 薛亮,戴城,韩江峡,等.油藏渗流物理和数据联合驱动的深度神经网络模型[J].油气地质与采收率,2022,29(1):145-151.
XUE Liang,DAI Cheng,HAN Jiangxia,et al.Deep neural network model driven jointly by reservoir seepage physics and data[J].Petroleum Geology and Recovery Efficiency,2022,29(1):145-151.
[12] 毕剑飞,李靖,吴克柳,等.数据驱动与物理驱动融合的双驱动渗流代理模型构建[J].油气地质与采收率,2023,30(3):104-114.
BI Jianfei,LI Jing,WU Keliu,et al.A data-driven flow surrogate model based on a data-driven and physics-driven method[J].Petroleum Geology and Recovery Efficiency,2023,30(3):104-114.
[13] 孙金声,刘凡,程荣超,等.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.
SUN Jinsheng,LIU Fan,CHENG Rongchao,et al.Research progress and prospects of machine learning in lost circulation control[J].Acta Petrolei Sinica,2022,43(1):91-100.
[14] RAISSI M,PERDIKARIS P,KARNIADAKIS G E.Physics-informed neural networks:a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J].Journal of Computational Physics,2019,378:686-707.
[15] 吴国正,王发杰,程隋福,等.基于物理信息神经网络的内部声场正反问题数值计算[J].计算物理,2022,39(6):687-698.
WU Guozheng,WANG Fajie,CHENG Suifu,et al.Numerical simulation of forward and inverse problems of internal sound field based on physics-informed neural network[J].Chinese Journal of Computational Physics,2022,39(6):687-698.
[16] 汤卓超,傅卓佳.基于物理信息的神经网络求解曲面上对流扩散方程[J].计算力学学报,2023,40(2):216-222.
TANG Zhuochao,FU Zhuojia.Physics-informed neural networks for solving convection-diffusion equations on surfaces[J].Chinese Journal of Computational Mechanics,2023,40(2):216-222.
[17] 余波,许梦强,高强.基于物理信息神经网络的功能梯度材料稳态/瞬态热传导分析[J].计算力学学报,2023,40(4):594-601.
YU Bo,XU Mengqiang,GAO Qiang.Physics-informed neural networks for solving steady/transient heat conduction problems of functionally graded materials[J].Chinese Journal of Computational Mechanics,2023,40(4):594-601.
[18] 余波,甘子玉,张森林,等.基于物理信息神经网络预测2D/3D非稳态温度场及热源[J/OL].工程力学,(2023-10-17). https://kns.cnki.net/kcms2/article/abstract?v=DMKM_QUxZ7Dcz 17zI5OICm645rehKIoHRAuXyIUepNleyU_ISLTKdUXz1kDVb Yr2xR15ASx9BMULzQapKpsW7BEYxupyWU0dNw282f8NOlz H8S7vkke2FpGqmAYt2LG6EV5DGUReWWvr411z62Tw-Qo4s-LXlx-jAg7ZaRMiylWfKGMF6fy8-S_t2miAlt0n& uniplatform=NZKPT&language=CHS.
YU Bo,GAN Ziyu,ZHANG Senlin,et al.Prediction of 2D/3D unsteady-state temperature fields and heat sources upon the physics-informed neural networks[J/OL]. Engineering Mechanics, (2023-10-17). https://kns.cnki.net/kcms2/article/abstract? v=DMKM_QUxZ7Dcz17zI5OICm645rehKIoHRAuXy IUepNleyU_ISLTKdUXz1kDVbYr2xR15ASx9BMULzQapKps W7BEYxupyW U0dNw282f8NOlzH8S 7vkke2FpGqmA Yt2LG 6EV5DGUReWWvr 411z62Tw-Qo4s-LXlx-jAg 7ZaRMiylWfKGM F6fy8-S_t2miAlt0n& uniplatform=NZKPT&language=CHS.
[19] YUAN Taikang,ZHU Junxing,WANG Wuxin,et al.A space-time partial differential equation based physics-guided neural network for sea surface temperature prediction[J].Remote Sensing,2023,15(14):3498.
[20] ZHAO Xiaoyu,GONG Zhiqiang,ZHANG Yunyang,et al.Physics-informed convolutional neural networks for temperature field prediction of heat source layout without labeled data[J].Engineering Applications of Artificial Intelligence,2023,117:105516.
[21] NAZARI L F,CAMPONOGARA E,IMSLAND L S,et al.Neural networks informed by physics for modeling mass flow rate in a production wellbore[J].Engineering Applications of Artificial Intelligence,2024,128:107528.
[22] FRANKLIN T S,SOUZA L S,FONTES R M,et al.A Physics-Informed Neural Networks (PINN)oriented approach to flow metering in oil wells:an ESP lifted oil well system as a case study[J].Digital Chemical Engineering,2022,5:100056.
[23] 徐宝昌,张学智,王雅欣,等.用于两相流环空压力预测的自适应物理信息神经网络模型[J].石油学报,2023,44(3):545-555.
XU Baochang,ZHANG Xuezhi,WANG Yaxin,et al.Self-adaptive physical information neural network model for prediction of two-phase flow annulus pressure[J].Acta Petrolei Sinica,2023,44(3):545-555.
[24] 张学智.钻井过程的动态模拟及解算新方法[D].北京:中国石油大学(北京),2022.
ZHANG Xuezhi.Dynamic simulation and new solution method of drilling process[D].Beijing:China University of Petroleum,2022.
[25] ALMAJID M M,ABU-ALSAUD M O.Prediction of fluid flow in porous media using physics informed neural networks[C]//Abu Dhabi International Petroleum Exhibition & Conference.Abu Dhabi:ADIPEC,2020.
[26] HAN Jiangxia,XUE Liang,WEI Yunsheng,et al.Physics-informed neural network-based petroleum reservoir simulation with sparse data using domain decomposition[J].Petroleum Science,2023,20(6):3450-3460.
[27] 刘合,李艳春,贾德利,等.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.
LIU He,LI Yanchun,JIA Deli,et al.Application status and prospects of artificial intelligence in the refinement of waterflooding development program[J].Acta Petrolei Sinica,2023,44(9):1574-1586.
[28] 田十方,李彪.基于梯度优化物理信息神经网络求解复杂非线性问题[J].物理学报,2023,72(10):9-19.
TIAN Shifang,LI Biao.Solving complex nonlinear problems based on gradient-optimized physics-informed neural networks[J].Acta Physica Sinica,2023,72(10):9-19.
[29] 韦昌,樊昱晨,周永清,等.基于龙格库塔法的多输出物理信息神经网络模型[J].力学学报,2023,55(10):2405-2416.
WEI Chang,FAN Yuchen,ZHOU Yongqing,et al.Multi-output physics-informed neural networks model based on the Runge-Kutta method[J].Chinese Journal of Theoretical and Applied Mechanics,2023,55(10):2405-2416.
[30] JAGTAP A D,KHARAZMI E,EM KARNIADAKIS G.Conservative physics-informed neural networks on discrete domains for conservation laws:applications to forward and inverse problems[J].Computer Methods in Applied Mechanics and Engineering,2020,365:113028.
[31] JAGTAP A D,EM KARNIADAKIS G.Extended physics-informed neural networks (XPINNs):a generalized space-time domain decomposition based deep learning framework for nonlinear partial differential equations[J].Communications in Computational Physics,2020,28(5):2001-2041.
[32] 李道伦,沈路航,查文舒,等.基于神经算子与类物理信息神经网络智能求解新进展[J].力学学报,2024,56(4):875-889.
LI Daolun,SHEN Luhang,ZHA Wenshu,et al.New progress in intelligent solution of neural operators and physics-informed-based methods[J].Chinese Journal of Theoretical and Applied Mechanics,2024,56(4):875-889.
[33] 刘涛,何淼,张亚,等.小井眼超深井井筒温度预测模型及降温方法研究[J].钻采工艺,2024,47(3):65-72.
LIU Tao,HE Miao,ZHANG Ya,et al.Research on wellbore temperature prediction model and cooling method of ultra-deep well with small hole[J].Drilling & Production Technology,2024,47(3):65-72.
[34] 李梦博,柳贡慧,李军,等.基于移动边界的钻井井筒动态温度场研究[J].钻采工艺,2015,38(6):11-14.
LI Mengbo,LIU Gonghui,LI Jun,et al.Research on dynamic temperature field model of drilling with a moving interface[J].Drilling & Production Technology,2015,38(6):11-14.
[35] WIGHT C L,ZHAO Jia.Solving Allen-Cahn and Cahn-Hilliard equations using the adaptive physics informed neural networks[J].Communications in Computational Physics,2021,29(3):930-954.
[36] WANG Sifan,YU Xinling,PERDIKARIS P.When and why PINNs fail to train:a neural tangent kernel perspective[J].Journal of Computational Physics,2022,449:110768.
[37] LIU Dehao,WANG Yan.A Dual-Dimer method for training physics-constrained neural networks with minimax architecture[J].Neural Networks,2021,136:112-125.
[38] LI Wensheng,ZAHNG Chao,WANG Chuncheng,et al.Revisiting PINNs:generative adversarial physics-informed neural networks and point-weighting method[J].arXiv preprint arXiv,2022,2205:08754.
[39] MCCLENNY L,BRAGA-NETO U.Self-adaptive physics-informed neural networks using a soft attention mechanism[J].Journal of Computational Physics,2023,474:111722.
[40] WANG Jiangshuai,LI Jun,LIU Gonghui,et al.Development of a wellbore heat transfer model considering circulation loss[J].Arabian Journal of Geosciences,2020,13(2):85.
[41] 王志远.含天然气水合物相变的环空多相流流型转化机制研究[D].青岛:中国石油大学(华东),2009.
WANG Zhiyuan.Study on annular multiphase flow pattern transition mechanism considering gas hydrate phase transition[D].Qingdao:China University of Petroleum(East China),2009.

物理信息神经网络驱动的井筒温度场求解方法

A method for solving wellbore temperature field driven by physical information neural network

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]	孙金声, 杨景斌, 吕开河, 白英睿, 刘敬平, 黄贤斌. 致密油气钻井液技术研究现状与展望[J]. 石油学报, 2025, 46(1): 279-288.
[2]	石祥超, 陈帅. 岩石可钻性分级标准的改进建议[J]. 石油学报, 2024, 45(9): 1432-1442.
[3]	徐宝昌, 尤香凝, 孟卓然, 刘伟, 尹士轩. 基于回压激励响应的实时地层压力反演方法[J]. 石油学报, 2024, 45(6): 1009-1018.
[4]	王海涛, 郑鹏, 邓青山, 孙立伟, 鲁港, 孟庆安, 覃吉. 三段制圆弧轨道设计参数的谱集计算[J]. 石油学报, 2024, 45(4): 708-717.
[5]	谢彬强, 陶怀志, 张俊, 陈晋东, 吕军贤, 赵林. 梳型温敏聚合物对无固相水基钻井液高温流变性的调控[J]. 石油学报, 2024, 45(2): 427-436.
[6]	杨宏伟, 李军, 张辉, 张更, 邢星. 考虑相间传质的全瞬态井筒气液两相流模型[J]. 石油学报, 2024, 45(11): 1680-1689.
[7]	祝效华, 刘伟吉, 石昌帅, 周伟. 重载冲击破岩提速机理实验[J]. 石油学报, 2024, 45(10): 1529-1537.
[8]	鲁港, 王海涛, 李杉, 李雪松, 杨志国, 王建华, 邱晨. 三维七段制圆弧型井眼轨道设计的拟解析解[J]. 石油学报, 2023, 44(9): 1545-1551.
[9]	石祥超, 陈帅, 孟英峰, 陈军海, 李皋, 杨昕昊, 焦烨. 岩石可钻性测定方法的改进和优化建议[J]. 石油学报, 2023, 44(9): 1562-1573.
[10]	熊亮, 赵勇, 魏力民, 庞河清, 慈建发. 威荣海相页岩气田页岩气富集机理及勘探开发关键技术[J]. 石油学报, 2023, 44(8): 1365-1381.
[11]	高永海, 尹法领, 张党生, 孙小辉, 赵欣欣, 陈野, 孙宝江. 水合物储层水平井钻井井筒-储层耦合模型与井壁稳定性分析[J]. 石油学报, 2023, 44(7): 1151-1166.
[12]	任妍君, 翟玉芬, 路岩岩. 抗高温高密度可逆油基钻井液体系[J]. 石油学报, 2023, 44(5): 841-851,861.
[13]	刘伟吉, 张有建, 罗云旭, 祝效华. 岩石内部高压电脉冲等离子体通道生成机理[J]. 石油学报, 2023, 44(4): 684-697.
[14]	徐宝昌, 张学智, 王雅欣, 刘伟, 孟卓然. 用于两相流环空压力预测的自适应物理信息神经网络模型[J]. 石油学报, 2023, 44(3): 545-555.
[15]	杨进, 李磊, 宋宇, 仝刚, 张明贺, 张慧. 中国海洋油气钻井技术发展现状及展望[J]. 石油学报, 2023, 44(12): 2308-2318.