|本期目录/Table of Contents|

[1]李彦尊 白玉湖 陈桂华 徐兵祥 陈 岭 董志强.基于人工神经网络方法的页岩油气产量预测新技术——以美国Eagle Ford页岩油气田为例[J].中国海上油气,2020,32(04):104-110.[doi:10.11935/j.issn.1673-1506.2020.04.012]
 LI Yanzun BAI Yuhu CHEN Guihua XU Bingxiang CHEN Ling DONG Zhiqiang.ANN method based on novel technology for production prediction of shale oil and gas:A case study in Eagle Ford[J].China Offshore Oil and Gas,2020,32(04):104-110.[doi:10.11935/j.issn.1673-1506.2020.04.012]
点击复制

基于人工神经网络方法的页岩油气产量预测新技术——以美国Eagle Ford页岩油气田为例()

《中国海上油气》[ISSN:1673-1506/CN:11-5339/TE]

卷:
第32卷
期数:
2020年04期
页码:
104-110
栏目:
油气田开发
出版日期:
2020-07-25

文章信息/Info

Title:
ANN method based on novel technology for production prediction of shale oil and gas:A case study in Eagle Ford
文章编号:
1673-1506(2020)04-0104-07
作者:
李彦尊 白玉湖 陈桂华 徐兵祥 陈 岭 董志强
(中海油研究总院有限责任公司 北京 100028)
Author(s):
LI Yanzun BAI Yuhu CHEN Guihua XU Bingxiang CHEN Ling DONG Zhiqiang
(CNOOC Research Institute Co., Ltd., Beijing 100028, China)
关键词:
页岩油气 人工神经网络 产量预测 产量递减
Keywords:
shale oil and gas artificial neural network production prediction production decline
分类号:
TE32+8
DOI:
10.11935/j.issn.1673-1506.2020.04.012
文献标志码:
A
摘要:
页岩油气产量受地质、工程等多重因素影响,常规产量预测方法难以反映其真实生产特征,因此引入了机器学习方法进行页岩油气产量预测。以美国Eagle Ford页岩某区块400余口生产井地质、油藏、工程数据为学习样本,对人工神经网络模型进行训练和优化,确定了最佳模型参数; 结合交叉验证等手段改进了训练方法,提高计算效率和预测精度,得到了初始产量、递减率、递减指数等产量递减参数与地质、油藏、工程参数之间的关系模型,进而形成了基于静态参数的页岩油气单井产量预测技术。实例应用表明,投产5年内,本文模型产量预测精度可达90%。在没有生产数据或生产数据较少情况下,本文模型预测产量具有突出优势
Abstract:
Offshore oilfields are mostly developed by horizontal wells with large well spacing, and there are obvious uncertainties in the quantitative scale and spatial distribution of different configuration units. In order to make full use of the abundant lateral information in horizontal wells, the spatial distribution characteristics of configuration units of sandy braided river reservoir are studied based on horizontal well information, by using core, logging, seismic and production performance data. Based on the division and analysis of configuration units, the logging facies of horizontal wells are calibrated by using imaging edge detection data. The logging facies model of horizontal wells with different configuration units in sandy braided river can be established by considering the factors such as structure, horizontal well trajectory and surrounding rock influence, the spatial combination relationship can be clarified, and the quantitative scale empirical formula of the configuration units can be validated by horizontal well information according to the statistical analysis of modern sedimentation and field outcrop. Guided by depositional model, taking geological knowledge base of configuration unit as scale constraint, the distribution of sand body is predicted by seismic impedance inversion constrained by horizontal well logging. The lateral delimitation of configuration unit is completed according to the idea of “well-seismic combination, thickness control and scale constraint”, the spatial distribution of internal configuration units of sandy braided river reservoir is clarified, and the reliability of prediction result is verified by reservoir numerical simulation. The research results show that the abundant lateral information of horizontal wells can greatly reduce the uncertainty of inter-well reservoir prediction and spatial distribution research of configuration units

参考文献/References:

[1] BOWKER K A. Barnett shale gas production,fort worth basin:Issues and discussion[J].AAPG Bulletin,2007,91(4):523-533.
[2] NEARING T R,STARTZMAN R A.Shale well productivity[R].SPE 18553-MS,1988.
[3] 王怒涛.实用气藏动态分析方法[M].北京:石油工业出版社,2011.
[4] 白玉湖,徐兵祥,陈岭,等.页岩油气典型曲线及解析模型产量预测新方法[J].中国海上油气,2018,30(4):120-126.DOI:10.11935 /j.issn.1673-1506.2018.04.014.
BAI Yuhu,XU Bingxiang,CHEN Ling,et al.New production prediction methods for typical curve and analytical model of shale oil and gas[J].China Offshore Oil and Gas,2018,30(4):120-126.DOI:10.11935 /j.issn.1673-1506.2018.04.014.
[5] MARSDEN J,KOSTYLEVA I,FASSIHI M R,et al.A conceptual shale gas model validated by pressure and rate data from the Haynesville shale[R].SPE 187456-MS,2017.
[6] 白玉湖,陈桂华,徐兵祥,等.页岩气产量递减典型曲线模型及对比研究[J].中国石油勘探,2016,21(5):96-102.
BAI Yuhu,CHEN Guihua,XU Bingxiang,et al.Comparison of typical curve models for shale gas production decline prediction[J].China Petroleum Exploration,2016,21(5):96-102.
[7] 郭长杰,王浩翔,刘晓,等.浅析机器学习技术在油气行业的应用场景[J].信息系统工程,2017(5):100-103.
GUO Changjie,WANG Haoxiang,LIU Xiao,et al.Application scenario of machine learning technology in oil and gas industry[J].China CIO News,2017(5):100-103.
[8] 马林茂,李德富,郭海湘,等.基于遗传算法优化BP神经网络在原油产量预测中的应用:以大庆油田BED试验区为例[J].数学的实践与认识,2015,45(24):117-128.
MA Linmao,LI Defu,GUO Haixiang,et al.BP neural network based on genetic algorithm applied in crude oil production forecast:Taking the BED test area of the Daqing oilfield as an example[J].Mathematics in Practice and Theory,2015,45(24):117-128.
[9] 檀朝东,BANGERT P,刘柏良,等.人工神经网络自学习方法在大港滩海油田的应用[J].中国石油和化工,2010(11):46-47.
TAN Chaodong,BANGERT P,LIU Boliang,et al.Application of artificial neural network self-learning method in Dagang oilfield[J].China Petroleum and Chemical Industry,2010(11):46-47.
[10] AHRIMANKOSH M,KASIRI N,MOUSAVI S M.Improved permeability prediction of a heterogeneous carbonate reservoir using artificial neural networks based on the flow zone index approach[J].Liquid Fuels Technology,2011,29(23):2494-2506.
[11] KHANAL A,KHOSHGHADAM M,LEE W J,et al.New forecasting method for liquid rich shale gas condensate reservoirs with data driven approach using principal component analysis[J].Journal of Natural Gas Science and Engineering,2017,38:621-637.
[12] CAO Q,BANERJEE R,GUPTA S,et al.Data driven production forecasting using machine learning[R].SPE 180984-MS,2016.
[13] 陈桂华,白玉湖,陈晓智,等.页岩油气纵向综合甜点识别新方法及定量化评价[J].石油学报,2016,37(11):1337-1342,1360.
CHEN Guihua,BAI Yuhu,CHEN Xiaozhi,et al.A new identification method for the longitudinal integrated shale oil/gas sweet spot and its quantitative evaluation[J].Acta Petrolei Sinica,2016,37(11):1337-1342,1360.
[14] 白玉湖,杨皓,陈桂华,等.压裂参数对页岩气井产量递减典型曲线影响分析[J].天然气与石油,2014,32(4):34-38.
BAI Yuhu,YANG Hao,CHEN Guihua,et al.Influence of fracturing parameters on typical shale gas well production decline curve[J].Natural Gas and Oil,2014,32(4):34-38.
[15] 李庆辉,陈勉,WANG F P,等.工程因素对页岩气产量的影响:以北美Haynesville页岩气藏为例[J].天然气工业,2012,32(4):54-59,123.
LI Qinghui,CHEN Mian,WANG F P,et al.Influences of engineering factors on shale gas productivity:A case study from the haynesville shale gas reservoir in north america[J].Natural Gas Industry,2012,32(4):54-59,123.
[16] 祝彦贺.北美某盆地Z区块页岩油气产量的影响因素[J].海洋地质前沿,2013,29(8):33-38,52.
ZHU Yanhe.Study of geologic parameters controlling production difference in Z block of a basin in north america[J].Marine Geology Frontiers,2013,29(8):33-38,52.
[17] 钱旭瑞,刘广忠,唐佳,等.页岩气井产能影响因素分析[J].特种油气藏,2012,19(3):81-83.
QIAN Xurui,LIU Guangzhong,TANG Jia,et al.Analysis of influencing factors on shale gas well productivity[J].Special Oil and Gas Reservoir,2012,19(3):81-83.
[18] 殷震.基于BP神经网络的电力变压器内部故障诊断方法研究[D].天津:天津大学,2007.
[19] 蒙西,乔俊飞,李文静.基于快速密度聚类的RBF 神经网络设计[J].智能系统学报,2018,13(3):331-338.
MENG Xi,QIAO Junfei,LI Wenjing.Construction of RBF neural networks via fast density clustering[J].CAAI Transactions on Intelligent Systems,2018,13(3):331-338.
[20] 张鲜,AWOTUNDE A A.油藏模拟历史拟合中Levenberg-Marquardt算法的改进[J].石油勘探与开发,2016,43(5):806-815.
ZHANG Xian,AWOTUNDE A A.Improvement of Levenberg-Marquardt algorithm during history fitting for reservoir simulation[J].Petroleum Exploration and Development,2016,43(5):806-815.
[21] 边耐政,李硕,陈楚才.加权交叉验证神经网络在水质预测中的应用[J].计算机工程与应用,2015,51(21):255-258.
BIAN Naizheng,LI Shuo,CHEN Chucai.Prediction of water quality using weighted cross-validation artificial network[J].Computer Engineering and Applications,2015,51(21):255-258.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期:2020-01-06 改回日期:2020-03-15*国家重点研发计划项目“煤层气、页岩气及现代煤化工关键技术标准研究(编号:2018YFF0213802)”部分研究成果。 第一作者简介: 李彦尊,男,工程师,2015年获中国石油大学(北京)油气田开发专业博士学位,主要从事页岩油气、致密气等非常规油气开发方面研究。地址:北京市朝阳区太阳宫南街6号院中海油大厦B 座304室(邮编:100028)。E-mail:liyz9@ cnooc.com.cn。
更新日期/Last Update: 2020-07-20