H油田侏罗系延9储层物性参数预测方法研究

doi:10.3969/j.issn.1672-9803.2023.03.013

摘要/Abstract

摘要： 孔隙度和渗透率是描述储层物性特征的重要参数。常规的孔隙度和渗透率计算方法基于线性拟合,模型较单一且准确度不足。因此,基于文献调研和鄂尔多斯盆地H油田延9储层概况,总结了一些常用的孔隙度和渗透率计算方法,进而提出一种梯度提升决策树预测储层物性的方法,该方法将得到的物性参数和多条相关测井曲线作为梯度提升决策树的多元输入信息,通过训练得到预测孔隙度和渗透率的模型。H油田的21口井中6口井有实验数据,利用其中4口井与用常规的统计回归法和孔渗拟合法得出的15口井的物性数据进行训练,另外2口井作为测试集,结果表明梯度提升决策树储层预测方法的孔隙度、渗透率结果与岩心分析结果相关性均为0.9,常规方法计算结果的相关性也较高,分别为0.88、0.87。为提高模型解释精度,将常规方法计算的物性和训练集的物性及其测井曲线作为梯度提升决策树的多元输入信息进行训练,优化后梯度提升树结果与岩心分析结果相关性达到0.93。

关键词: 梯度提升决策树, 孔隙度, 渗透率, 储层物性

Abstract: Porosity and permeability are important parameters to describe reservoir physical properties. Conventional calculation methods of porosity and permeability are based on linear fitting, and the model is single and the accuracy is insufficient. Therefore, based on the literature investigation and the general situation of Yan 9 reservoir in Ordos Basin, this paper summarizes some commonly used calculation methods of porosity and permeability, and then puts forward a method of predicting reservoir physical properties by gradient boosting decision tree. In this method, the physical properties and many related logging curves are used as multivariate input information of gradient boosting decision tree, and the models for predicting porosity and permeability are obtained through training. There are 21 wells in H Oilfield, among which 6 wells have experimental data, 4 wells are trained with the physical property data of 15 wells obtained by conventional statistical regression method and porosity-permeability fitting method and the other 2 wells are taken as test sets. The results show that the correlation coeffieient between porosity and permeability results of gradient boosting decision tree reservoir prediction method and core analysis results is 0.9, and the correlation coefficients calculated by conventional methods are also high, which are 0.88 and 0.87 respectively. In order to improve the interpretation accuracy of the model, the physical properties calculated by conventional methods, the physical properties of training sets and their logging curves are trained as multivariate input information of gradient boosting decision tree, and the correlation coefficient between the optimized gradient boosting decision tree results and core analysis results reaches 0.93.

Key words: gradient boosting tree decision, porosity, permeability, reservoir physical property

中图分类号:

TE132.1

白倩, 赵军龙, 黄千玲, 许鉴源. H油田侏罗系延9储层物性参数预测方法研究[J]. 录井工程, 2023, 34(3): 81-88.

BAI Qian, ZHAO Junlong, HUANG Qianling, XU Jianyuan. Study on prediction method of physical properties of Jurassic Yan 9 reservoir in H Oilfield[J]. Mud Logging Engineering, 2023, 34(3): 81-88.

参考文献

[1] 王亚兰,杨凯程,王河,等.苏南区块中子-密度交会计算孔隙度方法应用[J].石油化工应用,2022,41(9):72-77.
WANG Yalan, YANG Kaicheng, WANG He, et al.Application of neutron-density intersection method to calculate porosity in southern Sulige block[J]. Petrochemical Industry Application,2022,41(9):72-77.
[2] 田军,刘永雷,徐博,等.深埋储层孔隙度迭代反演方法[J].石油地球物理勘探,2022,57(3):666-675.
TIAN Jun, LIU Yonglei, XU Bo, et al.Iterative inversion method of porosity in deep buried reservoir[J]. Oil Geophysical Prospecting,2022,57(3):666-675.
[3] 张宇航,时保宏,张曰静,等.机器学习方法在浅层滩坝相薄储层孔隙度预测中的应用:以准噶尔盆地车排子地区白垩系为例[J/OL].沉积学报:1-12[2023-03-21].http://kns.cnki.net/kcms/detail/62.1038.P.20221026.1812.002.html.
ZHANG Yuhang, SHI Baohong, ZHANG Yuejing, et al. Application of machine learning method in porosity prediction of shallow beach-dam facies thin reservoir:A case study of Cretaceous in Chepaizi area of Junggar Basin[J/OL]. Acta Sedimentologica Sinica:1-12[2023-03-21].http://kns.cnki.net/kcms/detail/62.1038.P.20221026.1812.002.html.
[4] 魏国华,韩宏伟,刘浩杰,等.基于半监督高斯混合模型与梯度提升树的砂岩储层相控孔隙度预测[J].石油地球物理勘探,2023,58(1):46-55.
WEI Guohua, HAN Hongwei, LIU Haojie, et al.Prediction of facies-controlled porosity of sandstone reservoir based on semi-supervised Gaussian mixture model and gradient boosting tree[J]. Oil Geophysical Prospecting,2023,58(1):46-55.
[5] 张荣虎,姚根顺,寿建峰,等.沉积、成岩、构造一体化孔隙度预测模型[J].石油勘探与开发,2011,38(2):145-151.
ZHANG Ronghu, YAO Genshun, SHOU Jianfeng, et al.Integrated porosity prediction model of sedimentation, diagenesis and structure[J]. Petroleum Exploration and Development,2011,38(2):145-151.
[6] 杨丽,覃利娟,彭志春,等.乌石东区成岩演化数值模拟与孔隙度预测[J].地质科技通报,2022,41(3):150-157.
YANG Li, QIN Lijuan, PENG Zhichun, et al.Numerical simulation of diagenetic evolution and porosity prediction in eastern Wushi area[J]. Bulletin of Geological Science and Technology,2022,41(3):150-157.
[7] 韩宏伟,刘浩杰,桑文镜,等.基于半监督学习的井震联合储层横向孔隙度预测方法[J].地球物理学报,2022,65(10):4073-4086.
HAN Hongwei, LIU Haojie, SANG Wenjing, et al.Prediction method of reservoir lateral porosity based on semi-supervised learning combined seismic with well logs[J]. Chinese Journal of Geophysics,2022,65(10):4073-4086.
[8] 张言辉. 基于物性预测相对渗透率的改进神经网络方法[J].天然气与石油,2020,38(3):44-49.
ZHANG Yanhui.Improved neural network method for predicting relative permeability based on physical properties[J]. Natural Gas and Oil,2020,38(3):44-49.
[9] 曹茜,王志章,樊太亮,等.基于FSVM法分岩性解释砂砾岩储层渗透率[J].新疆大学学报(自然科学版),2018,35(3):264-271.
CAO Qian, WANG Zhizhang, FAN Tailiang, et al.Permeability interpretation of sandy conglomerate reservoir by lithology based on FSVM method[J]. Journal of Xinjiang University(Natural Science Edition),2018,35(3):264-271.
[10] 李雄炎,秦瑞宝,曹景记,等.复杂储层连通孔隙度评价与渗透率定量计算方法[J].石油地球物理勘探,2022,57(2):377-385.
LI Xiongyan, QIN Ruibao, CAO Jingji, et al.Evaluation of connected porosity and quantitative calculation method of permeability in complex reservoirs[J]. Oil Geophysical Prospecting,2022,57(2):377-385.
[11] 黄雨阳,冯进,宋伟,等.结合NMR横向弛豫时间谱与压汞资料的砂岩储层改进渗透率智能预测方法[J].物探化探计算技术,2020,42(3):338-344.
HUANG Yuyang, FENG Jin, SONG Wei, et al.Improved intelligent permeability prediction method for sandstone reservoirs based on NMR transverse relaxation time spectrum and mercury injection data[J]. Computing Techniques for Geophysical and Geochemical Exploration,2020,42(3):338-344.
[12] 杨坤,王付勇,曾繁超,等.基于数字岩心分形特征的渗透率预测方法[J].吉林大学学报(地球科学版),2020,50(4):1003-1011.
YANG Kun, WANG Fuyong, ZENG Fanchao, et al.Permeability prediction method based on fractal characteristics of digital core[J]. Journal of Jilin University(Earth Science Edition),2020,50(4):1003-1011.
[13] 王猛,董宇,蔡军,等.基于BP神经网络的储层渗透率预测及质量评价方法[J].地球物理学进展,2023,38(1):321-327.
WANG Meng, DONG Yu, CAI Jun, et al.Reservoir permeability prediction and quality evaluation method based on BP neural network[J]. Progress in Geophysics,2023,38(1):321-327.
[14] 谷宇峰,张道勇,鲍志东,等.利用梯度提升决策树(GBDT)预测渗透率:以姬塬油田西部长4+5段致密砂岩储层为例[J].地球物理学进展,2021,36(2):585-594.
GU Yufeng, ZHANG Daoyong, BAO Zhidong, et al.Permeability prediction using Gradient Boosting Decision Tree (GBDT): A case study of tight sandstone reservoir in Chang 4+5 Member in western Jiyuan Oilfield[J]. Progress in Geophysics,2021,36(2):585-594.
[15] 谷宇峰,张道勇,鲍志东.测井资料PSO-XGBoost渗透率预测[J].石油地球物理勘探,2021,56(1):26-37.
GU Yufeng, ZHANG Daoyong, BAO Zhidong, et al.Permeability prediction of logging data by PSO-XGBoost[J]. Oil Geophysical Prospecting,2021,56(1):26-37.
[16] 王谦,谭茂金,石玉江,等.径向基函数神经网络法致密砂岩储层相对渗透率预测与含水率计算[J].石油地球物理勘探,2020,55(4):864-872.
WANG Qian, TAN Maojin, SHI Yujiang, et al.Relative permeability prediction and water cut calculation of tight sandstone reservoir by radial basis function neural network method[J]. Oil Geophysical Prospecting,2020,55(4):864-872.
[17] 高强勇,王昕,高建英,等.致密岩心核磁共振孔隙度影响因素分析[J].测井技术,2021,45(4):424-430.
GAO Qiangyong, WANG Xin, GAO Jianying, et al.Analysis of influencing factors of NMR porosity of tight core[J]. Well Logging Technology,2021,45(4):424-430.
[18] 周新波,段迎利,袁伟,等.M油田渗透率计算方法研究[J].科技创新与应用,2014,108(32):76.
ZHOU Xinbo, DUAN Yingli, YUAN Wei, et al.Study on permeability calculation method of M Oilfield[J]. Technology Innovation and Application,2014,108(32):76.
[19] 路萍,王浩辰,高春云,等.致密砂岩储层渗透率预测技术研究进展[J].地球物理学进展,2022,37(6):2428-2438.
LU Ping, WANG Haochen, GAO Chunyun, et al.Research progress of permeability prediction technology for tight sandstone reservoirs[J]. Progress in Geophysics,2022,37(6):2428-2438.
[20] 汤潇,张超谟,周文银,等.Purcell模型与Thomeer模型相结合的高精度渗透率模型研究[J].当代化工,2019,48(12):2934-2938.
TANG Xiao, ZHANG Chaomo, ZHOU Wenyin, et al.Study on high precision permeability model based on Purcell model and Thomeer model[J]. Contemporary Chemical Industry,2019,48(12):2934-2938.
[21] 孙颖. 核磁共振在页岩储层参数评价中的应用综述[J].地球物理学进展,2023,38(1):254-270.
SUN Ying.Application of nuclear magnetic resonance in shale reservoir parameter evaluation[J]. Progress in Geophysics,2023,38(1):254-270.
[22] 胡婷婷,贾春明,余海涛,等.核磁共振测井法估算砂砾岩储层束缚水饱和度模型构建[J].核电子学与探测技术,2020,40(6):897-901.
HU Tingting, JIA Chunming, YU Haitao, et al.Construction of model for estimating irreducible water saturation of glutenite reservoirs by nuclear magnetic resonance logging[J]. Nuclear Electronics & Detection Technology,2020,40(6):897-901.
[23] 何玉春. 基于机器学习的储层渗透率预测方法[J].石化技术,2022,29(12):182-184.
HE Yuchun.Prediction method of reservoir permeability based on machine learning[J]. Petrochemical Industry Technology,2022,29(12):182-184.
[24] 沈夏炯,张俊涛,韩道军.基于梯度提升回归树的短时交通流预测模型[J].计算机科学,2018,45(6):222-227,264.
SHEN Xiajiong, ZHANG Juntao, HAN Daojun.Short-term traffic flow prediction model based on gradient boosting regression tree[J]. Computer Science,2018,45(6):222-227,264.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed