基于XGBoost算法的低渗透致密气井动静一体化分类模型

商永涛; 寨硕; 林新宇; 李相亮; 李辉; 冯青

doi:10.3969/j.issn.1006-6535.2023.05.018

特种油气藏 >

2023 , Vol. 30 >Issue 5: 135 - 143

DOI: https://doi.org/10.3969/j.issn.1006-6535.2023.05.018

油藏工程

基于XGBoost算法的低渗透致密气井动静一体化分类模型

商永涛 ,
寨硕 ,
林新宇 ,
李相亮 ,
李辉 ,
冯青

展开

1.中海油田服务股份有限公司,天津 300459;
2.成都理工大学,四川成都 610059

商永涛(1981—),男,高级工程师,2008年毕业于中国石油大学(华东)无线电物理专业,2011年毕业于该校无线电物理专业,获硕士学位,现主要从事油气田开发开采技术应用方面的研究工作。

收稿日期: 2022-08-14

修回日期: 2023-06-24

网络出版日期: 2023-12-25

基金资助

国家自然科学基金“页岩储层纳米孔隙结构表征及渗流机理研究”(51304032)

收起

Dynamic and Static Integrated Classification Model for Low Permeability Tight Gas Wells Based on XGBoost Algorithm

Shang Yongtao ,
Zhai Shuo ,
Lin Xinyu ,
Li Xiangliang ,
Li Hui ,
Feng Qing

Expand

1. China Oilfield Services Limited, Tianjin 300459, China;
2. Chengdu University of Technology, Chengdu, Sichuan 610059, China

Received date: 2022-08-14

Revised date: 2023-06-24

Online published: 2023-12-25

Fold

摘要

子米气田为典型的低渗透致密气田,不同气井储层物性及生产特征差异大,开发策略亟需改善。针对该问题,提出了一种基于XGBoost算法的致密气井分类方法。通过计算特征参数的皮尔逊相关系数,判断4个动态、5个静态特征参数与气井产能的相关程度,以此确定模型的输入特征。然后基于XGBoost算法,通过参数优化,完成模型训练并对子米气田进行气井分类。研究表明:影响气井分类的主要因素为生产配产、原始地层压力、有效厚度、孔隙度和无阻流量;子米气田1、2类气井产能主要受到储层厚度和基质渗透率的影响,3类气井产能主要影响因素为有效厚度和含气饱和度;与专家划分结果相比,模型分类结果准确率为92.3%。该研究提高了气井分类实效性,降低了人为主观性,分类结果切合矿场实际,对气井分类管理和开发策略的制订有一定的指导意义。

关键词： 机器学习; 气井分类; XGBoost算法; 子米气田

本文引用格式

商永涛 , 寨硕 , 林新宇 , 李相亮 , 李辉 , 冯青 . 基于XGBoost算法的低渗透致密气井动静一体化分类模型[J]. 特种油气藏, 2023 , 30(5) : 135 -143 . DOI: 10.3969/j.issn.1006-6535.2023.05.018

Abstract

The Zimi Gas Field is a typical low-permeability tight gas field with great differences in reservoir physical properties and production characteristics among different gas well reservoirs, so the development strategy is in urgent need of improvement. To address this problem, a classification method for tight gas wells based on XGBoost algorithm is proposed. The input features of the model are determined by calculating Pearson correlation coefficients of the feature parameters and judging the correlation degree of 4 dynamic and 5 static feature parameters with gas well productivity. Then, based on the XGBoost algorithm, the model training is completed and gas wells are classified in Zimi Gas Field through parameter optimization. The study shows that the main factors affecting gas well classification are production allocation, original formation pressure, effective thickness, porosity and open flow capacity; the gas well productivity of Category 1 and 2 in Zimi Gas Field is mainly affected by reservoir thickness and matrix permeability, while the main factors affecting the gas well productivity of Category 3 are effective thickness and gas saturation; the accuracy of the model classification results is 92.3% compared with the expert classification results. This study improves the effectiveness of gas well classification, reduces human subjectivity, and the classification results are in line with the actual mine field, which has a certain guidance for gas well classification management and development strategy formulation.

Key words： machine learning; gas well classification; XGBoost algorithm; Zimi Gas Field

参考文献

[1] 贾爱林,位云生,郭智,等.中国致密砂岩气开发现状与前景展望[J].天然气工业,2022,42(1):83-92.
JIA Ailin,WEI Yunsheng,GUO Zhi,et al.Development status and prospect of tight sandstone gas in China[J].Natural Gas Industry,2022,42(1):83-92.
[2] 赵群,杨慎,钱伟,等.中国非常规天然气开发现状及前景思考[J].环境影响评价,2020,42(5):34-37.
ZHAO Qun,YANG Shen,QIAN Wei,et al.Current situation and prospect of unconventional gas development in China[J].Environmental Impact Assessment,2020,42(5):34-37.
[3] 汪新光,郇金来,彭小东,等.基于数字岩心的致密砂岩储层孔隙结构与渗流机理[J].油气地质与采收率,2022,29(6):22-30.
WANG Xinguang,HUAN Jinlai,PENG Xiaodong,et al.Flow mechanism and pore structures of tight sandstone based on digital core analysis[J].Petroleum Geology and Recovery Efficiency,2022,29(6):22-30.
[4] 郭一凡,司马立强,王亮,等.基于偏最小二乘回归方法的毛管压力曲线预测超致密砂岩储层渗透率[J].油气地质与采收率,2022,29(6):67-76.
GUO Yifan,SIMA Liqiang,WANG Liang,et al.Prediction of ultra-tight sandstone reservoir permeability by capillary pressure curve based on partial least squares regression method[J].Petroleum Geology and Recovery Efficiency,2022,29(6):67-76.
[5] 王继平,张城玮,李建阳,等.苏里格气田致密砂岩气藏开发认识与稳产建议[J].天然气工业,2021,41(2):100-110.
WANG Jiping,ZHANG Chengwei,LI Jianyang,et al.Tight sandstone gas reservoirs in the Sulige Gas Field:development understandings and stable-production proposals[J].Natural Gas Industry,2021,41(2):100-110.
[6] 夏飞.浅谈我国致密气开发技术现状及未来发展潜力[J].石化技术,2017,24(9):176.
XIA Fei.Current situation and future development potential of tight gas development technology in China[J].Petrochemical Industry Technology,2017,24(9):176.
[7] 贾焰然,石军太,李星浩,等.低渗致密气井分类评价方法研究——以长庆子洲气田为例[J].地质与勘探,2021,57(3):647-655.
JIA Yanran,SHI Juntai,LI Xinghao,et al.Classification and evaluation methods for low-permeability tight gas wells in the Zizhou Gas Field of Changqing[J].Geology & Exploration,2021,57(3):647-655.
[8] 刘雨林,范凌霄,房大志,等.源—储分类新方法在川东地区页岩气井产量分析中的应用[J].油气藏评价与开发,2022,12(3):429-436.
LIU Yulin,FAN Lingxiao,FANG Dazhi,et al.Application of a new source-reservoir classification method in production analysis of shale gas wells in eastern Sichuan[J].Reservoir Evaluation and Development,2022,12(3):429-436.
[9] 陈芳芳,姜越,刘金滚,等.地震频谱衰减特征在基岩气藏气井分类评价中的应用——以东坪基岩气藏为例[J].新疆石油地质,2019,40(5):600-604.
CHEN Fangfang,JIANG Yue,LIU Jingun,et al.Application of characteristics of seismic frequency spectrum attenuation in classification and evaluation of gas wells in basement gas reservoirs:a case of Dongping Basement Gas Reservoir[J].Xinjiang Petroleum Geology,2019,40(5):600-604.
[10] 姜宝彦.页岩气评价标准与储层分类探讨[J].西部探矿工程,2021,33(5):76-77.
JIANG Baoyan.Discussion on shale gas evaluation standard and reservoir classification[J].West-China Exploration Engineering,2021,33(5):76-77.
[11] 侯晨虹,石军太,刘成,等.M区块煤层气井分类评价方法[J].天然气工业,2018,38(增刊1):74-79.
HOU Chenhong,SHI Juntai,LIU Cheng,et al.Classification evaluation method of M Block coalbed methane well[J].Natural Gas Industry,2018,38(S1):74-79.
[12] 梁治国.苏里格气田苏10区块气井生产动态特征分析[J].录井工程,2021,32(3):136-140.
LIANG Zhiguo.Analysis on dynamic production characteristics of gas wells in Block Su10 of Sulige Gas Field[J].Mud Logging Engineering,2021,32(3):136-140.
[13] 袁继明,夏勇,李杰,等.利用灰色关联法对低压产水气井进行分类研究[J].石油化工应用,2018,37(5):73-75.
YUAN Jiming,XIA Yong,LI Jie,et al.Classification of low pressure water producing gas wells by grey relational analysis[J].Petrochemical Industry Application,2018,37(5):73-75.
[14] 刘进博,朱志勇,洪将领,等.基于生产资料特征分析的气井分类方法[J].石油钻采工艺,2021,43(4):510-517.
LIU Jinbo,ZHU Zhiyong,HONG Jiangling,et al.Gas well classification method based on production data characteristic analysis[J].Oil Drilling & Production Technology2021,43(4):510-517.
[15] 马先林,周德胜,蔡文斌,等.基于可解释机器学习的水平井产能预测方法[J].西南石油大学学报(自然科学版),2022,44(4):81-90.
MA Xianlin,ZHOU Desheng,CAI Wenbin,et al.An interpretable machine learning approach to prediction horizontal well productivity[J].Journal of Southwest Petroleum University(Science & Technology Edition),2022,44(4):81-90.
[16] 田建华,朱博华,卢志强,等.基于井控多属性机器学习的缝洞型储层预测方法[J].油气地质与采收率,2023,30(1):86-92.
TIAN Jianhua,ZHU Bohua,LU Zhiqiang,et al.Fracture-cavity reservoir prediction based on well-controlled multi-attribute machine learning[J].Petroleum Geology and Recovery Efficiency,2023,30(1):86-92.
[17] LI Yingchang,LI Chao,LI Mingyang,et al.Influence of variable selection and forest type on forest aboveground biomass estimation using machine learning algorithms[J].Forests,2019,10(12):1-24.
[18] ZHONG Liheng,HU Lina,ZHOU Hang.Deep learning based multi-temporal crop classification[J].Remote Sensing of Environment,2019,221(2):430-443.
[19] 蔡林菲,吴达胜,方陆明,等.基于XGBoost的高分二号影像树种识别[J].林业资源管理,2019,19(5):44-51.
CAI Linfei,WU Dasheng,FANG Luming,et al.Tree species identification using XGBoost based on GF-2 images[J].Forest Resources Management,2019,19(5):44-51.
[20] FREEMAN E A,MOISEN G G,COULSTON J W,et al.Random forests and stochastic gradient boosting for predicting tree canopy cover:comparing tuning processes and model performance[J].Canadian Journal of Forest Research,2015,46(3):323-339.
[21] MAXWELL A E,WARNER T A,FANG F.Implementation of machine-learning classification in remote sensing:an applied review[J].International Journal of Remote Sensing,2018,39(9):2784-2817.
[22] 李春雷,曹小朋,张林凤,等.基于机器学习算法的水驱储层相渗曲线仿真预测[J].油气地质与采收率,2022,29(6):138-142.
LI Chunlei,CAO Xiaopeng,ZHANG Linfeng,et al.Simulation and prediction of water-flooding reservoir relative permeability curve based on machine learning[J].Petroleum Geology and Recovery Efficiency,2022,29(6):138-142.
[23] 卜亚辉.基于机器学习的高含水油田剩余油预测方法[J].油气地质与采收率,2022,29(4):135-142.
BU Yahui.Prediction of remaining oil in high water cut oilfield based on machine learning[J].Petroleum Geology and Recovery Efficiency,2022,29(4):135-142.
[24] 颜世翠.基于机器学习算法和属性特征双优选的砂体岩性预测方法[J].油气地质与采收率,2022,29(1):98-106.
YAN Shicui.Prediction method of sandstone lithology based on optimized machine learning algorithms and attribute features[J].Petroleum Geology and Recovery Efficiency,2022,29(1):98-106.
[25] 李艳辉,陶莹莹,丁锐. 考虑分布时滞和随机测量数据丢失的网络控制系统非脆弱L1滤波[J]. 东北石油大学学报,2019,43(1):117-124.
LI Yanhui,TAO Yingying,DING Rui. Robust non-fragile L1 filtering for networked control systems with stochastic missing measurements and distributed delays[J]. Journal of Northeast Petroleum University,2019,43(1):117-124.

Options

文章导航

模态框（Modal）标题

摘要

本文引用格式

Abstract

参考文献