摘要
参数拟合的传统页岩气井产能预测方法存在一定的局限性,引入基于支持向量机的非参数大数据分析方法进行页岩气井产能预测研究。根据生产数据记录以及井底压力随生产过程的变化规律,建立e-SVR支持向量回归模型,对长宁页岩气某区块实际生产数据分别进行了单井和多井的训练及预测检验,其中单井回归检验的相关度系数达到0. 958 556,体现了该方法优秀的回归能力;多井学习模型在前95 d区间内对单井数据的预测也达到了接近单井回归的效果,体现了该方法在密集数据区间内较好的预测能力,为页岩气产能预测提供了新的思路。
The traditional productivity forecast method of shale gas well based on parameter matching has some limitations. This paper introduced a non-parametric big data analysis method based on support vector machine( SVM). According to the recorded production data and the change law of bottomhole pressure with the production process,the e-SVR( support vector regression) model was established,and it was tested by single-well and multi-well training and forecasting based on actual production data of a block of Changning shale gas field. The correlation coefficient of single-well regression test reaches 0. 958 556,which reflects the excellent regression performance of this method. The forecasting of the multi-well learning model based on single well data in the first 95 d interval is close to the effect of the single-well regression model,which reflects the better forecasting performance of the method in the dense data interval.The proposed method provides a new idea for forecasting shale gas productivity.
引文
[1]ARPS J J.Analysis of decline curves[J].Transactions of the AIME,1945,160(1):228-247.
[2]FETKOVICH M J.Decline curve analysis using type curves[J].Journal of Petroleum Technology,1980,32(6):1065-1077.
[3]CARTER R D.Type curves for finite radial and linear gasflow systems:constant-terminal-pressure case[J].Society of Petroleum Engineers Journal,1985,25(5):719-728.
[4]PALACIO J C,BLASINGAME T A.Decline curve analysis using type curves-analysis of gas well production data[J].SPE 25909,1993.
[5]AGARWAL R G,GARDNER D C,KLEINSTEIBER S W,et al.Analyzing well production data using combined type curve and decline curve analysis concepts[C].SPE 49222,1998.
[6]肖阳,江同文,冯积累,等.不关井试井分析在桑南西生产动态分析中的应用[J].油气井测试,2012,21(1):19-21.XIAO Yang,JIANG Tongwen,FENG Jilei,et al.Application of well test analysis without closing well for production performance in the Sangnan Oilfield[J].Well Testing,2012,21(1):19-21.
[7]梅显旺,罗梅,马威奇,等.致密油水平井多级压裂后产能影响因素分析[J].油气井测试,2016,25(4):29-32.MEI Xianwang,LUO Mei,MA Weiqi,et al.Analysis of influencing factors on productivity after multistage fracturing to tight oil of horizontal well[J].Well Testing,2016,25(4):29-32.
[8]汪益宁,李洪,肖伟,等.生产数据分析在哈拉哈塘油田储层评价中的应用[J].钻采工艺,2016,39(6):46-49.WANG Yining,LI Hong,XIAO Wei,et al.Application of production data analysis in carbonate reservoir evaluation at Halahatang Oilfield[J].Drilling&Production Technology,2016,39(6):46-49.
[9]LI Q,LI P,PANG W,et al.A new method for production data analysis in shale gas reservoirs[J].Journal of Natural Gas Science and Engineering,2018,56:368-383.
[10]方瑞明.支持向量机理论及其应用分析[M].北京:中国电力出版社,2007:66-92,131-159.
[11]刘敬,谷利泽,钮心忻,等.基于单分类支持向量机和主动学习的网络异常检测研究[J].通信学报,2015,36(11):136-146.LIU Jing,GU Lize,NIU Xinqi,et al.Research on network anomaly detection based on one-class SVM and active learning[J].Journal on Communications,2015,36(11):136-146.
[12]薛磊,吴炜,杨晓敏,等.一种基于支持向量回归的多层原油分层产能贡献预测模型[J].石油化工高等学校学报,2006,19(4):88-92.XUE Lei,WU Wei,YANG Xiaomin,et al.A forecasting model of productivity contribution in single zone of multiple-zone production based on support vector regression[J].Journal of Petrochemical Universities,2006,19(4):88-92.
[13]李卓,柴滨景,胡继东.支持向量机在油田系统建模中的应用[J].现代电子技术,2007,30(1):162-164.LI Zhuo,CHAI Binjing,HU Jidong.A novel approach to model in oilfield based on support vector machine[J].Modern Electronics Technique,2007,30(1):162-164.
[14]石广仁.支持向量机在裂缝预测及含气性评价应用中的优越性[J].石油勘探与开发,2008,35(5):588-594.SHI Guangren.Superiorities of support vector machine in fracture prediction and gassiness evaluation[J].Petroleum Exploration and Development,2008,35(5):588-594.
[15]汪旭颖.基于支持向量回归机的油田异常井预警模型研究[D].大庆:东北石油大学,2015.WANG Xuying.Research on early warning model of oilfield abnormal wells based on support vector regression[D].Daqing:Northeast Petroleum University,2015.
[16]赵小龙.基于支持向量机的深层地应力预测模型[J].特种油气藏,2016,23(1):139-141.ZHAO Xiaolong.SVM deep geo-stress prediction model[J].Special Oil&Gas Reservoirs,2016,23(1):139-141.
[17]李航.统计学习方法[M].北京:清华大学出版社,2012:95-134.
[18]Vapnik V.The nature of statistical learning theory[M].Springer Science&Business Media,2013:133-140.