致密油藏水平井体积压裂初期产能预测  

Forecast of Initial Horizontal Well Productivity in Tight Reservoirs by Volumetric Fracturing Process

作　　者：王威[1] 

机构地区：[1]中国石油大庆油田有限责任公司勘探开发研究院,黑龙江大庆163712

出　　处：《新疆石油地质》2016年第5期575-579,共5页Xinjiang Petroleum Geology

基　　金：国家科技重大专项(2011ZX05014)

摘　　要：针对体积压裂井压后初期产能预测难度大的问题,依据大庆油田龙26井区和垣平1井区体积压裂地质、工程及动态数据,分析了压后初期产能的影响因素。采用灰色关联分析法、复相关系数法和熵值法,综合分析体积压裂后初期产能与影响因素的相关程度,确定了主要的影响因素。研究结果表明,影响体积压裂初期产能的主要因素依次为加砂量、加液量、地层压力和渗透率。在此基础上,应用支持向量机方法预测油井压裂后的初期产量。经实际压裂效果验证,支持向量机方法对压后初期产量的预测准确率较高,比传统的多元线性回归预测法更准确,此研究对类似的致密油气藏的开发具有一定的指导和借鉴意义。Aiming at the difficulties in initial well productivity forecast after volumetric fracturing, the article analyzes the influencing factors of initial well productivity after fracturing based on the data of geology, engineering and production performance by volumetric fracturing process in District Long-26 and Yuanping-1 of Daqing oilfield. Gray correlation analysis method, multiple correlation coefficient method and entropy evaluation method are used to analyze the degree of correlation between the initial well productivity after volumetric fracturing and the influencing factors, then the main factors are determined. The study result shows that the main factors influencing the initial well productivity after volumetric fracturing include sand adding volume, liquid adding volume, formation pressure and permeability, based on which the support vector machine （SVM） method is used to forecast the initial well production after fracturing. The actual fracturing effect verifies that SVM method can more accurately forecast the initial production after fracturing than conventional multiple linear regression （MLR） method, which may provide guidance and reference for similar tight oil and gas reservoir development.

关 键 词：致密油藏 水平井 体积压裂 影响因素 支持向量机方法 产能预测 

分 类 号：TE357.14[石油与天然气工程—油气田开发工程]