聚合物驱含水率变化定量表征模型  

A quantitative mathematic model for polymer flooding water-cut variation

作　　者：赵辉[1] 李阳[2] 曹琳[3] 

机构地区：[1]中国石油大学(华东) [2]中国石油化工股份有限公司 [3]中国石化胜利油田有限公司东辛采油厂

出　　处：《石油勘探与开发》2010年第6期737-742,共6页Petroleum Exploration and Development

基　　金：国家"863"计划"渤海油田深部调剖提高采收率技术研究"项目(2003AA602140)

摘　　要：基于实际聚合物驱含水率变化曲线的特征及定量表征模型的建立原则,建立聚合物驱含水率变化定量表征模型。刻画聚合物驱含水率变化曲线特征的主要参数包括含水率开始下降时注入聚合物孔隙体积倍数、含水率下降最大值时注入聚合物孔隙体积倍数、含水率恢复到初始注聚含水率值时对应的注入聚合物孔隙体积倍数及含水率下降最大值。基于粒子群优化算法给出了定量表征模型参数的拟合方法。同时,综合油藏数值模拟、正交设计和支持向量机方法建立了考虑多因素组合影响的聚合物驱含水率变化特征参数预测模型。计算实例表明,应用所建立的模型,能便捷地预测聚合物驱含水率变化,把握其整体趋势,指导聚合物驱方案的实施。On the basis of the dynamic features of polymer flooding water-cut variation curve and a couple of major establishment principles,a novel quantitative characterization model for polymer flooding water-cut variation was built.The model mainly includes four characteristic parameters with specific meanings,namely the polymer pore volume injected at the initial time of water-cut decrement,the corresponding polymer pore volume injected at maximum water-cut decrement,the polymer pore volume injected when water-cut recovering to the initial water-cut,and the maximum water-cut decrement.The automatic regression method for specific parameters of the model was proposed on the particle swarm optimization algorithm.Based on reservoir numerical simulation,in combination with orthogonal experimental design and support vector machine approach,the prediction model of characteristic parameters was presented which can fairly consider the influence of the combination of multi-factors.The practical application results indicated that the water-cut variation and performance trend can be easily determined for the scheme compilation and adjustment of polymer flooding by using the proposed method.

关 键 词：聚合物驱 含水率 粒子群算法 影响因素 支持向量机 

分 类 号：TE357.46[石油与天然气工程—油气田开发工程] TE331