基于储层物性时变的聚合物驱数值模拟

Numerical simulation of polymer flooding considering reservoir property time variation

姜瑞忠 崔永正 胡勇 乔欣 郜益华 徐建春1

中国石油大学（华东）石油工程学院，山东 青岛 266580 中国石油北京油气调控中心，北京 100007 中海油研究总院有限责任公司，北京 100028

School of Petroleum Engineering, China University of Petroleum, Qingdao 266580, China Beijing Oil and Gas Control Center, CNPC, Bejing 100007, China CNOOC Research Institute Co., Ltd., Bejing 100028, China

我国大部分油田现已进入开发中后期，聚合物驱作为有效提高采收率的方法被大量应用于矿场实践中。大量研究发现，高含水油田开发过程中的长期水冲刷会使储层物性发生变化，但目前考虑储层物性变化的聚合物驱数值模拟研究尚少。文中建立了考虑储层物性随冲刷变化的聚合物驱数学模型，利用全隐式有限差分方法进行数值离散，并运用FORTRAN语言编写数值模拟程序。不考虑储层物性随冲刷的变化情况，将计算结果与商业软件ECLIPSE预测值进行对比，二者一致性好。计算结果表明：考虑渗透率时变后，含水率上升变快，开发效果会变差；考虑相对渗透率时变后，含水率上升变缓，开发效果会变好。对2口井进行实例验证的结果证明，考虑储层物性变化后，历史拟合效果良好。文中建立的模型更加符合油藏开发的特点，对准确预测聚合物驱油藏开发指标和剩余油分布具有一定的应用价值。

At present, most of oil fields in our country have entered the middle and late development stages, polymer flooding as a method which can effectively enhance oil recovery is widely used in oil field practice. Lots of previous studies have found that physical properties of the reservoir rock will change a lot after long?鄄term water flooding, but now there are no studies of polymer reservoir simulation which considered the variation of physical properties. Then a polymer flooding mathematical model considering the variation of physical properties based on surface flux is proposed. The model is discretized using backward finite difference method, and a new numerical polymer simulator considering the time variation is programmed by FORTRAN language. The new simulator is validated against commercial software without considering the variation of physical properties, and the results perfectly match the results of ECLIPSE. Furthermore, the time?鄄variation effects of different parameters are investigated. After considering the time variation of absolute permeability, the rising speed of water cut slows down and the ultimate oil recovery declines. After taking the time variation of relative permeability into account, the rising speed of water cut speeds up and the ultimate oil recovery is found to increase. And the simulation results fits well with the two well′s actual historical production data after considering time variation, which verifies the correctness of the model. The model can successfully characterize the variation of reservoir properties during production, and it has important application value for accurately forecasting the production performance and remaining oil distribution.

10.6056/dkyqt201906015