微观剩余油仿真研究综述  

Method of 4D Reservoir Modeling

作　　者：张留美[1] 潘少伟[1] 

机构地区：[1]西安石油大学计算机学院,陕西西安710065

出　　处：《技术与创新管理》2010年第6期762-764,771,共4页Technology and Innovation Management

摘　　要：传统的建模方法包括确定性建模和随机建模,但它们不能实现储层参数在时间维的预测。建立储层参数的四维模型能够揭示储层参数在四维空间的分布和演化规律,对剩余油的进一步挖潜具有重要的理论和现实意义。本文在充分吸取已有两种储层四维建模方法优点的基础上,提出了一种建立储层四维模型的新方法:首先结合实验分析、生产动态等资料,来求取历史储层参数;接着利用人工神经网络方法,对历史储层参数进行学习与训练,总结出各井点储层参数随时间的演变规律,进而对未来的井点储层参数做出预测;然后建立起构造模型,应用随机模拟或克里金插值方法来预测井间的储层参数;最后应用三维数据场可视化技术,对各个开发时期的储层参数进行显示。通过该方法建立起江苏油田庄2断块E1f11-1小层含油饱和度的四维模型,结合生产动态数据分析,发现所建的四维模型较准确的反映了E1f11-1小层含油饱和度在三维空间的分布和演化规律。The determinable modeling and stochastic modeling that are of traditional methods can＇t forecast the changes of reservoir parameters in the domain of time.The establishing of 4D model of the reservoir parameters can reveal the distribution and evolution of the law of reservoir parameters in the 4D space,which are good for the further exploitation of the remained oil.There are two methods of establishing 4D reservoir model.One is establishing the 3D model of the reservoir parameters during the development period,and the other is establishing the 4D model of the reservoir parameters with the artificial neural network approach.A new method of establishing the 4D model of reservoir is developed on the basis of the advantages of the above two methods.First,the previous reservoir parameters are calculated by the experimental and producing information.Second,the previous reservoir parameters are learned and trained by the artificial neural network,and the law of evolution of reservoir parameters of the wells is calculated,so the forward reservoir parameters of the wells are forecasted.Third,the interwell reservoir parameters are forecasted by the methods of stochastic simulation and Kriging interpolation.Last,the reservoir parameters of different development period are displayed with the method of the 3D visualization.Then the 4D reservoir model is established successfully.

关 键 词：确定性建模 随机建模 储层四维建模 人工神经网络 

分 类 号：TE357.9[石油与天然气工程—油气田开发工程]