论文部分内容阅读
基于达西二维产量公式 ,通过研究储层产能的理论公式 ,并分析储层产能的两类主要影响因素 (人为因素和储层因素 ) ,认为在一个油区内各种作业方式等人为因素大致相同的前提下 ,储层产能主要取决于储层的性质。在此基础上 ,本文建立起储层产能与测井数据之间的关系 ,采用人工神经网络技术建立了储层产能预测系统。该系统采用了 5个评价参数 (有效孔隙率、渗透率、含油饱和度、泥质含量和产能系数 )作为输入节点 ,通过人工神经网络 (ANN)模型预测出表示储层动态特征的结果。将本方法用于预测新疆克拉玛依油田八区克上组油层的产能 ,取得了良好的效果 ,从而证实了本方法的有效性。
Based on the Darcy two-dimensional yield formula, by studying the theoretical formula of reservoir productivity and analyzing two main types of influencing factors (man-made factor and reservoir factor) of reservoir productivity, it is considered that man-made factors such as various operating modes in an oil zone Under roughly the same premise, reservoir productivity depends mainly on the nature of the reservoir. On this basis, this paper establishes the relationship between reservoir productivity and well logging data, and uses artificial neural network technology to establish a reservoir capacity prediction system. The system uses five evaluation parameters (effective porosity, permeability, oil saturation, shale content and productivity coefficient) as input nodes, and predicts the result of reservoir dynamic characteristics through artificial neural network (ANN) model. The method is applied to predict the productivity of Ke-Shek Formation in the eight districts of Karamay Oilfield in Xinjiang. The result shows that the method is effective.