论文部分内容阅读
油藏开发分类是当前一项重要研究课题。在大概念的地质分类之后,精确地划分亚类,是现代油藏工程学和油田开发地质学向着定量化方向发展所必需的。本文提出应用多变量统计学中的聚类分析方法进行油藏的开发数值分类。文中提出了以储层地质、工程四个方面特性的27项指标做为分类的依据和依据数字信息的参数预处理方法,及聚类分析的应用。文章以胜利油田特征不同的39个油藏为样本,进行了实例分类。分类结果可以反映各类油藏之间在开采特点及油水运动规律上的区别和联系,与多年来在开发实践上的认识一致。
Reservoir development classification is currently an important research topic. Subdividing the sub-classes precisely after the conceptual geological classification is necessary for the development of modern reservoir engineering and oilfield development geology towards quantitative direction. This paper presents a numerical classification of reservoir development using the method of cluster analysis in multivariate statistics. In this paper, 27 indexes based on four aspects of reservoir geology and engineering are put forward as the basis of classification and parameter preprocessing based on digital information, and the application of cluster analysis. In this paper, 39 reservoirs with different characteristics of Shengli Oilfield are taken as examples, and the examples are classified. The classification result can reflect the difference and connection between various types of reservoirs in the mining characteristics and the law of oil-water movement, which is consistent with the understanding of development practice over the years.