裂缝识别的动态模型与应用

Dynamic model for identification of fractures and its application

程时清 李菊花 李相方 陈思

CHENG Shi-qing LI Ju-hua LI Xiang-fang CHEN Si

中国石油大学石油工程教育部重点实验室, 北京 102249 大庆油田设计院, 黑龙江大庆 163712

大量低渗透油藏开发实践表明,油井动态特征与储层物性和裂缝分布有关,研究天然裂缝和注水诱发裂缝分布对注水开发井网调整具有非常重要意义。将油水井见水周期、水线推进速度、含水上升速度、累积注采比、注水压力等5个参数作为动态变量,基于Bayes后验概率公式建立动态裂缝识别模型,将待判别的样品与已知类型的样品进行类比,以确定待判别样品的归类。通过在油藏实际应用,分析了动态参数与裂缝分类的关系,确定该模型能识别地层裂缝分布和油水井连通状况,由此可以判别裂缝大致方位。与其他裂缝识别方法相比,该模型把动态数据作为重要依据,更能体现油水井裂缝的动态变化。

A host of research practices show that the behavioral characteristics of oil wells have relations with the distribution of formation physical property and fractures.So it is very important to clarify the distribution of natural fractures and artificially induced factures so as to adjust water flooding well patterns.This paper based on Bayes posterior probability formula,establishes a dynamic model for the identification of fractures by using five performance parameters such as water breakthrough period,waterline movement velocity,rising velocity of water cut,cumulative injection production ratio,and injection pressure.It carries out an analogy between the samples to be distinguished and the known ones to confirm how to classify them.The relationship between dynamic model parameters and fracture parameters is analyzed on the basis of practical application,showing that the dynamic model for identification of fractures can be used to identify the fracture distribution and the connection between oil and water wells.Then the approximate direction of fractures can be confirmed.The new model is built in accordance with dynamic data,and more fully embodies the dynamic change of well fractures as compared to other approaches.

10.3969/j.issn.1000-7393.2007.02.030