基于模糊C均值聚类的流动单元划分方法——以克拉玛依油田五3中区克下组为例  

A new method of classifying flow units with fuzzy C-mean clustering--a case study of Kexia formation in the middle Wu3 area,Karamay Oilfield

作　　者：唐衔[1,2] 侯加根[2] 邓强[2,3] 许凡[2] 

机构地区：[1]中化集团石油勘探开发公司,北京100031 [2]中国石油大学(北京)资源与信息学院,北京102249 [3]中国石油西南油气田分公司川西北气矿,四川江油621709

出　　处：《油气地质与采收率》2009年第4期34-37,40,共5页Petroleum Geology and Recovery Efficiency

基　　金：国家科技重大专项课题"大型油气田及煤层气开发"(2008ZX05014)

摘　　要：流动单元的划分方法直接影响着储层性质的评价及油藏后期开发。针对常用的流动单元评价方法——流动带指标法容易出现聚类指标优劣交叉的问题,提出了基于模糊C均值聚类模型划分流动单元的方法。该方法通过分析多种储层属性,寻找各类的最优聚类中心,并给出样本点属于每一类的隶属度,更准确地标定出每个样品的聚类域。以克拉玛依油田五3中区克下组油藏为例,通过在聚类过程中有效地实施人工干预,解决了流动单元划分指标不匹配的问题,在生产实际中取得了较好的应用效果。The method of classifying flow units directly influences on reservoir property evaluation and later development of oil pools.Aiming at the mismatch of the clustering parameters with each other caused by the general FZI method,the method for classifying flow units based on fuzzy C-mean clustering model was put forward.By analyzing kinds of reservoir attributes,the clustering center and the sample point's membership degree of every can be found out,and the sample's clustering will be determined.Combined with the practice in the T_2K_1 pool in the middle Wu3 area,Karamay Oilfield,this new method has been proved to be practical,and perfectly avoided mismatch of the parameters for classifying the flow units adopting certain manual intervention.Good results have been achieved in the field application.

关 键 词：流动单元 模糊C均值聚类 流动带指标 克下组 克拉玛依油田 

分 类 号：TE112.23[石油与天然气工程—油气勘探]