Title

Method to Identify Complicated Lithology of Clastic Rocks in Tahe Oilfield

作者

袁瑞

Authors

Yuan Rui

单位

长江大学地球科学学院, 湖北武汉 430100; 长江大学信息与数学学院, 湖北荆州 434023

Organization

School of Geoscience, Yangtze University, Wuhan, Hubei, 430100, China; School of Information and Mathematics, Yangtze University, Jingzhou, Hubei, 434023, China

摘要

为了准确识别塔河油田石炭系复杂碎屑岩储集层的岩性,研究了基于常规测井资料的岩性识别方法.根据取心资料将储集层的碎屑岩分为砂岩、含泥砂岩、含砾砂岩和含钙砂岩4种类型;从常规测井资料中提取对岩性反映相对敏感的多种信息,并进行归一化处理;采用常规支持向量机方法和最小二乘支持向量机方法对少量已知岩性的取心岩样进行学习,识别碎屑岩岩性,建立了适用于塔河油田石炭系复杂碎屑岩储集层的岩性识别方法.塔河油田X区块的石炭系碎屑岩岩性识别表明,基于最小二乘支持向量机方法的岩性识别符合率比常规支持向量机方法提高10百分点.研究结果表明,基于最小二乘支持向量机的岩性识别方法能有效识别塔河油田石炭系复杂碎屑岩储集层的岩性.

Abstract

In order to correctly identify the complicated lithology of clasolite in Carboniferous reservoir in the Tahe Oilfield,the identification method was studied on the basis of conventional logging data.According to coring data,the lithology of the reservoir was divided into four types:sandstone,shaly sand,pebbly sandstone and calcareous sandstone.Extensive pertinent information regarding the lithology was taken from conventional logging data before processing in a normalization of data phase.The information of some known lithologies of coring samples was studied with conventional SVM and LSSVM,an identification method of lithology that suitable for characterizing carboniferous clasolite of the Tahe Oilfield has been obtained.In X block,the accuracy of the identification of the lithology with LSSVM was 10 Percentage point higher than that of conventional SVM.The result of the lithology identification showed that LSSVM was effective for identifying the complicated Carboniferous clasolite in the Tahe Oilfield reservoir.

关键词：

石炭系; 岩性识别; 常规测井资料; 最小二乘支持向量机;

Keywords:

Carboniferous; lithologic identification; conventional logging data; least squares support vector machines;

基金项目

湖北省自然科学"随钻跟踪海量数据处理与成图关键技术研究"(编号:2013CFA053)资助.

DOI

https://doi.org/10.11911/syztjs.201406013

塔河油田复杂碎屑岩岩性识别方法