地球物理学进展 ›› 2018, Vol. 33 ›› Issue (5): 1911-1918.doi: 10.6038/pg2018BB0426

• 应用地球物理学Ⅰ(油气及金属矿产地球物理勘探) • 上一篇    下一篇

乍得B盆地密度和声波时差曲线异常自动识别与重构

何苗1,于海峰2(),田中元1,计智锋1   

  1. 1. 中国石油勘探开发研究院,北京 100083
    2. 大庆石油管理局钻探工程公司吉林探区事业部,吉林松原 138000
  • 收稿日期:2017-12-13 修回日期:2018-06-28 出版日期:2018-10-20 发布日期:2019-01-11
  • 作者简介:何苗,男,1992年生,硕士,主要从事地震反演储层预测研究工作.
  • 基金资助:
    国家重大科技专项“海外重点探区目标评价与未来领域选区选带研究”(2016ZX05029-005)

Automatic recognition and reconstruction of abnormal density and acoustic log curves: a case study from B basin,Chad

HE Miao1,YU Hai-feng2(),TIAN Zhong-yuan1,JI Zhi-feng1   

  1. 1. Research Institute of Petroleum Exploration and Development,Beijing 100083,China
    2. Daqing Oilfield Drilling Engineering Company, Jilin Songyuan 138000, China
  • Received:2017-12-13 Revised:2018-06-28 Online:2018-10-20 Published:2019-01-11

摘要:

密度和声波时差测井曲线是地震解释和储层反演中的两条关键曲线,它们的质量直接影响到合成地震记录的好坏,进而影响地震解释和储层反演的结果.目前测井曲线的异常分析建立在人工识别的基础上,工作量繁重且识别误差较大.本文为有效开展乍得B盆地地震储层反演工作,通过对目的层段测井资料质量分析,划分出密度和声波时差曲线的三种异常类型——Ⅰ、Ⅱ和Ⅲ型,并建立对应的测井响应标准,在此基础上编写测井曲线异常自动识别模块,实现了曲线异常井段及类型的快速识别.针对异常井段提出了利用改进的聚类分析方法(MMRGC)实现异常井段、异常曲线的重构.通过对该地区异常曲线重构前后地震合成记录标定以及测井约束反演质量分析,认为该套方法是可靠的,有效提高了工作效率,为地震反演工作奠定良好基础.

关键词: 自动识别, 曲线重构, 合成记录, 改进的聚类分析方法

Abstract:

Density and acoustic log curves are the two key curves in seismic interpretation and reservoir inversion, and the quality of logging curves would directly affect the results of synthetic seismic records, which would further influence relevant seismic interpretation and reservoir inversion. Currently, the analysis of abnormal log curve is based on artificial recognition, which is time-consuming and inaccurate. By analyzing the logging data of the target layer,three types of abnormal density and DT log curves were established—Ⅰ、Ⅱ and Ⅲ. Based on the abnormal logging curve automatical identification module, we identify the abnormal intervals and types quickly. We reconstruct abnormal curves via MMRGC curve reconstruction method. The analysis of the synthetics and logging constrained inversion accuracy of the reconstructed results demonstrated that these methods are reliable and efficient, thus it possess the potential to be extensively utilized in the study of seismic inversion.

Key words: automatic recognition, curve reconstruction, synthetic records, improved MRGC

中图分类号: