石油学报 ›› 2019, Vol. 40 ›› Issue (2): 182-189.DOI: 10.7623/syxb201902006

• 地质勘探 • 上一篇    下一篇

波形分类技术在缝洞型储层流体识别中的应用

逯宇佳1, 曹俊兴1, 刘哲哿1, 田仁飞1, 肖学2   

  1. 1. 成都理工大学地球物理学院 四川成都 610059;
    2. 中国石油化工股份有限公司河南油田分公司油气勘探管理部 河南南阳 473132
  • 收稿日期:2018-05-29 修回日期:2018-12-11 出版日期:2019-02-25 发布日期:2019-03-05
  • 通讯作者: 曹俊兴,男,1961年9月生,1983年获西北大学地质学系岩石矿物学专业学士学位,1996年获成都理工大学应用地球物理专业博士学位,现为成都理工大学教授,主要从事地球探测与信息技术研究。Email:caojx@cdut.edu.cn
  • 作者简介:逯宇佳,男,1995年5月生,2016年获成都理工大学勘查技术与工程专业学士学位,现为成都理工大学地球探测与信息技术专业硕士研究生,主要从事地震资料处理和解释研究。Email:1033415284@qq.com
  • 基金资助:

    国家重点研发项目(2016YFC0601100)资助

Application of waveform classification technology for fluid identification in fractured-vuggy reservoir

Lu Yujia1, Cao Junxing1, Liu Zhege1, Tian Renfei1, Xiao Xue2   

  1. 1. College of Geophysics, Chengdu University of Technology, Sichuan Chengdu 610059, China;
    2. Oil and Gas Exploration and Management Department, Sinopec Henan Oilfield Company, Henan Nanyang 473132, China
  • Received:2018-05-29 Revised:2018-12-11 Online:2019-02-25 Published:2019-03-05

摘要:

塔里木盆地塔河地区奥陶系碳酸盐岩缝洞型储层的缝洞空间分布散乱,缝洞的规模、形态和内部结构等特征差异明显,导致缝洞型储层的流体难以识别。缝洞型储层内部地震波形的微弱差异在一定程度上能够表征其内部流体特征。传统上沿层段提取地震波形的波形分类技术无法准确地定位缝洞型储层。应用波形分类与三维缝洞体雕刻相结合的新技术识别缝洞流体,可为降低勘探风险、提高钻井成功率提供一种新的思路。首先利用瞬时振幅属性建立缝洞体三维空间模型,确定缝洞储层的空间位置,从而获得缝洞储层内部的地震波形;使用特征加权K近邻法,结合油气井资料,对缝洞型储层进行波形分类,建立储层内流体类型与地震波形之间的网络关系,并将其应用于未知区域的储层流体识别;最终得到缝洞型储层的流体类型。理论模型和实际地震资料测试结果表明该方法具有一定的实用价值。

关键词: 三维缝洞体雕刻, 波形分类, 特征加权K近邻法, 流体识别, 监督学习

Abstract:

The fractured-vuggy reservoir of Ordovician carbonates in Tahe area, Tarim Basin is scattered in space, characterized by great differences in the size, shape and internal structure, leading to a difficulty in identifying the reservoir fluid. The weak difference of seismic waveforms in fractured-vuggy reservoir can characterize its internal fluid characteristics to some extent. The traditional waveform classification technology extracting seismic waves along the layer is unable to locate the fractured-vuggy reservoir accurately. A new method combining waveform classification technology with 3D fractured-vuggy body engraving technology is put forward to identify the fluid of fractured-vuggy reservoirs, which provides a new idea for reducing exploration risk and improving drilling success rate. Firstly, the instantaneous amplitude attribute is used to establish a 3D space model of the fractured-vuggy bodies. The spatial location of fractured-vuggy reservoirs will be determined to obtain the seismic waveform inside fractured-vuggy reservoirs. Then, the feature weighted K-nearest neighbor method in combination with oil and gas well data is applied to classify the waveforms of fracture-cavity reservoirs, so as to create the network relationship between fluid types and seismic waveforms in the reservoir and further apply it to reservoir fluids in unknown regions. Finally, the fluid type of the fractured-vuggy reservoir is obtained. The testing results of theoretical model and practical seismic data show that this method has a certain practical value.

Key words: three-dimensional fractured-vuggy body description, waveform classification, feature weighted K-nearest neighbor method, fluid identification, supervised learning

中图分类号: