石油学报 ›› 2015, Vol. 36 ›› Issue (s1): 59-67.DOI: 10.7623/syxb2015S1007

• 地质勘探 • 上一篇    下一篇

湖相致密复杂岩性地震识别技术

赵万金1, 唐传章2, 王孟华2, 王余泉2, 杨德相2, 赵伟森2, 周从安2, 常建华2, 卢昊2   

  1. 1. 中国石油勘探开发研究院西北分院 甘肃兰州 730020;
    2. 中国石油华北油田公司 河北任丘 062552
  • 收稿日期:2015-01-26 修回日期:2015-05-22 出版日期:2015-11-30 发布日期:2015-12-09
  • 通讯作者: 赵万金,男,1980年9月生,2003年获石油大学(华东)学士学位,2006年获中国石油大学(华东)硕士学位,现为中国石油勘探开发研究院西北分院工程师,主要从事地震储层预测技术研究与应用。Email:zhao_wj@petrochina.com.cn
  • 作者简介:赵万金,男,1980年9月生,2003年获石油大学(华东)学士学位,2006年获中国石油大学(华东)硕士学位,现为中国石油勘探开发研究院西北分院工程师,主要从事地震储层预测技术研究与应用。Email:zhao_wj@petrochina.com.cn
  • 基金资助:

    中国石油天然气股份有限公司重大科技专项(2014E-035)和中国石油天然气股份有限公司物探攻关项目(2013-03-06)资助。

A seismic recognition technology of lacustrine complex tight lithology

Zhao Wanjin1, Tang Chuanzhang2, Wang Menghua2, Wang Yuquan2, Yang Dexiang2, Zhao Weiseng2, Zhou Congan2, Chang Jianhua2, Lu Hao2   

  1. 1. Northwest Branch, PetroChina Research Institute of Petroleum Exploration and Development, Gansu Lanzhou 730020, China;
    2. PetroChina Huabei Oilfield Company, Heibei Renqiu 062552, China
  • Received:2015-01-26 Revised:2015-05-22 Online:2015-11-30 Published:2015-12-09

摘要:

针对湖相复杂致密岩性地震预测难点多的特点,提出一种在高分辨率层序约束下利用贝叶斯统计方法识别致密岩性的思路。应用地震时频旋回技术和Wheeler变换与反变换技术建立三维高分辨率层序格架,划分沉积体系域。运用中观尺度测井数据开展岩石物理交会分析,利用实验室X-射线衍射数据(XRD)和等效介质模型进一步从微观尺度分析,最终确定拟声波阻抗和纵、横波速度比是区分岩性的敏感参数。在此基础上,开展高分辨率层序约束下的波阻抗及纵、横波速度比反演,而后利用反演属性进行贝叶斯统计,在井上岩性识别精度高的前提下,利用得到的条件概率分布预测空间岩性最佳分布,最后对岩性分类结果进行对比分析。实际应用证明,在高分辨率层序约束反演的基础上利用贝叶斯统计识别复杂致密岩性具有较高的吻合率,该项技术适用于湖相复杂致密岩性识别。

关键词: 致密岩性, 高分辨率层序, 体系域, 岩石物理, 反演, 贝叶斯统计

Abstract:

Aiming at much more difficulties in predicting seismic prediction of complex lacustrine tight lithology, under constraint of high-resolution sequence, a method of recognizing tight lithology using Bayes statistics way was proposed. High-resolution 3D sequence framework was established using seismic time-frequency cycle technology together with forward/inverse Wheeler transform technology, and sedimentary system tracts were divided. Petrophysics cross-plot analysis was conducted with mesoscale logging data. Laboratorial X ray diffraction data and equivalent medium model were used for further micro-scale analysis, so as to determine pseudo acoustic impedance and ratio of P-and S-waves velocity (Vp/Vs)as sensitive parameters for lithology recognition. On this basis, acoustic impedance and Vp/Vs inversion was carried out under constraint of high-resolution sequence. Then Bayes statistics was conducted using inversed attributes. Under high-precision lithology recognition, the obtained conditional probability distribution was adopted to predict optimal distribution of spatial lithology, and classification results of lithology were conducted comparative analysis. The practical application shows that on the basis of constrained inversion with high-resolution sequence, the complex tight lithology recognized by Bayes statistics had a high consistency, and this technology is suitable for recognition of complex lacustrine tight lithology.

Key words: tight lithology, high-resolution sequence, system tract, petrophysics, inversion, Bayesian statistics

中图分类号: