桂金咏^1,2,,
高建虎²,
李胜军²,
陈康³,
王洪求²,
陈启艳²
1. 中国石油大学(华东)地球科学与技术学院, 青岛 266580
2. 中国石油勘探开发研究院西北分院, 兰州 730020
3. 中国石油西南油气田勘探开发研究院, 成都 610041

基金项目: 国家重大科技专项（2016ZX05007-006）与中石油科技专项（2019B-0607）联合资助


详细信息

作者简介: 桂金咏, 男, 1986年生, 博士研究生, 工程师, 主要从事叠前地震反演与地震储层预测研究.E-mail:guijy@petrochina.com.cn

中图分类号: P631

收稿日期:2018-09-25
修回日期:2019-12-10
上线日期:2020-01-05

The method of seismic lithofacies prediction based on weighted statistics of elastic parameters

GUI JinYong^1,2,,
GAO JianHu²,
LI ShengJun²,
CHEN Kang³,
WANG HongQiu²,
CHEN QiYan²
1. School of Geoscience, China University of Petroleum(Huadong), Qingdao 266580, China
2. PetroChina Research Institute of Petroleum Exploration & Development-NorthWest, Lanzhou 730020, China
3. PetroChina Exploration and Development Institute of Southwest Oil & Gas Field Company, Chengdu 610041, China


MSC: P631

--> Received Date: 25 September 2018
Revised Date: 10 December 2019
Available Online: 05 January 2020


摘要
摘要:岩相信息能够反映储层岩性及流体特征，在地震储层预测中具有重要作用.常规方法主要利用与岩相信息关系密切的弹性参数定性或定量地转化为岩相信息.在实际应用中，弹性参数的获取主要基于叠前地震反演技术.而不同弹性参数的叠前地震反演精度间存在着差异，势必影响岩相的整体预测精度.本文提出对弹性参数进行加权统计来预测岩相.首先，基于贝叶斯理论，引入权重系数来调节弹性参数信息的采用量，构建出最终的目标反演函数；其次，考虑到勘探初期缺少明确的测井岩相信息，提出利用高斯混合分布函数来自动估算岩相先验概率；最后，根据输入弹性参数的取值，计算每类岩相对应的后验概率密度，将目标反演函数取最大后验概率密度时对应的岩相类别作为最终预测的岩相.新方法旨在减少弹性参数精度间的精度差异对岩相预测结果的影响，以期提高地震岩相的预测精度.模型与实际资料测试均表明该方法可行、有效且预测精度较高.
关键词: 地震岩相/
弹性参数/
加权统计/
贝叶斯理论/
混合高斯分布

Abstract:The lithofacies information can reflect reservoir lithology and fluid characteristics, and play an important role in seismic reservoir prediction. In practice application, the elastic parameter is mainly obtained by prestack seismic inversion technology. However, there are differences in the accuracy of prestack seismic inversion with different kinds of elastic parameters, which will inevitably affect the overall prediction accuracy of the lithofacies. In this paper, a weighted statistics of elastic parameters method is proposed to predict lithofacies. Firstly, based on Bayesian theory, weighted coefficients are introduced to adjust the amount of elastic parameter information used in the final objective inversion function. Secondly, considering the lack of clear logging lithofacies information in the early exploration period, the Gauss mixture distribution function is proposed to estimate the lithofacies prior probability automatically. Finally, according to the value of the input elastic parameters, the posterior probability density of each lithofacies type is calculated, and the lithofacies type corresponding to the maximum posterior probability density of the objective inversion function is taken as the final predicted lithofacies. The new method aims to reduce the influence of the accuracy difference of elastic parameters on the lithofacies prediction results. Both the model and the actual data test show that the method is feasible, effective and has high prediction accuracy.
Key words:Seismic lithofacies/
Elastic parameter/
Weighted statistics/
Bayesian theory/
Gauss mixture distribution



PDF全文下载地址:

http://www.geophy.cn/data/article/export-pdf?id=dqwlxb_15306