节点文献

基于模拟退火算法的随机反演技术在砂体预测中的应用

Application of stochastic inversion based on simulated annealing algorithm for sand body prediction

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 杨立强邬长武董宁

【Author】 YANG Li-qiang,WU Chang-wu,DONG Ning(Exploration and Production Research Institute,SINOPEC,Haidian 100083,Beijing,China)

【机构】 中国石油化工股份有限公司石油勘探开发研究院

【摘要】 塔河油田Su地区石炭系卡拉沙依组砂体的埋藏深度较深,厚度一般小于10m,且横向变化较大,砂、泥岩之间的阻抗差异较小,地震资料不能区分砂泥岩,因此储层描述困难.为此,针对该区目的层段开展了测井约束地震反演研究.本文选用当前理论上较为突出,且实际应用中较为可靠的随机反演方法,同时将模拟退火全局优化方法引入进来,采用约束稀疏脉冲的反演结果作为随机反演的输入模型,大大提高了模型的精度,增强了随机反演的可靠性.通过对比分析,反演剖面的分辨率和信噪比有所提高,能较好地反映储层性质的横向变化,预测结果与实际钻井符合,证实该反演技术在该地区具有一定的实用性.

【Abstract】 The sand bodies of Carboniferous Karashayi formation in Tahe oil field are usually buried deeply,with a thickness of less than 10 m and a big lateral change.The resolution of seismic data can not differentiate sandstone from mudstone,making it difficult for reservoir description.In order to solve the problem,corresponding reservoir logging constrained seismic inversion methods are studied in the area.A new method proposed here by adopting inversion results of constrained sparse spike inversion as stochastic inversion input,would greatly enhance the modeling accuracy and stochastic inversion reliability.Simulated annealing algorithm of global optimization is also used.Comparison indicates that the resolution of Inversion and signal-noise ratio are raised,by which the lateral changes of reservoir properties are properly reflected with predicted result being correspondence to that of operation.It shows that the technique of high resolution logging constrained inversion is adaptive in the area.

  • 【文献出处】 地球物理学进展 ,Progress in Geophysics , 编辑部邮箱 ,2013年01期
  • 【分类号】P631.4
  • 【被引频次】11
  • 【下载频次】280
节点文献中: 

本文链接的文献网络图示:

本文的引文网络