基于岩石重构图像的核磁共振响应模拟  

Simulating NMR Responses in Porous Media Based on Reconstructed Digital Rock Image

作　　者：成家杰[1] 肖立志[1] 许巍[1] 张宗富[1] 安天琳[1] 邓峰[1] 刘化冰[1] 

机构地区：[1]油气资源与探测国家重点实验室,中国石油大学(北京),北京102249

出　　处：《波谱学杂志》2013年第3期336-344,共9页Chinese Journal of Magnetic Resonance

基　　金：国家自然科学基金资助项目(41074102,41130417)

摘　　要：能够反映孔隙介质内部真实复杂结构的CT图像经常被用于模拟研究中并能取得较好的结果.但是,在很多情况下缺少三维CT图像或者其分辨率受限,此时,更易于获取的二维薄片图像被用于构建三维的孔隙结构图像.假定孔隙介质各向同性,基于二维图像,使用多点统计方法可以构建岩石的三维孔隙结构.基于数字图像,采用随机游走方法模拟得到核磁共振响应,然后由模拟得到的回波串反演得到T2分布.研究表明,岩石重构图像中的模拟结果与CT图像中的模拟结果有较好的一致性.基于数字图像的核磁共振响应模拟为分析不同孔隙结构的核磁共振响应提供了便利,同时,模拟结果也为验证孔隙结构的重构效果提供了依据.Micro-CT images revealing pore space are widely used in simulation studies of porous media. Compared to three-dimensional （3D） CT images, two-dimensional （2D） thin section images are more readily available. In this paper, a multiple-point statistics （MPS） method was implemented to construct 3D pore space using a stack of 2D thin section images of a rock, assuming the medium is isotropic. A random walk model describing Brownian movement of particles in the rock was then used to simulate the NMR responses. The simulation results obtained with the reconstructed 3D pore space image were found to be similar to those obtained with the true 3D micro-CT image. It is coneluded that 3D image reconstructed from 2D thin section images shows similar pore space structure as than revealed by micro-CT images, and thus can be used for porescale simulation in porous media.

关 键 词：核磁共振(NMR) 数字图像 随机游走 多孔介质 多点统计 重构 

分 类 号：O482.53[理学—固体物理]