基于MCMC的数字岩心重建方法

Reconstruction of digital core based on Markov Chain-Monte Carlo method( MCMC)

常规的多孔介质重建方法主要考虑基质和孔隙两部分,而且编程难度大,计算速度慢。针对页岩气储层各向异性且含有有机质的特点,提出了一种基于改进后的马尔科夫链-蒙特卡洛(MCMC)方法的数字岩心重建方法:先从实际页岩气储层二维切片图像中提取孔隙度和有机质,分别重建孔隙和有机质的数字岩心,再将孔隙和有机质的初始模型组合在一起构成最终的页岩气储层数字岩心。研究结果表明:通过对比分析二维MCMC重建图像和原图自相关函数,发现该方法重建图像能够较好反映原图性质;该方法计算速度快,过程简单,方法所重建的模型能够较好体现页岩储层各向异性的特点且能反映页岩储层含有有机质的特性。

The matrix and pore are only considered in conventional porous medium reconstruction methods,their programming is usually difficult,and their computing speed is low. There is organic matter in the porous medium of shale gas reservoir,and the shale gas reservoirs are generally anisotropic. Aiming at the characteristics of the porous medium of shale gas reservoir,a porous medium reconstruction method based on improved Markov Chain-Monte Carlo( MCMC) method is proposed: First,pore and organic matter are extracted from two-dimensional slice images of actual shale gas reservoir,the digital cores of pore and organic matter are reconstructed separately. Then the initial models of pore and organic matter are combined together to form the final digital core of shale gas reservoir. An example shows that this digital core reconstruction method is simple and of high computation speed; the comparison of reconstructed 2D images with original 2D images shows the digital cores of shale gas reservoir reconstructed using this method can well present the characteristics of shale gas reservoir in which there is organic matter and anisotropy.