Reconstruction of 2D Images from 3D micro-CT Rock Images for Numerical Simulations and Reservoir-on-a-Chip Applications

Summary We have a generalized protocol that generates 2D reconstructed porous media from any 3D image stacks. By reducing the dimension of the problem, we gain a significant amount of computation time. Our initial results show that there exist few apriori unknown tuning parameters that have a strong influence on the accuracy of the macroscopic parameters such as porosity and permeability. By fine-tuning these parameters based on trial and error, we notice a good match in these macroscopic variables tested on 10 different 3D rock samples with an acceptable range of error. Further studies will focus on relating these tuneable variables according to the physical characteristics of the actual 3D porous media.