On the application of CFD for establishing a protocol for Representative Elementary Volume (REV) definitionin pore-scale modeling of reservoir rock

Flow and transport in porous media is vital to many areas of engineering and science, including groundwater hydrology, subsurface CO2 sequestration, and Hydrocarbon production from underground reservoirs. Image-based pore-scale modeling of transport phenomena has become an important tool for understanding the complicated relationships between the pore structure and the measurable macroscopic properties; in particular, the effective permeability. The main venues for pore-scale modeling are network extraction and the Computational Fluid Dynamics (CFD). The latter CFD approach has been rendered successful in handling cases with large heterogeneity, and attempts to solve the Navier-Stokes equations on the complex boundary conditions of the inner-rock pore spaces. The present work uses the OpenFOAM software to determine the effective permeability of (Bentheimer) sandstone rock. Our methodology was based on extraction of stereolithography (STL) surfaces from the CT-scan images of reservoir rock, which was implemented by our in-house developed computer code. The code is capable of preserving the precision of surface extraction by as low as the resolution of the CT-scan images (3.0035 micrometers). Simulation of fluid flow within the STL surfaces was later attempted by OpenFOAM. To explore the effect of image subscaling on effective permeability, different subsets of the entire micro-structure domain were used for CFD simulation. The results indicate that variation in effective permeability of subset rock samples decreases along with increase in the size of the pore-scale sample and favorably converges to the value established for the entire domain, as the size approaches 250x250x250 (voxels). Similar results were obtained as for the porosity of subset pore-scale samples. The results are invaluable in establishing an authentic protocol for definition of Representative Elementary Volume (REV) within the CFD pore-scale modeling context.

Keywords:

Computational Fluid Dynamics , Pore-scale modeling , OpenFOAM , Stereolithography , Sandstone reservoir rock