Analytical Upgridding of Geocellular Mode in the Presence of Gravity-Dominated Displacement

Abstract

Summary In simulating enhanced-oil-recovery (EOR) processes, it is critical that all the flow behaviors be properly accounted for in the simulation. Because of computation limitations, long calculation time, and complexity of physics, geological models cannot be directly used for fieldwide simulations. Upgridding reduces the number of gridblocks in the simulation model and therefore makes the simulation more efficient. An appropriate upgridding process needs to preserve the dynamic behavior of the fine-scale model. We propose such an analytical methodology. Our new technique is based on preserving the characteristics, which are based on the fractional-flow concept specifically modified for vertical flow between the layers. We develop our method with a specific application to gravity-dominated displacement. In upgridding the fine-scale model, we have developed a criterion by which the sequence in which the fine-scale layers are combined is proposed such that fractional-flow characteristics based on the fine-scale model are honored. Using this methodology, we can determine not only the sequence in which layers are combined, but also to what extent we can upgrid the fine-scale model. The proposed methodology is developed for two-phase, 2D flow under the effect of gravity-segregated displacement. However, it is also tested for three-phase, 3D flow in gravity-dominated displacement with moderate effect of viscous and capillary forces. The proposed solution is analytical; therefore, it is computationally efficient. We have validated the methodology with both synthetic and field examples and demonstrate that the proposed methodology is superior to conventional proportional layering and variance-based methodologies.