The use of numerical finite-difference models to substantiate EOR implementation decisions has significantly increased, along with the need for accurate numerical representation of the driving forces and processes dominating the fluid flow. As reservoir characterization improves, so does the need for more representative models, since the perception appears to exist that increased accuracy and therefore predictive power, is linearly dependent on model size. Technology advances in hardware and software have overcome many of the past limitations on numerical model size, but not eliminitated them; hence correct upscaling is not only relevant but still very necessary to breach measurement scale gaps which are not limited to the traditional geocellular-to-numerical processes but probably now critical at the core/log to geomodelling stage previously overlooked by most reservoir engineers.
Numerical simulation of chemical flooding incorporates different challenges to those observed in miscible and immiscible flooding; flow mechanisms and EOR agent interaction with the reservoir have a higher degree of complexity with advanced numerical formulations (along with resolution requirements) to depict the fluid flow in a finite-difference space. This paper is a continuation of our previous work on miscible and immiscible flooding; we investigate the effect of grid size on the displacement behavior of a polymer flood, in terms of force balances, concentration changes and polymer rheology using different reservoir architectures, and discuss the potential caveats for application of more complex chemical processes and highlight the scale at which such modeling retains the underlying physical character.