Evaluation of Transport Coefficient (Alpha Factors) for Upscaling Polymer Flooding Model

This study explores the adaptation and extension of alpha factors, traditionally applied in gas flooding, to polymer flooding. By leveraging the established methodologies of alpha factor integration in gas flooding systems, the approach aims to offer a potential framework for addressing upscaling challenges in polymer flooding. The methodology was evaluated using a one-dimensional (1D) homogeneous model, employing various upscaling approaches that incorporate pseudo-relative permeability curves and/or transport coefficients. Simulations were performed using the E300 compositional model, a widely recognised tool for modelling miscible gas injection processes. To simulate polymer flooding, the injected fluid (designated as ‘GAS’) and the displaced fluid (‘OIL’) were redefined to represent water and oil, respectively. The properties of the injected fluid were modified to mimic the behaviour of the water phase, enabling accurate representation of polymer flooding dynamics. The PR EOS, traditionally used for carbon dioxide (CO2) gas flooding, was adapted in this study to model the polymer flooding process. Three components were specified for the simulation: the water component represented by ‘H2O’, and the polymer component represented by ‘POLYMER’ and the oil component represented by ‘C15’. Pseudo-curves were derived using the pore-volume weighting average method, while transport coefficients were calculated following the methodology of Li and Durlofsky. Several upscaling combinations were evaluated, including single-phase (SP), two-phase (TP), and alpha factor (ALPHA) approaches. The results demonstrated that the SP_TP_ALPHA model, which integrates pseudo-curves with transport coefficients, provided the most accurate predictions. This model significantly reduced Normal Root Means Square Error (NRMSE) values and closely matched the components production profiles. In addition, the model effectively captured fluid flood front propagations and minimised deviations in recovery factors. These findings emphasise the critical role of combining alpha factors with pseudo-curves to maintain the integrity of the displacement front and optimise flow dynamics in upscaled models. The future work may explore the extension of this methodology to heterogeneous systems and multi-dimensional simulations to further refine its applicability in real-world scenarios.