Razieh Khosravi

Low-salinity polymer flooding (LSPF) is a promising technique for enhancing oil recovery in heterogeneous heavy oil reservoirs. This study presents a novel upscaling methodology that combines particle swarm optimization with a flow simulator to address challenges related to reservoir heterogeneity and wettability effects. The methodology was validated using a pilot-scale analytical model that integrates the Buckley-Leverett and Koval theories. Employing a particle swarm optimization algorithm (population size: 30), convergence was achieved in 33 iterations. History matching yielded estimates of 60 (μg/g rock) for maximum polymer adsorption and 360 for the Langmuir constant. The pseudo relative permeability curves needed to be more favorable than the rock curves due to decreased heterogeneity. Three dimensional LSPF exhibited a heightened sensitivity to injectivity loss compared to conventional polymer flooding, primarily due to the increased viscosity associated with lower salinity, which contributes to elevated injection pressures. Tracking oil saturation and polymer concentration over time revealed a slower change in polymer concentration compared to oil saturation in pilot area. LSPF (4000 ppm salinity and 1500 ppm polymer) provided an incremental oil recovery as much as 12 % on top of polymer injection under the condition of this study. The proposed upscaling methodology could provide broad applications for predicting and optimizing water-based oil recovery processes at pilot scale.