Abstract
Characterization of polymer flow behavior, using data from laboratory measurements, is an essential step in the development of predictive reservoir simulation models of polymer enhanced oil recovery (EOR). We describe a methodology for calibrating simulation parameters, using detailed observations from multiphase coreflood experiments and taking account of the core geometry and polymer rheology.
A series of multiphase brine and polymer floods were performed on an oil-filled sandstone core plug, at different concentrations and at different flow rates. Measured data included pressure drop along the core and oil saturation derived from nuclear magnetic resonance. A fine-scale, three-dimensional simulation grid was designed to capture accurately the geometry of the experimental apparatus, including the coreholder platens for fluid injection and extraction. The simulation input parameters were adjusted to match the experimental results of each coreflood, and sensitivity studies were performed to assess the impact of uncertainties.
In the coreflood experiments, different recovery efficiencies were observed, depending on the type of aqueous solution and the injection flow rates. An initial relative permeability model was defined by matching a constant-rate brine flood. Further tests were performed for brine, xanthan, and hydrolyzed polyacrylamide (HPAM) solutions, at incrementally increasing flow rates, and the resulting residual oil saturations were used to define capillary desaturation curves. Experimental data also showed that the apparent polymer solution viscosity at different shear rates differed from the values predicted by conventional rheology. To represent this behavior in the simulator, a new method for estimating the apparent aqueous-phase viscosity during multiphase flow has been developed and validated. At each step in the simulation study, sensitivity studies were used to check the quality of the experimental results, and unexpected behavior was explained or corrected.
High accuracy is required when designing EOR processes. In this project, close collaboration between research scientists and simulation experts led to development of innovative workflows to interpret experimental results, build the simulation grid, and characterize the polymer properties.
