A Predictive Model for Water and Polymer Flooding

A "predictive evaluation model" (PEM) has been developed for feasibility analysis of water and polymer flooding. It is designed to produce a reservoir performance prediction suitable for economic analysis, with small computing time and input data requirements. The tools previously available for this purpose range from "binary" screening guides to sophisticated reservoir simulators. Binary screening guides do not consider the composite effect of reservoir parameters, and offer little information about economic feasibility. Many simplified prediction methods are available for waterflooding, and some for special cases of polymer flooding; however, the assumptions inherent in these methods limit applicability. Mathematical reservoir simulators are excellent prediction tools, but operational costs are often prohibitive when screening prospective reservoirs. The PEM was developed to fill the gap between simplified methods and reservoir simulators. The PEM uses "vertical equilibrium" methods to generate pseudorelative permeability curves, which are then used in a one-dimensional finite-difference model; this accounts for vertical heterogeneity and crossflow between communicating layers. Areal sweep correlations for pattern floods are then applied, followed by injection rate calculations. The PEM is based on the assumption of incompressible oil-water flow, but includes a correction for initial gas saturation. The resulting output consists of cumulative produced volumes and producing rates as a function of time for oil, water, and gas, as well as injection rates and volumes. The PEM considers many important flow properties which usually are accounted for only in reservoir simulators and requires a small fraction of the computing time. Polymer solution flow properties accounted for in the PEM include permeability reduction, adsorption, viscous fingering of drive water into polymer slug, and viscosity, all as functions of polymer concentration; the 11 inaccessible pore volume11 effect is also included. Predictions can also be made for tertiary polymer floods initiated after waterflooding. Injection rate calculations account for variations with time due to reservoir flow characteristics; the nonNewtonian behavior of polymer solutions is also considered. Because the PEM is designed for preliminary analysis, where extensive reservoir and fluid data may not be available, it includes routines for estimating relative permeability and capillary pressure curves. Although it is based on a stratified model, it can generate layers of different permeability given the Dykstra-Parsons permeability variation. These features reduce data requirements to a minimum when necessary, but the PEH also accepts more extensive data if it is available. Sensitivity studies were conducted to show the effects of various reservoir and fluid parameters on oil recovery and injection rates, for both water and polymer flooding. The PEM was validated by matches with a cross-sectional polymer flood simulator and other published simulation results; good agreement was observed. History matching of actual field data was successfully performed for a pilot waterflood and a field-scale polymer flood.