Utilizing a Design of Experiment Approach in Reservoir Simulation for Pattern Balancing and Optimizing a Polymer Flood

Abstract

Improving the efficiency of polymer flood or enhancing the incremental barrel of oil per pound of polymer injected is one of the most important factors for a successful polymer project. In almost all the active polymer floods, total polymer injection and liquid processing capacity is limited with the original design. Utilizing polymer to its fullest extent is therefore imperative for effective reservoir management in a polymer flood. A full-field polymer flood is made up of several patterns, and consequently polymer flood efficiency depends on pattern efficiency. The scope for optimizing patterns include identifying priority for Optimizing injection/production rates.Optimizing polymer concentration. The key reservoir-level factors that affect a polymer flood are geology, voidage replacement, and in situ polymer viscosity and should be given due consideration in reservoir management. The paper outlines a novel way based on a straightforward simulation model for optimizing polymer flood. Base case reservoir simulation model is used to represent the fulfilled polymer flood, capturing geology and polymer rheology and overall flow behavior. Then, a design of experiment is created with individual pattern's injection rate and concentration as primary variables. The objective function is defined to maximize sweep efficiency, oil recovery and polymer flood efficiency for short term interval. More than 1000 experiments were created and run to identify sensitivity of individual patterns. Further using Sobol and Morris analysis, relative ranking of patterns is done for all patterns with reference to individual and global objective function. To further ascertain impact on ultimate recovery, cases with identified pattern changes have been ran till end of field life and compared with the base case.