An analysis of physical models and numerical schemes for polymer flooding simulations
Résumé
Chemical Enhanced Oil Recovery processes (CEOR) are nowadays commonly used by engineers to improve the recovery factor of an oil field. To ensure a successful recovery, they are helped by a sophisticated numerical simulator that supposedly leads to the best decision for these complex CEOR processes. In this paper, we propose to investigate the physical and numerical singularities arising in the numerical simulation of chemical enhanced oil recovery technique for oil fields. More specifically, we focus on polymer flooding, which is the most widely used CEOR technique, associated with a black oil model, which is the most basic but realistic multiphase flow system. We assume that the polymer is only transported in the water phase or adsorbed on the rock. The polymer reduces the water phase mobility and can change drastically the behavior of water oil interfaces. Due to its size, the polymer flows faster than water so that an inaccessible pore volume must be added in the model. We propose to review the various physical models for adsorption, mobility reduction and inaccessible pore volume. Then, a mathematical study of the simplified system will be carried out to identify its singularities and their impact on the numerical resolution. Then, considering an IMPES scheme, we propose an approximate CFL (Courant-Friedrichs- Levy) criterion which is required for the numerical stability of the simulation. Numerical polymer flooding experiments are computed with a complete reservoir simulator to illustrate the validity of our approximate CFL criterion: from core flood to 3D real case. A comparison with the fully implicit scheme is performed in order to investigate the numerical diffusion effects.