Two-dimensional polymer flood simulation

Abstract

The present experimental and theoretical investigation was concerned with the oil recovery and sweep efficiency obtained in a two-dimensional polymer flood. Both single phase (brine displaced by polymer solution) and two-phase (oil and brine displaced by polymer) displacement studies were conducted. The mathematical models developed utilize the concept of the polymer resistance factor, and its variation with throughput past any given point in the flow pattern. The models were validated on the basis of good agreement observed between the predicted and the experimental results. Experimental displacement studies were carried out in both homogeneous and heterogeneous two-dimensional transparent models, representing the quadrant of a five-spot pattern. The models were packed with glass beads of different sizes in a predetermined manner. Prior to a polymer flood, the porous medium was saturated with brine, or with brine and a refined oil, depending on the type of run. Next a polymer solution, containing a dye, was injected into the model. Effluent samples were collected, and the flood fronts were photographed at regular intervals. For the purpose of simulation, resistance factors under two-phase flow conditions were determined, and it was shown that the average mobility of the displacing phase across a linear element of the porous medium used can be calculated’ as a function of throughput across this element. This technique was generalized for areal elements, and was incorporated into the mathematical models. The single phase, two-dimensional model considered a flow system in which the mobility of the displacing liquid was changing with time behind the flood front.. The pressure distribution in the flow system at each time step was obtained, and was used to compute the sweep patterns for comparison with the experimentally determined patterns. The two-phase, two-dimensional model involved the simultaneous solution of the two-phase flow differential equations, incorporating the resistance factor concept in the solution scheme. An implicit-explicit technique was employed for solving the finite difference equations obtained. Good agreement was observed between the predicted and the experimental polymer floods. Under the experimental conditions involved the improvement due to the use of polymer was approximately 10% for oil recovery, and 5 to 15% for sweep efficiency. A number of computer runs conducted showed that for permeability contrasts greater than those studied experimentally, a polymer flood would be still more effective.