Affiliation:
1. Daqing Petroleum Institute
2. Daqing Oil Company
Abstract
Abstract
The Upper-Convected Maxwell (UCM)model has been normally used to describe the rheological properties of polymer solutions. However, this model considers only viscosity and elasticity of a fluid without including the non-Newtonian power law behavior, a very important property to polymer solutions in reservoirs. Therefore, we present a Modified Upper-Convected Maxwell (MUCM)in this paper to take into account this property.
The steady flow mathematical model of MUCM fluids is established. The flow characteristics of viscoelastic fluids in the dead ends of porous pores in reservoirs are simulated numerically. The contours of velocity and stream function are drawn and micro-scale sweep efficiency is calculated quantitatively. The various regularities, corresponding to the change in visco-elastic properties of the fluid, are studied.
Numerical results show the visco-elasticity of polymer solutions is the main factor influencing sweep efficiency. With increasing elasticity, the flowing area in the dead end of pores is enlarged significantly, thus the area with immobile zones becomes smaller. Flow velocity is much larger than that for a Newtonian fluid, the sweep area and displacement efficiency increase as theelasticityincreases. The visco-elastic nature of the displacing polymer fluids can in general improve the displacement efficiency in pores compared to using Newtonian fluids. This conclusion should be useful in selectingpolymer fluids and designing polymer flooding operations.
Introduction
In the process of water flooding alone, the residual oil remaining within porous media is difficult to be displaced or recovered. In comparison, polymer flooding, because of its favorable ratio in mobility between displacing and displaced fluids, is more effective. Experimental results indicated the visco-elasticity of polymer solutions can enhance the displacement efficiency, but there are few theoretical studies on this subject. Polymer flooding is one of the most important chemical flooding processes. It has long been believed that one mechanism of polymer flooding in enhancing oil recovery results from the viscolesticy of the polymer solutions. The viscoelastic fluid can improve the displacement efficiency, which has been proved byexperiments1,2. The viscoelastic HPAM (Hydrolyzed Polyacrylamide) solutions (concentration in series)and viscous glycerin solutions (no elasticity, viscosity in series) are studied by laboratory flooding at visual microscopic glass models and artificial cores3. The experimental results indicate that both the viscosity and the first normal stress differential of HPAM solution increase with the increase in viscoelasticity, and that the displacement efficiency is mainly related to the first normal stress differential of the fluid.
For the past several years, numerical simulation of viscoelastic flows has been carried out to understand the viscoelastic fluid flow behavior in a variety of processes of both industrial and scientific interest. Polymeric fluids, owing to their viscoelastic character, are of particular interest to the numerical simulation community because of their wide applications in material processing and their different behavior from that of Newtonian fluids in ways which are often complex and strikingly different. There are many forms of residual oil in the reservoir, for example, oil droplet, oil film, oil trapped in dead ends, and so on. The configurations of residual oil existing within actual pores may be in many different forms and more complicated. In theoretical studies, for simplicity, these micro-pores are often simplified, for example, assuming an abrupt axisymmetric expansion, an abrupt axisymmetric contraction, micropore with dead ends, etc. Many numerical studies have focused on the viscoelastic flow through an abrupt axisymmetric expansion4 and contraction5,6. Among them, the expansion model is a very typical model for studying the viscoelastic flows. To investigate micropores with dead ends a difference method is used in this paper. The model is easily implemented when the study area is regular. So the difference method is a simple and available method for solving mathematical and physical model in this paper.