Abstract
This paper describes an enhanced oil recovery (EOR) model involving use of chemical additives. The model is as general as possible in light of present phenomenological knowledge. It takes into consideration diffusion, exchanges between fluid phases. exchanges between fluids and solids, and convection (including gravity, relative permeabilities, viscosities, interfacial tensions (IFT's), and mobility reduction by polymers).
Its main properties arethree-phase flow (aqueous, oleic, and microemulsion);that each phase can be represented by a mixture of seven components:overall flow is one dimensional (ID),
but the sections of space considered may be constant or variable; and the porous medium may be heterogeneous.
In addition to its compositional nature. which is required for describing changes in phase properties, the model must also account for ionic environment, loss of chemicals, and capillary number distribution. For these reasons, it differs from conventional multicomponent models.
It is designed for simulating any EOR problem involving the addition of suitable chemicals to fluids to be injected continuously or in successive slugs. The model has been tested successfully by two-phase flow, four components (water, oil, polymer, and surfactant plus alcohol), and Winsor Type I environment experiments. It can be adapted to more and more complex phenomenological situations, depending on availability of data.
Introduction
Of the new techniques leading to increased oil production, EOR using chemical additives is one of the most promising. It involves adding sufficient amounts of chemical species to injected fluids to change phase properties, thereby enhancing oil flow. These changes occur in a highly complex way because of the superposition of several mass-transfer phenomena. The difficulties are increased because some of these phenomena, such as exchange between fluid phases and between fluids and solids, are not yet fully understood.
A great deal of experimental research is being done to clarify this situation. However, no matter what experimental progress is made, it will never be easy to analyze all the mechanisms that occur simultaneously and to predict the behavior of the recovery process by pure reasoning or by intuition. A mathematical approach becomes necessary, and the first numerical studies of the problem can be found in the literature.
Our contribution to this effort is described in this paper. It concerns developing a numerical model that is as general as possible within the scope of current phenomenological knowledge. The model is necessarily multicomponent and multiphase. However, it differs from ordinary compositional models by the number of new problems to be investigated. The principal problems include suitable ionic environment, loss of chemical additives, and capillary number distribution in space and time.
SPEJ
P. 461^
Numerical Study of Cooling by Ferrofluids in an Electrical Transformer using an Axisymmetric Model