Abstract This paper presents results from a series of single phase stable displacement experiments in a heterogeneous core assembly. This system consists of a 0.5 m cylindrical sandstone core with a central core removed and replaced with high permeability ballotini. Displacements at unit, 2:1 and 10:1 viscosity ratio were performed which demonstrated the importance of crossflow mechanisms in polymer flooding processes.
The experiments were modelled using a curvilinear grid in a chemical flood simulator. All of the main features observed in these experiments were successfully reproduced and analysed in terms of viscous and dispersive crossflow effects. Some discussion is presented on the significance of crossflow at the reservoir scale.
Introduction In waterflooding or polymer flooding heterogeneous systems many complex effects may occur. Some of these effects such as fluid crossflow, are associated with the heterogeneous nature of the reservoir. The most common feature in this respect is reservoir layering where the permeability contrast between layers is often large. Fluid flow in stratified systems has been studied by several workers and the importance of fluid crossflow between layers has been emphasised. In a polymer flood in such a system, it has been shown previously that the polymer acts by enhancing viscous crossflow effects thus improving the vertical sweep efficiency.
Experimental work in support of a waterflood or polymer flood in a heterogeneous system is usually performed in short one-dimensional cores. Whereas this may be appropriate to determine some properties of the flood e.g. linear displacement efficiency, polymer adsorption levels, etc, it cannot reproduce the essential features of the recovery mechanism in the heterogeneous system. In order to do this, core assemblies with some degree of known heterogeneity must be used. Most heterogeneous experimental systems reported in the literature which have been used to evaluate EOR fluids have used either radial or linear parallel core assemblies in which the different permeability layers were completely separated e.g.. However, these do not allow fluid to crossflow between layers and are therefore inappropriate to the majority of reservoir systems.
Some experimental studies have appeared in the literature on the core flooding of heterogeneous systems in which crossflow effects have been present. These studies have emphasised various effects such as that of gravity segregation, flow rate, transverse dispersion and slug breakdown. In some cases, the effects of polymer and surfactant have been demonstrated in layered syste. However, one of the key features of a non-unit mobility displacement in such a system is that the local flow rates in the layers and the effluent from these layers are changed. No detailed results on effluent analysis or flow ratios from each separate high/low permeability region have been reported. Nor has a complete computer simulation analysing such results been presented. It is in this respect that the work presented in this paper is novel.
The crossflow effects that may occur in a system either at the reservoir or experimental scale, are associated with viscous, capillary, gravity and dispersion effects. For a complete analysis of either a waterflood or a polymer flood in a heterogeneous system, it is also important to know how these various crossflow effects scale. We have used earlier work on scaling theory and applied it to polymer flooding in layered systems.
P. 341^