An Experimental Study of Waterflooding From a Two- Dimensional Layered Sand Model

Abstract

Summary The effect of flow rate on oil recovery by waterflooding is studied with a two-dimensional (2D), layered sand model that allows visual observation. The model consists of three communicating layers of equal thickness of water-wet sand with the permeability ratio of 2 : 4 : 1 from top to bottom. Three white mineral oils of viscosity 15. 30, and 150 cp [ 15, 30, and 150 mPa.s] have been used for immiscible liquid/liquid displacement under constant pressure drops covering a range of flow rates. The changes in flow regimes in the various layers are observed with respect to the variations in pressure drop across the model. The effects of capillary imbibition, gravity segregation, and viscous pressure gradient are observed and an attempt is made to calculate the instantaneous crossflow. Crossflow of oil from the tight layers to the most-permeable layer increases the intermediate oil recovery for a given volume of water injected. Oil recovery increases somewhat with a decrease in flow rate and considerably with a decrease in oil viscosity. Slow flow rates produce higher recovery but may not be feasible in the field for economic reasons. For M less than 1, viscous crossflow contributes very slightly to oil recovery until the time of breakthrough of water from the most permeable layer. After that. the viscous crossflow increases and creates a secondary oil stream close to the interfaces of the most permeable layer. The additional oil is produced until the least-permeable layer also breaks through. A sequence of slow flow rate followed by a fast flow rate is shown to improve oil recovery for a given volume of water injected, and this occurs within a reasonable experimental time frame. The slow flow rate is set from a criterion of stable displacement in the most-permeable layer based on stability analysis, and the fast flow rate is based on the maximum pressure the model can take. The effect of variation in interfacial tension (IFT) on the displacement behavior was also studied. IFT is reduced from 35 to 4 and 0.5 dynes/cm [35 to 4 and 0.5 mn/mi, respectively, by addition of two surfactants to the injected flood water. These IFT ranges are not sufficiently low to increase oil recovery from interfacial effects. The two surfactants influenced the recovery behavior differently. Introduction In multiphase flow, invariably more than one fluid is present in the porous medium. One of these fluids wets the rock matrix preferentially, which results in a capillary pressure difference preferentially, which results in a capillary pressure difference between the fluids. Also, the difference in the density of the fluids involved causes gravity segregation. Any displacement study representing reservoir conditions needs to consider the interplay of all the forces involved: applied pressure drop (viscous pressure), capillary pressure. and gravity segregation. For a given displacement system, the relative effects of viscous pressure, capillary imbibition, and gravity segregation on internal pressure, capillary imbibition, and gravity segregation on internal flow behavior apparently depend on the flow rate. If the flow rate is very high, the displacement is characterized by a viscous flow regime (with viscous fingers in the case of unfavorable mobility ratio). At slower flow rates, gravity underride may occur, indicating a gravity-dominated flow regime. If the flow rate is reduced further, a capillary flow regime may develop as a result of the capillary imbibition. In a single-layer porous medium, only one type of flow regime occurs for a specific flow rate. However, in a layered porous medium with crossflow, a single flow rate may produce different flow regimes in various layers. produce different flow regimes in various layers. During early water injection, the front moves faster in the most permeable layers. The different front positions create a saturation permeable layers. The different front positions create a saturation discontinuity in the transverse direction, which results in the imbibition of water from high-permeability to low-permeability layers. Capillary imbibition in the transverse direction is called capillary crossflow. In addition, if the mobility ratio is not unity, different pressure gradients in various layers result in different viscous pressure gradients in various layers result in different viscous pressures across the ]layers. This creates a transverse flow that is pressures across the ]layers. This creates a transverse flow that is called the viscous crossflow. Gravity segregation causes gravity crossflow across the interfaces. Gravity crossflow is uniform along the interface but both capillary and viscous crossflows are maximums in the vicinity of the leading front, for this is where the greatest differences occur in saturation and pressure. The object of this study is to investigate the effect of the variations in flow rate, oil viscosity, and IFT on recovery and internal flow behavior of a waterflood in a 2D, layered sand model with vertical communication. A visual model showing the vertical cross section of the reservoir was selected to observe the flow regimes, It includes the gravity effects. The number of layers and the permeability ordering were carefully selected so that the effect of vis permeability ordering were carefully selected so that the effect of vis cous, gravity, and capillary crossflows could be distinctly observed. Putting the most-permeable layer in the middle allows observation Putting the most-permeable layer in the middle allows observation of most of the mechanisms occurring in natural reservoirs. The experiment was performed at a low pressure and at ambient temperature.