Wettability Literature Survey Part 5: The Effects of Wettability on Relative Permeability

Publisher Website
Google Scholar
Abstract
Summary.: The wettability of a core will strongly affect its waterflood behavior and relative permeability. Wettability affects relative permeability because it is a major factor in the control of the location, flow, and distribution of fluids in a porous medium. In unfamiliar or fractionally wetted porous media, the water relative permeability increases and the oil relative permeability decreases as the system becomes more oil- wet. In a mixed-wettability system, the continuous oil-wet paths in the larger pores alter the relative permeability curves and allow the system to be waterflooded to a very low residual oil saturation (ROS) after the injection of many PV's of water. The most accurate relative permeability measurements are made on native-state core, where the reservoir wettability is preserved. Serious errors can result when measurements are made on cores with altered wettability, such as cleaned core or core contaminated with drilling-mud surfactants. Introduction: This paper is the fifth in a series of literature surveys covering the effects of wettability on core analysis. Wettability has been shown to affect waterflood behavior, relative permeability, capillary pressure, irreducible water saturation (IWS), ROS, dispersion, simulated tertiary recovery, and electrical properties. Earlier, but less complete, reviews covering the effects of wettability on waterflooding and relative permeability can be found in Refs. 6 through 16. Relative permeability is "a direct measure of the ability of the porous system to conduct one fluid when one or more-fluids are present. These flow properties are the composite effect of pore geometry, wettability, fluid distribution, and saturation history." Wettability affects relative permeability because it is a major factor in the control of the location, flow, and spatial distribution of fluids in the core. Craigs and Raza et al. have given good summaries of the effects of wettability on the distribution of oil and water in a core. Most experimental studies that examined fluid dis-tribution as a function of wettability used bead packs or othe micromodels, although some more recent studies have used reservoir rock and fluids such as epoxy or Wood's metal that can be solidified in situ (e.g., see Yadav et al.). Consider a strongly water-wet rock initially at IWS. Water, the wetting phase, will occupy the small pores and form a thin film over all the rock surfaces. Oil, the nonwetting phase, will occupy the centers of the larger pores. This fluid distribution occurs because it is the most energetically favorable. Any oil placed in the small pores would be displaced into the center of the large pores by spontaneous water imbibition, because this would lower the energy of the system. During a waterflood of a water-wet system, water moves through the porous medium in a fairly uniform front. The injected water will tend to imbibe into any small- or medium-sized pores, moving oil into the large pores where it is easily displaced. Only oil is moving ahead of the front. In the frontal zone, each fluid moves through its own network of pores, but with some wetting fluid located in each pore. In this zone, where both oil and water are flowing, a portion of the oil exists in continuous channels with some dead-end branches, while the remainder of the oil is trapped discontinuous globules. Fig. 1a, taken from Raza et al., shows water displacing oil from a water-wet pore. The rock surface is preferentially wetted by the water, so water will advance along the walls of the pore, displacing oil in front of it. At some point, the neck connecting the oil in the pore with the remaining oil will become unstable and snap off, leaving a spherical oil globule trapped in the center of the pore. After the water front passes, almost all the remaining oil is immobile. Because of such immobility in this water-wet case, there is little or no production of oil after water breakthrough. The disconnected, residual oil exists in two basic forms:small, spherical globules in the center of the larger pores, andlarger patches of oil extending over many pores that are completely surrounded by water. In a strongly oil-wet rock, the rock is preferentially in contact with the oil, and the location of the two fluids is reversed from the water-wet case. Oil generally will be found in the small pores and as a thin film on the rock surfaces, while water will be located in the centers of the larger pores. The interstitial water saturation appears to be located as discrete droplets in the centers of the pore spaces in some strongly oil-wet reservoirs. A waterflood in a strongly oil-wet rock is much less efficient than one in a water-wet rock. When the waterflood is started, the water will form continuous channels or fingers through the centers of the larger pores, pushing oil in front of it (see Fig. 1b). Oil is left in the smaller crevices and pores. As water injection continues, water invades the smaller pores to form additional continuous channels, and the WOR of the produced fluids gradually increases. When sufficient water- filled flow channels form to permit nearly unrestricted water flow, oil flow practically ceases. The remaining oil is found (1) filling the smaller pores, (2) as a continuous film over the pore surfaces, and (3) as larger pockets of oil trapped and surrounded by water. Because much of this oil is still continuous through the thin oil films and can be produced at a very slow rate, the ROS is not well-defined. In this paper, the terms "wetting" and "nonwetting" will be used in addition to water-wet and oil-wet. This will more easily enable us to draw conclusions about a system with the opposite wettability. For example, a waterflood in a system of one wettability will behave in the same manner as an oilflood in the same system with the wettabilities reversed. Relative permeability curves will also show that the fluids can exchange positions and flow behavior. Because...

This publication has 0 references indexed in Scilit: