Conditions for Generating Foam in Porous Media

Abstract

Abstract The generation of foam in porous media remains a controversial issue. It has been reported that foam is formed only above a minimum flow rate (or capillary number), above a minimum pressure gradient, or below a critical capillary pressure. In this paper we investigated experimentally how these and other parameters affect the formation of N2 and CO2 foams in Berea sandstones. We found that foam was readily formed by co-injecting gas and surfactant solution whenever the core was presaturated with surfactant, regardless of flow rate or pressure gradient. In general, foam generation can be initiated by first creating regions of high surfactant saturation in situ and then draining those regions; the onset of foam generation coincided with the onset of drainage. Somewhat surprisingly, it was easier to generate a stronger foam than a weaker foam, and fewer pore volumes of injection were needed to form foam in a longer core than in a shorter core. A consistent mechanism that emerges from these observations is that foam is formed by raising capillary pressure in a surfactant-saturated region. As gas enters such regions, individual lamellae are formed by snap-off in certain pores. These lamellae must be sufficiently stable to allow the local capillary pressure to increase in time, forcing gas to enter increasingly smaller pores and generate more lamellae. Presaturating the porous medium with surfactant ensures that stable lamellae are formed whenever snap-off occurs. In contrast, if continuous gas channels exist over sample-spanning distances, the injected gas preferentially flows through such channels, capillary pressure cannot rise significantly, and foam generation is delayed. pressure cannot rise significantly, and foam generation is delayed. Practical implications of these finding on applications of foam are Practical implications of these finding on applications of foam are discussed. Introduction Foam is used to improve the volumetric sweep efficiency of gas and steam floods by either reducing gas mobilities in depth or by plugging thief zones near the injector. However, such benefits may never be realized if foam cannot be formed under reservoir conditions. This has been a concern in applying foam in the field. Ransohoff and Radke were first to report that a minimum flow rate was needed to generate foam. They studied homogeneous beadpacks with permeabilities between 40 and 370 darcies. They injected gas or gas permeabilities between 40 and 370 darcies. They injected gas or gas and surfactant into beadpacks saturated with surfactant solution and found that strong foams were formed only above a critical "capillary number", which was defined differently from the usual capillary number, but was still proportional to displacement velocity and inversely proportional to the gas-liquid surface tension. The corresponding proportional to the gas-liquid surface tension. The corresponding critical velocity was not clearly given, but was in the neighborhood of 0.1 cm/sec (300 ft/day). Such high velocities are rarely achievable in the field. Fortunately, the beadpack results may not apply to consolidated porous media or sandpacks. Indeed, many studies reported in the literature, from earlier work: of Holm and co-workers to more recent works, do not indicate the existence of a critical velocity. Those studies that do show a critical velocity often involve high-permeability porous media, very dry foams (foam quality higher than 90%), and a porous medium that had been flooded by gas to low liquid saturations before injecting foam. Rossen and Gauglitz argue that a minimum pressure gradient is needed to generate a "strong" foam. Central to this argument is the assumption that strong foams are flowing foams, which are formed by dividing and multiplying lamellae generated by snap-off or leave-behind. Because a minimum pressure gradient is needed to displace lamellae out of pore throats, it is theorized that a minimum pressure gradient is needed to generate a strong foam. P. 353