Effects of Spreading and Nonspreading Oils on Foam Propagation Through Porous Media

Abstract

Summary. Experiments show that a spreading oil increases the time for foam generation and decreases the speed of foam propagation in a porous medium. It also breaks a foam faster than a nonspreading oil. These findings may be important in interpreting results of different foam displacement experiments and therefore surfactant selection. Introduction The use of foam for mobility control has become a subject of active research and field tests in the oil industry in the past several years. However, many mechanisms governing the flow of foam through porous media still need to be understood. One of these is the effect of oil on foam flow. That oil has a detrimental effect on foam flow through porous media has been recognized for a long time. We will mention several of the many reasons for this here. First, oil can scavenge surfactant from the gas/water interfaces because of the solubility of surfactant in oil and thus destabilizes the foam. This is especially important when the aqueous phase contains divalent cations, which can cause partitioning of surfactant into the oil. Second, the formation of an oil/water macroemulsion can also deplete surfactant from the gas/water interface. This is especially important if a fine-textured macroemulsion is formed that contains many oil/water interfaces for the adsorption of surfactant. Third, polar components in the oil may adsorb at the gas/water interface. And if these oil components have poorer foam-stabilizing properties than the surfactant, they will destabilize the foam. Fourth, the spreading of oil at the gas/water interface can reduce the local surface tension, thereby causing thinning and the rupture of bubble lamellae by the Marangoni effect. In this paper, we present results of experiments designed to study the effect of spreading and nonspreading oils on the macroscopic flow behavior of foam through porous media. The oil/surfactant systems were chosen so that other foam-debilitating mechanisms were absent or believed to be insignificant. Experimental Choice of Surfactant/Oil Systems. The surfactant used in our study was 0.5 wt% Siponate DS-10(TM) at 1.2 wt% NaCl. Siponate DS-10 is a commercial branched-side-chain dodecylbenzene sodium sulfonate. Two nonpolar oils were chosen that had the same viscosity but different spreading characteristics: hexadecane and a 30/70 mixture of Nujol(TM) mineral oil and Shell-Sol(TM) 71 saturated hydrocarbon solvent (Table 1). Interfacial tension (IFT) measurements at room temperature showed that the coefficient of spreading on the surfactant/air interface is negative for hexadecane and positive for the Nujol/Shell-Sol mixture, indicating that hexadecane is nonspreading and Nujol/Shell-Sol is spreading (Table 2). These spreading behaviors were confirmed by visual observations when small drops of the two oils were placed on the air/liquid interface of the surfactant solutions contained in shallow dishes. Hexadecane remained as droplets on the air/liquid interface. whereas Nujol/ Shell-Sol quickly spread into a very thin film. Results of partitioning experiments showed that at room temperature, Siponate DS-10 does not partition into either of the oils used in our studies. We also performed experiments where we put equal volumes of surfactant and oil in a separatory funnel, shook the funnel vigorously to form an oil/water macroemulsion, and then monitored the surfactant concentration in the aqueous phase. Results showed that for both oils. the surfactant concentration in the aqueous phase came back at 97 % of the initial value after 2 minutes. Thus, for practical purposes, depletion of surfactant from the aqueous phase as a purposes, depletion of surfactant from the aqueous phase as a result of the formation of a macroemulsion with either oil is negligible. From the observations presented above, it can be concluded that the spreading behavior of the two oils was the dominant factor controlling foam flow in our experiments. Foam Displacement Experiments. Fig. 1 is a schematic of the apparatus, which consisted of an unconsolidated flint-shot silica sandpack (20 × 3.94 × 0.78 in. [51 × 10 × 2 cm], porosity = 34 %, and permeability = 110 darcies) with transparent walls that made permeability = 110 darcies) with transparent walls that made possible visualization of the flow process. Pressure gauges mounted possible visualization of the flow process. Pressure gauges mounted on the sandpack measured the injection pressure and the internal pressures at-five locations in the sandpack. pressures at-five locations in the sandpack. In the beginning of each experiment, the sandpack was saturated with oil and interstitial water so that So=89% and Sw= 11 %. Foam components consisting of nitrogen and surfactant solution were injected into the sandpack through a common flowline connected to the middle port of the inflow sandface. All experiments used nitro-gen/surfactant flow rate ratio of 5 to 1 cm3/min. A flow rate of 1.5 cm3/min corresponded to an interstitial velocity of 11 ft/D [3.4 m/d] when only a single phase was flowing linearly through the sandpack. The nitrogen flow rate was measured at standard temperature and pressure. Still and time-lapse photography recorded the flow behavior. The volumes of liquid produced and injected were measured; from their difference, the average gas saturation in the sandpack was calculated. Produced liquid was collected in fractions. The oil/water emulsion was broken by heating the fractions in an oven and centrifuging. The average oil and water saturations in the sandpack were obtained by material balance. Results We performed two foam displacement experiments: one with the sandpack initially saturated with hexadecane and the other with Nujol/Shell-Sol (So=89% and Sw =11 %). Figs. 2 through 6 show the average saturation histories and pressure data. From these figures and from foam propagation visualizations, we can make certain observations.. In both experiments, a foam was formed. In the case of a spreading oil, however, it took a longer time to form a foam bank filling the whole vertical interval at the inflow sandface. 2. Once formed, the foam traveled at constant speeds across the sandpacks in a piston-like fashion. SPERE P. 893