Reducing Surfactant Adsorption in Carbonate Reservoirs
- Tabatabal Ahmadall (U. of Oklahoma) | Marla V. Gonzalez (U. of Oklahoma) | Jeffrey H. Harwell (U. of Oklahoma) | John F. Scamehorn (U. of Oklahoma)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Engineering
- Publication Date
- May 1993
- Document Type
- Journal Paper
- 117 - 122
- 1993. Society of Petroleum Engineers
- 5.4 Enhanced Recovery, 4.1.2 Separation and Treating, 4.1.5 Processing Equipment, 5.8.7 Carbonate Reservoir, 5.4.7 Chemical Flooding Methods (e.g., Polymer, Solvent, Nitrogen, Immiscible CO2, Surfactant, Vapex), 4.2.3 Materials and Corrosion, 5.4.9 Miscible Methods, 2.5.2 Fracturing Materials (Fluids, Proppant)
- 5 in the last 30 days
- 1,189 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 35.00|
This paper compares the adsorption of anionic and cationic surfactants on carbonate minerals. It shows that, in general, cationic surfactants may exhibit significantly less adsorption on carbonate minerals than that exhibited by anionic surfactants. It is also shown that cationic surfactant adsorption on carbonates may be reduced dramatically by the presence of divalent cations, an environment in which the precipitation of anionic surfactants might make their use impractical. It is recommended that cationic surfactants be considered for use in miscible flooding with CO2.
Under suitable conditions, CO2 can be an effective injection fluid for tertiary oil production by miscible flooding. The application of a CO2 flood to a reservoir, however, presents the problem of viscous instability and potentially poor mobility control. At typical reservoir conditions of 1,000 to 3,000 psia and 140°F [6.9 to 20.7 MPa and 60°C], the CO2 has a viscosity of a few hundredths of a centipoise, which is a hundred times less than the viscosity of the oil to be displaced.1 This difference in viscosities results in an unstable displacement front because CO2 fingers form in the oil phase, which decreases the efficiency of the process.
Different mobility control methods have been proposed to reduce this problem.2,3 Besides water-alternating-gas (WAG) processes and direct thickening with polymers, dispersion of CO2 in water with a surfactant has been considered an alternative to reduce CO2 mobility. The foam formed by the surfactant consists of gas macroscopically dispersed in brine and has a viscosity several orders of magnitude greater than the viscosity of either phase in the dispersion, even at very low surfactant concentration. The high viscosity of the dispersion stabilizes the displacement front and increases oil recovery. Several investigators1,4 have shown that the use of foam reduces gas mobility for the WAG process and for direct injection of the CO2/water dispersion.
Surfactant screening for application in CO2 flooding processes has been based primarily on their ability to produce stable dispersions of CO2 in brine.1,4,5 Besides having the ability to disperse the supercritical gas, however, the surfactant must survive the environment where it will be used.6,7 Surfactant adsorption on the reservoir minerals presents a potential problem because abstraction of the surfactant from the dispersion will destabilize the dispersion. Adsorption is potentially a greater problem in CO2-foam processes than in low-tension micellar/polymer processes. In micellar flooding, the surfactant-rich middle phase serves as a reservoir of surfactants to stabilize the process against the abstraction of surfactant by adsorption on the reservoir minerals. In CO2 dispersions in brine, however, there is no surfactant-rich third phase. The effect of the loss of surfactant on dispersion stability is magnified by the low concentration of surfactant at which stable dispersion forms. This means that a loss of surfactant as a result of its adsorption can reduce the stability of the CO2 dispersion, which increases gas mobility and thereby decreases the extent of tertiary oil production. Surfactant losses should be minimized, therefore, to increase the stability of the recovery process. Further, because the injected surfactant is a major portion of the overall cost of the project, its loss should be minimized to optimize process economics.
Fig. 1 shows a typical surfactant-adsorption isotherm for an anionic surfactant on a metal-oxide mineral. The adsorption isotherm shows the amount of surfactant adsorbed on the substrate vs. the equilibrium surfactant concentration. The typical isotherm exhibits four identifiable regions.8 Region 1 corresponds to low surface coverage by adsorbed surfactant monomers. In Region 1, also known as the Henry's law region, a linear relationship exists between the surfactant equilibrium concentration and adsorption density.8 The mechanism responsible for the surfactant adsorption is mainly the electrostatic attraction between the charged surface of the solid and the charged head group of the surfactant molecule. Region 2 is characterized by a sharp increase in the adsorption caused by the formation of local monolayer or bilayer aggregates on the surface, called hemimicelles9 or admicelles,10 respectively. Hydrophobic bonding between the surfactant tail groups contributes significantly to the aggregation phenomenon in this region. In Region 3, the forces influencing adsorption are the same as those in Region 2; the Region 2/Region 3 transition is identified by a decrease in the slope of the adsorption isotherm and may be either distinct or gradual. Region 4 begins at the critical micelle concentration (CMC) of the surfactant; in Region 4, micelle formation competes with surfactant adsorption and results in a plateau region where surfactant adsorption becomes nearly constant with increasing surfactant concentration.
While the adsorption of surfactants on minerals has been studied extensively, the vast majority of this work has focused on minerals typical of sandstone reservoirs because most reservoirs that are candidates for micellar flooding are sandstone reservoirs. Because of the high adsorption of cationic and nonionic surfactants that occurs on sandstone minerals under reservoir conditions, cationic surfactants have not been examined to any degree for application in tertiary recovery processes, even in carbonate reservoirs. In contrast, it has been estimated that 60% of total miscible-flooding EOR will occur in carbonate reservoirs.11 There has been very little study of surfactant adsorption on carbonate minerals.12-14
Other studies,15-21 however, have elucidated much about the surface chemistry of carbonate minerals, including effects of pH and lattice ions on the surface charge, which led us to suspect that cationic surfactants could show very low adsorption on carbonates. Calcite and dolomite are carbonate minerals with similar structures.22,23 Calcite is formed by alternate layers of calcium ions and carbonate-ion groups. Dolomite is composed of alternate layers of calcium, magnesium, and carbonate ions. Both solids are salt-type minerals; therefore, the solubility in water is higher than for oxides and silicates.23 The surface charge-on the two carbonate minerals in aqueous systems seems to be generated by the preferential dissolution of lattice ions, either Mg2+, Ca2+, or CO32-. This dissolution process is determined by interactions between the dissolved ions and the solution constituents; complexes formed by the reactions can then be adsorbed again or precipitated on the solid surface.17,19 Although not all work reported in the literature agrees, some studies indicate that hydrogen ions also appear to act as potential determining ions for carbonates,15,19,21 which means that the surface charge on the minerals also depends on the equilibrium pH of the solution. At high pH values, more hydroxyl anions are present in the bulk solution and a net negative surface charge is observed. At low pH in the presence of an excess of hydrogen cations, however, the surface charge is positive. These results suggest that cationic surfactants can exhibit very low adsorptions on carbonates - adsorptions significantly below those exhibited by anionics. This paper presents some results of a recent examination of this hypothesis.
|File Size||1 MB||Number of Pages||6|