Summary Foam performance was evaluated by several experimental methods (corefloods, bulk foam stability, micromodel observations, interfacial parameters) using three commercial surfactants, either by themselves or mixed with a fluorinated surfactant. The presence of oil reduced foam mobility reduction factors (Fmr) to different degrees; excellent Fmr were still attained with some surfactants in the presence of residual oil. Using the fluorinated surfactant as additive enhanced the oil tolerance of some, but not all, foams. The tested foams exhibited oil transporting properties. Each foam adjusted to its own residual oil saturation and corresponding level of mobility reduction. Foam performance in corefloods did not correlate in general with predictions based on the other experimental methods. Introduction Foams have been considered for mobility control in solvent, gas, or vapor injection improved oil recovery (IOR) processes, for blocking and diverting using either conventional or gelled foams, and for gas/oil ratio control at production wells. In a diverse range of applications, a foam encounters a range of oil saturations, which necessitates designing a foam with the required stability to oil. In applications where it is desirable to place a foam into swept (low oil saturation) zones, a foam with intermediate or low stability to oil may be adequate. If a foam is to be used as an oil displacing fluid or for gas/oil ratio control in producing wells, foam stability to oil is essential. Coreflood experiments by different investigators suggest that oil becomes detrimental to foam at oil saturations above 5 to 20%. 1 Among a number of mechanisms of foam/oil interaction suggested in the literature,1–5 three main models have emerged in attempts to predict foam stability to oil: spreading and entering coefficients, lamella number, and pseudoemulsion film models. The classical spreading and entering coefficients, based on interfacial tensions measured with bulk liquids, have been used with some success,6,7 but do not correlate with foam stability to oil in general.2-5,8,9 A geometry-dependent spreading coefficient,10 a "generalized entering coefficient,"4 and a "film excess energy,"5 have been defined in order to take into account thin-film forces important in foam/oil encounters in porous media. The lamella number attempts to quantify the observation that oil can become emulsified and imbibed into foam lamellae, which tends to destabilize a foam to various degrees.2,11,12 Pseudoemulsion film models state that a foam can only be stable in the presence of oil if the oil is wetted by the aqueous phase, i.e., if oil and gas phases remain separated by a film of aqueous phase (the pseudoemulsion film).3-5,13 Although different models have been successfully applied to different situations, translating the fundamental mechanisms of foam/oil interaction into generally applicable rules for field application remains difficult. The objective of this work was to evaluate the performance of six foams in the absence and presence of crude oil using different experimental techniques: corefloods in Berea and North Sea reservoir sandstone, bulk foam heights in a blender and in a high-pressure cell, lamella breakage frequency in an etched-glass micromodel, and interfacial parameters. The purpose of the corefloods was to screen a series of promising surfactant candidates for the application of foam in the North Sea. A large amount of experimental 14 and simulation15 work was simultaneously carried out by Norsk Hydro, and has led to a successful field test.16 The other experimental techniques have been used by others as screening tools or to evaluate foam/oil interactions. They were used in this work to study possible correlations with coreflood performance. Some fluorinated surfactants form foams that are very stable in the presence of oil.2,4,8,17 They are, however, more costly than hydrocarbon surfactants by one to two orders of magnitude. In this study, a fluorinated surfactant was used as an additive at relatively low concentration to improve the oil tolerance of three conventional surfactants. Experiment Materials. Cores were either Berea sandstone (length 30 cm, diameter 3.8 cm, porosity 23%, and absolute permeability to air 940 to 1,200 md), or reservoir sandstone (length 17 cm, diameter 3.7 cm, porosity 26%, and absolute permeability to air 3,400 to 3,900 md) from the Oseberg field (North Sea), supplied by Norsk Hydro. The reservoir cores were extracted in a chloroform/methanol mixture and dried before use. Four commercial surfactants were used: Chaser GR-1080 (Chaser International, proprietary blend containing mostly alpha olefin sulfonates), Enordet X-2001 (Shell Chemical Company, alcohol ethoxyglycerylsulfonate), Dow XSS-84321.05 (Dow Chemical, mixture of C10 diphenyletherdisulfonate and C 14-16 alpha olefin sulfonate), and Fluorad FC-751 (3M Company, fluoroalkylsulfobetaine). They will be referred to as Chaser, Enordet, Dow, and Fluorad in this paper. Cited concentrations are active concentrations in %w/v in sea water. The three hydrocarbon-based surfactants (Chaser, Enordet, and Dow) were used either by themselves or mixed with Fluorad (1:9 by mass Fluorad to Chaser, Enordet, or Dow). The brine was filtered (0.45 ?m) synthetic sea water (density 1.004 g/cm 3 and viscosity 0.38 mPa's, at 75°C and 13.8 MPa), and the gas was methane (CP grade, 99 vol% and density 0.0835 g/cm3 viscosity 0.0160 mPa's, at 75°C and 13.8 MPa). Crude oil (Oseberg Field, North Sea) was supplied by Norsk Hydro and cleaned by centrifugation and filtration (0.22 ?m). The viscosity and density of the dead oil at 23°C and ambient pressure were 9.5 mPa's and 0.87 g/cm3 respectively. Methane-saturated oil had a gas/oil ratio of 70 and a density of 0.75 g/cm 3 at 75°C and 13.8 MPa. Corefloods. The core was contained in a stainless-steel core holder within a lead sleeve to which confining pressure (24 MPa) was applied. Liquids were injected by displacement from floating piston vessels using HPLC pumps. Methane was supplied either from a cylinder, its flow rate being controlled by a mass flow controller, or from a Ruska pump. System pressure was controlled by a gas dome-type backpressure regulator. Pressure drops across the whole core and across the outlet half were monitored by differential pressure transducers. The half-core pressure drops were consistently about 70% of the full-core pressure drops, possibly because of capillary end effects or changing flow rates and foam qualities caused by gas expansion during flow through the core. Materials. Cores were either Berea sandstone (length 30 cm, diameter 3.8 cm, porosity 23%, and absolute permeability to air 940 to 1,200 md), or reservoir sandstone (length 17 cm, diameter 3.7 cm, porosity 26%, and absolute permeability to air 3,400 to 3,900 md) from the Oseberg field (North Sea), supplied by Norsk Hydro. The reservoir cores were extracted in a chloroform/methanol mixture and dried before use. Four commercial surfactants were used: Chaser GR-1080 (Chaser International, proprietary blend containing mostly alpha olefin sulfonates), Enordet X-2001 (Shell Chemical Company, alcohol ethoxyglycerylsulfonate), Dow XSS-84321.05 (Dow Chemical, mixture of C10 diphenyletherdisulfonate and C 14-16 alpha olefin sulfonate), and Fluorad FC-751 (3M Company, fluoroalkylsulfobetaine). They will be referred to as Chaser, Enordet, Dow, and Fluorad in this paper. Cited concentrations are active concentrations in %w/v in sea water. The three hydrocarbon-based surfactants (Chaser, Enordet, and Dow) were used either by themselves or mixed with Fluorad (1:9 by mass Fluorad to Chaser, Enordet, or Dow). The brine was filtered (0.45 ?m) synthetic sea water (density 1.004 g/cm 3 and viscosity 0.38 mPa's, at 75°C and 13.8 MPa), and the gas was methane (CP grade, 99 vol% and density 0.0835 g/cm3 viscosity 0.0160 mPa's, at 75°C and 13.8 MPa). Crude oil (Oseberg Field, North Sea) was supplied by Norsk Hydro and cleaned by centrifugation and filtration (0.22 ?m). The viscosity and density of the dead oil at 23°C and ambient pressure were 9.5 mPa's and 0.87 g/cm3 respectively. Methane-saturated oil had a gas/oil ratio of 70 and a density of 0.75 g/cm 3 at 75°C and 13.8 MPa.

No abstract

Summary Adsorption levels of two surfactants (anionic and amphoteric) suitable for foam flooding in high-salinity and high-hardness conditions were measured on sandstone and limestone with different brines. The anionic surfactant adsorbed more strongly on limestone than on sandstone, while the opposite behavior was observed with the amphoteric surfactant. The presence of divalent ions increased adsorption levels of both surfactants on both rocks. Adsorption mechanisms are suggested from measured surface charge properties of the rocks. Introduction In recent years, interest in surfactant-based EOR processes has focused on the application of mobility-control foams. Propagation of a foam depends, in part, on propagation of the surfactant, which in turn is strongly affected by adsorption losses at the solid/liquid interface. While adsorption of EOR surfactants on sandstones has been measured quite extensively, only limited information dealing with adsorption on carbonates is available in the petroleum literature. A survey shows not only that- the majority of EOR projects in Canada and the U.S. use gas or steam injection and thus are potential candidates for the application of foams, but also that a significant number of these projects (63% in Canada, 16% in the U.S.) are being conducted in carbonate reservoirs. The purpose of this work was to measure adsorption on both sandstone and limestone for two surfactants found to form effective mobility-control foams in extremely high-salinity conditions. One surfactant is anionic and the other amphoteric. Adsorption of the latter surfactant type has rarely been measured. Several mechanisms may be responsible for surfactant adsorption, including electrostatic interaction, van der Waals interaction between hydrocarbon chains of adsorbed surfactant molecules (hemimicelle formation), covalent bonding, nonpolar chain-solid interactions, hydrogen bonding, and salvation and desolvation of adsorbate and adsorbent species. Electrostatic and van der Waals forces are believed to be dominant in systems containing ionic surfactants and oxide surfaces (e.g., silicate minerals). Adsorption on such salt-like minerals as calcite and dolomite is a more complex process and is much less understood than adsorption on oxides because of the complexity of the dissolution products and the interfacial region that exists when these minerals products and the interfacial region that exists when these minerals are immersed in an aqueous medium. Additional surfactant-adsorption mechanisms suggested for these minerals include salt formation at the solid surface by surfactant binding to lattice ions, ion exchange of surfactant with lattice ions, and surface precipitation. Adsorption of EOR surfactants generally increases with increasing salt concentration and increasing divalent ion content. Some systems, however, exhibit only a small dependence of adsorption on salt concentration, and some studies show that inorganic monovalent ions may lower surfactant adsorption by competing with the surfactant for the surface or by shielding the surface charge on the solid. An increase in salt concentration may increase surfactant adsorption by decreasing the solvent power of the aqueous phase for the surfactant, thus driving the surfactant to the interface, or by decreasing headgroup/headgroup electrostatic repulsion in the adsorbed layer. Increasing the electrolyte concentration compresses the electrical double layer, the resulting change in adsorption depending on the sign of the charges of the solid surface and the surfactant. While monovalent inorganic ions change the magnitude of the solid's surface charge by compression of the electrical double layer, multivalent ions may specifically adsorb to a surface of opposite charge and reverse the sign of the surface charge, thereby affecting adsorption of ionic surfactants. As a step toward identifying adsorption mechanisms in the systems of interest, the surface-charge properties of the rocks used in the adsorption experiments (Berea sandstone and Indiana limestone) were determined through measurement of electrophoretic mobilities of dilute suspensions of rock particles in different brines. A review of literature that deals with the surface charge of the solids of interest in this work appears in Ref. 25. The isoelectric points (IEP's) of quartz and kaolinite (the principal components of Berea sandstone) in water or dilute electrolyte have been found to be at a pH from 1.5 to 3.7 and 4 or less, respectively. Quartz and kaolinite are thus negatively charged at neutral pH. Because of the complex dissolution behavior of calcite (the main component of Indiana limestone), much more diverging results have been obtained with this mineral. IEP's between pH 4 and 11 have been reported. Some workers found calcite to carry a positive surface charge over the complete pH range investigated, while others found it to be negatively charged. The addition of divalent cations to the aqueous medium changes the surface charge of both quart and calcite toward less negative or positive values. Surfactant adsorption is commonly measured either through static or dynamic (coreflood) experiments. In this work, we chose the latter method because it effectively allows a larger solid/liquid ratio to be used and thus results in greater sensitivity. Effluent profiles obtained from corefloods are matched by a numerical model. Details of the adsorption models are given elsewhere. The model used in this work combines the Gibbs surface excess concept with the pseudophase separation model of surfactant solutions. Surfactant adsorption is driven by the surfactant monomer concentration rather than by the total surfactant concentration. With the assumption that the commercial surfactants used in this work can be approximated as single-component surfactants, the surfactant monomer concentration is equal to the total surfactant concentration below the critical micelle concentration (CMC) but remains constant at the CMC value at total surfactant concentrations above the CMC. The parameters in the adsorption model are the selectivity, S; monolayer coverages of surfactant, m, and water, m; rate constants of adsorption, k, and desorption, k; and a dispersion parameter, . Fitting an experimentally determined effluent concentration profile results in a set of best-fit parameters that allows calculation of an adsorption isotherm through the following equation: (1) where n is the surface excess adsorption of surfactant and x are the mole fractions of surfactant (i=1) and water (i=2) in the monomer (dispersed) phase. The monolayer coverages are estimated from the molecular area of the surfactant molecules, determined from surface tension measurements and the Gibbs adsorption equation, and the specific surface area of the solids. The molecular area of water can be taken to be 0. 125 nm (Ref. 50).

No abstract

No abstract