Wettability: G.J. Hirasaki, SPE, Shell Development Co. Fundamentals and Surface Forces ‘Summary. The wetting of mineral surfaces by water and oil is described by models of surface forces that become important when two surfaces approach each other. Force components are eletrostatc, van der Waals, and structural. The electrostatic force depends ‘on brine pH and salinity, evade oll composition, and the mineral. The surface forces are expressed asa dsjoining pressure isotherm, and its integral is the specific interaction potential isotherm. The specific interaction potetial isotherm can be used to determine the Sable and metastable flm-hickness profiles at substrate, This profile gives the contact angle. ¢ three-phase contact region for a given capillary pressure and/or curvature ofthe Introduction ‘Wettbility has been recognized as an important factor in remsin- {ng oil saturation and in capillary pressure and relative permesbil- ity curves.12 This work describes some of the physics of the contact angle between mineral surfaces, water, and oil. Desrip- tion of a particular system requires a chemical description of the ‘mineral, brine, and ol. The wettability of a rock/brine/oil system. ‘cannot be described by a single contact angle because it isthe mul ttude of contact angles atthe various three-phase contact regions inthe pore spaces that determines system wettability. A complete wettability description requires a morphological description of the ‘pore space with the contact angles as a boundary condition forthe ‘uid distribution. Regardless ofthe morphology, however, the Wet- lability depends on the contact angles. This work focuses on the intermolecular surface forces that affect wetabilty. A tremendous amount of research on tis subject exists outside the petroleum in- dustry that canbe applied to petroleum reservoirs, The tate of the aris reviewed in Refs. 3 through 6. Mohanty” applied these con- cepts to the fluid distribution in petroleum reservoirs, and Hirasaki® elaborated on the development of Mohanty's applied ‘eo. ‘The intermolecular surface force approach originated with Der- §aguin and Landau's and Verwey and Overbeek's (DLVO) theory ‘of colloidal stability. This theory describes the stability and fle: ‘ulation of lyophobic (solvent-fearin) colloids, considering ele- trostatic and van der Waals interactions. This approach has been used I? to deseribe the stability of the wetting-water film. The present work goes beyond the stability ofthe weting-water fm to examine the value of the contact angle when the wetting film collapses, To describe the contact angle, it is necessary to consider ‘other surface forces, collectively called structral forces. Deserip~ tion of wetability in petroleum reservoirs requires the inclusion of capillary pressure and the curvature of the pore walls, ‘This papet reviews the extensive literature on the physics and chemistry of wettability that resides ouside the petroleum litera- ture. The new information describes how fundamental surface forces affect wettability in petroleum reservoirs and the magnitude of ad- vvancing and receding contact angles for ol/water/mineral systems interms of surface forees. This work is intended to provide a guide {or researchers investigating wettability mechanisms, ‘Thermodynamics of Wettabillty “The thermodynamics of wetailty*8.13.1 requires a description of the region where three phases come together at the contact ine, as shown in Fig. 1. When 2 pair of interfaces approach each other ‘ata three-phase contact line, the interfaces interact wit eachother, ‘and the distance separating the intecface affects the system energy. ‘Thus, this distance called thickness—is thermodynamic varia ble. The change in energy per unit area with change in distance 48 the pair of interfaces is brought from a large separation oa finite ‘thickness is expressed as a force per unit area or dsjoining pres- sure, II, The disjoining pressure isthe force that tends to disjoin or separate the two interfaces. A negative disoining pressure at- tracts the two interfaces, Conran 01 Saco Paoum Engrs ‘SPE Formation Evaluation, June 1991 ‘The conditions for equilibrium of a system with a pair of inter- faces are equality of temperature and chemical potetials between the phases and the augmented Young-Laplace equation, pp? =Il+2H=0%7, ves a where 107 =mean curvature, 0% ~intefucial tension (IFT), and —p = Laplace pressure or capillary pressure, P,. In the menis- us region where the separation ofthe interfaces is ange, I'=0 and Eq, I reduces to the usual Young-Laplace equation. If one ofthe bulk phases isa flat solid, the mean curvature ofthe fuid/Cuid in- terface is zero where the interfaces are parallel. Here, P.=II ‘A free-energy function is needed to determine the configuration ‘ofthe equilibrium interface in the three-phase contact region. The thermodynamic potential (or free energy) that is @ minimum for processes at fixed temperature, volume, and chemical and gravi tional potentials is the grand canonical potential, 2. This potential has been called the “interaction potential” in the DLVO theory ‘of colloidal science (defined with zero capillary pressure such that itis zero at infinite separation). 0, however, difers from the DLVO imeraction potential because the disjining pressure at equilibrium san be nonzero if the capillary pressure is nonzero. Here, @ describes the energy of interaction of two phases separated By a third phase as a function ofthe separation distance. The departure ofthe film thickness from the equilibrium thickness results in an increase of this potetil from the minimum value, described by ium | 9H, vere spatial pe nit rea (r specifi iteration potential) {od T= leo quam soning pessre that seis Tand chsh capary preset th sold subst a Fig 2 states the relsionsipbetmen the dining prs an ths pete teraction penal The otberm eqs Ty at hee thicknesses, fo. hy san usable eur om Sopose ta i fim i ake meastabeequloum thckoess, a The pte energy change fom hh Showa by he ‘Sete I the mpatv aren is reat tan he poste are, then wil havea ower ve of Sperifiinracton poeta hat fo agi metastable The energy bare betwen the 0 fo cal equilibria is shown by te peak in the potential at hy and by the postive re inte oper portion of Fig 2. The two cknsies Can coexist ithe poeta values reel “the magne of he minimum of th pce inracton poen- tal eave tif seperation determines the ae of the mc rescopi contact angle, a wl shown tr. When PC the requ vale of pei interaction peal athe minum = the work eqied separate the bk ors. Kean be nerpeted {the work of ahesio inthe presence of thd pase say ‘actor apo) othe sprening coin. Fe wor 0 oress, 8-0. IF ths work eqs the IFT, then 8=90" @ Contact Angle and Film Stability ‘The contact angle*19 isthe angle ofthe macroscopic meniscus ‘when extrapolated to zero thickness (Fig. 3). Young's equation of 217er 7 vob {RNa 4 + ne Sy 9 ROO na Fig. 1—Contact region between two fulds and a solid , and &, are constant, the contact angle, expressed as a function of the surface energies or IFT's, isa statement of mechanical equilibrium. The difference ‘of surface energies in Young's equation equals the difference in the specific interaction potential between the equilibrium thickness and infinite separation (when P.=0). ‘An aliernatve determination ofthe contact angle isto integrate the augmented Young-Laplace equation Eq. 1). 118 Inegra- tion of the mean curvature term gives the cosine ofthe angle of inclination of the microscopic meniscus from the flat solid. One ‘minus the cosine ofthe angle equals the specifi interaction poten tial divided by the fluid/fuid IFT (Fig. 4)- Angle increases or decreases identify the concave and convex repions ofthe mictoscop- je meniscus. Along the macroscopic meniscus, where IT=0, 1~cos ‘isa straight line with a slope equal to the capillary pressure divided by the IFT. This straight-line region can be extrapolated to 2270 thickness to determine the contact angle. 1-c0s o-[-| “ahh fo=—G ° It the extrapolated value of 1~c0s a is negative, then a nonzero ‘contact angle does not exist because the macroscopic meniscus be- comes tangent to the slid. Ifthe extrapolated value is positive, then a nonzero contact angle exists. Ith extrapolated value equals 1.0, then 6=90°. A thin film ofthe original film phase, however, may stil exist—e.g., water of hydration on an oil-wet mineral surface. ‘The system in Fig. 4 has two different locally stable film tick nesses, hy and hy. hy has a finite contact angle, and does not. In Fig. 4, the thinner film has lower energy and thus is more sta~ bile. Ifthe system is formed by thinning from a large separation, ‘however, the film may thin only to the metastable film at hy be ‘UNFORM —] hy Pag ENERGY DIFFERENCE BETWEEN LOCAL EOULIBRA ‘2—isotherms of disjoining pressure and specific Inter- | action potential. ‘cause ofthe energy barter separating the local equilibrium states. In this cave, 00. On the other han, ifthe film is formed by ad- sorption from zero thickness, then twill not increase beyond hy, ‘and the system will havea finite contact angle. The effect of metasta- ble film states results in contact angle hysteresis. The contact an- gle ata given capillary pressure depends on te history ofthe film's arrival at equilibrium. ru ) REGION TRANSTION REGION mewscus _—— | wi 1-008 (R-T1)}dh | CONCAVE || conve || concave rust | concave, | cowex | concave toosat a / h Hy ¢ a] he ve | A= tos 6 oun sunsTRATE [Fe spree at mimeniscuewanetion zone, | [la Dijon preeure and angi of ncination a8 ‘SPE Formation Evaluation, Jane 191As the capillary pressure increases for the system in Fig. 4, fy approaches ha, ad the energy barter separating the two metasta bile states decreases. Above some critical capillary pressure, the metastable film with zero contact angle will no longer exis Disjoining Pressure Components Disjoining pressure results from intermolecular or interionc forces. ‘The contributions ae identified as van der Waals, electrostatic, and structural o solvation forces. The fist two contributions were in- twoduced about 1940 wo explain colloid stability. The theory has been called the DLVO theory? after Derjaguin and Landau of the Soviet Union and Verwey and Overbeek of the Netherlands vvan der Waals Interactions. The Hamaker constant (coefficient of the dependence of energy on distance) of thin films ean be cal- ‘ulated from the refractive index, dielectric constant, and absorp- tion frequencies ofthe materials. The results show tha the Hamaker constant of mineral/wateroilsysiems strongly depends onthe ma- terials. van der Waals forces exist between all matter, and thus, are an important component of the surface forces i thin films. A modern approach to quantification ofthe van der Waals surface energy is the Lifshitz theory, which is based on quantum field theory. For ‘quantification, this approach requires the dielectric properties of the media, such asthe zero frequency dielectric constant, theretac~ tive index, and absorption frequencies. An approximation that uses only these parameters is used hereto calculate the Hamaker con- sant for minera/waterfoil systoms. ‘The traditional approach of van der Waals interaction is based ‘on the Hamaker theory, which assumes thatthe interactions are -wise additive and independent of the intervening media and {hat the interaction between two differen media i the geometic ‘mean of the interaction of each medium with itself. The modern approach shows that these assumptions are good if the only in factions are the London dispersion forces (induced dipole/induced Interaction for Plane-Paralel Fm. Israelachvli? reviews the approximations that lead tothe following expression forthe Hamak- ‘er constant of Materials 1 and 2 separated by Material 3 AmAyeo tron a) where 4 ese, ese2 elle #6)] (4) and Aiso= Pr, (e-n3)03-nB) BNE nf ond) Wn 4nd) (Cn? 4nd) +00 40d] o “The material parameters are ak follows: the ero-Frequency dilec- trie constant, ¢; the refractive index in the visible part ofthe spec- ‘rum extrapolated to zero frequency, nj; and the electronic absorption or ionization frequency, ve, whichis assumed tobe the same fr each material, This frequency is similar for water, quart, ‘and alkanes but is smaller for aromatic compounds. Tis the abso” Iute temperature, and k and P are Boltzmann's and Plank’s con- stants, respectively. The above relation does nt include the effect of retardation resulting from the finite speed of electromagnetic in- actions, bt this eflect only decreases the Hamaker constant for film thicknesses greater than about 5 nm. The interaction energy is calculated from the Hamaker constant andthe film thickness oh) = (N22, ° ‘The Hamaker constant inthe form given by Eq. 5 consist of two terms: (I) a zero-frequeney, dielectrie-constant-dependent term proportional tothe temperature, and (2) frequency-and-refractive- index-dependent term that dacs aot have an explicit dependence ‘on temperature. The first term isan entropic term resulting from the Keesom and Debye forces of polar molecules. The second term ‘results from London dispersion forces. The van der Waals conti- ‘bution othe disjoning pressure is negative or attractive when the ‘Hamaker constant is positive and is positive or repulsive when the ‘SPE Formation Evaluation, June 1991 ‘iamaker constant is negative. Each term in Eq. 4 the value ofthe material parameter of the film materi those ofthe bulk phases; otherwise, it is positive. “The traditional approsch'”-9 to calculating the Hamaker con- stant of a film assumed thatthe interactions are pair-wise additive ‘nd thatthe interaction between two different media isthe geomet- Fie mean of the interactions of each medium with itself. The ex: pression forthe Hamaker constant ofa film of Material 3 between Materials 1 and 2 with these assumptions is Anata je-Aast-Ay, o Eq. 6 shows that i the Hamaker constant of the film materia has «Value between those of the bulk phases, the the composite Hamak- er constant ofthe film is negative; otherwise, it is postive. This felation between the material paramctrs is similar to each term of the expressions in Bq, 4 Interaction With Adsorbed Layer on Substrate. The preceding liscuseed a film of Material 3 between bulk phases of Materials ‘and 2. The van der Waals interactions can also be described when the bul phase has one or more adsorbed layers. Parsegian and ‘Ninham2® described how to include adsorbed layers in the Lifshitz theory. Vold?" and Vincent2? included the effect of adsorbed lay- es, assuming pair-wise additivity and geometric mean interactions between dissimilar media. The latter approach will be used here. ‘Suppose thatthe bulk phase of Material | has an adsorbed layer ‘of Material 4 of thickness 8. The film phase is of Material 3 of thickness A, and the other bulk phase is of Material 2, The van der ‘Waals interaction potentials derived from summation of pair-wise interactions (including addition and subtraction of terms so that the remaining terms are interactions of semi-infinite bodies across a ‘up and the assumption of geometric mean interaction berween dis- similar media. The interaction potential for this case is negative if is between a ‘When 3s small compared with hte interaction approaches the casein the absence of an adsorbed layer—ie., film of Material 3 between bulk phases of Materials 1 and 2, When 8 is large com- pared with, the interaction approaches that fora film of Material $3 between bulk phases of Materials 4 and 2. Thus, as h changes from small to large compared with 6, the interaction changes from fone with Material 4 asthe bulk phase to one with Material 1 as the bulk phase. This can change the sign of the interactions. lectrostatic Interactions. A mode! ofthe electrical double-layer interaction as two surfaces approach each other in an aqueous medi- tum is discussed. The interaction can be repulsive, attractive, or ‘combination, The model is used to calculate the necessary condi- tion for the stability of a tick, weting-water fim, ‘When two charged bodies approach each other ina vacuum, the interactions are governed by Coulomb's law. When the interven- ing medium is water, the interactions are complicated by the pres- cence ofthe electrolyte ions. ‘The electrical field near a charged surface decays approximately exponentially with a decay length called the Debye longth that is inversely proportional wo the square root ofthe electrolyte concentration. Ths electrical field, o elec- trical double layer, extends about 1.0 nm in 0.1 Bf NaCl. When two charged surfaces approach each other in water, the overlap of the double layers will cause a change in the system energy. The energy differential with respect to the water-film thickness i force or, expressed per unit area, the electrical contribution tothe dis- Joining pressure. How the energy and disjoining pressure change 188 function of the film thickness and the system parameters is discussed below. ‘Model for Double-Layer Interactions. The interaction of two ‘identical double layers i the foundation of the DLVO theory for colloidal stability. The iteration of dissimilar double layers ap- proximated for low potentials and smeared charges by analytical 2109DDMENSIONLESS INTERACTION POTENTIAL, Fig. 5—Electrostatic specific interaction potentials for con ‘stant charge and potential models. solutions of Hogg er al.? for the constant surface clectical- potential case and by Usui2* for the constant surface electrical~ ‘charge case, These limiting cases of constant potential or charge as the surfaces approach each other bracket the case of constant electrochemical potential, or charge regulation, 2° in which both the potential and the charge change asthe surfaces approach each ‘other. The electrostatic specifi (uit area) interaction potential for ‘these limiting cases can be expressed by the following equation P21 401 HFe2 Neath hy ® ‘The dimensionless interaction potas p 42Fy, cosech hp) p= elegex® 2p, @) Where 4, is the surface electrical potential of Surface 1 at finite separation. The rai ofthe potentials at infinite separation is Fa, = %eo2/®eor CO) ‘The dimensionless distance is hp? . ay ‘The reciprocal of the Debye length is (e2EC Pe egekT) 2) ‘The plus sign in Eq. 8 applies for a constant surface charge, and the minus sign applies for a constant surface potential ‘The interaction potential forthe constant-charge and constant- potential cases are shown in Fig. § for five different potential ra- ‘0s. When both potentials have the same sign, the interaction poten tial with constant charge is always positive, but the interaction potential with constant potential becomes negative at some stall, thickness unless the two potentials are the same. This attraction of {wo surfaces of constant potential of the same sign but different ‘magnitudes results from a charge reversal on the surface with the lower potential when the electrical field is dominated by that of the surface with the higher potential Eq, 8 and Fig. 5 show tat if the constant potential and constant charge curves are mirror images of each other, then the sign of Fy, is reversed. 18 If the potential curves have’ opposite signs, he 20 ‘TABLE 1—ISOELECTRIC POINT AND POINT OF ZERO ‘CHARGE OF MINERALS" Miner | Quartz Za Calc Btoos ‘Alumina 709 Bruce ‘7 Hematite eto ‘Adsorption isotherms of water on slica show hat these forces are ¢greally diminished a the temperature is raised to 65°C,‘ as may be expected for hydrogen-bonding effects. Other measurements ‘demonstrat thatthe water near silica and cay surfaces i different from bulk water. Water in submicrometer silica capillaries has 4 viscosity as much as 40% greater than that of bulk water, but ‘this increase disappears atTO°C., Nuclear magaetic resonance meas urements ofthis boundary layer have reported thicknesses 10 10, tm, The static dielectric constant of water in Na montmorillonite and silica gel is sinificanly reduced from bulk values. Infrared spectra of weter adsorbed on silica show a strengthening of inter molecular hydrogen bonds, “The assumption thatthe properties ofthe bulk and film phases ‘are uniform upto the interface is certain tobe in error for mixtures because components are adsorbed atthe interfaces. Inthe case of electrolyte solutions, the effect of the ion concentration in- homogeneity is treated by the electrostatic interactions. Crude oll, contain many components, and adsorption effects could result in the oil next to the interface having different dielectric properties {rom the bulk ol. For example, adsogption of asphaltene could r= salt in a 1.0-nm layer having @ refractive index near 1.6, while the bulk oil has a refractive index of about 1.4. Ifthe thickness ‘and properties ofthe adsorbed layer are known, then its effect could ‘oe included in the model for van der Weal forces with an adsorbed layer, 2s discussed earlier. ‘Agel layer will form on the surface of quartz when iti exposed to water at an elevated temperature for several days.4.5 This gel layer shifts the adsorption isotherm of wate from 2.5 to 3.0 nm. ‘The van der Waals propeties ofthis gl layer should be between those of quartz and water. I these properties are equal to those of wate, then the ayer thickness will correspond tothe cutoff dis- tance oF limiting water-film thickness discussed late. ‘When the surface contains adsorbed polymers, the interfacial region is quite thick and the overlap ofthe adsorbed layers can con- tribute to additional forces called steric forces.? ‘The electrostatic models described earlier characterize the sut- faces as having an electrical potential and surface charge density "These models treat the surface charges or ios as though they are ‘uniformly distributed. The charge density on most surfaces is about fone charge/nm? or less. If two surfaces containing opposite ‘charges with at last one of the surfaces having mobile charges ap- proach within Im, the interaction may be much greater than if the charges Were uniformly dstributed.* SPE Formation Evaluation, June 1991 [[_TaBte 2—MoDeL PARAVETERS FOR BASE CASE ana oat J 0 Ser eon oor ~ | SeeSerelesn wy ° | SEGRE rn ae - ae Fetal & Stability of Wetting-Water Layer thick, wetting-water film will be (metastable if an adequate ener. gy barrier exists between the tick, weting film (with zero or small ‘contact angle) anda thin water film (with a significant contact a le). Inthe case of zero capillary pressure anda flat surface, the condition of marginal sibility is when the sum of the electrostatic and the van der Waals interaction potentials (structural forees are ‘not significant in thick films) has only a focal maximum (i.e. in- flection point) inthe thick-film region. This is equivalent to [=O and dTl/dh=0. This stability condition stil allows the possibility ‘ofa small contact angle at secondary minimum in the interaction poteatal curve. The DLVO stability condition for colloids differs from the stability condition for thin films. The stability condition {or colloids is that the local maximum inthe interaction potential is equal fo or greater than zero, The energy bartier is messured ‘with respect 10 zer0, ofthe value a infinite separation, because the particles approach each other from large separations and the kinetic energy may past it across the secondary minimum on its way tothe local maximum. Usuil® derived the conditions for col- Toda stability in an unsymametrical system using Eq 8 forthe elec- ‘tical interaction potential. It is assumed thatthe maximum occurs ata large thickness, compared withthe Debye thickness. Ths as- sumption simplifies Eq, 8 and gives a result independent ofthe as sumption of wheter the charge or potential is held constant. Usui's theory is modified forthe condition of film stability with zero capl- lary pressure; ic, the condition that the interaction potential has ‘local maximum between the thick, wetting film andthe thin film, or equivalently, the disjoining pressure has a local maximum equa to or greater than zero. This results in the following expression forthe minimum stable electrical potentials, which isthe same 2s that of Usui except for the numerical coefficient Bey beod =EXPVALB2A TEE 03) Fig. 6 shows the disjoining pressure isotherms for different clee- trolyte concentrations, assuming a charge density of 0.1 ion/nm? anda Hamaker constant of 110-20 J. The maximum ofthe dis: joining pressure isotherm decreases with increasing electrolyte con ‘centration and becomes negative slightly above 0.1 Mf. Above a 1LO.M concentration, the isotherm is dominated by the van dee Waals forces except at smu thicknesses, where the structural forces are imporcant. Buckley et al,!2 used adhesion tests to study the stability of| thick, weiting-water films between glass and crude oil. Their st bility condition was that the local maximmom ofthe disjoining pres- sure had to be 2er0 or a small positive value. This condition of ero 90". As the limiting thickness resulting from the stuc tural forees increases, the contact angle decreases because it limits ‘or cuts off tho interaction potential at this thickness. AS a refer. fence, one monolayer of water is about 0.3 nm thick. As the elec twical potential of the second surface becomes increasingly different from that of the first surface, the contact angle inereases because the electrical contribution to the interaction potential decreases and ‘becomes negative. This example ilustrates the importance of the presence of the hydration water on the mineral surface. Ifa de- hydrated mineral surface i fist contacted with ol large contact angle can result when itis subsequently contacted with water. Con- versely, these results show that ithe mineral has a layer of hydra tion water omy one of two monolayers thik, the contact angle wil be <90" ig. 11 is un example with an electrolyte concentration of 1.0 [M, Curves are shown for different values ofthe Hamaker constant and surface electrical potentials. The results at this high electro |e concentration are not very Sensitive to the electrical potential ‘SPE Formation Evaltion, June 1991 Water Receding on. waver | ‘SLCA Water Advancing wareR Ubi Beyer Sica charge‘ Mvstatlon Water Fig, 12—Etfect of surfactants on receding and advancing eon tact regions. value (assuming smeared charges) except for the smallest value of | the Hamaker constant. Again, the importance ofthe limiting water- layer thickness ilustrated. A discrete sito model may show that electrostatic interactions are important at small separations when the surfaces have charges of opposite sign. ll Films and Contact Angle In Oll Phase . The preceding diseussion of the contact angle when a thin water film exists between the mineral and ol applies during the accurnu- lation of oil in a reservoir when water isthe receding phase, Dur- ing the waterlood stage, surface forces, such as those from natural crude oil surfactants, can result in anol film remaining on the min- eral surface, as shown in Fig. 12. Brown and Neustater"® ob- served contact angle hysteresis ofan intial (water receding value ‘of 30° (in aqueous phase) and an advancing value of 180°. As- ‘une tha the Structural forces retain tin adsorbed layer of ydra- tion water of thickness 6 between the mineral surface and the of film, Also assume thatthe structural forces of the steric interac- tion type resulting from the surfactants (such as asphaltenes) ad sorbed on the two oil iterfaces limit the thinning of the ol film. For example, Christenson and Isaelachvili™’ measured a 3.5-am hard-wall adsorbed asphaltene layer between mica surfaces. Any electrostatic interactions within the ol film are neglected, The coo tact angle inthe oil phase is determined by the van der Waals in teractions, with the cutoff distance determined by the steric interactions. The Hamaker constants used to calculate the van der Waals interaction from Eq. 7 are 6.5% 10-2 J for quart, 3.710729 J for water, und $,0%10~2 J for ol. Fig. 13 shows the isjoining pressure isotherms for an oil film between bulk water and « quartz surface with an adsorbed hyda- tion water layer of thickness 8. If no hydration water is present, the disjoining pressure is positive or repulsive—ice., oil weting, ‘The Hamaker constant ofthe oils intermediate between quartz and water, which, fom Eq, 7, means tht this constant is negative. When the underlying water layer is thick, the disjoining pressure is nega tive, or atractive, Eq. 7 shows that when the water layer is very thick compared with the oil il, the Hamaker coastant approaches ‘that of an oil film between bulk water phases. For this case, the m3| OL FL THOS, rm Fig. 13—Disjoining pressure isotherm for el fim over hydra- ton water layer on allica substrate. ur 02. PL HCE, oe 14—Contact angles in oll phase for oll lm over hydra water layer on sllea substrate. UPON DISCOVERY ‘OL _ AFTER WATERFLOOD WATER Fig. 15—Thin fms and contact angles in rock pore space. | disjoining pressure is attractive, and theo film would thin to noth ing if nothing (¢.g., an adsorbed asphaltene layer) existed to limit ‘the thinning. At intermediate water-film thicknesses, the disjoin ing pressure changes sign with oi-film thickness. Eq, 7 shows that ‘the interaction potential can change sign as a function of the oil- film thickness. Fig. 14 shows the contact angle in the oil phase, if the oil-film thickness is limited by steri-type structural fores. If no adsorbed wate layer exists, the contact angle (in the ol phase) is zero regardless of the limiting oil-film thickness; ., the oils ‘Wetting the dehydrated quartz surface. In the absence of an adsorbed water layer, adsorption of surfactants in the oil and steric effects are not needed for oil wetting; ie., the van der Waals interactions fare sufficient. With increasing thickness ofthe adsorbed wate layer, {he Limiting thickness ofthe oil film must also increase for the con- tact angle to be zero. Ifthe limiting ol-flm thickness is less than the diameter of an alkane chain (the continuum assumption breaks down for submonolayers) of 0.4 nm, the contact angle is very de- pendent onthe value of the limiting thickness, Ifthe oi-ilm thick ‘ess is limited by adsorption of asphaltene aggregates of about 1.0 fhm diameter, then the contact angle is equal to or close to zero even with a thick adsorbed water layer. These results show that the contact angle in the ol phase isa function ofthe thickness of the adsorbed hydration water and the limiting cutoff thickness of the oil film, in addition to the van der Waals interactions 2 ‘This model of oil wetting and hysteresis resulting from a surfactant-stabilized film is consistent withthe experimental ob- Servations of Aronson et al.32 and Brown and Neustadter.°9 In Ref, 32, the oi soluble surfactant N.N-dimethyidocosylamine spread spontaneously on the silica surface, ‘McGuiggan and Pashley’ studied the effects of surfactants on contact ange hysteresis using aic/aqueous surfactant solution/mica. The surfactant was dihexadecyldimethylammonium acetate, a sur factant with extremely low monomer solubility, forming bilayer aggregates in solution, and which adsorbs onto mica as ether a ‘monolayer or bilayer, depending on solution concentration. The inal water advancing contact angle on a dry surface was about 90°, The state advancing contact angle slowly changed to some- ‘what less than the initial advancing contact angle. The initial reced- ing angle, measured. immediately after the contact line was, advanced, was also only somewhat less than the inital advancing contact angle. Ifthe system: was allowed to equilibrate 0.5 t03 hours however, then the waterreceding contact angle decreased to be- low 10°. These results demonstrat that significant time-dependent hysteresis effects take place when the surfactant is very large and has a very slow diffusion rat. Films and Contact Angles In Pore Space ‘Two types of films and contact angles illustrated in Fig 15) have been discussed: water film with a contact angle inthe Water phase ‘SPE Formation Evaluation, June 1991and an oil film with a contact angle in the oil phase. The water film with the contact angle in the water phase could describe the fluid distribution asthe ol initially enters the pore space. The oil film and the contact angle inthe oil phase could describe the Nuid distribution after a waterflood. Also, alter the waterflod, the contact angle in the water phase could change as 2 result of movement of the contact line. The interrelationship of these contact angles with the pore geometry is needed to describe the effect of wewtabiity ‘on the capillary pressure and relative permeability curves and the residual oi saturation Recent Results and Needed Research van der Waals Interactions. Ihave determined the van der Waals interactions, including the Hamaker constant, for combinations of 43 different materials using the complete Lifshitz theory 163° with the computer program provided by Gee.” The approximate for- mula (Eq. 4) is accurate at small thickness for quartz/water/alkane systems. It overestimates the exact (Lifshitz) value for aromatics Retardation effects should be included for films thicker than $ nm. ‘The appropriate refractive index in the approximate formula is the value extrapolated to zero frequency, not the usual tabulated value at the Na D line, Our best estimate of the Hamaker constant for ‘quartz/water nexadecane is 0.85% 10-2 5, Electrostatic Interactions. Recent results by Buckley and ‘Morrow? show thatthe wetting transition at low salinities corre- sponds to that predicted from the electrostatic and van der Waals calculations. However, high salinities indicate the importance of other mechanisms [My experience has shown tha extreme care must be taken o pre- pare and equilibrate samples for zeta potential measurements to have eproducible and correlatable results. Als, the ASTM D664-81 procedure cannot be used to measure the base content of cra oils. Zeta potential measurements are needed for minerals other than lass and quartz. In addition to the pH, such minerals as calcite have zeta potentials that area function ofthe calcium and carbonate concentrations. *? The carbonate concentration isa function of the (COs partial pressure ‘Traditional models of electrostatic interactions use the mean field approximation; e., surface charges are uniformly distributed and the ions are point charges, The thin films that exist in the ease of| finite contact angles may require consideration of such structural interactions as correlation of surface charges and specific ion ef- fects. Models are needed forthe charge regulation that mist occur «as (wo surfaces approach each other Structural and Solvation Interactions. Short-range structural and solvation interactions‘'- are needed to explain the magnitude of the contact angle when its nonzero. We have found thatthe inter- actions resuting from the dscreteness ofthe Muid ina submonolayer ‘ean be modeled with a 2D equation of state to predict the contact angles and spreading coefficients of van der Waals fluids (no elec- ‘wostatic or hydrogen-bonding interactions).® This establishes, in principe, thatthe contact angle can be predicted when the struc- tural interactions are known. The structural intraction information needed for petroleum reser- voirs is the magnitude of the hydrogen-bonding (Lewis acid! base)$?- interaction of water and polar crude oil functional groups with mineral surfaces, Some results have been reported for silica and alumina surfaces using heat of adsorption, heat of im ‘mersion, temperature-programmed desorption, and infrared spec- troscopy. Studies on minerals of interest t petroleum reservoirs are needed. Force measurements of cra il between two mica surfaces show that a thick, irreversibly adsorbed layer sometimes forms with time. 5 The thickness ofthis adsorbed oil layer isa parameter in the model for the water-advancing contact angle Rock Morphology. The rock morphology is beyond the scope of this paper. An understanding ofthe interactions between contact angle and rock morphology, however, is needed to provide a prac- ‘ical understanding for the practicing enginer. ‘SPE Formation Evaluation, une 1991 Conclusions 1. The Hamaker constant for van der Waals interactions can be calculated from the dielectric constant, refractive index, and ed- sorption frequency of the materials, 2. The interaction ofthe electrical double layers on two surfaces as they approach cach other result in interaction potentials that are cither postive (repulsive), negative (attractive), or a combination. 3. The electrostatic imeraction potential isa function ofthe mag~ nitude and sign of the electrical potentials ofthe surfaces in isols- tion and of how the charge and potential change as the surfaces interact with each other. 4. The condition forthe (meta)tability of a thick, wetting-water film on a at surface with zero capillary pressure can be estimated from the electrical surface potentials i isolation, the Hamaker con- stant, and the electrolyte composition 5. A thick, wetting-water film can be collapsed with increased capillary pressure and/or curvature (convex) ofthe solid substrate. 6. The contact angle with a thin water film isa function ofthe Hamaker constant and strongly depends onthe limiting thickness of the hydration water, 1. The contact angle is also a function ofthe electrical potentials of the two surfaces, but the electrostatic interactions (assuming smeared charges) become less significant at high electrolyte con- centrations. The thickness of the hydration water, however, may still epend on the mineral-surface charge and on the effect of dis crete charges. The adsorption of an oil film still depends on the interface charges. '8, When advancing water moves the contact ine, natural surfac> tants in the erude oil may retain an oil film on the substrate. In this case, the contact angle inthe oil phase depends on the Limiting thickness of both the hydration water and the oll film, in addition to the Hamaker constants. Nomenclature A= Hamaker constant, J C= concentration, molecules/m? ‘e = charge per electron, 1.602% 10-9 C = ratio of electrical potentials film thickness, m variable of integration, m equilibrium film thickness ‘mean curvature, m=! Boltzmann's constant, 1.381%10-2 1K refractive index pressure, Pa Planck's constant, 6.626% 10-4 J-s capillary pressure, Pa = equilibrium spreading coefficient, 1m? temperature, K valence of ion local inclination angle of interface thickness of absorbed layer, m 1c0s 8 relative permittivity or zero-frequency dielectric constant ‘ero-frequency dielectric constant of Material i permittivity of free space, 8.854%10 12 C2/J-m contact angle reciprocal Debye length, m-! ‘chemical potential, Jmol electronic absorption frequency, Hz disjoining pressure, Pa IFT, Vim? film tension, J/m? electrical potential, V gravitational potential, kg specific interaction potential, Jim? ‘Aw = departure from equilibrium is, /m? = grand canonical potential, 3 nsTE] George J. Hirasak! is » research ad- | Viser in the Reservolr Mechanism Re- ‘search Depl. of Shell Development Co. In Houston. He joined Shell in 1967 and ‘worked for 2 years Inthe West Coast Div. ‘a8 eectlon leader of reservolr engineer. Ing. He holds a BS degree from Lamar Usand a PhD degree trom Rlce U., both In.chemical engineoring. Recipent of the 1989 Lester C. Uren award and edltor of the SPE monograph Reservoir Simula- tion, Hirssak served on program committees for the 1874 and 1976 Annual Meetinge, on the 1981 Forum Series Commit- tee, and on the Monograph Committee in 1878 and trom 1982 »>0 = nonzero frequency 1,2,3 = materials or surfaces References 1. Anderson, W.G.: “WenubiltyLiteranure Survey—Par 4: fot of ‘Weetabity on Capillary Pressure," JPT (Oct. 1987) 1283-1300. 2, Anderson, W.G.:""Wetability Literature Survey—Pur 5: The Etfects ‘of Wetsbily on Relative Permeability," JPT (Nov. 1987) 1453-68 3. Iraelacvi, JIN Intermolecular and Surface Forces, Acad Pres, [New York City (1985). 4. Dedaguin, B.V., Churaev,N.V., abd Maller, V.M.: Surface Forces, CConsutanis Bureau, New York City (1987) ‘5. Deraguin, B.V.: Theory of Stability of Colloids and Thin Films, Con- sallants Bureau, New York City (1989). 6. Desaguin,B.V. and Chracy, NV: "Properties of Water Layers Ad neato nteraces," Fu Inerfcial Phenomena, C.A. Croton Solin Wiley & Sons, New York City (1986) Chap. 13 7. Moan K.K.: "Fuk in Poros Meda: Two-Phase Distribution snd Flow,” PRD diseration, U. of Minnesota, Minneapolis (1981). 8. Interfacial Phenomena in Oli Recovery, NR. Morrow (ed), Marcel DDeker Inc, New York City (1990) Chaps. 2 and 3. 9, Melrose, J.C: "Inerprettion of Mined Wetability tats in Reser voir Rock," paper SPE 10771 presented atthe 1982 SPE Annual Tech ‘ial Conference and Exhibition, New Orleans, Sept. 26-29. 10 Hall, A.C, Collis, SH, and Melrose, 3.C.: *Stabiy of Aqueous Wetng Fils in Ainbasca Tar Sands, SPBY (April 1983) 249-58 11. Takamura, K:"* Microscopie Strcture of Atabatcs Ol Sand,” Clk J. Ghent Eng. (Av. 1982) 60, 538-45 12 Buckley, 1.8, Tatamura,K., and Morrow, N.R: “IaNuence of Flec- ‘acl Surfce Charges onthe Weting Properties of Crude Oi,” SPERE (ug. 1985) 332-40. 13, Deragun, B.V. Zh. Fe. Khim, (1940) 14,137 14, Deraguin,B.V.: "Definition ofthe Concept of and Magnitude of the Dishining Presue ant Its Role in he Satis and Kinetes of Thin Layers of Liquids," Colloid J. (1955) 1, No. 3, 191-97 15, Fumkn, A\N. "Wetting and Adherence of Bubbles," Zh. Fl. Kim. (4938) 12, 337-45, 16, Hough, D.B. and White, LLR.: “The Calculation of Hamaker Con- stants fom Lifshitz Theory with Application to Weting Phenom, ‘Adv. Cold Inerface Sc. (1980) 14, 3-81 17, Bargeman, D. and Van Voorst Vader, Evan der Waals Forces Be tween Immers Partcle, "J. Blecroana. Chem. (1972) 37, 45-5. 18. Usui 5.2 "Heteocougulaion,” Prog. Surface Membrane Se. (1972) 5, 233-66 19, Richmond, P: "The Theory and Caleltion of va der Waals Forces," J. Colloid $e. (1978) 2, 130-7 20, Puseglan, V-A. snd Nisham, B.W.: “Toward the Coret Calcul ‘hon of van der Watle Interactions Between Lyophobic Cll nn ‘Aqueous Medium," J. ColltdInerface Sei. (1971) 31, No.2, 332-41. 21, Vol, MJ: "The Effect of Adsorption on te van det Waals Interac- tion of Spherical Colloidal Prices," J. Colloid Se. (1961) 16, 1-12. 26 22 Vincent, B: “The van der Waals Atracton Between Colloid Par les Having Adsorbed Layers I. Caleulaton of Interaction Curves,” J, Cold interface Si (1913) 42, No. 2, 270-85. 23. Hope, Re, Healy, T-W., and Fuerte, D.W.: "Mutual Coagula- ‘on of Colloidal Dispersions," Trane. Faraday Soe. (1966) 62, 1ese-st. 26, Usui, 8. “Tneraction of Electrical Double Layers t Constant Surface Charge.” J. Called Ierfoce Se. (1973) 44, No, 1, 107-13 25, Chat, D,, Healy, T.W.,and White, LR: “Electrical Double Layer Interactions under Regulation by ‘Surface Ionization Equira— Dissimilae Ampboteric Surfaces," J. Chem. Soe. Faraday Trans. 1 (1996) 72, 2884-65. 26, Huoter, RJ. “Zeta Potential in Colloid Science," Principles and Ap Dlcations, Academic Press, Washington, DC (1981) 27, Puersena, D.W.: "Inecacial Proceses in Mineral Water Systems," ure Applied Chem. (1970) 24, 135-64 28, James, FLO, and Parks, GA: Characterization of Aqueous Colloids by Their Bletrical Dole Layer and Insc Surface Chemical Prop- ties," Surface and Colloid Science, Maeve (ed), Peau Pes, New York Ciy (1982) 12, 119-216, 29, Somasundaran, P. nd Aga, GE: "The Zero Pint of Charge of Cl- cite," J. Colloid Incerface Si. (1967) 24, 433-40. 20, van Oighen, H.: dn Iroduction 1 Clay Colloid Chemistry, second ‘sion, Joba Wiley & Sons, New York City (1970. 31. Iseacvli J: "Solvation Fores and Liu Strocue as Probed by Direct Force Measurements," Accounts Chem es. (1987) 20, 815-21 5, Aronson, MLP., Ptko, MF. and Priacen, H.M.: "On the Subilcy ‘of Aqucous ime Between Ol and Silica,” J. Colloid Inerice Se (Gune 1978) 68, No. 2, 296-206. 38, Brown, CE. and Neuter, ELL: "The Weabity of OUWateSiica Systems with Reference to Oil Recovery," J. Cn. Pet. Tech. uly= Sept 1980) 100-10. 24, Christensen, H.K. apd Israclachi,.N.: “Direct Measurements of Interactions bad Viscosity of Crude Oils in Thin Fins Between Model Clay Suraces,” J. Colloid Interface Si. (Set. 1987) 119, No.1, 134-202, 35. McGuiggan, PM. and Pashley, RM: "A Study of Surfactant Sol tion Wetting on Mica,” Collis and Surfaces (1987) 27. 277-3 36, Dayalshinsi, LE, Lie, EM. and Piva, LP.:""The General ‘Theory of van der Waals Forces,” Advan. Phys, 10, 165-208. 37, Goe, ML "Surface Fores in Thin guid Flms on Qua,” PAD dsserticn, U. Melbourne, Melbourne, Australia (1987 38, Bucy, 1S. and Movs, NLR: "Characterization of Crue Ot Wet- ting Bevior by Adhesion Test, paper SPE 20263 presente at the 1980 SPEIDOE Synposium on Eancod Oil Recovery, Tulsa, Apel 22.25, 39, “ASTM D66t-81, Sundar Test Methods for Newalizaion Number by Potentiometric Traton" Annual Book of ASTM Standard, A Soc. Testing Materials, Phlaelphin (1983) 08.01, Sect 5, 327 40, Thompson, D.W- and Pownall P.G Surface Ble Properties of Cae," J; Colloid Inerface Si. (1989) 134, No. 1, 7482. 41, Christensen, H.K: “Now DLVO Forces Between Surfaces Solvation, Hyaration, and Capillary Bets," Disp. Sc. Tech (1988) 9, No. 2, 171-206, 42, Derjaguin, B.V. and Churaey, N.V.: “The Current State ofthe The- cn of Long-Range Surface Forces," Colloid and Surfaces (198) 4 2m, 43. yanOss, CJ, Chaodhory, M.K., and Good, RJ. “interacial Lisi ‘an der Weal and Polar Iteracons in Macroscopic Systeme,” Chem. ey. (1988) 88, 927-41 ‘4. Costanzo, PM, Giese, RF, and van Oss, CJ: “Determination of ‘the Ack Base Characavss of Clay Miner Surfces by Comact Angle “Measurement—Imglicatinne forthe Adsorption of Orgaic Solute rom Aqueous Media,” Adhesion Se. Teoh (1990) 4, No, 4, 267-75. 45, Fang J. and Christenson, HL: "Viscosity and Stadies ‘of Austilian Crude Oil in Thin Fs,” 1. Disp. Set. Tech (190) My No.2, 97-118 31 Metric Conversion Factors. A x 1.0" E-o1 Bru x 1.055086 E+00 ddyneslem x 1.0" E+00 f2 x 9.290 3088 E02 oF CF-32y/18 pei X 6.894757 E+00 = kPa “onan tr oa SPEFE ‘rial PE mars nse ren Mach 8 808 Paes x pttion SOx out ies mar nie or 2, 900, ow GPE PON ee ‘rate i88 SPEDE Eanes OF coy Sos on Tan, Ao 7-2 SPE Formation Evaluation, June 1991