44 Oilfield Review

Fundamentals of Wettability

Wael AbdallahEdmonton, Alberta, Canada

Jill S. BuckleyNew Mexico Petroleum Recovery Research CenterSocorro, New Mexico, USA

Andrew CarnegieKuala Lumpur, Malaysia

John EdwardsBernd HeroldMuscat, Oman

Edmund FordhamCambridge, England

Arne GraueUniversity of BergenBergen, Norway

Tarek HabashyNikita SeleznevClaude SignerBoston, Massachusetts, USA

Hassan HussainPetroleum Development OmanMuscat, Oman

Bernard MontaronDubai, UAE

Murtaza ZiauddinAbu Dhabi, UAE

For help in preparation of this article, thanks to Austin Boyd,Gabriela Leu and Romain Prioul, Boston; Ray Kennedy,Edmonton; Patrice Ligneul, Dhahran, Saudi Arabia; JohnMcCullagh, Sugar Land, Texas, USA; Guillemette Picard,Clamart, France; Raghu Ramamoorthy, Abu Dhabi, UAE; andAlan Sibbit, Moscow. Thanks also to the participants of theMay 2007 Schlumberger Wettability Workshop, Bahrain.ECLIPSE, RSTPro (Reservoir Saturation Tool) and WFL (WaterFlow Log) are marks of Schlumberger.

Understanding formation wettability is crucial for optimizing oil recovery. The oil-

versus-water wetting preference influences many aspects of reservoir performance,

particularly in water flooding and enhanced oil recovery techniques. Making the

assumption that a reservoir is water-wet, when it is not, can lead to irreversible

reservoir damage.

Wetting forces are in play all around us. Theyhave practical applications, such as making rainbead up on a freshly waxed car so it is protectedfrom rust. And they provide whimsy: wetting forces bind sand grains to hold the shape of achild’s sand castle.

Forces of wetting influence hydrocarbonreservoir behavior in many ways, including satura -tion, multiphase flow and certain log interpre -tation parameters. However, before getting intothese details, it is best to first establish whatwettability is.

Wettability describes the preference of a solidto be in contact with one fluid rather thananother. Although the term “preference” mayseem odd when describing an inanimate object,it aptly describes the balance of surface andinterfacial forces. A drop of a preferentiallywetting fluid will displace another fluid; at theextreme it will spread over the entire surface.Conversely, if a nonwetting fluid is dropped ontoa surface already covered by the wetting fluid, it

will bead up, minimizing its contact with thesolid. If the condition is neither strongly water-wetting nor strongly oil-wetting, the balance offorces in the oil/water/solid system will result ina contact angle, θ , between the fluids at thesolid surface (below).

In many oilfield applications, wettability istreated as a binary switch—the rock is eitherwater-wet or oil-wet. This extreme simplificationmasks the complexity of wetting physics inreservoir rock. In a homogeneous, porousmaterial saturated with oil and water, “stronglywater-wetting” describes one end member of acontinuum in which the surface strongly preferscontact with water. A strongly oil-wetting surfaceprefers contact with oil.1 Degrees of wettingapply along the continuum, and if the solid doesnot have a marked preference for one fluid overthe other, its condition is termed intermediate-wetting or neutral-wetting. Parameters thatinfluence where on the continuum a system liesare discussed later.

> Contact angle. An oil drop (green) surrounded by water (blue) on a water-wet surface (left) forms a bead. The contact angle θ is approximately zero. On an oil-wet surface (right), the drop spreads,resulting in a contact angle of about 180°. An intermediate-wet surface (center) also forms a bead, but the contact angle comes from a force balance among the interfacial tension terms, which are γsoand γsw for the surface-oil and surface-water terms, respectively, and γow for the oil-water term.

θ ~ 180°θ ~ 0°

θ

θ

γso = γsw + γow cos θ

γow

γswγso

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 44

Summer 2007 45

Reservoir rocks are complex structures, oftencomprising a variety of mineral types. Eachmineral may have a different wettability, makingthe wetting character of the composite rockdifficult to describe. Typically, the primary consti -tuents of reservoirs—quartz, carbonate anddolomite—are water-wet prior to oil migration.

This brings up a further complexity: thesaturation history of the material may influencesurface wetting, such that pore surfaces that hadbeen previously contacted by oil may be oil-wet,but those never contacted by oil may be water-wet. Various terms have been used to describeboth of these conditions, including mixed-,fractional- and dalmation-wetting. In this article,the general term “mixed-wetting” will be used forany material with inhomogeneous wetting. It isimportant to note the fundamental differencebetween intermediate-wetting (lacking a strongwetting preference) and mixed-wetting (having avariety of preferences, possibly includingintermediate-wetting) conditions.

Another important distinction is that apreferentially water-wetting surface can be incontact with oil or gas. Wettability does notdescribe the saturation state: it describes thepreference of the solid for wetting by a certainfluid, given the presence of that preferredwetting fluid. Thus, a water-wet rock can be

cleaned, dried and fully saturated with analkane, while the surfaces in the pores remainwater-wet. This can be easily seen: drop such anoil-saturated but water-wet rock into a beaker ofwater and it will spontaneously imbibe asignificant quantity of water and expel oil.

Strictly speaking, the term imbibition refersto an increase in the saturation of the wettingphase, whether this is a spontaneous imbibitionprocess or a forced imbibition process such as awaterflood in a water-wet material. Conversely,drainage refers to an increase in saturation ofthe nonwetting phase. However, in practice, theterm imbibition is used to describe a processwith increasing water saturation, and drainage isused to describe a process with increasing oilsaturation. Care should be taken when readingthe literature to determine which sense is being used.

This article outlines the effects of wettabilityin the oil field, and then describes the basicchemistry and physics of wetting that explainthese effects. The emphasis here is on oil/water/solid interactions, but there are also gas/liquid/solid systems for which wettability is important.The measurement methods are briefly described.Two case studies from the Middle East and aNorth Sea chalk laboratory study describescenarios that require an understanding of

wettability. Laboratory methods with thepotential to improve our ability to measure andmodel wettability conclude the article.

The Practical Importance of WettabilityThe current, favorable oil price has improved theeconomics of waterflooding and some enhancedoil recovery methods. With multiple phasesflowing in the reservoir, understanding wetta -bility becomes important.2 However, even during

1. Unless otherwise specified in this article, the terms“water-wet” and “oil-wet” are used to indicate “strong” preferences.

2. An extensive wettability literature survey was publishedin 1986 and 1987.Anderson WG: “Wettability Literature Survey—Part 1:Rock/Oil/Brine Interactions and the Effects of Core Handling on Wettability,” Journal of Petroleum Technology 38 (October 1986): 1125–1144.Anderson WG: “Wettability Literature Survey—Part 2:Wettability Measurement,” Journal of Petroleum Technology 38 (November 1986): 1246–1262.Anderson WG: “Wettability Literature Survey—Part 3:The Effects of Wettability on the Electrical Properties ofPorous Media,” Journal of Petroleum Technology 38(December 1986): 1371–1378.Anderson WG: “Wettability Literature Survey—Part 4:Effects of Wettability on Capillary Pressure,” Journal ofPetroleum Technology 39 (October 1987): 1283–1300.Anderson WG: “Wettability Literature Survey—Part 5:The Effects of Wettability on Relative Permeability,” Jour-nal of Petroleum Technology 39 (November 1987):1453–1468.Anderson WG: “Wettability Literature Survey—Part 6:The Effects of Wettability on Waterflooding,” Journal ofPetroleum Technology 39 (December 1987): 1605–1622.

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 45

primary recovery, wettability influences produc -tivity and oil recovery.3 The original wettability of a formation and altered wettability during and after hydrocarbon migration influence theprofile of initial water saturation, Swi, andproduction characteristics in the formation.

Most reservoirs are water-wet prior to oilmigration and exhibit a long transition zone,through which saturation changes gradually frommostly oil with irreducible water at the top of thetransition zone to water at the bottom. Thisdistribution is determined by the buoyancy-

based pressure difference between the oil andwater phases, which is termed the capillarypressure, Pc (below). Oil migrating into an oil-wet reservoir would display a different saturationprofile: essentially maximum oil saturation downto the base of the reservoir. This differencereflects the ease of invasion by a wetting fluid.

Layers within formations can also havedifferent wetting states because of lithologyvariations. A tight zone may remain water-wetting if little or no oil migrates into it, whilesurrounding formations are converted to a more

oil-wet state. Other wetting variations may not be so easily explained. Several carbonate reser -voirs in the Middle East are thought to havevariation of wettability by layer, but the cause isnot yet understood.

This wetting heterogeneity can affectrecovery. For example, models using ECLIPSEreservoir simulation software incorporatedparameters typical of a Middle East carbonatereservoir, with water-wet layers and oil-wetlayers having similar permeabilities. Underwaterflood, water penetrates the water-wetlayers more readily than the oil-wet layersbecause of capillary effects. The simulationshows that little oil would be recovered from theoil-wet layers.

Wettability also affects the amount of oil thatcan be produced at the pore level, as measuredafter waterflood by the residual oil saturation,Sor. In a water-wet formation, oil remains in thelarger pores, where it can snap off, or becomedisconnected from a continuous mass of oil, andbecome trapped. In an oil-wet or mixed-wetformation, oil adheres to surfaces, increasing theprobability of a continuous path to a producingwell, and resulting in a lower Sor.

Because the impact of wettability extendsfrom pore scale to reservoir scale, wettability canaffect project economics. Through the param -eters Swi and Sor, wettability influences oilrecovery, one of the most important quantities inthe E&P business. In addition, the relativepermeabilities of oil and water vary with forma -tion wettability. In projects with huge upfrontcapital expenditures for facilities, such as thosein deepwater areas, failure to understandwettability and its ramifications can be costly.

Wettability affects waterflood performance,which also can involve significant upfrontspending. Imbibition forces—the tendency of aformation to draw in the wetting phase—determine how easily water can be injected andhow it moves through a water-wet formation.Water breakthrough occurs later in a waterflood,and more oil is produced before the water breaksthrough in a water-wet reservoir than in an oil-wet reservoir.

Wettability can also influence gasfloodperformance. The gasflood front or oil bank canmove water, if it is mobile, again generating flowvariation based on oil/water wetting preferences.In addition, if asphaltenes are present in thecrude oil, contact by injected hydrocarbon gasalters the equilibrium condition and can lead toasphaltene precipitation. As discussed later, this precipitation can alter the wettability of thepore surfaces.

46 Oilfield Review

> Forming a transition zone. A homogeneous formation exhibits a zone of transition from high oilsaturation at the top to high water saturation at the bottom (blue curves). This saturation transitionhas its origin in the capillary pressure, Pc, which is the difference between the water and oil pressuresat the interface (equations, above). In a capillary tube, water-wetting (WW) surface forces causewater to rise (left inset), displacing oil, but if the tube inner surface is oil-wetting (OW), the oil willpush water down (right inset). The wetting force, and therefore Pc, is inversely proportional to thecapillary radius. The capillary rise, h, is determined by the balance of wetting forces and the weightof fluid displaced from the bulk-fluid interface. Translating this to a porous formation, there is a free-water level (FWL) defined where the capillary pressure between water and oil is zero. Since porousrocks have a distribution of pore and pore-throat sizes—similar to a distribution of capillary tubes—at any given height above the FWL, the portion of the size distribution that can sustain water at thatheight will be water-saturated. At greater height, the buoyancy of oil in water provides greatercapillary pressure to force water out of smaller voids. In a water-wet formation (left), the oil/watercontact is above the FWL, indicating that pressure must be applied to force oil into the largest pores.In an oil-wet formation (right), the contact is below the FWL, signifying that pressure must be appliedto force the water phase into the largest pores. The oil/water contact divides the zone containingmostly oil from the one containing mostly water.

Dept

h

Capi

llary

pre

ssur

e

Oil/water contact

Pc = 0, free-water level

Dept

h

Capi

llary

pre

ssur

e

Oil/water contact

Pc = 0, free-water level

θ = 0°

h

θ = 140°

hr

Pc = Pnw – Pw

Pc = ρ g hPc = 2 γ cosθ/r,

where Pc = capillary pressurePnw = pressure in nonwetting phasePw = pressure in wetting phase

ρ = density difference between phasesg = gravitational acceleration

h = height of capillary riseγ = interfacial tensionθ = contact angler = inner radius of capillary.

r

Water saturation

WW OW

Water saturation

Oil

Water

Water

Oil

59819schD7R1:59819schD7R1 10/12/07 3:39 PM Page 46

Summer 2007 47

Even in a gas reservoir, wettability or itsalteration can affect recovery. Condensateblockage near a wellbore decreases gasproductivity. Some recovery methods usechemical means to alter the wettability aroundthe wellbore to produce the oil and thereby clearthe blockage.4

Some enhanced oil recovery processes aredesigned to overcome the wetting forces thattrap oil. They aim either to alter the wettingpreference of the formation to be more oil-wetting or to decrease the interfacial tensionbetween the fluids, thereby decreasing thewetting forces.

Some logging methods are also dependent onwetting. Resistivity methods rely on a continuouselectrical path through the rocks, which isprovided by the water phase. In an oil-wetformation, the water may not be continuous. Thisinfluences the saturation exponent, n, in Archie’sequation relating saturation and resistivity.5 Inwater-wet conditions, n is ~2, but in oil-wetconditions, n is greater than 2. So if n is set to 2 inan oil-wet formation, a resistivity-based saturationassessment will likely be wrong.

Nuclear magnetic resonance (NMR) responsesalso depend on the position of the fluids withrespect to the pore surfaces. The nonwetting fluidexhibits relaxation rates similar to those of bulkfluid, because it sits in the middle of pores, whilethe wetting phase has shortened relaxation timesbecause of surface interactions.6

Wettability is of vital importance to drilling-fluid formulation, particularly in oil-base muds.For example, surfactants are included to keepsolids in suspension. An oil-external mud filtratecontaining oil-wetting surfactants invades thenear-well formation, potentially altering thewettability of the pores.7 This can change theposition of the fluids in the pore spaces, whichmay affect some logging responses. Because thealteration may not be permanent, differentmeasurements may be obtained in subsequentlogging runs.

Wettability ChangesWetting forces lead to an equilibrium conditionbetween at least three substances: a solid andtwo fluids.8 The constituents and conditions forall three substances influence the wetting prefer -ence. Thus, we must consider the oil compo nents,the brine chemistry and the mineral surface, aswell as the system temperature, pressure andsaturation history.9

Oil composition is key to changing thewettability of a naturally water-wet surface,because any wettability-altering components are

in the oil phase. These are polar compounds inresins and asphaltenes, both of which combinehydrophilic and hydrophobic characteristics.Bulk-oil composition determines the solubility of

the polar components. A crude oil that is a poorsolvent for its own surfactants will have a greaterpropensity to change wettability than one that is a good solvent (above).10 Temperature,

3. Morrow NR: “Wettability and Its Effect on Oil Recovery,”Journal of Petroleum Technology 42, no. 12 (December1990): 1476–1484.

4. For more on gas-condensate reservoirs: Fan L, Harris BW,Jamaluddin A, Kamath J, Mott R, Pope GA, Shandrygin Aand Whitson CH: “Understanding Gas-CondensateReservoirs,” Oilfield Review 17, no. 4 (Winter 2005/2006):14–27.For an example of wettability alteration in gas-condensate wells: Panga MKR, Ooi YS, Chan KS,Enkababian P, Samuel M, Koh PL and Chenevière P:“Wettability Alteration Used for Water Block Preventionin High-Temperature Gas Wells,” World Oil 228, no. 3(March 2007): 51–58.

5. Archie’s equation can be expressed as Sw = (Rt /Ro)n,where Rt is the formation resistivity at saturation Sw, andRo is the formation resistivity at 100% water saturation.

6. For more on NMR logging: Alvarado RJ, Damgaard A,Hansen P, Raven M, Heidler R, Hoshun R, Kovats J,

Morriss C, Rose D and Wendt W: “Nuclear MagneticResonance Logging While Drilling,” Oilfield Review 15,no. 2 (Summer 2003): 40–51.

7. In an oil-external emulsion, surfactant molecules formclusters called micelles consisting of an aqueous coreencapsulated in a surfactant monolayer, in which thehydrophilic parts of the molecules point inward towardthe aqueous core and the hydrophobic parts pointoutward toward the oil phase.

8. Wetting preference also can involve three immisciblefluids, such as mercury, water and air.

9. Buckley JS, Liu Y and Monsterleet S: “Mechanisms ofWetting Alteration by Crude Oils,” paper SPE 37230, SPE Journal 3, no. 1 (March 1998): 54–61.

10. Al-Maamari RSH and Buckley JS: “AsphaltenePrecipitation and Alteration of Wetting: The Potential forWettability Changes During Oil Production,” paper SPE84938, SPE Reservoir Evaluation & Engineering 6, no. 4(August 2003): 210–214.

>Wettability alteration from asphaltene precipitation. Contact angles were measured after exposureto several crude oils diluted with n-heptane to various oil-volume fractions (top). The contact angleincreased markedly near the asphaltene-precipitation point (large filled circles). Another way toinduce asphaltene precipitation is by decreasing the pressure (bottom). In a PVT vessel, asphaltenesbegin to flocculate, or clump together, as shown in high-pressure microscope photographs, as thepressure decreases to the asphaltene-precipitation point. As asphaltenes come out of solution, thelight transmittance decreases (blue).

Ligh

t tra

nsm

ittan

ce p

ower

, µW

Pressure, psi4,000 6,000 8,000 10,000 12,000 14,000

20

15

10

5

0

7,000 psi

5,500 psi

8,000 psi 8,500 psi 11,500 psi

Bubblepoint

Asphaltene-precipitation point

Wat

er-a

dvan

cing

con

tact

ang

le, d

egre

es

Oil-volume fraction

Mars-Yellow

Mars-Pink

Tonsleep

A-93

Lagrave

0 0.2 0.4 0.6 0.8 1.00

30

60

90

120

150

180

Mars-Pink

Mars-Yellow

Tensleep

A-93

Lagrave

47 µm

47 µm

47 µm 47 µm 47 µm

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 47

pres sure and composition of the crude oil affect asphal tene stability (see “Asphaltenes—Problematic but Rich in Potential,” page 22).

For components of an oil to alter wetting, theoil phase must displace brine from the surface.The surface of a water-wet material is coated bya film of the water phase.11 The part of this waterfilm that is closest to the surface forms anelectrical double layer: excess charges on thesolid surface are countered by electrolyte ions ofopposite charge. The first layer of water withthese ions is static, and the second layerexchanges ions with the bulk water.

When two interfaces—such as the solid-water and water-oil interfaces—are in proximity,the forces acting to keep them separated or drawthem closer together include van der Waals,electrostatic, and structural or solvation inter -actions.12 The net force is often expressed as aforce per unit area, termed the disjoining pres -sure. A positive disjoining pressure holds theinterfaces apart; a negative disjoining pressurebetween interfaces is attractive. The composi -tion of crude oil, and the pH and composition ofbrine influence the disjoining pressure.

Measurements of these quantities have beenused to predict water film stability, and thegeneral trends are upheld by experiment.Atomic-force microscopy (AFM) has been used toimage the solid surfaces after aging, providinggraphic illustration of the complexity of surfaceinteractions (above).13 When the film desta -bilizes, polar components of the crude oil canadhere to the surface and make the surface moreoil-wetting. Dissolved divalent ions, such as Ca2+,can also destabilize the film.

Because of the presence and nature ofionized sites on the solid surface, the range of pHleading to instability is different for carbonatesand sandstones. Silica surfaces are negativelycharged above a pH of about 2, so positivelycharged ions (base chemical species) canadsorb. Conversely, calcite surfaces may bepositively charged below pH 9.5, so negativelycharged ions (acidic species) can adsorb.14

Carbonate wettability is also influenced byspecific interactions with carboxylic acids and bythe reactivity of carbonate minerals.15

The existence of the double layers in thewater phase explains why there is a differencebetween a material that is saturated with crudeoil and one with surfaces that are wetted by oil.So long as the water film is stable, components of

the crude oil cannot attach to the solid surfaceand alter the wetting tendency toward oil-wet.One result of this surface interaction is contactangle hysteresis. The water-advancing contact-angle, which is present when bulk water displacesbulk oil from a surface, can be much larger thanthe water-receding angle, which occurs whenbulk oil displaces bulk water. Descriptions of the surface layers in these two conditions may be complex.16

This recalls another influence on the wettingpreference of a surface, its saturation history. Inan oil-bearing formation, the wettability can varywith depth, with a greater water-wettingpreference near the bottom of the transitionzone and a greater oil-wetting preference nearthe top.17 The higher zones have a greatercapillary pressure, which can counteract the dis -joining pressure and destabilize the water film,allowing surface-active components in the oil tocontact the solid. Lower in the structure, thesolid surfaces mostly retain the water film.

However, saturation in a reservoir is notstatic. Multiple phases of oil migration, develop -ment of a gas cap, leakage of oil and gas from thereservoir and tectonic activity all can affect thesaturation state of a reservoir. These changes willresult in different fluid saturations based in parton the wettability of the surface at the time.

This dependence of saturation on historyapplies not only over geologic time, but alsowithin drilling and production time scales.Drilling fluids, particularly oil-base muds,contain surfactants that can invade pore spaces.This invading fluid can alter wettability in thenear-well region, affecting flow when the well isput on production. Fluids used in workoveroperations can have a similar near-well impacton wettability.

During production, the parameters alreadydiscussed in the context of primary production orwaterflooding can also be changed by injectedfluid that alters formation wetting eitherdeliberately or inadvertently. This action mayresult in improved or damaged injectivity orproductivity. An injected brine whose dissolvedsolid content or pH differs from those of theformation brine can induce wetting changes.Surfactants, including those generated bymicrobial action, can decrease the interfacialtension between fluids and change the contactangle. Quartz tends to become more oil-wet athigher temperatures, but calcite tends tobecome more water-wet.18 Thus, thermalrecovery methods can change wettability.19

48 Oilfield Review

> Effect of brine chemistry on film stability and contact angle. A glass surface was conditioned inwater with a salt [NaCl] concentration of 0.01, 0.1 or 1.0 mol/m3, and a pH of 4, 6 or 8. This water-wetsurface was then aged in a crude oil known to contain components that can alter wettability. Contact-angle measurements showed oil-wetting behavior at low concentration and low pH, and water-wetting behavior at high concentration and high pH (left). The surface-water film retained its stabilityat high concentration and high pH. In related tests, freshly cleaved mica surfaces were aged invarious NaCl solutions and then in the crude oil for 11 to 14 days. When the brine conditions (0.01 mol/m3, pH = 4) allowed a change to oil-wetting state, a surface image from atomic-forcemicroscopy (AFM) shows a complex of micron-sized surface irregularities deposited on the surface(top right). These were thought to be asphaltic material because the irregularities were insoluble indecane. A similar image of a mica surface aged in brine (1.0 mol/m3, pH = 8) that retains a water-wetting surface film indicated no deposits (bottom right).

180

150

120

90

60

30

00 10 20

Aging time in oil, days

30 40 50

Wat

er-a

dvan

cing

con

tact

ang

le, d

egre

es

pH Ionic strength

4

6

8

0.01

0.1

1.0

0 2.5 5.0 7.5 10.0 12.50

2.5

5.0

7.5

10.0

12.5

µm

0 2.5 5.0 7.5 10.0 12.50

2.5

5.0

7.5

10.0

12.5

µm

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 48

Summer 2007 49

As a reservoir is produced, pressure depletioncan alter the composition of the crude, movingthe asphaltene-precipitation point, which canlead to asphaltene deposition in the reservoir.This can also occur due to formation pressure ortemperature decline, which, in addition toasphaltene dropout, can lead to wax formation,gas condensation, or formation of a gas cap; theseaffect the wettability distribution in a formation.

A Pore-Level ViewPore geometry complicates the application of thewetting principles discussed above. A contactangle is easiest to understand when the surfaceis a smooth plane. However, pore walls are notsmooth, flat surfaces, and typically more thanone mineral species composes the matrix sur -rounding the pores.

Surface rugosity confounds visualization of asimple contact angle in a pore, because theapparent contact angle (based on the averageplane of the surface) can differ markedly fromthe true contact angle, which is based on thelocal orientation of the surface (below). Sharppoints, or asperities, on the surface can also bethe loci for thinning of the water film coating thesurface, leading to the potential for wettingalteration at these points.

Conceptual or tutorial models of capillarity inporous media often refer to a “bundle ofcapillaries” model. The distribution of pore sizes

is modeled by a distribution of capillaries withvarious radii. Each capillary is invaded by anonwetting oil phase at a different capillaryentry pressure, which is inversely proportional tothat capillary’s radius. Once the entry pressure isovercome, the whole cross section of thecapillary is filled with oil.

In actuality, the complex geometry of a poreis defined by the grain surfaces surrounding it.The capillary entry pressure in this geometryrelates to the inscribed radius of the largestadjacent pore throat. Although most of the porebody may fill with oil, the interstices wheregrains meet do not fill, because the capillarypressure is insufficient to force the nonwettingoil phase into those spaces.

Thus, depending on the pore and pore-throatgeometry and the surface roughness, some partsof the pore space are oil-filled and the others arebrine-filled (assuming no gas saturation). Somesolid surfaces are in contact with oil, and forsome or all of those surfaces, the water film maynot be stable. Where the film is not stable, thesurface wetting preference can be changed. Thismay lead to a situation of mixed wettability,where some parts of the pore surface are water-wetting and others are oil-wetting. The generallyaccepted theory is that because of the way thiscondition arose, the large pore spaces are morelikely to be oil-wetting, and the small pore spacesand interstices within pores are more likely to bewater-wetting (above).20

11. The double layer of water is almost always present on a water-wet material. It can be removed at hightemperature, but soaking in water or condensation fromhumid air will replenish the double layer.

12. Hirasaki GJ: “Wettability: Fundamentals and SurfaceForces,” SPE Formation Evaluation 6, no. 3 (June 1991):217–226.

13. Buckley JS, Takamura K and Morrow NR: “Influence ofElectrical Surface Charges on the Wetting Properties ofCrude Oils,” SPE Reservoir Engineering 4, no. 4(August 1989): 332–340. For more on wettability based on atomic-forcemicroscopy studies: Buckley JS and Lord DL:“Wettability and Morphology of Mica Surfaces AfterExposure to Crude Oil,” Journal of Petroleum Scienceand Engineering 39, no. 3–4 (September 2003): 261–273.

14. Buckley et al, reference 9.15. Thomas MM, Clouse JA and Longo JM: “Adsorption of

Organic Compounds on Carbonate Minerals – 1. ModelCompounds and Their Influence on Mineral Wettability,”Chemical Geology 109, no. 1–4 (October 25, 1993): 201–213.

16. Hirasaki, reference 12.17. Okasha TM, Funk JJ and Al-Rashidi HN: “Fifty Years of

Wettability Measurements in the Arab-D CarbonateReservoir,” paper SPE 105114, presented at the 15th SPEMiddle East Oil & Gas Show and Conference, Bahrain,March 11–14, 2007.

Jerauld GR and Rathmell JJ: “Wettability and RelativePermeability of Prudhoe Bay: A Case Study in Mixed-Wet Reservoirs,” paper SPE 28576, SPE ReservoirEngineering 12, no. 1 (February 1997): 58–65.Marzouk I, Takezaki H and Miwa M: “Geologic Controlson Wettability of Carbonate Reservoirs, Abu Dhabi,U.A.E.,” paper SPE 29883, presented at the SPE MiddleEast Oil Show, Bahrain, March 11–14, 1995.Andersen MA: Petroleum Research in North Sea Chalk,Rogaland Research, Stavanger (1995): 53–54.

18. Rao DN: “Wettability Effects in Thermal RecoveryOperations,” SPE Reservoir Evaluation andEngineering 2, no. 5 (October 1999): 420–430.

19. Hamouda AA and Gomari KAR: “Influence ofTemperature on Wettability Alteration of CarbonateReservoirs,” paper SPE 99848, presented at the SPE/DOE Symposium on Improved Oil Recovery, Tulsa,April 22–26, 2006.

20. Salathiel RA: “Oil Recovery by Surface Film Drainage inMixed-Wettability Rocks,” Journal of PetroleumTechnology 25 (October 1973): 1216–1224.Kovscek AR, Wong H and Radke CJ: “A Pore-LevelScenario for the Development of Mixed Wettability in OilReservoirs,” American Institute of Chemical EngineersJournal 39, no. 6 (June 1993): 1072–1085.

> Pore surface roughness. The apparent contactangle, measured from the average surface plane,can differ significantly from the true contactangle at a locally inclined surface (top). Even if apore is water-wetting, the surface water may notbe a double layer, but could be thicker due topore rugosity (bottom). At an asperity, the surfaceforces are more favorable for displacing thedouble layer than elsewhere on the surface.

Oil

θapparent

θtrue

Water

Grain Asperity

>Wetting in pores. In a water-wet case (left), oil remains in the center of the pores. The reversecondition holds if all surfaces are oil-wet (right). In the mixed-wet case, oil has displaced water fromsome of the surfaces, but is still in the centers of water-wet pores (middle). The three conditionsshown have similar saturations of water and oil.

Oil

Water-wet Mixed-wet Oil-wet

Brine (water) Rock grains

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 49

This simple view assumes a homogeneousformation with oil migration from below. Mostformations are more complex than this, and thelithological complexity must be taken intoaccount when applying the migration history tothe current wetting state.

In addition to this mixed wettability based onsaturation history, there can be mineralogy-based mixed wettability. The pH and concen -tration conditions for a stable water film aredifferent for quartz, dolomite and calcitesurfaces, and for clays and other compoundswithin the pore space. Thus, different grains mayhave different wetting preferences.

Many experts today believe that most oilreservoirs have some mixed-wetting charac -teristics. The original, water-wet condition isaltered to some extent by oil migration. In thefollowing discussion, we examine the implica -tions of the wetting condition on multiphase flow

by examining the case of two-phase flow througha uniformly water-wet medium and then a mixed-wet medium (above).

Water-wet case—As the preferentiallywetting phase, water will be in the small spacesthat were not invaded by oil. Oil will be in thelarge pores. Before this formation is produced,both phases are continuous, although theconnate water phase in the highest part of theformation may have such a low saturation thatthe relative permeability to water, krw, isessentially zero. Since electrical resistivitylogging will respond to a continuous, conductingwater phase, the use of an Archie saturationexponent, n, of around 2 is valid.

Under natural or induced waterflooding, bothphases flow. The oil relative permeability, kro, ishigh, since oil flows through the largest pores,and decreases as oil saturation decreases. Thewater relative permeability, krw, starts low andincreases as water saturation increases.

Water saturation increases preferentially inthe smaller pore spaces first, due to wettingforces. As the displacement moves from smallerto larger pores, the water increasingly occupiespore throats that were formerly filled with oil.One pore or a group of pores containing oil canbecome cut off from the rest of the oil. Lackingsufficient driving pressure to overcome thecapillary entry pressure for the now water-saturated pore throat, the oil is trapped in place.

Eventually, all continuous flow paths arewater-filled, and oil stops flowing. The final krw islower than the original kro because of the oiltrapped in large pores.

This trapped oil is one target of enhanced oilrecovery methods. Some of these methods seek tomobilize the oil by lowering the interfacialtension or by changing the contact angle. Bothhave the effect of decreasing the capillary entrypressure. Another way to produce more oil is byincreasing the pressure gradient, or viscous force,within the pore. The ratio of viscous force due toa drive pressure and capillary force at the driveninterface is called the capillary number.21 A highcapillary number results in greater recovery; itcan be increased by lowering the interfacialtension or making the pressure drop larger.

Mixed-wet case—In this case, the oil likelymigrated into a water-wet formation, so theoriginal water and oil saturation distribution maybe macroscopically similar to the case describedabove. However, in a mixed-wet case, the oil occu -pying the large spaces of the pores has altered thewettability of the contacted pore surfaces.

As before, initially kro is high and krw is low.However, as the water saturation increases, itinvades the largest pores first and remains in thecenter of those pores, because of the oil-wetcondition of the surfaces surrounding thosepores. This causes a more rapid decline in kro asthe most permeable paths fill with water.However, the water does not trap the oil, becausethe oil-wet surfaces provide a path for the oil toescape from nearly water-filled pores. Theflooded water may not be in contact with theconnate water, which can yield an Archiesaturation exponent, n, greater than 2.

In this mixed-wet condition, when waterbreaks through to a producing well, oilproduction continues for a long time, althoughthe water cut increases. Laboratory tests oncores prepared with a procedure resulting inmixed-wet conditions of varying degrees showthat maximum oil recovery is obtained forslightly water-wet samples.22

50 Oilfield Review

> Capillary pressure and relative permeability for water-wet and mixed-wet conditions. This schematicrepresentation contrasts possible capillary-pressure, Pc, (red) and relative-permeability curves forwater, krw, (blue) and oil, kro, (green) for water-wet (left) and mixed-wet (right) reservoirs. The firstcurve to consider is the primary drainage Pc curve (dotted), which indicates a certain pressure in theoil phase that is required before a substantial displacement of water can occur. Since most reservoirsare considered to be water-wetting when oil first migrates, this curve is also used for the mixed-wetcondition. The other curves (dashed = increasing water saturation, solid = increasing oil saturation)differ based on the wettability change due to oil contact with the surfaces in the large pore spaces. In the strongly water-wet situation, the capillary-pressure curve stays positive over most of thesaturation range, while in the mixed-wet case its sign has both positive and negative portions,signifying that some parts of the surface imbibe water and others imbibe oil. The kro values are less at low water saturation in the mixed-wet case, because the oil is in competition with water in thelarge pores. Similarly, the krw at high water saturation is reduced in the water-wet case because theoil preferentially occupies the large pores.

Rela

tive

perm

eabi

lity,

%

50 0

–

+

00 100Water saturation, %

Water-wet100

Capi

llary

pre

ssur

e

Rela

tive

perm

eabi

lity,

%

50 0

–

+

00 100Water saturation, %

Mixed-wet100

Capi

llary

pre

ssur

e

Pc Pc

kro

krwkrw

kro

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 50

Summer 2007 51

In both water-wet and mixed-wet conditions,hysteresis in relative permeability and capillarypressure accompanies changes in saturation(above). This reflects the difference betweenwater-advancing and water-receding contactangles, and the locations of oil and water in thepore spaces.

Oil-wet case—The extreme of a completelyoil-wet reservoir is unlikely except in a reservoirthat is its own source rock. In that case, thekerogen—organic solids that can yield oil uponheating—in place and the oil-maturationprocess could result in oil-wet surfaces.

Measuring WettabilitySeveral methods are available to measure a reser -voir’s wetting preference. Core measure mentsinclude imbibition and centrifuge capillary-pressure measurements (below). An imbibitiontest compares the spontaneous imbibition of oiland water to the total saturation change obtainedby flooding.

The Amott-Harvey imbibition test is commonlyused.23 A sample at irreducible water saturation,Swirr, placed into a water-filled tube spontaneouslyimbibes water over a period of time—at least10 days, and sometimes much longer. Then thesample is placed in a flow cell and water is forcedthrough, with the additional oil recovery noted.The sample is now at residual oil saturation, Sor,and the process is repeated with an oil-filledimbibition tube, and then an oil-flooding appara -tus. Separate ratios of spontan eous imbibition tototal saturation change for water, Iw, and oil, Io, aretermed the water and oil imbibition indices,respectively. The Amott-Harvey index is the differ -ence between the water and oil ratios. The result isa number between +1 (strongly water-wetting)and –1 (strongly oil-wetting).

In a US Bureau of Mines (USBM) test, acentrifuge spins the core sample at stepwise-increasing speeds.24 The sample starts atirreducible water saturation, Swirr, in a water-filled tube. After periods at several spin rates,the sample reaches residual oil saturation, Sor,and it is placed into an oil-filled tube for anotherseries of measurements. The areas between eachof the capillary-pressure curves and the zerocapillary-pressure line are calculated, and thelogarithm of the ratio of the water-increasing tooil-increasing areas gives the USBM wettabilityindex.25 The measurement range extends from+∞ (strongly water wetting) to –∞ (strongly oilwetting), although most measurement resultsare in a range of +1 to –1. The centrifuge methodis fast, but the saturations must be correctedbecause the centrifuge induces a nonlinearcapillary-pressure gradient in the sample.

It is possible to combine the Amott-Harveyand USBM measurements by using a centrifugerather than flooding with water and oil to obtainthe forced flooding states. The Amott-Harveyindex is based on the relative change in satura -tion, while the USBM index gives a measure of

21. The Bond number is the ratio of gravitational force tocapillary force and is useful in determining equilibriumconditions in thick reservoirs.

22. Jadhunandan PP and Morrow NR: “Effect of Wettabilityon Waterflood Recovery for Crude-Oil/Brine/RockSystems,” SPE Reservoir Engineering 10, no. 1(February 1995): 40–46.

23. Amott E: “Observations Relating to the Wettability ofPorous Rock,” Transactions, AIME 216 (1959): 156–162.Boneau DF and Clampitt RL: “A Surfactant System forthe Oil-Wet Sandstone of the North Burbank Unit,”Journal of Petroleum Technology 29, no. 5 (May 1977):501–506.

24. The USBM wettability index can also be determinedusing a porous-plate method.

25. Donaldson EC, Thomas RD and Lorenz PB: “WettabilityDetermination and Its Effect on Recovery Efficiency,”SPE Journal 9 (March 1969): 13–20.

> Hysteresis in capillary pressure. The primarydrainage (red) and imbibition (black) curves boundthe capillary-pressure behavior. If the direction ofsaturation change is reversed at an intermediatesaturation, Pc will follow an intermediate path(green). Another reversal will take it back to thedrainage curve (yellow). This behavior couldoccur in the middle of a transition zone, or as aresult of oil banking during a waterflood.

Capi

llary

pre

ssur

e

Pc = 0

Water saturation 1000

>Measurement of core wettability. An imbibition cell contains a sample at Swirr in water (left). Expelledoil collects at the top of a graduated tube. A similar cell turned upside down can measure oil imbibition,starting at Sor. In a centrifuge, the graduated tube is at a larger radius than the core for collectingwater (right), and in an opposite configuration to collect oil. The measurements are illustrated on acapillary-pressure curve (center). Spontaneous water imbibition is from S1, which is Swirr, to S2 at zero capillary pressure. The core is waterflooded or spun in a centrifuge, moving along the negativecapillary-pressure curve to S4. Spontaneous oil imbibition is from S4 to S3, and then an oilflood takesthe sample back to S1, assuming there was no wettability change due to flooding. The imbibition indexis the ratio of spontaneous saturation change to spontaneous plus driven saturation change, separatelydetermined for water, Iw, and oil, Io. The Amott-Harvey index is Iw – Io. The USBM index uses the areas under the positive and negative capillary-pressure curves. This index is the logarithm of the ratio of the areas.

Core

Core

Water saturation, fraction

S1 S4S2

S3

+

–

Capi

llary

pre

ssur

e

IW = S2–S1

S4–S1

IUSBM = log IO = S4–S3

S4–S1

IAH = IW–IO

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 51

the energy needed to make the forceddisplacement, making them related but indepen -dent indicators of wettability.

Unfortunately, core wettability can be alteredat any of several handling stages before the corereaches the laboratory, even when steps aretaken to preserve its native wetting state. First, itcan be contaminated by drilling mud. On its tripto surface, temperature and pressure changescan lead to fluid composition changes, possiblycausing asphaltenes and waxes to precipitateand coat pore surfaces. Exposure to oxygen canalter the chemical composition of crude oil,generating surfactants that affect the core.These changes may also occur during storage andlater handling.

The alternative to using preserved core is torestore the core condition. First, a vigorouscleaning renders a core water-wet, then it issaturated in simulated formation brine and aged.Next, it is flooded with crude oil—typically deadoil—and aged for approximately 40 days,typically at reservoir temperature and pressure.More complex methods are available thatpreserve sensitive clays. The assumption is thatthe result of this procedure approximates the in-situ wetting state. However, variations in brine oroil composition between the formation—throughout its history—and the laboratory canaffect the resulting wetting state.

Measurements may also be made withoutusing core material from the formation. Oneexample is the contact-angle test (above). In thistest, a crystal of quartz or calcite is cleaned, or anew mica surface is cleaved, and aged in asimulated formation brine. A drop of crude oilplaced in contact with the surface is aged.

Several methods are used for creating movingcontact lines, from which the water-advancingand receding contact angles are measured. Theassumption in this test is that the crude oil willchange the model surface—under the conditionsof the brine temperature, pH and saltconcentrations—to that of the formation.

Wettability is often inferred from othermeasurements. Strongly water-wet and stronglyoil-wet materials display certain characteristicrelative-permeability curves, but intermediate-wetting and mixed-wetting states are not a simpleextrapolation between the wettability extremes.

No method for measuring wettability gives an absolutely accurate result, which drivesongoing research, as discussed later in “Newsfrom the Laboratory.”

Production in Transition ZonesPredicting the production of oil and water in atransition zone can be difficult when the crudeoil has altered formation wettability aftermigration. In its unperturbed state, a homoge -neous formation would exhibit a smooth transi -tion from dry oil production at the top of thetransition zone, increasing water cut deeper inthe formation, to no oil production at a pointabove the free-water level.

Unfortunately, drilling a well perturbs thefluid distributions in the near-well region unlessthe well is drilled underbalanced. Drilling-mud

filtrate invasion into a formation can alter thenear-well saturations, affecting shallow-readingwell logs. It also can increase near-well formationpressure in a process termed supercharging.26

Measuring pressure gradients helps evaluatereserves and productivity. In the oil zone, oildensity sets the pressure gradient; in the waterzone, water density controls it. However, filtrateinvasion can produce anomalous formationpressure measurements that, if misinterpreted,might condemn a prospect. Particularly trouble -some for interpretation are gradients indicativeof water but positioned high above the free-waterlevel, substantial shifts in pressure potentialsbetween the lower and upper parts of thetransition zone that can result in negativepressure gradients, and gradients implying an oildensity different from one that would normallybe expected.

Significant anomalies in the gradients of tran -sition zones of homogeneous limestone reservoirsare often found in the Middle East. In some ofthese formations, it is even possible to produce oilfrom zones in which both the pressure gradientand formation resistivity indicate a water zone.Schlumberger studied these phenomena using anECLIPSE 100 finite-difference numerical-flowsimulator. The engineers modeled drilling-fluidinvasion and the effect of hysteresis in relative-permeability and capillary-pressure curves onresulting near-well pressure and water cut.27

52 Oilfield Review

> Contact-angle measurement. Crystalsrepresentative of pore surfaces are aged insimulated formation brine. After an oil drop istrapped between the crystals, the system isaged again. Then, the bottom crystal isdisplaced. Oil moves onto a water-wet surface(lower left) providing a water-receding contactangle (θr). Water moves onto the surface aged in contact with oil (lower right) providing awater-advancing contact angle (θa).

Water-recedingcontactangle, θr

Aged region

Water-advancingcontactangle, θa

> Scanning curves for an intermediate-wet carbonate. Hysteresis between the primary drainage (red)and imbibition (black) curves can be represented by a series of scanning curves (gold). Eachscanning curve represents a different starting saturation point on the drainage or imbibition curve,which would correspond to different heights in the transition zone.

Capi

llary

pre

ssur

e, p

si

Water saturation, fraction

0 0.2 0.4 0.6 0.8 1.0–10

0

10

20

30

40

50

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 52

Summer 2007 53

The wetting character, being dependent onoriginal oil saturation, varies with height, withthe greatest amount of oil-wet rock surface at thetop of the transition zone and the water-wet stateat the bottom. One manifestation of this situationis a variation with depth of capillary entrypressure, or threshold pressure, for water. In thiscase, above a water-base mud-filtrate saturationof 30%, the threshold pressure becomessignificant: for the conditions of this model it wasabout 6 psi [40 kPa], similar to the value in theoil zone. The hysteresis in drainage andimbibition capillary pressure also depends on theinitial saturation at each height. In the model,this hysteresis is represented by a series ofcapillary-pressure curves termed scanningcurves (previous page, bottom).

There are three features of pressuremeasurements in these limestone reservoirs thatthe model sought to explain (right). First, thepressure gradient has an inflection with largecurvature, which is not a result of differentialsupercharging. Second, significant amounts of oilcan be produced below the inflection, in the zonewith a water-like gradient. Finally, just above theinflection, the gradient decreases.

The study had several findings. The first wasthat the combination of all these features is bestexplained by the formation being mixed-wet,with the saturation and wetting characteristicsdescribed above. Using a water-wet assumptionin the model did not result in the anomalies. Theanomalies are seen in the simulations when thedrilling mud used is water-base, but not when itis oil-base.

The production of oil below the inflection isalso a result of the original saturation profile,according to the results of this model (right). Useof the scanning curves leads to predicting lowerresidual oil saturation when the initial watersaturation is higher. Thus, oil at the bottom of thetransition zone may have a low initial saturation,but some of it remains mobile because of itssaturation history.

When faced with these apparent contradic -tions, an operator needs to know where theoil/water contact is within a transition zone, howmuch mobile oil and water are in the zone, andhow these fluids flow. Such issues may beaddressed to a certain degree with logging andformation tester data, and can be improved aftermatching observed measurements in single-wellsimulations. Resistivity logs will help identifylikely transition zones and possible locations ofcontacts. Density and neutron logs help deriveporosity and show similar locations of permeablelithology, which are then used to choose zones forsubsequent wireline formation tests. Wireline

26. Phelps GD, Stewart G and Peden JM: “The Analysis ofthe Invaded Zone Characteristics and Their Influence onWireline Log and Well-Test Interpretation,” paper SPE13287, presented at the SPE Annual Technical Conferenceand Exhibition, Houston, September 16–19, 1984.

27. Carnegie AJG: “Understanding the Pressure GradientsImproves Production from Oil/Water TransitionCarbonate Zones,” paper SPE 99240, presented at theSPE/DOE Symposium on Improved Oil Recovery, Tulsa,April 22–26, 2006.

> Transition-zone anomalies in Middle East carbonates. Pressure measurements (Track 1) in manyMiddle East carbonate transition zones have three unusual aspects: an inflection with largecurvature; a decrease in gradient just above the inflection; and potential production of significantamounts of oil below the inflection. In this case, the pressure measurements are unlikely to beaffected by supercharging, since the mobility is high, and the resistivity log indicates increasing oilsaturation moving upward through this zone (Track 2).

Formation pressure,psi

TVD,m

Drawdown mobility, mD/cpX,600 100X,500 1

Resistivity,10ohm.m0.1

Oil gradient

X,X40

X,X50

X,X60

X,X70

Decrease in gradientInflection

Water gradient

>Matching pressure anomalies. An ECLIPSE reservoir simulator can matchthe pressure anomalies using parameters typical of Middle East carbonatereservoirs and assuming mixed-wet conditions. Prior to invasion by drilling-mud filtrate, the reservoir pressure profile has distinct oil and water gradients(black). With hysteresis scanning curves in the model, the pressuredecreases above the inflection, matching observations (green). Withouthysteresis, the gradient above the inflection does not decrease (red).

Dept

h, ft

Pressure, psi

Oil gradient

4,560 4,580 4,600 4,620 4,640 4,660 4,680 4,700

9,292

9,212

9,132

9,052

8,972

8,892

Reservoir pressureFiltrate pressure (2 days static), no scanning curveFiltrate pressure (2 days static), scanning curve

Water gradient

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 53

formation testers can acquire openhole filtratepressures, formation permeabilities, fluidpressures, oil densities and fluid samples.28 NMRlogging can be used to discriminate pore types tohelp refine the points for making formationtester measurements.29

Detecting Water Zones in a Mixed-Wet CarbonatePetroleum Development Oman (PDO) operatesan onshore field producing from the Shuaibaformation, a Cretaceous limestone with porosityaround 30%. The field has been on production formore than 35 years, and recent infill drilling ofhorizontal wells exhibited logging anomalies thatpetrophysical experts sought to clarify. Somewells produced 100% water, even though the oilsaturation determined from resistivity was inexcess of 50%, which historically had been usedas the limiting value for water-producing zones.PDO suspected some intervals had been flushedwith water, which had not been seen in well logsbecause of hysteresis in electrical properties dueto a mixed-wetting character.30

In a mixed-wet formation, resistivity fromlogs may not provide accurate saturation esti -mates, depending on the saturation history.There are two saturation conditions to consider:

with original fluids and after waterflooding. Oiland brine in the original saturation configurationresulted from oil migration into a water-wetformation. The Archie saturation exponent, n,used to convert from resistivity to saturation istypically about 2; in this case it was 1.8.

However, laboratory studies have shown thatthe formation is mixed-wet, and that its Archieexponent is different when the water saturationincreases compared with when it decreases(above). At an increasing water saturation abovethe 50% cutoff value, the exponent n equals 4,leading to a significant difference in therelationship between resistivity and saturation.

With this insight, PDO sought a method thatwould discriminate zones that had beenwaterflooded from those that were in the originalstate with high oil saturation. They achieved thisby combining two logging methods for deter -mining saturation: resistivity and pulsed-neutroncapture (PNC) cross section, sigma.

Resistivity was measured as part of a stan -dard LWD suite, and the pulsed-neutron capturedevice was in an RSTPro Reservoir SaturationTool that was pumped inside the drill pipe to thebottom of the borehole. The two tools provideindependent measures of water satura tion. TheRSTPro tool could also be used while stationaryto measure water flowing in the well annulus; in

a second pass of the tool, this WFL Water FlowLog measurement identified inflow of fluids intothe borehole at a series of stations.

This approach is possible because the wells inthis field are drilled underbalanced. In under -balanced drilling (UBD), the wellbore pressurewhile drilling is kept below the formation pres sure.UBD avoids flushing the near-well region withdrilling fluid: a distinct advantage for saturationmeasurement. Formation fluid inflow mixes withthe drilling fluid, which in this case was crude oilfrom a neighboring field. The only source of waterin the wellbore annulus was the formation.

In this favorable environment of high-salinityformation water and high porosity, the accuracy ofthe oil saturation determined by both sigma andresistivity is about 5 to 7% of the pore space. Whencomparing the two logs, a 10% difference was usedas a conclusive indicator of saturation anomaly.

PDO logged 11 horizontal wells that weredrilled underbalanced. Some wells produced onlyoil, and the resistivity and sigma logs matchedwithin the 10% criterion. They examined twowells with no water production: the averagedifference between the methods was 0.1 and0.2 saturation units, with a standard deviation of4.5 saturation units. This agreement gave PDOconfidence in the approach.

54 Oilfield Review

> Calculating Archie’s exponent from resistivity index (RI) for a Shuaiba carbonate formation. As coresamples are drained (gold symbols), the cores behave as if they were water-wet. The Archie saturationexponent, n, is around 1.8 (dashed black line), given by the negative slope of the line on this logarithmicplot. For a water-imbibition process (blue symbols), the behavior departs significantly from the drainagecase, with n equal to 4 or greater for water saturation above about 50% (dashed blue line). A curveindicating imbibition behavior is included to guide the eye (solid blue). For an RI of 10, this representsan interpreted saturation difference of about 25 saturation units.

Resi

stiv

ity in

dex

(RI),

dim

ensi

onle

ss

Water saturation, fraction

n = 1.8 n = 4

0.1 11

10

100

Sample 1 drainageSample 4 drainageSample 6 drainageSample 1 imbibitionSample 3 imbibition

Sample 2 drainageSample 5 drainageSample 7 drainageSample 2 imbibitionImbibition

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 54

Summer 2007 55

The logging results were used directly inmaking completion decisions for some wells,such as Well E (below). This flank well had a TDonly 350 m [1,150 ft] away from a water-injectionwell. As soon as the well penetrated the reservoir,it started producing dry oil at a low rate. Asdrilling continued, water production began andincreased rapidly with further drilling. Two morewater inflow zones were found, with no otherzones flowing oil.

Without the additional information obtainedthrough the new method, several zones wouldhave been completed in this well, and it wouldhave been a prolific water producer. Instead, PDOabandoned the well beyond 1,850 m [6,070 ft] MDand left an openhole completion between 1,775and 1,825 m [5,824 and 5,988 ft] MD, with thepossibility of water shutoff later. A postcompletion

production test delivered a gross rate of 225 m3/d[1,415 bbl/d] with a 50% water cut.31

PDO feels this method, combining UBD andresistivity, sigma and WFL logging, provides ahigh-confidence approach to identify water-producing intervals in this mixed-wet formation.The basis of the problem was resistivity hysteresisthat is dependent on the most recent displacingphase. The key to the solution was combiningresistivity- and sigma-based logging in an UBDenvironment in which there is no invaded zone, soshallow- and deep-reading measurements areboth measuring undisturbed formation.

Flow Across FracturesChalks in the North Sea can vary in wettabilityfrom strongly water-wet to intermediate-wet.32

Since many of these chalks are fractured and

the fields undergo waterflooding, researchers atthe University of Bergen, Norway, investigatedthe effect of wettability on flow throughfractured chalk.33

A block of outcrop chalk approximately 20 cmlong by 10 cm high by 5 cm thick [7.9 by 4 by 2 in.]was tested in the original condition and comparedwith a similar block that had been aged in crudeoil. Water-imbibition tests on plugs from the samematerial and treated the same as the slabs had Iw values of 1 and 0.7, respectively, indicatingstrongly and moderately water-wet conditions.Water saturation was determined in these tests by2D nuclear tracer imaging, with a sodium-22[22Na] tracer in the water phase. In addition, afterthe flow tests, plugs were cut from the blocks forimbibition tests to confirm wettability.

28. Carnegie, reference 27.29. Gomaa N, Al-Alyak A, Ouzzane D, Saif O, Okuyiga M,

Allen D, Rose D, Ramamoorthy R and Bize E: “CaseStudy of Permeability, Vug Quantification, and RockTyping in a Complex Carbonate,” paper SPE 102888,presented at the SPE Annual Technical Conference andExhibition, San Antonio, Texas, September 24–27, 2006.

33. Graue A and Bognø T: “Wettability Effects on OilRecovery Mechanisms in Fractured Reservoirs,” paperSPE 56672, presented at the SPE Annual TechnicalConference and Exhibition, Houston, October 3–6, 1999.Graue A, Viksund BG, Baldwin BA and Spinler EA:“Large Scale 2D Imaging of Impacts of Wettability on Oil Recovery in Fractured Chalk,” paper SPE 38896,presented at the SPE Annual Technical Conference andExhibition, San Antonio, October 5–8, 1997.

30. Gauthier PJ, Hussain H, Bowling J, Edwards J andHerold B: “Determination of Water-Producing ZonesWhile Underbalanced Drilling Horizontal Wells—Integration of Sigma Log and Real-Time ProductionData,” paper SPE 105166, presented at the 15th SPEMiddle East Oil and Gas Show and Conference, Bahrain,March 11–14, 2007.

31. Gauthier et al, reference 30.32. Andersen, reference 17.

> Comparison of sigma and resistivity logging for saturation. Shuaiba horizontal Well E was drilled underbalanced. Both resistivity-based saturation measure- ments (black) and pulsed-neutron saturation measurements (green) were calculated (Track 3). The resistivity measurement beyond the typical 50% cutofflevel, where oil production is expected, is shaded (red). Without the new method, these zones would be completed, but the WFL results (Track 5) showwater influx in three zones beginning at 1,750 m [5,740 ft]. Dry oil flows only near the heel of the well above about 1,750 m. The water-influx zones correspondto regions with large differences between the two saturation measurements, indicating the log differences are a good discriminator of a saturation anomaly.

Dept

h,m

1,70

0

1,80

0

1,90

0

2,00

0

2,10

0

2,20

0

Gam

ma

Ray

gAPI

050

Poro

sity D

ensit

y0

0.4

Oil S

atur

ation

(Res

istivi

ty)Ps

eudo

-inflo

w

100% oil, low rate

Rapidly increasing waterproduction, high rate 100% water 100% water

01

Poro

sity

(PN

C)0

0.4

Oil S

atur

atio

n (P

NC)

01

fract

ion

1,470

1,450

m

040

m/mi

n

True V

ertic

alDe

pth S

ubse

a

Wat

erVe

loci

ty

fract

ion

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 55

Both slabs were sawn into three pieces, withthe saw cuts representing fractures (left). Theblocks were butted together, except for one cut,where the chalk pieces were separated by 2 mm[0.8 in.]: this represented an open fracture,while the other cuts represented closedfractures. The waterflood tests began with thematerial at Swirr.

The strongly water-wet blocks filled to 1-Sor

before the front crossed the fractures, regardlessof whether the chalk pieces abutted. The floodfront in the moderately water-wet material, incontrast, crossed closed fractures almost aseasily as it moved though the adjacent intactmaterial, and thus transferred a viscous pressuregradient into the adjacent matrix block.However, the open fracture was more of a barrierthan the closed fractures in the moderatelywater-wet system.

Bergen researchers examined the fracture-bridging phenomenon in a second set of tests.34

Chalk core plugs from the same outcrop were cutto 3.8-cm [1.5-in.] diameter. To alter wettability,crude oil was continuously flooded through theplugs for an extended period.35 For each wettingcondition, two core plugs were placed in serieswith a 2-mm [0.4-in.] spacer between theadjacent ends to represent an open fracture(next page). A magnetic resonance imaging(MRI) tomography method imaged the fluid-saturation distribution in the fracture at the endof the upstream core. Additional scans produceda saturation profile along the core length.

56 Oilfield Review

Length, cm

Height, cm

108

64

2

12

15

913

17Length, cm

Height, cm

108

64

2

12

15

913

17

Length, cm

Height, cm

108

64

2

12

15

913

17Length, cm

Height, cm

108

64

2

12

15

913

17

Length, cm

Height, cm

108

64

2

12

15

913

17Length, cm

Height, cm

108

64

2

12

15

913

17

Length, cm

Height, cm

108

64

2

12

15

913

17Length, cm

Height, cm

108

64

2

12

15

913

17

Length, cm

Height, cm

108

64

2

12

15

913

17Length, cm

Height, cm

108

64

2

12

15

913

17

Length, cm

Height, cm

108

64

2

12

15

913

17Length, cm

Height, cm

108

64

2

12

15

913

17

0.9 to 1.00.8 to 0.90.7 to 0.80.6 to 0.70.5 to 0.60.4 to 0.50.3 to 0.40.2 to 0.3< 0.2

SW

Bloc

k he

ight

, cm

Block length, cm

Inlet

A

C

B

109876543210

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Measured grid

Closed fractures

2-mm open fracture

Outlet

MWWWW

1

2

3

4

5

6< Waterflooding a fractured chalk block. An outcrop chalkslab was cut into three pieces and reassembled (bottom).Part A was abutted to both parts B and C, representingclosed fractures, but a 2-mm gap representing an open fracture separated part B from part C. The scan maps (top)indicate saturation at increasing injected volumes from top to bottom for a strongly water-wet (WW) slab (left sequence)and a moderately water-wet (MWW) slab (right sequence).The sharp flood front in WW Image 2 shows that part Afilled to its maximum water saturation, which is 1-Sor, beforewater crossed the closed fracture, but MWW Image 2shows water already in part B. Again, in WW Image 4 thereis a sharp front at the fracture, and then part C fills first fromthe open fracture and then from both A and B in the WWImage 5. In MWW Images 3 and 4, part C filled across theclosed fracture from A before filling from the open fracture.

59819schD7R1:59819schD7R1 10/12/07 3:40 PM Page 56

Summer 2007 57

The cores at Swirr were waterflooded. The lessstrongly water-wetting the material, the soonerwater appeared in the open fracture. For themoderately water-wet case, beads formed at thefracture face that bridged the 2-mm openfracture. With additional flow, the bridging beadsexpanded and eventually filled the fracture. Incontrast, the strongly water-wet material achieveda high saturation throughout the upstream corebefore any water entered the fracture, and then itfilled the fracture from the bottom up. Only thendid water begin flowing into the downstreamcore. A wider fracture gap of 3.5 mm [0.14 in.]prevented most bridging, even in the moderatelywater-wet case, with most water filling from thebottom of the fracture upward.

These results showed that wetting-phasebridges form through a combination of viscousforces, which control the growth of the waterdroplets making up the bridge, and interfacialtension between the water and oil phases, whichcontrols the droplet contact angle. Thus, openfractures in a chalk reservoir will have differenteffects on waterflood efficiency, depending onthe wettability of the chalk matrix.

News from the LaboratoryLaboratory techniques beyond the simple Amott-Harvey imbibition test have the potentialto expand our understanding of wettability.Surface examination by atomic-force microscopyand the fractured chalk blocks examined by 2Dnuclear tracer imaging are but two examples ofcurrent laboratory techniques. Many othertechniques have been used, and updates of theseand other new approaches are being tested intoday’s laboratories.

For example, although there is variation inboth water-wet and oil-wet rocks, the Archiesaturation exponent tends to be greater for oil-wet rocks. Recent work using ideas from

34. Aspenes E, Graue A, Baldwin BA, Moradi A, Stevens Jand Tobola DP: “Fluid Flow in Fractures Visualized byMRI During Waterfloods at Various WettabilityConditions—Emphasis on Fracture Width and FlowRate,” paper SPE 77338, presented at the SPE AnnualTechnical Conference and Exhibition, San Antonio,September 29–October 2, 2002.Graue A, Aspenes E, Moe RW, Baldwin BA, Moradi A,Stevens J and Tobola DP: “MRI Tomography ofSaturation Development in Fractures During Waterfloodsat Various Wettability Conditions,” paper SPE 71506,presented at the SPE Annual Technical Conference andExhibition, New Orleans, September 30–October 3, 2001.

35. Graue A, Aspenes E, Bognø T, Moe RW and Ramsdal J:“Alteration of Wettability and Wettability Heterogeneity,”Journal of Petroleum Science and Engineering 33, no. 1–3 (April 2002): 3–17.Aspenes E, Graue A and Ramsdal J: “In-Situ WettabilityDistribution and Wetting Stability in Outcrop Chalk Agedin Crude Oil,” Journal of Petroleum Science andEngineering 39, no. 3–4 (September 2003): 337–350.

> Bridging across an open fracture. Two core samples in a waterflooding cell were separated by a 2-mm gap. One was moderately water-wet (MWW, upper sequence) and the other strongly water-wet(WW, lower sequence). Saturations at various pore-volumes (PV) injected were measured along thecore (plots) and in 2D cross section at the upstream core face (photographs) using MRI. The MWWcore had earlier breakthrough, with water forming beads on the core face. The beads coalesced andeventually bridged the gap (inset schematic). Saturation profiles along the core length show thatwater is transported across the open fracture before Sor is reached. In the WW core, saturation scansshow that no water enters the open fracture before the upstream core reaches Sor. The drops on thecore face did not bridge across the gap, until the gap filled from the bottom up (inset schematic).

Wat

er s

atur

atio

n, %

PV

Length, cm0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0

0

10

20

30

40

50

60

70

80

90

100

Wat

er s

atur

atio

n, %

PV

Length, cm

0.20 PV injection0.15 PV injection0 PV injection 0.34 PV injection0.25 PV injection0.23 PV injection

0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3.0 3.3 3.6 3.9 4.20

10

20

30

40

50

60

70

80

90

100

0 PV injection 0.23 PV injection 0.28 PV injection 0.41 PV injection 0.63 PV injection

0 PV injection0.23 PV injection0.28 PV injection

0.41 PV injection0.63 PV injection

0 PV injection0.15 PV injection0.20 PV injection

0.23 PV injection0.25 PV injection0.34 PV injection

0 10 20 30 40 50 60

0 10 20 30Increasing water saturation

40 50 60

WW

WW

MWW

MWW

Increasing water saturation

59819schD7R1:59819schD7R1 9/18/07 8:30 AM Page 57

percolation theory provides a new approach tothe resistivity and saturation relationship. As anexample, one of these models introduces onlytwo parameters.36 One is an exponent, similar tothe Archie saturation exponent n. The other newparameter is the water connectivity correctionindex, which can be related to the oil-wetfraction of the pore surfaces. When this index iszero, the model reduces to Archie’s relationship.The new model matches the relationship seen incore measurements on oil-wet limestones.37

Archie’s equation uses resistivity, which is aDC, or zero-frequency, measurement. At highfrequency, materials exhibit a complex dielectricresponse, including both conductivity—theinverse of resistivity—and permittivity. Measure -ments of the formation permittivity are sensitiveto the formation water content because atambient conditions the permittivity of water is atleast an order of magnitude higher than thepermittivity of oil or the rock matrix. When totalformation porosity is known, water saturationcan be determined directly, avoiding the need foroften unknown cementation and saturationexponents in the Archie equation, which is usedto interpret resistivity measurements.

Dielectric measurement interpretationrequires that a relationship be establishedbetween the dielectric properties of rocks andtheir constituents. Multiple mixing models havebeen proposed to predict rock’s dielectric con -stant based on its volumetric composition.Experi mental data obtained on carbonate rockssaturated with both oil and brine showed that acomplex refractive index law (CRI) worked betterthan other mixing laws at a frequency of 1 GHz.38

However, permittivity is also influenced byfactors other than mineralogy and water content,especially at lower frequencies (above left).Although CRI is the best simple mixing model at1 GHz, it fails to accurately reconstruct dielectricand permittivity dispersion of rocks over a widefrequency range. A new model that includes rocktexture matches rock dielectric properties over awide frequency range more successfully.39 Thisnew model has an average or backgroundbehavior described by the CRI model, thenincorporates ellipsoidal grains and pores toreflect the influence of texture on dielectricdispersion (left).

Pores, grains and oil inclusions can be repre -sented in a simple way as oblate spheroids—ellipsoids with two longer axes of equal length.An advantage of using ellipsoids is that the modelcan be calculated analytically. One additionalgeometrical parameter is added for each phase:

58 Oilfield Review

> Permittivity dispersion for two carbonate rocks. The two brine-saturatedsamples with similar mineralogy and porosity have a similar permittivity at 1 GHz. The CRI model (black) matches dispersion for Carbonate 2, but rocktextural differences result in separation of the lower frequency response forCarbonate 1.

Rela

tive

perm

ittiv

ity

Frequency, Hz108 109

50

10

15

20

25

30

35

40

45

Carbonate 1Carbonate 2CRI

> The textural model. Water-wet pores filled with oil and water (top left) arerepresented in the textural model as randomly distributed oblate spheroidsplaced in a background of a CRI medium (top right). In an oil-wet rock, oil is incontact with the grains and surrounds conductive brine (bottom left). Brine ispredominantly situated in the center of the pores. In the textural model, this isrepresented as spheroids with oil surrounding water (bottom right).

Oil

Water-wet

Oil-wet

Brine (water) Rock grains CRI background

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 58

Summer 2007 59

the aspect ratio, or the ratio of the long to theshort axis of the oblate spheroid. Rocks withthinner insulating regions—flatter grains withhigher aspect ratio—exhibit greater dielectricand conductivity dispersion (right).

Wettability influences the rock dielectricresponse.40 It strongly affects the spatialdistribution of conductive and nonconductivephases—brine and hydrocarbon, respectively—within the pore space and, therefore, the rock’sdielectric properties. In a strongly water-wetrock, the hydrocarbon phase is predominantly inthe center of the pores surrounded by theconductive brine. The fluid-phase distribution isthe opposite in a strongly oil-wet sample, withthe hydrocarbon phase next to the pore walls.This reverse distribution of the conductive andnonconductive phases has several effects. In oil-wet rocks, the conductive brine phase does notform a continuous, connected network withincreasing brine isolation as the oil wettabilityincreases. This leads to a large decrease in rock conductivity.

Another laboratory technique that has aconsiderable potential to aid in characterizingwettability and pore geometry is NMR.41 The NMRsignal is a measure of the degree of relaxation ofmagnetic moments after an initial polarization.42

Fluids in direct contact with a rock surfaceundergo enhanced relaxation because of thepresence of paramagnetic ions or magnetic impur -ities on the rock surface.43 The wetting preferenceof the surface determines which of two availablefluids will be in contact, and therefore, which willbe influenced by the surface.

When a single fluid phase is present in therock, the relaxation time, or T2 distribution, isdominated by surface relaxation effects.However, when two phases are present, the NMRresponse can vary considerably, depending onthe rock wettability. After drainage with alaboratory oil, the rock remains water-wet with aremnant layer of water coating the rock surface.

36. Montaron B: “A Quantitative Model for the Effect ofWettability on the Conductivity of Porous Rocks,” paperSPE 105041, presented at the 15th Middle East Oil andGas Show and Conference, Bahrain, March 11–14, 2007.

37. Sweeney SA and Jennings HY: “The Electrical Resistivityof Preferentially Water-Wet and Preferentially Oil-WetCarbonate Rock,” Producer’s Monthly 24, no. 7 (1960):29–32.

38. Seleznev N, Boyd A, Habashy T and Luthi S: “DielectricMixing Laws for Fully and Partially Saturated CarbonateRocks,” Transactions of the 45th SPWLA Annual LoggingSymposium, Noordwijk, The Netherlands, June 6–9,2004, paper CCC.

39. Seleznev N, Habashy T, Boyd A and Hizem M:“Formation Properties Derived from a Multi-FrequencyDielectric Measurement,” Transactions of the 47thSPWLA Annual Logging Symposium, Veracruz, Mexico,June 4–7, 2006, paper VVV.

40. Bona N, Rossi E and Capaccioli S: “Electrical Measure -ments in the 100 Hz to 10 GHz Frequency Range forEfficient Rock Wettability Determination,” SPE Journal 6,no. 1 (March 2001): 80–88.

41. Allen D, Flaum C, Ramakrishnan TS, Bedford J,Castelijns K, Fairhurst D, Gubelin G, Heaton N, Minh CC,Norville MA, Seim MR, Pritchard T and Ramamoorthy R:“Trends in NMR Logging,” Oilfield Review 12, no. 3(Autumn 2000): 2–19.Chen J, Hirasaki GJ and Flaum M: “NMR WettabilityIndices: Effect of OBM on Wettability and NMR

Responses,” Journal of Petroleum Science andEngineering 52, no. 1–4 (June 2006): 161–171.

42. The typical sequence for polarization and detection isthe Carr-Purcell-Meiboom-Gill (CPMG) pulse-echomethod: Meiboom S and Gill D: “Modified Spin-EchoMethod for Measuring Nuclear Relaxation Times,” TheReview of Scientific Instruments 29, no. 8 (1958): 688–691.

43. Brown RJS and Fatt I: “Measurements of FractionalWettability of Oilfield Rocks by the Nuclear MagneticRelaxation Method,” Transactions, AIME 207 (1956): 262–264.Foley I, Farooqui SA and Kleinberg RL: “Effect ofParamagnetic Ions on NMR Relaxation of Fluids at SolidSurfaces,” Journal of Magnetic Resonance, Series A 123,no. 1 (November 1996): 95–104.

> Dispersion in a textural model. Two series are shown here, one withvarying aspect ratios (AR) and the other with varying wettability indices (WI).The water-wet case (WI = 1.0) with spherical grains (AR = 1) is the base case(red). In one series, the grain AR increases to 10 (green) and 100 (blue) whileremaining water-wet. As grains become flatter, or AR increases, rockconductivity (bottom) decreases significantly and relative permittivity (top)increases. This provides a crucial link between dispersion properties androck texture. A second series maintains spherical grains (AR = 1), butwettability changes from water-wet (red) to intermediate-wet (orange) to oil-wet (black). Increasing oil-wetting character leads to a strong decrease inthe rock conductivity. The wettability index here is based on the fraction ofoil-wet pores relative to the total pore volume, and ranges from 1 for stronglywater-wet, to –1 for strongly oil-wet. The pores in these models are spherical(AR = 1), with porosity of 30%, water saturation of 80%, and brineconductivity of 5 S/m.

Cond

uctiv

ity, S

/m

Frequency, Hz108 109

0.5

0.4

0.3

0.2

0.1

0

Flatter grainsMore oil-wet

Rela

tive

perm

ittiv

ity

Frequency, Hz108 109

30

10

15

20

25 Flatter grains

More oil-wet

110

100

1 0WI

Grai

n AR

–0.9

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 59

The T2 oil peak value is close to bulk value(below). In contrast, drainage with oil-base mud(OBM) results in shorter T2 values than for bulkOBM. This is an indication that the OBM, which

was made by adding substances, includingasphaltenes, to the same laboratory oil, iscontacting the rock surface and, therefore,wetting the core.

The wetting character can vary with poresize, often with microporosity remaining water-wet. Therefore, determining the microporousfraction can be crucial for analyzing formationsexhibiting complex wetting character. In thefield, T2 NMR logs are commonly used forestimating microporosity fraction. However, thisapproach can fail due to variations in surfacerelaxivity or pore geometry.

A different method, called restricted diffu -sion NMR, is unaffected by surface relaxivity andis sensitive to pore sizes, connectivity andtortuosity. The diffusion coefficient in a bulkfluid, D0, is a constant that measures how rapidlya concentrated group of molecules diffuses.However, inside a restricted space, such as in thepores of a rock, the diffusion, D, may be reducedfrom the bulk value because molecules arerestricted in their motion by the pore walls.

The diffusion coefficient is determined byanalysis of the NMR echo decay in the presenceof a nonhomogeneous magnetic field.44 Anexample for a carbonate rock shows the normal -ized D/D0 distribution at early diffusion times(below left).45 The distribution has a peak atsmall diffusion coefficients, which correspondsto molecules in restricted microporous space,and a second peak at higher diffusion coeffi -cients closer to D0, which corresponds to those inless restricted areas, or larger pores. Applying anempirical cutoff to the distribution separates themicroporosity: the percentage of the populationat D/D0 values less than the cutoff gives themicroporosity, which is in good agreement withmercury porosimetry data and irreducible watersaturation after centrifugation.

Some of these new methods are providinginput for pore-network modeling, which hasemerged as an effective way to investigate thecapillary, flow and transport properties of porous

60 Oilfield Review

> T2 decay-time distributions. The T2 distribution for a carbonate sample fullysaturated with brine (H2O) (solid black) is shifted to shorter time than thebulk-brine signal (dotted black) due to surface interactions. The brine isreplaced with a brine made from deuterated water (D2O), which has no NMRsignal other than a small amount of residual H2O (green). After the deuteratedsample is flushed with OBM, the peak (solid red) is shifted from the bulk OBM(dotted red), indicating that the OBM wets the rock. The sample was cleanedand prepared again in the deuterated state, then flushed with laboratory oil.The main peak (solid blue) aligns with the bulk-oil signal (dotted blue), andso with laboratory oil, the surface remains water-wet.

Dist

ribut

ion,

por

osity

uni

ts

T2 distributions, s10-4 10-3 10-2 10-1 100 101

0

2

4

6

8

10

12

16

18

20

14

22

Brine (H2O)Brine (D2O)Laboratory oilOBMBulk brine (H2O)Bulk laboratory oilBulk OBM

> Distinguishing microporosity. The mean normalized diffusion coefficient,D/D0, indicates the presence of two populations of diffusing molecules (blue).The measured diffusion coefficient is smaller in micropores (inset, left)because the diffusion path of molecules (red) is more tortuous. The peak atlarge values is a measurement of molecules in larger pores (inset, right) thathave D values closer to the bulk D0 value. The sharper peak at smaller Drepresents molecules in micropores. The area under the curve to the left ofthe fixed cutoff line (black), relative to the total area under the curve, is themicroporosity: 44% of the pore space in this example.

Nor

mal

ized

ampl

itude

D/D0

Microporosity

0 0.2 0.4 0.6 0.8 1.00

0.5

1.0

1.5

44. Stejkal EO and Tanner JE: “Spin Diffusion Measurements:Spin Echoes in the Presence of a Time-Dependent FieldGradient,” Journal of Chemical Physics 42, no. 1 (1965):288–292.

45. Hürlimann MD and Venkataramanan L: “QuantitativeMeasurement of Two-Dimensional Distribution Functionsof Diffusion and Relaxation in Grossly InhomogeneousFields,” Journal of Magnetic Resonance 157, no.1(July 2002): 31–42.

46. Valvatne PH and Blunt MJ: “Predictive Pore-ScaleModeling of Two-Phase Flow in Mixed Wet Media,”Water Resources Research 40, no. 7, W07406 (2004):doi:10.1029/2003WR002627.

47. Picard G and Frey K: “Method for Modeling Transport ofParticles in Realistic Porous Networks: Application tothe Computation of NMR Flow Propagators,” PhysicalReview E 75 (2007): 066311.

48. Knackstedt MA, Arns CH, Limaye A, Sakellariou A,Senden TJ, Sheppard AP, Sok RM, Pinczewski VW andBunn GF: ”Digital Core Laboratory: Properties ofReservoir Core Derived from 3D Images,” paper SPE87009, presented at the SPE Asia Pacific Conference onIntegrated Modelling for Asset Management, KualaLumpur, March 29–30, 2004.

59819schD7R1:59819schD7R1 8/31/07 7:09 PM Page 60

Summer 2007 61

media.46 Modeling the pore space with a networkof nodes and bonds enables a numericallyefficient calculation of the flow properties andprovides detailed understanding of processessuch as miscible and immiscible displacementthat are important in enhanced oil recovery.

A pore-network model is an idealized porespace, generally incorporating a pore-scaledescription of the medium and of the physicalpore-scale events. Complex multiphase flow andtransport processes in porous media can besimulated using pore-network models.47 For suchprocesses, pore-network models are faster to runon a computer than other approaches using moreexact models.

Although pore-network models are tradi tion -ally employed in qualitative studies, they have thepotential to become predictive if pore-structureparameters are properly assigned. Significantadvances have recently been made in predictivepore-network modeling, where geologi callyrealistic networks are constructed by analyzing 3Dimages that may be generated by 3D reconstruc -tion of X-ray microtomograms (left).48

However, the resolution of X-ray microto -mograms is currently limited to several micronsand, therefore, a proper description of micro -porosity with pore-network models is a challenge.Several techniques, such as NMR restricteddiffusion, confocal microscopy—useful for opticalimaging of thick specimens—and scanningelectronic microscopy, have potential to extendapplicability of pore-network models to micro -porous rocks.

Pore-network modeling is useful for studyingthe impact of wettability on oil recovery.Petrophysical parameters, such as capillarypressure, relative permeability and resistivity,are calculated under different wetting condi -tions. These conditions are given by contactangles that are randomly assigned based onselected distributions, providing the fluiddistribution and interface configurations in thenetwork for multiple realizations. Quasi-staticdrainage and imbibition simulations can beconducted to examine the result of the wettingconditions (left).

These laboratory techniques point the way tothe future of wettability application. Fieldsaround the world are maturing, and the industrywill extract as much of the hydrocarbon resourceas is economically possible before they areabandoned. All systems will need to be opti -mized to achieve this goal, and that requirescontinued improvement in applying a funda -mental parameter underlying recovery: rockwettability. —MAA

> Pore-network modeling. Microtomogram slices of a carbonate (top left) are partitioned (top middle)into grains (black) and pores (individually colored). Many slices form a 3D microtomogram that isconverted to a pore network (bottom left). A small subset of a model is shown; the pores are not toscale. The network comprises spheres and more complex shapes, such as tubes with triangular crosssections. Displacement from these structures is piston-like (top right). In this water-wet condition, oil is in the middle of the tube and water is at the apexes, as also shown in cross section (middle right).The model can allow oil to touch surfaces and alter the contact angle toward oil-wetting (shading onsurface). In a subsequent waterflood, oil layers may remain in the elements with high value of contactangle (bottom right).

Oil

Oil

Water

Water

> Pore-model results after a wettability shift. A micromodel had an initialdistribution of contact angles between 50° and 60°. In simulations, contactangle in 95% of the network elements shifts to higher values as a result ofwettability alteration. Simulated drainage and imbibition cycles allowedcalculation of the Amott-Harvey (red) and USBM (black) wettability indices,shown as a function of the altered contact angle.

Wet

tabi

lity

inde

x

Contact angle, degrees0 20 40 60 80 100 160120 140

2.5

2.0

1.5

1.0

0.5

0

–0.5

–1.0

Amott-Harvey

USBM

59819schD7R1:59819schD7R1 8/31/07 7:10 PM Page 61