Foams for Controlling Water Production

Copyright 2005, Society of Petroleum Engineers Inc.

This paper was prepared for presentation at the 2005 SPE InternationalSymposium on Oilfield Chemistry held in Houston, Texas, U.S.A., 2 – 4February 2005. This paper was selected for presentation by an SPE Program Committeefollowing review of information contained in a proposal submitted bythe author(s). Contents of the paper, as presented, have not beenreviewed by the Society of Petroleum Engineers and are subject tocorrection by the author(s). The material, as presented, does notnecessarily reflect any position of the Society of PetroleumEngineers, its officers, or members. Papers presented at SPE meetingsare subject to publication review by Editorial Committees of theSociety of Petroleum Engineers. Electronic reproduction, distribution,or storage of any part of this paper for commercial purposes withoutthe written consent of the Society of Petroleum Engineers isprohibited. Permission to reproduce in print is restricted to aproposal of not more than 300 words; illustrations may not be copied.The proposal must contain conspicuous acknowledgment of where and bywhom the paper was presented. Write Librarian, SPE, P.O. Box 833836,Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

AbstractThere is a great interest in the oilproduction industry in relativepermeability modifiers (RPM) that canpreferentially reduce water productionwithout significantly affecting oilproduction. Various hydrophilic polymershave been studied for their RPMproperties. This paper shows the potentialof foam as RPM. Foams stable for a longtime in presence of flowing water weredeveloped using polymeric surfactants.Scale-up calculations show foam lifetimeslong enough in field geometries for theprocess to be practicable. Foam toleranceto residual oil in water producing zoneswas produced by adding anionic surfactantto a nonionic polymeric surfactant.Resistance to flowing oil was, however,found to be unexpectedly high due tostable emulsion generation in porousmedium. A process modification is shown toreduce the severity of this problem.

Introduction

Reducing water production during oilproduction in a mature oilfield is animportant objective for the oil producingindustry. Produced water has a significantimpact on the profitability of the field.Producing water involves a waste of driveenergy in pumping water rather than oil,in addition to loss of oil production.Also, water production can lead to otherproblems such as corrosion of surfaceequipment, need for more separatorcapacity, and other disposal and handlingconcerns.1-3

Several techniques have beendeveloped to overcome this problem. Themost popular involve injection ofpolymer1-3, 6-9 or gel4,5 into the productionwells. These materials preferentiallyreduce water permeability relative to oiland are, hence, referred to as relativepermeability modifiers (RPM). Variousmechanisms have been proposed to explaintheir RPM property such asswelling/shrinkage, fluids partitioningand wall effects.6 Though none of them isuniversally applicable, the hydrophilicpolymers, in particular, are accepted toexhibit RPM behavior due to the walleffect, in which a hydrated layer ofpolymer adsorbs on the rock surface alongpore throats and inhibits water flow.

In this paper, we show the potentialof foam for use as RPM in oil-producingwells. Aqueous foam has properties thatmake it suitable for use as a RPM. Foamreduces water permeability in the regionswhere it is present by trapping gas in thepore spaces. In addition, foam is notexpected to impede flowing oil as aqueousfoam usually collapses in presence of

SPE 93273

Foams for Controlling Water ProductionVikram Bhide, SPE, George Hirasaki, SPE, Clarence Miller, SPE, and Maura Puerto, Rice University; Ian Robb, SPE, and Lewis Norman, SPE, Halliburton

2 V. BHIDE, G. HIRASAKI, C. MILLER, M. PUERTO, I. ROBB, L. NORMAN SPE 93273

large amount of oil.10, 11 However, foamsfrom conventional low molecular weightsurfactants are unsuitable for thisapplication as they tend to collapserapidly in presence of flowing water,12

presumably, due to surfactant desorptionfrom the air-water interfaces of foam.Hence polymeric surfactants were studiedfor foam generation, as these are notexpected to desorb as easily. In addition,for water-control application, foam has tobe stable to residual oil present in thewater producing zones and foam has tocollapse in presence of flowing oil in theoil zones. These properties were examinedin this study and are addressed in thefollowing sections.

Foam washout stability with polymericsurfactants

Screening tests. To test foam stability inpresence of flowing water in a shortamount of time, a screening test wasdeveloped. The experimental setup (Fig 1)consisted of a foot-long glass column,about an inch in diameter equipped at thebottom with a perforated metal platefitted with a 360-mesh screen. Air wasblown at 10 ml/min from the bottom througha fixed amount of surfactant solutionuntil foam was generated to fill the glasscolumn. About 15 ml of surfactant solutionwas used. The excess surfactant wasdrained and water (or brine) was injectedat a fixed rate of 3 ml/min from the top.The injected water flowed through theplateau border regions of the foam (wherethree foam films meet at 120o), carryingoff the surfactant present in thoseregions. This eventually caused the foamfilms to rupture, giving rise to foamcollapse from the top to the bottom of thecolumn. The time required for the foamheight to decrease from the upper to thelower mark in Fig 1 was denoted as thewashout time. The distance between the twomarks was about 9.5 inches and the volumebetween the marks was approximately 120ml.

Some of the conventional andpolymeric surfactants (triblock copolymersfrom BASF and Dow Chemical, andhydrophobically modified polymers suppliedby Halliburton) that were tested in theabove setup are listed in Tables 1 through3. In Table 1, SDS and C16TAB are purematerials and the rest are commerciallyavailable surfactant solutions that maycontain ingredients other than activematerials. Two types of triblockcopolymers were tested (Table 2) –ethylene oxide-propylene oxide-ethyleneoxide (EO-PO-EO) and ethylene oxide-butylene oxide-ethylene oxide (EO-BO-EO),wherein BO is more hydrophobic than PO.The hydrophobically modified (HM) polymerslisted in Table 3 (HMPA1 and HMPA3) havesome randomly distributed hydrophobicchains attached to a relativelyhydrophilic backbone. However, their exactcomposition and molecular weight are notknown.

The foam washout times in thescreening test are shown in Fig 2. Mostconventional surfactants produced foamthat washed out quickly (< 10 min) exceptfor CS-330. The triblock copolymers, ingeneral, showed much better washoutresistance than conventional surfactants.In presence of salt (5% NaCl), thetriblocks were even better foamers andalso better foam washout stabilizers.F108, with the highest molecular weight of~15,000, also had the highest washout timeof ~150 min. In presence of salt, the HMpolymers also showed excellent foamwashout stability. Foam produced fromHMPA1 required washout time close to 300min. This demonstrates that polymericsurfactants are more suited thanconventional surfactants for producingfoams in reservoirs that can last for along time in flowing water.

Sand pack experiments. Foam was generatedwith some of the above polymericsurfactants in a porous medium to evaluatethe water blocking property of foam. Theexperimental setup that was used is shown

SPE 93273 FOAMS FOR CONTROLLING WATER PRODUCTION 3

in Fig 3. The porous medium consisted of a2-ft long, 5 to 7 darcy sand pack.Pressure drops could be measured acrossfour equal sections of the pack. Foam wasgenerated in-situ by injecting air andsurfactant solution through the sand pack.Strong foam generation, i.e. foam withhigh resistance to flow, and simultaneouspropagation of foam has been observed tobe difficult to achieve in porous mediumin several studies including this one.13-16

A procedure of alternating injection ofsurfactant and air into an initiallysurfactant saturated sand pack was used toachieve both strong foam generation andpropagation. This sequence of air andsurfactant injection was continued till asufficiently high resistance to flow wasdeveloped.

Foam generation in the sand pack isshown in Fig 4a using the example of B20-5000 polymeric surfactant. In this case,after three cycles of air and surfactantinjection, an average mobility reductionfactor (MRF, defined below) of greaterthan 100 was developed.

(1)

Note that the corresponding MRFs reportedin the studies of polymers as RPM aregenerally only about 5 to 10.6 Due to someentrance and exit end effects,13, 15 MRFduring foam generation was usually lowerthan average in the first section andhigher in the fourth section as seen inFig 4a.

Following foam generation,surfactant solution was injected todisplace any flowing air and leave behindonly trapped air in the sand pack. Water(or brine) was then injected in the packand collapse of foam monitored with porevolumes (PV) of water throughput. Washoutof foam from B20-5000 is shown in Fig 4b.Foam was observed to collapse almostsequentially in the first three sectionsof the sand pack. However, MRF in the 4th

section actually increased initiallyduring washout. This is again believed tobe due to a capillary end effect whichcauses foam generation near the exit bythe mechanism of snap-off and traps someof the air from foam washed out in thepreceding sections near the exit. MRF in4th section decreased somewhat after foamin 3rd section washed out but nevercollapsed completely. The reason for thisslow washout in 4th section had to do withslow desorption of polymeric B20-5000 andair compressibility effects and willbecome clearer later in the section oncauses of foam washout. The end effectlength is, however, not significant infield length scales and can be neglectedfor the purpose of measuring foam washout.So, PV for foam washout for only the firstthree sections is considered. It requiredabout 22 PV of 5% NaCl throughput to washout foam from first three sections.

A practical problem with carryingout foam washout at constant rate of 6ft/d, as in Fig 4b, was that it requiredlong time durations (~6 days in Fig 4b).Hence, for most experiments washout wasswitched to a high rate pump from time totime to accelerate washout. This is shownin Fig 4c for B20-5000. Gaps in washout inthe Fig are periods when flow was switchedto higher rates and when pressuretransducers were isolated to keep themfrom going out of range. In this case,about 16 PV were required for foamwashout. Thus, though changing flow ratehad some effect on number of PV, it wasnot very substantial and did not affectany meaningful comparison of surfactants.

In contrast to polymericsurfactants, washout of foam with pureconventional surfactants such as C16TABwas quite rapid. In Fig 5, foam fromC16TAB collapsed in all sections of thesand pack (including 4th) in ~2.5 PV.Washout of foam from commercialsurfactants such as CS-330 required largerthroughput (~10 PV) because of someimpurities in the formulation, asdescribed in the following section. Fig 6


summarizes the foam washout results forseveral conventional and polymericsurfactants. Again, foam from F108required the largest number of PV of brinefor washout. The HM polymers tested inthis study were found to be relativelypoor foamers in sand pack experiments,with maximum MRF during brine flow of 50or lower. The foam films from HMAP1 andHMPA3 appeared to be unstable at thehigher capillary pressures in the sandpack in contrast to the screening testwhere capillary pressures are lower. Yet,the HM polymers show very good foamwashout stability as seen in Fig 6. Thereason that PV for foam washout for HMPA1was not higher than F108 (as in screeningtest) was because of foam dissolutionlimitation in sand pack as explained inthe next section.

The foam washout experiments in sandpack show that the screening testdeveloped was a good qualitative indicatorof foam washout in sand pack. The foamwashout experiments demonstrate that thepolymeric surfactants can produce foamswith much higher foam washout resistancethan most conventional surfactants. Alittle later, the section on scale-up willestimate washout times for fieldapplications.

Reasons for foam washout stability. Asmentioned earlier, the reason foam fromconventional surfactants is unstable towashout is because of desorption ofsurfactants from air-water interface intothe flowing water which destabilizes foamfilms. The desorption of conventional andpolymeric surfactants from air-waterinterface was measured in our research inthe following manner: A pendant bubble wascreated initially in a surfactant solutionand then the surfactant solution wasreplaced with water. The surface tensionof the bubble was monitored as a functionof surfactant concentration around thebubble. The experimental setup is shown inFig 7. More detailed description can befound elsewhere.17 A similar principle has

also been used by Svitova et al.18 tomeasure desorption of surfactants fromair-water and oil-water interfaces. Thedesorption results are summarized in Fig8. The surface tension of the bubble as afunction of replaced cell volumes ofsurfactant solution (and hence surfactantconcentration) is shown in the Fig. It isobserved that pure conventionalsurfactants, such as SDS desorbed easilyfrom the air-water interface as thesurfactant solution was diluted andeventually the surface tension reachedthat of pure water, indicating that allthe surfactant was removed from the air-water interface. For commercial surfactantCS-330, though the surfactant desorbedrapidly, there were other surface activeimpurities in the surfactant formulation(probably solubilized long chain alcohols)that did not desorb from the air-waterinterface. This is represented in the Figby a final surface tension significantlylower than that of pure water (~ 55 mN/m).This explains the higher washout time forfoam from CS-330 in the screening test. Asopposed to the conventional surfactants,most polymeric surfactants did not desorbsignificantly from the air-waterinterface. For example, the change insurface tension of the bubble initiallygenerated in F108 solution was only about5 mN/m. This explains why polymericsurfactants such as F108 and HMPA1 producesuch long lasting foams in presence offlowing water.

However, it was found thatdissolution of gas from foam into theflowing water was a limiting factor onfoam lifetime in sand pack for foam frompolymeric surfactants. In the sand packexperiments described in the previoussection, the water that was injectedduring foam washout was under-saturatedwith air at the inlet pressures of thesand pack (approximately 2 to 3 atm). Thiscan cause foam collapse due to gasdissolution. Using the method ofcharacteristics and ideal gas law, thefollowing expression for PV required for


foam dissolution in completely unsaturatedwater was derived:17

(2)where, Sg is the initial gas saturationbefore washout and H is the Henry’sconstant for the gas in foam. Using thisequation, one gets ~30 PV for foamdissolution in 3 sections for initial gassaturation of 70%. The observed PV formany polymeric surfactants was ~16-25,indicating that foam dissolution has avery important role in foam washout formany polymeric surfactants. This wasfurther verified from the following twoobservations: 1) foam from polymericsurfactants collapsed even when surfactantsolution was injected instead of water ina similar number of PV, and 2) whenadditional air was dissolved into theinjected liquid, foam washout was slowedconsiderably.

Therefore, foam dissolution canexplain why foam from HMPA1 does not lastlonger than foam from F108 as may beexpected from the screening test. Thisalso explains why very little washout offoam was observed in the 4th section withpolymeric surfactants as describedearlier. Near the exit pressure(atmospheric), the flowing water wasalways either saturated or oversaturatedwith air and no washout occurs by gasdissolution.

It must be noted that the aboveexpression for foam dissolution isindependent of pressure (provided Henry’slaw and ideal gas law hold) and foamlifetime is not expected to be a functionof depth of the reservoir. The expressiondescribes the worst case scenario ofunsaturated water. In oilfield reservoirs,one may expect dissolved gases,particularly hydrocarbon gases such asmethane, to be present in the producedwater. Thus, if foam is generated with N2

but reservoir brine contains CH4, some CH4

will be transferred to foam as N2

dissolves, thereby prolonging foam

lifetime. The expression also highlightsthe importance of the type of gas used togenerate foam. Hence, N2 rather than CO2

would be preferable to produce foambecause of its lower solubility in brine.

Scale-up to field geometries. From theabove experimental observations in onedimensional sand packs, the time scalesfor foam washout in radial oilfieldgeometries can be calculated. A parameterm is defined as follows,

m = vf/v, where vf is the velocity of foam frontduring washout and v is the interstitialwater velocity when no foam is present. Itcan be shown that for 1-D sand packexperiments

m = 1/(PV for foam washout).This can be computed from Fig 6 aftertaking into account that PV for washout inFig 6 is for three sections only forpolymeric surfactants.

A schematic of foam in the waterproducing zone is shown in Fig 9. Let rw

be the well radius and re be a positiondeep enough the formation such that changein pressure at that position (Pe) is notvery large at the time scales of washout.Let ri be the initial position of foamfront and rt the position at time t duringwashout. Let Pw be the pressure at thewell and Pt be the pressure at the foamfront at time t. Let the mobilityreduction factor be MRFf in the foamregion and MRFnf in the remaining regionwhere foam has washed out.

Thus,

(3)

using

Darcy’s law (4)Here, (= k/) is the water mobility information, is the porosity and MRFmean isthe harmonic average of MRFs from rw to re


at any given time t. A couple ofsimplifying assumptions can be made to getan analytical solution of (4). In 1-Dexperiments, the foam washout front wasobserved to travel with approximatelyconstant speed (hence, constant m) frominlet to outlet. This is shown for thefoam washout of B20-5000 for the firstthree sections in Fig 10 (from MRF versusPV data from Fig 4b). In addition, only aweak dependence of m on flow rate wasobserved from experiments in Figs 4b and4c. So, m can be considered independentof flow rate and time, and values from 1-Dexperiments can be used. In addition, fora deep enough re, PePw can be consideredapproximately constant (in any case, thiswill only underestimate foam washouttime). These assumptions give thefollowing expression for foam washout time(t):

(5)

Since, rw << ri, we get:

(6)

Foam washout time as a function ofinitial foam penetration (ri) is plottedfor F108 in Fig 11 for two values of MRFf

with the following typical parametervalues: k = 100 mD, = 0.3, rw = 0.5 ft, re

= 1000 ft, PePw = 1000 psi, m = 0.03,MRFnf = 1, MRFf = 50 or 100. The plot showsan approximately quadratic dependence ofwashout time on foam penetration length.Foam lifetime of the order of years can beseen for initial foam penetration of 20 ftor more with MRFf 50. What is moreimportant for field application is thefoam lifetime in relation to time for foamgeneration. Equation (6) can also be usedto compute time for fluid frontpropagation using appropriate values forfront velocity (m) and resistance factor(MRFf). To compute time for foamgeneration, it is assumed that foam isgenerated in the field by injecting oneslug of surfactant followed by one slug ofgas. It can be shown that time forsurfactant solution injection isnegligible compared to gas injection.Hence time for foam generation (tfg)equals time for gas injection to distanceri. That is,

tfg t from (6) with m = 1 and MRFf =100/n (7)

In 2-ft sand pack experiments, MRF duringair injection was about 50% of MRF duringliquid flow after foam generation (i.e. n 2 or MRFf 50 in (7)). Hence, it can beshown that

Time for foam washout (tfw) (2/m) tfg

(8)

For F108, tfw 66 tfg. That is, in thecase of F108, for every day of productiondowntime in generating foam one may expectat least 2 months of reduced waterproduction. Thus, with a few days of foamtreatment, water production can be reducedfor several months using foam from F108.

The MRF of foam in 3-D has beenobserved to be less than that in 1-D19, 20

because of more avenues to bypass achannel blocked by a foam lamella. Themaximum MRF achievable due to foam at 20ft or more in radial geometries may haveto be studied further, but there is some


evidence in literature that it is possibleto have high MRF in field conditions.21, 22

In the use of foam for gas blocking,22 forexample, Blaker et al. have estimatedmaximum foam MRF of 400 for a foam zonelonger than 280 ft.

Foam stability to residual oilThe water producing zones in an oilfieldmay also contain residual amounts of oilfrom past history. Hence, for the use offoam as RPM, it must have tolerance forresidual oil. Foam stability to residualoil in porous medium was tested using theexperimental set-up shown in Fig 12. Foamwas generated in a short sand pack (about1.5” long) by co-injecting air andsurfactant (80% air fraction at 1 atm).The texture of foam generated could beobserved in the transparent plastic tubingat the outflow end. The pressure dropacross the short pack was measured atsteady state and is an indicator of foamstability in the short sand pack. Pressuredrop across the pack when no oil waspresent was compared to pressure drop whenresidual oil was present. To prepare thepack at residual oil saturation, oil wasflowed through a water saturated pack andthen followed by water injection. Residualoil saturation of about 30% was observedin the pack.

Fig 13 shows the steady statepressure drop across the short pack at 80%foam quality (gas fraction) for the caseswith and without residual oil for non-ionic triblock copolymers and anionicsurfactants. Hexadecane was used as oil inthese experiments. It can be seen that thetwo polymeric surfactants tested - F108and B20-5000 - show little foam toleranceto residual oil and hardly any foam wasproduced with residual oil. In contrastboth anionic surfactants CS-330 and SDSproduced stable foams with residual oil.This indicated that the chargedsurfactants are better for producing foamstolerant to oil than unchargedsurfactants. This was further supported bythe entering (E), spreading (S) and

bridging (B) coefficients for these oil-water-surfactant systems as listed inTable 4. The various interfacial tensionslisted were measured using a pendant dropapparatus. The thermodynamic theories offoam stability predict that foam is likelyto be destabilized when any of thesecoefficients is positive and a highermagnitude of a coefficient indicatesgreater tendency for destabilization by aparticular mechanism.23, 10, 11 The positivevalues of all the coefficients for all thesurfactants in the Table show that all areexpected to form foam intolerant to oil.In addition, the magnitudes of thecoefficients for SDS are comparable tothose of F108 or B20-5000. The onlydifference between the stable and unstablesets of surfactants is that the stable setis negatively charged. The chargedsurfactants produce an electrostaticbarrier for the entry of an oil dropletinto an air-water interface and preventfoam destabilization by spreading orbridging mechanisms. This phenomenon hasalso been observed by other researchers.24

Because of the above observations, amixture of F108 and CS-330 was tested toproduce foam with both residual oiltolerance and good washout resistance. Theamount of CS-330 added had to be kept to aminimum to prevent displacement of F108 atthe air-water interface by the fasteradsorbing CS-330. But because of bindingof CS-330 molecules to F108 molecules insolution, a larger amount of CS-330 had tobe used for CS-330 to be available at thesurface. Finally, a 1:1 ratio by weight ofCS-330 and F108 was used. This 0.5% F108 +0.5% CS-330 mixture did produce foam thatwas tolerant to residual oil as seen inFig 13. In addition, foam washoutresistance was comparable to F108 in bothscreening test and 2-ft sand packexperiment. In the screening test, bothF108 + CS-330 mixture and F108 alonerequired close to 150 min for foam washoutwith 5% NaCl. In the 2-ft sand packexperiment, foam from F108 + CS-330mixture washed out in 18-19 PV which was


only slightly lower than 21-25 PV requiredfor F108.

Foam stability to flowing oilAnother criterion for the use of foam asRPM is that it should not offersignificant resistance to oil flow. Forthis to happen, foam should collapse withflowing oil. When oil was injected througha glass column containing foam, foam wasobserved to collapse rapidly within a fewminutes. Table 5 shows the time for foamcollapse with flowing hexadecane in thescreening test compared with foam washouttime with brine. For all the surfactantstested – anionic as well as non-ionic -foam collapsed in a very short timecompared to washout time with water, asdesired for the application.

In the porous media (2-ft sandpack), however, the surfactant present inthe foam led to a very stable oil-in-wateremulsion when oil was flowed through thefoam region. This emulsion produced alarge resistance for flow of oil (MRF>1000). The F108 + CS-330 combination, inparticular, turned out to be a much betteremulsifier than a foamer. Since the flowresistance was due to presence ofsurfactant which produced emulsions, theflow resistance can also be tested whenoil is flowed through a surfactantsolution saturated sand pack, even withoutthe presence of foam. This is depicted inFig 14. The F108+CS-330 combination gavehigher flow resistance (MRF/viscosity)than F108 itself, because the combinationproduced a more stable emulsion. Thisstable emulsion is undesirable for the RPMapplication of foam.

One way to reduce emulsion stabilityis by modification of oil + water +surfactant phase behavior. It has beenobserved by several researchers that therate of emulsion coalescence increases asone approaches the 3–phase region and ismaximum in the Winsor III system.25, 26

F108+CS-330 combination is veryhydrophilic. Hydrophilic-lipophilicbalance (HLB) closer to intermediate

values could be obtained by increasingsalt concentration or by replacing CS-330by a more lipophilic surfactant. This isalso shown in Fig 14 for a couple ofcases. Adding salt shifts the HLB of thesystem to more lipophilic values. However,for F108+CS-330 combination, adding even alarge amount of salt did not produce aWinsor II or Winsor III system because ofthe very hydrophilic nature ofsurfactants. But when the solution wassaturated with CaCl2, the rate of emulsioncoalescence increased substantially inbottle tests. Hence, as shown in Fig 14,this produced lower resistance to flowingoil than the system with 5% NaCl. Also,when CS-330 was replaced with morelipophilic TDA-4PO (which has a branchedC13 hydrocarbon chain), emulsioncoalescence was very rapid at 20% NaCl andvery little resistance was offered toflowing oil in Fig 14. Unfortunately, ineach of these cases foam stability wasalso affected adversely. Thus, there issome trade-off between foam stability andemulsion coalescence in using thisapproach.

Hence, a process modification isproposed to address this problem. Themodification would involve application ofa pre-flush of brine before foamgeneration in the field. The idea is thatwhen the well is brought back toproduction, this brine from the pre-flushwill flow through the foam regions andremove most of the surfactant present inthe aqueous phase, and this in turn, willreduce the amount and stability ofemulsion generated. This concept wastested in the 2-ft sand pack as shown inFig 15. Unlike in field conditions, theflow direction was not changed in thelaboratory experiments. So, foam wasgenerated first and then brine wasinjected. When oil was flowed through thesand pack after brine flush, some emulsionwas still generated, probably due to thepresence of some polymeric surfactant atthe air-water interfaces of foam. But thisemulsion flowed out or coalesced over


time. For oil flow amounting to 4-5 PV ofthe initial foam region, resistance factorwas reduced to less than 10. Thisremaining resistance appears not to be dueto emulsion, but instead due to some airtrapped in the porous media. One mayexpect this air to dissolve into theflowing oil over time and offer even lowerresistance to flow.

In summary, this paper demonstratesthe principle of using foam as RPM. Acombination of non-ionic polymeric F108and anionic CS-330 was shown to producefoam that would satisfy many of theimportant criteria required of a RPM. Butdifferent types of polymeric surfactantswill have to be developed for particularfield conditions. For example, polymericF108 is not suitable for reservoirs wheretemperatures exceed its cloud point (about90 oC). Also, a single anionic polymericsurfactant can be developed to generatefoam instead of using a combination ofnon-ionic and anionic surfactants.

Conclusions1. The potential of foam as a RPM for

reducing water production is shown.Mobility reduction factors greater than100 were achieved in 1-dimensional sandpack experiments.

2. Foams stable to washout with water wereproduced by using polymeric surfactantssuch as triblock F108 andhydrophobically modified HMPA1. Thewashout stability was due to almostirreversible adsorption of polymericsurfactants at the air-water interface.

3. Foam dissolution into flowing water wasobserved to be a factor limiting thefoam lifetime of polymeric surfactants.Foam lifetime for field conditions willdepend on dissolved gas content inproduced water and type of gas in foam,but not on the pressure in theformation, to a first approximation.

4. Even accounting for the gasdissolution, scale-up of foam washoutof polymeric F108 to field geometries

showed large enough time between re-treatment of production well for theprocess to be practically feasible.

5. Foam tolerance to residual oil wasproduced by a mixing anionic CS-330with non-ionic polymeric F108. Theincreased tolerance was due toelectrostatic repulsion between theair-water and oil-water interfacesproduced by the charged surfactant.

6. Flowing oil in the porous medium,however, produced a stable emulsionthat did not flow out easily. Theresistance to flowing oil was minimizedby a process modification involving apre-flush of brine before foamgeneration.

AcknowledgementsThe authors would like to thankHalliburton for funding this project andfor providing some of the surfactants usedin this study.

NomenclatureH = Henry’s constant, m/Lt2, atmk = absolute permeability, L2,

darcym = ratio of foam front tointerstitial water velocity,dimensionlessMRF = mobility reduction factor,

dimensionlessMRFf = MRF in foam regionMRFnf = MRF in region without foamMRFmean = mean MRFPw = well pressure for water zone,

m/Lt2, psiPt = pressure at foam front at time

t, m/Lt2, psiPe = pressure deep inside the

formation, m/Lt2, psiPV = pore volumes, dimensionlessR = gas constant, mL2/t2K, 0.082

lit-atm/K-molrw = well radius, L, ft ri = radial distance of initial foam

front, L, ft rt = radial distance of foam front at

time t, L, ft


re = radial distance where pressureis Pe , L, ft

Sg = gas saturation, dimensionlessT = temperature, K, oKvf = foam front velocity, L/t,

ft/dayv = interstitial water velocity,

L/t, ft/day

tfg = time for foam generation, t, days or months

tfw = time for foam washout, t, daysor months

= porosity, dimensionless = mobility, L3t/m = viscosity, m/Lt, cp

SI Metric Conversion Factorscp 1.0* E03 = Pa.sdarcy 9.869 233 E01 = m2

day 8.64* E+04 = sft 3.048* E01 = min 2.54* E+00 = cmpsi 6.894 757 E00 = kPa

*Conversion factor is exact.References

1. Dalrymple, E.D., and Brown, D.:“Effect of Relative-PermeabilityModifier Treatments in a Sandstone-Layered System and in a Sandstone-Homogeneous System”, paper SPE/DOE39636 presented at the 1998 SPE/DOEImproved Oil Recovery Symposium,Tulsa, April 19-22.

2. Eoff, L., Dalrymple, E.D., Reddy,B.R., and Everett, D.: “Structureand Process Optimization for the Useof a Polymeric Relative-PermeabilityModifier in Conformance Control”,paper SPE 64985 presented at the2001 SPE International Symposium onOilfield Chemistry, Houston,February 13-16.

3. Eoff, L., Dalrymple, D., Reddy,B.R., Morgan, J., and Frampton, H.:“Development of a HydrophobicallyModified Water-Soluble Polymer as aSelective Bullhead System for Water-Production Problems”, paper SPE

80206 presented at the 2003 SPEInternational Symposium on OilfieldChemistry, Houston, February 5-7.

4. Liang, J., Sun, H., and Seright,R.S.: “Why Do Gels Reduce WaterPermeability More Than OilPermeability”, SPERE (1995) 282.

5. Liang, J., and Seright, R.: “FurtherInvestigations of Why Gels Reduce kwMore Than ko”, paper 1997 SPE 37249presented at the SPE InternationalSymposium on Oilfield Chemistry,Houston, February 18-21.

6. Zaitoun, A., and Bertin, H.: “Two-Phase Flow Property Modifications byPolymer Adsorption”, paper SPE/DOE39631 presented at the 1998 SPE/DOEImproved Oil Recovery Symposium,Tulsa, April 19-22.

7. Zaitoun, A., and Kohler, N.:“Improved Polyacrylamide treatmentsfor Water Control in ProducingWells”, paper SPE 18501 presented atthe 1989 SPE International Symposiumon Oilfield Chemistry, Houston,February 8-10.

8. Zaitoun, A., and Kohler,N.: “Two-Phase Flow through Porous Media:Effect of an Adsorbed PolymerLayer”, paper SPE 18085 presented atthe 1988 SPE Annual TechnicalConference and Exhibition, Houston,October 2-5.

9. Zitha, P., Chauveteau, G., andZaitoun, A.: “Permeability DependentPropagation of Polyacrylamides UnderNear-Wellbore Flow Conditions”,paper SPE 28955 presented at the1995 SPE International Symposium onOilfield Chemistry, San Antonio,February 14-17.

10. Schramm, L.L.: “FoamSensitivity to Crude Oil in PorousMedia”, in Foams: Fundamentals andApplications in the Petroleum Industry, L.L.Schramm (ed.), American ChemicalSociety, Washington, D.C. (1994).

11. Wasan, D.T., Koczo, K., andNikolov, A.D.: “Mechanisms ofAqueous Foam Stability and


Antifoaming Action with and withoutOil”, in Foams: Fundamentals andApplications in the Petroleum Industry, L.L.Schramm (ed.), American ChemicalSociety, Washington, D.C. (1994).

12. Szafranski, R.: “LaboratoryDevelopment of the Surfactant/FoamProcess for Aquifer Remediation”,Ph.D. thesis, Rice University, 1997.

13. Myers, T.J., and Radke, C.J.:“Transient Foam Displacement in thePresence of Residual Oil: Experimentand Simulation Using a Population-Balance Model”, Ind. Eng. Chem. Res.(2000) 39, 2725.

14. Gauglitz, P.A., Friedmann, F.,Kam, S.I., and Rossen, W.R.: “FoamGeneration in Porous Media”, paperSPE 75177 presented at the 2002SPE/DOE Improved Oil RecoverySymposium, Tulsa, April 13-17.

15. Apaydin, O.G., and Kovscek,A.R.: “Transient Foam Flow inHomogeneous Porous Media: SurfactantConcentration and Capillary EndEffects”, paper SPE 59286 presentedat the 2000 SPE/DOE Improved OilRecovery Symposium, Tulsa, April 3-5.

16. Persoff, P., Radke, C.J.,Pruess, K., Benson, S.M., andWitherspoon, P.A.: “A LaboratoryInvestigation of Foam Flow inSandstone at Elevated Pressure”,SPERE (1991) 365.

17. Bhide, V.V.: “DevelopingStable Foams from PolymericSurfactants for Water ProductionControl”, Ph.D. thesis, RiceUniversity, 2004.

18. Svitova, T.F., Wetherbee,M.J., and Radke, C.J.: “Dynamics ofSurfactant Sorption at the Air/WaterInterface: Continuous-FlowTensiometry”, J. Colloid and InterfaceScience (2003) 261, 170.

19. Hirasaki, G.J., Miller, C.A.,Szafranski, R., Tanzil, D., Lawson,J.B., Meinardus, H., Jin, M.,Londergan, J.T., Jackson, R.E.,

Pope, G.A., and Wade, W.H.: “FieldDemonstration of the Surfactant/FoamProcess for Aquifer Remediation”,paper SPE 39292 presented at the1997 SPE Annual Technical Conferenceand Exhibition, San Antonio, October5-8.

20. Hirasaki, G.J., Jackson, R.E.,Jin, M., Lawson, J.B., Londergan,J., Meinardus, H., Miller, C.A.,Pope,G.A., Szafranski, R., Tanzil,D.: “Field Demonstration of theSurfactant/Foam Process forRemediation of a HeterogeneousAquifer Contaminated with DNAPL”, inNAPL Removal: Surfactants, Foams andMicroemulsions, S. Fiorenza, C.A.Miller, C.L. Oubre, C.H. Ward(eds.), CRC Press LLC, Boca Raton(2000).

21. Hanssen, J.E., and Dalland,M.: “Gas-Blocking Foams”, in Foams:Fundamentals and Applications in the PetroleumIndustry, L.L. Schramm (ed.),American Chemical Society,Washington, D.C. (1994).

22. Blaker, T., Celius, H.K., Lie,T., Martinsen, H.A., Rasmussen, L.,and Vassenden, F.: “Foam for GasMobility Control in the SnorreField: The FAWAG Project”, paper SPE56478 presented at the 1999 SPEAnnual Technical Conference andExhibition, Houston, October 3-6.

23. Garrett, P.R.: “The Mode ofAction of Antifoams”, in Defoaming:Theory and Industrial Applications, P. R.Garrett (ed.), Marcel Dekker, Inc.;New York (1993).

24. Zhang, H., Miller, C.A.,Garrett, P.R., and Raney, K.H.:“Mechanism for Defoaming by Oils andCalcium Soap in Aqueous Systems”, J.Colloid and Interface Science (2003) 263,633.

25. Baldauf, L.M., Schechter,R.S., Wade, W.H., and Graciaa, A.:“The Relationship between SurfactantPhase Behavior and the Creaming andCoalescence of Macroemulsions”, J.


Colloid and Interface Science (1982) 85,187.

26. Aveyard, R., Binks, B.P.,Fletcher, P.D.I., Ye, X., and Lu,J.R.: “The Resolution of Emulsions,Including Crude Oil Emulsions, inRelation to HLB Behavior”, inEmulsions – A Fundamental and PracticalApproach, J. Sjoblom (ed.), KluwerAcademic Publishers, Dordrecht(1991).

Table 1: Conventional surfactants tested for foam washout stability

Surfacta

nt

Active components

SDS Sodium dodecyl sulfate

[CH3(CH2)11-OSO3Na]

C16TAB n-Hexadecyltrimethylammonium bromide

[CH3(CH2)15-N(CH3)3Br]

CS-330 Sodium laureth sulfate

[CH3(CH2)11(OCH2CH2)3-OSO3Na]

AOS

16/18

C16-18 alpha olefin sulfonate

Foamer1 Alkyl polyoxyethylene etherFoamer2 Alpha olefin sulfonateFoamer3 Alkyl ether sulfate

Table 2: Triblock copolymers tested for foam washout stability

Product

name

Average

molecular

weight

% EO Average

molecular

formulaF108 14,600 80 EO132-PO56-EO132

F88 11,400 80 EO103-PO39-EO103

F68 8,400 80 EO76-PO30-EO76

P85 4,600 50 EO26-PO39-EO26

B20-5000 5,000 80 EO45-BO14-EO45

B20-3800 3,800 80 EO34-BO11-EO34

Table 3: Hydrophobically modified polymerstested for foam washout stability

HM

polyme

r

Description

HMPA1 Copolymer of methacrylic acid and ethyl

methacrylate + small amount of methyl

methacrylate, C18 hydrophobe; partially

crosslinked

HMPA3 Similar to above with lower molecular

weight and lower fraction of

crosslinking

Table 4: Entering, spreading and bridging coefficients

Surfacta

nt

a-w o-w o-a E S B

F108 37.2 6.9 23.5 20.6 6.8 879.2B20-5000 33.7 5.2 23.5 15.4 5.0 610.5CS-330 28.0 0.5 23.5 5.0 4.0 232.0SDS 36.8 4.8 23.5 18.1 8.5 825.0

All surfactants 0.5%. All except SDS contain 5%NaCl. = Interfacial tension of a-w (air-water), o-w(oil-water), o-a (oil-air) interfaces in mN/m.E = a-w + o-w - o-a S = a-w - o-w - o-a B = 2

a-w

+ 2o-w - 2

o-a

Table 5: Foam collapse time with flowing oil compared to washout with brine in the screening test

Surfactant Foam collapse

time with

hexadecane

(3 ml/min)

Washout time

with 5% NaCl

(3 ml/min)

0.5% F108, 5%

NaCl

3 min 160 min

0.5% CS-330, 5%

NaCl

3.5 min 60 min

0.5% F108, 0.5%

CS-330, 5% NaCl

6 min 161 min


W ater/brine injection

perforated plate fitted with a 360mesh screen

air injection

surfactant solution

M arks between which tim e forfoam collapse is recorded(Volume ~ 120 m l)

1ft 1 inch

3-way valve

foam

Fig. 1Setup used for foam washout screening test

0

50

100

150

200

250

300

SDS

0.5% C16TAB

CS-330

AOS 16/18

Foamer1

Foamer2

Foamer3

F108

F88

F68

P85

B20-5000

B20-3800

HMPA1

HMPA3

2% surfactant; w ashout w ith DI w ater0.5% surfactant, 5% NaCl; w ashout w ith 5% NaCl0.5% surfactant; w ashout w ith DI w ater

washout time (min) conventional

surfactantstri-block copolym ers

HM polym ers

Fig. 2Foam washout times for conventional andpolymeric surfactants in the screening test

HMPA1 and HMPA3 neutralized with dil. NaOH to pH ~7


3

Length=24 inches, dia=0.9 inch

6 inches

pressure transducer Isco pum ps and/or Syringe pum ps

sand pack: k = 5.5-7.5 darcy, porosity ~0.4

2 4 1

5

To com puter

Auto collector

Fig. 3Experimental setup for the foam washout insand pack

0

50

100

150

200

250

0 1 2 3 4 5 6 7PV

MRF

section 1section 2section 3section 4overall

air12 ft/d

air12 ft/d

air12 ft/d

surf12 ft/d

surf12 ft/d

surf12 ft/d

surf6 ft/d

Sg = 93%

Sg = 54% Sg = 62% Sg = 72%

Fig. 4aFoam generation with B20-5000 in 2-ft sand pack

-10

40

90

140

190

240

290

340

390

6.5 11.5 16.5 21.5 26.5 31.5 36.5 41.5 46.5 51.5PV

MRF


Fig. 4bFoam washout for B20-5000 (at constant brinerate)

0

50

100

150

200

250

300

350

400

6.5 11.5 16.5 21.5 26.5 31.5PV

MRF


266 ft/d 300 ft/d 309 ft/d 330 ft/d 377 ft/d

Fig. 4cFoam washout of B20-5000 using high ratepump (in gaps)

0

50

100

150

200

250

300

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5PV

MRF


air12 ft/d

surf12 ft/d

surf6 ft/d

air

surf

Sg = 74%

Sg = 43% Sg = 55%

Fig. 5aFoam generation with C16TAB


-20

30

80

130

180

230

280

330

380

430

480

4 5 6 7 8 9PV

MRF


DI water 6 ft/d

Fig. 5bFoam washout for C16TAB with DI water

0

5

10

15

20

25

30

CS-330

C16TAB

F108

B20-5000 P85

F88

HMPA

1

HMPA

3

PV

Fig. 6Pore volumes for foam washout in 2-ft sandpack for conventional and polymeric surfactants

Surfactant concentrations: HMPA3 = 2%, rest = 0.5%,all in 5% NaCl (except C16TAB). HMPA1& HMPA3

neutralized with dil. NaOH to pH ~7. HMPA1 andHMPA3 were relatively poor foamers compared to

W ater in

W ater out10 ml/min

Air bubble

Surfactant solution (30 ml)

Iscopump

Iscopump

Teflon coated needle tip

Glass cell2 cm x 4 cm x 8 cm

Lamp

To computer

Video camera

W ater in

W ater out10 ml/min

Air bubble


Iscopump

Iscopump



Lamp

To computer

Video camera

W ater in

W ater out10 ml/min

Air bubble


Iscopump

Iscopump



Lamp

To computer

Video camera

Fig. 7Schematic of pendant bubble setup to measuresurfactant desorption

0

10

20

30

40

50

60

70

80

0 5 10 15 20 25 30 35 40cell volum es

surface tension (mN/m)

SDS

CS-330

F108

HM PA1

Fig. 8Desorption of conventional and polymericsurfactants from air-water interface during washout

W ater zone

W ashed out foam MRF nf

Pw Pt Pi Pe

rw

re ri

rt

No foam MRF nf

Foam MRF f

Fig. 9Schematic of foam in water producing zonenear production well


0

20

40

60

80

100

120

140

0 5 10 15 20 25 30 35 40Foam penetration (ft)

Time for washout (months)

M RFf = 100

M RFf = 50

Foam from F108, m = 0.03

Fig. 11Time for foam washout as a function ofinitial foam penetration distance

AirsourceSurfactant

Filter packedwith sand

1/8” plastictubing

PR

Pressure transducer

Isco pum p

1.5 inch

BurettecontainingIPA-water

3-way

2-way

foam

air

liquidcheck valve

Fig. 12Short sand pack arrangement used for testingfoam stability to residual oil

0

5

10

15

20

25

pressure dr

op ac

ross short pack

(psi)

F108 B20-5000 CS-330 SDS F108+CS330

no oil

residualhexadecane

Fig. 13Steady state pressure drops for foam flow inthe short pack with and without residual oil

0

20

40

60

80

100

120

140

160

0 0.5 1 1.5 2 2.5 3 3.5PV (w .r.t. 1st section)

MRF/oil viscosity

F108+CS-330, 5% NaClF108, 5% NaClF108+CS-330, 36% CaCl2F108+TDA-4PO, 20% NaCl

Fig. 14Resistance to oil flow (MRF/oil viscosity)in 2-ft sand pack initially saturated with

0

20

40

60

80

100

120

140

160

180

200

0 5 10 15PV (w .r.t. 1st section)

MRF section 1

foam generation brine flush oil flow

air air

surf surf

6 ft/d

3 ft/d 3 ft/d

resistance factor for oil flow = 9

Fig. 15Reduced emulsion formation for flowing oilafter a foam process modification (oil flow following

a brine flush in 2-ft sand pack)

R 2 = 0.98

0

10

20

30

40

50

60

70

80

90

100

5 10 15 20 25 30PV

MRF

(fir

st 3

sec

tion

s)

Fig. 10Washout of foam from B20-5000 (data from fig4b)

Foams for Controlling Water Production

Documents

Transcript of Foams for Controlling Water Production