.

SPEH)OE
aoolstyof

Potrolwn Er@rrasra

U.S. Oapmtmant
of Ermrgy

SPE/DOE 20238
Design and Optimization of Alkaline Flooding Formulations
T. Ft. French and T,E, Burchfield,

NatL Inst. for Petroleum & Energy Research

SPE Membsrs

This paper waa prepared for preamlatlon at the SPE/DOESevanlh Sympoalumon EnhanoadOil Raooveryheld In Tulsa, Oklahoma,April 22-25, 1SS0.
Thla paper was aeleotedfo- preaarrtatlonby an SPE ProgramOommlffeefollowing review of information contained in an abatract submitted by the author(s).Contelt, of the paper,
ae prewnted, have not bean eviawad by the Sooletyof PetroleumEndnwrs and aro subleof to mmectlon bYthe author(a).The materfalt ~ PreWnf@ does not n~~*~lY ~fl~t
ormembara.ParmrafxewnteddtspE mwtinware8ubl@towbli~tionre~ewWEdltofialQJmminwbof fhe ~lav
anYpoaitlonof the Sooietyof Pmolaum Englnwra,ireofficers,
of PetroteumEnglnwra. Perrniwionto oopyis rwtrlotd to anabatrwf of not morethan SWwords.Illustration maynot be 00PIed.Theabafraotshouldoontainoormplouousacknowtadgment
of where and by whom the papar is presented. Write Publlcationa Manager, SPE, P.O. Box S3SSS6,Richardson,TX 75CSXWSS.Telex, 7S09S9SPEDAL.

The state+f-the-art for alkaline flooding technology is the

injeotion of combinations of aikaiis with synthetic surfactants.
Surlaotant-enhanced aikaiine flooding formulations with icw-pH
aikaline agents have potentiaifor irforeasedoii recovery. They reaot
iess with reservoir mirmrais and facilitate the use of iow
concentrations of surlaotants because surfactant adsorption is
reduced in the presanoeof aikaline agents.
Laboratory experiment and resuits from field projects have
been evaluated and used to establish guidelines for designing
optimumaikaiineficods. For a reservoirto be a osndidatefor alkailtte
flooding, the reservoirshould ccntakr iittle or no gypsum,the divaiertt
ion exchangeoapacity should be les~than 5 maqkg, and the in situ
pH shoukf be greater than 6.5.

Ths firtalseotion of this paper Is an example of how screening

criterfa, phaae behavior teats, and oorefloodscan be used to design
and optimize an aikalineflood for a speoifkcoil field, The example Is
for the design of a Iow-PH,surfactant-enhancedaikaiineflood in the
RangerZons of the Wilmington(CA), ffeld, where a high-pHaikaline
ficod was previously ccnductsd, That projeot ertoounteredsevere
probiemswith aikeli consumptionand coaling. A bw.pH aikali (such
as NaHC03 or NaHC03 + Na2C03) in ccmbinatbn with synthetic
surfactant should be effective for producing incremsntai oii in the
RangerZone of Wilmingtonfield, Problemsassociatedwith high pH
fioods, such as Intractable scsies and high consumptlcm,wiii be
mitigated.
~
Aikaiine agents have an appeal for enhanced oii recovery
because of their iow cost and favorable performance in laboratory
tests. Aikaiine ag~nts mobiiize oil efficiently in laboratory
experiments: however, fieid applications of aikaiine fiooding
processeshave usualiy been disappdnting.

Two optimizedaikalinesystemswh!ch exhibittypicai inteffaclal

tension (iFil behaviorare desofibed, One systemwas optimizedwith
a very iow acid (acid number = 0,13 mg KOHlg) crude oil from
Delsware-Chiicfere(OK), field, The other systemwas optimizedwith
an acidio (aCid number = 1.59 mg KOH/g)crude oii from Wilmington
(CA), oii field. A mixture of synthetic surfaotant and low-pH alkaii
producedbwer ilV and sustainedlow iFf bnger than either reactant
alone. This effsotwas obeenrsdwith the ackiiocrude and the slightly
acidic crude. it was ooncluded that surfactant-enhancsd alkaiine
fioodirtg with IOW-PHaikaiis shows promise with both acidic and
siightiyackfiooiie.

In the past 60 yeara, more than 50 alkaiine fiooding fieid

projects have been initiated in the United States. Deapfte this
extensive field testing, aikaiine fiooding ie not yet feasible as a
commercial.soaie operation. in many of the tests, aikaiine agents
were extensivelyoonsumadby mineraidissolutionand bn exchange
reaotions,and the depositionof scale at prcductbn welis was severa.
Recentiy,Wyoming, Louisiana,and Texas have been the most
active states for aikaiineflooding. Caiifomiawas a major site at one
time, but reoentEOR pilot teatson the West Coast have not included
aikaiineflooding.

Synthetic surfaotant is an expensive component of icw-pH

aikaiine fbcding formuiatione. Dilute (0.1 .0.2 %) oortcenfratbns of
the added synthetio surfactant are being used in some current
aikaiine flooding field projeots. Laboratory resuits ehowed that
surfactant iosses by adsorption are reduced under aikaiine
conditions. Losees by precipitatbn shoufd aiec be reduced due to
the bwering of divalent bn ocnoentrstionby aikalinepreflush.

Evaluationof past aikaiine field tests is difficult baoauesof the

scarcity of reported information. Data on reservoir mineralogy, oii
properties,and brine wa%es are often unavailable. HOIfever, when

Referertoss
andiiiusfrationa
at endof paper

616

DESIGNAND OPTIMIZATION OF ALKALINE FLOODINGFORMULATIONS

publishedflefd resutteare oomblnedwith laboratorydata, then there

la a auffiolenfiylargedata baw to cfeveiopgood criteria for aoreenhrg,
deslgnlng,and optimizing alkalinefloodingformulellons.

diameter of 0.457 cm. ~ tubes were 50 tt (18,23 m) in terrgthand

were paokedwith tofuenwrfraoted Wifmingfonsend to a porosityof
0.28%. Th8 sandpacks were eaturated wNI Mte folkwed by
Wilmingtoncrude 011A. After watertbodktg, tke afkaline aolutbrts
were injaotedat a frontal advenoerate of 2 tutl (0.61nvcf)oDifferentfaf
pressuresranged from 1.0 to 3.0 paiM (6.9 to 20.7 k P#rn).

This paper is dMded into three seotbna: the rote of syrtthetio

eurfactants in alkaline flooding, the use of aoreening crtterta to
evaluate fiekf prospeots, and the design and optimization of a
chemicaiformubtbn.

Oii Dis@eoementTeats

t P~

Oil dispfaoementteats were conduoted at reeewoirtemperature(52

C) in aandpacks (25.4 by 3.7 om) made with OfeanSdWWmington
serwf. The ooreswere peekedinfOTefion~ afeeveawith Tefbno end
irr a Haaafertype oora holder. E%!wepaokfrrg,
caps am then piti
the sandwas olemed by Soxhfetextreoth wtth tolueneamfwas not
fired. After determining initial brine permeability, the oores were
saturatad with tittered Wilmington oif A and weferfbodad to residual
oil saturetbn. Chemicalslugs were injected at a frontal edvanw rate
of 1 fVD,

Phase Be;wior
Equal volumes (4 mL) of Wilmingtonoil A and aqueous phase
were sealed in 10-mL grsduated ampoules. The samples were
placed in an oven at 52 C. Observationswere made after 1 week.
The sampleswere obsewed et that the and evaluated. The aan@es
were then evatuatedwhile being shaken. Finally, the sampleswere
evaluated again after 1 more day. The observation while shaking
provided infonnatbn about the type of emutsionformed (oil-!n.water
or water-ln+ii), the ease of emufaifbatlon,the stabilityrd the emulsbn
(while shaking), and the amount (qualitative) of 011em~dslfied. The
final evak@on (1 day after shaking) provides wfditional data on
emulsbn stability.

Surfactarrt
Anaiytbal methods used for measurement of surfaotant
oonoentratbn were two-phaee titretbn and HPLC. The two-phase
titratbn was described by Rosen and t30fdsmith,4and the HPLC
prooedureis sirnliarto that desoribedby Hofmanand Angstadt,6 The
wavelength used for absorbenoe measurement of Petmatep B-100
surfactant was 222 nm, whbh is very close to the wavelength
selectedby Hofmanand Arrgatadt.

The emulsions formed were evaluated according to the

followingcriteria

No visible emuisifkation

Blaokemulsbn

Brown emulsbn that changesto black

while ehaking

Brown emulsbn

Brown enwlsbn oontalning>90% of

the oii in the sample

SPE/DOE20238

Alkalinity
Effiuent sampleewere oolleoted in 5-ml incrementsand oerrtrituged
before the volume of oil ooifeoted was determined. The aqueous
phasewas then eeparatedand anaiyzedfor alkalinityby titratbn to pH
4 with 0,1 N HCL
ElementalAnalysis
The effluents were anaiyzed for siiloon with atomic absorption
spectroscopy,

Higher values of emuision quaiity correlate with greater

interracial activity and nearness to optimal conditions. Thie visuai
method for determining optimal saiinity for producing interfaclali
active eyatemsis only slightlydifferentfrom those used by others.2J
The bwer pH alkailsottenfail to aohievethe ultra-bw iFT values
necessaryfor sfgnitbant rnobiiizatkn of oil. It was discoveredin 1955
that the additfonof very small cmoentrafbna of aynthetiotrurfaotant
to alkalineSObtiOnscoutd signitkantty i~rove transientIFT behavbr
and oil rrmbiiizatbnwith wetkty atkalinechemkals.6 Two examplesof
optimized alkaline systems are presented whkh exhibit fypbel IFT
behavbr with and without added aurfaotant.

Interglacial
Tensbn
Irrterfacialpropertiesof seleotedsystemswere measuredwith a
spinning drop interfaclal tensbmeter at 52 C. The transient iFT
behavior of nonequilibriated samples was monitored over a time
intervalof severalhours.

Acidic Crude

SurfaotantTransporl through PorousMedia

Unfired, oil-free Berea sandstoneoores (25,4X 3.7 cm) were
mounted in Hassler-type ooreholders. Frontal advance rates were
approximately0.5 W(I (0.152rnld), provfdinga oore residencetime of
about 1.7 days, Surlaotant transport through the oii-free oores was
monitoredqt 52 C (overwontroiled).
Silm Tube Sandpaoke

I
I
I

Slim tube experimentswere performedat resewoir te~erature (52

C) to determine alkali consumption while propagating a pH front
through Wilmington sand, The slim tubes were oonstruofed of
stainlesssteel tubing with an outer diameterof 0.535 cm and an inner

I
616

The first example is for a pH 9.3 mixture of 0.095 N NaHC03

and 0.095 N Na2C03 that was tested at 5P C with oil B from the
Ranger Zone of Wilmington, CA field. The Wilmington orude oil is
acidio(Table 1) and quite reaotivewith aikaii. F&l showsthat alkali
oeusee a very rapid bwertng of IFT to 55 @4/min Only 2 minutes;
however, 1~ bagina increasingimmediately. Within 32 minutes, IFT
inoreasedto 500 @N/m.
When 0.1% Neodol 25-9 afoohol ethoxylate Surfactant and
alkali were oombfned(Fig. 1), the reauttwas a rapiddeoreasein IFT,
folbwed by a sustained,very sbw, irwreasein IFT. The minimum1~
was 4 @Urnfor the oornbinatbn of aurfaotantand aikali, much lower
than for surtaofant(shown in Fig. 2) or aikali abne. The iFT with the
oombinatbn of syntheticaurtaotantand alkali was atlli bebw 50 @J/m
after 16 mirrutee.

SPE)DOE 20238

Trov-- R. French and Thomas E. Btarchiield

Theconcfusbndrawn from Figs. 1 mcf2is0ratwnhWil

mfngton
011,
a conWMkm of sytlhetk aurfectantand atkaliproducesbwer In
and sustains fewer lfT fonger, than either component alone. It
appearsthat for this oil, very kw In la pradcmlnantfya result of the
natural aurfactanfs produced by reactions wlfh the alkall, ml the
sustained low IFT is prtmarily the rusuft of the added synthetto
eurtactant.7

eurfecfant retention (by edaorptbn) was reduced by 49% when the

pH was Inmeasadfrnm &3 to 10.2.
The questbn of how much surfactant Is rqtrlfad for 90Cd
propagatbn fnto a particular reservoir has not been aafkfacfotity
answered, Measurements on the adsorptbn O? two anionic
s::actanta onto Wilmington Ranger Zone sand are gtvan fn Table 3.
Cabulatbna based on these data indfcete that about 2 PV of 0.1%
surfacfant aofutbn would be rquired to meet raeervofr adsorption
raqulrements. Since adeorptbn la partialtyreverafbte,this me be an
overestimate. It fa hoped that at least one of the afkaUnefk L teats
currently bel)~ conducted at tow eurfactant fevelawill pfovfde more
Informatbn aboutthe actual magNtudaof eudacterUfoaaaa.

Low-Ackt Crude
The second example is for a light o!! %omDelsware-Chikfera
(OK), fkdd, (See Tabfe 1,) The alkali used was a pfi 9.5 mixture of
0.049 N NaHCO~,0.0078 N Na2C03, ml 0.0325 N Na2HP04, The
experiments were conducted % S2 C, the same temperature that
was used for the heaviercFJde. The Delswara.Chllderaoil is stightfy
acidicand relativelyunreacifveto akall. Fig,3 sh6w8that afkaliabn3
producedonfy a small reductbn in in.

Sciubilfty
AnoNrerway surfactant bases may ocour in a M88rvok is by
filtratbn.8 This effect shcukf be consideredwhen dsslgnlng @kalltWsurfactanffbode. Many EOR eurfecfants@ comptex mixturesthat
are sparfngly soluble in aqueous solutions. Many popular EOR
surfactants have bw aoiutrflttiesin brines. Forrnutstbne made wltrr
these surfactants are of,m cloudy dlsperebne, instead of tme
aolutlone. For example, the fittersbilfty of 0.1% Petroetep B-1OO
surfactarrtat 50 C was greatfy reducedwhen mixedwith a reletfvefy
high Ionk strength alkaline eotutbn. At 50 C, in 0.3% NaCl (fonk
strength = 0.051), 98.lA of the aurfactant worJtdpass through a
0.45 pfifter, However,when a mbrfureof 0.032NN@HP04 +0.032N
NaHC03 + 0.016N Na2C@ wtfh a pH of 10 waa added to the 0.3%
NaCl (total bnlc strength = 0.155), the fractbn of surfactant that
passedthrough the filter was reducedto 22.1%.

The additbn of 0.1% PetrcetepB-100 anionic surfactantto the

alkaliresultedIn a very rapidbwering of lfT to a bw value,whkh wee
sustained for a long perfcd. The minimum lFl of 2.9 mim for the
combination of synthetfc surfactant and alkali waa much bwer than
that & $eved with either component alone. The IFT behavbr for
0.1% aurfactantabne is shownin Fig.4.
The resutfswith the Delsware-Chltdemcrude indicatethat even
for crude olla with very ~w acid numbers, the combination of alkali
ati a bw concentratbn of syntheticSuffactant can producevery bw
Il=r Vsluee.a

Synthetic surfactant is an expensive component of IOW-PH

alkalinefboding formulations. Forthis reason,onlyxfilute (0.1-O.2%)
Concentration of added surfactant are used In current field
m@ts.l $ As ahcwn above, very low concerrtratbns of surfactant
!n alkaline solution are sufficient to produce an ultra-low IFT. A
potential problem is that the surfactant, at these low levels, may not
propagatevery far Into an oil raeewolr because of losses caused by
preclplfetbn, adaorptbn, and phase trepplng,
Precipitation
Calcium and magnesium ions present irr reservoka can
precipitate the aurfactants used in EOR applbatlons. Alkalis are
effectfve for reducing the levels of divalent ions in petroleum
reaervoirs,7*f~-11The effectivenessof different alkalis for reducing
dlvalent ion levels are: hydroxide < carbonate < silicate. The
preferred Injectionstrategy Is probablyto use an alkaline prefluahto
reducedivalent bn levels before Injectinga forrnulatbn that con!ains
surfacfant,

The effect of this behavbr k shown in Fig. 6, whkh la the

effluent analysis of a 52 C corefbod performed with the above
soiufbns in an oil-free, 600 md Bereacore. Total aurfactentratentbn
of 0.160 meqkg of core waa qufte bw. However,the passageof the
surfactant through the core occurs in two distinct 8UffaCtantpeaks.
The second peak occurred after V@ctbn of the bwer bnic strength
chaee brine was resumed. This effect is believed to be due to
filtratbn, ratherthan tnie adaofiwbn.
This parlicufar experiment was performed in an oil-free core
with a refetivelyhigh fonk strength alkalineeofutbn. ft is ptWWth3dto
show that the eolubilnieso! aurfactantsIn alkaline aohitbna need to
be considered. If oil had been present In the core, the manner in
which the surfactant was propagated through the core could have
occurred differently,
~B.
Reservoir mineral, brine, and oil propettiss that are most
Important for determining the feaaibillty of alkaiine fbodhg were
prevbusfy identified from a detailed study of alkaline field projects
conducted since 1960.1 Some of the data publishedin the chemical
literatureand reeuttsof Iatmratoryexperimentswere especiallyuseful
in correistbn of minerafcgyand the albweble pH level of the Injected
fluids. This is particularly knpcrlant bacauaethe veIy high pH alkalis
are completely consumed by some of the minerala in petroisum
resewoks. A summaryof the ecreenlngcriteriafolbwa.

An example of the use of sodium bicarbonate in a preflush ie

shown in Fig. 5.12 Sandstonecoreswere initiallysaturatedwith brine
contakrlng 1,000 ppm each of cskium and magnesiumions. With a
aallne preflush of ecdlum chloride, calcium and magnesium Ions
persisted In the effluent long after the original brfne was displaced
from the core, while the injection of sodium bicwbonate reduced
fhek concentrationmuch eartler. Onfy 0,25 PV of l% NaHC@ woutd
be requiredto preclpfteteall of the calcium (aa CaC@) in a brfnethat
contains600 ppm of caklum Ions.

The C02 content of petroleum reservoirs la repotted only

kifrequentfy, Yet, ft is one of the moat Important parameters for
determiningwhether a reservoir is a candkfate for alkaline fbcdlng.
Aikaiine flcwdlng ahoukf usually not be considered In reservolesof
high C02 content: mole fraction of C02 (in produced gas) >0.01 or
pH c 6.5.1)14

Adsorption
Surfacfantlevels In an injectedsolutbn can also be reducedby
adsorption onto reaewoir rock. Numerous researchers have
meaaured the adsorption of anionio aunactants under alkaline
condltiorre. These reauftscan be summarized by stating that alkali
reducedthe adsorptbn of anbnk surfactantsonto pure clay minerals
as much as 93% and reduced adsorption 7 to 49% when reservoir
sandstones were used.8*lo~12.13

Gypsum

Teble 2 summarizesthe resuftsof several corefkods at two pH

levels. hr oil.free, high permeability Berea sandstone cores,

The presence of gypsum (anhydrtte)Is a deterrent to any ~

of alkaline (or surfacfanf) ffoodlng. it has been shown that l%
9ypsum in a maewoir mn consume 10 PV of a We solution of
NaHCO$ Alkalinefiocding should be rejectedfor any reservclr with
617

DESIGN AND OPTIMIZATION OF ALKALINE FLOODING FOIWJIATIONS

SPE/DOE 20238

Oreeferttran O.l%9Ypsum. Thisamount ofgypaum can bainfened

from 1000 ppm sulfate bn in the btfne.l
Kaolinfte
The presenceof kaolinite Ie a serious deterrentto flcodtng with
higher pH alkalis.ls Only bw pH (about 8.2- 10) atkalla strcukf be
consideredfor use in reservoirsthat contain appreciableamcunte of
mollnite.

Examination of the brine analysia provldee a very postiiie

hctbetbn that thle field should be amenabteto alketineflooding. The
carbon dbxlde cmtent should beJecceptsbla,as indbatad by the pH
of 7.4. The total of 947 m#f divalent bns is weti wtthinthe ecreenlng
criterion of 0.4 wt. % (4,000 rn@t)which relates to the dfvatem ton
exchangecapacity.
Examination of the rock mineralogy reveals that there am not
slgnifkent amounts of any clays. Ontytraces of kaollnltearxf mbredIayer illiie/smactiteare present. (3ypaumis present,but onty in trace
amounts. Despfte the clean appearance of the brfne and rock,
severe probfems were encountered with alkali consumption and
scaling durfng the previous project that was conckwtedwith acdfum
ot.hosilkate.

?dontrnorillonite
Montmorillonite,whkh is leasreactivewith atkalie,la detrimental
to alkaline fboding processes because of tts high surface area arxl
vety high catbn exchangecapacity. High CEC velueeare asacciated
with high montmorillontfe content. If divalenf Ions OCCUPY
the
exchange eltes, much of the injected alkali can be consumed by
adverse precipitation reactions. The screening criterion for ion
exohange capacity Is 5 meqkg. This corresponds rcughly with the
presenceof 1% montrncrilbnite and 0.4 wl % of diva[ent bns in the
brine,1116

The low-clay, low-gypsum cartent of the Wilmington sand

indkates that reactionwith alkali should be smell. Yet, the published
data show that losses of orthcsilicate in the Ranger Zone could be
very high. Chemical consumption flow experiments showed that
after 36 days af reskfence time, losses of Otihosilkate were afmut
169 msqkg of rock.21

Limestone
High pH alkalis are reactive with limestone and should be
avoided in dolomitic reservoirs. Low (Nai-iC03) and mcderate
(Na2C03) pH carbonates have been shown to be compatible with
dobmitlc cores.1718

Much of the scale formation was attributed to magnesium

silicate. The apparent source of magnesiumis the brine. Laboratory
tests22 demonstrate the reaction of silicate with brine magnesium
ions, but it is not enough to account for the amount of Silicate
consumption. This, abne, suggeststhat the mineralanafyeiemay not
be fully representative; however, plagicclase is another possible
ecurce of magnesium.

AddJ@M&
Tne acid number of crude oil was previouslythought to be an
important parameter for alkaline flooding processes.
The
exparimentadescrfbed in the first section of this paper showed that
oil acid number may not bean extremely impcrtarrtparameter. The
detaiied study of 24 alkaline field projects that was menticned above
failed to chow a correlation between oil acid number and project
success.

Plagicclasefeldspar is a major componentin the RangerZone

sand analysis shown in Table 4. PiagloclaeeISsomewhat reactive
with alkali,15 and this reaction coufd account for some of the alkali
consumptbn tn the field test.21 In laboratorymeasurementsof alkal}
transmission through Wilmington sand22 there were both a fast
reaction, as if with exchange Ions, and a slow reaction, as if by
aluminosilicatedissolution. A comparisonof fast alkali ceneumptbn
for three alkalis, and for mixtures, is given in Table 5. Ali vafuee in
Table 5 were determined by NIPER. The value for OIthMOSte is an
order of magnttudefewer than that reported in Ref. 21. Among the
differences in experimental conditions is the time of residence, If
consumption had been evaluated after 36 days in these
experiments, as it was In Ref. 21, consumptbn cculd have been
higher.

it appearsthat oil acid nuntwr is Importantfor achievingtow In

in alkalinefbods that do not Includeadded syntheticsurfacfantIn the
aikaline fbcding fonnulatbns. Even then, mechanismsother than
IFT bwering may resutt in nrobliizatbn of residual oil during alkaline
fboding. The state-of-the-artin alkaline flooding is the combination
of bw-pH alkalis with small amounts of added synthetic surfactants.
Under thcae conditions, there is not enough informationto conclude
that oil acid number should be considered a criterion for alkaline
flooding,

The size of the alkaline slug required to satisfy the alkali

consumption capacity (ACC) of the Wilmington resewoir was
cakulated from the folbwlrrgr

Other parameters such aa temperature, penneablilty, salinity,

and oil viscosity have the familiar limits for chemwl flooding,whose
valuea depend on the specific limitations of the surfactant and
polymer, 1

m
wi ~nsumW Kg&f!o~kxpmk
dug size(PV)

--(1)

alkali consumed
...................(2)
alkali concentration x a

where equivalentsfig of rock is the experimentallymeasured alkali

consumption, p is bulk rock density In kg/dr#, Injected alkali
concentration b normality, and a is the fractional porosity of the
reservoir.

Wilmington (CA), field is presented as an example of the

design and optlmlzatbn of an aikallne flood. Wilmington field was
selected because it ie a large oil field that was the site of a previous
alkaline flood. An alkaline flooding prcject, which encountered
severe problemswith alkal!conaumptbn and scaling,waa previously
conducted In the RangerZone of the fletd.19-21

Cakulatbne with these equatbns were made with the data

given in Table 5 for exparfmerrtallydetermined atkali consumptbn
measurementsusing Wilmingtonflefd aand. Resultsfrom sandpack
floods with carbonate mixtures and results from. ellm-tube
experiments with pure carbonates were used for calculatbns.
Resktencetime in the slim tube was approximately1 rtwnth.

Resewoir parameters shown in Table 4 were taken from

publications by the City of Long Beach and Thums Long Beach
Company.l 9 The crude oil, furnished by Thums Long Beach
Mmpany, from Well A-52 and brine were analyzedafter delivery to
NIPER. The rIXMmineralogyIs of sendfromWell D-4924.

From Eqs. 1 and 2, the size of a slug of 0.095N NaHC03

neededto satisfythe ACC of the reeewcdrfa 0,18 PV. (This b only a
fifth of the amountof sodium silicatethat would be requfredto satfefy

61s

:PRAX3E
20238.
-.
,.
_,---

Trov. R. French

and Thomas E. RurcW#ld

that a favorable rrwbility ratio was assured. Ths shear rate expected
at a frontal advance rate of 1 WD (0.3 m/d) for a 32% porosity
sandpackwith 400 md pmneability was calculated:

the ACC). The size of a 0.095 N Na2C03 slug is 0.51 PV. The larger
value for Na2C03 reflects the increased reactivity of the Na2C03.
The sandpack flood results snow that only 0,22 PV of the 0.19 N
equinormal mixture of carbonates would be needed to satisfy
This value reflects the higher
consumption requirements.
concentratbn of the solution more than the shorter residencetime In
the cores.

shear rate = (266.4)(V) .~ .........................................(3)

T
= 7.7 sac-f

These results for alkali consumption are, at best, only

estimates. They are, however,very encouraging. It appearsthat the
lower pH alkalis ehould propagate into the resewolr a substantial
distance before being completelyconsumedby rook-alkalireacfbns.
Just as important,these bwer pH alkaliswill not cause the intractable
sillcste scalesthat w8re encounteredin the prevbus pibt pmject.21

where V . frontai advance rate (ff/D), 0 = porosity (fraction) and

K = pe~eabii~y (N).
The viscosity of Wilmingtoncrude oil A used in this study was
measuredto be 78.8 cp (0.0786 Pas)at the reservoirtemperatureof
52 C. This is much higher viscositythan that repotted prevbusly for
oil from the Ranger Zor 119 This difference in viscosftymay be me
to loss of solution gas. At the calculated shear rate, 3,000 ppm
4800CX biopolymer has a viscosity of 118 cp (0.118 Pae). Since
shear rate calculations for a sandpack are at best approximate, a
pciymer bwrcentratbn of 3,000 ppm is probabiy the minimum that
assuresa favorable rncbiiity ratio. me 3,000-ppm@ymer siug was
folbwed by two iower concentrationsiugfsof 45 cp (0.045 Pa.e)and
9 Cp(0.008 Pas) Vieccsny.

ebMmBsFwQmmwwl
One of the mechanisms for mobllizatbn of residual oil is IFT
reduction, and phase behavior tests can be used to optimize the
parametersthat will producelow IFT. The experimentalpmcadurefor
performing these tests was described above and in previous
p*,51ications,2-3,7-8Sinm the 10W-pl-falkalis such as NaHC03 and
Na2C03 dc not produce the uitra-bw IFT valuea of the higher-pH
alkalis,small amountsof syntheticeurfactantswere addedto the alkali
to promotefurther Iowerfngof IIT. Wilmingtonoil A was used for the
phase bahavbr tests.

The sandpack flood tests were conducted with and without

aikaii, with and without surfacfsnt, and with and without prefiush.
Resuttsof the sandpack fioods are ccmoined with ccnsunqNbn and
adsorptbn data in other sectionsof this paper In orderto remmrnend
Injecfbn strategiesthat maybe successfulin Wilmingtonfield.

Salinity requirementdiagrams (SRD) for two ionic surfactants

and one nonionic eurfactantwere prepsredfrom the phase behavbr
test results. Surfactant concentratbn was used for the abscissa in
the SRDS and NaCl concentration was used for the ordinate. All
sen@&sccrrfahwcf0.095N NaHC03 + 0.025NNa2C03 (PH= 9.3) or
0.19N Na2C03 (PH= 10.8). The areas indketed in the SRDSare the
areas where interracial activity was visually obsewed. The
experimentswere conductedat 52 C.

The amount of incremental oii produced during the sandpack

floods correlates with the interfaciai tension measurementsgiven in
Table 7. The three systems that had iFTs bebw 4 f.tN/mproduced
the most oil. me two eystemswith IFIs above 50 fdWmproduced
the ieast oil, Alkaline ficodirrgwith aikalinepreflush(Wii-3)produced
31A more oii than aikaiine fiocding without preflush. Aikailne
floodingwith 0,1% eurfactant(Wii-3) produced83% moreoif than the
equivalent flood without surfactant (Wil-4). Aikaline flooding with
0.1% surfactant (Wii-3) produced 49% more oil then flooding with
eurtactantabne (Wii-5).

The salinity requirement diagram for Petrostep B-105, an

anicnlc sutfactant, is shown in Figure 7. The areas indicatedare the
zones where interracial activity was observed. The figure shows
changes that occur as surfacfant concentration increases and pH
Increases, The additbn of sudactant to sodium cwbcnate or to
sodiumbicarbonatemixedwith sodium carbonatecausesthe optimai
salinityto incfeaseat any eurfactantccncentrstbn; the optimalsalinity
of the mixed carbonatesis higherthan the optimal salinityfor sodium
carbonate;the optimai salinity for mixed carbonatesis afways higher
and closerto the optimaleaiinityof the commercialsurfactants. This is
because a lower concentratbn of naturai eurfactantis availabiefrom
the cnrde oii at PH 9.3 than at pH 10.8. The equinormsi mixture of
NaHC03 and Na2C03 is especially interesting because of the
Interracialactivity that occurs between 2.4 and 3.8?40NaCi St O.1%
surfactant concentration, This optimai salinity brackets the
Wilmingtonresetvcir brine compositionwhkh la 2.7%.

In these iimited results,aikaivsurfacfarrvpciymer(ASP)fbcding

mobilized more oii than aikali/pclymer (AP) fboding, and rmnbnk
surfactant was superior to anionic surfactant. ASP flooding also
mcbiiiied more oil than surfactant@ciymer(SP) fiocding conducted
at equivalent ionic strength. These oil production results correlate
with other resultsthat ahowthat alkaii and surfactanfwork togetherto
mcbiiizeoii.
Sandpack fbcd effiuents were also analyzedfor IX ,, aikaiinffy,
and siibcn. Oniy backgroundlevels (5 ppm) of silken were detected,
The absence of dissolved silicon in the effluent is encouraging
becauseif confirmsthat siikste scale is not beingproduced.

Similar experiments (see Figs, 8 and 9) were performed with

Nsodcl 25-9, a nonionic alcohol ethoxylate, and AES 1215-3S, an
anbnb alcohol ethoxysulfate. Increasingthe concentratbn of either
of these surfactants resulted in an Increasedoptimal sailnity. This
behavbr is moredramatk than was observedIn the PetroetepB-105
system previouslydisramsed.

Wilmington field reservoir data were used for an example cost

calculation. For this example,the flooding sequenceis:
0.25 PV 0.025NNaHC03 + 0.095NNa2C03

The optimai eallnitywindow is an importantconsiderationwhen

choosing a surfactant to be used in a spscifk oil field. The AES
1215-3S eurfactant gave somewhat higher IFT in combinatbn with
mixed carbrates and aisc had a higherthan desiredoptimal salinity
window at 0.1%surfactanf ccncentratbn than the other surfactants.
Therefore, Neodol 25-9 and Petrcstep B-105 (in combinatbn with
the mixed carbonates)were selected for oii mcbiiizatlon tests. The
resufts of the salinity requirement tests and IFT tests at 0.19!0
surfactsntccncentratbn are summarizedin Tebte6.

Phase behavior results were ueed to design a series of

sandpack fbods. Injected chemicais and injection strategies are
summarizedin Table 7. The potymerconcentrationwas selected sc
619

0.25 PV 0.095N NaHC03 + 0.025NNa2C03 + 3,000 ppm

bia.?olymer+ 0.1% synthetb eurfactant
0.?0 PV 2,000 ppm bbpolymsr
005 PV 1,000 ppm bbpclymsr
The purpcae of the alkaline prefiush is to satisfy the alkali
consumption due to reaction with the reservoir rock and brine, to
pmtacf the bw ooncentretioneurfactantfmm precipttatbn by divatent
BSS@
ions,7*s,11,21end to reduce loseof eurfacfantby a~pfbn.
on a conservative38% oil recovery,the chemicalmet for recovering
crude oil would be $3.901bbl in this example, with 78gAof that
amount being sfxmt for polymer.

-------

1.

2.

----

--

.-.

For a reservoir to be a candidate for alkaline flooding, the

reservoir should contain little or nc gypsum, the divalent bn
exchange capacity should be fees than S mac#kg,and the In
situ pH ahoukfbe greaterthan 6.6.
A mixture of synthetk aurfactant ati bw-pH alkall produces
lower IFT and sustains fow IFT longer than either reactant
alone.Thlewas observedwith crudeshavingbw and high acid
numbers.

3.

An alkaline preflush protects synthetic surfaotants from

preclpltatbn by removal of divalent krns and reautts in higher
incremental oil production, although sweep efficiency of
preflush aoMbns may be paw.

4.

Alkali reducee the adaorptlon of surfactant; however, the

question of how much surfactant Is needed for good
prcI~I:~
through a reservoir has not been satisfactorily

5.

.-.r

A bw-pH alkali (euch as NaHC03 or NaHC03 \ Na2C03) In

ccm!alnatlonwith synthetic surfactant should b~ effective for
producing incremental oil In the Ranger Zone of the
Wilmingtonfield, which was given ae an exar@a.

The authors wish to thank Dr. Philip B. Lorenz for his work on
definingthe screeningcriteria that were describedIn this paper. The
authorsalSCwlahto thank Kerr.McGeeChemicalCorpcratbn and the
U.S. Departmentof Energyfor sponeortngthe work,

1.

Lorenz, P, B. and D. A. Peru. Guidelines Help Select

Reservoirsfor NaHC03 EOR, Oil and Gss J. Sept. 11,1989,
pp. 53-57.

2.

Nelson, R. C., J. B. Lawson, D, R. Thigpen, and G. L.

Stegemeier. Cosurfactant-EnhancedAlkaline Flooding, f%s.
at Fourth Joint SPE/DOE Symposium on Enhanced Oil
~~~2ery, Tulsa, OK, Apr. 16-18, 1964, SPEIDOE Paper
.

3.

Wasan, D, T,, Enhanced Oil Recovery Through In Situ

Generateed Surfaotants Augmented by Chemical Injection,
Illlnois Instituteof Technobgy Annual Report 1989.

4.

Rosen, M. J, and H. A. Goldsmith. Systematic Analysis of

Surface Active Agents. Wiley4nterscienoa, N. Y., 1972, pp.
423.424.

5.

Hofman, Y. L. and H. P. Agnstadt, Analysis of EnhancedOil

Recovery Formulations. Chmmatographia, v. 24, 1987, pp.
666-879.

6.

Peru, D. A. , Aqueous Flooding Methoda for Terliary Oil

Recovery, U.S. Patent 4,817,7f5, Apr. 1989.

7.

French,T. R., D. A. Peru, and S, D, Thornton,Low

Alkaline
Chemical Formulations, U. S. Dept. of Energy r e~rt No.
NIPER-375,Oct. 1956.

8.

French, T. R. ,Deslgn and Optlmizatlon of PhoaphateContainlrrg Alkaline Flooding Formulations, U. S. Dept. of

Energy FteporlNo. NIPER-446,Sept. 1989.

9.

Clark,S. R., L M. Pitts,and S. M. Smfth,Daetgnand Applkatbn

of an Atkallne-Surlactant-PolymerRecovery System to West
Kiehl Field, Pres. at SPE Rooky Mountain Ragbnal Meeting,
May 1688, SPE paper No.17538.

-----

--------

---,

--------

-.

-,

-v-

---.

I o.

Krumrine, P. H., J. S. Falcone, Jr., and T. C. CamPf@lt

Surfactant Ffcoding 1: The Effect of AtkalineAdditfveacn IFT,
Surfaotant Adsorption, and Recovery Efficiency, BOO.Pet.
Eng. J., Aug. 1982.

11,

Latrrid,J, and Bazin,B. AlkalinePreffuahIn a Lcw-PermaabUtty

Cfayey Sandstone.J. Pet. sol. Eng., 3,1989, pp. 111-120.

12.

Peru, D. A. ad P. B. Lorenz, Surfactant-EnhancedLow PH

Alkaline Flooding,paper SpE 17117.

13.

Hurd, B. G,, Adsorption and Transport of CfWMbSlS@cfes In

Latwratory Surfactant Waterfbodlng Exparfments, Pres. at
SPE Improved 011Recovery Symposium,
Tube, OK, Mar. 2224, 1976, SPE paper 5818.

14.

Collins, A. G., Geochemistry of Oitiiekf waters,

Scientifk Publishing CO.,1975.

15.

Thornton, S. L., Reaotbn of Sodium Hydroxide with Silicate

Minerals, U.S. Dept. of Energy Report No. NIPER-128,N-115
Order No. BE86000275,Apr. 1986.

18,

Smith, F, W., lon.Exchange Ckmdifbnintl of Sandstones for

Chemkal Flooding, J. Pet. Tech., v. 30, June 1976, p. 959968.

17.

Burk, J. H,, Comparison of Sodium Carbonate, Sodium

Hydroxkfe, and Scdlum 0rth0611kSt0for EOR, SPE Reservoir
Eng., v.2, No. f, Feb. 1987. PP.9-1$.

18.

Cheng, K, H. Chembal Consumptbn DuringAlkaline Ffcodlng:

A Comparative Evaluation, Pras. at Fiftfr Joint SPE/DOE
Symposium on Enhanced oil Recovery, Tuba, OK* Apr. 2023, 1986, SPEIDOE paper 14944.

19.

Gro$akurth, D. E., Cauatic Waterflcoding Demonstration

Project, Ranger Zone Lon9 Beach Unfft ~lmin9ton Fiefdc
California, Annual report, October 197wuns 1977, U.S. Dept.
of Energy Report No. SAN/1385-l.

20.

Mayer, E. H. and V. S. Breit, Alkaline Ffcod Predbtbn Studies,

Ranger w pilot, Wilmington FieMt cal~omla, p~a. at Third
Joint SPf3DOE Sympostumon EnharwedCMRecsvery,Tulsa,
OK, Apr. 4-7,1982, SPE/DOEpaper No. 10720.

21,

DaubanP. L., R. A. Eaaterty,and M. M. WesfemtAn Eva~@n

of the Alkaline Waterllooding Demonstration Project, Ranger
Zone Wilmington Ffetd,Ca[tfomia,US. De@,of EnergyReport
No. DOE/BC/10830-51,May 1887.

22,

Thornton, S. D. and P. B. Lorenz, Mineral-Alkali Reactbna

Under Dynamk Conditima. U.S. Dept. of Energy Report No.
NIPER-340, Aug.1988.

Elaevler

SPE 20238

TABLE 1.- Ctude Oik

oil

Gravity,0 APl

Totalaoid rwmber,mg KOH/g

Deiaware-Chikier

33.0

0.13

Wiknlngton011A

12.9

3.40

Wilmington011B

18.0

1.59

TABLE 2,- Injaofbn of 0.1% PetroafepB-1OOeurfaotantInto unfired Berea cores

Chemical
Frontal
Temp., advadoe
Coreflood C

Permeability,

Chemical

Chemioal
ini fLCt8s#~-

vo:~,
pH

Final

Inittal

Alkali Surfactant Alkali Surfactant

23

0.47

1.05

10.17

508

331

21.89

0.301

1.91

0.198

52

0.59

1.40

10.17

502

400

28.57

0.477

1.31

0.186

52

0.55

1.31

6.3

554

435

0.408

0.375

1Per kg of rook.

TABLE 3.- Sutfactant adsorption

Alkali

Sandstone

Surfaotant

Experiment

meq. adsorbecfl
kg sandstone

.0325NNa2HP04
.0162NNa2C03 +
.032NNaHC03
(0S% NaCl)

crushed
Berea

Petroatep B-1OO

Batch

0.834

,0325NNa2HP04+
.0162NNa2Cm +
,032NNaHC03
(0.3%NaCl)

Berea

Petroatep B-1OO

Coreflood

0.198

0.28N NaHC03
(1.48%NaCl)

Wiimingm
sand

Petrostep B-105

Batch

2.18

0.14NNaHC03 +
0,14N Na2C03
(1.26%NaCl)

Wilmington
eand

Petroatep B-105

0.085N NaHCOa
(2.4%NaCl)

Wiiminf$on
aand

Naodol 25-3S

slimtube(40 ft)

0.869

0.067N Na2C03
(2.4%NaCl)

Naodol 25-3S

Slimtube(40 ff)

0.622

sand

621

2.26

TAELE4.- WilmingtonRangerZone fletd data

OUA
0.9796
12.92
413.0
3,40
0.64
76.44
10.84

spaoifiogravity@ 60/~OF
API gravity, de9rees
KinernatioVISOOSW
@ 1250F,*
Total add nutir, rn@gm
Nitrogen (ohemiluminescem), wt %
Carton, wt %
Hydrogen,WI%
Brine

26,600
322
625
7.4

Total dissolved solids, mgA

Magnesium,IWL
Calcium,rr@
pH
Rook mineralogy,FO *ML 4660.8 to 46*4 n

74
23
trc
trc
3
tro
18
5

Quartz,%
Feldepar,%
Gypsum,%
Kaolinite,%
illite/mioe,%
Mixed-layerilliieMneotiie, %
Plagioolesefetdepar,L
FektsparK, %
Resewoirparameters

52
100 to 3,000
850
320

Temperature,C
Permeabilii, rnd
Gross zone thickness,ft
Net zone thickness,ft

TABLE 5.- Wilmingtonfield sand alkali consumption

(oil present)

pH

Experiment
Slim tube 1
Slim tube 2
Slim tube 3

0.087N NaHC03
0.067N Na2CQj
0.065 N Ne4Si04

SandpaokWil 1
SandpaokWil 2
Sandpackwit 3
Sariip20k Wil 4

0.095N NaHC03 + 0.095 N Na2C03

0.095N NaHC03 + 0.095 N Na2C03
0.095N NaHC03 + 0.095 N Na2C03
0.095N NaHC03 + 0.095 N Na2C03

meqlkg

8.24
10.79
11.72

3.1
6.4
16.0

9.3
9.3
9,3
9.3

5.3
6.3
10.9
4.9

ReSdence
time,
days
30
30
30
.8
.8
.8
.6

SPE 20238

,.

TASLE6.- Reeuke
ofSSHMY
r~lremti teeflng
et0.1%eurfeofent
ooncenfmfbn

Neodol Petmetep
S-102 12-15-3S
229

Neodol

22-9

N-+

Na~

5103

12.15-3s

3.0-3.S 2.4-3.s 3.0.4.1

2.9

5.3

3<0

0.5-2.0 o-2.6 1.5-3.0

N82C03

T~LE7- Injerxbnof optimizedASP

eoMione In aendpecke using WWngfOn fiefd send, oil and brine

N4Pd
3,000
PPX2,000
Wm 1,000
ppm
Teel

Prefluzih
W1.,1

no.
Wifl

POfW)W
Veklme,

Wmr

W12 0.25
W23 0.25
W24 0.26
w-is 0.25
1~~ ~
2M ~ne~

volume

0.2s
0.?5
0.25
0.25
0.25

0.25
0.25
0.25
0.25
0.25
~

chase

P@Y&

PvPv

Pv

Pv

SW,

Meffrw
oensumptk.n Reff0,4 lFr,
%
Mum

Km,

Pv%lmqmg

0.26
0.25
0.26
0,25
0.25

1.s
1.0
1.0
1.0
1.0

44.0 102I0.0053
41.5 1050 0.0083
42.4 460 0.0109
4s.s 975 0.0049
41.3 1400 MA.

41.8
39.9
55.0
30.0
36.9

o.(fg5
N N@.fQ,o,~ N ~~, and3.394
N@l,eX~ W 6W0fIWWW
25% MCI, and FbcOn 4S00
O.WN-,
o,~ Nf.&#q, 0.1%wt@mrii,

AsP

eoMbn
eutimaM2

3.3 Necdof
25.9
3.5 Pelroetep S.105
3.3 Neodol
25.9
%2.0 None
2S2.0Neodel26-S

AsP
t@fbn

atkei#

N@fQ+f@@2
~+

Na#%

f@+=2+w&
m+
f%wl
none

m aF@L
CX b@ZO~r.

VillStl sS@

wee

OKitfadfrOm

Inorder!Onwifnrehcwnperafh bnb efwin@h.

ffm ASP femsdefbn, NsOLaOmWMWZwaefnweaeed
20haeeeobOoneconfofn
2.6%NaC1.
4FfmWV

- (OUF401MW tine

ohamkd

tbrmu

reeking

amer weremd)xloo.

Soo.

23eoy
2400-

NOSURFACTANT
o 0.$%NEODOL26-9ETNOXYLATE

ec4-

ZiOO.
mOO-

4oQ-

1200.

g
%mg

&

13cQ -

~1400-

~.
/
m
o*
o

<~.~
10

20

30

TIME, mm

Flguro1. -Trwtont 1~ of WIlr JrfgtonRcngorZofm011,FM93

100

TIME, mhr

200

Ftgur22. Truwl.frt IFTOfWilmingtonRsngsrZOn2011and0.1%Wodol

2W EthoxylatoSurhotant

SPE 20258

4000

E
g

3000-

~~

NO2URFACTANT

o o.1%PETROWEP3-100

1000

1
9

800

600

400

200

TIME, mln

Ffguro3.. TmnsfontIn of De12wereChlldore

Oil,PH9.S

0 ~400
o
TIME, Mln

Flguro4.- Tmnslmt IFTof fwf2vmre-Chll*m011md 0.1% P2tfoM@Pslm Anionicsurwent

Preflueh
-3% Naci

0.5

1.5

2.5

PORE VOLUMES OF INJECTED FLUID

Ftgure 5. wSodiumSlowbonetees mflueh Addftfvo

634

S= 20238

O.ooc

2345678

Pore volume, PV

Figure6.- SutiaotantTransportIn an Oil-FreeBoraaCore

-i

[
o

0.2

0.4

0.6

0.6

I
1.0

PETROSTEP B-lo5, wt. %

Flgura7. -ScMtyrequir
omentdla
ramforananlonlo au~actantand
w:hrmaarbonats(PH1! .8) or mixtureof sodiumcarboneto
r@aodlumbloerborwt.(PH9.3) t 5% C.

I
o

0.5

1,0

1.5

2.0

2.5

3.0

AES 1215-2SCONCENTRATION, wt %

Ffgurs8.- SaWty mqufrsrwntdlsgmmfor n nlonlc surfaotsntand

sodiumoarbonsto(PH10.8)ors mlxtumof sodiumoarbonsto
nd sodiumMosrbonsw(@f9.S)at W C.

12

10

aQ

.I ,

0.5

1.0

1.5

2.0

2.5

3.0

NE,...>OI.25-9 CONCENTRATION. wt %

Ffgure9.- Ssllrrftyrqu&nantdl

agrsmfor a nonlonlesurfastsntand
sodium osr nsts (@f 10.8) ors mfxturs of sodiumcarbonata
andsodiumWuhonsts (#f 9.s)ats% C.