Guardians of E-Commerce: Harnessing NLP and Machine Learning Approaches for A...
79307422 2-wettability-literature-survey-part-1
1. Nettability Literature Survey—
Part 1: Rock/OiVBrine Interactions
and the Effects of Core Handling
on Wettabiiity
William G. Anderson, SPE, Conoco Inc.
s
Summary. Nettability is a major factor controlling the location, flow, +d .@stributiOn Of fl~d? @ a
reservoir. The wettabdity of a core will affect almost all types of core analyses, including capillary pressure,
refative permeability, waterflood behavior, electrical properties, and simulated tertiary recovery. The most
accurate resuks are obtained when natiye- ‘or restored-state cOr~ Me ~ with native cmde Oil~d brine at
reservoir temperattrre and pressure. Such conditions provide cores that have the same wettab~ky as the
reservoir. ~. .’
The wettabih@’ of, ori&lly water-yet reservoir rock can be altered by the adsorption of polar com~ounds
and/or the deposition of organic materiaf ‘that was originally in tie crude oil. The degree of alteration is deter-
mined by the interaction of the oil constituents, the mineral su~ace,. and tie brine chefi$~. The PrO:~ures
for obtaining native-irate,” clesned, and rsstored-state cores are diSCUSSe4ss we~ aS the eff@s Of cOnng,
preservation, and experimental conditions on nettability. Also reviewed are methods for artificially controlling
the wetmbflity during laboratory experiments.
htrodirotion
This paper is the first of a series of literature surveys
covering the effects”of nettability on core analysis. 1-3
Changes in nettability have been shown to affect capil-
Iq pressure, relative permeability, waterflood behavior,
dispersion of tracers, simulated terdaiy recovery, imedn-
cible water saturation (IWS), residual 01 saturation
(ROS), and electilcal properties. 4-26 For core analysis
to predict the behavior of a reservoir accurately, the net-
tability of a core must be the same as tbe nettability of
the undisturbed reservoir rock. A serious”problem occurs
because many aspects of core handling can drastically af-
fect nettability.
Water-Wet, Oil-Wet, and Neutrafly Wet. Wettabfity
is defined as “the tendency of one fluid to spread on or
adhere to a solid surface in the presence of other irmnis-
cible fluids. ” 7 In a rock/oif/brine system, it is a meas-
ure of the preference that tie rock has for either the oil
or water. When the rock is water-wet, there is a tenden-
cy for water to occupy the smsll pores and to contact the
majority of the rock surface. Siarly, in an oil-wet sys-
tem, the rock is preferentially in contact with the oil; the
location of the two fluids is reversed from the water-wet
case, and oil will occupy the small pores and con~ct tie
majority of the rock surface. It is importunt to note, how-
ever, that the teim wettabWy is used for the wetting
preference of the rockand does not necessarily refer to
the fluid that is in contact with tie rock at any given time.
For example, consider a clean sandstone core that is
saturated with a refined ofl. Even though the rock sur-
face is coated with oil, the sandstone core is still preferen-
tially water-wet. Thk wetting preference can be
coP&h!1986society.+Pe!role.mEngineers
lm’ral ofPetmlc.mTechnology,October19S6’
demonstmted by allowing water to imbibe into the core.
The water will displace the oil from the rock ,surface, ~-
dicating that the,rock sui’face “prefers’? to be in contact
with water rather than oil. Simjkwly, a cow samra!ed with
water is oil-wet if oil will imbibe into the core and dis-
place water from the rock surfuce. Depending on the spe-
citic interactions of rock, oil, and Mne, the wwab@
of a system can range from itrongly water-wet to str022g-
Iy oil-wet. When the rock his no strong preference for
either oif or water, the system is said to be of neufml (or
intermediate) wettabfi~. Besides strong and neutkd net-
tability, a third ~pe is fracdonsl nettability, where differ-
ent areas of the core have different wibing preferences. 27
The wettabfity of the rocklfluid system is impoitant
because it is a major factor controlling the location, flow,
snd distribution of fluids in a reservoir. .Jn general, one
of the flizids in a porous medium of uniform wettabilky
that contis at least two immiscible fluids will be the wet-
ting fluid. When the system is iu equilibrium, the wet-
ting fluid will completely OCCUPYthe smsllest pores and
be in contact with a majoriv of the rock s~face (ass~-
ing, of course, that the saturation of the weting fluid is
sufficiently high). The nonwetting fluid will occupy tie
centers of the larger pores and form globules that extend
over several pores.
In the remainder of this survey; the terms wetdizg snd
nonwetting fluid wilf be used in addkion to water-ivet and
oil-wet. This will help us to draw conclusions about a SYS-
tern with the oppositi wetibility. The behavior of oif in
a water-wet system is very similar to tle behavior of water
in an oil-wet one: For exmuple, it is generally assumed
that for a system with a strong wetting prefererice, the
tietting-phase relative permeab~ky is only a function of
1125’
2. ABLE 1—DISTRIBUTION OF RESERVOIR WETTABILITIES BASED ON CONTACT ANGLE34
Contact
Angle Silicate Carbonate Total
(degrees) Resewoirs Reservoirs Reservoirs “
Water-wet o to 75 13 2. 15
Intermediate wet 75 to 105
Oil-wet 105to 180 1: A 3;
Total 30 25 55
its own saturation-i. e., it shows no, hysteresis. 7,12,28 fected the wettsbflity behavior in the contsct-migle tests.
Owens and Arch&28 measured the gas/oil drainage per-
mea.bflily, where the oif was the strongly wetting fluid,
and compared it with the water/ofi imbibition relative por-
‘meahii@, where the water was tie strongly wetting fluid.
The water-imbibkioi reIative permeability (strongly
water-wet system) was a continuation of the oil-drainage
relative permeability (strongly oif-wet system), demon-
strating the amdogy between systems of opposite wetta-
bilities.
Historically, afl petroleum reservoirs were believed to
be strongly water-wet. This was based on tyo mjor fac~.
Fust, almost all”clean sedentary rocks are strongly
water-wet. Second, sandstone reservoirs were deposited
in aqueous erivironments into which oil later migrated.
It was assumed that the comate water woufd prevent the
oil from touching the rock surfaces. In 1934, Nuttingzg
realized that some producing reservoirs were, in fact, ac-
@ally strongly oil-wet. He found that the quaitz surfaces
of the Tensleep sandstone in Wyoming had adsorbed
heavy hydrocmbons in layers about 0.7pm tldck (about
1,000 molecules) so fdy that they could not be removed
by gasoline or vzrious solvents.”When $e hydrocarbon
film was removed by ,fuing the core, the film could be
restored by soaking the cores in crude oil overnight.
Examples of other reservoirs that are genertiy recog-
nized as sirongly oil-wet are the Bradford sands.of the
Bmdford pool,, Pennsylvania, 30-32 and the Ordovicisrr
ssmds of the Okkiborns City field. 33 More recentfy,
Treiber et aL % used the water advancing contact angle
to eiamine the wettabdity of 55 oil reservoirs. fn this
procedure, deoxygenated synthetic fofmation brine and
dead anaerobic crudes were tested on quartz and calcite
c~stafs at rescrvom temperamre. COnta~.angles (mea.+.
ured through the water) fromOto750 [0 to 1.3 rad] were
dipmed water:we~ from .75 to 105° [1.3 t6 1.83 rad],
intermediate wet; and from 105 to 180° [1.83 to 3:14 red],
oil-wet. As summarized in Table 1, 37 of the reservoirs
tested were classified as oil-wet, 3 were of intemnetilate
wegabtity, and 15 were water-w.et. Most of the oil-wet
reservoirs were mildly oil-wet, with a contact angfe be-
tieen 120 and 140° [2:1 and 2.4 rad]. Of the carbonate
resemoirs included, 8% were water-wet, 8% were inter-
mediate, and S4% were oil-wet. Most of the carbonate
. . reservoim were from the west Texas area, however, so
there is a geographical bias in the data.
Treiber et al. cautioned that these tindlngs could not
be considered representative of a trufy random samplhzg
of petroleum reservoirs. The sainples were bkmedbecause
(1) all were operations for the sarnb company, (2) most
were being considered for some type of flooding, and (3)
some of the reservoirs had demonstrated unusual be-
havior. A fourth consideration is how much the use of
degassed fluids rather than the real formation fluids af-
1126
As ,iiscussed later, this probably causes amoverestima-
ticm of the oil-wetness. Therefore, the large percentage
of reservoirs found to’be ol-wet is less significant than
the general indications that not all reservoirs are water-
wet and that the r&ervoir wettabilky varies widely.
Contact-sngJe measurements made by Clilingm snd
Yen35 suggest that most carbonate reservoirs ramgefiorn
neutrally to oif-wet. They measured the we~bdity of 161
limestone, dolomitic liiestone, calcitic dolomite, arid
dolomite cores. The cores tested included (1) 90 cores
from Asrnari limestones and dolomites from tie Mid$e
EasG (2) 15 dolotite cores from west Texas; (3)3 cores
of Madison liiestone from Wyoming; (4) 4 cirbonate
cores from Mexican oil fields; (5)”4carbonate cores from
the Rengiu’oil field in the People’s Republic of China;’
(6) 16 csrbonate cores from Albe~ (7) 19 chalk cores
from tie North.,SeT (8) 5 samples from India+ amd (9)
5 samples from Soviet oil fields in the Urals-Volga region.
Table2 shows the distribution of wettsbfities with 80%
of the reservoirs either oil-wet or strongly oil-wet. Some.
of the strongly oil-wet reservoirs were oil-wet bew+se
of a bitumen coating. Note that the range of contact an-
gles considered to be neutrslly wet is smaller thari the
range given in Table 1. This demonstrates the variation
, from paper to paper of the cutoff angles between ?Je
different wetting statea.
As discussed iu more detiril later, reservoir rock can
change from its &iginal, strongly water-wet condition by
adsorption of polar compounds and/or the de ositjon of
3Porgsnic matter originally in the crude oil. 7, 6-42 Some
crude oils make a rock oil-wet by depositing a thick or-
ganic film on the mineral surfaces. Ofher crude oils con-
ti pOlar compounds that cm be adsorbed to make the
rock more oif+vet. Some of these compounds are srrffi-
ciently water soluble to pass throu~ the aqueous phase
to the rock.
Fractiormf Wettabilfty. The retilzation that rock wetta-
bflity can be akercd by adsorbable crude oil components
lexito tie idea that heterogeneous forma of wettabiity exist
in reservoir rock. Generally, the intend surface of re.ser’-
voir rock is composed of many minerals with different
surface chemistry and adsorption properties, which may
lead to variations in nettability. Fractional nettability
slso called heterogeneous, spotted, or fhlmation
wettabifky-was proposed by Brown and Fatt27. aud
others. 43* fn fkactioml wegsbdity, crude oif compo-
nents are strongly adsorbed in certain areas of the rock,
so a portion of the rock is strongly oil-wet, whie the rest
@strongly water-wet. Note tit this is cqncepty+flydiffer-
ent from intermediate wettsbtity, which assumes that all
portioim of the rock surfsce have a slight but equal prefer-
ence to being wetted by water or oil.
lomnalofPetmlaunTechmlo8y,@to&r 1986
3. hfi@ Wettibifity. Salatfrie147 introduced the term
ririxed wettabilby for a special type of fractional netta-
bility iriwhich the oil-wet surfaces form continuous paths
through tie larger porca. 48-50The smafler pores remain
water-wet and contsin no oil. The fact that sll of the oil
in a miixed-wettabflity core is located in the lager oif-
tvet pores causes a.smdlbut finite oil permeabtlty to ?x-
ist down to very low oil sahrrdtiom. T@ in mm pertits
the drainage of oil during a waterflood to continue until
verj low oil saturations are reached. Note that the main
distirrctionbetywen mixed md fractional wettsbtity is that
@.elatter,implies neither specitic locations for the oil-wet
surfaces nor continuous oil-wet paths.
SalatMet visualizes the generation of mixed nettability
ih,,tie following manner, When od yufially invaded an
ciiiginilly water-wet reservoir, it displaced waler from the
larger pores, while the smaller pores remained water-fi!kd
becairse 6f capillary forces. A mixed-tiettabihty condl-:
tion occuried if the oif deposited a layer of oil-wet or-
ganic, material ody on those rock surfaces tliat were in
direct contsct with the oil but not on Me brine-covered
surfaces. Oil-wet “deposits would not be formed in the
small water-ffled. pores, snowing them to remain water-
wet. The question that Salatliel did not address was how
the oil first came into direct contact with the rock. As the
oil moyes into the larger pores, a tfdn layer of interstitial
water remains on @pore walls, preventing the oil from
contacting the rock. Under certain conditions, however,
@e waterfdrn separating the crude and the mineml sur-
face cm mpture. Hall et al. 51 sad Melrose52 recently
developed a theoretical model for the stability of these
thiri water films that shows fiat the water ftis become
thiier and thinner as more oil enters the rock. The water
fim is $tabflized by electrostatic forces arising from the
eIectficaf double layers at the oillwater and wateif rock
interfaces. 51’54As tie water “fifmthickness is further re-
duced, a critical tlickness is reached where the water films
in the larger pores become unstable. The films rupture
and are dkplaced, aflowing oil to contact ~e rock.
Native-State, Cleaned, and ReStored-State Coraa.
Cores in three different stitei .of ,presepatioh are used
in core anafysis: native .Xaie,cleaned, and restored state.
The best results for mrrftiphase-type flow mialyses are ob-
klincd with mtive-ststi cmes, where alterations to the wet-
tabfity of the un@rrbed reservoir rock are minimized. ‘.
Ii thk set of paper~ the term “native-state” is used for
niry core that was obtained and stored by methods that
preserve the Nettability of the reservoir. No distinction
is made between cores taken with oil- or water-baked
fluids, as long as the native wektabflity is maintained. Be
aware, however, that some papers dktin=gish on tie ba-
sis of drilliig fluid (e.g., see Treiber et al. 34), In these
papers, “native-state” refers only to cores taken with a
suit8bIeofl-fjjhte.t~e rhill:mg,mud, which nrtintains the
original connate water saturation. ‘:Fresh:state” refers
to a core with unaltered wettabtity that was taken with
a water-base d@ling mud that cbntsins no compounds that
can alter core tiettabdity. Here, the term native-state is
used for both cases.
The second type of ‘core is the cleaned core, where an
attempt is made to remove all the fluids and adsorbed or-.
ganic msterid by flowing solvents through the cores.
Cleared cores are ususlly strongly water-wet and should
Journalof PetroleumTechnology,October19S6
I TABLE 2–DiStribUtiOn OF dARBONATE
RESERVOIR WETTABILITIES35 I
Water-wet
Intermediate wet
oil-wet
Strongly oil-wet
‘contact
Angle
(degreas)
O to 80
.80 to 100
100to 160
160 to 180
Percent of
Reservoirs— .
a
:
15
be used.only for such measurements as,porosity and air
permeability where the wettab~lty will not affect the
results.
The third type of core is the restored-state core, in which
the mtive nettability is restored by a three-step process.
The core. is cleaned and then satrmtcd with brine, fol-
lowed by crude oil. F,@ally,@e core is aged at reservoir
temperature for about 1,000 hours. The methods used to
obtain the.three different types of cnres will be discussed
in more detail later.
Factors Affecting the Original
Reservoir Wenabiiity
The &igiaal strong wa~r-wetness of most,rcservoi ~n-
erals csn be altered by the adsorption of polar compounds.
and/or the deposition of qr tic matter that was origi-
$nally in the crude oil. 7,20,3,32,36+155-63‘fhe surface-
active agents io the oil aze generally believed to be p+w
com ounds that contsin oxygen, nitrogen, andlor SU1-..
!,@r. 6,37,40,41,SS,56,W68These compounds con@in both
a polar and i hydrocmbon end. The polar end adsorbs
on the rock surface, exposing the hydrocsrbqz end and
making the s+face more oil-wet. Experiments haveshown
fhat some.of these aaturaf surfacfants are sut%cientfysolu-
ble iizwater to adsorb onto tic rock surface aftcrpassing
through a thin layer of water. 42,60S69-71
In addition to the oif composition, the degree to.which
the wettabllity is aftcred by fhese surfactants is also deter-
minqi by the pressure, tcmpermmc, mineml surface, and
brine chemistry, including ionic composition and PH. The
effects of pressure and temperature will be discussed Mcr
@the section on experimental conditions. The importance
of the mineral surface is shown by the contact-angle @eas-
orementa +cbssed esrlier, ‘,3s in which a large majority
of tie carbonate reservoirs tested were. oil-wet, while
Many.of the sandstone reservoirs were water-wet: Seveml
researchers have found that some polar compounds af-
fect the wettabilky of sandstone and carbmaitc surfaces
in ~lfferent ~ays, 37,+42,6672-76 The. chemistiy of the
brine can also alter the. wettabtity. Mrdtivafent cations
sometimes enhance the adsorption of surfactants on the
finer~ ~“rface. %77-s3 ‘fhe brine pH is also impOrt8nt
in determination of the nettability and other interfacisf
properties of the cmdelbrinelrock system. 6Z26,~In afka-
Iine floodmE, for example, allmhne chemicals can react
with some c-fides to pr~duce surfactqts that sfter wett?-
b~lty, 6,26
!.
Surface-Active Compormda in Crude Oil. ~ie the
surface-acfive components of crude tie found in a wide
~ge ofWrolem fractions, 41 they are more prevalent
m gze heavier fractions of cmde, such as re$ina and
1:127
4. —
aaphaftenes. These surfactants are believed to be polar
compounds that contain oxygen, nit@gen, and/or suffer.
The oxygen compounds, which are usually acidic, include
the phenols and a bwge number of different carboxylic
acids. 67$5,86Seifert mid Howellsw showed that the car-
boxylic acids are titerfaciilfy active at alkaline PH. The
sulfor compounds include the sulfides and tilophenes,
with smaller amounts of other compormda, such as mer-
~ap~s ~d polysulfides. 87,88The nitrogen compo~da
are generaUy either basic or neutral and include csrba-
xoles, tides, pyridenes, quinoliies, and porphy-
~~. 40.s7-90me pVhYtis camf&m imerfadly active
metal/porphyrin complexes with a number of dii%erentm>
tds, including nickel, vanadium, iron, coppr, “tic, titani-
um, calcium, aud mirgnesinm. 9*-95
Because the .9zrfactarrtsin cnrde oil are composed of
a lsrge number of very complex chemicals that represent
onfy a small fraction of the,crude, identi@g which com-
beeipossible ‘6
pounds are im ortam in aftering the wettabtity has not
h addkion, attempts to correlate bulk
crude propert;ei with the abflity of the crude to ~ter wet-
tabiliw have been unauccessf+ McGhee et al. 62 satn-
rated Berek cores with brink, odflooded them to IWS with
different cnides, then incubated them at 140”F [60”C]
for 1,0+30hours to allow the wet@bfity to reach equilib-
rium. .TheU.S. BWearrof Mines (USBIvf)wettsbllily in-
dex ‘wasthen meaaured and compared with bulk propwties
of the crude. They found no correlation between the
USBM index and interfaci.d tension (IFT), organic acid
content, percerit nitrogen, or percent suIfnr of the cnrde.
oCuiec96 measured the Arnott nettability index of
restored-stnte cor6s and found no correlation between ivet-
tabilky and amonnta of acids, basea, aromatics, resins,
nitrogen, or srdfur. In all cases, when the restored-state
cores were water-wet, the crndes bad low sapha.ftenesnd
anffur contents., However, other low-asphaftene”xnd low-
mdtirr c@des rendered cores neutrally or oil-wet.
Experiments that determined the general mtnre of the
surfactxnts xnd the crude oil fractions in which they are
concentrated without attempting “~ determine exactly
which compounds cause wet@bifity alteration have been
more success~. Jolwnsen and Danningw,9s found that
asphzdtenea were responsible for changing some cmde-
oiWwater/glass systems from water-wet to oil-wet. The
system ,was oil-wet when the cmde was used but wat&r-
wet when the deasphalted crude was used; The addkion
of a very small amount (0.25%) of the whole crude to
Ozedeasphalted crnderestored the oif-wettingness of the
system. Donafdsbn aud Crocker55 ‘and Donafdson56
measured wepabilky alte@ion caused by the polar coin-
poaada extracted from aeveml different mincmf oifa. First,
the nettability of a series of uizcontanzinated Berea plugs
was mcasnred”with brine agd a refied mineral oiL The
average USBM wettabilky iizdex was 0.81, or strongly
water-wit. After cleanirig, the USBM nettability index
of the plugs was measured with brine and a 5 % mixture
of the extracted polar organic componnda in the retined
mineral oil., The phtgs were significantly leas water-wet,
with USBM wettabtity indices ranging from 0.45 (water-
wet) to -0.W (xretitmlfy.wet), demonstrating that polar
compmmda in cnide can alter the wettabtity. Note that
there was appsfentfy no aging time with the polar cOm-
ponnds in the plugs, so equilibrium wettxbifities may be
more strongly oil-wet.
/.,’
112s
,,
Several researchers 57,5s analyzed wetkbiliw-afterirw
compounds extracted from cor~. Jennings58 r~moved ~
portion of the nettability-altering cornpotirids by extract-
tig a non-water-wet core with tolnene,,followed by a cfdo-
rofordmethanol mixiure. An imblbitioiztest showed tbst
some of the wettabil@-idtering compounds had been re-
moved during the s~ond extraction because the core was
more water-wet. The material ‘remo+ed dtiring the sec-
ond extraction contained porphytis ~d bigh-moledor-
weight pamffinic and aromatic” cbmpoiihds.
Denekaa et al, 41 used a d~tiflntion pro;exs to separate
cmde oifs into fractions of different molecular weight.
A clean, dry core wax saturated with the crude oil frac-
tion to be tested, then aged for 24 hotii:. An imbibition
test based on the relative rate of imbibkion waa nsed to
determine the wettxbility Alteration..1“.1 The originxl
crqde oil and the heaviest residue left a,fterdistillation had
the greatest effect on the nettability; they were the only
fluids that nzadetbe rock oil-wet.’fhia fiplies that a con-
siderable portion of the surfact+ds, in the crude oiI had
a large molecular weight. Many of the lower-molecular-’,
weight fractions, however, also debreaaed the water-
nettability, demonstrating that the surfactants in crude
have a broad range of molecufa.r weights. Cuiecg6 ob-
tained similar results. Note that Denekas et al. and Cuiec
botlr used@ cores and that idso~tiob of the w&ttnbility-
.ikering cOnzpoundswoufd probably have been ilteredif
the cores contained brine during tie aging “process..
A number of roaearchera have examined the i@rfacinUy
active materiak that are cOncentrat&dat the Oil/water in-
terface. Generalfy, these materi~s can iilso be adsorb@
on tie rock surface to sfter wettsblli~. 3.7,8~g9-*mBar-
tell and Nlederhauser 103managed @ sepirate these,ma-
terials from ~e crude oil dqd f6und that they, formed a
hxfd, black. noncrys@fline substxnce tit was sapI@tic
in nature.
. .
Adsorption Through Water F~s. Experiments hove
shown tfiat nataml surfactants. ii cimde.are often suffl-
cienify soluble in water to a&sorb onto the rock surface ,-
&r passing tbrougb “a thin ld~er. of water. 42,60>69-71
Meaauremerits comparing aaphd~ene a+orption in cores
with and without water show tlwt m many.caaes a water
film wiUreduce but not completely ir@it =pk+lterfe”nd-
sowtion. 6M9,70 Be&use the wa~ei .qzdasphaltenes wM
coadsorb, however, the wat4.rfilm may alter tie detailed
adsorption mechanism. 70,1w Lyutin @d Burdynm found
that the asphalt&e adsorption f$ornA&h ,crude in an yn-
consolidated sandpack wai about 80% of the dry y~ue
at a water saturation of 107. PV, decreasing to 40 % when
the water sapration was incrth+ed to 30% PV. Berezin
et al. 69 exmzzizzedthe adsorption of .&phalten&s&d ;ex-
ins from crude onto cleaned sandstone cores. With Tui-
mazy crude, a water,satnration of about 17% reduced the
adaoz’ptionby about a factor. of three. With two other
crndes, a water saturation of about 20% completely. in-
hibited the adsorption. Such complete inhibition by tie,
water fifm woidd be expected in reservoirs that remain
water-wet, with no significant ,adsorption:from the cmde.
Reisberg ad Doscherbs aged clean glaas slides in
c~de ofi.floating above brine and observed the forma-
tion of oif-wet fti. The formation i.nd s~bility of the
oif-wet fti on the slide was observed by lowering .!-be
slide into the brine and observing whether the brine dis-
Joumalof PetroleumTechnology,October1986‘
5. “placed alf of the cmde ofl from the slide. They first aged
a clean glass slide in crude mrdfound that a film, d6posited
over several days, made the slide moderately oil-wet.
They modified the experiment by immersing the slide in
water before aging it in crude. Surprisingly, the oil-wet
fti formed much more rapidly. when a NaCl solution
was used instead of water, the slide also became oil-wet,
but it was necessary to age the slide for a longer period
of time.
Sandston& mrd Carbonate Surfaces. The types of min-
eral surfaces irr a reservoti are also im ortarrt in deter-
Emi@rg wettabfity. Both Treiber et cd. and ClriHngai
ad Yen35 found that carbonate reservoirs are wpi@y.
more oil-wet than sandstone ones. Two other sets of ex-
periments show that the mineral surface interacts with the
cmde oil imposition to detefine wettabWy. The first
set e~nes the adsorption onto silica and cm’bonatesur-
faces of relatively simple polar crrmpoundy the second
set exnmines the adsorption of crirde.
Simple Pokr Compounds. Whsin the effects of brine
chemistry are removed, silica tends to adsorb simple or-
gtic bases, wh~e the carbonates tend to adsorb simple
~rgaic tilds. 37,Q83 This occurs because silica normallY
has a negatively charged, weakly acidic surface in water
nenr neutral pH, while the carbonates have positively
charged, weakly bssie surfaces. 37,@,83,105
These surfaces will preferentially adsorb compounds
of the opposite polarity (acidity) by m acidbase reaction.
Wettabili~ of silica will be more strongly affected by the
organic bases, whfie the carbonates will be more strong-
ly affected by the organic acids. This was found to be the
CW+in experiments on the adsorption and nettability al-
te~tion of refntively simple polar ,compormds oir sand-
stone and carbonates. The compounds were dksolved in
a nonpolm oil, and the contact angkeof the oillwatcrhnin-
eral system was measured on mr initially km, strongly
water-wet aystal surface. &xreraHy, adsorption mrdwet-
tabifity nkemtion occurred with basic compounds on the
acidic silici surfaces and acidic compounds on the basic
ccrbomte surfaces. Acidic compounds had very little ef-
fect. on sifica, and basic com ounds had litde effect on
ffieCabonat=.37A42,66.7S.7’J.o~l~Note,however, that
most of the ,adsorbed compounds chmrged the wettabfiV
only from strongly to mildly water-wet, rather than to
oil-wet.
The acidic compounds that adsorbed and nftered the
wettab~l~ of the carbonates in preference to silica in-
cluded naphtherric acid 37.109mrd a number of carbox$;
ic acids (RCOOH), including capryfic (octanoic ,
psfmitic (.hexadecarroic), 42 stearic (@adecanoic), 210,10s
and oleic (cis-9-octadecnnoic) acids. 42 Basic compounds
that adsorbed on the acidic silica surfaces included iso-
qrdno~ne37and octadccykmzine [CH3(CH~17NH~. 1ffi,108
@mz40 and Morrow ei al. 66 exded dze adsorption
and wettabilhy alteration on quartz “imddolomite of a num-
ber of relatively low-molecukw-weight compounds found
in crude oils. Basic nitrogen compounds gave advancing
contact angles up to 66” [1.15 rad] (water-wet), with
higher angles for qrrnrtzthan dolumite. Sulfur compounds
tested provided arzgfes of 40” [0.7 rad] or less with
rto systematic dlffercrrces between the two surfaces, The
contact angles either were stable or decreased with time
(i.e., the system became more water-wet). The acidic
oxygen compounds gave higher airgles on dolumite.tbsn
quartz, up to 145° [2.5 rad] for. octanoic acid
[CH3(CH2)6COOHJ and up to 165° [2.9 rad] foc law-
ic acid [CH3 (CH2),0 COOH]. Note, however, that tbe
uxygen-containing acidic compounds appesr to react
gradualfy with the dolomite, so the contacf angles ae un-
stable and the sy$em gadu?dly becomes more water-wet.
Cram if.ail noted that none of the relatively simple com-
pounds they tested could create a stcble, oil-wet yu’face.
There fure, they cuncluded that the compounds respomi-.
blefor wettabihty alteration in crude were higher-weight
polar compounds and other po+ons of the asphaltenes
and resins.
In the inure complex crudeibrinelrock systems, the nrin-
er# surface will not necessarily have a preference for
compounds of the opposite acidhy. The simple systems
dkcussed here tested each surfactcrd individpaUy and re-
moved the effects of brine chemismy. @ tke section on
brine chemistry, it will be shown thnt multivalent cations
cnn promote the adsorption of surfactants with the same
acidigi as the surfcqe. fir addition, “theadsorption of auy
single surfactant in the crude might’ be enhanced or
depressed by the adsorpion of other compounds.’
Adsor@on From Crude. A nnmber of researchers
found differences in the adsorption of crude oil cum-
punerits onto dry sandstone and carbonate sur-
faces. 41,72-74,l@,110Denekss et al. 41 sepmated out the
acidic and basic orgaric compounds from crude arrd test-
ed them in initially clean, dry cores by tie method de-
scribed earlier. They found that the wettabfity of
snndstone was altered by both the acidic and basic com-
pounds, while the Iiiestone was more sensitive to the ba-
sic nitrogenous orgmric compounds.
Several experimenters have compared the adsorption
of asphnbenes from crude onto initially clean, dry sazrd-
packs composed of either quartz or disaggregared” core
material that contained both quartz and carbonate. 72,110
They found that adsorption was greater in disaggregate
core material. Tumasyan and Babclyan 110.measured the
adsorption, of asphaltenes from Kyarovdag cmde onto
quartz and cleaned, disaggregate Kynrovdag core m~-
terial @atconticd 10.4% carbonate. The adsorption wnc
about 8x 10’4 mg/cm2 for qumtz nnd about 18x 10’4
mglcmz for the core material, mr iacrease of more than
a factor of two: Abdurashitoi ef al. 72 meaarzzedthe ad-
sorption of asphnfte”nesonto sida-sized fractions of pure
quartz samdsand sends containing both quartz and car-
bonate. They found that the adsorption on the qunrtz sands
was as much as an order of magnimde lower Own the ad-
sorption on the sands containing both minerals. These re-”
srrks are very qualitative, however, because the speCXc
surface arei of the quartz packs was lower dum the area
of the mixed mineral smrdpacks, which afso reduces the
~onnt of adsorption.
Brine Chemistry. The salhrity nrzdpH of brine ae very
imw-t in determining wettabiilg because they strongly
affect the surface charge on the rock surface and fluid in-
terfaces, which in turn can affect the adsorption of .srrr-
factants. ‘o,‘m Positively charged, cationic srrrfactanta
wifl be attracted to negatively chnrged surfaces, while
negatively charged, anionic surfactcnts will be attracted
to positively charged sin-faces. The surface charge of Q.
ica and ccfcite in water is positive at low pH, but nega-
Journd of Petml..m T-hnology, October 1986 1129
6. tive at high PH. For sifica, the surface becomes negatively
char ed when the pH is increased above about 2 to
!3.7, 3,105 whfle calcite does not become negatively
char ed untif the pH is greater than about 8 to
9.5. &,’05.111As discussed in the previous section, sili-
ca is negatively charged near neutral pH and tends to ad-
sorb organic acids, whiIe cafcite is positively charged and
tends to adaorb organic bsaes. Calcite will adsorb cation-
ic sm’f@mJts rather than anionic surfactants., however,
if the pH of the solution iJJwhich it is immersed is in-
creased above 9.111
The pH also affects the iotiation of the snrface-active
organic acids and basea in the crude.’. fn afkahe water-
floo’iling, a relatively inexpensive caustic chemical–
typically sodium hydroxide or sndium ofiosificate-is
added to the injection ,water. 112The hydroxide ion mscts
with organic acids in acidic crude oils to produce surfac-
tanta that after the wettabiti~ ador adsorb at the oilhine
interface to lower f.FT. Seifert and How.ells85 exsmined
the interfaciafly active materials in a California cnrde oil.
They found that the crude contained. a large mount Of
carbnxylic acids that form soaps it ilkaline PH.
The possibfity of EOR during amalkaline flocaldepcnda
on the pH and salinity of the brine, the a.sidi of the crude,
?’and the origiwd wettabtity of the system.2,113,114Cooke
et al. 6 discussed the effects of salkhy nn wettxb~ky in
alkaline flonds where the soaps are formed by the inter-
action of the alkaline water with the acidic ‘crude oil. In
relatively fresh water, the snaps that are formed are soht-
ble in water; promoting water-wetness. If the system is
initially oif-wet, EOR may occur b a wettabdity rever-
.?~~ from ~fi..wet to ~a~r-wet. 17.2 ,114J 15 On the ~ther
hand, in high-salinity systems, EOR may occur as,a re-
sult nf awater-wet-to-oil-wet wettnbifky reveraaf. Aa the
saliniw is increased; the soaps become almost in.duble,
adsorb on the rock surfaces, and. promote” oil-
~en~g, 6.113If the system is initially water-wet, COok
et al. ststf that EOR in a highly sshme~ystem may oc,cur
by a water-wet-to-oil-wet tiettability reversaf mech-
~sm.6 ,113,114
In silkdoillbrine systems, multivalent metsl cations
in @ebrine can reduce the solubilky of the crude sur-
factants andlor promote adsorption at the mineraf sur-
faces, causin the system to become more oil-
?wet. 6,W,77,79,8.116,117Multivalent metal ions that have
altered the wettabifity of such systems include Ca ‘2,
Mg’2, Cu’2, Ni’2, and Fe’3. Treiber et al. 34 exam-
ificdthe effects of trace “metalions in the brine on the wet-
tab~hy. They meakured the contact angles on quartz of
dead Werobic cmdes in deoiygenated synthetic fofma-
tion brine and found that as little as 10 ppm of Cu’2 or
Ni’2 coufd change the wettab~ky fcom watdr-wet to oiI-
wet. Brown and Neusta.dterm placed cmde oil droplets
in a contact-angle apparatus tilled’ with dktifled water.
They found that Oreaddition of.less than 1 ppm of Ca’2
or Mg’2 would alter the nettability, making the system
more oil-wet. The addition of trace amounta of Fe’2 also
changed the wettabiIity with some of the crudes tested.
The.$emultivalent ions have SISObeen ahown to incresae
the od wetness of soils stab~ied with cut-back
xaphalt..11S,119 (Cutback ~~~t ‘is ml MPildt ==ted ‘iti
an inexpensive solvent, such as gasoline, to reduce the
viscosi~.) HaiIcnck11s - Seved strongly water-wet
soils with cut-back asphalt. He fokd that the oil wetness
of the anil after the.asphalt tre@zent waa greatly increaaed’
by pretreating the soif with a solution of ferric or @mi-
IJU222sulfate.
Morrow et al. xl aged glass sfidca in Moutray crude,
waahed the slides to remove the bufk crude, and then used
isooctafre and distilled water. to meaaure tie water-
advancing angle. They fonnd that the wettabifity strong-
ly depended on the sfnount of trace ions iJJthe system.
When the glaas slide was extremely clean, no residual fb
was depnsited by the crnde, and the system was water-
wet. Next, they treated the glass with femic (Fe’3 ) or”
orler transition metal ions before exposing it to drecnide.
They obtained contact angles up to 120 to 140” [2.1”to
2.4 rad], with the angle dependent on the choice of ion’
and its c&rcentration. The”ferric ion was pxrticulady ef-
fective in altering the wettabilky.
There appexr to be two related reasons for ~e effects
of these mnftiv#ent ions on the wettabili~. Fiiat, they ~,
can reduce the snlubility of the surfactants in the crude
andbrine, helping to ‘promote oil-wetting. 6,113Second,
they behave as “activators” for the smfactants in the
crnde. “Acdvator” .la a term used in the floption indus-
try for ions or cnmpounds thst, while not snrfact@ them-
selves, enhance surfactant adsorption on the mizreraf
smface snd increase the flnatabifity. Generally, the acti-
vators act like a bridge between the mineraI surface and
tie adsorbing surfactant, helping tn bmd the snrfactant
to the NIffsce. go AS shown previously, clean @J~ haa
a negatively charged surface nnd tends to adsorb @osi-
tively charged) orgnnic bases from solution. The (nega-
tively chsrged) acida in solution will not adsorb on the
surface because they will be repelled by the like charge
on the quartz surface. For example, clean quartz is not
floated by fatty acids, ideating that tie quartz remains
water-wet. At the proper pH conditions, however, the
wettabfity can be changed and the quartz can be floated
by the additidn of smafl nmounts of marry multivalent
metalfic cations, including Ca’2, Ba ‘2”, Cu’2, Al’3 i
ad Fe +3. WT%Q.107These ions adsorb on tie qun3t2
surface, providing positively char.r@ sites for the adso%
tion of”&e fstty ;~ds. -. -
,-.
For exam 1. Ga.din and Chang7s “imdGaudm and
Fuerstemm% ‘staled the adsorption of laurnte ions on
auartz. When sodium Iaurate, CHa (CH2.)toCOONa, is
~ded to the water, it dissociates @~ a ne~ati;ely charged
laumte ion and a positively charged Na’ ion. Because
quartz develops a negative surface charge as a result of
the dissociation of H + ions from the Si-OH groups on
ihe silica surface, the negatively charged laurate ion is
repelled ffom tie negatively charged quartz surface.
Hence no adsorption occurs. However, adsorption occurs
when, for example, divafent Ca’2 or Ba’2 ions are
added aa the activator. These positive djwdent ions can
adsorb on the surface, allowing the negatively chmged
surfaciant (in this caae, the laurate ions) to adaorb in as-
sociation wia them. Researchers with other experimen-
tal systems also state that divalent ions can bind to a
negatively charged”surfactant to fogn a positive, cationk
surfactant/metal ~mplex, w.bich is then attracted to and
adsorbs on the negatively charged quartz snrf?ce, 116,117
“CIays. Several researchers have studied tie adsorption
of aiphaftcnis Wd resipa onto clays, and found that.ad- ;
sorption can make the clays more oil-wet. 70,76.1~, 120-133
,.
1130’ JournalofPmol&mTechnology,October1986
7. Clementz %120,121 eximined adsorption under mhy-
droua conditions of the heafl ends—the nonvolndle, high-
molecuiar-weight fraction-of cmde oil, which are
primarily asphal~nes snd resins. He found that rhe com-
pounds adsorbed rapidly onto montmorillonhe, forming
a stable clay/organic compound and chsnging the netta-
bility from water-wet to oil-wet. Clementz also looked
at adsorption under anhydrous coiiditions of the heavy
ends onto Berw’corca that contain significant amounts of
kaolinite. The adsorption of tie heavy ends made the core
neutrally wet as determined by an imblbhion test. The ad-
sorption also,reduced the expansion of swelling clays, clay
surface area, cation exchange capacity, and water sensi-
tivity. The ‘materials that adsorbed onto both the mont-
morillonite and kaolinite were difficult to remove,
although most of them could be exmacted with a chloro-
formlacetone mixture.
Clementzused dry cores and clays. As discussed C&
er, the presence of a water fdm will gen@ly reduce the
adsorption of wettabfity-altering materials, typically by
a factor of two to four, although in some cases, it will
completely inhibk adsorption. ‘>@ Collins and Melmse70
measured the adsorption onto kaolinite of asphaltenes dk-
solved in tolucne. The dry clay adsorbed a maximum of
about 30 mg asphaltene/g clay. The addkion of 6.6%
water t? the clay reduced the adsorption to 13 mg/g. In
addition to reducing the adsorption, the water l@ may
alter the d:tied mechanism of asphaltene adsor@on be-
cause the asphaltenes and water will coadaorb. 7 For ex-
ample, in contrast to his work with anhydrous cores,
ClemenE found that the adsorption of asphaltene onto
Berea cores in the presence of water dld not reduce the
water sensitivity of the kaoliidte. lW
Non-Water-Wet Mineral.% When all of the surface con-
taminants are carefully removed, most minerals, includ-
ing quartz, carbonates, and sulfates, are wrongly
water-wet. go,1o7,lx From flotation studies, however, a
few minerals have been found that are naturally but weakly
water-wet or even oil-wet. These minerals include sul-
“fnr; graphite, talc, coal, snd msny sulfides. Pyrophyllhe
and other talc-like silicates (sificates with a sheet-liie
&rncture) are ~obably also neutrally wet to Oil-
~et, 30,107,134-12~ese finer~~ ~ ~OW~ tO be SO~e-
what hydrophobic because air can be used to float them
on water ir3froth flotation, implying a large waterlairhnin-
eral contact angle. Because they are non-water-wet with
air, it is probable that they are also oil-wet.
On the bsais of core-cleaning attempts in a limited num-
ber of reservoirs, it appears that cores containing conl are
sometimes na.tur@ly neutrally wet because they can be
cleaned only to a neutfsll wet condition rsther than a
z~Uon lY wa~r.wet one. 1 a,129 Cuiec96 and Cuicc et
8al. 13 cleaned unpreserved cores with different solvents
and then measured wettabdity. In four cases where cores
contained large ixnounts of unexmactable organic carbon,
they were able to clean the cores only ti neutmf nettabil-
ity. Wendel et al. 12s cleaned core from the Hutton reser-
v6ii contaminated with an invefi-oil-emulsion drilling
mud. Core from most zones in thk reservoir could be
cleaned to a water-wet state. However, in one zone that
contained si=tilcant amounts of coal, the core was neu-
trally wet after cleaning. About 50% of the rock surface
in tie neutrally wet zone was covered by a thin layer of
organic matter less than 300 ~ [30 mn] thick. ThISfayer
Journalof PetroleumTechnology,October1986
--
may be a:rcauftof dif3i3si0nof organic cmnpounda released
during diagenesis from the small, organic, detrital pard-
cles of cod scattered throughout the zone. Unforhmate-
ly, this is unclear at present. Thin sections from both
water-wet and neutrally wet (tier clcaqing) zones show
that both contain approqtely equal amounts and dis-.
@bution of woody coal; algae coal, and pyrite. Conse-
quently, it is unknown what causes the postcleaning
neutral tiettability of this neutrally wet zone.
Boneauand Cfampitt 131and Trantham and Cknnpitt 132
stite that the oil wetness of the North Burbank unit is
caused by a coating of chamosite clay (Fe3Al zSi2O IO
.3HzO) on the pore surfaces rather than the more com-
mon, strongly adsorbed organic coating. The chamosite
clay, which is iron rich, covers about 70% of the rock
surfaces It seems plausible that tbe chamosite cfay renders
the core oil-wet because, as discussed earlier, iron ions
are strong activators, promoting o~-weufng. Cl~pitt*
states that unpubliahti contact-angle measurements made
with all of the ~emls “inthe No&Burbank core ahowed
that chamosite M naturally oil-wet.
Artificial Variation of Wettabifity
Aadescribed previously, a native-state core contains a
cmmplexmixture of different compounds that can adsorb
and deaorb, possibly altering the wettabdity during an ex-
periment. Msny re8esrchers have tried to simpfify their
experiments by artificially control~mg the wettabfii~ to
some constant, uniform value. The three methods most
commonly used are (1) treabnent of a clean, dry core with
various chemicals, generally orgWOcMOrOsfl~es “fOr.
sandstone cores and naphtienic adds fOrwbOnat~ core~
(2) using sintered cores witi, pure fluids; and (3) adding
surfacthnts to the fluids. A sintered tdlon core with pure
fluids is the preferred method to obtain a uniformly wet-
ted core because tic wetrabili~ of these coma is constant
snd reproducible. The fiettabilhy of cores treatid with
brganochlorosilan:s, naphthenic acids, or stirfactanta is
much more variable because it afso depends on such vari-
ables”as the chemical uked, the concentration, tie treat-
ment time, the rock surface, and the brine PH. These
treatments have advantages, however, when he~rOgen~-
ous wettabtity or wettsbil~ alter~tion is studied.
Organochlorosffanes and Other Core Tr&dmMa. One.
method of making a sandatone core uniformly non-watcr-
wet is to treat it with a solution conpining an OrganO-
~~oro~~e ~ompo”nd, 133-139vm~tiom of ~~ tf~~t-
ment have also been used to create fractionally wetted
~mdpacf&&+$50.1@h!2~d mixed-wet cores. 143me ~r-
ganosilane compounds contain silicon molecules with at-
tached chlorines and non-water-wet organic groups, with.
the general formufa RnSiC14-n where R is usually
methyl or phenyl and n =0, 1, 2, or 3.133 These sub-
stances ,react with the hydroxyl (OH) groups on silicon
dioxide surfsces, exposing the organic groups and ren-
dering the surface non-water-wet. For example, dimerhyl-
dichlorosilane, (CH3)2SiC12 (Drifilm@ or Teddol@ ),
chemisorbs on the outside of the sificate lattice of glass,
eliminating HC1and exposing CH groups, which reduce
hthe wata-wetness of the surface 1 Other compounds il3-
clude heximethyldkilazane 138 and trimethyMdo!Osi-
kme. 145The wettabfily of the core is altered by flowing
.P.rsmtic.nnm.nimtimwithR.L.Clmnpiu,Ptiltip Petroleum,BaII[esvilleOK,
Dac.1S83.
1131
8. a solution of the organosilane through it, snowing a suffi-
cient time for the reaction to occur, and then flushing the
unreacfed compound from the core. Some control of the
change in we,ttabifi~ can be achieved by variation in the
concentration of organosilsne in the solution. For a com-
plete description of the method, see Ref. 134.
In addition tu unifoqnly treating cores, organocldorosi-
kmca are used to prepare fractionally wetted sand-
Packs, 43,46.50,140-142 Sand gtins txeated with 6rgm30-
chlorosilmies sre mixed witkuntrcated; water-wet sands.
‘The fraction of oil-wet surface is aasumed to be the sb.me
as the fraction of orgnndcblorosilane-treated sand. One
problem, however, is that some of the organocblorosi-
lane is known to be transferred to the water-wet sand
grains, likely changing their wettabtity.43 Auother
method of obtaining fractioml nettability is to form ‘tie
porous medium from water-wet (gkMs)beads and oil-wet
(tcflon) beads. 146
Moh~~ ad s~tcr 143have reqedfy publiahcd a tech-
nique to generate mixed-nettability cores so that the large
pores have continuous’ water-ivet surfacca, leaving the
smalf pores ofi-wet. Note that in these cores, the netta-
bility is reversed from Salatbiel’s47 mixed-wettability
cores. Cleaned cores are first treated witi orgagosikmes
to render them uniformly oil-wet. The treated cores are
saturated with oil, flooded with heptadecane to dkplace
the oil, and then flooded with brine to ROS. Because the
core is oil-wet, the large pores are fdlcd with brine, but
the small ones are fdled with oil. Brine and heptadecane
IMY then be injected simu&aneously to alter the fraction
of pores ffled with oil or water. After the desired satura-
tion is reached, the core is fust placed in a cold water
bath (50”F [1O”C])to freeze the heptsdecane, then an 11.5
pH sodium hydroxide solution is injected to diapfimc the
brine. Mohanty and Sslter state that the alkaline solution
removes the orgnnosilane coating from the lnrger, brine-
tilled pores, leaving them strongly water-wet, while the
frozen heptadecane prevents auy change in nettability in
tie small oiLffled, oil-wet pores. Folly, t+e alksdineso-
lutiori is displaced with brine, arid all of the fluids are re-
moved, leaving a mixed-wet@ ility core. After this
treatment, the cores imbibed both oil and water, indicat-
ing that areas of the. core were both water- and oil-wet.
Unfortunately, Moh&y and Salter did not test the cores
: by oil flooding them to determine whether they had a very
low water.saturation after tie injection of “manyPV’s of
oil. This would have verified the formation of continu-
ous water-wet paths through the large pores, which would
be analogous to oil-wet paths in Sala~el’i cores.
One problem with orgamochlorosilane treatments is that
the nettability of the tmatcd core varirs depending on such
variables as the orgtiocidorosil~e used, the concentra-
tion, the treatment iiine, the time elapsed since the sur-
face was treated, and the pH of the brine. 147 No
dependable treatment has been reported for acgeving a
given: core wettahilhy. Note that many organosilane-
treated cores ze only neutrally to mildly oil-wet, instead
of strongly ofl-wet. Coley et al. 134used General Elec-
tric Co. s~lcone fluid No. 99 in concentrations ranging
from ‘0.002sto 2.0% and were nhle to vmythe contact
angle, in. glass capillaries only from 95 to 115° [1.7 to
2 fad]. RathnieO et al. 137found that cores treated with
dmethyldichlorosikme would still slowly imbibe water,
indicating that the cores were at most neutrslly wet. In
1132
~onmt, Newcombe et d. 136 stated that contact an~es
sa large aa 154” [2.7 rad] could be obtained for aK1casur-
faces mated with different conqamations of methylsilOx-
ane polymer, but these contact angles tended to decrease
towsrd 90” [1.6 rid] as they aged. Memwat $?ral. l@
tmatcd silica surfaces with various concentrations of four
different organochlorosilanes and obtnincd contact angles
ffom 75 to 160” [1.3 to 2.8 rad] with water and xylene
on the treated surfaces. Depcndmg on the specific treat-
ment, they found that the contact angle could gradually
increase or decrcaae aa the system aged. Because&e wet-
tabtity of cores treated with organosikmes can range from
mildly water-wet to strongly oil-wet depending on the Spe:
cific treatment, the Amott or USBM method shoufd be
used to determine the wettabllity of the treated core.
Quilon@ treatments are mother method that has been
used to alter the wettabiIity of sandstofie cores. Tiffii hnd
Yellig 149treated Berea cores with Quiion-C@ to render
them uniformly oil-wet. Workers at tie Petroleum Recov-
ery Inst. have used Quilon-S@, i related com-
~md, 15~153 me won compounds consist of a chrome
complex containing a hydrophobic fatty acid group in an
isopropyl alcohol solution. When @iiori is injected into
tie core, the molecules bmd to the surface, expose the
fatty acid group, and render the rock surface oil-wet. 1X
Note that wettabtity of the treated core probably v~ies,
depending on concentration, tfcatment time, etc., so it
should be meaaured with the USBM or Amott metbnds.
In many crises, the treated core is probably OIIIYneutral-
ly to mildly oil-wet.
These trcatmeritahave been used on sandatone core with
the chemical binding to the s~Ica surfaces. Orgsno-.
chlorosikme treatments, which adsorb on silica surfaces
by reacting with the hydro+yl grou$a, are generally not
effective on carbonate surfacea. 1 5,1% A number of
~Wche~~ 109.155-157bWe used m hthenic acids to
render carbomtc cores more ofl-wet. ?7 The naphtle~c
acids react with the calcium carbonate to form cslcium
naphthenates, which are ofi-tiening. 109 Note that
naphthenic acida will not nlter the wettabtity of saudatone.
s~faces. ‘w
Sharma and Wuuderlich15s altered the nettability of
Berea pIugs by saturating’ them with au asphaltic crude:
Drv Dlum were vacuum-saturated with isrhaltic cmde oil,
th&’fl~shed with pentane, which ten& to precipita~
asphakenca onto the pore walls. 67 The pentane was”re-
moved in a vacuum, kwiug behind a layer of asphaltenes.
Tim plugs probably had mixed wettnb~ky after Ecatmenc
boti,oil snd water would imbibe spontaneously. 3 An ad-
vsutage of thk method is that it uses com@mda found
natnrslly in the resetioir and ,pight be a more realktic
ticatrnent than the other trcahnenta discussed above. Note,
however, that it is necessary to verify tiat the crude is
compatible with the pentie because some cmdes will plug
the core when pentane is injeded.
Artificial Coma. Several rcaearchera have used artificial
cores and pure fluida to control wettab~hy. The uniform
composition of tie core and the absence of surfactimts
combine to give a constant; uniform, and reproducible
nettability. The most popular material for the artificial
core baa been polytetiafluoroethylene (teflon). Stegemeier
fid Jess&n159used porous packa of tcflon particles. More
recent experiments hsve used consolidated teflon
Journal of Pekole.m Technology, October 1986
9. cores, 160-’68 which are prepared by compressing teflon
powder and sintering it at elevated temperatures to
produce a consolidated core. Mungan 167 completely
describes the process. Lefebvr.e du Prey lm has 81s0us<d
Sinter.Sd5tifle88 8tee] and altina cores. ~,
Teflon is preferred for two reaaons: it is chemically inert
and has a low surface energy. 16gMost minerals found
in reservoir rock have “ahigh sutface energy, so”abn08t
all liquid8 will spread,,op and wet them against $r. The
w&abiMy of such high-energy solids must bec$mtroUed
with either adsorbed fti on the solid 8urface or 8urfac-
tants in the fluids. Both of these inethod8 raise the prob-
lem of changes in the wettab~ty during the experiment
as a result of adsorptionldesorption phenomena. On the
other hand, the surface energy of teflOn iS low enouti.
that a wide range of contact angles can be obtained with
various combinations of pure fluids that do not contain
surfactants. The u8e of pwe flbids with teflon also avoids
difflcukies with contact-angle hysteresis associated with
adsorption/desorption equilibrium and the problems as-
sociated with contact angle and ‘IFT aging phenomena.
This”is dlscus8ed in more detail in Ref. 1. fiany experi-
ments in tetlon cores use air or Nz anfl various fluida to
va~” the contact angle. Contact angles from Oto 1080 [0
to 1.9 rad] can be obtained by the proper ChOiCeof liq-
161For ~x~ple, an air/water/teflOn sYs-ui~gas pati8.
tern has”a contact angle through the water of 108° [1.9
rad]. Lefebvre du Prey 160 used mixtures of water,
glycerol, glycol, and alcohols to represent the water pbaae
and mixtures of pure hydrowbons for the oil phaae. Con-
tact angles through the oil phase of from Oto 168° [Oto.
2.9 rad] were reported for hk teflm, steel, and alumina
cores.
Surface-Active Agents. The use of clean cores and pure
tkdd8 with various C0nCenmati013Sof a SiJigleSurfactant
‘isthe third way that re8e81Cher8have controlled the net-
tability of cores: Owens ,and Archerzs used biwium
dlnonyl 8ulfonate in the oil and reported stable contsct
angles up to 180° [3.1 rad] on a quartz crystal. Morrow
et al. 66 were uriable to reproduce this work, finding a
strong time dependence for the contact angle. They tried
to control the wetfabtity with octanoic acid, obtaining@-
gles from Oto 155” [0 to 2.7rad] on dolomite. They found
that the wettabflity could be maintained for less than a
day, however, after which tlesystem became increa.$ingly
water-wet as tie octanoic acid 81OWIYreacted with the
dolomite.
A number of res&hers 17,26.17&174haVe I18Sd8mkJeS,
R-NH2, to study EOR caused by Wettab~hy dtemtion iII
laboratory &#.wflmds. Wwtability reversal from Ofi-wet
to water-wet and. from water-wet to”oil-wet are two of
the proposed mechanism for enhanced recove~ during
alkaline waterflooding. 114 In Oiese laborato~ 8tudies,
clean core, a refined oil, and ibrine containing mines
were used. The wettabili~ was rever8ed by changing the
pH from alkaline. to acidic. When the PH was alkalime,
the amine group physically adsorbed on tie rock surface,
exposing t& hydrocwhn cl@ to make the sUrfaCe0~-
wet. 173The wettabilhy was altered when the PH became
acidic because tie mines formed water-soluble salts that
rapidly desorbed from the rock surfaces, leaving thern-
water-wet, Hence a core that is oil-wet becomes water-
Jownd of PetJolam Technology, October 1986
,.
wet when, water conta.inhg a mild acid is injected. ”The
most cofnrnonly used amines have been hexylamine and
n-@ylamine. Mungan 174measured the water-advancing
cOnmct a@ on a siliii, syficc u8ing water, n-
hexylamine, and a refined bid. Tbe contact angle with no
aminca present was about 60° [1 rad], or water-wet. As
the concentration of tines tias increaaed, tie contact an-
gle gradually changed to about 120° [2.1.rad], or mildly
oil-wet. In addition to iltering the wettab~ky, the amines
partition between the oil and water and lower IFT.
Alteration of the OrighraI‘Wettabflity
As mentioned previously, alterations in nettability can
affect the iesuks of moat core analyses. IdcaOy,tieie @-
yse8 shordd be mu with core wettabfity that is identical
to the nettability of the undisturbed reservoir rock. Un-
fortunately, many factors can significantly alter *i. wet-
tabilky of thecore. The8e factors can be divided into two
general categories (1) tho8e that influenc&core wettsbik
ity before testing, such as drilliig fluit is,’packaging,
preservation, and cleanin% and.(2) those that influence
wettabiitv dtig testing, such m test fluids, temperahue,
and pres&re. - -”
The wettabili~”of a.core can be altered during the drill-
ing process by the flushhg aitionsof driig fluids, par-
ticularly if the fluid contsim 8urfactimts 128>175or ha8 a
pH~~.1i~.1?6different 1% that of the reservoir fluids.
The w@@biiv may also be changed by the pressure and
temperature drop that occurs as the core is brought to the
surface. This action expels fluids, particularly the light
ends, and changes the spatial dkribution of the fluids.
In addition, asphaltenes and other heavy ends may deposit
on the rock suifaces, making them more oil-wet. The tech-
niques used in ba@.ling, packaging, and preserving the
core & slso alter the .wettability through a 10SSof light
ends, deposition of heavy ends, and ofidation. The labo-
ratory procedures for cleaning and preparing the core cm
change the wettabfity by altering the amount and type of
material adsorbed on tie rock surface.
Factors that can alter wettabfity diu’ingtesting include
the test temperature and pressure. Generally, cores nm
at atmospheric coiidition8 are more oil-wet than those mn
at ,reservoir conditiom becau8e of.the reduction iri”SOIU-.
bility of wettabfity-akering compotindi. An additional
factor iMuencing -he wettabfl~ is the choice Of test
fluids; certain migerd oils camalter the wettqbility. Core
analyses ze sometimes run with air/brine or airlmercu-
v,in place of ofl and brine. ,These analyses ~nplicitly as-
sume that wetfabfity effects are unimportant.
Currently, three different *of cores “ae used in core
analysis: (1) the mtive-state ,core, where every effort is
made to maintiin the nettability Ofthe in-situ roch (2)
the cleaned core, where the intent is to remove all of the
adsorbed compounds from the rock and to leave the core
strongly water-we! and (3) the restored-state mm, where
the core is firstcleaned apd ~en returned to”its original
wettabti~. These definitions are used in the majority of
the more recent literature. However, in some papers, par-
ticuk@y older ones, the term restored-state is used for
what are actually cleaned cores (e.g., see Craig7 ). The
work with native- and restored-state core is at either mm
bient or reservoir temperature and pressure, i.idfe cleaned
cores are usually tested at ambient temperature.
1133
10. Native.State Core
Coring. In a nativi-stite (fresh) core, every precaution
is taken to minimtie changes from the undkturbed reser-
voir nettability condhion, starting when the core ig frost
flushed by the dr~g mud. In pyticukw, a mid with s~-
factants or a pH that ‘differs greatly from the reservoir
fluids must be avoided. Ofl-based-emulsion muds ?nd
other muds containing‘surfactaits, caustics, mud thinners,
organic coriosion, in&itors, and lign@fonates must be
avoided. 175,177Note that, @hIlethey probably exist, no
commercially available oil-based muds have been m rted
P“thatcan preserve the reservoir nettability. 175>17’178
The different coring fluids for obtaining native-stati
cye have been recomriended. (1) synthetic formation
brine, (2).unoxidized lease crude oil, or (3) a water-breed
mud with a miniznuin of addhives. Bobek et al. 175rec-
ommend coring with brine and noadditives. If ,.hk is not
possible, a water-based mud containing only bentonite,
cnrboxymeibyl cellulose, rock salt,, and barite should be
used. This is recommended b~ause they found that this
would not alter the wembfiity of strongly water-wet cores.
Note, however, that the carboxymethyl celkdose,may ~ter
the wettabfit of oil-wet cores, rendering them more
2water-wet, 15 .~75Bbrlich and Wygsl 179recommend a
synthetic formation brine coritaining CaC12 powder for
fluid loss ‘control and no o~er additives. Mungait 180rec-
otiends coriug with lease crude oil. Note f.hatthere are
tio possible problems with the use of crude oil: (1) it
is flammable, and (2) surfactams can be formed by oxi-
dation. of the cmde, which could alter tie wetta-
bili@. ”,103
rhfortummly, very MtIework has been published about
the effects of individual drilling mud components cmwet-
tabiihy, particularly for oif-wet cores. Burkhardt .r al. 176
exaniined the effects of mud filtrate flushing on restored-
,stite”cores snd found no significant effects. Unfortunate-
ly, the cores were in contact with the crude oil for only
12 to 16.hours, so it is doubtfol that the wettabtity was
restored before testing.
Bobek ei al. 175tested several dit%r<nt drillkm mud
componerits used in water-baaemuds on both wat~r-wet
and oil-wet pings The drilling mud :omp6nen@ to be test-
ed were dksolved in or leachdwith distied water then
the resulting solntion was filtcreii. Concentrations of the
compamds were chosen to duplicate those encountered
in the field: Water-wet limestone and s~dstone plugs were
saturated .wltb the test solution and wetibflity alteration
monitored by the irgblbkion method., As,dkcussed earli-
er,’.they”fonnd flat rock salt, carboxymethyl celhdose,.
bentonite, and bake had ho effect on tie wetmbfli~ Of
these initially water-wet plugs. Starch, lime, tetmsoditirn
phosphate, and calcium lignosujfomte altered the wetta-
btity of the sandstone and/or limestone plugs.
Drilhg components that, did not.affect the water-wet
plugs were tested on oil-wet sandstone plugs. The dry,
initially water-wet plugs were made oil-wet before test-
ing by saturation with Elk Basin crude and aging for one
day. Notethit becanse of the short duration of the aging,
the wettabiity may not have been in equilibri~. The aged
cores were flushed witi a drilling mud component fkraW,
then the wettabfity was measured. by the imbibition
method. .Sa.kdid not affect the wettabfity, whfle carb-
oxymethyl cellulose made. the plugs” more water-wet
(bariti was not tested). Bobck et al. found that the PH
,.
1134
of the filtrates was an iinpotit factor in we~bdity al-
teration. The origti bentonite filtrate changed me wet-
tzbtity from oil-wet to water-wet. When the pH was
lowered into the neutml or acidic rsnge, however, no wet-
tabtitjI reversaI”occurr+.
Sharma snd Wunderlich 158meaaurcd the wettabi!.iwaf-
teration caused by different drilliig mu@components in
water-wet and oif-wet Berea plugs. The oil-wet Berea
plugs were prepared by treatment’witi an asphaltic crude
and pcntane, as discussed previously. Dry plugs were
saturated with brine, injected with 10 to 12 PV’s of the
drilling fluid component, aged for 15 hours; tien flushed
with 5 to 6 PV’s of brine. Wettabilky was measurd af-
ter contamination by a combined USBM/Aniotl method
developed by Shagna and Wunderlich 158and compared
with the we~bility of control samples. The driiing comp-
onents tested included bentonite,, carboxymethyl cellu-
lose, Dextrid@ (an organic polymer), D~spac@ (a
polyanionii ceflukiwe polymer), hydroxyetbylcelhdose,
pregelatinized starch, and xanthm gum. These compo-
nents are generally considered relatively bltid, with onfy
small effects on the wettiibilky. None of the components
affected the nettability of ~e water-wet plugs. However,
N of the components, with the exception.of the bentonite
fdtrate, made the oil-wet plugs significantly less oil-wet.
This indicat&stie need for further rese~ch on acceptable
drillings muds for obtaining native-state core.
Several researchers have attempted unsuccessfully to
fmd suitable commercially available oil-based muds for
~btig ~tive.sate core, 175,177,178,~ of tie ~fi.ba~~
drilling mud iiltrates tesied made water-wet cores more
oil-wet. Unfortunately,. none of the reports identify tlze
specific drilling mud components used. ,
Cm-e Packaging and Preierv&ion. Once the core is
brought to the surface, it must be protected from wetW
btity alteration caused by the loss of light ends or depo-
sition snd oxkldion of heavy ends., On exposure to air, .
subticei k crizdecan rapidly okidize to form wkw prod-
ucts that are surfactants, altering ihe wettabili-
,W,34,73,103,115,175,181,182 M ad&tiOn, a tick ofl-w@
residue from the crude will be deposited on &e rock sur-
face if the core is allowed to dry out. To prevent wetE-
bility alteration, Bobek et al. 175 recommended two
sltemitive packaging procedure: that are now gene@y
used for native-state cores. The fwst ia to wrap the cores
at the weUii@in polyethylene or polyvinylidene film and
then in aluminum foil. The wrapped cores we then sealed
with a thick layer of paraffin or a special plastic sealer
designed to exclnde oxygen aid prevent evaporation. The
second, preferred method is to immerse the cores at tip
wellsite in deoxygenated formation or synthetic brine in
a glass-lined steel or plastic tube, which is then seafid
to prevent leakage and the entrance of oxygen. ImbibG
tion wettabihw tests showed that the nettability of core
packaged by either of these WO methods was unchanged
from the wettabfily mesaimed at the wellsite. Instid of
deoxygenated brine, Mungari180 recommended tit the
cores be cut and stored in degassed l&ae crude oil. Mor-
gan and Gordan 183and McGhee et al. 62 recommended
that the cores be stored in their wetting fluid, either for-
ntstion brine or crude oil. “Thewettabiky would be,deter-
mined by an imblbhion test at the wellsite. Finally, not?
that cores taken in a robber sleeve, fiberglass, or PVC
Journalof Petroleum Tedmo108y, Octpber 1986
11. TABLE 3—EFFECTS OF EXPOSURE TO AIR AND PARTIAL DRYING ON
NATIVE-STATE CORE
Number Average . Average
of Cores Dlsplacement- Oisplacement-
Tested iescriplion by-Water Ratio by-oi Ratio
2 Native state 0,97 0,00
3 Exposed to air at
70 to 100nF for 1 day 0.63 0.00
2 Exposed to air at
75°F for 60 days 0.42 0.00
4 Exposed to air at
225°F for 7 days 0.18 0.00
C’ay$btdnedbyuseoftheAmattwellafl~v,,s, natNeNatecorefrom0,! Zone8., SterlingCo”nly,
L
1’
liner can be DESeNed if the ends are capped’and sealed. approaches one as the water-wetness increases. Siniiiar:
A number-of experiments have dem&strated that ex-
posure to air and drying cao alter the nettability of core.
As discussed earlier, Treiber et al. w measured the.net-
tability of 50 reservoirs usfig deoxygemted synthetic for-
mation brine and anaerobic crude. In some cases, the
contact angle showed that the reservoir was water-wet.
For some of those cnrdes, exposure to oxygen changed
the wettabili~ to ofi-wet. Bartell mrd Niederhauser103
stodied interracially active materials in crude, which con-
centrate and form solid fflms at the oil/water interface.
These materiils can also be adsorbed on the rock surface,
rendering it oil-wet. Crodes and brines were obteined and
stored without exposure to oxygen. Most of these crudes
showed very little interfaciel activity. On exposure to air,
the cmdes developed moderate-to-strong fdrn-fomning
tendencies, while the oil/water IFT was lowered by as
much as 50 %, indicating that surfactants were formed by
oxidatiori of the crude...
Richadson et al. 1s2 stored core’ from a mixed-
wettability reservoir47 using four different methods. Ox-
idation snd drying of the core tiere prevented with the
first two methods: (1) core wrapped in foil and scaled in
paraffh and (2) core stored in evacuated (deoxygenated)
formation water. The other methods were (3) core stored
in aerated formation water and (4) core stored in cloth
core bags. The cores were oilflooded with kerosene to
IWS and then waterflooded. The average ROS for the
samples protected from oxidation and drying (Methods
1 and2) was about 13%; forthesamples submergedin
aerated water, about 24%; and for the samples storedin
core bags, about 25%.
Bobciket al. 175used the imbibition method to compare
the nettability of native-state cores at the wellsite, cores
allowed to weather, and cores stored by the two recom-
mended metiods discnssed above. The nettability of the
cores stored by either of the two recommended methods
was the same as the nettability measured at the wellsite,
while most of the weathered cores became more oil-wet.
Am0tt177 used bismethtid”t ocompari”th ewettabihy
of native-state cores with similar cores that were exposed
tooxygen ayd~owed to partially dry, asshownin Ta-
ble 3. The native-stste cores were strongly water-wet, with
a dkplacement-by -water ratio of 0.97. In the Amott test,
the displacement-by-water ratio is the ratio of the oil
volume displaced by spontaneous imbibition to the total
oil volume displaced by .botb imbibition and forced dis-
placement. Itiszero forneuqally mdofl-wetcO~~s aod
J&mal of Petroleum .Tecbnology, October 1986
ly~the displacement-by-oil ratio is zero for neutrally and
water-wetcores aod approaches one as the oil-wetness
increases. The cores became more oil-wet as they were
either exposed to the air for longer periods of time, o!
at higher temperatures. Similar tests on an initially weakly
water-wet core showed elmost no change. On the other
hand, Mungan 115used the imbibition method to meas-
ure the wettabilhy of native-state cores. In contrast to tie
experiments discussed above, cores preserved in deaer-
ated water were oil-wet, but becsore water-wet when ex-
posed to ti for 1 week. Chfingar and Yen35 have also
reported that some cores became more water-wet on ex-
posure to sir, indicating that it is bnpossible to predict
how the wetfsbtity will be altered by tie oxidation of tie
cmde.
Mungao 180recnmmendsflushing native-state core with
five erode oil before sny flow studies are startsd. After
native-state cores .havc been prepared, they are usually
nm at reservoir conditions with crude oil and brine.
Probably the greatest, uncontrollable problem with
nstive-state core is the alteration of nettability as the core
is brought m the surface., When the pressure is lowered
to atmospheric, light ends are lost from the erode, chang-
ing its properties. In addition, heavy components cm come
out of solution and deposit on the rock, making it more
oil-wet. 137 The decrease in temperature wilf also
decrease the solubili~ of some wettebfli~-altering com-
pounds. Pressure coring prevents tie loss of light .erids.
However, the cores are frozen before removal, so
wetta.bflity-altering compounds ‘W deposit. Unfortunate-
ly, there is no experimental work avsilableon wettabfli-
ty alteration as the core is brought to the surface. ..
Cleaned Core .’
The second type of core used in core aoalyais is the
cleaned core. Cr&ig7 recommends that cleaned core be
used for multiphase flow measurements only when the
reservoir is known to be strongly water-wet because errors
iq the core am.lysis will be introduced otiimvise. There
are two main reasons to clean wre. The first is to remove
all liquids “fromthe core so that. porosity, permeability,
and fluid saturations can be measured. Core cleaning for
these roudoe core measurements will not be considered
in this paper. The second reason for cleaning is to obtain
a strongly water-wet core, generally as a first step in
restoring the wettabfity of a contaminated core.
1135
12. In obtaining a cleaned core, an attempt is made to re-
move all of the fluids and adsorbed material, leaving a
clean rock surface. Gant aid Anderson 129discuss the
me’ibcdsused,tu clean core. One common method is reflux
extraction (l)ean-Stark or Soxhfet) with a solvent such as
toluene, sometimes followed by extraction with cfdoro-
form or methanol. Alternatively, a flow-through system
where the solvents are injected under pressure is some-
times used. 57,&,65lf the cleaning procedure is success-
ful, the core is left strongly water-wet. Cuiec”$5 mid
others 57,1w discussed the chemicalreactions involved in
the cleanin process.
Cui&ca,& compared me efficiency of different solvents
in flow-through core cleaning. Initially water-wet outcrop
sandstone and limestone cores were saturated with differ-
ent cntdes (sometimes the cores also contained brine), then
aged. The aged cores ivere ncyrmfly neutr~- to oil-wet,
as determined. by the Amott wettabflity test. The cores
were then cletied with dlffererit solvents, and the Amott
test was used to determine cleauing efficiency. Cuiec
found that he could clean both sandstone and liiestone
cores by flowing the f?ll@ig seven solvents through the
core: pdarte, hexane, heptane, cyclobexane, beuzene,
pyridine, and ethanol. Chloroform, tohtene, and methanol
used singly were not very effective. Cuiec also looked
at several dtiercnt acidic and basic solvents used individu-
ally and found that the acidic solvents tended to be more
effective in cleaning sandstone, whiie the basic solvents
were better in cleaning Iimcstone. This difference was at-
tributed to the acidic nature of the sandstone surface and
the basic mture of the limestone surface. For examplei
because sandstone (silica) has a weakly acidic surface,
it tends to adsorb bases from tie crude oif. When a
stronger acid flows through the sys@m, it will gradually
react with and strip off the adsorbed bases, Ieaviug a clean
silica surface.
G@ and Anderson 129 surveyed most of the core-
cleouing experiments in “theliterature. They found that
the best choice of solvents depends heavily on the cmde
~d the mineral surfaces becavse they help determine the
amount and @pe of nettability-altering compounds ad-
sorbed. Solvents that give good results with some cores
and cmdes often faif in other cases. For example, Grist
et al. 1s4 and Holbrook and Beruard45 both found that
they could clean core to a strongly water-wet state using
a cliloroformfmerhanol mixture, while Jennings5s repOrt-
ed. fiat thk was unsuccessful. For cleaning for routine
core analysis, API 1s5 reports that cbfom form is excel-
lent for many midcontinent .ct’udes,wh~e toluene is us6-
ful for asphaltic cntd:s..
hi many cases, it appears that any single solvencis rela-
tively ineffective in core cleaning and that much better
results can be obtained wiih a mixtttfe or series of
solvents. 129 The followhg solvents have, been rCpOfi-
ed for specflc binbinations of crude and core to give
pcor resufts when used alone: chloroform, ‘,65 ben-
zene, 5S,1M,120~=bon di~~fide, lU,120 ~~mol, @ ~d
toluene.5@.@.65.1~.120,1T.l~,1s6
Many of the researchers cited above have found that
toluene used alone is one of the least effective solvents.
However, when combined with other solvents, such as
methanol (CHs OH) 184 or ethanol (CH3 CH zOH), 61
toluene is often ve~ effective. The toluene is effective
in removing the hydrocarbons, including asphal-
tenes 130JWand some of the weakly polar compcmmi.s,lW
wbile”tbemore,spongly ’polar methanol (ethanol) qcmoves
the strongly adsorbed polar compounds that are often
responsible for altering wettabilky. In addition to
toluenelmethanol and tolueneletbonol, successful clean-
ing has ako been reported with cbforoforndace-
tone 1wZ120.1= and cfdorofordmetfranol, 1s4 as well as
a number of different series of solvents. ‘,65
CuieCand his coworkers made the most extensive study
of core cleaning for nettability restoration. In a recent
paper, Cuiec et az. 130statkd that their core cleaning al-
ways begins with a toluene flush to remove hydrocarbons
and asphaltenes. A number of solvents are then.tested to
determine the most effective, including (1) a series of non-
polar solvents, e.g., cyclohexane or heptane; (2) acidic
solvents, e.g., cblorofonn, ethanol, or metbanoh (3) ba-
sic solvent’i, e.g., dioxane or pyridine; ond (4) mixtures
of solvents, e.g., methanollacetoneltoluene. When none
of these procedures are effective, other tests are performed
by combining the above procedural, using otler solvents,
ad incremfig the Circdatiog time.’
Tohtene is generally not a very effective solvent, but
it can”alter the nettability of some core. Jennings 186
cleaned sever,d cores by toluene extraction and found that
the wettabilities and relative penn.eabilities were not
changed. He stoted that this indicated that toluene-
extmcted core retoined the reservoir wettabi!ity and coufd
be used for relative p+mneabtity rneosurements, However,
this generally is not the case. Aftbough it is less et%cient
than other solvents, we have found that toluene extrac-
tion can alter the wettabili@ and relative penneabtitics
of native-state core. fn some cases, neutmlly wet or tidly
oil-wet mtive-state core becomes strongly water-wet ti-
ter extraction witi toluene. The relative permeability
cume~ ~~o ~E,fi. AMOU177 ako found that toluene ex-
traction can clean some cores, while it had Wt3eeffect
for other ones, such as the strongly oil-wet Bradford cores.
Therefore, because tolueneextraction will alter the wet-
tabilky and relative permeability of many. native-state
cores, m%uremcnts shouId be tie on mtive-state cores
before toluene extinction.
One problem with a cleaned core is that it is sometimes
difficult, if not impossible, to remove all of the adsorbed
material. If this occurs, the wetta.bfity of tie cleaned core
wifl be left in some indefinite staie, causing variations in
core analyses. Grist et al. lW cleaned cores by three cu-
rently used methods and then examined how ROS and end-
point effective permeabilities vtied after a waterflood.
ROS was very similar for W methods. However, the end-
point effectiye water permeability varied by more than
a factor of three betwtin different cleaning methods. Their
explanation for this behavior was that some methods were
able to extract more of the adsorbd. components, leav-
ing.the rock more water-wet. In the more water-wet cores,
the rwidti oil had a greater tendency to form trapped
droplets, blocking pore throats and lowering water per-
meabtity. The least effective of the three cleaniug me~ods
was overnight reflux extinction with toluene. More ef-
fective was reflux extraction with toluene followed by 2
days of extraction with a mixture of cbloroforin oud
methanol. Finally, tie most efficient method was reflux
extraction with tcduene followed by 3 weeks of extrac-
tion with chloroform and methanol. In the last stage of
cleaning, methanol was used alone.
1136 JournalofPetroleumTechnology,October1986
13. Another draivback of cleaned cores is that it is occas-
sionafly possible for cIeaning to change an originally
water-wet rock to an ofi-wet one. The extraction process
may quickfy boil off the connate water, allowing the re-
maining oif to contact the rock surface and form oil-wet
deposits that are afmost impossible to remove. 187
The cleaning experiments discussed examine the best
methods to remove. cmde oil constituents from the pore
walk. In many cases, core is also contaminated with drill-
ing mud surfactits, which “mustalso be removed before
the wettabifity of a“core can be restored. 12ar129The best
choice of solvents depends on the crude, the mineraf s~r-
faces, and the drilfiig mud surfactsnts. Gant and
Anderson 129cleaned Berea sandstone and Guelph (Bak-
er) dolomite plugs contaminated with an invert-oil-
emufsion rfrifling mud.filtrate. The best solvent fOr bOfh
rock types was a 50/50 mixture of toluene/methanol, oi
the equivalent, containing 1% ammonium hydroxide. A
three-step method (three successive Dean-Stark
extractions-toluene, foffowed by glacial acetic acid, fol-
lowed by ethanol) was the second best choice for Berea,
while 2-methoxyetiyl ether was the second best choice
for dolomite, demonstrating that the choice of solvents
can depend O? the mineral surfaces in the core.
Restored-State Core
,
If one conld be “positivethat the original reservoir wetta-
bilhy had”izotbeen inadvertently modified, a native-state
core would give resnks closest to those of the reservoir.
However, riative-st@ecores present scveraf problems. The
necessary procedures to preserve the wettabtiq’ are
troublesome and time-consuming. Even when sll of the
precautions iwe tsken, there is still a possibility that the
nettability has been chtiged through oxidation or through
deposition as ‘tie temperature and pressure dropped when
the core was brought to the surface. In addition, tic ques-
tion ‘arisesabout tie procedure to follow to obtah the most
reliable information from cores in which the nettability
WaS aftered
When onfy core with alt&ed wei@bfity is available,
the best possible mukiphase measurements ore obtained
by resto& the resefioir wettabfi~ with a three-step
F~roces~.47, 0,64,65.,96,115,128,130,1S0,18SThe f~st ste~ is to
~leaa the core to rernoVe’all compounds from th~ rock
surface. After the core is cleaned, the second step is to
flow reservoir fluids into the core sequentially. F~y,
the core is aged at the reservoir temperamre for a suKL-
cient dine b establish adsorption equilibrium. Seversl ex-
perimenters have compared measurements made on core
in the native, cleaned, and rsatored states. In each experi-
ment, measurements in the restored state were slmost
identical to the,previous mtive-sta.te “ones,demons@atin
that this procedure will restore nettability. 50,115Js0;18~
~,, The first and most @fficult step in nettability restora-
tion is to clean the contaminated core by use of the
methods described to remove all compounds adsorbed on
the surfaces and to make the core as water-wet as possi-
ble. All compounds must be removed from the core be-
cause we have no knowledge of which compounds were
adsorbed on the undisturbed reservoir rock and which
were deposited afterward. The USBM or Amott netta-
bility measurements are used to verify that the core is
strongly water-wet.,Unforttzmtely, detemining which sol-
vent wiII successfoffy clean the core is still a trial-and-
Fig. 1—Wettabitity changes for a restored-state core and
the effects of flushing restored-state cores with refined
OIIS. Berea core and Big Muddy crude.
error process because the best choice of solvents depends
heavily on the crude oil, the mineraf surfaces, and any
&idling mud contaminant. Further discussion can be’
found”in Ref. 129.
In the second step, sequentially flowing reservoir fftids
btto the core, the core is saturated with deoxygenated syn-
thetic or formation brine and then flooded with crnde oil
to simnla.tsthe intlow of oil into the resemoir. When cmde
oil for wettahifity restoration is obtained, precautions
should be taken to minimize alterations to the crude. The
hnple must be taken before tiy Sm&ct.mrtsor Ofier
chemicals ae added to treat the crude. It should be taken
as Iong as possibIe after any wefl treatments to aIIow time
for these chemicafs to be flushed from the well. Finally,
the cmde should be sealed in air:tight containers as soon
as possible to minimize oxidation and the 10SS of light.
ends.
The fired step in wettsbfity restoration is to age tie core
at the reservoir temperature for a stilcient time to es-
tablish adso~tion e&libzium. The aging time required
to re-eatabfish reservoir nettability varies, depending on
the crude, brine, and reservoir rock. Generally, we feel
&at core shoufd be aged for 1,000 hours (40 days) at the
reservoir temperature. 128This a&ng period was chrken
for two reasons several experiments have shown that up
to 1,000 hours is required to reach wetting equilibri-
um ~$s. 115,189-lgland 1,000 hours is roughly the length
of ~me required for the contact ~gle measnred on a flat
surface to approach its equilibrium value. 7>26,34.191III
some cases, the restoration time can be significantly less
than l,OW hours. MunganlsO was able to restore the wet-
tabilky after aging for 6 days, while the nettability of the
rocfuoilh’ine system used by Schmid50 and Riibl et
~. lss was restord after only 3 days. Salathie147was able
to restore a mixed-wettabilhy qate to samples after 3 days.
Cuiec et al. 130describes two reservoirs in which the wet-
tabtity was restored after only a few hours, with no fur, <
ther cbamgesin tie wettabiIity for aging times as long as
1,000 ho;rs.
There are two basic options to determine the aging time
to restore wettabllitv. We feel that it is ,most convenient
to age afl cores for l:W” hours, which is roughly the mm-
imum time that the experiments discussed previously re-
quired to achieve wetting equilibrium. While cores may
Iourmt of Petroleum Technology, October 1986 1137
14. be aged for a period longer than the minimum necessary,
thk is not a serious drawback because the aging cores re-
quire minimal attention. Another possibtity is to deter-
mine the minizhtzmaging time by measuring the wetrabfity
of the core with the USBM or Amen methods at frequent
intervals during the aging period. The aging is stopped
when the wettabflity reaches its cquilibtium value.
Although this minimizes aging time, it is much less con-
venient because it is labor intensive nnd requires frequent
disturbances to the plug.
The core is aged at either the reservoir pressure with
live cmdeso, 180,188,191or ambient pressure with dead
cmde. 115.128.190 When live crude oils a~d the reservoir
pressure are used, the solubflities of the wettabilhy-
altering compounds should have their reservoir values.
It is possible that the nettability will differ when dead
cmdes .atnmblent pressure are used. At the present time,
however; it is not known whether the difference is im-
portazm
Fig. 1 shows the chsnges in the USBM wettabfity in-
dex as a core was restored. * A series of Berea plugs was
saturated with brine and driven to IWS by centrifugation
in crude oil. Each core was aged in dead crude for a differ-
ent period,of time, after which the USBM wettabtity was
measured. As can be seen, the wettabilky changed from
water-wet (W =0.8) to moderately oil-wet ( W= – O.3)
over a 40-day period. The plugs flushed with Soltrol@
and Blnndol@ will be dkcussed later.
Lorcnzetal. 190and Cuiec65 found tbatitissometknes
possible to speed up the approach to wetting equilibrium
by saturating.thecore with oil alone. The approach to
equilibrium is fastix’ because the polar compounds no
longer have to dlffase across a water layer to adsorb on
the rock. This procedure should be avoided, however, be-
cause it can give iriaccuratc results. For example, con-
sider the restoration of a core that originally had
Salatbiel’s47 mixed wettab~ity, where &e large pores are
oil-wet nndthesma!.l ones are water-wet. During tbeag-
ing process, thesmall pores must contain comate.water
to prevent the deposition of au oil-wet fiim, leaving them
water-wet. Onthe other hand, ifaclcan core is saturated
only with oil, tbe entire core, includlu gtbesmallpores,
will become uniforndy oil-wet, whlchis the wrong wet-
tability. Anaddhionalp roblemw ithsaturadngt iec ore
solely with oil istbat the effects of brine chemistry are
ignored. As discussed previously, the wettabtity of the
core depends on the ionic composition and pH of the brine.
Finally, Clementz10?.120,121$howedtbat flowing cntde
oiltbrougha dry core camcause tbeformition of very
stable oil-wet, claylorganic complexes. Thepresenceof
an initial water film on the clay surfaces haa been shown
to reduce but not completely inhibit the adsorption of the
nettability-akering materkds. @@.70The effects of brine
OtI wearability make it necesssry to saturate the core with
brine, then oil, during the nettability restoration process.
Experimental Conditions
Once a mtive or restored-state core is obtained, core anal-
ysescan be performed. These tests can be mn with either
crude or refined oil at ambient or reservok’ temperature
and pressure. Because wettabfity effects are being ig-
‘Personalc.mm””lcati.nwl,hD..J.Wendel,Pe!r.aleumTestingSeMceS,Smta
FeSPringS,CA,No”.19S0.
nored, cleaned cores are generally mn with refined oif
(or even mercury or air) at room temperature and pres-
sure. From the viewpoint of titaining the nettability,
the best laboratory tests should be mn with native or re-
stored cores at reservoir conditions with live cntde oil and
brine because this is the best simulation of reservoir con-
ditions possible. Cor~ are generally more water-wet .at
rese~oir conditions tbzn they are at rooin temperature
ad pressure, 62,180,192-195The effects of the following
experimental conditions on nettability will be dkcussed:
(1) rcaervoir vs. room temperature, (2) live vs. &ad cmde
at reservoir pressure, and (3) refined vs. crude oils.
Changing the temperature has two different effects, both
of wtich tend to make the core mom water-wet at higher
temperamres. First, an increase in temperature tends to
increase the solubtity of wettabtity-altering cOm-
pounds. 196 some of ~eSe compounds will even desorb
from the surface as the temperature increases. Second,
the IFT and the contact angle measured through “-bewater
will decreaae as the temperate increases. This effect has
been noted in experiments with cleaned cores, minernl
oil, and brine, where it was found that cores at higher
temperatures were more water-wet even though there
were no compounds that could adsorb and desorb. 19-2?
For example, McCaffery201 measured the water-
advancing contact angle on qWrtz of n-tetradecane and
brine. The amglewas about 40° [0.7 rad] at 77°F [25”C],
but decreased to about 15°. [0.3 rad] as the temperature
was raised to 300”F [150” C].
when live crude oils at the reservoir pressure and tem-
perature are used, the solubilities of.dzewettabfity-afteting
compounds have their reservoir values. The use of dead
crude at ambient or reservoir prcasure may change the
nettability kmzse the properties of the crude are altered.
Light ends are lost from the crude, while the heavy ends
are,less soluble, which may make the core more oil-wet.
However, the effects of pressure are not known at pres-
ent. The two reported experiments found that ressure
is much less important than temperature. 18~,1~ Hje~.
meland and Lamondo 192found little difference in con-
tact angles measured using stock-tank vs. live cmde at
the reservoir temperature (190”F 88”C]) and pressure
i(3,800 psi [26.2 MPa]). Mtmgan 10 measured a water-
advancing contact angle ,of $7” [1.5 rad] using live reser-
voir crude and synthetic formation brine at resemok tem-
perikurc (138”F [59”C]) and pressure (1,200, psi [g.3
MPa]). The water-advwcing contact angle was almost
identical, 85” [1.48 rad], using degassed crude and brine
at ambient pressure and reservoir temperamre.
Because refined oifs are much easier to work with tlum
cmde, it is a common laboratory practice to flush native-
or restored-state cores with refined oil before testing.
However, there is a possibilky that this idters the netta-
bility. Craig7 poshdated that it would be possible, once
the original wettabtity was restored, to use refined nzin-
eml oil in place of crude oil in laboratory tests without
adversely affecting the wetk+bility. Test times are shor!
compared with the time it takes to achieve adsorption equi-
librium and obtain native wettsbtity (about 1,000 hours).
Craig hypothesized that the desorption of wettability-
Mbzencing materials would require a correspondingly
long period of time. If this is correct, @eorig@l wetta-
btity wotdd be unchanged if laboratory tests using refined
oil and brine were conducted quickly enough.
1138 Journal ofPetroleumTechnology,October1986
15. The onfy experinient to teat this hypothesis that we are
aware of was conducted by Wendel. * He aged ,Blg Mud-
dy crude in Berea sandstone at IWS to develop his
restorer-state cores. The cores. were flushed with one of
two refinkdoik, Soltrol 170 or Bkmdol, to detexmine how
they affected the wettabfily. The results&e show in Fig.
1. Bkmdol did not si@ficantly affect the we~btity, while
Soltml 170 changed tbe core from oil-wetto neutrally wet.
The wettabdity: alteration could be caused by either
surface-atilve impurities in the Soltrol 175or desorption
of previously depositwj oti-wetdng crude compounds from
the pore walls into tie Soltrol. It i; not known which ex-
planation is correct. Wendel did not attempt to fflter the
refined oils tfmugh a cbromatogmphic coltmm tn remove
surface-active compounds. These contamiim ts are known
to have a large effect on corit.w-migle measurements,
which are extremely seyitive to small amounts of con-
taminants. Wettabifity measurements in core should be
less sensitive, however, because the ratio of smface area
to volume is “much higher.
Conc133sioris
1. The nettability of a rsaeryoir ample affects ita capil-
lmy pressure, relative pe,rmwbtity, waterflood behavior,
dispersion, mid electrical properties. fn addhion, simu-
lated teftiary recovery can be 51ter,cd.The tcfi,~ recov-
e~ PrOcesses affected by we~biE~ include hot-water,
surfactant, miscible, aid caustm floodlng.
2. Cleaned, strongly water-wet cores should be used
onfy in such c6re analyses as porosity. and air permeabil-
ity, where the wettabilky is unhpportmrt. h addition, they
may be used in other tests when the reservoir is known
to be strongly water-wet.
3. The nettability of originally water-wet mineral sgr-
faces can be altered by the adsorption of pokw compounds
aui/or the deposition of orgariic rnatter””tlmtwas origi-
MUY in the crude oil. Fmrfactants in the crude oil are
generally believed to ,be polar compounds that contain
oxygen, nitrogen, ,mdlor sulfur. These compounds ~
most prevalent in the heavier fractions of crude oil, such
as the resins and asphaltenes.
4. Nettability alteration is determined by the interac-
tion of the oil constituents, the mineral surface, and the
brine chemistfy, including ionic composition and pH. In
siIicafoiVbrine systems, trace amounts of mukivalent me-
M.cations can alter *e nettability. The catiom can reduce
the solubti~ of crude oil surfactants and/or activate the
adsorption of aniotiic surfactants onto the silks. Mfdtiva-
lent ions:$tathave altered tie wettabfity of sihcafoil/brine
systems tnclude Ca’2, .Mg’2, Cu’2, Ni’2, and Fe’3.
5. Work on mineral flotation indicatca that coal,
graphite, sulfur, ‘talc, the talc-liie silicates, and many sul-
tidca are probably naturally,neutrally wet to oil-wet. Most
other minerals-includhg quartz, carbonates, and
sulfates—are strongly water-wet in their natural s@.te.
6. Contact-angle measurements suggest ti.at most car-
bonate reservoirs range frpm neutralIy to oil-wet as a re-
sult of the adsorption of surfactimts from the crude oil.
7. Very little work has been reported about the changes
in wetibility caused by drioiig mud addhives. Three
different “Cciring.flrtidshave been recommended to obtain
native-state core: (1) synthetic formation brine, (2) un-
-PemmalmmmunlcatimwilhD,J.Wend.at,PetroleumTestingS6wkeS,Santa
Fe SP,ingS,GA,N.”. 19S0,
oxidized lease crude oil, or (3) a water-based mud with
a minimum of iddkives. B&use of surfactits in the SYS-
tern, no conimercitiy available oil-based or oil-ernukion
muds are khown that preserve the native wt?ttability.
8. The wetr?bility of a native-state core. cambe altered
by loss of light ,ends and/or the deposition aid oxidation
of heayy ends. TWOalternative pac@ging procedures cam
be used to miniinize these effects. The first is to immerse
the corei in deoxygenatdd formation or synthetic brine
and place !hem in a glass-lined steel or plastic tube, which
is then seaIed against leakage aud the entrance of oxy-
gen. An alteinaiive procedure is to wrap the cores at the
welksite in polyethylene or pglyvinylidene fdm and tien.
‘in ihunirimn foil. The wrapped pore is tlyn coated with
i+thick layer of paraffin or a plastic sealer.
9. Because of the increased solublky of the wettability-
altering compounds at the higher temperature spd pres-
sure, the cmde-odtbrinetcore system is usually more
water-wet at reservoir condition than at ambient con&-
tions. In addition, the contact angle measured tluough the
water will generslly decrease as the tempetatufe is in-
creased, and the system will become more water-wet,
even if no surfactauts are present.
10. Extraction with toluene cti alter the wettabil.ity of
some iiative-state cores, causing some”initially neutrally
wet or qikfly oil-wet cores to become strongly water-wet.
Measurements on native-state iorei should be made be-
fore toluene extraction.
11. During the attempted restoration of a cleaned core
to ita orig$al wetrabtity, the core should be saturated with
brine, @lflooded, md then aged at the reservoir corrdi-
tiom for 1,COOhours. This will embie a inixed-wettabtity
condhkin to be restored, if thk was the original wettabd-
ity. In addition, it will allow the brine chemistry to influ-
ence tie r~tored nettability. An alternative procedure,
which completely saruratea the core with cmde oil, should,
be avoided.,
12. The three commonly used methods for artificially
conrrolEng wettabfity during laboratory experiments are
(1) treatment of the core with chemicals, generally or-
ghocblorosilane solutions for sandstone. cores and
naphtbenic acids for carbonate cores; (2) using sintered
teflon cores with pure fluids; and (3) adding surfacmnts
to the fluids. To obtain a uniformly wetted core, a sin-
teredteflon core with pure fluids is preferred because its
nettability is more constant and re reducible than the wet-
#tabilby of cores treated WI organochlorosilanes,
naphthenic acids, or smfactants. However, these treat-
ments have advantages when heterogeneous wettabllity
or nettability alteration is”studied.
Acknowledgments
I w grateful to Jeff Meyers for hk many helpful sugges-
tions and comments. I also thank the management of
Conoco h.c. for permission to publish this paper,
References
1. Andemm, W,G.: ‘<Weuabitily LiteratureSurvey-Parr 2 Wet-
Iabitim Measurenimt,’, to be published in JPT (Nov. 19S6).
2. Anderson, W.G.: .’Wetmbili~ Literature SUrvey-PaX % The Ef-
fectsof Watabtity on,the E2ectrkaJProperties of Porous Media,>,
to be published in JPT (Dec. 1986),
3. Anderson, W.G.: “WemabilityLiterature Smy–Pmt 4: The Ef-
fectsof We@ilify m CapiUaIYPressure,,xpaper SPE 15271waif-
able at SPE, E?icbardsm, TX.