Nettability Literature Survey

Part 2: Nettability Measurement

William G. Anderson, SPE, Conoco Inc.
5?E

13933

Many methuds have been used to measure wettabtity. This paper describes the three quantitative
Summary.
methods in use to&y: contact angle, Amott method, and the U.S. Bureau of Mines (USBM) method. The
advantages and fimitationa of alf the qualitative methods-inrblbition, microscope examination, flotation, glaas
slide, relative permeability curves, capilkmy pressure curves, capiffarinretric method, displacement capillary
pressure, penneabiLhy/saturation relationships, and reservoir logsare covered. Nuclear magnetic resomnce
(NMR) and dye adsorption, two methods for measuring fractional wettabflily, are also discussed. Fdy,
a
metbud is proposed to determine whether a core haa mixed wettabilhy.
Introduction
Thispaper is the second in a series of fitcrature surveys
covering the effects of nettability on core analysis. 1-3
Changes in the wettabtity of cores have been shown to
affect electrical propertica, capillary pressure, waterflood
behavior, relative penncabtity, dispersion, and simulated
EOR. For core analysis to predict tbe behavior of the
reservoir, the nettability of the core must be the sime
aa the nettability of the undisturbed reservoir reck.
When a drop of water is placed on a surface immersed
in oil, a cuntsct angle is formed that rsngis from Oto 180
[0 to 3.14 rad]. A typical oil/water/soIid system is shown
in Fig. 1, where the surface energies in the system are
reIated by Youngs equation, 4

rad], the greater the wetting preference for one fluid over
another. If 8 is exactly 90 [1.6 rad], neither fluid
preferentially wets the solid. As shown in Table 1, when
O is behveen O and 60 to 75 [0 and I twl.3 rad], the
system is defined as water-wet. When 6 is between 180
and 105 to 120 [3.1 and 1.8 tu 2.1 rad], the system is
defined m oil-wet. In the middle range of cnntsct angles,
a systcm is neutrally or intermediately wet. The contact
angfe that is chosen as the cutoff varies fmrn paper to
paper.
The term am n., is sometimes called the adhesion
tension, LTA
5:
~A=v m Ows=aowcose.

am Cos 19=uo$-u . . . . . . . . . . . . . . .

..

where
aow = interracial energy [interracial tension (IFT)]
between the oil and water,
LTO,= interfaced energy between the oil and
did,
o~, = interracial energy between the water and
solid, and
9 = contact angle, the angle of the
waterloillsolid contact line.
By convention, the contact angle, O, is me~ured
tbrougb the water. Tbe interracial energy uOWis equal to
o, the IFT.
As sbowmin Fig. 1, when the contact angle is less than
90 [1.6 rad], the surface is preferentially water-wet, and
when it is greater than 90 [1.6 rad], the surface is
preferentially oil-wet. For almoat all pure flrdda and clean
rock or F&shed crystal surfaces, IJo, and UWhave values
such that O=0 [0 rad]. When compounds such as crudeoil components are adaorbed on rock surfaces, these intcrfacisl energica arc changed unequalfy. fMs changes
Oand hence the wettabtity. The farther Ok from 90 [1.6
CQwrmltsm %e~.! Pemleum
Em.e.=
1246

. . . . . . . . . . . . . . . ...(2)

. . (1)
The adh+ion tension is positive when the system is waterwet, negative when the system is oil-wet, and near zero
when thesystem is neutmfly wet.
Mathods of Wettabitity Measurement
MaJJy different methods have been proposed for measuring the wettabtity of a system. 616 They include quantitative methods-contact angles, imbibition and forced
displacement (Amott), and USBM wettabfity method
and qualitative mcthoda-imbibition rates, micmscope examination, flotation, glaas slide method, relative p&rnreabdity curves, penneabilitykdumtion
relationships,
capillary pressure curves, capiflmimetric method, displacement capif.larypressure, reservoir logs, nuclear nrsgnetic resonance, and dye adsorption.
Although no single accepted method exists, tbrcc quantitative methods generally are used: (1) contact-angle
measurcnrent, (2) the Amntt 6 method (imbibition and
forced displacement), and (3) the USBM method. g.17.18
The contact angle mcasurea the wettsbilhy of a spccitic
su$ace, while the Amott and USBM methods measure
the average wettsbility of a core. A comparison of the
wettsbllhy criteria for the three metbuds is shown in Table 1. we remaining tests in the fist are qualitative, each
with somewhat different criteria to determine the degree
of water or oil wetness. Unfortunately, tbk leads to amJourmJo! PetroleumTechnology,November1986

bigrritieswhen expainrenta in the literature are compared.

Many of the wettsbfity measurements arc alao imprecise,
partimdzrly near neutral nettability. One, method may
show that a core is mildly oil-wet, while soother shows
that the core is mildly water-wet. In tlrk paper, the differcrrt methnds of wettabfity mcaaurement are described, beginning with the quantitative methods. The more
qualitative methods are then described, followed by the
metfroda for measuring the nettability of fm@onal and
mixed-wettabfity cores.
Quantitative Nettability Measurements
Contact Angle. The contact angle is the best wettabiity
rncasurement method when pure fluids and artificial cores
are used because there is no possibility of surfactants or
other compnrmda aitcring the wettabthy. The method is
also used to determine whether a crude oil can alter wettabilhv and to examine the effects of temperature. uressure, and brine chemistrv on wettsbility~ Howe&, as
discussed in more detaif _latcr, some dlff%ultiea areinvolved in applying contact-angle meaaurementa to reservoir cores. Marty methods of contact-angle measurement
have been used. They include the tilting plate method,
sessile drops or btibbles, vertical rod method, tensiometric method, cylinder method, and capillary rise method.
Descriptions of these methods can be found irr AdmnSon>4 Jobnaon and Dettre, 19 Good, 20 Neumann and
0ood,21 and Popiel. 22 MO~t of these methods are nOt
generally used in the petroleum industry, however, because they are better suited for pure fluids with no adsorption or desorption of surfactants. Because of the
surface-active agenta in crude, a significant length of tine
is needed for a contact angle to reach equ~lbrium.
The methods that aregenerally used in the petroleum
industry are the sessile drop method 21.23-25and a modified form of the sessile dro method described by Iezch
ef al. 26 and Treiber et al. ! 7 In both methods, the nrineral crystal m be tested is moumcd in a test cell compixed
entirely .of inert materisls to prevent contamination. The
sessife drop method uses a single flat, polished mineraf
cryataf (see Fig. 1). The modified sessife drop method
uses two flat, polisbcd mineral crystals that are mounted
parzllel to each other on adjustable posts, as shnvnr irr Fig.
2a. Becauae sandstnnea are ohcn composed primarily of
quartz and limestones of calcite, quartz or calcite crystals are used to simulate the pore snrfaces of the reservoir rock. Obviously, the nettability of clays in the
reservoir cannot be examined tit-h this method.
The first step in mcastuing contact angle is to cl+ the
app~ams thoroughly, because even tmce arnounta of cOntammants can alter the contact angle. Then the cell cOn-

F1.a. 1Wettabilitv

of the oil/water/rnck

tainin~ the mineral crystals is tilled with brine.

DenW-gemted synthetic ~ormztion brine is used to prevent the introduction of foreign metal ions, which irrconcentrations of only a few parts per million can alter the
~e~bfi~.127
For the mOdifid aessile chop method, an
oil drop is placed between the two crystals so fhat it contacts a large area of each crystal. After the oil/crystal interface hzs aged for a few days, the two crystals are
displaced pamflel to each other. As shown in Fig. 2b, this
shifts the oil drop and allows brine to move over a portion of the surface previously covered with oil. The corrtact angles measured in this fashion are called
water-advancing contact inglea. A nonequilibriunr angle is obacrvcd inrmediately.after the drop is moved. Thk
angle decreases for a day or two until a constant value
is obtained for that age of the oiUsolid irrterfacc. The
oilhnineral surface is then aged further, the water is advanced again, and a new value is obtained.
The procedures sre similzr in the sessife drop methnd.
A drop of crude oil is formed at the end of a fine capillary tube and brought into contact with the flat mineral
surface (see Fig. 1). The droplet is allowed to age on the
surface. The water-advancing and water-receding contact
angles are measurd by usirrg the capillary tube to expand
zrrd contract the volume of the cmde-oil drop. 23,24
When the crude oil containa naturaf surface-active substances; the water-advancing contact angle increases aa
the OiUcrystafinterface ages, approaching a iimhing value
as adsorption equilibrium is reached. To achieve this liiting vahre may require hmrdrcds or even tbOm~ds Of
hours of interface-aging time. Fig. 327 gives examples
of tbe change in the water-advancing contact angle as the
oilholid interfke ages. This demon~es that early mcaa-

TABLE 1APPROXIMATE
RELATIONBHIP BETWEEN WSTTABILITY,
CONTACT
AND THE USSM AND AMOTT WETTABIUTY
INDZXES

Contact angle
Minimum
Maximum
USBM nettability index
Amott nettability index
Displacement.by-water ratio
Displacement-by-nil ratio
Amoti-Harvey nettability index

Joumal

of Peuoleum Technology, .November

1986

svstem. 16

Water-Wet

Neutrally Wet

00
60 to 75
W naar 1

60 to 75
105 to 1200
W near O

Positive
Zero
0.3s/s1.0

Zem
Zero
-0.3 </<0.3

ANGLE,

Oil-Wet

to 1200
160
W near -1

105

Zero
Positive
-1.0</s
-0.3

1247

:E
i

CRKiTAL

WATER

OIL

*~ l-&#.
..

CRYSTAL

l.

____..

Fig. 3-Approach

__J

to equilibrium

contact angla.m

[a)

1-

.1

I
I
I

1~

L__.___;
(b)

Fig. 2Contar$angle

measurement.

urements can show that the system is water-wet even

though it is actnally oil-wet at equilibrium. Because it contains no surfactmrts, the contsct angle for pure &cane
(Curve A) did not change from zero as the interface was
aged.
Orie problem in contact-angle measurements is hysteresis, because it is genirally found experimentally that
a liquid drop on a surface cm have msny different stable
contact angles. The contact angles reported in the literature sre either the water-advancing or water-~~g
FOntact angle because these two angles are the most
reproducible. Tbe water-advancing angle cmr also be
thought of as the oil-receding one. The advmrcing angle,
Oad,, is measured by pulling the periphery of a drop over
a surface, while the reding contact angle, @~; is meaaured by pushing it back. The dMference, O,~, - @,e., is
the contact-angle hysteresis snd can be greater thsn 60
[1 rsd]. 19
Johnson and Dettre 19 and Adanrson4 state that there
appear to be three causes of contam-zngle hysteresis: (1)
smface roughness, (2) surface heterogeneity, and(3) surface immobility on a macromolecular scale. To see how
surface ruughness can cause hysteresis, consider a
horizontal but rough plate. Because tie rough surface contains pesks and vslfeys, a liquid drop WY generslly be
attached to a surface that is not horizontal. The macmscupically observed contact sngle will not be the same as
the true contact angle on a microscopic scale. The roughness of the surface will 2U0w a large number of metaatable states of the drop to exist with different contact angles.
The surface roughness wilJ generally diminish the apparent cuntsct nn e for water-wet mck and increase it for
oil-wet reck. 19 ,22
1248

Hysteresis resulting from surface heterogeneity can be

caused by either heterogeneity in the rock surface composition or differential adsorption of wettabifity-akcring
compounds. This problem is generally avoided by mcas@g tie angle on a singfe crystal and by rigorously &anmg the entire appsmms before measurement. Finslly,
surface immobiihy can cause hysteresis by preventing the
fluid motion necesssry for the contact angle to reach ita
equilibrium value. For example, slow adsorption of a arwfactant from the soIid/Iiquid interface into the brdk liquid
can cause hysteresis. Some cmde oils w even form a
solid film at the oil/water interface. 2s-30Tlwse Iihna sre
particularly likely to form if the crude has been exposed
to oxygen, but have even been found in some anaerobic
cmdes.
Even though it is possible, with great care, to get exact
and reproducible contact-mrgle measurements, the question of how representative these results are of the wettabtity of reservoir core arises. The contsct angle cannot
mke into account rhe roughness, heterogeneity, and complex geometg of reservoir reck. First, consider the problems caused by muglrness. Morrow31 bsa pointd out that
roughness and pore geumetry will intluence the
oiJ/water/solid contact line and can change the apparent
contsct srrgle. On a smooth surface, the contact angle is
tixcd. On the sharp edges found in reservoir rock, however, this condition is relaxed, and there is a wide range
of pussible contsct angles. 31.32Morrow postulates that
most of the oillwatcrfrock contact limes will be located
at the sharp edges because, at these edges, the cuntact an-.
gle can change without movirr~ the position of the contact line.
A second problem with applying contact-angle measurements to reservoir rocks is that the contact angle cannot take into account the heterogeneity of the reck surface.
Contact angles sre measured on a single mineral crystal,
while a core contains many different constituents. As discussed preciously, surfactmrts in the cmde can affect the
wettzbility of the sands and clays differently, causing
localized heterogencmrs wettabifity.
A Wlrd limitation is that no information can bs gaind
about the presence or absence of permanently attached
organic coatings on reservoir rocks. 33 These films can
be detected ordy by making other nettability measurements. This is particrdsrly importrurt when workirrg with
~e~mr~-mte ~ores. 3435 Before the original we~bfity
can be restored, all the adsorbed materials must be removed, wbicli will generally leave the cure in astfongly
water-wet stste. The only way to determine whether the
cleaning prccess b+ been successfid is to measure the wctJo.mal of Petrol..m Technology,November1986

tabiilty of the cleaned core. If it is not strongly ivaterwet, addhional cleaning is necessay.
AIuott Method. The Amott method 6.8.36combines im.
hibition and forced displacement to mcaaure the average
wetmbili~ of a core. Both reservoir core and fluids can
be used in the test. The Amott method is baaed on the
fact that the wetting fluid will generally imbibe spontaneously into the cure, dkplacing the nonwetting one. The
tatio of apontaneouaimbibition to fumed imMbition is used
tu kduCe the influence of other factors, such m Elative
permeability, viscosity, and the initial saturation of the
rock.
Core is prepared by centrifuging under brine until the,
residwd oil saturation (ROS) is rcabhed. The Amott wettabilhy measurement then consists of the following four
atcps: (1) immerse the core in oil, and meamre the volume
of water displaced by the spontaneous (free) imbibition
of oil after 20 hours; (2) centrifuge the core in oil until
the irreducible water saturation (lWS) is reached, and
measure the total amount of water displaced, including
the volume displaced by spontaneous imbibition; (3) immerre the core in brine, atrd measure the volume of 01
spontaneously displaced by imbibition of water after 20
hours; and (4) centrifuge the core in oil until ROS is
reached, atrd meaaure the total amount of oil displaced.
Note that the core may be driven to IWS and ROS by flow
rather than with a centrifuge. This is especially necessary
for unconsolidated material that cannot be centrifuged.
The tmt restdta are expressed by (1) the displacementby-oil ratio the ratio of the water volume displaced by
spnntatmms oil imbibition alone, VW,P to the total displaced by oil imbMion and centrifugal (forced) displacement, Vm,

~o=+, .............................(3a)
w,

and (2) the displacement-by-water ratio the ratio of

the oif volume displaced by spontaneous water imbibition, V..p, to the total oif volume dkplaced by itnbibition and ccntritkgal (forced) displacement, Vol:

L=+. ..
0,

. . . . . . . . . . . . . . . . . . . . ..(3b)
..

As shown in Table 1, preferentially water-wet cores

have a pusitive displacement-by-water ratio and a zero
value for the displacement-by-d ratio. The displacementby-water ratio approaches 1 aa the water wetness increases. Simifarly, oif-wet cores have a positive
displacement-by-nil ratio and a zero displacement-bywater ratio. Both ratios are zero for neutrally wet cores.
Amott chose an arbitraty time period of 20 hours for
the spmmmeous oil and water imbibition steps in his
metbcd. If possible, we recommend instead that the cores
be allowed to imbibe until either imbibition is complete
or a time liit of 1 to 2 weeks is reached. imbibition can
@e from several hour.. to more tlIan 2 months to complete. M If the imbibition is stop~d after a short period
of time, then the measured spontaneous itnblbition volume
will be lower than the equilibrium value for lowJaud

of

PetroleumTechnolo~,i%enber 1986

permeability samples, causing an underestimation of 60

or ~~. 8.36The mcaaurcd displacement ratios wilf underestimate tie water- or oil-wemess of the ruck. Of course,
it is necessafy to choose some upper time limit to finish
Ore measurement in a reasonable length of time. If the
core is still itnbibmg when the time limit is reached, however, then the measured spontaneous imbibition volume
will underestimate the rcaemoir wettabfity, and the AIZIOtt
ratios should be interpreted with caution. 8.36
A number of researchers37.3s used a modification of
the Amott nettability test called the AntomHarvey relative displacement index. This procedure haa an additional step in the core preparation before the testis run:
the cure is centrifuged first under brine and then under
crude to reduce the plug to fWS. The displacement-bywater and dkplacenzent-by-oil ratios are then calculated
by the Atnott method. The Amott-Hawey relative displacement index is the displacement-by-water ratio minus the displacement-by-oil ratio:
- . . . . . . . . . . . . . . . ...(4)

I=8W80=++.

01

This combmes the two ratios into a single wettabflity

index that varies from + 1 for complete water wetness
to 1 for cnmplete oif wetness. Cuiec 39 states that the
system is water-wet when +0.3s I< 1, intermediate wet
when 0.3<1<0.3,
and oif-wet when 1 S2S 0.3.
The Amott test measures the total volume of spnntaneoua and forced oif and water imbibition. Morrow et al. a
developed a modified form of rhe Amott test, which nfao
measures the imbibition rates during the sfmntammus imbibition measurement. In tbk test, the core is suspended
in oil or water from an electronic bafance by a small
monofilament line. Weight change is monitored ac a tilnction of time as the spontaneous imbibition occurs. The
nettability of the core is then determined from both the
Antott weteabifity index and the spontaneous imbibition
rates. This test may offer snme advantage over the atandad Amott tmt because it is based on additional .@ata.
The main problem with the Amott wetmbfity test and
ita modifications is that they are insensitive near neutral
wetmbfity. The teat measures the ease with which the wetting fluid can spontaneously displace the nonwetting one.
However, neither fluid will apnntanemrsly imbibe and dispkice the other when the contact angle varies tlom roughly
60 to 120 [1 to 2.1 rad]. 3.4145 In addition, the limiting
contact angfe above which spontammus intbibltion wilf not
occur depcnda on the initial saturation of the core.
McCaffery41 and Morrow and McCaffery4s examined
the spontaneous imbibition of pure fluids with a known
contact angle into a synthetic teflon core that initially cOntained air. IiI the experiments, n-dodecane @=42 [0.7
rad]) wotid always spontaneously imbibe into the core,
while ct-bromomphtbalene (0=73 [1.3 rad]) would never
spontartermsly imbibe. Dioctyl erfrer (0=49 [0.9 rad])
would not imbibe into a dry core but would imbibe into
a core with an initial diomyl ether saturation of 30 %. Tbe
Amott method would classi@ the teflon core with ndodecane aa mildly water-wet, while c+bromomphtfmlene
would be neutmlly wet. The core with dioctyl ether
(6=49 [0.9 rad]) would be classified as water-wet if it
already conmincd some fluid or as neutrally wet if the
core waa initialfy dry.
1249

*
10

WATER

SATURATION,

I 00
iVERAGE

~ OIL WET LOG A,,A2=-O.5+

PERCENT

I 00

AVERAGE

WATER

SATURATION

PERCENT

Ii

k/ERAGE

NEUTRAL

WATER

LOQA,/A~=

SAT URATION,

0.02

\ PERCENT

!
lg. 4-USBM
wettcbility measurement 9: (Ibrine drive. U-oil
osilanes, (c) core pretreated with oil for 324 hours at 140R

USBM Wettabiity Index. The third quantitative test that

is used to measure tie nettability is the USBM test developed by Donaldson et al. 9,17,18The USBM test olso
measures the average nettability of the core. The test is
relatively rapid, requiring a few &ys to test four to eight
plugs. A major advsntage it has over the Amott wettabflity test is its sensitivity near neutral nettability. A minor
disadvantage is that the USBM wettabili~ index can only
be measured on plug-size samples because the samples
must be.spun in a centrifuge. The USBM test compares
the work necessary for one fluid to displace the other.
Because

of the favorable

free-energy

change,

the work

required for the wetting fluid to displace the nonwetdng

fluid from the core is less than the work required for the
oppsite displacement. It has been shown that the required
work is pro rtionsl to the area under the capillary presaue ~umeY,46 ~ ~ti~~ ~~~~a, wh~~ a ..~~ is ~at~~.
1250

drive) (a) untreated core, (b) cora treated with organochlobrine contains 1,000 ppm sodium tripolyphosphdte.

wet. the area under the bfinedrive cntdlkwv messure

cu~e (when the water displaces the oifj is srn~ler than
the area under the capillary pressure curve for the reverse
displacement. In fact, if the water-wetting is stiong
enough, most of the water wifl spnntmeously inddhe into
the core, and the area under the brine-drive curve will
be very small.
Before the testis mn, plugs are prepared by centrifigation under oil at high speed to drive them to lWS. This
point is denoted by the satefiaka (*) in ,Flgs. 4a though
4c, which represent wettabfli~ test results in cores with
three tifferent sufface treatments. During the USBM
measurement, a modified version of the procedure describd by Hassler and Bnmner47 and Slobod et al. 4s is
used to calcnlate the centrifigaf capillary pressures. (The
USBM method uses the avefage saturations in the core. 17
J.o mntrsat, the centrifugal capillary pressure cufve is
Journalof PetroleumTechnology,November1986

based ontbe saturation at the face of the core, which is

czfculated from the ayerage saturation by the method
found in Ref. 47.) In the first step of the measurement,
cores are placed in brine and centrifuged at imcrementzlly increasing speeds until a capilla~ pressure of 10 psi
[70 kpa] is reached. This step is known as the brine
drive becauae brine displaces oil from the core. At each
incremental capiflary pressure, the average saturation of
the plug iz calculated from the volume of expelled oif.
Curve I (Figs. 4a through 4c) is a pIot of capiIlary przsaure vs. the average saturation for the brine drive.
In the second step, the core is placed in oif znd centrifuged. During this oil-drive step, oil dkplacea brine
from the core. As in the first step, the capillzry pressures
znd averzge saturation are measured until a capillary
pressure of 10 psi [70 kpa] is reached. In each czae, the
curves sre Iinezrly extrapolated or truncated if the last
pressure ia not exactly 10 psi [70 Wa]. The results of the
oil drive are plotted as Curve II in Figs. 4a through 4c.
The USBM method uses the ratio of zreas under the.
two capillary prezzure curves to czlctdate a we~btily index according to l?.q. 5..
w=10g(,4, /A*), . . . . . . . . . . . . . . . . . . . . . . . . ...(5)
where,4 ~ snd A2 we the areas under the oil- and brinedrive curves, respectively. As shown in Table 1,. when
W is greater thzn zero, the core is water-wet, znd when
W is less that zero, the core is oil-wet. A wettzbilty index nezr zero mezns that the cure is neutrzlly wet. The
lzrger the absolute vzlue of W, the greater the wetting
preference.
Exzmplcz of water-wet, oil-wet, znd neutrzlly wet cores
sre shown in Figs. 4a though 4C for an initizlly waterwet outcrop Torpedo sandstunc core. Fig. 4a shows the
USBM wettahility index of the untreated water-wet core.
The srea underthe oil-drive curve is much lsrger thsn
rhe zrez under fhe water-drive curve, yiehiing a wettzbility index of 0.79. In Fig. 4b, tbe core was treated with
zn organosibne compound, which rendered it oil-wet. Ike
zrea under the oil-drive curve is now much smzller thsn
the mea under the water-drive curve because oif is the
wetting fluid, yielding a wettabWy index of 0.51. In
Fig. 4c, the core was aged with crude, znd the brine wzz
treated with sodium tripoIyphosphate. The core is now
neutrzlly wet, znd both of the areas zre equal, mzking
the USBM wettabfity index zero.
A major advantzge of the USBM wettzbility test over
the Amott testis its sensitivity nenr neutrzl wettzbflity.
On the other band, the USBM test cannot determine
whether a system has fractions or mixed wettzb~ity,
while the Amutt test is sometimes senzitive. In some
fractional- or fixed-wet systemz, both water znd oil will
fiblbe frmly. 49-51me AMUti method will have pOSitiVe
displacement-by-wster znd displacement-by-d rstios, indicating that the system is nonuniformly wetted.
Cumbined Amutt/USBM
Methud.
Sharma and
Wunder1ich51 bzve recentfy developed a modification of
the USBM method that ZIIOWS,
the czdculation uf buth the
Amott ind USBM wettzbdity indices. The procedure,
shown in Fig. 5, haa tive steps: (1) initizl oil drive, (2)
spontaneous (free) imbibition of brine, (3) brine drive,
(4) spontanecma (free) imbibition of oil, znif (5) oil drive.
November19s36
Journal of PetroleumTechnology,

The zreas under the brine- znd oil-&lve curves zre used
to calcufate the USBM index, while the Amott index uses
the volumes of free and totzl water znd oil displacements.
During the initizl oil-drive step (Curve 1), the plugs are
driven to IWS. Next, the cores we immersed in water,
and the vulume of water that imbibes freely is measured
(Curve 2). During the brine-dive step (Curve 3), the average saturation of the plug is determined from the zmount
of expelled uil at each incremental capilla~ pressure.
These data zre used to cslculate the area under the brinedrive curve, A2, for the USBM method. At the end of
the brine-drive step, the plug is left at ROS. The Amott
displacement-by-water ratiu, 5 ~, is the ratig of the oil
vulume displaced by free brine imbibition to the totzl
volume displaced by free imbibition znd centrifugzJ dkplacement (Eq. 3a).
In the fourth step (Curve 4), the plug is immersed in
oil, znd the volume uf oil thzt imbibes spontaneously is
measured. The finzl step is the oil drive (curve 5), where
the czpillmy pressures znd average saturations are used
to czhxdate A, for the USBM method. Eq. 5 is then used
to calculate the USBM wettzbili~ index. At the end of
the oil drive, the plug is left at IWS. The Amott dkplacement-by-oil ratio, 60, is the ratio of the free oil imbibition to the totzl volume displaced by free imbibkion and
centiitlzgal displacement (Eq. 3b).
There zre two advzntzges of the combined USBM/
Amott method over the stzndard USBM method51: tbe
resolution of the USBM method is improved by accounting for the saturation chznges that occur at zero capillary
pressure, znd the Amott index is ZISOcalculated. As discussed eadier, the AMOttmethod will sometimes indicate
that a system is nonuniformly wetted.
Qualitative Wettabiti,tyMeaaurementi
Imbibition
Method. The most cummonly used qunMztive nettability measurement is the imbibition method, 52.57because it gives a quick but rough idea of the
wettsb~hy without requiring zny complicated equipment.
The originzl imbibition appzrztus tested the wettabfity
at mom temperature and pressure. 52
More recently, Kyre er al. 57 described a modification
of the apparatus that zllows wettzbility to be measured
at reservoir conditions. In zn imbibition test, a core at
IWS is first submerged in brine undementb a graduated
cyliider, znd the rzte and zmount of oil dkplzced by brine
imbbitiun zre measured. The core ii strongly watier-wet
if l.zrgevolumes of brine are rzpidly imbibed, while lower rates znd smzller volumes imply a more wezkfy waterwet core. If no water is imbibed, the core is either oilwet or neutrzlly wet. Non-water-wet cores zre then driven
to ROS znd submerged in oiI. The imbibition apparaNs
is inverted, with the grzduated cylinder below the core
to measure the rate and volume of water dkplaced by uil
imtdbltiom If the core imbibes oil, it is oil-wet. The
strength of uil-wetness is Miczted by the rste and volume
uf oif imbibition. If neither oil nor water is imbibed, the
cure.is neuuslly wet. Finzlly, some cores will imbibe both
water ~d ofi. 49-s] These cores hzve either fiaCtiOnidor
mixed wettab~ky. One problem with the imbibition
method is that, in addition to wettahiity, imbMion ratez
zlso depend on relative permeabfity, viscosity, IFT, pure
structure, znd the initial saturation of the core. 3,10 Frequently, this dependence on other vzriables ia reduced by
1251

1.0
0.8 -

0.6 -

zL
:

0.4-

$2
K

5
5
#
:

0.2WATER

WATER

SATURATION

PV

(0)

100
1(

AvERAGEWATERSATURATION,
PERCENTW
Fig. 5Combhmd

Amott/USBM

method.51

comparison of the measured imbibition rate with a reference rate measured when the core is strongly water-wet.
To do this, the core is cleaned by heating at 750F
[400C] for 24 hours to oxidize alf of the organic material, leaving the core strongly water-wet. The core is then
remturated m ita origiual oiJ aatnration with a refined white
oil having the same viscosity aa the crude oiJ, and fhe
reference imbibition rate is measured. Denekaa et al. 53
~~~
nettability changes in terms of the relative rate
of Imbibitiorx
WATER

R=<,
m.

. . . . . . . . ... . . ... . . . . . . . . . . . . . . . ...(6)

where
If = relative rate of imbibition,
A = initial imbibition rste of the cnre just after
it is submerged (cm3 /s), and
ritr = iuitial imbibition rate of the cleaned,
strongly water-wet core (cm3 Is).
If,the core is water-wet, m is the initial water imbMiOn
rate. If the core is oil-wet, ri is the initial oil imbibkion
rate, and the relative imbibition rate, R, is reported as
a negative number. Note, however, that whife the use of
a reference rate reduces tie effect of other variables, tie
irnbMion method still suffers from the same problem as
the Arnott method-insensitivity near neutraf wettabifily.
Bobek et al. 52 alau suggested an imbibition.teat for unconsolidated cores. In tbia test, a thin layer of sand is
spread on a microscope slide, after which the oif saturation is increased by addition of a refined minemf oil.
Dropleta of water am then placed on the surface of the
sand, srrd the movement of the fluid k observed by rnicm1252

SATURATION

O/. PV

(b)
Fig. 6Typh2al water/oil relative pemreabilily curvaa bc!
on the effective permeability to oil, at the resetvoir cl
nate water saturation; (a) strongly water-wet rock,
strongly oil-wet rock.

scope. Jf the ample is water-wet, the water wilf move

readiiy into the as.nd, displacing oil from the surface of
the sand mains. In addition. the oil will form sr3hericaI
droplets, ~dicadng that it is he nonwetdng phaaeuA sindIar procedure is used to test for oil wettabiMy. This test
is similar to the microscope examination method.
Microscope Exmrdrratiotr. Microscope examination is
sometimes used in laboratory flow viaualiiation studies.
The wettabtity is determined from w description of the
flow on a single pure level in an ideafized porous mediUITl during Watedooding. 58.59TM.Sdescription includes
the structure of the residual Oi and the changes in the locution of tbe oil and water that occur during watertlooding. ff the system is stfongly water-wet, the water
surrounds the groins as a thin fifrxr.The large pools of
residual oil rest on a water fti, while the smafIer drops
Journalof PetroleumTechnology,November1986

of residual oil form spherical drups in the center of the

pores. If the system is intermediately wet, ,both oil and
water will be found in contact with the rock surfaces, ,~d
both can be found in the smslf pores. Finally, if the system is oil-wet, the roles of the oil and water are reversed.
The ofl forms a fti around the grain surfaces and is found
in the small pores, whfle the water rests on an oil tilm
or forms small spheres.
The method of quzlitstively dete ruining the wettabfity by microscope examination is pmticukly important
in the study of wettabil@ reversaJs, 58>m&one of the
propoacd mechanisms for EOR that occurs dining alkaa chefi~
fine waterfloodmg. m In ~e~e ex~tienta,
that changes the nettability is injected into the porous
medium during a waterflood, causiug a zone of wettabdity reversal to propagate through the core. A microscope
is used to follow wettabtity changes and to determine
whether EOR will occur by thk mechaniam.
Methods.
Flotation methods are fast but work
only for strongly wetted systems. In the sinrpleat method,
water, oil, arrd sand are placed in a glass bottle. The bottle ia shaken, and tie experimenter observes the fate of
tie smd gains. 65-6s This method is recommended by
API for determining the effects of surfactarrts on nettability. C If the system is strongly water-wet, clean sand
grains will settle to the bottom of the bottle. Sand grains
placed in the oil wilI aggregate aud form small clumps
of grains surrounded by a thii layer of water. If the system is oil-wet, some of the grains can be suspended at
the oil/water interface. Oil-wet sand grains in the water
will clump together, forming small oil globuka coated
with sand. This flotation system is qualhative and worka
only for strongly wetfed systems.
SeVeFJ
experimenters 69.70 have used more elaborate
flotation tests developed in the minin industry that were
based on liquid/liquid extraction. 22,7? In these tests, particles are initially suspended in water. A second fluid,
either oil or sir, is bubbled from below. The parricles that
are water-wet remain in the water, while the hydrophobic, oil-wet particles adhere tu the oil (air) and rise to the
surface. The fraction of articles in each phaae can then
be measured. elenrmtz% used the flotation method to
meaaure the wettsbihty of small clay particka, which cmrnot be conveniently me$suked in any other way. Untreated, strongly water-wet particlea would not float. After
expusrue to cmde, the clay particles floated, demonatratirrg that their nettability had been altered:
Flotation tests based on liquid/liquid extraction appear
tn divide particles intu two categoria$ stron 1 water-wet
~~
and mildly water-wet tu strongly oil-wet. ~2~~BesidG
,
wettabilky, flotation of a particle also depends on particle size, particle density, and IFT. A small particle with
low density and high IFT might float if the contact angle
was ~eater than about300 [0.5rad]. On the other hsnd,
the minimum contact angle for flotation of a large, dense
particle could bc 90 [1.6 rad]. 72,73
Flotation

Glass SIide Method. Another esrly qualitative nettability meaaurcment tcduzique is the @as slide methnd, 30,m
which assumes thst a glass surfsce is representative of
the reservoir rock. A clean, dry, glaas microscope slide
is suapmrded in a layer of cmdc oif floating on water in
a transparent contsiuer ~d aged. The glaas slide is then
Journal of PetroleumTechnology, November 1986

lowered iuto the water. If tie slide is water-wet, the water

quickly displacea the oif on the slide. On the other hand,
if the slide is oil-wet, a stable oil-wet tilm is formed, and
the oil is very slowly d~placed. Reisberg and Doscher30
aged slides in crude oil snd found that it took up to 30
days for the 6MI wettabtiiy to be reached. Cwke st al. 5s
used a simple variation of the glass slide method as a
quick, qrmlitative test tu screen different acidic-oil/alkalii~water comblnationa for uae in zdkaliie waterflooding experiments They placed oil and water without
-g
fi a glass vkd and waited to see whether a stable
od-wet tihn formed on the vial. This was determined by
tilting the vial srrd seeing how the water and oil behaved
on the previously oil-covered surface.
Relative Permeabfity Methods. A number of quaMative methuda are baaed on the effects of wettabfity on relative penneabtity. However, they are all suitable ordy for
discriminating between strongly water-wet and strongly
oil-wet cores. A smaller change in nettabilitye.g.,between strongly and muderstely water-wetmay not k nntimd by these methcda. One method developd by Ehrlich
~d WYg~74 is based on the rules of thumb given by
Craig7 to differentiate be~een stmmgly water-wet aud
~mon~y ~fi-wet ~or=. C~g, ~ 16,52,75 fieS ~ f ~mb me
. .

aa fouows.
1. Connate water saturations are usually greater than
20 to 25% PV in a water-wet rock, but less than 10%
PV in an oil-wet rock.
2. Water aaturztion at which oil and water relative pcrmeabtities are equal is generally greater than 50% for
water-wet cores and less than 50% for oil-wet ones.
3. The relative permeabfity to water at floodout is
generally less than 30% in water-wet rocks, but frnm 50
to 100% in oil-wet ones.
These relative permeabtities are based on the oil permeabtity at the comate water saturation. Examples of
relative perrneabfity curves in strongly water-wet snd oilwct corer taken from Craig7 are given in Fig. 6. Note
that Raza et al. 16 state that there are exceptions tn the
gezieral rule that the connate water saturation is higher
for a warcr-wet rock than for an oil-wet one.
Treiber et al 27 pro~sed a secrmd quali@ve technique
for strongly wetted rocks The method comp=es the
oil/water, gas/oil, and gas/water relative permeabilkie.s
and takes advantage of the fact that relative permeability
:E.;E?JF2:EP;:,:EZZ$2:E;
water-wet, the relative perrneab~ky tu oil (the preferentially wetting phaae with respect to the gas) in the gas/oil
relative pcrmeabtity test shordd be a continuation of the
relative permeabdity to the water (the wetting phzae) in
the water/Oil relative pcmzeabil~ test. 76 If significant
differences are observed, the sample is not strongly
water-wet.
An example of the comparison of the relative perr22eabtity curves in a strongly water-wet core taken from
Owens and Archer76 is shown in Fig. 7. The gas/oil
drainsge relative perrneabfity, where the oil ia the stronglywetting fluid, is shown as the dotted Iiues. The water/oil
relative pcrrneabfity, where tbe water is the strmrgly wetdug fluid, ia shown aa the solid lines. Note that the water
relative penneabi3ity, where the wetting fluid saturation
is iricreasing, is a continuation of the oil relative pe2n2ea1253

o
CO

OIL

h
a%

.:

*3

AA

AAA

&

WAT&T
ROCK
(NUGGET SAND]

AA

\GA2
\

\
\
\

wATER

~3

C.L

00

./

OIL-WET

ROCK

[SWINGER ~D1

.0

Ok
1

20

40

60

60

0.1

WE771NGFtlXi5 SATURATION,PERCENTFORE ?PACE

i9. 7cOMpariSOfl of gasioil drainage and water/oil im.
ibition relative
andstone. 76

permeabltity

relationships.

Torpedo

bditv. where the wetting fluid saturation is decreasing.

Thi~ demonstrates that iite core is water-wet.
Ba~c~ et al. 78 developd a third wettSb~@ memu~-
ment technique that is based on unsteady-state relstive permeability. Their method uses the capillary end effect that
occurs when a core initially at IWS is waterflooded at a
constant, slow injection rate. The end effect is the accumtJIation of wetting phase near the outlet end of the core
caused by the discontinuity between the porous medium
~d the otfet pipe. 79 An ihcreased pressure drOP cm
occur because of tbia wetting fluid accumulation. Batycky
et al.s relative-permeab fitytwettabihty tests are run at
very slow flow rates, so end effects are very important
in determination of the pressure drop across the core. In
contrast, standard unsteady-state relative pertneabilty
measurements use high flow rates to minimize the end
effect.
Batycky et a-f.determined the wettabfky by waterflcoding the core at very low rstesuntil the ROS WS.V
reached.
The flow was stopped to alfow the fluid to redistribute,
then restarted in the reverse direction. The core k waterwet if there is no change in the pressure drop after the
flow reversal snd oil-wet if the pressure drop is reduced
immediately after the reversal. In a water-wet core at
ROS, the wetting fluid saturation will bc high throughout the cure, with no addkional water accumulation at ibe
outlet end. 7s,s0 There will be no redismibution of fluids
when the flow is stopprG consequently, the pressure drup
will not change. On the other hand, if the core is oil-wet,
capillaty fortes will cause oil (the wetting phase) to .accumtdate near the outfet. The pressure drop caused by
12s4

CONNATE

~9. 8Relatiom~~
!urpermeabltity.

WATER SNUR.T,ON % W

betwean connate water saturation at

this oil accumulation is detected by stopping the flow,

thereby allowing capillary forces to redistribute the oil
evenly throughout the core. When flow is started in the
reverse direction, the pressure drop will initially be lower, gradually rising to its original value as the end effect
is re-established on the opposite end of the core.
Permeahiity/Saturation
Relationships. Two qwdiative
methods based on sir permeability and fluid saturations
have been proposed. Both methods are statistical, require
a relatively tige number of sqnples, and give only a very
rough idea of the wettabfli~. The advantage of the
methods is that only routine core anafysis measurements
are required. However, the reliability of these methods
is unknown. The methods are afso limited to core samples without significant fractures or vugs, in which the
pore structure determines the air permeability.
~
et al. 16 proposed an empirical methud to determine reservoir nettability based on connate water saturation and air penneabihy. To obtain the connate water
sau.tration,core is obtained with an oil-based drilliig fluid,
then the freshly cut cores are analyzed for their water content. The cores are extracted and dried, and the air permeabi@ is messured. A qualitative measure of the
wettabtity ii obtained by plotting the comate water saturation vs. the sir permeability. Fig. 8 shows exanrples
of tlte plot for strongly oil-wet and smonglyw6ter-wet conditions. 16For the oil-wet case, the average connate water
saturation is generally relatively low. The curve is nearly vertical and extends over only a smsll saturation interval. Conversely, for the water-wet reservoir, the curwe
Jourmtof PetroleumTechnology,November1986

has a gentle slope and extends over a large saturation interval.

Frehse81 proposed a second statistical method .b,ascd
on the assumption that low-permeability core samples wiJl
have a higher wetting-phase saturation than the highperrnmbilhy ones. For a uniformly wetted rock, the small
pores are fdled with the wetting fluid, while the large
pores contain both the wetting and nonwetting fluids. In
comparison to higher-permeability samples, a lowpermeability sampIe will generally have spore structure
containing a larger number of small pores that are filled
with the wetting fluid. To determine the nettability,
Frehseclassifica the routine core mralysis samples into
different permcab~ky ranges. The saturation distributions
for the bighfit and lowest permeability ranges are then
compared. For example, consider a core taken with a
water-based mud, where the reaiduaJ oil saturations are
known. The rcacrvoir is assumed to be oil-wet if the 10Wpermeability samples haye a higher average ROS and
water-wet if the high-permeability sarnpIes have a higher
oil saturation. Currently, this method appears to be theoretical only. We are not aware of auy.tests comparing the
results of tbk method with more standard nettability
measurements, such as tbe Amott nr USBM indices.
We feel that wettabflty evaluations based on air permeabili~ and fluid saturations should not be used at present. Rara et d.s method is empirical, and it is not kuown
whether it is generally vrdid. Frebaes method hw not been
tested. UntiI tbesemethods are evaluated hy comparison
with standard wettabdity measurement, they should be
considered mrreliable.
Grigorevs2. proposed a theoretical method for determining an appwent contact angle based on the IWS and
ROS. The method is probably not generally valid. It is
based on a large number of unprovenasaurnptions about
the bebavior of the waterloillrock system. In addhion,
there do not appew to be my testa comparing thk method
with other wettabdit y measurements.

CapilJ~3 Pressure Curves. As far back as 1951,

suggested that tire entire capillary pressure
Calhoun
curve should be used to measure the nettability of the
core. Gatenby and MaradenM were the first to examine
the use of tbeareas under the capillary pressure curves
for thk purpose. The capillary pressure curves used were
the complete drainage and imbibition curves for both pesitive and negative capillary pressures measured by the
porous plate method. The two areai that they examined
were the total area surrounded by the drainage and imbL
bhion capillary pressure curves and the area under the
oil-drive c!rrye. They found that neither of these areas correlated well with the nettability of the cnre. However,
Donaldaon et at. 9 later showed that the areas that should
be measured were the areas under both the oil-drive and
brine-drive curves. Tbii is the basis of the quantitative
USBM method dkcussed earlier.
Capillarimetric Method. Johansen and Dunnirigg547 developed a qimhtative wettabdity measurement that measured the adhesion tension, a cos .9, in a glass capillary
Nbe. In this capilkwimetric method, the top of the tube
is connected to a column filled with oil, wh]le the bottum
is connected to a column filled with water (see Fig. 9).
The top of the water colunru cm be raised or lowered relaJoumalof PetroleumTcchnoloa, Novgmber1986

Fig. 9CapilIarlmetric

method.

tive to the oil cohunn, changing the hydrostatic head. As

the hydrostatic head is changed, the oil/water interfaces
will rise or fall in the tube until the capillary forces balance the gravitational forces:
P,=

20 cos e

=$(poho-p~hv+),
r

. . . ..

. . (7)

where
r = radius of the capillary tube,
P. = oil demi~,
p ~ = water density,
ho = height of the oil column above the
oil/water interface, and
h~ = height of the water column above the
o-illwater interface.
Eq. 7 can be rearranged to calculate the product of aand
cos 0, which Johansen and Dunning called the dkplacement energy (adhesion tension):.

ED=acos

@= T(poho-pWhJ.

. . . . . . . . . ...(8)

The dkplacement energy is positive if water wets the

glass and negative if oil weta it. If one of the liquids completely wets the glaas, then the contact angle is zero, cos
8 ia unity, and the displacement energy is equal to the IFT.
Johansen and Dunning usually changed the height of the
water column so that the interface moved over an area
previously cnvered by oil; hence the contact angle in Eq.
8 k water advancing. The capillarimetric methcd assumes
that glass is representative of the reservoir rock and there:
1255

fore is generslly only qualitative. Because this method

measures the product .s cm 0, the problems dkcussed in
the section on contnct angles SISOhinder this meifmd.
Displacement Capiffary Pressure. One of the earliest
wettabWy measurements was the displacement capillsry
pressure method, which uses the threshoId capillary pressure to calculate am apparent contact augle. 8894This
method is now used infrequently, however, because pore
geometry effects can cause the calcqated contsct angle
to differ greatly from the contact angle measured on a flat
plate. 3 The displacement (or threshold) capilla~ pressure is the capillaty pressure at which nonwetting fluid
will first enter a core initially 100% satorated with the
preferentially wetting fluid. h apparent contact sngle is
cdcnlatcd from the threshold capillary prssaure by mcdeling the mck as a straight, cylindrical capillary tube4, 10:

20 Cos 8=

P~=,

. . . . . . . . . . . . . . . . . . . . . . . ...(9)

Tmx
where PT is the displacement capillary pressure, a is the
IFT, O= is tbe apparent contsct nngle, and ris the
radius of the pure through which the nonwetting fluid begins to enter the core. Because the capillsry pressure needed to inject nonwetting fluid is reduced as the pore radius
is increased, rmax is m mfersge of the radii of the largest pores in the core. Note. @it one limitation of this
method is that it examines the wettabfity of only the lsrgest pores. Because Eq. 9 has two unknowns, 0. ~d
rm=, the oofy way to solve for the apparent contact angle is to mske additional assumptions. It is usually
assumed that some fluid exists that wfil completely wet
the core, so cos 8=1, and rmax csn be calculated. This
allows the contact angle, to be computed for other fluid
pairs.
Slobod and Blum~3 proposed two aenriqnantirativewettsbility measurements baaed on the displacement capjlfary pressure, the wettabfity number, and the apparent
contact angle. The wettabfity number is calculated by carving out two dkplacement experiments-first, water by
oil, and second, oil by sir. Eq. 9 for the oillwatcrlrock
system becomes
2U0-W Cos &o.w
P(o-w)r=

. . . . . . . . ... . ... (lOa)

TIM.

and for the nirJoil/rock system,

2U=.-0 Cos %.-O
P(.-O)T=

. . . . . . . . . .. . . . . ..(lOb)

rmax

In hth equations, the radius of the pore is assumed to

be the ssme. The wettrrbiMy number, N, is determined
by solving Eq. 10 for the ratio of the cos O terms:

cm O.-.
N==
Cos 6.-0
12.56

D.-OP(O-W)T

. . . . . . . . . . . . ..(11)

uo-w%m

Slobod snd Blum stated that if it were assumed that the

oif is.completely wetting in the oillairlrock system, then
cos .9..0 is unity. An apparent contnct angle for the
oil/water system can then be computed from Eq. 11:

cOs(oo_w)= =

C.-OP[O-W)T
... . . . . . . . . . . . . . ., (12)
Oo-wp(a-o)T

Slobod snd Bhuu resEzed that their nasumptions were

onfy approximately true and that the contact angle that
could be calculated from the displacement pressure was,
at best, only semiquantitative. In general, the apparent
contact angle measured from the displacement pressnfe
is not cqusl to the contact angle messured on a smooth
surface because of pore geometry effects. Morrow snd
MS c0workers4345 compared apparent contact ~~es
computed in sintercd teflon cores using pure fluid with
the true contact sngles measured on a smcotl teflon plate.
There wss no chnnge in the apparent contact sngle when
the tme contact angle was varied from Oto 22 [0 to 0.4
rsd]. In addition, when@ was greater than 22 [0.4 radl,
the apparent contact sngle wna always less thsn the true
contact angle. Fiiy,
in some cases, the apparent cOntact angle calculated from the dkplacement pressure csn
show the wrong fluid to be the wetting phase. Positive
displacement pressures for both fluids, particularly when
the mre is initially 100% saturated with the other fluid,
have been frequently
reported
in the literature. 41.U.B,W,W.95-W
When a positive displacement pre3aure is required for both fluids, the fluid with the lower
displacement pressure is the preferentially wetting flqd
because less energy is required to force. it into the
provides further discussion.
core, S3.98 &der~0n3
proposed a method to measLogs. Grshamw
ure the.nettability of in-situ reservoir mck with logs that
wsa baaed on rhe fact that the electrical resistivity of an
oil-wet rock is hi her than that of a water-wet rock at the
ssrne saturation. f In Grahams method, the formation is.
injectd with brine, and resistivity logs are run. The formation is then injected with the same brine containing a
reverse wetting agent, which will change a water-wet fornrstion to sn oil-wet one; if the formation is already oilwet, the reverse wetting agent will not alter the wettdility. After logs sre rerun, the nettability of the formation
cm be determined by comparing the two resistivity measurements. If the formation was originally water-wet, the
change to oil-wet will increase the resistivity. If the fOrmation was oil-wet, no change in rcsistivity will be observed.
Holmes and Tlppie lm proposed a second method that
compsres lugs with core data. The saturation in a formation is first measured with logs and the data converted
into a capillsry pressure curve. Next, the capillaW pressure is messured in a clean water-wet core where it is
sssumed that the contact angle is zero, snd the two cspiflnry pressure curves nre compared. If they agree, the
reservoir is strongly water-wet. If they do not agree,
Hofmes snd Tlppie model the porous medium as a series
of strsight cylindrical capillaries and determine the apparent contact angle with equations similsr to those discusacd in the previous section on displacement capiffary
pressure. Because of the number of approximations, this
Reservoir

Journalof PetroleumTechnolom..November1986

aPPafent cOntactangle wfi provide only a rough estimate

of tfre actual reservoir wettabdity.
hleaaurement of Fractional and
Mixed Nettability
Infractional wetted cores, a p@ion of the rock is strongly
water-wet, whtie the rest is strongly oif-wet. The term
mixed wettabifhy was introduced by SalatMeI lol to
refer to a special type of fractional wettablli~ in which
the oil-wet surfaces form continuous paths through the
farger pores. Additional information can be found in
Ref. 1.

found that aaphaftene adsorption in sandstone cores coufd

afso increase the relaxation time. In one set of experiments, clean smrdstone plugs were saturated with cm&
oil, then aged for several days. The bulk of the uil was
removed by flushing with cyclohexane, leaving bebind
a Iilm of asphaltenes on the ruck surfaces. The plugs were
saturated with water, and the relaxation time rneasurcd.
The adaorbed film increased the relaxation trme when
compared with the time for clean plugs. In another experiment, a plug was saturated with water and crude,
aged, flushed with cyclohexmre, then saturated with water.
The nuclear magnetic relaxation curve for this sample had
three components: (1) a fast component for water in the
small pores, (2) an intermediate component for water in
the large pures, mrd (3) a slow compmrent for water in
tire large pores that had been fdled with oil and coated
with asphaltenes. However; Devercaux dld not suggest
my way to use this to measure the wettabfity.
Brown and Fatt 102also proposed a nuclear magnetic
relaxation method to measure the wettabflity of reacrvuir
core, which app=ently has not actually been used. The
method compares the nuclear magnetic thermal relaxation rate of the untreated core with reference measurements on the same core in both strongly water-wet and
strongly oil-wet states. The core is. first flushed with
tohrene or hexane to displace all of the brine and oil. After vacuum drying, the core is saturated with distilled
water, and the therrnaf relaxation rate is measured. It is
assumed that the preparation procedure above IWSnot altered the wettahility of the core. Next, the core is made
strongly water-wet by flushing with methanol and chloroform or by ftig at 950F [51OC] tn remove all of
tbe adsorbed surface material. The nuclear magnetic thermal relaxation rate of the core in this water-wet ieference stare is mcasrrred. FAy,
the core is trcarcd with
an organocbforosikine, which renders it strongly oil-wet,
and the thermal relaxation rate is measured. The reference relaxation rates for the core when it ia strongly waterwet and oil-wet are plotted vs. the percent of oil-wet surface, and a straight line is drawn between them. Assuming a linear relationship between fractional wettabdity and
relaxation rate, the fictional wettabiity of Orenative-state
untreated core is then found by plotting ita relaxation rate
on this straight line.
Unfornrrrately, this proposed procedure suffers from
several problems. First, the functional relationship between relaxation rate and fractional nettability is not clear.
Brown and Fatt found a linear relationship between fractional wettabilhy and reaction rate, whale Kumar et al.
found a linear relationship using reaction time (the inverse
of reaction rate). Second, aa discussed in Anderson, 1 the
methnda that Bruwn and Fatt suggest to prepare the origimd core will generally after the mtive-state wettabflity.
Finally, it is not possible to tell whether the cleaning
method haa rendered the core totally water-wet or the organocblorosilane traamrent has rendered the core totally
uil-wet. Irr some cases, cure treated with an organocblorosikme is only neutrafly wet. I

Nuclear
Magnetic
Relaxation.
Brown and Fatr 102and
uthers 103,lW proposed a nuclear magnetic resonance
f,NMR) method for determining the fraction of the core
that is oil-wet vs. water-wet irr a core with fractional nettability. The method rrses the nuclear magnetic thermal
relaxation time for water protons (hydrogen) in poruus
media. To measure the relaxation time, the sample is first
expused to a smmrgmagnetic field, which makes the nuclei
of the hydrogen atoms line up with the field. The core
is then exposed to a much weaker field. The nuclear magnetic relaxation time, which is the time it takes for the
hydrogen nuclei to adjust (relax) to the new field, is meaaured. There are two relaxation times: relaxation of the
component parallel m the field is called thermal relaxation, and relaxation of the component pe errdicrdar to
the field is cabl transverse relaxation. T 05 The thermaf relaxation time is the time used to measure fractiOnSI wettabifity.
For thermal relaxation to occur after the magnetic field
is changed, Oreprutuns rnuat dissipate some of their energy
to random thermal motion of the molecufes. The protons
are only lurwely coupled to their environment, so they require a time on the order of seconds to adjust to the new
magnetic field, which is a verj long time for atomic
proceasca.
The use of nuclear magnetic relaxation times to measure wettabflity ia based on the observation that the surfaces of the porous media can significantly reduce the
measured relaxation time. 102.105When a proton is near
a surface, it cmr bccume temporarily bound to the surfiace,relaxing much faster than in tie bulk fluid. The wettabili
of the surface can influence the relaxation
dine. ? ~-1135 oil-wet surfaces cause ~ SmWer ~~ction
in relaxation time than water-wet surfaces.
Brown and Fatt 102 exmnincd 100% water-saturated
sand packs in which a fraction of the aand grains were
water-wet and the remainder bad been treated with an orgarmchforosilarre tu render them oil-wet. They found a
finear rclatiun between the rafaxation rate and the fraction of oil-wet surface area. f,The relaxation rate is the
inverse of the relaxation time.) The greater the fraction
of oil-wet grairra, the longer the relaxation time, and the
slower the relaxation rate. Krmrar et al. lM measured
relimtion times with 100% water-saturated bead packa
composed of water-wet glass beads and non-water-wet
polymetlryhnethacrylate beads. The relaxation time increased heady as the fmctiun of non-water-wet beads
Dye Adaurption. Holbruok and Bernard lM used the adincreased.
sorption of metbylene blue from an aqueous solution inBrown and Fattlm and Knmar et al. lM applied their
jected into a cure to measure fractional nettability. This
methud only to sandpack and beadpacka. Deveraaux los
method successfully mcaaured the wettabtity of fraction-

Journalof petroleum Technology,November1986

1257

ally wetted sandpacka containing mixtures of oil-wet and

water-wet sands. However, the method will probably not
work for reservoir cores mntaining krrgearnounKof clay.
Iz3 this merhod, water-covered rock surfaces are
assumed to be water-wet, whfle the oil-covered ones are
assumed to be oil-wet. The technique is based on the observation that a rock surface covered with water will adsorb a lsrge nmount of methylene blue, whereas one
covered with oil will not. The dye adsorptionof the test
core is measured at ROS, where essentially all of tie wet:
ting phase is continuous. 107-110 This enables the dye to
adsorb on essentially W of.the water-covered,
water-wet
surfac~.
A reference
dye adsorption
measurement
is
made on an adjacent core plug that is clcancd to render
it totally water-wet. The cleaned reference plug is sao+
rated with brine, so the entire rock surface is watercovered. The fractional wettabM~ is then established by
dividing the dye adsorption of the test core by that of the
100% water-wet reference core. When tl@ method wss
tested on fractionally wetted sandpacka containing mixtures of oil-wet and water-wet sands, a Iinear relationship between the fractional wettab@V and the dye
adsorption was found.
The dye adsorption test actually measures the fraction
of the total surface area of the core that is contacted by
the injected water. Because of this, both oil und water must
be present in the core when the dye adsorption is meaaured. The dye adsorption method mskes two additional
assumption: the water phase is continuous at ROS, so
the dye contacts all of the water-covered surfaces; and
the thin fk
of oiI and water coating the mck surfaces
are not affec@d by large changes in saturation. Tracer
experiments have shown that essentially elf of the water
is continuous at ROS for both water-wet snd oil-wet
~re~, I10 me ~~~uption that the thin fti
of@ ad
water are not affected by Iarge changes in

saturation seerrra

reasonable because the amount of liquid in the thin films

is very stmdl in comparison with the bulk tluida. Shankar
and DuMen111 exarrrincd how dye adsorption varied with
water saturation in Berea sandstone cores. They injected
oil and brine at currstaut rates and allowed the saturations
in the core to reach equilibrium. The injection wsa then
switched from brine to dyed brine, and the dye adsorption was measured. They found that the dye adsorption
was ahnoat conatsnt when tie water saturation wsa greater
thsn 40% PV. The dye adsorption drcreaaed at lower sanrrations, where the water stinted to lose continuity. These
measurements show that the wettabihty of a core measured by dye adsorption is not dependent on the saturation, except possibly at water saturations near IWS.
Although the fractionally wetted sandpack that were
used by Holbrook and Bernard did not contain clays, they
retilzed that clays would strongly affect dye adsorption
in reservoir cores. This occurs becanse the surface area
aud dye adsorption capacities of clays are much larger
ha those of sad g~ns, 111,112f+olbrook ~d Be~~d
stated that their test would measure the fraction

of the clay

water-wet irr a reservoir core. However,

they had problems when measuring the dye adsorption
snd fractional wettabiMy of cores containing a significant
amount of montnrorillouite. Irreversible chsnges in the
clay structnre of the water-wet reference core reaufting
from extraction and drying caused them to cuhxdate relative water wettabtitiea that were greater tbau lCO%. Note,
surface drat was

1258

however, that it maybe possible to use supercriticsl drying to avoid these problems. L13

Mixed Wettabfity. At the current time, there is no single nettability test that wifl determine whether a core has
Sslathiels lol mixed wettabiity. It appossible, however, to make this determination by examining the reardta
of (1) a glass slide nettability test, (2) a waterflood of
the native-state core, and (3) several waterflood of
restored-state coma that were aged with different brine
saturations. As discussed in the introductory paper, 1 in
a mixed-wettabdity core, the oil-wet rock surfaces form
continuous paths throughout the large pores, while the
smaller pures remain water-wet. Mined wettabilhy can
occur in a rock if the crnde forms a thick oil-wet layer
on the smface only in those places where it is in dirwt
contact. This can be tested with the glxas slide method,
with half the slide in crude and the other half iu brine.
Quartz or calcite cryatala could also be used to give a surfxce more representative of the reservoir. The core may
have mixed wettabtity if the haffof the slide aged in cru&
forms a thick, oil-wet layer, while the half aged in briue
remains water-wet. If the entire sfide remains water-wet
or becomes oil-wet, the core will probably have a uuiform wettsbdity.
The second measurement to iudicate mixed wettabiMy
is a waterflood of the native-state core. If the core has
mixed wettabilky, oil will be produced down to a very
low ROS as many PVs of water sre injected. Uniformwettabili~ cores wiR generally have a shorter duration
of production andlor a larger ROS. 6,7,114,115Finally, ,a
series of waterflood in restored-state cures cambe used
to confirm rhe mixed wearability of the core. A series of
cores is cleaned, saturated in brine, oilflonded with cmde
to dfierent brine saturations, then aged to restore its original wettabtity. Salathiel fonud that the recuve~ fmm his
restored-state nrixed-wettabdity cores had a maximum at
a pmticuka value of the brine saturation during aging.
When the water aahmmionwas lower than this value, some
of the arnidl pores becsme oil-wet, lowering recoveg.
Conversely, at larger water sahmations, the oil patiways
through the core became discontinuous.
Two other measurements that will sometimes help in
determining whether a core has mixed wettsbifity are imbibkion meaaurementa aud capillary pressure behavior. 3
Spontaneous (free) imbidti,on of both oil snd water h@
been reported for some cures with iiactional or mixed wettabiity. 49-51These corez will have positive diaplacementby-water aud displacement-by-uil ratios. Another indicakx
of mixed wettabtity is a comparison of Oil-displacingbrine capillary pressure measured on native-state plugs
va. meaauremenk on the ssnre plugs after they have been
clca.ned aud rendered water-wet. In some mixed-wet
plugs, the native-state capillary pressure curve wilf cross
over the cleaned curve m the capillary pressure ia increased. 116118Fticr discussion can be found in Ref. 3.
In summary, if the oil forms thick, oil-wet tilms only
on thoac pm-tioui of the glass afide with which it ia iu direct
contac~ if the core can be flooded down to very low oil
saturation, yet still produce small smouuta of ofi tid if
the oil recove~ from a rcatored-state core has a nraximum at a sp.%itlc brine saturation during its aging period, then the core Iiiely haa mixed wettabil@. ImbMion
Journalof PetroleumTechnology,November1986

and capillssy pressure mesanrementa can also help determine whether a core-has mixed nettability.
Conclusions
1. Three quantitative wettabdity measurements me in
use today contact angle, the Amott method, and the
USBM me~od. The contact angle measures the wettabflity of crude and brine on a polished mineral surface.
It is the best method to use when pure fluids and artificial
cores are used. It is also used to examine the effects on
nettability of experimental conditions, such m pressure,
temperature, and brine chemistry. The USBM and Amott
methods measure the average wettabil~ of core. They
are superior to the contact-angle method when the wettabflity of native- or restored-state core is measured. They
also must be used to determine whether a core has been
cleaned completely. The USBM method appsass to be superior to the Amott method, which is insensitive mar neutral wettab~lty. A modification of the USBM method,
developed by Sharnra and Wunderlich, 57 allows the calculation of both the USBM and Amott nettability indices.
2. A Iarge.number of qualitative wettabfity measurement methods are available. The imbibition method is the
most widely used because it is fast, does not require any
complicated equipment, mrd gives RJJidea of the average
wettabfi~ of the core. Tbe microscope examination
method is otlen used.in flow visualization studies. Finally, wettabilhy measurement methods based on relative
permeabdity curves are often used when these data are
available.
3. Two methods have been developed to measure the
fractional wettabtity: the NMR method and the dye adsorption method. Neither method is widely used today.
4. There is no mefhod m determine

whefher

a core has

mixed wettabfity. However, it appesra that it maybe pGssible to make such determination by examining the results of a glass slide nettability test, a waterflood of the
native-state core with many PVs of water, several waterflood of restored-state cores that were aged with different brine saturations, and imbibition and capillary pressure
measurements.
Nomenclature
,4~ = area under the oil-drive centrifigrd
capillary pressure curve, USBM method
,42 = area under the brine~tike cenfrifigal
capillary pressure curve, USBM method
ED = displacement energy (adhesion tension)
g = acceleration of gravity
ho = height of the oil column above the
Oil/water inter@ce
h. = height of the water column above the
Oil/water interface
I = Amots-Harvey relative displacement index
~ = initi~ imbibition ~k of ~ core just after it
is submerged in a fluid
fir, = initial imbibition rate of a core after it is
cleaned and rendered strongly water-wet
N = wettabfity number, Eq. 11
P, = capilla~ pressure
r = capillary tube ra~us
Journal of Petroleum Technology, November 1986

rmm = computed equivalent circular radius of the

largest pores in a core, Eq. 9
R = relative rate of imbibition, Eq. 6
VW = volume of oil displaced by sponfsneous
imbibition of water, Amott method
Vo, = toti volume of oil displaced, Amott
method

vWSp=

volume

of water

imbibition,

displaced

of oil,

Amen

by spontaneous
method

Vw = total volume of water dkplaced, Amen

method
W = USBM nettability index

60 = displacement-by-oil ratio, Amen method

8 ~ = displacement-by-water ratio, Amott method
@= contact angle
p. = oil density
P w = water density
~.=~
aA = adhesion tension
0., = interracial energy between the oil and solid
LT.. = interfaced energy between the oil and
water
f7w = intefiacid energy between the water and
solid
Subscripts
a =
adv =
a-o =
0-w =
rec =

apparent
advsncing
sir-oil
oil-water
receding

T = fhmahold

Acknow[sdgments
I am grateful to Jeff Meyers for his many helpful suggestions and comments. I also thank the management of
Cormco Inc. for permission to publish this paper.

1. Andersom W. G.: Wertability Literamre Survey-par! 1:

RccklOiUSrine Interactions, and the Effects of Core Handling on
Wetmbi[ily,,. JPT (Oct. 1986).
2. Anderson, W.O.: ,Wenability Literature Survey-ParI 3: llx Effectsof We.uabililyon tie Ekcuical Prqwties of Porous Media,%
JPT (Dec. 19S6).
3. Anderson,W.G.: WetfabilicyLiterature Survey-Pan 4 The Effects of Wmabilhy OCapillaryFtess.re,x>pap SPE 15271avaJable at SPE, Richardson, TX..
4. A&mson, A. W.: Physical C6emimy of Swfoct=s.fourlh edticm,
John Wiley and Sons Inc., New York City (1982) 332-6S;
5, Amy., J. W., Bass, D., and Whiting, R.L.: PeIroitvun Reservoir
Ef@eering: physical Propenies, McGraw Hill Beak Co. Inc.,
New York City (1960).
6. Arnon, E.: CObswvatiomRelating to the WettabiliV of Porous
Reck, Trans., AIME (1959) 216, 15&62.
7. Craig, F. F.: 2% Res.erwir Engineering Aspects of Wa@oodin~, Monograph Series, SPE, Richardson, TX (1971) 3, 12-44.
8. Cuiec, L.E. er aL: Detenninarion of the Wettabilhy of a Sm.
PI. of Reservoir Rwk,S Rev. 1.s. Frcmc. d. P&role (Sept.-Ott.
197S)33, No. 5,705-28. English fmnslatkm waihable from John
Cremr Libmrj, translation no. 81-13920-081.
9. Domldmn, E. C., Thomas, R. D:, and Lomnz, P. B., WemabiJily Determination and Its Effect on RecweIY Efliciency,z, SPEJ
(March 1969) 13-20.
1259

10. DuUien, F.A.L.: Porous Media: Fluid Tmnspon and Pow SrrucUtre, Academic Press, New York Cify (1979).
11. Hjelmekmd,O. and Torsae+er,0.: LWeOabiliW,
the Key to Proper
Lat%matov Waterfbdhg Expminurds, Intl. Energy Agency
Workshop on Enhanced 011Recovery, BarOesviUeEnergy T.%hmlosy cater (April24, 1980). CONF-S03$140, U.S. DOE Feb.
1981) 1-3A.
12. kmescu.e, E. and Maini, B.B.: A Review of Laboralov TechrnqUEX
for Measuring WeftabilifYof Petroleum Rcser.wir Recks,,,
Pefrolenm Recovery Jnst., rqmt 1983-3, CaJgary (Oct. 1983).
13. Marsden, S. S.: -Wetiabiliw-Im Measurement and Application
to Waterfkwding, J. Jap. Assoc. Pet. Tech. (Ian. 1965)30, No.
1, 1-10,
14. Marsdm, S.S,: GWettabiliiyThe Elusive Key to Wawflccding.
Pezroleum Engbwer (April 1965) 37, No. 4, 82-87.
M. E., and Domkdsoo, E. C.: Rela15. McGhee, J.W., Crder,
tive Wetting J@erties of Crude Oils in Berea Sandstone, BartIesville Energy Technolo~ Cm.ter, report BETCEG-7819, U.S.
DOE (Jan. 1979).
16, J&a, S.H., Treiber, L.E., ad Archer, D. L.: Wettab&ty Of
Reservoir R.xks and Its Evaluation, Producers Monthly (April
1968) 32, No. 4, 2-7.
17. Oonaldsm, E.C. et al,: C.Equipmentand Prxedures for Fluid
Flow and WettabiJ@ Tests of Geological Materials,, Bartksville
Enerps Technology Center, rqort DOE/BETCllC 79)5, U.S.
DOE @iS.y 1980),
18. Donaldson, E. C.: ;Oil-Water-Rock Nettability Measurement,
Pmt.. American Chemical Sac., Div. of Pefroleum Chemistry
(Wch 29-Auril
. 3. 1981). 26.. No. 1. 110-22.
19. Johnson,R.E. and Deffre, R, H.: .&Weoabilhy and Confact Angles, suofaceand Co#oidScience, E. Matijevic (cd.), WiJeYJn-

terscience, NW York Civ (1969) 2, 85-153.

20. God, R,J.: Confact Angles and the Surface Free Energy of
8oli&, Su@ce d Colkid Science, R.J. @cd and R.R. Suombxg (eds.) Plenum Press, New York Cify (1979) 11, 1-29.
21. Neumann, A.W. and @od, R. J.:<Techniquesof Measuring C.mtact Angles, Su,&ce and Cbllo)d Science, R.J, Good and R.R.
Smomtem (@.), Plenum FTe3s,New York City (1979) 11,31-91.
22.. Popiel, W.].: ln%ducdm M Colloid Soenc.; Exp&on FKSS,
Hicksvilk, N.Y. (1978).
23. Hjehneland, 0.S. and Larmmdo, L. E.: ?lxpsrinmmd fnvesdgation of Om Effects of Temperature, Pressure, and Cm& Oil
CompXition on Jnterfacial Properties, SPEFE (July 1986)
271.?!2
.-. . . .

24. M.Caffery, F.G. and Mngan, N.: .Comact A@. and fnterfacial Tension Studies of wc Eydmcakon-WmerS.iid Systems,
J. c& Pet. Tech, (July-Sept. 1970) 9, No. 3.185-96.
2s. McCaffe.IY,F. G.: .Measwemetx gf Inteifacid Tensions and Con-
mu Anglesat High T~mNie
and Pressure,J. C& Pet. Tech.
(]ldy-xt. 1972) 11, No. 3, 26-32.
26. Leach, R.O. m al.: A Laboratory and Field Study of Weuability Adjuslmerdin WaterOccding,,, JPT(Feb. 1962)205-IZ Trans..
AU4E. 22S.
27. Treiber, L. E., Archer, D. L., and Owens, W. W.: LAL+boratoIYKvaluafim of the We@dliw of Fti Od FmducrngR-oirs,
SPEI (Dec. 1972) 531-40, ,,
28. BarteO, F.E. and Niederbauser, D. O.: Fdm Forming Constimmd.sof Crude Peuoleum OiJs, Fwwkmenral Research on Occurrence and Recovery of Pefrofeum, API, New York City
(1946-W47J 57-80.
29. Dcdd, C. G.: ,The Rheologicd Behavior of FM at Crude
Petroleum-Water Lxerfaces; J. Phys, C%em.(1903) 64,344-50.
30. Reisbwg, J. and Dmcher, T. M.: %derfacird Phenomenain Cmde
Oil-Wafer Systems, Producers Monthly (Nov. 1956)21, No. 1,
43-50.
31, Morrow, N.R.: ,Physics and Thermodynamics of Capillary Action in Porous hfedia,, Jnd. Eng. Chem. (June 1970) 62, No. 6,

presenfed at fhe.1985 SPE Annwd Technical Coif&ence and Exhibition, la.s Vegas, Sept. 22-25.
,
36. Cuiec, L.E. et aL: Recomcn&tiom for the Dewminatim of
the WettaMhy of a Swimen of Reservoir Rwk, Rev. Insf.
Fmc. duP&03e fNov. -Dec. 1978) 33, No. 6,907-14. E@sh
translation available from Associated Technical Servicm; tramlation no. 33H159F.
37. Bmeau, D.F. and Ckunpin, R.L.: A Surfacfant System for the
Oif Wet Sandstone of tie North Burbank Utit, JPTiMay 1977)
501-06.
38. Tnmnham,J.C. and Clampitf, R.L.: Determination of Oti SaNration A!ier Watertlc.cdi.g in a!i Oil-Wet Reservoir-The North
Burbank Unit, Tract 97 Project,>, JPT (May 1977) 491-5CQ.
39. Cuiec, L.E.: RcckiCmde 011Jnteracdons and Weoabilify An
Aftempt to Understand Their Infemelation, paper SPE 13211
presented at the 1984 SPE AnnuaJTechnical Conference and Exhibition, Housfon, Sept. 16-19.
40. Morrow, N.R., Lim, H.T., and Ward, 1.S.: Effect of CmdeOfi-fnducedWermbiifYChanges on Oil R.ecovew,SPEFE&eb.
1986) s9-103.
41. McCaffery,F. G.: Tim Effect of Weffabilityon Relative Penn+
ability and Imbibition in Porous Media, PhD fbesi.s,U. of Calgary, Calgaly, Albafa (1973).
42. Mdmse, J.C.: Wetfab!lity as Relafed to Capihy Action in
Porous Media., SPEJ (Sept. 1965J 259-71.
43. Mmrmv, N.R. and MwIg.an,N.: WettabChy and CapiIIariWin
Porous Media, Peuolemn Recovery Research Just., Calgary,
report RR-7 (Jan. 1971).
44. Morrow. N. R.: CaDiJkuv Pressure Correlations for Uniformly
Wetted Porous fvfti~, J.CaJI.Per. Tech. (OcL-Dec. 1976) U,
No. 4, 49-69.
45. Morrow, N.R. and Mccaffew, F.G.: Displacement Studies in
UniformfyWetted Porous Media, WeIdn,$ Spreadin&andAdhesio., G.F. Pad&y (cd.), Academic FESS, New York CiV (1978)
289-319.
46. Le.veren, M. C.: .Capi21myBehavior in Porous Solds, TrCULS,
AJME (1941) 142, 152-69.
47. Hassler, G.L. and Bnumer, E.: Measurement of C.#la.ry EYessure in small Core Sam@, 9, Tmx.r., AJME (1P43J160.114-23.
48. Slotcd, R.L,, Chamters, A., and Prehn, W. L.: WI,, of Centrifuge for Determining CommteWater, Residual Oil, and Capillay Pressure Curves of .%akl core Samples, Trans., AIME
(1951) 192, 127-34.
49. Burkhardt,J.A., Ward, M.B., and McLan, RH.: <EffectOfGm&
Surfacing and Mud Fikrate Flushing on Reliability of Core&mlysis Conducted on Fresh Cores, paper SPE 1139G presented at
fhe 195S SPE Annuaf FalJ Meefing, Houston, Oct. 5-8.
50. Mohamy, K.K. and SaJter, S.J.: ,Muhiphax Fiow in Porous Meam,
OJ Mobd~ou,
Tmmvm DisWrsicm, and Weoabili-

51.

52.

53.

34.

55.

32-56,
32. Eick, J.D., Gw6, R,J., and Neumann, A.W.; Thermodynam-

ics of Confact A@K U. Rough Solid Surfaces,,. J. Colloid Inte$ace Sci fNov, 1975) 53, No. 2, 235-48.
33. WrignEX,O,R, and Leach, R. O.: Tmprcwing 0,1 Displacement
Efflciemy by WeltabfliV Adjustment, JPT(AprO 1959)65-72
Trans. , AIME, 216..
34. Cuiec, L. E.: Res.foration of the Natural State of Core .%mplm,
PSFUSPE 5634 presenfed at the 1975SPE .&mud TechnicaJConference and Exhibtion, Dallas, Sept. 28-Ott. 1.
1260

56.

57.
5S.

ty, Wer SPE 12127pwnted at the 19S3SPE ANIIUJTechnical

Conference and Exhibition, San Fnmcisco, Oct. 5-8.
Sharma, M.M. and Wundedi.h, R.W.: The Aheratio. of R.xk
Ihoperties Due fo Jmeractions wiih Drilling Fluid Components,
paper SPE 14302 presented at the 1985 SPE Annual TechnicaJ
Conference and Exhibition, Las Vegas, Sept. 22-25.
Botek, J.E., Maw, C.C., and Denekas, M. O.: ResavoirRwk
WetmbiMy-Its Signifknce and Evaluation, JPT (July 195S)
155-61Y Trans.. AIME. 223.
Dcmkas, M. O., Mmtax, C.C., am3Davis, G.T.: %ffccf of Cmde
Oil Compments on Reck Weuabiily, JPT(Nov. 1959)330-3%
Tram., AJME, 216.
Ehdich, R,, i-fasiba,HM., and Rainm@.6i,P.: <AlkalineWaferflccding for Weuebtity Alteration-Evaluating a Pofential Field
Amdicmion.
JPT Dec. 1974) 1335-!3.
. .. .
Gimnmdino;, Sh. K.: sThe Nature of the Surface of Minerals
of Oil-Beaing Rocks, Izv. Vyssh. Ucheb. .Zmedenii, Neft i GLU
(1963) 6, No. 7,3742, English uanskidon available from Assmiatcd Technical Services, fmmlation no. 23R75R.
Handy, L. L.: De&nmrmfion of Effective capilbmy FT&SUreS
for
Porous Media from imbibition Data, Trans., AJAIE(1960)219,
75-80.
KG~3.R., Naunrmn, V. O., and Manax, C.C.: :Effectof Rcsc.rvoir Emimnment on Water-fJL!DiiL?mmentz JPTfJune 1961)
57%92.
Cooke, C. E., Williams, R.E., and Ko1062ie,P. A.: oil ~Very by Make
Watmi%wding, JPT (Dec. 1974) 1365-74.
Journal of Pemolcum Technology, November 1986

59. OunaMson,E. C., andThomas, R. D.: Microsmpic Obsavatiom

of Oil Displacement.in Water-Wet and Oil-Wet Systems, paper

SPE 3555 presented at the 1971 SPE Annual Meeting, New
Orle.ms, Oct. 3-6.
.$3. Castor, T. P., Smnert.an, W, H,, and Kelly, J.F.: .$Recovery

Mechanisms of Alkaline Flocdine., Surface Phenomena in EnJkancedOil Recovery, D.O. Shah ?~.), Pienum Press, New York
City (1981) 249-91.
61. Michads, A.S. and Tinmdn.s, RS.: Chromatogiaphic TmnsFOrl of Reverse-Weting Agents ad Its Effect m OiJ Displacement in Porous Media, Trans., AJME (1960) 219, 15W37.
62, Michaels, A.S,, Smncell, A., and Por7er, M.C.: ,Effect of Chm
nmtographic Transpml i. Hexylandne on Displacement of Oil by
Water i Poro.s Media,,> SPEJ (Sept. 1964) 231-3% Tram.,
AJME, 231.

63. Micbaels, A.S. and Portez, M.C.: Water-Od DisplacementsFfom

Porous Media Using Transient Adhesion Tension Altermions.
,4JC6EJ. (hiy 1965) 11, No. 4, 617-24.
64. Morris, E.E. and Wieland, D.R.: <.A Microscopic Study of tie
Effect of Variable Weiiabilily Conditionson Immiscible Fluid D*
placmnemt,,s paper SPE 704 presemed at the 1963 SPE Annual
Fail Meedng, New Orleans, Oct. 6-9.
.65. <APIRemiunmded Practices for Laborafov Testing of Surface
Active Agenfs for Well Stimulation,,, American Pefrolcum Jnst.,
API RP 42, second edition, Dallas (Jam 1977).
66. Banell, F.E. and Osterhof, H.J.: <AdhesionTen?.iox FTess.re
of Displacement Methcd,>3J, Phys. Chem (1933) 37, No. 5,
543-52,
67. Nutting, P. G.: ChemicaJ Problems in fhe Water Driving of
Petroleum from Oil Sands,Indl E.g. Chem. (Oct. 1$05) 17, No.
10, 1033-36,
68. Rust, C.F.: A Lab.mabnyStudy of Wetfability Effects on Basic
Core Parapet.%>, paper SPE 9S6G presented.! the 1957 SPE
Venezuelan AnnuaJ Meedg, Caracas, NOV. 6-9.
69, Cefik, M..% and Somamn&mn, P.: WeOsbiMy of Resewoir
MinemJsby Flotationand CmreJadonwifh SurfacUntAdsmption,
paper SPE 9C02 presented at the 1980 SPE ML Symposium on
Oilfield and Geotherrmd Chemistry, Stanford, CA, May 28-30.
70. Cletnmtz, f3.M.: Altermicm of Reck Properdes by Adsorption
of Pefmleum Heavy Ends: Implication for Enhanced 0,1 Recove~, Paper SPE 10683presented at the 19S2SPE/DOE SymFOsmm . . Enhanced Oil Recovery, Tulsa, OK, April 4-7.
71. Lai, R.W.M. and Fuerstenau, D. W.: .,Liquid-Liquid Exuaction
of Ulmtilne Particles, Trans., ASME (1968) 241, 549-56.
72. Gaudim A.M.: Floration. second edition. McGraw HiJJBook Co.
Jnc., New York City (1957).
73, Glembotskii. V. A.. JGassen. V. I.. and Pkiksin. I.N.: Flo&uion,
Primary Sources, New York Citi (1972).
74. EbAich, R. and wy@, R,].:: Tntemelationof Cmde Oii and Rmk
Properties With the Recovery of Oil by Caustic WaterRooding,
SPFJ (Aug. 1977) 263-70.
nmh, 1.S.K. and Marsdm, S.S.: A Wett ability Scale for
75. Kam
Porous Media, 9>preprint 14D presented al fhe 1966 AJChE NatL
w..,;..
n.,,,,= n.. A..
.
76. Owens, W.W. and Archer, D.L.: The Effect of Rock W.@ility on Oil-Water Relative Petmeabilir$ Relationships, JPT (July
1971) 873-78: Trans.. ASME. 251.
77. Schneider, F.N. and Owens, WW.: Sandstone and Carbonate
Tw+ and Three-Phase Relative P.nneabiliw Chamcterisdcs,
SPEJ fblarch 1970) 75-.34 Tram., ASME, 249.
78. Batycky, J.P. er aL: Tmerpredng Relative Permeability and Wettabiiiy from Unsteady Sfa.fcDisplacement Measurements, SPEJ
(June 1981) 296-308.
79, Richardson, J.G. et af.: Laboratory Determination of ReJmive
Permeability, >Trans., AJME (1952) 19.5, 187-96.
80, tile, C.M.: Multiuhase Fkmvin Porous Media, Gulf Fublisbing
Co., Houston (.19S1)42-64.
81, Frehse, W.: Method for Determining the Sme of Wetfing of
Oil ReseIvoir Rd.% Zeitschti$@rAngewwdreGeologie (Feb.
1973) 19, No. 2, 86-88. English fmns!ation available fmm fhe
John Crerar Library, translation no. 74-10424-O8G.
82. Grigorev, S.N.: Procedure for Evaluating Relative Wetfa.bi!ify
of Resewoir Reck of Producing Depesifs, Izv. Vyssh.Ucheb.
Zawde#iL Nej?i Guz (Nov. 1980) No. 11,33-38. English translation avaib.ble from AsscciafeATechical Services, mnmlation
no. .59L169R.
83. Calhoun, l, C.: Criteria for Detaining Rock Welfabilhy, Oil
& Gas J. (May 10, 1951) 50, No. 1, 151.
..-...6.

-,

-----

.,

Journal of Petroleum TcchnolosY, November 1986

84. Gatenby, W.A. and Marsden, SS.: Some Wenabilify Characteristics of Synthetic Porous Media,- Pmduccm Monthly fNov,
1957) 22, No. 1, 5-12,
85. D.rmine. H.N. and Johamen. R,T.: Ca~iIlarimetric Methcd fm
Mca.sur&.nt
of Crude Oil Wetting l%n~ency,,, Per. Enz. (July
19S8) 30, No. 7, fi26-B27.
86, Johansen,R.T. and Dunning H, N.: Relative Wef6ng Tendencies of Crude Oil by fbe Capillmhnetric Method,,, Pmducem
Momhly (Sept. 1959}23, No. 11, 2%22.
S7. Johanse, R.T. and Dunning, H.N. :. Relative Wetting Tendencies of Crude 0,1s by Capilkirimetric Method,,> U.S. Dept. of
the Imerior, USBM, report Rf 5752 (1961).
88. Barfell, F.E. and Oscerhof, H.].: ..Determimdon of the Wettabtily of a Sofid by a Liquid Relatim of Adhesion Tensicmto Stability of Color Varnish and Lacquer Systems,>, Ind. Enz, C&m,
(1927) 19, 1277-80.
89. Bartell, F.E. and Miller, F. L.: .Degree of Wetdng of Silica by
Crude Petroleum 0,1s,, [rid. Ens. Chem. (July 1928) 20, No.
7.738-42.
90. %ner, F.C., Odd, C.O., and BwL4J,F.E.: Wkpkmement Fmssums for Petroleum Od-Water-8iJicaSystems, Oil& GasJ. (Nov.
12, 1942) 41, No. 27, 199-208.
91. Benner, F. C., Dcdd, C. G,, and Bartell, FE.: &EvailMiOn of EffecfiveDiwkmment Re.ssuresfor Petroleum Oi-Water-Silica Systems,,, F&dmneml Reteach on Occurrence and Recover3-of
Pefroleum, API, New York City (1943) 05-93.
92, SinghaJ,A.K. and Dramh.k, P. M.: Nettability Comrol of Glass
Beads, Cab. J. Chem. E.g. (Feb. 1975) 53, 3-8.
93. Slobod, R.L. and Blwn, H.A. : <Methodfor Demmining Nettability of Resewoir Rocks, Tram., AIME (1952) 195, 1-4.
94. Warren, J.E. and CaJhoun, J.C.: CASmdy of Watedlmd Efficiency in Oil-Wet Systems,,, JPT (Feb. 1955) 2-2% Trms.,
ASME, 204.
95. Kinney, P.T. and Nielsen, R.F.: <TheRole of Wet!abilhy in Oil
RecoveV; Producers Monrhly (Jan. 1950) 14, No. 3, 29-35.
96. Kinney, P.T. md Nielsen, R.F.: WettabiJi~ in Oil Recovery, >v
Worfd Oil (March 1951) 132, No. 4, 145-54,
97. Stahl, C.D. and Nielsen, R.F.: .Residual Water and Residua) Oil
By Capillary Pressure Technicpes,,, Producers Jfmubly (Jan.
1950) 14, No. 3, 19-22.
98. Kinney, P.T., Killins, C. R., and Nielsen, R. F.: Worn. ApplicaSands,,,
60., of C@@ Pressure Measurements0 P,wklli.
Bull., TedmicaJ Cofermce o Petroleum Prodmtio, Mineral
Jndu.sfdesExperiment Station, PennsylvaniaState U. (Oct. 24-26,
1951) No. 59, 52-61.
99. Graham, J.W.: Reverse-Wetd.g Logging,,, Trans., AIME
(1958) 233, 304-09.
100. Holmes. M. and TiDDie.D.B.: Comparison Between Loz and
Capillti Bessure Dafato Estimate Rsemoir Wetting, ~aper
SPE 6856 presemed at the 1977 SPE .kmual TechnicaJ Conference and Exhibifior,> Denver, Oct. 9-12.
10L Salathiel, R. A.: ,.Oil Recovery by Surface Film Drainage in
Mixed-Wetfabilhy RCC!S,9,JPT (Oct. 1973) 1216-2% Trans.,
. ASME, 25S.
102
Brow, R.J.S. and Fan, L: .Measnremems of Fmcdoml Werebility of Oilfield Rocks by tie Nuclear Magnetic Relaxation
M@A@SSTram. , A3ME (1956) 207, 262-&.
103, Devereaux, O. F.: .Effect of Cm& Oil o the NMR Relaxation
of Water Rotors in .%ndstonc,,, Nature (Au8. 5, 1967) 215,
614-15.
104. Kumar, J., Fa.tf,L, and Samf, D. N.: Nuclear Magnetic Reluxio Time of Water in a Porous Medium wifh HeterogeneousSurface Wel!nbdily, J, Appl. Physics (Sept. 1969) 40, No. 10,
4165-71.
105. Brown, R.J.S, and Ganwn, B.W.: Nuclear Magnetism Lo8Sing, Trans., ASME (1960) 219, 199-207.
106. JJolbmok, O.C, and Bernard, G, C,: Detmminatio of WeiW
biJity by Dye Adsorption, Trans., AJME (1958) 233, 261-64.
107. Brown, W.O.: The Mobiliy of Connate Water During a Watertlccd,,> JPT (July 1957) 190-95; Trans., AIME, 210,
Jones, S.C.: ,,Some Surprises in fhe Tmnsgmrtof Miscible Fluids
in he Freseceof a Second JnmdsciblePhase,,, SPEJ fFeb. 1985)
101-12.
109. P.aimondi, P., Torcaso, M., and Henderson, J.: sThe Effect of
Jnfemtitial Wafm on tie Mixing of Hydrocarbons During a M&
cible Dkplacement prccess, Mineral IA.mies Erpen.men?SIotion CircularNo. 61, PennsylvaniaSfafeU., University park (Oct.
23-25, 1961) 1-34.
1261

110 .%her, SJ. and Mohanly, K.K.: MuMphasz FIOWiu Porous Me-

111

dia 1. Macroscopic Observations and Mcdeling, paper SPE

11017 presented at fhe 1982 SPE Annual Technica3 Conference
and Exfi%ition, New Orleans, Sept. 26-29.
Shankar, P.K. and Dullien, F.A.L., .Eq$erimenti Investigation
of Twc-Liquid RelativePmneabilitj and Dye Adsapiion capaci~
VeraLSSamratiouRelatiomhips in UnUeaIedand Drifdm* Treated
%ndsfone Samples, St@c. PheIwmeIMin Enhanced Oil Recow
e?% DO. Shah (cd.), Plenum Press, New York CIIY (1981)
.<*T
. ..- , ,

112 Hang, P.T. and Brindley, G. W.: M.thylene B1.e Adsorptim

by clay &fineraSs.Detemdnation of Surface b
and CMionExchanze
Claw
. Csmcities.,
. .
. and I_%V
. Minerals [1970) 18.203-12,
113. Heaviside, J,, Langley, G. O., and Pallmt, N.: .~Pekeabtily
Charac&is6cs of Magms R-!voir Rcwk,$3Pw., Euro&an Formation Evaluation Symposium, London (iWwcb 1983).
114. Kennedy, H. T., Burja, E.O., and Boykin, R.S.: An Iuvestigatim of tie Effects of Wettabilily on the Recovery of Oil by Water
Flooding,), J. Phys. Chem. (Sept. 195S) 59, 867-69.
115. Lmenz, P.B., DmaSdwn, E.C., and Thomas, R.D.: WJS$of Centrifugal Measurements of Wettabi!iv to Predicf Oil RecovcIY,
Bartlesvilk Energy Technology Center, rep-m 7873, USBM
(1974).

1262

116. Richardson, J. G., Perkins, F. M., and Osota, 1.S.: &Differences

in Behavior of Fresh and Aged East Texas WcmdbineCore, JPT
(June 1955) 86-91; Trans., A2ME, 204.
117. Rf3d, W., Schmid, C., and Wissmam, W.: ccDisplacementTests
with Porom RcwkSamples Under Resermir Conditions,, Proc.,
.W%tiWorld PeuoleIJmCong., Frankfurt (1963) Sec. 2, papr 11,
Pi%
..,..d67-8. I.
118. Scbmid, C.: The WettabiiiV of Petroleum Rocks and tie Results
of Experiments to Study the Effects of Variations in WeuabtiV
of Core Samples, Erdoel und Kohle-Erdgas-Petroch.mfe (1964)
17, No. 8, 605-C9. English translation available from tie John
Crerar Library, Uanslation no. lT-65-12404.

S1 Metric Conversion Factors

X

degrees

F
psi

1.745329 E02 = rad

= e~
(F-32)/l.8
E+OO
= kPa
6.8947S7
JPT

receivedin tha Smlmy d Pmro[e.m Engineersoffic. Dec. 2S,

i 9s4. Papet(SPE 13933]accepmdforpubfimtionJuly23, 19s5. Revisedmanuscript
rece.w,d March3, 1986.

Ofig(nal !mnuscrlpl

Joum+of

Petroleum Technology, November 1986