I

A Laboratory and Field Study of Nettability

Adjustment in Water Flooding
R. O. LEACH
MEMBER AIME
0. R. WAGNER
AssocfA7E MEMBER AIME
H. W. WOOD
MEMBER AIME
C. F. HARPKE
JUNIOR MEMBER AlME

ABSTRACT
A field test has been made in which
additiona[ oil recovery was obtained
from a previously-water flooded oilwet sandstone
reservoir.
This recovery improvement was accomplished
by adjusting the reservoir nettability
throagh chemical treatment oj the
flood water,
The test was made in the Muddy
Jsand of the West Harrisburg Unit,
Neb. The chemical used, sodium hydroxide, was injected as a slug of dilute caustic solution through a waterinjection well, The natural nettability
of the reservoir and the citemical requirements for reversing wettabili(y
were determined
in the laboratory
from contact-angie studies using field,
water and oil santples. Laboratory
j?ood tests with a synthetic system had
shown that reversing the nettability of
an oil-wet consolidated core wou[d
Iecid to improved oi[ recovery. The
jield performance indicotes that the
mechanism by which increased oi[ recovery is obtained in the field is the
same as that observed in the laboratory:
Laboratory
studies indicate that
higher ultimate recoveries and decreaved water-injection requirements result
when the adjusting agent is added early
in the life of the flood. However, a
previously waterjlooded area was intentionally chosen for the jield test
so I,. .t unambiguous
conclusions
could be made about the eflects of
chemical injection on nettability and
the extent of oil-recovery improvement afforded by nettability reversal.
.
_
Original mawsc;ipk
received in Society of
Petroleum Engineers oflloe Aug. 1. 1961. R*
vised manuscript received Dec. 4, 1961. Paner
presented at S6th Annual Fall Meeting Of
SPE, Oct. 8-11, 1%S1,in IMIFw,
206

PAN AMERICAN PETI?OLEUM CORP.

TULSA, OKIA,

CASPER, WYO.

INTRODUCTION
Forces existing at the fluid-solid
and fluid-fluid interfaces in a porous
medium have an important effect on
oil recovery during a water flood.
Nlodifying these interracial forces in
the reservoir to improve oil recovery
has been the object of much research.
Several papers have discussed the
use of surfactants to lower the wateroil interracial tension. This paper describes work concerned with improving oil recovery by modifying the
forces at the fluid-solid interfaces
that is, by changing the preferential
nettability.
An earlier paper presented evidence
to show that some reservoir systems
could be changed from preferentially
oil-wet to preferentially water-wet by
the action of 8imp[e chemicals added
to the water to increase waterflood oil
recovery. The process was attractive
for further study because it used only
inexpensive chemicals and because it
called for a gross wettabllity rever8al
rather than for a precise adjustment.
The present paper is an extension
of the earlier work, Laboratory flow
tests were made using treated water
and oil in a consolidated core to determine the amount of additional Oil
recovery and the producing performance which might be expected from
reversing the wettabifity of an oil-wet
reservoir during a flood. A contactangle study using water and oil from
the Muddy J sand of the Harrisburg
field, Banner County, Neb,, indicated
the reservoir to be preferentially oilwet and susceptible to nettability reversal through chemical injection,
Based on these studies, a field trial
3Referencea given at end of mmer.

of wettability-reversal water flooding

was initiated in the Harrisburg field.
The primary purpose of the trial was
to determine if nettability-reversal
flooding would improve oil recovery in
an actual field situation. The field trial
was also an important test of the
laboratory technique used to determine the effect on oil recovery of
certain wettabllity manipulation,
as
well as a test of the contact-angle
method of determining reservoir nettability.
The paper is presented in two parts.
The first part covers the laboratory
experiments leading up to the field
trial, and the second part covers the
field trial.
PART 1LABORATORY

STUDY

The laboratory experiments consisted of displacement tests using an

idealized fluid system and of a contact-angle study with crude and water
from the Muddy J sand of the
West Harrisburg Unit, Banner County, Neb. The displacement tests were
designed to determine the effects of
a particular wettabllity manipulation,
oil-wet to water-wet, on oil displacement in a water flood. The contactangle measurements were made to
determine the natural nettability of
the reservoir and to determine if
beneficial wetting change8 could be
brought about by a chemical addition
to the flood water.
PROCEDURE
DISPLACEMENTTESTS
The displacement tests were performed in a consolidated sandstone
core using a refined oil and water. The
wetting properties of this system could
JOURNAL

OF

PETROLEUM

TECH

NOf. OGY

be changed at will by altering the composition of the fluids.

Fluids

The oil phase used in the displacement tests was Soltrol C which had
been treated to remove the unsaturates
and aromatics. To facilitate the X-ray
determination of saturation distribution in the core, 2.5 volume per cent
1, 3 diiodopropane was added to the
oil as an X-ray absorber. The viscosity
of the oil containing the absorber was
I:5 cp, To prepare a preferentially
oil-wet core, a small amount of noctylamine was added to the oil. The
contact angles of these fluids in equilibrium with water were determined on
quartz as a function of the waterphase pH, as previously described.:
Changing the pH from 9.o to about
1.0 causes the contact angle, read
through the water phase, to change
from about 160 to about 30. Thus,
the nettability of the laboratory test
system could be changed from strongly oil-wet to water-wet by adding dilute acid,
Core

The core was Berea (Ohio) sandtone, 1.875 in. in diameter and 18-in.
long (shown in Fig, 1). It was acidized
with 0,5 N HC1 to remove carbonates and then tired at 1,930F to reduce its sensitivity to water. On one
end a truncated cone with a minor
base diameter of 0.375 in. and a length
of Z in. was machined, The core was
then sealed in a Lucite sheath. The
PRODUCTION

---t

,;

INJECTION
~/..

FEBRUARY,

Core Salurution

The clean dry core was made preferentially oil-wet by flowing several
pore volumes of the oil-amine solutioti through it over a period of weeks
until the concentration of amine in the
effluent oil was the same as that being injected. Next, 2 PV (pore volumes ) of the amine solution containing the X-ray absorber were flowed
through the core. The core was then
scanned on an X-ray apparatus to
obtain a 100 per cent oil-saturation
reference scan for later determination
of saturation distribution in the core,
In preparing a preferentially waterwet core, the dry core was first saturated with distilled water and then
flooded with Soltrol C (containing
the X-ray absorber) to establish a
connate-water saturation. The initial saturation distribution in the core
was t~en determined by X-ray.
Displacement

Procedure

Fur three tests, the vertically mounted core was first flooded under its
initial oil-wet condition by injecting
water of pH 9 which had previously
been equilibrated with oil in the core.
Then, after some desired producing
water-oil ratio was reached, acid was
i~jected to reverse the core wettabllity
h-em oil-wet to water-wet. In one test,
made under water-wet conditions
throughout, the core was saturated
with oil containing no amine, and distilled water was used as the displacing
phase, The linear velocity of the displacing phase was 1 to 2 ft/day (calculated assuming the total porous
cross-sectional area was ava;lable to
flow). Periodically during each test,
saturation distributions were determined by the procedure described in the
following section.
To prepare the core for re-use, it
was acidized to remove the residual
amine, and then cleaned and dried
by conventional procedures,
Determination of Saturation
Distribution

f?:c.

conical end was provided with an end

section to minimize capillary-endeffect. This removable end section
was made of a 14-md-permeability,
ceramic porous material, 0,37S in, in
diameter and 1.5:in. long. To insure
capillary contact, a double layer of
Kleenex tissue was compressed between the core and end section. The
core had a permeability of approximately 600 md and a porosity of 23
per cent.

lBKan
1962

CORE ASSEMBLY.

The X-ray absorption method used

to determine fluid saturation distribu-

tions along the core is similar to that

described by Geffen and Gladfeltcr.
The run was interrupted at various
stages of the displacement tests, and
the core was scanned along its axis
by moving the core past the tube
aperture at a constant rate. The ratio
of incident to transmitted radiation
was recorded directly on a strip chart,
Contrast between water and oil saturations in the core was provided by
the 1, 3 diiodopropane in the oil. The
method used is sensitive to saturation
changes of about 0.5 per cent PV.
Oil-recovery determinations were made
by material balance, since the accuracy of the X-ray measurements
was not sufficiently good for absolute
saturation determination, Because of
apparatus difficulties, the absolute accuracy was not the same in each of
the four tests.
CONTACI.ANGLE
S1UflY
The contact angle observed for a
given system depends upon the manner in which the three phases are
brought into contact and also, upon
the time allowed for the interfaces to
reach equilibrium. A specific procedure is required to obtain valid indications of preferential wettablity.
In practice, the water-advancing contact angle is determined as a function of the age of the oil-quartz interface. Only when the water is advanced
over oil-quartz surfaces near adsorption equilibrium do the resulting con1

-.

k?+--fb

BACK PRESSUREREGULATORS

FILTER~

FLOW

PUMP

**M

S!wbu Qu
$,. ,,,,
1,3,.9,10 2,4,5,6,7,9 %%?
,!. $,. rwc.
060?,t* ,,,

2,4,6.810 1.%5.7,9

t.,.>...,, T{ !,),9

cm&79g#a&r
V*WW

2,4,5,6, u&.::.p
1.8,10

FIG. 2CONTACT-ANGLE APPARATUS.

207

tact angles indicate true preferential

wettabdity,
~. schematic diagram of the con@et-angle apparatus is shown in Fig.
2, The apparatus consists of a contactangle cell, a circulating pump, lines
and other auxiliary equipment for
bringing the phases into equilibrium
and for controlling the movement of
the water and oil. All of the metal
parts of the apparatus in contact with
the fluids are made from monel.
The cell contains two quartz-crystal
surfaces mounted rarallel to each
other on adjustable posts. The ceil
is filied with water. Then, with the
dropper tip an oii drop is placed
between the two quartz crystals so
that it contacts a large area of each
crystal. After the new oil-quartz interfaces thus formed have aged for a
whiie, the water is advanced over a
small portion of the surface w~ch
had been in contact with the oii. This
advance is caused either by pulling
the oil drop with the dropper tip or
by moving one crystal laterally with
respect to the other. The water-advancing contact angle (corresponding
to the particular age of this oii-quartz
surface ) is recorded when the interface stops moving. Then the oil-quartz
surface is aged further, the water is
advanced again and a new value is
obtained for the advancing contact
angle. Additional point~~e obtained
by repeating the proces~. When the
crude oil contains natural surfaceactive substances, the advsmcipg
angle increases as the oil-solid surface ages and tends to reach a constant value as adsorption equilibrium
is approached. Thk limiting value,
which may require several hundred
hours of oii-quartz aging time to
achieve, indicates reservoir wettabiiity.
The effect of chemically treated
water on the natural nettability then
was tested. To do this, the chemically
treated water was pumped through
the ceii without otherwise disturbing
the aged oii drop a~ady in place.
The effects of the chemical on wettabiiity were shown by the spontaneous
changes in the contact angle which
occurred.
The contact-angle
measurements
were made with wellhead samples of
water and oil from the Muddy J
sand, West Harrisburg Unit, Banner
County, Neb. The measurements were
made at a pressure of 250 to 300 psi
and at the r~ervoir temperature of
200F. The produced formation water
had a total salinity of 8,500 ppm,
predominantly
sodium bicarbonate
(2,600 ppm) and sodium chloride
(5,700 ppm). The supply, water for

the water flood had a total salkity of

less than 300 ppm.
RE23UL113AND DISCUSSION
~ISp~ACEMENT ~STS
In this study four displacement
tests were perfurrned in a single core.
In three of these tests the core initially was water-driven under oii-wet
conditions to simulate some degree of
prior water flooding and then subjected to a nettability-reversal flood.
The degree to which the oil saturation
was reduced in the initial water flood
was different in each test so that the
effectiveness of the nettability-reversal
flood would be determined over a
range of saturation conditions. For
comparison, the same core was also
water flooded under water-wet conditions throughout.
The oil-production performance of
the three nettability-reversal tests is
shown in Fig. 3, The time at which
the reversal agent was injected is indicated by the arrows, In one test,
injection of the. reversal agent was
begun when the producing water-oil
ratio reached 3:1 in the oil-wet water
flood. In the other tests, the WOR
reached 25:1 and 100:1, respectively,
before acid was injected to reverse
nettability. The highest oil recovery
with a minimum of total injection was
obtained when the reve~al agent was
injected early, that is; when there
was the least amount of water flooding
under the initial oil-wet conditions.
From the trend of the data in Fig. 3,
it is inferred that even less total injection would be required to produce
the oil if the nettability reversal were
initiated at the very outset, without
any prior waterflood depletion under
oii-wet conditions.
The producing WOR performance
curves for the threelaboratory wettability-reversal tests me shown in Fig.
4. Upon injection of the reversing
agent, the producing water-oil ratio
immediately decreased below the value

TEST I
*~r

x
TEST 3
TEST 2
.. .. ;.. ..-7. . .-.. _.__,

~ .7T

which had been reached under oilwet conditions.

Figs, 5, 6, 7 and 8 show saturation
changes during the laboratory core
tests. The shapes of the saturation prodies in the oil-wet water floods (Figs.
5, 6 and 7) confirm that the core
was initially oil-wet; i.e., as is typical
of oil-wet water floods, the oil saturation is continuously changing over the
entire length of the core during the
flood. In contrast, the oil behind the
flood front in the water-wet flood
(Fig. S) was reduced to its residual
saturation as the front passed, The.
saturation distribution curves for the
oil-wet floods show no evidence of
capillary end effect and, thus, indicate
that the core design was successful
in eliminating end effect.
When a nettability-reversal flood
is initiated foliowing a conventional
water-drive in core tests, the oil-production rate increases immediately (see
Figs. 3 and 4). The progress of the
reversing agent through the core is
clearly shown in Elgs. 5 and 6 by
the abrupt saturation changes which
occur at the front. No detectable increase in oil saturation occurred downstream of the front in any of the
tests. Instead, the oil saturation existing at any position downstream of the
reversal front remained constant until
the front arrived at that point. This
characteristic performance is significant in that it is contrary to what
would be expected from the usual
application of relative permeability
concepts.
As the extra oil is released by the
invading reversing phase, an equivalent amount of oil is produced simul-

Ix
.~

iEGENLi
TEST NO, l
TEST NO. 2 -TEST NO. 3
,.,

,,
i

INJECTKJN- PORE VOLUMES

S-9

G.
_,...
, AC~~NJECTEO
..: -----J ..+.

Yi
.:.l..._:.

;!.

-1

. .

!~(v~-~---j

<

0
=Io
, ~~
...*..
u
__._...___
.
g [.:1
.-4 . 4--4-

ii

1,

20

_:._--r-. . 3---_$.
,
., ~-.. ,_..
. .$ --+. J

-\

l-!--

!!

-I+_+j---f

,.

+--l
--

!,

30
40
50
60
TO
80
OIL R.ECOVERYIN PERCENT PORE VOLUME

I
90

FIG. 3-NETTABILITY.REVERSAL pERFOE?MFIG. 4-PRODUCING WATER.OIL RATIO

ANCEIN BEUEACoaz. WOR - WATER-OIL
PERFORMANCE
IN CORETESTS.O.W = OIL.
RATIO, W-R = jVE::~LITY-REVERSAL
WET WATER FLOOD.
.

we

.,

lNf~?ED7+

ZO~

JOUIZNAL

0S

,.

PETROLEUM.

TEC33NOLOGY

A water-drive performed in the

Berea core under water-wet conditions
resulted in about the same residual
oil saturation as did the early injection of a reversal agent, The residual
oil saturation was 25 per cent. The
fluid saturations during the water-wet
flood are shown in Fig. 8. Note the
similarity in the movement of the

taneously; therefore, an irnmed~ate increase in oil-production rate occurs.

The possible reasons for the absence
of an oil bank build-up have not yet
been investigated thoroughly. HOWever, the roles of capillary pressure
changes and hysteresis ~in the relative
permeability characteristics are being
considered.
~

OIL
OIL

SATURATION
SATURATION

AFTER
OIL-WET
wATERFLOOD
DURING wETTABILITY-REVERSAL

90

~g

70

;>
~g

60

+2
0

50

ok
p~

40 -

~;

30

ga.
z
K
&

..

CONTACT-ANGLESTUDY

Natural Nettability,
Harrisburg Field, Neb.

,
la2
d

3!

==%

_4&

&

20
10
o

I
OUTLET

INLET
o

flood front in this figure and the flood

front. in the nettability-reversal floods
(Figs, 5 and 6), Similarly, the injection requirements to flood-out and the
residual oil saturation were approximately the same in the water-wet flood
as in the early initiated tiettabilityreversal flood. Thus, it is concluded
that the residual oil saturation and
injection requirements in a wettabilityreversal flood initiated early are potentially as low as in a water-wet water
flood in the same porous media.

.2

.5

OISTANCE

FROM

CORE

INLET-

.8

,7

.6
FRACTIONAL

cORE

.9

I .0

LENGTH

FIG, 5SATURATION DISTRSBLITIONIN BEREA CORE DUBINGNETTABILITY REVERSAL(TEST

No. 1). (NUMBERSABOVE LINES INDICATETHE ORDERIN WHICH SATURATIONPROFILES
WERE DETERMINE.)
M
-A

OIL
OIL

SATURATION
SATURATION

OURING
OURING

OIL-WET
wATERFLOOD
NETTABILITYREvERSAL

Fig. 9 shows the water-advancing

contact angle on quartz vs the age of
the oil-quartz interface for water and
oil from the Muddy J sand, Harrisburg field, Neb, This figure shows that
a high contact angle is obtained when
the water is advanced over quartz
which has been in contact with the
oil for about 300 hours. The angle is
indicated to be in the range of 135
to 170, measured through the water
phase. Assuming the contact:angle surface to be a sufRciently good approximation of the reservoir surface, it
follows from these measurements that
the Muddy J reservoir in the Harrisburg field is oil-wet.
Pendant drop measurements showed
the interracial tension to be 25 dynes/
cm under the natural conditions.
Nettability Changes Caused
by Chemical Additives

OISTANGE FROM CORE INLET -

FRACTIONAL

CORE

Adding HCI to reduce the water

pi-I to 1.2 did not change the natural
oil-wet condition.
Fig. 10 shows the spontaneous
change in contact angle caused by
adding sodium hydroxide to simulated
formation water (5,500 ppm NaCl ).
AS the concentration uf sodium hydroxide was increased, the contact
angIe decreased, A c6ntact angle less
than 90 was produced at a NaOH
concentration as low as 0.1 molar.
Increasing the NaOH concentration to
0.4 molar produced a contact angle of
about 70C. To simulate the effects on
the wetting of adding caustic )0 the
fresh injection water used in the field,
0.4-nloiar NaOH was pumped through
the cell to replace the alkaline brine
solution, without otherwise disturbing
the oil drop still in place on the crystal.
This caused the contact angle to decrease spontaneously to below 40. ,
Assuming the quartz-crystal contactangle surface to be a sufficiently good
approximation of the reservoir surface,
it is apparent that a~ding NaOH to
the fresh injection water would pro-,

LENGTH

FIG. GSATURATION
DISTRXBUTIO~
IN B~REACOREDURINCNETTABILITY REVERSAL
(TEST No. Z).
~
ck
~
ai

OIL
OIL

SATURATION
SATURATION

OURING
OURING,

OIL- WET wATERFLOOD

WETTASILITY -REVERSAL

70
60

;:50
m
~ti40
~g
a
30
~1+2
S%20
SE
=&lo
%
Ico
k.
OISTANCE
FIG.

7SATURATION

FEBRUARY.

FROM

CORE

INLET-

FRACTIONAL

CORE

LENGTH

DXSTRIBUTXON
IN BEREACOREDURINGWETTABXLITYREVERSAL
(TEST No. 3).
,

209

1962

,.

00

I
.
INITIAL
~~*
..

70
\

60

~
--.

-,

PROCEDURE

I
OIL SATURATIO14

2
--.

50

.,

-~-+--.+
3

40

4i

-.

+.

/
30

FLOOD-OUT
OIL SATURATION
. __
i

20

10-

0 INLET

I
.1

I
.2

!
.3

OISTANCE

FROM

1,
t

I
.5

,4
CORE

INLET -

I
,7

6
FRACTIONAL

CORE

~OuTLET

I
.8

I .0

LENGTH

k%. 8-SATUIIArtO~ D[5TR[BUTIONIN BERSACOREDURUW A WATER-WETWATER FLOOD

(TEST No. 4).
cluce
a preferentially water-wet condition in the reservoir. Thus, a nettability reversal could be accomplished
in this field.
Pendant drop measurements indicated the interracial tension to be 9
dynes/cm between the crude and the
0.4-molar NaOH solution (containing
no NaCl ).

PART IIFIELD

TRIAL

A major consideration in designing

the field experiment was that it should
provide a definite answer about the
effects of chemical injection, Many
implicit and explicit assumptions had
been made in the research leading to
the field trial, and several unusual
properties of the experimental systems
had been observed, Therefore it was
considered essential that the field experiment should unambiguously confirm or deny the laboratory findings
that the Harrisburg Muddy J
reservoir is oil-wet and that injecting
sodium hydroxide would make the

reservoir water-wet and increase oil

recovery. Such an unambiguous answer would be forthcoming. if oil
could be made to flow again from
flooded-out oil welk which were producing only water,
Fig. 11 is a map of that part of
the field chosen for the test. The wells
in this area are located approximately
on 40-acre spacing. The field had
been under flood for about a year
prior to the present test, Some wells
had been completely flooded-out and
shut-in. On the basis of the laboratory tests, it was predicted that injesting caustic would cause oil production to start again from these wells,
The J sand is located at a depth
of 5,900 ft, and the average pay
thickness is approximately 10 ft; average porosity and permeability are 15
per cent and 119 md, respectively.

:RUN 2; QUARTZ CRYSTAL 7

.

a RuN 1,
RUN 1,
G RUN 2,
RUN 2,
g

QUARTZ
00ARTZ
QUARTZ
OUARTZ

CRYSTAL
CRYSTAL
CRySTAL
CRYSTAL

7.
T
2
7

.-

---- ---

+- -L-.

. . ..j..++_

1:
~5;oo;p;M

170

!+
1%
n 150

+--J.
I

LEGEND

;_l__

,,
.._{+.. _:

The injection wells in the test area

are Unit Wells 8 and 13. It was decided to inject a caustic slug followed
by untreated water into Well 13, while
injecting only untreated water into
Well 8. Balanced injection rates would
be used, Constant withdrawal rates
would be maintained at the offsetting
wells, Wells offsetting Well 13 were
expected to respond to the caustic,
but the wells offsetting Well 8 were
not expected to show increased production. The dashed diagOnd line
in Fig. 11 roughly separates the test
and control areas. The line is drawn
in a manner consistent with a permeability trend in the field and consistent
with the performance observed in the
test.
Wells which had been shut-in were
placed on production in Feb., 1960,
so that a production trend prior to
chemical injection could be established. A 40,000-bbl slug of 2 per cent
NaOH (0.5 molar) was injected into
WA1 13 beginning June 12, 1960.
After completing the caustic injection
on Aug. 6, 1960, injection of tmtreated flood water was again employed. Balanced injection rates of
approximately 800 to 1,000 BWPD
were maintained in Wells 8 and 13.
This bzdanced injection was maintained until Jan., 1961, when the injection into Well 13 was increased to
1,600 BWPD and the injection into
Well 8 was decreased to 400 BWPD.
At the same time, Wells 7 and 18
were shut in.
The pilot test was conducted at
minimum additional expense (excluding the cost of caustic) by using the
following procedure and equipment.
Sodium hydroxide was delivered to
the field by tank truck in a 50 per
cent aqueous solution and stored in
two 210-bbl welded-steel tanks equipped with heating coils. A 40,000-Btu/
hr steam generator was used to maintain the caustic solution at 80F, approximately 22F above the freezing

--i--

~o .
____-_~..-.--+...

~
g

__.

!1;

-., . .. .. .. .
,.

~
..

.
m

j
a

70

g
>

W,.

100

1000

MqASUREMENfS
ON

FOR HMUUSBURG,WATER AND OIL

,.

210

QUARTZ, pH = 8.5,

~~

:oRqQ-

.11

FRESH WATER
p-i----;,
I

i%
,12>3,

OAUSTIC%
~.-_-<l$_
@16
,7

,,
i 14
:

;->j<
19

15
20

,1
.2
.3
.4
CONCENTRATIONOF SOMUM HYDROXID
MOLES/LITER)

AGE OF OIL- QUARTZ INTERFACE (HOURS)

FIG. 9CONTACT.ANGLE

e4

Fm

Ftc. 1IPILOT AREA, HARRISBURG

IO-CHANGE IN CONTACT ANGLEwmi

ADDED CAUSTIC, HARRISBURGFIELD,Nm.

FIELII, NEIL

JOURNALOF PETROLEUMTE~~l$lO1.OC~
,

&-

point, Protective clothing was worn

by field personnel handling the chemical. An electrically-driven fluid proportioning pump was used to combine
the 50 per cent caustic solution and
the fresh injection water to a tlnal
2 per cent caustic concentration, immediately downstream from the waterinjection pump.
RESULTS
I,RopAGATION OF CAUSTIC
THROUGHTHE RESERVOIR
Caustic first appeared at Well 18
on Aug. 7, 1960, the day after caustic
injection into Well 13 was completed
(40,000 bbl), Chemical analysis of
the produced water indicated that the
injected NaOH had reacted with the
reservoir rock to produce a highly
caustic siiicate. The maximum pH
observed at Well 18 was 12.5, while
the maximum total alkalinity
was
equivalent to about 1 per cent NaOH.
The early appearance of caustic at
Well 18 indicated that the reservoir
was not being invaded uniformly. (In
the, initial water ffood of the fieId,
Well 18 flooded-out three weeks after
water injection started.) Caustic appeared at Well 12 in the week of Nov.
26, 1960, after the cumulative injection of 135,000 bbl (of water and
caustic) into Well 13.
RESPONSE OF WELLS
TO CHEkIICALINJECTION
Well 18 had been producing only
clear water prior to the appearance
of caustic. Afterwards, a small amount
of oil was produced, not exceeding
about 100 bbl. However, even this
smaI1 amount of oiI production was
considered to be an encouraging indication that the injected caustic was
affecting the formation in the manner
expected,
Test data for Wells 7 and 12 are
shown in Fig. 12. Prior to the injection of caustic, both wells were producing 100 per cent water. Soon after
NaOH injection, Well 12 started ~roducing oil at a low but gradually increasing rate. Eventually, a 14 per
cent level of oil production was reached. The total oil produced from Well
12 by nettability reversal was estimated from the test data to be 1,700
bbl. This well was shut in on April 8,
1961. By comparison, Well 7 produced only water until it was shut in
in Jan.,. 1961. This was as expected.
The reappearance and timing of oil
production at Well 12 is evidenu that
the injected caustic made the oil-wet
reservoir water-wet, as expected from
the laboratory experiments. The laboratory experiments also suggested that
FEBRUARY,

1962

NoOH INJECTION :--:

STARTEO 1

1960.

1961

- :
r-

FJSA12-TEsT DATA,UNIT WELLS7 AND 12.

the absolute amount of additional oil

obtainable from a very-highly-flooded
portion of the reservoir wouId be comparatively small. This also was observed to be the result in the field.
Greater oil recoveries were obtained from Well 19, which had not been
so completely flooded-out as Wells 7
and 12. Fig. 13 shows the test data
for Well 19, obtained before and after
caustic was injected into Well 13. An
estimate of what the oil production
would have been without chemical
injection is indicated by the dashed
line. This figure indicates also that
nettability-reversal
oil production
started early at a low rate which
gradually increased with time. Through
June 17, 1961, approximately, 5,700
bbl of nettability-reversal oil had been
produced from Well 19 without the
appearance of caustic. The rise in
water production, which began in
Jan,, 1961, is believed to be the result of increasing the Well 13 injections at that time, Very high water
productivity WSSalso observed at WeIls
24 and 25 beginning at that time.
The only. other well in which response to the injecied caustic could be
detected was Well 17. This response
was small, probably no more than
1,200 bbl of total oil by nettability
reversal. The movement of caustic
towards Well 17 probably was slowed
because of injections into Well 16
and because of the permeability trends
in the. field. As of June 17, 1961, the
total oil production attributed to nettability reversal was 8,700 bbl.
A more detailed comparison between the laboratory and field performance can be made. This comparison rests on several simplifying assumptions which may or may got be
warranted, but the resuks are thought
to be of sutilcient interest to justify
their presentation, As had been observed in the laboratory testa, ,increased &l production in the field
started very soon after wettabdity reversal was begun (Fig. 12). This sug-

Fit. 13TEsT DATA, UNIT WELL 19.

gests that in the field there was no

significant accumulation of the re- leased oil downstream from the caustic
water (no local increases in oil saturation). If so, then the rate of wettability-reversal oil production at a
given time will depend directly upon
the rate at which oil is being released
at that time by the caustic bank moving through the reservoir.
By comparing the slopes of Profiles
8 and 15 in the laboratory reversal
floods after prolonged oil-wet flooding
(Fig. 6), it is observed that the change
in oil saturation at the reversing front
increased approximately as the first
power of the distance the front moved
along the core, Thus, the rate of wettahility-reversal oil production increased
as the first power of the distance
traveled by the reversing front (until
,the front reached the end of the core).
In the field, it may be supposed that
the caustic moved out from Well 13
in an approximately radial fashion,
such that the average distance from
the injection well to the reversing front
can be considered to be proportional
to the square root of the volume of
water injected behind the leading
caustic edge.
It further may be supposed that,
before the test started, a saturation
gradient existed in the invaded area
such that (as in the laboratory tests)
the oil available for release was negligible at the injection well but increased
as the caustic bank moved out into the,
reservoir. As an additional approximation, suppose that the oil available
for release increases as the nth power
of the distance from the injection well
to the reversed front.
Based on the foregoing, the following equation relates the cumulative
volume of oil v. released and moved
in the pilot area (but not necessarily
captured) to the cumulative volume
of water W injected behind the leading edge of the caustic,

,*

Cw,w . . , , , (1)
v.
The eortstant c in J3q. 1 is a function
of porosity, reservoir thickness, saturation and conformance. If the amount
of fluids produced by a well is some
constant fraction of the fluids being
moved in the reservoir, then according to the above equation a log-log
plot of cumulative nettability-reversal
oil vs cumulative water injection would
be a straight line having a slope of
(r2/21- 1).

nettability adjustment can improve oil

reeovery from some reservoirs. A wettabllity-reversal flood now is being
planned for a portion of the Harrisburg field not previously flooded. The
results are expeeted to be more favorable than in the pilot because the
chemical witl be injected with the
first water into the reservoir.

Z!2

Fig, 14 is such a plot for the wettability-reversal

oil production from
Wells 12 and 19. This plot includes
data obtained only through Jan. 14,
1961, since the injection pattern was
changed after that time. A straight
line is obtained for each weU, as predicted by Eq, 1. This, then, is evidence
that the manner of oil release and oif
movement in the field experiment is
similar to that observed in the laboratory studies. Further, since the slope
of each line is approximately two,
the exponent ns 2.
Based on the general agreement between the field and laboratory results,
it is concluded that the injection of
NaOH into the J sand at Harrisburg
field reversed the wettabiiity and iJUproved oil reeovery from that po~ion
of the reservoir invaded. This con6rrns
the laboratory contact-angle measurements and flooding tests.
Application

The performance of the pilot area

has confirmed the Laboratory studies
kssofar as a comparison can be made,
that is, the response of a highly, flooded
area to a wettabdky change. Accordhtgly, the laboratory tests can be qsed
to indicate
recovery
improvement

CONCLUSIONS
s

CUMULATIVE WATER INJECTION INTO WELL 13

(EELS. )

FIG. l%RSLATIONSHW BETWEENWETTA.

BILtTY.REVERSAL OIL PRODUCTION AND

WATEaINJECTION,
HARRISBURG
FIELD,NEB.
under more favorable conditions, For
example, Figs. 3 and 4 indicate that,
through early nettability reversal, oil
recovery potentially could be as much
as 15 per cent PV more than would

be obtained by flooding under oil-wet

conditions to a producing water-oil
ratio of 25. The actual amount of improvement obtained would depend
upon the characteristics of the particular porous medium under consideration. In addition to increasing recovery, chemical injections to reverse
reservoir wettabiiity would decrease
the total water injection required to
reach abandonment conditions.
,
Also, in a three-dimensional system,
a better sweep pattern would be expected from reversing the nettability
than in flooding under oil-wet conditions. ~TIds would result because the
water-wet conditions produced would
lead to a more favorable relative permeability relationship and, hence, to
a more favorable mobiiity ratio,
Because of the favorable laboratory
and field results, it is concluded that

Based on the agreement between

the laboratory and field results presented in this paper, the following
conchtsions can be mrtde.
1, There are oil-wet reservoirs.
2. Reservoir nettability can be determined in the laboratory by contactangle measurements.
3. Additional oil can be recovered
from oil-wet reservoirs by adding a
wettability-reversal chemical to the
flood water.
4. If the nettability of an oiLwet
reservoir is reversed early in the flood,
the water-injection requirements and
the residual oil saturation are poten.
tially as low as in water flooding an
equivalent reservoir under water-wet
conditions.

REFERENCES
1, Slobod, R, L.: A Review of Methods
Used to Increase Oil Recovery, Prod.
Monthly (Feb., 1958) 22.
2, Taber, J. J.: The Injection of Detergent Slugs in Water Floods, Traas.,
AIME (1958) 213, 186.
3, Wa@er, O. R. and Leach, R, 0.: Improving Oil Displacement Efficiency by
Nettability
Adjustment,
Trans., AIME
(1959) 216, 65.
4. Ggffen, T. M. ari~ Gladfeher, R. E.:
X-ray Absorption Method of Determining Ffuid Saturation in Cores, Trans.,
AIME ( 1952) 195, 322.

,,

212

JOS.IRNAL OF PETSSOLEUM fTECklNOLOGY