SPE 113976

LoSalTM Enhanced Oil Recovery: Evidence of Enhanced Oil Recovery at the

Reservoir Scale
A. Lager*, BP, EPTG, Pushing Reservoir Limits, Sunbury, UK; K. J. Webb, BP, EPTG, Pushing Reservoir Limits,
Sunbury, UK; I. R. Collins, BP, EPTG, Pushing Reservoir Limits, Sunbury, UK; D. M. Richmond, BP, BP
Exploration (Alaska), Anchorage, AK, USA

Copyright 2008, Society of Petroleum Engineers

This paper was prepared for presentation at the 2008 SPE/DOE Improved Oil Recovery Symposium held in Tulsa, Oklahoma, U.S.A., 19–23 April 2008.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
For over 10 years research has been carried out on the impact of low salinity waterflooding on oil recovery. Data derived from
corefloods, single well tests, and log-inject-log tests have shown that injecting low salinity water into an oil reservoir should
result in a substantial increase in oil recovery in many cases. The results varied from 2 to 40% increases in waterflood
efficiency depending upon the reservoir and composition of the brine.

In 2005, a hydraulic unit was converted to inject low salinity brine into an Alaskan reservoir, by switching a single injection
pad to low salinity water from high salinity produced water. An injector well and 2 close production wells were selected within
a reasonably well constrained area. A surveillance programme was devised which included capturing produced water samples
at regular intervals for ion analysis and the capturing of production data.

Detailed analysis of the production data, and the chemical composition of the produced water, demonstrated an increase in oil
production and provided direct field evidence of the effectiveness of LoSal™ at inter-well scales. Additionally, the response of
the reservoir to low salinity water injection was confirmed by single well chemical tracer test.

In parallel, laboratory studies have led to mechanistic understanding of LoSal™ in terms of multiple-component ionic
exchange (MIE) between adsorbed crude oil components, cations in the insitu brine and clay mineral surfaces. The results
clearly show that the enhanced oil production and associated water chemistry response was consistent with the MIE
mechanism proposed.

The oil production data have been modeled using an in-house developed modification to Landmark’s VIPTM reservoir
simulation package. An excellent match for the timing of the oil response was obtained which provides a good basis for
predicting the result for large scale application of LoSal™ flooding.

Introduction
It has been more than 10 years since Yildiz and Morrow (1996) pushed forward the research started by Jadhunandan (1990;
1991; 1995) and published their paper on the influence of brine composition on oil recovery. This paper showed that changes
in injection brine composition can improve recovery. Since then, Tang & Morrow (1999) have progressed the research on the
impact of brine salinity on oil recovery, followed by other researchers such as Webb et al. (2004) and McGuire et al. (2005);
these authors performed an extensive research programme on low salinity injection (LoSal™). This programme included
numerous core flood experiments performed at ambient and reservoir conditions (at high temperature and pressure, with ‘live’
fluids) both in secondary and tertiary mode, single well tracer tests (SWCTT) and log inject log tests, showed a significant
increase in oil recovery due to low salinity brine injection.

Recently, it has been shown that multiple-component ionic exchange (MIE) between clay mineral surfaces and the injected
brine is responsible for the improvement in oil recovery arising from the injection of low salinity water (Lager et al., 2006).

*
Corresponding author, arnaud.lager@bp.com
2 SPE 113976

Adsoprtion of crude-oil components onto reservoir rock is generally understood as the mechanism that renders parts of
reservoir rock oil-wet. An important way in which this occurs is through multivalent cations at a clay surface bonding with
polar compounds present in the oil phase (resins and asphaltenes) and by the direct adsorbtion of organic polar componds onto
clay mineral surfaces by displacement of the most mobile cations present at the clay surface.

At low salinity, polar organic compounds are thought to be desorbed and replaced with multivalent cations. Data presented by
Lager et al (2006) demonstrated that when multivalent cations were present in the connate brine, flooding with low salinity
water lead to higher oil recovery. The residual oil saturation was roughly the same if the connate water was free of divalent
ions and flooded with high salinity brine or if low salinity brine was used to displace the connate brine in a tertiary low salinity
flood. The expansion of the electric double layer due to low salinity associated with MIE enables desorption of polar
compounds from the clay surface. This results in an increase in oil recovery because the bonds holding oil in contact with the
rock are broken.

Figure 1 illustrates how polar molecules from the oil are attracted to the negatively charged clay surface. Some mechanisms
involve divalent cations (calcium and magnesium) that act as bridges between the negatively charged molecules in the oil and
the negatively charged clay surface. When divalent ions, from the invading low salinity brine, exchange with either cationic
organic complexes or with bases due to the change in ion exchange equilibra, bound oil becomes mobile and increase in oil
recovery ensues.

It was still uncertain however how this mechanmism would scale to interwell distance.

In 2005 it was decided to convert a single hydraulic unit in an Alaskan oil field from brackish produced water injection to low
salinity water injection (Table 1). From mechanistic studies and the results of the SWCTT described below, it was known that
the low salinity brine used was a suboptimal LoSalTM brine, but due to operational constraints it was the only low salinity
brine available at the time the pad was converted. However, a significant increase in oil recovery was still expected even
though it would be suppressed compared to the response obtained with a fully optimized low salinity brine. The expected
response was thought to be significant enough to remove any uncertainty about MIE and how it would scale at inter-well
distance.

The targeted hydraulic unit consisted of an injector well and 2 close production wells within a reasonably well constrained area
(see figure 2).

Low salinity Brackish

brine injection brine
(ppm) (ppm)
Na 817 4,980
Ca 101 71
B 2 22
K 8 38
Al < 0.15 0.16
Mg 19 33
Ba 3.45 13.5
Sr 1.6 4.7
Salinity (TDS) 2,600 16,640
Table 1, Brine composition.

The production wells (MPL-11 and MPL-07) were closely monitored for oil rate and water cut using a portable test separator,
complemented by shake out samples obtained on a weekly basis. The produced water was analysed weekly to monitor any
change in the brine chemistry.

It can be seen from Figure 2 that well MPL-11 is closer to the low salinity water injector than MPL-07and it was this well that
SPE 113976 3

was expected to respond first to the low salinity water. However, little response has been seen at MPL-11, to date. The
reasons for this are discussed below.

Results:
Production Data:

The hydraulic unit that was selected for the low salinity water injection had been subjected to several different enhanced oil
recovery (EOR) processes previously during its production life (see figure 3 & 4). After 4 years of natural depletion, a water
injector was put on line (MPL-16A), resulting in an increase in oil production at well MPL-07 from 400 to 1100 bbls/day. The
production then sharply declined until 2002. During this period, the water cut increased to 95%. It was then decided to inject a
slug of miscible injectant (MI). The production increased from 200 to 500 bbls/day for a year and a half and the water cut
decreased from 95% to 80%. The production then fell back to 150 bbls/day. Because of its ready availability, in May 2005, it
was decided to inject low salinity water. This process was succseful with the oil production increasing to a peak of 320
bbls/day followed by a decrease to 200 bbls/day. At the same time, the water cut dropped from 92% to 87%. The total liquid
rate was maintained at a stable rate of 2500 bbls/day until May 2006; this then increased to 3000 bbls/day thereafter (figure 5).
The injectivity of the low salinity injector (MPL-16A) stayed constant, indicating that little or no formation damaged occurred
owing to the swelling of clays or the generation of fines (figure 6).

Produced water chemistry:

The salinity of the produced water before the start of injection of the low salinity brine was around 14000 ppm total dissplved
solids (TDS) (see figure 7). In October 2005, the salinity in MPL-07 started to decline whilst the salinity of the water produced
from well MPL-11 stayed above 13000 ppm TDS until May 2006 and stabilized around 12200 ppm.. This is attributed to the
presence of sealing fault between MPL-16A and MPL-11. This is confirmed by the presence of shadow on the seismic survey
between the injection well and MPL-11.

The magnesium ion (Mg) concentration, in MPL-07 produced water (see figure 8), increased from 0.4 to 0.75 meq/l and then
sharply decreased below the detection limit (<0.01 meq/l) during 5 months of production. The Mg concentration then climbed
back to 0.5 mol/l before again returning to a concentration lower than the detection limit. After July 2007, the Mg
concentration returned to a concentration of 0.4 meq/l. The calcium ion (Ca) concentration does not exhibit such a dramatic
change as magnesium (see figure 9). The concentration fluctuates between 1.2 and 1.7 meq/l and most significantly, the sharp
drop in concentration shown by magnesium was not observed.

Single well chemical tracer test:

A new well was drilled and cored at a pad near the hydraulic unit where low salinity brine was injected, although the area
around the selected pattern had seen no water (high or low salinity). Single well chemical tracer tests were performed on this
well by injecting 4 different brines in sequence. The water sequence comprised a high salinity brine to establish a base line
residual oil saturation (Sor), MPL-07 produced water (16000 ppm TDS), Prince Creek aquifer water referred as the non-
optimized low salinity brine (2600 ppm TDS) and an optimized low salinity brine. The residual oil saturations derived from
this test are given below in Table 2.

Test description SOR Measured

High Salinity water 0.30 ± 0.02
Produced water 0.30 ± 0.02
Non Optimised Low salinity brine 0.28 ± 0.02
Optimised Low salinity brine 0.20 ± 0.02
Table 2, Results from the single well chemical tracer test

The results show that there is no difference between high salinity water and the produced water. The non-optimized low
salinity water decreased the Sor by 2 saturation units whilst the optimized brine lowered the Sor by 10 saturation units
compared to the base line. The high salinity water and the produced water were used to establish a solid base line and try to
lower the inherent uncertainty in the results. From the mechanistic study it was known that no difference in Sor will occur due
to the injection of the produced water and only a small decrease in Sor would occur from the non optimized low salinity brine.
By establishing a solid base line it was possible to prove that the non optimized brine has a genuine impact on the Sor, albeit a
small one, despite the 2% uncertainty.
4 SPE 113976

Discussion:
Field data are notoriously difficult to interpret as it is extremely difficult to maintain all parameters constant (i.e. injection rate,
production rate, possible interaction with other injection wells). However, the drop in observed water-oil ratio (WOR) owing
to the increase in oil rate at MPL-07 is seen to be occurring at the same time as the dramatic change in the produced water
chemistry. We believe that this is a clear signature of a low salinity waterflood in operation. A VIPTM model using the LoSalTM
code developed by Jerauld et al (2006) was used to confirm the response due to low salinity brine injection. The model was
history match until May 2005 and was therafter run as a predictive model forward in time. Two models were employed: one
where the low salinity brine was injected and another where only high salinity brine was injected (see figure 10). According to
the model, the LoSalTM effect should start producing incremental oil after November 2005. This corresponds to the observed
drop in WOR seen in MPL-07 and the dramatic change in water chemistry.

Mg variation:

Conventionally, waterflooding is used as a physical means to displace the oil from reservoirs and as a means of maintaining
the reservoir pressure. As such, very little chemical interaction is expected between the injectant and the reservoir, with the
exception of formation damage arising from the interaction between swelling clays and fresh water or mineral scaling owing
to the reaction between barium (present in the connate water) and sulfate (present in the injection water when seawater is used
at the injectant for example; Collins, 2005). However, according to the hydrogeology literature, if an aqeous fluid having a
different composition from the indigenous connate brine is injected into an aquifer, cation exchange will occur between the
mineral surfaces and the injected brine resulting in a significant change in the water chemistry (Valocchi et al., 1981). To this
end, Appelo et al. (1994) successfully described and modelled the freshening of an aquifer. These authors found that the key
characteristics of freshening in a saline aquifer arising from the injection of a low salinity water, is a rapid decrease in divalent
cation concentration (calcium and magnesium) to a concentration below that of the concentration in the low salinity water
injected, in conjunction with a monotonic decline in the concentration of chloride ions. The very same behavior was observed
during reduced condition coreflood experiments, as reported previously (Lager et al, 2006), and in the produced water of well
MPL-07 during the low salinity flood. The complete removal of Mg2+ in the produced water is the proof of the strong
interaction between the injection brine and reservoir rock. Furthermore the sharp drop in Mg2+ concentration confirmed the
self sharpening front inferred from the MIE mechanism discussed further in Seccombe et al. paper (2008). A non sharpening
front would have resulted in a highly dispersed response where Mg2+ concentration would have decreased gradually as seen
generally in field tracers.

Injectivity index and pH variation:

The results from the field described here suggest that two previously cited possible mechanisms for the increase in oil recovery
arising from low salinity water injection: fines migration and increase in solution pH, are not active. First, the injectivity
index stayed constant throughout the low salinity water injection. If fines migration was occurring, the fines would have
blocked some of the pore throats, resulting in a drop in injectivity. It is also interesting to note that prior to low salinity water
injection, the injectivity of MPL-16A was slowly decreasing (the data start at around 0.8 and finish at 0.6 prior to low salinity
water injection). This is believed to be due to the re- injection of un-filtered produced water which was slowly plugging the
near-well bore. The low salinity brine, being cleaner than the produced water, on the contrary did not plug the well bore and
helped to stabilize the injection rate. Another interesting conclusion can be drawn from the constant injectivity index that arose
post low salinity water injection: thermal fracturing of the reservoir, which happens when the water injected is relatively
colder than the formation, is unlikely to be occuring. Such fracturing could have been responsible for the increase in oil
recovery owing to a possible increase in injectivity. Regarding potential pH changes occurring in the reservoir on low salinity
water injection, it was noticed during some early reduced condition coreflood experiments that the pH of the effluent brine
suddenly increased from 9 to 10.5. It has been postulated that such increases could be responsible for an increase in oil
recovery (McGuire et al, 2005). However, later coreflood experiments showed that the low salinity effect also worked under
acidic conditions and that pH raises were probably experimental artifacts arising from de-pressurisation of the samples. The
field data shows very little variation in pH (see figure 11) and certainly does show not an increase up to pH 10.5 effectively
eliminating the pH mechanism.

Timing:

Owing to the lack of sulfate present in the low salinity injection water, barium ions can be considered as a non-reactive
conservative tracer. Thus by monitoring the change in concentration of barium in the produced water, it was possible to assess
when 1 pore volume (PV) of low salinity water had been injected (see figure 12). These data show that 13 months after the
SPE 113976 5

start of the injection, the concentration of barium had started to level off and it was therefore assumed that 1PV of water had
been injected. By back calculation, it was found that the low salinity effect started after ca. 0.3 PV’s of low salinity water had
been injected (see figure 13). During corefloods performed at reservoir conditions, it was found that in tertiary mode, an
increase in oil recovery was achieved after injecting ca.0.4 of a pore volume of low salinity brine. Considering the inherent
uncertainty associated with field data, the results are in good agreement and the timing of the low salinity response can be
considered to be rapid.

Future development and optimization:

The increase in oil recovery owing to the injection of low salinity brine, in this field case, was relatively small; however, this
was expected from experiments performed at BP’s Sunbury laboratory which suggested that the water was not optimum for
this reservoir. This was later confirmed by the SWCTT performed in an adjacent well. This test reported only a 2% decrease in
SOR whereas by injecting a low salinity brine optimized for the reservoir, a 10% decrease in SOR was achieved. The
optimization was based on the mineralogy of the reservoir, its production history, and the composition of the connate brine.
The next step will be to treat the injection water to generate an optimized brine and to flood the whole reservoir with this low
salinity water.

Conclusion:
The injection of low salinity brine into an Alaskan reservoir was successful. The remaining uncertainty concerning how MIE
scaled to inter-well distances has been alleviated. A measurable drop in WOR was observed and the oil production rate was
doubled during nearly 12 months of production. Furthermore, at the same time, the water chemistry of the produced brine
changed dramatically following the trend predicted by the theory of low salinity waterflooding previously proposed by the
authors and as previously reported during reservoir condition coreflood experiments. The timing of the effect was fast, despite
being a tertiary flood, and was accurately predicted by BP’s in-house proprietary Low salinity VIPTM model. Incremental oil
was observed after ca. 0.3 pore volumes of low salinity water had been injected. No adverse effects such as clay swelling or
pore plugging owing to fines migration were observed.The next step will be to treat the injection water to generate an
optimized brine and to flood the whole reservoir with this low salinity water.

References:

Appelo, C.J., 1994. Cation and proton exchange, pH variations and carbonate reactions in a freshening aquifer. water resources
research, 36(10): 2793-2805.
Collins, I. R., 2005. Prediction the location of barium sulfate scale formation in production systems, SPE 95067.
Jadhunandan, P., 1990. Effects of brine composition, crude oil and aging conditions on wettability and oil recovery. PhD
dissertation.
Jadhunandan, P. and Morrow, N.R., 1991. Spontaneous imbibition of water by crude oil/brine/rock systems. In Situ, 15(4):
319-345.
Jadhunandan, P. and Morrow, N.R., 1995. Effect of wettability on waterflood recovery for crude oil/brine/rock systems.
SPERE, 10(1): 40-46.
Jerauld, G.R., Lin, C.Y, Webb, K.J., and Seccombe, J.C., “Modeling Low-Salinity Waterflooding” SPE 102239, presented at
the Annual Meeting of SPE, 2006.
Lager, A., Webb, K.J., Black, C.J.J., Singleton, M. and Sorbie, K.S., 2006. Low salinity oil recovery- An experimental
investigation, International symposium of the society of core analysts, Trondheim.
Mcguire, P., Chatham, J.R., Paskvan, F., Sommer, D. and Carini, F.H., 2005. Low salinity oil recovery: an exciting
opportunity for Alaska's North Slope. SPE, 93903.
Rueslatten, H.G., Hjelmeland, O. and Selle, O.M., 1994. Wettability of reservoir rocks and the influence of organo-metallic
compounds. North Sea oil and gas reservoir, 3: 317-324.
Seccombe, J.C., Lager, A., Webb, K.J., Jerauld, G. and Fueg, E., 2008. Improved waterflood recovery: LoSalTM EOR field
evaluation. SPE 113480.
Tang, G.-Q. and Morrow, N.R., 1999. Influence of brine composition and fines migration on crude oil brine rock interactions
and oil recovery. journal of petroleum science & engineering, 24: 99-111.
Valocchi, A.J., Street, R.I. and Roberts, P.V., 1981. Transport of ion exchanging solutes in ground water: chromatography
theory and field simulation. water resources research, 17: 1517-1527.
Webb, K.J., Black, C.J.J. and Al-Jeel, H., 2004. Low salinity oil recovery - log inject log. SPE, 89379.
Yildiz, H.O. and Morrow, N.R., 1996. Effect of brine composition on recovery waterflooding of Moutray crude oil by.
Petroleum science & engineering, 14: 159-168.
6 SPE 113976

OIL
OIL

OIL

C
C
OIL O - O
C O O
H + H
R1 R2 R3
O - O Ca
O-
R N+ Ca2+ Mg2+
O- O- O O-

Clay Clay Clay

Clay
Cation exchange Cation bridging Ligand bridging Water bridging
Figure 1, Clay/Oil Attraction by Divalent Cations

Figure 2, L16A waterflood pattern. Well spacing 1200' - 1700'. OOIP 9.9 mmstb
 SPE 113976

Oil rate
water cut

0
200
400
600
800
1000
1200
1400
1600

0%
20%
40%
60%
80%
100%
12/07/1995
12/07/1994

12/07/1995 12/06/1996

12/06/1996 12/06/1997

12/06/1997
Natural depletion

12/06/1998
12/06/1998
12/06/1999
12/06/1999

12/05/2000
12/05/2000

12/05/2001 12/05/2001

Date
Lowess = 0.05
MPL-07 Oil rate

L-16A Waterflood

12/05/2002
12/05/2002

Figure 4, Water cut of MPL-07

MI

MI
12/05/2003
Figure 3, Oil rate (bbls/day) of MPL-07
12/05/2003

L-16A water injection

12/04/2004
12/04/2004
12/04/2005

12/04/2006 12/04/2005

12/04/2007
Low salinity

12/04/2006

Low salinity
12/03/2008
12/04/2007
7
8 SPE 113976

1.2
1000 4000

1.0

800
0.8
3000

600 0.6

0.4
400 2000

0.2

200
0.0
1000

0 -0.2
04/23/2004 02/17/2005 12/14/2005 10/10/2006 08/06/2007
09/20/2004 07/17/2005 05/13/2006 03/09/2007
Oil Vol Metered(L) Gas Vol Metered(L) Mg(R) Total liquid(R 2)

Figure 1, Production data of MPL-07. Oil rate and total liquid in bbls/day, Gas volume in scf, Magnesium concentration in
mol/l

Injector MPL-16A
1.4
Shut in due to ESP failure in L-07
1.2
LoSal injection

1.0
Injectivity index

0.8

0.6

0.4

0.2

0.0
12/30/1998 12/30/2000 12/30/2002 12/30/2004 12/30/2006
12/31/1999 12/30/2001 12/31/2003 12/30/2005 12/31/2007

Figure 5, Injectivity index of MPL-16A

SPE 113976 9

16000

14000

12000
TDS (ppm)

10000

8000

Start of Low salinity brine injection, May 2005

6000

12/30/04 12/30/05 12/30/06 12/31/07

MPL-07 MPL-11

Figure 6, Change in produced water salinity in MPL-07 & MPL-11

0.8

0.7

0.6

0.5
Mg (meq/l)

0.4

0.3

0.2

0.1

0.0

-0.1
12/30/04 12/30/05 12/30/06 12/31/07
Date
Figure 7, Magnesium concentration in MPL-07 produced water with time
10 SPE 113976

1.8

1.6

1.4

1.2

1.0
mEq / l

0.8

0.6

0.4

0.2

0.0
12/30/2004 12/30/2005 12/30/2006 12/31/2007

Figure 8, Change in Calcium concentration in MPL-07 produced water with time

Figure 9, Comparison between production data (green triangle), LoSal VIP model (red line) and High salinity VIP model
(Black line)
SPE 113976 11

10

9.5

8.5

7.5

7
Low salinity injection
6.5

5.5

5
11/19/2000 11/19/2001 11/19/2002 11/19/2003 11/19/2004 11/19/2005 11/19/2006

Figure 10, Change in pH (Y-axis) during low salinity flood with time

0.12

0.1

1 Pore volume
0.08

0.06

0.04

0.02

0
4/5/2005
5/5/2005
6/5/2005
7/5/2005
8/5/2005
9/5/2005
10/5/2005
11/5/2005
12/5/2005
1/5/2006
2/5/2006
3/5/2006
4/5/2006
5/5/2006
6/5/2006
7/5/2006
8/5/2006
9/5/2006
10/5/2006
11/5/2006
12/5/2006
1/5/2007
2/5/2007
3/5/2007
4/5/2007
5/5/2007
6/5/2007
7/5/2007

Figure 21, Change in Barium concentration (in mmol/l) in the produced water. With time
12 SPE 113976

400

350

300

250
Oil rate (bbls/day)

200

150

100

50

0
-0.2 0.3 0.8 1.3 1.8
Pore volume injected

Oil rate Mg2+ concentration

Figure 32, Oil rate Vs Pore volume injected.