Journal of Petroleum Science and Engineering 156 (2017) 102–109

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

journal homepage: www.elsevier.com/locate/petrol

Impact of temperature on the low salinity EOR effect for sandstone cores
containing reactive plagioclase
Iv
~erez Torrijos, Tina Puntervold*, Skule Strand, Tor Austad, Vinh Vuong Tran,
an D. Pin
Kaia Olsen
University of Stavanger, 4036, Stavanger Norway

A R T I C L E I N F O A B S T R A C T

Keywords: Previous experimental work has indicated that the induced increase in pH as the high salinity formation brine is
Smart water displaced by low salinity brine, is a key factor for observing low salinity EOR effects. The pH gradient is mainly
Plagioclase due to desorption of Ca2þ from the clay surface in the presence of low salinity brine depleted in Ca2þ. Desorption
EOR
of Ca2þ from the clay surface is an exothermic process, which decreases the pH gradient as the temperature is
Sandstones
increased. Therefore, a decrease in the low salinity EOR effect at high temperatures is observed. Some types of
Low salinity
Clays reactive plagioclase, especially albite, may also influence the pH as the brine salinity is changed. The exchange of
Naþ with Hþ is less temperature sensitive.
Tertiary low salinity EOR effects from outcrop core material containing both clay, ≈10 wt%, and reactive albite,
≈30 wt%, have been studied at 60 and 120  C. The low salinity EOR effects at the two temperatures were quite
similar, 9.2 and 8.7% of OOIP, respectively. The oil recovery test results agreed well with the observed pH
gradients for the oil-free core material, which were also quite similar, 2.3 and 2.2 pH units at 60 and 120  C. The
oil recovery and pH screening results showed that ion exchange at the albite surface will partly compensate for the
decreased ion exchange at the clay surface at higher temperatures. Therefore, the presence of reactive plagioclase
in high temperature sandstone reservoirs seems to be favorable for observing LS EOR effects provided the initial
pH of formation water is low enough to make the rock mixed wet.

1. Introduction high FW salinity, and the reason is that the hydration number of metal
cations in aqueous solutions decrease as temperature increase, and the
Based on experimental data, it has previously been pointed out that cations become more reactive toward the surface. Increased reactivity of
the low salinity (LS) enhanced oil recovery (EOR) effect in sandstone was the inorganic ions prevents the adsorption of the polar organic molecules,
related to wettability alteration triggered by an increase in pH as the high and the surface becomes more water-wet (Aghaeifar et al., 2015; Zavit-
salinity (HS) formation water (FW) was displaced by the low salinity sas, 2005). The initial pH of the FW, which should be in the range of
brine. This effect is also sometimes referred to as a Smart Water EOR 5–6.5, and the clay content are very important for obtaining mixed-wet
effect (Austad et al., 2010; RezaeiDoust et al., 2010, 2011). Often, the reservoir conditions, which usually result in the best LS EOR response
induced pH change can be buffered, and in that case, no pH change (Pi~
nerez Torrijos et al., 2016a).
would be observed in the produced water. In BP's two-well field test in A pH gradient of 2–3 pH units, e.g. from 6 to 9, is related to a large
Alaska (Seccombe et al., 2010), no significant pH change was observed in decrease in the concentration of Hþ, i.e. from 106 to 109 M. The mass
the produced water, but compositional analysis of the produced water of material linked to this pH change is, however, extremely small, as
showed changes in HCO 2þ
3 , H2S, Fe , and Mg
2þ
concentrations. These indicated by the above given concentrations. Different minerals present
changes could be directly related to an increase in the pH during the LS in the porous medium can influence this pH gradient. It is therefore
slug injection, as discussed in a previous paper (RezaeiDoust et al., 2011). understandable that predicting possible LS EOR effects is difficult based
However, if the core material is too water-wet, a significant pH gradient on general reservoir data like clay content, crude oil properties (acid
but no LS EOR effect can be observed (Aghaeifar et al., 2015; RezaeiDoust and base number), temperature, and FW brine composition. A change of
et al., 2010). This could be the case at high temperatures combined with more than one pH unit close to pH 7, will have relatively small effects

* Corresponding author.
E-mail addresses: [email protected] (I.D. Pi~
nerez Torrijos), [email protected] (T. Puntervold).

http://dx.doi.org/10.1016/j.petrol.2017.05.014
Received 10 August 2016; Received in revised form 12 May 2017; Accepted 16 May 2017
Available online 17 May 2017
0920-4105/© 2017 Elsevier B.V. All rights reserved.
I.D. Pi~
nerez Torrijos et al. Journal of Petroleum Science and Engineering 156 (2017) 102–109

on the properties of the reservoir minerals, but could have dramatic reactions) are extremely fast. The question is: “How is the pH gradient,
effects on the reactivity of polar organic components present in the ΔpH, created as the HS formation water is displaced by the LS injection
crude oil. Crude oil adhesion tests performed by Buckley and Morrow brine?” Two important pH-influencing mineral reactions have been
on silica surfaces in the presence of various brines, showed that adhe- identified involving clay minerals and sometimes reactive plagioclase/
sion was dependent on pH (Buckley and Morrow, 1990). Similar ob- albite (NaAlSi3O8) as indicated by the following reactions (Friedman
servations were also made by Didier et al. (2015) for two different et al., 1992; Strand et al., 2014):
sands. They observed that at low salinity and acidic pH (pH < 7), oil
adhesion increased. Fogden (2012) measured deposition of crude oil Clay  Ca2þ þ H2 O ↔ Clay  Hþ þ Ca2þ þ OH þ heat (5)
onto a kaolinite surface at various pH values, and he also found that the
deposition was dependent on pH. NaAlSi3 O8 þ H2 O ↔ HAlSi3 O8 þ OH þ Naþ (6)
Acidic and basic organic molecules are acting as anchor molecules
towards the rock surface, especially towards the clay minerals. Molecules Both clay minerals and feldspar minerals contain permanent negative
containing both basic, R3N:, and acidic, RCOOH, functional groups, sites that must be charge-balanced by cations. If reactive cations are
quantified by acid and base numbers (AN and BN, given in mgKOH/g oil) absent or low in concentration, which is the case in LS brine, reactive
are surface-active towards negatively charged clay minerals, such as protons, Hþ, partly balance the negative mineral charges, resulting in a
kaolinite and illite, and the concentration of their reactive form is pH increase. Because of the strong hydration enthalpy of Ca2þ compared
depending on the pH, as illustrated in Eqs. (1) and (2): to that of Naþ, 380.6 and 96.9 kcal/mol, Eq. (5) is exothermic in
nature (Aksulu et al., 2012). This means that the pH gradient decreases as
Basic material : R3 N : þ H2 O ↔ R3 NHþ þOH (1) the temperature increases for core material without reactive plagioclase.
A large decrease in ΔpH was observed when the reservoir temperature
increased beyond 100  C in core material containing clay but negligible
Acidic material : RCOOH þ H2 O ↔ RCOO þ H3 Oþ (2)
amounts of reactive plagioclase (Aksulu et al., 2012). Using the same core
The most reactive form is shown in bold (Austad et al., 2010; Burgos material, a significant LS EOR effect was observed when aging and
et al., 2002; Madsen and Lind, 1998; RezaeiDoust et al., 2011). The flooding the core at 90  C, while no LS EOR effect was observed when
protonated base, R3NHþ, will adsorb onto negatively charged clays by aging and flooding at 130  C (RezaeiDoust et al., 2010). The fact that the
electrostatic interaction, as observed both on kaolinite and illite clay temperature could influence the LS EOR potential was also proposed by
minerals (Aksulu et al., 2012; RezaeiDoust et al., 2011). Even though the Gamage and Thyne (2011). Presence of dissolvable anhydrite, CaSO4, in
CEC and basal plane charges may be different for the two types of clay core material will suppress the desorption of Ca2þ from the clay surface,
minerals, and thus show slight differences in surface interactions with i.e. Eq. (5) is moved to the left, and the pH gradient decreases (Aksulu
brine and oil components, the clay minerals have been treated the same et al., 2012).
in the proposed chemical model for wettability alteration (Austad et al., If, however, the core material contains significant amounts of reactive
2010; RezaeiDoust et al., 2011). The adsorption of benzoic acid onto plagioclase/albite, the imposed pH gradient at high temperature may be
kaolinite at 32  C increased from 0.1 to 3.7 μmol/m2 as the pH decreased mostly determined by Eq. (6), i.e. the pH gradient will be nearly inde-
from 8.1 to 5.3 (Madsen and Lind, 1998). Thus, according to the data pendent of temperature. The LS EOR potential should then be the same at
gathered by Madsen and Lind (1998), the adsorption of benzoic acid onto both low and high temperatures. This hypothesis was addressed in this
kaolinite increased significantly, by a factor 37, as the pH decreased from paper. Outcrop sandstone core material containing significant amounts
8.1 to 5.3. These results confirm that it is the protonated acidic compo- of reactive plagioclase, that previously have shown good reproducibility
nents, RCOOH, which have the highest affinity to the kaolinite surface. (Austad et al., 2010; RezaeiDoust et al., 2011), was used in tertiary LS
Protonated acidic components are present in larger fraction at pH 5.3 EOR experiments at 60 and 120  C. Furthermore, a pH scan was per-
than at pH 8.1. On the basis of the reported adsorption data by Madsen formed by flooding a 100% brine-saturated (oil-free) core successively
and Lind (1998) in the pH range 5.3–8.1, and of reported anionic ex- with FW-LS-FW at the given temperatures. The effluent pH was measured
change capacities of kaolinite clay only increasing from 0.6 to 0.8 meq/ and the results were compared with the produced water (PW) effluent
kg in the same pH range (Allard et al., 1983), it has been suggested by samples collected during the oil recovery experiments.
Austad and co-workers that the adsorption of acidic components,
RCOOH, to the clay surface takes place by hydrogen bonding involving 2. Experimental
protons adsorbed on the clay surface (Austad et al., 2010; RezaeiDoust
et al., 2011). It is interesting to note that the pKa values of these active 2.1. Materials
compounds are very similar, pKa ≈ 5.5, meaning that the concentrations
change similarly by pH changes. As the pH increases, Eq. (1) is moved to 2.1.1. Outcrop core plugs
the left and Eq. (2) is moved to the right. Thus, both the basic and acidic Outcrop sandstone cores and their mineralogical composition were
materials are transformed into species of lower affinity towards the clay provided by TOTAL. The mineralogical composition is given in Table 1.
surface, and the water wetness of the clay minerals increases as indicated Physical core properties are listed in Table 2. Specific surface area
by the following equations between the adsorbed polar components and measurements by the BET method were carried out on small represen-
the hydroxide ions: tative pieces of the outcrop material. It is important to emphasize that the
samples were not crushed to avoid new matrix surfaces.
Clay  R3 NHþ þ OH ¼ Clay þ R3 N : þ H2 O (3)
Surface area measurements were carried out in a TriStar II PLUS in-
strument from Metromeritics®. The instrument allowed measurements at
Clay  RCOOH þ OH ¼ Clay þ RCOO þ H2 O (4) very low surface areas. The porous rock samples were not crushed prior
It is important to note that proton transfer reactions (acid-base to analysis.

Table 1
XRD/XRF analyses of mineralogical composition in wt% of the outcrop core material.

Core plugs Albite Quartz Calcite Apatite Pyrite Anatase Chlorite-C Illite Total Clays & Micas

B-21/B-22 32.9 56.2 0.3 0.2 0.1 0.5 1.7 8.1 9.8
C-2 30.6 56.6 0.3 0.2 0.1 0.6 2.2 9.4 11.6

103
I.D. Pi~
nerez Torrijos et al. Journal of Petroleum Science and Engineering 156 (2017) 102–109

Table 2 2.2. Methods

Physical properties of the outcrop cores.

B-21 B-22 C-2 2.2.1. Core flooding setup

Dry weight (g) 235.1 237.1 163.0 The core flooding setup consisted of a Gilson HPLC pump, stainless
Length (cm) 10.00 10.00 7.00 steel piston cells, a Hassler core holder, an oven, and an atmospheric
Diameter (cm) 3.78 3.78 3.79 separator/glass burette. A computer system controls the injection rate
Bulk volume (cm3) 112.2 112.2 78.9 and sample all relevant data such as core temperature, inlet pressure and
Pore volume (cm3) 23.3 23.0 15.9
Porosity (%) 20.8 20.5 20.1
differential pressure across the core. The same core flooding set up was
Permeability (md) 48.4 135.1 3.4 used in both the oil recovery and surface reactivity tests.
BET (m2/g) 1.81 ± 0.02
2.2.2. Core restoration
Table 3
Crude oil properties. 2.2.2.1. Mild core cleaning. After an oil recovery test, the outcrop core
was mildly cleaned by flooding at a rate of 0.1 ml/min with: (1) Kerosene
AN BN Density at 20  C, Viscosity at 25  C,
(mg KOH/g) (mg KOH/g) (g/cm3) (cP)
until a clear effluent; (2) Heptane to displace kerosene; (3) 4 PV of LS
brine to displace FW and easily dissolvable salts; (4) Finally, the cores
Crude Oil <0.10 1.8 0.846 17.6
were dried at 90  C to constant weight.

2.1.2. Crude oil 2.2.2.2. Initial water saturation. Initial FW saturation of 20% was
The stabilized crude oil was centrifuged for 45 min at a high rotation established using the desiccator technique in line with the procedure
speed and filtered through a 5-μm Millipore filter to avoid the presence of described by Springer et al. (2003). The principle of the method is to
impurities. saturate the core with diluted FW brine, then evaporated water from the
Acid and Base number analyses were performed at room temperature. core until the core reaches a target weight, achieving Swi ¼ 0.20 FW.
Potentiometric titrations were carried out according to the modified Afterwards, the core was left to equilibrate for 72 h. The benefits of using
methods of ASTM 665 and ASTM 2898 described by Fan and Buckley the desiccator method in comparison to e.g. the porous plate method, are
(2007). Repeatability of the analysis was ±0.02. that repeated consistent initial water saturations are generated, which is
Density measurement of the liquid samples were carried out at 20  C important in parametric studies, and the method is also time-efficient.
using a DMA-4500 density meter delivered by Anton Paar.
Viscosity measurements of brine and crude oil were determined using a 2.2.2.3. Crude oil saturation. The core with Swi ¼ 0.20 was placed in a
Physica MCR 302 rheometer from Anton Paar. The shear rates ranged core holder and evacuated for a short time prior to oil saturation. The
from 0.1 to 1000 s1. At lower temperatures the measurements were core was saturated and flooded at 50  C with 2 PVs of crude oil in each
performed with a plate and cone geometry, using a cone with a diameter direction, for even distribution of crude oil. Then the core was aged in
of 50 mm and 1 of inclination. At high temperature the measurements crude oil in a pressurized steel ageing cell at the respective test temper-
were performed in a pressure cell with a bob and cylinder configuration ature for 14 days. The core ageing at high temperature was performed
delivered by Anton Paar. N2 was used to pressurize the cell to the test using a back pressure of 10 bar.
pressure of 10 bars. The percent error for the data presented is below 1%.
The oil properties are displayed in Table 3. 2.2.3. Oil recovery test
The aged core was mounted in the core holder with a confining
2.1.3. Brines pressure of 20 bar and a back pressure of 10 bar. The core was left
All chemicals used to prepare the brines were reagent grade and overnight to achieve thermal stability at the test temperature. The core
delivered by Merck laboratories. Deionized (DI) water was used for the was successively flooded at a rate of 4 PV/D with FW and LS brine. At the
brine preparation. Brine compositions are provided in Table 4. Formation end of the test, the rate of the LS brine injection was increased to 16 PV/D
water (FW) and low salinity water (LS) were the injection fluids used in to detect possible viscous end effects. The cumulative oil produced was
the experiments. FW is a mixture between CaCl2 and NaCl with a con- monitored at room conditions in a graded glass burette. The accuracy of
centration of Ca2þ of 90 mM and a total salinity of 100000 ppm. The LS the burette reading is ±0.1 ml. Produced water samples were collected
brine is a 1000 ppm NaCl solution. No swelling of clays has been detected through the bottom valve. Thermal expansion of fluids during heating
when injecting the LS brine. accounts for a small fraction of the first oil produced during FW flooding,
pH measurements were performed using a Seveneasy™ pH meter from but during the following LS injection this effect did not influence the
Mettler Toledo. A Semi micro-pH combination electrode for small liquid results. The pressure drop values collected were average values during
volumes was used. The electrode was triple point calibrated with buffers 10 s of data observation.
at pH 4, 7 and 10. The relative accuracy of the measurements was ± 0.01
pH units at room temperature. pH measurements were performed on 2.2.4. pH screening test
newly produced water samples at ambient temperature. Chemical interactions between the core minerals and the injection
brine were investigated by performing a pH screening test. The mildly
cleaned core was successively flooded at a rate of 4 PV/D by FW- LS - FW.
Table 4
Brine compositions and properties. Effluent samples were collected and pH and density were plotted against
PV injected. The test was performed at 60, 90 and 120  C. Note that the
FW (mM) LS (mM)
pH screening test alone only indicates a possible EOR effect, as it only
þ
[Na ] 1540 17.1 reveals the brine-solid interactions taking place, without the presence of
[Kþ] – –
crude oil.
[Ca2þ] 90 –
[Mg2þ] – –
[Cl] 1720 17.1 2.2.5. Spontaneous imbibition test
Ionic strength 1810 17.1 A spontaneous imbibition test was performed on a mildly cleaned
Salinity (ppm) 100000 1000 sandstone core. The oil-saturated and aged core with Swi ¼ 0.20 FW was
Density (g/cm3) 1.067 0.999
pH 5.5 5.7
placed in a glass Amott imbibition cell surrounded by FW as imbibing

104
I.D. Pi~
nerez Torrijos et al. Journal of Petroleum Science and Engineering 156 (2017) 102–109

brine at 60  C. When the oil recovery plateau was reached with FW, the and LS brine at a rate of 4 PV/D. During the secondary flooding with FW
imbibing brine was changed to LS brine. Cumulative oil production was the oil recovery plateau of 40.1% of OOIP was reached after 2 PV, Fig. 1.
monitored and plotted against time. After 4 PV injected, the flooding fluid was switched to the LS brine No
swelling of clays took place during LS brine injection, as no increase in
3. Results and discussion pressure drop was observed. A tertiary LS EOR effect was observed, and
the oil recovery increased to 49.3 %OOIP after 12 PV injected. The LS
Plagioclase minerals, which can be present in sandstone rock, can EOR effect was 9.2 %OOIP. At the end, the flooding rate was increased to
influence the initial pH of the formation water, affect the initial wetting, 16 PV/D, and 4.8 %OOIP extra oil was produced, which may be linked to
and the pH development during a LS flood (Reinholdtsen et al., 2011; increased viscous forces and/or possible end effects. In a previous study
Strand et al., 2014; Xie et al., 2012). Albite has a CEC of 1.5–4 meq/Kg in on the same core material, using the same brines and test temperature,
the pH range of 4–8, which is lower than that of clays, and it also has a the flooding rate was increased prior to injection of LS brine, to eliminate
lower surface area than clays (Allard et al., 1983). Another important influence of possible end effects on the LS EOR effect (RezaeiDoust et al.,
characteristic of albite is that its zeta potential is negative in the pH range 2011). The results from that study showed that the LS EOR effect was
of reservoirs (Oelkers et al., 2009). Reactive albite, with the chemical real, and not influenced by end effects.
formula NaAlSi3O8, reacts as shown in Eq. (6), by exchanging Naþ with Produced water samples were collected and the pH was measured. A
Hþ in the presence of LS water, giving increased pH in the system (Blum, pH increase of 0.5 pH units was observed, from about 6.1 until it stabi-
1994). It has been stated for albite, that the initial cation exchange of Naþ lized at 6.6. During the LS flooding, the pH increased from 6.6 to 8.9, and
by Hþ from the aqueous phase is rapid, at equilibrium and forming it remained quite stable. A low initial pH close to 6 favors adsorption of
hydrogen feldspar on the surface (Chou and Wollast, 1985). For illite it organic material onto the clays and promotes initial mixed-wet condi-
has been stated that the interlayer Kþ-ions are exchangeable, but so tions. During the LS flooding, the pH increased and extra oil was pro-
slowly and by such strong solutions that it is usually described as “fixed duced because of a wettability alteration towards more water-wet
potassium” (Garrels, 1984). Therefore, even though the CEC of albite is conditions. During this process, not all mobilized oil by wettability
much smaller than that for clays, the feldspar can significantly contribute alteration is expelled from the core (Morrow, 1990). A fraction of this oil
to the pH increase observed, because the cation exchange is faster at the could be capillary trapped (Pi~ nerez Torrijos et al., 2016b; Shiran and
feldspar surface than in the interlayers of illite. Skauge, 2013). It should be noticed that the observed increase in oil
Naþ is not strongly solvated by water molecules, therefore the ion recovery is synchronized with the increase in pH, confirming the strong
exchange and the corresponding pH increase is probably not very tem- relationship between the LS EOR effect and the induced pH gradient. This
perature dependent. Kia et al. found that in a Berea core containing 8% relationship was also observed in a previous study (Pi~nerez Torrijos et al.,
clay, only 5% of the CEC was linked to hydrogen exchange and therefore 2016b). A strong link between salinity changes and pH changes during
pH change (Kia et al., 1987). Clays have a large surface area, and the CEC waterflooding has been pointed out previously by Vaidya and Fogler in
is also linked to adsorption of polar organic components from the crude work related to Berea sandstone cores (Vaidya and Fogler, 1992).
oil, and to adsorption of inorganic ions in the brine phase. The sandstone used in these experiments, is an outcrop material that
The temperature effect of the core material was tested using both oil has been exposed to rain water/percolating water over a long period of
recovery tests at 60 and 120  C, and pH screening tests at 60, 90 time. Albite minerals could then partly be transformed into their pro-
and 120  C. tonated form, Eq. (6). In the presence of high salinity (HS) FW, acidic
The outcrop core material has been used in numerous experiments conditions will be promoted as shown by Eq. (7). Initial acidic reservoir
previously, giving reproducible results (Austad et al., 2010; Pi~ nerez condition is favorable for adsorption of polar components onto the clay
Torrijos et al., 2016b; RezaeiDoust et al., 2011). The physical and surface (Austad et al., 2010; RezaeiDoust et al., 2011).
mineralogical data are given in Tables 1 and 2. Cores of the same size
were used in the viscous flooding experiments, while the core used for HAlSi3 O8 þ Naþ ↔ NaAlSi3 O8 þ Hþ (7)
spontaneous imbibition was a little shorter. The total clay content was The maximum pH value during the LS flooding was obtained as the oil
close to 10 wt%, mainly illite clay, which is within the range where recovery plateau of 49.3 %OOIP was reached, Fig. 1. The produced
significant LS EOR effects are observed. The core material also contained alkalinity, due to ion exchange at the clay and albite surfaces, Eqs. (5)
more than 30 wt% of albite, and provided that the albite is of the reactive and (6), promotes desorption of organic material from the clay surface,
type, it should have a significant effect both on the initial pH of the FW Eqs. (3) and (4). After the oil recovery plateau was reached, a steady state
and also on the induced pH gradient as the FW is displaced by the LS situation was established. This is confirmed by the fact that the pH was
brine. Calcite, which is also present in the core material can dissolve in stable during the high rate of LS flooding. The extra oil produced during
water and increase pH to some extent. Calcite is present in a small the high rate LS flooding is probably not related to wettability alteration,
amount, and the solubility of it in LS brine is quite low, thus its contri- but most likely linked to increased viscous forces and/or end effects.
bution to the pH increase is expected to be lower than that from the The oil recovery test at 120  C was performed on the outcrop core B-22.
plagioclase (Pi~nerez Torrijos et al., 2016c). The core was restored with Swi ¼ 0.20, and saturated, flooded and aged
The acid number, AN, of the crude oil was low, AN<0.10 mg KOH/g, with crude oil at the test temperature. Apart from the temperature, the
but the base number, BN, was high, BN ¼ 1.8 mg KOH/g, Table 3. conditions were similar to those of the test at 60  C.
Therefore, there should be enough polar components in the crude oil to The secondary oil recovery plateau with FW was 35.3 %OOIP, which
make the rock mixed wet, if the initial conditions are acidic, pH < 7. The is 4.8% OOIP lower compared to that of the experiment at 60  C, Fig. 2.
composition of the formation water (FW) and low saline brine (LS) is This result can be explained by the difference in pressure drop observed
shown in Table 4. The FW was composed of NaCl and CaCl2 with a total at 60 and 120  C. The difference in pressure drop at the two temperatures
salinity of 100 000 ppm. The concentration of Ca2þ was 90 mM, which can be a result of several factors; e.g. the absolute permeabilities of the
should give reasonable adsorption onto the clay minerals. The LS brine core samples, the fluid viscosities at various temperatures, and small
was 1000 ppm NaCl. differences in initial core wetting at the two different temperatures.
Adsorption of polar organic molecules onto the clay surface can decrease
3.1. Oil recovery tests because inorganic ions like Ca2þ become dehydrated at higher temper-
atures, and their reactivity toward the clay surface increase. Higher water
The oil recovery test at 60  C was performed on core B-21. The core was wetness can result in more trapped oil due to increased capillary forces.
restored with Swi ¼ 0.20, and saturated, flooded and aged with crude oil During the LS flood, the oil recovery increased to about 44.0 %OOIP,
at the test temperature. Then, the core was successively flooded with FW

105
I.D. Pi~
nerez Torrijos et al. Journal of Petroleum Science and Engineering 156 (2017) 102–109

Fig. 1. Oil recovery test at 60  C from core B-21. The core with Swi ¼ 0.20 was saturated and aged in crude oil. The core was successively flooded with FW-LS at a rate of 4 PV/D. At the end
the rate was increased to 16 PV/D. (a) Oil recovery and pH development, (b) Pressure drop vs PV injected.

Fig. 2. Oil recovery test at 120  C from core B-22. The core with Swi ¼ 0.20 was saturated and aged in crude oil. The core was successively flooded with FW-LS at a rate of 4 PV/D. At the
end the rate was increased to 16 PV/D. (a) Oil recovery and pH development, (b) Pressure drop vs PV injected.

showing a tertiary LS EOR effect of 8.7% of OOIP, which is in the same and small differences in initial wetting properties can influence the ul-
range as for the test at 60  C. Also in this case, no swelling of clays took timate recoveries at the two temperatures.
place during LS brine injection, as no increase in pressure drop was As noticed in Table 5, the viscosity ratio between crude oil and FW at
observed. When the flooding rate was increased to 16 PV/D, 6.6% OOIP 60 and 120  C is lower and therefore more favorable than the viscosity
extra oil was produced. ratio between crude oil and LS brine during the tertiary recovery.
For comparison, the oil recovery profiles at 60 and 120  C are pre- Pendant drop measurements performed on the same crude oil – brine
sented in the same chart, Fig. 3. systems showed only minor changes in interfacial tension (IFT) by
The oil recovery profiles are very similar, but the ultimate oil recovery changing salinity, temperature or pH, varying in the range 19–26 dyn/
was 4.8 %OOIP lower at high temperature. This difference in ultimate cm. Thus, IFT does not have a significant influence on the oil recoveries
recovery cannot be linked only to differences in fluid viscosities or to observed in this study (RezaeiDoust, 2011). This shows that the extra oil
mobility ratios between 60 and 120  C, listed in Table 5. Also the dif- recovered during the LS brine injection cannot be linked to an increase in
ference in core permeabilities, the resulting oil recovery pressure drop viscous forces nor to a change in IFT, but rather be related to increased
capillary forces (increased water wetness) and improved microscopic
sweep efficiency.
In addition, the pH change in the produced water during the flooding
sequence was very similar at both test temperatures as shown by Fig. 4.
The temperature does not seem to have any significant influence on the
pH gradient.

3.1.1. pH screening
To confirm the LS EOR potential for this outcrop material, a pH
screening test was performed on core B21 to study the interaction be-
tween the minerals at the pore surface and the different injection brines
used, FW and LS brine. The test was performed without the influence of
crude oil (Aksulu et al., 2012). Valuable information can be obtained

Table 5
Viscosities* of crude oil and brines at different temperatures.

Temp, ( C) Crude oil (cp) FW(cp) LS (cp) Viscosity ratio Viscosity ratio
(μo/μwFW) (μo/μwLS)

25 17.60 1.18 1.01 14.92 17.43

Fig. 3. Comparison of oil recovery profiles during oil recovery tests at 60  C and 120  C.
60 4.59 0.63 0.60 7.29 7.65
The cores were successively flooded with FW-LS at a rate of 4 PV/D. At the end the
120 2.02 0.29* 0.23* 6.97 8.78
flooding rate was increased to 16 PV/D.

106
I.D. Pi~
nerez Torrijos et al. Journal of Petroleum Science and Engineering 156 (2017) 102–109

Table 6
pH gradients observed in the pH scan and during the tertiary LS EOR test.

Temp. ΔpH (FW/LS) ΔpH (FW/LS) LS EOR effect

( C) pH scan Oil Recovery (% of OOIP)

60 3.0 2.3 9.2

90 3.0 – –
120 2.7 2.2 8.7

gradient can be observed, but no extra oil can be produced in a wetta-

bility alteration process.

3.1.2. Wettability alteration by spontaneous imbibition

To confirm wettability alteration, a mildly cleaned core saturated
with crude oil and FW (Swi ¼ 0.20), was put for spontaneous imbibition
at 60  C. The core was initially imbibed with formation water, FW, and
the resulting oil recovery reflected the initial wetting in the core. When
Fig. 4. Comparison of pH profiles during oil recovery tests at 60  C and 120  C. The cores oil recovery plateau was reached, the imbibing fluid was changed to LS
were successively flooded with FW-LS at a rate of 4 PV/D. At the end the flooding rate was brine to verify any LS EOR effects taking place by wettability alteration.
increased to 16 PV/D. The spontaneous imbibition results are presented in Fig. 6. With FW as
imbibing brine, the oil production increased slowly up to a plateau of
33.1 %OOIP after 32 days, a behavior consistent with a mixed core
regarding: (1) Initial pH with FW, which is important for the initial
wetting. After changing the imbibing fluid to LS brine, the oil recovery
wetting conditions, (2) pH gradient, which gives information about the
slowly increased with 4.7 %OOIP up to a production plateau at 37.8 %
degree of cation exchange at the mineral surfaces, and (3) size and rate of
OOIP, reached after 6 days of imbibing time. The slow production in-
the established pH gradient.
crease is consistent with wettability alteration, because it takes time for
The outcrop core B-21 was mildly cleaned after the oil recovery test
the new LS brine to imbibe into the core, change wetting, and imbibe into
and then used in a pH screening test. The core was successively flooded
previously un-swept oil-filled pores.
with FW-LS-FW, and the pH of the eluting brine was logged. The test was
performed at 60, 90 and 120  C, and the results are shown in Fig. 5.
No significant difference in the pH gradient was observed at 60 and 3.2. Relationship between pH, wettability and LS EOR effects
90  C. The pH changed from 6 to 9, i.e. ΔpH≈3. As expected, the pH
gradient at 120  C was somewhat smaller, ΔpH≈2.7. This can be Buckley and Morrow (1990) tested adhesion properties of 22 crude
explained by less desorption of Ca2þ ions from the clay at higher tem- oils onto silica surfaces as a function of brine composition and pH and
perature, as indicated by Eq. (5). The results from the pH screening test noticed remarkable similarities in the results. In the adhesion map, they
were compared with the oil recovery results in Table 6. observed characteristic pH values in the range of 6–7, above which,
The pH of the first eluted FW during the oil recovery tests at both 60 adhesion did not occur at different salinities, and they concluded that the
and 120  C was in the favorable region to induce adsorption of polar pH was the dominant factor for this behavior. It was also observed that
components, pH e 6. Furthermore, a distinct pH gradient during LS de-asphalted crude oil did not exhibit adhesion in the low pH range,
flooding promotes wettability modification and induces a large confirming that the pH effect on adhesion was primarily linked to
EOR effect. surface-active components in the crude oil. It should be noted, that the
The appropriate pH conditions observed in the oil recovery tests that surface charges of both clays and silicates are negative in the actual pH
triggered the increased oil recovery, were also reproducible in the oil-free range, but only clays have permanent negative charges. Similar results
system. Therefore, the pH-screening test can give valuable firsthand in- were recently confirmed by Didier et al. (2015) in adhesion studies of
formation about the potential for observing LS EOR effects for an actual crude oil and two different sands. At a given pH, it was observed that the
oil reservoir. However, if the reservoir system is already water-wet, a pH adhesion of oil increased by lowering the salinity, an observation, which
supported previous results showing increased adsorption of polar organic
molecules onto clay in a low salinity brine, compared to the adsorption in

Fig. 5. pH screening tests on core B-21. Mildly cleaned and 100% brine saturated core
was successively flooded with FW-LS-FW at a rate of 4 PV/D. pH in effluent samples was
Fig. 6. Spontaneous imbibition at 60  C on core C-2 with FW and LS as imbibing brines.
recorded. Tests were performed at 60, 90, and 120  C.

107
I.D. Pi~
nerez Torrijos et al. Journal of Petroleum Science and Engineering 156 (2017) 102–109

a high salinity brine (Burgos et al., 2002; Fogden, 2012; RezaeiDoust temperature. Therefore, the LS EOR effects were very similar, 9.2 and
et al., 2011). 8.7% of OOIP, respectively.
In a paper by Ehrlich et al. from 1974, alkaline water flooding for  Compared to previously published work on core material without
wettability alteration was evaluated for potential field application (Ehr- reactive plagioclase, the presence of reactive plagioclase/albite,
lich et al., 1974). A solution of NaOH was used as injection fluid, and they ≈30 wt%, contributed to a low initial pH of FW, 5.5–6.0, and to
were able to discriminate between EOR effects linked to wettability enhance the pH gradient bye2.3 pH units, during the LS flood both at
modification and to a decrease in interfacial tension, IFT. Compared to a 60 and 120  C.
NaCl-solution as injection fluid, an injection brine containing NaOH  The low initial pH was favorable for obtaining mixed wet conditions.
could increase the recovery up to 12–15% of the PV when using a light  Significant LS EOR effects can be observed at high temperatures if
crude oil with low AN. The EOR effect was mostly attributed to wetta- reactive plagioclase is present and the initial pH is low.
bility alteration because the decrease in IFT was quite small. As pointed  Wettability alteration was confirmed in a spontaneous imbibition
out by the authors, the challenge was to be able to transport the alkaline process at 60  C by switching imbibing brines from formation water
solution through the reservoir. The alkalinity will be consumed by: acidic to low salinity brine. The LS EOR effect was 4.9 %OOIP at 60  C.
gases, e.g. CO2 and H2S; and by precipitation of hydroxides, Mg(OH)2
and Ca(OH)2. According to the suggested LS Smart Water EOR mecha- Acknowledgements
nism for sandstone reservoirs (Austad et al., 2010; RezaeiDoust et al.,
2011), the alkalinity is created in the reservoir by ion exchange of The authors acknowledge TOTAL E&P Norge AS for financial support
reactive cations at active mineral surfaces as the HS brine is displaced by and for permission to publish this paper, and Jose Da Costa Ferreira for
the LS brine. No transportation of alkalinity through the porous medium his experimental support in the laboratory.
is therefore needed.
Recently, the first modelling of the LS EOR effect including a chemical
References
surface complexation model and an induced pH gradient for the wetta-
bility alteration process was presented by Brady et al. (2015). In the Aghaeifar, Z., Strand, S., Austad, T., Puntervold, T., Aksulu, H., Navratil, K., Storås, S.,
modelling, the possible release of fines from the clay minerals under LS Håmsø, D., 2015. The influence of formation water salinity/composition on the low
conditions and pH changes was discussed. Previous experimental results salinity EOR effect in high temperature sandstone reservoirs. Energy & Fuels 29,
4747–4754.
have also linked fines release to LS flooding and corresponding pH in- Aksulu, H., Håmsø, D., Strand, S., Puntervold, T., Austad, T., 2012. Evaluation of low
crease due to ion exchange in Berea sandstone containing kaolinite clay salinity EOR-effects in sandstone: effects of temperature and pH gradient. Energy &
(Vaidya and Fogler, 1992). Fuels 26, 3497–3503.
Allard, B., Karlsson, M., Tullborg, E.-L., Larson, S.Å., 1983. Ion Exchange Capacities and
It may seem like a contradiction that in a secondary flood with HS
Surface Areas of Some Major Components and Common Fracture Filling Materials of
brine, more water-wet conditions lead to a higher residual oil saturation Igneous Rocks. KBS Technical Reports, G€ oteborg, Sweden.
(Sor), but in a LS EOR flood, more water-wet conditions lead to a lower Austad, T., RezaeiDoust, A., Puntervold, T., 2010. Chemical mechanism of low salinity
water flooding in sandstone reservoirs. In: Paper SPE 129767 Prepared for
Sor. However, the principles for achieving a Smart Water EOR effect in a
Presentation at the 2010 SPE Improved Oil Recovery Symposium. Tulsa, Oklahoma,
tertiary flooding regime can be explained in the following way: A sec- USA, 24–28 April.
ondary water flood using HS brine, often represented by FW in laboratory Blum, A.E., 1994. Feldspars in weathering. In: Parsons, I. (Ed.), Feldspars and Their
experiments, displaces oil from the most accessible pores. Some oil-wet Reactions. The Netherlands. Kluwer Academic Publishers, pp. 595–630.
Brady, P.V., Morrow, N.R., Fogden, A., Deniz, V., Loahardjo, N., Winoto, 2015.
pores with high clay content are bypassed because the FW is already in Electrostatics and the low salinity effect in sandstone reservoirs. Energy & Fuels 29,
equilibrium with the formation and it is therefore not able to change 666–677.
wetting properties. By injection of LS Smart Water in a tertiary mode will Buckley, J.S., Morrow, N.R., 1990. Characterization of crude oil wetting behavior by
adhesion tests. In: Paper SPE/DOE 20263 Presented at the SPE/DOE Seventh
of course lead the Smart Water brine to follow the same path through the Symposium on Enhanced Oil Recovery. Tulsa, Oklahoma, April 22–25.
porous system, due to the viscosity being slightly lower than that of the Burgos, W.D., Pisutpaisal, N., Mazzarese, M.C., Chorover, J., 2002. Adsorption of
HS water. In addition, the Smart Water has the potential to imbibe into quinoline to kaolinite and montmorillonite. Environ. Eng. Sci. 19, 59–68.
Chou, L., Wollast, R., 1985. Steady-state kinetics and dissolution mechanisms of albite.
bypassed crude oil-containing pores by increasing the water wetness of Am. J. Sci. 285, 963–993.
the pore surface, mainly the clay minerals. Crude oil present in bypassed Didier, M., Chaumont, A., Thibaut, J., Bondino, I., Hamon, G., 2015. Contradictory trends
pores can only be mobilized by capillary forces. Therefore improved for smart water injection method: role of pH and salinity from sand/oil/brine
adhesion maps. In: Paper SCA2015-005 Presented at the International Symposium of
water wetness is crucial for obtaining positive capillary forces. The
the Society of Core Analysts. St. John's Newfoundland and Labrador, Canada, 16-21
spontaneous imbibition results in Fig. 6 confirmed wettability alteration August.
and increased capillary forces, by switching from FW to LS as imbibing Ehrlich, R., Hasiba, H.H., Raimondi, P., 1974. Alkaline waterflooding for wettability
alteration - evaluating a potential field application. J. Pet. Technol. 26, 1335–1343.
brines. A new bank of oil is mobilized due to improved microscopic
Fan, T., Buckley, J., 2007. Acid number measurements revisited - SPE 99884. SPE J. 12
sweep efficiency. Surely, the Smart Water will increase the water wetness (4) http://dx.doi.org/10.2118/99884-pa.
of the whole rock, and therefore some of the mobilized oil will be trapped Fogden, A., 2012. Removal of crude oil from kaolinite by water flushing at varying
due to increased capillary forces. If, however, the amount of oil mobilized salinity and pH. Colloids Surf. A Physicochem. Eng. Asp. 402, 13–23.
Friedman, G.M., Sanders, J.E., Kopaska-Merkel, D.C., 1992. Principles of Sedimentary
by Smart Water is higher than the trapping of oil by increased capillary Deposits: Stratigraphy and Sedimentology. Macmillan Publishing Company, New
forces, a net tertiary Smart Water EOR effect will be obtained. York, USA.
Gamage, P., Thyne, G., 2011. Systematic investigation of the effect of temperature during
aging and low salinity flooding of Berea sandstone and Minn. In: 16th European
4. Conclusions Symposium on Improved Oil Recovery. Cambridge, UK, 12–14 April.
Garrels, R.M., 1984. Montmorillonite/Illite stability diagrams. Clays Clay Miner. 22,
161–166.
Outcrop core material containing reactive plagioclase was tested for Kia, S.F., Fogler, H.S., Reed, M.G., 1987. Effect of pH on colloidally induced fines
LS EOR effects at 60 and 120  C using a crude oil with low acid number, migration. J. Colloid Interface Sci. 118, 158–168.
AN<0.10 mgKOH/g, and high base number, BN ¼ 1.8 mgKOH/g. The FW Madsen, L., Lind, I., 1998. Adsorption of carboxylic acids on reservoir minerals from
organic and aqueous phase. SPE Reserv. Eval. Eng. Febr. 47–51.
had a salinity of 100 000 ppm with a concentration of Ca2þ of 90 mM. Morrow, N.R., 1990. Wettability and its effect on oil recovery. J. Pet. Technol. Dec.
The LS brine contained pure NaCl with a salinity of 1000 ppm. The main 1476–1484.
results are summarized as follows: Oelkers, E.H., Golubev, S.V., Chairat, C., Schott, J., 2009. The surface chemistry of multi-
oxide silicates. Geochim. Cosmochim. Acta 4617–4634.
Pi~
nerez Torrijos, I.D., Austad, T., Strand, S., Puntervold, T., Wrobel, S., Hamon, G., 2016a.
 In the temperature range of 60–120  C, the pH gradient as the FW was Linking low salinity EOR effects in sandstone to pH, mineral properties and water
displaced with the LS brine, was not significantly affected by composition. In: Paper SPE 179625 Presented at the SPE Improved Oil Recovery
Conference. Tulsa, OK, USA, 11–13 April.

108
I.D. Pi~
nerez Torrijos et al. Journal of Petroleum Science and Engineering 156 (2017) 102–109

Pi~
nerez Torrijos, I.D., Puntervold, T., Strand, S., Austad, T., Abdullah, H.I., Olsen, K., EOR at interwell scale, Endicott field, Alaska. In: Paper SPE 129692 Presented
2016b. Experimental study of the response time of the low-salinity enhanced oil at the 2010 SPE Improved Oil Recovery Symposium. Tulsa, OK, USA, 24–28
recovery effect during secondary and tertiary low-salinity waterflooding. Energy & April.
Fuels 30, 4733–4739. Shiran, B.S., Skauge, A., 2013. Enhanced oil recovery (EOR) by combined low salinity
Pi~
nerez Torrijos, I.D., Puntervold, T., Strand, S., RezaeiDoust, A., 2016c. Optimizing the water/polymer flooding. Energy & Fuels 27, 1223–1235.
low salinity water for EOR effects in sandstone reservoirs - composition vs salinity. In: Springer, N., Korsbech, U., Aage, H.K., 2003. Resistivity index measurement without the
78th EAGE Conference & Exhibition 2016. Vienna, Austria, 30 May - 2 June. porous plate: a desaturation technique based on evaporation produces uniform water
Reinholdtsen, A.J., RezaeiDoust, A., Strand, S., Austad, T., 2011. Why such a small low saturation profiles and more reliable results for tight North Sea chalk. In: Paper
salinity EOR - potential from the Snorre formation?. In: 16th European Symposium Presented at the International Symposium of the Society of Core Analysts Pau,
on Improved Oil Recovery. Cambridge, UK, 12–14 April. France, 21–24 Sept.
RezaeiDoust, A., 2011. Low Salinity Water Flooding in Sandstone Reservoirs - a Chemical Strand, S., Austad, T., Puntervold, T., Aksulu, H., Haaland, B., RezaeiDoust, A., 2014. The
Wettability Alteration Mechanism. Faculty of Science and Technology. Stavanger, impact of plagioclase on the low salinity EOR-effect in sandstone. Energy & Fuels 28,
Norway: University of Stavanger. 2378–2383.
RezaeiDoust, A., Puntervold, T., Austad, T., 2010. A discussion of the low salinity EOR Vaidya, R.N., Fogler, H.S., 1992. Fines migration and formation damage: influence of pH
potential for a North Sea sandstone field. In: Paper SPE134459 Presented at the SPE and ion exchange. SPE Prod. Eng. Novemb. 325–330.
Annual Technical Conference and Exhibition. Florence, Italy, 19–22 September. Xie, Q., Liu, Q., Ma, D., Wu, J., 2012. Influence of brine composition on C/B/R
RezaeiDoust, A., Puntervold, T., Austad, T., 2011. Chemical verification of the EOR interactions and oil recovery in low permeability reservoir cores. In: The 33rd Annual
mechanism by using low saline/smart water in sandstone. Energy & Fuels 25, IEAEOR Conference and Symposium. Regina, Canada, 27–30 August.
2151–2162. Zavitsas, A., 2005. Aqueous solutions of calcium ions: hydration numbers and the effect of
Seccombe, J., Lager, A., Jerauld, G., Jhaveri, B., Buikema, T., Bassler, S., Denis, J., temperature. J. Phys. Chem. 109, 20636–20640.
Webb, K., Cockin, A., Fueg, E., Paskvan, F., 2010. Demonstration of low-Salinity

109