Sie sind auf Seite 1von 16

Journal of Dispersion Science and Technology

ISSN: 0193-2691 (Print) 1532-2351 (Online) Journal homepage:

Optimal design of pickering emulsions for heavy-

oil recovery improvement

Jungin Lee & Tayfun Babadagli

To cite this article: Jungin Lee & Tayfun Babadagli (2019): Optimal design of pickering emulsions
for heavy-oil recovery improvement, Journal of Dispersion Science and Technology, DOI:

To link to this article:

Published online: 12 Aug 2019.

Submit your article to this journal

Article views: 120

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

Optimal design of pickering emulsions for heavy-oil recovery improvement

Jungin Lee and Tayfun Babadagli
School of Petroleum Engineering, University of Alberta, Edmonton, Canada


This study focuses on the ability of complex colloidal solution to enhance heavy oil recovery by Received 5 February 2019
stabilizing a heavy oil-brine Pickering emulsion. After testing many different combinations of Accepted 26 July 2019
anionic surfactants, and cationic surfactants, and nano-particles, we formulated the best stability
options and created oil-in-water Pickering emulsions stabilized by silica, a cationic surfactant
Microemulsion formulation;
[dodecyltrimethylammonium bromide (DTAB)], and an anionic surfactant [alcohol propoxy sulfate emulsion stability;
(Alfoterra S23-7S-90)]. Then, we conducted various core flooding experiments in order to demon- surfactants and polymers;
strate the practical ability of the created emulsion system and observe its capacity to enhance oil nanofluids; cold heavy oil
recovery. Rate-dependency flooding tests were also conducted to determine the optimal flow rate recovery; dynamic flooding
(shear rate) required for heavy oil production through emulsification for different permeability experiments
media. Various permeability media were selected in this research to test the dynamic nature of
the solution, including slim tube sandpack of which provides more challenging conditions for
emulsification due to its length (1 m). We conducted the slim tube sandpack flooding experiment
was conducted at the optimal rate to confirm in-situ emulsion generation and to support the
potential use of the chemical combination in the heavy oil industry.


Introduction reduction and mobility ratio improvement,[24–29] SP (surfac-

tant-polymer) flooding by IFT reduction and mobility
Thermal methods predominantly applied for heavy oil
improvement,[30,31] ASP (alkaline-surfactant-polymer) flood-
recovery have many drawbacks, such as high consumption ing by IFT reduction, mobility improvement.[32,33]
of fresh water, energy, and CO2 emissions. These disadvan- Chemical flooding may become ineffective if the chemical
tages lead the industry to search for lowered cost and more is adsorbed on the rock surface or has reactions with the
efficient techniques such as non-thermal chemical meth- rock surface, leading to emulsion stability loss, particularly
ods.[1,2] Non-thermal chemical application methods for in the case of alkaline and surfactant flooding. Therefore, it
heavy oil include microemulsion (nanoemulsion) flooding is essential to generate an emulsion that is sufficiently stable
by in-situ emulsion generation and oil viscosity reduc- with little mechanical shear force in order to break the
tion,[3–9] surfactant flooding by IFT reduction and capillary heavy oil into fine droplets into the water phase, thereby
number increase,[10] polymer flooding by sweep efficiency achieving a significant viscosity reduction and improvement
improvement,[11–16] alkaline flooding by IFT reduction via to oil flow.
in-situ surfactant generation,[8,17–21] nanofluid (or nanopar- Conventional emulsions stabilized with only surfactants or
ticle) flooding by wettability alteration or Pickering emulsion polymers are known to be thermodynamically unstable. However,
generation,[22,23] AS (alkaline-surfactant) flooding by IFT emulsions, which have been stabilized by surface-active colloidal

CONTACT Tayfun Babadagli School of Petroleum Engineering, University of Alberta, Edmonton T6G 1H9, Canada.
Color versions of one or more of the figures in the article can be found online at
ß 2019 Taylor & Francis Group, LLC

particles called Pickering emulsions, are kinetically stable. The Table 1. SARA results.
irreversible adsorption of the colloid particles at the interface Components Mass recovered (wt.%)
between oil and water gives Pickering emulsions kinetic stability. Saturates 29.36
This mechanism can provide a barrier to the coalescence of drop- Aromatics 25.65
Resins 29.51
lets, indicating stability loss.[34,35] Asphaltenes 14.78
This work studies various types of cost-effective oil-in-
water Pickering emulsion systems formed with heavy oil of
11.74 API, in which the oil-water interface is structurally
stabilized by a colloidal layer. In the past, there have been
attempts to use Pickering emulsions stabilized with nanopar-
ticles to enhance oil recovery,[36–38] as well. Peng et al.[39] Figure 1. Dodecyl trimethyl ammonium bromide (DTAB).
and Suleimanov et al.[40] give an extensive review on the
nanoparticle application in enhanced oil recovery. The emul- commonly studied nanoparticles for EOR applications such
sion stability of a Pickering emulsion is dependent on the as alumina, zirconia, and nickel. Yoon et al.[41] determined
size, packing order, and wettability of the nanoparticles at the specific surface area of Ludox CL-X to be 130 m2g-1
the interface, and studies report that Pickering emulsions using the Brunauer-Emmet-Teller method and TEM images
display stronger stability when hybridized with surfac- are available in the cited paper. The cationic surfactants
tants.[41] Various types of colloidal layers were extensively DTAB (dodecyltrimethylammonium bromide), was obtained
investigated for their potential in stabilizing the heavy oil- commercially, and for surfactant, A noncommercial anionic
in-water interface as a result of the van der Waals attraction alcohol propoxy sulfate (Alfoterra S23-7S-90) was used. The
and electrostatic attraction; negatively charged nanofluids, surfactant was chosen based on its optimal synergy effects it
cationic, and anionic surfactants were used to create a struc- displayed with heavy oil from the previous literature at low
turally stable colloidal layer. To investigate the practical salinity conditions.[42] All listed chemicals were used without
applicability of the colloidal dispersions in enhanced heavy further purification.
oil recovery, sandpack flooding and core flooding tests were
conducted. Extracted emulsions were then visualized under
the transmitted light microscope to confirm their stability Oil and water
against flocculation and coalescence.
Heavy dead crude oil used was obtained from a field in
The results revealed that some colloidal dispersion
western Saskatchewan, Canada. At room temperature
formed with low-priced silica and surfactants can generate
(25  C), density and viscosity of this oil was measured to
and stabilize oil-in-water emulsions. The oil-water interface
be 981 kg/m3 and 4,812 cP, respectively. The API gravity of
formed with the complex colloidal layer can increase the
structural stability of the emulsion and help them maintain the oil was 11.74 . The heavy oil’s mean total acid number
their stability against coalescence or flocculation. (TAN) was measured to be 3.2 mg KOH/g. The acid number
Additionally, sandpack flooding and core flooding results is measured to determine the amount of soap that can be
also showed that the emulsions produced in-situ in the core generated; however, the amount of acid may not be corre-
can efficiently flow through the sandstone pores as a result lated to acid number possibly depending on the measure-
of low IFT and low viscosity. These findings are critical in ment methods. Some components may not be able to
reducing the overall cost of chemical EOR methods for generate soap and other application conditions may have a
heavy oil recovery under cold conditions. This paper intro- negative influence on the soap generation.[43] SARA analysis
duces a potential chemical combination selected based on its of the heavy oil is available in Table 1. The pH of the
ability to create a complex colloidal oil-water interface with 2.54 wt.% brine sample (NaCl) used throughout the experi-
the ability to stabilize an oil-in-water emulsion. The chemi- ments is 6.98. Based on its optimum compatibility with
cals and their potential optimal combinations were tested Alfoterra S23-7S-90 in creating a heavy oil-in-water emul-
for their application in cold water flooding for heavy-oil sion,[42] a 2.54 wt.% salinity brine sample was selected.
recovery through sandpack flooding, core flooding, and slim
sandpack flooding. We expect that the stable nanoscale
Pickering emulsion system created with low cost chemicals
developed in this work can find practical applicability in the Design of the pickering emulsion
heavy oil industry.
Commercially available inorganic nanoparticles such as
hydrophilic silica can be hydrophobized in situ with a con-
Experimental ventional surfactant such as DTAB, the structure of which is
shown in Figure 1.[41,44] Through the electrostatic inter-
action, the charged nanoparticles can adsorb oppositely
Silica nanofluid solution, Ludox CL-X from Sigma-Aldrich, charged surfactant; the surfactant monolayer formation can
with an average particle size of 22 nm was chosen (45 wt. % then increase the hydrophobicity of the particle surfaces,
nanoparticle suspension in H2O) because of its environmen- lead to surface-active particles. In-situ hydrophobization of
tal-friendliness and low cost when compared to other originally hydrophilic particles can allow particle attachment

Table 2. Solution A.
Solution A-1 Solution A-2 Solution A-3 Solution A-4
Silica (wt%) 1.00 1.00 1.00 1.00
Cationic DTAB (wt%) 0.00 0.02 0.04 0.06
Anionic Alfoterra (wt%) 0.50 0.50 0.50 0.50

at the oil-water interface,[34,45] and modifying the wettability

of colloidal particles in the liquid media can help prepare
particle-stabilized emulsions thus enhancing adsorption at
the liquid-liquid interface.[46]
Cationic surfactant can also help adsorb anionic surfac-
tant onto a negatively charged surface like silica, due to the
formation of ion pairs.[47] The anionic surfactant selected Figure 2. Core flooding set up.
for this test was Alfoterra as it has given optimal results in
our previous chemical screening test.[42] Mixing anionic sur- concentrations ranging from 0 to 0.06 wt% Table 2 was then
factant and cationic surfactant was also shown to be a novel incorporated into the solution to both observe its ability to
surfactant methodology leading to ultralow interfacial ten- form a stable emulsion and to find the optimal concentra-
sion (IFT) and high oil solubilization.[48] Mixing oppositely tion of DTAB in terms of costs and application (Table 2).
charged particles and surfactant is an important mechanism 0.5 wt.% anionic surfactant (Alfoterra S23-7S-90) was added
in the synergistic stabilization of emulsion drops. to effectively stabilize the colloidal solution, then oil was
Therefore, silica, DTAB, and Alfoterra S23-7S-90 were added and agitated manually (brine and oil ratio 1:1, 5 ml
used to create a cost-effective complex colloidal solution giv- each). The solution with the most optimal concentration of
ing optimal recovery results in various experiments, of DTAB was then used for various core flooding experiments
which include the glass tube experiment, sandpack flooding to test its ability to enhance heavy oil recovery.
experiment, slim tube sandpack flooding experiment, and
the Berea sandstone core flooding experiment.

Sandpack flooding
Zeta potential
Unconventional heavy oil and bitumen reservoirs in several
The zeta potential is a key indicator of the stability of col-
countries globally such as Canada and Venezuela, are fam-
loidal dispersions and is commonly used to quantify the
ously known for having huge resources of unconventional
magnitude of the charge. The zeta potential of silica particle
heavy oil and bitumen. These resources are characterized by
dispersions in aqueous phase (NaCl ¼ 1 mmol) was meas-
high porosity and high permeability; therefore, a high poros-
ured, as well as the zeta potentials of silica dispersion with
ity, high permeability sandpack was employed to simulate
the addition of DTAB, and zeta potentials of silica with the
addition of both DTAB and Alfoterra S23-7S-90 at pH ¼ 7 conditions of a typical heavy oil reservoir of which is a good
using a Malvern Zetasizer Nano ZS instrument employing a candidate for chemical flooding.[2] The core flooding appar-
dip cell. A minimum of two measurements were conducted atus was set up with an injection pump (Teledyne ISCO syr-
for each solution to confirm the validity of the results. 1% inge pump), core holder, and measuring cylinder or vials to
of heavy oil was added to the sample in order to provide collect the sample (Figure 2). The core holder was tightly
the turbidity required to measure the zeta potential and packed with heavy oil saturated sand of 11.74 API with
investigate the possible interaction between the oil and sandpack. Porosity was measured to be 35% prior to oil
chemicals used in the study. We selected a 1 mmol brine saturation. In the sandpack preparation, to create uniform-
concentration due to the limitation of the equipment to ity, the 500 - 800 mm diameter sands were first saturated
measure samples with higher electrolyte concentrations such with heavy oil. Saturated sands were then placed into the
as the 2.54 wt% NaCl brine sample used for our glass tube coreholder (1.5 in. diameter and 5.4 in. height) layer by
experiment, sandpack flooding(s), core flooding, and glass layer using approximately 5 g of saturated sands each time,
bead visualization in this study. Zeta potential measurements next a steel rod was used to apply force (three manual taps
were conducted in order to ensure the stability of the col- each time) to tightly pack the sands thus minimizing the
loidal solution as an injectant for the sandpack flooding, residual air in the system resulting in maintaining sandpack
slim tube sandpack flooding, and Berea sandstone core uniformity. The permeability of the sandpack (before oil sat-
flooding experiments. uration and using water) was measured to be 10 Darcy.
To measure the porosity and permeability of the sand-
pack and Berea sandstone cores, this measurement proced-
Glass tube concentration test
ure was followed:
1.0 wt% Ludox CL-X (silica) was mixed with a brine sample
of 2.54 wt% salinity (pH ¼ 6.98), and the mixture was cre- 1. The sandpack or core holder was vacuumed using a
ated with light manual agitation. DTAB of varying vacuum pump.

Figure 4. Top view of the glass bead model.

Figure 3. Schematic diagram of glass bead visualization setup.

2. Overburden pressure was applied to the annulus of the

sandpack/coreholder using a syringe pump; 100 psi for
sandpack and 200 psi for the core sample.
3. Water was injected to saturate the sandpack/core from
one end at a flow rate of 0.1 cc/min. A pressure trans-
ducer was connected to the injection end of the core-
holder and the other end of the coreholder was left at
atmospheric pressure after the pressure build-up.
4. After the core was saturated with water, the injection
rate was increased to 1 cc/min. The pressure increased
gradually and was stabilized after some time (a few
minutes for the sandpack and 30 min. for the
core sample).
5. The water injection rate was increased to 2, 3, and 4 cc/
min and kept for 30 minutes, with the pressure Figure 5. Schematic diagram of Berea sandstone core set up.
being recorded.
6. The porosity was calculated from the water volume the Carman-Kozeny equation, and for the 3 mm glass beads,
saturated in the sandpack/core and permeability was a permeability of 7,200 D can be estimated.[49] Colloidal
determined by using Darcy’s law. The porosity value solution was injected from the injection port of the glass
was verified by calculating the percent volume of oil bead model using an ISCO syringe pump (Figure 4), and in
used in the sand pack after sand pack preparation. order to both minimize the impact of the gravity drainage
process and mimic the conditions under which the core
2-D Glass bead visualization flooding test was conducted, the glass bead model was
placed horizontally (Figure 3).
A 2-D visual glassbead model (7 x 14 cm) was created, made
up of a single layer of sintered micro-scale glass beads. The
silica glass beads, approximately 3 mm in diameter, were Berea sandstone core flooding
packed and glued between two 9 mm Plexi-glass plates using It must be taken into consideration that the effectiveness of
an Epoxy glue, and the processes were recorded with a chemical flooding is reduced significantly during core flood-
high-speed camera to obtain visual data at pore scale. This ing when there is channeling;[50,51] therefore, in this Berea
experiment was conducted to investigate the dynamics of in- sandstone core flooding study, an overburden pressure of
situ emulsion generation at pore scale leading to the forma- 100 psi, which adds to the pressure measured in the core,
tion of residual oil saturation, and Figure 3 shows the sche- was applied to the core to control the possible channeling
matic diagram of the glass bead visualization setup. This that can be created during the core flooding process. The
experiment was conducted to visualize and confirm in-situ Berea sandstone core experiment set up is displayed in
oil-in-water emulsion generation under conditions closely Figure 5. The porosity and permeability of the cores used in
matching sandpack flooding conditions (Figure 2). The flow this study were on average, 19.1% and 273.9 mD, respect-
rate selected for this glass bead experiment was 0.2 ml/min ively. Little variation in the porosity and permeability
(0.88 ft/day), which was found to be the optimal flow rate in between the cores was observed as the fresh core sample cut
previous rate-dependency sandpack flooding experiments from the same block of Berea sandstone was used for each
(see details in section Sandpack Flooding Oil Recovery). The flooding experiment, and a fresh core sample was used for
porosity of the glass bead model used in this study was each experiment. 93.2 – 93.6 wt.% of chemical compounds
35%, of which was calculated by deducting the volume of found in the Berea sandstone cores were silica with alumina
glass beads from the total volume of the glass bead model constituting 3.6–4.6%. Other compounds include Fe2O3
and dividing the calculated value by the total volume of the (0.30–0.35%), CaO (0.25–0.65%), MgO (0.08–0.15%), Na2O
glass bead model. The permeability can be calculated using (0.75–0.85%), and TiO2 (0.1% max). Berea sandstone core

Figure 6. Slim tube experiment setup.

Table 3. Zeta potential results.

Brine Solutions Zeta potential STD. Deviation
lmM Silica (1%) þ heavy oil (1%) 106 mV 7.8 mV
Silica (1%) þDTAB (0.04%) þ heavy oil (1%) 11 mV 6.53 mV
Silica (1%) þDTAB (0.04%) þ Alfoterra S23-7S-90 (0.5%) þ heavy oil (1%) 92 mV 8.95 mV

flooding was conducted to remove any possible effects of Results and discussion
channeling and to test the capacity of the colloidal solution
in a setting with a much lower permeability and porosity Zeta potential measurements
than that of the aforementioned sandpack flooding study. Zeta potential (Table 3) is measured to characterize the dif-
We focused on rate-dependency experiments in the case of fuse part of the double layer, of which controls particle
Berea sandstone experiments considering the possible impact interaction.[41] At pH ¼ 6.98, adding cationic surfactant
of viscous forces, capillary forces, and the interaction time DTAB in the presence of silica (1 wt.%) and heavy oil
between the fluids and substrates on emuslification on emul- (1 wt.%) increased the absolute magnitude of the zeta poten-
sification The effect of flow rate on emulsification is further tial of the solution by 95 mV rendering the solution incipi-
elaborated in the “Results and Discussion” section (subsec- ently instable at -11 mV; a strong attraction between
tion of “Sandpack Flooding Oil Recovery”). The core sam- positively charged DTAB and negatively charged silica nano-
ples used were 1.5 inch diameter and 5.4 inch in height. particles can result in the aggregation of nanoparticles.[52]
Colloids with high zeta potential, whether negatively or posi-
Slim tube sandpack flooding tively charged, are electrically stabilized whereas colloids
with low zeta potentials coagulate or flocculate easily due to
A 1 m length and 1.5 cm diameter slim tube was used for the attractive forces that exceed the repulsion. The addition
the slim tube sandpack flooding experiment in order to test of an anionic surfactant in presence of silica (1 wt.%), DTAB
the in-situ emulsification capacity of the colloidal solution. (0.04 wt.%), and oil (1 wt.%) lowered the zeta potential of
This condition can better mimic real reservoir settings over
the complex colloidal dispersion to -92 mV thus stabilizing
a short sandpack flooding test (Figure 2), as well as works
the colloidal solution. This observation indicates the import-
to test the in-situ emulsification potential of colloidal solu-
ance of using an anionic surfactant to stabilize the colloidal
tion A-3 (Table 2). The colloidal solution was prepared with
dispersion at pH ¼ 6.98. Al-Anssari et al.[53] also reported
NaCl (2.54% salinity) before being injected into the system
that adding cationic surfactants in a solution with silica can
and the reservoir temperature was kept at a room tempera-
weaken the stability of solution, but addition of anionic sur-
ture (25  C). The 500–800 mm diameter sands were satu-
rated with heavy oil before being placed inside the tube, factant can improve its stability. Adsorption of anionic sur-
similar to the way the short sandpack experiment was pre- factant onto a negatively charged surface like silica is
pared, but with a longer steel rod appropriate for packing enhanced in the presence of cationic surfactant due to the
the tube to reduce the possible air trapped in the system. formation of ion pairs.[47]
Figure 6 displays the slim tube experiment setup; sand mesh
screens of the appropriate size (<400 mm) were inserted into Stabilized emulsions of an anionic surfactant alfoterra
the both ends of the slim tube to prevent possible sand pro- S23-7S-90 and silica nanoparticles hydrophobized
duction and the tube was plugged with rubber stoppers to with DTAB
prevent potential solution leakage. Two different flow rates
were employed to study the impact on oil recovery and in- Figure 7 displays the glass tube results of DTAB stabilized
situ emulsion generation and produced samples were then emulsions. Sample A-1 is an emulsion stabilized by silica
collected and visualized under a microscope. and anionic surfactant without any addition of cationic

Figure 7. Emulsion with DTAB results.

Table 4. DTAB emulsion evaluation.

A-1 A-2 A-3 A-4
Formation of emulsion (IFT reduction)    N/A
Stability 1 hour:  1 hour:  1 hour:  1 hour: N/A
5 hours: x 5 hours:  5 hours:  5 hours: N/A
24 hours: x 24 hours: x 24 hours:  24 hours: N/A (Severe coating)
Precipitation x x x x

surfactant (DTAB); stability loss was observed several optimal concentration of DTAB, together with injecting a
minutes after the initial manual agitation, and after 150 min, stable colloidal solution into the porous media, this adsorp-
the emulsion separated into two distinct phases. Sample A-2 tion issue can be minimized. It took over 24 hours for the
with a 0.02 wt % DTAB concentration displayed better sta- A-3 emulsion to completely lose stability, and creaming at
bility than sample A-1, but a thin unidentifiable white layer the oil and water interface was observed. The aqueous phase
was formed at the oil and brine interface after 4 h. The was observed to be light brown in color; lighter than that
white layer, caused by creaming, could be a result of the aqueous phase of A-2 with less DTAB concentration.
DTAB’s aggregation due to its reaction with salt in the brine The turbid nature of the samples is observed when adding
sample;[45] creaming is one of the indicators of emulsion DTAB, possibly due to the strong attraction between nano-
stability loss, and can be observed at the interface between particles and micelles leading to the micelle-mediated aggre-
the oil and aqueous phase. Creaming can be prevented by gation of nanoparticles.[52] Samples A-1 to A-4 are shown in
reducing the density difference between the two phases or Table 4.
by homogenizing the solution more vigourously. Sample A-
3 with 0.04 wt % DTAB concentration showed the most
Emulsion stability
optimal results out of all the samples tested on forming a
stable emulsion. Sample A-4 with 0.08 wt % DTAB could A Winsor type 4 emulsion sample created with the A-3
not be observed with the naked eye glass tube tests due to solution (Figure 7) was visualized under the microscope to
severe coating of the inner surface of the glass tube, caused observe the emulsion stability. It was observed that the par-
by the attraction forces existing between the cationic surfac- ticles remain consistent in size and distribution for 30 min
tant and the negatively charged glass surface. Cationic showing little-to-no indication of emulsion stability loss.
adsorption on silica is a common problem in real reservoir However, larger sized oil droplets began to emerge 30 min
situations, possibly caused by high concentration of a cat- after the injection of the emulsion sample into the glass
ionic surfactant leading the emulsion sample to coat the sil- chamber, due to the coalescence of oil droplets (Figure 8).
ica surface due to attraction forces; however, with an After some loss of emulsion stability at the 30 -min mark,

Figure 8. Microscopic visualization results.

Figure 9. Particle distribution of the Winsor type 4 emulsion samples (a): Immediately after (b): 3 hours later.

the emulsion appeared to remain stable, and no visible sign (2.54 wt.% NaCl) into the sandpack at 1 ml/min, oil recovery
of emulsion stability loss was observed until after 3 h passed, was observed to be 0.1% of OOIP (Figure 10).
as shown in Figure 9. Secondary colloidal dispersion flooding using A-3 after
low-salinity water flooding was then conducted at the same
injection rate (1 ml/min), leading to oil recovery of the ori-
Sandpack flooding oil recovery ginal oil in place of 13% (Figure 11), which is equivalent to
the oil recovery by colloidal dispersion flooding conducted
Low-salinity waterflooding is often appealing because of the as the primary oil recovery method (Figure 12). However,
low-cost and relatively simple design compared to chemical when the flow rate was reduced to 0.2 ml/min (0.88 ft/day),
EOR techniques.[54] A sandpack test was conducted to inves- oil recovery improved significantly by 27% for primary A-3
tigate the effect of low-salinity water flooding on heavy oil sandpack flooding (Figure 13). Flow rates similar to 0.2 ml/
recovery and compare its performance to sandpack flooding min, such as 0.19 ml/min and 0.3 ml/min, were also selected
with the A-3 solution. After injecting 1 PV low salinity brine for further rate-dependency sandpack flooding to observe

Figure 10. Produced samples from low salinity water flooding at 1 ml/min.

Figure 11. Produced samples from colloidal dispersion flooding after low salinity water flooding at 1 ml/min.

Figure 12. Produced samples from colloidal dispersion flooding at 1 ml/min.

Figure 13. Produced samples colloidal dispersion flooding at 0.2 ml/min.

lower oil recovery was observed. Therefore, it is essential to

find an adequate flow rate that can give the colloidal solu-
tion enough time to contact and interact with the oil and
not lose its stability in the process. Additionally, it must also
generate enough shearing force, of which is necessary for
emulsification and plays an equivalent role to manual agita-
tion applied in glass tube experiments.
It was also observed that in the 0.2 ml/min, 0.3 ml/min,
Figure 14. Sandpack flooding oil recovery at various rates. and 0.19 ml/min sandpack flooding tests, oil production was
observed to decrease after around 1/6 of 1 PV solution
the emulsion sensitivity to flow rates. Sandpack flooding injection. Change in produced fluid types at different stages
conducted with aforementioned flow rates all displayed simi- of production is plotted in Figure 15a. However, it should
lar oil recovery results of approximately 20%. All the sam- be noted that emulsion types indicated in the plot are based
ples observed in Figures 10–13 were collected in the 1 PVI on the produced samples after they have been agitated to
time frame at every 0.16 PVI interval. form Winsor type 4 emulsions and are therefore not an
Figure 14 further illustrates the oil recovery results from indication of the separated emulsion phase (aqueous phase)
the rate dependency sandpack flooding tests, with results of the Winsor type 1 samples before the agitation. Extra agi-
showing that appropriate flow rate is an important param- tation was required to observe the nature of the emulsions
eter in colloidal dispersion flooding. When the duration of reflect the condition of the produced samples immediately
the experiment (or contact time between the solution and after the collection and not after the phases have already
oil) was very short (1 ml/min) or too long (0.04 ml/min), been separated over time due to gravity. The curves in

Figure 15. (a) Heavy oil chemical flooding type curve generated from sandpack flooding results using an optimal flow rate (0.2 ml/min). (b) Heavy oil chemical
flooding type curves generated from sandpack flooding results (a: 0.04 ml/min, b: 0.2 ml/min, c: 0.6 ml/min, d: 1 ml/min).

Figure 15b display that at an optimal flow rate (0.2 ml/min), The possible loss of colloidal stability of A-3 solution in
oil-in-water emulsion (with a high concentration of oil) the sandpacks can be due to various factors including
period is longer and leads to higher oil recovery. However, change in brine pH, of which is normally incremental dur-
at low recovery flow rates, oil-in-water phase (with low con- ing the oil production process,[57] change in water con-
centrations of water) period appears to be longer, leading to tent,[58] loss of equilibrium ratio of brine and oil, due to the
higher water (brine) production. Despite the apparent dom- rapid brine production, and surfactant adsorption to the
inance of the oil droplets (dark yellow) due to the high con- rock surface. In EOR, surfactant adsorption onto the surfa-
centration of oil in Figure 15a, the viscosity of the produced ces of the pore space is a factor that can result in great
sample was measured to be lower than the original oil phase chemical loss, leading to decreased system performance.
itself, showing that formed emulsion is not a water-in-oil Therefore, surfactant adsorption on the surfaces should be
emulsion but rather an oil-in-water emulsion with a high taken into consideration when designing a complex colloidal
concentration of the oil phase. One of the most important solution system. However, surfactant adsorption may not be
characteristics of a water-in-oil emulsion is its high viscosity the biggest contributing factor in A-3 stability loss in this
in comparison to the original oil or water phase, due to the study. Supra-CMC mixture can diminish the adsorption of a
droplet crowding effect.[55,56] Despite the high concentration surfactant with a lower concentration than the main surfac-
of the dispersed phase, oil-in-water emulsion production tant with a higher concentration.[59]
could be achieved with the colloidal solution used. Figures Another possible reason for the solution stability loss is
15a and 15b display the relationship between cumulative oil the loss of equilibrium ratio of brine and oil due to the
recovery, WOR, and emulsion types. Produced emulsions rapid production of brine (potentially due to channeling).
were categorized into four types: I is oil and an oil-in-water Emulsions formed with nanofluids can be highly sensitive to
emulsion with an oil concentration higher than 95%; II is an the volume ratio of water to oil.[46] In the glass tube experi-
oil-in-water emulsion with an oil concentration higher than ments, the colloidal solution was shown to be stable when
70%; III is defined by an oil-in-water emulsion with an oil the brine to oil ratio was 1:1. Due to the production of
concentration of 50 – 70%; IV is characterized by an oil-in- brine, the equilibrium ratio of brine and oil can be altered,
water emulsion with an oil concentration less than 50%. leading to deterioration of system stability. These

Table 5. Microscopic visualization of oil droplets in the emulsion phase.

Flow rate Immediately after 30 days later



0.19 ml/min

aforementioned factors could contribute to the gradual loss that the emulsion phase did not completely lose its stability.
of production and therefore, solutions to these issues must Liquids of different properties can commonly be separated
be discussed. In this study, possible exaggerated production by gravitational force. The principles governing gravity sep-
of brine due to channeling is further discussed and studied aration of oil can be described by Stoke’s law (Equation 1).
by conducting Berea sandstone core flooding experiments As defined by Stoke’s law, oil droplet rise can be affected by
while simultaneously applying overburden to reduce the change in oil and water density and viscosity or the change
effect of channeling (Figure 5). in oil droplet size:
In order to observe the change in the emulsion phase
and determine its stability, produced emulsion phases were ðdw do ÞD2
observed under the microscope immediately after the flood- V¼g (1)
ing and then for 30 days. During those 30 days, some separ-
ation between the oil and water phases by gravity separation where V ¼ oil droplet rise rate, g ¼ acceleration due to
was observed; however, the aqueous (emulsion) phase did gravity, dw ¼ water density, do ¼ oil density, D ¼ oil drop-
not become transparent even after 30 days, which implies let diameter, and u ¼ water viscosity.

Figure 16. Zoomed in view of the glass bead model.

Figure 17. Berea sandstone core flooding results.

In this study, change in oil droplet size over time was whereas, for the 0.6 ml/min and 0.19 ml/min cases, the oil
investigated under the Axiostar plus transmitted-light micro- droplets appear to be black in color. This is due to the
scope. It is shown that 30 days later, the number and size of modification in the micrograph illumination setting, which
oil droplets decreased significantly as the oil separated fur- ultimately enhanced visualization in the case of emulsion
ther from the aqueous phase (emulsion phase). It should be samples from the 0.6 ml/min and 0.19 ml/min sandpack
noted that the samples visualized were collected directly flooding. The new method essentially improved the micro-
from the microemulsion phase of the produced Winsor type scopic recognition and visual definition of the oil droplets.
1 samples at different time intervals (immediately after and
30 days later) for visualization purposes (Table 5). The
2-D Glass bead visualization (in-Situ
recovery of the original oil in place for the injection of col-
emulsion generation)
loidal dispersion was then calculated by measuring produced
oil volume. It should be noted that there is a slight differ- Figure 16 shows the dynamic oil-in-water emulsification
ence between the color of oil droplets in the produced emul- process observed in the glass bead model. White spot indi-
sion samples of 1 ml/min, 0.2 ml/min sandpack flooding, cate glass beads that are pressed against the wall, due to the
and that of 0.6 ml/min, 0.19 ml/min. The color of oil drop- slight variation in size of between the glass beads used to
lets for the produced samples of 1 ml/min, 0.2 ml/min sand- create the model. A 3 mm diameter was consistently used,
pack flooding may appear to be yellow or light brown, but there were some inevitable erroneous bead diameters

yielded the poorest performance (1.7% recovery).

Increasing the amount of injected colloidal dispersion from
1 PV to 2 PV with initial injection at 0.01 ml/min (1 PV)
and 0.03 ml/min (1 PV) did not show any significant
enhancement in oil recovery (14% recovery) (Figure 17),
implying the significance of flow rate on the oil recovery,
especially at the initial stage.
Low flow rates such as 0.01 ml/min can be detrimental to
oil recovery as it can lead to colloidal dispersion stability
loss in the core before being able to achieve in-situ emulsion
generation. Flow rates had to be maintained at a low rate
for the sandstone core flooding experiments in order to con-
trol the pressure, of which increased greatly at times due to
the tight structure of the sandpack cores (low permeability
and porosity).

Slim tube sandpack flooding experiments

Two flow rates of 0.3 ml/min and 0.03 ml/min were selected
Figure 18. Visualization of emulsions.
for the slim tube sandpack flooding tests to observe the pos-
sible in-situ emulsion generation with the complex colloidal
solution (A-3) in a slim tube (Figure 18). 1 PV of the solu-
leading to some uneven lines in the glass bead sample
tion was injected in each case. The samples shown in
model. The black parts represent glass beads covered with
Figures 19 and 20 were collected at 0.1 PVI interval. The
heavy oil.
slim tube experiment results confirmed the capacity of the
Prior to solution injection, the glass beads were saturated
A-3 solution to generate emulsion at different flow rates
by injecting heavy oil into the model. Then, the solution
(Figures 19 and 20), despite challenges such as length of the
was injected into the injection port as displayed in Figure 4.
tube and the slow flow rate selected for the slim tube flood-
Produced oil and emulsion were then collected into a cylin-
ing. Both flow rates (0.3 ml/min, 0.03 ml/min) gave similar
der (Figure 3). Once the solution was injected, emulsion
oil recovery results of  30%. Through the microscopic
generation was observed by the obvious color change of the
visualization of the emulsion phase of produced samples,
black part (saturated with oil) of the glass bead sample.
emulsion formed at the late-mid production for the 0.3 ml/
Some oil saturated black parts begin to lighten in color, min case, which is similar in terms of the oil droplet size
changing from black, to dark brown, and even continuing and distribution to the emulsion generated at the early-mid
onto a light brown before turning into a transparent yellow- production stage for the 0.03 ml/min case (Figure 18). This
ish color, indicating the remaining brine after the emulsion observation is possibly due to the emulsion stability loss
and oil displacement (Figure 16). Previously, glass tube with the low flow rate. Emulsion samples observed in Figure
experiments have displayed that when oil emulsifies with 18 were extracted from the produced samples (aqueous
water, the color of the mixture becomes brown (Figure 7). phase) without any further agitation, as the produced emul-
The color change observed in this experiment confirms the sions did not initially form Winsor type 4 emulsions. The
capacity of solution A-3 to both generate and displace emul- relative emulsion stability loss of the 0.03 ml/min case did
sion, eventually leading to enhanced oil recovery. Yellow not seem to carry a significant impact on the final oil recov-
LED light, used behind the glass sample to improve the ery results from the slim tube experiment. The resilience of
visualization capacity, is responsible for the amber color of A-3 solution was validated in this slim tube experiment pos-
the remaining brine, of which should usually appear; as sibly implying that application of the A-3 solution in a
white-transparent. much larger setting such as a real reservoir would be effect-
ive; however, it must be noted that it is possible for the
emulsion phase with slow flow rates (slight loss of emulsion
Berea sandstone core flooding
stability) to have a more significant impact on the final
Berea sandstone core flooding experiments were conducted recovery during a larger scale operation, as the objective is
with the colloidal dispersion solution (A-3) consisting of to prevent oil and water phase separation before production.
1 wt.% silica nanoparticles (22 nm Ludox CL-X), 0.04 wt.%
DTAB, and 0.5 wt.% Alfoterra S23-7S-90 (Table 2). Colloidal
dispersions, when injected at 0.02 ml/min þ 0.04 ml/min,
were experimentally confirmed to give the best recovery This study showed that DTAB can significantly improve the
results (22% recovery) at 1 PVI. For dual flow rate experi- stability of a Pickering oil-in-water emulsion system formed
ments, the flow rate was changed when the injection reached with heavy oil of 11.74 , 2.54 wt% brine (NaCl), silica nano-
0.5 PVI. Injecting the colloidal dispersion at 0.01 ml/min particles, and anionic surfactant. Previous studies by Lee

Figure 19. Results from 0.3 ml/min.

Figure 20. Results from 0.03 ml/min.

and Babadagli [42] confirmed that hydrophilic surfactants Oil Conference Canada, Calgary, AB, Canada, June 12–14,
can create heavy oil-in-water emulsions more effectively 2012.
[2] Delamaide, E.; Bazin, B.; Rousseau, D.; Degre, G. Chemical
than hydrophobic surfactants. Therefore, a hydrophilic
EOR for Heavy Oil: The Canadian Experience. SPE EOR Oil
anionic surfactant was incorporated in the colloidal system
and Gas West Asia, Muscat Conference, Oman, Mar 31–Apr 2,
for these experiments. Zeta potential measurements demon- 2014. DOI: 10.2118/169715-MS.
strated the effective attachment of the cationic and anionic [3] Kaminsky, R.; Wattenbarger, R.; Lederhos, J.; Leonardi, S. A.
surfactants to the silica nanoparticles and proved their abil- Viscous Oil Recovery Using Solids-Stabilized Emulsions. SPE
ity to form a stable complex colloidal solution at low salinity Annual Technical Conference and Exhibition, Florence, Italy,
conditions. Sandpack flooding, Berea sandstone core flood- Sept 19–22, 2010. DOI: 10.2118/135284-MS
ing, and slim tube sandpack flooding experiments were con- [4] Qiu, F. The Potential Applications in Heavy Oil EOR With the
Nanoparticle and Surfactant Stabilized Solvent-Based Emulsion.
ducted using the colloidal dispersion solution (A-3), thus
Canadian Unconventional Resources and International
confirming the stability of emulsions and their potential in Petroleum Conference, Calgary, Alberta, Oct 19–21, 2010. DOI:
enhanced heavy oil recovery. By emulsification, Silica 10.2118/134613-MS.
(1 wt%) þ DTAB (0.04 wt%) þ anionic surfactant (0.5 wt%) [5] Zhang, D.; Du, X.; Song, X.; et al. Application of the
colloidal solution could improve the oil recovery by approxi- Marangoni Effect in Nanoemulsion on Improving
mately 27% when compared to the brine flooding case (base Waterflooding Technology for Heavy-Oil Reservoirs. SPE J.
case), when at the optimal flow rate. The glass visualization 2018, 23, 831–840. DOI: 10.2118/187953-PA.
[6] Lee, J.; Babadagli, T. Improvement of Microemulsion
study further confirmed the ability of the colloidal solution
Generation and Stability Using New Generation Chemicals and
to efficiently generate emulsion in-situ at the pore scale
Nano Materials During Waterflooding as a Cost-Efficient
under atmospheric pressure and temperature. The significant Heavy-Oil Recovery Method. SPE Trinidad and Tobago Section
increase in heavy oil recovery observed from the sandpack Energy Resources Conference, Port of Spain, Trinidad and
flooding and Berea sandstone core flooding through in-situ Tobago, June 25–26, 2018. DOI: 10.2118/191171-MS.
emulsification confirms the practical applicability of the [7] Lee, J.; Babadagli, T.; Giesbrecht, R. Impact of Divalent Ions on
nanoscale dispersion in the field of heavy oil recovery. Heavy Oil Recovery by Inemulsification. J. Surf. Deterg. 2019,
(to be published).
[8] Pei, H.; Shu, Z.; Zhang, G.; et al. Experimental study of nano-
Funding particle and surfactant stabilized emulsion flooding to enhance
heavy oil recovery. J. Pet. Sci. Eng. 2018, 163, 476–483. DOI:
This research was conducted under the second author’s (TB) NSERC 10.1016/j.petrol.2018.01.025.
Industrial Research Chair in Unconventional Oil Recovery (industrial [9] Arab, D.; Kantzas, A.; Bryant, S. Nanoparticle-Enhanced
partners are Petroleum Development Oman, Total E&P Recherche Surfactant Floods to Unlock Heavy Oil. SPE Improved Oil
Developpement, SiGNa Oilfield Canada, Husky Energy, Saudi Aramco,
Recovery Conference, Tulsa, Oklahoma, Apr 14–18, 2018. DOI:
Devon, APEX Eng. and BASF, and an NSERC Discovery Grant
(No: RES0011227).
[10] Mohammadzadeh, O.; Sedaghat, M. H.; Kord, S.; Zendehboudi,
S.; Giesy, J. P. Pore-Level Visual Analysis of Heavy Oil
Recovery Using Chemical-Assisted Waterflooding Process –
Use of a New Chemical Agent. Fuel 2019, 239, 202–218. DOI:
[1] Wu, Y.; Mahmoudkhani, A.; Watson, P. A Non-Thermal [11] Seright, R. Potential for Polymer Flooding Reservoirs with
Surfactant-Polymer Based Technology for Enhanced Heavy Oil Viscous Oils. Spe. Res. Eval. Eng. 2010, 13. DOI: 10.2118/
Recovery in Oil Sand and Ultra Shallow Reservoirs. SPE Heavy 129899-PA.

[12] Knight, B.; Rhudy, J. Recovery of High-Viscosity Crudes by [30] Feng, A.; Zhang, G.; Ge, J.; Jiang P.; Pei H.; Zhang, J. Q.; Li, R.
Polymer Flooding. J. Can. Pet. Technol 1977, 16, 46–56. Study of Surfactant-Polymer Flooding in Heavy Oil Reservoirs. SPE heavy Oil Conference Canada, Calgary, Alberta, Jun 12–14,
[13] Asghari, K.; Nakutnyy, P. Experimental results of polymer 2012. DOI: 10.2118/157621-MS.
flooding of heavy oil reservoirs. Canadian International [31] Rousseau, D.; Bekri, S.; Boujlel, J.; Hocine, S.; Degre, G.
Petroleum Conference, Calgary, Alberta, Jun 17–19, 2008. Designing Surfactant-Polymer Processes for Heavy Oil Reservoirs: Case Studies. SPE Canada Heavy Oil Technical
[14] Wang, J.; Dong, M. Optimum Effective Viscosity of Polymer Conference, Calgary, Alberta, March 13–14, 2018. DOI: 10.
Solution for Improving Heavy Oil Recovery. J. Pet. Sci. Eng. 2118/189745-MS.
2009, 67, 155–158. DOI: 10.1016/j.petrol.2009.05.007. [32] Huang, S.; Dong, M. Alkaline/Surfactant/Polymer (ASP) Flood
[15] Wassmuth, F.; Green, J.; Arnold, W.; et al. Polymer Flood Potential in Southwest Saskatchewan Oil Reservoirs. J. Cad. Pet.
Application to Improve Heavy Oil Recovery at East Bodo. J. Technol. 2004, 43, 1–15.
Can. Pet. Technol. 2009, 48. DOI: 10.2118/09-02-55. [33] Sedaghat, M.; Mohammadzadeh, O.; Kord, S.; Chatzis, I. Heavy
[16] Bondino, I.; Nguyen, R.; Hamon, G. et al. Tertiary polymer Oil Recovery Using ASP Flooding: A Pore-Level Experimental
flooding in extra-heavy oil: an investigation using 1D and 2D Study in Fractured Five-Spot Micromodels. Can. J. Chem. Eng.
experiments, core scale simulation and porescale network mod- 2016, 94, 779–791. DOI: 10.1002/cjce.22445.
els. International Symposium of the Society of Core Analysts, [34] Zhu, Y.; Fu, T.; Liu, K.; Lin, Q.; Pei, X.; Jiang, J.; Cui, Z.; Binks,
Austin, Texas, Sept 18–21, 2011. B. P. Thermoresponsive Pickering Emulsions Stabilized by Silica
[17] Pei, H.; Zhang, G.; Ge, J.; Jin, L.; Liu, X. Analysis of Nanoparticles in Combination with Alkyl Polyoxyethylene
Microscopic Displacement Mechanisms of Alkaline Flooding Ether Nonionic Surfactant. Langmuir 2017, 33, 5724–5733.
for Enhanced Heavy-Oil Recovery. Energy Fuels 2011, 25, DOI: 10.1021/acs.langmuir.7b00273.
4423–4429. DOI: 10.1021/ef200605a. [35] Chavalier, Y.; Bolzinger, M. A. Emulsions Stabilized with Solid
[18] Ge, J.; Feng, A.; Zhang, G.; Jiang, P.; Pei, H.; Li, R.; Fu, X. Nanoparticles. Colloids Surf. 2013, 439, 23–34. DOI: 10.1016/j.
Study of the Factors Influencing Alkaline Flooding in Heavy- colsurfa.2013.02.054.
Oil Reservoirs. Energy Fuels 2012, 26, 2875–2882. DOI: 10. [36] Pei, H.; Zhang, G.; Ge, J.; Zhang, J.; Zhang, Q. Investigation of
1021/ef3000906. Synergy between Nanoparticle and Surfactant in Stabilizing Oil-
[19] Dong, M.; Liu, Q.; Li, A. Displacement Mechanisms of in-Water Emulsions for Improved Heavy Oil Recovery. Colloids
Enhanced Heavy Oil Recovery by Alkaline Flooding in a
Surfaces A Physicochem. Eng. Asp. 2015, 484, 478–484. DOI: 10.
Micromodel. Particuology 2012, 10, 298–305. DOI: 10.1016/j.
[37] Sharma, T.; Velmurugan, N.; Patel, P.; Chon, B. H.; Sangwai,
[20] Ding, B.; Zhang, G.; Ge, J.; Liu, X. Research on Mechanisms of
J. S. Use of Oil-in-Water Pickering Emulsion Stabilized by
Alkaline Flooding for Heavy Oil. Energy Fuels 2010, 24,
Nanoparticles in Combination with Polymer Flood for
6346–6352. DOI: 10.1021/ef100849u.
Enhanced Oil Recovery. Pet. Sci. Technol. 2015, 33, 1595–1604.
[21] Tang, M.; Zhang, G.; Ge, J.; Jiang, P.; Liu, Q.; Pei, H.; Chen, L.
DOI: 10.1080/10916466.2015.1079534.
Investigation into the Mechanisms of Heavy Oil Recovery by
[38] Zhang, T.; Davidson, D.; Bryant, S. L.; Huh, C. Nanoparticle-
Novel Alkaline Flooding. Colloids Surf. A Physicochem Eng Asp.
Stabilized Emulsions for Applications in Enhanced Oil
2013, 421, 91–100. DOI: 10.1016/j.colsurfa.2012.12.055.
Recovery. SPE Improv. Oil Recover. Symp. 2010. DOI: 10.2118/
[22] Alomair, O. A.; Matar, K.; Alsaeed, Y. Nanofluids Application
for Heavy Oil Recovery. SPE Asia Pacific Oil & Gas Conference 129885-MS
[39] Peng, B.; Zhang, L.; Luo, J.; Wang, P.; Ding, B.; Zeng, M.;
and Exhibition, Adelaide, Australia, Oct 14–16, 2014. DOI: 10.
2118/171539-MS. Cheng, Z. A Review of Nanomaterials for Nanofluid Enhanced
[23] Corredor, L. M.; Aliabadian, E.; Husein, M.; Chen, Z.; Maini, Oil Recovery. RSC Adv. 2017, 7, 32246–32254. DOI: 10.1039/
B.; Sundararaj, U. Heavy Oil Recovery by Surface Modified C7RA05592G.
Silica Nanoparticle/HPAM Nanofluids. Fuel 2019, 252, [40] Suleimanov, B. A.; Ismailov, F. S.; Veliyev, E. F. Nanofluid for
622–634. DOI: 10.1016/j.fuel.2019.04.145. Enhanced Oil Recovery. J. Pet. Sci. Eng. 2011, 78, 431–437.
[24] Bryan, J.; Kantzas, A. Enhanced Heavy-Oil Recovery by Alkali- DOI: 10.1016/j.petrol.2011.06.014.
Surfactant Flooding. SPE Annual Technical Conference and [41] Yoon, K.; Son, H.; Choi, S.; et al. Core Flooding of Complex
Exhibition, Anaheim, California, Nov 11–14, 2007. DOI: 10. Nanoscale Colloidal Dispersions for Enhanced Oil Recovery by
2118/110738-MS. in-Situ Formation of Stable Oil-in-Water Pickering Emulsions.
[25] Pei, H.; Zhang, G.; Ge, J.; Tang, M.; Zheng, Y. Comparative Energy Fuels 2016, 3, 2628–2635. DOI: 10.1021/acs.energyfuels.
Effectiveness of Alkaline Flooding and Alkalinesurfactant 5b02806.
Flooding for Improved Heavy-Oil Recovery. Energy Fuels 2012, [42] Lee, J.; Babadagli, T. Screening of Chemicals for Low Cost
26, 2911–2919. DOI: 10.1021/ef300206u. Heavy Oil Recovery through Microemulsion Stability Tests.
[26] Nizamidin, N.; Weerasooriya, U.; Pope, G. Systematic Study of SPE/IATMI Asia Pacific Oil & Gas Conference and Exhibition,
Heavy Oil Emulsion Properties Optimized with a New Jakarta, Indonesia. Oct 17–19, 2017. DOI: 10.2118/186344-MS.
Chemical Formulation Approach: particle Size Distribution. [43] Sheng, J. Investigation of Alkaline-Crude Oil Reaction.
Energy Fuels 2015, 29, 7065–7079. DOI: 10.1021/acs.energyfuels. Petroleum 2015, 1, 31–39. DOI: 10.1016/j.petlm.2015.04.004.
5b01818. [44] Binks, B. P.; Rodrigues, J. A. Enhanced Stabilization of
[27] Kumar, R.; Mohanty, K. Sweep Efficiency of Heavy Oil Emulsions Due to Surfactant-Induced Nanoparticle
Recovery by Chemical Methods. SPE Annual Technical Flocculation. Langmuir 2007, 23, 7436–7439. DOI: 10.1021/
Conference and Exhibition, Denver, Colorado, Oct 30–Nov 2, la700597k.
2011. DOI: 10.2118/146839-MS [45] Goloub, T.; Koopal, L. Adsorption of Cationic Surfactants on
[28] Dong, M.; Ma, S.; Liu, Q. Enhanced Heavy Oil Recovery Silica. Comparison of Experiment and Theory. Langmuir 1997,
through Interfacial Instability : A Study of Chemical Flooding 13, 673–681. DOI: 10.1021/la960690d.
for Brintnell Heavy Oil. Fuel 2009, 88, 1049–1056. DOI: 10. [46] Chen, X.; Song, X.; Huang, J.; et al. Phase Behavior of Pickering
1016/j.fuel.2008.11.014. Emulsions Stabilized by Graphene Oxide Sheets and Resins.
[29] Liu, Q.; Dong, M.; Yue, X.; Hou, J. Synergy of Alkali and Energy Fuels 2017, 31, 13439–13447. DOI: 10.1021/
Surfactant in Emulsification of Heavy Oil in Brine. Colloids fuels.7b02672.
Surf. A. 2006, 273, 219–228. DOI: 10.1016/j.colsurfa.2005.10. [47] Paria, S.; Khilar, K. A Review on Experimental Studies of
016. Surfactant Adsorption at the Hydrophilic Solid-Water Interface.

Adv. Colloid Interface Sci. 2004, 110, 75–95. DOI: 10.1016/j.cis. Interface Sci. 2017, 508, 222–229. 0021–9797. DOI: 10.1016/j.
2004.03.001. jcis.2017.08.043.
[48] Li, Y.; Zhang, W.; Kong, B.; Puerto, M.; Bao, X.; Sha, O.; Shen, [54] Gandomkar, A.; Rahimpour, M. Investigation of Low-Salinity
Z.; Yang, Y.; Liu, Y.; Gu, S.; et al. Mixtures of Anionic/Cationic Waterflooding in Secondary and Tertiary Enhanced Oil
Surfactants: A New Approach for Enhanced Oil Recovery in Recovery in Limestone Reservoirs. Energy Fuels 2015, 29,
Low-Salinity, High-Temperature Sandstone Reservoir. SPE J. 7781–7792. DOI: 10.1021/acs.energyfuels.5b01236.
2016, 21, 1164. DOI: 10.2118/169051-MS. [55] Sandoval-Rodrıguez, L. S.; Ca~
nas-Marın, W. A.; Martınez-Rey, R.
[49] Arguelles, F. Pore Scale Investigations on the Dynamics of Rheological Behavior of Water-in-Oil Emulsions of Heavy and
SAGD Process and Residual Oil Saturation Development. Ph.D. Extra-Heavy Live Oils: Experimental Evaluation. CT&F Cienc.
Dissertation, Department of Civil and Environmental Tecnol. Futuro 2014, 5, 5–22. DOI: 10.29047/01225383.37.
[56] Nasery, S.; Hoseinpour, S.; Phung, L. T. K.; Bahadori, A.
Engineering, University of Alberta, 2015.
Prediction of the Viscosity of Water-in-Oil Emulsions. Pet. Sci.
[50] Ma, S.; Dong, M.; Li, Z.; Shirif, E. Evaluation of the
Technol. 2016, 34, 1972–1977. DOI: 10.1080/10916466.2016.
Effectiveness of Chemical Flooding Using Heterogeneous
Sandpack Flood Test. J. Pet. Sci. Eng. 2007, 55, 294–300. DOI:
[57] Binks, B. P.; Murakami, R.; Armes, S. P.; Fujii, S. Effects of pH
10.1016/j.petrol.2006.05.002. and Salt Concentration on Oil-in-Water Emulsions Stabilized
[51] Hankins, N.; Harwell, J. Case Studies for the Feasibility of Solely by Nanocomposite Microgel Particles. Langmuir 2006,
Sweep Improvement in surfactant-Assisted Waterflooding. J. 22, 2050–2057. DOI: 10.1021/la053017.
Pet. Sci. Eng. 1997, 17, 41–62. DOI: 10.1016/S0920- [58] Arla, D.; Sinquin, A.; Palermo, T. Influence of pH and Water
4105(96)00055-1. Content on the Type and Stability of Acidic Crude Oil
[52] Kumar, S.; Aswal, V.; Kohlbrecher, J. Size-Dependent Emulsions. Energy Fuels 2007, 21, 1337–1342. DOI: 10.1021/
Interaction of Silica Nanoparticles with Different Surfactants in ef060376j.
Aqueous Solution. Langmuir 2012, 28, 9288–9297. DOI: 10. [59] Upadhyaya, A.; Acosta, E. J.; Scamehorn, J. F.; Sabatini, D. A.
1021/la3019056. Adsorption of Anionic-Cationic Surfactant Mixtures on Metal
[53] Al-Anssari, S.; Arif, M.; Wang, S.; Barifcani, A.; Iglauer, S. Oxide Surfaces. J. Surfact. Deterg. 2007, 10, 269–277. DOI: 10.
Stabilizing Nanofluids in Saline Environments. J. Colloid 1007/s11743-007-1045-3.