SPE 110738

Enhanced Heavy-Oil Recovery by Alkali-Surfactant Flooding

J. Bryan and A. Kantzas, SPE, University of Calgary and TIPM Laboratory

Copyright 2007, Society of Petroleum Engineers

This paper was prepared for presentation at the 2007 SPE Annual Technical Conference and
Exhibition held in Anaheim, California, U.S.A., 1114 November 2007.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than
300 words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.

Abstract
This study presents the results of laboratory core studies
investigating the recovery mechanisms of alkali-surfactant
flooding in heavy oil reservoirs. Specifically, mixtures of
water and alkali-surfactant systems have been injected into
cores containing heavy oil (11 000 mPas and 15 000 mPas).
Salinity is varied in order to generate oil-in-water vs. water-inoil emulsion systems, and the effects of generating different
emulsions are compared.
The application of this work is for the many heavy oil
reservoirs in countries such as Canada and Venezuela
containing viscous oil that still has some limited mobility
under reservoir conditions. Alkali-surfactant (AS) flooding
has considerable potential for non-thermal oil recovery after
primary production.
Experiments were performed on cores with varying
permeability, at different AS injection rates. All tests were
performed on gas-free oil systems. The response from direct
injection of AS systems is compared to AS injection after
waterflooding. Pressure and oil recovery information is
obtained from core floods, and these results are interpreted
based on a semi-theoretical framework obtained from phase
behavior and bulk liquid studies. It is demonstrated that both
oil-in-water and water-in-oil emulsions can lead to the
recovery of additional oil.
Alkali-surfactant flooding is already an established
technique in conventional oil reservoirs, whereby enhanced oil
recovery is a result of reduced trapping of oil due to the
lowered oil/water interfacial tension. In addition, the injection
of these chemicals may lead to the formation of emulsions, as
has been documented by previous researchers. In our work,
we demonstrate that in heavy oil systems, emulsion formation
is a necessary requirement for the production of heavy oil.
When these emulsions form, AS injection can lead to
considerable improvements in the flooding response, even
without the addition of polymers to stabilize the flood.

Introduction
Several countries in the world, notably Canada and Venezuela,
contain massive resources of unconventional heavy oil and
and bitumen. With issues of resource stability and rising oil
prices, international interest is now shifting rapidly towards
Canadas oil sands. The oil sands are characterized as
unconsolidated, high porosity and high permeability
reservoirs. While ease of flow is therefore not a significant
concern, the single biggest obstacle to successful recovery
from the oil sands is the high oil viscosity.
Heavy oil reservoirs are a special subset of the oil sands,
whereby the oil viscosity at reservoir temperature and pressure
varies on the order of 50 50 000 mPas (cP). While this oil
is still very viscous, it does have some limited mobility at
resevoir conditions. As much as 20% of the oil may be
recovered by solution gas drive1, but in many cases the
recovery is much lower. At the end of primary production,
significant oil still remains in the reservoir, but the reservoir
energy has now been depleted. This is the target for enhanced
heavy oil recovery.
In order to recover additional heavy oil after primary
production, a fluid usually has to be injected in order to
displace oil to the production wells. However, mobility ratio
concerns dominate displacement of viscous oil, and most EOR
processes focus on reduction of the oil viscosity or
improvement in the mobility ratio. Unfortunately, many of the
heavy oil reservoirs in Canada are relatively small and thin,
making them poor candidates for expensive thermal processes.
Ideally, the displacement mobility ratio should be improved in
an inexpensive (i.e. non-thermal) fashion.
This work
investigates the potential for alkali-surfactant flooding to be
used for enhanced heavy oil recovery.
The injection of alkalis and/or surfactants into oil
reservoirs is not a new technology. As early as in the 1920s,
Nutting proposed the injection of alkaline solution into
reservoirs for oil recovery2. The injection of a combination of
alkali and surfactant was discussed in the 1950s by Reisberg
and Doscher3. Since then, chemical injection (alkali and/or
surfactant) has become an accepted enhanced oil recovery
methodology in many conventional oil applications.
Surfactants are a special class of molecule that are both
hydrophobic and hydrophilic, thus the most stable
configuration for these molecules is at the oil-water
interface4,5. In surfactant flooding, these molecules are
generally injected along with water in order to reduce the oilwater interfacial tension. This leads to a reduction in capillary
forces, that may trap oil in rock pores6,7. Alkali flooding is a
special subset of chemical flooding, whereby the surfactant is

generated in-situ through the reaction between the injected

alkali and the acidic crude oil8.
Upon either injection or formation of surfactants, several
different mechanisms can work synergistically to improve oil
recovery. Possible mechanisms are9: emulsification and
entrainment of droplets, emulsification and entrapment of
droplets, and wettability reversal. Different researchers have
proposed different causes for improved oil recovery.
Generally, reduction in oil-water interfacial tension is often
cited7,10, along with alteration of wettability to oil wet
conditions and oil film drainage8,10, or alternatively wettability
alteration to more water wet conditions11, which can alter the
displacement pattern and relative permeability curves. It is
likely that in actual reservoir applications, several processes
are working together to give the improved oil recovery. This
work considers the possible mechanisms that could be
responsible for enhanced recovery of high viscosity heavy oil.
Chemical Flooding in Heavy Oil
Several studies have been done in the past investigating the
effect of injecting alkali and/or surfactant into heavy oil
systems. In general, it has been shown that these chemicals do
lead to improved recovery compared to waterflooding, but the
mechanisms responsible for this recovery are less certain than
in conventional oil. Some researchers cite the low oil-water
interfacial tension12-14, similar to conventional oil.
Alternatively, the formation of water-in-oil (W/O) emulsions
have also been proposed as a possible recovery
mechanism15,16. These emulsions should be more viscous than
oil by itself, thus they could possibly lead to improvements in
the mobility ratio and sweep efficiency of the flood. Finally,
the formation of oil-in-water (O/W) emulsions has also been
suggested as a possible recovery mechanism17-21. In this
application, oil is emulsified into water and the oil droplets
either plug rock pores and give improved sweep efficiency, or
are perhaps even entrained along with the flowing aqueous
phase18,21.
The key to understanding how AS floods can lead to
improved oil recovery is to examine the state of the reservoirs
at the time of injection. In conventional oil linear systems,
residual oil at the end of waterflooding is trapped by capillary
forces, thus reducing the interfacial tension can lead to
reduced trapping of oil ganglia6. In heavy oil reservoirs,
however, residual oil after waterflooding is not mainly due to
capillary trapping. Rather, the oil is largely bypassed due to
viscous fingering caused by the adverse mobility ratio
between oil and water. Therefore, for chemical floods to give
improved heavy oil recovery, the AS solution must somehow
be improving the mobility ratio, and thus giving a more stable
displacement of oil to the production wells.
The mechanism by which the mobility ratio can be
improved is the creation of oil and water emulsions, which can
form under conditions of shear and low interfacial tension.
O/W emulsions may plug waterflooded channels, leading to
improved volumetric sweep of the reservoir16-18,20,21.
Additionally, the emulsified oil droplets could potentially be
entrained in the flowing aqueous phase19,21. In either case, the
goal is to create O/W emulsions, as opposed to more viscous
W/O emulsions. However, if W/O emulsions were injected or
formed in-situ then perhaps these fluids, being more viscous

SPE 110738

than the oil, could also lead to improvements in the mobility

ratio16,22 and more stable floods.
Proper design of an AS flood for heavy oil requires
understanding which emulsions are forming, and how these
emulsion systems are leading to improved oil recovery. This
work examines the results from laboratory core flood systems,
and interprets the recovery mechanisms based on a semitheoretical description developed from experiments in bulk
liquid systems.
Materials and Experiments
The heavy oil used in these experiments comes from a field in
Saskatchewan, Alberta, that is currently undergoing primary
production. At ambient temperature (23C) the oil has a
viscosity of around 11 000 mPas, and a density of 982 kg/m.
In some of the core floods, the heavy oil used was more
viscous around 15 000 mPas. This oil is all from the same
field, however differences in handling or possibly geological
heterogeneities could be responsible for the different measured
viscosity values. The aqueous phase used is either de-ionized
(DI) water or brine (Brn) containing 20 000 ppm NaCl in
water.
The preformed surfactant used in this study is a
commercial anionic surfactant that was supplied by Stepan
Company. This surfactant (Bio-Terge PAS-8S) is a sodium
alkane sulfonate, and has been used in heavy oil studies in the
past21,23. The alkali used is sodium carbonate, Na2CO3, which
was chosen over other alkali agents due to its buffering
capabilities and the fact that it is less sensitive to variations in
the water chemistry than NaOH24. The alkali will react with
organic acids present in the oil and should generate
monovalent anionic surfactants. In this manner, both the
preformed and in-situ surfactants are anionic.
In order to investigate the effect of changing salinity and
alkali concentration on the emulsions formed, a suite of
experiments were first carried out in bulk liquid systems.
Surfactant concentration was kept constant at 0.1 wt% (1000
ppm), and alkali was varied from 0%, 0.1%, 0.5%, 1.0% and
1.5% (0, 1000, 5000, 10 000 and 15 000 ppm). Phase
behavior studies were performed in 20 mL glass scintillation
vials, using the methodology outlined by Nelson et al.25.
Interfacial tension (IFT) measurements were also performed,
using a Kruss Site-04 Spinning Drop Tensiometer. This
apparatus is valid for measuring low IFT systems, but cannot
measure normal ranges of oil-water IFT (in the absence of
alkali and surfactant) due to the similar oil and water densities.
At the conclusion of the bulk liquid study, a suite of
corefloods was carried out to investigate if emulsions could
form and flow in porous media under reasonable reservoir
flow rates. In some experiments, DI water systems were
studied and in other cores, brine was injected. Tables 1 and 2
list the properties of the sand packs, the oil viscosity and the
injection rates for both DI AS and Brn AS systems. In total,
28 core floods were performed.
Cores A to I in Table 1 are actually the same core, which is
a stainless steel tube packed with LM 70 sand. Cores J to L
are a second core, that was packed with a sieved fraction (<
75m) of LM 70 sand. Cores M and N are a larger version of
cores A to I, which were flooded in order to get more accurate

SPE 110738

recovery and water cut information from the larger core size.
In cores M and N, overburden pressure of 2070 kPa (300 psi)
was also applied, which is responsible for the lower
permeability compared to cores A to I. Finally, core O is the
larger version of cores J to L, using the same sieved sand.
In the Brn AS floods studied in Table 2, Cores Q to X are
the same core as A to I in Table 1. Likewise, cores Y to AB
are the same core as J to L in Table 1, and core AC is the
larger version (cores M and N in Table 1). In between
experiments, cores were flooded with toluene to remove any
oil from the rock. The cores were then flushed using DI water
in order to remove any surfactant that may have adsorbed onto
the sand grains. Finally, the cores were flushed with methanol
and were then dried using air. The core porosity and
permeability was re-measured before starting a new
experiment, in order to ensure that the core properties
remained unchanged and results from different experimental
runs could be compared against one another. Since the cores
were packed using unconsolidated material and in many cases
overburden pressure was not present, the measured
permeability to water changed between experiments. This was
especially pronounced in the higher permeability systems, as
evidenced in Tables 1 and 2.
After the cores were saturated with either DI water or brine
and their permeability had been measured, heavy oil (viscosity
shown in Tables 1 and 2) was injected at a constant rate using
a digital ISCO pump until the cores were at their irreducible
water saturation. At this point, injection of either water or AS
solution was performed at the constant rates shown in Tables 1
and 2. It should be noted that the velocities listed in these
tables are Darcy velocities, calculated by dividing the injection
flow rate (set by the ISCO pump) by the cross-sectional area
of the core. In core floods with only one test, this means that
either a waterflood or a primary AS flood was performed. In
these floods, the experiment was terminated after around 5 8
pore volumes (PV) of fluid had been injected. In systems with
both waterfloods and AS floods, a waterflood consisting of
approximately 5 PV of fluid was injected, followed by around
2 PV of AS solution.
Produced fluid volumes were measured using low field
NMR, since the heavy oil is sticky and may adhere to the
walls of the glass collection containers, making visual
determination of produced oil and water cut difficult to
measure accurately. After the NMR spectra had been
analyzed, the oil recovery was verified by separation with
toluene. Pressures were recorded using transducers connected
to an automatic data acquisition system.
The results of the experiments are divided into findings
from the bulk liquid study and the core floods. The data
collected from the core floods consist of produced fluid
volumes and pressure, thus the actual recovery mechanisms
that are occurring at the pore scale can only be inferred based
on an understanding of how the systems behave in the absence
of porous media.
Bulk Liquid Study
Oil and water are immiscible fluids, therefore when they are
mixed energy has to be put into the system in order to
emulsify them. The energy required for emulsification is
proportional to the interfacial tension between the two

fluids5,26. For fluids that are dis-similar, such as oil and water,
the energy requirement can be significant. Shear or agitation
is required to overcome the Gibbs free energy of formation
and, more importantly, the Laplace equation of capillary
pressure in a droplet26. In order to form an O/W or W/O
emulsion, therefore, two requirements must be met: the oilwater IFT must be reduced, and shear must be added to the
system. The addition of alkali and surfactant is used to reduce
the oil-water IFT such that emulsification can occur with a
minimum of applied shear.
Figure 1 shows the measured oil-water IFT values as a
function of alkali content, with the surfactant fixed at 0.1 wt%
(1000 ppm). As mentioned previously, due to the similar
densities of oil and water, when interfacial tension is higher
than 1 mN/m the apparatus cannot measure the IFT even at its
maximum rotational speed. Several comments can be made
regarding Figure 1. First, considering the DI water data: in the
absence of any alkali the surfactant alone cannot reduce the
IFT down to measurable values. Even with 0.1 % alkali
(1000 ppm) added, the IFT is too high to be measured. After
0.2% alkali has been added to the aqueous phase, however, the
IFT values plateau at around 0.02 mN/m for the entire range
of alkali contents tested. For the brine system, even in the
absence of alkali it appears that the surfactant alone can
reduce IFT down to measurable values. In the presence of
salt, the IFT has a minimum at 5000 ppm alkali, and additional
alkali leads to increases in the IFT.
The salinity effect observed in Figure 1 has also been
previously noted by other researchers8,25. Salt tends to shift
the salinity minimum to lower alkali concentrations, and the
presence of additional alkali then acts like increased salinity.
Therefore, if injecting brine, the IFT in the system seems less
stable than for DI water. In either case, however, it can be
seen that with the addition of alkali and surfactant to the
system, the interfacial tension between oil and water is
reduced by two to three orders of magnitude.
While low IFT is certainly a requirement for
emulsification to occur, the results from Figure 1 do not
indicate which type of emulsion will be generated. When oil
and water are agitated at low IFT conditions, both O/W and
W/O emulsions are formed simultaneously. The emulsions
are thought to break exponentially, at a rate governed by the
frequency of droplet collisions and by the energy barriers to
droplets coalescing upon contact27. The emulsion that
separates more slowly is the emulsion that is macroscopically
said to have formed.
Due to the high viscosity of heavy oil, when the oil phase
is continuous, water droplets will collide at a much lower
frequency than oil droplets in a continuous, low viscosity
water phase. For this reason, W/O emulsions are much more
common than O/W emulsions in heavy oil systems5,27.
However, oil droplets in water can also be stabilized by
electrostatic repulsion of similarly-charged surfactant
heads26,28. Several other rules of thumb can also be used to
identify which emulsion type should form in bulk liquid
systems5. First, if the volume of one fluid is much larger than
the other fluid, it is more likely that the fluid with the larger
volume will be the continuous phase. Second, if the surfactant
is more soluble in one phase, that phase will also more easily
be continuous. Finally, surfactants made from monovalent

metal cations tend to produce O/W emulsions, while

polyvalent metal cations will more easily form W/O
emulsions. This is known as the oriented wedge theory5.
Phase behavior tests were carried out in order to ascertain
if heavy oil could be emulsified into water, despite the large
viscosity disparity that tends to favour W/O emulsions.
Following the aforementioned rules of thumb, systems were
generated with 90% water and 10% heavy oil, and the
preformed surfactant is water soluble. Also, the surfactant and
the alkali use monovalent metal cations (Na+), which should
favour O/W emulsions. In each system, the alkali and
surfactant were mixed into the aqueous phase and then oil was
placed on top of the AS solution. The vials were then
manually agitated for 60 seconds, after which point they were
examined visually to see if any emulsions had formed.
Figure 2 is a photograph of the control sample (no
chemical added) and a sample containing 0.1% surfactant but
no alkali in DI water. After agitation, it can be seen that the
oil and water quickly separated again, leaving a clear aqueous
phase with oil on top. The phase behavior therefore indicates
that in both systems the IFT was not low enough to sustain a
stable emulsion, and this is verified by examining Figure 1.
One other interesting note from Figure 2 is that in the control
sample, oil smeared the glass, indicating that the glass has at
least some tendency to be wet by the oil8. In the vial
containing surfactant, although no emulsion was formed, the
glass is clean, which indicates that the surfactant made the
glass more strongly water wet. This is the basis for
detergency4,27, and is still potentially useful for EOR
applications.
Figure 3 shows samples containing DI water and 0.1%
surfactant, with the alkali content being varied. In Figure 3,
the samples have not yet been agitated, however they already
appear to be different than in the control sample or the sample
with just surfactant. In these samples, the oil-water interface
appears to be flat instead of curved, indicating that there has
been a change in the interfacial properties of the system.
Figure 4 shows the samples after agitation. For all alkali
contents, the fluid appeared to be a single-phase, black, low
viscosity fluid. The low viscosity is the key point: water is the
continuous phase, thus O/W emulsions had been generated.
Over the course of 24 hrs the emulsions gradually separated.
The glass is no longer as clean as in the sample containing
only surfactant, but the oil smearing is not as bad as the
control sample.
Figure 5 shows samples after agitation with varying alkali
content, but with brine instead of DI water. The effect of the
salt is evident by comparing Figures 4 and 5. When salt is
present in the water, the aqueous phase is clear and the oil
phase is swelled and more viscous after agitation, indicating
the presence of W/O emulsions. Additionally, the glass is
now totally smeared with oil, and there is also some oil stuck
to the bottom of the glass. In DI water systems, oil droplets
were stabilized due to electrostatic repulsion between the
negatively-charged surfactant heads. In the presence of brine,
the electrical forces are reduced26, thus it becomes more
difficult to sustain O/W emulsions. Additionally, the salt may
have caused the surfactants to become more oil soluble8,
which will also promote the formation of W/O emulsions.

SPE 110738

It has been shown that with the choice of different

surfactants, O/W emulsions can be stabilized even in more
saline systems29. Using this chemical system, however,
changing the salinity allows for an examination of the effect of
generating different emulsion types in reservoir EOR
applications.
Extension of Bulk Liquid Results to Porous Media
The formation of different emulsion types becomes more
complex in porous media, due to the additional interaction
between the fluid and rock interfaces. On a flat solid surface,
the distribution of oil and water is given by Youngs
Equation4,27:
ow cos = os ws
(1)
Where ow = oil-water interfacial tension
os = oil-solid interfacial free energy
ws = water-solid interfacial free energy
= contact angle of the water-solid interface.
Youngs equation expresses the nature of the solid to be
wetted by either oil or water. The solid wettability is
expressed as the difference in the interfacial free energies (os
- ws). While these parameters cannot be measured directly,
they can be inferred from measurements of the fluids
interfacial tension and the contact angle.
If the contact angle between the glass, oil and water is
examined in Figure 2, it can be seen that the glass is
preferentially wet by water ( < 90). Despite this fact, it
appears that even at the water wet condition, oil can still
adhere to the glass in the absence of surfactant. This indicates
that in the control sample, water would not spread fully
between the glass and oil4. With the addition of alkali before
agitation (Figure 3), the remarkable difference is the flat oilwater interface. This is a characteristic of low interfacial
tension systems, and indicates that at this scale capillary
pressure has become negligible between the oil and water30.
In other words, the interfacial forces no longer control the
distribution of the fluids, and the system appears to be
neutrally wet.
It is important to realize that in porous media, even for low
IFT systems there may still be some capillary pressure present,
due to the small radius in the Young-Laplace equation4,30:
2 ow cos 2( os ws )
=
(2)
Pc =
R
R
Where Pc = capillary pressure
R = radius of the pore or capillary tube.
One of the major differences between bulk liquid experiments
and the behavior of the fluids in porous media is that the fluids
are in much smaller confined spaces while in the reservoir.
Therefore, although at low interfacial tension the solid appears
to be neutrally wet (i.e. fluids contact the solids independently
of the wettability), this may not be the case in porous media.
Additionally, although the fluids appear to be neutrally wet for
both the DI AS and Brn AS solutions, the oil adheres much
more strongly to the glass in the presence of salt, thus the glass
becomes less hydrophilic even if this is not expressed by the

SPE 110738

measured contact angles. In higher IFT systems (> 1 mN/m)

this has also been observed by other researchers11. Therefore,
phase behavior experiments still allow for at least a qualitative
understanding of the wetting nature of the glass.
The reason that solids wettability is important is that in the
presence of solids, the fluid that wets the solid will more
easily be the continuous phase in an emulsion system5.
Assuming a pore to be a cylindrical capillary tube, Jamaluddin
and Butler31 calculated the net work per unit volume of a
water droplet emulsified into oil to be:
3 R 2

(3)
Pemul = ow + cos
R r 3

Where Pemul = net work per unit volume of a droplet

r = radius of the emulsion droplet.
Equation 3 shows that if a water droplet were to form in a
continuous oil phase, the minimum net work requirement
occurs for a large emulsion droplet (radius equal to the pore
radius) and an oil wet rock ( 90). Likewise, for an oil
droplet to form in water, the minimum work requirement
would be when the rock is strongly water wet. Therefore, it
seems that the rock wettability has a strong influence over
which emulsion system will form in the reservoir.
Previous work with thermal heavy oil operations32,33 and
injection into heavy oil cores at very high rates34 has provided
evidence for the in-situ formation and flow of W/O emulsions.
As mentioned previously, due to the high viscosity of the oil
phase, these emulsion types will be much more common than
O/W emulsions. For all of these studies, however, the oilwater IFT was high, and solids were deemed to have become
oil wet. In the case of AS injection, where interfacial forces
are much smaller and the solids wettability is less clear, it
becomes very difficult to determine which emulsions will
form and how the solids will be wetted by oil and water.
From the bulk liquid study, the following inferences can be
made: at low interfacial tension, the distribution of fluids on
the solid surfaces is unclear, and cannot be easily explained by
contact angle/IFT measurements. It appears that for this
chemical system, in the absence of salt O/W emulsions can
form and be stabilized by electrostatic repulsion between
droplets. In this state, glass (and sand) appears to still be at
least partially wetted by water. In the presence of salt in the
aqueous phase, only W/O emulsions form and the glass
becomes much more hydrophobic, or oil wet. Although it is
not certain if these relationships will still hold in small rock
pores, these will still be the assumptions used to interpret the
pressure and production data from the core floods presented in
the following section.
AS Recovery Potential Through Core Flooding
In bulk liquid experiments, the systems were subjected to
significant shear during agitation.
The core flooding
experiments were designed to test if emulsions could form due
to the shear forces of flow through rock pores. At high
temperatures, emulsions can form in porous media due to
condensation of water in very small droplets32,33. Likewise, at
flow under very high shear (i.e. 25 250 m/day velocities),
emulsification can also occur34. At lower shear conditions and

ambient temperature, however, researchers have not found

evidence of significant emulsion generation and flow when
viscous oil and water mix35. The key question that had to be
answered was, under conditions of low IFT due to the
presence of AS in the water, could emulsions form at the
levels of shear from flow at normal reservoir rates. In order to
answer this question, a suite of waterfloods, primary and
secondary AS floods were carried out and are summarized in
Tables 1 and 2.
Figure 6 shows the measured pressure drop and oil
recovery profile for a heavy oil waterflood, performed at a
constant injection rate. At early times, water is injected into
the core but the viscous oil cannot flow out at the same rate,
thus pressure builds up in the core. During this region of high
pressure gradient, the oil recovery is high. After water breaks
through, the pressure in the system declines sharply, as
injected water can now flow along paths of low resistance, and
the oil recovery rate drops accordingly. Figure 7 compares the
pressure response in the waterflood to the produced water
cuts. Once water breaks through, the remaining oil is
produced at very high water cuts. Other researchers have
shown that for fast injection rates, the water cuts will be
higher and the oil recovery profile will be less efficient than
for lower injection rates36.
Figure 8 compares the pressure response for the waterflood
with that of a primary Brn AS flood. Similar results are
shown elsewhere21 for a DI AS system. When AS is injected
along with water, the pressure maximum occurs at later times,
indicating a delay in the water breakthrough. Even after water
breakthrough, pressure declines more gradually in the
chemical flood, indicating a more stable displacement front.
The total recovery efficiency (RF/PV inj) of waterfloods vs.
primary AS floods is shown in Tables 3 and 4 for the different
floods, and the results are summarized in Table 5. As
expected, due to the more stable displacement during primary
AS flooding, the recovery efficiency appears to be higher than
for waterflooding. These results will be further interpreted
later in the paper.
Figures 9 and 10 show the pressure response compared to
the oil recovery and water cut profiles for a waterflood
followed by a secondary DI AS flood. In this flood, AS
injection began after approximately 4.5 PV of water had been
injected at a constant rate. Upon injection of this chemical
solution, the water cuts drop significantly and the oil recovery
improves as well. It should be noted that despite the fact that
DI AS solution had been injected, and O/W emulsions had
been generated in the bulk liquid studies, no produced
emulsions were observed in the core flood. However, the
pressure response indicates that with chemical injection,
something had changed in the system. The mechanism
proposed in this work is that of emulsification and entrapment,
leading to improved sweep efficiency.
In order to model the flow of fluids during AS injection,
authors in the past have attempted to define effective
viscosity values, assuming that O/W emulsions can be
represented by a single-phase fluid of different viscosity to
that of the oil37. In AS flooding applications, however, this
approach is not valid. Emulsions are forming in-situ, flowing,
breaking and plugging rock pores.
Additionally, the
emulsions droplet sizes are likely in the same range as the pore

sizes31, making it incorrect to treat them as a pseudo-single

phase fluid. Modeling emulsion flow using two-phase flow
equations with relative permeability34 is likewise incorrect,
since oil and water are no longer flowing in separate flow
paths. Instead, both viscosity and relative permeability can be
incorporated into phase mobility, which is a simplified
approach to understanding the effect of the AS solution on the
recovery response21:
k
Q/A
(4)
x = x =
x P / L
Where kx = effective mobility to oil (o) or water (w)
x = effective viscosity of oil (o) or water (w)
x = oil (o) or water (w) effective mobility
Q/A = velocity of oil or water
P/L = pressure gradient across the core.
From the measured pressure data and the oil and water
production rates, effective phase mobilities can be calculated
using Equation 4. It should be noted that in this expression,
the measured pressure drop is applied to both the oil and
water; this is valid because of the low IFT in the system,
which leads to negligible differences in oil and water pressure
due to low capillary pressures. Once the individual phase
mobilities have been calculated, the effective mobility ratio
(Meff) can be found:

k
(5)
M eff = w = w o
o w ko
The effective mobility ratio is calculated for waterflooding
and DI AS flooding in Figure 11. In this figure, the injected
pore volumes have been normalized in order to reduce the
scatter in the floods. Waterfloods are carried out to a
normalized value of one PV injected, and secondary AS floods
then go to two PV injected. For the waterfloods, the effective
mobility ratio is low until water breakthrough, as expected.
After breakthrough, Meff is on the order of 40 50. Upon
injection of DI AS solution, a remarkable reduction in Meff is
observed. The effective mobility ratio decreases down to the
order of unity, meaning that the flood actually becomes stable
at early times. This corresponds to the region in Figures 9 and
10 where the pressure gradient increases in the core. In all of
the floods, no produced O/W emulsion was observed, thus the
proposed recovery mechanism is that of emulsification &
entrapment.
Effective mobility ratio for Brn AS floods are presented in
Figure 12.
Once again, during waterflooding Meff is
approximately 50 after one normalized PV of fluid has been
injected, and with Brn AS addition the mobility ratio improves
significantly. In both DI AS and Brn AS flooding, therefore,
the mobility ratio response is very similar. Even after the
mobility ratio increases once again, the final value of Meff is
around 20, indicating that even after pressure declines again in
the system the injected fluid is still less mobile than pure
water.
The recoveries from secondary DI AS flooding are shown
in Table 3, and those for secondary Brn AS flooding are in
Table 4. The average recovery efficiencies for both types of

SPE 110738

floods are listed in Table 5. These results show definitively

that secondary AS flooding, either using brine or DI water,
leads to improved oil recoveries above that of waterflooding.
Other Flooding Configurations
In certain cores that were flooded with DI AS solution, the
fluids that were produced appeared to be a single-phase, black,
low viscosity liquid. This was observed in cores J, K, L and
O, which were the cores that had much lower rock
permeability values. For these cores, it is expected that
injection at the same rates into a lower permeability rock leads
to much higher levels of shear. In response, the oil droplets
should be smaller, thus the O/W emulsion can actually be
produced. The recovery mechanism in this case is that of
emulsification & entrainment of droplets in the flowing
continuous phase.
Effective mobility plots for secondary AS flooding in the
low permeability cores are presented in Figure 13. Similar to
the previous plots, during waterflooding Meff is around 50, and
after AS flooding Meff has been reduced to 20. The main
difference is that during emulsification & entrainment Meff
never decreases down to the order of unity, the way it does
during the entrapment plots. Accordingly, the recovery
efficiency values summarized in Tables 3 and 5 are lower than
for the other secondary AS floods. These results show that,
while emulsification & entrainment still leads to improved oil
recovery after waterflooding, it does not seem to be as
effective as entrapment.
One other flooding configuration is shown in Figure 14:
that of tertiary Brn AS flooding after secondary DI AS
flooding. In this plot, a waterflood was performed for
approximately five PV, followed by a DI AS flood for one PV.
As seen previously, during the DI AS injection the pressure in
the core increased and the oil recovery increased accordingly
as well. Figure 14 is for core O, which was a low permeability
core and showed production of an O/W emulsion. A salinity
buffer (2% brine) was then injected, followed by Brn AS
solution. Upon injection of the Brn AS solution, the pressure
in the core increased once again, and more oil was recovered.
The produced oil adhered strongly to the walls of the glass vial
(similar to the pictures in Figure 5), and the aqueous phase
was now clear. It is inferred based on these observations that
during Brn AS injection, a viscous W/O emulsion was
produced along with free water.
The recovery efficiency of the tertiary Brn AS injection is
shown in Table 5. While this is not as effective as secondary
Brn AS injection, this result shows that even after the
conclusion of a DI AS flood, by altering the salinity and
generating different emulsion systems further recovery can
still be achieved.
Discussion of Core Flooding Results
One important point that should be made regarding the DI
AS results in Figure 11 and Table 3 is that no O/W emulsion
was observed in the produced fluids, despite the fact that these
emulsions were generated in the bulk liquid tests. This was
also previously observed by other researchers19 in their core
floods. However, it is certain that the chemical is causing
some change in the system, due to the fact that pressure builds
up in the core, accompanied by a decrease in the produced

SPE 110738

water cut. It is the pressure and water cut response that is key
to understanding what is happening: emulsion droplets are
forming due to the low IFT and shear from flow, but these
droplets are not being entrained in the flowing aqueous phase.
Rather, they must be plugging off the water channels, and this
is leading in turn to improved sweep efficiency and a reduced
effective mobility ratio.
In systems with high shear, O/W emulsions were actually
produced under DI AS flooding. In these same systems, if
flooded with Brn AS solution after DI AS flooding, again the
pressure in the core built up but now the produced fluid
appeared to be a W/O emulsion. Therefore, although in many
cases actual emulsions were not observed in the produced
fluids, the bulk liquid experiments are still useful in order to
determine which type of emulsion should be forming in the
porous medium. It is possible that even under DI AS
conditions multiple emulsion types may be forming in the
core38, but the results from low permeability experiments
show that the bulk liquid emulsion systems can be used at
least as an indication of what should be happening in reservoir
applications.
Figure 15 plots the total recovery efficiency against the
minimum value of Meff for the secondary DI and Brn AS
floods, the entrainment secondary DI AS floods and the
tertiary Brn AS floods. In all cases, the total recovery
efficiency is directly correlated to the minimum value of Meff.
This shows that improved recovery is due to an improvement
in the mobility ratio of the system.
Figure 15 shows that the smallest values for the minimum
Meff occur for secondary DI and Brn AS floods. In Table 5,
these systems also have the highest recovery efficiency values.
During emulsification & entrainment or tertiary Brn AS
flooding, Meff is not reduced to such low values, and the
recovery efficiencies in Table 5 are likewise not as high.
The recovery efficiency values in Table 5 are the average
from all of the core floods performed in this study. The
largest standard deviation in repeated measurements is 0.014
RF/PV inj, so this was used as the error bar to determine if
floods had higher or lower recoveries when compared to one
another. If two systems were different by > 0.014, they were
said to be different in their recovery efficiency. Table 6
compares the results from waterflooding in DI vs Brn systems.
The results differ by less than this error bar, so it can be said
that waterflooding at the same rate is not affected by the
salinity of the injected fluid. When comparing waterflooding
to primary AS flooding, however, the results differ by a larger
amount. This indicates that primary AS floods are more
effective than waterfloods for both DI AS and Brn AS
systems.
Obviously, Table 5 shows that secondary AS floods are
much more efficient than primary AS floods. This can be
understood through a simple Jamin pressure drop
calculation30. If an emulsion droplet is surrounded on both
sides by a low IFT fluid, then the pressure drop to force this
droplet through a pore constriction is small. This is the state
of a primary AS flood after breakthrough of the AS solution.
Alternatively, if the emulsified droplet is contacted on the
back end with a low IFT fluid but on the front end with a high
IFT fluid, as is the case for AS flooding after waterflooding,
the Jamin pressure drop is significantly higher in this situation.

Therefore, trapping of emulsion droplets will be more

pronounced, leading to improved sweep and the dramatic
improvements in mobility ratio observed in Figures 11 and 12.
What is interesting in Tables 5 and 6 is the comparison of
DI AS floods to Brn AS floods. The difference between
primary floods is exactly 0.014 RF/PV inj, which is right on
the border of the error bar. In this case, it is not clear whether
there is a difference in the primary flood response between the
different AS systems. For secondary AS flooding, however, it
appears that DI AS flooding is more efficient than Brn AS
flooding. This could possibly be due to more stable IFT in
this system (Figure 1). Additionally, if relationships between
rock wettability and emulsion type hold, O/W emulsions and
DI AS systems are preferable because in these cases the rock
should still be more water wet, which can allow for extended
waterflooding36 or tertiary Brn AS floods.
If emulsification & entrainment leads to a reduced
recovery efficiency, then it is valuable to be able to predict
when shear forces are high enough that entrainment will
occur. The shear rate generated by injected fluid at a constant
rate is given by37:
4
actual =
(6)
k
Where actual = actual shear rate from flow at a fixed rate
= injection fluid velocity
k = absolute permeability of the rock
= rock porosity.
At any given injection rate, the actual shear will be a
constant value, according to Equation 6. However, pressure
builds up in the core and then declines, thus in reality the
fluids in the core are subject to a changing shear. Apparent
shear rate of the oil is therefore defined as the shear stress37,
divided by the oil viscosity:
app =

4 k P

L o

(7)

Dimensionless shear rate can therefore be expressed as the

ratio between the actual shear rate and the apparent shear rate:

(8)
dim = actual
app
Figure 16 plots the total recovery efficiency of different
floods vs. dimensionless shear rate. For waterflooding and
primary DI AS flooding, the most effective oil recovery region
occurs before the breakthrough of the water or AS solution. If
pressure in the system builds up further, this leads to improved
flow rates by Darcys Law. This explains the weak trend
showing that at higher dimensionless shear rate, the recovery
efficiency is slightly higher for these systems. For secondary
AS flooding, however, the trend is opposite. The best
recovery is achieved under conditions of lowest dimensionless
shear rate. This indicates that entrapment is in fact the main
recovery mechanism for these DI AS systems. At lower shear
rate, emulsified oil droplets are formed but the shear is not
high enough to force these droplets through the pore throats.
At elevated shear, the trapping of the oil droplets is reduced,

and emulsification & entrainment is the extreme form of these

high shear conditions.
Figure 17 compares the recovery efficiency for secondary
DI AS and secondary Brn AS floods, both as a function of
dimensionless shear rate. It can be seen that both systems
follow the same trend, which indicates that the recovery
mechanism is the same. Whether trapping is due to emulsified
oil or emulsified water droplets, in either case the recovery
mechanism is that of plugging of pore throats, giving
improved sweep efficiency and reduced effective mobility
ratios.
One last consequence of the fact that recovery is due to
entrapment of droplets is that this indicates that emulsion
droplets are on the same order of pore sizes. In this situation,
it is not correct to think of the production mechanism being
the flow of low viscosity O/W emulsions, as has been cited in
previous work19. This concept only applies to a situation
where the emulsion droplets are much smaller than the pore
sizes, meaning that the emulsion can flow as a pseudo-single
phase fluid. This is not observed in core floods that show
significant improvements in recovery.
Conclusions
The results from the suite of core floods tested in this work
demonstrate irrefutably that alkali-surfactant flooding has
great potential for non-thermal heavy oil recovery. AS
solutions reduce the interfacial tension between oil and water
to values that allow for emulsions to form under the shear
from flow at normal reservoir rates. These emulsions plug
rock pores, leading to improved sweep efficiency and
reductions in the effective mobility ratio between water and
heavy oil.
In the bulk liquid study carried out, different emulsion
types were created by changing the water salinity. It was
observed that for systems forming O/W emulsions, the walls
of the glass vials remained at least partially water wet. When
salinity was increased and W/O emulsions were formed, the
glass appeared much more strongly oil wet. This emulsion
behavior was used to interpret the results from core floods.
Both DI AS and Brn AS systems led to improved recovery
through entrapment of droplets. It appears that the recovery
mechanism is the same for both systems. Under conditions of
high shear and O/W emulsification, O/W emulsions could
actually be entrained and produced with the flowing aqueous
phase, but this mechanism is not as efficient as entrapment.
If possible, it is preferable to keep the rock water wet and
generate O/W emulsions, at least initially. This gives good
secondary recovery, while still leaving the rock in a state that
can potentially still see additional recovery through extended
waterflooding or tertiary Brn AS flooding.
Acknowledgements
The authors wish to acknowledge the contributions of Xiao
Dong Ji, Jun Gao and Michael Erath of TIPM Laboratory in
helping to gather the experimental data for this project, and to
An Mai for help with the analysis. The oil used in this study
was generously donated by Nexen Inc., and the surfactant was
provided by Stepan Company. Financial support was received
from NSERC, COURSE, ISEEE and the Canada Research
Chair in Energy and Imaging and its Industrial Affiliates

SPE 110738

(Shell, Nexen, Devon, PetroCanada, Canadian Natural, ET

Energy, Suncor, Schlumberger, Laricina and Paramount).
Nomenclature
A
= cross-sectional area of the core
AS
= alkali-surfactant
Brn
= brine
DI
= de-ionized water
IFT
= interfacial tension between oil and water
k
= absolute permeability of rock
kx
= effective permeability to oil (o) or water (w)
L
= length of the core
Meff
= effective mobility ratio between water and oil
O/W
= oil-in-water emulsion
Pc
= capillary pressure (Poil Pwater)
Pemul
= net work per unit volume of an emulsion droplet
PV
= pore volume
Q
= flow rate of oil or water
r
= radius of an emulsion droplet
R
= radius of a pore or capillary tube
RF
= recovery factor
W/O
= water-in-oil emulsion

actual
app
dim
P

x
ow
os
ws

=
=
=
=
=
=
=
=
=
=
=

actual shear rate

apparent shear rate (shear stress/oil viscosity)
dimensionless shear rate
pressure drop across the core
contact angle of the water-solid interface
viscosity of oil (o) or water (w)
oil-water interfacial tension
oil-solid interfacial free energy
water-solid interfacial free energy
rock porosity
injection fluid velocity

References
1.

2.

3.

4.

5.

6.

7.

8.

Firoozabadi, A.: Mechanisms of Solution Gas Drive in

Heavy Oil Reservoirs, J. Can. Pet. Tech. (Mar 2001), 40
(3), pg. 15.
Nutting, P.G.: Chemical Problems in the Water-Driving of
Petroleum Reservoirs, Ind. And Eng. Chem. (Oct 1925),
17, pg. 1035.
Reisberg, J. and Doscher, T.M.: Interfacial Phenomena in
Crude Oil-Water Systems, Prod. Monthly (Nov 1956), 21,
pg. 43.
Rosen, M.J.: Surfactants and Interfacial Phenomena, 2nd
Edition, John Wiley & Sons, Inc., New York USA (1989),
pg. 3.
Schramm, L.L.: Petroleum Emulsions Basic Principles,
Emulsions Fundamentals and Applications in the
Petroleum Industry, Schramm, L.L., ed., Advances in
Chemistry Series 231, American Chemical Society,
Washington, DC USA (1992), pg. 21.
Moore, T.F. and Slobod, R.L.: The Effect of Viscosity and
Capillarity on the Displacement of Oil by Water, Prod.
Monthly (Aug 1956), pg. 20.
Wagner, O.R and Leach, R.O.: Effect of Interfacial
Tension on Displacement Efficiency, SPE 1564, SPEJ
(Dec 1966), pg. 335.
Cooke, C.E. Jr., Williams, R.E. and Kolodzie, P.A.: Oil
Recovery by Alkaline Waterflooding, J. Pet. Tech. (Dec
1974), 16, pg. 1365.

SPE 110738

9.

10.
11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

Johnson, C.E. Jr.: Status of Caustic and Emulsion

Methods, SPE Paper 5561, J. Pet. Tech .(Jan 1976), pg.
85.
Mungan, N.: Role of Wettability and Interfacial Tension
in Water Flooding, SPE 705, SPEJ (Jun 1964), pg. 115.
Wagner, O.R and Leach, R.O.: Improving Oil
Displacement Efficiency by Wettability Adjustment, SPE
1101-G, Pet. Trans. AIME (1959), 216, pg. 65.
Scott, G.R., Collins, H.N. and Flock, D.L.: Improving
Waterflood Recovery of Viscous Crude Oils by Chemical
Control, J. Can. Pet. Tech. (Oct Dec 1965), 4 (10), pg.
243.
Thomas, S., Farouq Ali, S.M., Scouler, J.R. and Verkoczy,
B.: Chemical Methods for Heavy Oil Recovery, J. Can.
Pet. Tech. (Mar 2001), 40 (3), pg. 56.
Pitts, M.J., Wyatt, K., and Surkalo, H.: Alkaline-Polymer
Flooding of the David Pool, Lloydminster Alberta, SPE
Paper 89386, 2004 SPE/DOE 14th Symposium on
Improved Oil Recovery, Tulsa, Oklahoma USA, April 17
21 2004.
Dranchuk, P.M., Scott, J.D. and Flock, D.L.: Effect of the
Addition of Certain Chemicals On Oil Recovery During
Waterflooding, J. Can. Pet. Tech. (Jul Sept 1974), pg.
27.
DElia, R. and Ferrer, J.: Emulsion Flooding of Viscous
Oil Reservoirs, SPE Paper 4674, 48th Annual Fall Meeting
of the Society of Petroleum Engineers of AIME, Las
Vegas, Nevada USA, Sept 30 Oct 3 1973.
McAuliffe, C.D.: Crude-Oil-in-Water Emulsions To
Improve Fluid Flow in an Oil Reservoir, J. Pet. Tech. (Jun
1973), pg. 721.
Jennings, H.Y., Johnson, C.E. and McAuliffe, C.D.: A
Caustic Waterflooding Process for Heavy Oils, SPE 4741,
J. Pet. Tech. (Dec 1974), pg. 1344.
Liu, Q., Dong, M. and Ma., S.: Alkaline/Surfactant Flood
Potential in Western Canadian Heavy Oil Reservoirs, SPE
99791, 2006 SPE/DOE Symposium on Improved Oil
Recovery, Tulsa, Oklahoma USA, April 22 26 2006.
Mai, A., Bryan, J., Goodarzi, N. and Kantzas, A.: Insights
Into Non-Thermal Recovery of Heavy Oil, WHOC Paper
2006-533, 1st World Heavy Oil Conference, Beijing China,
Nov 12 15 2006.
Bryan, J. and Kantzas, A.: Potential for Alkali-Surfactant
Flooding in Heavy Oil Reservoirs Through Oil-in-Water
Emulsification, CIM Paper 2007-134, 58th Annual
Technical Meeting of the Petroleum Society, Calgary, AB
Canada, Jun 12 14 2007.
Farouq Ali, S.M., Figueroa, J.M., Azuaje, E.A., and
Farquharson, R.G.: Recovery of Lloydminster and
Morichal crudes by caustic, acid and emulsion floods, J.
Can. Pet. Tech. (Jan Mar 1979), pg. 53.
Atkinson, I.: Interfacial Phenomena and Phase Behavior
Using Alkali-Surfactant-Polymer Enhanced Recovery
Methods in Medium and Heavy Crude Oils, M.Sc. Thesis
(2004), University of Calgary, Canada, pg. 122.
Martin, F.D. and Oxley, J.C.: Effect of Various Alkaline
Chemicals on Phase Behavior of Surfactant/Brine/Oil
Mixtures, SPE Paper 13575, International Symposium on
Oilfield and Geothermal Chemistry, Pheonix, Arizona
USA, Apr 9 11 1985.
Nelson, R.C., Lawson, J.B., Thigpen, D.R. and Stegemeier,
G.L.: Cosurfactant-Enhanced Alkaline Flooding,
SPE/DOE Paper 12672, SPE/DOE Fourth Symposium on
Enchanced Oil Recovery, Tulsa, OK USA, Apr 15 18
1984.

26. Isaacs, E.E. and Chow, R.S.: Practical Aspects of

Emulsion Stability, Emulsions Fundamentals and
Applications in the Petroleum Industry, Schramm, L.L.,
ed., Advances in Chemistry Series 231, American
Chemical Society, Washington, DC USA (1992), pg. 51.
27. Garrett, H.E.: Surface Active Chemicals, Pergamon Press,
Oxford England (1972), pg. 26.
28. Acevedo, S., Gutierrez, X., and Rivas, H.: Bitumen-inWater Emulsions Stabilized with Natural Surfactants, J.
Colloid and Interface Science (2001), 242, pg. 230.
29. Liu, Q., Dong, M., Yue, X., and Hou, J.: Synergy of alkali
and surfactant in emulsification of heavy oil in brine,
Colloids and Surfaces A: Physicochem. Eng. Aspects
(2006), 273, pg. 219.
30. Dullien, F.A.L.: Porous Media Fluid Transport and Pore
Structure, 2nd Edition, Academic Press, Inc., San Diego,
California USA (1992), pg. 118.
31. Jamaluddin, A.K.M. and Butler, R.M.: Factors Affecting
the Formation of Water-in-Oil Emulsions During Thermal
Recovery, AOSTRA J. Research (1988)., 4 (2), pg. 109.
32. Vittoratos, E.: Flow Regimes During Cyclic Steam
Stimulation in Cold Lake, J. Can. Pet. Tech. (1991), 30,
pg. 82.
33. Chen, T., Chakrabarty, T., Cullen, M.P., Thomas, R.R. and
Sieben, M.C.: Laboratory and Field Evidence of Emulsion
Flow in Porous Media, CIM/AOSTRA Paper 91-78, 1991
CIM/AOSTRA Technical Conference, Banff, AB Canada,
Apr 21 24 1991.
34. Cuthiell, D., Green, K., Chow, R., Kissel, G., and
McCarthy, C.: The In Situ Formation of Heavy Oil
Emulsions, SPE 30319, 1995 International Heavy Oil
Symposium, Calgary, AB Canada, Jun 19 21 1995.
35. Chen, T., Yuan, J.Y., Serres, A.J., and Rancier, D.G.:
Emulsification of Water into Bitumen in Co-injection
Experiments, J. Can. Pet. Tech. (Special Ed. 1999), 38
(13), pg. 1.
36. Mai, A. and Kantzas, A.: Heavy Oil Waterflooding:
Effects of Flow Rates and Oil Viscosity, CIM Paper 2007144, 58th Annual Technical Meeting of the Petroleum
Society, Calgary, AB Canada, Jun 12 14 2007.
37. Alvarado, D.A. and Marsden, S.S. Jr.: Flow of Oil-inWater Emulsions Through Tubes and Porous Media, SPE
Paper 5989, SPE Journal (Dec 1979), pg. 369.
38. Mai, A., Bryan, J. and Kantzas, A.: Investigation Into the
Mechanisms of Heavy Oil Recovery by Waterflooding and
Alkali-Surfactant Flooding, International Symposium of
the Society of Core Analysts, Calgary, AB Canada, Sept 9 14 2007

10

SPE 110738

Table 1. DI AS Core Flood Properties

Core ID D (m)
A
0.011
B
0.011
C
0.011
D
0.011
E
0.011
F
0.011
G
0.011
H
0.011
I
0.011
J
0.011
K
0.011
L
0.011
M
0.036
N
0.036
O
0.026

L (m)
0.103
0.103
0.103
0.103
0.103
0.103
0.103
0.103
0.103
0.105
0.105
0.105
0.211
0.211
0.532

0.45
0.45
0.45
0.45
0.45
0.45
0.45
0.45
0.45
0.49
0.49
0.49
0.45
0.45
0.41

Table 3. DI AS Flood Recovery Efficiency

WF rate AS rate
o
(mPas) (m/d) (m/d)
11000
0.26
11000
0.51
15000
0.51
11000 0.51
11000 0.51
0.51
15000 0.51
0.26
15000 0.51
1.02
15000 0.51
2.05
15000 0.51
2.05
11000
0.51
15000 0.51
0.51
15000 0.51
0.26
11000 0.42
0.42
11000 0.023 0.42
11000 0.51
0.51

k (D)
19.22
16.25
14.98
19.98
15.84
12.71
12.72
12.05
11.44
1.19
0.91
0.97
8.7
8.7
0.79

Core ID
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O

WF total AS
total
RF/PV inj RF/PV inj
0.098
0.064
0.046
0.037
0.035
0.127
0.043
0.133
0.038
0.093
0.037
0.095
0.031
0.094
0.066
0.050
0.066
0.047
0.055
0.050
0.512
0.051
0.208
0.048
0.055

Table 2. Brn AS Core Flood Properties

Table 4. Brn AS Flood Recovery Efficiency

Core ID D (m)
Q
0.011
R
0.011
S
0.011
T
0.011
U
0.011
V
0.011
W
0.011
X
0.011
Y
0.011
Z
0.011
AA
0.011
AB
0.011
AC
0.036

L (m)
0.103
0.103
0.103
0.103
0.103
0.103
0.103
0.103
0.105
0.105
0.105
0.105
0.211

0.45
0.45
0.45
0.45
0.45
0.45
0.45
0.45
0.49
0.49
0.49
0.49
0.45

k (D)
15.98
11.04
15.9
11.74
17.94
11.55
11.81
10.24
1.48
1.19
0.92
0.97
9.05

WF rate AS rate
o
(mPas) (m/d) (m/d)
11000
0.51
11000
0.51
11000
0.51
11000
0.51
11000 0.51
0.51
11000 0.51
0.51
15000 0.51
0.26
15000 0.51
1.02
11000
0.51
15000
0.51
15000 0.51
0.51
15000 0.51
0.26
11000 0.42
0.42

Core ID
Q
R
S
T
U
V
W
X
Y
Z
AA
AB
AC

WF total AS
total
RF/PV inj RF/PV inj
0.089
0.109
0.076
0.086
0.029
0.109
0.030
0.121
0.041
0.200
0.034
0.188
0.065
0.070
0.035
0.123
0.026
0.098
0.045
0.237

Table 5. Summary of Recovery Efficiency from Different Flood

Systems

System
DI waterflood
Prim DI AS
Sec DI AS
Entrainment
Brn waterflood
Prim Brn AS
Sec Brn AS
Tert Brn AS

Total
RF/PV inj
0.042
0.069
0.180
0.062
0.034
0.083
0.148
0.061

SPE 110738

11

Table 6. Comparison of Different Emulsion Systems

System
DI WF vs Brn WF
DI WF vs Prim DI AS
Brn WF vs Prim Brn AS
Prim DI AS vs Prim Brn AS
Sec DI AS vs Sec Brn AS

Difference
No
Yes
Yes
Yes?
Yes

Equilibrium IFT (mN/m)

0 ppm NaCl
20 000 ppm NaCl

Figure 3. DI samples with varying alkali before agitation

0.1

0.01
0

5000

10000

15000

20000

Alkali Content (ppm)

Figure 1. Oil-water IFT with varying alkali content

Figure 4. DI samples with varying alkali after agitation

Figure 2. Phase behaviors - control and no alkali

Figure 5. Brn samples with varying alkali after agitation

12

SPE 110738

0.25

1400
1200

500
Pressure
Oil RF

300

0.1

200
0.05

100
0

0
0

1000
800

0.4

600

0.3

400

0.2

200

0.1

0
0

PV Injected

500

0.7

Pressure

0.6

Water cut

0.5

300

0.4

200

0.3
0.2

100

0
4

0.9
0.8

1000

0.7
Pressure

800

0.6

Water cut

0.5

600

0.4

400

0.3
0.2

200

0.1

1200
Pressure (kPa)

Pressure (kPa)

0.8

Water Cut (fraction)

600

0.1

0
0

PV Injected

PV Injected

Figure 7. Heavy oil waterflood - pressure and water cut

Figure 10. Secondary Brn AS flood - water cuts

700

1000
Effective Mobility Ratio

600
Pressure (kPa)

1400

0.9

Figure 9. Secondary Brn AS flood - pressure and recovery

700

4
PV Injected

Figure 6. Heavy oil waterflood - pressure and recovery

400

0.5

Oil RF

Water Cut (fraction)

400

0.15

0.6
Pressure

Pressure (kPa)

0.2

Oil Recovery Factor

600
Pressure (kPa)

0.7

Oil Recovery Factor

700

500
Waterflood
400

AS Flood

300
200

100
10
1
0.1
0.01

100

0.5

1.5

2.5

Norm alized PV Injected

PV Injected

Figure 8. Waterflood vs. primary AS flood pressure response

Core M AS

Core N AS

Core E AS

Core G AS

Core H AS

Core I AS

Figure 11. DI AS floods - effective mobility ratio

Core F AS

13

1000

0.3

100

0.25

Total Oil RF/PV Inj

Effective mobility ratio

SPE 110738

10
1
0.1

0.5

1.5

0.15
0.1
0.05
0
0.01

0.01
0

0.2

2.5

Norm alized PV Injected

Core V

Core W

Core AB

Core AC

Core AA

Figure 12. Brn AS floods - effective mobility ratio

10

100

Secondary AS Floods

Entrainment

Secondary Brn AS

Tertiary Brn AS

100
10
1
0.1
0.01
0.001
0

0.5

1.5

2.5

Total Oil Recovery/PV Injected

Effective mobility ratio

Figure 15. AS flooding post WF - recovery efficiency

1000

Waterflood
Primary AS
Secondary AS
Entrainment
0.1

0.01

Normalized PV Injected

1
Core K AS

Figure 13.
ratio

0.1

Minimum Effective Mobility Ratio

Core L AS

10

100

Dimensionless Shear Rate

Core O AS

Emulsification and entrainment - effective mobility

Figure 16. DI systems - recovery efficiency vs. shear rate

0.4
Pressure

0.35

Pressure (kPa)

Oil RF

0.3

10000

0.25

8000

0.2
6000

0.15

4000

0.1

2000

0.05

Oil Recovery Factor

12000

Total Oil Recovery/ PV Injected

14000

0.1

Secondary DI AS
Secondary Brn AS
0.01

0
0

10

10

100

Dimensionless Shear Rate

PV Injected

Figure 14. Pressure and recovery response for tertiary Brn AS

flooding

Figure 17. Secondary DI AS and Brn AS flood recovery efficiency