SPE 144968

Alkaline Steam Foam: Concepts and Experimental Results

Hon Chung Lau, Shell International Exploration and Production

Copyright 2011, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Enhanced Oil Recovery Conference held in Kuala Lumpur, Malaysia, 19–21 July 2011.

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

Abstract

Experimental studies have shown that the steam foam process can be significantly enhanced by injecting a suitably formulated
alkali-surfactant mixture in the aqueous phase of steam. Emulsion screening tests, corefloods and flow visualization
experiments using an alpha olefin sulfonate (AOS) surfactant with 16 to 18 carbon, Na2CO3 and a heavy Californian crude
have shown that alkaline steam foam offers significant advantages over regular steam foam by combining the benefits of
thermal and chemical enhanced oil recovery (EOR) processes.

Firstly, Na2CO3 reduces surfactant consumption by adsorption by rendering the clay surface more negatively charged.
Secondly, by precipitating divalent ions that get ion-exchanged off the clays, Na2CO3 reduces surfactant consumption by
precipitation. Thirdly, a suitably formulated alkali-surfactant system reduces the oil-water interfacial tension (IFT) sufficiently
to enable the heavy oil to be emulsified into the aqueous phase in the presence of steam. This oil-in-water emulsion is less
viscous than that of the oil and can be readily transported. Consequently, the residual oil saturation is reduced to that below
steam. Fourthly, this lower residual oil saturation reduces the destabilizing effect of oil on foam resulting in stronger steam
foam that provides better mobility control than regular steam foam. Therefore, it has the potential to further reduce steam
gravity override. Fifthly, the reduction in gravity override also reduces loss of heat to the cap rock. Steam utilization is
thereby improved.

When applied to steam drive, cyclic steam soaks and steam-assisted gravity drainage (SAGD) processes, alkaline steam foam
has the benefits of increasing foam propagate rate, improving mobility control, improving steam utilization and reducing the
residual oil saturation. In the cases of steam soaks and SAGD, it also increases the gravity drainage rate of oil by reducing the
effective oil viscosity through emulsification.

Introduction

It is well known that steam drive and cyclic steam soaks suffer from gravity override of steam due to the unfavorable mobility
of the steam. Use of steam foam1-5 to mitigate this problem has been studied in the laboratory and field piloted since the
1980’s. However, so far steam foam has not gained wide usage because it requires propagation of the surfactant through
significant distances into the reservoir. This process is inherently slow and expensive due to surfactant consumption by
adsorption on clay surfaces, precipitation by divalent ions and partitioning into the oleic phase.

By combining the benefits of both thermal and chemical EOR processes, alkaline steam foam6-9 offers significant
improvements over regular steam foam. This paper documents experimental studies that showed how these improvements can
be achieved.

Alkaline Steam Foam Concept

The concept behind an alkali-surfactant chemical EOR process is that the injected alkali reacts with the petroleum acid in the
crude to form petroleum soap which acts as the primary surfactant. The injected surfactant acts as a cosurfactant to raise the
electrolyte concentration required for minimum IFT to alkali concentrations high enough for satisfactory propagation of the
alkali through the reservoir10. At suitable concentrations of primary surfactant and cosurfactant, an optimum system is formed
2 SPE 144968

where the oil-water IFT is significantly reduced by the formation of an optimum microemulsion stystem. This enables oil to
be miscibly displaced, thus reducing the residual oil saturation (ROS). When this concept is combined with the use of steam
foam for mobility control, the injected surfactant plays two roles. Firstly, it is the surfactant to stabilize the steam foam.
Secondly, it acts as the cosurfactant in the alkaline-surfactant system for ROS reduction.

Emulsion Screening Test

The ability of an alkali-surfactant system to form microemulsions with a heavy Californian crude from the San Joaquin Valley
was investigated by using test tube emulsion screening experiments. The San Joaquin crude used had an oil gravity of 13oAPI
and an acid number of 2.5 mg KOH/g (Table 1). The surfactant used was ENORDET AOS 1618 with 1 wt% NaCl. The alkali
used was Na2CO3. In the beginning of each experiment, a surfactant solution containing predetermined amounts of AOS,
NaCl, and Na2CO3 was prepared. Then 12.5 ml of the surfactant were put into a graduated vial. The solution was degassed by
subjecting it to a vacuum for 15 minutes. A nitrogen blanket was then put on top of the surfactant to exclude contact with
oxygen. Hot crude was then added to the vial to arrive at an oil-to-water ratio of 1:1 or 1:3. Again the vial was subjected to a
vacuum for 15 minutes to remove any dissolved gas in the oil. A nitrogen blanket was put on top of the oil and the vial was
sealed with an aluminum cap. These steps were taken to prevent the surfactant from contacting oxygen. A number of
surfactant samples, thus prepared, were then put in an oven and heated to 212oF for six days. The samples were shaken twice
daily except for the sixth day. At the end of six days, a visual inspection of the samples was made to determine the phase
volumes and whether the each sample was under optimum, optimum and over optimum.

An under optimum emulsion is one where surfactant resides mostly in the aqueous phase. As the system becomes more
optimum, increasing amount of oil is solubilized into the aqueous phase. An optimum system is one where a distinct middle
microemulsion phase is formed which consists of surfactant and solubilized oil. An over optimum system is one where the
surfactant preferentially partitions into the oleic phase.

Resultant activity diagrams are shown in Figs. 1 and 2. Results are summarized as follows. Firstly, the interfacial activity is
very sensitive to the oil-to-water ratio. When this ratio increases, the ratio of primary surfactant (petroleum soap) to
cosurfactant (AOS) increases. The system becomes more over optimum. The active region shifts downward to lower Na2CO3
concentrations. Secondly, at a fixed temperature, oil-to-water ratio and NaCl concentration, the Na2CO3 concentration at which
the system is optimum increases with increasing AOS concentration. This is because with increasing AOS concentration the
ratio of primary surfactant to cosurfactant decreases. The system becomes more under optimum. The active region shifts
upward to higher Na2CO3 concentrations. These results are consistent with those obtained by Nelson et al in their work with
cosurfactant-enhanced alkaline flooding10. They have important ramifications.

In a steam foam process, there exist three distinct regions from injector to producer (Fig. 12). Near the injector, there exists an
alkali steam foam region. Downstream of it is the oil bank. Downstream of the oil bank is the unswept region. In the unswept
region, the oil-to-water ratio may be closer to 1-to-3 if it has undergone a hot water flood. In the oil bank, the oil-to-water
ratio is one or higher. In the alkaline steam foam region, the oil-to-water ratio is lowest as the oil has been mobilized to form
the oil bank. Hence the system goes from under optimum in the unswept zone, to optimum or over optimum in the oil bank
and then to under optimum again in the alkaline foam region. There exists therefore in the oil bank, an optimum or near
optimum region with very low oil-water IFT (10-2-10-3 dyne/cm) capable of recovering additional oil over and above that due
to steam or regular steam foam.

However, it is worthwhile to note that attainment of this very low interfacial tension may not be necessary for reduction of
residual oil saturation. In the flow visualization experimental section, it will be shown that alkaline steam foam generates a
higher pressure gradient than regular steam foam. This increase in pressure gradient coupled with mixing energy of the steam
and a moderate reduction in oil-water IFT, say, from 30 to 0.1-1.0 dyne/cm may be adequate to emulsify the oil into the
aqueous phase thus lowering the ROS.

Surfactant Propagation Experiments

A coreflood was designed to study if alkali could reduce surfactant losses due to adsorption and precipitation on San Joaquin
reservoir sand. The reservoir sand had a cation exchange capacity of 9.87 meq/100g (Table 1). The experiment was
performed with the reservoir sand packed in a cylindrical stainless steel vessel 1 inch in diameter and 12.25 inch in length. It
had a permeability of 4 darcys and porosity of 36%. Experiments were conducted with the sand pack positioned horizontally,
at 212oF with a 10 psig backpressure. Fluids were injected at a frontal rate of 2 ft/day. This frontal rate was calculated by
multiplying the length of sand pack by the flow rate and dividing the product by the pore volume. Before the beginning of the
experiment, the sand pack was saturated with synthetic connate water which contained 0.0024 meq/g CaCl2, 0.0008 meq/g
MgCl2 and 0.0168 meq/g NaCl. At this composition, about 80 percent of the clays were in calcium form. No oil was used.
SPE 144968 3

The sand pack was flooded with 2.25 PV of 2.65 wt% Na2CO3. This was followed by continuous injection of 0.5 wt%
ENORDET AOS 1618 and 1 wt % NaCl. The effluent was collected and analyzed for surfactant, chloride and divalent ion
concentrations by titration. Results, plotted in Fig. 3, can be summarized as follows.

Firstly, the produced water was free of calcium after the connate water was displaced. In the absence of alkali, a calcium peak
of 2000 ppm was observed with the same surfactant formulation. This indicated that the calcium ions that got ion-exchanged
off the clays were precipitated by the carbonate ions in the injected fluid. Secondly, the surfactant front propagated at the
same speed as the chloride front indicating that there was no precipitation or adsorption. These results indicate that Na2CO3
can increase the speed of propagation of the AOS surfactant significantly by two suppressing adsorption and precipitation.

Residual Oil Saturation Experiments

Tertiary corefloods were performed in Ottawa sand packs to study the effect of Na2CO3 and ENORDET AOS 1618 on ROS.
In these experiments a one-foot-long, 1.5 inch diameter pressure vessel was packed under water with grade F-140 Ottawa
sands. The absolute permeability of the sand pack was 4 darcys, porosity was 0.31, and PV was 111 ml. The sand pack,
positioned vertically, was heated to 212oF in an oven with a back pressure of 100 psig. It was saturated with the San Joaquin
crude oil and then flooded with distilled water until no oil was produced. Fifty percent mass quality steam was injected into the
sand pack at a flow rate of 3 cc/min cold water equivalent (CWE). Five PVs of liquid (10 PV CWE) were injected.

This was followed by injection of steam foam consisting of 50% quality steam with a 0.006 mole fraction nitrogen in the vapor
phase, and predetermined concentrations of Na2CO3, AOS and NaCl in the aqueous phase. A flow rate of 3 cc/min (CWE)
was used. Experiments were terminated when 15 PV of liquid phase had been injected.

Results, given in Table 2 and Fig. 4, are summarized as follows. Firstly, alkaline steam foam formed by AOS concentration
of 0.05 wt% or higher and Na2CO3 concentration of higher than 1 wt% gave substantially lower ROS (0-5%) than that to
steam (13%). In several cases, a zero ROS was observed. Secondly, above these concentrations, ROS to alkaline steam foram
was not sensitive to AOS and Na2CO3 concentrations. Thirdly, ROS to alkaline steam foam was insensitive to oil saturations
at the start of foam injection.

These results are encouraging. They show that alkaline steam foam can reduce ROS over a wide range of Na2CO3 and AOS
concentrations. This suggests that the attainment of very low IFT is not crucial for lowering the ROS. To investigate the
reason for this, flow visualization experiments were conducted to observe both the microscopic and macroscopic flow pattern
of alkaline steam foam.

Flow Visualization Experiments

A diagram of the apparatus is shown in Fig. 5. The visual model had a working temperature of 400oF and working pressure of
300 psig. Visual observations, 35 mm photographs and video recordings were made through a 1.25 inch thick tempered Pyrex
glass.

The model was constructed with Inconel 600 with a 1.25 inch thick tempered Pyrex glass in the front and three
injection/production ports at each end (Figs. 6 and 7). The injection/production ports were fitted with 40-micron frits. The
model was heated to 400oF by cartridge heaters inserted into the model. Because the heat requirement varied within the
model, the heaters were split into several banks with individually controlled power stats to maintain near adiabatic conditions.
To diminish heat loss, a box constructed of Maronite surrounded the model. The space between the model and the box was
filled with vermiculite, another insulating material. A Maronite slab was placed on top of the vermiculite. Maronite blocks
were used to cover the Pyrex glass during heating.

The temperature and pressure at several locations in the sand pack were measured through 15 openings drilled through the
back side of the model (Fig. 6). These openings were fitted with 20-micron frits for pressure measurements. Temperature was
measured with thermocouples inserted through the middle of each frit. Pressure and temperatures were recorded on chart
recorders.

San Joaquin outcrop sand was packed manually in approximately 14 layers of 0.24 inch thickness with the top of sand visible
through the glass. After compaction with the aid of a steel bar and rubber mallet, each layer was disturbed before the next
layer was laid down to avoid sorting by grain size. A 200 psig overburden pressure was applied to a rubber bladder through the
top Inconel cover to compact the sand further. The sand pack was saturated with a least 1 PV of CO2 to displace trapped air,
followed by synthetic connate water. The sand was sealed with Dow Corning 3210 RTV with 10% catalyst S, followed by a
5% catalyst mixture. Copper sheeting, Viton rubber and the Inconel top cover were placed directly on top of the RTV after it
had cured for 12 hours. The model was assembled with bolts and fastened. Microscopic flow behavior was observed through
4 SPE 144968

a Zeiss microscope placed in front of the Pyrex glass. The microscope was equipped with a video camera, a time lapse video
recorder and a TV monitor for videotaping. A 35 mm camera was used to take still photographs. Macroscopic flow behavior
was recorded using the same camera and video system, but with the camera mounted on a tripod position 6-ft from the flow
apparatus.

Several experiments were conducted to study the flow behavior of regular and alkaline steam foams. In these experiments, the
model was packed with San Joaquin outcrop sand, saturated with synthetic connate water and heated to 200oF with a 5 psig
backpressure. Absolute permeability of the sand pack ranged from 0.4-4 darcys due to difference in packing. Average
porosity was 34 percent. At the beginning of each experiment, the sand pack was saturated with the San Joaquin crude oil and
then steam flooded until residual oil saturation to steam was established. Steam foam was then injected at velocities which
simulated near-injector conditions.

Regular Steam Foam Experiment

The surfactant solution consisted of 0.5% ENORDET AOS 1618 and 4 wt% NaCl. Vapor phase consisted of 50% quality
steam containing a 0.006 mole fraction nitrogen.

Steam Drive. Initial oil saturation was 49%. Residual oil saturation to steam was 15.7% after 12 PV of water (24 PV
CWE) was injected.
Steam Foam Drive-Macroscopic observations. Foam formed readily upon injection of foam components as evidenced by
the immediate appearance of foam near the injector followed by an increase in pressure at the first pressure tap (Fig. 8). A
distinct foam bank was observed within which the oil saturation was low. The foam first appeared as a finger but soon became
piston-like. After 0.6 PV of surfactant was injected, the foam began to under run, i.e. it propagated faster at the bottom of the
pack than at the top. It is worthwhile to note that under running of foam was observed in the Mecca and Bishop steam foam
pilots1. Produced oil was in the form of an oil-in-water emulsion. Surfactant broke through at 1.5 PV of surfactant injected.
Foam broke through at 1.6 PV indicating significant retardation of the surfactant and foam fornts. At foam breakthrough, there
was a 50 psi/ft pressure gradient across the sand pack. The pressure gradient continued to increase and stabilized at about 78
psi/ft. The average foam front propagation rate, calculated from the pressure response, was 14 ft/day whereas surfactant was
injected at a frontal rate of 19 ft/day. The remaining oil saturation during the steam foam injection was determined
periodically by material balance. Oil saturation was normalized to ROS to steam at the beginning of the experiment and
plotted in Fig. 10. Residual oil saturation to steam foam was measured after 12 PV of surfactant was injected. The sand pack
was divided into nine sections of equal volume for analysis. Results are shown in Fig. 11. The lower ROS of the bottom
sections was probably due to foam under running. Overall ROS to steam foam was 14.2% compared to 15.7% for steam.
Steam Foam Drive-Microscopic observations. During steam foam drive, observations through the microscope showed that
a single location witnessed the following sequence of events: unswept region, a region of high oil saturation or oil bank, an
emulsion bank in which emulsification took place, and a steam foam region. The mobile oil ahead of the foam front moved in
the form of ganglia, often continuous, and snapped off from pore to pore. The oil bank was visible macroscopically as a
distinct region of high oil saturation downstream of the foam front. The emulsion bank was not obvious in macroscopic
observations. However, observations through the microscope showed that the primary mechanism for oil displacement was
emulsification of the oil by the surfactant. Emulsified oil droplets often were smaller than the pore throat size and flowed
through the aqueous phase. Bulk foam with bubbles smaller than a pore channel was observed. Occasionally the foam
lamellae would come to a complete halt, plugging the flow channel and diverting steam and surfactant flow. Emulsified oil
traveled through the aqueous phase and through the foam lamellae. Oil did not spread at the gas-water interface.

Alkaline Steam Foam Experiment

The surfactant consisted of 0.33% AOS 1618, 9% wt. Na2CO3. The vapor phase consisting of 50% quality steam with 0.006
mole fraction nitrogen. Steam and surfactant were injected at the same flow rates as in the regular steam foam experiment
(Table 3).

Steam Drive. Initial oil saturation was 51%. Remaining oil saturation to steam flood was 27% after 8.3 PV of water
(16.6PV CWE). The difference in the remaining oil saturation to steam between this and the previous experiment was
probably caused by difference in packing and PV of steam injected.
Alkaline Steam Foam: Macroscopic Observations. Foam was visible immediately upon injection of foam components.
The front began to under run after less than 0.1 PV of surfactant was injected. The leading edges of the foam front fingered
into higher permeability regions until plugging diverted the foam to less permeable regions. A distinct oil bank was visible in
front of the foam (Fig. 12). Pressure response at all five pressure taps occurred within the first 0.2 PV of surfactant injected
(Fig. 9). The average foam front propagated at 50 ft/day compared to 14 ft/day for regular steam foam, as measured by the
pressure response. Oil broke through after 0.26 PV of surfactant injected, foam at 0.4 PV. At foam breakthrough the pressure
gradient was 89 psi/ft across the pack. The pressure gradient reached 115 psi/ft when the experiment was terminated. This
SPE 144968 5

indicated alkaline steam foam is stronger than regular steam foam. Remaining oil saturation during the experiment was
measured by material balance and normalized with respect to remaining oil saturation to steam at the beginning of the
experiment. Results are plotted in Fig. 10. Ninety one percent of the oil produced from alkaline steam foam was produced
during the first PV of injection. Residual oil saturation to alkaline steam foam was measured after 3.45 PV of surfactant was
injected. The sand pack was divided into nine sections for ROS determination. Results, shown in Fig. 11, indicate that foam
was well distributed throughout the sand pack. Overall ROS to alkaline steam foam was 4.0%.
Alkaline Steam Foam: Microscopic Observations. Observation through a microscope positioned at a fixed location
witnessed the unswept zone and then the arrival of an oil bank, an emulsion bank and then steam foam. In the oil bank, the oil
moved through pores in the form of ganglia which were often continuous. The presence of emulsion bank just ahead of the
steam foam front was evidenced by the emulsification of oil into small droops. Emulsification was noticeably quicker and
more efficient (smaller droplets) than that in regular steam foam. Oil droplets in the aqueous phase flowed easily through pore
throats. Emulsification became more rigorous upon the arrival of the foam front. Foam texture most often was not
homogeneous from pore to pore. Bubble size appeared to decrease with increasing interstitial velocity of the aqueous phase.
Bubble size of the alkaline steam foam was in general smaller than that of regular steam foam. Both stationary and moving
bubbles were observed. Moving bubbles remained distinct and spherical at high velocities. The oil did not spread at the gas-
water interface. The sand appeared to have a very low oil saturation shortly after passage of the foam front.

Discussion

Comparison of results of flow visualization experiments with regular and alkaline steam foam is given in Table 3. Residual
oil saturation to alkaline steam foam was 4% compared to 14% for regular steam foam. These values agreed well with those
measured from ROS experiments conducted in 1.5-inch diameter tubular sand pack using Ottawa sands under similar linear
superficial velocities. Furthermore, results indicate that alkaline steam foam propagates at a much faster rate than regular
steam foam. This increase in foam propagation rate was due, in part, to reduction in surfactant precipitation and adsorption.
Our microscopic observations also suggest that increase in foam propagation was also due to the very rapid emulsification of
oil by the alkali-surfactant system, thus reducing the local oil saturation to a sufficiently low value for foam generation. This
is because foam cannot be formed at a high oil saturation because oil destabilizes foam. Also the non-spreading behavior of
oil at the gas-water interface is another reason for strong steam foam11.

A lower ROS therefore leads to stronger foam which provides better mobility control, resulting in less steam override and less
loss of steam to the cap rock. Thus the heat efficienty is improved. Thus the ability of the alkali-surfactant system to emulsify
the oil into the aqueous phase results in not only a lower ROS but also the synergistic effects of faster foam propagation,
stronger foam, better mobility control and improved steam utilization.

Summary of Experiment Results

Results of experiments discussed above can be summarized as follows:

1. Emulsion screening tests showed that an alkali-surfactant system consisting of AOS 1618 and Na2CO3 was found to
generate classical under optimum, optimum, and over optimum systems with a heavy Californian crude oil at 212oF. It
was found that the optimum region where very low oil-water interfacial tension exists is sensitive to the oil-to-water ratio
and moves to higher Na2CO3 concentration as the oil-to-water ratio is reduced.
2. A surfactant propagation coreflood showed that Na2CO3 can substantially accelerate surfactant propagation through the
reservoir by precipitating divalent ions that are exchanged off the clays and by reducing adsorption by rendering the clay
surface more negatively charged.
3. Tertiary corefloods showed that an alkaline steam foam using AOS 1618 and Na2CO3 reduces residual oil saturation to
steam over a wide range of Na2CO3 and AOS concentrations.
4. These coreflood results were also supported by flow visualization experiments which showed that alkaline steam foam
propagation is characterized by an oil bank, an emulsion bank, and a steam foam region. Compared to regular steam
foam, alkaline steam foam shows an increase in degree of emulsification, much faster foam propagation rate, a higher
pressure gradient, and significant reduction in residual oil saturation.
5. Flow visualization studies also showed that rapid emulsification of oil into small droplets with size smaller than that of a
pore throat is the main mechanism of oil displacement by alkaline steam foam.

Applications of Alkaline Steam Foam

This study has shown that the steam foam process can be significantly enhanced by injecting an alkali-surfactant mixture in the
aqueous phase of steam. Suitably formulated alkaline steam foam combines the benefit of thermal and chemical EOR
processes. The injected alkali suppresses surfactant adsorption and precipitation. This increases the surfactant and therefore
foam propagation rate. In addition, it provides more mobility control by forming stronger foam. This reduces steam override,
reduces heat loss to the cap rock and improves heat efficiency. In addition, by emulsifying the oil into the aqueous phase,
6 SPE 144968

alkaline steam foam has the potential to reduce the residual oil saturation to steam. All these enhance the effectiveness of a
steam foam drive.

In the case of cyclic steam soak, there is an additional benefit. The emulsification of oil into the aqueous phase decreases the
effective viscosity of the oil and thereby increases the gravity drainage rate during the production cycle. The oil production
rate is thereby increased.

In the case of SAGD for very heavy crude, the formation of strong foam near the injector has the added benefit of reducing the
steam breakthrough to the producer12. In addition, the gravity drainage rate is increased by the emulsification process leading
to a faster oil production rate. Furthermore, the diverting effect of foam can mitigate steam channeling through high
permeability streaks and can result in more uniform steam chest development along the length of the horizontal steam injector.

References

1. Hirasaki, G. J.: “The Steam-Foam Process,” J. Petroleum Technology (May 1989), 449-456.
2. Dilgren, R. E. Deemer, A. R., and Owens, K. B.: “Laboratory Development and Field Testing of
Steam/Noncondensible Gas Foams for Mobility Control in Heavy Oil Recovery,” paper SPE 10774 presented at the
1982 SPE California Regional Meeting, San Francisco, March 24-26.
3. Patzek, T. W. and Koinis, M. T.: “A Californian Steam-Foam Pilots,” JPT (April 1990) 496-503.
4. Patzek, T. W. and Myhill, N. M.: “Simulations of the Bishop Steam-Foam Pilot,” SPE 18786 presented at the 1989
SPE California Regional Meeting, Bakersfield, April 5-7.
5. Lau, H. C. and O’Brien, S. M.: “Surfactant Transport Through Porous Media in Steam-Foam Process, SPERE (Nov.
1988) 1177-85.
6. Lau, H. C. and Borchardt, J. K.: “Improved Steam-Foam Formulations: Concepts and Laboratory Results,” SPERE
(Nov 1991) 470-476.
7. Lau, H. C.: “Alkali-Enhanced Steam Foam Oil Recovery Process,” U.S. Patent No. 4,609,044 (Sept. 2, 1986).
8. Lau, H. C.: “Alkali-Enhanced Steam Foam Oil Recovery Process,” Canadian Patent No. 1,259,476 (Sept. 19, 1989).
9. Lau, H. C.: “Trona-Enhanced Steam Foam Oil Recovery Process,” U.S. Patent No. 4,737,938 (March 1, 1988).
10. Nelson, R. C., Lawson, J. B., Thigpen, D. R., and Stegemeier, G. L..: “Cosurfactant-Enhanced Alkaline Flooding,”
paper SPE 12672 presented at the 1984 SPE/DOE Symposium on Enhanced Oil Recovery, Tulsa, April 15-18.
11. Lau, H. C. and O’Brien, S. M.: “Effects of Spreading and Nonspreading Oils on Foam Propagation Through Porous
Media,” SPERE (Aug. 1988) 893-96.
12. Chen, Q, Gerritsen, M. G., and Kovscek, A. R., “Improving Steam-Assisted Gravity Drainage Using Mobility Control
Foams: Foam Assisted-SAGD (FA-SAGD), paper SPE 129847 presented at the 2010 SPE Improved Oil Recovery
Symposium, Tulsa, April 24-28.
SPE 144968 7

San Joaquin sand properties

Cation exchange capacity, meq/100g 9.87
San Joaquin oil properties
o
Oil gravity, API 13
Acid number, mg KOH/g 2.5
Base number, mg KOH/g 2.5
o
Viscosity at 60 F, cp 43,400
o
Viscosity at 100 F, cp 2,690
o
Viscosity at 180 F, cp 96
Table 1 – Properties of San Joaquin sand and oil

Experiment AOS, wt% Na2CO3, wt% NaCl, wt% So at start of ROS to foam, %
foam, %PV PV
1 0.5 2.65 1 30 0
2 0.5 2.65 1 20 0
3 0.5 10 1 28 5
4 0.5 15 0 23 2
5 0.33 9 0 14 0
6 0.33 9 0 N 2
7 0.33 9 1 N 1
8 0.25 2.65 0 N 5
9 0.25 8 0 N 2
10 0.10 1 0 N 7
11 0.10 6 0 N 5
12 0.05 9 0 N 4
13 0.025 1 0 N 11

N – Not measured. ROS to steam under similar experimental conditions was 13%.

Table 2 – Residual oil saturation to alkaline steam foam in experiments using 1 ft long, 1.5 inch diameter
pressure vessel packed with grade F140 Ottawa sands.

Regular Steam Foam Alkaline Steam Foam

Surfactant formulation 0.5% AOS 1618, 4% NaCl 0.33% AOS 1618, 9%
Na2CO3
Steam components 50% quality steam, 0.006 50% quality steam, 0.006
mole fraction nitrogen mole fraction nitrogen
Frontal rate of surfactant = qL/Vp, ft/day 19 19
Emulsification of oil Slow Fast
Foam texture Fine Finer
Maximum pressure gradient, psi/ft 78 115
PV surfactant injected at foam breakthrough 1.6 0.4
Foam propagation rate measured by pressure 14 50
response, ft/day
Spreading characteristics Oil non-spreading Oil non-spreading
ROS (%) 14 4
Table 3 – Summary of flow visualization experiments with San Joaquin outcrop sands.
8 SPE 144968

Fig. 1 – Activity diagram of Na2CO3/AOS 1618/San Joaquin crude at 212oF; oil-to-water ratio=1; 1 wt% NaCl.

Fig. 2 – Activity diagram of Na2CO3/AOS 1618/San Joaquin crude at 212oF; oil-to-water=1/3; 1 wt% NaCl.
SPE 144968 9

Fig. 3 - Effluent concentration of surfactant propagation experiment at 212oF. The sand pack containing San Joaquin
sand was flooded with 2.25 PV of 2.65 wt% Na2CO3, followed by continuous injection of 0.5% AOS 1618 and 1 wt%
NaCl.

Fig. 4 – Residual oil saturation to alkaline steam foam.

10 SPE 144968

Fig. 5 – Schematic of flow visualization apparatus.

SPE 144968 11

Fig. 6 – Side view of flow visualization apparatus.

Fig. 7 – Front view of flow visualization apparatus.

12 SPE 144968

Fig. 8 – Pressure history of regular steam foam with 0.5 wt% AOS 1618, 4 wt% NaCl.

Fig. 9 – Pressure history of alkaline steam foam with 0.33 wt% AOS 1618, 9 wt% Na2CO3.
SPE 144968 13

Fig. 10 – Comparison of normalized remaining oil saturations to regular and alkaline steam foam.

Fig. 11 – Residual oil saturation by section at the end of regular and alkaline steam foam experiments.
14 SPE 144968

Fig. 12 – Macroscopic view of alkaline steam foam observed in flow visualization experiment.