See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/250093184

Insights Into Non-Thermal Recovery of Heavy Oil

Article in Journal of Canadian Petroleum Technology · March 2009

DOI: 10.2118/09-03-27

CITATIONS READS
103 872

4 authors, including:

J. Bryan
The University of Calgary
54 PUBLICATIONS 1,028 CITATIONS

SEE PROFILE

All content following this page was uploaded by J. Bryan on 18 May 2015.

The user has requested enhancement of the downloaded file.

 PAPER 2006-553

Insights Into Non-Thermal Recovery of Heavy

Oil
A. MAI1,2, J. BRYAN1,2, N. GOODARZI3, A. KANTZAS1,2
1. University of Calgary, Schulich School of Engineering
Department of Chemical and Petroleum Engineering
2. TIPM Laboratory
3. Formerly with University of Calgary and TIPM Laboratory, now with Nexen Inc.

This paper has been selected for presentation and publication in the Proceedings at the 1st World Heavy Oil Conference.
All papers selected will become the property of WHOC. The right to publish is retained by the WHOC’s Publications Committee.
The authors agree to assign the right to publish the above-titled paper to WHOC, who have conveyed non-exclusive right to the
Petroleum Society to publish, if it is selected.

Abstract
Many heavy oil reservoirs in the world contain oil that has Introduction
limited mobility under reservoir conditions. In these reservoirs,
a small fraction of the oil in place can be recovered using the Many countries in the world have heavy oil reservoirs.
internal reservoir energy through heavy oil solution gas drive Canada and Venezuela in particular contain some of the largest
(primary production). An integral part of this process is the so- heavy oil and bitumen resources in the world. Rising energy
called “foamy oil mechanism”, where oil is produced as a gas- demands, coupled with a decline in conventional oil reserves,
in-oil dispersion. At the end of primary production, the bulk of has led to increased interest in heavy oil recovery in recent
the oil is still in place, while the natural energy of the reservoir years. The size of these heavy oil deposits is immense, and
has been depleted. In this work, we present experimental these are likely one of the main future energy sources in the
results of foamy oil depletion at two different length scales, and years to come. Unfortunately, the oil in these reservoirs is
varying depletion rates. At the conclusion of primary recovery, highly viscous, and cannot easily flow to production wells
the potential for further non-thermal exploitation of these under normal reservoir conditions. Understanding the
reservoirs is explored. Results for waterflooding and chemical mechanisms by which heavy oil can be displaced in reservoirs
flooding are presented, demonstrating the viability of these is crucial to the successful recovery of this resource base.
techniques for heavy oil EOR. There are several displacement Heavy oil can be defined as a class of oils with viscosity
mechanisms identified through the secondary and tertiary ranging from 50 mPa⋅s (cP) up to around 50,000 mPa⋅s. This
processes that contribute to significant (although slow) oil has limited mobility under reservoir temperature and
incremental recovery of heavy oil. pressure, and Darcy’s Law predicts that the oil can flow slowly
under high-applied pressure gradients. However, it has been
observed that in these reservoirs, solution gas drive leads to
significantly higher rates and recoveries than what was expected
by conventional understanding of gas-oil relative permeability

1
behavior1. This behavior, first reported in Canadian heavy oil higher viscosity fluids4. It has also been observed10,11 that the
pools, has since been observed in many other reservoirs around applied pressure decline rate can also lead to differences in the
the world including South America, China and Albania. final distribution of gas and the gas relative permeability.
Investigations into the causes of this abnormal but fortuitous These observations have led to the development of pressure
primary production response have been the focus of many gradient-dependent or capillary number-dependent gas relative
publications in the past 25 years. permeability curves12-15. The inherent relationship explored in
The recovery from primary production in heavy oil all of this work is that foamy oil flow only appears to occur
reservoirs may be as high as 20%2, but is often lower. At the under the influence of high pressure gradients across the porous
conclusion of primary production, therefore, there is still a medium.
significant amount of oil in place in the reservoir, and the In the field, high pressure gradients can be developed due to
reservoirs have been stripped of their natural energy. In order the combined influence of the pressure decline rate and the
to recover additional oil, the reservoir energy has to be production of sand. Laboratory scale investigations of solution
replenished, and then oil has to be displaced to production gas drive in heavy oils requires the application of high depletion
wells. This is the field of improved/enhanced oil recovery rates in order to match the observations made in the field. At
(EOR). longer length scales, it becomes more difficult for pressure
EOR schemes often involve reduction of the oil viscosity gradients to develop across the entire physical model16-18,
through the addition of thermal energy or hydrocarbon solvents. meaning that gas evolution and foamy oil flow is in fact a
Many heavy oil reservoirs are relatively small or thin, however, localized phenomenon17,18.
making these EOR techniques uneconomic. In these reservoirs, In this study, experiments are performed at two different
inexpensive non-thermal methods are required to recover length scale systems: 1.2 m and 0.4 m. At this reduced length
additional oil. In this paper, we investigate the potential for scale, it is possible to develop the required pressure gradients
waterflooding and alkali-surfactant flooding, and demonstrate across the entire sand packs, thus foamy oil flow occurs
how both of these techniques, if properly designed and throughout the cores and appropriate rates and recoveries are
controlled, can lead to the recovery of high viscosity heavy oil. determined at different pressure depletion rates. The cores can
then be used for further improved oil recovery testing.

Primary Production in Laboratory Cores Primary Production: Experimental Results

Reservoirs that exhibit foamy oil behavior are characterized Three primary depletion experiments were carried out in this
by the appearance of a ‘chocolate mouse’ like, oil-continuous study, and the properties of the oil and the sand packs are
foam at the wellhead. Experience has shown that when oil is shown in Table 1. As can be seen, the short core has
produced as this non-equilibrium mixture, reservoirs can approximately 6 % of the pore volume (PV) of the two longer
perform with higher than expected rates of production: up to 30 cores, labeled “A” and “B”. The permeability of the short
times that predicted by Darcy’s Law, and lower than expected system is significantly higher than either of the two longer
production gas-oil-ratios3. In the field, it has been observed that cores, and despite the fact that the same heavy oil was used for
foamy oil flow is often accompanied by sand production along all three core floods, the live oil measured density and viscosity
with the oil and gas. Sand production leads to sand dilation and are all different for the different systems. This illustrates the
the presence of high porosity, high permeability zones difficulty in performing repeatable experiments using actual
(wormholes) in the reservoir4,5. It is generally believed that in crude heavy oil, as opposed to viscous mineral oil. The oil used
the field, the high rates and recoveries observed are the in these experiments was taken from a heavy oil field in
combination of the foamy oil mechanism and the presence of Saskatchewan, which is currently still undergoing primary
these wormholes. production.
The entire cold heavy oil production with sands (CHOPS) is In the primary depletion experiments, the core outlet
difficult to reproduce at the laboratory, due to difficulties in pressure was declined at a steady rate, which varied for the
performing constant rate or pressure depletion experiments different systems. The pressure differential across the core was
while producing sand co-currently with the oil and gas. monitored, along with the producing fluid rates and GOR
Therefore, most laboratory experiments focus on understanding values. Figure 1 shows the pressure drop and produced oil flow
the foamy oil mechanism, and how this flow of dispersed gas in rates across Long Core A. At early times, the pressure in the
oil is responsible for the high rates observed in the field. The system was still above the bubble point, therefore as pressure
value of laboratory data is that it can be collected at a much dropped linearly at the outlet of the core, the live oil expanded
higher frequency then field data, and under more controlled and a low flow rate of oil was induced. Once pressure dropped
settings. Pressure data can also be collected and monitored below the supersaturation point of the core, the pressure drop
along with production data, in order to better understand the across the core increased sharply, and the oil flow rate
foamy oil mechanism. correspondingly increased as well. All three core systems
Considerable research has been performed in the past 25 demonstrated this same behavior, which has also been observed
years to understand the mechanics of foamy oil flow. Early by other researchers10-15.
work focused on development of models to explain the flow of Figure 2 shows an example of the pressure and GOR
the gas-in-oil dispersion6-8. While some of this work was later response in the short core system. Similar to the production rate
disproved9, these early studies laid the groundwork for the response, when the pressure differential in the system increases,
development of new ideas in this area. Studies have since the solution gas-oil-ratio (GOR) increases as well. At this
focused on modeling foamy oil behavior, both mechanistically point, gas has evolved from the oil, and the gas bubbles have
and through modification of conventional simulation. Both grown enough that they are providing pressure support to the
approaches have led to an improved understanding of this system. During the period of the high pressure gradient, gas
phenomenon. and oil are both produced at high rates, and foamy oil is visually
One interesting observation that has been made is that when observed in the production vessels. Once the high pressure
comparing depletion experiments for different viscosity fluids gradient is depleted, the flow rate of oil declines rapidly. In the
in the same system, higher pressure gradients develop in the small core experiment, the gas was also depleted at this time,

2
but in longer core systems there would still be gas remaining in As water was injected into the cores, it was observed that the
the system even after the oil rate begins to decline. water immediately displaced the gas, water moved initially
Table 2 shows the maximum pressure gradient established in through the easiest flow channels in the core, and gas was
each core system, along with the oil recovery factor at the produced first. After water broke through, very little gas was
conclusion of primary recovery. The product of the pressure produced along with water and oil. Waterflooding after primary
gradient and the core absolute permeability divided by the live recovery was performed on the three different sand-packs.
oil viscosity is also shown. This product can be thought of as
the viscous force that will be proportional to flow of a single Waterflooding Results for Short Sand Pack
phase fluid in a core. Table 2 shows that as this “viscous force”
increases, so does the recovery in the system. Since foamy oil During primary production, a significant amount of oil was
is in reality a two-phase system with changing viscosity, this recovered (38 % of OOIP). This means that a large portion of
analysis is too simplistic to be useful for actually predicting the porous medium is occupied by gas. Before any effects of
recovery in primary production of heavy oil, but it does waterflooding can be seen, water first must be injected to
illustrate the importance of the development of the high displace this gas. This was done at a constant flow rate of 0.1
pressure gradient region. The oil recovery obtained in these mL/hr, which corresponds to an interstitial velocity of 2.11 E-3
short cores is significantly larger than what can be achieved in m/d. After water had displaced gas, water saturation amounted
longer core systems17,18, however these cores are still valuable to 44.53 %. Thus, in the first step of waterflooding, the water
for post-cold production tests. In these systems, there are large existing in the porous medium had formed a continuous path
areas of low oil/high gas saturation, in contact with areas of from the injector to the producer. Water injection was then
higher oil saturation, and the core pressures were depleted continued at the same flow rate, which is considerably smaller
during primary production. This is the state at which the than the range seen in normal reservoir practice30.
waterflooding experiments were begun. The products were collected in small vials, containing
between 15 to 20 g of oil and water. Heavy oil is sticky and
thick, making visual determination of the produced oil difficult
and inaccurate. Low field NMR was therefore used to monitor
Waterflooding After Primary Recovery the produced fluids. This technology uses the fact that oil and
Waterflooding of oil reservoirs is a well-recognized water spectra are distinct, as can be seen in the Figure 3. The
technique for oil recovery after primary production. In oil mass is proportional to the total amplitude in the first peak,
conventional oil, waterflooding theory is well documented19. whereas the water mass is proportional to the total amplitude in
The inherent assumption in conventional oil waterflooding the last peak. In this manner, even if fluids are mixed or oil
theory is a similarity in viscosity between oil and water20,21. In smeared the walls of the collection vial, the oil and water
heavy oil applications this is not the case. However, content can still be determined. Based on previous experiments
practitioners often still attempt to apply the same theoretical at room temperature, water cut measurements can be made to an
understanding to their fields. accuracy of ±1%.
There has been some limited documented experience for After a total of 1.5 pore volumes (PV) of water injected, the
waterfloods in heavy oil reservoirs22-25, but in general the additional recovery is 8.47 % of OOIP, as shown in Figure 4.
mechanism of viscous oil recovery by waterflooding has not During the course of the experiment it was observed that the oil
been explored. Waterflood recoveries are known to be low for was not produced continuously, which contributes to the
high viscosity heavy oil, but in many heavy oil fields scattering seen in the water cut trend. Water cut ranges from 92
waterflooding is still commonly applied, since it is relatively – 97 %, with a slow increase in water cut with time. Since
inexpensive and field operators have years of experience water is continuous throughout the porous medium, any water
designing and controlling waterfloods. injected would be easily produced at the other end, thus the
At the end of a conventional oil waterflood, residual oil is produced fluids will always have a high water cut.
left in place due to reservoir heterogeneities or capillary At this point, it is useful to consider what forces could be
trapping. In laboratory core flood experiments, capillary present in the system, leading to recovery of this viscous oil.
bypassing is the main mechanism responsible for trapping of There are several forces acting upon the oil and water in the
oil20,26. This is not the case in heavy oil reservoirs. In these rock pores that could potentially be responsible for the oil
reservoirs, the high oil viscosity (and hence the poor mobility recovery after breakthrough: gravity forces, capillary forces
ratio between displacing and displaced fluids) is the main cause (water imbibition) and inertial forces (oil stripping). It is
for oil bypassing and residual oil at the end of the waterflood. unlikely that oil is produced due to gravity due to the fact that
Previous investigations have focused on the oil/water mobility the diameter of this sand-pack is fairly small (0.0381 m), and
ratio, and how it relates to viscous fingering or instability of the the density contrast between oil and water is not significant.
water front27,28. In these studies, the investigators focused on Stripping tends to occur when water is injected at high flow
dead oil waterfloods, where a continuous water path does not rates; in this case, where water is injected at a very slow rate, it
exist between the producer and injector. Thus, the focus is to is unlikely that this mechanism could be responsible for the oil
control the water front before breakthrough occurs. recovery. The only remaining force is that of water imbibition
At the end of primary production, a significant amount of oil (capillary forces). The imbibition process in heavy oil is not
still remains in the reservoir along with gas that was formed well understood, however if water is injected over a long period
during primary depletion. In these cores, the high internal gas of time, this must have some contribution to the oil recovery
saturation has essentially formed a tortuous pathway of low process. Scott et al.31 suggested that waterflooding in heavy oil
resistance between the production and injection end of the sand can occur in different zones: a capillary dominated region or a
pack. The response of these cores to waterflooding will be viscous region, with a transition zone in between. While there
considerably different than that of dead oil waterfloods. There has been no data to support this theory, this provides a plausible
is only limited information published regarding the status of explanation of why oil can be recovered at low injection rates.
waterflooding after primary production24,29, and the Capillary forces, largely neglected in heavy oil reservoirs due to
mechanisms of oil recovery at this stage are poorly understood. the high oil viscosity, can in fact be the major contributing
factor to oil recovery after water breakthrough.

3
 The inlet pressure profile is shown in Figure 5. The time In the case where brine displaces a heavy oil via viscous
period shown includes the period in which gas was produced at forces, this displacement is very unstable. Water would be
the beginning of the waterflood. At the start of the experiment, expected to bypass oil and follow the continuous water channel.
the system still contains some pressure from gas remaining As the flow rate of water was reduced, the water cut is also
trapped in the sand pack. As water is injected, the inlet pressure reduced substantially. As seen previously in Figure 6, the slope
builds up until water breaks through at the production end. The of recovery also increases, implying a change in the recovery
inlet pressure then slowly decreases. As the oil saturation in the mechanisms. The reasons for the original reduction in the water
porous medium decreases and less oil is produced, water cut cut is that with the reduction of viscous force, capillary
increases correspondingly. imbibition pulls water away from its continuous flowing path
and into the oil saturated regions of the core. As water imbibes
Waterflooding Results for Long Sand Packs into the core oil is displaced. As water is injected at a slower
rate, the extent of viscous fingering is reduced, thus allowing
The same sand (Lane Mountain 70) was used to pack both of for the production of some of the bypassed oil. This effect
the long sand packs. It was desired to have two sand packs with diminishes as water continues to find paths to channel through
very similar characteristics in order to allow for a direct the porous medium.
comparison between them, however it is impossible to create
two identical sand-packs. The properties of the sand packs were Waterflooding Results from Sand Pack B
shown previously in Table 1, where it can be seen that the rock
properties are fairly similar. The properties of the different At the start of water injection, this system had a primary
batches of oil (dead and live) are also fairly consistent. The recovery of 26.98 % of OOIP. This experiment is also still
produced fluids from these sand-packs were collected and fluid ongoing, however for an injection period of about 2000 hrs,
content was determined using two methods: separation and which corresponds to a PV injected of 1.83, the waterflood
NMR. Separation was used when there was a significant recovery is 12.00 % of OOIP, as seen in Figure 8.
amount of water whereas NMR was done when the samples In this experiment, brine was injected to displace gas for
have a considerable fraction of oil. A comparison of both of approximately 200 hrs. At this point the water saturation in the
these methods shows that they provide consistent results of sand pack reached 33.23 %. Water was then injected at 2.75
produced water and oil volumes. mL/hr (1.06 E-2 m/d). The pressure and produced water cut
profile is shown in Figure 9. After approximately 950 hrs of
Waterflooding Results from Sand Pack A production (region 1), it was noticed that the water cut of the
product was consistently high, so the sand pack was shut in to
At the end of primary production, 20.4 % of the oil had been soak for a period of 350 hrs (region 2).
recovered. Once again, water was first injected in order to After soaking, water was injected at the same flow rate
displace gas and replace the voidage from primary production. again. Immediately the water cut decreased slightly and then
This experiment is still under production, however, after a total increased again to follow the same trend as before. This
injection period of 1627 hrs, which corresponds to a PV reduction in water cut gives a direct indication that imbibition
injected of 0.47, the recovery of oil from waterflooding is had occurred in the porous medium. During the soaking period,
8.29% of OOIP, as can be seen from Figure 6. oil and water have re-distributed in the porous medium due to
For this sand pack, water was injected at a rate of 2.75 mL/hr capillary forces. Water would move into the smaller pores,
(1.06 E-2 m/d) until a significant amount of gas was produced displacing oil into larger pores, and this oil could be produced
and water had broken through. At this stage the water once injection began again. However, at this same injection
saturation in the porous medium was 27.50%. Water was rate, the water cut quickly returns to its previous high values.
injected at the same rate for another 160 hrs, after which the Thus, the soaking period only affect the amount of oil recovery
flow rate of water was reduced to 0.55 mL/hr (2.13 E-3 m/d). immediately after re-injection. The ratio of the rate of
The change in the slope of recovery coincides with the change imbibition to viscous flow of water was unchanged after
in injection rate. soaking.
The inlet pressure profile and the water cut of the products It is important to note that in Figure 9, the slope of pressure
are shown in Figure 7. The two vertical lines on the figure decline is different in both regions; region 1 has a steeper slope
show the end of the voidage replacement region and the point than region 2. In region 1 where a steady increase in the WC
where the injection rate was decreased. From the pressure could be seen, as time progresses (to about 500 hrs) it is
profile it can be seen that when the injection rate is reduced, possible that the effect of viscous flow is diminishing. When
pressure sharply decreases and then continued to slowly the sand pack is shut in, water at the inlet still carry a certain
decrease further, as was also seen in the case of the short sand pressure; some water could thus still travel further into the sand
pack. Prior to the reduction in flow rate water cuts are fairly pack, in order to equalize the pressure. Since the outlet is
high. As the flow rate was reduced the water cut of the closed, water re-distributes itself within the porous medium.
products was also reduced but then gradually increased again. The pressure decline is due to capillary action.
After approximately 1500 hours (0.40 PV injected) the water
cut has increased back to its original value at the higher rate. Comparison of Long Sand Packs
Even though in this case water had again formed a
continuous path, the oil saturation prior to the second The pressure profiles and the water cut of the two systems
waterflooding step is fairly high (72.50 %). It is possible that at are compared in Figure 10. For case A, when the injection rate
the start of the second step of the waterflood, while injection is reduced, its pressure decreased immediately. The drop in
pressures are still high, some oil is still being produced under injection pressure is due to the decrease in flow rate. At this
viscous force. It was observed that the products being produced point, even though the water cut is reduced, an oil bank was not
at the outlet initially contains a high oil fraction. Again, due to forming, since pressure continues to decrease. Instead, the
the characteristics of heavy oil it could not be concluded visibly lower water cut is a reflection of more water moving into the
what the oil fraction was at very early times. It was not possible system through capillary forces, as opposed to flowing just
to collect small samples to evaluate this behavior. through the water channel via viscous forces.

4
 From the figure it can be seen that at early time when both Alkali-Surfactant Flooding of Heavy Oil
systems were having the same injection rate of 2.75 mL/hr, they
both had fairly similar water cut values, with system A having Reisberg and Doscher34 first proposed combining a small
slightly smaller values, as can be seen more clearly in Figure fraction of pre-formed surfactant with alkali in injected water,
11. It is important to note that prior to the waterfloods these and demonstrated how these chemicals together could lead to
systems had different oil saturation. There are two mechanisms additional recovery of oil. Since then, alkali-surfactant (AS)
interacting in the porous medium. Both of these sand packs flooding has become a well-recognized enhanced oil recovery
have fairly similar characteristics, however, prior to waterflood, technique in conventional oil reservoirs. The injected alkali
case B has a higher initial water saturation, which could reacts with naturally occurring acids in the oil, leading to the
potentially lead to faster bypassing, and a higher water cut. The generation of in-situ surface-active agents (soaps) at the oil-
effect of imbibition is harder to extrapolate. water interface35. These soaps lead to significant reductions in
In the literature, the behavior of water imbibition in heavy oil the oil-water interfacial tension36,37, which can greatly reduce
reservoir has not been reported. However, there are many the capillary forces trapping an oil ganglion in a rock pore 38,39.
studies regarding imbibition in conventional oil and gas It has been shown both experimentally and theoretically20 that
reservoirs. These literature sources reported inconsistent in conventional oil reservoirs, the majority of trapping occurs
observations regarding the effects of initial saturation on due to these capillary forces. Thus, a reduction in interfacial
recovery and production rate32. For cores that contain a higher tension can lead to enhanced recovery of oil, beyond the amount
initial water saturation, it is expected that imbibition rates will obtained from waterflooding.
be higher, due to increased contact between oil and water in the A secondary result of reducing the oil-water interfacial
rock pores32. On the other hand, a higher water saturation prior tension is the potential for emulsifying the fluids in one
to waterflooding may lead to more by-passing in viscous flow another40. When two immiscible liquids are mixed and shear is
through the rock pores. Core B, which contained a higher applied, if enough energy is added to the system the liquids will
initial water saturation, has marginally higher water cuts at early emulsify into small droplets that will quickly separate again
times when the injection rates are the same. This appears to once the agitation to the system stops. In the presence of
indicate the dominance of viscous flow effects at higher rates. emulsifying agents, however, the fluids may not separate as
At lower injection rates into core A, for the same pore volumes easily36. Additionally, with low interfacial tension systems
of water injected into either core, core A consistently has lower considerably less shear is required in order to emulsify the
water cuts and higher oil recovery. By reducing flow rates and fluids39. Therefore, emulsification of water into oil (W/O)35,41
allowing imbibition rates to be more significant in relation to and oil into water (O/W)42,43 have also been considered as
the viscous flow rates, improved waterflood recoveries can mechanisms by which enhanced oil recovery can occur.
therefore be obtained. There have been limited investigations into alkali and AS
From the pressure and production data, at the same pore flooding in heavy oil reservoirs23,31,41,44-47, with varying degrees
volume of water injection into both cores, the ratios of total of success. Most of these works have focused specifically on
production rates and pressure are compared in Table 3. In the oil-water IFT reduction as the mechanism for improved oil
case of oil and water flowing in the same channel, the total recovery23,31,41,45,46. The importance of emulsification has also
pressure drop is the sum of pressure drops due to viscous flow been mentioned by several researchers as well41,43,47. When
of water and oil, plus a term for the capillary effect: investigating chemical flooding, it is very important to
understand the mechanisms responsible for the success or
∆Ptot = ∆Pw + ∆Po – Pc………………………………(1) failure of the flood. Chemical flooding can be a complex
process, involving reactions between the oil, injected aqueous
The production rates in Table 3 appear to be slightly higher solution, and the porous medium. Without properly
than the injection rates, which is due to pump difficulties in understanding how the AS flood is leading to enhanced oil
controlling very low flow rates. As time progresses, however, recovery, it is difficult to design successful floods in heavy oil
the ratio of the produced volumes is approaching a factor of fields.
five, which is the ratio of the injected flow rates between the As mentioned previously, in heavy oil reservoirs the high oil
two systems. As the ratios of the flow rates decrease, it is viscosity (and hence the poor mobility ratio between displacing
expected that the ratio of the pressure drops should decrease as and displaced fluids) is the main cause for oil bypassing and
well, assuming that production is due only to viscous forces. residual oil in the reservoir. Enhanced oil recovery methods
Table 3 shows that in fact, the opposite trend is occurring. This therefore focus on reduction of the injected fluid mobility23 or a
can be explained by examining Equation 1. From imbibition in reduction in the oil viscosity. In smaller, thin heavy oil
gas and conventional oil, it has been reported that the volume of reservoirs, inexpensive non-thermal methods are required. AS
water imbibed is proportional to time33. Since this will be true flooding has been identified as a potential viable non-thermal
for heavy oil systems as well, then in case A the volume EOR technique47,48. Since heavy oil will not be trapped by
imbibed must be more significant than in case B. At faster flow capillary forces, the effect of the oil-water interfacial tension
rates, the viscous pressure drops required for oil and water to reduction must be somehow contributing to an improvement in
flow are high enough that the pressure reduction from water the mobility ratio. Specifically, the proposed mechanism is that
imbibition is less significant. At slower flow rates, however, the AS solution contacts the oil, leading to the formation of
this term has pressures in the same order of magnitude as the surfactant molecules at the oil-water interface. This, coupled
viscous pressure drops, and the pressure required for flow at a with shear applied from the fluids flowing through small rock
given rate will be less. pores, AS emulsifies the heavy oil into the water. O/W
The cumulative recovery of both cases is compared in Figure emulsions have lower viscosities than their constituent oil, thus
12. From the comparison above, it could be seen that reducing the oil mobility has been improved47,48.
the injection rate resulted in an improvement on the recovery. It Even after the optimization of a heavy oil waterflood, there
is not yet known what the final recovery from these systems may still be a significant fraction of oil remaining that can
will be. possibly be recovered by the addition of a small amount of
alkali and pre-formed surfactant to the injected water. The key
to the success of a heavy oil AS flood is that enough shear must

5
be applied in order to emulsify the oil into droplets that are to NaOH52. All the surfactants formed will therefore tend to be
small enough to flow through the pore throats. Unlike in anionic, which will reduce surfactant adsorption on negatively
previous applications42 where the emulsion is generated at the charged rocks53.
surface and injected into the reservoir, the shear forces There are many different combinations of alkalis and
generated due to flow at representative field rates must be surfactants, that together can be used to emulsify oil into any
sufficient to emulsify the oil in-situ and transport it in the brine48. For the purpose of this study, the goal was to develop a
flowing water phase. The focus of this investigation is to simple combination of alkali and surfactant that would lead to
examine the behavior of AS floods, both as a secondary and the generation of O/W emulsions. As such, the approach
tertiary recovery technique in unconsolidated sand, heavy oil outlined by Atkinson52 was followed. Namely, the surfactant
reservoirs. concentration was held fixed at 0.1 wt% (1000 ppm), while the
alkali content was varied from 0 wt % to 1.5 wt % (0 – 15,000
Experimental Methods and Procedures ppm) of the aqueous phase. Alkali is considerably cheaper than
pre-formed surfactant, thus it is better to hold the surfactant at a
The original intent for this work was to follow up the low low concentration and vary the amount of alkali added. Phase
rate waterfloods in the aforementioned models with AS floods behavior tests were performed, as outlined in the literature50,51,
of constant chemical composition. At this time, however, the and IFT tests were later performed in order to validate the
waterfloods are still ongoing, so experiments were instead results from the phase behavior tests.
performed on a small core that was packed using glass-beads. It was noted that with surfactant but in the absence of alkali,
The details of the porous medium are shown in Table 4. oil and water would mix when agitated, but the resultant
The oil is the same as that used in the primary production emulsion separated very quickly. In the presence of alkali and
and waterflooding experiments. The glass beads used as the surfactant, however, the resultant mixture appeared black, but
porous medium are also the same as the short core, however the flowed easily when tipped. This was therefore determined to be
permeability of this core is significantly higher due to the lack a low viscosity O/W emulsion. The black colour of the
of overburden pressure on the core. The dimensions of the core emulsion indicates that it is a macroemulsion50, which is not
are small enough that multiple experiments could be run to thermodynamically stable and will therefore separate again over
determine the mechanism by which oil is recovered by AS time. This is desirable for EOR applications since after
flooding. It is important to note that in order for O/W producing the heavy oil, it will have to be separated from water
emulsions to form in-situ, the emulsions must be able to form before it can be further refined.
under conditions of mild shear47. In this work, the experiments In the presence of 2 wt% NaCl (20,000 ppm) in the aqueous
were performed on a core with a very high permeability, which phase, there were no conditions under which an O/W emulsion
exposes them to lower shear than they would in a lower could be formed. After agitation, the aqueous phase remained
permeability rock. If emulsification can be validated in such a clear, indicating that it contained no emulsified oil. The oil
high permeability porous medium, then in a porous medium phase, on the other hand, appeared to have swelled, and when
containing even smaller pores, emulsification will occur more the bulk liquid specimens were tipped, the oil did not flow as
easily. easily as the control sample, which contained no alkali or
Before AS floods can be performed on the four models after surfactant. It was concluded, therefore, that in the presence of 2
primary production, the proper chemical composition must be wt% NaCl in the water, a more viscous W/O emulsion had
determined, and techniques for identifying when emulsions formed instead. This strong dependence of the emulsion type
were forming had to be developed and tested. Experiments on salinity has also been observed by other researchers48-51, and
were therefore first performed on bulk liquid systems, and then demonstrates the importance of performing these phase
multiple experiments were performed on the sand pack from behavior tests in order to choose the proper AS composition to
Table 4. In between experiments, the core was cleaned with use in heavy oil AS floods.
toluene and methanol. Its pore volume and permeability was Oil-water IFT was also measured using a spinning drop
re-measured each time, in order to ensure that the sand tensiometer, with and without the presence of salt. The results
properties did not change appreciably between experiments. of these measurements are shown in Figure 13. An examination
of the IFT behavior of this oil/aqueous system shows that in the
Bulk Liquid Testing presence of salt, the oil-water IFT is initially lower than the
When oil and water mix with one another under shear, both system containing no salt. At higher alkali concentrations,
O/W and W/O emulsions are initially formed. The final however, the interfacial tension is consistently slightly higher
emulsion state that characterizes the system is the emulsion that than in the salt-free system. It is important to note, however,
separates more slowly. Due to the viscous nature of heavy oil, that even in the presence of salt, the oil-water IFT is greatly
W/O emulsions are much easier to form that O/W reduced by the addition of alkali and surfactant. In the design
emulsions36,39. Determining how to generate O/W emulsions of chemical floods, authors commonly look only at the IFT
containing viscous heavy oil is non-trivial, and has been the values31,45,46 to determine the optimum flood composition. It is
topic of limited research in the past48,49. The emulsion type is true that low oil-water IFT is a requirement for successful AS
known to be a strong function of the surfactant type, the alkali flooding, however the IFT tests alone do not show which
content and the salt content in the system50,51. The first goal of emulsion type will be preferred. Therefore, a combination of
this study was therefore to determine how O/W emulsions could both IFT measurements and bulk liquid phase behavior studies
be generated in bulk liquid systems. should be performed in order to properly choose the chemical
The surfactant used in this study is a commercial anionic composition for a heavy oil AS flood.
surfactant supplied by Stepan Oil Company. This surfactant In the experimental work performed in this paper, the chosen
(Bioterge PAS-8S) is a sodium alkane sulfonate. Tests in the AS composition consisted of 0.1 wt% (1000 ppm) surfactant
past52 indicated that in the presence of alkali, ultra-low and 0.5 wt% (5000 ppm) alkali. From the IFT tests, this alkali
interfacial tension was possible with, even for heavier oils. The concentration was in the middle range of the low IFT plateau,
alkali used in this study is sodium carbonate (Na2CO3). It was and the phase behavior tests showed that with this composition
chosen because of its buffering capabilities and because it tends of chemicals, the oil could be emulsified into water and the
to lead to greater reductions in heavy oil-water IFT, compared emulsions remained stable for longer than higher alkali

6
contents. At 5000 ppm alkali, the emulsion takes approx. 48 emulsified oil. Due to a portion of the oil flowing as an
hours to separate, as compared to higher alkali content, where emulsion, the pressure requirement is lower in the AS flood.
the emulsion separates in under one day. All tests have been Figure 16 shows the complete pressure profile for the
performed to date only with de-ionized water, where phase waterflood compared to the AS flood, over the entire course of
behavior studies have shown that O/W emulsions will form. the experiment. The unique nature of the AS flood is apparent
from an examination of the pressures. In a waterflood, after the
Core Flood Experiments initial high pressure drop, water breaks through and the
injection pressure remains low in the system. At high flow
In all the core floods, the core was initially placed under
rates, the majority of water is then traveling through the
vacuum and then saturated with de-ionized water. Oil was then
continuous water channels; production water cuts will be high
injected at a fixed rate, and the fluids were collected in order to
and very little additional oil is recovered. In the AS flood,
determine the original oil in place (OOIP) in the core. At the
however, two other pressure peaks are observed to occur, at
end of the oil flood, water or AS fluid would be injected at a
around 2 PV and 5.5 PV of fluid injected. The resulting oil
fixed rate corresponding to a Darcy velocity of 0.51m/day,
recovery profiles are compared in Figure 17. The error bars
using a digital Isco pump. Pressures were monitored, and
were found from performing repeat experiments. The large
produced fluids were collected in vials that were measured
error in the recovery results is due to the low pore volume of the
using low field NMR. Figure 14 shows the pressure and oil
system; in such a small core, small inaccuracies in produced
recovery profile for a waterflood carried out in the small core
water cut are magnified.
from Table 4. In these experiments, water and AS injection
For the AS flood, the oil recovery is initially higher than that
were all performed at the same rate in order to compare the
of the waterflood, despite the lower injection pressure
flood responses. Figure 14 shows that initially the pressure of
requirement. The time up to the loss of the initial high injection
the system increased sharply, since water was being injected at
pressure is termed the “viscous displacement region”, and is
a fixed rate into viscous heavy oil that could not easily flow at
shown in Figure 17 and Table 5 as Region A. After the viscous
this same velocity. As the injection pressure in the system
region, the waterflood recovery rate is a constant of around 0.01
increases, oil is displaced ahead of the water front, and a
RF/PV inj for the remainder of the waterflood. In the AS flood,
significant fraction of the oil is produced. At a certain time, the
however, the two other high pressure regions lead to a time
injection pressure becomes a maximum and then drops quickly
where the oil recovery rate is still higher than that of the
down to low values. After injecting almost ten more pore
waterflood, shown as Region B in Figure 17 and Table 5.
volumes of water, only an additional 10% of oil is recovered.
When no more pressure increases in the system are observed
At high water injection rates, therefore, the majority of oil is
(Region C), the AS flood and waterflood recover oil at the same
recovered at these early times, before water breakthrough.
rate. At this point, there appears to be no more benefit to doing
It is worthwhile noting at this point that in these small core
an AS flood.
experiments, no attempt has been made to optimize the
The data presented in Figures 15 to 17 and in Table 5 are all
waterflood. As a result, after water breakthrough occurs, it
comparing primary recovery from waterflooding to that of AS
required almost ten additional volumes of water injection in
flooding. Although AS floods are shown to be more efficient,
order to recover the same fraction of oil as what could be
in most practical situations they will only be considered as a
recovered by under two pore volumes in the slower rate
tertiary oil recovery option, at the completion of a waterflood.
waterfloods discussed previously. This is yet more evidence
Therefore, at the end of the waterflood experiment shown in
that the low rate waterfloods are performing phenomenally
Figure 14, an AS flood was then continued, at the same
better than the high rate waterflood performed on the small
injection rate of 0.51 m/day. The pressure and oil recovery
core. At these high rates, viscous forces dominate since the
profile for this experiment are shown in Figure 18. The
goal is emulsification of oil. Without any added chemical,
associated produced water cuts are shown in Figure 19. AS
however, the waterflood performs very poorly at these high
injection begins at the vertical line labeled in both of these
injection rates.
figures. After approximately 0.15 PV of AS solution injected,
Figure 15 contrasts the pressure response of an AS flood (0.1
the pressure in the system began to increase once again, which
wt% surfactant and 0.5 wt% alkali), injected at the same rate as
correspondingly led to an increase in the oil recovery and a
the waterflood. It can be seen that the maximum pressure
decrease in the produced water cut. Similar results have also
requirement for the same injection rate is lower for the AS
been observed by other researchers47. From these figures,
flood, and that after the pressure reaches a maximum value, the
therefore, it appears that so long as the injected chemical
system takes longer to return back to very lower injection
composition is properly developed, AS flooding has
pressures.
considerable potential as an effective non-thermal EOR
The lower injection pressure for the same injection flow rate
technique at the end of a waterflood. AS flooding as a post-
leads to two possible explanations. The first is that the AS
waterflood EOR scheme appears to have at least a similar
flood is less efficient, and is channeling through the core and
recovery efficiency to AS flooding performed directly into dead
contacting a smaller area of the rock. If this were the case, then
oil.
the recovery during this viscous portion (i.e. the oil recovered
It should be noted in all of these tests that although the
up the point where the injection pressure returns back to low
mechanism of oil emulsification is used to explain the improved
values) would be less for the AS flood. Table 5 shows the
oil recoveries, no visible O/W emulsions were produced. The
recovery rates (oil RF/PV inj) for different portions of the
presence of O/W emulsions is inferred based on the measured
floods. During the viscous displacement portion, the AS flood
low IFT values and the results of the bulk liquid phase behavior
is recovering oil at a higher rate of 0.309 RF/PV inj, compared
tests. This was confirmed by other researchers47, who
to a value of 0.257 for the waterflood. This indicates that the
performed micromodel studies to image oil emulsification, but
reduced pressure requirement is not a reflection of a less
also did not report the presence of these emulsions in their
efficient flood. Indeed, the viscous portion of the flood
produced fluids.
recovers more oil than the waterflood. The lower pressure
One explanation for these results is that the heavy oil is
requirement can be inferred to be indicative of oil
emulsified in-situ and at first flows entrained with the water as a
emulsification along with viscous displacement of un-
low viscosity fluid, which explains why the pressure did not

7
begin to increase until after 0.15 PV of AS fluid were injected REFERENCES
after the waterflood. Since macroemulsions are not
thermodynamically stable, the oil droplets will later coalesce 1. SMITH, G.E., Fluid Flow and Sand Production in
into “oil banks”, and the subsequent increase in injection Heavy Oil Reservoirs Under Solution Gas Drive; SPE
pressure corresponds to the pressure required to displace these 15094, 56th California Regional Meeting of the Society
small regions of higher oil saturation. The recovery mechanism of Petroleum Engineers, Oakland, California USA, April
is therefore emulsification and coalescence, as opposed to the 2-4 1986.
entrainment mechanism described in the literature40. A second 2. FIROOZABADI, A., Mechanisms of Solution Gas
possible explanation for this response is emulsification and Drive in Heavy Oil Reservoirs; J. Can. Pet. Tech., Vol.
entrapment40,42,43, whereby the emulsified oil droplets plug off 40, No. 3, March 2001.
pores in the water channel, leading to improvements in sweep 3. POON, D. and KISMAN, K., Non-Newtonian Effects
efficiency in the reservoir. At this point, it cannot be verified on the Primary Production of Heavy Oil Reservoirs; J.
which mechanism is responsible for the improved oil recovery Can. Pet. Tech., Vol. 31, No. 7, September 1992.
observed. 4. MAINI, B.B., Foamy Oil Flow in Primary Production of
Heavy Oil Under Solution Gas Drive; SPE 56541, SPE
Annual Technical Conference and Exhibition, Houston,
Conclusion Texas USA, October 3-6 1999.
Primary production experiments show that, as expected, the 5. MAINI, B.B., Foamy Oil Flow; SPE 68885, SPE J. Pet.
recoveries and flow rates are dependent on the existence of a Tech., Distinguished Authors Series, , 54 – 64, October
high pressure gradient across the core. In the three systems 2001.
tested, recoveries from primary production are consistently 6. ISLAM, M.R. and CHAKMA, A., Mechanics of Bubble
large. Flow in Heavy Oil Reservoirs; SPE 20070, 60th Annual
Waterflooding after primary production shows improved oil Regional Meeting, Venture, California USA, April 4 – 6
recovery at slower injection rates. The effects of imbibition are 1990.
observed in these experiments, and it is concluded that 7. CLARIDGE, E.L. and PRATS, M., A Proposed Model
imbibition can be a mechanism leading to additional oil and Mechanism for Anomalous Foamy Oil Behaviour;
recovery during waterflooding. Heavy oil waterfloods can be SPE 29243, International Heavy Oil Symposium,
successfully controlled, so long as injection rates are kept low Calgary, Alberta Canada, June 19 – 21 1995.
enough to allow for capillary forces to aid recovery. 8. SHEN, C. and BATYCKY, J.P., Observation of
AS flooding is a potentially viable technique for recovering Mobility Enhancement of Heavy Oils Flowing Through
oil at the conclusion of waterflooding. The injected chemical Sand Pack Under Solution Gas Drive; J. Can. Pet. Tech.,
solution, consisting of 99.4% water, is able to recover additional Vol. 38, No. 4, 46 – 53, April 1999.
oil beyond that of waterflooding. The recovery of AS flooding 9. POOLADI-DARVISH, M. and FIROOZABADI, A.
after waterflooding appears to be at least as good as the Solution Gas Drive in Heavy Oil Reservoirs; J. Can.
potential for performing these floods directly into virgin oil. Pet. Tech., Vol. 38, No.4, April 1999.
The recovery mechanism is the in-situ emulsification of oil into 10. WALL, C.E. and KHURANA, A.K. Saturation:
the flowing aqueous phase. Permeability Relationships at Low Gas Saturations;
Journal Institute of Petroleum, Vol. 57 p.261-269,
September 1971.
11. KUMAR, R. and POOLADI-DARVISH, M. Solution-
Acknowledgement Gas Drive in Heavy Oil: Field Prediction and Sensitivity
The authors wish to acknowledge the help and support from Studies Using Low Gas Relative Permeability; J. Can.
John Schnitzler, Jianhua Gu, Xiao Dong Ji and Michael Erath Pet. Tech., Vol. 41, No. 3, 26 – 31, March 2002.
from TIPM Laboratory. The oil used in this study was 12. TURTA, A., FISHER, D.E., GOLDMAN, J., GREEN,
generously provided by Nexen Inc., and Stepan Oil Company K. and MAINI, B.B. Experimental Investigation of Gas
provided the surfactant. Financial support was received by Release and Pressure Response in Foamy Oil Depletion
NSERC, COURSE and the Canada Research Chair in Energy Tests; CIM 2002-186, Canadian International
and Imaging and its Industrial affiliates (Shell/Albian, Nexen, Petroleum Conference in Calgary, Alberta Canada,
Devon Canada, Petro Canada, Canadian Natural, ET Energy, 2002.
Suncor, Schlumberger, and Paramount). 13. TALABI, O. and POOLADI-DARVISH, M. Effect of
Rate and Viscosity on Gas Mobility during Solution-Gas
NOMENCLATURE Drive in Heavy Oils; SPE 84032, SPE Annual Technical
Conference Denver, Colorado USA, October 5-8 2003.
AS = Alkaline surfactant 14. OSTOS, A. and MAINI, B.B., An Integrated Experimental
CHOPS = Cold heavy oil production with sands Study of Foamy Oil Flow During Solution Gas Drive; CIM
EOR = Enhanced oil recovery 2003-049, Canadian International Petroleum Conference,
GOR = Gas oil ratio Calgary, Alberta Canada, June 10 – 12 2003.
IFT = Interfacial tension 15. OSTOS, A. and MAINI, B.B., Capillary Number in Heavy
NMR = Nuclear Magnetic Resonance Oil Solution Gas Drive and its Relationship with Gas-Oil
OOIP = Original Oil In Place Relative Permeability Curves; SPE 89430, 14th Symposium in
O/W = Oil in water emulsion Improved Oil Recovery, Tulsa, Oklahoma USA, April 17 – 21
Pc = Capillary pressure 2004.
PV = Pore volume 16. SHENG, J.J., MAINI, B.B., HAYES, R.E. and
RF = Recovery factor TORTIKE, W.S. Critical Review of Foamy Oil Flow;
W/O = Water in oil emulsion Transport in Porous Media, Vol.35, 157-187, 1999.
∆P = Pressure drop 17. GOODARZI, N.N. and KANTZAS, A., Observations of
Heavy Oil Primary Production Mechanisms from Long

8
 Core Depletion Experiments; CIM 2006-087, 57th 34. REISBERG, J. and DOSCHER, T.M., Interfacial
Annual Technical Meeting of the Petroleum Society of Phenomena in Crude Oil-Water Systems; Prod.
CIM, Calgary, Alberta Canada, June 13 – 15 2006. Monthly, November 1956, 43 – 50.
18. GOODARZI, N.N. and KANTZAS, A., The Effect of 35. COOKE, C.E. Jr., WILLIAMS, R.E. and KOLODZIE,
Scale on the Primary Depletion of Heavy Oil Solution P.A., Oil Recovery by Alkaline Waterflooding; J. Pet.
Gas Drive; CIM 2006-086, 57th Annual Technical Tech., Vol. 16, December 1974.
Meeting of the Petroleum Society of CIM, Calgary, 36. GARRETT, H.E., Surface Active Chemicals; Pergamon
Alberta Canada, June 13 – 15 2006. Press, 1972.
19. GREEN, D.W. and WILHITE, G.P., Enhanced Oil 37. ROSEN, M.J., Surfactants and Interfacial Phenomena,
Recovery; SPE Textbook Series Vol. 6, Society of 2nd Edition; John Wiley & Sons Inc., 1989.
Petroleum Engineers Inc., 1998. 38. MUNGAN, N., Role of Wettability and Interfacial
20. MOORE, T.F. and SLOBOD, R.L., The Effect of Tension in Water Flooding; SPE 705, SPE Journal, June
Viscosity and Capillarity on the Displacement of Oil by 1964.
Water; Prod. Monthly, 20 – 30, August 1956. 39. ISAACS, E.E. and CHOW, R.S., Practical Aspects of
21. DONG, M. and DULLIEN, F.A.L., Effect of Capillary Emulsion Stability; Emulsions Fundamentals and
Forces on Immiscible Displacement in Porous Media; Applications in the Petroleum Industry, Schramm, L.L.,
SPE 56676, 1999 SPE Annual Technical and Exhibition, ed., Advances in Chemistry Series 231, American
Houston, Texas USA, October 3-6, 1999. Chemical Society, 1992.
22. JENNINGS, H.Y., Waterflood Behaviour of High 40. JOHNSON, C.E. Jr., Status of Caustic and Emulsion
Viscosity Crudes in Preserved Soft and Unconsolidated Methods; SPE 5561, J. Pet. Tech., 85 – 92, January
Cores; SPE 1202, J. Pet. Tech., 116 – 120, January 1976.
1966. 41. DRANCHUK, P.M., SCOTT, J.D. and FLOCK, D.L.,
23. SELBY, R., ALIKHAN, A.A. and FAROUQ ALI, S.M., Effect of the Addition of Certain Chemicals On Oil
Potential of non-thermal methods for heavy oil Recovery During Waterflooding; J. Can. Pet. Tech.,
recovery; J. Can. Pet. Tech., Vol. 28, No. 4, July-August July – September 1974.
1989. 42. MCAULIFFE, C.D., Oil-in-Water Emulsions and Their
24. KUMAR, M., HOANG, V. and SATIK, C., High Flow Properties in Porous Media; J. Pet. Tech., 727-
Mobility Ratio Waterflood Performance Prediction: 733, 1973.
Challenges and New Insights; SPE 97671, SPE 43. JENNINGS, H.Y., JOHNSON, C.E. and MCAULIFFE,
International Improved Oil Recovery Conference, Kuala C.D., A Caustic Waterflooding Process for Heavy Oils;
Lumpur, Malaysia, December 5-6 2005. SPE 4741, J. Pet. Tech., 1344-1352, December 1974.
25. MILLER, K.A., Improving the State of the Art of 44. CHIWETELU, C.I., NEALE, G.H., HORNOF, V. and
Western Canadian Heavy Oil Waterflood Technology; J. GEORGE, A.E., Recovery of a Saskatchewan Heavy Oil
Can, Pet. Tech., Vol. 45, No. 4, 7 – 11, 2006. Using Alkaline Solutions; J. Can. Pet. Tech., Vol. 33,
26. CHATZIS, I., MORROW, N.R. and LIM, H.T., No. 4, 37 – 42, April 1994.
Magnitude and Detailed Structure of Residual Oil 45. THOMAS, S., FAROUQ ALI, S.M., SCOULER, J.R.
Saturation: SPE/DOE 10681, 1982 SPE/DOE 3rd Joint and VERKOCZY, B., Chemical Methods for Heavy Oil
Symposium on Enhanced Oil Recovery, Tulsa, Recovery; J. Can. Pet. Tech., Vol. 40, No. 3, March
Oklahoma USA, April 4 – 7 1982. 2001.
27. PETERS, E.J. and FLOCK, D.L., The Onset of 46. PITTS, M.J., WYATT, K., and SURKALO, H.,
Instability During Two-Phase Immiscible Displacement Alkaline-Polymer Flooding of the David Pool,
in Porous Media; SPE Journal, 249 – 258, April 1981. Lloydminster Alberta; SPE 89386, 2004 SPE/DOE 14th
28. BENTSEN, R.G., A New Approach to Instability Symposium on Improved Oil Recovery, Tulsa,
Theory in Porous Media; SPE Journal, 765 – 779, Oklahoma, USA, April 17 – 21 2004.
October 1985. 47. LIU, Q., DONG, M. and MA., S., Alkaline/Surfactant
29. KANTZAS, A. and BROOK, G., Preliminary Flood Potential in Western Canadian Heavy Oil
Laboratory Evaluation of Cold and Post-Cold Reservoirs; SPE 99791, 2006 SPE/DOE Symposium on
Production Methods for Heavy Oil Reservoirs Part B: Improved Oil Recovery, Tulsa, Oklahoma, USA, April
Reservoir Conditions; J. Can. Pet. Tech., Vol. 43, No. 22 – 26 2006.
10., 39 – 48, October 2004. 48. LIU, Q., DONG, M., YUE, X., and HOU, J., Synergy of
30. ROJAS, G.A. and FAROUQ ALI, S.M., Dynamics of alkali and surfactant in emulsification of heavy oil in
Subcritical CO2/Brine Floods for Heavy-Oil Recovery; brine; Colloids and Surfaces A: Physicochem. Eng.
SPE 13598, SPE REE., 35 – 44, February 1988. Aspects, Vol 273, 219 – 228, 2005.
31. SCOTT, G.R., COLLINS, H.N and FLOCK. D.L., 49. ACEVEDO, S., GUTIERREZ, X., and RIVAS, H.,
Improving Waterflood Recovery of Viscous Crude Oils Bitumen-in-Water Emulsions Stabilized with Natural
By Chemical Control; J. Can. Pet. Tech., 243 – 251, Surfactants”, J. Colloid and Interface Science, Vol. 242,
October-December 1965. 230 – 238, 2001.
32. LI, K., CHOW, K. and HORNE, R.N., Effect of Initial 50. NELSON, R.C. and POPE, G.A., Phase Relationships in
Water Saturation on Spontaneous Water Imbibition; Chemical Flooding; SPE 6773, SPE Journal, October
SPE 76727, SPE Western Regional/AAPG Pacific 1978.
Section Joint Meeting, Anchorage, Alaska USA, May 20- 51. NELSON, R.C., LAWSON, J.B., THIGPEN, D.R. and
22 2002. STEGEMEIR, G.L., Cosurfactant-Enhanced Alkaline
33. LI, K. and HORNE, R.N., Characterization of Flooding; SPE/DOE 12672, SPE/DOE Fourth
Spontaneous Water Imbibition into Gas-Saturated Symposium on Enchanced Oil Recovery, Tulsa, OK,
Rocks; SPE 74703, SPE Journal, 375 – 384, December USA, April 15 – 18 1984.
2001. 52. ATKINSON, I., Interfacial Phenomena and Phase
Behavior Using Alkaline-Surfactant-Polymer Enhanced

9
 Recovery Methods in Medium and Heavy Crude Oils;
M.Sc. Thesis, University of Calgary, Alberta Canada,
2004. 14 600
53. DEZABALA, E.F., VISLOCKY, J.M., RUBIN, E. and 12 500
RADKE, C.J., A Chemical Theory for Linear Alkaline Qo

Pressure drop (kPa)

Flow rate (mL/hr)
10
Flooding; SPE 8997, SPE Journal, April 1982, 245 – DP 400
8
258. 300
6
200
Sand pack dimensions Short Long A Long B 4
Length (m) 0.44 1.15 1.18 2 100
Diameter (cm) 3.81 8.89 8.89 0 0
Sand pack properties 0 50 100 150 200 250
Pore volume (cm³) 165 2592 2680 Tim e (hrs )
Porosity (fraction) 0.377 0.363 0.366
Permeability (D) 6.7 3.0 3.8
Swi (fraction) 0.110 0.089 0.086 Figure 1. Pressure drop and oil flow rate - System A.
Fluid properties (@ 23°C)
Dead oil density (kg/m³) 981.5 981.5 981.5
Dead oil viscosity (mPa⋅s) 11,000 11,000 11,000 60 400

GOR (m³/m³) 13.0 12.4 11.6 50

C um GO R 350
Live oil density (kg/m³) 970 973 963

Cumulative GOR (cc/cc)

DP 300

Pressure drop (kPa)

Live oil viscosity (mPa⋅s) 6478 3596 3303 40
250
Table 1. Sand and fluid properties of the systems.
30 200
150
20
Short Long A Long B 100
10
Max ∆P/L (kPa/m) 530 437 404 50
Max (k/µ)(∆P/L) 0.548 0.365 0.465 0 0
Oil recovery factor (fraction) 0.38 0.20 0.27 0 20 40 60 80 100
Table 2. Pressure gradient and primary oil recovery factor Time (hrs)
in the systems.

Pore Volume injected (QT)B / (QT)A Figure 2. Pressure drop and cumulative produced GOR –
(∆PB) / (∆PA)
short core.
0.35 5.89 4.74
0.40 5.75 5.94
0.44 5.70 6.46 0.14
Table 3. Comparison between flow rates and pressure 0.12 Bulk oil
drops – sand packs A & B. Oil+w ater
0.1

Core properties
Amplitude

0.08
w ater
Length (m) 0.103
0.06
Diameter (m) 0.011
Porosity (fraction) 0.45 0.04
Permeability (D) 17
0.02 oil
Table 4. AS flooding core properties.
0
Waterflood AS Flood 0.1 1 10 100 1000 10000

(RF/PV inj) (RF/PV inj) T2 (ms)

Region A (viscous disp.) 0.257 0.309
Region B (after breakthrough) 0.01 0.03
Region C (after breakthrough) 0.01 0.01 Figure 3. Determination of oil and water content
Table 5. Recovery efficiencies for waterflooding and AS from NMR.
flooding.

10
 12 1
100 1
10
0.9 80

Water cut (fraction)

Recovery (%OOIP)
8 0.9

Water cut (fraction)

Pressure (kPa)
60
6 0.8
0.8
Pin
4 Rec 40
WC
WC 0.7
2 0.7
20

0 0.6
0 0.5 1 1.5 0 0.6
0 500 1000 1500 2000
PV injected
Time (hrs)
Figure 4. Recovery and water cut of short sand pack.
Figure 7. The inlet pressure profile of sand pack A.

10 1
400 14
8 0.9

Cumulative production (mL)

12

Water cut (fraction)

Pressure (kPa)

Recovery (% OOIP)
300
6 0.8 10

8
4 0.7 200
6
Pin Cum prod
2 0.6 4
WC 100 Rec
2
0 0.5
0 500 1000 1500 2000 2500 0 0
0 0.5 1 1.5 2 2.5
Tim e (hrs)
PV injected
Figure 5. Inlet pressure profile of short sand pack.
Figure 8. Recovery profile of sand pack B.

250 12
180 1
Cumulative production (mL)

200 10
160
Recovery (% OOIP)

8 140
A

Water cut (fraction)

150
120
Pressure (kPa)

6 1 2
100 Pin
100
WC 0.8
4 80
Cum prod
50 60
Rec 2
40
0 0
20
0 0.1 0.2 0.3 0.4 0.5
0 0.6
PV injected
0 500 1000 1500 2000 2500
Tim e (hrs)
Figure 6. Recovery profile of sand pack A.
Figure 9. Inlet pressure profile of sand pack B.

11
 150 1 1

0 ppm NaCl

Minimum IFT (mN/m)

0.9
20 000 ppm NaCl

Water cut (fraction)

100 Pin_B
Pressure (kPa)

Pin_A 0.8 0.1

WC_B
WC_A 0.7
50

0.6 0.01
0 5000 10000 15000 20000
0 0.5 Alkali Content (ppm )
0 0.5 1 1.5 2
PV injected
Figure 13. Oil-water interfacial tension measurements: 0.1
Figure 10. Pressure and water cut comparison – sand packs wt% surfactant, varying alkali content.
A & B.

450 35
1 400
30
350

Recovery (% OOIP)
25
300

Pressure (kPa)
250 20
Water cut (fraction)

0.8 Pressure
200 15
150 Recovery
10
100
5
0.6 A 50

B 0 0
0 2 4 6 8 10
PV injected
0.4
0 0.1 0.2 0.3 0.4 0.5
PV injected Figure 14. Pressure and recovery profile for a linear dead
oil waterflood.
Figure 11. Effect of injection rate on produced water cut.

10 450
400
8 350
Recovery (% OOIP)

300
Pressure (kPa)

AS Flood
6 250
Waterflood
200
4 150
A 100
2 B 50
0
0 0 0.5 1 1.5
0 0.1 0.2 0.3 0.4 0.5 PV injected
PV injected

Figure 15. Pressure profile at early times for a waterflood

Figure 12. Recovery profile comparison between sand and an AS flood.
packs A & B.

12
 450 450 1
400 400 0.95
350 350

Water cut (fraction)

0.9
300 300

Pressure (kPa)
Pressure (kPa)
AS Flood 0.85
250 250
Waterflood Pressure 0.8
200 200
Water cut 0.75
150 150
100 100 0.7

50 50 0.65

0 0 0.6
0 2 4 6 8 10 0 5 10 15

PV injected PV Injected

Figure 16. Complete pressure profile for a waterflood vs. an Figure 19. Pressure and waterflood profile for an AS flood
AS flood. after a waterflood.

70

60 C
Recovery (% OOIP)

50 A B

40

30

20
AS Flood
10
Waterflood
0
0 2 4 6 8 10
PV injected

Figure 17. Oil recovery profile for a waterflood and an AS

flood at the same injection rate.

450 80
400 70
Pressure
350
Recovery (% OOIP)

60
Recovery
300
Pressure (kPa)

50
250
40
200
30
150
100 20

50 10
0 0
0 5 10 15
PV Injected

Figure 18. Pressure and production profile for an AS flood

after a waterflood.

13

View publication stats