SPE 146782

Laboratory and Simulation Study of Optimized Water Additives for Improved Heavy Oil Recovery X. Wang, SPE, P. Luo, SPE, Y. Zhang*, SPE, V. Charkovskyy, SPE, S. Huang, SPE, Saskatchewan Research Council *Now with Energy Resources Conservation Board of Alberta

Copyright 2012, Society of Petroleum Engineers This paper was prepared for presentation at the SPE Heavy Oil Conference Canada held in Calgary, Alberta, Canada, 1214 June 2012. 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

This paper discusses a laboratory evaluation of the feasibility of different chemical flooding strategies and a simulation study to

optimize the feasible strategies for a west-central Saskatchewan heavy oil reservoir. The integrated experimental approach was

composed of oil/brine interfacial tension (IFT) measurements, polymer viscosity measurements, wetting tendency measurements,

and sandpack coreflood tests. The experimental results showed that the equilibrium interfacial tension between reservoir oil and

formation brine could be lowered to an ultralow level (0.05 mN/m) by adding a certain concentration of alkali and surfactant into

the brine. The addition of alkali and surfactant caused the wettability characteristics in all tested systems to become oil-wet. All of

the polymer solutions exhibited pseudo-plastic behaviour, i.e., the apparent viscosity decreased with increasing shear rate. A series

of sandpack coreflood tests were carried out to investigate the recovery performance of alkali + surfactant, polymer, and

alkali + surfactant + polymer (ASP) floods. Enhanced oil recoveries (from the chemical flood and extended waterflood) varied

significantly from 0.71 to 14.65% OOIP. The coreflood results suggest that in enhanced waterflooding for recovering viscous

heavy oil, mobility control by polymer is more important than IFT reduction by alkaline/surfactant. In the simulation study, the

relative permeability curves were obtained through history matching. Then, as the sensitive operating parameters, the ASP slugs

and polymer concentrations were tuned to show their effects on enhanced heavy oil recovery (EHOR). In summary, ASP flooding

provides synergistic effects that can maximize the recovery performance.

Introduction At present, approximately 22.9 billion barrels of original oil in place (OOIP) have been discovered in west-central Saskatchewan.

These reservoirs are described as lying in the heavy oil belt, which extends from the Alberta border well into Saskatchewan

(100120 km). The oils deposited in this belt are very viscous: the viscosity ranges from 2,000 mPas to more than 10,000 mPas at standard conditions. As well, the majority of these heavy oil reservoirs have very thin pay zones (

2 SPE 146782

caused by the adverse mobility ratio between oil and water (Bryan and Kantzas 2007). Although low interfacial tension (IFT) is a

precondition for releasing the trapped residual oil from porous media, oil recovery is usually very low without mobility control in

heavy oil reservoirs (Wang et al. 2007; Zhang et al. 2010; 2009; 2006).

Since the addition of any chemical(s) to injected water is expensive, it is necessary to use chemical additives of low cost or in

small amounts, and to maximize recovery factors and ensure good economics. In this paper, the developed waterflood

enhancement technology combines 1) an interfacial instability mechanism created by the chemical (alkali, surfactant) solution as it

lowers interfacial tension, with 2) mobility improvement by the polymer solution. The synergistic effectiveness of both processes

causes the residual heavy oil to be broken up into small droplets, entrained in the reduced-mobility water phase, and moved toward

the production well.

Experimental Design

Sampling and characterization of crude oil and brine Crude oil and brine samples were obtained from a heavy oil reservoir in west-central Saskatchewan, which has been under

waterflood since 1988 and where the water cut increased sharply soon after the waterflood was started. The density and viscosity

of the cleaned dead oil were measured at the reservoir temperature of 30C. The true boiling point (TBP) distribution for the oil

was measured using gas chromatographic simulated distillation. In addition, the average molecular mass of the oil was determined

by freezing point depression. The formation brine, represented by produced water, was filtered four times through a 2.5-m pore

size filter, and then analyzed for density, viscosity, refractive index, conductivity, pH, total dissolved solids, and mineral

concentrations.

Chemical selection and fluid behavior measurements After a preliminary screening of alkalis on the basis of the collected oil/brine samples, sodium carbonate (Na2CO3) was mixed with

a surfactant Steol CS-460 (Stepan, USA) to evaluate the interfacial tension (IFT) between the sampled oil and aqueous phase. The

spinning drop technique (KRSS SITE100 spinning drop tensiometer), a convenient and accurate method to measure low IFT

between oil and aqueous phases, was used in this study.

Based on the properties of the collected oil/brine samples, a partially hydrolyzed polyacrylamide 3630S (SNF Floerger, France)

was selected to test its rheological behavior with the formation brine. This polymer has a molecular weight of 20 million Daltons,

with hydrolysis levels of 25% to 30%. Viscosities of polymer solutions were measured using a Brookfield DV-I+ cone/plate

viscometer at a shear rate range of 0 to 230 S-1 and temperature of 30C controlled by a water bath.

Live oil preparation and property characterization The reservoir fluid (live oil) was made by recombining methane (CH4) with the cleaned dead oil at a saturation pressure of 6.7

MPa. The following pressure-volume-temperature (PVT) properties of the reconstituted (live) oil were determined: viscosity,

density, bubblepoint pressure, gas/oil ratio (GOR), and formation volume factor (FVF). The viscosity of the live oil as a function

of pressure was measured using a precalibrated inline capillary tube viscometer. In the viscosity measurement, the test fluid was

pressurized above the bubblepoint in the PVT cell and then circulated through the capillary tube at a constant rate. The pressure

drop across the tube was recorded and the Hagen-Poiseuille equation was used to calculate the reservoir oil viscosity. The viscosity

of the fluid at the saturation pressure was obtained by extrapolation of the measured viscosity-versus-pressure data. Density as a

function of pressure was determined above the saturation pressure and at reservoir temperature in a high-pressure Anton Paar

densitometer DMA 512 cell, into which the single-phase fluid was charged. The GOR and FVF of the reservoir fluid were

determined by withdrawing a known volume of the single-phase fluid (at reservoir temperature) and flashing to atmospheric

conditions in a sampling bottle. The oil was accumulated in the bottle and the gas was collected in a gasometer. The collected oil

was weighed, and the gas volume was recorded and its composition analyzed by a gas chromatograph.

Wetting tendency study Wettability is commonly defined as the tendency of a fluid to spread on or adhere to a solid surface in the presence of other

immiscible fluids. It is critical to understand the wettability behaviour of porous media after the displacing fluid is injected in a

reservoir because it influences various important reservoir properties, such as the distribution of oil and water on reservoir rock,

relative permeability characteristics, and consequently the oil recovery. The contact angle is a direct measurement of the

wettability of a specific surface. The contact angle, , at the liquidsolid surface is, by convention, measured through the water phase and can range from 0 to 180. When the contact angle is between 0 and 60, the surface is preferentially water-wet and when

it is between 180 and 120, the surface is considered to be oil-wet. If the contact angle is between 60 and 120, neither fluid

preferentially wets the solid, i.e., wettability is intermediate. If contact angles are measured through the oil phase, as was the case

in this study, the inverse of the above rule defining wettability applies.

SPE 146782 3

The objective of the wetting tendency study was to determine the influence of chemical types (alkali and surfactant) and

concentrations on the wettability behaviour of the two immiscible fluids on a solid surface. In this study, wetting tendencies

between chemical solutions and the oil were determined by measuring the contact angle on two different solid surfaces (glass and

quartz). The measurements were conducted with the dead oil at 30C and atmospheric pressure using the KRSS DSA100 contact angle meter. Digital images of a sessile oil drop on the solid surface, as shown in Fig. 1, can be acquired and analyzed

automatically to obtain the contact angle using the software bundled with the meter.

Coreflood studies Four coreflood tests were conducted using the reconstituted reservoir fluid and formation water. The coreflooding appraratus is

depicted schematically in Fig. 2. It consists of: 1) a pump for injecting different fluids into the core through floating-piston cells;

2) pressure gauges and differential pressure transducers to measure the injection/production pressures and the pressure drop across

the core during the floods; 3) a core holder inside which the core is placed; and 4) produced fluids (oil/water) collectors at the

outlet. The dead volumes of all the flow lines were measured and accounted for in all the material balance calculations.

To obtain results that are more representative of actual field conditions, the four linear coreflood experiments were carried out

using unconsolidated sandpacks. These packs measured around 30.0 cm long and 5.1 cm in diameter. For each coreflood, field

sands were packed to ensure the same wettability for all the tests. The sands were packed in a lead sleeve that was placed in a

cylindrical core holder, and subjected to confining pressure to seal the assembly. A pneumatic vibrator was used to wet-pack five

segments of sands in sequence. Each segment was vibrated for about 15 minutes.

For measurement of pore volume, brine was displaced from the core with dry compressed air. Approximately 1.5 pore volumes

of acetone was then slowly pumped through the core and drained out. The remaining acetone was evaporated with dry compressed

air. The core was then evacuated to a pressure less than 75 torr and imbibed with brine to determine the pore volume (PV).

After the measurement of pore volume, the core pressure was increased to the operating pressure of 10.0 MPa. At the same

time, a net overburden pressure of 6.0 MPa was applied to the lead sleeve and was maintained during the pressurizing stage. The

sandpack was then saturated with brine, and the stabilized pressure drop across the sandpack was measured at the test temperature

of 30C. Then the core was aged for one week to re-establish ionic equilibrium. The differential pressures across the core during

brine injection were recorded, and the absolute permeability of the core to brine was calculated from Darcys law. The brine was displaced with the live oil to establish connate water saturation (Swc). The endpoint permeability of the core to the oil in the

presence of brine was estimated. The core was then aged for a recommended period of one week.

After aging and permeability measurement, an initial waterflood was started. This step was terminated when oil production

became negligible. The fluid recoveries were used to calculate waterflood residual oil saturation. Then, the core was flooded with

the various chemical solutions (alkali/surfactant/polymer) until the designated number of pore volumes was reached. The enhanced

recovery process was then followed by an extended waterflood (EWF). The injection rate of each stage was kept at 20 cm3/hr.

Throughout the experiments, produced fluids were flowed from the core through a backpressure regulator (BPR) and then flashed

to atmospheric pressure in a separator. The volume of produced liquids was recorded by collection in a graduated glass cylinder.

Pressures at both production and injection ends were monitored throughout the flooding process by pressure transducers.

Experimental Results and Discussion

Sampling and characterization of crude oil and brine The gravity of the dead crude oil was measured to be 16.5API, and its acid number is 1.09 mg KOH/g. The molar mass measured

using freezing point depression was determined to be 375 kg/kmol. The density and viscosity of the crude oil at 30C were measured to be 945.9 kg/m3 and 376 mPas, respectively. The carbon number distribution of the crude oil, characterized by the

equivalent carbon number pseudo-components up to C30, is shown in Fig. 3.

Table 1 shows the properties of formation brine. At atmospheric pressure and 30C, the brines density and viscosity were measured to be 1005.7 kg/m3 and 0.82 mPas, respectively. The brine contains a very high amount of total dissolved solids (13,600 mg/L at 180C) and is considered a hard water because of the high contents of calcium (40 mg/L) and magnesium (70 mg/L).

Thus, a quick precipitation test was carried out at room conditions to evaluate the precipitation tendency when adding 0.5 wt.%

Na2CO3 in the brine. After standing for 24 hours until the precipitates settled, the clear supernatant was then used for the

following tests.

Reservoir Fluid Characterization The cleaned dead oil was recombined with CH4 to give a saturation pressure of 6.7 MPa at the reservoir temperature of 30C. This

corresponded to the reservoir fluid saturation pressure (6.7 MPa) identified in the data provided by the client. The PVT properties

of the reconstituted oil (i.e., live oil) are presented in Table 2. The gas/oil ratio of the live oil was measured to be 18.96 sm3/m3.

The density and viscosity of the undersaturated live oil (i.e., pressure is higher than bubblepoint pressure) as a function of pressure

4 SPE 146782

are plotted in Fig. 4. These parameters were measured above the saturation pressure at the reservoir temperature, which made it

possible to determine the value at saturation pressure by a short extrapolation. The live oil viscosity at the saturation pressure of

6.7 MPa was estimated to be 169 mPas, down about half from the dead oil viscosity (376 mPas) at reservoir temperature.

Measurements of interfacial tensions A surfactant, either formed in-situ or added externally, reduces the interfacial tension (IFT) between oil and water as to displace

the discountinuous trapped oil that remains after waterflooding. An alkali reacts with the natural acids in oils, forming in-situ

surfactants. In general, the higher the acid number of the oil, the better candidate the reservoir might be for an alkaline flood. As

the acid number of the heavy oil was 1.09 mg KOH/g, it was expected that addition of a certain concentration of alkali and a small

amount of surfactant would produce a low oil/water interfacial tension.

Without addition of any chemicals, the IFT of the crude oil and brine was measured to be 8.8 mN/m using axisymmetric drop

shape analysis technique for the pendant drop case. When alkali was added to the brine, the IFT decreased with an increase in

alkali concentrations at a certain range, as shown in Fig. 5, which shows the estimated equilibrium IFTs based on the average

values of dynamic IFTs observed during the last 10 minutes of the measurements. A reaction between fatty acids in the crude oil

and Na2CO3 occurs rapidly at the oil-brine interface, whereas desorption of the active species from the interface is considerably slower (Shah 1998). In general, stabilized IFT values were reached within one hour. Surfactant (Steol CS-460) was mixed with 0.5

wt% Na2CO3. The measured equilibrium IFTs are shown in Fig. 6. Surprisingly, the measured IFT increased with increasing

surfactant concentration. The lowest IFT of 0.05 mN/m was achieved at 50 ppm. To ensure the benefit of maintaining low IFT

during flooding, a relatively higher surfactant concentration than that which provides the minimum interfacial tension shouble be

considered, so as to compensate for chemical loss through partitioning into the oil phase and adsorption on the formation sands.

Therefore, a combination of 0.01 wt% + 0.5 wt% Na2CO3 was chosen to play the role of IFT reducer in optimizing coreflood oil

recovery efficiency by augmenting injection water with chemical.

Effect of fluid characteristics on wetting tendency Reservoir wettability is an essential factor in enhancing oil recovery, significantly influencing oil and water flow and distribution

during flow through porous media. In a chemical flood, the injected chemicals will partition into the oil and water phases, and also

adsorb on reservoir rock surface, which might result in modifying the wetting preference of the reservoir rocks. Schramm et al.

(2003) showed that surfactant adsorption can shift wettabilitys from oil-wetting towards water-wetting. When the orientation of the

adsorbed surfactant is such that its hydrophobic tail groups point away from the rock surface or along the surface, it will result in a

decrease in water-wetting and an increase in oil-wetting. Xu et al. (2006) reported that the highest oil recoveries are associated

with intermediate and mixed-wettability states.

The baseline contact angles were determined by measuring a sessile dead oil drop on quartz and glass surfaces in the presence

of brine or water at room conditions (30C and 0.1 MPa). The contact angles (measured through the oil phase) of the dead oil on

glass and quartz solid surfaces in brine were determined to be 156 and 148, respectively, which indicated that oil is strongly non-wetting on the solid surface. When the chemicals (0.5 wt% Na2CO3 and 100 ppm Steol CS-460) were added to the aqueous phase,

the contact angles of the dead crude oil on both glass and quartz in the brine solution were measured to be 60 at the beginning,

and they gradually decreased to about 40 after one hour. The results indicate that after addition of chemicals, the wettability of original dead oil/brine systems on both surfaces (glass and quartz) was changed from intermediate oil-wet to strongly oil-wet. One

possible explanation for the wettability alterations in the presence of chemicals is that chemical adsorption occurred on the solid

surfaces, and surfactant partition occurred in both the oil and water phases.

Rheological measurements of polymer solutions Polymer is used to increase the displacing fluids viscosity to improve the mobility ratio and sweep efficiency of the flood. Because polymer is a relatively expensive material, the proper type and concentration of polymers are very important for designing

a cost-effective chemical EOR process. Critical polymer properties include cost-effectiveness (e.g., cost per unit of viscosity),

resistance to degradation (mechanical or shear oxidative, thermal, microbial), tolerance of reservoir salinity and hardness,

adsorption on rock, trapping by pore throat, inaccessible pore volume, permeability dependence of performance, rheology, and

compatibility with other chemicals that might be used (Wang et al. 2007).

On the basis of the physical properties of the sampled brine, Flopaam 3630S, a polyacrylamide manufactured by SNF Floerger

for EOR, was selected in this research. As shown in Fig. 7, the polymeric solutions exhibited pseudo-plastic behaviour, i.e.,

apparent viscosity decreases with increasing shear rate.

Coreflood Results A total of four linear corefloods were performed to investigate several waterflood enhancement processes, and to determine the

proper injection strategy. The injection strategies used for the four corefloods included alkalinesurfactant (AS) flooding for Run

SPE 146782 5

L1, alkalinesurfactantpolymer (ASP) flooding with higher polymer concentration for Run L2, polymer (P) flooding for Run L3, and alkalinesurfactantpolymer (ASP) flooding with lower polymer concentration for Run L4. In these tests, chemical slugs were injected into the sandpacked cores right after the initial waterflood when the residual oil saturation was reached. Table 3 lists the

detailed operating conditions and recovery performance for each run. The initial oil saturations of the four packed cores were in

the range of 82.3 to 86.3 % PV. The absolute permeabilities to brine of the cores varied from 2.08 to 2.78 m2, which are close to

that of the studied reservoir.

Run L1IWF + AS + EWF An initial waterflood (IWF) was conducted in the core, followed by injection of a 0.5 PV slug of 0.5 wt% Na2CO3 + 100 ppm

Steol SC-460 solution, then an extended waterflood (EWF) until irreducible water saturation was reached. The cumulative oil

production curve (Fig. 8) shows that the initial waterflood produced 50.52% OOIP. Incremental oil production by alkalinesurfactant injection and a final extended waterflood was very poor only 0.71% OOIP. Fig. 8 also shows the differential pressure recordings across the core and the oil recovery profile obtained at a constant injection rate. At the beginning of the waterflood, the

average differential pressure was about 10 kPa until the breakthrough of injection fluids. The average differential pressure during

the latter part of the IWF was about 6 kPa, and kept about the same during alkalinesurfactant injection and EWF. There are two possible reasons for the poor incremental oil recovery in this run. First, the effective alkalinity in the solution was low due to ion

exchange, precipitation of solids, and rock dissolution. As well, the surfactant concentration was rather low, such that favourable

interfacial activity of the slug against the trapped oil was only barely maintained. Second, without addition of polymer, the injected

solution flowed along the paths that had been established during the initial waterflood, leaving some areas poorly swept due to the

unfavourable mobility ratio between the water and the residual viscous heavy oil.

Run L2IWF + ASP + EWF The chemical slug for Run L2 was formulated to be 0.5 wt% Na2CO3 + 100 ppm Steol SC-460 + 0.1 wt% 3630S. The flood

sequence of Run L2 consisted of an initial waterflood, followed by ASP slug injection, then an extended waterflood. The endpoint

permeabilities to brine and initial oil saturations of the cores used in Run L2 were slightly lower than those in Run L1. The

cumulative oil production curve (Fig. 9) shows that the initial waterflood produced 49.09% OOIP, which was about 1.5% lower

than that in Run L1. The experimental data indicate that after the injection of 0.5 PV of ASP solution, the oil recovery by EWF

was 14.65% OOIP, which was significantly higher than the 0.71% OOIP in Run L1. This indicates that the polymer improved the

mobility ratio and was able to bank oil mobilized by the alkaline and surfactant. The additional oil recovery by ASP injection

might be largely attributed to the high differential pressures across the core due to reduction of the water mobility by addition of

the polymer. As plotted in Fig. 9, the average differential pressure during the latter part of the initial waterflood was about 10 kPa.

During the ASP slug injection, the differential pressure rapidly rose to over 150 kPa. After 0.5 PV of ASP slug injection, the

differential pressure increased continuously to over 200 kPa during the first PV of EWF. The reason for this was believed to be a

combination of several factors, such as in-situ formation of emulsions, high polymer solution viscosity, formation of an oil bank,

and polymer retention on the surfaces of the grains, thereby increasing the resistance to flow. The differential pressure then

gradually decreased to about 150 kPa until the end of the EWF.

Run L3IWF + P + EWF The chemical of Run L3 was chosen to consist of only 0.1 wt% polymer. The aim of injecting a polymer-alone slug was to

distinguish the effect of mobility improvement by the polymer to obtain incremental oil recovery. As listed in Table 3, the

endpoint permeabilities to brine and oil were 2.20 and 1.88 m2, respectively, which were very close to those in Run L2. The

initial waterflood produced 49.64% OOIP. Fig. 10 shows the cumulative oil recovered by IWF, EOR, and EWF, as well as the

differential pressures. During 0.5 PV of polymer slug injection, oil recovery was 0.21% OOIP. The EWF obtained an enhanced oil

recovery of 10.05% OOIP. Therefore, the total enhanced oil recovery of 10.26% OOIP was lower than the 14.65% OOIP in Run

L2 by ASP slug but still much higher than the 0.71% OOIP in Run L1 by AS slug. It was seen that the summation of enhanced oil

recoveries from Run L1 and Run L3, which were considered to be the respective contributions of AS and P, was less than the

enhanced oil recovery in Run 2. This indicates that the ASP system could introduce synergistic effects that maximized the

recovery efficiency by the chemical slug. The differential pressure behaviour of Run L3 (see Fig. 12) had a similar pattern to that

observed in Run L2 but with different scales. The average differential pressures were quite stable during the initial waterflood

stage (about 10 kPa). During polymer slug injection there was a sharp increase of average differential pressure. At the end of

polymer slug injection, the average differential pressure was about 100 kPa and kept increasing to about 140 kPa during the first

0.5 PV of EWF. It is noted that this average differential pressure was lower than that in the chemical slug injection process in Run

L2. It is speculated that W/O emulsions were formed in situ and plugged some of the flow channels so as to create extra

differential pressure.

6 SPE 146782

Run L4IWF + ASP + EWF To optimize the ASP formula, Run L4 was carried out using 0.5 wt% Na2CO3 + 100 ppm Steol SC-460 + 0.05 wt% 3630S. The

only difference between Runs L4 and L2 was that Run L4 applied half the amount of polymer. The injection stages included an

initial waterflood, followed by ASP injection, and then an extended waterflood. As shown in Fig. 11, the initial waterflood

recovered 49.26% OOIP in the core. ASP flooding followed by EWF produced an additional 5.06% OOIP, which was

considerably less than in Runs L2 and L3. This indicates that at such a low polymer concentration, polymer loss by adsorption and

degradation by shearing resulted in deteriorated performance of the polymer. Fig. 11 also presents the differential pressures during

IWF, ASP slug injection, and a final EWF in the coreflood. The differential pressures at the IWF stage of Runs L2 and L4 were

very similar. Surprisingly, when the ASP slug was injected in Run L4, the average differential pressures rose rapidly to as high as

370 kPa, much higher than the 200 kPa in Run L2, although a lower concentration of polymer was used in Run L4. It was

speculated that some fine sands migrated with the flowing fluids and accumulated downstream of the core or production line to

create extra resistance to flow.

Coreflood Summary

Low enhanced oil recovery (0.71% OOIP) by the AS flood in Run L1 indicates that oil/water IFT reduction by AS was not

sufficient to improve heavy oil recovery. The oil left behind in the waterflooded core was largely bypassed due to viscous

fingering caused by the adverse mobility ratio between oil and water. This suggests that the addition of polymer as a mobility

control agent was extremely important in enhanced waterflooding for recovering viscous heavy oil. For the polymer-alone flood in

Run L3, mobility control was the only recovery mechanism, and the result was rather satisfactory. For the ASP flood, both IFT

reduction and mobility control played important roles in recovering additional oil. Polymer loss and degradation in the core made

it insufficient to achieve desirable performance in Run L4. These coreflooding results indicate that in enhanced waterflooding for

viscous heavy oil, mobility control by polymer is more important than IFT reduction by alkaline/surfactant. ASP provides

synergistic effects that can maximize the recovery performance.

Simulation Study

Model creation and history match In order to test the effects of slug size and polymer concentration on oil recovery, the optimum coreflood test (Run L2) was

selected for building a simulation model and conducting further case studies. First, the reservoir fluid model was tuned using

CMGs WinProp module. The simulated live oil properties are listed in Table 2 for comparison with the experimental results. The experimental and simulated density and viscosity for undersaturated live oil at different pressure points are plotted in Figs. 12 and

13. The results were then output to CMGs STARS module for history matching. The history-matched results are given in Figs. 14 and 15. It can be seen that a good match was achieved for the both PVT and coreflood test data. According to different

capillary number (Nc) ranges in waterfloods and ASP floods (Van Quy et al. 1983; Amaefule et al. 1982), the three different

water/oil relative-permeability sets obtained are given in Fig. 16. It shows that when logNc is equal to or smaller than -5.9, the

corresponding curve set was used for simulating the waterflood process; when logNc is equal to or smaller than -3.0, the

corresponding curve set was used for simulating the chemical flood; when the flood becomes miscible (logNc < -1.1), the straight

line curve set was used. In general, the chemical flood will give relatively higher oil relative permeabilities but lower water relative

permeabilities at various water saturations. Based on the well-matched simulation Run L2, different study cases were performed.

Effects of chemical slug size and polymer concentration on oil recovery In order to understand the effects of slug size on oil recovery, different cases were simulated three different ASP slug sizes. The

cumulative oil production of each case is shown in Fig. 17. It can be seen that the larger slug size can result in higher oil recovery.

Regarding the oil recovery factor, slugs of 0.2 PV, 0.5 PV and 1.0 PV give, respectively, total oil recovery factors equal to 60.4%,

63.6%, and 68.0%.

ASP slugs with three different polymer concentrations of 0.05 wt%, 0.1 wt%, and 0.2 wt% were simulated The results of each

run are given in Fig. 18, which clearly shows that a higher polymer concentration can lead to higher oil production. The ASP slug

with 0.2 wt% polymer obtained 67.4% oil recovery, while the others with 0.1 wt% and 0.05 wt% polymer obtained only 63.6%

and 61.7% oil recovery, respectively.

Therefore, according to the discussion above, both the chemical slug size and polymer concentration will have significant

effects on oil recovery. It is necessary to balance these factors with the chemical cost so as to perform a successful and economic

field-scale chemical flood.

SPE 146782 7

Conclusions In this study, it was found that addition of polymer as a mobility control agent was extremely important in waterflood

enhancement for heavy oil recovery. The injection strategy must be properly designed to enhance waterflooding efficiency using

an economical chemical concentration. The following conclusions are made.

1. The 0.5 wt% Na2CO3 + surfactant Steol CS-460 system could generate favourably low interfacial tensions over a wide range of

surfactant concentrations (50 to 500 ppm). The minimum IFT was found to be 0.05 mN/m at the surfactant concentration of 50

ppm.

2. The contact angles (measured through the oil phase) of the dead oil in brine on glass and quartz solid surfaces were determined

to be approximately 156 and 148, respectively, which indicated a strong water-wetting tendency. The presence of chemicals caused the wettability characteristics in all chemical fluid systems to become oil-wet on glass and quartz.

3. Oil recovery was highest by ASP with the higher polymer concentration, followed by polymer alone, ASP with the lower

polymer concentration, and AS.

4. Two main mechanisms responsible for enhanced heavy oil recovery using chemical additives are reduction of oil/water

interfacial tension and improvement of the mobility ratio between oil and water. Nevertheless, the contribution of the latter

mechanism is more important than that of the former for enhanced heavy oil recovery.

5. Oil relative permeabilities in a chemical flood are higher than those in a waterflood, whereas the water relative permeabilities

are in the opposite trend. Furthermore, a larger chemical slug size and polymer concentration can contribute relatively higher oil

production, but at greater chemical cost.

Acknowledgements The authors acknowledge the financial support of the Petroleum Technology Research Centre and the participating oil companies

which include BP Exploration (Alaska) Inc., Canadian Natural Resources Limited, ConocoPhillips Company, Devon Energy

Corporation, Husky Oil Operations Limited, Nexen Inc., PennWest Exploration, Shell Canada Energy, and Total E&P Canada Ltd.

The authors also wish to express their sincere thanks to B. Schnell, K. Rispler and R. Shi for carrying out the experimental

measurements, and to B. Tacik for editorial support. Finally, we appreciate the permission granted by Saskatchewan Research

Council to publish this paper.

References Amaefule, J.O. and Handy, K.L. 1982. The Effect of Interfacial Tensions on Relative Oil/Water Permeabilities of Consolidated

Porous Media. Paper SPE 9783, SPE J. 22 (3): 371381.

Bryan, J. and Kantzas, A. 2007. Enhanced Heavy-Oil Recovery by Alkali-Surfactant Flooding. Paper SPE 110738 presented at the

SPE Annual Technical Conference and Exhibition, Anaheim, California, 1114 November.

Green, D.W. and Willhite, G.P. 1998. Enhanced Oil Recovery, TX: Society of Petroleum Engineers.

Liu, Q., Dong, M.Z., and Ma, S. 2006. Alkaline/Surfactant Flood Potential in Western Canadian Heavy Oil Reservoirs. SPE 99791

presented at the 2006 SPE/DOE Symposium on Improved Oil Recovery, Tulsa, 2226 April.

Miller, K.A. 2005. State of the Art of Western Canadian Heavy Oil Water Flood Technology. CIPC paper 2005-251 presented at

the 6th Canadian International Petroleum Conference (56th Annual Technical Meeting), Calgary, 79 June.

Renouf, G. 2007. Do Heavy and Medium Oil Waterfloods Differ? CIPC Paper 2007-055 presented at the 8th Canadian

International Petroleum Conference, Calgary, 1214 June.

Schramm, L.L, Stasiuk, E.N., and Marangoni, D.G. 2003. Surfactants and Their Applications. Annual. Rep. Prog. Chem., Sect. C

99, 348.

Shah, D.O. 1998. Micelles, Microemulsions, and Monolayers; Science and Technology, 249287. New York City: Marcel Dekker, Inc.

Van Quy, N. and Labrid, J. 1983. A Numerical Study of Chemical Flooding Comparison with Experiments. Paper SPE 10202,

SPE J. 23 (3): 461474.

Wang, D.M., Seright, R.S., Shao, Z.B., and Wang, J.M. 2007. Key Aspects of Project Design for Polymer Flooding. Paper SPE

109682 presented at the SPE Annual Technical Conference and Exhibition, Anaheim, California, 1114 November.

Wang, J., and Dong, M. 2007. A Laboratory Study of Polymer Flooding for Improving Heavy Oil Recovery. CIPC Paper 2007-

178 the 8th Canadian International Petroleum Conference, Calgary, 1214 June.

Xu, W., Ayirala, S.C., and Rao, D. 2006. Compositional Dependence of Wetting and Contact Angles in Solid-Liquid-Liquid

Systems under Realistic Environments. Canadian Journal of Chemical Engineering 84 (1): 4451.

8 SPE 146782

Zhang, Y.P., Sayegh, S., and Huang, S. 2010. Coupling Immiscible CO2 Technology and Polymer Injection to Maximize EOR

Performance for Heavy Oils. Journal of Canadian Petroleum Technology, 49 (5): 2533.

Zhang, H., Dong M., and Zhao, S. 2009. Which One Is More Important in Chemical Flooding for Enhanced Court Heavy Oil

Recovery, Lowering Interfacial Tension or Reducing Water Mobility? Energy & Fuels, 24, 18291836.

Zhang, Y.P., Sayegh, S., and Huang, S. 2006. Enhanced Heavy Oil Recovery by Immiscible WAG Injection. Paper SPE 2006-014

presented at the 7th Canadian International Petroleum Conference (57th Annual Technical Meeting), Calgary, 1315 June.

TABLE 1Properties of Produced Water

Density (kg/m3) at 15C 1009.1

at 20C 1008.1

at 30C 1005.7

Viscosity (mPas) at 15C 1.16

at 20C 1.02

at 30C 0.82

Refractive index at 25C 1.3353

Conductivity (S/cm) at 25C 17200

pH at 20C 7.78

TDS (mg/L) at 180C 13600

Chloride (Cl-) (mg/L) 8100

Sulfate (SO42-) (mg/L) 420

Sodium (Na+) (mg/L) 4900

Potassium (K+) (mg/L) 40

Magnesium (Mg2+) (mg/L) 70

Calcium (Ca2+) (mg/L) 40

Iron (Fe2+) (mg/L) 0.2

Barium (Ba2+) (mg/L) 3.8

Manganese (Mn2+) (mg/L) < 0.1

TABLE 2Physical Properties of Reservoir Fluid at 30C

Properties Reconstituted Oil

(Experimental Data) Reconstituted Oil (Simulated Data)

Error (%)

Saturation Pressure, Psat (MPa) 6.7 6.7 0.01%

Density, at Psat (kg/m3) 934.6 935.6 0.11%

Viscosity, at Psat (mPas) 169.0 165.8 1.82%

Formation Volume Factora (m3/m3) 1.0361 1.0425 0.61%

Gas/Oil Ratio (sm3/m3) 18.96 18.86 0.52%

SPE 146782 9

TABLE 3Operating Conditions & Recovery Performances of Corefloods

Run Number Run L1 Run L2 Run L3 Run L4

Fluids Used in Flood

Type of Oil Saturating Core Live Oil Live Oil Live Oil Live Oil

Core Properties

Core Sandpack Sandpack Sandpack Sandpack

Total Length of Core, cm 30 30 30 30

Diameter of Core, cm 5 5 5 5

Core Pore Volume, cm3 214.6 212.9 213.2 216.1

Coreflood Parameters

BPR Pressure, MPa 10 10 10 10

Operating Temperature, C 30 30 30 30

Brine Permeability, m2 2.78 2.22 2.20 2.08

Oil Permeability at Swc, m2 2.44 1.88 1.88 1.66

Initial Oil Saturation, %PV 82.3 83.1 84.5 86.3

Injection Fluid Properties

Brine Viscosity at 30C, mPas 0.82 0.82 0.82 0.82

Oil Viscosity at 10 MPa & 30C, mPas 192.3 192.3 192.3 192.3

Injection Rate of IWF, cm3/h 20 20 20 20

EOR Injection Sequence*2 A-SEWF A-S-PEWF PEWF A-S-PEWF

Chemical Solution A = Na2CO3; S = Steol SC-460; P = 630S

A = 0.5 wt% S = 100 ppm

A = 0.5 wt% S = 100 ppm P = 0.1 wt%

P = 0.1 wt% A = 0.5 wt% S = 100 ppm P = 0.05 wt%

Fluid Injection

Initial Waterflood, PV 4.07 4.01 3.98 3.67

Chemical Slug, PV 0.5 0.5 0.5 0.5

Extended Waterflood, PV 2.05 3.96 3.54 3.73

Recovery Data

IWF Recovery, %OOIP 50.52 49.09 49.64 49.26

Chemical Slug Recovery, %OOIP 0.42 0.37 0.21 0.36

EWF Recovery, %OOIP 0.29 14.28 10.05 4.70

Total Recovery, %OOIP 51.23 63.47 59.90 54.32

10 SPE 146782

= 148 = 156

(a) (b)

Fig. 1Photograph of Sessile Oil Drop on (a) Glass and (b) Quartz in Presence of Brine at Atmospheric Pressure and 30C.

Fig. 2Schematic of Coreflooding Apparatus.

SPE 146782 11

Carbon Number

0 5 10 15 20 25 30

Mo

lar

Pe

rce

nta

ge

(m

ol%

)

0

1

2

3

4

5

6

Dead Oil (C31+=20.53 mol%)

Fig. 3Carbon Number Distribution for Dead Oil.

Pressure (MPa)

6 9 12 15 18

Den

sity (

kg

/m3)

900

910

920

930

940

950

Vis

cosity (

mP

a s

)

0

100

200

300

400

500

Fig. 4Densities and Viscosities of Recombined Live Oil as a Function of Pressure at 30C.

Pressure (MPa)

0 2 4 6 8 10 12

Den

sity (

kg

/m3)

900

920

940

960

980

1000

Vis

cosity (

mP

a s

)

5000

6000

7000

8000

9000

10000

Live oil density

Live oil viscosity

12 SPE 146782

Na

2CO

3 Concentration (ppm)

0 1000 2000 3000 4000 5000 6000

Inte

rfacia

l T

ensio

n (

mN

/m)

0.1

1

10

Fig. 5Equilibrium IFTs of Crude Oil and Brine with Only Alkaline Additives (Varied Na2CO3 Concentrations) at 30C.

Steol CS-460 Concentration (ppm)

0 100 200 300 400 500

Inte

rfacia

l T

ensio

n (

mN

/m)

0.00

0.05

0.10

0.15

0.20

Fig. 6Equilibrium IFTs of Crude Oil and Brine with Alkaline and Surfactant Additives (0.5 wt% Na2CO3 + Steol CS-460 with Varied Concentrations) at 30C.

SPE 146782 13

Polymer in Brine Viscosities

Shear rate (s-1

)

1 10 100 1000

Vis

cosity

(mP

as

)

1

10

100

0.5 wt% 3630S

0.4 wt% 3630S

0.35 wt% 3630S

0.3 wt% 3630S

0.25 wt% 3630S

0.1 wt% 3630S

Fig. 7Viscosities of Polymer (Flopaam 3630S) in Brine at 30C.

Fluid Injected (PV)

0 3 6 9

Cum

ula

tive o

il R

ecovery

(%

OO

IP)

0

20

40

60

80

100

Diffe

rential P

ressure

s (

kP

a)

0

20

40

60

80

100

IWF

AS Slug

EWF

Differential Pressure

Fig. 8Oil Recovery and Differential Pressure vs. Pore Volume Injected for Linear Coreflood Run L1.

14 SPE 146782

Fluid Injected (PV)

0 3 6 9

Diffe

rential P

ressure

s (

kP

a)

0

100

200

300

400

500

Cum

ula

tive o

il R

ecovery

(%

OO

IP)

0

20

40

60

80

100

Differential Pressure

IWF

ASP Slug

EWF

Fig. 9Oil Recovery and Differential Pressure vs. Pore Volume Injected for Linear Coreflood Run L2.

Fluid Injected (PV)

0 3 6 9

Diffe

rential P

ressure

s (

kP

a)

0

100

200

300

400

500

Cum

ula

tive o

il R

ecovery

(%

OO

IP)

0

20

40

60

80

100

Differential Pressure

IWF

P Slug

EWF

Fig. 10Oil Recovery and Differential Pressure vs. Pore Volume Injected for Linear Coreflood Run L3.

SPE 146782 15

Fluid Injected (PV)

0 3 6 9

Diffe

rential P

ressure

s (

kP

a)

0

100

200

300

400

500

Cum

ula

tive o

il R

ecovery

(%

OO

IP)

0

20

40

60

80

100

Differential Pressure

IWF

ASP Slug

EWF

Fig. 11Oil Recovery and Differential Pressure vs. Pore Volume Injected for Linear Coreflood Run L4.

Pressure (MPa)

6 7 8 9 10 11 12 13 14

Den

sity (

kg

/m3)

900

910

920

930

940

950

Experimental_Density

Simulation_Density

Fig. 12Experimental and Simulated Density for Undersaturated Live Oil.

16 SPE 146782

Pressure (MPa)

6 7 8 9 10 11 12 13 14

Vis

co

sity (

mP

.s)

160

180

200

220

240

260

Experimental_Viscosity

Simulation_Viscosity

Fig. 13Experimental and Simulated Visosity for Undersaturated Live Oil.

Injected PV

0 2 4 6 8

Cu

mu

lative

Oil

Pro

du

ctio

n (

cm

3)

0

20

40

60

80

100

120

Experimental Data

Simulation Data

Fig. 14Experimental and Simulated Cumulative Oil Production (Run L2).

SPE 146782 17

Injected PV

0 2 4 6 8 10

Cu

mu

lative

Wa

ter

Pro

du

ctio

n (

cm

3)

0

200

400

600

800

1000

1200

1400

1600

1800

Experimental Data

Simulation Data

Fig. 15Experimental and Simulated Cumulative Water Production (Run L2).

Sw

0.0 0.2 0.4 0.6 0.8 1.0

Kro

w

0.0

0.2

0.4

0.6

0.8

1.0

Krw

0.0

0.2

0.4

0.6

0.8

1.0

Krw (waterflood) logNc = -5.9

Krow (waterflood) logNc = -5.9

Krw (Chemical flood) logNc = -3.0

Krow (chemical flood) logNc = -3.0

Krw (miscible flood) logNc = -1.1

Krow (miscible flood) logNc = -1.1

Fig. 16Water/Oil Relative Permeability Curves for Waterflood and Chemical Flood within Different Ranges of Capillary Number.

18 SPE 146782

Injected PV

0 1 2 3 4 5 6 7 8 9

Oil

Recovery

Facto

r (%

OO

IP)

0

20

40

60

80

100

ASP Slug 0.2 PV

ASP Slug 0.5 PV

ASP Slug 1.0 PV

Fig. 17Cumulative Oil Production at Different ASP Slug Sizes.

Injected PV

0 1 2 3 4 5 6 7 8 9

0

20

40

60

80

100

0.05 wt.% Polymer

0.1 wt.% Polymer

0.2 wt.% PolymerOil

Recovery

Facto

r (%

OO

IP)

Fig. 18Cumulative Oil Production at Different Polymer Concentrations.