A triblock copolymer for polymer flood in porous media

Krishna Panthi, Kishore K. Mohanty n

The University of Texas at Austin, United States

a r t i c l e i n f o

Article history:Received 12 December 2012Accepted 3 January 2014Available online 13 January 2014

Keywords:block copolymerscylindrical micellespolymer floodenhanced oil recoveryviscous oil recovery

a b s t r a c t

High molecular weight polymers are used in petroleum reservoir polymer floods to enhance oil recovery.The objective of this work is to evaluate small polymeric surfactants for their viscosifying capacity inreservoir brines and oil displacement ability. The phase behavior and viscosity of a triblock copolymer(P123) are studied as a function of brine salinity and temperature. Its flow through a porous rock and oildisplacement is evaluated and compared with that of a Newtonian fluid (glycerol) and a non-Newtonianfluid with a high molecular weight polymer (HPAM) of similar viscosity. P123 forms cylindrical micellesin brine to give high viscosity. The viscosity increases with salinity at a low salinity, but decreases at ahigher salinity. In the secondary mode, both the polymers (P123 and HPAM) and glycerol solutionsincrease the oil recovery significantly over the water flood. The oil recovery is similar for the threeviscous fluids. In the tertiary mode, none of the viscous fluids increased oil recovery over the waterfloodat typical field rates. Pressure drop during P123 flood is significantly lower than the pressure drop duringHPAM and glycerol floods of similar initial viscosity. Viscosity of the aqueous P123 solution decreaseswhen it is equilibrated with oil. Some of the cylindrical micelles are converted to spherical micelles in thepresence of solubilized oil. P123 is not as cost effective as HPAM because it is slightly more expensive andneeds a higher concentration for a similar viscosity.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

Since less than half of the original oil in a subterraneanpetroleum reservoir is produced through primary (depressuriza-tion) and second recovery (mostly water flooding) techniques, it isnecessary to utilize enhanced oil recovery (EOR) techniques toincrease the ultimate recovery of oil (Lake, 1989). Chemical,miscible gas, and thermal recovery techniques are being devel-oped to increase oil recovery. One of the promising techniqueswithin chemical methods is polymer flooding. Typically a highmolecular weight (about 10–20 million Dalton) polymer is mixedwith brine and injected (Sorbie, 1991). Polymer flooding increasesoil recovery by increasing the sweep efficiency and decreasingwater fractional flow because of favorable viscosity ratio betweenthe oil and the polymeric solution (Green and Willhite, 1998). It issuitable for viscous oils because waterflood sweep efficiency is lowdue to viscous fingering and permeability heterogeneity. Polymerflooding typically does not increase the microscopic displacementefficiency of light oils, though recent studies show that viscoelasticpolymeric solution may be able to improve the microscopicdisplacement efficiency over that of the waterflood (Delshadet al., 2008; Wang et al., 2010).

The typical polymers used for polymer flood are syntheticpolymer hydrolyzed polyacrylamide (HPAM) and biopolymerXanthan gum. There are temperature and salinity limits for the useof these polymers. Chemical, mechanical, and biological degradationsmust be avoided during the use of these polymers (Molloy et al.,2000; Levitt and Pope, 2008, 2010). The viscosity of the solutionincreases as the molecular weight increases at constant polymerconcentration. The molecular weight of the typical polymers used isin the range of 10–20 million Dalton (Donaldson et al., 1985). Thehydrodynamic radius of these polymers is of the order of micronsand they have difficulty flowing through low permeability porousrocks (e.g. 10 mD or lower).

It is possible to viscosify water with small molecules if theyassociate. Some cationic surfactants, such as hexadecyltrimetri-methylammonium bromide (CTAB) (Rehage and Hoffmann, 1991;Lin et al., 1994; Candau and Oda, 2001) and nonionic (Acharya andKunieda, 2003) surfactants, as well as some triblock polymers(Chu, 1995; Hamley, 2001; Nakashima and Bahadur, 2006; Aswalet al., 2007; Dreiss, 2007) can self-assemble into long, flexiblecylindrical micelles to impart useful viscoelastic properties. Unlikeordinary polymers, cylindrical micelles are in equilibrium withtheir monomers, and micellar chains can reversibly break andrecombine and are therefore called “living polymers” (Candauet al., 1985) or “equilibrium polymers”. It would be useful toidentify an assembled system which could break into smalleraggregates and pass through the pore throats of low permeabilityrocks as well as aggregate in pore bodies and give high viscosity.

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/petrol

Journal of Petroleum Science and Engineering

0920-4105/$ - see front matter & 2014 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.petrol.2014.01.001

n Correspondence to: Petroleum & Geosystems Engineering, CPE 4.168,The University of Texas at Austin, 200 E. Dean Keaton St., Austin, TX 78712,United States. Tel.: þ1 512 471 3077; fax: þ1 512 471 9605.

E-mail address: mohanty@mail.utexas.edu (K.K. Mohanty).

Journal of Petroleum Science and Engineering 114 (2014) 52–60

Triblock copolymers such as Pluronics (EOm–POn–EOm) aresmall molecules (MWo10,000); they are called “polymeric sur-factants” because of their amphiphilic nature. Ethoxy (EO) blocksare hydrophilic, whereas the propoxy (PO) blocks are hydrophobicand this hydrophobicity increases with temperature. The aqueoussolution of P123, a Pluronic, starts to become cloudy at about 50 1C(Ganguly et al., 2005; Chaibundit et al., 2008) and phase separatesat about 93 1C (Bharatya et al., 2007). Because of their amphiphilicnature triblock copolymers form micelles in water at certaintemperatures. In a salinity and temperature range, these micellesare cylindrical which leads to increased viscosity of the aqueoussolution. The goal of this work is to evaluate the applicability ofthese small block copolymers as a viscosifying agent for polymerfloods in porous rocks.

The aqueous phase behavior of a triblock copolymer is studiedin brines typical of oil fields. The viscosity behavior is observed asa function of brine salinity. The phase behavior of the equilibratedoil–brine-triblock copolymer is evaluated. The rheology of theaqueous solution is studied during flow in a porous rock. Then aviscous oil is displaced by this block copolymeric solution in aporous rock and this behavior is compared with those by thetraditional HPAM polymeric solution and a glycerol solution(Newtonian) of about the same viscosity. The next section outlinesthe methodology. The third section describes the results followedby the conclusions.

2. Experimental methodology

2.1. Materials

A tri-block copolymer P123 (PEO20–PPO70–PEO20) and gly-cerol (from Sigma Aldrich) were obtained at 99% purity. Hydro-lyzed polyacrylamide (HPAM3330) was obtained from SNF. Itsmolecular weight was 8�106 and it has 25–30% degree ofhydrolysis. Sodium chloride (NaCl), calcium chloride (CaCl2), andmagnesium chloride (MgCl2) were obtained from Fisher Scientific.Except glycerol, all the other chemicals were solid and wereprepared in distilled water. The base brine composition was setto 5 wt% NaCl, 1 wt% CaCl2, and 0.5 wt% MgCl2, typical of many oilreservoir brines. Oil was obtained from a petroleum reservoir andhad a viscosity of 103 cP at 25 1C.

2.2. Physical properties

Polymer P123 was added to the brine and its phase behaviorwas studied as a function of salinity and temperature. 2 ml oil wasmixed with 2 ml P123 solution in a thin, graduated, 5 ml borosi-licate glass pipette, which was flame sealed at the top. If not statedotherwise, the P123 solution refers to 1 wt% P123 in brine with5 wt% NaCl, 1 wt% CaCl2 and 0.5 wt% MgCl2. The sample wasequilibrated by shaking from time to time. The interfacial tension

(IFT) between oil and P123 solution was measured by in a Ramé-Hart Goniometer by using the pendent drop method. Viscosity ofaqueous solutions was measured with a Contraves low shearviscometer at 1 s�1 and with an ARG2 rheometer as a functionof shear rate from 1 to 100 s�1. Particle size within the aqueousphase was measured by a Delsa™ Nano C particle analyzer usingdynamic light scattering.

2.3. Core preparation

All 3 cores (Berea #1–#3) approximately 1 ft long and 1.5 in. indiameter were drilled, dried, and weighed. A vacuum pump wasused to evacuate air from the core and several pore volumes of CO2

were flushed to remove the air. Next, the cores were flooded withbrine containing 65,000 ppm total dissolved solids including50,000 ppm NaCl, 10,000 ppm CaCl2 and 5000 ppm MgCl2. Pres-sure data were recorded and brine permeability was calculated.The core #2 and #3 were then flooded with a viscous oil (103 cP)at a high pressure gradient (�100 psi) until no water wasproduced in the effluent. Table 1 describes the composition ofbrine and polymer solutions used and Table 2 lists the propertiesof all Berea sandstone cores. All coreflood experiments wereperformed at 25 1C.

2.4. Core flood

Seven different core floods were performed. One core flood wasconducted for P123 solution single phase flow in the absence ofoil. For each viscosifying agent (P123, HPAM, and Glycerol) corefloods were performed in similar cores, and under similar condi-tions. Three floods were conducted under secondary condition, i.e.,the core was at high oil saturation with connate brine when theviscous solution was injected. Three corefloods were conducted intertiary conditions, i.e., viscous solutions were injected after awater flood. Capillary end effect was minimal because of water-wet conditions, one-foot long cores, and high pressure drops.

2.4.1. Single phase flowThe Berea #1 core was injected with brine at the rate of

1 ft/day, 2 ft/day, 4 ft/day, and 8 ft/day and the pressure dropswere measured. After brine flow, P123 solution was injected at the

Nomenclature

CTAB hexadecyltrimetrimethylammonium bromideIFT interfacial tensionOOIP original oil in placeWOR water to oil ratioHPAM hydrolyzed polyacrylamideNaCl sodium chlorideCaCl2 calcium chlorideMgCl2 magnesium chloride

cP centipoiseCTAB hexadecyltrimetrimethylammonium bromidepv pore volumeSoi initial oil saturationWF water floodPF polymer floodRRF residual resistance factorSor residual oil saturationTHF tetrahydrofuran

Table 1Composition of brine and polymer solutions for core flood.

Chemical/viscosity Brine P123 HPAM Glycerol(wt%) (wt%) (wt%) (wt%)

Polymer 1 0.28 60NaCl 5 5 5 5CaCl2 1 1 1 1MgCl2 0.5 0.5 0.5 0.5Viscosity (cP) @ 10 s�1 1 11 11 11

K. Panthi, K.K. Mohanty / Journal of Petroleum Science and Engineering 114 (2014) 52–60 53

same rates (1–8 ft/day) followed by brine at the same flow rates tocalculate the residual resistance factor and in-situ polymerviscosity.

2.4.2. Secondary core floodAfter saturation of the Berea cores #2 and #3 with oil, they

were flooded in the secondary mode with different polymericsolutions (i.e., P123 and HPAM) or glycerol of equal viscosity fromthe bottom at the rate of 0.05 ml/min (which corresponds to aninterstitial velocity of 1 ft/day). About 2 pore volumes (pv) ofpolymer solution was injected followed by about 2pv of brine.After secondary polymer floods, these cores were cleaned by thereconditioning method (described later) and saturated with the oilagain. The effluents were collected in small glass vials for thedetermination of oil cut and viscosity.

2.4.3. Tertiary core floodCores (Berea #2 and #3) saturated with oil and connate water

was first flooded with about 2pv of brine. After water flooding,polymer solutions were injected in the tertiary mode to increaseoil recovery. First, the cores were flooded with about 3pv ofpolymer (P123, HPAM) or glycerol solution at the rate of 0.05 ml/min (or 1 ft/day). Then the flow rate was increased to 8 ft/day andone more pore volume of polymer solution was injected. This wasfollowed by 1pv brine at the same flow rate (8 ft/day) and then theflow rate was further doubled, i.e. increased to 16 ft/day, with oneor more pore volume of brine. The compositions of all the polymersolutions are presented in Table 1.

2.4.4. Reconditioning of coreThe Berea cores #2 and #3 were used many times with

cleaning after each polymer flood. The cores were injected with3–4pv of tetrahydrofuran (THF) to dissolve and remove aromaticcontents of the oil. When the effluent solution appeared clear, thecores were flooded with 3–4pv of chloroform to dissolve aliphaticpart of the oil which was followed by 3–4pv methanol to removeall solvents and brine. Then the cores were flooded with about10pv of brine and permeability was measured to make sure thatthe permeability was restored.

2.5. List of core flood experiments

Seven core flood experiments were performed as listed inTable 3. Viscosity of oil in all corefloods was 103 cP. Core flood#1 is the single phase flow for P123. Core flood #2 is P123 flood insecondary mode. Core floods #3 and #4 are secondary flood forHPAM and Glycerol, respectively. Core floods #5, #6, and #7 aretertiary floods for P123, HPAM, and glycerol, respectively. Coreflood #6 is also considered for water flood.

3. Results and discussion

3.1. Phase behavior of P123 in brine

The effects of block copolymer P123 and salinity on phasebehavior and viscosity are shown in Tables 4a and 4b. The brinecomposition for the base case has been assumed to be 5 wt% NaCl,1 wt% CaCl2, 0.5 wt% MgCl2 and 1 wt% P123. Table 4a shows theeffect of P123 concentration from 0.25 to 3 wt% on phase stabilityat the base case salinity. All solutions are clear up to 40 1C butthere is a temperature above which the aqueous solution of P123is not stable. This temperature decreases as the P123 concentra-tion increases. For example, 3 wt% P123 solution is hazy above60 1C, but 1 wt% P123 solution is hazy at 80 1C. These solutionshave a distinct phase separation at 91 1C. Viscosity is measured at25 1C and 10 s�1; it increases with the increasing concentration ofP123. Table 4b shows the effect of increasing the NaCl concentra-tion from 0 to 20 wt% with a fixed amount of block copolymer(1 wt% P123) and hardness (Caþ þ and Mgþ þ). As temperatureincreases, the amount of salinity that can be tolerated decreases.At 40 1C, 1 wt% P123, haziness sets in at about 17 wt% NaCl. At80 1C, 1 wt% P123, haziness sets in at about 4 wt% NaCl. The P123solution without any salt is stable at 91 1C, but the rest of thesolutions are phase separated at this temperature. As the salinityincreases, the viscosity at 25 1C increases up to the salt concentra-tion of about 6 wt% and then decreases.

3.2. Phase behavior of P123 with oil

The 1 wt% P123 was mixed in hard brine (1 wt% CaCl2 and0.5 wt% MgCl2) to which an additional amount of NaCl (0–20 wt%)was added. Equal amounts of viscous oil and P123 solution weremixed in pipettes, i.e., water–oil ratio (WOR)¼1. Fig. 1 shows thephase behavior as a function of the additional NaCl concentration(0–20 wt%) at 25 1C. The phase behavior appears to be of Winsortype II, i.e., water is solubilized into oil, at the salinities studied.Increasing amounts of water is solubilized in the oil phase withincreasing NaCl concentration beyond 6 wt% NaCl. This oil phase isnot a water-in-oil macroemulsion; it is a water-in-oil microemul-sion. The viscosity of this phase was about 3% lower than the pure

Table 2Core properties.

Core # Mineral Length Diameter Brine perm Porosity Pore volume(cm) (cm) (md) (%) (ml)

Berea 1 Sandstone 29.33 3.74 394.7 22 70Berea 2 Sandstone 29.4 3.72 343.3 21 66Berea 3 Sandstone 29.23 3.75 216.2 20.7 67

Table 3Core flood experiments.

Core flood # Core # Polymer Polymer–brine composition Fluid sequence Soi(%)

1 Berea 1 P123 1% P123, 5% NaCl 1% CaCl2, 0.5% MgCl2 2.17pv Brine 2.88pv P123 2.23pv Brine2 Berea 2 P123 1% P123, 5% NaCl 1% CaCl2, 0.5% MgCl2 2pv P123 2.1pv Brine 72.73 Berea 2 HPAM 0.28% HPAM, 5% NaCl 1% CaCl2, 0.5% MgCl2 2pv HPAM 2pv Brine 754 Berea 2 Glycerol 60% Glycerol, 5% NaCl 1% CaCl2, 0.5% MgCl2 2pv Glycerol 2pv Brine 665 Berea 3 P123 1% P123, 5% NaCl, 1% CaCl2, 0.5% MgCl2 3.25pv Brine 4.35pv P123 2.77pv Brine 726 Berea 2 HPAM 0.28% HPAM, 5% NaCl 1% CaCl2, 0.5% MgCl2 3pv Brine 4.6pv HPAM 1pv Brine 65.27 Berea 2 Glycerol 60% Glycerol, 5% NaCl 1% CaCl2, 0.5% MgCl2 3pv Brine 5pv Glycerol 1pv Brine 72

K. Panthi, K.K. Mohanty / Journal of Petroleum Science and Engineering 114 (2014) 52–6054

oil viscosity. Water-in-oil macroemulsions usually have a higherviscosity than the oil viscosity.

To measure the IFT between oil and P123 solution, a 1 wt% P123solution (having 5 wt% NaCl, 1 wt% CaCl2 and 0.5 wt% MgCl2) andoil were mixed in the ratio of 1:1, kept in an oven for 2 days at40 1C and then equilibrated at the room temperature. The equili-brated oil was taken in an inverted needle and immersed in theequilibrated P123 solution and pressurized in the syringe to formdrop of oil in P123 solution. Then the IFT was measured by theRamé-Hart contact angle goniometer using the pendent dropmethod. Thirty different values were obtained by changing thedrop and drop size; the mean IFT value was found to be 6.1 dyne/cm (compared to 21.9 dyne/cm for oil and brine). Thus the IFT isnot ultralow (o0.001 dyne/cm) with this polymeric surfactant atthese conditions. The PO units should be reduced and EO unitsshould be increased to make it less oil soluble.

3.3. Viscosity

The dynamic viscosity measurements were performed on P123with an ARG2 rheometer. Fig. 2a shows the viscosity of the basecase brine (5 wt% NaCl, 1 wt% CaCl2, 0.5 wt% MgCl2) at differentconcentrations of P123. The P123 solutions are shear thinning andas the block copolymer concentration increases, the viscosityincreases. Fig. 2b and Table 4b show the viscosity of 1 wt% P123for different concentrations of NaCl (with 1 wt% CaCl2, 0.5 wt%MgCl2). As the NaCl concentration increases, the solution viscosityincreases and then it decreases. The highest viscosity occurs at4 wt%, but the viscosity was maintained up to 8 wt% NaClconcentration. Fig. 2c shows the viscosity of 1 wt% P123 in thebase case brine as a function of shear rate along with the viscosityof HPAM and glycerol solutions. The HPAM and glycerol concen-trations were chosen such that the viscosities of all three solutionswere about the same at 10 s�1 shear rate. The polymer solutionsare shear thinning, but P123 was more shear thinning than HPAM.The shear rate in a Berea core at 1 ft/day interstitial velocity flow is

about 10 s�1. The viscosity at 10 s�1 is referred to as the viscositylater in this paper if the shear rate is not specified.

The viscosity data of P123 solution were then fitted to a Carreaunon-Newtonian viscosity model to determine a mathematicalrelationship between viscosity (m) and shear rate (γ), as shownin Fig. 2d. The Carreau model for the block copolymer is

m¼ m1þðmo�m1Þ½1þðKγÞ2�ðn�1Þ=2

where n (rate index)¼0.5905, K (consistency index)¼0.88475,mo¼0.02873 Pa s, m1¼2.56E�09 Pa s. The viscosity of P123 solu-tion equilibrated with the oil is shown in Fig. 2e. The data showsthat the viscosity of this solution is almost equal to that ofthe brine. The viscosity of an aqueous solution with the sameamount of P123 and salinity but not equilibrated with oil wasabout 11 cP. This implies that the non-equilibrated aqueous solu-tions of P123 have large-scale structures (perhaps cylindricalmicelles) whereas the solutions equilibrated with oil do not havelarge-scale structures.

3.4. Particle size

The size of micellar aggregates in the base case solution with1 wt% P123 was measured in a Delsa Nano Particle Analyzer(Beckman Coulter, Inc.) and the average particle size was obtainedto be 146.6 nm with a standard deviation of 95.6 nm and theintensity distribution is shown in Fig. 3. The aggregate sizedistribution of P123 solution equilibrated with oil was obtainedto be 143.2 nm with a standard deviation of 68.8 nm, shown inFig. 3. Furthermore, the size distributions of the effluents duringcore floods of P123 measured and the values were similar to thoseof the equilibrated sample. The small change in the size distribu-tion cannot explain the large viscosity change. It is hypothesizedthat some of the long cylindrical micelles are converted tospherical micelles because of solubilization of oil. The shape ofthe micelles needs to be measured to validate this hypothesis.

Table 4aStability and viscosity of different concentrations of P123 at a fixed salinity.

Experiment #I #II #III #IV #V #VI

P123 (wt%) 0.25 0.5 0.75 1 2 3NaCl (wt%) 5 5 5 5 5 5CaCl2 (wt%) 1 1 1 1 1 1MgCl2 (wt%) 0.5 0.5 0.5 0.5 0.5 0.5Viscosity (cP) (10 s�1) 1.3 2.8 5.6 11 44.7 65.7Aq. st. at 25 1C Clear Clear Clear Clear Clear ClearAq. st. at 40 1C Clear Clear Clear Clear Clear ClearAq. st. at 60 1C Clear Clear Clear Clear Hazy HazyAq. st. at 80 1C Clear Clear Clear Hazy Hazy HazyAq. st. at 91 1C Phase separate Phase separate Phase separate Phase separate Phase separate Phase separate

Table 4bStability and viscosity at different NaCl concentrations at 1 wt% P123.

Expt. #I #II #III #IV #V #VI #VII #VII #IX #X

P123 (wt%) 1 1 1 1 1 1 1 1 1 1NaCl (wt%) 2 4 6 8 10 12 14 17 20CaCl2 (wt%) 1 1 1 1 1 1 1 1 1MgCl2 (wt%) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5Viscosiy (cP) (10 s�1) 1.2 2.3 11.0 10.5 11.0 6.8 3.1 1.6 1.6 2.2Aq. st. (25 1C) Clear Clear Clear Clear Clear Clear Clear Clear Cloudy CloudyAq. st. (40 1C) Clear Clear Clear Clear Clear Clear Clear Clear Cloudy CloudyAq. st. (60 1C) Clear Clear Clear Clear Clear Hazy Cloudy Cloudy Cloudy CloudyAq. st. (80 1C) Clear Clear Hazy Hazy Cloudy Cloudy Cloudy Cloudy Cloudy CloudyAq. st. (91 1C) Clear Phase sep. Phase sep. Phase sep. Phase sep. Phase sep. Phase sep. Phase sep. Phase sep. Phase sep.

Expt. stands for experiment; Aq. st. stands for aqueous stability and phase sep. stands for phase separation.

K. Panthi, K.K. Mohanty / Journal of Petroleum Science and Engineering 114 (2014) 52–60 55

3.5. Single phase flow in porous media

The P123 solution was injected into a brine-saturated core atdifferent flow rates and the pressure drop was measured. Thepressure drop is shown in Fig. 4a. As the flow rate increases, sodoes the pressure drop, but at a rate lower than the flow rateincrease. The apparent viscosity for porous media flow wascalculated from Darcy0s law and is shown in Fig. 4c. The apparentviscosity in the porous medium was modeled using the para-meters derived from Carreau correlation for rheometer data andan effective shear rate proposed by Cannella et al. (1988), i.e.,

γapp ¼ C3nþ14n

� � nn� 1 uffiffiffiffiffiffiffi

Kϕp

where n is the rate index derived from rheometer viscosity(Fig. 2d) and C is the fitting parameter calculated to be 2.4 fromcoreflood pressure drop data. Fig. 4b shows a plot of apparentviscosity from the pressure drop data and measured viscosity fromthe rheometer. The two viscosities match at higher shear rates, butnot at the low shear rates. This discrepancy can be due to theinaccuracy of pressure measurement at low flow rates.

The polymer flow was followed by brine injection and thepressure drop is shown in Fig. 4b. Pressure drop falls with brineinjection and settles at a value slightly higher than the initialpressure drop with brine flow (before polymer flow). A residualresistance factor (RRF) is calculated from the brine pressure dropsbefore and after the polymer solution. RRF values are shown inFig. 4c at various shear rates (corresponding to flow rates 2 ft/day,4 ft/day and 8 ft/day).

3.6. Secondary oil recovery

The results of secondary polymer floods and water flood aretabulated in Table 5 and shown in Fig. 5. The coreflood numbersare indicated on the figure legends. For the secondary P123 flood(Core flood #2), the initial oil saturation was about 66.7%. P123flood at 1 ft/day rate reduced the oil saturation to 38.1% in 2pv

followed by 2pv of brine injected. Oil bank broke through at about0.5pv injected. Most of the oil was produced by 1pv. Cumulativeoil recovery for the P123 followed by brine flood was 47.6% of theoriginal oil in place (OOIP). The pressure drop first decreased andthen rose to about 7 psi at the end of the polymer injection andthen dropped to about 4 psi at the end of brine injection. ForHPAM flood (Core flood #3), the initial oil saturation was about75%. HPAM flood reduced the oil saturation to 42.2% at the end ofHPAM injection followed by the brine injection. Oil bank break-through was similar to P123 flood, but the pressure drop duringHPAM injection was about 14 psi and that of brine injection wasabout 7 psi. Cumulative oil recovery for the HPAM flood followedby brine flood was 43.6% OOIP. For glycerol (Core flood #4),the initial oil saturation was about 65.9%. Glycerol flood reducedthe oil saturation to 25.8% at the end of glycerol followed bybrine injection. Oil bank breakthrough was similar to P123 andHPAM floods, but the pressure drop during glycerol injection wasabout 25 psi and that of brine injection was about 8 psi. Cumula-tive oil recovery for the glycerol followed by brine flood was48% OOIP.

For the secondary water flood (Core flood #6), the initial oilsaturation was about 65.2%. Waterflood reduced the oil saturation to47.9% and 25.8% oil was recovered. The water flood was not veryefficient because of the adverse viscosity ratio (103) between oil andbrine. The oil cut decreased to a very low value soon after the waterbreakthrough, as shown in Fig. 5a. For the two polymers and theglycerol, the oil cut decreased more slowly. The oil recovery duringP123 flood and glycerol flood were almost the same, about 48% OOIP.The oil recovery during HPAM flood is 43.6%. The polymers andglycerol reduce the viscosity ratio to approximately (103/11) andimprove the oil recovery. The oil saturations during these floods areshown in Fig. 5b. The residual oil saturation during the water floodwas the highest at 47.9%. HPAM reduced the oil saturation to 42.2%.P123 and glycerol reduced the oil saturation to 38.1% and 34.2%,respectively. Glycerol solution maintains a high viscosity at high shearrates (because of its Newtonian viscosity) and thus displaces the oilmost efficiently. The polymers are shear thinning and thus less viscousat higher shear rates than glycerol and are less effective.

Fig. 1. Phase behavior of P123 and oil at 25 1C ((A) scans with 0–20 wt% NaCl and (B) phase behavior sample of P123 for IFT measurement).

K. Panthi, K.K. Mohanty / Journal of Petroleum Science and Engineering 114 (2014) 52–6056

The pressure drop data for all the polymers flood and brinewater flooding are presented in Fig. 5c. During the polymer/glycerol floods, the pressure drop decreased and then increasedwithin a factor of 2 of the initial pressure drop. For the water flood,the pressure drop gradually decreased with the recovery of oil by afactor of 3. This indicates that the flood was more unstable in caseof the water flood. The pressure drop was the highest for theglycerol flood and the lowest for the P123 flood. The low pressuredrop during P123 flood may be due to the lower viscosity of theP123 solution as it equilibrates with the oil. The core was reusedfor multiple floods; the permeability was not the same at the startof each flood.

3.7. Tertiary oil recovery

The results of the tertiary core floods of P123, HPAM, andglycerol are tabulated in Table 6 and shown in Figs. 6–8. Before the

tertiary flood of P123 (Core flood #5), water flood recovered about38% OOIP. After waterflood, 2.4pv of P123 was injected at 1 ft/daywhich increased the oil recovery only slightly to 39.5%. Then theflow rate was increased to 8 ft/day which further increased therecovery of 45.4%. After injection of one pore volume P123 at 8 ft/day brine was injected at the same flow rate which furtherincreased the recovery to 51%. At last, the flow rate was furtherincreased to 16 ft/day which further increased the net oil recoveryof 58% OOIP. The same procedure was repeated for both HPAM andglycerol. For HPAM (Core flood #6), water recovered 25.8% OOIPand HPAM increased the recovery to 28% at 1 ft/day rate whereasat 8 ft/day flow rate the recovery was increased to 39%. After thatno more oil was produced, though the flow rate was increased to16 ft/day. Similarly, for glycerol (Core flood #7), water recovered38% OOIP and glycerol flooding increased the recovery to 39% at1 ft/day rate whereas at 8 ft/day flow rate the recovery wasincreased to 45%. After that no more oil was produced, though

Fig. 2. (a) Viscosity of 0.25–3 wt% P123 at 25 1C at the base salinity. (b) Viscosity of 1 wt% P123 at 0–20 wt% NaCl at 25 1C. (c) Viscosity of P123, HPAM, and glycerol solutionsat the base salinity 25 1C. (d) Viscosity from rheometer and Carreau model fit. (e) Viscosity of aqueous P123 solution equilibrated with oil at 25 1C.

K. Panthi, K.K. Mohanty / Journal of Petroleum Science and Engineering 114 (2014) 52–60 57

the flow rate was further increased to 16 ft/day. The pore volumesof polymer injected are high, but they ensure that residualsaturations are established. Figs. 6 and 7 show that tertiary oilrecovery at typical field rates (1 ft/day) is minimal for all the threeviscosifiers, but the oil recovery can be increased if high flow ratesare imposed.

The pressure drops during the tertiary polymer floods are shown inFig. 8. The pressure drop during the P123 flood is relatively low incomparison to the pressure drop during HPAM and glycerol floods. At1 ft/day, the pressure drop for P123 is about 5 psi whereasthe pressure drops for HPAM and glycerol are about 18 psi, and25 psi, respectively. At 8 ft/day, the pressure drop of P123 is about

60 psi where that of the glycerol is about 100 psi and HPAM is about180 psi. Again, the low pressure drop during P123 flood may be due tothe lower viscosity of the P123 solution as it equilibrates with the oil.

3.8. Effluent brine viscosity

The injected P123 solution had a viscosity of 10 cP at 10 s�1.The viscosity of the effluent P123 solution was measured to beabout 1.5 cP at 10 s�1. This viscosity is similar to the viscosity ofthe P123 solution in equilibrium with oil. The particle sizedistribution was measured and was similar to that of the solutionequilibrated with oil. This change in viscosity may be due toconversion of some large cylindrical micelles into sphericalmicelles.

Fig. 3. Particle size distribution of the aqueous P123 solution and aqueous P123 inequilibrated oil at 25 1C.

Fig. 4. (a) Pressure drop of brine flood and P123 flood in single phase flow experiment (ab¼1 ft/day, bc¼2 ft/day, cd¼4 ft/day and de¼8 ft/day). (b) Viscosity fromrheometer and Carreau fit from single phase flow of P123. (c) In situ viscosity of P123 and RRF of brine during P123 flooding at 25 1C.

Table 5Secondary core flooding result.

Core flood#

Injectionfluid

Initial oilsaturation

Residual oilsaturation

Recovery

(%) (%) (% OOIP)

2 P123 72.73 38.1 47.63 HPAM 75 42.2 43.64 Glycerol 65.9 34.2 486 Brine 65.2 47.9 25.8

K. Panthi, K.K. Mohanty / Journal of Petroleum Science and Engineering 114 (2014) 52–6058

Fig. 5. (a) Cumulative oil recovery and water cut during secondary polymer floods. (b) Residual oil saturation for polymers during secondary floods. (c) Pressure drop duringsecondary floods.

Table 6Tertiary core flood result.

Core flood # Polymer Soi Recovery for WF Sor after WF Cum. recovery for WFþPF Sor after PF

5 P123 66.7 38 44.62 58 16.46 HPAM 65.2 25.8 47.95 39 397 Glycerol 72 37.9 38.6 45 33.5

Fig. 6. Oil saturation during water flood and tertiary polymer floods (ab: 1 ft/daywater flood, bc: 1 ft/day polymer flood, cd: 8 ft/day polymer flood, de: 8 ft/daybrine flood, and ef: 16 ft/day brine flood).

Fig. 7. Oil cut during water flood and tertiary polymer floods (ab: 1 ft/day waterflood, bc: 1 ft/day polymer flood, cd: 8 ft/day polymer flood, de: 8 ft/day brineflood, and ef: 16 ft/day brine flood).

K. Panthi, K.K. Mohanty / Journal of Petroleum Science and Engineering 114 (2014) 52–60 59

4. Conclusions

The phase behavior of a block copolymer, P123, solution wasstudied as a function of salinity and temperature, typical ofpetroleum reservoirs. Its flow through a Berea rock and displace-ment of a viscous oil were also studied. Its behavior was comparedwith that of a Newtonian fluid (glycerol) and a non-Newtonian(HPAM) fluid of similar viscosity.

Aqueous P123 solutions are unstable at high temperatures andhigh salinity. P123 forms cylindrical micelles in brine resulting inincreased viscosity. The viscosity increases with salinity at a lowsalinity, but decreases at a high salinity. In the secondary mode,both the polymers (P123 and HPAM) and glycerol solutionsincrease oil recovery significantly over the water flood. The oilrecovery is similar for the three viscous fluids. In the tertiarymode, none of the viscous fluids increased oil recovery over thewaterflood at 1 ft/day (a typical field rate). Oil recovery wasinduced at higher flow rates and pressure gradients greater than50 psi/ft. Pressure drop during P123 flood is significantly lowerthan the pressure drop during HPAM and glycerol floods of similarinitial viscosity. Viscosity of the aqueous P123 solution decreaseswhen it is equilibrated with oil. Some of the cylindrical micellesare possibly converted to spherical micelles in the presence ofsolubilized oil.

For field application of a viscosifying agent, the cost of thechemical is important. The cost of P123 is about $2.37 per lb forlarge amounts. This price is slightly higher than the price forHPAM which is under $2 per lb. The concentration needed forP123 is also higher than HPAM for the same viscosity. Thus thiscopolymer is not as cost-effective as HPAM for high permeabilityreservoirs. There are many block copolymers in the family ofPluronics; other copolymers can be evaluated. P123 has somebenefits in terms of a lower pressure gradient. The near-wellboreinjectivity is an issue for polymer floods in general; it should bestudied for P123. The near-wellbore region is often well swept foroil, in which case the copolymer viscosity may not be reduced.HPAM shows shear thickening at shear rates typical of near-wellbore regions due to viscoelasticity. Rheology of P123 in porousmedia at high shear rates should also be investigated.

Acknowledgments

The authors would like to thank the industrial affiliates of theChemical Enhanced Oil Recovery project at the University of Texasat Austin for the financial support.

References

Acharya, D.P., Kunieda, H., 2003. Formation of viscoelastic wormlike micellarsolutions in mixed nonionic surfactant system. J. Phys. Chem. B 107,10168–10175.

Aswal, V.K., Wagh, A.G., Kammel, M., 2007. Formation of rodlike block copolymermicelle in aqueous salt solution. J. Phys.: Condens. Matter 19, 116101.

Bharatya, B., Guo, C., Ma, J.H., Hassan, P.A., Bahadur, P., 2007. Aggregation andclouding behavior of aqueous solution of EO–PO block copolymer in presenceof n-alkanols. Eur. Polym. J. 43, 1883.

Candau, S.J., Hirsch, E., Zana, R., 1985. Light scattering investigations of the behaviorof semidilute aqueous micellar solutions of cetyltrimethylammonium bromide:analogy with semidilute polymer solutions. J. Colloid Interface Sci. 105,521–528.

Candau, S.J., Oda, R., 2001. Linear viscoelasticity of salt-free wormlike micellarsolutions. Colloids Surf. 183, 5–14.

Cannella, W.J., Huh, C., Seright, R.S., 1988. Prediction of xanthan rheology in porousmedia. In: Paper SPE 18089, Annual Technical Conference and Exhibition of theSociety of Petroleum Engineers, Houston, Texas, Engineers. October 2-5.2-5.Houston, Texas.

Chaibundit, C., Ricardo, Na´gila, M.P.S., Costa Fla´via de, M.L.L., Wong, M.G.P.,Hermida-Merino, D., Rodriguez-Perez, J., Hamley, I.W., Yeats, S.G., Booth, C.,2008. Effect of ethanol on the micellization and gelation of Pluronic P123.Langmuir 24, 12260–12266.

Chu, B., 1995. Structure and dynamics of block copolymer colloids. Langmuir 11,414–421.

Delshad, M., Kim, D.H., Magbagbeola, O.A., Huh, C., Pope, G.A., Tarahhom. F., 2008.Mechanistic interpretation and utilization of viscoelastic behavior of polymersolutions for improved polymer flood efficiency. In: Paper SPE 113620, SPE/DOEImproved Oil Recovery Symposium. April 19–23. Tulsa, Oklahoma.

Donaldson, E.C., Chilingarian, G.V., Yen, T.F., 1985. Enhanced Oil Recovery I.Fundamentals and Analyses, Developments in Petroleum Science. 17A. ElsevierScientific Publishing, Co., New York City, pp. 318–330

Dreiss, C.A., 2007. Wormlike micelles: where do we stand? Recent developments,linear rheology and scattering techniques. Soft Matter 3, 956–970.

Ganguly, R., Aswal, V.K., Hassan, P.A., Gopalakrishnan, I.K., Yakhmi, J.V., 2005.Sodium chloride and ethanol induced sphere to rod transition of triblockcopolymer micelles. J. Phys. Chem. B 109, 5653–5658.

Green, D.W., Willhite, G.P., 1998. Enhanced Oil Recovery, Society of PetroleumEngineers. Richardson, TX

Hamley, I.W., 2001. Structrure and flow behavior of block copolymers. J. Phys.:Condens. Matter 13, R643–R671.

Lake, L.W., 1989. Enhanced Oil Recovery. Prentice Hall, New JerseyLevitt, D.B., Pope, G.A., 2010. Chemical degradation of polyacrylamide polymers

under alkaline conditions. In: Paper SPE 129879, Improved Oil RecoverySymposium. April 24–28. Tulsa, Oklahoma.

Levitt, D., Pope, G.A., 2008. Selection and screen of polymers for enhanced-oilrecovery. In: Paper SPE 113845, SPE/DOE Symposium on Improved Oil Recovery.April 19–23. Tulsa, Oklahoma.

Lin, Z., Cai, J.J., Scriven, L.E., Davis, H.T., 1994. Spherical-to-wormlike micelletransition in CTAB solutions. J. Phys. Chem. 98, 5984–5993.

Molloy, P.J., Smith, M.J., Cowling, M.J., 2000. The effects of salinity and temperatureon the behavior of polyacrylamide gels. Mater. Des. 21 (3), 169–174.

Nakashima, K., Bahadur, P., 2006. Aggregation of water soluble block copolymers inaqueous solutions: recent trends. Adv. Colloid Interface Sci. 123–126, 75–96.

Rehage, H., Hoffmann, H., 1991. Viscoelastic surfactant solutions: model researchfor rheological system. Mol. Phys. 74, 933–973.

Sorbie, K.S., 1991. Polymer-Improved Oil Recovery, Blackie Technology and Engi-neering. Edinburgh, UK

Wang, D., Xia, H., Yang, S., Wang, G., 2010. The influence of visco-elasticity on microforces and displacement efficiency in pores, cores, and in the field. Paper SPE127453, SPE EOR Conference at Oil and Gas West Asia. April 11–13. Muscat,Oman.

Fig. 8. Pressure drop during water flood and tertiary polymer floods (ab: 1 ft/daywater flood, bc:1 ft/day polymer flood, cd: 8 ft/day polymer flood, de: 8 ft/day brineflood, and ef: 16 ft/day brine flood).

K. Panthi, K.K. Mohanty / Journal of Petroleum Science and Engineering 114 (2014) 52–6060