Coinjection of Seawater and Produced Water to Improve OilRecovery from Fractured North Sea Chalk Oil Reservoirs

Tina Puntervold,* Skule Strand, and Tor Austad

UniVersity of StaVanger, 4036 StaVanger, Norway

ReceiVed NoVember 26, 2008. ReVised Manuscript ReceiVed February 20, 2009

Carbonate oil reservoirs are often fractured with moderate water-wet conditions, which prevent spontaneousimbibition of water into the matrix blocks. Enhanced oil recovery by water flooding is therefore seldomsuccessful, and the average oil recovery from carbonates is usually much less than 30%. Hence, the improvedoil recovery potential is very high in these types of reservoirs. Recent studies on chalk cores from the NorthSea formations have shown that seawater is able to change the wettability toward a more water-wet conditionat high temperatures, >100 °C. The successful injection of seawater into the Ekofisk formation, where the oilrecovery is now estimated to reach 50%, is a good example. Seawater contains favorable concentrations of thepotential determining ions Ca2+, Mg2+, and SO4

2- that are active in the displacement of strongly adsorbedcarboxylic material from the chalk surface. The initial formation water will partly mix with seawater. Therefore,the amount and composition of the produced water will vary with time. Due to environmental reasons, theproduced water should be reinjected together with seawater into the chalk formation. The question that we askin this paper is: “Will mixtures of seawater and produced water displace the oil in a similarly good manner aspure seawater?” This paper contains: (1) calculations of the compatibility of different mixtures of SSW (syntheticseawater) and artificial PW (artificial produced water) regarding precipitation of CaSO4, SrSO4, and BaSO4 atvarious temperatures. In most cases the fluids were compatible when the PW was diluted at least 4 times withSSW. (2) Experimental work on oil recovery from chalk cores of moderate water-wetness using various mixturesof SSW and PW. Both spontaneous imbibition and viscous flooding were performed. At T > 100 °C, the oilrecovery by using PW:SSW mixtures in ratios ranging from 2:1 to 1:8 was significantly higher than by usingpure PW in a spontaneous imbibition process. In a viscous flood, SSW appeared to be much more efficientthan PW to displace the oil, and high oil recovery values were reached. This work is environmentally promising,because the conclusion drawn is that it should be possible to coinject produced water and seawater into aNorth Sea chalk oil reservoir, such as Valhall, without losing the good enhanced oil recovery properties ofseawater.

Introduction

Chalk is a carbonate rock, and carbonate reservoirs are usuallymoderately water-wet to oil-wet and highly fractured, whichmeans that the injected brine will follow the fractures from theinjector to the producer and will only displace the oil containedin the fractures. The orientation of the fractures relative to thefluid flow direction also appears to have an impact on oilrecovery.1 Therefore, the oil recovery from carbonates is usuallylow, well below 30% on average. In that way, the naturallyfractured Ekofisk chalk field in the North Sea appears to be anexception because the oil recovery is expected to exceed 50%by injection of seawater. The wetting condition of the Ekofiskfield varies from preferentially water-wet in the Tor formation,to moderately water-wet in the Lower Ekofisk, and to neutralto preferentially oil-wet in the Upper Ekofisk formation.2

Through a series of papers, we have recently documented thatseawater acts as an EOR-fluid in chalk formations by increasingthe water-wetness and improving spontaneous imbibition of

water into the matrix blocks.3-6 In that sense, injection ofseawater into high temperature chalk fields can be regarded asa tertiary oil recovery technique.

The chemical mechanism behind the wettability alterationcaused by seawater has been studied in detail.4,5 It was confirmedthat seawater contained potential determining ions Ca2+, Mg2+,and SO4

2-, which were able to displace strongly adsorbedcarboxylic material from the chalk surface at high temperatures,T > 90-100 °C. This effect increased drastically as thetemperature increased beyond 100 °C, due to increased adsorp-tion of SO4

2- onto the chalk surface.6 The presence of SO42-

is essential for the wettability alteration reactions to take place.In addition, the divalent cations Ca2+ and/or Mg2+ must bepresent. In seawater, all these ions are present in properconcentration ratios, that is, [Mg2+] ∼ 2 [SO4

2-] ∼ 4 [Ca2+].The formation water, FW, and produced water, PW, containsignificant amounts of Ca2+ and Mg2+, but they only containvery small amounts of SO4

2-, which is insufficient to promotea wettability modification toward a more water-wet condition.

* Corresponding author. E-mail: tina.puntervold@uis.no; phone: +4751832213;fax: +4751831750.

(1) Shedid, S. A. J. Pet. Sci. Eng. 2006, 50, 285–292.(2) Torsaeter, O. , An experimental study of water imbibition in chalk

from the Ekofisk field, In Paper SPE12688 presented at the SPE/DOEFourth Symposium on Enhanced Oil RecoVery, Tulsa, OK, USA, April 15-18, 1984; 1984.

(3) Zhang, P.; Austad, T. Colloids Surf., A 2006, 279, 179–187.(4) Zhang, P.; Tweheyo, M. T.; Austad, T. Colloids Surf., A 2007, 301,

199–208.(5) Zhang, P.; Tweheyo, M. T.; Austad, T. Energy Fuels 2006, 20, 2056–

2062.(6) Strand, S.; Høgnesen, E. J.; Austad, T. Colloids Surf., A 2006, 275,

1–10.

Energy & Fuels 2009, 23, 2527–2536 2527

10.1021/ef801023u CCC: $40.75 2009 American Chemical SocietyPublished on Web 03/20/2009

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u

Whether the fluid flow in a fractured chalk reservoir is relatedto spontaneous imbibition of water from the fractures into thematrix blocks or related to viscous flooding, the FW will bebanked up in front of the injected water. Thus, the producedwater at an early stage in the water flood will mainly be pureFW.7 Because of ion exchange between the injected seawaterand the chalk surface, that is, substitution of Ca2+ by Mg2+

and adsorption of SO42- on the surface,4,6 the composition of

the injected seawater will change in the flooding process. Theincrease in Ca2+sand decrease in Mg2+sconcentration may alsoresult in precipitation of CaSO4 in the formation at hightemperatures, as indicated in a previous simulation study.8

Consequently, the change in PW composition during a waterflood using seawater is rather complicated due to ion exchange,precipitation, and mixing with the FW. Furthermore, the injectedseawater may also release/dissolve traces of Ba2+ and Sr2+

present in the chalk formation as carbonates. The concentrationof these ions in the PW will normally not increase beyond thevalue in the initial FW because of precipitation of sulfate saltsin the reservoir. It is important to check the compatibility ofmixtures of SSW and PW prior to injection with regards topossible precipitation of CaSO4, BaSO4, and SrSO4.

9

Normally, the water-oil ratio increases during a water flood,and due to environmental reasons, the PW should be reinjectedinto the formation rather than deposed to the sea. It is well-known that PW contains significant amounts of low molecularweight aromatic material, which is known to be carcinogenic,and it can be bioaccumulated in living organisms. Evenextensive extraction of PW with a supercritical fluid (propane+ butane) will leave a significant amount of aromatic materialin the water phase. Thus, the easiest way to prevent pollutingthe sea is to reinject the PW. The question is: “Is it possible tore-inject the PW together with seawater and still maintain anoil recovery similar to what was obtained by seawater?” Thescope of this work was to look at oil recovery from moderatelywater-wet chalk cores using different mixtures of PW and SSWas injection fluids at temperatures ranging from 50 to 130 °C.The oil recovery was studied both by spontaneous imbibitionand viscous flooding. Some results from the compatibility studyof PW and SSW regarding precipitation of sulfate salts8 havealso been included.

Experimental Section

Rock. The porous material used is outcrop Stevns Klint chalkfrom a quarry nearby Copenhagen, Denmark. This chalk consistsof 98% pure biogenic CaCO3 with high matrix porosity (45-50%)and low matrix permeability (1-2 mD). It has a reactive largesurface area of about 2 m2/g.10,11 The properties of the Stevns Klintchalk are quite similar to those observed for the North Sea chalkoil reservoirs.

Oils. Two oils with different acid number (AN) and base number(BN), termed oil A (AN ) 1.8 mg KOH/g oil, BN ) 0.52 mgKOH/g oil) and oil B (AN ) 0.11 mg KOH/g oil, BN ) 0.10 mgKOH/g oil), were mixed in different ratios to create oils withdifferent AN and BN. Oil B was made from oil A by removingpolar components using silica as described previously.12 Densityand viscosity at 20 °C were quite similar for both oils A and B,

determined to 0.806 g/cm3 and 2.5 cP for oil A and 0.799 g/cm3

and 2.5 cP for oil B. Hence, since all oils used originate from oilsA and B, the viscosities of the different oils are believed to besimilar. The AN and BN of the oils used in the different tests canbe found in Table 1.

Brines. Synthetic seawater (SSW), artificial FW, and PW relatedto the Ekofisk field were used for the modeling studies, Table 2.The compositions of the different mixtures of PW and SSW, 1:1,1:2, 1:4, and 1:8, respectively, were calculated and are given inTable 2. In the experimental studies, artificial PW from the Valhallfield, PW, and SSW were mixed in different volume ratios, 2:1,1:1, 1:2, and 1:8 respectively. The mixtures were termed PW2SSW1,PW1SSW1, PW1SSW2, PW1SSW8, and the compositions aregiven in Table 3. Due to the early stage of seawater injection intothe Valhall field, the composition of PW was quite similar to thecomposition of initial FW. Ekofisk field FW and PW compositionswere provided by ConocoPhillips, and Valhall field FW and PWcompositions were received from BP.

Core Preparation. The chalk cores were prepared according tothe method described in a previous paper.13 The shaped cores, L )65 cm and D ) 38 mm, were flooded with 250 mL of distilledwater, corresponding to approximately 7 pore volumes, PVs, usinga traditional Hassler core holder in order to remove dissolvablesulfates, which will influence the wetting properties of the cores.12,13

Then, the cores were dried at 120 °C until a constant weight wasobtained. The dry weight was noted and was used to calculate theporosity after saturating the cores with Valhall FW (SO4

2-

removed), Table 3. The cores were drained to a residual watersaturation of about 10% on a porous plate with water-saturated N2

gas by gradually increased in pressure up to 10 bar. Capillarycontact between the cores and the porous plate was establishedthrough a thin layer of paste made from crushed core material andValhall FW. After drainage, the cores were again placed in theHassler core holder, put under vacuum for a maximum of 15 minto remove air, and then saturated with the actual oil. To obtain ahomogeneous wetting condition, the cores were flooded with 1.5PV of oil in each direction at 50 °C with a confining pressure notexceeding 15 bar. The cores were removed from the Hassler celland wrapped in Teflon tape to avoid adsorption of organic materialonto the core surface. After that they were aged in the same oilinside a sealed steel container at 90 °C for 4 weeks, which issufficient time to reach close to adsorption equilibrium.14 After the

(7) Nielsen, C. M.; Olsen, D.; Bech, N., Paper SPE 63226 presented atthe SPE Annual Tech. Conference and Exhibition, Dallas, Texas, 1-4October, 2000; 2000.

(8) Puntervold, T.; Austad, T. J. Pet. Sci. Eng. 2008, 63, 23–33.(9) Bader, M. S. H. Desalination 2007, 208 (1-3), 159–168.(10) Frykman, P. Mar. Pet. Geol. 2001, 18 (10), 1041–1062.(11) Røgen, B.; Fabricius, I. L. Pet. Geosci. 2002, 8 (3), 287–293.(12) Puntervold, T.; Strand, S.; Austad, T. Energy Fuels 2007, 21 (3),

1606–1616.

(13) Puntervold, T.; Strand, S.; Austad, T. Energy Fuels 2007, 21 (6),3425–3430.

(14) Anderson, W. G. , J. Pet. Technol. 1986, NoVember, 1246-1262.

Table 1. Core and Imbibition Data

testNo.

temp.(°C)

PW:SSW(v:v)

porosity(%)

Swi

(%)AN (mgKOH/g)

BN (mgKOH/g)

1 50 PW 49.5 8.9 0.39 0.112 50 2:1 48.7 8.5 0.39 0.113 50 1:1 48.3 8.6 0.39 0.114 50 1:8 49.4 8.3 0.39 0.115 70 PW 47.3 9.0 0.45 0.116 70 2:1 49.6 8.9 0.45 0.117 70 1:1 48.9 9.2 0.45 0.118 70 1:8 46.0 9.5 0.45 0.119 70 PW 48.7 8.6 0.69 0.2110 70 1:1 49.6 8.9 0.69 0.2111 70 1:8 48.8 8.8 0.69 0.2112 90 PW 47.6 9.1 0.69 0.2113 90 2:1 47.3 8.3 0.69 0.2114 90 1:1 46.4 9.3 0.69 0.2115 90 1:2 48.6 8.7 0.69 0.2116 90 1:8 47.8 8.5 0.69 0.2117 110 PW 47.9 8.7 0.70 0.2918 110 1:1 47.3 8.8 0.70 0.2919 110 1:2 44.9 9.1 0.70 0.2920 110 1:8 46.7 8.4 0.70 0.2921 130 PW 47.5 8.6 1.1 0.3122 130 2:1 48.5 9.2 1.1 0.3123 130 1:1 47.4 9.5 1.1 0.3124 130 1:8 48.7 8.8 1.1 0.31

2528 Energy & Fuels, Vol. 23, 2009 PunterVold et al.

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u

aging period, the Teflon tape was removed, and the cores wereready for spontaneous imbibition.

Spontaneous Imbibition. At low temperatures, 50 and 70 °C,spontaneous imbibition of water was performed in standard Amottglass cells, whereas at high temperaturess 90, 110, and 130 °Cssealed steel containers with ∼10 bar back pressure were used. Afteraging, the cores were transferred to the imbibition cell, surroundedwith imbibition brine, and the amount of oil produced (%OOIP)was collected in a buret and recorded as a function of time. Foreach test series, 3-5 cores were imbibed with different mixturesof PW and SSW. The different tests and core data are listed inTable 1.

Viscous Flooding. After reaching the oil production plateau inthe spontaneous imbibition process, that is, oil production hadstopped, the cores imbibed with PW at 90, 110, and 130 °C weretransferred from the imbibition cell to a Hassler core holder toundergo viscous flooding, sometimes also referred to as forcedimbibition. The cores were first flooded with PW, and then withSSW. The cores were placed in a rubber sleeve with a confiningpressure of ∼20 bar and a back pressure of ∼12 bar to preventboiling. The injection fluids were pumped at a nearly constantdifferential pressure, ∆P, across the core, which varied from 1 to9 psi depending on the fluid. The pressure gradient was adjustedso that the maximum rate was never higher than 10% of the corePV per day, that is, ∼3 mL per day. The produced oil was collectedin a buret, and the volume was recorded versus time.

Simulation Tool. The OLI Systems Stream Analyzer 1.3delivered by OLI Systems, Inc. was used for the simulations. Thisis a chemical model software based on thermodynamic equilibriumconditions using published experimental data.

Results and Discussion

Because of strict environmental regulations, oil companiesare required to clean the produced water before deposing it tonature. Reinjection of PW into the formation can solve thisgrowing and increasingly expensive problem provided that itdoes not harm the injection well. Loss of injectivity may be

the case if salts are precipitated from the injection fluid. For afractured reservoir, the injection well can tolerate considerablesolid deposition without injectivity problems.15 Furthermore, theoil recovery from a high temperature chalk reservoir by waterinjection is sensitive to the composition of the injected water.In this section, first the results from simulated compatibilitystudies of mixtures of PW and SSW will be discussed, then theexperimental oil displacement results will be presented usingPW:SSW mixtures as injection fluids.

Precipitation and Sulfate Scale. The compatibility studieswere based on compositions of PW and FW from the Ekofiskfield, as listed in Table 2. Injection of seawater into the Ekofiskfield has been going on since 1987, and the PW is therefore amixture of FW and seawater. The volumetric mixing of FWand SSW can easily be calculated by supposing that Na+ andCl- are inert ions, that is, they do not react with the formationin any way. Calculations based on Cl- showed that 0.267 L ofFW was mixed with SSW to form 1.00 L of PW. Similarcalculations using Na+ gave a consistent value of 0.262 L ofFW. On average, 0.264 L of FW was mixed with 0.736 L ofSSW to give 1.00 L of PW. On the basis of these values, theexpected compositions of the other ions can be calculated,(PW)calc, assuming that no interaction with the formation takesplace. A comparison with the experimentally measured com-positions of the different ions in the PW, (PW)exp, will thenindicate a possible interaction with the formation. Calculatedand measured data can be found in Table 4 and Figure 1.

An increase in the Ca2+sand a decrease inMg2+sconcentration is a clear indication that Mg2+ substitutesCa2+ at the chalk surface at high temperatures, which is in line

(15) Mackay, E. J.; Collins, I. R.; Jordan, M. M.; Feasey, N. , PWRI:Scale formation risk assessment and management, In Paper SPE80385presented at the SPE 5th International Symposium on Oilfield Scale,Aberdeen, U.K., January 29-30 2003; 2002.

Table 2. Molar Compositions (mol/l) of PW:SSW Mixtures for the Ekofisk Field

mixtures PW:SSW

SSW PWa FWa 1:1 1:2 1:4 1:8

Mixtures PW:SSWHCO3

- 0.002 0.008 0.004 0.005 0.004 0.003 0.003Cl- 0.525 0.765 1.423 0.645 0.605 0.573 0.552SO4

2- 0.024 0.007 0.000 0.015 0.018 0.021 0.022Mg2+ 0.045 0.021 0.022 0.033 0.037 0.040 0.042Ca2+ 0.013 0.049 0.100 0.031 0.025 0.020 0.017Na+ 0.450 0.635 1.156 0.543 0.512 0.487 0.471K+ 0.010 0.007 0.007 0.008 0.009 0.009 0.010Ba2+ 0.0 × 1000 8.2 × 10-6 0.002 4.1 × 10-6 2.7 × 10-6 1.6 × 10-6 9.1 × 10-7

Sr2+ 0.0 × 1000 2.2 × 10-3 0.009 1.1 × 10-3 7.4 × 10-4 4.4 × 10-4 2.5 × 10-4

Li+ 0.0 × 1000 9.8 × 10-4 0.000 4.9 × 10-4 3.3 × 10-4 2.0 × 10-4 1.1 × 10-4

ionic strength 0.657 1.150 1.559 0.904 0.821 0.755 0.711TDS (g/l) 33.39 45.81 83.09 39.60 37.53 35.87 34.77

a Data received from ConocoPhillips, Norway.

Table 3. Molar Compositions (mol/l) of PW:SSW Mixtures for the Valhall Field

mixtures PW:SSW

SSW PWa FWa 2:1 1:1 1:2 1:4 1:8

HCO3- 0.002 0.013 0.009 0.009 0.007 0.006 0.004 0.003

Cl- 0.525 1.096 1.066 0.905 0.811 0.715 0.639 0.589SO4

2- 0.024 0.001 - 0.008 0.012 0.016 0.019 0.021Mg2+ 0.045 0.008 0.008 0.020 0.026 0.032 0.037 0.041Ca2+ 0.013 0.031 0.029 0.025 0.022 0.019 0.017 0.015Na+ 0.450 1.027 0.996 0.834 0.738 0.642 0.565 0.514K+ 0.010 0.005 0.005 0.007 0.008 0.008 0.009 0.010ionic strength 0.657 1.150 1.112 0.985 0.904 0.821 0.755 0.711TDS (g/l) 33.39 64.96 62.8 54.38 49.18 43.91 39.70 36.89

a Data received from BP, Norway.

Oil RecoVery from North Sea Chalk Oil ReserVoirs Energy & Fuels, Vol. 23, 2009 2529

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u

with previous studies in the laboratory.4,16 The molar decreasein Mg2+ concentration is slightly higher than the increase inCa2+ concentration, which could point to some precipitation ofCaSO4, as indicated by previous simulation results,8 which arealso presented later in this paper.

The actual concentration of Ba2+ in the PW is significantlylower than that expected from calculations based purely onmixing FW and SSW. This can probably be ascribed to trappingof Ba2+ in the formation due to precipitation of BaSO4, whichis also indicated by the simulation studies presented later. Theconcentration of Sr2+ in the PW is quite similar for both (PW)exp

and (PW)calc, reflecting very little interaction with the formation.SO4

2-, however, is strongly retarded in the formation. Theexperimentally measured concentration in the PW is less thanhalf of the calculated concentration based on mixing FW andSSW. It has been shown earlier that SO4

2- adsorbs onto thechalk surface, and the affinity for the chalk surface increasesdrastically as the temperature increases beyond 100 °C.6

Furthermore, the decrease in the SO42- concentration can also

be linked to precipitation of BaSO4 and CaSO4. It is interestingto note that the interactions between the components in seawaterand the chalk surface observed in the laboratory by using outcropmaterial is similar to observations in the field.17

The reservoir temperature of the Ekofisk field, which is alsothe temperature of the producer, is 130 °C. The temperaturearound the injector is much lower due to the injection of largequantities of cold seawater. The solubility of anhydrous CaSO4,which is the most stable form at high temperatures, decreasesas the temperature increases, Figure 2. The produced waterappears to be saturated with CaSO4 as it enters the productionwell. The fact that the solubility of CaSO4 increases withdecreasing temperature will probably prevent CaSO4 scaling inthe producer due to the decrease of temperature in the productiontubing. The precipitation of CaSO4 from the various mixturesof PW and SSW is negligible below 100 °C. Provided that thetemperature around the injector is less than 100 °C, which isquite reasonable, precipitation of CaSO4 will not cause injectionproblems for the mixed fluids. It is interesting to note that evenSSW appeared to precipitate CaSO4 at 130 °C under athermodynamic equilibrium condition.

The trend in the solubility of SrSO4 is similar to CaSO4, butthe solubility is lower, Figure 3. Also in this case, the PW issaturated with SrSO4 when the fluid enters the production wellat 130 °C. The solubility of SrSO4 increases gradually as theamount of SSW increases at temperatures below 130 °C. If theproduced water is diluted more than 4 times with SSW, thenthe mixture could be injected without scale problems when thetemperature in the injection area is less that 80 °C.

The solubility of BaSO4 has an opposite temperature trendcompared to CaSO4 and SrSO4, Figure 4. The PW is saturatedwith BaSO4 at 130 °C, and the precipitation of BaSO4 willincrease with decreasing temperature in the production system.Thus, BaSO4 may cause a serious scale problem as the volumeof PW increases. However, if the PW is diluted 4 times withSSW, then the amount of precipitated BaSO4 is very small, infact, less than 10-6 mol/L in the entire temperature range30-130 °C.

On the basis of this simple screening, it appears that if thePW is diluted 4 times or more with SSW, then the mixture issuitable for injection into a fractured chalk reservoir withoutloss of injectivity due to scale formation. The question istherefore: “Will mixtures of PW and SSW displace the oilequally good as pure SSW?”.

Oil Recovery by Using Mixtures of PW and SSW.Experimental imbibition studies of oil recovery from moderatewater-wet chalk cores have been performed by using the variousmixed water compositions listed in Table 3 as imbibition fluids.Traces of Ba2+ and Sr2+ have been left out to avoid possibleprecipitation. The FW, used as the initial water in the corepreparation, did not contain any SO4

2-. The composition ofproduced water is quite similar to the FW, which indicates thatthe PW is sampled early in the seawater injection phase, thatis, before seawater breakthrough.

The oil recovery potential, when using different mixtures ofPW and SSW, was studied by spontaneous imbibition. Theimbibition rate and ultimate oil recovery indicate the relativeability of the fluids to displace the oil by wettability modifica-tion. It has been documented previously that spontaneousimbibition of SSW by wettability modification increases withincreasing temperature.3-5 It is also known that the water-wetness decreases when the AN of the crude oil increases.18 Itwas recently verified that the basic components in the crudeoil, represented by the base number, BN, were able to increasethe water wetness slightly for crude oils with a constant AN )0.50 mg KOH/g oil.12 To study the oil recovery from moderatewater-wet chalk at different temperatures, 50, 70, 90, 110, and130 °C, oils with different wetting potentials, that is, differentAN, were used. Higher AN was needed at higher temperaturesto see possible differences in the oil displacement potential forthe various mixtures of PW and SSW. For the different oilsused, AN varied between 0.39 and 1.1 mg KOH/g oil, whereasthe BN varied between 0.11 and 0.31 mg KOH/g oil, Table 1.The AN is, however, the most important wetting parameter, andit will dictate the wetting conditions.12

The spontaneous imbibition tests at 50 °C were performedwith an oil with AN ) 0.39 mg KOH/g oil, Figure 5. Thedifferent cores appeared to be moderately water-wet becauseof moderate imbibition rate and oil recovery. For a water-wetsystem, the oil recovery is fast, and close to 60% of the oil willbe recovered. No systematic trend in the imbibition behaviorwas detected for the different fluid mixtures, which indicatesthat the temperature was too low to activate the wettabilitymodification reaction during the imbibing time frame.

Also in the imbibition tests at 70 °C using an oil with AN )0.45 mg KOH/g oil, no discrimination in oil recovery was seenfor the different mixtures of PW and SSW, Figure 6. Theimbibition rate and the oil recovery were high, which indicatedthat the cores probably were too water-wet to see any effect ofthe water compositions used. As a result of that, another testseries was performed at 70 °C, Figure 7, this time with a more

(16) Korsnes, R. I.; Strand, S., Hoff, Ø.; Pedersen, T.; Madland, M. V.;Austad, T., EUROCK 2006 Multiphysics Coupling and Long TermBehaViour in Rock Mechanics; Taylor & Francis Group: London, 2006;ISBN: 0 415 41001 0.

(17) Mackay, E. J.; Jordan, M. M., Natural sulphate ion stripping duringseawater flooding in chalk reservoirs. In Chemistry in the oil industry VIIIsymposium, Manchester, UK, 3-5 November 2003.

(18) Standnes, D. C.; Austad, T. J. Pet. Sci. Eng. 2000, 28 (3), 111–121.

Table 4. Composition of PW from the Ekofisk Field, Data fromExperimental Analysis, and Calculated Data Based on Mixing of

FW and SSW

component(PW)exp

(mol/L)(PW)calc

a

(mol/L)[(PW)exp -

(PW)calc] (mol/L)

Ca2+ 0.049 0.036 + 0.013Mg2+ 0.021 0.039 - 0.018Ba2+ 8.2 × 10-6 5.3 × 10-4 - 0.000 52Sr2+ 0.0022 0.0024 - 0.0002SO4

2- 0.007 0.018 - 0.011

a Calculated based on mixing 0.264 L of FW and 0.736 L of SSW.

2530 Energy & Fuels, Vol. 23, 2009 PunterVold et al.

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u

acidic oil, AN ) 0.69 mg KOH/g oil, in order to obtain a lesswater-wet system. It can be seen from the results in Figure 7

that oil recovery was reduced to ∼30-35%. After 90 days ofimbibition the core imbibed with PW stabilized at ∼30%,

Figure 1. Difference between calculated and measured component concentration in PW, which is linked to substitution of Ca2+ by Mg2+ at the rocksurface, adsorption of SO4

2- onto the rock, and precipitation of CaSO4.

Figure 2. Precipitation of CaSO4 vs temperature for mixtures of seawater and PW from Ekofisk field - Ekofisk formation.

Figure 3. Precipitation of SrSO4 vs temperature for mixtures of seawater and PW from Ekofisk field - Ekofisk formation.

Oil RecoVery from North Sea Chalk Oil ReserVoirs Energy & Fuels, Vol. 23, 2009 2531

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u

whereas the cores imbibed with PW1SSW1 and PW1SSW8 hadreached ∼35 and ∼30%, respectively, and were still producingslowly. It seemed that a temperature of 70 °C was too low toproduce any significant difference in wettability propertiesbetween cores imbibed with and without SO4

2- present in thebrine.

The test series at 90 °C using an oil with AN ) 0.69 mgKOH/g oil, appeared to contain moderate water-wet cores, butin contrast to the tests at 50 and 70 °C, a small differences inthe oil-displacement potential for the different mixtures couldbe detected, Figure 8. The fluids containing SSW, thatis, withSO4

2- present, appeared to give 5-10% higher oil recoverycompared to PW, which stabilized at 30% recovery. Previousstudies have shown that chalk cores saturated with an initialbrine containing no SO4

2- respond poorly to SSW as awettability modifying fluid at temperatures below 100 °C.13

After spontaneous imbibition was completed for the coreimbibed with PW, the core was flooded with PW at the sametemperature. The core responded with a ∼4% increase in oilrecovery. When the flooding fluid was switched from PW to

SSW the oil recovery increased another ∼6%. Thus, viscousflooding is successful at 90 °C, and SSW is preferred over PW,due to the potential for wettability alteration. These results arebacked up in a similar study looking at the oil recoverydifferences between a spontaneous imbibition process and aviscous flood with respect to temperature and water composi-tion.19

The affinity of SO42- toward the chalk surface increases

drastically with temperatures >100 °C.6 The spontaneousimbibition tests performed at 110 °C, using an oil with AN )0.70 mg KOH/g oil, were sensitive to the composition of theimbibing fluid, Figure 9. After 20 days of imbibition, only 32%of the oil was recovered when imbibing pure PW. The differentmixtures with SSW gave an oil recovery around 55% for sametime interval. Obviously, the presence of SO4

2- in the imbibingfluid is crucial for improving the oil recovery by wettabilitymodification. These findings are in line with previous work.4

(19) Strand, S.; Puntervold, T.; Austad, T. Energy Fuels 2008, 22, 3222–3225.

Figure 4. Precipitation of BaSO4 vs temperature for mixtures of seawater and PW from Ekofisk field - Ekofisk formation.

Figure 5. Spontaneous imbibition of mixtures of PW and SSW into chalk cores at 50 °C saturated with a crude oil of AN ) 0.39 mg KOH/g oil.

2532 Energy & Fuels, Vol. 23, 2009 PunterVold et al.

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u

The different mixtures of PW and SSW appeared to containhigh enough concentration of all the important ions, Ca2+, Mg2+,and SO4

2-, to promote wettability modification. In fact, theimbibition curves were quite similar, Figure 9. After 20 daysof spontaneous imbibition, the core exposed to PW had stoppedproducing oil and was removed from the imbibition cell andsubjected to viscous flooding with PW at the same temperature.The differential pressure across the core was gradually increased.After 22 days, it had reached 4 psi, and the first drop of oil wasproduced. After the oil recovery plateau was reached, the totaloil recovery had increased to about 40%. Finally, PW wasexchanged with SSW as injection fluid in a tertiary water floodprocess. A sudden and drastic increase in oil recovery was seen,and the total oil recovery amounted to about 70%, which issomewhat higher than obtained for a spontaneous imbibitionprocess with a water-wet system. Thus, SSW also improves theoil recovery in a viscous flood due to its ability to improve thewater wetness of chalk. This agrees well with the results in a

similar study,19 and it is also in line with observations usingValhall cores at reservoir conditions.20

The imbibition tests at 130 °C were performed with an oilwith AN ) 1.1 mg KOH/g oil, Figure 10. The trend in theimbibition behavior was similar to the experiments performedat 110 °C, that is, the oil recovery by using pure PW was onlyhalf of the recovery obtained by the mixed fluids, 30 and 60%,respectively. The reason is that PW does not contain SO4

2-,and because of that wettability modification is prevented. Afterspontaneous imbibition was completed, the core imbibed withPW was subjected to viscous flooding with the same fluid. Adifferential pressure across the core was gradually increased to2 psi, and a very small increase in oil recovery, 3-4%, wasobtained. Tertiary displacement of oil by SSW, however,improved the oil recovery significantly, reaching close to 60%,

(20) Webb, K. J.; Black, C. J. J.; Tjetland, G., A laboratory studyinvestigating methods for improving oil recovery in carbonates In Inter-national Petroleum Technology Conference; IPTC: Doha, Qatar, 2005.

Figure 6. Spontaneous imbibition of mixtures of PW and SSW into chalk cores at 70 °C saturated with a crude oil of AN ) 0.45 mg KOH/g oil.

Figure 7. Spontaneous imbibition of mixtures of PW and SSW into chalk cores at 70 °C saturated with a crude oil of AN ) 0.69 mg KOH/g oil.

Oil RecoVery from North Sea Chalk Oil ReserVoirs Energy & Fuels, Vol. 23, 2009 2533

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u

which is more or less the same recovery level as obtained fromthe mixed fluid systems.

It is interesting to note that there appeared to be a certaintemperature between 90 and 110 °C where the wettabilitymodification was activated. The oil recovery increased drasti-cally when increasing the temperature from 90 to 110 °C. Inprevious experiments, wettability modifications at lower tem-perature, 70 °C, have been experienced,4,5 but in those experi-ments the cores were not preflushed with DW to removedissolvable sulfates. Thus, it appears that cores completely freefrom sulfate at initial conditions require a higher temperaturein order to undergo wettability modification during a spontane-ous imbibition process.

Another important observation is that the concentration ofthe active ionssCa2+, Mg2+, and SO4

2-sin the PW:SSWmixtures appeared to be high enough to promote the wettabilitymodification, provided that the temperature was above 100 °C.

No significant difference in oil recovery was observed for thedifferent mixtures at 110 and 130 °C, even with a rather largedifference in the AN of the crude oils, 0.70 and 1.1 mg KOH/goil, respectively.

Compaction. Previously it has been shown that the chemicalreactions that promote wettability modification by SSW at hightemperatures also cause enhanced weakening or compaction ofchalk under stress.16,21,22 The mechanism behind the chemicalweakening of chalk by SSW was described by a substitution ofCa2+ by Mg2+ at the intergrain contacts in the presence of

(21) Korsnes, R. I.; Madland, M. V.; Austad, T. , Impact of brinecomposition on the mechanical strength of chalk at high temperature, InEUROCK 2006 Multiphysics Coupling and Long Term BehaViour in RockMechanics; Taylor & Francis Group: London, 2006; ISBN: 0 415 410010.

(22) Korsnes, R. I.; Madland, M. V.; Austad, T.; Haver, S.; Røsland,G. J. Pet. Sci. Eng. 2007, 60 (3-4), 183–193.

Figure 8. Spontaneous imbibition of mixtures of PW and SSW into chalk cores at 90 °C saturated with a crude oil of AN ) 0.69 mg KOH/g oil.After spontaneous imbibition using PW, viscous flooding was performed with PW and SSW at 90 °C. ∆P ) 2-4 psi.

Figure 9. Spontaneous imbibition of mixtures of PW and SSW into chalk cores at 110 °C saturated with a crude oil of AN ) 0.70 mg KOH/g oil.After spontaneous imbibition using PW, viscous flooding was performed with PW and SSW at 110 °C. ∆P ) 4 psi.

2534 Energy & Fuels, Vol. 23, 2009 PunterVold et al.

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u

SO42-.21 Thus, injected fluids that promote wettability modifica-

tion also promote enhanced mechanical weakening of chalk.Therefore, based on the findings in this study, it is reasonableto conclude that PW:SSW mixtures will behave in similar waysas pure SSW regarding compaction. Due to capillary effectsand low initial water saturation, the chalk surface in the thinfilm close to the grain contacts is believed to stay water-wetregardless of the initial wetting condition of the chalk. As aresult of that, during a dynamic water flood with SSW or amixture of PW and SSW, the enhanced water weakening ofchalk will be nearly unaffected by the wetting conditions aslong as the supply of the active ions takes place. If, however,the chalk is subjected to SSW or a mixture of PW and SSWunder a stationary condition, that is, no flooding takes placeafter the oil was displaced to residual saturation, the mechanicalstrength of the chalk appeared to decrease somewhat as thewater-wetness decreased.23 In a water-wet system, the surfacearea of the chalk where aqueous chemical reactions are takingplace is greater than in the case of a mixed-wet system. As aconsequence, after chemical equilibrium has been established,the water-wet area of the mixed-wet chalk has been exposed togreater chemical substitution than for a completely water-wetsystem. Compaction as a drive mechanism is favorable regardingoil recovery, but this is an expensive IOR-method because ofwell damage and rig overhaul due to subsidence of the seafloor.Compaction can be decreased by removing SO4

2- from SSWor by injecting pure PW, but then the oil recovery will bedrastically decreased.

Conclusions

In this work, we have studied the compatibility of mixturesof PW and SSW by modeling solubility properties of actualsulfate scales at various temperatures. Experimental oil recoverytests were also performed. The main results from this work were:

• If the PW is diluted at least 4 times with SSW, the simulatedresults showed that scale problems related to precipitation ofCaSO4, SrSO4, and BaSO4 in the injection well appeared to bevery small.

• Above approximately 100 °C, the tested mixtures of PWand SSW were able to displace the oil in a spontaneousimbibition process from cores of low water-wetness. Nodifference in oil recovery was observed for the various PW:SSW mixturess2:1, 1:1, 1:2, and 1:8. The oil recovery was atleast as high as for a completely water-wet condition. Imbibitionof PW gave an oil recovery that was only half of the recoveryfrom imbibing PW-SSW mixtures.

• SSW also enhanced the oil recovery drastically by viscousdisplacement due to wettability modification at temperaturesabove 100 °C. The oil recovery by injecting SSW can then beregarded as a tertiary method.

• Below 100 °C, the oil was displaced in a spontaneousimbibition process without a significant wettability modification,that is, there was no significant difference in oil recoverybetween PW and mixtures of PW and SSW.

• Viscous flooding at 90 °C showed an extra ∼4% recoveryby injecting PW and another 6% by injecting SSW. The successof injecting SSW is greater above 100 °C.

• At high temperatures, the enhanced mechanical weakeningof chalk by mixtures of PW and SSW will be similar to theprevious observations with SSW. Due to the water-wet conditionof the chalk surface in the thin water film at the intergraincontacts, the decrease in mechanical strength will probably beindependent of the wetting conditions in a dynamic floodingprocess.

Nomenclature

AN ) Acid number, mgKOH/g oilBN ) Base number, mgKOH/g oilEOR ) Enhanced oil recoveryFI ) Forced imbibition (viscous flooding)FW ) Formation waterIOR ) Improved oil recoveryN2 ) Nitrogen gasOOIP ) Oil originally in placePV ) Pore volumePW ) Produced waterPWxSSWy ) Brine mixture of x parts PW and y parts SSW

(23) Strand, S.; Hjuler, M. L.; Torsvik, R.; Pedersen, J. I.; Madland,M. V.; Austad, T. Pet. Geosci. 2007, 13, 69–80.

Figure 10. Spontaneous imbibition of mixtures of PW and SSW into chalk cores at 130 °C saturated with a crude oil of AN ) 1.1 mg KOH/g oil.After spontaneous imbibition using PW, viscous flooding was performed with PW and SSW at 130 °C. ∆P ) 1-2 psi.

Oil RecoVery from North Sea Chalk Oil ReserVoirs Energy & Fuels, Vol. 23, 2009 2535

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u

SSW ) Synthetic seawaterSwi ) Initial water saturation∆P )Differential pressure

Acknowledgment. The authors acknowledge ConocoPhillips andthe Ekofisk coventurers, including TOTAL, ENI, StatoilHydro andPetoro, and the Valhall partnership, including BP Norge AS,

Amerada Hess Norge AS, A/S Norske Shell, and Total E&P NorgeAS for financial support. Also, thanks are due to the NorwegianResearch Council, NFR, for support. A. J. Reinholdtsen, I. Nazarovaand E. M. Nicolaisen are recognized for their technical help at thelaboratory.

EF801023U

2536 Energy & Fuels, Vol. 23, 2009 PunterVold et al.

Dow

nloa

ded

by P

OR

TL

AN

D S

TA

TE

UN

IV o

n Ju

ly 2

, 200

9Pu

blis

hed

on M

arch

20,

200

9 on

http

://pu

bs.a

cs.o

rg |

doi:

10.1

021/

ef80

1023

u