Mechanistic modeling of low salinity water flooding

Author’s Accepted Manuscript Mechanistic Modeling of Low Salinity Water Flooding Cuong Dang, Long Nghiem, Ngoc Nguyen, Zhangxin Chen, Quoc Nguyen www.elsevier.com/locate/petrol

PII: DOI: Reference:

S0920-4105(16)30131-0 http://dx.doi.org/10.1016/j.petrol.2016.04.024 PETROL3425

To appear in: Journal of Petroleum Science and Engineering Received date: 28 September 2015 Revised date: 22 February 2016 Accepted date: 18 April 2016 Cite this article as: Cuong Dang, Long Nghiem, Ngoc Nguyen, Zhangxin Chen and Quoc Nguyen, Mechanistic Modeling of Low Salinity Water Flooding, Journal of Petroleum Science and Engineering, http://dx.doi.org/10.1016/j.petrol.2016.04.024 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Mechanistic Modeling of Low Salinity Water Flooding Cuong Dang(1)*, Long Nghiem(2), Ngoc Nguyen(1), Zhangxin Chen(1), Quoc Nguyen(3) (1): Department of Chemical & Petroleum Engineering, University of Calgary (2) Computer Modelling Group Ltd. (3)Department of Petroleum & Geosystems Engineering, The University of Texas at Austin (*) Corresponding Author Abstract Low salinity waterflood (LSW) has become an attractive enhanced oil recovery (EOR) method as it shows more advantages than conventional chemical EOR methods in terms of chemical costs, environmental impact, and field process implementation. Extensive laboratory studies in the past two decades have proposed several pore-scale mechanisms of oil displacement during LSW flooding, which are still open for discussion. However, the capability of reservoir simulators to model accurately this process is very limited. This paper provides a critical review of the state of the art in research and field applications of LSW. The focus is on a widely agreed mechanism that is the wettability alteration from preferential oil wetness to water wetness of formation rock surfaces. Ion exchange and geochemical reactions have been experimentally found to be important in oil mobilization due to enhanced water spreading at low salinity. To evaluate the significance of this surface wetting mechanism, a comprehensive ion exchange model with geochemical processes has been developed and coupled to the multi-phase multi-component flow equations in an equation-of-state compositional simulator. This new model captures most of the important physical and chemical phenomena that occur in LSW, including intra-aqueous reactions, mineral dissolution/precipitation, ion exchange and wettability alteration. The proposed LSW model is tested using the low-salinity core-flood experiments reported by Fjelde et al. (2012) for a North Sea reservoir and the low-salinity and high-salinity heterogeneous core-flood experiments by Rivet (2009) for a Texas reservoir. Excellent agreements between the model and the experiments in terms of effluent ion concentrations, effluent pH, and oil recovery were achieved. In addition, the model was also proved to be highly comparable with the ion-exchange model of the geochemistry software PHREEQC for both low salinity and high salinity (Appelo, 1994). Important observations in laboratory and field tests such as a local pH increase, a decrease in divalent effluent concentration, mineralogy contributions, and the influence of connate water and injected brine compositions can be reproduced with the proposed LSW model. Built in a robust reservoir simulator, it serves as a powerful tool for LSW design and the interpretation of process performance in field tests. Although the advantages of LSW have been reported, none of the systematic approaches has addressed the issues of modeling and predicting LSW field-scale performance. The new modeling approach introduced in this research can effectively capture the critical effects of geology in the LSW process. This integrated modeling approach involves the use of geological software, a reservoir simulator, and a robust optimizer in a closed loop for sensitivity analysis, history matching, optimization, and uncertainty assessment. The field-scale benefits of secondary and tertiary LSW are then addressed. Keywords: Low Salinity Water Flooding, Enhanced Oil Recovery, Wettability Alteration, Geochemistry

1

1. State-of-the Art Smart LSW Low salinity water flooding (LSW) is an emerging EOR technique in which the salinity of the injected water is controlled to improve oil displacement efficiency without a significant loss of injectivity due to clay swelling. In particular, the presence of clay minerals is a favorable condition for the high efficiency of this process. This recovery concept is quite attractive as 50% of the world’s conventional petroleum reservoirs are found in sandstones that commonly contain clay minerals. It has been experimentally found that changes in the injected brine composition can improve waterflood performance by up to 38% (Webb et al. 2004), leading to a new concept of optimal injection brine composition for water flood. In the 1990s, Jadhunandan and Morrow (1995) and Yildiz and Morrow (1996) reported the influence of brine composition on oil recovery, which identified a possibility to improve waterfloods with optimized injection brine formulation. Numerous laboratory experiments (Tang and Morrow, 1997; Morrow et al., 1998; Tang and Morrow, 1999a; Tang and Morrow, 1999b; Zhang and Morrow, 2006; Zhang et al., 2007b; Buckley and Morrow, 2010; Kumar et al., 2010; Lohardjo et al, 2010; Morrow and Buckley, 2011) have later confirmed that EOR can be obtained in tertiary low salinity waterflood. The salinity in these tests was in the range of 1,000-2,000 ppm. Although the advantages of LSW have been realized, the mechanisms of incremental oil recovery by LSW are still open. Several mechanisms have been proposed during last two decades including fines migration, wettability alteration, multi-component ionic exchange (MIE), saponification and pH modification, and electrical double layer effects. Dang et al. (2013) provided a critical review of these mechanisms. Fines migration is one of the first explanations for the incremental oil recovery by LSW. Morrow and his research colleagues observed that fines (mainly kaolinite clay fragments) were released from rock surfaces, resulting in an increase in spontaneous imbibition recovery with a decrease in salinity for different sandstones. The fines migration was believed to improve mobility control of the conventional high salinity water flooding by plugging pore throats, a reduction in the effective permeability to water in a water swept zone, and diverting flow into un-swept zones. However, numerous researchers from industry (Lager et al., 2007, 2008a, 2008b; Webb et al., 2004, 2005, 2008; McGuire et al., 2005; Jerauld et al., 2008) reported that LSW has higher recovery without any observations of fines migration during their experiments and pilot tests. Based on these observations, people questioned about the link between fines migration and the additional oil recovery and it is not the direct cause for the benefits of LSW. In fact, a local increase in pH is usually observed during LSW (McGuire et al., 2005; Lager, 2006, Zhang and Morrow, 2006b, and Austad et al., 2010), and it is suggested that the EOR mechanisms of LSW appear similar to those of the alkaline flooding by generation of in-situ surfactants and a reduction in interfacial tension. Note that the acid number of crude oil should be larger than 0.2 mg KOH/g in order to generate enough surfactant to induce wettability reversal and/or emulsion formation in alkaline flooding (Ehrlich et al., 1974; Jensen and Radka, 1988), but most of crude oil samples have been used with acid number less than 0.05 mg KOH/g for LSW experiments. Therefore, it is difficult to conclude that additional oil recovery is mainly from in-situ surfactant generation. Among the proposed hypotheses, wettability alteration towards increased water wetness during the course of LSW is the widely suggested case of increased oil recovery. The effects of low salinity brine on wettability modifications have been reported by many authors (Jadhunandan and Morrow, 1995; Tang and Morrow, 1999; Drummond and Israelachvili, 2002 and 2004; Vledder et al., 2010; Zekri et al., 2011). It has been experimentally found that the low salinity brine has a significant effect on the shape and the end points of the relative permeability curves (Webb et al., 2004; Kulkarni and Rao, 2005; Rivet, 2009; 2

Fjelde et al., 2012), resulting in lower water relative permeability and higher oil relative permeability. This phenomenon can be physically explained by the ionic exchange between the injected brine and formation water, and mineral dissolution/precipitation in LSW. The ionic exchange during this process leads to the adsorption of divalent ions, promotes the mineral dissolution, and changes the ionic composition of formation water and the wettability condition. This is the most reliable suggestion for the underlying mechanisms of LSW and it is used to develop a mechanistic model of LSW in this paper. LSW has been evaluated at the field scale based on promising results from laboratory experiments (McGuire et al. 2005; Lager et al., 2008; Skrettingland et al., 2010; Thyne and Gamage, 2011). A summary of LSW field implementations is shown in Table 1. McGuire et al. (2005) and Lager et al. (2008) reported the LSW performance in Alaska’s oil fields using water injection salinity between 1,500 to 3,000 ppm. Based on single well chemical tracer tests (SWCTT), the reported incremental LSW EOR in these oil fields ranged from 6 to 12% OOIP, resulting in an increase in water flooding recovery by 8 to 19%. Lager et al. (2006) indicated that fines migration or significant permeability reduction due to LSW was never observed. They also achieved a 40% increase in recovery using North Sea crude oil with a very low acid number (Acid Number < 0.05). Another result from the Log-Inject-Log test (Webb et al., 2004) showed 25-50% reduction in residual oil saturation by LSW. Another successful LSW by a SWCTT test was described by Seccombe et al. (2010) in a mature offshore oil field located on the North Slope of Alaska. Additionally, historical field evidence in the Powder River basin of Wyoming reported by Robertson (2010) also showed that oil recovery tended to increase by about 12.4% as the salinity of injection brine decreases. Thyne and Gamage (2011) published a comprehensive evaluation of the effect of LSW for 26 field trials in Wyoming. Table 1: Summary of LSW Implementation in the field

Reservoir

Injected/Formation Formation Salinity (ppm) Damage

Incremental Oil Recovery (%OOIP)

References

Webb et al. Sandstone

3,000/ 220,000

No

20 -50 (2004)

Sandstone

McGuire et al. 150-1,500 /15,000

No

13

< Alaskan Oil Field >

(2005)

Sandstone

10,000/60,000



3,304/42,000



7,948/128,000

No

Recovery tends to decrease as the salinity ratio increases.

Sandstone

(2007) Lager et al.

2,600/ 16,640

No

10 (2008b)

Sandstone

Robertson

Veledder et al. 2,200/ 90,000

No 3

10 - 15 (2010)

Sandstone

Seccombe et al. No

12,000/ --

13 (2010)

Sandstone

Skrettingland et al. 500/50,000

No

No significant change.

(2010)

While extensive experimental studies of LSW have been reported, modeling work is rarely found in the literature. One of the earliest studies on modeling LSW was presented by Jerauld et al. (2008). They developed a new model for LSW with some modifications from a traditional water flooding model. In their model, salt was modeled as an additional single-lumped component in the aqueous phase, and relative permeability and capillary pressure were made functions of salinity, including the influence of connate water, hysteresis in water relative permeability between imbibitions, and dispersion phenomena. Unfortunately, this model used simple linear dependence of residual oil saturation on salinity. Rueslatten et al. (2008) performed a LSW experiment on a North Slope core sample. Then a model using the geochemical code, PHREEQC, was created to simulate the LSW. This model gave only an approximation of the pH variation as the mechanism of LSW. Subsequently, Sorbie and Collins (2010) extended their work by introducing a semi-quantitative model that describes a multicomponent ion exchange process at the pore scale. This model attempts to show the consequences of the change in the electrical double layers and the adsorption of polar organic species. However, further experimental studies are required to confirm this mechanism. Recently, Omekeh et al. (2012) presented a model of ion exchange and mineral solubility in LSW. They considered two-phase flow of oil and brine that contains Na+, Mg++, SO4--, and Cl-. Cations are involved in a fast ion exchange process with the negative clay surface, X -. In this model, the total release of divalent cations from the rock surface gives rise to a change in the relative permeability so that more oil is mobilized, and that the desorption of divalent ions is the main mechanisms of LSW. However, divalent ions such as Ca++ and Mg++ are expected to be adsorbed on the clay mineral during the course of LSW (Appelo, 1994) and the wettability alteration happens with the adsorption of divalent ions only (Suijkerbuijk et al. 2012). In addition, it would be more advantageous to model the process in a compositional simulator with full geochemical reactions instead of in a simplified two-phase flow model. Korrani et al. (2013) coupled IPHREEQC, an open-source geochemical module from the United States Geological Survey, with UTCHEM, a research chemical-flooding reservoir simulator developed at the University of Texas at Austin, for modeling LSW. The authors used only simple case studies to compare the ion exchange between UTCHEM-IPHREEQC and the original IPHREEQC. These comparisons have not provided additional insights since the ion-exchange model used in UTCHEMIPHREEQC was adapted from the one used in the original IPRHEEQC module. The results from this model should be validated against coreflood experiments and field observations. In the past few decades, LSW has received favorable attention mainly for sandstone reservoirs; however, the current interest of LSW applications for carbonate reservoirs also increased significantly from the oil and gas industry. One of the primary mechanisms proposed for the benefits by LSW in carbonate reservoirs is the wettability alteration. Several hypotheses have been suggested to explain for the wettability alteration, e.g., reactive potential determining ions, surface adsorption/desorption, mineral 4

dissolution, surface potential reduction, capillary pressure and temperature effects. (Zhang et al. 2006a; 2007a; Austad et al. 2008; Fathi et al. 2010; Hiorth et al. 2010; Evje and Hiorth, 2012; Nasralla et al. 2014; Qiao et al. 2015a). Based on those hypotheses, several approaches have been proposed in the scientific literature to modeling LSW in carbonates. Evje and Hiorth (2011) introduced a model that incorporates two-phase water-oil flow and water-rock interaction in carbonate reservoirs. This model focused mainly on calcite dissolution as the mechanism responsible for LSW. However, numerous LSW coreflood experiments have been conducted at low temperature with incremental oil recovery. Al-Shalabi et al. (2014) reported that there was no consistent relationship between mineral dissolution and incremental oil recovery by LSW. Thus the dissolution of calcite could play a role in LSW, but may not be the sole process that is responsible for wettability alteration during the LSW process. Al-Shalabi et al. (2014) used contact angles to represent surface wettability, however; the empirical correlation between salinity and a contact angle is non-predictive (Qiao et al. 2015b). Recently, Johns and his co-workers have proposed a new modeling approach for LSW in carbonate reservoirs using a multiphase reactive transport model with surface complexation and mineral dissolution and precipitation (Qiao et al. 2015a; 2015b). This model has been implemented in their in-house simulator to match experimental data with promising results. 2. Mechanistic Modeling of Low Salinity Water Flooding Several geochemical software programs have been developed including SOLMINEQ.88, KGEOFLOW, Geochemist’s Workbench, TOUGHREACT, PHREEQC, which are commonly used to model geochemical reactions. These programs can model complex phenomena such as mineral dissolution/precipitation and intra-aqueous species equilibrium reactions, which are not available in conventional reservoir or black-oil simulators. Unfortunately, these geochemical packages were developed for a specific chemical engineering purpose, and cannot adequately be extended to petroleum engineering. None of the programs could be used to explain the myriad experimental LSW laboratory observations, or are able to predict LSW performance at the field scale. As outlined earlier, LSW modeling is rarely discussed in the literature and the capability of reservoir simulators to accurately model this process is limited. The main target of this section, therefore, is to describe the development of a mechanistic model that can be used to evaluate the incremental oil-recovery benefit of LSW over conventional water flooding in oil reservoirs. Governing Equations The flow in porous media is governed by Darcy’s law and the mechanical dispersion/diffusion of components. The conservation equations for the species in a compositional model are described as follows. The conservation equations of the nh components in the oil and gas phases that may be soluble in the aqueous phase are:

i 

 Tu yiu pn 1  Pcu  ~u gd   Diqu yiqu 

  o, g , w



q  g , o, w



V n 1 1 Vin,aq  q in 1  Ni  Nin  0, i  1,..., n h ; t

5

(1)

The conservation equations for the na aqueous components (ions) are:





~ u gd  D u y u  Vn 1  V n 1  j  Twu y ujw p n 1   w iw iw j, aq j, mn  q nj 1 





(2)

V n 1 N ja  N nja  0, j  1,..., n a t

The conservation equations for the nm mineral components are: 1  k  Vnk,mn 





V n 1 N k  N nk  0, k  1,..., n m . t

(3)

The superscripts n and n+1 denote the old and new time levels, respectively. These equations are discretized in an adaptive implicit manner. The superscript u = n is for explicit gridblocks and n+1 for 1 1 ) and ( implicit gridblocks. The terms ( V nk ,mn ) correspond, respectively, to the intra-aqueous Vin,aq

reaction rates and mineral dissolution/precipitation rates and will be discussed in the next section. The equation for thermodynamic equilibrium is the equality of fugacities of the components between the oil, gas and aqueous phases, i.e., gi,1  fig  fio  0, i  1,..., n h ,

(4)

gi,2  fig  fiw  0, i  1,..., n h ,

(5)

The fugacity f ig is calculated from the Peng-Robinson equation of state. The fugacity f iw of gaseous components soluble in the aqueous phase is calculated using Henry’s law, i.e.,

fiw  yiw Hi

(6)

In compositional simulation, a volume constraint equation is normally also a part of the equation set:

 N qn 1    n 1   n 1  0 , q = g,w,o. q  q 

(7)

Intra-aqueous and Mineral Reactions Nghiem et al. (2004) proposed a fully-coupled geochemical equation-of-state model for the simulation of CO2 storage in a saline aquifer in a compositional simulator. It allows the modeling of intraaqueous reactions in the aqueous phase and reactions between minerals and aqueous species. Although this model was initially designed for the simulation of CO2 storage, it is also applicable to the modeling of geochemical reactions in the LSW process. Nghiem et al.’s model has been extended to model LSW, which includes multiple-ion exchange and wettability alteration in addition to intra-aqueous chemical reactions and mineral dissolution and precipitation. A chemical reaction is in equilibrium if the forward reaction-rate and backward reaction rates are equal giving a net reaction-rate of zero. Reactions in equilibrium satisfy the following condition: 6

Q  K eq ,  0 , α = 1,…,Raq

(8)

n aq

Q   (a k ) k

(9)

k 1

Qα is the activity product of a reaction where naq is the number of species in an intra-aqueous equilibrium reaction, ak is the activity of each species in the aqueous phase, Keq is the equilibrium constant for the chemical reaction, and  k is the stoichiometric coefficient of a component in the chemical reaction. The chemical equilibrium constant (Keq) values are provided as a function of temperature for many aqueous reactions by Kharaka et al. (1988). Activities for each species (αk) are related to their concentration (mk) in solution. The activities of water and the mineral species can be taken to be unity without introducing significant errors:

i   i mi

i = 1,…, naq

(10)

where  i is the activity coefficient. The activity coefficient in ideal cases can be taken to be approximately equal to the concentration of the species (in molality). Intra-aqueous reactions are modeled as instantaneous reactions. This increases the computational efficiency by permitting the use of an Equilibrium-Rate-Annihilation (ERA) matrix that reduces the number of flow equations (Nghiem et al., 2004, 2011). Reactions between the aqueous species and the reactive minerals in a reservoir are kinetically controlled. The reactions involving reservoir minerals cause the minerals to dissolve or precipitate when they are not in a state of equilibrium with the ions in the aqueous phase. The dissolution or precipitation rate of a mineral is calculated as:

 Q  ˆ k 1  ,   1,..., R mn r  A    K eq ,  

(11)

where: Âβ = reactive surface area of reactant mineral β per unit bulk volume of porous medium (m2/m3) kβ = rate constant of mineral reaction β (mol/m2·s). Keq,β = chemical equilibrium constant of mineral dissolution/precipitation reaction Qβ = activity product of mineral β dissolution reaction rβ = dissolution/precipitation rate per unit bulk volume of porous medium [mol/(m3·s)] Rmn = number of mineral reactions The reaction rates are normally stated at a reference temperature. The reaction-rate constant kβ at any desired temperature can be found using the following equation: 7

 E k   k 0, exp  a  R

 1 1      T T0 

(12)

where Ea is the activation energy for reaction (J/mol), k0,β is the reaction-rate constant for reaction at reference temperature, R is the universal gas constant (8.314 J/(mol K)), T is the temperature (K) and T 0 is a reference temperature (K). The reactive surface area of a mineral per unit bulk volume of a porous medium changes as the mineral dissolves or precipitates (Xu et al., 2003). Nghiem et al. (2004) employed the following equation to adjust the reactive surface area of a mineral: 



  0 .

N

(13)

N 0

where Â0 is the reactive surface area at time zero, Nβ are the moles of mineral β per grid block volume at current time, and Nβ0 are the moles of mineral β per grid block volume at time zero. Ion Exchange In a reservoir environment, there is a chemical equilibrium between ions in the solution and ions that sorb on the mineral surface, mainly on clay. As water with ion concentration different from the connate water ion concentration is injected, ion exchange reactions occur. In a LSW process, two typical ion exchange reactions that involve sodium, calcium and magnesium ions are:

Na  

1 Ca  X2   Na  X  1 Ca 2 2 2

(14)

1 1 Na   (Mg  X 2 )  ( Na  X)  Mg2 2 2

(15)

where X denotes the clay mineral in the reservoir rock. The above reaction is reversible and indicates that Na+ is taken up by the exchanger and Ca2+ and Mg2+ are released if we inject high salinity brine into a reservoir, while the reversible is true in LSW or there is a surplus of Ca2+ and Mg2+. The following selectivity coefficients are used to model ion exchanges in the clay surfaces:

K 'Na \ Ca

K 'Na \ Mg







  (Ca )



  (Mg )

( Na  X) m(Ca 2 )

0.5

Ca  X 2 0.5 m( Na  ) ( Na  X) m(Mg2 )

0.5

Mg  X 2 0.5 m( Na  )

2  0.5

 ( Na  )

(16)

2  0.5

 ( Na  )

(17)

where ( Na  X) , Ca  X2  , and Mg X2  are the equivalent fractions of Na+, Ca2+ and Mg2+ on the exchanger, respectively. The use of the selectivity coefficients in Equations (16) and (17) follows the 8

Gaines-Thomas convention (Gaines and Thomas, 1953). Note that the selectivity coefficients K 'Na \ Ca and K 'Na \ Mg are operational variables and not thermodynamic variables like the equilibrium constants. Modeling Wettability Alteration The effect of wettability alteration is modeled by shifting the relative permeability curves as schematically shown in Figure 1. Multiple relative permeability tables can be defined for a reservoir rock type, where each table corresponds to one value of a specified interpolant. Typically, two sets of relative permeability curves representing, respectively, low salinity and high salinity conditions are considered and an interpolation between these two curves is carried out. The interpolant is the equivalent fraction of an ion on the rock surfaces. Relative permeability curves are usually measured from laboratory experiments and serve as the input data for numerical simulation. It is shown in the coreflood simulation that the use of Ca++ equivalent fraction on the exchanger as the interpolant provides a good match of experimental data.

Figure 1: Schematic shifting of relative permeability curves in LSW 3. Model Validation This section presents the validation of the new model against (1) the geochemistry software PHREEQC and (2) laboratory measurements. The PHREEQC software and a series of North Sea sandstone and Texas’s reservoir coreflood experiments conducted by Fjelde et al. (2012) and Rivet (2009) served as the basis for this validation. The mechanistic simulations in this section include brine/rock interactions, dynamic single/multiple ion exchange, intra-aqueous reactions, mineral dissolution/precipitation, and CO2 solubility. 9

3.1 Comparison of Ion Exchange Model with PHREEQC Validation of the ion-exchange model in the GEM compositional simulator was carried out by comparing the results of the model with PHREEQC geochemistry software. PHREEQC is designed to perform a range of aqueous geochemical calculations and processes in natural-water and laboratory experiments. The program is based on the equilibrium chemistry of aqueous solutions interacting with minerals, gases, solid solution, exchangers, and sorption surfaces. It may be used to simulate 1D transport phenomenon with dispersion and diffusion. The base case for comparison of the ion exchange model developed in this work with that in PHREEQC is based on Appelo’s problem (Appelo, 1994). Either a low or high salinity brine slug was used to displace formation brine that initially contains sodium and/or calcium chloride in equilibrium with the cation exchanger (Figure 2). The aquifer was then flooded with brine that contains sodium chloride and calcium chloride at different concentrations.

Figure 2: Base case simulation for ion exchange validation The linear model in PHREEQC has 20 cells, and it requires that 20 solutions, numbered 1 through 20, be defined; the number of the solutions corresponds to the number of the cells in the model. In this validation, all cells contain the same solution, and the number of the exchangers corresponds to the number of the cells in a column. In PHREEQC, the ADVECTION data block needs to include the number of cells and the number of shifts for the simulation. The calculation only accounted for the number of pore volumes that flow through the cells with no explicit definition of time or distance. A one-dimensional linear homogeneous reservoir model that consists of 20 grid blocks with properties shown in Table 2 was used in this comparison. Since this is a displacement of a water slug, porosity was set at a value of 0.99. The ion exchange was allocated for all grid blocks with a constant value. This model involves some important geochemistry reactions during LSW. However, these geochemistry properties were not included in the base case simulation. We used the low salinity injected brine to displace formation water as shown in Table 3. The evolution between Ca2+ and Na+ in the effluent from GEM and PHREEQC are plotted in Figure 3. There is a very good agreement between GEM and PHREEQC. These results are also consistent with the analytical results by Appelo (1994) using mass balance over the front. Initially, the effluent is in equilibrium with the original exchanger composition, formation water is produced until 0.7 of injected pore volume, and then the fronts have a sharpening tendency because of a preference for the displacing ions.

10

Table 2: Basic Reservoir Properties for Base Case Parameter Grid blocks system Grid block sizes Horizontal permibilities Vertical permeabilities Porosity Ion exchange

Value for Base Case 20 x 1 x 1 ∆x = 3.66 m, ∆y = 30.48 m, ∆z = 15.24 m 2000 mD 2000 mD 0.99

CEC Selectivity coefficient

50 0.4 at 25oC (From Appelo, 1994)

1 2 1 Ca  Na _ X  Ca _ X 2  Na  2 2

Table 3: Brine Composition Used for Low Salinity Water Flooding Na+ 0.01326 1.326

Injected Brine (mol/l) Connate Water (mol/l)

Ca2+ 0.000148 0.148

Cl0.01622 1.622

The detailed exchange between Na+ and Ca2+ is shown in Figure 4. The equivalent fraction of Ca-X2 increases with a decrease in Na-X, which exactly follows the hypothesis of calcium absorption when low salinity brine is injected into the reservoir:

1 2 1 Ca  Na  X  Ca  X2  Na 2 2

(18)

1.4

1.2

Ca2+_GEM Ca2+_PHREEQC

Concentration (mol/l)

1

Na+_GEM Na+_PHREEQC 0.8

0.6

0.4

0.2

0 0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

Injected Pore Volume

Figure 3: Effluent evolution of Ca2+/Na+ during LSW - GEM vs. PHREEQC

11

Ca-X2 Na-X Figure 4: Evolution of the equivalent fractions of Ca-X2 and Na-X for various gridblocks The evolution of Ca2+ in the high and low divalent ions in formation water (Table 4) was also simulated and compared with PHREEQC. A very good agreement between GEM and PHREEQC on the effluent Ca2+ concentration after ion exchange (Figure 5) was obtained. Table 4: Connate Water Composition Used for Evaluating Calcium Ions Effects

High Calcium Connate Water (mol/l) Low Calcium Connate Water (mol/l)

Na+

Ca2+

Cl-

1.326

0.148

1.622

1.326

0.08

1.486

1.4

Ca2+_High Ca Connate_GEM Ca2+_High Ca Connate_PHREEQC Na+_High Ca Connate_GEM Na+_High Ca Connate_PHREEQC Ca_Low Ca Connate_GEM Ca_Low Ca Connate_PHREEQC Na_Low Ca Connate_GEM Na_Low Ca Connate_PHREEQC

Concentration (mol/l)

1.2

1

0.8

0.6

0.4

0.2

0 0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

Injected Pore Volume

Figure 5: Effluent evolution of Ca2+/Na+ for high and low divalent ion in formation water GEM vs. PHREEQC 12

3.2 Model Validation with North Sea Coreflood Experiment 3.2.1 Experimental setup Fjelde et al. (2012) conducted a typical LSW coreflood experiment and reported important data, such as formation/injection brine composition, relative-permeability curves for high- and low-salinity water floods, evolution of effluent divalent ions, average oil saturation, and pH. It is a good example, therefore, for the validation of the new LSW model. Comparing with these experimental results, not only the ion exchange but also the wettability alteration due to cation exchange and the effect of calcite dissolution could also be validated. The experimental setup is shown in Figure 6 and the compositions and properties of the synthetic formation water (FW), sea water (SW) and low-salinity water (LSW) used in the coreflood experiments are given in Table 5. The stock tank oil (STO) was pre-filtered through a 0.45μm oil filter before core flooding. The measured viscosity for the STO was 1.5 cP at 80°C. Core plugs from a sandstone oil reservoir in the North Sea were used in the experiments. The clay content in the rock was approximately 13% of the weight of the bulk sample. Important core properties are listed in Table 6. Reservoir core plugs were prepared with formation water (FW) and aged with crude oil at the initial water saturation (Swi). These core plugs were either flooded with FW, sea water (SW), or low-salinity water (LSW). The effluent samples were analyzed for ionic concentrations and pH. After aging, one composite core was flooded with FW, SW, and then LSW, using an injection rate of 6.323 cm3/d. For each of the brines, approximately 10 PVs were injected. Effluent samples were collected to determine oil production, ionic composition, and pH of the water phase.

Figure 6: Fjelde’s Experimental Setup (Modified from Fjelde et al., 2012)

13

Table 5: Brine composition used for coreflood experiments (Fjedel et al., 2012) Formation Water (FW) (moles/litre)

Sea Water (SW) (moles/litre)

Low Salinity Water flooding (LSW) (moles/litre)

Na+

1.32622

0.45011

0.01326

Ca++

0.14794

0.01299

0.00148

Cl

1.67773

0.52513

0.01661

HCO3-

0.000118

-

-

-

Table 6: Core properties used in Fjelde’s experiment Length of core (cm)

8

Porosity (fraction)

0.279

Absolutely permeability (mD)

150

3.2.2 1D North Sea Coreflooding Simulation A one-dimensional model has been setup in order to simulate the above coreflooding experiments. It consists of 40 gridblocks with the same properties as the sandstone core in Fjelde et al.’s experiment as indicated in Figure 7 and Table 7. In this model, we considered the reversible ion exchange between calcium and sodium as well as aqueous and mineral reactions. The reactions that are modeled are: CO2 (aq )  H2O  H  HCO3

(19a)

H  OH  H2O

(19b)

Calcite  H  Ca 2  HCO3

(19c)

1 2 1 Ca  Na  X  Ca  X2  Na  2 2

(19d)

The dissolution of calcite is one of the most important reactions in LSW. It provides a Ca2+ source for ion exchange that affects the reservoir wettability. Buckley et al. (1998) and Suijkerbuijk et al. (2012) indicated that the Ca2+ concentration is important in determining wettability and the relative permeability curves depend on formation brine composition. In the one-dimensional model, we used two relative permeability sets that were provided by Fjelde et al. (2012), as shown in Figure 8.

14

Figure 7: 1D Model for Validating Fjelde’s Coreflooding Experiment Table 7: Basic Reservoir Properties for Run 2 Grid blocks system Grid block sizes Horizontal permibilities Vertical Permeabilities (mD) Porosity Volume of Clay Injection Velocity

40 x 1 x 1 ∆x = 0.2 cm, ∆y = 2 cm, ∆z = 3.8 cm 150 mD 150 mD 0.279 0.13 6.323 cm/d

Figure 8: Relative Permeability Curves for 1D Model (Fjelde et al., 2012) 15

Dynamic Ion Exchange There are three stages of injection including formation water injection, sea water injection, and low salt water injection. Calcium ions are expected to exchange with sodium ions on the clay surface. We compare the evolution of effluent Ca2+ concentration from experimental measurements and from the simulator GEM. To examine the effect of ion exchange, two GEM runs were performed. In the first run, ion exchange was not included in the simulation. Figure 9 shows a good agreement between simulation and experimental results during the first injection stage with formation water. However, the simulated concentration of Ca2+ is lower than that in the experiments when the injected fluid switches from formation water into sea water and then to low salinity water. The calcium concentration was constant at the injected concentration value. This discrepancy is due to the absence of ion exchange consideration in different injected modes. A much better match for sea water and low salinity water injection stages could be obtained when the model includes the reversible ion exchange reactions, cation exchange capacity of clays, the amount of the clay in the rock, the concentrations of the cation, and the selectivity coefficients. Figure 10 shows an excellent match of the effluent calcium evolution between GEM with ion exchange and the experimental results. Figure 11 shows the dynamic ion-exchange process inside the core for the experiment described above. The equivalent fraction of Ca-X2 is constant during the first ten injected pore volumes of formation water because of the unvarying composition of the injected brine and formation water. The Ca2+ ions adsorb on the clay minerals during SW injection with a significant reduction of salinity concentration. It is important to note that the injected Ca2+ concentration in the Fjelde et al.’s experiment is extremely low, leading to the desorption of Ca2+ (decrease in Ca-X2 equivalent fraction) after the formation water had been flushed out at about eleven injected pore volumes. Ca++ Effluent - GEM vs Experiment (Fjelde) 0.16

0.14

0.12

Concentration (mol/l)

Fjelde_Experiment

GEM

0.1

0.08

0.06

0.04

0.02

0 0

5

10

15

20

25

30

35

Injected Pore Volume

Figure 9: Evolution of Ca2+ without Ion Exchange vs. Experimental Coreflooding (Fjelde et al., 2012)

16

Ca++ Effluent - GEM vs Experiment (Fjelde) 0.16

0.14

0.12

Concentration (mol/l)

Fjelde_Experiment GEM

0.1

0.08

0.06

0.04

0.02

0 0

5

10

15

20

25

30

35

Injected Pore Volume

Figure 10: Evolution of Ca2+ - GEM vs. Experimental Coreflooding (Fjelde et al., 2012)

a. Ca-X2

b. Na-X Figure 11: Evolution of Equivalent Fraction Ca-X2/Na-X inside Core Sample 17

pH Profile Without ion exchange and geochemical reactions, the simulation is not able to capture the change in pH during LSW. Figure 12 shows an excellent match of the effluent pH when using ion exchange coupled with geochemistry in LSW. The new model can match well the variance in the effluent pH that increases as the injected brine salinity decreases. Several researchers observed the same trend of an increase in the effluent pH during LSW (McGuire et al., 2005; Lager et al., 2006; Austad et al., 2010). McGuire (2005) and concluded that the local pH increase in LSW is relatively similar to the mechanisms in alkaline flooding and it may contribute to the additional oil recovery. However, a pH increase is not the main mechanism responsible for high oil recovery by LSW since pH is only 1 to1.5 units higher than the injected pH and the final pH value is much smaller than the one in alkaline flooding. The small increase in pH is doubtful to produce a sufficient amount of in-situ surfactant as McGuire et al.’s conclusion (RezaeiDout et al. 2009, Melberg, 2010). A local pH increase during the LSW process can be explained by the dissolution of calcite minerals at the expense of protons as shown by reaction (20c).

8.5

8

pH

7.5

7

6.5

Fjelde_Experiment GEM

6

5.5 0

5

10

15

20

25

30

35

Injected Pore Volume

Figure 12: pH - GEM vs. Experimental Coreflooding (Fjelde et al., 2012) Effects of Wettability Alteration Without ion exchange and its effect on wettability, the simulation cannot capture the LSW effect and leads to a difference between the average oil saturation prediction and measured values as shown in Figure 13. The experiment data clearly show a decrease in the average oil saturation with LSW. However, the average oil saturation in simulation is constant even when injection was changed from formation water injection to sea water injection and low salinity water injection. Since the concentration of Ca2+ in the connate water has a strong effect on the oil and water relative permeabilities (Dang et al. 2013., Suijkerbuijk et al. 2012., Rivet. 2009, Buckley et al. 1998), the equivalent fraction of Ca-X2 which represents the exchangeable amount of Ca2+ is used as the interpolant for the wettability alteration due to LSW effects. An excellent match of oil saturation can be obtained 18

with wettability alteration as a result of ion exchange and the two experimentally determined relative permeability sets from the work of Fjelde et al. (2012) for high/low salinity water flooding. The adsorption of Ca2+ leads to the wettability alteration towards more water wetness and is responsible for the additional oil recovery. The equivalent fraction of Ca-X2 increases as SW injection is switched to LSW. However, this change is relatively small with a minor effect on the oil recovery. This observation supports for our hypothesis: The adsorption of Ca2+ leads to a modification of relative permeabilities, resulting in a higher oil recovery. Oil Saturation - GEM vs Experiment (Fjede) 0.7

0.6

Fjelde_Experiment 0.5

Oil Saturation

GEM-Without Ion Exchange 0.4

0.3

0.2

0.1 0

5

10

15

20

25

30

35

Injected Pore Volume

Figure 13: Oil Recovery - GEM (without Modeling of Wettability Alteration) vs. Experimental Coreflooding (Fjelde et al., 2012) Most of the oil was produced at the initial time during FW injection. The oil saturation after FW injection is around 24.5% and it further decreased to 23.5% during sea water injection. By interpolating between the two relative permeability curves based on the net ion exchange, an excellent match of oil saturation was obtained (Figure 14). The final recovery factor (RF) after FW, SW, and LSW injections were shown in Figure 15. Noted that core samples were taken from a strong water wet sandstone reservoir. Therefore, the change in relative permeability is insignificant in this case.

19

Oil Saturation - GEM vs Experiment (Fjede) 0.7

0.6

Fjelde_Experiment GEM Oil Saturation

0.5

0.4

0.3

0.2

0.1 0

5

10

15

20

25

30

35

Injected Pore Volume

Figure 14: Average Oil Saturation - GEM vs. Experimental Coreflooding (Fjelde et al., 2012)

Figure 15: Final RF after FW, SW, and LSW Injection 20

3.3 Model Validation with Texas Sandstone Reservoir Core Flood Experiment 3.3.1 Experimental Setup Rivet (2009) presented a series of oil displacement corefloods with high- and low-salinity brine to study the effect of LSW on oil recovery, residual oil saturation, and relative permeability. Both Berea core and reservoir core plugs were used in his study. One set of experiments, with a sequence of high- and low-salinity water floods in a composite reservoir core, was selected to validate the LSW model. The core consisted of a stack of six core plugs, which were obtained from an unconsolidated sandstone oil reservoir. The core plugs were arranged in air permeability order such that air permeability decreased from the inlet towards the outlet. The plugs contained about 20 wt % clay according to an XRD analysis reported by the field operator. Properties of the core are given in Table 8. First, core samples were saturated with brine. After the pore volume was measured, temperature was elevated to the experiment temperature. The core was then flooded with at least five pore volumes of connate brine. Next, oil flood was conducted at constant injection pressure. Note that all of the oil samples were filtered with an OFITE filter press. The oil was forced through a Millipore HA, 47-mm, 0.45- micron nominal-pore-opening cellulose filter. Oil-saturated core was aged for 25 days at 85oC. After aging, experiment temperature was restored. Then different water flooding scenarios were conducted to evaluate the effects of brine composition on the oil recovery factor. Table 8: Rivet’s core property Plug Number

Porosity (%)

Permeability (mD)

Length (cm)

1-1

26.5

338

4.8

1-2

26.3

379

4.8

1-3

26.7

469

5.0

1-4

27.8

634

5.0

1-5

27.5

665

4.8

1-6

25.1

888

5.1

Two different brines were used in the water flood experiments. The Synthetic Reservoir Formation Brine (SRCFB)was used as the high-salinity brine, while the Synthetic Lake C Water (SLCW)was used as the low-salinity brine. Fluid properties are given in Table 9. Table 9: Fluid properties for Rivet’s coreflooding experiments SRCFB Concentration (g/L)

28.62 NaCl 0.65 KCl 2.71 CaCl2

SLCW Concentration (g/L)

0.7 NaCl 0.21 CaCl2

21

3.3.2 1D Texas Core flooding Simulation A one-dimensional model has been setup to simulate the above core-flood experiments. The model consists of 60 grid blocks, representing the six plugs in Rivet’s core sample with the reservoir properties shown in Table 10 and Figure 16. Two sets of relative permeability curves provided by Rivet (2009) were used for simulating the LSW effects (Figure 17). Table 10: Basic Reservoir Properties for Run 3 Grid blocks system Absolute permibilities

60 x 1 x 1 10*338 mD, 10*379 mD, 10*469 mD, 10*634 mD, 10*665 mD, 10*888 mD 10*0.265, 10*0.263, 10*0.267, 10*0.278, 10*0.275, 10*0.251 0.2 60.9 cm/d

Porosity Volume of Clay Injection Velocity

Figure 16: 1D Model for Validation with Rivet’s Coreflooding Experiment

Figure 17: Relative Permeability Curves for 1D Model (Rivet, 2009) 22

Figure 18 shows the average oil saturations from Rivet’s experiment and from the new LSW model for two different cases: (i) without ion exchange and (ii) with ion exchange and wettability alteration. The average oil saturation decreases from 0.6 to 0.235 by LSW. The LSW in Rivet’s experiment has a better effect on oil recovery in comparison with the one in Fjelde et al.’s experiment because of better injected brine composition. Figure 19 indicates that the ion exchange reaction supports the hypothesis that calcium ions adsorb on the rock surface, resulting in an increase in the oil relative permeability and a decrease in water relative permeability as experimentally shown by Rivet (2009). Therefore, the ion exchange during LSW contributes to the alteration of wettability towards more water wetness and thus to the additional oil recovery. With the new proposed LSW model, an excellent history match for both high salinity and low salinity water flood experiments from the work of Rivet (2009) could be achieved as shown in Figures 20 and 21. 0.7

LSW_Rivet's Experiment 0.6

Average Oil Saturation

LSW_GEM_60GB 0.5

LSW_GEM_No Ion Exchange 0.4

0.3

0.2

0.1

0 0

0.5

1

1.5

2

2.5

3

3.5

4

Injected Pore Volume

Figure 18: Average Oil Saturation - GEM vs. Experimental Coreflooding (Rivet, 2009)

Ca-X2 Na-X Figure 19: Evolution of Equivalent Fraction Ca-X2/Na-X inside Core Sample

23

0.7

0.6

Average Oil Saturation

HSW_Rivet's Experiment 0.5

HSW_GEM_60GB 0.4

0.3

0.2

0.1

0 0

0.5

1

1.5

2

2.5

3

3.5

4

Injected Pore Volume

Figure 20: Average Oil Saturation – HSW: GEM vs. Experimental Coreflooding (Rivet, 2009)

0.7

0.6

Average Oil Saturation

LSW_Rivet's Experiment 0.5

LSW_GEM_60GB 0.4

0.3

0.2

0.1

0 0

0.5

1

1.5

2

2.5

3

3.5

4

Injected Pore Volume

Figure 21: Average Oil Saturation – HSW: GEM vs. Experimental Coreflooding (Rivet, 2009) Figure 22 show the recover factors by LSW and HSW. In this experiment, LSW shows superior performance compared to HSW with a nearly 10% incremental recovery factor. Compared to Fjelde’s experiment, it is an important observation that LSW’s performance will strongly depend on the fluid properties and characterization of core samples. LSW shows much better performance in a mixed wet core (Rivet’s experiment), whereas only small incremental oil recovery was observed in Fjelde’s experiment using a strongly water wet core sample. In addition, the new mechanistic LSW modeling approach can adequately predict the performance of LSW in both scenarios.

24

Figure 22: Cumulative Oil Recovery – LSW vs. HSW The effluent pH profiles from the LSW model and Rivet’s experiment are shown in Figure 23. The pH was initially decreased from 7.1 to 6.3 because of calcite precipitation. As long as low salinity brine was continuously injected into the core sample, the pH effluent tended to increase due to the dissolution of the calcite and ion exchange effect. 9

8.5

8

7.5

pH

7

6.5

6

Rivet Experiment 5.5

GEM 5

4.5

4 0

0.5

1

1.5

2

2.5

3

3.5

4

Injected Pore Volume

Figure 23: Effluent pH History – GEM vs. Experimental Coreflood (Rivet, 2009)

25

4. LSW Field-Scale Modeling There are several key factors that affect the success of LSW at the field scale: (1) geology, (2) ion exchange and geochemistry, (3) relative permeability modification to reflect wettability alteration, and (4) injection scheme. Modeling and simulation processes must take these factors into account to achieve the best prediction of LSW performance. Since LSW is a process that depends critically on the reservoir geological characteristics, it is important that these key geological factors be incorporated in reservoir modeling and simulation. Multiple geological realizations (facies and clay distribution) are needed for history matching, optimization and uncertainty assessment of LSW. In practice, it would be inefficient, for example, if the geological modeling, reservoir simulation, history matching, and optimization tasks were processed separately. To overcome this, an objective of this study is to develop an innovative approach for effectively connecting these modeling steps in a closed-loop workflow. A comprehensive LSW modeling approach has been developed, as shown in Figure 24. First, the geological model is generated by considering all of the important factors, including facies, distribution and quantities of clay mineral, porosity and permeability. These geological properties are coupled with flow equations and mechanistic recovery mechanisms (e.g., ion exchange, geochemical reactions, and wettability alteration for LSW) in the reservoir simulator. The geological model and reservoir simulator are then automatically linked with a robust optimizer in a closed loop that allows fast, continuous geological model updating. As mentioned above, multiple geological realizations are required for LSW modeling and evaluation. By using simple scripts, the geological software can, under the command of the optimizer, geostatistically generate new geological realizations. This approach is powerful for large-scale LSW history matching when important data (well logs, core and fluid samples, and multiple relative permeabilities) is available, or when robust optimization is required for capturing geological uncertainty.

Figure 24: Closed-loop LSW Modeling

26

LSW has mainly been implemented in sandstone reservoirs where clay minerals act as ion exchangers. As discussed earlier, single- or multiple-ion exchanges between injected brine and clay minerals result in wettability alteration during LSW. In this case, clay minerals play a crucial role in LSW performance and it is, therefore, important to capture its effects in LSW prediction. Clay modeling has, however, rarely been taken into account in conventional reservoir-simulation frameworks. The content and distribution of clay minerals may significantly affect final field-scale recovery; therefore, it is essential that a new approach be developed which comprehensively considers clay this effects, particularly in the case of a geologically sensitive process such as LSW. To overcome the above limitations, a new feature for incorporating the effects of clay in LSW modeling is introduced in this study. The new technique, which models two important properties including clay distribution and cationexchange capacity, is implemented by combining petrophysical modeling, empirical correlations, and geostatistical methods. In this approach, generation of the structural model is also the starting point for the geological modeling process; however, we assign geologically consistent petrophysics information and add different facies characterizations to develop a more realistic reservoir model. It is important to note that there is a strong correlation between a grain size and clay content, as has been reported by several authors (e.g., Saner et al., 1996, Shahin et al., 2012). This information was incorporated in the model by assigning claydependent grain sizes to different facies where higher clay content results in a lowering of the mean grainsize distribution. Three main facies types that are important in LSW were defined including: (1) Finegrained Sandstone (FS), (2) Medium-grained Sandstone (MS), and (3) Coarse-grained Sandstone (CS). These facies will be used in the rest of this study. FS is a fine-grained sandstone with a mean grain-size of 80 µm. FS is used to simulate low-porosity, low-permeability sandstone with a very high clay content. CS is a coarse-grained sandstone with a mean grain-size of 250 µm. CS is associated with high-porosity, high-permeability sandstone with very low clay content. MS is a transitional facies, between FS and CS, with average values of porosity and permeability, and medium clay content. Clay minerals have a significant influence on effective reservoir porosity and permeability. The main types of clays are laminated, dispersed, and structural, and combinations thereof. Generally speaking, laminated clay replaces sand grains and pore space, dispersed clay replaces only pore space, and structural clay replaces only sand grains. In this research, the effect of dispersed clay on porosity is considered since it reduces the clean sandstone porosity фmax to effective reservoir porosity фe, as follows: Vcl = фmax - фe

(20)

where Vcl is the volume of the dispersed clay per bulk volume. Wettability alteration is caused mainly by ion exchange. For a given rock, the capacity for cationexchange is expressed as the cation-exchange capacity (CEC). CEC can be estimated based on a surface area, clay content, organic carbon, and pH. Seilsepour and Rashidi (2008) indicated that more than 50% of the variation in CEC can be explained by variation in the clay and/or organic carbon content of different rock samples. However, the percentage of organic carbon in the rock sample, which is not usually available for reservoir simulation, is relatively small (less than 0.5%) compared with the clay quantity. Thus a new correlation for the calculation of CEC based on clay percentage has been developed in this study, using the fifteen common reservoir rock samples: CEC [meq/kg] = 628.58 * (fraction of clay) + 48.863 27

(21)

The difference between the calculated CEC values for the new correlation and Breeuwsma et al.’s equation (1986) is acceptable for most of the rock samples. A new interpolant value, called the scaled-equivalent-fraction ion exchange, has been introduced to capture this phenomenon in LSW field-scale simulations. The scaled-equivalent-fraction ion exchange is computed as follows:



(22)

where  is the equivalent-fraction ion exchange, CECcal is the calculated cation-exchange capacity in each grid block, and CECmax is the maximum cation-exchange capacity of the active cells. Since the scaled-equivalent-fraction ion exchange is associated with the calculated CEC function, the relative permeability curves will be shifted based on both the equivalent-fraction ion exchange and the clay content in each grid block. 5. Field Application We extended our work to full-field-scale LSW simulation for the Brugge field, which is a sandstone reservoir in the North Sea. Data for field-scale modeling and simulation were provided by TNO (Netherlands Organization for Applied Scientific Research) for a benchmark project to test the efficiency of water flooding optimization and closed-loop reservoir management. Critical effects of clay mineral are captured throughout the modeling process. LSW was compared with conventional high salinity water flooding to determine the benefits of this process in the full-field scale. This section presents one of the first discussions on the LSW modeling at the full-field scale and provides a modeling and prediction tool for the oil and gas industry. The structure of the Brugge field consists of an E-W elongated half-dome with a large boundary fault at its northern edge and one internal fault, with a modest throw at an angle of approximately 20 degrees to the northern edge fault. The dimension of the field is about 10 km x 3 km. From top to bottom, the Brugge field consists of 9 layers of the 4 main formations: Schelde, Maas, Waal, and Schie. The Schelde formation corresponds to the top two layers (layers 1 and 2) of the simulation model, the Maas formation corresponds to layers 3, 4 and 5, and the Waal formation corresponds to layers 6, 7 and 8, and the Schie formation corresponds to layer 9. The original high-resolution model of the Brugge field consists of 20 million gridblocks. This high resolution model is upscaled to a simulation model with 44,550 active cells. The field is operated with 14 vertical producers and 16 vertical water injectors. To develop a realistic model, it is assigned geologically consistent petrophysics information and facies characterization. The reservoir consists of three facies: FS facies (fine-grained sandstone), MS facies (medium-grained sandstone), and CS facies (coarse-grained sandstone). The facies model is generated using the SIS technique for the 9 layers in the 4 main formations, based on their depositional environment and geological characterization, as shown in Table 11. Figure 25 shows the detailed facies distributions in these four formations. To model their geological characteristics, clay mineral was statistically populated in the three facies (CS, MS, and FS) using the Sequential Gaussian Simulation (SGS) technique, as indicated in Table 12. Again, it is important to note that there is a strong correlation between a grain size and clay content (Saner et al., 1996). This correlation was incorporated in the model by assigning clay-dependent grain sizes to the three facies such that the higher the clay content, the lower the mean grain-size distribution. In this 28

study, dispersed clay is defined as clay which replaces only pore space. Clay is distributed throughout the reservoir using a clay histogram and a clay variogram. The clay histogram is used to constrain the clay content within each facies, while the clay variogram helps distribute the clay mineral throughout the reservoir. Figure 26 shows the generated clay distribution in the Brugge Field model. A geological model of the Brugge field, imported using GOCAD™ software, served as the starting point for modeling the reservoir using GEM™, equipped with the new LSW model. The reservoir model consists of 60,048 active gridblocks including 139 gridblocks in the x direction, 48 gridblocks in the y direction, and 9 gridblocks in the z direction. All nine layers in the z direction honor the geological sequence in the Brugge field. Initial wettability is considered preferential mixed-wet (or weak oil-wet) based on the relative-permeability classification suggested by Honarpour et al. (1986). The injection composition and fluid property model are shown in Tables 13 and 14. Table 11: Facies petrophysical properties associated with Brugge Field model Formation/Facies

Coarse-Sand (CS)

Medium-Sand (MS)

Fine-Sand (FS)

Schelde

0.3

0.57

0.13

Maas

0.2

0.65

0.15

Waal

0.25

0.65

0.1

Schie

0.17

0.68

0.15

Table 12: Clay content associated with Brugge Field model Mean Grain Size

Formation/Facies

Clay Content

FS

0.25 – 0.34

80

MS

0.15 – 0.25

250

CS

0.0 – 0.15

500

(μm)

Table 13: Formation and Injected Brine Molality 2+

Ca Na+ H+ ClOH HCO3-

Formation 0.68178 1.32622 1.0E-07 1.67773 5.242E-6 0.1786

Injected Brine 0.068178 0.132622 1.0E-07 0.1661 1.705E-06 0.0282416

29

Table 14: Fluid Property Model Oil/Gas Components CO2 CH4 C3H8 FC6 FC10 FC15 FC20

Initial Oil Mole Fraction 0.01 0.19 0.13 0.17 0.3 0.15 0.05

Aqueous Components H+ Ca++ Na+ ClOHHCO3-

Schelde FM

Initial Aqueous Molality 1.0E-7 0.068178 0.132622 1.67773 5.242E-6 0.1786

Maas FM

Waal FM Schie FM Figure 25: Facies Distribution in Four Formations of the Brugge Field

30

Schelde FM

Maas FM

Waal FM Schie FM Figure 26: Clay Distribution in Four Formations of the Brugge Field 5.1 Secondary LSW Response In the first investigation, high salinity water flooding and LSW, separately applied for about one year following primary production, were considered the secondary recovery methods in the Brugge field. Figure 27 shows the performances of LSW and conventional high salinity water flooding in terms of the recovery factor. LSW gave about 6% incremental OOIP over high salinity water flooding. The simulation results, which show the benefits of LSW in the full-field scale, are similar to pilot tests conducted by operators in other oil fields. The additional oil recovery comes from favorable wettability alteration due to ion exchange and geochemical reactions, the main underlying mechanisms of LSW in sandstone reservoirs.

Figure 27: Oil Recovery by High and Low salinity Water flooding in Brugge Field 31

Figure 28 shows the oil saturation maps at the final stages of high salinity water flooding and LSW in two representative layers of the Maas and Wall formations, respectively. The regions with a significant difference in oil saturation are highlighted in red circles. This figure clearly shows that LSW has better performance on decreasing the remaining oil saturation compared with high salinity water flooding.

High Salinity Water Flooding

Low Salinity Water Flooding

Figure 28: Comparative Remaining Oil Saturation after High- and Low- Salinity Water Flooding LSW has the highest efficiency in terms of improved oil recovery (e.g., in layer 3 of the Mass formation) in relatively high dispersed-clay content regions of the MS facies. These regions have high ion-exchange capacity and the sufficient effective porosity and permeability, as indicated in Figure 29. Low-salinity brine can easily contact the clay mineral, promoting favorable ion exchange and wettability alteration, so mobile oil can easily flow to the producing wells without being impeded by high-claydensity barriers.

Scaled Equivalent Fraction Ion Exchange

Porosity Distribution

Figure 29: Scaled Equivalent Fraction Ion Exchange and Porosity Distribution, Maas FM Since LSW improves oil recovery by wettability alteration only, it has a difficulty producing oil from low-permeability zones with high-clay density in FS facies. In this case, injected brine will be bypassed into high permeability zones and will not contact the clay. These problems are clearly indicated in regions where LSW shows only small improvement in oil recovery compared with high salinity water flooding 32

(Figure 30). Although these regions contain high clay mineral in the top, oil recovery by LSW is very low due to extremely poor sweep efficiency. Furthermore, regions with very low clay content do not exhibit wettability alteration. In these regions, LSW acts similarly to conventional high salinity water flooding. From this, it is again worth noting that well-placement optimization is very important in LSW implementation.

High Salinity Water Flooding

Low Salinity Water Flooding

Figure 30: Poor Performance of LSW in High Density Clay Regions 5.2 Tertiary LSW Response A significant number of the producing sandstone reservoirs in the world have already been flooded using conventional high salinity water flooding in their secondary recovery stage. Thus it is necessary to investigate the benefits of LSW for tertiary recovery. To explore the potential of LSW for tertiary recovery, high salinity water flooding was injected in the Brugge field model for the first ten years and then LSW was used for the remainder of the operating life. The brine composition used in LSW for tertiary recovery is similar to the one used in the first test. Figure 31 shows that tertiary LSW had incremental oil recovery of 4.1% OOIP over the high salinity water flood baseline. It also had a higher oil production rate which could efficiently extend the production life of the water flooding project.

Figure 31: Comparative Results on RF by High and Low Salinity Water Flooding in Tertiary Mode 33

In the overall evaluation shown in Figure 32, secondary LSW is more effective than conventional high salinity water flooding and tertiary LSW in terms of timing and oil recovery. This is an important finding since LSW can be considered for the first stage of secondary recovery, instead of high salinity water flooding. It provides not only better oil recovery but also reduces the cost and complexity of the additional setup and operation required for new injected-water treatment in older platforms, which is currently a challenge for tertiary recovery development in offshore and deep-water reservoirs.

Figure 32: Comparative Results on Recovery Factor of Secondary LSW, Tertiary LSW, and High Salinity Water Flooding The effectiveness of EOR processes is related to the time they are implemented. Figure 33 shows the results of three LSW development scenarios, each with a different starting time: after 5 years, 10 years, and 15 years of high salinity water flooding. Generally, the results show that the sooner the EOR LSW process is started, the greater the benefits in terms of oil recovery. Brown water flooded reservoirs typically have water channels, which bypass injected water late in the production period, a problem which would significantly offset the benefits of LSW if it was delayed to tertiary recovery.

Figure 33: Comparative Results on Recovery Factor of Starting Time of LSW Project 34

6. Further Discussions for Future Development (1) Fines Migration in LSW: it has been widely agreed that wettability alteration is the main underlying mechanism of LSW and the benefits of LSW were observed even if there was no evidence for the existence of fines migration. As reported from the literature, fines migration is not a direct cause, but it might contribute to the LSW process in terms of conformance control. An abrupt decrease in a salinity gradient with a low contained calcium ion could mobilize the clay minerals, reducing the absolute permeability in watered-out zones and resulting in diversion of the injected water into an unswept zone. Bedrikovetsky and his co-workers (Santos and Bedrikovetsky, 2006; Bedrikovetsky et al., 2010; 2011; Zeinijahromi et al., 2011a; 2011b) have pointed out several advantages of taking the advantages of fines-migration-induced formation damage for improved water flooding, and suggested several approaches to capture this phenomenon in LSW modeling. Not only limited to pure LSW, Dang et al. (2014) also suggested to take the advantages of fines migration for improving the displacement efficiency of the hybrid low-salinity water-alternating-gas process. From a technical point of view, it is desired to incorporate fines migration to the current mechanistic LSW model presented in this study to efficiently quantify the effects of formation damage as well as the benefits of conformation control during the course of LSW. (2) LSW modeling for carbonate reservoirs: It is desired to extend the model presented in this paper to modeling LSW in carbonate reservoirs which are of current interest for the worldwide oil and gas industry. This will require consideration for carbonate surface charges, dissolution of anhydrite, adsorption of sulfate ions and carboxylic groups, as well as temperature effects on the geochemical reactions. Obviously, validation of the LSW model with experimental core flooding results will be required to prove the model’s performance in carbonate reservoirs.

7. Summary and Conclusions LSW is an important EOR technique as it shows more advantages than conventional chemical EOR methods in terms of chemical costs, environmental impact, and field process implementation. This paper presents a method for modeling this process with the following highlights: 





A comprehensive ion exchange model with geochemistry has been developed and coupled to the multi-phase multi-component flow equations in an equation-of-state compositional simulator. It can capture most of the important physical and chemical phenomena that occur in LSW, including intra-aqueous reactions, mineral dissolution/precipitation, and wettability alteration. The new model was successfully validated with (1) the ion-exchange model of the geochemistry software PHREEQC, (2) the low-salinity core-flood experiments for a North Sea reservoir, and (3) the low salinity and high salinity heterogeneous core-flood experiments for a Texas reservoir. Excellent agreements were obtained on the evolution of ions, experimental effluent ion concentrations, effluent pH and oil recovery. Modeling LSW at field scale requires a bridging of geology and engineering knowledge. The linking of geological modeling software, a reservoir simulator, and a robust optimizer provides the basis for an effective modeling process: generating multiple geologically driven realizations, updating the reservoir model, and performing history matching, sensitivity analysis, and optimization in a closed-loop reservoir management. 35



 

The critical effects of clay mineral (clay content and cation-exchange capacity) are incorporated in the new LSW model, and an effective wettability interpolant is proposed for capturing the effects of clay minerals on wettability alteration. Secondary and tertiary LSW produced 6% and 4.1% incremental OOIP over high salinity water flooding in the full-field scale, respectively. Simulation results also indicated that the sooner the LSW process is started, the more benefit can be obtained.

Acknowledgments The authors thank Dr. Vijay Shrivastava of Computer Modelling Group Ltd. for his valuable comments and assistance in this research. The authors also thank Paradigm for providing the GoCad™ software for this research. The compositional simulator GEM™ and the optimization software CMOST™ from Computer Modelling Group Ltd. were used in this paper. The authors also wish to thank TNO for providing the access to Brugge Field data. Nomenclature Symbol

Definition

Unit

ai

Activity of component i

A

Cross section

Aˆ 

Reactive surface area of mineral β

m2/m3

CEC

Cation exchange capacity

eq/m3

CECcal

Calculated cation exchange capacity

eq/m3

CECmax

Maximum cation exchange capacity

eq/m3

d

Depth

Ea

Activation energy

f

Fugacity

g

Gravity acceleration

Hi

Henry’s law constant of component i

k

Permeability

kr

Relative Permeability

Fraction

K’

Selectivity coefficient

-

k

Reaction rate constant of mineral reaction β

Molality cm2

m J/mol m/s2 mD

36

mol/m2·s

Keq

Chemical equilibrium constant

-

L

Length

mi

Molality of component i

na

Number of aqueous components

-

nct

Total number of components

-

nh

Number of hydrocarbon components

-

ng

Number of gaseous components

-

no

Number of oleic components

-

P

Pressure

qi

Injection/production rate of component i

Qβ

Activity product

rβ

Reaction rate of mineral β

mol/(m3·s)

R

Universal gas constant

Jmol-1K-1

Raq

Number of reactions in aqueous phase

-

Rmn

Number of mineral reactions

-

S

Saturation

T

Temperature

Tq

Molar transmissibility of phase q (q = o, g, w)

V

Gridblock bulk volume

Vcl

Clay volume

Volume Fraction

yiq

Mole fraction of component I in phase q (q = o, g, w)

Mole Fraction

∆

Difference operator

φ

Porosity

Volume Fraction

φmax

Clean sandstone porosity

Volume Fraction

φe

Effective porosity

Volume Fraction

i

Activity coefficient of component i

cm Molality

psi cc/min -

Fraction deg K m3

-

37

-

 k

Stoichiometric coefficient of a species for the Chemical reaction

~ 

Mass density

kg/m3



Reaction rate

mol/m3

µ

Viscosity



Ion exchange equivalent fraction

ψ

Constitutive equation

-

cp Fraction -

Reference 1. Al-Shakabi, E.W., Sepehrnoori, K., Delshad, M., Pope, G. 2015. A Novel Method to Model LowSalinity-Water Injection in Carbonate Oil Reservoirs. SPE Journal, 20(5):1154-1166. 2. Appelo, C.A.J. 1994. Cation and Proton Exchange, pH Variations, and Carbonate Reactions in a Freshening Aquifer. Water Resources Research Journal, 30(10):2893-2805. 3. Austad T., RezaeiDoust A., Puntervold T. 2010. Chemical Mechanism of Low Salinity Water Flooding in Sandstone Reservoir. Paper SPE 129767 presented at the SPE Improved Oil Recovery Symposium, Tulsa, 24-28 April. 4. Austad, T., Strand S., Madland, M., Puntervold T., Korsnes, R. 2008. Seawater in Chalk: An EOR and Compaction Fluid. SPE Reservoir Evaluation and Engineering, 11(4): 648-654. 5. Bedrikovetsky, P., Siqueira, F., Furtado, C., Souza, A. 2010. Modified Particle Detachment Model for Colloidal Transport in Porous Media. Transp Porous Med (2011) 86:353-383. 6. Bedrikovetsky, P., Zeinijahromi, A., Siqueira, F., Furtado, C., Souza, A. 2011. Particle Detachment under Velocity Alternation During Suspension Transport in Porous Media. Transp Porous Med (2012) 91:173-197. 7. Breeuwsma A., Wosten J.H.M., Vleeshouwer J.J., Slobbe V., Bouma J. 1986. Derivation of Land Qualities to Assess Environmental Problem from the Soil Survey. Soil Science Society of America Journal, 50, 186-190. 8. Buckley, J.S., Liu, Y., Monsterleet, S. 1998. Mechanisms of Wetting Alteration by Crude Oils. Paper SPE 37230 presented at the SPE International Symposium on Oilfield Chemistry, Houston, TX, USA, 18-21 February. 9. Buckley, J.S., Morrow, N.R. 2010. Improved Oil Recovery by Low Salinity Water flooding: A Mechanistic Review. Paper presented at 11th International Symposium on Evaluation of Wettability and Its Effect on Oil Recovery, Calgary, 6-9 September. 10. Dang T.Q.C., Nghiem, L.X., Chen, Z., Nguyen, Q.P., Nguyen, T.B.N. 2014e. CO2 Low Salinity Water Alternating Gas: A New Promising Approach for Enhanced Oil Recovery. Paper SPE 169071 presented at the 2014 SPE Improved Oil Recovery Symposium, Tulsa, OK, USA, 12-16 April. 11. Dang, T.Q.C., Nghiem, L.X., Chen, Z., Nguyen, T.B.N. 2013. State-of-the Art Smart Low Salinity Water flooding for Enhanced Oil Recovery. Paper SPE 165903 presented at the 2013 SPE Asia Pacific Oil & Gas Conference and Exhibition, Jakarta, Indonesia, 22-24 October. 12. Drummond C., Israelachvili J. 2002. Surface Forces and Wettability. Journal of Petroleum Science and Engineering, 33:123-133. 38

13. Drummond C., Israelachvili J. 2004. Fundamental Studies of Crude Oil-Surface Water Interactions and Its Relationship to Reservoir Wettability. Journal of Petroleum Science Engineering, 45:61-68. 14. Ehrlich R., Hasiba H.H., Raimondi P. 1974. Alkaline Water flooding for Wettability Alteration – Evaluating a Potential Field Application. Journal of Petroleum Technology, 1335-1343. 15. Evje S., Hiorth A. 2011. A Model for Interpretation of Brine Dependent Spontaneous Imbibition Experiments. Advances in Water Resources, 34(2011):1627-1642. 16. Fathi, S.J., Austad, T., Strand, S. 2010. “Smart Water” as a Wettablity Modifier in Chalk: The Effect of Salinity and Ionic Composition. Energy and Fuels, 24(4): 2514-2519. 17. Fjelde, I., Asen, S.V., Omekeh, A. 2012. Low Salinity Water Flooding Experiments and Interpretation by Simulations. Paper SPE 154142 presented at the Eighteenth SPE Improved Oil Recovery Symposium, Tulsa, OK, USA, April 14-18. 18. Hiorth, A., Cathles, L., Madland, M. 2010. The Impact of Pore Water Chemistry on Carbonate Surface Charge and Oil Wettability. Transport in Porous Media, 85(1): 1-21. 19. Jadhunandan, P.P., Morrow, N.R. 1995. Effect of Wettability on Waterflood Recovery for CrudeOil/Brine/Rock Systems. SPE Reservoir Engineering, 10(1):40-46. 20. Jensen J.A., Radke C.J. 1988. Chromatographic Transport of Alkaline Buffers Through Reservoir Rock. SPERE, 3(3):849-856. 21. Jerauld, G.R., Lin, C.Y., Webb, K.J., Seccombe, J.C. 2008. Modeling Low Salinity Water flooding. SPE Reservoir Engineering, 11(6):1000-1012. 22. Kharaka, Y.K., Gunter, W.D., Aggarwal, P.K., Perkins, E.H., DeBraal, J.D. 1988. SOLMINEQ.88: A Computer Program for Geochemical Modeling of Water-Rock Interactions. US Geological Survey, Water Resources Investigation Report 88-4227, Menlo Park, CA. 23. Korrani A.K.N., Sepehrnoori K., Delshad M. 2013. A Novel Mechanistic Approach for Modeling Low Salinity Water Injection. Paper SPE 166523 presented at the SPE ATCE, New Orleans, Louisiana, USA, 30 September – 2 October. 24. Kulkarni M.M., Rao D.N. 2005. Experimental Investigation of Miscible and Immiscible WAG Process Performance. Journal of Petroleum Science and Engineering, 48(2005):1-20. 25. Kumar, M., Fogden A., Morrow, N.R., Buckley, J.S. 2010. Mechanisms of Improved Oil Recovery from Sandstone by Low Salinity Flooding. Paper SCA 2010-25 presented at the 24th International Symposium of Core Analysts, Halifax, Canada, 4-7 October. 26. Lager, A., Webb, K.J., Black, C.J.J. 2007. Impact of Brine Chemistry on Oil Recovery. Paper A24 presented at the 14th EAGE Symposium on Improved Oil Recovery, Cairo, 22-24 April. 27. Lager, A., Webb, K.J., Black, C.J.J., Singleton, M., Sorbie, K.S. 2006. Low-Salinity Oil Recovery – An Experimental Investigation. Paper SCA 2006-36 presented at the International Symposium of the Society of Core Analysts, Trondheim, Norway, 12-16 September. 28. Lager, A., Webb, K.J., Black, C.J.J., Singleton, M., Sorbie, K.S. 2008a. Low-Salinity Oil Recovery – An Experimental Investigation. Petrophysics 49(1):28-35. 29. Lager, A., Webb, K.J., Collins, I.R., Richmond, D.M. 2008b. LoSalTM Enhanced Oil Recovery: Evidence of Enhanced Oil Recovery at the Reservoir Scale. Paper SPE 113976 presented at the SPE/DOE Improved Oil Recovery Symposium, Tulsa, 19-23 April. 30. Loahardjo, N., Xie, X., Morrow, N.R. 2010. Oil Recovery by Sequential Water flooding of Mixed-Wet Sandstone and Limestone. Energy & Fuels, 24(9):5073-5080. 31. McGuire, P.L., Chatam, J.R., Paskvan, F.K., Sommer, D.M., Carini, F.H. 2005. Low Salinity Oil Recovery: An Exciting New EOR Opportunity for Alaska’s North Slope. Paper SPE 93903 presented at the SPE Western Regional Meeting, Irvine, CA, USA, 30 March-1 April. 39

32. Morrow, N.R., Bukley, J.S. 2011. Improved Oil Recovery by Low-Salinity Water flooding. JPT, Distinguished Author Series, 106-112. 33. Morrow, N.R., Tang, G.Q., Valat, M., Xie, X. 1998. Prospects of Improved Oil Recovery Related to Wettability and Brine Composition. Journal of Petroleum and Engineering, 20(3-4):267-276. 34. Nasralla, R.A., Sergienko, E., Masalmeh, S.K., Van de Linde, H.A., Brussee, N.J., Mahani, H., Suijkerbuijk, B., Alqarshubi, I. 2014. Demonstrating the Potential of Low-Salinity Warerflood to Improved Oil Recovery in Carbonate Reservoirs by Qualitative Coreflood. Paper SPE 172010 presented at the Abu Dhabi International Petroleum Exhibition and Conference, Abu Dhabi, UAE, 10-13 November. 35. Nghiem, L.X., Sammon, P., Grabenstetter, J., Ohkuma, H. 2004. Modeling CO2 Storage in Aquifers with Fully-Coupled Geochemical EOS Compositional Simulator. Paper SPE 89474 presented at the SPE Fourteenth Symposium on Improved Oil Recovery, Tulsa, OK, USA, April 17-21. 36. Nghiem, L.X., Shrivastava, V., Kohse, B. 2011. Modeling Aqueous Phase Behavior and Chemical Reactions in Compositional Simulation. Paper SPE 141417 SPE Reservoir Simulation Symposium, The Woodlands, TX, USA, February 21-23. 37. Omekeh A., Friis H.A., Fjelde I., Evje S. 2012 Modeling of Ion-Exchange and Solubility in Low Salinity Water flooding. Paper SPE 154144 presented at the Eighteenth SPE IOR Symposium, Tulsa, OK, 14-18 April. 38. Qiao, C., Johns, R.T., Li, L., Xu, J. 2015a. Modeling Low Salinity Water flooding in Mineralogically Different Carbonates. Paper SPE 175018 presented at the SPE Annual Technical Conference and Exhibition, Houston, TX, USA, 28-30 September. 39. Qiao, C., Li, L., Johns, R.T., Xu, J. 2015b. A Mechanistic Model for Wettability Alteration by Chemical Tuned Water flooding in Carbonate Reservoirs. SPE Journal, 20(4):767-783. 40. Rivet, S.M. 2009. Coreflooding Oil Displacements with Low Salinity Brine. Master Thesis, University of Texas at Austin. 41. Robertson E.P. 2010. Oil Recovery Increases by Low Salinity Flooding: Minnelusa and Green River Formations. Paper SPE 132154 presented at the SPE Annual Technical Conference and Exhibition, Florence, Italy, 19-22 September. 42. Rueslatten, H.G., Hjelmeland, O., Selle, O.M. 2008. Wettability of Reservoir Rocks and The Influence of Organo-metallic compounds: North Sea Oil and Gas Reservoir. 3:317-324. 43. Saner S., Cagatay M.N., Sanounah A.M. 1996. Relationships between Shale Content and Grainsize Parameters in the Safaniya Sandstone Reservoir, NE Saudi Arabia. Journal of Petroleum Geology, 19:305-320. 44. Santos A., Bedrikovetsky, P. A Stochastic Model for Particulate Suspension Flow in Porous Media. Transp Porous Med (2006) 62:23-53. 45. Seccombe, J., Lager, A., Jerauld, G.R., Jhavier, B., Buikema, T., Bassler, S., Denis, J., Webb, K., Cockin, A., Fueg, E. 2010. Demonstration of Low Salinity EOR at Interwell Scale, Endicott Field, Alaska. Paper SPE 129692 presented at SPE/DOE Improved Oil Recovery Symposium, Tulsa, 24-28 April. 46. Seilsepour M., Rashidi M. 2008. Prediction of Soil Cation Exchange Capacity Based on Some Soil Physical and Chemical Properties. World Applied Science Journal 3(2):200-205. 47. Shahin A., Tatham R., Stoffa P., Spikes K. 2011. Optimal Dynamic Rock-fluid Physics Template Validated by Petroelastic Reservoir Modeling. Geophysics, 76(6):O45-O58.

40

48. Skrettingland, K., Holt, T., Tweheyo, M.T., Skjevrak. 2010. Snorre Low Salinity Water Injection – Core Flooding Experiments and Single Well Field Pilot. Paper SPE 129877 presented at the SPE Improved Oil Recovery Symposium, Tulsa, OK, USA, 24-28 April. 49. Sorbie K.S., Collins I.R. 2010. A Proposed Pore-Scale Mechanism for How Low Salinity Water flooding Works. Paper SPE 129833 presented at the SPE IOR, Tulsa, OK, USA, 24-28 April. 50. Suijkerbuijk, B.M.J.M., Hofman, J.P., Ligthelm, D.J., Romanuka, J., Brussee, H.A., Marcelis, A.H.M. 2012. Fundamental Investigation into Wettability and Low Salinity Water flooding by Parameter Isolation. Paper SPE 154204 presented at the Eighteenth SPE Improved Oil Recovery Symposium, Tulsa, OK, USA, April 14-18. 51. Tang G.Q., Morrow N.R. 1999b. Oil Recovery by Water flooding and Imbibition – Invading Brine Cation Valency and Salinity. Paper SCA-9911 presented at the 1999 International Symposium of the Society of Core Analysis, Golden, Colorado, USA, 1-4 August. 52. Tang, G.Q., Morrow, N.R. 1997. Salinity, Temperature, Oil Composition and Oil Recovery by Water flooding. SPE Reservoir Engineering, 12(4):269-276. 53. Tang, G.Q., Morrow, N.R. 1999a. Influence of Brine Composition and Fines Migration on Crude Oil/Brine/Rock Interactions and Oil Recovery. Journal of Petroleum Science and Engineering, 24(2-4):99-111. 54. Thyne, G., Gamage, P. 2011. Evaluation of the Effect of Low Salinity Water flooding for 26 Fields in Wyoming. Paper SPE 147410 presented at the SPE Annual Technical Conference and Exhibition, Denver, Colorado, USA, 30 October-2 November. 55. Vledder, P., Foncesca, J.C., Wells, T., Gonzalez, I., Ligthelm, D. 2010. Low Salinity Water flooding: Proof of Wettability Alteration on a Field Wide Scale. Paper SPE 129564 presented at the SPE Improved Oil Recovery Symposium, Tulsa, OK, USA, 24-28 April. 56. Webb, K.J., Black, C.J.J., Al-Ajeel. 2004. Low Salinity Oil Recovery – Log-Inject-Log. Paper SPE 89379 presented at the SPE/DOE Symposium on Improved Oil Recovery, Tulsa, 17-21 April. 57. Webb, K.J., Black, C.J.J., Edmonds, I.J. 2005. Low Salinity Oil Recovery: The Role of Reservoir Condition Corefloods. Paper C18 presented at the 13th EAGE Symposium on Improved Oil Recovery, Budapest, Hungary, 25-27 April. 58. Webb, K.J., Lager, A., Black, C.J.J. 2008. Comparison of High/Low Salinity Water/Oil Relative Permeability. Presented at the International Symposium of the Society of Core Analysts, Abu Dhabi, UAE, 29 October – 2 November. 59. Xu T., Apps J.A., Pruess K. 2003. Reactive Geochemical Transport Simulation to Study Mineral Trapping for CO2 Disposal in Deep Arenaceous Formations. Journal of Geophysical Research, 108(B2, 2071) 60. Yildiz, H.O., Morrow, N.R. 1996. Effect of Brine Composition on Recovery of Moutray Crude Oil by Water flooding. Journal of Petroleum Science and Engineering, 14:159-168. 61. Zeinijahromi, A., Lemon, P., Bedrikovetsky, P. 2011a. Effects of Induced Fines Migration on Water Cut during Water flooding. Journal of Petroleum Science and Engineering, 78(2011): 609617. 62. Zeinijahromi, A., Nguyen, P., Bedrikovetsky, P. 2011b. Taking Advantage of Fines-MigrationInduced Formation Damage for Improved Water flooding (Reservoir Simulation using Polymer Flood Option). Paper SPE 144009 presented at the SPE European Formation Damage Conference, Noordwijk, The Netherlands, 7-10 June.

41

63. Zekri A.Y., Nasr M.S., Al-Arabai Z. 2011. Effect of LoSal on Wettability and Oil Recovery of Carbonate and Sandstone Formation. Paper SPE 14131 presented at the International Petroleum Technology Conference, Bangkok, Thailand, 7-9 February. 64. Zhang, Y., Morrow N.R. 2006a. Comparison of Secondary and Tertiary Recovery with Change in Injection Brine Composition for Crude Oil/ Sandstone Combinations. Paper SPE 99757 presented at the SPE/DOE Symposium on Improved Oil Recovery, Tulsa, 22-26 April. 65. Zhang, P., Tweheyo, M.T., Austad, T. 2006b. Wettability Alteration and Improved Oil Recovery in Chalk: The Effect of Calcium in the Presence of Sulfate. Energy and Fuels 20(5): 2056-2062. 66. Zhang, P., Tweheyo, M.T., Austad, T. 2007b. Wettability Alteration and Improved Oil Recovery by Spontaneous Imbibition of Seawater into Chalk: Impact of the Potential Determining Ions Ca2+, Mg2+, and SO42-. Colloids and Surfaces A: Physicochemical and Engineering Aspects 301(1):199-208. 67. Zhang, Y., Xie, X., Morrow, N.R. 2007a. Waterflood Performance by Injection of Brine with Different Salinity for Reservoir Cores. Paper SPE 109849 presented at the SPE Annual Technical Conference and Exhibition, Anaheim, CA, USA, 11-14 November.

Research Highlights: In the past, waterflooding was designed largely without regard for the composition of the injected brine. LSW is an emerging EOR technique in which the salinity of the injected water is controlled to improve oil recovery. Corefloods and other tests indicated that changes in the injected brine composition can improve basic waterflood performance by 5% to 40% of the original oil in place (OOIP), thereby suggesting that perhaps brine composition could be varied to optimize waterflood recovery. Despite significant growing interest in LSW, a consistent mechanistic study has not yet emerged. This paper contributes to solve the current technical issues related to LSW modeling as below: 

Understanding the underlying mechanisms from a wide range results in the past.



Solving the challenges in modeling and numerical simulation.



Interpreting and calibrating between modeling and laboratory experiments.



Solving the challenges in capturing the critical effects of geology.



Field-scale implementation and evaluation.

42

43