Effect of injection brine composition on wettability and oil recovery in sandstone reservoirs

Fuel 182 (2016) 687–695

Contents lists available at ScienceDirect

Fuel journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Effect of injection brine composition on wettability and oil recovery in sandstone reservoirs Zhao Hua a,b,⇑, Mingyuan Li c, Xiaoxiao Ni c, Haoyi Wang c, Zihao Yang c, Meiqin Lin c a

CNOOC Research Institute, Beijing 100028, China Postdoctoral Research Station, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China c Research Institute of Enhanced Oil Recovery, China University of Petroleum, Beijing 102249, China b

h i g h l i g h t s 2+


Bridging action of Ca

results in more oil-wet surface and limited oil recovery.


Wettability alteration and oil layer instability are the mechanism to EOR by LSF.
LSF effect was investigated by surface force model in a cylindrical pore.

a r t i c l e

i n f o

Article history: Received 8 March 2016 Received in revised form 31 May 2016 Accepted 2 June 2016

Keywords: Low salinity waterflooding Ion composition Wettability alteration Oil recovery Surface force

a b s t r a c t Wettability alteration and enhanced oil recovery (EOR) of sandstone by low salinity waterflooding (LSF) have become the focus of many studies. However, there is no clear explanation of how the ion composition of brines influences the LSF effect. In this paper, imbibition, f-potential measurements, core displacement, and microscopic displacement experiments were adopted to investigate the effect of brine composition on the wettability and oil recovery of sandstone reservoirs and the Ca2+ bridge action between rock surface and the acidic components in crude oil. Besides, the surface force model in a cylindrical pore was also developed for investigating the mechanisms behind wettability alteration and EOR by LSF. The results show that Ca2+ in the brine acted like a bridge between the negatively charged sites on rock surface and dissociated acidic components in crude oil, and the acidic components in crude oil played an important role in the Ca2+ bridge action. Divalent cations (e.g. Ca2+) in the brine hindered the low salinity effect, and thus the low-salinity brine without divalent cations showed optimal low salinity effect. Lowering the salinity of 1:1 electrolyte solution (e.g., NaCl solution) caused the expansion of the electrical double layer and increasing disjoining pressure between rock/brine and oil/brine interfaces, and thus the rock surface changed to be more water-wet and a thicker wetting-water film could be stable between rock surface and crude oil. When injecting low salinity NaCl brine, more water-wet state improves the ability of water imbibition in small pore and a thicker wetting-water film promotes destabilization of oil layers adhering to sandstone surface, thus resulting in EOR. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction In the recent years, extensive laboratory and field tests have showed that LSF can act as an attractive enhanced oil recovery technology [1–6]. The popularity of this technology is mainly because of its efficiency in displacing crude oil, low investment, easy operation, ease of injection, and environmental protection, all of which bring economic benefits compared to other chemical

⇑ Corresponding author at: CNOOC Research Institute, Beijing 100028, China. E-mail address: [email protected] (Z. Hua). http://dx.doi.org/10.1016/j.fuel.2016.06.009 0016-2361/Ó 2016 Elsevier Ltd. All rights reserved.

EOR methods [7]. In most cases, the additional oil recovery from LSF was in the range of 5–20% of OOIP [7–9]. Over the last decades, the wettability alteration has been widely believed to be an important contributor to LSF [1,5,7,10]. Wettability alteration of sandstone rocks by LSF is related to the minerals on sandstone surface, the polar components in crude oil and the ion composition of injected water [8,11,12]. Many studies have been carried out on the wettability alteration of smooth mineral substrates (of quartz [10,13], glass [11], or mica [3,14]). The alteration was characterized by monitoring the bound oil residues and the macroscopic contact angle on the substrate in brines. High pH and low-salinity solutions can result in desorption of oil layer on

688

Z. Hua et al. / Fuel 182 (2016) 687–695

substrates, and thus alter the wet state from oil-wet or neutral-wet to water-wet. Divalent cations (e.g. Ca2+) in the aqueous solutions act like a bridge between the negatively charged sites on substrate surface and negatively charged polar groups in crude oil, opposing desorption. From the field observation, LSF in the Omar field in Syria results in a more water-wet rock surface, which leads to an incremental recovery of 10–15% of the Stock Tank Oil Initially In Place (STOIIP) [1]. Low-salinity water expanding the electrical double layer was regarded as the main reason behind wettability alteration and EOR by LSF [3,10]. Lowering the salinity of injected brine, especially reduction of the multivalent cations, changes the electrical charges at both brine/rock and oil/brine interfaces to strongly negative, which causes higher repulsive forces between two interfaces, and as a result, the wet state is changed towards more water-wet and oil recovery is improved. Lager et al. [6] proposed that multicomponent ionic exchange (MIE) was responsible for the wettability alteration and EOR. MIE would allow for desorption of organic polar compounds and organo-metallic complexes from the rock surfaces and replacing them with cations in injection brines, thus resulting in more water-wet surfaces and an increase in oil recovery. Austad et al. [2] explained that a local increase in pH during LSF led to water-wet state and EOR which could be related to alkaline flooding. However, the theory based on salting-in effect has been proposed that lowering the salinity of brines could enhance the water structure around oil and thus increased the solubility of oil in water [8,15]; therefore the wettability changed to be more water-wet and oil recovery was enhanced during LSF. Wettability alteration and enhanced oil recovery of sandstone reservoirs by LSF have become the focus of many studies. However, the LSF effect for different ion compositions and how the ion composition of low-salinity brines influences wettability and oil recovery were barely studied in previous literature. In this paper, the effect of brine composition on wettability and oil recovery and the mechanism behind the effect in sandstone reservoirs were investigated. Analyzing the effect and understanding the underlying mechanism enable the optimization of injection brine composition for EOR.

Table 2 Properties of brines. Brine LS-1 LS-2 LS-3 LS-4 LS-5 LS-6 HS-1 HS-2 HS-3

CaCl2 (mol L1)

NaCl (mol L1)

Salinity (ppm)

Ionic strength (mol kg1)

rOW

0 0 0 0 0 0 0.017 0.573 1.14

0 0.05 0.1 0.5 1 1.72 0 0 0

0 2925 5850 29,250 58,500 100,620 1887 63,603 126,540

0 0.05 0.1 0.5 1 1.72 0.05 1.72 3.42

22.5 21 17 16.5 13 10.5 18.0 16.0 13.0

(mN m1)

which are listed below in Table 2. All the reagents used in the study, such as NaCl, CaCl2, MgCl2, K2SO4 and FeCl3, were of analytical grade. 2.1.3. Cemented quartz cores Based on the technique of X-ray diffraction, the mineral contents of Tarim oilfield natural core were investigated. The result shows that the natural core is mainly composed of quartz and the quartz content is about 85%. Thus, in this study, the cemented quartz cores were used for imbibition and displacement experiments, and crushed for zeta potentials measurements. The artificial sandstone cores are cylindrical and their key parameters are presented in Table 3. In Table 3, the initial oil saturation (So) is percent content of crude oil in each core when the irreducible water saturation (Sw) is established, that is, So = 1  Sw. 2.2. Zeta potential measurement DelsaTM Nano analyzer was used to measure the zeta potential of crude oil/brine and rock/brine interfaces at 45 °C. The oil/brine emulsions were prepared at volume ratio of 1:100, and the rock/ brine suspensions were prepared by adding 0.2 wt% of crushed core powder to the brine. 2.3. Displacement and imbibition experiments

2. Experimental section 2.1. Materials 2.1.1. Crude oil and mixing oil The crude oil, obtained from Tarim oilfield, has a viscosity of 3.8 mPas (at 45 °C), a density of 0.829 g cm3 (at 45 °C), API of 35.76, a n-C7 asphaltene content of 0.54%, a resin content of 3.80%, and an acid number of 0.4 mg KOH/g oil. Kerosene added to white oil at a 1:3 ratio by volume is as mixing oil. White oil and kerosene were supplied by Sinopec Corp. In order to remove impurities with interfacial activity, white oil and kerosene were treated by silica gel adsorption for 60 h under room temperature before use. 2.1.2. Brines The ion composition of Tarim oilfield simulated formation water (SFW) is described in Table 1. Thus the ionic strength of SFW is 1.72 mol kg1. Other brines were prepared by dissolving desired amounts of salts in deionized water, the properties of

The cores were first dried at 105 °C to constant weight, and were then saturated with SFW under vacuum. To calculate the absolute permeability, SFW was injected at a rate of 0.10 mL/min at 45 °C and the pressure drop was monitored. The cores were flooded and saturated with crude oil and the initial water saturation (irreducible water saturation, Sw) was established. The cores were then aged at 45 °C for sixty hours in the core holder. After aging, imbibition and displacement experiments were performed at 45 °C with prepared brines. The displacement experiment was performed under the confining pressure of 10 MPa using a flow of 0.10 mL/min to simulate when the water flooding occurred. During waterflooding, the pressure drop across the cores was measured by a differential pressure transducer. 2.4. Microscopic displacement experiment The microscopic displacement experiment was performed at room temperature as described in Ref. [16]. The experimental

Table 1 Ion composition of Tarim model formation water. Ion Mass concentration/mg L

1

Ca2+

Fe3+

K+

Mg2+

Na+

SO2 4

Cl

5160.4

1.76

323.7

436.6

28983.9

398.3

55190.8

Z. Hua et al. / Fuel 182 (2016) 687–695 Table 3 Key parameters of artificial sandstone cores. Sandstone core

Length l (cm)

Diameter d (cm)

Porosity u (%)

Permeability K (103 lm2)

Initial oil saturation So (%)

R-1 R-2 R-3 R-4 R-5 R-6 R-7 R-8

7.07 7.04 7.04 7.08 7.05 7.08 7.06 7.07

2.51 2.51 2.51 2.51 2.51 2.51 2.51 2.51

21.03 22.72 22.91 20.70 20.63 20.84 22.90 20.91

9.40 12.72 7.77 14.35 13.27 10.36 10.46 8.80

59.81 59.12 60.78 67.59 66.86 66.15 61.05 61.59

setup consisted of a light microscope, a micro-visual glass model and a high-accuracy low-flow-rate Model 355 micro syringe pump which is used to control the injection flow rate of bines. A sketch of the flooding setup is shown in Fig. 1. In order to create a synthetic porous medium, first a pattern of a porous medium was designed according to the microscopic pore-throat structure in casting sheet image of a natural core, and then etched on a glass plate. Before

689

experiment, the micro-visual glass model was cleaned with petroleum ether, ethanol, and deionized water in turn. The micro-visual model was initially flooded with SFW to establish a water-wet surface, and then saturated with crude oil to establish the initial water saturation. Then brines were pumped into the micro-visual model at rate of 2.5 lL/min until no oil was produced. The displacements occurring in the micro-visual model at a magnification up to 20 times were recorded continuously with a light microscope. 3. Results and discussion 3.1. Polar components identified in crude oil The polar components in crude oil were identified using the negative electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI FT-ICR MS). All procedures for the negative ESI analysis by FT-ICR MS have been described in detail elsewhere [17,18]. The m/z value of crude oil between 200 and 780 with distribution peak centered at about m/z 400 are shown in Fig. 2. The inset in Fig. 2 is the close-up view of the

Fig. 1. Experimental device for microscopic visual displacement experiment.

Fig. 2. Broadband negative-ion ESI FT-ICR mass spectra of Tarim crude oil.

Z. Hua et al. / Fuel 182 (2016) 687–695

expanded mass spectrum obtained at m/z 393 and m/z 396. The analysis results show that polar components in crude oil are mainly naphthenic acids. 3.2. Influence of brine composition on rock wettability Brine imbibition was used to investigate the effect of bine composition (i.e., ion composition and salinity) on the wettability of sandstone surface. The core plugs, initially saturated with SFW and aged with crude oil, were subsequently imbibed with NaCl and CaCl2 solutions with different salinities, as shown in Table 4. As shown in Fig. 3, the oil recovery and imbibition rate for the six cores are obviously different. The oil recovery after imbibing for 72 h (Rf) and after imbibing for 60 days (Rs) were collected, as given in Table 4 where the oil recovery was negligible for cores 5 and 6 during imbibition. After imbibing in NaCl solutions for 60 days, the oil recovery increased from 4.2% to 8.8% when the salinity of NaCl solution was decreased from 100,620 ppm to 2925 ppm, and the oil recovery was up to 15.9% for deionized water. However, after imbibing in CaCl2 solutions for 60 days, low oil recovery of only 3.1% and about 0% were obtained when the salinity of CaCl2 solution were 1887 ppm and 63,603 ppm, respectively, which demonstrates that for the CaCl2 solution, the imbibition weakened obviously even in low salinity and did not occur nearly when the salinity was high enough. According to the comparison of imbibition in different brines, the oil recovery and imbibition rate in sodium chloride brines were obviously higher than those in calcium chloride brines with the same ionic strength. The results exhibit that the low salinity aqueous solutions without Ca2+ can obviously alter the sandstone surface to be more water-wet, while Ca2+, even at low salinity, can notably

weaken the above-mentioned effect of low salinity aqueous solutions. A possible explanation of this observation assumes that the decreased salinity of NaCl solution causes the expansion of the electrical double layer and greater repulsive forces between oil/ brine and rock/brine interfaces with negative electric charges, and thus low-salinity brine creates a water-wet state which is favorable for water imbibition and improving oil recovery. While the Ca2+ in brines can attract negatively charged sites on the rock surface and negatively charged polar groups at oil-water interface by electrostatic attraction, acting like a ‘‘bridge” between the rock surface and the oil-water interface, and as a consequence, the rock surface becomes more oil-wet. The higher the salinity of Ca2+ is, the more oil-wet the rock becomes, when the absorption of Ca2+ on rock surface does not achieve the level of saturation. These suggestions above are in accordance with the zeta potentials of rock/brine and oil/brine interfaces which are shown in Figs. 4 and 5, respectively. The marker in these figures expresses the standard deviation of the zeta potentials in three runs. The results demonstrate the influence of ion composition and salinity on zeta potentials at interfaces. High negative charges at rock/ deionized water and oil/deionized water interfaces are caused by dissociation of silanol groups on rock surface and carboxyl groups

20

Table 4 Effect of ion composition and salinity of imbibing brine on the oil recovery during imbibition. Core

Imbibition brine

Rf (%)

Rs (%)

R-1 R-2 R-3 R-4 R-5 R-6

LS-1 LS-2 LS-6 HS-1 HS-2 HS-3

4.5 4.3 2.1 0.6 – –

15.9 8.8 4.2 3.1 – –

R-1

(a)

R-2

R-3

R-4

(b)

126540 ppm 63603 ppm

10

Zeta potential (mV)

690

1887 ppm 100620 ppm

0 -10 -20

NaCl CaCl 2

2925 ppm -30 -40 0.0

0.4

0.8

1.2

1.6

3.6 4.0

Ion strength (mol/kg) Fig. 4. Zeta potential of rock/brine interface as a function of ion strength at 45 °C.

R-5

R-6

R-1

R-2

R-3

R-4

R-5

R-6

(c)

Fig. 3. Experimental device for imbibition test (a) and imbibition of sandstone cores in brines with different compositions after (b) 72 h and (c) 60 days at 45 °C. The close-up views of produced crude oil are shown above each picture.

691

Z. Hua et al. / Fuel 182 (2016) 687–695

Oil recovery or Water-cut/%

Zeta potential (mV)

0

100

126540 ppm 63603 ppm 1887 ppm 100620 ppm

-10 -20 -30

NaCl CaCl 2

2925 ppm

-40

300 80

LS-1

HS-2 60

200

40 100 Water-cut Oil recovery Pressure drop

20

Pressure drop / kPa

10

-50 0.0

0.4

0.8

1.2

1.6

3.6 4.0

0

0 0

Ion strength (mol/kg) Fig. 5. Zeta potential of oil/brine interface as a function of ion strength at 45 °C.

1

2

3

4

5

Injected volume / Vp Fig. 7. The displacement test curve of core after being saturated with mixing oil.

of naphthenic acids. Increasing salinity of NaCl solution results in the compression of the electrical double layer and decreasing negative charge of two interfaces. Calcium ions affect the charge at two interfaces more strongly than Na+ at the same ionic strength, so the weak positive charge at rock/brine interface and weak negative charge (close to zero) at oil/brine interface are caused by 63,603 ppm CaCl2 solution. These results suggest that the ion composition and salinity are very important in determining the interactions between the rock/brine and oil/brine interfaces because they can strongly affect the electrical charge at both interfaces, and hence affect rock wettability and the amount of produced oil by imbibition. 3.3. Enhanced oil recovery by LSF The flooding results of cores 7 and 8 showing water-cut, oil recovery and differential pressure are given in Figs. 6 and 7, respectively. Core 7 was first saturated with crude oil and SFW was injected until the water-cut was about 99%; then 100,620 ppm and 2925 ppm NaCl solutions were injected in turn to test the effect of LSF. As shown in Fig. 6, the oil recovery was improved by 10% when 100,620 ppm NaCl solution was injected, and a further 8% by 2925 ppm NaCl solution. Meanwhile the water content tended to drop when changed to inject NaCl solutions or decreasing the salinity of NaCl solution.

500

80

LS-6

SFW

400

LS-2

60

300

40

200

Water-cut Oil recovery 100 Pressure drop

20

Pressure drop/ kPa

Oil recovery or Water-cut/%

100

0

0 0

2

4

6

8

10

Injected volume / Vp Fig. 6. Impact of decreasing NaCl salinity on water-cut, oil recovery and pressure drop.

As mentioned in the imbibition experiment, the initial wettability of core 7 was dominantly controlled by the amount of divalent cations (e.g. Ca2+) in SFW; i.e., the presence of divalent cations in SFW notably promoted the adsorption of polar organic compounds in crude oil on the rock surface. Due to the resulting oil-wet state and the bridge action of divalent cations, some remaining oil was left in the sandstone reservoir after injecting SFW; and thus the oil recovery was relatively low. The rock surface changed to be more water-wet and the electrical double layer was expanded when changed to inject 100,620 ppm NaCl solution with the same ion strength of SFW, and the effect was more significant when decreasing the salinity of injecting NaCl solution; therefore the oil recovery was enhanced obviously by changing to inject NaCl solution and decreasing the salinity of NaCl solution. After changing to 100,620 ppm and 2925 ppm NaCl solutions injection, the pressure drop continued to rise up to 267.6 kPa and 299 kPa, respectively; followed by a gradual decrease and then remained stable at around 188 kPa and 258 kPa, respectively. When changed to inject low-salinity brine, the increase in pressure drop can be interpreted by increasing resistance due to more newly generated oil droplets passing through smaller pore throats. The pressure drop reaching equilibrium suggests that no new oil droplet further generates, which is consistent with the curve of oil recovery. Core 8 was first saturated with mixing oil; then 63,603 ppm CaCl2 solution was injected until the water-cut was about 99%; then deionized water was injected to test the effect of oil components on the bridge action of divalent cations and the efficiency of LSF. As shown in Fig. 7, the oil recovery reached up to 88% when high salinity of CaCl2 solution was injected at first and the subsequent deionized water flooding did not show any oil improvement and a substantial increase in the pressure drop. The water-cut kept in a low level before a steep rise to 95% when injecting CaCl2 solution, resulting in a high recovery during the initial stage; and then the water-cut remained very high values, even after changing to inject deionized water, which is consistent with the change of the oil recovery and the pressure drop. Cores 7 and 8 behaved differently during waterflooding, confirming the existence of Ca2+ bridge action for Core 7 and indicating direct influence of oil components on the initial Ca2+ bridge action and LSF effect. For the rock-brine-crude oil system, the acidic components in crude oil play an important role in the Ca2+ bridge action and the double-layer expansion at oil/brine interface during LSF. While for the core 8 saturated with mixing oil free of acidic components, the Ca2+ bridge was not established when injecting

692

Z. Hua et al. / Fuel 182 (2016) 687–695

high salinity of CaCl2 solution before LSF and thus the oil recovery reached up to a high value in the displacement stage, and the rather weak response from deionized water could be interpreted by the limited expansion of the electrical double layer at oil/brine interface. More direct evidence for enhanced oil recovery by LSF and the mechanism behind LSF effect were obtained by using the microscope to observe the flowing state of oil and water in microvisual model. The clean micro-visual model was first saturated with crude oil and SFW was injected until the water flooding reached equilibrium; then 2925 ppm NaCl solution was injected to test the effect of LSF. As observed in Fig. 8b, after brine flooding with SFW, much remaining oil with film and column structure distributed in small pores of the micro-visual model because the injected water preferentially flowed along the big channel. After changing to 2925 ppm NaCl solution injection, the remaining oil in small pores was driven to big channel and oil recovery was improved, as shown in Fig. 8c. The double-layer expansion by LSF alters the wettability of glass surface to be more water-wet, which is more favorable for the imbibition of low-salinity NaCl solution and displacing the remaining oil in small pores of the micro-visual model. Meanwhile the double-layer expansion leads to higher negative charges at oil/ brine and rock/brine interfaces, causing the increasing electrostatic repulsive forces between these interfaces, and hence promotes destabilization of oil layers adhering to glass surface. Therefore the oil recovery was enhanced during LSF in the micro-visual glass model.

3.4. Mechanism behind wettability alternation and enhanced oil recovery As mentioned in the experimental sections, the double-layer expansion by LSF not only alters the rock surface to be more water-wet but also promotes destabilization of oil layers adhering to rock surface, thus resulting in EOR. For rock-NaCl brine-crude oil system, the fluid-rock interaction and how the interaction links the double-layer expansion to wettability alteration and EOR by LSF were discussed in this part.

3.4.1. Disjoining pressure theory Irreducible water saturation of about 30–40% in artificial cores was learned from initial oil saturations shown in Table 3, which could inhibit the adsorption of active components in crude oil. A wetting-water film was expected to be left between the rock surface and the invading oil, and the reservoirs remain water-wet [19]. In this case, the interaction between oil and rock can be described by surface forces which are composed of electrostatic force (PE), van der Waals force (PVDW), and structural force (PS).

The sum of each component of surface forces is disjoining pressure (P) [20].

P ¼ PE þ PVDW þ PS From the linear superposition approximation (LSA), electrostatic force between rock/brine and oil/brine interfaces can be estimated using zeta potentials, f1 and f2, for the rock/brine and oil/ brine interfaces, respectively, and gives [21]

PE ðhÞ ¼ 64n1 kB T tanhðzef1 =4kB TÞ tanhðzef2 =4kB TÞejh where n1 is the ionic concentration in the bulk; kB is the Boltzmann constant; j is the Debye reciprocal length; z is the valence of the ion; e is the elementary charge; T is the ambient temperature; and h is the thickness of the wetting-water film between two parallel interfaces. For the oil/brine/rock system, the van der Waals force between two parallel interfaces is expressed as [22]

PVDW ðhÞ ¼ 

Aqwo 6ph

3

where Aqwo is the Hamaker constant for the oil/water/rock system, and Aqwo is approximately 1  1020 J [20,23–25]. Compared to electrostatic and van der Waals forces, the structural forces are short-range interactions, at a distance of less than 5 nm. The structural force is given by the following equations [23]: hh

PS ðhÞ ¼ AS e

S

where As is the coefficient and hs is the characteristic decay length for the exponential model. For the oil/water/rock system, it is supposed that As is 1.5  1010 Pa and hs is 0.05 nm [25]. 3.4.2. Influence of salinity on surface forces Zeta potential results above suggest that the low-salinity water causes expansion of the electrical double layer, thus results in a change in electrostatic repulsive force between oil and rock. Based on the zeta potentials of rock/brine and oil/brine interfaces shown in Figs. 4 and 5, the disjoining pressures between oil and rock were calculated as given in Fig. 9. The maximum of disjoining pressure isotherm (Pmax) decreases with increasing the salinity of NaCl solutions and becomes negative when the salinity is increased to 29,250 ppm. Above the salinity of 29,250 ppm, the isotherms are controlled by the van der Waals force except at small thicknesses, where the structural forces play a dominant role. The results demonstrate that decreasing the salinity of NaCl solution increases the magnitude of the negative charge at the rock/brine and oil/brine interfaces, resulting in increasing repulsive forces between rock/brine and oil/brine interfaces and increasing Pmax. And thus injecting low salinity NaCl solution will cause the destabilization of oil layers adhering to sandstone surface, which is a contributing mechanism to EOR.

Fig. 8. Microscopic distribution of remaining oil in micro-visual model, before (a) and after flooding with SFW (b) and 2925 ppm NaCl solution (c).

693

Z. Hua et al. / Fuel 182 (2016) 687–695

80

Πmax

40

Π (kPa)

(dashed line l); therefore, curved surface R and circular arc C

LS-2 LS-3 LS-4 LS-5 LS-6

_

(point B is the center of C) are gained, respectively, when arc OP and point O rotate around dashed line l, as shown in Fig. 10c. Because the intersecting line (line OA) of Sections 1 and 2 is per_

pendicular to both the tangents of arcs C and OP on curved surface R at point O, line OA overlaps the normal vector (n) of curved surface R at point O; thus line OA is also perpendicular to the tangent

0

_

_

of arc OQ at point O and the center of the curvature circle of arc OQ at point O lies on line OA. Because the meniscus of oil-brine interface is axially symmetric, it can be proved that the point A is the

-40

_

-80

center of the curvature circle of arc OQ at point O. Thus r2 is equal to the length of line AO, given by

0

2

4

6

8

10

h (nm)

r2 ¼ OA ¼

Fig. 9. Disjoining pressure between rock surface and oil drop surface in the NaCl solution as a function of film thickness at different salinity.

3.4.3. Influence of salinity on wetting-water film stability and wettability For a cylindrical pore system, the augmented Young-Laplace equation [20] can be given by




Pc ¼ PO  PW ¼ P þ

1 1 þ r1 r2



rOW

ð1Þ

where P is disjoining pressure; as shown in Fig. 10, r1 is the curva_

ture radius of oil-brine interface profile (OP) in the axial section at the point O, i.e., the section pass the symmetric axis (dashed line l) of the cylindrical pore; r2 is the curvature radius of oil-brine inter_

face profile (OQ ) at point O in Section 2 which the tangent of the arc _

OP at the point O is perpendicular to; Pc is Laplace pressure or capillary pressure; PO and PW are the oil phase pressure and the brine phase pressure, respectively; rOW is the oil-brine interfacial tension, as given in Table 2. The curvature of the oil-brine interface profile in Section 1 is equal to the rate of the change of the angle, a, of inclination with respect to the arc length, s. Thus the curvature is

1=r 1 ¼ da=ds

ð2Þ

The right hand side of Eq. (2) is

da dh da da d cos a  ¼ ¼ sin a  ¼ ds ds dh dh dh

ð3Þ

Therefore, the curvature is expressed as

1=r 1 ¼ d cos a=dh

ð4Þ _

For a cylindrical pore, the meniscus of oil-brine interface (OP) is axially symmetric, i.e., symmetric to rotation around the axis

OB rh ¼ cos a cos a

ð5Þ

where r is the average pore radius. For a cylindrical pore, it is assumed that the pore tortuosity   2 [26], the average pore s = 1.4. Based on the equation K ¼ /r 8s2

radius (r) of about 0.84 lm is therefore inferred from reservoir permeability (K) and porosity (u) as listed in Table 3. Cylindrical pores contains the hemispherical meniscus such that the wetting-water film on the pore wall has a curvature of r, and thus where r1 = +1, r2 = r, and Peq  Pc/2 inferred from Eq. (1). Meanwhile oP/oh < 0 is a requirement for locally stable points of the wetting-water film. The departure of the film thickness from the equilibrium point needs to do work, resulting in the increase of interaction potential (Dx), described by

Dx ¼

Z

heq h





PðhÞ  Peq dh

ð6Þ

Taking 2925 ppm NaCl solution as an example, the capillary pressure (Pc) in this pore is about 50 kPa calculated from YoungLaplace equation, and thus the locally stable thicknesses are determined by the intersections of the disjoining pressure isotherm and P = Pc/2. The locally stable points are depicted in Figs. 10b and 11a, and in both cases the isotherms of interaction potential are shown in Fig. 11b. As shown in Fig. 11, the wetting-water film can take two positions, heq,1 = 0.362 nm and heq,2 = 3.702 nm, if P(h) = Pc/2 and oP/oh < 0. Compared to the interaction potential at heq,2, the wetting-water film at heq,1 has lower energy. If the water film is formed by thinning from a large separation, however, a thicker metastable film at heq,2 may exist due to the energy barrier at h3 separating the locally stable points. The results suggest that the thicker wetting film has a trend to be stable at low salinity brine, which could promote destabilization of crude oil from rock surface. In the rock/brine/crude oil system, the surface force also controls the contact angle of meniscus. For a water-wet cylindrical pore in sandstone reservoirs, the thickness of the wetting-water

Fig. 10. Schematic diagram of oil/brine interface in a cylindrical pore.

694

Z. Hua et al. / Fuel 182 (2016) 687–695

1.5

80 40

Π (h)=Pc/2

Δ ω (mJ/m2)

Π (kPa)

heq=0.362 nm

1.0

Π (h)

0 -40

0.5

Energy barrier

-0.5

-80

heq=3.702 nm

0.0

Locally stable point

-1.0 3 heq, 2

0 heq, 1 h3

6

0 heq, 1 h3

9

3 heq, 2

6

h (nm)

h (nm)

(a)

(b)

9

Fig. 11. Disjoining pressure and interaction potential of rock/2925 ppm NaCl solution/crude oil system as a function of film thichness.

film (h) is relatively thin compared with the pore radius (r) and the angle of inclination (a) is small, so that

1 1 d cos a cos a d cos a 1 þ ¼ þ  þ r1 r2 r dh rh dh

ð7Þ

Substituting Eq. (7) into Eq. (1) gives

rOW d cos aðhÞ ¼


 Pc  PðhÞ dh 2

ð8Þ

Eq. (8) is integrated from the uniform film thickness, heq, where

a (heq) = 0 to yield 1  cos aðhÞ ¼

1

rOW

Z

heq




PðhÞ 

h

 Pc dh 2

ð9Þ

As shown in Eq. (9), the right hand side of Eq. (9) equals the ratio of the increase of interaction potential to oil-brine interfacial tension. The macroscopic meniscus, where P(h⁄) = 0, extrapolated to rock surface forms the contact angle; h⁄ is the starting point for extrapolating, as shown in Fig. 10b. By integrating the right hand side of Eq. (9) when h > h⁄, a linear relationship is obtained:

1  cos aðhÞ ¼ 1  cos aðh Þ þ

Pc ðh  h Þ 2rOW

When the salinity of NaCl brines is increased, the maximum of disjoining pressure isotherm decreases, i.e., the energy barrier separating the locally stable points decreases, and a thinner film at heq,1 may form. Then the extrapolated value (1  coshW) is positive and a finite contact angle exists. According to the surface force, the contact angle of brine in reservoir pores calculated from Eq. (11) depends on the salinity, as given in Fig. 12. Fig. 12 shows the contact angle of water on rock surface (hW, as shown in Figs. 10b and 12) increases as the salinity of NaCl solution is increased. When the salinity of NaCl solution is in the range of 29,250–100,620 ppm, the thickness of the wetting-water film is thinner (heq = heq,1). In this case, the water wettability of quartz surface is mainly controlled by the disjoining pressure between rock surface and crude oil, and the contact angle of water on rock surface (hW) decreases from 30.28 to 19.86 when the salinity of NaCl solutions is decreased from 100,620 ppm to 29,250 ppm. When the salinity of NaCl solutions is decreased to 5850 ppm, the disjoining pressure between rock/brine and oil/brine interfaces is so large that the LSF results in a thicker wetting-water film and a fully water-wet surface (i.e., hW = 0). The results suggest that lowering the salinity of NaCl solution alters the wettability of the rock surface to a more water-wet state favorable for water imbibition in small pores and EOR.

ð10Þ

32

Thus this straight line can be extrapolated to the vertical axis to determine the contact angle (hW) of water on rock surface:

Pc

2rOW

h

LS-5

ð11Þ

Lowering the salinity of brine increases the maximum of disjoining pressure isotherm (Pmax), as shown in Fig. 9. When Pmax > Peq, the wetting-water film may thin only to the outer locally stable point (heq,2). Then the extrapolated value (1  coshW) is negative that suggests the extrapolation of the meniscus is parallel to the rock surface, and hence the low-salinity brine wets fully rock surface and oil layer easily slides on the wetting-water film. As shown in Fig. 9, when the salinity of NaCl brines is less than or equal to 5850 ppm, Pmax > Peq, so that the low salinity NaCl brines will result in fully water-wet surface and a thicker wetting-water film, and meanwhile, the thicker wetting-water film can result in oil layer instability and prevent the destabilized oil layer from being trapped, which could be a contributing mechanism to EOR by LSF, as shown in Figs. 6 and 8.

θW ( )

1  cos hW ¼ 1  cos aðh Þ 

LS-6 24

LS-4 16

8

heq=heq,2

0

LS-2 LS-3 0.0

heq=heq,1

0.5

1.0

1.5

2.0

cNaCl / (mol/L) Fig. 12. Calculated contact angle of water on rock surface (hW) in the rock/NaCl solution/crude oil system as a function of salinity at 45 °C. (hO is the contact angle of oil on rock surface in three-phase system; hW is the contact angle of water on rock surface in three-phase system; and hW = 180°  hO).

Z. Hua et al. / Fuel 182 (2016) 687–695

4. Conclusions Imbibition, f-potential measurements, core displacement, and microscopic displacement experiments and a surface force model for determining the effect of injection brine composition on wettability and oil recovery in sandstone reservoirs and the mechanisms underlying the effect allow drawing the following conclusions: (1) Ca2+ in the brine acts like a bridge between negatively charged sites on the rock surface and negatively charged polar groups of dissociated acidic components in crude oil to make the sandstone surface hydrophobic. The addition of little of Ca2+ in brines could obviously alter the sandstone reservoir to more oil-wet, and the increased salinity of CaCl2 brine yields more oil-wet state. For example, after imbibing in CaCl2 solutions for 60 days, low oil recovery of only 3.1% and about 0% were obtained when the salinity of CaCl2 solution were 1887 ppm and 63,603 ppm, respectively. Therefore divalent cations (e.g. Ca2+) in the brine hinder the low salinity effect and low-salinity brine without divalent cations will show optimal low salinity effect. (2) Lowering the salinity of NaCl brine causes the expansion of the electrical double layer and increasing disjoining pressure between rock/brine and oil/brine interfaces, thus resulting in a more water-wet rock surface and improving oil recovery. The imbibition results in NaCl solution show that the oil recovery increased from 4.2% to 8.8% when the salinity of NaCl solution was decreased from 100,620 ppm to 2925 ppm, and the oil recovery was up to 15.9% for deionized water. LSF showed a high potential to improve oil recovery, decrease water-cut and increase pressure drop for the core saturated with crude oil; and the core flooding results show that the oil recovery was improved by 10% when changed to inject 100,620 ppm NaCl solution on the foundation of SFW flooding, and a further 8% by 2925 ppm NaCl solution. (3) Increasing disjoining pressure between rock surface and crude oil and a more water-wet rock surface occur when lowering the salinity of injected NaCl brines, which is supported by the calculation using the surface force model in a cylindrical pore. The calculation show that the contact angle of water on rock surface (hW) decreases from 30.28 to 19.86 when the salinity of NaCl solution is decreased from 100,620 ppm to 29,250 ppm. When the salinity of NaCl solution is decreased to 5850 ppm, the disjoining pressure between rock/brine and oil/brine interfaces is so large that the LSF results in a thicker wetting-water film and a fully water-wet rock surface (i.e., hW = 0). Therefore EOR by LSF can be attributed to two main reasons as following: LSF leads to more water-wet sandstone reservoirs, and thus enhances the ability of water imbibition; Meanwhile, LSF results in a thicker wetting-water film between rock surface and crude oil, thereby promoting destabilization of oil layers adhering to sandstone surface.

Acknowledgements The authors acknowledge the CNOOC Research Institute, the Research Institute of Enhanced Oil Recovery of China University

695

of Petroleum (Beijing), and the National Key Scientific and Technological Project (Grant No. 2011ZX05024-004) for financial support and permission to publish the results of this study. References [1] Vledder P, Gonzalez I, Carrera Fonseca J, Wells T, Ligthelm D. Low salinity water flooding: proof of wettability alteration on a field wide scale. In: SPE 129564, SPE improved oil recovery symposium. Tulsa, Oklahoma, USA; 2010. [2] Austad T, RezaeiDoust A, Puntervold T. Chemical mechanism of low salinity water flooding in sandstone reservoirs. In: SPE 129767, SPE improved oil recovery symposium. Tulsa, Oklahoma, USA; 2010. [3] Nasralla R, Nasr-El-Din H. Double-layer expansion: is it a primary mechanism of improved oil recovery by low-salinity waterflooding? In: SPE 154334, SPE improved oil recovery symposium. Tulsa, Oklahoma, USA; 2012. [4] Lager A, Webb K, Collins I, Richmond D. LoSal enhanced oil recovery: evidence of enhanced oil recovery at the reservoir scale. In: SPE113976, SPE/DOE improved oil recovery symposium. Tulsa, Oklahoma, USA; 2008. [5] Suijkerbuijk B, Hofman J, Ligthelm D, Romanuka J, Brussee N, van der Linde H, et al. Fundamental investigations into wettability and low salinity flooding by parameter isolation. In: SPE154204, SPE improved oil recovery symposium. Tulsa, Oklahoma, USA; 2012. [6] Lager A, Webb KJ, Black CJ, Singleton M, Sorbie KS. Low salinity oil recovery-an experimental investigation. Petrophysics 2008;49(1):28–35. [7] Al Shalabi EW, Sepehrnoori K, Delshad M. Mechanisms behind low salinity water injection in carbonate reservoirs. Fuel 2014;121:11–9. [8] RezaeiDoust A, Puntervold T, Strand S, Austad T. Smart water as wettability modifier in carbonate and sandstone: a discussion of similarities/differences in the chemical mechanisms. Energy Fuels 2009;23(9):4479–85. [9] Brady PV, Morrow NR, Fogden A, Deniz V, Loahardjo N. Electrostatics and the low salinity effect in sandstone reservoirs. Energy Fuels 2015;29(2):666–77. [10] Yang J, Dong Z, Dong M, Yang Z, Lin M, Zhang J, et al. Wettability alteration during low-salinity waterflooding and the relevance of divalent ions in this process. Energy Fuels 2015;30(1):72–9. [11] Fogden A. Effect of water salinity and pH on the wettability of a model substrate. Energy Fuels 2011;25(11):5113–25. [12] Hadia NJ, Hansen T, Tweheyo MT, Torsæter O. Influence of crude oil components on recovery by high and low salinity waterflooding. Energy Fuels 2012;26(7):4328–35. [13] Farooq U, Asif N, Tweheyo MT, Sjöblom J, Øye G. Effect of low-saline aqueous solutions and pH on the desorption of crude oil fractions from silica surfaces. Energy Fuels 2011;25(5):2058–64. [14] Nasralla R, Bataweel M, Nasr-El-Din H. Investigation of wettability alteration by low salinity water in sandstone rock. In: SPE146322, SPE offshore Europe oil and gas conference and exhibition. Aberdeen, UK; 2011. [15] Hamam SE, Hamoda MF, Shaban HI, Kilani AS. Crude oil dissolution in saline water. Water, Air, Soil Pollut 1988;37(1–2):55–64. [16] Meybodi HE, Kharrat R, Araghi MN. Experimental studying of pore morphology and wettability effects on microscopic and macroscopic displacement efficiency of polymer flooding. J Pet Sci Eng 2011;78(2):347–63. [17] Shi Q, Zhao S, Xu Z, Chung KH, Zhang Y, Xu C. Distribution of acids and neutral nitrogen compounds in a Chinese crude oil and its fractions: characterized by negative-ion electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Energy Fuels 2010;24(7):4005–11. [18] Shi Q, Hou D, Chung KH, Xu C, Zhao S, Zhang Y. Characterization of heteroatom compounds in a crude oil and its saturates, aromatics, resins, and asphaltenes (SARA) and non-basic nitrogen fractions analyzed by negative-ion electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Energy Fuels 2010;24(4):2545–53. [19] Anderson W. Wettability literature survey-part 1: rock/oil/brine interactions and the effects of core handling on wettability. J Pet Technol 1986;38 (10):1125–44. [20] Hirasaki GJ. Wettability: fundamentals and surface forces. SPE Formation Eval 1991;6(2):217–26. [21] Israelachvili JN. Intermolecular and surface forces: revised third edition. San Diego: Academic Press; 2011. [22] Gregory J. Approximate expressions for retarded van der Waals interaction. J Colloid Interface Sci 1981;83(1):138–45. [23] Takahashi S. Water imbibition, electrical surface forces, and wettability of low permeability fractured porous media. Stanford: Stanford University; 2009. [24] Takamura K. Microscopic structure of Athabasca oil sand. Can J Chem Eng 1982;60(4):538–45. [25] Morrow NR. Interfacial phenomena in petroleum recovery. Boca Raton: CRC Press; 1990. [26] Yang S. Fundamentals of petrophysics. Beijing: Petroleum Industry Press; 2011.