Advances in Water Resources 32 (2009) 98–109 Contents lists available at ScienceDirect Advances in Water Resources journal homepage: www.elsevier.com/locate/advwatres Interfacial tension measurements and wettability evaluation for geological CO2 storage C. Chalbaud a,1, M. Robin a, J-M Lombard a, F. Martin a, P. Egermann a,1, H. Bertin b,* a b Institut Français du Pétrole (IFP), 1 et 4, Avenue de Bois Préau, 92852 Rueil Malmaison Cedex, France Université de Bordeaux, Laboratoire TREFLE, Esplanade des Arts et Métiers 33405 Talence Cedex, France a r t i c l e i n f o Article history: Received 12 June 2008 Received in revised form 13 October 2008 Accepted 14 October 2008 Available online 29 October 2008 Keywords: CO2 Storage Interfacial tension Wettability Capillary Green house Aquifers a b s t r a c t Interfacial interactions, namely interfacial tension, wettability, capillarity and interfacial mass transfer are known to govern fluid distribution and behavior in porous media. Therefore the interfacial interactions between CO2, brine and oil and/or gas reservoirs have a significant influence on the effectiveness of any CO2 storage operations. However, data and knowledge of interfacial properties in storage conditions are scarce. This issue becomes particularly true in the case of deep saline aquifers where limited, economically driven, data collection and archiving are available. In this paper, we present a complete set of brine–CO2 interfacial tension data at pressure, temperature and salinity conditions, representative of a CO2 storage operation. A semi-empirical correlation is proposed to calculate the interfacial tension from the experimental data. Wettability is studied at pore scale, using glass micromodels in order to track fluids distribution as a function of the thermodynamic properties and wettability conditions for water– CO2 systems. With this approach, we show that, in strongly hydrophilic porous media, the CO2 does not wet the solid surface whereas; if the porous media has less hydrophilic properties the CO2 significantly wets the surface. Ó 2008 Elsevier Ltd. All rights reserved. 1. Introduction Large-scale subsurface storage of anthropogenic carbon dioxide is considered as a promising technology to stabilize greenhouse gas concentration in the atmosphere [1]. There are at least three options for CO2 geological storage [2]: oil and gas depleted reservoirs, deep saline aquifers and unminable coal beds. Successful CO2 sequestration in deep saline aquifers and different types of hydrocarbon reservoirs is largely governed by the fluid–fluid and fluid– rock interfacial interactions among which interfacial tension is of central importance. In this study we focused in two main interfacial properties: IFT and rock wettability because (i) they control the capillary-sealing efficiency with respect to CO2 of the caprock, a low permeable medium usually imbibed with water and (ii) they control the transport properties (relative permeabilities and residual saturations) of the water (or brine) and CO2 phases in the reservoir rock. In the case of deep saline aquifers, the interfacial tension corresponds to that of a brine–CO2 system. The reliability and accuracy of the interfacial tension data used in these studies are particularly important for CO2 storage, since it greatly influences the flow process and it controls the capillary-sealing efficiency (see Eq. (1)). * Corresponding author. E-mail address: henri.bertin@bordeaux.ensam.fr (H. Bertin). 1 Present address: Gaz de France, 361, bd Pres. Wilson 93211 Saint-Denis La Plaine, France. 0309-1708/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.advwatres.2008.10.012 This sealing efficiency is the consequence of a trapping mechanism sometimes called structural trapping or hydrodynamic trapping, a caprock or a sealing fault prevent the movement of the ascending CO2 plume and it concerns usually the largest part of the CO2 injected, for this reason it is also referred as primary trapping [3]. Capillary or residual (or mobility trapping) is among the so-called secondary trapping mechanism and it concerns the CO2 left during the re-imbibition occurring behind the ascending CO2 plume as residual disconnected bubble. In some cases this mechanism concerns a significant part of the injected CO2 having the particular interest that it does not present any leakage risk, unlike primary trapping [4,5]. As well as primary trapping the latter mechanism depends on water (or brine) IFT and reservoir rock wettability Pth c ¼ P CO2  P brine ¼ 2cb;CO2 cos h R ð1Þ where P th c is the threshold capillary pressure of the saturated brine caprock, which characterizes the ability of a non-wetting phase (CO2) to start flowing in a porous medium saturated with a wetting phase (brine), R is the largest connected pore throat in the caprock and h is the contact angle. Because there is no agreement concerning the use of interfacial tension or surface tension terms under reservoir conditions, a preliminary clarification of vocabulary is first made. Throughout this paper, we shall use ‘‘surface tension” only in the case of pure compounds and gas/liquid systems at ambient conditions and ‘‘interfacial tension” (IFT) in all other cases. C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 Interfacial tension experimental values of a pure water–CO2 system in reservoir conditions, cw;CO2 , have been reported by several authors since 1957 [6–9]. Nevertheless, there is some concern with the use of these data [9], related to the thermodynamic equilibrium, the type of image analysis method used and the failure to take into account the dissolved CO2 effect on the water phase density [10]. The cw;CO2 has also been investigated by molecular dynamics computer simulations, [11,12] but the differences between the calculated and the measured values by the same authors [11,12] were considerably wide (up to 30%). Yan et al. [13] calculated the cw;CO2 at 40 °C using the linear gradient theory model [14,15] but the authors concluded that this method was not suitable for the water–CO2 system because it overestimated the IFT at low and moderate pressures and it underestimated it at high pressures. In addition, this method presents limitations in the near-critical region [14]. To our knowledge, there is no complete set of experimental data dealing with the brine–CO2 system at high pressures, high temperatures and different salt types and concentrations which could be used to estimate cb;CO2 value at conditions relevant to field studies. Yang and Gu [16] reported cb;CO2 values at high pressures and different temperatures for low and constant salinity, 4270 ppm. For these values the salt effect is negligible, thus these data do not permit to evaluate the impact of salt concentration on interfacial tension. Although the study by Yang and Gu [16] does allow the evaluation of the mass transfer between water and CO2 phases, IFT data are not determined at the equilibrium. Therefore, the claim of a complete miscibility (0 mN/m interfacial tension) between water and CO2 at P = 122.38 bar and T = 58 °C is not consistent with values reported in other studies [6–8] nor with the values presented in this paper at similar thermodynamic conditions. Recently, Bennion and Bachu [17] reported nine cb;CO2 values for several reservoir brines with different salt concentrations and under different thermodynamic conditions. These data allow us to identify some trends related to the cb;CO2 evolution with the thermodynamic conditions. Data reported by these authors are however not enough, either to clearly understand the behavior of the cb;CO2 at reservoir conditions or to develop any quantitative relationship to estimate the cb;CO2 ; the authors emphasize the need for systematic measurement and analysis of the cb;CO2 under reservoir conditions. Recently Chiquet et al. [18] present interfacial tension measurement at reservoir pressure and temperature but with a very low salinity that is not representative of most of the deep saline aquifer conditions. The presence of salts in brine increases the interfacial tension compared to the pure water case at the same pressure and temperature. This behavior was identified for different thermodynamic conditions, different fluid systems and different chlorides [17,19– 25]. A summary of most of these data was carried out by Argaud [26]. As explained by Ralston and Healy [23], and Johansson and Eriksson [24], this increase is related to the location of the cations in the aqueous phase and how this affects the interface structure. Cations have no affinity for the gas–liquid interface hence they are accumulated in the bulk solution. Water molecules at the interface are attracted by cation solvation towards aqueous bulk phase. This attraction increases along with the concentration of cations and as the ratio of cation charges z+ to cation surface area r2 increases; thus the effect on IFT increases in the following sequence [19,26]: þ Csþ < Rb < NHþ4 < Kþ < Naþ < Li < Ca2þ Mg2þ þ ð2Þ Several authors have reported a linear relationship between the increase of the surface tension of water and molal salt concentration of NaCl at ambient conditions [19,23,24]. Massoudi and King [20] showed that this kind of relationship is also valid for the brine– 99 CO2 system for pressures up to 60 bars and NaCl molal concentrations up to 5 molal at a temperature of 25 °C. Similar measurements at higher pressures or temperatures are not available. Aveyard and Saleem [21], and Johansson and Eriksson [24] suggested that for a given salt concentration, the increase in interfacial tension is proportional to the Kelvin temperature. However, there is no experimental data to support this suggestion. The first part of this paper is dedicated to the experimental measurement of the interfacial tension of brine–CO2 systems in reservoir conditions and to the development of a correlation to facilitate the use of these results and to investigate the behavior of brine–CO2 systems in reservoir conditions compared to other water–gas systems. In Eq. (1) we introduced the contact angle, h, which is often related to the wettability of porous media even if the upscaling of this parameter to the scale of the reservoir can be seen as a subject of controversy [27]. Therefore, the second part of this paper is dedicated to the study of wettability at pore scale. Wettability is defined as the tendency of one fluid to spread on or adhere to a solid surface in the presence of other immiscible fluids [28]. In the case of a CO2 storage operation; it influences the fluid distribution in the reservoir because it affects the quantity displaced as well as how the displacement proceeds. The wetting fluid will occupy the smallest pores and will be present in the largest pores like films on the rock surface. The existence of such films enhances the continuity of the wetting phase, affecting the petrophysical properties of the reservoirs; the relative permeability and the capillary pressure curves, therefore the injectivity level associated to a reservoir formation [29]. Dealing with leakage of the stored CO2, its invasion in the caprock may occur according to different physical mechanisms: capillary breakthrough of the CO2 phase, diffusion of CO2 molecules into brine and migration through reactivated or induced fractures in the caprock, for instance after a pressure build up or a temperature decrease during CO2 injection. As can be seen in Eq. (1), wettability has a direct effect on capillary breakthrough which occurs when the pressure of the CO2 phase rises above a threshold value, which corresponds to the capillary threshold pressure. More information about this pressure value and how to measure it from laboratory tests can be found in Egermann et al. [3]. In two-phase and three-phase systems, the gas, in most cases, is considered to be the non-wetting fluid. This generalization usually leads to disregarding the possibility of a partial wetting (and its consequences) of the injected CO2, even if, in storage conditions, it is not a gas phase but rather a supercritical or a liquid phase. Recently, Chiquet et al. [30] measured the contact angle between brine, dense CO2 and minerals representative of shales, such as mica and quartz. The authors reported that water wettability in these minerals barely changes. According to the authors, such minor contact angle variations could be attributed to the reduction, at high pressures (low pH), of electrostatic interactions at the interfaces. These interactions tend to stabilize the brine film hence favoring water-wettability. At the core scale, Egermann et al. [31], by means of standard CO2 flooding experiments, in carbonate cores at reservoir conditions, observed that for pressures ranging from 80 to 180 bars and temperatures from 60 to 80 °C, the CO2 is clearly the non-wetting phase. A more pronounced change from water-wet to intermediate wet conditions, while increasing pressure, was reported by Siemons et al. [32] for coal–water–CO2 systems, by means of contact angle measurements. According to these results, the change in the contact angle reported by Chiquet et al. [30] does not have any significant impact at the core scale. This apparent discrepancy between both studies becomes more important because of the carbonate nature (98% calcite) of the samples used by Egermann et al. [31]; in carbonate rocks, the isoelectric point (the point where the surface charge 100 C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 density vanishes) occurs at a lower pressure. Therefore, if the wettability alteration while increasing pressure can be attributed to the electrostatic forces, this alteration should be more important for carbonate minerals [12], which is not the case for Egermann et al. [31]. This point needs further investigation since these experiments were conducted under different conditions which make it hard to directly compare them. Due to apparent contradictions in literature results concerning the possibility of a partial CO2 wetting, wettability on minerals at different physical scales has to be further investigated. The first part is dedicated to the interfacial tension, the description of our experimental device, the experimental conditions and the IFT measuring procedure. Special attention is paid to the characteristics of the device and the measuring procedure, which we consider to be the main cause of differences observed between reported IFT values for the pure water–CO2 system at high pressure and high temperature. Then we present our results which are discussed and compared with those reported in previous works, some of them being presented in this introduction. The application of these results to practical storage cases, especially into deep saline aquifers, is analyzed. The next section is devoted to cb;CO2 modeling. The second part of the paper deals with the study of the wettability at the pore scale following the same structure as in the first part. Finally, conclusions are drawn regarding the future work to be conducted. 2. Interfacial tension For a given salt concentration and a given temperature, an average of 10 values of cb;CO2 were measured for pressure values ranging from 45 bars up to 255 bars. Three temperatures 27, 71 and 100 °C and four salinities (5 g/L, 50 g/L, 100 g/L and 150 g/L) were investigated. The density values of brine solutions were calculated from Søreide and Whitson [33] and Rowe and Chou [34]. This calculation takes into account the effect of dissolved CO2 and presence of NaCl. The CO2 density was considered equal to that of pure CO2 according to King et al. [35]. 2.1. Experimental apparatus The framework is designed to work under representative pressure and temperature reservoir conditions, i.e. up to 260 bars and 120 °C (Fig. 1). It allows interfacial tension measurements for two-phase systems, using the pendant (or rising) drop method. Drop-shape techniques have been widely used and are among the most accurate methods. We used the axisymmetric drop-shape analysis (ADSA). The ADSA method consists in acquiring images of a drop and extracting the experimental drop profile using edge detection techniques. Details of the ADSA methodology can be found elsewhere [36]. The accuracy of the ADSA method depends on the quality of the extracted edge profile. There was no concern related to it for the brine–CO2 rising drop configuration chosen in this study. The diagram of the experimental device for the rising drop case is shown in Fig. 1. This device can be divided into three separate systems: the high pressure viewing cell, the fluid circuit and feeding system, and the imaging system. The feeding system consists in a volumetric pump operated with compression oil connected to two reservoirs cells equipped with a piston. A CO2 cap is positioned at the top of the viewing cell in order to improve thermodynamic equilibrium and the fluids are presaturated for measurement conditions. This means that enough CO2 was put into the brine feeding cell to saturate the brine solution and enough water was put into the CO2 feeding cell to saturate CO2. We use the term ‘‘presaturate” instead of ‘‘saturate” because experience showed that using this kind of experimental device for the water–CO2 stabilization in pressure is not sufficient to ensure complete saturation of the phases [9]. Since any shift in pressure and temperature has a significant influence on the results, it is important to carefully control any possible shift of these conditions in the entire system. The feeding system, the fluid circuit and the viewing cell are maintained at the same pressure and temperature conditions. The imaging system consists in a linear arrangement of a light source, a glass diffuser, the rising drop and the digital camera. 2.2. Experimental procedure To our knowledge, only Hebach et al. [9] have published a measuring procedure that can be applied for highly accurate measurements under high pressures, for the water–CO2 system. CO2 drops are generated at the tip of a glass capillary tube (outer diameter = 1.5 mm) in the viewing cell which had been previously filled with the brine solution. The capillary tube is also used as an internal metric standard to calibrate the drop size. It has been observed that the decompression process requires more time to achieve a stable thermodynamic equilibrium and the drop visualization may be delicate. At each pressure, the IFT was obtained by taking the average of at least five measured values. For each salinity a complete set of isotherms was measured. Prior to the first experiment and to a change of the brine solution (different NaCl concentration) the whole system was cleaned by circulating de-ionized water and the surface tension of water (air–water at ambient conditions) was measured until finding its typical value, 72 mN/m ± 2% [37]. Since the equilibrium of the system is of primary importance to the accuracy of the measurements, once the drop is generated and enlarged at the tip of the capillary, it is maintained in the cell for several minutes. Measurements were taken over time during this period as shown in Fig. 2. The equilibrium IFT is a static value that was reached between 8 and 15 min after the drop generation, depending on thermodynamic conditions. The period during which IFT decreases over time corresponds to the dissolution of CO2 in brine and vice versa. If CO2 is not pre-saturated before the drop formation and if the CO2 cap is not placed in the viewing cell, static IFT Saturated CO2 CO2 cap Saturated Brine H2O + NaCl P, T CO2 Drop Light Source Glass Diffuser Oven P, T Fig. 1. Experimental set-up. Camera Image Analyse 101 29.5 a 50.00 29 28.5 28 27.5 27 set point 26.7 mN/m 26.5 0 200 400 600 800 1000 1200 Interfacial Tension (mN/m) Interfacial Tension (mN/m) C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 45.00 27˚C 0.085 m NaCl 27˚C 0.87 m NaCl 27˚C 1.79 m NaCl 27˚C 2.75 m NaCl 40.00 35.00 30.00 25.00 t (s) 20.00 Fig. 2. Change of the cb;CO2 with time. P = 120 bars, T = 27 °C and [NaCl] = 5000 ppm. 50 100 150 200 250 300 250 300 Pressure (bar) b 50.00 Interfacial Tension (mN/m) values would be more difficult to achieve. Hebach et al. [9] showed that this evolution of the IFT corresponds to the evolution of the aqueous phase density due to the dissolution of the CO2. If IFT is calculated with the instantaneous density difference, it remains constant throughout. Therefore, the evolution the IFT presented in Fig. 2 is the consequence of using the saturated aqueous phase density value instead of the instantaneous ones. If densities cannot be measured in the viewing cell, we consider that the best approach to thermodynamic equilibrium is to use the saturated aqueous phase density and wait until a constant value of IFT is reached. 0 45.00 71˚C - 0 m NaCl 71˚C - 0.085 m NaCl 71˚C - 0.87 m NaCl 71˚C - 1.79 m NaCll 40.00 71˚C - 2.75 m NaCl 35.00 30.00 25.00 2.3. Brine–CO2 IFT vs. pressure and temperature 20.00 0 50 100 150 200 Pressure (bar) c 50 Interfacial Tension (mN/m) Fig. 3 shows cb;CO2 isotherms obtained at T = 27, 71 and 100 °C. The evolution of cb;CO2 with pressure and temperature is similar to that reported by other authors for the cw;CO2 under similar conditions [6–9]: at low pressures, cb;CO2 decreases with pressure; this decrease is more pronounced at low temperatures. At high pressures, a plateau value of cb;CO2 is reached for all NaCl concentrations. At T = 27 °C, this plateau is reached for P = 80 bars; for T = 71 °C it is reached at P = 150 bars. At T = 100 °C, it is not possible to claim that a plateau has yet been reached. It is interesting to note that the pressure at which this plateau is reached does not depend on the NaCl concentration neither at T = 27 °C nor at T = 71 °C. Nevertheless, the value of the cb;CO2 at the plateau does depend on the salt concentration. At these temperatures, the cb;CO2 value is the same once the plateau has been reached. Its value is approximately 26 mN/m for 5 g/L of NaCl. At the highest temperature investigated (100 °C), the minimal cb;CO2 is also close to 26 mN/m for 5 g/L. We call this value cWplateau, where the W subscript refers to pure water. For a similar range of temperatures and pressures, several authors [6–9] have already reported the presence of a plateau in the IFT value in the case of pure water–CO2 system. Fig. 3b presents the IFT of pure water–CO2 and of brine–CO2, for different salt concentrations at T = 71 °C. It can be seen from this figure that the difference between the measured IFT values for pure water and the lowest salinity brine (5 g/L or 0.085 m) is negligible. We therefore consider that our results at [NaCl] = 0.085 m can be regarded as analogous to those of pure water. For such a low NaCl concentration, a negligible effect of salt has already been reported for pressures up to 60 bars [20]. All the results are presented in NaCl concentration expressed in molal values for comparison with the linear relationship between the IFT increase and the molal concentration of salt reported by previous authors mentioned in the introduction. The maximum standard deviation of the presented experimental values is close to 3.5%. Fig. 4 shows the same experimental results in a different manner. Each figure represents a different salt concentration. What we want to emphasize with those figures is that at high pressures the 45 100 100 100 100 40 ˚C ˚C ˚C ˚C - 0.085 m NaCl 0.87 m NaCl 1.79 m NaCl 2.75 m NaCl 35 30 25 20 0 50 100 150 200 250 300 Pression (bar) Fig. 3. cb;CO2 as a function of pressure for different NaCl molal concentrations at (a) T = 27 °C, (b) T = 71 °C and (c) T = 100 °C. cb;CO2 reaches a constant value that does not depend on the pres- sure nor on the temperature. At high pressures, CO2 becomes nearly incompressible (constant Dq). At lower temperatures, this incompressibility is reached at a lower pressure. A similar evolution of CO2 solubility in water and brine with pressure has been reported [38–40] (additional CO2 solubility in brine from an increment in pressure greatly decreases at higher pressures). This clearly shows that there is an important reduction of phase and solubility effects on the interfacial tension, that could explain the existence of plateau values in the cb;CO2 after a given pressure. This value therefore depends only on the temperature. These results point out the possibility of extrapolating cb;CO2 for higher pressures and establishing a value of 26 mN/m (cWplateau) for pure water or very low salinities and other values according only to the salt a 50.00 b Interfacial Tension (mN/m) C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 Interfacial Tension (mN/m) 102 45.00 27˚C 0.085m NaCl 71˚C 0.085m NaCl 40.00 100˚C 0.085 m NaCl 35.00 30.00 25.00 50.00 45.00 27 ˚C 0.87 m NaCl 71 ˚C 0.87 m NaCl 100˚C 0.87 m NaCl 40.00 35.00 30.00 25.00 20.00 20.00 0 50 100 150 200 250 0 300 50 100 Pressure (bar) Interfacial Tension (mN/m) c 150 200 250 300 Pressure (bar) 50.00 45.00 27 ˚C 1.79 m NaCl 71 ˚C 1.79 m NaCl 100 ˚C 1.79 m NaCl 40.00 35.00 30.00 25.00 20.00 0 50 100 150 200 250 300 Pressure (bar) Interfacial Tension (mN/m) d 50.00 45.00 27 ˚C 2.75 m NaCl 71 ˚C 2.75 m NaCl 100 ˚C 2.75 m NaCl 40.00 35.00 30.00 25.00 20.00 0 50 100 150 200 250 300 Pressure (bar) Fig. 4. cb;CO2 as a function of pressure for different temperatures and NaCl concentrations (a) 0.085 m (5 g/L), (b) 0.87 m (50 g/L), (c) 1.79 m (100 g/L) and (d) 2.75 m (150 g/L). concentration. The applicability of this suggestion at temperatures near or above 100 °C would require cb;CO2 experimental values for pressures values greater than 300 bars. 2.4. Brine–CO2 IFT vs. salt concentration According to previous works mentioned in the introduction, we found a linear relationship between the increase in brine–CO2 interfacial tension and the molal salt concentration. This increase becomes more important at higher temperatures, as shown in Fig. 5a. When the cb;CO2 plateau is reached, at T = 27 °C and 71 °C, the IFT increase rate is reduced and the difference of these increase rates with temperature is negligible. Thus it is possible to establish an IFT trend, independent of the temperature, once the plateau is reached (Fig. 5b). Before reaching the plateau, at T = 27 °C, this linear relationship is as follows: dc ¼ 1:49 m ð3Þ where dc represents the increases of the IFT and m represents the molal concentration of salt. At T = 71 °C and 100 °C, slopes dc/m of the linear relationship are steeper (see Fig. 5). After reaching the plateau, we found a unique slope dc/m (independent of the temperature) equal to 1.43. Fig. 6 shows the brine surface tension increase as a function of the molal concentration of NaCl from Argaud [26]. According to this figure, the IFT increase is given by a slope dc/m equal to 1.63. In the case of NaCl, this relationship is almost C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 Average IFT Increase (mN/m) a 8 y = 2.5305x 2 R = 0.9729 Average 27°C 7 Average 71°C 6 Average 100°C y = 2.2188x 2 R = 0.9945 5 4 y = 1.4943x 2 R = 0.994 3 2 1 0 0 0.5 1 1.5 2 2.5 3 NaCl Concentration (molality) Average IFT Increase (mN/m) b 4.5 4 3.5 y = 1.4311x 2 R = 0.9865 3 2.5 2 1.5 1 0.5 0 0 0.5 1 1.5 2 2.5 3 NaCl Concentration (molality) Fig. 5. Average increase of the cb;CO2 as a function of the NaCl molality for different temperatures. (a) Before reaching a plateau and (b) after the cb;CO2 plateau. 103 quence of Eq. (2). For KCl, MgCl2 and CaCl2, this increase is not linear at high molalities (>1.0 m). Nevertheless, in most of the cases seen, the concentration of these salts in brine is within the range of molalities inside which linear relationships has been reported. There is no available data to evaluate the interfacial tension increase for a system which contains different salts, for example to know whether this increase would be additive or not. Massoudi and King [20] found cb;CO2 values at T = 25 °C, under pressures up to 60 bars, for the brine–CO2 systems; these show a linear relationship between the cb;CO2 and the molal concentration of NaCl (up to 5 m). The dc/m slope value was close to 1.58 instead of 1.49 as reported in Eq. (3). They also found a very slight (negligible according to the authors) decrease of this coefficient with pressure. The evolution of the interfacial tension increase with molal concentration of NaCl in water is analogous to that reported by Massoudi and King [20] and those summarized by Argaud [26]. At T = 27 °C, and after reaching the plateau, the value of the coefficient obtained for the linear relationship is very close to those reported by the above cited authors. These results do not support a proportional relation of the IFT increase with the Kelvin temperature suggested by Aveyard and Saleem [21], and Johansson and Eriksson [24]. In fact, the difference in the IFT increase observed between T = 71 °C and 100 °C is within the standard deviation of our results. Chiquet et al. [18] provided an isotherm of IFT measurement at salinity of 20 g/L and 34–35 °C for different temperatures which could be compared with our lowest salinity measurement. If we use the model to be presented in the next section to calculate the IFT at the same pressure, temperature and salinity conditions as Chiquet et al. [18], we find that the values estimated with our model are significantly higher than our values, about 10–15% higher. This difference could be attributed to differences in the density estimation, differences in the saturation state of the brine or the CO2 phase or to visualization artefacts. However since temperature and salinity of both studies are rather different it is difficult to go further in this comparison. In order to facilitate comparisons with other studies we showed in Table 1 all IFT obtained in this study including the density values used for each IFT. 2.5. IFT Modeling Fig. 6. Surface tension of brine as a function of the molal concentration of NaCl [26]. constant over a large range of molalities. Similar relationships exist for other chlorides at ambient conditions [26] and follow the se- The only existing model that can be found in the literature to predict cw;CO2 at storage conditions is, to our knowledge, the one described by Hebach et al. [9]. The proposed equation is obtained from a regression fit of experimental data. This equation presents several shortcomings. First, it is based on cw;CO2 calculated without taking into account the dissolved CO2 in the water phase. Second, it needs many fitting coefficients (nine parameters, including two exponential terms) and lacks physical basis (at very high pressures and low temperatures where the density difference is equal to zero the calculated cw;CO2 is also zero, which is not the case for the water–CO2 system). From our experimental results, we observed that the cb;CO2 is strongly correlated with the density difference, Dq (see Fig. 7). The density difference takes into account the effect of pressure and temperature, as well as the effect of the salt presence. Nevertheless, we consider that the Dq is not the only parameter to predict the cb;CO2 . In Fig. 8 we show that for the same Dq, different IFT can be obtained, and that this is mainly the consequence of salt concentration. We also show that the existence of a plateau in the cb;CO2 is strongly related with the density difference. For density differences below 0.6 g/cm3 a stabilization in the cb;CO2 has been observed where the only parameter that affects the cb;CO2 is the salt concentration. We believe that in order to model the interfacial tension, it is necessary to separate its variation from the density difference into two parts (see Fig. 8): 104 C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 Table 1 (continued) Table 1 Experimental interfacial tension values. Temperature (°C) Pressure (bar) Salinity (m) Brine density (g/cm3) CO2 density (g/cm3) IFT (mN/m) 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 27 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 71 240.1 205 175 120 106.5 82.2 68.5 61 48 247.5 194.7 163.9 126 109.7 95 82 68.5 60.3 52.7 226 206 151.9 127 109 93 81.5 69.2 245.5 225.5 202 176.1 147.4 123.6 107.3 82 242.7 207 175 142.9 119.8 107.1 91.3 67.7 61.7 51.1 252.6 225.1 197.2 166.7 135 95.6 81 66.5 59.9 246 225.5 201 175.5 151 130.5 114 95 82 67.7 61.8 50.7 232 204.7 174.5 149.3 121 107.8 94 82.1 67.7 0.085 0.085 0.085 0.085 0.085 0.085 0.085 0.085 0.085 0.87 0.87 0.87 0.87 0.87 0.87 0.87 0.87 0.87 0.87 1.79 1.79 1.79 1.79 1.79 1.79 1.79 1.79 2.75 2.75 2.75 2.75 2.75 2.75 2.75 2.75 0.085 0.085 0.085 0.085 0.085 0.085 0.085 0.085 0.085 0.085 0.87 0.87 0.87 0.87 0.87 0.87 0.87 0.87 0.87 1.79 1.79 1.79 1.79 1.79 1.79 1.79 1.79 1.79 1.79 1.79 1.79 2.75 2.75 2.75 2.75 2.75 2.75 2.75 2.75 2.75 1.0241 1.0223 1.0207 1.0177 1.0169 1.0154 1.0143 1.0135 1.0115 1.049 1.0477 1.0462 1.0443 1.0435 1.0428 1.0422 1.0411 1.0405 1.0397 1.0772 1.0764 1.0741 1.0729 1.0721 1.0714 1.0708 1.0701 1.1051 1.1043 1.1035 1.1024 1.1012 1.1003 1.0995 1.0985 1.0243 1.0225 1.0208 1.0191 1.0178 1.017 1.0161 1.0143 1.0137 1.0122 1.0296 1.0282 1.0268 1.0251 1.0232 1.0202 1.018 1.0176 1.0168 1.0564 1.0555 1.0542 1.053 1.0517 1.0506 1.0496 1.0483 1.0474 1.0463 1.0458 1.0448 1.0808 1.0807 1.0794 1.0782 1.0768 1.076 1.0752 1.0745 1.0735 0.92899 0.90782 0.8861 0.83071 0.81089 0.75705 0.68701 0.18805 0.11989 0.93326 0.90112 0.87723 0.83869 0.81611 0.79073 0.75701 0.68783 0.18336 0.14009 0.92147 0.90885 0.86616 0.83957 0.81586 0.7852 0.75475 0.69305 0.93155 0.92015 0.9054 0.88645 0.86143 0.83465 0.81205 0.75636 0.72255 0.66684 0.5905 0.46303 0.339 0.2759 0.21075 0.1365 0.12021 0.094864 0.73507 0.69725 0.64657 0.56352 0.42125 0.22692 0.1753 0.13293 0.11583 0.72684 0.69803 0.65445 0.59192 0.50142 0.39671 0.30879 0.22462 0.17863 0.13637 0.12056 0.093954 0.70776 0.66191 0.58879 0.49225 0.34501 0.27892 0.2205 0.17893 0.1362 26.00 25.77 25.73 26.72 27.93 27.53 27.51 33.37 37.75 27.53 27.52 28.77 27.88 29.40 29.15 28.80 28.21 35.43 38.38 28.32 28.16 27.54 27.00 30.05 28.23 26.65 30.30 30.02 29.78 29.58 29.79 30.29 29.67 29.26 30.64 24.78 26.78 26.52 29.24 29.80 30.30 32.59 37.50 39.50 41.13 27.13 27.59 28.18 27.74 28.09 32.50 37.54 40.96 41.79 28.67 29.69 29.04 29.25 30.04 32.15 33.54 36.66 39.45 42.68 44.92 46.77 29.96 29.95 29.19 30.19 33.87 35.95 38.29 40.54 43.79 Temperature (°C) Pressure (bar) Salinity (m) Brine density (g/cm3) CO2 density (g/cm3) IFT (mN/m) 71 71 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 62.2 51.1 236 226.2 202.3 170.5 143.7 122.5 96 68 258 214.1 181.5 153 119.5 103.3 83.5 63.2 254 189.2 169 137.5 109.3 84.1 63.5 227.6 207.3 161.5 131.2 106.7 84.8 64.3 2.75 2.75 0.085 0.085 0.085 0.085 0.085 0.085 0.085 0.085 0.87 0.87 0.87 0.87 0.87 0.87 0.87 0.87 1.79 1.79 1.79 1.79 1.79 1.79 1.79 2.75 2.75 2.75 2.75 2.75 2.75 2.75 1.0732 1.0723 0.9838 0.9832 0.9816 0.9794 0.9773 0.9756 0.973 0.9702 1.0111 1.0087 1.0067 1.0048 1.0024 1.001 0.9994 0.9974 1.0345 1.0348 1.0348 1.0319 1.03 1.0282 1.0266 1.0631 1.0622 1.06 1.0584 1.0569 1.0556 1.0543 0.12153 0.094864 0.56247 0.5424 0.4864 0.39624 0.31241 0.24948 0.17862 0.11567 0.60223 0.5157 0.42885 0.34175 0.24081 0.19696 0.14912 0.10595 0.59536 0.4512 0.39164 0.29365 0.2128 0.15036 0.1065 0.54541 0.49907 0.36846 0.27508 0.20579 0.15208 0.10815 44.79 47.87 26.03 26.14 26.98 27.94 29.90 31.33 35.12 38.15 25.97 27.30 29.23 29.88 32.38 34.85 38.26 41.47 27.65 31.41 32.45 34.78 37.67 42.00 44.63 29.57 31.11 33.64 35.96 39.43 43.19 45.13  the first part (high Dq) could be correlated with equations similar to the Parachor model for pure compounds [41–47];  the second part (low Dq), which corresponds to the plateau in interfacial tension, has to be correlated using the cw;CO2 , once the plateau is reached (cWplateau) and a linear relationship to match the increase due to dissolved salts. Then, cb;CO2 can be explicitly described as cb;CO2 ¼ cWplateau þ k xNaCl þ  g P ðDqÞ  T br M ð4Þ where k, b, g are regression coefficients obtained from a leastsquares fit of our experimental data. The values of these coefficients are presented in Table 1. P, M and cWplateau are constant values presented in Table 2. P and M are the Parachor number and molar mass of CO2, respectively. The use of an exponent different than four for the density difference can be seen as arbitrary (according to recent results, this exponent value is 3.8842 [48]); Schechter and Guo [41] explained in a comprehensive review of the Parachor model and its uses in IFT prediction of fluids, that the IFT of several systems can deviate from a slope of 3.88 to higher values. This behavior is attributed to some molecules which rapidly adsorb at the interface. In this case, the IFT significantly reduces, even if the density difference between the bulk phases only changes very slightly (Fig. 7, part 1). Dealing with the coefficient of Tr, it is only used as a correction factor in the correlation which leads to an improvement of the regression (lower deviation) of the experimental results. The parameter k (1.255) obtained from the regression is lower than those linear relationships presented in Eq. (3). This can be explained by the use of the Tr as a correction factor and by the fact that part of the salt effect on IFT is already contained in the brine phase density. The correlation proposed by Eq. (4) takes into account pressure, 105 C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 temperature and salt effects. Fig. 8a shows the modeled cb;CO2 using Eq. (4) vs. the experimental values in the same conditions. From the regression fit of our experimental data, we can predict the cb;CO2 using only three regression coefficients with a mean deviation of 2.5% and more that 96% of the calculated values below a deviation of 6% from the experimental values (see Fig. 8b). 60 50 IFT (mN/m) 2 1 40 30 Slope power law > 4 20 3. Wettability 10 0 0 0.2 0.4 0.6 0.8 1 Density Difference (g/cm3) Fig. 7. Variation of the interfacial tension of brine–CO2 with density difference. 3.1. Experimental set-up a 50 45 40 IFT (mN/m) 35 30 25 20 0.085 m 1.79 m 0.085 m modeled 1.79 m modeled 15 10 5 0.87 m 2.75 m 0.87 m modeled 2.75 m modeled 0 0 0.2 0.4 0.6 0.8 1 1.2 Density difference (g/cm3) 8 6 results (%) 4 2 0 -2 0 0.2 0.4 0.6 A two-dimensional heterogeneous pore structure is etched onto the surface of a completely flat glass plate. The size of the 2D pore network is 6.55 cm  1.25 cm. The thickness of the pore structure is only one pore size, about 0.3 mm. Glass plates are naturally water-wet. In order to change the surface to strongly oil-wet, we treated the surface with a silane. Another micromodel, naturally water-wet, was treated to intermediate-wet by ageing the surface with an asphaltic crude oil. In both cases, wettability conditions were tested by means of contact angle measurements on treated glass plates. The pore distributions of the oil-wet and the intermediate-wet micromodel were identical. A schematic diagram of the experimental set-up is given in Fig. 9. The micromodel can be operated up to 100 bars and 60 °C. Fig. 10 shows the contact angle measurements obtained at reservoir conditions for each micromodel. 3.2. Experimental procedure 10 b Deviation from experimental A pressurized glass micromodel was used to visualize the phase distribution and mobilization during flooding a water saturated porous media by CO2. Different thermodynamic conditions were investigated to cover the three physical states of CO2: gas, liquid and supercritical. Three wettability conditions were investigated: water-wet (WW), intermediate wet (IW) and oil wet (OW, the less hydrophilic surface). 0.8 1 -4 -6 -8 The following procedure was followed for all the reported tests. Initially, the micromodel was saturated with distilled water at ambient conditions. Once the micromodel and the CO2 were stabilized at a given pressure and temperature, the CO2 was injected at a low flow rate (1 cm3/h) and the evolution of the phase distribution was recorded. After about 30 min, the phase distribution stabilized. Changes observed during this lapse of time correspond to the evolution of the saturations of the pore network until a stabilized CO2 saturation. Those changes depend on the thermodynamic conditions and on the dissolution of CO2 into water. The results reported in this study are images obtained after this stabilization. Once this static state was recorded along the micromodel, a flush of distilled water was performed in order to obtain the initial state, and then the desired pressure value was adjusted. -10 Density Difference (cm3/g) Fig. 8. (a) Modeled cb;CO2 vs. experimental values. (b) Correlation % deviation from the experimental data. Camera CO2 Oven Table 2 Regression coefficients and constants used to model the interfacial tension. Coefficients of the regression fit (Eq. (4)) k g b 1.255 4.7180 1.0243 Moving Direction H2 O Micromodel & Housing Monitor Video Recorder Constant values used in Eq. (4) P M (g/mol) cWplateau (mN/m) 82 44.01 26 Light Fig. 9. Experimental set-up for micromodel visualization. 106 C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 Fig. 10. Contact angle measurements air–pure water at ambient conditions: (a) water-wet micromodel, (b) intermediate-wet micromodel and (c) oil-wet micromodel. 3.3. Experimental results Figs. 11–14 present the phase distribution after stabilization for CO2 flooding of water saturated micromodels. For gaseous and supercritical CO2, phases can be identified as follows: the darkest phase is always the CO2 phase. For liquid CO2, it is a little more difficult since its color is very close to the one corresponding to the water phase. For all micromodel pictures, the average size of the circular grains (circles) is 0.3 mm. 3.3.1. Experiments with water-wet micromodels For water-wet micromodels and low pressures (gaseous CO2), we have observed very thin water films surrounding the solid sur- face (Figs. 11 and 12a). At higher pressures (Fig. 12b and c), we cannot observe such films. Roughness could affect the observation of these films. However, it is important to note that the same micromodel was used in these three experiments. The evolution of the film thickness is directly related to the relative affinity between water, CO2 and the solid substrate. The estimation of this affinity is not straightforward since it requires many physicochemical parameters to be taken into account, most of them not being available in the literature. Chiquet et al.’s work [30] presented in the Introduction could contribute to explaining this thickness reduction. The authors attribute this behavior to a reduction in electrostatic interaction that tends to stabilize the water films. Despite this behavior, the shape of the interfaces (Fig. 12) Fig. 11. Gaseous CO2 (P = 5 bars, T = 20 °C): (a) water-wet micromodel; (b) intermediate-wet. Fig. 12. Water-wet micromodel: (a) gaseous CO2 (P = 57.9 bars, T = 20 °C), (b) supercritical CO2 (P = 105.4 bars, T = 60°C), (c) liquid CO2 (P = 100 bars, T = 23 °C). C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 107 Fig. 13. Oil-wet micromodel: (a) gaseous CO2 (P = 51.3 bars, T = 19 °C), (b) supercritical CO2 (P = 100 bars, T = 60 °C), (c) liquid CO2 (P = 100 bars, T = 23 °C). Fig. 14. Intermediate-wet micromodel: (a) gaseous CO2 (P = 60 bars, 25 °C), (b) supercritical CO2 (P = 100 bars, T = 60 °C), (c) liquid CO2 (P = 100 bars, T = 25 °C). shows that there is no transition from a water-wet system to an intermediate-wet system while increasing pressures up to 100 bars. That is to say that if the solid was originally water-wet, it keeps strong water wettability at higher pressures, and water always remains a connected phase. This is coherent with the wettability indices deduced by Egermann et al. [31] from the injection of CO2 performed in carbonate samples and with Chiquet et al. [30] results on mica and quartz substrates presented in the Section 1. probably be related to a higher dissolution of CO2 into water (lower pH) which reduces the electrostatic forces that tend to stabilize the water films. Our observations in intermediate-wet and oil-wet micromodels are consistent with contact angle observations from Dickson et al. [49] which show a larger contact angle variation with a more hydrophobic glass plate. 4. Discussion 3.3.2. Experiments with oil-wet and intermediate-wet micromodels In Fig. 11b, it is not possible to observe water films around the pores. This contrasts with the case observed under the same thermodynamic conditions in a water-wet micromodel (Fig. 11a). The results obtained for these two wettability conditions are similar: at low pressures, the water is still the wetting phase. This corresponds to the current assumption whereby gas is considered to be the non-wetting phase, compared to a liquid. Nevertheless, it is important to mention that some interfaces (Figs. 13 and 14a) show that this water wettability was weaker for the intermediate-wet micromodel. The experiments performed at higher pressures (Figs. 13b, 13c, 14b and 14c) show a different phase distribution. In some regions the water can be observed as a dispersed phase. Moreover, the shape of the interfaces shows that in this case the CO2 behaves as the wetting phase. In these figures it can be observed that the wettability of the CO2 is stronger at low temperatures. This could The first part of this paper presented a complete set of experimental values of interfacial tension for brine–CO2 systems at pressure and temperature conditions that can be considered as representative of those of a geological CO2 storage operation. The salt used for these results was NaCl. A correlation was proposed from experimental data, in order to model IFT values with a low deviation. The objective of this systematic study is to enable reservoir engineers to estimate the interfacial tension of brine–CO2 in reservoir conditions, it can also be used to evaluate and understand the evolution of this property with different physical parameters: temperature, pressure and salinity. The available data in the literature which could be used to estimate IFT of water–CO2 do not take into consideration the effect of dissolved CO2 on the aqueous phase density. At high pressures and low temperatures, the consideration of this effect becomes more important because the density difference between 108 C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 the brine and the CO2 phase is minimal. For example: at T = 27 °C and P = 240 bars the density difference is 14% lower if the dissolved CO2 is not taken into account in the estimation of brine phase density. Because density difference enters linearly in the IFT calculation, this factor could explain why at high pressures (once the plateau has been reached) there is a difference in cw;CO2 reported values between low and high temperatures [9]. This indicates that there is a strong risk that available cw;CO2 data are underestimated, especially at low temperatures. Concerning the implications of salt effects, we observed that for salt concentrations above 30 g/L of NaCl, the induced IFT increase is higher than the standard deviation, therefore it must not be considered as negligible. Additionally, CaCl2 or MgCl2 (both present in carbonate reservoirs) are reported to have an effect on surface tension [26]; if those values are extrapolated to high pressure and a wide range of temperatures for the brine–CO2 system (as we showed is the case of the NaCl), the increase due to the presence of these cations in brine would be near double that of NaCl in storage conditions. This means that the increase would be close to 8 mN/m for the same concentration of CaCl2 or MgCl2, instead of an increase of 4 mN/m measured for 150,000 ppm of NaCl, at pressures over 150 bars and temperatures ranging between 27 °C and 71 °C. This increase is as high as 30% of the cw;CO2 value. Therefore, the estimation of the cb;CO2 in storage conditions, without taking into account the dissolved CO2 in the aqueous phase and the effect of salts, strongly underestimates the interfacial tension. If the cb;CO2 is underestimated, the threshold capillary pressure (Pth c ), calculated from Eq. (1), which determines the CO2 breakthrough in the caprock and the height of CO2 column stored, is consequently underestimated and can lead to an underestimation of the storage capacity of a given reservoir. In the case of CO2 storage in a deep saline aquifer, cw;CO2 governs fluid distribution in the porous media and is taken by any reservoir simulation as a key parameter to estimate the water displacement by a CO2 injection. Therefore, any uncertainty related to the estimation of the interfacial tension could lead to misleading results in terms of storage capacity and sealing efficiency of a given storage site. Nevertheless, since CO2 flow in an aquifer is dominated by large viscous instability that is inherent in CO2 displacing brine it is difficult for the authors to establish quantitative effect in terms of sweep efficiency at the reservoir scale. At the core scale Egermann et al. [31] showed that the underestimation of cb;CO2 can lead to an overestimation of the displacing efficiency of a CO2 flooding in a deep saline aquifer, hence it increases the amount of displaced brine, and could erroneously increase the available reservoir storage volume for CO2. In the second part of this study, laboratory experiments were performed to provide data and knowledge in the case of CO2 injection in deep saline aquifers and depleted oil fields, in different wettability conditions. These experiments were carried out under thermodynamic conditions representative of CO2 and treated the pore scale by means of a qualitative study in glass micromodels. The observations at the pore scale in glass micromodels show a partial wetting behavior of the CO2 at reservoir conditions, in treated micromodels. At the pore scale, it is visible in the shape of the interfaces and the fact that the water phase is a dispersed phase. Among the effects of a partial wetting of the CO2 there is a lower capillary breakthrough pressure of the caprock as can be deduced from Eq. (1). This has negative impact on the capacity of the reservoir, hence on the amount of CO2 that can be stored. The partial wetting of the CO2 could lead to an improvement in the CO2 displacement efficiency compared to a displacement in a strongly water wet porous media. As we showed in a previous study [29], the partial wetting of CO2 has some effect at the reservoir scale; different wettability scenarios lead to different spatial extents of the injected CO2 and injectivity index. 5. Conclusions and future works In this study we showed that it is possible to predict the cb;CO2 using a correlation with few input variables. The developed correlation takes into account the influence of the thermodynamic conditions and the presence of salt. Bearing in mind that the maximum standard deviation of our experimental results is about 3.5% and the mean deviation of the correlation is 2.5%, we consider that this correlation is highly suitable for the cb;CO2 prediction in the classical conditions where CO2 is injected. Additional IFT experimental data at higher pressures are needed in order to know if the proposed correlation is suitable for deeper reservoir depths. This part of our study was focused on the influence of CO2 wettability on a drainage process. In our experiments we showed a partial wetting behavior of CO2 on hydrophobic glass surfaces at reservoir conditions. This has a positive impact in terms of a more uniform spatial distribution of the CO2 plume and lower injectivity indexes and a negative impact in terms of caprock sealing efficiency. The imbibition process has not been investigated. The knowledge and quantification of the kr hysteresis between the drainage and the imbibition curves is fundamental to estimate the capillary trapping of CO2 after the injection phase and to understand and quantify better the hysteresis of the kr curves [5,50]. This issue was investigated by simulation [5,51]. Nevertheless; few specific experimental relative permeability determinations are available. Therefore, future experimental works should be performed to further investigate drainage and imbibition processes, in order to improve the accuracy of the estimation of the amount of CO2 that can safely been injected. Acknowledgments The authors wish to thank IFP for permission to publish these results. The authors also acknowledge C. Féjean and V. Ruffier for the estimation of phase densities. M. Argaud for fruitful discussions and J. Béhot and S. Shaïek for her key contribution to the experimental work. References [1] International Energy Agency. Prospects for CO2 capture and storage, Paris: OECD/IEA; 2004. [2] Orr Jr. FM. Storage of carbon dioxide in geologic formations, paper SPE 88842. Distinguished author series; 2004. [3] Egermann P, Lombard J-M, Bretonnier P. A fast and accurate method to measure threshold capillary pressures under representative conditions, SCA A46. Paper presented at the 2006 SCA international symposium, Trondheim, September 18–22; 2006. [4] Spiteri EJ, Juanes R, Blunt MJ, Orr FM. Relative permeability hysteresis: trapping models and application to geological CO2 sequestration, SPE 96448 2005. Paper presented at the annual technical conference and exhibition, Dallas, October 9–12. [5] Juanes R, Spiteri EJ, Orr FM, Blunt MJ. Impact of relative permeability hysteresis on geological CO2 storage. Water Resour Res 2006;42:12. [6] Heuer GJ. Interfacial tension of water against hydrocarbon and other gases and adsorption of methane and solids at reservoir conditions, PhD dissertation, The University of Texas at Austin; 1957. [7] Chun BS, Wilkinson GT. Interfacial tension in high pressure carbon dioxide mixtures. Ind Eng Chem Res 1995;34:4371–7. [8] Harrison K. Interfacial tension measurements of CO2-polymer and CO2–water systems and formation of water-in-CO2 microemulsions, PhD dissertation, The University of Texas at Austin; 1996. [9] Hebach A, Oberhof A, Dahmen N, Kögel A, Ederer H, Dinjus E. Interfacial tension at elevated pressures-measurements and correlations in the water + carbon dioxide system. J Chem Eng Data 2002;47:1540–6. [10] Ennis-King J, Paterson L. Role of convective mixing in the long term storage of carbon dioxide in deep saline formations, SPE 84344. Paper presented at the 2003 SPE annual technical conference and exhibition, Denver, October 5–8. [11] Da Rocha SRP, Johnston KP, Westacott RE, Rossky. Molecular structure of the water–supercritical CO2 interface. J Phys Chem B 2001;105:1192–214. [12] Kuznetsova T, Kvamme B. Thermodynamic properties and interfacial tensions of a model water–carbon dioxide. J Phys Chem 2002;4:937–41. C. Chalbaud et al. / Advances in Water Resources 32 (2009) 98–109 [13] Yan W, Zhao G-Y, Chen G-J, Guo T-M. Interfacial tension of (methane + nitrogen) + water and (carbon dioxide + nitrogen) + water systems. J Chem Eng Data 2001;46:1544–8. [14] Zuo Y-X, Stenby E. A linear gradient theory model for calculating interfacial tension of mixtures. J Colloid Interf Sci 1996;182:126–32. [15] Zuo Y-X, Stenby E. Calculation of interfacial tensions of hydrocarbon–water systems under reservoir conditions. In Situ 1998;22:157–80. [16] Yang D, Gu Y. Interfacial interactions of crud oil–brine–CO2 systems under reservoir conditions, SPE 90198. Paper presented at the 2004 SPE annual technical conference and exhibition, Houston, September 26–29. [17] Bennion B, Bachu S. The impact of interfacial tension and pore-size distribution/capillary pressure character on CO2 relative permeability at reservoir conditions in CO2 brine system, SPE 99325. Paper presented at the 2006 SPE/DOE symposium on improved oil recovery, Tulsa, April 22–26. [18] Chiquet P, Daridon J-L, Broseta D, Thibeau S. CO2/water interfacial tensions under pressure and temperature condition of CO2 geological. Energy Convers Manage 2006;48(3):736–44. [19] Washburn ED. International critical tables, vol. 4. NY: McGraw-Hill; 1928. [20] Massoudi R, King Jr. Effect of pressure on the surface tension of aqueous solutions. Adsorption of hydrocarbon gases, carbon dioxide, and nitrous oxide on aqueous solutions of sodium chloride and tetra-n-butylammonium bromide at 25 °C. J Phys Chem 1975;79(16):1670–5. [21] Aveyard R, Saleem SM. Interfacial tension at alkane–aqueous electrolyte interfaces. J C S Faraday 1975;73:1613–7. [22] Cai B-Y, Yang J-T, Guo T-M. Interfacial tension of hydrocarbon + water/brine system under high pressures. J Chem Eng Data 1996;41:493–6. [23] Ralston J, Healy TW. Specific cation effects on water structure at the air–water and air–octadecanol monolayer-water interfaces. J Colloid Interf Sci 1973;42(3):629–44. [24] Johansson K, Eriksson JC. c and dc/dT Measurements on aqueous solutions of 1.1 electrolytes. J Colloid Interf Sci 1974;49(3):469–80. [25] Harkins WD, McLaughlin HM. The structure of films of water on salt solutions surface tension and adsorption for aqueous solutions of sodium chloride. J Am Chem Soc 1925;47:2083–9. [26] Argaud MJ. Predicting the interfacial tension of brine/gas (or condensate) systems, SCA. Paper presented at the 1992 European core analysis symposium, Paris, September 14–16. [27] Drummond C, Israelachvili J. Fundamental studies on crude oil surface interactions and its relationship to reservoir wettability. J Petrol Sci Eng 2004;45:61–81. [28] Craig FF. The reservoir engineering aspects of waterflooding. Monograph series SPE, vol. 3. Texas: Richardson; 1971. [29] Chalbaud C, Robin M, Békri S, Egermann P. Wettability impact on CO2 storage in aquifers: visualisation and quantification using micromodel tests, pore network model and reservoir simulations, SCA A16. Paper presented at the 2007 SCA international symposium, Calgary, September 12–15. [30] Chiquet P, Broseta D, Thibeau S. Wettability alteration of caprock minerals by carbon dioxide. Geofluids 2007;7:112–22. [31] Egermann P, Chalbaud C, Duquerroix J-P, Le Gallo Y. An integrated approach to parameterize reservoir models for CO2 injection in aquifers, SPE 102308. Paper View publication stats 109 presented at the annual technical conference and exhibition, San Antonio, September 24–27; 2006. [32] Siemons N, Bruining H, Castelijns H, Wolf K-H. Pressure dependence of the contact angle in a CO2-H2O-coa systel. J Colloid Int Sci 2006;297(2): 755–61. [33] Søreide I, Whitson CH. Peng-Robinson predictions for hydrocarbons, CO2, N2, H2S with pure water and NaCl brine. J Fluid Phase Equilib 1992;77:217–40. [34] Rowe AM, Chou JCS. Pressure–volume–temperature–concentration relation of aqueous NaCl solutions. J Chem Eng Data 1970;15:61–6. [35] King MB, Mubarak A, Kim JD, Bott TR. The mutual solubilities of water with supercritical and liquid carbon dioxide. J Supercrit Fluids 1992;5:296–302. [36] Lahooti S, del Río OI, Cheng P, Newmann AW. In: Neumann AW, Spelt JK, editors. Applied surface and thermodynamics, vol. 1. New York: Marcel Dekker; 1996. p. 441. [37] Jönsson B, Lindman B, Holmberg K, Kronberg B. Surfactant and polymers in aqueous solution. New York, NY: John Wiley and Sons Inc.; 1998. 263. [38] Wiebe R. The binary system carbon dioxide–water under pressure. Chem Rev 1941;29:475. [39] Malinin SD, Savelyeva NI. The solubility of CO2 in NaCl and CaCl2 solutions at 25, 50 and 75 °C under elevated CO2 pressures. Geochem Int 1972;9(1):410. [40] Malinin SD, Kurovskaya NA. Solubility of CO2 in chloride solutions at elevated temperatures and CO2 pressures. Geochem Int 1975;2(2):199. [41] Schechter DS, Guo B. Parachors based on modern physics and theirs uses in IFT prediction of reservoir fluids, SPE 30785. Paper presented at the 1995 SPE annual technical conference and exhibition, Dallas, October 22–25. [42] Firoozabadi A, Ramey HJ. Surface tension of water–hydrocarbon systems at reservoir conditions. J Can Petrol Tech 1988;27:3. [43] McLeod DB. On a relation between surface tension and density. Trans Farad Soc 1923;19:38–43. [44] Fowler RH. A tentative statistical theory of McLeod’s equation for surface tension and the Parachor. Proc Roy Soc London 1937:229–46. Series A. [45] Reno GJ, Katz DL. Surface tension of n-heptane and n-butane containing dissolved nitrogen. Ind Eng Chem 1943;35(10):1091–3. 46 Rowlinson J, Widow B. Molecular theory of capillarity. Oxford, UK: Oxford University Press; 1982. [47] Widow B. Phase equilibrium and interfacial structure. Chem Soc Rev 1985:121–39. [48] Moldover MR. Interfacial tension of fluids near critical points and two scalefactor universality. Phys Rev 1985;31:1022. [49] Dickson JL, Gupta G, Horozov TS, Binks BP, Johnston KP. Wetting phenomena at the CO2/water/glass interface. Langmuir 2006;22:2161. [50] Imbus S, Orr FM, Kuurskraa VA, Kheshgi H, Bennaceur K, Gupta N, et al. Critical issues in capture and storage: findings of the SPE advanced technology workshop (ATW) on carbon sequestration, SPE 102968. Paper presented at the annual technical conference and exhibition, San Antonio, September 24–27; 2006. [51] Flett M, Gurton R, Taggart I. The function of gas–water relative permeability hysteresis in the sequestration of carbon dioxide in saline formations, SPE 88485. Paper presented at the Asia Pacific oil and gas conference and exhibition, Perth, October 18–20; 2004.