Thermodynamic Properties of a Gas–Liquid–Solid System during the CO₂ Geological Storage and Utilization Process: A Review

Abstract

Emission reduction in the main greenhouse gas, CO₂, can be achieved efficiently via CO₂ geological storage and utilization (CCUS) methods such as the CO₂ enhanced oil/water/gas recovery technique, which is considered to be an important strategic technology for the low-carbon development of China’s coal-based energy system. During the CCUS, the thermodynamic properties of the CO₂–water–rock system, such as the interfacial tension (IFT) and wettability of the caprock, determine the injectability, sealing capacity, and safety of this scheme. Thus, researchers have been conducting laboratory experiments and modeling work on the interfacial tension between CO₂ and the water/brine, wettability of caprocks, the solubility of gas–liquid binary systems, and the pH of CO₂-saturated brine under reservoir temperature and pressure conditions. In this study, the literature related to the thermodynamic properties of the CO₂–water–rock system is reviewed, and the main findings of previous studies are listed and discussed thoroughly. It is concluded that limited research is available on the pH of gas-saturated aqueous solutions under CO₂ saline aquifer storage conditions, and less emphasis has been given to the wettability of the CO₂–water/brine–rock system. Thus, further laboratory and modeling research on the wettability alternations of caprock in terms of molecular dynamics is required to simulate this phenomenon at the molecular level. Moreover, simplified IFT and solubility prediction models with thermodynamic significance and high integrity need to be developed. Furthermore, interaction mechanisms coupling with multi-factors associated with the gas–liquid–solid interface properties and the dissolution and acidification process need to be explored in future work.

1. Introduction

Recently, countries worldwide have announced their goal of attaining a carbon peak and neutralization to tackle climate change caused by anthropogenic CO₂ emissions into the atmosphere using vast amounts of fossil fuel energy. In science, coal, oil, and gas still constitute a large percentage of the energy supply and conception chain, and the technical path of carbon emission reduction while ensuring that a safe energy supply is attracting increased attention in the field of energy usage. The growth in the proportion of renewable energy sources, such as wind and solar power, in the energy supply and conceptions chain, as well as the development of carbon capture, storage, and utilization (CCUS), are recognized as the key factors in achieving net zero carbon emissions [1,2]. Currently, deep saline aquifers are considered the most promising and effective method of CO₂ emission reduction via CO₂ geological storage because of their large capacity, which is estimated to be approximately 1000–10,000 GT, which is more than the total amount of anthropocentric CO₂ emissions produced this century. Another promising option is CO₂ storage in offshore reservoirs because of their vast uptake capacity and low risk of leakage into offshore aquifers [3].

Since 2000, China has been developing CCUS technology, which is considered to be one of the most important strategic technologies for the low-carbon development of China’s coal-based energy system. According to the 2021 China CCUS Annual Report, about 5.1 GT and 9.0 GT CO₂ can be sequestered in China via CO₂-Enhanced Oil Recovery (CO₂-EOR) [4] and CO₂-Enhanced Gas Recovery (CO₂-EGR) technologies, respectively [5]. Additionally, by utilizing depleted gas reservoirs, approximately 15.3 GT CO₂ can be stored, while injection into deep saline aquifers offers an even greater sequestration potential, estimated at around 2420 GT [5]. However, the implementation of domestic CO₂ geological utilization and storage technology currently faces significant challenges, including high energy consumption for carbon capture, source–sink mismatches, unclear long-term sequestration mechanisms, and underdeveloped leakage monitoring and emergency response systems. Moreover, domestic CCUS technology must integrate cutting-edge scientific and technological advancements from both local and international sources to establish industrial clusters that encompass CO₂ emission sources, capture, and geological storage, thereby creating a comprehensive and effective carbon emission reduction strategy.

At present, storage-related research has yielded promising results; however, demonstration projects and widespread applications are lacking. Thus, there is an urgent need to address key scientific questions concerning capture and storage, scale up the development of CO₂ geological utilization and storage technology, and ultimately provide a solid technical foundation for the low-carbon, clean, and efficient use of coal and other fossil energy sources, contributing to national energy security. One major reason for the launch of new CCUS projects is public concern about the environmental impacts and risks related to CCUS projects, which may directly influence the decision-making process during CCUS site selection [6]. Technical support is required for the development of CCUS, as expected, in terms of technologies supporting CO₂ saline aquifer storage, since leakage from the sealing site determines the flexibility of this scheme. Fundamental research on carbon migration in reservoirs, which is influenced by the thermodynamic properties of fluids, is essential to deepen the understanding of the mechanisms influencing CO₂ saline aquifer storage.

There are four main trapping mechanisms in the geological carbon storage process, including residual trapping, also known as capillary trapping, structural trapping, and solubility and mineral trapping, as shown in the Figure 1 [7,8,9]. Structural or hydrodynamic trapping plays an important role at the beginning of the injection of CO₂ into saline aquifers, where the injected CO₂ can be trapped and sealed via a low-permeability caprock or shale above the saline aquifers [10]. However, the trapped CO₂ is likely to escape into the atmosphere along the fractures of the caprock, leading to failure during structural trapping. Thus, structural integrity is crucial for the successful structural trapping of CO₂. In contrast, capillary trapping relies less on structural integrity, and the trapping of migrating CO₂ in pores can be achieved via snap-off effects [11]. Capillary trapping mechanisms have been widely studied because of their reliability, strong adaptability, and long-term efficiency during the CO₂ saline aquifers storage process [12]. It was found that the efficiency of capillary trapping is determined by the threshold capillary pressure, which, in turn, is a function of the interface properties of the CO₂-water-rock system, which determines the fluid distribution on the surface of the reservoir rocks [13]. Thus, the reservoir wettability and interfacial tension (IFT) between the CO₂ and reservoir brines are important thermodynamic properties of fluids that determine the injectability, sealing capacity, and safety of the CO₂ saline aquifer storage scheme.

Meanwhile, injected CO₂ dissolves in reservoir fluids and leads to a density gradient between different areas in the water layer [14], which eventually leads to the convective mixing of CO₂ and reservoir fluids. Then, the dissolution rate of CO₂ into the reservoir fluids is further increased due to the density gradient, resulting in CO₂ solubility trapping in the water layer of the reservoir [15]. During the process of CO₂ storage in saline aquifers, the process of CO₂ dissolving into reservoir brine lasts for most of the time. Thus, the solubility trapping mechanism plays a major role in this scheme. Moreover, the mutual solubility between CO₂ and reservoir fluids influences the long-term safe storage of CO₂ as well as the solid–liquid–gas interactions at the injected reservoir site [16]. Therefore, it is essential to investigate the mutual solubility of CO₂ and reservoir fluids during phase equilibrium under CO₂ saline aquifer storage conditions [17].

Owing to the slow chemical reactions with the reservoir rocks, the dissolved CO₂ is mineralized, which is known as mineralization trapping and is the most important trapping mechanism involved in CO₂ saline aquifer storage [18]. However, the dissolution of CO₂ into the reservoir fluids lowers the pH of the reservoir. The erosion of reservoir rocks and caprocks caused by the low pH of the reservoir results in the leakage of sealed CO₂ from the caprocks [19]. Generally, the pH of CO₂-saturated brine plays a key role in estimating the CO₂ storage capacity of the reservoir during mineralization trapping [20,21]. Therefore, fundamental research on the pH of CO₂-saturated brine during the CO₂ saline aquifer storage scheme may lead to stimulation failure or cause high risks in the operation process [22,23]. In addition to the main trapping mechanisms mentioned here, other CO₂ trapping mechanisms [24], such as hydrodynamic and carbon adsorption trapping and CO₂ trapping via changes in the partial capillary pressure [25] or permeability [26], have been studied.

The sealing site for the CCS covers extremely high temperatures and pressures, where CO₂ is in a supercritical state ($T_{Cr}=$ 304.2 K and $P_{Cr}=$ 7.4 MPa) and the salinity of the potential sealing site could be greater than 100 g/L [27,28,29]. The CO₂ phase diagram shown in Figure 2 illustrates that the injected CO₂ changes its phase from liquid/vapor to supercritical in the reservoir [30]. According to previous research, the CO₂ phase state determines its properties, leading to a change in the thermodynamic properties of reservoir fluids saturated with injected CO₂ [30]. Therefore, the laboratory apparatus must be designed to withstand the extremely high pressure, temperature, and corrosion associated with the high salt concentration.

After CO₂ is injected into a porous medium, a series of physical and chemical changes occur in the rock strata and brine [31]. These changes upset the balance between the existing rock strata and brine, resulting in a change in the trapping efficiency, as shown in Figure 3 [31].

Thus, fundamental research on the thermodynamic properties of gas–liquid–solid systems is important for estimating the safe and long-term storage of CO₂ saline aquifers. Thus, in this study, the experimental and predicted thermodynamic properties, such as the interfacial tension, wettability, pH, and solubility of the CO₂-brine-rock system, were reviewed carefully. Experimental and modeling methods were compared, methods for obtaining thermodynamic properties of a gas–liquid–solid system with high integrity were proposed, and trapping mechanisms were clarified. The results of this review provide theoretical support for the feasibility of a long-lasting and safe CO₂–saline aquifer storage scheme to provide a solid technical foundation for the low-carbon, clean, and efficient use of coal and other fossil energy sources, contributing to national energy security.

2. Thermodynamic Properties

2.1. Interfacial Tension between CO₂ and Water/Brine

The threshold capillary pressure that determines the efficiency of residual trapping is a function of the wettability and the interfacial tension (IFT) between the CO₂ and reservoir brines, as described by the standard Young–Laplace equation, as shown in Equation (1) [13,32]. This threshold capillary pressure characterizes the penetration ability of the non-wetting phase (CO₂) into the macropores and throats of the reservoir rock that were originally water saturated. Research findings suggest that lower IFT leads to enhanced CO₂ capillary trapping, as it reduces the capillary pressure required for CO₂ capture within pore spaces, thereby improving storage efficiency. Due to the complexity of the actual reservoir, the wettability of the sealing site can be represented by measuring the contact angles (CAs) of the CO₂-brine-rock system. Thus, the laboratory and modelling work of CO₂–brine IFTs and rock CAs have been reviewed, and the key findings are highlighted in this section.

$$P_{\mathrm{c}}^{\mathrm{th}}=\frac{2{\gamma }_{{\mathrm{w}/\mathrm{b},\mathrm{CO}}_2}\cos \theta }{R}$$

(1)

where P_c^th is the threshold capillary pressure of brine/water-saturated caprock ${\gamma }_{{\mathrm{w}/\mathrm{b},\mathrm{CO}}_2}$ (mN·m⁻¹) is the IFT between water/brine and CO₂, θ is the CA between the brine/water and rocks in a CO₂ atmosphere, and R is the maximum radius of cylindrical caprocks.

Different methods exist for measuring the IFT between gases and liquids. Among these, the pendent drop method, which is based on the axisymmetric drop shape analysis (ADSA) technique, is widely used because of its high accuracy and efficiency. Pictures of the pendent drop of the desired solution were captured using a CCD camera at a high resolution under the evaluated temperature and pressure conditions. However, the accuracy of the calculated IFTs based on this method strongly depends on the density difference between the gas and liquid, as illustrated in Equation (2).

$$\gamma =\frac{g\Delta \rho D_e^2}{H}$$

(2)

where g is the gravity constant; γ is the IFT; Δρ is the density difference; H is the shape facture, which is related to the drop shape parameters; and D_e is the horizontal diameter of the drop. Therefore, the accuracy of the obtained density difference determines the integrity of the IFT values measured using the pendent drop method.

Numerous experimental studies have been conducted on the IFT between CO₂ and reservoir fluids using the pendent drop method based on the ADSA technique [32,33,34,35]. Initially, most laboratory studies were conducted on CO₂ + water binary systems under reservoir conditions. Then, considering the effects of actual reservoir brine’s ionic strengths and types on the IFT between the CO₂-brine system, the IFTs of the CO₂ + brine binary system consistent with the reservoir conditions were measured, where temperatures and pressures ranged from 278 to 423 K and 0.1–70 MPa, respectively [32,33,34,36,37,38]. However, owing to the differences in the experimental procedures and methods, disparities and contradictions exist in the experimental results, which has led to doubts regarding the accuracy and reliability of the IFT data in the literature [39]. Accurate determination of the density difference between CO₂ and water/brine is critical to the IFT measurement results based on ADSA, as shown in Equation (1) [13]. Due to the mutual dissolution of CO₂ and water/brine, the density of the two-phase equilibrium must be considered when measuring the two-phase IFT [40]. Thus, the difference between density difference determination methods also leads to a disparity in the IFTs reported in the literature. It is found that the IFT of the CO₂ + brine system increases with molalities, and those with divalent ion is twice larger than the monovalent ions at the same temperature, pressure, and molality conditions [41]. The measurement time also influenced the accuracy of the measured IFT data [39]. Larger errors were observed in the obtained data when waiting too long for the measurement, where the drop shape presented less integrity. The poor results (excessively small IFT) in Chun’s [36] work resulted from waiting for days for the system to reach phase equilibrium, as observed in previous literature [39].

To supplement the IFT database between CO₂ and water/brines, the research scope was extended to higher temperatures and pressures via modelling methods. The molecular dynamics (MD) method was developed to better understand the relationship between the surface strength and IFT for the CO₂ + H₂O binary system; however, a relatively high deviation between the predicted and experimental values was found [42,43,44]. In recent years, an artificial neural network (ANN) method was presented to estimate the IFTs for a CO₂ + water/brine binary system [45,46,47,48,49]. The ANN method exhibited a high level of accuracy with respect to different systems, while it was strongly dependent on the experimental data; thus, the estimation is restricted to the experimental conditions and cannot be extended to predictions beyond the experimental temperature and pressure ranges with high integrity. In previous work conducted by Chow [50], a theoretical model uses of square gradient theory (SGT) applied to the ‘statistical associating fluid-theory variable-range Mie’ (SAFT-VR-Mie) Equation of State (EoS) was developed for the IFTs of the CO₂ + H₂O/CO₂ binary system and the CO₂ + N₂/Ar + H₂O ternary system. The model exhibited a high level of integrity for the CO₂ + H₂O/brine binary system. The SAFT theory proposed by Chapman et al. [51,52] is one of the most successful models for modern algebraic descriptions of chain fluids and has been applied to estimate the thermodynamic and phase equilibrium properties of complex fluid mixtures in recent years [53,54,55,56,57]. However, the SAFT EoS is too complicated and requires too many factors and parameters. Most recently, Jerauld et al. [58] proposed a revised correlation method for the accurate estimation of the CO₂-brine IFT based on over 1600 data points from the literature and concluded that the Kashefi [59] method has better qualitative behavior and improves predictions with an AARE of approximately 5%. Neural network- and machine learning-based models have high computing speeds, strong adaptability, and a fault-tolerant ability; however, the normalization method lacks thermodynamic meaning. Moreover, the thermodynamic relationships between the main CO₂ trapping mechanisms in terms of solubility and IFT for the CO₂ + H₂O/brine binary system need to be elaborated in detail [27]. A detailed literature review is provided in the following

Table 1. Literature review for the CO₂-pure solution IFTs.

References	Solutions	Method	T/K	p/MPa
Jho et al. (1978) [33]	H2O	CT	284–318	0–6
Chun et al. (1995) [36]	H2O/methanol	CT	278–344	0.1–19
Hebach et al. (2002) [34]	H2O	PD	278–333	0.1–20
Akutsu et al. (2007) [37]	H2O	PD	298–318	7.5–17
Kavame et al. (2007) [38]	H2O	PD	278–335	0.1–20
Chiquet et al. (2007) [60]	H2O	PD	308–383	0–50
Georgiadis et al. (2010) [39]	H2O/Alkane	PD/SAFT-VR/DG	298–374	1–60
Bikkina et al. (2011) [61]	H2O	PD	298–333	1.4–21
Li et al. (2012) [40]	H2O	PD	335–448	0–50
Saraji et al. (2013) [62]	H2O	PD	308–333	3–10.5
Khosharay et al. (2014) [63]	Complex system	PD	284–313	0–6
Pereira et al. (2016) [64]	H2O	PD	298–469	0–69
Liu et al. (2016) [65]	CH4/CO2-H2O	PD	299–346	0–30
Rocha et al. (2001) [43]	H2O	MD	318–338	20–28
Kuznetsov et al. (2002) [44]	H2O	MD	285–298	9–30
Sutjiadi et al. (2008) [66]	H2O/alcohol	PD	313	0–27
Nielsen et al. (2012) [42]	H2O	MD	298–383	0–40
Chow et al. (2016) [67]	Complex system	PD/EE	298–448	0–40
Chow et al. (2016) [50]	Complex system	SAFT-VR+SGT	298–473	0–60
Zhang et al. (2016) [46]	H2O	ANN	278–448	0.1–61

PD, pendent drop; ANN = artificial neural network; SGT = square gradient theory; MD = molecular dynamics; CT = capillary tube; EE = empirical equation.

and

Table 2. Literature review for the CO₂-brine IFTs.

References	Solutions	Molalities	Method	T/K	p/MPa
Yang et al. (2005) [68]	Complex brine	4.27 g·L−1	PD	300–331	0–30
Chalbaud et al. (2006) [69]	NaCl (aq)	0–150 g·L−1	PD	300–373	4–26
Pierre et al. (2007) [60]	NaCl (aq)	0–20 g·L−1	PD	308–383	5–45
Bachu et al. (2009) [70]	NaCl (aq)	0–334 g·L−1	PD	293–398	2–27
Chalbaud et al. (2009) [13]	NaCl (aq)	0.085–2.75 mol·kg−1	PD	300.15–373.2	4–25
Li et al. (2012) [40]	Complex brine	0.98–4.95 mol·kg−1	PD	298–448	2–50
Liu et al. (2015) [71]	NaCl (aq)	0.1–1.0 mol·L−1	PD	300–313	3–9
Liu et al. (2017) [72]	NaCl (aq)	0–1.80 mol·kg−1	PD	300–353	3–12
Arif et al. (2016) [73]	NaCl (aq)	25 wt %	PD	308–343	0–20
Pereira et al. (2017) [74]	NaCl (aq)	0.98–1.98 mol·kg−1	PD/DGT	300–423	0–70
Yekeen et al. (2021) [75]	NaCl (aq)	0–7 wt %	PD	353–453	8–22
Aggelopoulos et al. (2010) [41]	Complex brine	5–300 g·L−1	PD	300–373	5–25
Aggelopoulos et al. (2011) [76]	CaCl2 (aq)	0.045–1.50 mol·kg−1	PD	300–373	5–25
Li et al. (2012) [40]	Complex brine	2–5 mol·kg−1	PD	343.15–423.15	2–50
Liu et al. (2017) [72]	Complex brine	0.045–1.8 mol·kg−1	PD	284–353	2–12
Arif et al. (2017) [77]	Complex brine	0–30 wt %	PD	323	0.1–20
Wang et al. (2018) [78]	Complex brine	21.46 g·L−1	PD	370.65	0.1–33.9
Mutailipu et al. (2019) [32]	Complex brine	0.98–4.95 mol·kg−1	PD	298–423	2–15
Li et al. (2014) [45]	H2O/brine	0–1 mol·kg−1	GAC	278–403	0–50
Zhang et al. (2016) [46]	Complex brine	0–5 mol·kg−1	ANN	278–448	0.1–61
Partovi et al. (2017) [47]	Complex brine	0–5 mol·kg−1	ANN	278–448	0.1–61
Amooie et al. (2019) [79]	Complex brine	0–4.087 mol·kg−1	ML	290–420	5–25

PD, pendent drop; ANN = artificial neural network; DGT = density gradient theory; GAC = graphical alternating conditional; ML = machine learning.

.

It was concluded that the interfacial characteristics of rock–fluid systems have a significant influence on the efficiency of CO₂ capillary trapping during saline aquifer sequestration [73]. Thus, in previous studies, the IFT between CO₂ and brine of different concentrations under reservoir pressure and temperature conditions has been investigated systematically. The results indicate that the IFT of the CO₂-brine binary system depends on the temperature, pressure, and molality [13,40,74]. According to previous research, the CO₂-brine IFT has a very limited pressure dependence tendency after the pseudo-plateau is reached. As for the salt concentrations, a linear relationship exists between the average increase in the CO₂-brine IFT and the brine molality [32,40]. According to previous research, this can be expressed using the following equations [80]:

$$\gamma =A_{\gamma }[m^{+}]+B_{\gamma }$$

(3)

$$m^{+}={\displaystyle \sum _{i=1}^{\mathrm{n}}{z}_i{m}_i}$$

(4)

$$A_{\gamma }=a_1+a_2{p}_{\mathrm{r}}+a_3{T}_{\mathrm{r}}$$

(5)

$$B_{\gamma }=b_1+b_2{p}_{\mathrm{r}}+b_3{T}_{\mathrm{r}}+b_4{p}_{\mathrm{r}}^2+b_5{p}_{\mathrm{r}}{T}_{\mathrm{r}}+b_6{T}_{\mathrm{r}}^2$$

(6)

The details of this method were discussed in a previous study [32]. The parameters for a_1–3 and b_1–6 were obtained by fitting the predicted results to the experimental data, and the coefficients used to model the CO₂-brine IFT are listed in

Table 3. Coefficients used to model the CO₂-brine IFTs.

Parameter	a1	a2	a3	b1	b2	b3	b4	b5	b6
Coefficient	3.672	1.376	−3.232	−165.524	−3.663	382.276	1.760	13.217	−160.054

.

Then, a modified empirical correlation based on experimental data was presented in Figure 4, where Δγ is the deviation between the experimental and estimated values for the CO₂-brine IFT. The modified empirical correlation uses only a few regression coefficients with a relatively low error for most of the experimental data with around 600 data points. The prediction results reached good agreement with the experiment, as shown in Figure 4. Thus, IFTs between CO₂-brine can be estimated via the modified empirical correlation based on the linear relations of Δγ and salinity, covering a wide range of temperatures (373–423 K), pressures (3–30 MPa), and ionic strengths (0–4.9 mol·kg⁻¹) for brines containing Na⁺, K⁺, Ca²⁺, and Mg²⁺.

However, to predict the IFTs when the isotherms reached a plateau, the empirical correlations based on the linear relation method yielded larger deviations (Δγ larger than ±4 mN·m⁻¹) than expected, as shown in Figure 4. Moreover, a large discontinuity in the slope for the IFTs against temperature at a constant pressure when passing from the liquid/vapor to the super-critical/liquid region was observed. Therefore, a model that is smooth and continuous in slope and successfully fits the IFT data corresponding to the CO₂ geological storage conditions, especially those covering the CO₂ phase-change region with thermodynamic significance, needs to be developed in future work. The following conclusions are relevant based on the results discussed above:

• Different methods exist to measure the IFT between gases and liquids. Among these, the pendent drop method, which is based on the axisymmetric drop shape analysis (ADSA) technique, is widely used because of its high accuracy and efficiency.
• An accurate measurement of the IFT in CO₂ reservoir brines at the evaluated high-temperature and pressure ranges corresponding to actual reservoir conditions relies on the integrity of the experimental apparatus and approach.
• Because of the temperature and pressure limitations of the desired experimental apparatus, a prediction model for the IFT between CO₂ and reservoir brines in accordance with the actual reservoir conditions covering a wide range of temperatures, pressures, ionic types, and strengths is essential.
• The results indicate that the IFT of the CO₂-brine binary system depends on the temperature, pressure, and molality. However, the pressure has a limited influence on the CO₂-brine IFT after the pseudo-plateau has been reached.

A simplified IFT prediction model with thermodynamic significance and high integrity for predicting IFTs when a pseudo-plateau is reached needs to be developed. And the model needs to cover a wide range of temperatures and pressures corresponding to the CO₂ geological storage conditions in future work.

2.2. Wettability of the CO₂-Water/Brine-Rock System

Multiphase flows in porous media are of great importance in many industrial processes in terms of geological CO₂ sequestration and utilization, including enhanced oil/gas/water recovery [81,82]. The effects of fluid properties and flow conditions have been widely studied in previous research, whereas less emphasis has been given to the wettability of the CO₂–water/brine–rock system, which is a key parameter influencing the storage capacity and security of the CO₂ geological storage scheme [83].

The seepage characteristics of reservoir fluids are a function of wettability because the fluid distribution in the pore space of reservoir rocks is governed by rock wettability [81]. Therefore, it is important to accurately measure the rock wettability under CO₂ geological storage conditions. The wettability of the CO₂-water/brine–rock system was estimated by measuring the contact angles of individual rock sample surfaces under reservoir conditions using conventional methods, including the ADSA [62,73]. However, conventional contact angle and tomographic imaging methods have serious limitations [83]. The measurement methods are diverse, and the methods that have been commonly used in recent years are summarized in Table 4, and Figure 5 shows how the static normal contact angle changes with pressure.

Table 4. Literature review on the investigation of the wettability of CO₂-brine-rock systems.

Method	Analysis Mode	Resources	Schematic
Sessile drop	Experimental, manual/automatic analysis	Bikkina et al. (2011) [84] Iglauer et al. (2014) [85] Liu et al. (2015) [71] Mutailipu et al. (2019) [32]
Captive bubble	Experimental, manual/automatic analysis	Chiquet et al. (2007) [86] Broseta et al. (2012) [87] Farokhpoor et al. (2013) [88] Shojai Kaveh et al. (2016) [89]
Sessile drop/Captive bubble on a titled plate	Experimental, manual/automatic analysis	Saraji et al. (2013) [62] Arif et al. (2016) [73] Arif et al. (2017) [77]
Capillary meniscus analysis	Experimental, image analysis	Li et al. (2014) [90] Hobeike et al. (2017) [91] Al-Zaidi et al. (2018) [92]
2D micromodel	Injected fluid imaged via microscope/analyzed using ImageJ	Chalbaud et al. (2009) [13] Kim et al. (2012) [93] Zhao et al. (2016) [82]
3D μ-CT imaging	CT-scan under dynamic condition/image analyzed using an algorithm	Andrew et al. (2013) [94] Lv et al. (2017) [95] Iglauer et al. (2018) [96]
Molecular dynamics (MD)simulation	High power computing/numerical solution of force field algorithm	Iglauer et al. (2012) [97] Liang et al. (2017) [98] Wang et al. (2022) [99]

Table 4. Literature review on the investigation of the wettability of CO₂-brine-rock systems.

Method	Analysis Mode	Resources	Schematic
Sessile drop	Experimental, manual/automatic analysis	Bikkina et al. (2011) [84] Iglauer et al. (2014) [85] Liu et al. (2015) [71] Mutailipu et al. (2019) [32]
Captive bubble	Experimental, manual/automatic analysis	Chiquet et al. (2007) [86] Broseta et al. (2012) [87] Farokhpoor et al. (2013) [88] Shojai Kaveh et al. (2016) [89]
Sessile drop/Captive bubble on a titled plate	Experimental, manual/automatic analysis	Saraji et al. (2013) [62] Arif et al. (2016) [73] Arif et al. (2017) [77]
Capillary meniscus analysis	Experimental, image analysis	Li et al. (2014) [90] Hobeike et al. (2017) [91] Al-Zaidi et al. (2018) [92]
2D micromodel	Injected fluid imaged via microscope/analyzed using ImageJ	Chalbaud et al. (2009) [13] Kim et al. (2012) [93] Zhao et al. (2016) [82]
3D μ-CT imaging	CT-scan under dynamic condition/image analyzed using an algorithm	Andrew et al. (2013) [94] Lv et al. (2017) [95] Iglauer et al. (2018) [96]
Molecular dynamics (MD)simulation	High power computing/numerical solution of force field algorithm	Iglauer et al. (2012) [97] Liang et al. (2017) [98] Wang et al. (2022) [99]

The wettability alteration of caprocks as a function of the temperature and pressure in the presence of supercritical CO₂ has been systematically investigated in previous studies. The wettability of the CO₂-brine-mica/quartz system barely changes and remains wet, regardless of the conditions observed [86]. The substrates remained hydrophilic in most of the experimental temperature, pressure, and salinity ranges, whereas a sudden increase in the CAs (5–13°) was observed when CO₂ was converted from the subcritical to the supercritical region for different rock samples in our previous work [32]. Alterations in the wettability of the caprock in the presence of supercritical CO₂ have also been reported in other studies [62,71,86,87]. Based on previous research, wettability alternations of the caprock should be taken into consideration for accurate estimation of the CO₂ storage capacity of a potential sequestration site. The wettability of rock/CO₂/brine and rock/oil/CO₂-enriched brine systems has been analyzed critically, and a future outlook was proposed in previous literature [100]. Until very recently, only limited investigations have been conducted regarding the CAs between reservoir rocks and water or brine, although wettability has a profound influence on fluid–fluid interactions in the presence of a solid surface under CO₂ storage conditions [82].

Traditional contact angle measurements overlook the 3D pore geometry, surface roughness, and chemical heterogeneity, making them more prone to error. Hence, a micro-computed tomography (CT) scanner method was used to measure the local CAs of a CO₂-brine–glass beads system at the pore scale [95,101]. However, there are obvious limitations and disadvantages associated with using the CT method. It is expensive, time-consuming, and the medical CT resolution is too low, resulting in only averaged saturations being measured, and this curvature measurement may be biased owing to its voxelized nature [82]. To overcome some of the abovementioned limitations and to deepen the fundamental understanding of the flow processes of supercritical CO₂ via brine-saturated porous media, Nuclear Magnetic Resonance (NMR) with molecular resolution has recently been used to measure CO₂-rock wettability in recent work [83]. Despite recent advances in our ability to accurately measure the wettability under reservoir conditions and engineer wettability in the subsurface, the complex physics of wetting continues to challenge our microscopic and macroscopic descriptions [82], owing to measurement disparities, variation in the sample type, roughness, contamination, and cleaning procedures [73,77,84,87]. Stefan et al. [85] proposed that the broad spread of CA data in the literature was caused by surface contamination. Thus, establishing a practical cleaning procedure for the surfaces of sample substrates is of vital importance for the accurate measurement of rock wettability [32,62,84,85,88].

Overall, the data on CO₂-brine IFTs and rock CAs from laboratory studies are inadequate and contain discrepancies. Most of the experimental studies discussed above used the pendent drop method based on the ADSA method. Thus, experimental errors, inconsistent measurement procedures, and errors in the subsequent analytical techniques could explain these disparities. Therefore, to obtain adequate data with high integrity for CO₂-brine IFTs and rock CAs corresponding to reservoir conditions, further laboratory and modelling research in terms of molecular dynamics and simplified equations of state is required, and the thermodynamic relations between interphase properties need to be elaborated in particular.

Based on the results discussed above, the following conclusions are relevant:

• Conventional methods in terms of the contact angle and tomographic imaging measurements have serious limitations.
• The disparity in the CA results in the reported data can be attributed to the differences in the experimental procedures, substrates, roughness, and cleaning procedures used for the substrate surfaces.
• The effects of the ionic strength and ionic type on the CAs of different substrates need to be further investigated to verify the effects of salinity on CAs.
• Alternation in the wettability of the caprock in the presence of supercritical CO₂ has been observed, and a molecular dynamics method should be implemented to simulate this phenomenon at the molecular level.

2.3. The Mutual Solubility between CO₂ and Water/Brine

Solubility capture is one of the main long-term carbon capture mechanisms, accounting for approximately 20% of the total capture amount at the initial stage of injection. It can capture up to 90% of the CO₂ injection amount [11,102]. Thus, the mutual solubility between CO₂ and reservoir fluids is one of the key thermodynamic properties that determines the flexibility of the CO₂ saline aquifer sequestration scheme. Over the years, solubility data for CO₂ + H₂O systems have been reviewed and analyzed [103]. Laboratory solubility data for a CO₂ + H₂O system covering a wide range of temperatures and pressures are available in the previously published literature [104,105,106]. Recently, empirical equations and estimation models for water-saturated CO₂ were developed based on experimental data available in the literature [107,108]. Then, modeling for the estimation of mutual solubility in the CO₂ + H₂O binary system was conducted [109,110,111,112,113].In addition, estimation methods of the p-T-x properties of the CO₂ + H₂O binary system were proposed by others [114,115].

Considering the effects of salinity on solubility, solubility estimation models based on the hybrid G^E equation of state (EoS) [116], Anderko–Pitzer EOS [117], Pitzer EoS [118,119], PC-SAFT EoS [120], mixed solvent electrolyte (MSE) model, and Soave–Redlich–Kwong (SRK) EoS [121] were developed by previous researchers. Mutual solubility research on the CO₂ + brine/complex brine binary system containing ions such as Na⁺, K⁺, Mg²⁺, and Ca²⁺ as well as Cl¹⁻, SO²⁻₄ for a temperature (T) of 273 K < T < 469 K, a molality (m) of 0 mol·kg⁻¹ < m < 5.0 mol·kg⁻¹, and pressure (p) of up to 104 MPa has been carried out over the years and is presented in Table 5 and Figure 6. Hou et al. [17] performed mutual solubility measurements on the CO₂ + H₂O binary system, and considering the effects of the dissolved salts, vapor–liquid equilibria were measured for the CO₂ + H₂O + NaCl/KCl system using the same analytical apparatus [17]. Duan’s model [122] was found to be broadly satisfactory for measuring the solubility of CO₂ in NaCl (aq) but performed poorly for assessing the solubility of CO₂ in KCl (aq). Ratnakar et al. [123] developed a method that uses an ionic strength-based mixing rule and extended the well-known Setschenow relation to capture the impact of a mixture of salts containing mono- or multivalent ions to predict the gas solubility in brine solutions with single or mixed salts. Then, Zhao et al. [124] used the Setschenow coefficient of ions to calculate the solubility of CO₂ in mixed salt solutions. Most recently, Sun et al. [106] proposed a simple model for the prediction of mutual solubility in CO₂ + H₂O and CO₂ + brine systems.

In the meantime, a well-developed model was proposed by Spycher and Pruess [131] to calculate the composition of the compressed CO₂ and liquid H₂O phases at equilibrium based on a noniterative procedure to evaluate the feasibility of CO₂ geologic sequestration for temperatures between 285 K and 383 K and pressures of up to 60 MPa (i.e., 285 K ≤ T ≤ 383 K; p ≤ 60 MPa) by equating chemical potentials and using the Redlich–Kwong EoS. Because the equilibrium constant K (i.e., directly related to the standard Gibbs free energy of the reaction as ∆G⁰ = −RT ln K) is a more fundamental thermodynamic property than Henry’s law constant H, and because the formulation can be more easily extended to a non-ideal aqueous phase (i.e., resulting from the addition of salts) than formulations involving Henry’s law constant, K instead of H was used in their solubility model. In 2005, their [132] solubility model was extended to include the effect of chloride salts in the aqueous phase by combining the activity coefficient formulations of Duan and Sun [119] and Rumpf et al. [133] with their solubility correlations with an accuracy within the experimental uncertainty for solutions of up to 6 mol·kg⁻¹ NaCl and 4 mol·kg⁻¹ CaCl₂. Then, Spycher and Pruess extended the temperature range of their solubility model up to 573 K (285 K ≤ T ≤ 573 K; p ≤ 60 MPa) in 2010 [134]. The Redlich and Wong equation [135] was used in the Spycher and Pruess [134] solubility model. Nevertheless, the basic Redlich–Kwong [136] equation can only be applied to a few rather simple fluids. This is because the parameters of the equation are based entirely on the two critical constants, namely the critical temperature (T_c) and the critical pressure (p_c), and do not involve the acentric factor. The equation of Peng and Robinson [137] is structurally similar to the Redlich and Kwong [135] equation, but it contains an acentric factor for its application to a pure fluid. The content related to the equation of state can be found in the literature [138,139].

Therefore, to better estimate the phase equilibrium properties of mixtures composed of interacting molecules, the interaction parameter K_ij based on the EoS and mixing rules can be correlated with the modified Peng–Robinson [137] EoS and the mixing rules of Panagiotopoulos and Reid [140]. Furthermore, in the Spycher and Pruess [134] model, different coefficients and parameters were developed for temperatures of above 373 K and below 373 K, such as the Margules expression and the equilibrium constant K. The activity coefficients used for the effects of different chloride salts in the brine system were the same, although it has been proven that the ion type has a significant influence on the mutual solubility of the CO₂ + brine binary system. Thus, to develop a more reliable and simplified solubility model covering the P–T range of interest for applications to geologic CO₂ sequestration, the original solubility model of Spycher and Pruess needs to be further modified by replacing the Redlich–Kwong [135] EoS with the Peng–Robinson [137], and the calculation methods of the mole fraction, equilibrium constant, and fugacity coefficient also needed to be optimized. According to the literature review presented in Table 5, it can be concluded that the main methods for modeling the solubility of liquid or gas phases include an EoS that is usually based on φ-φ [108,112] or γ-φ [17,129,134] equations.

Table 5. Literature data for CO₂-H₂O/brine mutual solubility system.

References	Aqueous Phase	m/ (mol·kg−1)	Temperature T/K	Pressure p/MPa
Bamberger et al. (2000) [112]	H2O	0	323.2–353.1	4–14.2
Servio et al. (2001) [141]	H2O	0	278.05–283.15	2–3.7
Anderson et al. (2002) [125]	H2O	0	274.15–288.15	0.07–2.2
Kiepe et al. (2002) [129]	H2O	0	313.2–393.17	0.09–10
Valtz et al. (2004) [126]	H2O	0	278.2–318.2	0–8
Chapoy et al. (2004) [108]	H2O	0	274.14–351.31	0.19–7.31
Iwai et al. (2004) [136]	H2O	0	313.2	15
Zhang et al. (2005) [142]	H2O	0	304–313	1–4.6
Dalmolin et al. (2006) [127]	H2O	0	288–323	0.1–0.5
Qin et al. (2008) [143]	H2O	0	323.6–324.1	0.03–0.05
Pereira et al. (2009) [102]	H2O	0	298.2–323.2	0.05–0.6
Martin et al. (2009) [144]	H2O	0	353–393	10–30
Han et al. (2009) [145]	H2O	0	313.2–343.2	4.3–17.3
Dell’Era et al. (2010) [146]	H2O	0	298.48–298.63	0.2–0.8
Ferrentino et al. (2010) [147]	H2O	0	308.15–323.15	5.5–15
Ruffine et al. (2010) [148]	H2O	0	333	4.8–11.5
Liu et al. (2011) [149]	H2O	0	308.15–323.15	2–16
Tabasinejad et al. (2011) [150]	H2O	0	422.98–478.35	3.9–103.4
Schüler et al. (2012) [151]	H2O	0	308–333	0.101
Liu et al. (2012) [152]	H2O	0	308.15–318.15	8–16
Hou et al. (2013) [17]	H2O	0	298.15–448.15	1–18
Serpa et al. (2013) [153]	H2O	0	298–323	0.1–0.4
Wang et al. (2013) [154]	H2O	0	313.15–373.15	9–17.6
Tong et al. (2013) [155]	H2O	0	374.1–323.2	7.2–27.3
Jiang et al. (2014) [156]	H2O	0	333.15–373.15	8–18
Bastami et al. (2014) [157]	H2O	0	328.15–375.15	6.8–20.7
Al Ghafri et al. (2014) [158]	H2O	0	323.15	2–18.7
Krotov et al. (2014) [159]	H2O	0	353.3	0.6–9.7
Meyer et al. (2015) [160]	H2O	0	283.15–333.19	0–5
Muromachi et al. (2015) [161]	H2O	0	286.15–298.15	0.2–4.0
Tang et al. (2015) [128]	H2O	0	308.15–468.15	7.9–41
Foltran et al. (2015) [162]	H2O	0	313.15	8–12.5
Rumpf et al. (1994) [133]	NaCl	4–6	313–433	0.5–9.6
Kiepe et al. (2002) [98,129]	NaCl	0.5–4.3	313–352	0.1–10
Bando et al. (2003) [163]	H2O/NaCl	0–0.55	303–333	10–20
Kiepe et al. (2002) [98,129]	KCl	0.5–4.0	313–353	0.1–10.5
Hou et al. (2013) [17]	NaCl/KCl	2.5–4.0	323–423	2.6–18.2
Tong et al. (2013) [155]	Complex brine	1–5	310–425	1.2–34.9
Hoballah et al. (2017) [130]	NaHCO3	0.8	348–398	5–50

Table 5. Literature data for CO₂-H₂O/brine mutual solubility system.

References	Aqueous Phase	m/ (mol·kg−1)	Temperature T/K	Pressure p/MPa
Bamberger et al. (2000) [112]	H2O	0	323.2–353.1	4–14.2
Servio et al. (2001) [141]	H2O	0	278.05–283.15	2–3.7
Anderson et al. (2002) [125]	H2O	0	274.15–288.15	0.07–2.2
Kiepe et al. (2002) [129]	H2O	0	313.2–393.17	0.09–10
Valtz et al. (2004) [126]	H2O	0	278.2–318.2	0–8
Chapoy et al. (2004) [108]	H2O	0	274.14–351.31	0.19–7.31
Iwai et al. (2004) [136]	H2O	0	313.2	15
Zhang et al. (2005) [142]	H2O	0	304–313	1–4.6
Dalmolin et al. (2006) [127]	H2O	0	288–323	0.1–0.5
Qin et al. (2008) [143]	H2O	0	323.6–324.1	0.03–0.05
Pereira et al. (2009) [102]	H2O	0	298.2–323.2	0.05–0.6
Martin et al. (2009) [144]	H2O	0	353–393	10–30
Han et al. (2009) [145]	H2O	0	313.2–343.2	4.3–17.3
Dell’Era et al. (2010) [146]	H2O	0	298.48–298.63	0.2–0.8
Ferrentino et al. (2010) [147]	H2O	0	308.15–323.15	5.5–15
Ruffine et al. (2010) [148]	H2O	0	333	4.8–11.5
Liu et al. (2011) [149]	H2O	0	308.15–323.15	2–16
Tabasinejad et al. (2011) [150]	H2O	0	422.98–478.35	3.9–103.4
Schüler et al. (2012) [151]	H2O	0	308–333	0.101
Liu et al. (2012) [152]	H2O	0	308.15–318.15	8–16
Hou et al. (2013) [17]	H2O	0	298.15–448.15	1–18
Serpa et al. (2013) [153]	H2O	0	298–323	0.1–0.4
Wang et al. (2013) [154]	H2O	0	313.15–373.15	9–17.6
Tong et al. (2013) [155]	H2O	0	374.1–323.2	7.2–27.3
Jiang et al. (2014) [156]	H2O	0	333.15–373.15	8–18
Bastami et al. (2014) [157]	H2O	0	328.15–375.15	6.8–20.7
Al Ghafri et al. (2014) [158]	H2O	0	323.15	2–18.7
Krotov et al. (2014) [159]	H2O	0	353.3	0.6–9.7
Meyer et al. (2015) [160]	H2O	0	283.15–333.19	0–5
Muromachi et al. (2015) [161]	H2O	0	286.15–298.15	0.2–4.0
Tang et al. (2015) [128]	H2O	0	308.15–468.15	7.9–41
Foltran et al. (2015) [162]	H2O	0	313.15	8–12.5
Rumpf et al. (1994) [133]	NaCl	4–6	313–433	0.5–9.6
Kiepe et al. (2002) [98,129]	NaCl	0.5–4.3	313–352	0.1–10
Bando et al. (2003) [163]	H2O/NaCl	0–0.55	303–333	10–20
Kiepe et al. (2002) [98,129]	KCl	0.5–4.0	313–353	0.1–10.5
Hou et al. (2013) [17]	NaCl/KCl	2.5–4.0	323–423	2.6–18.2
Tong et al. (2013) [155]	Complex brine	1–5	310–425	1.2–34.9
Hoballah et al. (2017) [130]	NaHCO3	0.8	348–398	5–50

Based on the results discussed above, the following conclusions are relevant:

• Empirical equations and estimation models for CO₂ + H₂O systems have been widely implemented to calculate the solubility of water-saturated CO₂; however, they have limitations in estimating the solubility of H₂O in compressed CO₂.
• There are still too many parameters in the Duan et al. model, and most importantly, it is not intended to compute the solubility of H₂O in a compressed CO₂ gas phase and does not distinguish between ions of the same charge.
• Although the Krichevsky–Kasarnovsky (KK) approach provides a reasonably good representation of the data, it either fails to fit the data or yields an unphysically negative slope at higher temperatures.
• The well-developed Spycher and Pruess model is widely used to develop a more reliable and simplified model for determining the mutual solubility between CO₂ and brines containing different ionic species and of different strengths.
• In the Spycher and Pruess model, different coefficients and parameters have been developed for temperatures above 373 K and below 373 K, such as the Margules expression and equilibrium constant K, which have no thermodynamic meaning. Furthermore, the activity coefficients used to determine the effects of different chloride salts in the brine system were the same. The interaction parameter K_ij, based on the EoS and mixing rules, can still be correlated with the modified Peng–Robinson and the mixing rules of Panagiotopoulos and Reid.

The main methods used for modelling the solubility of liquid or gas phases include an EoS that is usually based on φ-φ or γ-φ equations. Previous solubility models were either limited to a certain range of temperatures or pressures, or were too complex. Most importantly, until very recently, limited research has been available for the simultaneous calculation of the mutual solubility of the CO₂-rich phase and the H₂O-rich phase with high integrity for the application of the CO₂ saline aquifer storage scheme. Furthermore, the interaction mechanism coupling with multiple factors of the gas–liquid–solid interface properties and the dissolution and acidification process needs to be explored in future work.

2.4. pH of CO₂-Saturated Brine

Modeling and experimental research have been carried out over the years to estimate the pH of the CO₂-saturated brine system. In general, electrometric or optical methods are used to measure the pH of CO₂-saturated brine or water, where the former is based on the electrode system [164], and the latter is based on UV–vis spectrophotometry [165,166,167]. Careful calibration of the measurement system is essential for both measurement approaches.

Previous studies on the pH of CO₂-saturated aqueous solutions have mainly focused on water [168,169,170,171,172], seawater [173], and NaCl brines [174,175,176,177,178,179,180,181,182]. However, reservoir brines contain Na⁺, Cl⁻, K⁺, Ca²⁺, Mg²⁺, SO₄²⁻, HCO₃⁻, and CO₃²⁻, which are expected to significantly influence the fluid properties [76,130,183]. Measurements of brines containing salts other than, or in addition to, NaCl were performed by Stefansson et al. [184], who studied complex systems containing Na₂CO₃, NaHCO₃, NaOH, HCl, and NaCl, and by Li et al. [182], who studied both NaCl and NaHCO₃ brines. An updated literature review on the pH measurement of CO₂-saturated aqueous solutions is presented in our previous work [23]. According to a review of the previous literature [23,168,171,172,173,174,175,177,180,182,185], it can be concluded that the electrometric method is often used to measure the pH of CO₂-saturated aqueous solutions covering a wide range of temperatures and pressures (pressures up to 35 MPa), due to its reliability and stability under reservoir conditions. While Toews et al. [169] and Parton et al. [170] measured the pH of the CO₂-H₂O system for the temperatures and pressures ranging from 298 to 343 K and from 7 to 20 MPa, respectively, via optical methods. Then, Shao et al. [179] used an optical method to measure the pH of a CO₂-NaCl system for temperatures and pressures ranging from 298 to 366 K and from 0 to 20 MPa, respectively, while the pressure was 0.1 MPa for the same measurement system in Millero et al.’s work [178]. Also, in Stefansson et al.’s [184] work, the optical method was used to measure the pH of a CO₂-saturated complex brine system at a low pressure range (pressures ranging from 1 to 2 MPa).

Typically, geochemical simulators such as PHREEQC [186] and EQ3NR [187] have been used to model the pH of gas-saturated aqueous solutions. These models implement the Pitzer model to calculate the activity coefficients of solutes in the aqueous phase, the correlations of Henry’s constant and an equation of state for gaseous components, and the correlations of standard equilibrium constants for aqueous-phase chemical reactions [188,189,190]. In a previous study on NaCl and NaHCO_3, as well as KCl brines, the Pitzer model accounted well for the measured pH over a wide range of temperatures, pressures, and ionic strengths [23,182]. This type of model offers a convenient route for predicting the pH of complex brines with dissolved CO₂; however, additional validation is required for brines other than NaCl (aq). Prior to estimating the pH, the solubility of gas-saturated aqueous solutions is calculated first in the PHREEQC; thus, the two important thermodynamic properties (pH and solubility) of gas-saturated aqueous solutions can be evaluated via geochemical simulators such as PHREEQC. However, there are limited studies available in the literature to investigate the thermodynamic relations between the pH and solubility of gas-saturated aqueous solutions.

Given the lack of experimental data, it is important to measure the pH of CO₂-saturated brines other than NaCl (aq) over wide ranges of temperatures, pressures, and molalities. The range of investigated conditions should include those relevant to CO₂ storage in deep-saline aquifers. In previous work, the pH of CO₂-saturated aqueous KCl solutions was measured over wide ranges of temperature and pressure and at salt molalities of 2 mol·kg⁻¹ and 4 mol·kg⁻¹ [23]. It was observed that the influence of KCl on the pH under conditions of a constant temperature and CO₂ partial pressure was different from that of NaCl, with the pH in the former being slightly higher. However, this difference in behavior is probably only significant in brines with high f K⁺ molalities [23]. The Pitzer model combined with the MacInnes convention [191], as implemented in PHREEQC version 3.5.0, provides a generally good prediction of the pH of the CO₂-saturated aqueous KCl solutions [23]. Hence, previous work served to validate a geochemical simulator based on the Pitzer model for the prediction of pH in systems involving CO₂ dissolved in water, NaCl (aq), NaHCO₃ (aq), and KCl (aq) [23,172,182]. The influence of divalent cations on the pH of CO₂-saturated aqueous solutions might be different from that of monovalent cations; thus, the model requires further validation for systems involving CO₂ dissolved in brine containing other types of ionic ions, such as Ca²⁺ and Mg²⁺. To the best of our knowledge, there are no publications on the pH of a saturated CO₂-CaCl₂ solution and a saturated CO₂-MgCl₂ solution under high-temperature and high-pressure conditions. Therefore, the aim of future research on the pH of saturated CO₂-brine is to validate the integrity of geochemical simulators based on the Pitzer model.

Based on the results discussed above, the following conclusions are relevant:

• Very limited research is available for the pH of gas-saturated aqueous solutions under CO₂ saline aquifer storage conditions, although it is considered to be one of the most important parameters of reservoir fluids.
• In general, electrometric or optical methods are used to measure the pH of CO₂-saturated brine or water. On the basis of the literature review, it can be concluded that the electrometric method is often used to measure the pH of CO₂-saturated aqueous solutions covering a wide range of temperatures and pressures (pressures of up to 35 MPa) due to its reliability and stability under reservoir conditions.
• Geochemical simulators based on the Pitzer model offer a convenient route for predicting the pH of complex brines with dissolved CO₂. However, additional validation is required for brines other than NaCl (aq).
• The pH and solubility of gas-saturated aqueous solutions can be evaluated via geochemical simulators such as PHREEQC. However, there are limited studies available in the literature to investigate the thermodynamic relations between the pH and solubility of gas-saturated aqueous. The mechanisms of dissolution and acidizing during the CO₂ saline aquifer storage process need to be studied systematically, and thermodynamics relations between the pH and solubility of gas-saturated aqueous solutions need to be clarified.
• Given the lack of experimental data, it is important to measure the pH of CO₂-saturated brines other than NaCl (aq) over wide ranges of temperatures, pressures, and molality. The range of investigated conditions should include those relevant to CO₂ storage in deep-saline aquifers. The results of these experiments were compared with calculations based on the Pitzer model to validate the geochemical simulators based on the Pitzer model.

3. Influence of Interface Properties on the Pore-Scale Multi-Phase Flow

3.1. Core Flooding Experiment and Modelling

The CO₂-water displacement process at the pore scale can be investigated via laboratory core flooding experiments under reservoir conditions using X-ray CT scanning technology [148,192,193,194]. The effects of different samples and flow rates on the CO₂/brine displacement efficiency have been studied [195,196]. Laboratory core flooding experimental systems and apparatus for CO₂ geological sequestration developed over the past decade have been reviewed carefully in previously published literature [197]. Meanwhile, the laboratory experiments have been replicated by implementing numerical models that can accurately interpolate the CO₂/brine relative permeability by matching the experimental data. Subsequently, a 2D semi-analytical model was developed to predict the brine displacement efficiency over a wide range of capillary numbers [198]. Magnetic resonance imaging (MRI) has been widely used to study the migration of injected CO₂ in porous media [199]. A multiphase flow experimental system based on MRI was used to study the multiphase flow and displacement process of CO₂ during the storage of saline water layers. The local porosity of a single layer can be obtained using the MRI method so that the internal pore structure of porous media can be comprehensively analyzed, and the visual detection of multiphase seepage in porous media can be realized. In addition, MRI phase migration velocimetry can accurately and quantitatively analyze the flow velocity and mass transfer of multiphase seepage in porous media [200].

The flow of carbon dioxide is affected by the anisotropy of the porosity and permeability of the porous media; therefore, it is necessary to reasonably analyze the pore structure of the sealed formation during the process of carbon dioxide storage. Recently, a pore-scale fluid flow method based on pore network models from sample X-ray computed tomography (micro-CT) was developed to analyze the CO₂/brine displacement process [71,201]. Pore network models have been widely used to analyze the core structure and multiphase flow simulation of oil/water to enhance oil recovery [202,203,204,205]. Currently, pore network multi-phase flow simulation models are being implemented in research on the CO₂/brine seepage characteristics of reservoir rock samples to provide a fundamental understanding of the CO₂/brine displacement process in porous media under CO₂ reservoir storage conditions. Other models applied in the research of multiphase flow problems include the Lattice Boltzmann model (LBM) of multiphase flow system [206], the Shan–Chen (S-C) model [207], the four-parameter random growth method (QSGS) [208], the generalized Lattice Boltzmann model (GLBM) [209], and the Homogenized Lattice Boltzmann model (HLBM) [210]. However, limited pore-scale numerical simulations are available in the literature.

On the other hand, the thermodynamic properties, such as the interfacial tension, wettability, and density of the gas–liquid–solid system, are important parameters of pore network multi-phase flow simulation models. Different equations have been developed to consider the heterogeneity of wettability in pore network multi-phase flow simulation models [204,211]. The effects of fluid properties and flow conditions have been widely studied in previous research, whereas less emphasis has been given to the wettability of the CO₂–water/brine–rock system, which is a key parameter that influences the storage capacity and security of the CO₂ geological storage scheme. The influence of the thermodynamic properties on muti-phase flow in porous media is discussed in this section in terms of the wettability and interface properties via pore network multi-phase flow simulation models.

3.2. Effects of Wettability

In general, the wettability of a given rock/fluid system controls the pore-scale arrangement of fluids within the rock pore space, which, in turn, affects the micro- and reservoir-scale multiphase flow properties [212]. Previous studies in this area have focused on water-driven oil repellency. Morrow et al. [213] found, by using different methods, that when the rock wettability changes from hydrophilic to neutral or even lipophilic, the water percolation resistance increases significantly, the oil percolation resistance decreases, and the oil repellency efficiency tends to increase, with the highest repellency efficiency occurring for weakly water-wet rock samples. Furthermore, it is proposed that for low-permeability reservoirs, the shift in wettability from hydrophilic to neutral is an important mechanism for improving the recovery rate.

The relative permeability of bead-filled porous media with altered wetting conditions was also investigated in recent years by Landry et al. [214] based on the Lattice Boltzmann method. They found that changes in the surface wettability did not significantly affect the relative permeability of the wetted phase but resulted in a significant decrease in the relative permeability of the non-wetted phase [214]. Hwang et al. [215] investigated the effect of the fracture wettability on the capillary pressure and relative permeability profiles. They found that a larger proportion of non-water-wetted solid surfaces led to a higher relative water permeability for a given water saturation level [215]. Guo et al. [216] focused on the effect of wettability inhomogeneity on percolation and found that, under both homogeneous and inhomogeneous wetting conditions, the inhalation process increased the relative permeability of water and decreased the relative permeability of scCO₂ compared to the repulsion process.

Thus, the wetting properties of the reservoir must be considered when studying its seepage characteristics to ensure the flexibility of the CO₂ deep saline aquifer storage scheme [13]. Investigations of the seepage characteristics of the CO₂-brine two-phase flow under CO₂ deep saline aquifer storage conditions have been carried out for years via both laboratory experiments and numerical simulations in terms of relative permeability and capillary pressure.

Recently, researchers measured the contact angles between brine, dense CO₂, and minerals representative of shales, such as mica and quartz [86,217]. They reported that the water-wettability of these minerals barely changed. Perrin and Benson [195] studied the CO₂-brine displacement process in rock samples at various flow rates. They observed that the CO₂-brine displacement efficiency depended on the heterogeneity of the cores and the flow rates. However, only limited laboratory experiments have been conducted regarding the wettability of reservoir rocks under CO₂ storage conditions owing to the complexity of the experimental process. With the fast development of pore-scale numerical simulations, the effect of the wettability on fluid flow in porous media can be investigated via pore network models combined with X-ray computed tomography (micro-CT) [202,203,204,205,218,219]. Chen and Liu measured the percolation properties of a CO₂-brine two-phase flow in porous media via pore network modeling [71,201]. It is believed that a two-phase flow simulation based on pore network modeling is a reliable, efficient, and economical method for investigating the reservoir CO₂ storage capacity.

In a previous study, the pore network flow simulation model was used to study the influences of the thermodynamic properties and pore structure on the CO₂-brine seepage characteristics at 323 K and 12.4 MPa with a 6 g/L NaCl solution, providing a fundamental understanding of the CO₂-brine displacement process under CO₂ reservoir storage conditions [220]. The results indicate that changing the rock-wetting behavior and interactions between the fluid and gas–rock has a significant effect on the gas-flooding process compared to water flooding as the temperature and pressure increase. To better understand the influence of wettability on the fluid flow in porous media, the CO₂-brine two-phase flow behavior in homogenous Berea sandstone was investigated using the pore network flow simulation method [221]. A single wetting system was used for the flow simulation, and the sample was considered to be homogeneously wetted Berea sandstone. The results indicate that the fluid flow of gas and water is much better in a strongly water-wet system than in a completely CO₂-wet system, and the entire system was in a stable displacement state.

However, most natural reservoir cores are heterogeneous wetting porous media, and their pore spaces are partly water-wet and partly hydrophilic. Thus, it is of great significance to study the effects of wettability heterogeneity on the gas–water flow under reservoir conditions. In our previous work, the influence of wettability on the CO₂-brine two-phase flow behavior in a Berea sandstone sample was studied using a pore network numerical flow simulation method [222]. The results indicate that the heterogeneity of wettability has a significant influence on the CO₂-brine two-phase flow behaviour during water flooding. When the non-wetting proportion of the system increases, the system becomes more CO₂ wet, and the liquidity of the aqueous phase is enhanced; in the end, more CO₂ is displaced by the aqueous phase. Therefore, saline with strong water-wetting properties might be the best site for CO₂ storage because of the better CO₂ flow characteristics of strong water-wetting porous media [222].

It seems that the numerical simulation based on pore network extraction is an efficient and economical method for studying the gas–water seepage characteristics under CO₂ reservoir storage conditions, especially for investigating the influence of wettability on the fluid flow in porous media. However, the integrity of this method needs to be validated by combining it with the results of laboratory core flooding experiments.

3.3. Effects of Interphase Properties

In addition to the wettability, the influence of the thermodynamic properties, such as the interfacial tension, fluid flow rate, and viscosity, on multiphase flow in porous media is also a popular research topic. Gu et al. [223] experimentally determined that residual brine saturation is a function of the viscous force at a certain flow rate. As the flow rate increased, the residual brine saturation decreased, and the endpoint CO₂ relative permeability increased. Lyi et al. [224] used the ANSYS-CFD technique to establish a numerical model for hot water injection in oil-bearing sandstone cores. It was found that the permeability (relative) of the oil flow curve was strongly influenced by the flow rate, whereas water had little or no effect on the relative permeability. A fully implicit isothermal compositional pore network model for gas and condensate flows was proposed by Reis et al. [225], which predicts the positive effect of flow on the relative permeability curve, and a higher relative permeability is expected at lower levels of IFT. Using an experimental method, Henderson et al. [226] found that the relative permeability of both phases increased with an increasing flow rate under both steady- and unsteady-state conditions. Increasing the value of the interphase IFT decreased the relative permeability of the gas phase more than that of the condensate phase.

The experimental data were analyzed by Balogun et al. [227] using the nonstationary water drive method and numerical calculations by history fitting, and it was found that the relative permeability curve increased for high-viscosity oil and decreased for low-viscosity oil as the flow rate increased. Esmaeili et al. [228] used the Johnson, Bossler, and Neumann (JBN) method and a history matching method to calculate the relative permeability of isothermal oil drive test results. The results showed that the relative oil-water permeability was almost insensitive to the viscosity ratio. Bennion et al. [229] found that the better the viscosity ratio between water and CO₂ (the more favorable), the better the replacement effect. Later, it was found that the IFT may play a more important role in the variation in the relative permeability and endpoint saturation characteristics than the viscosity ratio [230].

Babadagli [231] suggested that a decrease in the IFT during the imbcycle process leads to a decrease in the capillary pressure, which in turn facilitates percolation. Jamialahmadi et al. [232] later confirmed this conclusion using the free-energy Lattice Boltzmann method. The effect of gas–tension was found after extensive experimental studies by Bennion et al. [230]: the relative permeability of a given system increases with a decreasing IFT. Later, Gu et al. [223] reconfirmed this conclusion; the curves of both phases shifted toward an increase in brine saturation as the IFT increased. Using an ideal pore-scale numerical model, Iyi et al. [233] analyzed the VOF and found that, under all wettability conditions, the repulsion efficiency could be improved by sufficiently reducing the IFT between fluid phases. By reducing the IFT and changing the wettability to a water-wet state, recovery could be improved.

Based on the results discussed above, the following conclusions are relevant:

The effects of thermodynamic properties on pore-scale multi-phase flow can be investigated via pore flooding experiments based on X-ray computed tomography (micro-CT) and pore-scale numerical simulations. The latter is widely used because of its high efficiency and economy. However, the pore-scale simulation model can be modified in combination with laboratory research to explore the influence of interface properties on multiphase flow in porous media.

Limited pore-scale numerical simulations are available in the literature. On the other hand, thermodynamic properties, such as the IFT and wettability of the gas–liquid–solid system, are important parameters of pore network multi-phase flow simulation models. The effects of fluid properties and flow conditions have been widely studied in previous research, whereas less emphasis has been given to the wettability of the CO₂–water/brine–rock system, which is a key parameter influencing the storage capacity and security of the CO₂ geological storage scheme.

The multi-dimensional and multi-scale gas–liquid flow visualization experiment, as well as pore-scale seepage numerical modeling coupling with multi-parameters, need to be carried out. The variation and the interaction mechanism of pore seepage features caused by the physical and chemical interactions of the multi-phase interface properties need to be clarified in future work to provide a theoretical base and technical guidance on the achievement of the CO₂ geological storage and utilization project with high integrity and efficiency.

4. Conclusions

A detailed review devoted to the consideration and analysis of literature sources for the last 15 years on thermodynamic properties in the gas–liquid–solid system in the process of geological storage of carbon dioxide is presented in this study. After reviewing the previous literature, the following conclusion can be drawn:

(1)

A simplified IFT prediction model with thermodynamic significance and high integrity for predicting IFTs when a pseudo-plateau is reached needs to be developed in future work to cover a wide range of temperatures and pressures corresponding to CO₂ geological storage conditions.

(2)

Further laboratory and modelling research on the wettability alternations of caprock in terms of the molecular dynamics and simplified equations of state is required, and the thermodynamic relations between interphase properties need to be elaborated in particular.

(3)

The main methods for modelling the mutual solubility of liquid or gas phases include EoS, which is usually based on φ-φ or γ-φ equations. Until very recently, limited research has been available on the simultaneous calculation of the mutual solubility of the CO₂-rich phase and the H₂O-rich phase with high integrity for the application of the CO₂ saline aquifer storage scheme. Furthermore, interaction mechanism coupling with multi-factors associated with the gas–liquid–solid interface properties and the dissolution and acidification process need to be explored in future work.

(4)

Very limited research is available on the pH of gas-saturated aqueous solutions under CO₂ saline aquifer storage conditions. Given the lack of experimental data, it is important to measure the pH of CO₂-saturated brines other than NaCl (aq) over wide ranges of temperature, pressure, and molality relevant to CO₂ storage in deep-saline aquifers in future work.

(5)

Multi-dimensional and multi-scale gas–liquid flow visualization experiments, as well as pore-scale seepage numerical modeling coupling with multi-parameters, need to be carried out. The variation and the interaction mechanism of pore seepage features caused by the physical and chemical interactions of the multi-phase interface properties need to be clarified in future work to provide a theoretical base for and technical guidance on the achievement of the CCUS project with high integrity and efficiency.

The main findings of this study can provide a reference for future research on the interface properties of gas–liquid–solid systems and can illustrate the mechanisms of pore-scale multi-phase flow during the CO₂ geological storage and utilization process to provide a solid technical foundation for the low-carbon, clean, and efficient use of coal and other fossil energy sources, thereby contributing to national energy security.