The potential for carbon dioxide sequestration in subsurface

geological formations

Mohammad A. Nomeli, Amir Riaz

University of Maryland, College Park

Abstract

Fossil fuels are projected to serve as the major source of energy worldwide in the coming
decades. Emissions to the atmosphere of green house gases such as CO2 produced by the
consumption of energy derived from fossil fuels are now comparable to the natural carbon
cycle. Such emissions are already beginning to exert an observable influence on climate
change, perturbing natural climate stability. Carbon sequestration has been proposed to
gradually reduce and eventually eliminate anthropogenic emissions of carbon dioxide into
the atmosphere. The success of sequestration depends on the identification and manipula-
tion of appropriate storage sites where CO2 can be sequestrated over long periods of time.
Deep saline aquifers are the preferred storage sites because of their potential for secure
sequestration over long periods of time as well as their abundance and large capacity. The
primary mechanism for long term sequestration in saline aquifers is solubility trapping as-
sociated with the dissolution of CO2 into brine. The solution of saline water and dissolved
CO2 is negatively buoyant within the subsurface environment. The positively buoyant free
CO2 is thus trapped and loses its ability to escape from the aquifer. The key aspect of solu-
bility trapping is gravitational convection associated with the unstable density stratification
of dissolved CO2 . The rate of mixing of free CO2 with brine is enhanced significantly as
a result of gravitational convection compared with the extremely slow process of diffusion
limited mixing. The strength of convection depends on solubility, which is a function of
temperature and pressure and brine salinity. Dissolution of CO2 into brine leads further
to the formation of carbonic acid that provides carbonate ions needed for the formation
of mineral precipitates. This process of mineral trapping of CO2 is considered to be the
most permanent form of sequestration. The objective of the current study is to provide a
perspective on the progress made thus far towards the modeling of the solubility of CO2 in
brine and the dependence of pH on the saturated mole fraction as a function of pressure,
temperature and salinity. The study is an attempt to facilitate the development of an in-
tegrated model for the prediction of the amount of CO2 that can be safely sequestered in
saline aquifers.

1 Introduction

Emissions of CO2 into the atmosphere resulting from human industrial activity around the
world have been increasing steadily over the past century. The concentration of CO2 in
the atmosphere crossed the 400ppm mark in May 2013, representing an increase of around
10.8% resulting from a 52% increase in emissions over the last 20 years. This large increase
in concentration can be attributed to sustained high emissions of CO2 globally such as, for
example, 30Gt in 2010 alone (Olivier et al., 2013). Table 1 lists the total amount of CO2
emissions by country over the last two decades.

1
 1990 1995 2000 2005 2012
Global 22.7 23.6 25.4 29.3 34.5
United States 4.99 5.26 5.87 5.94 5.19
European Union 27 4.32 4.08 4.06 4.19 3.74
China 2.51 3.52 3.56 5.85 9.86
Russian Federation 2.44 1.75 1.66 1.72 1.77
India 0.66 0.87 1.06 1.29 1.97

Table 1: Emissions of CO2 in billion tonnes by top emitting countries from 1990 to 2012
reported by Olivier et al. (2013)

Carbon dioxide can be sequestered most effectively in deep saline aquifers composed of
sandstone and carbonate formations in sedimentary basins (Michael et al., 2010). The
geologic, chemical and thermodynamic properties of such aquifers are well suited for long
term storage of CO2 in an immobilized state. Carbon dioxide is injected into the aquifer
typically under supercritical conditions where it is lighter than the native brine, which is
made up of salt water and a variety of other dissolved minerals. The injected CO2 thus rises
through brine to either form a stationary pool in a dome or develop as a spreading gravity
current (DOE, 2007) as illustrated in Figs. 1 and 2. Permanent sequestration in such
geologic formations is expected to be achieved over a period of thousands of years through
physical processes such as capillary, solubility and mineral trapping (Bachu, 2003; Orr,
2009b). These mechanisms occur over increasingly longer time scales, ranging from years
to millennia and represent increasingly higher levels of security in terms of keeping the CO2
sequestered (Metz et al., 2005). As a result of these timescales, free CO2 must be prevented
from escaping to fresh water aquifers or the atmosphere by containment underneath an
impermeable cap rock until it is completely sequestered (Bachu et al., 1994). The total
amount of CO2 that can be safely sequestered is thus a function of the rate at which CO2 is
immobilized by the various trapping mechanisms, as well as the structural integrity of the
aquifer as judged by the relative impermeability of the cap rock and the absence of major
geologic faults and fissures (Orr, 2009b). The longer the injected CO2 remains in contact
with the cap rock and fractures due its positively buoyant state, the greater the risk of
leakage to fresh water aquifers and urban areas.
The solution of CO2 and brine next to the free CO2 -brine interface is heavier than the
underlying unsaturated brine. This gives rise to a form of hydrodynamic instability that
produces negatively buoyant plumes of saturated brine (Ennis-King et al., 2003; Riaz et al.,
2006), as sketched in Figs. 1 and 2. These sinking plumes set up a local recirculation of
unsaturated brine along the interface, leading to further dissolution of CO2 and the forma-
tion of additional plumes of saturated brine (Hassanzadeh et al., 2006; Tilton et al., 2013).
This cyclical process continues to convert supercritical CO2 into aqueous CO2 -in-brine so-
lution. The rate at which dissolution takes place is governed entirely by the strength of
convection (Hassanzadeh et al., 2007; Daniel et al., 2013). Instability characteristics such
as the time at which the plumes appear, the rate at which they grow and their late time
evolutionary features govern the effectiveness of solubility trapping (Tilton & Riaz, 2014).
Dissolved CO2 further promotes a more permanent trapping in the form of mineral precip-
itates. The formation of carbonic acid, H2 CO3 , upon dissolution of CO2 into brine raises
the acidity of the solution and provides free calcium for the dissolution of minerals from

2
Figure 1: Schematic of the CO2 sequestration in a structural dome. Supercritical free CO2
pools against the dome and slowly dissolves into the brine underneath. increased solution
results in a gravitational convection in the form of plumes that accelerate dissolution.

Figure 2: Schematic of the sequestration based on capillary and solubility trapping in an

aquifer with slight inclination. A portion of the spreading gravity current of supercritical
CO2 is immobilized through capillary trapping. Gravitational instability develops to further
deplete the gravity layer of supercritical CO2 .

the porous matrix and the formation of mineral precipitates, respectively. The formation
of such precipitates would change the local porosity that would in turn affect the pattern
of convection.
The process of solubility trapping driven by natural convection is the most effective mech-
anism for dissolving CO2 in brine and for preventing free CO2 from escaping to the atmo-
sphere (Riaz et al., 2006). Accurate prediction of the extremely large time scales associated
with complex physical processes is necessary for the identification of storage sites and their
engineering optimization as well as for understanding the risk of leakage (Orr, 2009a). In or-
der to achieve this objective, a rigorous characterization of physical and chemical processes
leading to capillary, solubility and mineral trapping is necessary (Ide et al., 2007; Riaz
& Cinar, 2013). Not only must the effects specific to individual trapping mechanisms be
quantified, but it must also be determined how these mechanisms interact to affect the long
term sequestration behavior in the aquifer. Such an understanding is being actively devel-
oped from a fundamental perspective. Reviews on gravity current migration can be found

3
 Saline Basins Low Estimate High Estimate
(billion tones) (billion tonnes)
Bighorn Basin 10.80 148.54
Montana Thrust Belt 2.43 33.43
North-Central Montana 67.38 926.494
Powder River Basin 14.46 198.89
Southwest Montana 2.10 28.88
Southwestern Wyoming 45.91 511.38
Williston Basin 58.37 802.66
Wind River Basin 13.65 71.01
Wyoming Thrust Belt 6.03 82.89
Total 221.15 2,804.21

Table 2: Estimated maximum CO2 storage capacity (GtCO2 ) in selected saline formations
in the United States, reported by National Energy Technology Laboratory (2012)

in Huppert & Woods (1995), Nordbotten & Celia (2006), Hesse et al. (2007), Hesse et al.
(2008), de Loubens & Ramakrishnan (2011) and Szulczewski & Juanes (2013). Similarly,
developments in the area of geochemistry relevant to mineral trapping are also reported
by Xu et al. (2005), Ennis-King & Paterson (2007), Steefel & Maher (2009), Molins et al.
(2012) and Jun et al. (2013). The progress in the area of solubility trapping is reviewed by
Neufeld et al. (2010), Ghesmat et al. (2011), Ghesmat et al. (2011), Cheng et al. (2012),
Elenius et al. (2012), MacMinn et al. (2012), Hewitt et al. (2013) and Tilton & Riaz (2014).
The current study provides an overview of the progress made in the two areas of thermody-
namic modeling for quantifying solubility trapping and the modeling of pH for predicting
the extent of mineral trapping. Because both the strength of natural convection and the
rate of precipitate formation in saline aquifers depend on the saturated mole fraction of
dissolved CO2 , accurate modeling of the thermodynamic basis of solubility is necessary for
developing reliable estimates of storage capacity and the associated time scales. The total
capacity available for sequestration in saline aquifers in the United States is listed in Table
2 at some specific locations based on the data reported by National Energy Technology
Laboratory (2012). Estimates of the fraction of this space that would be actually occupied
by the injected CO2 as well the associated time scales are not currently available due to
the complex nature of the overall coupled process of sequestration. However, the accuracy
of thermodynamic models for solubility and pH would play a key role in future integrated
assessments of actual storage capacity. The current study therefore provides an overview
of the present state of modeling of solubility and pH.

2 Calculation of dissolved mole fraction at saturation

Upon injection of CO2 into saline aquifers, production of weak carbonic acid (H2 CO3 ) results
from the dissolution of supercritical CO2 into aquifer brine. The subsequent increase in the
acidity of the solution induces primary minerals to dissolve into brine (He & Morse, 1993;
Palandri, 2004). Precipitation of secondary carbonate minerals follows when aqueous species
of dissolved CO2 react with dissolved minerals (Prigiobbe et al., 2009; Guyot et al., 2011); a

4
process known as mineral trapping. Solubility and mineral trapping are widely considered
to be the most promising long term solutions to geologic CO2 sequestration (Metz et al.,
2005).
In order to find the reaction rates associated with mineral trapping of CO2 , as a function
of temperature and pH, it is necessary to first quantify the solubility of CO2 in brine. We
consider the injection of high-pressure CO2 into a system made up of a relatively deep aquifer
hosting an aqueous solution primarily of high sodium chloride NaCl salinity. In order to
find the solubility of CO2 , we first need to evaluate the temperature and pressure conditions
that depend on the depth of the aquifer as well as on the local geothermal gradient and
pressure gradient. The depth of deep saline aquifers can be of the order of a kilometer.
Taking a surface temperature of 20 C and average geothermal gradient of 30 C km1 ,
the temperature can be between 50 C at 1 km and 110 C at 3 km. Further taking the
hydrostatic pressure gradient of 100 bar km1 , aquifer pressures can be around 101 bar at 1
km and 401 bar at 4 km. Therefore, the investigation of the sequestration potential of deep
saline aquifers requires knowledge of the thermodynamic properties of CO2 -H2 O mixtures
up to temperatures of 100 C and pressures of around upto several hundred bars. The effect
of dissolved NaCl on the solubility of CO2 also needs to taken into account. In the following
we first review the most effective approach for determining the mole fraction of dissolved
CO2 at high temperatures and pressures and subsequently associate this information with
the calculation of the pH of the solution at saturation that is needed for reaction rates of
dissolved CO2 and minerals dissolved in brine.
Assuming brine to be made up of H2 O and NaCl, the solubility of CO2(aq) in brine can be
quantified by the mole fraction, xCO2 , given by
mCO2
xCO2 = , (1)
mCO2 + mH2 O + 2 mNaCl

where mCO2 , mH2 O and mNaCl denote, respectively, the molality of CO2 , H2 O and NaCl in
the solution. The factor of 2 for mNaCl is due to the complete dissociation of dissolved NaCl
into Na+ and Cl ions. Molality of CO2 can be expressed as the ratio, mCO2 = aCO2 /CO2 ,
of the activity, aCO2 , of CO2 in the solution and the activity coefficient, CO2 . The activity
of dissolved CO2 and the fugacity in the CO2 rich phase, fCO2 , are related through the
equilibrium constant, K = fCO2 / aCO2 . Using these relations, the mole fraction of CO2 in
Eq. (1) can be expressed as
fCO2
xCO2 =   , (2)
fCO2 + mH2 O + 2 mNaCl KCO2

The effect of dissolved salts is important and increases with NaCl concentration. This effect
will be expressed through the activity of liquid water and the activity coefficient for the
aqueous CO2 . Accurate values of water activity for solutions of various electrolytes can be
obtained using the Pitzer & Mayorga (1973) ion interaction model (e.g., Rumpf & Maurer
(1993)). However, in the salinity range up to ionic strength around 6 molal (below halite
saturation), the model can be simplified by assuming that the water activity equals its mole
fraction on the basis of a fully ionized salt (Spycher & Pruess, 2005).
Figure 3 depicts the variation of activity coefficient of aqueous CO2 in NaCl solutions with
different molalities. Activity coefficient of CO2 increases with the molality of NaCl. In

5
 2.35 Duan and Sun (2003)
Helgeson (1969)
Rumpf (1994)
Drummond (1981)
2.05

CO2 1.75 m=3

1.45
m=1

1.15
320 330 340 350 360 370 380
Temperature (K)
Figure 3: Activity coefficients for aqueous CO2 from various sources for 1 and 3 mol/kg
NaCl aqueous solutions in the temperature range of 50-100 C. Activity coefficient of CO2
increases with molality of NaCl, m, and decreases with temperature.

most cases higher temperatures lead to a decrease in the activity coefficient, which would
indicate a lower solubility based on Eq. 2. However an increase in temperature also affects
fugacity, fCO2 , and the equilibrium constant, K, in Eq. 2, as described in more detail below.
Hence, an increase in temperature is not always accompanied by an increase in solubility.
In order to find the mole fraction of CO2 in the solution for particular molalities of H2 O and
NaCl, corresponding values of the activity coefficient, the equilibrium constant (Krauskopf
& Bird, 1995) and fugacity need to be determined.
The equilibrium constant, K(eq) , is related to the standard free energy change by G =
RT ln K. Considering
(G /T) H
 
= 2 . (3)
T P T
Here, H denotes the enthalpy of the reaction, T is the temperature, G denotes the gibbs
free energy. Vant Hoff equation has been employed to find the effect of temperature on the
equilibrium constant (Krauskopf & Bird, 1995).

H
 
ln K
= . (4)
T P RT2
Here R is the gas constant. The variation of the thermodynamic equilibrium constant with
respect to pressure is given by (Garrels & Christ, 1965).

V
 
ln K
= , (5)
P T RT

where V is the change in the molar volume of the reaction in the standard state. In
general, however, the change in the equilibrium with respect to pressure is quite small
(Van Eldik et al., 1989; Atkins, 1978).

6
 Table 3: Values of the coefficients used in the Redlich-Kwong equation.

Value Units
Q 7.54 107 3.96 104 T(K) bar cm6 K0.5 mol2
S 27.80(0.01) cm3 /mol

Following Marini (2007), the equilibrium constant, K, in Eq. 2 is given by

" #

VCO2
K = K exp (P P ) , (6)
RT
where, R, T and P are, respectively, the ideal gas constant, temperature and pressure. P
is the standard pressure and VCO2 = 32.6 cm3 /mol is the average partial molar volume
of CO2 in the temperature range, (T<100 C) and pressure range up to 600 bar (Marini,
2007). In this range of temperatures, the equilibrium constant at standard pressure, K ,
can be obtained for the supercritical, K(g) , and the subcritical, K(l) , states of the CO2 rich
phase as

logK(g) = 1.189 + 0.01304 T 5.446 105 T2 , (7)

logK(l) = 1.169 + 0.01368 T 5.380 105 T2 , (8)

where, K(g) is used when the temperature and the volume of the gas phase are above the
critical parameters of CO2 , 31 C and 94 cm3 mol1 , respectively. K(l) is given by Eq. (8), is
used when the temperature and the volume of the gas phase are below the critical parameters
of CO2 .
The fugacity for the CO2 rich phase in the supercritical state is determined from the fol-
lowing relation,
fCO2 = yCO2 P , (9)
where, and yCO2 are, respectively, the fugacity coefficient and the mole fraction of CO2
in the supercritical phase. For the range of pressures (up to 600 bar) and temperatures
(50-100 C) concerned with CO2 sequestration problem, for calculating the fugacity of CO2 ,
Spycher & Pruess (2010b) have shown that that mole fraction of H2 O in the compressed
gas phase can be neglected (yCO2 1) when T 100 C because of the negligible amount
of H2 O in the CO2 rich phase (Spycher et al., 2003). The fugacity coefficient of CO2 , , is
then given by

     
v S 3Q v+S S
ln = ln + ln ln Z , (10)
vS v S SRT3/2 v 3(v + S)
where v is the molar volume of CO2 , Z = Pv/RT is the compressibility factor of the CO2 rich
phase and Q and S are the correcting factors that are given in Table 3 (Spycher & Pruess,
2010a) for the temperature range of 283 380 K. Molar volumes of CO2 for temperatures
between 50 to 100 C and pressures up 600 bar in the gas phase and the aqueous phase can
be obtained with the equation of state given by Spycher & Pruess (2010b) and Duan et al.
(1992), respectively.

7
 3

2.5

2
1.5
This study
2
Koschel et al. (2006)
1 Mueller et al. (1988)
Shagiakhmetov & Tarzimanov (1981)
0.5 Malinin & Kurovskaya (1975)
Prutton & Savage (1945)
Wiebe & Gaddy (1939)
0
0 100 200 300 400 500 600
Pressure (bar)
Figure 4: Mole fraction of dissolved CO2 as a function of pressure at 100 C. The mole
fraction x2 , obtained from the modified Redlich-Kwong EOS agrees well with experimental
data for saturated binary solutions.

 
3 RT 2 RTS Q 2 QS
v v 1/2 + S v 1/2 = 0 . (11)
P P T P T P

The maximum root of Eq. 11, vmax , gives the molar volume of CO2 in the gas phase while the
molar volume in the liquid phase is supposedly given by the minimum root, vmin (Marini,
2007). The molar volume in the gas phase, vmax allows the calculation of the fugacity
coefficient, , from Eq. (10), which yields the value of fugacity, fCO2 , from Eq. (9) for a
given pressure, P.
The solubility of CO2 in pure water is compared with experimental data in Figure 4. Al-
though early experimental studies were focused mainly on pressures and temperatures much
higher than those of interest for the geological sequestration of CO2 , data have been reported
also in the pressure, temperature range relevant to the CO2 sequestration in brine aquifers.
Wiebe & Gaddy (1939) studied the solubility of CO2 in water at 100 C, at pressures to
700 atm. Malinin & Kurovskaya (1975) also studied solubility of CO2 in chloride solutions
at elevated temperatures and pressures. The results of the model proposed by Spycher &
Pruess (2010b), which is employed in the current study at lower pressure range, are in good
agreement with the data obtained by Muller et al. (1988). There are also some other valu-
able data given by Shagiakhmetov and Tarzimanov (1981), Koschel et al. (2006)., Prutton
and Savage (1945). Comparison of the results from Eq. 2 (solid line) with experimental
data (symbols) including the standard deviation at 100 C and up to 600 bar reveals that
this model is in good agreement with experimental data reported from the other researchers.
however, there is also some small discrepancy between the model and experimental data.
This small differences between experimental and calculated results of CO2 solubility may
arise due to factors not accounted for in the theoretical model such as hydrogen bond, hy-
dration, chemical reaction and ionization (Duan et al., 2008). The hypothesis of infinite

8
 2.5 2
(a) (b)
m=1
2 1.6
m=1

xCO2 (10-2)
m=3
xCO2 (10-2)

1.5 1.2

m=3
1 0.8

0.5 0.4

0 0
0 100 200 300 400 500 600 320 330 340 350 360 370 380
Pressure (bar) Temperature (K)

Figure 5: The calculated values of mole fraction of CO2 by means of Eq. 2 based on various
activity coefficient models; Duan & Sun (2003) (solid lines), Rumpf et al. (1994) (crosses),
Fournier (1985) (diamonds), Drummond (1981) (triangle). (a) T= 50 C and (b) P= 100
bar. The solubility of CO2 increases with pressure. Higher temperature and salinity values
lower solubility.

dilution of water in CO2 -rich liquid phase also may cause the deviations (Spycher & Pruess,
2005).
We evaluate the solubility of CO2 for various values of temperature, pressure and salinity
based on various activity coefficient models in saline water in Figure 5. We find that at
50 C the solubility based on the activity coefficients of Duan & Sun (2003) (solid lines),
Rumpf et al. (1994) (crosses) and Drummond (1981) (triangle) are in good agreement.
The solubility calculation based on the activity coefficients of Fournier (1985) (diamonds),
however, gives lower mole fractions of dissolved CO2 compared to others. This variation is
more pronounced at higher pressure, temperature and salinity values. Figure 5 also shows
that the mole fraction of CO2 increases with pressure. However, higher temperatures and
salinities lead to a decrease in the mole fraction. The solubility results from this model,
based on different activity coefficient models, are employed to find the pH of the system in
the following section.

3 Dependence of pH on pressure, temperature and salinity

Dissolution of CO2 into brine results in the production of carbonic acid (H2 CO3 ) leading
to the subsequent generation of [H+ ] ions. We determine the pH of the aqueous phase as a
function of the amount of CO2 dissolved, which in turn depends on conditions of pressure,
temperature and salinity of brine. The pH of CO2 enriched brine can be determined on the
basis of the electro-neutrality condition by combining the speciation equilibrium model of
Li & Duan (2007) with the solubility model of Spycher & Pruess (2005) as follows,

p
[H+ ] = (Ka1 + Ka2 ) CT + Kw , (12)
where, Ka1 and Ka2 are the first and second dissociation constants, respectively. Kw =
[H+ ][OH ] 1014 at standard pressure and temperature is the water equilibrium constant
(Quist, 1970). The total concentration of all species in the aqueous phase is denoted by CT .

9
 4.3 3.4
(a) (b)
m=3

3.95 3.3
m=1

m=0
pH 3.6 pH 3.2

3.25 m=3 3.1

m=1
m=0
2.9 3
0 100 200 300 400 500 600 320 330 340 350 360 370 380
Pressure (bar) Temperature (K)

Figure 6: The pH variation of binary solution, H2 O-CO2 , for different values of salinity
based on various activity coefficient models; model 1(solid line), model 2(cross), model
3(diamond) and model 4(triangle). Experimental data of Toews et al. (1995) for m = 0 is
shown with circles. (a) T= 50 C and (b) P= 100 bar.

The dissociation constants for the reactions considered must be valid for the temperature
and pressure range of the system.
The dissociation constant of water, Kw , has been studied by several researchers. The model
of Marshall & Franck (1981) is more reliable under aquifer conditions and will be employed
in the current study to find the total concentration of [H+ ]. The total concentration of
species in the aqueous phase is given by

CT = [H2 CO3 ] + [HCO 2

3 ] + [CO3 ] . (13)
Carbonic acid, H2 CO3 , has two protons which may dissociate from the original molecule
and make the brine acidic (Meyssami et al., 1992; Li & Duan, 2007). Thus, there are two
dissociation constants. The first represents the dissociation into the bicarbonate (also called
hydrogen carbonate) ion,

[HCO +
3 ][H ]
Ka1 = . (14)
[H2 CO3 ]
The first dissociation constant of carbonic acid Ka1 = kd /kr = 4.6 107 at 25 C, where kd
is the rate constant for dissociation and kr is the rate constant for acid recombination. The
dissociation rate of carbonic acid is very fast (kd = 8 105 s1 ) (Brezonik & Arnold, 2011),
therefore the kinetics of this reaction does not concern us further. In the current study, the
values for Ka1 are obtained from the model of Read (1975) at different temperatures and
pressures. The second dissociation constant, Ka2 , for the dissociation of bicarbonate ion
into carbonate ion, CO2 3 , is given by

[CO2 +
3 ][H ]
Ka2 = . (15)
[HCO 3]

The second dissociation constant is Ka2 = 4.681011 at 25 C, which is negligible compared

to Ka1 (Mehrbach et al., 1973) and can be neglected in Eq. 12.

10
At the macroscopic scale an aqueous solution is found to be electrically neutral (Stumm,
1970), therefore, charge must be conserved. The electro neutrality condition is

[H+ ] + [Na+ ] = 2[CO2
3 ] + [HCO3 ] + [OH ] + [Cl ] . (16)
Molarity of Na+ and Cl are equal to molality of NaCl (Li & Duan, 2007). Substituting
concentration of [H2 CO3 ] and [HCO
3 ] from Eq. 14 and 15 into the Eq. 13 we find

[H+ ]2 [CO23 ] [H+ ][CO2

3 ]
CT = + + [CO2
3 ]
Ka1 Ka2 Ka2
[H+ ] [H+ ]2
 
= [CO2
3 ] 1 + + . (17)
Ka2 Ka1 Ka2

Therefore, concentrations of deprotonated forms HCO 2

3 (bicarbonate) and CO3 (carbon-
ate) depend on the concentration of [H+ ] (Stumm, 1970).

Ka1 Ka2 CT
[CO2
3 ]= . (18)
[H+ ]2 + Ka1 [H+ ] + Ka1 Ka2

Ka1 [H+ ]CT

[HCO
3]= . (19)
[H+ ]2 + Ka1 [H+ ] + Ka1 Ka2
Substituting these into the electro neutrality condition, Eq. 16, and combining it with
Kw = [H+ ][OH ] we derive a 4th degree equation in [H+ ]:

2Ka1 Ka2 + Ka1 [H+ ] Kw

[H+ ] = CT + 2 + + + . (20)
[H ] + Ka1 [H ] + Ka1 Ka2 [H ]
The condition, Ka1  Ka2 and Ka1 [H+ ]  Ka2 [H+ ], allows the following simplification

2Ka1 Ka2 + Ka1 [H+ ] 2Ka1 Ka2 + Ka1 [H+ ] + Ka2 [H+ ]

[H+ ]2 + Ka1 [H+ ] + Ka1 Ka2 [H+ ]2 + Ka1 [H+ ] + Ka1 Ka2 + Ka2 [H+ ]
Ka1 Ka2
= + + . (21)
Ka1 + [H ] Ka2 + [H+ ]

Substituting Eq. 21 into the Eq. 20 we find

Ka1 Ka2 Kw
[H+ ] = + CT + + CT + . (22)
Ka1 + [H ] Ka2 + [H ] [H+ ]
Finally, pH is the negative logarithm of the activity of hydrogen ions in solution, pH=-Log
[H+ ]. For the acidic range (Hellmann, 1994) where the concentration of [H+ ] is higher,
assuming that Ka2  Ka1  [H+ ] can help simplifying the equation even further to Eq. 12.
Figure 6 depicts the variation of pH with respect to changes in pressure, for T=50 C, and
changes in temperature, for P=100 bar, in plots (a) and (b), respectively. Three different
levels of salinity, m=0, 1 and 3, are considered. As a function of pressure, pH drops rapidly
when the pressure, P, increases from 1 bar to around 100 bar for all values of m, as shown

11
in Fig. 6(a). For P> 100 bar, pH decays relatively slowly and at a constant rate. Values
of pH based on different activity coefficients are found to be quite similar for both cases of
m=1 and m=3. As a function of temperature, pH increases at a relatively constant rate, as
shown in Fig. 6(b). The difference in pH based on different models of the activity coefficient
is small for m=1 but increases for m=3, particularly at smaller values of the temperature.
The pH variation shown in Fig. 6 generally follows the trend of the underlying solubility
behavior observed in Figure 5, i.e., an increase in the mole fraction with pressure and a
decrease in the mole fraction with temperature. In the same range of pressure, larger values
of temperature and salinity tend to lower the solubility of CO2 in brine, which limits the
pH reduction.
Figure 6(a) also compares the computed values of pH with those measured experimentally
for the H2 O-CO2 system by Toews et al. (1995). The experimentally determined values of
pH at 50 C vary from 3.03 to 3.16 for the pressure range of about 70-200 bar. A compar-
ison with calculated pH values in Fig. 6(a) indicates a close similarity. Though a slight
deviation is observed to set in for larger values of the pressure. This small discrepancy is
due perhaps to different CO2 mole fractions and ionic strengths used in the experiment.
The [H+ ] generated by dissolution of CO2 will enhance dissolution of mineral compositions
to consume available [H+ ] in the aqueous phase which may buffers the pH of the system.
Due to the generation of cation-HCO 3 complexation, the capacity of the brine will also
increase to contain dissolved CO2 . Additionally, direct measurement of Ka1 and Ka2 itself
may involve errors. For example, Ellis (1959) cites errors up to 20% in conductivity mea-
surements used in the determination of Ka1 and Ka2 values. Nevertheless, this comparison
with experimental measurements of pH is encouraging and helps build confidence in the
application of the model to m>0 cases.

4 Conclusions

The rate of mineral trapping of CO2 in the form of minerals calcite in deep saline aquifer
depends on the knowledge of the amount of dissolved CO2 and the pH of the solution under
saturation conditions. A modified Redlich-Kwong equation of state can be used based
on various activity coefficient models to determine how temperature, pressure and salinity
affect the solubility of CO2 and the pH of the solution. The electro-neutrality condition can
be used to determine the pH.
The activity coefficients of CO2 due to the presence of NaCl affects solubility, pH and
kinetics of chemical reactions. However, pH and mineral kinetics are relatively less sensitive
to the activity coefficient calculations compared to solubility. The solubility of CO2 in liquid
phase increases with pressure while temperature and salinity have an opposite effect. Larger
values of the activity coefficients lead to a decrease in the mole fraction of CO2 in aqueous
phase. Dissolution of CO2 leads to the production of carbonic acid that decreases the pH of
the system. Larger value of temperature and salinity, however, tend to lower the solubility of
CO2 in brine, which limits the pH reduction. The modeling approach followed in this study
shows good agreement with experimental work reported in previously published studies and
can be used for the prediction of CO2 potential in deep saline aquifers.

12
References
(2007). Basic research needs for geosciences: Facilitating 21st century energy systems.
Tech. rep., Workshop sponsored by the US Department of Energy, Office of Basic Energy
Sciences, Bethesda MD. Page 53.

Atkins, P. (1978). Molecules in motion: ion transport and molecular diffusion. Physical
Chemistry. PW Atkins, editor. Oxford University Press. Oxford, UK , (pp. 819848).

Bachu, S. (2003). Screening and ranking of sedimentary basins for sequestration of co2 in
geological media in response to climate change. Environmental Geology, 44 (3), 277289.

Bachu, S., Gunther, W. D., & Perkins, E. H. (1994). Aquifer disposal of CO2 : Hydrody-
namic and mineral trapping. Energy Convers. Mgmt., 35 , 269279.

Brezonik, P., & Arnold, W. (2011). Water chemistry: an introduction to the chemistry of
natural and engineered aquatic systems. Oxford University Press.

Cheng, P., Bestehorn, M., & Firoozabadi, A. (2012). Effect of permeability anisotropy on
buoyancy-driven flow for CO2 sequestration in saline aquifers (vol 48, W09539, 2012).
Water Resources Research, 48 .

Daniel, D., Tilton, N., & Riaz, A. (2013). Optimal perturbations of gravitationally unstable,
transient boundary layers in porous media. Journal of Fluid Mechanics, 727 , 456487.

de Loubens, R., & Ramakrishnan, T. S. (2011). Analysis and computation of gravity-

induced migration in porous media. Journal of Fluid Mechanics, 675 , 6086.

Drummond, S. (1981). Boiling and mixing of hydrothermal fluids: chemical effects on

mineral precipitation. Pennsylvania State University.

Duan, Z., Hu, J., Li, D., & Mao, S. (2008). Densities of the CO2 H2 O and CO2 H2 ONaCl
systems up to 647 K and 100 MPa. Energy & Fuels, 22 (3), 16661674.

Duan, Z., Mller, N., & Weare, J. H. (1992). An equation of state for the CH4 CO2 H2 O
system: II. mixtures from 50 to 1000 C and 0 to 1000 bar. Geochimica et Cosmochimica
Acta, 56 (7), 26192631.

Duan, Z., & Sun, R. (2003). An improved model calculating CO2 solubility in pure water
and aqueous NaCl solutions from 273 to 533 K and from 0 to 2000 bar. Chemical Geology,
193 (3), 257271.

Elenius, M. T., Nordbotten, J. M., & Kalisch, H. (2012). Effects of a capillary transition
zone on the stability of a diffusive boundary layer. IMA Journal of Applied mathematicS ,
77 (6, SI), 771787.

Ellis, A. (1959). 750. The effect of pressure on the first dissociation constant of carbonic
acid. J. Chem. Soc., (pp. 36893699).

Ennis-King, J., & Paterson, L. (2007). Coupling of geochemical reactions and convective
mixing in the long-term geological storage of carbon dioxide. International J. Greenhouse
Gas Control , 1 (1), 8693.

13
Ennis-King, J., Preston, I., & Paterson, L. (2003). Onset of convection in anisotropic porous
media subject to a rapid change in boundary conditions. Phys. Fluids, 17 , Article no.
084107.

Fournier, R. (1985). Carbonate transport and deposition in the epithermal environment.

Reviews in economic geology, 2 , 6372.

Garrels, R., & Christ, C. (1965). Minerals, solutions, and equilibria. Minerals, solutions,
and equilibria.

Ghesmat, K., Hassanzadeh, H., & Abedi, J. (2011). The effect of anisotropic dispersion on
the convective mixing in long-term co2 storage in saline aquifers. AIChE Journal , 57 (3),
561570.

Ghesmat, K., Hassanzadeh, H., & Abedi, J. (2011). The impact of geochemistry on convec-
tive mixing in a gravitationally unstable diffusive boundary layer in porous media: CO2
storage in saline aquifers. Journal of Fluid Mechanics, 673 , 480512.

Guyot, F., Daval, D., Dupraz, S., Martinez, I., Menez, B., & Sissmann, O. (2011). CO2 geo-
logical storage: The environmental mineralogy perspective. Comptes Rendus Geoscience,
343 (2), 246259.

Hassanzadeh, H., Pooladi-Darvish, M., & Keith, D. W. (2006). Stability of a fluid in a

horizontal saturated porous layer: effect of non-linear concentration profile, initial, and
boundary conditions. Trans Porous Media, 65 , 193211.

Hassanzadeh, H., Pooladi-Darvish, M., & Keith, D. W. (2007). Scaling behavior of convec-
tive mixing, with application to geological storage of CO2 . AIChE J., 53 (5), 11211131.

He, S., & Morse, J. (1993). The carbonic acid system and calcite solubility in aque-
ous Na-K-Ca-Mg-Cl-SO4 solutions from 0 to 90 C. Geochimica et Cosmochimica Acta,
57 (15), 35333554.

Hellmann, R. (1994). The albite-water system: Part i. the kinetics of dissolution as a

function of pH at 100, 200 and 300C. Geochimica et Cosmochimica Acta, 58 (2), 595
611.

Hesse, M., Riaz, A., & Tchelepi, H. A. (2007). Convective dissolution of co2 in saline
aquifers. In Geochimica ET. Cosmochimica Acta, vol. 71 of 15 , (pp. A401A401). Amer-
ican Geophysical Union, Pergamon-Elsevier Science Ltd, Oxford.

Hesse, M. A., Orr, F. M., & Tchelepi, H. A. (2008). Gravity currents with residual trapping.
Journal Of Fluid Mechanics, 611 , 3560.

Hewitt, D. R., Neufeld, J. A., & Lister, J. R. (2013). Convective shutdown in a porous
medium at high Rayleigh number. Journal of Fluid Mechanics, 719 , 551586.

Huppert, H., & Woods, A. (1995). Gravity-driven flows in porous layers. Journal of Fluid
Mechanics, 292 , 5569.

Ide, S. T., Jessen, K., & Orr, F. M. (2007). Storage of co2 in saline aquifers: Effects of
gravity, viscous, and capillary forces on amount and timing of trapping. International
Journal Of Greenhouse Gas Control , 1 (4), 481491.

14
Jun, Y.-S., Giammar, D. E., & Werth, C. J. (2013). Impacts of Geochemical Reactions on
Geologic Carbon Sequestration. Environmental Science & Technology, 47 (1), 38.

Krauskopf, K., & Bird, D. (1995). Surface chemistry: the solution-mineral interface. Intro-
duction to geochemistry (Ed MG-HI Editions) Mc Graw-Hill International Editions edn,
Earth Sciences and Geology Series, (pp. 135163).

Li, D., & Duan, Z. (2007). The speciation equilibrium coupling with phase equilibrium in
the H2 O CO2 NaCl system from 0 to 250 C, from 0 to 1000 bar, and from 0 to 5
molality of NaCl. Chemical Geology, 244 (3), 730751.

MacMinn, C. W., Neufeld, J. A., Hesse, M. A., & Huppert, H. E. (2012). Spreading and
convective dissolution of carbon dioxide in vertically confined, horizontal aquifers. Water
Resources Research, 48 .

Malinin, S., & Kurovskaya, N. (1975). Solubility of CO2 in chloride solutions at elevated
temperatures and CO2 pressures. Geochem. Int, 12 (2), 199201.

Marini, L. (2007). Geological sequestration of carbon dioxide: thermodynamics, kinetics,

and reaction path modeling, vol. 11. Elsevier Science.

Marshall, W., & Franck, E. (1981). Ion Product of Water Substance, 0-1000 C, 1-10,000
bars: New International Formulation and Its Background . American Chemical Society
and the American Institute of Physics for the National Bureau of Standards.

Mehrbach, C., Culberson, C., Hawley, J., & Pytkowicz, R. (1973). Measurement of the
apparent dissociation constants of carbonic acid in seawater at atmospheric pressure.
Limnology and Oceanography, (pp. 897907).

Metz, B., Davidson, O., De Coninck, H., Loos, M., & Meyer, L. (2005). IPCC special report
on carbon dioxide capture and storage: Prepared by working group III of the intergovern-
mental panel on climate change. IPCC, Cambridge University Press: Cambridge, United
Kingdom and New York, USA, 2 .

Meyssami, B., Balaban, M. O., & Teixeira, A. A. (1992). Prediction of pH in model systems
pressurized with carbon dioxide. Biotechnology progress, 8 (2), 149154.

Michael, K., Golab, A., Shulakova, V., Ennis-King, J., Allinson, G., Sharma, S., & Aiken,
T. (2010). Geological storage of CO2 in saline aquifers-A review of the experience from
existing storage operations. International Journal of Greenhouse Gas Control , 4 (4),
659667.

Molins, S., Trebotich, D., Steefel, C. I., & Shen, C. (2012). An investigation of the effect of
pore scale flow on average geochemical reaction rates using direct numerical simulation.
Water Resources Research, 48 .

Muller, G., Bender, E., & Maurer, G. (1988). Das dampf-flussigkeitsgleichgewicht des
ternaren systems ammoniak-kohlendioxid-wasser bei hohen wassergehalten im bereich
zwischen 373 und 473 Kelvin. Berichte der Bunsengesellschaft fur Physikalische Chemie,
92 (2), 148160.

National Energy Technology Laboratory (2012). United States carbon utilization and stor-
age Atlas. Tech. rep., Department of Energy.

15
Neufeld, J. A., Hesse, M. A., Riaz, A., Hallworth, M. A., Tchelepi, H. A., & Huppert, H. E.
(2010). Convective dissolution of carbon dioxide in saline aquifers. Geophysical Research
Letters, 37 .

Nordbotten, J. M., & Celia, M. A. (2006). Similarity solutions for fluid injection into
confined aquifers. Journal Of Fluid Mechanics, 561 , 307327.

Olivier, J. G., Janssens-Maenhout, G., & Peters, J. A. (2013). Trends in global CO2 emis-
sions: 2013 report. PBL Netherlands Environmental Assessment Agency.

Orr, F. M. (2009a). Co2 capture and storage: are we ready? Energy & Environmental
Science, 2 (5), 449458.

Orr, F. M. (2009b). Onshore geologic storage of co2 . Science, 325 , 16561658.

Palandri, J. (2004). A compilation of rate parameters of water-mineral interaction kinetics

for application to geochemical modeling. Tech. rep., DTIC Document.

Pitzer, K., & Mayorga, G. (1973). Thermodynamics of electrolytes. II. activity and osmotic
coefficients for strong electrolytes with one or both ions univalent. The Journal of Physical
Chemistry, 77 (19), 23002308.

Prigiobbe, V., Hanchen, M., Werner, M., Baciocchi, R., & Mazzotti, M. (2009). Mineral
carbonation process for CO2 sequestration. Energy Procedia, 1 (1), 48854890.

Quist, A. (1970). Ionization constant of water to 800. deg. and 4000 bars. The Journal of
Physical Chemistry, 74 (18), 33963402.

Read, A. (1975). The first lonization constant of carbonic acid from 25 to 250 C and to
2000 bar. Journal of Solution Chemistry, 4 (1), 5370.

Riaz, A., & Cinar, Y. (2013). Carbon dioxide sequestration in saline formations: Part I -
Review of solubility modeling studies. Under review. Journal of Petroleum Science and
Technilogy.

Riaz, A., Hesse, M., Tchelepi, H., & Orr, F. (2006). Onset of convection in a gravitationally
unstable diffusive boundary layer in porous media. Journal of Fluid Mechanics, 548 (1),
87111.

Rumpf, B., & Maurer, G. (1993). An experimental and theoretical investigation on the
solubility of carbon dioxide in aqueous solutions of strong electrolytes. Berichte der
Bunsengesellschaft fuer physikalische chemie, 97 (1), 8597.

Rumpf, B., Nicolaisen, H., Ocal, C., & Maurer, G. (1994). Solubility of carbon dioxide in
aqueous solutions of sodium chloride: Experimental results and correlation. Journal of
solution chemistry, 23 (3), 431448.

Spycher, N., & Pruess, K. (2005). CO2 -H2 O mixtures in the geological sequestration of
CO2 . ii. partitioning in chloride brines at 12-100 C and up to 600 bar. Geochimica et
Cosmochimica Acta, 69 (13), 33093320.

16
Spycher, N., & Pruess, K. (2010a). The correcting factors are obtained from the private
communication with Dr. Nicolas Spycher which is slightly different from his paper a
phase-partitioning model for CO2 -brine mixtures at elevated temperatures and pressures:
Application to CO2 -enhanced geothermal systems. Transport in porous media, 82 (1),
173196.

Spycher, N., & Pruess, K. (2010b). A phase-partitioning model for CO2 -brine mixtures at
elevated temperatures and pressures: Application to CO2 -enhanced geothermal systems.
Transport in porous media, 82 (1), 173196.

Spycher, N., Pruess, K., & Ennis-King, J. (2003). CO2 -H2 O mixtures in the geological
sequestration of CO2 . i. assessment and calculation of mutual solubilities from 12 to
100C and up to 600 bar. Geochimica et cosmochimica acta, 67 (16), 30153031.

Steefel, C. I., & Maher, K. (2009). Fluid-rock interaction: A reactive transport approach.
Reviews in Mineralogy and Geochemistry, 70 (1), 485532.

Stumm, W. (1970). Aquatic Chemistry; An Introduction Emphasizing Chemical Equilib-

ria in Natural Waters by Werner Strumm and James J. Morgan. New York, Wiley-
Interscience.

Szulczewski, M. L., & Juanes, R. (2013). The evolution of miscible gravity currents in
horizontal porous layers. Journal of Fluid Mechanics, 719 , 8296.

Tilton, N., Daniel, D., & Riaz, A. (2013). The initial transient period of gravitationally
unstable diffusive boundary layers developing in porous media. Physics of Fluids, 25 (9).

Tilton, N., & Riaz, A. (2014). Nonlinear stability of gravitationally unstable, transient,
diffusive boundary layers in porous media. Journal of Fluid Mechanics, 745 , 251278.

Toews, K., Shroll, R., Wai, C., & Smart, N. (1995). pH-defining equilibrium between
water and supercritical CO2 . influence on SFE of organics and metal chelates. Analytical
Chemistry, 67 (22), 40404043.

Van Eldik, R., Asano, T., & Le Noble, W. (1989). Activation and reaction volumes in
solution. 2 [erratum to document cited in CA110 (24): 219738v]. Chemical Reviews,
89 (8), 19811981.

Wiebe, R., & Gaddy, V. (1939). The solubility in water of carbon dioxide at 50, 75 and
100, at pressures to 700 atmospheres. Journal of the American Chemical Society, 61 (2),
315318.

Xu, T., Apps, J. A., & Pruess, K. (2005). Mineral sequestration of carbon dioxide in a
sandstoneshale system. Chem. Geology, 217 , 295 318.

17