**4. Thermodynamics of CO2 dissolution into formation brine**

It has been reported that 0.9-3.6 mol% of CO2 can be dissolved in brine, depending on pressure, temperature and brine composition (Rumpf et al. 1994, Koschel et al. 2006, Bando et al.

Dissolution Trapping of Carbon Dioxide in

scheme (d) is 5.13 x <sup>11</sup> 10<sup>−</sup> (Hirai et al. 1997)

scheme (c) drops further at increased CO2 pressure.

with carbonate host rock or carbonate based cements.

which may render CCS schemes ineffective.

schemes more economical.

al. 2003).

Reservoir Formation Brine – A Carbon Storage Mechanism 237

H2CO3. The dehydration of H2CO3 is also relatively slow, with a dehydration rate constant of kde = 18 s-1 (Pocker and Bjorkquist 1977). The deprotonation rate in scheme (c) is koff = <sup>7</sup> 10 s-1 (Pocker and Bjorkquist 1977) and the associated pKa value is 3.45. Note that this pKa value published by Adamczyk et al. (2009) is considerably different from the normally assumed pKa = 6.35 for the CO2(aq)/H2O system at atmospheric pressure conditions. The protonation rate of scheme (c) then results in kon = koff/Ka. The K value at 50.66 MPa for the reaction in

In addition, one consequence of an increase in CO2 solubility with increasing pressure (or decreasing temperature or salinity) is that the aqueous phase increasingly acidifies because more CO2 is present in the aqueous phase and reaction (b) is shifted to the right side according to Le-Chatelier's principle. It can therefore be expected that the pKa value of

In laboratory measurements pH values between 3.2-3.6 were observed within a temperature range between 300-343 K, a pressure range between 4-11 MPa and a salinity range 1-4 M NaCl solutions (Schaeff and McGrail 2004). In siliclastic and carbonate gas fields however pH values between 5-5.8 have been observed (Gilfillan et al. 2009); the discrepancy between lab and field data is most likely caused by complex geochemical buffering reactions, e.g.

The increased proton concentration in the brine has significant implications for geochemical reactions (Stumm and Morgan 1996, Gauss 2010) generally leading to more rock dissolution and higher dissolution rates. When the pH value has increased again to sufficiently high levels CO2 can be trapped as a solid phase – so-called mineral trapping (IPCC 2005, Gauss 2010). This could in principle also be engineered in the future although the physical and

• It is possible that too much rock is dissolved and high permeability channels are formed; this is especially a problem in carbonates (Egermann et al. 2005, Luquot and Gouze 2009). Injected CO2 will preferentially flow through such channels, which are also termed "wormholes". This reduces reservoir sweep efficiency which again decreases capillary trapping as only low initial CO2 saturations are achieved. Low initial CO2 saturations however result in low residual CO2 saturations (Pentland 2010 and 2011a,b; Al-Mansoori et al. 2010; Iglauer 2009). In addition such high permeability flow paths increase the risk of CO2 leakage, especially if caprock material is affected. • Should so much host rock be dissolved that the mechanical rock integrity is affected,

• In case of precipitation of solid minerals due to geochemical reactions (when the pH value has increased again) rock permeability can be significantly reduced, e.g. by blockage of small pore throats (which determine the permeability value). This can result in serious injectivity problems, e.g. injection rates may have to be reduced dramatically

• Rock dissolution increases permeability and enhances injectivity rendering CCS

• Precipitation of CO2 in solid minerals (after chemical reactions) is the safest form of CO2 storage in CCS as the CO2 cannot escape to the surface anymore. This trapping mechanism is believed to take between thousands to billions of years (IPCC 2005, Xu et

chemical phenomena associated with this process are highly complex and coupled.

then this can result in wellbore instability or even landmass subsidence.

Such reactions bring a range of problems and advantages with them:

Fig. 2. The advance of the fastest finger front is shown for different permeabilities (represented by different lines) (from Riaz et al. 2006 with permission from Cambridge University Press)

2003, Kiepe et al. 2002). Before analyzing these relationships in more depth, it should be pointed out that CO2 and brine are a reactive system, CO2 reacts with water to form carbonic acid which subsequently dissociates (scheme 1) through a proton-relay mechanism that is catalyzed by several water molecules (Adamczyk et al. 2009) lowering the pH value of the brine.


Scheme 1. Formation and dissociation of carbonic acid. Reaction scheme (a) assumes that scCO2 is dissolved in an analogous way to gaseous CO2 (Adamczyk et al. 2009)

Adamczyk et al. (2009) studied these reactions at atmospheric pressure and found that the slowest step in scheme 1 is the forward reaction of (b), the hydration of CO2(aq) resulting in

Fig. 2. The advance of the fastest finger front is shown for different permeabilities (represented by different lines) (from Riaz et al. 2006 with permission from Cambridge

2003, Kiepe et al. 2002). Before analyzing these relationships in more depth, it should be pointed out that CO2 and brine are a reactive system, CO2 reacts with water to form carbonic acid which subsequently dissociates (scheme 1) through a proton-relay mechanism that is catalyzed by several water molecules (Adamczyk et al. 2009) lowering the pH value of the

CO2(sc) + 3H2O U CO2(aq) + 3H2O (a)

H2O + CO2(aq) U H2CO3 (b)

Scheme 1. Formation and dissociation of carbonic acid. Reaction scheme (a) assumes that

Adamczyk et al. (2009) studied these reactions at atmospheric pressure and found that the slowest step in scheme 1 is the forward reaction of (b), the hydration of CO2(aq) resulting in

scCO2 is dissolved in an analogous way to gaseous CO2 (Adamczyk et al. 2009)

<sup>−</sup> + H<sup>+</sup> (c)

<sup>−</sup> + H<sup>+</sup> (d)

H2CO3 U HCO3

<sup>−</sup> <sup>U</sup> <sup>2</sup> CO3

HCO3

University Press)

brine.

H2CO3. The dehydration of H2CO3 is also relatively slow, with a dehydration rate constant of kde = 18 s-1 (Pocker and Bjorkquist 1977). The deprotonation rate in scheme (c) is koff = <sup>7</sup> 10 s-1 (Pocker and Bjorkquist 1977) and the associated pKa value is 3.45. Note that this pKa value published by Adamczyk et al. (2009) is considerably different from the normally assumed pKa = 6.35 for the CO2(aq)/H2O system at atmospheric pressure conditions. The protonation rate of scheme (c) then results in kon = koff/Ka. The K value at 50.66 MPa for the reaction in scheme (d) is 5.13 x <sup>11</sup> 10<sup>−</sup> (Hirai et al. 1997)

In addition, one consequence of an increase in CO2 solubility with increasing pressure (or decreasing temperature or salinity) is that the aqueous phase increasingly acidifies because more CO2 is present in the aqueous phase and reaction (b) is shifted to the right side according to Le-Chatelier's principle. It can therefore be expected that the pKa value of scheme (c) drops further at increased CO2 pressure.

In laboratory measurements pH values between 3.2-3.6 were observed within a temperature range between 300-343 K, a pressure range between 4-11 MPa and a salinity range 1-4 M NaCl solutions (Schaeff and McGrail 2004). In siliclastic and carbonate gas fields however pH values between 5-5.8 have been observed (Gilfillan et al. 2009); the discrepancy between lab and field data is most likely caused by complex geochemical buffering reactions, e.g. with carbonate host rock or carbonate based cements.

The increased proton concentration in the brine has significant implications for geochemical reactions (Stumm and Morgan 1996, Gauss 2010) generally leading to more rock dissolution and higher dissolution rates. When the pH value has increased again to sufficiently high levels CO2 can be trapped as a solid phase – so-called mineral trapping (IPCC 2005, Gauss 2010). This could in principle also be engineered in the future although the physical and chemical phenomena associated with this process are highly complex and coupled.

Such reactions bring a range of problems and advantages with them:


Dissolution Trapping of Carbon Dioxide in

0

experimentally measured values

experimentally determined values.

0.2

CO2

solubility [mol CO

2/kg brine]

0.4

0.6

0.8

1

1.2

1.4

**4.2 Effect of temperature on CO2 solubility in brine** 

the pressure to 10 MPa and salinity to 1 mole NaCl/kg.

0 50 100 150

temperature [°C]

**4.3 Effect of brine salinity on CO2 solubility in brines** 

Fig. 4. CO2 solubility versus temperature at high pressures. The substantially lower value measured by Rumpf et al. (1994) is caused by the high brine salinity (cp. section 4.3). Duan

CO2 solubility decreases with increasing salinity as show in Figure 5. The open diamonds show simulated data calculated with Duan et al.'s (2003+2006) CO2 solubility calculator setting the temperature to 323 K and the pressure to 10 MPa. It appears that Duan et al's. (2003+2006) model slightly over predicts CO2 solubilities. The other points shown are

Moreover, the type of dissolved salt has an influence on CO2 solubility. Yasunishi and Yoshda (1979) studied CO2 solubilites at atmospheric pressure in a wide variety of salt solutions, these salts included NaCl, KCl, Na2SO4, MgCl2, CaCl2, K2SO4, MgSO4, BaCl2, AlCl3, Al2(SO4)3 among others. They found that for the same electrolyte concentration, KCl solutions can absorb more CO2 than NaCl solutions, while CaCl2 and MgCl2 solutions absorb approximately the same amount of CO2. Monovalent NaCl or KCl solutions with the same salt concentration absorb more CO2 than their divalent CaCl2 or MgCl2 counterparts. For example Yasunishi and Yoshda (1979) measured at atmospheric pressure and 298 K that a 4.216 mol/L NaCl solution absorbs L = 0.3144 (L is the Ostwald coefficient, L = Vg/Vl with

et al.'s (2003+2006) data is simulation data (open diamonds); the other points are

Reservoir Formation Brine – A Carbon Storage Mechanism 239

CO2 solubility decreases with increasing temperature as shown in Figure 4. Experimental data relevant for CCS and simulated data are displayed. The computational data curve (open diamonds) was calculated with Duan et al. (2003+2006)'s solubility calculator setting

Nighswander et al. (1989),

Rumpf et al. (1994), 23.3 wt% NaCl Kiepe et al. (2002), 3 wt% NaCl

Bando et al. (2003), 3 wt% NaCl

Koschel et al. (2006), 5.8 wt% NaCl

0 wt% salt

Sabirzyanov et al. (2003),

1wt% NaCl Li et al. (2004), 1 wt% NaCl

In light of the new results published by Adamczyk et al. (2009) it is important to note that carbonic acid has a considerable acidity as it acts like a carboxylic acid on nanosecond timescales; this may have significant implications for geochemical reactions, rock surface alterations and associated possible rock wettability changes. Rock wettability strongly influences multi-phase fluid dynamics and capillary trapping.

On an important side issue these chemical reactions also happen in the oceans when CO2 gas in the atmosphere dissolves in seawater thereby reducing its pH value. With the increasing CO2 concentration in the atmosphere (from 190 ppm in 1750 to 380 ppm in 2005, IPCC 2005) more carbonic acid is formed in the oceans and the seawater pH value decreases with possible massive effects on sea life, starting with the sensitive but all important sea plankton. Therefore disposing anthropogenic CO2 by dissolving it into the ocean seems to be a risky enterprise, as the pH value would drop further and locally reach substantially lower numbers.

#### **4.1 Effect of pressure on CO2 solubility in brine**

CO2 solubility (mole fraction of CO2 per mass unit of brine) in formation brine is a strong function of pressure as shown in Figure 3. The data curve (open diamonds) in Figure 3 was computed with Duan and Sun (2003) and Duan et al. (2006)'s online CO2 solubility calculator. The temperature was held constant at 323 K and brine salinity was 1 mol NaCl/kg brine. CO2 solubility rapidly increases when pressure is raised from 0.1 MPa to 10 MPa, then the increase flattens out although a slight solubility increase follows. Three experimentally measured points at CCS pressure conditions are also added to the graph.

Fig. 3. CO2 solubility increases with pressure increase. The data shown was computed with Duan and Sun (2003) and Duan et al. (2006)'s CO2 solubility calculator. The black squares show experimental data points measured by Nighswander et al. (1989), Li et al. (2004) and Kiepe et al. (2002)

In light of the new results published by Adamczyk et al. (2009) it is important to note that carbonic acid has a considerable acidity as it acts like a carboxylic acid on nanosecond timescales; this may have significant implications for geochemical reactions, rock surface alterations and associated possible rock wettability changes. Rock wettability strongly

On an important side issue these chemical reactions also happen in the oceans when CO2 gas in the atmosphere dissolves in seawater thereby reducing its pH value. With the increasing CO2 concentration in the atmosphere (from 190 ppm in 1750 to 380 ppm in 2005, IPCC 2005) more carbonic acid is formed in the oceans and the seawater pH value decreases with possible massive effects on sea life, starting with the sensitive but all important sea plankton. Therefore disposing anthropogenic CO2 by dissolving it into the ocean seems to be a risky enterprise, as the pH value would drop further and locally reach substantially

CO2 solubility (mole fraction of CO2 per mass unit of brine) in formation brine is a strong function of pressure as shown in Figure 3. The data curve (open diamonds) in Figure 3 was computed with Duan and Sun (2003) and Duan et al. (2006)'s online CO2 solubility calculator. The temperature was held constant at 323 K and brine salinity was 1 mol NaCl/kg brine. CO2 solubility rapidly increases when pressure is raised from 0.1 MPa to 10 MPa, then the increase flattens out although a slight solubility increase follows. Three experimentally measured points at CCS pressure conditions are also added to the graph.

0 5 10 15 20 25 30

pressure [MPa]

Fig. 3. CO2 solubility increases with pressure increase. The data shown was computed with Duan and Sun (2003) and Duan et al. (2006)'s CO2 solubility calculator. The black squares show experimental data points measured by Nighswander et al. (1989), Li et al. (2004) and

influences multi-phase fluid dynamics and capillary trapping.

**4.1 Effect of pressure on CO2 solubility in brine** 

lower numbers.

0

0.2

CO2

Kiepe et al. (2002)

solubility [mol CO2/kg brine]

0.4

0.6

0.8

1

1.2

1.4
