The Pennsylvania State University

The Graduate School

John and Willie Leone Family Department of Energy and Mineral Engineering

PHASE EQUILIBRIA IN CO2-BRINE SYSTEM FOR CO2 STORAGE

A Dissertation in

Energy and Mineral Engineering

by

Haining Zhao

© 2014 Haining Zhao

Submitted in Partial Fulfillment

of the Requirements
for the Degree of

Doctor of Philosophy

December 2014
The dissertation of Haining Zhao was reviewed and approved* by the following:

Serguei N. Lvov
Professor of Energy and Mineral Engineering
Professor of Material Science and Engineering
Director of Electrochemical Technologies Program
Dissertation Advisor
Chair of Committee

William D. Burgos
Professor of Civil & Environmental Engineering

Li Li
Assistant Professor of Petroleum and Natural Gas Engineering

John Yilin Wang

Assistant Professor of Petroleum and Natural Gas Engineering

Luis F. Ayala H.
Associate Professor of Petroleum and Natural Gas Engineering
Associate Department Head for Graduate Education

*
Signatures are on file in the Graduate School

ii
 ABSTRACT

The apparatus for measuring CO2 solubility at elevated temperatures and pressures were

developed in this study; new experimental CO2 solubility in the following brines were collected: (1).

Single-salt aqueous NaCl, CaCl2, Na2SO4, MgCl2, and KCl with ionic strength from 0.5 to 6 mol/kg; (2).

Synthetic mixed-salt (NaCl and CaCl2) brines with ionic strength 1.712 and 4.984 mol/kg; (3). Synthetic

mixed-salt (NaCl, CaCl2, Na2SO4, MgCl2, KCl, SrCl2, and NaBr) brines with ionic strength 1.712 and 4.984

mol/kg; (4). A natural Mt. Simon formation brine with ionic strength 1.815 mol/kg.

The experimental data are used to develop the proposed PSUCO2 model. The model is capable of

calculating mutual solubilities of CO2 and H2O for the system of CO2-salt-H2O containing NaCl, CaCl2,

Na2SO4, MgCl2, or KCl or a mixture of them. Comparisons against literature data reveal a clear

improvement of the proposed PSUCO2 model among the published models in predicting CO 2 solubility in

the aforementioned brines.

A comparison of modeling results with experimental values on the P-x diagram (Figure 2. 14)

revealed a pressure-bounded "transition zone" in which the CO2 solubility decreases to a minimum and

then increases as the temperature increases. CO2 solubility is not a monotonic function of temperature in

the transition zone but outside of that transition zone, the CO 2 solubility decreases or increases

monotonically in response to increased temperature. The similar phenomenon is also observed by plotting

CO2 solubility contours in P-T diagram (Figure 4. 9), where the path of the maximum gradient among the

CO2 solubility contours is defined to divide the P-T diagram into two regions: in Region I, the CO2

solubility in the aqueous phase decreases monotonically in response to increased temperature, but in region

II, the behavior of the CO2 solubility is the opposite of that in Region I as the temperature increases.

A web-based computation interface of the model PSUCO2 is developed by the author and can be

accessed via the link: http://www.carbonlab.org/psuco2/.

iii
 TABLE OF CONTENTS

LIST OF FIGURES ........................................................................................................................................vi

LIST OF TABLES ...................................................................................................................................... viii

PREFACE ......................................................................................................................................................xi

ACKNOWLEDGMENTS ............................................................................................................................ xii

CHAPTER 1 INTRODUCTION ...................................................................................................................1

CHAPTER 2 SOLUBILITY OF CO2 IN AQUEOUS NaCl SOLUTIONS...................................................4

Abstract .......................................................................................................................................................5

2.1 Introduction ...........................................................................................................................................6

2.2 Review of the experimental approaches................................................................................................ 8

2.3 Experimentation .................................................................................................................................. 10

2.3.1 Experimental P-T-x conditions .................................................................................................... 10

2.3.2 Chemicals .................................................................................................................................... 11

2.3.3 Apparatus ..................................................................................................................................... 11

2.3.4 Procedure ..................................................................................................................................... 12

2.3.5 Validation of the obtained data .................................................................................................... 14

2.3.6 Measured CO2 solubilities ........................................................................................................... 17

2.4. Thermodynamic model of the CO2-NaCl-H2O system ...................................................................... 21

2.4.1 Thermodynamic framework for vapor-liquid phase equilibrium ................................................. 21

2.4.2 Incorporating Pitzer's activity model ........................................................................................... 24

2.4.3. Improved mixing rule parameter for the CO 2-rich phase............................................................ 29

2.4.4 Results and discussion ................................................................................................................. 35

2.5 Conclusions ......................................................................................................................................... 52

CHAPTER 3 SOLUBILITY OF CO2 IN AQUEOUS SOLUTIONS OF CaCl2, MgCl2, Na2SO4 AND KCl

....................................................................................................................................................................... 53

Abstract ..................................................................................................................................................... 54

3.1 Introduction ......................................................................................................................................... 55

iv
 3.2 Experimentation .................................................................................................................................. 57

3.3 Theoretical essentials for PSUCO2 ..................................................................................................... 57

3.4 Results and discussion ........................................................................................................................ 64

3.4.1. Measured results ......................................................................................................................... 64

3.4.2. Modeling results ......................................................................................................................... 71

3.5 Conclusions ......................................................................................................................................... 86

CHAPTER 4 SOLUBILITY OF CO2 IN SYNTHETIC FORMATION BRINE ........................................ 87

Abstract ..................................................................................................................................................... 88

4.1 Introduction ......................................................................................................................................... 89

4.2 Materials and methods ........................................................................................................................ 91

4.2.1 Chemicals .................................................................................................................................... 91

4.2.2 Synthetic brine preparation. ......................................................................................................... 91

4.2.3 Apparatus and procedure ............................................................................................................. 93

4.3 Results and discussions ....................................................................................................................... 95

4.3.1 Experimental results .................................................................................................................... 95

4.3.2 Additivity rule of Setschenow Coefficients of the individual ions .............................................. 95

4.3.3 Comparison of model calculations against the experimental data ............................................... 99

CHAPTER 5 CONCLUISONS ................................................................................................................. 117

REFERENCES ............................................................................................................................................ 120

APPENDIXES ............................................................................................................................................ 134

Appendix A: Additional Model Equations.............................................................................................. 135

Appendix B: Pitzer Ion-Ion Interaction Model ....................................................................................... 138

Appendix C: Experimental Error Estimation .......................................................................................... 145

Appendix D: The PSUCO2 Web-Computational Interface .................................................................... 150

Appendix E. Original Experimental Data Records ................................................................................. 152

v
 LIST OF FIGURES

Figure 1. 1 CO2-brine system and energy production. ..............................................................................2

Figure 2. 1 Approaches for CO2 solubility measurement. ........................................................................9

Figure 2. 2 Selected experimental pressure and temperature conditions. ............................................... 10

Figure 2. 3 Schematic of Experimental apparatus .................................................................................. 12

Figure 2. 4 Comparison of experimental CO2 solubility to literature data.............................................. 19

Figure 2. 5 Comparison of experimental CO2 solubility with model calculations .................................. 20

Figure 2. 6 Henry's law constant............................................................................................................. 23

Figure 2. 7 Computational flow chart of the proposed PSUCO2 model................................................. 26

Figure 2. 8 The data fitting procedure .................................................................................................... 27

Figure 2. 9 Phase diagram for the CO2-H2O system up to 573 K and 2000 bar .................................... 35

Figure 2. 10 Phase diagram for the CO2-NaCl-H2O system up to 573 K and 2000 bar............................ 36

Figure 2. 11 Phase diagram for the CO2-H2O system from 288-373 K, up to 300 bar. ........................... 40

Figure 2. 12 Phase diagram for the CO2-H2O system at temperatures 373.15-473.15 K and pressures 1-

100 bar for the CO2-rich phase. ............................................................................................ 41

Figure 2. 13 Comparison between the SP2010 model and this model at 523-573 K................................ 41

Figure 2. 14 CO2 solubility transition zone .............................................................................................. 43

Figure 2. 15 Comparison of this model and the MZLL2013 model. ........................................................ 47

Figure 3. 1 Comparison of the experimental CO2 solubilities in various aqueous salt solutions at 298.15

K and 1.013 bar based on different concentration scales. .................................................... 68

Figure 3. 2 Comparison of the experimental CO2 solubilities in different aqueous salt solutions based

on different concentration scales at 323.15 K and 150 bar. .................................................. 69

Figure 3. 3 The average absolute deviation (AAD %) of the calculated CO2 solubility values in aqueous

CaCl2, Na2SO4, MgCl2 and KCl solutions by different models ............................................ 71

Figure 3. 4 Comparison of model calculated values against the experimental data ............................... 73

Figure 3. 5 Comparison of model calculations against the experimental CO2 solubility in aqueous

CaCl2 solutions ..................................................................................................................... 78

vi
Figure 3. 6 Comparison of the model calculations against the experimental CO2 solubility in aqueous

Na2SO4 solutions .................................................................................................................. 79

Figure 3. 7 Comparison of the model calculations against the experimental CO 2 solubility in aqueous

MgCl2 solutions. ................................................................................................................... 80

Figure 3. 8 Comparison of the model calculations against the experimental CO 2 solubility in aqueous

KCl solutions ........................................................................................................................ 82

Figure 3. 9 Comparison of model calculation of H2O solubility in the CO2-rich phase ......................... 84

Figure 4. 1 The schematic of the additivity rule of the Setschenow coefficients.................................... 97

Figure 4. 2 Comparison of model calculations with experimental data. ............................................... 101

Figure 4. 3 Comparisons of the modeling results against the experimental CO2 solubility data in both

synthetic reservoir brines and NaCl+CaCl2 brines ............................................................. 102

Figure 4. 4 Comparison of model calculations against the experimental data reported by Yasunishi et al.

(1979) at 1 atm and 298 K .................................................................................................. 104

Figure 4. 5 Comparisons of the model calculations (PSUCO2, SP2010 and OLI) against the

experimental CO2 solubility data in the Weyburn Formation brine .................................... 105

Figure 4. 6 Comparison of the model calculations against the experimental........................................ 108

Figure 4. 7 Comparison of PSUCO2 with published models up to 2000 bar at 523K.......................... 110

Figure 4. 8 The average absolute deviation (AAD %) of the calculated CO2 solubility in aqueous

mixed-salt solutions among different models compared to the experimental data ............. 110

Figure 4. 9 The CO2 solubility contours in pure water and synthetic Mt. Simon formation ............. brine

generated by the PSUCO2 model ....................................................................................... 116

Figure D. 1 User-interface of CO2-brine phase equilibria model. Part I: CO2 solubility in pure H2O and

in single-salt brine; Part II: CO2 solubility in mixed-salt brine. ......................................... 150

Figure D. 2 User-interface of pure CO2 EoS. ........................................................................................ 151

Figure D. 3 User-interface of brine density model. ............................................................................... 151

vii
 LIST OF TABLES

Table 2. 1 Comparison of measured CO2 solubility in the binary CO2-H2O system at 323.15 K and 100

bar with literature data and model calculations. ................................................................... 15

Table 2. 2 Experimental results of CO2 solubility in NaCl(aq) compared with literature data (collected

at the same P-T-x conditions) and model calculations. ........................................................ 16

Table 2. 3 Experimental CO2 solubility measurements in NaCl(aq) at 150 bar and temperatures from

323.15 to 423.15 K. The experimental data are compared with model calculations in terms

of the average absolute deviation (AAD %) between calculated and experimental results. . 18

Table 2. 4 The results vCO2 (cm3 mol-1) determined by this model. ....................................................... 22

Table 2. 5 Parameters of Henry's constant Eq. (2.3). ............................................................................ 22

Table 2. 6 Parameters of Eq. (2.13). ...................................................................................................... 27

Table 2. 7 Pitzer triple-ion interaction parameter .................................................................................. 28

Table 2. 8 The mixing rule binary interaction parameter ...................................................................... 28

Table 2. 9 The components, key parameters, approaches and limitations of the CO 2 solubility model

presented in this study. ......................................................................................................... 32

Table 2. 10 The average absolute deviation (AAD %) of CO2 solubility in the aqueous phase between

model calculations and the selected reliable experimental data over a wide P-T-x range (25-

2000 bar, 288.15-573.15 K, and 0-6 mol kg-1 NaCl) ............................................................ 38

Table 2. 11 The average absolute deviation (AAD %) of the H2O solubility in the CO2-rich phase

between model calculations and the selected reliable experimental data at 1-2000 bar and

278.15-573.15 K .................................................................................................................. 39

Table 2. 12 The average absolute deviation (AAD %) of calculated CO2 solubility (mol kg-1) in aqueous

phase from experimental data for the CO 2-H2O system at temperatures of 273.15-573.15 K

and pressures of 1-2000 bar: Comparison between this model and the MZLL2013 model. 49

Table 2. 13 The average absolute deviation (AAD %) of the calculated CO2 solubility (molality) in the

aqueous phase from experimental data for the CO2-NaCl-H2O system at 273.15-573.15 K,

viii
 1-1400 bar and 0-6 mol kg-1 NaCl: Comparison between this model and the MZLL2013

model. ................................................................................................................................... 51

Table 3. 1 Coefficients of Eqs. (3.15) and (3.16). ................................................................................. 62

Table 3. 2 The results of concentration-dependent of the Pitzer triple-ion interaction parameter ........ 63

Table 3. 3 Experimental results for the CO2 solubility in aqueous CaCl2, Na2SO4, MgCl2 and KCl

solutions at 323.15-423.15 K and 150 bar. ........................................................................... 64

Table 3. 4 Comparison of the model calculations with experimental data. ........................................... 74

Table 4. 1 Chemical composition of the natural Mt. Simon brine. ........................................................ 92

Table 4. 2 Chemical compositions of the synthetic Mt.Simon brine and Antrim Shale formation brines

and the corresponding synthetic NaCl+CaCl2 proxy brines. ................................................ 92

Table 4. 3 Experimental CO2 solubility results in the synthetic formation and NaCl+CaCl2 brines. .... 94

Table 4. 4 Brine density (g cm-3) correlation (Eq. (4.7)) at 298K and 1 bar. ......................................... 98

Table 4. 5 The parameters of Eq. (4.8) .................................................................................................. 99

Table 4. 6 The average absolute deviation of the PSUCO2 model calculated CO 2 solubility in

NaCl+CaCl2 and NaCl brines from the experimental CO2 solubility data in synthetic Mt.

Simon and Antrim Shale formation brines. ....................................................................... 100

Table 4. 7 The absolute average deviations (AAD %) of model calculations from the experimental

data. .................................................................................................................................... 111

Table 4. 8 The components, key parameters, and the limitations of PSUCO2. ................................... 112

Table 5. 1 P-T-x matrix of CO2 solubility study for single-salt system. .............................................. 119

Table 5. 2 P-T-x matrix of CO2 solubility study for mixed-salt system. ............................................. 119

Table A 1 Parameters for Eqs. (A.10-A.11). ....................................................................................... 137

Table B 1 Constants of Eqs. (B.1) – (B.4) .......................................................................................... 141

Table B 2 Constants in Eqs. (B.5) – (B.6). ......................................................................................... 142

Table B 3 Constants of Eqs. (B.7) – (B.10). ....................................................................................... 143

Table B 4 Constants of Eq. (B.12) ...................................................................................................... 144

Table E 1 Original Experimental CO2 solubility data record for the CO2-H2O system. ..................... 153

Table E 2 Original Experimental CO2 solubility data record for the CO2-NaCl-H2O system. ........... 154

ix
Table E 3 Original Experimental CO2 solubility data record for the CO2-CaCl2-H2O system. .......... 157

Table E 4 Original Experimental CO2 solubility data record for the CO2-Na2SO4-H2O system. ....... 159

Table E 5 Original Experimental CO2 solubility data record for the CO2-MgCl2-H2O system. ......... 162

Table E 6 Original Experimental CO2 solubility data record for the CO2-KCl-H2O system. ............. 164

Table E 7 Original Experimental CO2 solubility data record for the CO2-synthetic formation brine

system ................................................................................................................................ 165

Table E 8 Original Experimental CO2 solubility data record for the CO2-synthetic NaCl+CaCl2 brine

system ................................................................................................................................. 167

x
 PREFACE

This dissertation consists of the three peer-reviewed research papers either published or submitted.

Haining Zhao is the first author on all three papers presented herein. All the experimental CO2 solubility

data reported in this dissertation except the data listed in Table 4. 3 Method 2 (4 data points) were collected

by the author, Haining Zhao. The experimental CO2 solubility data listed in Table 4.3 Method 2 were the

contribution made by Robert Dilmore, Douglas E. Allen, Sheila W. Hedges, and Yee Soong. The author

(Haining Zhao) developed the proposed PSUCO2 model by improving the previous published models of

Spycher et al. (2003), Spycher and Pruess (2005, 2010) and Akinfiev and Diamond (2010). All the figures

in this dissertation are original art work made by the author.

xi
 ACKNOWLEDGMENTS

There are so many people I would like to thank for their contribution to this work. First of all, I

would like to express my deepest gratitude to my advisor, Dr. Serguei N. Lvov, for supporting and

providing me with the resources to pursue this meaningful research over the past four years, and for helping

and guiding me throughout the entire research process: experiment design, data analysis, model

development, and publications. I cannot list all the details here, but a few cases deserve this page.

In the year 2011, I had struggled for about half a year to get the experimental apparatus to work,

and after that I almost wasted 3 months and got incorrect measurements. I knew something was wrong but I

couldn’t figure it out at that moment. I told you (Dr. Lvov) that the time and money were wasted and the

obtained data were useless. Your generosity and encouragement helped me to overcome the frustrating

feelings and find a correct solution; in the year 2012, I was experiencing a hard time to making progress for

the model development, so you asked Dr. Nikolay N. Akinfiev for his computer code in order to help me

get out of trouble. Without your help and guidance, the development of the PSUCO2 model wouldn’t have

been accomplished. In the year 2013, you encouraged me to publish our research work to a prestigious

journal and we finally got there.

We experienced many difficulties and unexpected situations at each step towards making reliable

measurements and high quality modeling. At the moment I’m going to leave your research group, to leave

my forever home EME at Penn State as well, when I look back over the past four years, I realize what huge

help and support you have provided for me and for my family. Without your dedication and constructive

instruction/guidance, as well as the support from the EME department, I would not have the opportunity to

do quality research work here. Additionally, without encouragement and financial support from you, the

on-line computation tool of the proposed PSUCO2 model would not have been possible.

Secondly, I would like express my sincere gratitude to the rest of my dissertation committee

members:

 Dr. William D. Burgos (Professor in Charge of Graduate Programs, Professor of Civil &

Environmental Engineering, Penn State),

 Dr. Li Li (Assistant Professor of Petroleum and Natural Gas Engineering, Penn State)

xii
  Dr. John Yilin Wang (Assistant Professor of Petroleum and Natural Gas Engineering, Penn

State)

 Dr. Luis F. Ayala H. (Associate Department Head for Graduate Education, Associate

Professor of Petroleum and Natural Gas Engineering)

for their friendly guidance and thought-provoking suggestions that each of them offered to me

over the years.

This work was jointly-supported by the U.S. Department of Energy, National Energy Technology

Laboratory, the EMS Energy Institute, and the John and Willie Leone Family Department of Energy and

Mineral Engineering at the Pennsylvania State University. With the help of my advisor, Dr. Serguei N.

Lvov, several well-known experts reviewed parts of this dissertation (Chapters 2 to 4) and provided many

constructive suggestions/recommendations, which significantly improved the overall quality of my work.

Thirdly, I would like to express my sincere gratitude to the following experts:

 Dr. Robert Dilmore (Research Engineer, Geosciences Division, National Energy

Technology Laboratory, U.S Department of Energy), for his time and patience to read through

the manuscript of Chapters 2 to 4 carefully and provide useful discussions, many constructive

suggestions, and detailed scientific and editorial corrections;

 Dr. Alfonso Mucci (Professor, Department of Earth and Planetary Sciences, McGill

University), for his time and patience to read through the manuscript of Chapter 2 carefully,

and provided detailed scientific and editorial corrections/recommendations and comments;

 Dr. Nikolay N. Akinfiev (Professor, Russian Academy of Sciences), for his time and patience

to read through the manuscript of Chapter 2 carefully and provide many constructive

suggestions; for his insight to theoretically pointed out a few weakness of Chapter 2; I

particularly thank him for sharing the source code of the AD2010 (Akinfiev and Diamond,

2010) model, which gives me a great help during the development of the PSUCO2 model;

 Dr. Nicolas Spycher (Geological Scientist, Lawrence Berkeley National Laboratory), for his

time and patience to read through the manuscript of Chapter 2 carefully and provided many

constructive suggestions. I particularly thank him for allowing me use his computer code (the

SP2010 model) for the model comparison throughout the dissertation;

xiii
 Dr. Larryn W. Diamond (Professor for Geochemistry and Petrology, University of Bern),

for his time and patience to read through the manuscript of Chapter 2 carefully, and for his

suggestion to compare the PSUCO2 model with a recent published model MZLL2013 (Mao

et al., 2013);

 One anonymous reviewer, for his time and patience to read through the manuscript of Chapter

2 carefully and provided many constructive suggestions;

 Dr. Sheila W. Hedges (National Energy Technology Laboratory, U.S Department of

Energy), Dr. Yee Soong (Research Group Leader, National Energy Technology Laboratory,

U.S Department of Energy), and Douglas E. Allen (Professor, Department of Geological

Sciences, Salem State University), for their time and efforts to read through the manuscript of

Chapter 4 carefully, and provided detailed scientific and editorial corrections and comments;

 Dr. Andre Anderko (Managing Director and Chief Technology Officer, OLI System Inc.)

and OLI System Inc., for provided the powerful OLI Studio 9.0.6 software package to EMS

Energy Institute. I also thank Dr. Andre Anderko for his time and patience to read through the

manuscript of Chapter 2 and provided many useful suggestions;

 Mr. Derek Hall (PhD Candidate, Penn State), for useful discussions and proof-reading the

part of Chapter 2.

In addition, I am sincerely grateful to:

 Dr. Luis F. Ayala H

 Dr. Turgay Ertekin (Head, John and Willie Leone Family Department of Energy and

Mineral Engineering Professor of Petroleum and Natural Gas Engineering, Penn State)

 Dr. Zulima Karpyn (Associate Professor of Petroleum and Natural Gas Engineering, Penn

State)

 Antonio Nieto (Associate Professor of Mining Engineering)

 Dr. Zi-Kui Liu (Director, Center for Computational Materials Design, Professor of Materials

Science and Engineering, Penn State)

 Dr. Serguei N. Lvov (Director, EMS Electrochemical Technologies Program, Professor of

Energy and Mineral Engineering and Materials Science and Engineering)

xiv
  Mr. Pichit Vardcharragosad (EME Teaching Asisstant, Ph.D. Candidate)

for the highest quality classes I have ever attended. From these wonderful teachers I have learned

valuable knowledge in the classes of reservoir simulation, hydrocarbon thermodynamics, gas reservoir

engineering, theory of flow through porous media, materials thermodynamics, geostatistics methods,

electrochemical energy conversion and theory of corrosion. All of this knowledge brought me to a broad

view of energy and mineral engineering, and have prepared me continue to explore in the area of petroleum

engineering or other energy-related areas with confidence. Thanks again to all the great professors in the

Department of Energy and Mineral Engineering.

I would also like to thank Dr. Mark Fedkin (Research Associate, Penn State), Dr. Justin Beck

(Postdoctoral Fellow, Penn State), and Derek Hall for their help in designing experimental apparatus and

constructing the experimental system.

Thanks to all my friends for the share of memorable occasions during my stay at Penn State!

Finally, I wish to express my deepest gratitude to my wife, my lovely son, and my parents, for

their love, supports and encouragements throughout my life.

xv
 CHAPTER 1

INTRODUCTION

1
 CO2-brine system are the most naturally encountered fluid system in geological formations. CO2

and brine are considered immiscible fluids but they are mutually soluble. The dissolution of CO2 in

formation brine can cause the changes of brine density (usually increase), wettability, saturation, and pH.

The brine pH change due to the dissolution of CO2 will influence the reaction rate occurred at the brine-

rock interface. In addition, the solubility of H2O in the supercritical CO2-rich phase could be large at the

formation temperatures higher than 373 K. The formation dry-out (Pruess and Müller, 2009) and solids

precipitation from CO2 injection into subsurface reservoir, which changes the formation porosity and

permeability, could occur when a large amount CO2 is injected into a high temperature and high salinity

formation. Therefore, the mass transfer between the CO 2-rich and the aqueous phases will change the fluid

and rock properties for the involved formation. Knowledge of the mass transfer between the two fluid

phases (mutual solubilities) at geological temperatures and pressures is essential for us to model the various

geochemical and industrial processes for energy recovery (Figure 1. 1).

Rig

CO2-
Oil & Gas Enhanced
CO2 Geothermal
Production injection
Water & CO2 Energy
injection Production

Oil & Gas Reservoir

Saline Aquifier

Natural CO2 Reservoir

Figure 1. 1 CO2-brine system existed in several of engineered-natural systems, including oil and gas
reservoir, saline aquifer, geothermal reservoir and natural CO2 reservoir.

2
 This study experimental measured CO2 solubility in NaCl, CaCl2, Na2SO4, MgCl2, KCl and

mixed-salt brines at 100-200 bar, 323-423 K and ionic strength up to 6 mol kg-1. The experimental data, as

well as the data collected from literature, are used for model development. The proposed PSUCO2 model is

able to overcome the deficiencies of the previously published models for their limitations in P-T-x range,

accuracy, and salt species.

While the studies of the CO2 solubility in aqueous NaCl solutions are abundant, in Chapter 2, this

study added new experimental data for the CO2-NaCl-H2O system at 150 bar, 323-423 K and 0-6 mol/kg

NaCl. By using all available experimental data in both the aqueous and the CO 2-rich phases for the CO2-

NaCl-H2O, a 𝜑 − 𝛾 type thermodynamic model is developed to accurately compute the mutual solubilities

for the CO2-NaCl-H2O system at 1-2000 bar, 273-573 K, and NaCl concentration up to saturation condition.

The detail experimental methods and theoretical framework for the proposed thermodynamic model can be

found in Chapter 2.

In Chapter 3, the experimental CO2 solubility data in aqueous CaCl2, Na2SO4, MgCl2, KCl

solutions are given at 150 bar, 323.15-423.15 K, and total ionic strength up to 6 mol/kg. Using the newly

obtained experimental data, the thermodynamic model developed in Chapter 2 for the CO2-NaCl2-H2O

system are then extended to include the influence of different salt species on dissolved CO 2 in the aqueous

phase, but the model still not be able to calculate CO2 solubility in mixed-salt solution or natural formation

brine.

In Chapter 4, based on the experiences and results gained from Chapters 2 and 3, in order to make

the model computable for CO2 solubility in natural formation brines, the additivity rule of the Setschenow

coefficients of the individual ions (Na+, Ca2+, Mg2+, K+, Cl-, and SO42-) is successfully used to calculate

CO2 solubility in natural formation brine at elevated temperatures and pressures. The modeling results are

validated by newly obtained experimental CO2 solubility data in natural and synthetic formation brines, and

in synthetic NaCl+CaCl2 brines.

Chapter 5 summarizes the principle findings as a conclusion of this dissertation.

3
 CHAPTER 2

SOLUBILITY OF CO2 IN AQUEOUS NaCl SOLUTIONS1

1
The text for this chapter was originally prepared for the publication as "Haining Zhao, Mark V. Fedkin,
Robert M. Dilmore, and Serguei N. Lvov (2014) Carbon dioxide solubility in aqueous solutions of sodium
chloride at geological conditions: Experimental results at 323.15, 373.15, and 423.15 K and 150 bar and
modeling up to 573.15 K and 2000 bar, Geochimica et Cosmochimica Acta,
doi:10.1016/j.gca.2014.11.004".

4
Abstract

A new experimental system was designed to measure the solubility of CO2 at pressures and

temperatures (150 bar, 323.15-423.15 K) relevant to geologic CO2 sequestration. At 150 bar, new CO2

solubility data in the aqueous phase were obtained at 323.15, 373.15, and 423.15 K from 0 to 6 mol kg -1

NaCl(aq) for the CO2-NaCl-H2O system. A γ − φ (activity coefficient — fugacity coefficient) type

thermodynamic model is presented for the calculation of both the solubility of CO 2 in the aqueous phase

and the solubility of H2O in the CO2-rich phase for the CO2-NaCl-H2O system. Validation of the model

calculations against literature data and other models (MZLL2013, AD2010, SP2010, DS2006, and OLI)

show that the proposed model is capable of predicting the solubility of CO2 in the aqueous phase for the

CO2-H2O and CO2-NaCl-H2O systems with a high degree of accuracy (AAD < 3.9%) at temperatures from

273.15 to 573.15 K and pressures up to 2000 bar. A comparison of modeling results with experimental

values revealed a pressure-bounded "transition zone" in which the CO2 solubility decreases to a minimum

then increases as the temperature increases. CO2 solubility is not a monotonic function of temperature in

the transition zone but outside of that transition zone, the CO2 solubility is decrease or increase

monotonically in response to increased temperature. A web-based CO2 solubility computational tool can be

accessed via the link: www.carbonlab.org/psuco2/.

5
2.1 Introduction

The CO2-H2O and the CO2-NaCl-H2O fluid systems play an important role in various earth system

and engineered geologic processes (Schmidt and Bodnar, 2000; Dubacq et al., 2013). They represent the

most commonly encountered fluids in geological environments (Hu et al., 2007). In many technical

applications, such as CO2 storage and sequestration, CO2-enhanced oil recovery, and CO2-enhanced

geothermal systems (Spycher and Pruess, 2010), dissolution, precipitation, and ion-exchange processes

associated with these fluid systems can potentially impact the behavior of geologic media. Therefore, our

knowledge of the phase equilibria and properties of these two systems over a wide pressure-temperature-

composition (P-T-x) range is essential in the development of models for various geochemical and industrial

processes.

Numerous studies have contributed to phase equilibrium data for CO2-bearing binary and ternary

systems. The main goal of many of those works has been to measure the CO 2 solubility in the aqueous

phase and the water solubility in the CO2-rich phase (e.g. Tödheide and Franck, 1963). In this study,

research was oriented to CO2 solubility experiments in support of geological CO2 storage applications. The

solubility of CO2 in formation brines is a critical parameter to estimate the CO 2 storage capacity, the

mechanism of trapping over site life, and the potential reactivity of the fluid-rock interface, as well as to

understand the fate of the dissolved CO2 in subsurface formations. Previous data (reported in this work)

from experimental CO2 solubility measurements in aqueous electrolyte solutions, along with reliable phase

equilibrium models can serve as the foundation for modeling a variety of CO 2 storage and utilization

processes.

Hu et al. (2007) have pointed out that many cubic equations of state (EoS) and virial EoS

truncated at the second or third virial coefficient are adequate to predict the phase equilibria properties for

the CO2-H2O and CO2-NaCl-H2O systems at low to medium pressures (up to 500 bar), but the deviations of

these models from experimental values at high pressures were found to be unacceptable (Hu et al., 2007;

Spycher et al., 2003). Spycher and Pruess (2010) model allows computation of both the CO2 solubility in

the aqueous phase and the H2O solubility in the CO2-rich phase with high accuracy, but their model is

limited to pressures up to 600 bar. Akinfiev and Diamond (2010) developed a highly accurate CO2

6
solubility model for the CO2-H2O and CO2-NaCl-H2O systems, but the applicable temperatures and

pressures of their model are limited to 251.15-373.15 K and 1-1000 bar. In addition, Akinfiev and Diamond

(2010) model only provides a rough estimate of water solubility in the CO 2-rich phase. The model

developed by Duan et al. (2006) allows the CO2 solubility in NaCl(aq) to be calculated up to 533.15 K,

2000 bar, and 4.5 mol kg-1 of NaCl(aq), but their model is generally less accurate than the Akinfiev and

Diamond (2010) and Spycher and Pruess (2010) models (see section 2.4.4, Table 2. 10). Moreover, Duan et

al.'s (2006) model does not allow calculation of the H2O solubility in the CO2-rich phase. The commercial

software OLI studio 9.0.6 (Springer et al., 2012) performs well at predicting the H2O solubility in the CO2-

rich phase at low to moderate pressures, but its estimates of the CO 2 solubility in the aqueous phase are less

accurate than those of other models mentioned above - Akinfiev and Diamond (2010), Spycher and Pruess

(2010) and Duan et al. (2006) (see section 2.4.4, Table 2. 10 and Table 2. 11). Additionally, Mao et al.

(2013) developed a CO2 solubility model for the CO2-NaCl-H2O system in the temperature range of

273.15-723.15 K, pressures from 1-1500 bar and NaCl molality from 0-4.5 mol kg-1. Mao et al.'s (2013)

model is in excellent agreement with experimental CO2 solubility in the aqueous phase (see section 4.4,

Table 2. 10, Table 2. 12 and Table 2. 13), but this model does not allow calculation of H2O solubility in the

CO2-rich phase. Note also that these models cannot be used for reliable phase equilibrium calculations in

systems containing K+, SO2−

4 (e.g. CO2-KCl-H2O and CO2-Na2SO4-H2O) at elevated temperatures and

pressures. Detailed comparisons between the model developed in this study and previously published

models are provided in section 2.4.4.

In this Chapter, an improved 𝛾 − 𝜑 (activity coefficient - fugacity coefficient) type phase

equilibrium model is presented for the CO2-NaCl-H2O system. There are three major improvements in the

model relative to the previous models of Spycher et al. (2003) and Spycher and Pruess (2010). First, the

binary interaction parameter (k CO2−H2 O ) of the mixing rule was considered to be dependent on temperature

and pressure. The parameter k CO2−H2O in the mixing rule was carefully tuned against the experimental H 2O

solubility in the CO2-rich phase (Tödheide and Franck, 1963) at temperatures of 273.15-573.15 K and

pressures up to 2000 bar. This adjustment significantly improved the calculated water solubility in the CO2-

rich phase, as well as estimates of the CO2 solubility in the aqueous phase at high temperatures and

pressures. Second, in Eq. (2.1), the partial molar volume of CO2 at infinite dilution in water (v̅CO2 ) was

7
treated as an adjustable parameter to fit the experimental CO2 solubility in the aqueous phase by Tödheide

and Franck (1963). Third, the Pitzer activity model was used to calculate the activity coefficient of the

dissolved CO2 and osmotic coefficient of water. The triple-ion interaction parameter (ξnca , see section 2.4.2

Eq. (2.10)) in the Pitzer model was considered to be dependent on temperature and NaCl concentration.

ξnca values were determined using new experimental CO2 solubility data presented herein, assuming that

the pressure effect on this parameter can be neglected. These improvements allow the model to return more

accurate estimates of the activity of dissolved CO2 and the water osmotic coefficients for concentrated

NaCl(aq) as compared to the previous models of Spycher and Pruess (2010) and Duan et al. (2006).

The improved CO2 solubility model can reproduce literature data for the CO2-NaCl-H2O and CO2-

H2O systems with an overall average absolute deviation (AAD) of less than 3.9% over temperatures from

273.15 to 573.15 K and pressures up to 2000 bar. In Section 4.4, we compare our model with the models of

Mao et al. (2013), Akinfiev and Diamond (2010), Spycher and Pruess (2010), Duan et al. (2006) and OLI

Studio 9.0. This comparison shows that the model developed here provides reliable estimates of both the

solubility of CO2 in the aqueous phase and the solubility of H2O in the CO2-rich phase.

2.2 Review of the experimental approaches

In this and many previous studies, a high pressure/temperature-controlled stainless steel autoclave

with agitation was used to investigate the gas-liquid phase equilibria. The typical apparatus for such

measurements is described in Prutton and Savage (1945), Takenouchi and Kennedy (1964), Malinin and

Savelyeva (1972). The experimental systems used by Wiebe and Gaddy (1939), Nighswander et al. (1989),

and Stewart and Munjal (1970) are more complicated, but, in general, operate in a similar manner. For a

phase equilibrium experimental study, the autoclave is loaded with a CO 2-brine mixture and the solution is

stirred for 6 to 24 hours until equilibrium is reached. The autoclave temperature is measured by a

thermocouple, while the pressure is measured using a pressure gauge or transducer. Generally, the aqueous

phase sample is taken from the autoclave to perform the CO 2 solubility measurement. During the sampling

process, the pressure in the system should remain constant. Alternatively, static view-cell 'synthetic'

8
methods without sampling are employed to determine the CO 2 solubility (Rumpf and Maurer, 1993; Pérez-

Salado Kamps et al., 2007) and the dew-point of H2O in the CO2-rich phase (Mather and Franck, 1992).

There are two common methods to measure the CO2 solubility in a liquid sample. In the first

method, dissolved CO2 in a sample cell is expanded into a temperature-controlled pressure vessel, and the

mass of CO2 is determined volumetrically (Wiebe and Gaddy, 1939; Prutton and Savage, 1945;

Nighswander et al., 1989; Liu et al., 2011). In the second method, dissolved CO2 in a sample cell is released

from the aqueous phase and absorbed in a trap with NaOH(aq) or Ba(OH) 2(aq) (Takenouchi and Kennedy,

1964, 1965; Stewart and Munjal, 1970; Malinin and Savelyeva, 1972; Rochelle and Moore, 2002). In

Method (1), the amount of CO2 is determined from the PVT changes in the pressure vessel before and after

the gas expansion. This method was used by Kim et al. (2008) to measure the mass of the injected CO2 into

the autoclave. Liu et al., (2011) also demonstrated the effectiveness of this method. The method can

precisely sense the mass change within a pressure cell through the variation in pressure and temperature.

Therefore, this method was selected for sample analysis in this study. In Method (2), the dissolved CO 2 is

converted to CO32- in an alkaline absorbent. As long as the base is present in excess, the CO 32- is stable

(Rochelle and Moore, 2002). The amount of trapped CO2 can be determined by a standard HCl titration of

the excess base absorbent.

Experimental system for CO2- Sample analysis for determining the CO 2

brine phase equilibria solubility in aqueous phase
CO2 expands at
constant pressure
CO2 expansion
CO2 expands at
constant volume

CO2 absorbed by
Sampling Standard HCl titration
base absorbent

Gas phase and Gas chromatography

Stainless
Steel aqueous phase
circulation Ultra-low temperature
autoclave water trap

1. Optical cell is needed to observe the phase boundary

2. The exact amount of H 2O and CO2 loaded into system
No
should be known
sampling 3. The exact volume of CO2 phase is needed for the CO 2
solubility calculations

Figure 2. 1 Main approaches used to measure the CO2 solubility in an aqueous phase. The shaded boxes
represent the method implemented in this study.

9
 The main approaches to sample and measure the CO2 solubility in the aqueous phases are

summarized in Figure 2. 1. Other methods such as the solid absorbent system (Ellis and Golding, 1963) and

heat flow calorimetry (Koschel et al., 2006) are not widely used to measure the CO2 solubility in the

aqueous phase due to the complexity of the systems and insufficient accuracy of the measurements.

2.3 Experimentation

2.3.1 Experimental P-T-x conditions

In this work, data from all available literature for the CO 2-NaCl-H2O system at temperatures from

273.15 to 573.15 K and pressures up to 1500 bar were reviewed in order to identify possible gaps in the

experimental P-T-x matrix. A large amount of CO2 solubility data compiled by Scharlin (1996) was also

included in our data screening. New CO2 solubility measurements were performed in P-T-x space where

little or no data are available.

200
~Depth 1.8 km

150 Aqueous
phase
Supercritical
P / bar

region
~Depth 1.0 km
100
Critical point
Supercritical boundary

50
34.85 CO2-rich phase

0
273.15 323.15 373.15 423.15
T/K
Basins Selected experimental points Critical point
Supercritical boundary Depth indication

Figure 2. 2 The common envelope of CO2 phase behavior in sedimentary basins along with selected
experimental pressure and temperature conditions (150 bar, 323.15, 373.15 and 423.15 K) in this study
(shown by solid circles).

10
 A simplified approach proposed by Bachu (2003) was employed in this study to estimate a range

of temperatures and pressures of interest (Figure 2. 2). With respect to the salt concentration range, the

USGS-produced water database2 served to select the concentration of NaCl solutions. In addition, we took

into account the McIntosh et al. (2004) study where eighty five formation brine samples from the Michigan

Basin were collected and analyzed. The total ionic strength (Eq. (A9)) of formation brines in the Michigan

Basin was shown to range from zero to around 8 mol kg-1, with chloride, sodium, and calcium accounting

for more than 95 % of the ions in these brines (McIntosh et al. 2004); the composition of the reservoir

brines is dominated by NaCl. Based on the above discussion, the ionic strength of the NaCl(aq) solutions

for our CO2 solubility experiments was set to range from 0 to 6 mol kg-1.

2.3.2 Chemicals

The carbon dioxide used in all experiments was Coleman Instrument grade with a minimum purity

of 99.99 %. The water was purified by a Milli-Q system and degassed by boiling before loading into the

autoclave. The NaCl(aq) solutions were prepared using Milli-Q water (with conductivity below 610-6 S m-

1
) and ACS reagent grade Sodium Chloride (J.T. Baker, 3624-01).

2.3.3 Apparatus

A schematic of the experimental apparatus is shown in Figure 2. 3. The setup includes a high

pressure liquid-CO2 pump (SFT-10), a vacuum pump (LEYBOLD TRIVAC D2A), a 600-ml stainless steel

Parr Instrument Co. autoclave with a Teflon liner, and a 40-ml stainless steel liquid sample cell attached to

the autoclave. The autoclave was enclosed in a heating mantle, and the contained solution was stirred by a

four-blade aluminum impeller with a Parr magnetic drive. Temperature and the impeller rotation speed

2
http://energy.cr.usgs.gov/prov/prodwat/

11
were controlled by the Parr 4842 controller. The autoclave was equipped with ports for both supercritical

CO2 and aqueous phase sampling from the top of the vessel.

An Omega PX602 pressure transducer with an accuracy of 0.45 % ranging from 0 to 345 bar was

used to monitor the system pressure. A J-type thermocouple was introduced from the top of the autoclave

into the aqueous phase to measure the temperature of the solution. Calibration of the thermocouple was

performed in both boiling DI water and an ice-water mixture prior to use. The uncertainty of the

temperature measurements was less than ±1.0 K over the temperature range of interest (323.15 to 423.15

K).

V1 V2
1
2
V3 V4 V6

3 V5
V7

V8
4

5 9
8
6
11 10
7 12
15 13
14

Figure 2. 3 Schematic diagram of the experimental apparatus: (1) 0 to 345 bar pressure transducer, (2)
CO2-rich phase sample cell, (3) sample line for the CO2-rich phase, (4) rupture disc, (5) CO2-rich phase, (6)
thermocouple, (7) aqueous solution, (8) liquid phase sample cell, (9) liquid CO 2 pump, (10) liquid CO2
cylinder, (11) liquid CO2 inlet, (12) sample line for the aqueous phase, (13) impeller, (14) heat mantle, (15)
Teflon liner.

2.3.4 Procedure

Degassed Milli-Q water was used in all experiments. Sodium chloride was placed into the

autoclave before loading the water. The autoclave was thoroughly evacuated using a vacuum pump and

water was loaded into the autoclave under vacuum. The autoclave was then heated to the desired

12
temperature. Once the experimental temperature was achieved, the liquid CO2 was pumped into the

autoclave to the desired pressure. Subsequently, the contents of the autoclave were agitated by the impeller

at a speed of about 60 RPM (revolutions per minute). Initial equilibrium was assumed to be reached when

no cell pressure variation was observed over a 12-hour period. Before taking the first sample, a small

amount of liquid sample was drawn and discarded to purge the sample line.

Whereas either a water-rich or a CO2-rich sample can be taken using the developed system, this

study focused on the aqueous phase. The sample cell was first evacuated to a vacuum pressure of about

0.02 bar. At high pressures, withdrawing fluid from the autoclave can result in a significant reduction of the

total pressure of the system. Hence, the pressure drop during the sampling process was compensated for by

pumping additional liquid CO2 into the CO2-rich phase at a flow rate of about 20 ml min-1. Pressure

fluctuations inside the autoclave were kept to within ±2 bar of the set experimental pressure, and the

pumped liquid CO2 is assumed not to have disrupted the aqueous phase equilibrium due to the following

reasons. First, the liquid CO2 was injected in the upper layer of the CO2-rich phase, whereas the aqueous

sample was taken from the bottom of the autoclave (Figure 2. 3), so the pumped CO2 does not directly

contact the aqueous sample in the bottom of the autoclave. Second, the sampling time was very short (less

than 60 seconds) so there was no, or negligible, disturbance on the aqueous phase equilibrium.

After sample withdrawal was complete, the sample cell was disconnected and weighed on an

electronic balance (Veritas L403i) with an accuracy of ±0.001g. The mass of the sample was defined by

weighing the sample cell before and after sampling. The sample cell was then connected to a 300-ml

pressure cell, which was used to determine the amount of dissolved CO 2. The pressure cell volume was

calibrated by filling the cell with distilled water and measuring the mass when it was full and empty at a

well-defined temperature. The water density was calculated using IAPWS-95 EoS (Wagner and Pruß,

2002).

The pressure cell was first evacuated by a vacuum pump and the dissolved CO 2 was slowly

released into the pressure cell from the sample cell. The mass of the dissolved CO 2 was determined by the

pressure and temperature change before and after CO2 expansion. Three consecutive expansions were

sufficient to determine the total amount of CO2 dissolved in the sample cell. The remaining CO2 in the

sample cell after three expansions was estimated by Henry's law (see section 2.4.1, Eq. (2.3)) and was less

13
than 0.0007 mol kg-1, which is negligible compared to the experimental error. The EoS for pure CO2

developed by Span and Wagner (1996) was used to calculate the mass of CO2 in the pressure cell after each

expansion. Finally, the CO2 solubility in the aqueous phase was obtained by relating the mass of the total

released CO2 to the total mass of the aqueous sample.

A small amount of water vapor might be transferred to the pressure cell during the CO 2 expansion,

which could introduce a small error (approximately 0.0009 mol kg -1 CO2) in the CO2 solubility

measurement. To eliminate the error caused by water vapor during the sample analysis, the partial pressure

of water vapor in the CO2-H2O gas mixture was estimated by the model of Spycher et al. (2003) and

subtracted from the total pressure of the pressure cell. The error analysis based on the error propagation

theory showed that the instrumental error of the measurement of CO 2 solubility was around 0.7 % and the

random error was from 0.5 to 4.5 %. The detailed error analysis procedure is described in Appendix C.

2.3.5 Validation of the obtained data

A temperature and pressure of 323.15 K and 100 bar were selected for the CO 2-H2O system, and

several P-T-x conditions for the CO2-NaCl-H2O system, to validate the reliability of our apparatus and

procedures. In Table 2. 1 and Table 2. 2, the comparisons show that the experimental CO2 solubility at

these P-T-x conditions was in close agreement with the experimental solubility values found in the

literature, and was also consistent with CO2 solubility models (Mao et al., 2013; Akinfiev and Diamond,

2010; Spycher and Pruess, 2010; Duan et al., 2006; OLI Studio 9.0). Subsequently, the names of the

models are abbreviated as MZLL2013 for Mao et al. (2013), AD2010 for Akinfiev and Diamond (2010),

SP2010 for Spycher and Pruess, (2010), DS2006 for Duan et al. (2006) model, and OLI for the model

adopted in OLI Studio 9.0 software.

From Table 2. 1 and Table 2. 2, one can see that the measured CO2 solubilities are in good

agreement with both published experimental and model data. Thus, the reliability of the experimental

technique was confirmed before further measurements were performed at the target P-T-x conditions in the

CO2-NaCl-H2O system.

14
Table 2. 1 Comparison of measured CO2 solubility in the binary CO2-H2O system at 323.15 K and 100 bar
with literature data and model calculations.

T/K P / bar References mCO2 / mol kg-1

Experimental measurements
323.15 100 Bando et al. (2003) 1.116 ± 0.033
323.15 101 Liu et al. (2011) 1.133(a)
323.15 101 Dohrn et al. (1993) 1.176(a)
323.15 101.3 Briones et al. (1987) 1.180(a)
323.15 100.9 Bamberger et al. (2000) 1.162 ± 0.017
323.15 101.3 D’Souza et al. (1988) 1.121(a)
323.15 101.3 Wiebe and Gaddy (1939) 1.151(a)
323.15 100 This study 1.151 ± 0.028
Modeling(b)
323.15 100 DS2006 1.148
323.15 100 AD2010 1.156
323.15 100 SP2010 1.136
323.15 100 OLI Studio 9.0 1.146(c)
323.15 100 MZLL2013 1.147
323.15 100 This model 1.151
(a)
The experimental data without '±' mean that the error was not given in the references.
(b)
DS2006, AD2010, SP2010 and MZLL2013, respectively, stand for the model developed by Duan et al.
(2006), Akinfiev and Diamond (2010), Spycher and Pruess (2010) and Mao et al. (2013)
(b)
Calculations made by OLI are under the following conditions: (1) enable the second liquid phase, (2) use
MSE model, (3) stream inflows contains 55.5082 mol water and 5 mol CO 2.

15
Table 2. 2 Experimental results of CO2 solubility in NaCl(aq) compared with literature data (collected at the same P-T-x conditions) and model calculations.

T/K P/ 𝐦𝐍𝐚𝐂𝐥 / This study Literature References Model calculations(a), 𝐦𝐂𝐎𝟐 / mol kg-1
bar mol kg-1 𝐦𝐂𝐎𝟐 / 𝐦𝐂𝐎𝟐 / This OLI(b) AD201 SP2010 DS200 MZLL2
mol kg-1 mol kg-1 study 0 6 013
323.15 150 0.529 1.114±0.017 1.102 Bando et al. (2003) 1.108 1.150 1.117 1.108 1.095 1.123
373.15 48 0.998 0.361±0.006 0.361 Malinin and Kurovskaya (1975) 0.377 0.358 0.368 0.369 0.370 0.376
413.15 92.01 5.999 0.278±0.014 0.288 Rumpf et al. (1994) 0.276 0.259 0.280 0.292 - -
Average absolute deviation (AAD %) of model calculations from the experimental CO 2 1.88 3.63 1.03 2.59 2.05 2.48
solubility data
(a)
OLI: OLI Studio 9.0, which is a simulation software for electrolyte chemistry developed by OLI Systems Inc. (Springer et al., 2012); AD2010: Akinfiev and
Diamond (2010); SP2010: Spycher and Pruess (2010); DS2006: Duan et al. (2006).
(b)
Calculations made by OLI are under the following conditions: (1) enable the second liquid phase, (2) use MSE model, (3) stream inflows contains 55.5082 mol
water and 5 mol CO2

16
2.3.6 Measured CO2 solubilities

The measured CO2 solubilities (𝑚CO2 ) at 150 bar and 323.15-423.15 K for the CO2-NaCl-H2O

system at NaCl(aq) concentrations between 0 and 6 mol kg-1 are listed in Table 2. 3. Figure 2. 4 shows that

the measured CO2 solubility data fall closely in line with the data from Koschel et al. (2006), with the

model calculations of SP2010, and the model developed herein. In Figure 2. 5, the models AD2010,

SP2010, DS2006 and OLI were verified against new experimental data reported herein. The experimental

CO2 solubility data obtained in this study are in excellent agreement with both the MZLL2013 and AD2010

models at 150 bar and temperatures from 323.15 to 423.15 K. Given that the accuracy of the AD2010

model is claimed to be 1.6 % at temperatures from 251.15 to 373.15 K and pressures up to 1000 bar, it is

striking that deviations of the AD2010 model from experimental data at 423.15 K are still within its

computation error. The SP2010 model results also agree well with the measured CO 2 solubility data

obtained in this study. The deviation of the DS2006 model results from our experimental data increases

with increasing NaCl molality because the model is limited to a NaCl concentration of 4.5 mol kg -1. The

OLI studio 9.0 model also performs well in predicting CO 2 solubility in aqueous NaCl solutions. In general,

the models developed thus far agree well with the experimental CO 2 solubility obtained in this work. The

average absolute deviations of the calculated values from our CO 2 solubility data (given in Table 2. 3)

confirm this conclusion.

17
Table 2. 3 Experimental CO2 solubility measurements in NaCl(aq) at 150 bar and temperatures from 323.15 to 423.15 K. The experimental data are compared
with model calculations in terms of the average absolute deviation (AAD %) between calculated and experimental results.

Experimental CO2 solubility at 150 bar Calculated CO2 solubility: mCO2 / mol kg-1
(This study)
T/K 𝒎𝐍𝐚𝐂𝐥 / 𝒎𝐂𝐎𝟐 / 𝛅𝒎𝐫𝐚𝐧𝐝𝐨𝐦 / This study OLI(b) AD2010 SP2010 DS2006 MZLL2013
-1 -1(a)
mol kg mol kg -1 mol kg
323.15 0 1.245 0.006 1.244 1.271 1.256 1.233 1.223 1.253
323.15 1 1.016 0.010 1.016 1.055 1.017 1.012 0.995 1.024
323.15 2 0.859 0.007 0.853 0.889 0.849 0.845 0.822 0.853
323.15 3 0.734 0.015 0.727 0.761 0.724 0.716 0.690 0.723
323.15 4 0.623 0.006 0.629 0.661 0.629 0.618 0.587 0.625
323.15 5 0.551 0.005 0.552 0.582 0.553 0.541 - -
323.15 6 0.498 0.004 0.492 0.520 0.493 0.481 - -
AAD (%) at 323.15 K 0.59 4.17 0.80 1.66 3.96 0.79
373.15 0 1.020 0.006 1.012 1.020 1.010 1.001 0.984 1.032
373.15 1 0.836 0.008 0.842 0.839 0.827 0.819 0.808 0.843
373.15 2 0.706 0.007 0.714 0.704 0.699 0.686 0.675 0.704
373.15 3 0.618 0.006 0.614 0.600 0.603 0.588 0.575 0.601
373.15 4 0.527 0.005 0.536 0.520 0.529 0.515 0.499 0.526
373.15 5 0.469 0.005 0.474 0.458 0.471 0.463 - -
373.15 6 0.427 0.009 0.426 0.408 0.425 0.425 - -
AAD (%) at 373.15 K 0.90 1.67 0.98 2.26 4.75 1.03
423.15 0 1.001 0.007 0.994 0.975 0.969 1.028 0.961 0.992
423.15 1 0.800 0.017 0.794 0.794 0.793 0.824 0.780 0.798
423.15 2 0.647 0.018 0.660 0.659 0.670 0.679 0.647 0.659
423.15 3 0.566 0.025 0.565 0.558 0.578 0.577 0.549 0.560
423.15 4 0.498 0.010 0.494 0.480 0.507 0.503 0.476 0.488
423.15 5 0.440 0.004 0.438 0.419 0.451 0.450 - -
423.15 6 0.390 0.007 0.391 0.370 0.407 0.414 - -
AAD (%) at 423.15 K 0.74 2.87 2.62 3.12 2.81 1.26
Overall AAD (%) 0.74 2.91 1.47 2.35 3.84 1.03
(a)
δmrandom is the random error estimated on molality scale. The instrumental error was determined to be about 0.7 % of the measured values by application of
error propagation theory, with the procedure description detailed in the supplemental information; the δmrandom was determined by repeated measurements at
each given P-T-x condition. (b)Calculations made by OLI are under the following conditions: (1) enable the second liquid phase, (2) use MSE model, (3) stream
inflows contains 55.5082 mol water and 5 mol CO 2

18
 1.2
Koschel et al., 2006 m NaCl
323.15 K, 1
1.0
NaCl
3.15 K, 3 m
32

-1
0.8 Cl
373.15 K, 1 m Na

mCO2 / mol kg 0.6 373.15 K

,3m NaCl

SP2010
0.4 This model
This study
323.15 K, 1 mol/kg NaCl
0.2 323.15 K, 3 mol/kg NaCl
373.15 K, 1 mol/kg NaCl
373.15 K, 3 mol/kg NaCl
0.0
25 50 75 100 125 150 175 200 225
P / bar

Figure 2. 4 Comparison of experimental CO2 solubility in aqueous NaCl solutions (1 and 3 mol kg-1 NaCl)
at 323.15 and 373.15 K obtained in this study and values obtained by Koschel et al. (2006). 𝒎𝐂𝐎𝟐 is the
CO2 molality in the aqueous phase; the solid symbols stand for the experimental data of Koschel et al.
(2006), whereas the open symbols represent the experimental data obtained in this work at 150 bar. The
dash-dot lines are the results from the SP2010 model; the solid lines represent the results of the model
developed in this work.

1.4
1.3 323.15 K
AD2010
This model
1.2 This study
1.1 Malinin and Savelyeva,1972
mCO2 / mol kg-1

150 bar
1.0 OLI
0.9 This model and
0.8 AD2010 are
48 bar overlapped here
0.7
0.6
DS2006
0.5
0.4 OLI SP2010
0.3
0 1 2 3 4 5 6
mNaCl / mol kg-1

(a) 323.15 K

19
 1.1
373.15 K
1.0
This model
0.9 AD2010 This study
0.8 Malinin and Kurovskaya, 1975
150 bar

-1
mCO2 / mol kg
0.7 SP2010

0.6 This model

0.5
0.4 DS2006
0.3 48 bar OLI

0.2
OLI
0.1
0 1 2 3 4 5 6
mNaCl / mol kg-1

(b) 373.15 K

1.1
SP2010 423.15 K
1.0
0.9 This model
This study
0.8 Malinin and Kurovskaya, 1975
-1

150 bar
mCO2 / mol kg

0.7
0.6
AD2010
0.5
SP2010 DS2006
0.4
DS2006 OLI
This model
0.3
48 bar
0.2
0.1
This model
0.0
0 1 2 3 4 5 6
-1
mNaCl / mol kg
(c) 423.15 K

Figure 2. 5(a-c) Comparison of measured CO2 solubility with model calculations at (a) 323.15 K, (b)
373.15 K, and (c) 423.15 K at 150 bar and concentrations of NaCl(aq) up to 6 mol kg-1. AD2010: Akinfiev
and Diamond (2010); SP2010: Spycher and Pruess (2010); DS2010: Duan et al. (2006); OLI: OLI Studio
9.0.6. The MZLL2013 model is almost same with this model, so it was not shown in figure for the purpose
of clarity. Use of AD2010 at T > 373.15 K, as done in Figure 2.5(c), is outside the nominal validity of the
model stated by Akinfiev and Diamond (2010).

20
2.4. Thermodynamic model of the CO2-NaCl-H2O system

2.4.1 Thermodynamic framework for vapor-liquid phase equilibrium

The thermodynamic equilibrium for the CO2-NaCl-H2O system can be described in terms of the

extended Henry's law for CO2 dissolved in aqeuous solutions with NaCl(aq) (Rumpf and Maurer., 1993) or

KCl(aq) (Pérez-Salado Kamps et al., 2007):

𝑠)
𝑣̅𝐶𝑂2 (𝑃−𝑃𝑤
𝑜
𝜑𝐶𝑂2 𝑦𝐶𝑂2 𝑃 = 𝑘𝐻,𝐶𝑂2
exp⁡[ ]𝑎𝐶𝑂2 (2.1)
𝑅𝑇

and by the extended Raoult's law for water:

𝑠)
𝑣̅𝑤 (𝑃−𝑃𝑤
𝜑𝐻2𝑂 𝑦𝐻2𝑂 𝑃 = 𝑃𝑤𝑠 𝜑𝑤𝑠 exp⁡[ ]𝑎𝐻2𝑂 (2.2)
𝑅𝑇

where yCO2 and⁡yH2O are, respectively, the mole fraction of CO2 and H2O in the CO2-rich phase. In Eqs. (2.1)

and (2.2), the reference state for the chemical potential of water is chosen to be the pure liquid at the system

temperature and pressure, whereas the chemical potential of dissolved CO2 (or other ionic species) is a

hypothetical, ideal Henryan (infinitely diluted) 1 mol kg-1 solution of the solute in pure water at the

temperature and pressure of the system (Pérez-Salado Kamps et al., 2007). The pure water vapor pressure

(𝑃𝑤𝑠 ) and the fugacity coefficient of pure vapor water (𝜑𝑤𝑠 ) were calculated using the IAPWS-95 equation of

state (EoS) given by Wagner and Pruß (2002). The parameter 𝑣̅𝐶𝑂2 in Eq. (2.1) was determined by fitting

model calculated results to reliable experimental data. The values of 𝑣̅𝐶𝑂2 are presented in Table 2. 4. The

partial molar volume of water 𝑣̅𝑤 , was approximated by the molar volume of saturated liquid water, 𝑣𝑤𝑠 ,
𝑜
which was also calculated using the IAPWS-95 EoS. In Eq. (2.1), 𝑘𝐻,𝐶𝑂2
is Henry's constant (molality scale)

of carbon dioxide in pure water determined from the experimental CO2 solubility data in pure water at

temperatures up to 573.15 K. In order to achieve the best possible results over a wide range of pressures

(from 1 to 2000 bar) and temperatures (from 273.15 to 573.15 K), Henry's constant was fitted separately for

21
two distinct regions (Figure 2. 6, and Table 2. 5): (1) below 2 bar and temperatures between 273.16 and

363.15 K, and (2) the remaining region of temperatures up to 573.15 K and pressures up to 2000 bar. The
𝑜
expression for 𝑘𝐻,𝐶𝑂2
is as follows:

𝑜 𝑇 𝑇
ln⁡(𝑘𝐻,𝐶𝑂2
) = 𝑎 + 𝑏⁡ ( ) + 𝑐 ( 𝑐) + 𝑑⁡ln(𝑇/𝑇𝑐 ) (2.3)
𝑇𝑐 𝑇

where Tc = 647.096 K is the critical temperature of water; a, b, c, and d are empirical parameters given in

Table 2. 5.

̅𝑪𝑶𝟐 (cm3 mol-1) determined by this model.

Table 2. 4 The results of 𝒗

T/K P / bar
200 300 400 500 1000 1500 2000
* *
278.15 32.90 32.35 31.81 31.26 29.30 30.26 30.26
323.15 32.90 32.35* 31.81* 31.26 29.30 30.26 30.30
373.15 35.42 33.80* 32.19* 29.60 29.67 30.30 30.55
423.15 36.20 34.35* 32.05* 30.65 33.18 32.95 32.87
473.15 37.00 35.08* 33.15* 31.23 33.57 33.11 33.25
523.15 27.80 26.27* 24.73* 23.2 24.66 27.15 28.90
533.15 33.80 16.50* 22.00* 20.75 24.20 24.44 25.44
538.15 43.86 16.50* 22.00* 20.75 22.72 22.08 17.97
540.15 46.71 16.50* 22.00* 20.75 20.29 18.60
541.15 47.75 16.50 22.00 20.75 18.68
543.15 49.23 14.40 20.10 18.50 17.49
548.15 51.40 16.50 17.34 17.36 9.05
573.15 77.50 15.60 13.40 11.85
*
Indicates the data are interpolated by the method described in Table 2. 9. Other data in the table are
obtained by using the experimental CO2 solubility reported by Tödheide and Franck (1963).

Table 2. 5 Parameters of Henry's constant Eq. (2.3).

P / bar T/K a b c d
1-2 273.15-363.15 2.77716×102 -3.62801×102 5.77716×101 3.00135×102
2-2000 273.15-573.15 7.25021 1.15385×101 -1.51174×101 -3.05448×101

22
 4.5

4.0

H,CO2)
o
ln(k
3.5
This model (P>=2 bar)
Harvey (1996)
3.0 Rumpf and Maurer (1993)
Fernandez-Prini et al. (2003)
This model (P<2 bar)

2.5
0.4 0.5 0.6 0.7 0.8 0.9
Reduced temperature Tr = Tk / Tcr(H2O)

Figure 2. 6 Comparison of Henry's constant (𝒌𝒐𝑯,𝑪𝑶𝟐 / bar mol-1 Kg) obtained in this study (Eq. (2.3)) with
literature values. The Henry's constant correlation reported by Harvey (1996) and Fernández-Prini et al.
(2003) was converted from mole fraction scale to molality scale for comparison.

The fugacity coefficients of CO2 and H2O in the gas mixture were computed by a modified

Redlich-Kwong EoS given by Spycher et al. (2003). The activity coefficient of CO2(aq) and ionic species

on the molal concentration scale were calculated using Pitzer equations for aqueous electrolyte solutions

(Pitzer, 1991). By rearranging Eqs. (2.1) and (2.2), the mole fractions of CO 2 and H2O in the CO2-rich

phase can be presented as:

yCO2 = Aa CO2(aq) (2.4)

yH2O = Ba H2O(l) (2.5)

𝑠)
𝑣̅𝐶𝑂2 (𝑃−𝑃𝑤 𝑠)
𝑜 𝑣̅𝑤 (𝑃−𝑃𝑤
where 𝐴 = 𝑘𝐻,𝐶𝑂2
exp [ ] /(𝜑𝐶𝑂2 𝑃), 𝐵 = 𝑃𝑤𝑠 𝜑𝑤𝑠 exp [ ] /(𝜑𝐻2𝑂 𝑃). A and B are constants
𝑅𝑇 𝑅𝑇

at a specific system temperature and pressure. Since yH2 O + ⁡ yCO2 = 1 , Eqs. (2.4) and (2.5) can be

combined as:

23
𝐴𝑎𝐶𝑂2 (𝑎𝑞) + 𝐵𝑎𝐻2𝑂(𝑙) − 1 = 0 (2.6)

At a given salt molality, the Newton-Raphson method was used to solve Eq. (2.6) with respect to

mCO2 .

2.4.2 Incorporating Pitzer's activity model

The activity of dissolved CO2 and the activity of water in the aqueous phase can be calculated as

follows:

𝑎𝐶𝑂2 = 𝛾𝐶𝑂2 𝑚𝐶𝑂2 (2.7)

𝜙𝑀𝑊𝐻2 𝑂
𝑙𝑛𝑎𝐻2𝑂 = − (2𝑚𝑁𝑎𝐶𝑙 + 𝑚𝐶𝑂2 ) (2.8)
1000

The derivatives of Eqs. (2.7) and (2.8) with respect to mCO2 are,

𝜕𝑎𝐶𝑂2
= 𝛾𝐶𝑂2 + 𝛾𝐶𝑂2 𝑚𝐶𝑂2 (2𝜆𝑛𝑛 + 6𝑚𝐶𝑂2 𝜇𝑛𝑛𝑛 + 6𝑚𝑁𝑎𝐶𝑙 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑁𝑎𝐶𝑙 ) (2.9)
𝜕𝑚𝐶𝑂2

𝜕𝑎𝐻2 𝑂 𝑎𝐻2 𝑂 𝑀𝐻2 𝑂 2

=− [1 + 2(𝑚𝐶𝑂2 𝜆𝑛𝑛 + 3𝑚𝐶𝑂 𝜇
2 𝑛𝑛𝑛
+ 𝑚𝑁𝑎𝐶𝑙 𝐵𝐶𝑂2 −𝑁𝑎𝐶𝑙 + 6𝑚𝐶𝑂2 𝑚𝑁𝑎𝐶𝑙 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑁𝑎𝐶𝑙 +
𝜕𝑚𝐶𝑂2 1000

2
𝑚𝑁𝑎𝐶𝑙 𝜉𝑛𝑐𝑎 )] (2.10)

The Pitzer's expressions for the activity coefficient of CO2 and the osmotic coefficient of H2O are

given in Appendix A by Eqs. (A6) and (A7), respectively. By substituting Eq. (A6) into Eq. (2.7), the

partial derivative of 𝑎𝐶𝑂2 with respect to mCO2 can be obtained resulting in Eq. (2.9). The partial derivative

24
of 𝑎𝐻2𝑂 with respect to 𝑚𝐶𝑂2 resulting in Eq. (2.10) is obtained in the same manner by substituting Eq. (A7)

into Eq. (2.8). Thus, the mCO2 can be computed by substituting Eqs. (2.7)-(2.10) into Eq. (2.6). The overall

flow chart for calculating CO2 solubility, including the solution to Eq. (2.6), is shown in Figure 2. 7. In this

calculation, the initial guess of water content in the CO2-rich phase was made by using the ideal water

content, i.e., 𝑦𝐻2𝑂 = 𝑃𝑤𝑠 /𝑃. One of the main differences between our model and the SP2010 model is the

computation of the water activity. In this study, we used the Pitzer model with an iterative Newton-

Raphson procedure, whereas Spycher and Pruess (2010) assumed that the water activity equals the mole

fraction of water in the aqueous phase, which is a rough assumption for highly concentrated aqueous

solutions but allowed a non-iterative solution for Eq. (2.6).

The Pitzer equations for the activity coefficient of CO2 and the osmotic coefficient of water are

listed in Appendix A (Pitzer, 1991). In order to simplify the notation of Eq. (A6) and Eq. (A7), we defined

two parameters, 𝐵𝐶𝑂2 −𝑁𝑎𝐶𝑙 and 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑁𝑎𝐶𝑙 according to Akinfiev and Diamond (2010) as follows:

𝐵𝐶𝑂2 −𝑁𝑎𝐶𝑙 = 𝜆𝑛𝑐 + 𝜆𝑛𝑎 (2.11)

𝐶𝐶𝑂2 −𝐶𝑂2 −𝑁𝑎𝐶𝑙 = 𝜇𝑛𝑛𝑐 + 𝜇𝑛𝑛𝑎 (2.12)

These two parameters are taken from Akinfiev and Diamond (2010). The neutral component

interaction parameters 𝜆𝑛𝑛 and 𝜇𝑛𝑛𝑛 of the Pitzer's model in Eqs. (2.9) and (2.10) were determined using

experimental CO2 solubility data for the CO2-H2O system (data sources are listed in Table 2. 13). The

quadratic form of parameters 𝜆𝑛𝑛 and 𝜇𝑛𝑛𝑛 suggested by Akinfiev and Diamond (2010) was used during

the data fitting process, and the results were expressed using Eq. (2.13). The detailed data fitting procedure

to obtain 𝜆𝑛𝑛 and 𝜇𝑛𝑛𝑛 are illustrated in Figure 2. 8.

𝜆𝑛𝑛 ⁡(𝑜𝑟⁡𝜇𝑛𝑛𝑛 ) = a1 + a2T + a3 T2 (2.13)

25
 Input P-T-x condition

Initial IPAWS-95 Water EoS

Initial guess for yH2O
parameters

kCO2-H2O
Mixing rule, Eq. (2.14)
Table 2.7

Solve v-cubic form of the

IAPWS-95 Water EoS
modified RK-EoS

Calculate the fugacity Fugacity coefficient of pure water

coefficients, φH2O and φCO2, Vapor pressure
Eq. (A5) Saturation molar volume

Thermodynamic equilibrium 1. Henry’s constant, Eq. (2.3)

framework, Eqs. (2.1-2.2) 2. vCO2, Table 2.4

Solve Eq. (2.6), Newton Pitzer’s activity model

Raphson approach Eqs. (2.7-2.10), (A6) and (A7)

N Pitzer pure Pitzer triple-ion

If (mCO2(new)/mCO2(old)-1) < ε component interaction
parameter, Table parameter,
2.6, Eq.(2.13) Table 2.7
Y
Obtain mCO2 and aCO2 , aH2O
simultaneously

Use Eq. (2.5) to update yH2O

Obtain mCO2 and aCO2 simultaneously

Deviation = abs(yH2O(new)/yH2O(old)-1)

N
If Deviation < ε

Y
Output
mCO2 and yCO2

Figure 2. 7 The flow chart used to calculate the mutual solubilities of CO2 and H2O in the CO2-NaCl-H2O
system. The inner loop solves the CO2 solubility in aqueous phase; the outer loop solves the H2O solubility
in the CO2-rich phase. The convergence criteria is set to 𝛆=10-5 during the calculations.

26
 The parameters in Eq. (2.13) are given in Table 2. 6. The Pitzer binary ion–ion interaction

(0) (1)
parameters 𝛽𝑐𝑎 , 𝛽𝑐𝑎 and 𝐶𝑐𝑎 for the CO2-NaCl system were taken from Pitzer et al. (1984).

Initial Henry’s law Experimental CO2 solubility

constant (koH,CO2) 1 data (mCO2)
Fernández-Prini (2003)

Obtain a new Henry’s Activity of Activity of coefficient of

constant (koH,CO2) 6 dissolved 2 dissolved CO2
CO2 (aCO2) γCO2=aCO2/mCO2

5 3

Use Eq. (A6) to Least square fitting Eq. (A6)

for binary CO2-H2O system
calculate a new activity 4 using experimental data to
of dissolved CO2 determine λnn and μnnn

Figure 2. 8 The data fitting procedure for determining the neutral component parameters 𝝀𝒏𝒏 , 𝝁𝒏𝒏𝒏 (Eq.
(2.13)) and the Henry’s constant 𝒌𝒐𝑯,𝑪𝑶𝟐 (Eq. (2.3)): (1) Use Henry’s constant (Fernández-Prini et al., 2003)
as an initial value, the activity of dissolved CO2 can be calculated via Eq. (2.1); (2) Use experimental CO2
solubility data and the obtained activity of dissolved CO2, the activity coefficient of dissolved CO2 can be
calculated by a simple relation 𝜸𝑪𝑶𝟐 = 𝒂𝑪𝑶𝟐 /𝒎𝑪𝑶𝟐 ; (3) Neutral component parameters 𝝀𝒏𝒏 and 𝝁𝒏𝒏𝒏 are
obtained by least-square fitting of Eq. (A6) using experimental CO2 solubility data for the CO2-H2O system;
(4) Activity of dissolved CO2 is re-calculated using the newly obtained 𝝀𝒏𝒏 and 𝝁𝒏𝒏𝒏 via Eq. (A6) as well
as the experimental CO2 solubility data; 5) Obtain a new Henry’s constant from Eq. (2.1). Iterations of the
steps (2)-(3)-(4)-(5)-(6)-(2) are needed to get a good consistency between the new- and old-Henry’s
constant. This procedure allows us to obtain⁡𝒌𝒐𝑯,𝑪𝑶𝟐 , 𝝀𝒏𝒏 and 𝝁𝒏𝒏𝒏 simultaneously after a few iterations.
The experimental data used during the fitting procedure are listed in Table 2. 12 (No. 1, 2, 3, 5, 9, 10, 13,
15, 16, 18, 19, 21, 22, 26, 28, 30, 41).

Table 2. 6 Parameters of Eq. (2.13).

Parameters a1 a2 a3
𝜆𝑛𝑛 -1.6662×10-01 4.9482×10-04 -5.1643×10-07
𝜇𝑛𝑛𝑛 8.9196×10-02 -2.8829×10-04 2.3197×10-07
The units of parameters a1, a2, a3 for λnn are kg mol , kg mol K , kg mol K , respectively and for μnnn
-1 -1 -1 -1 -2

are kg2 mol-2, kg2 mol-2 K-1, kg2 mol-2 K-2, respectively

The Pitzer triple-ion interaction parameter 𝜉𝑛𝑐𝑎 in Eqs. (2.10), (A6) and (A7) was determined by

using the bisection search algorithm to find the exact values of 𝜉𝑛𝑐𝑎 against the experimental CO2 solubility

data collected in this study over a concentration range of 0-6 mol kg-1 NaCl and temperatures of 323.15-

27
423.15 K. In this study, the parameter 𝜉𝑛𝑐𝑎 was found to be insensitive to the pressure, but it is a function

of temperature and NaCl molal concentration. Using 𝜉𝑛𝑐𝑎 given in Table 2. 7, the cubic spline interpolation

was used to characterize the relationship between 𝜉𝑛𝑐𝑎 and NaCl molal concentration, the linear

interpolation for the temperature was used to determine values of 𝜉𝑛𝑐𝑎 at any temperatures between 273.15

and 573.15 K.

Table 2. 7 The Pitzer triple-ion interaction parameter (𝝃𝒏𝒄𝒂 / kg2 mol-2) for the CO2-NaCl-H2O system at
temperatures from 323.15 to 423.15 K and molal NaCl(aq) concentrations from 1 to 6 mol kg -1.

T/K NaCl molality (mol kg-1)

1 2 3 4 5 6
298.15 0.01617* 0.00250* -0.00111* -0.00289* -0.00390* -0.00450*
323.15 0.01213 0.00218 -0.00079 -0.00238 -0.00334 -0.00395
373.15 0.00420 0.00161 -0.00012 -0.00133 -0.00221 -0.00285
423.15 0.04899 0.02003 0.00950 0.00466 0.00208 0.00062
473.15 0.09379* 0.03845* 0.01921* 0.01065* 0.00638* 0.00409*
523.15 0.13858* 0.05687* 0.02888* 0.01664* 0.01067* 0.00755*
*
573.15 0.18337 0.07529* 0.03854 *
0.02263 *
0.01496* 0.01102*
*
Indicates the data interpolated at 150 bar and the corresponding temperatures (T / K) by the method
described in Table 2. 9.

Table 2. 8(a, b) The mixing rule binary interaction parameter, 𝒌𝑪𝑶𝟐 −𝑯𝟐𝑶 (dimensionless), determined from
experimental H2O solubility in the CO2-rich phase.

(a) T < 473.15 K

T/K 273.15 288.15 298.15 304.19 313.15 323.15 348.15 373.15 423.15 473.15
k CO2−H2O -25 -20.5 -17 -14.3 -11.5 -9 -5.6 2.1 1.9 0.32

(b) T ≥ 473.15 K
T/K 473.15 523.15 533.15 538.15 540.15 541.15 543.15 548.15 573.15
P / bar k CO2−H2O
200 0.436 -0.065 -0.235 -0.247 -0.255 -0.273 -0.291 -0.312 -0.272
300 0.548* 0.025* -0.050* -0.070* -0.070* -0.078 -0.085 -0.095 -0.226
400 0.646* 0.098* 0.037* -0.012 *
-0.009* -0.008 -0.019 -0.045 -0.215
500 0.698 0.136 0.056 0.003 0.001 -0.013 -0.032 -0.081 -0.324
600 0.697* 0.118* 0.036* -0.020* -0.031* -0.058 -0.092 -0.160
700 0.692* 0.089* -0.002* -0.070* -0.090* -0.124 -0.164 -0.252
800 0.685* 0.060* -0.051* -0.125 *
-0.158* -0.192 -0.242 -0.366
900 0.674* 0.034* -0.098* -0.172 *
-0.218* -0.251 -0.308 -0.492
1000 0.660 0.015 -0.138 -0.211 -0.265 -0.307 -0.375 -0.679
1100 0.644* 0.003* -0.174* -0.248* -0.307* -0.359 -0.450
1200 0.627* -0.004* -0.207* -0.284 *
-0.349* -0.413 -0.537
1300 0.608* -0.010* -0.235* -0.320 *
-0.392* -0.473 -0.674
1400 0.595* -0.014* -0.259* -0.357 *
-0.438* -0.536
1500 0.590 -0.019 -0.278 -0.396 -0.491 -0.627
2000 0.643 -0.042 -0.314 -0.609

28
*
Indicates the data interpolated by the method described in Table 2. 9. At T < 473.15 K, the values of
𝑘𝐶𝑂2 −𝐻2𝑂 were evaluated using experimental data of Weibe and Gaddy (1941), Malinin (1959), Tödheide
and Franck (1963), Müller et al. (1988), King et al. (1992). At T > 473.15 K, the values of 𝑘𝐶𝑂2 −𝐻2𝑂 were
evaluated using experimental data of Tödheide and Franck (1963).

2.4.3. Improved mixing rule parameter (𝒌𝑪𝑶𝟐−𝑯𝟐𝑶 ) for the CO2-rich phase

The Spycher and Pruess (2010) modified Redlich-Kwong equation of state (RK EoS) (see

Appendix A) was used in this study to calculate the fugacities of H2O and CO2 for the CO2-rich phase. A

combined rule for polar systems (e.g. CO2-H2O) proposed by Panagiotopoulos and Reid (1986) has been

proven successful when describing the gas mixture in the CO2-H2O system (Panagiotopoulos and Reid,

1986; Spycher and Pruess, 2010). In this approach, the binary interaction parameter, 𝑎𝑖𝑗 , takes the

following form:

𝑎𝑖𝑗 = √𝑎𝑖 𝑎𝑗 (1 − 𝑘𝑖𝑗 + (𝑘𝑖𝑗 − 𝑘𝑗𝑖 )𝑦𝑖 ) (2.14)

where 𝑦𝑖 ⁡is the mole fraction of component i in the gas mixture, and 𝑘𝑖𝑗 and 𝑘𝑗𝑖 are the binary interaction

parameters (in this study, i and j denote CO2 and H2O, respectively). The binary interaction parameters play

a critical role in describing the CO2-rich phase diagram. The parameters 𝑘𝐻2𝑂−𝐶𝑂2 ⁡and 𝑘𝐶𝑂2 −𝐻2𝑂 evaluated

by Spycher and Pruess (2010) were found to work fine at low to moderate pressures, but with these

parameters unchanged, the model cannot accurately predict the mutual solubilities of CO 2 and H2O at

pressures above 800 bar. Therefore, a modification of the parameters should be made to allow the model to

apply over a wider pressure range. It was decided to modify one of the parameters, and 𝑘𝐶𝑂2 −𝐻2𝑂 was

chosen because this parameter has the most pronounced influence on calculating the H 2O solubility in the

CO2-rich phase. In such an approach, 𝑘𝐻2𝑂−𝐶𝑂2 ⁡was treated in the same manner as Spycher and Pruess

(2010) did. The fitting results of the parameter 𝑘𝐶𝑂2 −𝐻2𝑂 are shown in Table 2. 8.

Determination of the binary interaction parameter 𝑘𝐶𝑂2 −𝐻2𝑂 was made using the experimental H2O

solubility data collected in the CO2-rich phases. The approach selected to obtain the 𝑘𝐶𝑂2 −𝐻2𝑂 was purely

29
empirical. At a given temperature and pressure, the experimental H 2O solubility data in the CO2-rich phase

was entered into the model, and the parameter 𝑘𝐶𝑂2 −𝐻2𝑂 was then manually adjusted until consistency was

reached between calculated and experimental values. At temperatures below 473.15 K, the experimental

data used to determine the 𝑘𝐶𝑂2 −𝐻2𝑂 parameter were taken from Wiebe and Gaddy (1941), Tödheide and

Franck (1963), Müller et al. (1988), and King et al. (1992). By fitting the experimental H2O solubility data

in the CO2-rich phase, we found that assuming a temperature dependence and pressure independence of

k CO2−H2O is suitable when the temperature is below 473.15 K.

The pressure influence on 𝑘𝐶𝑂2 −𝐻2𝑂 , however, becomes more important as the temperature

increases, especially above 473.15 K. Reliable experimental data are desirable to evaluate the parameter

k CO2−H2O at high P-T conditions. Both Takenouchi and Kennedy (1964) and Tödheide and Franck (1963)

reported experimental data for the CO2-H2O system at high temperatures and pressures using the sampling

method for both the CO2-rich and the aqueous phases (Mather and Franck, 1992). While the aqueous phase

compositions obtained by Takenouchi and Kennedy (1964) and Tödheide and Franck (1963) are in fairly

good agreement, the compositions of the CO2-rich phase are significantly different. In order to resolve this

discrepancy, Mather and Franck (1992) used the so-called 'synthetic' experimental method to study the

phase equilibria of the CO2-H2O system. Initially, known amounts of CO 2 and H2O are loaded into an

autoclave (with a sapphire window). By heating the components in the autoclave at constant volume,

homogeneous single-phase conditions can be observed through the window. By this means, Mather and

Franck (1992) measured the dew points (at 498.15-546.15 K and 1100-2640 bar) of H2O in the CO2-rich

phase to support the experimental data reported by Tödheide and Franck (1963). The experimental data of

Tödheide and Franck (1963) are also validated by the theoretical calculations of Gallagher et al. (1993) and

Shyu et al. (1997). Based on the above reasons, the data from Tödheide and Franck (1963) were selected

for the determination of the binary interaction parameter 𝑘𝐶𝑂2 −𝐻2𝑂 at temperatures above 473.15 K.

In order to correlate Tödheide and Franck’s (1963) data, 𝑘𝐶𝑂2 −𝐻2𝑂 were considered as a function

of both temperature and pressure. However, introducing a pressure-dependent binary interaction parameter

in the mixing rule will disrupt the cubic form of the original RK EoS, thus the expressions for the fugacity

coefficients of CO2 and H2O in Eq. (A5) may not hold true at temperatures higher than 473.15 K. In

addition, although 𝑘𝐶𝑂2 −𝐻2𝑂 was considered a function of pressure and temperature, the analytical

30
relationships between 𝑘𝐶𝑂2 −𝐻2𝑂 and (P, T) are still unknown; thus, it is impossible to derive a new fugacity

coefficient expression. By using the original fugacity coefficient expressions (Eq. (A5)), we compared our

calculated 𝜑𝐻2𝑂 and 𝜑𝐶𝑂2 for the CO2-rich phase with the SP2010 model at elevated temperatures; the

AAD(%) of calculated 𝜑𝐶𝑂2 and 𝜑𝐻2𝑂 from 473.15 to 573.15 K and up to 600 bar between the two models

are approximately 4.8% and 6.0%, respectively. Thus, the tuning of the parameter 𝑘𝐶𝑂2 −𝐻2𝑂 caused a small

change in the fugacity coefficient calculations, but these changes were shown to improve the overall

performance of the model for the calculation of the mutual solubilities in the CO 2-H2O system at elevated

temperatures and pressures (Figure 2. 9). The detailed descriptions of this new CO 2 solubility model can be

found in Table 2. 9.

31
Table 2. 9 The components, key parameters, approaches and limitations of the CO 2 solubility model presented in this study.

No. Model Description Sources of Applicable Ref.

Components Components P-T range
1 RK EoS A modified RK EoS was selected to calculate fugacities of CO 2 and H2O in the CO2- Spycher et al 273.15-573.15K Eqs. (A1),
rich phase in the CO2 solubility model. (2003) 1-2000 bar (A5)
2 Mixing rule The two-parameter mixing rule for polar molecules was used. Panagiotopou 273.15-573.15K Eqs. (14),
los and Reid 1-2000 bar (A3), A4
3 𝑘𝐻2𝑂−𝐶𝑂2 Mixing rule binary interaction parameter k H2 O−CO2 for the CO2-H2O system was (1986) 273.15-573.15K -
given by Spycher et al. (2003). 1-2000 bar
4 𝜑𝐻2𝑂 , 𝜑𝐶𝑂2 Fugacity coefficients of H2O and CO2 for the CO2-rich phase were calculated using 273.15-573.15K Eq. (A5)
equations provided by Spycher et al. (2003) and Panagiotopoulos and Reid (1986). 1-2000 bar
5 𝑘𝐶𝑂2 −𝐻2𝑂 Approach to obtain the 𝑘𝐶𝑂2 −𝐻2𝑂 was purely empirical. At T>473.15 K, the kCO2-H2O Fitted in this 273.15-573.15K Table 2. 8
was determined using experimental data of Tödheide and Franck (1963) for the study 1-2000 bar
CO2-rich phase. At 273.15<T<473.15 K, 𝑘𝐶𝑂2 −𝐻2𝑂 was evaluated using experimental
data of Wiebe and Gaddy (1941), Müller et al. (1988), and King et al. (1992).
6 Interpolation (1). At 273.15 K<T<473.15 K, 𝑘𝐶𝑂2 −𝐻2𝑂 was considered as only temperature- Fitted in this 273.15-573.15K Table 2. 8
methods used dependent. A cubic spline interpolation with a step of 1 K was used to get the study 1-2000 bar
to calculate relationship between 𝑘𝐶𝑂2 −𝐻2𝑂 ⁡and temperature (T). (2). At T>473.15 K, 𝑘𝐶𝑂2 −𝐻2𝑂
𝑘𝐶𝑂2 −𝐻2𝑂 was treated as a function of both temperature (T) and pressure (P). A cubic spline
at any interpolation with a step of 1 bar was used to get the relationship between 𝑘𝐶𝑂2 −𝐻2𝑂
targeted P-T and P at a constant T, then the linear interpolation was applied to calculate 𝑘𝐶𝑂2 −𝐻2𝑂
at any targeted P-T point.
7 Henry's law We have fitted Henry's law constant separately for two distinct regions: (1). P<2 bar Fitted in this 273.15-573.15K Eq. (2.3)
constant and temperatures of 273.16-363.15 K; (2). the remaining P-T region. By fitting the study 1-2000 bar Figure 2. 8
Henrys constant specific for low P conditions, the model calculated error at low Figure 2. 6.
pressures (e.g. 1.013 bar) can be reduced dramatically from ~5% to ~1%. Thus, it is
worthwhile to include two distinct Henry’s constant for two P regions in the model.
The detailed data fitting procedure is provided in Figure 2. 8.
8 𝜑𝑤𝑠 , Ps , The fugacity coefficient, vapor pressure, molar volume, and density of pure water Wagner and 238.15-873.15 K, Eqs. (2.1-2.2)
𝑣𝑠 ,⁡𝜖𝑟 , 𝜌𝐻2𝑂 were computed using IAPWS-95 EoS. The relative permittivity was calculated by Pruß (2002) 0.1-12000 bar for
equation given by Fernández et al. (1997) in order to compute the Debye-Hückel Fernández et 𝜖𝑟 , the others
slope in the Pitzer’s activity model. al.(1997) 251.15-1273.15 K
0.1-10000 bar
9 𝑣̅𝐶𝑂2 Our approach to obtain 𝑣̅𝐶𝑂2 was purely empirical. The experimental CO2 solubility Fitted in this 273.15-573.15K Table 2. 4
data given by Tödheide and Franck (1963) were used to determine the values of study 1-2000 bar

32
No. Model Description Sources of Applicable Ref.
Components Components P-T range
𝑣̅𝐶𝑂2 . The calculated CO2 solubility is not sensitive to 𝑣̅𝐶𝑂2 at low pressures (e.g.
P<300 bar), but it has a significant influence in predicting CO 2 solubility in the
aqueous phase at high pressures (e.g. P>1000 bar).
10 Interpolation At a given P-T point, the 𝑣̅𝐶𝑂2 can be obtained by: (1) using the cubic spline Fitted in this 273.15-573.15K Table 2. 4
methods used interpolation to get the relationship between 𝑣̅𝐶𝑂2 and P at a constant T; (2) and then study 1-2000 bar
to get v̅CO2 at applying the linear interpolation of T to get the 𝑣̅𝐶𝑂2 at any targeted P-T point.
targeted P-T
11(a) Pitzer activity The Pitzer activity model was incorporated into the CO 2 solubility model. The Fitted in this 273.15-573.15K Eqs. (A6)-
model benefits of incorporating the Pitzer activity model are: (1). Calculate phase equilibria study 1-2000 bar (A7)
of the CO2-H2O and CO2-NaCl-H2O system in a larger P-T region than previously
developed models (Spycher and Pruess, 2010; Akinfiev and Diamond, 2010, Duan Eqs.
et al., 2006); (2). Provide a higher accuracy than some of previously developed (2.7)-(2.10)
models (Spycher and Pruess, 2010; Duan et al., 2006); 3. The model can be
extended to other aqueous systems such as CaCl2(aq), MgCl2(aq), Na2SO4(aq) and
KCl(aq).
12 𝜆𝑛𝑛 , 𝜇𝑛𝑛𝑛 The Pitzer pure neutral components interaction parameters were determined in this Fitted in this 273.15-573.15K Eq. (2.13)
study by using the CO2 solubility data in pure water at elevated Ts and Ps. The study Not sensitive to Table 2. 6
detailed data fitting procedure and experimental data references are provide in pressure Figure 2. 8
Figure 2. 8.
13* (0) (1)
𝛽𝑐𝑎 , 𝛽𝑐𝑎 , The Pitzer binary ion-ion interaction model were taken from Pitzer et al. (1984). Pitzer et al. 273.15-573.15K -
𝐶𝑐𝑎 (1984) 1-1000 bar
0-6 mol kg-1 NaCl
14(b) 𝐵𝐶𝑂2 −𝑁𝑎𝐶𝑙 The combined Pitzer interaction parameters depend on T and NaCl concentration. Akinfiev and 273.15-373.15K Eqs. (A10)-
𝐶𝐶𝑂2 −𝐶𝑂2 −𝑁𝑎𝐶𝑙 They were taken from Akinfiev and Diamond (2010) Diamond Not sensitive to (A11)
(2010) pressure Table A1
15 ξnca Pitzer's triple-ion interaction parameter 𝜉𝑛𝑐𝑎 was purely determined from our Fitted in this 273.15-373.15K Table 2. 7
experimental data at 323.15, 373.15, 423.15 K, and can be extrapolated to any study Not sensitive to
temperature from 273.15 to 573.15 K. No other experimental data listed in Table 2. pressure
13 for the CO2-NaCl-H2O system were used to fit this parameter. A cubic spline
interpolation with a step of 0.001 mol kg-1 NaCl was used to get the relationship
between 𝜉𝑛𝑐𝑎 ⁡ and NaCl molal concentration at a constant T, then the linear
interpolation was applied to calculate 𝜉𝑛𝑐𝑎 ⁡at any targeted P-T-x point. The model
can accurately reproduce experimental data which are not used in the data fitting
procedure.

33
No. Model Description Sources of Applicable Ref.
Components Components P-T range
16 Phase The phase equilibria equations were solved by using the Netwon-Raphson method. Calculated in 273.15-573.15K Eqs. (2.1-
equilibria There are two iteration loops: the inner loop solves the CO2 solubility in aqueousthis study 1-2000 bar 2.10)
calculations phase (mCO2 ); the outer loop solves the H2O solubility in the CO2-rich phase (yH2O ) Figure 2. 7
based on the inner loop results of mCO2 .
(a)
The Pitzer model for the CO2-H2O system was confirmed at pressures up to 2000 bar and temperatures up to 573.15 K using the experimental data of Tödheide
and Franck (1963)
(b)
BCO2 −NaCl and CCO2−CO2−NaCl were determined by Akinfiev and Diamond (2010) at the temperatures from 273.15 to 373.15 K, these parameters were used in
Eq. (A6) to calculate the activity coefficient of the dissolved CO 2 in the aqueous phase. In this study, any uncertainties caused by these parameters at the
temperature above 373.15 K is absorbed by triple-ion interaction parameter ξnca during the data fitting process.

34
2.4.4 Results and discussion

The calculated phase diagrams of the CO2-H2O and CO2-NaCl-H2O systems for both the aqueous

and CO2-rich phases are shown in Figure 2. 9 and Figure 2. 10, respectively. In Figure 2. 9 and Figure 2.

10, at temperatures higher than 538.15 K, the modeling results begin to deviate from the experimental

values at pressures near the critical region, and finally the calculation does not converge when the pressure

increases further. The model is incapable of predicting the CO 2 solubility around critical points, most likely

because the 𝛾 − 𝜑 approach uses different models for the CO2-rich and aqueous phases (e.g. RK EoS for

the CO2-rich phase, Pitzer activity model for the aqueous phase).

K
473.15 K
423.15 K

533.15 K
523.15 K

2000
K

5K
5K

538.15

523.15 K

423.15 K
473.15 K
533.15

5 K 41.15 540.1
538.1

1750 540.15 K
K

541.15 K
1500
3.1 5

Aqueous 543.15 K Supercritical

1250 phase CO2-rich phase
54

line
P / bar

ical

548.15 K
1000
K
Crit
15
8.
54

750
573.15 K
500 573.15 K

250

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
xCO2

Figure 2. 9 Pressure-composition phase diagram for the CO2-H2O system at temperatures from 423.15 to
573.15 K and pressures up to 2000 bar for both the aqueous and the CO2-rich phases. The solid lines
represent the calculated results from our model and the lines on the left-hand of the dashed critical line
stand for the bubble-point curves of CO2-saturated H2O(l) whereas the curves on the right-hand of the
critical line represent the dew-point curves of H2O-saturated supercritical CO2. The open circles and
triangles represent the experimental data from Tödheide and Frank (1963), the solid circles and inverted
solid triangles show Malinin’s (1959) data, the solid triangles correspond to the results of Müller et al.
(1988). The model has AAD of 1.51% for the aqueous phase and 4.11% for the CO2-rich phase when
compared with the experimental data reported by Tödheide and Frank (1963).

35
 523.15 K
433.15 K

523.15 K
2000

473.15 K
5K

533.15 K

433.15 K
538.15 K

538.15
533.1
1750 540.15 K 540.15 K

1500 Aqueous541.15 K 541.15 K

phase 543.15 K Supercritical
1250 543.15 K

line
P / bar CO2-rich phase

ical
1000

Crit
548.15 K 548.15 K

750
573.15 K
500 573.15 K

250 4 mol kg-1 NaCl

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
x'CO2

Figure 2. 10 The model-calculated pressure-composition phase diagram of the CO2-NaCl-H2O system for 4
mol kg-1 NaCl at temperatures from 433.15 to 573.15 K, and pressures from 1 to 2000 bar for both the
aqueous phase and CO2-rich phase. 𝒙′𝑪𝑶𝟐 stand for salt-free mole fraction of CO2, and it is calculated as
𝒙′𝑪𝑶𝟐 = 𝒎𝑪𝑶𝟐 /(𝒎𝑪𝑶𝟐 + 𝟏𝟎𝟎𝟎/𝑴𝒘𝑯𝟐𝑶 ), where 1000 is the amount of water in gram. Open circles are the
experimental data for the CO2-H2O binary system from Tödheide and Frank (1963), served as the
benchmark for the comparisons of the phase behavior when NaCl is introduced into the system.

As shown in Figure 2. 9 and Figure 2. 10, convergence problems in the vicinity of critical region

are also due to relatively high temperatures and pressures where the cubic EoS has a single root (Veeranna

and Rihani, 1984), and the VLE calculation in the presence of H2O could be more complex around the

critical point due to the strong non-ideality of the polar molecules. Gallagher et al. (1993) used a Helmholtz

free energy model to calculate thermodynamic properties of the CO 2-H2O binary system up to 623.15 K

and observed a numerical instability near the critical region. Shyu et al. (1997) presented a calculated CO2-

H2O phase diagram at 623.15 K by using an EoS and the Wong-Sandler mixing rule. Their phase diagram

has a smooth transition through the no-convergence regions encountered by Gallagher et al. (1993) (as well

as by us), but no details on the calculation procedure are provided by Shyu et al. (1997).

The CO2 solubility model developed in this study for the CO 2-NaCl-H2O system can be used to

predict CO2 solubility in the aqueous phase, as well as H2O solubility in the CO2-rich The CO2 solubility

model developed in this study for the CO2-NaCl-H2O system can be used to predict CO2 solubility in the

aqueous phase, as well as H2O solubility in the CO2-rich phase, at a P-T-x range of 1-2000 bar, 273.15-

36
573.15 K and 0-6 mol kg-1 NaCl. In Table 2. 10, the calculated CO2 solubilities in the aqueous phase are

compared with selected reliable experimental data (Wiebe and Gaddy, 1939; Malinin and Savelyeva, 1972;

Malinin and Kurovskaya, 1975; Rumpf et al. 1994; Yasunishi and Yoshida, 1979a; Tödheide and Franck,

1963) and the published CO2 solubility models (MZLL2013, AD2010, SP2010, DS2006, and OLI). The
exp
100 mcalc
CO2 ,i −mCO2 ,i
average absolute deviation (AAD = ∑N
i=1 |
P
exp |%) of the CO2 solubility in the aqueous phase
Np mCO ,i
2

between the model calculations and the experimental data shows that our model provides the most reliable

performance among the aforementioned models (Table 2. 10).

The model also predicts the H2O solubility in the CO2-rich phase over a wide P-T range (1-2000

bar, 273.15-573.15 K) with a very good degree of accuracy (Table 2. 11, Figure 2. 9, Figure 2. 11 and

Figure 2. 13). A comparison of the literature models reveals that the AD2010 model only provides a rough

estimate of the H2O solubility in the CO2-rich phase and the DS2006 and MZLL2013 models do not

provide the H2O solubility in the CO2-rich phase. Therefore, thus far, the most reliable models to estimate

the H2O solubility in the CO2-rich phase are the SP2010 model and the commercial software package OLI

Studio 9.0. Table 2. 11 shows a comparison between the calculated results of three models (SP2010, OLI,

and our model) and the reliable experimental H2O solubility data in the CO2-rich phase (Wiebe and Gaddy,

1941; Tödheide and Franck, 1963; Müller et al., 1988; Valtz et al., 2004). One can see that these three

models (our model, OLI, SP2010) predict very similar H 2O solubility in the CO2-rich phase in low to

moderate P-T regions, but the model developed herein covers the largest P-T range (Table 2. 11).

37
Table 2. 10 The average absolute deviation (AAD %) of CO2 solubility in the aqueous phase between model calculations and the selected reliable experimental
data over a wide P-T-x range (25-2000 bar, 288.15-573.15 K, and 0-6 mol kg-1 NaCl). The values in parentheses represent the number of experimental data points
evaluated by each model.

Ref. Np(b) P / bar T/K 𝒎𝐍𝐚𝐂𝐥 Model deviation from experimental data, AAD %
No.#(a) / mol kg-1
This study OLI(c) AD2010 SP2010 DS2006 MZLL2013
1 21 150 323.15-423.15 0-6 0.74(21) 2.91(21) 1.47(15) 2.35(21) 3.84(15) 0.91(21)
2 29 25-709 323.15-373.15 0 1.90(29) 3.05(29) 1.82(29) 1.70(25) 1.66(29) 2.55(29)
3 32 48 298.15-423.15 0-5.91 1.86(32) 4.64(32) 1.48(28) 1.95(32) 1.74(28) 2.04(28)
4 63 1-100 313.15-433.15 4-6 1.62(63) 4.63(63) 1.46(38) 5.53(63) 3.37(35) 2.09(35)
5 27 1.013 288.15-308.15 0.45-5.73 1.83(27) 3.97(27) 2.65(27) 2.59(27) 2.82(23) 2.09(24)
6 78 200-2000 323.15-573.15 0 1.51(78) 9.98(56) 1.07(6) 3.94(35) 6.27(30) 7.03(69)
Overall AAD % (No.1-No.5) 1.59(172) 3.84(172) 1.78(137) 2.82(168) 2.67(130) 1.94(137)
Overall AAD % (No.1-No.6) 1.58(250) 4.86(228) 1.66(143) 3.01(203) 3.28(160) 2.79(206)
Total data coverage rate by each model (N/Np) 100% 91% 57% 81% 64% 82%
(a)
1. This study; 2. Wiebe and Gaddy (1939); 3. Malinin and Savelyeva (1972), Malinin and Kurovskaya (1975); 4. Rumpf et al. (1994); 5. Yasunishi and
Yoshida (1979); 6. Tödheide and Franck (1963).
(b)
Np is the total number of experimental data located in the targeted P-T range (P: 1-2000 bar, T: 298.15-573.15 K).
(c)
Calculations made by OLI are under the following conditions: (1) enable the second liquid phase, (2) use MSE model, (3) stream inflows contains 55.5082 mol
water and 15 mol CO2.

38
Table 2. 11 The average absolute deviation (AAD %) of the H 2O solubility in the CO2-rich phase between model calculations and the selected reliable
experimental data at 1-2000 bar and 278.15-573.15 K. The numbers in the round brackets are stand for the amount of experimental data points evaluated by the
model.

Ref. Np(a) P / bar T/K Model deviations from experimental data, AAD %(b)
This study OLI(c) SP2010
Valtz et al. (2004) 30 4-80 278.15-318.15 11.88(30) 11.92(30) 12.06(27)
Müller et al. (1988) 49 3-80 373.15-473.15 6.88(49) 5.66(49) 4.85(49)
Wiebe and Gaddy (1941) 39 1-709 298.15-348.15 7.88(16)(b) 4.42(16)(b) 5.67(16)(b)
(b)
Tödheide and Franck (1963) 78 200-2000 323.15-573.15 4.11(78) 7.02(3) 6.18(31)(b)
Overall AAD % 7.69 (173) 7.26(98) 7.79(123)
Total data coverage rate by each model (N/Np) 88.3% 50% 62.8%
(a)
Np is the total number of experimental data located in the targeted P-T range (1-2000 bar and 273.15-573.15 K).
(b)
For the H2O solubility in the CO2-rich phase, the data points which AAD is larger than 20 % were not included in the comparison.
(c)
Calculations made by OLI are under the following conditions: (1) enable the second liquid phase, (2) use MSE model, (3) stream inflows contains 55.5082 mol
water and 15 mol CO2

39
 300
From the right to the left From the right to the left
288.15 K 288.15 K
250 293.15 K 293.15 K
298.15 K 298.15 K
304.19 K(No exp. data) 304.19 K
200 323.15 K 308.15 K
348.15 K 313.15 K
P / bar
373.15 K 323.15 K
150
348.15 K
100 H2O(l)+CO2(l)+CO2(g)
three phase lines

50

0.000 0.010 0.020 0.030 0.990 0.995 1.000

xCO2

Figure 2. 11 Pressure-composition phase diagram for the CO2-H2O system from 288.15 to 373.15 K, and
pressures up to 300 bar for both the aqueous phase and CO 2-rich phase. The circles stand for the
experimental values for the aqueous phase, and the triangles represent the experimental values for the CO 2-
rich phase. For the aqueous phase, the solid circles are the experimental data from Wiebe and Gaddy (1939)
at 323.15, 348.15 and 373.15 K. The gray circles are the experimental data from King et al. (1992) at
288.15, 293.15 and 298.15 K. The white circles are the experimental data taken from and Tong et al. (2013).
For the CO2-rich phase, the solid triangles are from Wiebe and Gaddy (1941) at 298.15, 304.19, 323.15 and
348.15 K. The gray triangles are the data from King et al. (1992) at 288.15, 293.15, 298.15, 308.15, and
313.15 K; the white triangle are the data from Valtz et al. (2004) at 288.15, 298.15, 308.15 K. The solid
lines are the results of model calculations from this study. The calculations show the well-known three-
phase [H2O(l)+CO2(l)+CO2(g)] coexistence lines where the system pressure lies below the CO 2 critical
point (73.77 bar, 304.19 K).

Figure 2. 11 shows the phase diagram of the CO2-H2O system at low temperatures (less than

373.15 K). The model accurately predicts the phase behavior of both the CO 2-rich and aqueous phases. In

addition, the three-phase (H2O(l)+CO2(g)+CO2(l)) coexistence pressure at temperatures below the CO2

critical temperature (304.19 K) can be obtained by the model (Figure 2. 11). For example, the calculated

three-phase coexistence pressure at 298.15 K is 65.89 bar, and this result is supported by the experimental

result (64.08 bar) of Wendland et al. (1999) at 298.15 K.

40
 90
80 Müller et al., 1988
This model
70
60

5K
K

413.1
15
P / bar
50

3.

K
47

5
K

3.1
15
40

3.

K
43
45

.15
30

393
20

K
15
10

3.
37
0
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
xCO2

Figure 2. 12 Phase diagram of the CO2-H2O system at temperatures 373.15-473.15 K and pressures 1-100
bar for the CO2-rich phase. The solid lines are our model calculations and points are from experimental
work of Müller et al. (1988).

600
5K

K
523.15 K

548.1

15
3.

500
57

548.15 K

400

P / bar
5K

300
523.1

200 573.15 K

100

0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
xCO2
Figure 2. 13 Comparison between calculations using SP2010 and the model of this study at temperatures
from 523.15-573.15 K, and pressures up to 600 bar for the CO 2-H2O system. Solid lines are our calculation
results for both the aqueous phase and CO2-rich phases, dash-dot lines are the calculation results from
Spycher and Pruess (2010). The circles are experimental data taken from Tödheide and Frank (1963). The
solid circles denote the CO2 solubility in the aqueous phase, whereas the open circles stand for the water
solubility in the CO2-rich phase. The triangles stand for the calculated CO2 solubility in the aqueous phase
at 523.15 K by DS2006.

41
 Figure 2. 12 shows the comparison of the model calculation with the experimental H2O solubility

in the CO2-rich phase (Müller et al., 1988) at the moderate temperatures of 373.15-473.15 K. There is very

good agreement between experimental and calculated values at the highest and lowest temperatures in this

interval, but deviations appear at temperatures between 433.15 and 453.15 K.

Figure 2. 13 shows a comparison of three models (this model, SP2010, and DS2006) with the

experimental data of Tödheide and Franck (1963) for both phases at temperatures of 523.15-573.15 K. Our

model and SP2010 capture the phase behavior of the CO 2-H2O system very well at pressures up to 500 bar

at 523.15 K, whereas our model provides more accurate results of H 2O solubility in the CO2-rich phase

than the SP2010 model at higher temperatures (e.g. 548.15 and 573.15 K).

Figure 2. 14(a, b) shows a comparison between the calculated values of this model and the

experimental CO2 solubility data from various sources (including 4 data points of this study) at 323.15,

373.15, 423.15 K and 1-1000 bar for the CO2-H2O system. Our measured CO2 solubility data in water fall

closely in line with the data from the literature, and the model developed in this study describes the phase

behavior of the CO2-H2O system with good accuracy. In Region I (P < PA, Figure 2. 14a), the CO2

solubility in the aqueous phase decreases as the temperature increases, until at point A (at pressure P A

=163.1 bar, calculated by this model), the CO2 solubility in the aqueous phase at two different temperatures

(373.15 and 423.15 K) reached the same value (mCO2 |A =1.051 mol/kg). As pressure increases to PB =256.9

bar (Region II, PA < P < PC), although no experimental data are available to validate this observation, the

CO2 solubility in the aqueous phase at two different temperatures (323.15 and 423.15 K) is the same

( mCO2 |B =1.370 mol/kg) as shown by the calculated results of this model (Figure 2. 14b). A similar

observation can be found at the pressure PC = 466.2 bar, where the CO2 solubility again is identical at

323.15 and 373.15 K (mCO2 |C =1.569 mol/kg). When the pressure is larger than P C, the solubility of CO2

increases, as shown by the ascending trend in Figure 2.14a (Region III, P > PC), as the temperature

increases. The behavior of the CO2 solubility in response to temperature increase in Region III is the

opposite of that in Region I (Figure 2. 14a).

42
 2.5 323.15 K 
373.15 K
423.15 K
2.0 This study
This model

-1
mCO2 / mol kg
1.5 B m2CO2
mCO2 C

1.0
m1CO2
A
0.5

 P
0.0
PA PB PC
1 10 100 1000
P / bar

(a)

PA=163.1 bar PB=256.9 bar PC=466.2 bar

1.6
Region 
K
15

C
3.

1.5
42

B
-1

1.4
mCO2 / mol kg

K
1.3 323.
15 3.15
37

1.2 323.15 K
373.15 K
423.15 K
1.1 This study
This model
A
1.0
150 200 250 300 350 400 450 500 550
P / bar

(b)

Figure 2. 14 The phase diagram of the CO2-H2O binary system at 323.15-423.15 K and 1-1000 bar. Solid
lines are bubble-point curves of CO2-saturated H2O(l) at different temperatures corresponding to the
experimental CO2 solubility data. Two vertical lines (i.e. 𝐏𝐀 and 𝐏𝐂 ) create three distinct regions (I, II and
III) in the diagram. In region I (P<PA), the solubility of CO2 decreases as the temperature increases; In
region II (PA<P<PC), the CO2 solubility decrease to a minimum then start to increase as the temperature
increases; In region III (P>PC), the CO2 solubility in aqueous phase is increasing as the temperature
increases. Region II is a transition zone in which at certain pressure (e.g. P B), the bubble-point curves of
CO2-saturated water with different temperatures may have intersections. Point A, B and C in the region II

43
are the intersections between those bubble-point curves at temperatures 323.15, 373.15 and 423.15 K at
pressures PA, PB and PC, respectively. The experimental data at 323.15, 373.15 and 423.15 K were taken
from the references (No. 1, 2, 5, 6, 7, 9, 10, 11, 13, 15, 16, 18, 21, 22, 26, 28, 32, 33, 35, 36, 37, 38, 39, 40,
and 41) listed in Table 2. 12. Figure 2.14b is an enlarged view of the region II, the values of PA, PB and PC
can be determined via model calculations as shown in Figure 2.14b.

While in region I (or III) the CO2 solubility in the aqueous phase is decrease (or increase)

monotonically in response to increased temperature, at a given pressure in region II, the CO2 solubility is

not a monotonic function of temperature. In fact, region II serves as a transition zone for the temperature-

dependent behavior of CO2 solubility in the aqueous phase between regions I and III. Therefore, we define

the pressure-bounded region II in the P-x phase diagram as the CO2 solubility “transition zone” at given

temperatures. The lower- and upper-boundary pressures of the CO2 solubility transition zone (i.e., Region

II) can be founded by the intersection points of the CO2 solubility curves at given temperatures in the P-x

phase diagram (Figure 2. 14b). For example, there are three temperature pairs (a) 373.15 and 423.15 K, (b)

323.15 and 423.15 K, and (c) 323.15 and 373.15 K corresponding to the intersection points A, B and C as

shown in Figure 2.14b. It is implied that the CO2 solubility curves (bubble point curves of the CO 2-

saturated water) with the highest temperature combination (e.g. pair (a)) tend to intersect at the lower

pressure boundary (PA) of the transition zone; whereas the CO2 solubility curves with the lowest

temperature combination (e.g. pair (c)) tend to intersect at the upper pressure boundary (P C) of the

transition zone.

Figure 2. 15(a-f) shows a comparison between this model and the MZLL2013 model. The

calculated results of those two models are compared with the reliable literature data (In Table 2. 12: Ref.

No. 3, 5, 6, and 7 for Figure 2. 15a; In Table 2.13: Ref. No. 2, 3, 5, 7, 10, 12 for Figure 2. 15b, Ref. No. 15

for Figure 2. 15c and Figure 2. 15d, Ref. No. 8, 9 for Figure 2. 15e, and Ref. No. 19 for Figure 2. 15f). Both

models are in excellent agreement with the experimental data. Figure 2. 15 also illustrates that the

performance between the two models are close to each other in the 1-1500 bar, 298.15-573.15 K and 0-4.5

mol/kg NaCl region, but the model developed in this study has a wider pressure and salt concentration

coverage, whereas the MZLL2013 model covers a wider temperature range. In addition to Figure 2. 15, we

presented the detailed comparison results between the two models in Table 2. 12 and Table 2. 13 for the

systems of CO2-H2O and CO2-NaCl-H2O, respectively.

44
 14
Takenouchi and Kennedy, 1964
Takenouchi and Kennedy, 1965
12 Tödheide and Frank, 1963
Malinin, 1959
10 MZLL2013
-1
This model
mCO2 / mol kg
8 964
edy, 1
573.15 K
enn
Other data K andK
6 .15 chi
573 enou
Tak 473.15 K
4

2 423.15 K

0
300 600 900 1200 1500
P / bar

(a)

0.040
Markham and Kobe, 1941
0.035 Harned and Davis, 1943
Yeh and Peterson, 1964
Onda, et al., 1970
-1

0.030 Yasunishi and Yoshida, 1979

mCO2 / mol kg

Burmakina et al., 1982

This model
0.025 MZLL2013

0.020

0.015
1.013 bar
298.15 K
0.010
0 1 2 3 4 5 6

mNaCl / mol kg-1

(b)

45
 0.7
This model
MZLL2013
0.6 313.15 K
333.15 K
0.5 353.15 K

-1
mCO2 / mol kg
393.15 K
0.4 413.15 K
433.15 K

0.3

0.2

0.1 4 mol/kg NaCl

Rumpf et al., 1994
0.0
0 20 40 60 80 100 120
P / bar

(c)

0.5
This model
313.15 K
0.4 333.15 K
353.15 K
393.15 K
-1
mCO2 / mol kg

413.15 K
0.3 433.15 K

0.2

0.1
6 mol/kg NaCl
Rumpf et al., 1994
0.0
0 20 40 60 80 100
P / bar

(d)

46
 1.4
48 bar This model
Malinin and Savelyeva, 1972 MZLL2013
1.2
Malinin and Kurovskaya, 1975 298.15 K
323.15 K

-1
1.0

mCO2 / mol kg
348.15 K
373.15 K
0.8 423.15 K

0.6

0.4

0.2

0.0
0 1 2 3 4 5 6
-1
mNaCl / mol kg

(e)

1.4
0.529 mol/kg NaCl
Bando et al., 2003
1.3
-1

1.2
mCO2 / mol kg

1.1

303.15 K
1.0
313.15 K
323.15 K
0.9 333.15 K
This model
MZLL2013
0.8
80 100 120 140 160 180 200 220
P / bar

(f)

Figure 2. 15(a-f) Vapor-liquid equilibrium of the CO2-H2O and CO2-NaCl-H2O systems. Comparison of
this model and MZLL2013 (Mao et al., 2013) to experimental CO2 solubility data in aqueous phase.
Results are shown in (a) for the CO2-H2O system at 423.15-573.15 K, 1-1500 bar; (b) for the CO2-NaCl-
H2O system at ambient conditions, (c) for the CO2-NaCl-H2O system at 313.15-433.15 K, 1-120 bar and 4
mol kg-1 NaCl, (d) for the CO2-NaCl-H2O system at 313.15-433.15 K, 1-120 bar and 6 mol kg-1 NaCl (the
MZLL2013 model is not applicable in this case due to the high NaCl molality), (e) for the CO 2-NaCl-H2O
system at 298.15-423.15 K, 48 bar, and 0-6 mol kg-1 NaCl, and (f) for the CO2-NaCl-H2O system at
303.15-333.15 K, 80-220 bar, at 0.529 mol kg-1 NaCl.

47
 In Table 2. 12 and Table 2. 13, the calculated results of the two models (This model and

MZLL2013) are compared with available experimental CO2 solubility data for the CO2-H2O and CO2-

NaCl-H2O systems at elevated temperatures and pressures. The average absolute deviation (AAD%) of the

calculated CO2 solubility values of this model from the experimental data is 3.82 % and 3.83% for the CO 2-

H2O (Table 2. 12) and CO2-NaCl-H2O (Table 2. 13) systems, whereas the results from the MZLL2013

model (AAD%) are 4.18% and 4.11%, respectively. Therefore, the model developed in this work provides

reliable estimates of the CO2 solubility in NaCl(aq) solutions as well as the H2O solubility in the CO2-rich

phase of the CO2-NaCl-H2O system over a wide P-T-x range (273.15-573.15 K, 1-2000 bar, 0-6 mol/kg

NaCl(aq)).

48
Table 2. 12 The average absolute deviation (AAD %) of calculated CO2 solubility (mol kg-1) in aqueous phase from experimental data for the CO2-H2O system
at temperatures of 273.15-573.15 K and pressures of 1-2000 bar: Comparison between this model and the MZLL2013 model.

No. References P / bar T/K N AAD % AAD %(e)

(This model) (MZLL2013)
1 Wiebe and Gaddy (1939) 25-709 323.15-373.15 29 1.82 2.72
2 Prutton and Savage (1945) 23-654 373.15-393.15 26 5.27 6.96(f)
3 Malinin (1959) 98-490 473.15-573.15 16(a) 5.30 2.28
4 Ellis and Golding (1963) 25-200 450.15-538.15 10 9.16 3.50
5 Tödheide and Franck (1963) 200-2000 323.15-573.15 78 1.51 7.03(4.51)(b)
6 Takenouchi and Kennedy (1964) 100-1500 383.15-573.15 96 7.06 5.28(f)
7 Takenouchi and Kennedy (1965) 100-1400 423.15-573.15 35 4.90 5.18
8 Stewart and Munjal (1970) 10.1-45.6 273.15-298.15 12 11.06 10.85
9 Malinin and Savelyeva (1972) 48 298.15-348.15 3(a) 1.51 1.53(f)
10 Malinin and Kurovskaya (1975) 48 298.15-423.15 3(a) 1.60 3.24(f)
11 Drummond (1981) 40-180 303.15-573.15 50 5.00 9.01(f)
12 Cramer (1982)(c) 8-58 306.15-486.25 7 3.89 7.28(c)
13 Briones et al. (1987) 68-177 323.15 7 3.12 2.30
14 Nakayama et al. (1987) 36.3-109.9 298.25 6 4.06 4.11
15 D'Souza et al. 1988 101-152 323.15-348.15 4 4.73 5.07
16 Müller et al. (1988)(d) 3-81 373.15-473.15 49 3.11 3.27
17 Nighswander et al. (1989) 23-102 353.15-473.15 33 5.40 6.16
18 Sako et al. (1991) 100-200 348.15-423.15 7 6.79 7.31
19 King et al. (1992) 60-243 288.15-298.15 27 1.87 1.40
20 Mather and Franck (1992) 1614-1692 500.15-533.15 6 5.45 -
21 Dohrn et al. (1993) 101-301 323.15 3 1.87 1.16
22 Rumpf et al. (1994) 11-58 323.15 7 0.65 1.64(f)
23 Teng et al. (1997) 64.4-294.9 278.05-293.05 24 4.29 3.85
24 Gu (1998) 17.1-58.3 304.19-313.15 10 2.78 3.20
25 Dhima et al. (1999) 100-1000 344.15 7 5.37 5.03
26 Bamberger et al. (2000) 40-140 323.15-353.15 29 1.61 1.45
27 Teng and Yamasaki (2002) 75-300 298.05 6 0.44 0.89
28 Bando et al. (2003) 100-200 303.15-333.15 12 2.31 2.76
29 Chapoy et al. (2004) 1.9-93.3 273.85-351.35 27 2.76 3.02
30 Valtz et al. (2004) 5-80 278.15-318.15 47 4.19 4.65
31 Li et al. (2004) 33.4-198.9 332.15 6 3.09 7.24
32 Koschel et al. (2006) 20.6-194.7 323.15-373.15 8 2.97 3.72

49
No. References P / bar T/K N AAD % AAD %(e)
(This model) (MZLL2013)
33 Qin et al. (2008) 106-499 323.65-375.75 7 3.07 3.77
34 Ferrentino et al. (2010) 75-150 313.05 4 2.41 0.87
35 Han et al. (2011) 3.8-20.2 313.05-333.05 15 5.71 5.69(f)
36 Liu et al. (2011) 20-160 308.15-328.15 31 1.70 1.67
37 Yan et al. (2011) 50-400 323.15-413.15 18 2.33 3.39(f)
38 Hou et al. (2013) 25-175 283.15-448.15 42 7.74 7.49(f)
39 Tong et al. (2013) 72-273 374.15 5 2.08 5.06(f)
40 Serpa et al. (2013) 1.3-4.1 298.15-323.15 9 6.04 5.32(f)
41 This work 100-150 323.15-423.15 4 0.39 0.76(f)
Overall 825 3.81 4.18(4.11)
N is the number of data points evaluated by this model.
(a)
The repeated measurements were counted as 1 datum.
(b)
AAD=4.51% is calculated by Mao et al. (2013) using 39 experimental data points given by Tödheide and Franck (1963). We report the calculate AAD=7.03
for the MZLL2013 model by using 69 experimental data points from Tödheide and Franck (1963) within the P-T range of 1-1500 bar and 273.15-573.15 K.
(c)
Cramer's (1982) data are compared with our model calculated values using a modified Henry's law constant defined by 𝑘 = 𝜑𝐶𝑂2 𝑃𝐶𝑂2 /𝑥𝐶𝑂2 .
(d)
The experimental data were taken from a large data compilation given by Scharlin (1996).
(e)
Taken from Mao et al. (2013).
(f)
Calculated using the MZLL2013 model.

50
Table 2. 13 The average absolute deviation (AAD %) of the calculated CO2 solubility (molality) in the aqueous phase from experimental data for the CO2-NaCl-
H2O system at 273.15-573.15 K, 1-1400 bar and 0-6 mol kg-1 NaCl: Comparison between this model and the MZLL2013 model.

No. References P / bar T/K 𝐦𝐍𝐚𝐂𝐥 ⁡/ mol kg-1 N AAD % AAD %

(This model) (MZLL2013)
1(a) Setchenow (1892) 1.013 288.35 0.225-6.088 9 3.47 2.09(e)
2 Markham and Kobe (1941) 1.013 273.35-313.15 0.1-4 15 3.71 3.66(f)
3 Harned and Davis (1943) 1.013 273.15-323.15 0.2-3 55 2.65 3.14(f)
4 Ellis and Golding (1963) 54-210.1 449.15-603.15 0.5-2 29 4.19 4.48(f)
5 Yeh and Peterson (1964) 1.013 298.15-318.15 0.15 4 1.69 5.83(f)
6 Takenouchi and Kennedy (1965)(b) 100-1400 423.15-673.15 1.09-4.28 55 13.81(b) 5.18(f)
7 Onda, et al. (1970) 1.013 298.15 0.51-3.21 8 1.72 2.90(f)
(c)
8 Malinin and Savelyeva (1972) 48 298.15-348.15 0.36-4.46 13 1.57 1.19(e)(4.62)(f)
(c)
9 Malinin and Kurovskaya (1975) 48 298.15-423.15 1-5.91 13 2.18 2.20(e)(7.18)(f)
(c)
10 Yasunishi and Yoshida (1979) 1.013 288.15-308.15 0.45-5.73 27 1.82 2.09(e)(6.63)(f)
11 Drummond (1981) 34-387.6 292.85-673.15 1-6.3 339 5.73 6.62(f)
(a)
12 Burmakina, et al. (1982) 1 298.15 0.001-0.2 8 1.34 1.67(g)
(d)
13 Cramer (1982) 8-62 296.75-511.75 0.5-1.95 13 9.55 12.74(c)
14 Nighswander et al. (1989) 21-100 353.15-473.65 0.173 34 5.56 6.28(f)
15 Rumpf et al. (1994) 4.67-96.37 313.15-433.15 4-6 63 1.62 2.09(f)
(a)
16 Vázquez et al. (1994) 1.013 298.15-308.15 0.69-2.9 20 1.52 2.78(f)
17 Gu (1998) 17.7-59.0 303.15-323.15 0.5-2.0 60 6.55 5.66(f)(4.05)(f)
18 Kiepe et al. (2002) 1-100.61 313.15-353.15 0.52-4.34 64 5.15 5.78(f)
19 Bando et al. (2003) 100-200 303.15-333.15 0.17-0.53 36 1.02 2.30(f)
20 Koschel et al. (2006) 51-202.4 323.15-373.15 1-3 14 5.19 4.50(f)
21 Kim et al. (2008) 101-201 278.9-280.4 1 6 4.87 7.09(e)
22 Ferrentino et al. (2010) 100-150 313 0.17 2 2.57 2.89(f)
23 Liu et al. (2011) 21-158.3 318.15 1.9 8 3.62 4.84(f)
24 Yan et al. (2011) 50-400 323.15-413.15 1-5 36 3.80 3.73(f)
(d)
25 This study 150 323.15-423.15 1-6 18 0.78 1.06(e)
Overall 949 3.83 4.11(4.62)
N is the number of evaluated data points for each experimental work. (a)Experimental data taken from a large data compilation given by Scharlin (1996); (b)The
experimental data reported by Takenouchi and Kennedy (1965) has a large uncertainty when compared with our model; (c)Cramer's (1982) experimental CO2
solubility is expressed in terms of a modified Herny's constant therefore the value computed from Mao et al.'s (2013) model could not be used for a direct
comparison to the experimental data given by Cramer (1982). We have converted Cramer's data to molality using our model and then make comparison with the
calculated results of Mao et al.'s (2013) results; (d)The number differs from Table 2. 3 due to the removal of the computed results for the CO2-H2O system;
(e)
Calculated by using the MZLL2013 model; (f)Taken from Mao et al. (2013).

51
2.5 Conclusions

A new PVT apparatus to measure the CO2 solubility in aqueous solutions at temperatures and

pressures corresponding to CO2 storage conditions has been constructed and shown to provide accurate and

reproducible CO2 solubility measurements at temperatures up to 423.15 K and pressures up to 150 bar. At

150 bar, new CO2 solubility data in the aqueous phase were obtained for the CO2-NaCl-H2O system at

temperatures of 323.15, 373.15, and 423.15 K and from 0 to 6 mol kg-1 NaCl(aq). Error analysis showed

that the instrumental error of the observed CO2 solubility is around 0.7 % and the random error is from 0.5

to 4.5 % based on the molal concentration scale.

In Chapter 2, we improved upon the previously published model of Spycher and Pruess (2010) by

including Pitzer's model to calculate the activity coefficient of the dissolved CO 2 and the osmotic

coefficient of water in the CO2-NaCl-H2O system. The new model can accurately predict the CO2 solubility

in the aqueous phase for both the CO2-H2O and CO2-NaCl-H2O systems with an overall accuracy better

than 3.9 % at temperatures from 273.15 to 573.15 K and pressures up to 2000 bar. A comparison between

our model and four widely accepted models (AD2010, SP2010, DS2006, and OLI), as well as a recent

published model MZLL2013, showed that our model provides accurate predictions of CO 2-H2O mutual

solubilities for the CO2-H2O and CO2-NaCl-H2O systems.

A comparison of our model results at three different temperatures (323.15, 373.15 and 423.15 K)

with a large number of experimental data reveals a CO 2 solubility transition zone, which is bounded by a

low and a high pressure boundary (P low and PHigh, respectively). In the transition zone (Plow < P < PHigh), the

CO2 solubility decreases to a minimum before increasing as the temperature increases. The CO 2 solubility

is not a monotonic function of temperature in the transition zone. Conversely, outside of that transition

zone, the CO2 solubility decreases (P < Plow) or increases (P > PHigh) in response to increased temperature.

This phenomenon is clearly observed and validated by both experimental data and modeling results for the

CO2-H2O systems.

52
 CHAPTER 3

SOLUBILITY OF CO2 IN AQUEOUS SOLUTIONS OF CaCl2, MgCl2, Na2SO4

AND KCl3

3
The text for this chapter was originally prepared for the publication as “Haining Zhao, Robert Dilmore,
and Serguei N. Lvov, Experimental studies and modeling of CO2 solubility in high temperature aqueous
CaCl2, MgCl2, Na2SO4, and KCl solutions, Geochimica et Cosmochimica Acta, in review”

53
Abstract

New experimental CO2 solubility data in aqueous CaCl2, MgCl2, Na2SO4 and KCl solutions were

collected at temperatures 323.15, 373.15, and 423.15 K, pressure of 150 bar, and ionic strengths from 0 to 6

mol kg-1 of dissolved salt species. This study experimentally demonstrates that CO 2 solubility in aqueous

solutions with different salt species are significantly different, whether the solutions have the same percent

of salt by weight or have the same ionic strength. The experimental results also show that in the aqueous

phase a cation with a higher charge density will have more significant salting-out effect on dissolved CO2

than other ions. In addition, the measured CO2 solubility data are used to extend the capability of a

previously developed CO2 solubility model (for the CO 2-NaCl-H2O system) to account for aqueous

solutions containing salt species other than NaCl. Comparisons against literature data reveal a clear

improvement of the proposed PSUCO2 model among the published models in predicting CO2 solubility in

aqueous CaCl2, MgCl2, Na2SO4 and KCl solutions. Also, the modeling results show a positive correlation

between the H2O solubility in the CO2-rich phase and the activity of H2O in the aqueous phase, the

influence of salts species on solubility of H2O in the CO2-rich phase is clearly observed. A web-based CO2

solubility computational tool can be accessed via the link: www.carbonlab.org/psuco2/.

54
3.1 Introduction

CO2 capture and sequestration (CCS) in deep geological formations has emerged as an important

option for reducing CO2 emissions from industrial sectors (Benson and Cole, 2008). Important mechanisms

for CO2 trapping in geologic sequestration systems include: structure/stratigraphic and capillary (pore-scale)

trapping of dense-phase CO2, solubility trapping, and mineral trapping. These physical and chemical

processes due to the phase transition of the CO2-brine system at geological temperatures and pressures have

significant influences on formation fluid and rock properties because of 1) solubility trapping and its

influences on mineral deposition; 2) variations of the buoyancy of the CO 2-rich phase and the density in the

aqueous phase due to the dissolution of CO 2 in brine (Benson and Cole, 2008; Haugan and Drange, 1992;

Duan et al., 2008); 3) the swelling and shrinking effects of reservoir brine as a function of CO 2 saturation

(Bachu and Adams, 2003; Steele-MacInnis et al., 2013); 4) the wettability alteration of the brine due to the

dissolution of CO2 (Yang et al., 2005); and 5) the formation dry-out effect due to H2O evaporates into the

CO2-rich phase (Pruess and Müller, 2009). Therefore, the CO2 and H2O mutual solubilities are the critical

parameters for understanding these trapping mechanisms. Developing a robust characterization of CO 2

solubility in formation brine can help to improve evaluations of storage potential of a reservoir and to

identify the risks associated with long term storage by performing reservoir simulation studies (Pruess et al.,

2001; Kumar et al., 2005; Bickle et al., 2007).

The solubility of CO2 in NaCl(aq) solutions at elevated pressures has been widely studied from 1-

1400 bar and 273.15-673.15 K (Springer et al., 2012; Zhao et al., 2014a). However, CO2 solubility in other

salt solutions, such as CaCl2(aq), Na2SO4(aq), MgCl2(aq), and KCl(aq), has not been fully understood,

especially for the CO2-Na2SO4-H2O and CO2-MgCl2-H2O systems. High quality experimental data are still

necessary in order to develop a reliable CO2 solubility model for these CO2-salt-H2O systems.

Three available models capable of calculating CO2 solubility in CO2-salt-H2O systems containing

salt species other than NaCl were considered: (1) OLI: OLI Studio 9.0.6 (a commercial software developed

by OLI Systems Inc.) with its mixed-solvent electrolyte (MSE) thermodynamic framework (Springer et al.,

2012); (2) SP2010: the model developed by Spycher et al.(2003) and modified by Spycher and Pruess

(2010), and (3) DS2006: the model developed by Duan and Sun (2003) and improved by Duan et al. (2006).

55
 Each of these models has its own advantages and disadvantages in calculating CO 2 solubility in

aqueous salt solutions other than NaCl. OLI Studio 9.0.6 exhibits an excellent performance in predicting

H2O solubility in the CO2-rich phase (Zhao et al., 2014a), and also performs well in predicting CO2

solubility in NaCl(aq) and KCl(aq) brines. However, OLI Studio 9.0.6 shows a relatively larger error in

predicting CO2 solubility in CaCl2(aq) and MgCl2(aq) than other models (SP2010; DS2006) based on

comparison to previously published data (See section 3.4.2, Table 3. 4). The model DS2006 provides a very

good result in calculating CO2 solubility in NaCl(aq), CaCl2(aq), and MgCl2(aq) at temperatures from

273.15-533.15 K, pressures up to 2000 bar, and salt molality up to 4.5 mol kg -1, but it does not provide the

calculation of H2O solubility in the CO2-rich phase. Moreover, the CO2 solubility values calculated using

the DS2006 model in aqueous Na2SO4 and KCl solutions, as shown herein, significantly deviate from new

experimental values.

The SP2010 model has an excellent performance on predicting H 2O solubility in the CO2-rich

phase. As for CO2 solubility in aqueous phase, the performance between the models SP2010 and DS2006 is

similar. However, the applicable pressure range of the model SP2010 is limited to 600 bar and, similar to

the model DS2010, it is also incapable of accurately computing CO 2 solubility in aqueous solutions of

Na2SO4 and KCl (See section 3.4.2, Table 3. 4). Therefore, there is a need for a new model to accurately

estimate mutual solubilities of H2O and CO2 for various CO2-salt-H2O systems in geological CO2 storage

conditions.

Our previously developed model (Zhao et al., 2014a) for the CO2-NaCl-H2O system is based on

the thermodynamic framework given by Spycher et al. (2003), Spycher and Pruess (2010) and Akinfiev and

Diamond (2010). Akinfiev and Diamond's (2010) approach to estimate Pitzer parameters (𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 and

𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 ) is proven successful for the CO2-NaCl-H2O system. Therefore, this work was motivated by

Akinfiev and Diamond's (2010) approach and set out to extend the model (Zhao et al., 2014a) to other CO2-

salt-H2O systems. By doing so, the combined Pitzer interaction parameters 𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 and 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡

were determined using experimental CO2 solubility for each CO2-salt-H2O system. In addition, the new

experimental CO2 solubility values were used to evaluate the triple-ion interaction parameters (𝜉𝑛𝑐𝑎 ) for

each CO2-salt-H2O system. As a result, the previously developed CO 2 solubility model for the CO2-NaCl-

56
H2O system was successfully extended to the aqueous salt solutions of CaCl2, Na2SO4, MgCl2 and KCl.

The updated model is named as PSUCO2.

3.2 Experimentation

The experimental approaches and procedures used herein for measuring CO 2 solubility in aqueous

solutions at elevated temperatures and pressures were described and validated in Chapter 2 (Table 2. 1 and

Table 2. 2). The carbon dioxide used in all experiments was Coleman Instrument grade with a purity of

99.99 %. Water was purified by a Milli-Q system and was degassed before being loading into the autoclave.

The purified water conductivity was below 610-6 S m-1. All aqueous salt solutions were prepared using

this Milli-Q water and ACS Grade reagents: calcium chloride (CaCl2∙2H2O, Alfa Aesar, 99%), sodium

sulfate (Na2SO4, Amresco, 99%), magnesium chloride (MgCl2, Alfa Aesar, 99%), and potassium chloride

(Alfa Aesar, 99%). The experimental system consisted of a 600-ml stainless steel autoclave (Parr

Instrument Co.), a 40-ml stainless steel sample cell, a liquid CO2 pump and a 300-ml stainless steel

pressure cell for sample analysis. Details on the CO2 solubility measuring technique and the error analysis

approach can be found in Section 2.3, Chapter 2.

3.3 Theoretical essentials for PSUCO2

The thermodynamic framework for vapor-liquid-phase equilibrium (VLE) calculation was

described in Chapter 2 and the previous publications (Spycher et al., 2003; Spycher and Pruess; 2010). In

Chapter 3, the previously developed model for the CO2-NaCl-H2O system given by Chapter 2 has been

extended to aqueous systems containing other salt species: CaCl2, Na2SO4, MgCl2, or KCl. The Pitzer

equations for the activity coefficient of CO2 and the osmotic coefficient of water are shown below as Eqs.

(3.1) and (3.2), respectively (Pitzer, 1991).

𝑙𝑛𝛾𝐶𝑂2

2
= 2𝑚𝐶𝑂2 𝜆𝑛𝑛 + 3𝑚𝐶𝑂 𝜇
2 𝑛𝑛𝑛
+ 2𝜆𝑛𝑐 𝑚𝑐 + 2𝜆𝑛𝑎 𝑚𝑎 + 𝑚𝑎 𝑚𝑐 𝜉𝑛𝑐𝑎 + 6𝑚𝐶𝑂2 𝑚𝑐 𝜇𝑛𝑛𝑐 + 6𝑚𝐶𝑂2 𝑚𝑎 𝜇𝑛𝑛𝑎 (3.1)

57
 2 𝐼 1.5 (0) (1)
𝜙−1= {−𝐴𝜙 + 𝑚𝑎 𝑚𝑐 [𝛽𝑐𝑎 + 𝛽𝑐𝑎 𝑒𝑥𝑝(−𝛼1 𝐼 0.5 ) + (𝑚𝑎 |𝑧𝑎 | + 𝑚𝑐 |𝑧𝑐 |)𝐶𝑐𝑎 ] +
𝑚𝑎 +𝑚𝑐 +𝑚𝐶𝑂2 𝐼+𝑏𝐼 0.5

1 2 3 2 2
⁡ 𝑚𝐶𝑂 𝜆 + 𝑚𝐶𝑂
2 𝑛𝑛
𝜇
2 𝑛𝑛𝑛
+ 𝑚𝐶𝑂2 𝑚𝑐 𝜆𝑛𝑐 + 3𝑚𝐶𝑂2
𝑚𝑐 𝜇𝑛𝑛𝑐 + 𝑚𝐶𝑂2 𝑚𝑎 𝜆𝑛𝑎 + 3𝑚𝐶𝑂2
𝑚𝑎 𝜇𝑛𝑛𝑎 +
2

𝑚𝐶𝑂2 𝑚𝑎 𝑚𝑐 𝜉𝑛𝑐𝑎 } (3.2)

In these equations, 𝑚𝐶𝑂2 is the molality of CO2; 𝑚𝑎 ⁡and 𝑚𝑐 are, respectively, the molality of

𝜙
anions and cations in the aqueous phase. In Eq. (3.2), 𝐶𝑐𝑎 = 0.5𝐶𝑐𝑎 /|𝑧𝑐 𝑧𝑎 |0.5 , 𝐴𝜙 is the Debye-Hückel

slope for the osmotic coefficient, and 𝑏 = 1.2 kg1/2 mol-1/2, 𝛼1 = 2.0 kg1/2 mol-1/2. I is the ionic strength

based on the molality scale. The empirical equations for calculating the Pitzer binary ion–ion interaction

(0) (1)
parameters (𝛽𝑐𝑎 , 𝛽𝑐𝑎 and 𝐶𝑐𝑎 ) were taken from literature and summarized in Appendix B. The Pitzer pure

CO2 interaction parameters (𝜆𝑛𝑛 and 𝜇𝑛𝑛𝑛 ) in the temperature range 273.15-573.15 K can be found in Eq.

(2.13) and Table 2. 6.

According to Akinfiev and Diamond (2010), the combined Pitzer interaction parameters

(𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 and 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 ) can be written for any salt species as:

𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 = 𝜈 + 𝜆𝑛𝑐 + 𝜈 − 𝜆𝑛𝑎 (3.3)

𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 = 𝜈 + μ𝑛𝑛𝑐 + 𝜈 − 𝜇𝑛𝑛𝑎 (3.4)

Subsequently, the Eq. (3.1) can be simplified as

𝑙𝑛𝛾𝐶𝑂2

2 2
= 2𝑚𝐶𝑂2 𝜆𝑛𝑛 + 3𝑚𝐶𝑂 𝜇
2 𝑛𝑛𝑛
+ 2𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 𝑚𝑠𝑎𝑙𝑡 + 𝜈 + 𝜈 − 𝑚𝑠𝑎𝑙𝑡 𝜉𝑛𝑐𝑎 + 6𝑚𝐶𝑂2 𝑚𝑠𝑎𝑙𝑡 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 (3.5)

The combined Pitzer interaction parameters 𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 and 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 in Eq. (3.5) were

determined using experimental CO2 solubility data for the CO2-H2O and CO2-salt-H2O systems. In order to

58
evaluate those two parameters, we have made an assumption that the chemical potentials of CO2 between

the CO2-H2O and CO2-salt-H2O systems are the same at a given P-T condition based on the similar

approach used by Akinfiev and Diamond (2010). At equilibrium, the chemical potential of CO2 in the CO2-

(𝑔) (𝑎𝑞) (𝑔)′

rich phase is equal to that in the aqueous phase, given by 𝜇𝐶𝑂2 = 𝜇𝐶𝑂2 for the CO2-H2O system and 𝜇𝐶𝑂2 =

(𝑎𝑞)′
𝜇𝐶𝑂2 for the CO2-salt-H2O system. In fact, the chemical potentials of CO2 between salt-bearing and salt-

free systems are different, due to a small variation of water solubility in the CO2-rich phase between the

(𝑔) (𝑔)′
two systems. However, assuming that ⁡𝜇𝐶𝑂2 = 𝜇𝐶𝑂2 corresponds to a mean error of about 0.3 % of

calculated CO2 solubility in the aqueous phase (Akinfiev and Diamond, 2010), so this assumption is not

(𝑔) (𝑔)′
significantly impactful to the overall model results. Thus, from 𝜇𝐶𝑂2 = 𝜇𝐶𝑂2 , the equality of CO2 activity

between the CO2-H2O and CO2-salt-H2O systems can be described as follows:

𝑜
𝑙𝑛(𝑚𝐶𝑂2 𝛾𝐶𝑂2 ) = 𝑙𝑛(𝑚𝐶𝑂 𝛾𝑜 )
2 𝐶𝑂2
(3.6)

𝑜
𝑚𝐶𝑂 2 𝑜
𝑙𝑛 ( ) = 𝑙𝑛𝛾𝐶𝑂2 − 𝑙𝑛𝛾𝐶𝑂2
(3.7)
𝑚𝐶𝑂2

In Eqs. (3.6) and (3.7), 𝑚𝐶𝑂2 and 𝛾𝐶𝑂2 are the solubility of CO2 (mole CO2 per kilogram of H2O)

𝑜 𝑜
and the activity coefficient of CO2 in aqueous salts solutions, respectively; whereas 𝑚𝐶𝑂2
and 𝛾𝐶𝑂2
are

those in pure water. Eq. (3.5) can be reduced to Eq. (3.8) by setting the parameters 𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 , 𝜉𝑛𝑐𝑎 and

𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 equal to zero for the CO2-H2O binary system as below,

𝑜 𝑜 𝑜
𝑙𝑛𝛾𝐶𝑂2
= 2𝑚𝐶𝑂 𝜆 + 3(𝑚𝐶𝑂
2 𝑛𝑛 2
)2 𝜇𝑛𝑛𝑛 (3.8)

Substituting Eqs. (3.5) and (3.8) into Eq. (3.7), we obtain:

𝑜
𝑚𝐶𝑂 2
𝑙𝑛 ( ) − 𝐶 = 2𝑚𝑠𝑎𝑙𝑡 𝐹 (3.9)
𝑚𝐶𝑂2

59
where C and F in Eq.(3.9) are given by Eqs. (3.10) and (3.11), respectively.

𝑜 2 𝑜
𝐶 = ⁡2𝜆𝑛𝑛 (𝑚𝐶𝑂2 − 𝑚𝐶𝑂2
) + 3𝜇𝑛𝑛𝑛 (𝑚𝐶𝑂2
− (𝑚𝐶𝑂2
)2 ) (3.10)

𝑣+𝑣 −
𝐹 = ⁡ 𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 + 𝑚𝑠𝑎𝑙𝑡 𝜉𝑛𝑐𝑎 + 3𝑚𝐶𝑂2 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 (3.11)
2

And, from Eq. (3.9)

𝑜
𝑚𝐶𝑂
𝐹 = (𝑙𝑛 ( 2
) − 𝐶) /(2𝑚𝑠𝑎𝑙𝑡 ) (3.12)
𝑚𝐶𝑂2

By equating Eqs. (3.11) and (3.12), we obtain

𝑚𝑜
𝐶𝑂2
(𝑙𝑛( )−𝐶)
𝑣 +𝑣− 𝑚𝐶𝑂
2
𝐹 = 𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 + 𝑚𝑠𝑎𝑙𝑡 𝜉𝑛𝑐𝑎 + 3𝑚𝐶𝑂2 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 = (2𝑚𝑠𝑎𝑙𝑡 )
(3.13)
2

The parameter C was calculated by Eq. (3.10) using experimental CO2 solubilities data in both the

CO2-H2O and CO2-salt-H2O systems. In order to determine the combined Pitzer interaction parameters

𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 and 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 in Eq. (3.13), a function for F suggested by Akinfiev and Diamond (2010) was

chosen for the data fitting process as shown in Eq. (3.14).

100 𝑇 𝑣+𝑣 −
𝐹 = 𝑎1 + 𝑎2 + 𝑎3 + 𝑎4 𝑚𝑠𝑎𝑙𝑡 + 𝑎5 3𝑚𝐶𝑂2 + 𝑎6 𝑔(𝑥) (3.14)
𝑇−𝜃 1000 2

Comparing Eqs. (3.13) and (3.14), we obtain

100 𝑇
𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 = 𝑎1 + 𝑎2 + 𝑎3 + 𝑎6 𝑔(𝑥) (3.15)
𝑇−𝜃 1000

𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 = 𝑎5 (3.16)

60
 2
where 𝜃=228 K, 𝑔(𝑥) = (1 − (1 + 𝑥)𝑒 −𝑥 ) , 𝑥 = 𝛼1 𝐼 0.5 , and 𝛼1 = 2.0 kg0.5 mol-0.5. The parameters
𝑥2

𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 and 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 were determined from experimental CO2 solubility data for each CO2-salt-H2O

system, the data source are listed in Section 3.4.2, Table 3. 4. Suppose there are n sets of experimental CO2

solubility data available at different P-T-x points, the Eq. (3.14) can then be written in matrix form as:

𝐴𝑥 = 𝑓 (3.17)

where

100 𝑇 (1) 𝑣+𝑣 − (1) (1)

1(1) 𝑚𝑠𝑎𝑙𝑡 3𝑚𝐶𝑂2 𝑔(1) (𝑥)
𝑇 (1)−𝜃 1000 2
100 𝑇 (2) 𝑣+𝑣 − (2) (2)
1(2) 𝑚𝑠𝑎𝑙𝑡 3𝑚𝐶𝑂2 𝑔(2) (𝑥)
𝑇 (2)−𝜃 1000 2
𝐴= . . .
.. .. .. ;
. . .
. . . . . .
100 𝑇 (𝑛) 𝑣 +𝑣 − (𝑛) (𝑛)
(1
(𝑛) 𝑚𝑠𝑎𝑙𝑡 3𝑚𝐶𝑂2 𝑔(𝑛) (𝑥))
𝑇 (𝑛) −𝜃 1000 2 𝑛×6

𝑎1
𝐹 (1)
𝑎2
𝐹 (2)
𝑎3
𝑥= 𝑎 ;𝑓= ..
4
𝑎5 .
(𝑎6 )6×1 (𝐹 (𝑛) )𝑛×1

The least-squares solution (𝑥̂) (Schay, 2012) for Eq. (3.17) is

[𝐴𝑇 𝐴]6×6 [𝑥̂]6×1 = [𝐴𝑇 𝑓]6×1 (3.18)

where 𝐴𝑇 is the transpose of matrix A. The parameters (𝑎1 ~𝑎6 )⁡for each CO2-salt-H2O system are obtained

by solving the linear Eqs. (3.18), the results are shown in Table 3. 1. Thus, the parameters 𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 and

𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 can be calculated by Eqs. (3.15) and (3.16).

61
Table 3. 1 Coefficients of Eqs. (3.15) and (3.16) for calculating 𝑩𝑪𝑶𝟐 −𝒔𝒂𝒍𝒕 and 𝑪𝑪𝑶𝟐 −𝑪𝑶𝟐 −𝒔𝒂𝒍𝒕 as a function of temperature.

Salt species* a1 a2 a3 a4 a5 a6
CaCl2(a) -2.3219 10-1 1.3069 10-1 8.0693 10-1 -5.6456 10-3 7.0156 10-3 4.9642 10-2
Na2SO4 (b) 1.2396 10-1 1.1765 10-1 2.0199 10-1 -2.3589 10-2 -1.7424 10-3 1.4198 10-3
(c) -2 -2 -1 -4 -4
MgCl2 -5.0410 10 2.2742 10 2.3521 10 -7.9747 10 -2.8093 10 1.1069 10-1
(c) -1 -2 -1 -3 -3
KCl -1.8409 10 6.2729 10 5.0722 10 -4.4679 10 -2.6025 10 8.7653 10-2
(a)
For the CO2-CaCl2-H2O sytem, the parameters 𝑎1 to 𝑎6 were determined from the experimental data given by: Setchenow (1892), Prutton and Savage (1945),
Onda et al. (1970), Malinin and Savelyeva (1972), Malinin and Kurovskaya (1975), Yasunishi and Yoshida (1979a), Liu et al. (2011), and this work.
(b)
For the CO2-Na2SO4-H2O system, the parameters 𝑎1 to 𝑎6 were determined from the experimental data given by Yasunishi and Yoshida (1979a), Corti et al.
(1990), Rumpf and Maurer (1993), Bermejo et al. (2005), and this work.
(c)
For the CO2-MgCl2-H2O system, the parameters 𝑎1 to 𝑎6 were determined from the experimental data reported by Yasunishi and Yoshida (1979a) and this
work.
(d)
For the CO2-KCl-H2O system, the parameters a1 to a 6 were determined from the experimental data given by Setchenow (1892), Geffcken (1904), Findley and
Shen (1912), Markham and Kobe (1941), Gerecke (1969), Yasunishi and Yoshida (1979a), Burmakina et al. (1982), Kiepe et al. (2002), Pérez-Salado Kamps et
al. (2007); Liu et al. (2011) and this work.

62
 The Pitzer triple-ion interaction parameter 𝝃𝒏𝒄𝒂 in Eqs. (3.1) and (3.2) were then calculated by

performing a binary search for 𝒎𝑪𝑶𝟐 until consistency between calculated and experimental values was

obtained. The results of 𝝃𝒏𝒄𝒂 determined in that manner are listed in Table 3. 2. The parameter 𝝃𝒏𝒄𝒂 is

shown to have a strong dependence on temperature and salt concentration, especially at high temperatures

and low salt concentrations, but is insensitive to system pressure. Finally, by using the calculated results of

(𝟎) (𝟏) 𝝓
𝜷𝒄𝒂 , 𝜷𝒄𝒂 , and 𝑪𝒄𝒂 (See Appendix B) along with the parameters 𝑩𝑪𝑶𝟐 −𝒔𝒂𝒍𝒕 , 𝑪𝑪𝑶𝟐 −𝑪𝑶𝟐 −𝒔𝒂𝒍𝒕 and 𝝃𝒏𝒄𝒂 , the

CO2 solubility model is effectively extended to the aqueous CaCl2, Na2SO4, MgCl2, KCl solutions.

Table 3. 2 The results of concentration-dependent of the Pitzer triple-ion interaction parameter (𝝃𝒏𝒄𝒂 ).

CaCl2 ionic strength (mol kg-1)

T/K
1 2 3 4 5 6
323.15 0.77073 0.39436 0.26277 0.19510 0.15454 0.12769
373.15 0.92645 0.46400 0.30811 0.22918 0.18232 0.15141
423.15 1.103936 0.55441 0.36504 0.26831 0.21085 0.17319

T/K Na2SO4 ionic strength (mol kg-1)

1 2 3 4 5 6
323.15 -0.05670 -0.03783 -0.03180 -0.02852 -0.02644 -0.02508
373.15 -0.10374 -0.05646 -0.04081 -0.03287 -0.02802 -0.02489
423.15 -0.06269 -0.04135 -0.03205 -0.02657 -0.02293 -0.02038

T/K MgCl2 ionic strength (mol kg-1)

1 2 3 4 5 6
323.15 0.63298 0.32062 0.21067 0.15474 0.12163 0.09991
373.15 0.64031 0.32867 0.21797 0.16100 0.12696 0.10452
423.15 0.86806 0.40224 0.25688 0.18685 0.14620 0.11964

T/K KCl ionic strength (mol kg-1)

0.5 1 2 3 4 4.5
323.15 0.59800 0.32006 0.16333 0.10802 0.07958 0.07004
373.15 0.71435 0.36749 0.18537 0.12400 0.09210 0.08128
423.15 0.95168 0.45440 0.21731 0.14267 0.10498 0.09233

63
3.4 Results and discussion

3.4.1. Measured results

Table 3. 3(a-d) Experimental results for the CO2 solubility in aqueous CaCl2, Na2SO4, MgCl2 and KCl
solutions at 323.15-423.15 K and 150 bar.

(a) The CO2-CaCl2-H2O system

𝒎𝑪𝒂𝑪𝒍𝟐 𝒎𝑪𝑶𝟐 𝜹𝒎𝒓𝒂𝒏𝒅𝒐𝒎
T/K
/ mol kg-1 / mol kg-1 / mol kg-1
323.15 0.333 1.094 0.021
323.15 0.667 0.965 0.020
323.15 1.000 0.853 0.011
323.15 1.333 0.746 0.002
323.15 1.667 0.663 0.002
323.15 2.000 0.618 0.009
373.15 0.333 0.898 0.015
373.15 0.667 0.777 0.008
373.15 1.000 0.683 0.011
373.15 1.333 0.614 0.007
373.15 1.667 0.544 0.008
373.15 2.000 0.491 0.006
423.15 0.333 0.866 0.025
423.15 0.667 0.723 0.006
423.15 1.000 0.632 0.003
423.15 1.333 0.550 0.006
423.15 1.667 0.490 0.009
423.15 2.000 0.442 0.004

(b) The CO2-Na2SO4-H2O system

𝒎𝑵𝒂𝟐𝑺𝑶𝟒 𝒎𝑪𝑶𝟐 𝜹𝒎𝒓𝒂𝒏𝒅𝒐𝒎
T/K
/ mol kg-1 / mol kg-1 / mol kg-1
323.15 0.333 1.029 0.015
323.15 0.667 0.866 0.022
323.15 1.000 0.716 0.015
323.15 1.333 0.584 0.016
323.15 1.667 0.502 0.017
323.15 2.000 0.442 0.015
373.15 0.333 0.855 0.012
373.15 0.667 0.729 0.010
373.15 1.000 0.617 0.010
373.15 1.333 0.534 0.013
373.15 1.667 0.452 0.007
373.15 2.000 0.395 0.012
423.15 0.333 0.831 0.025
423.15 0.667 0.710 0.026
423.15 1.000 0.611 0.016
423.15 1.333 0.524 0.019
423.15 1.667 0.444 0.008
423.15 2.000 0.389 0.009

64
 (c) The CO2-MgCl2-H2O system
𝒎𝑴𝒈𝑪𝒍𝟐 𝒎𝑪𝑶𝟐 𝜹𝒎𝒓𝒂𝒏𝒅𝒐𝒎
T/K
/ mol kg-1 / mol kg-1 / mol kg-1
323.15 0.333 1.071 0.007
323.15 0.667 0.961 0.012
323.15 1.000 0.834 0.020
323.15 1.333 0.743 0.006
323.15 1.667 0.671 0.022
323.15 2.000 0.609 0.010
373.15 0.333 0.875 0.009
373.15 0.667 0.767 0.005
373.15 1.000 0.664 0.001
373.15 1.333 0.594 0.012
373.15 1.667 0.533 0.016
373.15 2.000 0.483 0.002
423.15 0.333 0.815 0.012
423.15 0.667 0.699 0.008
423.15 1.000 0.618 0.023
423.15 1.333 0.536 0.020
423.15 1.667 0.468 0.012
423.15 2.000 0.430 0.002

(d) The CO2-KCl-H2O system

𝒎𝑲𝑪𝒍 𝒎𝑪𝑶𝟐 𝜹𝒎𝒓𝒂𝒏𝒅𝒐𝒎
T/K
/ mol kg-1 / mol kg-1 / mol kg-1
323.15 0.500 1.174 0.018
323.15 1.000 1.112 0.007
323.15 2.000 1.005 0.014
323.15 3.000 0.929 0.003
323.15 4.000 0.880 0.014
323.15 4.500 0.855 0.017
373.15 0.500 0.941 0.011
373.15 1.000 0.895 0.002
373.15 2.000 0.805 0.015
373.15 3.000 0.740 0.030
373.15 4.000 0.683 0.017
373.15 4.500 0.657 0.005
423.15 0.500 0.881 0.008
423.15 1.000 0.816 0.012
423.15 2.000 0.724 0.020
423.15 3.000 0.651 0.004
423.15 4.000 0.593 0.014
423.15 4.500 0.565 0.006
Note: The instrument error 𝛿𝑚𝑖𝑛𝑠t𝑟 ⁡was not listed in the table and it was estimated approximately 0.7% of
the measured results (detailed error estimation please see Zhao et al., 2014). The range of random errors are:
1) CO2-CaCl2-H2O system, 0.31-2.90%, the average is 1.31%; 2) CO2-Na2SO4-H2O system, 0.74-3.58%,
the average is 2.39%; 3) CO2-MgCl2-H2O system, 0.17-3.76%, the average is 1.71%; 4) CO2-KCl-H2O
system, 0.29-4.01%, the average is 1.49%. 𝑚𝐶𝑂2 is the molality of CO2 in aqueous phase, mol kg-1; and
𝛿𝑚𝑟𝑎𝑛𝑑𝑜𝑚 ⁡denotes random error caused by repeated measurements.

65
 New experimental CO2 solubility data at 150 bar and 323.15-423.15 K for the systems of CO2-

CaCl2-H2O, CO2-Na2SO4-H2O, CO2-MgCl2-H2O and CO2-KCl-H2O at the ionic strength of the solutions up

to 6 mol kg-1 are reported in Table 3. 3.

At atmospheric pressure, Li and Tsui (1971) measured CO2 solubility in aqueous NaCl solution

(0.6413 mol kg-1 NaCl) and acidified seawater (with 10, 20, and 29‰ Chlorinity) at the temperature range

from 273.85 to 303.15 K. Based on their experimental results, Li and Tsui (1971) concluded that the

Buch’s assumption (Buch et al., 1932)—the effect of a given weight of sea salts on solubility of CO 2 is the

same as that of an identical weight of NaCl—is proved to be valid. However, in this study we have

experimentally demonstrated that CO2 solubilites in aqueous solutions with different salt species are

significantly different at 150 bar, whether the solutions have the same percent of salt by weight or have the

same ionic strength. This observation is also substantiated by experimental CO 2 solubility data given by

Yasunishi and Yoshida (1979a) and Onda et al. (1979) at ambient conditions (Figure 3. 1). Therefore, the

Buch’s assumption and the conclusion made by Li and Tsui (1971) are unable to extend to a more general

case: the effect of a given weight of any salt (or mixed-salt) on solubility of CO2 is the same as that of an

identical weight of NaCl. Figure 3. 1b and Figure 3. 2b show that, based on the same ionic strength, the

𝐾𝐶𝑙 𝐶𝑎𝐶𝑙2
magnitude of CO2 solubility in aqueous salt solutions follows a decreasing sequence of 𝑚𝐶𝑂2
> 𝑚𝐶𝑂2 >

𝑀𝑔𝐶𝑙2 𝑁𝑎𝐶𝑙 𝑁𝑎 𝑆𝑂4

𝑚𝐶𝑂2 > 𝑚𝐶𝑂2
> 𝑚𝐶𝑂22 .

Figure 3. 1 and Figure 3. 2 clearly show the trend of reduction of CO 2 solubility in the aqueous

phase as salt concentration increases, this 'salting-out' effect on CO2 solubility is the results of the various

interactions occurring in the aqueous phase, mainly due to the hydration of ions, and the interactions

between the hydrated ions and the dissolved CO2(aq). In addition to the 'salting-out' effect, these ion-water

molecule interactions dominate the CO2 solubility behavior in different aqueous salt solutions. For example,

on the basis of the number of ions produced when an electrolyte molecule dissociates in solution, MgCl2

and CaCl2 are the same type electrolytes (1-2 type), and the measured solubility of CO2 in CaCl2(aq) and in

MgCl2(aq) are quite close at a wide P-T-x range (Figure 3. 1 and Figure 3. 2). Moreover, the measured CO2

solubility in CaCl2(aq) is slightly greater than that in MgCl2(aq). Whereas, in spite of the same type (1-1

type) of electrolyte between NaCl and KCl, the effects of these two salts on the CO2 solubility in aqueous

66
solution are significantly different: The solubility of CO 2 in aqueous NaCl solution is much smaller than

that in aqueous KCl solution (Figure 3. 1 and Figure 3. 2).

The studies of ion hydration and ion–water molecule interactions (Samoilov, 1957; Collins, 1997)

demonstrated that small ions of high charge density bind water molecules strongly, whereas large ions of

low charge density bind water molecules weakly. In addition, Collins (1997) pointed out that ion effects on

water structure could be explained by a competition between charge density dominated ion–water

interactions and hydrogen bonding dominated water–water interactions. For example, the 1-2 type

electrolytes MgCl2 and CaCl2 share the same type of anion (Cl-). By assuming that the contributions of Cl-

to the salting-out effect between the salts MgCl2 and CaCl2 are the same, the small differences of the

salting-out effects between these two salts are mainly caused by the cations Mg2+ and Ca2+. The alkali

(Mähler and Persson, 2011) and alkaline earth metal ions tend to attract surrounding water molecules in the

aqueous phase to form a hydration shell—Mn+(H2O)m (M = Na, K, Ca, or Mg; n = 1 for alkali metals, n = 2

for alkaline earth metals; m is cluster number for H2O molecules) (Bush et al., 2009). The effective number

of water molecules in a hydration shell depends on the charge density of the involved cation. Mg 2+ and Ca2+

have the same charge number, but the ionic radius of Mg2+ (0.07±0.004 nm) is smaller than that of Ca 2+

(0.103±0.005) (Marcus, 1988). As such, the charge density of Mg2+ is slightly greater than that of Ca 2+,

thus Mg2+ tends to attract slightly more water molecules in its hydration shell than does Ca2+. Therefore, the

amount of free water molecules—not involved in the formation of the hydration shell and therefore

available to trap dissolved CO2 (aq) by forming hydrogen bonds4 between CO2(aq) and H2O molecules

(Moin et al., 2011) in aqueous MgCl2 solution—is smaller than those in aqueous CaCl2 solution based on

the same ionic strength. As a result, at the same ionic strength, the solubility of CO 2 in aqueous MgCl2

solution is smaller than that in aqueous CaCl2 solution. This explanation is corroborated by the

experimental CO2 solubility obtained in this study (Table 3. 3, Figure 3. 1b and Figure 3. 2b).

4
According to Moin et al.'s (2010) study, the average lifetime of hydrogen bonding between oxygen atoms
of CO2 and atoms of water molecules is higher than the hydrogen bonding lifetime of pure water. Their
simulation results are supported by experimental infra-red (IR) spectrum of a saturated aqueous CO2
solution.

67
 0.04
298.15 K, 1.013 bar NaCl
CaCl2
Na2SO4
0.03 MgCl2
-1
KCl
KCl
mCO2 / mol kg
PSUCO2

0.02
NaCl

0.01 Na2SO4

MgCl2
(a) CaCl2
0.00
0 10 20 30 40
Percent of salt by weight / %

(a)

0.04
298.15 K, 1.013 bar
NaCl
CaCl2
0.03 Na2SO4
KCl
-1

MgCl2
mCO2 / mol kg

KCl
PSUCO2
0.02
CaCl2
Na2SO4
0.01
NaCl

MgCl2
0.00
0 2 4 6 8 10 12 14 16 18
-1
Ionic strength / mol kg

(b)

Figure 3. 1 Comparison of the experimental CO2 solubilities in various aqueous salt solutions at 298.15 K
and 1.013 bar based on different concentration scales: (a) Percent of salt by weight; (b) Ionic strength. The
experimental data are taken from Setschenow (1892), Geffcken (1904), Findley and Shen (1912), Kobe and
Williams (1935), Onda et al. (1970), Yasunishi and Yoshida, (1979a), Burmakina et al. (1982), and the
large CO2 solubility data compilation given by Scharlin (1996). Solid lines represent the calculated values
from the PSUCO2 model.

68
 1.4
323.15 K, 150 bar
1.2 KCl

-1
1.0

mCO2 / mol kg
NaCl
MgCl2
0.8
Na2SO4 CaCl2
0.6 NaCl
CaCl2
Na2SO4
0.4 MgCl2
KCl
PSUCO2
0.2
0 5 10 15 20 25
Percent of salt by weight / %

(a)

1.4
323.15 K, 150 bar
1.2
KCl
-1

1.0
mCO2 / mol kg

CaCl2

0.8

NaCl MgCl2
0.6 CaCl2
Na2SO4
MgCl2
0.4 Na2SO4
KCl
PSUCO2 NaCl
0.2
0 1 2 3 4 5 6
-1
Ionic strength / mol kg

(b)

Figure 3. 2 Comparison of the experimental CO2 solubilities (this study) in different aqueous salt solutions
based on different concentration scales at 323.15 K and 150 bar (the comparison at 373.15 K and 423.15 K
were not shown due to the similar pattern, but the curves at the higher temperatures (i.e., 373.15 and 423.15
K) become more closer to each other): (a) Percent of salt by weight (%); (b) Ionic strength. Experimental
CO2 solubility in aqueous NaCl solution is given by Zhao et al. (2014a). Solid lines represent the calculated
values from the PSUCO2 model.

69
 In spite of the same type of electrolyte with the identical anion, CO2 solubility in aqueous

electrolyte solutions could be significantly different. The measured solubility of CO 2 in aqueous NaCl

solution was found to be considerably different from that in aqueous KCl solution (Figure 3. 1b and Figure

3. 2b). The calculated ion radius for K+ and Na+ are 0.141±0.008 and 0.097±0.006 nm (Marcus, 1988),

respectively. Na+ has a significantly greater charge density than K + and the hydration number of a hydrated

Na+ will be larger than that of a hydrated K +. Therefore, the number of free water molecules in aqueous

KCl solution is greater than those in aqueous NaCl solution, which allowing more CO2(aq) to be trapped by

aqueous KCl solution, and therefore the solubility of CO 2 in aqueous KCl solution is larger than that in

aqueous NaCl solution, especially at high KCl molality (Figure 3. 1 and Figure 3. 2).

Assuming that the volume occupied by an ion can be approximated by the volume of a sphere, the

charge density ratios for the ions of K+, Na+, Ca2+, and Mg2+ are estimated to be 1:3:5:16. The charge

densities, as well as the corresponding attraction forces between an ion and H 2O molecules, of K+ and Na+

are smaller relative to that of Ca2+ and Mg2+. Thus, the hydration shell surrounding these low charge

density ions (K+ or Na+) are loosely packed and have potential to hold more water molecules than does the

hydration shell of the high charge density ions (Ca 2+ or Mg2+). Therefore, the increase of charge density

from K+ to Na+ will result in a significant increase in the number of H 2O molecules in the hydration shell,

meanwhile cause a remarkable decrease in the number of free H 2O molecules in bulk solution, and

consequently lead to a sharp reduction of the CO2 solubility in aqueous NaCl solutions compared to that in

aqueous KCl solutions at the same ionic strength. This explanation elucidates why the CO2 solubility in

aqueous NaCl solution is much smaller than that in aqueous KCl solution at the same ionic strength.

However, the cations Ca2+ and Mg2+ exert a larger attraction force on surrounding water molecules

due to the charge density of these ions are larger relative to that of K + and Na+. As a result, the hydration

shells around these ions are tightly packed and therefore have no (or much less) place to hold additional

water molecules (i.e., the hydration shell is saturated). As a result, the increase in charge density for these

high charge density ions (e.g., change from Ca 2+ to Mg2+) effects no significant change on the number of

H2O molecules in its hydration shell, and the change of the amount of free H2O molecules in the bulk

solution is correspondingly small. This provides explanation for the small difference between the CO 2

solubilities in aqueous CaCl2 and MgCl2 solutions.

70
3.4.2. Modeling results

The calculated CO2 solubilities in the aqueous phase from various models (PSUCO2, SP2010,

DS2006 and OLI) were compared to the experimental data listed in Table 3. 4 by average absolute

deviation (AAD %), which is defined as below:

𝑐𝑎𝑙𝑐 𝑒𝑥𝑝
100 𝑚𝐶𝑂 −𝑚𝐶𝑂 ,𝑖
2 ,𝑖
𝐴𝐴𝐷(%) = ∑𝑁
𝑖=1 | 𝑒𝑥𝑝
2
| % (3.19)
𝑁𝑝 𝑚𝐶𝑂 ,𝑖
2

𝑐𝑎𝑙𝑐
where 𝑚𝐶𝑂2 ,𝑖
represents the calculated CO2 solubility values from the CO2 solubility models (PSUCO2,

𝑒𝑥𝑝
SP2010, DS2006, and OLI); 𝑚𝐶𝑂2 ,𝑖 denotes the experimental CO2 solubility taken from literature; and N

means the total number of experimental data of each work. In Table 3. 4, one can see that the model

PSUCO2 predicts the CO2 solubility in aqueous CaCl2, Na2SO4, MgCl2 and KCl solutions with a high

degree of accuracy over a wide P-T-x range (273.15-573.15 K, 1-2000 bar and ionic strength up to 6 mol

kg-1). Moreover, Figure 3. 3 clearly indicates that the proposed PSUCO2 model provides the best

performance in calculating the CO2 solubility in all CO2-salt-H2O systems considered herein compared

with the remaining models (SP2010, DS2006, and OLI).

12.2%

Duan et al. (2006)

9.3%

Spycher and Pruess (2010)

7.5%

OLI Studio 9.0

3.6%

PSUCO2

0 2 4 6 8 10 12 14
AAD (%)

Figure 3. 3 The average absolute deviation (AAD %) of the calculated CO2 solubility values in aqueous
CaCl2, Na2SO4, MgCl2 and KCl solutions by different models compared to the experimental data (See
section 3.4.2, all available experimental data listed in Table 3. 4).

71
 1.5
1.4 150 bar, CaCl2(aq) 323.15 K
373.15 K
1.3
423.15 K
1.2 SP2010

-1
1.1 DS2006

mCO2 / mol kg
OLI Studio 9.0
1.0 PSUCO2
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.0 0.5 1.0 1.5 2.0
-1
mCaCl2 / mol kg

(a)

1.5
1.4 150 bar, Na2SO4(aq)
323.15 K
1.3 373.15 K
423.15 K
1.2 SP2010
-1

1.1 DS2006
mCO2 / mol kg

OLI Studio 9.0

1.0
PSUCO2
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.0 0.5 1.0 1.5 2.0
mNa2SO4 / mol kg-1

(b)

72
 1.5
1.4 150 bar, MgCl2(aq) 323.15 K
1.3 373.15 K
1.2 423.15 K
SP2010
-1 1.1
mCO2 / mol kg
DS2006
1.0 OLI Studio 9.0
PSUCO2
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.0 0.5 1.0 1.5 2.0
-1
mMgCl2 / mol kg

(c)

1.6
1.5 150 bar, KCl(aq) 323.15 K
373.15 K
1.4
423.15 K
1.3 SP2010
-1

1.2 DS2006
mCO2 / mol kg

OLI Studio 9.0

1.1 PSUCO2
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

mKCl / mol kg-1

(d)

Figure 3. 4 Comparison of model calculated values against the experimental data (this study): (a) CO 2-
CaCl2-H2O system; (b) CO2-Na2SO4-H2O system (c) CO2-MgCl2-H2O system (d) CO2-KCl-H2O system.

73
Table 3. 4 Comparison of the model calculations with experimental data. The results are shown by average absolute deviation (AAD %, Eq. (3.19)). The
numbers in parentheses stand for the number of experimental data evaluated by each model.

No. References Systems P/ T/K msalt / 𝑵 Model calculations, AAD %

bar mol kg-1 PSUCO2 OLI(d) SP2010 DS2006
CO2-CaCl2-H2O 0.1-5.0 11 4.11(11) 5.58(9)(b) 4.34(9)(b) 7.08(10)(c)
1(a) Setschenow (1892) 1.01 288.35
CO2-KCl-H2O 1-4.3 3 1.92(3) 4.60(3) 21.47(3) 15.98(3)
(a)
2 Geffcken (1904) CO2-KCl-H2O 1.01 288.15-298.15 0.4-1.1 8 0.70(8) 2.29(8) 10.30(8) 6.37(8)
Findley and Shen
3 CO2-KCl-H2O 1.01 298.15 0.2-1 4 0.56(4) 3.40(4) 8.25(4) 6.18(4)
(1912)
Kobe and Williams CO2-Na2SO4-H2O 1.01 298.15 1.76 2 1.30(2) 26.82(2) N/A(0) N/A(0)
4
(1935) CO2-MgCl2-H2O 1.01 298.15 4.5 1 5.50(1) N/A(1)(b) N/A(0)(b) N/A(0)(c)
Markham and CO2-Na2SO4-H2O 298.15-313.15 0.2-1.5 8 0.61(8) 7.85(8) 7.03(4) 25.89(8)
5 1.01
Kobe (1941) CO2-KCl-H2O 273.35-313.15 0.1-4 16 0.84(16) 4.59(16) 15.65(10) 10.62(16)
Prutton and Savage
6 CO2-CaCl2-H2O 15-712 348.15-394.15 1-3.9 116 5.58(116) 3.86(116) 9.06(104) 9.22(116)
(1945)
(a)
7 Gerecke (1969) CO2-KCl-H2O 15-60 274.16 0.5-3.4 40 3.69(40) 7.19(40) 16.73(40) 13.01(40)
CO2-CaCl2-H2O 1.013 298.15 0.2-2.3 8 0.49(8) 5.94(8) 4.16(8) 1.00(8)
8 Onda et al. (1970)
CO2-Na2SO4-H2O 1.013 298.15 0.5-1.5 3 1.77(3) 10.60(3) 26.49(3) 30.33(3)
Malinin and
Savelyeva (1972)
9 CO2-CaCl2-H2O 48 298.15-423.15 0.2-6.9 25 4.32(25) 5.74(25) 1.64(25) 3.17(21)(c)
Malinin and
Kurovskaya (1975)
CO2-CaCl2-H2O 298.15-308.15 0.2-5.3 16 3.46(16) 6.14(16) 5.76(14)(b) 3.16(15)(c)
Yasunishi and CO2-Na2SO4-H2O 288.15-308.15 0.2-2.4 26 3.46(26) 11.95(26) 5.35(7) 28.49(25)
10 1.013
Yoshida (1979a) CO2-MgCl2-H2O 288.15-308.15 0.1-4.4 29 3.66(29) 9.72(29) 7.47(25) 2.63(29)
CO2-KCl-H2O 298.15-308.15 0.4-4.8 16 0.95(16) 8.01(16) 22.39(16) 17.05(14)(c)
Burmakina et al.
11 CO2-KCl-H2O 1 298.15 0.001-0.2 8 1.15(8) 4.48(8) 5.90(8) 3.93(8)
(1982)
12 Corti et al. (1990) CO2-Na2SO4-H2O 16-200 323.15-348.15 1-3.3 24 13.91(24) 20.50(15)(b) N/A(0) 38.87(8)
Rumpf and Maurer,
13 CO2-Na2SO4-H2O 4-97 313.15-433.15 1-2 102 2.73(102) 6.91(102) N/A(0) 27.57(102)
(1993)
14 Kiepe et al. (2002) CO2-KCl-H2O 1-105 313.15-353.15 0.5-4 88 10.16(88) 9.42(88) 13.58(87) 13.07(88)
Bermejo et al.
15 CO2-Na2SO4-H2O 20-131 287.15-369.15 0.25-1 113 6.70(113) 12.77(113) 12.73(24) 22.72(113)
(2005)
Pérez-Salado
16 CO2-KCl-H2O 4-94 313.15-433.15 2-4 98 1.65(98) 2.67(98) 18.35(98) 18.34(98)
Kamps et al.

74
 No. References Systems P/ T/K msalt / 𝑵 Model calculations, AAD %
bar mol kg-1 PSUCO2 OLI(d) SP2010 DS2006
(2007)
CO2-CaCl2-H2O 20-160 318.15 1 8 8.43(8) 11.26(8) 9.63(8) 11.76(8)
17 Liu et al. (2011)
CO2-KCl-H2O 20-160 318.15 1.5 8 2.93(8) 3.32(8) 13.78(8) 15.39(8)
CO2-CaCl2-H2O 15-380 308.15-424.15 1-5 36 7.36(36) 6.73(24)(b) 4.85(24)(b) 4.69(24)(c)
18 Tong et al. (2013)
CO2-MgCl2-H2O 12-350 308.15-424.15 1-5 39 5.14(39) 6.14(26)(b) 5.67(26)(b) 5.83(26)(c)
CO2-CaCl2-H2O 150 323.15-423.15 0.33-2 18 0.88(18) 2.13(18) 3.12(18) 3.61(18)
CO2-Na2SO4-H2O 150 323.15-423.15 0.33-2 18 0.75(18) 8.04(18) 7.50(3) 15.90(18)
19 This study
CO2-MgCl2-H2O 150 323.15-423.15 0.33-2 18 0.83(18) 2.30(18) 3.49(18) 3.20(18)
CO2-KCl-H2O 150 323.15-423.15 0.5-4.5 18 0.33(18) 4.21(18) 14.91(18) 16.71(18)
CO2-CaCl2-H2O 4.3(238) 5.9(224) 5.3(210) 5.5(220)
Overall AAD % for the
CO2-Na2SO4-H2O 3.9(296) 13.2(287) 11.8(41) 27.1(277)
system of
CO2-MgCl2-H2O 3.8(87) 6.1(73) 5.5(69) 3.9(73)
(Including this study)
CO2-KCl-H2O 2.3(307) 4.9(307) 14.7(300) 12.4(305)
Overall AAD % CO2-salt-H2O 3.6 7.5 9.3 12.2
Experimental data P-T-x
CO2-Salts-H2O 100% 96% 67% 94%
coverage rate
'N/A' in the table denotes the P-T-x region of the experimental data beyond the capacity of the corresponding model.
(a)
Indicates the experimental data are taken from large data compilation by Scharlin (1996).
(b)
The experimental P-T-x points at mCaCl2 > 3 mol kg-1 were removed from the comparison for both the OLI and SP2010 model due to the poorly results for these
model at high CaCl2 concentration.
(c)
The experimental P-T-x points at mCaCl2 > 4.5 mol kg-1 were removed due to the DS2006 model is limited to 4.5 mol kg-1 salt molality.
(d)
Calculations made by OLI are under the following conditions: (1) enable the second liquid phase, (2) use MSE model, (3) stream inflows contains 55.5082 mol
water and 5 mol CO2

75
 Figure 3. 4 and Table 3. 4(row 19) show that SP2010 and DS2006 models predict the new

experimental data obtained herein quite well for the CO2-CaCl2-H2O and CO2-MgCl2-H2O systems,

however, these same models predict the measured CO2 solubility poorly in aqueous Na2SO4 and KCl

solutions. The similar performance between the models SP2010 and DS2006 is mostly due to the fact that

both models use the same type of simplified Pitzer equation, which is proposed by Duan and Sun (2003), to

calculate activity coefficient of dissolved CO2. Note that the simplified Pitzer equation used in the SP2010

model is re-parameterized so it can be used up to 573.15 K (Spycher and Pruess, 2010). OLI Studio 9.0.6

performs very good in calculating CO2 solubility in the aqueous phase for the systems of CO 2-KCl-H2O,

CO2-CaCl2-H2O, and CO2-MgCl2-H2O, but the results for the CO2-Na2SO4-H2O system are significantly

deviated from the experimental CO2 solubility data (Table 3. 4, also Figure 3. 4b). In addition, some of the

experimental data (e.g., Corti et al., 1990 and Kiepe et al., 2002), are seriously deviate from calculated

results of all available models (Table 3. 4), additional experimental work should be pursued in the future to

explore this discrepancy.

0.05
1.013 bar, CaCl2(aq)
288.35 K, Setchenow, 1892
0.04
308.15 K, Yasunishi and Yoshida, 1979
298.15 K, Onda et al., 1970
-1
mCO2 / mol kg

298.15 K, Yasunishi and Yoshida, 1979

0.03 PSUCO2

0.02

0.01

0.00
0 1 2 3 4 5 6
-1
mCaCl2 / mol Kg

(a)

76
 1.2
48 bar, CaCl2(aq)
1.0 PSUCO2
298.15 K, Malinin and Savelyeva, 1972
323.15 K, Malinin and Savelyeva, 1972

-1
0.8 348.15 K, Malinin and Savelyeva, 1972

mCO2 / mol kg
373.15 K, Malinin and Kurovskaya, 1975
423.15 K, Malinin and Kurovskaya, 1975
0.6

0.4

0.2

0.0
0 1 2 3 4 5 6 7
-1
mCaCl2 / mol kg

(b)

1.4
394.15 K, CaCl2(aq)
PSUCO2
1.2 1.012 mol/kg CaCl2
2.281 mol/kg CaCl2
1.0 3.899 mol/kg CaCl2
-1
mCO2 / mol kg

0.8

0.6

0.4

0.2
Prutton and Savage, 1945
0.0
0 100 200 300 400 500 600 700
P / bar

(c)

77
 1.0
424.35 K, CaCl2(aq)
PSUCO2
0.8 1 mol/kg CaCl2
3 mol/kg CaCl2

-1
5 mol/kg CaCl2

mCO2 / mol kg
0.6

0.4

0.2

Tong et al., 2013

0.0
0 100 200 300 400
P / bar
(d)

Figure 3. 5(a-d) Comparison of model calculations against the experimental CO 2 solubility in aqueous
CaCl2 solutions: (a) 1.013 bar, 288.15-308.15 K, and 0.14-5.31 mol kg-1 CaCl2 (Setschenow, 1892; Onda et
al., 1970; Yasunishi and Yoshida, 1979a); (b) 48 bar, 298.15-423.15 K, and 0.16-6.95 mol kg-1 CaCl2
(Malinin and Savelyeva, 1972; Malinin and Kurovskaya, 1975); (c) 21-712 bar, 394.15 K, and 1.01-3.90
mol kg-1 CaCl2 (Prutton and Savage, 1945); (d) 44-380 bar, 424.35 K, and 1-5 mol kg-1 CaCl2 (Tong et al.,
2013).

The comparisons of the calculated results of PSUCO2 with literature data are also shown in Figure

3. 5 to Figure 3. 8. Figure 3. 5(a-d) shows an excellent agreement of our modeling results with the reliable

experimental CO2 solubility data in the aqueous CaCl2 solutions at 1-712 bar, 288.15-424.35 K, and CaCl2

molality from 0.1 to 7.0 mol kg-1. With the similar performance to the CO2-CaCl2-H2O system, the

proposed PSUCO2 model is capable of accurately calculating CO 2 solubility in aqueous Na2SO4, MgCl2,

and KCl solutions at elevated temperatures and pressures, the examples are shown in Figure 3. 6(a, b),

Figure 3. 7(a, b), and Figure 3. 8(a-c) for the systems of CO2-Na2SO4-H2O, CO2-MgCl2-H2O, and CO2-

KCl-H2O, respectively.

78
 0.8
1.011-1.013 mol kg-1 Na2SO4(aq)
0.7 313.15 K
323.15 K
0.6 333.15 K

-1
353.15 K

mCO2 / mol kg
0.5 393.15 K
413.15 K
0.4 433.15 K
PSUCO2
0.3

0.2

0.1
Rumpf and Maurer, 1993
0.0
0 20 40 60 80 100
P / bar

(a)

0.5
2.002-2.01 mol Kg-1 Na2SO4(aq)
313.15 K
0.4 333.15 K
353.15 K
-1

393.15 K
mCO2 / mol kg

0.3 413.15 K
433.15 K
PSUCO2
0.2

0.1

Rumpf and Maurer, 1993

0.0
0 20 40 60 80 100
P / bar

(b)

Figure 3. 6(a, b) Comparison of the model calculations against the experimental CO2 solubility in aqueous
Na2SO4 solutions: (a) 4-97 bar, 313.15-433.15 K, and ~1 mol kg-1 Na2SO4; (b) 7-92 bar, 313.15-433.15 K,
and ~2 mol kg-1 Na2SO4. The experimental data are reported by Rumpf and Maurer (1993).

79
 0.045
1.013 bar, MgCl2(aq)
Yasunishi and Yoshida (1979)

0.035 PSUCO2

-1
288.15 K

mCO2 / mol kg
298.15 K
303.15 K
0.025

0.015

0.005
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
-1
mMgCl2 / mol kg

(a)

0.4
5 mol/kg MgCl2(aq)
Tong et al., 2013
0.3
-1
mCO2 / mol kg

0.2
PSUCO2
309.6 K
344.68 K
0.1 374.6 K
424.4 K

0.0
0 50 100 150 200 250 300 350 400
P / bar

(b)

Figure 3. 7(a, b) Comparison of the model calculations against the experimental CO2 solubility in aqueous
MgCl2 solutions: (a) 1.013 bar, 288.15-303.15 K, and 0.15-4.45 mol kg-1 MgCl2 (Yasunishi and Yoshida,
1979); (b) 35-312 bar, 309.6-424.4 K, and 5 mol kg-1 MgCl2 (Tong et al., 2013).

80
 0.05
1.013 bar, KCl(aq)
PSUCO2
288.15, 288.35 K
298.15 K
0.04
308.15 K

-1
313.15 K

mCO2 / mol kg 0.03

0.02

0.01
0 1 2 3 4 5
-1
mKCl / mol kg

(a)

1.0
3.995-4.05 mol kg-1 KCl(aq)
313 K
0.8 333 K
353 K
-1

373 K
mCO2 / mol kg

393 K
0.6
413 K
433 K
PSUCO2
0.4

0.2

Pérez-Salado Kamps et al., 2007

0.0
0 20 40 60 80 100
P / bar

(b)

81
 1.6
Kiepe et al. (2002) KCl(aq)
1.4

1.2 PSUCO2, 313.15 K

PSUCO2, 353.15 K
-1
mCO2 / mol kg

1.0 313.15 K, 0.5 m KCl

313.15 K, 1.0 m KCl
0.8 313.15 K, 2.5 m KCl
313.15 K, 4.0 m KCl
353.15 K, 0.5 m KCl
0.6
353.15 K, 1.0 m KCl
353.15 K, 2.5 m KCl
0.4
353.15 K, 4.0 m KCl
0.2

0.0
0 20 40 60 80 100
P / bar

(c)

Figure 3. 8(a-c) Comparison of the model calculations against the experimental CO 2 solubility in aqueous
KCl solutions: (a) 1.013 bar, 288.15-313.15 K, and 0-4.75 mol kg-1 KCl (Setchenow, 1892; Geffcken, 1904;
Findley, 1912; Markham and Kobe, 1941; Yasunishi and Yoshida, 1979a; Burmakina et al., 1982); (b) 6-94
bar, 313.15-433.03 K, and ~4 mol kg-1 KCl (Pérez-Salado Kamps et al., 2007); (c) 1-105 bar, 313.15-
353.15 K, and 0.5-4 mol kg-1 KCl (Kiepe et al., 2002).

The comparison of the calculated results of the PSUCO2 model with the experimental data for the

CO2-KCl-H2O system at ambient conditions also shows a remarkable agreement (Figure 3. 8(a,b)), whereas

at elevated pressures and temperatures, a large discrepancy of experimental results for the CO 2-KCl-H2O

system between Pérez-Salado Kamps et al. (2007) (Figure 3. 8b) and Kiepe et al. (2002) (Figure 3. 8c) was

observed, the calculated results from the PSUCO2 model support the experimental CO 2 solubility given by

Pérez-Salado Kamps et al. (2007).

In Figure 3. 9a, the comparisons of calculated H2O solubility in the CO2-rich phase among the

different models (PSUCO2, OLI, and SP2010) were made at 298.15 and 423.15 K, and up to 500 bar. At

298.15 K, the differences among the three models are rather small (Figure 3. 9a). All three models are

capable of predicting the phase change (Figure 3. 9a, enlarged view) from gaseous to liquid CO2-rich phase

(Spycher et al., 2003) at 298.15 K as compared with the experimental data from Wiebe and Gaddy (1941).

At 423.15 K, both PSUCO2 and SP2010 models are shown agree well with the experimental data given by

Tödheide and Franck (1963), but the OLI shows a gradually deviation in predicting H 2O solubility in the

82
CO2-rich at 423.15 K as pressure increases, i.e. from 150 to 500 bar (Figure 3. 9a). Also, the calculated

results from OLI show a sharp phase change around 435-445 bar at 423.15 K, this change deviates from

both the experimental data (Tödheide and Franck, 1963) and the results calculated by the other two models

(SP2010 and PSUCO2).

25
CO2-H2O binary system at 298.15 and 423.15 K
0.6 Phase change
20 0.5 enlarged view at
yH2O %

0.4 298.15 K (0-120 bar) Tödheide and Franck, 1963

0.3 Weibe and Gaddy, 1941
15 0.2
SP2010
yH2O %

OLI Studio 9.0

0.1 PSUCO2
0 20 40 60 80 100 120
10 P / bar

5 423.15 K

298.15 K
0
0 100 200 300 400 500 600
P / bar

(a)

6.1
CO2-salt-H2O system at 423.15 K
Na2SO4(aq)
5.9
NaCl(aq)
KCl(aq)
5.7
yH2O %

MgCl2(aq) NaCl(aq)
and KCl(aq)
5.5 CaCl2(aq)
CaCl2(aq), MgCl2(aq)
and Na2SO4(aq) are
overlapped here
5.3
Salt molality: PSUCO2
2.0 mol kg-1 H2O SP2010
5.1
100 200 300 400 500
P / bar

(b)

83
 0.94
CO2-salt-H2O system at 423.15 K

0.92 Na2SO4(aq)

Activity of H2O (aH2O)

0.90 NaCl(aq)

KCl(aq)
0.88 CaCl2(aq)

0.86
Salt molality
2.0 mol kg-1 H2O MgCl2(aq)
0.84
100 200 300 400 500
P / bar

(c)

Figure 3. 9(a-c) Comparison of model calculation of H2O solubility in the CO2-rich phase: (a) model
calculations (SP2010, OLI, and PSUCO2) for the binary CO 2-H2O system, experimental data at 298.15 and
423.15 K are taken from Wiebe and Gaddy (1941) and Tödheide and Franck (1963), respectively; (b)
comparison of calculated H2O solubility in the CO2-rich phase at 423.15 K in the CO2-salt-H2O systems
between SP2010 and PSUCO2, the results from OLI deviates from these two models and therefore not
shown in the figure (b); (c) the PSUCO2 model calculated activity of H 2O in the aqueous phase for the
various CO2-salt-H2O system at 423.15 K. Calculations made by OLI Studio 9.0.6 are under the following
conditions: (1) enable the second liquid phase, (2) use MSE model, (3) stream inflows contains 55.5082
mol water and 5 mol CO2.

In Figure 3. 9b, the computed H2O solubilities in the CO2-rich phase for various CO2-salt-H2O

systems demonstrate that the different salt species presented in the aqueous phase will influence the H 2O

solubility in the CO2-rich phase. Given the same salt molality (e.g. 2 mol kg-1 in Figure 3. 9b), the

differences of H2O solubility in various single-salt aqueous solutions are quite large, whereas the SP2010

model reflects a relatively smaller variation of H2O contents in the CO2-rich phase caused by the presence

of different salt species than does PSUCO2. Additionally, the results from the SP2010 model show that the

H2O solubilities in the CO2-rich phase are overlapped for certain systems (Figure 3. 9b). This is because the

activity model used by the SP2010 model differentiates cations only based on their charge number, and

therefore all monovalent cations (e.g. K+ and Na+) are the same kind of cation (and all divalent cations (e.g.

Ca2+ and Mg2+) are also the same during the CO2 solubility calculations. Thus, for the SP2010 model, the

computed CO2 solubilities in aqueous NaCl and KCl solutions with same salt molality are the same,

84
consequently, the calculated H2O solubilities in the CO2-rich phase for the systems of CO2-NaCl-H2O and

CO2-KCl-H2O are also identical. The same situations occur for the CO 2-CaCl2-H2O and CO2-MgCl-H2O

systems by using the SP2010 model. Because the similar activity equation is used in both DS2006 and

SP2010 models, at the same salt concentration, the CO2 solubilities in aqueous NaCl and KCl solutions,

likewise in aqueous CaCl2 and MgCl2 solutions, are also indistinguishable by using DS2006 model. Unlike

the SP2010 and DS2006 models, the proposed PSUCO2 model is able to predict the mutual solubilities of

CO2 and H2O in a CO2-salt-H2O system by taking into account the effects caused by different salt species,

e.g., a strong dependence of H2O solubility in the CO2-rich phase on salt species are observed from the

PSUCO2 model (Figure 3. 9b).

Figure 3. 9c provides an explanation of the observed differences of H 2O solubility in the CO2-rich

phase among the various CO2-salt-H2O systems (Figure 3. 9b). The H2O solubility in the CO2-rich phase

are dominated by the activity of H2O in the aqueous phase. Even at the same P-T-x condition, the H2O

solubility in the CO2-rich phase is different for the various CO2-salt-H2O system due to each salt species

has distinctive influence on the activity of H2O (aH2O) in the aqueous phase. The modeling results shows

that a positive correlation between the H2O solubility in the CO2-rich phase and the activity of H2O in the

aqueous phase for the various CO2-salt-H2O system (Figure 3. 9c). In this case (100-500 bar, 423.15 K, and

2 mol kg-1 salt), the activity of H2O in aqueous Na2SO4 solution takes the greatest value among the others

(Figure 3. 9c), so that there is the corresponding largest H2O solubility in the CO2-rich phase (yH2O) for the

CO2-Na2SO4-H2O system at the same P-T-x condition. For other salt species (NaCl, CaCl2, MgCl2 and

KCl), the magnitude of H2O solubility in the CO2-rich phase is also preserved corresponding to the

variation of H2O activity in the aqueous phase in response to changed salt species (Figure 3. 9b and Figure

3. 9c). The observation of higher activity of H2O yields higher solubility of H2O in the CO2-phase is based

on the comparison among different CO2-salt-H2O systems, note that for the same salt species the increase

of the H2O activity does not guarantee the increased H2O solubility in the CO2-rich phase.

85
3.5 Conclusions

At 150 bar, new CO2 solubility data in the aqueous phase were obtained for the CO 2-CaCl2-H2O,

CO2-Na2SO4-H2O, CO2-MgCl2-H2O and CO2-KCl-H2O systems at temperatures 323.15, 373.15, and

423.15 K and ionic strengths from 0 to 6 mol kg-1 of dissolved salt species. Error analysis showed that the

instrumental error of the observed CO2 solubility is around 0.7% and the random error is from 0.1 to 3.0 %

based on the molal concentration scale.

The comparisons of experimental results based on the same concentration scale (ionic strength or

salt percent by weight) showed that (1) the solubilities of CO2 in aqueous solutions in the presence of

different salt species are quite different; (2) the solubilities of CO2 in CaCl2(aq) and MgCl2(aq) solutions

are close to each other; and (3) in the aqueous phase a cation with a higher charge density (a smaller radii

and a larger charge number) will have more significant salting-out effect on dissolved CO2 than other ions.

By using new experimental CO2 solubility data collected in different aqueous salt solutions, we

extended the previously developed CO2 solubility model for the CO2-NaCl-H2O system (see Chapter 2) to

include the salt species of CaCl2, Na2SO4, MgCl2 and KCl into the proposed PSUCO2 model. Comparisons

against literature data reveal a clear improvement of the proposed PSUCO2 model among the published

models in predicting CO2 solubility in aqueous CaCl2, MgCl2, Na2SO4 and KCl solutions at temperatures

from 273.15-573.15 K and pressures up to 2000. Also, the modeling results show a positive correlation

between the H2O solubility in the CO2-rich phase and the activity of H2O in the aqueous phase, the

influence of salts species on solubility of H2O in the CO2-rich phase is clearly observed.

86
 CHAPTER 4

SOLUBILITY OF CO2 IN SYNTHETIC FORMATION BRINE5

5
The text for this chapter was originally prepared for the publication as “Haining Zhao, Robert Dilmore,
Douglas E. Allen, Sheila W. Hedges, Yee Soong and Serguei N. Lvov, Measurement and modeling of CO 2
solubility in natural and synthetic formation brines, Environmental Science & Technology, in review”.

87
Abstract

CO2 solubility data in the natural formation brine, synthetic formation brine, and synthetic

NaCl+CaCl2 brine were collected at the pressures from 100 to 200 bar, temperatures from 323-423 K.

Experimental results demonstrate that the CO2 solubility in the synthetic formation brines can be reliably

represented by that in the synthetic NaCl+CaCl2 brines. We extended our previously developed model

(PSUCO2) to calculate CO2 solubility in aqueous mixed-salt solution by using the additivity rule of the

Setschenow coefficients of the individual ions (Na +, Ca2+, Mg2+, K+, Cl-, and SO42-). Comparisons with

previously published models against the experimental data reveal a clear improvement of the proposed

PSUCO2 model. Additionally, the path of the maximum gradient on the CO 2 solubility contour map is

defined to divide the P-T diagram into two regions: in Region I, the CO2 solubility in the aqueous phase

decreases monotonically in response to increased temperature, but in region II, the behavior of the CO 2

solubility is the opposite of that in Region I as the temperature increases. A web-based CO2 solubility

computational tool can be accessed via the link: www.carbonlab.org/psuco2/.

88
4.1 Introduction

Knowledge of the properties of the CO2-brine system is essential not only for the CO2 related

industrial applications but also for understanding its behavior in geochemical systems. Since the start of the

first CO2 enhanced oil recovery (CO2-EOR) project in the Kelly-Snyder Oil Field in West Texas began in

1971, CO2 has been injected into geological formations to economically improve recovery of incremental

hydrocarbon from mature oil-bearing reservoirs.6 Further, the world's first industrial-scale CO2 capture and

sequestration (CCS) project was initialed at the Sleiper Field in the North Sea since 1996. 7 As a part of a

portfolio of low-carbon technologies, 20 large-scale integrated CCS projects are currently operating or have

secured financial backing to proceed to construction (Global CCS Institute). 8 In addition to the CO2-EOR

and CCS, CO2 enhanced geothermal system (CO2-EGS) has been proposed — utilizing supercritical CO2 as

the heat transmission fluid to extract heat energy from geothermal reservoirs as a means to produce

renewable geothermal energy, and to concomitantly geologically store CO 2 through potential fluid losses at

depth (Brown, 2000; Spycher and Pruess, 2010; Tester et al., 2006). The CO2-brine system containing ions

of Na+, Ca2+, Mg2+, K+, Cl-, and SO42-, will be the most commonly encountered fluid system in all of above

mentioned engineered natural systems; a good understanding of the thermodynamic properties of the CO 2-

brine system is, therefore, critical for evaluating, designing and modeling these industrial processes

(Spycher and Pruess, 2010; Pruess et al., 2001; Kumar et al., 2005; Bickle et al., 2007).

In geosciences, there are various sources from which the CO2-brine system could be naturally

generated and presented in the sallow crust of the earth (Kaszuba et al., 2006), e.g., the CO2-brine fluid

generated from a hydrocarbon reservoir at intermediate levels of thermal maturity of organic matter (Cappa

and Rice, 1995), or generated by degassing of CO2 in magma-hydrothermal system (Lowenstern, 2001).

Under high temperatures and pressures, the evidence that the CO 2-brine systems transform the subsurface

rock matrix are abundant, such as the formation of ore deposits (Gize and Macdonald, 1993; Phillips and

Evans, 2004), the CO2-rich fluid inclusions in minerals (Lamb et al., 1987), and the participation of rock-

forming processes (Sisson et al, 1981). Therefore, over geologic time, the CO2-brine system can also be

6
Texas State Historical Association Website:https://www.tshaonline.org/handbook/online/articles/doksu
7
British Geological Survey Website: http://www.bgs.ac.uk/science/CO2/home.html
8
CCS Applications, Global CCS Institute:
http://www.globalccsinstitute.com/sites/default/files/pages/92241/projects-action-rgb-low-res.pdf

89
expected to have great influence on the rock and fluid properties in various geologic environments,

understanding the thermodynamic properties of this fluid system is of significant importance for

geosciences.

In spite of the fact that CO2 solubility in pure water (Spycher et al., 2003; Springer et al., 2012)

and in single-salt aqueous systems (Akinfiev and Diamond, 2010; Spycher et al., 2003; Springer et al.,

2012; Zhao et al., 2014a, Zhao et al., 2014b) has been widely studied, investigation of CO2 solubility in

mixed-salt solutions at elevated temperatures and pressures remains few and therefore the complexity of

ion-ion and ion-neutral interactions for the CO2-mixed-salt-H2O system is largely unknown. Correlating

and predicting CO2 solubility in such mixed-salt aqueous systems at elevated temperatures and pressures

has remained a difficult task (Yasunishi et al., 1979b; Rumpf et al., 1994). While some recent studies report

experimental CO2 solubility in aqueous mixed-salt solutions, they are unable or just using a simple

correlative equation to provide corresponding theoretical solubility calculations for the CO 2-mixed-salt-

H2O system (Liu et al., 2011; Tong et al., 2013).

Nevertheless, three previously published models are capable of computing CO 2 solubility in the

mixed-salt aqueous systems: the DS2006 model (Duan and Sun, 2003; Duan et al., 2006) the SP2010

model (Spycher et al., 2003; Spycher and Pruess, 2005, 2010) and the OLI Studio 9.0.6 — a commercial

software package developed by OLI Systems Inc (Springer et al., 2012). All three models show excellent

performance in calculating CO2 solubility in pure water and single-salt systems of CO2-NaCl-H2O, CO2-

CaCl2-H2O, and CO2-MgCl2-H2O (See Chapters 2 and 3) but their capacity for predicting CO2 solubility in

real brines remains to be validated, since: (1) the experimental data of CO2 solubility in mixed-salt brine at

elevated temperatures and pressures are too scarce to substantiate their performance, (2) for brine with a

relatively high concentrations of SO42- and K+ ions, the DS2006 and SP2010 models cannot be fitted well

to the experimental data (See Chapter 3). Therefore, a more accurate model is needed to predict the

solubility of CO2 in aqueous mixed-salt solutions.

Inspired by the studies of Görgényi et al. (2006) and Gordon and Thorne (1967) on salting-out

effects on non-electrolytes, we have demonstrated that the additivity rule of Setschenow Coefficients

(Setchenow, 1889) of individual ions can be used to accurately calculate the solubility of CO2 in mixed-salt

solutions at elevated temperatures and pressures. Without previous experimental and modeling efforts on

90
CO2 solubility in pure water and in five single-salt aqueous systems (Chapters 2 and 3), the goal of

accurately predicting CO2 solubility in mixed-salt solutions cannot be easily achieved. Throughout Chapter

4, PSUCO2 denotes our updated CO2 solubility model for aqueous mixed-salt solution.

4.2 Materials and methods

4.2.1 Chemicals

The carbon dioxide used in all experiments was Coleman Instrument grade with a purity of

99.99 %. Water was purified by Milli-Q system and was degassed before loading it into the autoclave. The

purified water conductivity was below 610-6 S m-1. All aqueous salt solutions were prepared using this

Milli-Q water and ACS Grade reagents: sodium chloride (NaCl, J.T. Baker, 99%), calcium Chloride

(CaCl2‧2H2O, Alfa Aesar, 99%), sodium sulfate (Na 2SO4, Amresco, 99%), magnesium chloride (MgCl2,

Alfa Aesar, 99%), potassium chloride (KCl, Alfa Aesar, 99%), Strontium chloride (SrCl 2, anhydrous, Alfa

Aesar, 99%) and Sodium bromide (NaBr, Alfa Aesar, 99%).

4.2.2 Synthetic brine preparation.

The ionic strength based on molality was used herein as a measure of the concentration of mixed-

salt aqueous solutions, it is defined as 𝐼 = 0.5 ∑𝑖 𝑚𝑖 𝑧𝑖2 , where 𝑚𝑖 and 𝑧𝑖 respectively represent the molality

and charge number of the i-th ion. The brines used for the CO2 solubility measurements include:

1) a natural Mt. Simon brine with IS = 1.815 mol kg-1;

2) a synthetic Mt. Simon formation brine with IS = 1.712 mol kg -1;

3) a synthetic NaCl+CaCl2 brine with IS = 1.712 mol kg-1;

4) a synthetic Antrim Shale formation brine35 with IS = 4.984 mol kg-1;

5) a synthetic NaCl+CaCl2 brine with IS = 4.984 mol kg-1.

91
 The chemical compositions of the brines used for experimental studies are listed in Table 4. 1 and

Table 4. 2.

Table 4. 1 Chemical composition of the natural Mt. Simon brine.

Species Chemical composition of the natural Mt. Simon brine

mg/L mol/kg H2O IS mol/kg H2O
Ba2+ 1.1000 10-1 8.2655 10-7 1.6531 10-6
+1 -3
B 1.3300 10 1.2696 10 6.3479 10-4
2+ +3 -1
Ca 6.8770 10 1.7706 10 3.5413 10-1
2+/3+ +1 -4
Fe 4.2000 10 7.7607 10 1.5521 10-3
2+ +3 -2
Mg 1.2670 10 5.3792 10 1.0758 10-1
Mn2+ 4.9000 9.2036 10-5 1.8407 10-4
K+ 7.1700 10+2 1.8923 10-2 9.4616 10-3
-4
Si 9.0000 3.3068 10 1.6534 10-4
+ +4
Na 2.3317 10 1.0466 5.2329 10-1
2+ +2 -3
Sr 1.6500 10 1.9423 10 3.8864 10-3
2- +2 -2
S 5.6800 10 1.8278 10 3.6557 10-2
- +2 -3
Br 2.4600 10 3.1769 10 1.5884 10-3
- +4
Cl 5.1068 10 1.4864 7.4319 10-1
I- 2.3100 1.8783 10-5 9.3916 10-6
SO42- 1.5110 10+3 1.6230 10-2 3.2459 10-2
-1
Total ionic strength (IS / mol kg ) 1.815
Note: Metals were determined by inductively coupled plasma optical emission spectrometry, anions were
determined by ion chromatography. The element boron (B) and silicon (Si) usually presented in the
aqueous phase as B(OH)4-(aq) and H3SiO4-(aq), so the ionic strengths of (B) and (Si) were treated as the
same as that of. B(OH)4- and H3SiO4-.The density of the Mt. Simon brine at 298 K and 1.013 bar is 1.0549
g/cm3, which was calculated by the empirical correlation (Eq.(4.7)).

Table 4. 2 Chemical compositions of the synthetic Mt.Simon brine and Antrim Shale formation brines and
the corresponding synthetic NaCl+CaCl2 proxy brines.

Salt species Synthetic Mt. Simon formation brine Synthetic Antrim Shale formation brine
/ mol kg-1 H2O /mol kg-1 H2O
Proxy brine I(a) Proxy brine II(b) Proxy brine I(a) Proxy brine II(b)
NaCl 1.0601 1.0601 2.9856 2.9856
CaCl2 0.1365 0.2172 0.3937 0.6661
Na2SO4 0.0165 - 0.0001 -
MgCl2 0.0544 - 0.2535 -
KCl 0.0188 - 0.0222 -
SrCl2 0.0021 - 0.0083 -
NaBr 0.0042 - 0.0090 -
IS 1.712 4.984
(a)
The synthetic formation brines and the corresponding synthetic NaCl+CaCl2 brines are have the same
ionic strength. During the PSUCO2 model calculation, the synthetic formation brine composed of seven
salt species were simplified to five salt species because the PSUCO2 model is not able to take SrCl2 and
NaBr into consideration. The amount of SrCl2 and NaBr in synthetic formation brines are converted to the
equivalent amount of NaCl based on the ionic strength, while the concentration of the other salt species
remains the same.

92
(b)
The proxy NaCl+CaCl2 brine has the same amount of NaCl as the corresponding synthetic formation
brine, while the concentrations of the other salt species are converted to the equivalent amount of CaCl 2
based on the ionic strength.

4.2.3 Apparatus and procedure

Because the experimental CO2 solubility data reported herein were collected at two different

laboratories (PSU Energy Institute and National Energy Technology Laboratory (NETL)) independently,

there are two different methods employed to measure the solubility of CO2 in the synthetic brines and in the

natural Mt. Simon brine. Method 1 (used by PSU Energy Institute): the experimental system consisted of a

600-ml stainless steel autoclave (Parr Instrument Co.), a 40-ml stainless steel sample cell, a liquid CO2

pump and a 300-ml stainless steel pressure cell for sample analysis. After sampling, the dissolved CO2 in a

40-ml sample cell was slowly released to a 300-ml pressure cell. The mass of the dissolved CO2 was

determined by the pressure and temperature change before and after CO2 expansion. Details on the CO2

solubility measuring technique of the Method 1 and the error analysis approach can be found in a previous

publication (See Section 2.3, Chapter 2). Method 2 (used by NETL): the experimental system contains a

flexible gold cell hydrothermal rocking autoclave reactor for gas-liquid equilibration. Fluid samples were

withdrawn directly into glass, gas-tight syringes containing a small quantity of 45% (w/w) KOH, which

prevents the loss of CO2 during sampling and analysis. After sampling, KOH-stabilized samples were

injected into the acidification module of a UIC coulometric carbon analyzer, titrated using a perchloric acid

solution, and the evolved CO2 was transferred via inert gas stream into a coupled coulometric CO 2 titration

cell. The UIC coulometric carbon analyzer is capable of detecting carbon in the range of 0.01 𝜇g to 100 mg,

the detailed experimental procedure can be found in Dilmore et al. (2008).

93
Table 4. 3 Experimental CO2 solubility results in the synthetic formation and NaCl+CaCl2 brines.

Method 1
-1
IS=1.712 mol kg IS=4.984 mol kg-1
T P Synthetic Mt. Simon Synthetic Antrim Shale
/K / bar Synthetic NaCl+CaCl2 brine Synthetic NaCl+CaCl2 brine
formation brine formation brine
mCO2 / err / mCO2 / err / mCO2 / err / mCO2 / err /
mol kg-1 mol kg-1 mol kg-1 mol kg-1 mol kg-1 mol kg-1 mol kg-1 mol kg-1
323 100 0.843 0.013 0.850 0.015 0.535 0.015 0.539 0.028
323 125 0.902 0.014 0.898 0.021 0.573 0.014 0.578 0.010
323 150 0.931 0.012 0.925 0.017 0.604 0.014 0.600 0.012
323 175 0.956 0.012 0.951 0.022 0.605 0.008 0.608 0.009
373 100 0.595 0.007 0.590 0.012 0.376 0.011 0.386 0.006
373 125 0.696 0.009 0.691 0.012 0.433 0.004 0.438 0.008
373 150 0.756 0.010 0.759 0.014 0.483 0.010 0.490 0.015
373 175 0.812 0.013 0.818 0.008 0.506 0.010 0.521 0.011
423 100 0.506 0.010 0.505 0.007 0.326 0.013 0.328 0.008
423 125 0.616 0.013 0.608 0.016 0.394 0.026 0.397 0.011
423 150 0.697 0.012 0.703 0.014 0.446 0.004 0.444 0.019
423 175 0.777 0.023 0.780 0.028 0.484 0.010 0.489 0.019
Method 2
IS=1.815 mol kg-1
T P Natural Mt. Simon formation
/K / bar brine
mCO2 / err /
mol kg-1 mol kg-1
328 50.5 0.488 0.030
328 99.8 0.785 0.036
328 150.4 0.849 0.026
328 200.7 0.927 0.011
Note: The instrumental error of Method 1 is about 0.7%. The chemical compositions of synthetic formation brines and NaCl+CaCl 2 brines are listed in Table 4. 2.
The chemical compositions of the natural Mt. Simon brine are listed in Table 4. 1.

94
4.3 Results and discussions

4.3.1 Experimental results

Experimental CO2 solubility data in synthetic formation brine, synthetic binary NaCl+CaCl 2 brine,

and the natural Mt. Simon brine of this study are listed in Table 4. 3.

4.3.2 Additivity rule of Setschenow Coefficients of the individual ions

The salting-out effects of a salt mixture with two or more salt species on a non-electrolyte have

been shown to be accurately additive at ambient conditions (Gordon and Thorne, 1967):

𝑜
𝑙𝑜𝑔 𝐶𝐶𝑂2 = 𝑙𝑜𝑔 𝐶𝐶𝑂2
− ∑𝑛𝑖=1 𝑁𝑠𝑖 𝐾𝑠𝑖 𝐶𝑡 (4.1)

𝑜
where 𝐶𝐶𝑂2 and 𝐶𝐶𝑂2
are, respectively, the CO2 molarity (mole per liter of solution) in a mixed-salt solution

and in pure water at the same temperature and pressure. 𝑁𝑖 is the fraction of the i-th salt species defined by

𝑁𝑖 = 𝐶𝑠𝑖 /𝐶𝑡 , with 𝐶𝑡 = ∑𝑛𝑖=1 𝐶𝑠𝑖 , where 𝐶𝑠𝑖 and 𝐶𝑡 are the molarity of the i-th salt species and the total molar

concentration of all dissolved salts, respectively. 𝐾𝑠𝑖 in Eq. (4.1) is the Setschenow coefficient for the CO2

solubility in a single-salt aqueous solution. 𝐾𝑠𝑖 can be evaluated by the classical Setschenow equation34 as

below:

𝑜
1 𝐶𝐶𝑂
𝐾𝑠𝑖 = 𝑙𝑜𝑔 𝑖
2
(4.2)
𝐶𝑠 𝐶𝐶𝑂
2

𝑜 𝑖
where 𝐶𝐶𝑂2
is the CO2 molarity in pure water, and 𝐶𝐶𝑂2
is the CO2 molarity in the single-salt aqueous

solution with i-th salt species. The Setschenow coefficient of a given salt species may be separated by the

contribution of individual ions as below (Görgényi et al., 2006):

95
 𝑖,𝑐 𝑖,𝑎
𝐾𝑠𝑖 = 𝜈 + 𝐾𝑠,𝑖𝑜𝑛 + 𝜈 − 𝐾𝑠,𝑖𝑜𝑛 (4.3)

𝑖,𝑐 𝑖,𝑎
where 𝐾𝑠,𝑖𝑜𝑛 and 𝐾𝑠,𝑖𝑜𝑛 are the separated Setschenow coefficients for cation and anion of the i-th salt

species; and 𝜈 + and 𝜈 − are the stoichiometric coefficients of the corresponding cation and anion,

respectively. In PSUCO2, an aqueous mixed-salt solution is treated as the mixture of the six types of ions

(Na+, Ca2+, Mg2+, K+, Cl- and SO42-) and H2O molecules. Other ions in a natural brine, such as Ba 2+, Sr2+,

Fe3+ and Br-, were neglected during the calculation due to their typically low concentrations in natural

solutions. Therefore, for a brine with a salt mixture of NaCl, CaCl 2, Na2SO4, MgCl2 and KCl, Eq. (4.3) can

be rewritten as below:

𝑁𝑎 +𝐶𝑙 −
𝐾𝑠𝑁𝑎𝐶𝑙 = 𝐾𝑠,𝑖𝑜𝑛 + 𝐾𝑠,𝑖𝑜𝑛

𝐶𝑎𝐶𝑙2 𝐶𝑎 2+ 𝐶𝑙 −
𝐾𝑠 = 𝐾𝑠,𝑖𝑜𝑛 + 2𝐾𝑠,𝑖𝑜𝑛

𝑁𝑎2 𝑆𝑂4 𝑁𝑎 + 𝑆𝑂2−

𝐾𝑠 = 2𝐾𝑠,𝑖𝑜𝑛 + 𝐾𝑠,𝑖𝑜𝑛
4

𝑀𝑔𝐶𝑙2 𝑀𝑔2+ 𝐶𝑙 −
𝐾𝑠 = 𝐾𝑠,𝑖𝑜𝑛 + 2𝐾𝑠,𝑖𝑜𝑛

𝐾 + 𝐶𝑙 −
𝐾𝑠𝐾𝐶𝑙 = 𝐾𝑠,𝑖𝑜𝑛 + 𝐾𝑠,𝑖𝑜𝑛

𝐶𝑙 −
𝐾𝑠,𝑖𝑜𝑛 =0 (4.4)

−
In Eq. (4.4), the Cl- was chosen as a reference ion by assigning 𝐾𝑠C𝑙 equals zero. Given that

Setschenow coefficient (𝐾𝑠𝑖 ) for CO2 solubility of each single-salt solution is known by Eq. (4.2), the

+ 2+ 𝑀𝑔2+ + 𝑆𝑂42−
Setschenow coefficients for the individual ions (𝐾𝑠𝑁𝑎 , 𝐾𝑠𝐶𝑎 , 𝐾𝑠 , 𝐾𝑠𝐾 , and 𝐾𝑠 ) can be computed

by solving the set of linear equations (Eq. (4.4)). The Setschenow coefficient (𝐾𝑠𝑀𝑖𝑥 ) on CO2 solubility due

to the addition of mixed salts in aqueous solutions can be calculated using the additivity rule for individual

ions as below:

𝐾𝑠𝑀𝑖𝑥 = ∑𝑛𝑖=1 𝑁𝑠,𝑖𝑜𝑛

𝑖 𝑖
𝐾𝑠,𝑖𝑜𝑛 (4.5)

96
 𝑖
where 𝐾𝑠𝑀𝑖𝑥 is the Setschenow coefficient of a salt mixture; 𝑁𝑠,𝑖𝑜𝑛 is the fraction of the i-th ion defined by

𝑖 𝑖
𝑁𝑠,𝑖𝑜𝑛 = 𝐶𝑠,𝑖𝑜𝑛 /𝐶𝑡,𝑖𝑜𝑛 and 𝐶𝑡,𝑖𝑜𝑛 = ∑𝑛𝑖=1 𝐶𝑠,𝑖𝑜𝑛
𝑖 𝑖
, where 𝐶𝑠,𝑖𝑜𝑛 is the molar concentration of the i-th ion, and

𝐶𝑡,𝑖𝑜𝑛 is the sum of the molar concentration of all ions dissolved in the solution. Once 𝐾𝑠𝑀𝑖𝑥 is obtained, the

CO2 molarity in a real brine can be calculated by Eq. (4.6) as below (Gordon and Thorne, 1967)

𝑜
𝑙𝑜𝑔 𝐶𝐶𝑂2 = 𝑙𝑜g 𝐶𝐶𝑂2
− 𝐾𝑠𝑀𝑖𝑥 𝐶𝑡,𝑖𝑜𝑛 (4.6)

For a given brine composed of the five most frequently encountered salt species (NaCl, CaCl2,

Na2SO4, MgCl2, and KCl), the calculated results from Eq. (4.6) and Eq. (4.1) are the same. However, the

concentration of a natural brine is usually reported by ion molarity (mol L -1), thus Eq. (4.6) provides a more

convenient way for users to input a brine composition. The schematic of the additivity rule of the

Setschenow coefficients is illustrated in Figure 4. 1.

Ks1
2 1

NaCl(aq) Brine density

CO
m

2
2
Ks2
m CO CaCl2(aq)
n
logCCO2  logCCO2   KsCs
0 i i
3
P, T and Brine mCO2 Ks3 i1
Na2SO4(aq)
composition
m
CO 4
2 Ks4
MgCl2(aq)
mC

CO2 solubility in
Ks5
O
5
2

KCl(aq) formation brine

Figure 4. 1 The schematic of the additivity rule of the Setschenow coefficients. At a given P-T-x condition,
firstly, the PSUCO2 model calculate the CO2 solubility (mCO2: mol kg-1) in each single salt aqueous
solution; secondly, the Setschenow coefficients (Ksi) for each single salt species on CO2 solubility can be
calculated by using Eq. (4.3); finally the additivity rule of the single salt (or single ion) Setschenow
coefficients is used to compute the CO2 solubility in mixed-salt solutions.

97
 The calculated CO2 molarity (𝐶𝐶𝑂2 ) can be converted to CO2 molality (𝑚𝐶𝑂2 ) by using the density

correlations of the different types of brine. The density of brine at 298K and 1.013 bar can be calculated by

Eq. (4.7).

𝜌𝑏𝑟𝑖𝑛𝑒 (𝑔⁡𝑐𝑚−3 ) = 𝑎𝑥 2 + 𝑏𝑥 + 𝑐 (4.7)

where coefficients a, b and c are constants (Listed in Table 4. 4) dependent on the specific brine, and they

were determined by fitting the literature experimental density of different brines (McIntosh et al., 2004;

Pitzer et al., 1984; Gates and Wood, 1985; Ghafri et al., 2012; and Dresel and Rose, 2010).

Table 4. 4 Brine density (g cm-3) correlation (Eq. (4.7)) at 298K and 1 bar.

Brine x(a) a b c
NaCl m, mol kg-1 -1.08 10-3 3.9376 10-2 9.9777 10-1
CaCl2 m, mol kg-1 -3.371 10-3 8.4788 10-2 1.0049
-1
Na2SO4 m, mol kg -8.2 10-3 1.2301 10-1 9.9775 10-1
-1
MgCl2 m, mol kg -3.967 10-3 7.8024 10-2 9.9673 10-1
KCl m, mol kg-1 -1.574 10-3 4.4934 10-2 9.9791 10-1
-1
Mixed-salt TDS, g L 0 6.5251 10-4 1.00097
(a)
m: salt molality, mol kg-1; TDS: total dissolved solids, g L-1.

In addition, at the pressures larger than 100 bar, an error correction term was used in the PSUCO2

model to compensate for the calculation error at pressures greater than 100 bar (Eq. (4.7)). The correction

term 𝛿𝑚𝐶𝑂2 is about 0-5% when compared to the calculated CO2 solubility (𝑚𝐶𝑂2 : mole CO2 per kilogram

of H2O) in the aqueous phase.

𝛿𝑚𝐶𝑂2 = 0⁡⁡(𝑃⁡ ≤ 100⁡𝑏𝑎𝑟)

𝑎2
𝛿𝑚𝐶𝑂2 = 𝑎1 + + 𝑎3 𝐼 0.5 + 10−3 𝑎4 𝑇 + 𝑎5 𝑚𝐶𝑂
′
2
⁡(𝑃⁡ > ⁡100⁡𝑏𝑎𝑟) (4.8)
𝑇−𝛩

The form of Eq. (4.8) is the same as the expression of Setchenow coefficients proposed by

Akinfiev and Diamond (2010). In Eq. (4.7), I is the total ionic strength (mol kg-1) of the brine, 𝛩 is a
′
constant, equal to 228 according to Akinfiev and Diamond (2010). T is temperature in K, and 𝑚𝐶𝑂2
is the

98
computed CO2 molality without any correction. The parameters a1 to a5 (given in Table 4. 5) were

determined by the least-squares fitting of the errors between the experimental CO2 solubility data (this

study; Li et al., 2004; Liu et al., 2011; Tong et al., 2013) and the PSUCO2 model calculated values.

Table 4. 5 The parameters of Eq. (4.8)

Parameter(a) value
a1 -0.6610
a2 17.4446
a3 1.0699
a4 0.0614
a5 0.0712
(a)
The parameter are determined only using the experimental data (Yasunishi et al., 1979; Li et al., 2004;
Liu et al., 2011; Tong et al., 2013)

As a result, the CO2 solubility (𝑚𝐶𝑂2 ) in the mixed-salt solutions can be obtained by adding the

correction term as below,

′
𝑚𝐶𝑂2 = 𝑚𝐶𝑂2
+ 𝛿𝑚𝐶𝑂2 (4.9)

4.3.3 Comparison of model calculations against the experimental data

The reliability of the experimental technique of Method 1 is confirmed in Chapters 2 and 3. Figure

4. 2 shows that the experimental data collected using Method 2 are in good agreement with the previously

published models (DS2006, SP2010, and OLI Studio 9.0.6). In Figure 4. 2, the experimental data collected

by Method 2 were slightly lower than the predicted values of the PSUCO2 model. There can be many

reasons for the small difference between the experimental CO2 solubility in natural Mt. Simon formation

brine and the modeling results. Firstly, the natural Mt. Simon formation brine contains some ions (e.g. Ba 2+,

Fe2+/3+, Sr2+, Br-, etc.) that is neglected in the PSUCO2 model and therefore the salting-out effect of these

ions on CO2 solubility are not accounted during the model calculations. Secondly, the influence of the

initial pH on the CO2 solubility in the aqueous phase is not considered during the model calculations. The

pH and composition of the natural Mt. Simon brine maybe changed after initial analysis, this variation in

99
natural brine can also leads to slightly different experimental results. Thirdly, a small difference between

experimental and modeling results can also result from both the experimental uncertainty and the

inaccuracy of the model calculations. Considering the two sets of experimental data were collected at

different temperatures (323 and 328K), brine concentrations, and even using different approaches (Method

1 and Method 2), based on the comparison between the experimental and modeling results (Figure 4. 2), the

Method 2 considered to be reliable for measuring CO 2 solubility in brine at the P-T-x conditions pertinent

to geologic CO2 storage. Figure 4. 2b shows the proposed PSUCO2 model is capable of predicting the

experimental results both in synthetic Mt. Simon brine (Method 1) and the natural Mt. Simon brine

(Method 2).

Sodium chloride (NaCl) is the dominant component of most reservoir brines and therefore the

single-salt aqueous NaCl solution is widely used as a proxy for natural formation brine for CO 2 solubility

studies. In this study, the experimental results corroborates that NaCl+CaCl 2 brine can be used as a proxy

for studying CO2 solubility in synthetic (or natural) formation brines (Figure 4. 3(a,b)), provided that the

same NaCl molality and the total ionic strength for both brines. As for NaCl+CaCl 2 and NaCl-only brines,

in order to identify a more realistic surrogate for studying CO 2 solubility in natural formation brines, a

comparison of the experimental CO2 solubilities between the two brines (NaCl+CaCl2 and NaCl-only) with

the same ionic strength is needed.

Table 4. 6 The average absolute deviation of the PSUCO2 model calculated CO2 solubility in NaCl+CaCl2
and NaCl brines from the experimental CO2 solubility data in synthetic Mt. Simon and Antrim Shale
formation brines.

AAD % (Compared to the experimental CO2 solubility

Experimental CO2 data in synthetic formation brine)
solubility in synthetic T/K PSUCO2 Experimental
PSUCO2(a)
formation brine (NaCl+CaCl2+MgCl2 data(b)
(NaCl–only)
+Na2SO4+KCl) (NaCl+CaCl2)
Synthetic Mt. Simon 323 0.57 3.34 0.61
formation brine, IS=1.712 373 2.34 1.29 0.67
mol/kg 423 0.59 0.98 0.69
Synthetic Antrim Shale 323 1.41 6.43 0.69
formation brine, IS=4.984 373 2.58 0.65 2.05
mol/kg 423 1.38 1.14 0.71
Overall AAD % 1.48 2.31 0.91
(a)
The PSUCO2 generated CO2 solubility data in aqueous NaCl solution provides comparable accuracy with
the corresponding experimental data.20
(b)
The AAD in column 5 is calculated between the experimental CO2 solubility in NaCl+CaCl2 and
synthetic formation brines

100
 1.2
1.1 328 K, IS=1.815 mol/kg
Natural Mt. Simon Brine
1.0
-1 0.9
mCO2 / mol kg
0.8
0.7
0.6
0.5 PSUCO2
DS2006
0.4 SP2010
OLI Studio 9.0
0.3 This study, Method 2
0.2
40 60 80 100 120 140 160 180 200 220
P / bar

(a)

1.2
1.1 Synthetic Mt. Simon brine
at 323K and IS=1.712 mol/kg
1.0
0.9
-1
mCO2 / mol kg

0.8
0.7 Natural Mt. Simon brine
at 328K and IS=1.815 mol/kg
0.6
0.5
323 K, 1.712 mol/kg, PSUCO2
0.4 328 K, 1.815 mol/kg, PSUCO2
323 K, 1.712 mol/kg, Method 1
0.3 328 K, 1.815 mol/kg, Method 2
0.2
40 60 80 100 120 140 160 180 200 220
P / bar

(b)

Figure 4. 2(a, b) Comparison of model calculations with experimental data. Figure. 4.2a: Comparison of
model calculations (PSUCO2, DS2006, SP2010 and OLI Studio 9.0.6) with experimental CO2 solubility
measured by Method 2. For the experimental data collected by Method 2, the synthetic brine recipe used
for model calculations is: 1.0547 mol/kg NaCl, 0.1771 mol/kg CaCl 2, 0.0162 mol/kg Na2SO4, 0.0538
mol/kg MgCl2, 0.0189 mol/kg KCl (IS=1.815). Figure 4.2b: Comparison of the PSUCO2 modeling results
with our experimental CO2 solubility data in natural Mt. Simon brine (IS=1.712 mol kg -1) at 323 K and in

101
synthetic Mt. Simon brine (IS=1.815 mol kg-1) at 328 K. For the experimental data collected by Method 1,
the synthetic brine recipe used for PSUCO2 calculation is listed in Table 4. 2(IS=1.712 mol kg-1).

1.05
Syn. Mt. Simon and NaCl+CaCl2 brines
IS=1.712 mol/kg

0.85
-1

3K
mCO2 / mol kg

32

0.65

PSUCO2
3K
37 SP2010
0.45 DS2006
OLI Studio 9.0
3K
42 Syn. Mt. Simon brine
Syn. NaCl+CaCl2 brine
0.25
60 80 100 120 140 160 180 200
P / bar

(a)

0.7
Syn. Michigan Basin and NaCl+CaCl2 brines
IS=4.984 mol/kg
0.6
-1

0.5
mCO2 / mol kg

3K
32
0.4

3K
0.3 37 PSUCO2
SP2010
DS2006
3K
0.2 42 OLI Studio 9.0
Syn. Michigan Basin brine
Syn. NaCl+CaCl2 brine
0.1
60 80 100 120 140 160 180 200
P / bar

(b)

Figure 4. 3(a, b) Comparisons of the modeling results against the experimental CO 2 solubility data in both
synthetic reservoir brines and NaCl+CaCl2 brines collected using the Method 1. (a) IS=1.712 mol kg-1; (b)
IS=4.984 mol kg-1.

102
 0.035
PSUCO2
SP2010
0.030 DS2006
OLI Studio 9.0
CaCl2+KCl (1:3)
-1
mCO2 / mol kg 0.025 CaCl2+KCl (1:1)
CaCl2+KCl (3:1)

0.020

0.015 298.15 K and 1.013 bar

Syn. CaCl2+KCl brine
Yasunishi et al. (1979)
0.010
0 50 100 150 200 250 300 350
TDS / g L-1

(a)

0.036
OLI Studio 9.0
PSUCO2
0.032 SP2010
DS2006
NaCl+Na2SO4 (1:3)
-1

0.028 NaCl+Na2SO4 (1:1)

mCO2 / mol kg

NaCl+Na2SO4 (3:1)
0.024

0.020

298.15 K and 1.013 bar

0.016 Syn. NaCl+Na2SO4 brine
Yasunishi et al. (1979)
0.012
0 50 100 150 200 250
TDS / g L-1

(b)

103
 0.036
PSUCO2
SP2010
0.032 DS2006
OLI Studio 9.0
NaCl+KCl (1:3)

-1
0.028 NaCl+KCl (1:1)

mCO2 / mol kg
NaCl+KCl (3:1)
0.024

0.020

0.016 298.15 K and 1.013 bar

Syn. NaCl+KCl brine
Yasunishi et al. (1979)
0.012
25 75 125 175 225 275
TDS / g L-1

(c)

0.032
PSUCO2
SP2010
DS2006
0.028 OLI Studio 9.0
NaCl+CaCl2+KCl (2:1:1)
-1
mCO2 / mol kg

0.024

0.020
298.15 K and 1.013 bar
Syn. NaCl+CaCl2+KCl brine
Yasunishi et al. (1979)
0.016
10 30 50 70 90 110
-1
TDS / g L

(d)

Figure 4. 4(a-d) Comparison of model calculations against the experimental data reported by Yasunishi et
al. (1979) at 1 atm and 298 K (a) synthetic CaCl2+KCl brine; (b) synthetic NaCl+Na2SO4 brine; (c)
synthetic NaCl+KCl brine; (d) synthetic NaCl+CaCl2+KCl brine. The ratios of the salt concentration in the
figure are based on the molarity (mole per liter of solutions).

104
 1.2
332 K, Weyburn Formation brine
1.0

-1
0.8

mCO2 / mol kg 0.6

0.4 PSUCO2
SP2010
OLI Studio 9.0
0.2 DS2006
Li et al. (2004)
0.0
0 50 100 150 200 250
P / bar

Figure 4. 5 Comparisons of the model calculations (PSUCO2, SP2010 and OLI) against the experimental
CO2 solubility data7 in the Weyburn Formation brine. The chemical composition of the Weyburn Formation
brine is: 1.3022 mol kg-1 Na+, 0.0505 mol kg-1 Ca2+, 0.0239 mol kg-1 Mg2+, 0.0119 mol kg-1 K+, 1.5255 mol
kg-1 Cl- and 0.0406 mol kg-1 SO42-.

However, in this study, we use the modeling results in place of the experimental CO 2 solubility in

aqueous NaCl-only solutions to save effort and time from additional experimental work for the CO 2-NaCl-

H2O system, because (1) the CO2 solubility in aqueous NaCl solution is well-understood at the P-T-x

conditions of interest (See Chapter 2) it can be accurately calculated by either previously published models

(DS2006, AD2010 (Akinfiev and Diamond, 2010), SP2010, and OLI Studio 9.0.6) or the PSUCO2 model,

and (2) we believe that the PSUCO2 calculated CO2 solubility in single salt aqueous NaCl solutions

(Column 4, Table 4. 6) provides comparable accuracy (average absolute deviation (AAD)<1%) to the

corresponding experimental data in aqueous NaCl solutions at 323-423 K, 100-175 bar and 0-6 mol kg-1

NaCl (See Chapter 2).

The overall AAD of the experimental CO2 solubility in NaCl+CaCl2 brine (Column 5, Table 4. 6)

and the PSUCO2 modeling results in NaCl–only brine (Column 4, Table 4. 6) from the experimental CO2

solubility data in the synthetic formation brines are 0.91% and 2.31%, respectively. This comparison

demonstrates that the NaCl+CaCl2 brine performs better than NaCl–only brine as a surrogate for studying

CO2 solubility in synthetic (or natural) formation brines. Therefore, for geological carbon storage and CO 2-

105
EGS applications, the synthetic NaCl+CaCl2 brine offers an accurate, accessible and inexpensive

alternative to researchers who may experience difficulty securing and preserving natural brine samples.

While each of the previously published CO2 solubility model is able to predict the CO2 solubility

in synthetic formation brines with a very good accuracy (AAD<3%) at the experimental P-T-x region

(Figure 4. 3), Figure 4. 4 demonstrates that the SP2010 and DS2006 models does not match the data from

Yasunishi et al. (1979), especially for the case of CaCl2+KCl (Figure 4. 4a) and NaCl+Na2SO4 (Figure 4.

4b). This difference arises from the fact that the brines used by Yasunishi et al. (1979) have a high

concentration of K+, SO42-. It is difficult for the DS2006 and SP2010 model to perform well for mixed-salt

aqueous solutions with relatively high concentrations of Na 2SO4 or KCl, or both of them (See Chapter 3)

The commercial software OLI Studio 9.0.6 performs much better than the DS2006 and SP2010 models in

predicting the CO2 solubility in aqueous solutions with high concentration of K+ and SO42- ions (See

Chapter 3). However, naturally encountered brine is usually dominated by NaCl, CaCl 2, and MgCl2, plus a

small amount of K+ and SO42- ions (McIntosh et al., 2004) the influences of K+ and SO42- ions on the CO2

solubility in commonly found natural brines are small. Therefore, although there is a lack of capacity to

deal with K+ and SO42- ions, the SP2010 and DS2006 models can still predict the experimental CO2

solubility data (Li et al., 2004) in typical formation brine with quite good accuracy as shown in Figure 4. 2a

and Figure 4. 5.

In Figure 4. 6a, with the same percent of salts by weight, among the systems of NaCl+KCl,

NaCl+CaCl2 and CaCl2+KCl, the experimental CO2 solubility in CaCl2+KCl aqueous solution has the

greatest value, whereas the CO2 solubility in the NaCl+CaCl2 solution has the smallest values (Liu et al.,

2011) The particular pattern on the magnitude of the CO2 solubility in these three mixed-salt solutions was

also shown in the calculated results of PSUCO2, but was not preserved in the results of the other models

(DS2006, SP2010 and OLI Studio 9.0.6). Figure 4. 6(b-d) shows that the DS2006 and SP2010 models tend

to systematically underpredict the experimental CO2 solubility given by Liu et al. (2011) more than does

PSUCO2. Additionally, at the pressures below 100 bar, the calculation results between the OLI Studio

9.0.6 and PSUCO2 are close to each other, at the higher pressures (P>100 bar), although the OLI Studio

9.0.6 tend to underpredict the experimental data at 308 K, it still performs better than the DS2006 and

SP2010 models for aqueous mixed-salt solutions.

106
 1.1
Liu et al. (2011)
1.0
318.15 K
0.9
-1 0.8 OLI Studio 9.0
mCO2 / mol kg SP2010
NaCl+KCl
NaCl+KCl
CaCl2+KCl
0.7 DS2006 CaCl2+KCl
NaCl+CaCl2
NaCl+KCl NaCl+CaCl2
0.6 CaCl2+KCl
PSUCO2
NaCl+CaCl2
0.5 PSUCO2
SP2010
CaCl2+KCl DS2006
0.4 NaCl+KCl OLI Studio 9.0
NaCl+CaCl2 NaCl+KCl
0.3 NaCl+CaCl2
CaCl2+KCl
0.2
20 40 60 80 100 120 140 160 180
P / bar

(a)

1.1
Liu et al. (2011)
14.3 wt% syn. NaCl+CaCl2+KCl brine
0.9
-1
mCO2 / mol kg

0.7

0.5 PSUCO2
SP2010
DS2006
OLI Studio 9.0
0.3 0.9518 mol/kg NaCl 308.15 K
0.5012 mol/kg CaCl2 318.15 K
0.2408 mol/kg KCl 328.15 K
0.1
0 50 100 150 200
P / bar

(b)

107
 1.5
Liu et al. (2011)
1.3 5 wt% syn. NaCl+CaCl2+KCl brine

1.1

-1
mCO2 / mol kg
0.9
PSUCO2
0.7 SP2010
DS2006
0.5 OLI Studio 9.0
0.3002 mol/kg NaCl 308.15 K
0.1581 mol/kg CaCl2 318.15 K
0.3
0.2353 mol/kg KCl 328.15 K

0.1
0 50 100 150 200
P / bar

(c)

1.3
Liu et al. (2011)
10 wt% syn. NaCl+CaCl2+KCl brine
1.1
-1

0.9
mCO2 / mol kg

0.7
PSUCO2
SP2010
0.5 DS2006
OLI Studio 9.0
0.6338 mol/kg NaCl 308.15 K
0.3 318.15 K
0.3337 mol/kg CaCl2
0.4968 mol/kg KCl 328.15 K
0.1
0 50 100 150 200
P / bar

(d)

Figure 4. 6(a-d) Comparison of the model calculations against the experimental data given by Liu et al.
(2011) for the synthetic brines at elevated temperatures and pressures. (a) results for the NaCl+KCl,
NaCl+CaCl2 and CaCl2+KCl brines; (b) results for 14.3 wt% NaCl+CaCl 2+KCl brine; (c) results for 5wt%
NaCl+CaCl2+KCl brine; (d) results for 10wt% NaCl+CaCl2+KCl brine.

108
 In addition, the calculated CO2 solubility of all available models (PSUCO2, DS2006, SP2010 and

OLI Studio 9.0.6) were compared at 1-600 bar and at 523 K in Figure 4. 7(a, b). The comparison showed

that over the P-T-x range of 1-600 bar, 523 K, and 0-5 mol/kg ionic strength, all models provide roughly

comparable results. Among these four models, only the PSUCO2 and DS2006 models can work at the

pressure range of 600-2000 bar, but the differences between the two models become quite large at higher

pressures for the formation brines. To the best of our knowledge, there is no experimental data available for

mixed-salt solutions to make a direct comparison between the experimental data and model calculations.

The experimental CO2 solubility in formation brines at high temperatures and pressures (e.g. 523 K, 1800

bar) is needed to resolve this discrepancy. However, the performance of the PSUCO2 and DS2006 models

for the CO2-H2O system at high temperatures and pressures were validated by the corresponding

experimental data (Tödheide and Franck, 1963; Takenouchi and Kennedy, 1965; Malinin, 1959).

12
OLI Studio 9.0
PSUCO2
DS2006
10 SP2010
Takenouchi and Kennedy (1965)
Tödheide and Franck (1963)
-1

8 Malinin (1959)
mCO2 / mol kg

r
wate
re 12
Pu = 1.7
6 e IS
brin
on
t . Sim
4 M
4.984
e IS =
n Basin brin
a
2 Michig

600 bar 523.15 K

0
0 500 1000 1500 2000
P / bar

(a)

109
 5
OLI Studio 9.0
PSUCO2
DS2006
4 SP2010
Takenouchi and Kennedy (1965)

-1
Tödheide and Franck (1963)

mCO2 / mol kg
3 Malinin (1959)

712
a ter I S = 1.
w rine
2 re
Pu imon b
S
Mt.
4.984
asin brine IS =
1 nB
Michiga

523.15 K
0
100 200 300 400 500 600
P / bar

(b)

Figure 4. 7(a, b) Comparison of PSUCO2 with the DS2006 and SP2010 models up to 2000 bar at 523K:
The experimental CO2 solubility in pure water are taken from literature (Tödheide and Franck, 1963;
Takenouchi and Kennedy, 1965; Malinin, 1959).

DS2006 6.4%

SP2010 6.3%

OLI Studio 9.0 5.0%

PSUCO2 4.3%

0 2 4 6 8
AAD %

Figure 4. 8 The average absolute deviation (AAD %) of the calculated CO2 solubility in aqueous mixed-
salt solutions among different models compared to the experimental data (Table 4. 7).

110
Table 4. 7 The absolute average deviations (AAD %) of model calculations from the experimental data.

References Systems P/ T/ IS / N(a) Model calculation(b), AAD %

No. bar K mol kg-1 PSUCO2 SP2010 DS2006 OLI Studio
22 (c) Seawater 1-45.6 273-298 0.73 37 8.52 (37) 12.46 (24) 13.09 (31) 9.48(37)
KCl+CaCl2
NaCl+Na2SO4
23 1.013 298 0.4-9.1 54 2.05(54) 8.66 (47) 8.26 (51) 3.84(54)
NaCl+KCl
KCl+NaCl+CaCl2
24 (d) Utsira porewater 80-120 291-353 0.57 37 13.45 (37) 14.22 (37) 13.66 (37) 15.30(37)
25 Weyburn formation brine 17.6-208.7 332 1.58 6 7.39 (6) 7.74 (6) 7.75 (6) 7.45(6)
NaCl+CaCl2
CaCl2+KCl
26 13-160 308-328 1.0-3.2 99 5.52 (99) 8.98 (99) 10.87 (99) 5.40(99)
KCl+NaCl
NaCl+CaCl2+KCl
27 NaCl+KCl 10.7-171.6 309-425 1.05 14 4.50 (14) 5.91 (14) 5.53 (14) 5.05(14)
NaCl+CaCl2
This study NaCl+CaCl2+Na2SO4+MgCl2 100-175 323-423 1.71-4.98 48 1.72 (48) 2.82 (48) 2.94 (48) 1.93(48)
+KCl
Natural Mt. Simon formation
This study 50-200 328 1.815 4 4.74 (4) 3.84 (4) 3.15 (4) 6.13(4)
brine
*
Overall AAD 4.3(225) 6.3(218) 6.4(222) 5.0(225)
(a)
N is the total number of experimental data reported in the literature.
(b)
Model calculation results were shown as average absolute deviation (AAD), the number in the parentheses represents the actual number of experimental data
𝑐𝑎𝑙𝑐 𝑒𝑥𝑝
100 𝑁 𝑚𝐶𝑂 −𝑚𝐶𝑂 ,𝑖
𝑝 2 ,𝑖 𝑐𝑎𝑙𝑐
evaluated by each model. The average absolute deviation (AAD) is defined as:𝐴𝐴𝐷(%) = ∑𝑖=1 | 𝑒𝑥𝑝
2
| %, where 𝑚𝐶𝑂2 ,𝑖
is the model calculated
𝑁p 𝑚𝐶𝑂 ,𝑖
2
𝑒𝑥𝑝
results; 𝑚𝐶𝑂2 ,𝑖 is the experimental CO2 solubility taken from literature; and Np represents the total number of experimental data evaluated for each work.
(c)
The data from Stewart and Munjal22 were excluded from the overall AAD due to all the models are significant deviate from the experimental data
(d)
The data from Rochelle and Moore24 were excluded from the overall AAD due to: 1) the random error of the repeated measurements at the same P-T-x
condition ranging from 2 to 20% so there is a large uncertainty of the experimental work; 2) all the models are significant deviate from the experimental data.

111
Table 4. 8 The components, key parameters, and the limitations of PSUCO2 (Modified from Table 2. 9).

No. Model Description Applicable Ref.

Components P-T range
273-573K
1 Modified RK EoS Spycher et al. (2003) modified RK EoS Spycher et al. (2003)
1-2000 bar
Panagiotopoulos and Reid
273-573K
2 Mixing rule A two-parameter mixing rule for polar molecules (1986)
1-2000 bar
Spycher et al. (2003)
273-573K
3 kH2O-CO2 Mixing rule parameter for the CO2-H2O system Spycher et al. (2003)
1-2000 bar
Empirical determined mixing rule parameter.
1. T<473 K, kCO2-H2O was considered as only temperature-
kCO2-H2O 273-573K
4 dependent. Table 2. 8
1-2000 bar
2. T>473 K, kCO2-H2O is the function of both temperature and
pressure.
Spycher et al. (2003)
Fugacity coefficients of H2O and CO2 for the CO2-rich 273-573K
5 𝜑𝐻2𝑂 , 𝜑𝐶𝑂2 Panagiotopoulos and Reid
phase 1-2000 bar
(1986)
The Henry's law constant of CO2 in H2O is used for phase 273-573K Fernández-Prini et al.(2003)
6 Henry's law constant
equilibrium calculations 1-2000 bar Eq. (2.3), Table 2. 5
238-873K, 0-1.2 104 bar
IAPWS-95 EoS for pure water
The fugacity coefficient, vapor pressure, molar volume, for 𝜖𝑟 , the others
7 𝜑𝑤𝑠 , Ps , 𝑣𝑠 ,⁡𝜖𝑟 , 𝜌𝐻2𝑂 (Wagner and Pruß, 2002)
relative permittivity and density of pure water 251-1273K
Fernández et al. (1997)
up to 104 bar
Tödheide and Franck's (1963) data were used to determine
273-573K
8 𝑣̅𝐶𝑂2 the values of 𝑣̅𝐶𝑂2 at temperatures from 323-573K and Table 2. 4
1-2000 bar
pressures from 200-2000 bar
Calculating the activity coefficient of dissolved CO2 and 273-573K
Akinfiev and Diamond (2010)
9 Pitzer activity model salts, and osmotic coefficient of water, at elevated 1-2000 bar
Section 2.4.2 and Section 3.3
temperatures and pressures.
10 𝜆𝑛𝑛 , 𝜇𝑛𝑛𝑛 Pitzer pure neutral components interaction parameters 273-573K Table 2. 6

112
No. Model Description Applicable Ref.
Components P-T range
273-573K
Holmes et al. (1994)
1-1000 bar
The Pitzer binary ion-ion parameters for the binaryH2O-salt Rogers and Pitzer (1981)
(0) (1) 0-6 mol kg-1 of NaCl,
11 𝛽𝑐𝑎 , 𝛽𝑐𝑎 , 𝐶𝑐𝑎 systems at elevated temperatures and pressure were taken Rumpf and Maurer (1993)
CaCl2, and MgCl2
from literature. Wang and Pitzer (1998)
0-2 mol kg-1 of Na2SO4
Holmes and Mesmer (1983)
0-4.7 mol kg-1 of KCl
𝐵𝐶𝑂2 −𝑠𝑎𝑙𝑡 The combined Pitzer interaction parameters for the CO2- Akinfiev and Diamond (2010)
12 273-573K
𝐶𝐶𝑂2 −𝐶𝑂2 −𝑠𝑎𝑙𝑡 salt-H2O system Table 3. 1
Determined 323, 373,
Pitzer's triple-ion interaction parameter determined by the
423K Table 2. 7
13 𝜉𝑛𝑐𝑎 experimental data given by Zhao et al. (2014a) and Zhao et
Extrapolated Table 3. 2
al. (2014b).
273-573K
Setschenow coefficients for individual ions, mixed-salts of
𝑎 𝑐 273-573K
14 𝐾𝑠,𝑖𝑜𝑛 , 𝐾𝑠,𝑖𝑜𝑛 , 𝐾𝑠𝑀𝑖𝑥 CO2 solubility in brines at elevated temperatures and Chapter 4
1-2000 bar
pressures
The error correction term for CO2 solubility in mixed-salt
P>100 bar
15 𝛿𝑚𝐶𝑂2 solutions when P > 100 bar. This term is generally around Chapter 4
Determined at 323-423K
5% of the calculated CO2 molality at high P-T-x reigons.
The additivity rule of 𝑎 𝑐 273-573K
16 Computing 𝐾𝑠𝑀𝑖𝑥 from 𝐾𝑠,𝑖𝑜𝑛 and 𝐾𝑠,𝑖𝑜𝑛 This study, Eqs. (4.1) and (4.6)
Setschenow coefficients 1-2000 bar
Solving phase The phase equilibria equations were solved by the Netwon- 273-573K
17 Figure 2. 7
equilibria equations Raphson approach with iterations. 1-2000 bar

113
 In sum, by using additivity rule of Setschenow Coefficients of the individual ions, the PSUCO2

model demonstrates an improvement (Table 4. 7 and Figure 4. 8) relative to the previously published

models (DS2006, SP2010, and OLI Studio 9.0.6) in predicting CO2 solubility in formation brines at the P-

T-x region pertinent to CCS and CO2-EGS. The components, key parameters, and limitations of the

PSUCO2 model are also summarized in Table 4. 8.

As an application of the proposed PSUCO2 model, the model calculated CO 2 solubility contours

on the P-T diagram at 100-1500 bar and 298.15-523.15 K are shown in Figure 4. 9. A clear evidence of the

strong salting-out effect on CO2 solubility in the Mt. Simon formation brine (Figure 4. 9b) can be observed

when compared to the CO2 solubility contours in pure water (Figure 4. 9a). The point A on each contour

map indicates a fixed P-T condition of 900 bar and 373.15 K. The contour for 1.8 mol kg -1 CO2 is located

on the left side of the point A for the CO 2-H2O system (Figure 4. 9a), when the aqueous phase is changed

from pure H2O to the Mt. Simon formation brine, the same contour (1.8 mol kg -1 CO2) is moved to the right

side of the point A (Figure 4. 9b).

In addition to the salting-out effect, each contour map can be divided into two regions (Region I

and Region II) based on the temperature-dependent behavior of CO2 solubility in the aqueous phase. The

boundary between Region I and Region II is defined as the maximum gradient path between the adjacent

CO2 solubility contours. In Figure 4. 9, the path of maximum gradient is roughly shown as the straight lines

between the neighbored contours. In Region I, the CO 2 solubility in the aqueous phase decreases

monotonically in response to increased temperature, but at a given pressure in region II, the behavior of the

CO2 solubility in response to temperature increase is the opposite of that in Region I. For instance,

following the isobaric paths a-b-c (Region I) and d-e-f (Region II) on the P-T diagram (Figure 4. 9a), as

temperature increases, CO2 solubility decreases along the path a-b-c, but increases along the path d-e-f. In

addition, at the points b and e, CO2 solubility at two different temperatures reached the same value

(𝑚𝐶𝑂2 |𝑏 =𝑚𝐶𝑂2 |𝑒 =1.8 mol kg-1). Along the isobaric path a-b-c-d-e-f, CO2 solubility decreases to a minimum

then increases in response to increased temperature.

We have discussed this temperature-dependent behavior of CO2 solubility by defining a “transition

zone” on the P-x diagram (Chapter 2, Figure 2. 14(a,b)), but the contours on the P-T diagram provide a

different perspective on this phenomenon. On the P-T diagram, the P-T region (100-400 bar, 298-373 K)

114
(Spycher and Pruess, 2010) for typical CCS project is located mainly in Region I, whereas the P-T window

(200-600 bar, 373-573 K) (Spycher and Pruess, 2010) for the CO2-EGS concept is located mainly in

Region II, as shown in Figure 4. 9b. Understanding the different temperature-dependent behavior of CO2

solubility in these two distinct regions may shed light on future work for evaluating and modeling the CCS

and CO2-EGS processes.

523.15 9.0
0.8 7.0
5.5
2.0
2.2
3.0 4.5
498.15 1.2
5.5
3.8
2.5
0.8 1.41.6 4.5
473.15 1.8
3.8
3.0
448.15

. 2.0 2.2 2.5 3.0

.
423.15 1.2 Region II
T/K

1.4 1.6 1.8 2.5

398.15 0.8

373.15 .. . f
2.0
X
A
2.2

. ...
e
348.15

323.15
1.2

1.4 1.6
d
c
Region I b .. 1.8
. 2.2

.
2.0 2.2
298.15 a
100 300 500 700 900 1100 1300 1500
P / bar
(a)

115
 523.15
4.0 5.0
1.8 4.5
1.6 3.5
2.2
3.0
498.15 4.0
1.0 2.5 3.5
1.2 3.0
1.4
473.15
2.2 2.5
CO2-EGS 1.8
448.15

423.15
. 1.6
Region II 2.2
T/K

1.0
1.2 1.4 1.8

.
398.15

1.6 A
373.15 X

348.15

323.15
CCS
1.0
1.2
. . . 1.4
1.8

.
Region1.6I 1.8
298.15
100 300 500 700 900 1100 1300 1500
P / bar
(b)

Figure 4. 9(a, b) The CO2 solubility contours in pure water and synthetic Mt. Simon formation brine
generated by the PSUCO2 model. Numbers on the contour indicate the CO 2 solubility value by molality
(mol kg-1): (a) pure water, IS = 0 mol kg-1; (b) synthetic Mt. Simon brine, IS = 1.712 mol kg-1.

116
 CHAPTER 5

CONCLUISONS

117
 A reliable PVT apparatus to measure the CO2 solubility in aqueous solutions at elevated temperatures

and pressures has been constructed and shown to provide accurate and reproducible CO2 solubility

measurements.

 At 150 bar, new CO2 solubility data in the aqueous phase were collected at 144 P-T-x conditions

(Table 5. 1 and Table 5. 2). The instrumental error is around 0.7 % and the random error is from 0.3 to

4.5 % of the measured CO2 solubility values.

 Experimental data demonstrates that CO2 solubility in aqueous solutions with different salt species are

significantly different, whether the solutions have the same percent of salt by weight or have the same

ionic strength. The experimental results also show that in the aqueous phase a cation with a higher

charge density will have more significant salting-out effect on dissolved CO2 than other ions.

 A 𝛾 − 𝜑 (activity coefficient - fugacity coefficient) type thermodynamic model (PSUCO2) is

developed and improved based on the measured CO2 solubility data. The model development starts

from the CO2-H2O and CO2-NaCl-H2O systems, then its capacity is extended to account for salt

species other than NaCl. Finally, the PSUCO2 model is capable of calculating CO2 solubility in natural

formation brine (or mixed-salt aqueous solution).

 Based on both experimental and modeling results, a CO2 solubility “transition zone” is found in the P-x

phase diagram for the CO2-H2O system. The CO2 solubility is not a monotonic function of temperature

in the transition zone. However, outside of that transition zone, the CO 2 solubility decreases or

increases in response to increased temperature.

 The path of the maximum gradient of the CO2 solubility contours is defined to divide the P-T diagram

into two regions: in Region I, the CO2 solubility in the aqueous phase decreases monotonically in

response to increased temperature; in region II, the behavior of the CO 2 solubility is the opposite of

that in Region I as the temperature increases. The P-T region for typical CCS project is located mainly

in Region I, whereas for the CO2-EGS concept it is located mainly in Region II.

 The modeling results show a positive correlation between the H 2O solubility in the CO2-rich phase and

the activity of H2O in the aqueous phase, the influence of salts species on solubility of H 2O in the CO2-

rich phase is clearly observed.

118
 The comparison of the PSUCO2 model to the widely-accepted models and commercial software

reveals an improvement of the proposed model. A web-based CO2 solubility computational tool is also

developed for reader’s convenience and it can be accessed via the link www.carbonlab.org/psuCO2/

Table 5. 1 P-T-x matrix of CO2 solubility study for single-salt system (96 P-T-x points were measured,
random error up to 5%).

System* P / bar T/K Salt Molality / mol kg-1

H 2O 0
NaCl 1 2 3 4 5 6
323.15
CaCl2 1/3 2/3 1 4/3 5/3 2
150 373.15
Na2SO4 1/3 2/3 1 4/3 5/3 2
423.15
MgCl2 1/3 2/3 1 4/3 5/3 2
KCl 0.5 1 2 3 4 4.5
*The CO2 and H2O presented in the experimental system are not shown in the table.

Table 5. 2 P-T-x matrix of CO2 solubility study for mixed-salt system (48 P-T-x points were measured,
random error up to 3%).

Experimental System* P / bar T/K Ionic strength / mol kg-1

100
NaCl+CaCl2+Na2SO4+MgCl2+KCl+SrCl2+NaBr 323.15
125 1.712
373.15
150 4.984
NaCl+CaCl2 423.15
175
*The CO2 and H2O presented in the experimental system are not shown in the table.

119
REFERENCES

120
Akinfiev N. N. and Diamond L. W.(2010) Thermodynamic model of aqueous CO 2-H2O-NaCl solutions

from -22 to 100 °C and from 0.1 to 100 MPa, Fluid Phase Equilibria, 295, 104-124.

Bachu S.(2003) Screening and ranking of sedimentary basins for sequestration of CO 2 in geological media

in response to climate change, Environmental Geology, 44, 227-289.

Bamberger A., Sieder G., and Maurer G.(2000) High-pressure (vapor+liquid) equilibrium in binary

mixtures of (carbon dioxide-water or acetic acid) at temperatures from 313 to 353 K, Journal of

Supercritical Fluids, 17, 97-110.

Bando S., Takemura F., Nishio M., Hihara E., and Akai M.(2003) Solubility of CO 2 in aqueous solutions of

NaCl at (30 to 60) °C and (10 to 20) MPa, Journal of Chemical & Engineering Data, 48, 576-579

Benson S.M. and Cole D.R.(2008) CO2 sequestration in deep sedimentary formations, Elements, 4, 325-

331.

Bermejo M.D., Martín A., Florusse L.J., Peters C.J., and Cocero M.J.(2005) The influence of Na 2SO4 on

the CO2 solubility in water at high pressure, Fluid Phase Equilibria, 238, 220-228.

Bickle M., Chadwick A., Huppert H.E., Hallworth M., and Lyle S.(2007) Modelling carbon dioxide

accumulation at Sleipner: Implications for underground carbon storage, Earth and Planetary

Science Letters, 255, 164-176.

Briones J.A., Mullins J.C., Thies M.C. and Kim B.-U.(1987) Ternary phase equilibria for acetic acid-water

mixtures with supercritical carbon dioxide, Fluid Phase Equilibria, 36, 235-246.

Brown, D.W.(2000) A hot dry rock geothermal energy concept utilizing supercritical CO 2 instead of water,

Proceedings, Twenty-fifth Workshop on Geothermal Reservoir Engineering, Stanford University,

Stanford, California.

Buch K., Harvey H. W., Wattenberg H. and Gripenberg S.(1932) Uber dem Kohlensauresystem im

Meerwasser, Rapports et procès-verbaux des reunions, Conseil Permanent International Pour

l'Exploration de la Mer, 79, 1-70.

Burmakina, G.V., Efanov, L.N., and Shnet, M.A.(1982), CO2 solubility in aqueous solutions of some

electrolytes and sucrose, Russian Journal of Physical Chemistry, 56, 705–707

121
Bush M.F., O’Brien J.T., Prell J.S., Wu C.-C., Saykally R.J., and Williams E.R.(2009) Hydration of

alkaline earth metal dications: effects of metal ion size determined using infrared action

spectroscopy, Journal of the American Chemical Society, 131, 13270-13277.

Cappa, J.A., and D.D. Rice (1995), Carbon dioxide in Mississippian rocks of the Paradox Basin and

adjacent areas, Colorado, Utah, New Mexico, and Arizona, in U.S. Geological Survey Bulletinn,

2000-H.

Chapoy A., Mohammadi A.H., Chareton A., Tohidi B., and Richon D.(2004) Measurement and modeling

of gas solubility and literature review of the properties for the carbon dioxide-water system,

Industrial & Engineering Chemistry Research, 43, 1794-1802.

Collins K.D.(1997) Charge density-dependent strength of hydration and biological structure, Biophysical

Journal, 72, 65-76.

Corti H.R., Krenzer M.E., De Pablo J.J., and Prausnitz J.M.(1990), Effect of a dissolved gas on the

solubility of an electrolyte in aqueous solution, Industrial & Engineering Chemistry Research, 29,

1043-1050.

Cramer S.D. (1982) The solubility of methane, carbon dioxide, and oxygen in brines from 0 to 300 oC,

Report of Investigations 8706, United States Bureau of Mines.

Dhima A., de-Hemptinne J.-C., and Jose J.(1999) Solubility of Hydrocarbons and CO 2 mixtures in water

under high pressure, Industrial & Engineering Chemistry Research, 38, 3144-3161.

Dilmore R.M., Allen D.E., Richard Mccarthy Jones J., Hedges S.W., and Soong Y.(2008), Sequestration of

dissolved CO2 in the Oriskany Formation, Environmental Science & Technology, 42, 2760-2766.

Dohrn R., Bünz A.P., Devlieghere F., and Thelen D.(1993) Experimental measurements of phase equilibria

for ternary and quaternary systems of glucose, water, CO2 and ethanol with a novel apparatus,

Fluid Phase Equilibria, 83, 149-158.

Dresel P.E. and Rose A.W.(2010), Chemistry and origin of oil and gas well brines in western Pennsylvania,

Open-File Oil and Gas Report 10-01.0, Pennsylvania Geological Survey, Fourth Series,

Harrisburg.

Drummond S.E.(1981) Boiling and mixing of hydrothermal fluids, Ph.D. Dissertation, The Pennsylvania

State University.

122
Duan Z. and 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, 257-271.

Duan Z., Sun R., Zhu C., and Chou I-M.(2006) An improved model for the calculation of CO 2 solubility in

aqueous solutions containing Na+, K+, Ca2+, Mg2+, Cl-, and 𝑺𝑶𝟐−
𝟒 , Marine Chemistry, 98, 131-139.

Duan Z., Hu J., Li D. and Mao S.(2008) Densities of the CO 2-H2O and CO2-H2O-NaCl system up to 647 K

and 100 MPa, Energy & Fuels, 22, 1666-1674.

Dubacq B., Bickle M.J. and Evans K.A.(2013) An activity model for phase equilibria in the H 2O-CO2-NaCl

system, Geochimica et Cosmochimica Acta, 110, 229-252.

D'souza R. Patrick J.R. and Teja A.S.(1988) High pressure phase equilibria in the carbon dioxide - n-

Hexadecane and carbon dioxide -water systems, Canadian Journal of Chemical Engineering, 66,

319-323.

Ellis A.J. and Golding R. M.(1963) The solubility of carbon dioxide above 100 °C in water and in sodium

chloride solutions, American Journal of Science, 261, 47-60.

Fernández D.P., Goodwin A.R.H., Lemmon E.W., Levelt Sengers J.M.H., and Williams R.C. (1997) A

formulation for the static permittivity of water and steam at temperatures from 238 K to 873 K at

pressures up to 1200 MPa including derivatives adn Debye-Hückel coefficients, Journal of

Physical and Chemical Reference Data, 26, 1125-1169.

Fernández-Prini R., Alvarez J. L., and Harvey A. H. (2003) Henry's constants and vapor-liquid distribution

constants for gaseous solutes in H2O and D2O at high temperatures, Journal of Physical and

Chemical Reference Data, 32, 903-916.

Ferrentino G., Barletta D., Donsi F., Ferrari G., and Poletto M.(2010) Experimental measurements and

thermodynamic modeling of CO2 solubility at high pressure in model apple juices, Industrial &

Engineering Chemistry Research, 49, 2922-3000.

Findley A. and Shen B.(1912), CLVI.—The influence of colloids and fine suspensions on the solubility of

gases in water. Part II. Solubility of carbon dioxide and of hydrogen, Journal of the Chemical

Society, Transactions, 101, 1459-1468.

123
Gallagher J.S., Crovetto R., and Levelt Sengers J.M.H.(1993) The thermodynamic behavior of the CO 2-

H2O system from 400 to 1000 k, up to 100 MPa and 30 % mole fraction of CO 2, Journal of

Physical and Chemical Reference Data, 22, 431-513.

Gates J.A. and Wood R.H.(1985) Densities of aqueous solutions of NaCl, MgCl 2, KCl, NaBr, LiCl, and

CaCl2 from 0.05 to 5.0 mol kg-1 and 0.1013 to 40 MPa at 298.15 K, Journal of Chemical &

Engineering Data, 30, 44-49.

Geffcken G.(1904) Beiträge zur Kenntnis der Löslichkeitsbeeinflussung, Ph.D. Dissertation, Leipzig

University.

Gerecke, J.(1969) Ein Beitrag zur Gaslöslichkeit in Elektrolytlösungen Untersucht am Beispiel der

Löslichkeit von H2, CO2 und NH3 in Wasser und Wässrigen Salzlösungen, PhD Dissertation,

Technische Hochschule für Chemie Carl Schorlemmer - Leuna Merseburg.

Ghafri S.A., Maitland G.C., and Trusler P.M.(2012) , Densities of aqueous MgCl 2(aq), CaCl2(aq), KI(aq),

NaCl(aq), KCl(aq), AlCl3(aq), and (0.964 NaCl+0.136 KCl)(aq) at temperatures between (283 and

472) K, pressures up to 68.5 MPa, and molalities up to 6 mol kg -1, Journal of Chemical &

Engineering Data, 57, 1288-1304.

Gize, A., and Macdonald R.(1993), Generation of compositionally atypical hydrocarbons in CO 2-rich

geoloic environments, Geology, 21, 129-132.

Görgényi M. Dewulf J., Van Langenhove H., and Héberger K.(2006), Aqueous salting-out effect of

inorganic cations and anions on non-electrolytes, Chemoosphere, 65, 802-810.

Gordon J.E. and Thorne R.L.(1967) Salt effects on non-electrolyte activity coefficients in mixed aqueous

electrolyte solutions-II. Artificial and natural sea waters, Geochimica et Cosmochimica Acta, 31,

2433-2443.

Gu F.(1998) Solubility of carbon dioxide in aqueous sodium chloride solution under high pressure, Journal

of Chemical Engineering of Chinese Universities, 12, 118-123.

Han X., Yu Z., Qu J., Qi T., Guo W., and Zhang G.(2011) Measurement and correlation of solubility data

for CO2 in NaHCO3 aqueous solution, Journal of Chemical & Engineering Data, 56, 1213-1219.

124
Harned H.S. and Davis R.(1943) The ionization constant of carbonic acid in water and the solubility of

carbon dioxide in water and aqueous salt solutions from 0 to 50, Journal of the American

Chemical Society, 65, 2030-2037.

Harvey A. H. (1996) Semiempirical correlation for Henry's constants over large temperature ranges, AIChE

Journal, 42, 1491-1494.

Haugan P.M. and Drange H.(1992) Sequestration of CO 2 in the deep ocean by shallow injection, Nature,

357, 318-320.

Holmes H.F. and Mesmer (1983) Thermodynamic Properties of aqueous solutions of the alkali metal

chlorides to 250 oC, Journal of Physical Chemistry, 87, 1242-1255.

Holmes H.F., Busey R.H., Simonson J.M., and Mesmer R.E.(1994), CaCl 2(aq) at elevated temperatures.

Enthalpies of dilution, isopiestic molalities, and thermodynamic properties, Journal of Chemical

Thermodynamics, 26, 271-298.

Hou S-X., Maitland G. C., and Martin Trusler J. P. (2013) Measurement and modeling of the phase

behavior of the (carbon dioxide + water) mixture at temperatures from 298.15 K to 448.15 K,

Journal of Supercritical Fluids, 73, 87-96.

Hu J., Duan Z., Zhu C., and Chou I-M. (2007) PVTx properties of the CO2-H2O and CO2-H2O-NaCl

systems below 647 K: Assessment of experimental data and thermodynamic models, Chemical

Geology, 238, 249-267.

Kaszuba, J. P., L. L. Williams, D. R. Janecky, W. K. Hollis, and I. N. Tsimpanogiannis (2006), Immiscible

CO2-H2O fluids in the shallow crust, Geochemistry Geophysics Geosystems, 7, Q10003.

Kiepe J, Horstmann S., Fischer K., and Gmehling J.(2002) Experimental determination and prediction of

gas solubility data for CO2 + H2O mixtures containing NaCl or KCl at temperatures between 313

and 393 K and pressures up to 10 MPa, Industrial & Engineering Chemistry Research, 41, 4393-

4398.

Kim Y.S., Lim B.D., Lee J.E., and Lee C.S.(2008) Solubilities of Carbon Dioxide, Methane, and Ethane in

Sodium Chloride Solution Containing Gas Hydrate, Journal of Chemical & Engineering Data, 53,

1351–1354.

125
King M.B., Mubarak A., Kim J.D., and Bott T.R.(1992) The Mutual Solubilities of Water with

Supercritical and Liquid Carbon Dioxide, Journal of Supercritical Fluids, 5, 296-302.

Kobe K.A. and Williams J.S.(1935) Confining liquids for gas analysis: Solubility of carbon dioxide in salt

solutions, Industrial and Engineering Chemistry, 7, 37-38.

Koschel D., Coxam J.-Y., Rodier L., and Majer V.(2006) Enthalpy and solubility data of CO 2 in water and

NaCl(aq) at conditions of interest for geological sequestration, Fluid Phase Equilibria, 247, 107–

120.

Kumar A., Noh M.H., Ozah R.C., Pope G.A., Bryant S.L., Sepehrnoori K., and Lake L.W.(2005) Reservoir

simulation of CO2 storage in aquifers, SPE Journal, 03, 336-348.

Lamb W.M., Valley J.W., and Brown P.E.(1987), Post-metamorphic CO2-rich fluid inclusions in

granulites, Contributions to Mineralogy and Petrology, 96, 485-495.

Li Z., Dong M., Li S., and Dai L.(2004) Densities and solubilities for binary systems of carbon dioxide +

water and carbon dioxide + brine at 59 oC and pressures to 29 MPa, Journal of Chemical &

Engineering Data, 49, 1026-1031.

Li Y-H and Tsui T-F.(1971), The solubility of CO2 in water and sea water, Journal of Geophysical

research, 76, 4203-4207.

Liu Y. Hou M., Yang G., and Han B.(2011) Solubility of CO 2 in aqueous solutions of NaCl, KCl, CaCl2

and their mixed salts at different temperatures and pressures, Journal of Supercritical Fluids, 56,

125–129.

Lowenstern J.B.(2001) Carbon dioxide in magmas and implications for hydrothermal systems, Mineralium

Deposita, 36, 490-502.

Mähler J. and Persson I.(2012) A study of the hydration of the alkali metal ions in aqueous solution,

Inorganic Chemistry, 51, 425-438.

Malinin S. D. (1959) The system water-carbon dioxide at high temperatures and pressures, Geochemistry,

3, 292-306.

Malinin S. D. and Savelyeva N. I.(1972) The solubility of CO2 in NaCl and CaCl2 solutions at 25, 50 and

75°C under elevated CO2 pressures, Geochemistry International, 9. 410-418.

126
Malinin S. D. and Kurovskaya N. A.(1975) Investigations of CO 2 solubility in a solution of chlorides at

elevated temperatures and pressures of CO2, Geochemistry International, 12, 199-201.

Mao S., Zhang D., Li Y., and Liu N.(2013) An improved model for calculating CO 2 solubility in aqueous

NaCl solutions and the application to CO2-H2O-NaCl fluid inclusions, Chemical Geology, 347,

43-58.

Marcus Y.(1988) Ionic radii in aqueous solutions, Chemical Reviews, 88, 1475-1498.

Markham A. E. and Kobe K. A.(1941) The solubility of carbon dioxide and nitrous oxide in aqueous salt

solutions, Journal of the American Chemical Society, 63, 449-454.

Mather A. E., and Franck E. U. (1992) Phase equilibria in the system carbon dioxide-water at elevated

pressures, Journal of Physical Chemistry, 96, 6-8.

McIntosh J.C., Water L.M., and Martini A.M.(2004) Extensive microbial modification of formation water

geochemistry: Case study from a Midcontinent sedimentary basin, United States. Geological

Society of America Bulletin, 116, 743-759.

Moin S.T., Pribil A.B., Lim L.H.V., Hofer T.S., Randolf B.R., and Rode B.M.(2011), Carbon dioxide in

aqueous environment — A quantum mechanical charge field molecular dynamics study,

International Journal of Quantum Chemistry, 111, 1370-1378

Müller G.; Bender E.; and Maurer G.(1988) Das Dampf-Flüssigkeitsgleichgewicht des ternären Systems

Ammoniak-Kohlendioxid-Wasser bei hohen Wassergehalten im Bereich zwischen 373 und 473

Kelvin, Berichte der Bunsengesellschaft für physikalische Chemie, 92, 148-160.

Nakayama T. and Sagara H.(1987) High pressure liquid-liquid equilibria for the system of water, ethanol

and 1,1-difluoroethane at 323.2 K, Fluid Phase Equilibria, 38, 109-127.

Nighswander J.A. Kalogerakis N. and Mehrotra A.K.(1989) Solubilities of carbon dioxide in water and 1

wt % NaCl solution at pressures up to 10 MPa and temperatures from 80 to 200°C, Journal of

Chemical & Engineering Data, 34, 355-360.

Onda K., Sada E., Kobayashi T., Kito S. and Ito K.(1970) Salting-out parameters of gas solubility in

aqueous salt solutions, Journal of Chemical Engineering of Japan, 3, 18-24.

127
Panagiotopoulos A.Z. and Reid R.C.(1986) New mixing rule for cubic equations of state for highly polar,

asymmetric systems, ACS Symposium Series 300: American Chemical Society, Washington D.C.,

571-682.

Pérez-Salado Kamps A. Meyer E., Rumpf B., and Maurer G.(2007) Solubility of CO 2 in Aqueous Solutions

of KCl and in Aqueous Solutions of K2CO3, Journal of Chemical & Engineering Data, 52, 817-

832.

Phillips, G.N., and K.A. Evans (2004), Role of CO 2 in the formation of gold deposits, Nature, 429, 860-

863.

Pitzer, K. S.(1991) Activity Coefficients in Electrolyte Solutions; CRC Press 2nd Ed. Boca Raton Ann

Arbor Boston London.

Pitzer K. S., Peiper J. C. and Busey R.H.(1984) Thermodynamic properties of aqueous sodium chloride

solutions, Journal of Physical and Chemical Reference Data. 13, 1-102.

Pruess K., Xu T., Apps J. and Garcia J.(2001), Numerical Modeling of Aquifer Disposal of CO 2,

SPE/EPA/DOE Exploration and Production Environmental Conference, 26-28 February, San

Antonio, Texas.

Pruess K. and Müller N.(2009) Formation dry-out from CO2 injection into saline aquifers:1. Effects of

solids precipitation and their mitigation, Water Resources Research, 45, W03402.

Prutton C. F. and Savage R. L.(1945) The solubility of carbon dioxide in calcium chloride-water solutions

at 75, 100, 120 and high pressures, Journal of the American Chemical Society. 67, 1550-1554.

Qin J., Rosenbauer R.J., and Duan Z.(2008) Experimental measurements of vapor-liquid equilibria of the

H2O+CO2+CH4 ternary system, Journal of Chemical & Engineering Data, 53, 1246-1249.

Redlich O. and Kwong J. N. S.(1949) On the thermodynamics of solution. V. An equation of state.

Fugacities of gaseous solutions, Chemical Reviews, 44, 233-244.

Rochelle C.A. and Moore Y.A.(2002) The solubility of supercritical CO 2 into pure water and synthetic

Utsiraporewater, British Geological Survey, Commissioned Report CR/02/052.

Rogers P.S.Z. and Pitzer K.S.(1981) High-temperatue thermodynamic properties of aqueous sodium sulfate

solutions, Journal of Physical Chemistry, 85, 2886-2895.

128
Rumpf B. and Maurer G.(1993) An Experimental and Theoretical Investigation on the Solubility of Carbon

Dioxide in Aqueous Solutions of Strong Electrolytes, Berichte der Bunsengesellschaft für

physikalische Chemie. 97, 85-97.

Rumpf B. Nicolaisen H., Öcal C., and Maurer G.(1994) Solubility of carbon dioxide in aqueous solutions

of sodium chloride: experimental results and correlation, Journal of Solution Chemistry, 23, 431-

448.

Sako T., Sugeta T., Nakazawa N., Okubo T., and Sato M.(1991) Phase equilibrium study of extraction and

concentration of furfural produced in reactor using supercritical carbon dioxide, Journal of

Chemical Engineering of Japan, 24, 449-455.

Samoilov O. Y.(1957) A new approach to the study of hydration ion in aqueous solutions, Discussions of

the Faraday Society, 24, 141-146.

Scharlin P. (1996), Carbon Dioxide in Water and Aqueous Electrolyte Solutions, Solubility Data Series vol

62. International Union of Pure and Applied Chemistry, Oxford University Press, ISSN 0919-

5622.

Schay G. (2012) A concise introduction to linear algebra, Brikhäuser Boston, ISBN 978-0-8176-8325-2.

Schmidt C. and Bodnar R. J. (2000) Synthetic fluid inclusions: XVI. PVTX properties in the system H 2O-

NaCl-CO2 at elevated temperatures, pressures, and salinities, Ceochimica et Cosmochimica Acta,

64, 3853-3869.

Serpa F.S., Vidal R.S., Amaral-Filho J.H.B., do-Nascimento J.F., Ciambelli J.R.P., Figueiredo C.M.S.,

Salazar-Banda G.R., Santos A.F., Fortuny M., Franceschi E., and Dariva C.(2013) Solubility of

carbon dioxide in ethane-1,2-diol-water mixtures, Journal of Chemical & Engineering Data, 58,

3464-3469.

Setschenow, J. (1889) Über die Konstitution der Salzlösungen auf Grund ihres Verhaltens zu Kohlensäure.

Zeitschrift für Physikalische Chemie, 4, 117-128.

Setschenow J.(1892) Über die konstitution der salzlosungen auf grund ihres verhaltens zu kohlensäure,

Annales de Chimie et de Physique. 25, 226-270.

Shepard D.(1968) A two-dimensional interpolation function for irregularly-spaced data, Proceedings of the

1968 ACM National Conference,517-524.

129
Shyu G-S., Hanif N.S.M., Hall K.R., and Eubank P.T.(1997) Carbon dioxide-water phase equilibria results

from the Wong-Sandler combining rules, Fluid Phase Equilibria, 130, 73-85.

Sisson V.B., Crawford M.L. and Thompson P.H.(1981) CO 2-brine inmmiscibility at high temperatures,

evidence from Calcareous Metasedimentary Rocks, Contributions to Mineralogy and Petrology,

78, 371-378.

Span R. and Wagner W.(1996) A new equation of state for carbon dioxide covering the fluid region from

the triple –point temperature to 1100 K at pressures up to 800 MPa, Journal of Physical and

Chemical Reference Data, 25, 1509-1596.

Springer R.D., Wang Z., Anderko A., Wang P., and Felmy A.R.(2012) A thermodynamic model for

predicting mineral reactivity in supercritical carbon dioxide: I. Phase behavior of carbon dioxide-

water-chloride salt systems across the H2O-rich to the CO2-rich regions, Chemical Geology, 322-

323, 151-171.

Spycher N. and Pruess K., and Ennis-King J.(2003) CO2-H2O mixtures in the geological sequestration of

CO2. I. Assessment and calculation of mutual solubilities from 12 to 100°C and up to 600 bar,

Geochimica et Cosmochimica Acta, 67, 3015–3031.

Spycher N. and Pruess K.(2005) CO2-H2O mixtures in the geological sequestration of CO 2. II. Partitioning

in chloride brines at 12-100 oC and up to 600 bar, Geochimica et Cosmochimica Acta, 69, 3309-

3320.

Spycher N. and Pruess K.(2010) A phase-partitioning model for CO2-brine mixtures at elevated

temperatures and pressures: Application to CO2-enhanced geothermal systems, Transport in

Porous Media, 82, 173-196.

Steele-MacInnis M., Capobianco R.M., Dilmore R., Goodman A., Guthrie G., Rimstidt D. and Bodnar

R.J.(2013) Volumetrics of CO2 Storage in Deep Saline Formations, Environmental Science &

Technology, 47, 79-86.

Stewart P. B. and Munjal P.(1970) Solubility of carbon dioxide in pure water, synthetic sea water, and

synthetic sea water concentrates at -50 to 250 °C and 10 to 45 atm. pressures, Journal of Chemical

& Engineering Data, 15, 67-71.

130
Takenouchi, S. and Kennedy G.C.(1964) The binary system H2O-CO2 at high temperature and pressures,

American Journal of Science, 262, 1055-1074.

Takenouchi S. and Kennedy G.C.(1965) The solubility of carbon dioxide in NaCl solutions at high

temperatures and pressures, American Journal of Science, 263, 445-454.

Taylor J.R.(1997) An introduction to error analysis: The study of uncertainties in physical measurements

(2nd ed.), University Science Books, Sausalito.

Teng H., Yamasaki A., Chun M.-K. and Lee H.(1997) Solubility of liquid CO 2 in water at temperatures

from 278 K to 293 K and pressures from 6.44 MPa to 29.49 MPa and densities of the

corresponding aqueous solutions, Journal of Chemical Thermodynamics, 29, 1301-1310.

Teng H. and Yamasaki A.(2002) Pressure-mole fraction phase diagrams for CO2-pure water system water

temperatures and pressures corresponding to ocean waters at depth to 3000 M, Chemical

Engineering Communications, 189, 1485-1497.

Tester, J.W., Anderson B.J., Batchelor A.S., Blackwell D.D., DiPippo R., Drake E.M., Garnish J., Livesay

B., Moore M.C., Nichols K., Petty S. (2006) The Future of Geothermal Energy: Impact of

Enhanced Geothermal Systems (EGS) on the United States in the 21st Century. Massachusetts

Institute of Technology (Open access report:

http://www1.eere.energy.gov/geothermal/pdfs/future_geo_energy.pdf).

Tong D., Martin Trusler J. P., and Vega-Maza D. (2013) Solubility of CO2 in aqueous solutions of CaCl2 or

MgCl2 and in a synthetic formation brine at temperatures up to 423 K and pressures up to 40 MPa,

Journal of Chemical & Engineering Data, 58, 2116–2124.

Tödheide K. and Franck E. U.(1963) Das Zweiphasengebiet und die kritische Kurve im system

kohlendioxid-wasser bis zu drucken von 3500 bar, Zeitschrift für Physikalische Chemie, Neue

Folge, 37, 387-401.

Valtz A., Chapoy A., Coquelet C., Raricaud P., and Richon D.(2004) Vapour-liquid equilibira in the carbon

dioxide-water system, measurement and modelling from 278.2 to 318.2 K, Fluid Phase Equilibria,

226, 333-334.

131
Vázquez, G., Chenlo, F., Pereira, G., Peaguda, J.(1994) Carbon dioxide solubility in aqueous solutions of

sodium chloride, copper(ii) sulfate, potassium iodide and sodium bromide, Anales de Quimica, 90,

324–328.

Veeranna D. and Rihani D.N.(1984) Avoiding convergence problems in VLE predictions using an equation

of state, Fluid Phase Equilibria, 16, 41-66.

Wagner W. and Pru 𝜷 A.(2002) The IAPWS formulation 1995 for the thermodynamic properties of

ordinary water substance for general and scientific use, Journal of Physical and Chemical

Reference Data, 31, 387-535.

Wang P. and Pitzer K.S.(1998) Thermodynamic properties of aqueous magnesium chloride solutions from

250 to 600 K and to 100 MPa, Journal of Physical and Chemical Reference Data, 27, 971-991.

Wendland M., Hasse H., and Maurer G. (1999) Experimental pressure-temperature data on three- and four-

phase equilibria of fluid, hydrate, and ice phases in the system carbon dioxide-water, Journal of

Chemical & Engineering Data, 44, 901-906.

Wiebe R. and Gaddy V. L.(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, 315-318.

Wiebe R. and Gaddy V.L.(1941) Vapor phase composition of carbon dioxide-water mixtures at various

temperatures and at pressures to 700 atmospheres, Journal of the American Chemical Society, 63,

475-477.

Yan W., Huang S., Stenby E.H.(2011) Measurement and modeling of CO 2 solubility in NaCl brine and

CO2-saturated NaCl brine density, International Journal of Greenhouse Gas Control, 5, 1460-

1477.

Yang D. Tontiwachwuthikul P. and Gu Y.(2005) Interfacial tensions of the Crude oil + Reservoir Brine +

CO2 systems at pressures up to 31 MPa and temperatures of 27 oC and 58 oC, Journal of Chemical

& Engineering Data, 50, 1242-1249.

Yasunishi A. and Yoshida F.(1979a) Solubility of carbon dioxide in aqueous electrolyte solutions, Journal

of Chemical & Engineering Data, 24, 11-14.

132
Yasunishi A., Tsuji M. and Sada E. (1979b) Solubility of carbon dioxide in aqeuous mixed-salt solutions,

In Thermodynamic Behavior of Electrolytes in Mixed Solvents-II; Furter ,W.F., Advances in

Chemistry, American Chemical Society: Washington, DC, 1979.

Yeh S-Y. and Peterson R.(1964) Solubility of Carbon dioxide, Krypton, and Xenon in aqueous solution,

Journal of Pharmaceutical Sciences, 53, 822-824.

Zhao H., Fedkin M.V., Dilmore R.M., and Lvov S.N.(2014a) Carbon dioxide solubility in aqueous

solutions of sodium chloride at geological conditions: Experimental results at 323.15, 373.15,

423.15 K and 150 bar and modeling up to 573.15 K and 2000 bar, Geochimica et Cosmochimica

Acta, doi:10.1016/j.gca.2014.11.004.

Zhao H., Dilmore R.M., and Lvov S.N.(2014b) Experimental studies and modeling of CO 2 solubility in

high temperature aqueous CaCl2, MgCl2, Na2SO4, and KCl solutions, Geochimica et

Cosmochimica Acta, in review.

133
APPENDIXES

134
Appendix A: Additional Model Equations

The fugacity coefficients for both CO2 and H2O in the CO2-rich phase were calculated using the

Redlich-Kwong equation of state (RK EoS) (Redlich and Kwong, 1949):

𝑅𝑇 𝑎
𝑃=( )−( ) (A.1)
𝑣−𝑏 𝑇 0.5 𝑣(𝑣+𝑏)

where 𝑣 is the molar volume (cm3/mol) and parameters a and b account for the attraction and repulsion

interactions between molecules. By tuning the EoS against the pure CO2 properties, Spycher and Pruess

(2010) set the pure component parameter a as a function of temperature:

a = ko + k1T (A.2)

The parameters 𝑘𝑜 , 𝑘1 , and b for pure components (CO2 and H2O) can be found in Spycher and

Pruess (2010). The standard mixing rules were applied for the CO2-H2O gas mixture as

𝑎𝑚𝑖𝑥 = ∑𝑛𝑖 ∑𝑛𝑗 𝑦𝑖 𝑦𝑗 𝑎𝑖𝑗 (A.3)

𝑏𝑚𝑖𝑥 = ∑𝑛𝑖 𝑦𝑖 𝑏𝑖 (A.4)

The cross parameters 𝑎𝑖𝑗 (see Eq. (2.14)) in Eq. (A.3) are a measure of the strength of attraction

between the molecules of the components i and j. The fugacity coefficient of the k-th component in the

CO2-H2O gas mixture can be obtained using the equations given in Panagiotopoulos and Reid (1986) and

Spycher and Pruess (2010):

𝑏𝑘 𝑃𝑣 (𝑣−𝑏𝑚𝑖𝑥 )
𝑙𝑛𝜑𝑘 = ( − 1) − 𝑙𝑛 (𝑃 )+
𝑏𝑚𝑖𝑥 𝑅𝑇 𝑅𝑇

∑𝑛 𝑛 𝑛 2 𝑛
𝑖=1 𝑦𝑖 (𝑎𝑖𝑘 +𝑎𝑘𝑖 )−∑𝑖=1 ∑𝑗=1 𝑦𝑖 𝑦𝑗 (𝑘𝑖𝑗 −𝑘𝑗𝑖 )√𝑎𝑖 𝑎𝑗 +𝑥𝑘 ∑𝑖=1 𝑥𝑖 (𝑘𝑘𝑖 −𝑘𝑖𝑘 )√𝑎𝑖 𝑎𝑘 𝑏𝑘 𝑎𝑚𝑖𝑥 𝑣
( − ) ×⁡( ) 𝑙𝑛 ( ) (A.5)
𝑎𝑚𝑖𝑥 𝑏𝑚𝑖𝑥 𝑏𝑚𝑖𝑥 𝑅𝑇 1.5 𝑣+𝑏𝑚𝑖𝑥

135
 The Pitzer equations for calculating the activity coefficient of dissolved CO 2, 𝛾𝐶𝑂2 ,⁡ and the

osmotic coefficient of H2O,⁡𝜙, are given in Pitzer (1991):

2 2
𝑙𝑛𝛾𝐶𝑂2 = 2𝑚𝐶𝑂2 𝜆𝑛𝑛 + 3𝑚𝐶𝑂 𝜇
2 𝑛𝑛𝑛
+ 2𝑚𝑁𝑎𝐶𝑙 (𝜆𝑛𝑐 + 𝜆𝑛𝑎 ) + 6𝑚𝐶𝑂2 𝑚𝑁𝑎𝐶l (𝜇𝑛𝑛𝑐 + 𝜇𝑛𝑛𝑎 ) + 𝑚𝑁𝑎𝐶𝑙 𝜉𝑛𝑐𝑎 ⁡(A.6)

2 𝐼 1.5 2 (0) (1) 1

𝜙−1= {−𝐴𝜙 + 𝑚𝑁𝑎𝐶𝑙 [𝛽𝑐𝑎 + 𝛽𝑐𝑎 𝑒𝑥𝑝(−𝛼1 𝐼 0.5 ) + 2𝑚𝑁𝑎𝐶𝑙 𝐶𝑐𝑎 ] + ⁡ 𝑚𝐶𝑂
2
𝜆 +
2 𝑛𝑛
2𝑚𝑁𝑎𝐶𝑙 +𝑚𝐶𝑂2 𝐼+𝑏𝐼 0.5 2

3 2 2
𝑚𝐶𝑂 𝜇
2 𝑛𝑛𝑛
+ 𝑚𝐶𝑂2 𝑚𝑁𝑎𝐶𝑙 (𝜆𝑛𝑐 + 𝜆𝑛𝑎 ) + 3𝑚𝐶𝑂2
𝑚𝑁𝑎𝐶𝑙 (𝜇𝑛𝑛𝑐 + 𝜇𝑛𝑛𝑎 ) + 𝑚𝐶𝑂2 𝑚𝑁𝑎𝐶𝑙 𝜉𝑛𝑐𝑎 } (A.7)

(0) (1) 𝜙
In Eqs. (A.6) and (A.7), the Pitzer ion-ion interaction parameters 𝛽𝑐𝑎 ,⁡𝛽𝑐𝑎 and 𝐶𝑐𝑎 for NaCl-H2O

system are taken from Pitzer et al. (1984). The molality of CO2 and NaCl in the aqueous phase are 𝑚𝐶𝑂2

and 𝑚𝑁𝑎𝐶𝑙 , respectively; 𝐴𝜙 is the Debye-Hückel slope for the osmotic coefficient, 𝑏 = 1.2 kg1/2 mol-1/2,

and 𝛼1 = 2.0 kg1/2 mol-1/2; and I is the ionic strength based on the molality scale. The Debye-Hückel slope

and the ionic strength are defined as follows:

3/2
1 1/2 𝑒2
𝐴𝜙 = (2𝜋𝑁𝐴 𝜌𝐻2𝑂 ) ( ) (A.8)
3 4𝜋𝜀𝑜 𝜀𝑟 𝑘𝑇

1
𝐼 = ∑𝑖 𝑚𝑖 𝑧𝑖2 (A.9)
2

where 𝜌𝐻2𝑂 is the density of water (g/L) calculated by the IAPWS-95 formulation (Wagner and Pruß, 2002);

𝜀𝑟 is the relative permittivity of pure water calculated from the equation of Fernández et al. (1997), which is

adopted as the IAPWS standard for 𝜀𝑟 ; 𝜀𝑜 = 8.85419⁡10-12 (F/m) is the vacuum permittivity; e = 1.60218 10-
19
(C) is the elementary charge; k = 1.38065 10-23 (J K-1) is the Boltzmann constant; and T is temperature in

K.

The combined Pitzer interaction parameters 𝐵𝐶𝑂2 −𝑁𝑎𝐶𝑙 and 𝐶𝐶𝑂2 −𝐶𝑂2 −𝑁𝑎𝐶𝑙 are adopted from

Akinfiev and Diamond (2010),

136
 100 𝑎3
𝐵𝐶𝑂2 −𝑁𝑎𝐶𝑙 = 𝑎1 + 𝑎2 + [1 − (1 + 2√𝐼)𝑒𝑥𝑝(−2√𝐼)] (A.10)
𝑇−𝜃 2𝐼

𝐶𝐶𝑂2 −𝐶𝑂2 −𝑁𝑎𝐶𝑙 = 𝑎4 (A.11)

where⁡𝜃 = 228⁡𝐾 and a constant, I is molality based ionic strength defined by Eq. (A.9). The parameters in

Eqs. (A.10) and (A.11) are given in Table A 1 (Akinfiev and Diamond, 2010).

Table A 1 Parameters for Eqs. (A.10-A.11).

Eqs. a1 a2 a3 a4
A10 5.7123 10-2 2.6994 10-02 4.5635 10-02 -
A11 - - - -5.76 10-04

137
Appendix B: Pitzer Ion-Ion Interaction Model

(0) (1) 𝜑
The empirical function f(T,P) to calculate Pitzer ion-ion interaction parameters (𝛽𝑐𝑎 , 𝛽𝑐𝑎 and 𝐶𝑐𝑎 )

at the desired temperatures and pressures for the system of CaCl 2-H2O, Na2SO4-H2O, MgCl2-H2O and KCl-

H2O are summarized as below:

1. CaCl2-H2O system (Holmes et al. 1994)

𝑓(𝑇, 𝑃) = 𝐹0 + 𝐹1 (𝑃) + 𝐹2 (𝑃)2 (B.1)

where P is the system pressure in bar. Fo, F1 and F2 are functions of temperature only and given by Eqs.

(B.2) to (B.4),

1 1 1 1 5 𝑇 3𝑇22 𝑇2 𝑇𝑥 𝑇𝑦
𝐹0 = 𝑞1 + 𝑞2 𝑇 + 𝑞3 𝑇 2 + 𝑞4 𝑇 3 + 𝑞5 𝑇 2 {𝑙𝑛𝑇 − } + 𝑞6 { + + 𝑙𝑛𝑇𝑥 } + 𝑞7 {2 + 1} 𝑙𝑛𝑇𝑦 (B.2)
2 6 12 6 6 2 2𝑇 𝑇 𝑇

1
𝐹1 = 𝑞8 + 𝑞9 + 𝑞10 𝑇 + 𝑞11 𝑇 2 + 𝑇𝑥−1 𝑃12 + 𝑇𝑦−1 𝑃13 (B.3)
𝑇

1
𝐹2 = 𝑞14 + 𝑞15 + 𝑞16 𝑇 + 𝑞17 𝑇 2 (B.4)
𝑇

where T1 = 647 K, T2 = 227 K, Tx = (T-T2), and Ty = (T1-T). The constants q1~q17 were listed in Table B 1.

2. Na2SO4-H2O system (Rogers and Pitzer, 1981; Rumpf and Maurer, 1993)

𝑇 1 1 1 1
𝑓(𝑇) = 𝑞1 + 𝑞2 (𝑇 2 − 𝑇𝑅2 ) + 𝑞3 (𝑇 − 𝑇𝑅 ) + 𝑞4 𝑙𝑛 ( ) + 𝑞5 ( − ) + 𝑞6 𝑇2 ( − )+
𝑇𝑅 𝑇−𝑇1 𝑇𝑅 −𝑇1 𝑇(𝑇−𝑇2 ) 𝑇𝑅 (𝑇𝑅 −𝑇2 )

1 1
𝐴( − ) (B.5)
𝑇 𝑇𝑅

138
 1 𝑇1 𝑇2 1
𝐴 = 𝑞7 + 𝑞8 𝑇𝑅 + 𝑞9 𝑇𝑅2 + 𝑞10 𝑇𝑅3 + 2𝑇1 𝑞11 ( + ) + 2𝑇2 𝑞12 ( − ) (B.6)
𝑇𝑅 −𝑇1 2(𝑇𝑅 −𝑇1 )2 2(𝑇𝑅 −𝑇2 )2 𝑇2 −𝑇𝑅

where 𝑇𝑅 is the reference temperature which is conveniently set at 298.15 K. 𝑇1 is 263 K, and 𝑇2 is 263 or

680 K depending on which ion-interaction parameter is to be calculated (Table B 2).

3. MgCl2-H2O system (Wang and Pitzer, 1998)

Wang and Pitzer (1998) developed a general model that describes the thermodynamic properties

of MgCl2(aq) based on an ion-interaction treatment of a variety of thermodynamic properties. The

(0) (1) (0) (1)

equations for calculating the ion interaction parameters (𝛽𝑐𝑎 , 𝛽𝑐𝑎 ,𝐶𝑐𝑎 , 𝐶𝑐𝑎 ) are given below,

𝑃 𝑃 2
𝑓(𝑇, 𝑃) = 𝐹0 + 𝐹1 ( ) + 𝐹2 ( ) /2 (B.7)
10 10

where P is the system pressure in bar. F0, F1 and F2 are functions of temperature only and given by Eqs.

(B.8) to (B.11)

1 2
𝐹0 = 𝑞1 + 𝑞2 𝑙𝑛𝑇 + 𝑞3 𝑇 + 𝑞4 𝑇 2 + 𝑞5 𝑇 3 + 𝑞6 𝑇 10 + 𝑞7 ( ) (B.8)
𝑇1 −𝑇

1 2
𝐹1 = 𝑞8 + 𝑞9 𝑙𝑛𝑇 + 𝑞10 𝑇 + 𝑞11 𝑇 2 + 𝑞12 𝑇 3 + 𝑞13 𝑇 10 + 𝑞14 ( ) (B.9)
𝑇1 −𝑇

1 2
𝐹2 = 𝑞15 + 𝑞16 𝑙𝑛𝑇 + 𝑞17 𝑇 + 𝑞18 𝑇 2 + 𝑞19 𝑇 3 + 𝑞20 𝑇 10 + 𝑞21 ( ) (B.10)
𝑇1 −𝑇

𝜙 (0) (1) (2)

𝐶𝑐𝑎 = 2[𝐶𝑐𝑎 + 𝐶𝑐𝑎 𝑒𝑥𝑝(−𝑥𝑐1 ) + 𝐶𝑐𝑎 𝑒𝑥𝑝(−𝑥𝑐2 )] (B.11)

where 𝑥𝑐1 = 𝛼𝑐1 𝐼; 𝑥𝑐2 = 𝛼𝑐2 𝐼, 𝛼𝑐1 = 0.4⁡𝑘𝑔⁡𝑚𝑜𝑙 −1 , 𝛼𝑐2 = 0.28⁡𝑘𝑔⁡𝑚𝑜𝑙 −1 , and T1 = 647K. The constants

q1~q21 are shown in Table B 3.

139
4. KCl-H2O system (Holmes and Mesmer, 1983)

1 1 𝑇
𝑓(𝑇) = 𝑞1 + 𝑞1 ( − ) + 𝑞3 𝑙𝑛 ( )
𝑇 𝑇𝑅 𝑇𝑅

+𝑞4 (𝑇 − 𝑇𝑅 ) + 𝑞5 (𝑇 2 − 𝑇𝑅2 ) + 𝑞6 𝑙𝑛⁡(𝑇 − 260) (B.12)

where 𝑞1 -𝑞6 are constants listed in Table B 4, and 𝑇𝑅 is 298.15 K.

140
 (𝟎) (𝟏)
Table B 1 Constants of Eqs. (B.1) – (B.4) for calculating Pitzer ion-interaction parameters (𝜷𝒄𝒂 , 𝜷𝒄𝒂 and
𝝓
𝑪𝒄𝒂 ) of CaCl2(aq) (Holmes et al., 1994).
(𝟎) (𝟏) 𝝓
Const. 𝒇(𝑻, 𝑷) = 𝜷𝒄𝒂 𝒇(𝑻, 𝑷) = 𝜷𝒄𝒂 𝒇(𝑻, 𝑷) = 𝑪𝒄𝒂

𝑞1 0 0 -1.3455 10-1
𝑞2 4.9213 10-3 -1.3814 10-1 0
𝑞3 -3.5512 10-5 1.6522 10-2 3.0401 10-4
𝑞4 4.7629 10-8 6.3784 10-6 1.3136 10-7
𝑞5 0 -3.1030 10-3 -5.8863 10-5
𝑞6 0 -2.0329 10-2 -6.4986 10-4
𝑞7 -3.5549 10-4 0 0
𝑞8 1.1021 10-3 0 -9.0317 10-7
𝑞9 -1.3924 10-1 0 0
𝑞10 -2.8663 10-6 1.0935 10-6 0
𝑞11 2.9609 10-9 -4.0084 10-9 5.9573 10-12
𝑞12 2.3285 10-3 0 0
𝑞13 -2.1508 10-2 0 0
𝑞14 0 0 -5.5630 10-9
𝑞15 0 0 1.7685 10-6
𝑞16 -1.2534 10-10 0 0
𝑞17 3.5462 10-13 0 0

141
 (𝟎) (𝟏)
Table B 2 Constants in Eqs. (B.5) – (B.6) for calculating Pitzer ion-interaction parameters (𝜷𝒄𝒂 , 𝜷𝒄𝒂 and
𝝓
𝑪𝒄𝒂 ) of Na2SO4(aq) (Rumpf and Maurer, 1993).
(𝟎) (𝟏) 𝝓
Const. 𝒇(𝑻) = 𝜷𝒄𝒂 𝒇(𝑻) = 𝜷𝒄𝒂 * 𝒇(𝑻) = 𝑪𝒄𝒂
𝑞1 1.869 10 -2
1.0994 5.54900 10 -3

𝑞2 -1.03611 10-5 -3.2355 10-4 0

𝑞3 3.00299 10-2 5.76552 10-1 5.14316 10-5
𝑞4 -1.43441 10 +1
-1.88769 10 +2
0
𝑞5 -6.66894 10-1 -2.05974 10-1 0
𝑞6 0 -1.46744 10+3 3.45791 10-1
𝑞7 -2.081437 10 +2
-5.29421 10 +2
4.25799 10+1
𝑞8 -1.43441 10 +1
-1.88769 10 +2
0
𝑞9 3.00299 10-2 5.76552 10-1 5.14316 10-5
𝑞10 -2.07222 10 -5
-6.471 10 -4
0
𝑞11 6.66894 10-1 2.05974 10-1 -3.45791 10-1
𝑞12 0 1.46744 10+3 0
(1)
*When calculating 𝛽𝑐𝑎 , T2 in Eqs. (B.5) and (B.6) is 680K, otherwise T2 is 263K

142
 (𝟎) (𝟏) 𝝓
Table B 3 Constants of Eqs. (B.7) – (B.10) for calculating Pitzer ion-interaction parameters (𝜷𝒄𝒂 , 𝜷𝒄𝒂 and 𝑪𝒄𝒂 ) of MgCl2(aq) (Wang and Pitzer, 1998).
𝝓 (𝟎) (𝟏) (𝟐)
(𝟎) (𝟏) 𝑪𝒄𝒂 = 𝟐[𝑪𝒄𝒂 + 𝑪𝒄𝒂 𝒆𝒙𝒑(−𝒙𝒄𝟏 ) + 𝑪𝒄𝒂 𝒆𝒙𝒑(−𝒙𝒄𝟐 )] (Eq. A11)
Coeff. 𝒇(𝑻, 𝑷) = 𝜷𝒄𝒂 𝒇(𝑻, 𝑷) = 𝜷𝒄𝒂 (𝟎) (𝟏) (𝟐)
𝒇(𝑻, 𝑷) = 𝑪𝒄𝒂 𝒇(𝑻, 𝑷) = 𝑪𝒄𝒂 𝒇(𝑻, 𝑷) = 𝑪𝒄𝒂
𝑞1 -5.50111455 10+1 7.21220552 10+1 5.92428240 0 0
𝑞2 1.50130326 10+1 -1.77145085 10+1 -1.65126386 -1.02256042 0
𝑞3 -1.58107430 10-1 1.14397153 10-1 1.89399822 10-2 3.77018617 10-2 -2.28040769 10-3
𝑞4 2.30409919 10-4 0 -2.99972128 10-5 -7.91682934 10-5 1.37425889 10-5
𝑞5 -1.31768095 10-7 -1.43588435 10-7 1.89174291 10-8 5.91314258 10-8 -1.94821902 10-8
𝑞6 -1.26699609 10-28 1.72952766 10-27 0 0 1.04649784 10-28
𝑞7 2.82197499 10+2 3.41920714 10+3 5.49030201 10+1 -2.28493084 10+2 0
𝑞8 0 0 4.50114048 10-2 0 0
𝑞9 0 2.28440612 10-4 -1.08427926 10-2 0 0
𝑞10 8.39661960 10-5 0 7.41041864 10-5 -7.79259941 10-5 0
𝑞11 -4.60207270 10-7 0 -5.99961498 10-8 4.28675876 10-7 0
𝑞12 6.21165614 10-10 0 0 -5.77509662 10-10 0
𝑞13 8.43555937 10-31 -1.77573402 10-29 0 0 0
𝑞14 0 -2.29668879 10+2 -4.60562847 0 0
𝑞15 0 0 0 -5.13962051 10-4 0
𝑞16 0 0 0 9.30761142 10-5 0
𝑞17 0 -2.71485086 10-7 0 0 0
𝑞18 0 0 0 0 0
𝑞19 0 0 -1.39016981 10-15 -7.43350922 10-13 0
𝑞20 0 0 0 0 0
𝑞21 -1.11176553 1.01000272 10+1 1.40556304 10-1 1.12721557 0

143
 (𝟎) (𝟏) 𝝓
Table B 4 Constants of Eq. (B.12) for calculating Pitzer ion-interaction parameters (𝜷𝒄𝒂 , 𝜷𝒄𝒂 and 𝑪𝒄𝒂 ) of
KCl(aq) (Holmes and Mesmer, 1983).
(𝟎) (𝟏) 𝝓
Const. 𝒇(𝑻) = 𝜷𝒄𝒂 𝒇(𝑻) = 𝜷𝒄𝒂 𝒇(𝑻) = 𝑪𝒄𝒂
𝑞1 4.808 10-2 4.76 10-2 -7.88 10-4
𝑞2 -7.5848 10+2 3.039 10+2 9.127 10+1
𝑞3 -4.7062 1.066 5.8643 10-1
𝑞4 1.0072 10-2 0 -1.298 10-3
𝑞5 -3.7599 10-6 0 4.9567 10-7
𝑞6 0 4.7 10-2 0

144
Appendix C: Experimental Error Estimation

2-D interpolation of scatter experimental data. Multiple measurements (usually 3 to 6

measurements) were made at each target experimental P-T-x point. The inverse distance weighted 2-D

interpolation method (Shepard, 1968) was used to get an approximation of the real value based on scattered

data generated by measurements at each interpolation point. Based on this approach, the value of the

interpolation point in a set, f(x, y), is defined as:

𝑓(𝑥, 𝑦) = ∑𝑛𝑖=1 𝑓𝑖 (𝑥𝑖 , 𝑦𝑖 ) 𝑊𝑖 (𝑥𝑖 , 𝑦𝑖 ) (C.1)

where n is the number of scatter (experimental) points in the set, f i(xi, yi) is the value of the i-th scatter

point, and Wi(xi, yi) is the weight function assigned to the i-th scatter point.

The weight function is defined as,

ℎ𝑖−2
𝑊𝑖 (𝑥𝑖 , 𝑦𝑖 ) = ∑𝑛 −2 (C.2)
𝑗=1 ℎ𝑗

where hi is the distance from the interpolation point to the i-th scatter point and is defined as:

ℎ𝑖 = √(𝑥 − 𝑥𝑖 )2 + (𝑦 − 𝑦𝑖 )2 (C.3)

Error estimation. The CO2 solubility in the aqueous phase is commonly expressed by either

molality (𝑚𝐶𝑂2 ) or mole fraction (𝑥𝐶𝑂2 ). The error estimation is performed on the mole fraction scale, but

the obtained results can be converted to the molality scale. The mole fraction of CO 2 in an aqueous sample

is defined as,

𝑀𝐶𝑂
2
𝑀𝑊𝐶𝑂
2
𝑥𝐶𝑂2 = 𝑀𝐶𝑂 𝑀 −𝑀𝐶𝑂 −𝑀𝑠𝑎𝑙𝑡 𝑀 (C.4)
2 + 𝑠𝑎𝑚𝑝𝑙𝑒 2 + 𝑠𝑎𝑙𝑡 (𝜈𝑎 +𝜈𝑐 )
𝑀𝑊𝐶𝑂 𝑀𝑊𝐻 𝑂 𝑀𝑊𝑠𝑎𝑙𝑡
2 2

145
where⁡𝑀𝐶𝑂2 is the mass of CO2 in the sample, ⁡𝑀𝑠𝑎𝑚𝑝𝑙𝑒 is the total mass of sample, ⁡𝑀𝑠𝑎𝑙𝑡 ⁡is the mass of

salt in the sample;⁡⁡𝑀𝑊𝐶𝑂2 , ⁡𝑀𝑊𝐻2𝑂 ⁡and ⁡𝑀𝑊𝑠𝑎𝑙𝑡 are the molar masses of CO2, H2O and NaCl, respectively;

𝜈𝑎 and 𝜈𝑐 are, respectively, the stoichiometric numbers of the anion and cation of the dissolved salt.

The error propagation theory described by Taylor (1997) is used to determine the instrumental

error of each measured point. From Eq. (C.4), the error in 𝑥𝐶𝑂2 due to the pressure variation in the pressure

cell is:

𝑃 𝜕𝑥𝐶𝑂2 𝜕𝑥𝐶𝑂2 𝜕𝑀𝐶𝑂2

𝛿𝑥𝑒𝑟𝑟 =| | 𝛿𝑃 = | | 𝛿𝑃 (C.5)
𝜕𝑃𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒⁡𝑐𝑒𝑙𝑙 𝜕𝑀𝐶𝑂2 𝜕𝑃𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒⁡𝑐𝑒𝑙𝑙

The error in 𝑥𝐶𝑂2 due to the temperature variation of pressure cell is:

𝑇 𝜕𝑥𝐶𝑂2 𝜕𝑥𝐶𝑂2 𝜕𝑀𝐶𝑂2

𝛿𝑥𝑒𝑟𝑟 =| | 𝛿𝑇 = | | 𝛿𝑇 (C.6)
𝜕𝑇𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒⁡𝑐𝑒𝑙𝑙 𝜕𝑀𝐶𝑂2 𝜕𝑇𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒⁡𝑐𝑒𝑙𝑙

The error in 𝑥𝐶𝑂2 due to the volume variation is:

𝑉 𝜕𝑥𝐶𝑂2 𝜕𝑥𝐶𝑂2 𝜕𝑀𝐶𝑂2

𝛿𝑥𝑒𝑟𝑟 =| | 𝛿𝑉 = | | 𝛿𝑉 (C.7)
𝜕𝑉𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒⁡𝑐𝑒𝑙𝑙 𝜕𝑀𝐶𝑂2 𝜕𝑉𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒⁡𝑐𝑒𝑙𝑙

The error in 𝑥𝐶𝑂2 due to the sample mass variation is:

𝑠𝑎𝑚𝑝𝑙𝑒 𝜕𝑥𝐶𝑂2
𝛿𝑥𝑒𝑟𝑟 =| | 𝛿𝑀𝑠𝑎𝑚𝑝𝑙𝑒 (C.8)
𝜕𝑀𝑠𝑎𝑚𝑝𝑙𝑒

The error in 𝑥𝐶𝑂2 due to the salt mass variation is:

𝑠𝑎𝑙𝑡 𝜕𝑥𝐶𝑂2
𝛿𝑥𝑒𝑟𝑟 =| | 𝛿𝑀𝑠𝑎𝑙𝑡 (C.9)
𝜕𝑀𝑠𝑎𝑙𝑡

146
 From Eq. (C.4), the derivative of 𝑥𝐶𝑂2 with respect to 𝑀𝐶𝑂2 is,

1 𝑀𝐶𝑂 𝑀 −𝑀𝐶𝑂 −𝑀𝑠𝑎𝑙𝑡 𝑀 𝑀𝐶𝑂 1 1

( 2 + 𝑠𝑎𝑚𝑝𝑙𝑒 2 + 𝑠𝑎𝑙𝑡 (𝜈𝑎 +𝜈𝑐 ))− 2 ( − )
𝜕𝑥𝐶𝑂2 𝑀𝑊𝐶𝑂 𝑀𝑊𝐶𝑂
2 2
𝑀𝑊𝐻 𝑂
2
𝑀𝑊𝑠𝑎𝑙𝑡 𝑀𝑊𝐶𝑂 𝑀𝑊𝐶𝑂
2 2
𝑀𝑊𝐻 𝑂
2
= 2 (C.10)
𝜕𝑀𝐶𝑂2 𝑀𝐶𝑂 𝑀
2 + 𝑠𝑎𝑚𝑝𝑙𝑒
−𝑀𝐶𝑂 −𝑀𝑠𝑎𝑙𝑡 𝑀
2
( + 𝑠𝑎𝑙𝑡 (𝜈𝑎 +𝜈𝑐 ))
𝑀𝑊𝐶𝑂 𝑀𝑊𝐻 𝑂 𝑀𝑊𝑠𝑎𝑙𝑡
2 2

1 2 3 1 2 3
𝑀𝐶𝑂2 = 𝑀𝐶𝑂2
+ 𝑀𝐶𝑂2
+ 𝑀𝐶𝑂2
= (𝜌𝐶𝑂2
+ 𝜌𝐶𝑂2
+ 𝜌𝐶𝑂2
)𝑉𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒⁡𝑐𝑒𝑙𝑙 (C.11)

1 2 3 1 2
In Eq. (C.11) 𝑀𝐶𝑂2
, 𝑀𝐶𝑂2
⁡and 𝑀𝐶𝑂2
represent the CO2 amount by mass, as well as 𝜌𝐶𝑂2
, 𝜌𝐶𝑂2
, and

3
𝜌𝐶𝑂2
represent the CO2 density in the pressure cell, due to the 1st, 2nd and 3rd expansion during sample

analysis process, respectively. Also, the Ppressure cell, Tpressure cell, and Vpressure_cell in Eqs. (C.5- C.11) denote the

pressure (bar), temperature (K) and volume (cm3) of the pressure cell. Other partial derivatives are obtained

as follows:

1
2 2
2 3 2
𝜕𝑀𝐶𝑂2 𝜕𝑀𝐶𝑂 𝜕𝑀𝐶𝑂 𝜕𝑀𝐶𝑂
( ) 𝛿𝑃 = √(( 2
) 𝛿𝑃1 ) + (( 2
) 𝛿𝑃2 ) + (( 2
) 𝛿𝑃3 ) (C.12)
𝜕𝑃 𝑉,𝑇 𝜕𝑃 𝑉,𝑇1 𝜕𝑃 𝑉,𝑇2 𝜕𝑃 𝑉,𝑇3

1
2 2
2 3 2
𝜕𝑀𝐶𝑂2 𝜕𝑀𝐶𝑂 𝜕𝑀𝐶𝑂 𝜕𝑀𝐶𝑂
( ) 𝛿𝑇 = √(( 2
) 𝛿𝑇1 ) + (( 2
) 𝛿𝑇2 ) + (( 2
) 𝛿𝑇3 ) (C.13)
𝜕𝑇 𝑉,𝑃 𝜕𝑇 𝑉,𝑃1 𝜕𝑇 𝑉,𝑃2 𝜕𝑇 𝑉,𝑃3

where 𝑇1 , 𝑇2 , 𝑇3 ⁡and 𝑃1 , 𝑃2 , 𝑃3 are the equilibrium temperatures and pressures of the pressure cell during 1st,

2nd and 3rd expansion; 𝛿𝑇1 , 𝛿𝑇2 , 𝛿𝑇3 and 𝛿𝑃1 , 𝛿𝑃2 , 𝛿𝑃3 are the temperature and pressure changes of the
1 2 3
𝜕𝑀𝐶𝑂 2
𝜕𝑀𝐶𝑂 2
𝜕𝑀𝐶𝑂 2
pressure cell during 1st, 2nd and 3rd expansion, respectively. The terms ( ) ,( ) ,( )
𝜕𝑃 𝜕𝑃 𝜕𝑃
𝑉,𝑇1 𝑉,𝑇2 𝑉,𝑇3

are determined by the slope of the isothermal curve generated by the EoS of Span and Wagner (1996).
1 2 3
𝜕𝑀𝐶𝑂 2
𝜕𝑀𝐶𝑂 2
𝜕𝑀𝐶𝑂 2
Similarly, the terms ( ) ,( ) ,( ) are determined by the slope of isobaric curve
𝜕𝑇 𝜕𝑇 𝜕𝑇
𝑉,𝑃1 𝑉,𝑃2 𝑉,𝑃3

generated by the same EoS.

147
 Substituting (C.10) and (C.12) into (C.5), (C.10) and (C.13) into (C.6), (C.10) and (C.11) into
𝑃 𝑇 𝑉
(C.7), respectively, the 𝛿𝑥𝑒𝑟𝑟 , 𝛿𝑥𝑒𝑟𝑟 , 𝛿𝑥𝑒𝑟𝑟 can be calculated. By differentiation of Eq. (C.4) with respect to

𝑀𝑠𝑎𝑚𝑝𝑙𝑒 and 𝑀𝑠𝑎𝑙𝑡 , the Eqs. (C.14) and (C.15) can be obtained as:

1 𝑀𝐶𝑂
2
𝜕𝑥𝐶𝑂2 𝑀𝑊𝐻 𝑂 𝑀𝑊𝐶𝑂
2 2
=− 2 (C.14)
𝜕𝑀𝑠𝑎𝑚𝑝𝑙𝑒 𝑀𝐶𝑂 𝑀
2 + 𝑠𝑎𝑚𝑝𝑙𝑒
−𝑀𝐶𝑂 −𝑀𝑠𝑎𝑙𝑡 𝑀
2
( + 𝑠𝑎𝑙𝑡 (𝜈𝑎 +𝜈𝑐 ))
𝑀𝑊𝐶𝑂 𝑀𝑊𝐻 𝑂 𝑀𝑊𝑠𝑎𝑙𝑡
2 2

1 𝜈 +𝜈 𝑀𝐶𝑂
( − 𝑎 𝑐 ) 2
𝜕𝑥𝐶𝑂2 𝑀𝑊𝐻 𝑂 𝑀𝑊𝑠𝑎𝑙𝑡 𝑀𝑊𝐶𝑂
2 2
= 2 (C.15)
𝜕𝑀𝑠𝑎𝑙𝑡 𝑀𝐶𝑂 𝑀
2 + 𝑠𝑎𝑚𝑝𝑙𝑒
−𝑀𝐶𝑂 −𝑀𝑠𝑎𝑙𝑡 𝑀
2
( + 𝑠𝑎𝑙𝑡 (𝜈𝑎 +𝜈𝑐 ))
𝑀𝑊𝐶𝑂 𝑀𝑊𝐻 𝑂 𝑀𝑊𝑠𝑎𝑙𝑡
2 2

𝑠𝑎𝑚𝑝𝑙𝑒 𝑠𝑎𝑙𝑡
Substituting (C.14) into (C.8), and (C.15) into (C.9), respectively, the 𝛿𝑥𝑒𝑟𝑟 and 𝛿𝑥𝑒𝑟𝑟 can be

calculated. Thus, the calculated instrumental error of the sample analysis can be estimated by the error

propagation as follows:

𝑆𝑎𝑚𝑝𝑙𝑒 2
𝛿𝑥𝑖𝑛𝑠𝑡𝑟 = √(𝛿𝑥𝑒𝑟𝑟
𝑃 )2 + (𝛿𝑥 𝑇 )2 + (𝛿𝑥 𝑉 )2 + (𝛿𝑥
𝑒𝑟𝑟 𝑒𝑟𝑟 𝑒𝑟𝑟 ) + (𝛿𝑥𝑒𝑟𝑟
𝑠𝑎𝑙𝑡 )2 (C.16)

In this study, the calculated instrumental error (δxinstr) is found to be below 0.7 % of the mole

fraction of CO2 in the aqueous phase. The instrumental error on the molality scale (𝛿𝑚𝑖𝑛𝑠𝑡𝑟 ) can be

converted from the mole fraction scale as below:

55.509𝛿𝑥𝑖𝑛𝑠𝑡𝑟 +(𝜈𝑎 +𝜈𝑐 )𝑚𝑠𝑎𝑙𝑡 𝛿𝑥𝑖𝑛𝑠𝑡𝑟

𝛿𝑚𝑖𝑛𝑠𝑡𝑟 = ⁡ (C.17)
1−𝛿𝑥𝑖𝑛𝑠𝑡𝑟

𝛿𝑚𝑖𝑛𝑠𝑡𝑟
and < 0.77%⁡ for all experimental CO2 solubility.
𝑚𝐶𝑂2

148
 The random errors were estimated from direct repetitions of the measurement results. The standard

deviation between 2-D interpolation point and measured experimental value represents the random error of

the value during the experiment:

1
𝛿𝑥𝑟𝑎𝑛𝑑𝑜𝑚 = ⁡ √ ∑𝑛𝑖=1(𝑥𝑖 − 𝑥2−𝐷 )2 (C.18)
𝑛−1

where 𝑥2−𝐷 ⁡is 2-D interpolation value of the measured solubility data. The random error based on molality

scale (𝛿𝑚𝑟𝑎𝑛𝑑𝑜𝑚 ) can also be converted from 𝛿𝑥𝑟𝑎𝑛𝑑𝑜𝑚 in the same manner as Eq. (C.17) does.

Symbols

hi distance between scattered (experimental) and interpolation points

𝑥2−𝐷 2-D interpolation value of scattered measured CO2 solubility, mole fraction

𝑀𝐶𝑂2 mass of CO2 in the sample cell, g

𝑀𝑠𝑎𝑙𝑡 mass of dissolved salt in the sample

𝑀𝑠𝑎𝑚𝑝𝑙𝑒 mass of sample, g

1 2 3
𝑀𝐶𝑂2
, 𝑀𝐶𝑂2
, 𝑀𝐶𝑂2
mass of CO2 of 1st, 2nd, and 3rd CO2 expansion

𝑀𝑊𝑖 molecular weight of a component, g mol-1

𝑃𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒⁡𝑐𝑒𝑙𝑙 pressure of the pressure cell during sample analysis process, bar

𝑃1 , 𝑃2 , 𝑃3 equilibrium pressure of pressure cell during 1st, 2nd, and 3rd CO2 expansion

𝑇1 , 𝑇2 , 𝑇3 equilibrium temperature of pressure cell during 1 st, 2nd, and 3rd CO2 expansion, K

𝑇𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒⁡𝑐𝑒𝑙𝑙 temperature of the pressure cell, K

𝑉𝑝𝑟𝑒𝑠𝑠𝑢𝑟e⁡𝑐𝑒𝑙𝑙 volume of the stainless steel vessel for measuring the CO2 solubility during sample

analysis process, 297.27 cm3

𝛿𝑃1 , 𝛿𝑃2 , 𝛿𝑃3 pressure difference of pressure cell during 1st, 2nd, and 3rd CO2 expansion

𝛿𝑇1 , 𝛿𝑇2 , 𝛿𝑇3 temperature difference of pressure cell during 1st, 2nd, and 3rd CO2 expansion

𝛿𝑉 estimated error for the pressure cell volume

149
Appendix D: The PSUCO2 Web-Computational Interface

Figure D. 1 User-interface of CO2-brine phase equilibria model. Part I: CO2 solubility in pure H2O and in
single-salt brine; Part II: CO2 solubility in mixed-salt brine.

Can be accessed at: http://carbonlab.org/psuco2/index.php.

150
Figure D. 2 User-interface of pure CO2 model, developed based on Span and Wagner’s (1996) EoS.

Can be accessed at: http://carbonlab.org/CO2eos/index.php.

Figure D. 3 User-interface of brine density model.

Can be accessed at: http://carbonlab.org/density/index.php.

151
Appendix E. Original Experimental Data Records

The experimental CO2 solubility data recorded herein were measured during the periods of August

2011- February 2014. The experimental apparatus is located in the Room 114, Academic Projects Building,

The Energy Institute, The Pennsylvania State University, University Park. The correction factor 0.992

should be applied to all obtained raw experimental data listed here to compensate water vapor error. Please

refer to Section 2.3.4 for the explanation of the error caused by water vapor during the sample analysis

process.

152
Table E 1 Original Experimental CO2 solubility data record at 1453-2190 psia and 322.15-423.15 K for the CO2-H2O system.

INVERSE WEIGHTING
𝒎𝒔𝒂𝒍𝒕 𝒎′𝑪𝑶𝟐
Run No #. P / psia T/K 𝒙𝑪𝑶𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝒉𝟐𝒊 , Eq. (C.3) Eq. (C.2)
1 2181 323.15 0 1.2550 2.2109 0.0339 0.5120
2 2170 323.15 0 1.2616 2.2223 0.0323 0.4880
3 2178 323.15 0 1.2642 2.2268 NA NA Discarded
4 2176 374.15 0 1.0342 1.8290 NA NA Discarded
5 2163 373.15 0 1.0297 1.8212 0.0063 0.4142
6 2165 373.15 0 1.0265 1.8157 0.0090 0.5858
7 2190 423.15 0 1.0168 1.7988 0.0048 0.0862
8 2180 423.15 0 1.0148 1.7954 0.0509 0.9138
9 2190 423.15 0 1.0254 1.8138 NA NA Discarded
10 1448 323.15 0 1.1296 1.9944 0.1770 0.4569
11 1453 322.15 0 1.1738 2.0707 0.1269 0.3276
12 1447 322.15 0 1.1626 2.0515 0.0806 0.2081
13 1469 323.15 0 1.1683 2.0612 0.0029 0.0074
14 2172 352.15 0 1.0789 1.9066 - -
15 2136 352.65 0 1.1158 1.9705 - -
′ ′
Note: 𝑚𝐶𝑂2
means the original measured data, the real CO2 solubility (𝑚𝐶𝑂2 ) obtained in this study equals to 𝑚𝐶𝑂2 = 0.992𝑚𝐶𝑂 2

153
Table E 2 Original Experimental CO2 solubility data record at 2150-2200 psia, 322.65-427.15 K, and from 1-6 mol kg-1 NaCl for the CO2-NaCl-H2O system.

INVERSE WEIGHTING
𝒎𝑵𝒂𝑪𝒍 𝒎′𝑪𝑶𝟐
Run No #. P / psia T/K 𝒙𝑪𝑶𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝒉𝟐𝒊 , Eq. (C.3) Eq. (C.2)
14 2177 323.15 1 1.0418 1.7793 NA NA Discarded
15 2173 323.15 1 1.0257 1.7523 0.1519 0.6589
16 2172 323.15 1 1.0216 1.7454 0.0786 0.3411
17 2153 373.15 1 0.8421 1.4432 0.0020 0.0004
18 2193 376.15 1 0.8269 1.4175 0.0032 0.0006
19 2181 374.15 1 0.8387 1.4374 0.0328 0.0061
20 2188 373.65 1 0.8447 1.4476 0.0065 0.0012
21 2176 373.15 1 0.8425 1.4438 5.3091 0.9917
22 2154 422.15 1 0.7970 1.3669 0.0021 0.0043
23 2161 424.15 1 0.8299 1.4226 0.0047 0.0095
24 2150 423.15 1 0.7918 1.3581 0.0015 0.0031
25 2177 423.15 1 0.8060 1.3821 0.4863 0.9831
26 2178 322.65 2 0.8772 1.4527 0.1620 0.0488
27 2170 323.15 2 0.8736 1.4468 0.0323 0.0097
28 2175 323.15 2 0.8651 1.4329 3.1215 0.9414
29 2174 373.15 2 0.7121 1.1825 0.4078 0.7579
30 2200 373.15 2 0.7236 1.2013 0.0017 0.0031
31 2171 373.15 2 0.7089 1.1772 0.0480 0.0892
32 2172 373.15 2 0.7097 1.1785 0.0786 0.1462
33 2153 371.15 2 0.7161 1.1890 0.0019 0.0036
34 2162 426.15 2 0.6654 1.1058 0.0052 0.2578
35 2163 426.15 2 0.6257 1.0405 0.0060 0.2982
36 2192 427.15 2 0.6581 1.0938 0.0035 0.1740
37 2162 422.65 2 0.6644 1.1041 0.0054 0.2701
38 2181 323.15 3 0.7167 1.1517 0.0339 0.0727
39 2165 323.15 3 0.7325 1.1768 0.0090 0.0192
40 2174 322.65 3 0.7434 1.1942 0.3700 0.7949

154
 INVERSE WEIGHTING
𝒎𝑵𝒂𝑪𝒍 𝒎′𝑪𝑶𝟐
Run No #. P / psia T/K 𝒙𝑪𝑶𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝒉𝟐𝒊 , Eq. (C.3) Eq. (C.2)
41 2171 323.15 3 0.7279 1.1695 0.0480 0.1030
42 2161 323.15 3 0.7267 1.1677 0.0047 0.0101
43 2192 373.15 3 0.6192 0.9966 0.0037 0.0291
44 2200 373.15 3 0.6208 0.9992 0.0017 0.0132
45 2172 373.15 3 0.6260 1.0075 0.0786 0.6185
46 2168 374.15 3 0.6223 1.0016 0.0172 0.1350
47 2181 376.15 3 0.6134 0.9874 0.0260 0.2041
48 2175 426.15 3 0.5739 0.9244 0.1073 0.0656
49 2175 422.15 3 0.5495 0.8855 0.7574 0.4630
50 2200 424.15 3 0.6030 0.9708 0.0017 0.0010
51 2195 426.15 3 0.5968 0.9609 0.0026 0.0016
52 2190 425.15 3 0.5921 0.9534 0.0047 0.0029
53 2190 424.15 3 0.5992 0.9648 0.0048 0.0029
54 2175 422.15 3 0.5903 0.9506 0.7574 0.4630
55 2178 322.65 4 0.6272 0.9780 0.1620 0.8513
56 2160 323.15 4 0.6347 0.9895 0.0041 0.0217
57 2182 323.15 4 0.6323 0.9857 0.0242 0.1270
58 2175 374.15 4 0.5316 0.8301 0.7574 0.2853
59 2175 373.65 4 0.5306 0.8285 1.7533 0.6604
60 2178 374.15 4 0.5380 0.8400 0.1444 0.0544
61 2176 422.15 4 0.5018 0.7839 0.8415 0.8228
62 2196 427.15 4 0.4744 0.7414 0.0023 0.0023
63 2173 423.65 4 0.5032 0.7861 0.1463 0.1431
64 2182 424.15 4 0.5042 0.7877 0.0236 0.0231
65 2165 423.15 4 0.5008 0.7824 0.0090 0.0088
66 2191 323.15 5 0.5483 0.8301 0.0042 0.0171
67 2172 323.15 5 0.5580 0.8445 0.0786 0.3212
68 2178 323.65 5 0.5544 0.8391 0.1620 0.6616

155
 INVERSE WEIGHTING
𝒎𝑵𝒂𝑪𝒍 𝒎′𝑪𝑶𝟐
Run No #. P / psia T/K 𝒙𝑪𝑶𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝒉𝟐𝒊 , Eq. (C.3) Eq. (C.2)
69 2183 374.15 5 0.4696 0.7117 0.0178 0.0803
70 2173 374.15 5 0.4756 0.7208 0.1319 0.5959
71 2173 376.15 5 0.4687 0.7104 0.0642 0.2900
72 2164 373.15 5 0.4658 0.7060 0.0075 0.0338
73 2181 424.15 5 0.4367 0.6622 0.0328 0.0179
74 2195 423.65 5 0.4457 0.6758 0.0026 0.0014
75 2175 423.65 5 0.4433 0.6722 1.7533 0.9558
76 2171 422.15 5 0.4435 0.6725 0.0458 0.0250
77 2176 323.15 6 0.4979 0.7321 5.3091 0.3333
78 2176 323.15 6 0.5056 0.7434 5.3091 0.3333
79 2176 323.15 6 0.5030 0.7395 5.3091 0.3333
80 2174 374.15 6 0.4182 0.6157 0.2897 0.0368
81 2176 373.65 6 0.4267 0.6281 2.2813 0.2895
82 2176 373.15 6 0.4320 0.6358 5.3091 0.6737
83 2176 424.15 6 0.3857 0.5681 0.8415 0.1472
84 2175 423.65 6 0.3993 0.5880 1.7533 0.3067
85 2175 423.15 6 0.3913 0.5763 3.1215 0.5461
′ ′
Note: 𝑚𝐶𝑂2
means the original measured data, the real CO2 solubility (𝑚𝐶𝑂2 ) obtained in this study equals to 𝑚𝐶𝑂2 = 0.992𝑚𝐶𝑂 2

156
Table E 3 Original Experimental CO2 solubility data record at 2171-2179 psia, 322.65-426.15 K, and from 0.3333-2.0000 mol kg-1 CaCl2 for the CO2-CaCl2-
H2O system.

INVERSE WEIGHTING
𝒎𝑪𝒂𝑪𝒍𝟐 𝒎′𝑪𝑶𝟐
Run No #. P / psia T/K 𝒙𝑪𝑶𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝒉𝟐𝒊 , Eq. (C.3) Eq. (C.2)
86 2176 323.15 0.3333 1.0969 1.9041 5.3091 0.4851
87 2177 324.15 0.3333 1.0742 1.8655 0.3272 0.0299
88 2176 323.15 0.3333 1.1115 1.9290 5.3091 0.4851
89 2174 373.15 0.3333 0.9106 1.5859 0.4078 0.7286
90 2173 373.15 0.3333 0.8909 1.5521 0.1519 0.2714
91 2176 425.65 0.3333 0.8486 1.4795 0.1553 0.1596
92 2173 423.15 0.3333 0.8652 1.5080 0.1519 0.1561
93 2177 424.65 0.3333 0.8452 1.4736 0.2322 0.2387
94 2177 423.65 0.3333 0.8986 1.5653 0.4336 0.4456
95 2177 323.15 0.6667 0.9640 1.6486 0.4863 0.4649
96 2173 323.15 0.6667 0.9513 1.6273 0.1519 0.1452
97 2174 323.15 0.6667 0.9915 1.6949 0.4078 0.3899
98 2177 373.15 0.6667 0.7758 1.3311 0.4863 0.5165
99 2174 373.15 0.6667 0.7923 1.3590 0.4078 0.4331
100 2171 372.65 0.6667 0.7838 1.3446 0.0474 0.0503
101 2178 425.15 0.6667 0.7240 1.2433 0.1008 0.4137
102 2172 423.15 0.6667 0.7367 1.2648 0.0786 0.3229
103 2173 426.15 0.6667 0.7277 1.2496 0.0642 0.2634
104 2173 322.65 1.0000 0.8502 1.4323 0.1463 0.4303
105 2171 323.65 1.0000 0.8532 1.4373 0.0474 0.1394
106 2173 323.65 1.0000 0.8720 1.4685 0.1463 0.4303
107 2173 373.15 1.0000 0.7019 1.1854 0.1519 0.3384
108 2177 375.15 1.0000 0.6826 1.1532 0.1651 0.3679
109 2173 374.15 1.0000 0.6812 1.1509 0.1319 0.2938
110 2174 425.15 1.0000 0.6365 1.0762 0.1550 0.2373
111 2174 423.15 1.0000 0.6357 1.0748 0.4078 0.6243

157
 INVERSE WEIGHTING
𝒎𝑪𝒂𝑪𝒍𝟐 𝒎′𝑪𝑶𝟐
Run No #. P / psia T/K 𝒙𝑪𝑶𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝒉𝟐𝒊 , Eq. (C.3) Eq. (C.2)
112 2177 426.15 1.0000 0.6416 1.0847 0.0904 0.1385
113 2173 323.15 1.3333 0.7493 1.2435 0.1519 0.2714
114 2174 323.15 1.3333 0.7525 1.2487 0.4078 0.7286
115 2174 373.15 1.3333 0.6116 1.0173 0.4078 0.3690
116 2174 372.15 1.3333 0.6219 1.0342 0.2897 0.2621
117 2174 373.15 1.3333 0.6247 1.0389 0.4078 0.3690
118 2175 423.15 1.3333 0.5543 0.9229 3.1215 0.8480
119 2174 423.15 1.3333 0.5515 0.9182 0.4078 0.1108
120 2173 423.15 1.3333 0.5627 0.9367 0.1519 0.0413
121 2176 323.15 1.6667 0.6684 1.0926 5.3091 0.8927
122 2173 323.15 1.6667 0.6712 1.0971 0.1519 0.0255
123 2177 323.15 1.6667 0.6668 1.0900 0.4863 0.0818
124 2176 372.65 1.6667 0.5490 0.8991 2.2813 0.8780
125 2176 375.15 1.6667 0.5406 0.8805 0.2388 0.0919
126 2179 374.15 1.6667 0.5524 0.9046 0.0782 0.0301
127 2177 423.65 1.6667 0.4878 0.7997 0.4336 0.5339
128 2173 423.65 1.6667 0.5022 0.8231 0.1463 0.1802
129 2177 424.65 1.6667 0.5015 0.8220 0.2322 0.2859
130 2176 323.15 2.0000 0.6226 1.0021 5.3091 0.8559
131 2177 323.15 2.0000 0.6344 1.0208 0.4863 0.0784
132 2174 323.15 2.0000 0.6169 0.9929 0.4078 0.0657
133 2176 374.15 2.0000 0.4937 0.7963 0.8415 0.8988
134 2171 372.65 2.0000 0.4974 0.8022 0.0474 0.0506
135 2171 372.65 2.0000 0.5053 0.8147 0.0474 0.0506
136 2174 423.15 2.0000 0.4447 0.7178 0.4078 0.2922
137 2173 423.65 2.0000 0.4398 0.7099 0.1463 0.1048
138 2176 422.15 2.0000 0.4469 0.7213 0.8415 0.6030
′ ′
Note: 𝑚𝐶𝑂2
means the original measured data, the real CO2 solubility (𝑚𝐶𝑂2 ) obtained in this study equals to 𝑚𝐶𝑂2 = 0.992𝑚𝐶𝑂 2

158
Table E 4 Original Experimental CO2 solubility data record at 1420-2178 psia, 322.15-425.15 K, and from 0.3333-2.0 mol kg-1 Na2SO4 for the CO2-Na2SO4-
H2O system.

INVERSE WEIGHTING
𝒎𝑵𝒂𝟐𝑺𝑶𝟒 𝒎′𝑪𝑶𝟐
Run No #. P / psia T/K 𝒙𝑪𝑶𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝒉𝟐𝒊 , Eq. (C.3) Eq. (C.2)
139 2175 323.15 0.3333 1.0154 1.7652 3.1215 0.3333
140 2175 323.15 0.3333 1.0138 1.7624 3.1215 0.3333
141 2175 323.15 0.3333 1.0287 1.7879 3.1215 0.3333
142 2176 372.15 0.3333 0.8477 1.4779 0.8415 0.2015
143 2174 371.65 0.3333 0.8714 1.5186 0.2127 0.0509
144 2175 373.15 0.3333 0.8553 1.4910 3.1215 0.7475
145 2176 422.65 0.3333 0.8311 1.4494 2.2813 0.7126
146 2177 423.15 0.3333 0.8684 1.5135 0.4863 0.1519
147 2177 423.65 0.3333 0.8195 1.4295 0.4336 0.1354
148 2175 323.15 0.6667 0.8241 1.4127 3.1215 0.3065
149 2176 323.15 0.6667 0.8663 1.4840 5.3091 0.5213
150 2175 323.65 0.6667 0.8361 1.4330 1.7533 0.1722
151 2175 323.15 0.6667 0.8731 1.4955 NA NA Discarded
152 2175 374.65 0.6667 0.6744 1.1591 NA NA Discarded
153 2174 371.65 0.6667 0.6713 1.1538 NA NA Discarded
154 2176 373.15 0.6667 0.6915 1.1881 5.3091 0.4932
155 2176 373.15 0.6667 0.7070 1.2144 5.3091 0.4932
156 2173 373.65 0.6667 0.7086 1.2172 0.1463 0.0136
157 2176 424.15 0.6667 0.6681 1.1484 0.8415 0.2809
158 2176 422.15 0.6667 0.6690 1.1499 0.8415 0.2809
159 2178 422.15 0.6667 0.6961 1.1959 0.1444 0.0482
160 2177 424.15 0.6667 0.7100 1.2195 0.3272 0.1092
161 2176 424.15 0.6667 0.7287 1.2513 0.8415 0.2809
162 2175 322.15 1.0000 0.6852 1.1575 0.7574 0.2194
163 2175 322.65 1.0000 0.7162 1.2093 1.7533 0.5079
164 2170 322.65 1.0000 0.6936 1.1716 0.0320 0.0093
165 2173 323.15 1.0000 0.6859 1.1587 0.1519 0.0440

159
 INVERSE WEIGHTING
𝒎𝑵𝒂𝟐𝑺𝑶𝟒 𝒎′𝑪𝑶𝟐
Run No #. P / psia T/K 𝒙𝑪𝑶𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝒉𝟐𝒊 , Eq. (C.3) Eq. (C.2)
166 2175 324.15 1.0000 0.6867 1.1601 0.7574 0.2194
167 2175 373.65 1.0000 0.5614 0.9503 NA NA Discarded
168 2174 372.15 1.0000 0.5748 0.9728 NA NA Discarded
169 2173 372.65 1.0000 0.5463 0.9251 NA NA Discarded
170 2176 373.65 1.0000 0.6054 1.0241 2.2813 0.2676
171 2175 373.15 1.0000 0.5966 1.0094 3.1215 0.3662
172 2175 373.15 1.0000 0.6173 1.0440 3.1215 0.3662
173 2176 423.65 1.0000 0.5537 0.9374 NA NA Discarded
174 2176 422.65 1.0000 0.5477 0.9273 NA NA Discarded
175 2175 423.15 1.0000 0.5914 1.0007 3.1215 0.8047
176 2175 424.15 1.0000 0.6108 1.0332 0.7574 0.1953
177 2135 323.15 1.3333 0.5283 0.8799 NA NA Discarded
178 2155 323.15 1.3333 0.5563 0.9262 0.0024 0.0007
179 2174 323.15 1.3333 0.5841 0.9720 0.4078 0.1155
180 2175 323.15 1.3333 0.5585 0.9298 3.1215 0.8839
181 2175 374.65 1.3333 0.5241 0.8730 0.3891 0.3394
182 2175 374.15 1.3333 0.5060 0.8431 0.7574 0.6606
183 2173 424.15 1.3333 0.5154 0.8587 0.1319 0.0548
184 2175 424.15 1.3333 0.5338 0.8890 0.7574 0.3151
185 2175 424.15 1.3333 0.5032 0.8385 0.7574 0.3151
186 2175 424.15 1.3333 0.4888 0.8147 0.7574 0.3151
187 2175 323.15 1.6667 0.4661 0.7644 3.1215 0.3662
188 2175 323.15 1.6667 0.4839 0.7934 3.1215 0.3662
189 2176 323.65 1.6667 0.5016 0.8222 2.2813 0.2676
190 2176 373.65 1.6667 0.4767 0.7817 NA NA Discarded
191 2174 373.15 1.6667 0.4441 0.7286 0.4078 0.0370
192 2176 373.15 1.6667 0.4366 0.7164 5.3091 0.4815
193 2176 373.15 1.6667 0.4509 0.7397 5.3091 0.4815

160
 INVERSE WEIGHTING
𝒎𝑵𝒂𝟐𝑺𝑶𝟒 𝒎′𝑪𝑶𝟐
Run No #. P / psia T/K 𝒙𝑪𝑶𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝒉𝟐𝒊 , Eq. (C.3) Eq. (C.2)
194 2175 425.15 1.6667 0.4302 0.7059 0.2315 0.0182
195 2175 423.15 1.6667 0.4346 0.7131 3.1215 0.2454
196 2175 423.15 1.6667 0.4390 0.7203 3.1215 0.2454
197 2176 422.15 1.6667 0.4518 0.7411 0.8415 0.0662
198 2175 423.15 1.6667 0.4331 0.7107 3.1215 0.2454
199 2176 422.65 1.6667 0.4314 0.7079 2.2813 0.1794
200 NA NA 2.0000 0.3643 0.5888 NA NA Discarded
201 2175 372.65 2.0000 0.3742 0.6047 1.7533 0.3597
202 2175 373.15 2.0000 0.3949 0.6379 3.1215 0.6403
203 2175 424.15 2.0000 0.3752 0.6063 0.7574 0.1345
204 2175 423.15 2.0000 0.3713 0.6000 3.1215 0.5542
205 2175 423.65 2.0000 0.3889 0.6283 1.7533 0.3113
206 1687 346.150 1.0000 0.6878 1.1618 - -
207 1687 346.150 1.0000 0.7148 1.2070 - -
208 1420 348.15 1.0000 0.5194 0.8798 - -
209 1420 348.15 1.0000 0.5367 0.9090 - -
210 1420 348.15 1.0000 0.5222 0.8847 - -
211 1419.9 348.15 1.0000 0.5261 0.8912 - -
′
Note: 𝑚𝐶𝑂2
means the original measured data, the real CO2 solubility (mCO2 ) obtained in this study equals to mCO2 = 0.992m′CO2

161
Table E 5 Original Experimental CO2 solubility data record at 2175-2176 psia, 323.15-424.15 K, and from 0.3333-2.0000 mol kg-1 MgCl2 for the CO2-MgCl2-
H2O system.

INVERSE WEIGHTING
T 𝐦𝐌𝐠𝐂𝐥𝟐 𝐦′𝐂𝐎𝟐
Run No #. P / psia 𝐱 𝐂𝐎𝟐 DISTANCE, FACTOR, Note
/K / mol kg-1 / mol kg-1
1/𝐡𝟐𝐢 , Eq. (C.3) Eq. (C.2)
212 2175 323.15 0.3333 1.0709 1.8599 3.1215 0.3333
213 2175 323.15 0.3333 1.0842 1.8825 3.1215 0.3333
214 2175 323.15 0.3333 1.0832 1.8808 3.1215 0.3333
215 2175 374.65 0.3333 0.8919 1.5538 0.3891 0.1816
216 2175 373.65 0.3333 0.8801 1.5336 1.7533 0.8184
217 2175 424.15 0.3333 0.8009 1.3975 0.7574 0.5000
218 2175 424.15 0.3333 0.8180 1.4269 0.7574 0.5000
219 2175 323.15 0.6667 0.9566 1.6362 3.1215 0.2500
220 2175 323.15 0.6667 0.9622 1.6456 3.1215 0.2500
221 2175 323.15 0.6667 0.9690 1.6570 3.1215 0.2500
222 2175 323.15 0.6667 0.9853 1.6844 3.1215 0.2500
223 2175 373.15 0.6667 0.7693 1.3200 3.1215 0.5000
224 2175 373.15 0.6667 0.7770 1.3331 3.1215 0.5000
225 2175 423.15 0.6667 0.7005 1.2034 3.1215 0.6403
226 2175 423.65 0.6667 0.7112 1.2216 1.7533 0.3597
227 2175 323.15 1.0000 0.8267 1.1575 3.1215 0.5000
228 2175 323.15 1.0000 0.8547 1.2093 3.1215 0.5000
229 2176 373.15 1.0000 0.6688 1.1302 5.3091 0.6297
230 2175 373.15 1.0000 0.6709 1.1337 3.1215 0.3703
231 2175 423.65 1.0000 0.6023 1.0189 1.7533 0.3597
232 2175 423.15 1.0000 0.6349 1.0735 3.1215 0.6403
233 2175 323.15 1.3333 0.7442 1.2351 3.1215 0.5000
234 2175 323.15 1.3333 0.7530 1.2495 3.1215 0.5000
235 2175 373.15 1.3333 0.6170 1.0262 3.1215 0.2500
236 2175 373.15 1.3333 0.5970 0.9932 3.1215 0.2500
237 2175 373.15 1.3333 0.5889 0.9799 3.1215 0.2500
238 2175 373.15 1.3333 0.5907 0.9829 3.1215 0.2500

162
 INVERSE WEIGHTING
T 𝐦𝐌𝐠𝐂𝐥𝟐 𝐦′𝐂𝐎𝟐
Run No #. P / psia 𝐱 𝐂𝐎𝟐 DISTANCE, FACTOR, Note
/K / mol kg-1 / mol kg-1
1/𝐡𝟐𝐢 , Eq. (C.3) Eq. (C.2)
239 2175 373.15 1.3333 0.5650 0.9405 NA NA Discarded
240 2175 423.15 1.3333 0.5357 0.8922 3.1215 0.8047
241 2175 422.15 1.3333 0.5605 0.9331 0.7574 0.1953
242 2175 323.15 1.6667 0.6608 1.0803 3.1215 0.5000
243 2175 323.15 1.6667 0.6927 1.1318 3.1215 0.5000
244 2175 373.15 1.6667 0.5365 0.8789 3.1215 0.2500
245 2175 373.15 1.6667 0.5587 0.9149 3.1215 0.2500
246 2175 373.15 1.6667 0.5175 0.8480 3.1215 0.2500
247 2175 373.15 1.6667 0.5357 0.8776 3.1215 0.2500
248 2175 423.15 1.6667 0.4687 0.7686 3.1215 0.7877
249 2176 424.15 1.6667 0.4844 0.7942 0.8415 0.2123
250 2175 323.15 2.0000 0.6215 1.0003 3.1215 0.5000
251 2175 323.15 2.0000 0.6064 0.9762 3.1215 0.5000
252 2175 372.15 2.0000 0.4849 0.7822 0.7574 0.1953
253 2175 373.15 2.0000 0.4878 0.7868 3.1215 0.8047
254 2175 423.15 2.0000 0.4239 0.6063 3.1215 0.5000
255 2175 423.15 2.0000 0.4433 0.6000 3.1215 0.5000
Note: m′CO2 means the original measured data, the real CO2 solubility (mCO2 ) obtained in this study equals to mCO2 = 0.992m′CO2

163
Table E 6 Original Experimental CO2 solubility data record at 2175-2176 psia, 323.15-424.15 K, and from 0.5-4.5 mol kg-1 KCl for the CO2-KCl-H2O system.

INVERSE WEIGHTING
𝐦𝐊𝐂𝐥 𝐦′𝐂𝐎𝟐
Run No #. P / psia T/K 𝐱 𝐂𝐎𝟐 DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝐡𝟐𝐢 , Eq. (C.3) Eq. (C.2)
256 2175 323.15 0.5 1.1832 2.050884 3.121527 1
257 2175 374.65 0.5 0.9371 1.631268 0.389051 0.181601
258 2175 373.65 0.5 0.9505 1.654209 1.753291 0.818399
259 2175 424.15 0.5 0.8882 1.547462 0.757371 1
260 2175 323.15 1 1.0971 1.871989 3.121527 0.5
261 2175 323.15 1 1.1081 1.890404 3.121527 0.5
262 2175 373.15 1 0.8846 1.514892 3.121527 0.5
263 2175 373.15 1 0.8874 1.519614 3.121527 0.5
264 2175 423.15 1 0.8223 1.409706 3.121527 1
265 2175 323.15 2 1.0135 1.674584 3.121527 1
266 2175 373.15 2 0.7878 1.306537 3.121527 1
267 2175 423.15 2 0.7298 1.211512 3.121527 1
268 2175 323.15 3 0.9205 1.474463 3.121527 1
269 2175 373.15 3 0.7463 1.198773 3.121527 1
270 2175 422.15 3 0.6318 1.016723 0.757371 1
271 2175 323.15 3 0.9205 1.474463 3.121527 1
272 2175 373.15 3 0.7463 1.198773 3.121527 1
273 2175 422.15 3 0.6318 1.016723 0.757371 1
274 2175 323.15 4 0.8686 1.349227 3.121527 1
275 2175 373.15 4 0.6791 1.057984 3.121527 1
276 2176 424.15 4 0.5977 0.932352 0.841499 1
277 2175 323.15 4.5 0.84 1.284844 3.121527 1
278 2175 373.15 4.5 0.6368 0.9775 3.121527 1
279 2175 423.15 4.5 0.5477 0.841881 3.121527 0.930968
280 2175 421.15 4.5 0.5414 0.832278 0.231462 0.069032
Note: m′CO2 means the original measured data, the real CO2 solubility (mCO2 ) obtained in this study equals to mCO2 = 0.992m′CO2

164
Table E 7 Original Experimental CO2 solubility data record at 1450-2538 psia, 323.15-424.15 K, and total ionic strength from 1.712 to 4.984 mol kg -1 H2O for
the CO2-synthetic brine system. The chemical composition of the dissolved species (NaCl, Na2SO4, CaCl2, MgCl2, KCl, SrCl2, and NaBr) are listed in Table 4. 2.

INVERSE WEIGHTING
𝐦𝐊𝐂𝐥 𝐦′𝐂𝐎𝟐
Run No #. P / psia T/K DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝐡𝟐𝐢 , Eq. (C.3) Eq. (C.2)
281 2538 422.15 1.712 0.7705 0.9750 0.5000
282 2538 424.15 1.712 0.7951 0.9750 0.5000
283 2538 373.15 1.712 0.8136 39.0625 0.5000
284 2538 373.15 1.712 0.8234 39.0625 0.5000
285 2538 323.15 1.712 0.9587 39.0625 0.5000
286 2538 323.15 1.712 0.9687 39.0625 0.5000
287 2175 423.15 1.712 0.6984 3.1215 0.5778
288 2176 423.65 1.712 0.7089 2.2813 0.4222
289 2175 373.15 1.712 0.7582 3.1215 0.5000
290 2175 373.15 1.712 0.765 3.1215 0.5000
291 2175 323.15 1.712 0.9331 3.1215 0.5000
292 2175 323.15 1.712 0.9433 3.1215 0.5000
293 1813 423.15 1.712 0.6147 1189.0606 0.5000
294 1813 423.15 1.712 0.6271 1189.0606 0.5000
295 1813 373.15 1.712 0.7043 1189.0606 0.5000
296 1813 373.15 1.712 0.6991 1189.0606 0.5000
297 1813 323.15 1.712 0.9015 1189.0606 0.5000
298 1813 323.15 1.712 0.9167 1189.0606 0.5000
299 1450 423.15 1.712 0.5051 7.0359 0.5000
300 1450 423.15 1.712 0.5149 7.0359 0.5000
301 1450 373.15 1.712 0.6019 7.0359 0.5000
302 1450 373.15 1.712 0.5979 7.0359 0.5000
303 1450 323.15 1.712 0.843 7.0359 0.5000
304 1450 323.15 1.712 0.8572 7.0359 0.5000
305 2538 423.15 4.984 0.4832 39.0625 0.5000
306 2538 423.15 4.984 0.4933 39.0625 0.5000
307 2538 373.15 4.984 0.5052 39.0625 0.5000

165
 INVERSE WEIGHTING
𝐦𝐊𝐂𝐥 𝐦′𝐂𝐎𝟐
Run No #. P / psia T/K DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝐡𝟐𝐢 , Eq. (C.3) Eq. (C.2)
308 2538 373.15 4.984 0.5151 39.0625 0.5000
309 2538 323.15 4.984 0.6062 39.0625 0.5000
310 2538 323.15 4.984 0.6132 39.0625 0.5000
311 2175 423.15 4.984 0.449 3.1215 0.5778
312 2176 423.65 4.984 0.4504 2.2813 0.4222
313 2175 373.15 4.984 0.4822 3.1215 0.5000
314 2175 373.15 4.984 0.4923 3.1215 0.5000
315 2175 323.15 4.984 0.5991 3.1215 0.5000
316 2175 323.15 4.984 0.6191 3.1215 0.5000
317 1813 423.15 4.984 0.374 NA NA Discarded
318 1813 423.15 4.984 0.397 1189.0606 1.0000
319 1813 373.15 4.984 0.4365 1189.0606 0.5000
320 1813 373.15 4.984 0.4372 1189.0606 0.5000
321 1813 323.15 4.984 0.5592 NA NA Discarded
322 1813 323.15 4.984 0.5732 1189.0606 1.0000
323 1450 423.15 4.984 0.3203 7.0359 0.5000
324 1450 423.15 4.984 0.3361 7.0359 0.5000
325 1450 373.15 4.984 0.3732 7.0359 0.5000
326 1450 373.15 4.984 0.3857 7.0359 0.5000
327 1450 323.15 4.984 0.5279 7.0359 0.5000
328 1450 323.15 4.984 0.5513 7.0359 0.5000
Note: m′CO2 means the original measured data, the real CO2 solubility (mCO2 ) obtained in this study equals to mCO2 = 0.992m′CO2

166
Table E 8 Original Experimental CO2 solubility data record at 1450-2538 psia, 323.15-424.15 K, and total ionic strength from 1.712 to 4.984 mol kg-1 H2O for
the CO2-synthetic brine system. The chemical composition of the dissolved species (NaCl and CaCl2) are listed in Table 4. 2.

INVERSE WEIGHTING
𝐦𝐊𝐂𝐥 𝐦′𝐂𝐎𝟐
Run No #. P / psia T/K DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝐡𝟐𝐢 , Eq. (C.3) Eq. (C.2)
329 2538 423.15 1.712 0.7705 39.0625 0.5000
330 2538 423.15 1.712 0.8024 39.0625 0.5000
331 2538 373.15 1.712 0.826 39.0625 0.5000
332 2538 373.15 1.712 0.8224 39.0625 0.5000
333 2538 323.15 1.712 0.9431 39.0625 0.5000
334 2538 323.15 1.712 0.9741 39.0625 0.5000
335 2175 423.15 1.712 0.7037 3.1215 0.5778
336 2176 423.65 1.712 0.7165 2.2813 0.4222
337 2175 373.15 1.712 0.7597 3.1215 0.5000
338 2175 373.15 1.712 0.7714 3.1215 0.5000
339 2175 323.15 1.712 0.9212 3.1215 0.5000
340 2175 323.15 1.712 0.943 3.1215 0.5000
341 1813 423.15 1.712 0.6043 1189.0606 0.5000
342 1813 423.15 1.712 0.6209 1189.0606 0.5000
343 1813 373.15 1.712 0.691 1189.0606 0.5000
344 1813 373.15 1.712 0.7018 1189.0606 0.5000
345 1813 323.15 1.712 0.8907 1189.0606 0.5000
346 1813 323.15 1.712 0.9206 1189.0606 0.5000
347 1450 423.15 1.712 0.5068 7.0359 0.5000
348 1450 423.15 1.712 0.5117 7.0359 0.5000
349 1450 373.15 1.712 0.5886 7.0359 0.5000
350 1450 373.15 1.712 0.6004 7.0359 0.5000
351 1450 323.15 1.712 0.8484 7.0359 0.5000
352 1450 323.15 1.712 0.8663 7.0359 0.5000
353 2538 423.15 4.984 0.481 39.0625 0.5000
354 2538 423.15 4.984 0.5039 39.0625 0.5000
355 2538 373.15 4.984 0.5198 39.0625 0.5000

167
 INVERSE WEIGHTING
𝐦𝐊𝐂𝐥 𝐦′𝐂𝐎𝟐
Run No #. P / psia T/K DISTANCE, FACTOR, Note
/ mol kg-1 / mol kg-1
1/𝐡𝟐𝐢 , Eq. (C.3) Eq. (C.2)
356 2538 373.15 4.984 0.5308 39.0625 0.5000
357 2538 323.15 4.984 0.6085 39.0625 0.5000
358 2538 323.15 4.984 0.6181 39.0625 0.5000
359 2175 423.15 4.984 0.4386 3.1215 0.5778
360 2176 423.65 4.984 0.4608 2.2813 0.4222
361 2175 373.15 4.984 0.4852 3.1215 0.5000
362 2175 373.15 4.984 0.5019 3.1215 0.5000
363 2175 323.15 4.984 0.5969 3.1215 0.5000
364 2175 323.15 4.984 0.6121 3.1215 0.5000
365 1813 423.15 4.984 0.3943 1189.0606 0.5000
366 1813 423.15 4.984 0.4064 1189.0606 0.5000
367 1813 373.15 4.984 0.4378 1189.0606 0.5000
368 1813 373.15 4.984 0.4455 1189.0606 0.5000
369 1813 323.15 4.984 0.5769 1189.0606 0.5000
370 1813 323.15 4.984 0.5886 1189.0606 0.5000
371 1450 423.15 4.984 0.3268 7.0359 0.5000
372 1450 423.15 4.984 0.3345 7.0359 0.5000
373 1450 373.15 4.984 0.387 7.0359 0.5000
374 1450 373.15 4.984 0.3914 7.0359 0.5000
375 1450 323.15 4.984 0.5188 7.0359 0.5000
376 1450 323.15 4.984 0.5675 7.0359 0.5000
Note: m′CO2 means the original measured data, the real CO2 solubility (mCO2 ) obtained in this study equals to mCO2 = 0.992m′CO2

168
 VITA

HAINING ZHAO

John and Willie Leone Family Department of

Energy and Mineral Engineering
The Pennsylvania State University, University Park, PA
E-mail: hzz5047@gmail.com; cell:(814) 954-2216
web: www.carbonlab.org

EDUCATION

The Pennsylvania State University, University Park, PA, USA

Ph.D. in Petroleum and Natural Gas Engineering, to be conferred 12/2014
Ph.D. Minor in Electrochemical Engineering, to be conferred 12/2014

Tianjin University, Tianjin, CHINA

M.S. in Mechanical Engineering 7/2010

Tianjin Institute of Urban Construction, Tianjin CHINA

B.S. in Gas Engineering 7/2004

PROFESSIONAL HISTORY

Beijing Fuhua Gas Co., Ltd. Beijing, CHINA

Compressed natural gas production & delivery 2004-2007

RESEARCH INTERESTS

Equation of State for reservoir fluid PVT modeling

CO2 Storage & EOR
Reservoir simulation; Description of flow through porous media
Electrochemical sensor/corrosion
Web-based computational interface

PUBLICATIONS

Haining Zhao, Mark Fedkin, Robert Dilmore, and Serguei N. Lvov (2014), Carbon dioxide
solubility in aqueous solutions of sodium chloride at geological conditions: Experimental results at
323.15, 373.15, 423.15 K and 150 bar and modeling up to 573.15 K and 2000 bar, Geochimica et
Cosmochimica Acta, doi: 10.1016/j.gca.2014.11.004.

Haining Zhao, Robert Dilmore, and Serguei N. Lvov (2014), Experimental studies and modeling
of CO2 solubility in high temperature aqueous CaCl2, MgCl2, Na2SO4, and KCl solutions,
Geochimica et Cosmochimica Acta, in review.

Haining Zhao, Robert Dilmore, Douglas E. Allen, Sheila W. Hedges, Yee Soong, and Serguei N.
Lvov (2014), Measurement and modeling of CO2 solubility in natural and synthetic reservoir
brines, Environmental Science & Technology, in review.