Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The study of CO₂ mass transfer in brine and in brine-saturated Mt. Simon sandstone and the CO₂/brine induced evolution of its transport and mechanical properties
(USC Thesis Other)
The study of CO₂ mass transfer in brine and in brine-saturated Mt. Simon sandstone and the CO₂/brine induced evolution of its transport and mechanical properties
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
The Study of CO
2
Mass Transfer in Brine and in
Brine-Saturated Mt. Simon Sandstone and the
CO
2
/Brine Induced Evolution of its Transport and
Mechanical Properties
by
Zhuofan Shi
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CHEMICAL ENGINEERING)
December 2019
ii
Table of Contents
Dedication ................................................................................................................................... v
Acknowledgments ...................................................................................................................... vi
List of Tables ...........................................................................................................................viii
List of Figures ............................................................................................................................. x
Abstract ..................................................................................................................................... xv
Chapter 1: Background and Motivation .................................................................................. 1
1.1. Global Warming and Carbon Dioxide Emission ................................................... 1
1.2. Carbon Capture and Sequestration ......................................................................... 4
1.3. Knowledge Gaps and Objectives of this Study ...................................................... 7
1.4. Organization of Dissertation .................................................................................. 9
Chapter 2: Measurement and Modeling of CO2 Mass Transfer in Brine at Reservoir
Conditions ................................................................................................................................. 11
2.1. Introduction and Literature Review ..................................................................... 11
2.2. Experimental Materials ........................................................................................ 17
2.3. Measurements of CO2 Solubility in Water/Brine ................................................. 18
2.3.1. Experimental Approach ....................................................................................... 18
iii
2.3.2. Interpretation of Experiments .............................................................................. 20
2.3.3. Results and Observations ..................................................................................... 22
2.4. Measurements of CO2 Mass Transfer in Water/Brine/Porous Medium ............... 27
2.4.1. Experimental Approach ....................................................................................... 27
2.4.2. Interpretation of Experiments .............................................................................. 28
2.4.3. Results and Observations ..................................................................................... 32
2.5. Natural Convection .............................................................................................. 40
2.6. FT-IR Study of CO2 Sorption on Mt. Simon Sandstone ...................................... 47
2.6.1. Experimental Approach ....................................................................................... 48
2.6.2. Results and Discussion ......................................................................................... 51
2.7. Discussion and Conclusions ................................................................................. 64
Chapter 3: CO2/brine Induced Evolution of the Transport Properties of Mt. Simon
Sandstone .................................................................................................................................. 69
3.1. Introduction and Literature Review ..................................................................... 69
3.2. Multiscale characterization of Mt. Simon Sandstone samples before and after
exposure to CO2/brine .......................................................................................... 73
3.2.1. Sample Preparation .............................................................................................. 73
3.2.2. Porosity and Permeability .................................................................................... 76
3.2.3. Ion Chromatography Analysis of Brine Composition ......................................... 81
3.2.4. Pore Size Distribution .......................................................................................... 87
iv
3.2.5. Scanning Electron Microscopy ............................................................................ 95
3.3. Discussion and Conclusions ............................................................................... 101
Chapter 4: CO2/brine Induced Evolution of the Mechanical Properties of Mt. Simon
Sandstone ................................................................................................................................ 105
4.1. Introduction and Literature Review ................................................................... 105
4.2. Experimental Approach ..................................................................................... 108
4.3. Experimental Results ......................................................................................... 112
4.4. Impact of Mechanical Tests on Permeability ..................................................... 122
4.5. Summary ............................................................................................................ 125
Chapter 5: Future Work ....................................................................................................... 127
5.1. CO2 Mass Transfer in brine ................................................................................ 127
5.2. Transport and Mechanical Characterizations of Rock Sample .......................... 128
References ............................................................................................................................... 133
Appendix ................................................................................................................................. 149
v
Dedication
To my loving parents,
Xijian Shi and Xiaoyan Miao
vi
Acknowledgements
Above all, I would like to thank my academic advisors Professor Kristian Jessen and Professor
Theodore Tsotsis for their invaluable guidance and generous support throughout my PhD studies
at the University of Southern California (USC). Without their encouragement, advice and
motivation, it would have been impossible for me to overcome all the challenges along the way.
They served as great mentors throughout my PhD journey. Their spirit will encourage me to
explore, discover and achieve in my next stage of life.
I am very grateful to my dissertation committee member Professor Doug Hammond. His advice
and support are gratefully appreciated. I would also like to thank Professor Iraj Ershaghi and
Professor Katherine Shing who served on my qualifying exam committee.
Additionally, I am thankful to Tina Silva and Shokry Bastorous for making sure that I always
followed the most appropriate safety procedures while working in laboratory. I greatly appreciate
Martin Olekszyk, Nicole Kerns and Monina Letargo for smoothly handling the monthly payment
and research ordering process. I also appreciate the help and advice from Andy Chen and Karen
Woo throughout my graduate studies. I would also like to thank the rest of the staff in MFD for
their kind help on numerous occasions
My sincerest gratitude goes to my former colleagues at USC, Dr. Yu Wang, Dr. Basabdatta
Roychaudhuri, Dr. Devang Dasani, Dr. Majid Monji and Dr. Huanhao Chen who helped me a lot
in my research journey. I also want to thank my current group members Dr. Mingyuan Cao and
Lin Sun for their assistance at various stages of this research. A special thank you to my roommate
Dr. Zhongtang Li for a lot of help and support along the course of my PhD studies.
vii
I am extremely grateful to the Viterbi School of Engineering, Southern California Gas Company
(SoCalGas) and Center for Geological Storage of CO2 (GSCO2) for their financial support at
various stages of my PhD.
Finally and most importantly, I would like to thank my parents, Xijian Shi and Xiaoyan Miao for
their endless love, encouragement and support for every day of my life.
viii
List of Tables:
Table 2.1 CO2 diffusivity from literature ................................................................................. 15
Table 2.2 Chemical composition of the brine .......................................................................... 17
Table 2.3 Properties of the porous media ................................................................................. 18
Table 2.4 The CO2 solubility in water at 50⁰C ......................................................................... 23
Table 2.5 The CO2 solubility in brine at 50⁰C ......................................................................... 23
Table 2.6 The initial and final pressures in the sample cell ..................................................... 32
Table 2.7 Parameters utilized for the sample cell in the 2-D simulations ................................ 43
Table 2.8 Chemical composition of the synthetic brine ........................................................... 49
Table 2.9 CO2 Solubility in brine at 50 ⁰C ............................................................................... 58
Table 2.10 Summary of the CO2 diffusivities measured in different experiments .................. 66
Table 3.1 List of core samples studied in this work including types of tests performed ......... 74
Table 3.2 Composition of the synthetic brine .......................................................................... 75
Table 3.3 Porosity of the specimens ........................................................................................ 79
Table 3.4 Permeability measurements before and after exposure to CO2/brine ...................... 81
Table 3.5 The concentration of the cations and anions in the fresh and aged brines
(6925-A1) ................................................................................................................................. 84
Table 3.6 The concentration of the cations and anions in the fresh and aged brines
(6927-A1) ................................................................................................................................. 85
Table 3.7 The concentration of the cations and anions in the fresh and aged brines
(6924-A1) ................................................................................................................................. 85
Table 3.8 Mineral composition analysis (XRD) of the Mt. Simon core at 2120 m
(6955.6 ft) ............................................................................................................................... ..87
ix
Table 3.9 BET analysis of sample 6927-C1 ............................................................................. 94
Table 3.10 Elemental concentrations (Wt%) for the fresh sample .......................................... 99
Table 3.11 Elemental concentrations (Wt%) for the aged sample ........................................... 99
Table 3.12 Changes (%) in the elemental concentrations between the fresh and the aged
sample .................................................................................................................................... 100
Table 4.1 Results of the dynamic mechanical properties for sample 6927-B2 (fresh) .......... 112
Table 4.2 Results of the dynamic mechanical properties for sample 6927-B4 (aged) .......... 113
Table 4.3 Permeability measurements before and after exposure to CO2/brine .................... 125
Table A-D. 1 Mineral composition from other depths in the Mt. Simon formations ............ 157
x
List of Figures:
Fig. 1.1 Emissions of key greenhouse gases in the United States from 1990 to 2014 ............... 1
Fig. 1.2 Climate influence by different climate drivers between 1750 and 2005 ...................... 3
Fig. 1.3 Schematic view of the CCS chain ................................................................................. 4
Fig. 1.4 Under way and proposed CCS projects ........................................................................ 5
Fig. 1.5 Schematic view of CO2 sequestration in deep saline reservoirs ................................... 6
Fig. 2.1 Schematic diagram of the experimental set-up ........................................................... 19
Fig. 2.2 Correlation of the solubility data for CO2/water at 50°C ............................................ 25
Fig. 2.3 Comparison with data from literature: CO2-water at 50° C ......................................... 26
Fig. 2.4 Correlation of the solubility data for CO2/brine at 50° C ............................................ 27
Fig. 2.5 Schematic of the assumed 1-D diffusion process ....................................................... 29
Fig. 2.6 Normalized pressure decay for CO2 diffusion in bulk water and in bulk brine .......... 33
Fig. 2.7 Normalized pressure decay for CO2 diffusion in a liquid-saturated porous medium
(1.6 mm glass beads) ................................................................................................................ 34
Fig. 2.8 Normalized pressure decay for CO2 diffusion in a liquid-saturated porous medium
(125-150 µ m quartz particles) .................................................................................................. 35
Fig. 2.9 Interpretation of the CO2 diffusivity experiment in bulk water .................................. 36
Fig. 2.10 Interpretation of CO2 diffusivity experiment in the bulk brine ................................ 37
Fig. 2.11 Interpretation of CO2 diffusivity experiment in a water-saturated porous medium
(1.6 mm glass beads) ................................................................................................................ 38
Fig. 2.12 Interpretation of CO2 diffusivity experiment in the brine-saturated porous
medium (1.6 mm glass beads) .................................................................................................. 38
xi
Fig. 2.13 Interpretation of CO2 diffusivity experiment in the water-saturated porous
medium (125-150 µ m quartz particles) .................................................................................... 39
Fig. 2.14 Data interpretation for the measurement of CO2 diffusivity in brine-saturated
porous media made of 125-150 µ m quartz particles ................................................................ 40
Fig. 2.15 Comparison of the amount of CO2 in the vapor phase with numerical simulations
in a water-saturated porous medium (1.6 mm glass beads) ..................................................... 44
Fig. 2.16: Comparison of CO2 dissolution rate with numerical simulations in a
water-saturated porous medium (1.6 mm glass beads) ............................................................ 45
Fig. 2.17 Snapshots of normalized CO2 concentration fields in liquid phase from the
numerical simulations in a water-saturated porous medium (1.6 mm glass beads) ................. 46
Fig. 2.18 SEM photographs of the sample powders ............................................................... 51
Fig. 2.19 Particle size distribution of the sample powders ....................................................... 51
Fig. 2.20 ATR-FTIR spectra of brine saturated Mt. Simon Sandstone ................................... 52
Fig. 2.21 Comparison of spectra of sample before exposure and exposure to CO2 at
8.3MPa … ................................................................................................................................. 53
Fig. 2.22 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone as a
function of increasing CO2 pressure up to 8.3 MPa ................................................................. 54
Fig. 2.23 ATR-FTIR spectra of adsorbed CO2 on dry Mt. Simon Sandstone as a function of
increasing CO2 pressure up to 8.3 MPa: CO2 asymmetric stretch (2342 cm
-1
) ....................... 55
Fig. 2.24 ATR-FTIR spectra of adsorbed CO2 on dry Mt. Simon Sandstone as a function of
increasing CO2 pressure up to 8.3 MPa: CO2 bending mode (630 to 680 cm
−1
) ...................... 56
Fig. 2.25 Absorbed CO2 peak area as a function of time ........................................................ 57
Fig. 2.26 Sorption isotherms of CO2 on brine saturated sample ............................................. 57
Fig. 2.27 CO2 Solubility in brine as a function of pressure ..................................................... 59
Fig. 2.28 Peak area vs. solubility in brine at the same pressure and temperature .................... 60
Fig. 2.29 Schematic of a flat sandstone sample sitting on the ATR crystal ............................. 61
xii
Fig. 2.30 Modeling result and experimental data for pressure of 0.7 MPa .............................. 64
Fig. 2.31 Modeling result and experimental data for pressure of 2.8 MPa .............................. 64
Fig. 2.32 Modeling result and experimental data for pressure of 5.5 MPa .............................. 65
Fig. 3.1 Pictures of 1cm×1cm×1cm sample (left), 1”×2” core sample (middle) and
0.25”×0.5” core sample (right) ................................................................................................. 75
Fig. 3.2 Schematic of incubation experiments ......................................................................... 76
Fig. 3.3 Schematic of the Helium pycnometry porosity measurement set-up ......................... 77
Fig. 3.4 Picture of the TKA-209 gas permeameter .................................................................. 78
Fig. 3.5 The schematic of the permeability measurement set-up ............................................. 78
Fig. 3.6 Photograph of the ICS-2100 IC system ...................................................................... 82
Fig. 3.7 Peaks of cations .......................................................................................................... 83
Fig. 3.8 Peaks of anions ........................................................................................................... 83
Fig. 3.9 Summary of the brine composition change ................................................................ 86
Fig. 3.10 Schematic of the flow perporometry experimental set-up ........................................ 89
Fig. 3.11 Experimental data of the flow perporometry test for sample 6927-A2 before (left)
and after (right) exposure ......................................................................................................... 91
Fig. 3.12 Experimental data of the flow perporometry test for sample 6925-A2 before (left)
and after (right) exposure ......................................................................................................... 91
Fig. 3.13 PSD of sample 6927-A2 before and after aging ....................................................... 93
Fig. 3.14 PSD of sample 6925-A2 before and after aging ....................................................... 93
Fig. 3.15 PSD extracted from the N2 adsorption tests .............................................................. 95
Fig. 3.16 SEM images of sample 6927-C1 before (a and c) and after (b and d) the
flow-through CO2/brine experiments ....................................................................................... 97
xiii
Fig. 3.17 Selected points for EDX analysis. Left image fresh sample, right image aged
sample ...................................................................................................................................... 98
Fig. 3.18 EDX results of point 1 (fresh) and point 26 (aged) .................................................. 99
Fig. 4.1 Photo of NER AutoLab 1500 system ....................................................................... 111
Fig. 4.2 Strategy of measurement of mechanical properties .................................................. 111
Fig. 4.3 Dynamic Young’s Modulus of Core 6927-B2 (fresh) .............................................. 114
Fig. 4.4 Dynamic Young’s Modulus of Core 6927-B4 (aged) .............................................. 115
Fig. 4.5 Dynamic Young’s modulus measured during the 1
st
loading process for the fresh
and aged samples .................................................................................................................... 117
Fig. 4.6 Dynamic Poisson’s ratio of Core 6927-B2 (fresh) ................................................... 118
Fig. 4.7 Dynamic Poisson’s ratio of Core 6927-B4 (aged) .................................................... 118
Fig. 4.8 Dynamic Bulk Modulus of Core 6927-B2 (fresh) .................................................... 120
Fig. 4.9 Dynamic Bulk Modulus of Core 6927-B4 (Aged) ................................................... 120
Fig. 4.10 Dynamic Bulk Modulus measured during the 1
st
loading stage for the fresh and
aged samples .......................................................................................................................... 121
Fig. 4.11 Dynamic Shear Modulus of Core 6927-B2 (fresh) ................................................. 121
Fig. 4.12 Dynamic Shear Modulus of Core 6927-B4 (aged) ................................................. 122
Fig. 4.13 Dynamic Shear Modulus measured during the 1
st
loading stage for the fresh and
aged samples .......................................................................................................................... 122
Fig. 4.14 The schematic of the experimental apparatus for permeability measurement
after the mechanical test. ........................................................................................................ 123
Fig. 4.15 Permeabilities of the 6927-B2 (fresh) and 6927-B4 (aged) cores after the
mechanical tests ..................................................................................................................... 124
Fig. 5.1 CT Image and its binary counterpart scanned with 16.2 µ m resolution ................... 129
Fig. 5.2 CT Image and its binary counterpart scanned with 1.66 µ m resolution ................... 129
xiv
Fig. 5.3 CT Image and its binary counterpart scanned with 3.97 µ m resolution ................... 130
Fig. 5.4 Evolution of the porosity in the axial direction ........................................................ 131
Fig. A-B. 1 Comparison of data (subset shown for clarity) and fitting function ................... 153
Fig. A-C. 1 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone
at pressure of 0.3 MPa ............................................................................................................ 154
Fig. A-C. 2 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone
at pressure of 0.7 MPa ............................................................................................................ 154
Fig. A-C. 3 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone
at pressure of 2.8 MPa ............................................................................................................ 155
Fig. A-C. 4 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone
at pressure of 5.5 MPa ............................................................................................................ 155
Fig. A-C.5 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone
at pressure of 8.3 MPa ............................................................................................................ 156
Fig. A-E.1 N2 adsorption isotherm of sample 6927-C1 before aging .................................... 158
Fig. A-E.2 N2 adsorption isotherm of sample 6927-C1 after aging ....................................... 158
Fig. A-F.1 EDX results for all the points ............................................................................... 161
Fig. A-G.1 Modeling result and experimental data for pressure of 0.7 MPa with εi=0 ......... 162
Fig. A-G.2 Modeling result and experimental data for pressure of 2.8 MPa with εi=0 ......... 163
Fig. A-G.3 Modeling result and experimental data for pressure of 5.5 MPa with εi=0 ......... 163
xv
Abstract
Emissions of greenhouse gases are thought to contribute to global warming. Geological
carbon sequestration (GCS) is currently considered a promising method to mitigate atmospheric
CO2 and, thus, to potentially minimize climate change. In this approach, CO2 is injected into the
subsurface and is trapped there by three main mechanisms, namely physical trapping, dissolution,
and mineral precipitation.
The present work focuses on two important aspects of GCS. First, we study mass transfer
and sorption phenomena in brine, which are the two key processes occurring during CO2
dissolution trapping in GCS. We employ pressure-decay experiments to measure CO2 solubility,
and mass transfer in water/brine systems at elevated pressures of relevance to CO2 storage
operations in saline aquifers together with modeling to delineate and interpret the experimental
data. Accurate measurements and modeling of mass transfer in this context are crucial to an
improved understanding of the long-term fate of CO2 that is injected into the subsurface for storage
purposes. Pressure-decay experimental data are presented for CO2/water and CO2/brine systems
with and without an unconsolidated porous medium being present. Companion high-resolution 2-
D numerical simulations demonstrate that natural convection will complicate the interpretation of
the experimental observations, if the size of the particles constituting the unconsolidated porous
medium is not sufficiently small. In such settings, we demonstrate that simple 1-D interpretations
based on diffusional transport alone can result in an overestimation of the uptake (diffusivity) by
two orders of magnitude. The high-resolution 2-D numerical calculations, on the other hand, agree
well with the experimental observations for conditions where natural convection contributes
substantially to the overall mass transfer process. The ATR-FTIR technique was also applied to
xvi
study the CO2 mass transfer and sorption phenomena in the Mt. Simon Sandstone. The
experimental observations are able to capture the CO2 diffusion in the brine saturated sample, and
are consistent with the aforementioned separate measurements of CO2 solubility in bulk brine and
in brine-saturated unconsolidated porous media.
We also study, in addition, rock-fluid interactions and their impact on the transport and
mechanical properties of the host rock, which are phenomena relevant to CO2 mineral trapping
during GCS. Specifically, the present study investigates the change in the flow-through
characteristics, porosity, and the mechanical behavior of Mt. Simon Sandstone samples caused by
exposure to brine/CO2. The cores being investigated were extracted from a certain depth interval
(2110.4 - 2111.4 m or 6924 - 6927 ft) in the Mt. Simon formation. Their mechanical and transport
properties were first characterized, and the cores were then aged in CO2-saturated brine at a
pressure of 17.24 MPa (2500 psi) and a temperature of 50⁰C for periods ranging from one to two
weeks. Following that, the change in transport and mechanical properties of the samples were
analyzed using He pycnometry, flow perporometry and triaxial testing. Our experiments show that
the porosity of the Mt. Simon samples slightly increases after exposure to CO2/brine, while the
permeability increases more substantially (depending on the confining pressure environment).
Measurements of the flow-through pore size distribution (PSD) are indicative of significant
changes occurring, consistent with the observed increases in permeability. Nitrogen adsorption
tests (BET), before and after incubation, show a significant loss of pore volume in the mesopore
range that is indicative of clay dissolution. Weakening of the materials was observed based on the
mechanical properties studied, a result that is consistent with the observed dissolution of clays that
play a central role in the cementation of the quartz grains. Finally, the analysis of the brine
xvii
compositions employed in the aging experiments reveals an increase in the concentration of most
cations after incubation with the Mt. Simon cores. This is also consistent with mineral/clay
dissolution, confirmed by the porosity, transport, and mechanical property measurements as well
as electron microscopy analysis of the same samples.
1
Chapter 1: Background and Motivation
1.1 Global Warming and Carbon Dioxide Emissions
It has become widely accepted in recent years that emissions of greenhouse gases contribute to the
observed global warming (Baird and Caan, 2005, Karl and Trenberth, 2003). The key greenhouse
gases emitted to the atmosphere, as a result of human activities, are carbon dioxide (CO2), methane
(CH4), nitrous oxide (N2O) and various fluorinated gases (F-gases)
(https://www.epa.gov/ghgemissions/global-greenhouse-gas-emissions-data). Fig. 1.1 shows the
emissions of all such gases in the United States from 1990 to 2014, expressed in million metric
tons of CO2 equivalents for consistency purposes.
Fig. 1.1 Emissions of key greenhouse gases in the United States from 1990 to 2014
(www.epa.gov/climate-indicators - Updated August 2016, 3)
2
According to Fig. 1.1, CO2 represents more than 70% of all the emitted greenhouse gases,
thus making it the primary greenhouse gas emitted due to human activities in the US. The
Intergovernmental Panel on Climate Change (IPCC) issued a global climate report in 2007 that
compared the relative influence of different climate drivers on the global climate between 1750
and 2005 (Forster et al., 2007). In this report, the IPCC calculated the radiative forcing (RF) of
each climate driver, which reflects its contribution to the net increase or decrease in the amount of
energy that is trapped on the Earth’s surface. Positive RF values represent a warming effect and
negative values represent a cooling effect. CO2 has by far the highest positive RF value (see Fig.
1.2) of all the human-induced climate drivers. Based on such data, one can reasonably conclude
today that CO2, among all climate change drivers, has contributed the most to global climate
change between 1750 and 2005.
3
Fig. 1.2 Climate influence by different climate drivers between 1750 and 2005 (Forster et al.,
2007)
Another reason why one should pay attention to emissions of CO2 more than those of any
other greenhouse gas emitted by human activities, is because CO2 remains in the atmosphere for a
much longer time than any of the other major greenhouse gases. It takes about 10 years for methane
and 100 years for nitrous oxide emissions to dissipate in the atmosphere (Forster et al., 2007). For
CO2, though about 80% of the emissions dissipate in 100 years, 20% of it will remain in the
atmosphere for nearly 800 years (Forster et al., 2007). This, then, means that the CO2 molecules
emitted today, from power plants and other anthropogenic sources, are going to continue to warm
4
the earth long into the future. Accordingly, it is clear that reducing the CO 2 emissions is a critical
issue that must be addressed in order to mitigate further global warming.
1.2 Carbon Capture and Sequestration (CCS)
Carbon capture and sequestration (CCS) is currently the most promising solution to mitigate
atmospheric CO2 emissions and to minimize climate change (Pires et al., 2011). Fig. 1.3 shows a
schematic view of the CCS chain.
Fig. 1.3 Schematic view of the CCS chain (http://www.co2crc.com.au/)
CCS comprises of three stages: (1) separation and capture, (2) transportation, and (3)
injection and storage. The separation and capture stage involve the separation of CO2 from other
5
gases, for example, from the flue-gas generated in power plants. The stage of transportation
includes moving large volumes of CO2 from its collection points to a site where it can be injected
into and stored underground. The final stage involves the injection of the CO2 into deep
underground storage reservoirs, and the ongoing management and monitoring of the storage site.
Fig. 1.4 shows a number of international CCS projects that are either ongoing currently or are
proposed in the future.
Fig. 1.4 Under way and proposed CCS projects (http://status.globalccsinstitute.com/?v=2016)
In particular, sequestration of CO2 in deep saline formations offers the potential of storing
billion of tons of injected CO2. In this approach, CO2 is injected into the subsurface and is trapped
by three main mechanisms: (1) physical trapping, (2) dissolution, and (3) mineral precipitation.
When CO2 is injected into a subsurface formation, it will migrate upwards due to buoyancy forces
until it is trapped (confined) in a region that is capped by an impermeable layer of rock/shale. This
is referred to as physical trapping, and is the least secure trapping mechanism because CO2 is in a
6
gaseous state and may, potentially, leak back to the surface and into the atmosphere. When CO2 is
trapped under an impermeable barrier, it remains in contact with the resident brine and will
dissolve gradually into the liquid phase at rates that are controlled by molecular diffusion.
Dissolution trapping is more permanent than physical trapping because the CO2 is held in solution
within the liquid brine. As the pH of the liquid phase decreases due to CO 2 dissolution,
geochemical reactions may also occur between the CO2/brine solution and the formation
rock/minerals, leading to dissolution/precipitation of salts and minerals.
Fig. 1.5 Schematic view of CO2 sequestration in deep saline reservoirs (https://ukccsrc.ac.uk/)
7
1.3 Present Knowledge Gaps and Objectives of this Study
During the dissolution trapping stage, the migration of the CO2 plume and the CO2 distribution in
the liquid phase, long after the injection stage ends, must be understood in order to accurately
assess the CO2 storage capacity and the safety of a certain storage operation. Although the
geological sequestration of CO2 is and has been studied intensely, there is still a lack of knowledge
regarding CO2 mass transfer in brine systems at reservoir conditions, especially in brine-saturated
porous media.
During the mineral trapping stage, fluid-mineral interactions may convert CO2 into
carbonate minerals, a process thought to be beneficial for the long-term storage. Such interactions
may, however, also impact a reservoir’s transport and mechanical properties via the destabilization
of clays, the corrosion of mineral surfaces and/or the precipitation of carbonates and other salts in
the porous structures. Therefore, a fundamental understanding of rock-fluid interactions and their
effect on the transport and mechanical behavior of the host rocks is necessary to ensure the integrity
of the sequestration system. Some of the knowledge gaps that need to be addressed include: (1)
the lack of accurate and reliable high-pressure thermodynamic equilibrium data of the CO2/brine
system at reservoir conditions; (2) an incomplete understanding of mass transfer of CO2 in brine
and brine-saturated porous media; (3) the current lack of modeling work to simulate the mass
transfer processes in CO2/brine systems; (4) lack of understanding of the impact of CO2-brine-rock
interactions on the transport and mechanical properties of sandstone rocks. In order to address
these knowledge gaps, we have conducted the following systematic studies:
8
Static experiments were carried-out in a typical PVT system to measure the solubility of
CO2 in de-ionized (DI) water and brine at relevant formation conditions. The experimental
phase equilibrium data obtained are important to describe the conditions prevailing at the
liquid-vapor interface during the modeling of transport phenomena in CO2/brine systems.
Dynamic experiments were conducted to measure the mass transfer of CO2 in brine and in
brine-saturated porous media. The experimental data generated provide valuable
information regarding the dynamics of the diffusion-sorption processes taking place in the
liquid phase. A mathematical model was developed that describes the diffusion-sorption
processes, and it was used to extract, via parameter fitting, the mass transfer properties.
Through the experiments and corresponding modeling, one can hence obtain the diffusivity
of CO2 in bulk brine and in brine-saturated porous media.
The transport and mechanical properties of Mt. Simon Sandstone samples were also
characterized before and after exposing them to CO2 and brine at relevant formation
conditions. The results generated here, provide evidence of CO2-brine-rock interactions
and will be of value in predictive simulations of geological carbon sequestration. Predictive
simulations can, in turn, be used in guiding actual storage design and operations.
In summary, the main objective of this work is to provide CO2 solubility and diffusivity
information that are currently not available in the literature, and to improve the fundamental
understanding of mass transfer and sorption phenomena in CO2/brine systems during the
dissolution trapping stage. Further, this work aims to provide information on CO2-brine-rock
interactions and to investigate their impact on the transport and mechanical properties of host rocks
during the mineral trapping stage. Based on the systematic studies of the CO2/brine/rock systems,
as outlined above, we aim to obtain a deeper understanding of mass transfer of CO2 in both bulk
9
brine and brine-saturated porous media, and of the change of transport and mechanical properties
of Mt. Simon Sandstone caused by rock-fluid interactions. Sorption and diffusion characteristics
will be also extracted from the experimental observations via “fit-for-purpose” modeling and
parameter estimation. The impact of fluid-rock interactions will be demonstrated from the
characterizations of sandstone samples before and after exposing to CO2 and brine. These results
will, in turn, provide a better understanding of the CO2 and host rock behavior after CO2 injection
into deep saline reservoirs.
1.4 Organization of the Dissertation
In Chapter 2, the CO2 diffusivity in water/brine and in water/brine-saturated porous media is
measured via pressure-decay observations in a PVT cell. The experimental observations are
discussed. Simple and more advanced mathematical models are developed to interpret the
experiments. Also, the CO2 mass transfer in the Mt. Simon Sandstone was also studied using ATR-
FTIR technique. A simple 1-D model was developed to interpret the experimental results.
In Chapter 3, the transport properties of Mt Simon Sandstone samples are characterized
before and after exposure to CO2 and brine at realistic formation conditions. A series of techniques
are applied to measure the porosity, permeability, pore structure, and also the brine composition
before and after the aging experiments. The results discussed in this chapter provide a consistent
description of the evolution of the transport behavior of Mt. Simon Sandstone due to CO2/brine
induced alteration.
In Chapter 4, the mechanical properties of the fresh and aged Mt. Simon Sandstone samples
are measured via tri-axial mechanical tests. The dynamic values of Young’s Modulus, Poisson
10
Ratio and Bulk Modulus, and observed changes therein are presented and discussed. The results
show a clear change in the mechanical behavior of the Mt. Simon Sandstone due to fluid rock
interactions.
Finally, Chapter 5 provides a summary of the presented work and outline a set of
recommendations for future work in this area.
11
Chapter 2: Measurement and Modeling of CO2 Mass Transfer in
Brine at Reservoir Conditions
1
2.1 Introduction and Literature Review
During dissolution trapping, when CO2 is trapped under an impermeable barrier, it remains in
contact with the resident brine and will dissolve gradually into the liquid phase at rates that are
controlled by molecular diffusion. To be able to accurately assess the CO2 storage capacity and
the safety of a given storage operation, the CO2 distribution in the liquid phase, long after injection
ends, must also be taken into consideration. Consequently, careful measurement of the CO 2 mass
transfer in water/brine at relevant subsurface conditions is critical in order to delineate the long-
term behavior of CO2 in the subsurface.
A good number of studies on CO2 solubility in single-salt and mixed-salt aqueous
solutions have been conducted to date. The CO2 solubility in single salt NaCl and CaCl2 solutions
has been studied for a wide range of pressures, temperatures and salt concentrations (Malinin and
Kurorskaya, 1975; Drummond, 1981; Rumpf et al., 1994; Kiepe et al., 2002; Koschel et al., 2006;
Gilbert et al., 2016; Tong et al., 2013 and Bastamin et al., 2014). Some data also exist for CO2
solubility in aqueous solutions of MgCl2 (Tong et al., 2013 and Zhao et al., 2015a), and Na2SO4
(Corti et al., 1990; Rumpf and Maurer, 1993; Bermejo et al., 2005; Zhao et al., 2015a and Gilbert
et al., 2016). CO2 solubility data in mixed-salt aqueous solutions are not so plentiful, however. Li
et al. (2004) measured densities and solubilities for the system of CO2 +Weyburn formation brine
at a temperature of 59 ° C and pressures up to 29 MPa. Liu et al. (2011) studied the CO2 solubility
1
Sections 2.1 -2.5 were done collaboratively with Dr. Wen of UT Austin and the research has already been
published (Shi et al., 2018)
12
in aqueous solutions of NaCl + CaCl2 (weight ratio = 1:1) at 45 ⁰C and a total salt concentration
of 10 wt%. Wang et al. (2014) investigated CO2 solubility in brine samples collected from five
formations at temperatures from 45 ⁰C to 75 ⁰C and and pressures from 8 to 11 MPa. Zhao et al.
(2015b) measured CO2 solubility in synthetic and natural Mt. Simon formation brine samples at
pressures from 10 to 20 MPa and temperatures from 50 to 100 ⁰C. Several thermodynamic models
for calculating CO2 solubility in mixed-salt aqueous solution have been developed and
demonstrated to provide accurate predictions. Duan et al. (2006) presented a model (DS2006) for
the calculation of CO2 solubility in mixed-salt solutions containing Na
+
, K
+
, Ca
2+
, Mg
2+
, Cl
-
and
SO4
2-
for temperatures from 0 ⁰C to 260 ⁰C, pressures from 0 to 200 MPa, and salt concentrations
of 0 to 4.5 mol/kg with 6.4% AAD (as reported by Zhao et al., 2015b). Spycher and Pruess (2010)
established a phase-partitioning model (SP2010) for CO2-NaCl brine mixtures at elevated
temperatures and pressures (12–300 ⁰C and 1–60 MPa) with 6.3% AAD (as reported by Zhao et
al., 2015b). Wang et al. (2014) developed a semi-empirical thermodynamic model to calculate CO2
solubility in aqueous solution containing Na
+
, K
+
, Ca
2+
, Mg
2+
, covering a temperature range range
and pressure range from 40 to 100 ⁰C and from 5 to 22 MPa, respectively, with 2% AAD. Zhao et
al. (2015b) proposed a model (PSUCO2) to predict CO2 solubility in natural and synthetic
formation brines in the P-T space of 0.1 to 20 MPa, 15 to 150 ⁰C, and salinity up to 5 mol/kg (with
4.3% AAD). Shi and Mao (2017) presented an activity-fugacity phase equilibrium model to
calculate CO2 solubility in brine solutions containing Na
+
, K
+
, Ca
2+
, Mg
2+
, Cl
-
and SO4
2-
, covering
a large P-T range (0- 170 ⁰C, 0.1- 50 MPa), and salinity up to 5 mol /kg, with 3% ADD.
Several experimental techniques have been proposed to measure gas diffusion in liquids
(Himmelblau 1964, Policarpo and Ribeiro 2011). These can be classified as conventional or
unconventional methods. Conventional methods are further classified as either direct or indirect
13
methods. In the conventional direct methods, one needs to analyze the composition of the liquid
phase by sampling at different times during an experiment. The conventional direct methods
include the Taylor-Aris dispersion method proposed by Frank et al. (1996), the capillary cell
method by Witherspoon and Saraf (1965), the diaphragm cell method by Gubbins et al. (1966),
and gas adsorption in a laminar jet as proposed by Tang and Himmelblau (1965). The main
limitation of these direct methods relates to the difficulty of intrusive sampling at high pressures
and temperatures. In the conventional indirect methods, one needs to correlate the measured
system properties to the composition of the liquid phase (Policarpo and Ribeiro, 2011). One of the
unconventional indirect methods is the so-called pressure decay method. In this method, the CO2
diffusivity is determined from experiments in a pressure-volume-temperature (PVT) cell by
monitoring and interpreting the observed pressure decay (Riazi, 1996) as CO2 dissolves and
diffuses into a given liquid. The primary advantage of the PVT-cell method is that intrusive
sampling is not required. However, in this approach if the mass transfer between the CO2-rich gas
phase and the water/brine phase is substantially slow then small gas leaks in the experimental set-
up may impact the interpretation of the results and contribute to the experimental error, particularly
when experiments are performed at elevated pressures (subsurface conditions), which exacerbate
such leaks. In addition, as the CO2-saturated brine is denser than fresh brine (Yan et al. 2011), the
density difference can induce density-driven convection and accelerate the mass transfer rate of
CO2 into the brine (Weir et al. 1996, Riaz et al. 2006, Yang and Gu 2006, Farajzadeh et al. 2009,
Wang et al. 2013). The onset of density-driven convection will further complicate the
interpretation of mass transfer experiments and can lead, if ignored, to an over-estimation of CO2
diffusivity, as further demonstrated and discussed in this work.
14
It is well-known that the diffusivity of CO2 in formation brine is highly dependent on the
prevailing temperature and pressure conditions. However, most of the reported measurements of
CO2 diffusivity in water/brine were conducted at atmospheric conditions. For example, Tang and
Himmelblau (1965) studied the diffusivity of CO2 in water at 25⁰C using a large liquid-jet method.
Tamimi et al. (1994) studied the diffusivity of CO2 in a temperature range of 20-95⁰C using a
wetted-sphere absorption apparatus. Frank et al. (1996) measured the diffusivity of CO2 in a
temperature range of 20-95⁰C using the Taylor-Aris dispersion method.
Only a few studies have been conducted at higher pressures of relevance to subsurface
storage of CO2. For example, Hirai et al. (1997) report the diffusivity of CO2 in water at 13⁰C and
at 9400 KPa2 and 39200 KPa using laser-induced fluorescence. Sell et al. (2013) measured the
CO2 diffusivity over a pressure range of 500-5000 KPa at 26⁰C using a microfluidic approach.
Raad et al. (2015) measured the diffusivity of CO2 at temperature range of 30-40⁰C and pressure
range of 5880-6265 KPa using a PVT cell.
The experimental work, reported to date, on the measurement of diffusivity of CO2 in saline
systems at reservoir conditions, is limited as well. Wang et al. (1996) measured the CO2 diffusivity
in 0.25 N NaCl solution over a pressure range of 1524-5178 KPa at 38⁰C, while Azin et al. (2013)
reported the CO2 diffusivity in brine samples from an oil field for a temperature range of 32-50⁰C
and a pressure range of 5900-6900 KPa. Zarghami et al. (2017) studied the effect of salinity on
CO2 diffusivity at a temperature of 68⁰C and a pressure of 17450 KPa.
A summary of CO2 diffusivity, as reported in the literature for water and assorted brines,
is provided in Table 2.1.
15
Table 2.1 CO2 diffusivity from literature
Source
Solutio
n
Temperature
(
o
C)
Pressure
(KPa)
Diffusivity (10
5
cm
2
/s)
(Azin et al. 2013) Brine 32-50 5900-6900 3.52-6.16*
(Raad et al. 2015) Brine 30-40 5880-6265 0.6-23*
(Renner 1988) Brine 38 1544-5833 3.07-6.86
(Yang and Gu 2006) Brine 27-58 2600-7500 170.7-269.8*
(Wang et al. 1996) Brine 38 1524-5178 2.925-4.827*
(Zarghami et al. 2017) Brine 50-75 17450 6.5-8.2*
(Zhang et al. 2015) Brine 25 1170 1.5-1.91*
(Belgodere et al. 2015) Water 21 4000 1.71
(Cadogan et al. 2014) Water 25-150 15000-45000 2.233-12.21
(Farajzadeh et al. 2007) Water 25 800-1200 2.75*
(Farajzadeh et al. 2009) Water 30 1000-5000 140-245*
(Frank et al. 1996) Water 25-55 101.325 1.97-3.67
(Hirai et al. 1997) Water 13 29400-39200 1-1.5
(Lu et al. 2013) Water -5-200 20000 0.7-16
(Mazarei and Sandall 1980) Water 25 16 1.88
(Sell et al. 2013) Water 26 500-5000 1.86
(Tamimi et al. 1994) Water 20-95 101.325 1.76-8.2
(Tse and Sandall 1979) Water 25 101.325 2.11*
(Unver and Himmelblau
1964)
Water 6-65 101.325 1.145-4.296
(Wang et al. 2013)
Water 45 3430-8020 233.6-251.34*
* Experimental data obtained from interpretation of pressure decay in a PVT cell.
In Table 2.1, the diffusion coefficients that are derived from interpretation of pressure
decay observations in PVT cells are labeled with an asterisk. From Table 2.1, we observe that the
reported values of the diffusion coefficient span two orders of magnitude for both water and brine.
For example, Yang and Gu (2006), Farajzadeh et al. (2009) and Wang et al. (2013) report CO2
16
diffusion coefficients in water and brine on the order of 10
-3
cm
2
/s, while most other experimental
observations are in the range of 10
-5
cm
2
/s. Azin et al. (2013), Farajzadeh et al. (2007) and
Zarghami et al. (2017) report that the mass transfer rate during the early stage of an uptake
experiment is one to two orders of magnitude greater than that during later times. They report the
values derived from later times as the CO2 diffusivity in the bulk liquid: Elevated values of the
diffusion coefficient suggest that density-driven convection or mechanical mixing may have
occurred that was not represented explicitly in the interpretation of the experiments.
In this Chapter, the CO2 diffusivity in water/brine and in water/brine-saturated porous
media has been measured at 50⁰C in the pressure range of 4245-5785 KPa via pressure decay
observations in a PVT cell. Our study has two key goals: (i) The experiments and their
interpretation provide valuable information on the solubility and the diffusion characteristics of
CO2 in brine of relevance to the Mt. Simon Formation, towards a better understanding of the CO2
behavior at later times in deep saline formations in the context of CCS processes. (ii) By
performing experiments with and without a porous medium being present, we investigate the role
that density-driven convection may play during diffusivity measurements in a PVT cell in order to
arrive at a set of guidelines for future experimental efforts.
We start by presenting our experimental procedure and interpretation via a relatively
simple mathematical model that considers diffusive mass transfer only. High-resolution numerical
calculations of CO2 transport in the water/brine are then reported in order to investigate the role of
density-driven convection with emphasis on the impact of a porous material on the uptake
dynamics. We conclude the manuscript with a discussion of the experimental observations and
their interpretation based on observations from simple and advanced modeling of the uptake
process.
17
2.2 Experimental Materials
The CO2 used in this study was Coleman-Grade with purity of 99.99%, while the brine
composition is based on information collected from the Illinois Basin - Decatur Project (IBDP)
Site in Decatur, Illinois: The composition of brine is reported in Table 2.2.
Table 2.2 Chemical composition of the brine
Salt g/L
NaCl 114.160
CaCl2 · 2H2O 73.367
MgCl2 · 6 H2O 16.729
KCl 40.296
KBr 1.117
LiCl 0.122
SrCl2 · 6H2O 2.434
Na2B4O7 0.766
In addition to CO2 diffusion in bulk water/brine, experiments were also performed for two
water/brine saturated unconsolidated porous media. The unconsolidated porous media were
created by packing dense solid particles with different particle-sizes: One porous medium was
generated from packing glass beads with a particle size of 1.6 mm, and a second porous medium
was prepared by packing quartz particles with particle sizes is the range of 125-150 µ m (National
Scientific Company). The properties of the porous media are given in Table 2.3.
18
Table 2.3 Properties of the porous media
Material Particle size Porosity (%)
Soda lime glass 1.6 mm 40
Quartz 125-150 µ m 45
The porosities of the porous media, reported in Table 2.3, were experimentally measured
via liquid infiltration and an accompanying material/volume balance (dimensions of packing,
mass/density of fluid).
2.3 Measurements of CO2 Solubility in Water/Brine
2.3.1 Experimental Approach
To interpret the diffusion experiments, the solubility of CO2 in water/brine must be known as a
function of pressure at the temperature of the experiments (performed at here 50⁰C). Hence, in an
initial set of experiments, the solubility of CO2 was determined in water and brine at 50⁰C for a
pressure range of 728-6074 KPa.
The solubility experiments were performed in a typical PVT system. The main equipment
(see Fig. 2.1) consists of a reference cell and a sample cell. The capacity of the reference cell and
the sample cell used for the solubility measurements was 14.7 cm
3
and 13.6 cm
3
, respectively.
Both reference and sample cells were placed in an air-bath to ensure that all experiments were
performed at a constant temperature of 50° C. High accuracy (0.03% full scale, Omega) pressure
19
Air bath
transducers, with a pressure range of 34,474 kPa, were used to monitor the pressure in each cell.
Thermocouples (Omega) were used to monitor the temperature in each cell. A data acquisition
system was used to record the pressure and temperature data throughout the experiments.
CO2 tank He tank
Vacuum
Sample cell
V-1
TC2
TC1
PG1
PG2
Reference cell
Fig. 2.1 Schematic diagram of the experimental set-up
To initiate the experiment, a known quantity of water (or brine) was loaded into the sample
cell. The temperature of the air-bath was then set to 50° C. To evaluate the gas leak-rate, the system
was evacuated at 933.2 Pa and then pressurized by helium. The pressure was then monitored for
24 h. Based on the observed He leak-rate, the leak-rate of CO2 is subsequently estimated via the
Dusty-gas model (see Appendix A). The reference and sample cells were then placed briefly under
moderate vacuum at 933.2 Pa to remove all air and residual helium from the cells. The reference
cell was then isolated from the sample cell and pressurized with CO2 to a certain pressure and
allowed to reach thermal equilibrium, with the pressure and temperature of the reference cell being
used to calculate the moles of CO2 initially present in the system. The CO2 in the reference cell
was then expanded into the sample cell (containing water or brine) and the pressures and
temperatures of the two cells were monitored/recorded until equilibrium was reached. Following
this initial step (with CO2 in the reference cell being expanded into the evacuated sample cell), the
reference cell was then again isolated from the sample cell and charged with additional CO 2 to
20
increase its pressure above the prevailing sample cell pressure. The CO2 in the reference cell was
then once more expanded into the sample cell with the pressures and temperatures of the two cells
being recorded until equilibrium was reached. This equilibration procedure was then repeated
several times to arrive at a relevant set of equilibrium pressure observations (at each stage, the
experiment was terminated once the system pressure was stabilized, i.e., when the pressure change
in the sample cell falls below 689.5 Pa/h, or when the pressure variation in the sample cell became
equal to the estimated leak-rate). The solubility of CO2 was then subsequently evaluated at each
pressure step from a material balance on the vapor phase using CO2 compressibility factors from
the NIST webbook (see Section 2.3.2 for additional details). It should be noted, that each pressure
step in the solubility measurements, effectively, corresponds to a diffusion experiment, which is
allowed to reach equilibrium.
2.3.2 Interpretation of Experiments
The purpose of the solubility experiments performed in this work, is solely to facilitate the
interpretation of the diffusivity experiments. Numerous authors have previously measured and
reported the solubility of CO2 in water (see, e.g., Carroll and Mather, 1992, Yan et al., 2011 – and
references therein) and in model brines (see, e.g., Yan et al., 2011 for NaCl-containing brines -
and references therein). However, the solubility of CO2 in any natural brine depends on the specific
salts in solution and on their concentrations. Accordingly, the CO2-brine solubility measurements
presented here are required to interpret our mass transfer experiments (since no prior data are
available with this particular brine), while the CO2-water solubility measurements serve as a
quality check for the accuracy of the experimental approach. For each stage in the solubility
measurements, the moles of CO2 in solution were calculated from the experimental observations
of pressure and temperature in the sample and reference cells along with the known quantity of
21
water (or brine) placed in the sample cell. The molar density of CO 2, as reported from the NIST
webbook, was used in the material balance. An assumption that is made in our evaluation of CO 2
solubility (justified by the small solubilities measured) is that the volume change, if any, of the
liquid phase, due to dissolution of CO2, is negligible and can thus be ignored. Based on this
assumption, the material balance for any given pressure stage is described by:
,1 ,2 ,1 ,1 ,2 ,2
( ) ( ) ( )
r r r s l s s l s
n V V V V V (2.1)
where n is the number of moles dissolved during that stage,
r
V ,
s
V and
l
V represent the
reference cell, sample cell and the liquid volumes, while
r
and
s
represent the vapor phase
molar density of CO2 in the reference and sample cells. Subscripts 1 and 2 refer to the initial
pressure and final pressure of any given stage of the experiment. We note, that the change in
density of water due to pressure variation is very modest over the relevant pressure range (0.5%
increase in water density between 10
2
kPa and 10
5
kPa at 50° C, based on data from the NIST
webbook) and can be ignored. Based on Eq. (2.1), one can calculate the number of moles of CO2
dissolved in the water (or brine) in the sample cell at the conclusion of each stage, and thus finally
arrive at the solubility of CO2 in water/brine corresponding to the observed final pressure in the
sample cell.
To interpret the solubility experiments, we start at the general requirement for
thermodynamic equilibrium in vapor-liquid systems (written in terms of the fugacity,
i
f , of
component i):
VL
ii
ff , 1,...,
c
in (2.2)
22
With our specific emphasis on the solubility of CO2 in an aqueous phase, at moderate
temperature and pressures, we apply the two-fluid (phi-gamma) formulation (e.g., Carroll and
Mather, 1992), to express the fugacity of CO2 in the vapor and liquid phases as follows:
1 1 1
V
f y p (2.3)
1 1 1 1
L
f x H (2.4)
0
2
0*
1 1 1
ln( ) ln( )
p
p
H H V dp
(2.5)
0
2
0*
1 1 1 1
ln( ) ln( ) ( )
p
p
V V dp
(2.6)
Here,
1
y and
1
x are the molar fractions of CO2 in the vapor and liquid phases, respectively,
p is the overall pressure,
1
is the fugacity coefficient of CO2 in the vapor phase,
1
H is the
Henry's constant at pressure p,
0
1
H is Henry's constant at the reference state (
0
1
p ),
1
is the
(asymmetric) activity coefficient of CO2 in the liquid phase,
0
1
is the activity coefficient of CO2
at the reference state.
1
V is the reduced partial molar volume (i.e., V/RT) of CO2,
1
*
V is the reduced
partial molar volume at infinite dilution, and
0
1
p is the vapor pressure of water at temperature T
(12.34 kPa at 50° C – NIST Webbook). The application of Eqs. (2.3)–(2.6) for data interpretation
is discussed in further detail in Section 2.3.3.
2.3.3 Results and Observations
The solubility of CO2 in water and brine was measured at 50° C for pressures up to 6074 kPa,
according to the procedure outlined in Section 2.3.1, and the data are reported as a function of
pressure in Tables 2.4 and 2.5.
23
Table 2.4 The CO2 solubility in water at 50⁰C
Pressure (KPa) xCO2 (%)
728 0.252
1384 0.486
2056 0.716
2872 0.970
3640 1.192
4573 1.419
5369 1.582
6074 1.704
Table 2.5 The CO2 solubility in brine at 50⁰C
Pressure (KPa) xCO2 (%)
926 0.178
3091 0.523
4966 0.764
With the sole purpose of interpreting the diffusivity experiments, we need functional
relationships between the gas phase pressure and the solubility of CO2 in water and brine. To this
end, we utilize Eqs. (2.3)– (2.6) as introduced in Section 2.3.2. To evaluate the fugacity of CO2 in
the vapor phase, we use the Peng–Robinson equation of state (EOS) and assume, that the partial
pressure of water in the vapor phase is sufficiently small, so that its impact on the fugacity
24
coefficient of CO2 is marginal. This assumption is supported by the measurements of Weibe and
Gaddy (1941) – the mole fraction of water in the vapor phase at 50° C never exceeds 0.01 (i.e.
2
CO
y > 0.99) at pressures between 1013.25–10,132.5 kPa. Furthermore, the mole fraction of water
in the vapor phase is evaluated from Raoult's law.
To arrive at a simple representation of the solubility data, we start by assuming that the
activity coefficient of CO2 (see Eq. (2.4)) in the liquid phase approaches its limiting value of 1
(asymmetric activity coefficient) for the range of solubilities/pressures of our experiments and
approximate the fugacity of CO2 in the liquid as
0
1 1 1
L
f x H : (2.7)
*0
12
exp( ( ) / ) V p p RT (2.8)
In this approximation, we furthermore assume that the partial molar volume of CO2 at
infinite dilution is a weak function of pressure as suggested by Van Ness and Abbott (1982), and
set the value of
*
1
V to 37.6 cm
3
/mol according to the work of Parkinson and DeNevers (1969).
Fig. 2.2 reports the fugacity (as calculated for the vapor phase – see discussion above)
divided by the pressure correction ( ) as a function of the mole fraction of CO2 in water (x1).
25
Fig. 2.2 Correlation of the solubility data for CO2/water at 50° C
We find from Fig. 2.2 that the approximate representation of the liquid phase fugacity
provides for a near perfect representation of the experimental observations. The slope of Fig. 2.2
provides an estimate for
0
1
H . We note, however, that this estimate does not represent the true
Henry's constant of the system, as the impact of composition on the non-ideal behavior (i.e.,
activity coefficient) was not included explicitly in the data reduction. Given that the purpose of the
solubility measurements is solely for use in the boundary condition for analysis of our diffusivity
measurements, we defer further use of the functional form of the excess Gibbs function (i.e.,
activity coefficient model) given the accuracy of the simplified representation (max difference
between model and data=1.2%). To further validate our experimental data (and model), Fig. 2.3
compares our measurements with data from the literature.
26
Fig 2.3 Comparison with data from literature: CO2-water at 50° C ( □ – Data from Houghton et
al., 1957, ○ and Δ – Data from IUPAC-NIST Solubility Database, NIST Standard Reference
Database 106)
Fig. 2.3 demonstrates that our measurements are in good agreement with literature data and
that the proposed simplified model performs well in the pressure range of relevance to our
solubility and diffusivity experiments.
A similar analysis was performed for our CO2 – brine solubility measurements, where we
assume that the partial molar volume of CO2, at infinite dilution, in brine is approximately the
same to that of the CO2/water system. Fig. 2.4 compares the experimental observations to the
simplified model and we observe a good agreement (max difference between model and data <
1%).
27
Fig. 2.4 Correlation of the solubility data for CO2/brine at 50° C
While a simplified representation of the solubility data works well for the pressure range
of this study (up to 6000 kPa), it is important to stress that solubility modeling at higher pressures
may require an explicit representation of the compositional effects on the nonideal behavior: In
such settings, one may not be able to ignore the impact of the activity coefficient in the phi-gamma
formulation.
2.4 Measurements of CO2 Mass Transfer in Water/Brine/Porous Medium
2.4.1 Experimental Approach
The diffusivity experiments were performed in the same experimental set-up that was used for the
solubility experiments (see Fig. 2.1 above), but employing different size reference and sample cells
83.3 cm3 and 54.8 cm3, respectively. A larger volume sample cell (height: 8.3 cm, inner diameter:
28
2.9 cm) was selected to accommodate the porous media. A known mass of water (or brine) was
first loaded into the sample cell. For the experiments involving a porous medium, a known mass
of solid particles was then added to the sample cell (wet packing) until the height of the particles
packed equaled the liquid level. The height of the liquid-saturated porous material was then
recorded for use in the subsequent interpretation (see Section 2.4.2 below). The system was then
allowed to reach thermal equilibrium at 50° C in the air-bath. Prior to a diffusion experiment, the
leak rate was measured using Helium following the procedure outlined in Section 2.3.1. The
sample and reference cells were then evacuated (as in the solubility measurements) and
subsequently isolated from each other. The reference cell was then charged with CO2 and allowed
to establish thermal equilibrium. The diffusion experiment was initiated by opening the valve in
between the reference and the sample cells in order to allow CO2 from the reference cell to enter
the sample cell (the pressure and temperature in both cells were recorded during the experiment
and their values are employed in the interpretation of the data – see additional discussion below).
When the pressure in the sample cell reached a certain value, the valve in between the two cells
was closed, thus isolating the reference cell. Past this point, the pressure in the sample cell
decreases due to CO2 diffusion and dissolution into the liquid phase. The pressure and temperature
of the system being monitored and recorded continuously allow for subsequent calculation of the
mass transfer rates of CO2 into the liquid phase (as discussed in Section 2.4.2).
2.4.2 Interpretation of Experiments
The experimental approach, as outlined in Section 2.4.1, provides a time series of pressures and
temperatures in the reference and sample cells. With known volumes of the reference cell and the
sample cell, we then calculate the number of moles of CO2 in the vapor phase via an equation of
state (the Peng–Robinson EOS is used here). This, in turn, allows us to calculate the number of
29
moles of CO2 in the liquid phase, ()
liq
nt , and in the vapor phase above the liquid in the sample cell,
()
s
nt , from the material balances:
( ) ( 0) ( ) ( ) ( )
liq r r r r s l s
n t V t V t V V t (2.9)
( ) ( 0) ( )
s r r liq
n t V t n t (2.10)
To relate the change in mole numbers with time to the diffusion of CO2 into the liquid
phase, a simple mathematical model was initially formulated based on Fick's law for diffusive
mass transfer. We assume that the diffusion process in this experimental system is a 1-D process
as illustrated in Fig. 2.5.
Fig. 2.5 Schematic of the assumed 1-D diffusion process
We further assume that the change in liquid volume due to evaporation and to swelling due to CO2
dissolution is negligible and can be ignored. This is a reasonable assumption given the low vapor
pressure of water at the experimental temperature (12.34 kPa at 50° C) and the low solubility of
CO2 in these liquids. Moreover, we assume (as it is commonly done in the technical literature to
date) that the diffusion coefficient is that for a single species (in fact, CO2 exists in various forms
30
in the solution, each potentially with its own diffusivity, so the measured diffusivity is an effective
diffusivity reflecting the overall transport of all carbon bearing species) and does not change
significantly with CO2 concentration in water/brine (a good assumption likely given the very dilute
CO2 concentrations in these liquids).
Based on these assumptions, the mass transfer of CO2 from the vapor phase into the liquid
phase is described by
2
2
CC
D
tz
(2.11)
subject to the following initial and boundary conditions
( , 0) 0 C z t (2.12)
(0, ) ( , )
eq s s
C t C p T (2.13)
( , ) 0
z
C L t (2.14)
Here, C denotes the concentration of CO2 in the liquid phase (mol/cm
3
), C eq is the solubility
(mol/cm
3
) of CO2 in the liquid phase at the pressure (ps) and temperature (Ts) of the sample cell,
D is the CO2 diffusivity (cm
2
/s - diffusion coefficient divided by the tortuosity in the experiments
with a porous medium being present) in the liquid. The equilibrium concentration of CO2 at the
liquid interphase is evaluated from our independent solubility measurements
1
/
eq w
C x M (2.15)
where is the density of the liquid phase (g/cm
3
), Mw is the molecular weight (g/mol) of the
mixture of CO2 and water/brine given by
31
1 ,1 1 ,
(1 )
w w w liq
M x M x M (2.16)
where
,1 w
M is molecular weight of CO2 and
, w liq
M the molecular weight of the fresh brine. The
mass density of the liquid phase ( ) increases slightly with CO2 dissolution as demonstrated by
Yan et al. (2011). They demonstrated, based on interpretation of literature data and new
measurements, that the correlation of Garcia (2001) for the apparent partial molar volume of CO2
in solution (cm
3
/mol)
2 4 2 7 3
1
37.51 9.585 10 8.74 10 5.044 10 V T T T
(2.17)
where T is the temperature (here in
ο
C), allows for accurate estimation of the liquid phase density
via
,1
1
1
w
M
V
(2.18)
11
1
1 1
b
ww
(2.19)
where
b
is the mass density (g/cm
3
) of the pure liquid (fresh brine) and
1
w is the mass fraction of
CO2 in solution.
The molar flow of CO2 (J1 - mol/s) into the liquid phase, across the vapor-liquid interface,
at any given time, is evaluated from
1
0 z
C
J A D
z
(2.20)
where A is the cross-sectional area of the PVT cell and ε is the porosity of the porous medium in
the cell (taken as unity for the bulk-liquid experiments with no porous medium being present).
32
Finally, the model is linked to the experimental observations by equating the change in moles of
CO2 in the sample cell to the molar CO2 across the interface
1
( ) ( )
ss
n t n t t J t (2.21)
The model was solved numerically via an implicit finite-volume method and was used to estimate
the diffusivity (D) from the experimental observations, i.e., by matching the moles of CO2 in the
vapor phase in the sample cell (evaluated from the changes in experimental pressure) over the time
of an experiment.
2.4.3 Results and Observations
Six diffusivity experiments were performed and initially interpreted according to the procedure
discussed in Section 2.4.1. The initial and final pressures in the sample cell for each experiment
are reported in Table 2.6. The observed pressure decay due to CO2 diffusion and dissolution in the
bulk liquids and in the liquid-saturated porous media is reported as a function of time in Figs. 2.6–
2.8, where we report the normalized pressure (i.e., the ratio of pressure to initial pressure) to
facilitate comparison.
Table 2.6 The initial and final pressures in the sample cell
Experiment Initial pressure (KPa) Final pressure (KPa)
Bulk water 5989 4250
Bulk birne 6031 5119
Water+beads 5646 4895
Brine+beads 6137 5702
Water+quartz 5900 5556
33
Brine+quartz 5918 5786
0 2 4 6 8 10 12 14 16 18
0.70
0.75
0.80
0.85
0.90
0.95
1.00
Bulk Water
Bulk Brine
Normalized Pressure
Time (h)
Fig. 2.6 Normalized pressure decay for CO2 diffusion in bulk water and in bulk brine
From Fig. 2.6, we observe that the pressure decay in the bulk water experiment is larger
than that in the corresponding bulk brine experiment. This is due to the larger solubility of CO2 in
water compared to that brine, and is consistent with the outcome of the CO2 solubility
measurements (see Section 2.3). From the slope of the pressure decay curves, it is also evident that
the pressure in the water experiment decays faster than in the brine experiment. This difference in
slope indicates that the CO2 diffusivity in water is larger than that in brine.
34
Fig. 2.7 reports the normalized pressure decay from the diffusivity experiments in a liquid-
saturated porous medium made from 1.6 mm glass beads. We observe that these experiments
behave similarly to the bulk liquid measurements (see Fig. 2.6).
0 5 10 15 20 25 30
0.86
0.88
0.90
0.92
0.94
0.96
0.98
1.00
1.02
Water+beads
Brine+beads
Normalized Pressure
Time(h)
Fig. 2.7 Normalized pressure decay for CO2 diffusion in a liquid-saturated porous medium (1.6
mm glass beads)
The normalized pressure decay for diffusivity experiments in a liquid-saturated porous
medium made from 125-150 µ m quartz particles is reported in Fig. 2.8. Here, we observe that the
pressure decreases much more slowly than for the corresponding experiments in bulk liquid and
in the liquid-saturated porous medium made of glass beads. At the end of the experiment, the
pressure is still reducing (far from equilibrium) even after 15 h. These observations indicate that
the diffusive mass transfer in the liquid-saturated porous medium made from quartz particles is
35
significantly smaller than that in the bulk liquid and in the liquid-saturated porous medium made
from 1.6 mm glass beads.
0 2 4 6 8 10 12 14 16 18
0.94
0.95
0.96
0.97
0.98
0.99
1.00
Water+Quartz
Brine+Quartz
Normalized Pressure
Time(h)
Fig. 2.8 Normalized pressure decay for CO2 diffusion in a liquid-saturated porous medium (125-
150 µ m quartz particles)
Figs. 2.9 and 2.10 compare the experimental observations with the simple 1-D
interpretations (Eqns. 2.11-2.16) for the bulk water and bulk brine experiments in terms of moles
of CO2 in the gas phase as a function of time. In addition, Fig. 2.10 includes error estimates
(bracketed by the broken lines) as obtained from error propagation analysis. We observe that the
model matches the data very well for both the bulk water and brine measurements, with estimated
values of CO2 diffusivity in water and brine of 2.93× 10
-3
cm
2
/s and 1.09× 10
-3
cm
2
/s, respectively.
The value of the diffusivity in water is found to be higher than that in brine, which is also consistent
36
with the difference in slope in the pressure decay curves. However, the order of magnitude of the
diffusivities obtained in bulk water and brine experiments (~10
-3
cm
2
/s) is larger than what is
expected for pure diffusive mass transfer in the liquid.
0 2 4 6 8 10
0.040
0.045
0.050
0.055
0.060
0.065
Experimental data
Predicted value
Equilibrium
Amount of CO2 in vapor phase (mol)
Time(h)
Fig. 2.9 Interpretation of the CO2 diffusivity experiment in bulk water
A potential explanation for the high values of diffusivity observed in the bulk
measurements is density-driven convection (e.g., Weir et al. 1996). As the density increases with
increasing dissolution of CO2 at the interface, the fluid on the top becomes denser. This density
gradient can then induce natural convection in the system which, in turn, accelerates the mass
transfer of CO2 into the water/brine and results in higher flux of CO2 at the interface. If this is the
case, we expect that the effect of convection will be reduced in experiments with a porous medium
because the porous structure can retard the natural convection. Additional analysis, to this end, is
provided in Section 2.5.
37
0 2 4 6 8 10 12 14 16 18
0.050
0.055
0.060
0.065
0.070
Experimental data
Predicted value
Equilibrium
Amount of CO2 in vapor phase (mol)
Time(h)
Fig. 2.10 Interpretation of CO2 diffusivity experiment in the bulk brine
Figs. 2.11 and 2.12 provide a comparison of the experimental observations (in terms of
moles of gas in the sample cell over time) and the interpretation for the water- and brine-saturated
porous medium made of 1.6 mm glass beads. The calculated and experimental behavior are in
good agreement for both experiments. The CO2 diffusivity estimated from the water experiment is
2.61× 10
-3
cm
2
/s while that for the brine experiment is 8.20× 10
-4
cm
2
/s.
We observe that the diffusivity values obtained from interpretation of these experiments
are smaller than those obtained from the bulk experiments due, likely, to the presence of the porous
medium and the fact that we estimate effective diffusivities for these experiments. However, the
magnitude of diffusivity is still on the order of 10
-3
cm
2
/s, which suggests that convection may still
contribute to the mass transfer in the glass-bead system: A particle size of 1.6 mm is, apparently,
not sufficiently small to prevent the onset of convection.
38
0 2 4 6 8 10 12
0.050
0.055
0.060
0.065
Experimental data
Predicted value
Equilibrium
Amount of CO2 in vapor phase (mol)
Time(h)
Fig. 2.11 Interpretation of CO2 diffusivity experiment in a water-saturated porous medium (1.6
mm glass beads)
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
0.060
0.065
0.070
Experimental data
Predicted value
Equilibrium
Amount of CO2 in vapor phase (mol)
Time(h)
Fig. 2.12 Interpretation of CO2 diffusivity experiment in the brine-saturated porous medium (1.6
mm glass beads)
39
Figs. 2.13 and 2.14 compare experimental observations with the interpretations for the
water- and brine-saturated porous medium made of quartz particles. Again, we observe that the
model matches the experiments very well for both the water and the brine experiments. Here, the
CO2 diffusivities for the water and brine experiments are estimated to 3.37× 10
-5
cm
2
/s and 1.25× 10
-
5
cm
2
/s, respectively. It is noticeable that the diffusivities obtained from these experiments are on
the order of magnitude of 10
-5
cm
2
/s, which agrees with other experimental observations reported
for water and brine (see Table 2.1). This also suggests that the presence of the quartz particle
significantly reduces the effect of natural convection, and that the mass transfer in this porous
system is dominated by molecular diffusion in the liquid phase.
0 2 4 6 8 10 12 14 16
0.058
0.060
0.062
Experimental data
Predicted value
Amount of CO2 in vapor phase (mol)
Time(h)
Fig. 2.13 Interpretation of CO2 diffusivity experiment in the water-saturated porous medium
(125-150 µ m quartz particles)
40
0 2 4 6 8 10 12 14 16 18
0.0630
0.0635
0.0640
0.0645
Experimental data
Predicted value
Amount of CO2 in vapor phase (mol)
Time(h)
Fig. 2.14 Data interpretation for the measurement of CO2 diffusivity in brine-saturated porous
media made of 125-150 µ m quartz particles
2.5 Natural Convection
As mentioned above, in some of our experiments the 1-D diffusion model predicts a diffusivity of
O (10
-3
cm
2
/s) magnitude, which is much larger than what is expected for purely diffusive mass
transfer. This implies that convection may have set-in for these cases, so that the 1-D diffusion
model is no longer suitable to describe the CO2 mass transport in the brine. In this section, we
perform high-resolution direct numerical simulations in a two-dimensional (2-D) rectangular
domain to predict the convective dissolution process in the liquid phase. Our simulations here only
focus on interpreting the experiments with the water saturated porous medium.
41
In an isotropic and homogenous porous medium, the flow is incompressible and obeys
Darcy’s law (Nield and Bejan 2006),
∇ ∙ 𝐮 = 0 (2.22)
𝐮 = −
𝐾 𝜇𝜀
(∇𝑃 − 𝜌𝑔 𝐞 𝑧 ) (2.23)
where u = (u, w) is the volume-averaged pore velocity, K is the medium permeability, is the
dynamic viscosity of the fluid, ε is the porosity, P is the pressure field, g is the acceleration of
gravity, ez is a unit vector in the z direction. The density is approximated as a linear function of
the concentration C of dissolved CO2:
𝜌 (𝐶 ) = 𝜌 0
+ ∆𝜌 0
𝐶 𝐶 𝑒𝑞 ,0
(2.24)
where 0 is density of the fresh water, 0 is the density change after the water becomes saturated,
and Ceq,0 is the concentration of CO2 in saturated water at time t = 0 (when the valve is closed).
The concentration field in the water phase is governed by the following advection-diffusion
equation
𝜕𝐶
𝜕𝑡
+ 𝐮 ∙ ∇𝐶 = 𝐷 ∇
2
𝐶 . (2.25)
For boundary conditions, the upper boundary of the liquid phase (i.e., the vapor-liquid
interface at z = 0) is always instantaneously saturated with CO2 and impermeable to fluid; the
lower boundary is impermeable to solute and fluid:
𝜕 𝑧 𝐶 = 0 and 𝑤 = 0 at 𝑧 = 𝐿 , 𝐶 = 𝐶 𝑒𝑞
(𝑡 ) and 𝑤 = 0 at 𝑧 = 0. (2.26)
And all the variables are periodic at the side walls. We note that the equilibrium concentration at
the interface varies with time as the pressure in the gas phase decays. The variation in pressure
42
with time is evaluated from a material balance on the moles of CO2 in the vapor phase for a fixed
vapor volume and compressibility factors from NIST Webbook.
In this study, Eqs. (2.22) – (2.26) are solved numerically using a Fourier-Chebyshev-tau
pseudo-spectral solver which was verified in Wen et al. (2013, 2015). For temporal discretization,
a third-order-accurate semi-implicit Runge-Kutta scheme (Nikitin 2006) was utilized for
computations of the first three steps, and then a four-step fourth-order-accurate semi-implicit
Adams-Bashforth/Backward-Differentiation scheme (Peyret 2002) was used for computation of
the remaining steps. For each time step, the saturated CO2 concentration at the interphase, Ceq, was
updated explicitly by using Henry’s law, as described in section 3.
The dynamics of this system, described by Eqs. (2.22) - (2.26), are governed by two
dimensionless parameters (assuming a constant concentration of CO2 at the top boundary): the
domain aspect ratio r = W/L, where W is the domain width or the diameter of the sample cell, and
the Rayleigh number
𝑅𝑎 =
𝐿𝐾𝑔 ∆𝜌 0
𝜇𝜀𝐷
(2.27)
representing the ratio of driving to damping forces. In an infinitely-wide domain ( r ), when
the Rayleigh number exceeds some critical value, mass transfer is primarily in the form of
convection. Although the specific value of this critical Rayleigh number for onset of convection is
still subject to debate in many studies (Ennis-King et al. 2005; Xu et al. 2006; Riaz et al. 2006;
Hassanzadeh et al. 2006; Kim et al. 2008; Kim and Choi 2012; Slim and Ramakrishnan 2010; Pau
et al. 2010; Elenius and Johannsen 2012; Slim 2014), there is no doubt that convection can be
observed when Ra > 100. Table 2.7 shows the specific values of diffusivity, permeability and
domain aspect ratio used in the 2-D simulations.
43
Table 2.7 Parameters utilized for the sample cell in the 2-D simulations
Materials Diffusivity× 10
5
(cm
2
/s) Permeability× 10
11
(m
2
) r Ra
Water-saturated 1.6 mm
glass beads
3.37 250.11 0.5 18442
Water-saturated 125-
150 µ m quartz particles
3.37 2.48 0.5 173
The permeability (K) of the unconsolidated porous materials, reported in Table 2.7, was
evaluated from the Kozeny-Carman correlation (Wyllie and Gregory, 1955):
𝐾 =
𝜀 3
𝑎 𝑆 𝑔𝑟
2
(1−𝜀 )
2
(2.28)
where Sgr is the surface area per unit grain volume and a is the Kozeny-Carman constant. In this
work, we set a = 5 and evaluate Sgr = 3/rp, where rp is the particle radius. The resulting Rayleigh
numbers for water-saturated 1.6 mm glass beads and water-saturated 125-150 µ m quartz particles
are, respectively, 18442 and 173. Obviously, for the water-saturated 1.6 mm glass beads, there
exists strong convection for dissolution.
We start by considering the water-beads experiment. Figs. 2.15 and 2.16 show the CO2
dissolution rate predicted by the 1-D diffusion model, using a diffusivity value of 3.37· 10
−5
cm
2
/s,
as obtained from the water-quartz experiment, is much smaller than that measured from the
experiments. In contrast, the 2-D convection model not only predicts the general evolution of the
quantity of CO2 in the vapor phase, but also captures the varying trends of CO2 dissolution rate for
different flow regimes.
44
Fig. 2.15 Comparison of the amount of CO2 in the vapor phase with numerical simulations in a
water-saturated porous medium (1.6 mm glass beads)
When CO2 is filled into the sample cell, it first accumulates near the vapor-liquid interface
and generates a thin CO2-saturated, diffusion boundary layer. As more CO2 dissolves into the water,
this diffusive layer becomes unstable under the influence of gravity and the CO 2-rich fluid sinks
in the form of slender fingers (see Fig. 2.17). Before the onset of convection, the mass transfer is
by diffusion and the CO2 dissolution flux declines as t
-1/2
; after that, the dissolution is greatly
enhanced by convection. Convection continuously brings new water/brine not saturated with CO2
to the interface and the flow transitions to a “quasi-steady” convective dissolution regime. The
dissolution of CO2 into the water will induce a pressure drop in the gas phase, which then reduces
the amount of CO2 dissolved in the liquid phase. However, in the current experimental
circumstance of 1.6 mm glass beads, the reduction of the pressure is rather small (see Table 2.6)
and as a result the relative change in the saturated CO2 concentration is also small. This results in
a dissolution rate that remains nearly invariant during the quasi-steady convective regime. As the
45
fresh water is depleted, the convection begins to shut-down and the dissolution rate declines
rapidly with time (see Fig. 2.16). The experimental dissolution rate, reported in Fig. 2.16, was
obtained from fitting the experimental data (moles of CO2 in gas phase) to a smooth function and
evaluating the derivative (See Appendix B).
Fig. 2.16: Comparison of CO2 dissolution rate with numerical simulations in a water-saturated
porous medium (1.6 mm glass beads)
In Fig. 2.16, the first vertical dashed line denotes the time t1=3000· Tad when the dissolution
rate starts to increase above that of pure diffusion. Tad=D/U
2
is the advection-diffusion-balanced
timescale and U=Kg 0/(µε) is the buoyancy velocity. Our experiments were not designed to
investigate the onset dynamics, and since t1 is usually very small at large values of Ra (t1 ≈ 9 s in
the water-bead case), it cannot be estimated accurately from the current experimental data (with
measurements every ∼2 s). We note that the transition time, t1, ranges from O(100Tad) to
46
O(1000Tad), depending on the magnitudes and types of perturbations (Slim, 2014) and that our
simulation results substantially agree with previous work on linear stability analysis (e.g., Riaz et
al., 2006, Javaheri et al., 2010, Elenius et al., 2014).
The second vertical dashed line in Fig. 2.16 denotes the time t 2=16· Ta, with Ta=L/U=the
advection time scale, where the mass transfer process transitions from the quasi-steady to the shut-
down regime according to previous studies by Hewitt et al. (2013) and Slim (2014): This estimate
of t2 also agrees well with both the experimental and numerical results.
Fig. 2.17 shows the evolution of the normalized CO2 concentration (i.e., C/Ceq,0) in
numerical simulations for convective CO2 dissolution in water-saturated 1.6 mm glass beads. For
this case, the onset of convection occurs in a very short time due to the large value of the Rayleigh
number: Apparent fingers can be observed beneath the CO2-water interface at t = 44 sec. At t =
6.2 min, the earliest fingertip reaches the bottom wall. And for t > 15 min, the convection begins
to be shut-down.
47
Fig. 2.17 Snapshots of normalized CO2 concentration fields in liquid phase from the numerical
simulations in a water-saturated porous medium (1.6 mm glass beads)
For the water-saturated 125-150 µ m quartz particles Ra=173, and convection should also
prevail in a sufficiently wide domain. Nevertheless, no convection was observed in our 2-D
simulations due to the limitation of the small domain width (corresponding to the experimental
configuration): Mass transfer in this case is purely diffusive. It is not surprising, therefore, that the
1-D diffusion model can reproduce the experimental observations and estimate a diffusivity in the
expected range, as shown in Fig. 2.13.
2.6 FT-IR Study of CO2 Sorption on Mt. Simon Sandstone
In the above discussion, we presented the study on the CO2 mass transfer in bulk water/brine and
in water/brine-saturated porous media via the pressure-decay method. In this section we study CO2
mass transfer and sorption in brine-saturated Mt. Simon sandstone rocks via a different technique,
namely in-situ ATR-FTIR. The collected spectra during the experiments under the Mt. Simon
Formation relevant conditions presented below provide valuable information towards
understanding the CO2 sorption and mass transfer in these brine-saturated sandstones.
Several experimental studies of CO2 diffusivity in bulk brines, at reservoir conditions, have
been reported by Wang et al. (1996), Yang and Gu (2006), Azin et al. (2013), Raad et al. (2015),
Zarghami et al. (2017) and Shi et al. (2018). However, the experimental work, reported to date, on
the measurement of CO2 diffusivity in brine saturated porous media is limited. For example,
Renner (1988) measured CO2 diffusivity in brine saturated consolidated porous media (Berea
48
sandstone core) up to 5.86 MPa at 38 ⁰C. Shi et al. (2018) report measurements of CO2 diffusivity
in unconsolidated porous media made from glass beads (2 different sizes) at 50 ⁰C.
Infrared spectroscopy has proven to be a valuable experimental technique for investigating
CO2 adsorption on coals (Goodman et al., 2005; Goodman et al., 2009 and Mastalerz et al., 2012),
polymers (Culp et al., 2010; Culp et al., 2013; Kauffman et al., 2011a and Kauffman et al., 2011b),
shales (Sanguinito et al., 2018) and clays (Krukowski et al., 2015). Falk and Miller (1992) studied
the infrared absorption band due to CO2 dissolution in bulk water. Schadle et al. (2016) developed
a portable infrared attenuated total reflection spectrometer to detect and quantify the CO2 dissolved
in brine at 20 ⁰C and pressures up to 11 MPa. Goodman et al. (2019) studied CO2 sorption on
hydrated shales and clays with in-situ Fourier Transform infrared spectroscopy (FT-IR) at 40 ⁰C
and pressures up to 10.3MPa.
In summary, the information of CO2 mass transfer and sorption in brine saturated natural
porous media is very limited. Accordingly, the role of the combined mechanism of CO2 sorption
and diffusion in brine/rock systems is still unclear due to the limited availability of experimental
data. In this work, we apply in-situ attenuated total reflectance-Fourier transform infrared
spectroscopy (ATR-FTIR) to investigate CO2 mass transfer and sorption in a brine-saturated Mt.
Simon sandstone. The amount of absorbed CO2 is quantified by correlating the integrated FTIR
peak area to the CO2 solubility in bulk brine, as measured separately via a PVT-cell approach. The
combined mechanism of CO2 sorption and diffusion is then deduced from the observed changes
in the infrared absorption band between experimental pressure stages.
2.6.1 Experimental Materials and Approach
49
The Mt. Simon sandstone samples investigated in this study were extracted from verification well
1 (VW1) at the depth of 6927 ft in the Illinois Basin - Decatur Project (IBDP). Their transport and
mechanical properties have been previously characterized by Shi et al. (2019). The CO2 used in
this study was Coleman-Grade with a purity of 99.99%. A synthetic brine was utilized (for its
composition see Table 2.8), simulating the brine collected in-situ at the Mt. Simon formation from
the IBDP Site in Decatur, Illinois (Labotka et al., 2015).
Table 2.8 Chemical composition of the brine
Reagent Concentration (g/L)
NaCl 111.65
Na 2SO 4 0.7
CaCl 2 59.15
MgCl 2·6 H 2O 16.75
KBr 6.45
LiCl 1.95
SrCl 2·6H 2O 2.45
Infrared spectra were recorded with a single-beam FT-IR spectrometer (Thermo Electron
Nexus 670 FTIR ESP) equipped with a wide-band mercury cadmium telluride (MCT) detector.
500 scans were collected at a resolution of 2.0 cm
−1
over the spectral range extending from 400 to
4000 cm
−1
. In-situ ATR-FTIR spectroscopy experiments were conducted in customized Spectra
Tech ATR cells that has been described previously (Goodman et al., 2005). The modified
equipment assembly consisted of two stainless steel high-pressure cells (up to 13.6 MPa) each
containing a cylindrical zinc selenide (ZnSe) ATR crystal. One cell contains a ZnSe crystal with
no sample present (blank) and the other cell contains a ZnSe crystal coated with the rock sample.
50
The two cells are connected via 1/16-in. stainless steel tubing, so that each cell experiences the
same pressure. The entire ATR cell assembly is enclosed in a heating jacket.
The Mt. Simon Sandstone samples were powdered by mortar and pestle and sieved to 73
μm. SEM photographs of the powder are shown In Fig. 2.18. Fig. 2.19 shows the particle size
distribution (PaSD) of the powdered and sieved sample which was generated using image analysis.
According to Fig. 2.18, the geometry of the particles is irregular, and their sizes differ substantially.
From the particle size distribution shown in Fig. 2.19, one notes that the PaSD follows a unimodal
distribution centered at ~40 μm. While particles are sieved using a 73 μm mesh, from the PaSD
one observes that the powder also contains particles larger than 73 μm. This may be explained
from Fig. 2.18 that indicates the presence of elongated particles which may be able to pass through
the mesh may be small enough to fit through the mesh, even though their major axis exceeds 73
μm. For the FTIR measurements, the sieved powders were mixed with brine to create a slurry. A
small quantity of that slurry (~100 mg) was used to generate with a paint brush a thin layer coated
onto the ZnSe ATR crystal. Subsequently, the ZnSe ATR crystal was loaded into the ATR cell
whose temperature was then set to 50 ⁰C. The cell was pressurized with nitrogen at 4.2 MPa in
order to conduct a leak test for a period of 24 hours. At the end of the leak test, the downstream
valve was opened slightly to allow the nitrogen to slowly flow out of the sample cell until
atmospheric pressure is reached. One spectra was then recorded as the starting point, with the
sample being exposed to N2 under atmospheric conditions. The cell was then pressurized with CO2
at 0.3MPa and the spectra was recorded every 2 hours for ~2 days, to make sure the system reached
equilibrium. Subsequently, the cell pressure was raised in a stage-wise manner to 0.7, 2.8, 5.5 and
8.3 MPa, and at each pressure stage the spectra was again recorded over ~2 days. Specifically, the
51
spectra were recorded with a time interval of 2-3 min during the first hour, of each pressure stage,
with subsequent scans at larger intervals (from several min to several hr).
Fig. 2.18 SEM photographs of the sample powders
Fig. 2.19 Particle size distribution of the sample powders
2.6.2 Results and Discussion
52
The ATR-FTIR spectra of a brine-saturated Mt. Simon sandstone sample, measured at 50⁰C with
N2 at atmospheric conditions (before being exposed to CO2), is reported in Fig. 2.20. Four major
peaks are observed. The peak near 1012 cm
-1
and the broad peaks between 620 cm
-1
and 700 cm
-
1
, correspond to Si-O vibration modes associated with clays. The peaks near 3385 cm
-1
and 1635
cm
-1
represent the H-O-H asymmetric and symmetric stretching modes, and the H-O-H bending
mode, respectively. In addition, a very small peak is observed near 2342 cm
-1
. This is caused by
CO2 in the brine (equilibrated with CO2 present in the atmosphere). The relatively simple
mineralogy detected from the IR spectra is consistent with what is reported by Shi et al. (2019):
The mineral composition of Mt. Simon Sandstone is dominated by quartz with at weight
percentage of ~ 80%.
Fig. 2.20 ATR-FTIR spectra of brine saturated Mt. Simon Sandstone
Fig. 2.21 compares the spectra for the sample before and after exposure to CO2 at 8.3MPa.
The total exposure time was 268 hr. No significant changes are observed for the peaks between
620 cm
-1
and 700 cm
-1
, and the peak at 1012 cm
-1
. This suggests, that during exposure to CO2, the
0
0.5
1
1.5
2
2.5
600 1100 1600 2100 2600 3100 3600
Absorbance
Wave number (cm
-1
)
53
mineralogy of the brine saturated sample, remains unchanged. In addition, no changes are observed
for the peaks at 1635 cm
-1
and 3385 cm
-1
, which means there is virtually no change in the IR peaks
corresponding to water. A single symmetrical absorption band, centered at 2342cm
−1
, is observed
after exposure to CO2. This location of the CO2 peak is consistent with the infrared band of CO2
absorbed in brine, as reported by Goodman et al. (2019) and Schadle et al. (2016). It indicates a
substantial increase in the quantity of CO2 that is either dissolved in the brine or adsorbed on the
surfaces of the sandstone sample itself. Aqueous carbonate and bicarbonate ions are both known
to absorb strongly in the 1300-1400 cm
-1
region (Falk and Miller, 1992). It is noticeable that no
infrared bands of aqueous carbonate and bicarbonate ions are observed in Fig. 2.21. It means that
most of the absorbed CO2 exists in the form of CO2 molecules. As reported by Sanguinito et al.
(2018), a CO2 bending mode peak between 655 and 670 cm
−1
was observed for CO2 adsorption on
the dehydrated shale and clay samples. However, the CO2 bending mode peak become extremely
weak when CO2 adsorption on identical but hydrated shale and clay samples (Goodman et al. 2019).
In our measurement, as shown in Fig. 2.21, no significant CO2 bending mode peak is observed
when the brine saturated sample was exposed to CO2.
54
Fig. 2.21 Comparison of spectra of sample before exposure and exposure to CO2 at 8.3MPa
The complete absorption spectra for CO2 sorption in brine-saturated Mt. Simon sandstone
during these experiments are shown, as a function of increasing CO2 pressure, in Fig. 2.22. As the
CO2 pressure increases, the intensity of the absorption band at 2342 cm
-1
, due to absorbed CO2,
also increases.
55
Fig. 2.22 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone as a
function of increasing CO2 pressure up to 8.3 MPa
Figs. 2.23 and 2.24 show the spectra of adsorbed CO2 on dry Mt. Simon sandstone as a
function of increasing CO2 pressure up to 8.3 MPa. Both the intensity of the CO2 asymmetric
stretch at 2342 cm
-1
(as shown in Fig. 2.23) and CO2 bending mode between 630 and 680 cm
−1
(as
shown in Fig. 2.24) increase with increasing pressure. However, the location and the shape of the
CO2 peak for the asymmetric stretch, in this case, is very different from that of CO 2 absorbed on
the brine saturated sample: For CO2 absorption in the brine saturated sample, the peak is centered
at 2342 cm
-1
and is symmetrical. When CO2 adsorbs (physisorption) on rock surfaces, with no free
water present, we observe multiple and asymmetrical peaks. According to Fig. 2.23, the
asymmetric peaks are centered at 2331 cm
-1
and 2357 cm
-1
. The multiple peaks for CO2 physically
adsorbed to the surfaces of the shale and clay samples were also reported by Sanguinito et al. (2018)
and Goodman et al. (2019). Goodman et al. (2019) monitored the asymmetric stretch peak of CO2
while dehydrating the sample. They found that the peak changed from a single symmetrical peak
centered at 2342cm
−1
, to multiple asymmetric peaks that were centered at 2343cm
−1
and 2331cm
−
1
. In addition, they found that the intensity of the peaks for the dehydrated sample was much
weaker than for the hydrated sample.
56
Fig. 2.23 ATR-FTIR spectra of adsorbed CO2 on dry Mt. Simon Sandstone as a function of
increasing CO2 pressure up to 8.3 MPa: CO2 asymmetric stretch (2342 cm
-1
)
Fig. 2.24 ATR-FTIR spectra of adsorbed CO2 on dry Mt. Simon Sandstone as a function of
increasing CO2 pressure up to 8.3 MPa: CO2 bending mode (630 to 680 cm
−1
)
The difference between the CO2 asymmetric stretch peaks for sorption on brine saturated
and dry samples can be caused by different mechanisms of CO2 sorption. In the measurement of
0.3
0.35
0.4
0.45
0.5
0.55
2250 2300 2350 2400 2450
Absorbance
Wave number (cm
-1
)
0.3 MPa
0.7 MPa
2.8 MPa
5.5 MPa
8.3 MPa
0.2
0.22
0.24
0.26
0.28
0.3
0.32
0.34
0.36
0.38
0.4
630 640 650 660 670 680
Absorbance
Wave number (cm
-1
)
0.3 MPa
0.7 MPa
2.8 MPa
5.5 MPa
8.3 MPa
57
CO2 sorption on the dry sample, CO2 is physically adsorbed to the surfaces of the clay minerals.
In contrast, when CO2 sorption occurs on the brine saturated sample, the brine occupies the surface
sites for CO2 adsorption and acts as a barrier for CO2 adsorption onto the surfaces of the sample:
Almost all the CO2 molecules are dissolved in the brine with no detectable CO2 physisorption on
the surface (with this FTIR approach).
Fig. 2.25 reports the absorbed CO2 peak area as a function of time, when the CO2 pressure
in the sample cell is increased in a step-wise fashion. The recorded spectra for each pressure step
can be found in the Appendix C. During transitions from a lower pressure to the next pressure
level, the recorded peak area first increases rapidly and finally approaches a new equilibrium value.
Fig. 2.25 Absorbed CO2 peak area as a function of time
Fig. 2.26 shows the integrated peak area at equilibrium, corresponding to each pressure step, as a
function of the pressure. We note that the CO2 uptake in the brine-saturated Mt. Simon sandstone
58
continues to increase with pressure and is consistent with the expected increase in solubility at
higher pressures.
Fig. 2.26 Sorption isotherms of CO2 on brine saturated sample
The solubility of CO2 in the brine was measured separately in a PVT cell, as described by
Chapter 2. In these solubility measurements, a magnetic stirrer was used during the experiment to
accelerate the fluid mixing and equilibration. The results are shown in Table 2.9 and Fig. 2.27.
Two solubility models (Zhao et al. 2015b and Duan et al. 2006) were used to predict the CO2
solubility in the Mt. Simon brine and are compared to our measurements in Fig. 2.27. The average
absolute deviation (AAD) of the model of Zhao et al. (2015) is 3.6%, while the AAD of the model
of Duan et al. (2006) is 4.1%: The relatively small values of AAD between the modeling results
and our experimental data demonstrate the reliability and accuracy of our measurements.
Table 2.9 CO2 Solubility in brine at 50 ⁰C
Pressure (KPa) XCO2 (%)
636 0.107
59
1511 0.247
2271 0.354
3045 0.452
3930 0.554
4775 0.627
5664 0.706
Fig. 2.27 CO2 Solubility in brine as a function of pressure
The CO2 solubility at a given pressure is plotted versus the FTIR peak area at the
corresponding pressure in Fig. 2.28. We observe that a linear model fits the spectral data with an
R
2
= 0.998, indicating that the IR measurements are suitable for a direct quantification of CO 2
dissolved in brine. This provides a pathway to study mass transfer via the transient period of the
IR scans as discussed further below.
0.000
0.100
0.200
0.300
0.400
0.500
0.600
0.700
0.800
0.900
0 1000 2000 3000 4000 5000 6000
More fraction of CO
2
in liquid (%)
Pressure(KPa)
Our data
Zhao2015
Duan2006
60
Fig. 2.28 Peak area vs. solubility in brine at the same pressure and temperature
In order to establish a model to describe/interpret the observed dynamics (as shown in Fig.
2.25), we consider a simple dual-porosity configuration: The system consist of packed particles
(see particle distribution above). The internal porosity of the particles is saturated with brine as is
the interparticle space (see Fig. 2.29).
61
Fig. 2.29 Schematic of a flat sandstone sample sitting on the ATR crystal
The brine saturated sample on the crystal includes two porosities: The porosity of the
particles, εi. We assume here that the particles maintain the porosity of the whole sample. A value
of εi of 0.15, as measured by Shi et al. (2019) for a core sample, is used in the modeling. The other
porosity is the porosity between particles, εm. The value of εm cannot be measured directly and
must be determined by fitting the model to the experimental data. The thickness of the sample, L,
as shown in Fig. 2.29, is correlated to the packing porosity, εm, based on the Equation (2.29).
1
ms
s
m
V
V
L
S
, (2.29)
where Vs is the volume (cm
3
) of the particles, S is the total area (cm
2
) of the slurry pasted on the
ATR crystal. Both Vs and S are evaluated from a material balance on the paste applied to the ATR
crystal. The conceptual model assumes that the CO2 molecules diffuse from the top surface of the
sample to the surface of the crystal. The CO2 concentration at the top surface (Z=0) is equal to the
62
equilibrium solubility of CO2 in the brine at the given pressure. A 1-D diffusion model that
describe the process can be written as:
2
2 eff
CC
D
tz
, (2.30)
(0, )
e
C t C , (2.31)
( , )
0
C L t
z
, (2.32)
0
( ,0) C z C , (2.33)
where Cs is the CO2 solubility (mol/cm
3
) at the given pressure (as reported in Fig. 2.27). C0 is the
initial CO2 concentration (mol/cm
3
) in the sample. Deff is the effective diffusivity of the porous
system, and is calculated via the Maxwell-Gamett equation (effective medium theory):
2(1 )( ) 2
2 (1 )( )
m i m i m
eff m
m i m i m
D D D D
DD
D D D D
, (3.34)
1
m
m
J
m
DD
D
, (3.35)
1
i i i
i
J
i
DD
D
, (3.36)
where Dm is the diffusivity (cm
2
/s) of the packing materials, Di is diffusivity of particles, and D is
CO2 diffusivity in bulk brine, that was measured by Shi et al. (2018). τm is the tortuosity of the
particle pack and τi is the tortuosity of the average particle. Finally, J is a power factor that is used
to relate porosity to tortuosity. We note that both J and εm must be determined by fitting the model
63
to the experimental data. The analytical solution to Equations (2.30) to (2.33), in terms of the
dimensionless concentration, is given by
22
0
2
0
0
4 1 (2 1) (2 1)
1 exp sin
2 1 4 2
eff
m
e
Dt
CC m m z
C C m L L
, (3.37)
To estimate the model parameters (J and εm), we use the experimental data from 0.7, 2.8 and 5.5
MPa and the following procedure: Based on the linear relationship between solubility and peak
area (see Fig. 2.28), we can calculate the dimensionless concentration from Equation (3.37) (given
initial estimates of J and εm) and compare the value to the experimental data,
0
0 e
AA
AA
. Here, A0 is
the integrated area for the peak at 2342 cm
-1
at the beginning of a pressure step, Ae the equilibrium
peak area at the end of the pressure step, and A is the peak area at any given time. We use an
unconstrained minimization routine to estimate the model parameters by reducing the relative error
at each time-point for the data set. It is important to note that we use the average dimensionless
concentration in a 2 µ m layer just above the crystal, since the IR penetration depth is ~1 to 2 µ m
(Goodman et al. 2005).
The modeling results and the experimental data for 0.7, 2.8 and 5.5 MPa are shown in
Figs.2.30-2.32. The parameter estimation results in values of J = 2.25 and εm =0.145. Based on
these values, the effective diffusivity of the overall porous material is 9.13× 10
-8
cm
2
/s. One should
notice that we made an assumption that the value of εi is 0.15 that is the same as measured for a
core sample. In addition, we collected the average dimensionless concentration in a 2 µ m layer
above the crystal to fit the experimental data. However, in fact, the IR penetration depth is ~1 to 2
µm. Therefore, some simulations have been conducted to test the sensitivity of the model to the
value of εi and the penetration depth. The results of sensitivity test is reported in Appendix G.
64
Fig. 2.30 Modeling result and experimental data for pressure of 0.7 MPa
Fig. 2.31 Modeling result and experimental data for pressure of 2.8 MPa
65
Fig. 2.32 Modeling result and experimental data for pressure of 5.5 MPa
2.7 Discussion and Conclusions
In the previous sections, we have described a number of experiments to measure solubility and
mass transfer of CO2 in water and brine based on pressure-decay observations in a PVT cell that
included measurements in bulk liquids and in two liquid-saturated unconsolidated porous media.
A simple 1-D mathematical model was initially utilized to interpret the pressure-decay
observations in terms of CO2 mass transfer in the liquids and in the liquid-saturated porous media.
CO2 diffusivities obtained by fitting the model to the experimental data demonstrate that a simple
representation allows for good empirical agreement between experiments and modeling. Table
2.10 provides a summary of the diffusion coefficients as extracted from the experiments (these,
however, are effective diffusion coefficients since they entail for the bulk liquids and the large
pore size media convective in addition to diffusive transport, as previously discussed).
66
Table 2.10 Summary of the CO2 diffusivities measured in different experiments
Materials Diffusivity× 10
5
(cm
2
/s)
Bulk water 293
Water-saturated 1.6 mm glass beads 261
Water-saturated 125-150 µ m quartz particles 3.37
Bulk brine 109
Brine-saturated 1.6 mm glass beads 82
Brine-saturated 125-150 µ m quartz particles 1.25
From Table 2.10, we observe that the ratios between the effective diffusivity in water and
the corresponding value in brine, as extracted from the water and brine experiments via the 1-D
model, are 2.69 (bulk), 3.18 (glass beads) and 2.67 (quartz), respectively. Several factors may
contribute to the observed lower effective diffusivities of CO2 in brine when compared to those in
the water systems. They include:
Impact of viscosity: Per the Stokes-Einstein equation, the diffusion coefficient is inversely
proportional to the dynamic viscosity of the bulk phase. The viscosities of water (0.547 cP) and
brine (0.893 cP) were measured in our laboratory via Poiseuille flow through a tube at 1atm and
50° C, their ratio being ∼1.6: Based on viscosity effects, alone, one would then expect the effective
diffusivity in water to be at least 1.6 times larger than that in brine. In addition to the impact of
viscosity on diffusive mass transfer, convective mass transfer (in bulk and glass beads experiments)
is also affected by the viscosity (as represented here via Darcy's equation).
67
Impact of multicomponent diffusion: In the interpretation of the uptake experiments, we
apply a simple Fick's law description of CO2 diffusion in a liquid. In reality, both the water-CO2
and (in particular) the brine-CO2 are multicomponent systems that, in principle, require a more
sophisticated representation of diffusive fluxes (e.g., a Maxwell–Stefan diffusion model that
properly accounts for the friction among the various species). In practice, that is rarely done as the
Maxwell–Stefan diffusion coefficients for each pair of species in solution must be known (or
determined via separate experiments). However, when one forgoes the use of the elaborate
Maxwell–Stefan description of transport in favor of the simplified 1-D Fickian description, the
additional drag between species in solution (and potentially also the charge imbalances between
diffusing ions) manifest themselves as an additional decrease in the effective diffusivity in the
CO2-brine system as compared to the CO2-water system.
The diffusivities (diffusion coefficient for bulk liquid and diffusion coefficient divided by
tortuosity for the porous materials) extracted from the pressure decay data span two orders of
magnitude (10
-3
– 10
-5
cm
2
/s). We explain this on the basis that measurements in bulk liquids and
in the (1.6 mm in diameter) bead packs are subject to accelerated mass transfer due to density-
driven natural convection. In contrast, measurements of mass transfer performed in quartz packs
(125-150 µ m diameter) provide diffusion coefficients that are consistent with the literature (10
-5
cm
2
/sec). Accordingly, the use of a porous material made of particles with suitable size can
eliminate the acceleration in CO2 mass transfer due to natural convection during diffusivity
measurement in PVT cells. Based on the diffusivity extracted from the water-quartz system, we
demonstrated that high-resolution numerical calculation accurately captures the uptake dynamics
observed in the experiments for settings where natural convection contributes substantially to the
overall mass transfer process.
68
The ATR-FTIR was applied to study the CO2 mass transfer and sorption in brine saturated
Mt. Simon sandstone. The CO2 sorption was characterized at elevated pressures from 0.3 to 8.3
MPa at a temperature of 50⁰C. A single symmetrical absorption band, centered at 2342 cm
−1
, was
observed for CO2 sorption in a brine saturated sample, while multiple asymmetrical peaks,
centered at 2331 cm
-1
and 2357 cm
-1
, were observed for CO2 adsorption on the surfaces of the dry
sample. The different CO2 peaks detected for dry and wet samples implies that when CO2 sorption
occur in the brine saturated sample, almost all CO2 molecules are dissolved into the brine, and that
no CO2 adsorption on the surfaces of the porous material is detectable via the FT-IR technique.
The amount of CO2 dissolution in the brine saturated sample can be quantified by a correlation of
the integrated peak area to the CO2 solubility in the bulk brine, as measured separately via a PVT-
cell approach. Finally, the CO2 mass transfer in the brine saturated sample, as detected by the FT-
IR, was subsequently delineated by a simple mathematical model, formulated based on effective
media theory and Fick's law for 1-D diffusive mass transfer. Our modeling and experimental
observations presented above clearly demonstrate the potential of using FT-IR as a tool of
measuring the CO2 diffusivity in porous materials.
69
Chapter 3: CO2/Brine Induced Evolution of the Transport Properties
of Mt. Simon Sandstone
2
3.1 Introduction and Literature Review
The processes at play during geological CO2 storage include CO2 flow and dissolution, diffusion,
and fluid–mineral interactions (Kampman et al., 2014). CO2 dissolution takes place during the
two-phase flow of CO2 and brine in the subsurface. Fluid-mineral interactions may convert
CO2 into carbonate minerals, a process thought to be beneficial for long-term storage, but may also
impact the reservoir’s transport and mechanical properties by destabilization of clays, corrosion of
mineral surfaces and/or associated precipitation of carbonates and other salts in the porous
structures. These processes contribute to CO2 immobilization in geological structures via (i)
residual trapping, involving CO2 being held immobile in the pore space by capillary forces (ii)
dissolution trapping, resulting from CO2 being dissolved into the formation brine, and (iii) the so-
called ionic trapping that results from the interaction of CO2 (as bicarbonate ions) with the
reservoir rock surfaces.
Brines in contact with a propagating CO2 plume quickly (time-scales of hours to days)
saturate with CO2, which lowers their pH, and promotes mineral dissolution as well as desorption
of exchangeable or mineral surface-bound ions. CO2-laden brines flowing through a reservoir
rapidly dissolve (and get saturated with) carbonate and Fe-oxyhydroxide minerals, followed by
slower reactions with silicate and phyllosilicate minerals that further raise the pH, and cause
precipitation of clays and carbonates in the more distant regions (Kampman et al., 2014). A similar
2
Some results in this chapter has already been published (Shi et al., 2019)
70
set of chemical reactions likely occurs in caprocks, albeit at a more sluggish pace, as CO2 transport
in the caprocks is by diffusion rather than convective transport. The growth of clay and carbonate
phases (Munz et al., 2012) within the reservoir (caprock), likely to impact both transport and
mechanical properties, is governed by the fluid pH and the availability of metal cations resulting
from the slow dissolution of silicate minerals (likely to be rate-limiting long-term).
Mouzakis et al. (2016) and Miller et al. (2016) investigated how caprock porosity is altered
by geochemical reactions induced from CO2 dissolution in pore fluids. They studied, under
geological carbon storage (GCS) relevant conditions, changes in caprock porosity and pore
geometry for two caprock formations (Gothic Shale and Marine Tuscaloosa). For Gothic Shale,
pore connectivity and total porosity both increased, but for Marine Tuscaloosa porosity increased,
while pore connectivity decreased (changes in these properties can have profound impact on
caprock sealing capacity in long-term GCS). The decrease in pore connectivity was attributed to
mineral dissolution and re-precipitation in pore throats and/or the hydration of clays, found as a
larger fraction in Marine Tuscaloosa.
Tutolo et al. (2015) carried out flow-through experiments with K-feldspar-rich cores (Eau
Claire caprock, overlying the Mount Simon formation). Their results confirm the mechanism
discussed above, that increased acidity from CO2 injection into feldspar-rich cores will dissolve
primary feldspars and will precipitate secondary aluminum minerals, both phenomena manifested
by corresponding changes in permeability. A core, through which CO 2-rich deionized (DI) water
was recycled for 52 days, showed a decrease in bulk permeability, exhibited lower porosity, and
the presence of an Al-hydroxide secondary mineral (boehmite). Two samples subjected to ∼3 day
single-pass experiments with a CO2-rich model brine (NaCl in DI water) also showed decreases in
71
bulk permeability, but elevated porosity, and the formation of a phase with kaolinite-like
stoichiometry.
Huq et al. (2015) studied, via flow-through experiments, the water–rock interactions during
CO2 injection in sandstones from the Altmark natural gas reservoir (major components include
quartz, feldspars, clay minerals like illite and chlorite, and cements of carbonates and anhydrite)
under simulated reservoir conditions. Experiments were conducted using CO2-saturated DI water,
and a CO2-saturated model brine (containing NaCl and CaCl2). Fluid analysis suggested the
concurrent dissolution of both calcite and anhydrite. Dissolution of feldspar and minor amounts of
clay (chlorite) was also evident. The permeability of the sample increased by a factor of two,
mostly due to the dissolution of rock cements during brine injection. Szabó et al. (2016) studied
three core samples from the Pannonian Basin in Hungary, which were crushed and exposed to
CO2 environments in (1-month long) laboratory batch experiments. At the end of the incubation
period, the brine composition and the mineral content of the reacted rock samples were analyzed.
Dissolution of the carbonate minerals was observed, and evidence was also found for feldspar
dissolution, both processes leading to secondary carbonate and clay mineral formation.
Ilgen and Cygan (2016) evaluated, via modeling, whether pyrite and oligoclase dissolution
can account for the observed geochemical changes recorded in brine samples from a monitoring
well during the Frio-I pilot CO2 injection experiment. Previous geochemical modeling had
concluded that dissolution of calcite and Fe-oxyhydroxides, or release of adsorbed Fe, are the most
likely sources of the increased ion concentrations in these brines. They concluded that due to
kinetic limitations, dissolution of pyrite and oligoclase cannot account for the increased Fe and Ca
concentrations observed on the time scale of the Frio-I pilot test (10 days), and in agreement with
72
previous studies, dissolution of calcite and iron oxide (resulting in ∼0.02 wt.% loss of the reservoir
rock mass) is, responsible for the observed changes
Liu et al. (2012) carried out batch laboratory experiments on the CO2–brine–caprock
interactions for the Eau Claire Formation at 200
o
C and 30 MPa. Scanning Electron Microscopy
(SEM) and X-ray diffraction (XRD) analysis indicate minor dissolution of K-feldspar and
anhydrite, and precipitation of pore-filling and pore-bridging illite and/or smectite, and siderite in
the vicinity of pyrite. Simon and Anderson (1990) studied the stability of clay minerals in
hydrochloric acid and reported that chlorite is less stable than illite, kaolinite and feldspars.
Chlorite dissolution in acid involved the simultaneous removal of iron from both the brucite and
octahedral layers. Kö hler et al. (2005) measured the dissolution rate of a natural illite in electrolyte
solutions over a pH range of 2 to 12, and found the clay dissolution rate decreases continuously
with time. Rosenbauer et al. (2005) carried out experiments to study CO2-brine-rock interactions
and found an increase in CO2 solubility in the presence of limestone and arkosic sandstone. Shao
et al. (2010) studied clay dissolution and precipitation on the surface of phlogopite and observed
nanoscale pitting on the surface over short period of time (5h) under GCS conditions. Yu et al.
(2012) conducted core flooding experiments on rock samples using a CO 2-saturated brine and
found calcite dissolution was more pronounced than ankerite, followed by feldspar which showed
the least significant dissolution.
In summary, there have been a number of studies to date, primarily with model systems,
focusing on the impact on the pore structure, transport characteristics, petrophysical, mineralogical
and mechanical properties of rock sample due to exposure to CO2/brine systems. In general,
increases in porosity and the formation of new pores are reported, but the impact on the transport
properties is far from clear. Analysis of the resulting brines in contact with the cores, points to
73
mineral dissolution occurring consistent with the observations of new porosity generation.
However, the picture is significantly more complex with secondary mineral precipitation also
occurring, which could lead to plugging of pores and decreases in permeability, as it has been
reported in some instances. Things are further complicated by the presence of a large content of
clays in some of the systems studied, which tend to swell upon exposure to CO 2 and can impact
both the transport and mechanical properties of these systems. Studies of mechanical properties,
of mostly model systems, when exposed to CO2, have indicated generally a negative impact on
these properties.
Despite all the progress made to date, there still remains a lot to be learned of how
geological CO2 storage media being considered today for GCS interact with CO 2/brine systems.
Specifically, we know of no comprehensive study looking at the impact of such exposure on the
pore structural and transport characteristics of cores from a formation utilized in large-scale CO2
injection operations. In this study, we provide a comprehensive investigation of the Mt. Simon
Sandstone when exposed to CO2/brine at formation-relevant conditions.
3.2 Multiscale Characterizations of Mt. Simon Sandstone Samples Before and After Exposure to
CO2/Brine
3.2.1 Sample Preparation
The rock samples investigated in this study are from the Mt. Simon Sandstone that constitutes the
~1500 ft (~460 m) thick storage zone (Labotka et al. 2015) in the Illinois Basin - Decatur Project
(IBDP). In that project, in order to monitor the state of operations, a verification well (VW1) was
drilled in 2010, at a distance of 305 m (1000 ft) north of the CO2 injection well (Labotka et al.
74
2015). The rock samples studied in this research were all extracted from the VW1 from a narrow
depth ranging between 2110.4 - 2111.4 m (6924 - 6927 ft). Specimens with three different
geometries and sizes are investigated in this study. Table 3.1 reports the various samples studied
in this work. Seven cubic samples, with size of 1cm× 1cm× 1cm, were prepared from the original
VW1 core via air cutting. Four cylindrical core samples, with a size of 1”×2”, were extracted with
liquid nitrogen drilling. In addition, a cylindrical sample with size 0.25”×0.5” was prepared with
air drilling. All cylindrical samples were extracted, from the main core, in the horizontal direction
(along the bedding plane). The various studies carried out with these core samples are summarized
in Table 3.1. The photographs of 3 of these samples (6927-A1, 6927-B2 and 6927-C1) are shown
in Fig. 3.1.
Table 3.1 List of core samples studied in this work including types of tests performed
Sample
Depth
(ft)
Dimension
Aging
time
Characterization tests
before aging
Characterization tests
after aging
6927-A1 6927
1cm× 1cm×
1cm
7 days
Porosity measurement,
IC analysis
Porosity measurement,
IC analysis
6927-A2 6927
1cm× 1cm×
1cm
14 days
Porosity measurement,
flow perporometry
Porosity measurement,
flow perporometry
6927-A3 6927
1cm× 1cm×
1cm
- Porosity measurement N/A
6925-A1 6925
1cm× 1cm×
1cm
7 days
Porosity measurement,
IC analysis
Porosity measurement,
IC analysis
6925-A2 6925
1cm× 1cm×
1cm
14 days
Porosity measurement,
flow perporometry
Porosity measurement,
flow perporometry
6924-A1 6924
1cm× 1cm×
1cm
7 days
Porosity measurement,
IC analysis
Porosity measurement,
IC analysis
6924-A2 6924
1cm× 1cm×
1cm
14 days Porosity measurement Porosity measurement
6927-B1 6927 1“×2” -
Porosity/permeability
measurement
N/A
6927-B2 6927 1“×2” -
Porosity/permeability
measurement,
Mechanical testing
N/A
6927-B3 6927 1“×2” 7days Porosity measurement, Porosity measurement,
6927-B4 6927 1“×2” 14 days
Porosity/ permeability
measurement
Porosity/ permeability
measurement,
Mechanical testing
6927-C1 6927 0.25”×0.5” 7days
Porosity measurement,
N 2 BET, SEM
Porosity measurement,
N 2 BET, SEM
75
Fig. 3.1 Pictures of 1cm×1cm×1cm sample (left), 1”×2” core sample (middle) and 0.25”×0.5”
core sample (right)
A synthetic brine whose composition is shown in Table 3.2, simulating the in-situ brine
found in the Mt. Simon formation (Labotka et al. 2015), was used in all experiments reported here.
All the chemicals used to produce the brine were at least ACS grade. Deionized water was used as
the solvent to dissolve all the solutes listed in Table 3.2.
Table 3.2 Composition of the synthetic brine
Reagent Concentration (g/L)
NaCl 111.67
CaCl 2· 2H 2O 78.35
MgCl 2· 6H 2O 16.53
KBr 6.48
Na 2SO 4 0.48
SrCl 2· 6H 2O 2.37
LiCl 9.36
As detailed in Table 3.1, several of the rock samples were exposed to CO2 /brine for varying
time periods. For some of these samples (6927-A1, A2, 6925-A1, A2, 6924-A1, A2 and 6927-B4)
exposure to CO2/brine refers to placing the samples in an autoclave (see Fig. 3.2) filled with brine,
76
pressurizing the autoclave with CO 2 at 17.24 MPa (2500 psi) and incubating them there at 50
o
C
for various time periods, as indicated in the Fig. 3.2. Other samples (6927-B3 and 6927-C1) were
exposed to CO2/brine during flow-through experiments at a constant back pressure of 14.48 MPa,
a confining pressure of 16.55 MPa and a temperature of 45
o
C. The duration (and hence the
exposure time) varied with the given experiment, as noted in Table 3.1. The aged samples, were
then dried at 80
o
C under vacuum for 48 hours and characterized for their pore structure, transport
and mechanical properties, as discussed in more detail below (we note that samples were not rinsed
after incubation to avoid any additional destabilization of clays).
Fig. 3.2 Schematic of incubation experiments
3.2.2 Porosity and Permeability Measurements
The porosity of the rock samples was measured via the Helium pycnometry method. This method
replaces the traditional method of liquid displacement by using a test gas and the volume-pressure
relationship known as Boyle’s Law. The schematic is shown in Fig. 3.3. In our measurement, the
volume of reference and sample cells, Vref and Vsam, are first calibrated. The specimen is placed in
77
the sample cell. The reference chamber is pressurized with Helium to a certain pressure, P 1. Then
the valve between the two cells is opened, and the gas in the reference cell expands into the sample
cell. Once the system becomes stable, the pressure in both cells is recorded as P2. The porosity of
the specimen is then calculated via Eqn. 3.1:
12
1
ref sam ref
b
V V P V P
V
(3.1)
where is the porosity, and
b
V is the bulk volume of the specimen.
He Tank
Vent
Sample Chamber
Reference Chamber
Pressure Gauge
Fig. 3.3 Schematic of the Helium pycnometry porosity measurement set-up
The permeability of select samples was measured via N2 gas flow-through experiments
using a TKA-209 gas permeameter. A picture of the permeameter and a schematic of the
experimental set-up are shown in Fig. 3.4 and 3.5, respectively.
78
Fig. 3.4 Picture of the TKA-209 gas permeameter
Nitrogen Tank
Pressure Gauge
Pressure Regulator
Core Holder
Air Tank
Mass Flow Meter
Fig. 3.5 The schematic of the permeability measurement set-up
For these tests, the core sample is placed in the core holder under a confining pressure of
1.38 MPa. The upstream pressure is controlled by a pressure regulator. The gas flow rate (mol/s)
and the differential pressure are recorded as the upstream pressure is increased step-wise. The gas
permeability is calculated via Eqn. 3.2.
avg
JRT L
k
PP
(3.2)
79
where k is the gas permeability (m
2
), J is the molar flux (mol/s/m
2
), R is the gas constant, T is
temperature (K), µ is N2 viscosity (Pa*s), L is the length of the core (m), Pavg is the average pressure
of upstream and downstream (Pa), ΔP is differential pressure (Pa).
To facilitate comparison of permeabilities measured at various pore pressures all values of
permeability reported in this work are the so-called “liquid permeability”, which is the gas
permeability extrapolated to infinite pore pressure. For this purpose, we plot k versus 1/P avg. The
intercept is the liquid permeability which means the value of gas permeability when the average
pressure is infinite.
Table 3.3 reports the measured porosity values for all samples prior to the initiation of any
of the experiments. For some of these samples the porosity was also measured after the samples
had been exposed to the CO2/brine environment.
Table 3.3 Porosity of the specimens
Sample Porosity
Fresh (%)
Porosity
Aged (%)
Aging
(weeks)
Relative
Change (%)
6927-A1 20.9 21.5 1 2.9
6927-A2 23.8 25.8 2 8.4
6927-A3 25.3 - - -
6925-A1 26.3 26.6 1 1.1
6925-A2 19.2 20.5 2 6.8
6924-A1 20.7 21.6 1 4.3
6924-A2 21.5 23.2 2 7.9
6927-B1 16.0 - - -
6927-B2 14.7 - - -
6927-B3 15.0 15.3 1 2
6927-B4 14.8 17.0 2 14.9
6927-C1 23.6 24.2 1 2.5
80
There is, generally, some variability in the porosity values among the samples from each
depth location, reflecting likely natural variability in the 1-ft section available from each depth and
from which these cores were extracted. When comparing the values of samples from the same
depth prepared via air drilling (6927-A1, 6927-A2, 6927-A3, 6927-C1) to those prepared via liquid
nitrogen drilling (6927-B1, 6927-B2, 6927-B3, 6927-B4) the latter core samples average 20-30%
less porosity than the former samples, reflecting likely a stronger impact of air cutting and the
likely creation of new macrofractures and cracks. Another potential explanation to the discrepancy
in porosity is that cutting may have a larger impact on the smaller samples (1cm
3
cubes) prepared
by air cutting than the larger samples (2”×1” cylinders) prepared by liquid nitrogen drilling. In
addition, any inaccuracies in bulk sample volume measurement may have a lager impact for these
smaller size samples. Exposure to CO2/brine is observed to increase the total (He) porosity, but
the effect is generally small. For the three samples that were exposed to CO 2/brine for 7 days
(6924-A1, 6925-A1, 6927-A1, 6927-B3, 6927-C1) the relative changes in porosity were 4.3%,
1.1%, 2.9%, 2% and 2.5%, respectively. For the cores that underwent a longer (14 days) exposure
to CO2/brine (6924-A2, 6925-A2, 6927-A2, 6927-B4) the relative changes in porosity were 7.9%,
6.8%, 8.4% and 14.9, respectively.
The measured permeabilities are reported in Table 3.4. The permeabilities of the selected
core samples, which are drilled from almost the same vertical location, are around 10 md before
exposure to CO2/brine. After core 6927-B4 was exposed to CO2/brine it was dried at 80⁰C under
vacuum for 48 hours and its permeability (with a confining pressure of 1.38 MPa) was again
measured and reported in Table 3.4. The (liquid) permeability of the same core increased by a
factor of 3.5, from 10.6 mD (1 mD = 9.869e-16 m
2
) to 37.7 mD, implying that whatever core
81
material dissolution is occurring is likely impacting pore throats connecting previously
disconnected and/or inaccessible pores to flow.
Table 3.4 Permeability measurements before and after exposure to CO2/brine
Sample Fresh sample (mD) Aged sample (mD)
6927-B1 9.0 -
6927-B2 9.8 -
6927-B4 10.6 37.7
3.2.3 Ion Chromatography Analysis of Brine Composition
The concentration of cations and anions in the fresh brine, as well as of the various brine solutions
after experimentation, was determined via IC analysis, employing an ICS-2100 IC system (Dionex)
(See Fig. 3.6). The IC analysis was performed by injecting 25 µ L of an appropriately diluted
sample into the Dionex ICS-2100 IC system, which was equipped with an Ion Pac CS12A column
for cation analysis, and an Ion Pac AS14 column for anion analysis. In the IC analysis, we utilized
20 mM Methanesulfonic acid solution as the eluent for cations and (3.5 mM Na2CO3 + 1mM
NaHCO3) as the eluent for the anions. The eluent flow rate for cation analysis was set to 0.8
mL/min, while that for anion analysis was set at 1.2 mL/min. The suppressor voltage was adjusted
to 42 mA for best peak separation for cations and to 24 mA for anions. We used DI water (Fluka)
to prepare all diluted solutions for the IC tests. Standard solutions from Dionex were used to
calibrate the instrument using the absolute calibration method for 3 anions (Cl
-
, Br
-
, and SO4
2-
) and
6 cations (Ca
2+
, Mg
2+
, K
+
, Sr
2+
, Na
+
and Li
+
). The peaks of cations and anions during calibration
are shown in Fig.3.7 and 3.8, respectively.
82
Fig. 3.6 Photograph of the ICS-2100 IC system
83
Fig. 3.7 Peaks of cations
Fig. 3.8 Peaks of anions
84
The concentration of cations and anions in the original brine (prior to being used for
incubation of the Mt. Simon cores) as well as in the brine after incubation were determined by ion
chromatography and the results are reported in Tables 3.5-3.7 for three different cores (6924-A1,
6925-A1, and 6927-A1). Small, but analytically significant (i.e., greater than the accuracy of the
IC analysis measurements) changes are observed in the aged brines (when compared to the fresh
brine) from all three cores. One observes a consistent decrease in the Li concentration among the
three cores (ranging from -0.27% for core 6927-A1, to -0.74% for core 6925-A1, to -3.23% for
core 6924-A1) and a consistent increase in the concentration of three other cations, namely K
(ranging from 4.57% for core 6925-A1, to 6.08% for core 6927-A1, to 10.38% for core 6924-A1),
Mg (ranging from 0.97% for core 6924-A1, to 2.46% for core 6925-A1, to 3.65% for core 6927-
A1), and Ca (ranging from 0.17% for core 6924-A1, to 1.79% for core 6925-A1, to 3.32% for
core 6927-A1).
Table 3.5 The concentration of the cations and anions in the fresh and aged brines (6925-A1)
6925-A1
Before (mg/L) After (mg/L) Change (%)
Na 44601± 45 45550± 45 2.13
K 1895± 7 1981± 7 4.57
Mg 2283± 13 2339± 13 2.46
Ca 19769± 78 20123± 78 1.79
Sr 873± 6 871± 6 -0.19
Li 1232± 2 1223± 2 -0.74
Cl 112693± 450 115218± 450 2.24
Br 3871± 19 3849± 19 -0.57
SO4 745± 4 740± 4 -0.65
85
Table 3.6 The concentration of the cations and anions in the fresh and aged brines (6927-A1)
6927-A1
Before (mg/L) After (mg/L) Change (%)
Na 44424± 45 45316± 45 2.01
K 1889± 7 2004± 7 6.08
Mg 2280± 13 2363± 13 3.65
Ca 19743± 78 20398± 78 3.32
Sr 870± 6 896± 6 2.98
Li 1227± 2 1224± 2 -0.27
Cl 113813± 450 113419± 450 -0.35
Br 3862± 19 3819± 19 -1.13
SO4 753± 4 734± 4 -2.58
Table 3.7 The concentration of the cations and anions in the fresh and aged brines (6924-A1)
6924-A1
Before (mg/L) After (mg/L) Change (%)
Na 44601± 45 43972± 45 -1.41
K 1895± 7 2092± 7 10.38
Mg 2283.± 13 2305± 13 0.97
Ca 19769± 78 19803± 78 0.17
Sr 873± 6 881± 6 0.96
Li 1232± 2 1192± 2 -3.23
Cl 112693± 450 111037± 450 -1.47
Br 3871± 19 3746± 19 -3.24
SO4 745± 4 743± 4 -0.22
For two other cations, Na and Sr, the behavior depends on the specific core. With respect
to Na, for core 6924-A1, one observes a decrease in concentration (-1.41%), whereas for the other
two cores one observes increases in concentration (2.01% for core 6927-A1 and 2.13% for core
6925-A1). For Sr, a slight decrease (-0.19%) is observed for core 6925-A1, while increases (0.96%
for core 6924-A1 and 2.98% for core 6927-A1) are observed with the other two samples. For two
of the anions, Br and SO4, the aged brines from all three cores show a consistent decrease in
concentration. For the Cl, core 6925-A1 shows an increase in concentration, while the other two
86
cores show a decrease in concentration. Fig. 3.9 summarizes the change of brine composition
before and after aging.
Fig. 3.9 Summary of the brine composition change
It is quite a complex undertaking to carry-out a quantitative mineral compositional analysis
for these complex geological materials. Such analysis for the cores under consideration here is
available only at select depths. Table 3.8 reports such a mineral composition analysis from a depth
of 2120 m (6955.6 ft), which is the closest to the one from which the cores studied here were
87
extracted from (in the Appendix D we include additional mineral compositional analyses from
other depths in the formation).
Table 3.8 Mineral composition analysis (XRD) of the Mt. Simon core at 2120 m (6955.6
ft)
Minerals Wt %
Quartz 80
K-feldspar 8
Plagioclase 4
Ankerite/Fe-dolomite 1
Dolomite 1
Pyrite 1
Barite 1
Clay 7
Clay Wt %
Illite-smectite 52
Illite 47
Chlorite 1
The increase in the concentration of K in the aged brine is consistent with the dissolution
of K-feldspar and illite, as reported in a number of other studies and discussed in Section 1
(Introduction). The increase in Mg concentration is consistent with dissolution of illite-smectite,
illite and potentially dolomite. The increase in Ca
2+
concentration may be caused by the dissolution
of illite-smectite, dolomite, ankerite and plagioclase, while the decrease in the SO4
2-
concentration,
may be caused by secondary precipitation of CaSO4.
3.2.4 Pore Size Distribution
For analyzing the pore structure characteristics of both the fresh and the aged samples, we utilized
here two different techniques: Flow perporometry and N2 adsorption (at 77 K = -196.15° C). These
88
two techniques are both capable of determining the pore size distribution (PSD) of the rock
samples, but typically probe two different regions of the pore space.
Flow perporometry is a technique that measures the so-called flow-through porosity of the
sample. It is based on the simple principle that when a pore is filled with a wetting liquid, the
pressure drop required to force gas through the pore is inversely proportional to the size of the
pore, and is specifically described by the Laplace equation below:
4 cos
D
p
(3.3)
where θ is the contact angle, D is the pore diameter, Δp is the pressure drop, and σ is the gas-liquid
interfacial tension.
The schematic of the flow perporometry experimental set-up is shown in Fig. 3.10. It
mainly consists of a pressure regulator, a sample holder and a data acquisition system. The rock
sample is placed in the sample holder in between two empty cells. The pressure of the top cell is
controlled by a pressure regulator and is monitored by a pressure transducer. The gas flow through
the sample, from the top to the bottom cell, is monitored by a series of mass flow meters (MFM’s),
each covering a narrow range of flow rates for added accuracy. The pressures of the top and bottom
cells and the gas flow rate exiting the bottom cell are all recorded by the data acquisition system
during the experiments.
89
Fig. 3.10 Schematic of the flow perporometry experimental set-up
First, one measures the so-called dry-curve. For that, Helium gas is pressurized in the top
cell at a certain pressure and flows through the sample, and the flow rate and the upstream and
downstream pressures (typically the downstream pressure remains atmospheric for the duration of
the test) are recorded as the upstream pressure is increased in steps. Then, one generates the wet-
curve. For that, the rock sample is taken out of the flow perporometry set-up and is completely
saturated with a wetting fluid (n-butanol, for the tests in this research,) and is then re-installed in
the apparatus. The top cell is again pressurized with Helium, but the liquid in the pores now acts
as a barrier, and no flow of gas occurs until the applied pressure reaches the capillary pressure of
the largest pores (see Eqn. 3.3 above). As the pressure is increased step-wise, more liquid is
expelled from progressively smaller and smaller pores, thus allowing more gas to permeate
through, which is measured and recorded for generating the sample’s wet-curve. Though, in
principle, the technique can access even the smallest of pores, practical consideration dictated by
the mechanical strength of the sample itself place a lower limit, which for the materials investigated
90
here is, typically, a few tenths of a micron. There are a number of methods for generating the PSD
from the experimental dry-curve and wet-curve data, ranging from the rudimentary to the more
involved 3-D pore-network models (e.g., Mourhatch et al., 2011). Since the goal here is to use the
flow perporometry technique to detect changes in the pore structure of the Mt. Simon formation
core samples upon exposure to CO2/brine mixtures, the well-accepted ASTM method (ASTM
D6767-16) is employed to analyze the experimental data and to derive the PSD. The fractional
pore size distribution can be obtained via Eqn. 3.4
__
1 100%
__
wet flow rate
Finer
dry flow rate
(3.4)
Nitrogen physisorption is another technique, known commonly as the BET method (e.g.,
Nia et al., 2016), for characterizing porous materials. In contrast to flow perporometry, the BET
technique accesses both the flow-through and dead-end pores. In addition, the BET method is
thought to access smaller pores, typically less than 0.1 µ m, but most likely less than 0.05 µ m.
The experimental data from the flow perporometry tests for core sample 6927-A2 and
6925-A2 are reported in Fig. 3.11 and 3.12, respectively (left figure for the fresh core and right
figure after the core was exposed to brine for 2 weeks, see Table 3.1). The dry and wet curves for
this sample are reported in terms of gas flow (mol/s) divided by the pressure drop (KPa) across the
sample (a quantity proportional to the gas permeance of the sample) versus the average of the
upstream and downstream pressures. From Fig. 3.11, one observes that the wet-curves become
linear after a certain pressure and do not meet the dry-curves. This is, typically, associated with
irreducible saturation phenomena whereby some of the liquid remains in small pores because the
maximum pressure applied is not sufficiently large to overcome the capillary pressure in these
pores. This then defines the indeterminate region of porosity where the pore sizes cannot be
91
calculated: An upper bound on the pressure one can employ is imposed by the mechanical strength
of the material.
0 200 400 600 800 1000
0.0
2.0E-7
4.0E-7
6.0E-7
8.0E-7
1.0E-6
1.2E-6
Dry curve
Wet curve
Flow / pressure drop (mol/s/KPa)
Average Pressure (KPa)
0 100 200 300 400 500
0.0
5.0E-7
1.0E-6
1.5E-6
2.0E-6
2.5E-6
3.0E-6
3.5E-6
4.0E-6
Dry curve
Wet curve
Flow / pressure drop (mol/s/KPa)
Average Pressure (KPa)
Fig. 3.11 Experimental data of the flow perporometry test for sample 6927-A2 before (left) and
after (right) exposure
0 100 200 300 400 500
0.0
5.0E-7
1.0E-6
1.5E-6
2.0E-6
2.5E-6
3.0E-6
3.5E-6
Dry curve
Wet curve
Flow / pressure drop (mol/s/KPa)
Average Pressure (KPa)
0 100 200 300 400
0.0
2.0E-6
4.0E-6
6.0E-6
8.0E-6
1.0E-5
Dry curve
Wet curve
Flow / pressure drop (mol/s/KPa)
Average Pressure (KPa)
Fig. 3.12 Experimental data of the flow perporometry test for sample 6925-A2 before (left) and
after (right) exposure
92
According to Fig. 3.11, the permeance for both the dry and wet sample 6927-A2 increases
after exposure by a factor of ~ 4 for this sample. For sample 6925-A2, shown in Fig. 3.12, the
changes in the dry curve permeance are also substantial (a factor of ~ 2). This observation is
consistent with our observations with sample 6927-B4, see Table 3.4. The experimental data in
Figs. 3.11 and 3.12 are analyzed by the ASTM method (ASTM D6767-16) to generate the pore
size distribution (known as the finer curve) of samples, as shown in Figs. 3.13 and 3.14. The
smallest detectable pore size, via this technique, is limited by both by the mechanical strength of
the sample and the upper limit of the mass flow meter(s). (Due to the small thickness of the samples
investigated here, the maximum differential pressure was set to 1.7 MPa, while the maximum flow
rate was 3000 sccm).
According to Figs. 3.13 and 3.14, the detectable pore range of 6927-A2 and 6925-A2
become wider after exposure. The broadening in PSD is consistent with our porosity measurements
in Section 3.2.2 that show that the porosities of all the samples increase after exposure to CO2/brine
accompanied by increases in permeability, which are indicative of dissolution occurring at the pore
throats and/or establishment of connections to additional previously inaccessible porosities.
93
7 6 5 4 3 2 1 0
0
10
20
30
40
50
60
70
80
90
100
PSD before aging
PSD after aging
Percent Finer (%)
Pore Diameter (um)
Fig. 3.13 PSD of sample 6927-A2 before and after aging
35 30 25 20 15 10 5 0
0
10
20
30
40
50
60
70
80
90
100
PSD before aging
PSD after aging
Percent Finer (%)
Pore diameter (um)
Fig. 3.14 PSD of sample 6925-A2 before and after aging
94
In addition to the flow perporometry experiments, N2 adsorption tests (BET) were
performed to probe the mesoporous range. The BET analysis of sample 6927-C1 is shown in Table
3.9 (The N2 adsorption isotherms can be found in the Appendix E). From Table 3.9, we observe
that the surface area and the mesopore volume are reduced by 53.5% and 56.6%, respectively. The
detectable pore range using the N2 adsorption technique is mesoporous, which for the sandstone
cores studied here mainly exists in the clays. The decrease in the mesopore surface area and pore
volume may indicate dissolution (removal) of the clays during the aging process. We note, that
clays are the main constituents of the cement that consolidates (binds together) the quartz particles,
and their dissolution may thus result in a reduction of the mechanical strength of the rock samples.
This observation is also in agreement with the changes observed with the flow perporometry tests
after incubation: the permeance increases and wider pore range is detected due potentially to the
formation of macropores and/or micro fractures along grain boundaries. The pore size distributions
extracted from the BET analysis are reported in Fig. 3.15. The results show a unimodal pore size
distribution (in the mesopore region) for this illite and illite-smectite rich samples, similarly with
Kuila and Prasad (2013) who also report a unimodal PSD for an (illite + smectite) rich shale sample
using the N2 adsorption technique. From Fig. 3.15, we observe a significant reduction in the pore
volume attributed to pores between 4 nm to 100 nm. The reduction in pore volume from this pore-
size-range is again consistent with the dissolution (or destabilization) of clays due to the exposure
to CO2/brine.
Table 3.9 BET analysis of sample 6927-C1
Fresh Aged
BET surface area 3.07 m² /g 1.43 m² /g
BJH mesopore volume 0.0081 cm³ /g 0.0035 cm³ /g
95
1 10 100
0.0000
0.0001
0.0002
0.0003
0.0004
0.0005
0.0006
PSD before aging
PSD after aging
Pore Volume (cm
3
/g/nm)
Pore Diameter (nm)
Fig. 3.15 PSD extracted from the N2 adsorption tests
3.2.5 Scanning Electron Microscopy
For one of the samples (6927-C1), we also employed scanning electron microscopy (SEM) to
image the same (flat end) surface before and after CO2/brine flow-through experiments had been
completed employing the sample, with the goal of directly visualizing pore formation and/or loss
during the experiment. These tests were carried out using a FEI Quanta 600 FEG instrument,
coupled with an energy dispersive spectroscopy (EDS), at the National Energy Technology
Laboratory (NETL) in Pittsburgh.
SEM images of sample 6927-C1 before and after the CO2/brine flow-through experiments
are shown in Fig. 3.16. From Fig. 3.16-a (for the sample before the flow-through experiments), we
observe that the quartz grains are surrounded by clays: The clays potentially play a central role in
the cementation of the grains. Some pores exist in the interspace between grains. Most of the pores
96
in Fig. 3.16-a are less than 10 µ m in size except a larger “fracture” at the top left corner, as
indicated by a yellow circle. Fig. 3.16-c (for the same sample at a higher magnification) shows the
clay crystals lining the pores in between the larger quartz particles. Most of pores in Fig. 3.16-c
are less than 2 µ m an observation that is consistent with our flow perporometry results for the fresh
sample 6927-A2 whose pore size lies in the range of 0.4 to 1.85 µ m. (Since the flow perporometry
technique only probes the connected pores, its detection range can be smaller than the pores we
observed from the 2D SEM image). Figs. 3.16-b and 3.16-d show the images of the aged sample
6927-C1. By comparing Figs. 3.16-a and 3.16-b, we observe some enlargement and formation
(due to clay removal) of pores after exposure to CO2/brine. Figs. 3.16-c and 3.16-d show this
change more clearly: The clay deposits in the red circles disappear after exposure and a larger pore
is uncovered in Fig. 3.16-d.
The removal of clay was also detected in the N2 adsorption tests (see Sec. 3.2.4).
Furthermore, alterations observed from the SEM analysis, due to the exposure to CO 2/brine, is
consistent with results from the flow perporometry test, where a wider pore size distribution is
observed after incubation (from 0.5-2 µ m to 1-6 µ m). In addition to the clay dissolution, a
compaction/rearrangement of larger grains is observed from Figs. 3.16-a and 3.16-b (indicated by
a yellow circle). The sample (6927-C1) shown in Fig. 3.16 was exposed to CO2/brine in a flow-
through experiment with a net stress of 2.78 MPa. The confining environment may result in
compaction of the grains, that is visually detectable from Fig. 3.16-a and 3.16-b. Such compaction
is further supported by the irreversible nature of the mechanical testing (see Section 4 below).
97
a b
c d
Fig. 3.16 SEM images of sample 6927-C1 before (a and c) and after (b and d) the flow-through
CO2/brine experiments
Complimentary EDX investigations were also carried out to identify potential chemical
alterations of the surface as a result of core exposing to the CO 2/brine environment. For the EDX
analysis, the following elements were measured: Ca, Mg, Al, Fe, Si, Sr, K, Na, I, P, S, O, Cl, and
98
Br. 8 points in Fig. 3.17 were selected to do an EDX analysis. In the left SEM image in Fig. 3.17,
(for the fresh sample) the points selected for analysis are noted as points 1 to 8 while in the right
image (for the aged sample) in the right figure Fig. 3.17 the same points are now indicated as
points 26 to 33.
Fig. 3.17 Selected points for EDX analysis. Left image fresh sample, right image aged sample
The EDX spectra for points 1 (fresh sample) and corresponding point 26 (aged sample) are
shown in Fig. 3.18 for illustration purposes (the EDX spectra for other points can be found in
Appendix F). Table 3.10 shows the elemental concentrations for the eight spots for the fresh
sample, while Table 3.11 shows the corresponding elemental concentrations for the same spots on
the aged sample. Table 3.12 shows the changes (%) in the elemental concentrations between the
fresh and the aged sample.
99
Fig. 3.18 EDX results of point 1 (fresh) and point 26 (aged)
Table 3.10 Elemental concentrations (Wt%) for the fresh sample
Element 1 2 3 4 5 6 7 8
O 62.33 39.99 31.68 44.44 50.23 52.19 49.87 55.85
Al 18.85 12.95 8.98 14.49 13.15 14.86 5.04
Si 4.06 27.27 45.64 34.14 23.31 23.42 23.45 33.41
P 9.77
S 2.31
K 1.06 12.23 8.33 12.44 6.91 6.52 7.15 4.21
Ca 1.01 1.92
Mg 1.11 1.22
Fe 3.41 3.85 2.31 2.34 1.5
Sr
Na
Cl 2.22 1.22 1.3 1.12
I
Br 14.36
Table 3.11 Elemental concentrations (Wt%) for the aged sample
Element 26 27 28 29 30 31 32 33
O 58.16 41.76 36.26 13.89 20.22 25.62 45.36 46.82
Al 15.09 15.13 6.56 11.69 13.62 11.97 15.05
Si 7.09 8.49 50.93 55.04 33.78 16.2 18.09 23.34
P 7.09
100
S 1.57
K 1.77 1.36 6.25 2.71 9.94 12.34 5.27 10.6
Ca 4.42 4.32 2.23
Mg 1.42
Fe 4.25 16.72 21.25 9.27
Sr 8.08
Na 1.66 3.33 6.92
Cl 1.15 14.4 2.03 3.11 15.32 12.39 1.97
I 7.12
Br 9.62
Table 3.12 Changes (%) in the elemental concentrations between the fresh and the aged sample
Element 1/26 2/27 3/28 4/29 5/30 6/31 7/32 8/33
O -6.69 4.43 14.46 -68.74 -59.75 -50.91 -9.04 -16.17
Al -19.95 16.83 -100 -19.32 3.57 -19.45 198.61
Si 74.63 -68.87 11.59 61.22 44.92 -30.83 -22.86 -30.14
P -27.43
S -32.03
K 66.98 -88.88 -24.97 -78.22 43.85 89.26 -26.29 151.78
Ca -100 130.21
Mg -100 -100
Fe 24.63 451.95 301.30 -100 -100
Sr
Na
Cl 548.65 154.92 1078.46 1006.25
I
Br -100
According to Table 3.10, the point 1 has the least content of Si at fresh condition. This
means the minerals located at point 1 contain least quartz, and should be the most reactive point
among the tested points. From Table 3.12 we can find that the concentration of Al and Ca decrease
while that of K increases at point 1 after exposure. The decrease in Al and Ca concentrations could
be caused by dissolution of K-feldspar, plagioclase and illite/smectite. The concentration of K
increases by 66.98%. Since the sample was dried out after aging without rinse, the increase in K
101
could be due to precipitation of the salts in the brine or the aging by-product. However, the weight
percent of K in brine recipe is only 1 Wt% that is less than the value of 1.06 Wt% repoeted in
Table 3.10. Therefore, the increase of K concentration can be caused by secondary precipitation
of illite. For the point 2/27, the concentration of Si and K decrease significantly. It is consistent
with dissolution of K-feldspar, plagioclase. The mild increase in concentration of Al is consistent
with illite dissolution. The concentration of Ca increases by 130.21%, and could be caused by
precipitation of salts in the brine because the weight percent of Ca in brine recipe is 10.9 Wt% that
is higher than the value of 4.42 Wt% reported in Table 3.11. For the point 3/28, the mild increase
in Si and decrease in K is consistent with dissolution of K-feldspar and secondary precipitation of
illite. The decrease of Al and K for point 4/29 can be due to dissolution of K-feldspar, plagioclase
and illite/smectite. For the point 5/30, the concentration of Al decreases while that of K increases,
which is consistent with dissolution of plagioclase and illite/smectite and precipitation of illite. For
the point 6/31, the increase in Al and K can be due to precipitation of illite, while the decrease in
Mg is consistent with dissolution of illite/smectite. The decrease in concertation of Al, K and Mg
for the point 7/32 is consistent with dissolution of K-feldspar, plagioclase and illite/smectite. For
the point 8/33, the significant increase in Al and K can be caused by precipitation of illite. As
previously indicated, samples were not rinsed after incubation to avoid any additional
destabilization of clays. The significant increase in the concentration of Cl could be due to salt
precipitation.
3.3 Discussion and Conclusions
The experimental observations presented in the previous sections in this Chapter show a significant
change in transport properties for the Mt. Simon Sandstone, when exposed to CO2/brine over a
102
relatively short period of time (1-2 weeks). The porosity for all the samples also increases after
exposure to CO2/brine but the changes are rather small. Similar size changes in porosity were
reported by Rimmelé et al. (2010), Nover et al. (2013) and Mouzakis et al. (2016), who attributed
the increase in porosity to the dissolution of mineral phases. The increase in permeability observed
for our samples correlates well with the results reported by Xie et al. (2011), Nover et al. (2013)
and Huq et al. (2015). In contrast, Yu et al. (2012) and Tutolo et al. (2015) reported a decrease in
permeability and explain the reduction via precipitation of new mineral phases. The analysis of the
brine composition (as shown in Table 3.5) before and after aging shows an increase in the
concentration of K, Mg and Ca, and provides direct evidence of some mineral/clay dissolution,
potentially, accounting for the increase in porosity and permeability. According to the mineral
composition of the samples (as reported in Table 3.8), the dissolution of K-feldspar, dolomite,
ankerite, plagioclase and clay minerals may account for the increase in the concentration of these
cations. Some recent studies on the evolution of mineralogy caused by CO2-brine-rock interaction
are consistent with our findings: Szabó et al. (2016) and Wang et al. (2013) reported dissolution
of dolomite when caprock/dolomite samples were exposed to a CO2 environment. Nover et al.
(2013) reported dissolution of K-feldspar and clay for cores exposed to CO2/brine. Mineral
dissolution can account for the increase in porosity, while secondary mineral precipitation inside
the pore space may decrease the porosity. However, in this study, all porosities increase after
exposure (Table 3.4), perhaps suggesting that for these rock samples and the relatively short period
of exposure to brine/CO2 mixtures, mineral dissolution dominates over secondary precipitation,
resulting in an increase in porosities. Because of the opening of the pore throats due to mineral/clay
dissolution, the permeability of the cores also increases after exposure.
103
The results of flow perporometry tests demonstrate an increase in the pore sizes that
contribute to flow after incubation. Consistent with our observation, Rimmelé et al. (2010) report
increases in pore sizes for Adamswiller sandstone after exposure to CO2-saturated water measured
by the MIP (Mercury Intrusion Porosimetry) technique. The analysis of N2 adsorption experiments
(BET) provides a measure of the pore size distribution on the mesoporous range. The observed
decrease in surface area and pore volume indicates loss of clay minerals due to the CO2/brine/rock
interactions. With clays playing a central role in the cementation of the larger (primarily) quartz
grains, the loss of clay minerals is likely to weaken the strength of sandstone and to provide
additional connectivity of the rock samples as indicated by the flow perporometry results. The
SEM images show an enlargement and formation/uncovering of pores after exposure of the rocks
to CO2/brine and provide visual evidence of clay dissolution. These observations are consistent
with the results of flow perporometry and N2 adsorption experiments. The compaction of larger
flow conduits, as evident from Figs. 3.16-a and 3.16-b, is attributed to the applied confining
environment and is consistent with the decrease in permeability observed with increasing net stress.
The EDX results are partially consistent with the brine composition analysis. In the brine
composition analysis, we conclude that the change of the brine composition after aging can be
explained by dissolution of K-feldspar, plagioclase and clay minerals. By comparing the EDX
result for the 8 points on the surface before and after exposure, we can see some results that is
consistent with dissolution of K-feldspar, plagioclase and illite/smectite. However,it seems like
the secondary precipitation of illte is observed in the EDX result which is not revealed in the brine
composition analysis. One should note that the weight percent of Ca in the brine recipe is 10.9 Wt%
that is higher than all the reported values in Tables 3.10 and 3.11. Therefore, the increase in Ca
concentration after exposure could be significantly affected by the precipitation of the salts in the
104
brine. The weight percent of Mg and K are both 1 Wt% and much less than all the values reported
in Tables 3.10 and 3.11. So the change of the Mg and K concentrations can be caused by
geochemical alteration. .
105
Chapter 4: CO2/Brine Induced Evolution of the Mechanical
Properties of the Mt. Simon Sandstone
3
4.1 Introduction and Literature Review
During the mineral trapping stage in geological carbon sequestration, mineral dissolution typically
occurs, as discussed in Chapter 3, which may result in a change of the mechanical properties of
the host rock. It is important, therefore, to further investigate the impact of exposure of rock
samples to CO2/brine on their mechanical properties. Better understanding of such behavior is
critical to ensure the integrity of the storage system.
There have been a number of studies to date, mostly with model systems, focusing on the
impact on mechanical properties of exposing various minerals to CO2. De Jong et al. (2014)
performed unconfined volumetric strain (dilatometry) measurements on compacted pellets of
MMT (Wyoming SWy-1 and Na-SWy-2 type smectites) upon exposure to CO2 in a high-pressure
optical cell (45° C, and CO2 pressures up to 15 MPa). The samples were heat-treated prior to
exposure to CO2 and their initial d001-spacing was determined via XRD. The SWy-1 and Na-SWy-
2 MMT both swell almost instantaneously (in a few sec) to an equilibrium state when exposed to
scCO2. Swelling strain increases systematically with CO2 pressure in the range 1–7 MPa, but only
slightly at higher pressures. The authors note that for smectite-bearing caprocks sealing carbon
sequestration reservoirs, these results imply that CO2 penetration can cause swelling, closing small
fractures or joints and thus reducing bulk permeability.
3
Some results in this chapter has already been published (Shi et al., 2019)
106
Zhang et al. (2017a,b) studied via molecular dynamics (MD) simulations the change in
elastic properties (including the elastic coefficients Cij, the bulk modulus K, the shear modulus G,
the Young’s modulus Eii, and the Poisson’s ratio Eij) of Ca-MMT exposed to CO2–H2O mixtures.
They report that the in-plane elastic constants, bulk modulus, shear modulus, and Young’s
modulus E33 correlated with the basal d001 spacing of the clay, with no such correlation found for
the Poisson’s ratio. Exposure of the MMT to the CO2/H2O mixture caused a decrease in the
strength of clay mineral. The same group also studied the elastic stiffness and hydrostatic
mechanical behavior of Na-MMT upon exposure to the CO2/H2O mixture. In the hydrostatic test
simulation of 1 W and 2 W hydration state of MMT, d001 linearly decreased with the increasing
hydrostatic pressure, with higher hydrated states being shown to be softer structures.
CO2 intercalation strongly affects the out-of-plane elastic stiffness, while has little influence on
the in-plane stiffness.
The effects of CO2-water-rock interactions on the mechanical properties of shale were
investigated by Lyu et al. (2016a). In their study, uniaxial compressive strength (UCS) tests
together with acoustic emission (AE) and SEM and EDS analysis were performed to investigate
the mechanical properties and microstructural changes of black shales exposed to CO2 (gaseous
and supercritical)-water mixtures for various periods (10, 20 and 30 days). The UCS, Young’s
modulus and brittleness index decreased gradually with increasing exposure time, and the impact
is greater for the scCO2-water case. AE analysis indicates that longer exposure times are
characterized by higher peak cumulative AE energy (Lyu et al. 2016b). SEM analysis reveals the
formation of new pores in the shale samples upon exposure to gaseous/scCO2-water mixtures,
while EDS indicates that samples that have been exposed contain on their surface more Ca and Fe
and less Al and K than unexposed samples. Hu et al. (2017) studied the strength, elasticity modulus
107
and brittleness of sandstones in contact with single fluids (CO2/N2/H2O) as well as a (CO2-H2O)
fluid mixture. The linear elastic modulus of the (CO2-H2O)-coupled sandstone was slightly greater
than that of H2O-bearing sandstone, but significantly less than that of CO2- or N2-bearing
sandstone. The latter are mainly subjected to the brittle failure and exhibit no obvious plastic
deformation.
Espinoza and Santamarina (2012) carried out sedimentation experiments to explore the
response of kaolinite and MMT to DI water, brine, liquid CO2 and scCO2. Both MMT and kaolinite
aggregate when submerged in CO2 and the final porosity in CO2 is smaller than that in brine.
Capillary effects induced by the water–CO2 interface were studied by clay–water paste desiccation
experiments conducted using scCO2: water dissolution into the surrounding CO2 causes suction
and capillary contraction, the invasion of the CO2–water interface into the sediment, and the
formation of desiccation cracks. Volume contraction and crack initiation are consistent with the
sediment response within an effective stress framework. These studies indicate that changing
capillary forces may lead to coupled chemo-hydro-mechanical phenomena in seal layers that could
facilitate CO2 breakthrough and advection through high porosity caprocks. Bemer and Lombard
(2010) conducted triaxial tests on the carbonate formations to study the acidification effect induced
by CO2 injection on their geomechanical properties and observed weakening due to chemical
alteration. Xie et al. (2011) and Zinsmeister et al. (2011) conducted triaxial tests on limestone and
exposed the rock sample to strong acid and CO2 dissolved in brine, respectively, and both of them
observed decreases in the elastic modulus of limestone.
Rimmele et al. (2010), Marbler et al. (2013) and Rathnaweera et al. (2015) conducted both
mineralogical and geomechanical characterization of a sandstone before and after exposure to
CO2/brine (water) to study the CO2-induced effect on its mechanical behavior. Marbler et al. (2013)
108
and Rathnaweera et al. (2015) found reduced strength and mineralogical alteration while Rimmele
et al. (2010) report that the mechanical and mineralogical properties are maintained after exposure
to CO2-saturated water despite an increase in porosity and permeability.
Though there have been a number of studies on the CO2-induced alteration of the
mechanical properties of rock samples, information concerning the evolution of the mechanical
behavior of the Mt. Simon Sandstone due to the fluid-rock interaction is limited. A fundamental
understanding of the interaction of CO2,-saturated in situ brine and the specific rock in the storage
zone is critical to evaluate the feasibility and safety of the IBDP CO2 sequestration project. In this
Chapter, the mechanical properties (Young’s Modulus, Bulk Modulus and Poisson’s Ratio) of the
fresh and aged Mt. Simon Sandstone samples were characterized by tri-axial mechanical tests. The
comparison between the results with fresh and aged samples provides valuable information on the
impact on the mechanical properties of the Mt. Simon Sandstone of exposure to CO2-saturated
brine.
4.2 Experimental Approach
Samples 6927-B2 and 6927-B4 (see Table 3.1) were selected for the mechanical tests. They were
generated by drilling in the horizontal direction (perpendicular to the main core axis) from
approximately the same vertical location within the 1-ft 6927 section of the original Mt. Simon
Sandstone core (2 cm in depth from each other). Their porosities measured (prior to the core
sample 6927-B4 being exposed to the CO2/brine environment for 14 days) were 14.7% and 14.8%,
respectively (see Table 3.3), and their bulk densities were 2.239 g/cm
3
and 2.233 g/cm
3
,
respectively. Their permeabilities measured before exposure of core 6927-B4 to CO2/brine are
109
reported in Table 3.4, and are also fairly close to each other (~7.5% difference). Note, that when
we tested core 6927-B1, which was also drilled from a nearby location (2 cm in depth from 6927-
B2), the difference between its permeability of 6927-B1 and that of core 6927-B2 was only ~8.1%.
This lends added credence to the assessment and premise that all these cores are similar geological
materials, and therefore one can use them to examine the impact that exposure to CO2/brine will
have on mechanical properties, i.e., by exposing core 6927-B4 to CO2/brine and comparing its
mechanical properties with those of core 6927-B2 – see further discussion to follow on the
mechanical property tests with these two cores.
For the measurement of mechanical or petrophysical properties (of samples 6927-B2 and
6927-B4) we utilized a NER AutoLab 1500 system (see Fig.4.1), which is capable of creating a
wide range of both lithostatic (confining) pressures and pore pressures in order to investigate the
variation and response of dynamic material properties to realistic formation conditions. For the
measurement of the mechanical properties, the instrument employs an ultrasonic system which
generates ultrasonic pressure and shear pulses (P and S waves) through a plate at one end of the
core and records the response at the opposing end. Dynamic values of the Young’s Modulus,
Poisson’s Ratio, Shear Modulus and Bulk Modulus are then determined from knowledge of the
density, shear and compressional wave velocities using Eqns. (4.1) to (4.4) below (Spaulding et
al., 2015):
2 2 2
22
34
s p s
ps
V V V
E
VV
(4.1)
2
2
12
21
sp
sp
VV
VV
(4.2)
110
22
E
(4.3)
3(1 2 )
E
K
(4.4)
where E = Young’s Modulus, ρ = Bulk density, Vs = Shear wave velocity, Vp = Compression wave
velocity, v = Poisson’s Ratio, K = Bulk Modulus, µ = Shear Molulus.
Of the two 1”×2” core samples we tested in the AutoLab system, core 6927-B2 was tested
without being exposed to CO2/brine (prior to testing) while the other core (6927-B4) was tested
after being exposed to the CO2/brine for two weeks. The comparison of the results between these
two cores (6927-B2 and 6927-B4) extracted from the same vertical position in the formation
provide valuable information about the impact of the exposure to CO2/brine on the mechanical
properties of the Mt. Simon sample. The strategy of this measurement is shown in Fig. 4.2. In our
measurement, we kept the pore pressure constant and raise the confining pressure in steps. We
carried out dynamic measurements (while maintaining the core temperature at 50⁰C and the pore
pressure at 8 MPa) that involved first increasing the confining pressure from 10 MPa to 28 MPa
step-wise in increments of 3 MPa, and then decreasing the confining pressure from 28 MPa to 10
MPa, again in increments of 3 MPa. Since the pore pressure during testing was held constant at 8
MPa, ramping the confining pressure from 10 MPa to 28 MPa is equivalent to changing the applied
effective (confining minus pore) pressure/stress on the core material from 2 MPa to 16 MPa. The
reason for cycling the pressure first up and then down is so that we observe any potential hysteresis.
In order to observe any potential irreversible changes that may occur, the dynamic mechanical
property measurement protocol was then repeated a second time.
111
Fig. 4.1 Photo of NER AutoLab 1500 system
0 50 100 150 200 250 300
10
20
30
Pressure (MPa)
Time (min)
Confining Pressure
Pore Pressure
Fig. 4.2 Strategy of measurement of mechanical properties
112
4.3 Experimental Results
The shear and compressional wave velocities were measured via the AutoLab instrument from
which the dynamic Young’s Modulus, Poisson’s Ratio, Shear Modulus and Bulk Modulus for the
cores were calculated using Eqns. (4.1) to (4.4). The results for the sample 6927-B2 (fresh) and
6927-B4 (aged) are reported in Table 4.1 and 4.2 respectively as a function of the effective pressure
in MPa. The effective pressure shown is simply the confining pressure minus the pore pressure,
at each measurement step.
Table 4.1 Results of the dynamic mechanical properties for sample 6927-B2 (fresh)
Effective pressure
(Mpa)
Young's Modulus
(GPa)
Poisson's
Ratio
Shear Modulus
(GPa)
Bulk Modulus
(GPa)
2.7 13.99 0.052 6.65 5.21
5.7 16.93 0.124 7.53 7.50
8.7 19.59 0.154 8.49 9.43
11.6 21.54 0.156 9.32 10.43
14.7 23.43 0.167 10.04 11.72
17.6 25.09 0.180 10.63 13.08
20.6 26.37 0.168 11.29 13.22
17.7 25.65 0.164 11.02 12.73
14.7 24.68 0.166 10.58 12.32
11.6 23.06 0.159 9.95 11.26
8.7 21.51 0.172 9.18 10.91
5.7 18.78 0.140 8.24 8.69
2.7 13.60 -0.014 6.89 4.41
5.7 17.86 0.138 7.85 8.22
8.7 20.07 0.157 8.67 9.76
11.6 22.05 0.160 9.50 10.81
14.7 24.19 0.180 10.25 12.59
17.7 25.61 0.175 10.89 13.15
20.6 26.58 0.169 11.37 13.39
17.6 26.19 0.183 11.07 13.79
14.7 24.94 0.171 10.65 12.63
11.6 23.49 0.171 10.03 11.88
8.6 21.54 0.155 9.32 10.41
5.7 19.04 0.156 8.24 9.21
113
2.7 15.33 0.128 6.80 6.87
Table 4.2 Results of the dynamic mechanical properties for sample 6927-B4 (aged)
Effective pressure
(Mpa)
Young's Modulus
(GPa)
Poisson's
Ratio
Shear Modulus
(GPa)
Bulk Modulus
(GPa)
2.7 13.22 0.048 6.31 4.88
5.7 15.84 0.119 7.08 6.93
8.7 18.13 0.144 7.92 8.48
11.6 20.29 0.155 8.78 9.81
14.7 22.00 0.161 9.47 10.81
17.6 23.75 0.174 10.12 12.15
20.6 25.29 0.174 10.77 12.91
17.7 24.48 0.159 10.56 11.95
14.7 23.68 0.161 10.20 11.66
11.6 22.45 0.154 9.73 10.82
8.7 20.71 0.148 9.02 9.79
5.7 18.65 0.142 8.17 8.67
2.7 15.39 0.105 6.96 6.50
5.7 17.56 0.151 7.63 8.38
8.7 19.60 0.163 8.43 9.69
11.6 21.20 0.154 9.19 10.20
14.7 22.74 0.156 9.83 11.03
17.7 24.49 0.174 10.43 12.53
20.6 25.25 0.158 10.90 12.32
17.6 25.01 0.169 10.70 12.59
14.7 23.83 0.154 10.32 11.49
11.6 22.72 0.158 9.81 11.06
8.6 20.93 0.152 9.08 10.02
5.7 18.87 0.143 8.25 8.82
2.7 15.77 0.126 7.00 7.03
From Table 4.1 and 4.2 we note that the values of Poisson's Ratio measured at effective
pressure of 2.7 Mpa are significantly lower than the values measured at other effective pressures.
This could be caused by the untight contact between the wave emission end and the end surface of
the core sample at such low net stress. Therefore, the abnormal values caused by the experimental
114
error will be ignored when we analyze the Poisson's Ratio data. The results for Young’s Modulus
of cores 6927-B2 (fresh) and 6927-B4 (aged) are plotted in Fig. 4.3 and 4.4 respectively.
2 4 6 8 10 12 14 16 18 20 22
12
14
16
18
20
22
24
26
28
1st cycle up
1st cycle down
2nd cycle up
2nd cycle down
Young's Modulus (GPa)
Effective Pressure (MPa)
Fig. 4.3 Dynamic Young’s Modulus of Core 6927-B2 (fresh)
115
2 4 6 8 10 12 14 16 18 20 22
12
14
16
18
20
22
24
26
1st cycle up
1st cycle down
2nd cycle up
2nd cycle down
Young's Modulus (GPa)
Effective Pressure (MPa)
Fig. 4.4 Dynamic Young’s Modulus of Core 6927-B4 (aged)
According to Figs. 4.3 and 4.4, both cores exhibit a similar behavior, with the Young’s
modulus increasing monotonically with effective stress (confining pressure minus pore pressure).
In other words, the stiffness of these Mt. Simon cores increases with effective pressure. We observe
from Figs. 4.3 and 4.4 that as the effective pressure increases, the net stress acting on the rock
samples increases. This, in turn, causes some of the compliant or penny shaped cracks and voids
in the sample to close, or decrease in size, and as a result the Young’s modulus will increase. It is
well known, that the presence of cracks and other void space in a material reduces Young’s
modulus, because stress cannot be transferred across the void space itself (Fjær et al., 2008).
Similar results have been reported previously for measurement of the Young’s modulus on shale
and cement samples (Spaulding et al. 2015, Villamor Lora et al., 2016). We note that the bulk
density used to calculate the dynamic Young’s modulus, as shown in Eqn. (4.1), is treated as a
constant and is evaluated at zero net stress. In fact, the pore volume may be reduced due to elevated
116
effective pressure during the test. According to Table 3.3, the porosities of samples 6927-B2 and
6927-B4 are around 15%, which means that the maximum change in Young’s modulus (if all
porosity was to be lost due to compaction) due to changes (increase) in the bulk density would be
in the same range, ~ 15%. However, the Young’s modulus increases by ~90% with increasing
effective pressure for both samples, as indicated by Figs. 4.3 and 4.4. Therefore, we conclude that
the effect of the change in bulk density on Young’s modulus is not significant compared to the
change caused by the increase in the effective pressure.
We note that hysteresis is observed in the Young’s modulus between loading and unloading
for both test cycles (broken lines added to guide the eye for the first cycle in Figs. 4.3 and 4.4).
This can be explained by the fact that an increase in the effective pressure induces both elastic and
inelastic strain, and that the unloading cycle does not provide sufficient time for the inelastic strain
to recover. Furthermore, at the end of the first loading/unloading cycle, the dynamic Young’s
modulus does not return to its original value (this is more obvious in Fig. 4.4 for the aged sample
6927-B4) signifying, potentially, that irreversible changes (plastic deformation) to the core’s pore
structure have occurred (consistent with observations from SEM images, see Fig. 3.16).
Comparing to Fig. 4.3 for the fresh sample, the aged sample indeed became weaker after exposing
to CO2/brine for 2 weeks.
Fig. 4.5 compares the dynamic Young’s modulus during the 1
st
loading process for core
6927-B2 and its equivalent aged counterpart (core 6927-B4). The dynamic Young’s modulus of
the aged sample is clearly lower than that of its fresh counterpart for the whole range of applied
effective pressures. The average relative difference is -5.8%. This observation is consistent with
the impacts that aging in CO2/brine has on the porosity (Table 3.3) and the permeability (Table
3.4) of these materials.
117
2 4 6 8 10 12 14 16 18 20 22
12
14
16
18
20
22
24
26
28
6927-B2
6927-B4
Young's Modulus (GPa)
Effective Pressure (MPa)
Fig. 4.5 Dynamic Young’s modulus measured during the 1
st
loading process for the fresh
and aged samples
The Poisson’s ratio values of cores 6927-B2 (fresh) and 6927-B4 (aged) are shown in Figs.
4.6 and 4.7, respectively. There is, generally, more scatter in the experimental data. The Poisson’s
ratio of both samples roughly increased with effective pressure, but shows a weaker correlation
with pressure than the Young’s modulus. One notes that the variation of Poisson’s ratio is not very
significant after the 1
st
loading process. This may be attributed to the irreversible closure of cracks
during the 1
st
loading. The average values of the Poisson’s ratio for the fresh and aged sample is
0.162 and 0.156, respectively.
118
4 6 8 10 12 14 16 18 20 22
0.12
0.14
0.16
0.18
1st cycle up
1st cycle down
2nd cycle up
2nd cycle down
Poisson's Ratio
Effective Pressure (Mpa)
Fig. 4.6 Dynamic Poisson’s ratio of Core 6927-B2 (fresh)
4 6 8 10 12 14 16 18 20 22
0.12
0.14
0.16
0.18
1st cycle up
1st cycle down
2nd cycle up
2nd cycle down
Poisson's Ratio
Effective Pressure (Mpa)
Fig. 4.7 Dynamic Poisson’s ratio of Core 6927-B4 (aged)
119
The results for the Bulk modulus of cores 6927-B2 (fresh) and 6927-B4 (aged) are reported
in Figs. 4.8 and 4.9, respectively. Similarly to Young’s modulus, the dynamic Bulk modulus of
both cores shows a monotonic increase with effective pressure. This result is not surprising based
on Eqn. 4.4 that dictates a Bulk modulus proportional to Young’s modulus, if the Poisson’s ratio
is constant. From our measurements, the Poisson’s ratio shows a weak dependence on the effective
pressure, with average values for the fresh and aged samples being 0.162 and 0.156, respectively.
Therefore, the Bulk modulus shows a similar trend with that of the Young’s modulus. Hysteresis
in the Bulk modulus is also observed between the loading/unloading cycles. Fig. 4.10 compares
the Bulk modulus for core 6927-B2 (fresh) and its equivalent aged counterpart (core 6927-B4)
during the 1
st
loading cycle. Similarly to the Young’s modulus, the Bulk modulus of the aged
sample is clearly lower than that of its fresh counterpart, with an average relative difference of -
6.7%, for the whole range of applied effective pressures. The results for the Shear modulus of
cores 6927-B2 (fresh) and 6927-B4 (aged) are reported in Figs. 4.11 and 4.12 respectively.
According to Figs. 4.11 and 4.12, the performance of the Shear modulus is similar to that of the
Young’s modulus and Bulk modulus. The comparison of the Shear modulus for core 6927-B2
(fresh) and that of its equivalent aged counterpart (core 6927-B4) during the 1
st
loading cycle is
shown in Fig. 4.13. Again, the Shear modulus of the aged sample is lower than that of its fresh
counterpart, with an average relative difference of -5.5%.
120
2 4 6 8 10 12 14 16 18 20 22
4
6
8
10
12
14
1st cycle up
1st cycle down
2nd cycle up
2nd cycle down
Bulk Modulus (GPa)
Effective Pressure (MPa)
Fig. 4.8 Dynamic Bulk Modulus of Core 6927-B2 (fresh)
2 4 6 8 10 12 14 16 18 20 22
4
6
8
10
12
14
1st cycle up
1st cycle down
2nd cycle up
2nd cycle down
Bulk Modulus (GPa)
Effective Pressure (MPa)
Fig. 4.9 Dynamic Bulk Modulus of Core 6927-B4 (Aged)
121
2 4 6 8 10 12 14 16 18 20 22
4
6
8
10
12
14
6927-B2
6927-B4
Bulk Modulus (GPa)
Effective Pressure (MPa)
Fig. 4.10 Dynamic Bulk Modulus measured during the 1
st
loading stage for the fresh and
aged samples
2 4 6 8 10 12 14 16 18 20 22
6
8
10
12
1st cycle up
1st cycle down
2nd cycle up
2nd cycle down
Bulk Modulus (GPa)
Effective Pressure (Mpa)
Fig. 4.11 Dynamic Shear Modulus of Core 6927-B2 (fresh)
122
2 4 6 8 10 12 14 16 18 20 22
6
8
10
12
1st cycle up
1st cycle down
2nd cycle up
2nd cycle down
Bulk Modulus (GPa)
Effective Pressure (Mpa)
Fig. 4.12 Dynamic Shear Modulus of Core 6927-B4 (aged)
2 4 6 8 10 12 14 16 18 20 22
6
8
10
12
6927-B2
6927-B4
Bulk Modulus (GPa)
Effective Pressure (Mpa)
Fig. 4.13 Dynamic Shear Modulus measured during the 1
st
loading stage for the fresh and
aged samples
123
4.4 Impact of Mechanical Tests on Permeability
For the samples subjected to mechanical property testing via ultrasonic analysis in the Autolab set-
up, their permeabilities were measured after the mechanical testing was completed. For those
samples, the permeability was measured via N2 gas flow-through experiments using an Ultra-perm
500 gas permeameter, while increasing the confining pressure on the sample in a step-wise manner
from 2.76MPa to 18.62MPa. A schematic of this apparatus is shown in Fig. 4.14.
Nitrogen Tank
Pressure Gauge
Pressure Regulator
Core Holder
Mass Flow Meter
Hydraulic Pump
Fig. 4.14 The schematic of the experimental apparatus for permeability measurement after the
mechanical test
The results are shown in Fig. 4.15, where the liquid permeabilities for the two cores are
reported as a function of the confining pressure. One observes that the permeability of core 6927-
B4 is lower than the permeability of core 6927-B2, typically, by a factor of ~1.5, which is
qualitatively consistent with the initial measurements made in the permeability apparatus at USC
124
(see Table 3.4) prior to the initiation of the mechanical property measurements. Further, one
observes that the measured values (particularly for core 6927-B4) in Fig. 4.15 are substantially
smaller than those reported in Table 4 (before the mechanical testing).
0 5 10 15 20
0
2
4
6
8
6927-B2
a
6927-B4
a
6927-B2
b
6927-B4
b
6927-B2
c
6927-B4
c
Permeability (mD)
Confining Pressure (MPa)
Fig. 4.15 Permeabilities of the 6927-B2 (fresh) and 6927-B4 (aged) cores after the mechanical
tests.
a
Measured with confining pressure from 2.76MPa to 18.62MPa - NETL set-up;
b
Measured
with confining pressure of 1.38MPa - USC set-up;
c
Measured with confining pressure of
2.76MPa - USC set-up
To make sure that these differences were not due to discrepancies between the
measurements made in two different experimental permeation systems, the permeabilities of the
core samples after being removed from the Autolab system were again measured in the
125
permeability set-up at USC. The results are shown in Table 4.3. For example, for core 6927-B4 its
(liquid) permeability measured at a confining pressure of 2.76 MPa at USC was 6.6 mD, while the
permeability measured in the AutoLab system at the same confining pressure of 2.76 MPa was 6.3
mD, a <5% difference. It is clear, therefore, that the differences in the permeability values
measured before and after the mechanical property testing are due to irreversible changes occurring
due to compaction/re-packing of the grains in these high-quartz content sandstone samples (see
SEM images in Section 3.2.5) as a result of the high net stresses employed. That the confining
pressure affects the transport properties of these cores is obvious from the data in Fig. 4.14.
However, based on the similarity of the permeability values measured in the beginning of the tests
shown in Fig. 4.14 (carried out in the AutoLab system at a confining pressure of 2.76 MPa) and in
the USC set-up after the conclusion of the mechanical tests at the same confining pressure, it would
seem that the impact of confining stress, applied during the permeability measurements after the
mechanical property tests were completed, is reversible (in contrast to the irreversible changes that
occurred during the testing of mechanical properties).
Table 4.3 Permeability measurements before and after exposure to CO2/brine
Sample
Fresh sample
(mD)
Aged sample
(mD)
After mechanical testing
(mD)
6927-B1 9.0
a
- -
6927-B2 9.8
a
- 5.0
b
/6.8
c
6927-B4 10.6
a
37.7
a
6.6
b
/8.4
c
a): Measured with a confining pressure of 1.38 MPa - USC setup. b): Measured with a confining
pressure of 2.76 MPa - USC setup - after mechanical testing. c): Measured with a confining
pressure of 1.38 MPa - USC setup - after mechanical testing.
126
4.5 Summary
The dynamic mechanical properties before and after aging, studied via the measurement of
ultrasonic P and S elastic waves, show a decrease in Young’s modulus Shear Modulus and Bulk
modulus that is indicative of a weakening of the Mt Simon samples after exposition to CO2/brine.
A reduction of strength of sandstone after exposure to CO2 was also reported by Marbler et al.
(2013) and Rathnaweera et al. (2015). One may then conclude, that the exposure of Mt Simon
sandstone samples to a CO2/brine environment over a relatively short period of time appears to
induce rock alterations that ultimately result in reduced mechanical strength. The observed
reduction is attributed to dissolution of clays that serve as main constituents of the cementation of
quarts grains. The results of the measurement show that the Young’s modulus, Shear Modulus and
Bulk modulus all increase with effective pressure. This is consistent with the impact of applied
effective pressure on the sample’s permeability, as shown in Fig. 4.14. The plastic deformation
detected in the measurements of the Young’s Modulus is also consistent with the irreversible
changes (decrease) in the permeabilities measured for these samples (see section 4.4) and further
supports the use of two geologically similar cores to determine the impact of exposure to CO2/brine
on mechanical properties (instead of using a single core and testing the mechanical properties
before and after exposure to CO2/brine solutions).
127
Chapter 5: Future Work
5.1 CO2 Mass Transfer in Brine
In Chapter 2, we presented our studies on CO2 mass transfer in bulk water/brine and water/brine-
saturated synthetic unconsolidated porous media carried-out in a PVT system. In addition, we also
presented a preliminary study of CO2 mass transfer in brine-saturated real Mt. Simon Sandstone
powders via a FTIR technique. A key difficulty with the FTIR technique is that the thickness of
the sample sitting on the ATR crystal cannot be measured directly and has to be evaluated by
fitting the experimental data (see Chapter 2). In addition, the CO2 adsorption on the surface of the
dry sample is not very significant due to the small fraction of the clay minerals. This may lead to
experimental error when calculating the integrated peak area. Therefore, a more sensitive
technique is needed to study the CO2 adsorption and mass transfer in the sandstone sample.
To overcome this challenge a series of experiments have been designed and are currently
under way in our group in which CO2 uptake experiments are carried-out with both dry and brine-
saturated powders of Mt. Simon Sandstone, as well as dry and brine-saturated whole cubic samples
from the same core. The reason behind the experiments with the dry powders and dry cubic
samples is that one can directly measure CO2 adsorption on the Mt. Simon Sandstone. The
experiments with the brine-saturated powders and cubic sample combined with the solubility
measurements of CO2 dissolution in brine (presented in Chapter 2) can then provide the
information on CO2 mass transfer in brine-saturated real rocks. The thermogravimetric analysis
(TGA) technique is employed whose ability to measure the mass transfer of gas in similar natural
porous media has been previously demonstrated by our team (e.g., see Wang et al., 2015). TGA
experiments with wet powders carried-out under the same conditions with the FTIR measurements
128
will allow us, for the first time, to delineate mass transfer, dissolution and adsorption in a real
porous medium.
5.2 Transport and Mechanical Characterizations of Rock Sample
In Chapters 3 and 4, we investigate the rock-fluid interactions and their impact on the
transport and mechanical properties of the Mt. Simon Sandstone. All the characterizations are
performed before and after exposing the core to CO2/brine. Of interest, however, is for one to
investigate the evolution of the pore structure during the exposure of the core to the CO 2/brine
fluid. For that, we carried-out a series of preliminary experiments on the fresh samples using the
in-situ CT scanning technique. The Sample 6927-B3 was scanned by an industrial CT (North Star
Imaging M-5000) with the resolution of 16.2 µ m, while the Sample 6927-C1 was scanned by a
micro-CT (ZEISS Xradia) with resolutions of 1.66 µ m and 3.97 µ m. The produced 2-D slices were
segmented and analyzed using ImageJ.
For the industrial CT imaging, 3222 2-D slices were created at a 16.2 µ m resolution. Fig.
5.1 shows one of the slices and its binary counterpart that can represent the pore structure. For the
micro CT imaging, 1929 and 1901 2D slices were created at 1.66 µ m and 3.97 µ m resolution,
respectively. Fig. 5.2 and 5.3 show the original images, and their segmented counterparts.
129
Fig. 5.1 CT Image and its binary counterpart scanned with 16.2 µ m resolution
Fig. 5.2 CT Image and its binary counterpart scanned with 1.66 µ m resolution
130
Fig. 5.3 CT Image and its binary counterpart scanned with 3.97 µ m resolution
According to Figs. 5.1 to 5.3, the quality of the images is good and the boundary between
the pore space and the solid is clear so that one can capture the pore structure of the rock sample.
After image segmentation, the porosity of each slice was calculated using ImageJ. The evolution
of the porosity value along the axial direction of the core is shown in Fig. 5.4. In Fig. 5.4, the
porosity is plotted versus the slice number that represents the axial distance from the inlet.
According to Fig. 5.4, the porosity of single slice fluctuates along the axial direction. However,
the amplitude is not large. It is safe to conclude that the porosity stays around an average value.
The average porosity calculated from the 1.66 µ m resolution micro CT scan is 14.6% while that
of 3.97 µ m resolution micro CT scan is 17.8%. The average porosity obtained from the 16.2 µ m
resolution industrial CT scan is 14.7%. Table 3.3 reports the porosity of the tested samples
measured using the helium pycnometry method. We can see the porosity of Sample 6927-C1
before exposure is 23.6% and that of Sample 6927-B3 is 15%. It is not surprising that the porosity
values obtained from CT scans are lower than those measured by the helium pycnometry method.
131
The CT scanner cannot detect the pore space whose dimension is lower than CT’s resolution.
However, the helium gas can detect almost all the pore space of the sample. According to Fig. 5.4,
the porosity calculated using the 16.2 µ m resolution images is less than that calculated using the
3.97 µ m images. This can be explained by the fact that the scans with better resolution can detect
more pore space and lead to higher porosity. However, we also note that the porosity calculated
from the 1.66 µ m images is less than that calculated using the 3.97 µ m images. This could be due
to the heterogeneity of the rock sample. The 3.97 µ m resolution imaging scans the whole
cylindrical Sample 6927-C1 with the dimension of 0.6 cm × 0.8cm while the 1.66 µ m resolution
imaging only scans a subset of the same sample with dimension of 0.3 cm × 0.3cm.
Fig. 5.4 Evolution of the porosity in the axial direction
Up to now, we have had the images of the fresh sample with good quality to extract the
information of pore structure. We have also done some scans of the brine saturated sample. Then
CO2 was injected to displace the brine, and some scans have been conducted during the core
132
flooding experiments. However, the quality of these scans are not good enough for us to extract
any information of the pore structure. It is necessary to redesign the experiment and improve the
scanning quality. In the future study, the sample will be saturated with brine and exposed to CO 2
in a core holder installed in the CT equipment. The evolution of porosity and pore size distribution
during exposure would be monitored by in-situ. However, one should be very careful when
choosing scanners with different resolutions. For a scanner with high resolution, such as micro-
CT, the time required for scanning is very long (a couple of hours). It is doubt that micro-CT
scanning can capture the dynamics of the evolution. For a scanner with lower resolution, such as
the industrial-CT, the time required is much shorter, but its resolution is too low (>15 µ m) to
capture the important changes at the grain level (e.g., clay dissolution). In such experiments the
change in permeability can also be detected in real time from core-flooding experiments based on
the images of pore structure. The evolution of brine composition during exposure can also be
determined via in situ sampling.
For our measurement of mechanical properties, the sample was tested at dry condition. To
measure the mechanical properties during exposure, the sample could be aged in a core-holder
under a confining environment. For such measurements, strain gauges are attached on the surface
of the sample to detect the change of the strain during exposure. The recorded signals from the
strain gauge can provide information of the deformation that takes place in real time. During our
experiments to date, we have found that the permeability and Young’s Modulus of the rock
samples change with applied net stress. A mathematical model describing the relationship between
the permeability, the Young’s Modulus and the applied net stress needs to be developed.
133
References
ASTM D6767-16 Standard Test Method for Pore Size Characteristics of Geotextiles by Capillary
Flow Test, ASTM International, West Conshohocken, PA, 2016,https://doi.org/10.1520/D6767-
16
Azin, R., Mahmoudy, M., Raad, S. and Osfouri, S., 2013. Measurement and modeling of CO2
diffusion coefficient in Saline Aquifer at reservoir conditions. Central European Journal of
Engineering 3 (4): 585-594.
Baird, C. and Cann, M., 2005. Environmental chemistry. New York, W.H. Freeman and Co.
Bastami, A., Allahgholi, M. and Pourafshary, P., 2014. Experimental and modelling study of the
solubility of CO2 in various CaCl2 solutions at different temperatures and pressures. Petroleum
Science 11 (4): 569-577.
Belgodere, C., Dubessy, J., Vautrin, D., Caumon, M. C., Sterpenich, J., Pironon, J., Robert, P.,
Randi, A., Birat, J. P., 2015. Experimental determination of CO2 diffusion coefficient in aqueous
solutions under pressure at room temperature via Raman spectroscopy: impact of salinity (NaCl).
Journal of Raman Spectroscopy 46(10): 1025-1032.
Bermejo, M. D., Martí n, A., Florusse, L. J., Peters, C. J. and Cocero, M. J., 2005. The influence of
Na2SO4 on the CO2 solubility in water at high pressure. Fluid Phase Equilibria 238 (2): 220-228.
Cadogan, S. P., Maitland, G. C., Trusler, J. P. M., 2014. Diffusion coefficients of CO2 and N2 in
water at temperatures between 298.15 K and 423.15 K at pressures up to 45 MPa. Journal of
Chemical and Engineering Data 59(2): 519-525.
134
Elenius, M.T. and Johannsen, K., 2012. On the time scales of nonlinear instability in miscible
displacement porous media flow. Comput Geosci. 16, 901911.
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 (6): 1043-1050.
Culp, J. T., Goodman, A. L., Chirdon, D., Sankar, S. G. and Matranga, C., 2010. Mechanism for
the dynamic adsorption of CO2 and CH4 in a flexible linear chain coordination polymer as
determined from in situ infrared spectroscopy. The Journal of Physical Chemistry C 114 (5): 2184-
2191.
Culp, J. T., Sui, L., Goodman, A. and Luebke, D., 2013. Carbon dioxide (CO2) absorption behavior
of mixed matrix polymer composites containing a flexible coordination polymer. Journal of
Colloid and Interface Science 393: 278-285.
de Jong, S. M., Spiers, C. J. and Busch, A., 2014. Development of swelling strain in smectite clays
through exposure to carbon dioxide. International Journal of Greenhouse Gas Control 24: 149-161.
Drummond, S.E., 1981. Boiling and mixing of hydrothermal fluids: chemical effects on mineral
precipitation. PhD thesis. The Pennsylvania State University.
Duan, Z., Sun, R., Zhu, C. and Chou, I. M., 2006. An improved model for the calculation of CO2
solubility in aqueous solutions containing Na
+
, K
+
, Ca
2+
, Mg
2+
, Cl
-
, and SO4
2-
. Marine Chemistry
98 (2-4): 131-139.
Ennis-King, J., Preston, I. and Paterson, L., 2005 Onset of convection in anisotropic porous media
subject to a rapid change in boundary conditions. Phys. Fluids 17, 084107.
135
Espinoza, D. N. and Santamarina J. C., 2012. Clay interaction with liquid and supercritical CO2:
The relevance of electrical and capillary forces. International Journal of Greenhouse Gas Control
10: 351-362.
Falk, M. and Miller, A. G., 1992. Infrared spectrum of carbon dioxide in aqueous solution.
Vibrational Spectroscopy 4 (1): 105-108.
Farajzadeh, R., Barati, A., Delil, H. A., Bruining, J., Zitha, P. L. J., 2007. Mass transfer of CO2
into water and surfactant solutions. Petroleum Science and Technology 25(11-12): 1493-1511.
Farajzadeh, R., Zitha, P. L. J., Bruining, J., 2009. Enhanced mass transfer of CO 2 into water:
Experiment and modeling. Industrial & Engineering Chemistry Research 48(13): 6423-6431.
Fjær, E., Holt, R. M., Horsrud, P., Raaen, A. M. and Risnes, R., 2008. Petroleum Related Rock
Mechanics (2nd ed.), Elsevier.
Forster, P., V. Ramaswamy, P. Artaxo, T. Berntsen, R. Betts, D.W. Fahey, J. Haywood, J. Lean,
D.C. Lowe, G. Myhre, J. Nganga, R. Prinn, G. Raga, M. Schulz and R. Van Dorland, 2007:
Changes in Atmospheric Constituents and in Radiative Forcing. In: Climate Change 2007: The
Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M.
Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge,
United Kingdom and New York, NY, USA.
Frank, M. J. W., Kuipers, J. A. M., vanSwaaij, W. P. M., 1996. Diffusion coefficients and
viscosities of CO2+H2O, CO2+CH3OH, NH3+H2O, and NH3+CH3OH liquid mixtures. Journal of
Chemical and Engineering Data 41(2): 297-302.
136
Garcia, J.E., 2001. Density of aqueous solutions of CO2, Lawrence Berkley National Laboratory
Report, LBNL 49023.
GCCSI, 2016. The global status of CCS: 2016 summary report. Global CCS Institute.
Gilbert, K., Bennett, P., Wolfe, W., Zhang, T. and Romanak, K., 2016. CO2 solubility in aqueous
solutions containing Na
+
, Ca
2+
, Cl
−
, SO4
2−
and HCO
3-
: The effects of electrostricted water and ion
hydration thermodynamics. Applied Geochemistry 67: 59-67.
Goodman, A. L., 2009. A comparison study of carbon dioxide adsorption on polydimethylsiloxane,
silica gel, and Illinois No. 6 coal using in situ infrared spectroscopy. Energy & Fuels 23 (2): 1101-
1106.
Goodman, A. L., Campus, L. M. and Schroeder, K. T., 2005. Direct evidence of carbon dioxide
sorption on argonne premium coals using attenuated total reflectance−Fourier transform infrared
spectroscopy. Energy & Fuels 19 (2): 471-476.
Goodman, A., Sanguinito, S., Tkach, M., Natesakhawat, S., Kutchko, B., Fazio, J. and Cvetic, P.,
2019. Investigating the role of water on CO2-Utica Shale interactions for carbon storage and shale
gas extraction activities – evidence for pore scale alterations. Fuel 242: 744-755.
Gubbins, K. E., Bhatia, K. K., Walker, R. D., 1966. Diffusion of gases in electrolytic solutions.
AIChE Journal 12(3): 548-552.
Hassanzadeh, H., Pooladi-Darvish, M. and Keith, D.W., 2006 Stability of a fluid in ahorizontal
saturated porous layer: effect of non-linear concentration profile, initial, and boundary conditions.
Transp Porous Med 65, 193-211.
137
Himmelblau, D. M., 1964. Diffusion of dissolved gases in liquids. Chemical Reviews 64(5): 527-
550.
Hirai, S., Okazaki, K., Yazawa, H., Ito, H., Tabe, Y., Hijikata, K., 1997. Measurement of CO 2
diffusion coefficient and application of LIF in pressurized water. Energy 22(2-3): 363-367.Karl,
T. R., Trenberth, K. E., 2003. Modern global climate change. Science 302(5651): 1719-1723.
Hu, S., Li, X. C., Bai, B., Shi, L., Liu, M. and Wu, H., 2017. A modified true triaxial apparatus for
measuring mechanical properties of sandstone coupled with CO2-H2O biphase fluid. Greenhouse
Gases: Science and Technology 7(1): 78-91.
Huq, F., Haderlein, S. B., Cirpka, O. A., Nowak, M., Blum, P. and Grathwohl, P., 2015. Flow-
through experiments on water–rock interactions in a sandstone caused by CO2 injection at
pressures and temperatures mimicking reservoir conditions. Applied Geochemistry 58: 136-146.
Ilgen, A. G. and Cygan, R. T., 2016. Mineral dissolution and precipitation during CO2 injection
at the Frio-I Brine Pilot: Geochemical modeling and uncertainty analysis. International Journal of
Greenhouse Gas Control 44: 166-174.
Kampman, N., Bickle, M., Wigley, M. and Dubacq, B., 2014. Fluid flow and CO2–fluid–mineral
interactions during CO2-storage in sedimentary basins. Chemical Geology 369: 22-50.
Kauffman, K. L., Culp, J. T., Allen, A. J., Espinal, L., Wong-Ng, W., Brown, T. D., Goodman, A.,
Bernardo, M. P., Pancoast, R. J., Chirdon, D. and Matranga, C., 2011a. Selective adsorption of
CO2 from light gas mixtures by using a structurally dynamic porous coordination polymer.
Angewandte Chemie 123 (46): 11080-11084.
138
Kauffman, K. L., Culp, J. T., Goodman, A. and Matranga, C., 2011b. FT-IR study of CO2
adsorption in a dynamic copper (II) benzoate−pyrazine host with CO2−CO2 interactions in the
adsorbed state. The Journal of Physical Chemistry C 115 (5): 1857-1866.
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 (17): 4393-4398.
Kim, M.C. and Choi, C.K., 2012. Linear stability analysis on the onset of buoyancy-driven
convection in liquid-saturated porous medium. Phys. Fluids 24, 044102.
Kim, M.C., Song, K.H., Choi, C.K. and Yeo, J. K., 2008. Onset of buoyancy-driven convection in
a liquid-saturated cylindrical porous layer supported by a gas layer. Phys. Fluids 20, 054104.
Kö hler, S. J., Bosbach, D. and Oelkers, E. H., 2005. Do clay mineral dissolution rates reach steady
state? Geochimica Et Cosmochimica Acta 69(8): 1997-2006.
Koschel, D., Coxam, J., 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 (1): 107-120.
Krukowski, E. G., Goodman, A., Rother, G., Ilton, E. S., Guthrie, G. and Bodnar, R. J., 2015. FT-
IR study of CO2 interaction with Na
+
exchanged montmorillonite. Applied Clay Science 114: 61-
68.
Kuila, U. and Prasad, M., 2013. Specific surface area and pore-size distribution in clays and shales.
Geophysical Prospecting 61(2): 341-362.
139
Labotka, D. M., Panno, S. V., Locke, R. A. and Freiburg, J. T., 2015. Isotopic and geochemical
characterization of fossil brines of the Cambrian Mt. Simon Sandstone and Ironton–Galesville
Formation from the Illinois Basin, USA. Geochimica Et Cosmochimica Acta 165: 342-360.
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 ° C and pressures to 29 MPa. Journal of Chemical
& Engineering Data 49 (4): 1026-1031.
Liu, F., Lu, P., Griffith, C., Hedges, S. W., Soong, Y., Hellevang, H. and Zhu, C., 2012. CO2–
brine–caprock interaction: Reactivity experiments on Eau Claire shale and a review of relevant
literature. International Journal of Greenhouse Gas Control 7: 153-167.
Liu, Y., Hou, M., Yang, G. and Han, B., 2011. Solubility of CO2 in aqueous solutions of NaCl,
KCl, CaCl2 and their mixed salts at different temperatures and pressures. The Journal of
Supercritical Fluids 56 (2): 125-129.
Lu, W. J., Guo, H. R., Chou, I. M., Burruss, R. C., Li, L. L., 2013. Determination of diffusion
coefficients of carbon dioxide in water between 268 and 473 K in a high-pressure capillary optical
cell with in situ Raman spectroscopic measurements. Geochimica Et Cosmochimica Acta 115:
183-204.
Malinin, S. D. and Kurovskaya, N. A., 1975. Solubility of CO2 in chloride solutions at elevated
temperatures and CO2 pressures. Geochem. Int. 12 (2): 199-201.
Marbler, H., Erickson, K. P., Schmidt, M., Lempp, C. and Pö llmann, H., 2013. Geomechanical
and geochemical effects on sandstones caused by the reaction with supercritical CO2: an
experimental approach to in situ conditions in deep geological reservoirs. Environmental Earth
Sciences 69(6): 1981-1998.
140
Mastalerz, M., Goodman, A. and Chirdon, D., 2012. Coal lithotypes before, during, and after
exposure to CO2: insights from direct Fourier transform infrared investigation. Energy & Fuels 26
(6): 3586-3591.
Mazarei, A. F., Sandall, O. C., 1980. Diffusion coefficients for helium, hydrogen, and carbon
dioxide in water at 25 ⁰C. AIChE Journal 26(1): 154-157.
Miller, Q. R. S., Wang, X., Kaszuba, J. P., Mouzakis, K. M., Navarre-Sitchler, A. K. , Alvarado,
V., McCray, J. E., Rother, G., Bañ uelos, J. L. and Heath, J. E., 2016. Experimental Study of
Porosity Changes in Shale Caprocks Exposed to Carbon Dioxide-Saturated Brine II: Insights from
Aqueous Geochemistry. Environmental Engineering Science 33(10): 736-744.
Mourhatch, R., Tsotsis, T. T. and Sahimi, M., 2011. Determination of the true pore size distribution
by flow perporometry experiments: An invasion percolation model. Journal of Membrane Science
367(1): 55-62.
Mouzakis, K. M., Navarre-Sitchler, A. K., Rother, G., Bañ uelos, J. L., Wang, X., Kaszuba, J. P. ,
Heath, J. E., Miller, Q. R.S., Alvarado, V. and McCray, J. E., 2016. Experimental Study of Porosity
Changes in Shale Caprocks Exposed to CO2-Saturated Brines I: Evolution of Mineralogy, Pore
Connectivity, Pore Size Distribution, and Surface Area. Environmental Engineering Science
33(10): 725-735.
Munz, I. A., Brandvoll, Ø, Haug, T. A., Iden, K., Smeets, R., Kihle, J. and Johansen, H., 2012.
Mechanisms and rates of plagioclase carbonation reactions. Geochimica Et Cosmochimica Acta
77: 27-51.
141
Nia, S. F., Dasani, D., Tsotsis, T. T. and Jessen, K., 2016. An Integrated Approach for the
Characterization of Shales and Other Unconventional Resource Materials. Industrial &
Engineering Chemistry Research 55(12): 3718-3728.
Nield, D. A. and Bejan, A., 2006. Convection in Porous Media, 3rd edn. New York: Springer.
Nikitin, N., 2006. Third-order-accurate semi-implicit Runge-Kutta scheme for incompressible
Navier-Stokes equations. Int. J. Numer. Meth. Fluids 51, 221-233.
Nover, G., von der Gö nna, J., Heikamp, S. and Kö ster, J., 2013. Changes of petrophysical
properties of sandstones due to interaction with supercritical carbon dioxide - a laboratory study.
Eur. J. Mineral. 25: 317-329.
Pau, G. S.H., Bell, J. B., Pruess, K., Almgren, A. S., Lijewski, M. J. and Zhang, K., 2010. High-
resolution simulation and characterization of density-driven flow in CO2 storage in saline aquifers.
Adv. Water Resour. 33, 443455.
Peyret, R., 2002. Spectral Methods for Incompressible Viscous Flow. New York: Springer.
Pires, J. C. M., Martins, F. G., Alvim-Ferraz, M. C. M., Simoes, M., 2011. Recent developments
on carbon capture and storage: An overview. Chemical Engineering Research & Design 89(9):
1446-1460.
Policarpo, N. A., Ribeiro, P. R., 2011. Experimental measurement of gas-liquid diffusivity.
Brazilian Journal Of Petroleum and Gas 5(3): 171-188.
Raad, S. M. J., Azin, R., Osfouri, S., 2015. Measurement of CO2 diffusivity in synthetic and saline
aquifer solutions at reservoir conditions: the role of ion interactions. Heat and Mass Transfer
51(11): 1587-1595.
142
Rathnaweera, T. D., Ranjith, P. G., Perera, M. S. A., Haque, A., Lashin, A., Al Arifi, N.,
Chandrasekharam, D., Yang, S. Q., Xu, T., Wang, S. H. and Yasar, E., 2015. CO2-induced
mechanical behaviour of Hawkesbury sandstone in the Gosford basin: An experimental study.
Materials Science and Engineering: A 641: 123-137.
Renner, T. A., 1988. Measurement and correlation of diffusion coefficients for CO2 and rich-gas
applications. Society of Petroleum Engineers: 517-523.
Riaz, A., Hesse, M., Tchelepi, H. A. and Jr, F. M. Orr, 2006. Onset of convection in a
gravitationally unstable diffusive boundary layer in porous media. J. Fluid Mech. 548, 87-111.
Riazi, M. R., 1996. A new method for experimental measurement of diffusion coefficients in
reservoir fluids. Journal of Petroleum Science and Engineering 14(3-4): 235-250.
Rimmelé , G., Barlet-Goué dard, V. and Renard, F., 2010. Evolution of the Petrophysical and
Mineralogical Properties of Two Reservoir Rocks Under Thermodynamic Conditions Relevant for
CO2 Geological Storage at 3 km Depth. Oil Gas Sci. Technol. – Rev. IFP 65(4): 565-580.
Rosenbauer, R. J., Koksalan, T. and Palandri, J. L., 2005. Experimental investigation of CO2–
brine–rock interactions at elevated temperature and pressure: Implications for CO2 sequestration
in deep-saline aquifers. Fuel Processing Technology 86(14): 1581-1597.
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 fü r
physikalische Chemie 97 (1): 85-97.
143
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 (3): 431-448.
Sanguinito, S. Goodman, A., Tkach, M., Kutchko, B., Culp, J., Natesakhawat, S., Fazio, J., Fukai,
I. and Crandall, D., 2018. Quantifying dry supercritical CO2-induced changes of the Utica Shale.
Fuel 226: 54-64.
Schä dle, T., Pejcic, B., Myers, M. and Mizaikoff, B., 2016. Portable mid-infrared sensor system
for monitoring CO2 and CH4 at high pressure in geosequestration scenarios. ACS Sensors 1 (4):
413-419.
Sell, A., Fadaei, H., Kim, M., Sinton, D., 2013. Measurement of CO2 diffusivity for carbon
sequestration: A microfluidic approach for reservoir-specific analysis. Environmental Science &
Technology 47(1): 71-78.
Shao, H., Ray, J. R. and Jun, Y., 2010. Dissolution and Precipitation of Clay Minerals under
Geologic CO2 Sequestration Conditions: CO2−Brine−Phlogopite Interactions. Environmental
Science & Technology 44(15): 5999-6005.
Shi, X. and Mao, S., 2017. An improved model for CO2 solubility in aqueous electrolyte solution
containing Na
+
, K
+
, Mg
2+
, Ca
2+
, Cl
−
and SO4
2−
under conditions of CO2 capture and sequestration.
Chemical Geology 463: 12-28.
Shi, Z., Sun, L., Haljasmaa, I., Harbert, W., Sanguinito, S., Tkach, M., Goodman, A., Tsotsis, T.
T. and Jessen, K., 2019. Impact of brine/CO2 exposure on the transport and mechanical properties
of the Mt Simon sandstone. Journal of Petroleum Science and Engineering 177: 295-305.
144
Shi, Z., Wen, B., Hesse, M. A., Tsotsis, T. T. and Jessen, K., 2018. Measurement and modeling of
CO2 mass transfer in brine at reservoir conditions. Advances in Water Resources 113: 100-111.
Simon, D. E. and Anderson M. S., 1990. Stability of Clay Minerals in Acid. SPE Formation
Damage Control Symposium. Lafayette, Louisiana, Society of Petroleum Engineers.
Slim, A. C. and Ramakrishnan, T. S., 2010. Onset and cessation of time-dependent, dissolution-
driven convection in porous media. Phys. Fluids 22, 124103.
Slim, A. C., 2014. Solutal-convection regimes in a two-dimensional porous medium. J. Fluid Mech.
741, 461-491.
Spaulding, R., Haljasmaa, I., Fazio, J., Gieger, C., Kutchko, B., Gardiner, J., Shine, J. M., Benge,
G., DeBruijn, G. and Harbert, W., 2015. An Assessment of the Dynamic Moduli of
Atmospherically Generated Foam Cements. Offshore Technology Conference. Houston, Texas,
USA, Offshore Technology Conference.
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 (1): 173-196.
Szabó , Z., Hellevang, H., Kirá ly, Cs., Sendula, E., Kó nya, P., Falus, Gy., Tö rö k, Sz. and Szabó ,
Cs. (2016). "Experimental-modelling geochemical study of potential CCS caprocks in brine and
CO2-saturated brine." International Journal of Greenhouse Gas Control 44: 262-275.
Tamimi, A., Rinker, E. B., Sandall, O. C., 1994. Diffusion coefficients for hydrogen sulfide,
carbon dioxide, and nitrous oxide in water over the temperature range 293-368K. Journal of
Chemical and Engineering Data 39(2): 330-332.
145
Tang, Y., Himmelblau, D., 1965. Effect of solute concentration on the diffusivity of carbon dioxide
in water. Chem. Eng. Sci. 20: 7.Tse, F. C., Sandall, O. C., 1979. Diffusion coefficients for oxygen
and carbon dioxide in water at 25 ⁰C by unsteady state desorption from a quiescent liquid.
Chemical Engineering Communications 3(3): 147-153.
Tong, D., Trusler, J. P. M. 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 (7): 2116-2124.
Tutolo, B. M., Luhmann, A. J., Kong, X. Z., Saar, M. O. and Seyfried Jr, W. E., 2015. CO2
sequestration in feldspar-rich sandstone: Coupled evolution of fluid chemistry, mineral reaction
rates, and hydrogeochemical properties. Geochimica Et Cosmochimica Acta 160: 132-154.
Unver, A. A., Himmelblau, D. M., 1964. Diffusion coefficients of CO2, C2H4, C3H6 and C4H8 in
water from 6 to 65 ⁰C. Journal of Chemical & Engineering Data 9(3): 428-431.
Veldsink, J. W., Vandamme, R. M. J., Versteeg, G. F., Vanswaaij, W. P. M., 1995. The use of the
dusty-gas model for the description of mass transport with chemical reaction in porous-media.
Chemical Engineering Journal and the Biochemical Engineering Journal 57(2): 115-125
Villamor Lora, R., Ghazanfari, E. and Asanza Izquierdo, E., 2016. Geomechanical
Characterization of Marcellus Shale. Rock Mechanics and Rock Engineering 49(9): 3403-3424.
Wang, L. S., Lang, Z. X., Guo, T. M., 1996. Measurement and correlation of the diffusion
coefficients of carbon dioxide in liquid hydrocarbons under elevated pressures. Fluid Phase
Equilibria 117(1-2): 364-372.
146
Wang, L., Shen, Z., Hu, L. and Yu, Q., 2014. Modeling and measurement of CO 2 solubility in
salty aqueous solutions and application in the Erdos Basin. Fluid Phase Equilibria 377: 45-55.
Wang, S., Hou, J., Liu, B., Zhao, F., Yuan, G., Liu, G., 2013. The pressure-decay method for nature
convection accelerated diffusion of CO2 in oil and water under elevated pressures. Energy Sources
Part A: Recovery Utilization and Environmental Effects 35(6): 538-545.
Wang, Y., Tsotsis, T. T. and Jessen, K., 2015. Competitive sorption of methane/ethane mixtures
on shale: measurements and modeling. Industrial & Engineering Chemistry Research 54 (48):
12187-12195.
Weir, G.J., White, S.P. & Kissling, W.M. Transp Porous Med (1996) 23: 37.
doi:10.1007/BF00145265
Wen, B., Chini, G. P., Dianati, N. & Doering, C. R. 2013 Computational approaches to aspect-
ratio-dependent upper bounds and heat flux in porous medium convection. Phys. Lett. A 377,
2931-2938.
Wen, B., Corson, L. T. & Chini, G. P. 2015 Structure and stability of steady porous medium
convection at large Rayleigh number. J. Fluid Mech. 772, 197-224.
Witherspoon, P. A., Saraf, D. N., 1965. Diffusion of Methane Ethane Propane and N-Butane in
Water from 25 to 43 ⁰. Journal of Physical Chemistry 69(11): 3752-+.
Wyllie, M. R. J., Gregory, A. R., 1955. Fluid Flow through Unconsolidated Porous Aggregates:
Effect of Porosity and Particle Shape on Kozeny-Carman Constants. Ind. Eng. Chem. 47 (7), 1379-
1388.
147
Xu, X., Chen, S. and Zhang, D., 2006. Convective stability analysis of the long-term storage of
carbon dioxide in deep saline aquifers. Adv. Water Resour. 29, 397-407.
Yan W., Huang S. and Stenby E.H., 2011, Measurement and modeling of CO2 solubility in NaCl
brine and CO2–saturated NaCl brine density. International Journal of Greenhouse Gas Control,
Vol. 5, No. 6, p. 1460-1477.
Yang, C. D. and Gu, Y. G., 2006. Accelerated mass transfer of CO 2 in reservoir brine due to
density-driven natural convection at high pressures and elevated temperatures. Industrial &
Engineering Chemistry Research 45 (8): 2430-2436
Yu, Z., Liu, L., Yang, S., Li, S. and Yang, Y., 2012. An experimental study of CO2–brine–rock
interaction at in situ pressure–temperature reservoir conditions. Chemical Geology 326-327: 88-
101.
Zarghami, S., Boukadi, F. and Al-Wahaibi, Y., 2017. Diffusion of carbon dioxide in formation
water as a result of CO2 enhanced oil recovery and CO2 sequestration. Journal of Petroleum
Exploration and Production Technology 7 (1): 161-168.
Zhang, W. D., Wu, S. L., Ren, S. R., Zhang, L., Li, J. T., 2015. The modeling and experimental
studies on the diffusion coefficient of CO2 in saline water. Journal of CO2 Utilization 11: 49-53.
Zhang, W. N., Hu, H. X., Li, X. C. and Fang, Z. M., 2017. Evolutions in the Elastic Constants of
Ca-Montmorillonites with H2O/CO2 Mixture under Supercritical Carbon Dioxide Conditions. The
Journal of Physical Chemistry C 121(5): 2702-2710.
148
Zhao, H., Dilmore, R. and Lvov, S. N., 2015a. Experimental studies and modeling of CO 2
solubility in high temperature aqueous CaCl2, MgCl2, Na2SO4, and KCl solutions. AIChE Journal
61 (7): 2286-2297.
Zhao, H., Dilmore, R., Allen, D. E., Hedges, S. W., Soong, Y. and Lvov, S. N., 2015b.
Measurement and modeling of CO2 solubility in natural and synthetic formation brines for CO2
sequestration. Environmental Science & Technology 49 (3): 1972-1980.
149
Appendix
Appendix A: Leak correction
The process for estimating leak rate during the CO2 diffusivity measurements are based on the
Dusty gas model (Veldsink et al., 1995)
𝑁 𝑖 = (−
1
𝑅𝑇
)[𝐷 𝑘 𝑒 𝑑𝑃
𝑑𝑧
+
𝐵 0
𝜇 𝑃 𝑑𝑃
𝑑𝑧
] , (s1)
𝐷 𝑘 𝑒 =
𝑑 𝑝 3
√
8𝑅𝑇
𝜋 𝑀 𝑤 , (s2)
where 𝐷 𝑘 𝑒 the Knudsen diffusivity in cm
2
/s, 𝑑 𝑝 the average diameter of the leak, 𝑀 𝑤 the molecular
weight of the gas, B0 the permeability in cm
2
(which only depends on porosity and geometry of
system), 𝜇 the viscosity in Pa s, Ni the flux of the leak gas. We assume that the leak occurs from
inside to the outside of the sample cell. Thus, the value of Ni is positive. By rearranging and
integrating Eq. s1, we arrive at
𝑁 𝑖 ∙ 𝐿 =
𝐷 𝑖 𝑒 𝑅𝑇
∫ 𝑑𝑃 𝑃 1
𝑃 2
+
𝐵 0
𝑅𝑇
∫
𝑃 𝜇 𝑑𝑃 𝑃 1
𝑃 2
, (s3)
where P1 is the pressure of the sample cell in psi and P2 the pressure of the atmospheric
environment.
Next, we define
−
𝑑𝑛
𝑑𝑡
= 𝑁 𝑖 𝐴 𝑐 , (s4)
where dn/dt is the gas leak rate in mol /s, 𝐴 𝑐 the cross section area in cm
2
. The leak rate is related
to the pressure in the sample cell via
150
𝑑𝑛
𝑑𝑡
=
𝑉 𝑠 𝑅𝑇
𝑑 (
𝑃 𝑍 )
𝑑𝑡
. (s5)
For Helium, the compressibility factor (Z) is assumed to be equal to unity.
By rearranging Eqs. s4-s5, we arrive at
−
𝑑 (
𝑃 𝑍 )
𝑑𝑡
=
𝐷 𝑖 𝑒 𝐴 𝑐 𝑉 𝑠 𝐿 (𝑃 1 − 𝑃 2) +
𝐵 0
𝐴 𝑐 𝑉 𝑠 𝐿 ∫
𝑃 𝜇 𝑑𝑃 𝑃 1
𝑃 2
, (s6)
where Vs is the void volume of the sample cell in cm
3
, L the hypothetical length of the leak path in
cm, R the gas constant and T the temperature in K. We further assume that the viscosity of He has
a weak dependence on pressure and set 𝜇 𝐻𝑒
=2.10x10
-5
pa s.
Next we introduce the groups
𝐴 1 =
𝐷 𝑘 𝑒 𝐴 𝑐 𝑉 𝑠 𝐿 , (s7)
𝐴 2 =
𝐵 0
𝐴 𝑐 𝑉 𝑠 𝐿 , (s8)
into Eqn. (6):
−𝑑 (
𝑃 𝑍 )/𝑑𝑡
(𝑃 1−𝑃 2)
= 𝐴 1 + (
𝐴 2
𝜇 )(
𝑃 1+𝑃 2
2
) (s9)
From measurements of the Helium leak rate at different pressures, we can evaluate A1 and A2/µ
from the intercept and slope from a plot of
−𝑑 (
𝑃 𝑍 )/𝑑𝑡
(𝑃 1−𝑃 2)
versus
𝑃 1+𝑃 2
2
. Once A1 and A2 are obtained,
the experimental data for CO2 experiments can be corrected as follows:
−
𝑑𝑛
𝑑𝑡
=
𝑉𝑠
𝑅𝑇
𝐴 1
√ 𝑀 𝑤 ,𝐻𝑒
√ 𝑀 𝑤 ,𝐶𝑂 2
(𝑃 1 − 𝑃 2) +
𝑉𝑠
𝑅𝑇
𝐴 2 ∫
𝑃 𝜇 𝐶𝑂 2
𝑑𝑃 𝑃 1
𝑃 2
(s10)
or
151
∆𝑛 = (
𝑉𝑠
𝑅𝑇
𝐴 1
√ 𝑀 𝑤 ,𝐻𝑒
√ 𝑀 𝑤 ,𝐶𝑂 2
(𝑃 1 − 𝑃 2) +
𝑉𝑠
𝑅𝑇
𝐴 2 ∫
𝑃 𝜇 𝐶𝑂 2
𝑑𝑃 𝑃 1
𝑃 2
) ∆𝑡 . (s11)
For each point in time, we can calculate Δn, that accounts for how many moles of gas leaks from
the sample cell during the time interval. During the interpretation of an experiment, we convert
the experimental pressure data into moles of CO2 in the vapor phase (ni) using a compressibility
factor from NIST. The corrected moles of CO2 for each time point is given by ni+ Δni, and the
corrected data can then be used to estimate the diffusivity of CO2. The experiment can be
terminated once the observed pressure change is similar to the predicted leak rate:
−
𝑑 (
𝑃 𝑍 )
𝑑𝑡
= 𝐴 1(𝑃 1 − 𝑃 2) + 𝐴 2 ∫
𝑃 𝜇 𝐶𝑂 2
𝑑𝑃 𝑃 1
𝑃 2
(s12)
References
Veldsink, J. W., Vandamme, R. M. J., Versteeg, G. F., Vanswaaij, W. P. M., 1995. The use of the
dusty-gas model for the description of mass transport with chemical reaction in porous-media.
Chemical Engineering Journal and the Biochemical Engineering Journal 57(2): 115-125
152
Appendix B: Evaluation of Dissolution Rate
To evaluate the experimental dissolution rate reported in Fig. A-B. 1, we convert the experimental
pressure/temperature data to moles of CO2 in the gas phase of the sample cell (nCO2) and fit the
following smooth function to the data (~40,000 points)
𝑛 𝐶𝑂 2
= 𝐶 0
+ 𝐶 1
∙ exp(𝐶 2
∙ 𝑡 ) + 𝐶 3
∙ exp (𝐶 4
∙ 𝑡 ) (s13)
The derivative with respect to time provides the uptake rate reported in Fig. 16. A comparison of
the of the data with Eq. (s13) along with the coefficients (including 95% confidence bounds) are
provided below.
Coefficient C0 C1 C2 C3 C4
Value 0.05081 0.004137 -0.4742 0.007118 -2.067
95% conf.
interval
0.05081 0.004113 -0.4766 0.007095 -2.076
0.05081 0.004161 -0.4719 0.007141 -2.058
153
Fig. A-B. 1 Comparison of data (subset shown for clarity) and fitting function
154
Appendix C: Recorded FT-IR Spectra for Each Pressure Step
Fig. A-C. 1 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone at
pressure of 0.3 MPa
Fig. A-C. 2 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone at
pressure of 0.7 MPa
155
Fig. A-C. 3 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone at
pressure of 2.8 MPa
Fig. A-C. 4 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone at
pressure of 5.5 MPa
156
Fig. A-C.5 ATR-FTIR spectra of absorbed CO2 on brine saturated Mt. Simon Sandstone at
pressure of 8.3 MPa
157
Appendix D: Mineral composition from other depths in the Mt. Simon formations
Table A-D. 1 Mineral composition from other depths in the Mt. Simon formations
Whole-rock mineralogy 5958.30 ft (%) 5971.60 ft (%) 6979.57 ft (%)
Quartz 81 90 73
K-feldspar 2 6 12
Plagioclase 0 0 7
Calcite 0 0 0
Siderite 0 0 0
Ankerite/Fe-dolomite 0 0 1
Dolomite 0 0 0
Pyrite 1 0 1
Barite 0 0 0
Fluorapatite 0 0 1
Hematite 0 0 0
Clay 15 3 4
Clay % % %
Illite-smectite 39 24 68
Illite 49 42 31
Chlorite 2 26 1
Kaolinite 10 7 1
158
Appendix E: N2 adsorption isotherms
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
Volume Adsorbed (cm
3
/g STP)
Relative Pressure (P/P
0
)
Fig. A-E.1 N2 adsorption isotherm of sample 6927-C1 before aging
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.5
1.0
1.5
2.0
Volume Adsorbed (cm
3
/g STP)
Relative Pressure (P/P
0
)
Fig. A-E.2 N2 adsorption isotherm of sample 6927-C1 after aging
159
Appendix F: EDX results
160
161
Fig. A-F.1 EDX results for all the points
162
Appendix G: Sensitivity test of the model for FT-IR experiments
To test the sensitivity of the model (as shown by Equations 2.29-2.37) to the porosity of the
particles εi, we consider a case that the value of εi is 0 that means the particles are nonporous. This
turns out the value of J = 4.07, εm = 0.268, and the effective diffusivity of the overall porous
material is 1.3135E-07cm
2
/s. The modeling and experimental results are plotted in Figs. A-F.1 to
3. The sensitivity of the model to the IR penetration depth was also tested. We found that the
change of the penetration depth from 1 to 2 µm made no change to the values of J and εm.
Figs. A-F.1 Modeling result and experimental data for pressure of 0.7 MPa with εi=0
163
Figs. A-F.2 Modeling result and experimental data for pressure of 2.8 MPa with εi=0
Figs. A-F.3 Modeling result and experimental data for pressure of 5.5 MPa with εi=0
Abstract (if available)
Abstract
Emissions of greenhouse gases are thought to contribute to global warming. Geological carbon sequestration (GCS) is currently considered a promising method to mitigate atmospheric CO₂ and, thus, to potentially minimize climate change. In this approach, CO₂ is injected into the subsurface and is trapped there by three main mechanisms, namely physical trapping, dissolution, and mineral precipitation. ❧ The present work focuses on two important aspects of GCS. First, we study mass transfer and sorption phenomena in brine, which are the two key processes occurring during CO₂ dissolution trapping in GCS. We employ pressure-decay experiments to measure CO₂ solubility, and mass transfer in water/brine systems at elevated pressures of relevance to CO₂ storage operations in saline aquifers together with modeling to delineate and interpret the experimental data. Accurate measurements and modeling of mass transfer in this context are crucial to an improved understanding of the long-term fate of CO₂ that is injected into the subsurface for storage purposes. Pressure-decay experimental data are presented for CO₂/water and CO₂/brine systems with and without an unconsolidated porous medium being present. Companion high-resolution 2-D numerical simulations demonstrate that natural convection will complicate the interpretation of the experimental observations, if the size of the particles constituting the unconsolidated porous medium is not sufficiently small. In such settings, we demonstrate that simple 1-D interpretations based on diffusional transport alone can result in an overestimation of the uptake (diffusivity) by two orders of magnitude. The high-resolution 2-D numerical calculations, on the other hand, agree well with the experimental observations for conditions where natural convection contributes substantially to the overall mass transfer process. The ATR-FTIR technique was also applied to study the CO₂ mass transfer and sorption phenomena in the Mt. Simon Sandstone. The experimental observations are able to capture the CO₂ diffusion in the brine saturated sample, and are consistent with the aforementioned separate measurements of CO₂ solubility in bulk brine and in brine-saturated unconsolidated porous media. ❧ We also study, in addition, rock-fluid interactions and their impact on the transport and mechanical properties of the host rock, which are phenomena relevant to CO₂ mineral trapping during GCS. Specifically, the present study investigates the change in the flow-through characteristics, porosity, and the mechanical behavior of Mt. Simon Sandstone samples caused by exposure to brine/CO₂. The cores being investigated were extracted from a certain depth interval (2110.4 - 2111.4 m or 6924 - 6927 ft) in the Mt. Simon formation. Their mechanical and transport properties were first characterized, and the cores were then aged in CO₂-saturated brine at a pressure of 17.24 MPa (2500 psi) and a temperature of 50⁰C for periods ranging from one to two weeks. Following that, the change in transport and mechanical properties of the samples were analyzed using He pycnometry, flow perporometry and triaxial testing. Our experiments show that the porosity of the Mt. Simon samples slightly increases after exposure to CO₂/brine, while the permeability increases more substantially (depending on the confining pressure environment). Measurements of the flow-through pore size distribution (PSD) are indicative of significant changes occurring, consistent with the observed increases in permeability. Nitrogen adsorption tests (BET), before and after incubation, show a significant loss of pore volume in the mesopore range that is indicative of clay dissolution. Weakening of the materials was observed based on the mechanical properties studied, a result that is consistent with the observed dissolution of clays that play a central role in the cementation of the quartz grains. Finally, the analysis of the brine compositions employed in the aging experiments reveals an increase in the concentration of most cations after incubation with the Mt. Simon cores. This is also consistent with mineral/clay dissolution, confirmed by the porosity, transport, and mechanical property measurements as well as electron microscopy analysis of the same samples.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Impact of exposure to brine/CO₂ on the mechanical and transport properties of the Mt. Simon sandstone
PDF
Investigation of gas transport and sorption in shales
PDF
A study of diffusive mass transfer in tight dual-porosity systems (unconventional)
PDF
Mass transfer during enhanced hydrocarbon recovery by gas injection processes
PDF
Investigation of gas sorption and mass transfer in gas shales
PDF
The study of mass transfer in gas shales and the optimization of hydraulic stimulation processes via additives
PDF
The role of counter-current flow in the modeling and simulation of multi-phase flow in porous media
PDF
Modeling and simulation of multicomponent mass transfer in tight dual-porosity systems (unconventional)
PDF
Chemical and mechanical deformation of porous media and materials during adsorption and fluid flow
PDF
Efficient simulation of flow and transport in complex images of porous materials and media using curvelet transformation
PDF
Methanol synthesis in the membrane reactor
PDF
Hydrogen storage in carbon and silicon carbide nanotubes
PDF
Modeling and simulation of complex recovery processes
PDF
In situ studies of the thermal evolution of the structure and sorption properties of Mg-Al-CO3 layered double hydroxide
PDF
Waterflood induced formation particle transport and evolution of thief zones in unconsolidated geologic layers
PDF
Calcite and aragonite dissolution in seawater: kinetics, mechanisms, and fluxes in the North Pacific
PDF
Studies of combustion characteristics of heavy hydrocarbons in simple and complex flows
Asset Metadata
Creator
Shi, Zhuofan
(author)
Core Title
The study of CO₂ mass transfer in brine and in brine-saturated Mt. Simon sandstone and the CO₂/brine induced evolution of its transport and mechanical properties
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Chemical Engineering
Publication Date
12/12/2019
Defense Date
09/13/2019
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
alteration of mechanical and transport properties,carbon sequestration,CO₂-brine-rock interactions,diffusion,natural convection,OAI-PMH Harvest
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Jessen, Kristian (
committee chair
), Hammond, Doug (
committee member
), Tsotsis, Theodore (
committee member
)
Creator Email
zhuofans@usc.edu,zhuofanshi@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-252010
Unique identifier
UC11673331
Identifier
etd-ShiZhuofan-8030.pdf (filename),usctheses-c89-252010 (legacy record id)
Legacy Identifier
etd-ShiZhuofan-8030.pdf
Dmrecord
252010
Document Type
Dissertation
Rights
Shi, Zhuofan
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
alteration of mechanical and transport properties
carbon sequestration
CO₂-brine-rock interactions
diffusion
natural convection