Reacciones de Equilibrio

Storage of Gases in Deep Geological Structures: Spatial and
Temporal Hydrogeochemical Processes Evaluated and Predicted

by the Development and Application of Numerical Modeling
Doctoral Thesis
(D i s s e r t a t i o n)
to be awarded the degree

Doctor rerum naturalium (Dr. rer. nat.)
submitted by
M. Sc. Christina Hemme
approved by the Faculty of Energy and Management Science,

Clausthal University of Technology,
Date of oral examination

26.02.2019
Chairperson of the Board of Examiners
Professor Dr. Hans-Jürgen Gursky (TU Clausthal)
Supervising tutor
Professor Dr. Klaus-Jürgen Röhlig (TU Clausthal)
Reviewer
PD Dr. Hans-Martin Schulz (Helmholtz Centre Potsdam)
Declaration
I hereby declare that I have created this Ph.D. thesis completely by my own and used no other
sources that the ones listed, and I have marked any citations accordingly.
Christina Hemme Clausthal-Zellerfeld, December 2018
I
Acknowledgements
I would like to thank Prof. Dr. Wolfgang van Berk †, who supervised me until July 2018, for his
scientific guidance, advice and helpful discussions throughout the years. He was always
interested in my work and willing to share his scientific knowledge.
I highly appreciate the help of Prof. Dr. Klaus-Jürgen Röhlig who kindly took over supervision
after the death of Prof. Dr. Wolfgang van Berk. Thank you for your valuable assistance and
scientific support in the end phase of my Ph.D. studies.
A big thank to PD. Dr. Hans-Martin Schulz for preparing the reviews of my thesis.
Special thanks to all colleagues of the Institute of Geology and Paleontology and of the Institute
of Disposal Research, especially Dr. habil. Elke Bozau.
I would like to thank my family for their help and support and - especially Nils Kollert - for your
proof-reading, patience and understanding.
† Prof. Dr. Wolfgang van Berk passed away on 29.07.2018
II
Kurzfassung
Bei der Speicherung von Kohlenstoffdioxid und Wasserstoff in ausgeförderten Erdöl- und
Erdgaslagerstätten und von Erdgas in Salzkavernen treten vielfältige hydrogeochemische
Prozesse auf, die zum Verlust der Gase, zu Veränderungen der Gaszusammensetzung oder
zu Problemen in der Abdichtung führen können.
Diese hydrogeochemischen Prozesse werden in der vorliegenden Arbeit identifiziert, mit

reaktiven 1D- und 3D-Stofftransportmodellen quantifiziert und mit ihrer räumlich-zeitlichen
Entwicklung abgebildet – auf der Grundlage i) chemisch-thermodynamischer
Gesetzmäßigkeiten, ii) der Gesetzmäßigkeiten der Reaktionskinetik der bakteriellen Sulfat-
Reduktion und Methanogenese und iii) der Gesetzmäßigkeiten des diffusiven Stofftransports.
Die verwendeten Modellierungsprogramme sind PHREEQC und PHAST, die beide vom
Geologischen Dienst der USA zur Verfügung gestellt werden.
In den Modellen werden hydrogeochemische Prozesse wie die Gas-Wasser-Gesteins

Reaktionen unter gegebenen Druck- und Temperaturverhältnissen simuliert, die durch die
Einspeicherung von Gasen im tiefen Untergrund ausgelöst werden. Ein mögliches Risiko
entsteht durch die Auflösung von Mineralphasen in den abdichtenden Gesteinsschichten, die
zu einer erhöhten Porosität und damit zu einer Verminderung der Abdichtung führt. Unter
bestimmten hydrogeochemischen und Druck- und Temperaturbedingungen kann es durch
Bakterien zu einer Umwandlung der gespeicherten Gase und damit zum Verlust der
gespeicherten Energie kommen. Dabei nutzen Sulfat-reduzierende Bakterien das
gespeicherte und gelöste Gas (Methan und/oder Wasserstoff), um vorhandene Sulfat-Ionen in
den wässrigen Lösungen (SO42-(aq)) zu Sulfid-Schwefel zu reduzieren – und damit Energie für
ihre Lebensprozesse zu gewinnen.
Die Modellierungsergebnisse erlauben es, Bedingungen zu ermitteln, die eine sichere

Speicherung der Gase gewährleisten. Zudem können anhand der Modellierungsergebnisse
diejenigen Faktoren identifiziert und deren Wirkungen quantifiziert werden, die auf die
räumlich-zeitliche Entwicklung der Gasverluste einwirken. Mit der Hilfe solcher
hydrogeochemischen Modellierungen kann die Speicherung von Gasen im tiefen Untergrund
im Hinblick auf Sicherheit und Produktivität optimiert werden.
III
Abstract
Storage of carbon dioxide and hydrogen in depleted hydrocarbon reservoirs and natural gas
storage in salt caverns trigger hydrogeochemical processes which can lead to a loss of the
stored gas, a change in the gas composition, and a loss of the sealing capacities of the cap
rock.
In this study, these different hydrogeochemical processes will be identified, quantified by 1D

and 3D reactive mass transport models, and predicted in terms of their spatial-temporal
development based on i) equilibrium thermodynamics, ii) kinetic reactions for sulfate reduction
and methanogenesis, and iii) diffusive mass transport. The modeling codes used are
PHREEQC and PHAST, both provided by the US Geological Survey.
In these models, hydrogeochemical processes, like gas–water–rock interactions under given

pressure and temperature conditions induced by the storage of gases in deep geological
structures, are simulated. A potential risk arises from the dissolution of mineral phases in the
cap rock that can lead to an increased porosity and, therefore, to a decrease in the sealing
capacity. Under specific pressure and temperature conditions, bacterial conversion of the
stored gases can lead to a loss of the stored energy. Sulfate-reducing bacteria use the stored
and dissolved gas (methane and/or hydrogen) to reduce available sulfate-ions in the aqueous
solutions (SO42-(aq)) to sulfide-sulfur—to produce energy for growth.
The modeling results allow the identification of storage conditions that ensure safe
underground gas storage. Furthermore, the modeling results help to identify the factors that
influence the spatial and temporal development of gas loss and quantify their effects.
Hydrogeochemical modeling provides an efficient tool for safe and productive underground
gas storage in deep geological structures.
IV
Preface
This cumulative Ph.D. thesis is based on three published (and peer-reviewed) articles. The
topics were also presented at national and international conferences.
Papers
• Hemme, C., van Berk, W., 2018. Hydrogeochemical modeling to identify potential risks
of underground hydrogen storage in depleted gas fields. Applied Sciences, 8(11), 2282.
doi.org/10.3390/app8112282
• Hemme, C., van Berk, W., 2017. Potential risk of H2S generation and release in salt
cavern gas storage. Journal of Natural Gas Science and Engineering, 47, 114–123.
doi.org/10.1016/j.jngse.2017.09.007
• Hemme, C., van Berk, W., 2017. Change in cap rock porosity triggered by pressure
and temperature dependent CO2-water-rock interactions in CO2 storage systems.
Petroleum, 3, 96–108. doi.org/10.1016/j.petlm.2016.11.010
Oral Presentations at Conferences
• Hemme, C., van Berk, W., 2017. Risk of H2S contamination of natural gas stored in salt
caverns. GeoBremen (2017).
Poster Presentations at Conferences
• Hemme, C., van Berk, W., 2017. H2S Generation and Release in Salt Cavern Gas
Storage. AAPG 2017 Annual Convention & Exhibition in Houston, Texas.
http://www.searchanddiscovery.com/pdfz/documents/2017/42082hemme/ndx_hemm
epdf.html
• Hemme, C., van Berk, W., 2015. One-Dimensional Reactive Transport Modeling of CO2
Storage Systems - Change in Cap Rock Porosity Triggered by Pressure and
Temperature Dependent CO2-Water-Rock Interactions. AGU Fall Meeting, San
Francisco 2015.
V
Contents
Declaration ........................................................................................................................................... I
Acknowledgements ............................................................................................................................. II
Kurzfassung ........................................................................................................................................ III
Abstract............................................................................................................................................... IV
Preface ................................................................................................................................................ V
1 Introduction .................................................................................................................................... 1
1.1 Review of Underground Gas Storage—State of Research ..................................................... 1

1.1.1 Underground Gas Storage—Risks and Potentials .......................................................... 3
1.1.2 Hydrogeochemical Processes—Mineral Dissolution and Precipitation ........................... 4
1.1.3 Pressure and Temperature Conditions ........................................................................... 4
1.1.4 Microbial Processes in Underground Gas Storage ......................................................... 5
1.1.5 Chemical and Physical Properties of the Stored Gases ................................................. 6
1.2 Aims of This Study ................................................................................................................... 7
1.3 Hydrogeochemical Reactive Mass Transport Modeling .......................................................... 8
1.3.1 Basics of Hydrogeochemical Modeling ........................................................................... 8
1.3.2 Thermodynamic Equilibrium Reactions ........................................................................... 9
1.3.3 Reaction Kinetics ........................................................................................................... 10
1.3.4 Reactive Mass Transport ............................................................................................... 10
1.3.5 Modeling Tools .............................................................................................................. 11
1.3.5.1 PHREEQC .............................................................................................................................. 12
1.3.5.2 PHAST .................................................................................................................................... 12
1.3.5.3 Databases ............................................................................................................................... 13
1.4 Hydrogeochemical Modeling Approaches—Similarities and Differences ............................. 14
2 Carbon Capture and Storage (CCS) .......................................................................................... 18
Change in Cap Rock Porosity Triggered by Pressure and Temperature Dependent CO 2–Water–
Rock Interactions in CO2 Storage Systems ....................................................................................... 18
Summary ........................................................................................................................................... 19
Abstract.............................................................................................................................................. 20
2.1 Introduction and Aim .............................................................................................................. 20
2.2 Kinetic or Thermodynamic Equilibrium Modeling for Long-Term CO2 Storage? ................... 22
2.3 Conceptual Numerical Model and Modeling Parameters ...................................................... 23
2.3.1 Model Setup................................................................................................................... 23
2.3.2 Initial Pressure and Temperature Conditions ................................................................ 25
2.3.3 Initial Mineralogical Composition of the Reservoir and Cap Rock and Initial Brine
Compositions ................................................................................................................................. 26
2.4 Methodology .......................................................................................................................... 28
2.4.1 Discretization ................................................................................................................. 29
VI
2.5 Results and Discussion ......................................................................................................... 31
2.5.1 Reference Scenario ....................................................................................................... 31
2.5.2 Alternative Scenarios ..................................................................................................... 35
2.5.2.1 Temperature / Pressure .......................................................................................................... 36
2.5.2.2 Veins ....................................................................................................................................... 38
2.5.2.3 Initial Cap Rock Porosity ......................................................................................................... 38
2.5.2.4 Effective Diffusion Coefficient ................................................................................................. 38
2.5.2.5 Injected Fluid Composition ...................................................................................................... 39
2.5.2.6 Initial Conditions According to Published Case Study ............................................................. 39
2.5.2.7 Advection as Transport Mechanism ........................................................................................ 40
2.5.2.8 Changing Temperature at a Constant Pressure...................................................................... 41
2.5.2.9 Changing Pressure at Constant Temperature......................................................................... 41
2.5.3 Sensitivity Analysis ........................................................................................................ 42
2.5.4 Modeling Limitations ...................................................................................................... 42
2.6 Conclusions ........................................................................................................................... 43
2.7 Appendix ................................................................................................................................ 46
2.8 Acknowledgements ............................................................................................................... 46
2.9 Supplementary Data (on CD) ................................................................................................ 46
3 Salt Cavern Gas Storage (SCGS) ............................................................................................... 47
Potential Risk of H2S Generation and Release in Salt Cavern Gas Storage .................................... 47
Summary ........................................................................................................................................... 48
Abstract.............................................................................................................................................. 49
3.1 Introduction ............................................................................................................................ 49
3.2 Methodology .......................................................................................................................... 52
3.2.1 Modeling Tools .............................................................................................................. 52
3.2.2 Model Setup................................................................................................................... 54
3.3.1 H2S Generation and Release—Reference Scenario ..................................................... 58
3.3.2 Factors Influencing H2S(g) Generation and Release ...................................................... 60
3.3.2.1 Gas Pressure Changes in Salt Cavern Gas Storages ............................................................ 60
3.3.2.2 Natural Gas Composition ........................................................................................................ 61
3.3.2.3 Kinetic Rate Constants ........................................................................................................... 61
3.3.2.4 Tortuosity ................................................................................................................................ 62
3.3.3 Inhibition of H2S(g) Generation and Release in Salt Caverns ........................................ 64
3.4 Conclusion ............................................................................................................................. 67
4 Underground Hydrogen Storage (UHS) .................................................................................... 69
Hydrogeochemical Modeling to Identify Potential Risks of Underground Hydrogen Storage in

Depleted Gas Fields .......................................................................................................................... 69
VII
Summary ........................................................................................................................................... 70
Abstract.............................................................................................................................................. 71
4.1 Introduction and Aims ............................................................................................................ 71
4.2 Methodology .......................................................................................................................... 75
4.2.1 Modeling Tools .............................................................................................................. 75
4.2.2 Model Setup................................................................................................................... 76
4.3.1 Loss of H2(g/aq) by Bacterial Conversion to CH4(g) and H2S(g) ......................................... 81
4.3.2 Hydrogeochemical Effects of Hydrogen Storage on Reservoir Rock and Cap Rock .... 85
4.3.3 Loss of Aqueous H2(aq) by Diffusion through the Cap Rock ........................................... 87
4.3.4 Influencing Factors ........................................................................................................ 89
4.3.4.1 Storage Time .......................................................................................................................... 90
4.3.4.2 Pressure and Temperature Conditions in Gas Reservoirs ...................................................... 91
4.3.4.3 Kinetic Rate Constant ............................................................................................................. 93
4.3.4.4 Stored Gas Composition ......................................................................................................... 94
4.4 Conclusions ........................................................................................................................... 95
5 Final Discussion .......................................................................................................................... 97
5.1 General Model Similarities and Differences .......................................................................... 98

5.2 Influencing Factors .............................................................................................................. 100
5.3 Optimal Storage Conditions ................................................................................................. 102
5.4 Model Limitations ................................................................................................................. 102
6 Final Conclusion and Outlook.................................................................................................. 104
List of Figures .................................................................................................................................. 105

List of Tables ................................................................................................................................... 107
Abbreviations ................................................................................................................................... 108
References ...................................................................................................................................... 109
VIII
Introduction
1 Introduction
Storage of gases in deep geological formations is of increasing interest due to the new German
energy concept, which aims to reduce greenhouse gas emissions and expand renewable
energies. Furthermore, increasing natural gas consumption leads to the strategic importance
of future energy supply and access to efficient energy storage systems (Federal Ministry for
Education and Research, 2011). This work was created in light of these debates and
transformation processes. This study focuses on the hydrogeochemical processes relevant to
the underground storage of gases in deep geological structures and related potentials and
risks. These processes were evaluated and predicted in space and time by the development
and application of numerical modeling. Therefore, the study is divided into three subtopics:
(i) carbon capture and storage (CCS),
(ii) salt cavern gas storage (SCGS),
(iii) underground hydrogen storage (UHS).
These technical means of storing gases in the underground trigger a series of

hydrogeochemical processes. For successful underground gas storage, the analysis of gas–
water–rock interactions is important because they can affect the storage efficiency and the
sealing capacity of the cap rocks as well as the stored gas composition.
1.1 Review of Underground Gas Storage—State of Research
Storing gases in the underground has a long history starting with the first gas implementation
in 1915 in Welland County, Ontario (Canada) in a depleted gas reservoir (Evans, 2008). Since
that point, both research and practical experience in underground gas storage have evolved.
Research in underground gas storage links a wide field of disciplines, including geology,
geochemistry, oil and gas geology, engineering, mechanics, and social sciences.
Natural gas storage in the underground is used to provide flexibility in times of increasing
dependence on natural gas imports. Germany imports around 91% of its natural gas and
operates 57 storage facilities in total; 20 depleted oil and gas fields, 31 salt caverns, and 6
aquifers in 2016 (LBEG, 2016). In comparison, the global number of underground storage
facilities amounts to 498 depleted oil and gas fields, 97 salt caverns, and 77 aquifers at the
end of 2016 (Cornot-Gandolphe, 2017). The total volume of working gas capacity is 23.9 bcm
in Germany versus 424 bcm worldwide (Cornot-Gandolphe, 2017). The storage of natural gas
1
Introduction
in salt caverns has been studied with respect to the design of storage caverns (Crotogino et
al., 2006), the geomechanics to analyze potential cavern instability and collapse (DeVries et
al., 2002), and the economic aspects regarding low investment costs (Farahani et al., 2015).
Furthermore, the interaction of storage gas with surface facilities, including computational-fluid-
dynamics and heat-transfer-modeling, have been analyzed (Barajas and Faruk, 2014).
In addition to natural gas, the storage of carbon dioxide in the underground has gained in
importance because of rising global carbon dioxide emissions. With CCS, the consequences
of the additional greenhouse gases on climate change could be mitigated (IPCC, 2005).
Germany operated only the in-situ testing site Ketzin, whereas worldwide carbon dioxide is
stored in several projects e.g. Sleipner (North Sea), Weyburn (Canada) or In Salah (Algeria).
Intensive studies on CCS were performed, including numerical simulation studies of CO 2
leakage (Pruess, 2013), related monitoring studies to detect leaks of carbon dioxide (Arts et
al., 2004; Blackford et al., 2015), and the migration of carbon dioxide (Arts et al., 2004).
Furthermore, seismic studies were conducted to discuss seismic activity due to CCS (Verdon
and Stork, 2016). CCS technologies, like post-combustion, pre-combustion and oxy-
combustion, are also the subject of debate (Araújo and de Medeiros, 2017; Leung et al., 2014).
The possibility of enhanced oil recovery (EOR) by injection of compressed carbon dioxide into
oil reservoirs is widely discussed (Emberley et al., 2005; Reichl et al., 2015).
Another gas that is discussed for potential storage in deep geological structures is hydrogen
to balance the demand for energy from renewable energy sources. Hydrogen storage is known
from town-gas storage, where a gas composed of H2(g), CO2(g), CO(g), CH4(g) and N2(g) is stored
in an underground reservoir (Šmigáň et al., 1990). It is also known from Russia where it is
stored for the aerospace industry (Panfilov, 2010; Ponomarev-Stepnoi, 2004) and from the US
and the UK where hydrogen is already stored in salt caverns (Crotogino et al., 2010; Henkel
et al., 2013; Panfilov, 2010; Stone et al., 2009). Special focus is also given in the literature on
the capacity and efficiency of UHS (Crotogino et al., 2010), on the effects of hydrogen storage
on the reservoir and cap rocks (Henkel et al., 2014; Truche et al., 2013), and on the integrity
of wells as well as on material durability (Somerday and Gangloff, 2012). Furthermore,
experimental studies have been performed in which the reactivity of hydrogen in sandstone
was evaluated (Yekta et al., 2018). Additionally, possible biotic and abiotic reactions are
analyzed in detail by Panfilov et al. (2006) and Panfilov (2016).
2
Introduction
1.1.1 Underground Gas Storage—Risks and Potentials
Generally, gases can be stored in depleted oil and gas fields, aquifers, and salt caverns. All
these storage options show advantages and disadvantages. The main advantages for
depleted oil and gas fields are that the fields are fully explored, the technical installations and
infrastructure are already available from production and the sealing capacity over geological
timescales is verified. Disadvantages are the low operational flexibility and a high amount of
cushion gas (Crotogino, 2011). Furthermore, hydrocarbons remaining from production are still
present in the fields and could contaminate the stored gas and pose a risk for its quality.
Another disadvantage is the possible risk of reactions between the stored gas and the rocks,
fluids, and microorganisms present in depleted oil and gas fields (Crotogino, 2011).
Storing gases in aquifers entails the risk of reaction between the stored gas and the original
rocks, fluids, and microorganisms as well. A high amount of cushion gas and low flexibility are
further disadvantages of storing gas in aquifers. Furthermore, the pressure loss is increased
when the injection and production rates are high. The information density about the aquifers is
very low in most cases; therefore, the costs of exploring and verifying the sealing capacity are
high (Crotogino, 2011).
The main advantage of storing gases in salt caverns is the high sealing capacity of the salt
rock and its impermeability below 300 m (Evans, 2008; Yang et al., 2013). Another advantage
is the high deliverability rate because of the high pressure in salt caverns. Furthermore, the
ability to shift between injection and withdrawal is possible in a short period (Foster Associates
Inc., 1995). The main disadvantages are that the caverns must be created in advance and that
the construction period could last a few years. Additionally, the created brine must be disposed
of in an environmentally friendly way (Crotogino, 2011).
The health, environmental and economic risks that could arise when storing gases
underground are discussed in detail by Evans (2008), who focuses on the risk of stored product
leakage to overlying formations or to the surface. However, all storage options have in common
that the underground structures selected for gas storage should have the capacity to store the
gases with no significant leakage. If large quantities of gas have to be stored, it is cheaper to
store it underground than in containers (Panfilov, 2010). Storage in salt caverns is the cheapest
solution, followed by storage in aquifers. The most expensive is storage in depleted gas
reservoirs (Taylor et al., 1986).
A general problem arising from underground gas storage is the social implications. Storage of
gases underground leads to an area of conflict between science, politics, economy, and public
opinion. Public acceptance is a fundamental prerequisite for underground gas storage. Stone
3
Introduction
et al. (2009) describe that the public of the UK accepts underground gas storage in principle
as long as it does not affect their locality. For that reason, salt cavern gas storage in Thornton
has been stopped (Beutel and Black, 2005; Stone et al., 2009). In general, acceptance
depends mainly on the societal benefit and the political conflict (Pietzner, 2012).
Comprehensive communication with the public is recommended (Pietzner, 2012), as has been
done in the German Ketzin CCS pilot-project (Schilling et al., 2009). However, public
acceptance and the concrete inclusion of CCS in the political energy concept is missing in
Germany (Pietzner, 2012; Vögele et al., 2018). The difficulties when striving for acceptance of
underground gas storage is also seen in other underground operations like fracking and
radioactive waste disposal.
In addition to natural gas, hydrogen and carbon dioxide, substances like compressed air,
thermal energy, crude oil, geothermal energy, and radioactive waste must be stored in the
underground. Therefore, the demand for underground storage options rises. In some cases,
the same storage options are possible for different storage mediums (Crotogino, 2011) so that
multiple conflicts of interest regarding the demand of underground storage capacities come up
(Zander-Schiebenhöfer et al., 2014). To counter this, subsurface spatial planning is required,
(Schulze et al., 2015) and a range of conditions for safe storing options for each storage
product must be identified.
1.1.2 Hydrogeochemical Processes—Mineral Dissolution and Precipitation
An important aspect of underground gas storage is the reaction of the stored gases with the
minerals of the reservoir rock and the cap rock, which can lead to mineral dissolution and
precipitation and resulting in porosity changes, as is known from CCS (e.g. Bildstein et al.
(2010), Gaus et al. (2008), Mohd Amin et al. (2014)) and from UHS (Amid et al., 2016). Porosity
changes in the cap rock can either improve or deteriorate its sealing capacity (Gaus et al.,
2005). Furthermore, the precipitation of minerals at the well equipment may cause scaling
(Reitenbach et al., 2015).
1.1.3 Pressure and Temperature Conditions
In Germany, salt caverns are at depths of 500–2000 m and depleted gas fields are at depths
of 350–2930 m (LBEG, 2016). Temperature and pressure conditions are correspondingly high,
considering the geothermal gradient of 33.3 °C per kilometer of depth and the pressure
4
Introduction
increase of 100 atm per kilometer of depth under hydrostatic conditions. These higher pressure
and temperature conditions must be considered when modeling CCS, SCGS, and UHS
because temperature and pressure could strongly affect the mass-action law constants of the
equilibrium reactions and the kinetic rate constants. Authors like Bildstein et al. (2010) and Van
Pham et al. (2012) considered higher pressure and temperature conditions at starting
conditions when modeling CO2 storage in deep geological structures but disregarded that
temperature and pressure conditions change along the path of diffusive transport of the stored
gases.
1.1.4 Microbial Processes in Underground Gas Storage
Some important processes when storing gases in the underground are the microbial activities,
including bacterial sulfate reduction, methanogenesis, acetogenesis, and iron(III)-reduction,
summarized by Hagemann et al. (2016). The stored gases are microbially catalyzed and
converted by bacteria. The activity of the different bacteria depends on the availability of
different electron acceptors such as sulfate or carbon dioxide (Ebigbo et al., 2013). The energy
that is gained from conversion is used by the bacteria for cell growth (Cord-Ruwisch et al.,
1987; Šmigáň et al., 1990; Whitman et al., 2006). The influences of these processes in
underground gas storage systems are extensively discussed in the literature (Ebigbo et al.,
2013; Hagemann et al., 2016; Kleinitz and Böhling, 2005; Panfilov, 2010).
An important microbial process when storing natural gas and hydrogen in deep geological
structures is the bacterial sulfate reduction (BSR). BSR occurs in low-temperature geological
settings ranging from 0 °C to a maximum of 110 °C (Bernardez et al., 2013; Ehrlich, 1990;
Jorgensen et al., 1992; Machel, 2001; Postgate, 1984). The sulfate-reducing bacteria (SRB)
use sulfate as an electron acceptor to oxidize organic compounds and generate sulfide in
aqueous anoxic environments. Even if the optimal growth conditions for SRB are at near-
neutral pH (Cord-Ruwisch et al., 1987), SRB also occur in more acidic environments of pH 3–
4 (Church et al., 2007; Tuttle et al., 1969). The activity of SRB has been observed in a wide
range of environments, for example, in hydrocarbon reservoirs (Machel, 2001), in underground
gas storage in porous media (Kleinitz and Böhling, 2005), in the underground storage of town
gas (Postgate, 1979) and in salt caverns filled by hydrogen gas (Panfilov, 2016).
Another microbial process that is important for hydrogen storage is methanogenesis. Here,
hydrogen is the electron donor and carbon dioxide the electron acceptor, which is reduced to
form methane. Methanogenic bacteria have been found in environments with temperatures up
to 97 °C (Gusev and Mineeva, 1992; Panfilov, 2016). The existence of methanogenic bacteria
5
Introduction
has been observed, for example, in anoxic sediments, flooded soils (Whitman et al., 2006),
and in town gas storages (Šmigáň et al., 1990).
1.1.5 Chemical and Physical Properties of the Stored Gases
The chemical and physical properties of the stored CO2(g), CH4(g) and H2(g) are extensively
discussed in the literature, including e.g. the effective diffusion coefficients, the viscosity, and
the solubility in water. For a comparison of these gases, the properties are summarized in
Table 1.1.
Table 1.1 Chemical and physical properties of the stored gases.
Carbon dioxide Methane Hydrogen

Chemical formula CO2 CH4 H2
2.35-2.49 × 10-10 3.0 × 10-11 (in
Effective diffusion coefficient
3.33 × 10-10 a (in clayey rocks clayey rocks at
(m2 s-1)
at 21 °C)b 25 °C)c
Diffusion coefficient in pure
1 × 10-9 a 1.85 × 10-9 d 5.13 × 10-9 d
water at 25 °C (m2 s-1)
Solubility in pure water at
3.5 × 10-2 d 1.4 × 10-3 d 7.9 × 10-4 d
25 °C and 1 atm (mol kgw-1)
9.41 × 10-5 (at 1.81 × 10-5 (at 1.01 × 10-5 (at
-1 -1
Viscosity (kg m s ) 197.4 atm; 197.4 atm; 167.8 atm;
26.85 °C)e 76.85 °C)f 76.85 °C)g
Solubility constant log K -1.468d -2.8502d -3.1050d
Critical temperature (°C) 31.05d -82.55d -239.95d
Critical pressure (atm) 72.86d 45.40d 12.80d
a
: based on assumptions made by Mohd Amin et al. (2014): diffusion coefficient in pure water of 1 × 10-9 m2 s-1
divided by an averaging tortuosity of 3
b
: Jacops et al. (2013)
c
: Krooss (2008)
d: based on phreeqc.dat
(e): Laesecke and Muzny (2017)

f
: Friend et al. (1989)
g
: McCarty et al. (1981)
Considered as a whole, the modeling of hydrogeochemical processes induced by gas storage

becomes increasingly important. Hydrogeochemical studies were performed for CCS by e.g.
Gaus et al. (2005) and Bildstein et al. (2010) and for UHS by Amid et al. (2016). These
processes are still not fully understood and need further research to gain a comprehensive
process understanding.
6
Introduction
1.2 Aims of This Study
The aim of this study is to contribute to a general system and process understanding of the
hydrogeochemical effects related to underground gas storage in deep geological structures on
aqueous solutions, mineral phases and gas compositions in the storage systems. Furthermore,
one main objective is to provide hydrogeochemical model approaches which can be used and
adapted to various underground gas storage systems and conditions to answer and solve
practical oriented questions and problems. This includes the identification and analysis of the
hydrogeochemical processes induced by the storage of CO2(g), CH4(g), and H2(g) in deep
geological formations. Special attention is given to the gas–water–rock interactions, spatial
and temporal dependence, and on temperature and pressure influenced reactions. Thereby,
the risks and potentials of storing gases in the underground will be analyzed with the aim to
answer the following main questions:
1) How does the storage of CO2(g/sc) influence the sealing capacity of the cap rocks?
To answer this question, the hydrogeochemical processes, including gas–water–rock

interactions, induced by carbon capture and storage are identified. The change in porosity of
the cap rock because of geochemical reactions are quantified in dependence on pressure and
temperature.
2) Is H2S(g) generation a risk when storing natural gas in salt caverns?
The aim of the second part of this study is to draw the attention to the possible risk of H 2S(g)
pollution in salt caverns filled with natural gas, to clarify and quantify time-dependent H2S(g)
generation processes, and to analyze the limiting factors for H2S(g) generation and release. In
the course of this, the viability of reducing or inhibiting H2S(g) generation and release is tested.
3) Which factors influence the loss of hydrogen in depleted gas storage?
The aim of the third part of the study is to investigate the loss of hydrogen by bacterial
conversion over storage times at reservoir scale and to describe qualitatively and quantitatively
which reservoir and cap rock minerals dissolve or newly form because of hydrogen storage.
Furthermore, the loss of hydrogen by diffusion is analyzed.
7
Introduction
1.3 Hydrogeochemical Reactive Mass Transport Modeling
1.3.1 Basics of Hydrogeochemical Modeling
A model is a representation of a real system or process (Konikow and Bredehoeft, 1992). The
basis of a model is the definition of a conceptual model with the hypothesis for how a system
or process operates (Konikow and Bredehoeft, 1992). Modeling is based on a multistage
process of abstraction whereby the results of each stage can be understood as a model of the
previous stage. A conceptual model can be expressed quantitatively by mathematical models,
where abstractions replace objects, events, and forces, and also by using expressions that
involve mathematical constants, variables, and parameters (Konikow and Bredehoeft, 1992;
Krumbein and Graybill, 1965). The results of the numerical computer model are subsequently
compared with the real system or process.
Hydrogeochemical systems are defined by one or more phases which can be aqueous(aq),
gaseous(g) and/or solid(s) (Stumm and Morgan, 1981). Reactions take place in single
phases(aq, g, s) as well as between the phases like i) aqueous species(aq) and solid(s) phases, ii)
aqueous species(aq) and gases(g) and iii) solid(s) phases and gases(g), the so-called gas–water–
rock interactions.
For numerical modeling of gases stored in deep geological formations, chemical-

thermodynamic or kinetically based model approaches can be chosen. Thermodynamic
systems are defined by variables like pressure, temperature, density, volume, concentration
of dissolved components, viscosity, etc. Chemical-thermodynamic based models describe the
initial and final state of irreversible reactions. The kinetic-based model must be chosen, when
the change in concentrations of the involved chemical substances in an irreversible reaction
must be considered.
Models should always be tested with sensitivity and parameter studies to identify the main
influencing factors (Konikow and Bredehoeft, 1992). For building confidence in the model,
different approaches can be applied. This includes quality assurance (of e.g. the scientific
bases), comparison with other models (e.g. code verification), and real systems (including
laboratory experiments, field tests, and natural analogs). With the help of models, a critical
analysis is provided (Konikow and Bredehoeft, 1992) and modeling approaches could be
tested.
8
Introduction
1.3.2 Thermodynamic Equilibrium Reactions
With the thermodynamics of chemical equilibrium, reactions and their direction and conversion
until equilibration are described. Chemical equilibrium is defined by the law of mass action
(Equation (1.1)) and related equilibrium constants. Equilibrium constants are based on the
properties of the atoms, molecules, ions, and phases which are involved in the reactions. They
are functions of temperature, pressure, and composition (Stumm and Morgan, 1981) and are
independent of the hydrogeochemical environment (van Berk and Hansen, 2006). The data
required for equilibrium models, apart from equilibrium constants, are the Gibbs free energies,
enthalpies, entropies, and heat capacities (Stumm and Morgan, 1981).
The law of mass action describes reversible equilibrium reactions (Equation (1.1)).
{𝐶}𝑐 ∙ {𝐷}𝑑
𝐾= {𝐴}𝑎 ∙ {𝐵}𝑏
(1.1)
where K is the thermodynamic equilibrium constant (-), a, b, c, and d are stoichiometric

coefficients (-), A and B are the reactants (mol L-1), and C and D are the reaction products
(mol L-1).
The dissolution of a mineral AB to its components A and B is described by the law of mass
action as follows (Equation (1.2)):
{𝐴} ∙ {𝐵}
𝐾= {𝐴𝐵} (1.2)
Because the activity of solid phases is assumed to be constant at 1, the solution product L or
the ion activity product IAP is derived from the thermodynamic equilibrium constant K
(Equation (1.3)).
𝐿 = 𝐼𝐴𝑃 = {𝐴} ∙ {𝐵}

(1.3)
Based on L and IAP, the saturation index (SI) of a mineral phase can be calculated. SI
describes if a mineral phase is in equilibrium with the aqueous solution or if the aqueous
solution is over- or undersaturated regarding a specific mineral phase (Equation (1.4)). If SI < 0
the mineral phase dissolves. If SI = 0 the phase is in equilibrium and with SI > 0 the mineral
phase precipitates.
𝐼𝐴𝑃
𝑆𝐼 = 𝑙𝑜𝑔10 ( ) (1.4)
𝐿
9
Introduction
IAP is calculated from the concentrations of a given water sample by recalculation as activity
and L is defined as the maximal solubility (given in databases). The temperature dependence
must be considered.
1.3.3 Reaction Kinetics
While the thermodynamic-equilibrium reaction assumes a time-independent stable state for a

system, with reaction-kinetics an unstable chemical state at any time of reaction can be
calculated. This can be relevant if slow irreversible, reversible or heterogeneous reactions
(reactions between the different phases: aqueous, gaseous, and solid) are included in the
system (Merkel and Planer-Friedrich, 2002). For kinetic-based models, either the values for
the rate constants or half-lives are needed. The rate constants are functions of pressure,
temperature and solution composition (Stumm and Morgan, 1981). These rate constants and
half-lives describe the reaction rates and the mechanisms of the defined reaction. Because
they are dependent on pressure, temperature, and solution composition, they are not
universally valid constants. These constants were determined in laboratory and field studies
and are, unlike the equilibrium constants, dependent on the hydrogeochemical environment.
1.3.4 Reactive Mass Transport
Reactive mass transport models simulate the transport of aqueous species coupled with the
chemical reactions to show the temporal and spatial development of hydrogeochemical
processes. With the help of reactive mass transport coupled chemical, physical and biological
processes can be analyzed (Steefel et al., 2005). The main transport processes of aqueous
species are advection, dispersion, and molecular diffusion. Advection describes the movement
of aqueous species with the velocity of the flowing fluid (Equation (1.5)).
𝑗𝑎𝑑𝑣 = 𝜈 ∙ 𝐶
(1.5)
where jadv is the advective mass transport (mol s-1 m-²), v is the flow velocity (m s-1), and C is
the species concentration (mol L-1).
10
Introduction
Dispersion describes the changes in aqueous species concentration by advective mass

transport in the fluid induced by differences in geometry and structure of the rock (Equations
(1.6) and (1.7)).
𝛿𝐶
𝑗𝑑𝑖𝑠𝑝𝐿 = −𝐷𝐿 (1.6)
𝛿𝑠𝐿
𝛿𝐶
𝑗𝑑𝑖𝑠𝑝𝑇 = −𝐷𝑇 (1.7)
𝛿𝑠𝑇
where jdispL is the longitudinal dispersive mass transport (mol s-1 m-²), jdispT is the transversal
dispersive mass transport (mol s-1 m-²), DL is the longitudinal dispersion coefficient (m² s-1), DT
is the transversal dispersion coefficient (m² s-1), δC/δsL is the species concentration gradient in
longitudinal direction (mol L-1), and δC/δsT is the species concentration gradient in transversal
direction (mol L-1).
The diffusion is described by Fick´s law (Equation (1.8)). The diffusive mass transport is
induced by concentration gradients.
𝛿𝐶
𝑗𝑑𝑖𝑓𝑓 = −𝐷𝑒 (1.8)
𝛿𝑥
where jdiff is the diffusive mass transport (mol s-1 m-²), De is the effective diffusion coefficient
(m² s-1), and δC/δx is the species concentration gradient in x-direction (mol L-1).
Reactive mass transport can be explained by transport equations in 1D (Equation (1.9),
Section 1.3.5.1) or 3D (Equation (1.10), Section 1.3.5.2).
1.3.5 Modeling Tools
The programs used in this study for hydrogeochemical modeling are PHREEQC (Parkhurst
and Appelo, 2013) and PHAST (Parkhurst et al., 2010), both provided by the US Geological
Survey. PHREEQC and PHAST connect hydrogeochemical and geohydraulic models to
reactive mass transport. Further programs with the capability for hydrogeochemical modeling
are The Geochemist´s Workbench (Bethke and Yeakel, 2009), EQ 3/6 (Wolery, 1992),
MINTEQA2 (Allison et al., 1991), and WATEQ4F (Ball and Nordstrom, 1991).
11
Introduction
1.3.5.1 PHREEQC
PHREEQC (= pH-Redox-Equilibrium written in the C programming language) is a computer

simulation program with the capability to calculate batch-reaction, one-dimensional (1D)
transport, speciation, and inverse geochemical calculations. It is written in the programming
languages C and C++ and can be used for aqueous geochemical modeling in a wide field of
application and in a wide range of hydrogeochemical environments (Parkhurst and Appelo,
2013). The basis of PHREEQC is the equilibrium chemistry of aqueous solutions that interact
with gases, minerals, solid solutions, sorption surfaces, and exchangers, but it is also possible
to include kinetically-controlled reactions whereby the interconnection of equilibrium and
kinetic reactants is possible.
The transport calculations (1D) in PHREEQC are used to model the movement of solutions
through a column, where the initial compositions of aqueous solutions, minerals, and gases
are pre-defined and the changes induced by transport (Appelo and Postma, 2005) are
simulated and are coupled with irreversible and reversible chemical reactions (Parkhurst and
Appelo, 2013). PHREEQC has the capability to simulate advective-transport calculations as
well as advective-dispersive and/or diffusive transport calculations (Parkhurst and Appelo,
2013). Reactive mass transport in one dimension can be explained by the Advection-Reaction-
Dispersion Equation (ARD) (1.9), which is a second order partial differential equation.
𝜕𝐶 𝜕𝐶 𝜕 2 𝐶 𝜕𝑞
= −𝑣 + 𝐷𝐿 2 − (1.9)
𝜕𝑡 𝜕𝑡 𝜕𝑥 𝜕𝑡
where 𝐶 is the concentration in water (mol kgw-1), t is the time (s), 𝑣 is the pore water flow
velocity (m s-1), 𝐷𝐿 is the hydrodynamic dispersion coefficient [m2 s-1, 𝐷𝐿 = 𝐷𝑒 + 𝛼𝐿 𝑣, with 𝐷𝑒
the effective diffusion coefficient and 𝛼𝐿 the dispersivity (m)], x is the distance (m), and 𝑞 is the
concentration in the solid phase (expressed as mol kgw-1 in the pores). Numerical solution
methods for the mass-transport-equation coupled to chemical reactions use the finite volume,
finite element, or finite difference method.
1.3.5.2 PHAST
With PHAST (= PHREEQC And HST3D), 3D models with 3D mass transport, diffusion, and
chemical reactions can be simulated. Using PHAST, multicomponent geochemical reactions,
solute transport, and groundwater flow can be modeled (Parkhurst et al., 2010). The
geochemical reactions in PHAST are simulated with PHREEQC, and the flow and transport
12
Introduction
calculations are based on HST3D; both programs are embedded in PHAST (Parkhurst et al.,
2010). The results are visualized using the software Model Viewer (Hsieh and Winston, 2002).
The transport equation used in PHAST is different from that in PHREEQC because a 3D
transport equation, which considers equilibrium and kinetic reactions, is applied (Equation
(1.10)). The equation is modified by Parkhurst et al. (2010) from Bear (1972) and Kirkner and
Reeves (1988).
𝑁𝐸 𝑁𝐾
𝛿 𝐸 𝛿 𝐾
(𝜀𝜌𝑐𝑗 ) = 𝛻 ⋅ 𝜀𝐷𝛻𝜌𝑐𝑗 − 𝛻 ⋅ 𝜀𝑣𝜌𝑐𝑗 − ∑ 𝑣𝑗,𝑒 (𝜀𝜌𝑐̅𝑒 ) + ∑ 𝑣𝑗,𝑘 𝜀𝜌𝑅𝑘 + ɋ𝜀𝜌𝑐̂𝑗 (1.10)
𝛿𝑡 𝛿𝑡
𝑒=1 𝑘=1
𝑗 = 1, … , 𝑁𝐶
where 𝜀 is the saturated porosity (-), ρ is the density of water (kg m-3), 𝑣 is the interstitial velocity
(m s-1), cj is the total aqueous concentration of component j (mol kgw-1), t is the time (s), D is
the tensor of the dispersion-coefficient (m2 s-1), NE is the number of heterogeneous equilibrium
𝐸
reactions, 𝑣𝑗,𝑒 is the stoichiometric coefficient of component j in the heterogeneous equilibrium
reaction e (-), 𝑐̅𝑒 is the concentration of solid reactant e (mol kgw-1), NK is the number of kinetic
𝐾
reactions, 𝑣𝑗,𝑘 is the stoichiometric coefficient of component j in kinetic reaction k (-), Rk is the
kinetic reaction rate k (mol kgw–1 s–1), 𝑐̂𝑗 is the total aqueous concentration of component j in
the source water (mol kgw-1), and NC is the number of the chemical components in the system
(Parkhurst et al., 2010).
For both, PHREEQC and PHAST, the accuracy of the modeling results must be manually
checked by evaluating the mass balance. Numerical stability/instability must be proven
manually by discretization studies and by refining grid and time steps. For spatial and temporal
discretization, finite-difference techniques are used in PHREEQC and PHAST. Further detailed
information about PHREEQC and PHAST are given in Parkhurst and Appelo (2013) and
Parkhurst et al. (2010).
1.3.5.3 Databases
The programs PHREEQC and PHAST are based on databases and come with nine different
databases (Parkhurst and Appelo, 2013): phreeqc.dat (Parkhurst et al., 1980), wateq4f.dat
(Ball and Nordstrom, 1991), pitzer.dat (Pitzer, 1973), llnl.dat (derived from databases for
13
Introduction
The Geochemist´s Workbench and EQ3/6 using thermodynamic data prepared by the
Lawrence Livermore National Laboratory), minteq.dat (Allison et al., 1991), minteqv4.dat (U.S.
Environmental Protection Agency, 1999), sit.dat (Grenthe et al., 1997), iso.dat (Thorstenson
and Parkhurst, 2002; Thorstenson and Parkhurst, 2004), and amm.dat (same as phreeqc.dat,
but ammonia is considered as a separate component). Additional developed databases are
the “gebo”-database from Bozau (2013) which extends pitzer.dat for solutions with high ionic
strength and the thermochemical database thermoddemv1.10_11dec2014 from the BRGM
institute, French Geological Survey (Blanc et al., 2012). The database used in this study is
phreeqc.dat which is extended by data from llnl.dat.
1.4 Hydrogeochemical Modeling Approaches—Similarities and Differences
The numerical modeling of the hydrogeochemical processes induced by the storage of gases
in the underground is based on the equilibrium thermodynamics of chemical gas–water–rock
interactions, including kinetic reactions. The diffusive transport of aqueous species through the
solutions in the respective underground gas storage system is modeled to evaluate the time-
and space-dependent development in such reactive mass transport models. With the help of
the codes PHREEQC and/or PHAST, the effects of the hydrogeochemical processes induced
by gas storage on the hydrogeochemical compositions of the aqueous solutions, partial
pressure of stored gas phases, and on the mineralogical composition of the reservoir and cap
rocks are calculated, partly considering the changing porosities in reservoir and cap rocks. The
similarities and differences in these three subtopics are summarized in Figure 1.1 and details
are given in Table 1.2.
14
Introduction
• Fixed effective diffusion

Carbon capture and coefficient for all aqueous
• Depleted hydrocarbon reservoir storage species
• 1D
• porosity creation
• Underground gas storage

• Risks and potentials
• Gas loss
• Gas-water-rock interactions
• High P-T conditions
• Diffusive mass transport
Underground Salt cavern • Thermodynamic equilibrium
hydrogen storage gas storage • Semi-generic modeling approach
• Modeling software: PHREEQC
• Database: phreeqc.dat
• Gas phases: CH4, CO2, H2S, (H2)
• Influencing factors
• Kinetics
• Storage time: 30 years
Figure 1.1 Intersections of the three subtopics CCS, SCGS, and UHS showing conceptual similarities
and differences.
In all three subtopics of this study—CCS, SCGS and UHS—hydrogeochemical processes in

deep geological formations are modeled and analyzed. The modeling approaches are semi-
generic in nature but, whenever data from real geosystems and/or experimental data are
available, these data are used to check the plausibility of the models. In CCS a depleted
hydrocarbon reservoir, in SCGS a salt cavern, and in UHS a depleted gas reservoir is
considered. The mineralogical compositions of the reservoir rocks and the cap rocks are
different in CCS and SCGS and, of course, different from the mineralogical composition of the
salt cavern. However, typical minerals like quartz, calcite, pyrite, and barite are considered in
all three models. The considered brines are pre-calculated in separate batch models (CCS and
SCGS) and in a one-dimensional reactive transport model for UHS.
The multi-component gas compositions that are analyzed when storing gases in the
underground are different in their partial pressures but not in their composition. In all three
modeled gas storage systems CO2(g), CH4(g), and H2S(g) are considered. Additionally, N2(g) is
given in SCGS and UHS models and, of course, H2(g) in the UHS system.
All three systems are modeled as one-dimensional reactive mass transport (1DRMT) models.
Additionally, the SCGS system is modeled using a three-dimensional reactive mass transport
15
Introduction
(3DRMT) model as well. The modeling tool for the 1DRMT models is PHREEQC, while PHAST
is used for the 3DRMT model.
The mass transport mechanism in the underground gas storage models is molecular diffusion
of all aqueous species. Dispersion is not considered and advection is only considered in an
alternative model scenario in CCS. While the UHS model considers multi-component diffusion
(where each species has an individual temperature-dependent diffusion coefficient), the CCS
model uses an effective diffusion coefficient for all aqueous species of 3.33 × 10-10 m2 s-1, and
the SCGS model uses 5.0 × 10-9 m2 s-1.
The prerequisite for all considered models (and PHREEQC applicability) is the availability of
water; it is given as irreducible water and, additionally, as a brine in SCGS.
In all three models, high pressure and temperature conditions (> 40 °C and > 40 atm) are
simulated. Using the phreeqc.dat database, the temperature and pressure effect on the
equilibrium constants are considered for almost all phases and aqueous species—with the
exception of the acetate-bearing species and a few aqueous silicon- and aluminum-bearing
species.
The changes in the porosity of the cap rock are addressed for CCS and UHS. Additionally,
porosity changes in the reservoir rock are addressed for UHS. These changes in porosity are
calculated based on volume changes from the PHREEQC modeling results.
To calculate the loss of the stored gas is, from different points of view, the aim for all modeled
systems. In CCS, the potential loss of carbon dioxide is investigated by the analysis of gas–
water–rock interactions which could lead to higher porosity in the cap rock and favor CO2(sc/g)
pathways. In SCGS the loss of natural gas is analyzed by regarding the conversion of the
stored CH4(g) to H2S(g) by bacterial sulfate reduction. In UHS the loss of hydrogen is modeled
by aqueous diffusion, gas–water–rock interactions that bond the stored H2(g), and by bacterial
conversion.
All models are tested with sensitivity studies to identify the main influencing factors, including
pressure and temperature conditions (SCGS, CCS, UHS), kinetic rate constants (SCGS,
UHS), tortuosity (SCGS), gas compositions (SCGS, CCS, UHS), mineralogical compositions
(CCS), initial cap rock porosities (CCS), effective diffusion coefficients (CCS), transport
mechanisms (CCS), and storage time (CCS, UHS).
Differences in the three subtopics arise in the modeled storage time. While CCS is the final
disposal of a waste product, SCGS and UHS are used for seasonal storage. The modeling
16
Introduction
time considered is 1,000,000 years for CCS and 30 years for SCGS and UHS. The 1,000,000
years for CCS is chosen based on the discretization results of the model and follows the safety
regulations for final disposal of radioactive waste in Germany, where the evidence of long-term
safety for one million years has to be provided (BMU, 2010). The chosen time of 30 years for
SCGS results from the typical operating lifetime of a salt cavern and for UHS from the
equipment lifetime.
Table 1.2 Underground gas storage models and their parameters.
CCS SCGS UHS

Model type / Batch/PHREEQC Batch/PHREEQC 1D-
modeling tool 1D- 1D- transport/PHREEQC
transport/PHREEQC transport/PHREEQC
3D-transport/PHAST
Database Extended Extended Extended
phreeqc.dat phreeqc.dat phreeqc.dat
Reaction modeling Spontaneous Spontaneous Spontaneous
type equilibration equilibration and equilibration and
kinetics for BSR kinetics for BSR and
methanogenesis
Storage facility Depleted Salt cavern Depleted gas
hydrocarbon reservoir reservoir
Gases considered CO2(g/sc), CH4(g), H2S(g) CO2(g), CH4(g), H2S(g), CO2(g/sc), CH4(g),
N2(g) H2S(g), N2(g), H2(g)
Number of 11 14 12
aqueous species
Effective diffusion 3.33 × 10-10 5.0 × 10-9 Multicomponent
coefficient (m2 s-1) diffusion
Pressure (atm) 100 180 40
Temperature (°C) 40 50 40
Storage time (a) 1,000,000 30 30
Transport Diffusion Diffusion Diffusion
mechanisms
17
Carbon Capture and Storage (CCS)
2 Carbon Capture and Storage (CCS)
Change in Cap Rock Porosity Triggered by Pressure and Temperature

Dependent CO2–Water–Rock Interactions in CO2 Storage Systems
Publication I
Authors of the original paper: Christina Hemme, Wolfgang van Berk
Originally published in: Petroleum 3 (2017), 96–108
Summary added for the purpose of this thesis.
Author Contributions:
Conceptual model: Christina Hemme1

Modeling: Christina Hemme1 and Wolfgang van Berk2
Result evaluation: Christina Hemme1 and Wolfgang van Berk2

Analysis of results: Christina Hemme1 and Wolfgang van Berk2
Visualization of results: Christina Hemme1
Writing—original draft preparation: Christina Hemme1
Writing—review and editing: Christina Hemme1 and Wolfgang van Berk3
Supervision: Wolfgang van Berk1

1
substantial; 2 medium; 3 minor
18
Summary
In this part of the study, the gas–water–rock interactions triggered by carbon dioxide storage
are considered. The focus is on the diffusive mass transport of CO2 along a gradient of
changing pressure and temperature conditions through the cap rock brine. In addition, the
change in porosity due to mineral dissolution and precipitation is simulated to analyze the
sealing capacity of the cap rock.
In carbon capture and storage systems, supercritical CO2 is injected into a reservoir and
dissolves in the reservoir brine. Subsequently, aqueous CO2(aq) diffuses into the cap rock to
regions of lower total pressure and temperature and triggers CO2–water–rock interactions that
are coupled with mass transport and result in precipitation and/or dissolution of minerals along
the CO2 migration path. Such hydrogeochemical interactions cause porosity changes and are
responsible for the improvement, or the risk of deterioration, of the long-term integrity of the
system.
This part of the study presents a semi-generic hydrogeochemical model based on chemical
equilibrium thermodynamics, data from several CO2 storage systems, and plausible
assumptions regarding non-available data. One-dimensional reactive transport modeling is
performed using the PHREEQC code (phreeqc.dat database) to develop a general system and
process understanding of the hydrogeochemical effects of underground CO2 storage in
depleted hydrocarbon reservoirs on the aqueous solutions, mineral phases, and gas
compositions in the storage systems. Furthermore, the loss or gain of total porosity affected
by hydrogeochemical reactions driven by diffusive mass transport exposed to pressure and
temperature gradients is identified and quantified. A fine spatial and temporal discretization
and the use of non-reactive tracers enable calculation over the relevant timescale for
simulations of long-term storage of CO2.
Modeling results show that the relevant timescale for simulations of long-term storage of CO2
is on the order of 106 years, and that pressure/temperature conditions, heterogeneities (veins
and fractures), and the mineralogical composition of the cap rock have the strongest influence
on the increase in cap rock porosity (maximum increase from the initial 5% to 7.5%). Critical
parameter combinations could put long-term integrity at risk. Nevertheless, a wide range of
conditions and parameter combinations for safe CO2 storage are identified by other modeling
scenarios. This model approach can be used and adapted to various underground CO2 storage
conditions to answer and solve practically oriented questions and problems (Hemme and van
Berk, 2015).
19
Abstract
Carbon capture and storage in deep geological formations is a method to reduce greenhouse
gas emissions. Supercritical CO2 is injected into a reservoir and dissolves in the brine. Under
the impact of pressure and temperature (P–T) the aqueous species of the CO2-acidified brine
diffuse through the cap rock where they trigger CO2–water–rock interactions. These
geochemical reactions result in mineral dissolution and precipitation along the CO2 migration
path and are responsible for a change in porosity and therefore for the sealing capacity of the
cap rock. This study focuses on the diffusive mass transport of CO 2 along a gradient of
decreasing P–T conditions. The process is retraced with a one-dimensional hydrogeochemical
reactive mass transport model. The semi-generic hydrogeochemical model is based on
chemical equilibrium thermodynamics. Based on a broad variety of scenarios, including
different initial mineralogical, chemical and physical parameters, the hydrogeochemical
parameters that are most sensitive for safe long-term CO2 storage are identified. The results
demonstrate that P–T conditions have the strongest effect on the change in porosity and the
effect of both is stronger at high P–T conditions because the solubility of the mineral phases
involved depends on P–T conditions. Furthermore, modeling results indicate that the change
in porosity depends strongly on the initial mineralogical composition of the reservoir and cap
rock as well as on the brine compositions. Nevertheless, a wide range of conditions for safe
CO2 storage is identified.
2.1 Introduction and Aim
A solution to reduce CO2 emissions into the atmosphere is the storage of CO2 in the reservoir
rocks of depleted hydrocarbon reservoirs, coal seams and deep saline aquifers (Holloway,
1997). For the storage of CO2 four different trapping mechanisms are known: (i) structural
trapping, (ii) residual trapping, (iii) dissolution trapping, and (iv) mineral trapping (Gaus et al.,
2008; IPCC, 2005). Black et al. (2015) stated that the effects of mineral dissolution and
precipitation are the key factors for safe CO2 storage. The injection of CO2 leads to CO2–water–
rock interactions that result in the precipitation and dissolution of minerals which in turn result
in a change in porosity and permeability of the cap rock. This change leads to the improvement
or deterioration of the sealing capacity of the cap rock. Mineral dissolution results in an
increase in porosity and enables the creation of pathways for CO2 migration. Mineral
precipitation leads to a decrease in porosity and increases the sealing effect (Gaus et al.,
2005). In the discussion about carbon capture and storage (CCS) many studies focus on CO2–
water–rock interactions in reservoir rocks but much less attention has been paid on these
20
interactions in the cap rock. Consequently, forecasting and quantifying such CO 2–induced
processes in cap rocks rank high on the agenda for successful carbon capture and storage in
deep geological formations. Thus, various modeling and experimental studies aim to retrace
and reproduce such hydrogeochemical reactions in cap rocks (Alemu et al., 2011; Bildstein et
al., 2010; Black et al., 2015; Bolourinejad and Herber, 2015; Credoz et al., 2009; Fleury et al.,
2011; Gaus et al., 2005; Gherardi et al., 2007; Szabó et al., 2016; Tian et al., 2014; Tian et al.,
2015; Wollenweber et al., 2010).
Knowledge of cap rock properties is important for these studies. Cap rocks are characterized
by specific mineralogical and physical properties like low porosity and permeability. The often
clay-bearing cap rocks have to be saturated with brine and must have a capillary entry pressure
that prevents the cap rock from supercritical CO2 entering (Mohd Amin et al., 2014) and
prevents CO2 from escaping towards overlying aquifers needed for water supply (Liu et al.,
2015).
Our study is based on equilibrium thermodynamics of chemical CO 2–water–rock interactions

to fundamentally reveal where a carbon capture and storage system inevitably develops in
terms of porosity alteration resulting from mineral dissolution and precipitation. It is not our aim
to reproduce a real complex geological system. Correspondingly, our modeling approach is of
semi-generic nature. We aim to identify long-term inorganic hydrogeochemical processes
including CO2–water–rock interactions and to quantify the change in porosity due to the effects
of geochemical reactions in dependence on pressure and temperature. This process is
retraced with a one-dimensional hydrogeochemical reactive mass transport model based on a
hypothetical concept. Authors like Bildstein et al. (2010) have modeled porosity creation in the
cap rock at constant temperature and pressure conditions. Generally, temperature and
pressure conditions change along the path of diffusive transport of CO2 (van Berk et al., 2015)
and could strongly affect the mass-action law constants for the equilibrium reactions between
mineral phases and CO2. Therefore, it is important to consider not only different temperature
and pressure conditions at the starting point of diffusive CO2 migration, like those modeled by
Van Pham et al. (2012), but also the change of temperature and pressure along the diffusive
pathway of CO2 through the cap rock. Therefore, this study will focus on the diffusive mass
transport of CO2 along a gradient of decreasing pressure and temperature conditions. A broad
variety of modeling scenarios, including different initial mineralogical and physical parameters,
is calculated to identify the hydrogeochemical parameters most sensitive for safe CO2 storage.
In addition, the numerical accuracy of our model is tested by a detailed grid convergence study.
21
2.2 Kinetic or Thermodynamic Equilibrium Modeling for Long-Term CO2

Storage?
Various modeling studies related to cap rocks in CCS systems apply batch and/or reactive
transport modeling approaches that include reaction kinetics (Gaus et al., 2005; Mohd Amin et
al., 2014; Tian et al., 2015). These studies focus on the time-dependent progress of mineral
dissolution and precipitation reactions which are characterized by associated reaction rate
constants. An overview over the rate parameters of water-mineral interaction kinetics are given
in a report by the US Geological Survey (Palandri and Kharaka, 2004). Kinetic simulations still
show uncertainties regarding the poor knowledge of the kinetic rate constants at elevated
levels of temperature, total and CO2 partial-pressure. However, such knowledge is essential
for kinetically based modeling of CCS systems, especially in this study which focuses on the
diffusive mass transport of CO2 along a gradient of decreasing P–T conditions. In addition, the
rates obtained in laboratory strongly depend on the experimental devices and the chosen
boundary or environmental conditions, and therefore they are difficult to compare with another
(Marini, 2007). Ignoring the facts that i) hydrogeochemical conditions (e.g., pH, pε, ionic
strength) evolve and change in time and space and ii) kinetic rate constants must change
accordingly, makes the modeling results meaningless–especially when thermodynamically
impossible reactions are modeled.
Approaches based on kinetics start from the assumption that kinetically characterized
reactions actually proceed–whether any driving force in terms of system thermodynamics
allows these reactions to proceed or not. In contrast, thermodynamically based modeling
approaches only calculate the reactions, which are thermodynamically possible, until
spontaneous equilibrium conditions are established. The term spontaneous implies that the
equilibrium is established for the residence time of CO2 in a spatially discrete compartment.
Stumm and Morgan (1981) argued that applying thermodynamically based approaches allow
to “identify reactions that are possible”. In consequence, the authors developed “first
equilibrium and then kinetic frameworks for natural water systems”. Moreover, Marini (2007)
stated that equilibrium thermodynamics seem to be a good modeling tool for the analysis of
chemical reactions in CO2 storage systems. This appraisal is based on the studies performed
by Helgeson (Barnes, 1979; Helgeson, 1967, 1968; Helgeson et al., 1969; Helgeson et al.,
1970; Helgeson, 1979) which show that a separation of the overall reaction path into a series
of partial equilibrium states enables successful modeling based on chemical equilibrium
thermodynamics.
One of the main differences between thermodynamically based modeling studies and
approaches including reaction kinetics (Gaus et al., 2005; Mohd Amin et al., 2014; Tian et al.,
22
2015) is the total time considered by the modeling of specific reactive transport processes. In
their kinetically based modeling study, Mohd Amin et al. (2014) considered a total time of
10,000 years referring to the fact that Bowden and Rigg (2005) and Credoz et al. (2009)
asserted that such a timescale is relevant for the long-term safety assessment of CO2 storage.
However, this determination of Bowden and Rigg (2005) and Credoz et al. (2009) is presented
without any retraceable justification. Tian et al. (2015) performed all their simulations
considering a total time of 1000 years while Gaus et al. (2005) considered 15,000 years. Gaus
et al. (2005) concluded that when using kinetically based calculations the system is still far
from equilibrium after 15,000 years. However and in contrast, to justify a relevant timescale
our study calculates the diffusive transport of a non-reactive tracer through the model system
for various total times and time step lengths–prior to any reactive transport modeling (for
details, see Section 2.4.1). From the time-dependent spatial spreading of this tracer we
conclude that a total time of 1,000,000 years is the timescale relevant for the long-term safety
assessment of CO2 storage systems similar to our semi-generic modeling systems.
Considering these facts, we decide to perform modeling based on thermodynamics, although

we are aware that the progress of various heterogeneous reactions is actually controlled by
kinetics. It is not our aim to connect thermodynamically driven hydrogeochemical processes
with temporal aspects. Our aim is to identify and quantify the long-term hydrogeochemical
processes. Provided that thermodynamic equilibrium is established, the thermodynamically
based modeling approach enables us to test where the CCS system is inevitably going in terms
of porosity alteration.
2.3 Conceptual Numerical Model and Modeling Parameters
2.3.1 Model Setup
The hydrogeochemical, one-dimensional reactive mass transport (1DRMT) modeling

approach presented in this study is based on a semi-generic concept. It includes available data
from several real CO2 storage systems and is based on assumptions regarding non-available
data for the sake of simplicity. Alternative scenarios for selected parameters are performed to
reveal how strongly the parameters affect the CO2–water–rock interactions (for details, see
Section 2.5.2).
The geological structure in this model is a depleted hydrocarbon reservoir in an anticlinal fold
(Figure 2.1). Carbon dioxide is injected into the reservoir rock at a depth of about 800 m. At
this depth CO2 is supercritical (CO2(sc)) and dissolves until the reservoir brine is saturated with
23
respect to CO2(aq). The critical point for CO2 lies at a temperature of 304.21 K (31.06 °C) and
at a pressure of 73.825 bars (72.859 atm) according to Angus et al. (1976) and Salimi et al.
(2012). At temperature and pressure conditions above this critical point, the CO 2(sc) is present
in a dense phase which is lighter than the brine in the reservoir rock and migrates from the
point of injection to the reservoir/cap rock contact where a CO2(sc) plume builds up (Figure 2.1)
(Lindenberg and Wessel-Berg, 1997). During this process the reservoir brine is displaced by
the supercritical CO2(sc). At the contact with the cap rock CO2 is available as CO2(sc) in the plume
and as dissolved CO2 (CO2(aq) and corresponding aqueous species e.g. HCO3-, CO32-,
CaHCO3+, H+) in the reservoir brine beneath the plume. The CO2(sc) dissolves in the cap rock
brine and an acidic solution is created. Molecular diffusion of aqueous species through the cap
rock brine leads to CO2(aq)–water–rock interactions in the cap rock that cause mineral
dissolution and precipitation and therefore a change in the porosity and permeability of the cap
rock. The mass-action law constants for the equilibrium reactions between mineral phases and
CO2(aq) depend on temperature and pressure, and therefore they change along the CO 2(aq)
migration path (van Berk et al., 2015).
Figure 2.1 System sketch of the model. CR = cap rock, RR = reservoir rock, CO 2(sc) = supercritical
CO2 and CO2(aq) = aqueous CO2.
Our hydrogeochemical model is based on thermodynamic equilibrium reactions and the local
equilibrium assumption. Kinetic aspects of mineral dissolution/precipitation are not considered.
Even if CO2 is supercritical under the assumed reservoir conditions, this model uses gaseous
CO2 as an analog. This simplifying assumption is based on the study of Rochelle and Moore
24
(2002) that shows results of supercritical CO2 solubility experiments at different pressure and
temperature conditions. We retraced these experiments using PHREEQC Interactive 3.1.4-
8929 and its database phreeqc.dat, with the difference that we used gaseous CO 2 instead of
supercritical CO2. The solubility constants for CO2(g) used in PHREEQC and CO2(sc) measured
by Rochelle and Moore (2002) are similar and consequently the following reactions apply
(Equations (2.1) and (2.2)):
CO2(g) = CO2(aq) log K = -1.468; at 25 °C; 1 bar

(2.1)
CO2(sc) = CO2(aq) log K = -1.468; at 25 °C; 1 bar

(2.2)
Furthermore, to prove and quantify this assumption a separate sensitivity analysis is presented
in Section 2.5.3.
In the model the migration path is demonstrated by a column of 101 cells, with a cell length of
1 m. One cell is located in the reservoir rock and one hundred cells are located in the cap rock
(Figure 2.1). For both ends of the column the modeling boundary conditions are defined as
flux. Due to a constant 𝑝CO2 in the plume, the lower boundary can be interpreted as a constant
concentration boundary. By adding five extra cells at the upper end of the column, the influence
of the upper boundary condition can be minimized, considering that only the 101 cells are
evaluated. The only mass transport mechanism in this model is molecular diffusion of all
aqueous species through the cap rock brine. The process of diffusion of aqueous species is
slow (Gaus et al., 2005) but allows the mass transport of dissolved species between all cells
(Arning et al., 2011), depending on time and diffusivity. One diffusion coefficient for all aqueous
species of 3.33 × 10-10 m2 s-1 is used. This value is based on the assumption made by Mohd
Amin et al. (2014) that the diffusion coefficient is made up of the diffusion coefficient in pure
water of 1 × 10-9 m2 s-1 divided by an averaging tortuosity of 3 (Cussler, 2009).
2.3.2 Initial Pressure and Temperature Conditions
The pressure and temperature conditions (P–T conditions) are similar to the conditions
predominating at the CCS storage system at Sleipner in Norway, starting with 40.0 °C (37.0 °C
at Sleipner) and 100.0 atm at reservoir conditions (100.0 atm at Sleipner). The temperature
decreases from the reservoir depth up along the geothermal gradient, at 33.3 °C per kilometer
of depth, and the pressure decreases accordingly with 100.0 atm per kilometer under
hydrostatic conditions. Consequently, each of the 101 cells is associated with a specific
pressure/temperature condition (for example, cell 1: 40.0 °C and 100.0 atm, cell 2: 39.967 °C
25
and 99.9 atm). We assume a constant partial pressure of the CO2(sc) plume over time because
CO2(sc) is delivered continuously by the plume for a long time and, for the sake of simplicity, we
equate the partial pressure with the hydrostatic pressure.
2.3.3 Initial Mineralogical Composition of the Reservoir and Cap Rock and Initial
Brine Compositions
Each cell is defined by specific mineralogical, hydrochemical and physical properties. A

homogeneous distribution of relevant parameters in the cap rock (cells 2–101) such as
mineralogical composition is assumed. Table 2.1 summarizes the selected initial mineralogical
composition in all cells. Assuming an initial total porosity of 25.0%, cell 1, representing the
reservoir rock, is filled with 1.0 L of irreducible water, whereas 4.0 L of pore space are occupied
by injected CO2 (CO2(sc) plume).
Cells 2–101 are located in the cap rock and are characterized by a representative volume of
20.0 L with an initial porosity of 5.0% (1.0 L irreducible water and 19.0 L solid phases), where
the type of porosity is neglected. The assumption of 5% initial porosity in the cap rock is based
on observations from the Nordland Shale in UK Quadrant 16, northern North Sea (Gaus et al.,
2005; Kemp et al., 2001) as an analog for the cap rock porosity in the CCS system Sleipner.
Taking into account the specific density of each mineral phase (in g cm-3), the amount of single
mineral phases is calculated in mole per kg of pore water for different reservoir parts (Table
2.1). It is assumed that a small amount of CO2 prevails in the cap rock and the reservoir rock
during burial prior to CO2(sc) injection (𝑝CO2 = 1.0 atm). Before CO2 injection and the triggered
diffusion, a separate batch modeling calculates the initial composition of brines in both
reservoir parts (the detailed input file for the separate batch model to calculate the cap rock
brine is given in the supplementary data Input file S1). The brine composition of cell 1 (reservoir
rock) is calculated by equilibrating 1 L of 1.5 M Na+/Cl--dominated solution with the defined
primary minerals and gas phase under 40.0 °C and 100.0 atm (Table 2.2). For calculating the
initial brine composition in cells 2–101 (cap rock), a 1.5 M Na+/Cl--dominated solution including
0.15 mol kgw−1 K+ is equilibrated with the mineral phases of the cap rock and CO2 under the
pre-assigned reservoir P−T gradient (cell 2: 39.967 °C and 99.9 atm; cell 3: 39.934 °C and
99.8 atm; Table 2.2).
26
Table 2.1 Mineralogical composition of the reservoir rock and the cap rock.
Primary minerals Weight percent (wt. %) Amount mol kgw−1
Reservoir rock
K-feldspar 14.0 19.952
Albite 5.0 7.563
Quartz 75.0 495.130
Calcite 6.0 23.778
Cap rock
Albite 3.0 5.833
Quartz 55.0 466.745
Calcite 5.0 25.471
Kaolinite 25.0 49.377
Pyrite 2.0 8.499
Chlorite 2.0 1.835
Barite 2.0 4.369
Table 2.2 Initial irreducible water composition of the reservoir rock and cap rock.
Parameter Reservoir rock Cap rock
pH 6.064 7.912
Temperature (°C) 40.0 39.967
Concentration Concentration
Elements
(mol kgw−1) (mol kgw−1)
Al 3.544 × 10-06 5.13 × 10-07
Ba -d 3.46 × 10-04
Ctota 4.254 × 10-02 2.52
Ca 1.365 × 10-02 1.96 × 10-05
Clb 1.500 1.61
Fe -d 2.91 × 10-07
K 7.649 × 10-03 2.36 × 10-02
Mg -d 5.75 × 10-03
Nab 1.492 4.47
Stotc -d 3.46 × 10-04
Si 1.105 × 10-04 6.56 × 10-05
a C: summed concentration of aqueous CH4 and C(+IV) species
b: Due to high Na+ and Cl- concentrations and the resulting high ionic strength the Pitzer database and
Pitzer equations should be used but the Pitzer database does not include Si- und Al3+-containing
aqueous species and silicate minerals, which are important components of the cap rocks of CO2
storage systems. Therefore, the phreeqc.dat is used, which includes such species and silicate
minerals and uses the Debye–Hückel equation, even if the activity coefficients for all aqueous species
are overestimated (Mohd Amin et al., 2014). Parkhurst and Appelo (1999) assert that “in sodium
chloride dominated systems, the model may be reliable at higher ionic strengths”. Furthermore, the
results of the different sensitivity scenarios are compared in relation to one another and not in an
absolute context
c S: summed concentration of aqueous S(+VI) and S(-II) species
d: not present
27
2.4 Methodology
The software PHREEQC Interactive 3.1.4-8929 (Parkhurst and Appelo, 2013) and its database
phreeqc.dat are the modeling tools for the one-dimensional reactive transport modeling. This
software is provided by the Geological Survey of the United States. PHREEQC is based on an
ion-association aqueous model and has the capability to simulate speciation, batch-reaction,
one-dimensional transport and inverse geochemical calculations. The calculations are based
on mass action laws that include all species used in this study (Al, Ba, C, Ca, Cl, Fe, K, Mg,
Na, S, Si) and their corresponding equilibrium constants. A basic of the modeling is the
thermodynamic database which includes the elements used in the model with their species
(aq, s, g), mass-action equations, and equilibrium constants. The activity coefficients of
species are calculated by using the Debye-Hückel equation. The one-dimensional reactive
transport model calculation in PHREEQC is based on the Advection-Reaction-Dispersion
Equation (2.3) (ARD):
𝜕𝐶 𝜕𝐶 𝜕 2 𝐶 𝜕𝑞
= −𝑣 + 𝐷𝐿 2 − (2.3)
𝜕𝑡 𝜕𝑡 𝜕𝑥 𝜕𝑡
where 𝐶 is the concentration in water (mol kgw−1), t is the time (s), 𝑣 is the pore water flow
velocity (m s-1), 𝑥 is the distance (m), 𝐷𝐿 is the hydrodynamic dispersion coefficient [m2 s-1, 𝐷𝐿
= 𝐷𝑒 + 𝛼𝐿 𝑣, with 𝐷𝑒 the effective diffusion coefficient, 𝛼𝐿 the dispersivity (m)], and 𝑞 is the
concentration in the solid phase (expressed as mol kgw−1 in the pores).
Our model considers a multiple web of hydrogeochemical reactions including 119 aqueous
species, 10 mineral phases and their interplay. We use PHREEQC Version 3 which considers
the pressure and the temperature dependence of the mass-action law constants for the
equilibrium reactions of the aqueous, gaseous and solid species involved. Table 2.3 shows the
equilibrium phases, mass-action equations and equilibrium constants used in the model.
Phases like dawsonite and nahcolite that are not present in the phreeqc.dat database are
taken from the llnl.dat database and added to the input files. However, the pressure
dependence of both minerals is not considered.
Numerous publications confirm that PHREEQC produces correct results by comparing the
results with other codes (De Windt et al., 2003; Gundogan et al., 2011; Nowack et al., 2006).
A detailed comparison of modeling results of different software packages like
The Geochemist´s Workbench, EQ3/6, FactStage/ChemApp and PHREEQC is given by
Haase et al. (2013) with the focus on geochemical modeling of CO2 and calcite dissolution in
NaCl solutions. Furthermore, the user´s manual of PHREEQC provides calculations to verify
their results (Parkhurst and Appelo, 2013).
28
Table 2.3 Equilibrium phases, mass-action equations and equilibrium constants used in the model.
Data from phreeqc.dat, but dawsonite and nahcolite are from llnl.dat (Parkhurst and Appelo, 2013).
Equilibrium Equilibrium reaction log K
phase at 25 °C, 1 bar
K-feldspar KAlSi3O8 + 8H2O = K+ + Al(OH)4- + 3H4SiO4 -20.573
Albite NaAlSi3O8 + 8H2O = Na+ + Al(OH)4- + 3H4SiO4 -18.002
Quartz SiO2 + 2H2O = H4SiO4 3.98
Calcite CaCO3 = CO32- + Ca2+ 8.48
Kaolinite Al2Si2O5(OH)4 + 6H+ = H2O + 2H4SiO4 + 2Al3+ 7.435
Pyrite FeS2 + 2H+ + 2e- = Fe2+ + 2HS- -18.479
Mg5Al2Si3O10(OH)8 + 16H+ = 5Mg2+ + 2Al3+ + 3H4SiO4 + 68.38
Chlorite(14A)
6H2O
Barite BaSO4 = Ba2+ + SO42- -9.97
Dawsonite NaAlCO3(OH)2 + 3H+ = Al3+ + HCO3- + Na+ + 2H2O 4.35
Nahcolite NaHCO3 = HCO3- + Na+ -0.11
CO2(g) CO2 = CO2 -1.468
2.4.1 Discretization
To check the numerical accuracy of the results, the total time, number of shifts and number of
cells are refined, and the model calculation for one (the same) scenario is rerun and the results
are compared (Parkhurst and Appelo, 1999, 2013). Using a non-reactive tracer, induced as a
peak injection into the first cell of the model, the progress of diffusion can be demonstrated. A
concentration gradient for the non-reactive tracer is chosen based on CO2 molality at the
temperature and pressure conditions in the reservoir rock and in the cap rock. Thereby the
diffusive mass transport progress of CO2 can be tested, if the CO2 would not react with the
brine and the mineral phases of the cap rock. The total time, number of shifts and number of
cells are refined until ions moved from regions of high concentration (cell 1) to regions of lower
concentration (cells 2–101), resulting in an equal concentration in all cells (1–101).
To define the total time a non-reactive tracer is induced in the model to calculate the time at
which the tracer concentration is equal in all cells. The model calculation is repeated for a total
time of 104, 105, 106 and 107 years with a constant number of shifts and cells. The results show
that the tracer concentration is equal in all the cells in the models at times of one million and
ten million years (Table 2.4). In the following models the time scale of one million years is
chosen.
In the second step of refinement the number of shifts is varied. The time is kept constant to
one million years but the number of shifts (100, 1000, 10,000) and therefore the length of the
time steps is varied, although PHREEQC may subdivide the “time step into smaller dispersion
time steps if necessary to calculate dispersion accurately” (Parkhurst and Appelo, 2013). Table
2.4 shows identical results for all the modeled numbers of shifts in the last shift. When
29
comparing the tracer concentration at a fixed time for all models, differences in the progress
of the diffusion are identifiable. The tracer concentration 1000 years later is equal in all cells in
the 1000-shifts model to that in the model with 100 shifts. With 10,000 shifts the tracer
concentration 400 years later is equal in all cells to that with 1000 shifts. The stopping criterion
is set to 1 per mil, so 400 years fall below this cut-off. Therefore, the relevant number of shifts
is set to 1000.
In the last step of refinement, the number of cells is varied (11, 21, 51, 101). The total time is
kept constant to one million years and the number of shifts are kept constant to 1000. Table
2.4 depicts identical results for all the modeled numbers of cells in the last shift. The same
procedure as for identification of the number of shifts applies here. When comparing the tracer
concentration at a fixed time for all models, differences in the progress of the diffusion are
identifiable. Considering the stopping criterion of 1 per mil, the results show that the relevant
number of cells is 101 (Table 2.4).
These results are valid for the model of the reference scenario only and discretization studies
should be done for each 1DRMT model. Numerical stability/instability is time and space
dependent. To achieve numerical stability, the grid and time steps have to be refined, but due
to the smaller time steps the number of solution calculations increases (Parkhurst and Appelo,
2013). Moreover, the change in concentration, which results from reactions, is small from one
time step to the next (Schäfer, 2004). A disadvantage of this method is that it is not clear how
many cells and which length of time step are necessary to achieve numerical stability (Schäfer,
2004).
Similar to our study, other studies like that of Mohd Amin et al. (2014) consider numerical
dispersion as well, but they assert that their “results are not sensitive to numerical dispersion
and other errors in the discretization of the transport and reaction equations” (Mohd Amin et
al., 2014). They substantiate their assertion with a half of the grid spacing and the outcome of
identical results. The results presented here show the opposite effect with a strong sensitivity
to numerical dispersion in the sense of Mohd Amin et al. (2014). The reason for that difference
is the assumption of a non-reactive tracer. CO2 (used by Mohd Amin et al. (2014)) reacts with
the reservoir and cap rocks and the respective formation waters, so an equal tracer
concentration in all cells is reached earlier than without a non-reactive tracer. The advantage
of a non-reactive tracer is that the progress of diffusion over the total time and the total distance
(= number of cells) can be demonstrated independently from the chemical reactions that may
occur. Furthermore, Credoz et al. (2009) assume a total time for the long-term safety
assessment of CCS of 10,000 years. Based on the results of the discretization for this model,
a total time of 1,000,000 years is defined. However, we are aware that the assumed physical
30
and chemical conditions, as well as the use of a non-reactive tracer, do not comply with a real
system and cannot represent the complexity of the system. Nevertheless, checking the
numerical accuracy of the modeling results is essential for all modeling studies.
Table 2.4 Results of temporal-spatial refinement rounded to two decimal places. The parameters used
for the reference scenario are shown in bold.
Tracer concentration C(tracer) (mol kgw−1)
Parameters Cell 1 Last cell ∆Ctracer
Time total (years)

0.00 0.85 0.01a 0.84
10,000 0.06 0.01a 0.05
100,000 0.03 0.01a 0.02
1,000,000 0.02 0.02a 10-6
10,000,000 0.02 0.02a 10-6
Shifts
100 0.02 0.02a 10-6
1000 0.02 0.02a 10-6
10,000 0.02 0.02a 10-6
Cells
11 0.10 0.09b 0.01
21 0.06 0.04c 0.02
51 0.06 0.01d 0.05
101 1.04 0.99a 0.05
a: last cell: 101
b: last cell: 11
c: last cell: 21
d: last cell: 51
2.5 Results and Discussion
2.5.1 Reference Scenario
The equilibrium species distribution and the mass conversion are calculated based on the
chemical thermodynamics of aqueous equilibrium reactions. The results of the 1DRMT model
after 106 years of diffusive mass transport are shown in Figure 2.2 where the column of 101
cells and the change of porosity for the reference scenario are depicted. Cell 1 is neglected
because it is positioned in the reservoir rock and has a different initial mineralogical
composition, whereas all the other cells are located in the cap rock. The modeling results show
that the mineral dissolution and precipitation differ from cell to cell, whereat mass conversion
31
occurs predominantly in cells 2–9. Therefore, the calculated results within the first ten cells are
shown enlarged and are discussed in detail (Figure 2.2).
The cap rock brine dissolves the CO2(sc) and free H+ are released because of carbonic acid
dissociation. Due to acid attack, albite, K-feldspar and calcite dissolve and the corresponding
pH increases to 8.4. Cell 2 is completely cemented after 106 years. The porosity is reduced
from the initial 5.0% to -0.5% (∆ -5.5%). The negative value (-0.5%) implies that the amount of
precipitated phases is higher than the available pore space. Consequently, the phases cement
also the available pore space in the next cell. Mineral dissolution and precipitation reactions
are expressed by the Saturation Index (SI) = log (IAP/K) (IAP: ion activity product; K: solubility
product). When CO2(aq) migrates into cell 2 the cap rock brine is temporarily oversaturated with
respect to quartz and dawsonite (SI > 0) and undersaturated regarding K-feldspar, albite,
kaolinite, calcite, chlorite and barite (SI < 0). Thereupon, quartz and dawsonite precipitate in
high amounts (+124.3 mol in cell 2) and K-feldspar, albite, kaolinite, calcite, chlorite and barite
dissolve in small amounts (-32.6 mol in cell 2). The dissolution of albite and K-feldspar
contributes Na+, K+, Al(OH)4- and H4SiO4 to the cap rock brine. To form dawsonite, Al3+ and
Na+ have to react with HCO3- and 2H2O and reach saturation with SIdawsonite > 0. The respective
equilibrium reactions are given in Table 2.3. Figure 2.2 shows that with increasing dissolution
of albite and K-feldspar the precipitation of dawsonite gets stronger. Calcite dissolves in cell 2
in small amounts only. Calcite dissolves slightly more strongly when albite and K-feldspar are
completely dissolved, but the amount of both in this model is high enough to buffer the pH and
protect calcite against stronger dissolution.
In cell 3 the porosity decreases from the initial 5.0% to 2.7% (∆ -2.3%). The precipitation of
quartz, dawsonite, calcite and kaolinite is responsible for this decrease. The dissolution of K-
feldspar, albite and chlorite in comparison with the precipitation of quartz, dawsonite, calcite
and kaolinite is too weak to create an increase in porosity. In cell 4 a weaker decrease in
porosity from the initial 5.0% to 4.44% is observable. Quartz, dawsonite, K-feldspar, chlorite
all precipitate and albite, kaolinite and calcite dissolve.
From cell 5 to cell 9 the porosity decreases from the initial 5.0% to ~4.5%. Quartz, dawsonite,
K-feldspar and chlorite precipitate in these cells, whereas albite dissolves. Calcite dissolves in
cells 5 and 6 but precipitates in small amounts in cells 7–9. Kaolinite precipitates in cells 5–7
and dissolves in cells 8 and 9. Cells 10 to 101 show similar results. Dawsonite precipitates as
well as quartz, albite, K-feldspar, chlorite and barite. Kaolinite and calcite dissolve. Albite is
completely dissolved from cells 2–8 but precipitates in small amounts from cell 9 onwards.
Simultaneously, dawsonite precipitation decreases and kaolinite dissolution starts.
32
Precipitation and dissolution of pyrite, barite and chlorite has a low intensity and can be
neglected.
Hellevang et al. (2005) discussed the potential of dawsonite to permanently trap CO2 and
assumed that “if dawsonite does precipitate at nonalkaline conditions it is only an ephemeral
phase which decomposes when CO2 pressure drops”. Using nahcolite (NaHCO3) as potential
secondary phase instead of dawsonite (NaAlCO3(OH)2), the decrease in porosity is with a
mean of -1.0% in cells 2–10 and of -0.3% in cells 11–101 smaller than using dawsonite (-1.2%
in cells 2–10 and -0.4% in cells 11–101). The difference in the results can be explained by a
smaller amount of precipitated nahcolite in comparison with dawsonite. By dissolution of albite
and K-feldspar, Na+ and Al3+ ions are available. Nahcolite can use Na+ ions only, whereas
dawsonite uses Na+ and Al3+ ions. It is striking that calcite and kaolinite precipitate when
previously formed secondary nahcolite is available. The pH is 8.1 when using nahcolite as the
potential secondary phase, whereas dawsonite precipitation leads to a pH of ca. 8.4.
33
34
Figure 2.2 Initial conditions in the column of 101 cells and calculated distribution of changes in porosity in %, mineral amounts as well as pH for the reference
scenario after 106 years. RR = reservoir rock, CR = cap rock. “Initial” shows the initial mineralogical composition of the cap rock and the initial pH of th e cap
rock brine. Cell 2 is completely cemented. The porosity is reduced from the initial 5.0% to -0.5% after 106 years (= decrease in porosity of --5.5%). The amount
of precipitated phases is higher than the available pore space. Consequently, the phases will use the available pore space in the next cell.
2.5.2 Alternative Scenarios
Table 2.5 shows the input parameters used in the alternative scenarios. All the alternative
scenarios are based on the reference scenario, and single parameters are changed to show
their effects on the modeling results of each parameter. The modeling results of the alternative
scenarios are presented in Figure 2.3.
Table 2.5 Parameter sensitivity analysis in alternative scenarios.

Scenario Parameter
reference 40.0–36.7 °C / 100.0–90.0 atm
1a 150.0–146.7 °C / 470.0–460.0 atm
Temperature/pressure
1b 60.0–56.7 °C / 150.0–140.0 atm
1c 34.0–30.7 °C / 60.0–50.0 atm
reference No veins
2a Quartz vein (100% quartz, 5.0% porosity)
2b Veins Calcite vein (100% calcite, 5.0% porosity)
2c Calcite vein (100% calcite, 5.5% porosity)
2d Calcite vein (100% calcite, 40.0% porosity)
reference 5.0%
3a 40.0%
Initial cap rock
3b 10.0%
porosity
3c 6.0%
3d 2.0%
reference 3.33 × 10-10 m2 s-1
4a Effective diffusion 1.00 × 10-10 m2 s-1
4b coefficient 4.50 × 10-11 m2 s-1
4c 1.00 × 10-11 m2 s-1
reference 100% CO2
5a Injected fluid 99.0% CO2 + 1.0% CH4
5b composition 50.0% CO2 + 50.0% CH4
5c 99.0% CO2 + 1.0% H2S
reference –
Initial conditions Mineralogical data from Bildstein et al. (2010) (see
6a according to Appendix Table A 2.1, Table A 2.2)
published case study Mineralogical data + P–T conditions from Bildstein et al.
6b
(2010) (see Appendix Table A 2.1, Table A 2.2)
reference Diffusion
Transport mechanism
7 Advection
reference 40.0–36.7 °C / 100.0–90.0 atm
8a Temperature 150.0–146.7 °C + constant pressure (469.9 atm)
8b 34.0–30.7 °C + constant pressure (59.9 atm)
reference 40.0–36.7 °C / 100.0–90.0 atm
9a Pressure 470.0–460.0 atm + constant temperature (149.967 °C)
9b 60.0–50.0 atm + constant temperature (33.967 °C)
35
2.5.2.1 Temperature / Pressure
Scenarios 1a–c (Figure 2.3a) cover a spectrum of temperature and pressure conditions,
comparable to CO2 storage systems like those at Ketzin (Germany), Frio (USA) and Lacq-
Rousse (France). Changing temperature and pressure will affect the equilibrium constants, K,
of the mass-action laws of minerals, aqueous species and gases in such CO2 dominated
systems (van Berk et al., 2015). Thus, scenarios 1a–c show the effect of temperature and
pressure on the change in porosity triggered by mineral dissolution and precipitation. The pH
varies for the different scenarios. It decreases in scenario 1a to 6.7 but increases in scenario
1b to 8.0 and in scenario 1c to 8.5–8.6. At a temperature of 150.0 °C and a pressure of
470.0 atm the porosity of the cap rock increases from the initial 5.0% to up to 6.7% in cells 12–
101. At a temperature of 60.0 °C and a pressure of 150.0 atm the porosity of the cap rock
decreases from the initial 5.0% to up to 4.8% in cells 12–101. At P–T conditions of 60.0 atm
and 34.0 °C the porosity decreases from the initial 5.0% to 4.5% in cells 12–101. At
temperatures between 34.0 °C and 135.0 °C and pressures between 60.0 atm and 440.0 atm
the porosity decreases from the initial 5.0% to an average of 4.4%. With the help of various
further temperature/pressure scenarios, a critical point is revealed where the porosity
development changes from a decrease to an increase. This critical point lies between the
temperatures of 135.0 °C and 150.0 °C and corresponding pressures of 440.0 atm and
470.0 atm.
Changes in temperature and pressure affect the mineral behavior. When comparing scenario
1a (150.0 °C / 470.0 atm; cell 1) with the reference scenario (40.0 °C / 100.0 atm; cell 1),
differences in the porosity change are identifiable. At higher temperature and pressure
conditions (scenario 1a) the porosity increases. Unlike the reference scenario where the pH
increases from the initial 7.9 to 8.4, the pH decreases to 6.7 in scenario 1a. In contrast to the
reference scenario, in scenario 1a chlorite is completely dissolved in cells 2−101. K-feldspar
is completely consumed in the first cells but is available from cell 5, one cell earlier than in the
reference scenario. Calcite, quartz and kaolinite dissolution and precipitation are less strong
than in the reference scenario. The same applies for dawsonite, which precipitates with
44.5 mol in cell 2 in the reference scenario and with 23.3 mol in cell 2 of scenario 1a.
36
Figure 2.3 Gain or loss of total porosity for scenarios 1 to 9 (a–i) in a column of 101 cells, with a cell
length of 1 m. 100 m are located in the cap rock and 1 m in the reservoir rock. The values are given as
mean values for every 10 meters.
37
2.5.2.2 Veins
Cap rock properties like mineralogical compositions rarely show a homogeneous distribution.
Quartz veins and calcite veins occur frequently. For this, scenarios 2a−d with mono-
mineralogical compositions of quartz and calcite are calculated (Figure 2.3b, Table 2.5). In a
quartz vein, a small decrease in porosity is observed in cell 2 (-0.01% from the initial 5.0%),
while the porosity change in all other cells is <0.01%. The same results apply for scenario 2b
considering a calcite vein and an initial porosity of 5.0%. Calcite veins with a higher initial
porosity of 5.5% (scenario 2c) lead to a decrease in porosity in cell 2 (-0.01%) and to a weak
increase in all other cells (∆ +0.01%). This trend is measurable up to an initial porosity of
10.0%. Another scenario (2d) considers an initial porosity of 40.0% in the calcite veins, as
modeled by Bildstein et al. (2010). Despite such a high initial porosity of 40% in calcite veins,
the modeling results show a decrease in porosity of -0.05% in cell 2 but constant porosities in
the following cells. In summary, secondary porosity can be created when the initial porosity
ranges between ~5.2% and ~10%. The pH in the cap rock brine decreases from 7.9 to 6.8 in
all four scenarios (2a–d) and is lower than in the reference scenario (pH 8.4).
2.5.2.3 Initial Cap Rock Porosity
With a scenario series (3a–d) different initial cap rock porosities between 2.0% and 40.0% are
calculated to test the effects of this parameter (Figure 2.3c). The high initial cap rock porosity
of 40% conceptually represents that open fractures occur in the cap rock. The modeling results
with an initial cap rock porosity of 40.0% show a strong decrease in porosity in cell 2 (-16.34%),
a decrease of between 8.34% and 31.66% (ø -9.5%) in cells 3–12 and a -3.4 to -3.7% loss in
porosity for cells 13–101. This trend is also traceable for initial porosities of 10.0%, 6.0% and
2.0%, as well as for the reference scenario with an initial porosity of 5%. The pH of the cap
rock brine lies between 8.4–8.5 in all scenarios (3a–d).
2.5.2.4 Effective Diffusion Coefficient
Scenarios 4a–c aim to unravel the effects of diffusion intensity by varying the effective diffusion
coefficients from 1.00 × 10-10 m2 s-1 (Bildstein et al., 2010), 4.50 × 10-11 m2 s-1 (Gaus et al.,
2005) to 1.00 × 10-11 m2 s-1 (Bildstein et al., 2010). As in the reference scenario, the pH in the
cap rock brine increases in all scenarios (4a–c) to ~8.4. Despite a broad range of diffusion
coefficients, similar results indicate that the diffusion intensity has no effect on changes in
38
porosity (Figure 2.3d). However, this conclusion is only valid for the assumed chemical,
physical and mineralogical conditions.
2.5.2.5 Injected Fluid Composition
Based on the study by Mohd Amin et al. (2014), several scenarios consider various
compositions of injected fluids, including CH4 and H2S mixtures with the main component CO2.
Assuming an ideal gas behavior in the gas mixtures, the fugacity coefficients will be calculated
for the gas components. Modeling results of scenario 5a (99.0% CO2 + 1.0% CH4) show a
strong decrease in porosity in cells 2 and 3 (from the initial 5.0% to 0.0%–3.07%) and a
moderate decrease in porosity in the remaining cells (down to ~4.57%). An increase in the
concentration of CH4 in the injected fluids from 1.0% to 50.0% leads to a weaker decrease in
the porosity in cells 2 and 3 but similar values in the remaining cells. The addition of 1.0% H2S
in scenario 5c decreases the porosity in cell 2, down to 2.79% in cell 3, whereas the porosity
increases slowly up to 4.57% in cell 10 and stays constant in the remaining cells. Summarizing,
the main differences in porosity change, when using the various fluid compositions, are
identifiable in the first cells (2–10) where the change varies from -0.6% to -1.2%. The porosity
changes in cells 11–101 is independent of the injected fluid composition (from pure CO 2 to
CH4 and H2S mixtures; Figure 2.2 and Figure 2.3e). Similar to the reference scenario, the pH
increases in scenarios 5a–c to ~8.4. As mentioned by Mohd Amin et al. (2014), when using
CH4 and H2S mixtures with the main component CO2, the porosity in the cap rock decreases
and the sealing capacity of the cap rock increases.
2.5.2.6 Initial Conditions According to Published Case Study
In comparison with our reference scenario, Bildstein et al. (2010) used a different mineral
assemblage in the cap rock, including illite, anhydrite, kaolinite, quartz, pyrite, ankerite, Ca-
montmorillonite and calcite (for percentages see Appendix 2.7). Scenario 6a considers the
same mineral assemblage under the pressure and temperature conditions of the reference
scenario. In scenario 6a the cap rock porosity decreases in cell 2 (∆ -0.7%), to 2.0% in cell 3
and to 4.3% in cell 4 (Figure 2.4). In the following cells, the porosity increases to 5.9% (cells 5
and 6) and remains constant at ~5.6% (cells 5–101). In cells 5–101 illite, quartz, ankerite, and
dawsonite precipitate and anhydrite, kaolinite, calcite, and Ca-montmorillonite dissolve.
Another scenario (6b) uses the P−T conditions and the mineral assemblage given by Bildstein
et al. (2010). In this scenario, the porosity of the cap rock decreases to 1.11% in cell 2 but
39
shows a strong increase in cell 2 to 7.46%. The average for the cells 11–101 lies at 6.2%
(Figure 2.3f). In scenario 6a the pH in the cap rock brine decreases from 6.54 to ~5.3 and in
scenario 6b the pH increases to 10.4.
Figure 2.4 Gain or loss of total porosity and mass conversion of the minerals for scenario 6a in the first
10 cells with a cell length of 1 m. 9 m are located in the cap rock (CR) and 1 m in the reservoir rock
(RR).
2.5.2.7 Advection as Transport Mechanism
The ability of the cap rock to safely store CO2 depends, amongst other factors, on the reservoir
pressure. If the reservoir pressure becomes too large, new fractures and faults are created
(Rutqvist et al., 2007) and pose a high risk of CO2 leakage. Scenario 7 simulates an intensive
CO2 leakage in fractures and faults by considering advection as the transport mechanism
instead of diffusion in the reference scenario (Figure 2.3g). Compared to diffusion, advection
enables a faster, stronger and more expansive spread of CO2. This leads to a lower pH in
scenario 7 (to 3.3–5.0 in cells 2–13 and 6.9–7.1 in cells 14–101) compared to 8.4 in all cells of
the reference scenario. Such a strong decrease in pH accompanies the complete dissolution
of K-feldspar, albite, kaolinite, calcite, chlorite in cells 2–13. However, no secondary porosity
is created in these cells because the volume of newly formed quartz and dawsonite
40
overcompensates the volume resulting from dissolution of K-feldspar, albite, kaolinite, calcite,
chlorite.
2.5.2.8 Changing Temperature at a Constant Pressure
Scenarios 1a–c reveal the change in porosity in dependence on temperature and pressure. In
contrast, in scenarios 8a+b only the temperature is changed and the pressure is set constant
to show the effect of temperature on the porosity change (Figure 2.3h). However, a constant
pressure assumed in these scenarios does not comply with a real system. In scenario 8a,
change in porosity is limited to cells 2 to 11: increases by 0.7% in cell 2 and decreases by -1.1%
on average in cells 3–11. Such calculated changes in porosity are less intensive when
compared to scenario 1a. When comparing these results with the results of scenario 1a strong
differences are observed in cell 2. In scenario 1a the porosity decreases and in scenario 8a
the porosity increases. In cell 3 to 11 the opposite effect is detectable. Whereas the porosity
increases in scenario 1a, it decreases in scenario 8a. The pH of the cap rock brine for both
scenarios (8a and 1a) decreases from the initial 7.9 to 6.7. In scenario 8b the calculated
porosity decreases in all cells, which is similar to scenario 1c. The difference with scenario 1c
(in which the temperature and pressure are changed) is detectable in cell 2 only where a
difference in the third decimal place is observed. Even the pH of the cap rock brine (8.5–8.6)
is the same in scenario 8b and 1c. In summary, scenarios 8a and 8b show that the temperature
has a stronger effect on mineral dissolution and precipitation at high pressure and temperature
conditions.
2.5.2.9 Changing Pressure at Constant Temperature
Scenarios 9a+b aim to test the influence of pressure on mineral dissolution and precipitation
(Figure 2.3i). In scenario 9a, the porosity increases in cell 2 but decreases in cell 3–11,
whereas no change in the remaining cells is observable. In contrast, in scenario 1a the porosity
decreases in cell 2 only, while all other cells show a porosity increase. As in scenario 1a, the
pH of the cap rock brine decreases from the initial 7.9 to 6.7 in scenario 9a.
The modeling results of scenario 9b show that the porosity decreases in all cells. When
comparing the results between scenario 9b and scenario 1c (where the temperature and
pressure are changed), differences are detectable in the third decimal place in cells 2 and 3
only. The same applies for the pH of the cap rock brine, which increases to 8.5 in scenario 9b
41
and to 8.5–8.6 in scenario 1c. Like the effect of temperature (scenarios 8a+b), the effect of
pressure on the change in total porosity is stronger at higher temperature and pressure
conditions. Comparing the effect of temperature with the effect of pressure on porosity change,
only small differences are identifiable. The pressure effect is slightly stronger than the
temperature effect but there are differences in the second decimal place.
2.5.3 Sensitivity Analysis
The sensitivity analysis aims to show the effect of different solubility constants for CO 2(g) on
the modeling results. We vary the log K values for CO2(g) from initial -1.468 to -1.769 and -1.167
for the reference scenario. These values are chosen by doubling the solubility constant and by
reducing the solubility constant by 50%. The modeling results show that the solubility constant
of CO2(g) influences the final simulation results. The strongest change in porosity is identifiable
in cells 2–4 whereby the change in the following cells (5–10) is at most in the second decimal
place (Table 2.6). However, the general trend of decreasing porosity in the reference scenario
does not depend on the chosen solubility constant.
Table 2.6 Comparison of solubility constants of CO2(g) based on the change in total porosity in the
reference scenario for cells 2–10. log K -1.468 is the solubility constant used in the reference scenario.
Solubility log K -1.769 log K -1.468 log K -1.167
constant CO2(g)
Cells Decrease in porosity (%)
2 -7.2 -5.5 -4.4
3 -1.1 -2.3 -2.0
4 -0.4 -0.6 -1.6
5 -0.5 -0.5 -0.5
6 -0.5 -0.5 -0.5
7 -0.5 -0.5 -0.5
8 -0.5 -0.5 -0.5
9 -0.4 -0.4 -0.4
10 -0.4 -0.4 -0.4
2.5.4 Modeling Limitations
The model presented here uses thermodynamic equilibrium reactions and the local equilibrium
assumption for CO2 storage systems. It does not aim to reflect the complexity of the whole
system. However, there is a general agreement that models enable critical analysis of a system
and that “model testing and the evaluation of predictive errors lead to improved models and
better understanding of the problem” (Konikow and Bredehoeft, 1992).
42
Our modeling approach is based on assumptions regarding chemical and physical parameters.
It excludes kinetic aspects of mineral dissolution/precipitation. The used database phreeqc.dat
includes the pressure and temperature dependence of the mass-action law constants for the
equilibrium reactions of many aqueous, gaseous and solid species involved but with the
exception of the pressure dependence of dawsonite and nahcolite which may be of great
relevance for simulating CO2 storage systems.
For the sake of simplicity, it is neglected that the modeling results of one cell could change the
initial parameters for the neighboring cells. Changes in porosity and permeability affect the
tortuosity. Such changes should be adapted in every single cell after every time step of the
modeling. There is no valid relationship between permeability and porosity because the
permeability-porosity correlation is site-specific and depends on many geometric factors
(Verma and Pruess, 1988). However, it is shown in laboratory experiments (Vaughan, 1987)
as well as in field data (Pape et al., 2000) that even small reductions in porosity, resulting from
mineral precipitation, can significantly reduce permeability. Furthermore, newly formed solid
phases close pore throats and lead to stronger reductions in permeability (Verma and Pruess,
1988).
Even if the diffusion coefficient is specific to each aqueous species, depending on the species
type, pressure, temperature and interaction with other species (Gaus et al., 2005), the used
computer code PHREEQC only allows a uniform diffusion coefficient for all aqueous species
of 3.33 × 10-10 m2 s-1. Nevertheless, the modeling results indicate that a change in the diffusion
coefficient has no effect on the calculated results. Regarding the diffusion of dissolved CO2 in
cap rock, Fleury et al. (2009) suggest that the diffusion coefficients depend primarily on the
porosity. Consequently, in real CO2 storage systems the effective diffusion coefficient must be
calculated not only in dependence on the species and temperature of each cell but also on the
newly formed porosity, changed by CO2−water−rock interactions, in each cell. Furthermore, a
homogeneous distribution of relevant cap rock parameters is considered in our study. In
contrast, a heterogeneous distribution of relevant cap rock parameters can lead locally to
different results. Moreover, tests of different boundary conditions (constant, closed and flux)
show no effects on the modeling results.
2.6 Conclusions
The application of the 1DRMT modeling presented in this study provides a good example of
how such a modeling approach could (i) identify mineral dissolution and precipitation in the
43
cap rock triggered by molecular diffusion of CO2 and (ii) quantify the resulting porosity change,
even allowing for several limitations.
Because of hydrogeochemical reactions, the results of the reference scenario show a change
in porosity from cell to cell over the length of 101 m after 106 years of diffusive mass transport.
Quartz and dawsonite precipitate, whereas K-feldspar, albite, kaolinite, calcite, chlorite and
barite dissolve. Compared to these dissolved minerals, the more intensive precipitation of
quartz and dawsonite causes a decreasing cap rock porosity in the reference scenario.
Generally, the main changes in porosity are identifiable close to the cap rock/reservoir rock
contact.
Alternative scenarios for selected physical and chemical parameters were performed in this
study to identify the key factors controlling the change in porosity. The modeling results show
that the creation of porosity depends on different factors. The following conclusions are based
on the results of the alternative scenarios.
• Mineral dissolution and precipitation and resulting porosity changes are temperature
and pressure dependent, because temperature and pressure affect the mass-action
laws of minerals and aqueous species involved in the equilibrium reactions. In the
tested range of 30.7–150.0 °C and 50.0–470.0 atm, both temperature and pressure
have a stronger effect on mineral dissolution and precipitation at higher P–T conditions.
Therefore, it is important to consider that the initial temperature and pressure conditions
can differ from location to location in a single system (i.e. in folded structures).
Compared to temperature, the effect of pressure on the change of porosity is slightly
stronger.
• Regarding temperature and pressure conditions, there is a critical point where the
porosity development changes from decrease to increase. For the reference scenario
the critical point lies between 135.0 °C and 150.0 °C and 440.0 atm and 470.0 atm.
This critical point would vary when the initial conditions change, specifically the initial
mineralogical compositions of the cap and reservoir rocks.
• Depending on the initial porosities, calcite, which frequently occurs as veins in cap
rocks, can dissolve. Therefore, calcite veins could be a risk for the sealing capacity of
the cap rock.
• When using initial cap rock porosities between 2.0% and 10.0%, a decrease in total
porosity over the length of one hundred meters of modeled cap rock is identifiable.
These results indicate that under the assumed physical, chemical and mineralogical
conditions, the sealing capacity of the cap rock is improved.
44
• Using the mineralogical phase assemblage from Bildstein et al. (2010), an increase in
porosity from the fifth meter onwards is identifiable, so the sealing capacity of the cap
rock decreases. This result indicates that porosity creation depends strongly on the
mineralogical composition of the reservoir and cap rock as well as on the brine
compositions.
• Even over a broad range tested, the intensity of diffusive mass transport has negligible
effects on mineral precipitation and dissolution as well as on resulting changes in
porosity.
• The injected fluid compositions affect the sealing capacity of cap rocks which can
increase by mixing CO2 with CH4 and H2S in comparison with pure CO2 injection.
• An intensive CO2 leakage in fractures and faults enables a faster and broader spread
of CO2, and consequently, could cause an earlier and stronger sealing risk.
• Compared to nahcolite, dawsonite precipitates in higher amounts and is one reason for
the decrease in porosity. However, the potential of dawsonite to trap CO2 permanently
is still a point of discussion (Hellevang et al., 2005).
Therefore, the chemical and physical data of the cap and reservoir rocks and brine must be
measured under in situ conditions to develop a numerical model for evaluating a specific CO2
storage system. Various cap rock mineral phase assemblages should be exposed to different
aqueous solutions, partial pressures of carbon dioxide and P–T conditions in long-term
laboratory batch experiments to observe mineral dissolution and/or precipitation. These results
should be reproduced by batch modeling using identical initial parameters. Such a combination
of laboratory experiments and modeling should be performed to test the plausibility of the
applied hydrogeochemical equilibrium model.
45
2.7 Appendix
Table A 2.1 Data from Bildstein et al. (2010).

Temperature = 80 °C, ρCO2 = 150 bar
Initial mineralogical composition of Reservoir water Acidified reservoir Initial cap rock water
the cap rock water
Weight percent pH 6.24 pH 4.75 pH 6.54
(wt. %)
Calcite 50 Species Molality Species Molality Species Molality
Ankerite 5 Al 5.622 × 10-8 Al 1.251 × 10-7 Al 1.531 × 10-7
Montmorillonite 25 C 4.895 × 10-3 C 1.141 C 2.180 × 10-3
Kaolinite 3 Ca 1.612 × 10-2 Ca 3.204 × 10-2 Ca 1.528 × 10-2
Illite 2 Cl 3.014 × 10-1 Cl 3.015 × 10-1 Cl 2.601 × 10-1
Quartz 10 Fe 2.137 × 10-7 Fe 1.751 × 10-6 Fe 1.534 × 10-5
Anhydrite 3 K 2.374 × 10-3 K 2.375 × 10-3 K 1.190 × 10-2
Pyrite 2 Mg 1.282 × 10-2 Mg 2.424 × 10-2 Mg 8.937 × 10-4
Na 2.594 × 10-1 Na 2.595 × 10-1 Na 2.543 × 10-1
S 7.642 × 10-3 S 7.649 × 10-3 S 1.841 × 10-2
Si 8.994 × 10-4 Si 8.833 × 10-4 Si 5.371 × 10-4
Table A 2.2 Additional equilibrium phases, mass-action equations and equilibrium constants used in
scenarios 6a–b. Data from phreeqc.dat (Parkhurst and Appelo, 2013), but ankerite is from Bildstein et
al. (2010).
Equilibrium phase Equilibrium reaction log K
at 25 °C,
1 bar
Ankerite CaFe0.7Mg0.3(CO3)2 + 4H+ + 0.7Fe2+ + 0.3Mg2+ + 2H2O + 2CO2 12.14a
Ca0.165Al2.33Si3.67O10(OH)2 + 12H2O = 0.165Ca2+ + 2.33Al(OH)4- + 3.67H4SiO4 + -45.027
Montmorillonite
2H+
K0.6Mg0.25Al2.3Si3.5O10(OH)2 + 11.2H2O = 0.6K+ + 0.25Mg2+ + 2.3Al(OH)4- + -40.267
Illite
3.5H4SiO4 + 1.2H+
Anhydrite CaSO4 = Ca2+ + SO42- -4.39
a
: log K at 80 °C and 1 bar
2.8 Acknowledgements
We would like to thank the anonymous reviewers for their constructive reviews that
considerably improved the manuscript.
2.9 Supplementary Data (on CD)
Input file S1 for the batch model to calculate the cap rock brine and Input file S2 for the transport
model to calculate the reference scenario SCGS are available (for PHREEQC Version 3).
46
Salt Cavern Gas Storage (SCGS)
3 Salt Cavern Gas Storage (SCGS)
Potential Risk of H2S Generation and Release in Salt Cavern Gas Storage
Publication II
Originally published in: Journal of Natural Gas Science & Engineering 47 (2017), 114–123



1
47
Summary
In this part of the study, the gas–water–rock interactions triggered by natural gas storage are
considered, and the risks of H2S(g) generation in salt caverns, which are used for natural gas
storage, are analyzed. The focus is on the spatial- and time-dependent H2S(g) generation
processes in salt caverns due to microbial activity (bacterial sulfate reduction) and on technical
methods to decrease or inhibit H2S(g) generation and release.
Storage of natural gas in salt caverns entails risks of hydrogen sulfide (H2S(g)) generation.
H2S(g) is toxic, leads to gas souring and the corrosion of the storage facilities. Therefore,
technical regulations prescribe that the H2S(g) concentration in stored gas is limited to 5 mg m-3.
H2S is generated by bacterial sulfate reduction (BSR) in the brine and the upper sections of
the sump of a salt cavern. Sulfate-S is used by the bacteria as an electron acceptor to oxidize
organic compounds and sulfide is generated. In salt caverns, the sulfate source is provided by
anhydrite layers within the salt rock. Stored methane (the main component of natural gas)
dissolves in the brine until saturation is established and continuously serves as a reductant for
BSR. The H2S(g) generated in brine and sump contaminates the stored gas by outgassing.
These hydrogeochemical processes are quantitatively retraced using a one- and three-
dimensional mass-transport model based on chemical equilibrium thermodynamics using
PHREEQC and PHAST software with the phreeqc.dat database. Reaction kinetics of BSR are
integrated into the model. The modeling approach simulates a semi-generic salt cavern at high
pressure and temperature conditions with data from several caverns. The modeling results
give basic and quantitative insights into the mechanisms of H2S(g) generation induced by BSR
in salt caverns and lead to a general system- and process-understanding of the
hydrogeochemical effects of natural gas storage in salt caverns.
By varying the input parameters, the factors controlling H2S(g) generation are identified and
explain why H2S(g) is only generated in some salt caverns. An important factor is the availability
of anhydrite as a sulfate source. Whereas the occurrence of Fe-bearing minerals like goethite
inhibits the release of BSR-generated H2S into the stored gas. The sulfur (S-II/-I) reacts with
Fe+II, and mackinawite or pyrite are formed. To identify early H2S generation and protect the
stored gas from souring, a monitoring system should be installed in the brine of the salt cavern.
If the aqueous H2S concentration increases, the addition of dissolved ferrous iron into the brine
and sump is a potential method to reduce H2S release. The hydrogeochemical model approach
can be used and adapted to various salt cavern conditions to answer and solve practically
oriented questions and problems (Hemme and van Berk, 2017b).
48
Abstract
The storage of natural gas in salt caverns can entail the risk of H2S generation, which in turn
leads to gas pollution. H2S is generated by bacterial sulfate reduction. The bacteria use
aqueous sulfate(aq) as an electron acceptor to oxidize the dissolved hydrocarbons and generate
sulfide. Anhydrite is available in the rock salt surrounding the cavern and acts as a sulfate (aq)
source. The stored natural gas, with its main component, methane, is in solubility equilibrium
with the brine and is additionally delivered by diffusion into the brine. The generated H2S
reaches the stored gas by outgassing from the brine. In this study, these processes are
simulated by one- and three-dimensional hydrogeochemical diffusive mass transport models,
which are based on equilibrium reactions for gas–water–rock interactions and kinetic reactions
for sulfate reduction. Modeling results show that the greatest amount of H2S is generated in
the brine. The amount of generated H2S(g) is mainly controlled by the amount of available
sulfate(aq) as well as the rate of diffusion, which is coupled with the maximum operating live
time of salt caverns. Additionally, the amount of generated and released H2S(g) is sensitive to
the chosen kinetic rate constant. To ensure constant gas quality over time, the gas and the
brine must be analyzed continuously and technical methods must be applied when the H2S(g)
concentration increases. According to the modeling results, H2S(g) generation is inhibited by
addition of dissolved ferrous iron to the brine. Dissolved ferrous iron reacts with sulfide-sulfur
to form mackinawite (FeS(s)) so that aqueous sulfide is no longer available for H2S(g) generation.
Another method is the addition of NaOH to increase the pH of the brine. Then, higher fractions
of generated sulfide-sulfur are transformed to free S2- (aq) instead of H2S(g) and H2S(aq).
3.1 Introduction
Natural gas is stored in salt caverns to balance the supply and demand of natural gas
throughout the year. Salt caverns are highly qualified for hydrocarbon storage because of
numerous physical properties and mechanical behaviors of the rock salt halite, like its self-
healing forces and its impermeability below 300 m (Evans, 2008; Yang et al., 2013). However,
Evans (2008) has stated that “there is a need to assess the safety record of previous and
existing underground fuel storage facilities.” One risk is the potential generation and release
of gaseous hydrogen sulfide (H2S(g)) in natural gas storage systems. H2S is toxic if inhaled, is
aggressive towards storage facilities (Cord-Ruwisch et al., 1987; Kleinitz and Böhling, 2005),
and can pose a threat to the environment (Reitenbach et al., 2015). The presence of H2S can
lead to corrosion of metallic iron under anaerobic conditions and to the precipitation of
amorphous ferrous sulfide, which in turn may cause plugging (Cord-Ruwisch et al., 1987).
49
Even more importantly, H2S contaminates the stored gas and can affect the gas quality (Cord-
Ruwisch et al., 1987). Therefore, in Germany technical regulations determine that the
concentration of 5 mg m-3 H2S(g) in stored gas must not be exceeded (DVGW, 2013).
There are considerable indications that H2S(g) generation could be a potential risk in salt
caverns used for gas storage. First, H2S is observed in hydrocarbon reservoirs where it
originates from sulfate reduction (Machel, 2001), either via abiotic reactions or via reactions
catalyzed by bacteria. The abiotic reaction, so-called thermochemical sulfate reduction (TSR),
is common in geological settings with temperatures ranging from 100 to 180 °C, while the
bacterial sulfate reduction (BSR), occurs in low-temperature geological settings ranging from
0 °C to 60–80 °C (Ehrlich, 1990; Machel, 2001; Postgate, 1984). In some cases, BSR has
been observed above 80 °C. Hyperthermophilic sulfate-reducing bacteria may live at
temperatures up to 110 °C (Jorgensen et al., 1992). However, BSR does not necessarily occur
in all hydrocarbon-bearing geosystems with temperatures below 80 or 110 °C. Otherwise, all
hydrocarbon reservoirs below this temperature would be sour, meaning H2S-bearing (Machel,
2001), and would display higher total sulfide concentrations in the aqueous and the gas phase.
Therefore, in this study, we focus on salt caverns filled by natural gas and exposed to
temperatures ranging from 50 to 80 °C. Our focal point is the formation of sulfide-sulfur (S-II)
from sulfate-sulfur (S(+VI)) via BSR and the subsequent release of formed sulfide-sulfur (S-II)
as H2S(g) into the stored natural gas.
An additional indication of H2S generation by BSR in salt caverns is the possible anaerobic
oxidation of methane (AOM), which is observed in marine as well as in non-marine
environments (Meulepas et al., 2010). In aqueous anoxic environments, sulfate-reducing
bacteria (SRB) use sulfate as an electron acceptor to oxidize organic compounds and generate
sulfide (Equation (3.1)). This generated sulfide-S could be available as aqueous H2S, HS-(aq),
and S2-(aq) and gaseous H2S. SRB use the produced energy from sulfate reduction to sulfide
for cell growth (Cord-Ruwisch et al., 1987). The sulfate for BSR could be derived from the
aqueous dissolution of calcium sulfide-sulfur mineral phases like gypsum (CaSO4[2H2O])(s) and
anhydrite (CaSO4)(s).
H2O
(3.1)
SO42-(aq) + CH4(aq) → H2S(aq) + CO32-(aq) + H2O
The increasing demand for storage capacity in salt caverns requires the utilization of less
favorable salt formations, including inhomogeneous salt structures with larger proportions of
insolubles like anhydrite layers (Schneider and Crotogino, 2010). Drilling operations and/or
workover operations may lead to bacterial contamination of hydrocarbon reservoirs, or SRB
50
populations may pre-exist in such reservoirs (Kleinitz and Böhling, 2005). The optimal growth
temperature for SRB is 38 °C (Bernardez et al., 2013) at near-neutral pH conditions (Cord-
Ruwisch et al., 1987). However, SRB also occur in more acidic environments of pH 3 (Tuttle
et al., 1969) and pH 4 (Church et al., 2007).
Furthermore, BSR is observed in saline environments where high rates of sulfate reduction
are measured (Kjeldsen et al., 2007). The activity of most SRB decreases if the Na+/Cl-
concentrations are above 50 to 100 g L-1 (Cord-Ruwisch et al., 1987; Postgate, 1984; ZoBell,
1958) but activity of SRB is even found in salt lakes and brines near “salt saturation” (ZoBell,
1958). Even if these conditions are not the optimum for SRB growth, a few SRB tolerate the
high salt (NaCl) concentrations and live near salt saturation (Cord-Ruwisch et al., 1987).
Additionally, H2S(g) is detected in underground storage systems of town gas (Crotogino, 2016)
and in underground gas storage in porous media (Kleinitz and Böhling, 2005). Furthermore,
the activity of sulfate-reducing bacteria is observed in salt caverns filled by hydrogen gas.
There, the SRB live in the sump and in the brine, generating biofilms at the cavern walls
(Panfilov, 2016).
This study focuses on H2S generation by bacterial sulfate reduction in a salt cavern that is
described by one- and three-dimensional hydrogeochemical reactive transport models. It is
based on thermodynamic equilibrium reactions for gas–water–rock interactions and kinetic
reactions for sulfate reduction. The aims of this study are (i) to draw the attention of the possible
risk of H2S(g) pollution in salt caverns, (ii) to clarify and quantify time-dependent H2S(g)
generation processes in salt caverns filled with natural gas, (iii) to analyze the limiting factors
for H2S(g) generation and release in salt caverns, and (iv) to identify technical methods to
decrease or inhibit H2S(g) generation and release.
51
Figure 3.1 System sketch of the model. Black box = reference scenario, orange box = alternative
scenario.
3.2 Methodology
The one- and three-dimensional reactive mass transport models are based on chemical-
thermodynamically principles, the reaction kinetics of BSR and the principles of diffusive mass
transport.
The modeling tool for the 1D model in this study is the computer program PHREEQC version 3
provided by the US Geological Survey. PHREEQC is based on an ion-association aqueous
model and can simulate batch-reaction, speciation, inverse geochemical and one-dimensional
transport calculations (Parkhurst and Appelo, 2013). The calculations are based on mass
action laws including all species and their corresponding equilibrium constants. The activity
coefficients of species are calculated by the Debye-Hückel equation. The equilibrium phases,
mass-action equations, and equilibrium constants used in the model are shown in Table 3.1.
52
The computer program PHAST (version 3.3.7-11094), provided by the US Geological Survey,
is the modeling tool for the 3D model. Using PHAST, multicomponent geochemical reactions,
solute transport and groundwater flow can be simulated (Parkhurst et al., 2010). The
geochemical reactions in PHAST are simulated with PHREEQC and the flow and transport
calculations are based on HST3D; both programs are embedded in PHAST (Parkhurst et al.,
2010). The results are visualized using the software Model Viewer (Hsieh and Winston, 2002).
The combined application of PHREEQC/PHAST and the Model Viewer software enables the
visualization of the temporal and spatial development of H2S generation in salt cavern gas
storages. Detailed information about PHREEQC and PHAST are given in Parkhurst and
Appelo (2013) and Parkhurst et al. (2010).
Table 3.1 Equilibrium phases, mass-action equations, and equilibrium constants (log K, at 25 °C and
1 bar). Data are from phreeqc.dat, except for CH4(g), H2S(g), N2(g) which are from llnl.dat (Parkhurst and
Appelo, 2013).
Equilibrium phase Equilibrium reaction log K
Halite NaCl = Cl- + Na+ 1.570

Anhydrite CaSO4 = Ca2+ + SO42- -4.39
Siderite FeCO3 = Fe2+ + CO32- -10.89
Quartz SiO2 + 2H2O = H4SiO4 -3.98
Barite BaSO4 = Ba2+ + SO42- -9.97
Pyrite FeS2 + 2H+ + 2e- = Fe2+ + 2HS- -18.479
Dolomite CaMg(CO3)2 = Ca2+ + Mg2+ + 2CO32- -17.09
Mackinawite FeS + H+ = Fe2+ + HS- -4.648
Sulfura S + 2H+ + 2e- = H2S 4.882
Calcite CaCO3 = CO32- + Ca2+ -8.48
CH4(g) CH4 = CH4 -2.8502
CO2(g) CO2 = CO2 -1.468
H2S(g) H2S = H+ + HS- -7.9759
N2(g) N2 = N2 -3.1864
a: Sulfur = elemental sulfur
The thermodynamic database; which includes all elements used in the model with their species
(aq, s, g), mass-action equations, and equilibrium constants; is essential for modeling with
PHREEQC and PHAST. The database used for 1D and 3D modeling is phreeqc.dat. A more
suitable database for the high Na+ and Cl- concentrations and the high ionic strength in the
model could be the Pitzer database (pitzer.dat), but pitzer.dat does not include Si-containing
aqueous species, Al3+, and silicate minerals, which are important factors when modeling H2S
generation in salt caverns. To validate that PHREEQC (using phreeqc.dat) produces correct
results, even under high Na+ and Cl- concentrations and high ionic strength, the salt solubility
in PHREEQC (using phreeqc.dat) is compared with salt solubility data from literature. In
Zimmermann et al. (1986), the solubility of Na+/Cl- is given in dependence of the temperature.
At 40 °C, 36.42 g Na+/Cl- (halite) dissolves in 100 g water (= 6.23 mol kgw−1 for Na+ and Cl-).
53
This measurement is consistent with the results of Na+/Cl- solubility in PHREEQC using
phreeqc.dat (= 6.32 mol kgw−1 for Na+ and Cl-), with the discrepancy that the temperature in
the model is 50 °C. NaCl(solid) solubility data in pure water at different temperatures from
literature (experimental data) are compared with modeled data (PHREEQC with phreeqc.dat)
and are summarized in Table 3.2.
Table 3.2 NaCl(solid) solubility (mol kgw−1) in pure water at different temperatures from literature
(experimental data) and modeled data. The database phreeqc.dat is used in the PHREEQC model.
Solubility NaCl(solid) - Data 0 °C 20 °C 40 °C 60 °C
(mol kgw−1) (mol kgw−1) (mol kgw−1) (mol kgw−1)
Zimmermann et al. (1986) 6.09 6.13 6.23 6.34
PHREEQC modeling 6.00 6.10 6.24 6.39
3.2.2 Model Setup
Salt caverns are divided into three parts (Figure 3.1): the stored gas at the top, which takes up
the largest volume; the brine in the middle, with a thickness of only a few meters; and the sump
at the bottom, which can occupy one third of the total cavern volume (Bérest and Brouard,
2003; Panfilov, 2016). The sump is composed of insoluble residues (Evans, 2008) and a
residual pore-filling aqueous solution that is not discharged after leaching (Panfilov, 2016). Due
to technical reasons, residual brine is situated above the sump (Haddenhorst, 1989). This brine
is in solubility equilibrium with the gas stored above (Fontenot, 1981; Réveillère et al., 2016)
and the surrounding mineral phases of the rock salt formation.
The initial gas composition in the model is based on data from DVGW (2013). The main
component is CH4(g) (partial pressure of CH4 (ρCH4) = 178.128 atm), with minor amounts of
N2(g) (ρN2(g) = 1.548 atm), CO2(g) (ρCO2(g) = 0.324 atm), and H2S(g) (ρH2S(g) = 0.00018 atm),
hereinafter called “initial gas”. For the sake of simplicity, the partial pressure is assumed to be
equal to the fugacity f (ρ = f). The total gas pressure is 180 atm, and the temperature in the
salt cavern is 50 °C.
The mineralogical composition of the sump and the brine are characterized by the composition
of the surrounding rock salt formation and the solution used for leaching. The mineralogical
composition of the rock salt formation is based on data from Kyle and Posey (1991) and is
shown in Table 3.3. For leaching, the chemical composition of groundwater taken from a ca.
20 m depth quaternary aquifer from Lower Saxony, northern Germany (NLWKN, 2015) is used
(Table 3.4).
54
Table 3.3 Initial mineralogical composition of the rock salt formation used for the sump with a porosity
of 20%. The data are modified after Kyle and Posey (1991). Mackinawite, elemental sulfur, and calcite
are potential secondary phases which may form at saturation.
Primary mineral phases Weight percent (wt. %) Amount (mol kgw−1)
Halite 97.0 144.62

Anhydrite 2.75 1.76
Siderite 0.05 0.04
Quartz 0.05 0.07
Barite 0.05 0.02
Pyrite 0.05 0.04
Dolomite 0.05 0.02
A salt cavern with a height of 350 m, divided into 300 m stored gas, 2 m brine, and 48 m sump,
is assumed. The focal point of the model is the brine-gas interface and the upper meters of the
sump (Figure 3.1). Therefore, the model is divided into a column of 8 cells, with a cell height
of 1.0 m each (in the z-direction). The boundary at the upper end of the column represents a
constant amount of stored gas. At the lower end of the column, the boundary condition is
defined as diffusive flux. The only mass transport in this model is molecular diffusion of all
aqueous species. One diffusion coefficient for all aqueous species of 5.0 × 10-9 m2 s-1 is used.
Cell 1 delivers CH4 (C(-IV)) continuously by diffusion. Cell 2 is located at the brine-gas
interface, cell 3 in the brine, and cells 4–8 in the sump. Each cell is defined by specific
mineralogical and hydrochemical properties. The brine and residual aqueous solution
compositions are calculated in separated batch models with PHREEQC (Table 3.4).
55
Table 3.4 Groundwater composition used for leaching (NLWKN, 2015), initial composition of the brine,
and the aqueous solution in the sump.
Groundwater Solution in the sump Brine
pH 6.4 8.2 5.7
Temperature (°C) 10.1 50.0 50.0
Elements Concentration Concentration Concentration

(mol kgw−1) (mol kgw−1) (mol kgw−1)
Ba -a 8.136 × 10-07 9.097 × 10-07

C -a 2.901 × 10-05 7.077 × 10-03
Ca 1.622 × 10-03 5.488 × 10-02 6.333 × 10-02
Cl 8.380 × 10-04 6.306 6.310
Fe 1.522 × 10-06 2.257 × 10-03 1.415 × 10-03
K 1.010 × 10-04 1.010 × 10-04 1.010 × 10-04
Mg 3.150 × 10-04 1.464 × 10-02 1.315 × 10-03
Mn 9.100 × 10-07 9.100 × 10-07 9.100 × 10-07
Ntotb 1.152 × 10-03 1.152 × 10-03 3.008 × 10-04
Na 7.050 × 10-04 6.306 6.310
O(0) 6.600 × 10-05 -a -a
P 4.840 × 10-07 4.840 × 10-07 4.840 × 10-07
Stotc 8.540 × 10-04 7.108 × 10-02c 6.262 × 10-02c
Si 2.560 × 10-04 5.242 × 10-05 3.368 × 10-02
a: Not present
b: summed concentration of aqueous N(-III), N(+III), N(+V) species
c: summed concentration of aqueous S(+VI) and S(-II) species
The brine (cell 2) is composed of the solution used for leaching and is equilibrated with the
mineralogical composition of the rock salt and the initial gas composition. It has a total volume
of 1 L, which is equivalent to 1 kg of pure water (kgw). The gas phase in cell 2 has a total
volume of 270,000 L (based on the model height for gas of 300 m, a base area of 0.005 m2,
and a pressure of 180 atm) and represents the stored gas volume at 180 atm under the
assumed salt cavern design. For modeling purposes, this gas is a tracer gas with the same
chemical properties as CH4(g). It is used instead of CH4(g) because CH4 is induced and used as
reactant in the calculation kinetics. The partial pressure of the tracer gas is equivalent to the
total pressure in the cavern so that the generated H2S is released as gas bubbles. The brine
in cell 3 is composed of the solution used for leaching equilibrated with the mineralogical
composition of the rock salt but without being in equilibrium with a gas phase. However, the
possibility of H2S(g) outgassing is assumed, and calcite, mackinawite, and elemental sulfur are
potential secondary phases that may form at saturation.
The residual aqueous solution in the sump (cells 4–8) is composed of the solution used for
leaching equilibrated with the mineralogical composition of the rock salt formation (Table 3.4).
An initial porosity of 20% is assumed in the sump. The amount of each mineral phase is
calculated in moles per kg of pore water by considering the specific density of each mineral
56
phase (g cm-3) and is summarized in Table 3.3. Additionally, calcite, mackinawite, and
elemental sulfur are potential secondary phases that may form at saturation. No gas phase is
available, but the possibility of H2S(g) outgassing is assumed in all cells.
The oxidation of methane by sulfate in the model system is kinetically controlled. The sulfate
reduction rate of 6.46 × 10-10 mol kgw−1 s−1 at 55 °C is derived from Adams et al. (2013). The
initial amount of CH4(aq) in the brine is 0.027 mol kgw−1. That corresponds to the maximum
amount of CH4(g) in the brine under assumed solubility–equilibration–conditions at a pCH4(g) of
180 atm and 50 °C. A separate transport model provides the initial amount of CH4(g) in the
sump for the kinetic calculation. This amount corresponds to the maximum available amount
of CH4(g) that could be delivered to the sump by diffusion in a timeframe of 30 years (because
the typically operating lifetime of a salt cavern is 30 years). In 30 years, 0.007 mol kgw−1 CH4
is available in the first meter of the sump, 0.004 mol kgw−1 CH4 in the second meter,
0.003 mol kgw−1 CH4 in the third meter and 0.001 mol kgw−1 CH4 in the fourth meter of the
sump.
The diffusive mass fluxes are calculated by multiplying diffusivity with tortuosity. Diffusivity is
defined by the coefficient of molecular diffusion of aqueous species in dilute solutions without
a porous matrix (Fu et al., 2016). A diffusivity of 10-9 m2 s-1 for all aqueous species involved in
the model is assumed. Tortuosity describes the relation of effective diffusion mass fluxes in a
porous medium to ideal diffusion mass fluxes in dilute solutions without a rock matrix and varies
between 0.0 and 1.0. To show the influence of the diffusive mass transport on the modeling
results, the values of tortuosity in different modeling scenarios are systematically varied
(Section 3.3.2.4).
Alternating operating phases characterize the total gas pressure in a salt cavern, divided into
a minimum pressure phase, injection pressure phase, maximum pressure phase, and
production pressure phase (Wang et al., 2015). The pressure is varied between a minimum of
60 atm and a maximum of 180 atm in different modeling scenarios (Section 3.3.2.1). Because
PHAST is not capable of calculating the pressure dependence of solubility equilibrium
constants for minerals, gases and aqueous species, the pressure dependence is simulated in
a separate model using the software PHREEQC and the database phreeqc.dat (Parkhurst and
Appelo, 2013).
57
3.3.1 H2S Generation and Release—Reference Scenario
The results of the modeled reference scenario show that CH4(g) dissolves from the stored gas
into the brine (at the brine-gas interface), according to the pressure/temperature conditions.
CH4(aq) diffuses through the brine and the sump, where SO42-(aq)-ions are available for BSR.
The SO42-(aq)-ions result from anhydrite dissolution and diffuse from the sump to the brine-gas
interface, where CH4(aq) is available. The diffusive mass transport is induced by concentration
gradients and causes a complex web of hydrogeochemical reactions and processes, including
BSR. Modeling results indicate that H2S(g) generation in salt caverns by bacterial sulfate
reduction mainly occurs on the diffusive path where methane and sulfate meet and react with
each other. The highest total H2S concentrations are at this meeting point. Therefore, how
much H2S(g) is generated over time depends on diffusion as well as on the reaction kinetic of
BSR, both depending on temperature and/or pressure. Figure 3.2 shows the results of the 3D
model and indicates the location of H2S(g) generation, as well as increasing H2S(g) concentration
with ongoing time and BSR. The maximum amount of generated H2S(g) is 0.0515 mol kgw−1 (=
6.50 mg m-3 H2S(g)) after 30 years, in total.
Comparing the results of the reference scenario modeled in 3D in PHAST to the results of the
reference scenario modeled in 1D in PHREEQC, slight differences are identifiable. After 30
years, 7.19 mg m-3 H2S(g) are generated in PHREEQC and 6.50 mg m-3 H2S(g) are generated
in PHAST. Additionally, the results of the 1D model show that H2S(g) is released into the stored
gas and how much is released after 30 years and the results of the 3D model show the location
where H2S(g) is generated. Furthermore, PHAST includes flow and transport calculations
(including parameters like tortuosity and dispersivity). Nevertheless, the differences are small
and the chemical processes are the same because the geochemical reactions in PHAST are
simulated with PHREEQC. The following results refer to the 1D model to make them
comparable with the results of the following scenarios (Section 3.3.2). Another reason to use
PHREEQC is the higher interest for industry in H2S(g) concentration in the stored gas than the
point where it is generated.
H2S is formed as a gas bubble in the brine and released into the stored gas above if the sum
of the partial pressure of all dissolved gases(aq) is greater or equal to the total gas pressure in
the stored gas. At 50 °C and 180 atm, a total of 7.19 mg m-3 H2S(g) is released into the stored
gas after 30 years. The value lies above the allowed limit of 5 mg m-3 (defined by DVGW
(2013)).
58
Aqueous sulfate is available in the brine from anhydrite dissolution (during leaching and
operation) and is additionally delivered by diffusion from the sump. Furthermore, the
dissolution of anhydrite contributes Ca2+(aq) into the brine and, in consequence, calcite
precipitates. In 30 years, 0.035 mol kgw−1 calcite precipitate at the brine-gas interface and
buffers the effects of BSR on the pH of the brine. Therefore, the increase in the pH of the brine
over 30 years is small (from 5.7 to 6.1). S(-II), a product of BSR, reacts with the available
aqueous Fe(+II) and mackinawite precipitates (0.004 mol kgw−1 in 30 years), inhibiting H2S
generation. Another possible reaction is S(-II) with Fe(+II) to form pyrite. Using pyrite instead
of mackinawite as a potential secondary phase, the amount of generated H2S(g) is smaller
(5.96 mg m-3). However, the stabilities of pyrite and/or mackinawite at this pressure and
temperature conditions are not identified. Furthermore, a small amount of pure water (H2O)
forms as a product of BSR with a maximum of 0.0004 kg kgw-1. Elemental sulfur is included in
the model as a potential secondary phase but does not precipitate, even if pyrite and
mackinawite are not included as potential secondary phases.
The reference scenario considers the inner part of the cavern and neglects the cavern wall
with its equilibrium phases of the rock salt formation which are in contact with the brine (Figure
3.1, orange box). To cover this part of the salt cavern, an alternative scenario is modeled.
Modeling results show that H2S(g) generation at the cavern wall is less intensive (5.15 mg m-3)
than in the part of the cavern that is not in contact with the cavern wall (the reference scenario,
where no equilibrium phases are available in the brine). This is tested by adding the same
equilibrium phases used in the sump to the brine. Conceptually, a water-to-rock ratio
equivalent to a theoretical porosity of 80% is assumed. The decrease in H2S(g) generation can
be explained with the occurrence of siderite, which is part of the initial rock salt formation. As
long as siderite is available, S(-II) is bound to newly formed mackinawite and inhibits H2S(g)
generation stronger than in the reference scenario. The initial amount of 0.001 mol kgw−1
siderite is completely dissolved in less than 3 years. Furthermore, more calcite precipitates
(0.06 mol kgw−1), which more strongly buffers the effect of BSR on the pH, so that the pH
increase is smaller compared to the H2S(g) generation in the center of the cavern (from 5.7 to
5.8). After less than 12 years, the available anhydrite at the brine-gas interface (cell 2) is
completely dissolved. Furthermore, irreducible water is available in the surrounding rock salt
formation so that diffusional transport of CH4(aq) and available SO42-(aq)-ions is possible even at
the cavern wall in the part of the cavern where the gas is stored. Consequently, H2S(g) could
be generated even there.
59
Gas
Brine H2S(g) (mol kgw-1)

0.0515
0.0386
0.0258
Sump 0.0129
0.000
1m
0 years 5 years 10 years 15 years 20 years 25 years 30 years
Figure 3.2 H2S(g) generation and increasing H2S(g) amount with ongoing time and bacterial sulfate
reduction.
3.3.2 Factors Influencing H2S(g) Generation and Release
Generic model scenarios show the consequences of varying conditions in salt caverns on the
H2S generation and release. In the following Sections, 3.3.2.1–3.3.2.4, the impacts of different
factors are analyzed.
3.3.2.1 Gas Pressure Changes in Salt Cavern Gas Storages
Special consideration is given to the effect of pressure changes in the cavern as a

consequence of gas injection, storage, and production phases. This scenario is based on
Wang et al. (2015) where a 1-year cycle is assumed, divided into 3 months of storage at
minimum pressure (60 atm), 3 months of injection (120 atm, as a “mean” between 60 atm and
180 atm), 3 months of storage at maximum pressure (180 atm), and 3 months of production
(120 atm). The modeling tool for this scenario is PHREEQC, with the database phreeqc.dat
where the pressure-dependent mass-action law constants for the equilibrium reactions of the
involved gaseous, aqueous, and solid species are included. Figure 3.3a shows H2S(g)
generation and release versus pressure.
After the first 3 months of storage at maximum pressure, a total of 0.59 mg m-3 H2S(g) is
generated and released. During each of the subsequent phases (3 months of production, 3
60
months of storage at low pressure, and 3 months of injection), 0.3 mg m-3 H2S(g) is generated
and released. The cycle starts again with a high-pressure storage phase where only 0.3 mg m-3
H2S(g) is generated and released (Figure 3.3a). The high amount of H2S(g) (0.59 mg m-3)
generated and released in the first maximum pressure phase can be explained with the higher
amount of aqueous sulfate in the brine at the beginning of storage, which results from anhydrite
dissolution during leaching. The amount of aqueous sulfate available from leaching in the brine
decreases with ongoing BSR, but additional aqueous sulfate is delivered by diffusion from the
sump, where anhydrite is still available. When the amount of sulfate in the brine is consumed,
the sulfate is delivered only by diffusion. Therefore, the change in pressure conditions has
minor influence on H2S(g) generation and release, but diffusion of dissolved sulfate and
dissolved methane through the brine is the limiting factor. To confirm these results, additional
modeling scenarios were performed with an extended range of pressure conditions. Modeling
results show that even under very low-pressure (25 atm) and high-pressure conditions
(300 atm) the amount of generated and released H2S(g) is nearly constant at 0.3 mg m-3
(0.2999 mg m-3 at 25 atm and 0.30002 mg m-3 at 300 atm).
3.3.2.2 Natural Gas Composition
The chemical composition of the stored natural gas can differ. An alternative gas composition,
based on data from the DVGW (2013), with 175.032 atm ρCH4, 1.476 atm ρN2, 3.492 atm
ρCO2, and 0.000288 atm ρH2S is used to show the influence of the initial gas composition on
H2S(g) generation and release. The results of the 1D model (in PHREEQC) show a slight
increase in H2S(g) generation and release from 7.19 mg m-3 H2S(g) (reference scenario) to
7.22 mg m-3 H2S(g) using this alternative gas composition (Figure 3.3b). The pH decreases to
5.9 from 6.1 in the reference scenario. The amount of precipitated calcite increases slightly
(from 0.035 mol kgw−1 in the reference scenario to 0.037 mol kgw−1) but the amount of
precipitated mackinawite is nearly constant (0.004 mol kgw−1 in 30 years). Generally, the initial
gas composition can influence the final H2S(g) generation and release, but in this case, the
effect is minor.
3.3.2.3 Kinetic Rate Constants
The kinetic rate constant of BSR is an important factor controlling H2S(g) generation and
release. The rate constant is varied from an initial 6.46 × 10-10 mol kgw−1 s−1 (based on Adams
et al. (2013)) to 1.04 × 10-09 mol kgw−1 s−1 (based on Kallmeyer and Boetius (2004)) and
61
5.95 × 10-08 mol kgw−1 s−1 (based on Timmers et al. (2016)). All rate constants used apply to
temperature and pressure conditions comparable to salt cavern conditions. The presented
results are based on the 1D model in PHREEQC. As shown by Figure 3.3c, with increasing
rate constant the H2S(g) generation and release increases. The elevated rate constant of
1.04 × 10-09 mol kgw−1 s−1 causes an increase from 7.19 to 7.59 mg m-3 H2S(g) and the rate
constant of 5.95 × 10-08 mol kgw−1 s−1 an increase to 8.88 mg m-3 H2S(g) after 30 years. The
results show that by varying the rate constant, other interconnected reactions are affected as
well. Comparing the results of the reference scenario to the results of the scenario with the
rate constant of 5.95 × 10-08 mol kgw−1 s−1, the following changes are observable at the brine-
gas interface: the amount of the secondary phase calcite decreases slightly from
0.035 mol kgw−1 to 0.032 mol kgw−1 and mackinawite formation increases from
0.0038 mol kgw−1 to 0.0047 mol kgw−1 (whereas the pH is constant at 6.1). In consequence,
the variations in the amount of generated and released H2S(g) indicate that it is important to
determine the kinetic rate constants at elevated levels of temperature of 50 °C and other
relevant site-specific cavern conditions as accurately as possible. However, doubling the rate
constant does not double the amount of generated and released H2S(g) because diffusion is
the main limiting factor.
3.3.2.4 Tortuosity
Another influencing factor is the tortuosity, included in the 3D PHAST model. With increasing
tortuosity in brine and sump (from 0.001 to 0.01, 0.1 and 1.0), bacterial sulfate reduction
increases due to more effective diffusive transport of CH4(aq) and SO42-(aq). Therefore, the
maximum value of generated H2S(g) increases from 2.17 mg m-3 (tortuosity of 0.001) to
7.98 mg m-3 (tortuosity of 1.0; Figure 3.3d) after 30 years. The higher the tortuosity, the lower
the pH (from pH 9.5 with a tortuosity of 0.001 to pH 8.2 with a tortuosity of 1.0). Furthermore,
the amount of precipitated calcite increases with increasing tortuosity (from 0.02 mol kgw−1 to
0.05 mol kgw−1) and at higher tortuosity, intensified anhydrite dissolution is identified in the
sump. The lower BSR rate at low tortuosity influences the amount of precipitated mackinawite,
which is smaller at a tortuosity of 0.001 (0.002 mol kgw−1) than at tortuosity of 1.0
(0.003 mol kgw−1). The most significant difference in the amount of generated H2S(g) is
identifiable from a tortuosity of 0.001 to 0.01 (Figure 3.3d). The differences in H2S(g) generation
from the tortuosity of 0.01 to 0.1 and to 1.0 are minor. These results also indicate that diffusion
of dissolved methane and dissolved sulfate through the brine is the major limiting factor.
62
time (months) a)
3 6 9 12 15
0.8
0.5944
0.6
H 2S(g) (mg m -3)
0.4
0.3012 0.2999 0.3001 0.3002
0.2
0.0
180 120 60 120 180
storage at production storage at low injection storage at
high pressure pressure high pressure
Pressure (atm)
b) 8.88 c)
8.0 8.0 7.59
7.19 7.22 7.19
6.0
H2S(g) (mg m -3)
6.0
H2S(g) (mg m -3)
4.0 4.0
2.0 2.0
0.0 0.0
reference alternative 6.46e-10 1.04e-09 5.95e-08
Natural gas composition Kinetic rate constant (mol kgw -1 s-1)
d)
7.98
8.0
6.28 6.51
6.0
H2S(g) (mg m -3)
4.0
2.17
2.0
0.0
0.001 0.01 0.1 1
Tortuosity ( - )
Figure 3.3 H2S(g) generation influenced by a) pressure changes (after 3 months) b) stored gas
composition (reference = typical composition of natural gas from Russia, alternative = typical
composition of natural gas from the North Sea (after 30 years) c) kinetic rate constant (after 30 years)
d) tortuosity (after 30 years). H2S(g) in mg m-3 in the stored gas. Yellow = reference scenario, blue =
modified parameters. a), b), and c) are modeled 1D in PHREEQC, and d) is modeled 3D in PHAST.
63
3.3.3 Inhibition of H2S(g) Generation and Release in Salt Caverns
If the H2S(g) concentration in the stored gas exceeds the limit of 5 mg m-3 (DVGW, 2013), the
stored gas is polluted and cannot be used anymore. Therefore, additional modeling scenarios
show three possible methods to inhibit H2S(g) generation and release. Generally, salt caverns
should be constructed in rock salt formations with low amounts of anhydrite and other sulfate
sources (e.g., gypsum). However, with increasing demand for storage capacity in salt
structures, this is not always possible. In case the geogenic conditions favor H2S(g) generation,
as a first method a preventative measure should be applied. In this measure, Fe(+II) (aq)
(dissolved FeCl2) is added to the brine directly after leaching, before the first gas injection, and
inhibits the H2S(g) generation. The modeling results show a decrease in H2S(g) generation from
7.19 mg m-3 in the reference scenario to 1.52 mg m-3. This can be explained by the higher
amount of available Fe(+II), which reacts with aqueous sulfide to form mackinawite so that
aqueous sulfide is no longer available for H2S(g) generation (Figure 3.4a). The amount of
precipitated mackinawite increases from 0.004 mol kgw−1 (in the reference scenario) to
0.09 mol kgw−1 in the operating time of 30 years. The minimum concentration of FeCl2 that
must be added before the first gas injection to decrease the amount of generated H2S(g) below
the allowed limit of 5 mg m-3 is 0.025 mol kgw−1 (assuming a homogeneous distribution of the
injected FeCl2).
Two inhibition methods could be used when an increase in H2S(g) is detected in the stored gas.
One method is to add Fe(+II)(aq) (dissolved FeCl2) to the brine after first gas injection. The
modeling results show that the available aqueous sulfide reacts with aqueous ferrous iron, and
mackinawite precipitates. When adding 0.1 mol kgw−1 FeCl2 after 5 years of storage, a total of
only 3.47 mg m-3 H2S(g) is released after 30 years (Figure 3.4b). The pH decreases from 6.1 to
5.8, less calcite precipitates (from 0.035 to 0.008 mol kgw−1), and mackinawite precipitation
increases strongly from 0.004 to 0.05 mol kgw−1. The addition of 0.05 mol kgw−1 FeCl2 after
first gas injection is sufficient enough to reduce the H2S(g) concentration just below the allowed
limit of 5 mg m-3 (assuming a homogeneous distribution of the added FeCl2).
Another inhibition method is the addition of NaOH to the brine. Modeling results show that this
method increases the pH and inhibits H2S(g) generation. The pH influences the speciation type
of the generated sulfide according to the mass action laws (included in the database
phreeqc.dat). At neutral pH, most of the sulfide is available as H2S and HS-. In higher pH
environments, HS- and S2- predominate, and at low pH, hydrogen sulfide occurs as H2S.
Therefore, when increasing the pH by addition of NaOH, more S2- is generated (at constant
S2- in total), and gas pollution by H2S(g) is inhibited. When 1.0 mol kgw−1 NaOH is added after
5 years of storage, the total amount of generated H2S(g) after 30 years decreases to
64
0.74 mg m-3 (Figure 3.4c). The pH increases from 6.1 (in the reference scenario) to 11.5, calcite
precipitation increases from 0.035 to 0.09 mol kgw−1, and less mackinawite precipitates (from
0.004 to 0.002 mol kgw−1). Another possibility is the addition of 0.1 mol kgw−1 NaOH after 5
years of storage and repetition of the same input after 10 years of storage. With these
conditions, the total amount of generated H2S(g) decreases to 3.86 mg m-3 after 30 years
(Figure 3.4c). By addition of only 0.13 mol kgw−1 NaOH after 5 years of storage the H2S(g)
concentration is reduced to 4.99 mg m-3. This amount is the minimum amount of NaOH that is
needed to reduce the H2S(g) concentration below the allowed concentration of 5 mg m-3 under
assumed storage conditions and homogeneous distribution of added NaOH.
To ensure constant gas quality over time, the gas and brine must be analyzed continuously
and the inhibition methods must be applied when the H2S(g) concentration increases.
Therefore, H2S(g) measurements and monitoring in the gas as well as in the brine could be
used as an early warning system. The application of sulfide scavengers is widely adopted in
the oil and gas industry. Amosa et al. (2010) summarized the different scavenger technologies
from copper-base scavengers to zinc-base scavengers and up to iron-base scavengers
(among others). Less attention has been paid on the application of scavengers in underground
gas storage systems especially in salt cavern gas storages. Some authors suggest to decrease
the bacterial activity in the underground gas storage in porous media by “simple substitution”
of products like heptamethyl nonane, dioxin, furan (Kleinitz and Böhling, 2005). They focus on
procedures which prevent bacterial growth. However, the here presented study focuses on
hydrogeochemical methods to decrease the generation and release of H2S(g) in salt cavern gas
storage systems and on the quantitative description of the amount of FeCl2 or NaOH that must
be added to decrease the H2S(g) concentration in the stored gas below the allowed limit of
5 mg m-3.
65
a)
8.0 7.19
6.0
H2S(g) (mg m -3)
5.0
4.0
2.0 1.52
0.0
reference 0.1
FeCl2 (mol kgw-1) before first gas injection

b)
8.0 7.19
6.0
H2S(g) (mg m -3)
5.0
4.0 3.47
2.0
0.0
reference 0.1
FeCl2 (mol kgw -1) after first gas injection
c)
8.0 7.19
6.0
H2S(g) (mg m -3)
5.0
3.86
4.0
2.0
0.74
0.0
reference 1.0 2 x 0.1
NaOH (mol kgw-1) after first gas injection
Figure 3.4 H2S(g), in mg m-3, in the stored gas after 30 years. Inhibition of H2S(g) generation and release
by addition of a) FeCl2 before first gas injection, b) FeCl2 after gas injection c) 1.0 mol kgw−1 NaOH after
5 years of storage, and 0.1 mol kgw−1 NaOH after 5 years and after 10 years of storage to the brine.
Reference is without any inhibition methods. Yellow = reference scenario, green = inhibition factors. Red
line = maximum allowed H2S(g) concentration in stored gas.
66
3.4 Conclusion
This study indicates that H2S(g) generation and related pollution of the stored gas are risks in
salt caverns. With increasing demand for storage capacity in salt caverns, the use of less
favorable salt formations will increase, and the potential risk of H2S(g) generation and release
will rise. H2S(g) in salt caverns is generated by bacterial sulfate reduction. The amount of
available sulfate in the rock salt formation and diffusional transport are the main limiting factors
of this process. The kinetic rate constant for bacterial sulfate reduction influences the amount
of generated and released H2S(g). Therefore, determining the kinetic rate constants at elevated
levels of temperature and other relevant cavern conditions as accurately as possible is
required. Experimental procedures to determine kinetic rate constants under consideration of
site-specific conditions are presented e.g. by Adams et al. (2013), Jakobsen and Postma
(1994), Timmers et al. (2016), Kallmeyer and Boetius (2004). Changing pressure conditions
during injection, storage, and production phases has a minor influence on generated H2S(g).
Furthermore, the composition of the stored natural gas can influence H2S(g) generation and
release. The tortuosity effects the amount of generated H2S(g), but the diffusion of dissolved
methane and dissolved sulfate through the brine dominates this factor. In case of a site-specific
model, the tortuosity can be roughly estimated by empirical equations based on petrophysical
rock properties. Examples are given by Attia (2005), Shen and Chen (2007), Ewing et al.
(2010), Ghassemi and Pak (2011).
Different hydrogeochemical methods inhibit H2S(g) generation and release in salt caverns. One
method is a prevention measure that could be applied before first gas injection if geogenic
conditions favor H2S(g) generation. In that method, FeCl2 is added to the brine directly after
leaching. Two further technical methods should be used for the inhibition of H2S(g) generation,
if during the time of storage an increase of H2S(g) is observed in the gas or brine. The addition
of aqueous FeCl2 to the brine after gas injection leads to a reduction of H2S(g) generation
because the available aqueous sulfide reacts with the aqueous ferrous iron and mackinawite
precipitates. Alternatively, NaOH could be added to the brine. This method increases the pH
and inhibits H2S(g) generation because the pH influences the speciation type of generated
sulfide.
67
This research did not receive any specific grant from funding agencies in the public,
commercial, or not-for-profit sectors.
Input file S3 for the transport model to calculate the reference scenario SCGS is available (for
PHREEQC Version 3).
68
Underground Hydrogen Storage (UHS)
4 Underground Hydrogen Storage (UHS)
Hydrogeochemical Modeling to Identify Potential Risks of Underground

Hydrogen Storage in Depleted Gas Fields
Publication III
Originally published in: Applied Sciences 8(11) (2018), 2282



1
69
Summary
In this part of the study, the gas–water–rock interactions triggered by hydrogen storage in
depleted hydrocarbon reservoirs are considered. The focus is on the loss of hydrogen by
bacterial conversion over storage times at reservoir scale. Furthermore, mineral dissolution
and precipitation due to hydrogen storage are qualitatively and quantitatively described and
the loss of hydrogen by diffusion along a gradient of decreasing pressure and temperature
conditions is analyzed.
The storage of hydrogen in the underground causes hydrogeochemical gas–water–rock

interactions. In view of these processes, this study is performed based on the principal laws of
chemical-mass conversion and diffusive mass transport through the cap rock. The aim is to
obtain an understanding of the spatial- and temporal-dependent hydrogeochemical processes
induced by hydrogen storage and to analyze related risks.
Hydrogen is injected into a depleted gas reservoir and dissolves in the brine until saturation is
established and continuously serves as a reductant for BSR and methanogenesis. At the
interface with the cap rock, the aqueous hydrogen diffuses through the cap rock brine and
triggers H2–water–rock interactions. These processes are retraced using one-dimensional
(1D) reactive mass transport models and the modeling tool PHREEQC (phreeqc.dat
database). Up to 12 aqueous species and 22 different mineral and gas phases are used to
form a semi-generic but well-defined hydrogeochemical modeling approach.
Modeling results show that the total loss of hydrogen gas by aqueous diffusion, gas–water–
rock interactions, and bacterial conversion is 76% over 30 years under modeling conditions.
Therefore, the loss by aqueous diffusion is minor. Underground hydrogen storage induces a
slight decrease of the reservoir rock porosity. A large part of the stored gas is converted by
bacterial activity (bacterial sulfate reduction and methanogenesis), limited by the amount of
available electron acceptors (sulfate and carbon dioxide), the kinetic rate constants, and
maximum storage time.
1D mass transport modeling scenarios help to identify the factors dominating the loss of
hydrogen by varying single parameters of the reference model. The modeling approach
considers storage times of 30–300 years. Gas–water–rock interactions that may be triggered
by the withdrawal or production of hydrogen are considered, as well as different gas
compositions and kinetic rate constants. From the modeling results, it seems reasonable to
choose depleted gas fields with low temperature and pressure conditions as well as low
amounts of carbonate- and sulfate-bearing minerals for hydrogen storage. The
hydrogeochemical model approach can be used and adapted to various underground
hydrogen storage conditions to answer and solve practically oriented questions and problems.
70
Abstract
Underground hydrogen storage is a potential way to balance seasonal fluctuations in energy

production from renewable energies. The risks of hydrogen storage in depleted gas fields
include the conversion of hydrogen to CH4(g) and H2S(g) due to microbial activity, gas–water–
rock interactions in the reservoir and cap rock, which are connected with porosity changes,
and the loss of aqueous hydrogen by diffusion through the cap rock brine. These risks lead to
loss of hydrogen and thus to a loss of energy. A hydrogeochemical modeling approach is
developed to analyze these risks and to understand the basic hydrogeochemical mechanisms
of hydrogen storage over storage times at the reservoir scale. The one-dimensional diffusive
mass transport model is based on equilibrium reactions for gas–water–rock interactions and
kinetic reactions for sulfate reduction and methanogenesis. The modeling code is PHREEQC
(pH-REdox-EQuilibrium written in the C programming language). The parameters that
influence the hydrogen loss are identified. Crucial parameters are the amount of available
electron acceptors, the storage time, and the kinetic rate constants. Hydrogen storage causes
a slight decrease in porosity of the reservoir rock. Loss of aqueous hydrogen by diffusion is
minimal. A wide range of conditions for optimized hydrogen storage in depleted gas fields is
identified.
4.1 Introduction and Aims
Underground hydrogen storage (UHS) is used to store large amounts of hydrogen generated
from renewable energy sources (such as wind and solar) to compensate for seasonal
fluctuations in the supply and demand of energy (Ebigbo et al., 2013; Reitenbach et al., 2015).
Large amounts of hydrogen can be stored in depleted oil and gas fields, in salt caverns, and
in aquifers (Ebigbo et al., 2013; Stone et al., 2009). Nevertheless, practical experience of
underground hydrogen storage is still rare. In the US and UK, hydrogen is currently stored in
salt caverns (Crotogino et al., 2010; Henkel et al., 2013; Panfilov, 2010; Stone et al., 2009),
but hydrogen storage in depleted oil and gas fields is still under research and discussion.
Depleted oil and gas fields have a huge storage capacity, are well known from former
exploration and production, and qualify therefore for hydrogen storage. However, the existing
underground gas storages (UGS) are designed for the storage of natural gas, which does not
contain hydrogen (or only very low amounts). Because the chemical and physical properties
of hydrogen are different to those of methane (CH4)—the main component of natural gas—the
effects of hydrogen on the reservoir rock, cap rock, and storage facilities must be analyzed
before injecting hydrogen into these storages (Reitenbach et al., 2015).
71
Compared to methane, hydrogen has a higher diffusivity in pure water. The diffusion coefficient
for hydrogen is 5.13 × 10−9 m2 s−1 and for CH4 it is 1.85 × 10−9 m2 s−1, both in pure water at
25 °C. The data are given in the database phreeqc.dat, that is provided by the US Geological
Survey. Generally, the diffusion process of an aqueous species is described by its diffusion
coefficient in pure water (m2 s−1). However, in any porous system, where solids are available,
the influence of porosity on the diffusion must be considered by scaling the diffusion coefficient
with tortuosity (Shen and Chen, 2007), resulting in an effective diffusion coefficient (Equation
(4.1)), where d′m is the effective diffusion coefficient, dm is the diffusion coefficient in pure water,
and 𝜏 is the tortuosity. The tortuosity is the ratio of the effective diffusion mass fluxes in a
porous medium to ideal diffusion mass fluxes in solutions without a rock matrix.
′
𝑑𝑚
𝑑𝑚 = (4.1)
𝜏2
Therefore, the effective diffusion coefficient for methane is 2.35−2.49 × 10−10 m2 s−1 in clayey
host rocks at 21 °C (Jacops et al., 2013). The effective diffusion coefficient for hydrogen is
3.0 × 10−11 m2 s−1 in clayey host rocks saturated with water at 25 °C (Krooss, 2008). However,
there is a lack of data regarding effective diffusion coefficients of hydrogen at underground
storage conditions with higher temperatures and pressures (Panfilov, 2016).
Compared to methane, hydrogen has a lower viscosity. The viscosity of hydrogen at the two-
phase boundary at 17 MPa and 350 K is 1.01 × 10−5 kg m−1 s−1 (McCarty et al., 1981) and
methane fluid at 20 MPa and 350 K is 1.81 × 10−5 kg m–1 s–1 (Friend et al., 1989). Furthermore,
hydrogen has a lower density than methane. At normal conditions, the density of hydrogen is
eight times smaller than that of methane and the difference increases by 20–30% under
storage conditions (Reitenbach et al., 2015). The solubility of H2(g) in pure water is smaller than
the solubility of CH4(g). At 25 °C and 1.0 atm, 7.9 × 10−4 mol kgw−1 H2(g) and
1.4 × 10−3 mol kgw−1 CH4(g) dissolve in pure water (data based on phreeqc.dat).
The properties of hydrogen—being different from those of methane—could lead to risks when
storing hydrogen in depleted gas fields, which were constructed for storing methane
(underground gas storage). Potential risks include (i) the conversion of hydrogen to CH 4 and
H2S due to microbial activity, (ii) chemical reaction of hydrogen with the minerals of the
reservoir rock/cap rock and thus potential resulting porosity changes, and (iii) the loss of
aqueous H2(aq) by diffusion through the cap rock.
72
The first risk (i) arises from microbial activities in the underground. The two processes relevant
here are bacterial sulfate reduction (Equation (4.2)) and methanogenesis (Equation (4.3)),
where hydrogen is microbially catalyzed by bacteria and is converted to H2S(aq) or CH4(aq). The
activity of the different bacteria depends on the availability of the different electron acceptors
such as sulfate or carbon dioxide (Ebigbo et al., 2013). Other possible microbial processes are
acetogenesis and iron(III)-reduction, summarized by Hagemann et al. (2016).
Bacterial sulfate reduction (BSR): SO42−(aq) + 5H2(aq) → H2S(aq) + 4H2O (4.2)
Methanogenesis: CO2(aq) + 4H2(aq) → CH4(aq) + 2H2O (4.3)
In aqueous anoxic environments, the sulfate-reducing bacteria (SRB) use sulfate as an

electron acceptor to oxidize hydrogen and generate sulfide-S which could be available as
aqueous S2−(aq), HS−(aq), and H2S(aq) and as gaseous H2S(g) as well. The sulfate is derived from
the aqueous dissolution of mineral phases like anhydrite (CaSO4(s)). The energy gained from
sulfate reduction is used by the SRB for cell growth (Cord-Ruwisch et al., 1987). SRB prefer
temperatures around 38 °C (Bernardez et al., 2013) and near-neutral pH conditions (Cord-
Ruwisch et al., 1987), but are active even at extreme habitat conditions such as at great depths
and temperatures, ranging from 0 °C to 60–80 °C (Ehrlich, 1990; Machel, 2001; Postgate,
1984) and in some cases up to 110 °C (Jorgensen et al., 1992). SRB activity was observed,
for example, in hydrocarbon reservoirs (Machel, 2001) and in the underground storage of town
gas (Postgate, 1979). The risk arising from bacterial sulfate reduction lies in the produced H2S,
which is corrosive towards the storage facilities, toxic if inhaled, and can pose a risk to the
environment.
Typical environments for methanogenic bacteria are, for example, anoxic sediments and
flooded soils (Whitman et al., 2006). However, the existence of methanogenic bacteria was
also observed in town gas storages (Šmigáň et al., 1990). Methanogenic bacteria prefer
temperatures of 30–40 °C for growth (Panfilov, 2016), but they also have been found at higher
temperatures of 80 °C (Davydova-Charakhch`yan et al., 1992; Magot et al., 2000) and up to
97 °C (Gusev and Mineeva, 1992; Panfilov, 2016). With these bacteria, H2(aq) is the electron
donor and CO2(aq) is the electron acceptor, which is reduced to form CH4(aq). The methanogenic
bacteria obtain their energy for cell growth from this conversion (Šmigáň et al., 1990; Whitman
et al., 2006). The problem associated with methanogenesis lies in the loss of hydrogen and
the related energy loss (Reitenbach et al., 2015). This phenomenon has already been
observed in the underground storage of town gas. A well-known example is the Czech town
gas storage at Lobodice, where the 54 vol. % injected H2(g) diminished to 37 vol. % H2(g).
73
Concurrently, CH4(g) increased from 21.9 vol. % to 40.0 vol. % within a time span of 7 months
(Šmigáň et al., 1990).
Another risk of underground hydrogen storage (ii) is the reaction of H2(g/aq) with the minerals of
the reservoir rock and the cap rock, which can lead to mineral dissolution and precipitation and
resulting porosity changes as known from carbon capture and storage (e.g. Bildstein et al.,
2010; Gaus et al., 2005; Hemme and van Berk, 2017a; Mohd Amin et al., 2014). Changes in
the porosity of the cap rock can either improve or deteriorate its sealing capacity. Furthermore,
precipitation of minerals at the well equipment may cause scaling (Reitenbach et al., 2015).
In depleted gas fields, the high diffusivity of hydrogen could be the reason for hydrogen loss
(iii). The hydrogen is dissolved in the formation water of the cap rock and diffuses through the
cap rock (Reitenbach et al., 2015). Reitenbach et al. (2015) stated that a significant loss of
hydrogen can be expected. The high diffusivity, low viscosity, and low density of hydrogen
leads to a high mobility and therefore the potential loss due to leakage should be considered
(Ebigbo et al., 2013).
A natural analog for an unintended and unpredictable leakage of hydrogen can be found in
hydrogen anomalies associated with faults. These so called natural hydrogen seeps have been
described by Larin et al. (2015), Sato et al. (1986), Wakita et al. (1980), Ware et al. (1985),
and Zgonnik et al. (2015). In these cases, faults act as “fluid conduits” (Larin et al., 2015). An
increased CH4 concentration is also observed within these anomalies because of the increased
microbial activity stimulated by the increased amount of hydrogen (Larin et al., 2015). If
hydrogen reaches the surface, it could inhibit the growth of trees, underbrush, and grass (Larin
et al., 2015).
There are experimental studies concerning kinetics of pyrite and pyrrhotite reduction by
hydrogen at high temperatures (Truche et al., 2010), the reactivity of hydrogen in sandstones
(Yekta et al., 2018), and the petrographic and petrophysical variation in reservoir sandstones
due to hydrogen storage (Flesch et al., 2018). However, modeling studies of hydrogen storage
to predict “long-term” behaviors are still rare. Therefore, this study is performed to model the
loss of hydrogen (and related energy loss) as a combination of (i) the conversion by bacteria,
(ii) gas–water–rock interactions in the reservoir and cap rock, which are connected with
porosity changes, and (iii) loss by aqueous diffusion along a gradient of decreasing pressure
and temperature conditions.
These processes are retraced by a one-dimensional reactive mass transport model to simulate
gas–water–rock interactions resulting from hydrogen storage in depleted gas fields. The aim
of this study is to investigate the basic mechanisms of BSR and methanogenesis in an
74
integrated way over storage times at the reservoir scale and to describe qualitatively and
quantitatively which reservoir rock and cap rock minerals dissolve or precipitate because of
hydrogen storage as well as the related porosity changes. Furthermore, the parameters that
influence the loss of hydrogen are determined.
Therefore, a model is presented, in which the hydrogeochemical mechanisms of underground

hydrogen storage are simulated in a reference scenario (Sections 4.3.1–4.3.3). In further
scenarios, single parameters will be changed in the model to show their effects on the modeling
results and to identify the parameters that are most sensitive for underground hydrogen
storage (Section 4.3.4).
4.2 Methodology
The modeling program for the one-dimensional reactive mass transport (1DRMT) model is
PHREEQC version 3 (PHREEQC = pH-REdox-EQuilibrium written in the C programming
language), provided by the US Geological Survey (Parkhurst and Appelo, 2013). PHREEQC
has capabilities to simulate (i) speciation and saturation-index calculations, (ii) batch-reaction
and 1D transport calculations, and (iii) inverse modeling (Parkhurst and Appelo, 2013). The
used database is phreeqc.dat extended by dawsonite, nahcolite, CH4(g), H2S(g), N2(g), which are
taken from llnl.dat. The 1DRMT model considers equilibrium reactions for gas–water–rock
interactions, and kinetic reactions for sulfate reduction and methanogenesis. The equilibrium
calculations are based on the mass action law including all species used in this study (Al, Ba,
C, Ca, Cl, Fe, K, Mg, N, Na, S, Si) and their corresponding equilibrium constants. The
equilibrium phases, mass action equations, and equilibrium constants used in the model are
summarized in Table 4.1. For detailed information about PHREEQC, see Parkhurst and
Appelo (2013).
75
Table 4.1 Equilibrium phases, mass action equations, and equilibrium constants used in the model.
Data from phreeqc.dat, except for dawsonite, nahcolite, CH4(g), H2S(g), N2(g), which are from llnl.dat
(Parkhurst and Appelo, 2013).
Equilibrium log K
Equilibrium Reaction
Phase at 25 °C, 1 bar
K-feldspar KAlSi3O8 + 8H2O = K+ + Al(OH)4− + 3H4SiO4 −20.573
Albite NaAlSi3O8 + 8H2O = Na+ + Al(OH)4− + 3H4SiO4 −18.002
Kaolinite Al2Si2O5(OH)4 + 6H+ = H2O + 2H4SiO4 + 2Al3+ 7.435
Quartz SiO2 + 2H2O = H4SiO4 3.98
Calcite CaCO3 = CO32− + Ca2+ 8.48
Pyrite FeS2 + 2H+ + 2e− = Fe2+ + 2HS− −18.479
K0.6Mg0.25Al2.3Si3.5O10(OH)2 + 11.2H2O = 0.6K+ +
Illite −40.267
0.25Mg2+ + 2.3Al(OH)4− + 3.5H4SiO4 + 1.2H+
Dawsonite NaAlCO3(OH)2 + 3H+ = Al3+ + HCO3− + Na+ + 2H2O 4.35
Mackinawite FeS + H+ = Fe2+ + HS− −4.648
Dolomite CaMg(CO3)2 = Ca2+ + Mg2+ + 2CO32− −17.09
Nahcolite NaHCO3 = HCO3− + Na+ −0.11
Anhydrite CaSO4 = Ca2+ + SO42− −4.39
Halite NaCl = Cl− + Na+ 1.570
Gypsum CaSO4·2H2O = Ca2+ + SO42− + 2H2O −4.58
Sulfura S + 2H+ + 2e− = H2S 4.882
Barite BaSO4 = Ba2+ + SO42− −9.97
Goethite FeOOH + 3H+ = Fe3+ + 2H2O −1.0
H2(g) H2 = H2 −3.1050
CO2(g) CO2 = CO2 −1.468
CH4(g) CH4 = CH4 −2.8502
H2S(g) H2S = H+ + HS− −7.9759
N2(g) N2 = N2 −3.1864
a: Sulfur = elemental sulfur
4.2.2 Model Setup
Gaseous hydrogen (H2(g)) is injected into the reservoir rock where residual gas from the
previous natural gas reservoir is still available and the reservoir brine is in equilibrium with
these gases and the reservoir rock. When H2(g) is injected, the available brine is saturated with
H2. With ongoing injection of H2(g), the initial reservoir brine is displaced and H2(g) builds up a
plume. The only available water in the reservoir rock is then the irreducible water. On contact
with the cap rock, H2(g) dissolves into the cap rock brine and diffuses through the cap rock
brine. These processes are retraced with a hydrogeochemical, 1D reactive mass transport
modeling approach, which is of semi-generic nature. The conditions assumed in the model are
based on conditions of a depleted gas field with a temperature of 40 °C and a reservoir
pressure of 40 atm. The model is divided into a column of 1488 cells, with a cell height of 1.0 m
each (in the z-direction). The cap rock is assumed to have a height of 182 m (cells 1–182), the
reservoir rock consists of 667 m (cells 183–850), and the underlying rock of 637 m (cells 851–
1488; Figure 4.1). For the sake of simplicity, a constant partial pressure in the H2(g) plume is
76
assumed and the partial pressure is set to be equal to the hydrostatic pressure. Even during
production, an amount of hydrogen remains in the reservoir.
(cells/meters
H2O
Cap rock
1-182)
CaSO4(s) ↔ SO42- + Ca2+
H2O CaCO3(s) ↔ Ca2+ + CO32-

H2(g) ↔ 2H+(aq) + 2e- CO32- + 2H+(aq) ↔ H2CO3
CaSO4(s) ↔ SO42- + Ca2+
(cells/meters 183-850)
CO2(g) ↔ CO2(aq) H2CO3 ↔ CO2(aq) + 4H2O

Reservoir rock
SO42-(aq) + 5H2(aq) → H2S(aq) + 4H2O

H2S(aq) ↔ H2S(g) CO2(aq) + 4H2(aq) → CH4(aq) + 2H2O
CH4(aq) ↔ CH4(g)
CO2(aq) + H2O ↔ H2CO3 FeOOH(s) + 3H+(aq) ↔ Fe3+ + 2H2O

H2CO3 ↔ CO32- + 2H+ Fe3+ + e- ↔ Fe2+ CH4(aq) + 2H2O ↔ CO2(aq) + 8H+ + 8e-
Ca + CO3 ↔ CaCO3(s)
2+ 2- Fe2+ + 2HS- ↔ FeS2(s) + 2H+ + 2e-
Underlying rock
(cells/meters
850-1488)
H2O
CaSO4(s) ↔ SO42- + Ca2+ CH4(g) ↔ CH4(aq)
Figure 4.1 Selected reactions and processes associated with bacterial sulfate reduction and
methanogenesis (purple). Single arrow = kinetic-controlled reactions; double arrow = equilibrium
reactions; blue triangles = time-dependent diffusive transport of aqueous components; bold = injected
gas for storage.
The mineralogical compositions of the cap rock, the reservoir rock, and the underlying rock are
based on data of the Röt Formation (Dersch-Hansmann et al., 2010; Feist-Burkhardt et al.,
2008), Buntsandstein Formation (Soyk, 2015), and Zechstein Formation (Biehl et al., 2016;
Schreiber and Bąbel, 2007) and are summarized in Table 4.2. The reactive amount of each
mineral phase is calculated in moles per kg of pore water (mol kgw−1) with the consideration
of the specific density of each mineral phase (g cm−3). Dawsonite, mackinawite, elemental
sulfur, albite (and pyrite) are considered as potential secondary phases, which may form at
saturation. The clay minerals are considered as follows: illite is defined as a primary mineral
phase, and the cation exchange capacity of chlorite, illite, montmorillonite, and kaolinite are
estimated based on the data given by Appelo and Postma (1999). An initial small amount of
CO2 (pCO2 = partial pressure of CO2 = 1.0 atm) is assumed to be available in the cap rock,
reservoir rock, and underlying rock from burial. The possible outgassing of CH4(g), N2(g), CO2(g),
H2S(g), and H2(g) is assumed in all cells.
The reservoir rock (cells 183–850) has an initial porosity of 10% (n = 0.1), 2.5% (n = 0.025) of
the pore space is filled with irreducible water, which is equal to 1 L H2O, and 7.5% (n = 0.075)
of the pore space is filled with gas. The total volume is 40 L (1 L/0.025) and therefore, 3 L are
77
filled with gas (40 L × 0.075). The porosity is filled with 3 L gas and 1 L H2O, so 36 L are
occupied by solids in each cell of the reservoir rock. The gas consists of 10% residual gas (that
was not extracted during natural gas production) and 90% stored hydrogen gas. The
composition of the residual gas in the reservoir rock is based on data from Bischoff and Gocht
(1984) with pCH4(g) = 3.56 atm, pN2(g) = 0.4 atm, and pCO2(g) = 0.04 atm. The stored hydrogen
gas is composed of 96% H2(g) (pH2(g) = 34.56 atm) and 4% CO2(g) (pCO2(g) = 1.44 atm), these
values are modified from data presented by Panfilov (2010). For modeling purposes, the stored
hydrogen gas is additionally induced as a tracer gas with the same chemical properties as
H2(g). It is used instead of H2(g) because H2(g/aq) is induced and used as the reactant in the
calculation kinetics. The amount of tracer gas corresponds to 4.3 mol L−1 of gas in each cell of
the reservoir. The summed volume of all gases in the reservoir rock is 3 L per cell and
represents the stored gas volume at 40 atm under the assumed reservoir conditions. The sum
of the partial pressure of all available gases (including tracer gas) is equivalent to the total
pressure (40 atm) in the reservoir so that the generated gases (H2S(g), CH4(g)) are released as
gas bubbles.
Table 4.2 Mineralogical composition of the reservoir rock, cap rock, and underlying rock. Dawsonite,
nahcolite, mackinawite, sulfur, albite (and pyrite) are potential secondary phases.
Primary Minerals Weight Percent (wt. %) Amount (mol kgw−1)

Cap rock
Halite 5.0 76.74
Quartz 50.0 746.42
Illite 20.0 46.73
Dolomite 5.0 24.32
Anhydrite 15.0 131.77
Reservoir rock
Kaolinite 1.0 3.73
Quartz 55.0 882.43
Calcite 0.5 4.82
Dolomite 0.5 0.03
Anhydrite 0.5 0.132a
Illite 11.5 28.88
Barite 0.5 0.0009
Goethite 0.5 0.002
Underlying rock
Halite 50.0 758.46
Quartz 8.0 118.04
Calcite 6.0 53.14
Dolomite 10.0 0.03
Pyrite 1.0 7.39
Anhydrite 25.0 66.11
a: Reactive amount of anhydrite
78
The cap rock (cells 1–182) is defined by an initial porosity of 5.0% (n = 0.05), with 1 L
irreducible water and 19 L solid. The underlying rock (cells 851–1488) is defined by an initial
porosity of 10% (n = 0.1), divided into 2.5% (n = 0.025) irreducible water, which is equal to 1 L
H2O, and 7.5% (n = 0.075) gas. The porosity is filled with 3 L gas and 1 L H2O, so 36 L are
occupied by solids in each cell of the underlying rock. The assumed natural gas composition
in the underlying rock is pCH4(g) = 81.1 atm, pN2(g) = 4.3 atm, pH2S(g) = 10.7 atm, and pCO2(g)
= 10.7 atm (Bischoff and Gocht, 1984).
The initial brine compositions of the cap rock, reservoir rock, and underlying rock are pre-
calculated in a separate transport model to simulate the million-years-long interaction of the
different brines with each other. The cap rock brine is composed of a 0.5 M Na+/Cl−-dominated
solution equilibrated with the mineral phases of the cap rock under cap rock pressure and
temperature conditions. The temperature and pressure conditions change along the diffusive
pathway of the aqueous species through the cap rock, starting with 40 °C and 40 atm at the
reservoir depth (data from Pudlo et al. (2010)). A gradient of decreasing temperature and
pressure conditions according to the geothermal gradient (33.3 °C km−1 depth) and a pressure
gradient of 100 atm km−1 depth under hydrostatic conditions is assumed. Therefore, each cell
is defined by a specific temperature and pressure condition (e.g., cell 182: 39.967 °C and
39.9 atm; cell 181: 39.934 °C and 39.8 atm; and so on). The initial reservoir rock brine
composition is a 0.5 M Na+/Cl−-dominated solution, which is in equilibrium with the reservoir
rock and the initial gas composition at 40 atm and 40 °C. The underlying rock is dominated by
a 6.7 M Na+/Cl−-dominated solution equilibrated with the underlying rock and gas composition
under underlying rock pressure and temperature conditions (106.7 atm and 60 °C). To
simulate the million-years-long interaction, molecular diffusion of all aqueous species through
the cap rock brine, reservoir rock brine, and underlying rock brine is simulated over one million
years. The brine compositions are summarized in Table 4.3.
The high Na+ and Cl− concentrations and high ionic strength in the brines require the use of
the Pitzer database. However, the Pitzer database does not include Al3+, which is important
when modeling hydrogen storage in depleted gas fields. Therefore, the database phreeqc.dat
is used. The validation that PHREEQC using phreeqc.dat simulates correct results with high
Na+ and Cl− concentrations and high ionic strength is shown by Hemme and van Berk (2017a).
Furthermore, Parkhurst and Appelo (1999) stated that models are reliable at higher ionic
strength if the system is sodium chloride dominated.
79
Table 4.3 Composition of the initial irreducible water in the reservoir rock, the cap rock brine, and the
underlying rock brine.
Cap Rock Irreducible Water in the Underlying Rock
Parameter
Brine Reservoir Rock Brine
pH 6.4 6.4 5.9
Temperature (°C) 37.0 40.0 60.0
Elements Concentration Concentration Concentration
(mol kgw−1) (mol kgw−1) (mol kgw−1)
Al 1.31 × 10−7 2.209 × 10−8 1.776 × 10−8
Ba 5.85 × 10−7 3.922 × 10−7 2.206 × 10−5
Ctota 3.11 × 10−2 1.762 × 10−2 7.405 × 10−3
Ca 3.63 × 10−2 1.186 × 10−2 1.562 × 10−2
Cl 1.27 1.123 5.396
Fe 5.69 × 10−2 6.572 × 10−2 4.263 × 10−11
K 5.28 × 10−1 6.151 × 10−1 4.604 × 10−1
Mg 8.73 × 10−4 2.579 × 10−3 1.142 × 10−2
N 5.61 × 10−2 6.485 × 10−2 5.081 × 10−2
Na 1.27 1.123 5.396
Stotb 1.01 1.143 7.540 × 10−1
Si 8.73 × 10−5 9.878 × 10−5 1.509 × 10−4
a: Summed concentration of aqueous CH4 and C(+IV) species
b: Summed concentration of aqueous S(+VI) and S(−II) species
Molecular diffusion of all aqueous species is the only mass transport accounted for by the
model. Multicomponent diffusion is calculated where each “solute can be given its own
diffusion coefficient, allowing it to diffuse at its own rate, but with the constraint that overall
charge balance is maintained” (Appelo and Wersin, 2007; Parkhurst and Appelo, 2013;
Vinograd and McBain, 1941). Fluid flow is not considered in the model due to the lack of total
hydraulic head differences in depleted gas reservoirs. A homogeneous distribution of the
relevant parameters in the reservoir and cap rock such as mineralogical composition is
assumed. Furthermore, no fractures or natural faults are considered. Because of the lack of
data, the storage time of hydrogen in depleted oil and gas fields is based on the equipment
lifetime of 30 years (Lord et al., 2014), but is varied in an additional scenario (Section 4.3.4.1).
The oxidation of hydrogen by SO42− and/or CO2 in the model is kinetically controlled. The
reaction kinetics describe the time-dependent changes of reaction products and educts of a
chemical reaction. To calculate the irreversible reactions, catalyzed by microorganisms like
sulfate-reducing bacteria and methanogenic bacteria, the Monod equation (Equation (4.4)) is
used. Bacterial concentrations are kept constant in the model with the aim to model the “worst-
case” scenario.
80
growth 𝑐s
𝜓 growth = 𝜓max ( ) (4.4)
𝛼 + 𝑐s
growth
where 𝜓max is the maximum specific growth rate (mol L−1 s−1), cs is the concentration of the
limiting substrates (mol L−1), and α is the half-velocity constant. The maximum specific growth
rate that is assumed in the model is 2.30 × 10−9 mol kgw−1 s–1 for methanogenesis (at 37 °C)
(Šmigáň et al., 1990) and 9.26 × 10−8 mol kgw−1 s−1 for BSR (at 30 °C) (Herrera et al., 1997).
For bacterial sulfate reduction, the limiting substrate is sulfate and for methanogenesis it is
carbon dioxide.
Panfilov (2010) defined a new model to describe the substrate-limited growth models (Equation
(4.5)), where te is the “individual time of eating” (s) and nmax is the maximum population size
(m−3). Due to the lack of data regarding the characteristic time of eating and the maximum
population size, the Panfilov model is not considered.
1 𝑛 𝑐𝑠
𝜓 𝑔𝑟𝑜𝑤𝑡ℎ = ( ) (4.5)
𝑡𝑒 𝑛2 𝛼 + 𝑐𝑠
1+ 2
𝑛𝑚𝑎𝑥
The lower boundary of the column (in the underlying rock) and the upper boundary of the
column (in the cap rock) are defined as diffusive flux. To check the numerical accuracy of the
results, discretization studies for one scenario are performed, in which the number of shifts
and the number of cells are refined. The model calculation for one scenario is rerun and results
are compared (Parkhurst and Appelo, 2013).
4.3.1 Loss of H2(g/aq) by Bacterial Conversion to CH4(g) and H2S(g)
The modeling results of the reference scenario (defined in Section 4.2.2; input file S4) show
that the maximum amount of bacterial-converted H2(g/aq) to CH4(g) and H2S(g) depends on the
amount of available (and reactive) electron acceptors, carbon dioxide and sulfate, and on the
amount of available H2(g/aq). Figure 4.1 shows selected reactions and processes that are
coupled to bacterial sulfate reduction and methanogenesis.
When CO2 and SO42− are fully consumed, methanogenesis and BSR will not proceed. The
stored gas is composed of 96% hydrogen and 4% carbon dioxide (residual gas is still
available). Therefore, CO2 is available as electron acceptor for the methanogenesis and CH4(g)
81
is generated. After less than 3 years of storage, the injected CO 2(g) is completely consumed,
but an ongoing supply of CO2 is provided by the dissolution of carbonate-bearing minerals
(such as calcite, for example). Consequently, the amount of generated CH4(g) is limited by the
assumed storage time of 30 years and the time-dependent kinetics of methanogenesis.
The available sulfate for BSR in the reservoir rock results from anhydrite dissolution. If
anhydrite is still present in a depleted natural gas reservoir, it is not reactive (e.g., it is coated
by other minerals). Otherwise, the sulfate would have been catalyzed by bacterial sulfate
reduction, with methane—the main component of natural gas—as electron donor. The same
applies to hydrogen storage in depleted natural gas reservoirs. However, the creation of new
fractures and natural faults or the dissolution of carbonate cements (pore-filling cements) leads
to contact of reactive anhydrite with the stored hydrogen. To simulate this “worst-case”
scenario, a small amount of reactive anhydrite is assumed in this model. The dissolution of
anhydrite provides SO42−(aq) for H2S(g) generation and Ca2+(aq) for calcite precipitation.
Additionally, the dissolution of anhydrite creates porosity, which in turn can be superseded by
the precipitation of other minerals (see Section 4.3.2). The reactive anhydrite is in solution
equilibrium with the brine. The consumption of sulfate(aq) by BSR is accompanied by ongoing
anhydrite dissolution and more sulfate becomes available for BSR until the reactive anhydrite
is completely dissolved.
As a product of BSR and methanogenesis, small amounts of water (0.007 kg) are formed
(Equations (4.2) and (4.3)). This increases the water content in the reservoir rock (1.0 kg initial
water + 0.007 kg generated water = 1.007 kg total water mass per cell after 30 years). The
water will probably be suppressed by the gas along the reservoir/cap rock boundary. However,
the small increase in the mass of water could increase the anhydrite dissolution, resulting in a
higher amount of sulfate ions that are available for BSR. Additional sulfate is delivered by
diffusion from the cap rock and the underlying rock, where the dissolution of anhydrite acts as
a sulfate source. The amount of generated H2S(g) depends on the storage time, the amount of
available and reactive anhydrite, the diffusion, and the time-dependent kinetic rate constant.
However, the last three mentioned factors depend on pressure and/or temperature.
Two hotspot regions with higher amounts of generated H2S(g) (related to a higher loss of
hydrogen) are identified: the contact area of the reservoir rock and the cap rock and the contact
area of the reservoir rock and the underlying rock. These are the regions where additional
sulfate is delivered into the reservoir rock by diffusion. However, the highest CH4(g)
concentrations can be found at the contact area of the reservoir rock and the underlying rock,
where additional CO2 is delivered by diffusion of the gas in the underlying rock into the reservoir
rock.
82
The 1DRMT modeling results show an average loss of 3.3 mol H2(g/aq) after 30 years in each
cell of the reservoir rock (cells 183–850). Over the same period, on average, 0.2 mol L−1 of gas
CH4(g) and 0.005 mol L−1 of gas H2S(g) are generated and 0.06 mol L−1 of gas CO2(g) is
consumed, which corresponds to the total amount of CO2(g) that was available as injected and
residual gas. The changes in gas composition in the reservoir rock in cell 183, located at the
contact with the cap rock, with ongoing time are shown in Figure 4.2a. The mass of lost H2(g/aq)
that is not converted to CH4(g) or H2S(g) is bound in aqueous species (OH−, CH4, HCO3−,
Al(OH)4−, CaOH+, CaHCO3+, MgOH+, MgHCO3+, H2, NH4+, NH3, H2S, HS−, H4SiO4, and
H3SiO4−) or in mineral phases (see Section 4.3.2).
83
Reference scenario a) 300 years b)

1.6 1.6
1.4 1.4
1.2 1.2
mol L-1 of gas

mol L-1 of gas
1 1
CH4(g)
CH4 CH4(g)
CH4
0.8 0.8
CO2(g)
CO2 CO2
CO 2(g)
0.6 H2(g)
H2 0.6 H2H
2(g)
0.4 H2S(g)
H2S 0.4 H2S
HS 2 (g)
0.2 0.2
0 0
0 5 10 15 20 25 30 0 50 100 150 200 250 300
time (years) time (years)
High pressure and temperature conditions (161 atm / 80 °C) c) Kinetic rate constant: BSR 2.40 10–7 mol kgw–1 s –1 d)
1.6 1.6
1.4 1.4
1.2 1.2
mol L-1 of gas
mol L-1 of gas

1 1
CH4(g)
CH4 CH4(g)
CH4
0.8 0.8
CO2(g)
CO2 CO2(g)
CO2
0.6 H2(g)
H2 0.6 H2(g)
H2
0.4 H2S(g)
H2S 0.4 H2S(g)
H2S
0.2 0.2
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Kinetic rate constant: BSR 9.26 10–9 mol kgw–1 s –1 e) Kinetic rate constant: Methanogenesis 2.30 10–6 mol kgw–1 s –1 f)
1.6 1.6
1.4 1.4
1.2 1.2
mol L-1 of gas
mol L-1 of gas
1 1
CH4(g)
CH4 CH4(g)
CH4
0.8 0.8
CO2(g)
CO2 CO2(g)
CO2
0.6 H2
H2(g) 0.6 H2(g)
H2
0.4 H2S
H2S(g) 0.4 H2S(g)
H2S
0.2 0.2
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Kinetic rate constant: Methanogenesis 2.30 10–12 mol kgw–1 s –1 g) Without CO2(g) h)
1.6 1.6
1.4 1.4
1.2 1.2
mol L-1 of gas
mol L-1 of gas
1 1
CH4(g)
CH4 CH4(g)
CH4
0.8 0.8
CO2(g)
CO2 CO2(g)
CO2
0.6 H2(g)
H2 0.6 H2H2(g)
0.4 H2S(g)
H2S 0.4 H2 S(g)
H2S
0.2 0.2
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Figure 4.2 Changes in gas composition in the reservoir rock with ongoing time in (a) reference scenario
and influenced by (b) increased storage time of 300 years, (c) higher pressure and temperature
conditions, (d) higher kinetic rate constant for BSR of 2.40 × 10-7 mol kgw–1 s-1, (e) lower kinetic rate
constant for BSR of 9.26 × 10-9 mol kgw−1 s-1, (f) higher kinetic rate constant for methanogenesis of
2.30 × 10−6 mol kgw−1 s-1, (g) lower kinetic rate constant for methanogenesis of
2.30 × 10−12 mol kgw−1 s-1, (h) varied stored gas composition (without CO2(g)).
84
The competition between methanogenic microorganisms and sulfate-reducing bacteria is

widely discussed in the literature (Adler et al., 2011; Kalyuzhnyi and Fedorovich, 1998;
Robinson and Tiedje, 1984; Timmers et al., 2015) and is not the object of this study. However,
in PHREEQC, BSR occurs prior to methanogenesis. This reaction order of H2(g/aq) with SO42−(aq)
(BSR) and CO2(aq) (methanogenesis) is tested in a separate batch model. It includes H2O(aq),
CO2(aq), and SO42−(aq) (in equal amounts) and a sufficient amount of hydrogen. H2(g/aq) reacts
primarily with SO42−(aq) to form S2−(aq) until the total amount of the initial SO42−(aq) is consumed.
This is followed by the reaction of H2(g/aq) with CO2(aq) to CH4(aq). Thereby, the equilibrium
constant (log K) is the crucial value. In a separate model, the log K value for SO42− (33.65;
Equation (4.6)) is reduced to below the log K value of CO2 (16.861; Equation (4.7); data from
phreeqc.dat). In this case, the methanogenesis takes place before BSR.
SO42−(aq) + 9H+(aq) + 8e– = HS−(aq) + 4H2O

(4.6)
CO32−(aq) + 2H+(aq) = CO2(aq) + H2O

(4.7)
4.3.2 Hydrogeochemical Effects of Hydrogen Storage on Reservoir Rock and Cap

Rock
Storage of hydrogen in depleted gas fields can lead to dissolution and precipitation of minerals,
which in turn could change the porosity of the reservoir rock and cap rock. An increase in the
cap rock porosity would decrease the sealing capacity and increase the risk of pathways for
the stored hydrogen to reach, for example, overlying aquifers. On the contrary, a decrease of
the cap rock porosity would increase its sealing capacity.
Truche et al. (2013) argued that the influence of hydrogen on sulfur-bearing minerals like
framboidal pyrite is high but the effect on other minerals available in claystones like clay
minerals, quartz and calcite is minor at “low temperatures”. By contrast, Reitenbach et al.
(2015) stated that the addition of hydrogen causes abiotic reactions of hydrogen with minerals
of the reservoir and cap rocks which leads to dissolution of sulfate and carbonate-bearing
minerals, clay minerals of the chlorite group and feldspars and to precipitation of iron-sulfide
bearing minerals, illite, and pyrrhotite. However, both studies agree that pyrite is a potential
oxidant for hydrogen and is reduced to pyrrhotite (FeS1 + x) (Reitenbach et al., 2015; Truche
et al., 2013). The 1DRMT modeling results of the reference scenario of this study show that at
40 °C and 40 atm the storage of hydrogen induces K-feldspar, kaolinite, and dolomite
precipitation in the reservoir rock. Quartz, calcite, and illite are dissolved. The reactive amounts
of barite, goethite, and anhydrite are completely consumed during the storage time of 30 years.
85
After the complete consumption of reactive anhydrite, the only sulfate source comes from the
cap rock and the underlying rock by diffusion and consequently this limits the loss of hydrogen
by bacterial sulfate reduction. High concentrations of S(−II)(aq) species in the reservoir brine
can be observed from the modeling results. The assumed reactive amount of goethite is
completely consumed in less than 3 years. The goethite dissolution buffers the effect of H2S(g)
generation because the amount of Fe(+II), which has been reduced from Fe(+III), reacts with
the aqueous sulfide to form pyrite (a potential secondary phase in the model) so that aqueous
sulfide is no longer available for H2S(g) generation. The modeling results show also that the
higher pH conditions (pH increases from initial 6.4 to 8.2–8.7) and the high sodium
concentrations in the brine induce weak albitization in the hotspot areas (the contact area of
the reservoir rock and the cap rock and the contact area of the reservoir rock and the
underlying rock). Except for pyrite and albite, the other potential secondary phases, dawsonite,
mackinawite, nahcolite, and sulfur, are not formed under reservoir conditions. Elemental sulfur
does not form, even if mackinawite and pyrite are not considered as potential secondary
phases. Small amounts of calcite are precipitated because the dissolution of anhydrite
contributes Ca2+(aq) into the reservoir brine. However, at the same time, calcite is dissolved and
delivers CO2 for methanogenesis so that the total amount of calcite decreases. The volume
changes of the mineral phases are shown in Figure 4.3. The changes in porosity are calculated
from the PHREEQC modeling results. The porosity in the reservoir rock decreases from initially
10% to 9.79–9.95%. This means that the available pore space for hydrogen storage decreases
over 30 years such that (i) the same amount of hydrogen is moved to greater distances from
the bore hole or (ii) less hydrogen can be stored in future injection phases (max. 0.2%).
The mineralogical changes in the cap rock, induced by diffusion of aqueous H2(aq) from the
reservoir rock, are minimal and cause no changes to the initial porosity. The reasons for that
are the short storage time of 30 years and the slow diffusion process. Furthermore, the
hydrogen reacts with the mineralogical assemblage of the reservoir rock and brine and it is
converted by BSR and methanogenesis.
The used modeling program, PHREEQC, has no capabilities to simulate gas transport
between the different cells (multiphase flow). To overcome this limitation, a separate transport
model provides an initial amount of H2(g) in the cap rock for the kinetic calculation to simulate
the influence of H2(g) on the cap rock properties in the case of a gas loss from the reservoir
rock by, for example, new fractures or natural faults. The initial amount corresponds to the
maximum available amount of H2(g) that could be available per cell in the cap rock and
represents a worst-case scenario. Modeling results indicate an increase in pH from 6.4 to 7.8
in the cap rock brine after 30 years of storage. The intense contact of hydrogen with the cap
86
rock results in a total loss in porosity, which increases the sealing capacity of the cap rock.
This decrease in porosity is caused mainly by albitization, whereas the dissolution of halite,
quartz, illite, and anhydrite counteracts a stronger porosity decrease. Precipitation of dolomite
is negligible. The potential secondary phase pyrite is precipitated whereas dawsonite,
mackinawite, and sulfur are not formed.
0.5
10-1
0.4
4.3
0.3
10 -2
(+)
0.2
10 -2
9.9
10-4
10-5
0.1
2.6
[L kgw-1]
7.7
5.7
10-5
10-5
10-3
10 -2
-0.1
-4.7
-4.4
-6.1
10-1
-5.2
-0.2
-1.1
10-1
(-)
-0.3
-3.7
-0.4
-0.5
Albite
Calcite
Quartz
Dolomite
Anhydrite
Illite
K-feldspar
Kaolinite
Barite
Goethite
Pyrite
Figure 4.3 Volume changes of mineral phases (L kgw−1) in the reservoir rock after 30 years of hydrogen
storage (reference scenario).
4.3.3 Loss of Aqueous H2(aq) by Diffusion through the Cap Rock
By assuming specific diffusion coefficients for the different aqueous species, the diffusion of
hydrogen and related products is modeled. A non-reactive tracer, with the same diffusion
coefficient as hydrogen of 5.13 × 10−9 m2 s−1, shows that significant amounts of hydrogen
(greater than 4.0 × 10−7 mol kgw−1) were calculated only for distances less than 4 m through
87
the cap rock (in the z-direction) over 30 years, assuming an initial cap rock porosity of 5%, no
faults, and non-reactive hydrogen (Figure 4.4a). However, hydrogen reacts with the
mineralogical assemblage of the reservoir rock and brine and is converted by BSR and
methanogenesis. The loss of aqueous hydrogen by diffusion is minor. Dissolved hydrogen and
corresponding aqueous species concentrations (OH−, CH4, HCO3−, Al(OH)4−, CaOH+,
CaHCO3+, MgOH+, MgHCO3+, H2, NH4+, NH3, H2S, HS−, H4SiO4, and H3SiO4−) in the cap rock
brine show no significant changes. The most significant difference has been calculated for the
NH4+ concentration, which rises to around 4 × 10−4 mol kgw–1 in the first meter of the cap rock
(cell 182). Therefore, the effect of hydrogen storage on the cap rock is only visible in the first
meter of the cap rock, which is in contact with the reservoir rock (cell 182), after 30 years. Even
though each aqueous species migrates with a different rate along its concentration gradient,
as described by Buzek et al. (1994), the difference in migration is too small to see deviations
in the model over the modeled time of 30 years. The loss of the non-reactive hydrogen tracer
by diffusion into the cap rock is 25% and represents the maximal loss of aqueous hydrogen by
diffusion over 30 years.
In summary (Sections 4.3.1–4.3.3), the loss of hydrogen gas by bacterial conversion, gas–
water–rock interactions, and aqueous diffusion is 76% over 30 years.
88
a)
(cells /
meters)
(mol kgw-1 )
174
0.0010
176 0.0008
178 0.0006
180 0.0004
rock
Cap
182 0.0002
Reservoir
184 0.0000
rock
0 6 12 18 24 30
Time (years)
(cells / b)
meters) (mol kgw-1)
174 0.0010
176 0.0008
178 0.0006
180 0.0004
rock
Cap
182 0.0002
Reservoir
184 0.0000
0.00
0 60 120 180 240 300
rock
Time (years)
Figure 4.4 Diffusion of a non-reactive hydrogen tracer through the cap rock over (a) 30 years and (b)
300 years.
4.3.4 Influencing Factors
To identify the controlling factors of hydrogen loss and the parameters that affect the modeling
results, simulation of generic model scenarios was performed.
89
4.3.4.1 Storage Time
Because of the lack of data, the storage time has been chosen based on the equipment lifetime
of 30 years (Lord et al., 2014). To show the influence of longer storage times on the system
environment, in this scenario a longer storage time of 300 years is modeled. With increasing
storage time, the dissolved hydrogen has more time to diffuse into higher regions of the cap
rock. After 300 years, the non-reactive tracer diffuses 10 m through the cap rock (Figure 4.4b),
whereas the reactive hydrogen affects just the first 5 m of the cap rock.
When considering longer storage times, it becomes clear that the mass of generated water is
increased to 1.03 kg (1 kg + 0.3 kg = 1.03 kg water in total) in the reservoir rock because the
water-producing processes, BSR and methanogenesis (Equations (4.2) and (4.3)), are time-
dependent. In the reservoir rock, the pH is increased to 8.7. The porosity loss from 10% to
9.19–9.99% is greater than that after 30 years of storage. The reason for the greater decrease
is the intensified dissolution of quartz and illite. Calcite is completely consumed in the hotspot
areas, the contact areas of the cap rock to the reservoir rock and of the underlying rock to the
reservoir rock. In all other cells, calcite is still available, but it is only partly dissolved. On the
other hand, the albitization process is intensified in the hotspot areas and more kaolinite is
formed. The reactive amounts of anhydrite and goethite are completely consumed. The only
sulfate source for BSR is delivered by diffusion from the underlying rock and the cap rock into
the reservoir rock. The mineralogical changes in the cap rock are minimal and show that even
after 300 years of hydrogen storage, the effects of diffusion of aqueous hydrogen are small.
When the storage time is increased to 300 years (under the assumption that no new gas
mixture of H2(g) and CO2(g) is injected), the total amount of newly generated CH4(g) is increased
to an average of 0.25 mol L−1 of gas and the H2S(g) concentration is increased to an average
of 0.006 mol L−1 of gas (Figure 4.2b). After 300 years, the total amount of stored H2(g) is
consumed.
In the case of storage cycles, including gas injection, storage, and production phases, the
newly injected hydrogen gas is again accompanied by a new amount of CO 2(g), which is
available as an electron acceptor for methanogenic bacteria. This aspect is modeled in a
separate transport model. The total loss of hydrogen by conversion to CH4(g) is higher if the
storage cycle has a higher frequency, because CO2(g) is co-injected and is available for
methanogenesis. With a total storage time of 30 years and a dwelling time of 6 months, a
maximum of 60 injections of “fresh gas” (composed of H2(g) and CO2(g)) can be stored in the
reservoir. This leads to a total generation of 6.72 mol L−1 of gas CH4(g) after 30 years. However,
the loss of hydrogen after 6 months, the shorter storage period, is of course smaller
90
(0.112 mol L−1 of gas in 6 months) than after 30 years of storage. Consequently, a shorter
storage period leads to a lower amount of generated CH4(g) (and a lower loss of hydrogen)
because methanogenesis is a time-dependent process. However, over the total time, the loss
of hydrogen sums to a higher amount. This holds true only for methanogenesis. For bacterial
sulfate reduction, the amount of generated H2S(g) depends not only on the amount of available
sulfate but also on the time-dependent kinetic rate constant.
To summarize, it is clear that longer storage times increase the risk of loss of hydrogen. On
the other hand, the reactive anhydrite and calcite will be consumed after a few years of storage
and safer storage conditions arise because BSR and methanogenesis will be limited by the
amount of sulfate and carbon dioxide delivered by diffusion (and diffusion is a slow process).
In this case, the hotspot areas gain in importance. However, assuming shorter storage periods
of 6 months, the CH4(g) concentration in the stored gas will be lower than after a storage time
of 30 years, but the total loss of hydrogen sums to higher amounts regarding shorter storage
periods (60 injections in 30 years).
4.3.4.2 Pressure and Temperature Conditions in Gas Reservoirs
Each depleted gas field has specific pressure and temperature conditions. To analyze the
influence of pressure and temperature on the modeling results of hydrogen storage, the
pressure and temperature conditions are varied.
The higher pressure and temperature conditions chosen are 161 atm and 80 °C. At even
higher temperature conditions, the sulfate-reducing bacteria and methanogenic bacteria are
mostly no longer active. The maximum temperature for SRB and methanogenic bacteria is
80 °C (Jorgensen et al., 1992; Magot et al., 2000). At higher pressure and temperature
conditions, thermochemical sulfate reduction, as known from deeply buried hydrocarbon
reservoirs, must be considered.
At 161 atm and 80 °C (at invariable kinetic rate constants for BSR and methanogenesis), the
amount of consumed H2(g/aq) is slightly lower (75%) than that in the reference scenario (76%;
modeled at 40 atm and 40 °C). The increased pressure results in a decreased gas volume and
in higher gas concentrations. The H2S(g) concentration is increased to an average of
0.009 mol L–1 of gas and the CH4(g) concentration is increased to an average 0.54 mol L–1 of
gas after 30 years (Figure 4.2c).
91
The porosity loss (from initial 10% to 9.52–9.86%) in the reservoir rock is, on average, higher
than in the reference scenario. This can be explained by the stronger albitization at this higher
pressure and temperature conditions and the pyrite formation increases as well. On the other
hand, the dissolution of quartz, calcite, and illite is higher. However, the amount of precipitated
minerals is higher than the amount of dissolved minerals and this causes the decrease in
porosity. The pH in the reservoir rock brine is 7.5–8.4, which is lower than in the reference
scenario. In addition, the amount of the aqueous sulfide sulfur (S(–II)) in the reservoir brine is
decreased at these conditions. The mass of generated water is, despite BSR and
methanogenesis, decreased from 1 kg to 0.96 kg. A reason for this could be the stronger
albitization process, binding the water.
The changed pressure and temperature conditions affect the equilibrium constants of the mass
action laws of gases, minerals, and aqueous species. Here, a modeling limitation is achieved
because the phreeqc.dat database does not contain pressure dependencies for the equilibrium
constants for all mineral phases that are used in the model. The pressure dependency of the
following mineral phases in the model are included: calcite, dolomite, anhydrite, barite, quartz,
and halite.
As in the reference scenario, the temperature and pressure conditions change along the
diffusive pathway of the aqueous species through the cap rock according to the geothermal
gradient of 33.3 °C km–1 depth and 100 atm km–1 depth under hydrostatic conditions, starting
with 80 °C and 161 atm at the reservoir depth. Each cell is defined by a specific temperature
and pressure condition. The influence of these changing pressure and temperature conditions
on the cap rock is minor. The concentration of aqueous sulfate sulfur S(+VI) is higher and
aqueous sulfide sulfur (S(–II)) is lower than in the reference scenario. Different from the
reference scenario, albitization occurs at these higher pressure and temperature conditions
even in the cap rock, and dawsonite precipitation and halite dissolution are slightly increased.
To summarize, at higher pressure and temperature conditions (while the kinetic rate constants
are not changed), the concentration of CH4(g) and H2S(g) is increased due to a lower gas
volume. Furthermore, the loss in porosity in the reservoir rock is increased. However, the
kinetic rate constant is pressure- and temperature-dependent; therefore, this influence is
tested in a further scenario (Section 4.3.4.3).
92
4.3.4.3 Kinetic Rate Constant
Special consideration is given to the influence of the kinetic rate constants of bacterial sulfate
reduction and methanogenesis on the loss of hydrogen during storage. Therefore, the rate
constants are varied. As a reminder, the initial kinetic rate constant in the reference scenario
for BSR is 9.26 × 10−8 mol kgw−1 s−1 and 2.30 × 10−9 mol kgw−1 s−1 for methanogenesis.
By increasing the kinetic rate constant for BSR slightly to 2.40 × 10−7 mol kgw−1 s−1, based on
data from Herrera et al. (1997), the total amount of stored hydrogen is lost in less than 20
years. The amount of generated H2S(g) does not change because the reactive amount of
anhydrite in the reservoir rock is completely consumed and the only sulfate source comes from
the cap rock and underlying rock by diffusion and diffusion is a slow process. By increasing
the storage time to 300 years (with a kinetic rate constant of 2.40 × 10−7 mol kgw−1 s−1 for
BSR), the effect of diffusion is recognizable and the H2S(g) generation increases in the hotspot
areas. The amount of generated CH4(g) increases from 0.2 mol L−1 of gas to an average of
0.25 mol L−1 of gas in each cell of the reservoir rock after 30 years (Figure 4.2d). The reason
for this could be that methanogenesis starts as soon as all the sulfate for BSR is consumed.
Because sulfate is consumed faster, due to the increased kinetic rate constant, the timeframe
in which the methanogenesis can take place is expanded. On decreasing the kinetic rate
constant for BSR to 9.26 × 10−9 mol kgw−1 s−1, the amount of generated H2S(g) does not change
because the reactive amount of anhydrite in the reservoir rock is completely consumed. Less
H2(g/aq) is consumed (14% loss) and a smaller amount of CH4(g) (on average 0.013 mol L−1 of
gas) is generated (Figure 4.2e).
At a higher kinetic rate constant for methanogenesis of 2.30 × 10−6 mol kgw−1 s−1, the total
amount of stored hydrogen is lost in less than 6 years (Figure 4.2f). CH4(g) production increases
to an average of 0.25 mol L−1 of gas and the mass of water rises to 1.03 kg per cell in the
reservoir rock after 30 years of storage. The amount of CO 2(g) from the injected gas mixture
and the residual gas is completely consumed in less than 3 years of storage and is not the
limiting factor for methanogenesis. If no CO2(g) is available anymore from the injected gas
mixture and residual gas, the dissolution of carbonate-bearing minerals begins and delivers
CO2 for methanogenesis. With a smaller kinetic rate constant for methanogenesis of
2.30 × 10−12 mol kgw−1 s−1, the loss of H2(g/aq) is slightly lower (75.7%). Less CH4(g) is generated
(but the differences are in the third decimal place, only from 0.192 mol L−1 of gas in the
reference scenario to 0.191 mol L−1 of gas) and the H2S(g) generation is constant (Figure 4.2g).
The produced mass of water and the pH conditions show no changes compared to the
reference scenario.
93
It can be concluded that the kinetic rate constants are important factors controlling the loss of
hydrogen storage. Knowledge of the kinetic rate constants for BSR and methanogenesis at
elevated temperature and pressure are required. Nevertheless, the maximal amount of
generated H2S(g) is, if all reactive sulfate-bearing minerals (e.g., anhydrite) are consumed,
limited by diffusion. On the other hand, the maximal amount of generated CH4(g) depends on
the kinetic rate constant and the storage time.
4.3.4.4 Stored Gas Composition
In the “POWER-to-GAS-to-POWER” concept for the storage and usage of hydrogen, the
hydrogen is stored purely in underground storage systems (Hagemann et al., 2016). As a
consequence of pure H2(g) injection and storage (no CO2(g) is co-injected), the amount of
generated CH4(g) decreases to an average of 0.32 mol L−1 of gas after 30 years. H2S(g)
generation is constant with an average of 0.005 mol L−1 of gas after 30 years (Figure 4.2h).
The loss of stored H2(g/aq) is slightly increased to 77% after 30 years of storage.
An alternative CO2 source for methanogenesis is the residual gas that was not removed during
natural gas production. Only if the residual gas is consumed or not reactive, does carbonate-
bearing mineral dissolution increase and delivers CO2 for methanogenesis, which is even more
pronounced at a longer storage time of 300 years. Finally, the amount of available and reactive
carbonate-bearing minerals (here calcite and dolomite) limits the process of methanogenesis
if no CO2(g) is stored with H2(g/aq). On the other hand, a separate test model shows that a higher
amount of co-injected CO2(g) (14.8 atm; 10 times higher than in the reference scenario)
increases the amount of generated CH4(g) to an average of 1.0 mol L–1 of gas. If the amount of
available CO2(g) would be infinitely large, the time-dependent kinetic rate constant and the
storage time limit the loss of hydrogen by conversion to CH4(g).
Without CO2(g) co-injection, the porosity in the reservoir rock decreases from 10% to 9.82–
9.98%. This loss in porosity is slightly smaller than in the reference scenario in which 4% CO2(g)
is co-injected. The changes in mineral dissolution and precipitation are minor in comparison to
the reference scenario. The amounts of precipitated kaolinite, K-feldspar, and dolomite are
slightly smaller than in the reference scenario and less calcite and illite are dissolved. No
changes in the pH are observable. The smaller amount of precipitated K-feldspar and kaolinite
leads to a smaller amount of bound water in these minerals. Consequently, the mass of water
increases to 1.025 kg in the reservoir rock and is higher than the increase in the reference
scenario (1.007 kg). The changes in the mineralogical composition of the cap rock are minor.
94
4.4 Conclusions
The simulation results show that the underground storage of hydrogen in depleted gas fields
entails the risk of hydrogen loss (and related energy loss) by bacterial conversion to CH4(g) and
H2S(g) and gas–water–rock interactions, which in turn lead to changes in the porosity of the
reservoir rock. The modeling results of the one-dimensional reactive mass transport model
identify the factors that control the loss of hydrogen:
• The loss of hydrogen by bacterial conversion to CH4(g) via methanogenesis is limited

mainly by the amount of co-injected CO2(g), the reaction kinetics, and the connected
maximal storage time of 30 years. Less co-injected CO2(g) will reduce H2(g) loss but
cannot inhibit the conversion to CH4(g) if further CO2 sources are available in the form
of residual gas and carbonate-bearing minerals. The generation of CH4(g) by
methanogenesis where CO2 is only delivered by the dissolution of carbonate-bearing
minerals is slower because the dissolution process limits the conversion to CH4(g).
• The loss of hydrogen by bacterial conversion to H2S(g) via bacterial sulfate reduction is
limited mainly by the amount of available sulfate in the reservoir. After complete
consumption of reactive anhydrite, the only sulfate source comes from the cap rock
and the underlying rock by diffusion and consequently limits the loss of hydrogen by
bacterial sulfate reduction because the process of diffusion is slow.
• The mass of generated water, as a product of BSR and methanogenesis, increases
the pressure in the system and diffusion can be intensified.
• The mineralogical changes in the reservoir rock result in a small decrease in porosity.
As a consequence, the available pore space for hydrogen storage decreases over 30
years such that (i) the same amount of hydrogen is moved to greater distances from
the bore hole or (ii) less hydrogen can be stored in future injection phases.
• The loss of aqueous hydrogen by diffusion and related effects on the cap rock
mineralogy is negligibly small, with a storage time of 30 years, because hydrogen
storage causes gas–water–rock interactions in the reservoir rock and brine and is
converted by BSR and methanogenesis to a greater extent. Furthermore, the process
of diffusion is slow.
• A longer storage period increases the loss of the stored hydrogen. Shorter storage
periods lead to less hydrogen loss for each period, but over the total time the summed
loss of hydrogen is higher because new CO2(g) is available for methanogenesis after
each gas injection.
95
• At higher pressure and temperature conditions, the concentrations of CH4(g) and H2S(g)
increase due to a lower gas volume. Furthermore, the loss in porosity in the reservoir
rock increases as well.
• Knowledge of kinetic rate constants for bacterial sulfate reduction and methanogenesis
at elevated levels of pressure and temperature are required as accurately as possible.
It is recommended to choose depleted gas fields for hydrogen storage where the residual gas
has low CO2(g) concentrations. The mineralogical composition of the reservoir rocks should
contain low amounts of sulfate- and carbonate-bearing minerals but high amounts of reactive
iron-bearing minerals. Reservoirs with low pressure and temperature conditions are
recommended as well as storage gas compositions with low amounts of CO 2(g). From the
modeling results, it seems reasonable to implement a multicomponent transport model and to
conduct a joint variation of several parameters as a next step.
We thank Leo Fuhrmann for technical assistance. We would like to thank three anonymous
reviewers for their constructive reviews.
Input file S4 for the transport model to calculate the reference scenario UHS is available (for
PHREEQC Version 3).
96
Final Discussion
5 Final Discussion
When addressing the increasing carbon dioxide emissions and related climate changes, it is
necessary to draw attention to a shift in energy production (Henkel et al., 2016). This also
comprises the research about safe underground gas storage, including carbon capture and
storage, natural gas storage in salt caverns, and underground hydrogen storage.
These underground gas storage options are important for the so-called German
“Energiewende”. CCS is of relevance for the reduction of greenhouse gas emissions, SCGS
is relevant to being more independent from gas suppliers and to compensate seasonal
fluctuations in the supply and demand of energy, and UHS is relevant for storing energy
generated from renewable energy sources like wind and solar but also, like SCGS, to
compensate for seasonal fluctuations.
This study aims at analyzing the hydrogeochemical processes that are induced by
underground gas storage. The chemical reactions between the available gases, aqueous
solutions, and mineral phases (gas–water–rock interactions) in the considered systems are
modeled. The tools for modeling the reactive mass transport are PHREEQC (1D) and PHAST
(3D), both are provided by the US Geological Survey. The semi-generic, hydrogeochemical
models are based on chemical equilibrium thermodynamics and kinetic reactions for sulfate
reduction in SCGS and UHS, as well as kinetic reactions for methanogenesis in UHS.
In the field of CCS, many studies focus on CO2–water–rock interactions in reservoir rocks in
depleted hydrocarbon reservoirs but much less attention has been paid to interactions in the
cap rocks. However, this is of great importance for risk assessment. To fill this gap, the first
part of this study (Section 2) focuses on the long-term change in the sealing capacity of the
cap rock in CO2 storage systems triggered by mineral dissolution and precipitation. The main
difference to other studies is the consideration of the diffusive mass transport of CO 2 along a
gradient of decreasing pressure and temperature conditions. The results of this model are
discussed and allow the identification and quantification of the loss or gain of total porosity
affected by hydrogeochemical reactions driven by diffusive mass transport.
In the second part of this study (Section 3), modeling-based evidence is presented that H2S(g)
generation in salt cavern gas storage is possible in terms of chemical thermodynamics,
reaction kinetics, and coupled diffusive mass transport. It is a potential risk for the gas quality
of the stored gas. Several studies on H2S(g) generation in hydrocarbon reservoirs and
hydrocarbon storage environments exist. However, this one is the first that clarifies and
quantifies H2S(g) generation processes in salt caverns filled with natural gas. Because of the
97
Final Discussion
increasing demand for storage capacity in salt caverns, the use of less favorable and sulfate-
bearing salt formations will increase, and the potential risk of H2S(g) generation will rise in the
future. Therefore, the interpretation of the modeling results allows the analysis of the limiting
factors for H2S(g) generation in salt caverns and the identification of technical approaches to
decrease or inhibit H2S(g) generation and release into the stored gas.
The third part of this study (Section 4) deals with the loss of hydrogen during underground
storage in terms of chemical thermodynamics, reaction kinetics, and coupled diffusive mass
transport. This study is the first one that clarifies and quantifies the loss of hydrogen (and
related energy loss) as a combination of (i) the conversion by bacteria, (ii) gas–water–rock
interactions in the reservoir and cap rock, which relate to porosity changes, and (iii) loss by
aqueous diffusion along a gradient of decreasing pressure and temperature conditions. The
interpretation of the modeling results allows analysis of the limiting factors of the loss of
hydrogen and the identification of storage conditions optimized for hydrogen storage in
depleted gas fields.
5.1 General Model Similarities and Differences
In all three subtopics of this study, numerical models are developed to simulate the
hydrogeochemical processes induced by underground gas storage in space and time. In each,
the gas–water–rock interactions that could lead to risks when storing gases in deep geological
formations are analyzed, followed by the description of possible consequences and potential
inhibitions methods. This study also shows new dependencies (like the pressure and
temperature changes along the diffusional transport) that were not considered in previous
studies and opens new applicability (like inhibitions methods for H2S(g) generation in salt
caverns). All discussed models are set up so that they can be adapted and applied to other
underground gas storage systems.
Panfilov (2010) stated that the hydrodynamic and chemical behavior of hydrogen stored in the
underground has little in common with the storage of natural gas or carbon dioxide.
Underground storage of hydrogen comes with more leakage problems than natural gas
storage due to its chemical and physical properties, such as high diffusivity, low viscosity, and
low density, resulting in a higher mobility (Ebigbo et al., 2013). Nevertheless, the principal
hydrogeochemical modeling approaches and general basics, such as induced gas–water–rock
interactions, the pressure and temperature dependence, dissolution and precipitation of
minerals, gas solubilities, diffusional transport, and so on, are the same in all three gas storage
98
Final Discussion
systems (see Table 1.2 and Figure 5.1). The differences between the systems are e.g. the
dimension (CCS and UHS: 1D, SCGS: 1D and 3D), the kind of modeling (CCS thermodynamic
equilibrium modeling, SCGS and UHS thermodynamic equilibrium and kinetic modeling), the
transport mechanisms (SCGS and UHS diffusion only, in CCS diffusion and advection in an
alternative model scenario), and the modeling time (CCS is constructed for long-term storage,
SCGS and UHS are designed for seasonal storage).
Storage of Gases in Deep Geological Structures:

Spatial and Temporal Hydrogeochemical Processes Evaluated and Predicted by the Development and
Application of Numerical Modeling
Reactive mass transport, equilibrium thermodynamics, kinetic reactions, gas–water–rock

interactions, porosity changes, high pressure and temperature conditions, modeling codes:
PHREEQC, (PHAST)
Carbon Capture Salt Cavern Underground

and Storage Gas Storage Hydrogen Storage
(CCS) (SCGS) (UHS)
1DRMT 1DRMT 1DRMT

3DRMT
Diffusion Diffusion Diffusion

(advection)
Loss via cap rock Loss via Loss via cap rock
conversion and conversion
Common conclusions: identify mineral dissolution and precipitation, quantify porosity

changes, indicate gas loss and related gas pollution, identify influencing factors and
conditions for safe underground gas storage
Figure 5.1 Overview of the three subtopics CCS, SCGS, and UHS.
99
Final Discussion
Furthermore, the analyzed gases are the same that occur in natural oil and gas fields (like
CH4(g), CO2(g), and in small amounts H2(g)), in which, under same reservoir conditions, similar
problems occur. For example, the problem of H2S(g) generation in natural gas fields is already
known but has not been analyzed in salt caverns when storing natural gas.
Additionally, in all three underground gas storage systems “hotspot” areas are identified and
the factors that could influence the modeling results are compared. These hotspot areas are
regions where intensified gas–water–rock interactions occur. In CCS the main changes in
porosity are identifiable close to the cap rock/reservoir rock contact. The hotspot area for
SCGS, in which the highest total H2S concentrations occur, is identified in the brine on the
diffusive path where methane and sulfate meet and react with each other. In UHS, two hotspot
regions with higher amounts of generated H2S(g) are identified. The contact area of the
reservoir rock and the cap rock and the contact area of the reservoir rock and the underlying
rock. The highest CH4(g) concentrations can be found at the contact area of the reservoir rock
and the underlying rock. In summary, hotspot areas in all three underground gas storage
systems are identified at the interface of two units with different mineralogical compositions
and/or brine compositions.
5.2 Influencing Factors
In all three subtopics, the influencing factors that affect the modeling results are analyzed and
identified. Table 5.1 shows the comparison of the considered parameters for each
underground gas storage system. Tortuosity, initial cap rock porosities, and the advection
transport mechanism are only considered in a single study and are not compared.
Table 5.1 Effect of the different influencing factors on the modeling results in comparison of the three
underground gas storage systems. Red = high influence, green = low influence, grey = not considered.
CCS SCGS UHS

Pressure/temperature
Kinetic rate constant
Tortuosity
Gas composition
Mineralogical compositions
Initial cap rock porosities
Effective diffusion coefficient
Advection
Storage time
100
Final Discussion
The influence of pressure and temperature conditions when storing gases in deep geological
structures is of great interest. When storing carbon dioxide in depleted hydrocarbon reservoirs,
mineral dissolution and precipitation and resulting porosity changes are pressure and
temperature dependent. Higher pressure and temperature conditions have a stronger effect
on mineral dissolution and precipitation. Compared to temperature, the effect of pressure on
the change of porosity is slightly stronger in CCS systems. When storing natural gas in salt
caverns, the changing pressure conditions during the injection, storage, and production phases
have a minor influence on generated H2S(g). However, the pressure and temperature
dependent kinetic rate constant for bacterial sulfate reduction influences the amount of
generated and released H2S(g). The pressure and temperature conditions are also of interest
when storing hydrogen in depleted gas reservoirs because at higher pressure and temperature
conditions, the concentrations of CH4(g) and H2S(g) increase due to a lower gas volume and the
loss in porosity of the reservoir rock increases. Similar to SCGS, the kinetic rate constant
influences the loss of the stored hydrogen. Therefore, accurate knowledge of the kinetic rate
constants at elevated pressure and temperature conditions is required.
In CCS systems, the injected gas composition affects the sealing capacity of the cap rock,
which can increase by mixing CO2 with CH4 and H2S in comparison to pure CO2 injection. In
SCGS systems, the composition of the stored natural gas can influence the amount of
generated and released H2S(g). In UHS systems, less co-injected CO2(g) will reduce H2(g) loss,
and the initial residual gas in the reservoir should have low CO2(g) concentrations as well.
The results of the CCS model indicate that porosity creation depends strongly on the
mineralogical composition of the reservoir and cap rock as well as on the corresponding brine
composition. This is also shown in SCGS where the amount of available sulfate in the rock salt
formation is one of the main limiting factors of the H2S(g) generation process. This is
comparable with the results of the UHS study in which the loss of hydrogen by bacterial
conversion to H2S(g) is limited mainly by the amount of the available sulfate in the reservoir
rock, the cap rock, and the underlying rock.
The effective diffusion coefficient is only varied in the study of CCS, but the influence of
diffusion is analyzed in SCGS and UHS as well. In CCS, even over the broad range of effective
diffusion coefficients tested, the intensity of the diffusive mass transport has negligible effects
on mineral precipitation and dissolution and on resulting porosity changes. In SCGS, the
tortuosity affects the amount of generated H2S(g), but the diffusion of dissolved methane and
dissolved sulfate through the brine have a stronger effect on the generation of H2S(g). In UHS,
the loss of H2(aq) induced by BSR and methanogenesis predominates, assuming a storage time
of 30 years. Whereas the loss of aqueous hydrogen by diffusion is negligible. Furthermore, the
101
Final Discussion
loss of hydrogen by BSR is limited mainly by the amount of anhydrite (the sulfate source in the
reservoir rock). After complete consumption of reactive anhydrite, the only sulfate supply into
the reservoir is via diffusion from the cap rock and the underlying rock. This consequently limits
the loss of hydrogen by bacterial sulfate reduction because the process of diffusion is slow.
In CCS, a storage time of one million years is simulated because it is considered as final
disposal and the long-term effects must be analyzed. In SCGS and UHS the typical 30 years
lifetime of the salt caverns or the equipment is modeled. Additionally, in UHS a longer storage
time of 300 years is modeled which increases the loss of the stored hydrogen.
5.3 Optimal Storage Conditions
With the increasing demand for underground storage capacities and the related conflict of
interest, it is of rising importance to figure out which storage option is optimal for which kind of
gas. With the results of the hydrogeochemical modeling in this study, safe conditions for each
considered gas can be defined.
For the storage of natural gas in salt caverns, salt structures with low amounts of anhydrite
and high amounts of iron-bearing minerals should be chosen. The pressure conditions have a
minor influence on the loss of the stored natural gas by conversion to H2S(g).
For the storage of hydrogen in depleted gas reservoirs, similar problems as for SCGS in salt
caverns arise. Therefore, the conditions in the depleted gas reservoirs must be similar to those
for SCGS in salt caverns. The mineralogical composition of the reservoir rocks should contain
low amounts of sulfate- and carbonate-bearing minerals but high amounts of reactive iron-
bearing minerals. Depleted gas reservoirs with low pressure and temperature conditions are
recommended. The stored hydrogen gas, as well as the initial gas in the reservoir, should have
low amounts of CO2(g).
When storing carbon dioxide in depleted hydrocarbon reservoirs, the cap rock should contain
low amounts of calcite. Furthermore, low initial cap rock porosities are recommended as well
as low pressure and temperature conditions in the reservoir.
5.4 Model Limitations
All models have their limitations. Modeling is always an attempt to reproduce reality but the
interlocking of all the parameters that must be considered are too complex and only single
102
Final Discussion
parts of a system can be displayed. Nevertheless, models enable a critical analysis and lead
to a better understanding of the analyzed system (Konikow and Bredehoeft, 1992).
Even if PHREEQC is a general hydrogeochemical program, which is applicable to various

hydrogeochemical environments, some limitations occur (Parkhurst and Appelo, 2013).
• High-salinity waters can be calculated with the pitzer.dat database when the Debye-
Hückel equation is no longer applicable. The Pitzer approach calculates ion activity
coefficients at high ionic strength but the pitzer.dat database includes only a limited
number of elements and is therefore limited in application.
• Mass transport modeling does not consider gas-transport between different cells (multi-
phase flow). To overcome this limitation, separate transport models can be simulated
that provide initial amounts of the gases in the appropriate cells.
• The changes in porosity must be calculated manually from PHREEQC results. The
effects of porosity and permeability changes during modeling on diffusion are,
therefore, not considered.
• PHREEQC cannot calculate in 3D.
• The database phreeqc.dat used includes the pressure and temperature dependence
of the mass-action law constants for the equilibrium reactions of many but not all
aqueous, gaseous, and solid species involved in the models.
The reactive amount of each mineral in the rock assemblages are precalculated based on the
respective porosity and is included in the input file. PHREEQC and PHAST allow mineral
precipitation until the available water is no longer oversaturated with any mineral (Fu, 2014).
That could result in a higher amount of newly formed minerals than the available pore space
allows. This effect should be considered at the evaluation and re-calculation phase of the
PHREEQC and PHAST modeling results when dealing with porosity changes. Furthermore,
PHAST is not capable of calculating the pressure dependence of solubility equilibrium
constants for minerals, gases, and aqueous species.
The data required for kinetic models (like kinetic rate constants, activation energies, and
experimental rate laws) are learned from experiment and are still rare. The corresponding
information required for equilibrium models (equilibrium constants, enthalpies, entropies,
Gibbs free energies, and heat capacities) are more abundant (Stumm and Morgan, 1981).
Kinetically-based models are limited by the rare rate constants for the modeling conditions.
103
Final Conclusion and Outlook
6 Final Conclusion and Outlook
The hydrogeochemical one-dimensional reactive mass transport modeling approaches

presented in this study link the sciences of hydrogeology, petroleum geology, and engineering,
including safety and geochemical aspects related to underground gas storage systems.
This study contributes to the process-understanding of underground gas storage systems. The
application of the reactive mass transport modeling presented provides a good example of how
such modeling approaches could (i) identify mineral dissolution and precipitation in the cap
and reservoir rocks triggered by gas storage and molecular diffusion of stored gases, (ii)
quantify the resulting porosity changes, (iii) indicate gas loss and related gas pollution by
bacterial conversion via BSR and methanogenesis, (iv) identify hydrogeochemical methods to
inhibit H2S(g) generation and release, and (v) identify conditions for safe underground gas
storage.
With the results of this study, the discussion of CO2–water–rock interactions in the cap rocks
of carbon capture and storage systems may be stimulated, attention may be drawn to the risk
of H2S(g) generation in salt caverns, and the hydrogeochemical risks of underground hydrogen
storage may be highlighted, concluding that hydrogeochemical modeling provides an effective
tool for efficient underground gas storage management.
In future studies, the focal point could be on specific modeling approaches considering
propane-butane liquid storage in relation to the H2S(g) generation in salt caverns. The existing
model can be adapted to different cavern conditions, various hydrocarbons, and is able to
consider—besides anhydrite—further sulfur minerals. For upcoming studies, it seems
reasonable to implement a multicomponent transport model and to conduct a joint
(probabilistic) variation of several parameters as a next step. The software platform RESUS
(Bojanowski, 2017; Ghofrani, 2016; Li, 2015)—originally developed for the final disposal of
radioactive waste from the Clausthal University of Technology—could be applied. For the
evaluation of a specific gas storage system with the development of a numerical model, the
physical and chemical data of the underground gas storage system must be measured under
in situ conditions. In a next step, the mineral phase assemblages of the reservoir should be
exposed to different partial pressures of the stored gases, aqueous solutions, and pressure
and temperature conditions in long-term laboratory batch experiments. The mineral dissolution
and precipitation, as well as gas compositions, should be observed and the results retraced by
batch modeling using identical initial parameters. Such a combination of laboratory
experiments and modeling is recommended to test the plausibility of the applied
hydrogeochemical equilibrium models.
104
List of Figures
Figure 1.1 Intersections of the three subtopics CCS, SCGS, and UHS. .............................................. 15
Figure 2.1 System sketch of the model. CR = cap rock, RR = reservoir rock, CO2(sc) = supercritical
CO2 and CO2(aq) = aqueous CO2. ......................................................................................... 24
Figure 2.2 Initial conditions in the column of 101 cells and calculated distribution of changes in
porosity in %, mineral amounts as well as pH for the reference scenario after 106 years.
RR = reservoir rock, CR = cap rock. “Initial” shows the initial mineralogical composition of
the cap rock and the initial pH of the cap rock brine. Cell 2 is completely cemented. The
porosity is reduced from the initial 5.0% to -0.5% after 106 years (= decrease in porosity
of --5.5%). The amount of precipitated phases is higher than the available pore space.
Consequently, the phases will use the available pore space in the next cell. ..................... 34
Figure 2.3 Gain or loss of total porosity for scenarios 1 to 9 (a–i) in a column of 101 cells, with a cell
length of 1 m. 100 m are located in the cap rock and 1 m in the reservoir rock. The values
are given as mean values for every 10 meters. ................................................................... 37
Figure 2.4 Gain or loss of total porosity and mass conversion of the minerals for scenario 6a in the
first 10 cells with a cell length of 1 m. 9 m are located in the cap rock (CR) and 1 m in the
reservoir rock (RR). .............................................................................................................. 40
Figure 3.1 System sketch of the model. Black box = reference scenario, orange box = alternative
scenario. ............................................................................................................................... 52
Figure 3.2 H2S(g) generation and increasing H2S(g) amount with ongoing time and bacterial sulfate
reduction. ............................................................................................................................. 60
Figure 3.3 H2S(g) generation influenced by a) pressure changes (after 3 months) b) stored gas
composition (reference = typical composition of natural gas from Russia, alternative =
typical composition of natural gas from the North Sea (after 30 years) c) kinetic rate
constant (after 30 years) d) tortuosity (after 30 years). H2S(g) in mg m-3 in the stored gas.
Yellow = reference scenario, blue = modified parameters. a), b), and c) are modeled 1D in
PHREEQC, and d) is modeled 3D in PHAST. ..................................................................... 63
Figure 3.4 H2S(g), in mg m-3, in the stored gas after 30 years. Inhibition of H2S(g) generation and
release by addition of a) FeCl2 before first gas injection, b) FeCl2 after gas injection c)
1.0 mol kgw−1 NaOH after 5 years of storage, and 0.1 mol kgw−1 NaOH after 5 years and
after 10 years of storage to the brine. Reference is without any inhibition methods. Yellow =
reference scenario, green = inhibition factors. Red line = maximum allowed H 2S(g)
concentration in stored gas. ................................................................................................. 66
Figure 4.1 Selected reactions and processes associated with bacterial sulfate reduction and
methanogenesis (purple). Single arrow = kinetic-controlled reactions; double arrow =
equilibrium reactions; blue triangles = time-dependent diffusive transport of aqueous
components; bold = injected gas for storage. ...................................................................... 77
Figure 4.2 Changes in gas composition in the reservoir rock with ongoing time in (a) reference
scenario and influenced by (b) increased storage time of 300 years, (c) higher pressure and
temperature conditions, (d) higher kinetic rate constant for BSR of 2.40 × 10-7 mol kgw–1 s-1,
(e) lower kinetic rate constant for BSR of 9.26 × 10-9 mol kgw−1 s-1, (f) higher kinetic rate
constant for methanogenesis of 2.30 × 10−6 mol kgw−1 s-1, (g) lower kinetic rate constant for
methanogenesis of 2.30 × 10−12 mol kgw−1 s-1, (h) varied stored gas composition (without
CO2(g))................................................................................................................................... 84
Figure 4.3 Volume changes of mineral phases (L kgw−1) in the reservoir rock after 30 years of
hydrogen storage (reference scenario). ............................................................................... 87
Figure 4.4 Diffusion of a non-reactive hydrogen tracer through the cap rock over (a) 30 years and (b)
300 years. ............................................................................................................................ 89
Figure 5.1 Overview of the three subtopics CCS, SCGS, and UHS. ................................................... 99
106
List of Tables
Table 1.1 Chemical and physical properties of the stored gases. .......................................................... 6
Table 1.2 Underground gas storage models and their parameters. ..................................................... 17
Table 2.1 Mineralogical composition of the reservoir rock and the cap rock. ....................................... 27
Table 2.2 Initial irreducible water composition of the reservoir rock and cap rock. .............................. 27
Table 2.3 Equilibrium phases, mass-action equations and equilibrium constants used in the model.
Data from phreeqc.dat, but dawsonite and nahcolite are from llnl.dat (Parkhurst and Appelo,
2013). ................................................................................................................................... 29
Table 2.4 Results of temporal-spatial refinement rounded to two decimal places. The parameters used
for the reference scenario are shown in bold. ...................................................................... 31
Table 2.5 Parameter sensitivity analysis in alternative scenarios. ....................................................... 35
Table 2.6 Comparison of solubility constants of CO2(g) based on the change in total porosity in the
reference scenario for cells 2–10. log K -1.468 is the solubility constant used in the
reference scenario. .............................................................................................................. 42
Table 3.1 Equilibrium phases, mass-action equations, and equilibrium constants (log K, at 25 °C and
1 bar). Data are from phreeqc.dat, except for CH4(g), H2S(g), N2(g) which are from llnl.dat
(Parkhurst and Appelo, 2013). ............................................................................................. 53
Table 3.2 NaCl(solid) solubility (mol kgw−1) in pure water at different temperatures from literature
(experimental data) and modeled data. The database phreeqc.dat is used in the PHREEQC
model. .................................................................................................................................. 54
Table 3.3 Initial mineralogical composition of the rock salt formation used for the sump with a porosity
of 20%. The data are modified after Kyle and Posey (1991). Mackinawite, elemental sulfur,
and calcite are potential secondary phases which may form at saturation. ........................ 55
Table 3.4 Groundwater composition used for leaching (NLWKN, 2015), initial composition of the brine,
and the aqueous solution in the sump. ................................................................................ 56
Table 4.1 Equilibrium phases, mass action equations, and equilibrium constants used in the model.
Data from phreeqc.dat, except for dawsonite, nahcolite, CH 4(g), H2S(g), N2(g), which are from
llnl.dat (Parkhurst and Appelo, 2013)................................................................................... 76
Table 4.2 Mineralogical composition of the reservoir rock, cap rock, and underlying rock. Dawsonite,
nahcolite, mackinawite, sulfur, albite (and pyrite) are potential secondary phases. ............ 78
Table 4.3 Composition of the initial irreducible water in the reservoir rock, the cap rock brine, and the
underlying rock brine. ........................................................................................................... 80
Table 5.1 Effect of the different influencing factors on the modeling results in comparison of the three
underground gas storage systems. Red = high influence, green = low influence, grey = not
considered. ......................................................................................................................... 100
107
Abbreviations
aq Aqueous
BSR Bacterial sulfate reduction
CCS Carbon capture and storage
CR Cap rock
dm Diffusion coefficient in pure water
d´m Effective diffusion coefficient
ƒ Fugacity
g Gaseous
IAP Ion activity product
K Equilibrium constant
kgw Kilogram water
M Molal
P Pressure
ρ Partial pressure
pε Negative log of the activity of the electron (-)
pH Negative logarithm of the activity of aqueous H+ ions (-)
RR Reservoir rock
s Solid
SCGS Salt cavern gas storage
SI Saturation index, (-) (ion activity product/solubility constant)
SRB Sulfate-reducing bacteria
T Temperature
TSR Thermochemical sulfate reduction
UHS Underground hydrogen storage
wt % Weight percent
1D One-dimensional
3D Three-dimensional
1DRMT One-dimensional reactive mass transport
3DRMT Three-dimensional reactive mass transport
108
References
Adams, M.M., Hoarfrost, A.L., Bose, A., Joye, S.B., Girguis, P.R., 2013. Anaerobic oxidation
of short-chain alkanes in hydrothermal sediments: potential influences on sulfur cycling and
microbial diversity. Front. Microbiol. 4 (110). doi:10.3389/fmicb.2013.00110.
Adler, M., Eckert, W., Sivan, O., 2011. Quantifying rates of methanogenesis and
methanotrophy in Lake Kinneret sediments (Israel) using pore-water profiles. Limnol.
Oceanogr. 56 (4), 1525–1535. doi:10.4319/lo.2011.56.4.1525.
Alemu, B.L., Aagaard, P., Munz, I.A., Skurtveit, E., 2011. Caprock interaction with CO 2: A
laboratory study of reactivity of shale with supercritical CO2 and brine. Appl. Geochem. 26
(12), 1975–1989. doi:10.1016/j.apgeochem.2011.06.028.
Allison, J.D., Brown, D.S., Novo-Gradac, K.J., 1991. MINTEQA2/PRODEFA2, a geochemical
assessment model for environmental systems: Version 3.0 user's manual. Environmental
Research Laboratory, Office of Research and Development, Athens, Georgia, 107 pp.
Amid, A., Mignard, D., Wilkinson, M., 2016. Seasonal storage of hydrogen in a depleted natural
gas reservoir. Int. J. Hydrogen Energy 41 (12), 5549–5558.
doi:10.1016/j.iJhydene.2016.02.036.
Amosa, M.K., Mohammed, I.A., Yaro, S.A., 2010. Sulphide scavengers in oil and gas industry:
A review. NAFTA 61 (2), 85–92.
Angus, S., Armstrong, B., De Reuck, K.M., 1976. Carbon Dioxide – International
thermodynamic tables of the fluid state. Pergamon Press Ltd, Oxford–New York, 385 pp.
Appelo, C.A.J., Postma, D., 1999. Geochemistry, groundwater and pollution, 4th ed. Balkema,
Rotterdam, 536 pp.
Appelo, C.A.J., Postma, D., 2005. Geochemistry, groundwater and pollution, 2nd ed. CRC
Press, Leiden, 649 pp.
Appelo, C.A.J., Wersin, P., 2007. Multicomponent diffusion modeling in clay systems with
application to the diffusion of tritium, iodide, and sodium in opalinus clay. Environ. Sci.
Technol. 41 (14), 5002–5007. doi:10.1021/es0629256.
Araújo, O.d.Q.F., de Medeiros, J.L., 2017. Carbon capture and storage technologies: Present
scenario and drivers of innovation. Curr. Opin. Chem. Eng. 17, 22–34.
doi:10.1016/j.coche.2017.05.004.
Arning, E.T., Fu, Y., van Berk, W., Schulz, H.-M., 2011. Organic carbon remineralisation and
complex, early diagenetic solid–aqueous solution–gas interactions: Case study ODP Leg
204, Site 1246 (Hydrate Ridge). Mar. Chem. 126 (1–4), 120–131.
doi:10.1016/j.marchem.2011.04.006.
Arts, R., Eiken, O., Chadwick, A., Zweigel, P., van der Meer, L., Zinszner, B., 2004. Monitoring
of CO2 injected at Sleipner using time-lapse seismic data. Energy 29 (9–10), 1383–1392.
doi:10.1016/j.energy.2004.03.072.
Attia, A.M., 2005. Effects of petrophysical rock properties on tortuosity factor. J. Petrol. Sci.
Eng. 48 (3–4), 185–198. doi:10.1016/j.petrol.2005.06.012.
Ball, J.W., Nordstrom, D.K., 1991. User's manual for WATEQ4F, with revised thermodynamic
data base and text cases for calculating speciation of major, trace, and redox elements in
natural waters: U.S. Geological Survey Open-File Report 91–183, 189 pp.
Barajas, P., Faruk, C., 2014. Effective modeling and analysis of salt-cavern natural-gas
storage. SPE Prodcution and Operations 29 (01), 51–60.
Barnes, H.L. (Ed.), 1979. Geochemistry of hydrothermal ore deposits, 2nd ed. Wiley, New
York, 816 pp.
109
Bear, J., 1972. Dynamics of fluids in porous media. American Elsevier, New York, 764 pp.
Bérest, P., Brouard, B., 2003. Safety of salt caverns used for underground storage: Blow out,
mechanical instability, seepage, cavern abandonment. Oil Gas Sci. Technol. 58 (3), 361–
384.
Bernardez, L.A., de Andrade Lima, L.R., de Jesus, E.B., Ramos, C.L.S., Almeida, P.F., 2013.
A kinetic study on bacterial sulfate reduction. Bioprocess. Biosyst. Eng. 36 (12), 1861–
1869. doi:10.1007/s00449-013-0960-0.
Bethke, C.M., Yeakel, S., 2009. The Geochemist´s Workbench release 8.0 hydrogeology
program. University of Illinois, Urbana, 110 pp.
Beutel, T., Black, S., 2005. Salt deposits and gas cavern gas storage in the UK with a case
study of salt exploration from Cheshire. Oil Gas Eur. Mag. 1, 31–35.
Biehl, B.C., Reuning, L., Schoenherr, J., Lewin, A., Leupold, M., Kukla, P.A., 2016. Do CO 2-
charged fluids contribute to secondary porosity creation in deeply buried carbonates? Mar.
Pet. Geol. 76, 176–186. doi:10.1016/j.marpetgeo.2016.05.005.
Bildstein, O., Kervévan, C., Lagneau, V., Delaplace, P., Crédoz, A., Audigane, P., Perfetti, E.,
Jacquemet, N., Jullien, M., 2010. Integrative modeling of caprock integrity in the context of
CO2 storage: Evolution of transport and geochemical properties and impact on
performance and safety assessment. Oil Gas Sci. Technol. – Rev. IFP 65 (3), 485–502.
doi:10.2516/ogst/2010006.
Bischoff, G., Gocht, W., 1984. Energietaschenbuch, 2nd ed. Vieweg+Teubner Verlag,
Wiesbaden, 454 pp.
Black, J.R., Carroll, S.A., Haese, R.R., 2015. Rates of mineral dissolution under CO 2 storage
conditions. Chem. Geol. 399, 134–144. doi:10.1016/j.chemgeo.2014.09.020.
Blackford, J., Bull, J.M., Cevatoglu, M., Connelly, D., Hauton, C., James, R.H., Lichtschlag, A.,
Stahl, H., Widdicombe, S., Wright, I.C., 2015. Marine baseline and monitoring strategies
for carbon dioxide capture and storage (CCS). Int. J. Greenh. Gas Con. 38, 221–229.
doi:10.1016/j.ijggc.2014.10.004.
Blanc, P., Lassin, A., Piantone, P., Azaroual, M., Jacquemet, N., Fabbri, A., Gaucher, E.C.,
2012. Thermoddem: A geochemical database focused on low temperature water/rock
interactions and waste materials. Appl. Geochem. 27 (10), 2107–2116.
doi:10.1016/j.apgeochem.2012.06.002.
BMU, 2010. Sicherheitsanforderungen an die Endlagerung wärmeentwickelnder radioaktiver
Abfälle: Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit, 22 pp.
Bojanowski, R., 2017. Probabilistische Modellierung hydrogeochemischer Prozesse mit dem
Code PHAST und der Softwareplattform RESUS. Bachelor Thesis TU Clausthal.
Bolourinejad, P., Herber, R., 2015. Chemical effects of sulfur dioxide co-injection with carbon
dioxide on the reservoir and caprock mineralogy and permeability in depleted gas fields.
Appl. Geochem. 59, 11–22. doi:10.1016/j.apgeochem.2015.03.003.
Bowden, A.R., Rigg, A., 2005. Assessing reservoir performance risks in CO2 storage projects,
in: Greenhouse gas control technologies. Elsevier, pp. 683–691.
Bozau, E., 2013. Prozessmodellierung hochsalinarer Wässer mit einem erweiterten
PHREEQC-Datensatz. Grundwasser 18 (2), 93–98. doi:10.1007/s00767-013-0222-8.
Buzek, F., Onderka, V., Vanĉura, P., Wolf, I., 1994. Carbon isotope study of methane
production in a town gas storage reservoir. Fuel 73 (5), 747–752.
Church, C.D., Wilkin, R.T., Alpers, C.N., Rye, R.O., McCleskey, R.B., 2007. Microbial sulfate
reduction and metal attenuation in pH 4 acid mine water. Geochem. T. 8 (10).
doi:10.1186/1467-4866-8-10.
110
Cord-Ruwisch, R., Kleinitz, W., Widdel, F., 1987. Sulfate-reducing bacteria and their activities
in oil production. J. Pet. Technol. 39 (01), 97–106.
Cornot-Gandolphe, S., 2017. Underground Gas Storage in the World - 2017 Status. Cedigaz
Insights 22, 13 pp.
Credoz, A., Bildstein, O., Jullien, M., Raynal, J., Pétronin, J.-C., Lillo, M., Pozo, C., Geniaut,
G., 2009. Experimental and modeling study of geochemical reactivity between clayey
caprocks and CO2 in geological storage conditions. Energy Proced. 1 (1), 3445–3452.
doi:10.1016/j.egypro.2009.02.135.
Crotogino, F., 2011. Wasserstoffspeicherung im geologischen Untergrund: Stand der Technik
und Potential. KBB Underground Technologies GmbH Hannover. Fachkonferenz
Energiespeicher für Deutschland, 2011, Köln.
Crotogino, F., 2016. Larger scale hydrogen storage, in: Letcher, T.M. (Ed.), Storing energy.
With special reference to renewable energy sources, 1 ed. Elsevier, Amsterdam, Oxford,
Cambridge, pp. 411–429.
Crotogino, F., Donadei, S., Bünger, U., Landinger, H., 2010. Large-scale hydrogen
underground storage for securing future energy supplies, in: Stolten, D., Grube, T. (Eds.),
18th World Hydrogen Energy Conference 2010 - WHEC 2010. Proceedings.
Forschungszentrum IEF-3, Jülich.
Crotogino, F., Köckritz, V., Reinhold, S., 2006. Conceptual design of storage caverns for an
LNG receiving terminal in Europe. SMRI Spring Meeting Brussels (Technical conference
paper).
Cussler, E.L., 2009. Diffusion: Mass transfer in fluid systems, 3rd ed. Cambridge Univ. Press,
Cambridge, 631 pp.
Davydova-Charakhch`yan, I.A., Kuznetsova, V.G., Mityushina, L.L., Belyaev, S.S., 1992.
Methane-forming bacilli from oil fields of Tataria and western Siberia. Microbiology 61, 299–
305.
De Windt, L., Burnol, A., Montarnal, P., van der Lee, J., 2003. Intercomparison of reactive
transport models applied to UO2 oxidative dissolution and uranium migration. J. Contam.
Hydrol. (61), 303–312.
Dersch-Hansmann, M., Hug-Diegel, N., Wonik, T., 2010. Ein vollständiges Röt-Profil (Oberer
Buntsandstein) in Nordhessen - Lithostratigraphie, Sedimentfazies, Geochemie und
Geophysik der Kernbohrung Fürstenwald. Geol. Jahrb. Hessen 136, 65–107.
DeVries, K.L., Mellegard, K.D., Callahan, G.D., 2002. Salt damage criterion: Proof-of-concept
research. Prepared for the United States Department of Energy and National Energy
Technology Laboratory, Rapid City, 90 pp.
DVGW, 2013. Technische Regel-Arbeitsblatt DVGW G 260 (A), Gasbeschaffenheit. Deutscher
Verein des Gas und Wasserfaches.
Ebigbo, A., Golfier, F., Quintard, M., 2013. A coupled, pore-scale model for methanogenic
microbial activity in underground hydrogen storage. Adv. Water Res. 61, 74–85.
doi:10.1016/j.advwatres.2013.09.004.
Ehrlich, H.L., 1990. Geomicrobiology, 2nd ed. Dekker, New York, 646 pp.
Emberley, S., Hutcheon, I., Shevalier, M., Durocher, K., Mayer, B., Gunter, W.D., Perkins,
E.H., 2005. Monitoring of fluid–rock interaction and CO2 storage through produced fluid
sampling at the Weyburn CO2-injection enhanced oil recovery site, Saskatchewan,
Canada. Appl. Geochem. 20 (6), 1131–1157. doi:10.1016/j.apgeochem.2005.02.007.
111
Evans, D.J., 2008. An appraisal of underground gas storage technologies and incidents, for
the development of risk assessment methodology. British Geological Survey for the Health
and Safety Executive, HSE Books, 264 pp.
Ewing, R.P., Hu, Q., Liu, C., 2010. Scale dependence of intragranular porosity, tortuosity, and
diffusivity. Water Resour. Res. 46 (6). doi:10.1029/2009WR008183.
Farahani, A.F., Rahimi, G., Hosseini, S.E., 2015. Economical evaluation of different methods
of underground storage of natural gas. Res. J. Rec. Sci. 4(2), 114–119.
Federal Ministry for Education and Research, 2011. Förderinitative Energiespeicher.
https://www.bmbf.de/foerderungen/bekanntmachung.php?B=639. Accessed 12
September 2018.
Feist-Burkhardt, S., Götz, A.E., Szulc, J., Borkhataria, R., Geluk, M., Haas, J., Hormung, J.,
Jordan, P., Kempf, O., Michalik, J., Nawrocki, J., Reinhardt, L., Ricken, W., Röhling, H.-G.,
Rüffer, T., Török, A., Zühlke, R., 2008. Triassic, in: McCann, T. (Ed.), The geology of
Central Europe. Geological Society, London, pp. 749–822.
Flesch, S., Pudlo, D., Albrecht, D., Jacob, A., Enzmann, F., 2018. Hydrogen underground
storage—Petrographic and petrophysical variations in reservoir sandstones from
laboratory experiments under simulated reservoir conditions. Int. J. Hydrogen Energy 43
(45), 20822–20835. doi:10.1016/j.ijhydene.2018.09.112.
Fleury, M., Berne, P., Bachaud, P., 2009. Diffusion of dissolved CO2 in caprock. Energy
Proced. 1 (1), 3461–3468. doi:10.1016/j.egypro.2009.02.137.
Fleury, M., Pironon, J., Le Nindre, Y.M., Bildstein, O., Berne, P., Lagneau, V., Broseta, D.,
Pichery, T., Fillacier, S., Lescanne, M., Vidal, O., 2011. Evaluating sealing efficiency of
caprocks for CO2 storage: An overview of the Geocarbone Integrity Program and results.
Energy Proced. 4, 5227–5234. doi:10.1016/j.egypro.2011.02.501.
Fontenot, D.J., 1981. Degassing method and apparatus, US4417907 A.
Foster Associates Inc., 1995. Profile of underground natural gas storage facilities and market
hubs, Washington, DC, 256 pp.
Friend, D.G., Ely, J.F., Ingham, H., 1989. Thermophysical properties of methane. J. Phys.
Chem. Ref. Data 18 (2), 583–638. doi:10.1063/1.555828.
Fu, Y., 2014. Development and application of numerical modeling for evaluating and predicting
hydrogeochemical processes temporally and spatially evolving in petroleum reservoirs.
Ph.D. Thesis, TU Clausthal, Clausthal-Zellerfeld.
Fu, Y., van Berk, W., Schulz, H.-M., 2016. Hydrogen sulfide formation, fate, and behavior in
anhydrite-sealed carbonate gas reservoirs: A three-dimensional reactive mass transport
modeling approach. AAPG Bull. 100 (05), 843–865. doi:10.1306/12111514206.
Gaus, I., Audigane, P., André, L., Lions, J., Jacquemet, N., Durst, P., Czernichowski-Lauriol,
I., Azaroual, M., 2008. Geochemical and solute transport modelling for CO 2 storage, what
to expect from it? Int. J. Greenh. Gas Con. 2 (4), 605–625. doi:10.1016/j.ijggc.2008.02.011.
Gaus, I., Azaroual, M., Czernichowski-Lauriol, I., 2005. Reactive transport modelling of the
impact of CO2 injection on the clayey cap rock at Sleipner (North Sea). Chem. Geol. 217
(3–4), 319–337. doi:10.1016/j.chemgeo.2004.12.016.
Ghassemi, A., Pak, A., 2011. Pore scale study of permeability and tortuosity for flow through
particulate media using Lattice Boltzmann method. Int. J. Numer. Anal. Met. Geomech. 35
(8), 886–901. doi:10.1002/nag.932.
Gherardi, F., Xu, T., Pruess, K., 2007. Numerical modeling of self-limiting and self-enhancing
caprock alteration induced by CO2 storage in a depleted gas reservoir. Chem. Geol. 244
(1–2), 103–129. doi:10.1016/j.chemgeo.2007.06.009.
112
Ghofrani, J., 2016. Conceptualisation and software development of a simulation environment

for probabilistic safety assessments of radioactive waste repositories. Ph.D. Thesis, TU
Clausthal, Clausthal-Zellerfeld.
Grenthe, I., Plyasunov, A.V., Spahiu, K., 1997. Estimations of medium effects on
thermodynamic data, in: Grenthe, I., Allard, B. (Eds.), Modelling in aquatic chemistry.
OECD, Paris, pp. 325–426.
Gundogan, O., Mackay, E., Todd, A., 2011. Comparison of numerical codes for geochemical
modelling of CO2 storage in target sandstone reservoirs. Chem. Eng. Res. Des. (89), 1805–
1816.
Gusev, M.V., Mineeva, L.A., 1992. Microbiology, Moscow Lomonosov University, Moscow,
464 pp.
Haase, C., Dethlefsen, F., Ebert, M., Dahmke, A., 2013. Uncertainty in geochemical modelling
of CO2 and calcite dissolution in NaCl solutions due to different modelling codes and
thermodynamic databases. Appl. Geochem. 33, 306–317.
doi:10.1016/j.apgeochem.2013.03.001.
Haddenhorst, H.-G., 1989. Storage of natural gas in salt caverns, in: Tek, M.R. (Ed.),
Underground storage of natural gas. Kluwer Academic Publishers, pp. 177–195.
Hagemann, B., Rasoulzadeh, M., Panfilov, M., Ganzer, L., Reitenbach, V., 2016.
Hydrogenization of underground storage of natural gas. Comput. Geosci. 20 (3), 595–606.
doi:10.1007/s10596-015-9515-6.
Helgeson, H.C., 1967. Solution chemistry and metamorphism. Res. Geochem. 2, 362–404.
Helgeson, H.C., 1968. Evaluation of irreversible reactions in geochemical processes involving
minerals and aqueous solutions–I. Thermodynamic relations. Geochim. Cosmochim. Ac.
32 (8), 853–877.
Helgeson, H.C., 1979. Mass transfer among minerals and hydrothermal solutions, in: Barnes,
H.L. (Ed.), Geochemistry of hydrothermal ore deposits, 2nd ed. Wiley, New York, pp. 568–
610.
Helgeson, H.C., Brown, T.H., Nigrini, A., Jones, T.A., 1970. Calculation of mass transfer in
geochemical processes involving aqueous solutions. Geochim. Cosmochim. Ac. 34, 569–
592.
Helgeson, H.C., Garrels, R.M., Mackenzie, F.T., 1969. Evaluation of irreversible reactions in
geochemical processes involving minerals and aqueous solutions–II. Applications.
Geochim. Cosmochim. Ac. 33, 455–481.
Hellevang, H., Aagaard, P., Oelkers, E.H., Kvamme, B., 2005. Can dawsonite permanently
trap CO2? Environ. Sci. Technol. 39 (21), 8281–8287. doi:10.1021/es0504791.
Hemme, C., van Berk, W., 2015. One-dimensional reactive transport modeling of CO2 storage
systems - change in cap rock porosity triggered by pressure and temperature dependent
CO2-water-rock interactions. AGU Fall Meeting, San Francisco.
Hemme, C., van Berk, W., 2017a. Change in cap rock porosity triggered by pressure and
temperature dependent CO2–water–rock interactions in CO2 storage systems. Petroleum
3 (1), 96–108. doi:10.1016/j.petlm.2016.11.010.
Hemme, C., van Berk, W., 2017b. H2S generation and release in salt cavern gas storage.
AAPG 2017 Annual Convention & Exhibition in Houston, Texas.
Henkel, S., Pudlo, D., Enzmann, F., Reitenbach, V., Albrecht, D., Ganzer, L., Gaupp, R., 2016.
X-ray CT analyses, models and numerical simulations: A comparison with petrophysical
analyses in an experimental CO2 study. Solid Earth 7 (3), 917–927. doi:10.5194/se-7-917-
2016.
113
Henkel, S., Pudlo, D., Gaupp, R., 2013. Research sites of the H2STORE project and the
relevance of lithological variations for hydrogen storage at depths. Energy Proced. 40, 25–
33. doi:10.1016/j.egypro.2013.08.004.
Henkel, S., Pudlo, D., Werner, L., Enzmann, F., Reitenbach, V., Albrecht, D., Würdemann, H.,
Heister, K., Ganzer, L., Gaupp, R., 2014. Mineral reactions in the geological underground
induced by H2 and CO2 injections. Energy Proced. 63, 8026–8035.
doi:10.1016/j.egypro.2014.11.839.
Herrera, L., Hernández, J., Bravo, L., Romo, L., Vera, L., 1997. Biological process for sulfate
and metals abatement from mine effluents. Environ. Toxicol. Water Qual. 12 (2), 101–107.
Holloway, S., 1997. An overview of the underground disposal of carbon dioxide. Energ.
Convers. Manage. 38, 193–198.
Hsieh, P.A., Winston, R.B., 2002. User´s guide to Model Viewer, a program for three-
dimensional visualization of groundwater model results. U.S. Geological Survey Open-File
Report 02–106, 18 pp.
IPCC, 2005. Special report on carbon dioxide capture and storage. Prepared by Working
Group III of the International Panel of Climate Change. Cambridge University Press,
Cambridge, United Kingdom/New York, NY, USA.
Jacops, E., Volckaert, G., Maes, N., Weetjens, E., Govaerts, J., 2013. Determination of gas
diffusion coefficients in saturated porous media: He and CH4 diffusion in Boom Clay. Appl.
Clay Sci. 83–84, 217–223. doi:10.1016/j.clay.2013.08.047.
Jakobsen, R., Postma, D., 1994. In situ rates of sulfate reduction in an aquifer (Rømø,
Denmark) and implications for the reactivity of organic matter. Geology 22 (12), 1101–
1106. doi:10.1130/0091-7613(1994)022<1103:ISROSR>2.3.CO;2.
Jorgensen, B.B., Isaksen, M.F., Jannasch, H.W., 1992. Bacterial sulfate reduction above
100 °C in deep-sea hydrothermal vent sediments. Science 258 (5089), 1756–1757.
doi:10.1126/science.258.5089.1756.
Kallmeyer, J., Boetius, A., 2004. Effects of temperature and pressure on sulfate reduction and
anaerobic oxidation of methane in hydrothermal sediments of Guaymas Basin. Appl.
Environ. Microbiol. 70 (2), 1231–1233. doi:10.1128/AEM.70.2.1231-1233.2004.
Kalyuzhnyi, S.V., Fedorovich, V.V., 1998. Mathematical modelling of competition between
sulphate reduction and methanogenesis in anaerobic reactors. Biores. Technol. 65 (3),
227–242. doi:10.1016/S0960-8524(98)00019-4.
Kemp, S.J., Bouch, J., Murphy, H.A., 2001. Mineralogical characterisation of the Nordland
Shale, UK Quadrant 16, northern North Sea, 52 pp.
Kirkner, D.J., Reeves, H., 1988. Multicomponent mass transport with homogeneous and
heterogeneous chemical reactions: Effect of the chemistry on the choice of numerical
algorithm: 1. Theory. Water Resour. Res. 24 (10), 1719–1729.
doi:10.1029/WR024i010p01719.
Kjeldsen, K.U., Loy, A., Jakobsen, T.F., Thomsen, T.R., Wagner, M., Ingvorsen, K., 2007.
Diversity of sulfate-reducing bacteria from an extreme hypersaline sediment, Great Salt
Lake (Utah). FEMS Microbiol. Ecol. 60 (2), 287–298. doi:10.1111/j.1574-
6941.2007.00288.x.
Kleinitz, W., Böhling, E., 2005. Underground gas storage in porous media - operating
experience with bacteria on gas quality. Society of Petroleum Engineers SPE 94248 (SPE
Europe/EAGE Annual Conference Madrid 13-16 June 2005).
Konikow, L.F., Bredehoeft, J.D., 1992. Ground-water models cannot be validated. Adv. Water
Res. 15 (1), 75–83. doi:10.1016/0309-1708(92)90033-X.
114
Krooss, B., 2008. Evaluation of database on gas migration through clayey host rocks. Belgian
National Agency for Radioactive Waste and Enriched Fissile Material (ONDRAF-NIRAS),
RWTH Aachen.
Krumbein, W.C., Graybill, F.A., 1965. An introduction to statistical models in geology. McGraw-
Hill, New York, 475 pp.
Kyle, J.R., Posey, H.H., 1991. Chapter 5 Halokinesis, cap rock development, and salt dome
mineral resources, in: Melvin, J.L. (Ed.), Evaporites, petroleum and mineral resources, vol.
50. Elsevier, Amsterdam, Oxford, New York, Tokyo, pp. 413–474.
Laesecke, A., Muzny, C.D., 2017. Reference correlation for the viscosity of carbon dioxide. J.
Phys. Chem. Ref. Data 46. doi:10.1063/1.4977429.
Larin, N., Zgonnik, V., Rodina, S., Deville, E., Prinzhofer, A., Larin, V.N., 2015. Natural
molecular hydrogen seepage associated with surficial, rounded depressions on the
European Craton in Russia. Nat. Resour. Res. 24 (3), 369–383. doi:10.1007/s11053-014-
9257-5.
LBEG, 2016. Untertage-Gasspeicherung in Deutschland, Landesamt für Bergbau, Energie
und Geologie. Erdöl Erdgas Kohle 132 (11), 409–417.
Leung, D.Y.C., Caramanna, G., Maroto-Valer, M.M., 2014. An overview of current status of
carbon dioxide capture and storage technologies. Renew. Sust. Energ. Rev. 39, 426–443.
doi:10.1016/j.rser.2014.07.093.
Li, X., 2015. Entwicklung der Softwareplattform RESUS: Repository Simulation, Uncertainty
Propagation and Sensitivity Analysis. Ph.D. Thesis, TU Clausthal, Clausthal-Zellerfeld.
Lindenberg, E., Wessel-Berg, D., 1997. Vertical convection in an aquifer column under a gas
cap of CO2. Energ. Convers. Manage. (38), 229–234.
Liu, H., Hou, Z., Were, P., Gou, Y., Xiong, L., Sun, X., 2015. Modelling CO 2-brine-rock
interactions in the Upper Paleozoic formations of Ordos Basin used for CO2 sequestration.
Environ. Earth Sci. 73 (5), 2205–2222. doi:10.1007/s12665-014-3571-4.
Lord, A.S., Kobos, P.H., Borns, D.J., 2014. Geologic storage of hydrogen: Scaling up to meet
city transportation demands. Int. J. Hydrogen Energy 39 (28), 15570–15582.
doi:10.1016/j.iJhydene.2014.07.121.
Machel, H.G., 2001. Bacterial and thermochemical sulfate reduction in diagenetic settings —
old and new insights. Sedimentary Geol. 140 (1–2), 143–175. doi:10.1016/S0037-
0738(00)00176-7.
Magot, M., Ollivier, B., Patel, B.K.C., 2000. Microbiology of petroleum reservoirs. Antonie van
Leeuwenhoek 77 (2), 103–116. doi:10.1023/A:1002434330514.
Marini, L., 2007. Geological sequestration of carbon dioxide: Thermodynamics, kinetics, and
reaction path modeling. Elsevier, Amsterdam, 453 pp.
McCarty, R.D., Hord, J., Roder, H.M., 1981. Selected properties of hydrogen (engineering
design data). National Bureau of Standards Monograph 168 - U.S. Department of
Commerce, Boulder, CO, 523 pp.
Merkel, B.J., Planer-Friedrich, B., 2002. Grundwasserchemie: Praxisorientierter Leitfaden zur
numerischen Modellierung von Beschaffenheit, Kontamination und Sanierung aquatischer
Systeme. Springer, Berlin, 219 pp.
Meulepas, R.J.W., Stams, A.J.M., Lens, P.N.L., 2010. Biotechnological aspects of sulfate
reduction with methane as electron donor. Rev. Environ. Sci. Bio./Techn. 9 (1), 59–78.
doi:10.1007/s11157-010-9193-8.
Mohd Amin, S., Weiss, D.J., Blunt, M.J., 2014. Reactive transport modelling of geologic CO 2
sequestration in saline aquifers: The influence of pure CO2 and of mixtures of CO2 with CH4
115
on the sealing capacity of cap rock at 37°C and 100bar. Chem. Geol. 367, 39–50.
doi:10.1016/j.chemgeo.2014.01.002.
NLWKN, 2015. Grundwassermessstelle Brauel I Stade. Niedersächsiches Ministerium für
Umwelt, Energie und Klimaschutz. www.umweltkarten-niedersachsen.de. Accessed 5
January 2017.
Nowack, B., Mayer, K.U., Oswald, S.E., van Beinum, W., Appelo, C.A.J., Jacques, D.,
Seuntjens, P., Gérard, F., Jaillard, B., Schnepf, A., Roose, T., 2006. Verification and
intercomparison of reactive transport codes to describe root-uptake. Plant Soil (285), 305–
321.
Palandri, J.L., Kharaka, Y.K., 2004. A compilation of rate parameters of water-mineral
interaction kinetics for application to geochemical modeling, Menlo Park, California, 64 pp.
Panfilov, M., 2010. Underground storage of hydrogen: In situ self-organisation and methane
generation. Transp. Porous Med. 85 (3), 841–865. doi:10.1007/s11242-010-9595-7.
Panfilov, M., 2016. Underground and pipeline hydrogen storage, in: Gupta, R.B., Basile, A.,
Veziroğlu, T.N. (Eds.), Compendium of hydrogen energy. Volume 2: Hydrogen storage,
distribution and infrastructure. Elsevier, Cambridge, Waltham, Kidlington, pp. 91–115.
Panfilov, M., Gravier, G., Fillacier, S., 2006. Underground storage of H2 and H2-CO2-CH4
mixtures, in: ECMOR X - 10th European Conference on the Mathematics of Oil Recovery,
Amsterdam, Netherlands. 04.09.2006 - 07.09.2006. EAGE Publications.
Pape, H., Clauser, C., Iffland, J., 2000. Variation of permeability with porosity in sandstone
diagenesis interpreted with a fractal pore space model. Pure Appl. Geophys. 157 (4), 603–
619.
Parkhurst, D.L., Appelo, C.A.J., 1999. Users guide to PHREEQC Version 2— A computer
program for speciation, batch-reaction, one-dimensional transport and inverse
geochemical calculations.: U.S. Geological Survey, Denver, Colorado, 326 pp.
Parkhurst, D.L., Appelo, C.A.J., 2013. Description of input for PHREEQC Version 3—A
computer program for speciation, batch-reaction, one-dimensional transport, and inverse
geochemical calculations: U.S. Geological Survey Techniques and Methods 6-A43,
519 pp.
Parkhurst, D.L., Kipp, K.L., Charlton, S.R., 2010. PHAST Version 2—A program for simulating
groundwater flow, solute transport, and multicomponent geochemical reactions: U.S.
Geological Survey Techniques and Methods 6-A35, 235 pp.
Parkhurst, D.L., Thorstenson, D.C., Plummer, N., 1980. PHREEQE—A computer program for
geochemical calculations: U.S. Geological Survey Water-Resources Investigations Report
80–96, 195 pp.
Pietzner, K., 2012. Folgeprojekt CCS-Kommunikation - Multivariate Analysen der
Einflussfaktoren auf die Akzeptanz von CCS. Wuppertal Institut für Klima, Umwelt, Energie
GmbH, 70 pp.
Pitzer, K.S., 1973. Thermodynamics of electrolytes. I. Theoretical basis and general equations.
J. Phys. Chem. 77 (2), 268–277. doi:10.1021/j100621a026.
Ponomarev-Stepnoi, N.N., 2004. Nuclear-hydrogen power. Atomic Energy 96 (6), 375–385.
Postgate, J.R., 1979. The sulphate-reducing bacteria. Cambridge University Press,
Cambridge, 151 pp.
Postgate, J.R., 1984. The sulphate-reducing bacteria, 2nd ed. Cambridge University Press,
Cambridge, 208 pp.
116
Pruess, K., 2013. Numerical simulation of CO2 leakage from a geologic disposal reservoir,
including transitions from super- to subcritical conditions, and boiling of liquid CO2. SPE
Journal 9 (02), 237–248. doi:10.2118/86098-PA.
Pudlo, D., Ganzer, L., Henkel, S., Kühn, M., Liebscher, A., De Lucia, M., Panfilov, M., Pilz, P.,
Reitenbach, V., Albrecht, D., Würdemann, H., Gaupp, R., 2010. The H2STORE Project:
Hydrogen underground storage – A feasible way in storing electrical power in geological
media?, in: Hou, M.Z., Xie, H., Yoon, J.S. (Eds.), Underground storage of CO2 and energy.
CRC Press, Taylor & Francis Group, Boca Raton, Fla., pp. 395–412.
Reichl, A., Schneider, G., Schliepdiek, T., Reimuth, O., 2015. Carbon capture and storage for
enhanced oil recovery: integration and optimization of a post-combustion CO2-capture
facility at a power plant in Abu Dhabi. Oil Gas Facili. 05 (4), 37–46. doi:10.2118/171692-
PA.
Reitenbach, V., Ganzer, L., Albrecht, D., Hagemann, B., 2015. Influence of added hydrogen
on underground gas storage: A review of key issues. Environ. Earth Sci. 73 (11), 6927–
6937. doi:10.1007/s12665-015-4176-2.
Réveillère, B., Azimi, B., Arnold, C., 2016. Prevention of stored gas humidification: lessons
learnt and review of possibilities. Solution Mining Research Institute Spring 2016 Technical
Conference, Galveston, Texas, USA.
Robinson, J.A., Tiedje, J.M., 1984. Competition between sulfate-reducing and methanogenic
bacteria for H2 under resting and growing conditions. Arch. Microbiol. 137 (1), 26–32.
doi:10.1007/BF00425803.
Rochelle, C.A., Moore, Y.A., 2002. The solubility of supercritical CO2 into pure water and
synthetic Utsira pore water: British Geological Survey report CR/02/052N, 28 pp.
Rutqvist, J., Birkholzer, J., Cappa, F., Tsang, C.-F., 2007. Estimating maximum sustainable
injection pressure during geological sequestration of CO2 using coupled fluid flow and
geomechanical fault-slip analysis. Energ. Convers. Manage. 48 (6), 1798–1807.
doi:10.1016/j.enconman.2007.01.021.
Salimi, H., Wolf, K.-H., Bruining, J., 2012. The influence of capillary pressure on the phase
equilibrium of the CO2–water system: Application to carbon sequestration combined with
geothermal energy. Int. J. Greenh. Gas Con. 11, 47–66. doi:10.1016/j.ijggc.2012.09.015.
Sato, M., Sutton, A.J., McGee, K.A., Russell-Robinson, S., 1986. Monitoring of hydrogen along
the San Andreas and Calaveras faults in central California in 1980-1984. J. Geophys. Res.
Solid Earth 91 (B12), 12315–12326. doi:10.1029/JB091iB12p12315.
Schäfer, D., 2004. Numerische Modellierung reaktiver Prozesse organischer Kontaminanten
in Grundwasserleitern: Habilitationsschrift. Institut für Geowissenschaften, Christian-
Albrechts-Universität zu Kiel.
Schilling, F., Ossing, F., Würdemann, H., CO2SINK Team, 2009. Public acceptance for
geological CO2-Storage. Geophysical Research Abstracts. Vol. 11, EGU2009-13801-2,
2009.
Schneider, R., Crotogino, F., 2010. New developments in Europe´s salt cavern projects,
in: Hou, M.Z., Xie, H., Yoon, J.S. (Eds.), Underground storage of CO 2 and energy. CRC
Press, Taylor & Francis Group, Boca Raton, Fla., pp. 199–205.
Schreiber, B.C., Bąbel, M., 2007. Evaporites through space and time. Geol. Soc. Sp. 285 (1).
doi:10.1144/SP285.
Schulze, F., Keimeyer, F., Schöne, R., Westphal, I., Janssen, G., Bartel, S., Seiffert, S., 2015.
Unterirdische Raumplanung – Vorschläge des Umweltschutzes zur Verbesserung der
über- und untertägigen Informationsgrundlagen, zur Ausgestaltung des
Planungsinstrumentariums und zur nachhaltigen Lösung von Nutzungskonflikten:
117
Teilvorhaben 2: planerische und rechtliche Aspekte. Umweltbundesamt, Dessau-Roßlau,

324 pp.
Shen, L., Chen, Z., 2007. Critical review of the impact of tortuosity on diffusion. Chem. Eng.
Sci. 62 (14), 3748–3755. doi:10.1016/j.ces.2007.03.041.
Šmigáň, P., Greksák, M., Kozánková, J., Buzek, F., Onderka, V., Wolf, I., 1990. Methanogenic
bacteria as a key factor involved in changes of town gas stored in an underground reservoir.
FEMS Microbiol. Lett. 73 (3), 221–224. doi:10.1111/j.1574-6968.1990.tb03944.x.
Somerday, B.P., Gangloff, R.P. (Eds.), 2012. Mechanisms, modelling and future
developments. Woodhead Pub, Cambridge, U.K, Phildelphia, Pa, 500 pp.
Soyk, D., 2015. Diagenesis and reservoir quality of the Lower and Middle Buntsandstein
(Lower Triassic), SW Germany. Ph.D. Thesis, University of Heidelberg.
Steefel, C.I., DePaolo, D.J., Lichtner, P.C., 2005. Reactive transport modeling: An essential
tool and a new research approach for the Earth sciences. Earth Planet. Sc. Lett. 240 (3–
4), 539–558. doi:10.1016/j.epsl.2005.09.017.
Stone, H.B.J., Veldhuis, I., Richardson, R.N., 2009. Underground hydrogen storage in the UK.
Geol. Soc. Sp. 313 (1), 217–226. doi:10.1144/SP313.13.
Stumm, W., Morgan, J.J., 1981. Aquatic Chemistry: An introduction emphasizing chemical
equilibria in natural waters, 2nd ed. John Wiley & Sons, Inc., New York, 780 pp.
Szabó, Z., Hellevang, H., Király, C., Sendula, E., Kónya, P., Falus, G., Török, S., Szabó, C.,
2016. Experimental-modelling geochemical study of potential CCS caprocks in brine and
CO2-saturated brine. Int. J. Greenh. Gas Con. 44, 262–275.
doi:10.1016/j.ijggc.2015.11.027.
Taylor, J.B., Alderson, J.E.A., Kalyanam, K.M., Lyle, A.B., Phillips, L.A., 1986. Technical and
economic assessment of methods for the storage of large quantities of hydrogen. Int. J.
Hydrogen Energy 11 (1), 5–22.
Thorstenson, D.C., Parkhurst, D.L., 2002. Calculation of individual isotope equilibrium
constants for implementation in geochemical models: U.S. Geological Survey Water-
Resources Investigations Report 02-4172, 129 pp.
Thorstenson, D.C., Parkhurst, D.L., 2004. Calculation of individual isotope equilibrium
constants for geochemical reactions. Geochim. Cosmochim. Ac. 68 (11), 2449–2465.
doi:10.1016/j.gca.2003.11.027.
Tian, H., Pan, F., Xu, T., McPherson, B.J., Yue, G., Mandalaparty, P., 2014. Impacts of
hydrological heterogeneities on caprock mineral alteration and containment of CO2 in
geological storage sites. Int. J. Greenh. Gas Con. 24, 30–42.
doi:10.1016/j.ijggc.2014.02.018.
Tian, H., Xu, T., Li, Y., Yang, Z., Wang, F., 2015. Evolution of sealing efficiency for CO 2
geological storage due to mineral alteration within a hydrogeologically heterogeneous
caprock. Appl. Geochem. 63, 380–397. doi:10.1016/j.apgeochem.2015.10.002.
Timmers, P.H., Suarez-Zuluaga, D.A., van Rossem, M., Diender, M., Stams, A.J., Plugge,
C.M., 2016. Anaerobic oxidation of methane associated with sulfate reduction in a natural
freshwater gas source. ISME J 10 (6), 1400–1412. doi:10.1038/ismej.2015.213.
Timmers, P.H.A., Gieteling, J., Widjaja-Greefkes, H.C.A., Plugge, C.M., Stams, A.J.M., Lens,
P.N.L., Meulepas, R.J.W., 2015. Growth of anaerobic methane-oxidizing archaea and
sulfate-reducing bacteria in a high-pressure membrane capsule bioreactor. Appl. Environ.
Microbiol. 81 (4), 1286–1296. doi:10.1128/AEM.03255-14.
Truche, L., Berger, G., Destrigneville, C., Guillaume, D., Giffaut, E., 2010. Kinetics of pyrite to
pyrrhotite reduction by hydrogen in calcite buffered solutions between 90 and 180°C:
118
Implications for nuclear waste disposal. Geochim. Cosmochim. Ac. 74 (10), 2894–2914.
doi:10.1016/j.gca.2010.02.027.
Truche, L., Jodin-Caumon, M.-C., Lerouge, C., Berger, G., Mosser-Ruck, R., Giffaut, E.,
Michau, N., 2013. Sulphide mineral reactions in clay-rich rock induced by high hydrogen
pressure. Application to disturbed or natural settings up to 250 °C and 30 bar. Chem. Geol.
351, 217–228. doi:10.1016/j.chemgeo.2013.05.025.
Tuttle, J.H., Dugan, P.R., Randles, C.I., 1969. Microbial sulfate reduction and its potential utility
as an acid mine water pollution abatement procedure. Appl. Microbiol. 17 (2), 297–302.
U.S. Environmental Protection Agency, 1999. MINTEQA2/PRODEFA2, A geochemical
assessment model for environmental systems: User manual supplement for version 4.0.
National Exposure Research Laboratory, Ecosystems Research Division, Athens, Georgia,
76 pp.
van Berk, W., Fu, Y., Schulz, H.-M., 2015. Creation of pre-oil-charging porosity by migration of
source-rock-derived corrosive fluids through carbonate reservoirs: one-dimensional
reactive mass transport modelling. Petrol. Geosci. (21), 35–42. doi:10.1144/petgeo2014-
065.
van Berk, W., Hansen, C., 2006. Hydrogeochemische Stoffflussmodelle: Leitfaden zur
Modellierung der Beschaffenheitsentwicklung von Grund- und Rohwässern. Springer-
Verlag, Berlin, Heidelberg, 226 pp.
Van Pham, T.H., Aagaard, P., Hellevang, H., 2012. On the potential for CO2 mineral storage
in continental flood basalts - PHREEQC batch- and 1D diffusion-reaction simulations.
Geochem. T. 13 (5).
Vaughan, P.J., 1987. Analysis of permeability reduction during flow of heated, aqueous fluid
through westerly granite, in: Tsang, C.-F. (Ed.), Coupled processes associated with nuclear
waste repositories. Acad. Pr, Orlando, Fla., pp. 529–539.
Verdon, J.P., Stork, A.L., 2016. Carbon capture and storage, geomechanics and induced
seismic activity. J. Rock Mechan. Geotechnol. Eng. 8 (6), 928–935.
doi:10.1016/j.jrmge.2016.06.004.
Verma, A., Pruess, K., 1988. Thermohydrological conditions and silica redistribution near high-
level nuclear wastes emplaced in saturated geological formations. J. Geophys. Res. 93,
1159–1173. doi:10.1029/JB093iB02p01159.
Vinograd, J.R., McBain, J.W., 1941. Diffusion of electrolytes and of the ions in their mixtures.
J. Am. Chem. Soc. 63 (7), 2008–2015. doi:10.1021/ja01852a063.
Vögele, S., Rübbelke, D., Mayer, P., Kuckshinrichs, W., 2018. Germany’s “No” to carbon
capture and storage: Just a question of lacking acceptance? Appl. Energ. 214, 205–218.
doi:10.1016/j.apenergy.2018.01.077.
Wakita, H., Nakamura, Y., Kita, I., Fujii, N., Notsu, K., 1980. Hydrogen release: New indicator
of fault activity. Science 210 (4466), 188–190. doi:10.1126/science.210.4466.188.
Wang, T., Yang, C., Yan, X., Daemen, J.J.K., 2015. Allowable pillar width for bedded rock salt
caverns gas storage. J. Petrol. Sci. Eng. 127, 433–444. doi:10.1016/j.petrol.2015.01.040.
Ware, R.H., Roecken, C., Wyss, M., 1985. The detection and interpretation of hydrogen in fault
gases. PAGEOPH 122 (2–4), 392–402. doi:10.1007/BF00874607.
Whitman, W.B., Bowen, T.L., Boone, D.R., 2006. The methanogenic bacteria, in: Dworkin, M.,
Falkow, S., Rosenberg, E., Schleifer, K.-H., Stackebrandt, E. (Eds.), The Prokaryotes. A
handbook on the biology of bacteria. Springer, New York, pp. 165–207.
119
Wolery, T.J., 1992. EQ3/6, A software package for geochemical modeling of aqueous systems:
Package overview and installation guide (Version 7.0). Lawrence Livermore National
Laboratory, Livermore, California, 66 pp.
Wollenweber, J., Alles, S., Busch, A., Krooss, B.M., Stanjek, H., Littke, R., 2010. Experimental
investigation of the CO2 sealing efficiency of caprocks. Int. J. Greenh. Gas Con. 4 (2), 231–
241. doi:10.1016/j.ijggc.2010.01.003.
Yang, C., Jing, W., Daemen, J.J.K., Zhang, G., Du, C., 2013. Analysis of major risks associated
with hydrocarbon storage caverns in bedded salt rock. Reliab. Eng. Syst. Safe. 113, 94–
111. doi:10.1016/j.ress.2012.12.017.
Yekta, A.E., Pichavant, M., Audigane, P., 2018. Evaluation of geochemical reactivity of
hydrogen in sandstone: Application to geological storage. Appl. Geochem. 95, 182–194.
doi:10.1016/j.apgeochem.2018.05.021.
Zander-Schiebenhöfer, D., Donadei, S., Schneider, G.-S., 2014. Mögliche
Nutzungskonkurrenzen bei Energiespeichern im Untergrund. Akademie für
Geowissenschaften und Geotechnologien e.V. 30, 37–42. doi:10.7788/figurationen-2014-
0104.
Zgonnik, V., Beaumont, V., Deville, E., Larin, N., Pillot, D., Farrell, K.M., 2015. Evidence for
natural molecular hydrogen seepage associated with Carolina bays (surficial, ovoid
depressions on the Atlantic Coastal Plain, Province of the USA). Prog. Earth Planet. Sci. 2
(31). doi:10.1186/s40645-015-0062-5.
Zimmermann, G., Jacob, G., Lautenschläger, K.-H., Richter, G., Teschner, J., 1986.
Wissensspeicher, 7th ed. Dt. Verl. für Grundstoffindustrie, Leipzig, 71 pp.
ZoBell, C.E., 1958. Ecology of sulfate reducing bacteria. Prod. Monthly 22 (7), 12–29.
120

Reacciones de Equilibrio

Hochgeladen von

Copyright:

Verfügbare Formate

Reacciones de Equilibrio

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Reacciones de Equilibrio

Hochgeladen von

Copyright:

Verfügbare Formate

Storage of Gases in Deep Geological Structures: Spatial and

Temporal Hydrogeochemical Processes Evaluated and Predicted

to be awarded the degree

M. Sc. Christina Hemme

approved by the Faculty of Energy and Management Science,

Date of oral examination

Christina Hemme Clausthal-Zellerfeld, December 2018

† Prof. Dr. Wolfgang van Berk passed away on 29.07.2018

Diese hydrogeochemischen Prozesse werden in der vorliegenden Arbeit identifiziert, mit

In den Modellen werden hydrogeochemische Prozesse wie die Gas-Wasser-Gesteins

Die Modellierungsergebnisse erlauben es, Bedingungen zu ermitteln, die eine sichere

In this study, these different hydrogeochemical processes will be identified, quantified by 1D

In these models, hydrogeochemical processes, like gas–water–rock interactions under given

Oral Presentations at Conferences

Poster Presentations at Conferences

1.1 Review of Underground Gas Storage—State of Research ..................................................... 1

2 Carbon Capture and Storage (CCS) .......................................................................................... 18

3 Salt Cavern Gas Storage (SCGS) ............................................................................................... 47

4 Underground Hydrogen Storage (UHS) .................................................................................... 69

Hydrogeochemical Modeling to Identify Potential Risks of Underground Hydrogen Storage in

5 Final Discussion .......................................................................................................................... 97

5.1 General Model Similarities and Differences .......................................................................... 98

6 Final Conclusion and Outlook.................................................................................................. 104

List of Figures .................................................................................................................................. 105

(i) carbon capture and storage (CCS),

(ii) salt cavern gas storage (SCGS),

(iii) underground hydrogen storage (UHS).

These technical means of storing gases in the underground trigger a series of

1.1 Review of Underground Gas Storage—State of Research

1.1.1 Underground Gas Storage—Risks and Potentials

1.1.2 Hydrogeochemical Processes—Mineral Dissolution and Precipitation

1.1.3 Pressure and Temperature Conditions

1.1.4 Microbial Processes in Underground Gas Storage

1.1.5 Chemical and Physical Properties of the Stored Gases

Table 1.1 Chemical and physical properties of the stored gases.

Carbon dioxide Methane Hydrogen

(e): Laesecke and Muzny (2017)

Considered as a whole, the modeling of hydrogeochemical processes induced by gas storage

1.2 Aims of This Study

To answer this question, the hydrogeochemical processes, including gas–water–rock

2) Is H2S(g) generation a risk when storing natural gas in salt caverns?

3) Which factors influence the loss of hydrogen in depleted gas storage?

1.3 Hydrogeochemical Reactive Mass Transport Modeling

1.3.1 Basics of Hydrogeochemical Modeling

For numerical modeling of gases stored in deep geological formations, chemical-

1.3.2 Thermodynamic Equilibrium Reactions

where K is the thermodynamic equilibrium constant (-), a, b, c, and d are stoichiometric

𝐿 = 𝐼𝐴𝑃 = {𝐴} ∙ {𝐵}

1.3.3 Reaction Kinetics

While the thermodynamic-equilibrium reaction assumes a time-independent stable state for a

1.3.4 Reactive Mass Transport

Dispersion describes the changes in aqueous species concentration by advective mass

1.3.5 Modeling Tools

PHREEQC (= pH-Redox-Equilibrium written in the C programming language) is a computer

1.4 Hydrogeochemical Modeling Approaches—Similarities and Differences

• Fixed effective diffusion

• Underground gas storage

In all three subtopics of this study—CCS, SCGS and UHS—hydrogeochemical processes in

Table 1.2 Underground gas storage models and their parameters.

CCS SCGS UHS