See discussions, stats, and author profiles for this publication at: [Link]

net/publication/283756800

Carbon capture and storage (CCS)

Article · January 2015

DOI: 10.1007/978-3-642-34132-8_24

CITATIONS READS

2 268

1 author:

Helge Hellevang
University of Oslo
111 PUBLICATIONS   1,057 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Petroleum system analysis; characterisation of unconventional reservoirs; stratigraphic modeling View project

"FME-SUCCESS" - Subsurface CO2 Storage-Critical Element and Superior Strategy View project

All content following this page was uploaded by Helge Hellevang on 01 December 2015.

The user has requested enhancement of the downloaded file.

Chapter 24

Carbon Capture and Storage (CCS)

Helge Hellevang

24.1 Introduction that the reservoir has a seal (cap rock), and sufficient
injectivity (porosity and permeability) to store large
Carbon Capture and Storage (CCS) involves capturing volumes of CO2. These issues are discussed below.
CO2 and then storing it so that it is not emitted to the
atmosphere. The background for CCS is a concern that
increasing atmospheric CO2 concentrations will cause 24.2 CO2 Capture
global climate changes, ocean acidification and a sea
level rise, with dramatic negative consequences for Capture refers to the separation of CO2 from natural
large populations (IPCC 2013). gas or flue (exhaust) gas where the concentration of
Although capture technologies in principle can be CO2 may be as low as 3–4%. In CCS, capture is
implemented for any natural gas or flue gas sources, followed by transport and finally underground storage.
costs put practical constrains on what kind of There is considerable research activity to improve the
operations that can implement CCS. Generally, any efficiency of present-day capture methods and to
large point source of CO2 (>1 Mt CO2/year), such as invent new and better technologies, the main motiva-
coal- or gas-fired power plants, natural gas producers tion being to reduce the cost of the capture process.
and cement factories, has sufficient CO2 production to Capture, whether it is by pre- or post-combustion,
justify installation of CCS technology. The cost may requires significant amounts of energy. For a coal-
however be rather prohibitive. fired power plant, an extra 20–25% of energy is typi-
CO2 may be captured from major point sources. If cally required for capturing and compressing the CO2.
several point sources are in the same geographical There are already several technologies available for
region, then captured CO2 should be transported to a capturing CO2. A review of the advances in these
common hub before being injected into a geological technologies is provided by Figueroa et al. (2008)
formation (Fig. 24.1). This lowers the cost of and only a brief overview will be given here. In post-
implementing the pipeline network and thereby the combustion capture, CO2 is removed from flue gases
total cost of the CCS operation. The geological stor- formed from the combustion of primarily coal. The
age units can be of various kinds, such as saline coal is burned in an air atmosphere, in contrast to oxy-
aquifers, coal seams and abandoned oil and gas combustion that utilizes an oxygen atmosphere with
fields. The requirement is that CO2 is stored as a little nitrogen. In pre-combustion capture, CO2 is
dense supercritical phase (scCO2), that typically removed from the fuel before combustion. This tech-
requires more than 700–800 m burial depth, and nology is applied to gasification plants. In separating
CO2 from the gas streams, various technologies are
being used, such as gas-phase separation, sorption to
liquids or solids, and hybrid processes such as sorp-
H. Hellevang (*)
Department of Geosciences, University of Oslo, Oslo, Norway tion/membrane systems. Each of these technologies
e-mail: [Link]@[Link] has their strengths and weaknesses, and the choice of

K. Bjørlykke (ed.), Petroleum Geoscience: From Sedimentary Environments to Rock Physics, 591
DOI 10.1007/978-3-642-34132-8_24, # Springer-Verlag Berlin Heidelberg 2015
592 H. Hellevang

Fig. 24.1 CO2 can be captured at major point emitters, such as coal-fired power plants, and piped through a network of pipelines to
major storage units, such as the Utsira Sand

technology must be based on the processing plant, and several minor CO2 point sources (<1 Mt/year) are in
whether the technology will be implemented on a geographically limited area, the sources (flue gas or
existing plants (retrofit), or integrated into new plants. natural gas) can be piped to a common capture facility.
In addition to the size of the facilities, geographical
location plays a role. Because of the costs and risks
associated with pipelines, the distance to the storage
24.3 What Is Required to Store CO2? reservoir should be as short as possible.

24.3.1 CO2 Point Sources

24.3.2 The Storage Reservoir: Injectivity
Size matters in CCS. A large part of the cost of CCS is and CO2 Storage Capacity
related to the capture facilities and the transport net-
work to the CO2 storage site (steel pipelines and The requirements of a CO2 storage deposit are to a large
compressors). The possibility of capturing large extent the same as those for petroleum reservoirs. Injec-
amounts of CO2 at one plant lowers the cost of CCS tion requires that the pressure of the injection well is
($/ton stored CO2). This is critical in order to imple- larger than the reservoir pressure. The amount of over-
ment CCS globally at a scale required to meet the pressure required is given by the injectivity of the
IPCC 2
C target. CCS therefore requires point sources reservoir, which is determined by several factors, such
of CO2, which can be cement factories, coal-fired as the size of the reservoir unit and its hydraulic con-
power plants, or production facilities for natural gas ductivity. The size of the reservoir unit refers to the
(all with a potential of about 1–3 Mt CO2/year). If volume that can absorb the pressure build-up resulting
24 Carbon Capture and Storage (CCS) 593

from the CO2 injection and formation water The requirement of a seal or cap-rock is therefore
mobilisation. For extensive reservoir units such as the analogous to petroleum reservoirs. Chapter 1 gives
Utsira Sand, pressure is dispersed through a large vol- an overview over different structural and stratigraphic
ume and pressure build-up is minimal, and the traps, and Chap. 12 explains the detailed tectonic and
injectivity is therefore very high. The Tubåen Forma- structural settings of hydrocarbon traps.
tion at Snøhvit, on the other hand, is divided into Although CO2 storage is in many ways analogous
pressure compartments by steep faults with low perme- to hydrocarbon storage, there is one important differ-
ability and hydraulic conductivity, resulting in rapid ence: CO2 is reactive and forms carbonic acid that may
pressure build-up (Pham et al. 2011). Hydraulic con- react with, and change the properties of, the cap-rock
ductivity depends on the permeability of the reservoir (Alemu et al. 2011). The effect of this alteration on an
and the viscosity of the fluid, with high permeabilities intact caprock at ambient temperatures (for most cases
(and low viscosities) favouring water mobilisation and probably at less than 100
C) is however likely very
preventing pressure build-up. limited as the reactions are slow and diffusion of CO2
A storage deposit must in addition to injectivity into a tight shale will be very slow as well (Gaus et al.
have sufficient capacity. If pressure build-up is not 2005; Lu et al. 2009). If CO2 is allowed to migrate into
considered, CO2 storage capacity for saline aquifers fractures or faults that penetrate the seal, then the
can be estimated by the following relationship (see picture is more uncertain, and it is not well understood
CO2 Storage Atlas Norwegian Sea, 2012): if and when reactions will be self-sealing, and when
reactions will enhance flow.
mCO2 ¼ V b  ϕ  N=G  ρCO2  χ (24.1)

where Vb, ϕ, and N/G are the bulk volume, porosity 24.4 The Fluid Properties of CO2
and Net/Gross ratio of the storage reservoir, ρCO2 is the
density of CO2, and χ is the storage efficiency factor. CO2 may occur as a solid, a liquid or a vapour,
We see that injectivity and storage capacity are linked determined by the temperature and pressure
through the reservoir volume and porosity (affecting (Fig. 24.2). Properties such as density and viscosity
permeability; see Chap. 10). The density of CO2 varies vary greatly between these phases. At 304.25 K and
from about 2 kg/m3 at the Earth’s surface (a bit heavier 7.38 MPa, CO2 is at the so-called critical point (Crp),
than air) to more than 500 kg/m3 at conditions deep and above this point CO2 is termed supercritical. At
underground where CO2 is supercritical. The CO2 temperatures above the Crp (but pressure below), CO2
properties are described in more detail in section is a critical gas, whereas if only pressure is above the
24.4. From equation (24.1), we see that the storage Crp, CO2 will be a critical liquid with high density
capacity is proportional to the CO2 density. The final similar to natural gas. The viscosity of scCO2 is about
variable is the storage efficiency factor. This is defined 0.05–0.06 cp at 100 bar and 37
C (Zabaloy et al.
as the fraction of the total reservoir pore space that can 2005) (corresponding to the Utsira storage reservoir),
be filled with scCO2, with values typically being a few slightly lower than for water under the same con-
percent or lower. The most important factor ditions. The storage capacity is a function of CO2
controlling the storage efficiency factor is reservoir density and therefore also of burial depth. At the
heterogeneity. Spatial differences in permeability Earth’s surface, CO2 takes up approximately 22.4 l/
lead to flow of CO2 along preferential fluid pathways, mol gas (about 0.51 l/g). At these conditions, only a
and CO2 is prevented by the capillary entry pressure couple of grams of CO2 could have been stored per
from migrating into fine-grained units. litre pore space. The situation is however very differ-
ent at the higher temperatures and pressures found
underground. The density of CO2 increases with pres-
24.3.3 The Seal (Cap Rock) sure and decreases with temperature in a quite com-
plex way (Fig. 24.3). At low temperatures, below the
As for oil and natural gas, the density of CO2 stored Crp, pressure increases leads to a phase transition from
underground is lower than formation water, so CO2 gas to liquid and a large increase in the CO2 density.
will migrate buoyantly towards the surface. Above the Crp, no such phase transitions occur and
594 H. Hellevang

Fig. 24.4 Mutual solubilities of CO2 and H2O. (a) The amount
Fig. 24.2 CO2 phase stability diagram (s ¼ solid, l ¼ liquid, of CO2 dissolved in the water-rich phase increases with pressure
g ¼ gas). Phase boundaries are given by the solid curves, but decreases with temperature; (b) the amount of H2O
whereas the triple point (T and P where s + l + g are all stable) dissolved in the CO2-rich phase increases with temperature,
and critical point (Crp) are indicated by solid circles whereas pressure first has a positive effect before the solubility
decreases at increasing pressures. Calculations done by R. Miri
(UiO) using the SAFT-EOS approach

reservoirs, which most likely will lie at depths greater

than 1 km, CO2 will have densities between about 200
and 700 g/cm3 compared to >1,000 g/cm3 for brine,
giving a significant buoyancy, i.e. upward force.
To avoid corrosion of pipelines and formation of
hydrates (ice-like particles of water and CO2), CO2
should be injected as a dry phase with as little water as
possible in the stream. However, as soon as CO2 is
injected into a reservoir, CO2 mixes and dissolves in
water and some water dissolves back into the CO2
phase and reacts to form carbonic acid. The mutual
solubilities of CO2 and water depend on temperature
and pressure and on the salinity of the formation water.
Fig. 24.3 CO2 densities as a function of temperature and pres- Although the mutual solubilities are quite limited
sure (from 0 to 70 Mpa and 273 to 453 K). Superimposed are (Fig. 24.4), they have profound effects on the storage
CO2 density curves estimated by three different geothermal
operation both during the injection period (typically
gradients; one being a ‘normal’ North Sea gradient (35
/km)
and two other curves being 10
/km from the normal gradient. lasting for a few decades) and during the subsequent
Density data from the NIST database long-term storage.
During injection, water that is close to the well
pressure increases lead to a steady increase in density. will dissolve into the dry CO2 and the salinity of the
If CO2 is stored underground at normal pressure brine will increase. This may result in the formation
(hydrostatic; see Chap. 10), the geothermal gradient of evaporites, clogging the pore space and reducing
determines the phase properties. Superimposed on the the injectivity. The potential that one pore volume of
density plot in Fig. 24.3 are three geothermal brine has to change the porosity depends on the initial
gradients, one being quite average for the North Sea salinity. If the salinity is 1 mol NaCl/L pore space, a
(35
C/km) and the two others being 10
C/km. We complete dry-out would clog approximately 3% of
see that there are large differences in density, espe- the pore space (assuming a halite molar weight and
cially at shallow depth. It is also clear that CO2 storage density of 58.44 g/mol and 2.17 g/cm3 respectively).
24 Carbon Capture and Storage (CCS) 595

This appears to be quite limited. However, as pores

dry out, additional brine may be brought in by capil-
lary suction and osmotic pressure differences, lead-
ing to further salt precipitation. Dissolution of CO2
into water affects the storage operation on both short
and long time-scales. First, dissolved CO2 reacts with
water and forms carbonic acid, and the pH may drop
to 3 or even lower depending on the buffer capacity
and CO2 pressure. This results in dissolution of the
minerals present in the reservoir and formation of
new secondary phases. In hundreds to thousands of
years, these mineral reactions may form carbonates,
immobilising CO2 for long time spans (see next
section).
A second effect of dissolved CO2 is that it Fig. 24.5 Solubility of CO2 in NaCl-dominated brines with salt
increases the density of the water. The solubility of contents from 0.0 to 5.8 moles per kg water (kgw). Note the
CO2 at 100 bars and 37
C in a solution of normal “salting out effect”, how the solubility decreases with increasing
salt content. Calculations done by R. Miri using the SAFT-EOS
seawater salinity (~0.5 mol NaCl per kg water (kgw)) approach
is approximately 1.1 mol/kgw. At full CO2 satura-
tion, density is about 1,034 kg/m3, which is 1.6%
higher than the solution without CO2 (about
1,018 kg/m3). It has been suggested that the density
contrast between CO2-saturated formation water and
normal brine may result in convection cells, making
the heavier CO2-charged solutions sink downwards,
and bringing ‘fresh’ brine upwards which increases
the dissolution rate of the scCO2 (Elenius and Gasda
2013). Viewed in the light of earlier studies on the
possibility of density currents in stratified and verti-
cally heterogeneous sandstones (Bjørlykke et al.
1988), conclusions on the extent of such mixing
still have to be taken with caution. Moreover, the
solubility of CO2 in water is strongly a function of
salt content. Higher salinities results in less dissolved
CO2 (Fig. 24.5), and less density contrast between
fully saturated CO2-brine and brine with no CO2
(Fig. 24.6). The potential for density currents for Fig. 24.6 Density difference (ΔDensity) defined as the differ-
ence between the density of fully CO2-saturated brines and
high-salinity brines may therefore be even lower. brines without dissolved CO2. The difference is getting lower
and lower as the salinity increases, mainly because of the
salting-out effect leading to decreasing CO2 solubility with
24.5 CO2 Trapping Mechanisms increasing salinity. Calculations done by R. Miri using the
SAFT-EOS approach

Four mechanisms are commonly considered responsi-

ble for keeping CO2 underground after injection: (1) 24.5.1 Structural Trapping
structural trapping; (2) capillary or residual trapping;
(3) solubility trapping; and (4) mineral trapping At short time-scales (during the injection period), most
(Benson and Cole 2008). The risk of leakages and CO2 will be a separate buoyant phase. The vertical
the time dependency of risk are related to these four rock permeability and the density contrast between the
mechanisms (see Fig. 6 in Benson and Cole 2008).
596 H. Hellevang

formation water and scCO2 (see Fig. 24.3) will deter- where x is mole fraction, ϕ and γ are the fugacity and
mine the upward flux of CO2. Chapter 12 and Sect. 1.8 activity coefficients of CO2, and KH is the Henry’s law
provide an overview of structural elements that can coefficient. The molecular CO2 reacts with water to
prevent vertical migration of CO2 or hydrocarbons. form carbonic acid, which dissociates in two steps to
form bicarbonate and carbonate:

24.5.2 Capillary (Residual) Trapping CO2 þ H2 O ¼ H2 CO3

H 2 CO3 ¼ HCO þ
3 þH ; (24.3)
As CO2 flows upwards and laterally below structural
or stratigraphic elements, water must be displaced. HCO
3 ¼ CO2
3 þH þ

Such displacement of a wetting fluid by a non-wetting

fluid is termed drainage. In siliciclastic reservoirs, where the equal sign (‘¼’) denotes equilibrium. Both
mineral grains will be water-wet, and a thin film of the Henry’s law constant and the equilibrium
water will therefore remain on the mineral grains after constants for the carbon speciation are temperature
a CO2 sweep. If water later flows through the area and pressure dependent. As a result of the CO2 disso-
rich in free CO2, some bubbles of CO2 may not be lution, pH drops to 3 or even lower at the CO2
able to overcome the capillary entry pressure of the pressures encountered during CO2 storage.
smaller pore throats, and will be trapped. The latter Because new interfaces between the undissolved
process is termed imbibition. In modelling drainage CO2 and formation water are constantly made through
and imbibition, the relation between the water and the mobilisation of the free-phase CO2, or through
CO2 saturations and the relative permeabilities must density-driven mixing of CO2 and formation water,
be known. The van Genuchten model is one of several more and more CO2 dissolves with time. Dissolution
ways of relating relative permeabilities to CO2 and is therefore favoured in sloping aquifers where the
H2O saturations (van Genuchten 1980). The model is CO2/water interface is large and in heterogenous
empirical and requires experimental data. Values for reservoirs where the plume spreads through a larger
irreducible saturations, which are the minimum satu- volume (Fig. 24.7). If vertical mixing through density-
ration values CO2 and H2O can have, and the shape of driven plume formation can occur (see last part of
the saturation/permeability curves, depend on factors Sect. 24.4), then this will promote further dissolution.
such as sediment sorting and mineral grain size.

24.5.4 Mineral Trapping

24.5.3 Solubility Trapping (CO2 Carbonatisation)
Injected CO2 will first occur primarily as a free phase, The increased acidity of the formation water resulting
but will start to dissolve in the formation water. The from CO2 dissolution increases the solubility of the
rate of dissolution depends on factors such as the CO2/ primary minerals (i.e. minerals in the reservoir prior to
water interface surface area, and diffusion of dissolved CO2 injection), leading to dissolution of these phases
CO2 away from the interface. Solubility trapping will and precipitation of a new stable mineral assemblage.
therefore be most efficient in heterogeneous reservoirs The rate of dissolution varies from mineral to mineral
where the CO2 plume spreads over larger volumes of and is strongly a function of the reservoir temperature.
the reservoir providing more interfacial surface area From the primary phases, the clay minerals, feldspars
between CO2 and formation water. The amount of and carbonates play important roles during CO2 stor-
dissolved molecular CO2 depends on factors such as age. Primary carbonates such as calcite dissolve fast
the CO2 partial pressure, temperature and salinity, and (within days) and transform some of the carbonic acid
can be generally expressed by: to alkalinity (HCO3), buffering some of the pH drop
induced by the CO2:
PCO2 ϕ
xCO2 ¼ ; (24.2)
KH γ
CaCO3 þ H 2 CO3 ¼ Ca2þ þ 2HCO3  : (24.4)
24 Carbon Capture and Storage (CCS) 597

Fig. 24.7 Injection of CO2 in the Johansen Formation #2 Harmonic mean averaged model, Kz ¼ 0.082 mD. Including
(modified from Sundal et al. 2014). Plume distributions after the flow baffles leads to significantly larger lateral spread of the
20 years. (a, b) Case #1 includes 15 thin flow baffles. (c, d) Case plume

The pH commonly increases from about 3 to 5 in this (Hellevang and Aagaard 2013, Hellevang et al. 2013).
reaction (Pham et al. 2011). The rates of feldspar and Feldspars release Ca2+ and Na+, supporting calcite,
clay mineral dissolution are many orders of magnitude ankerite and dawsonite (NaAl(OH)2CO3) formation.
slower than for carbonate dissolution, especially at In many quartz-rich reservoirs that are poor in Me2+,
low temperatures. These reactions however play both simulations (e.g. Johnson et al. 2004) and natural
important roles in the immobilisation of CO2 in the observations (e.g. Gao et al. 2009) have suggested that
long term, i.e. mineral trapping of CO2. Clay minerals, dawsonite may be the dominant secondary carbonate
such as chlorites, contain divalent metal cations (Me2+) that forms and traps CO2.
that can react with dissolved CO2 and form secondary Estimates of the mineral-trapping potential of a
carbonates (the inverse of reaction 24.4). The most reservoir require prediction of the following: (1) the
common secondary Me2+-carbonates predicted to amount (volume) of formation water being affected by
form during CO2 storage are ankerites the CO2 injection, which depends on the drainage
(CaFexMg1x(CO3)2, with x 1), siderite (FeCO3), pathway; (2) the representative average primary and
and in some cases disordered dolomite and magnesite. secondary mineralogy; and (3) the dissolution rates of
The formation of the latter two in numerical simulations the primary minerals and growth rates of secondary
may however be an artifact of the kinetic models being phases. All these points require a good understanding
used overpredicting rates of Mg-carbonate formation of the depositional environment and the diagenetic
598 H. Hellevang

Fig. 24.8 CO2 carbonatisation potential for (a) Utsira Sand; sediments) depends strongly on the initial sediment composi-
and (b) Gulf Coast sediments (modified from Hellevang and tion, and that the range in predictions using different kinetic
Aagaard 2013). The figure shows that the total amount of models is very large. ‘Alpha’ here denotes fraction of the min-
carbonates that can form (~10 times more in Gulf Coast eral surface area being reactive (see Hellevang et al. 2013)

overprint. In the kinetic models, reactive surface areas 1,030–1,100 kg/m3) is used to map the extent of the
are commonly regarded as a highly uncertain parame- CO2 plume. This method has been very successful in
ter. Knowledge on how much of the surfaces of reac- monitoring the CO2 plume migration for the Sleipner
tive minerals (e.g. feldspars and chlorites) are in injection (Eiken et al. 2011). Microseismicity caused
contact with the formation water (provides the specific by fracture or fault activation may also provide valu-
surface area as m2/L porewater) is therefore essential. able information on the effect of pressure build-up and
Figure 24.8 shows predictions of the long-term fluid migration in the sub-surface. The status of CO2
trapping potential for two reservoirs, the Utsira Sand, storage operations is obtained indirectly by monitoring
and a representative of a feldspar-rich Gulf Coast the reservoir pressure. If the pressure approaches the
sediment (Frio Formation, Lower Tertiary). The figure estimated cap-rock fracture pressure, preventative
shows how reducing the reactive surface area by one measures have to be taken. Such measures can be
order of magnitude (parameter α) has a large impact reduction of the injection rate or water production.
on the predictions, and that the three different kinetic However, in some cases, such as for the Tubåen For-
models used also lead to a large range in predicted mation at Snøhvit, the storage capacity is very limited
carbonatisation. Because of the uncertainties in the and alternative storage reservoirs have to be found. In
kinetic models, there is at present a focus to improve most cases, a combination of seismic monitoring and
both the kinetic models and the parameters used in the measurements of the well pressures is employed.
models.

24.7 Examples of CO2 Storage Projects

24.6 How Can We Know that CO2 Is
Safely Stored? There are today (October 2013) 12 CCS projects that
are operational, but a large number of others are in the
To ensure that the CO2 is safely stored underground, planning and executive phases (Global CCS Institute).
various monitoring technologies are in use, with seis- One of the projects that has received most attention is
mic and well-pressure monitoring being two important at Sleipner (North Sea), being the first industrial-scale
methods. Seismic monitoring is built on comparing dedicated CCS project started in 1996. Some projects
seismic signals in time and space, so called time- such as In Salah, Algeria, were running until recently,
lapse or 4D seismics (see Chap. 19). By doing so, the but that project ended after illustrating how several
difference in the density of supercritical CO2 (approx- monitoring technologies can be used to successfully
imately 400–800 kg/m3) and water (approximately track the fate of stored CO2. Below is a short overview
24 Carbon Capture and Storage (CCS) 599

of the Sleipner, Snøhvit (Barents Sea), In Salah, and the seismic images is the presence of a chimney-like
Weyburn-Midale (Saskatchewan, Canada) CCS structure in the storage reservoir just above the injector.
projects. This structure possibly formed shortly after injection
commenced, by liquefaction of the sediments breaking
up the thin intra-reservoir shale layers (Hermanrud et al.
24.7.1 Sleipner CCS Project 2009). Although CO2 has migrated through what was
thought to be a tight shale layer in the cap-rock and into
Natural gas produced at Sleipner contains about 9.5% an overlaying sand wedge, there are no indications of
CO2. In order to ship this natural gas in a liquid form further leakage to shallower depths.
(LNG) to customers it has to be cooled to 162
C,
which may cause formation of CO2 ice aggregates. To
avoid this, most of the CO2 is captured during produc- 24.7.2 Snøhvit CCS Project
tion, i.e. removed from the natural gas stream. Before
1996, CO2 captured from the Sleipner field was Gas produced from the Lower to Middle Jurassic Stø
released directly to the atmosphere. A few years Formation at Snøhvit contains 5–8% CO2. This gas is
before, in 1991, Norway had established a tax on transported by pipelines to the processing plant at
CO2 emissions (in 1999, about 18 US$/ton CO2 on Melkøya, where CO2 is captured and transported back
emissions of ~1 Mt CO2/year). This represented a underground. CO2 storage started in 2008 with injection
considerable financial levy on the operators of into the early Jurassic Tubåen Formation, which lies
Sleipner. In 1996, Statoil and partners started to inject just below the gas-producing Stø Formation In contrast
CO2 captured at Sleipner into the Utsira Formation, to the Utsira Sand, the stratigraphic sequences in the
and the injection has continued successfully at a rate of Barents Sea region have experienced a complex history
approximately 1 Mt CO2/year. of burial and later tectonic uplift. The present-day depth
The CCS operation at Sleipner was feasible of the Tubåen is about 2,400 m at the injection site, and
because of the properties of the Utsira storage reser- temperature and pressure before injection were 95
C
voir located near the Sleipner field. The Utsira Sand (Hansen et al. 2013) and 240 bars (assuming hydro-
sediments are basin-restricted marine lowstand static). The depth and temperature at maximum burial
deposits of Miocene-Pliocene age (Zweigel et al. may however have been more than 3,200 m and 130
C.
2004), only weakly compacted and with high porosity After a short period of injection, the operation experi-
(30–40%). The sand is very extensive (see Fig. 24.1) enced problems with rapid pressure build-up,
and is in many ways an optimal storage reservoir. At interpreted to be caused by reactions clogging the
800 m below seafloor, temperatures and pressures are pore space and reducing the permeability around the
about 37
C and 100 bar, which is above the critical well (Hansen et al. 2013). In the Tubåen Formation,
point of CO2. As overburden, the Utsira Formation has these reactions could be: (1) salt formation (scale)
a 250 m clay-rich sequence of the Nordland group caused by the drying-out of highly saline brine around
which provides the seal, with a further several hundred the well; or (2) CO2 hydrate (clathrate) formation. As
metres of coarser-grained Quaternary sediments above CO2 entered the reservoir at low temperature (4
C) and
(Zweigel et al. 2004). The mudstones deform in a pressures were higher than 290 bars (Hansen et al.
ductile manner and stress changes are therefore not 2013), the latter is more likely (Kvamme and Tanaka
expected to lead to fractures that could provide leak- 1995, Duan and Sun 2006), although the highly saline
age pathways for the CO2. brine depresses the freezing point of CO2 hydrate (Duan
The injected CO2 has been monitored by 3D and Sun 2006). Salt (halite) formation caused by injec-
seismics surveys carried out before, and several times tion of dry CO2 has been predicted numerically (Muller
after, injection started in 1996. The seismic images et al. 2009) and in laboratory experiments (Bacci et al.
show that CO2 has spread out laterally below nine 2013), and although not yet observed in situ, could also
distinct stratigraphic layers. A bulk of the CO2 is still very well be the reason for the pressure build-up. After
trapped under these layers and this has reduced the the initial pressure build-up was solved by chemical
amount of free CO2 that has reached the cap rock treatment, the operation continued. Unfortunately, the
(Hermanrud et al. 2009). A much discussed feature of tight and compartmentalised Tubåen Formation turned
600 H. Hellevang

out to have only a limited storage capacity, and pressure capacity of the Weyburn-Midale reservoirs, including
continued to increase until injection was abandoned in both the EOR phase and later CO2 storage, is
2011. Injection was then moved to the shallower Stø estimated to be between 50 and 60 Mt. The
Formation, below the gas reserves. As CO2 migrates Weyburn-Midale CO2 monitoring and storage project,
updip and will mix with the natural gas, there is an supported by the International Energy Agency Green-
expectance that additional storage sites will be needed house Gas R&D program (IEA GHG), has invested in
in the future to continue the CO2 storage operation. an extensive geophysical and geochemical monitoring
programme, including passive seismic and soil gas and
fluid geochemistry monitoring (Whittaker et al. 2011).
24.7.3 In Salah CCS Project

The In Salah CCS project in central Algeria is an 24.8 What Is the Future of CCS?
example of an onshore CCS operation. During the
lifetime of the project, between 2004 and 2011, 3.8 The main obstacles to implementing CCS on a large
Mt CO2 was stored in a Carboniferous sandstone unit scale today are high costs and, for many regions, a lack
1.9 km below the surface. The project demonstrated a of a market for CO2. In North America, CO2 has been
large number of monitoring technologies, such as the used for enhanced oil recovery for decades, and this
use of Interferometric Synthetic Aperture Radar illustrates that the costs of capture and transport can be
(InSar) to monitor surface heave, 4D seismics to more than compensated if a market exists to sell the
track the underground CO2 migration, geophones to CO2. A second challenge is that the cost of releasing
detect microseismicity, well-head sampling of geo- CO2 to the atmosphere is still too low to make it
chemical tracers, etc. (Ringrose et al. 2013 and financially viable for most emitters to store CO2.
references therein). The combination of all these Below is a short review of present-day activities to
technologies enabled the tracking both of possible utilise and store CO2 in the USA, and a suggested
hydrofractures and potential migration of the CO2 solution to lower the cost and allow large-scale storage
outside the Krechba hydrocarbon lease, and remedial by gathering emissions from many point sources into
action (operation and monitoring responses). The In one shared pipeline network and storage site.
Salah CCS project has therefore provided highly valu-
able information on the whole chain, from planning to
operation and abandonment. 24.8.1 CCUS/EOR

CCUS (and CCUS/EOR), is short for Carbon Capture

24.7.4 Weyburn-Midale CCS-EOR Project Use (or Utilization) and Storage, and Enhanced Oil
Recovery (EOR). By using captured CO2 in industrial
The Weyburn-Midale project is at present the world’s processes such as EOR, CCS may become profitable.
largest CCS operation, and is an example of a CCS There are at present nine CCUS/EOR projects in the
Enhanced Oil Recovery (EOR) project. The CCS operational phase and several in the planning and exec-
operation started in 2000, and currently nearly 3 Mt utive stage, most of them in the USA (Global CCS
CO2 are annually injected in multiple wells and used Institute, Oct. 2013). Of these, the largest-scale is the
for enhanced oil recovery (Whittaker et al. 2011). At Weyburn-Midale CCS-EOR project described above.
the planned end of the operation, in 2035, the amount Onshore CCUS/EOR has been a great success. The
of stored CO2 is expected to be more than 40 Mt. The same is not yet true for offshore CCUS/EOR. There are
injected CO2 is a byproduct of coal gasification at the several reasons for this: (1) seawater is readily available
Great Plains Synfuels Plant in North Dakota, USA. offshore and can be used instead of the less available
The target reservoirs are of lower Carboniferous CO2; (2) the costs of installing CO2 capture facilities
age, at a present-day depth of 1,300–1,500 m. The may be considerable; (3) injected CO2 can lead to
rocks were deposited in a peritidal sequence, with corrosion of already existing infrastructure (wells etc.)
carbonates deposited at highstands and evaporitic and investment in new infrastructure (in addition to the
dolomites formed at lowstands. The total storage capture facilities) may be required; (4) stored CO2 may
24 Carbon Capture and Storage (CCS) 601

be produced back along with the oil. Whether CCS/ 24.9 Summary
EOR will be implemented offshore on a larger scale in
the future depends on the cost-benefit picture and the Carbon capture and storage is one out of several
availability of CO2. There are limits to the degree that strategies to mitigate rapidly increasing atmospheric
water can enhance oil recovery, and CO2 is known to CO2 levels. CCS involves capture of CO2 from natural
further increase the amount of oil that can be extracted gas or flue gas, and transport and storage in geological
and thus extend the lifetime of oil fields. Moreover, new formations. There are several technologies available for
technologies using CO2 for EOR are being developed. capture, and the choice of technology depends on the
One recent discovery is that CO2 foams (rather than gas processing plant and the cost of operation. Because of
or supercritical) can significantly improve the recovery the relatively high costs of capture, there are large
of medium to heavy oils and that foams are twice as research programmes underway to improve the
effective compared to water or gas/supercritical CO2 energy-efficincy of current methods and to come up
(Rassenfoss 2012). with new and more cost-effective capture technologies.
The captured CO2 can be stored underground in geo-
logical formations, preferentially as a dense supercriti-
24.8.2 Large Scale Transport and Storage cal phase, but this also requires large amounts of
energy. The main requirements of a storage formation
Most of the discussion so far has been on capture and are that it has sufficient injectivity and storage capacity
storage from single point sources. This limits CCS and a tight cap rock. Moreover, as water is mobilised by
projects to a few Mt/a. The total emissions from point the injected CO2 and the pressure waves caused by the
sources in larger regions in Europe, Asia, and America injection, one has to ensure that freshwater resources
may however reach hundreds of Mt/a. By connecting are not affected by the storage operation. In the future,
smaller (<20–40 Mt) CO2 suppliers through hubs, and to meet the IPCC target of globally storing Gt of CO2
transporting large amounts (hundreds of Mt/a) to per year rather than the few Mt/a stored today, storage
reservoirs such as the Utsira Sand with very large stor- operations will have to be scaled up. One way to do this
age capacity, the cost of stored CO2 per ton may be is to collect CO2 from several large point sources and
reduced significantly. The cost of pipelines and infra- store it in sandstones, such as the Utsira Sand, that
structure is shared among several stakeholders, and possess a very high storage capacity (Fig. 24.1).
larger dimensions of pipelines allow a greater volume
CO2 to be transported per ton steel used (volume/cross-
section ratio of pipelines increases). The amount of
pressure a pipeline can hold is a function of the diameter
of the pipeline, with smaller pipes withstanding higher References
pressures. The upper limit of pipeline dimensions must
therefore be weighed against the pressures required to Alemu, B.L., Aagaard, P., Muntz, I.A. and Skurtveit, E. 2011.
transport the CO2 as a dense phase. Caprock interaction with CO2: A laboratory study of reac-
tivity of shale with supercritical CO2 and brine. Applied
To allow such large-scale CCS in Europe requires
Geochemistry 26(12), 1975–1989.
regulations across national boundaries, and there are Bacci, G., Durucan, S. and Korre, A. 2013. Experimental and
several complicated issues that must be sorted out. numerical study of the effects of halite scaling on injectivity
One is the ownership and risk responsibility of the and seal performance during CO2 injection in saline aquifers.
Energy Procedia 37, 3275–3282.
transport and storage of a mixture of CO2 from several
Benson, S.M. and Cole, D.R. 2008. CO2 sequestration in deep
sources. Who will be held responsible in case of a sedimentary formations. Elements 4, 325–331.
pipeline rupture? Who owns the stored CO2 and will Bjørlykke, K., Mo, A. and Palm, E. 1988. Modelling of thermal
be responsible for a monitoring programme? It is convection in sedimentary basins and its relevance to diage-
netic reactions. Marine and Petroleum Geology 5, 338–351.
natural that in Europe the European Union and
Duan, Z. and Sun, R. 2006. A model to predict phase equilib-
associated nations such as Norway, should facilitate rium of CH4 and CO2 clathrate hydrate in aqueous electro-
such large scale CCS. lyte solutions. American Mineralogist 91, 1346–1354.
602 H. Hellevang

Eiken, O., Ringrose, P., Hermanrud, C., Nazarian, B., Torp, T. sequestration partitioning. In: Bains, S.J. and Worden, R.H.
A. and Høier, L. 2011. Lessons learned from 14 years of (eds.), Geological Storage of Carbon Dioxide. Geological
CCS operations: Sleipner, In Salah and Snøhvit. Energy Society Special Publication 233, pp. 107–128.
Procedia 4, 5541–5548. Kvamme, B. and Tanaka, H. 1995. Thermodynamic stability of
Elenius, M.T. and Gasda, S.E. 2013. Convective mixing in hydrates for ethane, ethylene, and carbon dioxide. Journal of
formations with horizontal barriers. Advances in Water Physical Chemistry 99, 7114–7119.
Resources 62, 499–510. Lu, J., Wilkinson, M., Haszeldine, R.S. and Fallick, A.E. 2009.
Figueroa, J.D., Fout, T., Plasynski, S., McIlvried, H. and Long-term performance of a mudrock seal in natural CO2
Srivastava, R.D. 2008. Advances in CO2 capture technol- storage. Geology 37, 35–38.
ogy—The U.S. Department of Energy’s carbon sequestra- Muller, N., Qi, R., Mackie, E., Pruess, K. and Blunt, M.J. 2009.
tion program. International Journal of Greenhouse Gas CO2 injection impairment due to halite precipitation. Energy
Control 2, 9–20. Procedia 1, 3507–3514.
Gao, Y., Liu, L. and Hu, W. 2009. Petrology and isotopic geo- Pham, T.H.V., Maast, T.E., Hellevang, H. and Aagaard, P. 2011.
chemistry of dawsonite-bearing sandstones in Hailaer basin, Numerical modeling including hysteresis properties for CO2
northeastern China. Applied Geochemistry 24, 1724–1738. storage in Tubåen Formation, Snøhvit field, Barents Sea.
Gaus, I., Azaroual, M. and Czernichowski-Lauriol, I. 2005. Energy Procedia 4, 3746–3753.
Reactive transport modelling of the impact of CO2 injection Rassenfoss, S. 2012. Increased oil production with something
on the clayey cap rock at Sleipner (North Sea). Chemical old, something new. Journal of Petroleum Technology, Oct.
Geology 217, 319–337. issue, 36–39.
Global CCS Institute. [Link] Ringrose, P.S., Mathieson, A.S., Wright, I.W., Selama, F.,
Hansen, O., Gilding, D., Nazarian, B., Osdal, B., Ringrose, P., Hansen, O., Bissell, R., et al. 2013. The In Salah CO2 storage
Kristoffersen, J.-B., et al. 2013. Snøhvit: The history of project: lessons learned and knowledge transfer. Energy
injection and storing 1 Mt CO2 in the fluvial Tubåen Fm. Procedia 37, 6226–6236.
Energy Procedia 37, 3565–3573. Sundal, A., Hellevang, H., Miri, R. and Aagaard, P. 2014. On
Hellevang, H. and Aagaard, P. 2013. Can the long-term poten- layer specific CO2 plume distributions and variability in min-
tial for carbonatization and safe long-term CO2 storage in eralization potential. Submitted to the Fourth EAGE CO2
sedimentary formations be predicted? Applied Geochemis- Geological Storage Workshop, Stavanger, April 23–25, 2014.
try 39, 108–118. van Genuchten, [Link]. 1980. A closed-form equation for
Hellevang, H., Pham, V.H. and Aagaard, P. 2013. Kinetic predicting the hydraulic conductivity of unsaturated soils.
modeling of CO2-water-rock interactions. International Jour- Soil Science Society of America Journal 44, 892–898.
nal of Greenhouse Gas Control 15, 3–15. Whittaker, S., Rostron, B., Hawkes, C., Gardner, C., White, D.,
Hermanrud, C., Andresen, T., Eiken, O., Hansen, H., Janbu, A., Johnson, J., et al. 2011. A decade of CO2 injection into
Lippard, J., et al. 2009. Storage of CO2 in saline aquifers— depleting oil fields: monitoring and research activities of
lessons learned from 10 years of injection into the Utsira the IEA GHG Weyburn-Midale CO2 monitoring and storage
Formation in the Sleipner area. Energy Procedia 1, 1997–2004. project. Energy Procedia 4, 6069–6076.
IPCC. 2013. Climate change 2013: The physical science basis. In: Zabaloy, M.S., Vasquez, V.R. and Macedo, E.A. 2005. Viscos-
Stocker, T.F., Quin, D., Plattner, G.-K., Tignor, M., Allen, ity of pure supercritical fluids. Journal of Supercritical Fluids
S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V. and Midgley, 36, 106–117.
P.M. (eds.), Contribution of Working Group I to the Fifth Zweigel, P., Arts, R., Lothe, A.E. and Lindeberg, E.G. 2004.
Assessment Report of the Intergovernmental Panel on Climate Reservoir geology of the Utsira Formation at the first
Change. Cambridge: Cambridge University Press, 1535 pp. industrial-scale underground CO2 storage site (Sleipner
Johnson, J.W., Nitao, J.J. and Knauss, K.G. 2004. Reactive area, North Sea). In: Baines, S.J. and Worden, R.H. (eds.),
transport modeling of CO2 storage in saline aquifers to Geological Storage of Carbon Dioxide. Geological Society
elucidate fundamental processes, trapping mechanisms and Special Publication 233, pp. 165–180.

View publication stats