international journal of greenhouse gas control 2 (2008) 502510

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/ijggc

Intermediate storage of carbon dioxide in geological

formations: A technical perspective
Semere Solomon *, Michael Carpenter, Todd Allyn Flach
Det Norske Veritas, DNV Research and Innovation, NO-1322 Hvik, Norway

article info

abstract

Article history:

Enhanced oil recovery (EOR) through CO2 flooding has been practiced on a commercial basis

Received 4 January 2008

for the last 35 years and continues today at several sites, currently injecting in total over 30

Received in revised form

million tons of CO2 annually. This practice is currently exclusively for economic gain, but

3 April 2008

can potentially contribute to the reduction of emissions of greenhouse gases provided it is

Accepted 4 April 2008

implemented on a large scale. Optimal operations in distributing CO2 to CO2-EOR or

Published on line 20 May 2008

enhanced gas recovery (EGR) projects (referred to here collectively as CO2-EHR) on a large
scale and long time span imply that intermediate storage of CO2 in geological formations

Keywords:

may be a key component. Intermediate storage is defined as the storage of CO2 in geological

Intermediate CO2 storage

media for a limited time span such that the CO2 can be sufficiently reproduced for later use

Geological formations

in CO2-EHR. This paper investigates the technical aspects, key individual parameters and

Enhanced oil recovery

possibilities of intermediate storage of CO2 in geological formations aiming at large scale

Flow simulation

implementation of carbon dioxide capture and storage (CCS) for deep emission reduction.
The main parameters are thus the depth of injection and density, CO2 flow and transport
processes, storage mechanisms, reservoir heterogeneity, the presence of impurities, the
type of the reservoirs and the duration of intermediate storage. Structural traps with no flow
of formation water combined with proper injection planning such as gas-phase injection
favour intermediate storage in deep saline aquifers. In depleted oil and gas fields, high
permeability, homogeneous reservoirs with structural traps (e.g. anticlinal structures) are
good candidates for intermediate CO2 storage. Intuitively, depleted natural gas reservoirs
can be potential candidates for intermediate storage of carbon dioxide due to similarity in
storage characteristics.
# 2008 Elsevier Ltd. All rights reserved.

1.

Introduction

Emissions of CO2 due to fossil fuel burning are the dominant

contributor to increasing trends in atmospheric CO2 concentration because 80% of the global energy consumption is based
on coal, oil, and natural gas (IEA, 2005).
Carbon dioxide capture and storage (CCS) is a promising
technology option to continue exploiting fossils and reduce
their emissions of GHG, provided that CCS is implemented on

a large enough scale. Several types of storage reservoirs may

provide storage capacities of the required magnitude. Geological storage of CO2 can be undertaken in a variety of
geological settings in sedimentary basins. Within these
basins, oil fields, depleted gas fields, deep coal seams and
saline formations are all possible storage formations.
This study assesses the geological and reservoir engineering aspects of using geologic formations for intermediate
storage of CO2. It is important to note that the storage

* Corresponding author. Fax: +47 67579911.

E-mail addresses: Semere.Solomon.Foto@dnv.com (S. Solomon), Mike.Carpenter@dnv.com (M. Carpenter), Todd.Flach@dnv.com
(T.A. Flach).
1750-5836/$ see front matter # 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijggc.2008.04.004

503

international journal of greenhouse gas control 2 (2008) 502510

requirements for storing CO2 in geological formations are

equally significant when using these sites for intermediate
storage. These include: (1) adequate porosity and thickness
(for storage capacity) and permeability (for injectivity), (2) a
satisfactory sealing caprock or confining unit and (3) a
sufficiently stable geological environment to avoid compromising the integrity of the storage site. The difference is that
intermediate storage is meant for short-term storage which
can be reproduced when the need arises.
Intermediate geologic storage of CO2 could promote
continued and wider, larger scale implementation of both
CCS and EHR for a number of intuitive reasons. Intermediate
storage could help to overcome the inherent differences in the
lifetime of EHR projects (typically 1025 years) and the lifetime
of CO2 capture projects (typically 2550 years), as well as the
differences in CO2 production rates from capture plants and
the injection rates needs for CO2 of EHR projects. Wider scale
application of EHR and CCS also implies the need for more
integrated distribution infrastructures, which intuitively
require some level of short-term buffering to avoid problems
with production interruptions at both ends, i.e. at the capture
plants producing the CO2 and at the EHR sites receiving the
CO2. And finally, intermediate storage could be an effective
hedging strategy against future price increases of CO2 and
otherwise price volatility in a market for CO2 under more
general cap/trade and supply/demand constraints. In short, to
promote long-term emissions reductions in application of CCS
may require well-engineered intermediate geologic storage of
CO2. Temporary, intermediate storage will promote long-term,
permanent storage.
Intermediate storage in this paper means the storage of CO2
in geological media for limited span of time that can be
sufficiently reproduced for later use. After this brief introduction, the study approach is highlighted in Section 2. Results of
this investigation is first presented in Section 3, analyses of the
parameters and possibilities of intermediate storage of CO2 in
geological formations and then key trapping mechanisms
such as the effects of residual gas trapping, dissolution on
intermediate storage are assessed under simulation results in
Section 4. In Section 5 the discussions and the conclusions
hereby drawn are presented. In addition changes in porosity
and permeability are evaluated in the later section. It is
assumed that the storage sites are well characterized.
Important issues such as the geographical disposition and
the costs involved require additional studies.

2.

Study approach

The study is carried out based on literature survey supported

by our own simulation study of CO2 injection on an idealized
model shown in Fig. 1. CO2 is injected into a porous formation
200 m thick (20 model layers, 10 m thick each) located at a
depth of 900 m. The formation is assumed homogeneous with
the porosity of the formation 38%, horizontal and vertical
permeability 4000 millidarcy (md) each. To evaluate the effects
of porosity and permeability changes on the models tested,
both parameters are changed to 15% for the porosity and
1000 md for the permeability (horizontal and vertical). Table 1
provides the properties assumed for the model in the various

Fig. 1 Three-dimensional block diagram of the

axisymmetric model used for the numerical simulations.

simulations. The porosity and permeability values may not be

typical for clastic sedimentary rocks found throughout the
world; however, they are idealized for testing different cases.
Initially, the brine saturation is 100% everywhere in the
model with a pressure of 80 bar at the top and 100 bar at the
bottom. A geothermal gradient of 25 8C/km is assumed, with
the temperature at the surface and base of the model held
constant at 6 and 37 8C, respectively. The salinity of the pore
water is assumed to be 3% by weight. The numerical
simulations begin with injection of 500,000 Sm3/day of CO2
into the porous formation. It is injected at a constant rate for 20
years. Simulations continue for 80 years. The numerical
simulations were carried out using 3D axisymmetric orthogonal, 1/4 model composed of 20 layers containing
175 m  175 m grid blocks. All grid blocks are 10 m thick.
The numerical simulator used for this work was GEM-GHG
from the Computer Modelling Group Ltd. GEM-GHG is a stateof-the-art compositional flow simulator which can model the
most relevant physical processes related to long-term storage
of CO2 in saline aquifer systems as illustrated in Fig. 2. GEMGHG is an equation of state geochemical compositional
simulator for modelling CO2 storage processes. The simulator
uses an adaptive implicit discretization technique to model
the component transport in porous media.
Together, capillary pressure and relative permeability are
known as characteristic curves; they control the way the liquid
and gas phases interact. Characteristic curves can be consid-

Table 1 Simulations parameter input for 3D model

Model type
Grid type
Storage structure
Grid cell size
Depth to bottom of aquifer (centre)
Pressure top and bottom
Temperature top and bottom
Vertical permeability
Horizontal permeability
Porosity
Injection rate
Salinity
Aquifer thickness
a
b

Low porositypermeability cases tested.

Standard.

Saline aquifer
Orthogonal, 1/4 model
Dome anticline
175 m  175 m  10 m
900 m
80 and 100 bar
6 and 37 8C
4000 md (1000 md)a
4000 md (1000 md)a
38% (15%)a
500,000 Sbm3/day for
20 years (7.25 Mt total)
3% by weight
200 m

504

international journal of greenhouse gas control 2 (2008) 502510

Fig. 2 Some of the chemical reactions that need to be

modelled for CO2 storage.

ered as hysteretic functions of liquid saturation. The simplest

characteristic curves are non-hystereticthe capillary pressure
and relative permeabilities depend only on the local saturation
at the current time. A more sophisticated approach is a
hysteretic formulation, in which capillary pressure and relative
permeabilities depend not only on the current value of the local
saturation, but on the history of the local saturation and the
process that is occurring: drainage (replacement of wetting
phase, water, with non-wetting phase, CO2) or wetting
(replacement of non-wetting phase with wetting phase, also
known as imbibition) (Doughty, 2006, Fig. 3).
The use of hysteretic characteristic curves is not so critical
for the simulation of CO2 injection periods when the plume is
continuously growing, because all locations follow the
primary drainage branch of the capillary pressure curve at
all times, and this branch can be replicated using a nonhysteretic formulation (Doughty, 2006). However, for postinjection periods, when the CO2 plume moves upward due to
buoyancy forces, different locations experience drainage and
wetting at different times, necessitating the use of a hysteretic
formulation (Doughty, 2006). This study employs hysteretic

Fig. 3 Illustrations of hysteretic paths (capillary pressure/

relative permeability) for the non-wetting phase, where
Sgmax is the maximum gas saturation.

functions (relative permeability and capillary pressure) used

by Doughty (2006).
Four different cases are studied. In case A the relative
permeability during drainage without capillary pressure
hysteresis is considered while in case B relative permeability
hysteresis with a maximum residual gas saturation of 0.25 is
considered. In case C both relative permeability and capillary
pressure hysteresis without considering the maximum residual gas saturation is tested. Finally, case D is the same as case
C with maximum residual gas saturation set to 0.25. Cases A
and B are non-hysteric formulation while cases C and D
hysteric formulation. All these cases were tested at porosity of
38% and permeability of 4000 md and for cases A and D, and
the simulations are repeated using low values of porosity (15%)
and permeability (1000 md) to evaluate the effects of changes
on the modelling output. The objective of these exercises is to
assess how the changes of these parameters can affect the
intermediate storage of carbon dioxide.

3.

Parameters for intermediate storage

Reproducing CO2 back from a reservoir thus to a large extent

depends upon several factors which need to be considered. It
is assumed that storage sites are well characterized so that
extensively faulted and fractured sedimentary basins or parts
thereof, are avoided and the presence of wells penetrating the
subsurface that can create potential CO2 leakage pathways are
well identified at the storage site. It is further assumed that
injection of CO2 into formations is well controlled so that
exceeding pressures that can damage the cap rock are
avoided. The parameters for intermediate storage are highlighted in this section.

3.1.

Depth of injection and density

The depth of CO2 injection and density of CO2 are important

parameters to consider for intermediate storage of carbon
dioxide. The two parameters are interrelated and both depend
on the temperature and pressure conditions in the subsurface.
To geologically store CO2, it must first be compressed, usually
to a dense-fluid state known as supercritical, i.e., at a
temperature and pressure above the critical temperature and
pressure of carbon dioxide (31.1 8C and greater than 73.9 bar).
The critical point represents the highest temperature and
pressure at which the substance can exist as a vapour and
liquid in equilibrium. Depending on the rate that temperature
increases with depth (the geothermal gradient), the density of
CO2 will increase with depth, until at about 800 m or greater,
the injected CO2 will be in a dense supercritical state.
The efficiency of CO2 storage in geological media, defined as
the amount of CO2 stored per unit volume (Brennan and
Burruss, 2003), increases with increasing CO2 density. Cold
sedimentary basins, characterized by low temperature gradients, are more favourable for CO2 storage (Bachu, 2003)
because CO2 attains higher density at shallower depths (700
1000 m) than in warm sedimentary basins, characterized by
high temperature gradients where dense-fluid conditions are
reached at greater depths (10001500 m). Storage in reservoirs
at depths less than approximately 800 m may be technically

international journal of greenhouse gas control 2 (2008) 502510

and economically feasible, but the low storage capacity

of shallow reservoirs, where CO2 may be in the gas phase,
could be problematic to consider this option for intermediate
storage.

3.2.

CO2 flow and transport processes

Understanding the CO2 flow and transport processes is also

important when selecting the sites to be used for intermediate
storage. The rate of fluid flow depends on the number and
properties of the fluid phases present in the formation, where
the fluid phases may be miscible or immiscible. The presence
of several different phases may decrease the permeability and
slow the rate of migration. This can in turn enhance the
dissolution process, which is not favourable for intermediate
storage.
As CO2 migrates through the formation, some of it will
dissolve into the formation water and some of it is retained in
the pore space by capillary forces, which can also dissolve with
time, commonly referred to as residual CO2 trapping, which
may immobilize significant amounts of CO2 (Obdam et al.,
2003; Kumar et al., 2005). Water saturated with CO2 is slightly
denser (approximately 1%) than the original formation water,
depending on salinity (Enick and Klara, 1990; Bachu and
Adams, 2003). With high vertical permeability, this may lead to
free convection, replacing the CO2-saturated water from the
plume vicinity with unsaturated water, producing faster rates
of CO2 dissolution (Lindeberg and Wessel-Berg, 1997; EnnisKing and Paterson, 2003).

3.3.

Storage mechanisms

Storage mechanisms are the effective means of securing

permanent storage of CO2 in geological formations. It is the
most irreversible of these that must therefore be avoided or
minimized when considering intermediate storage, typically
by applying reservoir engineering principles. The storage
mechanism known as physical trapping of CO2 below lowpermeability seals or cap rocks (stratigraphic trapping), such
as very-low-permeability shale or salt beds, is the principal
means to store CO2 in geological formations. Structural traps
include those formed by folded or fractured rocks. Both of
these types of traps are suitable for intermediate storage of
CO2, although, care must be taken not to exceed the allowable
overpressure to avoid fracturing the cap rock or re-activating
faults.
Hydrodynamic trapping can occur in saline formations that
do not have a closed trap, but where fluids migrate very slowly
over long distances. When CO2 is injected into a formation, it
displaces saline formation water and then migrates upwards
due to its buoyancy relative to water. When it reaches the top
of the formation, it continues to migrate as a separate phase
until it is trapped as residual CO2 saturation or in local
structural or stratigraphic traps within the sealing formation.
In the short term this may not affect intermediate storage, but
in the longer term significant quantities of CO2 can dissolve in
the formation water and then migrate with the groundwater.
Carbon dioxide in the subsurface can undergo a sequence
of geochemical interactions with the rock and formation
water, a mechanism known as geochemical trapping. First,

505

when CO2 dissolves in formation water, a process commonly

called solubility trapping occurs. With solubility trapping once
CO2 is dissolved, it no longer exists as a separate phase,
thereby eliminating the buoyant forces that drive it upwards.
This mechanism while very important for permanent storage
of CO2, it does not favour intermediate storage. Mineral
trapping is believed to be comparatively slow, potentially
taking thousands of years or longer so it will not have an
immediate effect on the intermediate storage of CO2.

3.4.

Reservoir heterogeneity

Reservoir heterogeneity also affects CO2 storage efficiency and

hence intermediate storage. The density difference between
the lighter CO2 and the reservoir oil and/or saline water leads
to movement of the CO2 along the top of the reservoir,
particularly if the reservoir is relatively homogeneous and has
high permeability it is well suited for intermediate storage of
CO2. On the other hand, reservoir heterogeneity may have a
positive effect, slowing down the rise of CO2 to the top of the
reservoir and forcing it to spread laterally, giving more
complete invasion of the formation and greater storage
potential (Bondor, 1992; Kovscek, 2002; Flett et al., 2005). This,
however, is not suitable for intermediate storage purposes due
to its contribution to enhanced dissolution.

3.5.

The presence of impurities

The presence of impurities (e.g. SOx, NOx, H2S) in the CO2 gas
stream affects the engineering processes of capture, transport
and injection, as well as the trapping mechanisms and
capacity for CO2 storage in geological media (IPCC, 2005).
Gas impurities in the CO2 stream affect the compressibility of
the injected CO2 (and hence the total volume to be stored) and
reduce the capacity for storage in free phase, because of the
storage space taken by these gases. This will influence
intermediate storage by competing the storage space to be
occupied by the CO2.

3.6.

Types of reservoirs

3.6.1.

Saline aquifers

Saline formations are deep sedimentary rocks saturated with

formation waters or brines containing high concentrations of
dissolved salts. At depths below about 8001000 m, CO2 has a
liquid-like density that provides the potential for efficient
utilization of underground storage space in the pores of saline
formations. In saline formations, the comparatively large
density difference between CO2 and formation water creates
strong buoyancy forces that drive CO2 upwards. The upward
migration of the buoyant plume of injected CO2, however, may
not be evenly distributed. This is because the presence of a
lower permeability layer acts as a barrier and causes the CO2 to
migrate laterally, filling any stratigraphic or structural trap it
encounters as in the case of the Sleipner storage site. This
creates a condition where part of the injected CO2 to become
dissolved in the formation water due to solubility trapping
coupled with delayed upward migration.
Furthermore solubility trapping eliminates the buoyant
forces that drive free-phase CO2 upwards. Also free convection

506

international journal of greenhouse gas control 2 (2008) 502510

replaces the CO2-saturated water from the plume vicinity with

unsaturated water, producing faster rates of CO2 dissolution.
Once CO2 is dissolved in the formation fluid, it migrates along
with the regional groundwater flow and thus will not be
available for retrieval or reproduction. Moreover residual CO2
trapping may immobilize significant amounts of CO2 (see
Section 5). The combined effects of dissolution, solubility
trapping and residual trapping thus could limit the use of deep
saline aquifers for intermediate storage. Proper injection
planning such as gas-phase injection may reduce the
dissolution of CO2 in the formation water and thus facilitates
intermediate storage in saline aquifers. If the injected CO2 is
contained in a closed structure (no flow of formation water), it
will take much longer for CO2 to completely dissolve because
of reduced contact with unsaturated formation water and thus
facilitating intermediate storage in deep saline aquifers
(Section 4). Moreover if gas impurities exist this could affect
the rate and amount of CO2 storage through dissolution and
precipitation.

The depth of injection and hence the density can vary when
storage is considered in depleted gas fields. If CO2 is injected
into a gas reservoir, a single miscible fluid phase consisting of
natural gas and CO2 is formed locally. Usually in natural gas
reservoirs, CO2 is more viscous (flows less easily) than natural
gas. Depleted gas reservoirs can be potential candidates for
intermediate storage of carbon dioxide due to similarity in
storage characteristics as natural gases.

3.6.2.

A number of simulations with idealized aquifers with open

boundaries and a closed aquifer (anticlinal) structure are first
considered to assess their suitability for intermediate storage.
Fig. 4 shows simulation results of both aquifer types with the
relative permeability during drainage accounted but the
capillary pressure hysteresis neglected. It can be seen that
when the injected CO2 is contained in a closed structure (no
flow of formation water) it will take much longer for the CO2 to
dissolve because of reduced contact with unsaturated formation water and this facilitates intermediate storage in deep
saline aquifers and similar structures in depleted oil and gas
fields.
The closed structure is taken as the reference (case A) to
assess the effects of change of parameters cases BD which is
presented in Fig. 5. Consideration of the relative permeability
during drainage only (case A), relative permeability hysteresis
during drainage with a maximum residual gas saturation of

Depleted oil and gas fields

The depths below about 8001000 m are ideal for CO2 injection
because CO2 attains a liquid-like density that provides the
potential for efficient storage in oil and gas fields. Depths up to
1500 m can be tolerated because the density and specific
volume become nearly constant. Carbon dioxide injected into
an oil reservoir may be miscible or immiscible, depending on
the oil composition and the pressure and temperature of the
system. In oil reservoirs, the buoyant plume of injected CO2
migrates upwards, but may not be evenly distributed due to
reservoir heterogeneity. If the reservoir is relatively homogeneous and has high permeability, it provides ideal site for
intermediate storage of CO2. Depleted fields will not be
adversely affected by CO2 (having already contained hydrocarbons) and if hydrocarbon fields are still in production, a CO2
storage scheme can be optimized to enhance oil (or gas)
production (IPCC, 2005).

3.7.

Duration of intermediate storage

Ideally the duration of intermediate storage should be short to

minimize the effects of residual saturation and dissolution.
The fate of the CO2 stored after the end of the injection period
is discussed in Section 4 including the optimum period for
intermediate geological storage.

4.

Simulation results

Fig. 4 Mass of CO2 for the immiscible and dissolved components in the case of closed and open boundary aquifer types.

international journal of greenhouse gas control 2 (2008) 502510

507

Fig. 5 Mass of CO2 for the immiscible and dissolved components showing hysteric models (cases A and B) and non-hysteric
models (cases C and D).

Fig. 6 Plan view of simulation output that shows the top

most layer of the models of non-hysteric (case A) and
hysteric (case D).

0.25 (case B), both relative permeability and capillary pressure

hysteresis (drainage and imbibition) without a maximum
residual gas saturation (case C) and case D is the same as case
C with maximum residual gas saturation of 0.25. Setting
values for the maximum residual gas saturation whether nonhysteretic (cases A and B) or hysteretic (cases C and D) have no
significant effect on the simulation output. In the nonhysteretic cases in general the mass of the free CO2 is large
and the mass of the soluble CO2 is small relative to the cases C
and D, where both the effects of relative permeability and
capillary pressure hysteresis assumed (Fig. 5). The results
show that consideration of the relative permeability hysteresis during drainage alone with or without setting the
maximum gas saturation values does not contribute much
to the residual or phase trapping, cases A and B. Rather more
CO2 is trapped when CO2-water capillary pressure hysteresis
effects are taken into account, cases C and D (e.g. Fig. 6). This
will affect intermediate storage by reducing the mass of the
immiscible component that would otherwise be available.
By comparing modelling results of cases A (reference, nonhysteretic) and D (hysteretic accounted) one can roughly
estimate the relative amount of trapped CO2 due to the

Fig. 7 Relative mass fraction of the immiscible CO2 for cases BD relative to case A.

508

international journal of greenhouse gas control 2 (2008) 502510

Fig. 8 Mass of CO2 for the immiscible and dissolved components comparisons of low and high porosity cases A and D.

capillary forces, which may exist as a residual gas or joined the

dissolved component after injection has ceased. Fig. 7 shows
the relative mass of the supercritical CO2 (immiscible or free
phase) normalized to the amount of CO2 originally available in
the non-hysteric case A at the end of the injection period (20
years). The difference between the curves show that in all
cases (BD) relative to case A, residual gas continues to
increase while the supercritical CO2 mass decreasing with
time after the injection period and constitutes roughly about

36% (Figs. 6 and 7). This can be used to roughly estimate the
duration of intermediate storage, in this modelling the
optimum being less than 15 years. The trapped CO2 become
dissolved as shown in Fig. 5 by the increase in the amount of
dissolved component when hysteric formulation considered
in the modelling.
A change in the porosity from 38 to 15% and permeability
from 4000 to 1000 md on the simulation results were tested in
the non-hysteretic (case A) and hysteretic (case D) models to
assess the effects of these parameters. Permeability changes
alone do not have an effect on the model results since relative
permeability is considered every time in the simulation.
However, the change in the porosity from 38 to 15% with low
permeability (1000 md) slightly increases the amount of the
supercritical CO2 and significantly reduces the dissolved
component due to probably less contact areas whether the
hysteresis effects are included or omitted (e.g. Figs. 8 and 9).
Thus increasing the residual gas trapping since the maximum
residual gas saturation is inversely correlated to the porosity
(Holtz, 2002). This implies that determination of accurate
porosity value is significant during simulation of CO2 storage
in geological formations, especially when intermediate storage is considered because it significantly affects the residual
gas saturation or trapping.

5.
Fig. 9 Plan view of simulation output that shows the top
most layer of the models which demonstrate effects of
parameter changes in case D (hysteric): low porosity
permeability (increased mass of supercritical CO2 and
hence relatively high residual gas saturation versus high
porositypermeability (decreased mass of supercritical
CO2 and hence relatively low residual gas saturation).

Discussions and conclusions

Reproducing CO2 back from a reservoir depends very much up

on the parameters such as the depth of injection and density,
CO2 flow and transport processes, storage mechanisms,
reservoir heterogeneity, the presence of impurities, the type
of the reservoirs and the duration of intermediate storage. At
depths below about 8001000 m where CO2 has a liquid-like
density provides the potential for intermediate storage. Once

international journal of greenhouse gas control 2 (2008) 502510

CO2 is injected into the geological formations, the primary flow

and transport mechanisms that control the spread of CO2 is
governed by many processes (Section 3) that can affect
intermediate storage in geological media.
For example in the case of the Utsira Formation up to 18% of
the CO2 injected can be dissolved in the formation water in the
lifetime of the project (Holloway et al., 2004). This implies
roughly about 1% of the injected CO2 can be dissolved in the
formation water annually. Similar results are obtained from
this study for the non-hysteric cases A and B and slightly
higher 1.52% in the hysteric formulation, cases C and D. In
systems with slowly flowing water, reservoir-scale numerical
simulations show that, over tens of years, a significant
amount, up to 30% of the injected CO2, will dissolve in
formation water (Doughty et al., 2001). This emphasizes the
need for short duration of intermediate storage, not more than
15 years.
Residual trapping with CO2 saturations can cause significant reduction in the amount of CO2 that can be reproduced
and thus affects intermediate storage regard less the reservoir
type. While this effect is formation-specific, Holtz (2002) has
demonstrated that residual CO2 saturations may be as high as
1525% for many typical storage formations. This study
showed a conservative estimation of 36%, which is under
estimated because it assumes relative value instead of
absolute calculation of the mass-balance due to limitations
in the software used. However, the amount of trapped gas
increases with the storage time (Fig. 7) which strengthens
further the need for short duration of storage to minimize the
mass of residual gas trapped. This study also demonstrates the
use to appropriate values of porosity as model out puts are
sensitive to this parameter which is inversely correlated to the
amount of maximum gas trapped (Holtz, 2002). The permeability should be high to allow injection of the CO2 but should
not be extremely high since this can affect the rate of
dissolution as well as the amount of trapped gas (Fig. 8). The
porosity and permeability used in the models tested may not
be representative for a particular storage site but can provide a
basis for parameter selection when actual sites are to be
considered for intermediate storage.
Storage mechanisms such as dissolution due to free
convection and solubility trapping do not favour especially
the use of deep saline aquifers with open boundaries for
intermediate storage. In the absence of impurities homogeneous reservoirs with high permeability for efficient
injection of CO2 are ideal for intermediate storage over short
period of time to reduce dissolution of the stored CO2 and
residual gas trapping.
Finally, the type of reservoir is among the major parameter
for selecting sites for intermediate storage of CO2. Therefore, it
is concluded that if the injected CO2 is contained in a closed
structure (structural traps) where no flow of formation water
occurs it will take much longer for CO2 to completely dissolve
because of reduced contact with unsaturated formation water.
In combination with proper injection planning such as gasphase injection may reduce the dissolution of CO2 in the
formation water and thus facilitating intermediate storage in
deep saline aquifers. Depleted oil and gas homogeneous
reservoirs with high permeability forming structural traps (e.g.
anticlinal structures) are good candidates for intermediate CO2

509

storage. Depleted gas reservoirs can be potential candidates

for intermediate storage of carbon dioxide due to similarity in
storage characteristics as natural gases.

Acknowledgement
The first author would like to thank The Bellona Environmental Foundation in Norway for the encouragements and
support to prepare this manuscript.

references

Bachu, S., 2003. Screening and ranking of sedimentary basins

for sequestration of CO2 in geological media. Environmental
Geology 44 (3), 277289.
Bachu, S., Adams, J.J., 2003. Sequestration of CO2 in geological
media in response to climate change: Capacity of deep
saline aquifers to sequester CO2 in solution. Energy
Conversion and Management 44 (20), 31513175.
Bondor, P.L., 1992. Applications of carbon dioxide in enhanced
oil recovery. Energy Conversion and Management 33 (5),
579586.
Brennan, S.T., Burruss, R.C., 2003. Specific Sequestration
Volumes: A Useful Tool for CO2 Storage Capacity
Assessment. USGS OFR 03-0452, available at http://
pubs.usgs.gov/of/2003/of03-452/.
Doughty, C., 2006. Modelling geologic storage of carbon dioxide:
comparison of non-hysteretic and hysteretic characteristic
curves. In: Proceedings, Tough Symposium 2006, Lawrence
Berkeley National Laboratory, Berkeley, California, May, pp.
1517.
Doughty, C., Pruess, C.K., Benson, S.M., Hovorka, S.D., Knox,
P.R., Green, C.T., 2001. Capacity investigation of brinebearing sands of the Frio Formation for geologic
sequestration of CO2. In: Proceedings of First National
Conference on Carbon Sequestration, CD-ROM USDOE/
NETL-2001/1144, Paper P.32, 1417 May 2001, Washington,
DC, United States Department of Energy, National Energy
Technology Laboratory, p. 16.
Enick, R.M., Klara, S.M., 1990. CO2 solubility in water and brine
under reservoir conditions. Chemical Engineering
Communications 90, 2333.
Ennis-King, J.P., Paterson, L., 2003. Role of convective
mixing in the long-term storage of carbon dioxide in
deep saline formations. Presented at Society of
Petroleum Engineers Annual Technical Conference and
Exhibition, Denver, Colorado, 58 October 2003, SPE Paper
No. 84344.
Flett, M.A., Gurton, R.M., Taggart, I.J., 2005. Heterogeneous
saline formations: Long-term benefits for geo-sequestration
of greenhouse gases. In: Proceedings of the 7th
International Conference on Greenhouse Gas Control
Technologies (GHGT-7), vol. I, September 59, 2004,
Vancouver, Canada, pp. 501510.
Holloway, S., Chadwick, A., Lindeberg, A.E., CzernichowskiLauriol, I., Arts, R., 2004. Best Practice Manual. Saline
Aquifer CO2 Storage Project (SACS), 53 pp.
Holtz, M.H., 2002. Residual gas saturation to aquifer influx: a
calculation method for 3-D computer reservoir model
construction. SPE Paper 75502, Presented at the SPE Gas
Technologies Symposium, Calgary, Alberta, Canada, April
2002.
IEA, 2005. World Energy Outlook 2004, OECD and International
Energy Agency Report, Paris, France.

510

international journal of greenhouse gas control 2 (2008) 502510

IPCC, 2005. In: Metz, B., Davidson, O., de Coninck, H.C., Loos, M.,
Meyer, L.A. (Eds.), IPCC Special Report on Carbon Dioxide
Capture and Storage. Prepared by Working Group III of the
Intergovernmental Panel on Climate Change, Cambridge
University Press, Cambridge, United Kingdom/New York,
NY, USA, 442 pp.
Kovscek, A.R., 2002. Screening criteria for CO2 storage in oil
reservoirs. Petroleum Science and Technology 20 (78),
841866.
Kumar, A., Noh, M.H., Sepehrnoori, K., Pope, G.A., Bryant, S.L.,
Lake, L.W., 2005. Simulating CO2 storage in deep saline
aquifers. In: Benson, S.M. (Ed.), Carbon Dioxide Capture for
Storage in Deep Geologic FormationsResults from the CO2

Capture Project, v.2: Geologic Storage of Carbon Dioxide

with Monitoring and Verification. Elsevier, London, pp.
898977.
Lindeberg, E., Wessel-Berg, D., 1997. Vertical convection in an
aquifer column under a gas cap of CO2. Energy Conversion
and Management 38 (Suppl.), S229S234.
Obdam, A., van der Meer, L.G.H., May, F., Kervevan, C., Bech, N.,
Wildenborg, A., 2003. Effective CO2 storage capacity in
aquifers, gas fields, oil fields and coal fields. In: Gale, J., Kaya,
Y. (Eds.), Proceedings of the 6th International
Conference on Greenhouse Gas Control Technologies
(GHGT-6), vol. I, 14 October 2002, Kyoto, Japan, Pergamon,
pp. 339344.