Abstract

The feasibility of carbon dioxide (CO

2
) injection into an extensive aquifer in the North Sea is assessed. The impact of seal
permeability, horizontal permeability, perforation interval, number of wells, aquifer size and cap rock size on the effectiveness
of large-scale storage is assessed. A compositional numerical simulation is performed in the period of 50 years (30 years of
injection and 20 years post-injection). We study volumes equivalent to the emissions of a large power station in the UK, 10
Mtonnes/year and find the necessary aquifer dimensions, cap rock size and horizontal permeability to allow safe storage
assuming that the pressure increases by no more than 10%.
We find no impact of the number of wells on field-average pressure response as long as the total injection rate remains
constant. However, using more wells enables CO
2
to be trapped by immobilization and dissolution while giving poor sweep
efficiency. In order to achieve the target injection rate with 5 injectors, two appropriate aquifer dimensions are proposed: an
area of 3,850 km
2
with thickness of 1,260 m; and an area of 11,550 km
2
with thickness of 630 m. For the same aquifer volume,
thickness plays a more important role on pressure response than area. Only deeper layers of the aquifer should be perforated in
order to minimize pressure build-up and enhance CO
2
displacement efficiency.
Introduction
CO
2
emissions contribute towards the greenhouse effect and climate change (Ghanbari et al., 2006). The majority of
anthropogenic CO
2
comes from power and industry sectors, for example, fossil fuel combustion (IPCC, 2005). Carbon Capture
and Storage or CCS has generated considerable interest because it is a way of reducing these emissions (Holloway et al.,
2006). Various geological sites are considered suitable for storage in CCS, including depleted oil and gas reservoirs, deep
saline aquifers and deep unminable coal seams (Gale, 2004). In the past, the main interest of CO
2
injection relied on oil or gas
reservoirs as an enhanced oil recovery (EOR) technique where CO
2
and the remaining oil in place become miscible and the oil
can therefore be extracted from the reservoir. However, saline aquifers are currently considered as potential sequestration sites
since they have a large estimated capacity and wide distribution throughout the globe (Gale, 2004, Nicot, 2008). Saline aquifer
sequestration was first mentioned in 1992 (Van der Meer, 1992) and currently, there are several successful projects of aquifer
injection, for example, Sleipner (Norway), In Salah (Algeria), Ketzin (Germany), and K12B (Netherlands) which prove the
feasibility of this emerging storage option. However, saline aquifer storage may cause several problems including migration
into groundwater leading to contamination (Gale, 2004) and risk of overpressure, causing fracturing and possible leakage due
to its shallow position. Pressure build-up is also one of the associated risks for aquifer storage, as large amounts of fluid are
added. This could eventually induce fracturing or aquifer deformation which negatively impacts storage security.
Four types of trapping mechanism occur at different timescales when CO
2
is sequestered; structural trapping, residual
trapping, solubility trapping and mineralization. Structural trapping involves capturing the majority of injected CO
2
in the form
of mobile gas beneath structural or stratigraphic traps; seal integrity is therefore the most significant factor in determining the
storage security for both short and long term (Kumar et al., 2004; Ngheim, 2009). However, the risk of this storage method is
the highest due to the uncertainties of the field characteristics and and relating geomechanic effects. Residual trapping occurs
when CO
2
is trapped in the pore space as an immobile phase by taking the advantage of the capillary and wettability effects
(Gale, 2004; Ngheim et al., 2009). This process occurs when brine starts invading the CO
2
plume and traps the supercritical
CO
2
in the pores space (Kumar et al., 2004). Capillary trapping is considered as the most rapid trapping mechanism (Qi et al.,
2009). Also, as CO
2
is highly soluble in brine, solubility trapping can occur (Ngheim et al., 2009). The degree of CO
2
solubility depends on several factors including salinity, temperature and pressure. When there is a drop in pressure, CO
2
can
IPTC 14464
Carbon Dioxide Storage Potential in the North Sea
Nalinee Chamwudhiprecha, Martin J Blunt, SPE - Imperial College London
Copyright 2011, International Petroleum Technology Conference
This paper was prepared for presentation at the International Petroleum Technology Conference held in Bangkok, Thailand, 79 February 2012.
This paper was selected for presentation by an IPTC Programme Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the International Petroleum Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily
reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at IPTC are subject to publication review by Sponsor Society
Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the International Petroleum Technology
Conference is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435
2
come out of solution leading to further migration. Solubility and residual trapping are considered as safe storage mechanisms
as the risk of leakage does not depend directly on the integrity of the cap rock (Suekane, 2007). CO
2
can also react with solid
minerals present in formation water and form precipitates, which considered as the ultimate desirable method of storage due to
its long-term integrity. However, mineralization is not considered in this study as it takes up to hundreds or thousands of years
to have a significant effect in the sandstone aquifers we consider due to the slow kinetics of precipitation reactions (Kumar et
al., 2004).
The North Sea is potentially one of the most effective storage sites with, for instance, the Inner Moray Firth basin and the
Bunter Sandstone, containing very large volumes of saline water-bearing reservoir rocks (Holloway et al., 2006). A number of
simulation studies have been performed to study different aspects of storage design e.g. Pruess et al., 2003; Kumar et al., 2004;
Mo & Akervoll, 2005; Qi et al., 2007; Kartikasurja et al., 2008; Primera et al., 2009; Sifuentes et al., 2009; Nghiem et al.,
2009; Vandeweijer et al., 2009. However, a limited amount of research has concentrated on the pressure response. Van der
Meer (1992) first suggested the idea of the injection limitation due to an increase in pore pressure. Van der Meer et al. (2006)
emphasized the effect of injecting additional fluid into an aquifer on fluid volumes and pressures in the total storage system by
focusing on CO
2
that dissolves in water. Yang (2008) proposed that rapid pressure build-up during injection greatly limited
storage volume for a closed system and in order to relieve the injection pressure and the storage capacity, brine could be
produced from the reservoir. Birkholzer et al. (2009) investigated the region of influence from CO
2
injection in terms of brine
displacement and pressure perturbation. Oruganti and Bryant (2009) studied the effect of the existence of sealing faults
towards pressure build-up.
This paper assesses the feasibility of CO
2
injection into an extensively large aquifer in the North Sea. Required aquifer
properties and operational constraints, including seal permeability, horizontal permeability, perforation interval, number of
wells, aquifer size and cap rock size are studied by studying the field pressure response, well injectivity and CO
2
distribution.
Numerical Simulations
Numerical simulations are performed using the compositional Eclipse E300 software. A base case model is constructed by
using aquifer dimensions and properties from an existing aquifer in the North Sea, the Bunter Sandstone (Bentham, 2006). The
model is assumed to be homogeneous. Initially, injection rate and number of wells are studied in order to define the injection
capability of this aquifer. By considering the target injection rate to be 10 Mtonnes/year (10
10
kg/year) which corresponds to
the estimation of CO
2
emission rate from large power stations, supplying 1-3 GW of coal-powered electricity, in the UK; for
example, the West Burton or Ratcliffe on Soar power stations (Holloway, 2006). However, the simulation results will show
that the base case aquifer is not sufficiently large to provide the planned storage capacity. Aquifer dimensions are therefore
varied and delineated based on the properties of Bunter Sandstone to allow sufficient storage volume. As cap rock is one of the
most significant factors for a potential storage site, cap rock size and permeability are studied to justify the appropriate
thickness, area and permeability. Also, aquifer permeability is varied in order to determine the appropriate value to store the
planned volume without negatively affecting aquifer pressure. In terms of operational constraints, perforation interval is
studied in order to optimize CO
2
underground distribution.
For every case, the injection strategy is controlled by injection rate. The simulation period is 50 years including 30 years of
pure CO
2
injection period and 20 years post-injection. Pressure response, CO
2
leakage and CO
2
phase distribution are
observed. Pressure response is observed in the form of the field pressure increase and average bottomhole pressure to study the
effect on pressure build-up and injectivity, respectively. Field pressure increase is calculated by considering the percentage
difference between the maximum field average pressure occurring at any time during the simulation period and the average
initial pressure of the field. In practical operations, field pressure has to be controlled so as not to exceed the fracture pressure
which could cause CO
2
migration out of the storage site. Values for the fracture-closure pressure gradients are site-specific and
can be determined from direct or indirect testing, or estimated using formation-specific default values (Zhou, 2008). For the
study in this paper, feasible schemes are determined based on a field pressure increase of up to 10% of the initial pressure.
Also, CO
2
phase distribution and CO
2
distribution around the injector are observed in order to assess the security of the storage
over time. Overall, the required dimensions and properties of a saline aquifer that is capable of supporting the target CO
2
storage are provided.
Model Description
A homogeneous box-shaped model of 215,600 grid cells is constructed as a base case to represent a laterally extensive aquifer.
Compartmentalization is not considered in order to allow fluid transmission without excessive pressure increase. As CO
2
injected in an unconfined aquifer, it will likely disperse over a large area and in low concentrations over time (Bentham, 2005).
The aquifer thickness is 126 m with an overlying seal which forms the first layer of the model. The seal layer has low
permeability and high capillary entry pressure to enable CO
2
trapping. The model size and properties are gathered from the
Bunter Sandstone which proves to have high storage potential amongst the reservoir rocks of the Southern North Sea
(Holloway et al., 2006). In terms of perforation policy, half of the aquifer layers are completed. The input parameters for the
base case simulation are shown in Table 1.
Relative permeability curves are based on experimental data of supercritical CO
2
-brine at in-situ conditions of the Viking
Formation sandstone, Alberta, Canada (Bennion & Bachu, 2006). Hysteresis effects are taken into account by using both
IPTC 14464
3
drainage and imbibition curves. Drainage occurs at the leading edge of the CO
2
plume where gas displaces water, while at the
trailing edge water displaces gas in an imbibition process (Juanes et al., 2006). By using both curves, the residual trapping
mechanism is activated. Relative permeability curves are shown in Figure 1. Also, a reversal of the drainage is taken into
account by calculating a scanning curve with Killoughs method (1976). Capillary pressure is calculated by the method of Van
Genuchten (1980). Capillary pressure for the seal layer is scaled with permeability by the use of the Leverett J-function.
Parameter Value
Grid cells (i, j, k) 140 110 14
Grid block size 500 m 500 m 9 m
Aquifer volume 485 km
3
Porosity 0.18
K
v
/K
h
ratio
0.1
K
h
500 mD
Temperature 62.8 C
Top depth 1717 m
Rock compressibility
1.510
5
bar
-1
Pressure at 3200 m 225 bars
Table 1 - List of aquifer properties for the base case.
Figure 1 - Relative permeability curves (Bennion & Bachu, 2006).
Fluid Properties
CO
2
has a critical pressure of 73.8 bars and a critical temperature of 304.2 K and is a supercritical fluid at aquifer conditions
(Ngheim et al., 2009). Two phases are considered in the simulation studies: a CO
2
-rich phase (gas phase), and an H
2
O-rich
phase (liquid phase). Three compositions are included: CO
2
, H
2
O and NaCl while salts are assumed to be in the liquid phase
only. By the use of CO2STORE function, mutual solubilities of both CO
2
in water and water in CO
2
are calculated to include
the effect of chloride salts in the aqueous phase (Spycher & Pruess, 2005). An accurate prediction of the aqueous phase density
is another important aspect for modeling CO
2
storage in saline aquifers (Nghiem et al., 2009). Gas density is calculated by a
cubic equation of state. Brine density is first approximated by pure water density using the method presented by Kell &
Whalley (1975) and then the effects of salt and CO
2
are corrected by using the Ezrokhis method (Zaytsev & Aseyev, 1993).
The CO
2
gas viscosity is calculated using methods of Vesovic et al. (1990) and Fenghour et al. (1999). Formation brine
salinity is assumed to be constant throughout the aquifer at 175,000 ppm.
Results
CO
2
Injection Rate and Number of Injectors
The study initially aims to find the optimum injection rate
per well in order to evaluate the amount of CO
2
that can
practically be stored in the base case model. For every
case, only half of the aquifer layers are perforated. By
varying the injection rate, the pressure response is
observed. The results in Table 2 show that in order to
maintain the field pressure not to exceed 10% of its
original value, only 440,000 tonnes/year of CO
2
can be
injected into this base case model using one well.
Furthermore, if the planned injection rate of 10
Mtonnes/year is to be injected, the pressure will increase
by nearly 100%. Therefore, in order to reach the target
injection rate, either many wells or a larger aquifer has to
be used.
Case
No.
Injection rate Field pressure
increase m
3
/day tonnes/year
1 34,250 23,000 0.5 %
2 148,000 100,000 2.3 %
3 650,000 440,000 10.0 %
4 700,000 473,000 10.7 %
5 800,000 540,000 12.1 %
6 1,000,000 675,000 14.0 %
7 2,070,000 1,400,000 20.0 %
8 2,960,000 2,000,000 30.0 %
9 15,000,000 10,000,000 96.4 %
Table 2 - Results of field pressure increase from injection rate
variation.
The results (Table 3) show that the number of wells (shown in Figure 2) does not have a significant effect on the average
field pressure response. Allowing more wells to be used does not lower the pressure buildup as the total injection rate remains
equal for all the cases. However, the number wells used affect the storage mechanisms as shown in Figure 3. A larger number
of wells lead to more immobile and dissolved CO
2
, hence less mobile CO
2
. As more wells are used, there is more contact area
between CO
2
and water which enhances the rate of dissolution and capillary trapping. The rate of dissolution depends on the
amount of mixing of CO
2
and formation water (Kumar et al., 2004). In terms of CO
2
distribution, for every case, CO
2
tends to
accumulate around the wellbore along the injection period, then gradually spreads out to form a gas layer below the overlying
seal after the injection ceases. Figure 4 shows the CO
2
distribution at the injector for plume radius comparison. For the case
with 1 well, CO
2
is forced to be injected at high injection rate which leads to wider plume radius. For the case with more wells,
0
0.2
0.4
0.6
0.8
1
0 0.2 0.4 0.6 0.8 1
R
e
l
a
t
i
v
e

p
e
r
m
e
a
b
i
l
i
t
y
,

f
r
a
c
t
i
o
n
Water saturation, fraction
Krw drainage
Krg drainage
Krg imbibition
IPTC 14464
4 IPTC 14464
CO
2
tends to accumulate and concentrate around the wellbore and water is not entirely swept in the lower layers. The high rate
in the single-well system makes gravity forces relatively small which enhances sweep efficiency, while in multiple well
scenarios, low rates allow time for the CO
2
to accumulate at the top of the formation.
In conclusion, a sufficient number of wells should be used in order to optimize both the storage mechanism by enhancing
both the immobile and solubility trapping, and the sweep efficiency. However, it is not feasible to inject 10 Mtonnes/year into
the base case aquifer even with more wells as pressure build-up exceeds the 10 % limit of fracture prevention. Therefore, the
remaining option is choose a larger aquifer.
Figure 2 - 2D top view of the base case model showing well locations for each case.
Case No. No. of wells
Injection rate per well
Field pressure increase
m
3
/day Mtonnes/year
1 1 well 15,000,000 10 96.4 %
2 3 wells 5,000,000 3.5 103.9 %
3 5 wells 2,960,000 2 102.6 %
4 7 wells 2,150,000 1.5 103.7 %
5 9 wells 1,670,000 1.1 103.3 %
Table 3 - Injection rate for cases with number of wells variation.
Figure 3 - Results of injected CO2 in different phases for different numbers of injection wells.
PermX(MDARCY)
0.00 125.00 250.00 375.00 500.00
0.00E+00
1.00E+09
2.00E+09
3.00E+09
4.00E+09
5.00E+09
6.00E+09
0 10 20 30 40 50 60
A
m
o
u
n
t

o
f

m
o
b
i
l
e
,

i
m
m
o
b
i
l
e

a
n
d

d
i
s
s
o
l
v
e
d

C
O
2
(
K
G
-
M
)
Time (years)
1 well
3 wells
5 wells
7 wells
9 wells
Mobile CO
2
Dissolved CO
2
Immobile CO
2
5
Figure 4 - 2D cross-section view of CO2 distribution at the injector for different numbers of wells Left: at the end of the injection period (30 years)
Right: at the end of the simulation period (50 years) Top to bottom: no. of wells 1, 3, 5, 7, 9.
Aquifer Dimensions
In order to store 10 Mtonnes/year of CO
2
for 30 years, the aquifer must be sufficiently large in order to accommodate the
storage volume and overcome the limitations on pressure increase. The base case aquifer model is enlarged with the same
properties by using five injectors with the total injection rate of 10 Mtonnes/year. Two means of investigating the size of the
aquifer are used: controlling lateral area and varying thickness and controlling thickness and varying lateral area. For both
cases, required area and thickness are determined by observing the average field pressure to increase within the limit of 10%.
For the first case, controlling area and varying thickness, the area is fixed at 3,850 km
2
according to Bunter Sandstone
aquifers dimensions (Bentham, 2006). The base thickness value is 126 m, according to Bunter Sandstone aquifers
dimensions (Bentham, 2006), which is multiplied by 3, 5, 7, 9 and 10. According to the results of field pressure (Table 4) , it
can be observed that the thickness of 1,260 m (10 times of the base thickness value) gives a pressure increase within the limit
of 10% as thick aquifers allow pressure to dissipate more vertically and hence alleviate pressure build-up. Considering the
bottomhole pressure, the case with the thickness of 126 m gives the highest bottomhole pressure, in other words, the lowest
injectivity owing to the difficulty of the injection into a thin aquifer while the other cases show relatively similar bottomhole
pressure. Therefore, if the aquifer thickness is sufficiently large, there will be no effect on injectivity. In terms of phase
distribution, in the case with high value of thickness, CO
2
tends to be less mobile and more dissolved in water as seen in figure
5. Thicker aquifers lead to more contact area between CO
2
and water which enhances the rate of dissolution. Also, longer
distance of migration enables water to trap CO
2
which results in higher residual gas. The results also correspond to the
previous study stating that the transition of CO
2
from the free gas phase into the dissolved phase in water has a pressure-
reducing effect (Van der Meer & Van Wees, 2006); this is because the effective density of CO
2
in brine is much higher than its
own phase. Considering the CO
2
plume radius, the case with higher value of thickness gives narrower plume radius as CO
2
is
trapped as an immobile phase before reaching the top of the aquifer as seen in Figure 6. Areal displacement is greatly efficient
for the case with the thickness of 126 m while other cases show similar results.
Case No. Thickness Thickness Field pressure increase Average bottomhole pressure (Bars)
1 Thickness 1 126 m 102.6 % 152
2 Thickness 3 375 m 39.4 % 127
3 Thickness 5 630 m 23.6 % 124
4 Thickness 7 882 m 15.6 % 125
5 Thickness 9 1,134 m 11.3 % 125
6 Thickness 10 1,260 m 9.6 % 126
Table 4 - Results of pressure increase and average bottomhole pressure from aquifer thickness variation.
1 well
7 wells
9 wells
5 wells
3 wells
Distance from injection well (km)
-7.5 -5 5 7.5 -2.5 0 2.5
1 well
7 wells
9 wells
5 wells
3 wells
-7.5 -5 5 7.5 -2.5 0 2.5
Distance from injection well (km)
GasSat
0.0000 0.1443 0.2885 0.4328 0.5770
IPTC 14464
6
Figure 5 - Results of injected CO2 in different phases from aquifer thickness variation.
Figure 6 - 2D cross-section view of CO2 distribution at the injector from thickness variation Left: at the end of the injection period (30 years) Right: at
the end of the simulation period (50 years) Top to bottom: Thickness 1, 3, 5, 7, 9, 10.
0
1E+09
2E+09
3E+09
4E+09
5E+09
6E+09
0 10 20 30 40 50 60
A
m
o
u
n
t

o
f

m
o
b
i
l
e
,

i
m
m
o
b
i
l
e

a
n
d

d
i
s
s
o
l
v
e
d

C
O
2
(
K
G
-
M
)
Time (years)
Mobile CO
2
Dissolved CO
2
Immobile CO
2
Thickness x1
Thickness x3
Thickness x5
Thickness x7
Thickness x9
Thickness x10
Thickness x 5
Thickness x 10
Thickness x 9
Thickness x 7
Thickness x 1
Distance from injection well (km)
3 4.5 0 1.5 -3 -4.5 -1.5
Thickness x 3
Distance from injection well (km)
3 4.5 0 1.5
Thickness x 5
Thickness x 10
Thickness x 9
Thickness x 7
Thickness x 1
-3 -4.5 -1.5
Thickness x 3
GasSat
0.0000 0.1443 0.2885 0.4328 0.5770
IPTC 14464
7
For the case that the aquifer thickness is controlled while the area is varied, the thickness is fixed to be 5 times of the base
thickness value, 630 m, three different sizes of aquifer were studied:- 2 times, 2.5 times and 3 times of the base area value
(3,850 km
2
). It can be concluded that with the given thickness of 630 m, the appropriate area is 11,550 km
2
(3 times of the
base area value) as field pressure increase does not exceed the control limit. The results of pressure increase and average
bottomhole pressure are shown in Table 5. Large aquifers can withstand higher overall pressure build-up and also contributes
to higher injectivity as pressure is allowed to disperse throughout a larger area. Lower bottomhole pressure is observed in the
case with larger area which results in higher injectivity. However, there is no significant difference on phase distribution as
the contact area remains the same for all cases. Also, the results show no significant difference on CO
2
underground
distribution.
Case No. Area Area Field pressure increase Average bottomhole pressure (Bars)
1 Area 1 3,850 km
2
23.6 % 124
2 Area 2 7,700 km
2
13.6 % 120
3 Area 2.5 9,625 km
2
10.9 % 119
4 Area 3 11,550 km
2
10.0 % 118
Table 5 - Results of pressure increase from aquifer area variation.
Finally, the effect of thickness and area are compared by fixing the volume of the aquifer to be 4,851 km
3
and varying the
aquifer thickness and area. By maintaining the volume, the results show that thickness plays a more important role on pressure
response than area according to table 6. In other words, for the same value of aquifer volume, the aquifer with larger thickness
value gives a lower pressure increase. In terms of CO
2
distribution, gas tends to be stored more vertically in the case with
higher thickness which leads to narrower plume radius as shown in Figure 7. Also, CO
2
phase distribution is shown in Figure
8. More immobile gas and less mobile gas are present in the case with higher thickness value. However, there is no difference
in the amount of gas dissolved in water. Therefore, the aquifer thickness is one of the most important criteria when
determining a suitable storage site so as to minimize the pressure effect together with increasing the amount of trapped gas.
Case No. Thickness & area Area Thickness Volume Field pressure increase
1 Thickness1 Area1 3,850 km
2
126 m 485.1 km
3
102.6 %
2 Thickness10 Area1 3,850 km
2
1,260 m 4,851 km
3
9.6%
3 Thickness5 Area2 7,700 km
2
630 m 4,851 km
3
13.6%
Table 6 - Properties and results from aquifer thickness and variation.
Figure 7 - CO2 distribution at the injector from area and thickness variation Left: at the end of the injection period (30 years) Right: at the end of the
simulation period (50 years) Top to bottom: Thickness x1 Area x1, Thickness x10 Area x1, Thickness x5 Areax2.
Thickness x10 Area x1
Distance from injection well (km)
-4.5 -3 3 4.5 -1.5 0 1.5
Thickness x5Area x2
Thickness x1 Area x1
Distance from injection well (km)
-4.5 -3 3 4.5 -1.5 0 1.5
Thickness x5 Area x2
Thickness x10 Area x1
Thickness x1Area x1
GasSat
0.0000 0.1443 0.2885 0.4328 0.5770
IPTC 14464
8
Figure 8 - Results of injected CO2 in different phases from aquifer thickness and area variation.
In conclusion, in order to store 10 Mtonnes/year of CO
2
by using 5 wells the aquifer area should be at least 3,850 km
2
with the
thickness of 1,260 m or 11,550 km
2
with the thickness of 630 m to withstand the pressure limitation. The model with the
dimensions of 70 km 55 km 1,260 m will be used for further study of the influence of other constraints (cap rock size,
perforation interval, horizontal permeability, seal permeability).
Seal Permeability
Cap rock integrity is one of the most important factors which helps reduce the risk of CO
2
leakage. By varying the
permeability of the cap rock to be 10 mD, 1 mD, 0.5 mD, 0.1mD and 0.01 mD while maintaining the seal thickness at 126 m,
gas saturation is observed in the uppermost layer (seal layer) as shown in Figure 9. The results show that CO
2
leakage is
observed at the seal permeability of 0.5 mD or higher. Capillary entry pressure is dependent on permeability (pore throat size)
as gas requires greater pressure to displace brine in a microscopic scale. For cases with low seal permeability, CO
2
tends to
migrate horizontally and forms more lateral CO
2
plume while cases with high seal permeability CO
2
is allowed to be stored in
the seal layer which results in thicker CO
2
plume. However, by observing gas distribution in the lower layers, all of the cases
appear to have comparatively similar displacement efficiency. For CO
2
phase distribution, the results are inconclusive.
In terms of pressure effect, Table 7 shows inconclusive results on average field pressure increase. On one hand, a low
permeability seal layer forms a pressure barrier which causes the pressure to buildup extensively higher than the case with
higher permeability. On the other hand, although a low permeability seal prevents CO
2
leakage because of its high capillary
pressure, it also allows a small amount of water to flow over the very large area of the cap rock which helps release the
pressure build-up. Overall, seal permeability can affect field pressure response both positively and negatively. Furthermore, no
effect on bottomhole pressure can be observed. The most appropriate value for seal permeability is 0.1 mD or less to prevent
CO
2
from escaping and provide reasonable pressure increase.
0.00E+00
1.00E+09
2.00E+09
3.00E+09
4.00E+09
5.00E+09
6.00E+09
0 10 20 30 40 50 60
A
m
o
u
n
t

o
f

m
o
b
i
l
e
,

i
m
m
o
b
i
l
e
,

d
i
s
s
o
l
v
e
d

C
O
2
(
K
G
-
M
)
Time (years)
Thickness x1 Area x1
Thickness x10 Area x1
Thickness x5 Area x2
Mobile CO
2
Immobile CO
2
Dissolved CO
2
IPTC 14464
9
Figure 9 - Results of CO2 migration into the seal layer from seal permeability variation Left: at the end of the injection period (30 years) Right: at the
end of the simulation period (50 years) Top to bottom: seal permeability 10 mD, 1 mD, 0.5 mD, 0.1 mD, 0.01 mD and 0.001 mD
Case No. Seal permeability Field pressure increase Average bottomhole pressure (Bars)
1 10 mD 9.7 % 126
2 1 mD 9.2% 126
3 0.5 mD 9.2 % 126
4 0.1 mD 9.6 % 126
5 0.01 mD 9.8 % 126
6 0.001 mD 9.6 % 126
Table 7 - Results of field pressure and injection pressure increase from seal permeability variation.
Perforation Interval
Different perforation intervals of injection wells are studied in order to investigate the effect on underground CO
2
distribution
for storage optimization. The perforation interval is varied to be from layer 2 to 20 (all sand layers) and from layer 10 to 20
(half of the aquifer thickness). As seen in Figure 10, in the case which all the layers are perforated, gas tends to accumulate
mostly in the upper layers. While in the case which half of the layers are perforated, gas is also distributed in the lower layers
which enhances displacement efficiency. Comparatively larger plume radius can be observed in the case which all of the layers
are perforated. CO
2
phase distribution (Figure 11) shows that gas is preferably stored as an immobile phase in the case which
half of the layers are perforated but less gas is in a dissolved phase.
In terms of pressure build-up, lower pressure increase is observed in the case which only half of the aquifer thickness is
perforated as shown in Table 8. As gas is also stored in lower layers, it is trapped before reaching the upper layers which
increases the rate of immobilization. This also results in less field pressure build-up. In terms of injectivity, higher bottomhole
pressure is observed when only half of the layers are perforated which lowers injectivity. This is due to the fact that the same
amount of CO
2
has to be injected through less perforated area. Therefore, in order to optimize the storage capacity and sweep
efficiency with less pressure build-up, only deeper parts of the aquifer should be perforated.
10 mD
1 mD
0.5 mD
0.1 mD
0.01 mD
0.001 mD
Distance from injection well (km)
-4.5 -3 3 4.5 -1.5 0 1.5
Distance from injection well (km)
-4.5 -3 3 4.5 -1.5 0 1.5
10 mD
1 mD
0.5 mD
0.1 mD
0.01 mD
0.001 mD
GasSat
0.0000 0.1443 0.2885 0.4328 0.5770
IPTC 14464
10
Figure 10 - CO2 distribution at the injector from perforation interval variation Left: at the end of the injection period (30 years) Right: at the end of the
simulation period (50 years) Top: layer 2 to 20 perforated Bottom: layer 10 to 20 perforated.
Figure 11 - Results of injected CO2 in different phases from perforation interval variation.
Case No. Perforation policy Field pressure increase Bottomhole pressure (bars)
1 Layer 2 to 20 perforated 10 % 109
2 Layer 10 to 20 perforated 9.6 % 126
Table 8 - Results of field pressure from perforation interval variation.
Horizontal Permeability
For an aquifer to be a potential candidate, permeability and porosity have to be high to provide sufficient storage volume for
planned CO
2
storage. Seven different values of horizontal permeability are used: 50 mD, 100 mD, 200 mD, 250 mD, 300 mD,
400 mD and 500 mD for a constant K
v
/K
h
ratio of 0.1.
High permeability enables CO
2
plume to migrate more laterally as seen in Figure 12. The gas plume beneath the seal layer
is also formed earlier which results in bypassed lower layers which negatively affects sweep efficiency. On the other hand, for
low permeability cases, CO
2
tends to accumulate around the wellbore and the areal extent of CO
2
plume is reduced due to the
difficulty of migration. CO
2
is therefore stored in the lower layers which increases displacement efficiency. In terms of CO
2
phase distribution, for high permeability aquifers, CO
2
is stored more in gas phase and less dissolved in water or immobile
phase, according to Figure 13 as gas rapidly moves through high permeability layers without being trapped by water. Results
of pressure response are shown in Table 9, in terms of injectivity, higher bottomhole pressure is observed for low permeability
aquifers as CO
2
migration path is disrupted and fluid pressure is increased. Therefore, low permeability gives poor injectivity.
However, the effect on field pressure gives a converse result as higher permeability allows pressure to dissipate more quickly
and laterally which leads to higher average field pressure increase.
In conclusion, horizontal permeability is one of the critical parameters which could positively or negatively impact CO
2
storage regarding different perspectives. Low permeability aquifers benefits average pressure build-up, sweep efficiency and
storage efficiency but worsen injectivity and vice versa.
All layers perforated
Half layers perforated
Distance from injection well (km)
-4.5 -3 3 4.5 -1.5 0 1.5
Distance from injection well (km)
-4.5 -3 3 4.5 -1.5 0 1.5
All layers perforated
Half layers perforated
GasSat
0.0000 0.1443 0.2885 0.4328 0.5770
0.00E+00
5.00E+08
1.00E+09
1.50E+09
2.00E+09
2.50E+09
3.00E+09
3.50E+09
4.00E+09
0 10 20 30 40 50 60
A
m
o
u
n
t

o
f

i
m
m
o
b
i
l
e
,

m
o
b
i
l
e

a
n
d

d
i
s
s
o
l
v
e
d

C
O
2
(
K
G
-
M
)
Time (years)
Layer 2 to 20 perforated
Layer 10 to 20 perforated
Mobile CO
2
Immobile CO
2
Dissolved CO
2
IPTC 14464
11
Figure 12 - CO2 distribution at the injector from horizontal permeability variation Left: at the end of the injection period (30 years) Right: at the end of
the simulation period (50 years) Top to Bottom: Kh 50 mD, 100 mD, 200 mD, 250 mD, 300 mD, 400 mD, 500 mD.
Figure 13 - Results of injected CO2 in different phases from horizontal permeability variation.
Distance from injection well (km)
-4.5 -3 3 4.5 -1.5 0 1.5
50 mD
100 mD
200 mD
250 mD
300 mD
400 mD
500 mD
Distance from injection well (km)
-4.5 -3 3 4.5 -1.5 0 1.5
50 mD
100 mD
200 mD
250 mD
300 mD
400 mD
500 mD
GasSat
0.0000 0.1443 0.2885 0.4328 0.5770
0.00E+00
5.00E+08
1.00E+09
1.50E+09
2.00E+09
2.50E+09
3.00E+09
3.50E+09
4.00E+09
4.50E+09
0 10 20 30 40 50 60
A
m
o
u
n
t

o
f

m
o
b
i
l
e
,

i
m
m
o
b
i
l
e
,

d
i
s
s
o
l
v
e
d

C
O
2
(
K
G
-
M
)
Time (years)
Kh=50 mD
Kh=100 mD
Kh=200 mD
Kh=250 mD
Kh=300 mD
Kh=400 mD
Kh=500 mD
Mobile CO
2
Immobile CO
2
Dissolved CO
2
IPTC 14464
12
Case No. Kh Field pressure increase Average bottomhole pressure (Bars)
1 50 mD 6.7 % 140
2 100 mD 7.9 % 133
3 200 mD 8.9 % 130
4 250 mD 9.2 % 129
5 300 mD 9.4 % 128
6 400 mD 9.5 % 127
7 500 mD 9.6 % 126
Table 9 - Results of pressure increase and average bottomhole pressure from horizontal permeability variation.
Discussion
By observing CO
2
underground distribution at the injector, along with phase distribution over time, it can be concluded that
gas initially accumulates around the wellbore and moves vertically until reaching the seal layer. The majority of CO
2
is stored
in the form of mobile gas under the structural trap. Then, CO
2
plume is formed beneath the seal layer and gas tends to spread
more horizontally. Highest gas saturation is always observed at the top of the aquifer layer at the injector. After the injection
ceases, gas tends to spread more horizontally with higher gas saturation at the tail of the plume. Immobilization is dramatically
enhanced at later times as water imbibes CO
2
and traps gas in the pore space. Dissolved gas is also present in relatively high
amount due to the CO
2
solubility in brine with a gradual increase. In practical operations, storage security is the primary goal
of CCS. The risk of CO
2
leakage can be alleviated by eliminating any escape paths which could possibly result from high seal
permeability.
Overall, three phases of gas are considered: mobile, immobile and dissolved. The most preferable trapping mechanism is in
the form of an immobile phase since gas is securely stored in the pore space. While gas could find an escape path when stored
in a mobile phase and the rate of dissolution relies on several conditions. Trapping mechanisms influence the storage security
over time. Therefore, in earlier times of trapping process, cap rock integrity is very significant. Long term storage security can
be enhanced by using high number of wells with only lower sections of the wells completed. Also, aquifer thickness and
permeability play an important role: thicker aquifer and lower permeability give higher extent of immobilization. In practical
operations, several perspectives have to be taken into account including commercial aspects. Storage optimization can be done
by improving injectivity with fewer wells. Drilling more wells is therefore not commercially viable even though
immobilization can be enhanced. One possibility is to use horizontal well in order to increase contacts between CO
2
and brine.
Areal sweep efficiency is small compared to vertical due to the effects of gravity segregation and viscous fingering.
Displacement process is preferable when CO
2
is allowed to be stored in all of the layers and not aggregated around the
wellbore or the top layers. Larger plume radius indicates that more gas easily migrates to the top part of the aquifer without
being trapped which reduces the sweep efficiency. In order to obtain effective displacement, less injectors are to be used with
only lower sections perforated. Also, aquifer thickness should be high, but with low permeability. The displacement process
can be further optimized by injecting CO
2
/brine mixture as mentioned in previous literature (Qi et al., 2009).
The main focus on this paper relies on the effect of several factors on pressure response. Considering practical operations,
field pressure is normally maintained not to exceed the fracture pressure which results in the lack of CO
2
storage security.
Pressure build-up occurs more extensively compared to the CO
2
plume. The effect of pressure can be observed beyond the
radius of tens of kilometers while CO
2
disperses in the aquifer in the extent of 4-5 kilometers. This indicates the significance
of pressure response on a laterally extensive aquifer area. The maximum field pressure is observed at the end of the injection
period. Pressure build-up is expected to be relieved after a certain amount of time after the injection ceases, however, the
pressure effect remains visible. In order to relieve pressure build-up, the aquifer should be enlarged in both area and thickness
to allow higher extent of pressure dissipation in horizontal and vertical direction, respectively. Large aquifer volume is
required in order to overcome the pressure limitation and thickness plays a more important role on pressure than area. Seal
properties also impact pressure build-up: the thicker the seal, the lower the pressure build-up. However, the results are
inconclusive for seal permeability since it could both positively and negatively impact the pressure response. Aquifer
permeability has a negative influence on pressure build-up as high permeability allows pressure to dissipate more quickly
resulting in higher average pressure build-up. In terms of completion policy, only lower sections of the wells should be
perforated.
Pressure build-up greatly affects the maximum injection rate as if the pressure limitation were to be relaxed to 20% instead
of 10%, the base case aquifer would be able to handle double the amount of injection. Also, a smaller aquifer would be
required to safely store the target injection rate when the pressure limitation was relaxed. The proposed aquifer dimensions for
this case (20% limit) would be the thickness of 882 m with the area of 3,850 km
2
or the thickness of 630 m with the area of
7,700 km
2
, whereas the case with pressure limitation of 10%, the aquifer dimensions are the thickness of 1,260 m with the area
of 3,850 km
2
or the thickness of 630 m with the area of 11,550 km
2
. Moreover, if the pressure limit were to be 30%, the
appropriate aquifer thickness would be half the size of the case with 10% limitation while the aquifer area would be one-third.
This indicates that the aquifer dimensions are greatly sensitive to pressure response limitation. This limitation also relies on the
integrity of the seal overlying the aquifer which should be properly defined by geomechanical tests.
Considering the well injectivity, higher injectivity results from lower bottomhole pressure which is preferable for
IPTC 14464
13
operational perspective. This can be achieved by expansive aquifer area, high aquifer thickness and high horizontal
permeability. In terms of well completion, more perforated area is preferable.

Summary and Conclusions
From the simulation studies, we conclude the following:
Number of wells has no impact on pressure response as long as the total injection rate remains constant. However,
larger number of wells enhances immobilization and dissolution, but lowers sweep efficiency.
In terms of aquifer dimensions, a thick aquifer benefits pressure build-up, immobilization, dissolution and injectivity.
Lateral extensive aquifer lowers pressure effect and enhances injectivity without any impact on storage mechanisms.
For the same aquifer volume, thickness plays a more important role on pressure response than area. By increasing the
aquifer thickness, pressure impact can be more efficiently minimized compared to increasing the aquifer area.
Therefore, the aquifer thickness is one of the most significant parameters on determining the injection feasibility.
CO
2
leakage is observed when seal permeability exceeds 0.5 mD. Seal permeability could affect pressure build-up
both positively and negatively. Furthermore, seal permeability has no effect on bottomhole pressure or displacement
efficiency.
For well completion, only deeper layers of the aquifer should be perforated in order to minimize pressure build-up
and enhance long-term storage security and sweep efficiency but this results in a reduced well injectivity.
High value of horizontal permeability benefits higher injectivity but results in higher pressure build-up, poor sweep
efficiency and greater amount of mobile gas.
In order to store 10 Mtonnes/yr of CO
2
with the pressure limitation of 10%, the proposed aquifer dimensions are the
area and thickness of at least 3,850 km
2
, 1260 m or 11,550 km
2
, 630 m, respectively. Appropriate cap rock area and
thickness are 1,411 km
2
and 63 m, relatively with permeability of 0.1 mD or less.
Suggestions for Further Work
A regional aquifer model is required in order to assess the real fracture pressure and large-scale pressure response.
Previous studies already focused on the effect of injecting CO
2
/brine in order to enhance sweep efficiency and hence
increase storage capacity. However, this could cause excessive pressure according to the incompressibility of brine
which needs further study.
Heterogeneity should also be considered in the next step as K
v
/K
h
ratio tends to affect CO
2
migration path which
results in pressure response.
Well placement should also be studied as it may affect pressure distribution throughout the aquifer and vary the cap
rock size.
Nomenclature
K
v
Vertical permeability, mD
K
h
Horizontal permeability, mD

References
Baklid, A., Korbol, R., & Owren, G. (1996, October 6-9). Sleipner Vest CO
2
Disposal,CO
2
Injection into a Shallow Underground
Aquifer. SPE 36600 SPE Annual Technical Conference and Exhibition .
Bennion, D. B., & Bachu, S. (2006). Supercritical CO
2
and H
2
S-Brine Drainage and Imbibition Relative Permeability Relationships for
Intergranular Snadstone and Carbonate Formations. SPE 99326 .
Bentham, M. (2006). An Assessment of Carbon Sequestration Potential in the UK-Southern North Sea Case Study. Tyndall Centre for
Climate Change Research.
Bentham, M., & Kirby, G. (2005). CO
2
Storage in Saline Aquifers. Oil & Gas Science and Technology - Rev. IFP , 60 (2005) (No.3),
559-567.
Fenghour, A., Wakeham, W. A., & Vesovic, V. (1999). The Viscosity of Carbon Dioxide. J. Phys. Chem. Ref. Data , 27.
Flett, M., Gurton, R., & Taggart, I. (2004). The Function of Gas-Water Relative Permeability Hysteresis in the Sequestration of Carbon
Dioxide in Saline Formations. SPE 88485 .
Gale, J. (2004). Geological Storage of CO
2
: What do we know, where are the gaps and what more needs to be done? Energy 29 (2004)
1329-1338 .
Ghanbari, S., Al-Zaabi, Y., Pickup, G. E., Mackay, E., Gozalpour, F., & Todd, A. C. (2006). Simulation of CO
2
Storage in Saline
Aquifers. Chemical Engineering Research and Design (84(A9)), 764-775.
Holloway, S., Vincent, C. J., & Kirk, K. L. (2006). Industrial carbon dioxide emissions and carbon dioxide storage potential in the UK.
Nottingham: British Geological Survey.
IPCC. (2005). Special Report on Carbon dioxide Capture and Storage. (B. Metz, Ed.)
Juanes, R., Spiteri, E. J., & Blunt, M. J. (2006). Impact of Relative Permeability Hysteresis on Geological CO
2
Storage. Water Resour.
Res. , 42.
Kartikasurja, D. O., Lin, T. G., Sukahar, M. W., & Viratno, B. (2008). Study of Produced CO
2
Storage into Aquifer in an Offshore Field,
Malaysia. SPE 114553 .
Killough, J. E. (1976). Reservoir Simulation with History-dependent Saturation Functions. Trans. AIME 261 , 37-48.
IPTC 14464
14
Kumar, A., Noh, M., Pope, G. A., Sepehmoori, K., Bryant, S., & Lake, L. W. (2004, April 17-21). Reservoir Simulation of CO
2
Storage
in Deep Saline Aquifers. SPE 89343 SPE/DOE Fourteenth Symposium on Improved Oil Recovery .
Mo, S., & Akervoll, I. (2005, March 7-9). Modeling Long-term CO
2
Storage in Aquifer with a Black-oil Reservoir Simulator. SPE 93951
SPE/EPA/DOE Exploration and Production Environmental Conference .
Nghiem, L., Shrivastava, V., Kohse, B., Hassam, M., & Yang, C. (2009). Simulation of Trapping Processes for CO
2
Storage in Saline
Aquifers. Canadian International Petroleum Conference .
Nicot, J.-P. (2008). Evaluation of large-scale CO
2
Storage on Fresh-water Sections of Aquifers: An Example from the Texas Gulf Coast
Basin. International Journal of Greenhouse Gas Control 2 , 582-593.
Oruganti, Y., & Bryant, S. L. (2009). Pressure Build-up During CO
2
Storage in Partially Confined Aquifers. Energy Procedia 1 , 3315-
3322.
Primera, A., Sifuentes, W., & Rodriguez, N. (2009, June 8-11). CO
2
Injection and Storage: A New Approach using Integrated Asset
Modeling. SPE 121970 SPE EUROPE/EAGE Annual Conference and Exhibition .
Pruess, K., Xu, T., Apps, J., & Garcia, J. (2003, February 26-28). Numerical Modeling of Aquifer Disposal of CO
2
. SPE 83695
SPE/EPA/DOE Exploration & Production Environmental Conference .
Qi, R., LaForce, T., & Blunt, M. J. (2008, October 31). Design of Carbon dioxide Storage in Aquifers. International Journal of
Greenhouse Gas Control 3 (2009) 195-205.
Sifuentes, W., Blunt, M. J., & Giddins, M. A. (2009). Modeling CO
2
Storage in Aquifers: Assessing the Key Contributors to Uncertainty.
SPE 123582 2009 SPE Offshore Europe Oil & Gas Conference & Exhibition, 8-11 September 2009, Aberdeen, UK .
Spycher, N., & Pruess, K. (2005). CO
2
-H
2
O mixtures in the Geological Seqeustration of CO
2
II Partitioning in Chloride Brines at 12-100
C and up to 600 bar. Geochimica et Cosmochimica Acta , 69 (13), 3309-3320.
Suekane, T., Nobuso, T., Hirai, S., & Kiyota, M. (2008). Geological Storage of Carbon dioxide by Residual gas and Solubility trapping.
International Journal of Greenhouse Gas Control 2 (2008) , 58-64.
van der Meer, L. (1992). Investigations regarding the Storage of Carbon dioxide in Aquifers in the Netherlands. Energy Conversion &
Management , 33 No. 5-8.
van der Meer, L. (1992). Investigations Regarding the Storage of Carbon dioxide in the Netherlands. Energy Conversion Management ,
33 (No. 5-8), 611-618.
van der Meer, L., & van Wees, J. (2006, September 24-27). Limitations to Storage Pressure in Finite Saline Aquifers and the Effect of
CO
2
Solubility on Storage Pressure. SPE 103342 SPE Annual Technical Conference and Exhibition .
Vandeweijer, V., Van der Meer, B., Kramers, L., Neele, F., Maurand, N., Gallo, Y. L., et al. (2009). CO
2
Storage in Saline Aquifer: In the
Southern North Sea and Northern Germany. Energy Procedia 1 (2009) , 3079-3086.
Vesovic, V., Wakeham, W. A., Olchowy, G. A., Sengers, J. V., Watson, J., & Millat, J. (1990). The Transport Properties of Carbon
Dioxide. J. Phys. Chem. Ref. Data , 19.
Yang, Q. (2008, October 20-22). Dynamic Modelling of CO
2
Injection in a Closed Saline Aquifer in the Browse Basin, Western
Australia. SPE 115236 SPE Asia Pacific Oil & Gas Conference and Exhibition .
Zaytsev, I. D., & Aseyev, G. G. (1993). Properties of Aqueous Solutions of Electrolytes. CRC Press .
Zhou, Q., Birkholzer, J. T., Tsang, C.-F., & Rutqvist, J. (2008). A Method for Quick Assessment of CO
2
Storage Capacity in Closed and
Semi-closed Saline Formations. International Journal of Greenhouse Gas Control 2 , 626-639.
IPTC 14464