Marine and Petroleum Geology 28 (2011) 1083e1093

Contents lists available at ScienceDirect

Marine and Petroleum Geology

journal homepage: [Link]/locate/marpetgeo

Calculation of CO2 column heights in depleted gas fields from known

pre-production gas column heights
M. Naylor*, M. Wilkinson, R.S. Haszeldine
School of GeoSciences, University of Edinburgh, EH9 3JW, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history: Depleted gas fields have been identified as potential targets for CO2 storage. In order to maximise storage
Received 20 March 2010 capacity, a target field must be deep enough to ensure that the CO2 is in a dense phase (either liquid or
Received in revised form supercritical). Accurate assessment of the storage capacity also requires an estimation of the amount of
30 September 2010
CO2 that can be safely stored beneath the reservoir seal which can be estimated in several ways. In this
Accepted 21 October 2010
paper we develop a methodology to convert known pre-production gas column heights into CO2 column
Available online 4 November 2010
heights in order to estimate storage capacities. Several authors have correctly identified that the
differences in interfacial tension and wettability act to reduce the threshold capillary entry pressure for
Keywords:
Carbon dioxide storage
CO2 compared to natural gas, consequently reducing column height estimates. However, under reservoir
Capacity conditions the density of CO2 is substantially higher than natural gas so the buoyancy force on the seal
Gas fields for a fixed column height is much lower for dense phase CO2 increasing column height estimates; we
Column height investigate the effects of this trade off on storage estimates and apply it to a typical dataset, in this case
the UK North Sea.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction buoyancy drive on the seal than the same column height of natural
gas. Thus, the phase behaviour of CO2 counteracts the reduction in
Depleted gas fields are good contenders for engineered CO2 capillary entry pressure. This balance between entry pressures and
storage, having both proven seals and well characterised geology. buoyancy, and the conversion of gas column heights to CO2 column
Assessment of the suitability of a field for CO2 storage requires the heights is the topic of this paper.
estimation of storage capacity. The most simplistic approach Our approach is to take the known, pre-production column
assumes volume for volume replacement, with a discount factor for height of natural gas within a reservoir and use the known differ-
water invasion into the storage zone (Bachu and Shaw, 2003). ences in material properties between the original gas and the
The physical properties of CO2 differ from those of hydrocarbon stored CO2 to convert the pre-production natural gas column height
gas, so that there is the possibility that capillary seals which are to a safe CO2 column height. The advantage of this technique is that
secure for hydrocarbons will not be secure for CO2. A number of it does not require knowledge of intrinsic properties of the seal,
authors have highlighted that the differences in interfacial tension such as the effective (maximum) pore throat radius, which are
(IFT) and wettability between CO2 and natural gas act to reduce the generally unknown and have very large uncertainties. It requires
threshold capillary entry pressure for CO2 compared to gas, that the drive on production was primarily hydraulic so that we can
reducing storage estimates (e.g. Chiquet et al., 2007a; Chiquet et al., assume the force on the seal due to the presence of CO2 can be
2007b). However, for cases with good hydraulic connectivity to the expressed as buoyancy derived from displacing formation water.
aquifer, the full assessment of the storage capacity beneath Our terminology is hence as follows:
a capillary seal also requires knowledge of the buoyancy drive
exerted on the seal by the stored fluid. Since the density of dense
Pre-production column height e the column height corre-
phase CO2 is substantially greater than that of natural gas under sponding to the reserves of gas in the trap prior to extraction
reservoir conditions, a column height of CO2 exerts a lower which may be capillary limited or may be limited by a spill
point or by gas supply in the case of an under filled field.

Maximal CO2 column height eThis is the upper bound on the
* Corresponding author. CO2 column height estimate and is calculated from the pre-
E-mail address: [Link]@[Link] (M. Naylor). production gas column height taking into account buoyancy

0264-8172/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/[Link].2010.10.005
1084 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093

 
and interfacial tension differences between natural gas and cosqseal cosqres gcosq
Pc;th ¼ g  z (2)
CO2, both properties that are relatively well characterised. rt rp R
Thus, this estimate is purely a function of the material prop-
erties of CO2 and natural gas. Where the natural gas field was where rp and rt are the pore and throat radii respectively. The
capillary limited, the calculated maximal CO2 column height contact angle arises as a surface chemistry effect and is therefore
corresponds to the point where the buoyancy pressure of the a function of the rock as well as the two phases; it is close to zero
CO2 equals (actually is infinitesimally less than) the entry when the substrate is strongly water wet.
pressure of the seal such that the seal would leak by capillary It follows that the threshold pressure, or displacement pressure,,
failure. in the non-wetting phase is equal to Pw þ Pc;th , where Pw is the

Safe CO2 column height - the maximal column height pressure in the water (or brine) phase, often equal to the hydro-
neglecting the mineralogical dependent effects of contact static pressure.
angles on the column height estimate. Depending upon the Since the limiting pore and throat radii are impossible to
mineralogy of the seal, the inclusion of contact angles will measure directly, a common way to proceed is to estimate the
reduce the column height estimate. This term has experi- pore throat radii using a porosity-permeability relationship (e.g.
mental, upscaling and theoretical uncertainties. Berg, 1975; Bloch, 1991; Rezaee et al., 2006). For example, an

Column height conversion factor (J) e the ratio of the pre- effective grain size, De can be calculated from core analysis using
production natural gas column height to the CO2 column an empirical permeability equation such as De ¼ ð1:89 k n5:1 Þ0:5
height. In this paper a methodology is developed to calculate (cm), where k is the permeability (mD) and n is the porosity
this value from typical hydrocarbon field data. Conversion (percent) (Berg, 1975). The pore and throat sizes can then be
factors of less than 1 indicate a reduction of the CO2 column estimated using De and assuming a theoretical packing geo-
height compared to the gas column height. This term can also metry. In a rhombohedral packing with perfect spheres
be prefixed with safe or maximal to describe whether contact rp ¼ 0:5  0:414De and rt ¼ 0:5  0:154De . Whilst this may prac-
angles are taken or are not taken into account respectively. tically be the only way forward for an exploration scenario, it is
susceptible to large uncertainties in upscaling from sparse core
The structure of this paper is to first develop the theory behind samples to the whole reservoir and the predictive capability of any
the calculation of CO2 column heights from gas column heights such empirical relationship is strongly limited by the calibration
identifying a conversion factor, next we investigate the controls on data.
that conversion factor, then a discussion informed by typical data The buoyancy pressure exerted on the seal due to a column of
and finished off with a discussion of some secondary, related height, h of less dense fluid under hydrostatic conditions is simply
effects. given by,

Pb ¼ Drgh (3)
2. Column height conversion factor - capillary model
Where Dr is the absolute density difference and g the acceleration
due to gravity.
Across a curved interface between two immiscible fluids
Equating the pressures in Eqs. (2) and (3), the capillary limited
a discontinuity in fluid pressure occurs as a consequence of the
column heights for the gas/water can be expressed as,
interfacial tensions at the interface. The capillary pressure, Pc is the
difference in pressure between a non-wetting phase, pnw and
ggas=water cosqgas=water
a wetting phase, pw , ¼ Drgas=water g hgas=water (4)
R
Pc ¼ pnw  pw (1) We can construct the analogous equation for the CO2/water
The seal is usually a very tight porous material, typically imbibed systems as for the gas/water system in (4),
with water, beneath which the hydrocarbon or CO2 sits in a more
porous and permeable reservoir rock. The threshold pressure, gCO2 =water cosqCO2 =water
¼ DrCO2 =water g hCO2 =water (5)
marks the point above which the non-wetting phase (here, the R
hydrocarbon gas or CO2) displaces the (typically) water phase that The geological factor in Eqs. (4) and (5) is the pore geometry,
imbibes the seal; in terms of basic physics it is the value at which expressed in the term R. This is an unknown factor in almost all
the pressure gradient in both phases becomes continuous. The storage settings, but can now be simply removed from further
pressure where a phase first starts to enter a caprock is referred to consideration. Utilising the fact that both fluids interact with the
as the entry pressure (limited by the largest pore throat at the same pore geometry, we can divide (4) and (5) to remove this term.
reservoir caprock interface) and the pressure where a phase can This gives us an expression for the CO2 column height as a function
travel through the entire thickness of the caprock is the break- of the measured gas column height and physical/chemical prop-
through pressure and is generally slightly larger than the entry erties of the fluid-rock systems.
pressure (limited by the largest pore throat in the least resistance
pathway through the caprock). We assume the capillary threshold Drgas=water cosqCO2 =water gCO2 =water
pressure to be the limiting factor controlling the amount of stored hCO2 =water ¼ hgas=water (6)
DrCO2 =water cosqgas=water ggas=water
natural gas or CO2 in the systems considered here. Other first order
controls of a column height may include being filled to a structural This gives the conversion ratio for gas/water column heights to
spill point, formation reworking or that the trap is under filled CO2/water column heights as
because of a finite supply of hydrocarbon charge. The threshold
capillary pressure, Pc;th is controlled by the surface tension between Drgas=water cosqCO2 =water gCO2 =water
the two fluids (g), the contact angle in the wetting phase (i.e. water)
Jgas=CO2 ¼ (7)
DrCO2 =water cosqgas=water ggas=water
with the caprock (qseal ) and reservoir (qres ) as well as the geomet-
rical properties of the pore network in the seal and reservoir Thus, the conversion factor is a combination of the ratios of the
(described by rp and rt or combined in an effective pore radius, R), density differences, contract angles and interfacial tensions for the
 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093 1085

gas/water e CO2/water systems. We discuss each of these key ratios However, Chiquet et al (2007b) demonstrated that previous anal-
in turn. ysis of the CO2/water IFT values was biased towards lower values as
previous workers had assumed an incorrect density for the CO2
saturated water phase. Figure 1 shows a compilation of data for the
2.1. Interfacial tension (IFT) ratio (CO2þCH4)/water system. We choose to linearly interpolate
between the pure methane/water and CO2/water IFT values,
At the reservoir/seal boundary, the IFT between the wetting and consistent with this observation, since there is not sufficient data to
non-wetting fluids linearly increases the capillary entry pressure do better.
that must be overcome in order for the wetting fluid to invade the Thus we suggest a correction to the ratio that accounts for
seal. The IFT for hydrocarbon/water systems has been used exten- known amounts of CO2 in the original gas. We define the mol% of
sively for many years in the industry and has been extensively CO2 in the natural gas as xCO2 and CH4 as xCH4 . Using this we esti-
characterised. We use methane as an example here because it is at mate the IFT of the natural gas/water system,
the upper limit of hydrocarbon IFTs and is thus provides a worst
case scenario for the effects of this term. The IFT for CO2/water ggas=water zxCO2 gCO2 =water þxCH4 gCH4 =water zxCO2 gCO2 =water
systems has not been investigated as extensively and consensus

þ 1xCO2 gCH4 =water (8)
regarding its behaviour is only just starting to emerge (Chiquet
et al., 2007b; Ren et al., 2000). For both methane-water and CO2- Therefore, the ratio of the IFTs is given by
water, the IFTs increase rapidly at low pressures but become
increasingly stable with increasing pressure (Fig. 1). As an gCO2 =water gCO2 =water
¼

approximation, we can take the ratio at 10 MPa as representative of ggas=water xCO2 gCO2 =water þ 1  xCO2 gCH4 =water
typical reservoir depths where gCO2 =water z25 mN/m (Chiquet et al.,
1 1
2007b) and gCH4 =water z50 mN/m (Danesh, 2007; Ren et al., 2000). ¼
gCH z (9)
xCO2 þ 1  xCO2 g 4 =water 2  xCO2
This gives an IFT ratio w 0.5. CO2 =water

The effect of the IFT ratio term is to lower the capillary entry
pressure for pure CO2/water systems by up to 50% compared to the The ranges of this ratio are from 0.5 if the natural gas was pure
pure gas/water system (Li et al., 2006). Thus the effect of this term is methane (i.e. xCO2 ¼ 0) to 1 in the (unrealistic for our purposes)
to decrease the maximum column height of CO2 relative to the end-member that the natural gas was initially pure CO2 (i.e.
methane gas column height. xCO2 ¼ 1).
The methaneewater system that existed prior to CO2 injection
may have contained significant CO2, and this must be accounted for
in calculating the change in IFT due to CO2 injection. It has been 2.2. Density difference ratios
experimentally demonstrated that in mixtures of CO2 and methane,
higher proportions of CO2 reduce the IFT (Ren et al., 2000), The density difference ratio describes the relative difference in
consistent with the CO2 IFT being lower than that of methane. the buoyancy drive between the CO2/water and natural gas/water
systems. Since buoyancy in the subsurface is generally derived from
displacing formation water, this approach is appropriate for
systems with good hydraulic connectivity to the aquifer. This
buoyancy force provides the drive that must overcome the capillary
entry pressure if the non-wetting fluid is to invade the caprock. The
greater the buoyancy contrast, the stronger the drive.
We demonstrate typical subsurface buoyancy depth profiles
using two end-member subsurface scenarios corresponding to
different heat flow regimes and sub-aerial and submarine condi-
tions (Fig. 2).
We calculate pressure P, temperature T and salinity ws as
a function of depth from defined surface pressure, temperature and
salinity values and corresponding geothermal, hydrostatic gradi-
ents and salinity gradients. Salinity tends to increase with depth
from sea water concentrations offshore and freshwater concen-
trations onshore (Moss et al., 2003); the salinity gradient is highest
when salt layers are present in the subsurface (e.g. Southern North
Sea). These profiles are then used to calculate how the phase
properties of CO2, natural gas and water vary with depth. We define
the depth, z to be the distance beneath the surface, which corre-
sponds to either the topographic elevation or the seafloor. In the
submarine case, the surface pressure also includes the pressure
exerted by the water above the seabed.
Fig. 1. Compilation of IFT data for (CO2þCH4)/water mixtures (Ren et al., 2000) and for
These calculations are typically conditioned to initial reservoir
the pure CO2/water (Chiquet et al., 2007b). For the data between 0 to 80 mol% CO2, at
each pressure measurements are taken at 4 temperatures (298.15 K, 313.15 K, 333.15 K, conditions and neglect coupled processes associated with the
353.15 K and 373.15 K). For the data at 100 mol% CO2, measurements were taken at 5 injection of CO2 which requires a full reservoir simulation. They are
different temperatures. Chiquet suggested that previous measurements of IFT were useful for the rapid initial assessment of a wide range of storage
biased to lower values because they assumed a density of the CO2 saturated water that scenarios and should be used in conjunction with geological and
proved inaccurate; we interpret it as being likely that the IFT values for the CO2þCH4
mixtures lie closer to the lines fitted between the pure CO2 and pure CH4 IFT values.
hydrodynamic information (such as the water drive) to identify
This interpretation is supported by the tight clustering of the Chiquet measurements which coupled processes will be important for each specific target.
compared to Ren. Thus this methodology provides a technique for the preliminary
1086 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093

Fig. 2. Phase properties and buoyancy drives for two different subsurface scenarios. The left hand figures refer to a sub-aerial case and the right a submarine case. (a,b) Vapour
pressure curves (solid black) with subsurface pressureetemperature trajectory (solid red). Contours of constant density as a function of pressure and temperature have been added
in dashed lines; the corresponding density (kg/m3) has been written in the same color. Dotted blue lines are the critical pressure and temperature. (c,d) Subsurface densities (solid
red line) plotted on a pressure-density phase diagram. The black solid line shows the saturated vapour curve and the dew point curve. Some isotherms have been added near the
critical point for reference as dashed lines. Dotted blue lines show the critical density and pressure. The dashed red line in (d) is the phase change pressure. (e,f) Natural gas, CO2,
fresh water and saline density profiles. The blue dotted line is the depth corresponding to the critical pressure and the red dashed line in (f) is the depth at which the phase change
occurs. (g,h) Density differences between the (1) natural gas and fresh water (black solid) or saline (black dashed) (2) CO2 and fresh water (red solid) or saline (red dashed). (i,j)
Density difference ratios between (1) gas/fresh water and CO2/fresh water systems (solid lines) and (2) gas/saline and CO2/saline systems (dashed lines).(For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article).

site assessment prior to a more complete analysis using reservoir rw ðP; T; ws Þ, the produced gas (assuming a specific gravity of 0.65
simulations. and a compressibility factor of 0.85), rgas ðP; T; Mgas Þ and the injec-
We use standard calculations (See appendices) to determine the ted CO2, rCO2 ðP; TÞ. The form of the density profiles all trade off
densities of the fresh (where ws ¼ 0) and saline formation water, between increases in pressure that tend to increase densities and
 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093 1087

increases in temperature which tend to decrease densities. We change regime are; higher geothermal gradients, higher surface
assume that the density will be fairly constant in the column which temperatures and lower surface pressures. The density of CO2
is reasonable when CO2 persists in the dense phase or for relatively Understanding the variation in CO2 density with depth is key to
short column heights; for a full treatment on real data this determining the conversion factor and the degree to which the
assumption can be relaxed by integrating over an actual column. decreased buoyancy compensates for the lower IFT in the CO2/
We neglect the effects of dissolution of CO2 in the formation water water system. Factors which increase the likelihood of being in
as the small increase in saturate brine density is second order to the a phase change regime are; lower geothermal gradients, lower
main effects described here. surface temperatures and higher surface pressures, i.e. moving the
The CO2 vapour pressure curves are plotted with pressuree well head pressure and temperature towards the top left in Figure 2
temperature pairs through the subsurface in Figure 2(a,b). In Case 2, (a,b). Factors which decrease the likelihood of being in a phase
where the pressureetemperature path crosses the vapour pressure change regime are; higher geothermal gradients, higher surface
curve, a phase transition from the gas to liquid phases is predicted temperatures and lower surface pressures. The density of CO2 at
at depths less than w300 m below the seabed (shallower than the depths corresponding to pressures and temperatures near either
depth corresponding to the critical pressure for CO2); this is higher the vapour pressure curve or the critical point will be subject to
than would be expected for a sub-aerial case because of the higher very high uncertainties and propagate through to high uncer-
surface pressure which is site specific because of the dependency tainties in the conversion factor.
upon the water depth. For a range of realistic parameterisations, the
density in the supercritical phase actually increases with shallower 2.3. Contact angle ratio
depths, acting to decreasing the buoyancy drive. In contrast, in Case
1 which corresponds to higher heat flow and lower surface pres- The contact angle is the angle at which a liquid/vapour interface
sure (sub-aerial) the pressureetemperature path does not intersect (here either hydrocarbon/water or CO2/water) meets the solid
the vapour pressure curve and the predicted CO2 densities corre- surface (the surrounding rock). Contact angles arise from surface
spond to a smooth trajectory from the supercritical density to gas chemistry effects. This effect cannot be easily eliminated because
densities over a much wider depth range always increasing the the interaction of natural gas with a specific rock mineralogy does
buoyancy drive. not necessarily act as a proxy for how CO2 will behave; scaling from
In contrast to the CO2 density profiles, the natural gas and water mineral scale experiments to core scale to reservoir scale is difficult.
density profiles (Fig. 2i,j) are better behaved as they are in more We need to know the contact angles for both the CO2/water and
stable regions of their respective phase spaces. The formation water gas/water systems for a given mineralogy in order to make the best
density is considerably less variable with typical values in the range estimate. CO2 wettability is a complex story and a clear picture has
960e1030 kg/m3 for fresh water and upto 1200 kg/m3 for saline not yet emerged as to the importance of wettability effects (Chiquet
water with 25% weight percent of salt dissolved. et al., 2007a; Hildenbrand et al., 2004; Li et al., 2005; Yang et al.,
Using these density profiles we calculate the density differences 2008).
(Fig. 2 g,h) and density difference ratios (Fig. 2i,j) which is the key The quantity of interest is the ratio of the cosines of both contact
controlling ratio that characterises the relative buoyancy drive angles, which is hard to constrain precisely, but will be
cosqCO2 =water
Drgas=water rwater  rgas cosqgas=water
w1. Given an initial water wet gas/water system, it will be
¼
DrCO2 =water rwater  rCO2 less than 1.
Since there is a range of uncertainty in the contact angle of the
Whilst we chose end-members here, it should be noted that CO2/water system, we prefer to report a range of conservative
relatively subtle changes in the thermal and pressure regimes (termed “safe”) and optimistic (termed “maximal”) values. The
propagate through to a significant differences in the density uncertainty in this term is significantly less than the uncertainty in
difference profiles (Fig. 2.g,h), and corresponding predictions of the the pore radius term R which contained order of magnitude
buoyancy drive. In the subsurface, the density of CO2 is always uncertainty.
greater than that of typical natural gas compositions. Since In the presence of the hydrocarbon phase, many minerals will
DrCO2 =water is lower than Drgas=water ; the density difference term in generally be water wet giving a contact angle of 0o giving
the conversion factor increases the potential CO2 column height. a contribution to the denominator of the ratio of
The density difference ratio is always greater than w2 when CO2 is cosðqgas=water Þ ¼ cosð0Þ ¼ 1. The use of this value provides the
in the dense phase and greatest when the pressureetemperature maximal CO2 column height conversion factor. The presence of
profile produces a phase change. The density of liquid phase CO2 is dense phase CO2 has been shown to alter the water-wettability of
nearly always higher than that in the supercritical phase, so we various organic materials present in the subsurface, such as coal,
expect the safe CO2 column height to be greater for subsurface resulting in an increase in the contact angle. Many seals are more-
conditions that encourage a phase change in the depth profile, i.e. or-less organic-rich mudrocks (e.g. 5-10% organic content for the
cooler (lower surface temp and geothermal gradients) and Kimmeridge Clay (Brown, 1984)) so that the behaviour of this
submarine (higher surface pressure). component of the rock is of potential importance.
Since the typical minimum for the IFT ratio is 0.5, the buoyancy In contrast, laboratory studies of contact angles (Chiquet et al.,
term should always at least counteract the decrease in column height 2007a) backed up by breakthrough studies (Hildenbrand et al.,
due to the IFT effect for regions where CO2 is in the dense phase. 2004; Li et al., 2005) suggest that the presence of mica and
Understanding the variation in CO2 density with depth is key to quartz, minerals commonly present in shaley caprocks, reduce the
determining the conversion factor and the degree to which the contact angle (cosqz0:66; cosq z0:85 respectively). Espinoza and
decreased buoyancy compensates for the lower IFT in the CO2/ Santamarina (2010) look at contact angles in a range of systems;
water system. Factors which increase the likelihood of being in they show that for the contact angles are relatively pressure inde-
a phase change regime are; lower geothermal gradients, lower pendent on quartz (and cosqbrine=CO2 z0:77) and calcite
surface temperatures and higher surface pressures, i.e. moving the (cosqwater=CO2 zcosqbrine=CO2 z0:77). Since these angles were
well head pressure and temperature towards the top left in Figure 2 measured on pure samples, and real rocks are composed of a range
(a,b). Factors which decrease the likelihood of being in a phase of minerals, they too provide a conservative estimate for estimation
1088 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093

of the safe CO2 column height conversion factor. The scaling of

Safe column
contact angles from a single mineral to multi-mineral reservoir is

correction
a non-trivial problem.

Height

factor
0.755
0.817
0.793
0.731
0.818
0.922
0.696
0.712
1.326
1.260
0.796
A CO2 storage study for the Rousse depleted gas field in France
(Tonnet et al., 2010) has analysed slices of the calcite and quartz rich
caprock substrate, which is low in mica, and show it to remain

Maximal column
water wet in the presence of CO2.
A close look at the work by Chiquet et al. (2007a) shows that the

correction
water-wettability of quartz is barely altered in the presence of

Height

factor
1.079
1.167
1.133
1.045
1.168
1.318
0.995
1.016
1.894
1.800
1.137
dense phase CO2, as opposed to light phase CO2 (see Fig. 8 in
Chiquet et al., 2007a), whereas the contact angle values obtained
with mica indicate that mica is becoming less water wet when the

0.502
0.501
0.500
0.501
0.501
0.501
0.502
0.502
0.503
0.502
0.502
Ratio
CO2 gets denser (i.e., pressure increases), even though there seems

IFT
to be a strong dispersion in the contact angle values (see Fig. 7 in
Chiquet et al., 2007a). With mica, less definitive conclusions were

Density
drawn in a subsequent study by the same laboratory (Shah et al.,

ratio

2.15
2.33
2.26
2.08
2.33
2.63
1.98
2.03
3.76
3.59
2.27
2008a).
The most conservative approach is to assume that the seal was
completely water wet in the presence of the hydrocarbon and to

Gas/water

difference
(kg/m^3)
983.42
963.87
977.77
1077.44
965.36
1010.11
1069.86
1042.05
1020.51
1011.49
984.77
density
use the most extreme reduction factor of 0.7. The most optimistic
approach is to assume that the caprock will remain completely
water wet, corresponding to a contact angle ratio of 1.0. Here we
report the results of calculations for both scenarios.

CO2/water

difference
(kg/m^3)
density

457.25
413.70
431.71
516.80
414.18
384.10
540.26
514.13
271.10
282.04
434.61
3. Southern North Sea example

We have applied this analysis to gas fields from the UK Southern

(kg/m^3)
North Sea, as a uniform publicly available dataset is available that

166.361
159.126
135.539
172.857
158.027
159.293
169.600
181.384
137.896
143.356
167.441
density
includes gas expansion factors that are not otherwise commonly
published, and chemical analyses of the gas (Abbotts, 1991). The Gas
fields all contain dry gas, with methane contents of 91e95 mol %,
Molecular
Weight of
and CO2 contents of 0.4e1.2 mol % (Table 1). The seal in this

16.962
16.650
16.915
16.975
16.535
16.667
16.932
17.010
17.209
17.005
16.787
instance is the Zechstein halite, so that capillary leakage is impos-
Gas

sible, and the column height ratios would not need to be calculated
in as real-life assessment of storage capacity. However, it does
compressibility

provide a realistic reference dataset for global submarine reservoir

conditions that are also relevant to a number of cases where
capillary leakage is the limiting factor; results on commercial
0.960
0.947
0.870
0.957
0.947
0.807
1.014
0.966
2.428
2.273
0.989
datasets where capillary leakage is likely to be the limiting control
Gas

are comparable. A histogram of the maximal and safe column

height ratios calculated for these data is shown in Figure 3 for all
expansion

the fields with sufficient information (Table 1).

Calculated column height ratios for gas fields from the UK Southern North Sea.

factor

The maximal column height ratios, neglecting contact angle

235.0
229.0
192.0
244.0
229.0
229.0
240.0
255.5
192.0
202.0
239.0
Gas

changes, are all greater than 0.995 and less than 2. When experi-
mental contact angles are included, the safe conversion factors are
Corrected

in the range of 0.7e0.92; the exceptions being the North and South
(kg/m^3)
1149.79
1123.00
1113.31
1250.30
1123.38
1169.40
1239.46
1223.43
1158.40
1154.84
1152.21
density

Sean fields. In the latter, the predicted CO2 densities are higher than
water

other fields because the CO2 is in the liquid phase with a density
that is substantially higher than the supercritical CO2 in the higher
pressures and temperatures examples.
(kg/m^3)
Density

692.54
709.30
681.60
733.50
709.20
785.30
699.20
709.30
887.30
872.80
717.60

The conversion factor is more sensitive to changes in the CO2

CO2

density than the gas density because the density difference

between the CO2 and water is smaller. The fields with the poorest
conversion factors correspond to hot, low pressure reservoirs
360.93
352.59
338.71
352.59
352.59
324.82
366.48
364.82
362.04
367.59
358.15
Temp

where the CO2 is less dense, such as Ravenspurn South.

(K)

The above analysis indicates that the results of this assessment

are crucially dependant upon the wetting behaviour of minerals in
Pressure

a CO2-rich setting. It is suggested that more research is required in

(MPa)

28.25
26.53
19.64
28.58
26.53
20.82
30.94
31.23
58.57
58.57
29.39

this field.
Where the conversion factor is confidently assessed as greater
than 1, a further assessment of the structural geometry of the traps
Ravenspurn S

could be performed to assess whether any extra storage volume

West Sole
Amythest

Camelot
Cleeton

would be gained by an increased column height, i.e. filling any

Clipper
Barque

Sean N
Leman

Sean S
Rough
name
Table 1

Field

under filled structures to spill point, subject to pressure increase

limitations that are outwith the scope of this paper. If the safe
 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093 1089

Fig. 3. Histogram of (a) maximal and (b) safe (including experimental contact angle data) column height ratios for the Southern North Sea dataset described in the text.

column height conversion factors are taken to be the more realistic NaCl they found that the induced IFT increase was outwith
of the values calculated here, then this leads to a reduction in the experimental error.
CO2 storage volume estimate of the fields. In geometrically complex This effect alone has the potential to increase the IFT in saline
reservoirs, each trap requires individual assessment to calculate the formations at depths where CO2 would be in the dense phase by
storage capacity with the safe CO2 column height. w4-5 mN/m making the ratio,
In order to calculate the safe and maximal column heights, we
multiply the conversion factors by the pre-production gas column
gCO2 =brine
z3=5 ¼ 0:6
height. Where the gas column was capillary limited, the CO2 gCH4 =brine
column heights are also capillary limited. However, the pre-
compared to w0.5 for the pure water case. However, increasing
production column height might not have been capillary limited. In
salinity can also increase the IFT for light hydrocarbon/water
an under filled gas field, the derived CO2 column height will be
systems (e.g. Cai et al., 1996), which would moderate the potential
a lower estimate of the true capillary limited capacity. Similarly,
increase. The effect of brine salinity on the IFT ratio is therefore
where the gas field is actually limited by a structural spill point, the
generally a second order feature.
derived CO2 column height will also be a lower estimate.
A comprehensive description of the effect of brine salinity on
This analysis has not taken into account irreversible invasion of
the hydrocarbon/brine IFT is a complicated problem due to the
water during the production phase which does not affect the esti-
diversity of the composition of hydrocarbon phases. The salinity of
mate of the safe column height since buoyancy is purely derived
the brine increases the IFT between the brine and either a gas or
from the vertical extent of a connected stringer of the buoyant fluid
non-polar oil. However, polar fractions such as resins and asphal-
within the reservoir. However, the irreversible water saturation
tenes may contain groups that tend to decrease the IFT. The
does affect the assessment of capacity, which is beyond the scope of
magnitude of these effects will be on the same order as the effects
this paper. Capacity is reduced by the irreversible invasion of water
CO2 saturation in the brine.
since it decreases the proportion of pore space available for CO2
Yang et al. (2005) measured the dynamic contact angle between
compared to what was available for the initial gas in place.
CO2, brine and crude oil systems. They noted that the measured
dynamic IFT rapidly decays to a constant value explained by the
absorption of CO2 molecules and the reorientation of water mole-
4. Second order factors cules at the interface. CO2 dissolution is also a factor with the crude
oil. The IFTs are slightly lower in the presence of CO2. These effects
4.1. Effect of brine salinity on IFT are pressure and temperature dependent.

Thus far, we have not considered the effect of brine salinity on 4.2. Generalisation to sour-gas systems
the IFT, the data in Figure 1 was for low salinity cases. The effects
of brine salinity on measurements of IFT and wettability for the Thus far, this paper has concentrated on the geological storage of
geological storage of CO2 have been investigated by several CO2 in a depleted field which contained either a pure hydrocarbon
authors (Aggelopoulos et al., 2010; Chalbaud et al., 2009; Espinoza mixture or a hydrocarbon þ CO2 mixture. Now we include a brief
and Santamarina, 2010; Massoudi and King, 1975). They came to discussion on how this system can be generalised to include H2S in
several important conclusions. Firstly, the basic form of the IFT the predominantly CO2 injection stream. Technically, an acid gas is
curve is controlled primarily by the density difference between any gas that contains significant amounts of acid gases such as CO2
the CO2 and brine. Further, increasing salinity linearly shifts the or H2S; a sour as is a natural gas that specifically contains significant
IFT curve to higher values at around a gradient of 1.3 mN/m per amounts of H2S. Many of the fields yet to be developed contain sour
molal salt concentration. For salt concentrations above 30 g/L of gas and the disposal of H2S is problematic. The conventional
1090 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093

process is to separate the sulphur using the Claus process but, as using Eq. (10). This IFT can then be used to find the column height
well as several environmental challenges regarding this process, conversion factor Eq. (7) and estimate the new column height. The
the production of sulphur by this mechanism is already in excess of estimated maximum CO2 column height will be reduced compared
world demand. An alternative way to deal with the H2S is to reinject to injecting pure CO2, again assuming that the capillary entry
it into the subsurface, potentially mixed with CO2, as part of a CCS pressure of the seal was the limiting factor in determining the pre-
project which is the context we are interested in here. Here we production column height.
explore briefly how the terms of Eq. (7) would be altered by the
presence of H2S. 4.2.3. Case (iii): where a sour CO2 stream is injected into a sour gas
The wettability alteration induced by H2S has been investigated reservoir
on several substrates (Shah et al., 2008a); it was found that with Here we need to estimate both the injected sour CO2 IFT and the
quartz and a caprock sample no significant wettability alteration produced sour gas IFT. The magnitude of the ratio of these IFTs will
was observed. With a mica substrate the water wettability was determine whether there will be a net increase or decrease in
moderately altered in the presence of gaseous H2S, and drastically column height compared to non-sour case. The calculation of the
in the presence of dense (liquid) H2S. IFT relies on knowing the mole fraction of each component, which
The density of H2S þ CO2 mixtures has to be considered in place must be assessed on a case-by-case basis.
of the density of pure CO2 in the density difference ratio; data and Here we present one specific case to demonstrate the non-linear
models exist for predicting this density (e.g., Stouffer et al., 2001). trend in IFT ratio with depth that emerges as a consequence of the
In order to estimate the impact of H2S on CO2 storage in non-linear equations of state. We consider the case where we
a depleted gas field, we need to be able to assess the impact of H2S extract H2S from a produced gas and 100% of this H2S is re-injected
on the IFT of gas mixtures. Pure H2S has a lower IFT than either within the injection gas, and assuming 100% efficient volume for
methane or CO2 at approximately 20% of the pure methane value. It volume replacement of total gas within the reservoir; this
has been shown that we can estimate the IFT of mixtures including conserves the molar concentration of H2S. The produced gas is
H2S, by generalising the gas/water IFT estimation (Eq. (8)) (Shah composed purely of CH4 and H2S. First we need to calculate the
et al., 2008b), total number of moles per unit volume of H2S in the produced gas
under reservoir conditions,
ggasMixture=water zxH2 S gH2 S=water


þxCO2 gCO2 =water þxCH4 gCH4 =water
z 0:2 xH2 S þ0:5 xCO2 þxCH4 gCH4 =water produced produced
mCH4 x H2 S
produced
xH2 S rsubsurface
produced CH4
mH2 S ¼ z (11)
(10) produced
1  xH2 S 1  xproduced MCH4
H2 S
As before, the influence of the IFT from different components is
The injection stream is a single phase composed purely of CO2 and
dependent upon the mole fraction of that component in the multi-
H2S and is assumed to have the same number of moles of H2S per
component mixture. Thus the impact of the addition of a finite
unit volume under reservoir conditions as were present in the
number of moles of H2S on the IFT of the mixture is dependent on
produced gas, so that 100% of the produced H2S is re-injected.
the absolute number of moles of the other components through the
Assuming any density change in the injection stream due to the
calculation of mole fractions.
presence of H2S is negligible, the mole fraction of H2S in the
We consider the implications for the range of likely storage
injection stream can be calculated
scenarios involving a sour gas stream. There are a number of
different sour gas injection scenarios relevant to CO2 storage; (i)
mproduced mproduced
injection of pure CO2 into a field which initially contained a sour xinjected ¼
H2 S
z
H2 S
(12)
þ mCO2 mproduced þ rCO2
H2 S produced subsurface
gas, (ii) injection of a CO2 þ H2S mixture into a field which initially mH2 S
H2 S MCO
contained low amounts of H2S, and (iii) injection of a CO2 þ H2S 2

mixture into a field which initially contained a sour gas. These two equations are non-linearly related due to the effect of the
density of CO2 and CH4 at reservoir pressure and temperature. In
4.2.1. Case (i): the initial gas in place was sour Figure 4 we explore this non-linear relation as a function of depth
The IFT of the initial gas in place would be lower than for a pure using the two scenarios presented earlier in Figure 2 to define the
hydrocarbon gas resulting in the pre-production column height subsurface pressure and temperature trends.
being lower than it would have been for a purer hydrocarbon, We calculate the mole fraction of H2S in the injection stream for
assuming that the capillary entry pressure of the seal was the mole fractions in the produced gas of 0.01, 0.02, 0.05 and 0.1 using
limiting factor in determining the pre-production column height. Eqs. (11) and (12) over 2 km depth range. These are plotted in
Essentially, the reduced seal performance due to the presence of Figure 4a where the diagonal dashed lines show lines of equiva-
H2S is recorded in the pre-production column height. For a known lence between gas compositions where IFTsubsurface ¼ IFTsubsurface
gas composition, the IFT of this pre-production mixture can again in black and IFTsubsurface ¼ 2 IFTsubsurface in red; the latter being
be estimated using Eq. (10) and applied in Eq. (6) to convert the equivalent to the non-sour case presented earlier in this paper
sour pre-production column height to a pure CO2 column height. where the IFT of pure CO2 was noted to be half that of pure CH4
Where a pure CO2 stream is injected into a sour gas reservoir, we which is confirmed by this line of equivalence passing through the
can expect larger column heights of CO2 than would have been origin where there is no H2S in either stream.
predicted using the pure methane IFT for the initial gas in place, In the inset in Figure 4a and b we zoom in on the data and add
again assuming that the capillary entry pressure of the seal was contours of constant depth. In both cases the mole fraction in the
the limiting factor in determining the pre-production column injection stream initially decreases with depth before increasing
height. again, reflecting the trade off in density variations between CO2 and
methane with depth (Figure 1). This is shown more clearly in
4.2.2. Case (ii): a sour CO2 stream is injected in a field that did not Figure 4c and d where the mole fraction of H2S in the injection
initially contain sour gas stream is plotted as a function of depth (solid line) compared to the
This is the simplest case to consider. The IFT of a CO2þH2S mole fraction of H2S in the produced gas (dashed line) for the same
injection mixture, with a known composition, can be calculated range of H2S contents in the produced gas.
 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093 1091

Fig. 4. Generalisation to acid gasses. Here we present an evaluation of the impact of H2S on the interfactial tension ratio assuming the same molarity of H2S in a produced methane
stream and in an injected CO2 stream. (a,b) Plot showing the produced and injected molar fractions of H2S. The line of equivalence where the gas compositions have the same IFT is
shown with the black dashed line. The line of equivalence where the injected (CO2þH2S) stream has half the IFT of the produced (CH4þH2S) gas is shown with the red dashed line.
Calculations have been performed to calculate the molar fraction of H2S in the injection stream assuming a constant molar fraction in the produced gas as a function of depth. The
inset figure zooms in on the data and adds contours of depth. Note how the injected molar fraction decreases with depth then increases again. (c,d) Plot of the molar fraction of H2S
in the injection stream (solid line) given some molar fraction in the produced gas (dashed line). The molar fractions vary non-linearly as they are dependent upon the densities of
the CO2 and methane which vary non-linearly with depth (Fig. 2c,d). (e,f) The IFT ratio is calculated as a function of depth using Eq. (10). The IFT of the injection stream is always
slightly greater than half the IFT of the produced gas. These ratios also vary non-linearly with depth because they are functions of the molar fractions.

Having calculated the mole fraction of H2S in both the produced utility than estimation from first principles, as pore throat radii do
and injected streams as a function of depth, we can use Eq. (10) to not need to be estimated, a procedure which would introduce order
find the ratio of the IFTs in the injected and produced gasses, which of magnitude uncertainties. The measurement of this property
is required to determine the column height conversion factor (Eq. requires specific core which may not be available. The conversion
(7)). In this example, mole for mole conservation of H2S in the factor takes into account the differences in buoyancy drive, inter-
produced and injection streams assuming a total volume for facial tensions and contact angles between the two systems.
volume replacement of the gasses produces a weak increase in the Where CO2 is in the dense phase, the reduction in capillary entry
IFT ratio tending to increase column heights. pressure as a consequence of interfacial tension differences at the
Note that the case of large proportions of H2S requires a full seal is at least balanced by the decrease in buoyancy of supercritical
calculation of the density of the mixture, which remains beyond the CO2 compared to natural gas. Thus actual reductions in the column
scope of this paper. We have not considered the modifications to height are then derived from changes in contact angles, about
the phase diagram associated with multi-component gasses, such which there are large theoretical and experimental uncertainties.
as the change in bubble point because such a complete treatment is Neglecting the role of contact angles, an optimistic approach, for
non-unique and substantially beyond the scope of this paper; the fields with CO2 in the dense phase, typical conversion factors are in
concepts outlined here remain pertinent to the full analysis. the range of 1.0e2.0 for the UK Southern North Sea case studies. For
shallow-buried fields with low pressure CO2 in the gas phase, the
5. Conclusions column height conversion factor can be as low as 0.5.
A more conservative estimate includes an assessment of fluid-
We have demonstrated the utility of a simple conversion factor to mineral contact angle ratios, about which there is significant
estimate safe CO2 column heights in depleted gas fields from known, uncertainty but is a number less than but on the order of 1. For
pre-production gas column heights. This technique has greater shaley caprocks this may be as low as 0.7, but this end-member
1092 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093

estimate was based on a single mineral analysis and scaling up to

a complex reservoir caprock is likely to increase this number Bw ¼ 1 þ DVw;P 1 þ DVw;T
towards 1 as suggested by other studies. For example, experiments
This expression contains a pressure and temperature correction
on carbonate rich caprocks indicate that caprock will remain water
factor,
wet. Taking the most extreme value for the single mineral reduces
conversion factors to between 0.7-1.3 for the Southern North Sea DVw;P ¼ Pð3:58922E  7 þ 1:95301E  9 TÞ  P 2 ð2:25341E
case studies. This suggests that more research should be directed
towards understanding wetting behaviour in CO2-rich systems to  10 þ 1:72834E  13TÞ
determine whether wettability is a significant control in real
systems.
DVw;T ¼ 1:001E  2 þ 1:33391E  4 T þ 5:50654E  7 T 2
Acknowledgements
Produced gas density (dry gas), rgas
MN is funded by the Scottish Government and the Royal Society
of Edinburgh. MW was funded by the UKERC. Using the ideal gas formula the density under reservoir condi-
tions (Danesh, 2007) is expressed as,
Variable Glossary
PMgas
rgas ¼
Zgas Rgas T
P Reservoir pressure (MPa) Where,
T Reservoir temperature (K) The gas formation factor, Bgas describes the ratio of the gas
Sg The specific gravity volume at reservoir conditions compared to the gas volume at
Bgas The gas formation factor standard conditions and is frequently a recorded reservoir param-
Zgas Compressibility of produced gas eter. This can be used to the compressibility of the gas,
ws Weight percent of dissolved salt in formation water
my Molar concentration of component y(mol/m3) Bgas P
Zgas ¼
xy Mole fraction of component y in a mixture 3:47  104 T
jgas=CO2 Natural gas to CO2 column height conversion factor
Where the molecular weight is known, this can be used directly. For
rw ,rgas ,rCO2 Density of formation water, produced gas, carbon
pure methane, Mgas ¼ 16. For natural gas mixtures it will be
dioxide (kg/m3)
higher, typically w22 (Selley, 1998). For gasses with heavier
DrX=Y Density difference between substances Y and X (kg/m3)
hydrocarbon molecules, such as pentane, it can be as high as w43
gX=Y Interfacial tension between substances X and Y
(Kapuni field, NZ (Selley, 1998)).
cos qX=Y Contact angle for substances X and Y on some implicit
The specific gravity is often the quoted parameter, which relates
solid matrix
the molecular weight of gas at standard conditions to the molecular
hX=Y Column height of substance X with substance Y above
weight of air at standard conditions, Mgas ¼ Sg Mair ¼ 28:96Sg .
My Molecular weight of component y(g/mol)
rp Pore radius
rt Throat radius References
Rgas ¼ 8:3144E  3 Universal gas constant for P and T in (MPa)
and (K) Abbotts, I. L. (1991), United Kingdom Oil and Gas Fields 25 Years Commemorative
Volume.
Aggelopoulos, C.A., Robin, M., Perfetti, E., Vizika, O., 2010. CO2/CaCl2 solution
Appendices interfacial tensions under CO2 geological storage conditions: influence of cation
valence on interfacial tension. Advances in Water Resources 33 (6), 691e697.
Bachu, S., Shaw, J., 2003. Evaluation of the CO2 sequestration capacity in Alberta’s oil
Formation water density, rw and gas reservoirs at depletion and the effect of underlying aquifers. Journal of
Canadian Petroleum Technology 42 (9), 51e61.
Berg, R.R., 1975. Capillary pressures in Stratigraphic traps. Aapg Bulletin-American
All of the calculations in this section are using T(F), P(psia) and
Association of Petroleum Geologists 59 (6), 939e956.
density(lbm/ft3) as described by Danesh (2007). The resulting Bloch, S., 1991. Empirical prediction of porosity and permeability in sandstones.
density, rw needs to be converted back into SI units to integrate AAPG Bulletin 75, 1145e1160.
Brown, S., 1984. Jurassic. In: Glennie, K.W. (Ed.), Introduction to Petroleum Geology
with the rest of the analysis.
of the North Sea. Wiley-Blackwell.
The formation water density at standard conditions expressed in Cai, B.Y., Yang, J.T., Guoet, T.M., 1996. Interfacial tension of hydrocarbonþ water/
the weight percent of dissolved salt, ws can be estimated in the brine systems under high pressure. Journal of Chemical and Engineering Data
form, 41 (3), 493e496.
Chalbaud, C., Robin, M., Lombard, J.-M., Martin, F., Egermann, P., Bertin, H., 2009.
Interfacial tension measurements and wettability evaluation for geological CO2
rwðstandardÞ ¼ 62:368 þ 0:438603ws þ 1:60074  103 ws storage. Advances in Water Resources 32 (1), 98e109. 10.1016/
[Link].2008.10.012.
Neglecting the mass of dissolved gas in the water at reservoir Chiquet, P., Broseta, D., Thibeau, S., 2007a. Wettability alteration of caprock minerals
by carbon dioxide. Geofluids 7 (2), 112e122.
conditions, the water density at reservoir conditions can be calcu-
Chiquet, P., Daridon, J.L., Broseta, D., Thibeau, S., 2007b. CO2/water interfacial
lated as, tensions under pressure and temperature conditions of CO2 geological storage.
Energy Conversion and Management 48 (3), 736e744.
rwðstandardÞ Danesh, A., 2007. PVT and phase behaviour of petroleum reservoir fluids. Elsevier.
rw ¼ Espinoza, D.N., Santamarina, J.C., 2010. Water-CO2-mineral systems: interfacial
Bw
tension, contact angle, and diffusion-Implications to CO2 geological storage.
Where Bw is the formation water volume factor at standard Water Resour Res 46, W07537.
Hildenbrand, A., Schlomer, S., Krooss, B.M., Littke, R., 2004. Gas breakthrough
conditions which can be calculated knowing the reservoir P and T experiments on pelitic rocks: comparative study with N-2, CO2 and CH4. Geo-
(McCain, 25). fluids 4 (1), 61e80.
 M. Naylor et al. / Marine and Petroleum Geology 28 (2011) 1083e1093 1093

Li, S., Dong, M., Li, Z., Huang, S., Qing, H., Nickel, E., 2005. Gas breakthrough pressure Shah, V., Broseta, D., and Mouronval, G., (2008a) Capillary alteration of caprocks
for hydrocarbon reservoir seal rocks: implications for the security of long-term with acid gases. SPE paper 113353, presented at the 2008 SPE IOR Symposium,
CO2 storage in the Weyburn field. Geofluids 5 (4), 326e334. Tulsa.
Li, Z.W., Dong, M.Z., Li, S.L., Huang, S., 2006. CO2 sequestration in depleted oil and Shah, V., Droseta, D., Mouronval, G., Montel, F., 2008b. Water/acid gas interfacial
gas reservoirs - caprock characterization and storage capacity. Energy Conver- tensions and their impact on acid gas storage. International Journal of Green-
sion and Management 47 (11e12), 1372e1382. house Gas Control 2, 594e604.
Massoudi, R., King, A.D., 1975. Effect of pressure on surface tension of aqueous Stouffer, C.E., Kellerman, S.J., Hall, K.R., Holste, J.C., Gammon, B.E., Marsh, K.N., 2001.
solutions. Adsorption of hydrocarbon gases, carbon dioxide, and nitrous oxide Densities of carbon dioxide plus hydrogen sulfide mixtures from 220 K to 450 K
on aqueous solutions of sodium chloride and tetra-n-butylammonium bromide at pressures up to 25 MPa. Journal of Chemical and Engineering Data 46 (5),
at 25 . Journal of Physical Chemistry 79 (16), 1670e1675. 1309e1318.
Moss, B., Barson, D., Rakhit, K., Dennis, H., Swarbrick, R., 2003. Formation pore Tonnet, N., Broseta, D. and Mouronval, G. (2010) Evaluation of the petrophysical
pressures and formation waters. In: Evans, D., Graham, C., Armour, A., properties of a carbonate-rich caprock for CO2 geological storage purposes, SPE
Bathurst, P. (Eds.), The Millennium Atlas: Petroleum Geology of the Central and 131525.
Northern North Sea. Geological Society, London, pp. 317e329. Yang, D.Y., Gu, Y.A., Tontiwachwuthikul, P., 2008. Wettability determination of
Ren, Q.Y., Chen, G.J., Yan, W., Guo, T.M., 2000. Interfacial Tension of (CO2þCH4) plus the crude oil-reservoir brine-reservoir rock system with dissolution of CO2 at
water from 298 K to 373 K and pressures up to 30 MPa. Journal of Chemical and high pressures and elevated temperatures. Energy and Fuels 22 (4),
Engineering Data 45 (4), 610e612. 2362e2371.
Rezaee, M.R., Jafari, A., Kazemzadeh, E., 2006. Relationships between permeability, Yang, D.Y., Tontiwachwuthikul, P., Gu, Y.G., 2005. Interfacial tensions of the crude oil
porosity and pore throat size in carbonate rocks using regression analysis and plus reservoir brine þ CO2 systems at pressures up to 31 MPa and temperatures
neural networks. Journal of Geophysics and Engineering, 370. 4. of 27 degrees C and 58 degrees C. Journal of Chemical and Engineering Data 50
Selley, R.C., 1998. Elements of Petroleum Geology. Academic Press. (4), 1242e1249. 10.1021/je0500227.