PROCEEDINGS, 1st ITB Geothermal Workshop 2012
Institut Teknologi Bandung, Bandung, Indonesia, March 6-8 , 2012
GEOTHERMAL ENERGY COMBINED WITH CO
2SEQUESTRATION:
AN ADDITIONAL BENEFIT
Hamidreza Salimi, Karl-Heinz Wolf, and Johannes Bruining Department of Geotechnology, Delft University of Technology
Stevinweg 1, 2628 CN, Delft, The Netherlands
e-mail: h.salimi@tudelft.nl; k.h.a.a.@tudelft.nl; j.bruining@tudelft.nl
ABSTRACT
In this transition period from a fossil-fuel based society to a sustainable-energy society, it is expected that CO2 capture and subsequent sequestration in
geological formations plays a major role in reducing greenhouse gas emissions. An alternative for CO2
emission reduction is to partially replace conventional-energy for heating and cooling buildings (e.g., cogeneration units) with geothermal energy.
A mixture of CO2 with cold return water injected into
geothermal reservoirs can be the integration of geothermal-energy production and subsurface CO2
storage. In this process, mixed CO2-water is injected,
while hot water is simultaneously produced from the production well. The process may end when CO2
either in the aqueous phase or in the CO2-rich phase
breaks through. It depends on the function of the CO2
as being a stored medium or being a source for gas-drive water production.
In this study, we discuss the influence of various CO2
injection volumes on low-enthalpy production mechanisms/storage of CO2 in relation to the
production of hot water from the geothermal aquifer systems. Furthermore, we provide injection-screening conditions for optimal geothermal recovery, maximal storage of CO2 and/or re-use of a CO2-water cycle.
For any energy source, one important attribute is the recovery efficiency, i.e., how much energy can be extracted from this source with respect to the amount of energy invested during the process of energy extraction. In this work, we estimate the total amount of energy invested for mixed CO2-water injection
into the geothermal reservoir, using an “effective-energy” analysis. Furthermore, we provide a cursory evaluation of the economics of the proposed project assuming that we can relate the energy balance to an economic analysis. In such a conversion, the notion intensity of embodied energy plays a central role. We also introduce a plot of the heat-energy extraction and the storage capacity, which can be used to locate
optimal in-situ conditions. These results, which are plotted in a heat-energy/storage-capacity diagram, are discussed in detail. Aspects regarding specific geo- and technical infrastructure are ignored.
Papers should follow standard technical paper format, an abstract followed by the more detailed presentation. The abstract should be typed on this area of the first page, with the presentation following after two lines of space.
INTRODUCTION
Concern about global warming is driving research and development aimed at reducing the emissions of greenhouse gases such as CO2. One way of reducing
CO2 emissions is to partly replace conventional
energy sources for heating buildings (e.g., cogeneration units) with geothermal energy. To reduce CO2 emission further, a feasibility study is
ongoing to also capture the CO2 and coinject it with
the cooled-down-return water of the geothermal system. In this way, synergy is established between geothermal energy production and subsurface CO2
storage. However, major issues such as how to estimate CO2-breakthrough time, and the influence of
CO2 coinjection on heat extraction have not been
completely addressed. In this paper, we use the simulation results of mixed CO2-water injection for
various injected CO2 concentrations to give a
complete overview of optimal heat recovery and maximally stored CO2 for a selected heterogeneity
structure derived from the Delft Sandstone Member. Injection of CO2 into geothermal reservoirs is a new
research topic that emerged with the increased interest in geothermal energy production (Wolf et al. 2008; Brown 2000; Pruess 2006, 2008). Brown (2000) proposed a novel enhanced geothermal systems (EGS) concept that uses CO2 instead of
water as heat transmission fluid, to achieve geologic sequestration of CO2 as an ancillary benefit.
Following up on his suggestion, Pruess (2006, 2008) evaluated the thermophysical properties and performed numerical simulations to explore the fluid
dynamics and heat transfer issues in an engineered geothermal reservoir that would be operated with CO2. In addition, Xu and Pruess (2010) performed
chemically reactive transport modeling to investigate fluid-rock interactions in an EGS operated with CO2.
Our work differs from the mentioned works, because we do not use CO2 as a working fluid, but we coinject
CO2 into a geothermal reservoir to store CO2
simultaneously with hot-water extraction.
Our rationale for investigation of the process of mixed CO2-water injection into geothermal reservoirs
is as follows. For high-injection-rate pure CO2
sequestration processes, there are issues of formation dry-out (salt precipitation), high risk of leakage because of buoyancy effects, inefficient CO2
-trapping, over-pressurizing the aquifers such that the reservoir pressure exceeds the maximum formation stress and consequently possible formation damage and fracturing occur that enhance the risk of CO2
leakage to the surface.
However, when we inject moderate amounts of a mixture of CO2 and colder return water into
geothermal reservoirs, we circumvent those drawbacks. In addition, the entire (or a part of the) energy needed to capture and compress CO2 is
compensated by produced geothermal energy. From the economic point of view, there are two separate expensive projects (one for CO2 storage and the other
one for geothermal); it would be more economically viable to combine these two projects as some of equipments and facilities are common between these two projects.
Even without considering the geothermal option, several researchers have proposed to inject mixtures of water/brine and CO2, instead of pure CO2, to
improve aquifer storage of CO2 (Bryant et al. 2008;
Delshad et al. 2010). The combined injection enhances residual trapping, the reduced mobility ratio enhances spreading, and single-phase dissolved CO2
injection avoids confining the CO2 to the upper part
of the reservoir decreasing the leak risk via the cap rock. We consider only capillary pressure and gravity, and take into account heat conduction within the porous rock. However, effects of salinity and reactions are not taken into account. Including chemical reactions is kept for future work. A complete set of simulations can be used to obtain a first estimate of conditions for either optimal heat-energy recovery or optimal CO2 storage capacity.
The objective of this paper is to investigate the influence of injected CO2 concentrations on the
efficiency of CO2 sequestration and heat extraction in
aquifers, and to assess the applicability of the non-isothermal NegSat solution approach. Therefore, the solution approach must be capable to model the phase disappearance/appearance as well as the phase
transition between subcooled and supercritical behavior. From the mathematical point of view, there is no difference between a single-phase subcooled region and a single-phase supercritical region. By subcooled, we mean below the critical temperature of CO2 but above the critical pressure. Supercritical
means both above the critical temperature and above the critical pressure. However, from the thermodynamic point of view, the model must be able to distinguish between subcritical and supercritical single-phase regions. Subcritical CO2
(i.e., below the critical pressure and temperature) is not considered in this paper. Note that the phase states involved are liquid, gas and supercritical. We also introduce a plot of the heat-energy extraction (energy that can be exploited) and the storage capacity, which can be used to locate optimal in-situ conditions.
The paper is organized as follows. First, we briefly review the history and the framework of the Delft geothermal project. Then, we discuss the results of a study on the reservoir geology and petrophysics of the aquifer, from which we obtained the heterogeneous porosity and permeability distribution in the aquifer. After that, we explain our novel negative saturation (NegSat) solution approach for non-isothermal compositional flow to deal with problems that involve phase appearances and disappearances. Subsequently, we give a set of simulation results that show the possible bifurcations of cold mixed CO2-water injection into a geothermal
reservoir. These results, which are plotted in a heat-energy/storage capacity diagram, are discussed in detail. Finally, we summarize our findings in the conclusion section.
THE DELFT GEOTHERMAL PROJECT
The Delft Geothermal Project is a consortium of governmental and industrial partners that aims to develop an innovative geothermal system at the campus of Delft University of Technology (DUT). Annually, DUT consumes 11 million m3 of gas, thus producing approximately 22 ktonnes of CO2 of which
15 ktonnes are attributed to the generation of electricity. The planned geothermal system is designed to contribute up to 5 MW, which results in a CO2 emission reduction of approximately 10 ktonnes.
To reduce CO2 emission further, a feasibility study is
ongoing to also capture the CO2 and coinject it with
the cooled-down-return water of the geothermal system. In this way, synergy is established between geothermal energy production and subsurface CO2
storage.
Students of DUT, Department of Geotechnology, initiated the Delft Geothermal Program (DAP). DAP started a feasibility study where innovation (casing drilling technology using composite materials, and
optional CO2 coinjection) is combined with the
commercial production of geothermal energy for heating offices and student houses on the campus. In 2009, the University was granted an exploration and production license for geothermal energy by the government. In this 60 km2-licensed area, three projects are under development: two doublets for heating glasshouses and public buildings and one for the combined geothermal/cogeneration power plant for the university. Drilling of the four wells for the two doublets of the glasshouse farmers is almost completed. Subsurface information from these wells is and will be used to further improve the reservoir geological and petrophysical knowledge of the aquifers present under the licensed area.
RESERVOIR CHARACTERIZATION
The geothermal doublet will consist of two wells: a hot-water production well, and a cold-water reinjection well. The wells target the Delft Sandstone Member, a fluvial sandstone formation contained in a structural low at a depths ranging from 1.7 to 2.3 km below surface. The Delft Sandstone Member is deposited in the western and central parts of the West Netherlands Basin. It is a light-grey, fine to coarse-gravelly, massive sandstone sequence with abundant lignitic matter, which varies in thickness between 0 (absent) and 130 m (Smits 2008).
The geothermal potential of this aquifer strongly depends on two key parameters: the temperature of the formation water in the aquifer (heat in place), and the permeability of the aquifer. Temperature of the formation water is a function of depth below surface. Bottomhole temperature readings indicate that the geothermal gradient in the area is approximately 3K/100m (Smits 2008), which results in a temperature estimate of 338−353 K for the formation water contained in the Delft Sandstone Member. This is sufficient for use in a low-temperature grid-heating network or glasshouse heating system. The second parameter to be estimated is the permeability of the aquifer, which directly influences the production (and reinjection) rate that can be reached. For this purpose, a reservoir characterization study of the Delft Sandstone Member was done to assess the spatial distribution of the fluvial sandstone bodies, their connectivity, and their internal permeability heterogeneity (Groenenberg et al. 2010). In this section, the results of this study and their implications for compositional flow simulations of cold CO2-water
injection into a geothermal reservoir are briefly elucidated.
Knowledge on the spatial variability in reservoir properties, such as porosity and permeability, is of primary importance in evaluating the flow of hot and cold water through formations. Appraisal drilling for hydrocarbons in and around the target area has
indicated that the Delft Sandstone Member has promising reservoir qualities (i.e., high porosity and permeability). DAP_GT_01 DAP_GT_02 10600 10800 11000 11200 11400 11600 11800 10600 10800 11000 11200 11400 11600 11800 Distance, [m] -2 1 6 0 -2 0 8 0 -2 0 0 0 -1 9 2 0 -1 8 4 0 -1 7 6 0 -2 1 6 0 -2 0 8 0 -2 0 0 0 -1 9 2 0 -18 4 0 -1 7 6 0 Z , [m ] 10600 10800 11000 11200 11400 11600 11800 10600 10800 11000 11200 11400 11600 11800 Distance, [m] -2 1 6 0 -2 0 8 0 -2 0 0 0 -1 9 2 0 -1 8 4 0 -1 7 6 0 -2 1 6 0 -2 0 8 0 -2 0 0 0 -1 9 2 0 -18 4 0 -1 7 6 0 Z , [m ] 10800 11200 11600 Length, m -2 2 0 0 -2 0 0 0 -1 8 0 0 -2 2 0 0 -2 0 0 0 -1 8 0 0 D e p th , m Sand Silty Sandstone Siltstone Coal Facies
Fig. 1: Cross-section through the reservoir model between the planned injection well
(DAP_GT_01) and production well
(DAP_GT_01). Colors represent the type of sediment (facies) present in the cell (see legend).
5 10 15 20 25 30 35 5 10 15 20 Nx Nz 20 40 60 80 100 120
Fig. 2: Permeability distribution (md) of the vertical x-z cross-section.
5 10 15 20 25 30 35 5 10 15 20 Nx Nz 0.05 0.1 0.15 0.2 0.25
Fig. 3: Porosity distribution of the vertical x-z cross-section.
A three-dimensional reservoir model was constructed based on the available seismic data and well data of existing oil- and gas wells in the region. Object-based facies modeling was applied to fill in the space
between the wells with different types of sediment. Based on the log, core, and cuttings analyses on the wells in the Delft and Moerkapelle oilfields (Drost and Korenromp 2009; Loerakker 2009), four different types of sediment were defined: sand, silty sandstone, siltstone, and coal. Each type has its own characteristic range of porosities and permeabilities attached to it, which were obtained from the core and cuttings analyses. Furthermore, object-based facies modeling requires input on the expected geometry of the fluvial sand bodies that is partly derived from the depositional setting, such as the direction and size of the meandering channel belt. Channel thickness is set as measured in cores, and varies from 1.5 to 4.5 m with an average of 3 m (Loerakker 2009). Fig. 1 displays a cross section through the facies model between the planned injection and production well of the Delft geothermal doublet, which shows the level of heterogeneity in the Delft Sandstone Member. Figs. 2 and 3 display the permeability and porosity distribution of a cross section through the facies model between the planned injection and production well of the Delft geothermal doublet.
RESERVOIR SIMULATION
In cold mixed CO2-water injection into geothermal
reservoirs, often regions of two-phase flow are connected to regions of single-phase flow. Different systems of equations apply for single-phase and for two-phase regions. We apply the non-isothermal negative saturation (NegSat) solution approach (Salimi et al. 2011b, 2011c, 2012) to solve efficiently non-isothermal compositional flow problems (e.g., CO2-water injection into geothermal reservoirs) that
involve phase appearance, phase disappearance, and phase transitions. The advantage of this solution approach is that it circumvents using different equations for single-phase and two-phase regions and the ensuing unstable switching procedure. In the NegSat approach, a single-phase multi-component fluid is replaced by an equivalent fictitious two-phase fluid with specific properties. The equivalent properties are such that in the single-phase aqueous region, the extended saturation of a fictitious gas is negative. The saturation of the equivalent gas Ŝg is called the extended gas saturation; it reads
ˆ ˆ , 1, 2,..., . ˆ ˆ i iw g c ig iw z x S i N x x − = = − (1)
In Eq. 1, zi is the overall mole fraction of component i and ˆxiαis the mole fraction of component i in phase α.
When the extended gas saturation is between zero and one, it is the same as the actual gas saturation and there are two phases. If the extended gas saturation is above one, we have a single gaseous phase and the actual gaseous saturation is one. If the extended gas saturation is below zero, we have a single liquid phase and the actual gas saturation is zero. Further
details about the non-isothermal NegSat solution approach and non-isothermal compositional multi-phase equations can be found in Salimi et al. (2011b, 2011c, 2012).
Numerical Model
We consider cold mixed CO2-water injection into a
geothermal reservoir modeled as a two-dimensional (2D) vertical porous medium initially filled with hot liquid water. In the model, we consider capillary pressure and gravity, but we disregard the diffusion flux of components in phases. Indeed, the diffusion coefficient in the temperature range of interest is between 10−8–10−9 m2/s (see, e.g., Gmelin 1973). For a typical Darcy velocity (0.04 m/day ~ 5 × 10−7 m/s) and a reservoir length of 1500 m, the Peclet number is in the range of 106–107. As we are solving the equations numerically, we inevitably introduce numerical diffusion. We take into account heat conduction within the porous rock, but for illustration purposes, we avoid the complication of heat loss to the overburden and underburden as it has minor effects until CO2 breakthrough.
We use empirical data for the transport parameters. The viscosities of the liquid and gaseous phase are approximated by the viscosities of CO2 and water as
functions of temperature (Perry and Green 1997); they read 3 6 6 2 12 3 1.6128 10 9.0436 10 0.0135 10 1.9476 10 , g T T T µ − − − − = × − × + + × − × (2) (247.8 /( 140)) 5 2.414 10 10 T . l µ = × − × − (3)
In Eqs. 2 and 3, T is the absolute temperature in Kelvin and the viscosity is in Pa·s. These equations disregard the compositional dependence as being negligible (Sayegh and Najman 1987; Chang et al. 1988; Frank et al. 1996). To calculate the heat capacity of phase α, we assume that the heat capacity of phase α is the sum of the molar heat capacities of its components (i.e., we disregard the enthalpy of mixing). We use the following expressions for the heat capacity of CO2 and water (Perry and Green
1997): 2 3 ,CO 5 2 45.369 8.6881 10 9.6193 10 , p C T T − − = + × − × (4) 2 2 ,H O 3 2 3 3 9 4 2.7637 10 2.0901 8.125 10 0.014116 10 + +9.3701 10 . p C T T T T − − − = × − + + × − × × (5)
Eqs. 4 and 5 express the heat capacity in (J/mol/K); T is the absolute temperature in Kelvin. The gas and
liquid Brooks-Corey relative-permeability functions used in the numerical model are
(
)
2 3 1 , rl g k S λ λ + = − (6) 2 2 1 (1 ) , rg g g k S S λ λ + = − − (7)where λ is the sorting factor. We use λ = 2 in the numerical model. We use a zero connate water saturation Swc = 0.
We consider a geothermal reservoir with a length of 1500 m, a width of 1500 m, and a height of 60 m. Initially, the reservoir is saturated with hot water and by that, the initial gas (CO2) saturation in the
geothermal reservoir is equal to zero. A cold mixture of CO2-water is injected through the entire
cross-section of the reservoir from the left side and, subsequently, water and CO2 are produced through
the entire cross-section of the reservoir at the opposite side. Table 1 shows the basic input data for the numerical simulations. The geothermal-reservoir properties correspond to a real geothermal reservoir that is located in the West Netherlands basin, in particular the early Cretaceous Delft sandstone member, below the city of Delft, The Netherlands. Figs. 2 and 3 display the permeability and porosity distribution of a cross section through the facies model between the planned injection and production well of the Delft geothermal doublet, which show the level of heterogeneity in the Delft Sandstone Member (Gilding 2010; Salimi and Wolf 2012).
Table 1: Data used in the numerical simulations.
Maximum injection pressure (bar) 255 Bottomhole production pressure (bar) 205 Initial Temperature (K) 353.15 Injection Temperature (K) 293.15 Maximum water–Injection rate (m3/s) 0.04167 Rock grain density (kg/m3) 2650 Rock specific heat capacity (J/kg/K) 1000 Total thermal conductivity (W/m/K) 2.1 Geometric mean permeability (mD) 21.6
Mean porosity 0.17
Residual water saturation 0 Residual gas saturation 0 Number of grid cells (Nz×Nx) 23×35 We discretize the geothermal reservoir into 23×35 grid-cells in the vertical and horizontal direction. The number of grid cells is the same as that used for the permeability and porosity distribution (Figs. 2 and 3). For discretization in space, we use the implicit upwind finite-volume method with the cell-centered scheme. For each simulation case, the water-injection rate is uniform. Moreover, the initial target mode for the injection well is the uniform water-injection rate.
However, a maximum bottomhole pressure of 255 bars acts as a constraint for the injection well. If the bottomhole pressure of the injection well exceeds 255 bars during the simulation, the maximum bottomhole pressure becomes the target mode for the injection well and the injection rate is reduced accordingly.
RESULTS AND DISCUSSION
In the simulations, we vary the overall mole fraction of the injected CO2. Based on the possible phase
sequences for CO2-water injection into a geothermal
reservoir and flow patterns, we discuss the results of two cases in terms of the overall injected CO2 mole
fraction (viz., 0.02 and 0.2) for the discretized heterogeneous porosity and permeability field shown in Figs. 2 and 3 (Salimi et al. 2012). A complete set of simulations is described in Salimi and Wolf (2012). For a low overall injected CO2 mole fraction
of 0.02, we have only a single phase in the entire reservoir from the injection to the production side. For a high overall injected CO2 mole fraction of 0.2,
there is also a two-phase region at the injection side. In all cases, we continue to inject cold mixed CO2
-water into the reservoir and produce hot -water until either CO2 breakthrough or cold-water breakthrough
(for pure cold-water injection) occurs.
Low-CO2-Injection-Concentration Case
This case has an overall injected CO2 mole fraction
of 0.02 (or about 49.9 kg of CO2 per ton of water) in
the heterogeneous permeability and porosity field shown in Figs. 2 and 3. The extended gas saturation is below zero for the entire time span considered, indicating that all the injected CO2 is completely
dissolved into the aqueous phase. Fig. 4a shows the extended-gas-(CO2)-saturation distribution at t = 30
yr for this case. Fig. 4a thus essentially displays a single-phase displacement process in the entire domain. Note that when the extended gas saturation becomes equal to zero, the phase state of the reservoir would be at the bubble-point state. Fig. 4a clearly reveals that the saturation distribution is spread out (mixed) by the permeability and porosity distribution of the reservoir, and thus the computed distribution for the extended gas saturation is dispersive.
Fig. 4b shows the temperature profiles at t = 30 yr for case 2. Evidently, the temperature distribution is smoothed. This is attributed to the high value of the thermal diffusion coefficient of the reservoir rock and, therefore, the temperature profile in the highly permeable zones is retarded and it is accelerated in the less permeable zones (see Fig. 4b).
By combining the temperature distribution with the extended-gas-saturation distribution, we obtain the corresponding overall CO2-mole-fraction distribution,
computation of the densities that are uniquely determined by the extended gas saturation, temperature and pressure. Fig. 4c clearly illustrates again that the CO2 front moves ahead of the
temperature front. As can be seen in Fig. 4c, CO2
breakthrough has already occurred before 30 years. Furthermore, the maximum value of the overall CO2
mole fraction at t = 30 yr is 0.02. Because the highest value is the same as the overall injected mole fraction, there is no CO2 bank in this case. The
distribution shown in Fig. 4c is dispersive and therefore the mixing is more efficient. This is attributed to a very large variation permeability values that are smaller than 0.01 mD and larger than 382 mD in combination with a low mobility ratio. The mobility ratio, M, is the ratio of the injected fluid mobility (kr/µ) to the initial fluid mobility.
Length, m D e p th , m 214 429 643 857 1071 1286 1500 13 26 39 52 -0.02 -0.018 -0.016 -0.014 -0.012 -0.01 -0.008 -0.006 -0.004 -0.002
Fig. 4a: Extended–gas–saturation distribution at t =
30 yr for zinj = 0.02. The extended gas
saturation is below zero, meaning the absence of a gas phase.
Length, m D e p th , m 214 429 643 857 1071 1286 1500 13 26 39 52 300 310 320 330 340 350
Fig. 4b: Temperature distribution (in K) at t = 30 yr
for zinj = 0.02. The high value of the
thermal diffusion coefficient smoothes the
temperature variations due to
permeability contrasts. Length, m D e p th , m 214 429 643 857 1071 1286 1500 13 26 39 52 0 0.005 0.01 0.015 0.02
Fig. 4c: Overall–CO2–mole–fraction distribution at t
= 30 yr for zinj = 0.02.
We note that the compositional front shown in Fig. 4c is far ahead of the temperature front shown in Fig. 4b, because the ratio of the compositional wave to the thermal wave in this single-phase displacement example would be
(
)( )
( )
1 1 1, c s th l c v v c ϕ ρ ϕ ρ − = + >> (8)where vc and vth are the speeds of the compositional wave and thermal wave respectively, φ is the porosity, (ρc) is the heat capacity per unit volume, and the subscripts s and l denote the solid and the liquid, respectively. We compute that approximately
vc/vth = 4.1.
High-CO2-Injection-Concentration Case
Here we consider an overall injected CO2 mole
fraction of zinj = 0.20 in the heterogeneous permeability and porosity field. With an overall CO2
mole fraction of zinj = 0.20 and, considering the injection temperature and pressure, the extended gas saturation is above zero at the injection side (Ŝg,inj = 0.32). It implies that there are two phases at the injection side, as opposed to the previous case. Fig. 5a shows the extended-gas-saturation distribution at t = 6.5 yr. Furthermore, it illustrates a channeling pattern for the extended gas saturation as opposed to the previous case. The extended gas saturation attains a maximum value of Ŝg,max = 0.92 at t = 6.5 yr. The maximum value is larger than the maximum extended gas saturation obtained with an overall injected mole fraction of zinj = 0.02. This is attributed to a larger ratio of the CO2-injection rate to the water-injection
rate in the case with zinj = 0.20. The supercritical CO2
is a light (xg/xl > 1) component compared to water. As a result, when the injection ratio increases for approximately the same range of pressure and temperature variations, the gas (supercritical CO2)
Fig. 5b illustrates that the cold-temperature front does not considerably penetrate into the reservoir. This is attributed to the fact that as the amount of injected CO2 increases, the difference between the speed of
the thermal front and the speed of the compositional front becomes larger. The reason for this is that heat transfer by the aqueous phase is more efficient than heat transfer by the gas phase.
Length, m D e p th , m 214 429 643 857 1071 1286 1500 13 26 39 52 0 0.05 0.1 0.15 0.2 0.25 0.3
Fig. 5a: Extended–gas–saturation distribution at t =
6.5 yr for zinj = 0.2. The displacement
indicates appreciable channeling.
Length, m D e p th , m 214 429 643 857 1071 1286 1500 13 26 39 52 335 340 345 350
Fig. 5b: Temperature at t = 6.5 yr for zinj = 0.2.
Length, m D e p th , m 214 429 643 857 1071 1286 1500 13 26 39 52 0 0.05 0.1 0.15 0.2
Fig. 5c: Overall-CO2-mole-fraction distribution for
zinj = 0.2 at t = 6.5 yr. A zero overall CO2
mole fraction downstream is indicated by a white color.
Fig. 5c shows the overall-CO2-mole-fraction
distribution at t = 6.5 yr. The overall CO2 mole
fraction reaches a maximum of zmax = 0.76 at t = 6.5 yr. In addition, Fig. 5c illustrates that a channeling regime occurs for this case. With channeling, we mean that the CO2 plume develops along the highly
permeable streaks (i.e., the progress of CO2 plumes is
dominated by the permeability distribution in combination with a high mobility ratio). Due to channeling, water is bypassed and thus dissolution is not efficient. This leads to earlier breakthrough and a small sweep efficiency. By comparing the CO2
concentrations shown in Figs. 4c and 5c, we notice that the displacement mechanism changes from dispersive to channeling as the overall injected CO2
mole fraction increases from 0.02 to 0.20 for the same heterogeneous reservoir structure. This is because of the difference in mobility ratios. The mobility ratio for the case with zinj = 0.03 is approximately M = 1, because CO2 was completely
dissolved into water at the injection side. On the other hand, the mobility ratio for the case with zinj = 0.20 is approximately M = 5.7 because two phases (aqueous phase and CO2-rich phase) were injected.
As is well known, adverse-mobility-ratio (M > 1) displacements will always have smaller sweep and displacement efficiency than those corresponding to
M ≤ 1 displacements. Further details about the effect of heterogeneity on the character of flow in porous media can be found in (Koval 1963; Waggoner et al. 1992; Farajzadeh et al. 2011).
Energy Invested
For any energy source, one important attribute is the recovery efficiency, i.e., how much energy can be extracted from this source with respect to the amount of energy invested during the process of energy extraction. Here we estimate the total amount of energy invested for mixed CO2-water injection into
the geothermal reservoir, using a useful energy analysis (Dincer and Rosen 2007).
We prefer to do the full calculations, because otherwise we need to resort to imprecise results of rounded computations, which may be physically reasonable but difficult to follow. Indeed, we are well aware that the numbers quoted below have at most two significant digits.
We assume that the total energy invested in this process consists of energy for producing materials (steel and cement), drilling, CO2 capture and
compression, and circulation pumping power. First, we calculate the energy consumed to produce materials (steel and cement) for the production and injection well. Based on the depth of the geothermal reservoir (2 km), the amount of steel needed for the injection-well casing is approximately 248572 (250000) kg and for the production-well casing is 392549 (400000) kg (De Mooij 2010). The total weight of steel for both wells is 641121 kg. Typical
energy required to produce steel is 60 MJ/kg (Costa et al. 2001). Therefore, the total energy consumed for the steel tubing is 60 MJ/kg × 641121 kg = 38467.26 GJ. Secondly, we obtain the energy used to produce cement. The total amount of cement needed for the injection and production well is 223207 kg (De Mooij 2010). The specific energy for producing cement is approximately 2950 kJ/kg. Hence, the total energy consumption for production of cement for two wells is 223207 kg × 2950 kJ/kg = 658.46 GJ. Thirdly, the energy invested in drilling the geothermal doublet is about 8712 GJ (Samuel and McColpin 2001). Fourthly, the energy demand for CO2 capture and compression is 1.2 MJ/kg CO2
(Freeman and Rhudy 2007). Fifthly, a pump is needed to inject a mixture of water and CO2, via dual
completion, into the reservoir. The pumping power should be such that it compensates the power loss along the well tubes and the power loss within the reservoir. The power loss in the tubes is a result of friction of the fluid to the tubing surface of both wells. Based on an injection rate of 0.04167 m3/s (Table 1) and a pump efficiency of 0.36, the pumping power is 0.30 MW (0.09 MW for the power loss in the wells and 0.21 MW for the power loss in the reservoir).
Note that the energy invested in materials and drilling is assumed to be constant during time. The energy demand for CO2 capture and compression and
pumping power is demand controlled and therefore depends on time. Moreover, the energy invested in materials, drilling, and pump is relatively small compared to the energy needed for CO2 capture and
compression. Below, we present the total energy invested for each case considered. The only source of energy production is the amount of heat produced from the geothermal reservoir.
Recovered Heat Energy Versus Maximally Stored CO2
Fig. 6 plots two curves: the blue-star points that show the recuperated heat energy versus the maximally stored CO2 at the end of the process and the
black-solid triangular points that represent the total energy consumed for each case. The figure includes the results for various overall injected CO2 mole
fractions. For all cases, the initial reservoir conditions, water-injection rate, and injection temperature are the same. However, the overall injected CO2 mole fraction is different for each case.
Along the blue-star curve, the overall injected CO2
mole fraction essentially increases from left to right. With no added CO2, the criterion to end the project is
cold-water breakthrough, while if any amount of CO2
is added, the criterion to end the project is when CO2,
dissolved into the aqueous phase, breaks through. If the entire pore volume of the reservoir were filled with CO2 at T = 353.15 K and P = 220 bars, a total
CO2-storage capacity of 13,678 ktonnes would be
attained. When no CO2 were added, a total
geothermal-energy production of 16,492 TJ (1 T = 1012) could be achieved. Table 2 shows the CO2
-breakthrough time, cumulative CO2 mass injected up
to the CO2-breakthrough time (i.e., stored CO2), and
cumulative heat-energy production for each case. First, we describe the blue-star-point curve. Then, we discuss the energy balance over the entire process for each case. After that, we provide a cursory evaluation of the economics of the proposed project.
The blue-star-point curve consists of six regions in terms of heat energy/stored CO2. Region 1 pertains to
cases of small overall CO2 mole fractions (0.002 < zinj < 0.02). In region 1, the heat energy decreases, but the stored CO2 increases because the ratio of the
CO2-injection rate to the water-injection rate
increases. In region 1, the overall CO2 mole fraction
is low. Consequently, the displacement process is in the aqueous single–phase region for the entire domain. From region 1 to region 2 (0.02 < zinj ≤ 0.03), the heat–energy/stored–CO2 behavior changes
and a transition occurs. In region 2, there is a drastic increase in both the heat-energy production and the CO2 storage capacity. In this region, the injected CO2
concentration is still low such that CO2 remains
completely dissolved into the aqueous phase at the injection side. However, gaseous CO2 is formed
further downstream. The sharp transition (significant increase) in heat energy and stored CO2 occurs for
the injected CO2 concentration (zinj = 0.03) that is very close to its maximum solubility limit in water at the injection temperature (293 K) and pressure (255 bars). Therefore, gaseous CO2 is first formed because
the temperature starts to increase, but it disappears further downstream because of lack of available CO2,
and consequently the CO2 concentration again drops
to below its solubility limit. We infer that there is a competition between evaporation and condensation. The frequent evaporation and condensation is the main reason to delay CO2 breakthrough and
consequently increase considerably both the cumulative heat–energy production and the maximally stored CO2. 0 500 1000 1500 2000 2500 0 1000 2000 3000 4000 5000 Stored CO 2, kton H e a t E n e rg y , T J Heterogeneous Energy Invested z=0.002 z=0.025 z=0.03 z=0.09 z=0.0275 z=0.02875 z=0.04 z=0.02 z=0.3 z=0.99 z=0.13 z=0.15
Fig. 6: Cumulative heat–energy production and energy invested versus maximally stored
CO2 at CO2 breakthrough either in the
aqueous or in the gaseous phase. We use z
to denote the overall injected CO2 mole
fraction. The trend from left to right
represents an increase in the injected CO2
mole fraction. The point that corresponds
to cold–water injection (i.e., no CO2) is
located on the y–axis (not shown here).
As the overall CO2 mole fraction increases from zini = 0.03 to zinj = 0.04, a gas (supercritical CO2) phase is
formed at the injection side (i.e., two-phase injection) and another bifurcation occurs. In region 3 (0.03 < zinj < 0.05), both the recovered heat energy and the maximally stored CO2 decrease drastically from 5044
TJ to 1902 TJ and from 1429 ktonnes to 694 ktonnes, respectively. The main reason for this sharp change in the heat-energy/stored-CO2 behavior is that the
competition between evaporation and condensation, which slows down the CO2 movement towards the
production well, weakens because of higher CO2
supply and the existence of two phases at the injection side. Consequently, it leads to earlier breakthrough and a shorter mixed-CO2
-water-injection period.
In region 4 (0.05 < zinj≤ 0.09), the recuperated heat energy remains more or less constant (i.e., it changes from 1902 TJ to 1893 TJ), but the stored CO2
increases from 694 ktonnes to 1099 ktonnes. The reason for this behavior is as follows. When the concentration is still low, the single–phase compositional–rarefaction wave is still the fastest (leading) wave and the sequence of waves has not yet changed.
At still higher CO2 injection (0.09 < zinj < 0.13), compositional–rarefaction wave is overtaken by a compositional shock, which moves much faster than the compositional–rarefaction wave in region 4. Therefore, in region 5, the time of breakthrough gradually decreases and both the possible amount of stored CO2 and the heat energy decrease. Note that
still the mobility ratio M between the injected fluid and the initial reservoir fluid is below one. This continues until zinj = 0.13.
Finally, from region 5 to region 6 (0.13 < zinj < 1), the heat-energy/stored-CO2 behavior changes because the
sequence of waves changes and a Buckley-Leverett wave is formed. Furthermore, a transition from dispersive to channeling occurs because of high mobility ratios (M > 1). For that reason, the breakthrough time of the compositional shock decreases. However, there is an increase in storage capacity due to the higher CO2/water injection ratios,
which sufficiently compensates earlier breakthrough. Moreover, in region 6, the energy recovery remains
approximately constant. This is because of the higher productivity index due to well constraints, which amply compensates earlier breakthrough.
In summary, including heterogeneity, we have a choice of an optimal energy recovery of 5044 TJ with a limited storage capacity of 1429 ktonnes or a minimal energy recovery of 812 TJ with a maximal storage capacity of 2433 ktonnes.
The regions, as discussed previously, can be changed by different heterogeneous structures, well outlays, well constraints, and salinity; however, they are useful for showing the prevailing features.
Net Energy Balance
In Fig. 6, the black-solid-triangular points represent the total energy consumed for each case. Fig. 6 shows that those points with an overall injected CO2 mole
fraction less than 0.10 are above their corresponding triangular points, which represent the invested energy. This indicates that these cases produce more energy than they consume. However, the cases with
zinj > 0.10, which fall below the energy-invested triangular points, eventually consume more energy than they produce.
Fig. 7 shows the net energy balance for all injection mixtures except for pure cold-water injection (no added CO2, zinj = 0) in the heterogeneous reservoir. With no added CO2, the net energy recovery attains a
maximal value of 14552 TJ, because the process is not stopped at CO2 breakthrough and no energy is
required for CO2 capture, transport, and storage. As
the overall injected CO2 mole fraction increases, the
net energy balance decreases. This behavior continues until zinj = 0.02 at which value there is a bifurcation (i.e., a marking point between different types of solutions). For 0.02 < zinj <0.03, the net energy balance increases as the overall injected CO2
mole fraction increases. For higher injection concentrations (zinj > 0.03), again the solution behavior changes (i.e., the net energy balance decreases as the overall injected CO2 mole fraction
increases). The net energy balance reaches zero at zinj = 0.1. For overall injected CO2 mole fractions larger
than 0.1, the net energy balance continues to decrease and consequently remains negative. This case may still represent a viable option as the energetically most efficient way to store CO2.
From the energetic point of view, we note that even for cases with a negative energy balance, still the amount of CO2 storage is higher than the amount of
CO2 emitted during this process. Therefore, if we
look at this process (mixed CO2-water injection into
geothermal reservoirs) from a CO2-storage point of
emission is still negative, even though more energy is needed than the amount of energy produced.
Economic Analysis
In a first approximation, we can relate the energy balance to an economic analysis (Dincer and Rosen 2007; Granovskii et al. 2006). In such a conversion, the notion intensity of embodied energy plays a central role. It is more or less a constant for a particular activity (e.g., metal manufacturing). Indeed, for most materials, the intensity of embodied energy is 20 US$–1992/GJ. This represents the costs of construction materials and devices per consumed fossil fuel energy to produce them. The production cost of steel plus cement would be (38467.26 GJ + 658.46 GJ) × 20.1 US$/GJ = 786427 US$. Moreover, we assume that natural gas with an embodied energy intensity of 3.16 US$/GJ (Granovskii et al. 2006) is used to run the CO2 capture–compression unit.
Therefore, the cost of CO2 capture and compression
is 1.2 GJ/ton CO2–captured × 3.16 US$/GJ = 3.792
US$/ton CO2–captured.
In other cases, we can find the costs directly. For example, the cost of drilling and completion of wells can be estimated by (Klein et al. 2004); drilling cost (US$) = 240785 + 210 × depth (m) + 0.019069 × depth2 (m2). For a depth of 2 km, the drilling cost for the production and injection well would be 737061×2 US$. This would correspond to an intensity of embodied energy of 169 US$/GJ. Operational costs are determined by experience (e.g., the estimated operational cost is 2000 US$/well/month (Sedillos et al. 2010). 0 0.1 0.2 0.3 0.4 0.5 -3000 -2000 -1000 0 1000 2000 3000 Overall Injected CO 2 Mole Fraction N e t E n e rg y B a la n c e , T J
Fig. 7: Net energy balance versus the overall injected
CO2 mole fraction. The inserted plot
shows an expanded part of the plot with injected mole fractions between zero and 0.1.
Without the geothermal-energy option, natural gas can be used to produce heat energy. When we use heat energy produced from the geothermal reservoir,
an amount of natural gas that provides the same amount of heat energy as geothermal energy will be avoided. Consequently, we assume that the embodied energy intensity of geothermal energy is the same as natural gas (3.16 US$/GJ). Therefore, the revenue of the amount of produced geothermal energy is 3.16US$/GJ heat-production.
By this approximation, for overall injected mole fractions lower than 0.06, the net revenue is positive, indicating that the process is economically beneficial. However, for overall injected mole fractions larger than 0.06, the net revenue is negative, implying that the investment is larger than the income of the process. A detailed economic evaluation needs a thorough investigation, which is outside the scope of this paper. Our analysis shows that in principle such a storage option is economically acceptable.
CONCLUSIONS
This paper has shown that it is beneficial to integrate heat-energy recovery and carbon sequestration with cold mixed CO2-water injection into a geothermal
reservoir.
• In all cases considered, a compositional wave that runs ahead of the thermal wave, limits the period of simultaneous CO2 storage and heat extraction
to the end of the project. Therefore, CO2 injection
reduces the lifetime of a comparable conventional geothermal aquifer.
• Permeability and porosity heterogeneities in a geothermal aquifer significantly influence both heat extraction and CO2 storage. Hence, reservoir
characterization plays an important role in assessing the benefits of CO2 storage and energy
extraction.
• Increasing the amount of CO2 in the injection
mixture leads to bifurcation points at which the character of the solution in terms of energy production and CO2 storage changes. Hence,
increasing the amount of injected CO2 will not
always reduce the maximum amount of heat extraction and CO2 storage monotonically. • Heat transfer is more efficient in an aqueous
phase compared to a supercritical-CO2 phase. In
addition, heat-transfer heterogeneities are averaged out because of the large value of the thermal diffusion coefficient.
• The character of heterogeneity and the mobility ratio control the displacement regime. For the heterogeneous-reservoir structure considered here, a transition from a dispersive to channeling regime occurs as the mobility ratio increases from
M ≤ 1 to M > 1.
• The frequent occurrence of evaporation and condensation substantially delays CO2
breakthrough and consequently leads to a larger
0 0.02 0.04 0.06 0.08 0.1 -1000 0 1000 2000 3000 Overall Injected CO 2 Mole Fraction N e t E n e rg y B a la n c e , T J
amount of heat-energy production and CO2
storage.
Based on the simulations, it is possible to construct a plot of the recuperated heat energy versus the maximally stored CO2. The plot for the homogeneous
reservoir structure differs from the plot of the heterogeneous reservoir structure. Including heterogeneity, we have a choice of an optimal energy recovery of 5044 TJ with a limited storage capacity of 1429 ktonnes or a maximal CO2 storage capacity
of 2433 ktonnes with a minimal energy recovery of 812 TJ. Moreover, for overall injected CO2 mole
fractions smaller than 0.1, the net energy balance is positive, indicating that the process produces more energy than consumes. However, the net energy balance becomes negative for overall injected CO2
mole fractions larger than 0.1. The relation between useful energy and costs suggested in the literature makes it plausible that processes with a positive energy balance are close to economically viable. These data exclude heat gain from the surrounding layers because these are minor effects for thick layers. They are relevant in the case of pure cold-water injection. The most important purpose of the paper is to elucidate the prevailing mechanisms that influence heat-energy production and storage capacity.
ACKNOWLEDGEMENTS
The main part of this paper is reprinted from the International Journal of Greenhouse Gas Control, Vol 6, Hamidreza Salimi and Karl-Heinz Wolf, Integration of heat-energy recovery and carbon sequestration, 56–68, Copyright (2012), with permission from Elsevier. The work has been performed within the framework of the Dutch National CO2-storage program CATO-2,
work-package 3.5.
GENERAL INSTRUCTIONS
Please use the <styles> as specified in this template (e.g. <heading 1>, <normal> etc.). Paragraphs should be typed in <Normal> style, and justification is preferred. Do not put line spaces before or after headings.
The first line of the second column must start at the same location from the top as the first line of the first column. If a figure or table begins the second column it should also start at this location. When filling a page partially with text (i.e., the last page), make the two columns of approximately the same length (you can do this by inserting a Section Break – Continuous).
MAIN HEADINGS <HEADING 1> STYLE
Subheadings Like This <Heading 2> Style
Subheadings within sections follow the same format as section titles, but with upper and lower case. Subsubheadings like this <heading 3> style
Subsubheadings have no underline, and only the first letter capitalized.
FIGURES, TABLES, AND EQUATIONS
Standard terminology should be used in equations. If possible, keep equations to one column. All equations should be numbered.
Original drawings and photos may be included in your paper as figures. Diagrams may be in color. Figures and tables should immediately follow first mention of them in the text and can be placed in one column, or in the case of wider figures, can cover two columns (but only at the top or bottom of the page). The caption should start with either the word “figure” spelled out or abbreviated, followed by the figure number (i.e., Figure 1 or Fig. 1). The figure caption should be followed by a single line of space before the text resumes.
Figure 1: Illustration of placement of figure and caption. <FigureCaption style> Use <Figure> style for figure itself.
Although nobody understands why, table captions always precede the table (unlike figure captions that follow the figure). Use <TableCaption> style for the table captions, as in Table 1..
Table 1: Illustration of placement of table and caption. <TableCaption> style. The table should not “float”, make sure the table properties place it inline with text.
A 100
B 200
REFERENCES
Brown, D. (2000), “A Hot Dry Rock Geothermal Energy Concept Utilizing Supercritical CO2
Instead of Water,” In: Proceedings of the Twenty-Fifth Workshop on Geothermal Reservoir Engineering, Stanford University, pp. 233–238.
Bryant, S.L., Lakshminarasimhan, S. and Pope, G.A. (2008), “Buoyancy-Dominated Multiphase Flow and Its Effect on Geological Sequestration of CO2,” SPE Journal 13 (4), 447–454.
Chang, Y.-B., Coats, B.K. and Nolen, J.S. (1988), “A Compositional Model for CO2 Floods Including
CO2 Solubility in Water,” SPE Reservoir Evaluation & Engineering 1 (2), 155–160.
Costa, M.M., Schaeffer, R. and Worrell, E. (2001), “Exergy Accounting of Energy and Materials Flows in Steel Production Systems,” Energy 26, 363–384.
De Mooij, J.W.C. (2010), “Exergy Analysis of the Use of Geothermal Energy and Carbon Capture, Transportation and Storage in Underground Aquifers,” MS Thesis, Delft University of Technology, Delft, The Netherlands.
Delshad, M., Wheeler, M.F. and Kong, X. (2010), “A Critical Assessment of CO2 Injection Strategies
in Saline Aquifers,” In: Paper SPE 132442 Presented at the SPE Western Regional Meeting, Anaheim, California, 27–29 May.
Dincer, I. and Rosen, M.A. (2007), “Exergy, Energy, Environment and Sustainable Development,” Elsevier, London.
Drost, G.I.A. and Korenromp, M.H.A. (2009), “Sediment-Petrographical Analysis of the Delft Sandstone,” BSc. Thesis, Delft University of Technology, Delft, The Netherlands.
Farajzadeh, R., Ranganathan, P., Zitha, P.L.J. and Bruining, J. (2011) “The Effect of Heterogeneity on the Character of Density-Driven Natural Convection of CO2 Overlying a Brine Layer,” Advances in Water Resources 34, 327–339.
Frank, M.J.W., Kuipers, J.A.M. and van Swaaij, W.P.M. (1996), “Diffusion Coefficients and Viscosities of CO2 + H2O, CO2 + CH3OH, NH3 +
H2O, and NH3 + CH3OH Liquid Mixtures,” Journal of Chemical and Engineering Data 41,
297–302.
Freeman, B. and Rhudy, R. (2007), “Assessment of Post-Combustion Carbon Capture Technology
Developments,” Report No. 1012796. Electric Power Research Institute, Palo Alto, California. Gilding, D.T. (2010), “Heterogeneity Determination
of the Delft Subsurface for Heat Flow Modelling,” MS Thesis, Delft University of Technology, Delft, The Netherlands.
Granovskii, M., Dincer, I. and Rosen, M.A. (2006), “Environmental and Economic Aspects of Hydrogen Production and Utilization in Fuel Cell Vehicles,” Journal of Power Sources 157 (1), 411–421.
Groenenberg, R.M., Gilding, D.T., Donselaar, M.E., Drost, G.I.A., Korenromp, M.H.A., Loerakker, M., Wiggers, C.J.I., Wever, A.K.T. and Wolf, K.A.A. (2010), “Targeting for Geothermal Energy Production: Reservoir Characterization and Geothermal Potential of the Delft Sandstone,” In: Conference Paper Presented at the 72th EAGE Conference and Exhibition, Barcelona, Spain, pp. 3814–3818.
Klein, C.W., Lovekin, J.W. and Sanyal, S.K. (2004), “New Geothermal Site Identification and Qualification,” California Energy Commission, Sacramento.
Koval, E.J. (1963), “A Method for Predicting the Performance of Unstable Miscible Displacement in Heterogeneous Media. SPE Journal 3, 145– 154.
Loerakker, M. (2009), “A Core Study to Determine the Heterogeneities in the Delft Sandstone Member in the Moerkapelle Field,” BSc. Thesis, Delft University of Technology, Delft, The Netherlands.
Perry, R.H. and Green, D.W. (1997), “Perry’s Chemical Engineers’ Handbook,” seventh ed. McGraw-Hill.
Pruess, K. (2006), “Enhanced Geothermal Systems (EGS) Using CO2 as Working Fluid – A Novel
Approach for Generating Renewable Energy with Simultaneous Sequestration of Carbon,”
Geothermics 35, 351–367.
Pruess, K. (2008), “On Production Behavior of Enhanced Geothermal Systems with CO2 as
Working Fluid,” Energy Conversion and
Management 49, 1446–1454.
Salimi, H. and Wolf, K-H. (2012), “Integration of Heat-Energy Recovery and Carbon Sequestration,” International J. of Greenhouse
Gas Control 6 (3), 56–68.
Salimi, H., Groenenberg, R. and Wolf, K-H. (2011a), “Compositional Flow Simulations of Mixed CO2-Water Injection into Geothermal
Reservoirs: Geothermal Energy Combined with CO2 Storage,” In: Proceedings of the
Thirty-Sixth Workshop on Geothermal Reservoir Engineering Stanford University, Stanford, California, January 31-February 2.
Salimi, H., Wolf, K.-H. and Bruining, J. (2011b), “Negative Saturation Approach for Compositional Flow Simulations of Mixed CO2–
Water Injection into Geothermal Reservoirs Including Phase Transition and Disappearance,” In: Paper SPE 142924 Presented at the SPE EUROPEC/EAGE Annual Conference and Exhibition, Vienna, Austria, 23–26 May, doi:10.2118/142924-MS.
Salimi, H., Wolf, K.-H. and Bruining, J. (2011c), “Negative Saturation Approach for Compositional Flow Simulations of Mixed CO2–
Water Injection into Geothermal Reservoirs Including Phase Appearance and Disappearance,” SPE Journal,
doi:10.2118/142924-PA, SPE-142924-PA. Salimi, H., Wolf, K.-H. and Bruining, J. (2012),
“Negative Saturation Approach for Nonisothermal Compositional Two-Phase Flow Simulations,” Transport in Porous Media 91 (2), 391–422.
Samuel, G.R. and McColpin, G. (2001), “Welbore Hydraulic Optimization with Positive Displacement Motor and Bit,” In: Paper SPE 72320 Presented at the SPE/IADC Middle East Drilling Technology Conference, Bahrain, 22–24 October, doi:10.2118/72320-MS.
Sayegh, S.G. and Najman, J. (1987), “Phase Behavior Measurements of CO2-SO2-Brine Mixtures,” Canadian Journal of Chemical Engineering 65
(2), 314–320.
Sedillos, L., Roundtree, R., Wright, J. and Miskimins, J. (2010), “Unconventional Well Profitability vs Asset Profitability,” In: Paper SPE 135329 Presented at the SPE Annual Technical Conference and Exhibition, Florence, Italy, 19–22 September, doi:10.2118/135329-MS.
Smits, P. (2008), “Construction of an Integrated Reservoir Model Using the Moerkapelle Field for Geothermal Development of the Delft Sandstone,” MS Thesis, Delft University of Technology, Delft, The Netherlands.
Waggoner, J.R., Castillo, J.L. and Lake, L.W. (1992), “Simulation of EOR Processes in Stochastically Generated Permeable Media,” SPE Formation
Evaluation 7, 173–180.
Wolf, K.-H.A.A., Willemsen, A., Bakker, T.W., Wever, A.K.T. and Gilding, D.T. (2008), “The Development of a Multi-Purpose Geothermal Site in Urban Area,” In: Proceedings of the
SPE-70th EAGE Conference and Exhibition, Rome, Italy, vol. 2, pp. 1018–1102.
Xu, T. and Pruess, K. (2010), “Reactive Transport Modeling to Study Fluid-Rock Interactions in Enhanced Geothermal Systems (EGS) with CO2
as Working Fluid,” In: Conference Paper Presented at the World Geothermal Congress, Bali, Indonesia, 25–29 April.