- Email: [email protected]
Water condensation in carbon-dioxide-based engineered geothermal power generation
Geothermics 51 (2014) 397–405
Contents lists available at ScienceDirect
Geothermics journal homepage: www.elsevier.com/locate/geothermics
Water condensation in carbon-dioxide-based engineered geothermal power generation Aleks D. Atrens ∗ , Hal Gurgenci, Victor Rudolph Queensland Geothermal Energy Centre of Excellence, The University of Queensland, Brisbane, QLD 4067, Australia
a r t i c l e
i n f o
Article history: Received 13 October 2011 Accepted 9 March 2014 Available online 29 March 2014 Keywords: H2 O CO2 Carbon dioxide Water Phase diagram Supercritical Geothermal Thermosiphon EGS Engineered geothermal systems
a b s t r a c t Engineered geothermal systems (EGS) may utilise carbon dioxide as a heat extraction fluid instead of water. Nevertheless, water present in the geothermal reservoir will be extracted into the working fluid, affecting fluid flow behaviour and the required surface plant design for such a system. Dissolved water in a carbon dioxide-rich phase changes thermodynamic properties, and causes corrosion, and erosion where water droplets condense. The conditions for condensation of water in such a system have not been examined. We present condensation curves that predict conditions for water condensation, and bubble curves that predict carbonic acid concentration in a condensed H2 O-rich phase. These diagrams predict concentration thresholds for condensation in the production wellbore and surface equipment. Predicted concentration thresholds for condensation do not change significantly in response to change in water content. The probable minimum CO2 concentration allowable for direct use of carbon dioxide as a working fluid is 95% for the turbine. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Carbon dioxide has been examined previously as a heat extraction and power cycle fluid for geothermal applications (Brown, 2000; Pruess, 2006, 2008; Pruess and Azaroual, 2006; Gurgenci et al., 2008; Atrens et al., 2009a,b, 2010b). It has been identified as a suitable fluid due to its: • Low viscosity and density within the reservoir. • Low solubility of polar (ionic) solids. • Suitability for concurrent sequestration.
A schematic of the CO2 -based EGS concept is shown in Fig. 1. Since the CO2 -based EGS concept was first proposed site trials of engineered geothermal systems (EGS) have revealed that most are likely to contain some water (Gurgenci, 2009). The interaction between carbon dioxide and water consequently is an important concern if CO2 is to be used in EGS. Three aspects of carbon dioxide–water interactions are important:
∗ Corresponding author. Tel.: +61 439 662 384; fax: +61 7 33654799. E-mail address: [email protected] (A.D. Atrens). http://dx.doi.org/10.1016/j.geothermics.2014.03.008 0375-6505/© 2014 Elsevier Ltd. All rights reserved.
• Reservoir interactions: the interactions between carbon dioxide, water, and reservoir rocks including both reactions and twophase fluid flow phenomena. • Thermodynamic interactions: how water effects the thermodynamic properties of the CO2 -rich phase, and the consequent implications for power plant design. • Liquid water condensation: the conditions under which a liquid H2 O-rich phase condenses.
Small amounts of water dissolved in the carbon dioxide flowing from the production well will have minimal process effects due to a small impact on fluid properties, and will not cause corrosion (Seiersten, 2001). If sufficient water is present to form a condensed H2 O phase, the water droplets carried in a predominantly gas flow may cause erosion or corrosion. The corrosion potential of a condensed water phase is more significant than the solution pH would suggest, due to the buffering provided by dissolved CO2 and CO2 -related species (Linter and Burstein, 1999). This makes condensation of water a vital consideration for well casing and surface plant design for a CO2 -based EGS. Where water is present in the reservoir it is necessary to identify conditions under which condensation occurs, as they may impact material selection, process design, and determination of when power generation directly with CO2 may be possible (q.v. Atrens et al., 2011).
398
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
Fig. 1. A depiction of a two-well CO2 -based geothermal system. Points 1–5 are: (1) the injection wellhead, (2) injection well-reservoir interface, (3) reservoirproduction well interface, (4) production wellhead, and (5) turbine exhaust.
The aim of this paper is to determine the concentration threshold, or the required minimum carbon dioxide concentration to prevent condensation of liquid water in important parts of the power system. Furthermore, the impact of dissolved water on the thermodynamic properties of a carbon dioxide-rich fluid is also discussed. Finally, the carbon dioxide content in the condensed water, present as various carbonate species is determined. The purpose is to provide a basis for examining the feasibility of a carbon dioxide cycle for application in EGS reservoirs. 2. Methods 2.1. Dew and bubble lines The conditions under which water condenses from a carbon dioxide–water fluid mixture can be calculated from equilibrium data. To assess these conditions for a range of temperatures and pressures relevant for geothermal power plant designs, a diagram of iso-compositional dew lines has been constructed. These dew lines provide a method to assess the concentration threshold for dewing at any pressure and temperature to prevent condensation in any section of the well and plant system. A diagram of iso-compositional bubble lines has also been constructed to examine CO2 concentration in the resulting H2 Orich liquid phase (the condensate). The concentration of carbonic acid can be determined from the CO2 concentration in the H2 Orich phase. Materials selection and corrosion tolerance calculation utilise carbonic acid concentration information. 2.2. Diagram construction Use of equations of state to derive dew and bubble lines was considered. There are a range of equations of state dealing with the CO2 –H2 O system, and additionally in some cases examine the influence of salts. However these are limited compared to using dew point data directly (for example, (Spycher et al., 2003) and (Spycher and Pruess, 2005) only examine temperatures below 100 ◦ C, (Duan et al., 2008) does not address dew point curves, and comprehensive studies such as (Ji et al., 2007) have large deviations at higher pressures).
Data on CO2 –H2 O mutual solubility was obtained from (Wiebe and Gaddy, 1939, 1940, 1941; Wiebe, 1941; Takenouchi and Kennedy, 1964; King and Coan, 1971; Patel and Eubank, 1988; Fenghour et al., 1996; Bamberger et al., 2000; Han et al., 2009). Some of these sources discuss the CO2 –H2 O system in detail (particularly Takenouchi and Kennedy, 1964). None of these sources individually covered the full range of temperatures and pressures relevant to CO2 -based EGS, and equilibrium dew and bubble line information was typically at varying compositions. Data from these sources were transformed to give dew and bubble lines at constant composition. This was accomplished by linear interpolation, or, on occasion, linear extrapolation. Not all compositional data has been reported in mol%; where necessary to convert units, a Standard Temperature and Pressure of 0 ◦ C and 0.101325 MPa (abs) was used. The original data ranged in temperature from 25 ◦ C to 350 ◦ C, and pressures from 0.1 MPa to 150 MPa. The relevant range for usage of CO2 in EGS is from 0 ◦ C to 300 ◦ C and 5 MPa to 60 MPa. Some data outside those ranges were used to construct the isocomposition curves. The triple line for the liquid CO2 –liquid H2 O-hydrate equilibrium (Sloan and Koh, 2008) has been included on dew and bubble line diagrams to indicate the conditions under which a H2 O-rich liquid phase solidifies into a hydrate phase. 2.3. Iso-compositional dew line contours Constant composition dew lines are presented in Fig. 2. These lines represent concentration thresholds. That is, they indicate the pressure–temperature conditions at which a H2 O-rich liquid phase will first condense from a fluid of the specified overall composition. For any overall fluid composition, a dew line divides the pressure–temperature plane into a region where the compounds are fully miscible, and a region where two phases are present. To the right of the dew line, a single phase exists. To the left of the dew line, two phases coexist. The two phases are a CO2 -rich fluid phase and a H2 O-rich liquid phase. The H2 O-rich phase also contains carbonic acid in equilibrium with the dissolved carbon dioxide, and associated carbonate, bicarbonate, and hydrogen ions, resulting in acidity. The H2 O-rich liquid phase condenses if temperature or pressure conditions of a fluid move from the right side to the left side of the dew line. 2.4. Iso-compositional bubble line contours Constant composition bubble lines are presented in Fig. 3. These lines provide the maximum mole fraction of CO2 that can be dissolved in a condensed liquid water phase at a given temperature and pressure. If the conditions change so that the CO2 dissolution threshold is exceeded then CO2 effervesces from the solution. When both a CO2 -rich fluid and a condensed H2 O-rich phase are present, the CO2 -rich phase has a composition corresponding to its dew point, and the H2 O-rich phase has a composition corresponding to its bubble point. This allows determination of CO2 solubility in a condensed H2 O-rich liquid within the CO2 -based EGS. That compositional information can be used with temperaturedependant equilibrium constants to determine the carbonic acid content and pH of the liquid phase. 3. Results 3.1. Liquid water condensation in power system operation The dew line diagram can be used to determine concentration thresholds for the CO2 -based EGS power system. The thresholds can be found by overlaying the pressure–temperature range for a section of the process onto the dew line diagram. The
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
399
Fig. 2. Dew lines for liquid phase condensation at different H2 O concentrations in supercritical CO2 in the pressure–temperature range for CO2 -based geothermal systems. Each line defines the CO2 mole fraction conditions for dew formation. The hydrate line shows the conditions at which a H2 O-rich phase will transform to a carbon dioxide hydrate. Circles indicate actual data points. Squares indicate interpolated or extrapolated data points.
highest concentration contour not intersected by the section’s pressure–temperature conditions gives a guideline for the CO2 concentration threshold for that process section. This can be extended by overlaying the entire pressure–temperature path for the CO2 -based power plant onto the dew line diagram. This is shown in Fig. 4 for a typical power system with a reservoir temperature of 225 ◦ C, depth of 5000 m, and hydraulic impedance comparable to Soultz-sousForêt, and using other reference parameters given in Table 1. The changes in thermodynamic conditions are calculated through the process as described in previous works (Atrens et al., 2010b). Corresponding to Fig. 1, the model encompasses fluid flow from the injection wellhead (1), to the bottom of the production well (2), through the reservoir to the base of the production well (3), up the production well to the production wellhead (4), calculates expansion in the turbine to give turbine exhaust conditions (5), and then evaluates the cooling system to return to injection wellhead conditions (1). The sharp gradients present in the curve between points 2 and 3 are due to the large pressure drop in the reservoir near the wellbores because of a small cross-section for fluid flow.
Fig. 4 allows evaluation of concentration thresholds for the power system. For the reference case, concentration thresholds are approximately 89 mol% CO2 for the production wellbore, and 97 mol% CO2 for the turbine. The concentration threshold for the cooling system is higher than 99.5 mol% CO2 . Thus if the CO2 concentration in the reservoir is 95 mol% CO2 (for the reference case), Table 1 Reference parameters. Mass flow rate Injection pressurea Reservoir temperature Ambient temperature Wellbore diameter Reservoir impedance ÄH Reservoir pressure Wellbore roughness Depth Reservoir length Maximum reservoir width a
120 kg/s 11.89 MPa 225 ◦ C 25 ◦ C 0.23125 m 0.2 MPa s L−1 8.603 × 10−11 m3 49.05 MPa 4 × 10−4 m 5000 m 1000 m 500 m
Calculated based on mass flow rate and reservoir parameters.
400
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
Fig. 3. Bubble lines for H2 O-rich liquid phase containing CO2 . Each line defines the CO2 mole fraction conditions for bubble formation. Squares indicate interpolated or extrapolated data points. The hydrate line is as described for Fig. 2.
flow in the production well is a single dry fluid phase and a H2 O-rich phase condenses within the turbine at the point A.
The pressure–temperature range of individual pieces of process equipment or the temperature–pressure path of the entire system can also be superimposed onto bubble line contours, as shown in Fig. 5. This allows prediction of the CO2 concentration in the H2 O-rich liquid phase inside process equipment. Using the above example of 95 mol% CO2 in the reservoir, condensation in the turbine occurs at approximately 15 MPa, 120 ◦ C (point A in Fig. 4). At this temperature and pressure Fig. 5 indicates that the CO2 concentration in the H2 O-rich liquid phase is estimated at slightly greater than 1 mol% CO2 .
the mol-fraction-adjusted average of water and carbon dioxide viscosities (Vesovic et al., 1998; IAPWS, 2003). This equation of state was used to evaluate the single-phase CO2 -rich fluid properties, as the effect of two-phase flow is beyond the scope of this work. The multi-component model was used to evaluate the deviation in pressure and temperature conditions of CO2 –H2 O mixtures from that of pure CO2 . Fig. 6 shows the production wellbore pressure–temperature conditions for flows containing pure CO2 , 90 mol% CO2 , and 80 mol% CO2 . Fig. 6 shows that pressure and temperature conditions for CO2 containing water for the range of compositions under which no condensation occurs deviate by a relatively small amount from the curve for 100% CO2 . The curves show the fluid behaviour assuming no condensation. The 80 mol% CO2 fluid would start dewing at approximately 200 ◦ C and 40 MPa.
3.3. Validity of usage
3.4. Cycle variation
The influence of dissolved water on the temperature–pressure path of the power cycle was evaluated using a CO2 –H2 O multicomponent equation-of-state (Lemmon and Jacobsen, 1999; Paulus and Penoncello, 2006). Viscosities used were calculated as
The dew and bubble contour diagrams allow evaluation of proposed power plant designs for the likelihood of condensation of liquid water. The sensitivity of condensation to parameter changes can identify important design choices. Five parameters have been
3.2. Water phase composition
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
401
Fig. 4. Typical CO2 thermosiphon cycle relative to liquid phase dew lines; cycle points correspond to: (1) cooling system outlet and injection wellhead, (2) injection well to reservoir interface, (3) reservoir to production well interface, (4) production wellhead and turbine inlet, and (5) turbine outlet and cooling system inlet. See the text for reference to point A. Other aspects of the figure are as described in Fig. 2.
examined to assess the sensitivity of concentration thresholds to change in system parameters. Four parameters (reservoir temperature, reservoir pressure, reservoir permeability, and wellbore diameter) are system constraints and cannot be changed during power plant operation. The fifth parameter examined, mass flow rate, can be varied during operation of the power plant. The system is modelled in a manner such that mass flow rate and injection pressure are not independent. Their relationship is a function of reservoir parameters. Increasing mass flow requires an increase in injection pressure and reduces production pressure and temperature. For any given set of system constraints, there will be a mass flow rate that provides the optimum energy recovery, and a mass flow rate that provides the optimum economic return. When varying the four system constraints, mass flow rate is also varied, to keep turbine work output at a maximum. For discussion of optimising mass flow rate, see (Atrens et al., 2010a,b). Mass flow rate is varied instead of being held constant to reflect the way a plant would be designed for the different system constraints.
The sensitivity of concentration thresholds at the production wellhead and turbine exhaust to variation in these parameters is shown in Figs. 7 and 8. Increases in reservoir pressure, reservoir temperature, and well diameter reduce concentration thresholds throughout the system. This is because they lead to higher temperatures and pressures at the production wellhead and turbine exhaust. The effects are most pronounced from changes to reservoir temperature, where small changes result in large differences in condensation concentration thresholds. The effect is relatively minor for changes in well diameter. Concentration thresholds are effectively insensitive to reservoir permeability. The reason for these effects is discussed in Section 4. Higher mass flow rates increase CO2 concentration thresholds for condensation in the production well, but reduce them for the turbine. The change for the production well concentration threshold is due to a larger frictional pressure drop, and therefore larger Joule-Thomson cooling effect. The turbine concentration threshold change is due to an increase in injection pressure to force the larger mass flow rate into the reservoir. This defines the exhaust
402
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
Fig. 5. Typical CO2 thermosiphon cycle relative to bubble lines for H2 O-rich liquid phase. Cycle points 1–5 are identical to Fig. 4.
pressure of the turbine, which also affects the exhaust temperature; increases to both of these improve the miscibility of H2 O in CO2 . The effect of variation of the four process constraints was also examined across a range of values that seem reasonable for CO2 -based EGS, to give an estimate of possible variation between different reservoirs or sites. The effect of variation was examined for changes of reservoir temperature from 195 to 255 ◦ C, of reservoir pressure from 39 to 59 MPa, of permeability from 50% to 150% of the reference case, and of well diameter up to 0.4 m (reductions in well diameter were not considered). Concentration thresholds for the production wellhead were found to vary from 84 to 95 mol% CO2 within these ranges of parameters, and mainly between 86 and 93 mol% CO2 . Concentration thresholds for the turbine were found predominantly to be within a range of 94–99 mol% CO2 , with the most extreme temperatures extending the minimum and maximum to 93 and 99.2 mol% CO2 , respectively. The range of concentration thresholds resulting from different mass flow rates was also assessed. Mass flow rates were varied across the range of flow rates that produce a positive output from the turbine. Variation across this range was found to result in concentration thresholds from 87 to 91 mol% CO2 for the production
wellhead, and from 92 to 99 mol% CO2 for the turbine exhaust. The extreme values required a significant deviation from the optimum mass flow rate. 4. Discussion The underlying reasons for sensitivity of concentration thresholds to process parameters are not immediately apparent. While the response to reservoir temperature, reservoir pressure, and wellbore diameter is qualitatively similar, it is due to different underlying characteristics of the system. Increases in reservoir pressure change concentration thresholds primarily due to denser fluid in the production well and the requirement of a higher injection pressure. The former results in an increased static pressure drop in the reservoir, causing a higher production temperature. The higher reservoir pressure leads to an increase in production pressure despite the increased static pressure drop. A higher injection pressure raises the turbine exhaust pressure, leading to a higher exhaust temperature. These effects improve CO2 –H2 O miscibility. When reservoir pressure is decreased, these effects are inverted, and concentration thresholds increase.
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
403
Wellbore Flow Behaviour and Influence on Condensaon 70
60
Hydrate
0.70
50
3
0.80
0.90
0.95
0.98
Pressure, MPa
0.85
40
30
0.97
0.99
4
20
0. 99 5
0.
60
10
Pure CO2 90 mol% CO2 80 mol% CO2
0
50
100
150
200
250
300
o
Temperature, C Fig. 6. Comparison of temperature–pressure change in production wellbore for different fluid compositions: (a) 100% CO2 ; (b) 90% CO2 and 10% H2 O; and (c) 80% CO2 and 20% H2 O.
Fig. 7. Percentage change in process parameters required for a reduction in concentration threshold for condensation at the production wellhead from 88.5 to 87.5 mol% CO2 ; a small bar indicates high sensitivity to that variable.
Fig. 8. Percentage change in process parameters required for a reduction in concentration threshold for condensation at the turbine exhaust from 98 to 97 mol% CO2 .
404
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
Increases in reservoir temperature have competing negative and positive effects on concentration thresholds. On the one hand, higher reservoir temperature favours higher mass flow rates, and increases frictional losses due to the lower fluid density in the production well. On the other hand, it increases temperatures in the surface equipment and production wellbore. The temperature increase is the dominant effect, so increases in reservoir temperature reduce the thresholds for condensation in the turbine and production well. Higher reservoir temperatures will also increase miscibility of CO2 and H2 O in the reservoir, which has not been examined in this work. This may increase water content in produced fluid, and will likely modify how water content in the reservoir changes over the lifetime of the power plant. Well diameter has a minor effect on concentration thresholds. The reduction in thresholds for larger well diameter size is due to a reduction in frictional pressure drop. The benefits of this are small because as diameter is increased, a mass flow rate increase also becomes desirable, to maximise electricity generation. Condensation thresholds are barely affected by changes in permeability. This is because a permeability change favours a different mass flow rate, but after mass flow rate is optimised for electricity generation, the injection pressure and production pressure are almost identical to the reference case. Mass flow rate can be varied for an in-place power plant, so it provides the possibility of adjusting concentration thresholds after construction. The disadvantage to varying the mass flow rate is that this moves the system away from an optimised design point that maximises electricity generation. There is, however, an opportunity to vary mass flow rate for an existing plant if a small change will be critical in preventing condensation. A secondary disadvantage is that the capacity to increase flow likely requires additional capital expenditure during construction to ensure sufficient equipment size. There are a number of sources of error in this assessment of concentration thresholds that should be noted. The role of diagram construction, mixed-phase thermodynamics, and turbine isentropic efficiency are important to consider. The accuracy of the concentration thresholds calculated from the dew and bubble lines in this work is dependent on the quality of the data used to construct them, and the temperature and pressure range covered by that data. The experimental data used for dew and bubble line construction are reported with much higher precision than the possible precision from graphically determining concentration from the diagrams. Systematic errors in measurement are expected to be small due to the use of a large number of different data sources. The data used for construction are sparse in some regions, however. Regions with limited data are the low-temperature region for CO2 -rich fluids, and the lowtemperature/high-pressure regions for the H2 O-rich phase. This may lead to some inaccuracies in those regions of the diagrams. The diagrams could be improved with additional experimental data for those regions. Inaccuracy in those regions is not expected to result in errors in predicted concentration thresholds, however, as the regions do not cover typical operating conditions for the surface equipment. Some inaccuracy is inherent in the diagrams due to the use of linear interpolation in their construction. Linear interpolation was used for its clarity, and is only used for small deviations from experimental data. It is not expected to result in meaningful errors in the predicted concentration thresholds. The use of the thermodynamic properties of pure CO2 for prediction of concentration thresholds was discussed and noted to have insignificant differences compared to using the thermodynamic properties of CO2 –H2 O mixtures. The calculations for mixtures used in that comparison did include one potentially significant simplification, however. Viscosity of the CO2 –H2 O mixture was based on a compositional average of the individual viscosities of water and
carbon dioxide. Viscosities of mixtures do not always behave in this manner, and in some cases can be higher or lower than those of the pure fluids. To evaluate whether this simplification had an impact on the result, the comparison was repeated with calculated viscosity increased or decreased by an order of magnitude. This was found to have only a minor effect on production pressure and temperature (approximately ±0.1 MPa, ±0.1 ◦ C). The small magnitude is because wellbore flow is fully turbulent, and the viscosity change is not sufficient to change the flow regime from turbulent flow. The turbine exhaust temperature calculation assumed isentropic turbine behaviour. This assumption is not reflective of reality, but was used because isentropic efficiency is unknown, and an efficiency of 100% provides the most conservative result for the turbine concentration threshold. A reduction in isentropic efficiency will reduce the concentration threshold for the turbine. 5. Conclusions A method has been presented to calculate the minimum CO2 concentrations required in different sections of an EGS power plant to prevent condensation of water out of the CO2 phase. Condensation of water should be avoided because it would cause rapid corrosion of the system components such as the steel well casings or the surface plant equipment. The minimum concentration thresholds calculated with this method can be used to determine if condensation is likely for a specific design and reservoir. They may also be used to predict if parameter changes will prevent condensation, or if the only recourse is configuration changes such as incorporation of dehydration mechanisms or a binary power generation system. The method presented includes bubble line diagrams which can be used to predict the CO2 content in a condensed H2 O phase. This compositional information can be used for prediction of material requirements for surface equipment or wellbore casings. The sensitivity of concentration thresholds to reservoir temperature, reservoir pressure, permeability, wellbore diameter, and mass flow rate were examined, and it was found that: • Higher reservoir temperatures or pressures were found to have a favourable effect on concentration thresholds. • Increases in wellbore diameter were also found to have a favourable effect, but the magnitude was small. • Increases in permeability were found to have an unfavourable effect on concentration thresholds, but the system was relatively insensitive to this parameter. • Increases in mass flow rate were found to reduce production wellhead concentration thresholds, but to have the inverse effect on turbine concentration thresholds. • Decreases in each parameter led to the opposite effect as an increase, although response of concentration thresholds was reduced as they approached 100 mol% CO2 . Mass flow rate may be changed to control concentration thresholds in process equipment, but this will move the system away from the optimum conditions for electricity generation. The capacity for mass flow rate increases during operation would increase cost of power plant equipment. The likely range for concentration thresholds in CO2 -based EGS was assessed for a range of system parameter values, and it was found that: • Concentration thresholds for the production wellhead are likely to range from 86 to 93 mol% CO2 . • Concentration thresholds for the turbine are likely to range from 94 to 99 mol% CO2 .
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
• Concentration thresholds for the cooling equipment were found to be greater than 99.5 mol% CO2 . • Values beyond these ranges are possible for systems operating far from optimal conditions for electricity generation, or for systems with particularly unusual site/reservoir constraints. A condensed H2 O-rich liquid phase present in the surface equipment is likely to have a composition of CO2 in the range of 1–3 mol%. To prevent condensation of water entirely (i.e. not even in the cooling equipment), very high purities are necessary. A CO2 -based EGS should have cooling equipment designed to withstand condensation of liquid water containing of 2–3 mol% CO2 . Acknowledgements We acknowledge the support of the Queensland State Government whose funding of the Queensland Geothermal Energy Centre of Excellence made this work possible. We also thank Andrejs Atrens for useful discussion. References Atrens, A.D., Gurgenci, H., Rudolph, V., 2009a. CO2 thermosiphon for competitive power generation. Energy & Fuels 23, 553–557. Atrens, A.D., Gurgenci, H., Rudolph, V., 2009b. Exergy Analysis of a CO2 Thermosiphon. Thirty-Fourth Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 6. Atrens, A.D., Gurgenci, H., Rudolph, V., 2010a. Economic Analysis of a CO2 Thermosiphon. World Geothermal Congress, Bali, Indonesia, pp. 10. Atrens, A.D., Gurgenci, H., Rudolph, V., 2010b. Electricity generation using a carbondioxide thermosiphon. Geothermics 39, 161–169. Atrens, A.D., Gurgenci, H., Rudolph, V., 2011. Removal of Water for Carbon Dioxide-Based EGS Operation. Thirty-Sixth Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 8. Bamberger, A., Sieder, G., Maurer, G., 2000. High-pressure (vapor + liquid) equilibrium in binary mixtures of (carbon dioxide + water or acetic acid) at temperatures from 313 to 353 K. Journal of Supercritical Fluids 17, 97–110. Brown, D.W., 2000. A Hot Dry Rock Geothermal Energy Concept Utilizing Supercritical CO2 Instead of Water. Twenty-Fifth Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 6. Duan, Z., Hu, J., Li, D., Mao, S., 2008. Densities of the CO2 –H2 O and CO2 –H2 O–NaCl systems up to 647 K and 100 MPa. Energy & Fuels 22, 1666–1674. Fenghour, A., Wakeham, W.A., Watson, J.T.R., 1996. Densities of (water + carbon dioxide) in the temperature range 415–700 K and pressures up to 35 MPa. Journal of Chemical Thermodynamics 28, 433–446. Gurgenci, H., 2009. Electricity Generation Using a Supercritical CO2 Geothermal Siphon. Thirty-Fourth Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 10. Gurgenci, H., Rudolph, V., Saha, T., Lu, M., 2008. Challenges for Geothermal Energy Utilisation. Thirty-Third Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 7.
405
Han, M.J., Shin, H.Y., Min, B.-M., Han, K.-H., Cho, A., 2009. Measurement and correlation of high pressure phase behavior of carbon dioxide + water system. Journal of Industrial and Engineering Chemistry 15, 212–216. IAPWS, 2003. Revised Release on the IAPS Formulation 1985 for the Viscosity of Ordinary Water Substance. The International Association for the Properties of Water and Steam, Vejle. Ji, Y., Ji, X., Feng, X., Liu, C., Lu, L., X.L., 2007. Progress in the study on the phase equilibria of the CO2 –H2 O and CO2 –H2 O–NaCl systems. Chinese Journal of Chemical Engineering 15, 439–448. King, A.D.J., Coan, C.R., 1971. Solubility of water in compressed carbon dioxide, nitrous oxide, and ethane. Evidence for hydration of carbon dioxide and nitrous oxide in the gas phase. Journal of the American Chemical Society 93, 1857–1862. Lemmon, E.W., Jacobsen, R.T., 1999. A generalized model for the thermodynamic properties of mixtures. International Journal of Thermophysics 20, 825–835. Linter, B.R., Burstein, G.T., 1999. Reactions of pipeline steels in carbon dioxide solutions. Corrosion Science 41, 117–139. Patel, M.R., Eubank, P.T., 1988. Experimental densities and derived thermodynamic properties for carbon dioxide–water mixtures. Journal of Chemical and Engineering Data 33, 185–193. Paulus, M.E., Penoncello, S.G., 2006. Correlation for the carbon dioxide and water mixture based on the Lemmon–Jacobsen mixture model and the Peng–Robinson equation of state. International Journal of Thermophysics 27, 1373–1386. 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 behaviour of enhanced geothermal systems with CO2 as working fluid. Energy Conversion and Management 49, 1446–1454. Pruess, K., Azaroual, M., 2006. On the Feasibility of Using Supercritical CO2 as Heat Transmission Fluid in an Engineered Hot Dry Rock Geothermal System. Thirty-First Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 8. Seiersten, M., 2001. Material Selection for Separation, Transportation and Disposal of CO2 . Corrosion 2001. NACE International, Houston, TX, 10 pp. Sloan, E.D., Koh, C.A., 2008. Clathrate Hydrates of Natural Gases. CRC Press, Boca Raton, FL. Spycher, N., Pruess, K., 2005. CO2 –H2 O mixtures in the geological sequestration of CO2 . II: Partitioning in chloride brines at 12–100 ◦ C and up to 600 bar. Geochimica and Cosmochimica Acta 69, 3309–3320. Spycher, N., Pruess, K., Ennis-King, J., 2003. CO2 –H2 O mixtures in the geological sequestration of CO2 . I: Assessment and calculation of mutual solubilities from 12 to 100 ◦ C and up to 600 bar. Geochimica and Cosmochimica Acta 67, 3015–3031. Takenouchi, S., Kennedy, G.C., 1964. The binary system H2 O–CO2 at high temperatures and pressure. American Journal of Science 262. Vesovic, V., Fenghour, A., Wakeham, W.A., 1998. The viscosity of carbon dioxide. Journal of Physical and Chemical Reference Data 27, 31–44. Wiebe, R., 1941. The binary system carbon dioxide–water under pressure. Chemical Reviews 29, 475–481. Wiebe, R., Gaddy, V.L., 1939. The solubility in water of carbon dioxide at 50, 75 and 100◦ , at pressures to 700 atmospheres. Journal of the American Chemical Society 61, 315–318. Wiebe, R., Gaddy, V.L., 1940. The solubility of carbon dioxide in water at various temperatures from 12 to 40◦ and at pressures to 500 atmospheres. Critical phenomena. Journal of the American Chemical Society 62, 815–817. Wiebe, R., Gaddy, V.L., 1941. Vapor phase composition of carbon dioxide–water mixtures at various temperatures and at pressures to 700 atmospheres. Journal of the American Chemical Society 63, 475–477.
Contents lists available at ScienceDirect
Geothermics journal homepage: www.elsevier.com/locate/geothermics
Water condensation in carbon-dioxide-based engineered geothermal power generation Aleks D. Atrens ∗ , Hal Gurgenci, Victor Rudolph Queensland Geothermal Energy Centre of Excellence, The University of Queensland, Brisbane, QLD 4067, Australia
a r t i c l e
i n f o
Article history: Received 13 October 2011 Accepted 9 March 2014 Available online 29 March 2014 Keywords: H2 O CO2 Carbon dioxide Water Phase diagram Supercritical Geothermal Thermosiphon EGS Engineered geothermal systems
a b s t r a c t Engineered geothermal systems (EGS) may utilise carbon dioxide as a heat extraction fluid instead of water. Nevertheless, water present in the geothermal reservoir will be extracted into the working fluid, affecting fluid flow behaviour and the required surface plant design for such a system. Dissolved water in a carbon dioxide-rich phase changes thermodynamic properties, and causes corrosion, and erosion where water droplets condense. The conditions for condensation of water in such a system have not been examined. We present condensation curves that predict conditions for water condensation, and bubble curves that predict carbonic acid concentration in a condensed H2 O-rich phase. These diagrams predict concentration thresholds for condensation in the production wellbore and surface equipment. Predicted concentration thresholds for condensation do not change significantly in response to change in water content. The probable minimum CO2 concentration allowable for direct use of carbon dioxide as a working fluid is 95% for the turbine. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Carbon dioxide has been examined previously as a heat extraction and power cycle fluid for geothermal applications (Brown, 2000; Pruess, 2006, 2008; Pruess and Azaroual, 2006; Gurgenci et al., 2008; Atrens et al., 2009a,b, 2010b). It has been identified as a suitable fluid due to its: • Low viscosity and density within the reservoir. • Low solubility of polar (ionic) solids. • Suitability for concurrent sequestration.
A schematic of the CO2 -based EGS concept is shown in Fig. 1. Since the CO2 -based EGS concept was first proposed site trials of engineered geothermal systems (EGS) have revealed that most are likely to contain some water (Gurgenci, 2009). The interaction between carbon dioxide and water consequently is an important concern if CO2 is to be used in EGS. Three aspects of carbon dioxide–water interactions are important:
∗ Corresponding author. Tel.: +61 439 662 384; fax: +61 7 33654799. E-mail address: [email protected] (A.D. Atrens). http://dx.doi.org/10.1016/j.geothermics.2014.03.008 0375-6505/© 2014 Elsevier Ltd. All rights reserved.
• Reservoir interactions: the interactions between carbon dioxide, water, and reservoir rocks including both reactions and twophase fluid flow phenomena. • Thermodynamic interactions: how water effects the thermodynamic properties of the CO2 -rich phase, and the consequent implications for power plant design. • Liquid water condensation: the conditions under which a liquid H2 O-rich phase condenses.
Small amounts of water dissolved in the carbon dioxide flowing from the production well will have minimal process effects due to a small impact on fluid properties, and will not cause corrosion (Seiersten, 2001). If sufficient water is present to form a condensed H2 O phase, the water droplets carried in a predominantly gas flow may cause erosion or corrosion. The corrosion potential of a condensed water phase is more significant than the solution pH would suggest, due to the buffering provided by dissolved CO2 and CO2 -related species (Linter and Burstein, 1999). This makes condensation of water a vital consideration for well casing and surface plant design for a CO2 -based EGS. Where water is present in the reservoir it is necessary to identify conditions under which condensation occurs, as they may impact material selection, process design, and determination of when power generation directly with CO2 may be possible (q.v. Atrens et al., 2011).
398
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
Fig. 1. A depiction of a two-well CO2 -based geothermal system. Points 1–5 are: (1) the injection wellhead, (2) injection well-reservoir interface, (3) reservoirproduction well interface, (4) production wellhead, and (5) turbine exhaust.
The aim of this paper is to determine the concentration threshold, or the required minimum carbon dioxide concentration to prevent condensation of liquid water in important parts of the power system. Furthermore, the impact of dissolved water on the thermodynamic properties of a carbon dioxide-rich fluid is also discussed. Finally, the carbon dioxide content in the condensed water, present as various carbonate species is determined. The purpose is to provide a basis for examining the feasibility of a carbon dioxide cycle for application in EGS reservoirs. 2. Methods 2.1. Dew and bubble lines The conditions under which water condenses from a carbon dioxide–water fluid mixture can be calculated from equilibrium data. To assess these conditions for a range of temperatures and pressures relevant for geothermal power plant designs, a diagram of iso-compositional dew lines has been constructed. These dew lines provide a method to assess the concentration threshold for dewing at any pressure and temperature to prevent condensation in any section of the well and plant system. A diagram of iso-compositional bubble lines has also been constructed to examine CO2 concentration in the resulting H2 Orich liquid phase (the condensate). The concentration of carbonic acid can be determined from the CO2 concentration in the H2 Orich phase. Materials selection and corrosion tolerance calculation utilise carbonic acid concentration information. 2.2. Diagram construction Use of equations of state to derive dew and bubble lines was considered. There are a range of equations of state dealing with the CO2 –H2 O system, and additionally in some cases examine the influence of salts. However these are limited compared to using dew point data directly (for example, (Spycher et al., 2003) and (Spycher and Pruess, 2005) only examine temperatures below 100 ◦ C, (Duan et al., 2008) does not address dew point curves, and comprehensive studies such as (Ji et al., 2007) have large deviations at higher pressures).
Data on CO2 –H2 O mutual solubility was obtained from (Wiebe and Gaddy, 1939, 1940, 1941; Wiebe, 1941; Takenouchi and Kennedy, 1964; King and Coan, 1971; Patel and Eubank, 1988; Fenghour et al., 1996; Bamberger et al., 2000; Han et al., 2009). Some of these sources discuss the CO2 –H2 O system in detail (particularly Takenouchi and Kennedy, 1964). None of these sources individually covered the full range of temperatures and pressures relevant to CO2 -based EGS, and equilibrium dew and bubble line information was typically at varying compositions. Data from these sources were transformed to give dew and bubble lines at constant composition. This was accomplished by linear interpolation, or, on occasion, linear extrapolation. Not all compositional data has been reported in mol%; where necessary to convert units, a Standard Temperature and Pressure of 0 ◦ C and 0.101325 MPa (abs) was used. The original data ranged in temperature from 25 ◦ C to 350 ◦ C, and pressures from 0.1 MPa to 150 MPa. The relevant range for usage of CO2 in EGS is from 0 ◦ C to 300 ◦ C and 5 MPa to 60 MPa. Some data outside those ranges were used to construct the isocomposition curves. The triple line for the liquid CO2 –liquid H2 O-hydrate equilibrium (Sloan and Koh, 2008) has been included on dew and bubble line diagrams to indicate the conditions under which a H2 O-rich liquid phase solidifies into a hydrate phase. 2.3. Iso-compositional dew line contours Constant composition dew lines are presented in Fig. 2. These lines represent concentration thresholds. That is, they indicate the pressure–temperature conditions at which a H2 O-rich liquid phase will first condense from a fluid of the specified overall composition. For any overall fluid composition, a dew line divides the pressure–temperature plane into a region where the compounds are fully miscible, and a region where two phases are present. To the right of the dew line, a single phase exists. To the left of the dew line, two phases coexist. The two phases are a CO2 -rich fluid phase and a H2 O-rich liquid phase. The H2 O-rich phase also contains carbonic acid in equilibrium with the dissolved carbon dioxide, and associated carbonate, bicarbonate, and hydrogen ions, resulting in acidity. The H2 O-rich liquid phase condenses if temperature or pressure conditions of a fluid move from the right side to the left side of the dew line. 2.4. Iso-compositional bubble line contours Constant composition bubble lines are presented in Fig. 3. These lines provide the maximum mole fraction of CO2 that can be dissolved in a condensed liquid water phase at a given temperature and pressure. If the conditions change so that the CO2 dissolution threshold is exceeded then CO2 effervesces from the solution. When both a CO2 -rich fluid and a condensed H2 O-rich phase are present, the CO2 -rich phase has a composition corresponding to its dew point, and the H2 O-rich phase has a composition corresponding to its bubble point. This allows determination of CO2 solubility in a condensed H2 O-rich liquid within the CO2 -based EGS. That compositional information can be used with temperaturedependant equilibrium constants to determine the carbonic acid content and pH of the liquid phase. 3. Results 3.1. Liquid water condensation in power system operation The dew line diagram can be used to determine concentration thresholds for the CO2 -based EGS power system. The thresholds can be found by overlaying the pressure–temperature range for a section of the process onto the dew line diagram. The
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
399
Fig. 2. Dew lines for liquid phase condensation at different H2 O concentrations in supercritical CO2 in the pressure–temperature range for CO2 -based geothermal systems. Each line defines the CO2 mole fraction conditions for dew formation. The hydrate line shows the conditions at which a H2 O-rich phase will transform to a carbon dioxide hydrate. Circles indicate actual data points. Squares indicate interpolated or extrapolated data points.
highest concentration contour not intersected by the section’s pressure–temperature conditions gives a guideline for the CO2 concentration threshold for that process section. This can be extended by overlaying the entire pressure–temperature path for the CO2 -based power plant onto the dew line diagram. This is shown in Fig. 4 for a typical power system with a reservoir temperature of 225 ◦ C, depth of 5000 m, and hydraulic impedance comparable to Soultz-sousForêt, and using other reference parameters given in Table 1. The changes in thermodynamic conditions are calculated through the process as described in previous works (Atrens et al., 2010b). Corresponding to Fig. 1, the model encompasses fluid flow from the injection wellhead (1), to the bottom of the production well (2), through the reservoir to the base of the production well (3), up the production well to the production wellhead (4), calculates expansion in the turbine to give turbine exhaust conditions (5), and then evaluates the cooling system to return to injection wellhead conditions (1). The sharp gradients present in the curve between points 2 and 3 are due to the large pressure drop in the reservoir near the wellbores because of a small cross-section for fluid flow.
Fig. 4 allows evaluation of concentration thresholds for the power system. For the reference case, concentration thresholds are approximately 89 mol% CO2 for the production wellbore, and 97 mol% CO2 for the turbine. The concentration threshold for the cooling system is higher than 99.5 mol% CO2 . Thus if the CO2 concentration in the reservoir is 95 mol% CO2 (for the reference case), Table 1 Reference parameters. Mass flow rate Injection pressurea Reservoir temperature Ambient temperature Wellbore diameter Reservoir impedance ÄH Reservoir pressure Wellbore roughness Depth Reservoir length Maximum reservoir width a
120 kg/s 11.89 MPa 225 ◦ C 25 ◦ C 0.23125 m 0.2 MPa s L−1 8.603 × 10−11 m3 49.05 MPa 4 × 10−4 m 5000 m 1000 m 500 m
Calculated based on mass flow rate and reservoir parameters.
400
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
Fig. 3. Bubble lines for H2 O-rich liquid phase containing CO2 . Each line defines the CO2 mole fraction conditions for bubble formation. Squares indicate interpolated or extrapolated data points. The hydrate line is as described for Fig. 2.
flow in the production well is a single dry fluid phase and a H2 O-rich phase condenses within the turbine at the point A.
The pressure–temperature range of individual pieces of process equipment or the temperature–pressure path of the entire system can also be superimposed onto bubble line contours, as shown in Fig. 5. This allows prediction of the CO2 concentration in the H2 O-rich liquid phase inside process equipment. Using the above example of 95 mol% CO2 in the reservoir, condensation in the turbine occurs at approximately 15 MPa, 120 ◦ C (point A in Fig. 4). At this temperature and pressure Fig. 5 indicates that the CO2 concentration in the H2 O-rich liquid phase is estimated at slightly greater than 1 mol% CO2 .
the mol-fraction-adjusted average of water and carbon dioxide viscosities (Vesovic et al., 1998; IAPWS, 2003). This equation of state was used to evaluate the single-phase CO2 -rich fluid properties, as the effect of two-phase flow is beyond the scope of this work. The multi-component model was used to evaluate the deviation in pressure and temperature conditions of CO2 –H2 O mixtures from that of pure CO2 . Fig. 6 shows the production wellbore pressure–temperature conditions for flows containing pure CO2 , 90 mol% CO2 , and 80 mol% CO2 . Fig. 6 shows that pressure and temperature conditions for CO2 containing water for the range of compositions under which no condensation occurs deviate by a relatively small amount from the curve for 100% CO2 . The curves show the fluid behaviour assuming no condensation. The 80 mol% CO2 fluid would start dewing at approximately 200 ◦ C and 40 MPa.
3.3. Validity of usage
3.4. Cycle variation
The influence of dissolved water on the temperature–pressure path of the power cycle was evaluated using a CO2 –H2 O multicomponent equation-of-state (Lemmon and Jacobsen, 1999; Paulus and Penoncello, 2006). Viscosities used were calculated as
The dew and bubble contour diagrams allow evaluation of proposed power plant designs for the likelihood of condensation of liquid water. The sensitivity of condensation to parameter changes can identify important design choices. Five parameters have been
3.2. Water phase composition
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
401
Fig. 4. Typical CO2 thermosiphon cycle relative to liquid phase dew lines; cycle points correspond to: (1) cooling system outlet and injection wellhead, (2) injection well to reservoir interface, (3) reservoir to production well interface, (4) production wellhead and turbine inlet, and (5) turbine outlet and cooling system inlet. See the text for reference to point A. Other aspects of the figure are as described in Fig. 2.
examined to assess the sensitivity of concentration thresholds to change in system parameters. Four parameters (reservoir temperature, reservoir pressure, reservoir permeability, and wellbore diameter) are system constraints and cannot be changed during power plant operation. The fifth parameter examined, mass flow rate, can be varied during operation of the power plant. The system is modelled in a manner such that mass flow rate and injection pressure are not independent. Their relationship is a function of reservoir parameters. Increasing mass flow requires an increase in injection pressure and reduces production pressure and temperature. For any given set of system constraints, there will be a mass flow rate that provides the optimum energy recovery, and a mass flow rate that provides the optimum economic return. When varying the four system constraints, mass flow rate is also varied, to keep turbine work output at a maximum. For discussion of optimising mass flow rate, see (Atrens et al., 2010a,b). Mass flow rate is varied instead of being held constant to reflect the way a plant would be designed for the different system constraints.
The sensitivity of concentration thresholds at the production wellhead and turbine exhaust to variation in these parameters is shown in Figs. 7 and 8. Increases in reservoir pressure, reservoir temperature, and well diameter reduce concentration thresholds throughout the system. This is because they lead to higher temperatures and pressures at the production wellhead and turbine exhaust. The effects are most pronounced from changes to reservoir temperature, where small changes result in large differences in condensation concentration thresholds. The effect is relatively minor for changes in well diameter. Concentration thresholds are effectively insensitive to reservoir permeability. The reason for these effects is discussed in Section 4. Higher mass flow rates increase CO2 concentration thresholds for condensation in the production well, but reduce them for the turbine. The change for the production well concentration threshold is due to a larger frictional pressure drop, and therefore larger Joule-Thomson cooling effect. The turbine concentration threshold change is due to an increase in injection pressure to force the larger mass flow rate into the reservoir. This defines the exhaust
402
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
Fig. 5. Typical CO2 thermosiphon cycle relative to bubble lines for H2 O-rich liquid phase. Cycle points 1–5 are identical to Fig. 4.
pressure of the turbine, which also affects the exhaust temperature; increases to both of these improve the miscibility of H2 O in CO2 . The effect of variation of the four process constraints was also examined across a range of values that seem reasonable for CO2 -based EGS, to give an estimate of possible variation between different reservoirs or sites. The effect of variation was examined for changes of reservoir temperature from 195 to 255 ◦ C, of reservoir pressure from 39 to 59 MPa, of permeability from 50% to 150% of the reference case, and of well diameter up to 0.4 m (reductions in well diameter were not considered). Concentration thresholds for the production wellhead were found to vary from 84 to 95 mol% CO2 within these ranges of parameters, and mainly between 86 and 93 mol% CO2 . Concentration thresholds for the turbine were found predominantly to be within a range of 94–99 mol% CO2 , with the most extreme temperatures extending the minimum and maximum to 93 and 99.2 mol% CO2 , respectively. The range of concentration thresholds resulting from different mass flow rates was also assessed. Mass flow rates were varied across the range of flow rates that produce a positive output from the turbine. Variation across this range was found to result in concentration thresholds from 87 to 91 mol% CO2 for the production
wellhead, and from 92 to 99 mol% CO2 for the turbine exhaust. The extreme values required a significant deviation from the optimum mass flow rate. 4. Discussion The underlying reasons for sensitivity of concentration thresholds to process parameters are not immediately apparent. While the response to reservoir temperature, reservoir pressure, and wellbore diameter is qualitatively similar, it is due to different underlying characteristics of the system. Increases in reservoir pressure change concentration thresholds primarily due to denser fluid in the production well and the requirement of a higher injection pressure. The former results in an increased static pressure drop in the reservoir, causing a higher production temperature. The higher reservoir pressure leads to an increase in production pressure despite the increased static pressure drop. A higher injection pressure raises the turbine exhaust pressure, leading to a higher exhaust temperature. These effects improve CO2 –H2 O miscibility. When reservoir pressure is decreased, these effects are inverted, and concentration thresholds increase.
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
403
Wellbore Flow Behaviour and Influence on Condensaon 70
60
Hydrate
0.70
50
3
0.80
0.90
0.95
0.98
Pressure, MPa
0.85
40
30
0.97
0.99
4
20
0. 99 5
0.
60
10
Pure CO2 90 mol% CO2 80 mol% CO2
0
50
100
150
200
250
300
o
Temperature, C Fig. 6. Comparison of temperature–pressure change in production wellbore for different fluid compositions: (a) 100% CO2 ; (b) 90% CO2 and 10% H2 O; and (c) 80% CO2 and 20% H2 O.
Fig. 7. Percentage change in process parameters required for a reduction in concentration threshold for condensation at the production wellhead from 88.5 to 87.5 mol% CO2 ; a small bar indicates high sensitivity to that variable.
Fig. 8. Percentage change in process parameters required for a reduction in concentration threshold for condensation at the turbine exhaust from 98 to 97 mol% CO2 .
404
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
Increases in reservoir temperature have competing negative and positive effects on concentration thresholds. On the one hand, higher reservoir temperature favours higher mass flow rates, and increases frictional losses due to the lower fluid density in the production well. On the other hand, it increases temperatures in the surface equipment and production wellbore. The temperature increase is the dominant effect, so increases in reservoir temperature reduce the thresholds for condensation in the turbine and production well. Higher reservoir temperatures will also increase miscibility of CO2 and H2 O in the reservoir, which has not been examined in this work. This may increase water content in produced fluid, and will likely modify how water content in the reservoir changes over the lifetime of the power plant. Well diameter has a minor effect on concentration thresholds. The reduction in thresholds for larger well diameter size is due to a reduction in frictional pressure drop. The benefits of this are small because as diameter is increased, a mass flow rate increase also becomes desirable, to maximise electricity generation. Condensation thresholds are barely affected by changes in permeability. This is because a permeability change favours a different mass flow rate, but after mass flow rate is optimised for electricity generation, the injection pressure and production pressure are almost identical to the reference case. Mass flow rate can be varied for an in-place power plant, so it provides the possibility of adjusting concentration thresholds after construction. The disadvantage to varying the mass flow rate is that this moves the system away from an optimised design point that maximises electricity generation. There is, however, an opportunity to vary mass flow rate for an existing plant if a small change will be critical in preventing condensation. A secondary disadvantage is that the capacity to increase flow likely requires additional capital expenditure during construction to ensure sufficient equipment size. There are a number of sources of error in this assessment of concentration thresholds that should be noted. The role of diagram construction, mixed-phase thermodynamics, and turbine isentropic efficiency are important to consider. The accuracy of the concentration thresholds calculated from the dew and bubble lines in this work is dependent on the quality of the data used to construct them, and the temperature and pressure range covered by that data. The experimental data used for dew and bubble line construction are reported with much higher precision than the possible precision from graphically determining concentration from the diagrams. Systematic errors in measurement are expected to be small due to the use of a large number of different data sources. The data used for construction are sparse in some regions, however. Regions with limited data are the low-temperature region for CO2 -rich fluids, and the lowtemperature/high-pressure regions for the H2 O-rich phase. This may lead to some inaccuracies in those regions of the diagrams. The diagrams could be improved with additional experimental data for those regions. Inaccuracy in those regions is not expected to result in errors in predicted concentration thresholds, however, as the regions do not cover typical operating conditions for the surface equipment. Some inaccuracy is inherent in the diagrams due to the use of linear interpolation in their construction. Linear interpolation was used for its clarity, and is only used for small deviations from experimental data. It is not expected to result in meaningful errors in the predicted concentration thresholds. The use of the thermodynamic properties of pure CO2 for prediction of concentration thresholds was discussed and noted to have insignificant differences compared to using the thermodynamic properties of CO2 –H2 O mixtures. The calculations for mixtures used in that comparison did include one potentially significant simplification, however. Viscosity of the CO2 –H2 O mixture was based on a compositional average of the individual viscosities of water and
carbon dioxide. Viscosities of mixtures do not always behave in this manner, and in some cases can be higher or lower than those of the pure fluids. To evaluate whether this simplification had an impact on the result, the comparison was repeated with calculated viscosity increased or decreased by an order of magnitude. This was found to have only a minor effect on production pressure and temperature (approximately ±0.1 MPa, ±0.1 ◦ C). The small magnitude is because wellbore flow is fully turbulent, and the viscosity change is not sufficient to change the flow regime from turbulent flow. The turbine exhaust temperature calculation assumed isentropic turbine behaviour. This assumption is not reflective of reality, but was used because isentropic efficiency is unknown, and an efficiency of 100% provides the most conservative result for the turbine concentration threshold. A reduction in isentropic efficiency will reduce the concentration threshold for the turbine. 5. Conclusions A method has been presented to calculate the minimum CO2 concentrations required in different sections of an EGS power plant to prevent condensation of water out of the CO2 phase. Condensation of water should be avoided because it would cause rapid corrosion of the system components such as the steel well casings or the surface plant equipment. The minimum concentration thresholds calculated with this method can be used to determine if condensation is likely for a specific design and reservoir. They may also be used to predict if parameter changes will prevent condensation, or if the only recourse is configuration changes such as incorporation of dehydration mechanisms or a binary power generation system. The method presented includes bubble line diagrams which can be used to predict the CO2 content in a condensed H2 O phase. This compositional information can be used for prediction of material requirements for surface equipment or wellbore casings. The sensitivity of concentration thresholds to reservoir temperature, reservoir pressure, permeability, wellbore diameter, and mass flow rate were examined, and it was found that: • Higher reservoir temperatures or pressures were found to have a favourable effect on concentration thresholds. • Increases in wellbore diameter were also found to have a favourable effect, but the magnitude was small. • Increases in permeability were found to have an unfavourable effect on concentration thresholds, but the system was relatively insensitive to this parameter. • Increases in mass flow rate were found to reduce production wellhead concentration thresholds, but to have the inverse effect on turbine concentration thresholds. • Decreases in each parameter led to the opposite effect as an increase, although response of concentration thresholds was reduced as they approached 100 mol% CO2 . Mass flow rate may be changed to control concentration thresholds in process equipment, but this will move the system away from the optimum conditions for electricity generation. The capacity for mass flow rate increases during operation would increase cost of power plant equipment. The likely range for concentration thresholds in CO2 -based EGS was assessed for a range of system parameter values, and it was found that: • Concentration thresholds for the production wellhead are likely to range from 86 to 93 mol% CO2 . • Concentration thresholds for the turbine are likely to range from 94 to 99 mol% CO2 .
A.D. Atrens et al. / Geothermics 51 (2014) 397–405
• Concentration thresholds for the cooling equipment were found to be greater than 99.5 mol% CO2 . • Values beyond these ranges are possible for systems operating far from optimal conditions for electricity generation, or for systems with particularly unusual site/reservoir constraints. A condensed H2 O-rich liquid phase present in the surface equipment is likely to have a composition of CO2 in the range of 1–3 mol%. To prevent condensation of water entirely (i.e. not even in the cooling equipment), very high purities are necessary. A CO2 -based EGS should have cooling equipment designed to withstand condensation of liquid water containing of 2–3 mol% CO2 . Acknowledgements We acknowledge the support of the Queensland State Government whose funding of the Queensland Geothermal Energy Centre of Excellence made this work possible. We also thank Andrejs Atrens for useful discussion. References Atrens, A.D., Gurgenci, H., Rudolph, V., 2009a. CO2 thermosiphon for competitive power generation. Energy & Fuels 23, 553–557. Atrens, A.D., Gurgenci, H., Rudolph, V., 2009b. Exergy Analysis of a CO2 Thermosiphon. Thirty-Fourth Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 6. Atrens, A.D., Gurgenci, H., Rudolph, V., 2010a. Economic Analysis of a CO2 Thermosiphon. World Geothermal Congress, Bali, Indonesia, pp. 10. Atrens, A.D., Gurgenci, H., Rudolph, V., 2010b. Electricity generation using a carbondioxide thermosiphon. Geothermics 39, 161–169. Atrens, A.D., Gurgenci, H., Rudolph, V., 2011. Removal of Water for Carbon Dioxide-Based EGS Operation. Thirty-Sixth Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 8. Bamberger, A., Sieder, G., Maurer, G., 2000. High-pressure (vapor + liquid) equilibrium in binary mixtures of (carbon dioxide + water or acetic acid) at temperatures from 313 to 353 K. Journal of Supercritical Fluids 17, 97–110. Brown, D.W., 2000. A Hot Dry Rock Geothermal Energy Concept Utilizing Supercritical CO2 Instead of Water. Twenty-Fifth Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 6. Duan, Z., Hu, J., Li, D., Mao, S., 2008. Densities of the CO2 –H2 O and CO2 –H2 O–NaCl systems up to 647 K and 100 MPa. Energy & Fuels 22, 1666–1674. Fenghour, A., Wakeham, W.A., Watson, J.T.R., 1996. Densities of (water + carbon dioxide) in the temperature range 415–700 K and pressures up to 35 MPa. Journal of Chemical Thermodynamics 28, 433–446. Gurgenci, H., 2009. Electricity Generation Using a Supercritical CO2 Geothermal Siphon. Thirty-Fourth Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 10. Gurgenci, H., Rudolph, V., Saha, T., Lu, M., 2008. Challenges for Geothermal Energy Utilisation. Thirty-Third Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 7.
405
Han, M.J., Shin, H.Y., Min, B.-M., Han, K.-H., Cho, A., 2009. Measurement and correlation of high pressure phase behavior of carbon dioxide + water system. Journal of Industrial and Engineering Chemistry 15, 212–216. IAPWS, 2003. Revised Release on the IAPS Formulation 1985 for the Viscosity of Ordinary Water Substance. The International Association for the Properties of Water and Steam, Vejle. Ji, Y., Ji, X., Feng, X., Liu, C., Lu, L., X.L., 2007. Progress in the study on the phase equilibria of the CO2 –H2 O and CO2 –H2 O–NaCl systems. Chinese Journal of Chemical Engineering 15, 439–448. King, A.D.J., Coan, C.R., 1971. Solubility of water in compressed carbon dioxide, nitrous oxide, and ethane. Evidence for hydration of carbon dioxide and nitrous oxide in the gas phase. Journal of the American Chemical Society 93, 1857–1862. Lemmon, E.W., Jacobsen, R.T., 1999. A generalized model for the thermodynamic properties of mixtures. International Journal of Thermophysics 20, 825–835. Linter, B.R., Burstein, G.T., 1999. Reactions of pipeline steels in carbon dioxide solutions. Corrosion Science 41, 117–139. Patel, M.R., Eubank, P.T., 1988. Experimental densities and derived thermodynamic properties for carbon dioxide–water mixtures. Journal of Chemical and Engineering Data 33, 185–193. Paulus, M.E., Penoncello, S.G., 2006. Correlation for the carbon dioxide and water mixture based on the Lemmon–Jacobsen mixture model and the Peng–Robinson equation of state. International Journal of Thermophysics 27, 1373–1386. 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 behaviour of enhanced geothermal systems with CO2 as working fluid. Energy Conversion and Management 49, 1446–1454. Pruess, K., Azaroual, M., 2006. On the Feasibility of Using Supercritical CO2 as Heat Transmission Fluid in an Engineered Hot Dry Rock Geothermal System. Thirty-First Workshop on Geothermal Reservoir Engineering. Stanford University, Stanford, CA, pp. 8. Seiersten, M., 2001. Material Selection for Separation, Transportation and Disposal of CO2 . Corrosion 2001. NACE International, Houston, TX, 10 pp. Sloan, E.D., Koh, C.A., 2008. Clathrate Hydrates of Natural Gases. CRC Press, Boca Raton, FL. Spycher, N., Pruess, K., 2005. CO2 –H2 O mixtures in the geological sequestration of CO2 . II: Partitioning in chloride brines at 12–100 ◦ C and up to 600 bar. Geochimica and Cosmochimica Acta 69, 3309–3320. Spycher, N., Pruess, K., Ennis-King, J., 2003. CO2 –H2 O mixtures in the geological sequestration of CO2 . I: Assessment and calculation of mutual solubilities from 12 to 100 ◦ C and up to 600 bar. Geochimica and Cosmochimica Acta 67, 3015–3031. Takenouchi, S., Kennedy, G.C., 1964. The binary system H2 O–CO2 at high temperatures and pressure. American Journal of Science 262. Vesovic, V., Fenghour, A., Wakeham, W.A., 1998. The viscosity of carbon dioxide. Journal of Physical and Chemical Reference Data 27, 31–44. Wiebe, R., 1941. The binary system carbon dioxide–water under pressure. Chemical Reviews 29, 475–481. Wiebe, R., Gaddy, V.L., 1939. The solubility in water of carbon dioxide at 50, 75 and 100◦ , at pressures to 700 atmospheres. Journal of the American Chemical Society 61, 315–318. Wiebe, R., Gaddy, V.L., 1940. The solubility of carbon dioxide in water at various temperatures from 12 to 40◦ and at pressures to 500 atmospheres. Critical phenomena. Journal of the American Chemical Society 62, 815–817. Wiebe, R., Gaddy, V.L., 1941. Vapor phase composition of carbon dioxide–water mixtures at various temperatures and at pressures to 700 atmospheres. Journal of the American Chemical Society 63, 475–477.