Proceedings of SCCO2 Power Cycle Symposium 2009 RPI, Troy, NY, April 29-30, 2009

Supercritical CO2 for Application in Concentrating Solar Power Systems Craig S. Turchi NATIONAL RENEWABLE ENERGY LABORATORY 1617 Cole Blvd, Golden, CO 80401 Tel: 303-384-7565 Fax: 303-384-7495 [email protected] Abstract – Current Concentrating Solar Power (CSP) plants utilize oil or steam to transfer solar thermal energy to a steam Rankine Cycle power block. These heat transfer fluids (HTFs) have properties that limit plant performance, for example, the synthetic oil has an upper temperature limit of 400°C while direct steam generation requires complex control and has limited storage capacity. Accordingly, alternative fluids are under investigation by research teams worldwide. Supercritical carbon dioxide (s-CO2) has been identified as a potential HTF for CSP plants. Supercritical CO2 could be deployed as the HTF in the solar field or utilized as both HTF and power cycle fluid in a Brayton Cycle power block. If used simply as an HTF, s-CO2 provides advantages of higher operating temperatures, no threat of freezing, and single-phase operation. Applying s-CO2 in a Brayton Cycle offers additional potential for system efficiency and savings by eliminating heat exchangers and reducing the size of the power block equipment. Challenges include the high pressure necessary for s-CO2 operation, interfacing s-CO2 with thermal energy storage, and the lack of operating experience with s-CO2 Brayton Cycles. The presentation will summarize ongoing research by the Department of Energy, via the National Renewable Energy Laboratory (NREL), Sandia National Laboratories, and private contractors to investigate s-CO2 use in parabolic trough and solar power tower systems. Specific attention is given to the match between s-CO2 properties and typical solar thermal operating conditions. Opportunities for further investigation and collaborative research will be discussed.

I. INTRODUCTION Concentrating Solar Power utilizes solar thermal energy to drive a thermal power cycle for the generation of electricity. CSP technologies include parabolic trough, linear Fresnel, central receiver or “power tower,” and dish/engine systems. The parabolic trough is the most common system with nine Solar Electric Generating Stations (SEGS) operating in southern California for over two decades, and new plants online in Nevada and Spain. The resurgent interest in CSP has been driven by renewable portfolio standards in southwestern states and renewable energy feed-in tariffs in Spain. CSP is less expensive than solar photovoltaic systems for large, centralized power plants. Certain CSP systems, in particular parabolic troughs and power towers, are also amenable to the incorporation of thermal energy storage. Thermal energy storage is much less expensive than electric storage and allows CSP plants to increase capacity factor, cover cloud transients, and dispatch power as needed – for example, to cover an evening demand peak.

1

However, CSP costs are still higher than conventional fossilfired generation. While costs are expected to fall as manufacturing scale and deployment increase, further improvements in system efficiency are also necessary to improve the economics. Linear Fresnel and Power Towers are less mature and in addition to technical advances, operating experience is needed to prove system reliability and cost. For all three CSP technologies (Figure 1), improvements in the thermophysical properties of the HTF are one avenue for increased system efficiency. Dish/engine systems are modular CSP technologies that use a Stirling engine for power generation. These designs do not employ a circulating HTF and will not be discussed here.

Proceedings of SCCO2 Power Cycle Symposium 2009 RPI, Troy, NY, April 29-30, 2009

range and capable of tolerating higher temperatures are under investigation. II. SUPERCRITICAL CO2 AS AN HTF The current state-of-the-art HTF in parabolic trough systems is synthetic oil made of a biphenyl/diphenyl oxide eutectic (e.g., Therminol™ VP-1 or Dowtherm™ A). A qualitative comparison of this oil and leading candidates for higher temperature fluids is shown in Table 1.

Table 1. Qualitative comparison of solar power HTFs HTF Synthetic oil

Molten nitrate salt

Ambient pressure operation Temperature to ~600°C possible Industrial use Direct molten salt storage possible

Direct steam generation

Inexpensive Nontoxic No upper temperature limit Direct use in Rankine cycle power block possible Common industry use Inexpensive Nontoxic No upper temperature limit Direct use in Brayton Cycle power block possible Brayton cycle has smaller power block mass and less complexity vs. Rankine cycle Single phase throughout system Can use sensibleheat thermal storage (indirect molten salt)

Supercritical CO2

Figure 1. CSP technologies that utilize flowing heat transfer fluids include power towers (top), linear Fresnel (middle) and parabolic troughs (bottom). Current CSP plants utilize oil or steam to transfer solar energy to the power block. These fluids have properties that limit plant performance, for example, the synthetic oil has an upper temperature limit of 400°C while direct steam generation requires complex control and has limited storage capacity. Higher operating temperatures generally translate into higher thermal cycle efficiency and often allow for more efficient thermal storage. While molten nitrate salts have attractive thermal properties and have been used in solar applications, their high freezing point necessitates cumbersome freeze protection measures. Accordingly, HTFs with a wide fluid

2

Advantages Proven (current state-of-the-art) Moderate operating pressure (~15 bar)

Disadvantages Expensive Flammable Limited to <400°C max temperature H2 degradation product degrades plant efficiency Requires elaborate freeze protection Moderately expensive Corrosive (especially for seals) High pressures in field Phase-change adds complexity to system design and control Multi-hour storage options unproven High pressures in field Unproven (under development for nuclear power industry)

Proceedings of SCCO2 Power Cycle Symposium 2009 RPI, Troy, NY, April 29-30, 2009

Each of the candidate HTFs has advantages. Steam and s-CO2 are cheap and nontoxic, but necessitate high pressures in the solar collector. Molten salts have virtually no vapor pressure, but freeze at temperatures above 100°C and require cumbersome freeze protection methods. Both direct steam and s-CO2 offer the potential to eliminate expensive heat exchangers by feeding directly into the power block. s-CO2 has the added advantage of single-phase operation but at pressures even higher that that used for superheated steam. Specific physical properties of the different HTFs are shown in Table 2 for temperature conditions of interest for parabolic troughs. Linear Fresnel systems run at lower temperatures, but have fewer moving parts versus parabolic trough fields and so may have lesser issues with high pressure seals. Power Towers would likely run at higher temperatures – in the range of 550 to 700°C and could be the best match for s-CO2. Although s-CO2 has not been considered in solar plants until recently, significant research into s-CO2 power cycles is underway related to next generation nuclear power plants. Carbon dioxide is of interest in sodium-cooled fast reactors due to CO2’s lesser reactivity with sodium when compared to steam, while operating at temperatures much lower than helium-based cycles. [4]. Brayton-cycle systems using CO2 also have smaller and less complex power blocks versus Rankine cycles [5]. Fortuitously, the temperatures of these proposed CO2-Brayton cycles match well with solar applications and the CO2-Brayton cycle has a theoretical efficiency that exceeds the superheated steam-Rankine cycle at temperatures above about 450°C (see Figure 2). III. CLOSED-LOOP BRAYTON POWER CYCLE Brayton power cycles are well-known in the utility industry due to their use for gas turbines. In contrast to these open-loop Brayton cycles, s-CO2 power plants would rely on a closedloop Brayton cycle with recompression of the supercritical fluid near its critical point. Operation remains supercritical throughout the entire cycle and recompression near the critical point takes advantage of the fluid’s relative high density to minimize compressor power. Conditions shown in Table 2 for the s-CO2 Brayton cycle are taken from a study performed by Dostal at MIT [5] and utilized by Argonne National Lab in a recent report [4]. In general, higher temperatures provide better efficiency, but higher pressures provide little advantage. Dostal indicates that a recompression ratio of 2.6 is close to optimum in a simple recompression Brayton cycle [10]. Figure 3 is a modified version of the cycle diagrams shown in references [4,10], and shows how a solar field could be integrated with a s-CO2 Brayton cycle. Parabolic troughs are indicated, but the solar energy could also be provided by a power tower. A two-tank thermal energy storage system is depicted. This process heats a molten salt by counter-current heat exchangers to charge the storage and reverses flow to discharge the stored thermal

3

energy back to the s-CO2. Based on the reference case provided in [4], the turbine inlet conditions are approximately 20 MPa and 470°C with a net efficiency is 38.3%. This value is slightly lower than indicated by Dostal, but indicates the sCO2 Brayton cycle efficiency is comparable with superheated and supercritical steam in the temperature region of 450°C to 550°C. At higher temperatures the Brayton cycle has an efficiency advantage, see Figure 2.

Figure 2. Theoretical efficiencies of different thermal power cycles [5]. More complex Brayton cycles have been proposed, but the efficiency benefits do not appear to justify the additional mechanical complexity [5]. In fact, one advantage of the scheme shown in Figure 3 is the relative simplicity versus a superheated or supercritical steam Rankine cycle. IV. CO2 HTF IN CSP PLANTS – ONGOING RESEARCH In 2008, the US Department of Energy issued 15 contracts examining innovative thermal energy storage processes for CSP. Two of these awardees, the City University of New York (CUNY) and Abengoa Solar, are investigating CO2. The CUNY team is investigating the use subcritical CO2 as an HTF in trough or power tower systems, with emphasis on interfacing the CO2 with a packed-bed thermal energy storage system, referred to as a thermocline. Abengoa Solar is examining the use of s-CO2 in a power tower and parabolic trough systems coupled with heat exchangers to heat molten salt that is either stored in a two-tank storage system (like depicted in Figure 3, above) or used to charge a packed-bed thermocline. Both approaches avoid the need for high-pressure storage vessels (as in the CUNY case), but do require salt-to-CO2 heat exchangers. The two-tank system is less operationally complex, while the thermocline is potentially smaller and more cost effective. Both of these contractors are focusing on utilization of CO2 as an HTF, but

Proceedings of SCCO2 Power Cycle Symposium 2009 RPI, Troy, NY, April 29-30, 2009

without its use in a Brayton power cycle. Expanding the longterm applicability to include a Brayton cycle power block offers the potential to increase overall efficiency and reduce capital cost. Interfacing a Brayton cycle into a CSP plant will require the combined expertise of power cycle developers and solar power experts.

3.

Coastal Chemicals Hitec product data sheet

4.

Argonne National Laboratory, "Performance Improvement Options for the Supercritical Carbon Dioxide Brayton Cycle," ANL-GenIV-103, 2007.

5.

Dostal, V., PhD Thesis: “A Supercritical Carbon Dioxide Cycle for Next Generation Nuclear Reactors,” Massachusetts Institute of Technology, January 2004.

6.

Parfomak, P., and P. Fogler, “Carbon Dioxide (CO 2) Pipelines for Carbon Sequestration: Emerging Policy Issues,” Congressional Research Service Report to Congress, April 19, 2007.

7.

Towler, B., D. Agarwal, and S. Mokhatab, “Modeling Wyoming’s Carbon Dioxide Pipeline Network,” Energy Sources, Part A, Vol. 30, 259-270, 2008.

8.

For example, see http://www.swepco.com/news/hempstead/ultrasuper. asp

9.

Zarza, E., L. Valenzuela, J. Leon, H.D. Weyers, M. Eickhoff, M. Eck, K. Hennecke, “The DISS Projects: Direct Steam Generation in Parabolic Trough Systems. Operation and Maintenance Experience and Update on Project Status,” ASME Journal of Solar Energy Engineering, Vol. 124, May 2002.

V. CONCLUSIONS & FUTURE WORK Supercritical CO2 operated in a closed-loop recompression Brayton cycle offers the potential of equivalent or higher cycle efficiency versus supercritical or superheated steam cycles at temperatures relevant for CSP applications. A single-phase process using s-CO2 as both HTF and thermal cycle fluid would simplify the power block machinery and is compatible with existing sensible-heat thermal energy storage processes. The greatest uncertainties in the utilization of such a cycle are the high pressure required and lack of experience with closedloop Brayton cycles – although this is an area of active research for next-generation nuclear power plants. The Department of Energy, through NREL, Sandia National Laboratories, and subcontractors is exploring the potential of s-CO2 for use in parabolic trough and power tower applications. This assessment includes modeling s-CO2 heat transfer properties and power cycles, investigating how best to incorporate thermal energy storage, and estimating the cost of s-CO2 piping systems for parabolic trough fields and power tower receivers. REFERENCES 1.

NIST thermophysical properties website, http://webbook.nist.gov/chemistry/fluid/

2.

Solutia Therminol VP-1 product data sheet

4

10. Dostal, V., P. Hejzlar, and M.J. Driscoll, “The Supercritical Carbon Dioxide Power Cycle: Comparison to Other Advanced Power Cycles,” Nuclear Technology, vol. 154, 283-301, June 2006.

Proceedings of SCCO2 Power Cycle Symposium 2009 RPI, Troy, NY, April 29-30, 2009

Table 2. Physical properties of HTF candidates in the range of interest for parabolic troughs. Field Inlet

Field Outlet

HTF Properties measured at 393°C and field outlet P

Temp (C)

Press. (bar)

Temp (C)

Press. (bar)

Visc. (cp)

Density (kg/m3)

Cp (J/g-K)

Thermal Cond. (W/m-K)

Kinematic Viscosity (m2/s)

Energy Density (kJ/m3-K)

Synthetic Oil Ref [2]

293

2

393

10

0.15

704

2.60

0.077

2.1E-07

1830

Direct Steam Generation

126

85

400

65

0.024

23.4

2.65

0.062

1.0E-06

62

Molten Salt (Hitec™) Ref [3]

350

1

450

1

1.7

1790

1.56

0.33

9.5E-07

2790

s-CO2 Ref [1]

323*

200*

472*

200*

0.033

158

1.22

0.052

2.1E-07

193

HTF

* Temperature and pressure range from example in [4].

Precooler

Generator

Solar Field

Turbine

Recompression Compressor

Main Compressor

Cold Tank Storage HXC High Temp Recuperator

Low Temp Recuperator

Hot Tank

Figure 3. Simplified schematic of a solar thermal s-CO2 recompression Brayton cycle power plant based on reference [10]. Typical turbine inlet conditions are approximately 20 MPa and 470°C [4].

5