Contents lists available at ScienceDirect

Solar Energy

journal homepage: www.elsevier.com/locate/solener

Techno-economic assessment of technological improvements in thermalenergy storage of concentrated solar power

Riezqa Andikaa, Young Kimc,d,⁎, Seok Ho Yoonc,d, Dong Ho Kimc, Jun Seok Choic,Moonyong Leea,b,⁎

a School of Chemical Engineering, Yeungnam University, Gyeongsan 38541, Republic of Koreab Sherpa Space, Daejeon 34051, Republic of Koreac Department of Thermal Systems, Korea Institute of Machinery &Materials, Daejeon 34103, Republic of Koread Plant Systems and Machinery, University of Science and Technology, Daejeon 34113, Republic of Korea

A R T I C L E I N F O

Keywords:CSP plantsLCOEsCO2 cycleSimulationSteam cycleThermal energy storage

A B S T R A C T

The technological and economic impact of design changes in thermal energy storage of concentrated solar power(CSP) systems is assessed. It is shown that the system costs change with the types of storage tanks and also thatthe operation temperature is limited by the thermal properties of the thermal storage medium. In addition, thecost of energy can be substantially reduced by replacing the conventional power cycle with more advancedpower cycles, such as a supercritical carbon dioxide power cycle. Using two types of thermal storage tanks andtwo thermal storage media, cases are generated incorporating combinations of the design options. A sensitivityanalysis is used to investigate the impacts of each technological improvement. The results of this work willcontribute to predicting the impact of research and improving the economics of the CSP system.

1. Introduction

Climate change and depleted fossil fuel reserves are drawing in-creased attention to renewable energy sources. The sun is a majorsource of energy for Earth, with an average of approximately 342 wattsof solar energy for every square meter per year. In total, the sun delivers4.4 × 1016 watts of solar energy to the Earth (NASA, 2005). Con-centrated solar power (CSP) is one of the highly promising technicaloptions for supplying baseload power with solar energy. In CSP, sun-light is collected and intensified so that the temperature can be highenough to make it applicable to the conventional steam cycle and to themore efficient power cycles, such as supercritical steam or carbon di-oxide cycles.

However, the operation efficiency of the CSP process is considerablylower than that of fossil fuel-based power generation systems due to theintermittent supply of solar energy. This can be resolved by reservingthe excess heat in a thermal energy storage (TES) unit for later useduring nighttime or cloudy periods. In this way, a CSP plant with a TESunit may provide electricity with greater flexibility for planning a ba-lanced system (Fthenakis et al., 2009; Izquierdo et al., 2010; Viehbahnet al., 2011; Lilliestam et al., 2012; Trieb et al., 2012; Usaola, 2012;Pfenninger et al., 2014; Casati et al., 2015; Petrollese et al., 2017).

Nevertheless, introducing a TES unit also requires additional capitaland operational costs. Since the levelized cost of electricity (LCOE) of aCSP plant is still higher than that of a fossil fuel-based power plant(Black and Veatch, 2012), care must be taken when designing a CSPplant integrated with a TES.

The evaluation of the LCOE may facilitate the assessment of po-tential cost reduction through technology improvements. Hernández-Moro and Martínez-Duart (2013) presented analytical expressionsshowing the future evolution of the LCOE of CSP based on the learningcurve approach. Researchers at Sandia National Laboratories identifiedsectoral technology improvement opportunities and discussed theirpotential impacts on cost reduction (Kolb et al., 2011). Hübner et al.(2016) discussed the techno-economic optimization of heat transferstructures in large scale latent heat energy storage systems.Nithyanandam and Pitchumani (2014) discussed the cost of CSP plantwith integrated EPCM-TES (encapsulated phase change materials basedthermal energy storage) system, tank based HP-TES (latent thermalstorage embedded with heat pipes) and 2-tank sensible storage system.However, the effect of molten salts properties on TES, power cycleunits, and LCOE has not been reported yet.

This work evaluates the technological and economic potential ofemploying technological improvements in thermal storage and power

http://dx.doi.org/10.1016/j.solener.2017.08.064Received 27 June 2017; Received in revised form 1 August 2017; Accepted 16 August 2017

⁎ Corresponding authors at: Department of Thermal Systems, Korea Institute of Machinery &Materials, Daejeon 34103, Republic of Korea (Y. Kim). School of Chemical Engineering,Yeungnam University, Gyeongsan 38541, Republic of Korea (M. Lee).

E-mail addresses: ykim@kimm.re.kr (Y. Kim), mynlee@yu.ac.kr (M. Lee).

Solar Energy 157 (2017) 552–558

0038-092X/ © 2017 Elsevier Ltd. All rights reserved.

MARK

cycle units. Technological improvements in thermal storage media andstorage tank type are investigated for developing high efficiency CSPprocesses. Their contributions to LCOE reduction are assessed based onthe simulation of an integrated CSP process.

2. Methodologies

2.1. Process simulation

An integrated CSP process simulated in this study consists of acollector, a receiver, a thermal storage unit, and a power cycle unit asshown in Fig. 1. A list of assumptions is given for the design of theintegrated CSP process. The solar field data are assumed to be those ofthe 279 MW CSP plant in Daggett, California (NREL, 2010; Turchi andHeath, 2013), where the direct normal isolation (DNI) is 950 W/m2,tower height 203 m, and full load thermal storage is 10 h. To estimatethe pressure and heat loss, the pipeline is specified to be of schedule40 S stainless steel (SS), 12 in in diameter with a roughness of0.0675 mm. The heat transfer coefficient is assumed to be 0.05 W/m2K.

The process was simulated in Aspen Plus V9. The Peng-Robinsonequation of state with the Boston-Mathias alpha function (PR-BM) wasused as the fluid package for the molten salt system. The IAPWS-95 andREFPROP models are used to simulate the steam and supercriticalcarbon dioxide (sCO2) cycle properties, respectively, in the wide

temperature and pressure range of interest (Aspentech, 2013). Theturbine efficiency is assumed to be 80% in the steam cycle and 93% inthe sCO2 cycle (Neises and Turchi, 2014; Elliott Turbo, 2015).

2.2. Case study

A commercially viable CSP plant with a thermal energy storage unitis taken as the base case. In this system, the TES unit is operated withSolar Salt and consists of hot and cold storage tanks. A steam cycle, inwhich a steam-driven turbine is used to generate electricity, is used forpower generation as it is in most demonstration and commercial plants(IEA, 2010; Kolb et al., 2011; IRENA, 2013).

In the following case studies, the process design is improved withtechnological changes to lower the cost of energy from that of the basecase. First, a change of the storage media is considered to increase theoperation temperature of TES unit, and, therefore, increase the powercycle efficiency. The operation of the TES unit is limited to 550 °C withSolar Salt due to salt decomposition above this temperature. Recently, afew salt mixtures have been proposed to resolve this issue, and one ofthem is shown in Table 1. The operation of the TES unit operation athigher temperatures is expected to elevate the power generation effi-ciency of the CSP plant by introducing a high efficiency power cycle,such as supercritical carbon dioxide power cycle (Kelly, 2010; Kolbet al., 2011). Even while maintaining the steam power cycle, the

Fig. 1. Process flowsheet of CSP with TES.

Table 1Properties of molten salts for case study.

Name Composition (wt%) Melting temperature (°C) Decomposition temperature(°C)

Density (kg/m3) Specific heat capacity (kJ/kgK)

Price (US$/ton)

Solar Salta NaNO3–KNO3 (60–40) 220a 600a 1910–1720b 1.49–1.55b 600c

Salt Ac KCl–Na2CO3–K2CO3 (20–40–40) 427c 877c 2054d 1.6 (at 450 °C)c 640c

a Heller (2013), Li et al. (2014).b Heller (2013).c Kim et al. (2016).d Estimated value.

R. Andika et al. Solar Energy 157 (2017) 552–558

553

efficiency may be increased by operating it at a higher temperature.Second, the introduction of a thermocline tank is considered as

shown in Fig. 2 to reduce the installation cost by replacing two separatetanks for charging and discharging the heat (Pacheco et al., 2002; EPRI,2010; Cocco and Serra, 2015). The combined and individual impactsare evaluated based on the integrated process design. Ten cases aregenerated accordingly.

2.3. Operating conditions

The operation temperature of the salts are set to be 5 °C below theirdecomposition temperature and 70 °C above their melting temperature.According to this rule, the temperature differences between the highestand lowest operating temperatures are 305 °C and 375 °C for Solar Saltand Salt A, respectively. The operating temperature of steam in a powercycle is determined by assuming a 10 °C minimum temperature ap-proach in the heat exchanger. In the steam cycle, the pressure of thesteam is adjusted according to its temperature. A linear regressionequation is established, based on the data in Table 2, to determine theoperating pressure (y) based on the operating temperature (x) as fol-lows:

= −y x0.6893 237.24 (1)

Reflecting the current turbine capability and availability in the si-mulations, one additional case is considered for Salt A, 280 bar, and

610 °C of steam (Cziesla et al., 2009). For the sCO2 cycle, a 5 °Cminimum temperature approach in the heat exchanger is assumed, and200 bar is selected as the operating pressure (Neises and Turchi, 2014).The operating conditions selected for the case studies are provided inTable 3. Note that the two-tank arrangement and the thermocline tankare designed to provide the same operating conditions for the powercycle.

2.4. LCOE evaluation

The LCOE is calculated with the equation:

=+

=

=

LCOEC C

EΣ ( )Σ ( )

I tn

O M t

tn

t

1 & ,

1 (2)

where CI is the total capital (investment) cost (US$), CO&M,t is theoperation and maintenance cost in year t(US$), and Et is the electricitygenerated in year t (kWh). The plant life is assumed to be 30 years. Thecost values are taken from the published literature (Dostal et al., 2006;EPRI, 2010; NREL, 2010; Turchi and Heath, 2013). All components andtanks operated at 454 °C or above are assumed to be fabricated from347 SS, and those operated below 454 °C are assumed to be carbon

Fig. 2. Pipeline layout of CSP plant with a thermocline tank.

Table 2Operating condition of steam turbine (Siemens, 2011, 2013a,b).

Type Temperature (°C) Pressure (bar)

SST-040 400 40SST-050 500 101SST-300 520 120SST-110 and SST-120 530 130SST-400, SST-600, and SST-800 540 140SST-900 585 165

Table 3Operating conditions of the power cycles in the case studies.

Salt Thermal storage tank Power cycle Cycle operating condition

Solar salt Two-tanks Steam 166 bar, 585 °CsCO2 200 bar, 590 °C

Thermocline Steam 166 bar, 585 °CsCO2 200 bar, 590 °C

Salt A Two-tanks Steam 280 bar, 610 °C,Steam 357 bar, 862 °CsCO2 200 bar, 867 °C

Thermocline Steam 280 bar, 610 °CSteam 357 bar, 862 °CsCO2 200 bar, 867 °C

R. Andika et al. Solar Energy 157 (2017) 552–558

554

steel. Note that the total capital cost may increase when a CSP usingmolten salt consisting chloride and/or carbonate. However, the totalcapital cost constitutes a suitable performance measure to evaluate CSPprocesses in a preliminary design stage where an order-of-magnitudeestimate of 25–40% error level is normally acceptable (Biegler et al.,1997).

3. Results and discussion

3.1. Power generation efficiency

The results of the energy analysis are summarized in Table 4. Thetable shows that cycle efficiency and net power generation is higher inthe sCO2 cycle than in steam cycle operation, even though pressurizingthe supercritical carbon dioxide with a pump requires more power thanfor steam. Overall, Salt A with the sCO2 cycle has the highest electricityproduction.

Table 4 also presents the required pumping power, which varieswith the selection of the salt. In addition to the difference in fluidproperties, the required pumping power is also affected by theirthermal properties. The heat supplied from the heat source is the samefor all cases, so a basic thermodynamics equation, as follows, is used:

=Q m c TΔ · ·Δ (3)

where ΔQ is the heat flow to the CSP system, m is the mass of moltensalt, c is the specific heat capacity of the molten salt, and ΔT is thechange in temperature (operation temperature range). It can be de-duced from Eq. (3) that the required amount of salt can be reduced ifthe operation temperature range is wider, reducing pumping powerrequirement.

3.2. Economic evaluation

The results in Table 5 and Fig. 3 indicate that the LCOE can bereduced by 3.9–28% by using Salt A instead of Solar Salt. While thecosts are similar for both Solar Salt and Salt A, the LCOE is still lowerfor cases with Salt A due to the reduction in the pumping energy andhigher power generation efficiency. It should be noted that this com-parison is conducted only to examine the impact of thermal storagetemperature change. Replacing two tanks with one thermocline tankmay contribute to reducing the LCOE if the tank has a high thermalefficiency. The details will be given in the next section.

3.3. Sensitivity analysis

In a thermocline tank, the hot and cold molten salt are separated bya layer formed by differences in density while in a two-tank system, thehot and cold molten salts are kept in separate tanks. Regardless of thetank wall insulation, the hot and cold molten salt may possibly bemixed so that the hot temperature can be lowered, decreasing thetemperature of heat delivered to the power cycle accordingly (Biencintoet al., 2014; Cocco and Serra, 2015). The LCOEs for the cases with thethermocline tank of varied efficiencies are further calculated to studythe impact of efficiency change of the thermocline tank. The efficiencyof a thermocline storage tank is defined to be (Thot,t–TAT)/(Thot,0–TAT)for illustration. Thot,0 and Thot,t are the hot salt temperatures at time 0and t, and TAT is the ambient temperature that is set at 28.6 °C in thiswork. As can be discerned from Table 6 and Fig. 4, even with a 2% loss,the thermocline tank loses its economic advantage over the two-tanksystem. A thermocline tank should have efficiency higher than 98% toprovide the same performance as the two-tank system.

We also studied the sensitivity of the LCOE to the shift in operatingtemperature by assuming a virtual salt. The melting and decompositiontemperatures of the salt are set to be the values shifted by the sameextent from that of Solar Salt and Salt A, as shown in Table 7. Althoughthe shifted temperatures are not the properties of an actual salt, we canassess the impact of operation temperature change on the LCOE fromthis study. Other operation conditions were determined by the sameassumptions as in Section 2.3 Table 8.

From Table 7 and Fig. 5, it can be inferred that there is a relation-ship between operating temperature shifts and the LCOE. The tem-perature shift affects the total capital cost (tank size), annual operationand maintenance cost, and annual electricity generation. Therefore, thecost change of each shift in temperature is calculated in the form of apercentage that is then divided by the temperature shift. As a result, therelationship between the variables is formulated as:

=+ + +

+

=

=

LCOEC C s x C C s y

E E s z( ( · · )) Σ ( ( · · ))

Σ ( ( · · ))sI I t

nO M t O M t

tn

t t( )

1 & , & ,

1 (4)

Table 4Energy analysis results of the simulated processes.

Salt Powercycle(operatingpressure)

Power cycle results CSP plant results

Turbineoutput(MW)

Pumppower(MW)

Cycle netefficiency(%)

Pumppower(MW)

Netelectricitygeneration(MW)

Solar salt Steam(166 bar)

99.52 1.39 35.16 5.86 92.27

sCO2 140.28 12.54 45.76 5.86 121.88

Salt A Steam(280 bar)

103.81 2.40 36.33 5.03 96.37

Steam(357 bar)

114.33 2.54 40.05 5.03 106.76

sCO2 153.90 8.09 52.24 5.03 140.77

Table 5Economic analysis result of case studies.

Parameter Solar salt Salt A

Two-tanks Thermocline Two-tanks Thermocline

Steam sCO2 Steam sCO2 Steam280 bar

Steam356.9 bar

sCO2 Steam280 bar

Steam356.9 bar

sCO2

Total direct cost (million US$) 589.63 625.81 583.29 619.47 589.25 603.41 635.46 583.90 598.05 630.10Total indirect cost (million US$) 107.38 112.78 106.44 111.83 107.33 109.44 114.21 106.53 108.64 113.41Total installed cost (million US$) 697.01 738.58 689.73 731.30 696.58 712.84 749.67 690.43 706.69 743.51Operation and maintenance costs (million US

$/year)9.09 10.10 9.08 10.09 9.21 9.52 10.47 9.20 9.51 10.46

Daily generated electricity (MWh) 1938 2559 1938 2559 2024 2242 2956 2024 2242 2956Annual generated electricity (MWh) 484,442 639,874 484,441 639,875 505,965 560,486 739,062 505,965 560,486 739,062LCOE (US cents/kWh) 6.67 5.43 6.62 5.39 6.41 5.94 4.80 6.37 5.90 4.77Cost reduction (%) Basis 18.6 0.7 19.2 3.9 10.9 28.0 4.5 11.5 28.5

R. Andika et al. Solar Energy 157 (2017) 552–558

555

where x, y, and z are the constants for total installed cost (°C−1), annualoperation and maintenance cost (°C−1), and annual generated power(°C−1). The variable s is the temperature shift (°C). The values for x, y,and z for Solar Salt are 1.423 × 10−4, 2.064 × 10−4, and5.993 × 10−4 °C−1, respectively. For Salt A, the x, y, and z values are1.009 × 10−4, 1.451 × 10−4, and 3.924 × 10−4 °C−1, respectively.Fig. 5 shows that the LCOE is lowered by increasing the operatingtemperature even when the range of operating temperature is fixed. As

long as we can use the same material for the thermal storage tank, theeffect of the temperature shift on power generation efficiency is large sothat the LCOE can be reduced although the installed costs increase withthe positive temperature shift.

One might suggest that the application of a cheaper material mayresult in a larger advantage over the operation at high temperature.Another sensitivity analysis was conducted in this regard to determinethe effect of a change in the materials in correspondence to the oper-ating temperature change as shown in Table 7. It was hypothesized thatthere is a point where the LCOE becomes lower with the choice of thecheaper material. The hot tank temperature was reduced to theminimum limit to use 347 SS (∼455 °C) or to use carbon steel(∼450 °C). It was estimated that the price of carbon steel is 35.7% of347 SS (Oshwin Overseas, 2017; Reliable Pipes and Tubes, 2017).

Figs. 6 and 7 showed that even though we changed the material forthe hot tank, the LCOE for the case with the carbon steel hot tank is stillhigher than with the 347 SS. This is because as the operating tem-perature range becomes smaller, a larger volume of molten salt is re-quired as shown in Section 3.1. Therefore, the size of the tank should bebigger for the lower temperature operation. The required tank size andthe amount of salt was reduced to 45.9% by increasing the operatingtemperature range from 455 °C to 595 °C (ΔT = 140 °C). In other

590 626 589 603 635

107113

107 109114

273

303

276286

314

6.67

5.43

6.41

5.94

4.80

0

1

2

3

4

5

6

7

0

200

400

600

800

1000

1200

1400

1600

Steam(166 bar)

sCO2 Steam(280 bar)

Steam(356 bar)

sCO2

Solar Salt Salt A

LCO

E (U

S ce

nts/

kWh)

Cost

s (m

illio

ns U

S$)

Total Direct Cost Total Indirect Cost O era on and Maintenance Costs

Fig. 3. Results of economic evaluation for the cases witha two-tank TES unit.

Table 6Comparison of LCOEs of the thermocline tank system scenarios.

Efficiency ofthermocline tank(%)

Solar salt Salt A

Hot salttemperature (°C)

LCOE (UScents/kWh)

Hot salttemperature (°C)

LCOE (UScents/kWh)

100% 595 6.62 872 5.9099% 589.34 6.66 863.57 5.9498% 583.67 6.71 855.13 5.9897% 578.01 6.75 846.70 6.0296% 572.34 6.80 838.26 6.0695% 566.68 6.85 829.83 6.10

6.67

6.62

6.66

6.71

6.50

6.55

6.60

6.65

6.70

6.75

Two Tanks Thermocline (100%) Thermocline (99%) Thermocline (98%)

LCO

E (U

S ce

nts/

kWh)

Fig. 4. Comparison of LCOE for cases with the thermocline tanks.

R. Andika et al. Solar Energy 157 (2017) 552–558

556

words, for every 5 °C range in operation temperature, there can be a1.64% reduction in tank size and the amount of salt.

4. Conclusions

This work demonstrates the potential for available improvements inthermal energy storage for concentrated solar power (CSP) systems. TheLCOE can be significantly reduced by changing the salt. When thedifference between the hot and cold operating temperature becomeslarger, a smaller amount of salt is required, and the energy required forpumping the salt is reduced. The CSP system with Salt A requires 28%less molten salt than the system with Solar Salt. When the salt cansupport the operation at the higher temperature, the power generationefficiency can be elevated. With a large enough temperature increase,the efficiency of the power cycle can be substantially improved by re-placing the steam cycle with a supercritical carbon dioxide cycle. Athermocline tank should have high efficiency for its introduction to bebeneficial in reducing the LCOE. Moreover, it has been found that due

Table 7Sensitivity analysis on temperature change.

Name Melting temp.(°C)

Decomposition temp. (°C) Hot operating temp.(°C)

Cold operatingtemp. (°C)

Steam temperature(°C)

Steam pressure(bar)

LCOE change(%)

Solar salt 220 600 595 290 585 166 0Solar salt with +5 °C

shift225 605 600 295 590 169 −0.22

Solar salt with +10 °Cshift

230 610 605 300 595 173 −0.44

Solar salt with −5 °Cshift

215 595 590 285 580 163 0.22

Solar salt with −10 °Cshift

210 590 585 280 575 159 0.44

Salt A 427 877 872 497 862 356.9 0Salt A with +5 °C shift 432 882 877 502 867 360 −0.13Salt A with +10 °C

shift437 887 882 507 872 364 −0.26

Salt A with−5 °C shift 422 872 867 492 857 353 0.15Salt A with −10 °C

shift417 867 862 487 852 350 0.29

Table 8Sensitivity analysis for material change.

Name Hot operatingtemp. (°C)

Cold operatingtemp. (°C)

Hot tankmaterial

Cold tankmaterial

Solar salt(BaseCase)

595 290 347 SS Carbon steel

Solar salt (SS-CS)

455 290 347 SS Carbon steel

Solar salt (CS-CS)

450 290 Carbon steel Carbon steel

712 713 713 713 714

285 285 286 286 286

5.95 5.95 5.94 5.93 5.92

4.5

5

5.5

6

695

745

795

845

895

945

995

-10°C Shi -5°C Shi No Shi +5°C Shi +10°C Shi

Salt A

LCO

E (c

ents

/kW

h)

Inst

alle

d/O

&M

Cos

ts (m

illio

n $)

Total Installed Cost O era on and Maintenance Costs

Fig. 5. Sensitivity results of temperature change related toSalt A.

R. Andika et al. Solar Energy 157 (2017) 552–558

557

to the aforementioned factors, employing a cheaper material for theTES does not necessarily ensure a lower LCOE.

Acknowledgments

This research was supported by the research program NK203D ofKorea Institute of Machinery and Materials funded by NationalResearch Council of Science and Technology. This research was alsosupported by Basic Science Research Program through the NationalResearch Foundation of Korea (NRF) funded by the Ministry ofEducation (2015R1D1A3A01015621), and by the Priority ResearchCenters Program through NRF, funded by the Ministry of Education(2014R1A6A1031189).

References

Aspentech, 2013. Aspen Physical Property System: Physical Property Methods.Aspentech, Burlington.

Biencinto, M., Bayón, R., Rojas, E., González, L., 2014. Simulation and assessment ofoperation strategies for solar thermal power plants with a thermocline storage tank.Sol. Energy 103, 456–472.

Black & Veatch, 2012. Cost and Performance Data for Power Generation Technologies.Black & Veatch, Overland Park.

Biegler, L.T., Grossmann, I.E., Westerberg, A.W., 1997. Systematic Methods of Chemical

Process Design. Prentice Hall, New Jersey.Casati, E., Casella, F., Colonna, P., 2015. Design of CSP plants with optimally operated

thermal storage. Sol. Energy 116, 371–387.Cocco, D., Serra, F., 2015. Performance comparison of two-tank direct and thermocline

thermal energy storage systems for 1 MWe class concentrating solar power plants.Energy 81, 526–536.

Cziesla, F., Bewerunge, J., Senzel, A., 2009. Lünen – State-of-the Art Ultra SupercriticalSteam Power Plant Under Construction (presented in POWER-GEN Europe 2009,Cologne, Germany; May 26–29, 2009).

Dostal, V., Hejzlar, P., Driscoll, M.J., 2006. The supercritical carbon dioxide power cycle:comparison to other advanced power cycles. Nucl. Technol. 154, 283–301.

Electric Power Research Institute (EPRI), 2010. Solar Thermocline Storage Systems:Preliminary Design Study. EPRI, Palo Alto.

Turbo, Elliott., 2015. Steam Turbine Calculator. <http://www.elliott-turbo.com/TurbineCalculator> (accessed 16.11.30).

Fthenakis, V., Mason, J., Zweibel, K., 2009. The technical, geographical, and economicfeasibility for solar energy to supply the energy needs of the US. Energy Pol. 37,387–399.

Heller, L., 2013. Literature Review on Heat Transfer Fluids and Thermal Energy StorageSystems in CSP Plants. Solar Thermal Energy Research Group. StellenboschUniversity, Stellenbosch.

Hernández-Moro, J., Martínez-Duart, J.M., 2013. Analytical model for solar PV and CSPelectricity costs: Present LCOE values and their future evolution. Renew. Sust. EnergyRev. 20, 119–132.

Hübner, S., Eck, M., Stiller, C., Seitz, M., 2016. Techno-economic heat transfer optimi-zation of large scale latent heat energy storage systems in solar thermal power plants.Appl. Therm. Eng. 98, 483–491.

International Energy Agency (IEA), 2010. Technology Roadmap: Concentrating SolarPower. IEA, Paris.

International Renewable Energy Agency (IRENA), 2013. Concentrating Solar Power:Technology Brief. IRENA, Abu Dhabi.

Izquierdo, S., Montañés, C., Dopazo, C., Fueyo, N., 2010. Analysis of CSP plants for thedefinition of energy policies: the influence on electricity cost of solar multiples, ca-pacity factors and energy storage. Energy Pol. 38, 6215–6221.

Kelly, B., 2010. Advanced Thermal Storage for Central Receivers with SupercriticalCoolants. Abengoa Solar, Lakewood.

Kim, Y., Yoon, S.H., Choi, J.S., Noh, Y., Kim, D.H., Song, C.H., 2016. Selection of a moltensalt mixture for thermal storage in high temperature concentrated solar power sys-tems. High Temp.-High Press. 45, 461–469.

Kolb, G.J., Ho, C.K., Mancini, T.R., Gary, J.A., 2011. Power Tower Technology Roadmapand Cost Reduction Plan. Sandia National Laboratories, Albuquerque and Livermore.

Li, C.-J., Li, P., Wang, K., Molina, E.E., 2014. Survey of properties of key single andmixture halide salts for potential application as high temperature heat transfer fluidsfor concentrated solar thermal power systems. AIMS Energy 2, 133–157.

Lilliestam, J., Bielicki, J., Patt, A., 2012. Comparing carbon capture and storage (CCS)with concentrating solar power (CSP): potentials, costs, risks, and barriers. EnergyPol. 47, 447–455.

National Aeronautics and Space Administration (NASA), 2005. NASA Facts: The Balanceof Power in the Earth-Sun System. <https://www.nasa.gov/pdf/135642main_balance_trifold21.pdf> (accessed 16.11.30).

National Renewable Energy Laboratory (NREL), 2010. System Advisor Model 2016.3.14.< https://sam.nrel.gov/> (accessed 16.10.18).

Neises, T., Turchi, C., 2014. A comparison of supercritical carbon dioxide power cycleconfigurations with an emphasis on CSP applications. Energy Proc. 49, 1187–1196.

Nithyanandam, K., Pitchumani, R., 2014. Cost and performance analysis of concentratingsolar power systems with integrated latent thermal energy storage. Energy 64,793–810.

Oshwin Overseas, 2017. ASTM A312 TP 347 Stainless Steel Seamless Pipe. <http://www.oshwin.com/stainlesssteel-piping-tubing-pipes-tubes/stainless-steel-pipes-tubes/ss347-pipes-type-astma312-tp347-stainlesssteel-seamless-pipes/#price> (ac-cessed 17.05.30).

Petrollese, M., Cocco, D., Cau, G., Cogliani, E., 2017. Comparison of three different ap-proaches for the optimization of the CSP plant scheduling. Sol. Energy 150, 463–476.

Pacheco, J.E., Showalter, S.K., Kolb, W.J., 2002. Development of a molten-salt thermo-cline thermal storage system for parabolic trough plants. J. Sol. Energy – T. ASME124, 153–159.

Pfenninger, S., Gauché, P., Lilliestam, J., Damerau, K., Wagner, F., Patt, A., 2014.Potential for concentrating solar power to provide baseload and dispatchable power.Nat. Clim. Chang. 4, 689–692.

Reliable Pipes & Tubes, 2017. Price List of Carbon Steel Pipes% Tubes, Seamless Pipe,Welded Pipe.< http://www.reliablepipestubes.com/blog/price-pricelist-carbon-steel-pipe-seamless-pipe-welded-pipe/> (accessed 17.05.30).

Siemens, 2011. Steam Turbines for CSP Plants. <https://www.energy.siemens.com/co/pool/hq/power-generation/steam-turbines/downloads/steam-turbine-for-csp-plants-siemens.pdf> (accessed 16.11.30).

Siemens, 2013a. Industrial Steam Turbines: The Comprehensive Product Range from 2 to250 Megawatts. <http://www.energy.siemens.com/hq/pool/hq/power-generation/steam-turbines/Industrial_Steam_Turbines_en.pdf> (accessed 16.11.30).

Siemens, 2013b. Pre-Designed Steam Turbines: The Comprehensive Product Range Up to12 Megawatts. <http://www.energy.siemens.com/hq/pool/hq/power-generation/steam-turbines/Pre-Designed_Steam_Turbines_en.pdf> (accessed 16.11.30).

Trieb, F., Schillings, C., Pregger, T., O’Sullivan, M., 2012. Solar electricity imports fromthe Middle East and North Africa to Europe. Energy Pol. 42, 341–353.

Turchi, C.S., Heath, G.A., 2013. Molten Salt Power Tower Cost Model for the SystemAdvisor Model (SAM). National Renewable Energy Laboratory (NREL), Golden.

Usaola, J., 2012. Operation of concentrating solar power plants with storage in spotelectricity markets. IET Renew. Power Gen. 6, 59–66.

Viehbahn, P., Lechon, Y., Trieb, F., 2011. The potential role of concentrated solar power(CSP) in Africa and Europe. Energy Pol. 39, 4420–4430.

697

715

695

6.67

13.65

15.06

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

680 Base Case SS-CS CS-CS

LCO

E (U

S ce

nts/

kWh)

Tota

l Ins

talle

d Co

st (m

illio

ns U

S$)

Fig. 6. Sensitivity results for the material change for the thermal storage tank.

Fig. 7. Selection of materials according to thermal storage temperature and the corre-sponding LCOE.

R. Andika et al. Solar Energy 157 (2017) 552–558

558