CC BY-NC-ND
0CrossMark
Available online at www.sciencedirect.com
ScienceDirect
Energy Procedía 49 (2014) 1117 - 1126
SolarPACES 2013
Solar-only parabolic trough plants with high steam parameters
J. Derscha*, T. Vogelb, T. Polklasc, C. Tummersc
a DLR, German Aerospace Center, 51170 Köln, Germany b University of Duisburg-Essen, Leimkugelstrasse 10, 45141 Essen, Germany c MAN Diesel & Turbo SE, Steinbrinkstrasse 1, 46145 Oberhausen, Germany
Abstract
A technical and economical comparison of two parabolic trough CSP plants is made using full thermodynamic models. One plant is a hybrid plant with a natural gas fired booster in order to reach 540°C live steam temperature. The design of this plant is similar to that of the Shams One plant and a meteorological dataset for the site in Abu Dhabi was used for the annual performance calculations. The second plant is a parabolic trough plant using molten salt as heat transfer fluid in the solar field as well as storage material. Both plants are equipped with almost the same power block and show similar annual performance. The molten salt plant shows a cost reduction potential and the same or even higher flexibility and dispatch ability as the hybrid plant. © 2013 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/3.0/).
Selection andpeer reviewbythescientific conference committee ofSolarPACES2013underresponsibilityofPSEAG. Final manuscript published as received without editorial corrections. Keywords: parabolic trough; storage, molten salt, hybrid, annual performance
1. Introduction
In solar-only CSP plants without storage the electricity production depends completely on the instantaneous direct irradiation and thus may vary significantly during the day. Furthermore electricity demand after sunset cannot be fulfilled. Hybridization or addition of thermal storage can be used to overcome these drawbacks. The former solution is applied at Shams One [1] where natural gas burners are used for final superheating and to replace the solar heat during times when the input from the solar field is not sufficient.
* Corresponding author. Tel.: +49-2203-6012219; fax: +49-2203-6014141. E-mail address: juergen.dersch@dlr.de
1876-6102 © 2013 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/Kcenses/by-nc-nd/3.0/).
Selection and peer review by the scientific conference committee of SolarPACES 2013 under responsibility of PSE AG. Final manuscript published as received without editorial corrections. doi: 10.1016/j.egypro.2014.03.121
Hybridization in a manner like that applied at Shams One has two positive effects: the plant output is enhanced by fossil fuel utilization and the cycle efficiency is increased by higher live steam temperatures. The impacts of this measure as well as further possible variants of hybrid CSP plants are investigated in a separate paper [2]. Though hybridization is considered as one solution, it is not feasible for all countries, either due to the lack of fossil fuels or due to legal restrictions. Therefore the present paper is dealing with a solar-only solution in combination with thermal storage and molten salt as heat transfer fluid to provide almost the same flexibility and cycle efficiency as the hybrid power plant. Although molten salt is not the state-of-the-art heat transfer fluid in parabolic trough CSP plants, it offers promising enhancements for the future and this paper is going to investigate this in detail.
2. Methodology
The basis of this study is a CSP plant similar to Shams One described in [1]. Since many details of Shams One are not published, a lot of additional assumptions are necessary in order to complete the model. The power block is very similar to that used at Shams One, but details of the solar field and the operating strategy in particular is solely based on author's assumptions. DNI and ambient temperature data are taken from METEONORM 6.1 and scaled linearly to the 1925 kWh/m2 per year, a value published in [1]. Furthermore Shams One uses Abengoa Solar's ASTRO 150 collector and this study uses data of the SKAL-ET collector since important performance details of the ASTRO 150 type were not available. In principle both trough types should performs very similar provided that they are equipped with the same HCE type.
The complete plant is modeled in Ebsilon®Professional using the EbsSolar library [3]. This software is a thermodynamic cycle balance program with graphical user interface. The simulation is based on mass and energy balances with pressure drop information and fluid property functions. Originally made for steady state simulations of power plants in a few operating states, it currently offers also annual yield calculations by utilization of the times series option in combination with a script language. In the current study results from the annual performance calculations are compared using levelized cost of energy (LCOE) as figure of merit. Using a full thermodynamic model of the plant offers the option to include several restrictions concerning operation states which are caused by the operating limits of the equipment. These limits are sometimes difficult to consider in models based on lookup tables. The limits considered here in particular are steam quality at turbine exit and condenser pressure.
3. Description of the CSP plant models
Fig. 1 shows the process flow diagram of the reference plant as modeled in Ebsilon®Professional. In order to ease the overview, the cycle has been simplified and the HTF heater as well as the booster are shown as grey boxes.
The turbine is a single case turbine with 540°C/100 bar live steam conditions. The plant is equipped with an air cooled condenser and the design condenser pressure is 0.13 bar. The turbine has 6 steam extractions for preheating and the nominal output is 119 MWel. The solar field is used to heat up the heat transfer fluid (eutectic mixture of biphenyl and diphenyl-oxide) from 295°C to 393°C under design conditions. The upper HTF temperature is kept constant while the lower temperature may vary with power block load and ambient conditions. The solar steam generator delivers superheated steam of 380°C which is afterwards further superheated up to 540°C by the booster. The HTF heater may be used during time periods when the heat delivered by the solar field is not sufficient to produce the required electrical output. In [1] an annual fraction of 3 % from the HTF heater and 18 % from the booster of the total heat utilized by the plant of heat is given.
The annual fraction of heat produced by fossil fuel depends heavily on the operating strategy of the plant. Since the operating strategy of Shams One is not published, a reasonable strategy was assumed. It may be described as follows:
• The solar field output determines the power block operation. Whenever the heat delivered by the solar field is sufficient the PB is operated.
• The booster is switched on full load when the heat input to the power block is higher than 240 MWth.
• The HTF heater is only operated when DNI of the current hour is more than 250 W/m2 lower than DNI of the preceding hour. Heat delivered by the HTF heater is calculated in a manner that the total HTF heat (from solar
and fossil) is equal to the nominal HTF heat. Furthermore the HTF heater is only operated before 4 p.m. since otherwise it would go into operation almost every day during late afternoon hours when DNI naturally decreases.
• Maximal condenser pressure is 0.23 bar, once the pressure is reaching this upper limit, the fossil and eventually the solar heat input to the cycle is reduced.
• Minimal steam quality at turbine outlet is 0.85. Once this lower limit is reached, the booster is used to increase the live steam temperature
Fig. 1. Process flow diagram of the hybrid reference CSP plant
The process flow diagram of the molten salt plant is shown in Fig. 2. The main differences to the hybrid plant are in the solar field. The molten salt plant has two storage tanks in order to store the heat delivered by the solar field, a hot tank at 520°C and a cold tank at 270°C. The fluid used in the solar field and as storage medium is HITEC® heat transfer salt (eutectic mixture of potassium nitrate, sodium nitrite and sodium nitrate). The allowed operating temperature of this molten salt is between 149°C and 538°C [4]. Thus the design outlet temperature of the solar field is set to 520°C and the design live steam temperature is set to 500°C for the molten salt CSP plant. The molten salt CSP plant has no booster but only a heater with limited thermal output. This heater is only used as emergency heater when the temperature of the salt system is going to fall below the lower temperature limit. It is not used for electricity production and also not used for normal freeze protection of the molten salt system. This is done by heat from the thermal storage. According to the assumed operating strategy for this plant, a certain amount (approx. 10%) of the storage content is not used for steam production but is utilized to keep the salt system above 180°C during non-sunshine hours.
The power block is almost the same as for the hybrid reference plant with the only exception that the live steam temperature is lower (500°C instead of 540°C). The solar field area is slightly increased in order to deliver excess heat for charging the thermal storage. Furthermore the type of parabolic troughs has been changed from Eurotrough in the hybrid case to Ultimate Trough®. This was done since the Ultimate Trough® offers higher concentration ratio which is an advantage when operation at higher temperatures. Comparing figs. 1 and 2 it becomes obvious that the graphical representation of the solar fields look somewhat different. The reason for that is that the hybrid plant model uses a simple model for the whole solar field, whereas the molten salt plant model uses models for each collector in a representative loop of the solar field. The underlying equations are the same but the approach used for the molten salt plant allows the utilization of more temperature nodes for heat loss calculation. This is necessary since the heat loss equation is a 4th order equation and the temperature spread in the solar field is about 2.5 times higher for the molten salt plant.
Fig. 2. Process flow diagram of the molten salt CSP plant
The operating strategy of the molten salt CSP plant may be described as follows:
• In the morning hours heat produced by the solar field is primarily used to charge the storage.
• The power block is not operated until the storage content reaches 60% of its nominal capacity.
• Once the power block has been started, it is operated at maximum possible load until the storage reaches the minimum level for PB operation.
• During non-sunshine hours salt from the cold tank is circulated through the solar field in order to keep the temperature always above 180°C. The return flow from the solar field is mixed with a small salt stream from the hot tank in order to reach a mixing temperature of 270°C prior to the cold tank entrance.
This is a simple operating strategy and many others are possible since the thermal storage decouples electricity production from solar heat production. Table 1 gives a summary of the main technical parameters of both plants.
J. Dersch et al. /Energy Procedía 49 (2014) 1117 - 1126 Table 1. Design data for both CSP plants.
Plant Hybrid reference plant Molten salt CSP plant Unit
Annual sum of DNI 1925 1925 kWh/m2
Design ambient temperature 24 24 °C
Solar field net aperture area 627786 648595 m2
Trough type Eurotrough Ultimate Trough® -
HCE type Schott PTR 70 Schott PTR 70 -
Optical efficiency 75 77.5 %
Power block nominal gross output 119.4 118.1 MWel
Live steam temperature 540 500 °C
Live steam pressure 100 100 bar
Nominal gross power block efficiency 39.8 38.8 %
Cooling type ACC ACC -
Storage capacity - 1063 MWhth
4. Economic model
The LCOE are calculated using a simplified method proposed by Roy et al. [6]
Cinvest ' (xCRF + xins ) + CO
LCOE =
el ,net
(1 + i)n • i (1+i)n -1
Cinvest (ASF ' cSF + cland ' ASF ' xland + Pnom,PB ' cPB ) ' (1 + xEPC )
C = C + C + C
O&M ^staff ^ ^ water ^ ^spare
~ (xSFstaff ' ASF + n
SF lPB-employees
). c + x • E + x • C
/ staff water el ,net,ann spare i
spare invest
Almost identical economic parameters are used for both plants. The values are given in table 2 and they are based on reasonable assumptions of the authors. For the Ultimate Trough® cost advantages of up to 25% compared to the Eurotrough have been published [7]. However in the current study identical specific cost assumptions are used for both plants since for the molten salt plant more expensive materials must be used due to increased operating temperatures and corrosion issues. Furthermore extensive heat tracing will be necessary for the molten salt plant. On the other hand side, the molten salt plant does not need a separate HTF system and it has no booster.
Table 2. Parameters of the economic model
Plant Hybrid Molten salt Unit
reference plant CSP plant
Specific investment cost solar field 210 210 €/m2
Specific investment cost power block 650 650 €/kWel
Specific investment cost BOP 100 100 €/kWel
Specific investment cost HTF system 80 - €/m2
Specific investment cost electric installation 170 170 €/kWel
Land cost 4 4 %
Specific cost for site preparation, buildings, etc. 50 50 €/m2
Specific investment cost booster 70 - €/kWth
Specific investment cost HTF heater 50 50 €/kWth
Specific investment cost storage - 30 €/kWth
Surcharge for development, engineering, risk, etc. 30 30 %
O&M material cost 1 1 % per year
Fuel cost (natural gas) 18 18 €/MWh
Number of employees for PB 25 25 -
Specific number of employees for SF 0.03 0.03 1/1000m2
Average cost per employee 40000 40000 €
Specific water consumption 0.3 0.3 m'/MWh
Water cost 2 2 €/m3
Real interest rate 8 8 %
Depreciation period 25 25 years
Annual insurance cost 0.7 0.7 %
5. Results
Table 3 gives a survey about the main results of the annual performance calculations. The molten salt CSP plant may be considered as solar-only plant since a negligible amount of natural gas is used only for freeze protection during time periods without any DNI for one or more days. The electrical output of the molten salt plant is about 5 % higher compared to the reference plant. This is due to a larger aperture area (+ 3 %) but also due to higher solar heat utilization factor and higher annual net power block efficiency. Net power block efficiency on annual basis of the molten salt plant is higher though the design efficiency is lower. This is because the power block of the molten salt plant runs near full load for most of the time. The resulting LCOE for the molten salt plant are about 12% lower than those for the hybrid reference plant.
Figure 3 shows a comparison of the monthly electricity production for both plants. The operating strategy has of course an impact on the annual electricity output but the strategy chosen here is very similar for both plants: DNI determines electricity production and the stored energy is converted to electricity as soon as possible. The monthly distribution is quite similar, showing that the molten salt solar-only plant is able to provide almost the same performance as the hybrid plant. For most of the months the output of the molten salt plant is slightly higher, except for winter months, when shorter days are leading to higher heat consumption for freeze protection.
Table 3. Results of the annual performance simulations
Plant Hybrid reference plant Molten salt CSP plant Unit
Net electricity output 247.3 259.2 GWhei
Total heat used in PB 714.6 719.9 GWhth
Net power block efficiency 34.6 36.0 %
Total fuel consumption 89735 324 MWh
PB operating hours 3201 2706 h
Equivalent PB full load hours 2190 2277 h
Solar heat utilized 92.2 99.3 %
Solar fraction 87.4 99.9 %
LCOE 206 181 €/MWhei
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Fig. 3. Monthly net electricity production of the hybrid reference and molten salt CSP plants.
Figure 4 shows daily performance curves for two sunny days, one summer day and one winter day. Again the performance is similar and the molten salt plant shows somewhat higher total output. The output of the hybrid reference plant naturally follows the DNI curve while the output of the molten salt plant is shifted towards the evening hours by means of the storage. This could be an advantage when the national demand curve shows an evening peak. The output of the molten salt plant for the summer day suffers from the high ambient temperatures in the afternoon, causing a load reduction due to the condenser pressure limit. The hybrid reference plant shows this load reduction too but it coincides with decreasing DNI and thus the impact in late afternoon is not that obvious. The hybrid plant shows one more operation hour for both days but this could be easily reached by the molten salt plant with a modified operating strategy and e.g. slightly reduced load during the day.
21. Jun
□solar field
output □heat to power block ■•□net electricity
£ ¿V ^ local time
23. Dec
□solar field
output □heat to power block -•□net electricity
400 : c
300 200 100 0
S ^ local time
21. Jun
□solar field
output □heat to power
block □net electricity
£ ^y local time
<9 ❖
1000 5.0E+05
800 4.0E+05
600 Hi £ 3.0E+05
500 i g
400 z " 2.0E+05
300 tí
200 o 1.0E+05
0 0.0E+00
□solar field output
□ heat to power block
□ net electricity
1000 900 800 700
400 : 1
300 200 100 0
1000 900 800 700
400 : I
300 200 100 0
local time
Fig. 4. Performance during two individual days. Left hand side: hybrid CSP plant; right hand side: molten salt CSP plant.
1.20 1.15 1.10
g 1.05
1 1.00
" 0.95 0.90 0.85 0.80
-♦-I Solar field -■-I Power block
-±-I Storage Interest rate
O&M replacement rate
0.7 0.8 0.9 1 1.1
relative change of input parameter
23. Dec
Fig. 5. Sensitivity analysis for the molten salt CSP plant concerning some economic parameters
The cost assumptions for both plants are of course associated with high uncertainties, thus a sensitivity analysis was made varying some economic assumptions of the molten salt plant systematically. Figure 5 shows the result of this analysis. As known from other studies, the interest rate has a severe impact on LCOE since CSP plants need a lot of capital. The second important parameter is the cost assumption for the solar field. Reduction of the specific
investment costs of the solar field by 20% gives about 8% lower LCOE values. Assumptions for specific power block investment costs, thermal storage investment costs, and annual O&M spares show lower impact on LOCE.
6. Conclusions
The technical and economical comparison of two parabolic trough CSP plants is made using full thermodynamic models. One plant is a hybrid plant with a natural gas fired booster in order to reach 540°C live steam temperature. This plant is designed similar to the Shams One plant and a meteorological dataset for the site in Abu Dhabi was used for the annual performance calculations. The second plant is a parabolic trough plant using molten salt as heat transfer fluid in the solar field as well as storage material. Both plants are equipped with almost the same power block and show similar annual performance.
The molten salt plant shows a cost reduction of about 10 % compared to the reference plant. Furthermore the molten salt plant offers the option to shift electricity production towards the evening hours since it has a thermal storage. Decoupling electricity production from solar resource by using a storage results in a higher capacity factor since the power block may be operated near full load most of the time. The utilization of a full thermodynamic model for annual performance calculations has the advantage that one can easily consider restrictions caused by operating limits of individual parts like condenser pressure or steam quality.
Nomenclature
CSP concentrated solar power
C total cost in €
c specific cost in €/m2 or €/kW
LCOE levelized cost of energy in €/MWh
i interest rate in %/a
n number of persons
PB power block
DNI direct normal irradiance
SF solar field
BOP balance of plant
HTF heat transfer fluid
Acknowledgements
The authors would like to thank the German Federal State of North Rhine-Westphalia and the European Regional Development Fund for the financial support of the project TURIKON in the frame of the program progress.NRW and the goal 2-program 2007-2013, Phase VI (Grant No. 64.65.69-EN-2019).
References
[1] Göbel, O.: Shams One 100 MW CSP Plant in Abu Dhabi - Update on Project Status; Proceedings of the SolarPACES 2010 Conference, Perpignan, France, September 21-24, 2010
[2] Vogel, T.; Oeljeklaus, G.; Görner, K.; Dersch, J.; Polklas, T.: Hybridization of Parabolic Trough Power Plants with Natural Gas, Paper submitted to the SolarPACES 2013 Conference, Las Vegas, Nevada, USA, September 17-20, 2013
[3] Pawellek, R.; Löw, T.; Hirsch, T.: EbsSolar - A Solar Library for Ebsilon®Professional. Proceedings of the SolarPACES 2009 Conference, Berlin, Germany, September 15-18, 2009
[4] HITEC® Heat Transfer Salt, Product data sheet, Coastal Chemical Co., L.L.C., 5300 Memorial Drive, Suite 250, Houston, TX 77007
[5] Ultimate Trough® Solar Collectors for Concentrated Solar Power, Product information, FLABEG Holding GmbH, Waldaustraße 13, 90441 Nürnberg, Germany
[6] Roy A., Meinecke W., Blanco Muriel M. (editors): Introductory Guidelines for Preparing Reports on Solar Thermal Power Systems, SolarPACES Report No. III-3/97, 1997
[7] Riffelmann, K.J.; Graf, D.; Nava, P.: Ultimate Trough - The new parabolic Trough Collectors Generation for Large Scale Solar Thermal Power Plants. Proceedings of the ASME 2011 5th International Conference on Energy Sustainability, ES2011, August 7-10, 2011, Washington, DC, USA
