1

2

Ident if icat ion of opt im um m olt en salt s for use as heat t ransfer f luids inparabolic t rough p lant s. A t echno-econom ic com parat ive opt im izat ion

1 Solar Therm al Energy Research Group (STERG), St ellenbosch Universit y, Privat e Bag X1, 7602, M at ieland, Sout h Africa2 Depart m ent of M echanical Engineering, DTU, Technical Universit y of Denm ark, 2800 , Kgs. Lyngby, Denm ark

3 Depart m ent of Energy Technology, KTH Royal Inst it ut e of Technology, 100 44, St ockholm , Sw eden

Christ oph A. Pan1, Davide Ferruzza2, Rafael Guédez3, Björn Laum ert 3, Fredrik Haglind 2, Frank Dint er1

3

1. Int roduct ion2. Background3. Research quest ion4. Heat t ransfer f luids5. Advant ages of m olt en salt s6. Techno- econom ic opt im izat ion7. Result s8. Next st eps9. Conclusion10. References

4

.00

500.00

1000.00

1500.00

2000.00

2500.00

3000.00

3500.00

4000.00

Operational Underconstruction

Development Planned

inst

alle

d ne

t cap

acity

of c

omm

erci

al

pow

er p

lant

s [M

We]

Parabolic trough Solar tower Linear Fresnel

• PT m ost m at ure and bankable CSP t echnology

• Tow ers are cat ching up• Increase com pet it iveness of PT by

reducing LCOE• Higher cycle ef f iciency t hrough

larger cycle t em perat ure d if ference needed

• M olt en salt linked t o challenges• New opport unit ies for CSP in

Sout hern Africa

Source: CSP World, 2017

5

• LCOE reduct ion by sw it ching f rom t herm al oil t o m olt en salt– 12.1 % (Dersch et al., 2014)– 16.4 % (Boukelia et al., 2015)– 20 % (Ruegam er et al., 2014)

• LCOE reduct ion by using d if ferent m olt en salt s– 3 t o 5 % (Ruegam er et al., 2014)

• LCOE variat ions in lit erat ure– 7.59 t o 23.6 $cent / kWhe

• St ud ies only look at one pow er p lant conf igurat ion and com pare it w it hot her f luids, but w hat about d if ferent conf igurat ions?

large variat ions caused by dif ferent assum pt ions and locat ions (Abu Dhabi, Béchar, Dagget t )

sm all variat ions but st ill signif icant

6

• Which m olt en salt as heat t ransfer f luid (HTF) is m ost suit ab le t o im prove t he perform ance of a parabolic t rough pow er p lant f rom a t echno- econom ic st andpoint for a variet y of conf igurat ions and locat ions?

7

-50 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900

NaKSodium

Na-K-Zn chloridesLi-Na-K flour./carbon.

Li-Na-K carbonatesSaltstream 700

Na-K-Li nitrates/nitritesK-Li-Ca nitratesNa-K-Li nitratesSaltstream 500

Sandia MixHitec XL

HitecSolar Salt

Dowtherm ATherminol VP-1

Temperature [°C]

Thermal oils

Molten salts

Liquid metals

Commercially used

Investigated in this study

Sources: Dow Chemical Company, 2001; Bradshaw & Siegel, 2008; Vignarooban et al., 2015

8

Adapted from: Archimede Solar Energy, 2016

Cycle efficiency up to 38 % Cycle efficiency over 40 %

9

• Opt im izat ion of t echnical and f inancial ind icat ors– LCOE– CAPEX– IRR

• Ident if icat ion of opt im al HTF for– various com ponent sizes (TES, solar f ield , t urb ine capacit y, et c.)– various operat ing st rat egies (base load or peaking)– various locat ion- based f ram ew ork cond it ions (solar resource, feed in t arif f ,

f inancial param et ers)• Populat ion based evolut ionary algorit hm• Range of solut ions t hat are opt im al• Com prom ise bet w een cost and qualit y

10

Source: Spelling, 2013

11

• Choice of HTF is dependent on DNI and syst em size• Solar Salt of fers low est LCOE

500 600 700 800 900 1000 1100

CAPEX [milUSD]

70

80

90

100

110

120

130

140

150

LCO

E [U

SD/M

Whe

]

South Africa

500 600 700 800 900 1000 1100

CAPEX [milUSD]

70

80

90

100

110

120

130

140

150

LCO

E [U

SD/M

Whe

]

Spain

500 600 700 800 900 1000 1100

CAPEX [milUSD]

70

80

90

100

110

120

130

140

150

LCO

E [U

SD/M

Whe

]

Chile

Thermal oil Solar Salt Hitec Hitec XL

12

500 600 700 800 900 1000

CAPEX [milUSD]

100

110

120

130

140

150

160

LCO

E [U

SD/M

Whe

]

Hitec XL

500 600 700 800 900 1000

CAPEX [milUSD]

100

110

120

130

140

150

160

LCO

E [U

SD/M

Whe

]Solar Salt

500 600 700 800 900 1000

CAPEX [milUSD]

100

110

120

130

140

150

160

LCO

E [U

SD/M

Whe

]

Thermal oil

500 600 700 800 900 1000

CAPEX [milUSD]

100

110

120

130

140

150

160

LCO

E [U

SD/M

Whe

]

Hitec

• Freeze alert t em perat ure variat ion in Sout h Africa• Set point should be close t o solar f ield out let t em perat ure for sm all

syst em s• Tem perat ure can be low ered signif icant ly in large syst em s

13

• Vary ot her param et ers (t urb ine capacit y, locat ion)• Analyze d if ferent operat ing st rat egies (base load, peaking)• Opt im ize IRR, capacit y fact or• Fut ure applicat ions of opt im izat ion

– TES t echnology (d irect st orage, t herm ocline)– Wat er consum pt ion– St eam generat or (ram p- up rat es, p inch point s)

14

• Opt im al HTF dependent on syst em size and DNI• Freeze alert t em perat ure has signif icant im pact on p lant econom ics• M odel can be used t o f ind opt im al pow er p lant conf igurat ion in t erm s of

– locat ion– operat ing st rat egy– feed in t arif f– cost s

• New salt m ixt ures needed– cheaper– w ider operat ing t em perat ure range

• Increase com pet it iveness of parabolic t rough• Support for policy m akers, developers and p lant operat ors• Technology prom ot ion & developm ent opport unit ies in Sout hern Africa

15

• Archim ede Solar Energy. 2016. M olt en salt s vs Oil. [Online], Available: ht t p :/ / w w w .archim edesolarenergy.it / m olt en_ salt _ vs_ oil.ht m [2017, January 13].

• Boukelia, T.E., M ecibah, M .S., M eriche, I.E. & Belgroun, Z. 2015. Com parat ive st udy of parabolic t rough pow er p lant w it h int egrat ed TES and BS using synt het ic oil and m olt en salt . In Tlem cen, Algeria: IEEE 3rd Int ernat ional Conference on Cont rol, Engineering & Inform at ion Technology (CEIT).

• CSP World . 2017. CSP World M ap. [Online], Availab le: ht t p :/ / w w w .cspw orld .org/ cspw orldm ap [2017, January 12].

• Dersch, J., Vogel, T., Polklas, T. & Tüm m ers, C. 2014. Solar- only Parabolic Trough Plant s w it h High St eam Param et ers. Energy Procedia. 49:1117–1126.

• Dow Chem ical Com pany. 2001. DOWTHERM A - Synt het ic Organic Heat Transfer Fluid - Liquid and Vapor Phase Dat a. [Online], Availab le: ht t p :/ / m sdssearch.dow .com / PublishedLiteratureDOWCOM / dh_ 097a/ 0901b8038097ad39.pdf?f ilepath=heat t rans/ pdfs/ noreg/ 176- 01463.pdf&from Page=Get Doc [2017, January 12].

• Ruegam er, T., Kam p, H., Kuckelkorn, T., Schiel, W., Weinrebe, G., Nava, P., Rif felm ann, K. & Richert , T. 2014. M olt en Salt for Parabolic Trough Applicat ions: Syst em Sim ulat ion and Scale Ef fects. Energy Procedia. 49:1523–1532.

• Spelling, J.D. 2013. Hybrid Solar Gas- Turbine Pow er Plant s: A Therm oeconom ic Analysis. PhD Thesis, KTH Royal Inst it ute of Technology. St ockholm , Sw eden.

• Vignarooban, K., Xu, X., Arvay, A., Hsu, K. & Kannan, A.M . 2015. Heat t ransfer f luids for concent rat ing solar pow er syst em s – A review . Applied Energy. 146:383–396.

16

ACKNOWLEDGEMENTS:

CONTACT DETAILS:

visit us: concent rat ing.sun.ac.za

Solar Therm al Energy Research Group (STERG)Stellenbosch Universit y South Africa

Christ oph A. Pan