Everything You Always Wanted to Know About CSP and Perspectives of... · 2017. 12. 8. · > Farrington Daniels Lecture 2017 > Robert Pitz -Paal • CSP Technology > 31.10.2017 . 1

Robert Pitz-Paal

Everything You Always Wanted to Know About CSP * *But Were Afraid to Ask

1. Characteristics of CSP

2. Market und Cost Development

3. Benefits for a mix of PV und CSP

4. Scientific Challenges in CSP Development

• Shape Accuracy of Solar Concentrators • Controlling the Solar Flux Distribution • Stable and efficient Volumetric Receivers

5. Conclusions

Outline

> Farrington Daniels Lecture 2017 > Robert Pitz-Paal • CSP Technology > 31.10.2017 DLR.de • Chart 2






5. Conclusions

Outline


Thermal Storage vs. Electric Storage

CSP with thermal storage and fossil back provides reliable dispatchble power at no additional cost

2000 h

+2000 h

η >95 %

η = 75%

200 h

Firm capacity


CSP only suitable in areas with high direct normal radiation > Farrington Daniels Lecture 2017 > Robert Pitz-Paal • CSP Technology > 31.10.2017 DLR.de • Chart 5





• Shape Accuracy of Solar Concentrators • Controlling the Solar Flux Distribution • Stable and efficient volumetric receivers

5. Conclusions

Outline


Global expansion of CSP in three phases > Farrington Daniels Lecture 2017 > Robert Pitz-Paal • CSP Technology > 31.10.2017

Lilliestam, J., Labordena, M., Patt, A. & Pfenninger, S. Nat. Energy 2, 17094 (2017).

1988 -1992 PhD

DLR.de • Chart 7



Since 2003 More volumetric receivers in DLR

DLR.de • Chart 8



1997 Sabbatical @ Sandia (US) TRNSYS STEC Library

DLR.de • Chart 9



2003 DLR Head of Divison and Prof. @RWTH Aachen

DLR.de • Chart 10



2006: 1,5 MW volumetric receiver demo plant planed with research & industry partners

DLR.de • Chart 11



2008 QUARZ®: DLR Qualification Lab for Solar Components

DLR.de • Chart 12



2011: Founding Director of new DLR Institute of Solar Research Transfer of Jülich Demo Plant to DLR as research platform

DLR.de • Chart 13



2016: SynLight® Largest artificial sun for solar chemical fuel production

DLR.de • Chart 14

Cost reduction over last 5 years at a learning rate of > 25%

> Farrington Daniels Lecture 2017 > Robert Pitz-Paal • CSP Technology > 31.10.2017


DLR.de • Chart 15

Cost for CSP and PV have dropped dramatically

• Installed CSP capacity is more than an order of magnitude smaller than PV capacty

Energy sales price PV

Energy sales prices CSP

2014

2007

2014

Module price PV

Dubai 2017 Chile 2016

2016

Australia 2017


More than 5000 full load hours

Solar Electricity cheaper than power from gas! 700 MW @ 5500 h CSP á 7,3 $cents/kWh + 800 MW @ 2300 h PV a 3 $Cents/kWh

= 5,95 $cents/kWh = 5,07 €cents/kWh for 24/7electricity







5. Conclusions

Outline


0

1

2

3

4

5

6

[MW

]

x 1

00

00

Installed Capacity

PumpBatteryCSP_SolarPV_SolarWindOtherHydro_NCREHydro_RORHydro_DamDieselLNGGeothermalBiomassCoal

2029

Social acceptance

Energy demand

Technological change in BESS

Externality costs

RE investment costs

Fossil fuel costs

CSP LCOE

Scenario B High High Low High Low High USD 50 /MWh by 2025

Chile Scenario Results – Expansion Model Scenario 1

PV

CSP

0

5

10

15

20

25

[GW

h]

x 1

00

00

Generation

BatteryPumpDieselPV_SolarCSP_SolarWindHydro_RORHydro_NCREGeothermalOtherBiomassLNGHydro_DamCoal

Chile Scenario Results – Expansion Model Scenario 1

2029

Social acceptance

Energy demand

Technological change in BESS

Externality costs

RE investment costs

Fossil fuel costs

CSP LCOE

Scenario B High High Low High Low High USD 50 /MWh by 2025

CSP PV

0

5000

10000

15000

20000

Pow

er (M

W)

PV_SOLCSP_SOLLNG_CCGTLNG_GTWINDHYDRO_RORHYDRO_DAMGEODIECOGBIOCOAL

Chile Szenario results: Short Term Simulation

2035 summer week dispatch by technology

CSP

PV






5. Conclusions

Outline


Introduction: Shape and Slope Deviations

Deviations of the ideal shape of curved mirrors for CSP applications can have a significant impact on the optical efficiency and thus the performance of the power plant. Critical measure is slope deviation, not shape deviation: -> slope needs to be measured accurately, shape is only secondary

Slide 23 www.dlr.de/enerMENA

Deflectometry: Measurement Principle

Projector

Camera

Mirror

Projection screen

α α

i

n

r


Measurement Set-Up for Individual Mirror Panels

Reflected horizontal and vertical stripe patterns

Projected horizontal and vertical stripe patterns

QDec set-up for horizontal and vertical measurement


Example Result for ParabolicTrough Mirror Panel

SDx = 2.5 mrad RMS value of slope deviation in curved direction

SDy = 2.2 mrad RMS value of slope deviation in non-curved direction

FDx = 9.5 mm RMS value of focus deviation

Intercept = 96.8% Expected intercept considering sunshape and additional typical collector errors


X

Z

camera

dabs

dcam

∆z

∆x

reflector

f

nu

nl

∆αu

α1,u

αu

α1,l

∆αl

αl

TARMES (Trough Absorber Reflection Measurement System): Basic idea and set-up of measurement system


Measurement: Turning of collector with camera at close distance (~17m)


Evaluation

• Correction of lens distortion

• Image rectification

• Image treatment

• Edge detection

• Input of geometrical set-up

• Calculation of slope errors


QFly – airborne prediction of the optical performance of parabolic trough collector fields


50 MW PTC solar field (Andasol I) QFly UAV

DLR.de • Chart 30

4. QFly - High Resolution Raw Data


Individual unprocessed photos in 5 min. time lapse:

DLR.de • Chart 31

4. QFly - High Resolution Result: Mirror Shape Maps


Raytracing software to determine intercept / optical performance

Result: Absorber Position

0 6 12 18 La

tera

l dev

iatio

n [m

m]

-20

20

0

10

-10

-30

30

Module length [m]

Accuracy RMS deviation ~1.5 mm

SDx in mrad

0

5

10

-10

-5

Accuracy RMS 0.1 mrad Local ±1 mrad

DLR.de • Chart 32

Gravity Load on Parabolic Trough Refectors


DLR.de • Chart 33

Photogrammetry to measure shape


Facet mounted in collector Simulation Measurement


DLR.de • Chart 36 > Farrington Daniels Lecture 2017 > Robert Pitz-Paal • CSP Technology > 31.10.2017

Final quality inspection …

400°C






5. Conclusions

Outline


Controlling the Solar Flux Distribution


?

optical efficiency safe operation

? problem is complicated by • high degree of freedom • different size and shape of focal images • size and shape varying with time • tracking uncertainty

DLR.de • Chart 38

2. State of the Art Existing tools for central receiver systems:

• Flux density distribution of central receiver systems with different approaches

analytical, e.g. HflCal (Kierra, 1986), Helios (Biggs et al. 1976)

raytracing, e.g. Mirval (Leary et al., 1979)

• Flux density distribution of secondary concentrators

e.g. SORSIM (Denk, 1992)

• Receiver performance

e.g. VOLREC (Hoffschmidt, 1997), VORECO (Buck, 2000)

Aim point strategies for central receiver systems:

• primarily operated to avoid deterioration of the receiver

• increase of the receiver performance was not intended

• more or less provisional

• e.g. PHOEBUS-receiver system on CESA-1-Tower (Berenguel et al., 1999)

Measured vs. simulated flux density distribution of a single heliostat:

measured simulated with a statistical mirror error

low conformity between reality and simulation


Measurement of Heliostat Slope using Deflectometry


Automated deflectometry measurement system Jülich

2014/15

• automatic selection of single heliostats/groups • automatic measurement and data processing • performance: ~60sec./hel.

DLR.de • Chart 40

Validation by Comparison of Ray Tracing Calculations to Flux Measurement Data

Flux Measurement Simulation


Optimization of Heliostat-Aim Point Assignment


continuous optimization: 𝑑𝑑𝑑𝑑𝑑𝑑 𝑆𝑆 = 2 ⋅ 𝑛𝑛𝐻𝐻 discrete optimization: 𝑆𝑆 = 𝑛𝑛𝑍𝑍𝑛𝑛𝐻𝐻 → Ant Colony Optimization Meta-Heuristic (ACO)*

? x

y

H1 H2 H3 H4 … HnH

A1 11 12 13 14 … 1nH

A2 21 22 23 24 … 2nH

… … … … … … …

AnZ nZ1 nZ2 nZ3 nZ4 … nZnH *Belhomme, B.et al. : Optimization of Heliostat Aim Point Selection for Central Receiver Systems Based on the Ant Colony Optimization Metaheuristic. Journal of Solar Energy Engineering, 2014. 136(1).

Natural Role Model: •Ants excrete pheromone on their trails •Pheromone on the trails evaporates over the time •Ants chose their way randomly mixed with a kind of short visibility (myopic) •Ants are strongly attracted by pheromone

DLR.de • Chart 42

Modulnummer horizontal

Mod

ulnu

mm

er v

ertik

al

1 5 10 15 20 25 30 35 40 45

1

6

12

18

24

30

36100kW

300kW

500kW

700kW

900kW

1100kW

Aim Point Optimization @ Solar Tower Jülich

Reference Case: Operator‘s experience Power Output = 100%

Intercept – Optimization Power Output 111.31 %


Modulnummer horizontal

Mod

ulnu

mm

er v

ertik

al

1 5 10 15 20 25 30 35 40 45

1

6

12

18

24

30

36100kW

300kW

500kW

700kW

900kW

1100kW






5. Conclusions

Outline



Energy from the sun: Open volumetric solar receiver

Sour

ce: d

lr.de

Volumetric effect

HiTRec-II SiSiC

honeycomb

DLR.de • Chart 45

What is the perfect absorber?


Honeycomb

Foam

Wire mesh

DLR.de • Chart 46

local hot spots → instable flow at

• high temperatures • linear pressure drop characteristics • low thermal conductivity

viscosity increases with increasing temperature

hot zones are badly cooled

Different Characteristics affecting Flow Stability

IR-Camera

Heater

HeaterPorousMonolith

Cold air

How can instable flow be visualized?

by thermograph monitoring of the cooling of a heated porous monolith

V=const.


cordierite honey comb

SiC foam

geometry/pressure loss characteristics influences flow stability

heat conductivity influences flow stability

v = const. v = 0 in hot channels



Optimizing the Absorber Design

State-of-the-art Unit element

Unit element

Increase cellularity and porosity

Unit element

Decrease inlet radiative losses

DLR.de • Chart 50


Optimizing the Absorber Design Numerical Simulation Innovative

geometry

Tair-out: 1149 K η = 90 %

HiTRec-II

Tair-out: 1012 K η = 72%

Optimized design

State of the arrt design

DLR.de • Chart 51


Prototype sample production by 3D printing

Cylindrical prototype test-sample: Ti6Al4V 3:1 scaled up geometry

Front view Top view Bottom view

DLR.de • Chart 52


Experimental Validation of Prototype

Thermal efficiency evaluation 20 kW solar simulator

DLR.de • Chart 53






5. Conclusions

Outline


• CSP troughs and towers with large thermal energy storage systems are commercial products today

• In combination with PV, CSP is competitive to 24/7 power from natural gas under favorable conditions

• Today we better understand how to measure, model and optimize • large solar fields of parabolic troughs and heliostats, • solar receivers and storage systems with different heat transfer fluid, • the impact of environmental effects like sunshape and aerosols to

maximize the performance and lifetime of a CSP plant.

• With 5 GW installed the technology is very young and significant further improvement is feasible

• Major future challenges are related to integrate new power cycles that operate at elevated temperatures and require new heat transfer and storage fluids

Conclusion


Everything You Always Wanted to Know About CSP * *But Were Afraid to AskOutlineOutlineThermal Storage vs. Electric StorageCSP only suitable in areas with high direct normal radiation OutlineGlobal expansion of CSP in three phasesGlobal expansion of CSP in three phasesGlobal expansion of CSP in three phasesGlobal expansion of CSP in three phasesGlobal expansion of CSP in three phasesGlobal expansion of CSP in three phasesGlobal expansion of CSP in three phasesGlobal expansion of CSP in three phasesCost reduction over last 5 years �at a learning rate of > 25%Foliennummer 16Foliennummer 17OutlineFoliennummer 19Foliennummer 20Foliennummer 21OutlineFoliennummer 23Foliennummer 24Foliennummer 25Example Result for ParabolicTrough Mirror PanelTARMES (Trough Absorber Reflection Measurement System):�Basic idea and set-up of measurement systemMeasurement:�Turning of collector with camera at close distance (~17m)�EvaluationQFly – airborne prediction of the optical performance of parabolic trough collector fields 4. QFly - High Resolution�Raw Data4. QFly - High Resolution�Result: Mirror Shape Maps�Gravity Load on Parabolic Trough RefectorsPhotogrammetry to measure shape Facet mounted in collector�Simulation MeasurementFinal quality inspection … OutlineControlling the Solar Flux Distribution Foliennummer 39Measurement of Heliostat Slope using DeflectometryValidation by Comparison of Ray Tracing Calculations �to Flux Measurement DataOptimization of Heliostat-Aim Point AssignmentAim Point Optimization @ Solar Tower JülichOutlineFoliennummer 45What is the perfect absorber?Different Characteristics affecting Flow StabilityFoliennummer 48Foliennummer 49Foliennummer 50Foliennummer 51Foliennummer 52Foliennummer 53OutlineConclusion

Documents

Everything You Always Wanted to Know About CSP and Perspectives of... · 2017. 12. 8. · > Farrington Daniels Lecture 2017 > Robert Pitz -Paal • CSP Technology > 31.10.2017 . 1