Advances in Concentrating Solar Power

Collectors: Mirrors and Solar Selective

Coatings
C.E. Kennedy
1National

Renewable Energy Laboratory (NREL), 1617 Cole Boulevard, M/S 3321, Golden, CO
80401-3393, 303-384-6272, 303-384-6103 (fax), cheryl_kennedy@nrel.gov

AIMCAL
Scottsdale, AZ
October 10, 2007
NREL/PR-550-43695
Presented at the Association of Industrial Metallizers, Coaters and Laminators (AIMCAL) 2007 Fall Technical Conference,
21st International Vacuum Web Coating Conference held October 7-10, 2007 in Scottsdale Arizona

Outline

Solar Market Potential

Solar Reflectors
Solar Selective Coating
Conclusion

Concentrating Solar Power Technologies

Parabolic Trough

Compact Linear
Fresnel Reflector
(CLFR)

Power Tower

Dish-Stirling

Solar concentration
allows tailored design
approaches
100kW LCPV Tracking

CPV Heliostat

CPV Winston Collector

DOE & WGA determined feasibility of

1000MW in SW
Southwest Solar Resources

DOE & WGA determined feasibility of

1000MW in SW
Southwest Solar Resources
Transmission Overlay

DOE & WGA determined feasibility of

1000MW in SW
Southwest Solar Resources
Transmission Overlay
Eliminate locations
< 6.75 kwh/m2/day

DOE & WGA determined feasibility of

1000MW in SW
Southwest Solar Resources
Transmission Overlay
Eliminate locations
< 6.75 kwh/m2/day
Exclude environmentally
sensitive lands, major
urban areas, and water
features

DOE & WGA determined feasibility of

1000MW in SW
Southwest Solar Resources
Transmission Overlay
Eliminate locations
< 6.75 kwh/m2/day
Exclude environmentally
sensitive lands, major
urban areas, and water
features
Remove land areas >
(3%) & 1% average
land slope
Remove land <5 contiguous km2

SW Solar Energy Potential

State
AZ
CA
CO
NV
NM
TX
UT
Total

Land Area
(mi2)
19,279
6,853
2,124
5,589
15,156
1,162
3,564
53,727

Solar
Capacity
(MW)
2,467,663
877,204
271,903
715,438
1,939,970
148,729
456,147
6,877,055

Solar
Generation
Capacity
GWh
5,836,517
2,074,763
643,105
1,692,154
4,588,417
351,774
1,078,879
16,265,611

The table and map represent land that has no primary use today,
exclude land with slope > 1%, <5 contiguous km2, & sensitive
lands.
Solar Energy Resource 6.75 kwh/m2/day
Capacity assumes 5 acres/MW
Generation assumes 27% annual capacity factor

Current total generation in the

U.S. is 1,000GW w/ generation
approximately 3,800 TWh

Renewable Portfolio Standards

WA: 15%
by 2020

MN: 25% by 2025;

Xcel: 30% by 2020

MT: 15%
by 2015

OR: 25%
by 2025

VT: 10%by 2013*

ND

ID

WI: 10%
by 2015

SD
WY

NV: 20%
by 2015
CA:
20%by 2010
33% by 2020

CO: 20%
by 2020

AZ: 15% NM: 20%

by 2025
by 2020

IL: 25%IN
MO: by 2025
11%by
2020*

KS

OK

NY: 24%
by 2013

MI

IA:
105 MW

NE
UT

NH: 16%
by 2025

PA: 18%
by 2020

OH

WV

NC

TN
AR

SC
MS

TX: 5,880 MW
by 2015

VA

KY

AL

ME: 10%
by 2017
MA: 4% new by 2009
RI: 15%by 2020
CT: 23%by 2020
NJ: 22.5%by 2021
DE: 20%by 2019
DC: 11%by 2022
MD: 7.5% by 2019
VA: 12%by 2022*
NC: 12.5% by 2021

GA

LA
AK

FL

* Voluntary Goals

HI: 20%by 2020

State RPS mandates successfully jump-starting desirable growth

Market for Solar in US SW

10,000 MW of CSP by 2020

California:

500 MW by 2010
8,000 MW by 2020 peaking demand

354 MW SEGS trough plants in CA

2 PPAs for 1.75 GW Dish Stirling plants in Southern
CA

500 MW (option to expand to 850 MW) Mojave Desert

300 MW (two options to expand to 900 MW) Imperial
Valley

553 MW PPA signed PGE, CA

300 MW PGE, CA Pending contractual announcement
175 MW PGE/FPL CLFR (commitment)
200 MW FPL CLFR (commitment)
1000 MW PGE (commitment) probably in CA

Arizona: 2,000 MW

Florida: 300 MW CLFR (FPL Commitment)

1 MW trough plant in AZ

Nevada: 1,500 MW

64 MW trough project in NV

New Mexico: TBD

West Texas: 1,000 + MW
Colorado:500 MW after 2010

Numerous RFPs in CO, TX, AZ,

10 MW initial (w/ option to expand to 300 MW)

500 MW FPL (commitment) in CA, FL, & other states

International CSP Project

Developments

1000MW CSP USA

30MW ISCCS Mexico

500MW CSP Spain

30MW ISCCS Morocco

30MW ISCCS Egypt
250MW SEGS Israel
400MW ISCCS Iraq
30MW ISCCS Algeria

100MW CSP South Africa

720 kW CPV Australia

154MW CPV Australia

Parabolic Trough Plants

Source: KJC Operating Company

Parabolic Trough
Cost Reduction Scenario
0.25

Advanced
Technology

0.20

Learning &
Competition
Increasing Plant
Size
Alternative
Financing
Tax Neutrality for
Solar Fuels
Tax Incentives

Nominal (LCOE)

Good Solar
Resource Site

0.236

Competitive Range
CA MPR Range
Gas Price: $6 /MMBtu

Future
0.162
0.143

0.15

0.120
0.10

0.110

0.102

0.094

0.05

0.00

1 - 50 MW
Plant (NV)

1 - 100 MW
Plant (CA)

4 x 100MW
Solar Park

Advanced
Technology

@ 1000 MWe @ 2000 MWe @ 4000 MWe

Deployment Deployment Deployment

Location: Barstow, CA
Incentives: Current California
Deployment Assumes:
- 90% PR in Solar Field
- 95% PR in Power Plant

Goals for Improved

Optical Materials

>90% Specular reflectance

into a 4-mrad cone angle
Unofficially 95%

10 - 30 year lifetime
Unofficially 30 y

Manufacturing cost
$10.76/m2 ($1/ft2)
1992 Cost Goal
Adjusted for inflation to
$15.46/m2 ($1.44/ft2)
Structural (self-supporting)
mirror to $27/m2 ($2.50/ft2)

Technical Approach

Samples supplied by:

Optical Characterization:

Industry
Subcontracts
Developed in-house

Perkin-Elmer (PE) Lambda 9 & 900 UV-VIS-NIR

spectrophotometers (250-2500 nm) w/ integrating
spheres
PE IR 883 IR spectrophotometer (2.5-50 m)
Devices & Services (D&S) Field Portable
Specular Reflectometer (7, 15, & 25-mrad cone
angle at 660 nm)

Outdoor (OET) & Accelerated Exposure

Testing (AET):

Atlas Ci65 & Ci5000 WeatherOmeters (WOM) (1X

& 2X Xenon Arc/60C/60%RH)
QPanel QUV (UVA 340@ 290- 340 nm/ 4 h UV at
40 / 4 h dark at 100%RH)
1.0 & 1.4 kW Solar Simulators (SS) ( 5X Xenon
300-500 nm. 1.0-kW SS 80C/ 80% RH,1.4 kW-SS4 quadrants 2 RH &T, light /dark)
BlueM damp heat (85C/85%RH/dark)
3 meterologically monitored sites at Golden,
Colorado (NREL), Miami, Florida (FLA), and
Phoenix, Arizona (APS)

3
2
1

Parabolic Trough Glass Mirror

Architecture
Low-iron Glass (4- or 5-mm thick)
Reflective Layer (wet-silver)
Back Layer (Cu)
1st coat Paint Layer (heavy Pb)
(2.5% Pb)
2nd coat Paint Layer (heavy Pb)
(1% Pb)

Acrylic (w/ high UV stability)

Thick glass is slumped
Flabeg mirrors still use Cu back protection
Three-coat paint system designed for outdoor applications
Mactac adhesive
Ceramic pad

Parabolic Trough Glass Mirror

Architecture
Low-iron Glass (4- or 5-mm thick)
Reflective Layer (wet-silver)
Back Layer (Cu)
1st coat Paint Layer (heavy Pb)
(2.5% Pb)
2nd coat Paint Layer (heavy Pb)
(1% Pb)

Acrylic (w/ high UV stability)

Thick glass is slumped
Flabeg mirrors still use Cu back protection
Three-coat paint system designed for outdoor applications
Mactac adhesive
Ceramic pad

Original Flabeg Mirror

Equivalent NREL Exposure Time (years)

% Hemispherical Reflectance

100

12

15

18

21

24

30

40

50

60

70

80

95

90

APS
FLA
NREL
Ci65

85
0

10

- OLD
- OLD
- OLD
- OLD
20

Total UV Dose (100 x MJ/m2)

Original vs. New Flabeg Mirror

% Hemispherical Reflectance of Old Flabeg (w/Cu & Pb paint) vs New Flabeg (w/ Cu & low-Pb
paint) Mirrors as a function of accelerated exposure in Ci65 WOM (65C/65%RH/~3sun light
exposure) and BlueM (85C/85%RH/dark), and outdoors in Colorado
100

% Hemispherical Reflectance

95

90

SWV SWV SWV SWV SWV SWV -

85

Old - BlueM
New - BlueM
Old - Ci65
New - Ci65
Old - NREL
New - NREL

80
0

12

16

Exposure Time (Months)

20

24

Alternate Thick Glass Mirror

Architecture
Low-iron Glass (3- or 4-mm thick flat)
Reflective Layer (wet-silver)
Back Layer (Cu-less)
1st coat Paint Layer (lead-free <0.15% Pb)
2nd coat Paint Layer (lead-free <0.15% Pb

Adhesive (PS, spray)

Substrate (SS, Al)

Alternate Thick Glass Mirror

100 0

Equivalent NREL Exposure Time (years)

2

10

11

12

13

14

15

16

% Hemispherical Reflectance

95

90

85

Pilkington: 4-mm glass

copper-free mirrors

80

Spanish: Cristaleria
Espanola S.A. (Saint
Gobain) 3-mm glass,
copper-free, lead-free
paint mirrors

NREL - Pilkington
NREL - Spanish
Ci65 - Pilkington
Ci65 - Spanish

75

70

65
0.0

3.3

6.7

10.0

13.3

16.7

20.0

23.3

26.6

30.0

33.3

36.6
2

Total UV Dose (100 x MJ/m )

40.0

43.3

46.6

50.0

53.3

Effect of Adhesive on Thick Glass Mirror

NREL Exposure Time (years)

% Hemispherical Reflectance

100 0

95

90

85

80
0.0

SPA/GE TSE 392-C ADH

PIL/GE TSE392-C ADH

SPA/GE D1-SEA210B

Pil/GE D1-SEA210B ADH

SPA/KRAFF SILKRAF ADH

PIL//KRAFF SILKRAF ADH

SPA/DOW Q3-6093 ADH

PIL/DOW Q3-6093 ADH

3.3

6.7

10.0

Total UV Dose (100 X MJ/m2)

13.3

16.7

Effect of Adhesive on Thick Glass Mirror

Equivalent NREL Exposure Time (years)

% Hemispherical Reflectance

100 0

95

10

11

12

13

SPA/GE TSE 392-C ADH

PIL/GE TSE392-C ADH

SPA/GE D1-SEA210B ADH

PILS/GE D1-SEA210B ADH

SPA/KRAFFT SILKRAF ADH

PIL/KRAFFT SILKRAF ADH

SPA/DOW Q3-6093 ADH

PIL/DOW Q3-6093 ADH

14

90

85

80
0.0

3.3

6.7

10.0 13.3 16.7 20.0 23.3 26.6 30.0 33.3 36.6 40.0 43.3 46.6

Total UV Dose (100 X MJ/m2)

Thin Glass Mirror Architecture

Low-iron Glass (~1 mm- thick)
Reflective Layer (wet-silver)
Back Layer (Cu)
(Cu-less)
Paint Layer (Pb)
(Pb-free)
Adhesive (PS, spray)
Substrate (SS, Al)
Thin glass mirrors are designed for indoor applications.

Thin Glass Corrosion

Thin Glass Mirror Matrix

D-optimal fractional factorial algorithm using Design-Expert software
Factors
Levels

Mirror
Type

Back
Protection

Adhesive /
Substrate

Edge
Protection

Substrate
Cleaning

Back
Priming

Naugatuck/Cu

Epoxy

3M504FL/AL steel

None

SAIC

3M

Naugatuck/ No Cu

Polyurethane

3M504FL/AL

Exuded Adh.

SES

None

Glaverbel

None

3M966/AL steel

CPFilm

3M966/AL

Mactac/AL steel

Mactac/AL

Epoxy/AL steel

Epoxy/AL

Urethane /AL steel

10

Urethane /AL

11

Contact /AL steel

12

Contact /AL

13

None

B: Back Protect

100

60

50

40

C: Adh/SS

None

Contact/Al

Contact/Al steel

Urethane/Al

Urethane/Al steel

Epoxy/Al

Epoxy/Al steel

Mactac/Al

Mactac/Al steel

30

3M 966/Al

B1 Epoxy
B2 Polyurethane
B3 None
3M 966/Al steel

Epoxy-based adhesive
probably good choice
No additional back
protection - survive the
longest
Polyurethane poor choice
BlueM - more accelerated
exposure chamber

70

3M 504FL/Al

Commercial vs. prototype

1- vs. 2-coat paint system
Difference in EU and US lead-free
regulations

80

3M 504FL/Al steel

Glaverbel - best overall

mirror in Mirror matrix test

90

Reflectance (SolWt)

ANOVA Analysis

Actual Factors
A: Mirror = Glaverbel D: Test method = Ci5000

Damp-Heat results similar but ~6X

faster than Ci5000
Discontinued in
Damp-Heat 5.9 MO

Discontinued in
Ci5000 18.9 MO

Thin Glass Mirror

Spectral Reflectance of Naugatuck copperless mirrors with 1 coat paint system after
accelerated exposure in Blue M (dark / 85oC / 85%RH) chamber
100

% Reflectance

80

60

1-coat paint system formulated for Cu free

mirrors.

0.0 MO
1.72 MO
2.70 MO

40

3.99 MO
5.24 MO
7.51 MO

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

Thin Glass Mirror

100

% Reflectance

80

60

(Naug/Clearcoat/966

0.0 MO
BlueM 3.52 MO
BlueM 7.21 MO
Ci65 3.16 MO
Ci65 6.15 MO
NREL 3.82 MO
NREL 9.57 MO

40

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

Aluminized Reflector Architecture

Protective Overcoat
Protective Oxide Topcoat
Enhanced Al Reflective Layer
Polished Aluminum Substrate

Aluminized Reflectors
NREL Exposure Time (y)

% Hemispherical Reflectance

100 0

Original
Improved Miro2
Improved Miro2 Set#2
Miro/4270kk

95

90

85

80
0

333

666

999

1332

1665
2

Total UV Dose (MJ/m )

1998

2331

Aluminized Reflector Specularity

FLA 11.8 m

NREL 11 m

APS 27.7 m

WOM 10.2 m

7-mradian Specular Reflectance at 660 nm

Equivalent NREL Exposure Time (years)

100 0

APS
FLA
NREL
Ci65

80

60

40

20

0
0.0
3.3
Alanod
4270/kk

6.7

10.0

13.3

16.7

Total UV Dose (100 x MJ/m2)

20.0

23.3

26.6

Aluminized Reflector
Spectral Reflectance of Alanod MiroSun mirrors after outdoor exposure in Phoenix, AZ at APS
100

% Reflectance

80

60

0.0 MO
10.72 MO

40

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

Aluminized Reflector
Spectral Reflectance of Alanod MiroSun mirrors after outdoor exposure in Miami, FL at FLA
100

% Reflectance

80

60

0.0 MO
12.0 MO

40

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

Aluminized Reflector
Spectral Reflectance of Alanod MiroSun mirrors after outdoor exposure in Golden, CO at
NREL
100

% Reflectance

80

60

0.0 MO
5.16 MO
11.37 MO

40

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

Aluminized Reflector
Spectral Reflectance of Alanod MiroSun mirrors after accelerated exposure in Ci65 WOM
(1 sun / 60oC / 60%RH) chamber
100

% Reflectance

80

60

0.0 MO
3.51 MO
6.63 MO
11.54 MO

40

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

Aluminized Reflector
Spectral Reflectance of Alanod MiroSun mirrors after accelerated exposure in Blue M (dark /
85oC / 85%RH) chamber
100

% Reflectance

80

60
0.0 MO
4.47 MO
6.66 MO
9.41 MO
13.84 MO
16.95 MO

40

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

Aluminized Reflector
Specular Reflectance at 7- and 25-mradians at 660 nm of Alanod MiroSun mirrors after
accelerated exposure in Blue M (dark / 85oC / 85%RH), WOM (1 sun / 60oC / 60%RH)
chambers, and outdoor exposure at NREL, APS, FLA, and Sandia

100

90

% Specular Reflectance

80

70

NREL - 25 mr
NREL - 7 mr
NREL - SWV
APS - 25 mr
APS - 7 mr
APS - SWV
FLA - 25 mr
FLA - 7 mr
FLA - SWV
WOM - 25 mr
WOM - 7 mr
WOM - SWV
Blue M - 25 mr
BlueM - 7 mr
BlueM - SWV
Sandia -25mr

60

50

40

30
0

12
EXposure Time (Months)

15

18

21

24

ReflecTech - Silvered Polymer

Reflector Architecture
UV-Screening Superstrate
Bonding Layer
Base Reflector
Flexible Polymer Substrate

ReflecTech Prototypes
Equivalent NREL Exposure Time (years)

% Hemispherical Reflectance

100

95

UV-Screen/SS95-NREL
ReflecTech A-NREL
ReflecTech B-NREL
UV-Screen/SS95-WOM
ReflecTech A-WOM
ReflecTech B-WOM

90

85

80

75

70
0

3.3

6.6

9.9

13.2

16.5

Total UV Dose (100 x MJ/m2)

19.8

23.1

26.4

ReflecTech III -NREL

Spectral Reflectance of ReflecTech pilot-run#3 (06-48) silver polymer mirrors after outdoor
exposure in Golden, CO at NREL

100

% Reflectance

80

60

ReflecTech (06-48)

40

0.0 MO
3.82 MO

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

ReflecTech III -NREL

Spectral Reflectance of ReflecTech pilot-run#3 (06-60) silver polymer mirrors after outdoor
exposure in Golden, CO at NREL

100

% Reflectance

80

60

0.0 MO
3.82 MO
ReflecTech (06-60)

40

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

ReflecTech III Ci65 WOM

Spectral Reflectance of ReflecTech pilot-run#3 (06-48) silver polymer mirrors after

accelerated exposure in Ci65 (1 sun / 60oC / 60%RH) chamber

100

% Reflectance

80

60

0.0 MO
3.16 MO
6.15 MO

ReflecTech (06-48)

40

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

ReflecTech III Ci65 WOM

Spectral Reflectance of of ReflecTech pilot-run#3 (06-60) silver polymer mirrors after

accelerated exposure in Ci65 (1 sun / 60oC / 60%RH) chamber

100

% Reflectance

80

60

ReflecTech (06-

40

0.0 MO
3.16 MO
6.15 MO

20

0
250

500

750

1000

1250

1500

Wavelength (nm)

1750

2000

2250

2500

Front Surface Solar Reflector

Architecture
Top Protective Layer (1-4m Al2O3)

IBAD Al2O3

Front Surface Solar Reflector

Architecture
Top Protective Layer (1-4m Al2O3)

Reflective Layer (100 nm Ag)

Substrate (PET)

Front Surface Solar Reflector

Architecture
Anti-soiling Layer (100 nm TiO2)
Top Protective Layer (1-4m Al2O3)
Adhesion Promoting Layer (APL) (1-10 nm)
Reflective Layer (100 nm Ag)
Metal Back Layer (30 nm Cu optional)
Substrate (PET)

(Chrome Plated Steel,

Leveled Stainless Steel,
or Aluminum)

Outdoor exposure at NREL of

Roll-Coated IBAD Al2O3 Samples
NREL Exposure Time (y)

% Hemispherical Reflectance

100

95

Both adhesion-promoting interlayer and

Ti backlayer were among most durable samples but:
Adhesion layer may slightly improve durability
Ti backlayer may slightly degrade durability

90

2um Al2O3/Ag/Cu
13AUG02-3 20nm/s
3.5um Al2O3/Ag/Cu
2AUG02 20nm/s
1.5um Al2O3/PL/Ag/Cu 13AUG02-1 20nm/s
4.5um Al2O3/PL/Ag/Cu 15AUG02-1 20nm/s
Al2O3/Ag/Ti
27AUG02 20nm/s
Batch
1 nm/s

85

80
0.0

3.3

6.7

10.0

Need more
exposure time to
determine lifetime

13.3
2

Total UV Dose (100 x MJ/m )

16.7

20.0

Cost Analysis
1 zone
2 zones
3 zones
1992 Cost Goal
Cost Goal in 2004$
1 zone Ann. Prod.
2 zones Ann. Prod
3 zones Ann. Prod.

40

3.0
2.5
2.0

30
1.5
20
1.0
10

0.5

0
0

50

100

150

200

250

Al2O3 Deposition Rate (nm/s)

0.0
300

Annual Production (1x10 6 m 2)

Total Production Costs ($/m )

50

30% yield
Coating 79%
time
10 to 200 nm/s
rate
Machine cost:
$2M-$4.1M

Loan%/length:
12% for 5 yrs
PET substrate
1-m Al2O3
Modified ASRM
$200/h machine
burden
1200-mm web
High-purity
High-volume
(i.e.,$200/kg)
Al2O3
1 vs. 2 vs. 3
zones in 1
machine

 Receivers:

Field Requirements for

Advanced Receivers

4 m (13.1 ft) long

70 mm (2.25 in) diameter
New 64 MWe Nevada plant
820 collectors and each
collector has 24 (96 m)
receivers
19,680 receivers
82 km of receivers (50 mi)

Existing SEGS plants have

5x this many receivers
New 553 MW plant will
need 8.5x this many
receivers
3-4%/yr Failure Rate
~$1000/tube

Advanced Selective Coating Goals

4.0

Ideal Solar Selective

100

500C (3.56 kW/m )

3.0
2.5

80

60

2.0

2
400C (1.78 kW/m )

40

1.5
Direct AM 1.5 (0.77 kW/m2)

1.0
300C (0.80 kW/m2)

0.5
0.0
100

20

200C (0.31 kW/m2)

1000

10000

Wavelength (nm)

0
100000

% Reflectance ()

Blackbody Irradiance (W/m2-nm)

3.5

Advanced Selective Coating Goals

Good optical and

thermal performance:
absorptance () 96%,
& emittance () 7%
>400C
High temperature
stability in air at
temperatures 550C
Manufacturing
processes with
improved quality control
Lower cost

Ideal Solar Selective

100

3.5
500C (3.56 kW/m2)

3.0
2.5

80

60

2.0

400C (1.78 kW/m2)

40

1.5
2

Direct AM 1.5 (0.77 kW/m )

1.0
300C (0.80 kW/m2)

0.5
0.0
100

20

200C (0.31 kW/m2)

1000

10000

Wavelength (nm)

0
100000

% Reflectance ()

coatings that have:

4.0

Blackbody Irradiance (W/m2-nm)

To develop receiver

High Temperature Solar

Selective Coating Development

Selective coating
properties impact
collector optical
performance and
thermal losses.
Improvements in
the receiver can
enhance collector
efficiency & lower
cost.
The international
community
currently leads this
area and there
exists minimal US
research & no US
manufacturer of
high-temperature
selective coatings.

0.15

Levelized Cost of Energy

0.14

= 0.25

0.13
0.12
0.11

Reduced
Thermal
Losses
(lower )

Black Chrome

Mo-Cermet

= 0.15

UVAC (t)

= 0.05

Increased Optical Properties

(higher )

0.10
0.80

0.82

0.84

0.86

New
Cermet (s)

UVAC (s)
Goal

0.88

0.90

Energy Absorbed by Receiver

0.92

Types of Selective Coatings

Intrinsic selective
material

Intrinsic absorber

Substrate
Dielectric
Metal
Dielectric
Metal
Substrate

Double cermet

Multilayer absorbers

AR
AR
AR
AR

Metal

Surface texturing

LMVF cermet

Substrate

HMVF cermet
LMVF cermet

Graded metal
dielectric composite

HMVF cermet
Metal

Metal
Substrate

Graded cermet

AR
LMVF cermet
HMVF cermet
Metal
Substrate

Substrate

Multiple cermet

Literature Review of Candidate

High-temperature (> 400C) Solar
Selective Materials

Graded Mo,W, ZrB, Pt- Al2O3 cermets

Si tandem absorber
Black Co, Mo,W

Double cermets- SS-AlN, AlN/Mo, or AlN/W

4-layer V-Al2O3, W-Al2O3, Cr-Al2O3, Co-SiO2, Cr-SiO2, NiSiO2
Double AR
Multilayers; Al-AlNx-AlN
Au/TiO2 cermet
ZrCxNy/Ag
Ti1-xAlxN
Quasicrystals multilayers & cermets
Surface Texturing

Desirable Properties for Stable

Coating in Air > 400C

High thermal & structural stabilities for combined & individual layers
Elevated melting points
Large negative free energies of formation
Materials that form a multicomponent oxide scale
Single-compound formation

Lack of phase transformations at elevated temperature

Suitable texture to drive nucleation, subsequent growth of layers with suitable

morphology
Stable nanocrystalline or amorphous materials

Excellent adhesion between the substrate and the adjacent layers

Enhanced resistance to thermal and mechanical stresses
Acceptable thermal and electrical conductivities
Higher-conductivity materials have improved thermal shock resistance

Some ductility at room temperature reduces thermal-stress failures

Good continuity and conformability over the tube

Compatibility with fabrication techniques

NREL Modeled Selective Coating

Comparison of theoretical optical properties for NRELs modeled prototype solar selective
coating with actual optical properties of existing materials.

Commercial (as tested)

Solar
Absorptance

Modeled

Black Cr

MoCermet

UVAC

# 6A

# 6B

0.916

0.938

0.954

0.959

0.950

Thermal Emittance@
25C

0.047

0.061

0.052

0.013

0.027

100C

0.079

0.077

0.067

0.017

0.033

200C

0.117

0.095

0.085

0.028

0.040

300C

0.156

0.118

0.107

0.047

0.048

400C

0.216

0.146

0.134

0.074

0.061

500C

0.239

0.179

0.165

0.110

0.073

Modeled NREL Selective Coating

1.00

Reflectance & E (norm.)

ASTM G173-03
AM 1.5
0.80 Direct Normal

Ideal
Selective
Coating

UVAC A
Advanced
NREL #6A

0.60

400C Black Body

0.40

0.20

0.00
0

10

Wavelength, um

100

Modeling Key Results

Solar Selective Coating Development

Modeled solar-selective coatings with =0.959 and
=0.061 that meet CSP goals
Emittance excellent & absorptance of modeled
coatings is very good but further improvements are
expected. However, trade-off exists between
emittance and absorptance.

Deposition Capabilities

Three-Chamber In-line System

Load-Lock Chamber
Pulsed DC Sputtering Chamber

3 - linear arrays of 5 - 1.5 Mini-mak

guns
2 - 12 planar cathodes

Electron-Beam/IBAD Chamber
6 multi-pocket e-beam source
Co-deposition bottom plate
IBAD w/ 12 Linear Ion Gun

System

12x12 ambient or heated substrate

4 Reactive Gases
Turbo molecular drag pumps
2x10-8 torr

Monitoring

RGA
Quartz Crystal Monitor
Pressure/Gas
Computer

Prototyping Key Results

Key issue is making deposited coating

XPS showed evaporation from compounds produced
layered stoichiometry

Despite depositing layers with over- and under-thickness and

compound layered structure, the optical performance of the
prototype NREL#6A was quite encouraging.

Need to codeposit materials

Required significant upgrade to equipment

Installed codeposition guns & sweeps

Pneumatic shutters
Second quartz crystal sensor
Upgrade computer & RGA software
+ associated air, water, & electrical
Automating control

Prototyping Key Results

Codeposit individual layers and modeled coating
Codeposition development
Deposited individual layers
Deposited modeled structure
Characterize properties
Optical performance lower than modeled
Typically optical coating need error <1%
Thickness error was >5% because of manual control
Install optical monitor
Provide positive feedback between quartz crystal and optical
monitor
Automate control remove human error and provide steering
and cutting at sensitive turning points allowing mid-course
corrections to be made
Compositional errors because stoichiometry not optimized
Composition with highest reflectance
Phase formation from Pretorius effective heat of formation
model & TGA
Optimize morphology with ion assist

Selective Coating Performance

can be measured at higher temperatures but is typically reported based on
calculations from reflectance measurements fitted to the black body curve
Actual performance of the absorber at high temperatures commonly does
not correspond to the calculated
Small errors in lead to large errors in
is a surface property & depends on surface condition of material and substrate
Surface roughness
Surface film
Oxide layers

Selective coatings can degrade at high T due to

Thermal load (oxidation)

High humidity or water condensation on the absorber surface (hydratization and
hydrolysis)
Atmospheric corrosion (pollution)
Diffusion processes (inter-layer substitution)
Chemical reactions
Poor interlayer adhesion

Therefore it is important that is measured accurately and to measure of

the selective coating at operating temperatures & conditions before using
calculated

Round Robin &

Purchase Perkin Elmer 883 IR spectrophotometer

Thermal Stability

Thermal stability is sometimes given based on the thermal

properties of the individual materials or the processing
temperature parameters
Actual durability data is uncommon for high temperature
absorber coatings
Durability or thermal stability is typically tested by heating the
selective coating, typically in a vacuum oven but sometimes
in air, for a relatively short duration (100s of hours)
compared with the desired lifetime (5-30 years)
IEA Task X performance criterion (PC) developed for flat plate
collector absorber testing (i.e., non-concentrating, 1-2X sunlight
intensity)
No analogous criterion known for testing high-temperature selective
coatings for CSP applications

Building capability for long term testing of thermal stability

Purchased & installed high-temperature

(600C) inert gas oven

Conclusion

DOE, the WGA, state RPS mandates, and feed-in tariffs

have successfully jump-started growth in CSP technologies
that would require 7 to 10 million square meters of reflector
and more than 600,000 HCEs over the next 5 years.
Commercial glass mirrors, Alanod, and ReflecTech may
meet the 10-yr lifetime goals based on accelerated
exposure testing. Predicting an outdoor lifetime based on
accelerated exposure testing is risky because AET failure
mechanisms must replicate those observed by OET.
Experimental IBAD Al2O3 front surface mirror has high
potential to meet need; but needs development by rollcoating company
None of the solar reflectors available have been in test long
enough to demonstrate the 10-year or more aggressive 30year lifetime goal, outdoors in real-time

Conclusion

Modeled solar-selective coatings with =0.959 and

=0.061 that meet CSP goals
Emittance excellent & absorptance of modeled coatings
is very good but further improvements are expected.
However, trade-off exists between emittance and
absorptance.
Key issue then becomes trying to make the coating
Prototype development underway. Individual and
modeled structure deposited by e-beam compound and
elemental codepostion & characterized. Need to
eliminate thickness errors by upgrading monitor and
control and determine optimum stoichiometry.
Purchased & installed PE 883 IR Spectrophotometer
(2.5- 50) and high-temperature inert gas oven. Roundrobin data being analyzed and commercial & prototyped
coating samples being put into test
Patent being pursued

Acknowledgments
Alanod, Glaverbel, Naugatuck, ReflecTech, SAIC, and SES for providing
solar reflective samples.
Schott and Solel for providing solar selective samples.
AZ Technology and Surface Optics Corporation for high-temperature optical
measurements .
Armstrong World Industries: Dr. J. S. Ross
Northeastern University:
Dr. Jackie Isaacs
Penn State University:
Prof. Singh, Tom Medill, and Dale Donner
SAIC:
Dr. Russell Smilgys and Steve Wallace
Stat-Ease:
Wayne F. Adams
Swisher and Associates:
Dick Swisher
NREL:
Lynn Gedvilas, Gary Jorgensen, Mark Mehos, Judy Netter, Craig
Perkins, Hank Price, Kent Terwilliger, and Student interns: Micah
Davidson, Anthony Nelson, Michael Milbourne, and Christopher, and
Andrea Warrick.
DOE supported this work under Contract No. DE-AC36-99GO10337.