is it possible to design accelerated service life tests for pv … · 2015. 9. 14. · m.koehl...
TRANSCRIPT
© Fraunhofer ISE
Is it possible to design accelerated servicelife tests for PV modules?
Michael Köhl
Fraunhofer Institute for Solar Energy Systems
Freiburg, Germany
Presented at the EMPA Workshop
Durability of Thin Film Solar Cells
April 2011
© Fraunhofer ISE
General methodology for service life assessment
Assumptions
1. Only long-term wear out degradation is considered
2. The primary degradation factors are due to weathering
3. The stress-levels depend on local climate and installation
4. The stress-levels depend on the micro-climate at the module
5. The test samples (PV-modules or components) have to be considered as a black-box
6. The modelling is based on investigation of the degradation kinetics of real state-of-the-art modules
7. A service life of 25 years is required
© Fraunhofer ISE
General methodology
Local climate
PV-modules
Micro-climateMaterials
ALT conditions
Service life
Modeling the Accelerated Life Test conditions based on realistic loads
Models
Degradation
Performance change
© Fraunhofer ISE
General methodology Step 1: Outdoor exposure and climate monitoring
City or reference:
Freiburg Germany
Desert
Sede Boqer
Israel
Alpes
Zugspitze
Germany
Tropical
Serpang
Indonesia
Maritime
Pozo Izquierdo
Gran Canaria
Monitoring degradation factors for modelling degradation
Measurement of module performance over time for validation of ALT
© Fraunhofer ISE
Monitoring in Cologne (one year)
0 50 100 150 200 250 300 3500
20
40
60
80
100
RH
[%]
Zeit [d]0 50 100 150 200 250 300 350
-20
-10
0
10
20
30
40
50
60 T Modul T ambient
Tem
pera
tur [
°C]
Zeit [d]
Ambient humidity ambient and module temperatures
© Fraunhofer ISE
Monitoring in the Alpes (one year)
0 50 100 150 200 250 300 3500
20
40
60
80
100
RH
[%]
Zeit [d]0 50 100 150 200 250 300 350
-30
-20
-10
0
10
20
30
40
50
60 T Modul T ambient
Tem
pera
tur [
°C]
Zeit [d]
Ambient humidity ambient and module temperatures
© Fraunhofer ISE
0 50 100 150 200 250 300 3500
20
40
60
80
100
RH
[%]
Zeit [d]0 50 100 150 200 250 300 350
-20
-10
0
10
20
30
40
50
60
70 T Modul T ambient
Tem
pera
tur [
°C]
Zeit [d]
Ambient humidity ambient and module temperatures
Monitoring in the desert (one year)
© Fraunhofer ISE
Ambient humidity ambient and module temperatures
0 50 100 150 200 250 300 3500
20
40
60
80
100
RH
[%]
Zeit [d]0 50 100 150 200 250 300 350
10
20
30
40
50
60
70 T Modul T ambient
Tem
pera
tur [
°C]
Zeit [d]
Monitoring in tropical Serpong (one year)
© Fraunhofer ISE
General methodology Step 2: Micro - climate
Micro-climate : = stresses for the material caused by interaction of materials and ambient climate
1. Module temperature modeled by solar irradiation, ambient temperature, wind speed
2. Module surface humidity modeled by module temperature, ambient temperature and humidity
3. UV-radiation modeled from solar irradiation and spectral transmittance of laminated materials
4. Temperature cycles of module temperature
5. Leakage current as function of voltage, module temperature and surface humidity
6. Salt concentration correlated with wetting/drying cycles
7. …….
© Fraunhofer ISE
Micro-climate of modules
Module – Temperature in the desert
Day
Night
© Fraunhofer ISE
-20 -10 0 10 20 30 40 50 60 700
200
400
600
800
1000
1200
1400
1600 tropisch arid gemäßigt alpin
freq
uen
cy [
h]
Temperature / °C
Outdoor weathering with temperature monitoring
Ambient temperature Average module temperatur e(c-Si)
Macro-climate => Micro-climate
-20 -10 0 10 20 30 40 50 60 700
200
400
600
800
1000
1200
1400
1600
freq
uen
cy [
h]
Temperature / °C
tropical arid moderate alpine marine
Micro-climate of modules
© Fraunhofer ISE
Histogram of measured module temperatures in Cadarache, F
-20 0 20 40 60 80
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
coun
t
temperature in °C
Avancis Flexcell FirstSolar Schott Sharp UniSolar Wuerth SchottCSIRef
Cadarache thin film moduleshistogramm of measured module temperature
Special thanks to Antoine Guerin de Montgareuil,
CEA-INES, Cadarache, France
a-Si 1 CIS 1 c-Si
a-Si 3 CIS 2
a-Si 4 CdTe
What about Thin Film Modules?
Module temperature for one year
© Fraunhofer ISE
Physical modeling of module temperature for each of the different modutypes using David Faiman‘s approach (could be King, Fuentes……as well)
v10mod
UU
HTT amb
Tmod module temperature
Tamb ambient temperature
v wind velocity
H solar radiation
U0 , U1 = module dependent parameters
Neglected: IR-radiation exchange and natural convection
U1 U0a-Si 1 10,7 25,7a-Si 3 5,8 25,8a-Si 4 4,3 26,1
CIS 1 3,1 23,0CIS 2 4,1 25,0CdTe 5,4 23,4c-Si 6,2 30,0
The parameters U are module-specific but location independent
M.Koehl et.al.: Modelling of the nominal operating cell temperature based on outdoor weathering, Sol. Energy Mat. Sol. Cells (2011)
Macro – climate => Micro – climate
Irradiation, wind, ambient temperature => Tmod
© Fraunhofer ISE
0 20 40 60 800
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
Freu
ency
Temperature in °C
a-Si 1 a-Si 2 a-Si 3 a-Si 4 CIS 1 CIS 2 CdTe c-Si
Histogram of simulated module temperatures for one year in the Negev
c-Si CIGS
Micro-climate of Modules
© Fraunhofer ISE
0 10 20 30 40 50 600
10
20
30
40
50
60
Daily module temperature difference /K
Gran Canaria 2011
0 10 20 30 40 50 600
10
20
30
40
50
60
Prob
abili
ty d
ensi
ty
Daily module temperature difference /K
Negev 2009
0 10 20 30 40 50 600
10
20
30
40
50
60
Daily module temperature difference /K
Cologne 2009
0 10 20 30 40 50 600
10
20
30
40
50
60
Prob
abili
ty d
ensi
ty
Daily module temperature difference /K
Indonesia 2007
Micro-climate of modules
Daily temperature cycling during one year
0 10 20 30 40 50 600
10
20
30
40
50
60
Prob
abili
ty d
ensi
tyDaily module temperature difference /K
Alpes 2008
Type approval testing:
200 cycles from -40 to 85°C
Max. changes 100 K/h
With current injection
© Fraunhofer ISE
Surface humidity – desert
Night-time = condensation
Day-time = Drying
Micro-climate of modules
© Fraunhofer ISE
00:00 04:00 08:00 12:00 16:00 20:00 24:000,1
1
10
100
1000
10000
I leak
age
/ nA
Time / hh:mm
M17 - sunny day M18 - sunny day M17 - rainy day M18 - rainy day
Micro-climate of modules
Leakage currents as source of Potential Induced Degradation
Leakage currents depend on the module temperature, the voltage and the surface humidity
© Fraunhofer ISE
Leakage current as function of potential, relative humidity and temperature
I =G/(1+ exp(-G*(rh-0,35)*1,75) *(G/f(0)-1))with G = 13μA
Humidity and potential-induced degradation (PID) of modules
0 100 200 300 400 500 6000
5
10
15
20
25
30
25-25 55-25 85-25 85-40 85-55 85-70 85-85
I/µA
U / V
T[°C] rh [%]
0
2
4
6
8
10
12
14
0 20 40 60 80 100
Leak
age
curr
ent /
µA
Rel. humidity / %
Hoffmann, S., M. Koehl, Effect of Humidity and Temperature on the Potential Induced Degradation, accepted by PIP, 2012
Micro-climate of modules
© Fraunhofer ISE
General Methodology Step 3: Time-transformation functions
Time-transformation functions
1. Module temperature: Arrhenius, Eyring
2. Module surface humidity impact: power law, TOW
3. UV-radiation: Dose-function, reciprocity?
4. Temperature cycles of module temperature: Coffin-Manson
5. Potential induced degradation
6. Salt concentration correlated with wetting/drying cycles: ?
7. …….
Modelling of the degradation processes as function of the degradation factors
© Fraunhofer ISE
Time-transformation functions
Changes of performance or degradation indicator P = ti *(
Temperature + A exp[-EA /RTi]
Humidity + B f(rh)i exp[-EB /RTi]
UV-radiation + C Ini exp[-EC /RTi]
Temperature cycles + D f(T)i exp[-ED /RTi]
Potential-induced Deg. + E f(P)i fp(rh)i exp[-EE /RTi]
Salt + F f(S)i fp (rh)i exp[-EF /RTi]...........)
© Fraunhofer ISE
Degradation factor:
Temperature
Moisture
UV-Radiation
T cycles
Potential I D
Salt
………
Simple deterministic model for aging processes: Time-transformation functionsChanges of property P after the testing time ti
Sample dependent degradatioprocess parameters
P = mi=1 {ti (
+ A exp[-EA /RTi]
+ B f(rh)i exp[-EB /RTi]
+ C Ini exp[-EC /RTi]
+ D f(T)i exp[-ED /RTi]
+ E f(P)i fp(rh)i exp[-EE /RTi]
+ F f(S)i fp (rh)i exp[-EF /RTi]
+…… X Ini f(X) i exp[-EX /RTi] …..)}
© Fraunhofer ISE
ALT – conditions for different locations: temperature impact
Accelerated life testingEquivalent lab tests
(same changes of performanceor degradation indicator as after service life)
by integration of the outdoor stresses
Difference in testing time between 8 and 20
0 50 100 150 200100
1000
10000
100000
8000 h
2100 h
4200 hTe
stin
g tim
e at
85°
C fo
r 25
year
s in
h
Activation energy in kJ/mol
South France Tropic Arid Moderate, urban Alpine
16000 h
Temperature stress
General methodology Step 4: Accelerated life testing conditions
© Fraunhofer ISE
Corresponding temperature testing times at 85°C for 25 a exposure in Cadarache, France
based on monitored module temperatures
50 100 150 200 250 3001
10
100
1000
10000
100000
log
test
ing
time
[h]
activation energy [kJ/mol]
Avancis Flexcell First Solar Schott Sharp Unisolar Wuerth SchottCSIRef
testing times @ 85°C for different thin film modules exposed in Cadarache
Different cell-types:
Factor 2 – 4 in testing time(depending on the degradation processes)
In the range of damp/heat
7 different module types
© Fraunhofer ISE
2025 50 75 100 125 15010
100
1000
10000equivalent testing times for five suns UV test for desert climate loads
60 °C 70 °C 80 °C 90 °C
activation energy [kJ/mol]
test
ing
time
[h]
1700 h
5700 h
Cumulated dose of UV- and solar radiation:
120 kWh/m² (about 8 x IEC)
Outdoor testing with radiation monitoring for one year in the desert
0
1
2
3
4
5
6
-10 0 10 20 30 40 50 60 700
20
40
60
80
100
120 Global UV
glob
al ir
radi
atio
n pe
r tem
p.-b
in[k
Wh/
m²a
]
Module temperature [°C]
UV
dose
pro
tem
p.-b
in [k
Wh/
m² a
]
UV = 5.X % of solar radiation
Reciprocity: p = 1
ttest = (Ii / Itest)p ti exp [-(Ea / R)(1/Ttest - 1/Ti)]
ALT – conditions for different locations: UV-radiation impact
© Fraunhofer ISE
Effective surface humidity
rheff =1/(1+ 1000/exp(rh*k))
20 30 40 50 60 70 80 90 100 110 120 130 140100
1000
100002000 h
test
ing
time
@85
°C/8
5h fo
r 25
a se
rvic
e lif
e [h
]
activiation energy [kJ/mol]
Tropic Arid Alpine
2000 h
Humidity
M. Koehl et. al., Solar Energy Materials & Solar Cells 99 (2012) 282–291
ALT – conditions for different locations: humidity impact
00,10,20,30,40,50,60,70,80,9
11,1
0 0,5 1
Rel. surface humidity
© Fraunhofer ISE
Simulated histograms of the relative humidity
Ambient humidity = partial pressure / saturation pressure (Tamb)
Surface humidity = partial pressure / saturation pressure (Tmodul)
-20 0 20 40 60 801
10
100
1000
Goodwin Creek 2007-01-01-bis-2007-12-31_ (type cSi)histogram of the periods with high moisture
frequ
ency
dis
tribu
tion
in h
ours
module temperature in °C
rheff (sambient) rh (surface) rheff (surface) = 85%
-20 0 20 40 60 801
10
100
1000
rheff (sambient) rh (surface) rheff (surface) = 85%
Desert Rock 2007-01-01-bis-2007-12-31_ (type TF)histogram of the periods with high moisture
module temperature in °C
Humidity dose: teff = t * rheff / 0.85
Eff. Humidity: rheff = 1/(1+ exp(-rh*k) *(1/f(0)-1))
© Fraunhofer ISE
-500 0 500 1000 1500 2000 2500 3000 3500 4000 45000,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
1,1
1 2 3 4 5 6 7
Nor
mal
ised
pow
er
Exposure /h
Damp-Heat at 85°C and 85% rel. humidity
-500 0 500 1000 1500 2000 2500 3000 3500 4000 45000,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
1,1
1 2 3 4 5 6 7
Nor
mal
ised
pow
er
Exposure /h
General methodology Step 5: Accelerated life testing
Damp-Heat at 90°C and 85% rel. humidity
Seven different commercial c-Si modules
M. Koehl et. al., PV reliability (Cluster II): Results of a German four-year joint project - Part I, results accelerated ageing tests and modelling of degradation, 25th EU-PVSEC (2010)
© Fraunhofer ISE
-500 0 500 1000 1500 2000 2500 3000 3500 4000 45000,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
1,1N
orm
alis
ed p
ow
er
Testing time /h
Damp-Heat-Prüfung 85%rh @ 85°C
Damp-heat testing at 85%rh and 85°C, module 1
Induction phase
Degradation phase
Degradation function
© Fraunhofer ISE
-500 0 500 1000 1500 2000 2500 3000 3500 4000 45000,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
1,1
No
rmal
ised
po
wer
Testing time /h
2300 h3200 h
Service lifetime @ 0.8 Pmpp
Damp-heat testing at 85%rh and 85°C, module 1 and module 2
© Fraunhofer ISE
-500 0 500 1000 1500 2000 2500 3000 3500 4000 45000,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
1,1
No
rmal
ised
po
wer
Exposure time /h
Damp-heat tests at 85%rh@85°C and 85%rh@90°C (large dots), Module 1 and module 2
© Fraunhofer ISE
-500 0 500 1000 1500 2000 2500 3000 3500 4000 45000,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
1,1
Module 1 65 kJ/mol Module 2 33 kJ/mol
No
rmal
ised
po
wer
Exposure time /h
a = exp [-(Ea / R)(1/T1 – 1/T2)]
Damp-heat tests at 85%rh@85°C and 85%rh@90°C (large dots), Module 1 and module 2
© Fraunhofer ISE
20 40 60 80 100 120100
1000
10000
100000
test
ing
tim
e @
85°C
[h
]
Activation energy [kJ/mol]
tropical arid alpine
Equivalent testing times (25 years at 85%rh @ 85°C
Module 1 would be suitable for all climates
Module 2 would survive in the mountains
Service lifetime
© Fraunhofer ISE
-1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 150
5
10
15
20
25
30
35
40
45
50
55
60
65
70
Displacement on cell (cm)
Inte
nsity
(arb
. uni
ts)
Average Ref._I(1660) Average 4000h d.h._I(1660)
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4 Average Ref._I(2863)/I(2913) Average 4000h d.h._I(2863)/I(2913)
Comparison of Vinyl-Band (red) and fluorescence back-ground (black)initial data and after 4000h damp-heat testing
Polymer Analysis with Raman-Spectroscopy and cell analysis by electroluminescence
Elektroluminescence-image of degraded cells
Location on cell [cm]
Inte
nsi
ty [
ar. u
nit
s]
Peike et.al.: Non-destructive degradation analysis of encapsulants in PV modules by Raman Spectroscopy, Sol. En. Mat. Sol. Cells (20
General methodology Step 6: Analysis of materials degradation
© Fraunhofer ISE
0 1000 h
GlasEVA EVA SiBS
3 mm 0,5 mm
Numerical simulation of energy and mass transport
Water vapour permeation and -diffusion in the back-sheet and the Encapsulant during damp-heat testing (85%rh @85°C)
General methodology Step 7: Simulation of materials degradation
© Fraunhofer ISE
time/d Tropic (hot and humid)
Between cell and glass
Behind the cell
time [d]
Simulation with real climate data
Alps (cold and dry)
Water vapour permeation and - diffusion in the Back-sheet and in the Encapsulant during damp-heat testing (85%rh @85°C)
time [d]
Numerical simulation of energy and mass transport
Wat
er c
once
ntra
tion
[rel
. uni
ts]
© Fraunhofer ISE
What happens with PV – modules in operation?
Degradation processes are induced or caused by transport phenomena
Time constant energy mech. tension charge mass
about zero light
seconds vibrations
minutes heat thermo-mech
hours to years ions oxygen, water vapo
years pollutants (salt, etc.
© Fraunhofer ISE
UV-source for combined UV and humidity tests
200-250 W/m²
2*3*1,6m²
Multiple stress testing
0 – 100°C
0 – 85%rF
Physical limits of combined testing
Testing with radiation requireslarge areas (expensive)
Concentration is needed foracceleration (4-5 X)
Highly efficient irradiation sourcesrequire lamp cooling
Water cooling limits the freezingat temperature cycling
Solar simulator (incl. VIS and NIR) causes heating of the samples
Cooling of the samples spoils high humidity by condensation at theheat exchanger
How to integrate mechanicalloads
General methodology Step 8: Service life testing
© Fraunhofer ISE
Summary
Time-series of climatic data ambient temperature and humidity, solar irradiation, wind speed
Modeling the module temperatures ambient temperature, solar irradiation, wind speed, module-specific coefficients (mounting situation might be considered)
Modeling the UV-radiation 5.5% of the solar radiation, module temperature
Modeling the effective surface humidity ambient temperature and humidity, module temperature
Modelling the micro-climatic stress conditions
© Fraunhofer ISE
Summary
Use a simple time-transformation function (Arrhenius based, eg)Time, module temperature and other degradation factors, but separately first
Modeling the module temperature stress as function of the material-specific activation energy,
(could be eventually included in damp-heat testing)
Modeling the UV-radiation impactas function of the material-specific activation energy (which is low, UV-dose more importan
Modeling the moisture test Higher test temperatures needed, as function of the material-specific activation energy,
Modelling the ALT conditions
© Fraunhofer ISE
Conclusions
One test:
Infant mortality, quality Ieststype approval testing acc. to IEC or UL
Enhanced stress testing:
Infant mortality, higher quality requirementsoffered by a number of test labs
Degradation over time:
Performance, materials or degradation indicator over time=> Changes of micro-climate (stress) because of material changesstability beyond infant mortality, induction periods, stress factor sensitivity
Single constant stress testing
Needed for service life testing:
Performance or degradation indicator over time until failure
© Fraunhofer ISE
Conclusions
Temperature cycling:
Thermo-mechanical stressNo scientific base for type approval testing acc. to IEC or UL
Which relaxation time at which temperature?
Temperature cycling with humidity:
Closer to reality, takes into account temperature dependence of water vapour permeation
Voltage cycling or UV-radiation cycling:
Dark periods allow recovery or diffusion of reactants
Single cyclic stress testing
Needed for service life testing:
Investigation of relaxation times and diffusion processesFrequency and amplitudes of dynamic mechanical testing
© Fraunhofer ISE
Conclusions
Reasons:
Material changes caused by a degradation process due to stress factor 1 might changethe micro-climate from stress factor 2
A combination of stress factors might cause new degradation processes(Photodegradation and hydrolysis, hydrolysis and corrosion)
Problems:
How to design life-tests for cdegradation changed micro-climatic stress?
How to define accelerated life tests with similar acceleration factors for all stress factors taking into account different time constants?
Multiple stress testing
Needed for service life testing:
A big number of unknown factors have to be determined:P = A ti exp[-EA /RTi] + B ti f(rh)i exp[-EB /RTi] + C ti Ini exp[-EC /RTi] + D ti f(T)i exp[-ED /RTi] + E ti f(P)i fp(rh)i exp[-EE /RTi] + Fpq ti f(Sp)i f (Sq)i exp[-Epq /RTi]
Experimental design for a respective number of tests at different stress levels
© Fraunhofer ISE
for your attentionThanksTo my colleagues
Daniel Philipp
Franz Brucker
Stefan Hoffmann
Philipp Huelsmann
Markus Heck
Stefan Brachmann
Karl-Anders Weiss
Stefan Wiesmeier
To our partners
TÜV Rheinland
Schott Solar
Solarfabrik
Solarwatt
Solarworld
Solon
© Fraunhofer ISE
Lugano
Switzerland
Mai 3 – 4 2012
Testing
Analysing
Simulating module - reliabilty
Meeting of the IEC TC82 WG2
After the Workshop in Stresa
Workshop on Reliability of PV-Modules
www.supsi.ch/go/pv-module-reliability
Organised by Fraunhofer ISE and SUPSI
© Fraunhofer ISE
Structure of the workshop
The topics will be presented by experts and further developed in small discussion groups.
Block I: Mechanics
Block II : PID -Humidity (Potential induced degradation)
Block III : UV –Humidity
Block IV : Failure modes and effects
.
Block V: Materials
Plenary discussion with presentation of discussion results
Optional: Visit of the outdoor exposure test site of ISAAC Supsi