understanding the temperature and humidity … · a pet based back-sheet will be exposed to...
TRANSCRIPT
NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.
Understanding the Temperature and Humidity Environment Inside a PV Module
2013 NREL PVMRW
Michael Kempe
Wednesday, February 27, 2013
NREL/PR-5200-58375
2
Introduction
• Many degradation processes within a PV module are driven by moisture.
• The concentration of moisture in a module is a complex function of the use environment and the module construction.
• In accelerated stress testing one must know how water affects degradation to determine what temperature and humidity conditions to use.
• Here we show that by choosing humidity conditions that more closely match the use environment, one can minimize the uncertainty associated with moisture induced degradation modes.
3
Outline
• Describe moisture on the backside of a module. • Look at the hydrolysis of a typical back-sheet
made of PET as a case study for comparing 85 ⁰C/85% RH to outdoor exposure.
• Examine the moisture and temperature environment on the front of a module as a worst case scenario.
• Show how good choices for RH testing will minimize uncertainty.
4
Representative Module Environment
• Use either IWEC or TMY-3 data for select environments.
• Use the model of King et al.* for module temperature.
• This produces “representative” data intended to generally duplicate a use environment
20
30
40
50
60
70
80
0 1 2 3 4 5
Mod
ule
Tem
pera
ture
(⁰C)
Days
Bangkok Thailand Module Temperature
Insulated Back, Glass/Polymer Close Roof, Glass/Glass Open Rack, glass/glass open Rack, Glass/Polymer
*D. L. King, W. E. Boyson, and J. A. Kratochvil, "Photovoltaic array performance model," SAND2004-3535, Sandia National Laboratories, Albuquerque, NM, 2004.
0102030405060708090
100
0 1 2 3 4 5Rela
tive
Hum
idity
Mod
ule
Surf
ace
(%)
Days
Bangkok Thailand Ambient Relative Humidity at Module Temperature
Insulated Back, Glass/Polymer Close Roof, Glass/Glass Open Rack, glass/glass open Rack, Glass/Polymer
5
Moisture in the Back-EVA Layer
• Assume diffusivity in EVA is much greater than in the back-sheet.
• Also assume transient moisture gradient in the back-sheet is unimportant.
𝑑𝐶𝐸𝑑𝑑
= 𝑊𝑉𝑉𝑉𝐵,𝑆𝑆𝑆𝐶𝐸,𝑆𝑆𝑆𝑙𝐸
𝐶𝐸,𝐸𝐸 − 𝐶𝐸
00.00020.00040.00060.0008
0.0010.00120.00140.00160.0018
0.002
0 1 2 3 4 5Abso
lute
Hum
idity
(g/c
m3 )
Days
Bangkok Thailand Module Back-EVA Absolute Humidity
Insulated Back,Glass/Polymer
Close Roof, Glass/Glass
Open Rack, glass/glass
open Rack, Glass/Polymer
Back-Sheet
Glass
H2O
Back-EVA Front-EVA PV
Cell
6
Back-Sheet Exposure
• A PET based back-sheet will be exposed to humidity between that outside and inside the module.
0102030405060708090
100
0 1 2 3 4 5
Rela
tive
Hum
idity
(%
)
Days
Bangkok Thailand Back-Sheet Relative Humidity
Insulated Back, Glass/Polymer Inside
Close Roof, Glass/Glass Inside
Open Rack, glass/glass Inside
open Rack, Glass/Polymer Inside
Insulated Back, Glass/Polymer Outside
Close Roof, Glass/Glass Outside
Open Rack, glass/glass Outside
open Rack, Glass/Polymer Outside
7
Pet Hydrolysis Kinetics
𝑙𝑙𝑙 𝐶𝐶−𝑥
= 𝐴 ∙ 𝑡 ∙ 𝑅𝑅2 ∙ 𝑒−𝐸𝑆𝑘𝑘
Ea=129.4 kJ/mol (1.340 eV), A=2.84·1010 1/day, RH expressed as a percentage. *PET becomes brittle (1/3 initial tensile strength) and “failed” when log(C/C-x)=~0.0024, or about 0.55% hydrolysis of ester bonds. **Pickett et. al saw the activation energy vary between 125 and 151 kJ/mol with an average of 136±13 kJ/mol for four different PET grades.
O
O O
O
PolyEthylene Terepthalate (PET)n
HHO+ O
O O
O H
n-x
HO O
O O
O
x
+
*W. McMahon, H. A. Birdsall, G. R. Johnson, and C. T. Camilli, "Degradation Studies of Polyethylene Terephthalate," Journal of Chemical & Engineering Data, vol. 4, pp. 57-79, 1959. **J. E. Pickett and D. J. Coyle, "Hydrolysis Kinetics of Condensation Polymers Under Humidity Aging Conditions," Submitted to the Journal of Polymer Degradation and Stability, 2013.
8
PET Hydrolysis Results
𝑙𝑙𝑙 𝐶𝐶−𝑥
= 𝐴 ∙ 𝑡 ∙ 𝑅𝑅2 ∙ 𝑒−𝐸𝑆𝑘𝑘
Open Rack
Insulated Back
Open Rack
Insulated Back
Open Rack
Insulated Back
Open Rack
Insulated Back
Denver, Colorado 13,000 4,900 6,500 2,400 5.3 8.7 45 49Munich, Germany 11,000 4,400 5,100 2,100 6.0 9.2 47 50Albuquerque, New Mexico 9,000 3,200 4,400 1,500 6.4 11 48 52Riyadh, Saudi Arabia 8,200 3,000 4,000 1,500 6.7 11 48 52Phoenix, Arizona 3,400 1,300 1,700 630 10 17 54 58Miami, Florida 1,100 510 530 250 19 27 62 65Bangkok, Thailand 700 310 320 150 24 34 66 69
Relative Humidity at 85⁰C so that
1000 h equals 25 years exposure
(%)
Temperature at 85% RH so that
1000 h equals 25 years exposure
(⁰C)
Years to 0.55% degradation (i.e.
Hydrolysis Service Life) (y)
1000 Hours 85⁰C/85% RH
Years equivalent (y)
PET is predicted to “fail” after 2064 h of 85 ⁰C and 85% RH.
9
Site Specific Equivalent T and RH
𝑅 = 𝐴 ∙ 𝑅𝑅𝑛𝑒 −𝐸𝐸𝑘𝑉
𝑅𝑅𝑤𝑤𝑤𝑤𝑤𝑑𝑤𝑑 𝐸𝑎𝑤𝑎𝐸𝑤𝑤 = 𝑅𝑅𝑊𝑊 =∑𝑅𝑅𝑛𝑒 −𝐸𝐸𝑘𝑉
∑ 𝑒 −𝐸𝐸𝑘𝑉
1𝑛
This tells you what the relative humidity is at the temperatures where the most damage is done.
𝑅𝑅𝑊𝑊𝑛𝑒
− 𝐸𝐸𝑘𝑉𝑒𝑒 =
∑𝑅𝑅𝑛𝑒 −𝐸𝐸𝑘𝑉
𝑁
∴ ∑𝑒 −𝐸𝐸𝑘𝑉
𝑁 = 𝑒− 𝐸𝐸𝑘𝑉𝑒𝑒
The equivalent temperature (Teq) gives the temperature at RHWA for which constant conditions will produce a degradation rate equivalent to the yearly average.
These terms cancel out
=∑𝑅𝑅𝑛𝑒 −𝐸𝐸𝑘𝑉
∑ 𝑒 −𝐸𝐸𝑘𝑉
1𝑛
𝑛
𝑒− 𝐸𝐸𝑘𝑉𝑒𝑒
∴ 𝑇𝑤𝐸 = −𝐾𝐸𝐸
𝑙𝑙∑ 𝑒 −𝐸𝐸𝑘𝑉
𝑁
10
PET Hydrolysis Equivalent T and RH
Open Rack
Insulated Back
Open Rack
Insulated Back
Open Rack
Insulated Back
Open Rack
Insulated Back
Denver, Colorado 13,000 4,900 6,500 2,400 33 54 14 4.6Munich, Germany 11,000 4,400 5,100 2,100 28 46 25 8.4Albuquerque, New Mexico 9,000 3,200 4,400 1,500 37 58 13 4.2Riyadh, Saudi Arabia 8,200 3,000 4,000 1,500 48 70 5.6 2.0Phoenix, Arizona 3,400 1,300 1,700 630 46 68 9.8 3.3Miami, Florida 1,100 510 530 250 37 54 36 14Bangkok, Thailand 700 310 320 150 41 59 33 12
Years to 0.55% degradation (i.e.
Hydrolysis Service Life) (y)
1000 Hours 85⁰C/85% RH
Years equivalent (y)
Teq for Ea=129.3 kJ/mol (⁰C)
RH, at Teq for 2nd order Kinetics of
PET (%)
𝑙𝑙𝑙 𝐶𝐶−𝑥
= 𝐴 ∙ 𝑡 ∙ 𝑅𝑅2 ∙ 𝑒−𝐸𝑆𝑘𝑘
11
What Are Relevant Activation Energies
*R. R. Dixon, "Thermal Aging Predictions from an Arrhenus Plot with Only One Data Point," Electrical Insulation, IEEE Transactions on, vol. EI-15, pp. 331-334, 1980.
Degradation activation energy from Dixon*. Based on RTI testing for properties such as: Elongation at Break Flexural strength Tensile Strength Shear Strength Burst Strength Weight Loss Dielectric Strength Imp. Strength
12
For Diffusion Controlled Processes
0
2
4
6
8
10
12
0 10 20 30 40 50 60 70 80 90 100 110
Coun
ts
5 kJ/mol Activation Energy Bin
Diffusivity
Permeability
Solubility
Histogram for moisture ingress activation energy for PV polymeric materials.
13
Thermal Stress by Location and Mounting
0102030405060708090
0 50 100 150 200
Equi
vale
nt T
empe
ratu
re (°
C)
Activation Energy (kJ/mol)
Equivalent Temperature
RiyadhBangkokDenverMunich
Dashed Lines:Insulated Back
Solid Lines:Rack Mounted
14
Modeling Moisture in the Front-EVA
Back-Sheet
Glass
H2O
Back-EVA Front-EVA PV
Cell
The Back-EVA equilibrates with a characteristic time of about a day. The Front-EVA equilibrates with halftimes of between a day and several years depending on the mounting configuration, location, and the position in front of the cell. Uses the backside water concentration at the perimeter in a 2-D diffusion finite element algorithm. The cell size is 156+2 mm to account for water diffusing from the back to the front.
15
×
Front Encapsulant Water Content
The front encapsulant traps in moisture seasonally making the center of the cell front the most hydrolytically damaging area. The remainder of this presentation focuses on the center of the front side to evaluate the most stressful position in the module.
Rack mounted, Glass/Polymer modules
0
0.0002
0.0004
0.0006
0.0008
0.001
0.0012
0 1 2 3 4 5 6 7
Wat
er C
once
ntra
tion
(g/c
m3 )
Time (Years)
0
1.56
3.12
4.68
6.24
7.80
DistanceFrom
Cell Edge (cm)
Riyadh, Saudi Arabia
Munich, Germany
PV Cell
Profiles along this line are plotted.
16
RH Not Very Dependent Kinetics or Ea
0
10
20
30
40
50
60
0 25 50 75 100 125 150 175 200
Wei
ghte
d RH
(%)
Activation Energy (kJ/mol)
Bangkok, Thailand
12345
n
Insulated Back
Rack Mounted
17
Small RH Dependence in All Climates
0
10
20
30
40
50
60
0 25 50 75 100 125 150 175 200
Wei
ghte
d RH
(%)
Activation Energy (kJ/mol and eV)
Solid Lines:Rack Mounted
Dashed Lines:Insulated Back
Darker Line:n=2
Lighter Line:n=1
Bangkok
Riyadh
Denver
Munich
18
85 ⁰C/85% RH Equivalent Time-Bangkok
0.1
1
10
100
1000
10000
100000
0 25 50 75 100 125 150 175 200
1000
h, 8
5 ⁰C
, 85%
RH,
O
utdo
or E
quiv
alen
t (y
)
Activation Energy (kJ/mol)
Bangkok, Thailand
Rack, n=2
Rack, n=1
Rack, n=0
Insulated, n=2
Insulated, n=1
Insulated, n=0
Solid Lines: Rack Mounted
Dashed Lines: Insulated Back
25
85% RH
19
85 ⁰C/85% RH Equivalent Time-Riyadh
The unknown humidity dependence results in a 1000× uncertainty in the acceleration
0.1
1
10
100
1000
10000
100000
0 25 50 75 100 125 150 175 200
1000
h, 8
5 ⁰C
, 85%
RH,
O
utdo
or E
quiv
alen
t (y
)
Activation Energy (kJ/mol)
Riyadh, Saudi Arabia
Rack, n=2
Rack, n=1
Rack, n=0
Insulated, n=2
Insulated, n=1
Insulated, n=0
Solid Lines: Rack Mounted
Dashed Lines: Insulated Back
25
85% RH
20
Good RH Choice Reduces Uncertainty
0.1
1
10
100
1000
10000
100000
0 25 50 75 100 125 150 175 200
1000
h, 8
5 ⁰C
, 5%
RH,
O
utdo
or E
quiv
alen
t (y
)
Activation Energy (kJ/mol)
Riyadh, Saudi Arabia
Rack, n=2
Rack, n=1
Rack, n=0
Insulated, n=2
Insulated, n=1
Insulated, n=0
Solid Lines: Rack Mounted
Dashed Lines: Insulated Back
25
5% RH
Testing using a chamber humidity of 5% vs. 85% significantly reduces the variability in the acceleration factor.
21
The Highest RH You Might Want is ~25%
0.1
1
10
100
1000
10000
100000
0 25 50 75 100 125 150 175 200
1000
h, 8
5 ⁰C
, 25%
RH,
O
utdo
or E
quiv
alen
t (y
)
Activation Energy (kJ/mol)
Bangkok, Thailand
Rack, n=2
Rack, n=1
Rack, n=0
Insulated, n=2
Insulated, n=1
Insulated, n=0
Solid Lines: Rack Mounted
Dashed Lines: Insulated Back
25
25% RH
22
Damp Heat vs. Low RH Stress Test
Without knowing the moisture induced degradation kinetics, it is better to use a low RH and accelerate processes principally by thermal acceleration.
0.1
1
10
100
1000
10000
100000
0 25 50 75 100 125 150 175 200
1000
h, 8
5 ⁰C
, 85%
RH,
O
utdo
or E
quiv
alen
t (y
)
Activation Energy (kJ/mol)
Riyadh, Saudi Arabia
Rack, n=2
Rack, n=1
Rack, n=0
Insulated, n=2
Insulated, n=1
Insulated, n=0
Solid Lines: Rack Mounted
Dashed Lines: Insulated Back
25
85% RH
0.1
1
10
100
1000
10000
100000
0 25 50 75 100 125 150 175 200
1000
h, 8
5 ⁰C
, 85%
RH,
O
utdo
or E
quiv
alen
t (y
)
Activation Energy (kJ/mol)
Bangkok, Thailand
Rack, n=2
Rack, n=1
Rack, n=0
Insulated, n=2
Insulated, n=1
Insulated, n=0
Solid Lines: Rack Mounted
Dashed Lines: Insulated Back
25
85% RH
0.1
1
10
100
1000
10000
100000
0 25 50 75 100 125 150 175 200
1000
h, 8
5 ⁰C
, 25%
RH,
O
utdo
or E
quiv
alen
t (y
)
Activation Energy (kJ/mol)
Bangkok, Thailand
Rack, n=2
Rack, n=1
Rack, n=0
Insulated, n=2
Insulated, n=1
Insulated, n=0
Solid Lines: Rack Mounted
Dashed Lines: Insulated Back
25
25% RH
0.1
1
10
100
1000
10000
100000
0 25 50 75 100 125 150 175 200
1000
h, 8
5 ⁰C
, 5%
RH,
O
utdo
or E
quiv
alen
t (y
)
Activation Energy (kJ/mol)
Riyadh, Saudi Arabia
Rack, n=2
Rack, n=1
Rack, n=0
Insulated, n=2
Insulated, n=1
Insulated, n=0
Solid Lines: Rack Mounted
Dashed Lines: Insulated Back
25
5% RH
23
Conclusions
• With respect to PET hydrolysis, 85 ⁰C/85% RH, may be equivalent to hundreds or thousands of years.
• For thermal and/or moisture induced failure, the mounting configuration can be as important as the location.
• Care must be taken in accelerated stress testing to account for the variable relative acceleration of the different degradation modes.
• Choosing the right humidity level for accelerated stress testing can dramatically decrease the uncertainty in the results.