Reliability of Power Rating and Labeling
Christos Monokroussos, Kengo Morita, Werner Herrmann
TÜV Rheinland Group
Email: [email protected]
www.tuv.com
1
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Introduction to TÜV Rheinland
TÜV RHEINLAND worldwide operates 6 accredited PV laboratories
Testing and certification of PV
modules and PV components:
- IEC 61215 (crystalline PV)
- IEC 61646 (thin-film PV)
- IEC 62108 (concentrating PV)
- IEC 61730 (safety)
- ANSI / UL 1703 (safety US)
2
11.10.2016
> 30 years
experience and
research in PV
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Active in national and
international standardization
committees (IEC, CENELEC,
IECEE, DKE)
Reliability of Power Rating and Labelling
Content
Introduction to Production Tolerance
Tolerance seen in the Market
Laboratory Experience
Review of relevant Standards
Measurement Uncertainty
Manufacturing Sorting
3
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Content
Introduction to Production Tolerance
Tolerance seen in the Market
Laboratory Experience
Review of relevant Standards
Measurement Uncertainty
Manufacturing Sorting
4
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Uncertainty Factors of Inline PV Module Power Rating
 Module sorting
into power
classes
Light-Induced
Degradation (LID)
 Calibration
accuracy of the
reference device
 Traceability of
calibration to
World PV Scale
(WPVS)
5
11.10.2016
Internal
Factor:
Production
External
Factors
Internal
Factor:
Measurement
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Accuracy of measuring
equipment:
 Data acquisition,
sensors
 Solar simulator
performance
Measuring technique:
 Design and use of
reference device
 Equipment connectors
 Module temperature
 Capacitive effects of
modules
 I-V correction
Reliability of Power Rating and Labelling
Loss of Money due to Measurement Uncertainty
Global PV Production Capacity in 2014: 46.7 GW 1
Module Price: 0.38-0.76 $/Wp 2
Loss in bn $
2.50
2.00
1.50
PT = ±2%
PT = ±3%
PT = ±4%
PT = ±5%
1.00
0.50
0.00
 Approx. 265 Mio $ lost per 1% of uncertainty per year
 High uncertainty leads to money loss
1
2
6
Trends 2015 in Photovoltaic Applications, IEA International Energy Agency (2015)
PVXchange Module Price Index, PV Magazine (Jul. 2016)
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Content
Introduction to Production Tolerance
Tolerance seen in the Market
Laboratory Experience
Review of relevant Standards
Measurement Uncertainty
Manufacturing Sorting
7
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Example Datasheets and Labels
8
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Label Tolerance seen in the Market
No. of PV Manufacturers
25
20
Sample size: 87 c-Si PV
module manufacturers
Origin of data
15
10
5
0
Claimed Tolerance in PV Module Labels [%]
9
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Content
Introduction to Production Tolerance
Tolerance seen in the Market
Laboratory Experience
Review of relevant Standards
Measurement Uncertainty
Manufacturing Sorting
10
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Laboratory Experience from Measurements at STC
14%
Unique SNs: 7179
Unique Module Classes: 935
Unique Manufacturers: 190
Test Period: 2015-2016
Test Lab: TR-SHG
12%
Rel. Frequency
10%
71.7%
28.3%
Measurement
Uncertainty: ±2.50%, k=2
Mean Deviation to Label:
0.74%±4.10%, k=2
8%
6%
4%
2%
0%
-10%
-8%
-6%
-4%
-2%
0%
2%
4%
PMAX Deviation to Low Tolerance Limit [%]
6%
 Significant number of modules underperforms at STC tests.
 Modules are typically not solarized (LID not included).
11
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
8%
10%
Reliability of Power Rating and Labelling
Experience from Pre-shipment Counter-Flash Measurements
Rel. Frequency
30%
25%
Test Period: 2015-2016
Test Lab: TR-SHG
Measurement Uncertainty: ±2.50%, k=2
20%
Modules: 2295; Batches: 30; Manufacturers: 9
2.50%±1.99%, k=2
15%
98.9%
1.1%
10%
5%
0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
PMAX Deviation to Low Tolerance Limit [%]
4.0%
 The majority of PV-modules for buyer’s projects are within tolerance.
 Modules are typically not solarized (LID not included).
12
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
5.0%
Reliability of Power Rating and Labelling
Experience from Pre-shipment Counter-Flash Measurements
35%
LID determined in project
30%
LID not determined in project
Rel. Frequency
25%
Test Period: 2015-2016
Test Lab: TR-SHG
Measurement Uncertainty:
±2.50%, k=2
20%
15%
No. of Batches: 30
No. of Manufacturers: 9
No. of Modules: 2295
10%
5%
0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
PMAX Deviation to Low Tolerance Limit [%]
 When LID is determined in the project phase the rating is often higher
(approx. 1.5%) comparing to projects that LID is undermined.
13
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
5.0%
Reliability of Power Rating and Labelling
Laboratory Experience from Buyer’s Projects
Deviation to label tolerance for different batches sampled in pre-shipment stage
3.5%
3.5%
LID determined in project
LID not determined in project
10
3.0%
50
2.5%
2.5%
2.0%
100
2.0%
1.5%
1.5%
Sample size
1.0%
1.0%
0.5%
0.5%
0.0%
0.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0% 0.0%
Mean PMAX Deviation to Low Tolerance Limit [%]
1.0%
2.0%
3.0%
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
5.0%
Mean PMAX Deviation to Low Tolerance Limit [%]
Test Period: 2015-2016 Test Lab: TR-SHG Meas. Unc.: ±2.50%, k=2
No. of Batches: 30 Manufacturers: 9 Sample: 2295 Dev.: 2.50%±1.99%, k=2
14
4.0%
Coef. of Variation per Batch [%], k=2
Coef. of Variation per Batch [%], k=2
3.0%
Reliability of Power Rating and Labelling
Content
Introduction to Production Tolerance
Tolerance seen in the Market
Laboratory Experience
Review of relevant Standards
Measurement Uncertainty
Manufacturing Sorting
15
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
PV Module Qualification Testing and Certification
Actual status
Combined structure under
new IEC 61215 series
Crystalline silicon PV modules
IEC 61215 Ed. 2
 Test Methods
 Test Requirements
Part 1 – General Requirements
Part 1-1 c-Si
Part 1-2 CdTe
Part 1-3 a-Si & µc-Si
Thin-film PV modules
Part 1-4 CIS & CIGS
IEC 61646 Ed. 2
 Test Methods
 Test Requirements
Part 1-x New technologies
Part 2 – Test Methods
16
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Uncertainty of Power Rating in Production
New Test Requirements of IEC 61215-1
 Information on tolerances for PMAX, ISC, VOC must be stated on the type
label of the PV module.
 Module type label information on tolerances will be confirmed within the
new test sequence of IEC 61215 for product qualification testing.
 All test samples of a IEC 61215 certification project undergo electrical
stabilization (MQT 19).
 Electrical performance at STC (MQT 6.1) after stabilization is
referenced for confirmation of type label information.
 As nominal output power within a type family is typically classified in
steps of 5 WP, additionally two samples of lowest, highest and middle
power class of the module type family will included in the test program.
 PASS criterion: With consideration of the lab measurement uncertainty,
stabilized STC values of PMAX, ISC, VOC of all test samples must lie within
the specified tolerances.
17
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
New Test Requirements of IEC 61215-1
Gate #1 Requirements
Criterion C1:
Stabilized PMAX of each sample  Nominal Power
(NP) with consideration of minus-tolerance
Criterion C2:
Average of stabilized PMAX values of all samples
 Nominal power
Criterion C3:
Stabilized VOC ≤ nominal value on type label with
consideration of plus-tolerance
Criterion C4:
Stabilized ISC ≤ nominal value on type label with
consideration of plus-tolerance
|mX |
|tX |
= lab measurement uncertainty
= manufacturers tolerance stated on type label
PMAX: x=1; VOC: x=2; ISC: x=3
18
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
New Test Requirements of IEC 61215-1
Deficits of the Gate #1 practice:
 The standard does not define what is meant by tolerance: Is it based on
manufacturers flash lists only? Is it corrected for potential LID effects (if c-Si
technology)? Does it include the measurement uncertainty in the production
line/s?
 High risk for module manufacturer and buyer
 2 samples per power class means a lack of statistical relevance
Statistical relevance for the benefit of the manufacturer and buyer is not
considered in PASS/FAIL criteria
 Sample picking is done by manufacturer  No objective results
 Unclear how FAIL of a single power class shall be treated
 How to handle transport damage?  Risk for module manufacturer
19
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Introduction on Acceptance Sampling Criteria
Probability of Acceptance
100%
95%
90%
80%
70%
60%
50%
40%
30%
20%
10%
10%
0%
0%
20%
AQL
40%
RQL60%
80%
100%
Proportion of Defective
 The Operating Characteristic (OC) curves describes how well a sampling plan
discriminates between good and bad lots.
 ISO 2859 specifies an acceptance sampling system for inspection. It is indexed in
terms of the Acceptable Quality Level (AQL). Closely related term is the
Rejectable Quality Level (RQL).
20
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Statistical Significance of Pass/Fail Criteria
1
AQL 0.95
0.9
0.8
 AQL assumes a 5% risk for the Producer that a lot with defects <AQL is rejected
 RQL assumes a 10% risk for the Consumer that a lot with defects >RQL is accepted
Probability
0.7
6 samples, no fail (IEC 61215-1, Gate#1): AQL=0.82%, RQL=31.9%
14 samples, 1 fail (IEC 61215-1, All Samples): AQL=2.6%, RQL=25%
32 samples, 2 fails, GIL 2; G, (ISO 2859-1): AQL=2.6%, RQL=15.8%
500 samples, 21 fails, GIL 2; N (ISO 2859-1): AQL=3.0%, RQL=5.6%
0.6
0.5
0.4
0.3
0.2
RQL 0.1
0
0
10
20
30
40
50
60
70
Defective proportion of PV-Modules [%]
80
90
100
 The sampling plan of IEC 61215-1, Gate#1 implies an unrealistically high quality control
for production lines (AQL approx. 0.82%) and does not provide confidence for buyers.
 A production line with a realistic acceptable quality level of 2.5% would have approx.
15% risk to fail gate#1 criteria.
21
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
AQL 0.95
0.9
Reliability
of Power Rating and Labelling
0.8
1
AQL 0.95
Statistical
Significance of Pass/Fail Criteria
0.7
0.9
Probability
ProbabilityProbability
ProbabilityProbabilityProbability
0.6
1
0.8
AQL 0.95
0.5
0.9
0.7
0.4
0.8
0.6
1
AQL 0.95
0.3
0.7
0.9
0.5
1
AQL 0.95
0.2
0.6
0.8
0.4
0.9
RQL0.5
0.1
0.7
0.3
0.8
0
0.4
0.6
0.2
0.7
0
RQL 0.3
0.5
0.1
0.6
0.2
0.4
0
0.5
0
RQL 0.1
0.3
0.4
0.2
0
0.3
0
RQL 0.2
0.1
1
 AQL assumes a 5% risk for the Producer that a lot with defects <AQL is rejected
 RQL assumes a 10% risk for the Consumer that a lot with defects >RQL is accepted
6 samples, no fail (IEC
AQL
61215-1,
Gate#1): AQL=0.82%, RQL=31.9%
0.95
14 samples, 1 fail (IEC 61215-1, All Samples): AQL=2.6%, RQL=25%
 AQL assumes
a 5% risk 2forfails,
the GIL
Producer
that a2859-1):
lot with AQL=2.6%,
defects
<AQL
is
rejected
32 samples,
Decreasing
producers’
risks
2; G, (ISO
RQL=15.8%
 RQL assumes
10% risk21
forfails,
the Consumer
that2859-1):
a lot withAQL=3.0%,
defects >RQL
is accepted
500 asamples,
GIL 2; N (ISO
RQL=5.6%
0.9
3
4
30
40
50 0
601
702
80
90
6 samples,
no
fail
(IEC
61215-1,
Gate#1):
AQL=0.82%,
RQL=31.9%
Defective proportion of PV-Modules [%]
14 samples, 1 fail (IEC 61215-1, All Samples): AQL=2.6%, RQL=25%
samples, 2 fails, GIL 2; G, (ISO 2859-1): AQL=2.6%, RQL=15.8%
Increasing buyers’32
comfort
10
20 500 samples,
30
70
80RQL=5.6%
90
2140
fails, GIL 50
2; N (ISO 60
2859-1): AQL=3.0%,
Defective proportion of PV-Modules [%]
10
10
20
20
30
40
50
60
70
Defective proportion of PV-Modules [%]
80
90
100
100
100
 RQL
Thesampling
plan of IEC 61215-1, Gate#1 implies an unrealistically high quality control
0
0.1
0
10
20
30
40
50
70
80
90
100
for production lines (AQL
approx.
0.82%)
and
does 60
not provide
confidence
for buyers.
Defective proportion of PV-Modules [%]
0
0
10 with 20
30 acceptable
40
50
60
70
80
90 approx.
100
 A production
line
a realistic
quality
level
of 2.5%
would
have
Defective proportion of PV-Modules [%]
15% risk to fail gate#1 criteria.
 Increasing the sampling size would reduce risks for both producers and consumers.
22
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Content
Introduction to Production Tolerance
Tolerance seen in the Market
Laboratory Experience
Review of relevant Standards
Measurement Uncertainty
Manufacturing Sorting
23
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Solar Simulator Performance
IEC 60904-9 Classification
Component
Spectral Irradiance
Non-uniformity of Irradiance
Short and Long Temporal Instability
(LTI and STI)
24
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Example
Reliability of Power Rating and Labelling
Factors affecting the Uncertainty of Power Measurement
Solar Simulator
Performance
Reference Solar Device

Calibration uncertainty of
reference device







25
Implementation of reference
device
Accuracy and calibration of
temperature sensor
Long-term drift and stability
of reference device
Repeatability of reference
irradiance acquisition
Reproducibility of reference
irradiance acquisition
11.10.2016





Non-Uniformity of Irradiance:
Non-uniformity
Measurement accuracy
Temporal instability of
irradiance (LTI & STI)
Electronic accuracy of I-V
measurement channels
Accuracy and calibration of
temperature sensors
Repeatability of solar
simulator
Reproducibility of solar
simulator
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Operator Technique /
Measurement Procedure

Calibration and
measurement procedure

I-V corrections

Spectral mismatch

Optical mismatch between
reference module and test
module
Temperature:
Control

Temperature non-uniformity


Connection technique
Operation:

System maintenance

Human error
Spectral mismatch/
Calibration accuracy of
reference
Reliability of Power Rating and Labelling
Uncertainty impacts on I-V curve
Non-uniformity of irradiance/
cell cracks, Isc variation of cells
Overall impact:
Temperature,
irradiance correction,
I-V data acquisition accuracy
Temperature
measurement
26
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
SI units
Reliability of Power Rating and Labelling
Traceability practice
Metrology
authority
Test Laboratory
Primary
calibration
Solar simulator 1:
Reference module
Calibration
(GOLDEN MODULE)
End user
WPVS cell
27
11.10.2016
PV module manufacturer
Production line/s
Test laboratory
Solar simulator 3:
Production line
power measurement,
power classification
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Solar simulator 2:
Working standards
for daily use
(SILVER MODULES)
Reliability of Power Rating and Labelling
Uncertainty Assessment Service (UAS) Objective
 UAS shall give confidence in stated manufacturer’s tolerances for PMAX,
VOC and ISC, which are subject to IEC 61215 certification.
AIST, NREL, PTB
Primary calibration
Reference cell (WPVS Design)
± 0.6% to 1%
PV Test Institute
Secondary calibration of
reference module (GOLDEN MODULE)
± 1.6% to 3.0%
Working reference
(SILVER MODULE)
Production line output power
measurement
28
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
???
UAS
PV module
manufacturer
???
Reliability of Power Rating and Labelling
Solar simulator performance
Variation in solar simulator performance: Spatial
non-uniformity and temporal instability of
irradiance, AM1.5 match of spectral irradiance)
29
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Measurement
uncertainty
Reliability of Power Rating and Labelling
Work Programme
• Documentation of measurement equipment
• Performance check of all solar simulators in manufacturers test
lab and production lines
• Evaluation of PV module designs: Reference modules (REF) vs.
production line modules (DUT)
• Review of calibration protocols for reference modules and
measurement equipment (I-V load, temperature sensors, etc.)
• Evaluation of specific effects: I-V hysteresis, LID
• Random sample picking from production and comparison of
manufacturer rating with TÜV Rheinland measurements
30
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Calculation Method

ISO/IEC Guide to the Uncertainty of Measurement (GUM)
Definition of general principles for Uncertainty of Measurement
uP MAX  k  u1  ...  u N
2
2
uPMAX
ui
k
=
=
=
Expanded overall measurement uncertainty
Standard uncertainty for uncertainty source
Coverage factor (k=2 for 95% confidence interval)
 PMAX uncertainty sources ui








31
Calibration uncertainty of REF
Measurement equipment: Current, voltage, temperature measurement
Repeatability of measurement, drift of irradiance sensor
Spatial non-uniformity of irradiance
I-V curve temperature and irradiance correction
Electrical mismatch between REF and DUT (variation in spectral response)
Optical mismatch between REF and DUT (variation in angular response)
Measurement technique: Misalignment between REF and DUT, transient
effects, temperature measurement, contacting technique, traceability practice
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Example of Uncertainty in Silver Module Calibration
Reference Calib.
IMPP=2. 64%, k=2
0.12%
Case of Study:
Class AAA system
PMAX=2.8%, k=2
Reference Integration
0.37%
Non-Uniformity of Irradiance
0.80%
2.00%
Spectral Mismatch
Other Optical
Factor
0.23%
Repeat.
0.24%
Reproducibility
0.35%
0.78%
0.64%
Electronic Related
Temp. Related
VMPP=1.09%, k=2
0.03%
Electronic Related
0.23%
0.40%
Temp. Related
Elect. Connection
0.27%
Repeatability
0.95%
Other
32
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Uncertainty
Source
Reference Calib. PMAX
±2.00%, k=2
Non-uniformity of
irradiance
±1.1% (Rectangular)
Temp. Accuracy
±1.0ºC (Rectangular)
Temp. non-uniformity
±1.0ºC (Rectangular)
Reproducibility
Electronic Accuracy
±0.8%, k=2
±0.2% (Rectangular)
Spectral mismatch
±0.35%, k=2
Non-uniformity of
irradiance unc.
±0.25%, k=2
Contact resistance
±10mΩ (Rectangular)
Reliability of Power Rating and Labelling
Content
Introduction to Production Tolerance
Tolerance seen in the Market
Laboratory Experience
Review of relevant Standards
Measurement Uncertainty
Manufacturing Sorting
33
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Module Sorting into Power Classes
45
40
Frequency
35
30
25
20
15
160W
165W
170W
10
5
0
158 159 160 161 162 163 164 165 166 167 168 169 170 171 172
Power [W]
 Sorting tolerance into power classes contributes to the production tolerance
 Sorting tolerance of 95% confidence level can be achieved  95% of
measurements lie within PAVG ± 2.
34
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
165W Power Class
0.04
150
Probability
Number of PV-Modules
Reliability of Power Rating and Labelling
Factors affecting Sorting Tolerance
100
50
0
230
235
240
245
Rated Power
250
163W
164W
165W
166W
167W
0.03
0.02
0.01
0
155
160
165
170
175
Power [W]
 The sorting tolerance of a power class is influenced by both sorting bins size
and production frequency
 Each power measurement has a degree of uncertainty, which will also affect the
label tolerance
35
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Production Tolerance: Combining Sorting and Measurement
0.3
Sorting Tolerance:  ST95=1.74%
Probability
Measurement Tolerance: 
Total Tolerance: 
0.2
TT95
MT95
=3.46%
Limiting Factor
0.1
0
150
-2
155
+2
160
165
170
Power [W]
175
 Total tolerance is mainly limited by measurement uncertainty when module
sorting is done per 5 Watt classes.
36
=3%
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
180
Reliability of Power Rating and Labelling
Conclusions
 The declared production tolerance of PV-modules in the market is
sometimes unclear and can be inaccurate.
 The gate #1 requirements of IEC 61215-1 aim to provide confidence in PVmodule labeling. However, the sampling size is currently low and does not
provide statistical significance in the evaluation.
 To reach maturity PV community needs a common language, which is clear,
transparent, and if possible consistent, on how production tolerances are
declared on PV-module nameplates and what they cover.
 Measurement uncertainty analysis is complex and requires assumptions
and experience for accurate estimates. Sorting tolerance and LID must be
considered.
 Production tolerance in PV industry lies in the range 3% to 5% for c-Si
modules, if appropriate quality assurance is in place.
 UAS aims in increasing confidence for consumers and reducing risks for
producers.
 To reduce risks it is recommended to perform counter-flash measurements
at accredited, laboratories for a small but representative PV-module
sample.
37
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Reliability of Power Rating and Labelling
Acknowledgements
 Mr. Joerg Althaus from TÜV Rheinland Energie und Umwelt GmbH
 Dr. Sherry Zhang, Mr. Sebastian Petretschek, Mr. Damien Etienne and Mr.
Victor Feng from TÜV Rheinland (Shanghai), Co. Ltd
for helpful discussions and data contribution
38
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016
Thank you for your attention
39
11.10.2016
Sayuri 2016 Workshop, AIST, Japan, October 4th – 5th, 2016