Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/280739603

COMBINED STANDARD FOR PV MODULE DESIGN QUALIFICATION AND TYPE


APPROVAL: NEW IEC 61215 -SERIES

Conference Paper · September 2014

CITATIONS READS

5 7,354

7 authors, including:

Bengt Jaeckel Arnd Roth


Fraunhofer Center for Silicon Photovoltaics VDE Association for Electrical, Electronic & Information Technologies
84 PUBLICATIONS   599 CITATIONS    5 PUBLICATIONS   12 CITATIONS   

SEE PROFILE SEE PROFILE

Gerhard Kleiss Peter Seidel


Leibniz Universität Hannover First Solar GmbH
48 PUBLICATIONS   251 CITATIONS    6 PUBLICATIONS   20 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Photovoltaic O&M and asset management View project

PV Standardization View project

All content following this page was uploaded by Bengt Jaeckel on 07 August 2015.

The user has requested enhancement of the downloaded file.


COMBINED STANDARD FOR PV MODULE DESIGN QUALIFICATION AND TYPE APPROVAL:
NEW IEC 61215 - SERIES

Bengt Jaeckel1, Arnd Roth2, Guido Volberg3a, Joerg Althaus3b, Gerhard Kleiss4, Peter Seidel5, and Markus Beck6
1
UL International GmbH, Admiral-Rosendahl-Strasse 9, 63263 Neu-Isenburg, Germany, bengt.jaeckel@gmail.com
2
VDE Prüf- und Zertifizierungsinstitut GmbH, Merianstraße 28, 63069 Offenbach, Germany, arnd.roth@vde.com
3a
TÜV Rheinland LGA Products GmbH, Am Grauen Stein, 51105 Köln, Germany, volberg@de.tuv.com
3b
TÜV Rheinland Energie und Umwelt GmbH, Am Grauen Stein, 51105 Köln, Germany, Joerg.Althaus@de.tuv.com
4
SolarWorld AG, Martin-Luther-King-Straße 24, 53175 Bonn, Germany, Gerhard.Kleiss@solarworld.de
5
First Solar GmbH, Rheinstraße 4, 55116 Mainz, Germany, peter.seidel@firstsolar.com
6
Siva Power, 2387 Bering Drive, San Jose, CA 95131, United States, markus.beck@sivapower.com

ABSTRACT: The commercial success of photovoltaics (PV) is largely based on the long-term reliability of the PV
modules. Current PV modules tend to carry a performance warranty of 25 years. These modules are typically
qualified to IEC 61215 or IEC 61646 – design qualification and type approval for terrestrial crystalline Si (c-Si) or
thin-film (TF) technologies, respectively. These qualification tests have shown to adequately identify design,
material, and process flaws that could lead to premature field failures. Consequently, PV module customers have
come to appreciate the criticality of the tests set forth in IEC standards 61215 and 61646. The PV market has come to
trust these IEC standards, and as such they represent an essential component of the success of the rapid growth of the
PV market.
Since the last revisions of the standards the knowledge of PV failure modes increased dramatically. To retain market
trust in the IEC standards, maintenance of the pertinent standards needs to assure compliance with state-of-the-art
knowledge and best practices. To be prepared for the upcoming technology developments this new structure allows
fast changes to keep the standards up-to date.
Over the past two years Working Group 2 (WG2) of the technical committee for Solar Photovoltaic Energy Systems
(TC82) within the International Electrotechnical Commission (IEC) invested a considerable effort to update and
merge the terrestrial PV module design qualification and type approval standards into a single IEC 61215 series. Part
1 addresses the minimum requirements for testing with subparts 1-1 through 1-4 addressing PV technology specific
deviations – e.g. stabilization – while Part 2 deals with the required tests protocols and test sequences. In the current
committee draft of edition 3 the power output verification has been completely overhauled, new tests have been
added, and pass/fail criteria revised.
This paper explains the most crucial changes to edition 3 of IEC 61215 and provides the rationale for the applied
changes. There are multiple reasons for the changes, but one significant justification is that the new standard series
allows more flexibility and faster responses to changes in any given PV technology.

Keywords: IEC, IEC 61215, IEC 61646, design qualification, type approval, c-Si, Thin-film

1. HISTORIC BACKGROUND through Block V


Outdoor (OD) deployment predominantly in
IEC 61215 edition 1 was published in 1993 based on desert environments
prior investigations as to typical observed field failure Block VI fell victim to Regan administration
modes – most prominently the JPL Block Buys I-V budget cuts
(1975-1981). The main focus during this effort was on 1981-1991: ESTI – EU Specifications 501-503
crystalline PV modules with solar cells of typically Similar to JPL Block V plus UV and OD
500µm thickness. The standard was revised as edition 2 exposure tests; TC maximum reduced to +85°C
in 2005 to take new developments into account and better EU 503 draft for IEC 61215
serve the fast growing industry. 1990: SERI IQT modifications for TF (a-Si)
As thin-film modules exhibit different and/or additional Added wet insulation resistance test, bypass diode
failure modes a separate standard – IEC 61646 edition 1 test, cut susceptibility test and ground continuity
– was released in 1996 and updated in 2007 as edition 2. test (last two from UL 1703)
This work was primarily based on work done at SERI on 1993: IEC 61215 Ed. 1 (c-Si)
a-Si and the standard evolved out of SERI IQT, IEEE Incorporated best practices from around the world
1262 and IEC 61215. 1995-2000: IEEE 1262 – all PV technologies
Hybrid between IQT and IEC 61215
a. Timeline 1996: IEC 61646 Ed. 1 (TF – a-Si)
The following list summarized the most important steps Based on IEEE 1262 plus light soak and anneal
during the development of both standards [1][2]. 1975- cycles
1981: JPL ‘Block Buys’ I-V (c-Si) 2005: IEC 61215 Ed. 2 (c-Si)
Based on NASA tests for space arrays Added wet leakage test from IEC 61646
Thermal cycling extremes (-40 & +90°C) based on Added bypass diode test from IEEE 1262
assumed worst case for terrestrial deployment Added Imp during TC 200
Short exposure to humidity rooted in space array 2007: IEC 61646 Ed. 2 (TF – a-Si, CdTe, CIGS)
payload pre-launch time Attempt to adapt Ed. 1 to different types of TF
Mechanical load, hail and high pot testing added modules (CdTe, CIGS)
Modified pass/fail criteria to rated module power enabling customers and investors as well as other
after HAST and post-stabilization stakeholders up and downstream the value chain to assess
Modified hot spot test the standards applied to certain components of the PV
Added bypass diode thermal test system – in this case the PV module. Given the important
2014: IEC 61215 Ed. 3 CDV role IEC 61215 and IEC 61646 played in achieving broad
acceptance of PV modules and the resulting trust in these
Both IEC standards employ very similar, well-defined standards from the customer base, educating the latter on
accelerated stress tests designed to reveal failure modes key changes is a critical mission of IEC WG2 and, in
mainly observed in the field on first generation PV particular, of this paper.
modules. In order to develop accelerated stress tests they
have to cause degradation to verify that they duplicate 3. TERMS AND DEFINITIONS
failure modes observed under outdoor deployment
conditions. Those failure modes are under investigation The new standard series required new definitions. In
as a global effort and the review is outside the scope of addition to the ones below the Terms and Definitions of
this paper. upcoming IEC 61215 ed.3 and IEC 61730 ed.2 should be
It is, however, important to point out that PV module considered for better understanding.
technology changed significantly since the beginning of
the JPL-Block Buys. Thus, a comprehensive correlation Module Quality Test (MQT): To better reference
cannot always be established between past observed field certain tests that are defined in the IEC 61215 series each
failures to newly observed failures in accelerated stress test is assigned a designated MQT-number similar to
tests. Future standards must exhibit the flexibility to MST (Module Safety Test) from IEC 61730. E.g. MQT 2
quickly act on technology advances to serve and support is equivalent to 10.2 of IEC 61215 ed. 2 as well as MST
the industry. 03 of IEC 61730 ed.2.

b. Historic failure rates of IEC PVmodule Tolerances on nameplate: “Uncertainty”, or better


standards tolerance statement related to values of electrical
Failure rates during IEC testing are published in the parameters on the nameplate of the PV module as given
literature on a regularly basis [3][4][5]. Below is a brief by the manufacturer. This will also be important for the
survey of the highest failure rates from [5] section 3, upcoming revision of EN 50380.
figures 3.5 and 3.6:
1.~50% of test failures were climate tests Uncertainty of Laboratory: Uncertainty of the
a. Thermal cycles stress solder connections, higher laboratory for a certain measurement, e.g. power
failure rate for c-Si compared to thin-film measurement. This value includes an assessment of all
b. Damp heat test failures more often for thin- inputs related to the measurement and their tractability as
film modules required by ISO 17025.
2.~10% of faults occurred prior to load tests
3.~10% workmanship quality – initial measurements Power Classes: Power – typically maximum power –
that include visual inspection, insulation resistance sorting criteria from the PV module manufacturer.
and performance Classes are usually spread 2.5W (thin-film) or 5W (c-Si)
4.~10% hot-spot failures in width. A representative value of the class is printed on
5.~10% mechanical load issues the modules nameplate. E.g. a +5W/-0W sorting of 250W
(nameplate) modules leads to modules ranging from 250
Generally speaking, most of today’s modules are being to 254.9W in one power class.
constructed to fulfill the design qualification and type
approval standards IEC 61215 and IEC 61646. 4. MAJOR CHANGES

2. MOTIVATION TO CHANGE This chapter highlights the most important changes of the
STANDARD STRUCTURE new IEC 61215 series. The changes as well as the
rationale are presented to further motivate the transition
The harmonization and, more precisely speaking, the to the next level of PV module type approval.
combination of 61215 and 61646 is motivated by the
following items: a. New standard structure
- Eliminate the perception that TF technologies Photovoltaic is a very fast evolving technology. New cell
are inferior to crystalline technologies. and module concepts emerge frequently. Besides the first
- Establish identical requirements for all (c-Si) and second (Thin-Film) generation of photovoltaic
technologies. technologies the next, third generation, is on the horizon.
- Subject all technologies to the same tests. Standards should not only address current products, but
- Simplification for customer: product research should be able to be proactive as well as react in a quick
and selection is based on one PV module and flexible fashion to technology advances, like the
qualification standard. introduction of organic PV devices (OPV).
Figure 1 shows a comparison between the old structure
Consequently the test protocols will be better aligned (left) and the new structure (right). In the past both
between the different solar cell and module technologies standards had general requirements and test protocols that
as well as have more synergies with IEC 61730. The were very similar, but due to updates they were usually
latter has always been a challenge in the past. not aligned (a prominent example are the different UV-
The new structure greatly enhances transparency tests). The new structure will ensure that all PV modules
have to pass through the same battery of tests and their
severities. updates based on new data as well as the possibility to
Moreover, the old structure of the IEC module design and swiftly add standards related to new technologies.
type approval standards only differentiated between c-Si While Part 1 and its technology specific subparts
and TF modules. No technology specific distinction was addresses the minimum requirements for testing
made within the thin-film category – i.e. a-Si, CdTe, including relevant pass/fail criteria. Part 2 deals with the
CIGS were treated as identical. However, knowing that required tests protocols and test sequences. This revised
each TF technology has its own and unique properties led structure greatly enhances transparency enabling
to unavoidable penalties. customers and investors as well as other stakeholders up
To avoid unequal certification conditions the technology and downstream the value chain to assess the standards
relevant differences are addressed within the sub- applied to certain components of the PV system – in this
standards of Part 1 for all current state of the art case the PV module.
technologies. Finally, the new structure allows for easy

Figure 1: Old structure (left), new structure (right) for c-Si and thin-film PV modules, including options for new
technologies – e.g. OPV.

b. Unified Power Output Pass/Fail via the introduction of the new Gate #1 criterion to verify
Criteria a correct initial rating of the PV module type. During
One major technical change is the pass/fail criterion for Gate #1 evaluation the manufacture tolerance (e.g. 3%,
PV module output power. In the past there was a relative light blue boxes around the binning classes in Figure 2) is
5% power loss by test and an 8% overall power loss used as measurement uncertainty to define the pass/fail
criterion for crystalline modules whereas thin-film criteria. Modules from different power classes must be
modules could exhibit a power loss of 10% with respect used for this investigation, subsequently determining the
to their nameplate rating. This obviously is not judging power range of the manufacturer’s “IEC 61215 type
PV modules by the same metric. Moreover, there had approval”. Figure 2 gives two possible examples of
been a long discussion within the community that the 8% laboratory measurement. As indicated in the figure only
or 10% criteria might be too weak. While the project the module type passes where all laboratory
team proposed an 8% power loss across the board, WG2 measurements fall within the manufactures nameplate
elected during its semi-annual meeting in the spring of rating including stated tolerances (indicated with the blue
2013 to standardize to 5%. boxes with “+” and “-”; those tolerances need not
Given that PV module prices are commonly based on the necessarily be symmetric).
power output (Wp) it is important that the nameplate Gate #2 is comparable to the old scheme of IEC 61646
rating is correct. This concept already inherent in IEC comparing nameplate rating and allowed maximum
61646, but not in ed.2 of IEC 61215 was used to overhaul power degradation; per WG2 decision equal to 5%.
the power pass/fail criteria for the new IEC 61215 series. The minimum allowed power output is determined by
The new structure has two stages – so called Gates – for first finding the minimum power allowed as per the
power evaluation. The first gate (Gate #1) verifies the nameplate rating and manufacturers tolerance. This value
nameplate or label of the module and ensures correct is set to 100%. Subsequent to all stress tests applied to
sorting within the factory. Gate #2 determines the modules, their related power output after final
degradation during testing. stabilization is compared to the initial value. A maximum
In the past, regardless of technology, the origin for the power loss of 5% is permissible.
lower measured power could be twofold: degradation or To better illustrate the method, the flowchart in Figure 3
incorrect nameplate rating. The origin of this ambiguity includes two examples with different power output and
was due to the absence of initial power determination for different manufacturer tolerances.
TF modules to verify nameplate rating, while in the case
of c-Si modules only relative values were determined.
Differentiating the two mechanisms has been addressed
Figure 2: Gate #1: Output power verification with respect to nameplate rating. STC power output of PV modules is
determined (MQT 6.1) using the given stabilization procedure (MQT 19). Nameplate rating is then compared to laboratory
measurements (green and red arrows indicate measured Pmax).

Figure 3: Definition of Gate #2 and how to apply it. Right side gives two examples where the concept of manufacturer
tolerance and excepted power degradation during testing is demonstrated.

c. Uncertainty Discussion be reasonably set since a higher uncertainty means less


A further extremely important change in the new control and therefor a lower valued product. Typical
standard relates to the application of measurement values for c-Si modules are in the range of ± 3-5% and
uncertainties; in particular the measurement uncertainties ± 5-10% for TF modules. This tolerance is related to the
during power determination at the manufacturer (MQT 2) measurement uncertainty and is different from power
and at the laboratory (MQT 2 and MQT 6.1)[6], [7], classes.
[8],[9]. The measurement uncertainty of the certification
The tolerance or uncertainty within the manufacturing laboratory tends to be lower than the manufacturer’s
line must be determined by the manufacturer and should uncertainty due to much better controlled environments
and more sophisticated calibration chains required under Following this guideline the Gate #1 & 2 values (see also
ISO 17025 accreditation. IEC Certification Bodies Test Figure 2 and Figure 3) are straight forward. However, as
Laboratories (CBTL) are accredited in accordance to ISO a result of this arbitrary zeroing of CBTL uncertainty the
17025 and as such need to know the uncertainty of their pass/fail criteria levied on the PV modules within the new
measurements. To maintain their ISO 17025 accreditation approach are significantly more severe than in the prior
CBTLs participate in round robins. An example is editions of IEC 61215 and IEC 61646.
presented in Figure 4.
Maintaining a traceable calibration chain is critical, but it The following presents aspects of the ongoing discussion
is equally important to understand the impact of different within WG2 on the topic of uncertainties.
solar cell technologies on the outcome, in particular, the The inputs for both determined uncertainties
effect of the spectrum of the solar simulator [10], [11], (manufacturer and laboratory) are mainly based on the
[12], [13]. The latter can be compensated with a spectral same physical properties and usually are traced back to
mismatch correction, but this requires knowledge of the the same 1st order references.
spectral response that is not always known [15], [16]. An Assuming the same measurement hardware in the factory
example of the effect of spectral mismatch is shown in and the laboratory the same trend (lower or higher
Figure 5. measurement values) should occur. In this case the
Similar plots as in Figure 4 and Figure 5 can be created approach from IEC guide 115 is reasonable as one cannot
for all measurement parameters and taking all effects of take the same problem into account twice. For different
the calibration chain into account laboratories state MQT hardware at the two locations a detailed investigation is
6.1 uncertainties in the range of ± 1.8 to 4%. necessary; unfortunately this is not a practical approach.
Another important concept to understand is in this
context relates to statistics. Most would agree that the use
of the ISO Guide 115 for a single module measurement is
questionable. But module manufacturing has long entered
the high volume stage – i.e. thousands of modules are
produced on one manufacturing line. From a statistical
viewpoint the mean of all power output measurements
should represent the promised value nameplate rating.
From this perspective of statistical sampling size it makes
sense to include only one uncertainty, either that of the
manufacturer or of the CBTL. As the manufacturer
uncertainties tend to be larger it would thus be practical
to use the respective value and follow IEC Guide 115.
The reverse would also be permissible, the key is to use
Figure 4: Boxplot of Isc values determined in the only one tolerance and not to sum up both as was the case
different laboratories with respect to the average value in prior editions of the type approval standards.
[17]. Transparency has to be given in the report as to which
uncertainty statement was used – the manufacturer’s or
that of the laboratory.
To consider the problems illustrated in Figure 4 and
Figure 5 the following is proposed to improve the IEC
61215 ed.3 draft as well as be aligned with the proposal
of EN 50380.
All modules are measured individually in their stabilized
state (MQT 19 + MQT 6.1) yielding their respective
PMeasurement values. For all modules under investigation the
arithmetic mean output power P is
determined. This value is subsequently compared to the
stated nameplate rating PMax taking into account the
manufacture’s tolerance m in %. The following applies:

%
∙ 1+ !≥
100 #

Each individual module must meet the following


Figure 5: Overview of measurements of the same set of criterion:
modules on different solar sun simulators. Values are % %%
∙ 1+ !≥ # ∙ 1− !
100 100
normalized to the values of Simulator 1. Deviations of up
to 5% difference in Pmax measurement between
simulators of one test laboratory are visible clearly [17].
where t in % is the laboratory’s stated uncertainty. Recall,
The previous details are essential to understand the tolerance is not equal to uncertainty as per the definition.
discussion within WG2 on how to handle laboratory
uncertainties. Per IEC Guide 115 – “Application of
uncertainty of measurement to conformity assessment
activities in the electro technical sector” – any laboratory
uncertainty can be in essence set equal to zero.
d. Test Sequence Modifications Measurement, 51 (2014), 451-456
Besides the standard structure modification presented at [7] Dirnberger, D., et al, “Uncertainty in PV Module
the beginning of this section the following lists other Measurement - Part I: Calibration of Crystalline and
proposed changes for edition 3 of IEC 61215 and Thin Film Modules, 39th IEEE Photovoltaic
required editorial modifications resulting from the new Specialists Conference Tampa, Florida (2013)
structure. [8] Dirnberger, D., et al, “Uncertainty in PV Module
Measurement - Part II: Verification of Rated Power
Unique test identifiers MQTs and Stability Problems”, 40th IEEE Photovoltaic
Unique sequence identifiers Sequences A, B… Specialists Conference Denver, Colorado (2014)
New Test sequence A to prove Gate #1 [9] Mihaylov, B., et al., “Uncertainty Considerations of
compliance Indoor PV Module calibration Based on Monte Carlo
NOCT replaced by NMOT (MQT 5) technical Simulations”, 28th European Photovoltaic Solar
change proposed for edition 3 Energy Conference – Paris (2013)
Hotspot test (MQT 9) is technology dependent [10] Bliss, M., et al, “Measurement of Flash Solar
each technology can display different sensitivity to Simulator Output Spectra over Bulb Lifetime and the
hot-spot heating; handled individually in Parts 1-x. Effect on Spectral Missmatch”, 28th European
After MQT 9 no STC (MQT 6.1) pass/fail criterion Photovoltaic Solar Energy Conference – Paris (2013)
exists. [11] Bogomolova, S.A., et al., “Evaluation of the
UV test (MQT 10) Modules now operate at Pmpp Uncertainty in Measuring IV Characteristics of The
conditions during test Thin-Film PV modules”, 28th European Photovoltaic
New J-Box Tests Technical change proposed for Solar Energy Conference – Paris (2013)
edition 3; MQT 14.1 and 14.2 were created [12] Chandel, R., et al, “Uncertainty Analysis of
Stabilization (MQT 19): Preconditioning (c-Si) and Photovoltaic Power Measurements using Solar
Lightsoaking (TF) redefined as Stabilization. Simulators”, Energy Technology, 1 (2013),763-769
Modules are operated at Pmpp conditions to make it [13] Herrmann, W., et al, “Uncertainty of Solar Simulator
consistent with MQT 10. Spectral Irradiance Data and Problems with Spectral
Match Classification”, 27th European Photovoltaic
5. SUMMARY Solar Energy Conference – Frankfurt (2012)
[14] Muellejans, H., et al, “Analysis and mitigation of
On the one hand the new structure is very different from measurement uncertainties in the traceability chain
that of editions 1 and 2. The implementation of sequence for the calibration of photovoltaic devices,
and test identifiers (MQT) greatly simplifies to refer to Measurement Science and Technology, 20, (2009),
individual tests defined in IEC 61215 from other PV 075101
module standards. [15] Qiu, Y.N., et al., “Uncertainty in the Calibration of
On the other hand, the technical changes implemented in Thin film based Silicon Multi-junction Devices using
this revision are not dramatic, as summarized in the single source Solar Simulators”, 23rd European
section 4d, and are mainly based on new knowledge Photovoltaic Solar Energy Conference (2008)
gained since the last revision. [16] Virtuani, A., et al., “Comparison of Indoor and
The major difference of edition 3 relates to the power Outdoor Performance Measurements on Amorphous
pass fail criteria. The proposal is based on both the c-Si Silicon and Micro-morph (a-Si/µc-Si) Large-area
and thin-film standards and combines the best of both Thin Film Solar Modules”, 24th European
concepts. Photovoltaic Solar Energy Conference (2009)
[17] Lokanath, S., First Solar, private communication
6. REFERENCES 2014

[1] Ross, R. G., et al, “Engineering Sciences and


Reliability”, JPL - Flat-Plate Solar Array Project
(1986)
[2] Osterwald, C. R., et al, “History of Accelerated and
Qualification Testing of Terrestrial Photographic
Modules: A Literature Review”, Progress in
Photovoltaics: Research and Applications, 17 (2009),
11-33
[3] TamizhMani, G., et al, “Failure Analysis of Module
Design Qualification Testing – III: 1997-2005 vs.
2005-2007 vs. 2007-2009”, 35th IEEE Photovoltaic
Specialists Conference – Hawaii (2010)
[4] TamizhMani, G., et al, “Regional Influence on
Module Design Quality: Qaulification Testing failure
Rate Results from Six Regional Labos of TüV
Rheinland around the world”, 40th IEEE
Photovoltaic Specialists Conference – Denver (2014)
[5] Koentges, M., et al, „Review of Failures of
Photovoltaic Modules”, PVPS - Photovoltaic power
Systems Program, Report IEA-PVPS T13-01:2014
[6] Dubard, J. et al, “Photovoltaic Module Performance
Measurements Traceability: Uncertainties Survey”,

View publication stats

You might also like