Progress in Energy

TOPICAL REVIEW • OPEN ACCESS You may also like

- Modeling and analysis of the solar
Accelerated aging tests vs field performance of PV photovoltaic levelized cost of electricity
(LCoE) - case study in Kupang
modules Rusman Sinaga, Nonce F. Tuati, Marthen
D.E. Beily et al.

- An Investigation of the Recovery of Silicon

To cite this article: Karl-Anders Weiß et al 2022 Prog. Energy 4 042009 Photovoltaic Cells By Application of an
Organic Solvent Method
Vaidyanathan Subramanian, Prichard
Tembo and Milan Heninger

- Performance of Artificial Neural Network

View the article online for updates and enhancements. and Particle Swarm Optimization
Technique based Maximum Power Point
Tracking for Photovoltaic System Under
Different Environmental Conditions
Said Zakaria Said and Lamine Thiaw

This content was downloaded from IP address 153.96.32.62 on 07/12/2022 at 08:37

 Prog. Energy 4 (2022) 042009 https://doi.org/10.1088/2516-1083/ac890a

Progress in Energy

TOPICAL REVIEW

Accelerated aging tests vs field performance of PV modules

OPEN ACCESS
Karl-Anders Weiß1,∗, Elisabeth Klimm2 and Ismail Kaaya3,4,5
RECEIVED 1
26 February 2022 Fraunhofer-Institut für Solare Energiesysteme ISE, Heidenhofstrasse 2, 79110 Freiburg, Germany
2
EK-Reliability, Kirchweg 4, 79299 Wittnau, Germany
REVISED 3
15 June 2022
IMEC, imo-imomec, Thor Park 8320 Hasselt University, EnergyVille, 3600 Genk, Belgium
4
EnergyVille, imo-imomec, Thor Park 8320, Genk 3600, Belgium
ACCEPTED FOR PUBLICATION 5
Hasselt University, imo-imomec, Martelarenlaan 42, Hasselt 3500, Belgium
11 August 2022 ∗
Author to whom any correspondence should be addressed.
PUBLISHED
26 August 2022 E-mail: [email protected]

Keywords: photovoltaics, accelerated testing, environmental engineering, field performance, service life prediction, quality assurance
Original content from
this work may be used
under the terms of the
Creative Commons
Attribution 4.0 licence.
Abstract
Any further distribution The solar conversion market with photovoltaic (PV) systems has experienced and is experiencing
of this work must an enormous growth worldwide and—according to the agreed goals of many countries to protect
maintain attribution to
the author(s) and the title the climate—will further grow over the coming decades. Investments related to PV became an
of the work, journal
citation and DOI. important financial product with the special feature of very long contract durations. A typical
setting is an operation of more than 20 years, during which the generation of electricity and also
revenues are expected. Sometimes Power Purchase Agreements with durations of up to 50 years
have been reported. Taking these long operational times into account, quality, durability, reliability,
and degradation rates become a crucial topic for the investment and for all stakeholders. PV
modules are the dominating components in this regard since they prevail the investment and—due
to their sheer quantity—are in most cases hard to replace if a system has PV module immanent
issues. Accelerated ageing tests are in general used to ensure the quality of PV components. These
tests are partly standardized, for PV mainly by the International Electrotechnical Commission IEC
and related national committees. These standards are used for type approval testing or safety
testing, which can also address legal issues. Accelerated ageing tests are also adapted to specific
needs and for example used for quality assurance (QA) of manufacturers or service life prediction
(SLP) by manufacturers or research institutes. All the efforts are taken to gain more knowledge
about the behavior of PV modules in operation and thus the accelerated tests have to be related to
normal operation. Since PV is used around the globe, the conditions vary significantly depending
on the location of installation, from dry and sunny deserts to mountain climates to tropical or
maritime climates. In addition, the type of installation has severe influence on the operational
conditions of PV modules i.e. mounted on a roof, roof integrated, open rack—or even in upcoming
applications like floating PV. The papers attempt is to give an overview on the state of the art of
accelerated testing and field performance analysis of PV modules with focus on the developments
over the last five to ten years. Developments are described and the status is analyzed regarding the
significance of tests including the latest developments and open scientific gaps related to the
envisaged correlation of accelerated tests with field performance. In the end the reader is enabled to
differentiate between reliability testing and service life prediction. The understanding for a
comprehensive approach of reliability testing including field evaluation data will be developed.

1. Introduction

The market for photovoltaic (PV) systems has experienced and is experiencing an enormous growth
worldwide and—according to the agreed goals of many countries to protect the climate—will further grow
over the coming years and decades. Investments related to PV became a key financial product with the special
attribute of very long contract durations. Typically, more than 20 years of operation are expected. Taking

© 2022 The Author(s). Published by IOP Publishing Ltd

Prog. Energy 4 (2022) 042009 K-A Weiß et al

these long operational times into account, quality, durability, reliability, and degradation rates become a
crucial topic for the economic success of PV.
Recently also sustainability assessments Life Cycle Assessment (LCA), Levelized Cost of Electricity
(LCOE) became of major significance for the renewables industry. For such assessment models the
degradation effects are essential, hence reliability and degradation data is of highest value for all investments
and stakeholders. Based on the concern for reliable PV modules and reliable statements on durability and
degradation, big efforts have been taken during the last decade. The Know-How on degradation effects and
rates as well as on failure modes of PV modules in the field and related accelerated tests were improved.
Accelerated ageing tests, with subsequent characterization, are in general used to ensure and measure the
quality of PV components and are used for a long time [1]. These tests are adapted to specific requirement
and for example, used for quality assurance (QA) [2] of manufacturers or service life prediction (SLP) [3] by
manufacturers and research institutes.
These efforts are taken to gain further knowledge about the behavior of PV modules in operation and
therefore accelerated testing need to be related to normal operation. It will enable a phenomenological
correlation of degradation effects occurring under basic laboratory conditions compared to complex
outdoor conditions. Since PV is used around the globe, the conditions vary significantly depending on the
location of installation [4]. In addition, the type of installation has high impact on the operational
conditions. Whether PV modules are mounted on a roof, or in open rack, are roof integrated—or even in
upcoming applications like floating PV, vehicle integrated PV, or in agricultural implementation—it has
significant influence on the operational conditions.
All these variations make it necessary to relate accelerated tests to field performance of PV modules. This
paper attempts to give an overview on the state of the art of accelerated testing and field performance analysis
of PV modules including the latest developments and open gaps. For detailed information on physical
degradation mechanisms and related chemistry, it is recommended to consult further literature, e.g. [5, 6],
since this would be beyond the possibilities of one publication. The paper enables to differentiate between
reliability testing and service life prediction.

2. Background

Based on the demand for reliable PV modules and reliable statements on durability and degradation, big
efforts have been taken during the last decade to improve the know-how on degradation effects and rates, as
well as failure modes of PV modules in the field and related accelerated tests.
Different directions of scientific and technological developments in the area of field performance analysis
and accelerated testing can be reported, including numerous publications reporting on degradation of fielded
PV modules and systems, as e.g. [7–10]. The ‘classical’ accelerated tests like the standardized damp heat (DH)
or UV or thermal cycling testing have been applied extensively on big numbers of modules and module types.
Testing has been done also far beyond the testing durations/cycles given in the IEC standards, delivering lots
of data which can e.g. be analyzed for a correlation between specific loads and specific failures. But still, these
tests have not been developed to simulate normal outdoor operation. Therefore, several research groups have
tried to develop accelerated tests with the aim to directly simulate the outdoor conditions and accelerate the
related effects. In this area of activity significant progress can be seen. Still, there is a lot of work to be done to
make such tests easier and especially more common, since they are still rarely applied. Other approaches aim
at developing tests to screen for specific failures. Here especially established tests for potential induced
degradation (PID) and light and elevated temperature induced degradation (LeTID) can be named.
Being aware that the degradation and reliability of PV modules depends on the specific conditions of the
location and the installation they are operated in, great efforts can be documented in the analysis of field
performance of PV modules, including correlation of operational loads with failures. Recently, approaches to
map climatic conditions and relate the local climatic conditions to degradation rates or failure risks using
degradation modelling have been proposed. These approaches and all the Service Life Prediction efforts
bring together know-how and data from accelerated testing and field performance analysis. Respectable
progress is noted but still many research questions are open.

3. Field performance of PV modules

Field performance of PV modules is the one relevant parameter, which is crucial for durability and reliability.
In the end, field performance determines the bankability and economic success of PV plants. Field
performance is also key for all tasks related to testing, qualification, and benchmark setting. This
benchmarking is used for activities and analysis of accelerated ageing. Accordingly, the modules are
differentiated regarding their field performance, and the single module performance (over time) depends on

2
Prog. Energy 4 (2022) 042009 K-A Weiß et al

the specific conditions of the location and the application/installation. The reasons for testing modules in the
field are various and rarely limited to only one single scope. But for sure field performance measurements are
an essential part to determine conditions for artificial degradation testing and for the design of such tests.
And for sure, the most important point, accelerated testing needs to be validated by field performance. The
optimization and adaption of standard test conditions relies on the understanding of effects from the fields
in different locations. The statistics of enormous numbers of accelerated ageing tests are worthless, if none of
those effects can be reproduced in the site of interest. In addition, not every module has necessarily to
withstand all the weathering conditions from extremely dry and cold to hot-humid climatic conditions. An
adaption of material is highly recommended for reliability and for substantiality aspects and for the
optimum in field performance.
This chapter analyses the state of the art of loads and operational conditions for PV modules. Included
are sections on different load parameters. Special focus is on location specific load effects and the progress of
work related to these topics.

3.1. Loads and conditions

For PV, there are no standardized radiation, temperature, or humidity conditions for operation or testing in
outdoor installations. This implies that we are dealing with non-standardized conditions and loads, when
analyzing the conditions of PV systems in operation mode. There are also neither ‘damp-heat’ conditions
without irradiation, nor cyclical temperature loads with constant humidity in the field. The various stressors
should always be considered. Beyond that, there are special cases, such as condensation and frost or soiling.
These cases are sometimes as unstandardized as bird droppings, or mechanical impacts from close traffic
which are not covered in the established tests and investigations. This makes research on loads and
conditions in outdoor applications essential to develop an understanding for degradation and reliability of
exposed PV modules. A good overview on work and results on loads and conditions and their impact on PV
reliability can usually be found in reports of IEA PVPS, mainly covered by Task 13 [11] as e.g. [3, 5, 6].
A survey in the IEA report [12] also hints to the fact, that among the vast number of laboratories and
industry with climatic cabinets not many studies have been conducted for long term PV module field
exposure. From the 15 international participants only 5 survey participants are veterans with outdoor testing
of PV modules with more than 10 years of experience and more than 100 modules tested, and on the other
side 5 survey participants are ‘newcomers’ with 2–4 years’ experience and mostly smaller test facilities.
Despite the restriction of the survey to the IEA PVPS Task 13 consortium, it represents well the state of art of
today’s module outdoor testing and experience. The survey includes 33 test facilities in different climates and
all relevant climatic zones: warm and tempered, arid, continental, tropical and alpine.
Performance and energy yield measurements of single modules is typically used in combination with
monitoring of ambient climatic conditions to analyses the correlation of PV modules degradation to
operational loads. Sometimes, but unfortunately not always, microclimatic conditions e.g. module
temperatures, are monitored in parallel. Measurements under real conditions should be done over at least
one year for representative purposes under any climatic condition. The result of such a monitoring activity is
only valid for the specific site on which it has been measured and it is also important to check whether the
monitoring time is representative for the local climate. There are some research activities trying to correlate
local climatic conditions with degradation effects, focussing on module effects [13] or on materials or
material classes [14].
The following sub-chapters will give an overview on the different climatic stressors and how they are
determined and analyzed regarding degradation effects.

3.1.1. Temperature
Temperature loads are of major importance for more or less all degradation effects since it directly influences
physical and chemical reactions, and elevated temperatures accelerate such processes [15]. Especially organic
materials are affected by temperature impacts [14]. Due to this effect, temperature is one of the most relevant
stressors influencing degradation of modules in operation. PV modules always operate at elevated
temperatures, compared to ambient conditions [16] if they are not technically cooled. For the determination
of the temperature load, it is important to use the module temperature (microclimate) and not the ambient
temperature (macroclimate) since they differ significantly. The module temperature mainly depends on the
ambient temperature, irradiation, module design, wind, and installation. There are models which allow the
calculation of module temperatures with adequate accuracy using ambient climatic conditions and type
specific parameters as input factors [16, 17]. Comparing the situation at one specific location, the type of
installation has the biggest influence [18, 19]. Open rack installations typically ensure relatively low module
temperatures while roof integrated installations lead to extremely high operational temperatures due to the
insulation effect. In extreme conditions even temperatures above 100 ◦ C have been reported [20], which is of

3
Prog. Energy 4 (2022) 042009 K-A Weiß et al

twofold severity. On one hand, degradation effects are extremely accelerated in such conditions. On the other
hand, the previous module qualification with standardized type approval tests becomes invalid, since type
approval only uses ‘accelerated tests’ with up to 85 ◦ C (see section 6.1). This means that these tests, compared
to the real operation conditions, not only lose entirely their acceleration, but they might also not even trigger
all degradation processes which take place under normal operation conditions.
For the most typical PV module technologies and designs, lots of measured and calculated temperature
load data is available, partly even transferred into global load maps [21] but for new module types or
application types like floating or vehicle integrated PV modules, no reliable data base for module
temperatures is available, yet.

3.1.2. Humidity
PV modules are exposed to different humidity levels in different climates. The presence or absence of water in
the ambience is together with the temperature load one of the main factors for degradation effects. Relative
humidity as main driver in several degradation mechanisms e.g. corrosion, embrittlement, discoloration,
optical and adhesion losses directly affect the lifetime and energy yield of PV modules. Many activities have
been reported on theoretical moisture ingress into the polymeric components [22] and the influence of the
ingress mainly as material defect-based research [23]. However, the interlink and proof of actual water
concentration in the fielded modules and reliable measurements methods are still under investigation [24].
Further, only few papers can be found which analyze the influence of the sole humidity conditions and their
effects on degradation/performance of modules in the field in a detailed way. Some papers include humidity
related research [4, 25] but none has been found, which has its main focus on humidity effects for PV
modules. Humidity data of local climates are available, since it is included in all geographic related data sets
and maps as e.g. the most known Köppen-Geiger climate categorization system [26]. For PV modules, the
microclimatic load is not identical with the ambient humidity conditions hence calculations using irradiation
and temperature data are necessary to estimate microclimatic humidity loads. Such calculations are partly
used to create degradation risk data sets and maps [4, 24] and supported by measurement technologies [27].
Publications and worldwide activities currently report on humidity effects on polymeric materials of PV
modules, specifically used in back sheets [14, 28]. The reported results show the importance of including
humidity loads in accelerated tests for qualifying materials for PV applications and a tendency towards test
with combined loads can be recognized. Nevertheless, it must be mentioned that the integral effect of
humidity loads on PV modules requires more dedicated research.

3.1.3. (Ultraviolet) Irradiation

UV radiation has been identified as a major reason for degradation of PV module materials, especially the
polymeric components, for a long time now [14, 29]. Research mainly focussed on the specific local loads
[25], sometimes including mapping or transfer into risk maps [4]. Issues with cracking of backsheet
materials in the fields after relatively short exposure times led [30] to the awareness that UV testing has to be
combined with appropriate humidity exposure to generate meaningful results [14]. Taking this into account,
it can be realized that many groups work on the development of appropriate combined testing sequences
including UV loads (section 4). It is also common knowledge in the meantime, that the standardized UV
exposure according to IEC 61215 with little as 15 kWh m−2 is in approximation only simulating 3 months of
normal operation in moderate climates, which is insufficient in relation to more than 25 years of outdoor
operation. Due to the fast-growing market of PV installations in desert regions, the extreme UV doses in
many deserts, often accompanied by high temperatures [25, 31] should be integrated in the testing and
qualification with more emphasis, to address the sensitivity of polymeric materials [14].

3.1.4. Temperature changes

Fielded PV modules, in addition to the above-mentioned loads, experience alternating temperature
cycles at different resolutions; hourly, daily, monthly, or annually. These temperature cycles induce
thermo-mechanical stress/fatigue on a PV module, which mostly leads to degradation modes such as solder
interconnections breakage, cell cracks, and glass breakage and fatigue of polymeric components [30, 32].
Additionally, temperature cycles could also accelerate degradation mechanisms induced by other stresses.
The IEC 61215 thermal cycling test (TC 200) was designed to evaluate the ability of the module to withstand
thermal mismatch, fatigue and other stresses caused by temperature fluctuations. Like other IEC tests, the
question always remains how these tests are comparable to outdoor performance. The authors in [33]
performed a comparative study of modules subjected to TC200 and TC50HF10 tests which are described in
sections 10.11 and 10.12 of the IEC 61215 standard respectively, to outdoor performance. Their findings
show that the trends in degradation where not perfectly comparable. In this regard, the development of
indoor tests that can accurately simulate real outdoor conditions is still under research. There are also
improved approaches based on finite element method (FEM) calculations used to simulate

4
Prog. Energy 4 (2022) 042009 K-A Weiß et al

thermo-mechanical behaviors of different PV modules [34, 35]. Simulation approaches are important
because of their flexibility and ability to quantify stress levels for a large variety of scenarios ranging from
process-induced stress up to field conditions.

3.2. Climate specific loads and effects

Climatic conditions are depending on the location of the installation of the PV modules and on their type of
installation, as already described before. Several research groups have analyzed the effects of the different
climatic loads, as in [36] or [37]. The papers show that there is a significant influence of the local climate on
the degradation rate but the real correlations between the stressors and the degradation processes are not
really understood. This is also valid for a correlation to specific failure modes. There are several contributions
reporting on the correlation of material degradation effects with climatic impacts, as in [38] or [39] and
some groups even try to define risk priority numbers according to the local climatic conditions [40]. The
results show that the climatic conditions, which PV modules experience, are different and cannot be simply
translated into standardized tests. As these—by nature—require that the typical operational load is
comparable in all places of application.
Apart from the analysis of specific data, the availability of data for local loads is continuously advancing.
To be recognized is that loads for PV modules differ from ambient conditions, consequently the usability of
classical weather service data is limited. Different maps have been developed, which use models to translate
the ambient climatic conditions into microclimatic loads for PV modules [41]. The approaches partly get
even further and include kinetic reaction models for different typical degradation processes and create PV
risk maps or even typical degradation rate maps covering big parts of the globe [4].
The described approaches work with the distinctive stressors, temperature, humidity, and irradiation,
typical present at all locations. Despite that, there are other load factors, which are only relevant in specific
locations. But since they can become predominant loads at these locations they are not to be neglected, as the
following sub sections show.

3.2.1. Soiling loads

The term soiling describes the dust settlement and accumulation onto the surface of PV modules. It is among
the prevalent environmental loads, of temperature, humidity, and irradiance, one of the major concerns in
reliability and yield of the exposed components and systems in the harsh environmental condition of the arid
climate. Arid areas, rich in sunshine, are of special interest, because only a small fraction of those uninhibited
deserts regions could theoretically provide the entire world with renewable solar energy. The pressing
problem is how to minimize or mitigate yield losses caused by soiling. Now, the focus of research has been on
high potential regions, such as the Middle East North Africa (MENA) as well as other desert locations i.e. in
Australia, the United States or Chile. Even tropical areas i.e. in India, rich in solar irradiance, can be utilized.
Hence, there is literally a boom of research on soiling related reliability issues in arid, and to some extent also
tropical locations in the last decade. Further comprehensive reviews can be found in [42–44] or [45] and its
update in [46]. We try to give an overview on the dominating issues and topics of the tremendous number of
publications related to soiling of the last years.
Although, worldwide R&D efforts have been intensified in the last decade, still the effects of dust
accumulation on PV performance cannot be predicted but are measured and analyzed by show cases in
defined locations. Quantification measurements with local focus e.g. in dry-maritime environment show
yield loss even beyond 80%, enhanced by a roadside construction site [47]. Yield losses up to 50% are not
uncommon for hot and dry desert environments. Outdoor measurements on utility scale are conducted to
further qualify different PV module technologies in spite of the soiling effects [48–50]. Daily soiling losses
can range from as low as 0.01% d−1 (Europe, North America) up to 1%–2% d−1 in particular arid regions of
MENA or East Asia [51]. Soiling rates at particular locations in different climate zones of South America
(Brazil) are measured to be in the moderate range of 0.15% d−1 [52].
A commonly assumed degradation rate for c-Si technologies of 0.5% yr−1 , used for warranty and
reliability calculation, has to be adjusted to 1.4% yr−1 for desert environments. This high degradation rate
can be explained by synergies of different influencing stressors, including soiling loads. Especially corrosion
effects (see section 3.2.3) are enhanced by soiling. Common knowledge is that the infrastructure of the
solar farm and installation conditions (tilt, orientation) does strongly influence the soiling effects [53].
Instrumentation to measure and monitor yield losses include soiling sensors [54], or the comparison with
the electrical parameters of frequently cleaned reference cells/modules [55, 56].
Certain is, there are location-depended dust adherence mechanisms, traced back to different dust
attributes. Composition and particle morphology of dust grains and soiling components are under research
with different analytical tools e.g. by microscopy, environmental scanning electron microscopy, x-ray
diffraction or x-ray fluorescence [57–59]. To minimize the issue of the location dependent yield loss due to

5
Prog. Energy 4 (2022) 042009 K-A Weiß et al

dust composition, suitable mitigation approaches of active and passive nature are to be selected. As active
methods cleaning solutions are applied, like manual or automatized mechanical cleaning or more advanced
electrodynamic systems [60, 61]. Optimization scenarios and economic aspects are considered within the
cleaning studies [50, 60]. Still under research are the long-term effects by the different cleaning systems on
the durability of the system and surfaces since cleaning requires cycles at least 2–4 cleanings per month in
high soiling-risk locations. In such arid regions water scarcity, high wind velocities and an increased
concentrations of airborne particles are also factors to be included in reliability considerations. Ambient
conditions and environmental loads play a major role [50, 60].
Passive mitigation solutions, such as ‘easy-to-clean’ coatings, interact with the dust layers itself to
supposedly prevent the dust to build strong bonds with the surface. There are hybrid coating solutions, being
under test [62, 63]. Interactions of coatings and deposited dust [60, 64], but also with the mechanical
cleaning methods are simulated in indoor testing. This enables to better study soiling behavior e.g. with fine
particles [65], the transmittance loss and reflection properties of solar glass with dust coverage [66–68].
The durability of material e.g. surface roughening by abrasion [58, 61, 69] or degradation effects by
accelerated testing is also under investigation, such as accelerated ageing tests of anti-soiling coatings
[70, 71].
Modelling soiling further includes approaches, such as mathematical models, numeric approximations
[72], trend analysis with data mining [73] or machine learning methods [74], as well as artificial neural
networks [75–77]. Maps showing approximations of the regional soiling risks can be generated with
Geographical Information Systems [78, 79]. Still, further research is necessary to completely cover and
understand the multidimensional issue of soiling.

3.2.2. Mechanical and snow loads

Mechanical loads, such as wind, snow and ice accumulation on PV module surfaces have an influence on
performance and reliability. They cause among other things cracked cells, cracked glass and bent alloy
frames [80].
In three of the five climate zones, the moderate, subpolar, and polar climate zones, snow fall, icy
precipitation, and accumulation are to be expected and their harmful consequences to be mitigated. Those
consequences involve damage to the infrastructure, yield losses, mechanical stress on the module frames,
glass, and PV cells as well as electrical safety issues due to harmed insulation. Energy losses of up to 15%
annually, or 90% per month, are reported for the US [81, 82]. Studies of optical effects on transmittance and
albedo by coverage and shading reveal differences in the uniform and partial shading [83]. Also, the tilt angle
has a strong influence on a snow coverage, such that inclinations up to 80◦ and 90◦ show the least coverage
[84]. Although the mechanical load tests on PV modules, according to IEC 61215, are carried out, potential
failures of non-homogeneous loads on inclined modules rather than horizontally oriented ones give
completely different load characteristics. After only 15% of the load used in the standardized test, frames start
to deform [5]. IEC 62938 provides a method to simulate non-uniform load characteristics and defines a test
procedure accordingly for framed PV modules since 2020 [85]. In the coming decades, North Europe awaits
increasingly heavy snow loads on buildings according to reports in the architectural sector, due to warmer
climate, which enforces wet winter-precipitations [86]. Hence, the mechanical damage through long-term
heavy snow loads on different module construction styles is still under research [87]. The aim is to qualify PV
modules as snow-load resistant. Module properties less disposed to snow load damage might be, but are not
exclusively, modules with silicone-based adhesives, frameless and of shorter height [5]. Also, the sliding
behavior off and on module surfaces are under research in connection to the melting behavior of snow on
modules. Results from this research is the recommendation for rather larger and frameless modules [88].
Diverse mitigation solutions are suggested [89] i.e. to mitigate effects of heavy snow loads, other than
shoveling the snow from the modules, using reversing current to heat up the modules [90, 91]. Although, the
freezing of meltwater poses a risk in damaging the PV modules, it seems that the issue was solved and a
working system is implemented [92]. Besides that, a removal of snow always poses the risk to damage the
solar module [93].
Furthermore, wind loads are also a relevant aspect for mechanical loads placed on PV modules as they
occur in almost all applications. The specific characteristic of wind loads is the highly dynamic behavior [94,
95]. Highest wind loads occur as gusts and consequently are hard to measure. Gusts can also trigger
vibrations of modules [96]. Numerical tools are used to analyze the reaction of PV modules on such wind
loads [97], sometimes supported by wind-tunnel experiments [98]. It is reported that the type of installation
and the design of the modules, especially the aspect ratio has strong influence on the occurring loads [99]. It
has also been shown that the design and the aspect ratio lead to type specific eigenfrequencies of modules
[96, 100], which can be correlated to the susceptibility to gusts with specific properties. Typical failure modes
of dynamic mechanical loads are cell cracks [101].

6
Prog. Energy 4 (2022) 042009 K-A Weiß et al

3.2.3. Corrosive loads

One of the chemical degradation processes of PV modules, determining the whole reliability, is the corrosion
of the single components [102]. Corrosion research in the field of PV reliability is especially in maritime
climates of significance, due to location-specific weather conditions. Methods for mapping atmospheric
corrosivity have been presented [103]. Further, exposing components in multiple climates to investigate
reactions make reliability and corrosion stability testing more realistic, since the corrosion rate of the
atmospheric corrosivity is controlled by time of wetness [104], temperature and electrolyte composition
[105]. That makes the atmospheric corrosion in maritime locations more likely to be destructive for the bill
of materials (BoM) of PV modules, caused by present salt, high levels of airborne chlorine, paired with high
moistures levels and usually also elevated temperatures. Research on aluminum and alloys present these
effects on module frames [106].
Ambient environment may further contain corrosive gases. The most publicly noted one is the bad
smelling and hygroscopic Ammonia. Questions arose, if PV modules are affected in the presence of elevated
concentrations, since Ammonia attacks and corrodes specifically copper and its alloys in the presence of
moisture [107]. Up to now, the authors know of at least two in the last decade designed tests to qualify
modules for such environment but have not seen reports of degradation caused by those specific
environmental conditions under normal operation conditions.
Humidity (section 3.1.4) is crucial for corrosion effects within the module. Moisture ingress is the
precursor of most degradation effects, such as metal grid corrosion and loss of adhesion [23]. However,
humidity is the carrier for pollutants and hence activator for electrochemical reactions, too. Degradation
effects in aquatic systems have been of interest long before and are still of interest for system materials around
and at the PV module [108]. Now, with the implementation of floating PV technology, PV modules are
exposed to constant wetting. A controlled and reproducible test with salt water for the exposure offshore is
applied to understand the effects. Up to 14% performance loss after 30 days were stated for all technologies.
Whereas alterations of transmittance through a salt deposition does not have significant effects [109]. Under
research are also the effects of desert conditions on the corrosion of modules, where also humidity and
condensation next to ambient pollutants are present [110] (see also section 3.2.1 soiling).
Corrosive loads have been studied in harsh environment and have been shown to have influence on the
PID risk, hence the overall performance of the system [111, 112, 112].
Measurement of ambient chloride deposition can be done by the wet candle method, according to ISO
9225, with a rain-protected wet textile surface [113, 114]. Other analysis majorly focuses on the impacts of
the degradation on solar cells. Contact failures of front contacts [115], and soldered bonds are used as basis
for lifetime prediction of PV modules [116]. Surface corrosion on solar cell level can become noticeable as
snail tracks [117].
PV module degradation analysis can be carried out with Raman spectroscopy on the polymeric
encapsulant to understand the decomposition and the release of acetic acid under high moisture levels
[118, 119]. The release of acid and its corrosive effects in contact with surrounding module component, such
as interconnectors have been analyzed and described [120].

4. Accelerated aging: definitions and purpose

Accelerated ageing tests are widely used and address different purpose. Some general terms are important to
understand to evaluate tests and data. This chapter gives an overview of the general purpose of accelerated
aging tests and its basic terms.

4.1. Accelerated aging

Accelerated ageing describes tests which aim to trigger degradation processes of materials or products and to
identify design flaws during product development, to prevent failures during field operation and improve
product reliability. The tests are meant to speed up possible destructive processes and loss of functionality of
material and take place significantly faster than under normal operation condition. Reliable results shall be
achieved in a very short period and at low costs. The tests can be performed with different purpose and
scientific approaches, as described in sections 4.4, 6 and 7. A very generic description given by the
Confederation of European Environmental Engineering Societies CEEES can be found in [121]. To enable
accelerated ageing, the degradation provoking factors must be increased. There are in general three
possibilities to achieve the acceleration:

• constant, representative high stress level of one or more degradation factors for a long period, in PVs for
example irradiation around the clock or constantly high temperatures

7
Prog. Energy 4 (2022) 042009 K-A Weiß et al

• Increasing the stress factor(s) above normal operational stress level to apply higher doses to the sample, in
PVs for example testing at higher than usual UV-intensity or relative humidity,
• Increasing the temperature to accelerate chemical degradation processes.

All types of acceleration the natural aging come along with risks. So, it is important to be aware that
increased loads can provoke processes and effects which are not relevant for normal operation [122]. It is
equally important to consider that in most accelerated tests, the relative intensity of the different present
stressors to each other is changed in comparison to the normal operation.

4.2. Acceleration factor

Following the objective of accelerated ageing, processes that occur in normal operation shall be accelerated.
But typically, the aim is to keep the process and reaction types similar. Similarity is represented by the
acceleration factor (AF), which is defined as
ttest
AF =
t0
with t test the testing time during accelerated ageing; and t 0 the exposition time under normal conditions,
which is necessary to provoke similar effects.
Since different physical and chemical processes show different dependencies on stress factors, they then
have different acceleration factors. Stay aware, that one acceleration factor is typically only valid for one
specific process. In a given material sample many (chemical and physical) processes take place during
accelerated testing. For each of the processes one specific acceleration factor is effective, which means that
only one general determined or chosen acceleration factor is invalid for many of the ongoing processes. For
the generation of measurable degradation effects, typically several processes interact. Still, to be able to use
the concept of accelerations factors in applied testing, often one acceleration factor is calculated and applied.
It describes the acceleration of one occurring and targeted degradation effect, which can be measured. It is
usually not a specific (chemical) degradation process. In PV e.g. yellowing of encapsulation materials or
performance loss of PV modules are often chosen for such measurements and not the (chemical) process
itself behind.
For mathematical modelling the idea of the rate domination process has been developed [25]. This
approach estimates that one physical or chemical process e.g. hydrolysis or UV drive chain scission, is
dominating the development of the degradation effect of interest. So, one relevant acceleration factor can be
determined, which is representing the behavior of the sample regarding one specific degradation effect.

4.3. Stress factors

PV module degradation is dependent on multiple factors such as manufacturing processes, installation sites,
mounting conditions and the type of the module. To understand the long-term behavior of PV modules, it is
a prerequisite to assess the stress factors acting on the modules and correlate the module’s response to the
applied stresses. Stress factors determine the degradation of PV module components and hence the service
life of PV modules. Accelerated aging tests are designed to simulate the effects of different stress factors on
different degradation mechanisms, identified in the field. Several stress factors relevant for PV reliability are
discussed in literature [3]. In table 1 major stress factors and the corresponding accelerated aging tests are
given based on IEC 61215 for crystalline silicon modules, IEC 61646 for thin film modules and IEC 62108 for
CPV modules. The emergence of new photovoltaic designs and applications (e.g. offshore/floating PV,
vehicle integrated PV) calls for new tests beyond existing standards to best describe their degradation
behaviors under different stress factors.
In the evaluation of PV module reliability, two distinctions are made from climatic stresses: macroclimate
and microclimate stresses. Macroclimate stresses describe the ambient weather conditions around the PV
module and microclimate describes the conditions experienced by the specific PV module components. For
example, ‘ambient temperature is macroclimate, and module temperature is microclimate’. This means that
microclimate stressors have a direct influence on PV module degradation mechanisms. In degradation
evaluation and predictions, microclimates are used to describe the degradation kinetics. However, it should
be noted that there is a strong correlation between macroclimatic and microclimatic conditions. In most
cases, microclimatic conditions are not directly measured but estimated from macroclimate conditions using
different mathematical models [123, 124]. Therefore, the accuracy of these transformation models is crucial
for understanding and correlating the degradation kinetics to stress factors [125].

4.4. Quality assurance

An important driver for the development and application of accelerated ageing tests is quality assurance
(QA). In all established industries manufacturers use accelerated ageing tests to ensure a defined quality level.

8
Prog. Energy 4 (2022) 042009 K-A Weiß et al

Table 1. Stress factors vs accelerated aging test and degradation modes for PV modules.

Stress factor Accelerated aging test Degradation mode/mechanism

Temperature and Humidity Damp Heat Corrosion

Encapsulant adhesion
Junction box adhesion
Delamination
Temperature cycles Thermal Cycles Broken cells
Junction box adhesion
Broken interconnect
Solar Irradiance UV test Encapsulant and backsheet discoloration
Cracks in polymeric backsheets
Delamination and bubbles
Frost Humidity Freeze Junction box adhesion
Delamination
Inadequate edge deletion
Wind and snow load Static Mechanical Load Broken glass
Broken interconnect ribbons
Broken cells
Structural failures
Electrical bond failures
Wind Gusts Dynamic Mechanical Load Broken glass
Broken interconnect ribbons
Broken cells
Electrical bond failures
Hailstones Hail Test Broken glass
Broken cells
Salt water and salt mist Salt Mist Spray Corrosion

Such tests shall help to identify issues of products during production or directly after production. The aim is
to identify problematic products, and to be able to take measures in production or stop products before they
enter the market. For that purpose, it is essential that tests are quick, highly reproducible, easy to apply and
not too costly. This limits possibilities for the definition of QA tests, which is especially problematic for
products with very long expected lifetimes—as PV [126]. Nevertheless, most of the standardized tests, as
described in section 6.1 are also applied for QA purpose. In addition, several of the failure mode specific tests
described in section 6.2 are typically additionally applied for QA e.g. PID testing to qualify cells. In most
cases are DH, thermal cycling (TC) and UV testing applied by the manufacturers, sometimes even with
testing durations exceeding the requirements of the standards [127]. Due to their test durations of typically
several weeks, they do not fulfill the requirement of delivering results during production.
Another aim of QA is to test as many samples as possible. In ideal case, each sample should be tested
before it leaves production. This is unfortunately not possible with accelerated ageing tests, since they affect
the sample irreversibly and can degrade the product severely.
Besides manufacturers, sometimes QA type testing is also applied by/for investors, importers, or traders
to ensure quality, or it is requested by banks or insurance companies. Many of them have specific testing
protocols [2] which have to be applied to come into business. All these QA tests or testing sequences just
define a specific level of loads for product quality evaluation, but they do not correlate with specific
applications, climates, locations, or service lifetimes.
In PV industries, the importance and awareness of QA and especially meaningful and comparable QA
testing procedures has been increased significantly over time. A catchable result is the meanwhile established
institution of the International PV Quality Assurance Task Force (PVQAT) which leads global efforts to craft
quality and reliability standards for solar energy technologies. Now 12 Working Groups are active in PVQAT
addressing different important QA topics [128]. PVQAT already developed several inputs and proposals for
standardization of IEC level and established numerous round-robin tests or other measures to improve
comparability and know-how on accelerated testing and QA in PV.

5. Testing equipment, conditions, and related limitations

Accelerated testing requires adequate equipment, which comes with specific properties and yet limitations.
Depending on the purpose of testing, different equipment has to be used. The following sections give
insights in the world of accelerated testing equipment with special focus on developments over the last years.
Furthermore, set-up related testing conditions, approaches and limitations are presented.

9
Prog. Energy 4 (2022) 042009 K-A Weiß et al

As side note, soiling testing equipment and conditions are not included since accelerated (laboratory)
soiling testing is not seen as a typical accelerated ageing test. Soiling typically causes (partly) reversible effects
and so not classical degradation. Soiling nevertheless can influence degradation and ageing. Please see
sections 3.2.1 and 3.2.3 regarding these topics. Description of laboratory soiling setups can be found in the
literature given in section 3.2.3.

5.1. Loads and conditions

For all accelerated ageing tests, the testing conditions must be carefully determined (see sections 3.1 and 4.3).
The selected test conditions then specify the needed test equipment. Selection criteria are e.g. the number
of controlled load factors or the size of the samples. In many cases this logical decision process can
unfortunately not be used. There are two reasons. First, available equipment is limiting the flexibility of the
test set-up. Second, technological limitations of equipment make some tests impossible or increases the effort
inappropriately. The following sub-sections analyze the situation for the different stress factors.

5.2. Temperature testing

Temperature has a major influence on the reaction rate and reaction kinetics of all chemical processes and for
that reason, also on degradation processes of PV modules, as described in section 3.1.1. The temperature in
accelerated ageing tests is therefore of highest importance and influencing all processes. Sole thermal ageing
by exposing products to elevated temperature conditions is done in oven like thermal chambers. Small
chambers can heat up to 750 ◦ C or even above. Typical standardized testing temperatures in PV are around
80 ◦ C–90 ◦ C.
In irradiation testing, the sample temperature can differ from the ambient chamber temperature.
Irradiation can significantly elevate sample temperatures. For evaluation of the loads, the sample
temperature is a key factor. It must be analyzed or taken as control parameter. For several chemical processes,
especially for organic materials—in PV mainly the polymeric encapsulation and backsheet—specific
degradation processes can only commence above a material-specific threshold temperature. So, if the device’s
normal operation is below such a threshold temperature and testing is above, different processes will be
triggered. Then the accelerated test cannot deliver any reasonable information on temperature degradations
the normal operation mode [129]. For the most common encapsulation material EVA such a threshold
temperature is in the range of 85 ◦ C–90 ◦ C [130]. Due to the high impact of temperature on reaction rates,
also temperature homogeneity within the tests is very important to ensure the comparability of results.

5.3. Relative humidity testing

Relative humidity exposure in accelerated ageing tests on PV modules is facilitated in climatic cabinets,
which are available in all sizes. Most make sure to fit at least 2 PV modules simultaneously. The commonly
used test for humidity exposure is the so-called DH test at 85 ◦ C and 85% r.h. for 1000 h, as defined in IEC
61215. This test is typically applied for quality assurance purpose, often also in extended versions with up to
3000 h or 4000 h. There are only few developments recognizable which try to vary the conditions e.g. to
determine acceleration factors, which would be necessary to correlate the test with outdoor exposure [131,
132]. Unfortunately, no reasonable progress regarding correlation of DH-tests to real outdoor conditions can
be mentioned, which is a severe shortcoming since DH testing is probably the most used accelerated test in
PV industries. Phenomenological degradation effects on PV modules e.g. interfaces (EVA-Glass) after DH
exposure are reported [133]. The material-oriented research shows, that moisture often plays a synergetic
role with UV, as e.g. for surface crack formation of backsheets [134].
It is important to set and design the accelerated ageing tests properly and adjust them for real loads not to
rule out the good material (see sequential combined stress testing) [135].
When talking about humidity testing, it is important to mention the water vapor permeation rate
properties of PV modules, mostly the polymeric components, such as backsheets or encapsulants for
Glass-Backsheet modules or edge sealants for Glass-Glass modules. Degradation effects of water permeation
into encapsulant or backsheet during damp-heat testing is described in recent publications [133] by a
consortium of PV and polymer research institutions, in the framework of the Solar-Train project
(www.solar-train.eu). There are set-ups to determine the WVTR and barrier properties of polymeric films,
which consist of a permeation cell connected to a mass spectrometer. The tests are performed under elevated
DH conditions and can vary, but usual settings are at 38 ◦ C and 90% r.h.

5.4. UV testing
UV testing of PV modules and materials is commonly used for qualification testing for a long time. On
module level, typically the UV exposure with a dose of 15 kWh m−2 at 60 ◦ C defined by IEC 61215 is often
used as a benchmark. The dose of the test is only comparable to approximately three months of outdoor

10
Prog. Energy 4 (2022) 042009 K-A Weiß et al

exposure in moderate climates. The test is therefore only named ‘UV preconditioning’ in the IEC standards.
Different set ups are used to apply the irradiation, using Metal-Halide or Xenon lamps (Xe), fluorescent
tubes or as coming technology, LED light sources [136]. Xe sources provide a spectrum, which is close to the
solar spectrum. Consequently, Xe sources transport alongside much heat to the sample, because of its
longwave IR irradiation. To avoid the extra temperature influence by heating the sample, fluorescent tubes or
LEDs are used, even with higher UV intensities. They make it easier to test at controlled sample
temperatures. The different light sources also deliver different spectra [137], which can be problematic
especially, if results of different tests shall be compared. That is a major shortcoming for current UV testing.
Since the PV community is aware of the shortcoming of the IEC type UV exposure, typically multiples of
the 15 kWh m−2 are applied. But usually, the used doses are still far below the typical lifetime doses of
modules. One reason for that is the high cost of testing, since samples cannot be stacked as in a DH or TC
testing set-up. They rather need the full frontside area exposed to the irradiation. A further reason is the
limitation of acceleration, caused mainly by thermal effects of light sources. The limited size of many UV
testing cabinets and the importance of UV testing for polymeric components [14] are major reasons why
high dose UV testing is often only applied to smaller material samples or mini sample modules. Currently, in
the last five years since approximately 2016, polymeric backsheet materials are of special interest for industry
and research. Since backsheets are directly exposed to the environment including UV loads, occurring
failures e.g. cracks, are now studied intensively. Crack formation, one of the most common failure modes of
the backsheets, leads to safety concerns. A special type of crack, channel cracks (fragmentation) were only
observed on samples treated with UV [134].

5.5. Thermal stress testing

Temperature cycling (TC) tests are mainly applied to analyses the effects of internal mechanical stresses due
to different thermal expansion coefficients of module materials. These tests are regularly applied to qualify
material, but there is not much development notable regarding scientific analysis of the tests or development
of new or adapted TC tests. The temperature cycling sequence defined by the standard IEC 61215, which uses
temperature ramps from −40 ◦ C to +85 ◦ C and back for 50 or 200 times, is almost exclusively used. There
are only few activities reported, which vary the conditions significantly. Typical variations of the IEC based
tests only touch the numbers of cycles. Often extended IEC 61215 thermal cycling (TC) tests are reported
with up to 800 cycles or even more. One approach would be to use highly accelerated thermal cycling tests,
which apply faster temperature changes [138]. Other researchers propose increasing the temperature range
or ramp rates [139]. Others have conducted studies to generate temperature cycle profile from in-situ
climatic conditions to improve the accuracy of thermo-mechanical degradation [140].
More dedicated work to develop correlations between the accelerated TC tests and thermal stress effects,
which occur at normal operation mode, would be necessary, since no validated model for this question has
been reported so far.

5.6. Corrosion testing

Accelerated corrosion tests aim to simulate degradation effects of atmospheric corrosivity in fielded PV
modules. These tests shall help to determine the corrosive effects during service life and qualify material for
corrosive atmosphere. To start and accelerate corrosion processes the established standardized test method is
an exposure to enhanced aggressive conditions compared to which a module will experience in normal
operation. In general, full-size modules or components are exposed to the corrosive effects of salt mist in
climatic chambers. During those salt spray tests a combination of saline atmosphere, high temperature and
high relative humidity is applied. However, long term studies found poor correlation between such
accelerated corrosion tests and field exposure, still they are incorporated in many product qualification
specifications [102]. Most studies in R&D on PV modules only investigate the effects of single loads, i.e.
temperature cycles [141] or moisture levels, including condensation, on the corrosion process
[115, 118, 120, 142].
Atmospheric corrosion is an electrochemical process, which combines humidity as electrolyte and
ambient pollutants. Such conditions are simulated in salt spray chambers [143]. Accelerated tests with such
specific climate chambers are based on normative standards [144–146].
These testing conditions were, originally intended for other purposes, i.e. DIN 60068-2-52 (cyclic testing)
for corrosion resistance in marine environment, transferred to test PV modules. Even if it is known that such
tests do not imitate the various climatic conditions worldwide and are less inspired by the atmospheric
corrosion of fielded PV modules, they are used. However, indications can be made on corrosive effects e.g. on
the metallic components such as the module frame or interconnectors, or on other components, i.e. coatings
or glass [115]. Also, effects of acetic loads on metallic cell materials have been analyzed, as e.g. on silver-ink
and silver-based metallization in PV modules [147].

11
Prog. Energy 4 (2022) 042009 K-A Weiß et al

There are many specifications made for the salt spray test. Grouped are the tests into neutral salt spray
test, acetic acid salt spray test, copper accelerated acetic acid salt spray test, and cyclic salt spray test. The
classic salt spray test, called neutral salt spray test, originally designed for paints and coatings is commonly
used for electronical devices. It is carried out under constant spraying conditions at 35 ◦ C and with a 5% salt
solution of sodium chloride (NaCl) for several hours and up to one week. However, PV modules are found to
be quite durable in such testing conditions [110]. That, and the fact that there are limitations to the chamber,
make such test long and costly. Limitation to the chamber account to space restrictions, since modules
cannot be stacked in the chambers, because the salt mist settles vertically onto the surfaces. In recent studies,
alternatives are presented, and conditions are optimized in regard of adding demanding and realistic
alternating procedures, i.e. temperature cycling (TC) and a sequence of humid and dry times. Several UV/salt
mist/DH sequences are established to test the impact of weathering factors in combination with corrosive
atmosphere on PV modules [110]. Tests and sequence from other sectors, such as the car industry or solar
thermal research [148], do influence the test design. Tests, which are characterized by several salt spray stages,
variations of salt concentration, stages of higher and lower levels of temperatures, as well as higher and lower
levels of humidity with subsequent resting stages have been reported. They put the sample under additional
stress due to the frequently changing environment [149]. These tests can also include humidity-freeze
cycles [150].

5.7. Mechanical testing

Accelerated mechanical testing is mainly done using the standardized tests according to IEC 61215, as
described in section 7.1. These tests are sometimes varied using higher loads or more cycles to increase the
effect of testing. Nevertheless, the test still has significant weaknesses. On one hand, it applies static loads,
which would typically represent snow loads. But the test is applied at room temperature, which means
material behavior can be different than in outdoor operation when snow loads occur. Especially polymeric
materials show temperature dependent mechanical properties. On the other hand, it applies loads in
different directions (up and down) which would represent wind loads but wind loads occur as gusts which
means highly dynamic. Some work has been reported addressing these issues [151]. The impact of testing at
cold temperatures has been analyzed [152], as well as effects of dynamic mechanical testing [153, 154]. Some
groups also addressed effects of non-uniform loads or asymmetric module designs [155] and the effects on
specific failures, as cell cracks [156].
Hail impact testing is not described here since hail typically leads to spontaneous catastrophic failures
and so the testing can also not be seen as a classical test, because accelerating degradation effects take place
over time. There is also no relevant development in hail testing to be noted.
Abrasion testing is typically addressed in the framework of soiling analysis and is also not covered here in
detail. Nevertheless, an adapted abrasion testing standard for PV has been published recently: IEC
62788-7-3:2022 ‘Measurement procedures for materials used in photovoltaic modules—Part 7–3: Accelerated
stress tests—Methods of abrasion of PV module external surfaces’.

6. Testing approaches for accelerated ageing tests

The stipulation of appropriate accelerated ageing tests must take several parameters into account, as
described in sections 3 and 4. It has a significant influence on the outcome and therefore also on the usability
of results. This chapter will give an overview on general approaches and terms as well as ongoing
developments and discussions.

6.1. Test design

Sections 3 and 4 describe possible purposes for accelerated testing and the stressors/load parameters and
conditions, which must be included in accelerated tests, depending on the specific goal. These parameters
open a multidimensional space of loads, intensities, durations, and sample types, which must be included
when a specific test is designed. Furthermore, appropriate characterization methods have to be identified. It
is therefore recommended to carefully analyze the needs and possibilities defined by the testing purpose, the
application, and the available testing infrastructure in the experimental design phase. Few guidelines
dedicated to PV can be found in literature [122] but besides that, literature from other industries using
environmental engineering and accelerated ageing tests for long time can be used. Especially automotive
industries offer a lot of experience in this topic [157].
As soon as the testing purpose and the technical capabilities are known, the sample design must be fixed,
if no full-size samples shall be used due to technical or financial limitations. The samples must include all
materials/parameters, which are relevant for the specific question. So, if chemical reactions/degradation shall
be provoked, it is important that all reaction partners are available in comparable amounts as in normal

12
Prog. Energy 4 (2022) 042009 K-A Weiß et al

operation. For tests addressing mechanical topics, especially the aspect ratio and often also the total
dimensions are necessary to be fixed values. The sample design also must include requirements, which occur
from the planned characterization methods to. This is of special importance if samples are used with designs
different from normal modules to enable e.g. humidity ingress monitoring.
In the next step the test parameters have to be set to ensure that the expected reaction of the sample is
enabled. Therefore, all relevant influencing loads for the physical or chemical reactions must be included. For
example, if a UV-driven degradation process is expected, but it requires the presence of H2 O for the reaction
to happen, a plausible level of water and irradiation must be ensured. The determination of testing
parameters can be supported by numerical simulation, typically using FEM or computational fluid dynamics
(CFD) [158].
Looking on the BoM of the samples, specific threshold loads for the included materials must be
identified. To know and respect the threshold loads avoids provoking processes in the accelerated tests, which
are not occurring in reality. It must be ensured that the test is performed in a load regime, which triggers
only wanted and similar to the real operational degradation. Similar induced degradation processes,
compared to the ones of outdoor operation, warrant meaningful results. The most critical load parameter for
this issue is typically temperature. Testing at too high temperatures is especially critical for organic materials.
Using too high intensities provokes unrealistic effects and leads to failures, which are not representative and
relevant for a normal operation. With testing parameters exceeding the threshold values, it is more likely to
terminate all types of samples instead of generating valuable results. The aim remains, to address a functional
and sustainable development of PV industry, see also section 6.4.

6.2. Sequential vs. combined testing

There are different approaches to address the issue of serving multiple load factors/stressors in accelerated
ageing tests (see also sections 2–4). One approach to meet the requirement is designing tests where the
different load factors are applied in a serial way and still be effective, even though they are not happening at
the same time [159]. Sometimes serial testing is used to narrow down on influencing factors to address
specific failure modes one at a time [160]. This approach is called sequential testing. It is used in most tests of
the IEC 61215 series and is also most common in the laboratories testing PV modules. The reason for the
dominance of sequential tests is the reduced requirements for testing equipment and less complex test
designs and so lower costs.
During the second approach samples are exposed to all relevant load factors for the degradation effect of
interest in parallel at the same time. The method is named combined testing. The load factors, which must be
included during combined test setups can be numerous. Especially if the overall degradation of a PV module
is under research, without a pre-defined focus on a specific failure mode, all load factors must be included.
This makes testing equipment extremely complex and is—in full consequence implemented—technically
almost beyond the realm of possibility. In application typically, the trinity of load factors temperature,
humidity and UV radiation are combined in test sequences. These combined tests use specially designed
equipment [104]. They are often named UV-Damp-Heat tests since the applied conditions are close to the
IEC type damp-heat and UV exposure conditions. These tests are, due to high hassles, only sporadically
applied for full size modules. However, combined testing of small samples is in common usage, since small
climatic cabinets are easier to employ and less expensive [133]. These combined tests could prove, that they
are able to reveal effects which are not present when load factors are applied separately. In [133] is described,
not surprisingly but now proven, that EVA as standard encapsulant for PV cells suffers the strongest chemical
and optical degradation when high UV, high temperature and high humidity are combined simultaneously.
Sometimes so-called highly accelerated stress tests (HAST) are also mentioned when talking about
combined testing. These tests typically do not address the purpose, idea, and approach as mentioned above.
HAST also combines some loads but typically use intensities far above normal operation, including the risk
of testing above threshold intensities (see section 6.1). Probably the most common HAST is the
pressure-cooker test sometimes used for material testing in PV [161, 162].
Several research groups are working on the development of combined testing sequences, to simulate
outdoor conditions more realistically. There are premature efforts of addressing different climates and using
full size climatic cabinets. As leading target value is for example the microclimatic humidity level within the
modules used [104]. It has also been reported that the exposure to combined loads leads to degradation
patterns which are close to patterns after outdoor operation [163]. Further work analyzed the pros and cons
of the different testing approaches [164]. Latest developments combine ideas of sequential and combined
testing to reach the goal of ensuring realistic load levels of humidity, temperature, and UV. Throughout the
hybrid tests, there are parts with combined loads (with UV, T, and RH) and sequential parts (in this case
temperature cycling) to keep effort and technical requirements limited. For the module accelerated sequential

13
Prog. Energy 4 (2022) 042009 K-A Weiß et al

test, a development over time is recognized [165–167] for example, with tests designed for loads on backsheet
materials. Work has also been reported including external mechanical loads into combined tests [168].
All the work shows that the definition and development of the tests, sequential or combined, require
numerical simulations to calculate the microclimatic loads during field exposure or accelerated testing. The
reason behind their need is that these parameters are defining the relevant loads for each sample and
therefore must be used as guiding parameters.

6.3. Test to failure vs fixed test duration

The total duration of accelerated tests can be defined with different approaches. On one hand a pre-defined
sequence or number of sequences, defined according to section 5.1, can be applied. Doing so, all parameters
of the implementation of the test, namely duration, intensities, total loads, intermediate steps etc are
pre-defined and the test is terminated when the planned end is reached. The samples are typically
characterized initially and at intermediate steps and/or at the end of the test sequence using pre-defined
measurement technologies. The sample behavior is then described using measured degradation effects or
performance parameters and, if necessary, sample rankings are generated accordingly. This approach is used
when the test is expected to simulate a specific threshold load level. Samples withstanding these loads are
seen as qualified for the application.
On the other hand, samples can be exposed to a pre-defined test or testing sequence without pre-defined
duration. In this case a specific failure effect or failure mode is defined before starting the test and the samples
are exposed until they show the defined failure. This so called ‘Test to failure’ approach is often used if failure
kinetics are not known, but the failure type is seen as critical. They are there to identify samples which can
withstand the testing environment as long as possible without showing the specific failure. Samples are
ranked afterwards according to the exposure time until failure. For the test to failure approach, suitable time
steps to take breaks from testing for (material) characterization must be defined. Suitable non-destructive
and non-disruptive characterization methods must be defined to identify the failure as soon as it occurs.

6.4. Limits for acceleration

One of the genuine goals of accelerated testing is to use an acceleration factor as high as possible. It makes
tests shorter, typically cheaper and delivers results faster, what is a key element in development processes.
Especially in PVs, with expected operational times of several decades, high acceleration factors are desired
and crucial. Unfortunately, the possibilities for acceleration are limited if meaningful results shall be
generated. Most important are material related limitations for acceleration (compare section 4.1). All
materials have different load regimes and, in the regimes, different physical and/or chemical processes occur.
To ensure meaningful results of accelerated testing, the accelerated test must trigger similar processes as in
normal operation, which requires the sample to be in the same load regime as in outdoor operation.
Prominent examples for such material-related limitations are temperature limits of polymeric materials.
Testing Polyolefins at temperatures above 100 ◦ C will for example certainly provoke degradation, but not
within the similar processes as in operation at below 80 ◦ C. That means a transfer of the accelerated test
results to normal operation conditions is not permitted [1, 166, 168, 169].
Further limitations can be due to technical limitations of equipment if e.g. climatic cabinets cannot
ensure constant or required temperature or humidity levels because of the parallel use of high intensity
irradiation.

7. Established accelerated tests

Accelerated testing has been used in PVs for a long time and there are several very established tests, which are
applied all around the world. This chapter gives general information about new developments of the most
established tests and testing purposes, which are type approval standards and failure specific tests.

7.1. Type approval and safety standards

Accelerated ageing tests, included in the international type approval and safety standards, are the most
commonly applied tests and sequences for PV modules. The tests are defined by the Technical Committee TC
82, Working Group WG 2 of the International Electrotechnical Commission IEC. Their latest version is the
IEC 61215–2:2021 Terrestrial photovoltaic (PV) modules—Design qualification and type approval—Part 2: Test
procedures.
The safety standard IEC 61730–2:2016 Photovoltaic (PV) module safety qualification—Part 2:
Requirements for testing included nearly the similar accelerated ageing tests on module level. Thus, they are
discussed together.

14
Prog. Energy 4 (2022) 042009 K-A Weiß et al

The standardized tests aim to ensure a certain quality and safety level, which is not directly linked to a
specific load or operational situation. There can neither a correlation be withdrawn beyond the title
‘Terrestrical photovoltaic…’ nor a defined lifetime. The tests focus solely on quality and design flaws and
help to identify early default and deficiency, also known as ‘infant mortality’ issues. The most common
standardized accelerated aging tests are the following:

• Damp Heat (DH) [Module Quality Test MQT 13]: exposure at 85% r.h., at 85 ◦ C for 1000 h. DH is often
used to identify degradation effects caused by high humidity and temperatures e.g. in coatings or polymers.
• Temperature Cycling (TC) [MQT 11]: exposure to 50 or 200 temperature cycles between −40 ◦ C and 85 ◦ C.
TC is used to provoke degradation related to internal mechanical stress e.g. in electrical connections or
tension in polymeric materials.
• UV preconditioning (UV) [MQT 10]: exposure to 15 kWh m−2 of UV light. It is used to provoke UV degrad-
ation effects, mainly in polymeric materials.
• Hot Spot Testing [MQT 9]: exposure to full illumination with partial shading of one cell to provoke hot
spots and test safety under these conditions.
• Humidity Freeze (HF) [MQT 12]: exposure to 10 cycles from −40 ◦ C to 85 ◦ C, at partially 85% r.h. HF is
used to test effects of freezing water and humidity on and in materials, especially porous material.
• Mechanical Load (ML) [MQT 16]: exposure to an external mechanical load of at least ±2400 Pa to test
effects of snow or wind loads.

The majority of these tests have not gone through significant changes during the last years or even
decades and they most likely neither will, because they do not aim at simulating outdoor operation. But still,
they are applied mostly all companies and institutions working on PV reliability issues. Further tests are
defined in the standards, but they are mostly not accounted as accelerated tests. A description can for
example be found in [122].
Often the standardized tests, mainly DH, TC and UV, are applied for even longer times and more cycles
as requested by the standards. Test results from DH4000, with elevated humidity and temperature conditions
for 4000 h, or TC600 and TC800, with 600 or 800 thermal cycles, are published very often with the aim to
state an elevated quality level. But it is important to mention again, that they are not directly linked to real
operational load conditions, and they especially do not differentiate between different load situations of
different applications or locations. Instead, the tests should help to identify weaknesses and issues of
materials and designs as statistical analysis shows [170]. Some tests, here especially the UV test should be
mentioned, only apply a relatively low dose of stress. The required 15 kWh m−2 only correspond to
approximately three months of outdoor operation in moderate climates. A correlation to mid- and
long-term exposure is not given, nor a correlation to more intense irradiations in other climate zones. As
infant mortality screening it might not even be sufficient for other than moderate climate zones.
Since the tests are standardized and applied for decades, comparability of results is in general expected by
the public, especially if tests are performed by accredited test labs. Nevertheless, there are severe differences,
which can lead to different results, as exemplarily older comparison of UV tests among different test labs
showed [137]. It is expected that the situation did not improve in the meantime. Recommendation on that
issue is not to change methods during the qualification of devices and not expect results to be easy to analyze
after a change. Many attributes of testing equipment and conditions have to be considered otherwise.
There have also several failure specific tests been developed and standardized but they are not completely
established. Descriptions of failure specific test can be found in section 7.2.

7.2. Failure specific tests

There are several specific failure modes which can occur in PV modules under different conditions [6]. For
some of them, the most relevant failures for the PV technology, specific tests have been developed to analyze
the modules susceptibility. The sensitivity towards specific loads or the susceptibility for the specific failures
are targeted. A very wide-ranging overview on failure modes of PV modules including the driving loads and
the measurement and—whenever available—the physical or chemical explanation of the failure modes can
be found in [5]. Some failure modes have been already analyzed in depth and they are seen as so important
that related testing standards have been established. One of these failure modes which was analyzed during
the last decade is PID [164, 171, 172]. It is reported that in the meanwhile developed tests and the established
standard ‘IEC TS 62804–1:2015 Photovoltaic (PV) modules—Test methods for the detection of
potential-induced degradation—Part 1: Crystalline silicon’ are helpful to identify PID prone modules, but the
relevant processes are not completely understood or modelled. That leading to incidents where qualified
modules still developed PID.

15
Prog. Energy 4 (2022) 042009 K-A Weiß et al

A degradation effect that became very important in the last couple years is the LeTID. LeTID is currently
major focus topic of research among scientific institutions and test labs [173–175]. Here, several tests and
even standards have been developed, as to mention IEC TS 63202-4 ED1—Photovoltaic cells—Part 4:
‘Measurement of light and elevated temperature induced degradation of crystalline silicon photovoltaic cells’ on
cell level. But this effect typically takes very long time to develop, and thus it is still hard to differentiate
between LeTID effects and ‘normal’ degradation effects especially during simple type approval testing.

8. Service life prediction and correlation of accelerated testing on real operation

The final goal of accelerated testing is predicting the service life of PV modules under normal outdoor
conditions. The chapter gives an overview on developments related to service life prediction (SLP) of PV
modules using data of accelerated ageing tests and the correlation of these tests with outdoor operation and
effects.

8.1. Service life prediction

PV module components degrade over time during operation, which leads to electrical performance loss
[7–10]. The service life of a PV modules depends mainly on the BoM, PV technology and the climate which
the PV module operates in, as these two factors highly influence the type and rate of degradation
mechanisms. PV reliability research mainly aims at: developing models to predict the degradation rates as
well as the service life of PV modules with acceptable accuracy, and to propose lab-based accelerated ageing
tests that can simulate near to real-world conditions and their impact on PV modules.
There is an increasing interest in understanding how PV modules and components degrade over time
and service life prediction of PV modules. Indeed, this can be shown by the trends in the number of
publications focusing on these topics over the years (see figure 1). From the figure it is clear though that there
is a slow development on lifetime prediction models on comparison to degradation studies.
To develop degradation rate or service life prediction models, a more detailed understanding of the
specific degradation mechanisms and their interactions need to be investigated. The dedicated accelerated
ageing tests help to identify degradation mechanisms in isolation by applying specific stresses that induce
these degradation mechanisms. The unsolved puzzle is on how to combine the induced degradation
mechanisms from indoor accelerated ageing based-on specific stressors to predict the service life of PV
modules in outdoor operation with combined stressors. According to the recent IEA-report on service life
estimation of PV modules [3] the use of accelerated ageing tests to predict the service life of PV modules is
still a challenge. Although, different approaches are available, based on physical and statistical methods to
extract the degradation rates from operational PV modules, not so much progress has been done to develop
service life models based on physical- and chemical-degradation of PV components that can allow ‘what if ’
simulations.
The few authors [176, 177] that have developed physics-based models use an empirical approach with a
general assumption that the degradation rate kinetics depend on a single dominating degradation
mechanism neglecting the intermediate steps involved in degradation kinetics (see also section 4.2). To
develop more reliable service life prediction models, degradation pathway network models for specific
degradation modes using network structural equation modelling (netSEM) analysis (mapping stressors,
mechanism, and responses) [178], need to be integrated in degradation rate models. The netSEM analysis is
useful to understand the complex interactions between variables and includes metrics that highlight
statistically significant relationships. This provides an in-depth insight into the mechanisms of degradation
in PV modules.
The assessment and evaluation of degradation of PV modules with different BoM, PV technologies and
in different climates is crucial input to validate and develop realistic hypothesis of service life prediction
models. Comprehensive studies such as the one by Jordan et al [179] where more than 11 000 degradation
rates from over 200 publications were assessed and categorized according to PV technology and climate
zones need to be updated to keep track of the degradation behavior of new PV technologies and BoMs in the
field. This can be useful to validate the sensitivity analysis of service life prediction models based on
technology and climate zones.

8.2. Correlation of accelerated testing on real operation

Apart from the standardized pass/fail criteria, accelerated testing is aimed at investigating the degradation
behavior of PV module components under different testing conditions. The degradation assessment from
indoor accelerated testing should mimic the degradation of PV modules in real operating conditions. For
example, the schematic diagram figure 2 shows an ideal cycle for PV module reliability studies and
correlation of accelerated testing to real operation.

16
Prog. Energy 4 (2022) 042009 K-A Weiß et al

Figure 1. Trends on the number of publications grouped into five years periods on PV/photovoltaic module degradation and
PV/photovoltaic lifetime prediction/forecast. Data generated from google scholar with the keywords only in the title to filter
papers mainly focusing on the topics.

Figure 2. Correlation cycle of accelerated testing to real operation of PV modules.

Every block of this cycle is an input for the others e.g. starting at block one, degradation modes can be
identified based on failure mode and effect analysis studies [180] and using models, experimental design to
induce the degradation modes can be achieved. In block two, the tests are verified using accelerated testing
conditions and the generated data help to calibrate the models. The calibrated models are then applied to
predict the service life of PV modules in real operating conditions using real environmental data in block
three. The analysis can also start from the analysis of performance data (block three) of already installed
modules if available, to developing validated models (block one).
But the question remains ‘to what extent is this cycle valid?’ or ‘to what extent are the accelerated indoor
tests comparable to real life exposure?’. This indeed remains a challenging question due to the complexity of
PV module designs. Firstly, PV modules come in different technologies, with different bills of materials and
operate in different locations with very different climatic conditions. All these differences lead to a variety of
degradation modes and degradation rates [170]. Secondly, frequent changes of PV module designs,
technologies, and materials [181], represent an additional lack of understanding of the degradation modes
since new materials and technologies usually come with reliability challenges. For example, improving a
specific PV module component to be resilient to a specific degradation mode might induce a completely new
degradation mode. Such challenges and the lack of long-term understanding about the performance and
reliability of these new materials hinders the correlation of accelerated testing to real operation. In addition,
in outdoor operation PV modules experience a combination of numerous climatic stressors at the same time
in comparison to controlled conditions in indoor testing. The combination of different stressors leads to
synergetic tendencies of the induced degradation modes and might induce degradation modes not identified
during accelerated aging testing. This means that the induced degradation modes during indoor experiments

17
Prog. Energy 4 (2022) 042009 K-A Weiß et al

could differ from the ones induced in real life operation hindering the correlations. The vice-versa is true,
too, due to the extreme stressors applied during accelerated aging testing, some degradation modes might be
triggered which might not really be observed in real operations.
Despite the challenges in the indoor-outdoor correlations, several research groups are working towards
these challenges, for example in [182], the authors adapted/advanced the existing standard procedures for PV
modules/components testing to optimize the testing that can simulate certain climatic conditions. They
applied time-dependent and repeating exposure to combined climatic and environmental stressors
(temperature, temperature cycles, humidity, irradiation, mechanical load, salt mist) to induce performance
losses, material degradation, and failures in test modules that resemble those effects occurring in real-life PV
installations under comparable climatic and environmental conditions. In another study [168], the authors
proposed a combined-accelerated stress test, that can simultaneously combine multiple stress factors to
mimic natural environment. This kind of simultaneous combined stress factors testing improves the
simulation of near-to field degradation modes hence allowing correlation of indoor testing to real life
operation. The bottom line here is that, to allow a good correlation of indoor accelerated testing to real life
PV module operation, accelerated testing beyond the standardized pass/fail criteria should be used.
Additionally, the authors in [25, 151, 176], have proposed and applied empirical physical models that can
help to translate indoor testing conditions to outdoor conditions. However, the models only focus on specific
stress factors applied during the accelerated testing and not on a combination of different stresses. A model
that combines numerous climatic stress-factors together with a combined stress-testing protocol could
provide a good solution for correlation of indoor testing to real-life operation and could allow more accurate
lifetime prediction based on indoor testing.

9. Conclusion and outlook

The research activity on accelerated ageing and field performance of PV modules has significantly increased
during the last decade. The increasing interest finds its causes in the growing market accompanied with the
technological development and diversification, along with the rising importance of PV for the financial
sector. The latter is demanding reliable quality and performance data. The need for such data has been
increased since reliability data has major impact on the sustainability and economy of the generated
electricity.
Field performance is the crucial parameter for all PV installations since it determines the economic
success. It is therefore essential to provide relating data in general for a successful deployment of PV modules
and in particular for quality and reliability assurance purposes. Progress has been noted in determining
operational load conditions in field use. This leads to improved availability of data sets on (environmental)
climatic conditions and microclimatic loads on the sample. That can be used to overall describe the
parameters and sample itself. This progress enables improved load and risk analysis and for example, the
creation of maps for load factors and expected degradation rates. But there is still a clear shortcoming
notable regarding the public availability of validated data. These include microclimatic data and the direct
translation from module and local data into degradation effects and degradation rates. However, there are
estimations available on the dominating processes but concluding and sound scientific explanations on
dose-response relationships and reaction kinetics are still missing. Comprehensive analysis and data sets for
upcoming technologies for example, floating PV or vehicle integrated PV, are not yet available.
Recent progress in accelerated ageing research mainly addresses the lacking correlation of accelerated tests
to real operational conditions. Different load factors have been tried to combine in tests or test sequences to
approximate the ultimate benchmark, which is the real outdoor conditions. Also, tests have been designed to
address specific failure modes and for identifying degradations effects in materials. Meanwhile, some of such
tests are commonly used and have partly even been transferred into standardized tests. In quality assurance,
typically the tests defined in IEC 61215 are still dominating, even if they cannot be related to Service Lifes or
translated into operational conditions. Often the test durations are simply multiplied to target higher quality
levels, which is critical since they anyway cannot certify for a specific degradation rate nor lifetime.
Numerical simulation tools are regularly used to support the pre-definition of testing conditions.
Especially in the ongoing development of combined and sequential testing programs with different loads
factors, simulations can help to find meaningful and (cost) efficient test designs.
Substantial progress was made in modelling of degradation rates and predicting service lifetimes. These
models include multi-step degradation calculations for PV modules. Field data are analyzed and the
(expected) dominating degradation processes predicted with statistical and physical models. For an efficient
use of these models, validated high quality data from outdoor operation is as crucial as module type specific
data from accelerated ageing tests, and derived kinetics data.

18
Prog. Energy 4 (2022) 042009 K-A Weiß et al

It is expected that there will be a continuously increasing interest in the field of accelerated ageing and
field performance research in future. This development is already driven by the financial sector and by
required sustainability assessments. If sufficient emphasis is placed on the aforementioned shortcomings and
presented limitation are carefully handled, a sound state of the art of model-based reliability, yield and
service life predictions seem within reach in the coming ten years. That said, this prediction is made under
the condition of available data, validated (material) degradation parameters and field performance data sets,
which is only to be accomplished with comprehensive and suitable field data for statistical analysis.

Data availability statement

No new data were created or analyzed in this study.

ORCID iDs

Karl-Anders Weiß  https://orcid.org/0000-0003-0425-1045

Elisabeth Klimm  https://orcid.org/0000-0002-0897-9853
Ismail Kaaya  https://orcid.org/0000-0003-4524-0975

References
[1] Osterwald C R and McMahon T J 2009 History of accelerated and qualification testing of terrestrial photovoltaic modules: a
literature review Prog. Photovolt., Res. Appl. 17 11–33
[2] Hacke P, Lokanath S, Williams P, Vasan A, Sochor P, TamizhMani G, Shinohara H and Kurtz S 2018 A status review of
photovoltaic power conversion equipment reliability, safety, and quality assurance protocols Renew. Sustain. Energy Rev.
82 1097–112
[3] Weiss K A et al 2021 Service life estimation for photovoltaic modules IEA-PVPS (available at: https://iea-pvps.org/key-topics/
service-life-estimation-for-photovoltaic-modules/) (Accessed 29 January 2022)
[4] Ascencio-Vásquez J, Kaaya I, Brecl K, Weiss K-A and Topič M 2019 Global climate data processing and mapping of degradation
mechanisms and degradation rates of PV modules Energies 12 4749
[5] Köntges M et al 2014 Review of failures of photovoltaic modules IEA International Energy Agency, Report (available at: http://
repository.supsi.ch/9645/) (Accessed 30 October 2019)
[6] Köntges M, Oreski G, Jahn U, Herz M, Hacke P and Weiß K-A 2017 Assessment of photovoltaic module failures in the field:
International Energy Agency Photovoltaic Power Systems Programme: IEA PVPS Task 13, Subtask 3: Report IEA-PVPS T13-09:2017
(Paris: International Energy Agency)
[7] Lillo-Sánchez L, López-Lara G, Vera-Medina J, Pérez-Aparicio E and Lillo-Bravo I 2021 Degradation analysis of photovoltaic
modules after operating for 22 years. A case study with comparisons Sol. Energy 222 84–94
[8] da Fonseca J E F, de Oliveira F S, Massen Prieb C W and Krenzinger A 2020 Degradation analysis of a photovoltaic generator after
operating for 15 years in southern Brazil Sol. Energy 196 196–206
[9] Del Cueto J A, Rummel S, Kroposki B, Osterwald C and Anderberg A 2008 Stability of CIS/CIGS modules at the outdoor test
facility over two decades 2008 33rd IEEE Photovoltaic Specialists Conf. (May) pp 1–6
[10] Kraus P, Massué C, Heumann S and Schlögl R 2019 Reliable long-term performance assessment of commercial photovoltaic
modules tested under field conditions over 5 years J. Renew. Sustain. Energy 11 063501
[11] Jahn U and Farnung B 2022 Performance, operation and reliability of photovoltaic systems IEA-PVPS (available at: https://iea-
pvps.org/research-tasks/performance-operation-and-reliability-of-photovoltaic-systems/) (Accessed 31 January 2022)
[12] Friesen G et al 2018 Photovoltaic module energy yield measurements existing approaches and best practice IEA PVPS Task 13,
Subtask 3 (available at: https://iea-pvps.org/key-topics/photovoltaic-module-energy-yield-measurements-existing-approaches-
and-best-practice/) (Accessed 15 February 2022)
[13] Kim J, Rabelo M, Padi S P, Yousuf H, Cho E-C and Yi J 2021 A review of the degradation of photovoltaic modules for life
expectancy Energies 14 14
[14] Omazic A, Oreski G, Halwachs M, Eder G C, Hirschl C, Neumaier L, Pinter G and Erceg M 2019 Relation between degradation of
polymeric components in crystalline silicon PV module and climatic conditions: a literature review Sol. Energy Mater. Sol. Cells
192 123–33
[15] Blanks H S 1990 Arrhenius and the temperature dependence of non-constant failure rate Qual. Reliab. Eng. Int. 6 259–65
[16] Faiman D 2008 Assessing the outdoor operating temperature of photovoltaic modules Prog. Photovolt., Res. Appl. 16 307–15
[17] Lee K, Monokroussos C and Herrmann W 2021 Comparison of different approaches to determine the nominal pv module
operating temperature (NMOT) 38th Eur. Photovolt. Sol. Energy Conf. Exhib. (November) pp 550–5
[18] Jaszczur M, Teneta J, Hassan Q, Majewska E and Hanus R 2019 An experimental and numerical investigation of photovoltaic
module temperature under varying environmental conditions (https://doi.org/10.1080/01457632.2019.1699306)
[19] Kurnik J, Jankovec M, Brecl K and Topic M 2011 Outdoor testing of PV module temperature and performance under different
mounting and operational conditions Sol. Energy Mater. Sol. Cells 95 373–6
[20] Kempe M D, Holsapple D, Whitfield K and Shiradkar N 2021 Standards development for modules in high temperature
micro-environments Prog. Photovolt., Res. Appl. 29 445–60
[21] Ascencio-Vásquez J, Brecl K and Topič M 2019 Methodology of Köppen-Geiger-Photovoltaic climate classification and
implications to worldwide mapping of PV system performance Sol. Energy 191 672–85
[22] Hülsmann P, Heck M and Köhl M 2013 Simulation of water vapor ingress into PV-modules under different climatic conditions J.
Mater. 2013 e102691
[23] Segbefia O K, Imenes A G and Sætre T O 2021 Moisture ingress in photovoltaic modules: a review Sol. Energy 224 889–906
[24] Mitterhofer S, Barretta C, Castillon L F, Oreski G, Topič M and Jankovec M 2020 A dual-transport model of moisture diffusion in
PV encapsulants for finite-element simulations IEEE J. Photovolt. 10 94–102

19
Prog. Energy 4 (2022) 042009 K-A Weiß et al

[25] Koehl M, Heck M and Wiesmeier S 2018 Categorization of weathering stresses for photovoltaic modules Energy Sci. Eng. 6 93–111
[26] Kottek M, Grieser J, Beck C, Rudolf B and Rubel F 2006 World map of the Köppen-Geiger climate classification updated
Meteorol. Z. 15 259–63
[27] Slapšak J, Mitterhofer S, Topič M and Jankovec M 2019 Wireless system for in situ monitoring of moisture ingress in PV modules
IEEE J. Photovolt. 9 1316–23
[28] Lyu Y et al 2020 Impact of environmental variables on the degradation of photovoltaic components and perspectives for the
reliability assessment methodology Sol. Energy 199 425–36
[29] Weiß K-A and Klimm E 2021 4 Loads for PV Modules (De Gruyter) pp 49–74
[30] Eder G C, Voronko Y, Oreski G, Mühleisen W, Knausz M, Omazic A, Rainer A, Hirschl C and Sonnleitner H 2019 Error analysis
of aged modules with cracked polyamide backsheets Sol. Energy Mater. Sol. Cells 203 110194
[31] Felder T C, Gambogi W J, Phillips N, MacMaster S W, Yu B-L and Trout T J 2017 Comparison of higher irradiance and black
panel temperature UV backsheet exposures to field performance Proc. SPIE 10370 1037002
[32] Khudiakova A, Brunner A J, Wolfahrt M and Pinter G 2021 Quantification approaches for fatigue crack resistance of
thermoplastic tape layered composites with multiple delaminations Materials 14 1476
[33] Jones C B, Lavrova O, Hobbs W B, Hamzavy B and Libby C 2018 IEC 61215 qualification tests vs outdoor performance using
module level in situ I-V curve tracing devices SAND2018-2229C (March) (Albuquerque, NM: Sandia National Laboratories
(SNL-NM)) (https://doi.org/10.1109/PVSC.2018.8548222)
[34] Nivelle P, Tsanakas J A, Poortmans J and Daenen M 2021 Stress and strain within photovoltaic modules using the finite element
method: a critical review Renew. Sustain. Energy Rev. 145 111022
[35] Neuhaus H, Aktaa J, Mittag M, Heinrich M, Romer P and Beinert A J 2019 Thermomechanical evaluation of new PV module
designs by FEM simulations 36th Eur. Energy Conf. and Exhibit. (October) pp 783–8
[36] Frick A, Makrides G, Schubert M, Schlecht M and Georghiou G E 2020 Degradation rate location dependency of photovoltaic
systems Energies 13 6751
[37] Halwachs M et al 2019 Statistical evaluation of PV system performance and failure data among different climate zones Renew.
Energy 139 1040–60
[38] Han H et al 2021 Aging behavior and degradation of different backsheets used in the field under various climates in China Sol.
Energy Mater. Sol. Cells 225 111023
[39] Markert J, Kotterer S, Mansour D E, Philipp D, Gebhardt P, Kenny R and Serra J M 2021 Advanced analysis of backsheet failures
from 26 power plants EPJ Photovolt. 12 7
[40] Tatapudi S, Sundarajan P, Libby C, Kuitche J and Tamizhmani G 2018 Risk priority number for PV module defects: new concepts
in solar and thermal radiation conversion and reliability 2018 Proc. SPIE 10759 1075907
[41] Karin T, Jones C and Jain A 2019 Photovoltaic degradation climate zones (https://doi.org/10.1109/PVSC40753.2019.8980831)
[42] Mani M and Pillai R 2010 Impact of dust on solar photovoltaic (PV) performance: research status, challenges and
recommendations Renew. Sustain. Energy Rev. 14 3124–31
[43] Sarver T, Al-Qaraghuli A and Kazmerski L L 2013 A comprehensive review of the impact of dust on the use of solar energy:
history, investigations, results, literature, and mitigation approaches Renew. Sustain. Energy Rev. 22 698–733
[44] Sayyah A, Horenstein M N and Mazumder M K 2014 Energy yield loss caused by dust deposition on photovoltaic panels Sol.
Energy 107 576–604
[45] Costa S C S, Diniz A S A C and Kazmerski L L 2016 Dust and soiling issues and impacts relating to solar energy systems: literature
review update for 2012–2015 Renew. Sustain. Energy Rev. 63 33–61
[46] Costa S C S, Diniz A S A C and Kazmerski L L 2018 Solar energy dust and soiling R&D progress: literature review update for 2016
Renew. Sustain. Energy Rev. 82 2504–36
[47] Schill C, Brachmann S and Koehl M 2015 Impact of soiling on IV-curves and efficiency of PV-modules Sol. Energy 112 259–62
[48] Urrejola E, Antonanzas J, Ayala P, Salgado M, Ramírez-Sagner G, Cortés C, Pino A and Escobar R 2016 Effect of soiling and
sunlight exposure on the performance ratio of photovoltaic technologies in Santiago, Chile Energy Convers. Manage. 114 338–47
[49] Cordero R R et al 2018 Effects of soiling on photovoltaic (PV) modules in the Atacama desert Sci. Rep. 8 13943
[50] Micheli L, Fernández E F, Aguilera J T and Almonacid F 2021 Economics of seasonal photovoltaic soiling and cleaning
optimization scenarios Energy 215 119018
[51] Mithhu M H, Rima T A and Khan M R 2021 Global analysis of optimal cleaning cycle and profit of soiling affected solar panels
Appl. Energy 285 116436
[52] Costa S C S, Kazmerski L L and Diniz A S A C 2021 Impact of soiling on Si and CdTe PV modules: case study in different Brazil
climate zones Energy Convers. Manage. X 10 100084
[53] Xu R, Ni K, Hu Y, Si J, Wen H and Yu D 2017 Analysis of the optimum tilt angle for a soiled PV panel Energy Convers. Manage.
148 100–9
[54] Stueve W, Farina F, Bourne B and Gostein M 2019 Field evaluation of MarsTM optical soiling sensor 36th Eur. Photovolt. Sol.
Energy Conf. Exhib. (October) pp 1471–3
[55] Klimm E and Weiss K A 2019 Performance testing of functional coatings with highly reproducible artificial soiling and outdoor
monitoring Presented at the Solar World Congress (Santiago, Chile, November)
[56] Ilse K et al 2019 Techno-economic assessment of soiling losses and mitigation strategies for solar power generation Joule
3 2303–21
[57] Figgis B 2018 Investigation of PV soiling and condensation in desert environments via outdoor microscopy Undefined (available
at: www.semanticscholar.org/paper/Investigation-of-PV-soiling-and-condensation-in-via-Figgis/7e469f3bb0bcc6f093a0e3
5014f462891bce0087) (Accessed 23 February 2022)
[58] Blanco L, Klimm E and Weiß K-A 2020 Degradation analysis of anti-soiling coatings in solar glass with AFM phase shift data the
use of data analytical methods 2020 47th IEEE Photovoltaic Specialists Conf. (PVSC) (June) pp 1584–8
[59] Figgis B, Ennaoui A, Ahzi S and Rémond Y 2017 Review of PV soiling particle mechanics in desert environments Renew. Sustain.
Energy Rev. 76 872–81
[60] Lorenz T, Klimm E and Weiss K-A 2014 Soiling and anti-soiling coatings on surfaces of solar thermal systems—featuring an
economic feasibility analysis Energy Proc. 48 749–56
[61] Heck M, Kaltenbach T, Siess C and Klimm E 2019 Artificial soiling testing and performance determination of functional coatings
36th Eur. Photovolt. Sol. Energy Conf. Exhib. (October) pp 1059–61
[62] Nayshevsky I, Xu Q and Lyons A M 2019 Hydrophobic–hydrophilic surfaces exhibiting dropwise condensation for anti-soiling
applications IEEE J. Photovolt. 9 302–7

20
Prog. Energy 4 (2022) 042009 K-A Weiß et al

[63] Nayshevsky I, Xu Q, Newkirk J M, Furhang D, Miller D C and Lyons A M 2020 Self-cleaning hybrid hydrophobic–hydrophilic
surfaces: durability and effect of artificial soilant particle type IEEE J. Photovolt. 10 577–84
[64] Klimm E 2015 Reliability and functionality tests of anti-soiling coatings for arid regions Presented at the AMI’s Polymers in
Photovoltaics (Düsseldorf )
[65] Klimm E, Weiss K-A and Spiegelhalter B 2017 Conceptual construction for the suspension of fine mineral particles to simulate
dust deposition, European federation of clean air and enviromental protection associations (EFCA) Presented at the 6th Int. Symp.
‘Ultrafine Particles’ (Brussels, Belgium)
[66] Klimm E, Ost L, Köhl M and Weiß K-A 2016 Microscopic measurement and analysis of the soiling behavior of surfaces with
standardized and real dust—a parameter study Energy Proc. C 338–45
[67] Heimsath A et al 2016 Specular reflectance of soiled glass mirrors—study on the impact of incidence angles AIP Conf. Proc.
1734 130009
[68] Smestad G P et al 2020 Modelling photovoltaic soiling losses through optical characterization Sci. Rep. 10 58
[69] Kaltenbach T, Klimm E, Philipp D, Masche M and Weiss K-A 2017 Soiling and abrasion testing of surfaces for solar energy
systems adapted to extreme climatic conditions Presented at the 8th EWS Symp.
[70] Bhaduri S, Mallick S, Shiradkar N and Kottantharayil A 2021 Characterization of reliability of anti-soiling coatings using tapping
mode-AFM phase imaging J. Renew. Sustain. Energy 13 023702
[71] Pescheux A-C, Raccurt O, Bourdon D and Le Baron E 2020 Accelerated aging tests and characterizations of innovated anti-soiling
coatings for solar receiver glasses Mater. Chem. Phys. 256 123646
[72] Klimm E, Kochersperger S and Weiss K-A 2018 Simulation of fluid flow in an artificial soiling device Presented at the COMSOL
Conf. (Lausanne) (available at: www.comsol.com/paper/simulation-of-fluid-flow-in-an-artificial-soiling-device-66411) (Accessed
17 February 2022)
[73] Coello M and Boyle L 2019 Simple model for predicting time series soiling of photovoltaic panels IEEE J. Photovolt. 9 1382–7
[74] Javed W, Guo B and Figgis B 2017 Modeling of photovoltaic soiling loss asa function of environmental variables Sol. Energy
157 397–407
[75] Pulipaka S and Kumar R 2018 Analysis of soil distortion factor for photovoltaic modules using particle size composition Sol.
Energy 161 90–99
[76] Pulipaka S, Mani F and Kumar R 2016 Modeling of soiled PV module with neural networks and regression using particle size
composition Sol. Energy 123 116–26
[77] Laarabi B, May Tzuc O, Dahlioui D, Bassam A, Flota-Bañuelos M and Barhdadi A 2019 Artificial neural network modeling and
sensitivity analysis for soiling effects on photovoltaic panels in Morocco Superlattices Microstruct. 127 139–50
[78] Hermann J, Lorenz T, Slamova K, Klimm E and Weiss K-A 2014 Modeling and testing the effect of soiling on pv modules
Presented at the The 6th World Conf. on Photovoltaic Energy Conversion WCPEC-6 (Tokyo, Japan)
[79] Hermann J et al 2017 Dust mitigation on PV modules in western Saudi Arabia Presented at the 8th EWS Symp.
[80] Fotak K et al 2012 Experimental testing of PV modules under inhomogeneous snow loads 27th Eur. Photovolt. Sol. Energy Conf.
Exhib. (October) pp 3414–7
[81] Marion B, Schaefer R, Caine H and Sanchez G 2013 Measured and modeled photovoltaic system energy losses from snow for
Colorado and Wisconsin locations Sol. Energy 97 112–21
[82] Andrews R W, Pollard A and Pearce J 2013 The effects of snowfall on solar photovoltaic performance Sol. Energy 92 84–97
[83] Andenæs E, Jelle B P, Ramlo K, Kolås T, Selj J and Foss S E 2018 The influence of snow and ice coverage on the energy generation
from photovoltaic solar cells Sol. Energy 159 318–28
[84] Granlund A, Narvesjö J and Malou Petersson A 2019 The influence of module tilt on snow shadowing of frameless bifacial
modules pp 1650–4 (available at: http://urn.kb.se/resolve?urn=urn:nbn:se:ri:diva-42554) (Accessed 1 February 2022)
[85] IEC 62938:2020 Photovoltaic (PV) modules—non-uniform snow load testing iTeh Standards Store (available at: https://standards.
iteh.ai/catalog/standards/iec/58917a89-8a74-4c67-aba9-8e20df586705/iec-62938-2020) (Accessed 1 February 2022)
[86] Flyen C, Almås A-J, Hygen H O and Sartori I 2010 Klima-og sårbarhetsanalyse for bygninger i Norge: utredning som grunnlag for
NOU om klimatilpassing p 64 (available at: https://sintef.brage.unit.no/sintef-xmlui/handle/11250/2435234) (Accessed 1
February 2022)
[87] Riley D, Burnham L, Walker B and Pearce J M 2019 Differences in snow shedding in photovoltaic systems with framed and
frameless modules 2019 IEEE 46th Photovoltaic Specialists Conf. (PVSC) (June) pp 0558–61
[88] Bogenrieder J, Camus C, Hüttner M, Offermann P, Hauch J and Brabec C J 2018 Technology-dependent analysis of the snow
melting and sliding behavior on photovoltaic modules J. Renew. Sustain. Energy 10 021005
[89] Pawluk R E, Chen Y and She Y 2019 Photovoltaic electricity generation loss due to snow—a literature review on influence factors,
estimation, and mitigation Renew. Sustain. Energy Rev. 107 171–82
[90] Frimannslund I and Thiis T 2019 A feasibility study of photovoltaic snow mitigation systems for flat roofs Czas. Tech. 7 81–96
[91] Marstein E S et al 2018 Mitigating snow on rooftop PV systems for higher energy yield and safer roofs 35th Eur. Photovolt. Sol.
Energy Conf. Exhib. (November) pp 1630–5
[92] Innos Solar energy innovations (available at: www.innos.no/) (Accessed 1 February 2022)
[93] Brearley D 2017 Designing PV systems for environmental extremes—solarpro magazine | PDF | photovoltaic system | corrosion
Scribd (available at: www.scribd.com/document/413418038/Designing-PV-Systems-for-Environmental-Extremes-SolarPro-
Magazine) (Accessed 1 February 2022)
[94] Chung P-H, Chou C-C, Yang R-Y and Chung C-Y 2019 Wind loads on a PV array Appl. Sci. 9 2466
[95] Klimm E, Guischard F and Weiss K-A 2017 Analysis of wind direction and speed measurements in arid region—a site evaluation
using data with low temporal resolution Presented at the IEA SHC Conf. (Abu Dhabi)
[96] Aßmus M, Jack S, Weiß K and Koehl M 2011 Measurement and simulation of vibrations of PV-modules induced by dynamic
mechanical loads (https://doi.org/10.1002/PIP.1087)
[97] Fogaing M B T, Hemmati A, Lange C F and Fleck B A 2019 Performance of turbulence models in simulating wind loads on
photovoltaics modules Energies 12 3290
[98] Meroney R and Neff D 2010 Wind effects on roof-mounted solar photovoltaic arrays: CFD and wind-tunnel evaluation
Undefined (available at: www.semanticscholar.org/paper/Wind-effects-on-roof-mounted-solar-photovoltaic-CFD-Meroney-Neff/
147d81257cb5197651d4c3b9a14212db870797a2) (Accessed 24 February 2022)
[99] Alrawashdeh H and Stathopoulos T 2020 Wind loads on solar panels mounted on flat roofs: effect of geometric scale J. Wind. Eng.
Ind. Aerodyn. 206 104339

21
Prog. Energy 4 (2022) 042009 K-A Weiß et al

[100] Aßmus M, Naumenko K and Altenbach H 2017 Mechanical behaviour of photovoltaic composite structures: influence of
geometric dimensions and material properties on the eigenfrequencies of mechanical vibrations Compos. Commun. 6 59–62
[101] Jahn U, Kunze I, Kajari-Schröder S and Köntges M 2011 Crack statistic of crystalline silicon photovoltaic modules 26th Eur.
Photovolt. Sol. Energy Conf. Exhib. (October) pp 3290–4
[102] Mansour D E 2021 Evaluation of the effect of backsheet properties on PV encapsulant degradation under ultraviolet-damp-heat
conditions EPFL (Lausanne) (https://doi.org/10.5075/epfl-thesis-8789)
[103] Slamova K, Schill C, Wiesmeier S, Köhl M and Glaser R 2012 Mapping atmospheric corrosion in coastal regions: methods and
results J. Photon. Energy 2 022003
[104] Koehl M, Heck M, Philipp D, Weiss K-A, Ferrara C and Herrmann W 2008 Indoor and outdoor weathering of PV-modules Proc.
SPIE 7048 37–44
[105] Ahmad Z 2006 Atmospheric corrosion Principles of Corrosion Engineering and Corrosion Control ed Z Ahmad (Oxford:
Butterworth-Heinemann) ch 10, pp 550–75
[106] Munson C A, Westre S and Westre A 2021 Electrochemical testing to predict corrosion performance on various aluminum alloys
Presented at the CORROSION 2021 (April) (available at: https://onepetro.org/NACECORR/proceedings/CORR21/5-CORR21/
D051S023R002/463755) (Accessed 1 February 2022)
[107] Petzold K 2011 Salmiakgeist greift Module an pv magazine Deutschland (4 May) (available at: www.pv-magazine.de/2011/05/05/
salmiakgeist-greift-module-an/) (Accessed 15 June 2022)
[108] Guo M 2021 Corrosion behavior of 316 stainless steel exposed to a simulated salt lake atmospheric environment under UV
illumination Int. J. Electrochem. Sci. 16 210457
[109] Gretkowska L 2018 A study on the performance of solar photovoltaic modules exposed to salt water (available at: www.um.edu.
mt/library/oar/handle/123456789/37502) (Accessed 1 February 2022)
[110] Lelièvre J-F et al 2022 Desert label development for improved reliability and durability of photovoltaic modules in harsh desert
conditions Sol. Energy Mater. Sol. Cells 236 111508
[111] Hoffmann S and Koehl M 2014 Effect of humidity and temperature on the potential-induced degradation Prog. Photovolt.
22 173–9
[112] Daoren G, Yingnan C, Gang S, Wenjing W and Zhenshuang J 2018 Research on potential induced degradation (PID) of PV
modules in different typical climate regions IOP Conf. Ser.: Mater. Sci. Eng. 330 012041
[113] ISO 9225:2012 ISO (available at: www.iso.org/cms/render/live/en/sites/isoorg/contents/data/standard/05/35/53501.html)
(Accessed 1 February 2022)
[114] Santucci R J, Davis R S and Sanders C E 2021 Atmospheric corrosion severity and the precision of salt deposition measurements
made by the wet candle method Corros. Eng. Sci. Technol. 57 147–58
[115] Iqbal N et al 2022 Characterization of front contact degradation in monocrystalline and multicrystalline silicon photovoltaic
modules following damp heat exposure Sol. Energy Mater. Sol. Cells 235 111468
[116] Asadpour R, Patel M T, Clark S, Bosco N, Silverman T J and Alam M A 2022 Worldwide physics-based lifetime prediction of c-Si
modules due to solder-bond failure IEEE J. Photovolt. 12 1–7
[117] Elgin M 2021 Analysis of snail tracks impact on photovoltaic module’s parameters Instytut Techniki Cieplnej (available at: https://
repo.pw.edu.pl/info/master/WUT833ad5acdd35472295fcfc9990590692/) (Accessed 1 February 2022)
[118] Hara K 2022 Raman spectroscopic analysis of encapsulants in aged photovoltaic modules J. Photochem. Photobiol. Chem.
425 113721
[119] Peike C 2015 Degradation analysis of the encapsulation polymer in photovoltaic modules by raman spectroscopy Fraunhofer
Bookshop (available at: www.bookshop.fraunhofer.de/buch/degradation-analysis-of-the-encapsulation-polymer-in-photovoltaic-
modules-by-raman-spectroscopy/242784) (Accessed 1 February 2022)
[120] Kumar R E et al 2022 Glass vs. backsheet: deconvoluting the role of moisture in power loss in silicon photovoltaics with correlated
imaging during accelerated testing IEEE J. Photovolt. 12 285–92
[121] CEEES 2009 Reliability—for a mature product from the beginning of useful life—the different type of tests and their impact on
product reliability, home (available at: www.ceees.org/downloads/publications/CEEES_Publication_No_9.pdf) (Accessed 31
January 2022)
[122] Weiss K A 2021 Photovoltaic Modules: Reliability and Sustainability (Berlin: De Gruyter) (https://doi.org/10.1515/9783110685558)
[123] Lamaamar I, Tilioua A, Ben Zaid Z, Babaoui A, Ettakni M and Hamdi Alaoui M A 2021 Evaluation of different models for
validating of photovoltaic cell temperature under semi-arid conditions Heliyon 7 e08534
[124] Koehl M, Heck M and Wiesmeier S 2012 Modelling of conditions for accelerated lifetime testing of humidity impact on
PV-modules based on monitoring of climatic data Sol. Energy Mater. Sol. Cells 99 282–91
[125] Kaaya I, Ascencio-Vásquez J, Weiss K-A and Topič M 2021 Assessment of uncertainties and variations in PV modules degradation
rates and lifetime predictions using physical models Sol. Energy 218 354–67
[126] Kurtz S 2018 Advance solar technology by marrying quality assurance with design engineering Nat. Energy 3 1016–8
[127] Gottschalg R, Ebert M, Dittmann S and Jäckel B 2020 Systematic causes of problems in operation of PV systems 2020 47th IEEE
Photovoltaic Specialists Conf. (PVSC) (June) pp 1675–9
[128] PVQAT 2022 International PV quality assurance task force (available at: www.pvqat.org/index.html) (Accessed 31 January 2022)
[129] Meeker W and Escobar L 1993 A review of recent research and current issues in accelerated testing Int. Stat. Rev. 61 147
[130] Pern F J 1997 Ethylene-vinyl acetate (EVA) encapsulants for photovoltaic modules: degradation and discoloration mechanisms
and formulation modifications for improved photostability Angew. Makromol. Chem. 252 195–216
[131] Subramaniyan A B, Pan R, Kuitche J and TamizhMani G2018 Prediction of acceleration factor for accelerated testing of
photovoltaic modules installed around the world 2018 Annual Reliability and Maintainability Symp. (RAMS) (January)
pp 1–6
[132] Bheemreddy V, Bill Liu J J, Wills A and Murcia C P 2018 Life prediction model development for flexible photovoltaic modules
using accelerated damp heat testing 2018 IEEE 7th World Conf. on Photovoltaic Energy Conversion (WCPEC) (A Joint Conf. of 45th
IEEE PVSC, 28th PVSEC 34th EU PVSEC) (June) pp 1249–51
[133] Mansour D E, Barretta C, Pitta Bauermann L, Oreski G, Schueler A, Philipp D and Gebhardt P 2020 Effect of backsheet
properties on pv encapsulant degradation during combined accelerated aging tests Sustainability 12 5208
[134] Lin C-C, Lyu Y, Jacobs D S, Kim J H, Wan K-T, Hunston D L and Gu X 2019 A novel test method for quantifying cracking
propensity of photovoltaic backsheets after ultraviolet exposure Prog. Photovolt., Res. Appl. 27 44–54
[135] Miller D C et al 2015 Degradation in PV encapsulation transmittance: an interlaboratory study towards a climate-specific test
2015 IEEE 42nd Photovoltaic Specialist Conf. (PVSC) (June) pp 1–6

22
Prog. Energy 4 (2022) 042009 K-A Weiß et al

[136] Tavakoli M, Jahantigh F and Zarookian H 2021 Adjustable high-power-LED solar simulator with extended spectrum in UV
region Sol. Energy 220 1130–6
[137] Koehl M and Philipp D 2015 Inter-laboratory comparison of UV-light sources for testing of PV-modules Prog. Photovolt., Res.
Appl. 23 1815–9
[138] Lv R, Tang J, Jaubert J-N and Xing G 2019 Highly accelerated thermal cycling test for new type of crystalline silicon photovoltaic
modules 2019 IEEE 46th Photovoltaic Specialists Conf. (PVSC) (June) pp 1991–4
[139] Gottschalg R, Betts T R, Zhu J and Owen-Bellini M 2015 Environmental stress potentials different climatic regions 31st European
Photovoltaic Solar Energy Conf. and Exhibition (November) pp 1854–7
[140] Nyarko F K A, Takyi G, Amalu E H and Adaramola M S 2019 Generating temperature cycle profile from in-situ climatic condition
for accurate prediction of thermo-mechanical degradation of c-Si photovoltaic module Eng. Sci. Technol. Int. J. 22 502–14
[141] Gawlinska-Necek K et al 2021 Silicon solar cells and modules with front contact paste containing copper-based component Prog.
Photovolt., Res. Appl. 29 1008–19
[142] Curran A J 2021 Reliability of commercially relevant photovoltaic cell and packaging combinations in accelerated and outdoor
environments (Case Western Reserve University) (available at: https://etd.ohiolink.edu/apexprod/rws_olink/r/1501/10?clear=
10&p10_accession_num=case1626783783924982) (Accessed 1 February 2022)
[143] Xu J and Bu J 2019 Preparation and properties of solar composite materials for marine environment Appl. Ecol. Environ. Res.
17 14217–29
[144] ISO 9227 Corrosion tests in artificial atmospheres (available at: www.micomlab.com/micom-testing/iso-9227/) (Accessed 29
January 2022)
[145] IEC 60068-2-52:2017 European Standards (available at: www.en-standard.eu/iec-60068-2-52-2017-environmental-testing-part-
2-52-tests-test-kb-salt-mist-cyclic-sodium-chloride-solution/) (Accessed 1 February 2022)
[146] IEC 61701:2020 Photovoltaic (PV) modules—salt mist corrosion testing iTeh Standards Store (available at: https://standards.iteh.
ai/catalog/standards/iec/e4e0c49e-4272-41dd-bc47-5c600c025a9a/iec-61701-2020) (Accessed 1 February 2022)
[147] Jeffries A M, Nietzold T, Schelhas L T and Bertoni M I 2021 Corrosion of novel reactive silver ink and commercial silver-based
metallizations in diluted acetic acid Sol. Energy Mater. Sol. Cells 223 110900
[148] Sutter F, Fernández-García A, Wette J, Reche-Navarro T J and Martínez-Arcos L 2019 Acceptance criteria for accelerated aging
testing of silvered-glass mirrors for concentrated solar power technologies Sol. Energy Mater. Sol. Cells 193 361–71
[149] Wang X F, Bao Y W, Liu X G and Qiu Y 2018 Effects of salt spray corrosion on durability of laminated photovoltaic glass Key Eng.
Mater. 768 246–50
[150] Kuo C-W et al 2019 Impact of the combination-accelerated stress testing for 4 solar cell mini modules 2019 IEEE 46th
Photovoltaic Specialists Conf. (PVSC) (June) pp 1979–81
[151] Sinha A, Gopalakrishna H, Bala Subramaniyan A, Jain D, Oh J, Jordan D and TamizhMani G 2020 Prediction of climate-specific
degradation rate for photovoltaic encapsulant discoloration IEEE J. Photovolt. 10 1093–101
[152] Schneller E J, Seigneur H, Lincoln J and Gabor A M 2019 The impact of cold temperature exposure in mechanical durability
testing of PV modules 2019 IEEE 46th Photovoltaic Specialists Conf. (PVSC) (June) pp 1521–4
[153] Hsu S-T and Wu T-C 2017 Simulated wind action on photovoltaic module by non-uniform dynamic mechanical load and mean
extended wind load Energy Proc. 130 94–101
[154] Rabelo M, Zahid M A, Khokhar M Q, Sim K, Oh H, Cho E-C and Yi J 2022 Mechanical fatigue life analysis of solar panels under
cyclic load conditions for design improvement J. Braz. Soc. Mech. Sci. Eng. 44 87
[155] Gabor A M, Janoch R, Anselmo A, Lincoln J L, Seigneur H and Honeker C 2016 Mechanical load testing of solar panels—beyond
certification testing 2016 IEEE 43rd Photovoltaic Specialists Conf. (PVSC) (June) pp 3574–9
[156] Sillerud C, Lee K and Crimmins J 2019 Sequential application of static and dynamic mechanical stresses for electrical isolation of
cell cracks 2019 IEEE 46th Photovoltaic Specialists Conf. (PVSC) (June) pp 2561–5
[157] Dodson H S B 2006 Accelerated Testing: A Practitioner’s Guide to Accelerated and Reliability Testing (Warrendale, PA: SAE
International)
[158] Anagha E R, Kulkarni S V and Shiradkar N 2021 Moisture ingress modeling in c-Si PV modules using finite element simulations
based on dual transport diffusion 2021 IEEE 48th Photovoltaic Specialists Conf. (PVSC) (June) pp 2470–4
[159] Kobayashi Y, Morita H, Mori K and Masuda A 2018 Effect of barrier property of backsheet on degradation of crystalline silicon
photovoltaic modules under combined acceleration test composed of UV irradiation and subsequent damp-heat stress Jpn. J.
Appl. Phys. 57 127101
[160] Tanahashi T et al 2019 Reproducing the ‘framing’ by a sequential stress test 2019 IEEE 46th Photovoltaic Specialists Conf. (PVSC)
(June) pp 1403–9
[161] Thellen C, Santoleri D, Rothacker A and Kim D 2020 Cost efficient and highly weather-resistant solar panel backsheet produced
through continuous co-extrusion processing Final-Report-DOE-Tomark-Worthen-08143-1 (April) (Chadds Ford, PA:
Tomark-Worthen LLC) (https://doi.org/10.2172/1615916)
[162] Ino Y, Asao S, Shirasawa K and Takato H 2018 Effect of soldering on the module degradation along bus bar in DH test and PCT
for crystalline Si PV modules 2018 IEEE 7th World Conf. on Photovoltaic Energy Conversion (WCPEC) (A Joint Conf. of 45th IEEE
PVSC, 28th PVSEC 34th EU PVSEC) (June) pp 3552–7
[163] Köhl M, Bogdanski N, Weiß K-A, Dürr I, Hoffmann S and Peike C 2014 PV module degradation in the field and in the lab—how
does it fit together? 29th Eur. Photovolt. Sol. Energy Conf. Exhib. (November) pp 3110–4
[164] Hacke P 2019 Combined and sequential accelerated stress testing for derisking photovoltaic modules Advanced Micro- and
Nanomaterials for Photovoltaics ed D Ginley et al (Amsterdam: Elsevier) ch 11, pp 279–313
[165] Gambogi W J et al 2018 Sequential stress testing to predict photovoltaic module durability 2018 IEEE 7th World Conf. Photovoltaic
Energy Convers. WCPEC Jt. Conf. 45th IEEE PVSC 28th PVSEC 34th EU PVSEC (https://doi.org/10.1109/PVSC.2018.8547260)
[166] Gambogi W et al 2019 Development of accelerated test sequences to assess long term durability of PV modules 2019 IEEE 46th
Photovoltaic Specialists Conf. (PVSC) (June) pp 2406–10
[167] Gebhardt P et al 2021 Optimizing combined accelerated stress testing for solar backsheet qualification 2021 IEEE 48th
Photovoltaic Specialists Conf. (PVSC) (June) pp 0322–5
[168] Owen-Bellini M, Hacke P, Miller D C, Kempe M D, Spataru S, Tanahashi T, Mitterhofer S, Jankovec M and Topič M 2021
Advancing reliability assessments of photovoltaic modules and materials using combined-accelerated stress testing Prog.
Photovolt., Res. Appl. 29 64–82
[169] Tatapudi S, Kuitche J and Tamizhmani G 2018 A novel climate-specific field accelerated testing of PV modules: new concepts in
solar and thermal radiation conversion and reliability 2018 Proc. SPIE 10759 1075908

23
Prog. Energy 4 (2022) 042009 K-A Weiß et al

[170] Gebhardt P, Mülhöfer G, Roth A and Philipp D 2021 Statistical analysis of 12 years of standardized accelerated aging in
photovoltaic-module certification tests Prog. Photovolt., Res. Appl. 29 1252–61
[171] Virtuani A, Annigoni E and Ballif C 2019 One-type-fits-all-systems: strategies for preventing potential-induced degradation in
crystalline silicon solar photovoltaic modules Prog. Photovolt., Res. Appl. 27 13–21
[172] Islam M A, Hasanuzzaman M and Rahim N A 2018 A comparative investigation on in-situ and laboratory standard test of the
potential induced degradation of crystalline silicon photovoltaic modules Renew. Energy 127 102–13
[173] Dupuis J, Plessis G, El Hajje G, Lajoie-Mazenc E, Sandré E, Radouane K and Dupeyrat P 2021 Light- and elevated
temperature-induced degradation impact on bifacial modules using accelerated aging tests, electroluminescence, and
photovoltaic plant modeling Prog. Photovolt., Res. Appl. 29 694–704
[174] Repins I L, Kersten F, Hallam B, VanSant K and Koentopp M B 2020 Stabilization of light-induced effects in Si modules for IEC
61215 design qualification Sol. Energy 208 894–904
[175] Philipp D, Geipel T, Gebhardt P, Schmid A, Naeem T and Fokuhl E 2019 LeTID—a comparison of test methods on module level
36th Eur. Photovolt. Sol. Energy Conf. Exhib. pp 816–21
[176] Kaaya I, Koehl M, Mehilli A, de Cardona Mariano S and Weiss K A 2019 Modeling outdoor service lifetime prediction of PV
modules: effects of combined climatic stressors on PV module power degradation IEEE J. Photovolt. 9 1105–12
[177] Bala Subramaniyan A, Pan R, Kuitche J and TamizhMani G 2018 Quantification of environmental effects on PV module
degradation: a physics-based data-driven modeling method IEEE J. Photovolt. 8 1289–96
[178] Huang W-H et al 2018 netSEM: network structural equation modeling (available at: https://CRAN.R-project.org/package=
netSEM) (Accessed 21 September 2020)
[179] Jordan D C, Kurtz S R, VanSant K and Newmiller J 2016 Compendium of photovoltaic degradation rates Prog. Photovolt., Res.
Appl. 24 978–89
[180] Colli A 2015 Failure mode and effect analysis for photovoltaic systems Renew. Sustain. Energy Rev. 50 804–9
[181] Oreski G et al 2021 Designing new materials for photovoltaics: opportunities for lowering cost and increasing performance
through advanced material innovations SAND2021-4837R, 1779380, 695676 (April) (https://doi.org/10.2172/1779380)
[182] Eder G C, Voronko Y, Dimitriadis S, Knöbl K, Újvári G, Berger K A, Halwachs M, Neumaier L and Hirschl C 2019 Climate
specific accelerated ageing tests and evaluation of ageing induced electrical, physical, and chemical changes Prog. Photovolt., Res.
Appl. 27 934–49

24