plant performance Technical Briefing

Benchmarking inverter
performance and reliability
with a new PVEL Scorecard
Inverters performance | Inverters are the leading source of corrective maintenance activity in PV
power plants, yet independent testing to inform procurement decisions remains the exception rather
than the norm. Michael Mills-Price and Jenya Meydbray of PVEL describe how a new inverter testing
regime is seeking to set quality benchmarks for this increasingly critical part of the PV system
Credit: PVEL

I
nverters are the number one driver electronics, these components degrade PVEL’s inverter the industry is beginning to recognise
of PV project profitability. Every time over time. But what is a reasonable testing aims to the importance of inverter selection
provide investors
a solar inverter underperforms or lifetime expectation – and how do PV to a project’s long-term economic
with better intelli-
shuts down unexpectedly, the entire PV inverter buyers generate reliable predic- gence on inverter performance. Low-performing invert-
system produces less energy – or none tions? Only about 25% of the world’s performance and ers that generate unexpected, ongoing
at all. Maintenance costs are compound- total installed PV capacity has operated reliability maintenance expenses can ultimately
ed by the financial consequences of for more than five years, so the industry become costly for asset owners. Due to
energy shortfall. lacks long-term real-world data. While the underlying complexity of inverter
Modern inverters contain hundreds most inverters today are warranted for design and construction as well as their
of complex, software-driven compo- 10 years, results from the field show that broad functionality, inverters are also
nents that monitor and control the most many products do not last that long. more vulnerable to reliability issues than
vital operations of a PV system. Like all As systems age and inverters degrade, any other PV system component. Devel-

94 | August 2019 | www.pv-tech.org

 Technical Briefing plant performance

oping accurate inverter lifetime and cost

of ownership predictions should not be
an afterthought.
PV Evolution Labs (PVEL) is one of
very few independent labs that performs
extended reliability and performance
testing on PV inverters. We found that
one-third of the products we tested
through our PV Inverter Product Quali-
fication Program (PQP) failed key safety
and performance tests – even though
all of the tested products were certified
by IEC and/or UL. In response to these
Figure 1. The figure shows an inverter that failed to operate after only 30% of the powered thermal cycling
findings and to growing demand for test sequence was complete. It was unable to return to operation
inverter data, PVEL published its first PV
Inverter Scorecard in May 2019.
This first Scorecard was developed
with two main goals: first, to educate
PV asset owners, project developers
and investors about the complexity and
inherent risks associated with invert-
ers and, second, to introduce the PV
buyer and asset owner community to
inverter reliability and performance
testing that provides critical insights
for inverter diligence. It is also the first
inverter benchmarking report based on
independent test data that is available
to the public. This article highlights key
insights from PVEL’s Scorecard to explain
why and how PV equipment buyers can
use objective reliability and performance
data to mitigate the financial conse-
quences of technology risks inherent to Figure 2. The figure shows an inverter that de-rated to avoid significant temperature increases of internal
inverters. components during the high temperature test – even though the ambient temperatures sustained were
within its operational window
Inverter procurement today
Although inverters are the leading source weather the financial impact of higher materials packages through every step
of corrective maintenance tickets in PV than expected operations and mainte- of the process, including final packaging
power plants and the top cause of energy nance expenses. with tamper-proof tape. This ensures
outages [1], very few PV project develop- that hand-picked samples are not sent
ers, financial institutions and asset owners About the PV Inverter Scorecard for testing. The inverters that ranked as
to date require independent testing The 2019 Scorecard is based on Top Performers for each test are listed
that assesses inverter reliability and independent testing of 35 inverter by name and model in the Scorecard,
performance. Historically, due diligence models produced by 12 different which is available as a free download at
expectations and testing requirements for manufacturers. Tested products include www.pvel.com/inverter-scorecard, and
inverters have been much less rigorous string inverters (both three-phase and summarised in the box towards the end
than those for PV modules. single-phase), microinverters and power of this article.
The challenge is that many project optimisers. Each inverter evaluated
stakeholders lack the institutional for the Scorecard underwent testing Test results: thermal performance
knowledge and data required for through PVEL’s Product Qualification Temperature directly impacts an invert-
in-depth technical due diligence of Program (PQP) for inverters. Results from er’s electrical performance and long-
inverters. Instead, buyers and inves- 14 tests are presented in the Scorecard. term reliability; therefore, all inverter
tors rely on certifications, brand names, Highlighted results from seven of these manufacturers provide product-specific
datasheets and warranties to evaluate tests are discussed below. maximum and minimum ambient
inverter bankability. PVEL’s PV Inverter All inverters were tested in the same temperatures for operation. The individu-
Scorecard proves that these data sources way, leveraging consistently calibrated al electrical components within inverters
are not sufficient for strategic inverter equipment and in consistent laboratory also have maximum and minimum
procurement where long-term financial environments. Inverters submitted for temperatures at which they operate.
returns are at stake. This is especially true testing through PVEL’s PQP are witnessed Temperature conditions can vary widely
for cost-sensitive projects that cannot in production – from the opening of raw in the field and inverter components

www.pv-tech.org | August 2019 | 95

 plant performance Technical Briefing

couple sensors. These temperatures

are then compared to the component
datasheets to verify design parameters
and determine whether the inverters
de-rated while operating within their
specified temperature windows. Not all
inverters tested were able to operate
without de-rating – and some were not
Figure 3. The share of inverters that failed to operate following each passive chamber test. Many of the
able to continue operating during the
inverters that were still operational following testing operated in a reduced capacity
tests (See Figures 1 and 2).
As pricing pressure on inverter
manufacturers intensifies, some
producers may utilise smaller, less
expensive and less robust components.
For example, using a silicon chip with
narrower temperature or voltage limits
may reduce costs in the short term, but
could ultimately cause problems for
system owners as the cheaper compo-
nents prove less reliable when exposed
to various thermal conditions.

Test results: passive chamber

testing
Inverters contain circuit boards, silicon
chips and integrated products that can
age and fail when exposed to sunlight,
rain, temperature swings, humidity,
snow and other common environmental
Figure 4. Delamination and internal corrosion in an inverter following passive chamber testing conditions. Unpowered environmental
chamber testing evaluates the impact of
may be susceptible to thermal drift, a Thermal performance tests are used these environmental stresses on invert-
phenomenon which results in individual to measure and document the thermal ers and their components. PVEL’s goal
components performing differently de-rating of inverters. The tests are with these tests is to assess the product
than expected. Inverters are designed among the best methods of determin- construction, Bill of Materials and
with built-in safety mechanisms that ing whether an inverter’s performance product design of an inverter.
prevent these internal components accurately reflects the temperature Passive chamber tests include thermal
from reaching their maximum allowable specifications on its datasheet. To cycling, humidity freeze and damp heat.
temperatures – but these safety mecha- perform these tests, the inverter is The test procedures align with and
nisms should only be triggered when placed in an environmental chamber expand upon the IEC 61215 test stand-
absolutely necessary because of their and connected to a solar array simula- ard, one of the most common certifica-
impact on energy yield. tor. Next, the following conditions are tion requirements for determining the
Manufacturers design inverters with applied: safe operation of PV modules. Important-
these safety mechanisms because degra- • Powered thermal cycling. Thermal ly, PVEL’s chamber tests reproduce failure
dation rates accelerate when components cycling is performed across the full modes and reliability issues commonly
exceed their temperature limits during operational temperature range while observed in the field, including coating
operation. Operating beyond allow- the inverter is powered from minimum delamination, corrosion, water conden-
able limits reduces the lifetime of the ambient temperature to maximum sation in wiring compartments, discol-
component and ultimately the inverter ambient temperature. oration and melting of external displays
itself. To avoid this, inverters de-rate, or • High temperature operation. The and controls, and electro-mechanical
reduce power output, when the tempera- maximum operational temperature fatigue of solder joints and electrical
ture limits of an internal component is sustained while the inverter is connections.
or subsystem are exceeded. While this powered. Passive chamber testing results
de-rating process is important for inverter • Low temperature operation. The (Figures 3 and 4) indicate:
reliability, it should only occur when the minimum operational temperature • 25% of inverters failed to operate after
inverter is exposed to conditions outside is sustained while the inverter is damp heat;
of its operational window. When inverters powered. • 21% of inverters failed to operate after
de-rate, they convert less energy than humidity freeze;
expected from the PV array. This results in As these tests are conducted, PVEL • A significant population of inverters
reduced energy yield and financial losses measures the temperatures of multiple that operated following these tests
for the asset owner. individual components using thermo- were only able to operate in degraded,

96 | August 2019 | www.pv-tech.org

 Technical Briefing plant performance

shut the system down. In extreme cases

such as with fires, electrical arcs pose
considerable safety risks. They can also
result in significant, irreversible damage
to the entire PV system. Inverters detect
arcs by sensing their characteristic
signature, or fingerprint, in the frequency
domain. However, this fingerprint may be
masked because it is dependent on the
location of the arc in the PV array. PVEL’s
ground and arc fault testing exposes the
inverters to arcs at several locations within
the array and documents the response of
the inverter.
PVEL’s fault tests are conducted at
our outdoor test site. The inverters are
Figure 5. The figure shows a properly detected and interrupted arc fault. It highlights the ability of some subjected to multiple ground and arc
inverters to effectively identify and respond to arc faults within the required time range, before the device fault conditions on a grid-connected PV
or the array incurs any damage system. The inverter is monitored to track
proper system shutdown. One third of the
less efficient states; Test results: ground and arc fault inverters evaluated for PVEL’s Scorecard
• The most common failure modes Safe operation is fundamental to the failed to detect at least one fault during
were faulty moisture protection of a economic success of a PV system, but testing – even though they passed
component, delamination and internal electrical arcs can occur if electrical certification testing (see Figure 5). This
corrosion. conductors are exposed to the environ- finding is alarming given the importance
ment. Exposure can occur as systems of fault detection to the safe operation of
Successful performance in passive age; for example, when insulation around PV systems.
chamber tests indicates that products system wiring degrades, connectors age
have robust construction and design or come loose on module backsheets fail Independent test results in context
that can withstand common field and start to crack. Properly detecting arcs Every solar project’s financial model
conditions. While all inverters that are is part of an inverter’s core operation. depends on energy yield forecasts that
deployed to PV sites should pass these In PV systems, electrical arcs can predict safe, reliable power generation for
tests, PVEL’s data shows that this is not manifest as fires– but only when the decades. They also depend on reason-
always the case. inverter fails to detect them and rapidly ably accurate estimates of operations and
maintenance costs. PVEL’s testing shows
that inverters are not always equipped to
meet these expectations. Some do not
even meet the minimum requirements
specified by datasheets and certifications.
Many buyers and investors rely on
warranties to protect them financially
when inverters fail. This strategy can
quickly backfire when an inverter
manufacturer exits the market. Addition-
ally, replacing inverter products may be
nontrivial. Imagine searching for a new
600V central inverter today. If a suitable
replacement is not available, the entire PV
system may require rewiring to prevent
electrical mismatch.
A 2017 study assessed the true cost of
inverter ownership using data from 400
failure reports (see Figure 6) and found
that two of the four inverter manufac-
turers assessed generated far higher
operations and maintenance costs than
predicted [2]. In some cases, annual
maintenance expenses were underesti-
mated by more than 500%.
Figure 6. Comparison of the cumulative cost of inverter ownership to cost estimates provided by manufac- The study goes on to note: “In view of
turers. Actual costs are only provided for years where data is available [2] the high costs associated with inverter

www.pv-tech.org | August 2019 | 97

 plant performance Technical Briefing

PVEL’s Top Performer inverters

The scorecard’s 14 tests are broken down into five categories: passive Top Performers in the conversion efficiency tests were Huawei’s SUN2000-
chamber testing; thermal performance characterisation; performance 30KTL-US and SUN2000-375W-USP0, and Schneider’s Context CL-60A.
testing – efficiency; performance testing – operational window; and field In the energy harvest tests, Top Performers were Huawei’s SUN2000-
testing. Where appropriate, ‘Top Performers’ are identified for each test 28KTL and SUN2000-30KTL-US, also accompanied by Schneider’s Context
– specific inverter makes and models that have performed particularly CL-60A.
well. The makers and inverter models awarded Top Performer status in the
inaugural scorecard are detailed below: Performance testing: operational
In the performance operational tests, which include operational envelope
Passive chamber and transient response, PVEL noted that “inverters that have wide DC input
In the Passive Chamber tests, only two companies were awarded Top ranges can support a more diverse set of possible stringing configurations,
Performer status: Delta’s single phase string inverter M8-TL-US and SMA allowing flexibility for the designer or installer of the system”.
Solar’s SB7.7-ISP-US-40, another single phase string inverter. PVEL only scored Delta’s M80U inverter in the AC operational envelope
test and the same inverter in the DC operational envelope category with
Thermal performance characterisation Schneider’s Conext CL 25000NA inverter.
Three different companies’ products were Top Performers in the powered In the transient response test, Top Performer status was given to Delta’s
thermal cycling category – Delta’s single-phase string inverter M8-TL-US, M8-TL-US, Fronius’ Symo 24.0-3 and Huawei’s SUN2000-11.4KTL-US inverter.
Schneider’s Conext CL-60A string inverter and SMA Solar’s SB7.7-ISP-US-40.
PVEL highlighted an example of an inverter that failed 30% through the test Field testing
sequence with an inability to return to operation. In the field testing category, PVEL said that the PQP determined whether an
In the high temperature operation category, three different companies’ inverter would operate safely and continuously in real-world conditions. The
products were Top Performers: Delta M8-TL-US, Fronius’ Symo 24.0-3, a tests include a ground and arc fault tests and a 30-day runtime in operation.
transformerless three-phase string inverter, and Huawei’s SUN2000-11.4KTL- The three Top Performers in the ground and arc fault tests were Delta’s M8
US string inverter. TL-US and Solivia 3.8 TL inverters, as well as Fronius’ Symo 24.0-3 inverter.
There were two Top Performers in the low temperature operation test: the In the 30-day runtime tests, only two companies’ products were given
Delta Solivia 3.8 TL and the Huawei SUN2000-11.4KTL-US. Top Performer accreditation: Huawei’s Sun2000-11.4KTL-US and SMA Solar’s
SB7.7-1SP-US-40.
Performance testing: efficiency
In the performance efficiency test category, which analyses MPPT efficiency, Summary
conversion efficiency and energy harvest, PVEL noted the tests set out to In summary, the three most-cited Top Performer companies were Delta with
demonstrate whether or not an inverter can actually perform as expected 10, Huawei with nine and Schneider with five.
based on product datasheets when deployed in the field. On a product basis, the three most cited Top Performers were Delta’s M8
Three different companies’ products were Top Performers in the TL-US inverter with four accreditations, Huawei’s Sun2000-11.4KTL-US also
MPPT efficiency test: Delta’s M8-TL-US, Huawei’s SUN2000-30KTL-US and with four and Fronius’ Symo 24.0-3 with thre accreditations.
Schneider’s Conext CL-60A string inverter. PVEL noted that Top Performers
in this test sequence had a 98-99% response rate for all three test By Mark Osborne
conditions.

failures, understanding the root cause of gies have not evolved significantly References
component failures, methods to access over the past decade. This is a risky [1] Hacke, P. Lokanath, S. Williams, P., Vasan, A., Sochor, P., TamizhMani,
or ensure reliability and forecast lifetime approach for PV asset owners and G., Shinohara, H., Kurtz, S. 2017 “A status review of photovoltaic
power conversion equipment reliability, safety, and quality
of the power conversion electronics and investors because independent testing
assurance protocols”, Renewable and Sustainable Energy
their components through testing and proves that not all inverters live up to Reviews, Vol. 82, No. 1, pp. 1098
quality standards becomes vital.” [2] The expectations. Instead there is a range [2] Hacke, P. Lokanath, S. Williams, P., Vasan, A., Sochor, P., TamizhMani,
G., Shinohara, H., Kurtz, S. 2017 “A status review of photovoltaic
testing conducted through PVEL’s PQP of performance, functionality, efficien- power conversion equipment reliability, safety, and quality
and the results presented in the Scorecard cy and reliability across commercially assurance protocols”, Renewable and Sustainable Energy
Reviews, Vol. 82, No. 1, pp. 1098
provide the data developers, banks and available products.
asset owners need to better predict the When PVEL began testing PV modules
long-term reliability of inverters. in 2010, we observed tremendous Authors
variability in performance and reliability
Jenya Meydbray is CEO of PV Evolution Labs (PVEL),
Next steps across manufacturers and tests. The
which he cofounded in 2010 as the first independ-
As the solar industry matures and asset product landscape was very similar to ent lab dedicated to supporting solar project devel-
owners focus more on total system the market for inverters today. As the opers, financial institutions and asset owners. Jenya
lifetime performance – and not just initial buyer community began to recognise developed among the first extended reliability and
costs – inverter reliability is becoming the variability of PV modules and the performance test protocols for the downstream PV industry
as well as innovative methods of evaluating PV performance
increasingly vital. PV inverter service life advantages of independent testing in
for power plant-level risk assessment and mitigation. He has
expectations began at less than five years identifying the best products to buy – nearly 15 years of experience in the solar PV industry.
in the 1990s. Today the market expects as opposed to brand, warranty terms
a central inverter to last at least 20 years. and datasheets alone – module quality Michael Mills-Price is head of PVEL’s inverter and
Inverters were once expected to convert has improved. We hope to see inverter energy storage business. He has nearly 20 years
of renewable energy industry experience and has
DC electricity to AC electricity – and not quality similarly improve over time. 
authored over 30 technical innovations, patents and
much else. Today they communicate whitepapers to advance the state of the industry.
with complex monitoring systems and To learn more about PVEL’s inverter testing He specialises in power electronic devices and the
diagnose system performance problems services or to access the full test reports interface of renewable technologies to the broader electrical
in real-time. behind PVEL’s PV Inverter Scorecard, contact power grids. Michael joined PVEL team in 2014 to create a suite
of performance and resiliency inverter testing strategies to
Despite this dramatic technical Michael Mills-Price at michael.millsprice@
benchmark commercially available products.
evolution, inverter procurement strate- pvel.com

98 | August 2019 | www.pv-tech.org