Conference PaperPDF Available

Deviations of results for energy yield from efficiency rankings of micro-inverters

Authors:

Abstract and Figures

To determine and rank the performance of micro-inverters, DC-AC conversion efficiency over the full range of load conditions (including inherent MPPT accuracy) is usually used as the principal indicator. However, for the end-user the AC energy yield fed into the grid is the most important value for benchmarking. To compare efficiency and yield of most micro-inverters available on the world market in 2014, an in-and outdoor test field at the University of Paderborn have been set up. The inverters have been fed by identical and calibrated crystalline silicon PV modules of 215 W at STC. To monitor accurately AC power output and yield, each of the micro-inverters has been equipped with a calibrated precision electricity meter. For micro-inverters that require control units for grid-feeding this has been purchased also. The test field is equipped with pyranometers in the module plane. The comparison includes energy yield over almost one year, efficiency-load characteristics, recovery times after low irradiance levels, cost consideration: While the market is quite new, the range of purchase costs varies considerably between the models in comparison, sometimes inverter costs are higher than module costs, particularly if an additional grid-connection or interface device is necessary for operation. The weighted conversion efficiency according to EU and CEC standards has been measured and compiled. While some inverters have been optimized for high irradiance levels, they ranked better at the CEC-efficiency, others performed very well also for low irradiance levels, thus ranking higher at in the EU-efficiency tables. These results deviated from the actual energy yield measurements, which showed a slightly different ranking. One device showed an extreme deviation of the efficiency rating from the yield results. An accurate but very slow MPPT algorithm that hardly could follow quickly changing irradiance conditions has possibly caused this effect. Apparently, some inverters are been optimized to show excellent EU and CEC efficiency ratings. Further investigations are on the way, including temperature effects.
Content may be subject to copyright.
DEVIATIONS OF RESULTS FOR ENERGY YIELD FROM EFFICIENCY RANKINGS
OF MICRO-INVERTERS
Stefan Krauter and Jörg Bendfeld
University of Paderborn, Electrical Energy Technology – Sustainable Energy Concepts
Pohlweg 55, D-33098 Paderborn, Germany
stefan.krauter@upb.de
ABSTRACT: To determine and rank the performance of micro-inverters, DC-AC conversion efficiency over the full
range of load conditions (including inherent MPPT accuracy) is usually used as the principal indicator. However, for
the end-user the AC energy yield fed into the grid is the most important value for benchmarking. To compare
efficiency and yield of most micro-inverters available on the world market in 2014, an in- and outdoor test field at the
University of Paderborn have been set up. The inverters have been fed by identical and calibrated crystalline silicon
PV modules of 215 W at STC. To monitor accurately AC power output and yield, each of the micro-inverters has
been equipped with a calibrated precision electricity meter. For micro-inverters that require control units for grid-
feeding this has been purchased also. The test field is equipped with pyranometers in the module plane. The
comparison includes energy yield over almost one year, efficiency-load characteristics, recovery times after low
irradiance levels, cost consideration: While the market is quite new, the range of purchase costs varies considerably
between the models in comparison, sometimes inverter costs are higher than module costs, particularly if an
additional grid-connection or interface device is necessary for operation. The weighted conversion efficiency
according to EU and CEC standards has been measured and compiled. While some inverters have been optimized for
high irradiance levels, they ranked better at the CEC-efficiency, others performed very well also for low irradiance
levels, thus ranking higher at in the EU-efficiency tables. These results deviated from the actual energy yield
measurements, which showed a slightly different ranking. One device showed an extreme deviation of the efficiency
rating from the yield results. An accurate but very slow MPPT algorithm that hardly could follow quickly changing
irradiance conditions has possibly caused this effect. Apparently, some inverters are been optimized to show excellent
EU and CEC efficiency ratings Further investigations are on the way, including temperature effects.
Keywords: Inverter, Small Grid-connected PV Systems, Power Conditioning
1 INTRODUCTION
Micro-inverters are inverters that are connected
principally to a single PV module (sometimes to two
modules), so each module-inverter combination acts as
an independent power plant. The micro-inverter consists
of a maximum power point tracker (MPPT), the DC-AC
inverter, and an islanding protection unit (see e.g., [1]).
To achieve higher power outputs, several module-inverter
combinations are interconnected in parallel on the AC
output side. This configuration offers various advantages:
Easier planning and easier installation, easy up- and
downscaling of the power plant, including extensions or
repair that could be carried out during power plant
operation. Effect of shadowing is very limited, and due to
low system voltages, potential induced degradation (PID)
does not occur. Logistics is simplified. However, costs of
power plants based on micro inverters are about 20%
higher. Some of the inverters cannot be operated by
themselves and require a control unit (often combined
with a remote shutdown option and a monitoring system),
thus adding extra costs. Also, conversion efficiency may
not be as high as for central inverters. Due to smart
master-slave concepts centralized solutions with multiple
but relatively large inverters may offer higher yields
under weak light conditions.
2 INDOOR TESTS
Due to the reproducible test conditions in the indoor
lab, the inverters have been examined individually with
predefined and controlled input data. Input has been a PV
module simulator with data set according to the modules
used in the outdoor test. The principal output data
recorded has been the delivered AC power of the
inverters. Besides input power, output is also a function
of input voltage. If input voltage is getting too low, the
inverters even stop operating. The following
examinations are based on the possible range of input
data (including voltage) given the specific PV module
also used for the outdoor investigation.
2.1 Peak and Weighted Efficiency
Peak efficiency is often reached close to the
maximum load of the inverter. Peak efficiency in data
sheets is not a helpful value since most of the time the
inverters operate in the range of 20% to 30% of their
rated power. Consequently, an adequately weighted
efficiency is a more adequate measure. One type of
weighted efficiency is the so-called “European
Efficiency”; the other is the “CEC efficiency” by the
California Energy Commission (CEC). CEC efficiency is
computed as an average value of DC-AC conversion
efficiencies at six pre-defined outputs between 10% and
100% of the rated power. For the European Efficiency
the weighting factors are lower for the high relative
power values.
2.2 Test Setup & Results
Ten different output power values from zero up to the
maximum operational power of the PV modules have
been considered (they are roughly identical for all seven
micro inverters). The values used (adjusted by controlling
the current) are: 10 W, 25 W, 50 W, 75 W, 100 W, 125
W, 150 W, 175 W, 200 W, and 225 W. The averaging
time for the power meter used (Zimmer LMG670) was
set to 500 ms. One micro inverter showed an unstable AC
power output for values below 100 W, for this device the
averaging time had to be extended to 2 s.
Proceedings of 32nd European Photovoltaic Solar Energy Conference (EUPVSEC), 20-24 June 2016, Munich, Germany
Figure 1: Measured DC-AC conversion efficiencies as a
function of DC input power for ten micro-inverters.
The European conversion efficiency ηEuro is
calculated according to:
ηEuro = 0.03 · η5% + 0.06 · η10% + 0.13 · η20% (1)
+ 0.1 · η30% + 0.48 · η50% + 0.2 · η100%
Table I: Ranking of all micro-inverters by European
conversion efficiency (same types as Table III)
Rank Manufacturer European
efficiency
Relative
eff. vs #1
1 SMA 95.4 % 100.0 %
2 Enphase 95.2 % 99.8 %
3 Hoymiles 95.0% 99.5 %
4 Power One 94.6 % 99.2 %
5 Envertech 93.2 % 97.7%
6 Involar 92.7 % 97.2 %
7 Changetec 90.9 % 95.3 %
8 AEconversion 90.3 % 94.7 %
9 Enecsys 90.3 % 94.7 %
10 Ienergy 89.9 % 94.3 %
The weighting for ηCEC, as proposed by the California
Energy Commission (CEC) is determined by:
ηCEC = 0.04 · η10% + 0.05 · η20% + 0.12 · η30% (2)
+ 0.21 · η50% + 0.53 · η75% + 0.05 · η100%
Table II: Ranking of all micro-inverters by CEC
conversion efficiency (same types as Table III)
Rank Manufacturer CEC
efficiency
Relative
eff. vs #1
1 Enphase 95.6 % 100.0 %
2 Power One 95.5 % 99.9 %
3 Hoymiles 95.4 % 99.8 %
4 SMA 95.1 % 99.5 %
5 Envertech 94.1% 98.4 %
6 Involar 93.9 % 98.2 %
7 Enecsys 92.0 % 96.3 %
8 Ienergy 91.5 % 95.7 %
9 AEconversion 91.2 % 95.5 %
10 Changetec 90.9 % 95.1 %
For each power-step 25 values have been recorded.
The measured DC-AC conversion efficiencies of all
inverters are shown in Fig. 1. Based on those
measurements, the European efficiencies and the CEC
efficiencies for the micro inverters have been calculated
according to (1) and (2). The ranking considering the
European conversion efficiency is shown in Table I, the
ranking for the CEC conversion efficiency is given in
Table II.
3 OUTDOOR TESTS
3.1 Set-up for testing
For the PV outdoor test lab (see Fig. 1) installed on
the roof of the University of Paderborn (51.707°N,
8.771°E), a specific test system comprising of eight
photovoltaic modules with eight micro-inverters has been
employed. To examine ten inverters, two of them have
been interchanged.
Figure 2: PV outdoor laboratory at the University of
Paderborn for micro-inverter comparison using eight
equal, calibrated PV modules.
Each PV module (QC-CO2 by Hanwa Q-cells)
consists of 60 solar cells. In stated plant, these equal and
calibrated modules are the input for each micro inverter.
The goal of this investigation is to analyze the
performance of the inverters in the lab and under real
operating conditions comparing their energy yield
simultaneously at the same climatic conditions and solar
irradiance. The climatic conditions have been monitored
during the whole test period. The meteorological
monitoring equipment consists of two calibrated
pyranometers in the plane of the module (CMP 21 and
SP2 lite by Kipp&Zonen), a 3-D ultrasonic anemometer
(by Thies), and a thermo-hygro sensor (by Thies).
Figure 3: Scheme of the outdoor test setup for the micro-
inverters (MI) investigated.
Fig. 3 shows the electrical measurement setup. Each
micro inverter has been directly connected to a calibrated
electrical energy meter with a S0-interface. All S0-
interfaces have been connected to an efficiency recording
data server. Fig. 5 shows an example of the collected data
for single day (data recording every 5 min). Due to the
high time resolution, some specific characteristics could
be observed: E.g., starting behaviors, MPP tracking
(accuracy and speed), dropouts and performance in high
resolution can be perceived.
Figure 4: Actual recorded data of AC energy generation
(over an integral of 5 min) for seven micro-inverters
during the day of 10/31/2013.
Figure 5: Actual recorded data of AC energy generation
of seven micro-inverters (over an integral of 5 min)
during the day of 4/6/2015.
Figure 6: Actual recorded AC-energy generated within a
5 min interval during a day with changing irradiance
levels. One inverter adapts quickly (blue line;
ABB/Aurora), the other not (red; Hoymiles MI 250).
Fig. 5 shows an example of the collected data for
single day during spring of the year 2015 (including the
new SMA micro-inverter). It is visible that some
inverters have difficulties to follow the quick change of
irradiance levels during that day (e.g. Ienergy, and
Hoymiles – see also Fig. 6)
A ranking list according to the total AC energy yield
of the micro inverters during the common operation
period is shown in Table III (July 2015 until June 2016,
(Envertech 6/1/2016 until 6/15/2016 only) relative energy
yield vs. #1.
Table III: Ranking of all micro-inverters by measured
AC energy yield (in Paderborn 2015/16, 50 weeks, *2
weeks only)
Rank Manufacturer Type, Model rel. yield
vs #1
1 Power One /
Aurora / ABB Micro-0.25-i 100.0 %
2 SMA Sunny Boy 240 95.2 %
3 Involar MAC 250 94.4 %
4 Enphase M 215 95.0 %
5 AEconversion /
Aptronic INV 250-45 92.4 %
6 IEnergy GT 260 91.8 %
7 Enecsys SMI-S-240W 88.8 %
8 Envertech EVT248 88.0 %*
9 Hoymiles MI 250 78.0 %
10 Changetec ELV 300-25 75.6 %
4 DISCUSSION AND CONCLUSION
4.1 Results
Some of the results could be expected by taking a
close look at the efficiency curves of Fig. 2: For example:
If highest performance occurs at relatively low input
power (and the corresponding low irradiance levels), the
ranking for the European Efficiency (which puts a higher
weight on lower irradiance levels) is higher, while the
CEC ranking is worse due to the higher weighting of high
irradiance levels (this is the case for the Changetec
inverter). The best in the CEC ranking has been the
Enphase inverter, which was second in European
Efficiency ranking. For the SMA inverters the ranking
positions have been the other way round: SMA is
optimized for European irradiance conditions, Enphase
for Californian conditions.
4.2 Correlation of energy yield with efficiency
Some of the rankings by yield (as shown in Table III)
cannot be explained by the curves of Fig. 2 and the
European or the CEC efficiency. We observed significant
differences of the MPPT devices in their capabilities to
adapt to fast changes of irradiance. Best in this discipline
has been the Power One / Aurora inverter, which has not
been the best in the static conversion efficiency
measurements, but in energy yield. The device MI 250
showed an extreme deviation of its efficiency rating from
the yield results. An accurate but slow MPPT algorithm
that had difficulties to follow changing irradiance
conditions has possibly caused this effect.
Fig. 7a shows the relative energy yield as a function
of European efficiency (which is often an important
criterion for the purchase decision of a potential
customers). As can be observed, there is just a quite
vague correlation between those two values with high
deviations. For the example indicated at Fig. 7b it can be
observed that there may not be just a significant
quantitative deviation, but also a qualitative one. For an
increase of 0.4% in Euro-efficiency, a decrease of -22%
in yield may occur. The outcome for yield vs. CEC
efficiency is similar.
Figure 7a: Electrical energy yield as a function of
European efficiency for all tested microinverters.
Figure 7b: Electrical energy yield as a function of
European efficiency: The user cannot expect a higher
yield if he purchased an inverter with a higher European
efficiency – the yield can be significantly lower.
4.3 Outlook
To overcome the considerable deviation between
yield and efficiency measurement results, a
dynamic test will be deployed to find a better
indicator for the relative yield to be expected by the
different inverters.
While the lab tests have been made at a
laboratory temperature conditions (20°-22°C) and
the yield measurements at actual operating
temperature, additional efficiency tests at elevated
temperatures (e.g. 40°C and 60°C) will be carried
out.
5 REFERENCES
[1] W-F. Lai, S-M. Chen, T-J. Liang, K-W. Lee and A.
Ioinovici, "Design and Implementation of Grid Connection
Photovoltaic Inverter," in IEEE Energy Conversion Congress
and Exposition (ECCE), 2012, pp. 2426-2432.
[2] S. Krauter; J. Bendfeld, “Cost, performance, and yield
comparison of eight different micro-inverters”, in 42nd IEEE
Photovoltaic Specialist Conference (PVSC), 2015.
Conference Paper
Full-text available
To compare efficiency and yield of most micro-inverters available on the world market in 20142017, an in- and outdoor test laboratory at the University of Paderborn have been set up. The inverters have been fed by identical and calibrated crystalline silicon PV modules of 215 Wp. To monitor accurately AC power output and yield, each of the micro-inverters has been equipped with a calibrated precision electricity meter. For micro-inverters requiring control units for grid-feeding that has been acquired also. The comparison includes energy yield over one to two years and efficiency-load characteristics. Purchase costs vary considerably between the models in comparison, sometimes inverter costs are higher than module costs, particularly if an additional grid-connection or interface device is necessary for operation. The weighted conversion efficiency according to EU and CEC standards has been measured and compiled. While some inverters have been optimized for high irradiance levels, they ranked better at the CEC efficiency, others performed very well also for low irradiance levels, thus ranking higher at in the EU-efficiency tables. These results deviated from the actual energy yield measurements, which showed a slightly different ranking. One device showed an extreme deviation of the efficiency rating from the yield results. An accurate but very slow MPPT algorithm that hardly could follow quickly changing irradiance conditions has possibly caused this effect. Apparently, some inverters are been optimized to show excellent EU and CEC efficiency ratings. The new inverters featuring two inputs did not show an exceptional performance at the EU- and CEC-ratings (ranking 58 only), but they achieved rank 1 and 2 at the AC energy yield ranking. For the customer the AC yield is a major performance indicator of the microinverter and should be included in the datasheet. Further investigations are on the way, including degradation and temperature effects.
Conference Paper
Full-text available
To investigate efficiency and yield of most micro-inverters available on the world market in 2013-14, an outdoor test field on the University of Paderborn has been set up. The inverters have been fed by identical & calibrated crystalline silicon PV modules of 215 W at STC. To monitor accurately AC power output and yield, each of the micro-inverters has been equipped with a calibrated precision electricity meter. For micro-inverters that require control units for grid-feeding this has been purchased also. The test field is equipped with pyranometers in the module plane for global and diffuse irradiance, temperature, wind speed and wind direction measurement. The comparison includes energy yield over almost one year, efficiency-load characteristics, recovery times after switch-off, cost comparisons (including control units necessary and wiring), basic safety, reliability, warranty duration comparisons, and lifetime costs for generated electricity. While the market is quite new, the range of purchase costs varies considerably between the models in comparison, sometimes inverter costs are higher than module costs. Index Terms — CEC efficiency, energy yield, decentralized PV, European efficiency, module inverters, performance comparison.
Conference Paper
A 200 W photovoltaic (PV) micro inverter is designed and implemented in this paper. The proposed inverter includes a high step-up DC-DC converter and a SPWM H-bridge inverter. It transfers solar energy into sinusoidal voltage waveform. The operating principles and the steady-state analysis of continuous conduction modes of the high step-up converter are discussed. The DSP controller TMS320LF2407A is used to achieve the system starting check, abnormal state detection, and maximum power point tracking as well as islanding detection. Finally, a prototype circuit with input range from 20 V to 40 V and 220 Vrms/200 W output is implemented.