Professional Documents
Culture Documents
net/publication/228652218
Article
CITATIONS READS
22 1,659
6 authors, including:
F. P. Baumgartner
Zurich University of Applied Sciences
42 PUBLICATIONS 240 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
European Liaison on Electricity Committed Towards long-term Research Activities for Smart Grids
(ELECTRA IRP) View project
All content following this page was uploaded by Roland Bründlinger on 16 May 2014.
ABSTRACT: This paper provides an overview of the prEN 50530, the upcoming European Standard for measuring the
overall efficiency of PV inverters. It explains in depth the approach and methodology introduced in the standard for a com-
bined assessment of the conversion as well as the Maximum Power Point tracking efficiency. In addition, a background in-
sight on the test conditions, the necessary test equipment and interpretation of the results is provided.
1 INTRODUCTION
2 APPROACH AND METHODOLOGY
For an accurate estimation of the performance and
economy of PV systems, the close-to-reality characterisa- Based on the approach published in [5], taking into
tion of the performance of the balance of system compo- account the results of recent research activities ([6][7][8])
nents is essential. With respect to PV inverters, manufac- and experiences of a number of European inverter labora-
turers of PV inverters have been well aware that per- tories, a comprehensive methodology has been developed
formance, together with reliability is one of the main fac- which allows to characterise the overall efficiency of PV
tors to which customers pay great attention to. inverters.
However, due to the absence of broadly accepted This methodology is based on the idea of dividing the
standard test procedures, performance testing of PV in- PV inverter system into two functional blocks, the MPP
verters up to now has been done in various ways, such as Tracker, which aims to maximise the power delivered by
e.g. specified in [1] or [2], leading to inconsistent results the PV array, as well as the DC to AC converter, which
and making the comparison of the performance of differ- converts the DC into AC power. To both blocks, an effi-
ent inverter models difficult for the user. ciency value can be assigned and by combining both val-
To overcome this unsatisfying situation, an interna- ues (ηMPPT and ηCONV) for a given operating condition, the
tional working group1 has been set up in late 2006 to start overall efficiency ηt can be calculated.
with the development of a new standard for the perform- However, in order to get correct results, it is crucial
ance characterisation of PV inverters which has been sent to consider the fact that the MPP Tracker also determines
out for comments as draft European Standard in January the input voltage VDC of the DC/AC converter, which
2009 [3]. then influences the DC/AC conversion efficiency ηCONV =
The objective of the work was to define an overall f(PDC,VDC) [7]. For the characterisation of the whole in-
efficiency, taking into account both, conversion effi- verter, it is therefore necessary to simultaneously meas-
ciency as well as the Maximum Power Point Tracking ure both efficiency values in order to ensure that ηMPPT
(MPPT) accuracy and to develop methods and proce- and ηCONV are measured at the same input voltage (VDC).
dures for the characterisation of these parameters in the The methodology in the new standard emphasises
laboratory. By this, a more differentiated assessment of this fact by requiring a combined measurement of the
the performance characteristics of PV inverters will be performance under steady state conditions, covering
possible, compared to current methods defined in exist- steady state MPPT accuracy and DC to AC conversion
ing standards such as [1] or [2]. efficiency for the determination of the overall efficiency.
In summer 2009, the draft standard was positively In addition to the performance under steady state
voted on the European level with a large majority of 90% conditions, i.e. constant irradiance and temperature of the
of the weighted votes. Also a number of comments were PV array, the new standard also introduces a test proce-
brought forward by experts from the national commit- dure for assessing the characteristics under dynamically
tees, which will be taken into account for the final docu- changing conditions.
ment, which is expected to be published by the end of Besides the methodology, also the measurement con-
2009. ditions and the test equipment used play an important
In addition, the new performance standard will com- role and fundamentally influence the quality of the re-
plement the recently published standard for datasheet and sults, a fact which has e.g. been highlighted in [8].
name plate information for PV inverters [4]. Hence, one of the key objectives of the new prEN 50530
is also to introduce and define appropriate requirements
1 for the conditions, under which measurements have to be
DKE AK 373.0.3 PV Systems Technology
made as well as the equipment to be used. 2.2 Test environment
In the following sections, the paper will describe the Fig. 1 shows the recommended test setup for the
new procedures, test conditions and environment as well measurement of the overall efficiency. In addition to the
as data analysis and presentation. devices already specified in the previous standard for
efficiency measurement [1], the new prEN 50530 re-
2 THE prEN 50530 TEST SEQUENCE quires the use of a PV simulator as DC source, which
provides information on its current MPP conditions
2.1 Definitions (specified by VMPP,PVS and IMPP,PVS).
For a common understanding of the parameters and On the AC side, either a (stable) AC grid or a grid
correct interpretation of the measurements, adequate, simulator can be used.
clear and targeted definitions are necessary. Accordingly,
the new standard [3] introduces a comprehensive set of
definitions required for the characterisation of the overall
performance of the inverter.
∫p
0
MPP (t ) ⋅ dt ∫p
0
DC (t ) ⋅ dt Parameter cSi I/V
curve
TF I/V
curve
ηt Overall efficiency, as ratio of the AC energy out- Irradiance dependency of
put to the theoretically available energy in the VMPP
MPP within a defined measuring period TM. VMPP G = 200W / m 2 0,95 0,98
TM
vL 2 H =
∫p AC (t ) ⋅ dt
η t = η CONV ⋅η MPPT
VMPP G = 1000W / m 2
ηt = T 0
M
or Fill factor (voltage)
∫ pMPP (t ) ⋅ dt
0
FFV =
VMPP; STC 0,8 0,72
VOC , STC
where pMPP(t), pDC(t) and pAC(t) are the instantaneous val-
ues of the theoretically available MPP power, the DC Fill factor (current)
input power, and the AC output power of the device un- I MPP; STC 0,9 0,8
der test, respectively. FFI =
I SC , STC
These parameters provide the basis for computing the (simulated) MPP voltage (VMPP,PVS), which covers the
I/V curves necessary for the test. For this purpose the voltage range of typical MPPT operation under normal
prEN 50530 defines a common I/V curve model specifi- conditions. In case the MPPT operates outside the volt-
cally suitable for MPPT performance tests. age range mentioned above, the requirement is extended.
The model, which is principally based on a single In addition to the steady-state requirements, the PV
diode equivalent circuit allows an explicit calculation of simulator must also be capable to emulate the I/V curves
the PV current IPV as a function of the PV voltage VPV for transient changing values of irradiance and tempera-
and thus overcomes the problems associated with the im- ture. In this case, it is important to minimise any steps
plicit nature of the formulas of the classical single diode which may occur during the transient change.
models.
Fig. 2 and Fig. 3 show some exemplary I/V and P/V Regarding to its dynamics, no definitive limits or
curves of the two different cell technologies specified in minimum requirements are mentioned, mainly due to the
Table 1 for an MPP voltage of 100 V and an MPP power fact that the actual demand on the dynamic characteris-
of 1000 W. tics of the PV simulator varies strongly depending on the
properties of the device under test. Therefore, it is practi-
12
cally impossible to set minimum requirements which
10
cover all possible input and MPPT characteristics of all
8 devices under test. However, it is clearly noted that an
Current [A]
works as intended with the device under test and that the
400
actual parameters generated by the PV simulator suffi-
200 ciently match the settings.
0
0 20 40 60 80
Voltage [V]
100 120 140
AC power supply
On the AC side of the device under test preferably a con-
Fig. 2 Exemplary I/V curves (c-Si characteristics) for trollable AC source (e.g. grid simulator) should be used,
MPPT measurements according to [3] that provides a stable AC voltage. If no controlled AC
source is used it is necessary to verify the influence of
e.g. changing AC voltage on the performance of the de-
vice under test.
MPP voltage of the simu- Simulated I/V characteristic 2.5 Test conditions for dynamic tests
lated I/V-characteristic of (see I/V curve model in sec- The dynamic MPPT efficiency und varying irradi-
the PV generator tion 2.2) ance conditions is characterised by using a ramp se-
quence consisting of irradiance ramp with different gra-
dients as well as irradiance levels.
VMPPmax (0,8·VDCmax a) c-Si In detail, ramp gradients ranging from 0,5 W/m²/s up
VDC,r c-Si to 100 W/m²/s are used for the tests with two different
irradiance levels, from 100 to 500 W/m² (low to medium
VMPPmin c-Si irradiance) as well as from 300 to 1000 W/m² (medium to
a high irradiance). A setting of 1000 W/m² is used as a ref-
VMPPmax (0,7·VDCmax ) TFb
erence to the rated DC power PDC,r of the device under
VDC,r TFb test.
Fig. 4 and Fig. 5 show the principle of the ramp test
VMPPmin TFb sequences for both irradiance levels, low-medium and
a medium-high.
The value whichever is lower shall be used.
b Each test sequence consists of n consecutive ramps
For devices under test, that are not intended for the operation
(up and down) for every gradient of x W/m²/s as men-
with thin-film technologies, these tests need not be performed.
tioned previously. At the high as well as the low level,
the irradiance is kept constant for a certain dwell time.
MPP power of the simulated I/V characteristic normal- Before each test, an initial settling time allows the MPP
ised to rated DC power, PMPP,PVS/PDC, r Tracker to stabilise. For the analysis and calculation of
the dynamic MPPT efficiencies, only the periods of the
0,05 0,10 0,20 0,25 0,30 0,50 0,75 1,00 ramps (marked areas in Fig. 4 and Fig. 5) are taken into
account, the initial settling time is not considered.
For each of the I/V curve settings in Table 2, the PV Ramp tests (low - medium irradiance)
simulator is programmed accordingly and for each condi-
Time considered
tion, the parameters needed for the determination of the for evaluation
MPPT as well as the conversion efficiency are measured Dwell time Simulated
irradiance
(high level)
over a period of 10 min. The sampling as well as the ac-
Irradiance (W/m2/s)
Ramp
quisition rate must be sufficient to cover the specific Ramp up
down n Repetitions
x W/m²/s
MPPT and input behaviour of the device under test. This x W/m²/s
500
covers in particular the input ripple typically appearing at
Initial
PV inverters. settling
Based on the data recorded during the measurements, time
the steady state MPPT efficiency (ηMPPTstat) as well as the
100
conversion efficiency (ηCONV) are computed as (ener- Dwell time
0
getic) averages over the measurement period of 10 min 0 t_0 t_up t_dn
Test time (s)
according the definitions (see 2.1). In addition, the over-
all efficiency ηt is calculated as the product of ηMPPTstat Fig. 4 Ramp test sequence (low-medium irradiance)
and the ηCONV according to for the characterisation of the MPPT efficiency under
PAC changing irradiance conditions.
η t = η CONV ⋅η MPPTstat =
PMPP, PVS
From the efficiencies measured at part-load condi- Ramp tests (medium - high irradiance)
300
as well as the CEC efficiency (California Energy Com- Dwell time
Initial
mission) settling time
n Repetitions
0
η CONV ,CEC = 0.04η CONV ,10% + 0.05η CONV , 20% + 0.12η CONV ,30% + 0.21η CONV ,50% 0 t_0 t_up t_dn
Test time (s)
+0.53η CONV ,75% + 0.05η CONV ,100%
Fig. 5 Ramp test sequence (medium-high irradiance)
for the characterisation of the MPPT efficiency under
For the time being, the traditional weightings for the changing irradiance conditions.
European Efficiency are used, however it should be
noted that recent research reported in [9] and [10] ques-
editorial and technical comments from the national stan-
dardisation committees have been received which will be
In addition to the dynamic ramp test sequences, considered for the final standard by the working group.
prEN 50530 also describes a test sequence at very low However the procedure outlined in this paper will not be
irradiance, which is used to characterise the MPPT per- affected. The final document is expected to be published
formance of the device under test during start-up and by the end of the year 2009.
shut-down periods (Fig. 6). The new standard marks an important step compared
to the current situation and will enable a more differenti-
ated assessment and benchmarking of the performance
100
characteristics of PV inverters.
Ramp up Ramp down
at 0.1 W/m²/s at 0.1 W/m²/s
Irradiance (W/m²/s)
80
6 REFERENCES
60
[1] EN 61683 PV systems - Power conditioners - Proce-
40 dure for measuring efficiency, CENELEC, 2001
Initial Time period of [2] W. Bower et al., Performance Test Protocol for Eva-
settling measurement
luating Inverters Used in Grid-Connected PV Systems
20
time Simulated
irradiance [3] Final draft FprEN 50530, Overall efficiency of PV
0
0 500 1000 1500 2000 2500
inverters, CENELEC, December 2008
Test time (s) [4] prEN 50524: Data sheet and name plate for photo-
voltaic inverters
Fig. 6 Additional start-up and shut down test [5] H. Häberlin et al., Total Efficiency η-tot – A new
For each ramp gradient and irradiance level, the dy- Quantity for better Characterisation of Grid-Connected
namic MPPT efficiency ηMPPT,dyn is calculated according PV Inverters, 20th EUPVSEC, Barcelona, 2005
to the definition. [6] B. Bletterie, R. Bründlinger, S. Spielauer, Quantify-
Generally it is important to consider that the dynamic ing dynamic MPPT performance under realistic condi-
MPPT test sequences are not intended to provide an es- tions first test results – The way forward, 21st
timation of the “average” MPPT efficiency under real EUPVSEC, Dresden, 2006
irradiance conditions. Accordingly, the standard does not [7] F. Baumgartner et al., Status and relevance of the DC
require calculating a single “dynamic MPPT efficiency” voltage dependency of the inverter efficiency, 22nd
nor does it define weighting factors for the calculation of EUPVSEC, Milano, 2007
a weighted average. [8] B. Bletterie et al., I-V curve modelling for inverter
Instead, the aim of the test sequence is to provide a testing – a critical review of current approaches, 22nd
means for a fundamental characterisation of the proper- European Photovoltaic Solar Energy Conference, Mi-
ties and performance of the MPPT function integrated lano, 2007
into the device under test. [9] Bletterie et. al.: Redefinition of the European effi-
More details on these issues, including exemplary ciency – Finding the compromise between simplicity and
test results can be found in [12]. accuracy, EUPVSEC, Valencia 2008
[10] B. Burger et. al.: Are we benchmarking PV inverters
on basis of outdated definitions if the European and CEC
3 CONCLUSIONS efficiency?, EUPVSEC, Hamburg, 2009
[11] C. Bendel, P. Funtan, T. Glotzbach, J. Kirchhof, G.
This paper provides an overview of the prEN 50530, Klein, Results of the OPTINOS Project – Deficits and
the upcoming European Standard for measuring the over- unsureness in Test Procedures of Photovoltaic Inverters,
all efficiency of PV inverters. The new testing introduced 23rd European Photovoltaic Solar Energy Conference
in the document provides the basis for a comprehensive and Exhibition, Valencia, Spain, 2008.
characterisation of the performance of PV inverters. [12] H. Häberlin, New test procedure for Measuring Dy-
The prEN 50530 introduces the definition of the namic MPP Tracking Efficiency at Grid connected PV
overall efficiency, taking into account both, conversion inverters, EUPVSEC, Hamburg, 2009
efficiency as well as the Maximum Power Point Tracking
(MPPT) efficiency. Clearly defined test methods and de-
tailed requirements for the measurement equipment and
test setup allow an accurate and reproducible characteri-
sation of the performance of PV inverters in the labora-
tory.
For developers of inverters or MPP Trackers, the
ramp tests show the behaviour of the inverters under
various conditions and improper control behaviour under
certain conditions can be identified easily. This helps to
improve the MPP Tracking algorithms and the control
schemes of the inverters.
Within a single-stage procedure (UAP, Unique Ac-
ceptance Procedure) the draft standard has been accepted
by the majority of the European countries. Some minor