MAXIMUM POWER POINT TRACKING IN THE PV GRID-CONNECTED INVERTERS OF

5KW
v. Salas and E. Olias
Electronic Technology Department, Universidad Carlos III de Madrid, Leganes (Madrid), SPAIN

ABSTRACT Pnom, for values from 0.1 Pnom to 1.0 Pnom. Although
there is no known how those results have been reached.
The aim of this article is to describe how closely PV In other papers [5], [6], so called dynamic MPP-Tracking
grid-connected inverters operate at the actual maximum Efficiency has been analysed although they have only
power point. To carry this study out six 50 Hz single-phase presented methods to measure it in a very short time
inverters were selected from European market, around of (seconds) and using some PV array simulator.
5 kW. Each one of them was put into an outdoor grid-
connected system installed in Leganes (Madrid), Spain. However, in this paper there are been carried out an
Different electrical signals with respect to irradiance, PV analysis of the maximum power point tracking in function
module temperature and local time were measured. DC of the power for an entire day, using an outdoor PV grid-
voltage and maximum power point tracking were connected system. That analysis has been done under
analysed. different atmospheric condition.

INTRODUCTION PHOTOVOLTAIC GRID-CONNECTED SYSTEMS

Currently the management of energy sources The main objective of the PV grid system is to
represents a fundamental problem for the development evacuate the greatest quantity possible of energy.
and prosperity of any community. As a result, two major Consequently the energy production of the photovoltaic
problems exist: energy sources, and the atmospheric systems is the parameter that judges the efficiency of
pollution from the residues from conventional sources. these systems. Nonetheless, the photovoltaic generator
Taking this into consideration, it is necessary to which is the only element which generates energy, shows
optimize energy resources such as the use of alternative a nonlinear characteristic and their production depends
energy sources. The main characteristics of such sources mainly on the irradiance (intensity and spectral), ambient
include their renewability and their limited contribution to temperature and the equivalent impedance (inverter +
contamination. Photovoltaic solar energy is in this grid) connected. But this equivalent impedance is hardly
category and its use has also notably increased in industry ever equal to the equivalent impedance to reach the
over the past few years, [1], [2], [3]. maximum energy from the PV generator. In this case, a
Common distributed generators, particularly PV system configured with PV generator + DC/AC converter +
generators, are increasingly being connected to utility grid presents production loss because it does not operate
grids to contribute electrical power to the utility grid so as in the maximum power point (MPP). That is to say, there
to meet power demands as well as to meet electric will not be optimum utilisation of the PVarray.
consumer demands for alternate sources of power, With the purpose of solving this problem it is possible
However, the performance of a PV system relies on the to incorporate in the control of the inverter a maximum
operating conditions. Subsequently, the maximum power power point tracking (MPPT) algorithm to maximize the
extracted from the PV generator depends strongly on energy harvested from a PV array. Besides including the
many factors. In this way, the maximum power point typical functions assigned to the inverters, this control will
(MPP) is variable throughout the day. Hence, PV grid be able to adapt the equivalent impedance so that the
connected systems have some MPP tracking method. maximum power can be extracted from PV generator,
Many proposals with reference to maximum power PMPP(t), for every weather condition.
point tracking (MPPT) in those systems have been done. One of the first tasks to be completed in this article has
In fact, in the catalogue and handbook of PV inverters a been to review the state of the technique inverter. This
MPPT efficiency (that is the ratio of the energy obtained by was carried out by compiling three hundred and eighty
the inverter from a PV array, to the energy obtained with seven different models below 10 kW, [7], [8]. From this
ideal MPP tracking over a defined period of time) hardly compilation it has been possible to check that in all
ever is given. However, when that number is given it is inverters a maximum power tracking module has been
assumed for the rated power. Notwithstanding, on the few found.
occasions when this is reached and nothing is said Many MPP tracking algorithms have been proposed
regarding what occurs for the rest of the power. applied to PV stand-alone systems [9] and PV grid-
Some papers have measured the MPPT efficiency. For connected systems, [10], [11], amongst which are the
example [4] gives MPPT efficiencies (0/0) for five different Perturb and Observe, Incremental Conductance, ripple
PV grid-connected inverters at different nominal power,

978-1-4244-2950-9/09/$25.00 2009 IEEE 000193

based, Sliding-Mode observer, and OCC (One cycle Sunny day (stable conditions):
control) methods.
Currently, there are plenty of standards relating to Figures 2, 3 and 4 show the evolution of parameters. In
quality of PV modules and inverters. However, Figure 2, the PV voltage and PV power generator are
unfortunately until now there has been no standard shown for a full day. As can be seen, the PV voltage
relating to MPPT measured applied to PV grid inverters, remains more or less constant up to a determined
even though under draw. irradiance value of approximately 200 W/m 2 Such values
remain constant during the day until nightfall. The concave
EXPERIMENTAL RESULTS AND DISCUSSION form that can be observed is due to the rising of the
ambient temperature of the PV cells. It yields a diminishing
In order to evaluate how the maximum power point of the PV current.
varies in real inverters, six 50 Hz single-phase inverters, of
around 5kW, have been selected, from the European
market. They have been chosen according to different
transformer options: three inverters have an LF
transformer, four of them an HF transformer and the
remaining four have no transformer.

To this end, every PV grid inverter was put in a low

voltage, outdoor PV system of 3 kW,connected to the grid
and located at the Universidad Carlos III de Madrid
(Leganes-Madrid- 40 31' North, Spain). The PV system
was set up with 6 serial PV modules of SANYO HIP190
BE. The same PV array was used for all inverters.
PHOTOVOLTA IC
GENERATOR
Figure 2 PV voltage and PV power generator versus
Local Time for a full day (stable conditions)
PV inverter UTI LI TY
GR ID In order to carry out a more detailed study of the MPP
(device under test)...-. f -..;T
tracking, the data from the Figure 2 was divided into three
sections: start, disconnection and rapid variations of the
irradiance.
Measurin g syste
300 120

High precision lPVVOitag~adla!!~eJ

Power Analyzer

I .--J
250 100
LMG500 Zimmer

Figure 1 Measure system used for gathered _ 200 E

z
/
80

meteorological variables and electric signals m ~

io m
.g
--I'
150 60

"~
>
This installation was conn1ected to a high precrsron

->
6: 100 40
Power Analyzer, LMG500 Zimmer, (basic accuracy 0.03
%) which received and gathered, every 0.05 seconds,
.s-:
50 20
values from meteorological variables and electric signals
at the input and output of the inverter, such as DC voltage
:
o~
and current, irradiance (G) and PV array temperature (Tc),
Figure 1. Indeed, a great deal of measurements were Local time (hh:mm:ss) "
gathered under different sky conditions. In order to Figure 3 PV voltage and PV power generator versus
analyze the maximum power point tracking, two Local Time for a transition period during sunset (stable
parameters have been analyzed: maximum voltage point conditions)
and MPPT efficiency.
In this regard, Figure 3 shows a transition period during
Different meteorogical conditions will be analysed. The sunset when an operation voltage is established in the last
experimental results that will be shown are grouped into few minutes (this is assumed to be the maximum voltage).
two groups (in terms of irradiance): sunny day (stable In addition, in Figure 4 the disconnection process is shown
conditions) and cloudy day (non-stable conditions). during which various oscillations can be observed until the
inverter is switched off definitively.

978-1-4244-2950-9/09/$25.00 2009 IEEE 000194

 On the other hand, Figure 5 shows the evolution of ~g!. genetatot Irradiance pv powet g~
t

-
ratio between irradiance and PV current and PV voltage ,
and irradiance. This new parameter, that is the ratio
between irradiance and PV current, has been taken into .~ .... -.roLL
t

account as a merit parameter for the study of the evolution t

of the MPP tracking. By observing the transition of A ... -

connection of the PV inverter, it can be seen how the ratio
remains constant due to the ambient temperature (with a
~~ ~
slight increase during the day).
250 70
~

Li
225 [) J 'PI v-,.".
200
..\c 60
o
,,- - - - - -
\. i
-,
175 50 N
. Loca l Time (hh:mm:ss)
~ 150 40 ~
'\.. ae
~
co
~
"0
>
125
-, 30 :5
>
a-
100
-, ~ Figure 6 The evolution of PV voltage and power and
<,
75 20
irradiance for an none-stable day
50

25
<, 10

IP\IVOibge ~

- - - - . . . . . ."e . . . . .
Irr ad ianc e !

:<: e e 1<: ~ 1400 N

Local Time (hh :mm: ss) E
~+~~$i~~\...t-"1 1 200 ~
Figure 4 PV voltage and PV power generator versus 1000 B
Local Time for a transition period during disconnection
C

'D "

-
process (stable conditions) ]

_.
"7
~

/
"/
/
/
""'\.
\.

\
,."

'-
Local Time (hh :m m:ss)

Figure 7 The evolution of PV voltage, ratio between

irradiance and the PV current and and irradiance for an
/ \ none-stable day
I \
.:> ...
I I/lrradiance PV voltag e Irradiance I '\..
...... . . . . . .
However, in Figure 7 it can be observed how the ratio
between irradiance and the PV current does not remain
constant throughout the whole day (as was the case with
l oc aI Tlme j hh:mm :ss )
the irradiance on sunny day when there were no suddenly
Figure 5 The evolution of ratio between irradiance and changes).
PV current and PV voltage and irradiance

Non-stable conditions:
,...
L

It is necessary to continue the study of the behaviour ~~ I N

1ZOO .E

-
of the PV inverter with respect to maximum power point J V
I
~
~
tracking in Figures 6,7,8 and 9. I 1000

--- - "" .., .~

~
In order to do so, the temporal evolution of different -----~-- -
variables that can be observed are taken from a highly
variable day (in terms of irradiance). _ 1 -1
L.....--'
-
~
I.
'-Ji..... '\:
' _ I_L _ _ _1
U\"",. 1
i Irradiance/l PV voltage Irradiance i
Figure 6 represents the evolution of PV voltage and
power and irradiance. As can be seen, this is quite similar Local Time (hh:mm:ss)

to that shown for a sunny day. That is to say, there is a Figure 8 The evolution of PV voltage, ratio between
connection process after which the voltage operates irradiance and the PV current and and irradiance for an
around the supposed maximum voltage. none-stable day (much more limited temporal window)

978-1-4244-2950-9/09/$25.00 2009 IEEE 000195

 It can be seen that the more different such a ratio is, [6] H. Haberlin and L. Borgna, "A new Approach for Semi-
the greater the variation of the irradiance is. These Automated Measurement of PV Inverters, especially MPP
variations are better seen in Figure 8, where the same tracking Efficiency, using a Linear PV Array Simulator with
parameters that Figure 7 are shown, but with a much more High Stability". Proceedings 19th EU PV Conf., Paris,
limited temporal window. France, 2004.

CONCLUSIONS [7] Salas, V., Olias, E., "Overview of the state of technique
for PV inverters used in low voltage grid-connected PV
This paper reports on the analysis carried out for systems: Inverters below 10 kW", Renewable and
maximum power point tracking of low voltage grid Sustainable Energy Reviews, Volume 13, Issues 6-7,
inverters. Firstly, real measurements on DC voltage were August-September 2009, Pages 1541-1550
taken under different sky conditions. In some cases an
instability was found which causes power loss. From [8] V. Salas et. aI., "An analysis of the technical
measurements taken for two types of days, the following exigencies and CE marking relative to low voltage (less
conclusions can be extracted. than 5 kW) photovoltaic inverters marketed in Spain".
Renewable and Sustainable Energy Reviews, Volume 13,
Firstly, there is start-up process in the early morning Issues 6-7, August-September 2009, Pages 1635-1640
which lasts various minutes and by which a PV voltage is
established. This is assumed to be the maximum voltage. [9] M. Calais, J. Myrzik, T. Spooner, and V. G. Agelidis,
"Inverters for single-phase grid connected photovoltaic
Secondly, a new merit parameter has been defined. It systems-An overview," in Proc. IEEE PESC'02, vol. 2,
relates the irradiance and the PV current, which is 2002,pp.1995--2000.
assumed to be the current in the maximum power point.
This parameter will help to verify anomalies in the PV [10] F. Blaabjerg, Z. Chen, and S. B. Kjaer, "Power
inverter. The evolution of a parameter for stable conditions electronics as efficient interface in dispersed power
has been proved, slightly rising for the duration of the day. generation systems," IEEE Trans. Power Electron., vol.
However, for non-stable conditions some instabilities were 19,no.5,pp.1184--1194,Sep.2004.
noted regarding the inverter measurements. These were
somewhat unexpected. [11] M. Calais and V. G. Agelidis, "Multilevel
converters for single-phase grid connected photovoltaic
ACKNOWLEDGEMENTS systems-an overview," in Proc. IEEE ISIE'98, vol. 1,
1998, pp. 224-229.
Part of this work was supported by the NUMANCIA
Project, Code: S-0505/ENE/031 0, funded by the
Community of Madrid, Spain.

REFERENCES

[1] "Trends in Photovoltaic Applications Survey Report".

International Energy Agency. 2009

[2] Salas, V., Olias, E., "Overview of the photovoltaic

technology status and perspective in Spain", Renewable
and Sustainable Energy Reviews, Volume 13, Issue 5,
June 2009, Pages 1049-1057

[3] "An argument for Feed-in Tariffs", EPIA January 2008

[4] Wildenbeest, E.J. et. ai, "Recent test results of 6

commercial inverters for grid connected PV systems",
Photovoltaic Energy Conversion, 1994, Volume: 1,
page(s): 909-912 vol.1

[5] H. Haberlin et. ai, "Measurement of Dynamic MPP-

Tracking Efficiency at Grid-Connected PV Inverters". 21st
EU PV Conf., Dresden, Germany, September 2006.

978-1-4244-2950-9/09/$25.00 2009 IEEE 000196