252  Measurement Automation Monitoring, Jul. 2017, no. 07, vol.

63, ISSN 2450-2855

Piotr ZIMOCH
SILESIAN UNIVERSITY OF TECHNOLOGY, DEPT. OF POWER ELECTRONICS, ELECTRIC DRIVES AND ROBOTICS
2 Krzywoustego St., PL-44100, Gliwice, Poland

Modeling and simulation of grid connecting converters for

photovoltaic applications in MATLAB/Simulink
Abstract

This paper presents simulation models of a grid-connecting converter for

photovoltaic applications using MATLAB/Simulink’s Simscape
SimPowerSystems toolbox. Manufacturers of these converters often do not
provide any information related to the topology and control algorithms,
thus impeding the modeling. Simulation models of two possible topologies
of SVM based current controlled three-phase converters are shown and
compared. The simulation results are compared to a 9.6 kW rooftop PV
plant built in Novi Sad, Serbia, by connecting a corresponding PV array
model at the input of the converter models. Simulation results have shown,
that the created models produce results similar to the physical system, and
that both modelled topologies are possible and viable to implement.

Keywords: grid connected converter, solar panel, photovoltaics,

MATLAB/Simulink.

1. Introduction
Grid-connecting converters are inevitable parts of photovoltaic
Fig. 2. One of the two strings consisting of 20 PV modules [3]
(PV) installations, as PV technology produces DC output, thus an
inverter is needed to convert the DC electrical energy to AC.
Specifications of the PV modules and the grid-connecting
Therefore connecting a photovoltaic installation to an electrical
converter are given in tables 1 and 2 respectively.
grid requires a converter. Often the data provided by the producers
of grid-connecting converters is insufficient – there is no Tab. 1. JKM-240P-60 PV module specification [4]
information on the topology or the control algorithms
Quantity Value
implemented, thus modeling of these converters is impeded.
One approach to modeling and simulating PV plants and grid- Maximum power 240 Wp
connecting converters has been described by Ropp and Gonzalez Maximum power voltage 30 V
[1], but the developed model may seem too complicated to an Maximum power current 8.01 A
implementation engineer of a small scale PV installation. A more
Open-circuit voltage 37.2 V
intuitive modeling approach is given in [2], where PSIM software
is used. Short-circuit current 8.56 A
This paper presents simulation models developed in Cell efficiency 17%
MATLAB/Simulink software using SimPowerSystems toolbox. Module efficiency 14.66%
The toolbox provides users with elements such as transistors, Maximum system voltage 600 V (UL)/1000 V (IEC) DC
diodes, sources, etc., therefore modeling is user-friendly, because
Maximum rated current series 15 A
basic modeling comes to drawing the schematic. Implementation
of control algorithms is also intuitive as Simulink is component- Power tolerance ±3%
based and follows a hierarchical structure.
The simulation results are compared to an existing PV
installation designed and implemented in Novi Sad, Serbia [3]. It Tab. 2. Sunny Tripower 8000TL inverter specification [5]
consists of 40 JKM-240P-60 photovoltaic modules divided into
Input
two parallel strings of 20 panels in series, connected to the grid
through a Sunny Tripower 8000TL converter. A single JKM- Quantity Value
240P-60 module can reach up to 240 W of power, thus the whole Maximum DC power 8200 W
installation of 40 modules produces up to 9.6 kW. Figure 1 shows Maximum DC voltage 1000 V
the structure of the system and Figure 2 presents a photograph of MPP voltage range (320-800) V
one of the two parallel strings.
DC nominal voltage 600 V

String A Minimum DC voltage/start voltage 150 V/188 V

Converter
Grid Maximum input current 33 A
Output
Quantity Value
AC nominal power @ 230 V, 50 Hz 8000 W
String B Maximum AC apparent power 8000 VA
Nominal AC voltage 3/N/PE, 230 V/400 V
AC grid frequency; range 50 Hz, 60 Hz; -6 Hz, +5 Hz
Maximum output current 16 A
Power factor 0.8 leading … 0.8 lagging
Fig. 1. Structure of the existing PV installation
Measurement Automation Monitoring, Jul. 2017, no. 03, vol. 63, ISSN 2450-2855  253

2. Simulink models q-axis current’s reference value is set to zero, as that corresponds
to zero reactive power fed to the grid. The outputs of currents’
Two simulation models were prepared in MATLAB/Simulink controllers are transformed to αβ reference frame and used for
using SimPowerSystems toolbox. One of them is simpler - the PV SVM time calculation [2, 7]. The inverter transistors are switched
source is directly connected to the DC link capacitance which acts according to the calculation results.
as a voltage source for an inverter, connected to an infinite grid
a)
through an LC filter (Fig. 3a). The other has an intermediary
DC/DC BOOST converter between the PV source and the DC link
capacitance (Fig. 3b). Both models use a perturb and observe
maximum power point tracking (P&O MPPT) algorithm [6], but
with different output variables. The simpler model uses an MPPT
algorithm that outputs the reference value of the DC link voltage,
while the other model’s MPPT outputs the duty cycle value for the
DC/DC converter. In both cases a PV array built-in model was
used as a source. It takes irradiance and temperature as inputs and
contains simulation models of many available PV modules,
including Jinko Solar JKM-240P-60.
Perturb and observe MPPT algorithm is one of the easiest
b)
methods of tracking the maximum power point of a PV array. It is
based on the principle of adjusting the output value according to
the current and previous PV array’s output power. The output is
changed with a fixed step. If the power rises, the output is further
altered in the same direction, if the power lowers, the direction of
changing the output is reversed. Perturb and observe MPPT
algorithm’s advantage lies in its ease of implementation, while its
major drawback is the fact, that in steady state the operating point
oscillates around the maximum power point, thus wasting some
amount of available energy [6].

a)

Fig. 4. MATLAB/Simulink inverter control subsystem of the model (a) with the PV
array directly connected to the DC link capacitance (b) with an intermediary
BOOST converter between the PV array and the DC link

Depending on the model, the DC link voltage reference value is

different. In the model with the PV source directly connected to the
DC link it is controlled by an MPPT algorithm, while in the model
with an intermediary BOOST converter it is set to constant 600 V.

3. Simulation results
The models of the PV plant with grid-connecting converters
were simulated with nominal irradiation and cell temperature
b) values - 1000 W/m2 and 25Ԩ, respectively.
b) Figures 5 and 6 present the phase currents waveforms, while
tables 3 and 4 show the RMS and THD values of the waveforms.
Power grid is modelled as an infinite power bus, thus the phase-to-
phase voltages are 400 V RMS, 50 Hz sine waves. There is less
power transferred from the circuit with a DC/DC converter as
there is additional power loss in it, but the THD value of the
current is lower. Due to the differences in control algorithms and
additional switching, the model with a BOOST converter is
noticeably slower in reaching steady-state, as it takes about 100
ms, while in the other model it only takes about 40 ms.

Fig. 3. MATLAB/Simulink model of the grid-connecting converter (a) with the PV

array directly connected to the DC link capacitance (b) with an intermediary
BOOST converter between the PV array and the DC link

Both inverter models are controlled by a current control

subsystem with a space vector modulation (SVM) algorithm [2,
7]. Structures of the control subsystems of both models are shown
in Fig. 4. Phase currents and phase-to-phase voltages (Iabc and
Vabc) are measured and fed to the control subsystem along with the
DC link voltage value. Grid voltages are input to a phase-locked
loop (PLL) block, which is the basis of grid synchronization [2].
Currents are transformed to dq reference frame and regulated by
PI controllers. DC link voltage is regulated with a PI controller, Fig. 5. Phase currents waveforms obtained by simulating the model of the grid-connecting
converter with the PV array directly connected to the DC link capacitance
which outputs a reference value of the d-axis current, while the
254  Measurement Automation Monitoring, Jul. 2017, no. 07, vol. 63, ISSN 2450-2855

Tab. 3. RMS and THD values of phase currents waveforms obtained by simulating
the model of the grid-connecting converter with the PV array directly
connected to the DC link capacitance

Ia Ib Ic
RMS 16.13 A 16.18 A 16.13 A
THD 0.72% 0.73% 0.67%

Fig. 6. Phase currents waveforms obtained by simulating the model of the grid-
connecting converter with an intermediary BOOST converter between the PV
array and the DC link Fig. 8. Charts of efficiency and current THD as functions of solar irradiation
obtained from the model of the grid-connecting converter with an
intermediary BOOST converter between the PV array and the DC link
Tab. 4. RMS and THD values of phase currents waveforms obtained by simulating
the model of the grid-connecting converter with an intermediary BOOST As there are additional losses in a grid-connecting converter
converter between the PV array and the DC link with a BOOST converter, efficiency is noticeably larger in the
model with the PV array connected directly to the DC link. On the
Ia Ib Ic
other hand, THD is lower at the output of the converter with an
RMS 15.64 A 15.69 A 15.64 A intermediary DC/DC converter, especially for low values of
THD 0.65% 0.63% 0.62% irradiance. As it is expected from a PV system, current THD
lowers with the increase of solar irradiation [8].
In addition to the above simulation, a set of efficiency and
current THD measurements were obtained as a function of solar 4. Comparison of simulation models and
irradiation, for values ranging from 100 to a 1000 W/m2, while the measurement results
cell temperature was set to the nominal value 25Ԩ. Figures 7 and
8 present the results for the model with the PV array connected Phase current waveforms, along with their RMS and THD
directly to the DC link capacitance and the model with an values, were measured using a Chauvin-Arnoux CA 8332 power
intermediary BOOST converter, respectively. quality analyzer [9], during midday in the summer. Additionally,
irradiance and PV modules temperature were measured using
a Solar Survey 200R instrument [10]. Results of 1005 W/m2 and
53Ԩ were obtained, respectively. The same values were used in
both simulation models as inputs to the PV array Simulink block.
Figures 9 – 11 present the results of measurement and simulations.
Tables 5 and 6 compare the current RMS and THD values
obtained from simulating both models and measurement.
Results show, that there is about 0.3 – 0.4 A difference in RMS
values between simulation and the physical object, which is an
error of about 3 – 4%. However, THD values differ strongly, as
results of about 3.7% were measured, while the simulations
resolved to about 0.9 – 1%.

Fig. 7. Charts of efficiency and current THD as functions of solar irradiation Fig. 9. Current waveforms obtained with the CA 8332 power quality analyzer
obtained from the model of the grid-connecting converter with the PV array
directly connected to the DC link capacitance
Measurement Automation Monitoring, Jul. 2017, no. 03, vol. 63, ISSN 2450-2855  255

directly to the DC link was also shown to have better dynamic

Fig. 10. Phase currents waveforms obtained by simulating the model of the grid-
connecting converter with the PV array directly connected to the DC link
capacitance with irradiation and cell temperature set as measured (1005
W/m2 and 53°C)
properties, as it reached steady-state over twice as quickly.
The simulation results were compared to a 9.6 kW rooftop PV
plant built in Novi Sad, Serbia, by connecting a corresponding PV
array model at the input of the converter models. It was shown
through simulation, that the created models produce results similar
to the physical system in terms of current RMS values. THD
values, however, differ strongly, possibly due to the grid being
modelled as infinite, thus neglecting the effect of loads and other
sources connected to the grid.
In summary, although simplified, the created models represent
a physical system on a good level and may be used to predict
possible power produced and transferred to the grid. Their main
downside is predicting the impact of the PV array and the
converter on THD in the grid, as one of the simplifications is
modelling the grid as infinite.

6. References
[1] Ropp M., Gonzalez S.: Development of a MATLAB/Simulink Model
of a Single-Phase Grid-Connected Photovoltaic System, IEEE
Transactions on Energy Conversion, vol. 24, no. 1, pp. 195-202,
March 2009.
[2] Abdalrahman A., Zekry A., Alshazly A.: Simulation and
Implementation of Grid-connected Inverters, International Journal of
Computer Applications, vol. 60, no. 4, pp. 41–49, Dec. 2012.
[3] Katic V., Corba Z., Milicevic D., Dumnic B., Popadic B.: Realization
and Operation of Roof-top Photovoltaic Power Plant at the Faculty of
Technical Sciences in Novi Sad, Tehnika, vol. 70, no. 4, pp. 655–662,
2015.
[4] JKM-245P-60, 220-245 Watt Poly Crystaline Module Datasheet,
Fig. 11. Phase currents waveforms obtained by simulating the model of the grid- Jinko Solar.
connecting converter with an intermediary BOOST converter between the [5] Sunny Tripower 8000TL Inverter Datasheet, SMA Solar Technology
PV array and the DC link with irradiation and cell temperature set as AG.
measured (1005 W/m2 and 53°C)
[6] Femia N., Petrone G., Spagnuolo G., Vitelli M.: Optimization of
Perturb and observe maximum power point tracking method, IEEE
Tab. 5. Comparison of phase currents RMS and THD values obtained through
Transactions on Power Electronics, vol. 20, no. 4, pp. 963–973, Jul.
simulating the model with the PV array connected to directly to the DC link 2005.
and measurement [7] Handley P. G., Boys J. T.: Space vector modulation: an engineering
review, 1990 Fourth International Conference on Power Electronics
Measured Simulated and Variable-Speed Drives (Conf. Publ. No. 324), London, 1990, pp.
Ia Ib Ic Ia Ib Ic 87-91.
RMS 10.7 A 10.7 A 10.7 A 11.1 A 11.1 A 11.1 A [8] Patsalides M., Evagorou D., Makrides G., Achillides Z., Georghiou
THD 3.7% 3.6% 3.6% 0.96% 0.90% 0.92%
G.E., Stavrou A., Efthymiou V., Zinsser B., Schmitt W., Werner J.H.:
The effect of solar irradiance on the power quality behaviour of grid
connected photovoltaic systems, Renewable Energies & Power
Quality Journal, vol. 1, no. 5, pp. 323-330, Mar. 2007.
Tab. 6. Comparison of phase currents RMS and THD values obtained through
simulating the model with an intermediary BOOST converter and [9] Three Phase Power Quality Analyser C.A 8332, C.A 8334, User's
measurement manual, Chauvin-Arnoux.
[10] Solar Survey 100/200R Datasheet, Seaward.
Measured Simulated _____________________________________________________
Ia Ib Ic Ia Ib Ic Received: 05.04.2017 Paper reviewed Accepted: 02.06.2017
RMS 10.7 A 10.7 A 10.7 A 10.3 A 10.4 A 10.3 A
THD 3.7% 3.6% 3.6% 1.02% 0.96% 0.99%
Piotr ZIMOCH, MSc, eng.

The author works in the Department of Power

5. Conclusions
Simulation models of two possible topologies of SVM based
current controlled three-phase converters were shown and
compared. While both topologies are efficient and viable to
implement in physical form, there are differences between them.
A converter with the PV array directly connected to the DC link
was shown to be more efficient at all levels of solar irradiance.
However, the grid current THD values are lower when using
a topology with an intermediary DC/DC converter, especially at
low irradiance levels. A converter with the PV array connected
Electronics, Electrical Drives and Robotics at the
Silesian University of Technology in Gliwice, Poland.
He received the M Sc degree in electrical engineering at
the Silesian University of Technology in 2016. The
main areas of research interest are power electronics in
photovoltaic applications, particularly resonant DC/DC
converters and digital control of power electronic
converters.
e-mail: Piotr.Zimoch@polsl.pl