Trends of power electronics on
renewable energy systems
P. Gazis1, G. Α. Vokas2, St. Papathanasiou3,
1
Electronic Engineer, T.E.I. Piraeus, Tel: +30 6934120442, E-mail: fi_epe@hotmail.com
2
3
Ass. Prof. T.E.I. Piraeus, Dept. of Electronics, Aigaleo - Athens, Greece
Tel: +30 210 5381354, E-mail: gvokas@teipir.gr
Ass. Prof., NTUA, Dept. of Electrical & Computer Engineering - Athens, Greece
Tel: +30 210 7723658, E-mail: st@power.ece.ntua.gr
Abstract
The paper focuses on the power electronics used in renewable energy
systems and especially in wind and photovoltaic (PV) applications. During the last years
there was a broad development in the field of power electronics which led to more
efficient systems and reduction of the cost per installed kW. The inverters became more
efficient and reached efficiencies in excess of 98%, while commercial solar modules
reached almost 17% efficiency. Furthermore, the wind turbines use inverters of
improved efficiency, reliability and regulation capabilities. In this paper the recent trends
of power electronics topologies used in such systems are presented.
1. Introduction – Photovoltaic and wind energy systems
The Kyoto agreement renewed interest in renewable energy systems worldwide.
Many renewable energy technologies today are well developed, reliable, and cost
competitive with conventional generators. The cost of renewable energy
technologies is on a falling trend and is expected to fall further as demand and
production increases [1], [4]. There are many renewable energy sources such as
biomass, solar, wind, mini-hydro, and tidal power [6]. Power electronics find
applications in most RES technologies, solar and wind energy systems being the
most important applications.
During the last years, there is a constant effort to improve each part of a
photovoltaic (PV) and wind turbine (WT) application. The efficiency of commercial
PV modules now exceeds 17%, inverters have reached almost 99% European
efficiency and there are new topologies found which make WTs more efficient and
flexible in their operation. Due to the increased demand, each manufacturer is trying
to find new concepts in order to achieve better system yield, which results in
increased economic returns for the investor. Most of the systems used in such
applications produce DC current, so inverters are required to convert this power to
AC, which is needed in most applications and definitely for grid connection. There
are two types of PV systems: stand alone and grid connected. The first is used on
remote locations, where the utility grid is not present. The grid connected systems
inject power and energy directly to the utility grid. These systems have different
structure and the inverters which are used have different methods to synchronize
and produce clean AC power.
2. Power Electronics for PV applications
The PV modules and the power electronics that convert the produced electric power
by the PV modules are the basic parts of a PV installation. The PV modules
comprise several solar cells which convert the energy of the sunlight directly into
electricity, and are connected in a proper way (typically in series), to provide desired
levels of DC current and voltage. They produce electricity due to a quantummechanic process known as the “photovoltaic effect” [1]. A presentation of this
conversion is shown on Figure 1a. There are many semiconductor materials
suitable for solar cells manufacturing. The most commonly used are monocrystalline
Si cells, polycrystalline Si cells and amorphous Si cells, although several other thin
film technologies exist in the market. The efficiency of monocrystalline Si cells is
almost 17%, for polycrystalline cells reaches almost 15%, while an efficiency of 10%
is achieved in the case of amorphous Si PV cells. All PV modules have a typical
current-voltage characteristic curve, used to make all necessary calculations, as
shown on Figure 1b [4].
Figure 1a: Photovoltaic effect
Figure 1b: PV module curve
Off-grid PV systems are used in cases, where the grid is not present and the use of
batteries to store energy is required, in order to cover the demand during the night
or whenever energy is needed. Blocking diodes are used to prevent the batteries to
discharge on the modules during the night, while they also protect the batteries from
short circuit. If more than one string is used, they also provide over-current
protection of the strings in case of short circuits. Charge regulators control the
charging of the batteries [1], [6]. In off-grid systems, there is the need to use dc
voltage and current with stable characteristics, independent from irradiance
fluctuations. Therefore, a DC – DC conversion topology is used. Switch mode DC –
DC converters [1] are used to match the dc output of a PV generator to a variable
load. Three different topologies are mostly used; step down converters, step up
converters and a combination of these two. In Figure 2 simplified diagrams of these
three topologies are presented.
Figure 2: Simplified diagrams of DC/DC converter topologies.
In order to maximize the performance of the string, in most charge regulators a
maximum power point tracker (MPPT) controller is used. The MPPT applies
heuristic algorithms to track the array voltage which results in maximum power,
given a solar irradiance level. The efficiency of modern MPPTs is between 92-97%,
getting a typical 20-45% power gain in winter and a 10-15% in summer. Actual gain
can vary widely depending on temperature, battery state of charge, and other
factors.
The MPPT is often performed via a high frequency DC to DC converter. Its input is
the output of the solar panels strings. It converts the DC input to high frequency AC,
and then back to a different DC voltage and current in order to match the panel
voltage to that of the batteries. MPPTs operate at very high (audio range)
frequencies, usually in the 20-80 kHz range. The advantage of high frequency
circuits is that they can be designed with high efficiency and small volume
transformers and other components. The design of high frequency circuits can be a
very difficult task because of EMI/EMC considerations (e.g. problems with circuit
parts that act as antennas, causing radio and TV interference). Noise isolation and
suppression issues become very important [6], [4].
Inverters convert DC to AC. In off-grid systems, stand alone self-commutated
inverters producing AC current without synchronisation with a reference signal are
used. These inverters have the responsibility to produce AC voltage and current
characteristics (sinusoidal 230V/50Hz) same as those of a typical grid in order to
supply off-grid loads. Otherwise, the inverter is not suitable for most electronic
devices [1]. Several different semiconductor devices such as MOSFETs and IGBTs
are used. The first are used in units up to 3KW, because they have the advantage
of low switching losses at higher frequencies. At higher voltages and powers IGBTs
are used [2]. These inverters can be single phase or three phases. A common
switching technique in order to eliminate higher frequencies is the SPWM method. A
general layout of a single phase inverter with half bridge and full bridge topology is
shown in figure 3a, b. In the half bridge topology, the two switches S1 and S2, the
capacitors C1 and C2 are connected in series with the dc source (batteries). The
center point between the two capacitors is at mid-potential. The voltage across each
capacitor is VDC / 2. The switches S1 and S2 switch on and off periodically to
produce the ac voltage. A filter (Lf and Cf) is used to reduce high switching
frequency components and to produce sinusoidal output from the inverter.
Figure 3a: Single phase half bridge inverter
Figure 3: Single phase full bridge inverter
The output of the inverter is connected to the load through a transformer. The full
bridge inverter has a similar function, but the output voltage is higher than the half
bridge inverter. In figure 4 we can also see the layout of a three phase stand alone
inverter [3], [4].
Figure 4: Three phase inverter
In grid-connected applications the energy is provided directly to the grid and the
necessary parts are the PV modules and the inverters. This reduces the cost of the
system and it also reduces the necessary maintenance, as the batteries are the
most maintenance-demanding components. The inverters for grid connected
applications may have different topology and operation than off-grid ones. They
have to produce excellent quality sine wave output, follow the frequency and
voltage of the grid and extract maximum power from the PV modules through the
MPPT. The inverter input scans the I–V curve of the string until the maximum power
point is found. [2], [1] The grid inverter always monitors the grid and the output
voltage and frequency must be controlled. The most common modulation is the
PWM modulation and operates at a range of 2 to 20 KHz [1]. Grid connected
inverters are classified as voltage source inverters (VSI) and current source
inverters (CSI). However, in PV applications VSI inverters are used. The layout of
the VSI topology is shown in Figure 5.
Figure 5: VSI inverter
During the years, the inverter topologies and the technology of power electronics
has improved. Until some years ago, the common practice was to use central
inverters for most PV applications. The PV modules were divided into series
connections (called strings), each generating a sufficiently high voltage to avoid
further amplification. Then all the strings were connected in parallel through string
diodes in order to reach high power levels. The use of central inverter has many
drawbacks such as MPPT power losses, losses from differentiations between the
modules and high voltage DC cables from the PV panels to the inverter. In the
beginning, line commutated inverters using thyristors were applied, characterized by
poor harmonic performance. Today, the most popular inverter topology is the string
inverter. One or two strings of crystalline modules are connected to each inverter
which has its own MPP tracker and the power losses are significantly lower [11].
The use of more than one MPP trackers in the power plant is necessary in the case
of different module orientation and shading. This maximizes each string’s I-V curve
power output [10]. From tests that have been performed on string inverters, the
development through the years is obvious. According to [H. Haeberlin, Berner
Fachhochshule], the inverter efficiency in 1988 was in the order of 85.5 – 90%, in
the mid 90’s was increased to 90 – 92% and nowadays it has reached 98% [12].
The most popular string inverter is the transformerless one, because the
transformers that operated at grid frequencies are bulky, expensive and cause
losses. Furthermore, the transformers impose limitations in the control of grid
current by the inverter. Especially at low load, the reactive power for the
magnetization of the transformer leads to a lower power factor [10]. Another
important factor that affects the system design optimization is the maximum input
voltage (Vmax-in) of the inverters as well as their input voltage (Vin) bandwidth. These
two characteristics (Vmax-in and Vin) are getting wider and still rise, a fact that allows
the designer to perform more efficient and flexible combinations in order to achieve
the desired power. The Vmax-in of the string inverters kept rising from 600V up to
900V in 2009, while the in 2010 inverters with Vmax-in=1000V allowing even bigger
strings are already in the market. The higher the Vmax-in is, the less strings of more
modules are used, so the losses are further decreased as less cables are used [11].
The IGBTs and MOSFETs with high pulsing frequencies provide improved power
quality in compliance with the regulations of the utility grid. The high frequency used
has led to the use of high frequency transformers with lower weight. This fact
reduced the total weight of the inverters significantly (up to 20%). The today’s string
inverters vary from 22 - 65 kg. The lower the weight is, the easier the installation
and the lower the transportation costs are.
According to the power categorization, the string inverters are now available in the
range of 2 to 30 kWp. Until 2008, string inverters were not produced at more than
the 5kWp. Single phase ones are used in PV stations of even 2MWp. However, a
more recent trend is the development of the three phase string inverters, at almost
the same power range. In 2010 the first 3-phase inverters became available in the
market providing easier design and electrical connections , as well as a completely
symmetric power output, an important factor for the utility grid operator.
The multistring inverter is a development of the string inverter. A combination of
strings are connected to separate DC/DC converters and then to a common DC/AC
converter. This is beneficial in comparison to the central inverter because each
string is controlled individually. This results to higher efficiency and flexibility of the
system [11], [9].
Central inverters are used in larger scale applications, offering Operation and
Maintenance contracts for the plant owners. The operation availability of such
inverters is warranted up to 99% throughout a complete year of operation.
Until 2008 the power range of the central inverters was from 100kW but not more
than 500kW. However, the Inverter manufacturers developed larger inverters mostly
using scalable techniques up to 1,25MWp.
The efficiency of Central Inverters has climbed from 92% since the 1990’s to 98.8%
in 2010. Their main advantage is their high reliability providing maximum
operational life. Recently, the trend is to offer Central Station Inverters (CSI), which
include the house, the transformer, the medium voltage switchgear, the monitoring
system, the cooling concept and the wiring and they arrive on the installation preassembled, which minimizes all the required tasks and connections.
Such inverters are used mostly in PV parks higher than 2MWp, however lower size
installations are also common.
3. Power Electronics for WT applications
There are two types of wind turbines (WT), the horizontal axis and the vertical axis.
They are both shown on Figure 6. However, the horizontal axis WT is by far the
most popular design [4].
Figure 6a: Horizontal axis WT
Figure 6b: Vertical axis WT
A large number of designs are available, ranging from 50W up to 7MW size. The
number of blades can vary but the most commonly seen are with 2 or 3 blades.
There are three types of wind power systems, the stand alone, the hybrid and the
grid connected.
The stand alone ones are mostly used for household applications and require
batteries in order to store the produced energy and an inverter to convert to AC
current. This system requires the use of a charge regulator which will feed the
power from the wind generator to the battery bank in a controlled manner. The most
common turbine which is used in such applications is the permanent magnet
generator and the charging control is made through controlled rectifiers. The charge
regulator should be programmed to limit the current into the batteries, to reduce
current when the batteries are charged and to maintain a trickle charge during full of
charge periods [1], [4]. The hybrid systems could also include other renewable
sources such as PV systems or even a diesel generator and feed the energy to a
household or even to the grid.
The grid connected WT are connected to the utility grid either directly or through
power electronics, feeding the produced energy to the grid. On this type of WT all
manufacturers are trying to increase the size and efficiency of the machines. Many
studies have been made on the speed control part and on ways to reduce the cost
of the unit. There are several types of inverters which are used on wind turbine
installations, such as PWM – VSI converters and matrix converters. However the
matrix converter is not widely used. The back – to back PWM VSI is a bi-directional
power converter consisting of two PWM – VSI inverters. The topology of this
inverter is shown in Figure 7a.To achieve full control of the grid current, the DC link
voltage must be boosted to a level higher than the amplitude of the grid line voltage.
The power flow of the grid side converter is controlled in order to keep the DC – link
voltage constant, while the control of the generator is set to suit the magnetization
demand and the reference speed.
The matrix inverters can efficiently convert the three phase electrical output of the
WT to the requested electrical grid characteristics for a proper connection. They use
an array of controlled bidirectional switches to convert AC power from one
frequency to another. They produce a variable output voltage with unrestricted
frequency. The privilege compared to other topologies is that the matrix converters
do not have a dc link circuit and they do not use large energy storage elements. A
figure of the matrix inverter is shown in Figure 7b where the switching devices, the
input filter and the clamp circuit are presented.
Figure 7a: Back to back PWM - VSI converter
Figure 7b: Matrix converter
MOSFETs for low power and IGBTs for higher power enable the implementation of
bidirectional switches, which make the inverter topology very attractive for AC
power handling.
The input filter minimizes the high frequency components in the input currents and
reduces the disturbances of input power. The input filter can be designed with a
combination of inductors and capacitors with parallel damping resistors where on
the other hand the clamp circuit provides overvoltage protection and uses fast
recovery diodes. The general operation of the inverter is to convert the AC current
of the WT to the AC current that the utility grid demands. One example of this taken
from L.H.Hansen is the conversion of 300 Hz current coming from the WT to 60Hz
which is the grid frequency for American utility. A 7200 Hz switching frequency is
applied on the switches and the result of the conversion is shown on Figure 8.
There is a duty cycle factor that can be adjusted to regulate the ratio of output to
input voltage, up to a maximum value. Finally, the output is passed through a filter
to eliminate high frequency harmonics [8].
Figure 8: Matrix converter steady state simulations
The driving circuit of a typical IGBT is the same as MOSFET, and the linkage
capacitors between an IGBT’s terminals are rather low. In PWM type rectifiers,
when the switching frequency increases, the power loss becomes high during the
deactivation of the switching element and the commutation diode. This case limits
the usage of an IGBT with 50 kHz as the switching element. Thus, in high frequency
resonance-type inverters, it can be practically used at the frequency of 250 kHz. For
50 kHz, the power loss during conduction of an IGBT is approximately 6 watts
where for 250 kHz it is approximately 0.8 watts.
Apart from the matrix converters there are also other three suggested topologies,
namely the tandem converter, the multilevel converter and the resonant converter
(NCC). The advantage of the matrix converter is that it consists only of active
components on the power part where the others consist of passive components too.
The matrix converter can omit the transformer without requiring high voltage ratings
for the components. However, on the harmonic performance the best topology is
the multilevel converter which shows the best spectra on both the grid and
generator side. Finally, the NCC topology is the most efficient followed by the
multilevel and the tandem converter. Further information can be found on [7].
Apart from the above topologies, what really makes the difference between the
manufacturers is the control concept. Constant research and development is being
performed on this field in order to develop better modulation concepts and to
eliminate harmonics, so as to minimize the cost for the filters.
A common application of the static converters is the switching to the grid of WT
equipped with induction generators (soft starting). Direct connection of the WT to
the grid causes high inrush currents which are undesirable especially for weak
grids, also severe torque pulsations and damage to the gearbox. For this reason the
soft starter is used which regulates the applied stator voltages.
The commutating devices are two anti-parallel thyristors per phase. The relationship
between the firing angle (α) and the resulting amplification of the soft starter is nonlinear and depends additionally on the power factor of the connected element. In the
case of a resistive load, α may vary between 0 (full on) and 90 (full off) degrees, in
the case of a purely inductive load between 90, (full on) and 180 (full off) degrees.
For any power factor between 0 and 90 degrees, α will be somewhere between the
limits sketched in Figure 9b.
Usually, the turbine accelerates under pitch-control to the synchronous speed
through the wind power alone and only then is it switched onto the grid. AC
controllers have been also reported for the connection to the grid at zero speed and
subsequent fast acceleration to the operating speed, which is particularly useful in
stall controlled WT [13].
When the generator is connected to the grid a contactor bypasses the soft starter to
reduce losses. This is also shown in figure 9a (Kbyp). It is possible to utilize the
controller in normal operation in order to reduce the static voltage and resulting to
reduce magnetizing losses when operating under light wind, at the expense of high
harmonic distortion [7].
A recent trend in WT technology is the use of multi-pole low speed generators
which permit the direct coupling of the turbine rotor to the electrical generator. This
eliminates the gearbox, the interconnection axes and the coupling and results to
price and weight reduction and improved reliability [13].
Figure 9a: Soft starter
Figure 9b: Control characteristic for a fully
controlled soft starter.
4. Conclusion
In this paper the main trends of the power electronics used in PV and WT
applications are presented. Due to the high demand for renewable energy sources
applications, there is a continuing research for improving the total efficiency of
these applications and by improving each electronic part included. As far as the PV
systems are concerned the inverter’s efficiency is continuously improving and ways
to minimize the weight of the devices are tested so as to decrease transportation
costs and ease the installation. Moreover, the power and voltage range of the string
and central inverters is increased, so that more efficient and cheaper PV
installations can be realized using a relatively low number of inverters. The power
electronics for WT systems are subject to extensive R&D, especially about more
efficient control concepts and even more efficient converters.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
Muhammad H. Rashid, “Power Electronics Handbook,” Ph.D., Fellow IEE, Fellow IEEE, Academic
press, Copyright 2001.
nd
Mohan, Undeland, Robbins, “Power Electronics” 2 edtion 1995, A.Tziolas A.E.
Manias Stefanos, “Power Electronics”, Professor of National University of Athens, Simeon press,
Coryright 2000.
Papadopoulos Michalis, “Energy production from renewable energy applications”, Professor of
National University of Athens, Copyright 1997 by National University of Athens.
Simone Buso and Paolo Mattavelli, “Digital Control in Power Electronics”, Department of Information
Engineering, University of Padova, Italy, Morgan and Claypool publishers, Copyright 2006.
Fragkiadakis I., “Photovoltaic systems”, Professor of T.E.I. of Crete, Zito press, Copyright 2004.
L. H. Hansen, L. Helle, F. Blaabjerg, E. Ritchie, S. MunkNielsen, H. Bindner, P. Sørensen and B.
Bak-Jensen, “Conceptual survey of generators and power electronics for wind turbines”, Riso
National Laboratory, Roskilde, Denmark, December 2001
Pradeep Bhatta, Michael A. Paluszek and Joseph B. Mueller, “Individual blade pitch and camber
control for vertical axis wind turbines”.
J. M. A. Myrzik, and M. Calais, Member, IEEE, “String and module integrated inverters for single
phase grid connected photovoltaic systems – A review”, Paper accepted for presentation at 2003
IEEE Bologna PowerTech Conference, June 23-26, Bologna, Italy
[10] Fritz Schimpf, Lars E. Norum, “Grid connected converters for photovoltaic, state of the art, ideas for
improvement of transformerless inverters”, NORPIE/2008, Nordic Workshop on Power and
Industrial Electronics, June 9-11, 2008
[11] Soeren Baekhoej Kjaer, Member, IEEE, John K. Pedersen, Senior Member, IEEE, and Frede
Blaabjerg, Fellow, IEEE, “A review of Single – phase grid connected inverters for photovoltaic
systems”, IEEE transactions on industry applications, vol 41, No. 5, September / October 2005.
[12] H. Haeberlin, Berner Fachhochshule, “Evolution of inverters for grid connected PV – Systems from
th
1989 to 2000”, 17 European Photovoltaic Solar Energy conference, Munich, Germany, Oct 22 –
Oct 26, 2001.
[13] S.A.Papathanasiou, G.A.Vokas, M.P.Papadopoulos, "Use of Power Electronic Converters in Wind
and Photovoltaic Generators", IEEE International Symposium on Industrial Electronics ISIE'95, Vol.1,
p.p.254-259, Athens, Greece, July 10-14, 1995.