Download Power Converters and Control of Renewable Energy Systems

1
Power Converters and Control of Renewable Energy Systems
Frede Blaabjerg, Remus Teodorescu, Zhe Chen
Aalborg University, Institute of Energy Technology,
Denmark
[email protected], [email protected], [email protected]
Marco Liserre
Politecnico di Bari, CEMD research group
Italy
[email protected]
Abstract — The global electrical energy consumption is steadily
rising and consequently there is a demand to increase the
power generation capacity. A significant percentage of the
required capacity increase can be based on renewable energy
sources. Wind turbine technology, as the most cost effective
renewable energy conversion system, will play an important
part in our future energy supply. But other sources like
microturbines, photovoltaics and fuel cell systems may also be
serious contributors to the power supply. Characteristically,
power electronics will be an efficient and important interface
to the grid and this paper will first briefly discuss three
different alternative/renewable energy sources. Next, various
configurations of the wind turbine technology are presented, as
this technology seems to be most developed and cost-effective.
Finally, the developments and requirements from the grid are
discussed.
Finally, a general discussion about interface issues of
renewable energy sources is done.
I. INTRODUCTION
The energy comsumption is steadily increasing and the
deregulation of electricity has caused that the amount of
installed production capacity of classical high power stations
cannot follow the demand. A method to fill out the gap is to
make incentives to invest in alternative energy sources like
wind turbines, photovoltaic systems, microturbines and also
fuel cell systems. The wind turbine technology is one of the
most promising alternative energy technology [1]-[3]. The
modern development started in the 1980’s with sites of a
few tens of kW to Multi-MW range wind turbines today.
E.g. Denmark has a high penetration (> 20%) of wind
energy in major areas of the country and in 2003 15% of the
whole electrical energy consumption was covered by wind
energy. A higher penetration level will even be seen in the
near future. The technology used in the early developed
wind turbines was based on a squirrel-cage induction
generator connected directly to the grid. That almost directly
transfers the wind power pulsations to the electrical grid.
Furthermore, there was no fast control of the active and
reactive power, which typically are the key parameters to
control the frequency and the voltage of the grid. As the
power range of the wind turbines increases those parameters
become more and more important. The power electronics is
the key-technology to change the basic characteristic of the
wind turbine from being an energy source to be an active
power source. Such possibilities are also used to interface
other renewable energy sources [4]-[8].
This paper will first explain the basic principles of wind
power conversion, fuel cells and photovoltaics. Next, the
trend in power electronics is outlined. Diffe-rent wind
turbine configurations are reviewed, as they are the most
promising alternative energy technologies today.
II. RENEWABLE ENERGY SOURCES
Three different renewable energy sources are briefly
described. They are wind power, fuel cell and photovoltaic.
A. Wind power conversion
The function of a wind turbine is to convert the linear
motion of the wind into rotational energy that can be used to
drive a generator, as illustrated in Fig. 1. Wind turbines
capture the power from the wind by means of
aerodynamically designed blades and convert it into rotating
mechanical power. At present, the most popular wind
turbine is the Horizontal Axis Wind Turbine (HAWTs)
where the number of blades is typically three.
Wind turbine blades use airfoils to develop mechanical
power. The cross-sections of wind turbine blades have the
shape of airfoils as the one shown in Fig. 2.
Airflow over an airfoil produces a distribution of forces
along the airfoil surface. The resultant of all these pressure
and friction forces is usually resolved into two forces and a
moment, lift force, drag force and pitching moment, as
shown in Fig. 2.
The aerodynamic power, P, of a wind turbine is given by:
1
2 3
P = ρπR v C p
(1)
2
where ρ is the air density, R is the turbine radius, v is the
wind speed and CP is the turbine power coefficient which
represents the power conversion efficiency of a wind turbine.
CP is a function of the tip-speed ratio (λ), as well as the
blade pitch angle (β) in a pitch controlled wind turbine. λ is
defined as the ratio of the tip speed of the turbine blades to
wind speed, and given by:
λ=
R⋅Ω
(2)
v
where Ω is the rotational speed of the wind turbine.
The Betz limit, CP,max (theoretical) =16/27, is the maximum
theoretically possible rotor power coefficient. In practice
three effects lead to a decrease in the maximum achievable
power coefficient [1]:
•
•
•
Rotation of the wake behind the rotor
Finite number of blades and associated tip losses
Non-zero aerodynamic drag
2
Electrical Power
Wind power
Gearbox (optional)
Rotor
Generator
Power converter
(optional)
Supply grid
Consumer
Power conversion &
Power conversion &
Power transmission
Power conversion
power control Fig. 1. Conversion from wind power to electrical power in apower
control [11].
wind turbine
Power transmission
Fig. 1. Conversion from wind power to electrical power in a wind turbine [11].
A typical CP-λ curve for a fixed pitch angle β is shown in
Fig. 3. It can be seen that there is a practical maximum
power coefficient, CP,max. Normally, a variable speed wind
turbine follows the CP,max to capture the maximum power up
to the rated speed by varying the rotor speed to keep the
system at the optimum tip-speed ratio, λopt.
As the blade tip-speed typically should be lower than half
the speed of sound the rotational speed will decrease as the
radius of the blade increases. For MW wind turbines the
rotational speed will be 10-15 rpm. A common way to
convert the low-speed, high-torque power to electrical
power is to use a gear-box and a normal speed generator as
illustrated in Fig. 1. The gear-box is optional as multipole
generator systems are alternative solutions.
Lift force
Pitching moment
Drag force
β
Trailing edge
φ
α
Leading edge
Angle of attack:α
Pitch angle: β
wind
Fig. 2. A simple airfoil used in wind turbines.
Fig. 3. Typical Cp-λ curve for a wind turbine for a fixed angle β.
The development in the wind turbine systems has been
steady for the last 25 years and four to five generations of
wind turbines exist. It is now a proven technology.
It is important to be able to control and limit the power at
higher wind speeds, as the power in the wind is a cube of the
wind speed.
Wind turbines have to be cut out at a high wind speed to
avoid damage. A turbine could be designed in such a way
that it converts as much power as possible in all wind
speeds, but then it would have to be too heavy. The high
costs of such a design would not be compensated by the
extra production at high winds, since such winds are rare.
Therefore, turbines usually reach maximum power at a
much lower wind speed, the rated wind speed (9-12 m/s).
The power limitation may be done by one of the
aerodynamic mechanisms: stall control (the blade position is
fixed but stall of the wind appears along the blade at higher
wind speed), active stall (the blade angle is adjusted in order
to create stall along the blades) or pitch control (the blades
are turned out of the wind at higher wind speed).
B. Fuel Cell power conversion
The fuel cell is a chemical device, which produces
electricity directly without any intermediate stage and has
recently received much attention [7]. The most significant
advantages are low emission of green house gases and high
power density. For example, a zero emission can be
achieved with hydrogen fuel. The emission consists of only
harmless gases and water. The noise emission is also low.
The energy density of a typical fuel cell is 200 Wh/l, which
is nearly ten times of a battery. Various fuel cells are
available for industrial use or currently being investigated
for use in industry, including
• Proton Exchange Membrane
• Solid Oxide
• Molten Carbonate
• Phosphoric Acid
• Aqueous Alkaline
The efficiency of the fuel cell is quite high (40%-60%). Also
the waste heat generated by the fuel cell can usually be used
for cogeneration such as steam, air-conditioning, hot air and
heating, then the overall efficiency of such a system can be
as high as 80%.
3
iPV
iSC
id
uPV
(a)
Fig. 4. V-I characteristics of a fuel cell [12].
A typical curve of the cell electrical voltage against current
density is shown in Fig. 4. It can be seen that there exists a
region where the voltage drop is linearly related with the
current density due to the Ohmic contact.
Beyond this region the change in output voltage varies
rapidly. At very high current density, the voltage drops
significantly because of the gas exchange efficiency. At low
current level, the Ohmic loss becomes less significant, the
increase in output voltage is mainly due to the activity of the
chemicals. Although the voltage of a fuel cell is usually
small, with a theoretical maximum being around 1.2 V, fuel
cells may be connected in parallel and/or in series to obtain
the required power and voltage.
The power conditioning systems, including inverters and
DC/DC converters, are often required in order to supply
normal customer load demand or send electricity into the
grid.
C. The photovoltaic cell
Photovoltaic (PV) power supplied to the utility grid is
gaining more and more visibility due to many national
incentives [7]. With a continuous reduction in system cost
(PV modules, DC/AC inverters, cables, fittings and manpower), the PV technology has the potential to become one
of the main renewable energy sources for the future
electricity supply.
The PV cell is an all-electrical device, which produces
electrical power when exposed to sunlight and connected to
a suitable load. Without any moving parts inside the PV
module, the tear-and-wear is very low. Thus, lifetimes of
more than 25 years for modules are easily reached.
However, the power generation capability may be reduced to
75% ~ 80% of nominal value due to ageing. A typical PV
module is made up around 36 or 72 cells connected in series,
encapsulated in a structure made of e.g. aluminum and
tedlar. An electrical model of the PV cell is depicted in Fig.
5.
iSC IPV
(uMPP, iMPP)
pMPP
PPV
uOC
UPV
(b)
Fig. 5. Model and characteristics of a PhotoVoltaic (PV) cell.
(a) Electrical model with current and voltages defined.
(b) Electrical characteristic of the PV cell, exposed to a given amount
of sunlight at a given temperature.
Several types of proven PV technologies exist, where the
crystalline (PV module light-to-electricity efficiency: η =
10% - 15%) and multi-crystalline (η = 9% - 12%) silicon
cells are based on standard microelectronic manufacturing
processes. Other types are: thin-film amorphous silicon (η =
10%), thin-film copper indium diselenide (η = 12%), and
thin-film cadmium telluride (η = 9%). Novel technologies
such as the thin-layer silicon (η = 8%) and the dye-sensitised
nano-structured materials (η = 9%) are in their early
development. The reason to maintain a high level of
research and development within these technologies is to
decrease the cost of the PV-cells, perhaps on the expense of
a somewhat lower efficiency. This is mainly due to the fact
that cells based on today’s microelectronic processes are
rather costly, when compared to other renewable energy
sources.
The series connection of the cells benefits from a high
voltage (around 25 V ~ 45 V) across the terminals, but the
weakest cell determines the current seen at the terminals.
4
6
2
1000 W/m
This power electronic system can be used with many loads
and generators.
o
15 C
40 oC
Cel
l 4
cur
ren
t 2
[A]
2
600 W/m
III. SINGLE-PHASE PV-INVERTERS
75 oC
2
200 W/m
0
0
0.1
0.2
0.3
0.4
Cell voltage [V]
0.5
0.6
0.7
The general block diagram for single-phase grid connected
photovoltaic systems is presented in Fig. 1a. It consists of
PV array, PV inverter, controller and grid.
(a)
2.5
PV
Array
15 oC
2
Cel
l
1.5
po
we
1
r
[W]
0.5
40 oC
PV Inverter
+ LP Filter
Grid
o
75 C
0
0
Control
0.1
0.2
0.3
0.4
Cell voltage [V]
0.5
0.6
0.7
a)
(b)
Fig. 6. Characteristics of a PV cell. Model based on the British Petroleum
BP5170 crystalline silicon PV module. Power at standard test condition
(1000 W/m2 irradiation, and a cell temperature of 25 °C): 170 W @ 36.0 V.
Legend: solid at 15 oC, dotted at 40 oC, and dashdot at 75 oC.
The series connection of the cells benefits from a high
voltage (around 25 V ~ 45 V) across the terminals, but the
weakest cell determines the current seen at the terminals.
This causes reduction in the available power, which to some
extent can be mitigated by the use of bypass diodes, in
parallel with the cells. The parallel connection of the cells
solves the ‘weakest-link’ problem, but the voltage seen at
the terminals is rather low. Typical curves of a PV cell
current-voltage and power-voltage characteristics are plotted
in Fig. 6a and Fig. 6b respectively, with insolation and cell
tempera-ture as parameters. The graph reveals that the
captured power is determined by the loading conditions
(terminal voltage and current). This leads to a few basic
requirements for the power electronics used to interface the
PV module(s) to the utility grid.’
The job for the power electronics in renewable energy
systems is to convert the energy from one stage into another
stage to the grid (alternative voltage) with the highest
possible efficiency, the lowest cost and to keep a superior
performance. The basic interfacing is shown in Fig. 10.
Power flow
Loads
Appliance
Industry
Communication
Power converter
Load /
generator
2-3
2-3
Generators
Wind
Photo-voltaic
Fuel cell
Other sources
Control
Reference (local/centralized)
Fig. 10. Power electronic system with the grid, load/source, power
converter and control.
b)
c)
d)
Fig 1. General schema for single-phase grid connected photovoltaic
systems. a) Block diagram; b) Central inverter; c) String inverter; d)
Module integrated inverter
The PV array can be a single panel, a string of PV panels
or a multitude of parallel strings of PV panels. Centralized
or decentralized PV systems can be used as depicted in the
Fig. 1,b,c,d .
Central inverters
In this topology the PV plant (> 10 kW) is arranged in
many parallel strings that are connected to a single central
inverter on the DC-side (Fig. 1b). These inverters are
characterized by high efficiency and the lowest specific cost.
However, the energy yield of the PV plant decreases due to
module mismatching and partial shading conditions. Also,
the reliability of the plant is limited due to the dependence
of power generation on a single component: the failure of
the central inverter results in the whole PV plant being out
of operation.
String inverter
Similar to the central inverter, the PV plant in this concept is
divided into several parallel strings. Each of the PV strings
is assigned to a designated inverter, the so-called "string
inverter"(Fig. 1c). String inverters have the capability of
separate MPP tracking of each PV string. This increases the
5
energy yield via the reduction of mismatching and partial
shading losses. These superior technical characteristics lead
to a reduction in the system cost, increase the energy yield
and enhance the supply reliability. String inverters have
evolved as a standard in PV system technology for grid
connected PV plants.
An evolution of the string technology applicable for higher
power levels is the multi-string inverter [1]. It allows the
connection of several strings with separate MPP tracking
systems (via DC/DC converter) to a common DC/AC
inverter. Accordingly, a compact and cost-effective solution
which combines the advantages of central and string
technologies is achieved. This multi-string topology allows
the integration of PV strings of different technologies and of
various orientations (south, west and east). These
characteristics allow time shifted solar power which
optimizes the operation efficiencies of each string
separately. The application area of the multi-string inverter
covers PV plants of 3-10 kW.
Module integrated inverter
This system uses one inverter is used for each module (Fig
1d). This topology optimizes the adaptability of the inverter
to the PV characteristics, since each module has its own
MPP tracker. Although the module integrated inverter
optimizes the energy yield, it has a lower efficiency than the
string inverter. Module integrated inverters are characterized
by more extended AC-side cabling, since each module of
the PV plant has to be connected to the available AC grid
(e.g. 230 V/ 50 Hz). Also, the maintenance processes are
quite complicated, especially for facade-integrated PV
systems. This concept can be implemented for PV plants of
about 50- 400 W peak.
PV inverter
The PV inverter technology has evolved quite a lot during
the last years towards maturity [2]. Still there are different
power configurations possible as shown in the Fig. 2.
on the LF side
with DC-DC
converter
PV
Inverters
without DC-DC
converter
with isolation
on the HF side
without isolation
with isolation
without isolation
Fig. 2: Power configuration for PV inverters
The question of having or not a dc-dc converter is first of
all related to the PV string configuration. Having more
panels in series and lower grid voltage, like in US and
Japan, it is possible to avoid the boost function. Thus a
single stage PV inverter can be used leading to higher
efficiency.
The issue of isolation is mainly related to safety standards
and is for the moment only required in US. The drawback of
having so many panels in series is that MPPT is harder to
achieve especially during partial shading, as demonstrated in
[3]. In the following, the different PV inverters power
configurations are described in more detail.
PV inverters with DC-DC converter with isolation
The isolation is typically acquired using a transformer that
can be placed on either the grid frequency side (LF) as
shown in the Fig. 3a or on the high-frequency (HF) side in
the dc-dc converter as shown in the Fig. 3b. The HF
transformer leads to more compact solutions but high care
should be taken in the transformer design in order to keep
the losses low.
DC
PV
Array
DC
DC
Grid
AC
(a)
PV
Array
DC
AC
AC
DC
DC
Grid
AC
(b)
Fig. 3. PV inverter system with DC-DC converter and isolation transformer
a) on the LF side b) n the HF side
In the Fig.4 is presented a PV inverter with HF
transformer using an isolated push-pull boost converter [4]
Fig. 4. PV inverter with HF transformer in the dc-dc converter
Also, the dc-ac inverter in this solution is a low cost
inverter switched at the line frequency. The new solutions
on the market are using PWM dc-ac inverters with IGBT
switched typically at 10-20 kHz leading to better power
quality performances.
Other solutions for high frequency dc-dc converters with
isolations includes: full-bridge isolated converter, singleinductor push-pull converter (SIC) and double-inductor
converter (DIC) as depicted in Fig. 5 [5]
6
L
PV
Array
DC
DC
DC
Grid
AC
(a)
(b)
i
Fig. 5 dc-dc converter topologies with isolation. a) full-bridge; b) singleinductor pushh-pull; c) double-inductor push-pull
In order to keep the magnetic components compact high
switching frequencies in the range of 20 – 100 kHz are
employed.
The full-bridge converter is usually utilized at power levels
above 750W. The advantages of this topology are: good
transformer utilization – bipolar magnetization of the core,
good performance with current programmed control –
reduced DC magnetization of transformer. The main
disadvantages in comparison with push-pull topology are the
higher active part count and the higher transformer ratio
needed for boosting the dc voltage to the grid level.
The single inductor push-pull converter can provide
boosting function on both the boosting inductor and
transformer, reducing thus the transformer ratio. Thus higher
efficiency can be achieved together with smoother input
current. On the negative side it can be mentioned that higher
voltage blocking switches are required and the transformer
with tap point puts some construction and reliability
problems.
Those shortcomings can be alleviated using the double
inductor push-pull converter (DIC) where the boost inductor
has been split in two. Actually this topology is equivalent
with two interleaved boost converters leading to lower
ripple in the input current. The transformer construction is
more simple not requiring tap point. The single disadvantage
of this topology remains the need for an extra inductor.
PV inverters with DC-DC converter without isolation
In some countries as the grid-isolation is not mandatory,
more simplified PV inverter design can be used, as shown in
Fig. 6
Fig 6. PV inverter system with DC-DC converter without isolation
transformer
a) General diagram
b) Practical example with boost converter and full-bridge inverter [4]
In Fig. 6b a practical example [4] using a simple boost
converter is shown.
Another novel transformerless topology [4] featuring a
high efficiency time-sharing dual mode single-phase
partially controlled sinewave PWM inverter composed of
quasi time-sharing sinewave boost chopper with a new
functional bypass diode Db in the boost chopper side and
complementary sinewave PWM full-bridge inverter (Fig. 8)
.
a)
b)
Fig.8. Time-sharing dual-mode sinewave modulated
single-phase inverter with boost chopper [6]
a) Circuit system configuration. b) Operating principle
7
PV inverters without DC-DC converter
The block diagram of this topology is shown in the Fig.
9a.
DC
PV
Array
Grid
AC
(a)
In Fig. 10b, a typical transformerless topology is shown
using PWM IGBT inverters. This topology can be used
when there are available a large nr. of solar panels
connected in series producing in exces of the grid voltage
peak at all times.
Another interesting PV inverter topololgy without boost
and isolation can be achieved using multilevel concept. Grid
connected photovoltaic systems with a five level cascaded
inverter is presented in Fig 10c [7]. The redundant inverter
states of the five level cascaded inverter allow for a cyclic
switching scheme which minimizes the switching frequency,
equalizes stress evenly on all switches and minimizes the
voltage ripple on the DC capacitors.
IV. CONTROL OF SINGLE-PHASE PV-INVERTERS
(b)
Fig. 9. PV inverter system without DC-DC converter
and with isolation transformer
a) general diagram b) practical example with full-bridge inverter and gridside transformer [4]
Control DC-DC boost converter
In order to control the output dc-voltage to the desired value,
a control system is needed which can automatically adjust
the duty cycle, regardless of the load current or input
changes. There two types of control for the dc-dc
converters: the direct duty-cycle control and the current
control [8]. As shown in the Fig. 11.
In Fig.9b are presented two topologies of PV inverters
where the line frequency transformer is used. For higher
power levels, self-commutated inverters using thyristors are
still being used on the market [4]
vFC(t)
Error
vref + signal
Reference
input
Control
signal
Compensator
PV inverters without DC-DC converter without isolation
The block diagram of this topology is shown in the
Fig.10a
iload(t)
Pulse-width
modulator
Converter
vDC(t)
d(t)
Sensor gain
(a)
vFC(t)
PV
Array
DC
Grid
AC
(a)
Error
signal
vref +
Reference
input
-
Control
signal
Compensator
iload(t)
Comparator and
controller
iswitch_ref(t)
vDC(t)
Converter
d(t)
iswitch(t)
iswitch(t)
Sensor gain
(b)
Fig. 11. Control strategies for switched dc-dc converters
a) direct duty-cycle control b) current control
(b)
(c )
Fig. 10. Transformerless PV inverter system without DC-DC converter
a) general diagram b) typical example with full-bridge inverter [4]
c) multilevel [7]
Duty-Cycle control
The output voltage is measured and then compared to the
reference. The error signal is used as input in the
compensator, which will compute from it the duty-cycle
reference for the pulse-width modulator.
Current Control
The converter output is controlled by choice of the
transistor peak current. The control signal is a current and a
simple control network switches on and off the transistor
such its peak current follows the control input. The current
control, in the case of an isolated boost push-pull converter
has some advantages against the duty-cycle control like
simpler dynamics (removes one pole from the control-to
output transfer function) . Also as it uses a current sensor it
8
can provide better protection of the switch by limiting the
current to acceptable levels.
Another issue is the transformer saturation. In the
transformer it can be induced a dc bias current generated by
small voltage imbalances due to the small differences in
boost inductors and/or switches. This dc current bias
increases or decreases the transistor currents. The current
control will alter the switch duty cycles in a way that these
imbalances tend to disappear and the transformer voltsecond balance to be maintained. Finally, the current
control is better suited to modularity where current sharing
needs to be solved when running in parallel.
Among the drawbacks of the current control it can be
mentioned that it required an extra current sensor and it has
a susceptibility to noise and thus light filtering of feedbcak
signals is required. The current control become unstable
whenever the duty-cycle becomes larger than 0.5 but this
drawback can be overcome by ramping the reference current
signal.
Considering the above-enumerated arguments, the current
control seems to be more attractive for PV inverter
applications and it is widely used.
two well known drawbacks: inability of the PI controller to
track a sinusoidal reference without steady-state error and
poor disturbance rejection capability. This is due to the poor
performance of the integral action.
Control of DC-AC grid converter
For the grid-connected PV inverters in the range of 1-5
kW, the most common control structure for the dc-ac grid
converter is using a current-controlled H-bridge PWM
inverter having a low-pass output filter. Typically L filters
are used bu the new trend is to use LCL filters that being a
higher order filter (3rd) leads to more compact design. The
drawback is that due to its own resonance frequency it can
produce stability problems and special control design is
required [9]. A typical dc-ac grid converter with LCL filter
is depicted in the Fig. 12
The PI current controller GPI(s) is defined as:
Fig.12. The H-bridge PV inverter connected to the grid
through an LCL filter
The harmonics level in the grid current is still a
controversial issue for PV inverters. The IEEE 929 standard
from 2000 allows a limit of 5% for the current total
harmonic distortion (THD) factor with individual limits of
4% for each odd harmonic from 3rd to 9th and 2% for 11th
to 15th while a recent draft of European IEC61727 suggests
something similar. These levels are far more stringent than
other domestic appliances such as IEC61000-3-2 as PV
systems are viewed as generation sources and so are subject
to higher standards than load systems.
Classical PI control with grid voltage feed-forward
[10],[11] as depicted in Fig. 13a is commonly used for
current-controlled PV inverters, but this solution exhibits
ii*
GPI(s)
ii
ui *
Gd(s) Gf(s)
ii
ug
(a)
ii*
ui *
Gc(s)
ii
Gd(s)
Gf(s)
ii
Gh(s)
(b)
a)
Fig. 13.The current loop of PV inverter.
with PI controller; b) with P+Resonant (PR) controller
KI
(1)
s
In order to get a good dynamic response, a grid voltage
feed-forward is used, as depicted in Fig. 15a. This leads in
turn to stability problems related to the delay introduced in
the system by the voltage feedback filter.
In order to alleviate these problems, a second order
generalized integrator (GI) as reported in [12] can be used.
The GI is a double integrator that achieves an infinite gain at
a certain frequency, also called resonance frequency, and
almost no attenuation exists outside this frequency. Thus, it
can be used as a notch filter in order to compensate the
harmonics in a very selective way. This technique has been
primarily used in three-phase active filter applications as
reported in [12] and also in [13] where closed-loop
harmonic control is introduced. Another approach reported
in [14] where a new type of stationary-frame regulators
called P+Resonant (PR) is introduced and applied to threephase PWM inverter control. In this approach the PI dccompensator is transformed into an equivalent accompensator, so that it has the same frequency response
characteristics in the bandwidth of concern. The current loop
of the PV inverter with PR controller is depicted in the Fig.
13b.
The P+Resonant (PR) current controller Gc(s) is defined
as [12], [15]:
GPI ( s ) = K P +
Gc ( s ) = K P + K I
s
s 2 + ωo2
The harmonic compensator (HC) Gh(s) as defined in [10]:
9
∑
Gh ( s ) =
h =3,5,7
K Ih
25
s
s 2 + (ωo h )
15
10
is designed to compensate the selected harmonics 3rd, 5th and
7th as they are the most prominent harmonics in the current
spectrum . A processing delay typical equal to Ts for the
PWM inverters [8] is introduced in Gd ( s) . The filter transfer
-10
function Gf(s) is expressed in (4) [16].
-15
(
(
2
2
i ( s)
1 s + z LC
=
G f (s) = i
2
ui ( s ) Li s s 2 + ω res
Ig (exp) [5A/div]
Ug (exp) [100/div]
20
2
5
0
-5
-20
)
)
-25
0
0.005
0.01
0.015
0.02 0.025
time[sec]
0.03
0.035
0.04
(a)
25
( Li + Lg ) ⋅ z
−1
2
2
where z LC
=  Lg C f  and ωres
=
L
Ig (exp) [5A/div]
Ug (exp) [100/div]
20
2
LC
15
10
i
5
The current error - disturbance ratio rejection capability at
null reference is defined as:
0
-5
-10
ε ( s)
ug ( s)
=
ii* = 0
G f (s)
-15
1 + ( Gc ( s ) + Gc ( s ) ) ⋅ Gd ( s ) ⋅ G f ( s )
-20
-25
0
where: ε is current error and the grid voltage ug is
considered as the disturbance for the system.
The Bode plots of disturbance rejection for the PI and PR
controllers are shown in Fig 14. As it can be observed, The
PR provides much higher attenuation for both fundamental
and lower harmonics then PI. The PI rejection capability at
5th and 7th harmonic is comparable with that one of a simple
proportional (P) controller, the integral action being
irrelevant.
0.005
0.01
0.015
0.02 0.025
time[sec]
0.03
0.035
0.04
(b)
25
Ig (exp) [5A/div]
Ug (exp) [100/div]
20
15
10
5
0
-5
0
-10
-15
-50
PR+HC
PI
P
-20
-25
-100
0
0.005
0.01
0.015
0.02
time[sec]
0.025
0.03
0.035
0.04
(c)
Fig.15. Experimental results at 3kW. Grid voltage and current. a) with PI
controller. B) with PR; c) with PR+HC
-150
-270
-360
-450
-540
10
1
10
2
10
3
Fig. 14. Bode plot of disturbance rejection (current error ratio disturbance)
of the PR+HC, P and PR current controllers.
Thus it is demonstrated the superiority of the PR controller
respect to the PI in terms of harmonic current rejection. In
[15] the discrete implementation on a low-cost fixed-point
DSP is demonstrated. In Fig. 15 some experimental results
with a 3kW PV inverter are shown demonstrating the
harmonic compensation.
The issue of stability when several PV inverters are
running in parallel on the same grid is becoming more and
more important especially when LCL filters are used. In
[17] it is shown that in the case of a concentration of several
hundreds of solar roofs in Holland, resonance frequencies in
the range of 1-2 kHz are occurring as a resulat of the grid
interaction with the LCL filters. Thus, special attention is
required when designing the current control. A method for
designing both the controller and LCL filter ensuring
stability is shown in [9].
MPPT
In order to capture the maximum power, a maximum
power point tracker (MPPT) is required. The maximum
power point of solar panels is a function of solar irradiance
10
and temperature as depicted in Fig..16. This function can
be implemented either in the dc-dc converter or in the dc-ac
converter. Several algorithms can be used in order to
implement the MPPT as followings.
(a)
a)
(b)
Fig. 16: PV characteristics.
Irradiance dependence b) temperature dependence
Perturb and Observe
The most commonly used MPPT algorithm is Perturb and
Observe (P&O), due to its ease of implementation in its
basic form [17]. Figure 16 shows the P vs. V and I curves of
a PV array, which has a global maximum at the MPP. Thus,
if the operating voltage of the PV array is perturbed in a
given direction and dP/dV > 0, it is known that the
perturbation moved the array's operating point toward the
MPP. The P&O algorithm would then continue to perturb
the PV array voltage in the same direction. If dP/dV < 0,
then the change in operating point moved the PV array away
from the MPP, and the P&O algorithm reverses the direction
of the perturbation. A problem with P&O is that it oscillates
around the MPP in steady state operation. It also can track in
the wrong direction, away from the MPP, under rapidly
increasing or decreasing irradiance levels. There are several
variations of the basic P&O that have been designed to
minimize these drawbacks. These include using an average
of several samples of the array power and dynamically
adjusting the magnitude of the perturbation of the PV
operating point.
Incremental Conductance
The incremental conductance algorithm seeks to overcome
the limitations of the P&O algorithm by using the PV array's
incremental conductance to compute the sign of dP/dV
without a perturbation [17]. It does this using an expression
derived from the condition that, at the MPP, dP/dV = 0.
Beginning with this condition, it is possible to show that, at
the MPP dI/dV = -I/V. Thus, incremental conductance can
determine that the MPPT has reached the MPP and stop
perturbing the operating point. If this condition is not met,
the direction in which the MPPT operating point must be
perturbed can be calculated using the relationship between
dI/dV and -I/V. This relationship is derived from the fact
that dP/dV is negative when the MPPT is to the right of the
MPP and positive when it is to the left of the MPP. This
algorithm has advantages over perturb and observe in that it
can determine when the MPPT has reached the MPP, where
perturb and observe oscillates around the MPP. Also,
incremental conductance can track rapidly increasing and
decreasing irradiance conditions with higher accuracy than
perturb and observe. One disadvantage of this algorithm is
the increased complexity when compared to perturb and
observe. This increases computational time, and slows down
the sampling frequency of the array voltage and current.
Parasitic Capacitance
The parasitic capacitance method is a refinement of
theincremental conductance method that takes into account
the parasitic capacitances of the solar cells in the PV array
[17]. Parasitic capacitance uses the switching ripple of the
MPPT to perturb the array. To account for the parasitic
capacitance, the average ripple in the array power and
voltage, generated by the switching frequency, are measured
using a series of filters and multipliers and then used to
calculate the array conductance. The incremental
conductance algorithm is then used to determine the
direction to move the operating point of the MPPT. One
disadvantage of this algorithm is that the parasitic
capacitance in each module is very small, and will only
come into play in large PV arrays where several module
strings are connected in parallel. Also, the DC-DC converter
has a sizable input capacitor used filter out small ripple in
the array power. This capacitor may mask the overall effects
of the parasitic capacitance of the PV array.
Constant Voltage
This algorithm makes use of the fact that the MPP voltage
changes only slightly with varying irradiances, as depicted
in Fig. 16a. The ratio of VMP/VOC depends on the solar
cell parameters, but a commonly used value is 76% [17]. In
this algorithm, the MPPT momentarily sets the PV array
current to zero to allow a measurement of the array's open
circuit voltage. The array's operating voltage is then set to
76% of this measured value. This operating point is
maintained for a set amount of time, and then the cycle is
repeated. A problem with this algorithm is available energy
is wasted when the load is disconnected from the PV array,
also the MPP is not always located at 76% of the array’s
open circuit voltage.
Anti-islanding
In addition to the typical power quality regulations
concerning the harmonic distortion and EMI limits, the grid-
11
connected PV inverters must also meet specific power
generation requirements like the islanding detection, or even
certain country-specific technical recommendations for
instance the grid impedance change detection (in Germany).
Such extra-requirements contribute to a safer grid-operation
especially when the equipment is connected in dispersed
power generating networks but impose additional effort to
readapt the existing equipments.
The European standard EN50330-1 (draft) [19] describes
the ENS (the German abbreviation of Mains monitoring
units with allocated Switching Devices) requirement, setting
the utility fail-safe protective interface for the PV
converters. The goal is to isolate the supply within 5 seconds
after an impedance change of Z = 0.5 W, which is associated
with a grid failure. The main impedance is typically detected
by means of tracking and step change evaluation at the
fundamental frequency. Therefore, a method of measuring
the grid impedance value and its changes should be
implemented into existing PV-inverters.
One solution is to attach a separate device developed only
for the measuring purpose as depicted in Fig. 17a.
(a)
(b)
Fig. 1. Grid-impedance measurement for PV inverters. a) using external
device; b) embedded on the inberter control using harmonic injection
This add-on option is being commonly used in the
commercial PV inverters, but the new trend is to implement
this function embedded in the inverter control without extra
hardware. Numerous publications exist in this field, which
offer measuring solutions for the grid impedance for a wide
frequency range from dc up to typically 1 kHz [20].
Unfortunately, not always can these methods easily be
embedded into a non-dedicated platform, i.e. PV-inverters
featuring typically a low-cost DSP. Specific limitations like
Harmonic
Injection
Anti-aliasing Sample&Hold
A/D
real-time computation, A/D conversion accuracy and fixedpoint numerical limitation, are typically occurring.
A novel approach presented in [21], [22] estimates the grid
impedance on-line with the purpose of detection the step
change of 0.5 Ω as required in [19] as shown in Fig. 17b..
The solution is found by injecting a test signal trough the
inverter modulation process. This signal, an interharmonic
current with a frequency close to the fundamental,
determines a voltage drop due to the grid impedance, which
is measured by the existing PV-inverter sensors. Then, the
same CPU unit that makes the control algorithm carries out
the calculations and gives the grid impedance value. The
principle of this method is shown in Fig. 18.
This approach provides a fast and low cost solution to meet
the required standards and was succesfully implemented on
a TMS320F24x 16-bit fixed point DSP platform as an addon to the existing control.
V. CONTROL OF THREE-PHASE INVERTERS
The control of a three-phase inverter connected to the grid
has more in common with the control of an active
rectifier/filter rather than with the control of an ac drive. In
fact with the first the distributed inverter shares the
characteristic to be connected to the grid on the ac side,
while with second it shares the common characteristic to
have less responsibilities in the management of the dc-link
voltage that is usually controlled by another converter stage.
Hence from the control perspective the three-phase
distributed inverter as an advantage over the rectifier and a
disadvantage over the motor inverter.
Its control issues will be discussed starting from its
mathematical model both with L-filter and LCL-filter on the
grid side.
Then simple controls as well as few advanced ones will be
introduced and briefly discussed. Finally some advanced
topics and some experimental results will close this Section.
Mathematical Model of the L-filter inverter
The state of the three-phase inverter is modelled by means
of a switching space-vector defined with the switching
functions p j (t ) (j = a, b, c)
p(t) = 2 ( pa(t) + α ⋅ p b(t) + α 2 ⋅ pc(t) )
3
then if the inverter is connected to the grid through an Lfilter (Fig. 1)
Windowing Pre-processing
DFT
Harmonic Voltage
Impedance Post-processing Internal Logic
Voltage Signal
U
Z= —
I
Grid
Anti-aliasing Sample&Hold
A/D
Windowing Pre-processing
DFT
Harmonic Current
Current Signal
Fig. 18. Implementation structure of grid impedance estimation
Tripping
&
Display
12
v (t ) = e (t ) + Ri (t ) + L
d i (t )
dt
(2)
Mathematical Model of the LCL-filter inverter
In the following the LCL-filter based inverter model is
reported in order to highlight the increased complexity of
the system. The system is shown in Fig. 2.
v(t) = 1 p(t)v o(t)
2
Fig. 1 L-filter inverter connected to the grid.
assuming to neglect the dc voltage dynamic such as the dc
voltage vo(t) is ain input for the system. Moreover v (t ) is
Fig. 2 LCL-filter inverter connected to the grid.
the space-vector of the inverter input voltages; i (t ) is the
space-vector of the inverter input currents; e (t ) is the
space-vector of the input line voltages.
The mathematical model written in the state space form is
AC Current control
The ac current control (CC) is usually adopted because the
current controlled converter exhibits, in general, better
safety, better stability and faster response [1].
This solution ensures several additional advantages. The
feedback loop also results in some limitations, such as that
fast-response voltage modulation techniques must be
employed, like PWM. Optimal techniques, which use
precalculated switching patterns within the ac period, cannot
be used, as they are not oriented to ensure current waveform
control [1].
Generally the current control is the most inner loop of a
cascade control that employ a dc-link voltage level
management system and active and reactive power
controller as reported in Fig. 3
The use of an ac LCL-filter claims for a deep dynamic and
stability analysis of the current control loop [2]. In order to
highlight the stability problems that arise from the use of an
LCL-filter it is sufficient to show the d or q system plant in
Laplace domain. If the converter side current is sensed, the
system plant is
d i (t ) 1 
1

=  − Ri (t ) − e (t ) + p (t )vo (t ) 
L
dt
2

(4)
A commonly used approach in analysing three-phase
systems is to adopt a dq-frame that rotates at the angular
speed ω (where ω = 2πf and f is the fundamental frequency
of the power grid’s voltage waveform). The space-vectors
which express the inverter electrical quantities are projected
on the d-axis and q-axis. As a consequence if a space-vector
with constant magnitude rotates at the same speed of the
frame, it has constant d- and q- components while if rotates
at a different speed or it has a time-variable magnitude it has
pulsating components. Thus in a dq-frame rotating at the
angular speed ω (2) becomes
 did ( t )
1
1

− ωiq ( t ) =  − Rid ( t ) − ed ( t ) + pd ( t ) vo ( t ) 

L
2

 dt

di
t
(
)
1
1

 q
 dt + ωid ( t ) = L  − Riq ( t ) − eq ( t ) + 2 pq ( t ) vo ( t ) 

(5)
(5) shows how in the dq-frame the d- and q- differential
equations for the current are dependent due to the crosscoupling terms ωiq(t) and ωid(t).
 R1
− L
 1

−ω
 i1d  
 i  
 1q   1

 
d  vC f d   C f
=
dt  vC f q  

  0
 i2 d  
i  
 2q   0


 0

ω
−
1
L1
0
0
1
L1
0
R1
L1
0
0
0
ω
1
Cf
−ω
0
1
L2
0
0
1
L2
−
0
0
−
−
−
1
Cf
0
−
R2
L2
−ω
G ( s) =
(
(
2
2
i( s)
1 s + z LC
=
2
v( s ) L2 s s 2 + ωres
)
)
(7)
If the grid side current is sensed, the plant for control is
2
z LC
i( s)
1
G ( s) =
=
(8)
2
v( s ) L2 s ( s 2 + ωres
)
2
=  L1C f 
where z LC

0 


 1
0  i
 1d   − L

  1
  i1q  
0 
0
v  
  Cfd  + 
1   vC f q   0
−

 
Cf  i   0
2d

  0
ω   i2 q  

 0
R2 
−
L2 
−1
2
2
and ωres
= ( L1 + L2 ) z LC
L2 .

 0
0 
 0


1
 0
− 
L1  ed   0

+
0   eq   1


0 
 L2

0 

 0
0 

0 
0 

0 

0   vd 
  vq 
0  

1 

L2 
(6)
13
In both the cases the two poles related to the resonance of
the LCL-filter challenge the current control instability,
particularly the second one (sensing of the grid current)
generally lead to a more stable behaviour.
Two axis-based current control
The most used control technique is the two axis-based [1].
Then if the two-axis system is a stationary αβ-frame, the
proportional plus resonant controller can be adopted and it is
DPR ( s )αβ
Ki s

K p + 2
+ ω0 2
s
=

0





Ks 
Kp + 2 i 2 
s + ω0 
0
(9)
If the frame is a rotating dq-frame, classical PI can be used
DPI ( s ) dq
Ki

K p + s
=

0




Ki 
Kp + 
s 
0
grid current. This problem can be solved both in a stationary
αβ-frame both in a rotating dq-frame. In the first case it is
sufficient to plug in other resonant controller also called
harmonic compensators
GR ( s )αβ =
∑
h = 3,5,7
kih
s
s + ( ω0 ⋅ h )
2
2
(12)
where h is the order of the harmonic to be compensated.
If the controller adopts a rotating dq-frame approach it is
possible to introduce other dq-frame rotating at multiple
speed respect to the fundamental one and adopting standard
PI in each of them. In both the cases it is necessary that the
harmonics to be compensated stay into the bandwidth of the
current controller otherwise stability problems arise.
Current control active damping
(10)
If this controller is transformed into an αβ-frame then
DPI ( s )αβ
Ki s

K p + 2
s
+ ω0 2
=

ω0 Ki
 − 2
s + ω0 2

ω0 Ki



Ks 
Kp + 2 i 2 
s + ω0 
s 2 + ω0 2
(11)
(11) it is equal to (9) except for non-diagonal terms. Hence
the PI controller in dq-frame and PR controller in αβ-frame
can achieve similar performances.
Fig. 1. The block diagram of a typical three-phase distributed inverter.
In the case of dq-frame, if it is oriented such as the d-axis
is aligned on the grid voltage vector the control is called
Voltage Oriented Control (VOC) (Fig 4). The reference
current d-component i*d is controlled to manage the active
power flow while the reference current q-component i*q is
controlled to mange the reactive power flow. To have the
grid current vector in phase with the grid voltage vector, i*q
should be zero.
Grid voltage harmonic compensators
The grid voltage is usually affected by a background
distortion that can result in a high harmonic distortion of the
Fig. 4. Voltage Oriented Control based on the use of a rotating dq-frame.
This solution seems very attractive especially in applications
above several kW, where the use of a damping resistor
increases the encumbrances, the losses could claim for
forced cooling and the efficiency decrement becomes a key
point. In [3] a lead-lag network has been used on the filter
capacitor voltage and it is possible to avoid the use of new
sensors because this voltage is near to the grid which is
normally sensed. Moreover, in [4] an interesting approach to
perform active damping has been proposed: a virtual resistor
is added. The virtual resistor is an additional control
algorithm that makes the LCL-filter behaving as if there was
a real resistor connected to it. However, an additional
current sensor is needed if the virtual resistor is connected in
series to the filter inductor or capacitor and an additional
voltage sensor is needed, if it is connected in parallel.
Basically all these approaches are multiloop-based [5] while
an alternative solution consists in adopting a more complex
controller acting as a digital filter around the resonance
frequency of the LCL-filter [2].
Direct power control
In the last years the most interesting emerging technique has
been the direct power control developed in analogy to the
well known direct torque control used for drives. In DPC
there are no internal current loops and no PWM modulator
block because the converter switching states are
14
appropriately selected by a switching table based on the
instantaneous errors between the commanded and estimated
values of active and reactive power [1], [6] Fig 5. The main
advantage of the DPC is in its simple algorithm instead the
main disadvantage is in the need for high sampling
frequency required to obtain satisfactory performances.
Reduction of the number of sensors
The basic number of needed sensors is 4 (two ac currents
and two ac voltages). However this number can be reduced
avoiding the use of grid voltage vector implementing a
virtual sensor or using a zero crossing detector in order to
have the phase reference for the current in order to have
unity power factor. Moreover if a feedforward current
control technique is adopted the grid current sensors can be
avoided but it is essential to provide a method for
overcurrent protection in industrial applications.
EMC-issues
The main EMC-issues are related to the low frequency
range and thus to the correct control to the current. Thus the
use of a LCL-filter on the ac side is an interesting solution:
reduced values of the inductance can be used and the grid
current is almost ripple free. The design of the LCL-filter
has been investigated [11].
Future research topics
Some intriguing topics of research are:
•
•
•
•
•
•
the immunity of the inverter to the presence of
polluting loads connected to the same PCC
compliance with international standards/need for new
standards respect to the harmonics due to the
switching;
to reduce the number of components;
whether use or not the Phase Locked Loop;
whether to use grid voltage feedforward or not;
grid current sensor position
Results
Some tests results, obtained on the set-up shown in Fig. 6,
are reported in order to evaluate the impact of the non-ideal
conditions on the behaviour of a PR-based controller in a
αβ-frame (Fig. 7), the use of harmonic compensator in a
stationary αβ-frame to mitigate these effects (Fig. 8) and
finally the effect of active damping (Fig. 9).
Fig. 5. Direct Power Control based on the active and reactive power
calculation.
In [8] an algorithm to estimate position of line voltage is
presented. The proportional-plus-integral current regulator is
modified to obtain the angle error signal driving an
observer, similar in structure to a phase-locked loop, which
provides the angle of line voltages.
Non-ideal conditions
The non-ideal conditions are many and they can affect
very much the overall system performance. They are too
long computation time, presence of acquisition filters, ac
phase unbalance, location of the grid voltage sensors after a
dominant reactance and passive damping if an LCL-filter is
used. A proper design to take into consideration them should
be provided [9].
It is well known that the grid unbalance causes even
harmonics at the dc output and odd harmonics in the input
current [10]. Some solutions have been studied such as the
use of negative sequence in the reference current that
unfortunately leads to uncontrollability of the power factor
or the use of two current controllers for positive and
negative sequences, which also can create stability
problems.
Fig. 6. Laboratory set-up The block diagram of a typical three-phase
distributed inverter.
15
Fig. 7. Compensation of grid background distortion: grid currents [2 A/div]
and grid voltage [100 V/div] (sampling/switching 10 kHz, active power 2
kW, PR-controllers in a αβ-frame).
generators are used in fixed speed systems, which are almost
independent of torque variation and operate at a fixed speed
(slip variation of 1-2%). Fig. 11 shows different topologies
for the first category of wind turbines.
All three systems are using a soft-starter (not shown in
Fig. 11) in order to reduce the inrush current and thereby
limit flicker problems on the grid. They also need a reactive
power compensator to reduce (almost eliminate) the reactive
power demand from the turbine generators to the grid.
It is usually done by continuously switching capacitor
banks following the production variation (5-25 steps). Those
solutions are attractive due to cost and reliability but they
are not able (within a few ms) to control the active power
very fast. The generators have typically a pole-shift
possibility in order to maximize the energy capture.
The next category is variable speed systems [6]-[35] where
pitch control is typically used. Variable speed wind turbines
may be further divided into two parts, one with partially
rated power electronic converters and one with fully rated
power electronic converters.
I
Induction
generator
Grid
Gear
Pitch
Reactive
compensator
(a)
II
Induction
generator
Fig. 8. Compensation of grid background distortion: grid currents [2 A/div]
and grid voltage [100 V/div] (sampling/switching 10 kHz, active power 2
kW, PR-controllers in a αβ-frame with 5th and 7th harmonic compensators).
Grid
Gear
Stall
Reactive
compensator
(b)
III
Induction
generator
Grid
Gear
Active
Stall
Fig. 9. Control change from active damping to no damping (t=40 ms): grid
currents [2 A/div] (sampling/switching 10 kHz, active power 2 kW, PRcontrollers in a αβ-frame).
VI. CONVERTER TOPOLOGIES FOR WIND TURBINES
In a fixed speed wind power conversion system, the
power may be limited aerodynamically either by stall, active
stall or by pitch control [6], [7]. Normally induction
Reactive
compensator
(c)
Fig. 11. Wind turbine systems without power converter but with
aerodynamic power control.
Pitch controlled (System I) b) Stall controlled (System II) c) Active stall
controlled (System III)
16
IV
W ounded Rotor
Induction
generator
Grid
Gear
Resistance
control
with PE
Pitch
Reactive
compensator
(a)
V
Doubly-fed
induction generator
Grid
Gear
Pitch
DC
AC
DC
Pref
AC
Qref
(b)
Fig. 12. Wind turbine topologies with partially rated power electronics and
limited speed range, (a) Rotor-resistance converter (System IV) (b) Doublyfed induction generator (System V).
Fig. 12 shows wind turbines with partially rated power
electronic converters that are used to obtain an improved
control performance. Fig. 12a shows a wind turbine system
where the generator is an induction generator with a
wounded rotor. An extra resistance is added in the rotor,
which can be controlled by power electronics. This is a
dynamic slip controller and it gives typically a speed range
of 2-10 %. The power converter for the rotor resistance
control is for low voltage but high currents. At the same
time an extra control freedom is obtained at higher wind
speeds in order to keep the output power fixed. This solution
still needs a soft-starter and a reactive power compensator.
A second solution of using a medium scale power
converter with a wounded rotor induction generator is
shown in Fig. 12b [18]-[26]. Slip-rings are making the
electrical connection to the rotor. A power converter
controls the rotor currents. If the generator is running supersynchronously electrical power is delivered through both the
rotor and the stator. If the generator is running subsynchronously electrical power is only delivered into the
rotor from the grid. A speed variation of ±30 % around
synchronous speed can be obtained by the use of a power
converter of 30 % of nominal power.
Furthermore, it is possible to control both active (Pref) and
reactive power (Qref), which gives a better grid performance,
and the power electronics enable the wind turbine to act
more as a dynamic power source to the grid. The solution
shown in Fig. 12b needs neither a soft-starter nor a reactive
power compensator. The solution is naturally a little bit
more expensive compared to the classical solutions shown
in Fig. 11 and Fig. 12a. However, it is possible to save
money on the safety margin of gear, reactive power
compensation units and it is possible to capture more energy
from the wind.
The wind turbines with a full-scale power converter
between the generator and grid give extra losses in the
power conversion but it may be gained by the added
technical performance [9]. Fig. 13 shows four possible
solutions with full-scale power converters.
The solutions shown in Fig. 13a and Fig. 13b are
characterized by having a gear. A synchronous generator
solution shown in Fig. 13b needs a small power converter
for field excitation. Multi-pole systems with the
synchronous generator without a gear are shown in Fig. 13c
and Fig. 13d.
The last solution uses permanent magnets, which are still
becoming cheaper and thereby more attractive. All four
solutions have the same controllable characteristics since the
generator is decoupled from the grid by a dc-link. The
power converter to the grid enables the system very fast to
control active and reactive power. However, the negative
side is a more complex system with a more sensitive
electronic part.
By introducing power electronics many of the wind
turbine systems get a performance like a power plant. In
respect to control performance they are faster but of course
the produced real power depends on the available wind. The
reactive power can in some solutions be delivered without
having any wind.
VI
Induction
generator
AC
Gear
DC
AC
Qref
Pref
Pitch
Grid
DC
(a)
DC
AC
V II
AC
G ear
P itc h
S y n c h ro n o u s
G e n e ra to r
G rid
DC
DC
AC
P re f
Q re f
(b)
DC
VIII
AC
Grid
AC
DC
DC
Pitch
AC
Synchronous
Generator
Multi-pole
P ref Q
ref
(c)
17
IX
PM-synchronous
G enerator
Multi-pole
AC
Pitch
G rid
DC
DC
The power control system should also be able to limit the
power. An example of an overall control scheme of a wind
turbine with a doubly-fed generator system is shown in Fig.
10.
Below maximum power production the wind turbine will
typically vary the speed proportional with the wind speed
and keep the pitch angle θ fixed. At very low wind the speed
of the turbine will be fixed at the maximum allowable slip in
order not to have overvoltage.
A pitch angle controller will limit the power when the
turbine reaches nominal power. The generated electrical
power is done by controlling the doubly-fed generator
through the rotor-side converter. The control of the grid-side
converter is simply just keeping the dc-link voltage fixed.
Internal current loops in both converters are used which
typically are linear PI-controllers, as it is illustrated in Fig.
11a. The power converters to the grid-side and the rotor-side
are voltage source inverters.
Another solution for the electrical power control is to use
the multi-pole synchronous generator. A passive rectifier
and a boost converter are used in order to boost the voltage
at low speed. The system is industrially used today. It is
possible to control the active power from the generator. The
topology is shown in Fig. 11b. A grid inverter is interfacing
the dc-link to the grid. Here it is also possible to control the
reactive power to the grid. Common for both systems are
they are able to control reactive and active power very fast
and thereby the turbine can take part in the power system
control.
AC
P ref
Q ref
(d)
Fig. 13. Wind turbine systems with full-scale power converters.
a) Induction generator with gear (System VI)
b) Synchronous generator with gear (System VII)
c) Multi-pole synchronous generator (System VIII)
d) Multi-pole permanent magnet synchronous generator (System IX)
Fig. 13 also indicates other important issues for wind
turbines in order to act as a real power source for the grid.
They are able to be active when a fault appears at the grid
and so as to build the grid voltage up again quickly; the
systems have the possibility to lower the power production
even though more power is available in the wind and
thereby acting as a rolling capacity. Finally, some are able to
operate in island operation in the case of a grid collapse.
VII. CONVERTER TOPOLOGIES FOR WIND TURBINES
Controlling a wind turbine involves both fast and slow
control. Overall the power has to be controlled by means
of the aerodynamic system and has to react based on a setpoint given by dispatched center or locally with the goal to
maximize the production based on the available wind power.
Measurement
grid point M
N
DC
AC
DC
T
U
PWM
meas
dc
AC
PWM
I acmeas
θ
meas
I rotor
Rotor side
converter controller
Grid side
converter controller
meas
Pgrid
meas
Qgrid
DFIG control
ω
Power controller
meas
gen
cross -coupling
conv , ref
Pgrid
conv , ref
Q grid
U dcref
meas
Pgrid
Speed controller
rated , ref
Pgrid
Wind turbine control
Fig. 10. Control of wind turbine with doubly-fed induction generator system [35 ].
Grid
operators
control system
18
DFIG
Transformer
Grid
Gear
Rotor-side
converter
θr
Grid-side
converter
Inductance
P grid
Q
grid
3
vra, vrb, vrc
i ra, irb, i rc
vga,v gb, v gc
iga, igb, i gc
v DC
Rotor
Pref control Q ref
PMG
Grid
control
Generator
rectifier
Grid
inverter
Inductance
Grid
vDC
Pref
vDC
Power
control
Grid
control Q
vga,v gb, v gc
iga, igb, igc
ref
Fig. 11. Basic control of active and reactive power in a wind turbine [17].
a) Doubly-fed induction generator system (System V)
b) Multi-pole synchronous generator system (System VIII)
SUMMARY
REFERENCES
[1]
J.F. Manwell, J. G. McGowan, A. L. Rogers, Wind Energy Explained,
Theory, Design and Application. Wiley, 1998.
[2] S. Heier. Grid integration of wind energy conversion
systems/translated by Rachel Waddington. John Wiley, 1998.
[3] T. Burton, D. Sharpe, N. Jenkins, E. Bossanyi. Wind Energy
Handbook. John Wiley, 2000.
[4] P. Thoegersen, F. Blaabjerg. “Adjustable Speed Drives in the Next
Decade. Future Steps in Industry and Academia”. Journal of Electric
Power Components and Systems, Vol. 32, No. 1, 2004, pp. 13-32.
[5] B.J. Baliga, “Power IC’s in the saddle”, IEEE Spectrum, July 1995,
pp. 34-49.
[6] F. Blaabjerg, Z. Chen, “Power Electronics as an enabling technology
for Renewable Energy Integration”, Journal of Power Electronics,
Vol. 3, No. 2, April 2003, pp. 81-89.
[7] F. Blaabjerg, Z. Chen, S.B. Kjaer, “Power Electronics as Efficient
Interface in Dispersed Power Generation Systems”, IEEE Trans. on
PE, Vol. 19, No. 4, 2004 (in press).
[8] O. Carlson, J. Hylander, K. Thorborg, “Survey of variable speed
operation of wind turbines”, Proc. of European Union Wind Energy
conference, Sweden, 1996, pp.406-409.
[9] M.R. Dubois, H. Polinder, J.A. Fereira, J.A, “Comparison of
Generator Topologies for Direct-Drive Wind Turbines”, IEEE Nordic
Workshop on Power and Industrial Electronics, NORPIE 2000, 2000,
pp. 22-26.
[10] L.H. Hansen, L. Helle, F. Blaabjerg, E. Ritchie, S. Munk-Nielsen, H.
Bindner, P. Sørensen, B. Bak-Jensen, “Conceptual survey of
19
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
Generators and Power Electronics for Wind Turbines”, 2001, Risø-R1205 (EN).
M.P. Kazmierkowski, R. Krishnan, F. Blaabjerg. Control in Power
Electronics Selected problems. Academic Press, 2002.
Z. Chen, E. Spooner, ”Wind turbine power converters: a comparative
study”, Proc. of Power Electronics and Variable Speed Drives, 1998,
pp. 471 – 476.
L.H. Hansen, P.H. Madsen, F. Blaabjerg, H.C. Christensen, U.
Lindhard, K. Eskildsen, “Generators and power electronics
technology for wind turbines”, Proc. of IECON '01, Vol. 3, 2001, pp.
2000 – 2005.
J. Rodriguez, L. Moran, A. Gonzalez, C. Silva, “High voltage
multilevel converter with regeneration capability”, Proc. of PESC
‘99, 1999, Vol.2, pp.1077-1082.
E.N. Hinrichsen, “Controls for variable pitch wind turbine
generators”, IEEE Trans. On Power Apparatus and Systems, Vol.
103, No. 4, 1984, pp. 886-892.
T. A. Lipo, “Variable Speed Generator Technology Options for Wind
Turbine Generators”, NASA.Workshop on HAWTT Technology,
May 1984, pp. 214-220.
Z. Chen, E. Spooner, “Grid Interface Options for Variable-Speed,
Permanent-Magnet Generators”, IEE Proc. -Electr. Power
Applications, Vol. 145, No. 4, July 1998, pp. 273-283.
J.D. van Wyk, J.H.R. Enslin, “A Study of Wind Power Converter
with Microcomputer Based Maximal Power Control Utilising an
Oversynchronous Electronic Schertives Cascade”, Proc. of IPEC '83,
Vol. I, 1983, pp. 766-777.
D. Arsudis, W. Vollstedt, “Sensorless Power control of a Double-Fed
AC-Machine with nearly Sinusoidal Line Currents”, Proc. of EPE '89,
Aachen 1989, pp. 899-904
D.
Arsudis,
“Doppeltgespeister
Drehstromgenerator
mit
Spannungszwischenkreis Umrichter in Rotorkreis für Wind
Kraftanlagen, Ph.D. Thesis, 1998, T.U. Braunschweig, Germany.
D. Arsudis, “Sensorlose Regelung einer doppelt-gespeisten
Asynchronmaschine mit geringen Netzrückwirkungen“, Archiv für
Elektrotechnik, Vol. 74, 1990, pp. 89-97.
S. Bhowmik, R. Spee, J.H.R. Enslin, “Performance optimization for
doubly fed wind power generation systems”, IEEE Transactions on
Industry Applications, Vol. 35, No. 4 , July-Aug. 1999, pp. 949 – 958.
J.B. Ekanayake, L. Holdsworth, W. XueGuang, N. Jenkins, “Dynamic
modelling of doubly fed induction generator wind turbines”, Trans. on
Power Systems, Vol. 18, No. 2, May 2003, pp. 803- 809.
Yamamoto, O. Motoyoshi, “Active and Reactive Power control for
Doubly-Fed Wound Rotor Induction Generator”, Proc. of PESC '90,
Vol. 1, pp. 455-460.
E. Bogalecka, “Power control of a doubly fed induction generator
without speed or position sensor”, Proc. of EPE’93, Vol. 8, pp. 224228, 1993.
R. Pena, J.C. Clare, G.M. Asher, “Doubly fed induction generator
using back-to-back PWM converters and its application to variable
speed wind-energy generation”. IEE Proc. on electronic power
application, 1996, pp. 231-241.
Z. Chen, E. Spooner, “Voltage Source Inverters for High-Power,
Variable-Voltage DC Power Sources”, IEE Proc. –Generation,
Transmission and Distributions, Vol. 148, No. 5, September 2001, pp.
439-447.
M. Dahlgren, H. Frank, M. Leijon, F. Owman, L. Walfridsson, “Wind
power goes large scale”, ABB Review, 2000, Vol.3, pp. 31-37.
A.K. Wallace, J.A. Oliver, “Variable-Speed Generation Controlled by
Passive Elements”, Proc. of ICEM ’98, Vol. 3, 1998, pp. 1554-1559.
Z. Chen, E. Spooner, “Grid Power Quality with Variable-Speed Wind
Turbines”, IEEE Transactions on Energy Conversion, Vol. 16, No.2,
June 2001, pp. 148-154.
S. Bolik, “Grid Requirements Challenges for Wind Turbines”, Proc.
of Fourth International Workshop on Large-Scale Integration of Wind
Power and Transmission Networks for Offshore Windfarms, 2003.
Å. Larsson, The Power quality of Wind Turbines, Ph.D. thesis,
Chalmers University of Technology, Göteborg, Sweden, 2000.
Z. Saad-Saoud, N. Jenkins, “The application of advanced static VAr
compensators to wind farms”, Power Electronics for Renewable
Energy (Digest No: 1997/170), 1997, pp. 6/1 - 6/5.
P. Kazmierkowski, R. Krishnan, F. Blaabjerg, “Control in Power
Electronics”, Academic Press, 2002, ISBN 0-12-40277205.
[35] M. Liserre, A. Dell’Aquila and F. Blaabjerg, “Genetic algorithm
based design of the active damping for a LCL-filter three-phase active
rectifier,” IEEE Trans. on Power Electr. Vol. 19, pp. 76 -86, Jan 2004.
[36] V. Blasko and V. Kaura, “A novel control to actively damp resonance
in input LC filter of a three-phase voltage source converter”, IEEE
Trans. on Ind. App., vol. 33, 1997, pp. 542-550.
[37] P. A. Dahono, “A control method to damp oscillation in the input LC
filter of AC-DC PWM converters,” in Proc. of PESC 2002, June
2002, pp. 1630-1635,.
[38] P. C. Loh, M. J. Newman, D.N .Zmood, D.G. Holmes, “A
Comparative Anlysis of multiloop Voltage Regulation Strategies for
Single and Three-Phase UPS Systems” IEEE Trans. on Power Elect,
vol. 18, no. 5. Sept. 2003.
[39] T. Ohnishi, “Three phase PWM converter/inverter by means of
instantaneous active and reactive power control” in Proc of IECON
91, pp. 819-824.
[40] T. Noguchi, H. Tomiki, S. Kondo, I. Takahashi: “Direct power
control of PWM converter without power-source voltage sensors”,
IEEE Trans. on Ind. App., vol. 34, May/June 1998, pp. 473-479.
[41] I. Agirman and V. Blasko, “A novel control method of a VSC without
AC line voltage sensors”, IEEE Trans. on Ind. App., vol. 39,
March/April 2003, pp.519-524.
[42] M. Liserre A. Dell’Aquila, F. Blaabjerg. “Design and control of a
three-phase active rectifier under non-ideal operating conditions” in
Proc. of IAS 2002, pp. 1181-1188.
[43] A. Moran, P. D. Ziogas, G. Joos: “Design aspects of synchronous
PWM rectifier-inverter system under unbalanced input voltage
conditions”, IEEE Trans. on Ind. App., vol. 28, Nov./Dec. 1992, pp.
1286-1293.
[44] M. Liserre, F. Blaabjerg, S. Hansen: “Design and Control of an LCLfilter Based Active Rectifier”, Conf. Rec. 36th IAS Ann. Meeting,
Chicago (USA), Sept./Oct. 30-4, 2001
[45] Teodorescu,R., Blaabjerg,F., Liserre.M – “Stability of GridConnected PV Inverters with Large Grid Impedance Variation” –
Proceedings of PESC’04, Aachen
[46] T.Shimizu, M.Hirakata, T.Kamezawa, H.Watanabe - “Generation
Control Circuit for Photovoltaic Modules“ IEEE Trans. On Power
Electronics, Vol. 16, No. 3, May, 2001, pp 293 – 300
[47] Calais, M.; Myrzik, J.; Spooner, T.; Agelidis, V.G.: “Inverters for
single-phase grid connected photovoltaic systems-an
overview”, Power Electronics Specialists Conference, 2002. pesc 02.
2002 IEEE 33rd Annual, Volume: 4, 23-27 June 2002 Pages: 1995 –
2000.
[48] K. Ogura; T. Nishida; E. Hiraki; M. Nakaoka and Shinichiro Nagai:
“Time-Sharing Boost Chopper Cascaded Dual Mode Single-Phase
Sinewave Inverter for Solar Photovoltaic Power Generation System”,
35th Annuak IEEE Power Electronics Specialists Conference, Aachen,
Germany, 2004, Pages: 4763-4766.
[49] Calais, M.; Agelidis, V.G.; Borle, L.J.; Dymond, M.S.: “A
transformerless five level cascaded inverter based single
phase photovoltaic system”, Power Electronics Specialists
Conference, 2000. PESC 00. 2000 IEEE 31st Annual, Volume: 3, 1823 June 2000 Pages: 1173 - 1178 vol.3.
[50] Frede Blaabjerg; Zhe Chen; Soeren Baekhoej Kjaer: “Power
Electronics as Efficient Interface in Dispersed Power Generation
Systems”, IEEE Transactions on Power Electronics, Vol. 19, No.5,
September 2004
[51] H.Haeberlin – “Evolution of Inverters for Grid connected PV
systems from 1989 to 2000“- Proceedings of 17-th European
Photovoltaic Solar Energy Conference, Munich, Oct. 2001
[52] [1]Frede Blaabjerg; Zhe Chen; Soeren Baekhoej Kjaer: “Power
Electronics as Efficient Interface in Dispersed Power Generation
Systems”, IEEE Transactions on Power Electronics, Vol. 19, No.5,
September 2004
[53] [2]H.Haeberlin – “Evolution of Inverters for Grid connected PV
systems from 1989 to 2000“- Proceedings of 17-th European
Photovoltaic Solar Energy Conference, Munich, Oct. 2001
[54] [3]T.Shimizu, M.Hirakata, T.Kamezawa, H.Watanabe - “Generation
Control Circuit for Photovoltaic Modules“ IEEE Trans. On Power
Electronics, Vol. 16, No. 3, May, 2001, pp 293 – 300
[55] [4]Calais, M.; Myrzik, J.; Spooner, T.; Agelidis, V.G.: “Inverters for
single-phase grid connected photovoltaic systems-an overview”,
20
[56]
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
[72]
[73]
Power Electronics Specialists Conference, 2002. pesc 02. 2002 IEEE
33rd Annual, Volume: 4, 23-27 June 2002 Pages: 1995 – 2000.
[5]Ned Mohan, Tore Undeland, William P. Robbins: Power
Electronics. Converters, Applications and Design. [Mohan] – John
Wiley & Sons 2003, ISBN: 0-471-22693-9
[6]K. Ogura; T. Nishida; E. Hiraki; M. Nakaoka and Shinichiro
Nagai: “Time-Sharing Boost Chopper Cascaded Dual Mode SinglePhase Sinewave Inverter for Solar Photovoltaic Power Generation
System”, 35th Annuak IEEE Power Electronics Specialists
Conference, Aachen, Germany, 2004, Pages: 4763-4766.
[7]Calais, M.; Agelidis, V.G.; Borle, L.J.; Dymond, M.S.: “A
transformerless five level cascaded inverter based single phase
photovoltaic system”, Power Electronics Specialists Conference,
2000. PESC 00. 2000 IEEE 31st Annual, Volume: 3, 18-23 June 2000
Pages: 1173 - 1178 vol.3.
[8]Robert W. Erickson, Dragan Maksimovic: Fundamentals of Power
Electronics [Erickson] - Kluwer Academic Pub; March 1, 1997,
ISBN: 0-412-08541-0, 773 pages.
[9]Teodorescu,R., Blaabjerg,F., Liserre.M – “Stability of GridConnected PV Inverters with Large Grid Impedance Variation” –
Proceedings of PESC’04, Aachen
[10]M. Kazmierkowski, R.Krishnan, F.Blaabjerg, Control in Power
Electronics. Selected Problems, Academic Press 2002, ISBN 0-12402772-5.
[11]C. Cecati, A. Dell'Aquila, M. Liserre and V. G. Monopoli,
"Design of H-bridge multilevel active rectifier for traction systems",
IEEE Trans. on Ind. Applicat., vol. 39, Sept./Oct. 2003, pp. 15411550.
[12]X. Yuan, W. Merk, H. Stemmler, J. Allmeling – “StationaryFrame Generalized Integrators for Current Control of Active Power
Filters with Zero Steady-State Error for Current Harmonics of
Concern Under Unbalanced and Distorted Operating Conditions”
IEEE Trans. on Ind. App., vol. 38, no. 2, Mar./Apr. 2002, pp.523 –
532.
[13]P. Mattavelli “A Closed-Loop Selective Harmonic Compensation
for Active Filters” IEEE Trans. on Ind. App., vol. 37, no. 1,
january/february, 2001, pp. 81 – 89.
[14]D. N. Zmood, D. G. Holmes, “Stationary Frame Current
Regulation of PWM Inverters with Zero Steady-State Error” IEEE
Trans. on Power Electr., vol. 18, no. 3, May 2003, pp. 814 – 822.
[15]Teodorescu,R., Blaabjerg,F., Liserre.M., Borup,U.,-” A New
Control Structure for Grid-Connected PV Inverters with Zero SteadyState Error and Selective Harmonic Compensation” – Proceedings of
APEC’04, Anaheim, CA
[16]M. Liserre, F. Blaabjerg, and S. Hansen, “Design and control of
an LCL-filter based active rectifier,” IEEE Trans. on Ind. App., vol.
38, no. 2 Sept./Oct. 2001, pp. 299-307.
[17]J. H. R. Enslin and P. J. M. Heskes, “Harmonic interaction
between a large number of distributed power inverters and the
distributed network”, PESC 2003, pp. 1742-1747.
[18]Hohm, D.P.; Ropp, M.E. – “Comparative study of maximum
power point tracking algorithms using an experimental,
programmable, maximum power point tracking test bed”,
Photovoltaic Specialists Conference, 2000. Conference Record of the
Twenty-Eighth IEEE, 15-22 Sept. 2000 Pages:1699 – 1702.
[19]European Standard EN 50330-1, Photovoltaic semiconductor
converters Part 1: Utility interactive fail safe protective interface for
PV-line commutated converters - Design qualification and type
approval, 1999.
[20]M. Sumner, B. Palethorpe, D.W.P. Thomas, P. Zanchetta, M.C.Di
Piazza, "A technique for power supply harmonic impedance
estimation using a controlled voltage disturbance", Trans. on Power
Electronics, Vol. 17, Issue 2, 2002, pp. 207 – 215.
[21]Teodorescu,R., Asiminoei,L., Blaabjerg, Borup,U.,-” A new
method of on-line grid impedance estimation for PV inverters” –
Proceedings of APEC’04, Anaheim, CA
[22]A. Timbus, R. Teodorescu, F. Blaabjerg, U. Borup. ”Online Grid
measurement and ENS detection for PV inverters running on highly
inductive grid” – IEEE Trans. On Power Electronics. Letters (in
process