IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 20, NO. 2, JUNE 2005 435
Ridethrough of Wind Turbines with Doubly-Fed
Induction Generator During a Voltage Dip
Johan Morren, Student Member, IEEE, and Sjoerd W. H. de Haan, Member, IEEE
Abstract—In this paper, a solution is described that makes it pos-
sible for wind turbines using doubly-fed induction generators to
stay connected to the grid during grid faults. The key of the solution
is to limit the high current in the rotor in order to protect the con-
verter and to provide a bypass for this current via a set of resistors
that are connected to the rotor windings. With these resistors, it is
possible to ride through grid faults without disconnecting the tur-
bine from the grid. Because the generator and converter stay con-
nected, the synchronism of operation remains established during
and after the fault and normal operation can be continued imme-
diately after the fault has been cleared. An additional feature is
that reactive power can be supplied to the grid during long dips in
order to facilitate voltage restoration. A control strategy has been
developed that takes care of the transition back to normal opera-
tion. Without special control action, large transients would occur.
Index Terms—Doubly-fed induction generator (DFIG), protec-
tion, wind power generation.
I. INTRODUCTION
T
HE worldwide concern about the environment has led
to increasing interest in technologies for generation of
renewable electrical energy. One way of generating electricity
from renewable sources is to use wind turbines. The most
common type of wind turbine is the fixed-speed wind turbine
with the induction generator directly connected to the grid.
This system has a number of drawbacks, however. The re-
active power and, therefore, the grid voltage level cannot be
controlled; the blade rotation causes power variations and,
therefore, causes voltage variations from 1 to 2 Hz in the grid.
The induction generator dynamics have resonance peaks of
approximately 10 Hz. The sensitivity to flicker is high at this
frequency [1].
Most of the drawbacks that are mentioned are avoided when
variable-speed wind turbines are used. These turbines improve
the dynamic behavior of the turbine and reduce the noise at low
wind speeds. The power production of variable-speed turbines
is higher than for fixed-speed turbines, as they can rotate at the
optimal rotational speed for each wind speed. Other advantages
of variable-speed wind turbines are that they reduce mechanical
stresses, they improve power quality, and that they compensate
for torque and power pulsations [2].
The disadvantage of the variable-speed turbine is a more com-
plex electrical system, as a power-electronic converter is needed
Manuscript received August 4, 2003; revised January 23, 2004. This work
was supported by the IOP-EMVT program of the Dutch Ministry of Economic
Affairs. Paper no. TEC-00197-2003.
The authors are with the Electrical Power Processing Group of the
Delft University of Technology, Delft 2628CD, The Netherlands (e-mail:
J.Morren@its.tudelft.nl).
Digital Object Identifier 10.1109/TEC.2005.845526
Fig. 1. Wind turbine with DFIG.
to make variable-speed operation possible. Due to the converter
that is needed, the price of variable-speed turbines tends to be
higher than constant speed turbines. A straightforward way to
obtain variable speed is to connect a converter directly between
the stator circuit of the generator and the grid. This converter has
to be designed for the rated power of the turbine. An alternative
concept, shown in Fig. 1, is a wind turbine with a doubly-fed in-
duction generator (DFIG), where the converter is connected to
the rotor windings. Compared to the turbines with the converter
connected to the stator, the DFIG has a number of advantages.
The converter is much cheaper, as the inverter rating is typically
25% of total system power, while the speed range of the gener-
ator is
33% around the synchronous speed. Also, the inverter
filters are much cheaper as they are also rated at 25% of the
total power. Further, power-factor control can be implemented
at lower cost, because the DFIG basically operates similar to a
synchronous generator [2].
A major drawback of variable-speed wind turbines, espe-
cially for turbines with DFIGs, is their operation during grid
faults. Faults in the power system, even far away from the
location of the turbine, can cause a voltage dip at the connec-
tion point of the wind turbine. The dip in the grid voltage will
result in an increase of the current in the stator windings of
the DFIG. Because of the magnetic coupling between stator
and rotor, this current will also flow in the rotor circuit and
the power-electronic converter. This can lead to the destruction
of the converter. It is possible to try to limit the current by
current-control on the rotor side of the converter; however, this
will lead to high voltages at the converter terminals, which
might also lead to the destruction of the converter.
A possible solution that is sometimes used, is to short cir-
cuit the rotor windings of the generator with so-called crowbars.
Resuming normal operation without transients when the fault is
cleared is not properly feasible with this solution, however. Most
0885-8969/$20.00 © 2005 IEEE
436 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 20, NO. 2, JUNE 2005
Fig. 2. E.On Netz requirements for wind park behavior during faults.
turbines using DFIGs therefore are automatically disconnected
from the grid nowadays, when such a fault occurs [3].
Worldwide, there is an ambition to install a large amount of
wind power and to increase the share of energy consumption that
is produced by wind turbines. The interaction with the grid be-
comes increasingly important then [4]. This can be understood
as follows. When all wind turbines would be disconnected in
case of a grid failure, these renewable generators will—unlike
conventional power plants—not be able to support the voltage
and the frequency of the grid during and immediately following
the grid failure. This would cause major problems for the sys-
tems stability [5]. It is therefore worldwide recognized that to
enable large-scale application of wind energy without compro-
mising system stability, the turbines should stay connected to the
grid in case of a failure. They should—similar to conventional
power plants—supply active and reactive power for frequency
and voltage support immediately after the fault has been cleared,
which is normally within a fraction of a second.
In Northern Germany, where the concentration of wind tur-
bines is high, the grid operator (E.On Netz) already has set re-
quirements for the behavior of wind turbines. Instead of discon-
necting them from the grid, the turbines should be able to follow
the characteristic shown in Fig. 2. Only when the grid voltage
goes below the curve (in duration or voltage level), the turbine
is allowed to disconnect. When the voltage is in the gray area,
the turbine should supply reactive power.
The behavior during grid faults has already been studied for
different types of turbines [6], [7]. When it is not possible to
keep DFIGs connected to the grid, large-scale introduction of
DFIG turbines does not seem feasible.
In this paper, a method is proposed that makes it possible for
wind turbines using DFIGs to stay connected to the grid during
grid faults. The key of the solution is to limit the high current
in the rotor and to provide a bypass for it via a set of resis-
tors that are connected to the rotor windings. With these resis-
tors, it is possible to survive grid faults without disconnection
of the turbine from the grid. Because the generator and con-
verter stay connected, the synchronism of operation remains es-
tablished during and after the fault and normal operation can be
continued immediately after the fault has been cleared. A con-
trol strategy has been developed that takes care of the transition
back to normal operation. Without this transition control, large
transients would occur.
When the dip duration is longer than a few hundred mil-
liseconds, the short-circuit resistors can be disconnected and the
system can resume normal operation at reduced grid voltage. It
can even supply reactive power to the grid during the fault.
Recently, some papers have been published that discuss the
protection of DFIGs during grid disturbances [8]–[10]. How-
ever, most papers give little information on the way the protec-
tion scheme is implemented. Further, they give only limited in-
formation on the behavior of the rotor voltage and current during
disturbances, while these signals are important during distur-
bances. Rotor currents or voltages that are too high might de-
struct the converter in the rotor circuit. In this paper, the focus is
on the rotor-side signals of the DFIG. Further, this paper is the
first that also discusses the possibility to supply reactive power
to the grid during faults in order to support voltage restoration.
First, shortly, some information is given on the modeling of
the system. Then, the controller is described. After a short dis-
cussion of voltage dips, simulation results are presented that
show the effectiveness of the protection scheme. The paper fin-
ishes with a discussion of the results and a conclusion and rec-
ommendations for further research on this topic.
II. M
ODELING OF THE
DFIG
A large number of papers describe the modeling of DFIGs
[11]–[14]. Only the most important aspects of the modeling will
be presented here. The system has been modeled and simulated
in the Simulink toolbox extension of Matlab.
A
reference frame is chosen to model the DFIG. The
model of the induction machine is based on the fifth-order two-
axis representation commonly known as the “Park model.” A
synchronously rotating
reference frame is used with the
direct
-axis oriented along the stator flux position. In this way,
decoupled control between the electrical torque and the rotor
excitation current is obtained. The reference frame is rotating
with the same speed as the stator voltage. When modeling the
DFIG, the generator convention will be used, which means that
the currents are outputs and that real power and reactive power
have a positive sign when they are fed into the grid. Using the
generator convention, the following set of equations results:
(1)
with
being the voltage (V), is the resistance , is the
current (A),
and are the stator and rotor electrical angular
velocity (rad/s), respectively, and
is the flux linkage (Vs). The
indices
and indicate the direct and quadrature axis com-
ponents of the reference frame and
and indicate stator and
rotor quantities, respectively. All quantities in (1) are functions
of time.
MORREN AND DE HAAN: RIDETHROUGH OF WIND TURBINES WITH DOUBLY-FED INDUCTION GENERATOR 437
Fig. 3. Cascade control for rotor speed.
A converter is used to connect the rotor circuit of the DFIG
to the grid, whereas the stator circuit is connected to the grid
directly. The converter must be able to transfer energy in both
directions. The grid-side converter has to control the dc-link
voltage, regardless of the magnitude and direction of the rotor
power and the rotor-side converter has to control the rotor cur-
rents.
For the converter model, it is assumed that the converters are
ideal. They exactly make the reference voltage signal that is set
by the controller. It has been shown in [13] that such a model
gives good simulation results for the simulation of voltage dips.
III. C
ONTROLLER
The electrical and mechanical dynamics of a wind turbine are
in different time scales. The electrical dynamics are much faster
than the mechanical ones. Therefore, it is possible to control
the machine in a cascade structure as shown in Fig. 3. The fast
electrical dynamics can be controlled in an inner loop and a
speed controller can be added in a much slower outer loop.
Due to the chosen orientation of the reference frame, the ac-
tive and reactive power control are decoupled. The active and
reactive power of the DFIG can be controlled by the
- and the
-axis component of the rotor current, respectively. The voltage
equations of the rotor are given in (1). Since for small values
of
the stator flux is mainly determined by the stator voltage,
it is practically constant. This implies that the derivative of the
stator flux is close to zero and can be neglected [11]. The rotor
voltage equations of (1) can then be written as [12]
(2)
The last term in both equations represents a cross-relation be-
tween the two current components. Reference voltages to obtain
the desired currents can be written as [12]
(3)
with
(4)
The
and errors are processed by a PI controller to give
and , respectively. To ensure good tracking of these currents,
the cross-related flux terms are added to
and to obtain
the reference voltages.
The internal-model-control (IMC) principle [15] has been
used to design the controllers. For a first-order system, the
controller becomes a proportional-integral (PI) controller [16]
(5)
where
is the proportional gain and is the integral gain.
Treating
and in (3) as a disturbance, the transfer
function from the rotor voltage
to the rotor current and
from the rotor voltage
to the rotor current is given by
(6)
Using the IMC principle, the current controllers become
(7)
where
is the bandwidth of the current control loop, is
the proportional gain, and
is the integral gain. The two gains
become [11]
(8)
The rotational speed is given by
(9)
It is assumed that the current controller is much faster than the
speed controller. The electrical torque is then
. The
reference torque is set to
(10)
where
is an “active damping torque” [11]. The transfer func-
tion from rotational speed to electrical torque becomes
(11)
Using again the internal model control method, the following
gains of the controller are obtained:
(12)
where
is the desired closed-loop bandwidth of the speed con-
troller. When
is chosen to be , changes in the me-
chanical torque are damped with the same time constant as the
bandwidth of the speed control loop [11]. The
-axis rotor cur-
rent directly controls the reactive power of the stator.
IV. V
OLTAGE DIP BEHAVIOR OF DFIG WITHOUT PROTECTION
A voltage dip (also the word voltage sag is used) is a sudden
reduction (between 10% and 90%) of the voltage at a point in
the electrical system, which lasts for half a cycle to 1 min [17].
There can be many causes for a voltage dip: short circuits some-
where in the grid, switching operations associated with a tem-
porary disconnection of a supply, the flow of the heavy currents
that are caused by the start of large motor loads, or large currents
drawn by arc furnaces or by transformer saturation. Voltage dips
438 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 20, NO. 2, JUNE 2005
Fig. 4. Rotor currents
i
(bottom) and
i
(top) for a voltage dip of 85%, 0.2 s
without protection.
due to short-circuit faults cause the majority of equipment trips
[18] and are therefore of most interest. Faults are either sym-
metrical (three-phase or three phase-to-ground faults) or non-
symmetrical (single-phase or double-phase or double-phase-to-
ground faults). Depending on the type of fault, the magnitudes
of the voltage dips of each phase might be equal (symmetrical
fault) or unequal (nonsymmetrical faults). The magnitude of a
voltage dip at a certain point in the system depends mainly on
the type of the fault, the distance to the fault, the system config-
uration, and the fault impedance.
The dynamics of the DFIG have two poorly damped poles
in the transfer function of the machine, with an oscillation fre-
quency close to the line frequency. These poles will cause os-
cillations in the flux if the doubly-fed induction machine is ex-
posed to a grid disturbance [11]. After such a disturbance, an
increased rotor voltage will be needed to control the rotor cur-
rents. When this required voltage exceeds the voltage limit of
the converter, it is not possible any longer to control the current
as desired.
This implies that a voltage dip can cause high induced volt-
ages or currents in the rotor circuit. The high currents might
destroy the converter, if nothing is done to protect it. In Fig. 4,
the rotor currents of the machine are shown for a voltage dip
of 85%, implying, that only 15% of the grid voltage remains.
The
-axis and -axis component of the rotor current are shown
in the figure. It can be seen that the rotor currents oscillates to
about four times the rated current. If nothing is done to protect
the converter, it will be destroyed completely.
V. V
OLTAGE DIP BEHAVIOR OF DFIG WITH PROTECTION
In this section, a description will be given of a technique that
has the objective to keep the generator connected to the grid in
case of a grid failure. With this technique, the turbine can re-
sume power generation after clearance of the fault within a few
hundred milliseconds. It is assumed that the DFIG is part of an
offshore wind farm and that the voltage dip occurs somewhere
in the 150-kV transmission grid. The DFIG is connected to this
grid through two transformers and a cable as shown in Fig. 5.
The 960-V stator voltage of the DFIG is first transformed to 34
kV, the rated voltage of the cable. Afterwards, it is transformed
Fig. 5. Simulation model setup.
Fig. 6. DFIG bypass resistors in the rotor circuit.
to 150 kV. Data of a 2.5-MW wind turbine with a DFIG have
been used during the simulations. The machine parameters, as
well as the controller parameters, can be found in the Appendix.
The key of the protection technique is to limit the high cur-
rents and to provide a bypass for it in the rotor circuit via a set
of resistors that are connected to the rotor windings (Fig. 6).
This should be done without disconnecting the converter from
the rotor or from the grid. Thyristors can be used to connect
the resistors to the rotor circuit. Because the generator and con-
verter stay connected, the synchronism of operation remains es-
tablished during and after the fault. The impedance of the by-
pass resistors is of importance but not critical. They should be
sufficiently low to avoid too large of a voltage on the converter
terminals. On the other hand, they should be high enough to
limit the current. A range of values can be found that satisfies
both conditions. In the simulation, a value of 0.86 p.u. was ap-
plied. When the fault in the grid is cleared, the wind turbine is
still connected to the grid. The resistors can be disconnected by
inhibiting the gating signals and the generator resumes normal
operation. A control strategy has been developed that takes care
of the transition to normal operation. Without special control ac-
tion, large transients would occur.
In order to show the effectiveness of the protection scheme,
the behavior of the DFIG during a voltage dip of 85% (15%
remaining voltage) and 200 ms is simulated. This dip could, for
example, be caused by a short circuit somewhere in the grid.
The resulting stator voltage is shown in Fig. 7.
In Fig. 4, the rotor current was shown for a voltage dip in a
case without protection. The same rotor current, but now with
protection, is shown in Fig. 8.
Whereas the rotor current without protection oscillates to
about four times the nominal current, it now only becomes
slightly higher than the nominal current. This current is not
flowing through the converter, but through the short-circuit
resistances. For the rotor of the generator, it is allowable that
the current becomes slightly higher than the nominal current
for a short time. The rotor has a sufficiently large thermal time
MORREN AND DE HAAN: RIDETHROUGH OF WIND TURBINES WITH DOUBLY-FED INDUCTION GENERATOR 439
Fig. 7. Stator voltage for a voltage dip of 85%, 0.2 s.
Fig. 8. Rotor currents
i
(bottom) and
i
(top) for a voltage dip of 85%, 0.2 s
with protection.
constant to cope with these currents. The voltage across the
short-circuit resistors and, thus, the voltage across the rotor and
converter terminals is shown in Fig. 9. The
-axis and -axis
components of the voltage are shown. It can be seen that the
voltage remains below the rated voltage.
When the whole turbine would be disconnected from the grid,
it can become difficult to control the mechanical rotation of the
wind turbine, as it is not possible any longer to develop an elec-
trical torque to counteract the mechanical torque provided by
the wind power. With the bypass resistors connected to the rotor,
the turbine stays connected to the grid, and it is still possible to
develop an electrical torque. The rotational speed of the turbine
during the dip is shown in Fig. 10.
When the dip is cleared, the wind turbine should resume
normal operation. Special control action should be taken,
however, because otherwise large transients can occur. These
transients occur when the current controllers resume their
operation without taking into account the occurrence of the dip.
Due to the mismatch between the signals that are expected by
the controllers and the real signals, windup of the integrators
will occur, resulting in large transients when normal operation
is resumed. In order to prevent these transients, a soft transition
to normal operation should be made. To get this transition, the
reference values are set to the actual values of the currents at
Fig. 9. Rotor voltages
v
(bottom) and
v
(top) for a voltage dip of 85%, 0.2 s
with protection.
Fig. 10. Mechanical speed for a voltage dip of 85%, 0.2 s with protection.
Fig. 11. Reactive power supply for a voltage dip of 40%, 2 s with protection.
the moment the fault is cleared. The reference values are then
slowly changed to the reference values that are needed to obtain
the required behavior of the turbine.
When the dip holds on for a longer time, it can be required
that the generator supplies reactive power. The proposed protec-
tion scheme also offers the possibility to supply reactive power
during the dip and immediately after the dip as shown in Fig. 11.
