IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 27, NO.

2, MAY 2012 713

Advanced Control Strategy of DFIG Wind Turbines

for Power System Fault Ride Through
Lihui Yang, Zhao Xu, Member, IEEE, Jacob Østergaard, Senior Member, IEEE,
Zhao Yang Dong, Senior Member, IEEE, and Kit Po Wong, Fellow, IEEE

Abstract—This paper presents an advanced control strategy for power grid connection codes in most countries require that WTs
the rotor and grid side converters of the doubly fed induction gen- should remain connected to the grid to maintain the reliability
erator (DFIG) based wind turbine (WT) to enhance the low-voltage during and after a short-term fault [1]. The ability of WT to
ride-through (LVRT) capability according to the grid connection
requirement. Within the new control strategy, the rotor side con-
stay connected to the grid during voltage dips is termed as the
troller can convert the imbalanced power into the kinetic energy of low-voltage ride-through (LVRT) capability.
the WT by increasing its rotor speed, when a low voltage due to a In order to fulfill the LVRT requirement for DFIG WTs, there
grid fault occurs at, e.g., the point of common coupling (PCC). The are two major issues to be addressed properly under a fault
proposed grid side control scheme introduces a compensation term condition. The first one is the over-current that can occur in
reflecting the instantaneous DC-link current of the rotor side con- rotor and stator circuits, while the second one is the DC-link
verter in order to smooth the DC-link voltage fluctuations during
the grid fault. A major difference from other methods is that the over-voltage. Both can be attributed to the excessive energy that
proposed control strategy can absorb the additional kinetic en- cannot be transmitted into the grid during the faults. As the
ergy during the fault conditions, and significantly reduce the os- power electronics converters in the DFIG system have relative
cillations in the stator and rotor currents and the DC bus voltage. lower power rating compared with the WT with fully rated con-
The effectiveness of the proposed control strategy has been demon- verters, the LVRT capability of the DFIG WT is of special in-
strated through various simulation cases. Compared with conven-
tional crowbar protection, the proposed control method can not
terest with respect to the stability issue of such system. Quite a
only improve the LVRT capability of the DFIG WT, but also help few studies have been carried out to improve the LVRT capa-
maintaining continuous active and reactive power control of the bility of the DFIG WT [3]–[19].
DFIG during the grid faults. Among the available control strategies, the crowbar protec-
Index Terms—Doubly fed induction generator (DFIG), low tion is the mostly used. During the faults, the rotor side con-
voltage ride through, power system fault, wind turbine. verter will be blocked, and the crowbar circuit, installed across
the rotor terminals, will be triggered to damp the over-current
in the rotor circuit. Consequently, the generator operates as a
I. INTRODUCTION conventional induction machine, which absorbs reactive power
from the faulted grid [2]. The chopper circuit, with a resistor
D OUBLY fed induction generator (DFIG) is a popular wind
turbine (WT) system due to its high energy efficiency,
reduced mechanical stress on the WT, and relatively low power
across the DC bus, is usually used along with the crowbar to
smooth the DC-link voltage by dissipating the excessive power
rating of the connected power electronics converter of low costs. over the DC bus [3], [4].
With increasing penetration level of WTs into the grid, the wind Based on the conventional crowbar protection, some im-
proved crowbar solutions have been proposed to enhance the
Manuscript received December 15, 2010; revised May 05, 2011 and July
LVRT performance of the DFIG WT [5]–[10]. Although the
27, 2011; accepted October 24, 2011. Date of publication December 23, 2011; crowbar circuits are able to protect the machine and the con-
date of current version April 18, 2012. This work was supported by the Hong verter during the faults, the controllability of the rotor converter
Kong Polytechnic University via projects A-PJ81 and 4-ZZ7Q, State Key Lab-
oratory of Electrical Insulation and power Equipment (EIPE11309), and the
with respect to the active and reactive power of the DFIG is
Fundamental Research Funds for the Central Universities, China. Paper no. temporarily lost. Moreover, the usage of crowbar and chopper
TPWRS-00989-2010. actually installs extra hardware in the DFIG that can increase
L. Yang is with the State Key Laboratory of Electrical Insulation and the costs and decrease the system reliability.
Power Equipment, Xi’an Jiaotong University, Xi’an 710049, China (e-mail:
lihui.yang@mail.xjtu.edu.cn). Considering these drawbacks, some researchers have pro-
Z. Xu is with the Department of Electrical Engineering, Hong Kong Poly- posed new solutions for reducing the inrush currents in the
technic University, Hong Kong (e-mail: eezhaoxu@polyu.edu.hk). rotor as well as the DC-link over-voltage during the faults, by
J. Østergaard is with the Centre for Electric Technology, Department of Elec-
trical Engineering, Technical University of Denmark, DK-2800 Lyngby, Den- means of designing more advanced control strategies for the
mark (e-mail: joe@elektro.dtu.dk). rotor and grid side converters [11]–[19]. However, some of
Z. Y. Dong is with the Centre for Intelligent Electricity Networks, The Uni- these algorithms are too complicated to implement in industrial
versity of Newcastle, Newcastle, Australia (e-mail: Joe.Dong@newcastle.edu.
au).
applications and depend strongly on the proper design of the
K. P. Wong is with the School of Electrical, Electronic and Computer control parameters or the estimation of certain parameters,
Engineering, The University of Western Australia, Perth, Australia (e-mail: which may have adverse effects on its robustness.
kitpo@ieee.org).
Color versions of one or more of the figures in this paper are available online
This paper presents an innovative control strategy for both
at http://ieeexplore.ieee.org. the rotor and grid side converters to enhance the LVRT capacity
Digital Object Identifier 10.1109/TPWRS.2011.2174387 of the DFIG WT, without the need of additional current and
0885-8950/$26.00 © 2011 IEEE
714 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 27, NO. 2, MAY 2012

where and are the stator

and rotor current vectors, respectively; and
are the stator and rotor voltage vectors,
respectively; and are the
stator and rotor flux vectors, respectively; , , and are
the base, stator, and rotor angular frequencies, respectively.
The system model is written in the p.u. system with the time,
, in seconds.

B. Drive Train
When studying the dynamic stability of the DFIG WT, the
Fig. 1. Schematic diagram of the DFIG WT system. two-mass model of the drive train is important due to the wind
turbine shaft is relatively softer than the typical steam turbine
shaft in conventional power plants [2]. The equations which
voltage protections. The key idea is to increase the generator represent the two-mass model of the drive train are expressed
rotor speed through proper control of the rotor side converter as
during a grid voltage dip. Different from conventional methods,
the proposed control strategy is based on a simple concept to
(2)
transform the unbalanced energy into the kinetic one, rather than
being dissipated otherwise. For the grid side control scheme, a
compensation item, which reflects the variation of the DC-link (3)
current of the rotor side converter, is added during the fault to
smooth the fluctuations of the DC-link voltage. Compared with (4)
the LVRT solution in [19], the proposed control strategy for the
grid side controller can effectively reduce the extremely high where is the wind turbine speed. and [SI unit (s)]
transient when the stator voltage dips to zero during faults. Sim- are the generator and turbine inertia constants, respectively.
ulation studies using Matlab/Simulink have been conducted on is the friction coefficient of the generator. is the shaft twist
a 1.5-MW DFIG WT to validate the effectiveness of the pro- angle, which is in radian (rad.). The electromagnetic torque ,
posed control strategy. the shaft torque , and the mechanical torque , which is
the torque input of the wind turbine, are
II. MODELING OF THE DFIG WIND TURBINE
The schematic diagram of a grid-connected DFIG WT system (5)
is shown in Fig. 1. The DFIG WT system, including the wind (6)
turbine, the drive train, the induction generator, the back-to-back
PWM converters, and the control system, is connected to the (7)
grid through a transformer. The control system consists of two
control levels including the WT control and the DFIG control. where is the air density, is the turbine radius, is the pitch
The WT level controls the output mechanical power of the wind angle, is the wind speed, is the power coefficient, and
turbine through the pitch angle and generates the reference value
for the rotor speed of the DFIG based on the measured wind
(8)
speed and optimum power-speed characteristic curve. A two-
stage control strategy is used to implement the power optimiza-
(9)
tion strategy below the rated wind speed, and the power limita-
tion strategy above the rated wind speed [20]. The DFIG control
level, including the rotor and grid side controllers, is to control
where is the blade tip speed ratio, has
the active and reactive power of the DFIG using the vector con-
a maximum value for the optimal tip speed ratio
trol technique. In Sections II-A–II-C, a detailed dynamic model
and optimized pitch angle . The wind turbine control is
of the DFIG WT will be presented.
achieved by driving the generator/turbine speed along the op-
A. Generator timum power-speed characteristic curve, illustrated in Fig. 2
[20], which corresponds to the maximum energy capture from
The voltage equations of the stator and rotor circuits of the in- the wind. In this curve, when the wind speed is larger than the
duction generator can be given in a - reference frame rotating cut-in wind speed and less than the lower limit
at the synchronous speed [21], [22] (zone A–B), the generator rotor speed reference for the rotor
side controller is set as the minimal value ,
, to ensure the generator slip smaller than 0.3. When
(1) the wind speed is between the lower limit and the rated value
(zone B–C), the DFIG is operated in the variable speed
mode, where is kept constant to (usually ), while
YANG et al.: ADVANCED CONTROL STRATEGY OF DFIG WIND TURBINES FOR POWER SYSTEM FAULT RIDE THROUGH 715

power from the WT, leading to significantly increased fluctua-

tions of the currents and voltages in the DFIG system. In order
to protect the power electronics devices that are sensitive to
over-currents and over-voltages, the purpose of introducing the
LVRT strategy is to ensure the DFIG can stay connected to the
faulted grid by effectively limiting the stator and rotor circuit
currents as well as the DC-link voltage.
The conventional crowbar control realizes the LVRT protec-
tion by blocking the rotor side converter circuit and introducing
additional rotor winding resistances that actually dissipate the
produced energy. In contrast, the proposed control strategy can
transform the additional output power into the WT kinetic en-
Fig. 2. Two static curves used in the design of the DFIG WT. (a) Mechanical
power versus wind speed. (b) Electrical power versus generator speed.
ergy by temporarily increasing the generator rotor speed during
the grid faults, thus effectively limiting the oscillations in the
currents. If the rotor speed increases above the rated value, the
pitch control will be triggered to decrease the power extraction
from the wind. This can restrict the over-speed of the rotor and
therefore limit the excessive mechanical stress applied to the
turbine system. Moreover, a compensation item, which reflects
the variation of the DC-link current of the rotor side converter,
is added to the grid side control scheme during the fault so as to
Fig. 3. Schematic diagram of the pitch control. smooth the fluctuation of the DC-link voltage.
The proposed control strategy makes full use of existing
resources within the DFIG WT system to realize the LVRT
is tuned to over different wind speeds by adapting the without the need of additional components like the crowbar.
generator speed to , which can be obtained by substituting It can effectively protect the power electronics devices against
into (7), expressed by [23] current and voltage fluctuations during grid faults to ensure
the LVRT of the DFIG. Several advantages can be achieved by
(10) using the new control strategy. From the energy perspective,
the new control strategy can absorb the additional output power
during the faults that will be otherwise dissipated in the crowbar
where is the optimal constant of
based control. Notably, the increased kinetic energy of the WT
wind turbine. Therefore, can be maintained at and the
can be released into the grid slowly after the fault clearance.
maximal power is achieved by tracking in (10). When the
From the control perspective, the new control strategy can era-
wind speed is higher than the rated value (zone C-D), is set
sure the back-to-back converters well connected to the DFIG
as the rated value of the generator speed , the over rated
as well as the grid without loss of the controllability during
turbine power production will be restrained by the pitch control,
the faults. This potentially can enable both active and reactive
which can therefore limit the over-speed of the generator.
supports to the faulted grids from WT which is difficult for
C. Pitch Control the crowbar based control. The detailed control scheme will
be illustrated in Sections III-B and III-C, and the simulation
The pitch angle of the blade is controlled to optimize the validation will be given in Section IV.
power extraction of the WT as well as to prevent over rated
power production in high wind. The pitch servo is modeled as
B. Control of Rotor Side Converter (RSC)
(11)
In normal operation, the control scheme of the RSC is illus-
trated in Fig. 4. In order to decouple the electromagnetic torque
The pitch control scheme is shown in Fig. 3 [2]. When the and the rotor excitation current, the induction generator is con-
generator speed exceeds , the pitch control is active and trolled in the stator-flux oriented reference frame, which is a
the pitch angle is tuned so that the turbine power can be re- synchronously rotating reference frame, with its -axis oriented
stricted to its rated value. along the stator-flux vector position (in this paper, the stator-flux
vector is calculated using ) [23]. The typical proportional-in-
III. PROPOSED LOW-VOLTAGE RIDE-THROUGH STRATEGY tegral (PI) controllers are used for regulation in the rotor speed
and reactive power (outer) control loops as well as the rotor cur-
A. Overview of the Proposed Control Strategy rent (inner) control loops. The values of , generated by the
When the power grid is subject to, e.g., a short-circuit fault, wind turbine control level, at different operation conditions have
the bus voltage at the PCC drops, and therefore introducing un- been defined in Section II.
desirable transients in the stator and rotor currents. The low When a short-term low-voltage fault occurs, the incoming
voltage also prevents the full transmission of generated active power from the wind and the power flowing into the grid are
716 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 27, NO. 2, MAY 2012

Fig. 4. Control scheme of the rotor side converter in normal operation. Fig. 6. Control scheme of the grid side converter in normal operation.

to the proposed control strategy can be little to the safe operation

of WTs. This analysis will be verified by the simulation results
in Section IV.

C. Control of Grid Side Converter (GSC)

Fig. 6 shows the control scheme of the GSC in normal opera-
tion, where and are the grid-side converter voltage vector
and the grid-side inductor current vector, respectively. In order
to obtain the independent control of active and reactive power
Fig. 5. Control scheme of the rotor side converter during a grid fault. flowing between the grid and the GSC, the converter control op-
erates in the grid-voltage oriented reference frame, which is a
synchronously rotating reference frame, with its -axis oriented
imbalanced instantaneously, resulting in the transient excessive along the grid-voltage vector position [23]. Similarly, the typ-
currents in the rotor and stator circuits. Therefore from the en- ical PI controllers are used for regulation in the DC-link voltage
ergy balance point of view, the key point of suppressing the (outer) control loop and the grid side inductor current (inner)
over-currents in the rotor and stator circuits is to reduce the im- control loops.
balanced energy flowing through the DFIG WT system. The equation which describes the dynamic behavior of the
When at least one of the monitored parameters, including the DC-link capacitor can be expressed as
rotor current, stator current, DC-link voltage, and grid voltage,
exceeds its respective protection setting due to the grid fault, (12)
the proposed LVRT control strategy will be triggered. With this
control strategy, the rotor side controller will increase the gener- where and are the DC-link currents of the grid and rotor
ator rotor speed by reducing the generator torque to zero during side converters, respectively. Neglecting the losses in the in-
the fault, in order to absorb and convert the incoming energy ductor resistances and power electronics converters, we have
from the wind into the kinetic energy in the WT inertia. The in-
creased kinetic energy can be transformed and released into the
grid after the fault clearance. The control scheme of the RSC
against grid faults is illustrated in Fig. 5.
The parameters of the PI controllers are tuned using the pole (13)
placement methodology [24] in order to ensure the rising times
of the inner and outer control loops are less than 10 ms and 50 where is the magnitude of the active power of the generator
ms, respectively; and the maximum deviations of the inner and rotor.
outer control loops are both less than 20%. In normal operation, when the power flowing through the grid
The new control strategy will not cause excessive mechan- and rotor side converters is balanced, is equal to , so the
ical stress to the WT system. When the DFIG WT is operating DC-link voltage is constant. When the grid voltage dips, may
at or close to the rated speed, the acceleration due to the pro- not be equal to due to the instantaneous unbalanced power
posed control scheme during the fault may accelerate the wind flow between the grid and rotor side converters, and therefore
turbine speed above its rated value. This can increase the thrust the DC-link voltage may fluctuate. In [19] in order to reduce the
and centrifugal forces, applied to the rotor construction that may fluctuation of the DC-link voltage, the item reflecting
endanger the wind turbine mechanical system [2]. However, the the instantaneous variation of the output power of the rotor side
over-speed of the WT can be effectively restrained by the pitch controller is directly set as the reference of the during the
control, which will be activated immediately when the rotor grid fault. However, the stator voltage may reduce to zero
speed becomes higher than the rated value. Moreover, the du- during the grid fault. This will introduce an extremely high tran-
ration of the over-speed is short, due to the short duration and sient value of .
rare occurrence of grid faults. According to the expressions of In this paper, when the rotor current, stator current, DC-link
the thrust and centrifugal forces in [2], the negative impacts due voltage, or grid voltage exceeds its respective relay setting due
YANG et al.: ADVANCED CONTROL STRATEGY OF DFIG WIND TURBINES FOR POWER SYSTEM FAULT RIDE THROUGH 717

Fig. 9. DFIG WT with the crowbar and DC chopper protection circuits.

Fig. 7. Control scheme of the grid side converter during a grid fault.

Fig. 8. Single line diagram for the studied system.

to the grid voltage dip, the term describing is rep-

resented as a disturbance to compensate the instantaneous rotor
power in the control scheme. In such a way, the can be regu-
lated smoothly during a grid fault. The detailed control scheme
of the GSC during the grid fault is shown in Fig. 7.
Similarly, the pole placement methodology is applied to de-
sign the PI controller parameters of the grid side controller so
as to obtain the same rising time and maximum deviation as the
rotor side controller.

IV. ANALYSIS OF LOW-VOLTAGE RIDE-THROUGH STRATEGIES

The DFIG WT under study is connected to the grid transmis-
sion level via a radial link as shown in Fig. 8. In order to evaluate
the proposed control strategy, the complete DFIG WT system
model has been developed and simulated in Matlab/Simulink
environment. The components of the simulation model are built Fig. 10. Simulated transient responses of the studied DFIG WT at wind speed
with standard electrical component blocks from the SimPow- of 13 m/s with the control strategy B and C during a 100-ms fault. (a) DC-link
erSystems block in Matlab/Simulink library. The parameters voltage. (b) Active power to the grid. (c) Magnitude of the rotor current.
of the studied DFIG WT, which are mostly based on existing
information from industries and literatures, are listed in the
Appendix. A. Comparing the LVRT Behaviors With Strategy B and C
Three different control strategies of LVRT on the perfor- At time , a three-phase short-circuit fault with a du-
mance of the studied DFIG WT during different symmetrical ration of 100 ms occurs in the transmission power grid and the
three-phase short-circuit faults are investigated to study their PCC voltage drops to about 0.2 p.u. The simulated transient be-
effects, including haviors of the studied DFIG WT operating at the wind speed of
Strategy A: The conventional protective system equipped 13 m/s and installed with the control strategy B and C are shown
with crowbar and DC-link chopper. in Fig. 10.
Strategy B: The proposed control strategy of RSC only. Fig. 10 shows that with the strategy B, the fluctuation of the
Strategy C: The proposed control strategy of RSC and GSC. DC-link voltage is not damped sufficiently. As the rotor current
The DFIG WT with the crowbar and DC chopper protection control might be affected by the DC-link voltage fluctuation, the
circuits of the strategy A is shown in Fig. 9. The crowbar circuit transient performances of active power and rotor current con-
includes the controllable switches and resistors. The crowbar trolled by the strategy B are worse than with the strategy C.
resistors are all chosen as . The DC chopper resistor is se- Due to the additional grid side control scheme in the strategy
lected as 0.5 p.u. The protection thresholds of rotor and stator C, the fluctuation of DC-link voltage has been effectively re-
currents are both set to 1.5 p.u. The detailed scheme of the duced and the LVRT behavior of the DFIG WT is better than
crowbar protection is described in [2]. with the strategy B. This is because when the wind speed is
718 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 27, NO. 2, MAY 2012

13 m/s, which is higher than the rated wind speed, the DFIG
WT will be operating at the rated rotor speed accordingly. The
over-speed of the WT caused by the proposed control strategy
of the RSC can be effectively restrained by the pitch control.
This will limit the transformation of the electric energy into the
kinetic one. So when the pitch control is triggered, introducing
the compensation term in the grid side controller can
help reducing the over-voltage on the DC bus. Hence the neces-
sity of including the GSC in the proposed control strategy has
been justified in order to achieve a better LVRT performance of
the DC-link voltage during the grid faults.

B. LVRT Behaviors at Low Wind Speed

In the following analyses, in order to meet the ride-through
requirement of a widely referred grid code [25], two typical
three-phase short-circuit faults are imposed on the transmission
power grid at , the PCC voltage drop to 0 for 150 ms
and to 15% of its nominal value for 625 ms, respectively. The
LVRT capability of the proposed control strategy C is simulated
and compared with the performance of the conventional crowbar
protection based the strategy A.
The LVRT behaviors of the studied DFIG WT operating at
wind speed of 8 m/s, when the voltage at the PCC drops to 0
for 150 ms are shown in Fig. 11. During the fault, the acceler-
ation of the generator rotor speed with the control strategy C is
faster than the cases with the strategy A. The active accelera-
tion of the rotor speed in the strategy C can transform the ad-
ditional energy into the kinetic one; therefore, the fluctuations
of the active power, reactive power, stator and rotor current, and
DC-link voltage are significantly reduced than the case of the
crowbar based strategy A. For example, the rotor current in the
case with the strategy A shoots up to 2.2 p.u. at the instant of
the grid fault, while the proposed control strategy C can sup-
press the rotor current no more than 0.96 p.u., which is below
the activation threshold of the crowbar. Therefore, the proposed
control strategy can enable the WT to ride through this fault
even without triggering the crowbar protection. It is also ob-
served that after the fault clearance, the rotor speed decreases
back to the reference value and the oscillations are well damped
under the proposed control strategy C.
Fig. 12 shows the LVRT behaviors of the studied DFIG WT
operating at wind speed of 8 m/s, when the voltage at the PCC
drops to 15% for 625 ms. This illustrates that with the pro-
posed control strategy C, the DFIG WT can remain connected to
the grid during a longer duration of fault without triggering the
crowbar protection and the LVRT behaviors of the active power,
reactive power, stator and rotor current, and DC-link voltage are
better than the case with the strategy A.

C. LVRT Behaviors at High Wind Speed

Fig. 13 presents the LVRT behaviors of the studied DFIG WT
operating at wind speed of 13 m/s with the conventional crowbar
control based strategy A and with the new control strategy C, Fig. 11. Simulated transient responses of the studied DFIG WT at wind speed
when the voltage at the PCC drops to 0 for 150 ms. of 8 m/s with the control strategies A (crowbar) and C (the newly proposed
When the DFIG WT is operating at the wind speed of 13 m/s, LVRT strategy), when the voltage at the PCC drops to 0 for 150 ms. (a) Voltage
at the DFIG terminal. (b) Generator rotor speed. (c) Active power to the grid. (d)
which is higher than the rated wind speed, the DFIG WT is Reactive power to the grid. (e) Magnitude of the stator current. (f) Magnitude
operating at the rated rotor speed. The acceleration of the rotor of the rotor current. (g) DC-link voltage.
YANG et al.: ADVANCED CONTROL STRATEGY OF DFIG WIND TURBINES FOR POWER SYSTEM FAULT RIDE THROUGH 719

Fig. 12. Simulated transient responses of the studied DFIG WT at wind speed Fig. 13. Simulated transient responses of the studied DFIG WT at wind speed
of 8 m/s with the control strategies A (crowbar) and C (the newly proposed of 13 m/s with the control strategies A (crowbar) and C (the newly proposed
LVRT strategy), when the voltage at the PCC drops to 15% of its nominal LVRT strategy), when the voltage at the PCC drops to 0 for 150 ms. (a) Active
value for 625 ms. (a) Voltage at the DFIG terminal. (b) Generator rotor speed. power to the grid. (b) Reactive power to the grid. (c) Magnitude of the stator
(c) Active power to the grid. (d) Reactive power to the grid. (e) Magnitude of current. (d) Magnitude of the rotor current. (e) DC-link voltage. (f) Generator
the stator current. (f) Magnitude of the rotor current. (g) DC-link voltage. rotor speed. (g) Pitch angle. (h) Shaft torque.
720 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 27, NO. 2, MAY 2012

speed due to the proposed control scheme during the fault will
make the rotor speed above its rated value, which will trigger
the pitch control immediately. This can limit the transforma-
tion of the electric energy into the kinetic one. Therefore in
Fig. 13(e), (f), and (h), the differences of the DC-link voltage,
the rotor speed, and shaft torque during the fault between the
strategies A and C are smaller than the case when the DFIG WT
operating at low wind speed. However, the transient behaviors
of the active and reactive power, stator and rotor current, and
pitch angle in the case of the proposed control strategy C are still
observed better than with the conventional crowbar protection.
Especially, the proposed control strategy can suppress the tran-
sient stator and rotor currents lower than the crowbar threshold;
thus, the DFIG can maintain uninterrupted control of active and
reactive power, potentially enabling the grid supporting services
during the fault.
Fig. 14 shows the LVRT behaviors of the studied DFIG WT
operating at wind speed of 13 m/s, when the voltage at the PCC
drops to 15% for 625 ms. This illustrates that with the proposed
control strategy C, the DFIG WT operating at high wind speed
can remain connected to the grid during a longer duration of
fault without triggering the crowbar protection and has better
LVRT behaviors of the active power, reactive power, stator and
rotor current, DC-link voltage, pitch control, and shaft torque
than the case with conventional crowbar protection.
In Figs. 13(f) and 14(f), the rotor speed slightly exceeds the
rated value. However, the maximum over-speed is only 2.72%
and the period of the whole violation is 230 ms when the voltage
at the PCC drops to 0 for 150 ms; the maximum over-speed is
only 6.06% and the period of the whole violation is 1.5 s when
the voltage at the PCC drops to 15% of its nominal value for
625 ms. The corresponding increase of the thrust and centrifugal
forces due to the over-speed is little according to their expres-
sions [2], and the over-speed duration is short, indicating a very
little impact on the wind turbine construction, which verifies the
analysis in Section III.

V. CONCLUSIONS
High penetration of WTs imposes a significant challenge to
the safe operation of power systems. To ensure the security of
electricity supply with substantial wind power, the WTs must
ride through and even contribute to supporting the grid operation
under the fault conditions.
This paper proposes a new and efficient control strategy for
both the rotor and grid side converters to improve the LVRT
capability of the DFIG WT. The new control strategy enables
the DFIG to continue the electricity production, and absorb the
excessive energy by increasing the generator rotor speed tem-
porarily when a fault occurs at, e.g., the PCC. The new strategy
also introduces a compensation item to the grid side controller
in order to suppress the DC-link over-voltage during the faults.
The simulation results show that the proposed control strategy
is able to effectively suppress the transients in the rotor circuit Fig. 14. Simulated transient responses of the studied DFIG WT at wind speed
current and the DC-link voltage. Compared with the conven- of 13 m/s with the control strategies A (crowbar) and C (the newly proposed
tional crowbar protection, the DFIG WT installed with the pro- LVRT strategy), when the voltage at the PCC drops to 15% of its nominal value
for 625 ms. (a) Active power to the grid. (b) Reactive power to the grid. (c)
posed control strategy gives a better transient behavior in event Magnitude of the stator current. (d) Magnitude of the rotor current. (e) DC-link
of short-term grid voltage dip. By using the new control strategy, voltage. (f) Generator rotor speed. (g) Pitch angle. (h) Shaft torque.
YANG et al.: ADVANCED CONTROL STRATEGY OF DFIG WIND TURBINES FOR POWER SYSTEM FAULT RIDE THROUGH 721

little impact will be resulted to the WT mechanical construc- [13] A. Hansen and G. Michalke, “Fault ride-through capability of DFIG
tion and the occurrence of the crowbar interruption can also be wind turbines,” Renew. Energy, vol. 32, no. 9, pp. 1594–1610, Jul.
2007.
minimized. [14] J. Lopez, P. Sanchis, X. Roboam, and L. Marroyo, “Dynamic behavior
of the doubly fed induction generator during three-phase voltage dips,”
IEEE Trans. Energy Convers., vol. 22, no. 3, pp. 709–717, Sep. 2007.
APPENDIX [15] L. Xu and P. Cartwright, “Direct active and reactive power control of
DFIG for wind energy generation,” IEEE Trans. Energy Convers., vol.
The parameters of the studied DFIG WT are as follows: 21, no. 3, pp. 750–758, Sep. 2007.
[16] M. Rahimi and M. Parniani, “Transient performance improvement of
Wind turbine: cut-in wind speed: 4 m/s; lower limit of the wind turbines with doubly fed induction generators using nonlinear
wind speed: ; rated wind speed: 12 m/s; inertia control strategy,” IEEE Trans. Energy Convers., vol. 25, no. 2, pp.
constant: ; damping coefficient: ; 514–525, Jun. 2010.
[17] F. K. A. Lima, A. Luna, P. Rodriguez, E. H. Watanabe, and F. Blaab-
shaft stiffness coefficient: ; time constant of the
jerg, “Rotor voltage dynamics in the doubly fed induction generator
pitch servo: . during grid faults,” IEEE Trans. Power Electron., vol. 25, no. 1, pp.
DFIG: rated power: 1.5 MW; rated voltage: 575 V; rated cur- 118–130, Jan. 2010.
rent; 1505 A; rated rotor speed: (with the [18] J. Liang, W. Qiao, and R. G. Harley, “Feed-forward transient current
control for low-voltage ride-through enhancement of DFIG wind tur-
synchronous speed as the base value); inertia constant: bines,” IEEE Trans. Energy Convers., vol. 25, no. 3, pp. 836–843, Sep.
; friction coefficient: ; stator resistance: 2010.
; rotor resistance: ; stator [19] J. Yao, H. Li, Y. Liao, and Z. Chen, “An improved control strategy of
leakage inductance: ; rotor leakage inductance: limiting the DC-link voltage fluctuation for a doubly fed induction wind
generator,” IEEE Trans. Power Electron., vol. 23, no. 3, pp. 1205–1213,
; mutual inductance: . May 2008.
Converters: resistance of grid side inductor: [20] A. D. Hansen, P. Sørensen, F. Iov, and F. Blaabjerg, “Control of vari-
; inductance of grid side inductor: ; able speed wind turbines with doubly-fed induction generators,” Wind
Eng., vol. 28, no. 4, pp. 411–434, 2004.
DC-link capacitor: . [21] F. Mei and B. C. Pal, “Modal analysis of grid-connected doubly fed
induction generators,” IEEE Trans. Energy Convers., vol. 22, no. 3, pp.
728–736, Sep. 2007.
REFERENCES [22] F. Wu, X. P. Zhang, K. Godfrey, and P. Ju, “Small signal stability anal-
ysis and optimal control of a wind turbine with doubly fed induction
[1] A. Thomas, Wind Power in Power Systems. New York: Wiley, 2005. generator,” IET Gen., Transm., Distrib., vol. 1, no. 5, pp. 751–760, Sep.
[2] V. Akhmatov, Induction Generators for Wind Power. Brentwood, 2007.
CA: Multi-Science, 2005. [23] R. Pena, J. C. Clare, and G. M. Asher, “Doubly fed induction gener-
[3] I. Erlich, J. Kretschmann, J. Fortmann, S. Mueller-Engelhardt, and H. ator using back-to-back PWM converters and its application to variable
Wrede, “Modeling of wind turbines based on doubly-fed induction speed wind-energy generation,” Proc. Inst. Elect. Eng., Elect. Power
generators for power system stability studies,” IEEE Trans. Power Appl., vol. 143, no. 3, pp. 231–241, May 1996.
Syst., vol. 22, no. 3, pp. 909–919, Aug. 2007. [24] J. H. Chow, “A pole-placement design approach for systems with mul-
[4] J. Yang, J. E. Fletcher, and J. O’Reilly, “A series dynamic resistor based tiple operating conditions,” IEEE Trans. Autom. Control, vol. 35, no.
converter protection scheme for doubly-fed induction generator during 2, pp. 278–288, Feb. 1990.
various fault conditions,” IEEE Trans. Energy Convers., vol. 25, no. 2, [25] M. Tsili and S. Papathanassiou, “A review of grid code technical re-
pp. 422–432, Jun. 2010. quirements for wind farms,” IET Renew. Power Gen., vol. 3, no. 3, pp.
[5] A. H. Kasem, E. F. EI-Saadany, H. H. EI-Tamaly, and M. A. A. Wahab, 308–332, Sep. 2009.
“An improved fault ride-through strategy for doubly fed induction gen-
erator-based wind turbines,” IET Renew. Power Gen., vol. 2, no. 4, pp.
201–214, Dec. 2008.
[6] S. Foster, L. Xu, and B. Fox, “Coordinated reactive power control for
facilitating fault ride through of doubly fed induction generator- and
fixed speed induction generator-based wind farms,” IET Renew. Power Lihui Yang received the Ph.D. degree in electrical
Gen., vol. 4, no. 2, pp. 128–138, Mar. 2008. engineering from Xi’an Jiaotong University, Xi’an,
[7] L. Peng, B. Francois, and Y. Li, “Improved crowbar control strategy of China, in 2010.
During 2008–2009, she was a visiting Ph.D.
DFIG based wind turbines for grid fault ride-through,” in Proc. IEEE
student with the Centre for Electric Technology,
24th Annu. Applied Power Electronics Conf. Expo., Washington, DC,
Technical University of Denmark, Lyngby, Denmark.
Feb. 2009, pp. 1932–1938.
She was a Research Assistant with Hong Kong Poly-
[8] M. B. C. Salles, J. R. Cardoso, A. P. Grilo, C. Rahmann, and K. technic University during June–November 2010.
Hameyer, “Control strategies of doubly fed induction generators to She is currently an Assistant Professor with Xi’an
support grid voltage,” in Proc. IEEE Int. Electric Machines and Drives Jiaotong University. Her research interests include
Conf., May 2009, pp. 1551–1556. stability and control of wind power generation.
[9] J. López, E. Gubía, E. Olea, J. Ruiz, and L. Marroyo, “Ride through
of wind turbines with doubly fed induction generator under symmet-
rical voltage dips,” IEEE Trans. Ind. Electron., vol. 56, no. 10, pp.
4246–4254, Oct. 2009.
Zhao Xu (S’00–M’06) received the Ph.D. degree
[10] L. G. Meegahapola, T. Littler, and D. Flynn, “Decoupled-DFIG fault in electrical engineering from The University of
ride-through strategy for enhanced stability performance during grid Queensland, Brisbane, Australia, in 2006.
faults,” IEEE Trans. Sustain. Energy, vol. 1, no. 3, pp. 152–162, Oct. From 2006–2009, he was an Assistant and later
2010. Associate Professor with the Centre for Electric
[11] D. Xiang, L. Ran, P. J. Tavner, and S. Yang, “Control of a doubly-fed Technology, Technical University of Denmark,
induction generator in a wind turbine during grid fault ride-through,” Lyngby, Denmark. Since 2010, he has been with
IEEE Trans. Energy Convers., vol. 21, no. 3, pp. 652–662, Sep. 2006. Hong Kong Polytechnic University. His research in-
[12] M. Rathi and N. Mohan, “A novel robust low voltage and fault ride terests include demand side, grid integration of wind
through for wind turbine application operating in weak grids,” in Proc. power, electricity market planning and management,
IEEE Industrial Electronics Society Conf., Nov. 2005, pp. 6–10. and AI applications.
722 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 27, NO. 2, MAY 2012

Jacob Østergaard (M’95,–SM’09) is Professor Kit Po Wong (M’87–SM’90–F’02) He received the

and Head of the Centre for Electric Technology, M.Sc, Ph.D., and higher doctorate D.Eng degrees
Department of Electrical Engineering, Technical from the University of Manchester, Institute of
University of Denmark, Lyngby, Denmark. His Science and Technology, Manchester, U.K., in 1972,
research interests include integration of renewable 1974, and 2001, respectively.
energy, control architecture for future power system, He was with The University of Western Australia,
and demand side. Perth, Australia, from 1974 until 2004 and presently
Prof. Østergaard is serving in several professional an Adjunct Professor there. From 2002 to 2011, he
organizations including the EU SmartGrids advisory was Chair Professor and Head (2002–2007) of the
council. Department of Electrical Engineering, The Hong
Kong Polytechnic University. His current research
interests include power system analysis, planning and operation, as well as
computation intelligence applications to power engineering.
Prof. Wong received three Sir John Madsen Medals (1981, 1982, and
Zhao Yang Dong (M’99–SM’06) received the Ph.D. 1988) from the Institution of Engineers Australia, the 1999 Outstanding
degree from The University of Sydney, Sydney, Aus- Engineer Award from IEEE Power Chapter Western Australia, and the 2000
tralia, in 1999. IEEE Third Millennium Award. He was General Chairman of IEEE/CSEE
He is Ausgrid Chair of Intelligent Electricity PowerCon2000, and of IEE (IET) APSCOM 2003 and 2009. He was one of the
Networks at the University of Newcastle, Newcastle, Founders of the international DRPT conference series. Presently, he is serving
Australia. He previously held academic positions as Editor-in-Chief of the IEEE PES LETTERS. He was an Editor-in-Chief of
with Hong Kong Polytechnic University, and The IEE Proceedings in Generation, Transmission and Distribution and Editor
University of Queensland, Brisbane, Australia. (Electrical) of the Transactions of Hong Kong Institution of Engineers. He is a
He also held industrial positions with Transend Fellow of IET, HKIE, and IEAust.
Networks, Tasmania, Australia. His research interest
includes power system planning, smart grid, power
system security assessment, power system stability and control, power system
load modeling, electricity market, and computational intelligence and its
application in power engineering.
Dr. Dong is an editor of the IEEE TRANSACTIONS ON SMART GRID.