1296

IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 29, NO. 3, MAY 2014

A Novel Fault-Tolerant DFIG-Based Wind Energy

Conversion System for Seamless Operation
During Grid Faults
Parag Kanjiya, Bharath Babu Ambati, and Vinod Khadkikar, Member, IEEE

AbstractA novel fault-tolerant configuration of doubly fed

induction generator (DFIG) for wind energy conversion systems
(WECSs) is proposed in this paper for the seamless operation
during all kinds of grid faults. The proposed configuration is developed by replacing the traditional six-switch grid-side converter
(GSC) of DFIG with a nine-switch converter. With the additional
three switches, the nine-switch converter can provide six independent output terminals. One set of three output terminals are
connected to the grid through interfacing inductors to realize
normal GSC operation while, the other set of three output terminals are connected to neutral side of the stator windings to provide
fault ride-through (FRT) capability to the DFIG. An appropriate
control algorithm is developed for the proposed configuration
that: 1) achieves seamless fault ride-through during any kind of
grid faults and 2) strictly satisfies new grid codes requirements.
The effectiveness of the proposed configuration in riding through
different kind of faults is evaluated through detailed simulation
studies on a 1.5-MW WECS.
Index TermsDoubly fed induction generator (DFIG), grid
faults, neutral side converter, seamless fault ride-through (seamless FRT), unbalance, wind turbine (WT).

I. INTRODUCTION
HE large-scale integration of wind power in todays
power system is increasing its proposition in total electricity generated. To improve the reliability of the power system
with large-scale wind power integration, different countries
have specified fault ride-through (FRT) requirements for wind
turbine (WT) in their respective grid codes [1][4]. A typical
FRT requirement curve for WT, as per German and Irish grid
codes, is shown in Fig. 1. According to Fig. 1(a), WTs must stay
connected when the terminal voltage remains above the bold
line. In addition to remaining connected to the transmission
system, recent grid codes put stringent requirements on the
supply of reactive current by WTs as per Fig. 1(b), in order to
improve the voltage security of the system. The general grid
code requirements can be summarized as follows.
1) The WTs must stay connected during voltage dips above
the specified level for the time duration stated in Fig. 1(a).

Manuscript received May 23, 2013; revised September 12, 2013; accepted
October 11, 2013. Date of publication December 05, 2013; date of current version April 16, 2014. Paper no. TPWRS-00651-2013.
The authors are with the Institute Center for Energy, Masdar Institute
of Science and Technology, Abu Dhabi, United Arab Emirates (e-mail:
pkanjiya@masdar.ac.ae; bambati@masdar.ac.ae; vkhadkikar@masdar.ac.ae).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TPWRS.2013.2290047

Fig. 1. Recent grid code requirements for grid connected wind turbines.
(a) FRT requirement curve. (b) Reactive support requirement curve.

2) The WTs must support the grid voltage with additional reactive current during voltage dips, as per Fig. 1(b). Moreover, the voltage control must take place within 20 ms after
the fault recognition.
3) The active power output of the WTs must be continued
immediately after the fault clearance and increased to the
original value with a gradient of at least 20% of the rated
power per second.
The doubly fed induction generator (DFIG) is widely used
for the grid-connected variable-speed WTs because of high efficiency and independent control of active and reactive power
using partial capacity converters. In the conventional architecture of a DFIG, the stator windings are directly connected to the
grid and the rotor windings are connected to the grid through
back-to-back connected voltage-source converters (VSCs) [5],
[6]. These two VSCs can be identified as: 1) rotor-side converter (RSC), connected between the rotor and dc link and 2)
grid-side converter (GSC), connected between the grid and dc
link through interfacing inductors. As the stator windings are
directly connected to the grid, any amount of voltage dip at
the DFIG terminals resulting from the grid fault directly affects
the air-gap flux and, hence, the energy conversion process. Depending on the type of fault, the voltage dip may introduce dc
component or combination of dc and reverse rotating ac component in the air-gap flux [7]. These flux components induces high
voltage in the rotor windings at rotational and/or double the rotational frequency. The RSC itself cannot limit these high-frequency voltages due the modulation index constraint and hence
loses its current control capabilities. Unless the proper mitigating measures are employed, the rotor currents under the grid
fault condition can exceed the transient current rating of the
RSC. Grid faults also cause severe mechanical stress on the
bearings and the gear box of WECS due to torque pulsation.

0885-8950 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

KANJIYA et al.: NOVEL FAULT-TOLERANT DFIG-BASED WECSS FOR SEAMLESS OPERATION DURING GRID FAULTS

1297

Fig. 2. Proposed fault-tolerant DFIG wind turbine configuration for seamless FRT.

Many solutions have been proposed to provide or improve the

FRT capability of the DFIG wind turbines. To limit the transient
currents, the rotor crowbar that disables the RSC and short circuit the rotor windings through resistors during the grid faults is
discussed in [8] and [9]. The main drawback of this scheme is
that the DFIG consumes huge reactive power during grid fault
and further aggravates the voltage dip. Another configuration
that uses series dynamic resistors is proposed in [10] and [11].
This can maintain the stator voltage or rotor currents within the
limits but not appropriate if the DFIG is being controlled to
supply the reactive power to the grid. Furthermore, above mentioned fault ride through techniques fail to maintain the pre-fault
voltage across the stator windings which causes undesired electrical and mechanical transients in the system.
To obtain transient-free FRT performance, it is necessary to
keep the pre-fault voltage across the stator windings during the
grid faults. To achieve this, the dynamic voltage restorer (DVR)
arrangement using an additional VSC with an output filter
and series transformer with bypass switches is investigated in
[12][14]. To provide the protection against dead short circuit at
DFIG terminals, the DVR should be rated for 100% of the full
DFIG rating. Moreover, this technique involves an operational
delay of auxiliary semiconductor switches (used to bypass
the series transformer during normal condition) and hence the
electrical transients are unavoidable. Furthermore, this solution
is very costly as it involves many auxiliary components. Alternately, the use of parallel grid side rectifier and series inverter at
the Y-point of the stator windings is proposed and investigated
during balanced faults in [15]. This scheme effectively reduces
the number of passive and active components used to achieve
fault-tolerant operation of DFIG. However, the major drawback
of this scheme is that the stator windings need to carry the slip

power during super-synchronous speeds and inability to ride

through unbalanced faults.
With the aim of achieving seamless FRT operation in line
with recent grid codes using minimum additional components,
a novel fault-tolerant configuration of DFIG using a nine-switch
converter is proposed in this paper. In the proposed configuration, a traditional six-switch GSC of DFIG is replaced with a recently proposed nine-switch converter [16][18] to provide two
independent three-phase outputs. One of these three-phase outputs is connected to grid via an interfacing inductor to realize
normal GSC operation, while the second output is connected
to the neutral side of the stator windings to offer series voltage
compensation capability to DFIG for riding through any kind
of grid faults. An appropriate control algorithm for the control
of a nine-switch converter is developed to achieve the seamless
FRT operation of DFIG. Moreover, to provide reactive current
support in line with recent grid code requirements during the
grid fault conditions, a coordinated reactive power controller is
developed to share reactive current between GSC and RSC. It
is worth noting that the proposed fault-tolerant DFIG configuration uses only three extra switches to achieve transient-free
operation during any kind of grid fault.
II. PROPOSED DFIG CONFIGURATION AND MODELING
A. System Description
The schematic of the proposed fault-tolerant DFIG WT configuration to achieve seamless FRT operation is shown in Fig. 2.
Similar to a conventional DFIG, the stator windings are directly connected to grid and the rotor windings are connected
to the RSC (switches R1R6) through slip rings. The RSC in
the proposed configuration shares the dc link capacitor (C) with

1298

IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 29, NO. 3, MAY 2014

the dual-output nine-switch converter (switches G1G9) instead

of conventional six-switch converter. The upper output of the
nine-switch converter is connected to grid via interfacing inductor
while, the lower output is connected to the neutral
side of the stator windings of the DFIG. During the normal operation of the DFIG the lower three switches (G3, G6, and G9) of
the nine switch converter are short circuited to form the Y-point
of the stator windings while, the upper six switches (G1, G2, G4,
G5, G7, and G8) are controlled as a conventional GSC to regulate the dc-link voltage. During the voltage dip resulting from
the grid faults, the lower three switches (G3, G6, and G9) also
start switching to generate the compensating voltages on the
neutral side of the stator winding to maintain pre-fault voltage
across it. This operation of generating the compensating voltages on the neutral side of the stator winding is termed here as
neutral side converter (NSC) operation. During grid faults the
NSC will absorb part of the active power generated by DFIG
and pumps it to the dc link, hence, dc-link voltage tends to rise
if any preventive measures are not taken. To protect the dc-link
capacitor from over voltage during the grid faults, a dynamic
braking resistor (DBR) is connected across the dc-link capacitor as shown in Fig. 2.

developed and stator reactive power in terms of rotor variables

can be written as
(3)
(4)
From (3) and (4), it can be seen that, by aligning the -axis
of the reference frame to the air-gap flux, the electromagnetic
torque developed is directly proportional to the -axis rotor current while the stator reactive power is directly proportional to
the -axis rotor current. Hence, by employing the decoupled
control of
and
, the electromagnetic torque (or active
power) and the stator reactive power can be controlled independently by RSC.
III. CONTROL OF THE DFIG WIND TURBINE
The control of the RSC and nine-switch converter (GSC and
NSC) in the proposed DFIG configuration has two modes of
operation: 1) normal mode and 2) fault mode. The details of the
control strategies for both the converters during these modes of
operation are discussed here.
A. Normal Mode Operation

B. Modeling of the DFIG

To analyze the transient and steady-state performance of the
wound rotor induction machine with the six-terminal stator, it is
modeled in d-q reference frame rotating at synchronous speed
following the procedure given in [20]. The voltage and flux
equations of the induction machine with the six-terminal stator
in d-q reference frame can be written as follows:

(1)

(2)
where suffix and represents the -axis and -axis components of respective variables,
and
represent the voltages
available at the grid side of the stator terminals,
and
represent the voltages available at the neutral side of the stator
terminals, and
and
represent the rotor terminal voltages.
The variables
and
represent the stator currents, while
and
represent the rotor currents.
and
represent
the stator and rotor resistances referred to stator while
,
,
and
represent the stator self-inductance, rotor self-inductance, and mutual inductances referred to stator, respectively.
is supply angular frequency while
is the rotor angular frequency in electrical radians per second.
By aligning the -axis of the reference frame to the air-gap
flux, the simplified expressions for the electromagnetic torque

In normal mode of operation, the DFIG is controlled to supply

maximum available active power by the WT and user-defined
reactive power.
1) Rotor-Side Converter Control During Normal Mode: To
control the electromagnetic torque and the reactive power produced by DFIG independently, in normal mode of operation
RSC is controlled in a synchronously rotating d-q reference
frame with d-axis aligned to stator flux vector. The expressions
for the electromagnetic torque and reactive power developed
by DFIG with d-axis oriented to stator flux are given in (3)
and (4), respectively. The electromagnetic torque and hence active power produced by DFIG is proportional to
and can
be regulated by controlling
. On the other hand, reactive
power produced by DFIG is proportional to
and can be
regulated by controlling
. The detailed control diagram of
RSC is shown in Fig. 3. The reference -axis rotor current
is calculated from the reference torque command
generated using a PI controller over the rotor speed. The reference
rotor speed command can be generated using an appropriate
maximum power point (MPPT) algorithm. The reference -axis
rotor current
is generated using a PI controller over reactive power supplied by the DFIG. In the normal mode of operation,
is allowed to take any value within the limits in order
to extract the maximum power from WT while
is limited as
per
(5)
to ensure RSC current is within the safe limit
.
The reference rotor currents
and
are tracked using decoupled PI current control by regulating
and
as per (2).
The gate signals (R1 to R6) for RSC are then issued by comparing the reference RSC voltage
with a triangular carrier wave using a sinusoidal PWM technique.
2) Nine-Switch Converter Control During Normal Mode:
The control of a nine-switch converter consists of two parts:

KANJIYA et al.: NOVEL FAULT-TOLERANT DFIG-BASED WECSS FOR SEAMLESS OPERATION DURING GRID FAULTS

1299

Fig. 3. RSC control.

Fig. 4. Nine-switch converter control.

1) GSC control and 2) NSC control. The detailed schematic of

nine-switch converter control is shown in Fig. 4.
The objective of the GSC is to regulate the dc-link voltage
to its reference value irrespective of direction of rotor power
flow. The voltage balance equation across interfacing inductor
with internal resistance
in synchronous reference frame
can be written as

(6)

By aligning the -axis of the reference frame of the GSC control with the grid voltage, active and reactive power supplied or
absorbed by the GSC can be written as

(7)
The active power and hence the dc-link voltage is proportional to
, and can be regulated by controlling
while,
the reactive power is proportional to
and can be regulated

1300

IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 29, NO. 3, MAY 2014

by controlling
. In normal mode of operation, the reference -axis GSC current
is generated using PI controller
over average dc-link voltage and the reference -axis GSC current
is set equal to zero. The reference voltage
that
has to be generated by GSC in order to control GSC current to
its reference value is computed by employing decoupled current
control as per (6).
In normal steady state operation, the voltage to be injected
by NSC at neutral side of the stator winding is zero. Hence, in
the normal mode of operation, the NSC controller is inactive
and the reference voltage to be generated by the NSC
is
kept equal to zero.
Both the GSC and NSC operation has to be carried out by
a single nine-switch converter in the proposed fault-tolerant
DFIG configuration. Like most reduced component converter
topologies, the nine-switch converter faces limitations imposed
on its allowable switching states. Both of the output terminals
of the same phase in the nine-switch converter can only connect
to either
[for phase-a:
-ON,
-ON, and
-OFF] or
0 V [for phase-a:
-OFF,
-ON, and
-ON], or its upper
output terminal to
and lower output terminal to 0 V [for
phase-a:
-ON,
-OFF, and
-ON]. The combination
where the upper output terminal needs to be connected to 0 V
and lower to
is not allowed [for phase-a:
-ON,
-ON,
and
-ON] as this short-circuits the dc link. To resolve this
limitation, two modulating references of the same phase should
share the modulation space without intersecting each other
[16]. This can be achieved by placing the reference for upper
terminal always above that of the lower terminal by adding
offsets to both the references. The adjustment of three-phase
modulating reference signals in modulation space by adding
the same offset to all the three phases does not show any impact
on the output (line voltages) of the converter (for example,
third-harmonic injection PWM method) [19]. The modulating
reference signal adjustment method for nine-switch converter
based on 120 -discontinuous modulation [19] is derived in
[18]. The offset reference voltage signals for GSC and NSC as
per [18] can be written as

(8)
The offset reference signals
and
are fed to the
individual three-phase PWM generators (PWM-I and PWM-II)
which generate the two different sets of six PWM signals. If the
GSC and NSC operations are to be achieved using two separate six-switch inverters, the PWM signals generated by PWM-I
and PWM-II can be directly given to corresponding inverters.
However, in a nine-switch inverter, the middle three switches
are shared by GSC and NSC, so their gate pulses are generated
by logical O operation of PWM signals corresponding to lower
three switches by PWM-I and upper three switches by PWM-II
as shown in Fig. 4. The PWM signals corresponding to the upper
three switches by PWM-I and lower three switches by PWM-II
are directly issued to upper three and lower three switches, respectively, of the nine-switch converter. Additional information
on the nine-switch inverter operation can be found in [16][18].
It is important to note that during normal mode of operation
the reference voltages
for NSC are zero and hence

offset reference signals

are minus one. This means
that the PWM signals for lower three switches of the nine-switch
converter are always at logic high and corresponding switches
are ON. This effectively short circuit the stator windings on neutral side and form the Y-point.
B. Fault Mode Operation
During the fault mode of operation, the NSC is controlled
to keep a pre-fault voltage across the stator winding while the
RSC and GSC control is switched to supply reactive current in
line with the requirement of the recent grid codes [Fig. 1(b)]. To
achieve transient-free operation, it is necessary to detect the grid
fault with least possible delay. The fault in the system can be
detected almost instantaneously by measuring the absolute error
between reference grid voltage magnitude (1 p.u.) and actual
grid voltage magnitude (in ) as follows:
(9)
The grid fault is detected whenever
exceeds the
threshold (typically 0.1 p.u.). Once the grid fault is detected,
the
signal in Fig. 3 and 4 changes its logic from low to
high. As
will oscillate above and below the threshold
during the unsymmetrical faults, the fault removal is detected
when the one cycle average of
reduces below the
threshold. The details of the actions taken by the controller
of different converters during the grid fault condition are
discussed below.
1) RSC Control During Fault Mode: The rotor side converter
has the capability to provide full reactive support required by the
recent grid codes during the grid fault condition provided that
the NSC keeps the pre-fault voltage across the stator winding.
However, if the RSC control is switched to supply full reactive
current, the active current supplied by the stator has to be reduced to keep RSC switch currents below the safe limit. This
leads to the over speeding of the rotor and, hence, increased
mechanical stress due to the storage of the energy produced by
the WT as kinetic energy. To reduce the negative impact of increased mechanical stress on WTs construction, it is advisable
to maximize active power extraction along with reactive current supply without exceeding the rating of any component in
the system. To achieve this, the reactive currents required to fulfill the grid code requirements is shared by the GSC and RSC
during grid faults in proportion to their ratings. The duty of supplying two third of the reactive current (maximum 0.6 p.u.) is
assigned to stator and hence RSC, while, one third (maximum
0.3 p.u.) is assigned to the GSC. To supply the reactive current
assigned to the stator during fault condition, the
is computed
using the PI controller over reference stator reactive current
( -axis current is proportional to reactive power as the reference
-axis is oriented to stator flux) as shown in Fig. 3. The
is
calculated using
(10)
to support two thirds of the reactive current required by grid
code as per Fig. 1(b).
In (10),
is the positive sequence grid voltage magnitude
and can be calculated within 20 ms to fulfill the grid code requirement. In the fault mode of operation, the priority is given

KANJIYA et al.: NOVEL FAULT-TOLERANT DFIG-BASED WECSS FOR SEAMLESS OPERATION DURING GRID FAULTS

to
to strictly support reactive current and the limit on
calculated as

is

1301

TABLE I
SYSTEM SPECIFICATIONS

(11)
2) Nine-Switch Converter Control During Fault Mode: As
discussed before, the DBR comes into picture during fault condition to keep the dc-link voltage below the safe limit which
comforts the GSC from the duty of keeping the dc-link voltage
constant. Therefore, during the grid fault condition to assist the
RSC, the GSC controller is switched to supply one third of the
reactive current required to fulfill the grid codes requirement.
The reference reactive current
for the GSC during fault
mode of operation is computed as per
if
if
and the GSC current is limited to
active current
as per

(12)
by putting limits on
(13)

The objective of the NSC control during fault condition is

to keep the pre-fault voltage across the stator winding for the
proper functioning of the RSC control. The logic high on the
signal during fault condition activates the NSC controller
as shown in Fig. 4. The compensating voltage
that has to
be injected by the NSC on the neutral side of the stator winding
during fault condition can be computed as
(14)
In (14),
and
are the present grid voltages and the
pre-fault grid voltages computed using pre-fault grid voltage
vector angle
. The pre-fault grid voltage vector angle
is obtained by saturating the phase-locked loop (PLL) over the
grid voltage using
signal as shown in Fig. 4. The pre-fault
grid voltages
are computed by holding the samples of two
cycle mean value of
when fault is detected.
The compensating voltages
can be directly used as the
reference voltages to control NSC in open-loop by converting
them to stationary reference frame. The open-loop control can
work satisfactorily for the NSC due to the absence of output
filter and series injection transformer. However, there will be
series voltage drop across the switches of the NSC which may
affect the compensation. To achieve perfect compensation of
the voltage dip, the switching voltage drop is added to
by
estimating it using PI controller which maintains the stator flux
at its pre-fault value
as shown in Fig. 4. The
d-q axis stator flux components in Fig. 4 are estimated using (1)
and (2) while, the pre-fault stator flux components are obtained
by holding the samples of two cycle mean value of
when
fault is detected.
IV. SIMULATED SYSTEM AND PERFORMANCE EVALUATION
To verify the effectiveness of the proposed DFIG wind
turbine configuration and its control to achieve fault tolerant
operation in line with the recent grid codes, an extensive simulation study is carried out in MATLAB. The detailed model of
DFIG with six-terminal stator and the power electronics converters are developed using SIMULINK and SimPowerSystems

tool-boxes. To represent the actual operating conditions, it is

assumed that the DFIG wind turbine is connected to a medium
voltage network (120 kV) through a step-up transformer
(575 V-Yg/25 kV
), a transmission line (25 kV50 km),
and another step-up transformer (25 kV
/120 kV-Yg) as
shown in Fig. 2. The wind turbine specifications and machine
parameters used are tabulated in Table I.
The performance of the proposed fault-tolerant DFIG configuration is evaluated under LLLG, LLG and LG faults at 120 kV
bus and the corresponding results are shown in Figs. 57. To
prove the compatibility of the proposed configuration with recent grid codes for the effective fault ride through, faults are
emulated for the duration of 150 ms (nine cycles of supply). All
of the results are represented in per unit (p.u.) for a wind speed
of 15 m/s.
The detailed response of the proposed fault-tolerant DFIG
configuration during the LLLG fault is shown in Fig. 5. The
system is under steady state before the simulation time of 6 s.
A LLLG fault lasting for 150 ms is applied on 120 kV bus at
6 s, which can be observed from the dip in the grid voltages
from 1 to 0.1 p.u. The moment fault is detected, fault mode
operation is enabled by a logic high of the
signal. This
enables the NSC which injects compensating voltage (PWM
switching voltage switched between
on the neutral side
of the stator winding to maintain pre-fault voltage across it.
For better visualization of NSC functionality in maintaining
pre-fault voltage across the stator winding, in Fig. 5 an average
(over a switching period) compensating voltage
is shown
instead of the actual PWM switching voltage. Note that the average voltage across the stator winding
does not experience any change due to the instantaneous voltage compensation by the NSC (without any delay as no transfer switches are

1302

IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 29, NO. 3, MAY 2014

Fig. 5. Different system variables to study the behavior of proposed fault-tolerant DFIG configuration during LLLG fault at 120-kV bus.

involved). The effectiveness of the NSC in keeping pre-fault

voltage across the stator winding is also evident from the trace
of the stator flux vector Sye
as it is maintained constant
throughout the fault duration.
As discussed in the control section, during the grid fault condition, the RSC control is switched to supply reactive current
by giving priority to . With priority to , the reference current
(responsible for maintaining WT speed and hence active power extraction) is limited to (11) with
set equal to
1.1 p.u. (with rated rotor current as base current). The evolution
of variables
and
during fault mode operation is depicted
in Fig. 5. To track these reference currents, the RSC injects an
average rotor voltage, across the rotor terminals which forces
the rotor current, to flow in rotor winding and corresponding
stator current,
in stator winding. Note that the sinusoidal
stator current
is achieved with permissible switching ripples even though the voltage injected by NSC on the neutral side
of the stator winding is the PWM switching voltage. The sufficiently high stator self-inductance served the purpose of filter
inductance. This shows that there is no need of an output filter
with NSC similar to RSC.
The GSC also switched to reactive current control mode
during fault condition (Fig. 4). The resulted GSC current due
to this control switch over is depicted in Fig. 5 as
. The
injection of the reactive current assigned to GSC as per (12)
can be noticed from the phase jump and increase in magnitude
of GSC current.

To illustrate the effectiveness of reactive current control mode

precisely, the active
and reactive
components of the
stator and GSC currents during LLLG fault are presented in
Fig. 5. As the positive sequence grid voltage falls to 0.1 p.u.
(well below 0.5 p.u.) during the fault condition, the DFIG injects a total of 0.9 p.u. reactive current
as per Fig. 1(b).
Note that this total reactive current is shared by stator and GSC
in 2:1 ratio. Also note that, due to the increase in stator and
GSC reactive currents, corresponding active currents
are
reduced for the safe operation of the converters.
To better understand the functionality of different converters,
the active and reactive powers absorbed by the different components of the DFIG system are also plotted in Fig. 5. Wherein,
positive sign indicates power absorption while negative sign indicates power generation. From the active power plot, it can be
noticed that during steady state (before 6 s) the active power absorbed by the grid
is equal to active power generated by
the stator
plus GSC
. Note that the active power produced by the rotor
is exactly equal to
as the objective
of the GSC is to transfer
into the grid via a dc link during
the normal mode of operation. The power absorbed by the NSC
can be observed to be zero during steady state as the NSC is not
active in normal mode of operation. Whereas during the grid
fault, as discussed in the control section, with the reduction in
active component of stator and rotor currents to accommodate
the necessary reactive current, a little dip in the
and
can

KANJIYA et al.: NOVEL FAULT-TOLERANT DFIG-BASED WECSS FOR SEAMLESS OPERATION DURING GRID FAULTS

1303

Fig. 6. Important system variables to study the behavior of proposed fault-tolerant DFIG configuration during LL fault at 120-kV bus.

Fig. 7. Important system variables to study the behavior of proposed fault-tolerant DFIG configuration during LG fault at 120-kV bus.

be observed although the stator voltage

is maintained at its
pre-fault value by the NSC. The active power injected into the
grid by the GSC
is zero as the GSC is injecting reactive
current up to its limit during the grid fault. With the reduction in
grid voltage during grid fault, there is a proportionate reduction
in the active power delivered to the grid
. Therefore, there
is a net amount of excess stator active power equal to deference
between absolute values of
and . This excess power is absorbed by the NSC
and dissipated in the DBR at dc link.
The active power produced by the rotor
is also dissipated
in the DBR as the power absorbed by the GSC
from the
dc link is zero. A similar analysis can also be drawn for the reactive power of different components from the plot of reactive
powers.
Because of the reduction in active power of the stator to accommodate the required reactive current component during grid
fault, a little rise in rotor speed (w) can be observed due to the accumulation of kinetic energy. Moreover, there is a net amount of
excess power pumped into the dc link by NSC and RSC during

fault condition; the DBR comes into the picture to prevent the
dc link from overvoltage (not more than 1250 V). As soon as
fault is removed from the system, the active current
of the
stator again increases within 20 ms to feed the full active power
produced by the WT which effectively fulfills the recent grid
code requirements.
The performance of proposed fault-tolerant DFIG configuration is also evaluated for the unsymmetrical faults. The responses of the DFIG during LL and LG faults are shown in
Figs. 6 and 7, respectively. It can be seen that the DFIG seamlessly ride through both the faults without noticeable transients.
It can also be noticed that the currents injected by the DFIG to
the grid are balanced although the faults are unsymmetrical in
nature. Moreover, the DFIG effectively injects the reactive current corresponding to positive sequence grid voltage according
to Fig. 1(b).
After evaluating the performance of proposed fault-tolerant
DFIG during different fault conditions, the superiorities and
benefits associated with it are summarized here.

1304

IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 29, NO. 3, MAY 2014

The proposed DFIG configuration can effectively keep the

pre-fault voltage across the stator winding of the DFIG
during any type of fault condition to achieve transient-free
FRT.
It is an economic solution to achieve series voltage compensation as only three full-rated extra switches are used
in the proposed solution compared with the DVR solution
where a full-rated six-switch converter is required.
With the proposed reactive current sharing control, the proposed fault-tolerant DFIG is able to fulfill all of the requirements of the recent grid codes.
In the proposed DFIG configuration, as the compensation
of grid voltage variations is achieved from the neutral side
of the stator winding, there is no need for a full-rated series
injection transformer as in case of the DVR solution.
There is no need for an output filter for the injected voltages
as the stator self-inductance will be sufficiently high.
The proposed fault-tolerant DFIG seamlessly rides through
any kind of grid fault as it does not involve any transfer
switch (required to short circuit series injection transformer during steady-state operation in the case of DVR
solution) operation during FRT.
V. CONCLUSION
A novel fault-tolerant DFIG configuration with minimum
additional components (only three extra switches) is proposed
for seamless operation during grid faults. Furthermore, the
detailed control strategies for the RSC and nine-switch converters during steady-state and fault conditions are developed to
fulfill recent grid codes requirements. The effectiveness of the
proposed fault-tolerant DFIG in riding through different types
of grid faults (LLLG, LL and LG) is evaluated through detailed
digital simulation studies. The simulation studies show that the
proposed fault-tolerant DFIG seamlessly rides through all of
the faults while fulfilling grid codes requirements. The advantages and benefits associated with the proposed configuration
over other FRT techniques can be summarized as seamless
FRT, strict satisfaction of grid code requirements, economical
solution with only three additional switches, no need of output
filter, series injection transformer, and its bypass switches.
REFERENCES
[1] W. Christiansen and D. R. Johnsen, Analysis of requirements in selected grid codes, section of electric power engineering, Tech. Univ.
of Denmark, 2006, Tech. Rep..
[2] M. Tsili and S. Papathanassiou, A review of grid code technical requirements for wind farms, IET Renew. Power Gener., vol. 3, no. 3,
pp. 308332, Mar. 2009.
[3] E.ON Netz GmbH., Bayreuth, Germany, Grid codeHigh and extra
high voltage, Apr. 2006.
[4] EirGrid Grid Code. ver. 3.5, EirGrid plc, Ireland, Mar. 2011.
[5] R. Pena, J. Clare, and G. Asher, Doubly fed induction generator using
back-to-back converters and its application to variable-speed wind-energy generation, Proc. Inst. Electr. Eng.Elect. Power Appl., vol.
143, no. 3, pp. 231241, May 1996.
[6] S. Muller, M. Deicke, and R. W. De Doncker, Doubly fed induction
generator systems for wind turbine, IEEE Ind. Appl. Mag., vol. 8, no.
3, pp. 2633, MayJun. 2002.
[7] 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. 709717, Sep. 2007.

[8] S. Seman, J. Niiranen, and A. Arkkio, Ride-through analysis of doubly

fed induction wind-power generator under unsymmetrical network disturbance, IEEE Trans. Power Syst., vol. 21, no. 4, pp. 17821789,
Nov. 2006.
[9] J. Morren and S. W. H. de Haan, Ride through of wind turbines with
doubly-fed induction generator during a voltage dip, IEEE Trans. Energy Convers., vol. 20, no. 2, pp. 435441, Jun. 2005.
[10] A. Causebrook, D. J. Atkinson, and A. G. Jack, Fault ride-through of
large wind farms using series dynamic braking resistors, IEEE Trans.
Power Syst., vol. 22, no. 3, pp. 966975, Aug. 2007.
[11] J. Yang, J. E. Fletcher, and J. OReilly, A series-dynamic-resistorbased converter protection scheme for doubly-fed induction generator
during various fault conditions, IEEE Trans. Energy Convers., vol.
25, no. 2, pp. 422432, Jun. 2010.
[12] A. O. Ibrahim, T. H. Nguyen, D.-C. Lee, and S.-C. Kim, A fault
ridethrough technique of DFIG wind turbine systems using dynamic
voltage restorers, IEEE Trans. Energy Convers., vol. 12, no. 3, pp.
871882, Sep. 2011.
[13] C. Wessels, F. Gebhardt, and F. W. Fuchs, Fault ride-through of a
DFIG wind turbine using a dynamic voltage restorer during symmetrical and asymmetrical grid faults, IEEE Trans. Power Electron., vol.
26, no. 3, pp. 807815, Mar. 2011.
[14] D. Ramirez, S. Martinez, C. A. Platero, F. Blazquez, and R. M. de
Castro, Low-voltage ride-through capability for wind generators
based on dynamic voltage restorers, IEEE Trans. Energy Convers.,
vol. 26, no. 1, pp. 195203, Mar. 2011.
[15] P. Flannery and G. Venkataramanan, A fault tolerant doubly fed induction generator wind turbine using a parallel grid-side rectifier and
series grid-side converter, IEEE Trans. Power Electron., vol. 23, no.
3, pp. 11261135, May 2008.
[16] C. Liu, B. Wu, N. R. Zargari, D. Xu, and J. Wang, A novel threephase three-leg ac/ac converter using nine IGBTs, IEEE Trans. Power
Electron., vol. 24, no. 5, pp. 11511160, May 2009.
[17] T. Kominami and Y. Fujimoto, Inverter with reduced switching-device count for independent ac motor control, in Proc. IEEE-IECON,
2007, pp. 15591564.
[18] L. Zhang, P. C. Loh, and F. Gao, An integrated nine-switch power
conditioner for power quality enhancement and voltage sag mitigation, IEEE Trans. Power Electron., vol. 27, no. 3, pp. 11771190, Mar.
2011.
[19] A. M. Hava, R. Kerkman, and T. A. Lipo, Carrier-based PWM-VSI
overmodulation strategies: Analysis, comparison, design, IEEE
Trans. Power Electron., vol. 13, no. 4, pp. 674689, Jul. 1998.
[20] P. C. Krause, O. Wasynczuk, and S. D. Sudhoff, Analysis of Electric Machines and Drive Systems, 2nd ed. New York, NY, USA:
WileyIEEE, 2002.
Parag Kanjiya received the B.Eng. degree in electrical engineering from the B.V.M. Engineering College, Sardar Patel University, V.V. Nagar, India, in
2009, and the M.Tech. degree in power systems from
the Indian Institute of Technology Delhi (IITD), New
Delhi, India, in 2011.
Since October 2011, he has been a Research
Engineer with the Masdar Institute of Science and
Technology, Abu Dhabi, United Arab Emirates.
His research interests include applications of power
electronics in distribution systems, power quality
enhancement, renewable energy, FACTS, and power system optimization.
Mr. Kanjiya was the recipient of the K.S. Prakasa Rao Memorial Award for
getting the highest C.G.P.A at IITD in August 2011.

Bharath Babu Ambati received the B.E. degree

in electrical and electronics engineering from Sir
C. R. Reddy College of Engineering (affiliated
with Andhra University), Eluru, India, in 2009, and
the M.Tech degree in power electronics, electrical
machines and drives from the Indian Institute of
Technology Delhi (IITD), New Delhi, India, in 2011.
He is currently working toward the Ph.D. degree at
Masdar Institute of Science and Technology, Abu
Dhabi, United Arab Emirates.
From July 2011 to June 2012, he was with
Schneider Electric India Private Ltd. as a Product Expert of Motion & Drives.
His current research interests include power electronics, electrical machines,
renewable energy generation, and power quality improvement.

KANJIYA et al.: NOVEL FAULT-TOLERANT DFIG-BASED WECSS FOR SEAMLESS OPERATION DURING GRID FAULTS

Vinod Khadkikar (S06M09) received the B.E.

degree from the Government College of Engineering, Dr. Babasaheb Ambedkar Marathwada
University, Aurangabad, India, in 2000, the M.Tech.
degree from the Indian Institute of Technology Delhi
(IITD), New Delhi, India, in 2002, and the Ph.D.
degree from the cole de Technologie Suprieure
(E.T.S.), Montral, QC, Canada, in 2008, all in
electrical engineering.
From December 2008 to March 2010, he was a
Postdoctoral Fellow with the University of Western

1305

Ontario, London, ON, Canada. Since April 2010, he has been an Assistant Professor with the Masdar Institute of Science and Technology, Abu Dhabi, United
Arab Emirates. From April 2010 to December 2010, he was a Visiting Faculty Member with the Massachusetts Institute of Technology, Cambridge, MA,
USA. His research interests include applications of power electronics in distribution systems and renewable energy resources, grid interconnection issues,
power quality enhancement, active power filters, and electric vehicles.