1

Wind Turbine Generator Low Voltage Ride

through Requirements and Solutions
G. Joós, Fellow, IEEE

flow from the wind farm to the grid is partly or in the worst
Abstract--Utilities have responded to the significant increase in case, totally interrupted and the power produced by the wind
the number of installed and proposed wind farms by developing farm has to be curtailed, dissipated or controlled by the wind
specific grid codes derived from those used for conventional turbine controllers. The control options depend on the type of
generators. The presentation summarizes the main grid code
wind turbine generator used.
requirements, related to reactive power and voltage control, low
voltage ride through (LVRT), and other issues such as control of The more common wind turbine implementations are
real power. It provides a technical overview of the LVRT based on induction generators, with or without slip energy
requirements in different jurisdictions. It reviews some of the control, full converter based generators and the doubly fed
LVRT solutions available for wind turbines based on doubly fed induction generator (DFIG), the latter being a common
induction generators (DFIG). The impact of these solutions on implementation. This paper deals with methods of managing
the DFIG response is evaluated under different fault and
the power produced by DFIGs during the low voltage ride
operating conditions. The investigation is carried out on a
representative power system benchmark. Solutions are compared through periods. This is done using among other the following
from the machine performance and grid voltage response methods: (i) the active crowbar [3, 4, 5]; (ii) an energy
viewpoints. The parameters affecting the performance of the management system connected to the intermediate dc bus
solutions are presented. (energy storage or dump load) [6, 7]; (iii) a rotor flux control
technique [8].
Index Terms--Wind power generation. Doubly-fed induction This paper presents and compares the results obtained
generators. Grid code. Low voltage ride through. Power system
benchmark. Short circuits.
using the three different technologies designed to handle the
LVRT requirement. The following issues are examined, with
I. INTRODUCTION reference to the requirements of the grid code: (i) the
operation of the DFIG as seen from the grid, namely the
A S wind farms are installed in larger numbers and the
penetration level increases, utilities are looking at means
to ensure that these new types of power plants behave as
voltage recovery following the clearing of the fault; (ii) the
evolution of the generator variables, namely speed and of the
converter dc bus voltages. It points out the strengths and
much as possible as conventional power generating stations.
weaknesses of the methods and suggests solutions. It
However, to accommodate the specific operating
examines the behaviour of these methods under different
characteristics of wind farms, utilities have generally
conditions including: (i) the type of short circuit; (ii) the depth
established specific grid codes for wind farms.
of the voltage drop, which depends upon the position of the
Grid codes cover among others, issues related to frequency
short circuit in the system.
operating ranges, reactive power capability, and voltage
In order to create a wide range of operating conditions, the
operating ranges under steady state and transient conditions.
DFIGs are operated in a benchmark electric power system,
One requirement is that the turbine remain connected to the
originally developed for testing FACTS devices and later used
grid within a certain voltage range and for a given time
for analyzing the behaviour of wind farms in a power system.
duration, a requirement expressed in the form of the Low
Voltage Ride Through (LVRT) curve [1]. Combined with the
II. GRID CODES
high voltage withstand requirement, it gives the complete
voltage operating range of the turbine, Fig. 1. Grid codes vary in scope and specific details from one
Low voltage occurrences are usually associated with grid jurisdiction to another. The main requirements of typical grid
disturbances, mostly in the form of short circuits occurring on codes are summarized below:
the lines connecting the wind farm to the main grid or at (i) Continuous voltage operating range - The wind turbines
remote locations within the grid. In such situations, the power are expected to operate within typical grid voltage variations.
In most Canadian utilities, the continuous voltage range is
from 0.9 to 1.1 pu.
This work was supported in part by the National Sciences and Engineering
Research Council of Canada. (ii) Frequency operating range - The wind turbines are
G. Joós is with McGill University, Montreal, Canada (e-mail: expected to operate within typical grid frequency variations.
geza.joos@mcgill.ca). In Canada, each utility can define its own frequency range

©2008 IEEE.
 2

where the turbines should normally operate. In Manitoba, the 1.2

100 ms
range is broadest with normal operation required while the 1.1
1
system frequency varies from 59.0 to 61.5 Hz.
(iii) Reactive power capability - The wind farms are
required to control their reactive power output. Specifically, in No trip

V (pu)
Wind
Canada, the power factor range is typically between 0.9 (lag) generator
to 0.9 (lead), and may depend upon the province. The turbines may trip

can be required to regulate their terminal voltage.

(iv) Power control - A certain control over the real power 0.150 1.5 1800 (30 min)
time (s)
output of the wind farm is also required. Maximum ramp rates
Fig. 1. Typical low voltage ride through requirement - Europe
are imposed on the wind turbine. In Canada, Hydro-Quebec
requires ramping up or down capability from 0 to Pmax, or
from Pmax to 0, in an adjustable 15 to 60 minutes interval. In
Manitoba, the ramp rate is 20 MW/min with the capacity to
reduce output to 0 MW in 5 minutes. Also turbines are
required to not exceed the maximum output power specified
by the system operator.
(v) High voltage operation - In the event the voltage goes
above 1.1 pu, the turbines should be capable of to staying on
line for a given length of time. The exact ranges are utility
specific. As an example, Hydro-Quebec specifies a voltage to
1.4 pu for 30 ms, Fig. 2. The time increases as the voltage Fig. 2. Typical low voltage ride through requirement - Canada
decreases until the voltage is in the continuous operating
region. Initially, the minimum voltage for the curve in Fig. 1 was
(vi) Low voltage operation or Low Voltage Ride Through 0.15 pu, but was subsequently lowered to 0 pu, which has
(LVRT) - In the event of a voltage drop, the turbines are been adopted by most codes. The typical duration is 0.15 s,
required to remain connected for a specific amount of time but some codes have different values and specify the time in
before being allowed to disconnect. This requirement is to cycles (Manitoba, 6 cycles, 0.10 s).
ensure that there is no loss of generation for normally cleared In general, the specific characteristics of the LVRT are
faults. Disconnecting a wind generator too quickly could have adapted from the requirements for conventional generators,
a negative impact on the grid, particularly with large wind complemented with studies of the behavior of the power
farms. In addition, some utilities require that the wind system in a specific jurisdiction. Grid codes are posted on the
turbines help support the grid during faults. In Germany, the web site of the transmission system operator in a given
E.on grid code (2006) specifies that the wind turbines should geographic jurisdiction.
supply 1.0 pu of reactive current when the voltage is below Finally, some grid codes specify an additional lower
0.5 pu. In Canada, Alberta, BC, and Quebec, among others, voltage no trip zone for individual wind turbine under
have specific requirements concerning low voltage operation. transient conditions. However reduced power output is
Other generation controls, typically required from permissible in this region (Manitoba), with a return to full
conventional generators, are usually not, at this time, required power, with application of the allowed power gradient limits.
from wind turbine generators, including frequency regulation, The LVRT requirement demands that the turbine remain
power system stabilizers and automatic generation control. connected to the grid while the voltage is within the LVRT
curve. Disconnection is allowed outside this range. It is
III. LVRT REQUIREMENTS therefore important to verify that the turbine operating limits
One of the first formulations of the LVRT requirement was are not violated while the PCC voltage is inside the LVRT
originally proposed in Europe in 2003 by E.On. This curve. As well, it is important to determine which factors
requirement has evolved and a later version is shown in Fig. 1. impact these turbine variables the most. The purpose of this
Since then, many jurisdictions have proposed their own work is to test three candidate LVRT solutions for DFIG
version of the LVRT curve, the one applying to Quebec is based wind turbine. The solutions are evaluated based on the
illustrated in Fig. 2. The formulation of the requirement is following criteria:
often based on the specific characteristics and configurations (i) Grid variables - Voltage during fault, voltage overshoot
of the electric power system into which the wind farm is at recovery, response time.
connected. In particular, long radial connections of a farm to (ii) Turbine variables - Rotor speed, rotor current, dc bus
the grid require that the farm remain connected even if the voltage.
voltage at the farm terminal drops to 0 during a three phase
short circuit, as shown in Fig. 2.
 3

IV. DFIG LVRT TECHNOLOGIES V. WIND FARM MODELING

The DFIG is a commonly used variable speed generator in The wind turbine generators are modeled using detailed
wind turbine applications. It consists of an induction machine three phase models of all components, including the induction
with back-to-back voltage source converters linking the rotor generator transients and the power electronic converters. The
to the grid, Fig. 3. The rotor-side converter is used to control models represent accurately the performance of the actual
the machine speed and the reactive power supplied through system in steady state and under disturbances and
the machine stator, while the grid-side converter is used to perturbations, such as single and three phase short circuits.
maintain the dc bus voltage constant. The reactive power of Detailed modeling appears to be the preferred approach for
this converter can be controlled, and one strategy is to the study of wind farm integration with some utilities.
maintain the power factor of this converter at unity. This It is argued that the transient modeling approach does not
minimizes the current flowing in the grid side converter. accurately reflect the details of the operation of the power
During a fault, the transient currents on the stator are electronics converters and their control systems [9, 10]. In
reflected on the rotor windings. The resulting rotor currents particular, the controller can be represented without requiring
may exceed the rotor converter current rating. The transients approximation, including all the non-linearities. In addition,
can also result in a rise in the dc bus voltage during the fault, asymmetrical faults are better studied in a three phase detailed
due to the power being generated. Voltage variations also representation rather than sequence components (including
occur during the voltage recovery period. The rotor-side the positive sequence). Detailed modeling could result in
converter must be designed to either withstand these significant execution times. A solution to this problem is to
conditions, or protection circuits must be used on the dc bus. use more powerful real time simulators. These also offer the
The following three solutions are compared and analyzed: possibility of replacing some of the controllers with real
(i) the active crowbar rotor circuit [3, 4, 5], Fig. 3; (ii) the dc prototype or production controllers.
bus energy storage circuit [6, 7], Fig. 4; (iii) the rotor current The system was implemented in the Matlab/Simulink/
control through control of the machine flux [8]. Details of the SimPowerSystem environment. The benchmark system was
operation of the three solutions are beyond the scope of this subjected to a variety of low voltage events in order to
paper. However, the performance of each solution with compare the performance of the different LVRT solutions.
reference to the LVRT requirements will be presented and Both balanced and unbalanced faults were considered, at
discussed. various distances from the wind farm terminals. Since voltage
events of less than 700 ms are considered, which is typical of
short circuits, the wind speed is assumed constant throughout
the duration of the event. Different wind speeds were
considered, as they change the DFIG operating conditions.

VI. BENCHMARK POWER SYSTEM

The benchmark power system is the IEEE 12 bus FACTS
benchmark, Fig. 5, developed for investigating the
performance of FACTS devices and adapted to study the
impact of wind farms on power systems. It consists of two
345 kV buses, four 230 kV buses, and four low voltage buses
Fig. 3. DFIG equipped with DC converter and energy storage system
connected to the generators (G1 to G4) [11]. The system is
divided into three areas. Area 1 is a generation area with
hydro generator (G2) and a small load. Area 1 also contains
infinite bus, represented by generator G1. Area 3, located
between area 1 and 2 is equipped with a hydro generator
whose capacity is not sufficient to meet its own demand.
Area 2 consists mainly of loads with a thermal generator.
Areas 2 and 3 have shunt capacitors to support the voltage.
The parameters of the system are given in [11].
The wind farm replaces the hydro generator connected at
bus 12 in Fig. 5. At rated wind speed, the wind farm can
supply the same amount of power as the conventional unit it
replaces. The wind farm is represented by five equivalent
turbines, each using a DFIG. The reactive power capability of
Fig. 4. DFIG equipped with dc bus converter and energy storage system the DFIG is used to regulate the wind farm terminal voltage.
 4

Power system operators may require that results from the

modeling and simulation tests be verified by field tests, either
during commissioning or in special field tests.
The wind farm performance for the three LVRT solutions
is simulated for different types of short circuit, namely three
phase short circuit at the point of common connection (PCC)
of the wind farm to the grid (the worst case in terms of power
flow interruption), as well as other types of faults (double
line, single phase to ground), and with faults occurring at
different distances from the farm, resulting in various low
voltage levels at the PCC. The power output, dependent on
the wind speed in the farm, is also varied. The duration of the
Fig. 5. Benchmark power system - Single line diagram – Wind farm at Bus 12
low voltage event (short circuit) is varied from 150 ms (9
cycles), Fig. 6, to 667 ms (40 cycles, extreme case), Fig. 7, to
VII. WIND FARM LVRT RESPONSE
evaluate the response of the different solutions.
A. General Considerations B. Wind Turbine Operation
The purpose of the parametric study is to determine The following summarizes the findings:
compliance with the LVRT requirements and the variations of (i) Generator control - The change of operating conditions
the variables of the wind farms and verify that: (i) the wind associated with the use of crowbar or flux method resulted in
farm meets the LVRT requirements for a number of operating a loss of real and reactive power (P-Q) control over the
conditions; (ii) the variables within the wind farm remain machine during the fault. Hence, at recovery, these two
within acceptable values. The following operating solutions could draw from the grid the reactive power
characteristics are verified: required to magnetize the generator. In the ESS method the
(i) Overall response – Verification that the turbine stays control over the machine was retained. However this control
connected and variables are within acceptable limits. occasionally led to overvoltages upon return to normal
(ii) Operation during the fault and during recovery – operation as the controller adjusted its operation to the new
System external parameters include voltage (low and high conditions.
voltage) at the PCC, current flowing into the Electric Power (ii) Energy management - Different technologies to ride
System (EPS), and power injected into the EPS, reactive through the fault handle the energy from the turbine in
power injected into the EPS. Internal turbine generator different ways. The energy distribution was considered over
parameters include dc bus voltage and rotor speed. the period covering the fault and voltage recovery. In the ESS
method, the generator energy is stored in the energy storage,
and subsequently returned to the grid. The flux method had its
energy stored in the rotor inertia, in the dc link voltage, and
subsequently sent to the grid. The crowbar method had its
energy stored in the rotor inertia, dissipated in the rotor
resistance (through the crowbar), in the dc link capacitor,
subsequently sent to the grid. The dc link capacitor was only
affected when the crowbar was not active, that is during
resumption of normal operation. For the three methods, the
major part of the energy was stored in the rotor inertia (for a
400 ms three phase short circuit at PCC: 70 %, 82 %, 84 %
for crowbar, ESS, and flux respectively). Rotor speed during a
fault should be monitored to avoid overspeed. The energy
distribution was dependent on the nature of the fault: the ESS
method was able to maintain rotor speed more easily for
remote faults, where as the flux and crowbar experienced
greater acceleration.
(iii) Rotor current - Both the ESS and the flux method
experienced large rotor current. The maximum current
observed varied from 2 pu (crowbar) to 4 pu (ESS). The
maximum current observed for longer faults were within the
Fig. 6. Response of wind farm variables to a 150 ms, three phase fault, at the same range. As can be expected, the large currents were
PCC. PCC voltage, line current, power flowing into the system, reactive power mostly observed for faults located close to the wind farm.
injected into the system, dc bus voltage, rotor speed. High wind conditions.
Activation of the crowbar limited the rotor current rapidly.
 5

value. Both the voltage and power generation recovery were

dependent on the protection removal. In some cases the
crowbar was reactivated at recovery, increasing the recovery
time. When using this method, it may be required to oversize
the rotor-side converter.
(vi) ESS operation and response - The ESS method had the
advantage of retaining control over the machine during the
fault. Hence there is a possibility of supporting the grid
voltage, through the rotor-side converter, as well as the grid-
side converter. The continued control over the machine
resulted in fast recovery after the fault was cleared.
However, some overvoltage (within the allowed range) was
observed. Under severe faults resulting in low terminal
voltage at the PCC, the DFIG could not export much power to
the grid. Hence the ability to control the machine real power
output had no advantage in this case. For less severe fault
however, the DFIG was able to export some of its real power
to the grid. In some case, the DFIG continued normal
operation, minimally affected by the fault. Hence, for those
less severe faults, this method has an advantage over the other
two. It was found that this method is affected more by the
fault severity than the type of fault. The method requires the
addition of a dc converter and an energy storage system as
well as appropriate sizing of the rotor-side converter.
(vii) Flux control and response - The flux method had the
Fig. 7. Response of wind farm variables to a 667 ms, double line to ground
fault, at a remote bus (Bus 1). PCC voltage, line current, power flowing into the advantage of not requiring any additional equipment to meet
system, reactive power injected into the system, dc bus voltage, rotor speed, rotor the LVRT requirement. This method was usually activated
angle. High wind conditions quickly to allow the rotor-side converter enough time to
oppose the fault related current components induced by the
(iv) DC bus voltage variations and control - Both the stator flux. During the fault, the DFIG response was smooth,
crowbar and ESS methods had an alternative path for the rotor with little oscillation in the output power and terminal voltage.
current, effectively preventing them from affecting the dc link The recovery was similar to that of the crowbar method in
capacitor voltage significantly. The dc voltage experienced most cases. As P-Q control over the machine is lost during the
less voltage variations during the fault. The crowbar operation fault, the DFIG consumed reactive power at grid recovery.
did result in dc voltage drops as attempts were made to The voltage recovery was slower than with the ESS; however
resume normal operating conditions, but these drops were this method did not exhibit grid overvoltages. The power
successfully corrected by the grid-side converter. In the ESS generation was also resumed faster than for the crowbar, but
method had any changes in the dc link capacitor were more slowly than for the ESS. It was found that this method is
corrected by the dc storage side converter, as long as the more affected by the fault severity than the type of fault
energy storage was within its limits. The flux method had no
other path for the current, resulting in the charging of the dc C. Impact of External Factors
link capacitor if the current could not be limited. For faults The responses observed were also influenced by external
close to the PCC, the dc link voltage increased significantly factors, in some cases outlining differences in the methods.
(up to 2.0 pu) during the fault. The larger variation in the dc (i) Fault duration - Longer faults causing different types of
voltage was observed for faults located close to the wind voltage drops were also considered, namely to 400 ms and up
farm. to 1000 ms. It was found that the ESS method had a good
(v) Crowbar control and response - The crowbar method response in the different regions of the LVRT curve, giving
had a fairly consistent fault response, both in term of wind results well inside the curve. The crowbar method had a good
farm output power and terminal voltage. When the crowbar is response, similar to the ESS method for the severe faults
in operation, the DFIG behaves as a squirrel-cage induction outside the LVRT curve. However, for the less severe faults
machine. Hence the operating conditions may vary within LVRT curve, the crowbar protection was activated and
significantly. However, the rotor current is well managed. The attempts to resume normal operation occasionally failed. As
dc bus voltage exhibited some oscillations, mostly due to the such its performance was not as good as the ESS method
crowbar being removed before the fault was cleared. The loss during longer fault within the LVRT curve. The flux method
of PQ control also resulted in reactive power consumption, had good response for long faults within the LVRT curve. Its
hence a slower return of the terminal voltage to nominal protection was activated quickly, resulting in a decrease in
 6

generation. For severe and very unlikely fault (three phase) of conditions to attenuate electric power fluctuation on the grid,
1000 ms, the recovery was difficult, leading to large due to wind variations [6]. The use of ESS however requires
oscillations. that practical issues be considered (cost, size, and life cycle).
(ii) Wind speed and turbine output power - It was observed The rotor-side converter needs to be sized accordingly.
that all three protection methods were needed more often (ii) Crowbar – The control of the operation of the crowbar,
when the wind farm was operated at full power, which is for and the impact of the setting of the point at which it is
high wind conditions. Under low wind conditions the operated should be considered. The possibility of multiple
machines are operated farther away from their limits and crowbar operations should be managed.
handling faults resulted in smaller variations of turbine and (iii) Flux – Overvoltages can be present on the dc bus.
system variables. In particular, it took more time for the These can be handled using a dc bus dump load (dc bus
monitored variables to reach their limits. In many cases the crowbar) or an energy storage system (ESS).
limits were not reached, and the protections were not (iv) Reactive power management – Reactive power can be
activated. Under low wind speed, the protections were mainly supplied by both the machine stator and the grid side
activated for faults closer to the wind farm. converter. Additional reactive power, particularly on a short
term basis, can be injected into the grid from the grid-side
VIII. DISCUSSION converter to help voltage recovery after fault clearing
The analysis of the simulation results, such as shown in (dynamic VARs).
Fig. 6, allow the following conclusions to be drawn:
(i) The three methods can meet typical LRVT requirements IX. CONCLUSIONS
(Figs. 1 and 2). The wind farm can remain connected to the This paper compares the performance of three LVRT
grid in all cases examined. solutions for DFIG based wind generators operating within a
(ii) The voltage during the recovery period remains within wind farm connected to a transmission grid. The electric
the acceptable range specified by Fig. 1, for under voltages as power network is subjected to a variety of faults, both
well as for overvoltages. Overvoltages are an important issue balanced on unbalanced. The wind farm is operated under
in the design of the turbine generator LVRT equipment and different wind conditions.
associated control and should be managed appropriately. The results show that all three methods meet the LVRT
Overvoltages generated by the turbine can affect not only the requirement and allow the wind farm to remain connected to
turbine but other equipment connected to the grid at the PCC. the grid for low voltage events up to 1000 ms, depending on
(iii) The generator rotor speed remains under control, that the type of fault. Generator acceleration is controlled and the
is the overspeed, if it occurs, is within acceptable limits. speed remains within limits.
(iv) The wind farm internal variables generally remain The three methods however differ in the manner in which
within converter and machine ratings. the external and internal wind turbine variables evolve over
In general terms, the three methods differ in the evolution time. There may also be a need to add supplemental
of the external variables (voltage response, real and reactive equipment and controls to limit the variations of some of the
power injection) and internal variables (converter and system variables to acceptable values, among others the dc
machine currents, dc bus voltage) during low voltage events. bus voltage.
Some of the variables, namely the dc bus voltage, may Overall, the ESS offers a better performance for the wind
need to be limited. This can be done by using a dc bus turbine, allowing a tighter control of real and reactive power
crowbar, consisting of switched resistors, which is activated and controlled recovery of voltage after fault clearing.
when the voltage exceed a predetermined value (typically 1.2 Oscillations in the farm output power and voltage may need to
pu). This is pertinent in the case, for example, of the flux be damped by proper design of the control loops.
control method. The crowbar method is the easiest to manage and offers
Based on the observations of the simulation results, the the more consistent operation, predictably limiting the rotor
following general comments can be made on the impact of a current. Issues to be considered include reactive power
particular method on system and component performance: consumption after fault clearing and the possibility of
(i) Energy Storage System (ESS) - There is an advantage in repeated unwanted crowbar operations.
using an energy storage component (battery or The flux method was successful in most cases in limiting
supercapacitor) instead of a simple dump resistor to store, as the rotor currents. Control of turbine variables for operation
the energy produced by the turbine during faults can be stored in the case of faults close to the PCC may be problematic,
and later released. A combination of storage and dump including possible oscillations in the terminal voltage and
resistance can be used. In addition to dissipating turbine output power, and dc bus overvoltages.
power, the dump resistance can be used to control the dc bus
overvoltages. The energy stored in the ESS can be released X. ACKNOWLEDGMENT
when the grid recovers to accelerate the return to normal The author gratefully acknowledges the contributions of J.
operation. The ESS can also be used under steady state Morneau, S. Nowak and C. Abbey, from the McGill
 7

University Electric Energy Systems Research Laboratory, to

this work.

XI. REFERENCES
[1] I. Erlich, W. Winter, and A. Dittrich, “Advanced grid requirements for the
integration of wind turbines into the German transmission system,” in 2006
IEEE Power Engineering Society General Meeting, 7 pp., June 2006.
[2] I. Erlich, H. Wrede and C. Feltes, “Dynamic Behavior of DFIG-Based
Wind Turbines during Grid Faults,” in 2007 Power Conversion
Conference PCC'07, pp. 1195-1200, Nagoya, April 2007.
[3] J. Niiranen, “Voltage Dip Ride Through of a Doubly-fed Generator
Equipped with an Active Crowbar,” in Nordic Wind Power Conf.,
Chalmers University of Technology, Mar. 2004
[4] J. Morren and S. W. H. de Hann, “Ride Through of Wind Turbines with
Doubly-fed Induction Generator during a Voltage Dip,” IEEE Trans.
Energy Conversion, vol. 20, no. 2, pp. 435–441, June 2005.
[5] B. Xie, B. Fox, and D. Flynn, “Study of fault ride-through for DFIG based
wind turbines,” in IEEE DRPT Conf., Hong Kong, Apr. 2004.
[6] C. Abbey, W. Li, and G. Joos, “Power Electronic Converter Control
Technique for Improved Low Voltage Ride through Performance in Wind
Turbine Generators,” in IEEE Power Electronics Specialists Conf. PESC'
2006, June 2006.
[7] C. Abbey and G. Joos, “Effect of low voltage ride through (LVRT)
characteristic on voltage stability,” in IEEE Power Electronics Specialist
Conference PESC'2005, San Francisco, June 2005.
[8] D. Xiang, L. Ran, P. J. Tavner, and S. Yang, “Control of a Doubly fed
Induction Generator in a Wind Turbine During Grid Fault Ride-through,”
IEEE Trans. Energy Conversion, vol. 21, no. 3, pp. 652–662, Sept. 2006.
[9] I. Erlich, and F. Shewarega, “Modeling of Wind Turbines Equipped with
Doubly-Fed Induction Machines for Power System Stability Studies,” in
2006 Power Systems Conference and Exposition PSCE '06, pp. 978 –
985, Oct.-Nov. 2006.
[10] C. Abbey, J. Morneau, J. Mahseredjian, and G. Joos, “Modeling
Requirements for Transient Stability Studies for Wind Parks,” in IEEE
2006 Power Engineering Society General Meeting, 6 pp, 18-22 June
2006.
[11] S. Jiang, U. D. Annakkage, and A. M. Gole, “A platform for validation of
FACTS models,” IEEE Trans. Power Delivery, vol. 21, no. 1, pp. 484–
491, Jan. 2006.

XII. BIOGRAPHY
Geza Joos (M’82, SM’89, F’06) graduated from
McGill University, Montreal, Canada, with M.Eng.
and Ph.D. degrees in Electrical Engineering.
He has been a Professor with McGill University
since 2001, and holds a Canada Research Chair in
Power Electronics applied to Power Systems (2004).
He is involved in fundamental and applied research
related to the application of high-power electronics to
power conversion, including distributed generation and
wind energy, and to power systems. His previous
employment includes ABB, the Ecole de technologie supérieure and Concordia
University. He has been involved in consulting activities in Power Electronics
and Power Systems, and with CEA Technologies as Technology Coordinator of
the Power Systems Planning and Operations Interest Group.
He is active in a number of IEEE Industry Applications Society committees,
including Chair of the Industrial Power Converter Committee and in IEEE Power
Engineering Society working groups. He is a Fellow of the Canadian Academy
of Engineering and of the Engineering Institute of Canada.