Eknath Vittal - The Impact of Reactive Power From Wind Generation On Power System Stability

The Impact of Reactive Power From Wind
Generation on Power System Stability

Eknath Vittal (B.S., M.Sc.)
A thesis submitted to the
School of Graduate Studies
in fulfilment of the
requirements for the degree of
Philosophi Doctor
University College Dublin
College of Engineering, Mathematical & Physical Sciences
School of Electrical, Electronic & Mechanical Engineering
Supervisor:
Dr Andrew Keane
Co-supervisor and Nominating Professor:
Prof. Mark OMalley
September 2011
Abstract
Wind generation in power systems is becoming an increasingly prevalent resource
as the push to cut CO
2
emissions and reduce fossil fuel dependency grows. To securely
accommodate the growing penetration levels of wind generation into power systems,
extensive stability studies must be completed. Power systems are complex non-linear
systems that rely on the synchronism and control systems provided by conventional
synchronous generators. As wind penetration increases, it will be the lack of control
that will cause the most signicant impact on power system stability. Wind generation
presents challenges in that not only is it inherently variable and uncertain, but it also
needs to be properly utilized from a mitigation perspective to maintain system stability
as conventional generation is displaced. The purpose of this thesis is to develop a series
of methodologies that will treat wind generation as an integral part of the power system,
and to assess the stability impacts wind generation will have. By applying fundamental
analysis techniques along with novel methodologies, the voltage stability and rotor angle
stability of power systems with large penetrations of wind generation will be assessed.
Traditionally, the voltage stability and rotor angle stability have been analyzed as
separate and distinct issues. This is due to the presence of automatic voltage regulation
in conventional synchronous generators. When synchronous generation is displaced by
wind, the loss of this regulation capability will have signicant impacts, particularly on
the voltage stability of the system, which will directly impact the rotor angle stability
of the remaining conventional synchronous units present in the system. This thesis will
rst show that the voltage stability of the power system can be signicantly improved
if reactive power control from wind generation is utilized to control the bus voltages of
transmission systems. This control is adept at improving system voltage security when
coupled with the variable and uncertain nature of the active power injections from wind
generation.
Next, it will demonstrate that by maintaining robust voltages, the rotor angle sta-
bility of conventional generators is improved. By achieving balanced scheduled voltages
at buses across the system, the likelihood of under-excitation of a synchronous gener-
ator is greatly reduced. Consequently, when a contingency occurs, the eld voltage
of the machine is better supported and angular separation is minimized. This greatly
I
improves the rotor angle stability of the system, and synchronism is maintained across
the network.
Finally, a strategy is developed here to identify locations in the network that will be
critical for the system to maintain voltage stability. By supporting the voltage at these
selected locations using wind generators, the voltage stability and rotor angle stability
can be maintained while minimizing the levels of control required by the system. This
will be particularly benecial in systems where 100% of wind generation present in the
system cannot be controlled.
The novel methodologies presented in this thesis will contribute to the continued
secure and reliable operation that is required of modern power systems while achiev-
ing government targets for emission reduction and increased sustainability in power
systems.
II
Contents
Abstract I
Acknowledgements VI
List of Publications VII
List of Acronyms VIII
List of Figures IX
List of Tables XI
1 Introduction to Power System Stability and Wind Generation 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Power System Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Wind Turbine Models and Impacts . . . . . . . . . . . . . . . . . . . . . 5
1.4 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2 Small-Disturbance, Long-Term Voltage Stability 9
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Common Analyses Techniques . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.1 Voltage Stability Mitigation Techniques for Wind Generation . . 12
2.3 The Time-Series Power Flow . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3.1 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3.2 The Irish Electricity System Model . . . . . . . . . . . . . . . . . 20
2.3.3 Implementation of Voltage Control on the Irish Power System . . 24
2.3.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . 27
2.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3 Large Disturbance, Short-Term Voltage Stability 35
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
III
3.2 Wind Turbine Modeling Considerations . . . . . . . . . . . . . . . . . . 36
3.3 The Impact of Wind Turbine Types on Short-Term Large-Disturbance
Voltage Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4 Rotor Angle Stability With High Penetrations of Wind Generation 42
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.2.1 Active Power Analysis . . . . . . . . . . . . . . . . . . . . . . . . 45
4.2.2 Reactive Power Analysis . . . . . . . . . . . . . . . . . . . . . . . 47
4.3 Test System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.4.1 Active Power Results and Discussion . . . . . . . . . . . . . . . . 50
4.4.2 Reactive Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.4.3 Fault Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5 A Robust Reactive Power Control Strategy for Wind Generation 67
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.2.1 Small-Signal Analysis . . . . . . . . . . . . . . . . . . . . . . . . 69
5.2.2 Transient Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.2.3 Voltage Stability Comparison . . . . . . . . . . . . . . . . . . . . 74
5.3 Test System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.4.1 Small-Signal Analysis . . . . . . . . . . . . . . . . . . . . . . . . 76
5.4.2 Transient Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.4.3 Impact of Wind Power Output Level . . . . . . . . . . . . . . . . 81
5.4.4 Voltage Stability Analysis . . . . . . . . . . . . . . . . . . . . . . 83
5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
6 Conclusions and Scope for Future Work 86
6.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.2 Scope for Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
References 92
IV
The outlet?
Thats where the electricity comes out.
Oh, you mean the holes.
Conversation between G. Costanza and S. Pete
V
Acknowledgements
First and foremost, I would like to thank my parents for their constant support
throughout my education. They have always pushed me to strive for more and have
provided constant encouragement for not only the past four years, but throughout my
life.
My mother Sunanda, has been there to proofread papers, provide encouragement
and support me throughout my time in Ireland. I will always be grateful for her kind
words and unyielding support.
My father Vijay, himself somewhat well known in the eld, has always expected the
best from me and always pushed me to achieve more. He has always been there for
support and has been a great sounding board for discussion and advice.
These two individuals have guided me throughout the process of writing this thesis
and I owe everything that I have accomplished to them.
Next, I would like to thank Prof. Mark OMalley. Meeting him while I was in
the process of completing my Masters has seemed like a profound stroke of luck in
hindsight. Without him I would not have come to Ireland and this thesis would not
have been written. His support and kindness is greatly appreciated and I thank him
greatly for providing me with this wonderful opportunity.
Dr Andrew Keane has been my primary supervisor throughout my time in Ireland.
He has been a great advisor and friend. His door was always open for discussions about
work, my future and any other random topic I could come up with. His advice and
guidance was invaluable and helped me achieve my goal of completing this thesis.
I would also like to thank my friends at the ERC: Pete Richardson for putting up
with my constant chats of links and whatnots that aided in my procrastination abilities,
Niamh Troy for putting up with none of that, Aonghus Shortt for our inordinately long
discussions about fantasy football and to all of my other colleagues and friends over
my four years at the ERC.
VI
List of Publications
I. Journal Publications
E. Vittal, M. OMalley, and A. Keane, A Steady-State Voltage Stability Analysis
of Power Systems with High Penetrations of Wind, IEEE Transactions on Power
Systems, vol. 25, no. 1, pp. 433-442, February 2010
A. Keane, L.F. Ochoa, E. Vittal, C.J. Dent and G.P. Harrison, Enhanced Utilization
of Voltage Control Resources With Distributed Generation, IEEE Transactions on
Power Systems, vol. 26, no. 1, Feb. 2011
E.Vittal, M. OMalley, and A. Keane,Rotor Angle Stability With High Penetrations
of Wind, IEEE Transactions on Power System, In Press
E. Vittal, M. OMalley, and A. Keane, A Robust Reactive Power Control Strategy for
Wind Generation For Improved Rotor Angle Stability, IEEE Transactions on Power
Systems, In Review
II. Conference Publications
E. Vittal, M OMalley, and A. Keane, Varying Penetration Ratios of Wind Turbine
Technologies for Voltage Stability and Frequency Stability, IEEE PES General Meet-
ing, Pittsburgh, PA, 2008
E. Vittal, M OMalley, and A. Keane, Impact of Wind Turbine Control Strategies on
Voltage Performance, IEEE PES General Meeting, Calgary, CA, 2009
E. Vittal, M OMalley, and A. Keane, A Small-Signal Stability Analysis of Wind
Generation, 8th International Workshop on Large-Scale Wind Integration into Power
Systems, Bremen, DE, 2009
E. Vittal, M OMalley, and A. Keane, A Steady State Voltage Stability Analysis of
Power Systems with High Penetrations of Wind, IEEE PES General Meeting, Min-
neapolis, MN, 2010
E. Vittal, L. Meegahapola, A. Keane, and D. Flynn, Optimal Allocation of Reactive
Power Resources to Minimize Losses and Maintain System Security, 9th International
Workshop on Large-Scale Wind Integration into Power Systems, Quebec, CA, 2010
III. Book Chapters
Power System Stability With High Penetrations of Wind Power, E. Vittal, A. Keane, H.
Slootweg, and W.L. Kling, Wind Power in Power Systems, Ed. Thomas Ackermann,
Wiley, 2011
VII
List of Acronyms
AVR - Automatic Voltage Regulation
DFIG - Doubly-Fed Induction Generator
FSIG - Fixed Speed Induction Generator
IVGTF - Integration of Variable Generation Task Force
PDF - Probability Density Function
PF - Power Factor
PLF - Probabilistic Load Flow
PMSG - Permanent Magnetic Synchronous Generator
PMU - Phasor Measurement Unit
PV - Power Voltage
NREL - National Renewable Energy Laboratory
SCADA - Supervisory Control and Data Acquisition
SNV - Summer Night Valley
TSPF - Time-Series Power Flow
VSC - Voltage Source Full-Convertor
VSWT - Variable Speed Wind Turbine
WP - Winter Peak
WTG - Wind Turbine Generator
VIII
List of Figures
1.1 Classications of Power System Stability (Kundur et al., 2004) . . . . . 4
2.1 Flow chart describing the methodology used in the analysis. . . . . . . . 15
2.2 The probability density function for instantaneous wind penetration for
2004 - 2008. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3 One-line diagram of the transformer connections for wind farms in the
all-island Irish power system. . . . . . . . . . . . . . . . . . . . . . . . . 26
2.4 The 20 kV bus PV curve for Case A, Oct. 31, 2006. Since the DFIGs
are operated at a 0.95 inductive PF, as the power output from the wind
turbine increases the bus voltage begins to drop. . . . . . . . . . . . . . 28
2.5 The 20 kV bus PV curve for Case B, Oct. 31, 2006. Now the PV curve
is drastically dierent. Terminal voltage control is enabled in all DFIGs,
the bus voltage is controlled to 1.075 PU. . . . . . . . . . . . . . . . . . 29
2.6 The PV curve for the 110 kV bus connected to the 20 kV bus for Case
A, Oct. 31, 2006. Here the voltage begins to decrease as power output
of the wind farm increases. . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.7 The PV curve for the 110 kV bus, but for Case B, Oct. 31 2006. The
increased control allows for improved voltages at higher power output
levels, implying an increased voltage stability margin. . . . . . . . . . . 30
2.8 A zoom of the nose of the PV curve from 110 kV bus for Case C, WP
2006 . Here at the same power output levels the voltages are higher even
when no control is implemented. . . . . . . . . . . . . . . . . . . . . . . 31
2.9 A zoom of the nose of the PV curve at the 110 kV bus from Case A, Oct.
31, 2006. As power output from the wind farm increases the voltage falls. 31
2.10 The PV curve when voltage control was implemented at the wind farm,
Case D, WP 2006. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.11 Here the voltage control is applied directly at the transmission level. The
target voltage is set at 1.05 PU and as a result all voltages are controlled
at or above that level regardless of loading scenario. . . . . . . . . . . . 33
3.1 The IEEE 30 bus test system. . . . . . . . . . . . . . . . . . . . . . . . . 39
IX
3.2 The IEEE 30 bus test system. . . . . . . . . . . . . . . . . . . . . . . . . 40
4.1 One-line diagram of the New England 39 bus test system. . . . . . . . . 49
4.2 Rotor angle traces from the synchronous wind case and the capacitive
case at generator 34 for the loss of generation contingency. . . . . . . . . 51
4.3 Active power response from the nine original synchronous units that are
still online for the synchronous wind case (solid line) and the capacitive
case (dashed line). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.4 Rotor angle traces for generator 34 from the capacitive case and the
unity case for the loss of generation contingency. . . . . . . . . . . . . . 54
4.5 Active power response from the nine original synchronous generators
online from the capacitive case (solid line) and the unity case (dashed
line). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.6 Rotor angle traces for generator 34 from the capacitive case and the
terminal voltage case for the loss of generation contingency. . . . . . . . 56
4.7 Active power response from the nine original synchronous generators
online from the capacitive case (solid line) and the terminal voltage case
(dashed line). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.8 The rotor angle of generator 34 for the three terminal voltage control
cases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.9 The reactive power output of generator 34 for the three wind cases. . . . 60
4.10 The voltage of bus 34 following the contingency for the three wind cases. 60
4.11 The eld voltage of generator 34 following the contingency for the three
wind cases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.12 The eld current of generator 34 for the three wind cases . . . . . . . . 63
4.13 The rotor angle of generator 33 following a 7 cycle fault at bus 33 for
the capacitive case and terminal voltage case. . . . . . . . . . . . . . . . 64
4.14 The reactive power collapse of generator 33 following the clearance of
the fault at bus 20. The limit on generator 33 is -300 MVAr. . . . . . . 65
4.15 The reactive power output of the wind farm at bus 33. . . . . . . . . . 65
5.1 The owchart describing the process used to identify the critical set of
wind farms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
5.2 The rotor angle for generator 34 following the loss of generator 32. . . . 79
5.3 The rotor angle for generator 34 following the loss of generator 32 at the
20.7% level. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
X
List of Tables
2.1 Critical Operating Points in the Irish System . . . . . . . . . . . . . . . 22
2.2 2013 Irish Power System Wind Characteristics . . . . . . . . . . . . . . 24
2.3 Wind Turbine Control Distribution . . . . . . . . . . . . . . . . . . . . . 25
3.1 Critical Clearing Time for 0.2j Fault . . . . . . . . . . . . . . . . . . . . 39
3.2 Minimum Bus Voltage During Faulted Condition . . . . . . . . . . . . . 40
4.1 Active Power Comparison Scenarios . . . . . . . . . . . . . . . . . . . . 47
4.2 Reactive Power Comparison Scenarios . . . . . . . . . . . . . . . . . . . 47
4.3 Wind Generation Locations . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.4 Dominant Mode Comparison . . . . . . . . . . . . . . . . . . . . . . . . 52
4.5 Rate of Decay Coecients . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.6 Dominant Mode for the Unity Case . . . . . . . . . . . . . . . . . . . . . 54
4.7 Rate of Decay Coecients for the Unity Case . . . . . . . . . . . . . . . 54
4.8 Dominant Mode for the Terminal Voltage Case . . . . . . . . . . . . . . 57
4.9 Rate of Decay Coecients for the Terminal Voltage Case . . . . . . . . 57
4.10 Dominant Mode for the WECC Wind Model Cases . . . . . . . . . . . . 59
4.11 Rate of Decay Coecients for the WECC Wind Model Cases . . . . . . 59
5.1 Studied Wind Control Cases . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.2 Wind Generation Locations . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.3 System Scenarios for Small-Signal Analysis . . . . . . . . . . . . . . . . 76
5.4 Modes and Participation Factors for the Displacement of Generator 30
for Scenario A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.5 Average Participation Factors for the Synchronous Wind Farms . . . . . 78
5.6 Reactive Power Production for the Synchronous Wind Farms . . . . . . 79
5.7 Prony Analysis of Generator 34 for the Synchronous Case at 41.3% . . . 80
5.8 Prony Analysis of Generator 34 for the Full Control Case at 41.3% . . . 80
5.9 Prony Analysis of Generator 34 for the Critical Control Case at 41.3% . 80
5.10 Prony Analysis of Generator 34 for the Synchronous Case at 20.7% . . . 82
5.11 Prony Analysis of Generator 34 at 20.7% . . . . . . . . . . . . . . . . . 83
XI
5.12 Critical Clearing Times at 41.3% for Fault at Bus 14 . . . . . . . . . . . 83
5.13 Critical Clearing Times at 20.7% for Fault at Bus 14 . . . . . . . . . . . 84
XII
CHAPTER 1
Introduction to Power System Stability and Wind Generation
1.1 Introduction
T
HE nature of modern power systems has changed due to a variety of factors:
the increased demand for sustainability, rises in the price of oil and the need
for the reduction of greenhouse gases, all of which have driven a large increase in the
level of wind generation in the power system. The Intergovernmental Panel on Climate
Change has cited that wind energy will be the primary source of renewable generation
in the electricity sector (IPCC, 2011). In both Europe and the United States, wind
generation is the dominant renewable resource currently present in power systems. In
Europe, wind energy is set to triple in penetration by the year 2020, with 15.7% of the
continents total energy provided by wind generation (EWEA, 2011). In the United
States, there is currently 42,432 MW of installed capacity providing 2.3% of the U.S.
electricity mix, with the number set to rise to 25% by the year 2025 (AWEA, 2011).
With wind generation set to become a signicant generation resource in power systems
1
Chapter 1. Introduction to Power System Stability and Wind Generation 2
around the world, it will become increasingly important to fully understand its impacts
and interaction with the conventional elements in power systems.
Fundamentally, power systems have been designed and operated around the concept
of generation delivery from large synchronous machines (Kimbark, 1956; Anderson and
Fouad, 2003; Machowski et al., 2008; Sauer and Pai, 1998). These machines have high
levels of reliability and complex control systems that allow the system to maintain
high levels of operational security. The correct operation and control of these machines
across the full spectrum of time-frames is critical for maintaining reliable power system
operation and stability (Steinmetz, 1920). The time-frames associated with power
systems occur over a variety of periods; long-term planning examines power systems
several months or even several years into the future. It focuses on how the power
system should be developed in order to accommodate specic types of generation or
how it will be expanded based on the availability of resources. Power system operation
studies are concerned with the day-to-day operation of power systems, focusing on
the commitment and dispatch of generating units in the system, how reserve resources
are determined, and a variety of other operational considerations. Both planning and
operational studies are important in achieving a highly reliable and securely operated
power system. The decisions coming out of these studies are the ones that will impact
the operational stability of the system. The stability of a power system will occur across
an operational time-frame of hours down to milliseconds (Van Cutsem and Vournas,
1998, 1996; Kundur, 1994). Maintaining power system stability requires that the various
components and elements of the system can interact without issue across all of the time-
frames of the stability spectrum. Wind generation will have signicant impact across
the power system stability time-frame and as wind generation becomes a more common
source of generation in the system, new mitigation techniques will be necessary in order
to continue operating the power system in a secure and stable manner.
1.2 Power System Stability
Power system stability is the broad denition given to the wide variety of criteria
required for a power system to operate safely and securely while minimizing the risk of
blackout. Maintaining power system stability for any contingency is the fundamental
goal of power system operation, which will be challenged with the introduction of
increasing levels of wind generation into the system. Traditionally, power systems have
consisted of large synchronous generating units with control systems in place that aid in
system operation. Synchronous generating units are synchronized with other machines
in the system, which rotate together at 50 Hz or 60 Hz, depending on the system
(Kimbark, 1956; Steinmetz, 1920) . This synchronism is the fundamental driver of
alternating current (AC) power system operation. Three-phase power, the control of
reactive power and inertial response, are all a result of synchronous power generation. In
general, power system stability is divided into three distinct classications of stability:
Voltage Stability
Rotor Angle Stability
Frequency Stability
These three areas of stability are further broken down into the type of event that
occurs, i.e. whether there is a small disturbance or a large disturbance, and the time-
frame in which the event occurs, either short-term or long-term. A ow chart describing
the classications of power system stability is given in Fig. 1.1
This thesis will focus on the voltage stability and rotor angle stability of power sys-
tems, in particular how they interact with high penetrations of wind. Although, voltage
stability and rotor angle stability are classied as two distinct aspects of power system
stability, the changing nature of the power system results in an increased coupling be-
tween the two concepts. As the presence of asynchronous generation coming from wind
generation grows, the link between voltage stability and rotor angle stability increases.
Figure 1.1: Classications of Power System Stability (Kundur et al., 2004)
Traditionally, these two areas of stability have been isolated and independent in mod-
ern power systems due to the presence of automatic voltage regulation (AVR) (Taylor,
1993; Kundur, 1994; Anderson and Fouad, 2003; Machowski et al., 2008; Sauer and Pai,
1998). AVR control mitigates many of the voltage issues that arise and impacts the
rotor angle stability of synchronous generation units by controlling their reactive power
injection into the system. As wind penetration levels increase, synchronous generation
and their accompanying AVR capability will be displaced. This will have signicant
impacts on both the voltage stability and rotor angle stability of the power system.
The displacement of conventional synchronous generation will lead to a decrease in the
ability to regulate and control the reactive power in the system. This lack of control will
have signicant impacts on both the voltage stability and rotor angle stability of the
system, particularly if the new wind generation is not controlled appropriately. As the
nature of the power system changes, it will become increasingly critical for wind gener-
ation to provide the necessary support and control, particularly from a reactive power
standpoint for continued secure system operation (EirGrid, 2010a). Overall, wind gen-
eration can be a valuable resource to the power system if it is controlled appropriately,
specically if wind can support the reactive power ows in the system.
1.3 Wind Turbine Models and Impacts
The analysis methods and techniques for determining power system stability are well
established in Kundur et al. (2004) for systems that consist mainly of large synchronous
generating units. As power systems move further into the 21
st
century, the push for
greener more ecient systems is growing. In response to rising CO
2
emissions, one
of the solutions is the installation of large levels of wind generation, most often an
asynchronous source of generation. This means wind generation operates at a speed
independent of the system frequency (50 Hz or 60 Hz), resulting in signicant impacts on
system stability. Wind generation, unlike conventional generation is also highly variable
and uncertain, which implies the need for changes in the way systems are operated and
controlled. The Integration of Variable Generation Task Force (IVGTF) along with the
National Renewable Energy Laboratory (NREL) have identied several key aspects of
power system operation that will need to adapt to accommodate higher penetrations of
wind generation (NERC, 2010; NREL and EnerNex, 2011; NREL and Energy, 2010).
Forecasting of wind power output and loads will require increased accuracy (Matos and
Bessa, 2011). The operation and dispatch of conventional generation will need to be
adapted to the variable nature of wind generation (Tuohy et al., 2009; Meibom et al.,
2011). Ancillary services will need to provide the necessary levels of reserve in order
to maintain system security in case wind generation is not available (Ding et al., 2011;
Soder, 1993). System planning and the selection of plants will have to take into account
the nature of wind generation as well (Ummels et al., 2007; Bakirtzis and Dokopoulos,
1988). These issues focus on longer time-frames of study and will not directly impact
the stability of power systems. Rather, the changes in power system stability that
occur in the short-term operational time-frame, that arise as a result of increased wind
penetration will be the focus of this thesis. Large penetrations of wind generation will
have signicant impacts across all three classications of power system stability. To
properly assess how wind generation will impact modern power systems, it is important
to understand how they operate (Coughlan et al., 2007). Wind turbines can be broken
down into four basic types of machines:
Type-I Wind Turbine Turbine (WTG): The Type-I machine is a xed-speed squir-
rel cage induction generator (FSIG) turbine. Power is generated at a nearly
constant speed by controlling the blade pitch angle of the turbine (Ackermann,
2005).
Type-II WTG: The Type-II machine utilizes a squirrel cage induction generator
with a variable rotor resistance to optimally produce power for varying rotor
speeds (Ackermann, 2005).
Type-III WTG: The Type-III machine is a doubly-fed induction generator (DFIG)
with power electronics that control the reactive power output of the machine. A
DFIG is an induction machine, where both sets of windings are excited. This
allows the machine to operate at super-synchronous speed and provides increased
power production over a variety of rotor speeds. This is ideal for wind generation
where the wind speed varies greatly; however the design of the Type-III machine
includes a rotor-side power electronic convertor that fully decouples the machine
for the power system. This means that reactive power can be controlled and power
is produced optimally over a variety of operational speeds, but the machine cannot
respond to changes in system frequency (Ackermann, 2005).
Type-IV WTG: The Type-IV machine can be either a DFIG or a type of syn-
chronous machine, often a permanent magnet synchronous generator (PMSG). It
uses a full-convertor to produce power over a variety of operational speeds rather
than a single rotor-side converter like the Type-III machine (Ackermann, 2005).
Type-I and Type-II WTGs were the rst generation of wind turbines installed in
power systems. They are still present, but a majority of the new installations in power
systems today are of Type-III or Type-IV. Based on this trend, the main focus the
analyses will be on the stability impacts of Type-III and Type-IV turbines, however,
it is also important to consider the impact of the Type-I and Type-II WTGs as well.
Therefore, the inuence of wind generation on the voltage stability and rotor angle
stability across both the short-term and long-term stability time-frames will be assessed.
Utilizing the traditional methods and techniques that are used to study conventional
synchronous generation, the signicant impacts of wind generation on power system
stability will be determined.
1.4 Outline
To examine stability impacts in detail, this thesis is organized as follows: Chapter
2 and Chapter 3 discuss the issues related to wind generation and voltage stability.
Chapter 2 focuses on the long-term voltage stability of a power system and how the
voltage security is impacted by type of reactive power control implemented by wind
turbines in the system. It utilizes a network model of the full Irish power system and
examine of how the variable and uncertain nature of wind generation will inuence
voltage stability. In particular, it will look at how the minute to minute variations of
wind generation and load can aect the voltage security margin of buses in the system,
and how that margin can be extended if reactive power injections from wind generation
are controlled to achieve a specic voltage.
Chapter 3 examines the short-term dynamic impacts that wind generation will have
on a power system. Rather than using a network model of the Irish system, Chapter
3 utilizes the IEEE 30 bus test system. The goal of Chapter 3 is to demonstrate how
the reactive power capabilities of variable speed wind turbines can benet the system
during a fault event and can help improve the critical clearing time of the network.
Chapter 4 introduces issues related to rotor angle stability in systems with high pen-
etrations of wind and will demonstrate how voltage stability and rotor angle stability
are becoming more coupled due to the introduction of wind generation in the system.
Again, Chapter 4 utilizes a test system, this time the New England 39 bus test system,
to demonstrate that the active power injected into the system by wind generation is
fundamentally dierent when compared to active power injected by conventional syn-
chronous generation. This dierence in active power places an increased burden on
the conventional synchronous units in the system which means that additional support
is necessary to maintain the rotor angle stability of the system. This support will be
delivered by wind generation in the form of reactive power control, specically terminal
voltage control. Chapter 4 demonstrates that by maintain consistent and balanced volt-
ages across the system, rotor angle stability of the conventional synchronous generation
in the system is improved.
Chapter 5 expands on the growing relationship between voltage stability and rotor
angle stability. With wind farm installations occurring at both the distribution and
transmission level, complete control of all farms in the system may not be possible.
Chapter 5 presents a methodology that will examine the network characteristics to
determine critical nodes in the system that require reactive power control. By only
controlling voltage in specic locations in the network, Chapter 5 will demonstrate that
complete control of all wind generation, while benecial to the system, is unnecessary
and high levels of voltage stability and rotor angle stability can be maintained through
the control of the critical farms in the system.
Although modern power systems are constantly evolving, the stability concepts asso-
ciated with renewable generation are fundamentally the same as for systems in the past.
The analysis techniques established over the years can be applied to the new resources
that are a part of power systems today and help maintain system security and oper-
ational stability. In order to continue secure service and power delivery to customers,
power systems will have to adapt their control methods and strategies, particularly
in the area of reactive power control. This thesis will demonstrate the impact reac-
tive power control can have on power system stability and develop methodologies and
strategies that can improve system operation and stability as wind penetration levels
continue to grow in power systems across the world. The overall goal is to demonstrate
the changing nature of the power system and as wind penetration levels increase, novel
methods and strategies will need to be implemented in order to maintain the high levels
of security that are required of power systems.
Chapter 6 will conclude the thesis and provide a scope for future work.
CHAPTER 2
Small-Disturbance, Long-Term Voltage Stability
2.1 Introduction
F
ROM Kundur et al. (2004) voltage stability is dened as the ability of a power
system to maintain steady voltages at all the buses in the system after being
subjected to a disturbance from a given initial operating condition. As seen in Fig. 1.1,
voltage stability is broken down into further distinct categories:
Large-disturbance voltage stability is the ability of the system to maintain voltage
stability following a signicant event such as a loss generation, system faults, or
circuit contingencies (Kundur et al., 2004).
Small-disturbance voltage stability is the ability of the system to maintain steady
voltage as a result of small variations in load and generation (Kundur et al., 2004).
These two types of stability are then studied in two time-frames:
9
Chapter 2. Small-Disturbance, Long-Term Voltage Stability 10
Short-term voltage stability involves the dynamics of fast-acting load components,
electronically controlled loads, and HVDC convertors. The analysis of short-term
voltage stability events are carried out in the time-domain (Kundur et al., 2004).
Long-term voltage stability involves the slower acting equipment in the power
system such as tap-changing transformers, thermostatically controlled loads, and
generator current limiters. The analysis of long-term voltage stability events are
carried out in the steady-state, generally through power ow analysis (Kundur
et al., 2004).
This chapter will focus on the small-disturbance, long-term voltage stability of the
power system. Chapter 3 will discuss the issues associated with the large-disturbance,
short-term voltage stability of the power system.
Maintaining voltage stability is critical for secure operation of a power system, and
when new elements are introduced into the system it is necessary to run detailed studies
to assess any changes in system security. The introduction of wind generation into
power systems complicates determining the security of a power system due to the
fact that wind generation is both variable and uncertain. In the steady-state, wind
generation will have constant uctuations as the wind speed changes and the power
outputs in dierent regions of the system will vary based on the atmospheric conditions.
How the system responds to these constant variations, not only from wind generation
output but the small changes in the load as well, will be what determines the small-
disturbance, long-term voltage stability of the system.
Traditional power ow studies are completed for a system around a single operating
point at a time when the system is most stressed. This becomes dicult with high
penetrations of wind generation since the time when the system is most stressed may
not coincide with a traditional study scenario. This is due to the variability of wind.
There have been signicant advances that incorporate the variability of wind into power
ow studies. These are described in the next section.
2.2 Common Analyses Techniques
I
N order to properly assess the voltage stability and in particular the voltage stability
margin, a detailed AC power ow analysis of the transmission system is necessary.
However, there are several issues that arise when completing power ow studies involv-
ing wind generation. Power ow studies have traditionally focused on a single operating
point in the system. However, challenges arise in assessing the true impact of wind gen-
eration on a power system. In particular, the variable nature of wind necessitates new
techniques to assess its impact.
The use of statistical techniques to analyze power systems is a well established
concept known as probabilistic load ow (PLF), and has been successfully modied
to model wind. The foundation for PLFs was established in (Borkowska, 1974) and
showed how transforming the input variables into random variables (RV), a resulting set
of output RV can be achieved. Generally, the form of both the input and output RV is
given as a probability density function (PDF), and will relay information about several
operational aspects of the power system. In (Allan and Leite da Silva, 1981; Leite da
Silva and Arienti, 1990; Leite da Silva et al., 1990), the focus was on the traditional
operation of power systems, and required non-linear optimization methodologies in
order to achieve a solution. As a result additional techniques must be implemented in
PLFs when incorporating a variable resource such as wind.
In (Billinton and Wangdee, 2007) and (Billinton et al., 1996), PLFs were used to
assess system reliability with large levels of wind generation, while (Burke and OMalley,
2008) examined the transmission planning aspect of power systems by incorporating
a sequential time-series along with a PLF in order to maximize the rm connection
of wind generation into the system. In (Bakirtzis et al., 1989) and (Hatziargyriou
et al., 1993) the aspects of probabilistic load modeling and the incorporation of wind
generation in power systems was examined. Since in PLF analyses, the historical data
input is treated as a RV, it will include only the probability that the worst-case point will
be captured during the simulation and that the system is secure for all contingencies.
In comparison to PLFs, time-series load ow simulations will deterministically model
the power system and will explicitly capture the system response at the worst-case
point. Studies have previously applied the time-series approach to model power sys-
tems. In (Boehme et al., 2007), it was shown how time-series power ows can be
applied to systems to determine overload conditions and specify non-rm connection
agreements for new generators on the distribution system. In (Thomson and Ineld,
2007), a time-series analysis was implemented in modeling variable resources such as
solar photovoltaic and gas-red micro-CHP (cogeneration) in low voltage networks.
Both studies examined the application of historical time-series data on the distribution
level of the power system and the use of time-series data will be critical in the future
for the analysis of power systems with high penetrations of wind.
2.2.1 Voltage Stability Mitigation Techniques for Wind Generation
T
HIS section will examine what the available mitigation techniques are for wind
turbines in power systems today. Since the Type-III WTG is the predominant
technology installed in wind farms, this chapter will focus on the changes in a power
systems steady-state voltage stability in response to an increase in DFIG wind genera-
tion. The main advantage of the DFIG turbine is the ability to provide reactive power
control without installing additional capacitive support. The DFIG can be operated
in one of two control modes; rstly, xed power factor (PF) control, where the turbine
controls reactive power production in order to achieve a specied power factor. Sec-
ondly, terminal voltage control, where the reactive power is controlled to meet a target
voltage. Using these two control schemes this chapter will assess the impact of DFIG
reactive power control on the systems voltage stability margin.
Various technologies and strategies have been developed in order to implement ter-
minal voltage control. Work in Kayikci and Milanovic (2007) has shown the capabilities
of reactive power control in DFIGs using combinations of grid side control and rotor
side control. In Akhmatov and Eriksen (2007), a case study on the Danish power
system demonstrated the importance of reactive power and voltage control in main-
taining system stability. In Tapia et al. (2007) PI-based control algorithm is described
and implemented to manage the reactive power out of a DFIG wind farm. A coordi-
nated voltage control strategy is applied using a DFIG wind farm in Cartwright et al.
(2004a). In Xu and Cartwright (2006), a novel algorithm for direct active and reactive
power control was implemented. In Ackermann (2005), several generalized methods of
reactive power control are provided.
Work in Vittal et al. (2009) showed that increased voltage control will improve
the probability that bus voltage will lie within a specied range and improve voltage
performance within the power system. Voltage performance refers to achieving desired
voltages within a specied operating range, and while the improvement in voltage
performance indicates increased system robustness, it is not a true measure of the
power systems stability. In assessing a systems stability, a measure such as a power-
voltage (PV) curve is a much better indicator of voltage stability (Kundur, 1994). Using
a time-series data set for wind speeds and loads for multiple years, this chapter will
analyze the data a priori, and build a sequential simulation around the single worst-
case point contained in those years, called a time-series power ow (TSPF) for a future
power system at the transmission level.
By utilizing time-series data as the input, the worst-case point within the data set
will be deterministically modeled. This allows the simulation to maintain the correla-
tion between the wind speeds and load levels seen throughout the year. Since this is a
sequential simulation that models the variability of wind generation in a power system,
it is necessary to maintain the balance between the changing generation and the load
in the system. This will be achieved by re-dispatching the conventional units in the
system using a merit-order economic dispatch. It is also crucial that the correct units
are scheduled to be online during the period of simulation. As such, a unit commitment
is required to determine the online plants during the worst-case point, while taking into
consideration the forced outage rates and availability of the units.
By incorporating AC power ow along with economic dispatch and unit commit-
ment, this chapter will produce a realistic simulation of a transmission system that
captures the variable behavior of the wind energy resource that will provide insight
into the systems steady-state voltage stability.
2.3 The Time-Series Power Flow
2.3.1 Methodology
T
HIS section will develop a methodology for completing a TSPF analysis based
on an analysis of time-series wind and loading data, and is described in the ow
chart presented in Fig. 2.1. Each element in the ow chart represents a critical aspect
of the methodology and will be described further in the following subsections. The
goal of this chapter is to establish a methodology that will properly assess the voltage
stability of the system while taking into account the variable nature of wind generation.
_
_
P
1
Q
1
.
.
.
P
n
Q
n
_
_
=
_
_
P
1
1
P
1
|V
1
|
. . .
P
1
n
P
1
|V
n
|
Q
1
1
Q
1
|V
1
|
. . .
Q
1
n
Q
1
|V
n
|
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
P
n
1
P
n
|V
1
|
. . . . . . . . .
Q
n
1
Q
n
|V
1
|
. . . . . . . . .
_
_
_
1
V
1
.
.
.
n
V
n
_
_
(2.1)
|V| =
_
_
|V
2
|
.
.
.
|V
n
|
_
_
(2.2)
=
_
2
.
.
.
n
_
_
(2.3)
The analytical description of the power ow used in these simulations is given in
(5.1), (Wood and Wollenberg, 1996). The P
i
and Q
i
values are updated every time-step
Figure 2.1: Flow chart describing the methodology used in the analysis.
from t
init
, the starting time of the simulation, to t
final
, the ending time of the simulation
time-period. The values of P
i
and Q
i
are used to solve for voltage, V, (5.2) and angle,
, (5.3). It should be noted that in (5.1) the initial voltage and angle value used to
solve the power ow come from the previous time-step, t-1. This aids in convergence
and simulation time and is completed for the n buses in the system.
2.3.1.1 Importance of Time-Series Analysis and Historical Data
Traditionally, the worst case operating point of the system generally occurs when the
transmission system is most stressed (maximum load) or when generation is at a mini-
mum and system voltages are much lower. These two scenarios are when transmission
overloads and low voltage collapse are most probable. Prior to wind becoming a signif-
icant proportion of power system generation portfolios, the worst case operating point
was easily identied based on traditional weather and loading patterns.
The methodology established in this chapter approaches the problem from a dif-
ferent perspective in comparison to PLFs. By analyzing sucient time-series data
pre-simulation, the data itself is reduced to capture the single worst-case operating
point and simulated around that point for a shorter period of time. This reduction
of data reduces computation time, and allows for a robust assessment of the power
systems voltage stability under high penetrations of wind generation. It is important
to make sure that the loading and wind power output data are chronologically synchro-
nized in order to ensure it captures the complex underlying relationship between wind
and load. By identifying the worst-case point for multiple years of data and directly
inputting that into a TSPF, a more thorough analysis is possible and a true measure
of a systems stability and robustness can be realized.
In high wind penetration systems, the worst-case operating point for voltage stabil-
ity studies occurs when wind generation serves the largest proportion of the systems
demand, and system stability support mechanisms are at a minimum. Transmission
overload studies would focus on the point where wind generation and demand were the
greatest and the transmission system is the most stressed. Frequency stability studies
would focus on the period when spinning reserve and inertial support is at a minimum.
2.3.1.2 Resource Analysis and System Model Setup
The improvement in bus voltage performance due to the addition of wind generation, in
particular when utilizing the terminal voltage control capability on DFIG wind turbines,
is a highly localized phenomena. This is due to the fact that the terminal voltage control
will most signicantly impact the region in which the wind generation is located. More
signicant than the locality of the control, is the power output dependence on the wind
speeds of a particular region. Using the same average wind power output across an
entire system will not capture the true variability of the resource as one area of the
system will not see the same wind speeds as another. As a result the use of regionally
specic wind power output data for dierent areas of the system is required.
Using an appropriate resource assessment will aid in the placement of wind farms in
the test system and provide the highest level of accuracy in the results of the simula-
tions. A thorough assessment will be based on the transmission capacity of the system
as well as the availability of high annual wind speeds. By utilizing a resource assessment
in conjunction with geographically diverse wind power output data the most realistic
simulation can be achieved that captures the correlation between wind and load for
any given time.
The resource analysis will provide wind power outputs and allow for the calculation
of new power output levels from the farms in the system. In (5.4), the regional wind
power output data, P
region
, is used with the installed wind capacity, C
wind
, to build
wind power matrix, P
wind
. This matrix is updated every time-step and represents the
variability of the wind resource.
_
_
P
wind
1
.
.
.
P
wind
n
_
_
=
_
_
P
region
C
wind
1
.
.
.
P
region
C
wind
n
_
_
(2.4)
2.3.1.3 Balancing Load and Variable Generation
The main diculty in achieving a realistic simulation that incorporates signicant levels
of wind generation is capturing the variability of the wind resource. Using historical
time-series data captures the variability of the wind however it presents signicant
challenges in an AC power ow model.
As wind farm power output and system loading changes at each time-step in the
simulation, action must be taken in order to maintain system-wide load/generation
balance. In a TSPF, the commitment and power output of the conventional generation
units are inter-temporally dependent, i.e. the past state of the unit will impact any
future state. However, there are two distinct levels of this inter-temporal dependence;
the rst is between each individual time-step. As wind generation and system loading
are updated continuously, load/generation balance will need to be maintained using
an economic dispatch algorithm (Wood and Wollenberg, 1996). This allows the online
conventional generation to ramp up or ramp down output levels in order to achieve a
system balance.
The second issue arises in the determination of which conventional units are online
and available to ramp their output levels. This is dealt with by using the wind power
output and loading data in conjunction with a unit commitment algorithm to determine
a commitment schedule (Baldick, 1995). This takes into account the minimum start-up
times and up and down times of the generating units and makes sure that a generator
does not start-up or shut-down outside of its operating limitations. A unit commitment
can be completed as often as necessary based on the make-up of the systems genera-
tion portfolio. Using economic dispatch and unit commitment together will facilitate
load/generation balance within the system as wind generation varies across the system.
By using (5.4), in conjunction with an economic dispatch, (ED), and unit commit-
ment algorithm, (UC), the power output from the conventional units in the system is
determined, (5.5). In (5.5), f represents a function of the UC, ED, and P
wind
. Next,
by combining the power outputs from P
conv
and P
wind
, a matrix containing the power
generated at all units in the system is achieved, (5.6).
_
_
P
conv
1
.
.
.
P
conv
n
_
_
= f
_
_
_
_
_
UC, ED,
_
_
P
wind
1
.
.
.
P
wind
n
_
_
_
_
_
_
_
(2.5)
_
_
P
1
.
.
.
P
n
_
_
=
_
_
P
wind
1
.
.
.
P
wind
n
_
_
+
_
_
P
conv
1
.
.
.
P
conv
n
_
_
(2.6)
Utilizing (5.6) and the desired terminal voltage set-points for the voltage controlled
buses in the system, V
term
, the reactive power levels at all the machines in the system
can be determined, (2.7). V
term
will vary based on the type of voltage control algorithm
implemented. If the operator desires to achieve secondary or tertiary voltage control a
broader area-wide or regional control scheme must be implemented (El Moursi et al.,
2008; Ilic et al., 1995). In (2.7), g represents a voltage control algorithm that behaves
similar to a simple excitation system, (Bergen and Vittal, 2000), found in conventional
synchronous machines. Voltage is controlled to a specied control level, and imple-
mented locally to achieve a specic target voltage at a particular bus. Finally, the
determination of the active power, (5.6), and the reactive power (2.7), allows for the
solution of (5.1), and the determination of the voltage levels, (5.2), and angles, (5.3),
for all the buses in the system.
_
_
Q
1
.
.
.
Q
n
_
_
= g
_
_
_
_
_
_
_
P
1
.
.
.
P
n
_
_
,
_
_
V
term
1
.
.
.
V
term
n
_
_
_
_
_
_
_
(2.7)
2.3.1.4 Data Analysis and PV Curves
Voltage stability is a crucial component of system stability and will be impacted with
the addition of a variable generation resource such as wind generation. PV curves are
an indication of a systems voltage stability as active power injection increases in the
system (Kundur, 1994).
In the case of this analysis, the active power injection is given by the wind generation
produced at each bus and the voltage stability is reected in the bus voltages at the
varying voltage levels. As the control strategy shifts from xed PF to terminal voltage
control, the bus voltages will vary greatly as the wind power output changes. The use of
the TSPF analysis allows the PV curves to represent the changes in wind power output
and the resulting voltages that occur due to the variance in the power generation.
The behavior of PV curves as well as the sensitivity of voltage to reactive power and
their relation to voltage stability is a well established concept (Kundur, 1994; Taylor,
1993; Begovic and Phadke, 1992). The PV curve is inuenced by the PF of the system.
More inductive PFs limit the power transfer capability of the bus, and lower the value
at which the critical voltage is reached. The opposite is true for capacitive PFs, where
the critical voltage or the point of voltage collapse is extended and allows for increased
power transfer in the system. This extension of the critical voltage point is known
as the voltage stability margin, and is a measure that directly reects an increase in
voltage stability in the power system and indicates that the system is more secure.
Since the maximum power transfer for a particular bus is limited to the size of the
connected wind farm, the voltage value reached at maximum power will indicate an
increase or decrease in the voltage stability margin of a bus.
2.3.2 The Irish Electricity System Model
This section will describe the application of the method developed in Section II to the
2013 all-island Irish power system. The Irish system serves as an excellent test system
for power systems studies, especially those that involve wind generation. The small size
and islanded nature of the system provide heightened responses to both voltage and
frequency studies that can provide valuable insight into the future behavior of larger
power systems.
As wind generation increases, several issues regarding stability and reserve will come
to the forefront of operating the Irish system with high penetrations of wind. Large
penetrations of wind generation will displace signicant portions of dynamic reactive
power support and spinning inertia. This will require study of not only the voltage
stability impacts, but also the need for increased reserve requirements for secure system
operation (Tuohy et al., 2008); changes in unit commitment schedules to handle the
uncertainty of wind and manage its variability (Tuohy et al., 2009); improvements in
the methods of frequency control and regulation (Lalor et al., 2005; Doherty et al.,
2005); and increased accuracy in capacity value calculations (EirGrid, 2010b).
The model used in this analysis is the full island model, meaning that the system
includes both the Northern Ireland (NI) power system along with the Republic of
Ireland (ROI) power system. It is important to note that these two synchronous systems
interconnect wind generation at two dierent voltage levels; the NI system interconnects
wind generation at the 33 kV level while the ROI system interconnects at 20 kV. In the
all-island model, the transmission system is considered to be at voltages greater than
and including 110 kV, while the distribution system lies at all voltages below the 110 kV
level. Wind generation was added to the system based on the resource analysis compiled
in the All-Island Grid Study (ESB, 2008). Overall, 2188 MW of wind generation was
installed across the island. Of this 1930 MW was DFIG generation, while 258 MW
were existing FSIG machines.
2.3.2.1 Identication of Worst-Case Scenarios
In the ROI and NI system, the two loading scenarios that are traditionally incorpo-
rated into power ow studies are the Summer Night Valley (SNV) and the Winter
Peak (WP). The SNV represents the minimum loading and generation operating point.
Traditionally this is where the system is most susceptible to low voltage collapse and
occurs during the warm summer night generally between July and August. The WP
is the maximum loading and generation operating point and is when the transmission
system is most severely stressed. The WP occurs during the colder winter evenings
from December to January. Studies around these two operating points were sucient
in determining the Irish systems steady-state stability. However, as wind penetration
increases and the load changes, the worst case operating point shifts. Often, it will not
coincide with the traditional loading scenarios and an analysis of the time-series data
needs to be completed to determine the new worst-case point.
For this study, 15 minute interval wind power output and loading data was gath-
ered for several farms across the country from 2004 until 2008. The number of farms
ranged from seven in 2004, to 74 in 2008. The 15 minute wind power output data was
deemed sucient in order to assess the voltage stability of the system for a time-series
power ow simulation based on the denition of small-disturbance, long-term voltage
stability (Kundur et al., 2004). In (Kundur et al., 2004), this type of voltage stability
encompasses the small-disturbances in the system, such as changes in load or genera-
tion over the slower acting equipment in the system, such as tap-changing transformers,
thermostatically controlled loads, and generator current limiters.
It should be noted that the time-series data used in these simulations is based on
the wind power output of the farms rather than wind power output data for partic-
ular regions. The use of wind power output from actual farms accounts for several
characteristics that may be lost if only wind speed data was used. These include the
topology of the farm and the aggregation of wind turbines and allows for the linear scal-
ing procedure used to grow the systems wind generation levels in these simulations.
In a dynamic simulation the need for an appropriate aggregation technique would be
necessary, (Steurer et al., 2007).
For each year, the loading data was scaled to meet loading levels in 2013 using a
2.6% growth factor (EirGrid, 2010b). For each data point, the instantaneous wind
penetration, i.e. the amount of load served by wind generation was calculated and can
be seen along with the corresponding system load in Table 2.1.
Table 2.1: Critical Operating Points in the Irish System
Date Load Wind Penetration
(MW) (Instantaneous Penetration)
May 5, 2004 2564 73%
January 1, 2005 2374 79%
October 31, 2006 2491 86%
May 19, 2007 2596 67%
July 1, 2008 2417 76%
December 15, 2006 (WP) 5848 2.4%
As seen in Table 2.1, the worst-case points vary from year to year and arent con-
tained to the SNV or WP of the respective years. Table 2.1, demonstrates the need
to analyze a power system over multiple years in order to determine when to run the
studies. The worst-case point in each year moves from season to season between the
Figure 2.2: The probability density function for instantaneous wind penetration for
2004 - 2008.
years and has little correlation with traditional loading patterns. This study will focus
on the October 31, 2006 point, as this is the worst-case operating point in the years be-
tween 2004 to 2008. Also included in Table 2.1 is the WP loading and penetration from
2006. As seen in the table, the wind penetration seen on October 31 is dramatically
higher than the 2.4% instantaneous penetration for the WP of the same year.
The probability density functions for instantaneous wind penetration were calculated
for each year, and can be seen in Fig. 2.2. Fig. 2.2, demonstrates that the 5 years of data
have similar distributions and the point observed in this study, October 31, 2006, is the
worst-case operating point in terms of maximum wind penetration during that 5 year
period. Rather than combining the multiple years of data, into a single probabilistic
time-series, the methodology used in this chapter, deterministically simulates around a
single point.
There are many more high wind days included in the two-week simulation period in
October and as such, the simulations contain many other extreme operating points of
the system above the 79% level seen on January 1, 2005. The loading scenario around
October 31, 2006 was compared to the traditional worst-case study for a given year,
the WP. This allowed for a proper comparison of traditional steady-state techniques
and the methodology presented in this chapter.
2.3.3 Implementation of Voltage Control on the Irish Power System
DFIGs have two main control schemes; terminal voltage control and xed PF control.
In this chapter, four TSPF cases were simulated, cases A, B, C and D, two for each of
the two selected loading scenarios; October 31, 2006 and the WP from 2006, were all
scaled and applied to the 2013 Irish power system. The wind penetration levels and the
loadings scenarios associated with each of the cases A, B, C and D can be seen in Table
2.2. In Table 2.2, although the WP has a higher average level of wind generation, the
load demand in the system at this time is much greater. As a result, the wind serves a
smaller proportion of the demand in the WP scenario. This demonstrates the need to
identify the appropriate loading scenario for study when examining systems with high
penetrations of wind.
Table 2.2: 2013 Irish Power System Wind Characteristics
Case Loading Max. Wind Mean Wind Min. Wind
Scenario (MW) (MW) (MW)
A & B Oct. 31, 2006 2051 649.2 13.14
C & D WP 1998 770.1 10.39
In all four cases transmission level voltage control was implemented by DFIG farms
larger than 35 MW. Based on recommendations and practice of the Irish transmission
system operator, Eirgrid, 35 MW was deemed to be a large enough wind farm to cope
with the MVA losses across the two levels of transformers through which wind farms
were connected to the transmission system. These farms were operated from 0.95 PF
for Q
max
and 0.9 PF for Q
min
, providing a range of reactive power from 11.50 MVAr
to -16.95 MVAr for 35 MW farms. These values were scaled based on the size of the
farm. Farms below 35 MW did not have the required reactive power capabilities to
provide sucient control at the 110 kV level. In the test system, there were 973.90
MW of DFIG generation capable of providing transmission level control.
Wind farms were modeled as follows: The turbines were rst aggregated at a low
voltage, 0.4 kV, collector bus. This bus was then connected through a transformer to
a distribution voltage level bus, generally 20 kV in the ROI and 33 kV in NI. It was
then connected to the transmission system through another transformer, going from
the distribution level to the transmission level of 110 kV. A diagram representing the
general method of connection of the wind farms can be seen in Fig. 2.3. Farms larger
than 35 MW controlled the terminal bus voltage of the 110 kV bus in all the cases.
However, in the cases where distribution level control was implemented, the 20 kV bus
became the terminally controlled bus. Along with the operation of the transmission
connected wind farms, xed speed turbines were operated at a xed 0.95 inductive
PF in all four cases. As such, there is 258.44 MW of wind generation that is never
controlled.
The control varied between the rest of the DFIG wind farms smaller than 35 MW. In
cases A and C, no voltage control was implemented and the DFIG wind farms operated
using a xed 0.95 inductive PF. In cases B and D, DFIGs utilized the voltage control
feature and applied terminal voltage control at the distribution level bus of 20 kV or 33
kV. The distribution of voltage control strategies in DFIG wind turbines for the four
study cases can be seen in Table 2.3.
Table 2.3: Wind Turbine Control Distribution
Case Terminal Voltage 0.95 PF
Control (MW) Inductive (MW)
DFIGs at xed PF (A & C) 973.90 1214.55
DFIGs controlled (B & D) 1930.01 258.44
Figure 2.3: One-line diagram of the transformer connections for wind farms in the
all-island Irish power system.
2.3.3.1 Unit Commitment and Economic Dispatch
In order to achieve a load/generation balance between each time-step of the TSPF a
unit commitment and economic dispatch was necessary. Using the WILMAR planning
tool, (Meibom et al., 2008; Tuohy et al., 2009), along with the corresponding time-
series data for load and wind, a unit commitment schedule was developed. The unit
commitment schedule determined the units that would be available during a particular
day.
The generation load balance between the 15 minute time-steps required an economic
dispatch. This was accomplished using the automated economic dispatch application
built into PSS/E by Siemens PTI (Siemens, 2005). Heat rate curves for each of the
conventional units in the system were written into the economic dispatch application
and based on the unit commitment schedule provided the load/generation balance
between each time-step (Meibom et al., 2008).
2.3.3.2 Data Recording and Analysis
Four two-week periods were simulated for a total of 5376 individual power ow analy-
ses. The voltage at all of the 110 kV, 33 kV and 20 kV buses at which wind generation
was interconnected at was recorded along with the power from each wind farm for every
power ow. As such, for each two-week period, 1344 data points were recorded for the
bus voltage and wind power output. From these data records PV curves were con-
structed at the varying bus voltages that allowed for the determination of the systems
steady-state voltage stability.
2.3.4 Results and Discussion
The main application of this methodology is to obtain PV curves that provide insight
into the voltage stability margin of the Irish system.
For large farms that were connected directly to the transmission system, the goal
was to control the terminal voltage of the 110 kV bus at a specic target voltage.
Whereas wind farms that connected in the distribution system (below the 110 kV
level) controlled at either the 33 kV or the 20 kV level, the goal was to improve the
stability margin at the 110 kV level. The reasoning behind this was that higher and
more predictable voltages in the distribution system would provide greater benet for
the transmission system.
In the current operation of wind turbines in the Irish system, large levels of reactive
power are consumed by the distribution system in order to operate the wind farms
at 0.95 inductive PF. This draws large levels of reactive power resources away from
the transmission system, thus negatively impacting the transmission level voltages.
To illustrate this result, the bus observed in this section has three DFIG wind farms
connected totalling 95 MW. As such, control at each farm is applied at the 20 kV
distribution bus.
2.3.4.1 Case A and B: October 31, 2006
The plots in Fig. 2.4 and 2.5 show the same 20 kV bus under the two dierent voltage
control scenarios for the two week period around October 31, 2006. The curve in
Fig. 2.4, represents when a majority of the DFIG turbines are operated at the 0.95
inductive PF, Case A. As seen in the curve, the voltages are approaching signicantly
lower levels as the wind farms reach their maximum combined power output of 95 MW.
In comparison, the curve in Fig. 2.5 is dramatically dierent. Here, every one of the
1344 power ows has achieved the target voltage of the 20 kV bus, Case B, and as such
the low voltages seen in Fig. 2.4 are eliminated.
Figure 2.4: The 20 kV bus PV curve for Case A, Oct. 31, 2006. Since the DFIGs are
operated at a 0.95 inductive PF, as the power output from the wind turbine increases
the bus voltage begins to drop.
The eect of the increased control is demonstrated in the PV curves of the 110 kV
bus connected to the 20 kV bus through a transformer. Fig. 2.6 shows the resulting
PV curve for Case A, while Fig. 2.7 shows the same for Case B. The impact of the
increased control is evident in Fig. 2.7, as the power generated by the wind farm
increases, voltages actually increase at the 110 kV bus. The voltages in Fig. 2.6 are still
within the range of stability, but are seen to be noticeably falling as power production
increases from the wind farm leading to a decrease in system security. System security
is dened as the ability of the power system to withstand a sudden loss or unanticipated
loss of system components (Kundur et al., 2004). In Fig. 2.7, the voltages increase as
power production from the wind farm increases. This is due to the implementation of
terminal voltage control from the DFIG farm and results in a more secure system that
Figure 2.5: The 20 kV bus PV curve for Case B, Oct. 31, 2006. Now the PV curve is
drastically dierent. Terminal voltage control is enabled in all DFIGs, the bus voltage
is controlled to 1.075 PU.
is able to cope with a unexpected contingency. As such when terminal voltage control
is implemented, voltages increase as power production from the wind farm increases
resulting in a more secure system that is able to cope with an unexpected contingency.
Figure 2.6: The PV curve for the 110 kV bus connected to the 20 kV bus for Case A,
Oct. 31, 2006. Here the voltage begins to decrease as power output of the wind farm
increases.
Figure 2.7: The PV curve for the 110 kV bus, but for Case B, Oct. 31 2006. The
increased control allows for improved voltages at higher power output levels, implying
an increased voltage stability margin.
Based on (Kundur, 1994), this implies that the voltage stability margin for that
particular bus is extended and stability across the system is improved when terminal
voltage control is implemented. Not only does the implementation of terminal voltage
control improve the PV curves voltage stability margin, it also controls the range of
voltage at the 110 kV bus between a smaller bandwidth and increases system security as
seen in Fig. 2.7. This demonstrates the value of increased terminal voltage control and
allows for more predictable voltages at the transmission level leading to more robust
system operation.
2.3.4.2 Case C and D: The Winter Peak
Cases C and D simulated the application of xed PF control and terminal voltage
control around the WP operating point for the Irish power system. The nose of the PV
curve from Case C, where the power output of the farm begins to reach a maximum,
will be compared to the nose of the PV curve from Case A. Observing the nose of
the PV curve focuses on the systems voltage stability at the point of maximum power
transfer. In the case of this study, maximum power transfer is dened by the maximum
size of the wind farm.
Figure 2.8: A zoom of the nose of the PV curve from 110 kV bus for Case C, WP 2006
. Here at the same power output levels the voltages are higher even when no control is
implemented.
Figure 2.9: A zoom of the nose of the PV curve at the 110 kV bus from Case A, Oct.
31, 2006. As power output from the wind farm increases the voltage falls.
Fig. 2.8 in contrast to Fig. 2.9, demonstrates a much shallower nose, i.e. for
increased levels of power output from the wind farm, the bus is able to achieve higher
voltage levels. This is due to the lower relative penetration levels of wind generation in
the system, seen in Table 2.1. Since the loading is so much higher in the WP, there are
more conventional generation units available to provide support to the system. As a
result, the PV curve shows an increased stability margin even when no voltage control
is implemented at the DFIG farm. When control is implemented the PV curves in the
two cases are very similar, as seen in Fig. 2.7 and 2.10. This further demonstrates
the signicant impact that voltage control can have on bus voltages regardless of the
loading scenario.
Figure 2.10: The PV curve when voltage control was implemented at the wind farm,
Case D, WP 2006.
2.3.4.3 Transmission Level Voltage Control
As mentioned previously, the larger wind farms in the Irish system provided voltage
control directly at the transmission level. Here the wind farms provide reactive power
support across two transformers and control the voltage at the 110 kV bus. Fig. 2.11
provides an example of this control, and demonstrates that large farms can provide
improved voltage performance and stability directly to the transmission system.
In comparison with Fig. 2.6, 2.7, and 2.10 the control at the farm in Fig. 2.11 is
more direct and as such the voltage is more robust. This is only achieved due to the
large size of the transmission connected wind farm. The size of the farm needs to be
sucient enough to provide the necessary reactive support at the transmission level.
Figure 2.11: Here the voltage control is applied directly at the transmission level. The
target voltage is set at 1.05 PU and as a result all voltages are controlled at or above
that level regardless of loading scenario.
2.4 Conclusion
As wind generation becomes a signicant portion of generation portfolios, how it is
modeled and studied will become increasingly important. As demonstrated by this
chapter, focusing studies around traditional worst-case operating points may not cap-
ture the true worst-case scenario in systems with high penetrations of wind. Along with
utilizing historical time-series data for wind and load, the observation variables must
also be carefully selected in order to determine when the worst-case scenario would
occur. With constant variations in active power and load occurring across the system,
there is a high likelihood that modeling around the point of maximum wind penetration
as a percentage of demand captures the worst-case operating point in the steady-state.
However, in terms of the dynamic stability of the system this may not be the case, and
it may be more appropriate to complete a Monte-Carlo type analysis, that captures
more of the uncertainty associated with a variable resource such as wind generation. By
combining power ow, economic dispatch, unit commitment and historical time-series
data that captures the variability of the wind, into a single large-scale simulation, this
chapter presents a methodology that is suitable for analyzing a large power system and
assessing its voltage stability as well as system response to other conditions under large
penetrations of wind generation.
This chapter also shows that utilizing the control features of the DFIG wind turbine
improves the voltage stability margin of both the distribution level and transmission
level buses in the system. The increase in the stability margin is given as a function of
the wind power output in this chapter. In order to achieve a specic value associated to
the increase in the stability margin, a detailed analysis such as continuation power ow
needs to be completed. However, in the scope of this chapter, the TSPF provided the
relative increase in system stability with the application of terminal voltage control.
Therefore, this means that based on voltage stability criteria, larger levels of wind
generation could be connected without degrading the voltage stability of the system.
Maintaining system stability is crucial and in particular, utilizing voltage control in
DFIGs may prove benecial in system operation. This is particularly important in a
power system such as the all-island Irish system that is largely isolated and depends
heavily on the import of fossil fuels.
CHAPTER 3
Large Disturbance, Short-Term Voltage Stability
3.1 Introduction
T
HE large-distrubance, short-term voltage stability of a power system is the ability
of the power system to maintain voltage stability following a signicant event
such as a loss of generation, system fault, or some other circuit contingency. The
short-term time frame requires that voltage stability be maintained by the dynamic,
fast acting elements in the system. Unlike the analysis completed in Chapter 2, which
assessed the small-disturbance, long-term voltage stability of the system using power
ow analysis dynamic time-domain analysis is necessary to assess the large-disturbance,
short-term voltage stability (Van Cutsem and Vournas, 1998, 1996). A time-domain
analysis requires detailed models of all the elements within the power system. These
models are based on the mathematical description of the specic elements contained
in the system and will include: generators, excitation systems, governors, stabilizers,
loads, and exible AC transmission system devices. These dynamic model represen-
35
Chapter 3. Large Disturbance, Short-Term Voltage Stability 36
tations will couple with the power ow equations to create a time-domain analysis of
the power system. The use of time-domain simulations is critical in determining sys-
tem behavior and stability and is a fundamental component of power system stability
analysis (Kundur, 1994; Kundur et al., 2004).
This chapter will focus on the system response to a fault when there is a high
penetration of wind generation present in the system. It will examine both the Type-
I and Type-III WTG models and demonstrate how both can be utilized to improve
voltage stability during a severe system contingency.
3.2 Wind Turbine Modeling Considerations
The variability of wind and its impact on the long-term small-disturbance voltage
stability of the system has been discussed in the Chapter 2. The discussion here will
focus on the Type-I WTG, the FSIG and the Type-III WTG, the DFIG and their
inuence on system stability. In systems that have established wind power installations,
there may often be large penetrations of the FSIG wind farms. In the FSIG, the rotor
of the turbine blades is coupled to the rotor of the electric machine through a gearbox.
This design allows the turbine to utilize the kinetic energy stored in the turbine blade,
and contribute to system frequency stability by providing a level of spinning inertia
(Littler et al., 2005; Kennedy et al., 2011). The disadvantage of the FSIG arises in
the fact that the machine is not capable of reactive power control, but rather absorbs
large levels of reactive power during normal operation. The reactive consumption of
the FSIG is generally oset using capacitor banks to achieve a constant power factor.
However, capacitor banks are slow acting and do not prove eective during low voltage
events and as a result, FSIG farms often trip out of service. This is often referred to as
the fault-ride through capability of the FSIG (Holdsworth et al., 2003; Papathanassiou
and Papadopoulos, 2001), and is of particular importance when installing FSIGs into
a power system. In comparison, as discussed in Chapter 2, DFIG machines are able to
provide reactive power control and support the system voltages during a fault event.
The response of FSIGs and DFIGs to a fault event will be compared in the next
section. It will seek to determine whether there is an appropriate ratio of DFIGs
and FSIGs that will help maintain voltage stability during a contingency event. By
achieving an appropriate balance between FSIGs and DFIGs that goal would be to
achieve improved levels of voltage stability and support the system through the natural
inertial response present in FSIGs. The inertial response in FSIGs is due to the fact that
the blades are coupled to the rotor of the induction generator. The mechanical design of
the FSIG allows for a natural inertial response to a loss of generation event. The FSIG
acts in a similar manner to a synchronous generator and attempts to provide inertial
response by utilizing the stored kinetic energy in the blades of the turbine. The level
of inertia present in the blades combined with the inertia in the shaft will determine
how quickly the FSIG will respond to the event. To determine this level, the inertial
mass of the blades is calculated as follows. The inertia of any given body is:
J =
m
i
r
2
i
(3.1)
In (3.1), m
i
is the mass of object i and r
i
is the radius from the center of the object.
For the blades of a wind turbine this yields:
J = 3m
b
r
3
2
=
1
9
m
r
r
2
(3.2)
Here, m
b
represents the mass of each blade and m
r
is the cumulative sum of the
blade and rotor structure, equivalent to 3m
b
. This value of inertia, J is converted to
the inertial constant, H as follows:
H =
J
2
m
2S
(3.3)
This value of inertia is coupled through the shaft of the turbine and is represented at
the generator side of the turbine as a single inertia. This representation is referred to as
a lumped or single mass model. This contribution of inertia aids the frequency stability
of the system. DFIG wind generation has a decoupled system, where the inertia of the
blades is separated from the system due to power electronics that allow for the control
of reactive power and has been discussed previously. By maintaining a balance between
the level of FSIGs and DFIGs a system with improved voltage stability and frequency
response can be achieved.
3.3 The Impact of Wind Turbine Types on Short-Term
Large-Disturbance Voltage Stability
This section will focus on how a system with high penetrations of wind generation
responds in the short-term to a severe system event, in this case a fault. This analysis
will be completed using the IEEE 30 bus test system, Fig. 3.1 (Pai, 1979). To examine
this eect, the generation and load in the IEEE 30 bus test system were modied
appropriately to achieve a realistic wind penetration level. The structure of the system
remained the same, generation and load were scaled up to allow for a larger penetration
of wind generation in the system.
Using PSS/E (Siemens, 2005), the system was modeled with a wind penetration
level of 15.6% of the total system generation of 1478.29 MW, approximately 230.6 MW
of wind generation. The system was divided into six operating areas. A single farm
was placed in areas 2, 3, 4, and 5, while areas 1 and 6 operated using only normal
conventional generation. This was done to mimic a real system where wind generation
will be spatially distributed across the network. The generation at each wind farm was
kept constant through all the simulations; only the ratio of DFIGs to FSIGs within
each individual farm varied from trial to trial. The rest of the generation in the system
was provided by 5 conventional units. The wind farms and conventional units supplied
the system load of 1270 MW and 295 MVArs.
Starting with a 50:50 ratio of DFIG to FSIG wind turbines and moving to a 70:30
ratio of DFIG to FSIG, the short-term large disturbance voltage stability of the system
Figure 3.1: The IEEE 30 bus test system.
was assessed. This was achieved by applying a 0.2j line fault at the branch connecting
buses 10 and 21. The results of this fault analysis are given in Table 3.1. Beginning
from the 50:50 ratio, it can be seen that increasing the ratio of DFIGs in the system by
10% corresponds to a 50ms increase in the critical clearing time of the fault. However,
another 10% increase in the level of DFIGs, results in the same critical clearing time.
The bus voltages under faulted conditions were plotted using the clearing times from
Table 3.1 and can be seen in Fig. 3.2. The voltages were monitored at all of the FSG
interconnections. Fig. 3.2 contains the voltage plots for bus 14, one of the wind farm
interconnections. Although there may not be a signicant dierence in the critical
clearing times as the level of DFIGs changed from 60% to 70%, the minimum voltage
reached shows signicant improvement as can be seen in Table 3.2.
Table 3.1: Critical Clearing Time for 0.2j Fault
Ratio (DFIG:FSIG) 50:50 60:40 70:30
Critical Clearing Time (ms) 540 590 590
Based on the analysis of short-term large-disturbance voltage stability the 60:40 ratio
Figure 3.2: The IEEE 30 bus test system.
Table 3.2: Minimum Bus Voltage During Faulted Condition
Ratio (DFIG:FSIG) 50:50 60:40 70:30
V
min
0.431 0.451 0.498
and 70:30 ratio provide the system with the most support during faulted conditions.
Although the 70:30 ratio improves the minimum voltage dip, the clearing times for
system stability remain the same for both ratios. This indicates that DFIG generation
can provide signicant improvements in terms of maintaining the stability of the system
during a signicant fault. As most modern wind farms are now DFIGs this issue is less
of a concern, however in older system where there are still a signicant level of FSIGs,
this shows that utilizing the capabilities of DFIGs can improve system stability.
3.4 Conclusion
As demonstrated in this chapter, the older generation of wind turbines, the Type-I
FSIG can be safely and securely accommodated into the power system by utilizing the
reactive power control capabilities of the Type-III DFIG. This is particularly benecial
in systems with large eets of FSIG generation, which would also provide the benet
of an increased inertial response during a low frequency event, but require additional
reactive power support in order to ride through a low voltage fault event.
In terms of the short-term large-disturbance voltage stability of the system, DFIG
generation provides signicant improvement over the FSIG. However most new wind
generation that is installed in power systems today is either of the Type-III WTG or
the Type-IV WTG, and would have the necessary control capability to provide voltage
support to the system. The ability of these turbines to provide voltage support will be
a signicant mitigation technique not only for voltage stability but rotor angle stability.
The benets of voltage support from DFIGs in terms of rotor angle stability will be
discussed in Chapter 4 and Chapter 5
CHAPTER 4
Rotor Angle Stability With High Penetrations of Wind Generation
4.1 Introduction
R
OTOR angle stability is classied into two distinct subcategories:
Small-disturbance (or small-signal) rotor angle stability is the ability of the sys-
tem to maintain synchronism during small disturbances. These disturbances are
suciently small that linearization of the system is possible. Small-signal insta-
bility arises when the system is insuciently damped and the oscillation grows,
resulting in a severe disturbance.
Large-disturbance rotor angle stability or transient stability is the ability of the
system to maintain synchronism following a large disturbance such as a fault or
loss of generation.
42
Chapter 4. Rotor Angle Stability With High Penetrations of Wind Generation 43
Both classications of rotor angle stability occur in the short-term time-frame, usu-
ally requiring analysis in the time-domain. In Chapters 2 and 3 it was shown that
the wind generation has a signicant impact on the voltage stability of the power sys-
tem. In traditional power systems, with low penetrations of wind, the rotor angles
of the synchronous generators are impacted by changes in active power ows in the
system. When there is a change in active power, the synchronous generators in the
system will respond with an electromagnetic torque that will dampen and minimize
rotor angle deviations. The presence of this electromagnetic torque, along with the
fact that bus voltages are tightly controlled by synchronous generation through the
use of automatic voltage regulation (AVR) essentially decouples the behavior of rotor
angle from the voltage stability of the system. However, as wind generation increases
and synchronous generation is displaced, the coupling between rotor angle stability and
the voltage stability of the system will strengthen. There will be less electromagnetic
torque present in the system to dampen angular deviations due to the asynchronous
nature of wind generation. There will also be a degradation in the voltage stability
of the system due to lack of reactive power control as AVRs are displaced in order to
accommodate the new wind generation. This will result in a decreased ability to control
voltage, which will directly inuence the rotor angle stability of the remaining conven-
tional synchronous units in the system. This chapter will demonstrate that there is a
fundamental dierence between the active power injections of conventional synchronous
generation and asynchronous wind generation. This dierence will place an increased
burden on the remaining conventional synchronous units in the system that will lead
to a decrease in the rotor angle stability of the system if there is a disturbance on the
system. By utilizing the reactive power control capabilities of wind generation, the
rotor angle stability of the system can be improved by supporting the bus voltages in
the system.
4.2 Methodology
Variable speed wind turbines (VSWT) provide electrical synchronism with the power
system through power electronic convertors (Kayikci and Milanovic, 2007; Petersson
et al., 2005; Xu and Cartwright, 2006); however, this power electronic coupling inhibits
mechanical synchronism with the system eectively rendering wind inertia-less. How
wind generation displaces conventional synchronous generation will signicantly impact
various stability aspects of the power system. This chapter will assess how reactive
power production from wind generation will directly inuence the rotor angle stability
of the system. However, it is rst necessary to understand how wind generation interacts
with the active power ows in the system as this is how rotor angle stability is impacted
in traditional power systems with low penetrations of wind.
Work has been completed that shows how wind generation will inuence the inertial
behavior of a power system. In Gautam et al. (2009), the impact of VSWTs on the
small-signal stability of a large power system was assessed. The work in Gautam et al.
(2009) shows the sensitivity change of the inertia with respect to wind generation in
the system. By replacing VSWT generation with equivalently rated synchronous units,
the small-signal stability and transient stability of the system was assessed. It was
determined that the active power delivered from VSWT generators is dierent from an
inertial aspect to that delivered by synchronous generation. Wind generation controls
can be altered to emulate an inertial response for frequency stability, but have not
been implemented widely in power systems (Morren et al., 2006; Lalor et al., 2005).
The work completed here looks to expand on the fundamental dierence between the
active power produced by VSWTs and that produced by conventional synchronous
generators, particularly how they interact with the rotor angle stability of the system.
Due to the fact that wind generation is inertia-less, the synchronous units that co-
exist in the system with wind will be forced to provide the necessary resources, i.e.
inertia and damping torque, required to mitigate any instability events. Carrying this
extra burden, will stress the synchronous units and could lead to less secure system
operation. By utilizing the built-in capabilities of wind generation, specically reactive
power control, the requirements placed on conventional synchronous generation could
be eased and system security could be improved.
This chapter will examine how reactive power from wind generation can be used
as a mitigation tool to ease the stress on synchronous generation and increase system
security and improve the rotor angle stability of the system. The aim is to show that
when active power ows change minimally, the manner in which the wind generation
provides reactive power support to the system is critical in maintaining rotor angle sta-
bility of conventional units in the system. The improvement in stability is achieved
by supporting bus voltages using reactive power injections from wind generation, in
particular utilizing the terminal voltage control capabilities of VSWT wind turbines.
This reduces the reactive power requirement from conventional synchronous genera-
tion and minimizes deviation in the eld voltage. This allows synchronous generators
to maintain their reactive power output inside their limits. By preventing reactive
power from the synchronous generation from collapsing, the balance between electrical
power output and mechanical input is maintained. This balance minimizes rotor angle
deviation and improves rotor angle stability. To ease the reactive power burden on syn-
chronous generation, the control strategy employed by the wind turbines is varied and
the impacts on the angular stability of the conventional generation in the system are
observed. The next section will develop a methodology that will clarify the distinction
between active power from wind generation and that from synchronous generation. It
will also, detail how the rotor angle will be inuenced and how signicant events can
be mitigated using the built-in capabilities of the wind turbine.
4.2.1 Active Power Analysis
Here wind generation is rst compared directly to synchronous generation in order to
achieve a baseline comparison for the rest of the analyses. This is achieved by creating a
base case consisting of DFIG wind farms operating at a xed 0.95 capacitive power fac-
tor spread across the system. Next, a second case is created where the wind generation
is replaced by equivalently sized and rated synchronous machines with exciter systems;
however no governors or stabilizers are modeled. The synchronous wind machines are
modeled in this manner in order to see how they respond in comparison to an asyn-
chronous wind generator, which cannot increase its active power output by providing a
governor response. The exciter is included to provide control for the eld current and
increase stability. The reactive power output of the synchronous units is xed at the
same 0.95 capacitive power factor as the wind generation. A brief description of the two
cases in the active power analysis can be seen in Table 4.1. A transient analysis is then
completed for a loss of generation event and the rotor angle, active and reactive power
outputs are monitored for each of the synchronous units in order to assess the impact
of wind generation on the system. The physical dierences between the synchronous
generators and wind generators, i.e. inertial contribution of the rotating mass, dictates
that there will be signicant variations in the active power ows across the system,
particularly, the ability to provide electromagnetic torque which is resolved into two
components;
The synchronizing torque component, is in phase with the rotor angle deviation.
The lack of synchronizing torque leads to aperiodic or non-oscillatory stability
(Kundur et al., 2004).
The damping torque component, is in phase with the speed deviation. The lack
of damping torque leads to oscillatory instability (Kundur et al., 2004).
Wind generators have very limited mechanical interaction with the rest of the power
system due to the power electronic decoupling of the blades and rotor, and as a result
do not have the capability to provide the system synchronizing torque or damping
torque. In order to characterize the dierences between synchronous generation the
active power analysis will examine what aspects of the system are inuenced by the
change of generator type for the two cases.
Table 4.1: Active Power Comparison Scenarios
Case Wind Generation Control Type
Capacitive Case Fixed 0.95 capacitive power factor
Synchronous Wind Case Fixed 0.95 capacitive power factor
4.2.2 Reactive Power Analysis
The active power analysis quanties the impact of the active power delivered by wind
generation and determines whether it is fundamentally dierent in comparison to the
active power delivered by synchronous generation (Gautam et al., 2009). Reactive
power however, is a purely electrical injection into the system, i.e. there is no mechanical
input required to create or deliver reactive power. As such, the reactive power delivered
by a synchronous unit can be compared directly to that delivered by a wind generator.
This analysis builds upon this concept by analyzing the impact that varying the reactive
power control strategy of the wind farms has on the system. By only changing the
reactive power output from the wind farms, the active power ows across the system
will remain xed. The resulting change in rotor angle deviation between the cases can
then be attributed to the changes in the systems reactive power ows.
In this section two cases examine the behavior of the wind generation and how they
control their reactive power output. In the rst wind case, wind generation operates at a
unity power factor, i.e. no MVArs are injected into the system. In the second wind case,
wind generation is operated using terminal voltage control, where the reactive power
is quickly controlled in order to achieve a specic voltage at a target bus (Akhmatov
and Eriksen, 2007; Tapia et al., 2007; Cartwright et al., 2004b). The studied cases are
listed in Table 4.2
Table 4.2: Reactive Power Comparison Scenarios
Case Wind Generation Control Type
Unity Case Fixed unity power factor
Terminal Voltage Case Reactive power controlled to target voltage
Similar to the active power analysis, a transient analysis is completed for a loss of
generation event and the active and reactive power ows are monitored along with the
rotor angle stability for the most impacted machine. Any changes in system stability
can be attributed to changes in reactive power ows and the stability of the system
under the varied reactive power control operating conditions is assessed.
A fault analysis is also completed for the capacitive case and the terminal voltage
case in order to compare the generator response to a severe low voltage event. A bus
fault is applied and cleared at a load bus in the system, and the rotor angle, bus voltage
and reactive power output for the generators are monitored. This allows for a further
insight into how reactive power interacts with the rotor angle stability of synchronous
machines.
4.3 Test System
The New England 39 bus system was used as the test system in this analysis, Fig. 4.1
(Pai, 1979). The ten synchronous units in the system were modeled as salient pole gen-
erators (GENSAL), with AC excitation systems (IEEEX1), steam turbine governors
(TGOV1), and stabilizers (STAB1) (Kundur, 1994). The load was modeled to include
33% constant current, 33% constant impedance, and 33% constant power (IEEE, 1995).
DFIG wind generation was added to the buses listed in Table 4.3, to achieve a instan-
taneous penetration level of 21.6% (1250 MW) for a demand level of 5785 MW. The
DFIG model used was a generic model of 1.5 MW GE DFIG machine and operated at a
xed 0.95 capacitive power factor for the base line analysis for the active power analysis.
The control parameters of the wind turbine are the standard parameters as described
by (Clark et al., 2010). The wind turbines were operated at 100% of their capacity
and each farm was 70 turbines aggregated in a 104 MW farm. Operating at 100%
of rated power has no impact on the study since the transient analysis is completed
as a snapshot of the systems most stressed operating point, in this case the point of
maximum wind penetration. Active power that was displaced by the wind generation,
was balanced by reducing the active power output of the ten original synchronous units
uniformly, as a result none of the units are fully displaced. The reactive power limits
Figure 4.1: One-line diagram of the New England 39 bus test system.
of the ten original synchronous units were left unchanged in order for the system to
reach a solution.
Table 4.3: Wind Generation Locations
Connection bus 2, 6, 8, 12, 16, 18, 24, 26, 32, 33, 35, and 37
The type of control strategy employed by the DFIG farms was varied based on Tables
4.1 and 4.2 and a transient analysis of the system was completed. The contingency
examined in all of the cases was the loss of the synchronous generator located at bus
33 operating at 632 MW. The synchronous unit that responded with the largest initial
angular deviation was the generator located at bus 34, and thus was monitored for all
of the comparison cases. The reference angle used for the studies was the angle of the
machine located at bus 31. The active and reactive power ows at the ten original
synchronous units from the test system were also monitored, the results of which are
included in the following section.
4.4 Results and Discussion
The results and discussion is divided into two sections. The rst assesses the dierence
between active power from wind generation and synchronous generation and how it in-
teracts with the rotor angle stability of the system based on the methodology described
in Section 4.2.1. The second section will examine the mitigation strategies available to
wind generation, i.e. reactive power control, and how they can contribute to improving
the system stability, in particular the rotor angle stability based on the methodology
described in Section 4.2.2. It will examine how reactive power support from wind gen-
eration interacts with synchronous generation. All simulations were completed in the
DSATools software package (PowerTech, 2011).
4.4.1 Active Power Results and Discussion
In Fig. 4.2 the rotor angle for generator 34 can be seen for each of the cases from Table
4.1. In comparing the rotor angles, it can be seen that there is a variation between
the two cases, but they are not dramatically dierent. The dierence in the two cases
is due to the fact that when wind generation is interconnected it is not providing any
electromagnetic torque to the system. When synchronous units replace the wind farms,
there is an inertial response and as such, there is a change in active power ows for the
nine original synchronous generators that are still online for the two cases following the
contingency event, Fig. 4.3. The change in active power ow is signicant, but as seen
in Fig. 4.2, the resulting impact on rotor angle is relatively benign.
A Prony analysis was completed on the rotor angle signal for generator 34 for each
of the two cases in order to analyze the rotor angle dierences in greater detail. By
comparing the dominant mode for each of the two cases, the magnitude and relative
damping of the oscillatory signal can be determined. These two characteristics of
Figure 4.2: Rotor angle traces from the synchronous wind case and the capacitive case
at generator 34 for the loss of generation contingency.
Figure 4.3: Active power response from the nine original synchronous units that are
still online for the synchronous wind case (solid line) and the capacitive case (dashed
line).
the mode are critical in determining how large the initial swing is and how quickly
the oscillations are damped out. By examining the properties of the mode, further
conclusions regarded the relative stability of each case can be determined, Table 4.4.
From the table, it can be seen that the modes display similar magnitudes and damping
levels in each of the cases, with the synchronous wind case providing slightly improved
system damping.
Table 4.4: Dominant Mode Comparison
Case Mag. Phase Freq (Hz) Damping
Synchronous Wind Case 10.95 -1.17
0.844 8.14 %
Capacitive Case 10.41 -1.589
0.860 7.95 %
This is also reected in the calculation of the exponential rate of decay given in
(5.1).
y = e
x
+ (4.1)
The coecients for each case can be seen in Table 4.5. Once again, the magnitude
and rate of decay coecients are similar, reecting the behavior seen in the rotor angle
traces, Fig. 4.2.
Table 4.5: Rate of Decay Coecients
Case
Synchronous Wind Case 42.4 -0.49 23.3
Capacitive Case 39.1 -0.46 23.4
The analysis completed in this section demonstrates that the capacitive generation
and synchronous wind generation are comparable from a rotor angle perspective, with
the synchronous wind case providing slightly improved damping levels. The more sig-
nicant system impact of synchronous generation is demonstrated in the change of the
active power ows, which is a result of the physical dierence between synchronous
generation and asynchronous wind generation. This demonstrates, that in order to
achieve damping levels comparable to the synchronous wind case, the 10 original syn-
chronous units in the capacitive case are required to provide larger active power outputs
and increased damping support. In a system with high levels of wind generation, this
places an increased burden on conventional synchronous generation and could lead to
less secure system operation. With the continued installation of wind generation in
power systems, it will become important to understand and improve system security
utilizing the built-in capabilities of wind generation. As such, the next section focuses
on using reactive power control as a mitigation technique that can help improve system
stability.
4.4.2 Reactive Power
4.4.2.1 Loss of Generation Contingency
The previous section demonstrated that with increasing penetrations of wind gener-
ation, that increased responsibility will be placed on existing synchronous generation
in a system. With wind generation composing a signicant percentage of a systems
generation portfolio, it will become critical to utilize the available mitigation techniques
available from wind generation in order to improve system stability.
Modern DFIG wind turbines have the capability to provide the system with large
levels of reactive power regardless of the level at which they are producing active power.
In the previous section the wind turbines were operated at a 0.95 capacitive power
factor. This control strategy will now be compared to the two others described in
Table 4.2, rst the unity case and then the terminal voltage case. In Fig. 4.4, the
rotor angle trace for the capacitive case is plotted along with the unity case. It is clear
from Fig. 4.4 that there is a signicant change in the behavior of the synchronous
machine at bus 34. Unlike the synchronous generation case, the active power ow from
the original synchronous units do not change greatly. This is conrmed by the active
power outputs for the nine original synchronous units that are still online for the two
cases seen in Fig. 4.5. For the loss of generation contingency, the conventional units
respond with nearly the same active power outputs, the average deviation in relation to
Figure 4.4: Rotor angle traces for generator 34 from the capacitive case and the unity
case for the loss of generation contingency.
the generation output for the capacitive case across the time steps is 0.33%. Since there
is a very small change in the active power ow, the change in rotor angle seen in Fig.
4.4 must be due to the change in reactive power production from the wind generation.
Once again, a Prony analysis was completed in order to identify the dominant mode
for the unity case and compared to the dominant mode for the capacitive case. The
details of the mode can be seen in Table 4.6.
Table 4.6: Dominant Mode for the Unity Case
Unity Case 10.11 -10.99
0.855 7.40%
Table 4.7: Rate of Decay Coecients for the Unity Case
Case
Unity Case 31.4 -0.48 24.8
The mode for the unity case is similar to the capacitive case (Table 4.4). This is
further conrmed by tting an exponential decay for the unity case, Table 4.7. Here,
the rate of decay coecients are very similar to the capacitive case (Table 4.5). These
Figure 4.5: Active power response from the nine original synchronous generators online
from the capacitive case (solid line) and the unity case (dashed line).
results indicate that providing bulk MVArs to the system is nearly as eective as
providing no reactive power support to the system whatsoever. There is little to no
improvement between the two cases, and as such a third case is studied, terminal
voltage control by the wind farm. In this case, MVArs are dynamically controlled by
the wind farm in order to achieve a target voltage, 1.0 pu in this case, at a designated
bus within the time-frame of contingency events. This provides more direct support
to the synchronous units in the system, allowing for more robust system operation.
The terminal voltage case was compared to the capacitive case for the same loss of
generation contingency and the resulting rotor angle traces for generator 34 can be
seen in Fig. 4.6.
As done in the previous cases, the active power output of the conventional units
is monitored to ensure there are minimal changes in active power ows, Fig. 4.7. As
with the unity case, the average deviation in relation to the generation output of the
capacitive case was calculated and determined to be 0.23%. The change in active power
for this case is even less than for the unity case. The mode studied for the terminal
voltage case from the Prony analysis can be seen in Table 4.8. In comparison to the
Figure 4.6: Rotor angle traces for generator 34 from the capacitive case and the terminal
voltage case for the loss of generation contingency.
Figure 4.7: Active power response from the nine original synchronous generators online
from the capacitive case (solid line) and the terminal voltage case (dashed line).
other two reactive power control strategies for the wind generation (Tables 4.4 and 4.6),
there is a signicant improvement in the damping of the mode, given the similar initial
swing magnitudes. This indicates that when reactive power is controlled to achieve a
specic target voltage, the oscillation is more eectively damped. When examining the
rate of decay for the terminal voltage case (Table 4.9) an interesting change is observed
in comparison to the other two cases (Tables 4.4 and 4.7).
Table 4.8: Dominant Mode for the Terminal Voltage Case
Terminal Voltage Case 10.2 49.3
0.872 10.8%
Table 4.9: Rate of Decay Coecients for the Terminal Voltage Case
Case
Terminal Voltage Case 22.0 -0.39 24.7
The rate of decay for the terminal voltage case is signicantly less in comparison to
the three cases that have already been examined, -0.39 sec
1
, however, the magnitude
of that decay, is signicantly less as well. This demonstrates a distinct dierence in
system behavior when MVArs are actively controlled; the initial swing in rotor angle
immediately following the contingency event is considerably reduced, by 3.85
, Fig.
4.6. This results in the lower rate of decay, as the system is nearer to the last stable
operating point.
Since the only variable that sees any dramatic change in the three wind cases is the
manner in which reactive power is produced by the wind generation the increase in
damping can be attributed to providing dynamic reactive power support. By quickly
controlling voltage to a specic set-point value, the system is better able to damp
oscillatory behavior that may lead to rotor angle instability. This demonstrates the
critical role that reactive power support can play in improving rotor angle stability.
Figure 4.8: The rotor angle of generator 34 for the three terminal voltage control cases.
4.4.2.2 The WECC Full Voltage Controller and DFIG
The previous analyses relied on utilizing the GE 1.5 DFIG wind turbine and controller
model. This section will explore whether the improved results with the implementation
of terminal voltage control are a result of the model specics or a fundamental system
response. As such, the GE 1.5 DFIG model was compared to the WECC voltage
source full-convertor turbine model (WECC VSC) and the WECC DFIG model and
convertor using the standard control parameters dened by WECC (2006a,b). The
same contingency as before, the loss of generator 33, was applied and the rotor angle
of generator 34 was observed. The resulting rotor angle traces were compared for each
of the three terminal voltage cases and can be seen in Fig. 4.8.
As seen in Fig. 4.8 the rotor angles behave in a relatively similar manner for all
three cases; there is very little dierence in the initial swing immediately following the
contingency, 0.26
and 0.11
for the WECC VSC and WECC DFIG cases respectively.

The main variation in the rotor angle is the damping level seen by the two WECC wind
models. This is further reected by examining the Prony mode in greater detail, Table
4.10. In each of the two additional cases, the magnitude of the mode is reduced in
comparison to the terminal voltage case (Table 4.8), and correspondingly the damping
level is reduced. As seen earlier with the rates of decay, this indicates that there is less
of a burden on the system to respond and return the system to a stable operating point.
This is further conrmed by examining the rate of decay for the two new cases, Table
4.11. As with the behavior of the terminal voltage case (Table 4.9), the magnitudes of
the decay are similar, and the rates of decay are slightly improved. This reects what
is seen in the rotor angle trace as well and demonstrates that the implementation of
terminal voltage control as a transient stability mitigation strategy is eective and not
isolated to a specic model type.
Table 4.10: Dominant Mode for the WECC Wind Model Cases
WECC VSC Case 6.86 -7.80
0.852 6.93%
WECC DFIG Case 8.04 -11.4
0.850 6.96%
Table 4.11: Rate of Decay Coecients for the WECC Wind Model Cases
Case
WECC VSC Case 26.1 -0.44 24.1
WECC DFIG Case 22.6 -0.40 24.3
4.4.2.3 Reactive Power and Field Voltage
As seen in Fig. 4.5 and 4.7, the active power ows of the conventional generation units
are not impacted by the type of reactive power control employed by wind generation,
however, there is a signicant impact on the reactive power ows of the conventional
generation units. In Fig. 4.9, the reactive power ows for generator 34 following the
loss of generation contingency are seen. For each wind case, generator 34 has a dierent
initial reactive power set point and responds to the event with a varying reactive power
response, Fig. 4.9. This is due to the change in reactive power output from the wind
farms.
In Fig. 4.10, the loss of generation event degrades the bus voltage the greatest for
the unity case, where wind generation is providing no reactive power support to the
Figure 4.9: The reactive power output of generator 34 for the three wind cases.
Figure 4.10: The voltage of bus 34 following the contingency for the three wind cases.
system. In the capacitive case, the support in the form of bulk MVAr production from
wind generation improves bus voltage compared to the unity case. In the terminal
voltage case, where MVArs are controlled to a target value, the bus voltage performs
the best following the contingency and the voltage recovers quicker to a steady-state
value. Fig. 4.10 indicates that the bus voltage suers due to the excessive reactive
power production requirements placed on the conventional generation by the reactive
power control strategy employed by wind generation. The variation in bus voltage
is directly reected in the eld voltage of the generator at bus 34, Fig. 4.11. The
variation in the machines eld voltage is due to the change in reactive power ows,
since there is no change in active power ows across the system. The stress placed
on conventional generation by the wind generations reactive power control strategy
degrades the eld voltage of the machine, directly impacting angular separation within
the machine. In further examining Fig. 4.11, the unity case suers from the largest
eld voltage deviation and as a result has the worst rate of decay in its rotor angle.
The terminal voltage case has the smallest deviation and the best rate of decay. The
capacitive case lies between the two other control cases. This demonstrates how built-
in mitigation techniques of wind generation can be utilized in order to improve rotor
angle stability in a system with a high penetration of wind.
This is the manner in which reactive power and rotor angle can interact and impact
the stability of the system. This relationship between reactive power and rotor angle
has been discussed in (Taylor, 1993), where a clear link between voltage instability
and rotor angle instability is presented. In modern power systems, this issue has been
mitigated using automatic voltage regulation (AVR) and as a result, has not been
a signicant issue. With higher penetrations of wind generation there will be more
remotely located generation units and a weakening of the systems AVR capability. As
such, the control of reactive power by wind generation in the system will become more
critical for system operation under high penetrations of wind generation.
Further insight into the impact on eld voltage can be seen by examining the peak
to peak deviation in the eld current, Fig. 4.12, of generator 34 for each of three wind
cases. The eld current set-points before the contingency reect the reactive power
Figure 4.11: The eld voltage of generator 34 following the contingency for the three
wind cases.
set-points of generator 34. The peak to peak deviation was 0.26 pu, 0.25 pu, and 0.26
pu for the capacitive, unity and terminal voltage cases respectively. Since they were all
very similiar, this indicates that generator 34 changed its reactive power at the same
level for all three cases and the large change in eld voltage, Fig. 4.11, is not a result
of over-excitation, but due to the deterioration in bus voltage as a result of inadequate
reactive power support.
4.4.3 Fault Analysis
The previous sections demonstrated that for a large system disturbance, i.e. the loss of a
generator, the terminal voltage cases decrease the burden on the remaining conventional
synchronous units and helps improve the rotor angle stability of the system. In this
section, a localized voltage event will be analyzed in order to further reinforce the
importance of appropriate reactive power control. For a loss of generation contingency,
the impacts are felt across the system, and all of the synchronous units in the system
respond to the event. A fault contingency is a much more localized event, and as such
Figure 4.12: The eld current of generator 34 for the three wind cases
the system reacts in a dierent manner. Here, a three phase-to-ground fault is applied
at bus 20, a load bus in the system, for seven cycles. Following the fault clearance, the
impact on rotor angle stability is observed for two wind cases, the capacitive case and
the terminal voltage case (GE DFIGs only).
In Fig. 4.13, the rotor angles for the capacitive case and terminal voltage case are
seen. Following the clearance of the fault, the rst machine to lose synchronism in the
capacitive case is located at generator 33. The same machine in the terminal voltage
case is able to return to a stable operating point. The synchronous machine in both
cases is loaded at the same level and capable of providing the necessary dynamic support
services, but in the capacitive case the machine loses synchronism. This is a result of
reactive power collapse at the synchronous machine, Fig. 4.14. In the capacitive case,
the wind farms are providing bulk levels of reactive power to the system. Initially this
eases the reactive power burden on generator 33, i.e. the reactive power set-point is
lower for the capacitive case than in the terminal voltage case, however, the machine
is unable to overcome the fault. This is seen in the reactive power production of the
wind farm located at bus 33, Fig 4.15. In the terminal voltage case, the reactive power
of the farm is initially absorbing a large amount of reactive power, and at the time of
Figure 4.13: The rotor angle of generator 33 following a 7 cycle fault at bus 33 for the
capacitive case and terminal voltage case.
the fault suddenly increases its reactive power output in order to help maintain bus
voltage and aid the stability of generator 33. In the capacitive case this is not possible
and as a result the machine continues to absorb MVArs beyond its limit and eventually
loses synchronism.
4.5 Conclusion
The asynchronous nature of wind generation places an increased responsibility on con-
ventional synchronous generation to provide the necessary resources to mitigate a con-
tingency event. By utilizing the built-in capabilities of wind generation to provide the
synchronous generation with reactive power support, the onus on synchronous genera-
tion can be eased. It is shown that the rotor angle of synchronous generators is directly
inuenced by the type of reactive power control employed by the wind generation. The
implementation of appropriate control strategies in wind farms, particularly the im-
plementation of terminal voltage control, can lessen the reactive power requirements
of conventional synchronous units and help mitigate large rotor angle swings and aid
Figure 4.14: The reactive power collapse of generator 33 following the clearance of the
fault at bus 20. The limit on generator 33 is -300 MVAr.
Figure 4.15: The reactive power output of the wind farm at bus 33.
conventional generation in damping the oscillatory signal following a loss of generation
event. Furthermore, it is shown that reactive power support from wind generation
can aid in mitigating severe low voltage events, thus minimizing angular separation in
synchronous units.
The displacement of the conventional synchronous units will have consequences be-
yond loss of inertia and synchronizing and damping torque. The loss of mitigation
capabilities such as AVRs, dynamic VAR support, and governor action will also have
signicant impacts on system stability.
CHAPTER 5
A Robust Reactive Power Control Strategy for Wind Generation
5.1 Introduction
C
HAPTER 4 has established the relationship between voltage stability and rotor
angle stability and has shown the impact reactive power control from wind gen-
eration can have on system security. This chapter will further expand upon the impact
that reactive power can have on the rotor angle stability of the system. Chapter 4
has noted that there is an interaction between voltage stability and rotor angle, but
it is somewhat decoupled thanks to the presence of AVR. As wind generation pene-
tration increases in the system, conventional synchronous generation will be displaced,
along with their AVRs. This will limit the capability of the system to regulate reac-
tive power. However Chapter 4 has shown that this can be mitigated if wind farms
implement terminal voltage control.
Implementing a coordinated control strategy such as terminal voltage control in
67
Chapter 5. A Robust Reactive Power Control Strategy for Wind Generation 68
wind turbines can be dicult in power systems with high penetrations of wind. Of-
ten, wind generation can be embedded within distribution networks and coordinated
voltage control strategies can be ill-dened. Work in Vovos et al. (2007); Madureira
and Pecas Lopes (2009) has shown how these limitations can be overcome and voltage
control can be coordinated in distribution networks to improve voltage stability. Keane
et al. (2011) has shown that implementing voltage control in the distribution network
can have signicant benets on the voltage stability of the transmission network. The
presence of distributed wind generation resources may not be the only limiting factor
in developing a robust reactive power control strategy for wind generation in power
systems; limited availability of reactive power resources at the wind farm or the lack
of supervisory control and data acquisition (SCADA) systems may inhibit the ability
of the wind farm from providing adequate voltage support to the transmission system.
From this the following question arises: Are all wind farms in the system required to
implemented terminal voltage control in order to maintain the rotor angle stability of
the system?
To answer this question, this chapter develops a robust reactive power control strat-
egy that will support the rotor angle stability of the synchronous generation in the
system. The methodology presented here will assess the electro-mechanical character-
istics of a system utilizing small-signal analysis. By analyzing the characteristics of
the system from an electro-mechanical perspective, critical wind farm locations that
have signicant interactions with synchronous generation in the network can be identi-
ed. By having these designated wind farms implement terminal voltage control, while
the rest of the farms operate using a much simpler control scheme, i.e. xed power
factor control, the change in rotor angle stability will be assessed using a transient
analysis. To ensure that the decision to implement reactive power control based on
electro-mechanical indicators does not degrade the voltage stability of the system, crit-
ical clearing times (CCTs) for the application of a fault will also be determined. This
will ensure that applying terminal voltage control at a limited set of farms does not
decrease the overall security of the system.
5.2 Methodology
This section will develop a methodology that will identify wind farm locations within a
system that are critical from an electro-mechanical viewpoint. By utilizing this method-
ology the rotor angle stability of the synchronous generators is improved without sac-
ricing the voltage stability of the system.
5.2.1 Small-Signal Analysis
A small-signal analysis examines the response of a linear model of the system to small
perturbations of the operating point (Kundur, 1994; Gao et al., 1992). The small-signal
analysis will identify modes that arise as a result of an electro-mechanical interaction
between synchronous generators in the system. As described in Kundur (1994); Gao
et al. (1992), the dynamics of the system can be expressed as (5.1), where x represents
the state variables of the system, and u and y represent the inputs and outputs respec-
tively. The expression in (5.1) can be linearized around a single operating point and
represented as Jacobian matrices of the states A, B, C and D along with the change
in states, x, and the inputs and outputs, u and y, respectively. The solution of
(5.2) yields to the eigenvalues, , of the linearized system (5.3).
x = f(x, u)
y = g(x, u)
(5.1)
x = Ax + Bu
y = Cx + Du
(5.2)
det(AI) = 0
for
1
,
2
,. . . ,
n
eigenvalues
(5.3)
From the eigenvalues the right eigenvector,
i
, and left eigenvector,
i
, of each oscilla-
tory mode are identied in (5.4) and (5.5).
A
i
=
i
i
for i =1,2,. . . ,n (5.4)
i
A=
i
i
for i =1,2,. . . ,n (5.5)
The right eigenvector will determine the relative activity of the element in the i th
mode, while the left eigenvector weighs the contribution of activity to the i th mode.
Oscillatory modes that occur at low frequencies, between 0.1 and 2.0 Hz, and have low
levels of damping below 10%, can often lead to instability in machines or lines in the
system Kundur (1994). As such, these are the modes that will be of interest in this
method.
Utilizing the right and left eigenvectors, the participation factor of each element,
dened as k, can be identied. The participation factor, P
ki
within (5.6), is the measure
that relates the states and modes of the eigenvectors that measures the net participation
of each element, a
kk
, in the system.
i
a
kk
=
ik
ki
= P
ki
(5.6)
In (5.6), a
kk
can represent any state of any element in the system, e.g., genera-
tor speed, controller gains, DFIG states, etc. Further examining a particular electro-
mechanical mode associated with a particular generator, the other generators that have
Create base case.
Model wind farms as
synchronous wind farms.
Displace single existing
synchronous unit.
Redispatch remaining
generators.
Run small signal
analysis
Identify low frequency,
undamped modes.
Examine participation
factors of generators
in identied mode.
Record synchronous
wind farms with high
participation factors.
Put displaced unit
into service.
Is there another
unit to displace?
Yes
No
Dened critical set of
wind farms
completed.
Select study scenarios.
Determine what the system conditions should be for the studied scenarios
Figure 5.1: The owchart describing the process used to identify the critical set of wind
farms.
an oscillatory relationship, i.e. the oscillatory behavior of one generator is inuencing
the rotation of another, can be identied. These generators are identied using partici-
pation factors. By identifying generators with high participation factors, the machines
that play a signicant role in maintaining the rotor angle stability of the system can
be identied.
In a system with high penetrations of wind this becomes more complicated since wind
farms are asynchronous and do not have any electro-mechanical interaction with the
system. The power electronic convertors in wind turbines decouple the mechanical side
of the system from the electrical side. In order to identify which farms will play a critical
role in maintaining the rotor angle stability of conventional generation a systematic
approach is necessary. The process for identication is given in the owchart in Fig.
5.1 and described in the steps below.
1. Select the scenarios that the study will focus on. This determines what the
system conditions will be during the small-signal analysis. The choice of the
study scenarios is up to the user. The methodology to determine which farms
will be critical is independent of the study scenarios and can be applied to any
set of system conditions.
2. Model all existing and potential wind farms as full synchronous units with all of
the necessary control systems, i.e. governors, excitation control and stabilizers.
These generators will be known as the synchronous wind farms and the case will
be dened as the Synchronous Wind Case.
3. Displace a single conventional generator that previously exists in the system and
solve the power ow. This will mimic a unit being taken o-line in order to
accommodate the larger penetration of wind generation. Each unit in the test
system is displaced in turn to fully assess the response of the system and mini-
mize the rotor angle deviations and maintain synchronism if the generator is not
committed to the dispatch.
4. Run the small-signal analysis and identify the machines that present low-frequency
undamped rotor angle modes. These are the oscillatory modes that will have
large impacts in the transient simulation on the rotor angle stability of con-
ventional generation. These modes will be exclusive to the existing conventional
synchronous generators in the system, any modes that arise from the synchronous
wind farms are ignored since they are replaced by wind farms in the following
steps.
5. Each mode identied from the previous step needs to be examined in greater
detail. These modes are exclusive to the existing conventional synchronous units
and within each mode the participation factors will be observed. Now, the syn-
chronous wind machines with the highest participation factors will be identied.
6. This process is repeated for the displacement of each of the existing synchronous
units in the test system, until all of the synchronous wind farms with consis-
tently high average participation factors are identied. These farms are critical
to maintaining rotor angle stability and the case will be dened as the Critical
Wind Case.
The methodology presented in this chapter is applied to a series of N-G scenarios,
where wind generation progressively displaces conventional synchronous generation. It
can generically be applied to any combination of N-1 or N-2 scenario. The fundamental
driver of the analysis is the network conguration and can thus be used to identify
network vulnerabilities regardless of the initial conditions.
5.2.2 Transient Analysis
A transient analysis is completed based on the results of the small-signal analysis. This
is done in order to conrm the small-signal results hold true in a time-domain analysis.
5.2.2.1 Full Control Analysis
To create a base line comparison for the synchronous case, a full control case is created.
Here, all of the wind farms in the synchronous wind case are replaced with doubly-fed
induction generator (DFIG) wind farms operating with terminal voltage control. A
loss of generation contingency is completed and the stability is assessed using Prony
analysis. The comparison between the synchronous wind case and full control case
demonstrates the impact the type of machine from which active power is provided has
on the rotor angle stability of the system.
5.2.2.2 Critical Farm Analysis
The previous section identied the critical set of synchronous wind farms that partic-
ipate highly with the existing generators when a unit is displaced due to high pene-
trations of wind generation. These interactions are purely electro-mechanical, however
based on Chapter 4, if the synchronous wind farm is replaced with an asynchronous
wind farm the manner in which reactive power is controlled can have serious impacts
on the rotor angle stability of the system and more secure system operation is achieved.
The reactive power provided by the wind farm, specically utilizing the reactive power
to achieve a target voltage is critical in improving rotor angle stability. Therefore, it is
hypothesized that when the synchronous wind farms with high participation factors are
replaced by a DFIG wind farm with terminal voltage control enabled, the rotor angle
stability of the system will be improved. All of the remaining synchronous wind farms,
i.e. those with low participation factors are then modeled as DFIGs with a xed 0.98
leading (capacitive) power factor. The stability of the system for a loss of generation
event will then be assessed by running a transient time domain analysis. The results
will be analyzed using Prony analysis techniques to compare the stability of the full
control case and the synchronous wind case. In Table 5.1, all three of the studied cases
and their description are presented.
Table 5.1: Studied Wind Control Cases
Case Description
Synchronous Wind Case
All wind farms modeled as full synchronous
units with terminal voltage control
Full Control Case
DFIGs modeled with terminal voltage
control implemented at all farms
Critical Wind Case
DFIGs modeled with terminal voltage
control implemented at selected farms
5.2.3 Voltage Stability Comparison
Finally, having completed the rotor angle stability assessment, a voltage stability anal-
ysis is completed for each of the three cases. This ensures that implementing terminal
voltage control at selected farms does not compromise the voltage security of the sys-
tem. To complete this analysis, a fault will be applied at a load bus in the system.
The critical clearing times for each of the three cases will be determined. Based on
these three analysis techniques a comprehensive assessment regarding the rotor angle
stability and voltage stability of the system can be completed.
5.3 Test System
As in Chapter 4, the New England 39 bus system was used as the test system in this
analysis, Fig. 4.1 (Pai, 1979). The ten synchronous units in the system were modeled
as salient pole generators (GENSAL), with AC excitation systems (IEEEX1), steam
turbine governors (TGOV1), and stabilizers (STAB1) (Kundur, 1994). The load was
modeled to include 33% constant current, 33% constant impedance, and 33% constant
power (IEEE, 1995). Wind farms were added to the buses listed in Table 5.2, to achieve
a instantaneous penetration level of 41.3% (2400 MW) for a demand level of 5811 MW.
This was later halved to 1200 MW and a penetration level of 20.7% in order to observe
the system response to various penetration levels. Load was also varied by 1000 MW in
order to conrm the robustness of the methodology. In the synchronous wind case the
wind farms were modeled in the same manner as the ten original units in the system,
salient pole generators with AC excitation systems, steam turbine governors and stabi-
lizers. Each machine had a maximum active power production of 200 MW and with a
reactive power capability of 100 MVAr. The active power that was displaced by the
addition of the wind farms was dispatched equally across the ten original synchronous
units present in the New England 39 bus test system.
Table 5.2: Wind Generation Locations
Connection bus 2, 6, 8, 12, 16, 18, 24, 26, 32, 33, 35, and 37
For the critical farm case and the full control case, the DFIG model used was a generic
model of 1.5 MW GE DFIG machine. The control parameters of the wind turbine are
the standard parameters as described by (Clark et al., 2010). The 1.5 MW GE DFIG
has the capability to control reactive power in three ways, terminal voltage control, xed
power factor control and xed reactive power control. The terminal voltage control and
xed power factor control will be the two control modes that are studied in this chapter.
The wind turbines are initially operated at 100% of their capacity and each farm is 134
turbines aggregated into a single 200 MW farm. This conguration allowed for single-
point voltage control to applied by the entire farm, simulating operation through a
SCADA system.
5.4 Results and Discussion
5.4.1 Small-Signal Analysis
Based on the methodology described in Section 5.2.1, the small signal analysis was
completed for the synchronous wind case. The studied scenarios are listed in Table 5.3.
Here, four scenarios are studied with varied wind and load conditions. This was done
in order to ensure that the small-signal methodology was robust and the results will be
valid regardless of the system conditions.
Table 5.3: System Scenarios for Small-Signal Analysis
Scenario A B C D
Wind 2400 MW 2400 MW 1200 MW 1200 MW
Output Level
Load 5811 MW 6811 MW 5811 MW 6811 MW
Level
The analysis was completed using the DSATools software package (PowerTech, 2011).
Each of the original ten synchronous generators present in the New England 39 bus
test system were individually displaced and the active and reactive power ows were
redispatched. For each of the ten units this was completed and the modes and the
accompanying participation factors for the synchronous wind farms were observed. An
example of this process is presented in Table 5.4.
The same was completed for the displacement of each generator and the synchronous
wind farms that displayed consistently high participation factors were recorded, based
on the methodology described in Fig. 5.1. In each iteration of the small-signal analysis,
synchronous wind farms with larger than average participation factors were noted.
There was no limit on the value of the participation factor, however for each system
conguration, the value of each participating generator was compared to the others
that were present. In the example provided in Table 5.4, generators 24, 26, and 33 had
signicantly higher participation factors than the other wind farms and were therefore
determined to be signicant for the displacement of generator 30. After completing
Table 5.4: Modes and Participation Factors for the Displacement of Generator 30 for
Scenario A
Generator Participating Participation
Mode Wind Farm Factor
31 26 0.04
33
26 0.81
24 0.13
33 0.11
33
24 0.91
26 0.88
33 0.39
2 0.18
34 33 0.06
35 33 0.02
37
37 0.11
2 0.04
38 32 0.07
39
26 0.06
24 0.04
ten small-signal analyses, six farms presented consistently high average participation
factors and were identied as critical farms. The average participation factors for each
scenario are presented in Table 5.5. The farms in bold are deemed critical.
The average participation factors for the farms located at buses 2, 24, 26, 32, 33 and
37 are noticeably higher in comparison to the other farms for each of the four scenarios.
This is due to the fact that in the power ow dispatch, these farms have a signicantly
higher reactive power margin, Table 5.6 in each scenario. As seen in the table, the
farms with the greatest capability to provide the system with reactive power have the
largest participation factors for each of the scenarios regardless of active power ows and
loading level and is driven by the network conguration and impedances. This indicates
that when, active power ows are redispatched, the ability of a farm to provide reactive
power support to the system is critical and as such, these farms should have terminal
voltage control enabled. This will allow the farm to provide dynamic reactive power
Table 5.5: Average Participation Factors for the Synchronous Wind Farms
Wind Farm Scenario Scenario Scenario Scenario
Bus Locations A B C D
2 0.305 0.343 0.398 0.411
6 0.058 0.069 0.041 0.038
8 0.079 0.088 0.066 0.053
12 0.044 0.037 0.012 0.022
16 0.088 0.091 0.093 0.082
18 0.063 0.053 0.044 0.033
24 0.321 0.367 0.366 0.287
26 0.334 0.346 0.329 0.372
32 0.301 0.391 0.402 0.311
33 0.322 0.331 0.377 0.310
35 0.121 0.091 0.098 0.102
37 0.299 0.312 0.353 0.363
support at a designated remote bus, in this case, the 110 kV bus to which the wind farm
is connected. This is due to the fact that the reactive power requirement is a function
of system impedances and loading rather than active power ows. In order to compare
the stability of one case to another, it was necessary to complete a transient analysis
to assess the rotor angle stability of the system in the time domain. The results of this
analysis are discussed in the next section.
5.4.2 Transient Analysis
The transient analysis was completed for each of the three wind cases dened in the
previous section. For each analysis, the contingency was the loss of the generator
located at bus 32. The generator that responded with the largest deviation following
the event was identied and used to present the results. In Fig. 5.2 the rotor angle for
a generator at bus 34 for the three cases is given following the contingency. To further
analyze the response of the machine, a Prony analysis of each trace was completed.
The results of the synchronous case are seen in Table 5.7.
Table 5.6: Reactive Power Production for the Synchronous Wind Farms
Wind Farm Scenario Scenario Scenario Scenario
Bus Locations A B C D
2 -90.16 -98.59 -86.04 -93.37
6 100.0 100.0 100.0 100.0
8 100.0 100.0 100.0 100.0
12 100.0 100.0 100.0 100.0
16 100.0 100.0 100. 0 100.0
18 60.39 45.22 100.0 7.53
24 -75.04 -85.20 -67.60 7.53
26 -85.21 -97.06 -78.24 -86.15
32 9.76 -0.24 9.76 1.11
33 -40.94 -51.25 -40.94 -51.25
35 41.03 32.90 42.0 21.0
37 -41.71 -51.23 -41.71 -51.23
Figure 5.2: The rotor angle for generator 34 following the loss of generator 32.
From the Prony analysis it is observed that there are four modes of magnitude greater
than 1.0 that have damping levels below 10%. To compare the stability of the syn-
chronous case to the full control case, a Prony analysis was also completed for the
Table 5.7: Prony Analysis of Generator 34 for the Synchronous Case at 41.3%
Magnitude Phase Freq. (Hz) Damping (%)
4.32 175.2
0.981 6.02
1.97 -13.12
1.197 6.84
1.73 151.3
0.885 6.50
1.05 -167.9
1.044 5.54
rotor angle trace in Fig. 5.2. The results are given in Table 5.8. In comparing the
details from the Prony analysis for the two cases it is seen that the magnitude of the
largest mode present is signicantly decreased in the full control case compared to the
synchronous case, 2.87 to 4.32. The damping of the mode in the full control case is
also increased. This combined with the decrease in modal magnitude along with the
increase in damping level suggests for the loss of generation contingency, the full control
case provides increased stability over the synchronous wind case. This demonstrates
the value of providing active power from asynchronous generation. To determine if
the methodology presented in Section 5.2.1 is eective, the Prony analysis of the full
control case is compared to the Prony analysis of the critical control case, Table 5.9.
Table 5.8: Prony Analysis of Generator 34 for the Full Control Case at 41.3%
2.87 -175.6
1.010 6.29
1.46 153.9
0.918 6.30
Table 5.9: Prony Analysis of Generator 34 for the Critical Control Case at 41.3%
3.43 178.6
1.010 6.44
1.47 151.4
0.918 6.20
In comparing the full control case and the critical control case very similar results
are noted for the two modes present in each case. The magnitude of the mode at
1.010 Hz is slightly decreased in the full control case, 2.87 compared to 3.43 for the
critical control case; however the damping of the mode in the full control case is also
decreased, 6.29% compared to 6.44% for the critical control case. The characteristics
of the second mode present in both cases are also very similar. This indicates that
providing full terminal voltage control at all the wind farms in the system provides a
minimal increase in rotor angle stability compared to controlling a critical set of farms
identied based on network characteristics. This is particularly benecial in systems
where it is not possible to control all of the farms or for planning purposes. Identifying
locations where providing control is benecial before wind farms are installed will ease
the operational burden on system operators as well as improve the stability of the
remaining synchronous generation.
The results demonstrate, that by identifying locations in the system that are par-
ticularly vulnerable from an electro-mechanical viewpoint and supplementing those
locations with additional reactive power support is particularly benecial in improving
the rotor angle stability of the system. The targeted reactive power control from wind
generation in the system, eases the reactive power burden on the remaining synchronous
units and prevents under-excitation. The improvement in rotor angle stability is also
reected in the behavior of the participation factors of the existing synchronous units.
5.4.3 Impact of Wind Power Output Level
The previous analysis was completed at a high penetration level of 41.3%. To see if
these results are dependent on the active power output level of wind, the production of
wind farms was halved to achieve a instantaneous penetration level of 20.7% and the
transient analysis was complete for all three cases. The same contingency as before was
applied, the loss of generator 32, and the rotor angles for generator 34 can be seen in
Fig. 5.3. A Prony analysis was completed for each of the rotor angles, with the results
for the synchronous case presented in Table 5.10.
The low frequency, under-damped modes that are present in the higher penetration
case (41.3%) and in the lower penetration case (20.7%) are very similar, indicating that
the rotor angle stability is unaected by penetration level when all of the active power
in the system is provided by synchronous generation. The Prony analysis was also
completed for the critical farm case and the full control case at the 20.7% penetration
Figure 5.3: The rotor angle for generator 34 following the loss of generator 32 at the
20.7% level.
Table 5.10: Prony Analysis of Generator 34 for the Synchronous Case at 20.7%
4.29 170.9
0.983 5.71
1.86 151.2
0.881 6.85
1.25 -3.49
1.156 7.03
1.22 169.9
1.049 6.38
level and are presented in Table 5.11. Once again, a similar response to the synchronous
case is seen; the change in penetration level has very little impact on the modal behavior
of the rotor angle and both the full control case and critical control case show improved
levels of rotor angle stability in comparison to the synchronous wind case. As a result,
providing terminal voltage at the farms specied in Table 5.5 provides improved rotor
angle stability regardless of wind power output level. This conrms the results seen
in Section 5.4.1, and once again demonstrates that the selection of the critical wind
farm buses is a function of the electromagnetic interaction between machines and the
network impedances. This allows reactive power control from the wind farms to aid
in improving the rotor angle stability of the system regardless of load or wind output
level.
Table 5.11: Prony Analysis of Generator 34 at 20.7%
Full Control Case
3.27 -178.9
1.017 5.77
1.79 151.3
0.920 6.49
Critical Control Case
3.15 -177.0
1.012 6.24
1.47 151.4
0.918 6.15
5.4.4 Voltage Stability Analysis
The nal analysis completed was a voltage stability analysis to ensure that the critical
control case did not decrease the level of voltage stability in the system due to the
decreased level of control. To complete this, the three studied cases were subjected to
a CCT analysis. A three-phase bus fault was applied at bus 19, and the three CCTs
were recorded, Table 5.12, initially for the higher penetration level (41.3%).
Table 5.12: Critical Clearing Times at 41.3% for Fault at Bus 14
Case Critical Clearing Time (cycles)
Synchronous Wind Case 14.18
Full Control Case 15.35
Critical Control Case 15.35
From Table 5.12, it can be seen that the critical control case and the full control case
are more secure from a voltage stability viewpoint in comparison to the synchronous
wind case. This is due to the fact that the synchronism of the system is increased
when there is less active power provided by the synchronous generation in the system.
Thus replacing synchronous generators with asynchronous wind farms increases the
synchronous strength of the system and as such, introducing wind generation increases
the CCT of the system. There is no dierence between the critical control case and full
control case, indicating that the stability of the system has not decreased as less control
is applied by the wind generation. This also holds at the lower penetration level, Table
5.13
Table 5.13: Critical Clearing Times at 20.7% for Fault at Bus 14
Case Critical Clearing Time (cycles)
Synchronous Wind Case 13.01
Full Control Case 14.96
Critical Control Case 14.40
At the 20.7% penetration level, the same trend as the 41.3% holds and the critical
control case and full control case show improved CCTs over the synchronous wind case.
However, the CCTs are signicantly lower at the lower penetration level. This is due to
the fact that larger percentage of the systems active power is provided by synchronous
generation than at the 41.3% penetration level. As a result, the system is more sensitive
to application of a bus fault and the CCT decreases at the 20.7% penetration level.
Finally, a case with a 0% wind penetration was run, to determine the critical clearing
time of the original New England 39 bus test system and was determined to be 7.93
cycles.
5.5 Conclusion
This chapter has developed a methodology that identies the areas of a system that are
critical from an electro-mechanical viewpoint for a fully synchronous system. Based on
the known locations of wind farms, reactive power can be controlled to achieve a target
voltage, allowing for improved rotor angle stability in the existing synchronous units.
While it is benecial to control the reactive power output of all the wind farms in the
system, it does not present a signicant improvement over the methodology described
in this chapter. It is more critical to system stability to control the voltage at specic
locations. The selection of these locations is determined by the conguration of the
network, i.e. the location of the load and their reactive power requirements. This will
be benecial to the system operator when reactive power control capability is limited,
and critical nodes of the network need to be identied. The criticality of a farm is a
function of the network and the available reactive power capability at the farm and is
reected in the average participation factor of the farm following a contingency event.
This chapter also demonstrates the signicant impact wind penetration and loading
will have on system stability and why it is often even more critical to control voltage
at lower penetration levels. When more synchronous units are online, the synchronism
tying the system together is stronger. Therefore, the importance of providing dynamic
reactive power support to allow the system to ride through a fault and maintain syn-
chronism is greater. However, once again, the nodes at which the reactive power control
needs to be provided will be determined by the impedance of the system and how the
load is distributed across the system. The methodology presented here can be applied
to any system and will determine critical control locations based on the specic net-
work congurations. The characteristics of the network determine the need for reactive
power support, and by providing robust and reliable voltages at these locations, more
secure system operation is achieved.
CHAPTER 6
Conclusions and Scope for Future Work
6.1 Conclusions
A
S wind generation continues to grow, it can prove to be a valuable resource in
modern power systems. With zero fuel cost, it can be a sustainable, zero emission
source of generation that will help power systems across the world meet their goals of
CO
2
reduction and reduce their reliance on other carbon intensive fuels. To achieve this
goal, it will be necessary to properly understand the impacts wind generation will have
on the stability of the power system. The electrical and mechanical synchronism that
ties power systems together is achieved by the presence of conventional synchronous
generation. As wind generation increases, the asynchronous active power wind gen-
erators inject into the system will have consequences that will need to be addressed.
These issues can be mitigated by utilizing the reactive power capabilities of variable
speed wind turbines present in power systems today.
86
Chapter 6. Conclusions and Scope for Future Work 87
Chapter 2 addressed how the variable and uncertain nature of the wind generation
will impact the small-disturbance, long-term voltage stability of the system as well as
how traditional study scenarios can often be insucient in providing the full picture
in terms of the impact on stability. Power ow studies have been completed based on
snapshots of the power system under their maximum stressed conditions. Generally
this occurs under maximum or minimum load. When large levels of wind generation
are introduced into the power system this point may no longer be valid in terms of
determining the change in the systems operational security. To understand the full
impacts on the systems security, it will be critical to utilize historical data from both
wind and load to identify when the system will truly be the most severely stressed.
By completing an analysis a priori, this point can be identied and by utilizing the
methodology developed in Chapter 2, the time-series power ow, the full impact of
wind generation on the small-disturbance, long-term voltage stability can be assessed.
Further, by utilizing wind generation to provide reactive power support in the form of
terminal voltage control, the voltage stability margin of the system can be extended and
provide the system with increased operational security. The production and control of
reactive power from wind generation will be a key tool in achieving high levels of not
only voltage stability, but also across the power system stability spectrum.
Chapter 3 examined how large-disturbance, short-term voltage stability of the power
system will be inuenced by the presence of Type-I and Type-III WTGs. Type-I
turbines are among the rst generation of wind turbines and often present in power
systems. They provide the system with a natural inertial response and can aid in im-
proving the systems frequency stability. However, they are large consumers of reactive
power and have negative impacts on the voltage stability of the power system. By
utilizing the reactive power capabilities of the Type-III, these negative impacts can
be mitigated and the power system can have the benet of improved frequency re-
sponse without sacricing the voltage stability of the system. In smaller, more isolated
power systems with moderate penetration levels of Type-I wind generation, this can be
of particular benet since the sensitivity of the frequency response will be increased.
Once again, the control of reactive power by wind generation is a very useful tool in
maintaining the stability of the system.
Chapter 4 introduces issues related to wind generation and the rotor angle stability
of conventional synchronous generators in the system. The active power injections
provided by wind generation are asynchronous and do not provide the system with any
damping torque. As a result, greater stress is placed on the remaining conventional
synchronous generators when there are higher penetration levels of wind generation
present in the system. The displacement of conventional synchronous generators by
wind generators will have signicant impacts on the power system stability, particularly
the rotor angle stability which will be greatly inuenced by the voltage stability of the
system. Traditionally, the issues of rotor angle stability and voltage stability have
been decoupled in power systems due to the presence of automatic voltage regulation
from conventional synchronous generation. However, as wind penetration levels grow,
conventional synchronous generation will be displaced along with their AVRs. This
displacement of AVR capability will mean that the voltage stability of the power system
will be signicantly impacted by the type of reactive power support provided by wind
generation. In many power systems, grid code simply stipulates that wind generation
need only provide bulk reactive power support to the system, similar to the manner
of switched shunt capacitors. However, this type of control coupled with a decrease
in the presence of AVRs may lead to under-excitation in the remaining conventional
synchronous units. This will lead to larger deviations in the machines eld voltage
which in turn will lead to larger angular deviations. However, this separation can be
minimized and improved system security can be achieved if wind generators provide
terminal voltage control. By directly supporting and controlling the voltage of buses in
the system, the voltage stability of the conventional synchronous generators is improved
and the likelihood of under-excitation is decreased. This results in a direct improvement
in the rotor angle stability of the conventional synchronous generators.
Finally, Chapter 5, develops a robust reactive power control strategy for wind gen-
eration for improved rotor angle stability. Based on the relationship established in
Chapter 4, Chapter 5 provides a methodology that identies nodes in the system that
are weak from a voltage stability viewpoint. By providing reactive power support and
controlling the voltage of these selected nodes using wind generation, the rotor angle
stability of the system can be improved without sacricing the level of voltage security
in the system. This is particularly useful in systems where it may not be possible to
control all of the wind generation. This could be due to several factors; the wind tur-
bines are older models where reactive power cannot be controlled, or SCADA systems
may not be present, or wind generation is embedded within the distribution network
and coordinated control is not possible. The methodology presented in Chapter 5,
identies critical locations in the system and demonstrates that controlling the voltage
at these selected locations is sucient to maintain the rotor angle stability of the con-
ventional synchronous generators in the system. This allows the remaining turbines to
be operated using a simpler control strategy such as xed power factor control without
any adverse impacts on system stability.
The overall goal of this thesis has been to demonstrate the changing nature of the
power system and how as wind penetration levels increase novel methods and strategies
will need to be implemented in order to maintain the high levels of security that are
required of power systems. The traditional study methods and mitigation techniques
that have been utilized may not be sucient to achieve stable power system operation.
To sustain high levels of operational security and stability the ability to control reactive
power by wind generation will become critically important. By maintaining robust and
balanced voltage across a power system two key aspects of power system stability are
improved, the voltage stability and the rotor angle stability. These two issues are
critical in maintaining the synchronism that binds together the power system, and as
wind generation continues to become an integral element in power systems, it will be
essential that reactive power from wind generation is utilized to its full extent in order
to maintain the stringent levels of security and reliability that are required of modern
power systems.
6.2 Scope for Future Work
In the future, analysis in this area will further explore the role reactive power plays in
maintaining the operational security of power systems. Utilizing data from synchro-
nized phasor measurement units (PMUs) can prove to be a valuable tool in assessing
power system stability as more wind generation is introduced into the system. PMUs
have the capability to provide synchronized data that can be used to trace the reactive
power ows in the power system and can be used to identify locations in the network
that are weak from a voltage stability viewpoint. This can lead to signicant improve-
ments in voltage stability and aid system operators in deploying mitigation resources.
How wind generation will play a role as one of those mitigation resources will also be a
signicant area of research. Developing coordinated control strategies that allow wind
turbines and wind farms to work together to provide reactive power support will be
critical for continued stable system operation.
As more asynchronous active power ows from wind generation are introduced into
the power system, work will need to be completed that will examine how reactive
power can be utilized to reduce active power losses in the system. Similar to the
methodologies discussed in this thesis, a system optimized to minimize losses will seek
to achieve balanced voltages across the system. Since these two concepts are inherently
related, the benets of loss optimization along coupled with reactive power support will
be studied in the future. By minimizing system losses with the resources available in
the system, more ecient and secure system operation can be achieved.
This thesis focused on how reactive power can be a valuable resource in minimizing
rotor angle deviation in conventional synchronous generators in the system and can aid
in maintaining the electrical synchronism of the power system. With additional control
systems, wind turbines could be congured to provide some mechanical interaction
with the power system (Lalor et al., 2005; Miao et al., 2009; Ekanayake and Jenkins,
2004). This can aid in providing oscillation stabilization similar to a power system
stabilizer found in a conventional synchronous generator. A controller congured in
this manner is very similar to the inertial response emulation controller in development
by several turbine manufacturers today. By mimicking stabilizers and synchronous
inertial response, wind generators could potentially provide the system with signicant
mitigation tools that will aid in maintaining the mechanical synchronism of the power
system. As wind penetration continues to increase in power systems, it will be critical
that wind turbines continue to evolve from simple sources of active power to signicant
contributors to the entire spectrum of power system stability.
References
Ackermann, T. (Ed.), 2005. Wind Power in Power Systems. John Wiley and Sons.
Akhmatov, V., Eriksen, P., August 2007. A large wind power system in almost island
operation; a danish case study. IEEE Transactions onPower Systems, 22 (3), 937
943.
Allan, R., Leite da Silva, A., September 1981. Probabilistic load ow using multilin-
earisations. IEE Proceedings on Generation, Transmission and Distribution 128 (5),
280 287.
Anderson, P. M., Fouad, A. A., 2003. Power System Control and Stability. IEEE Press.
AWEA, 2011. Wind industry statistics.
URL http://www.awea.org/learnabout/industry_stats/index.cfm
Bakirtzis, A., Dokopoulos, P., August 1988. Short term generation scheduling in a
small autonomous system with unconventional energy sources. IEEE Transactions
on Power Systems, 3 (3), 1230 1236.
Bakirtzis, A. G., Dokopoulos, P. S., Gavanidou, E. S., Ketselides, M. A., March 1989. A
probabilistic costing method for the evaluation of the performance of grid-connected
wind arrays. IEEE Power Engineering Review, 9 (3), 28 29.
Baldick, R., 1995. The generalized unit commitment problem. IEEE Transactions on
Power Systems 10 (1), 465475.
Begovic, M., Phadke, A., February 1992. Control of voltage stability using sensitivity
analysis. IEEE Transactions on Power Systems, 7 (1), 114 123.
Bergen, A., Vittal, V., 2000. Power Systems Analysis. Prentice Hall.
92
References 93
Billinton, R., Chen, H., Ghajar, R., 1996. A sequential simulation technique for ade-
quacy evaluation of generating systems including wind energy. IEEE Transactions
on Energy Conversion, 11 (4), 728734.
Billinton, R., Wangdee, W., February 2007. Reliability-based transmission reinforce-
ment planning associated with large-scale wind farms. IEEE Transactions on Power
Systems, 22 (1), 34 41.
Boehme, T., Wallace, A., Harrison, G., August 2007. Applying time series to power
ow analysis in networks with high wind penetration. IEEE Transactions on Power
Systems, 22 (3), 951 957.
Borkowska, B., 1974. Probabilistic load ow. IEEE Transactions on Power Apparatus
and Systems, (3), 752759.
Burke, D., OMalley, M., May 2008. Optimal wind power location on transmission
systems - a probabilistic load ow approach. In: Probabilistic Methods Applied to
Power Systems, 2008. PMAPS 08. Proceedings of the 10th International Conference
on. pp. 1 8.
Cartwright, P., Holdsworth, L., Ekanayake, J., Jenkins, N., July 2004a. Co-ordinated
voltage control strategy for a doubly-fed induction generator (dg)-based wind farm.
IEE Proceedings on Generation, Transmission and Distribution, 151 (4), 495 502.
Cartwright, P., Holdsworth, L., Ekanayake, J. B., Jenkins, N., 2004b. Co-ordinated
voltage control strategy for a doubly-fed induction generator (dg)-based wind farm.
IEE Proceedings on Generation Transmission and Distribution 151.
Clark, K., Miller, N., Sanchez-Casca, J., 2010. Modeling of GE Wind-Turbine Genera-
tors for Grid Studies Version 4.5,. GE Energy.
Coughlan, Y., Smith, P., Mullane, A., OMalley, M., August 2007. Wind turbine mod-
elling for power system stability analysis; a system operator perspective. IEEE Trans-
actions on Power Systems, 22 (3), 929 936.
Ding, Y., Wang, P., Goel, L., Loh, P. C., Wu, Q., May 2011. Long-term reserve expan-
sion of power systems with high wind power penetration using universal generating
function methods. IEEE Transactions on Power Systems, 26 (2), 766 774.
Doherty, R., Lalor, G., OMalley, M., 2005. Frequency control in competitive electricity
market dispatch. IEEE Transactions on Power Systems, 20 (3), 15881596.
EirGrid, October 2010a. Facilitation of renewables wp3 nal report.
URL http://www.eirgrid.com/media/FacilitationRenewablesFinalStudyReport.
pdf
EirGrid, December 2010b. Generation adequacy report.
URL http://www.eirgrid.com/media/GCS%202011-2020%20as%20published%
2022%20Dec.pdf
References 94
Ekanayake, J., Jenkins, N., December 2004. Comparison of the response of doubly fed
and xed-speed induction generator wind turbines to changes in network frequency.
IEEE Transactions on Energy Conversion, 19 (4), 800 802.
El Moursi, M., Joos, G., Abbey, C., 2008. A secondary voltage control strategy for
transmission level interconnection of wind generation. IEEE Transactions on Power
Electronics, 23 (3), 11781190.
ESB, 2008. All-island grid study workstream 1 renewable energy resource assessment.
URL http://www.dcenr.gov.ie/Energy/North-South+Co-operation+in+the+
Energy+Sector/All+Island+Electricity+Grid+Study.htm
EWEA, 2011. Pure energy; wind energy targets for 2020 and 2030.
URL http://www.ewea.org/fileadmin/ewea_documents/documents/
publications/reports/Pure_Power_III.pdf
Gao, B., Morison, G., Kundur, P., November 1992. Voltage stability evaluation using
modal analysis. IEEE Transactions onPower Systems, 7 (4), 1529 1542.
Gautam, D., Vittal, V., Harbour, T., August 2009. Impact of increased penetration of
dg-based wind turbine generators on transient and small signal stability of power
systems. IEEE Transactions on Power Systems, 24 (3), 1426 1434.
Hatziargyriou, N., Karakatsanis, T., Papadopoulos, M., February 1993. Probabilistic
load ow in distribution systems containing dispersed wind power generation. IEEE
Transactions on Power Systems, 8 (1), 159 165.
Holdsworth, L., Wu, X., Ekanayake, J., Jenkins, N., May 2003. Comparison of xed
speed and doubly-fed induction wind turbines during power system disturbances.
IEE Proceedings on Generation, Transmission and Distribution, 150 (3), 343 352.
IEEE, August 1995. Standard load models for power ow and dynamic performance
simulation. IEEE Transactions on Power Systems, 10 (3), 1302 1313.
Ilic, M., Liu, X., Leung, G., Athans, M., Vialas, C., Pruvot, P., 1995. Improved
secondary and new tertiary voltage control. IEEE Transactions on Power Systems
10 (4), 18511862.
IPCC, 2011. Special report on renewable energy resources and climate change.
URL http://srren.ipcc-wg3.de/report/IPCC_SRREN_TS
Kayikci, M., Milanovic, J., June 2007. Reactive power control strategies for dg-based
plants. IEEE Transactions on Energy Conversion, 22 (2), 389 396.
Keane, A., Ochoa, L., Vittal, E., Dent, C., Harrison, G., February 2011. Enhanced uti-
lization of voltage control resources with distributed generation. IEEE Transactions
on Power Systems, 26 (1), 252 260.
Kennedy, J. M., Fox, B., Littler, T., Flynn, D., August 2011. Validation of xed speed
induction generator models for inertial response using wind farm measurements.
IEEE Transactions on Power Systems, 26 (3), 1454 1461.
References 95
Kimbark, E., 1956. Power System Stability: Volume III Synchronous Machines. IEEE
Press.
Kundur, P., 1994. Power system stability and control. McGraw-Hill New York.
Kundur, P., Paserba, J., Ajjarapu, V., Andersson, G., Bose, A., Canizares, C., Ha-
tiziargyriou, N., Hill, D., Stankovic, A., Taylor, C., Cutsem, T. V., , Vittal, V., May
2004. Denition and classication of power system stability. IEEE Transactions on
Power Systems 19 (2), 13871401.
Lalor, G., Mullane, A., OMalley, M., November 2005. Frequency control and wind
turbine technologies. IEEE Transactions on Power Systems, 20 (4), 1905 1913.
Leite da Silva, A., Arienti, V., July 1990. Probabilistic load ow by a multilinear sim-
ulation algorithm. IEE Proceedings on Generation, Transmission and Distribution,
137 (4), 276 282.
Leite da Silva, A., Ribeiro, S., Arienti, V., Allan, R., Do Coutto Filho, M., November
1990. Probabilistic load ow techniques applied to power system expansion planning.
Littler, T., Fox, B., Flynn, D., june 2005. Measurement-based estimation of wind farm
inertia. In: 2005 IEEE Power Tech. pp. 1 5.
Machowski, J., Bialek, J., Bumby, J., 2008. Power system dynamics: stability and
control. Wiley.
Madureira, A., Pecas Lopes, J., December 2009. Coordinated voltage support in distri-
bution networks with distributed generation and microgrids. IET Renewable Power
Generation, 3 (4), 439 454.
Matos, M., Bessa, R., May 2011. Setting the operating reserve using probabilistic wind
power forecasts. IEEE Transactions on Power Systems, 26 (2), 594 603.
Meibom, P., Barth, R., Brand, H., Hasche, B., Swider, D., Ravn, H., Weber, C., 2008.
All Island Grid Study. Wind Variability Management Studies.
Meibom, P., Barth, R., Hasche, B., Brand, H., Weber, C., OMalley, M., August
2011. Stochastic optimization model to study the operational impacts of high wind
penetrations in ireland. IEEE Transactions on Power Systems, 26 (3), 1367 1379.
Miao, Z., Fan, L., Osborn, D., Yuvarajan, S., June 2009. Control of dg-based wind
generation to improve interarea oscillation damping. IEEE Transactions on Energy
Conversion, 24 (2), 415 422.
Morren, J., de Haan, S., Kling, W., Ferreira, J., February 2006. Wind turbines emulat-
ing inertia and supporting primary frequency control. IEEE Transactions on Power
Systems, 21 (1), 433 434.
NERC, 2010. Integration of variable generation task force.
URL http://www.nerc.com/filez/ivgtf_Interconnection.html
References 96
NREL, Energy, G., 2010. Western wind and solar integration study.
URL http://www.nrel.gov/wind/systemsintegration/pdfs/2010/wwsis_
final_report.pdf
NREL, EnerNex, 2011. Eastern wind integration and transmission study.
URL http://www.nrel.gov/wind/systemsintegration/pdfs/2010/ewits_
final_report.pdf
Pai, M., 1979. Computer techniques in power system analysis. Tata McGraw-Hill.
Papathanassiou, S., Papadopoulos, M., December 2001. Mechanical stresses in xed-
speed wind turbines due to network disturbances. IEEE Transactions on Energy
Conversion, 16 (4), 361 367.
Petersson, A., Thiringer, T., Harnefors, L., Petru, T., December 2005. Modeling and
experimental verication of grid interaction of a dg wind turbine. IEEE Transactions
on Energy Conversion, 20 (4), 878 886.
PowerTech, 2011. Dsatools user manual.
URL http://www.dsatools.com/html/prod_overview.php
Sauer, P., Pai, M. A., 1998. Power System Dynamics and Stability. Prentice Hall.
Siemens, P., 2005. Pss/e 30.2 program operational manual. Volume II.
Soder, L., May 1993. Reserve margin planning in a wind-hydro-thermal power system.
Steinmetz, C. P., July 1920. Power control and stability of electric generating stations.
Transactions of the American Institute of Electrical Engineers, XXXIX (2), 1215
1287.
Steurer, M., Langston, J., Suryanarayanan, S., Ribeiro, P., Meeker, R., Srensen, P.,
2007. Model validation and voltage deviation analysis of an existing wind farm using
high delity real time digital simulation. In: Proceedings 19th International Confer-
ence on Electricity Distribution. pp. 2124.
Tapia, G., Tapia, A., Ostolaza, J., June 2007. Proportional integral regulator-based
approach to wind farm reactive power management for secondary voltage control.
IEEE Transactions on Energy Conversion, 22 (2), 488 498.
Taylor, C., 1993. Power System Voltage Stability. McGraw-Hill.
Thomson, M., Ineld, D., August 2007. Network power-ow analysis for a high pene-
tration of distributed generation. IEEE Transactions on Power Systems, 22 (3), 1157
1162.
Tuohy, A., Denny, E., Meibom, P., Barth, R., OMalley, M., July 2008. Operating
the irish power system with increased levels of wind. In: Power and Energy Society
General Meeting - Conversion and Delivery of Electrical Energy in the 21st Century,
2008 IEEE. pp. 1 4.
References 97
Tuohy, A., Meibom, P., Denny, E., OMalley, M., 2009. Unit commitment for systems
with signicant wind penetration. IEEE Transactions on Power Systems, 24 (2),
592601.
Ummels, B., Gibescu, M., Pelgrum, E., Kling, W., Brand, A., March 2007. Impacts of
wind power on thermal generation unit commitment and dispatch. IEEE Transac-
tions on Energy Conversion, 22 (1), 44 51.
Van Cutsem, T., Vournas, C., February 1996. Voltage stability analysis in transient
and mid-term time scales. IEEE Transactions on Power Systems 11 (1), 146 154.
Van Cutsem, T., Vournas, C., 1998. Voltage stability of electric power systems. Kluwer
Academic Publishers.
Vittal, E., OMalley, M., Keane, A., July 2009. Impact of wind turbine control strategies
on voltage performance. In: Power Energy Society General Meeting, 2009. PES 09.
IEEE. pp. 1 7.
Vovos, P., Kiprakis, A., Wallace, A., Harrison, G., February 2007. Centralized and
distributed voltage control: Impact on distributed generation penetration. IEEE
Transactions on Power Systems, 22 (1), 476 483.
WECC, 2006a. Generic type-3 wind turbine-generator model for grid studies.
WECC, 2006b. Generic type-4 wind turbine-generator model for grid studies,.
Wood, A., Wollenberg, B., 1996. Power Generation Operation and Control. John Wiley
and Sons.
Xu, L., Cartwright, P., September 2006. Direct active and reactive power control of
dg for wind energy generation. IEEE Transactions on Energy Conversion, 21 (3),
750 758.

Eknath Vittal - The Impact of Reactive Power From Wind Generation On Power System Stability

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Eknath Vittal - The Impact of Reactive Power From Wind Generation On Power System Stability

Uploaded by

Copyright:

Available Formats

The Impact of Reactive Power From Wind

Generation on Power System Stability

for the WECC VSC and WECC DFIG cases respectively.

You might also like