559069

September 2019
2018
UNIVERSITY OF GAZİANTEP
M.Sc. in Electrical and Electronics Engineering
GRADUATE SCHOOL OF
NATURAL & APPLIED SCIENCES
Design and Simulation of a Back-to-Back Modular Multilevel

Converter Based Permanent Magnet Synchronous Generator
Wind Energy System Having Ultracapacitor Energy Storage
M.Sc. THESIS
IN
ELECTRICAL AND ELECTRONICS ENGINEERING
NISSAN SABBAR THUMAIL
BY
NISSAN SABBAR THUMAIL

September 2019
Design and Simulation of a Back-to-Back Modular Multilevel Converter Based
Permanent Magnet Synchronous Generator Wind Energy System Having
Ultracapacitor Energy Storage
M.Sc. Thesis
in
Electrical and Electronics Engineering
University of Gaziantep
Supervisor
Assoc. Prof. Dr. Ahmet Mete VURAL
By
Nissan Sabbar Thumail
September 2019
II
© 2019 [Nissan Sabbar THUMAIL]
III
REPUBLIC OF TURKEY
UNIVERSITY OF GAZIANTEP
GRADUATE SCHOOL OF NATURAL & APPLIED SCIENCES
ELECTRICAL AND ELECTRONICS ENGINEERING
Name of the thesis: Design and Simulation of a Back-to-Back Modular Multilevel

. Converter Based Permanent Magnet Synchronous Generator
. Wind Energy System Having Ultracapacitor Energy Storage
Name of the student: Nissan Sabbar THUMAIL
Exam Date: 05/09/2019
Approval of Graduate School of Natural and Applied Sciences
Prof. Dr. A. Necmeddin YAZICI

Director
I certify that this thesis satisfies all the requirements as a thesis for the degree of Master
of Science.
Prof. Dr. Ergun ERÇELEBİ

Head of Department
This is to certify that we have read this thesis and that in our consensus; it is fully
adequate, in scope and quality, as a thesis for the degree of Master of Science.
Assoc.Prof.Dr. Ahmet Mete VURAL

Supervisor
Examining Committee Members: Signature
Assoc. Prof. Dr. Ahmet Mete VURAL …………………..
Assist. Prof. Dr. Serkan ÖZBAY …………………..
Assist. Prof. Dr. Kadir Sercan BAYRAM …………………..
IV
I now declare that all information in this document has been obtained and
presented in agreement with academic rules and ethical conduct. I also declare
that, as required by these rules and conduct, I have fully cited and referenced all
material and results that are not original to this work.
Nissan Sabbar Thumail
V
ABSTRACT
Design and Simulation of a Back-to-Back Modular Multilevel Converter Based
Permanent Magnet Synchronous Generator Wind Energy System Having
Ultracapacitor Energy Storage
THUMAIL, Nissan Sabbar
M. Sc. in Electrical and Electronics Engineering
Supervisor: Assoc. Prof. Dr. Ahmet Mete VURAL
September 2019
92 Pages
Wind energy technology is in progress with increasing output power and high
controllability. It has numerous aptitudes to be utilized in micro and smart grids due
to its cleanness and high level of availability. Large turbines have been progressively
designed with permanent magnet synchronous generators (PMSG) due to their higher
efficiency, reliability, and controllability compared with other generator types.
Modular multilevel converter (MMC) has been proposed for the transfer of huge
amounts of power over long distances under high DC voltage. It has a modular
structure, low harmonic content, and high controllability. On the other hand,
ultracapacitor is the energy storage device which is used in electrical vehicles and
electrical energy grid in order to increase the reliability and stability. In this thesis a
wind energy conversion system (WECS) based on PMSG, back-to-back MMCs, and
ultracapacitor is designed and modelled. It is shown by simulation studies that using
ultracapacitor in the WECS improves the dynamic performance of the system under
different types of disturbance conditions.
Key Words: Wind Energy Conversion System; Renewable Energy; Permanent

Magnet Synchronous Generator; Modular Multilevel Converter, Ultracapacitor.
VI
ÖZET
Ultrakapasitör Enerji Depolamaya Sahip Sırt-Sırta Bağlı Modüler Çokseviyeli
Çevirgeç Tabanlı Sabit Mıknatıslı Senkron Jeneratör Rüzgar Enerjisi
Sisteminin Tasarımı ve Benzetimi
THUMAIL, Nissan Sabbar
Yuksek Lisans Tezi, Elektrik ve Elektronik Mühendisliği
Danışman: Doç. Dr. Ahmet Mete VURAL
Eylül 2019
92 Sayfa
Rüzgar enerjisi teknolojisi artan çıkış gücü ve yüksek denetlenebilirliği ile sürekli
gelişmektedir. Temizliği ve yüksek seviyede kullanılabilirliğinden dolayı mikro ve
akıllı şebekelerde kullanılabilecek birçok avantajlara sahiptir. Büyük türbinler artan
bir şekilde diğer jeneratör tipleri ile karşılaştırıldığında yüksek verimlerinden,
güvenirliklerinden ve denetlenebilirliklerinden dolayı sabit mıknatıslı senkron
jeneratörler (SMSJ) ile tasarlanmaktadırlar. Modüler çokseviyeli çevirici (MÇÇ)
büyük miktarlardaki gücü uzun mesafeler üzerinden yüksek DC gerilim altında
transfer için önerilmiştir. Modüler bir yapıya, düşük harmonic içeriğe ve yüksek
denetlenebilirliğe sahiptir. Diğer taraftan, ultrakapasitör, elektrik araçlarında ve
güvenirliği ve kararlılığı artırmak için elektrik enerjisi şebekesinde kullanılan bir enerji
depolama aygıtıdır. Bu tezde, SMSJ, sırt-sırta-bağlı MÇÇler ve ultakapasitör tabanlı
bir rüzgar enerjisi dönüşüm sistemi (REDS) tasarlanmış ve modellenmiştir. Benzetim
çalışmaları ile REDS’de ultrakapasitör kullanımının değişik tiplerdeki bozucu
koşullarında sistemin dinamik performansını iyileştirdiği gösterilmiştir.
Anahtar Kelimeler: Rüzgar Enerjisi Dönüşüm Sistemi; Yenilenebilir Enerji; Sabit

Mıknatıslı Senkron Jeneratör; Modüler Çokseviyeli Çevirici; Ultrakapasitör.
VII
“Dedicated To My Family And Beloved Parents”
VIII
ACKNOWLEDGEMENTS
In the name of Allah, the most gracious, the most merciful. First of all, I would like to
thank Allah who granted me health and strength to overcome life’s difficulties and for
his guidance to complete this master thesis.
I would like to express my gratefulness to my supervisor Assoc. Prof. Dr. Ahmet Mete
VURAL for his guidance, patience, kindness, and encouragement throughout this
research.
I would like to express my thanks to the staff members of the Department of Electrical
and Electronics Engineering at the University of Gaziantep and my thanks to all other
friends for their helping me in preparing this research.
Finally, my grateful thanks to all my family and to my wife for providing support to
me in my study.
IX
TABLE OF CONTENTS
ABSTRACT .............................................................................................................. VI
ÖZET....................................................................................................................... VII
ACKNOWLEDGEMENTS ..................................................................................... IX
TABLE OF CONTENTS.......................................................................................... X
LIST OF FIGURES .............................................................................................. XIV
LIST OF TABLES ............................................................................................. XVIII
LIST OF SYMBOLS / ABBREVIATIONS ........................................................ XIX
CHAPTER 1 ............................................................................................................... 1
INTRODUCTION…………………………………………………………………...1
1.1 Introduction ...................................................................................................... 1
1.2 A Brief History of Wind Energy Systems ........................................................ 2
1.3 Objective of the Thesis ..................................................................................... 5
1.4 Organization of the Thesis ............................................................................... 5
CHAPTER 2 ............................................................................................................... 7
LITERATURE REVIEW.......................................................................................... 7
2.1 Introduction ...................................................................................................... 7

2.2 Literature Review ............................................................................................. 7
2.3 Place of Work in the Literature ...................................................................... 10
CHAPTER 3 ............................................................................................................. 12
THEORETICAL BACKGROUND........................................................................ 12
3.1 Overview ........................................................................................................ 12

3.2 Multilevel Converters ..................................................................................... 13
3.3 The Topologies of Multilevel Converter ........................................................ 14
3.3.1 Flying Capacitor (Capacitor-Clamped) Multilevel Converter ................ 14
3.3.2 Diode-Clamped (Neutral Point Clamped) Multilevel Converter ............ 15
X
3.4 Submodule Based Multilevel Converters ....................................................... 17
3.4.1 Cascaded H-Bridges ................................................................................ 19
3.4.2 Modular Multilevel Converter (MMC) ................................................... 21
3.4.3 Advantage and Disadvantages of MCC .................................................. 25
3.4.4 Other Submodule Topologies.................................................................. 26
3.5 An Overview of Energy Storage Devices ...................................................... 26
3.5.1 Ultracapacitor .......................................................................................... 27
3.5.2 Why Are Ultracapacitors Used in WECS ............................................... 28
3.6 MMC. Switching. Methods ............................................................................ 29
3.6.1 High. Frequency Switching (Carrier Based Switching). ......................... 30
3.6.1.1 Scalar PWM ..................................................................................... 31
3.6.1.1.1 Sub‫ـ‬Harmonic )Level-shift) Methods ........................................... 31
3.6.1.1.2 Phase‫ـ‬Shift Method ....................................................................... 33
3.6.1.2 Space Vector PWM .......................................................................... 34
3.6.2 Low Frequency Switching ...................................................................... 35
3.6.2.1 Selective. Harmonic. Elimination. ................................................... 35
CHAPTER 4 ............................................................................................................. 36
MODELING OF THE WECS ................................................................................ 36
4.1 Introduction .................................................................................................... 36

4.2 Proposed System Modeling ............................................................................ 36
4.2.1 The PMSG ............................................................................................... 38
4.2.2 MMC ....................................................................................................... 39
4.3 The Control Scheme of the Proposed Topology ............................................ 42
4.3.1 Generator Side Model ............................................................................. 42
4.3.1.1 Wind Turbine ................................................................................... 43
4.3.1.2 Drive Train ....................................................................................... 44
4.3.1.3 PMSG ............................................................................................... 44
4.3.1.4 Pitch Angle Controller ..................................................................... 46
4.3.2 MMC Controller...................................................................................... 46
4.3.3 Mathematical Model of the MMC .......................................................... 48
XI
CHAPTER 5 ............................................................................................................. 50
SIMULATION RESULTS ...................................................................................... 50
5.1 Introduction .................................................................................................... 50

5.2 Simulation Software and Model Parameters .................................................. 50
5.2.1 Simulation Parameters............................................................................. 50
5.3 PMSG Connected to Power Grid Tests .......................................................... 51
5.3.1 Rotor Speed, Mechanical torque and the Active/Reactive power of
PMSG ............................................................................................................... 52
5.3.2 PMSG/Grid Voltage, Current and VI power ........................................... 54
5.4 PMSG Connected to Power Grid via back-to-back MMCs Tests .................. 55
5.4.1 Rotor Speed, Mechanical torque and the Active/Reactive power of
PMSG ............................................................................................................... 56
5.4.2 PMSG/Grid Voltage, Current and VI power ........................................... 58
5.4.3 Capacitor Balance and DC Link Voltage and Current ............................ 60
5.4.4 PMSG/Grid Active and Reactive Power ................................................. 61
5.5 PMSG MMC System with Ultracapacitor Tests: ........................................... 62
5.6 Simulation results of PMSG based WECS under variable wind speed
conditions .............................................................................................................. 63
5.6.1 Voltage and current waveforms .............................................................. 63
5.6.2 Active power and reactive power waveforms ......................................... 66
5.6.3 Apparent power waveforms .................................................................... 67
5.6.4 DC link voltage waveforms..................................................................... 68
5.6.5 Voltage, current, and SOC of ultracapacitor ........................................... 68
5.7 Simulation results of PMSG based WECS under phase A to ground fault .... 69
5.7.1 Voltage and Current Waveforms............................................................. 71
5.7.2 Active Power and Reactive Power Waveforms ...................................... 72
5.7.3 DC Link Voltage Readings ..................................................................... 73
5.7.4 Voltage, Current, and SOC of Ultracapacitor ......................................... 74
5.8 Simulation Results of PMSG Based WECS Under Three-phase Fault.......... 74
XII
5.9 Simulation Results of PMSG-based WECS Under Short Circuit in DCLink.80
5.10 Discussion of The Results ............................................................................. 84
CHAPTER 6 ............................................................................................................. 86
CONCLUSION......................................................................................................... 86
6.1 Conclusion ....................................................................................................... 86

6.2 Suggestions for Future Work ......................................................................... 87
REFERENCES ......................................................................................................... 89
XIII
LIST OF FIGURES
Figure 1.1 Comparison between the sources of electric energy in terms of the
amount of emitted co2 per kwh [1] .............................................................................. 2
Figure 1.2 Wind power generation system topologies [8] ………………………….. 4
Figure 3.1 Main parts of a WECS [6] ....................................................................... 12
Figure 3.2 (a) Output voltage waveform of a single-phase five-level flying capacitor
converter (b) The circuit diagram of a single-phase five-level flying capacitor
converter..................................................................................................................... 14
Figure 3.3 (a) Output voltage waveform of a single‫ـ‬phase five-level diode‫ـ‬clamped

converter (b) The circuit diagram of a single‫ـ‬phase five-level diode‫ـ‬clamped
converter..................................................................................................................... 16
Figure3.4 (a) SSBC topology (b) SDBC topology.................................................... 18
Figure 3.5 (a) DSCC and DSBC topology with coupled arm inductors ................... 18
(b) Chopper cell (DSCC) (c) bridge cell (DSBC) ...................................................... 18
Figure 3.6 The topology of the full-bridge circuit .................................................... 20
Figure 3.7 Typical submodule, arm, and converter arrangement for 400 mw MMC
by siemens [27] .......................................................................................................... 22
Figure 3.8 Three-phase MMC circuit topology ........................................................ 23
Figure 3.9 Half‫ـ‬bridge circuit. ................................................................................... 23
Figure 3.10 States of half-bridge circuit and current paths ....................................... 24
Figure 3.11 (a) Diode-clamped. three-level converter (b) Capacitor-clamped .

three-level. converter. (c) Clamp double circuit (d) Five-level cross connected.
circuit [28] .................................................................................................................. 26
Figure 3.12 MMC switching. methods. .................................................................... 30
Figure 3.13 The carriers of phase disposition switching technique .......................... 31
Figure 3.14 The carriers of phase opposition disposition switching technique ........ 32
XIV
Figure 3.15 The carriers of alternative phase opposition disposition switching
technique .................................................................................................................... 32
Figure 3.16 The carriers of phase‫ـ‬shifted switching technique ................................. 33
Figure 3.17 The carriers of sawtooth rotation method .............................................. 33
Figure 3.18 Space vector diagrams of two and five-level converters ....................... 35
Figure 3.19 Waveform of output voltage for five‫ـ‬level converter with selective
harmonic elimination method [33] ............................................................................. 35
Figure 4.1 PMSG-based wecs based on back-to-back MMCs .................................. 36
figure 4.2 The overall design of wecs with back-to-back MMCs and ultracapacitor 37
Figure 4.3 The overall structure of the proposed MMC ........................................... 40
Figure 4.4 One arm structure of the mcc ................................................................... 41
Figure 4.5 Generator side model ............................................................................... 43
Figure 4.6 Drive train ................................................................................................ 44
Figure 4.7 Pitch angle controller ............................................................................... 46
Figure 4.8 MMC control ........................................................................................... 47
Figure 4.9 Circuit structure of MMC [43]................................................................. 48
Figure 5.1. PMSG simulation model that connected with grid under variable wind
speed conditions. ........................................................................................................ 52
Figure 5.2: Simulation results of PMSG under variable wind speed conditions. from
top to bottom: wind speed, rotor speed, mechanical torque, active power, reactive
power .......................................................................................................................... 53
Figure 5.3 Simulation results of PMSG/Grid under variable wind speed conditions.
From top to bottom: voltage, current, apparent power.................................................54
Figure 5.4 Simulation results of PMSG for a period of 9.85s-10s: from top to
bottom: voltage, current, apparent power .................................................................. 54
Figure 5.5 PMSG-MMC system model .................................................................... 56
Figure 5.6 Simulation result for PMSG, from up to down: wind speed (m/sec), rotor
speed (rad/s), torque mechanical (nm ), active power (mw) and reactive power
(mvar) ......................................................................................................................... 57
XV
Figure 5.7 Simulation results of MMCs under variable wind speed conditions. from
top to bottom: voltage, current, apparent power. PMSG side (left) and grid side
(right).......................................................................................................................... 58
Figure 5.8 Simulation results of MMCs under variable wind speed conditions for a
period of 5.94s-6.14s. from top to bottom: voltage, current, apparent power. PMSG
side (left) and grid side (right). .................................................................................. 59
Figure 5.9 Capacitor voltages in each submodule in one arm of both MMCs. from
top to bottom: PMSG side (supply), (b) Grid side (load). ......................................... 60
Figure 5.10 Capacitor voltages in each submodule in one arm of both MMCs for a
period of 17.3s-18s. from top to bottom: PMSG side (supply), (b) grid side (load). 60
Figure 5.11 The variation in dc link current (upper) and dc link voltage (lower). ... 61
Figure 5.12 PMSG side and grid side active and reactive power. ............................ 62
Figure 5.13 Voltage and current waveforms of the supply side of the PMSG based
wecs with no ultracapacitor (left) and with ultracapacitor (right) .............................. 64
wecs with no ultracapacitor (left) and with ultracapacitor (right) for a time period
between 3 and 3.2 s .................................................................................................... 65
Figure 5.15 Active power and reactive power waveforms of the supply side of the
PMSG based wecs with no ultracapacitor (left) and with ultracapacitor (right)........ 66
Figure 5.16 Apparent power waveforms of the supply side of the PMSG based wecs
with no ultracapacitor (left) and with ultracapacitor (right)....................................... 67
Figure 5.17 DC link voltage waveforms of the back-to-back MMCs of the PMSG
based WECS with no ultracapacitor (left) and with ultracapacitor (right)………..….68
Figure 5.18 The voltage, current and SOC readings of the ultracapacitor………….69
Figure 5.19 The modification of the simulation model for phase a to ground fault . 70
Figure 5.20 The parameters of the fault block for phase a to ground fault ............... 70
Figure 5.21 Voltage and current waveforms of the supply side and the load of the
PMSG based wecs under phase a to ground fault with no ultracapacitor (left) and
with ultracapacitor (right) .......................................................................................... 71
Figure 5.22 Active power and reactive power waveforms of the supply side and the
load of the PMSG based wecs under phase a to ground fault with no ultracapacitor
(left) and with ultracapacitor (right) ........................................................................... 72
XVI
Figure 5.23 DC link voltage waveform of the back-to-back MMCs of PMSG based
WECS under phase A to ground fault with no ultracapacitor (left) and with
ultracapacitor (right)…………………...……………………………………………73
Figure 5.24 The voltage, current and soc readings of the ultracapacitor under phase a
to ground fault ............................................................................................................ 74
Figure 5.25 The modification of the simulation model for three-phase to .
ground fault ................................................................................................................ 75
Figure 5.26 The parameters of the fault block for three-phase to ground fault ........ 75
PMSG based wecs under three-phase to ground fault with no ultracapacitor (left) and
load of the PMSG based wecs under three-phase to ground fault with no
ultracapacitor (left) and with ultracapacitor (right) .................................................... 77
wecs under three-phase to ground fault with no ultracapacitor (left) and with
ultracapacitor (right)................................................................................................... 78
Figure 5.30 The voltage, current and soc readings of the ultracapacitor under three-
phase to ground fault .................................................................................................. 79
Figure 5.31 The modification of the simulation model for short circuit at.
the dc link ................................................................................................................... 80
Figure 5.32 The parameters of the fault block for short circuit at the dc link........... 80
PMSG based wecs under short circuit at the dc link with no ultracapacitor (left) and
load of the PMSG based wecs under short circuit at the dc link with no ultracapacitor
(left) and with ultracapacitor (right) ........................................................................... 82
wecs under short-circuit at the dc link with no ultracapacitor (left) and with
ultracapacitor (right)................................................................................................... 83
Figure 5.36 Ultracapacitor readings, from up to down: voltage, current and state of
charge (SOC).............................................................................................................. 84
XVII
LIST OF TABLES
Table 1.1 The effect of different design choices for rotor diameters, hub heights and
turbine sizes on the produced energy. .......................................................................... 3
Table 3.1 The output voltage and switching states for five‫ـ‬level flying capacitor.
converter..................................................................................................................... 15
Table 3.2 The output voltage and switching states for five‫ـ‬level diode-clamped
converter..................................................................................................................... 16
Table 3.3 The possible variations of MMC [25] ....................................................... 17
Table 3.4 The switching states of full-bridge circuit................................................. 20
Table 3.5 Switching. states of half-bridge circuit...................................................... 24
Table 4.1 Wind turbine parameters ........................................................................... 43
Table 5.1 Simulation parameters for PMSG ............................................................. 51
Table 5.2 Simulation parameters for MMC .............................................................. 51
Table 5.3 Simulation parameters for ultracapacitor .................................................. 51
XVIII
LIST OF SYMBOLS / ABBREVIATIONS
AC Alternative Current
APOD Alternative Phase Opposition Disposition
CCC Capacitor Commutated Converters
CO2 The Carbon Dioxide
CSC Current Source Converters
DC Direct Current
DFIG Doubly-Fed Induction Generate
DSBC Double-Star Bridge Cells
DSCC Double-Star Chopper Cells
EMT Electro-Magnetic Transient
ESS Energy Storage System
FACTS Flexible Alternating Current Transmission Systems
FC Fixed Capacitor
FCC Flying Capacitor Converter
FFM Fundamental Frequency Modulation
FSWT Fixed Speed Wind Turbine Generators
G Generator
GSC Grid-Side Converter
GTO Gate Turn-off Thyristor
HAWT Horizontal Axis Wind Turbines
HV High Voltage M
HVAC High Voltage Alternating Current
HVDC High Voltage Direct current
IG Induction Generator
IGBT Insulated Gate Bipolar Transistor
MMC modular Multilevel Converter
N, n Number of Level, capacitor
NPC Neutral-Point Clamped
OWF Offshore Wind Farms
Pu Per Unit
XIX
PCC Point of Common Coupling
PD Phase Disposition
PD-PWM Phase Disposition Pulse Width Modulation
PLL Phase-locked Loop
PMSG Permanent Magnet Synchronous Generator
POD Phase Opposition Disposition
PS Phase Shift
PV Photovoltaic
PWM-VSI Pulse Width Modulation for three-Phase Voltage
Source Inverter
RSC Rotor-Side Converter
SCIG Squirrel Cage Induction Generator
SDBC Single-Delta Bridge Cells
SM Submodules
SSBC Single-Star Bridge Cells
SSCI Sub-Synchronous Control Interaction.
SSO Sub-Synchronous Oscillation.
SVM Space Vector Modulation
THI Third Harmonic Injection
V_dc DC Link Voltage
Vc Voltage in the Submodule Capacitor
VSC Voltage Source Converter
WRIG Wound Rotor Induction Generator
WRSG Wound Rotor Synchronous Generator
Icirc Circulating Current
Varm Arm Voltage
XX
CHAPTER 1
INTRODUCTION
1.1 Introduction
Nowadays experts say that the biggest threat to the planet is the threat of global
warming which is caused by the emissions of the carbon dioxide. The two-thirds of
the carbon dioxide emissions are coming from energy consumption, distribution, and
production, so lessening the carbon emissions from the energy sector is probably the
fastest way to eliminate pollution [1]. Figure 1.1 shows a brief comparison between
the sources of electric energy in terms of the amount of emitted CO2 per kWh. Solar
panels and wind farm turbines are common sight increasingly being used as renewable
energy sources. By looking at the latest statistics, we can clearly compare the
traditional production technologies and renewable energy technologies. Solar is
emitting only 0.07 to 0.2 pounds of CO2 per kilowatt-hour (kWh), while wind is
responsible for only 0.02 to 0.04 pounds of CO2 per kWh. On the other hand, coal
emits between 1.4 and 3.6 pounds of CO2 emissions per kWh. Thus, we see that wind
energy is the best choice for reducing global warming. Another advantage of wind
energy turbine is that it does not need for the water to operate and thus do not waste
water resources or competing with the other important water needs such as drinking
water or agriculture. For these reasons, wind energy has seen a significant growth in
the last years. As a result of the previous reasons, wind turbines have gained increasing
importance throughout the world especially for the power system planners, electricity
wholesale market designers, operators, and policymakers. On the other side, there are
some limitations in the operation and the practical implementation of the wind
turbines. The immaturity in production technologies and initial costs are the two major
disadvantages of wind power systems [2]. The initial costs of wind turbines are still
higher than fossil-fueled based and other traditional power generating techniques. The
other disadvantage is the technology immaturity whereas we still need technology
innovations that increase reliability and energy output while lowering system capital
expenses.
1
Figure 1.1 Comparison between the sources of electric energy in terms of the
amount of emitted CO2 per Kwh [1]
1.2 A Brief History of Wind Energy Systems
Since a long time ago windmills, or rather wind turbines have been used by the human
beings in various fields mainly for water pumping and crop grinding purposes. The
first wind turbine had been built and used for electricity generating purposes in 1887.
It was used for battery charging. After that, the capacity of wind turbines increased
and became more suitable for connecting it with the electrical grid and for generating
energy for traditional purposes. In 1931, USSR erected a 100 kW wind turbine but the
first-megawatt range wind turbine was erected by the USA in 1941 [3]. However, wind
turbines did not occupy an important position among electrical power production
techniques during much of the 20th century because of their high cost and very large
amounts of available fossil fuels. But this did not last long, as with the happening of
oil crises in the 1970s, while attention was again drawn to clean energy, especially to
the wind energy in many countries and the modern era of wind turbines started. For
instance, USA and Denmark energy companies started to build and erect more modern
wind turbines. By the beginning of the 1980s, the maximum available size of wind
turbines was 50 kW and it became 200 kW at the end of 1980. The size of the wind
turbines continued to increase over the previous years. Nowadays, the average size of
the wind turbine is between 2 MW and 3 MW with the possibility of building larger
2
turbines with greater efficiency and economy of scale, but they are still more complex
to build. The maximum obtained energy from a wind turbine is roughly proportional
to the swept area of the rotor. Hence the technology developments in the blade design
increased the capacity of wind turbines. For example, the power output can be
increased by a factor of four by doubling the rotor diameter, which increases the swept
area. Table 1.1 shows the effect of different design choices for rotor diameters, hub
heights and turbine sizes on the energy produced in Denmark which is a dominant
country in this field [4].
Table 1.1 The effect of different design choices for rotor diameters, hub heights and
turbine sizes on the produced energy.
Power Rating (MW) Rotor (m) Hub Height (m) Annual Energy
Production (MWh)
3.0 90 90 7497
3.0 112 94 10384
3.0 90 80 7089
1.8 80 80 6047
To be able to understand this type of systems, we need to get a knowledge from several
areas such as aerodynamics, electric machinery, power systems, and mechanics. All
the mentioned parts must be integrated with each other in order to design and
implement a cost‫ـ‬efficient wind power system that has low environmental impact and
transfer the produced power from the wind power plants to the costumer. First of all,
the electrical system of wind turbines is considered as the heart of the wind energy
system. The function of this system in wind turbines is to convert the mechanical
energy in the form of kinetic energy in wind currents into electrical energy that can be
utilized and deliver into the electrical grid. However, there are many factors that should
be considered in the design of this electrical system. The first factor is less losses and
the size of the system should be optimized. The second condition is that the system
should be able to regulate the speed and torque on the machine shaft in order to keep
the stresses on the structure to a minimum value and get the maximum captured energy.
The third condition is that the system should be able to inject high quality and
continuous electrical energy fed from the turbine to the electrical grid [5]. Over the
previous years, the turbines were fixed speed turbines but in the 1990s the variable
speed turbine generator was produced as a commercial product after as it was only a
research topic early. As a result for this development in manufacturing techniques,
3
many wind based power generation systems have been produced such as Wound Rotor
Induction Generator (WRIG), Wound Rotor Synchronous Generator (WRSG), Doubly
Fed Induction Generator (DFIG), and Permanent Magnet Synchronous Generator
(PMSG). There are other types of generators such as brushless doubly fed induction
generator, direct-drive synchronous generator, Squirrel Cage Induction Generator
(SCIG), and the multiple-stage geared system. The most common types are the
direct‫ـ‬drive synchronous generator system and (DFIG) due to their simplicity and
reliability [6]. We will explain and discuss the commonly used technologies for wind
generating systems with their design, function, and properties in detail in Chapter 3.
Over the previous years, most of the installed turbines were onshore turbines but the
growth of this type is limited because of the visual pollution that is caused by large
wind turbines and the lack of inexpensive land near major population centers.
Therefore, attention began to be directed towards the off‫ـ‬shore wind turbines [7]. The
offshore wind farms (OWF) flow at higher speeds than onshore ones and the produced
energy by the offshore wind turbine is more than that of onshore ones. This indicates
that a quick and big growth in wind energy systems can be observed over the next
decades, especially in the area of OWF.
Figure 1.2 Wind power generation system topologies [8]
Figure 1.2 shows the different power generation topologies for wind systems. The
power converters are considered as a fundamental element in wind turbine systems. It
4
converts fixed frequency and voltage to a variable frequency and voltage in case of
using an induction generator. This could be performed in many different ways such as
the soft starter, which reduces the transient current during disconnection or connection
of the generator to the grid, the back-to-back pulse width modulation (PWM) for three-
phase voltage source inverter (VSI) which can be a bi‫ــ‬directional power converter,
matrix converter, or tandem converter [9]. The modular multilevel converter (MMC)
is one of the most preferable converter topologies due to its advantages such as
minimum inherent harmonic voltage content and low switching losses. It contains five
different topologies with a common idea which creates a sinusoidal voltage from
several discrete levels of voltages. We will discuss and analyze all these converters in
detail in Chapter 3.
1.3 Objective of the Thesis
The increase in power demand in the world consider an important factor to find
efficient, cost effective and environmentally friendly power generation systems. This
thesis aims to design, analyze and simulate efficient wind energy system. Permanent
magnetic synchronous generator and back to back modular multilevel converter
system has been modeled, designed, and simulated for balanced operation with 3-
phase. The Permanent magnetic synchronous generator feed the network, and back to
back modular multilevel converter have been positioned between the permanent
magnetic synchronous generator and separate generation centers. An ultracapacitor
was added to the model and the results shows the effect of adding this capacitor for
improving the performance of the system.
1.4 Organization of the Thesis
We divided this thesis into 6 Chapters, given as follows:

• Chapter 1: In this Chapter, we made a short-term introduction to the recent
developments of wind turbine energy systems and their importance in power
generation systems. We also introduced some different classifications of
turbines as well as their control systems.
• Chapter 2: In this Chapter, we made a literature review about this topic in the
last few years and offered a brief background by investigating the related
5
publication and researches. Also, we described the research aim and the place
of this work in the literature.
• Chapter 3: This Chapter provides a general review of different types of wind
power generation systems as well as the advantages and disadvantages of these
systems. The second section in this Chapter will discuss the control techniques
and the most appropriate type for the PMSG-based wind power systems.
• Chapter 4: This Chapter explains the steps of design and the simulation
building blocks which were used in the design of the PMSG-based wind
turbine, MMC, and the control scheme.
• Chapter 5: This Chapter presents the results and provides a brief explanation
for these results.
• Chapter 6: This Chapter provides conclusions and suggests some proposals
for future works that can improve what was done in this study.
6
CHAPTER 2
LITERATURE REVIEW
2.1 Introduction
In the last few years, there has been a rapid development of wind based power
generation systems, various technologies, and topologies which have been introduced.
In addition to these progresses, many control technologies were introduced that help
to connect these turbines to the national grid. The most common one of these types is
the PMSG-based wind turbine with back-to-back MMCs. The efforts of a lot of
researchers were applied to model and simulate a PMSG-based wind turbine systems
with energy storage systems. In this Chapter, we will review and briefly summarize
the previous related literature studies in this regard.
2.2 Literature Review
In (2016), Yupu Lu, Da Xie, Chenghong Gu, and Junbo Sun studied the small signal
stability of PMSG-based wind turbines to simplify damping strategy design. They
developed unified small‫ــ‬signal models for different types of PMSG and studied the
small-signal stability of these models. The designed model consists of electrical,
mechanical and control systems. The mechanical system of the model has two‫ــ‬mass
shaft specialized for analyzing the steady‫ــ‬state and dynamic performance of the
generator rotor and the turbine. The researchers modeled the transmission line, PMSG,
and its converter separately in order to build combined models of small‫ــ‬signal PMSG-
based wind turbine system. After that, the relation between state variables and different
oscillation modes were determined by analyzing the state variables based on the
unified small‫ــ‬signal models. All types of oscillation modes in PMSG-based turbines
could be found by the small signal models of PMSG when the PMSG is connected to
the grid. The oscillation modes that could be found by the small signal models are
subsynchronous control interaction, Subsynchronous oscillation, and low‫ـ‬frequency
oscillation including damping and frequency of each oscillation mode. It was shown
7
that both the damping ratios and frequencies of the oscillation methods are influenced
by the initial various values of the small signal models [10].
In (2016), J. Mahseredjian , H. Saad, S. Nguefeu, and S. Dennetière analyzed the

interaction and operation of MMC-HVDC links when they are embedded in an AC
grid. They designed an MMC-HVDC link suitable for small-signal analysis and they
had approved the validity of this model against an electromagnetic transient EMT-type
model for AC systems, which have different short circuit ratio. At the end of the study,
the effect of the AC line connected to the HVDC in parallel was studied by conducting
a modal analysis and parametric studies [11].
In (2015), M. Reza Abedi and, Y. Lee, showed the integration between PMSG and a
smart energy storage system with a long-term energy management. This could be
implemented by utilizing a fuel cell/electrolyzer with the smart energy storage system.
They had used an ultracapacitor in order to make compensation for the fluctuations of
short‫ــ‬term transient energy that is caused by the wind speed variations or by the
changes in the load. The control process in the model enables the integration between
the wind turbine system and the smart energy storage system and determines the smart
energy storage system power that should be generated pursuant to the available energy
in the DC link. In this design, the inverter is immune to fluctuations and is not affected
by load changes and the wind power interruption due to maintain the DC link voltage
as constant as possible. The design was tested in Matlab/Simulink program package
against wind speed variations and sudden changes in load. The results demonstrate the
applicability of approving the control architecture and the overall operation [12].
In (2015), Kun Zhang, Chengxiong Mao and, Jiming Lu proposed an integrated control
strategy based on the features of direct driven PMSG to overcome the technical
problems resulting from increasing wind energy penetration in power systems. In this
design, a hybrid battery/ultracapacitor energy storage system was incorporated with
the PMSG in order to enhance the capability of the low voltage ride‫ـ‬through in the
wind generation system and smooth the output power. In addition, better power and
energy performances were achieved by the battery/ultracapacitor complementary
technologies [13].
8
In (2014), Guoqing Li, Xiangzhen Ren, and House Chen designed a model with
ultracapacitor in order to smooth the output curve of wind power. The researchers
benefit from the property of ultracapacitors to accomplish this issue. The fluctuation
of wind power and state of charge (SOC) level of ultracapacitor are considered in order
to make the transmission more flexible and optimize the time constant in real time.
The PSCAD/EMTDC program was used to simulate this energy storage device, which
is integrated with the wind power generation system. The efficacy of this control
strategy was approved by the simulation results [14].
In (2017), Karaagac U, Mahseredjian J, Cai L, and Saad H. reviewed the problem of

modeling the multi-terminal MMC‫ــ‬HVDC systems, when they are integrated with
Offshore Wind Farms (OWF) depending on power electronics based converters. The
computational performances and modeling accuracy were compared in this article for
different combinations of OWF and MMC models. The researchers performed an
OWF with DFIG which is connected to an AC grid by a multi-terminal MMC‫ــ‬HVDC
system. Then AC fault simulations was applied in case of they applied offshore and
onshore wind farms. The associated overcurrent protection and the detailed
representation of the offshore collector grid were included in this OWF system model.
The MMC controls include fast OWF power generation reduction based fault ride-
through function and offshore fault current limiter [15].
In (2011), Takahashi R, Nakatani M, and Tamura J proposed a new connection scheme

of the energy storage system to wind farm. They did not use a PWM based DC-DC
converter to connect the ESS to the DC bus circuit of the wind generator. They
analyzed the loss characteristics and compared the implemented system and the case
when using a PWM based DC-DC converter. The results showed that the proposed
system can improve efficiency and simplify the overall control scheme significantly.
The research team used the wind speed data, which were obtained practically in
Hokkaido, Japan to study the loss characteristics by simulation cases in the
PSCAD/EMTDC program [16].
In (2018), Gaber El-Saady, and El-Nobi A. Ibrahim studied and performed modeling
operation and simulation of a gearless PMSG-based wind generation system connected
to the grid. They applied a maximum power point tracking method developed by
9
implementing the control strategy of machine-side converter. The injected reactive and
active power to the grid side are controlled by the converter so the DC link voltage
would stay constant. A fully controlled back‫ـ‬to‫ـ‬back converter with a VSI was used to
connect the PMSG with the grid. This back‫ـ‬to‫ـ‬back converter consists of an
intermediate DC link circuit and a PWM scheme. The system under fault conditions
was protected by the overvoltage protection of DC link. This article discussed the
effects, faults, and change in wind speed on the system operation. The overall design
was made by using Matlab/Simulink program package [17].
In (2006), Chinchilla, Mónica, Burgos, Juan Carlos, and Arnaltes Santiago studied the
control and operation principle of the direct driven PMSG. They connected this
generator to the grid by using a full controlled type frequency converter. This
frequency converter was made up of an intermediate DC circuit, PWM rectifier, and a
PWM inverter. The correction and regulation of power factor are done in this model
depending on the vector control of the grid‫ــ‬side inverter. The model was designed to
achieve maximum power under different load conditions from the incident wind. The
complete system was tested and the result showed the dynamic performance of the
system and the benefits of the proposed system were verified [18].
In (2014), Ko, Hee Sang presented a dynamical model of the PMSG wind turbine and
its control scheme. The design contains a full‫ـ‬scale back‫ـ‬to‫ـ‬back MMC with a PMSG
variable speed wind turbine. The control system was used to control the power
converter and the wind turbine dynamically. Characteristics of the wind turbines in
terms of ability to control active and reactive power allowed researchers to control
both of them independently. He added the facility of controlling the reactive power
from remote location so as to contribute and regulate the voltage. Finally, he discussed
the ability of the control scheme by implementing simulations for a site of the OWF
in Jeju, Korea [19].
2.3 Place of Work in the Literature
In this thesis, a PMSG and back-to-back MMCs based wind energy system has been
designed and simulated for achieving balanced operation with three-phase power
system. In this work, an ultracapacitor is also added to the DC link of MMCs in order
10
to investigate any possible potential benefits and improvements especially in the
dynamic states of the wind energy system. Four cases have been proposed and tested
under normal operation at variable speed, phase-A to ground fault, three-phase to
ground fault, and short circuit condition at the DC link. The simulation results show
that the added ultracapacitor to the wind energy system did not make any significant
changes to the voltage and current waveforms for both supply and load sides.
However, reductions of the fluctuations of active and reactive power are observed
during startup when compared with the case without ultracapacitor. This study
investigates the application of ultracapacitors in wind turbine system for power
smoothing. A control algorithm is designed based on the dynamic of each part of the
wind turbine system. The wind turbine system with purposed algorithm is simulated.
The results show the significant improvement of the wind turbine output power. With
the improvement of the ultracapacitor technology, the application of the
ultracapacitors in wind turbine system as an energy storage will be unavoidable and
more than ever.
11
CHAPTER 3
THEORETICAL BACKGROUND
3.1 Overview
In this chapter, we will explain all theoretical aspects of wind power generation
systems and explain all the parts that make up the wind energy conversion system
(WECS). We will explain the most common types of wind power generators as well
as the advantages and disadvantages of each type. We will also show the advantages
that characterize the PMSG from the others, components of the PMSG as well as the
control techniques for this type. The second section will show an overview of energy
storage devices such as ultracapacitor, batteries and the superconducting magnetic
energy storage system. The advantages of adding the ultracapacitor for the design is
explained in this section. Finally, we will review MMC topologies as well as its
semiconductor switching techniques. The main parts of a WECS are shown in Figure
3.1.
Figure 3.1 Main parts of a WECS [6]
12
3.2 Multilevel Converters
Nowadays, medium voltage high power applications need attractive solution to

achieve the increased demand on energy. Multilevel converter is the solution for AC
to DC and DC to AC power conversion [20]. In its simplest form, a negative or a
positive DC link voltage are connected to the output of a two‫ـ‬level voltage source
converter (VSC) by appropriate switching of the power semiconductors. In case of a
three-level VSC, an additional neutral voltage (zero) is added to the output which
reduces harmonic content. Therefore, in two or three-level VSCs, the generated output
voltage waveform is a combination of pulses that contains the desired fundamental
frequency with higher order harmonics which needs appropriate filtering. On the other
hand, multilevel converters generates stair-case voltage waveform with much smaller
harmonic content by realizing many separate DC sources or capacitors. This is the
main advantage of multilevel converters compared to two or three-level ones. The
voltage ratings of semiconductors in multilevel converters are reduced by the
increasing in voltage level step size. Hence, this property results in reducing the rates
of failure in multilevel converters. However, in case of two or three-level converters,
several semiconductors must be connected in series in order to share the voltage. Any
firing delay among one of them may cause failure. In case of multilevel converters,
also the rate of change in voltage (dv/dt) is reduced and hence the problems of
electromagnetic compatibility will be significantly reduced. On the other hand, the
voltage and power rating of a multilevel converter can be upgraded easily by its
modular structure if it is designed by submodules. Moreover, it has lower output
common mode voltage and input current distortion as compared with two-level
converters. Generally, it has lower switching frequency than other types of VSCs for
the same harmonic content thus, the switching losses per device is reduced and the
thermal management of the semiconductors can be handled much easier. Despite the
many advantages mentioned above, there are two main drawbacks of multilevel
converters. The first one is the difficulty of the design of the switching scheme and its
control due to their complexity in the topological structure. The second one is the high
cost which results from the usage of many semiconductors and circuit elements.
13
3.3 The Topologies of Multilevel Converter
3.3.1 Flying Capacitor (Capacitor-Clamped) Multilevel Converter
This type was proposed by Foch and Meynard in 1992. It uses capacitors for clamping
purpose [21]. In this topology, The DC link is subdivided into n-levels by n-1 number
𝑉𝑑𝑐
of capacitors. For example, if there are five levels at the output, these levels are ,
2
𝑉𝑑𝑐 𝑉𝑑𝑐 𝑉𝑑𝑐
− , ,− and 0. The common DC link is shared by each of the phases. The
2 4 4
number of the required auxiliary capacitors is calculated by equation 3.1 and the
number of required DC link capacitors is N-1 for N level converter according to the
following Eq:
Number of aux capacitors =

( n − 1)  ( n − 2) (3.1)
2
The main advantage of this design is that it needs a little number of switches so that
two or more switching combinations can give the required output voltage. It could benefit
from this advantage properly for balancing of capacitor voltage. The disadvantage of this
topology is that it needs complicated control circuit and the capacitor voltage balancing
also becomes more complicated when the number of voltage levels are increased. On other
side, increasing voltage level will increase the number of the required capacitors also and
that will increase the cost and the dimensions of the design [22][23]. The semiconductor
switching states for a five-level flying capacitor converter are shown in Table 3.1. The
output voltage waveform and the circuit diagram for a single-phase five-level flying
capacitor converter are shown in Figure 3.2.
(a) (b)
Figure 3.2 (a) Output voltage waveform of a single-phase five-level flying capacitor
converter (b) The circuit diagram of this converter
14
Table 3.1 The output voltage and switching states for five‫ـ‬level flying capacitor.
converter
Output Voltage S1 S2 S3 S4 S1’ S2’ S3’ S4’
𝑉𝑑𝑐
− ON OFF OFF OFF ON ON ON OFF
2
𝑉𝑑𝑐
− OFF OFF OFF OFF ON ON ON ON
4
0 ON ON OFF OFF ON ON OFF Off
𝑉𝑑𝑐
ON ON ON ON OFF OFF OFF OFF
2
𝑉𝑑𝑐
ON ON ON OFF ON OFF OFF OFF
4
3.3.2 Diode-Clamped (Neutral Point Clamped) Multilevel Converter
This topology was proposed by Akagi and Nabae Takahashi in 1981 [24]. The
researchers published several papers that explain and cover the different levels of
diode‫ـ‬clamped multilevel converter such as four, five and six‫ـ‬level. This topology is
used in several applications such as static VAR compensation, high voltage DC
transmission system interconnections and the drives for controlling the variable speed
motors. Each of the phases shares a common DC link and four capacitors are used to
𝑉𝑑𝑐 𝑉𝑑𝑐 𝑉𝑑𝑐 𝑉𝑑𝑐
subdivide this DC link into five levels, such as, ,− , ,− and 0. The
2 2 4 4
clamping diodes are used to limit the voltage stress across the switching devices so all
𝑉𝑑𝑐
the capacitors are charged to . The different voltage levels should be blocked by
4
3 𝑉𝑑𝑐
the clamping diodes. D2 and D2' block Vdc/2, D1' and D3 for blocking and D1
4
𝑉𝑑𝑐
and D3' for blocking . Figure 3.3 shows the output voltage waveform and the circuit
4
diagram of a single-phase five-level diode‫ـ‬clamped converter. The desired output
voltage of the five‫ـ‬level converter is also shown in Table 3.2. When the diode-clamped
multilevel converter is used as a three‫ـ‬phase converter, it will have 2𝑁 − 1 voltage
levels in the line-to-line voltage, while it will have N voltage levels for the phase-to-
neutral voltage. Thus, five-level single-phase converter which is shown in Figure 3.3
has nine-level in the line‫ـ‬to‫ـ‬line voltage when it is used as three‫ـ‬phase. The
diode‫ـ‬clamped multilevel converter has been designed initially so as to increase the
range of power ratings to megawatt ranges. Nowadays, the diode-clamped multilevel
15
converters are utilized in many applications, especially in the adjustable speed drives
applications with a range from several kilowatts to megawatt. The main disadvantage
of this type of converter is the capacitor voltage balancing problem and the complexity
of the control scheme when the voltage levels increase.
Figure 3.3 (a) Output voltage waveform of a single‫ـ‬phase five-level diode‫ـ‬clamped

converter (b) The circuit diagram of a single‫ـ‬phase five-level diode‫ـ‬clamped
converter
Table 3.2 The output voltage and switching states for five‫ـ‬level diode-clamped
converter
O/P Voltage S1 S2 S3 S4 S1’ S2’ S3’ S4’

𝑉𝑑𝑐
− OFF OFF OFF ON ON ON ON OFF
2
𝑉𝑑𝑐
− OFF OFF OFF OFF On On ON ON
4
0 OFF OFF ON ON ON ON OFF Off
𝑉𝑑𝑐
ON ON ON ON OFF OFF OFF OFF
2
𝑉𝑑𝑐
OFF ON ON ON ON OFF OFF OFF
4
16
3.4 Submodule Based Multilevel Converters
The submodule based multilevel converters are considered as the most interesting area
for the researchers among the different conventional multilevel converter topologies.
It is also called as “Modular Multilevel Converter (MMC)”. When we classify this
type of converters, there are three issues that should be taken into consideration. The
first issue is the connection between the submodules between each other. It is either
delta or star connection. The second issue is related with the usage point (whether it is
used as a single cluster or dual clusters). The third is the submodules topologies. It is
either full-bridge (H-bridge) or half-bridge (chopper cell). Finally, we will get eight
different converter topologies for submodule based multilevel converters. But
practically not all of them are implementable. The possible variations of this topology
are shown in Table 3.3. We mentioned the implementable topology by YES and the
irrelevant topology by NO.
Table 3.3 The possible variations of MMC [25]

Delta Star
Single Double Single Double
Full-bridge YES NO YES YES
Chopper NO NO NO YES
The list below shows four different types of MMC that can be implemented practically.
On the other hand, Figure 3.4 shows SSBC and SDBC topologies for three-phase
operation [26]. Figure 3.5 shows DSCC and DSBC shows SSBC and SDBC topologies
for three-phase operation.
I. Single-Star Bridge Cells (SSBC)

II. Double-Star Chopper Cells (DSCC)
III. Double-Star Bridge Cells (DSBC)
IV. Single-Delta Bridge Cells (SDBC)
17
(a) (b)
Figure3.4 (a) SSBC topology (b) SDBC topology
(b)
(c)
Figure 3.5 (a) DSCC and DSBC topology with coupled arm inductors
(b) Chopper cell (DSCC) (c) Bridge cell (DSBC)
18
We can represent the double‫ـ‬delta structure as two isolated and parallel connected
single‫ـ‬delta circuit, since the common neutral point is not available in delta connection
like in star connection. In case of single-delta or single-star structures, it is not possible
to generate AC voltage at the output because we should use bridge cell with chopper
cells in order to generate AC voltage at the output. As a result, four types of MMCs
are not implementable practically among eight theoretical combinations. These
topologies are single‫ـ‬delta chopper cells, single‫ـ‬star chopper cells, and the bridge &
chopper double delta circuits.
Specifically, these four types can be classified into MMC or “cascaded H-bridges.
Some researchers refer only for the chopper cell-based double star circuits as MMC,
while some other researchers refer to both chopper and bridge-cell circuits as MMC.
We will explain the bridge circuits (DSBC, SDBC and SSBC) as a cascaded H‫ـ‬bridges.
Nonetheless, there are two common features among these topologies. The first feature
is that in all topologies, either being bridge cell or chopper cell, the submodules are
connected in series. The second feature is about the arm inductor which is connected
in series with the submodules of all these topologies. In the double star case, we have
two methods for connecting the arm inductor; the first one is to connect two separate
inductors (one for each arm) or we can connect single inductor for both phase arms.
The weight and size of the total of two uncoupled inductors are more than single
coupled inductor. Next section will discuss the characteristics of MMCs and cascaded
H-bridges.
3.4.1 Cascaded H-Bridges
The cascaded H‫ـ‬bridges can be designed with three different cases. These are SSBC,
SDBC, and DSBC. The common DC link is not available in SSBC and SDBC
topologies as in case of DSBC. So, when we have an application that needs active
power transfer from output to the DC link, or in an inverse direction such as in motor
drive applications, we cannot use either SSBC or SDBC topologies due to need for an
energy source for the DC link. The SSBC and SDBC can be used in this case if a
separated and isolated DC sources are connected in parallel with the bridge capacitors.
On the other side, using the DC sources in this way requires heavy and bulky
transformers which will cause increasing the cost and the weight of the design. The
19
DSBC structure has a neutral point of the star connection which enables the designer
to connect it to a single DC link. The full-bridge circuit generates three different
voltage levels, such as, −𝑉𝑐 , 0 and +𝑉𝑐 where 𝑉𝑐 is the voltage of submodule capacitor.
The topology of the full‫ـ‬bridge circuit is shown in Figure 3.6. The states of the switches
in term of these three levels are clarified in Table 3.4. The output voltage waveform of
cascaded H-bridges is formed by summing the AC output of all the full-bridge
inverters by connecting them in series. The number of output voltage levels can be
increased by two when we add one full-bridge in series. This topology can be used
practically in HVDC and motor drive applications.
Table 3.4 The switching states of full-bridge circuit.

𝑉𝑜 𝑆1 𝑆2 𝑆3 𝑆4
𝑉𝑐 1 1 0 0
−𝑉𝑐 0 0 1 1
0 1 0 1 0
0 0 1 0 1
.
Figure 3.6 The topology of the full-bridge circuit
The battery energy storage systems and STATCOM mainly use SDBC and SSBC
topologies, while DSBC topology is used in motor drive applications such as blowers
and fans. The DSBC is also used in solar and wind power systems where DC link
voltage changes with the weather conditions and wind. Since it has a buck and boost
20
function which gives it the characteristic of tolerating a wide range of variation in the
DC link voltage [26]. By comparing the chopper cells with the bridge cells we see that
the bridge cells have the advantage of doubling the level if output voltage for the same
number of submodules for each phase. The other advantage over chopper cells is that
using separate DC sources in submodules produce balance between voltages of
submodule capacitors and this simplifies the control circuits of the converter. The main
disadvantage of the DSBC topology is the need for a group of external DC sources
which will increase the cost, weight, and the size of the design, since the power
electronic switches and transformer are generally found in DC sources.
3.4.2 Modular Multilevel Converter (MMC)
In 2003, Marquardt and Lesnicar designed modular multilevel converter (MMC)

topology. MMC was investigated by many researchers and became one of the most
important and attractive one among other multilevel converters topologies especially
for high power applications. The typical application areas of MMC are connection of
renewable energy systems with the national grid, STATCOM applications, medium
voltage motor drives and VSC based HVDC transmission systems. Figure 3.7 shows
the typical arrangement for a 400 MW MMC as well as its submodule and converter
arm segment developed by Siemens. Figure 3.8 shows the circuit topology for a three-
phase MMC. In MMC, the converter consists of three-phase legs which share a common
DC link. Each phase leg in the converter has a "lower phase arm" and "upper phase arm".
The phase arm in the phase leg of the converter has an arm inductor,
𝐿𝑎𝑟𝑚 and (N) cascade connected submodules. The function of submodules is to produce
the required AC voltage for each phase. The function of the arm inductor is to filter the
high frequency component in arm current, limit the fault current, and controls the
circulating current. The output AC voltage is taken from the middle point between the two
inductors of the lower and upper arms. Figure 3.9 shows the half-bridge circuit and Figure
3.10 shows the states of half‫ـ‬bridge circuit and current paths. The half-bridge consists
of a submodule capacitor and two power semiconductor switches in cascade. The capacitor
is connected in parallel to the switch. This module is the most common submodule
structure used in MMC topologies due to its simple and efficient structure [42].
21
Figure 3.7 Typical submodule, arm, and converter arrangement for 400 MW MMC
by Siemens [27]
One of the half-bridge submodule switches should be ON at any time. If s1 in Figure 3.8
is in ON state and s2 in OFF state, then the half‫ـ‬bridge circuit is switched on or in
“current path”. If s1 is in OFF state and s2 is in ON state then the submodule is
switched off or “bypassed”. The submodule capacitor voltage is equal to the terminal
voltage of half‫ـ‬bridge circuit. It is zero in case of switching off and
𝑉𝑐 in case of switching ON. When unwitting switching happens and both switches are ON,
then the capacitor of the submodule will be short circuit. The other case of unwitting
switching happens when both switches are OFF, submodule voltage will be undetermined
and dissimilar voltages may appear at the terminals of the submodule according to the
direction of the current. Table 3.5 shows the charge/discharge conditions for the capacitor
and terminal voltage.
22
Figure 3.8 Three-phase MMC circuit topology
Figure 3.9 Half‫ـ‬bridge circuit

. .
23
Figure 3.10 States of half-bridge circuit and current paths
Table 3.5 Switching states of Half-bridge circuit

. .
State number in 𝑺𝟐 𝑺𝟏 𝑽𝒐 Current Current Capacitor

Figure 3.9 direction path Charge
1 1 0 0 𝑖𝑜 > 0 𝑆2 -
2 1 0 0 𝑖𝑜 < 0 𝐷2 -
3 0 1 𝑉𝑐 𝑖𝑜 > 0 𝐷1 Charge
4 0 1 𝑉𝑐 𝑖𝑜 < 0 𝑆1 Discharge
In MMC, the energy is stored in submodule capacitors of the whole converter unlike
the conventional VSC where the energy of the DC link is stored in a big and central
capacitor. The capacitance of the capacitors must be equal as much as possible in order
to eliminate voltage unbalance which produces extra currents inside the converter.
There are several methods to obtain the submodule capacitor voltage balancing. In
steady‫ـ‬state open loop operation, for each phase leg that contains (2N) submodules,
(N) of them are inserted to the current path. According to this explanation, the natural
average voltage of the submodule capacitors, Vc, could be expressed as following
Vdc
Vc = (3.2)
N
24
When the number of inserted submodules in a specific leg change, the imbalance
between the DC source and phase leg will happen. This imbalance will create internal
circulating current which charges and discharges the capacitors of submodules
continuously. So it is necessary to achieve the natural balance point by achieving this
condition in converter’s open loop operation. In some specific closed loop operation,
the submodule number in each phase can be decreased or increased from (N),
temporarily and intentionally.
3.4.3 Advantage and Disadvantages of MCC
The MCC has a lot of advantages and features as compared with the other VSC
topologies and multilevel converters. The most important advantages are listed below:
1. MMC has higher efficiency when compared to other multilevel converter

topologies.
2. The harmonic distortion level is low which will decrease the cost and the size of
the required filters.
3. The modular structure gives high redundancy. The converter keeps working
normally even there is a faulty submodule which will be bypassed.
4. The apparent switching frequency of the converter is high despite the low switching
frequency of the submodules.
5. This topology can be designed by different power and voltage levels with ease
of flexibility for scaling.
6. When the short circuit happens at the DC link, the AC side current stays normal due
to arm inductors.
7. It has high independence in term of centralized circuit element, like separate DC
sources or DC link capacitor as in case of cascaded H‫ـ‬bridges.
The semiconductors of the MMC require low voltage rating because it is limited to the
𝑉𝑑𝑐
voltage of submodule capacitor ( ). On the other hand, MMC has also some
𝑁
disadvantages as compared with the other multilevel converter topologies such as:
1. The voltage drop across arm inductors may cause reactive power losses in high
current applications.
25
2. It needs a high number of power electronic devices compared with the other
topologies.
3. Maintaining the balance of the submodule capacitor voltage is relatively difficult
and needs measurement and computational requirements depending on the used
method.
3.4.4 Other Submodule Topologies
There are some other topologies that are used instead of half-bridge and full-bridge
topologies as submodules for the MMC. These alternative circuits are listed below and
shown in Figure 3.11.
1. Five‫ـ‬level cross‫ـ‬connected circuit,

2. Three‫ـ‬level converter circuits; either diode‫ـ‬clamped or capacitor‫ـ‬clamped,
3. Clamp double circuit.
Figure 3.11 (a) Diode-clamped three-level converter (b) Capacitor-clamped three-

.
level converter (c) Clamp double circuit (d) Five-level cross connected circuit [28]
. .
3.5 An Overview of Energy Storage Devices
Energy storage devices give a lot of benefits when it is used in power systems such as
reliability of supply, power quality, and improving stability. In order to meet the
requirements of practical power systems applications, energy storage technologies are
developed continuously. Renewable energy sources change widely during the day and
they are intermittent that is why the energy storage technologies is so important in the
26
power systems which contain a renewable energy source. If storage technologies of
electricity are integrated and designed properly, this will enable the power plants from
dispatching the unused electricity later and lessen the variability in load. Energy
storage technologies can be classified into three main categories according to their
applications and system requirements. These are real long‫ـ‬term (many hours to days),
long‫ـ‬term (minutes or hours) and short‫ـ‬term (a few seconds or minutes) [29]. Major
. .
energy storage technologies used in power systems are as follows:

• Flywheel
• Ultracapacitor (Supercapacitor)
• Superconducting Magnetic Energy Storage System
. .
• Battery
• Fuel Cell
• Compressed Air Energy Storage
.
• Pumped Hydro Energy Storage
3.5.1 Ultracapacitor
Ultracapacitor or supercapacitor is an energy storing device that stores energy in the

form of electric field and usually consists of electrochemical double layer capacitors.
The stored energy in an ultracapacitor is given by
1
𝑆𝑡𝑜𝑟𝑒𝑑 𝐸𝑛𝑒𝑟𝑔𝑦 = 𝐶𝑉2 (3.3)
2
Where (V) is the voltage across the parallel plates and (C) is the capacitance. The
ultracapacitors are appropriate for short‫ـ‬duration and fast‫ـ‬response applications
. ..
because it has a short-term response device. They are excellent for stabilizing
frequency and voltage of power systems. The ultracapacitors give much higher energy
density as compared with traditional capacitor or batteries [30]. The typical power
.
storage capacity of the ultracapacitors is varying from 1kW to 250kW, while their .
energy storage capacity is usually less than 3MWh. The main characteristics of
ultracapacitors can be summarized as below [31]:
• Energy: < 3 MWh

• Lifetime:  30-40 years
• Power: 1.0-250 kW
27
• Price:  85-480 $/kWh
• Charging-Discharging Efficiency:  95%
3.5.2 Why Are Ultracapacitors Used in WECS
Battery-based and ultracapacitor-based storage systems are considered as the main

types of storage systems that can be used with the pitch control. Although the batteries
have good efficiency in different applications, the ultracapacitors are more efficient
with the appropriate control system of the WECS. The pitch control system adjusts the
pitch of the blade in line with the operation strategy of the wind turbines and this leads
to prevent the mechanical stress on the turbine. This kind of operation gives constant
power output, extends wind turbine service life, and reduces the mechanical stress on
the turbine. The energy storage system also provides emergency backup plan in case
of grid general failure where it helps the blades to return to its neutral position for safe
shutdown. The return of turbine blades to their natural position in case of shutdown
prevents the total damage of turbine or the blades because of uneven or strong wind
force. By making a brief comparison between the different storage techniques we see
that ultracapacitor technology has a lot of advantages as compared with the other
storage technologies such as:
1. Although the cost of installing battery-based storage systems is approximately equal

to the cost of installing storage systems based on capacitor technology, the
ultracapacitor needs simpler and cheaper monitoring and charging systems than the
battery-based pitch control system. Also, it needs less equipment for vibration
damping and fewer components for mechanical mounting. As a result, the total cost of
the design will be less than the other storage technologies.
2. The ultracapacitor-based storage system has a power density higher than the battery-
based one, so it has the ability to provide a burst power to return the blades to its off-
condition position in case of the national grid failure. On the other hand, it has high
reliability for short‫ـ‬term mismatches between power availability and power demand.
3. The batteries require cooling and heating systems because they are vulnerable to the
changes of weather. This situation will add additional cost, hence the overall design
28
cost will be higher. On the other hand, ultracapacitor-based systems can afford the
different weather conditions so it will eliminate the extra cost of cooling system.
4. The ultracapacitor can work under different weather conditions and temperatures
ranging from -40°C to 65°C unlike batteries which work under a narrow operating
temperature range.
5. The ultracapacitor has long term life cycle from 500000 up to 1000000
charges/discharge cycles. This property gives the ultracapacitor predictable aging on
average and long service life. On the other hand, the batteries have relatively short
term life cycle. This issue makes the replacement of batteries necessary every three or
four years.
6. In emergency shutdown cases, the energy storage system should be able to provide
the maximum required power demand even if it is required for several second so if we
use battery-based system, we will need a big size and heavy weight groups of batteries
to achieve this demand, while ultracapacitor can provide a burst power during peak
power demand without the need for oversizing efforts.
3.6 MMC Switching Methods

. .
MMC switching techniques can be classified into two main parts, such as, high
. . .
frequency switching and low frequency switching. The waveform of high frequency
carrier is compared in magnitude with the waveform of the reference voltage
waveform which is sinusoidal based in high frequency switching and the switching
logic signal is generated according to this comparison. On the other side, a pre-
calculated pulse pattern is used to drive the converter in low frequency switching. In
order to improve the performance of the converter, harmonic elimination methods can
be used in the system. This method has high efficiency due to its low switching count.
In general, these two methods produce low harmonic distortion at the output voltage
and current waveform and this is a big advantage because the design will not require
a big and expensive passive filter to solve the harmonic distortion problem. Moreover,
it will have a fast dynamic response against source or load changes. Theoretically, the
high frequency switching techniques have a drawback of high switching losses due to
29
the high switching count but with a suitable design the losses could be decreased and
this drawback could be overcome. Recently new switching technique has been
proposed, called as “mixed switching”. As the name refers, this switching technique
takes the advantages of the two methods mentioned above. It has the properties of both
high and low frequency switching. Figure 3.12 shows a schematic diagram for MMC
switching methods.
Figure 3.12 MMC switching methods

. .
3.6.1 High Frequency Switching (Carrier Based Switching)

. .
Due to their ease of implementation and satisfactory performance, high frequency

switching methods are considered as the dominant type for MMC. The frequency of
the carrier is constant in high frequency switching during all switching periods. The
output voltage and the reference mean values are adjusted to be equal, that’s why the
space vector method or scalar method can be used for this method. In the scalar
switching techniques, a reference (modulation) waveform is used. This waveform has
a high frequency carrier waveform. The switching occurs by comparing the
magnitudes of the carrier and the reference waveforms and the switching occurs at the
crossover points. In space vector method, each switching condition of the converter
needs a vector diagram and the output reference is represented by a vector. The
30
converter is driven by the related switching pulse pattern in order to get the desired
output voltage. This switching pulse pattern is generated according to the switching
condition vectors and the reference vector. Finally, the design and execution of the scalar
method are much easier than the space vector method though both methods can give
similar switching pulse pattern.
3.6.1.1 Scalar PWM
3.6.1.1.1 Sub‫ـ‬Harmonic )Level-shift) Methods
𝑁 identical triangular carriers are used in these methods in order to displace the whole
DC link voltage contiguously. In these methods, the circuit elements which are
responsible for creating different voltage levels get a balanced exploitation by making
the carriers peak‫ـ‬to‫ـ‬peak amplitudes equal to each other. The different in the phase‫ـ‬shift
of carriers in relation with each other will produce three different sub‫ـ‬methods. These
methods are:
• Phase opposition disposition switching technique
• Phase disposition switching technique
• Alternative phase opposition disposition switching technique
• Phase Disposition Switching Technique

All the carriers should be in‫ـ‬phase in this switching technique. In this method. If the
MMC has 4 sub‫ـ‬modules per phase arm (N=4), the carriers will be displaced in the
𝑉𝑑𝑐 band. Figure 3.13 illustrates this method. We should note that all the figures from
3.13 to 3.17 we have 1 pu in the vertical axis corresponds to 𝑉𝑑𝑐 .
Figure 3.13 The carriers of phase disposition switching technique

31
• Phase Opposition Disposition Switching Technique
In this technique, the carriers which are located below and above the zero axis are 180°
out-of-phase. Figure 3.14 shows the switching waveforms for an MMC with 4
submodules per phase arm (N=4). Note that the carriers are displaced in the 𝑉𝑑𝑐 band.
Figure 3.14 The carriers of phase opposition disposition switching technique
• Alternative Phase Opposition Disposition Switching Technique

In this method, the contiguous carriers are 180° out of phase. Figure 3.15 shows the
. .
switching waveforms for an MMC with 4 submodules per phase arm (N=4). Note that
.
the carriers are displaced in the 𝑉𝑑𝑐 band.
Figure 3.15 The carriers of alternative phase opposition disposition switching

technique
32
3.6.1.1.2 Phase‫ـ‬Shift Method
360°
In this method, a displacement by should be achieved between all the identical
𝑁
.
triangular carriers (N) in the whole DC link voltage. The frequency of the carrier is
. . .
𝐹𝑐 and its voltage is the peak-to-peak amplitude of 𝑉𝑑𝑐 . Figure 3.16 shows the
switching waveforms for an MMC with 4 submodules per phase arm (N=4). Note that
the carriers are displaced in the 𝑉𝑑𝑐 band.
Figure 3.16 The carriers of phase‫ـ‬shifted switching technique
In phase-shifted method, the sawtooth waveform can also be used instead of triangular
waveform as the carrier waveform. It is called "sawtooth rotation". Figure 3.17 shows
.
the switching waveforms for an MMC with 4 submodules per phase arm (N=4). Note
that the carriers are displaced in the Vdc band.
Figure 3.17 The carriers of sawtooth rotation method
33
3.6.1.2 Space Vector PWM
Space vector PWM is one of the high frequency switching methods. It is used only
with three-phase power systems unlike scalar PWM methods which could be used for
.
both single and three‫ـ‬phase systems. It is used with a three-phase system because three
.
reference modulation signals will be transformed into a reference vector in this

technique. In space vector PWM, the vector is positi ned in a space vector diagram.
o
Figure 3.18 shows the sample space vector diagrams for five‫ـ‬level and two‫ـ‬level
.
converters. The nearest three triangle corners (switching states of the converter) are
identified according to the magnitude and angle of the reference vector. The angle and
magnitude are depending directly on the three reference modulation signals. In these
hexagon formed diagrams, each corner of the triangles belong to a switching state of
the converter. The converter is switched according to switching states. Three adjacent
vectors are given below to synthesize the reference vector.
(T j v j + T j +1v j +1 + T j + 2v j + 2 (3.4)
Vref =
T
T = Tj + Tj +1 + Tj +2
T : total switching peri d o
T j , T j +1 , T j + 2 : the calculated switching periods of adjacent vectors

. .
Space vector diagram can be used with any type of multilevel converters so it is
.
considered as the universal diagram. It has several advantages such as good utilization .
of DC link voltage creates a low current ripple and easy hardware implementation by
. . . .
a DSP processor. On the other hand, when the number of voltage levels increases, the .
complexity of choosing states increases due to the increase in the redundancy of

switching states.
34
Figure 3.18 Space vector diagrams of two and five-level converters
3.6.2 Low Frequency Switching
3.6.2.1 Selective Harmonic Elimination

. . .
The working principle of this method is depending on the harmonic elimination theory
[32]. In order to cancel the predominant low frequency harmonics in the output voltage
. . .
and set the desired amplitude of fundamental component, the switching angles are pre-
. . . .
calculated. Figure 3.19 shows the waveform of output voltage for five‫ـ‬level converter
where selective harmonic elimination method determines the switching angles.
. . . .. . .
Figure 3.19 Waveform of output voltage for five‫ـ‬level converter with selective
harmonic elimination method [33]
35
Chapter 4
MODELING OF THE WECS
4.1 Introduction
In this chapter, the PMSG-based WECS equipped with back-to-back MMC and the
ultracapacitor is proposed. The control strategy of the MMC has also been proposed.
The MMC consists of three legs, one for each phase and each leg has an upper and a
lower arm. The variable speed wind turbine has a horizontal axis that directly drives
the PMSG rotor via a gearless drive train. The field excitation is obtained to the
generator via permanent magnets which is placed on the rotor of the generator [34].
4.2 Proposed System Modeling
The proposed system structure consists of two main parts; a wind turbine with PMSG
and back-to-back MMCs, which are illustrated in Figure 4.1.
Figure 4.1 PMSG-based WECS based on back-to-back MMCs
In the design, the MMC has six-arms; three arms for upper (positive) side and three
arms for lower (negative) side. Each arm consists of a submodule (cell) which is the
basic construction block of the MMC. The cell is a simple half-bridge circuit with two
IGBT switches and a capacitor. The cell capacitor can be charged or discharged based
on the arm current direction. The number of the voltage levels is depending upon the
36
number of cells in each arm. In addition, an RL filter is used in each phase in order to
reduce the voltage ripple. Figure 4.2 summarizes the overall design of the WECS
which is proposed in this work. The parts of the design are explained briefly below.
Figure 4.2 The overall design of WECS with back-to-back MMCs and ultracapacitor
37
4.2.1 The PMSG
The PMSG is the permanent magnet machine combined with sinusoidally distributed
windings. When ignoring the damper windings, the 𝑑𝑞 reference frame expression is
as follows [35]:
1 𝑑𝐼𝑑
𝑚𝑑 𝑉𝐷𝐶 = 𝑅𝐼𝑑 + (𝐿𝑀𝑑 + 𝐿𝐿𝑆 ) − 𝜔𝑟 (𝐿𝑀𝑞 + 𝐿𝐿𝑆 )𝐼𝑞 (4.1)
2 𝑑𝑡
1 𝑑𝐼𝑞
𝑚𝑞 𝑉𝐷𝐶 = 𝑅𝐼𝑞 + (𝐿𝑀𝑞 + 𝐿𝐿𝑆 ) − 𝜔𝑟 (𝐿𝑀𝑑 + 𝐿𝐿𝑆 )𝐼𝑑 + 𝜔𝑟 𝜓𝑓 (4.2)
2 𝑑𝑡
𝑑𝜔𝑟 𝑃
= (𝑇 − 1.5 ((𝐿𝑀𝑑 − 𝐿𝑀𝑞 ) 𝐼𝑑 + 𝜓𝑓 ) 𝐼𝑞 ) (4.3)
𝑑𝑡 2𝐽 𝑊𝑇
where 𝑑 is the direct-axis component, 𝑞 is the quadrature-axis component, 𝑚𝑑 𝑎𝑛𝑑 𝑚𝑞

are the d and q axis modulation indices of machine side converter, 𝑉𝐷𝐶 is the DC link
voltage, 𝐿𝑀𝑑 and 𝐿𝑀𝑞 are the mutual inductances between PMSG coils, 𝑅 is the stator
resistance 𝐿𝐿𝑆 is the stator leakage inductance, 𝜓𝑓 is the permanent magnet flux, 𝐼 is
the stator current 𝑃 is the poles number, 𝜔𝑟 is the electrical speed of the rotor, 𝐽 is the
moment of inertia of the rotor and turbine, and 𝑇𝑊𝑇 is the mechanical torque of the
turbine. The developed torque 𝑇𝑒 has two factors: The first factor is the component
because of field flux, and the other factor is because of reluctance torque. The
developed torque expression is:
𝑇𝑒 = 1.5𝑃(𝜓̂𝑓 𝐼𝑞 + (𝐿𝑀𝑑 − 𝐿𝑀𝑞 )𝐼𝑑 𝐼𝑞 ) (4.4)
where 𝜓̂𝑓 = √2𝜓𝑓 is the peak permanent flux.
In case of the permanent magnet machine 𝐿𝑀𝑞 > 𝐿𝑀𝑑 . This means that 𝐼𝑑 will reduce
the total electrical torque. However, if saliency is neglected, the 𝑇𝑒 expression in (4.4)
is reduced to:
𝑇𝑒 = 1.5𝑃(𝜓̂𝑓 𝐼𝑞 ) (4.5)
38
Wind Turbine
The mechanical power that is gained from the wind turbine is computed by the
following equation [36],
1 (4.6)
𝑃𝑚 = 𝜌𝐴𝐶𝑝 (𝜆, 𝛽)𝑉𝑤3
2
where 𝜌 is the density of the air, 𝐴 is the area swept by the blades, 𝐶𝑝 (𝜆) is the power
coefficient, 𝑉𝑤 is the wind speed, λ is the tip speed ratio, and β is the blade pitch angle.
The power coefficient 𝐶𝑝 (𝜆) can be computed by using the following equation [37],
𝐶2 𝑥
𝐶
− 6
(𝜆,
𝐶𝑝 𝜃) = 𝐶1 ( )
− 𝐶3 𝜃 − 𝐶4 𝜃 − 𝐶5 𝑒 𝛽 (4.7)
𝛽
where 𝐶1 to 𝐶6 and 𝑥 are constants. 𝛽 can be calculated by [36],
1 1 0.03 (4.8)
= −
𝛽 𝜆 + 0.8𝜃 1 + 𝜃 3
where. 𝛽 is the blade pitch angle between the rotation plane and the cross-section of
the blade and 𝜃 is the azimuth angle. The tip speed 𝜆 can be calculated by the following
equation [38],
𝑅𝑤𝑚 (4.9)
𝜆=
𝑉𝑤
where 𝑅 is the blade’s radius and 𝑤𝑚 is the angular velocity of the generator rotor.
4.2.2 MMC
The proposed WECS is based on two MMCs which are connected as back-to-back.
The MMC in the generator side converts the AC voltage output of the PMSG to DC
voltage, while the MMC in the grid side converts the DC voltage to a three-phase
voltage back again. Figure 4.3 shows the overall structure of the proposed MMC. One
arm structure of the MMC can be seen as shown in Figure 4.4.
39
Figure 4.3 The overall structure of the proposed MMC
40
Figure 4.4 One arm structure of the MCC
The voltage of the arm 𝑉𝑖𝑗 can be calculated with considering the status of the cell
switches 𝑆𝑖𝑗𝑘 , as follows [39]:
𝑉𝑖𝑗 = ∑ 𝑆𝑖𝑗𝑘 𝑉𝑐𝑖𝑗𝑘 (4.10)
where i = u (upper) and ℓ (lower) arm, j = a, b, c (the 3phase); and the k = 1, 2, ..., n
denotes the submodule. The currents of the arm can be calculated by :
𝐼𝑗 𝐼𝑑𝑐
𝐼𝑢𝑗 = + + 𝐼𝑐𝑖𝑟𝑐𝑗 (4.11)
2 3
41
𝐼𝑗 𝐼𝑑𝑐
𝐼𝑙𝑗 = − + + 𝐼𝑐𝑖𝑟𝑐𝑗 (4.12)
2 3
where 𝐼𝑐𝑖𝑟𝑐𝑗 is the circulating current for each phase and:
(4.13)
𝐼𝑐𝑖𝑟𝑐𝑎 + 𝐼𝑐𝑖𝑟𝑐𝑏 + 𝐼𝑐𝑖𝑟𝑐𝑐 = 0
These circulating currents influence submodule capacitor voltages [40]. At last, the
voltages of AC and DC side of the MMC can be computed by
𝑉𝑑𝑐 𝑑𝐼𝑙𝑗 (4.14)

𝑉𝑗 = − 𝑉𝑙𝑗 − 𝐿𝑎𝑟𝑚
2 𝑑𝑡
𝑑𝐼𝑢𝑗 𝑑𝐼𝑙𝑗
𝑉𝑑𝑐 = 𝑉𝑢𝑗 + 𝑉𝑙𝑗 + 𝐿𝑎𝑟𝑚 ( + ) (4.15)
𝑑𝑡 𝑑𝑡
4.3 The Control Scheme of the Proposed Topology
The control scheme of the proposed WECS can be divided into two parts: The PMSG
(machine) side MMC control and the grid side MMC control.
4.3.1 Generator Side Model
The proposed WECS is a variable speed one which is based upon a PMSG. Because
of the relatively low speed of the generator, the rotor shaft is directly coupled to the
generator, so the gearbox is not needed. The power generating part of the WECS is
composed of wind turbine, PMSG, pitch angle controller, and drive train. Figure 4.5
illustrates the generator side model [41].
42
Figure 4.5 Generator side model
4.3.1.1 Wind Turbine
The wind turbine is the first element of the proposed WECS. The main input to the
wind turbine is the wind speed. In this work, we have used the model of a variable
pitch wind turbine from the simulation software library. The wind turbine gets three
variables; wind speed, pitch angle, and generator speed. The output of the wind turbine
to the drive train is the torque (𝑇𝑚 ). The parameters of the wind turbine are given in
Table 4.1.
Table 4.1 Wind Turbine parameters
Name of Parameter Value and Unit

Nominal Mechanical Output Power 8.5 kW
Base Power of the PMSG 8.5 /0.9 kVA
Base Wind Speed 12 m/s
Maximum Power at Base Wind 0.8 pu of nominal mechanical
Speed power
Base Rotational Speed 1.2 pu of base generator speed
43
4.3.1.2 Drive Train
The output torque information from the wind turbine is sent to drive train. Drive train
represents the inertia of the wind turbine and the generator shaft. The simulation
software model of the drive train is shown in Figure 4.6.
Figure 4.6 Drive train
4.3.1.3 PMSG
The output torque information from the drive train is given as input to the PMSG. The
rotor speed is given to the pitch angle controller and the three-phase output is
connected to the PMSG side MMC. he Permanent Magnet Synchronous Machine
block operates in either generator or motor mode. The mode of operation is dictated
by the sign of the mechanical torque (positive for motor mode, negative for generator
mode). The electrical and mechanical parts of the machine are each represented by a
second-order state-space model.
The sinusoidal model assumes that the flux established by the permanent magnets in
the stator is sinusoidal, which implies that the electromotive forces are sinusoidal.
The trapezoidal model assumes that the winding distribution and flux established by
the permanent magnets produce three trapezoidal back EMF waveforms.
The block implements the following equations.
These equations are expressed in the rotor reference frame (qd frame). All quantities
in the rotor reference frame are referred to the stator.
d 1 R Lq
id = vd − id + Pmiq (4.16)
dt Ld Ld Ld
44
d 1 R L  pm
iq = vq − iq + d Pmid − (4.16)
dt Lq Lq Lq Lq
Te = 1.5 p[iq + ( Ld − Lq )id i ] (4.17)
Where
𝐿𝑞 , 𝐿𝑑 q and d axis inductances
𝑅 Resistance of the stator windings
𝑖𝑞 𝑖𝑑 q and d axis currents
𝑣𝑞 𝑣𝑑 q and d axis voltages
𝜔𝑚 Angular velocity of the rotor
𝜆 Amplitude of the flux induced by the permanent magnets of the rotor in the
. stator phases
P Number of pole pairs
T Electromagnetic torque
The 𝐿𝑞 𝑎𝑛𝑑 𝐿𝑑 inductances represent the relation between the phase inductance and
the rotor position due to the saliency of the rotor. For example, the inductance
measured between phase a and b (phase c is left open) is given by:

Lab = Ld + Lq + ( Lq − Ld ) cos(2e + ) (4.18)
3
Where 𝜃𝑒 represents the electrical angle.
For a round rotor, there is no variation in the phase inductance.
Lab
Ld = Lq = (4.19)
2
For a salient round rotor, the dq inductances are given by:
max( Lab )
Ld = (4.20)
2
And
min( Lab ) (4.21)
Lq =
2
45
4.3.1.4 Pitch Angle Controller
The pitch angle controller model is shown in Figure 4.7. The pitch angle controller has
been employed in the. The pitch angle controller will be only active in case when the
rotor speed is twice the rated value.
Figure 4.7 Pitch angle controller
4.3.2 MMC Controller
The PMSG output is characterized by varying frequency and voltage magnitude. For
the grid synchronization purpose, the PMSG output is interfaced with the grid through
a power electronic converter that converts the AC voltage to DC voltage for control
purposes. Then it inverts these DC power to AC power with 50 Hz by the another
MMC. The voltage/current controller shown in Figure 4.8 provides the voltage
compensation. It will generate the voltage compensation which is synthesized by
PWM controller, then it inserted in series with the supplier voltage in order to make
the voltage sinusoidal in shape and stay balanced. The supply voltages 𝑉𝑆𝑎𝑏𝑐 are
converted to dq0 coordinates as follows:
2 2𝜋 2𝜋
𝑉𝑑 = [𝑉𝑎 sin(𝑤𝑡) + 𝑉𝑏 sin (𝑤𝑡 − ) + 𝑉𝑐 sin (𝑤𝑡 + )] 𝑉𝑞 (4.22)
3 3 3
2 2𝜋 2𝜋
= [𝑉𝑎 cos(𝑤𝑡) + 𝑉𝑏 cos (𝑤𝑡 ) + 𝑉𝑐 cos (𝑤𝑡 + )] (423)
3 3 3
1 (4.24)
= (𝑉𝑎 + 𝑉𝑏 + 𝑉𝑐 ))
3
The d-axis voltage consists of oscillating and average components of the source
voltage. The average voltage has been calculated by using the second order low pass
filter. The reference voltages at the load side have been calculated. The switching
signals have been assessed by comparing the load voltages with the reference voltages
46
and with the sinusoidal signal of the PWM controller. Then dq0 are transformed into
𝑉𝑆𝑎𝑏𝑐 coordinates as follows:
(4.25)
𝑉𝑎 = [𝑉𝑑 sin(𝑤𝑡) + 𝑉𝑞 cos(𝑤𝑡) + 𝑉0 ]
2𝜋 2𝜋
𝑉𝑏 = [𝑉𝑑 sin (𝑤𝑡 − ) + 𝑉𝑞 cos (𝑤𝑡 − ) + 𝑉0 ] (4.26)
3 3
2𝜋 2𝜋
𝑉𝑐 = [𝑉𝑑 sin (𝑤𝑡 + ) + 𝑉𝑞 cos (𝑤𝑡 + ) + 𝑉0 ] (4.27)
3 3
These three-phase load reference voltages that are produced will be compared with
load line voltages. Then the errors will be processed by the sinusoidal PWM controller
in order to generate the switching signals required for IGBT switches. Figure 4.8
shows the Simulink model of MMC control circuit.
Figure 4.8 MMC control
47
4.3.3 Mathematical Model of the MMC
As shown in Figure 4.9, the three-phase MMC consists of a number of 2N series

connected switching cells or SMs that form a phase-leg [43].
Figure 4.9 Circuit structure of MMC [43].
In the phase leg, SMs are sorted in lower and upper arms with either single coupled
inductor or one inductor per arm. The SM switches are controlled in a complimentary
method to ensure that at any immediate, the SM is either bypassed or connected to the
arm. Every one of the arms in the converter works as a controlled voltage source, which
has a variable voltage ranging from zero to Vdc. The arm voltage is determined via the
switching case of the SM of arm, which can be calculated by Equation 4.28:
n
dirm
Varm =  S sm *Vc , sm + Lrm (4.28)
i =1 dt
𝑖: is the upper arm current.
𝑖𝑙𝑑: the current cost of half an output (ac side 𝐼𝑟𝑐: circulating current.
48
The current pass through arm consisted from two parts, circulating current and the load
current throughout the phase-legs in the converter, as shown in Equations (4.29) and
(4.30).
ild (4.29)
ip = + I rc
2
ild (4.30)
io = − I rc
2
This is an equation voltage upper arm, lower arm beside the circulating voltage and
evaluated: The voltages aggregate in the arm should be egalitarian to DC link voltage
also the voltage should be regularly distributed between SMs of the lower and upper
arms in addition to the phase-leg. Additionally, when SM is linked inside the arm, the
load current pass through from the capacitor produce a ripple in the SM voltage. The
active power (Pac) and reactive power (Qrc) of three phase system able calculated
utilizing equation (4.31) and (4.32) as follows:
i p = io (4.31)
I rc =
2
ild = i p − io (4.32)
49
CHAPTER 5
SIMULATION RESULTS
5.1 Introduction
This chapter provide the simulation results validating the usage of PMSG and back-
to-back MMC based WTGS. The simulation model designed to test the proposed
control strategy for PMSG with back-to-back MMC. In addition, we have investigated
the improvement in PMSG MMC system when added ultracapacitor to DC-Link of the
back-to-back MMC and test the system stability when subject to different types of
fault.
5.2 Simulation Software and Model Parameters
The model is designed and simulated by used MATLAB Simulink. The tests are five
cases study; The first case is to test PMSG generator connected directly to power grid.
The second case is to test the PMSG generator when connected to power grid via back-
to-back MMC. The third case study is to test the improvement in system when added
ultracapacitor to DC-Link. The fourth, fifth and sixth case study is to test system when
subjected to fault which are: 3 Phase to ground short circuit, Phase A to ground short
circuit and DC Link short circuit.
5.2.1 Simulation Parameters
The main parameters of the system that are used in simulation environment are shown
in Table 5.1, Table 5.2 and Table 5.3
50
Table 5.1 Simulation parameters for PMSG
Parameter Name Value

Number of wind turbines 1
Rated line-to-line voltage 1150 Volt
Rated power 1.5 MVA
Rated frequency 60 Hz
Number of pole pairs 48
Direct and quadrature components of the stator
0.3 mH
inductances (𝐿𝑑) & (𝐿𝑞)
The wind speeds used in the simulations 5,12, 7, 10, 15, 5 and 12 m/s
Table 5.2 Simulation parameters for MMC

Nominal voltage 575 Volt
Nominal system frequency 60 Hz
Number of power modules per arm 12
DC link nominal voltage 1150 Volt
The carrier frequency for PWM 600 Hz
The capacitance of the capacitor in each cell 0.5 mF
Table 5.3 Simulation parameters for ultracapacitor

Rated capacitance 99.5 F
Rated voltage 48 Volt
Number of series capacitors 23
5.3 PMSG Connected to Power Grid Tests
In this case, the output of the PMSG has been connected to grid directly via
transformer. The wind turbine of PMSG is fed by variable wind speed that changes in
every 5 seconds. These instants are 5, 12, 7, 10, 15, 5 and 12 m/s. Figure 5.1 shows
the PMSG generator model that is connected to grid directly.
51
Figure 5.1. PMSG simulation model that connected with grid under variable wind
speed conditions.
The test results are in follows:
5.3.1 Rotor Speed, Mechanical torque and the Active/Reactive power of PMSG
This part simulates the variation of IG speed with wind speed. The simulation results
for different wind speed are shown in Figure 5.2.
52
Figure 5.2 Simulation results of PMSG under variable wind speed conditions. From
top to bottom: wind speed, rotor speed, mechanical torque, active power, reactive
power
53
5.3.2 PMSG/Grid Voltage, Current and VI power
Figure 5.3 Simulation results of PMSG/Grid under variable wind speed conditions.
From top to bottom: voltage, current, apparent power
Figure 5.4 Simulation results of PMSG for a period of 9.85s-10s: From top to
bottom: voltage, current, apparent power
54
From Figures 5.2, 5.3 and 5.4, the wind speed start from 5 m/s and it takes about 4.5 s
to be in steady-state. The rotor speed is about 4.17 rad/s and the mechanical torque is
about 24.2 KNm. The active power is about 0.1 MW and reactive power is about 0.7
MVAR. The voltage is about 370 V and the current is about 1280 A. The apparent
power is about 0.545 MVA. After 5 s, the wind speed is raised to 12m/s and as a result
the rotor speed is increased to about 6.32 rad/s at 6.5 s. The mechanical torque is raised
to reach 160 KNm and the active power is raised to 0.86 MW. The voltage and current
reach to 770 V and 2900 A, respectively. The apparent power is about 0.81 MVA.
After 10 s, the wind speed is dropped down to 7m/s and as a result the rotor speed is
dropped to 4.95 rad/s at 13 s. The mechanical torque is reduced to 51.1 KNm and the
active power is raised to 0.23 MW. The voltage and current are dropped to 500V and
1510A, respectively. The apparent power is about 0.69 MVA. After 15 s, the wind
speed is raised up to 10m/s and as a result the rotor speed is raised up to reach 5.82
rad/s at 17 s. Thus, the mechanical torque is raised to reach 110.5 KNm and the active
power is raised to 0.5 MW. The voltage and current are raised up to 660 V and 2325
A, respectively. The apparent power is about 0.705 MVA. After 20 s, the wind speed
is raised up to 15m/s and as a result the rotor speed is raised up to reach about 6.74
rad/s at 21s. The mechanical torque is raised to reach 251 KN m and the active power
is raised to 1.48 MW. The voltage and current are raised up to 930 V and 232 5A,
respectively. The apparent power is about 0.965 MVA. After 25 s, the wind speed is
reduced again to 5m/s and hence the rotor speed is dropped down to 4.1rad/s at 29.5s.
The mechanical torque is also reduced to reach 24.25 KNm. The active power drop to
0.1 MW. The voltage and current are raised up to 369 V and 998A, respectively.
Finally, the apparent power is about 0.9 MVA. As seen from the simulated results, the
model is effectively changed with wind speed. However, the voltage and current
waveforms shown in Figure 5.4 are uniform sinusoidal waveforms with no ripple.
From these results it can be inferred that we can use the PMSG model in order to
further simulate PMSG-MMC system.
5.4 PMSG Connected to Power Grid via back-to-back MMCs Tests
In this part, we have connected PMSG to back-to-back MMCs in which the PMSG is
connected to AC side of one MMC, while the AC side of the other MMC is connected
to the grid. We have also tested the system dynamics with variable wind speed as in
section A. Figure 5.5 shows the overall PMSG-MMC model. From Figure 5.5, the
55
wind turbine block model has three inputs; one is wind speed that is changed in every
5 s, the second parameter is the generator speed in pu, the third value is the pitch angle
and we assume its value to be zero. Each MMC has 12 cell submodules in each arm
and there are two VSC controllers in both sides. Phase shift PWM is used to control
switching operation of half bridge IGBTs. In addition, there is an RL filter in grid side
of MMC to smooth the signal and reduce the ripples of current/voltage waveforms.
The grid is modelled using 3-phase 1150 V ideal voltage source. A 3-phase load is also
connected in parallel with the grid phases.
Figure 5.5 PMSG-MMC system model
5.4.1 Rotor Speed, Mechanical torque and the Active/Reactive power of PMSG
This part simulates the variation of IG speed with wind speed. The Simulation results
for different wind speed are in Figure 5.6.
56
Figure 5.6 Simulation Result for PMSG, from up to down: Wind Speed (m/sec),
Rotor Speed (Rad/s), Torque Mechanical (Nm ), Active Power (MW) and Reactive
Power (MVAR)
The simulation results are as follows. As shown in Figure 5.6, the wind speed starts
from 5m/s, then increases in each 3 s. The wind speed is varied as follows: 5, 12, 7,
10, 15, and 5 m/s. The rotor speed and torque are varied based on wind speed changes.
These parameters increase as wind speed increases and decrease as wind speed
decreases.
57
5.4.2 PMSG/Grid Voltage, Current and VI power
For testing the efficiency of MMC controller, we have examined the voltage, current,
individual capacitor voltage in each cell, and DC link voltage of back-to-back MMCs.
Figure 5.7 and 5.8 show the 3-phase voltage and current waveforms as well as apparent
power for PMSG side and grid side under variable wind speed conditions for a period
of 5.94s-6.14s.
Figure 5.7 Simulation results of MMCs under variable wind speed conditions. From
top to bottom: voltage, current, apparent power. PMSG side (left) and grid side
(right).
58
Figure 5.8 Simulation results of MMCs under variable wind speed conditions for a
period of 5.94s-6.14s. From top to bottom: voltage, current, apparent power. PMSG
side (left) and grid side (right).
From the observations of Figures 5.7 and 5.8, it is clear that the waveforms of voltage
and current are sinusoidal almost with no harmonics. The apparent power of both
MMCs is observed to be about 1 MW in steady-state. This value is stable in both sides.
These results prove that the VSC controller operates effectively.
59
5.4.3 Capacitor Balance and DC Link Voltage and Current
The capacitor voltages of each submodule in phase A of both MMCs are shown in
Figure 5.9. Whereas, the capacitor voltages of each submodule in one arm of both
MMCs for a period of 17.3s-18 s are shown in Figure 5.10.
Figure 5.9 Capacitor voltages in each submodule in one arm of both MMCs. From
top to bottom: PMSG side (supply), (b) grid side (load).
Figure 5.10 Capacitor voltages in each submodule in one arm of both MMCs for a
period of 17.3s-18s. From top to bottom: PMSG side (supply), (b) grid side (load).
60
As observed from Figure 5.9 and 5.10, it is visually approved that each submodule
voltage in both MMCs is close to each other and the waveform shape is relatively
uniform. Low ripples are apparent because of ON/OFF IGBT switching and the
capacitor charging/discharging. Figure 5.11 shows that both the current and voltage of
the DC link of back-to-back MMCs.
As observed From Figure 5.11, the DC link voltage reaches to 1100 V in steady-state,
which is determined by the reference voltage setting. On the other hand, the DC link
current has very little variations due to effective 𝑉𝑐 balance strategy that is set by VSC
controller and the modulating strategy used in the simulation
Figure 5.11 The variation in DC link current (upper) and DC link voltage (lower).
5.4.4 PMSG/Grid Active and Reactive Power
The active and reactive power of the both sides of back-to-back MMCs are presented
in Figure 5.12.
61
Figure 5.12 PMSG side and grid side active and reactive power.
From the observations it can be inferred that the active and reactive power are affected
with the variation in wind speed. The active power is about 0.5 MW at 3 s at starting
and the active power rises up to 0.56 MW at 6s when wind speed is 12m/s and then it
reduces to 0.25 MW at 9 s when wind speed is 7m/s. It raises again to reach 0.32 MW
at 12 s when wind speed is 10m/s and it raised up to reach 0.9 MW at 15 s when wind
speed is 15m/s then it reduced to 0.2MW at 15 s when wind speed is 5m/s.
5.5 PMSG MMC System with Ultracapacitor Tests:
A total of four different simulation cases both without/with ultracapacitor has been
investigated in this work. These are as follows:
1) PMSG based WECS under variable wind speed conditions,
2) One-phase to ground AC fault (Phase A to ground fault) occurs at the converter
AC side of PMSG based WECS system.
3) Three-phase to ground AC fault occurs at the converter AC side of PMSG
based WECS system.
4) Short circuit at the DC link of back-to-back MMCs.
62
5.6 Simulation results of PMSG based WECS under variable wind speed
conditions
In this part we test the PMSG based WECS system, in this part, we have connected
PMSG to back-to-back MMCs in which the PMSG is connected to AC side of one
MMC, while the AC side of the other MMC is connected to the grid. We have also
tested the system dynamics with variable wind speed. There are three inputs; one is
wind speed that is changed in every 3s which its values are (5,12, 7, 10, 15, 5 and 12)
m/s. The second parameter is the generator speed in pu, the third value is the pitch
angle and we assume its value to be zero. Each MMC has 12 cell submodules in each
arm and there are two VSC controllers in both sides. Phase shift PWM is used to
control switching operation of half bridge IGBTs. In addition, there is an RL filter in
grid side of MMC to smooth the signal and reduce the ripples of current/voltage
waveforms. The grid is modelled using 3-phase 1150 V ideal voltage source. A 3-
phase load is also connected in parallel with the grid phases.
5.6.1 Voltage and current waveforms
The voltage and current waveforms for both supply (wind turbine side) and load side
(grid side) of the PMSG based WECS under variable speed conditions. The wind
turbine is fed by variable wind speed, and we have used step block in order to change
wind speed every 3 s. These wind speed instants are 5, 12, 7, 10, 15, 5 and 12 m/s,
which are presented in Figure 5.13 and for a time period between 3 and 3.2 s in Figure
5.14. The simulated voltage and current waveforms show that adding ultracapacitor
did not make any significant changes to the voltage and current waveforms for both
supply and load sides. For both cases, it can be understood that the WECS is working
well in three-phase balanced conditions.
63
WECS with no ultracapacitor (left) and with ultracapacitor (right)
64
WECS with no ultracapacitor (left) and with ultracapacitor (right) for a time period
between 3 and 3.2 s
65
5.6.2 Active power and reactive power waveforms
Figure 5.15 Active power and reactive power waveforms of the supply side of the
PMSG based WECS with no ultracapacitor (left) and with ultracapacitor (right)
The active power and reactive power readings for both sides of MMC (supply side and
load side) are shown in Figure 5.15. The simulation results show that after adding
ultracapacitor reduces active and reactive power fluctuations during startup when
compared with the case without ultracapacitor. As soon as the system reaches steady-
state condition, the transients of active power and reactive power waveforms become
similar for both cases, namely without and with ultracapacitor.
66
5.6.3 Apparent power waveforms
The simulated apparent power waveforms for both sides of MMC (supply side and
load side) are shown in Figure 5.16. As can be seen from Figure 5.16, the apparent
power of the load with ultracapacitor becomes stable in 0.2 s, while it needs about 2.5
s to reach the stability condition in the WECS without ultracapacitor. On the other
hand, the supply side apparent power during startup is reduced with ultracapacitor
from around 5 MVAR to 2.25 MVAR. This reduced startup transient can benefit the
power system to which WECS is connected by not disturbing the stable operation of
the power system. This situation can avoid overloading of the transformers and any
possible damaging of the protection devices. Finally, we can see that, at the load side,
the apparent power transient of the load during startup is reduced with the utilization
of the ultracapacitor.
Figure 5.16 Apparent power waveforms of the supply side of the PMSG based
WECS with no ultracapacitor (left) and with ultracapacitor (right)
67
5.6.4 DC link voltage waveforms
DC link voltage waveform of the back-to-back MMCs of the PMSG based WECS is
shown in Figure 5.17. The obtained waveforms illustrates that adding ultracapacitor to
the system significantly reduces the startup transients occurred at the DC link voltage
by approximately 5 times. This positive situation will protect the equipment, especially
the capacitors of the DC link voltage from a possible damaging during the startup of
the WECS.
Figure 5.17 DC link voltage waveforms of the back-to-back MMCs of the PMSG
based WECS with no ultracapacitor (left) and with ultracapacitor (right)
5.6.5 Voltage, current, and SOC of ultracapacitor
The voltage, current and SOC readings of the ultracapacitor are shown in Figure 5.18.
As can be seen from Figure 5.18, the voltage and current waveforms of the
ultracapacitor are steady-state and stable. The ultracapacitor voltage remains at the
reference value of 1150 volt. In other hand, the ultracapacitor currents at the 320kA.
The SOC of the ultracapacitor stays in the range between 95% and 96.5%.
68
Figure 5.18 The voltage, current and SOC readings of the ultracapacitor
5.7 Simulation results of PMSG based WECS under phase A to ground fault
In this part, we have investigated the effect of phase-A to ground fault at the load side
(grid side). For this test we have added the “fault block” to the PMSG based WECS
system explained in section 5.2. The wind speed is varied every 3 s and the total time
of test is 10 s. The wind speed instants are 5, 12, 7, and 10. The “fault block” set at the
load side and connected to the three-phase of the grid as shown in Figure 5.19. The
phase A has been selected to emulate this type of fault in the fault block. On the other
hand, the selected parameters of the fault block are presented in Figure 5.20. The fault
switching time is selected to be in period between 4.98 and 5 s (200 ms).
69
Figure 5.19 The modification of the simulation model for phase A to ground fault
Figure 5.20 The parameters of the fault block for phase A to ground fault
70
5.7.1 Voltage and Current Waveforms
(grid side) of the PMSG based WECS under the fault condition are illustrated in Figure
5.21. The results show that after the fault is cleared at 𝑡 = 5 s, the ultracapacitor was
able to damp out the fluctuations efficiently in the voltage waveforms at the supply
side. From the presented results it can also be inferred that using ultracapacitor greatly
improved the voltage profile of the PMSG based WECS dynamically under such a
strong disturbance. However, at the load (grid) side, there were not too many changes
in the voltage and current waveforms of the load following after the fault.
PMSG based WECS under phase A to ground fault with no ultracapacitor (left) and
with ultracapacitor (right)
71
5.7.2 Active Power and Reactive Power Waveforms
load side) under the fault condition is shown in Figure 5.22. As can be seen from Figure
5.22, the active power and reactive power start to suffer from strong oscillations at the
fault time. Especially, at the load side, active power and reactive power without
ultracapacitor are fluctuating, while in case with ultracapacitor, the fluctuations are
relatively small and they are damped out at about 0.02 s. Following after the fault when
it is cleared at t=5 s, it can be seen that there are no significant dynamic changes in the
waveforms of active power and reactive power of the PMSG based WECS for both
supply side and load side following after the fault for both cases of without and with
ultracapacitor.
load of the PMSG based WECS under phase A to ground fault with no ultracapacitor
(left) and with ultracapacitor (right)
72
5.7.3 DC Link Voltage Readings
The DC link voltage waveform of the back-to-back MMCs of PMSG based WECS
under phase A to ground fault is shown in Figure 5.23. From the presented simulated
waveforms, it is apparent that the DC link voltage of the system is robust with the
ultracapacitor under such a strong disturbance and it was able to keep the DC link
voltage at the reference value. However, without ultracapacitor, the DC link voltage
has a ripple around at 1430 volts. We can comment that using ultracapacitor in the
PMSG based WECS can avoid strong ripples at the DC link voltage of the system and
thus protect the capacitors from a possible damage.
WECS under phase A to ground fault with no ultracapacitor (left) and with
ultracapacitor (right)
73
5.7.4 Voltage, Current, and SOC of Ultracapacitor
The voltage, current and SOC readings of the ultracapacitor under phase A to ground
fault are shown in Figure 5.24. As can be seen from the results, the voltage and current
waveforms of ultracapacitor are smooth during the disturbance. The ultracapacitor
voltage stays at its reference value of 1150 volt. On the other hand, the SOC value of
the ultracapacitor remains in the range between 95% and 96.5%.
Figure 5.24 The voltage, current and SOC readings of the ultracapacitor under phase
A to ground fault
5.8 Simulation Results of PMSG Based WECS Under Three-phase Fault
In this part, we have investigated the effect of adding ultracapacitor to the PMSG based
WECS under three-phase fault to ground fault at the load side. For this test we have
added the “fault block” to the PMSG based WECS system explained in section 5.2.
The wind speed is varied every 3 s and the total time of test is 10 s. The wind speed
instants are 5, 12, 7, and 10. The connection of the fault block to the system is shown
in Figure 5.25. The fault is applied at 𝑡 = 4.98 s and cleared at 𝑡 = 5.0 s for a duration
of 200 ms. The parameters of the fault block are shown in Figure 5.26.
74
Figure 5.25 The modification of the simulation model for three-phase to ground fault
Figure 5.26 The parameters of the fault block for three-phase to ground fault
75
5.27. From Figure 5.27, it can be observed that following after the three-phase to
ground fault, the supply current waveform swell is well compensated by the
ultracapacitor when compared with the case without ultracapacitor. However it is
observed that there is no noteworthy changes in the voltage and current waveforms of
the load due to this fault. This is due to the fact that the load is connected to the stiff
power system very closely.
PMSG based WECS under three-phase to ground fault with no ultracapacitor (left)
and with ultracapacitor (right)
76
load side) under the fault condition is shown in Figure 5.28. From Figure 5.28, it can
be observed that there are no significant differences in the waveforms of the load side
between without-ultracapacitor and with-ultracapacitor cases following after the fault.
However, when the fault is cleared at 𝑡 = 5 sec, it can be clearly observed that the
active power fluctuations at the WECS side due to the fault are significantly reduced
by the ultracapacitor.
load of the PMSG based WECS under three-phase to ground fault with no
ultracapacitor (left) and with ultracapacitor (right)
77
under three-phase to ground fault is shown in Figure 5.29. As can be observed from
the simulated waveforms, there is almost no change in the DC link voltage following
after the fault in case when ultracapacitor is used. However, when there is no
ultracapacitor is used, the DC link voltage goes temporary down to around 700 volts.
In addition, at the time of the fault the DC link voltage of the system with ultracapacitor
is little affected, while in case when the system is without ultracapacitor, the DC link
voltage is temporary dropped and then returned to its reference value back again.
WECS under three-phase to ground fault with no ultracapacitor (left) and with
78
The voltage, current and SOC readings of the ultracapacitor under three-phase to
ground fault are shown in Figure 5.30. As can be seen from the results presented in
Figure 5.30, the voltage and current waveforms of the ultracapacitor are robust and
they are at their steady-state values. The ultracapacitor voltage stays at the reference
value of 1150 volt. The SOC of the ultracapacitor remains in the range between 95%
and 96.5%.
Figure 5.30 The voltage, current and SOC readings of the ultracapacitor under three-
phase to ground fault
79
5.9 Simulation Results of PMSG Based WECS Under Short Circuit at DC Link
Figure 5.31 The modification of the simulation model for short circuit at the DC link
In this part, we have investigated the effect of adding ultracapacitor to the PMSG based
WECS under a short circuit at the DC link of the back-to-back MMCs. For this test,
we have added the “fault block” to the system and connected it to the positive and
negative lines of the DC link, which is shown in Figure 5.31. Fault occurring and
clearing times are selected as 𝑡 = 4.98𝑠 and 𝑡 = 5.0𝑠, respectively. The parameters of
the fault block are shown in Figure 5.32.
Figure 5.32 The parameters of the fault block for short circuit at the DC link
80
5.33. As can be observed from the results presented in Figure 5.33. At the time of the
fault, the voltage and the current waveforms of the supply and the load have ripples.
Moreover, it can be noticed that the supply voltage without ultracapacitor is dropped
to 0 volts. However, in case with the ultracapacitor, the supply voltage is only dropped
to 50 volts peak and after the fault is cleared, it reaches its steady-state condition. On
the other hand, the load current waveform is rippled during the fault period. However,
it returns to the steady-state condition after 0.04 s, while the load current waveform
for the system without ultracapacitor stays rippled. Following after the fault when it is
cleared at 𝑡 = 5 sec, the transients are reduced by the ultracapacitor at the WECS side.
However, at the load side, there is no significant change observed in the voltage and
current waveforms of the load.
PMSG based WECS under short circuit at the DC link with no ultracapacitor (left)
and with ultracapacitor (right)
81
load side) under the fault condition is shown in Figure 5.34. The simulation results for
the active and reactive power (Figure 5.34) show that the adding ultracapacitor reduces
active and reactive power fluctuations during startup when compared with the case
without ultracapacitor. As soon as the system reaches to steady-state operation, the
transients of active and reactive power waveforms are similar to each other for both
cases.
load of the PMSG based WECS under short circuit at the DC link with no
ultracapacitor (left) and with ultracapacitor (right)
82
under short circuit at the DC link is shown in Figure 5.35. From the presented results
in Figure 5.35, the DC link voltage of the system with ultracapacitor stays at its
reference value, while the DC link voltage of the system without ultracapacitor needs
about 3 seconds to reach its steady-state value following after the fault. During the
fault, the DC link voltage of the system without ultracapacitor is dropped to 0 volt,
while the DC link voltage of the system with ultracapacitor is dropped to about 50
volts.
WECS under short-circuit at the DC link with no ultracapacitor (left) and with
This section presents voltage, current and SOC readings of the ultracapacitor in case
of short circuit at the DC link. As can be seen from the presented results in Figure 5.36,
the ultracapacitor voltage and current, as well as the SOC, are at their steady-state
values before the fault. However, during the fault condition, the voltage and the
current, as well as the SOC of the ultracapacitor are temporary dropped to 0 volt.
83
Figure 5.36 Ultracapacitor readings, from up to down: voltage, current and state of
charge (SOC)
5.10 Discussion of the Results
From results we can conclude the following:
1- The variable wind speed is affected by power generation in which the power is
raised and reduced based on win speed value.
2- Adding ultracapacitor to PMSG based WECS with MMC do not make any
significant changes to the voltage and current waveforms for both supply and load
sides in steady-state conditions.
3- Adding ultracapacitor to PMSG based WECS with MMC reduces both active and
reactive power fluctuations during startup when compared with the case without
ultracapacitor. However, as soon as the system reaches steady-state operation, the
transients of active and reactive power waveforms are similar to each other for both
cases.
84
4- Adding ultracapacitor can reduce the supply apparent power during startup. This
reduced startup transient can benefit the power system to which WECS is connected
by not disturbing the stable operation of the power system. This situation can avoid
overloading of the transformers and any possible damaging of the protection
devices. Also, adding ultracapacitor to the load side is reduced apparent power
transient of the load during startup too.
5- Adding ultracapacitor is significantly reduces the startup transients of the DC link
voltage approximately 5 times. This situation will protect the equipment, especially
the capacitors of the DC link voltage from a possible damaging.
6- For the case of phase-A to ground fault, the advantage of adding ultracapacitor was
able to damp out the fluctuations efficiently in the voltage waveforms at WECS
side greatly improved the voltage profile of the WECS dynamically under such a
strong disturbance, but in load side there were no too much changes in the voltage
and current waveforms of the following after the fault.
7- For case of 3-Phase to ground short circuit fault, the three-phase current, supply
current waveform swell is well compensated by ultracapacitor when compared to
without ultracapacitor case. However, there are no noteworthy changes in the
voltage and current waveforms of the load.
8- For the case of Short circuit at DC-Link fault, the supply voltage for both systems
(with ultracapacitor and without it) is drop-down toward 0 volts. Moreover, the load
current waveform is rippled at fault period, however in case of system with
ultracapacitor to system it has been returns to uniform and sinusoidal after little
milliseconds second, while in case of system without ultracapacitor the load current
waveform for stay rippled. In addition, after the fault when it is cleared, the
transients are reduced by ultracapacitor at the WECS side. However, at the load
side, there is no significant change can be occurred in the voltage and current
waveforms of the load side.
85
CHAPTER 6
CONCLUSION
6.1 Conclusion
With the improvement of the ultracapacitor technology, the application of the

ultracapacitors in wind turbine systems as energy storage elements will be
unavoidable. In this work, at first, a simulation model of a PMSG based WECS has
been designed with back-to-back MMCs in order to control the system variables
efficiently. Then, an ultracapacitor has been added to the proposed system in order to
investigate the improvements and potential benefits of using the ultracapacitor in the
WECS. Four different simulation cases have been considered for both without and
with ultracapacitor. These are normal operating of the WECS under variable wind
speed conditions, phase-A to ground fault at the load side, three-phase to ground fault
at the load side, and short circuit at DC link of the back-to-back MMCs of the PMSG
based WECS. The simulation results show that the adding of ultracapacitor did not
make any significant changes to the voltage and current waveforms for both supply
and load sides. However, for the active power and reactive power, there is a
momentous reduction in the active power and reactive power fluctuations during
startup when the ultracapacitor is utilized. When the system reaches its steady-state
condition after clearing the disturbances, the active power and reactive power
waveforms are similar to each other for both cases (without and with ultracapacitor).
It is also observed from the simulation studies that the supply apparent power during
startup is reduced with the addition of the ultracapacitor from around 5 MVAR to 2.25
MVAR. This reduced startup transients can benefit the power system to which the
WECS is connected by not disturbing the stable and balanced operation of the rest of
the power system. Moreover, this situation can avoid overloading of the transformers
and any possible damaging of the protection devices. For the load side, it is also
observed that the apparent power transient during startup is reduced with the addition
of ultracapacitor. The simulation results of the DC link fault show that the startup
transients of the DC link voltage reduce approximately by 5 times when the
86
an ultracapacitor is used. Again, this situation will protect the equipment, especially
the capacitors of the DC link voltage from a possible damaging. For line faults at the
AC sides, it is noticed that after the fault is cleared, the ultracapacitor was able to damp
out the fluctuations efficiently in the voltage waveforms at WECS side. From the
presented results it can be inferred that using ultracapacitor greatly improves the
voltage profile of the WECS dynamically under such strong disturbances. At the load
(grid) side, there were not too many changes in the voltage and current waveforms of
the load following after the fault. On the other hand, the simulation results show that
load’s active power and reactive power without ultracapacitor fluctuates in case when
ultracapacitor is not used. But when the ultracapacitor is utilized, the active power and
reactive power recovers their steady-state values after just a little transient time around
20 ms. The DC link voltage of the back-to-back MMCS of the PMSG based WECS
with the ultracapacitor is able to stay at its reference value when there is a short-circuit
at the DC link. However, in case when no ultracapacitor is used, a transient time of
approximately 260 ms is needed for the restoration of the DC link voltage. In summary,
it can be concluded that adding ultracapacitor to a wind energy system can improve
the dynamic conditions of the system under disturbances such as faults.
6.2 Suggestions for Future Work
The following suggestions can be provided related to the wind energy systems:
1. Prepare similar studies and simulate it depending on the other types of wind
generators like DFIG and compare the efficiency of both systems.
2. Use artificial intelligence-based control techniques such as deep learning and
new types of fuzzy logic and neural networks can be proposed in order to
enhance the dynamic operating conditions of the wind energy system.
3. Verify the analytical results obtained from such modeling and simulation with
the real-world on-line testing.
4. Study the hybrid design of a micro-grid that includes other power sources.
5. Make some Improvements in MMC by Developing dynamic controls that are
able to provide a broad variety of voltage conversion ratios.
87
6. Integrate ultracapacitor storage system using the proposed control strategy into
PMSG- based wind farm consists of a group of wind turbines. Each PMSG-based
wind turbine is equipped with ultracapacitor.
88
REFERENCES
[1] B. Andersen, “A comparison of two and three bladed floating wind turbines.”
University of Toledo, 2010.
[2] R. E. Use and C. Scientists, “Benefits of Renewable Energy Use | Union of
Concerned Scientists,” Benefits Renew. Energy Use, 2013.
[3] J. Vestergaard, L. Brandstrup, and R. D. Goddard, “A brief history of the wind
turbine industries in Denmark and the United States,” in Academy of
international business (Southeast USA Chapter) Conference proceedings, 2004,
pp. 322–327.
[4] T. Thiringer, “Wind Power Generator Systems and Local Power System
Interconnection,” Encycl. Sustain. Sci. Technol., pp. 12134–12174, 2012.
[5] IRENA-5, “RENEWABLE ENERGY TECHNOLOGIES: COST ANALYSIS
SERIES; Wind Power,” IRENA, 2014.
[6] Z. Chen and H. Li, “Overview of different wind generator systems and their
comparisons,” IET Renew. Power Gener., 2008.
[7] M. Bilgili, A. Yasar, and E. Simsek, “Offshore wind power development in
Europe and its comparison with onshore counterpart,” Renew. Sustain. Energy
Rev., vol. 15, no. 2, pp. 905–915, 2011.
[8] Z. Q. Zhu and J. Hu, “Electrical machines and power-electronic systems for
high-power wind energy generation applications: Part II - Power electronics and
control systems,” in COMPEL - The International Journal for Computation and
Mathematics in Electrical and Electronic Engineering, 2013.
[9] L. H. Hansen, L. Helle, F. Blaabjerg, E. Ritchie, H. Bindner, and P. Sørensen,
“Conceptual survey of Generators and Power Electronics for Wind Turbines,”
2001.
[10] D. Xie, Y. Lu, J. Sun, and C. Gu, “Small signal stability analysis for different
types of PMSGs connected to the grid,” Renew. Energy, vol. 106, pp. 149–164,
2017.
[11] H. Saad, J. Mahseredjian, S. Dennetière, and S. Nguefeu, “Interactions studies
of HVDC-MMC link embedded in an AC grid,” Electr. Power Syst. Res., vol.
89
138, pp. 202–209, 2016.
[12] M. R. Abedi and K. Y. Lee, “Smart energy storage system for integration of
PMSG-based wind power plant,” IEEE Power Energy Soc. Gen. Meet., vol.
2015-Sept, no. 1, pp. 2–6, 2015.
[13] K. Zhang, C. Mao, J. Lu, D. Wang, J. Zhang, and X. Chen, “Power control of
directly driven wind generation system based on superconducting magnetic
energy storage,” Electr. Power Components Syst., vol. 43, no. 7, pp. 747–758,
2015.
[14] G. Q. Li, X. Z. Ren, and H. H. Chen, “A control strategy to mitigate the
fluctuation of wind power with ultra-capacitor,” Appl. Mech. Mater., vol. 448–
453, pp. 1742–1746, 2014.
[15] U. Karaagac, J. Mahseredjian, L. Cai, and H. Saad, “Offshore Wind Farm
Modeling Accuracy and Efficiency in MMC-Based Multiterminal HVDC
Connection,” IEEE Transactions on Power Delivery, vol. 32, no. 2. pp. 617–
627, 2017.
[16] R. Takahashi et al., “Smoothing control of wind farm output fluctuation by the
new scheme with energy storage system,” p. 9 pp., 2011.
[17] G. El-sandy, E. A. Ibrahim, H. Ziedanand, and M. M. Soliman, “Modeling and
Operation of Permanent Magnet Synchronous Generator Wind Energy
Conversion System Connected with Grid,” no. May 2018.
[18] M. Chinchilla, S. Arnaltes, and J. C. Burgos, “Control of permanent-magnet
generators applied to variable-speed wind-energy systems connected to the
grid,” IEEE Trans. Energy Convers., 2006.
[19] H. S. Ko, “Modeling and control of PMSG-based variable-speed wind turbine,”
in Advances in Industrial Control, 2014.
[20] R. H. Baker and L. H. Bannister, “Electric power converter.” Google Patents,
18-Feb-1975.
[21] T. A. Meynard and H. Foch, “Multilevel conversion: high voltage choppers and
voltage-source inverters,” in PESC’92 Record. 23rd Annual IEEE Power
Electronics Specialists Conference, 1992, pp. 397–403.
[22] J.-S. Lai and F. Z. Peng, “Multilevel converters-a new breed of power
converters,” in IAS’95. Conference Record of the 1995 IEEE Industry
Applications Conference Thirtieth IAS Annual Meeting, 1995, vol. 3, pp. 2348–
2356.
90
[23] L. M. Tolbert and F. Z. Peng, “Multilevel converters for large electric drives,”
in APEC’98 Thirteenth Annual Applied Power Electronics Conference and
Exposition, 1998, vol. 2, pp. 530–536.
[24] A. Nabae, I. Takahashi, and H. Akagi, “A new neutral-point-clamped PWM
inverter,” IEEE Trans. Ind. Appl., no. 5, pp. 518–523, 1981.
[25] B. Çiftçi and A. M. Hava, “Modüler çok seviyeli dönüştürücüler: topolojiler ve
donanımsal özellikler,” V. Enerj. Veriml. ve Kalitesi Sempozyumu, pp. 141–145,
2013.
[26] H. Akagi, “Classification, terminology, and application of the modular
multilevel cascade converter (MMCC),” IEEE Trans. power Electron., vol. 26,
no. 11, pp. 3119–3130, 2011.
[27] “Siemens ‘The Smart Way HVDC Plus - One Step Ahead,’” 2014. [Online].
Available:
http://www.energy.siemens.com/hq/pool/hq/powertransmission/HVDC/HVD
C-PLUS/HVDC-PLUS_The-Smart-Way.pdf.
[28] S. Debnath, J. Qin, B. Bahrani, M. Saeedifard, and P. Barbosa, “Operation,
control, and applications of the modular multilevel converter: A review,” IEEE
Trans. power Electron., vol. 30, no. 1, pp. 37–53, 2014.
[29] S. Faias, P. Santos, J. Sousa, and R. Castro, “An overview on short and long-
term response energy storage devices for power systems applications,” system,
vol. 5, p. 6, 2008.
[30] T. M. Masaud, K. Lee, and P. K. Sen, “An overview of energy storage
technologies in electric power systems: What is the future?,” in North American
Power Symposium 2010, 2010, pp. 1–6.
[31] S. C. Smith, P. K. Sen, and B. Kroposki, “Advancement of energy storage
devices and applications in electrical power system,” in 2008 IEEE Power and
Energy Society General Meeting-Conversion and Delivery of Electrical Energy
in the 21st Century, 2008, pp. 1–8.
[32] H. S. Patel and R. G. Hoft, “Generalized techniques of harmonic elimination
and voltage control in thyristor inverters: Part I--Harmonic Elimination,” IEEE
Trans. Ind. Appl., no. 3, pp. 310–317, 1973.
[33] J. Rodriguez, J.-S. Lai, and F. Z. Peng, “Multilevel inverters: a survey of
topologies, controls, and applications,” IEEE Trans. Ind. Electron., vol. 49, no.
4, pp. 724–738, 2002.
91
[34] M. Y. Worku, “Power smoothing control of PMSG based wind generation using
supercapacitor energy storage system,” Int. J. Emerg. Electr. Power Syst., vol.
18, no. 4, 2017.
[35] A. Cimpoeru, “Encoderless Vector Control of PMSG for Wind Turbine
Applications: Encoderless, Control, PMSG,” Aalborg Univ. Inst. Energy
Technol. Reino da Dinamarca, 2010.
[36] F. Obeidat, X. Lie, and L. Yongdong, “Simulation of grid connected HVDC
offshore wind farm topologies,” in 2013 IEEE 10th International Conference
on Power Electronics and Drive Systems (PEDS), 2013, pp. 897–902.
[37] S. Ghasemi, A. Tabesh, and J. Askari-Marnani, “Application of fractional
calculus theory to robust controller design for wind turbine generators,” IEEE
Trans. energy Convers., vol. 29, no. 3, pp. 780–787, 2014.
[38] M. Anburaj, T. Rameshkumar, and A. Ajitha, “Wind Energy Conversion Based
On Matrix Converter,” Int. J. Power Electron. Drive Syst., vol. 5, no. 1, p. 119,
2014.
[39] A. Cimpoeru and K. Lu, “Encoderless vector control of PMSG for wind turbine
applications,” Inst. Energy Technol., 2010.
[40] M. D. Kharad and N. Kumar, “Modeling and Simulation of Unified Power
Quality Conditioner (UPQC),” Int. J. Eng. Adv. Technol. ISSN, pp. 2249–8958.
[41] J. Rodriguez, P. Correa, and L. Moran, "A vector control technique for medium
voltage multilevel inverters," Applied Power Electronics Conference and
Exposition, 2001. APEC 2001. Sixteenth Annual IEEE, vol.1, pp.173-178
vol.1, 2001.
[42] Ciftci, B. (2014). Selection of suitable PWM switching and control methods for
modular multilevel converter drives. Norwegian University of Science and
Technology, NTNU.
[43] Ilves, K., Taffner, F., Norrga, S., Antonopoulos, A. (2013 September). A
Submodule Implementation for Parallel Connection of Capacitors in Modular
Multilevel Converters, Conference Paper in IEEE Transactions on Power
Electronics, DOI: 10.1109/EPE.2013.6634325.
92
93

559069

Uploaded by

Copyright:

Available Formats

559069

Uploaded by

Document Information

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

559069

Uploaded by

Copyright:

Available Formats

September 2019

Design and Simulation of a Back-to-Back Modular Multilevel

NISSAN SABBAR THUMAIL

Name of the thesis: Design and Simulation of a Back-to-Back Modular Multilevel

Prof. Dr. A. Necmeddin YAZICI

Prof. Dr. Ergun ERÇELEBİ

Assoc.Prof.Dr. Ahmet Mete VURAL

Examining Committee Members: Signature

Assoc. Prof. Dr. Ahmet Mete VURAL …………………..

Assist. Prof. Dr. Serkan ÖZBAY …………………..

Assist. Prof. Dr. Kadir Sercan BAYRAM …………………..

Nissan Sabbar Thumail

Key Words: Wind Energy Conversion System; Renewable Energy; Permanent

Anahtar Kelimeler: Rüzgar Enerjisi Dönüşüm Sistemi; Yenilenebilir Enerji; Sabit

LIST OF FIGURES .............................................................................................. XIV

LIST OF TABLES ............................................................................................. XVIII

LIST OF SYMBOLS / ABBREVIATIONS ........................................................ XIX

2.1 Introduction ...................................................................................................... 7

3.1 Overview ........................................................................................................ 12

MODELING OF THE WECS ................................................................................ 36

4.1 Introduction .................................................................................................... 36

SIMULATION RESULTS ...................................................................................... 50

5.1 Introduction .................................................................................................... 50

6.1 Conclusion ....................................................................................................... 86

Figure 1.2 Wind power generation system topologies [8] ………………………….. 4

Figure 3.1 Main parts of a WECS [6] ....................................................................... 12

Figure 3.3 (a) Output voltage waveform of a single‫ـ‬phase five-level diode‫ـ‬clamped

Figure3.4 (a) SSBC topology (b) SDBC topology.................................................... 18

(b) Chopper cell (DSCC) (c) bridge cell (DSBC) ...................................................... 18

Figure 3.6 The topology of the full-bridge circuit .................................................... 20

Figure 3.8 Three-phase MMC circuit topology ........................................................ 23

Figure 3.9 Half‫ـ‬bridge circuit. ................................................................................... 23

Figure 3.10 States of half-bridge circuit and current paths ....................................... 24

Figure 3.11 (a) Diode-clamped. three-level converter (b) Capacitor-clamped .

Figure 3.12 MMC switching. methods. .................................................................... 30

Figure 3.13 The carriers of phase disposition switching technique .......................... 31

Figure 3.16 The carriers of phase‫ـ‬shifted switching technique ................................. 33

Figure 3.17 The carriers of sawtooth rotation method .............................................. 33

Figure 4.1 PMSG-based wecs based on back-to-back MMCs .................................. 36

Figure 4.3 The overall structure of the proposed MMC ........................................... 40

Figure 4.4 One arm structure of the mcc ................................................................... 41

Figure 4.5 Generator side model ............................................................................... 43

Figure 4.6 Drive train ................................................................................................ 44

Figure 4.7 Pitch angle controller ............................................................................... 46

Figure 4.8 MMC control ........................................................................................... 47

Figure 4.9 Circuit structure of MMC [43]................................................................. 48

Figure 5.5 PMSG-MMC system model .................................................................... 56

Figure 5.25 The modification of the simulation model for three-phase to .

ground fault ................................................................................................................ 75

Table 3.3 The possible variations of MMC [25] ....................................................... 17

Table 3.4 The switching states of full-bridge circuit................................................. 20

Table 3.5 Switching. states of half-bridge circuit...................................................... 24

Table 4.1 Wind turbine parameters ........................................................................... 43

Table 5.1 Simulation parameters for PMSG ............................................................. 51

Table 5.2 Simulation parameters for MMC .............................................................. 51

Table 5.3 Simulation parameters for ultracapacitor .................................................. 51

1.2 A Brief History of Wind Energy Systems

Figure 1.2 Wind power generation system topologies [8]

1.3 Objective of the Thesis

1.4 Organization of the Thesis