Design of A Generator Excitation System

U NIVERSITY OF B ERGEN
Geophysical Institute
Faculty of Mathematics and Natural Sciences
M ASTERS T HESIS
Design of a
Generator Excitation System
Author: Jonas Fotland Fænn
June 1, 2021
ii
iii
Abstract
The world requires more and more energy, and the synchronous generator is one of the most
widely used machine to create electricity. With all the debate about climate changes, and
electrification of the power industry, it is therefore important to get a deep comprehension
of how the machine works, and how to get the machine to work in reliable conditions. The
excitation system provides many features such as simplicity of control, stability and reliabil-
ity under all conditions. A fully working excitation system is with other words essential to get
a reliable power system. This thesis will take a look at the theory surrounding the excitation
system, all from the synchronous machine itself, the power electronics and the excitation
system. The thesis focus on designing an excitation system for a synchronous machine us-
ing Matlab ® /Simulink® . The machine is suppose to be as similar as possible to the Terco
MV1027-235 synchronous machine stationed at Western Norway University of Applied Sci-
ences.
The work resulted in a working excitation system based on an single-phase thyristor recti-
fier. Along with that two control systems were made, one which the field voltage could be
manually adjusted, and one which would automatically control the terminal voltage output
of the synchronous machine.
iv
Acknowledgment
This master thesis is written in collaboration of the University of Bergen and Western Norway
University of Applied Sciences. It marks the end of my two year master study, and a total of
five year studying electrical engineering.
I would like to thank my supervisor, Associate professor Emil Cimpan, for great help and
guidance during this year. I would also like to thank my co-supervisor Mostafa Paskyabi for
help with the report, and at last my three fellow friends; Gabriel Abildgaard, Marius Reigstad
and Fredrik Storebø for all discussions and help throughout this five year period.
(JF)
v
Contents
Abstract ii
Acknowledgment iv
List of Figures vii
List of Tables x
Acronyms xi
Symbols xii
1 Introduction 1
1.1 Literature review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 State of the art 4

2.1 SM Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1.1 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1.2 Voltage Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.3 Park Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Excitation System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.1 Type DC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.2 Type AC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Stationary rectifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Rotating rectifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.3 Type ST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3 Rectification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3.1 Single-Phase Uncontrolled Rectifier . . . . . . . . . . . . . . . . . . . . . . 16
2.3.2 Three-Phase Rectifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
vi CONTENTS
Capacitor filter, C and source inductor, L s . . . . . . . . . . . . . . . . . . 18

2.4 Control Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.4.1 Automatic Reactive Power Regulator (AQR) . . . . . . . . . . . . . . . . . . 19
2.4.2 Automatic Power Factor Regulator (APFR) . . . . . . . . . . . . . . . . . . 20
2.4.3 Under-excitation Limiter (UEL) . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4.4 Overe-xcitation Limiter (OEL) . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.4.5 Power System Stabilizer (PSS) . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.5 Basic Control Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.5.1 Proportional (P) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.5.2 Integral (I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.5.3 Differential (D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.5.4 Proportional-Integral (PI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.5.5 Proportional-Differential (PD) . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.5.6 Proportional-Integral-Differential (PID) . . . . . . . . . . . . . . . . . . . 26
3 Method 27
3.1 Full ® /Simulink® Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2 Synchronous Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.3 SCR Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.4 Control Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.4.1 Semiconductor Controlled Rectifiers (SCR) Control . . . . . . . . . . . . . 32
3.4.2 Terminal Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.4.3 Pre-charge Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.5 Step Response SCR Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.6 Step Response Full Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.6.1 Additional PI-Tuning for Terminal Control . . . . . . . . . . . . . . . . . . 36
4 Results 38
4.1 SM Input/Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.2 Relationship between i f and Vt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.3 Bridge Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.4 SCR Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.5 SCR to Terminal Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.6 Mechanical Input Variations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.6.1 PI Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5 Discussion 48
6 Conclusions 51
CONTENTS vii
Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
References I
Appendices V
A Park transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V
B Simulink Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII
C Ziegler-Nichols & Cohen-Coon Tuning . . . . . . . . . . . . . . . . . . . . . . . . XII
D General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XIII
Index XV
viii
List of Figures
2.1 Simplified schematic of a single-phase Synchronous Machine (SM) with salient

poles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Simplified schematic of salient and cylindrical rotor. . . . . . . . . . . . . . . . . 5
2.3 Simplified schematic of damper winding. . . . . . . . . . . . . . . . . . . . . . . . 6
2.4 Air-gap flux due to i a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.5 Equivalent circuit for a salient-pole rotor SM. . . . . . . . . . . . . . . . . . . . . 8
2.6 Schematic of a salient-pole SM, adapted from [1]. . . . . . . . . . . . . . . . . . . 9
2.7 Block diagram of full excitation system, consisting of, exciter, control units and
SM [10]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.8 Simplified schematics of system with amplidyne. Adapted from [7]. . . . . . . . 12
2.9 Field-controlled alternator with non-controlled rectifiers, adapted from [7]. . . 13
2.10 Alternator-supplied controlled-rectifier system, adapted from [7]. . . . . . . . . 14
2.11 Simplified schematics of brushless system, adapted from [7]. . . . . . . . . . . . 14
2.12 Simplified schematics of system, adapted from [7]. . . . . . . . . . . . . . . . . . 15
2.13 Single Phase Rectifier and its current flow. . . . . . . . . . . . . . . . . . . . . . . 16
2.14 Input and output voltage curves for the Single Phase Rectifier. . . . . . . . . . . 17
2.15 Three Phase Rectifier. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.16 Three Phase Rectifier. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.17 Illustration of how the capacitor smooth out the rectified voltage. . . . . . . . . 19
2.18 Single-phase bridge with source inductor L s . . . . . . . . . . . . . . . . . . . . . . 19
2.19 AQR Transfer function. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.20 Transfer function. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.21 UEL Transfer Function Model: Q = a + bP linear characteristics. . . . . . . . . . 21
2.22 OEl example model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.23 ∆P Type Transfer Function Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.24 Block diagram of transfer function of a Proportional Integral Differential (PID)
control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.1 Sketch of Matlab ® /Simulink® Model. . . . . . . . . . . . . . . . . . . . . . . . . . 28

LIST OF FIGURES ix
3.2 Thyristor block: a) Equivalen circuit b)VI characteristic. . . . . . . . . . . . . . 29

3.3 Topology of SCR bridge and pre-charge system. . . . . . . . . . . . . . . . . . . . 30
3.4 Sketch of full control system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.5 Sketch of: a) Input voltage b)Gate current . . . . . . . . . . . . . . . . . . . . . 32
3.6 Sketch: Open VS Closed Loop for SCR control. . . . . . . . . . . . . . . . . . . . . 32
3.7 Sketch: Open VS Closed Loop for Terminal Control. . . . . . . . . . . . . . . . . . 33
3.8 Control for the precharge system. . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.9 Step response for SCR bridge. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.10 Dynamic response test to find the 1.order transfer function for the system. . . . 35
3.11 Reaction Curve. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.1 Constant mechanical input at 1500 Rotations per Minute (RPM). . . . . . . . . . 39

4.2 Field voltage set at a reference of 90 V. . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.3 Va ,Vb and Vc as a result of the mechanical input and the field voltage. . . . . . . 39
4.4 Relationship between Field Current and Terminal Voltage. . . . . . . . . . . . . . 40
4.5 Input voltage over a period of 40 ms, and α = 180°. . . . . . . . . . . . . . . . . . 41
4.6 Output voltage over a period of 40 ms, and α = 180°. . . . . . . . . . . . . . . . . 41
4.7 Gate currents over a period of 40 ms, and α = 180°. . . . . . . . . . . . . . . . . . 41
4.8 Input voltage over a period of 40 ms, and α = 0°. . . . . . . . . . . . . . . . . . . . 42
4.9 Output voltage over a period of 40 ms, and α = 0°. . . . . . . . . . . . . . . . . . . 42
4.10 Gate currents over a period of 40 ms, and α = 0°. . . . . . . . . . . . . . . . . . . . 42
4.11 Field voltage step up, 15 V each step. . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.12 Voltage step up on the terminal of the machine. . . . . . . . . . . . . . . . . . . . 43
4.13 The current through SCR1 while stepping up the voltage in steps of 15 V. . . . . 44
4.14 Field voltage variations, switching from SCR to terminal control. . . . . . . . . . 44
4.15 Terminal voltage variations, switching from SCR to terminal control. . . . . . . 45
4.16 Field voltage response from decreasing the mechanical input to 1400 RPM. . . 45
4.17 Terminal voltage response from decreasing the mechanical input to 1400 RPM. 46
4.18 Field voltage response from increasing the mechanical input to 1500 RPM. . . . 46
4.19 Terminal Voltage response from increasing the mechanical input to 1500 RPM. 46
4.20 Switching from SCR control to terminal control. . . . . . . . . . . . . . . . . . . . 47
4.21 Reference change from 200 V to 230 V. . . . . . . . . . . . . . . . . . . . . . . . . . 47
B.1 Full Simulink Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII

B.2 Full control system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX
B.3 Thyristor Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X
B.4 Synchronous Machine connected to load . . . . . . . . . . . . . . . . . . . . . . . XI
x
List of Tables
2.1 Parameters for figure 2.19. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.3 Parameters for figure 2.21 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1 Parameters in Matlab ® /Simulink® for the SM. . . . . . . . . . . . . . . . . . . . . 28

3.2 Parameters for SCR(Adapted from datasheet.) . . . . . . . . . . . . . . . . . . . . 29
3.3 Parameters for SCR bridge. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.4 Parameters for pulse generator A . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.8 Dynamic response on the terminals. . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.9 Dynamic response on the terminals. . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.10 Readings for figure 3.11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.11 Parameters for Ziegler-Nichols tuned regulator. . . . . . . . . . . . . . . . . . . . 37
3.12 Parameters for Cohen-Coon tuned regulator. . . . . . . . . . . . . . . . . . . . . . 37
C.1 Ziegler-Nichols tuning by using reaction curve. . . . . . . . . . . . . . . . . . . . XII

C.2 Cohen-Coon tuning by using reaction curve. . . . . . . . . . . . . . . . . . . . . . XII
D.1 Parameters for Terco MV1027-235. . . . . . . . . . . . . . . . . . . . . . . . . . . . XIII

D.2 Parameters for Thyristor/SCR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XIV
D.3 Parameters for Terco MV1027-235. . . . . . . . . . . . . . . . . . . . . . . . . . . . XIV
xi
Acronyms
AC Alternating Current
APFR Automatic Power Factor Regulator
AQR Automatic Reactive Power Regulator
AVR Automatic Voltage Regulator
D Differential
DC Direct Current
EMF Electromotive Force
I Integral
IEEE Institute of Electrical and Electronics Engineers
OEL Overe-xcitation Limiter
P Proportional
PD Proportional Differential
PI Proportional Integral
PID Proportional Integral Differential
PMSG Permanent Magnet Synchronous Generator
PSS Power System Stabilizer
RMS Root Mean Square
RPM Rotations per Minute
SCR Semiconductor Controlled Rectifiers
SM Synchronous Machine
Type AC AC Excitation System
Type DC DC Excitation System
Type ST Static Excitation System
UEL Under-excitation Limiter
xii
Symbols
C capacitor vi, 18, 30, 34

Ea open-circuit voltage 7, 8
f frequency 5, 17
fN nominal frequency 28
ia stator current phase-a viii, 6, 7
ib stator current phase-b 6
ic stator current phase-c 6
id rotor current 6
if field current 5, 6
iq stator current 6
Kd differential gain 26
Ki integral gain 24, 26
Kp proportional gain 24, 26
L inductor 6
Lf field inductance 28, 30
L md Direct-axis inductance 28, 49
L mq Quadrature-axis inductance 28, 49
Ls source inductor for the thyristor bridge vi, viii, 18, 19, 30
L ssm leakage inductance 28
n rotations per minute 5
p poles 5
PN nominal power, W 28
Rf field resistance 28, 30
RL load resistor 16, 30, 34
Rs stator resistance 28
T period 17
Vi input voltage for the rectifiers 16, 17
VN nominal voltage 28
Symbols xiii
Vo average output voltage for the single-phase diode bridge 16, 17

ε electromotive force 6
Φ flux 6
λag total air gap flux 7
λaa 0 total air gap flux linkages of phase a 7
ω angular frequency 7, 17
α firing angle 32, 40, 48
ε electromotive force 6
λ flux linkage 6
1
1 Introduction
A power system is a network of electrical components used to supply, consume and trans-
fer electrical energy. The generating power plant feeds the electrical grid with power which
then is consumed by the customers. When a phone charger is connected, or a big factory is
starting up their machinery, everyone expects to get the power they pay for at the instant it is
needed. The most common power plants convert energy from fossil fuels, nuclear fuels, and
falling water, into electrical power [1].
In a power system, a lot of things can go wrong. The system is disturbed at all times by
a variety of loads, and natural disasters may occur. Lightning strikes, falling trees, ice or
cyclones damaging essential equipment may lead to blackout [2], [3]. Blackouts can both
be expensive and dangerous, like the one in August 2003 which shut down big areas of U.S.
and Canada, affecting around fifty million humans and costing several billion dollars [4],
[3]. Natural disasters and events like this can never be prevented with 100% security, but
protective and reliable equipment can diminish the harm.
Stable and reliable operation of power systems is a precondition for today’s society. This is
also a huge challenge to electrical power engineers. Deep comprehension of the power sys-
tem and its components is important to maintain the system’s stability. The synchronous
generator is the generator that produces the majority of the electrical power [5]. In a power
system, it is essential to control the voltage and ensure that the generator is always synchro-
nized.
To keep the power system reliable, national demands and international agreements regulate
both voltage and frequency in the system. In Norway, the voltage must always be within
±10% of the agreed voltage, and the frequency must be within 50 ± 2% Hz [6]. Europe has
agreed upon a 50 Hz grid, but this is not the case all over the world; for example, in U.S., they
use 60 Hz.
2 CHAPTER 1. INTRODUCTION
1.1 Literature review

The excitation system provides the field current to the rotor of the SM and by that induces
the magnetic field. It provides both control and protective functions for the machine, and by
this means, it provides reliability for the generator. The excitation system consists of exciter,
regulators, stabilizers, limiters and protective circuits. It also protects components from ex-
ceeding their capability limits [7]. In other words, it is a composite system consisting of more
than just the exciter that provides the field current. The excitation system accounts for ap-
proximately one to a few percent of the rated values for the main generator [8], so in large
SMs, the excitation system may require to produce a field current of thousands of amperes,
at a few hundreds of voltage [1]. The excitation system is meant to be reliable under all con-
ditions, to be a simplicity of control, and to be easy to maintain.
In the early stage, the SM were usually implemented with a Direct Current (DC) generator as
exciter source [9], and according to Institute of Electrical and Electronics Engineers (IEEE),
few new ones are being built today [10]. In these kinds of systems, the regulation of excitation
was usually carried out on the field winding side of the exciter, and this resulted in slow sys-
tems [5]. As technology evolved, electronic and electromagnetic regulators were introduced,
and the power systems started to grow significantly and their functionalities increased. In-
stead of just keeping the voltage constant, they were now designed to improve the static and
dynamic stability. This resulted in a more stable power system. On the other hand, the way
the DC current is fed into the SM has changed little during the past century. There are some
exceptions, but most of the systems still feed the field current via brushes and slip-rings. In
some of the biggest SMs out there, hundreds of brushes need to be checked and maintained
[11].
1.2 Objective
The objective of this master dissertation is to design an excitation system for a SM using
Matlab ® /Simulink® . It will take a look on the excitation systems used in power plants, which
will provide substantial information surrounding the topic. A controlled rectifier will be con-
nected to the field winding of the SM, and a sufficient control system will be implemented.
• Implementation of a SM in Matlab® /Simulink® , to simulate the Terco MV1027-235

SM, stationed at Western Norway University of Applied Sciences.
• Design of AC/DC rectifier as exciter for the simulated SM.
• Design of two control systems for regulation of rectifier output, as well as controlling
the three-phase terminal voltage of the SM.
1.3. STRUCTURE 3
1.3 Structure
The thesis will be divided into six chapters as follows:
2 State of the art This chapter takes a look into the theory behind a syn-
chronous generator and its excitation system.
3 Method This chapter will dig further into the theory behind the
chosen excitation systems and the technology used to
solve the problem.
4 Results This chapter will present the simulated model.
5 Discussion This chapter will discuss the results obtained from the
simulated model.
6 Conclusion This chapter will be a conclusion of the work and will

present future work.
4
2 State of the art
This chapter will introduce some theories and existing technologies, evolving the different
parts of the system designed in this thesis. The chapter starts with an introduction of the SM.
Additionally, it will describe and compare different excitation systems and divide them from
each other structurally and transparently. Then a brief description of the control units for
the Automatic Voltage Regulator (AVR) follows. Many of the excitation systems use different
kinds of rectifiers, a device that converts Alternating Current (AC) current to DC current. The
theory is described in its own section. The chapter will end with an explanation about the
PID controller and how different parts of the controller work.
2.1 SM Theory
An electrical machine is a device that converts electrical energy into mechanical energy or
vice versa. The SM is the most common machine used as a generator to feed electrical energy
into the society [12], from different energy sources suchs as hydro, steam, and nuclear to
windmills. High reliability and cost-effectiveness are two keywords when describing this
machine. Synchronous in SM adresses the fact that the rotor, along with the magnetic field
created in the rotor, rotates at the same speed as the rotating magnetic field produced by the
armature currents in the stator [8].
Figure 2.1: Simplified schematic of a single-phase SM with salient poles.

2.1. SM THEORY 5
2.1.1 Construction
The SM consists of two main parts, the rotor, and the stator. The rotor is the rotating part
of the machine, and the stator is the stationary part. The voltage is generated in the arma-
ture winding, which is found on the stator. Further on, the rotor produces the main field
flux. Figure 2.1 shows a cross-section of a single-phase SM with two salient poles. The SM
would normally have more than two poles. With a frequency, f = 50Hz, the two-pole ma-
chine would have a speed of 3000 RPM. This can be calculated by equation 2.1 [13]. The
synchronous machine usually does not operate at these speeds [14].
2 · 60 · f
n= (2.1)
p
There are two kinds of rotors, the salient pole rotor and the round rotor, as shown in figure
2.2. In a salient pole rotor, the poles projects out of the rotor core. The rotor is made out of
steel lamination to reduce eddy current losses. The salient pole synchronous machine has a
non-uniform air gap. Salient pole rotors are typically used at low and medium speeds, and
they usually have a large number of poles. In hydroelectric power plants, the salient pole is
generally used [8]. In the round rotor, there are no projected poles. The poles in this type
of rotor are formed by the current flowing through the rotor excitation winding. The round
rotor is typically used at high speed, and they usually use two or four poles. This makes them
mechanically robust. Round rotors are often used in steam and gas power plants [8].
Figure 2.2: Simplified schematic of salient and cylindrical rotor.
A SM operates with two separate magnetic fields, the stator field, and the rotor field. These
two rotating magnetic fields will affect each other, and an armature reaction will occur. Fig-
ure 2.1 shows where the field winding is located in the machine. This winding is getting fed
by a DC current from the excitation system. This current is labeled as i f , and the function of
this current is to induce a magnetic flux in the rotor. Both the stator field and the rotor field
will rotate at synchronous speed.
6 CHAPTER 2. STATE OF THE ART
In a three-phase SM, there will be three armature windings, a, b and c. Figure 2.1 represents
a by a coil in a single slot, and its output is represented by a 0 . In real life, these windings will
be evenly distributed around the stator, and not only in single slots as the figure display [8].
The windings (a, b, and c) will be shifted 120◦ apart from each other, and they will produce
a voltage that resembles a sinusoidal waveform at the terminal.
Figure 2.3: Simplified schematic of damper winding.
Damper windings are often used in SM. Their function is to damp out any oscillations that
might occur. These windings are often included to help the SM to regain its synchronism
after a disturbance [14]. The damper windings have a significant effect during transient
conditions but in steady-state calculations the damper windings are usually ignored, hence
i q = i d = 0. In figure 2.3 a simplified schematic of the windings is shown.
2.1.2 Voltage Generation
The generation of voltage in a SM is based on Faraday’s law. This law states that the Electro-
motive Force (EMF) or the induced voltage is given by the rate of change of the magnetic flux.
The ε is given by equation 2.3. The voltage induced in a closed-loop inductor is proportional
to the total flux change through the inductor loop. Since this is true, the flux Φ will be equal
to the total flux linkages λ. The flux linkages are given by equation 2.3, where the inductor L
is the proportionally constant between the current and the flux linkages.
d Φ(t ) d Φ(t )
ε(t ) = − =− (2.2)
dt dt
Φ(t ) = Li (t ) (2.3)
Neglecting the leakage flux, the flux which does not cross the air gap, and considering only
the air-gap flux, the air-gap flux is only dependent on the field current i f and the stator cur-
rents i a , i b and i c [1]. To find the total air gap flux, an assumption that the magnetic circuit
is linear is made. This makes it possible to use superposition by considering the flux link-
ages due to i f alone(i a = i b = i c = 0), and then to i a , i b and i c alone(i f = 0). These two flux
2.1. SM THEORY 7
Figure 2.4: Air-gap flux due to i a .
linkage components will give two voltage components, the open-circuit voltage and the ar-
mature reaction voltage. These voltages will, in sum, give us the air-gap voltage due to the
total air-gap flux [1]. The open-circuit voltage is the voltage in the SM while no load is con-
nected and no current flowing through the stator coils. The armature voltage is the voltage
that is induced due to the current that flows in the stator coils. Three fluxes are defined [1].
λag is defined as the total air gap flux linkages of coil aa’, λaa 0 is defined as the total air gap
flux linkages of phase a, and λar is defined as the armature reaction flux linkages. The air
gap flux and the air gap voltage are written in phasor form in equation 2.4 and equation 2.6.
Equation 2.5 gives the armature reaction flux linkages in phasor form. These equations are
valid for a round rotor SM since the air gap is uniform(see figure 2.4).
λag = λaa 0 + λar (2.4)
λar = L s I a (2.5)
Vag = E a − j ω0 L s1 I a (2.6)
where E a is the open circuit voltage, Vag is the generated voltage, and L s1 is a fictitious in-
ductance. Accounting for the resistance, r and leakage reactance, X s , in the windings, the
terminal voltage, Va , for a round rotor machine is given:
Va = E a − r I a − j X s I a (2.7)
In a salient pole SM where the air gap is not uniform, see figure 2.2, another set of equa-
tions is needed. This means that the length of the air gap will change with the rotation of
the rotor, which makes the flux linkage vary. To handle this problem, the armature reac-
tion is decomposed along two axes, one along the direct axis of the rotor and another along
the quadrature axis, which is orthogonal to the direct axis. The direct axis components are
marked with subscripts letter d , and the quadrature axis components are marked with the
subscript letter q. This is called a park-transformation, which is a way of representing a set of
three sinusoidally-varying phase currents and voltages(a, b and c), using only two values (d
and q)[15] [16]. The full park transformation can be found in appendix A. The corresponding
flux linkages of coil aa 0 in phasor form is given by equation 2.8.
Λar = Λad + Λaq (2.8)
Λad = L d 1 I ad (2.9)
where L d 1 is the inductance and, I ad is the current, on fictitious winding d .
Λaq = L q1 I aq (2.10)
where L q1 is the inductance and, I aq is the current, on fictitious winding q.
The terminal phase voltage of phase a for a salient-pole SM is given by equation 2.11 and the
equivalent circuit are shown in figure 2.5.
Va = E a − r I a − j X d I ad − j X q I aq (2.11)
where X d and X q is the reactance on the direct and quadrature axis, respectively.
Figure 2.5: Equivalent circuit for a salient-pole rotor SM.

2.1. SM THEORY 9
2.1.3 Park Transformation
The park transformation represents a set of 3 sinusoidally-varying phase currents and volt-
ages using only two values. This method greatly simplifies the modeling of the synchronous
generators. In 1929 and 1933, R.H. Park published two articles considering the Park transfor-
mation. These articles proposed a method that converted the stator variables into equivalent
rotor variables [15], [16]. This was a whole new approach to electric machine analysis.
Figure 2.6: Schematic of a salient-pole SM, adapted from [1].
Figure 2.6 shows an three-phase synchronous machine with two fictitious windings, the di-
rect and quadrature axes. The voltages on the terminal and for the direct and quadrature
axis are shown in equation 2.13.
dλ
v = -Ri − (2.12)
dt
Where v is the voltage vector consisting of the terminal voltages (v a , v b , v c ), the field voltage(v F )
and the damper voltages(v D and vQ ). i is the corresponding current vector, R is a 4x4 matrix,
and λ is the flux linkages. The equation can be written as:
      
va r ia λaa 0
      
 v   r i  λ 0 
 b    b   bb 
      
 v   r i  d λ 0 
 c    c   cc 
 =  −   (2.13)
 −VF   rF  iF  d t  λF F 
      
      
−VD   rD  i D  λDD 
      
−VQ rQ iQ λQQ
Where λaa 0 , λbb 0 , λcc 0 , λF F , λDD 0 , λQQ 0 is the flux linkages. Where the variables with lower
case letters is the stator variable, and the variable with upper case letters is the rotor vari-
able. By using equation 2.3, equation 2.14 represents the flux linkages. The majority of the
inductances shown in this equation are dependent on the placement of the rotor, and thus
they will change with time. All of the equations for these are shown in appendix A.
    
λaa 0 L aa L ab L ac L aF L aD L aQ ia
    
λ 0   L L bQ   
 bb   ba L bb L bc L bF L bD   ib 
    
λ 0  L L cQ   
 cc   ca L cb L cc L cF L cD   ic 
 =   (2.14)
 λF F   L F a L F b L F c LF F LF D L FQ   
    iF 
    
λDD  L D a L Db L Dc L DF L DD L DQ   
    i D 
λQQ L Qa L Qb L Qc L QF L QD L QQ iQ
Where i a , i b , i c , i F , i D and i Q is the current flowing in each winding. Applying the park trans-
formation on equation 2.14 (the transform can be found in appendix A), the time-dependent
variables are eliminated and the equation can be rewritten to equation 2.15. The matrix now
consists of only constant variables.
    
λ0 L0 0 0 0 0 0 ia
    
λ   0 Ld 0 kM F kM D 0   
 d    ib 
    
λ   0 0 Lq 0 0 kMQ   
 q    ic 
 =   (2.15)
 λF   0 kM F 0 LD LF D 0   
    iF 
    
λD   0 kM D 0 MR LF 0   
    i D 
λQ 0 0 kMQ 0 0 LQ iQ
Where λ0 , λd , λq , λF , λD , λQ is the flux linkage of each winding. L 0 , L d , Lq, L D , L F , L Q is

the self inductance of each winding. M F , M D , MQ , M R represents the mutual inductance
between the windings [17]. Windings are shown in figure 2.6.
2.2. EXCITATION SYSTEM 11
2.2 Excitation System

The SM is dependent on a magnetic field to be able to generate power. The excitation system
does this, by feeding DC current into the field winding of the machine. The system works as
a feedback control system and is usually operated with various kinds of PID controllers. As
technology evolves, it has gotten more and more functions in the power system. Today the
excitation system controls the voltage, the field current, the power factor, and the reactive
power. It also functions as a protection of the SM and for the excitation system itself [18] [5].
Hence the excitation system carries out both control and protective functions.
The excitation system is used to describe an entire system usually consisting of an AVR, ex-
citer, measuring elements, power system stabilizer, excitation limiters, and protection units.
See figure 2.7. The exciter is the source of the excitation current, and this source varies from
system to system. The source can be realized by a DC generator, which feeds the main SM
directly, or it can be an AC generator which produces a three-phase current which is then
rectified by the use of diodes, thyristor or transistor bridge [19]. There are also self-excited
systems that get their excitation current from the main generator itself, based on controlled
rectifiers [9].
Figure 2.7: Block diagram of full excitation system, consisting of, exciter, control units and
SM [10].
This thesis will classify the systems based on the excitation power source, in the same man-
ner as IEEE Std 421.5TM IEEE:
• DC Excitation System (Type DC) - DC generator as the source of excitation power.
• AC Excitation System (Type AC) - AC generator as the source of excitation power.

• Static Excitation System (Type ST) - Excitation power is supplied via transformer and
rectifiers.
2.2.1 Type DC
This kind of system uses a DC generator as the source of excitation power. The output of the
exciter is fed into the main SM through slip rings [20], [7]. In large systems where the main
generator produces a lot of power, a pilot exciter might be needed for the exciter. In these
cases, the system will have one main exciter, which provides the excitation current to the
main SM and the pilot exciter, a self-excited machine that provides field current to the main
exciter. In some cases, the exciter is self-excited. Many systems in service today still use this
excitation system, but few new ones are being built [10]. The area for this kind of system was
between the 1920s and 1960s, and by the mid-60s they were more and more replaced by the
Type AC systems [7].
A simplified block-diagram of a Type DC system is represented in figure 2.8. It shows the

main exciter as a DC generator which supplies the DC current into the field winding of the
controlled SM. In this example, an amplidyne is feeding the DC generator with field current.
An amplidyne is the most common form of the metadyne, which is an electrical machine
with two pairs of brushes. The device got high amplification and fast speed of response [21].
The amplidyne provides the exciter field, in a so called "buck-boost" scheme, and the rest of
the required field is made by self-excitation [7]. If the amplidyne in figure 2.8 is out of service,
the system can adjust the exciter field manually through the rheostat.
Figure 2.8: Simplified schematics of system with amplidyne. Adapted from [7].
2.2.2 Type AC
In this system the source of excitation power comes from an AC machine with rectifiers con-
nected to the output of the exciter. The rectifiers can either be stationary or rotating. The
rectifiers converts the AC to DC and they can be both controlled or non-controlled, depend-
2.2. EXCITATION SYSTEM 13
ing on the system. This system can be built in various forms, depending on rectifiers used,
source of excitation for the exciter, and the applied methodology of control [7]. It is common
to divide these system into two groups; stationary rectifiers and rotating rectifiers.
Stationary rectifiers
The systems with stationary rectifiers are in need of sliprings and brushes for the delivery
of excitation current to the generator. If the system is to use non-controlled rectifiers (e.g.
diodes), a regulator is used to control the field of the AC-exciter [7]. An example of a system
like this is represented in figure 2.9. In this example, the exciter is self-excited by the means of
controlled-rectifiers (e.g. thyristors), and the three-phase output of the exciter is rectified by
the non-controlled bridge. Later on the DC is then fed into the field winding of the controlled
SM through brushes and slip rings.
Another example of a stationary Type AC system is shown in figure 2.10. In this system the
exciter is no longer self-excited, but the exciter is in need of an pilot-exciter. The pilot-exciter
is self-excited through controlled-rectifiers and controlled by the regulator. As shown in the
figure, the AC exciter is feeding its power to another set of controlled-rectifier, which then
convert the three-phase AC into DC. The most common controlled-rectifiers are based on
thyristors, because of their high power handling capabilities [22]. The regulators in figure
2.10 control the DC directly through the thyristor-bridge.
Figure 2.9: Field-controlled alternator with non-controlled rectifiers, adapted from [7].
Figure 2.10: Alternator-supplied controlled-rectifier system, adapted from [7].
Rotating rectifiers
In the manner of rotating rectifiers, there are no need of brushes [7], [23]. This system is most
often referred to as a brushless excitation system. These brushless systems are often used in
smaller hydro-power plants, maritime facilities, and in industrial systems [23]. They still use
an AC generator as source of excitation, and this power is then rectified to a DC current. The
DC current is then fed directly into the field winding of the controlled SM. The name comes
from the fact that the rectifier along with the armature of the exciter, rotates, see figure 2.11.
This system requires a pilot-exciter, which feeds the AC exciter, in figure 2.11 a Permanent
Magnet Synchronous Generator (PMSG) is uses as a pilot exciter.
Figure 2.11: Simplified schematics of brushless system, adapted from [7].
This system was originally developed to avoid the use of brushes and sliprings. Nevertheless,
it seems like the brushless Type AC performs equally well as the Type AC systems that uses
brushes [7].
2.3. RECTIFICATION 15
2.2.3 Type ST
The static excitation system is one of the most typical excitation system in big hydro-power
plants [23]. The name of this system comes from the fact that all the components are static or
stationary. Common for these systems is that the excitation power is fed either from the main
terminals of the controlled SM or from the station auxiliary bus. This system can in other
words be described as a self-excited system. This power is fed through a transformer which
then feeds either a controlled, or a non-controlled rectifier which regulates the voltage. Since
this is a self-excited system, the system needs an additional source of power to get it started.
The SM cannot produce any voltage without any field current. This process of build-up for
the generator field-flux is termed field flashing [24], [7].
Figure 2.12 shows an example of a Type ST system. The power for the exciter is taken from
either the terminals of the main SM itself, or the station auxiliary bus. An excitation trans-
former is connected to the output of the terminal and steps down the voltage and feeds the
AC voltage into a controlled-rectifier. This system is also known as a bus-fed or transformer-
fed excitation system. The controlled-rectifier is then transforming the AC current to DC
current and feeds the field winding of the main SM via brushes. This system is both inexpen-
sive and it is easily maintainable[7].
Figure 2.12: Simplified schematics of system, adapted from [7].
2.3 Rectification
Rectifiers are widely used in practical applications, such as computers and battery charg-
ers. [25]. Both Type AC and Type ST systems use rectifiers in their excitation systems. In the
excitation systems out there, the three-phase bridge is the most common to use and has cur-
rently become the only selected one [5]. The function of a rectifier is to convert AC current to
DC current. If the operation is reversed the device is called an inverter. Rectifiers are made
possible by the unique features of semiconducting devices. In simple terms, a semiconduct-
ing device can pass the current in one direction and block it or slow it down in the opposite
direction. They can also be controlled or uncontrolled, depending on the chosen semicon-
ductor. In this subsection both the single- and three-phase rectifier, based on diodes, will be
discussed.
2.3.1 Single-Phase Uncontrolled Rectifier
The circuit in figure 2.13 represents an uncontrolled single-phase full wave rectifier. This is
an uncontrolled device, since it uses the simplest form of semiconductors, four diodes. The
diode cannot be controlled, and in simple terms it can conduct in one direction (forward
biased) and block in the opposite direction (reverse biased).
The uncontrolled single-phase full-wave rectifier rectifies the full sinusoidal AC current and
converts it into a DC current. The bridge configuration makes it possible to convert both the
positive and negative half cycle of the sinusoidal wave. In the circuit two of the four diodes
will always be forward bias, and two of them will always be reverse biased. When a diode is
forward bias, it conducts current, and when its reverse biased it blocks the current.
Figure 2.13 shows how the current flow in a single-phase rectifier. The left part of the figure
represents the positive half cycle of the sinusoidal AC. In this state D1 and D4 are forward bi-
ased, and D2 and D3 are reversed biased. The current will then flow from the source, through
D1 , R L , D4 and back to the source. The right part represents the negative half cycle. In this
state, D1 and D4 turn reverse biased and block the current, while D2 and D3 switching to
forward bias. Here the current will flow from the source, through D3 , R L , D2 and back to the
source [26].
Figure 2.13: Single Phase Rectifier and its current flow.
Figure 2.14 represents the waveform of both the input and the output voltage. The sinusoidal
black curve represents the input voltage, Vi , and the below curve represents the output,Vo .
As shown in this figure, the output voltage will not be a pure straight line, as we might think
2.3. RECTIFICATION 17
of a DC. The voltage from the rectifier will have some ripples as a result of the alternating
supply. These ripples can be modified by the use of different kind of snubber circuits.
Figure 2.14: Input and output voltage curves for the Single Phase Rectifier.
The bridge above represents an ideal circuit. The average voltage of an ideal single-phase
rectifier is given by equation by:
Z T /2 p
1 1 p 2p
Vo = 2Vi si nωt d t = ( 2Vi cosωt ) = 2Vi (2.16)
T /2 0 ωT /2 π
2p
Vo = 2Vi = 0.9Vi (2.17)
π
Where Vi is the AC input measured in rms, ω = 2π f and ωT /2 = π [22] [26].
2.3.2 Three-Phase Rectifier
The three phase rectifier is based on the same principle as the single phase one. This rectifier
is widely used in practice [27]. These voltages can be seen on like three phases that are 120
electrical degrees ahead of each other. The advantages here is that the output voltage will
have less ripple in comparison to a single-phase input, see figure 2.15.
The rectifier which will be described in this section is also an uncontrolled-rectifier made
out of diodes. In the same manner as the single-phasen it can be made out of transistors and
thyristors, and it will then require some sort of control to function. This circuit is widely used
in both the Type AC and Type ST excitation systems [5].
Figure 2.15 shows the full bridge consisting of six diodes, which must be considered to be an
idealized circuit. To get to the average output DC only one of the segments in figure 2.16 is
considered, that means a period of 360°/6 = 60° is set to be the interval. To get the average
voltage for an idealized circuit represented in equation 2.18 to 2.20 [22]. VLL is the rms value
Figure 2.15: Three Phase Rectifier.
Figure 2.16: Three Phase Rectifier.
on the line-to-line voltage and A is the area under the chosen segment.
p 1 1
v o = v ab = 2VLL cos ωt − π < ωt < π (2.18)
6 6
Z π/6 p p
A= 2VLL cos ωt d t (ωt ) = 2VLL (2.19)
−π/6
Z π/6
p
1 3p
Vo = 2VLL cos ωt d (ωt ) = 2VL L = 1.35VLL (2.20)
π/3 −π/6 π
Capacitor filter, C and source inductor, L s
In practical rectifiers a capacitor filter is often used for smoothing the output voltage. The
rectified voltage will energize the capacitor as long as the input voltage is superior to the
capacitor voltage. The capacitor will charge until it reaches the peak value for the input
voltage, and discharge when the AC voltage falls under the capacitor voltage. This will result
in a more stable output for the system.
2.4. CONTROL UNITS 19
Figure 2.17: Illustration of how the capacitor smooth out the rectified voltage.
Figure 2.17 shows how the capacitor filter changes the output of the bridge. The stippled
lines represents the rectified waveform without the filter, and the black line represents the
capacitor voltage. When the voltage is rising the capacitor is charged up and when the volt-
age decreases the capacitor is decharging. By increasing the capacitor value the output volt-
age will be more and more similar to a DC current.
To get a understanding of why a source inductor is needed a simplified circuit is shown in

figure 2.18.
Figure 2.18: Single-phase bridge with source inductor L s .
When a diode bridge, described in section 2.3, is running without any source inductance, an
assumption has been stated that the currents transfer from one diode to another immedi-
ately. This is not true for practical rectifiers with a AC source that got some inductance L s
to it. The change of current must now take some time to happen. This is called commuta-
tion or overlap process, and this forces all the diodes in the circuit to conduct at the same
time. When all the diodes are conducting the output voltage will be zero and all of the input
voltage will apply across L s .
2.4 Control Units
2.4.1 AQR
As the name suggest, the AQR is applied to regulate the reactive power of the controlled
unit. The use of both AQR and APFR, has its origin in industrial applications of synchronous
motors and generators. Large generators connected to the bulk power systems, the use of
both AQR and APFR are usually forbidden according to constrains [10]. The expression of
the AQR is given in equation 2.21 [5]. a,b and c are coefficients determined on the basis of
the operational requirements. P and Q is the active and reactive power, respectively.
Q = a + bP + cP 2 (2.21)
Figure 2.19: AQR Transfer function.
Table 2.1: Parameters for figure 2.19.
Symbol Meaning Symbol Meaning
T (s) Rate of changes for the voltage Q LI M (p.u.) AQR output

setting device
QD Deadband zone ±UL (p.u.) Integral limit
2.4.2 APFR
The APFR is a system which is used to control the power factor of the synchronous generator.
This could be to keep the power factor at a set value or within a set range. The transfer
function model is shown in figure 2.20, the output of the block is going to the AVR.
Figure 2.20: Transfer function.

2.4. CONTROL UNITS 21
T (s) Rate of change for the voltage P F LI M (p.u.) APFR output

setting device
P FD Deadband zone ±UL (p.u.) Integral limit
2.4.3 UEL
The UEL is used to protect the generator from losing synchronism and to prevent operation
that could lead to overheating in the stator region of the synchronous machine [10]. There
are different kind of UELs, where some uses a combination of voltage and current, and some
uses a combination of active and reactive power. -The linear expression of the UEL is ex-
pressed in equation 2.22 . In this equation, a and b are parameters determined by the limit
characteristic [5]. For the linear limit characteristics, the transfer function model is shown in
figure 2.21. The output of this block is sent to the AVR.
Q = a + bP (2.22)
Figure 2.21: UEL Transfer Function Model: Q = a + bP linear characteristics.
Table 2.3: Parameters for figure 2.21
TR1 (s) Time constant for the P trans- KU E L UEL gain

ducer
TR2 (s) Time constant for the Q trans- TL (s) Time constant for the leading
ducer phase compensation
U E max Upper output limit TG (s) Time constant for the lagging
phase compensation
U E mi n Lower output limit

There are different kinds of UELs but these will not be discussed here.
2.4.4 OEL
OELs are often referred to as maximum excitation limiters and field current limiters [10]. The
OEL is used as protection to prevent the generator rotor excitation winding from overheating
[5]. The detection signal varies with different excitation modes. Generator excitation current
and voltage, and the exciter excitation current are included in these excitation modes. An
example of an OEL model is shown in figure 2.22. According to IEEE Std421.5, OEL modeling
should not be required in most system studies [10].
Figure 2.22: OEl example model.
TR (s) Time constant for the trans- a 1 (p.u.) Return difference value for
ducer operation reset
r OE L (p.u.) Limited target value TLI M (p.u.) Integral limit value
r OE D (p.u.) Initial value for OEL operation Q E LI M (p.u.) Max.limit value
ts Timing value for OEL opera- TG (s) Time constant for the lagging
tion phase compensation
K OE L OEL gain TOE L Integral time constant
TL (s) Time constant for the leading d1 Min. detection value for guar-
phase compensation anteed operation
2.4.5 PSS
The PSS are used to enhance damping of power system oscillations. This is done through
excitation control. The most common used inputs are shaft speed, terminal frequency and
2.5. BASIC CONTROL LAW 23
power [10]. PSS can then be divided into three types. ∆P -type which uses the power as
input, ∆ω-type which uses the shaft speed and ∆ f -type which uses the frequency as input.
The transfer function model of an example of a ∆P -type is shown in figure 2.23 [5].
Figure 2.23: ∆P Type Transfer Function Model.
TP (s) Time constant for the active T3 (s) Time constant for the leading
power P detector phase compensation
TR (s) Time constant for the isola- T4 (s) Time constant for the lagging
tion filter phase compensation
T1 (s) Time constant for the leading K P SS PSS gain

phase compensation
T2 (s) Time constant for the lagging UP L (p.u.) PSS output limit
phase compensation
2.5 Basic Control Law

The basic control law of the excitation controller has three factors controlling the system [5].
The proportional (P), the integral (I), the derivative (D), and the combination of these. Com-
bination of these would result in the Proportional Integral (PI)-, Proportional Differential
(PD)- and the PID - controllers. PID controller comes with several advantages: ease of use,
stable and reliable and simple structure. These controllers are used in all kinds of industrial
control systems [28]. The PID controller is the most common control used in excitation sys-
tems [29]. Each letter in the name got their own function, and if one function is not needed
in the PID controller it can simply be set to zero.
2.5.1 P
A proportional controller is based on a response in proportion to the difference between

the desired value and the current value of the variable. This kind of control system is used
in many industrial settings and it is also used in some smart devices. Proportional control
is a feedback control system, where correction is applied to the controlled variable which
is proportional to the difference between the set value. The proportional term is given by
equation 2.23. The tansfer function of the proportional controller is given by equation 2.24.
u(t ) = K p e(t ) (2.23)
U (s)
G c (s) = = Kp (2.24)
E (s)
Where U (s) is the Laplace transform of the output, E (s) is the laplace transform of the input
and K p is the proportional gain of the controller [5]. By increasing K p , the sensitivity of the
system increases, and by that the system will respond faster compared to a lower K p . This
does not mean that the higher K p the better, because if it gets too high values the system will
become unstable [28].
2.5.2 I
The integral term I, is proportional to the magnitude and the duration of the error signal. It
is the sum of the instantaneous error over time and gives the offset that should have been
corrected previously. The error is then multiplied by the integral gain K i and then added
to the output of the system. The function of adding the integral term is to accelerate the
movement of the process towards set-point and the term eliminates residual steady-state
error that occurs with a pure proportional controller. The integral controller can however
cause the present value to overshoot the set-point value. Too high integral gain will also here
lead to an unstable system [28]. The integral term is shown in equation 2.25. The transfer
function is expressed in equation 2.26 [5]. The disadvantages of the pure integral controller
is that it cannot keep the output u and the input e synchronized and quick to respond. The
output changes will always lag behind the deviation changes [5].
Z t
u(t ) = K I e(τ)d τ (2.25)
0
U (s) K i
G c (s) = = (2.26)
E (s) s
2.5.3 D
Some controlled systems require a more advanced regulation. These objects are usually
adopted according to the variation trend of the controlled variable to avoid generating much
2.5. BASIC CONTROL LAW 25
greater deviation during the regulation [5]. This leads to the differential control law which is
expressed by equation 2.27. The controller output u(t ) and the rate of change in the input
deviation d e(t )/d t are in proportion, and the proportional factor TD is the differential time
constant. The differential controller cannot be used as a standalone control system, it is al-
ways combined with the proportional controller as a PD- or a PID-controller. This controller
is able to predict the systems behaviour and thus improve settling time and the stability of
the system. The transfer function is expressed in equation 2.28:
d
u(t ) = TD e(t ) (2.27)
dt
G c (s) = TD s (2.28)
2.5.4 Proportional-Integral (PI)
The PI-controller is a combination of the proportional and the integral controller. This con-
troller combines the quick response of the proportional controller, and the elimination of
steady-state error by the integral controller. Equation 2.29 expresses the control law of the
PI-controller, while equation 2.30 expresses the transfer function. The proportional term is
Rt
represented by K P e(t ) and the integral term is represented by KTPI 0 e(t )d t . T I is the integra-
tion time constant [5].
· Z ¸
1 t
u(t ) = K P e(t ) + e(t )d t (2.29)
TI 0
· ¸
U (s) 1
G c (s) = = KP 1 + (2.30)
E (s) TI s
2.5.5 Proportional-Differential (PD)
The PD-controller is the sum of the proportional and the differential controllers. A PD-
controller is often adopted for an excitation control system [5]. The control law for the PD-
controller is expressed in equation 2.31, while the transfer function is expressed in equation
2.32. The differential action in the system will always try to prevent the controlled variable to
change, this will improve the system stability. At the same time if the differential time con-
stant TD is too large, the differential action will be too strong, which will influence the system
stability.
· ¸
d e(t )
u(t ) = K P e(t ) + TD (2.31)
dt
TD s + 1
G c (s) = K P TD
(2.32)
KD
s +1
2.5.6 Proportional-Integral-Differential (PID)
The PID controller is a combination of proportional-, integral- and derivative-control. It is

most often used as a feedback controller, where it delivers the output at desired levels. If one
want to use a proportional controller, or a PI-controller, one simply set the other parameters
to zero. The expression of the transfer function of the PID control is as follows:
U 1
= (K p + K D s) (2.33)
∆Ui 1 + KI s
Where K p , K i and K d are the proportional, integral and differential regulation factors, re-
spectively.
Figure 2.24: Block diagram of transfer function of a PID control.

27
3 Method
In this chapter, an excitation system for a SM is designed. The system will be based on a
single-phase controlled rectifier, built up by thyristors, often termed SCR [22]. The SCR was
chosen because of the need of a controllable rectifier, and its wide use in excitation systems
[19]. The source of excitation power will be the single-phase AC power in the socket on the
wall.
Matlab ® /Simulink® is a commercial tool, which provides convenient solutions for system
simulation technology [30]. Simulation studies comes with a couple of challenges. It can be
hard to balance the system, so that it accounts for all physical phenomena that would occur
in a physical model. Also systems which embrace multiple physical domains can come with
some challenges. Matlab ® /Simulink® along with Simscape ElectricalTM , library used for all
components, is both adressing these needs [31], [30].
3.1 Full ® /Simulink® Model
Figure 3.1 display a sketch of the system designed in Matlab ® /Simulink® . Full simulation is
found in appendix B. The simulated model consists of a SM, excited by a controllable recti-
fier. Two control systems, are implemented for controlling the rectifier output, as well as the
terminal voltage. The system consists of the following elements:
SM. Modelled to behave as the Terco MV 1027-235 SM.

SCR bridge. Single-phase controlled rectifier.
SCR control. Controls the rectified DC voltage.
SM control. Controls the terminal voltage of the SM.
28 CHAPTER 3. METHOD
Figure 3.1: Sketch of Matlab ® /Simulink® Model.
3.2 Synchronous Generator

The SM SI Fundamental block, from Simscape ElectricalTM , is used to simulate the generator
of the system. It models a three-phase machine, with either round rotor or salient poles. The
block uses parameters in SI units and can be driven as a motor or generator, depending on
the mechanical input. The full block can be found in [32].
The machine is suppose to behave as the Terco MV1027-235 SM stationed at University of

Bergen and Western Norway University. The datasheet for the Terco MV1027-235 can be
found in appendix D.1. This is a small generator with a nominal power of 1.2 kVA, and the
excitation source is set to 230 V AC Root Mean Square (RMS). A SCR-bridge is connected to
the field winding of the machine, and will rectify the AC current into a controlled DC current.
The parameters used in Matlab ® /Simulink® is displayed in table 3.1.
Table 3.1: Parameters in Matlab ® /Simulink® for the SM.
Parameters Value Description Obtained from

PN 1.2 kVA Nominal power. From ap. D.1
VN 230 V Nominal voltage From ap. D.1
fN 50 Hz Nominal frequency From ap.D.1
Rf 129.5Ω Field resistance. From [33].
Lf 7.7 H Field inductance. From [33].
Rs 1.4Ω Stator resistance. From [33].
L ssm 25 mH Stator inductance. From [33].
L md 1.2 H Direct-axis mag. inductance. Estimated.
L mq 1.2 H Quadrature-axis mag. inductance. Estimated.
3.3. SCR BRIDGE 29
As the table displays, the parameters are collected both from the datasheet for the Terco
MV1027-235 SM, previous tests on the machine and some are estimated.
3.3 SCR Bridge

The thyristor block, from Simscape ElectricalTM , is used to model the SCR. It represents a
macro model of the real SCR, and does not account for the geometry or the complex physical
processes. The macro model is simulated as a resistor, R on , inductor, L on , DC voltage source,
V f , and a switch connected in series.
Figure 3.2: Thyristor block: a) Equivalen circuit b)VI characteristic.
Figure 3.2 a) displays the inside of the block. The block turns on via its gate, when a positive
pulse signal is applied, given Vak > V f , and that the pulse last long enough so that I ak be-
comes greater than I l . It turns off when I ak = 0 A, and a negative voltage, Vak last equal or
longer than the turn off time T q .
A SCR is found for the system, so that practical parameters are implemented in the model.
The parameters can be found in the datasheet, see appendix ??
Table 3.2: Parameters for SCR(Adapted from datasheet.)
Parameters Value Comment

R on 0.001Ω Internal resistance.
L on 0H Internal inductance.
Vf 0.8 V Forward voltage.
Il 80 · 10−3 H Latching current.
−6
Tq 100 · 10 s Turn-off time.
Ic 0A Initial current.
Rs 500Ω Snubber resistance.
−9
Cs 250 · 10 F Snubber capacitance.
The bridge configuration and a pre-charge system is displayed in figure 3.3. The four SCRs
work parallel to each other:
SCR1 and SCR4 . Rectifies the positive sinusoidal input voltage. Gate A.
SCR2 and SCR3 . Rectifies the negative sinusoidal input voltage. Gate B.
Figure 3.3: Topology of SCR bridge and pre-charge system.
Table 3.12 display the parameters for the different components in the bridge. Since the
bridge is suppose to feed the field winding of the SM, the load resistor, R L and load induc-
tor, L L is set equal to the field resistance, R f and field inductor L f , respectively. The voltage
over R L and L L , represents the output voltage of the bridge. In Matlab ® /Simulink® this volt-
age signal is measured, and sent to the SM. The current through this RL - load, has been
measured and compared to the field current in the SM.
Table 3.3: Parameters for SCR bridge.

AC source 230 V AC 50 Hz RMS value of input voltage.
Ls 1 mH Source inductor.
C 50 mF Smoothing capacitor.
RL 129.5 Ω Field resistance.
LL 7.7 H Field inductance.
3.4 Control Systems

Figure 3.6 display both the SCR control, and the terminal control, and the switching mecha-
nism between them. They are both based on phase control for the SCRs. Phase control work
by sending small pulses of current, to trigger the gates of the SCRs, and will stop conducting
when the input voltage falls below the holding current, a term called natural commutation
[34] [22].
3.4. CONTROL SYSTEMS 31
Figure 3.4: Sketch of full control system.
Table 3.4, show the parameters for the two pulse generators. The output of these blocks
simulates the gate current which trigger the SCR pairs. The amplitude is set to one, and the
period, T is calculated:
1 1
f = → T = = 0.02 (3.1)
T f
Table 3.4: Parameters for pulse generator A
Pulse generator 1 Pulse generator 2

Amplitude 1 Amplitude 1
Period 0.02 s Period 0.02 s
Pulse width 30 % Pulse width 30 %
Phase delay 0s Phase delay 0.01 s
The pulse width, tell us how long the signal lasts in % of the period. Both pulse generators, are
firing one pair of SCR each. The only difference between the pulse generators, is the phase
delay for the second one. These pulses are delayed by 0.01 s, which makes the second pair of
SCRs, always lag behind by half a period and thus only conduct for the negative sinusoidal
input voltage, and contrary the other pair will only conduct for the positive.
Figure 3.5: Sketch of: a) Input voltage b)Gate current
Figure 3.5, show that by delaying the gate current, the SCRs can be fired at desired angles of
the input voltage. This angle is called the firing angle, α. By controlling the firing angle, the
output is controlled.
3.4.1 SCR Control
The SCR bridge is controlled by a PI controller. The reason for not adding the derivative gain
in the system, is from the fact that there were no significant difference whether it was added
or not. Matlab® /Simulink® offers it own way of tuning the PI regulator. It is a model-based
tuning method, which computes a linear model of the plant from the output and the input.
This plant includes all of the blocks in the control loop, besides from the PI controller itself.
Further on the tuner accumulate all of the input/output data by simulating the model over a
short period of time, and both P and I is calculated. The parameters for the PI controller are
displayed in table 3.5.
Figure 3.6: Sketch: Open VS Closed Loop for SCR control.

3.4. CONTROL SYSTEMS 33
Parameters Value
P −0.08833
I −0.07221
3.4.2 Terminal Control
The terminal control system, controls the three-phase output of the SM. Figure 3.7 show the
block diagram of the system. The process represents the whole model. The amplitude of the
three-phase output voltage, is measured and compared to the reference.
Figure 3.7: Sketch: Open VS Closed Loop for Terminal Control.
The PI parameters are displayed in table 3.6.
Parameters Value
P −0.0264
I −0.00839
3.4.3 Pre-charge Control
The control of the pre-charge system is shown in 3.8. It is based on logic operators, and is
made so that it will charge the capacitor to a desired value, before the bridge start its rec-
tification. No pre-charge for the capacitor can lead to large inrush current, and by that the
system will act as a protective circuit.
Figure 3.8: Control for the precharge system.
If the bridge is set to start at a desired output voltage, the DC source is set equal, so that the
system will charge the capacitor before the system starts running. In theory, the SCR bridge
could run without a pre-charge, since it is in fact controllable. On the other hand, from a
practical perspective, pre-charge of the capacitor should be implemented.

DC source V DC Set to the desired capacitor voltage.
R PC 1 mH Source inductor.
C 50 mF Smoothing capacitor.
RL 129.5 Ω Field resistance.
LL 7.7 H Field inductance.
3.5 Step Response SCR Bridge

A dynamic response is done on the SCR bridge, to find the 1.order transfer function.
64
Output Voltage [V]
62
60
58
56
9.5 10 10.5 11 11.5 12 12.5 13 13.5 14 14.5 15 15.5 16

Time [s]
Figure 3.9: Step response for SCR bridge.

3.6. STEP RESPONSE FULL PROCESS 35
Table 3.8: Dynamic response on the terminals.
Output voltage Time Delay

Vo = 56.05 V t 1 = 10.004 s D 0 = 0.009
Vo f = 63.91 V t 2 = 14.022 s D f = 0.0089
∆Vo = 7.86 V ∆D = −1.0 · 10−4
The gain and time coefficient for the 1.order system, is calculated in equation 3.2 and 3.3
respectively.
∆Vo
K= = −1.2723 · 10−5 (3.2)
∆α
Vt 63% = Vo + 63% · ∆Vo = 61.0018V → T = 2.9540 s (3.3)
The first order transfer function for the SCR bridge is expressed in equation 3.7.
K −1.272 · 10−5
G aut = = (3.4)
Ts +1 2.954s + 1
3.6 Step Response Full Process

A dynamic response is done for the terminal control system, to obtain the transfer function
of the full process. The input delay variable was set to a value that corresponded to 200 V
on the terminal. The delay were then applied a step so that the output rised to a new value.
Figure 3.10 displays the response on the terminals. Full calculations are shown in appendix
(WILL AD).
215
Terminal Voltage [V]
210
205
11 11.5 12 12.5 13 13.5 14 14.5 15 15.5 16 16.5 17 17.5 18

Time [s]
Figure 3.10: Dynamic response test to find the 1.order transfer function for the system.
Table 3.9: Dynamic response on the terminals.
Terminal Voltage Time Delay

Vt 0 = 203.1 V t 1 = 12.734 s D 0 = 0.0086
Vt f = 214.2 V t 2 = 16.677 s D f = 0.00855
∆Vt = 11.1 V D = −5 · 10−5
The gain and time coefficient for the 1.order system, is calculated in equation 3.5 and 3.6
respectively.
∆Vt
K= = −4.5045 · 10−6 (3.5)
∆D
Vt 63% = Vt 0 + 63% · ∆Vt = 210.093V → T = 2.667 s (3.6)
The first order transfer function for the automatic control system is found in equation 3.7.
K −4.5045 · 10−6
G aut = = (3.7)
Ts +1 2.667s + 1
3.6.1 Additional PI-Tuning for Terminal Control
The Ziegler-Nichols and Cohen-Coon tuning methods will also be tested on the automatic
regulator, and compared against the regulator tuned in Matlab® /Simulink® . These methods
are based on the reaction curve found by the open loop experiment. The input that controls
the system, is forced to a desired value, and applied a step. The output is measured and the
reaction curve can be obtained. The PI parameters can then be calculated from the tables
found in appendix C.
Table 3.10: Readings for figure 3.11
Terminal Voltage Time Delay

Vt 0 = 203.1 V t 0 = 12.61 s D 0 = 0.0086 s
Vt f = 214.2 V t 1 = 12.84 s D 0 f = 0.00855 s
∆Vt = 11.1 V t 2 = 14.25 s ∆D 0 = −5 · 10−5 s
3.6. STEP RESPONSE FULL PROCESS 37
216
214
212
210
208
206
t0 t1 t2
204
202
200
11.5 12 12.5 13 13.5 14 14.5 15 15.5 16 16.5
Time [s]
Figure 3.11: Reaction Curve.
0.9v 0
K p Z i eg l er −N i chol s = = 2.3842 Tr Z i eg l er −N i chol s = τ0 = 0.273 (3.8)
K 0 τ0
Table 3.11: Parameters for Ziegler-Nichols tuned regulator.
Parameters Value
P −2.4853 · 10−5
I −3.6019 · 10−5
Table 3.12: Parameters for Cohen-Coon tuned regulator.
Parameters Value
P −2.5228 · 10−5
I −4.4115 · 10−5
38
4 Results
The tests have been done while the generator has been in an open-circuit condition. This
condition has been simulated by a very low load, so that approximately no current was
drawn. The mechanical input has been set at a constant at 1500 RPM, so that the terminals
produce 50 Hz. The results will be presented in the same order, as the method has presented
the system.
4.1 SM Input/Output
This section shows the mechanical input, the field voltage and the terminal voltages, shown
in figure 4.1, 4.2 and 4.3, respectively. The period in each sinusoidal voltage was measured
to 20 ms which proves that the frequency on the terminal is 50 Hz, see 3.1. This is a result of
the machine having 2 poles and the mechanical input set to 1500RMS, see equation 2.1.
4.1. SM INPUT/OUTPUT 39
1,600
¤
Mechanical Input rpm
1,550
£
1,500
1,450
1,400
20 20.01 20.01 20.02 20.02 20.03 20.03 20.04 20.04 20.05 20.05
Time [s]
Figure 4.1: Constant mechanical input at 1500 RPM.
100
Field Voltage [V]
95
90
85
80
20 20.01 20.01 20.02 20.02 20.03 20.03 20.04 20.04 20.05 20.05
Time [s]
Figure 4.2: Field voltage set at a reference of 90 V.
200
Terminal Voltage [Vabc]
100
−100
−200
20 20.01 20.01 20.02 20.02 20.03 20.03 20.04 20.04 20.05 20.05
Time [s]
Figure 4.3: Va ,Vb and Vc as a result of the mechanical input and the field voltage.
40 CHAPTER 4. RESULTS
4.2 Relationship between i f and Vt

Figure 4.4 shows the relation between the field current and the peak value of the terminal
voltage. The designed system is simulated with no saturation. The stippled line in the fig-
ure display the magnitude of the field-current, measured to 0.673 A, when the peak of the
terminal voltage is 200 V.
250
200
150
100
50
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Field Current [A]
Figure 4.4: Relationship between Field Current and Terminal Voltage.
4.3 Bridge Output

In this section the SCR bridge is tested in an open loop condition. The delay, or the firing
angle, α, is set to fixed values.
Figures 4.5, 4.6 and 4.7 show the input voltage, output voltage, and the two gate currents that
are sent to the SCR pairs. The delay is fixed so that the firing angle, α = 180°, and this gives
zero voltage output as expected, see figure 4.6.
Similarly, figures 4.8, 4.9 and 4.10, show the input voltage, output voltage, and the two gate
currents when the delay gives firing angle, α = 0°. This gives the maximum output of 310.9 V
DC, displayed in figure 4.9. Figure 4.10 shows that the gate currents are now firing the SCRs
so that the they conduct for the entire sinusoidal input wave.
4.3. BRIDGE OUTPUT 41
200
Input Voltage [V]
−200
0 20 40
Time [ms]
Figure 4.5: Input voltage over a period of 40 ms, and α = 180°.
1
Output Voltage [V]
0.8
0.6
0.4
0.2
0
0 20 40
Time [ms]
Figure 4.6: Output voltage over a period of 40 ms, and α = 180°.
1.5
Gate Current [V]
0.5
0
0 20 40
Time [ms]
Figure 4.7: Gate currents over a period of 40 ms, and α = 180°.

200
Input Voltage [V]
−200
1,000 1,020 1,040

Time [ms]
Figure 4.8: Input voltage over a period of 40 ms, and α = 0°.
314
Output Voltage [V]
312
310
308
306
1,000 1,020 1,040
Time [ms]
Figure 4.9: Output voltage over a period of 40 ms, and α = 0°.
1.5
Gate Current [V]
0.5
0
1,000 1,020 1,040
Time [ms]
Figure 4.10: Gate currents over a period of 40 ms, and α = 0°.

4.4. SCR CONTROL 43
4.4 SCR Control

Figures 4.11 and 4.11, show how the voltage is stepped up to a desired voltage. The system
is in this state controlling the output of the rectifier, and hence the field voltage for the ma-
chine. The voltage is stepped up in steps of 15 V until 90 V is reached, see figure 4.11. This
gives a peak-value of 206.3 V on the terminals as shown in figure 4.12.
100
80
Field Voltage [V]
60
40
20
0
0 2 4 6 8 10 12 14 16 18 20 22 24
Time [s]
Figure 4.11: Field voltage step up, 15 V each step.
200
Terminal Voltage Vabc [V]
100
−100
−200
0 2 4 6 8 10 12 14 16 18 20 22 24
Time [s]
Figure 4.12: Voltage step up on the terminal of the machine.
Throughout this test the current through SCR1 , is measured and shown in figure 4.13. The
peak of the current, is a inrush current that occurs, when the voltage is stepped up. The volt-
age were stepped up in steps of 15 V, and the figure represents the highest one that occurred.
140
120
Thyristor Current [A]
100
80
60
40
20
0
10 10.5 11 11.5 12 12.5 13 13.5 14 14.5 15
Time [s]
Figure 4.13: The current through SCR1 while stepping up the voltage in steps of 15 V.
4.5 SCR to Terminal Control

The SCR control is set to hold the peak of the terminal voltage at a reference of 200 V peak-
value. The system is still connected to a very low load, to simulate open-circuit conditions,
so that the load draws approximately zero current. Figures 4.14 and 4.15 show the field and
terminal voltage, and their variations, respectively. The field voltage is stepped up to 90 V,
and stabilized before the switching are done. Shown in section 3.6, the peak value of the
terminal voltage is stabilized at 206.3 V with a field voltage of 90 V, and when the terminal
voltage is set to 200 V the figures show how the field voltage is stabilizing at around 87.3 V.
110
Field Voltage [V]
100
90
80
70
30 31 32 33 34 35 36 37 38 39 40
Time [s]
Figure 4.14: Field voltage variations, switching from SCR to terminal control.
4.6. MECHANICAL INPUT VARIATIONS 45
Terminal Voltage Vabc [V] 200
100
−100
−200
30 31 32 33 34 35 36 37 38 39 40
Time [s]
Figure 4.15: Terminal voltage variations, switching from SCR to terminal control.
4.6 Mechanical Input Variations

Some mechanical input variations are done while the system is in the state of automatic
control. The control system is still set to hold the peak value of the terminals at 200 V. This
test shows how the control systems respond when the RPM first drops, and then rises again.
Figures 4.16 and 4.17 again show the field and terminal voltages. The system is here dropped
at a constant speed of 1500 RPM to 1400 RPM in a step, and the figures shows how the volt-
ages respond. As figure 4.16 shows, the field voltage increases from 87.3 V to approximately
93.6 V.
110
Field Voltage [V]
100
90
80
70
50 51 52 53 54 55 56 57 58 59 60
Time [s]
Figure 4.16: Field voltage response from decreasing the mechanical input to 1400 RPM.
Terminal Voltage Vabc [V] 200
100
−100
−200
50 51 52 53 54 55 56 57 58 59 60
Time [s]
Figure 4.17: Terminal voltage response from decreasing the mechanical input to 1400 RPM.
Figures 4.18 and 4.19 show the field and terminal voltages, when the system increases the
mechanical input as a step from 1400 RPM to 1500 RPM. The field voltage is now decreasing
from 93.6 V back to 87.3 V.
110
Field Voltage [V]
100
90
80
70
70 71 72 73 74 75 76 77 78 79 80
Time [s]
Figure 4.18: Field voltage response from increasing the mechanical input to 1500 RPM.
200
Terminal Voltage Vabc [V]
100
−100
−200
70 71 72 73 74 75 76 77 78 79 80
Time [s]
Figure 4.19: Terminal Voltage response from increasing the mechanical input to 1500 RPM.
4.6. MECHANICAL INPUT VARIATIONS 47
4.6.1 PI Comparison
In this section the Ziegler-Nichols and Cohen-Coon PI tuning are used and compared against
the autotuned PI regulator. All measures has been done at the exact same moment for all of
the controllers, so that they could easily get compared.
208
Matlab Autotune
Ziegler Nichols
206 Cohen-Coon
204
202
200
198
196
18 18.5 19 19.5 20 20.5 21 21.5 22 22.5 23 23.5 24 24.5 25 25.5 26
Time [s]
Figure 4.20: Switching from SCR control to terminal control.
260
Matlab Autotune
Ziegler Nichols
250
Cohen-Coon
240
230
220
210
200
190
23 24 25 26 27 28 29 30 31 32 33 34 35
Time [s]
Figure 4.21: Reference change from 200 V to 230 V.

48
5 Discussion
The excitation system plays an essential role in the stability of modern power systems. They
have recieved a lot of protective and control functions that result in a more stable system.
The main goal of this thesis was to look into the existing excitation systems out there and
design an excitation system for a small synchronous generator. The SM, stationed at the
Western Norway University of Applied Science, parameters has been used to make the sim-
ulated SM, designed in Matlab ® /Simulink® , to behave the same.
The excitation system design can come close to a Type AC system, since a single-phase AC
source is used as the excitation source. Further more a SCR bridge rectifier was chosen to,
rectify the AC current into DC current. This choice was made because of the need of a con-
trolled rectifier and that most literature used SCRs for this purpose.
The rectifier was analyzed by forcing the delay, and hence the firing angle, α to some desired
angles. Section 4.3 showed that when the firing angle α is set to 180°, the output of the bridge
reached to 0V . That shows that the SCRs are blocking for the whole sinusoidal AC input. The
maximum output was achieved at α equal to zero degrees, and the maximum voltage was
measured at approximately 310.9V .
This show that the bridge manage to control voltages as high as the maximum excitation
voltage that the Terco MV1027-235 can handle, that is 220 V, as shown in the datasheet in
appendix D.1. The saturation limiter can then be programmed, so that the bridge will never
let through more excitation power then the SM can handle.
To get the output of the bridge, the field voltage, to start at zero voltage output, the PI con-
troller for the bridge controller, were given an initial value that correspond to the firing angle
α equal to 180°. Figure 4.7 show that the firing pulses are shifted 180°. For the automatic
control system, the PI controller got an initial value corresponding to the reference termi-
nal peak voltage at 200 V AC. This make the switching, from bridge controller to automatic
controller, more smooth bigger overshoots avoided.
The designed SM were simulated to behave as the Terco MV1027-235 as much as possible.
Parameters used in the simulation were found from the datasheet and results from former
49
students, as shown in appendix D.1. By constant mechanical input of 1500 RPM, 50 Hz on the
terminals were achieved, as shown in figure 4.3. The relationship between the field voltage
and the peak terminal voltage of the simulated machine, is not quite similar to the actual
machine. Studying the open circuit test, see [33], the simulated SM got a slight higher ratio
between the voltages. This is probably a result of the estimations of the parameters, L md and
L mq .
The relationship between the field current and the terminal voltage is displayed in figure 4.4.
This figure shows how the peak voltage on the terminals increases with the field current and
shows the field current when the system is running at reference terminal voltage at 200 V.
The generator is simulated without taking the saturation into account. In a real generator,
the field flux will increase linearly with the field current until a certain point, but the terminal
voltage will start to degrade when the machine is overexcited. Since the SM operates way
below the maximum field current given in the datasheet, the saturation is not accounted for.
The bridge controller is designed to control the output voltage of the bridge to desired values.
In this manner, the user can set the field voltage to desired values. A PI controller was used in
series with a saturation limiter so that the firing of the SCRs would always happen within it
limits. A flowchart was implemented in Matlab ® /Simulink® , that delayed each step by 0.25
s. In this way, the inrush-currents over the thyristors got minimized and, in that manner, pro-
tect them. The highest inrush current that was measured when the field-voltage stepped up
to 90 V, is shown in figure 4.13. This inrush-current was measured to approximately 141.2 A.
This should not be a problem for the SCRs because of their high power handling capabilities.
The switching from controlling the bridge to controlling the terminals, happens after the
desired field voltage is reached. As figure 4.14 shows, the system spends a few seconds to
stabilize. The terminal voltage drops from 206.3 V AC peak, to 200 V AC peak, as shown
in figure 4.15. These results show that the PI controller overshoots a bit when the systems
are switched, but not very much. This could probably be fixed by tuning the PI controller
even better, but the terminals are staying between the ±10% of the reference voltage. Hence,
they follow the Norwegian standard, written about in chapter 1. Also, tests, which the RPM
varied were executed to see how the control system reacts. The system was both exposed by
decreasing the mechanical input from 1500 RPM to 1400 RPM, and vice versa. The results
show a bit overshoot coming from the PI controller. Still, one could argue that in real life,
the mechanical input would never drop or increase by 1000 RPM at an instance of time, and
hence, the control system would get a bit more time to adjust both on the increase, and
decrease, of the mechanical input.
Three different PI controllers got tested on the automatic control system. These are all com-
pared in figure 4.20, where the control is switched from controlling the SCR to controlling the
50 CHAPTER 5. DISCUSSION
terminals. In figure 4.21, they have all stabilized and applied a step. The results show that the
calculated PI parameters perform as good, or if not better then the autotuned parameters.
The Ziegler-Nichols and Cohen-Coon both responds quicker and shows less overshoot.
Generally the control system is pretty slow. This is because of the slow switching mecha-
nisms that comes with the SCR, and by using some other switches the system could be made
faster.
51
6 Conclusions
Design of an excitation system for a SM was investigated and presented in this thesis. A rec-
tifier based on SCRs worked as the exciter, which rectified the AC current and fed the SM
with DC current to magnetize the field winding of the machine. Therefore, this excitation
system could classified as an Type AC excitation system, based on the source of excitation. A
control system to regulate the output voltage of the SCR were developed. This system makes
it possible to control the field voltage directly, and step it up to a desired value. An automatic
control system was also developed which controlled the terminal voltage output of the SM.
Both control systems, were based on controlling the firing angle α, by the means of phase
control, for the SCR bridge, and both seemed to work fine when exposed with different tests.
A comparison between the Ziegler-Nichols and Cohen-Coon tuning method, and the auto-
tuning implemented in Matlab® /Simulink® was performed and shows pretty similar results.
The thesis has covered a wide range of topics, all the way from control systems, power elec-
tronics, excitation systems and the SM itself. It provide information about the most essential
components and systems evolving the excitation system. The work did result in a working
excitation system in Matlab® /Simulink® , which could be implemented in future projects. It
can however be challenging to balance the simulation, in the manner of including all physi-
cal phenomenon, and contrary not make it too complex for practical applications. The sim-
ulation has included parameters from different datasheet, and components, that are used in
practical applications, accounting for some of these challenges.
Future Work
The simulated SM could be optimized by more exact parameters.
• The control system proves to be functional, but could be improved by more advanced
control loops.
• Implement the system in the laboratory. Proposed component list can be found in
apendix C.
• Implementation of control units, described in section 2.4.

I
References
[1] A. Bergen and V. Vittal, Power Systems Analysis, en, Second, ser. 2. United States of
America: Tom Robbins, 2000, vol. 2, ISBN: 0-13-691990-1.
[2] Z. Guan, X. Wang, X. Bian, L. Wang, and Z. Jia, “Analysis of causes of outdoor insula-
tors damages on HV and UHV transmission lines in China,” en, in 2014 IEEE Electrical
Insulation Conference (EIC), Philadelphia, PA, USA: IEEE, Jun. 2014, pp. 227–230, ISBN:
978-1-4799-2789-0 978-1-4799-2787-6 978-1-4799-2786-9. DOI: 10.1109/EIC.2014.
6869381.
[3] Y. Qun and G. jianbo, “Self-organized Criticality and Its Application in Power System
Collapse Prevention,” en, in 2006 International Conference on Power System Technol-
ogy, Chongqing, China: IEEE, Oct. 2006, pp. 1–3, ISBN: 978-1-4244-0110-9 978-1-4244-
0111-6. DOI: 10.1109/ICPST.2006.321740.
[4] J. Hauer, N. Bhatt, K. Shah, and S. Kolluri, “Performance of "WAMS East" in provid-
ing dynamic information for the North East blackout of August 14, 2003,” en, in IEEE
Power Engineering Society General Meeting, 2004., vol. 2, Denver, CO, USA: IEEE, 2004,
pp. 1685–1690, ISBN: 978-0-7803-8465-1. DOI: 10.1109/PES.2004.1373161.
[5] J. Li, Design and Application of Modern Synchronous Generator Excitation Systems, en.
John Wiley & Sons, Jun. 2019, ISBN: 978-1-118-84087-0.
[6] Olje og energidepartementet, “Forskrift om leveringskvalitet i kraftsystemet - Lovdata,”
lovdata, Dec. 2004.
[7] Kundur, Power System Stability And Control, en. McGraw-Hill, 1994, ISBN: 978-0-07-
063515-9.
[8] A. E. Fitzgerald, C. Kingsley, and S. D. Umans, Electric Machinery, en, 6th ed, ser. McGraw-
Hill Series in Electrical Engineering. Boston, Mass: McGraw-Hill, 2003, ISBN: 978-0-07-
366009-7 978-0-07-112193-4.
[9] R. Hongtao, “Analysis of the Short Circuit at the Outlet of Self-excited Synchronous
Generator,” en, p. 4,
[10] “IEEE Std 421.5™-2016 (Revision of IEEE Std 421.5-2005),” en, p. 207, 2016.
II REFERENCES
[11] W. Wright, R. Hawley, and J. Dinely, “Brushless Thyristor Excitation Systems.,” en, IEEE
Transactions on Power Apparatus and Systems, vol. PAS-91, no. 5, pp. 1848–1854, Sep.
1972, ISSN: 0018-9510. DOI: 10.1109/TPAS.1972.293509.
[12] A. Ahmadi, S. Mahbubi, and M. Shahnazari, “Rotating rectifier fault detection in brush-
less excitation system of synchronous generators,” en, in 2019 27th Iranian Confer-
ence on Electrical Engineering (ICEE), Yazd, Iran: IEEE, Apr. 2019, pp. 1–5, ISBN: 978-1-
72811-508-5. DOI: 10.1109/IranianCEE.2019.8822368.
[13] B. M. Weedy, B. J. Cory, N. Jenkins, J. B. Ekanayake, and G. Strbac, Electric Power Sys-
tems, en. John Wiley & Sons, Jul. 2012, ISBN: 978-1-118-36108-5.
[14] “IEEE Guide for Synchronous Generator Modeling Practices and Parameter Verifica-
tion with Applications in Power System Stability Analyses,” en, IEEE, Tech. Rep. DOI:
10.1109/IEEESTD.2020.9020274.
[15] R. H. Park, “Two-reaction theory of synchronous machines generalized method of anal-
ysis-part I,” en, p. 12, 1929.
[16] ——, “Two-reaction theory of synchronous machines-II,” en, Transactions of the Amer-
ican Institute of Electrical Engineers, vol. 52, no. 2, pp. 352–354, Jun. 1933, ISSN: 0096-
3860. DOI: 10.1109/T-AIEE.1933.5056309.
[17] P. Vernekar, Z. Wang, A. Serrani, and K. Passino, “Feedback Control Methods for a Sin-
gle Machine Infinite Bus System,” en, p. 144,
[18] M. Basler, “Excitation systems: The current state of the art,” en, in 2006 IEEE Power
Engineering Society General Meeting, Montreal, Que., Canada: IEEE, 2006, 7 pp. ISBN:
978-1-4244-0493-3. DOI: 10.1109/PES.2006.1709610.
[19] M. Kumar Sharma, R. Pathak, M. Kumar Jha, and M. Qureshi, “Excitation control of
a power plant alternator using interval type-2 fuzzy logic controller,” en, Advances in
Modelling and Analysis C, vol. 73, no. 4, pp. 182–188, Dec. 2018, ISSN: 12404535. DOI:
10.18280/ama_c.730407.
[20] A. Vahedi, “Rotary diode failure detection in brushless exciter system of power plant
synchronous generator,” en, in 2016 6th Conference on Thermal Power Plants (CTPP),
Tehran, Iran: IEEE, Jan. 2016, pp. 6–11, ISBN: 978-1-5090-0372-3. DOI: 10.1109/CTPP.
2016.7482926.
[21] J. R. Williams, “The amplidyne,” en, Electrical Engineering, vol. 65, no. 5, pp. 208–213,
May 1946, ISSN: 0095-9197. DOI: 10.1109/EE.1946.6441670.
[22] N. Mohan, T. M. Undeland, and W. P. Robbins, Power Electronics: Converters, Applica-
tions, and Design, en, 3rd ed. Hoboken, NJ: John Wiley & Sons, 2003, ISBN: 978-0-471-
22693-2.
[23] L. Sivertsen, Elektriske Maskiner, no, 1. Vigmostad & Bjørke AS, Jul. 2019.
REFERENCES III
[24] E. Irmak, N. Guler, and M. Ersan, “PI controlled solar energy supported static excita-
tion system desing and simulation for synchronous generators,” in 2016 IEEE Interna-
tional Conference on Renewable Energy Research and Applications (ICRERA), Birming-
ham, United Kingdom: IEEE, Nov. 2016, pp. 1024–1028, ISBN: 978-1-5090-3388-1. DOI:
10.1109/ICRERA.2016.7884489.
[25] A. Mansoor, W. Grady, R. Thallam, M. Doyle, S. Krein, and M. Samotyj, “Effect of supply
voltage harmonics on the input current of single-phase diode bridge rectifier loads,”
en, IEEE Transactions on Power Delivery, vol. 10, no. 3, pp. 1416–1422, Jul. 1995, ISSN:
08858977. DOI: 10.1109/61.400924.
[26] G. V. Aparna, G. Suresh Babu, and T. M. Krishna, “Simulation and Analysis of Sin-
gle Phase Full Bridge Diode Rectifier with Different Passive Power Factor Correction
Techniques,” en, in 2018 International Conference on Recent Innovations in Electrical,
Electronics & Communication Engineering (ICRIEECE), Bhubaneswar, India: IEEE, Jul.
2018, pp. 2351–2355, ISBN: 978-1-5386-5995-3. DOI: 10.1109/ICRIEECE44171.2018.
9008888.
[27] S.-I. Motegi and Y. Nishida, “Three-Phase PFC Rectifier by Means of Three Single-
Phase Passive PFC Units,” en, in 2019 10th International Conference on Power Elec-
tronics and ECCE Asia (ICPE 2019 - ECCE Asia), Busan, Korea (South): IEEE, May 2019,
pp. 829–832, ISBN: 978-89-5708-313-0. DOI: 10.23919/ICPE2019- ECCEAsia42246.
2019.8797054.
[28] Z. Meng-di, W. Xi-xiu, Q. Wei, L. Fei, and L. Bo, “Application of PID control based on
BP-NN for marine generator excitation system,” en, in 2012 Power Engineering and
Automation Conference, Wuhan, Hubei, China: IEEE, Sep. 2012, pp. 1–4, ISBN: 978-1-
4577-1600-3 978-1-4577-1599-0. DOI: 10.1109/PEAM.2012.6612457.
[29] Y. Cao and J. Ma, “Research on PID parameters optimization of synchronous generator
excitation control system,” en, in 2010 5th International Conference on Critical Infras-
tructure (CRIS), Beijing, China: IEEE, Sep. 2010, pp. 1–5, ISBN: 978-1-4244-8080-7. DOI:
10.1109/CRIS.2010.5617581.
[30] S. Yuqiu, Z. Jibin, and S. Jianxin, “The Analysis and Simulation of Closed-loop Paper
Tension Control System Based on MATLAB in the Web Offset,” en, in 2012 Third Inter-
national Conference on Digital Manufacturing & Automation, Guilin, China: IEEE, Jul.
2012, pp. 881–883, ISBN: 978-1-4673-2217-1 978-0-7695-4772-5. DOI: 10.1109/ICDMA.
2012.207.
[31] R. A. Hyde, “Using SimscapeTM to Support System-Level Design,” en, p. 6,
[32] “Simscape Electrical Reference (Specialized Power Systems),” en, p. 980,
[33] T. Fischer, “Terminal Voltage and Reactive Power Control of a Synchronous Generator
Through Transmission Lines,” en, p. 127,
IV REFERENCES
[34] P. Dedie, V. Brommer, and S. Scharnholz, “ICCOS Countercurrent-Thyristor High-Power

Opening Switch for Currents Up to 28 kA,” en, IEEE Transactions on Magnetics, vol. 45,
no. 1, pp. 536–539, Jan. 2009, ISSN: 0018-9464. DOI: 10.1109/TMAG.2008.2008428.
V
Appendices
A Park transformation
P shows the matrix for the park transformation, that transforms the stator variables into rotor
variables. These steps are shown in equation A.1, A.2, A.3 and A.4.
 
r p1 p1 p1
3 3 3
2
cosθ cos(θ − 2π ) cos(θ + 2π )

P= (A.1)
3 3 3 
2π 2π
si nθ sin(θ − 3 ) sin(θ + 3 )
 
1
r  p2 cosθ si nθ
2 1 
2π 
P−1 = PT = p cos(θ − 2π
3
) si n(θ − )
3 
(A.2)
3 2 
p1 cos(θ + 2π ) si n(θ + 2π
)
2 3 3
" # " #
−1 P 0 −1 T PT 0
P , , B =B = (A.3)
0 1 0 1
vB , Bv , iB , Bi , λB , B λ (A.4)
The time various expressions for the stator and exciter coil, quadrature and direct axis induc-
tances. All of these are variouse by the placement of the rotor, and the placement is given by
θ. [1]
λaa 0
L aa = = L s + L m cos(2θ) (A.5)
ia
λbb 0 2π
L bb = = L s + L m cos(2θ − (A.6)
ib 3
VI APPENDIX A. PARK TRANSFORMATION
λcc 0 2π
L cc = = L s + L m cos(2θ + ) (A.7)
ic 3
λaa 0 π
L ab = L ba = = −[M s + L m cos2(θ + )] (A.8)
ib 6
λbb 0 π
L bc = L cb = = −[M s + L m cos2(θ + )] (A.9)
ic 2
λcc 0 5π
L ca = L ac = = −[M s + L m cos2(θ + )] (A.10)
ia 6
L aD = L D a = M D cosθ (A.11)
2π
L bD = L Db = M D cos(θ − ) (A.12)
3
2π
L cD = L Dc = M D cos(θ + ) (A.13)
3
L aF = L F a = M F cosθ (A.14)
2π
L bF = L F b = M F cos(θ − ) (A.15)
3
2π
L cF = L F c = M F cos(θ + ) (A.16)
3
L aQ = L Qa = MQ cosθ (A.17)
2π
L bQ = L Qb = MQ cos(θ − ) (A.18)
3
2π
L cQ = L Qc = MQ cos(θ + ) (A.19)
3
L DF = L F D = M R , L FQ = L QF = 0, L DQ = L QD = 0 (A.20)
VII
L DD = L Q , LF = LF , L QQ = L Q (A.21)
Expressions for the transformed voltages, fluxes and inductances for the park transforma-
tion. The zero sequence is given by equation A.22. [1]
d α0
v 0 = −r i 0 − (A.22)
dt
The direct axis equations are given in equations A.23, A.24 and A.25.
dθ d αd
v0 = −r i d − αq − (A.23)
dt dt
d αF
vF = rF iF + (A.24)
dt
d αD
vD = rD iD + =0 (A.25)
dt
The quadrature axis equations are given in equations A.26 and A.27.
dθ d αq
v q = −r i q − αd − (A.26)
dt dt
d αQ
vQ = −r Q i Q + =0 (A.27)
dt
L 0 = L s − 2M s (A.28)
3
Ld = L s + Ms + Lm (A.29)
2
3
L q = L s + Ms − Lm (A.30)
2
VIII
B Simulink Model
Figure B.1: Full Simulink Model.

IX
Figure B.2: Full control system.

X APPENDIX B. SIMULINK MODEL
Figure B.3: Thyristor Bridge

XI
Figure B.4: Synchronous Machine connected to load

XII
C Ziegler-Nichols & Cohen-Coon Tuning
Table C.1 and C.2 shows the equations for calculating the P, I and D for Ziegler-Nichols and
Cohen-Coon, respectively.
KP Tr Td
v0
P
K 0 τ0
0.9v 0
PI 3τ0
K 0 τ0
1.2v 0
PID 2τ0 0.5τ0
K 0 τ0
Table C.1: Ziegler-Nichols tuning by using reaction curve.
KP Tr Td
v0
P
K 0 τ0
· ¸
v0 τ0 30v 0 + 3τ0
PI 0.9 +
K 0 τ0 12v 0 9v 0 + 20τ0
· ¸
v 0 4 τ0 τ0 [32v 0 + 6τ0 ] 4τ0 v 0
PID +
K 0 τ0 3 4v 0 13v 0 + 8τ0 11v 0 + 2τ0
Table C.2: Cohen-Coon tuning by using reaction curve.

XIII
D General
D.1 Terco MV1027-235

The SM that the simulated machine is suppose to behave like is a Terco MV1027-235 with
serial number: 39333. This machine consists of two damper windings and all parameters
from the datasheet are listed in table D.1.
Technical Specifications Datasheet Name plate

PowerGener at or 1.2 kVA 1.2 kVA
cosφ 0.8 0.8
PowerMot or 1.0 kW 1.0 kW
VoltageSt ar 220 − 240 V 220 − 240 V
CurrentSt ar 3.5 A 3.5 A
VoltageDel t a 127 − 140 V 220 − 240 V
CurrentDel t a 6.1 A 3.5 A
DCE xci t at i onV ol t ag e 220 V 220 V
DCE xci t at i onCur r ent 1.4 A 1.4 A
Temp.Class F (155·)
Protection I P 23
Duty Type S260mi n
Norm E N 60034 : 1993
Table D.1: Parameters for Terco MV1027-235.
D.2 Littelfuse S6065KTP

The SM that the simulated machine is suppose to behave like is a Terco MV1027-235 with
serial number: 39333. This machine consists of two damper windings and all parameters
from the datasheet are listed in table D.2.
XIV APPENDIX D. GENERAL
Table D.2: Parameters for Thyristor/SCR.
Technical Specifications Datasheet

Rated Average On-State Current 65 A
Thyristor Type SCR
Package Type TO-218AC
Repetitive Peak Reverse Voltage 600 V
Surge Current Rating 950A
Mounting Type Through Hole
Maximum Gate Trigger Current 50 mA
Maximum Gate Trigger Voltage 2V
Maximum Holding Current 80 mA
Pin Count 3
Dimensions 16x4.78x21.21 mm
Peak On-State Voltage 1.8 V
Minimum Operating Temperature −40° C
Maximum Operating Temperature +125° C
Repetitive Peak Off-State Current 0.02 mA
D.3 Component List

A component list has been made in table D.3, for future students.
Table D.3: Parameters for Terco MV1027-235.
System Part Component Quantity Manufacturer Name Specification

AC/DC Stage S6065KTP 4 Littelfuse 65A, 600V
AC/DC Stage Capacitor 1 ALS70H273QW250 27000µF, 250V
AC/DC Stage Resistor 1 TE500B100RJ 100Ω, 500W W
Precharge Circuit IGBT 3 STGF20H60DF 40A, 600V
Basic Current Sensor 2 LEM LAH 50-P 0 − 50 A
Basic Voltage Sensor 3 LEM LV 25-P 10 − 500 V
XV
Index
AC Excitation System (Type AC), 11–15, 17, Root Mean Square (RMS), 28, 30, 38
48, 51 Rotations per Minute (RPM), ix, 5, 38, 39, 45,
Alternating Current (AC), 4, 11–19, 27, 28, 30, 46, 49
48, 49, 51
Semiconductor Controlled Rectifiers (SCR), vi,
Automatic Power Factor Regulator (APFR), vi,
ix, x, XIV, 27–35, 40, 43–45, 47–51
19, 20
Static Excitation System (Type ST), 12, 15, 17
Automatic Reactive Power Regulator (AQR),
Synchronous Machine (SM), viii, x, XIII, 2, 4–
vi, viii, 19, 20
9, 11–15, 27–30, 33, 48, 49, 51
Automatic Voltage Regulator (AVR), 4, 11, 20,
21 Under-excitation Limiter (UEL), vi, 21, 22
DC Excitation System (Type DC), 11, 12

Differential (D), vi, XII, 24
Direct Current (DC), 2, 4, 5, 11–17, 19, 27–29,
34, 40, 48, 51
Electromotive Force (EMF), 6
Institute of Electrical and Electronics Engineers

(IEEE), 2, 11
Integral (I), vi, XII, 24, 32, 33, 37
Overe-xcitation Limiter (OEL), vi, 22
Permanent Magnet Synchronous Generator

(PMSG), 14
Power System Stabilizer (PSS), vi, 22, 23
Proportional (P), vi, XII, 23, 32, 33, 37
Proportional Differential (PD), 23, 25
Proportional Integral (PI), XII, 23, 25, 26, 32,
33, 36, 47–50
Proportional Integral Differential (PID), viii,
XII, 4, 11, 23, 25, 26

Design of A Generator Excitation System

Uploaded by

Copyright:

Available Formats

Design of A Generator Excitation System

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Design of A Generator Excitation System

Uploaded by

Copyright:

Available Formats

U NIVERSITY OF B ERGEN

Author: Jonas Fotland Fænn

List of Figures vii

2 State of the art 4

Capacitor filter, C and source inductor, L s . . . . . . . . . . . . . . . . . . 18

2.1 Simplified schematic of a single-phase Synchronous Machine (SM) with salient

3.1 Sketch of Matlab ® /Simulink® Model. . . . . . . . . . . . . . . . . . . . . . . . . . 28

3.2 Thyristor block: a) Equivalen circuit b)VI characteristic. . . . . . . . . . . . . . 29

4.1 Constant mechanical input at 1500 Rotations per Minute (RPM). . . . . . . . . . 39

B.1 Full Simulink Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII

2.1 Parameters for figure 2.19. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3.1 Parameters in Matlab ® /Simulink® for the SM. . . . . . . . . . . . . . . . . . . . . 28

C.1 Ziegler-Nichols tuning by using reaction curve. . . . . . . . . . . . . . . . . . . . XII

D.1 Parameters for Terco MV1027-235. . . . . . . . . . . . . . . . . . . . . . . . . . . . XIII

C capacitor vi, 18, 30, 34

Vo average output voltage for the single-phase diode bridge 16, 17

1.1 Literature review

• Implementation of a SM in Matlab® /Simulink® , to simulate the Terco MV1027-235

• Design of AC/DC rectifier as exciter for the simulated SM.

4 Results This chapter will present the simulated model.

6 Conclusion This chapter will be a conclusion of the work and will

2 State of the art

Figure 2.1: Simplified schematic of a single-phase SM with salient poles.

Figure 2.2: Simplified schematic of salient and cylindrical rotor.

Figure 2.3: Simplified schematic of damper winding.

2.1.2 Voltage Generation

Figure 2.4: Air-gap flux due to i a .

λag = λaa 0 + λar (2.4)

terminal voltage, Va , for a round rotor machine is given:

Λar = Λad + Λaq (2.8)

where L d 1 is the inductance and, I ad is the current, on fictitious winding d .

where L q1 is the inductance and, I aq is the current, on fictitious winding q.

Figure 2.5: Equivalent circuit for a salient-pole rotor SM.

2.1.3 Park Transformation

Figure 2.6: Schematic of a salient-pole SM, adapted from [1].

Where λ0 , λd , λq , λF , λD , λQ is the flux linkage of each winding. L 0 , L d , Lq, L D , L F , L Q is

2.2 Excitation System

• DC Excitation System (Type DC) - DC generator as the source of excitation power.

• AC Excitation System (Type AC) - AC generator as the source of excitation power.

A simplified block-diagram of a Type DC system is represented in figure 2.8. It shows the

Figure 2.10: Alternator-supplied controlled-rectifier system, adapted from [7].

Figure 2.11: Simplified schematics of brushless system, adapted from [7].

Figure 2.12: Simplified schematics of system, adapted from [7].

2.3.1 Single-Phase Uncontrolled Rectifier

Figure 2.13: Single Phase Rectifier and its current flow.

Where Vi is the AC input measured in rms, ω = 2π f and ωT /2 = π [22] [26].

2.3.2 Three-Phase Rectifier

Figure 2.15: Three Phase Rectifier.

Figure 2.16: Three Phase Rectifier.

Capacitor filter, C and source inductor, L s

To get a understanding of why a source inductor is needed a simplified circuit is shown in

Figure 2.18: Single-phase bridge with source inductor L s .

2.4 Control Units

Figure 2.19: AQR Transfer function.

Table 2.1: Parameters for figure 2.19.

Symbol Meaning Symbol Meaning

T (s) Rate of changes for the voltage Q LI M (p.u.) AQR output

QD Deadband zone ±UL (p.u.) Integral limit

Figure 2.20: Transfer function.