Вы находитесь на странице: 1из 135

SPRINGER BRIEFS IN

ELEC TRIC AL AND COMPUTER ENGINEERING

SKrishna

An Introduction
to Modelling of
Power System
Components
123

SpringerBriefs in Electrical
and Computer Engineering

For further volumes:


http://www.springer.com/series/10059

S Krishna

An Introduction
to Modelling of Power
System Components

123

S Krishna
Department of Electrical Engineering
Indian Institute of Technology Madras
Chennai, Tamil Nadu
India

ISSN 2191-8112
ISSN 2191-8120 (electronic)
ISBN 978-81-322-1846-3
ISBN 978-81-322-1847-0 (eBook)
DOI 10.1007/978-81-322-1847-0
Springer New Delhi Heidelberg New York Dordrecht London
Library of Congress Control Number: 2014933281
 The Author(s) 2014
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed. Exempted from this legal reservation are brief
excerpts in connection with reviews or scholarly analysis or material supplied specifically for the
purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the
work. Duplication of this publication or parts thereof is permitted only under the provisions of
the Copyright Law of the Publishers location, in its current version, and permission for use must
always be obtained from Springer. Permissions for use may be obtained through RightsLink at the
Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Preface

The book is based on the notes prepared for the courses taught by the author at
Indian Institute of Technology Madras. The book gives the derivation of the model
of power system components such as synchronous generator, transformer, transmission line, DC transmission system, flexible AC transmission systems, excitation system, and speed governor. The model of load and prime movers are given
without derivation. The book can serve as a text for a short graduate course on
power system modelling, or as a supplement for graduate courses on power system
stability and flexible AC transmission systems.
Chennai, India, December 2013

S Krishna

Contents

Synchronous Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.1 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2 Model with Three Damper Windings . . . . . . . . . . . . . . . .
1.2.1 Voltage, Current, and Flux Linkage Relationships. .
1.2.2 Expression for Inductances. . . . . . . . . . . . . . . . . .
1.2.3 Equations of Motion . . . . . . . . . . . . . . . . . . . . . .
1.2.4 Expression for Electrical Torque. . . . . . . . . . . . . .
1.3 Parks Transformation . . . . . . . . . . . . . . . . . . . . . . . . . .
1.4 Transformation of Rotor Variables. . . . . . . . . . . . . . . . . .
1.4.1 Transfer Function Id s=Wd sjvf 0 . . . . . . . . . . . .
1.4.2 Transfer Function Vf s=Wd sid 0 . . . . . . . . . . . .
1.4.3 Transfer Function Iq s=Wq s . . . . . . . . . . . . . . .
1.5 Tests for the Determination of Parameters . . . . . . . . . . . .
1.5.1 Determination of d-Axis Parameters . . . . . . . . . . .
1.5.2 Determination of q-Axis Parameters . . . . . . . . . . .
1.6 Time Domain Model with Standard Parameters. . . . . . . . .
1.7 Time Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.7.1 Short-Circuit Time Constants . . . . . . . . . . . . . . . .
1.7.2 Open-Circuit Time Constants . . . . . . . . . . . . . . . .
1.8 Model in Per Unit Quantities . . . . . . . . . . . . . . . . . . . . .
1.9 Other Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.9.1 Model in the Absence of Zero Sequence Variables
00
and with Tdc
Td00 . . . . . . . . . . . . . . . . . . . . . . . .
1.9.2 Model with Stator Transients Neglected. . . . . . . . .
1.9.3 Two Axis Model. . . . . . . . . . . . . . . . . . . . . . . . .
1.9.4 One Axis (Flux Decay) Model . . . . . . . . . . . . . . .
1.9.5 Classical Model . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Transformer, Transmission Line, and Load
2.1 Transformer . . . . . . . . . . . . . . . . . . . .
2.1.1 Single-Phase Transformer . . . . .
2.1.2 Three-Phase Transformer . . . . . .

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

1
1
1
3
4
20
21
22
26
28
28
29
29
29
32
33
36
36
37
38
39

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

40
40
41
42
43
43

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

45
45
45
47

vii

viii

Contents

2.2

Transmission Line . . . . . . . . . . . . . . . . .
2.2.1 Inductance . . . . . . . . . . . . . . . . .
2.2.2 Capacitance . . . . . . . . . . . . . . . .
2.2.3 Transmission Line Model. . . . . . .
2.3 Krons Transformation . . . . . . . . . . . . . .
2.3.1 Definitions . . . . . . . . . . . . . . . . .
2.3.2 Application to Transformer . . . . .
2.3.3 Application to Transmission Line .
2.4 Load . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

50
50
56
60
65
65
68
69
71
72

DC and Flexible AC Transmission Systems. . . . .


3.1 Power Semiconductor Devices . . . . . . . . . . .
3.2 DC Transmission System . . . . . . . . . . . . . . .
3.2.1 Line-Commutated Converter. . . . . . . .
3.2.2 12-Pulse Line-Commutated Converter .
3.3 FACTS. . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.1 SVC . . . . . . . . . . . . . . . . . . . . . . . .
3.3.2 TCSC . . . . . . . . . . . . . . . . . . . . . . .
3.3.3 VSC . . . . . . . . . . . . . . . . . . . . . . . .
3.3.4 STATCOM . . . . . . . . . . . . . . . . . . .
3.3.5 SSSC. . . . . . . . . . . . . . . . . . . . . . . .
3.3.6 Multi-Converter FACTS Controllers . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

73
73
74
75
83
86
86
94
98
111
112
114
114

Prime Movers and Excitation System .


4.1 Prime Movers . . . . . . . . . . . . . . .
4.1.1 Steam Turbine . . . . . . . . .
4.1.2 Hydraulic Turbine . . . . . . .
4.2 Torsional Dynamics . . . . . . . . . . .
4.3 Speed Governor. . . . . . . . . . . . . .
4.4 Excitation System . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

117
117
117
117
118
119
121
123

Appendix A: Solution of Linear Ordinary Differential Equations


with Constant Coefficients . . . . . . . . . . . . . . . . . . . . . . .

125

Appendix B: Fourier Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

127

About the Author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

129

About the Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

131

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

133

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.

Chapter 1

Synchronous Generator

1.1 Construction
The synchronous generator converts mechanical energy into electrical energy. It has
a stationary component or stator, and a rotating component or rotor. The stator is
an annular structure made up of iron and has slots. Insulated coils are placed in
the slots, and these coils are connected to obtain a three-phase winding. The rotor
has electromagnets which are known as field poles. The rotor is placed within the
stator. The cross section of a synchronous generator with two field poles is shown
in Fig. 1.1. a, b, and c are the three stator windings which are 120 apart. f is the
field winding. A dot indicates that positive current flow is out of the paper/screen,
while a cross indicates that positive current flow is into the paper/screen. If the field
winding is excited by a DC source, then as the rotor rotates, an emf is induced in the
armature winding according to Faradays law.
The number of field poles depends on the speed of the prime mover. If the number
of field poles is p f , then one rotation of the rotor induces p f /2 cycles of emf in the
armature winding. The hydraulic turbines operate at low speed. Therefore, to obtain
the rated frequency, the synchronous generator driven by a hydraulic turbine has a
large number of field poles. On the other hand, steam and gas turbines operate at
high speeds; hence, the synchronous generator driven by these turbines has two or
four field poles.
The rotors often have amortisseur or damper circuits in the form of copper or
brass rods. These rods are short-circuited and are intended to damp out oscillations
in speed.

1.2 Model with Three Damper Windings


The model is initially developed for a synchronous generator having a pair of field
poles; generalization to any number of field poles is done later. An axis is defined
for each stator winding as shown in Fig. 1.2. Two axes, namely direct axis (d-axis)
S Krishna, An Introduction to Modelling of Power System Components,
SpringerBriefs in Electrical and Computer Engineering,
DOI: 10.1007/978-81-322-1847-0_1, The Author(s) 2014

1 Synchronous Generator

Fig. 1.1 Synchronous


generator

Fig. 1.2 Axes of


a synchronous generator

d axis

b axis

a axis
q axis

c axis

and quadrature axis (q-axis), are also defined. The d-axis is along the field pole.
The q-axis lags the d-axis by 90 [1]. a-axis, b-axis, and c-axis are stationary, and
d-axis and q-axis rotate at the speed of the rotor. The rotor is assumed to rotate in the
counterclockwise direction, and is the angle by which the d-axis leads the a-axis.
The amortisseur circuits and the eddy current effects in the rotor are represented
by two equivalent sets of short-circuited damper windings [13]. One set of windings
is oriented such that the flux in the rotor due to current in these windings is along
d-axis; these windings are said to be on the d-axis. The other set of windings is
oriented such that the flux in the rotor due to current in these windings is along
q-axis; these windings are said to be on the q-axis. The damper winding 1d is on the
d-axis, and the damper windings 1q and 2q are on the q-axis.

1.2 Model with Three Damper Windings

Fig. 1.3 Circuit diagram of a


winding

i
+

+
d

dt

1.2.1 Voltage, Current, and Flux Linkage Relationships


The circuit diagram of a synchronous generator winding (a, b, c, f, 1d, 1q, or 2q) is
shown in Fig. 1.3 [1]. , R, i, and are flux linkage, resistance, current, and voltage,
respectively.
By Kirchhoffs voltage law,
da
dt
db
dt
dc
dt
d f
dt
d1d
dt
d1q
dt
d2q
dt

= Ra i a va

(1.1)

= Ra i b vb

(1.2)

= Ra i c vc

(1.3)

= R f i f + v f

(1.4)

= R1d i 1d

(1.5)

= R1q i 1q

(1.6)

= R2q i 2q

(1.7)

The polarity of v f is such that v f and i f are positive in steady state.


The flux linkages, currents, and inductances are related as follows:

La
a
b Mba


c Mca


f = Mfa


1d M1da


1q M1qa
2q
M2qa

Mab
Lb
Mcb
Mfb
M1db
M1qb
M2qb

Mac
Mbc
Lc
Mfc
M1dc
M1qc
M2qc

Ma f
Mb f
Mc f
Lf
M1d f
M1q f
M2q f

Ma1d
Mb1d
Mc1d
M f 1d
L 1d
M1q1d
M2q1d

Ma1q
Mb1q
Mc1q
M f 1q
M1d1q
L 1q
M2q1q


ia
Ma2q
ib
Mb2q


Mc2q
ic

M f 2q
if

M1d2q
i 1d
M1q2q i 1q
L 2q
i 2q

(1.8)

1 Synchronous Generator

Fig. 1.4 Closed path chosen for the determination of


magnetic field intensity
Closed
path

a axis

1.2.2 Expression for Inductances


The stator windings are assumed to be filamentary. Let each stator winding have N
turns. The permeability of the stator core and the rotor core is assumed to be infinite.
Therefore, the magnetic field intensity is nonzero only in the air gap. If is the angle
measured from the a-axis in the counterclockwise direction, the air gap length g is
a periodic function of which satisfies the condition
g( + ) = g()

(1.9)

g is small compared to the inner radius of the stator. Therefore, it is assumed that
the magnetic field intensity in the air gap is radial [4]. From (1.9), it follows that the
magnetic field intensity Ha due to i a , in the air gap, in the radial outward direction,
satisfies the following equation:
Ha ( + ) = Ha ()

(1.10)

The expression for Ha is obtained from Amperes law applied to the closed path
shown in Fig. 1.4. The closed path consists of a semicircle in the stator core and a
straight line passing through and perpendicular to the rotor axis of rotation.
Therefore,

N ia
if 2 < < 2
Ha = 2gN ia
(1.11)
2g if 2 < < 3
2
A quantity called air gap magnetomotive force (MMF) is defined as the product of
air gap magnetic field intensity and air gap length. The air gap MMF due to i a is
Fa  Ha g

(1.12)

1.2 Model with Three Damper Windings

Fa

Nia /2

Nia /2
120

60

60
120
(degree)

180

240

300

Fig. 1.5 Air gap MMF due to i a

The variation in Fa with is shown in Fig. 1.5.


Let the air gap MMF be approximated by its fundamental component with peak
value 2N i a / [4]. This approximation can be done even for a practical stator winding
placed in the stator core having many slots and not restricted to six slots as in Fig. 1.1.
In fact, the air gap MMF for a practical stator winding is closer to a sinusoidal waveform than that for the one shown in Fig. 1.1. In order to establish a sinusoidal air gap
MMF waveform, it will be shown that the equivalent winding must be sinusoidally
distributed. Suppose the number of turns per radian at any location is

na =

2 N sin if 0
2
N sin if 0

(1.13)

The winding distribution is shown in Fig. 1.6. The current i a in the turns for
< < 0 is into the paper/screen, whereas the current in the turns for 0 < <
is out of the paper/screen. The number of turns of the equivalent sinusoidally distributed winding is

Na =

n a d =

4
2
N sin d = N

(1.14)

Substituting this in (1.13) gives

na =

N2a sin if 0
Na
2 sin if 0

(1.15)

1 Synchronous Generator

na

2N/

0
180

120

60

0
(degree)

60

120

180

Fig. 1.6 Winding distribution for sinusoidal air gap MMF

The air gap magnetic field intensity Ha1 in the radial outward direction, due to i a
with a winding distribution given by (1.15), is obtained from Amperes law applied
to the closed path shown in Fig. 1.4. For 0,

Ha1 =

1
i a
2g

0
n a d + i a

n a d

+

0 

Na
1
Na

sin d
=

sin d + i a
i a
2g
2
2

Na i a
cos
=
2g

(1.16)

Since Ha1 ( + ) = Ha1 (), the expression for Ha1 given by (1.16) is applicable
for any . The air gap MMF due to i a with the winding distribution given by (1.15)
is
Na i a
cos
(1.17)
Fa1 = Ha1 g =
2
The variation in Fa1 with is shown in Fig. 1.7. Hence, the winding distribution
given by (1.15) results in the air gap MMF equal to the fundamental component of
Fa .
The distribution of the other equivalent windings (b, c, f, 1d, 1q, and 2q) and the
air gap MMFs are given by

1.2 Model with Three Damper Windings

Fa1

Naia/2

Naia/2
180

120

60

0
(degree)

60

120

180

Fig. 1.7 Fundamental component of air gap MMF due to i a



2
Na
2
5
if
nb =
sin

2
3
3
3


2

2
Na
sin +

if
nc =
2
3
3
3
Nf
nf =
sin ( ) if +
2
N1d
sin ( ) if +
n 1d =
2
N1q

n 1q =
cos ( ) if +
2
2
2
N2q

n 2q =
cos ( ) if +
2
2
2


Na
2
i b cos
Fb1 =
2
3


Na
2
i c cos +
Fc1 =
2
3
Nf
Ff 1 =
i f cos( )
2
N1d
i 1d cos( )
F1d1 =
2
N1q
i 1q sin( )
F1q1 =
2
N2q
F2q1 =
i 2q sin( )
2

(1.18)
(1.19)
(1.20)
(1.21)
(1.22)
(1.23)
(1.24)
(1.25)
(1.26)
(1.27)
(1.28)
(1.29)

1 Synchronous Generator

The air gap length is a periodic function of satisfying (1.9). If the higher-order
harmonic components are neglected [4], then
1
= a0 + a2 cos(2 2 )
g

(1.30)

where a0 > a2 > 0. The flux density in the radial outward direction in the air gap
due to i a is
Fa1
(1.31)
Ba = 0 Ha1 = 0
g
where 0 is the permeability of free space; permeability of air is almost equal to
that of free space. The air gap flux, due to i a , linking a turn of winding a whose

sides are at = and = , is Ba rld; l is the length of the stator
and the rotor, and r is the inner
radius of the stator. The air gap flux linkage of

winding a, due to i a , is
na
Ba rld d. The ratio of this flux linkage to

i a gives the self inductance of winding a due to the flux crossing the air gap. The total
self inductance of winding a is obtained by adding to this, the leakage inductance
due to the leakage flux. If the leakage inductance of winding a is L al , the total self
inductance of winding a is

L a = L al +

Na

sin
2

Na

cos {a0 + a2 cos (2 2 )} rl d d


2

= L a0 + L a2 cos (2 )
where
L a0  L al +

(1.32)
Na2
0 rla0 ,
4

L a2 

Na2
0 rla2
8

(1.33)

The expressions for other inductances are obtained as follows:

Mab =
0

Na

sin
2

Na
2
cos
2
3



2
= Mab0 + L a2 cos 2
3

Mac =
0

Na

sin
2

{a0 + a2 cos (2 2 )} rl d d

(1.34)


Na
2
cos +
2
3

{a0 + a2 cos (2 2 )} rl d d

1.2 Model with Three Damper Windings



2
= Mab0 + L a2 cos 2 +
3

Ma f =

Na

sin
2

(1.35)

Nf

cos ( ) {a0 + a2 cos (2 2 )} rl d d


0
2

= Ma f 1 cos

Ma1d =

(1.36)

Na

sin
2

N1d

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2

= Ma1d1 cos

Ma1q =

(1.37)

Na

sin
2

N1q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

= Ma1q1 sin

Ma2q =

(1.38)

Na

sin
2

N2q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

= Ma2q1 sin
5/3

Mba =

(1.39)



Na
2
sin
2
3

2/3

Na

cos {a0 + a2 cos (2 2 )} rl d d


2



2
= Mab0 + L a2 cos 2
3
5/3

L b = L al +
2/3



Na
2
sin
2
3

(1.40)

10

1 Synchronous Generator

Na
2
cos
2
3

{a0 + a2 cos (2 2 )} rl d d



2
= L a0 + L a2 cos 2 +
3
5/3

Mbc =

(1.41)



Na
2
sin
2
3

2/3

Na
2
cos +
2
3

{a0 + a2 cos (2 2 )} rl d d

= Mab0 + L a2 cos (2 )
5/3

Mb f =



Na
2
sin
2
3

2/3

(1.42)

Nf

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2



2
= Ma f 1 cos
3

5/3

Mb1d =



2
Na
sin
2
3

2/3

(1.43)

N1d

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2



2
= Ma1d1 cos
3
5/3

Mb1q =
2/3



2
Na
sin
2
3

(1.44)

1.2 Model with Three Damper Windings

11

N1q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2



2
= Ma1q1 sin
3
5/3

Mb2q =

(1.45)



2
Na
sin
2
3

2/3

N2q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2



2
= Ma2q1 sin
3
/3
Mca =
2/3



Na
2
sin +
2
3

Na

cos {a0 + a2 cos (2 2 )} rl d d


2



2
= Mab0 + L a2 cos 2 +
3

/3
Mcb =
2/3

(1.46)

(1.47)



2
Na
sin +
2
3



2
Na

{a0 + a2 cos (2 2 )} rl d d
cos
0
2
3

= Mab0 + L a2 cos (2 )

(1.48)

12

1 Synchronous Generator

/3
L c = L al +

2/3



2
Na
sin +
2
3



Na
2

{a0 + a2 cos (2 2 )} rl d d
0
cos +
2
3



2
= L a0 + L a2 cos 2
3
/3
Mc f =
2/3

(1.49)



2
Na
sin +
2
3

Nf

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2



2
= Ma f 1 cos +
3

/3
Mc1d =
2/3

(1.50)



2
Na
sin +
2
3

N1d

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2



2
= Ma1d1 cos +
3
/3
Mc1q =
2/3



2
Na
sin +
2
3

(1.51)

N1q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2



2
= Ma1q1 sin +
3

(1.52)

1.2 Model with Three Damper Windings

/3
Mc2q =
2/3



2
Na
sin +
2
3

N2q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

2
= Ma2q1 sin +
3

Mfa =

Nf

sin ( )
2

= Ma f 1 cos

Mfb =

Na

cos {a0 + a2 cos (2 2 )} rl d d


2
(1.54)



2
Na

{a0 + a2 cos (2 2 )} rl d d
cos
0
2
3



2
= Ma f 1 cos
3

Mfc =

(1.53)

Nf
sin ( )
2

13

(1.55)

Nf
sin ( )
2



2
Na

{a0 + a2 cos (2 2 )} rl d d
cos +
0
2
3



2
= Ma f 1 cos +
3

(1.56)

14

1 Synchronous Generator
+

L f = L fl +

Nf
sin ( )
2

= L fl +

M f 1d =

N 2f
8

0 rl (2a0 + a2 )

(1.57)

Nf
sin ( )
2

Nf

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2

N1d

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2

N f N1d
0 rl (2a0 + a2 )
8
+

M f 1q =

Nf
sin ( )
2

(1.58)

N1q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

=0

(1.59)

M f 2q =

=0

Nf
sin ( )
2

N2q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


0
2
(1.60)

1.2 Model with Three Damper Windings


+

M1da =

N1d

sin ( )
2

15

Na

cos {a0 + a2 cos (2 2 )} rl d d


2

= Ma1d1 cos

M1db =

(1.61)

N1d
sin ( )
2



Na
2

{a0 + a2 cos (2 2 )} rl d d
0
cos
2
3



2
= Ma1d1 cos
3
+

M1dc =

(1.62)

N1d
sin ( )
2


2
Na
cos +
2
3

{a0 + a2 cos (2 2 )} rl d d



2
= Ma1d1 cos +
3

M1d f =

(1.63)

N1d
sin ( )
2

Nf

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2

N f N1d
=
0 rl (2a0 + a2 )
8

(1.64)

16

1 Synchronous Generator
+

L 1d = L 1dl +

N1d
sin ( )
2

= L 1dl +

2
N1d
0 rl (2a0 + a2 )
8

M1d1q =

(1.65)

N1d
sin ( )
2

N1d

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2

N1q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

=0

(1.66)

M1d2q =

N1d
sin ( )
2

N2q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

=0

(1.67)

+/2

M1qa =
/2

N1q
cos ( )
2

Na

cos {a0 + a2 cos (2 2 )} rl d d


2

= Ma1q1 sin

(1.68)

1.2 Model with Three Damper Windings


+/2

M1qb =
/2

17

N1q
cos ( )
2



Na
2

{a0 + a2 cos (2 2 )} rl d d
0
cos
2
3



2
= Ma1q1 sin
3
+/2

M1qc =
/2

(1.69)

N1q
cos ( )
2


2
Na
cos +
2
3

{a0 + a2 cos (2 2 )} rl d d



2
= Ma1q1 sin +
3
+/2

M1q f =
/2

(1.70)

N1q
cos ( )
2

Nf

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2

=0

(1.71)

+/2

M1q1d =
/2

N1q
cos ( )
2

=0

N1d

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2
(1.72)

18

1 Synchronous Generator
+/2

L 1q = L 1ql
/2

2
N1q

= L 1ql +

+/2

M1q2q =
/2

(1.73)

N1q
cos ( )
2

N1q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

0 rl (2a0 a2 )

N1q
cos ( )
2

N2q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

N1q N2q
0 rl (2a0 a2 )
8
+/2

M2qa =
/2

N2q
cos ( )
2

Na

cos {a0 + a2 cos (2 2 )} rl d d


2

= Ma2q1 sin
+/2

M2qb =
/2

(1.74)

(1.75)

N2q
cos ( )
2



Na
2

{a0 + a2 cos (2 2 )} rl d d
0
cos
2
3



2
= Ma2q1 sin
3

(1.76)

1.2 Model with Three Damper Windings


+/2

M2qc =
/2

N2q
cos ( )
2


19

Na
2
cos +
2
3

{a0 + a2 cos (2 2 )} rl d d



2
= Ma2q1 sin +
3
+/2

M2q f =
/2

(1.77)

N2q
cos ( )
2

Nf

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2

=0

(1.78)

+/2

M2q1d =
/2

N2q
cos ( )
2

N1d

cos ( ) {a0 + a2 cos (2 2 )} rl d d


2

=0

(1.79)

+/2

M2q1q =
/2

N2q
cos ( )
2

N1q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

N1q N2q
=
0 rl (2a0 a2 )
8

(1.80)

20

1 Synchronous Generator
+/2

L 2q = L 2ql
/2

= L 2ql +

N2q
cos ( )
2

2
N2q

N2q

sin ( ) {a0 + a2 cos (2 2 )} rl d d


2

0 rl (2a0 a2 )

(1.81)

where
Na2
0 rla0
8
Na N f
0 rl(2a0 + a2 )
8
Na N1d
0 rl(2a0 + a2 )
8
Na N1q
0 rl(2a0 a2 )
8
Na N2q
0 rl(2a0 a2 )
8

Mab0 

(1.82)

Ma f 1 

(1.83)

Ma1d1 
Ma1q1 
Ma2q1 

(1.84)
(1.85)
(1.86)

It is to be noted that the order of the subscripts in the notation for mutual inductances
does not affect the expression for mutual inductance. For example, Mab = Mba .
In general, the number of field poles is p f . If the total number of turns in a winding
is N , the number of turns per field pole pair is 2N / p f . The inductance of the part of a
winding with 2N / p f turns is equal to 2/ p f times the expression derived above, if
is the electrical angle. The total inductance is obtained by multiplying this expression
by p f /2. Hence, the inductance expressions derived above are valid even if p f > 2.
The mechanical angle m is related to the electrical angle by
=

pf
m
2

(1.87)

1.2.3 Equations of Motion


If friction and windage losses are neglected, by Newtons law,
J

d2 m
= Tm Te
dt 2

(1.88)

where J is the combined moment of inertia of the rotor and the prime mover, Tm is
the mechanical torque, Te is the electromagnetic or electrical torque, and m is the
position of the rotor in mechanical radians. Equation (1.88) in electrical angle is

1.2 Model with Three Damper Windings

21

2 d2
J
= Tm Te
p f dt 2

(1.89)

Multiplying by 2/ p f gives


2
pf

2
J

d2
2
2
=
Tm
Te
2
dt
pf
pf

(1.90)

2

Defining J   2/ p f J , Tm  2Tm / p f , and Te  2Te / p f and substituting in
(1.90) give
J

d2
= Tm Te
dt 2

(1.91)

J  , Tm , and Te are the moment of inertia, mechanical torque, and electrical torque,
respectively, of an equivalent synchronous generator with two field poles [1].
Equation (1.91) can be written as the following two first-order equations.
d
=
dt
d
1
=  (Tm Te )
dt
J

(1.92)
(1.93)

is the speed of the rotor in electrical radian per second.

1.2.4 Expression for Electrical Torque


Let We be the total electrical energy supplied to the magnetic field of all windings.
From the circuit diagram in Fig. 1.3,
d f
d1q
d2q
dWe
da
db
dc
d1d
= ia
+ ib
+ ic
+if
+ i 1d
+ i 1q
+ i 2q
(1.94)
dt
dt
dt
dt
dt
dt
dt
dt

Let W f be the energy stored in the magnetic field of all windings.


T
ia
ib

ic

1
if
Wf =

2 i

1d
i 1q
i 2q

La
Mba

Mca

Mfa

M1da

M1qa
M2qa

Mab
Lb
Mcb
Mfb
M1db
M1qb
M2qb

Mac
Mbc
Lc
Mfc
M1dc
M1qc
M2qc

Ma f
Mb f
Mc f
Lf
M1d f
M1q f
M2q f

Ma1d
Mb1d
Mc1d
M f 1d
L 1d
M1q1d
M2q1d

Ma1q
Mb1q
Mc1q
M f 1q
M1d1q
L 1q
M2q1q

Let Wm be the mechanical work done by the magnetic field.


ia
Ma2q
ib
Mb2q


Mc2q
ic

M f 2q
if

M1d2q
i 1d
M1q2q i 1q
L 2q
i 2q
(1.95)

22

1 Synchronous Generator

dWm
dm
d
= Te
= Te
dt
dt
dt

(1.96)

By the law of conservation of energy,


dW f
dWe
dWm
=

dt
dt
dt

(1.97)

From (1.94) to (1.97),

T
ia
ib

ic

1
if
Te =

2
i 1d

i 1q
i 2q

La
Mba

d Mca

d M f a
M1da

M1qa

M2qa

Mab
Lb
Mcb
Mfb
M1db
M1qb
M2qb

Mac
Mbc
Lc
Mfc
M1dc
M1qc
M2qc

Ma f
Mb f
Mc f
Lf
M1d f
M1q f
M2q f

Ma1d
Mb1d
Mc1d
M f 1d
L 1d
M1q1d
M2q1d

Ma1q
Mb1q
Mc1q
M f 1q
M1d1q
L 1q
M2q1q

Ma2q
ia

Mb2q
i b
ic
Mc2q

M f 2q
i f (1.98)
i 1d
M1d2q

M1q2q i 1q
L 2q
i 2q

1.3 Parks Transformation


In the model obtained in the last section, there are inductances which are timevarying since they depend on , and varies with time. The model can be simplified
by transformation of stator variables. Let

f
if
a
ia
va
1d
i 1d


s  b , is  ib , vs  vb , r 
1q , ir  i 1q
c
ic
vc
2q
i 2q
(1.99)
Equations (1.1)(1.3) and (1.8) can be written as

ds
= Ra i s vs
dt
s = L s i s + Msr ir
T
r = Msr
i s + L r ir

(1.100)
(1.101)
(1.102)

where

L a0 Mab0 Mab0
L s  Mab0 L a0 Mab0
Mab0 Mab0 L a0

cos(2 )
cos(2 2/3) cos(2 + 2/3)
(1.103)
cos(2 )
+ L a2 cos(2 2/3) cos(2 + 2/3)
cos(2 + 2/3)
cos(2 )
cos(2 2/3)

1.3 Parks Transformation

23

Ma f 1 cos
Ma1d1 cos
Msr  Ma f 1 cos( 2/3) Ma1d1 cos( 2/3)
Ma f 1 cos( + 2/3) Ma1d1 cos( + 2/3)

Ma2q1 sin
Ma1q1 sin
Ma1q1 sin( 2/3) Ma2q1 sin( 2/3)
Ma1q1 sin( + 2/3) Ma2q1 sin( + 2/3)

L f M f 1d 0
0
M f 1d L 1d
0
0

Lr 
0
0
L 1q M1q2q
0
0 M1q2q L 2q

(1.104)

(1.105)

Let s , i s , and vs be transformed to dq0 , i dq0 , and vdq0 , respectively, as follows:


dq0  TP s , i dq0  TP i s , vdq0  TP vs

(1.106)

where TP is a 3 3 nonsingular matrix and

dq0

d
id
vd
 q , i dq0  i q , vdq0  vq
0
i0
v0

(1.107)

Using this transformation in (1.101) gives


TP1 dq0 = L s TP1 i dq0 + Msr ir

(1.108)

Pre-multiplying by TP gives
dq0 = TP L s TP1 i dq0 + TP Msr ir

(1.109)

Let TP be chosen such that TP L s TP1 is a diagonal matrix [5]; then, the transformation
is known as Parks transformation. The columns of TP1 are right eigenvectors of
L s .TP is given by

TP1

kd cos
kq sin
k0
= kd cos ( 2/3) kq sin ( 2/3) k0
kd cos ( + 2/3) kq sin ( + 2/3) k0

(1.110)

where kd , kq , and k0 can be chosen arbitrarily. The diagonal elements of TP L s TP1


are the eigenvalues of L s .
The power at the synchronous generator terminals is

T
T
TP1 TP1 i dq0
P = vsT i s = vdq0

(1.111)

24

1 Synchronous Generator

T

The matrix TP1 TP1 is diagonal, given by


TP1

T

TP1

3kd2 /2 0
0
= 0 3kq2 /2 0
0
0 3k02

(1.112)

T i
The transformation is said to be power invariant if P = vdq0
dq0 [1, 2]. Power

invariance is satisfied if TP1 = TPT . The values of kd , kq , and k0 for power invariance
are



2
2
1
, kq =
, k0 =
(1.113)
kd =
3
3
3

If positive values are used,

2 cos 2 cos ( 2/3) 2 cos ( + 2/3)


1
TP =
2 sin 2 sin ( 2/3) 2 sin ( + 2/3)
3
1
1
1

(1.114)

Using this TP in (1.109) gives





if
Ld 0 0
Md f Md1d 0
id
d
0
i 1d

q = 0 L q 0 i q + 0
0 Mq1q Mq2q
i 1q
0
i0
0
0
0
0
0 0 L0
i 2q
(1.115)

where
L d  L a0 Mab0 +

Md f 

3
Ma f 1 ,
2

3
L a2 ,
2


Md1d 

L q  L a0 Mab0
3
Ma1d1 ,
2


Mq1q 

3
L a2 ,
2

L 0  L a0 + 2Mab0

3
Ma1q1 ,
2


Mq2q 

(1.116)
3
Ma2q1
2
(1.117)

L d , L q , and L 0 are called d-axis inductance, q-axis inductance, and zero sequence
inductance, respectively. From (1.102) and (1.106),


f
Md f
0
1d Md1d 0

=
1q 0 Mq1q
2q
0 Mq2q

if
0
0
L f M f 1d 0
i
d

0
0
0
i 1d (1.118)
i + M f 1d L 1d
0
0
L 1q M1q2q i 1q
0 q
i0
0
0
0 M1q2q L 2q
i 2q

1.3 Parks Transformation

25

From (1.100) and (1.106),




d TPT dq0
= Ra TPT i dq0 TPT vdq0
dt

(1.119)

Pre-multiplying by TP gives
ddq0
= Mdq0 Ra i dq0 vdq0
dt

(1.120)

where

010
M 1 0 0
000

(1.121)

From (1.98),
Te

 

1  T T  dL s /d dMsr /d
is
= i s ir
T /d
i
dM
O
2
r
sr

(1.122)

where O is a null matrix. Substituting TPT i dq0 for i s in (1.122) and simplifying give
dL s T
dMsr
1 T
T
T i dq0 i dq0
ir = d i q q i d
TP
TP
Te = i dq0
2
d P
d

(1.123)

If o is the steady-state or operating speed, then


= o t +

(1.124)

is the angular position of the rotor in electrical radian with respect to a reference
rotating at speed o .
With the transformation of stator variables and use of instead of , the equations
governing the synchronous generator are
dd
dt
dq
dt
d0
dt
d f
dt
d1d
dt

= q Ra i d vd

(1.125)

= d Ra i q vq

(1.126)

= Ra i 0 v0

(1.127)

= R f i f + v f

(1.128)

= R1d i 1d

(1.129)

26

1 Synchronous Generator

d1q
dt
d2q
dt
d
dt
d
dt
d

= R1q i 1q

(1.130)

= R2q i 2q

(1.131)

= o

(1.132)


1  
Tm d i q + q i d

J
= L d i d + Md f i f + Md1d i 1d

(1.133)

(1.134)

q = L q i q + Mq1q i 1q + Mq2q i 2q
f = Md f i d + L f i f + M f 1d i 1d

(1.135)
(1.136)

1d = Md1d i d + M f 1d i f + L 1d i 1d
1q = Mq1q i q + L 1q i 1q + M1q2q i 2q
2q = Mq2q i q + M1q2q i 1q + L 2q i 2q

(1.137)
(1.138)
(1.139)

0 = L 0 i 0

(1.140)

Parks transformation results in constant inductances. It is apparent that Parks transformation results in the replacement of the stator windings by three windings whose
voltage, current, and flux linkage are vdq0 , i dq0 , and dq0 , respectively. From (1.134)
to (1.139), it is apparent that one winding is on the d-axis and one winding is on the
q-axis; hence, the voltage, current, and flux linkage have notations with the subscripts
d and q. However, the presence of the first term on the right-hand side of (1.125)
and (1.126) suggests that this is not true [6]. The quantities having a notation with
the subscript 0 are zero sequence quantities.

1.4 Transformation of Rotor Variables


In order to obtain a model whose parameters can be determined, the model obtained
in the previous section is partitioned into three parts as shown in Fig. 1.8.
An equivalent synchronous generator model is shown in Fig. 1.9. Parts 2 and 3 in
Fig. 1.8 are replaced by the following equations in Fig. 1.9.
Id (s) = G 1 (s)d (s) + G 2 (s)V f (s)
Iq (s) = G 3 (s)q (s)

(1.141)
(1.142)

where Id (s), d (s), V f (s), Iq (s), and q (s) are Laplace transform of i d , d , v f ,
i q , and q , respectively, and G 1 (s), G 2 (s), and G 3 (s) are transfer functions. The
following three transfer functions are derived below.
G 1 (s) =


Id (s) 
d (s) v f =0

(1.143)

1.4 Transformation of Rotor Variables

27
vf

vd

vq

v0

d f

Tm

dt
d 1d

d d
dt
d q
dt
d 0
dt
d
dt
d

= q R a i d v d

f = M df i d + L f i f + M f 1d i 1d

id

Part 1

d 1q

= o

dt
J
0 = L 0 i0

= R 1d i 1d

1d = M d 1d i d + M f 1d i f + L 1d i 1d

= R a i0 v0

Part 2

d = L d i d + M df i f + M d 1d i 1d

= d R a i q v q

dt

= Rf if + vf

dt
d 2q

Tm d i q + q i d

dt

iq

Part 3

= R 1q i 1q
= R 2q i 2q

q = L q i q + M q1q i 1q + M q2q i 2q
1q = M q1q i q + L 1q i 1q + M 1q2q i 2q
2q = M q2q i q + M 1q2q i 1q + L 2q i 2q

Fig. 1.8 Partition of the synchronous generator model


vd
d d
dt
d q
dt
d 0
dt
d
dt

vq

v0

Tm

= q R a i d v d
= d R a i q v q

vf

d
Id ( s)= G 1 ( s) d ( s)+ G 2 ( s)V f ( s)
id

= R a i0 v0
= o

1
Tm d i q + q i d
=
dt
J
0 = L 0 i0

Iq ( s)= G 3 ( s) q ( s)
iq

Fig. 1.9 Equivalent synchronous generator model


V f (s) 
G 1 (s)
=

G 2 (s)
d (s) id =0
Iq (s)
G 3 (s) =
q (s)

(1.144)
(1.145)

28

1 Synchronous Generator

1.4.1 Transfer Function Id (s)/d (s)|v f =0


With v f = 0, the Laplace transform of equations in part 2 of Fig. 1.8 gives
s f (s) = R f I f (s)
s1d (s) = R1d I1d (s)
d (s) = L d Id (s) + Md f I f (s) + Md1d I1d (s)
f (s) = Md f Id (s) + L f I f (s) + M f 1d I1d (s)
1d (s) = Md1d Id (s) + M f 1d I f (s) + L 1d I1d (s)

(1.146)
(1.147)
(1.148)
(1.149)
(1.150)

Since the purpose is to obtain a transfer function, f (0) and 1d (0) do not appear
in (1.146) and (1.147), respectively. Elimination of I f (s), I1d (s), f (s), and 1d (s)
from (1.146) to (1.150) gives

L f L 1d M 2f 1d + (L f R1d + L 1d R f )/s + R f R1d /s 2
Id (s) 
= 

2
d (s) v f =0
L d L f L 1d L d M 2f 1d L 1d Md2 f + 2Md f M f 1d Md1d L f Md1d

2 R )/s + L R R /s 2
+(L d L f R1d + L d L 1d R f Md2 f R1d Md1d
f
d f 1d

(1.151)
This transfer function can be written as






1 + sTdo
1 + sTdo
Id (s) 



=
d (s) v f =0
L d 1 + sTd 1 + sTd


1.4.2 Transfer Function V f (s)/d (s) i

(1.152)

d =0

Laplace transform of equations in part 2 of Fig. 1.8, with i d = 0, gives


s f (s) = R f I f (s) + V f (s)
s1d (s) = R1d I1d (s)
d (s) = Md f I f (s) + Md1d I1d (s)

(1.153)
(1.154)
(1.155)

f (s) = L f I f (s) + M f 1d I1d (s)


1d (s) = M f 1d I f (s) + L 1d I1d (s)

(1.156)
(1.157)

Elimination of I f (s), I1d (s), f (s), and 1d (s) from (1.153) to (1.157) gives

L f L 1d M 2f 1d + (L f R1d + L 1d R f )/s + R f R1d /s 2
V f (s) 
=
(1.158)
d (s) id =0
(Md f L 1d Md1d M f 1d )/s + Md f R1d /s 2

1.4 Transformation of Rotor Variables

29

This can be written as








1 + sTdo
R f 1 + sTdo
V f (s) 


=

d (s) id =0
Md f 1 + sTdc

(1.159)

1.4.3 Transfer Function Iq (s)/q (s)


Laplace transform of equations in part 3 of Fig. 1.8 gives
s1q (s) = R1q I1q (s)

(1.160)

s2q (s) = R2q I2q (s)


q (s) = L q Iq (s) + Mq1q I1q (s) + Mq2q I2q (s)

(1.161)
(1.162)

1q (s) = Mq1q Iq (s) + L 1q I1q (s) + M1q2q I2q (s)


2q (s) = Mq2q Iq (s) + M1q2q I1q (s) + L 2q I2q (s)

(1.163)
(1.164)

Elimination of I1q (s), I2q (s), 1q (s), and 2q (s) from (1.160) to (1.164) gives
2
L 1q L 2q M1q2q
+ (L 1q R2q + L 2q R1q )/s + R1q R2q /s 2
Iq (s)

= 
2
2 + 2M
2
q (s)
L 2q Mq1q
L q L 1q L 2q L q M1q2q
q1q M1q2q Mq2q L 1q Mq2q
2 R M 2 )/s + L R R /s 2
+(L q L 1q R2q + L q L 2q R1q R2q Mq1q
q 1q 2q
1q q2q

(1.165)
This can be written as





1
+
sT
1
+
sT
qo
qo
Iq (s)



=
q (s)
L q 1 + sTq 1 + sTq

(1.166)

1.5 Tests for the Determination of Parameters


The parameters in (1.152) and (1.159) are called d-axis parameters, and those in
(1.166) are called q-axis parameters. These parameters can be determined by tests.
This section describes one type of tests known as standstill frequency response tests
in which the rotor is standstill at a certain position.

1.5.1 Determination of d-Axis Parameters


There are two tests to be conducted to determine the d-axis parameters. The circuit
diagram for the first test is shown in Fig. 1.10 [1, 6]. The rotor position should be

30

1 Synchronous Generator
i
A
ib

+
v

vb

vc
+

va +

ia
f

ic

Fig. 1.10 Circuit diagram for the first test for the determination of d-axis parameters

such that = 0. Since the rotor is at standstill, = 0. From the circuit diagram in
Fig. 1.10,
v = va vb

(1.167)

vb = vc
i = i a = i b + i c

(1.168)
(1.169)

vf = 0

(1.170)

Hence,


2
v
3

3
id =
i
2
Id (s)
3 I (s)
=
Vd (s)
2 V (s)

(1.172)

dd
= Ra i d vd
dt

(1.174)

vd =

(1.171)

(1.173)

Since = 0, from (1.125),

Laplace transform of this equation gives



d (s) 
Ra
Vd (s)

=

Id (s) v f =0
s Id (s)
s
From (1.173) and (1.175),

(1.175)

1.5 Tests for the Determination of Parameters


Fig. 1.11 Circuit diagram
for the second test for the
determination of d-axis
parameters

31

vf

+
+
vb

V
a

vc
+

va


Id (s) 
1
=

d (s) v f =0
[2/(3s)]V (s)/I (s) Ra /s

(1.176)

The magnitude and phase angle of this transfer function are obtained at different
frequencies.
The circuit diagram for the second test is shown in Fig. 1.11 [6]. The rotor position
should be such that = 0. From the circuit diagram in Fig. 1.11,

vd =
id = 0

2
v
3

(1.177)
(1.178)

Since = 0 and i d = 0, from (1.125),


dd
= vd
dt

(1.179)



V f (s) 
3 V f (s)
s
=

d (s) id =0
2 V (s)

(1.180)

From (1.177) and (1.179),

The magnitude and phase angle of this transfer function are obtained at different frequencies. From the values of the magnitude and
 the phase angle of the two transfer
functions Id (s)/d (s)|v f =0 and V f (s)/d (s)i =0 at different frequencies, the vald
 , T  , T  , T  , R /M , and T  are estimated.
ues of the d-axis parameters L d , Tdo
f
df
do d
d
dc
 , and T  are real and positive, and
It can be verified from the tests that Td , Td , Tdo
do
 is positive. The notations are such that T  > T  and T  > T  .
Tdc
d
d
do
do

32

1 Synchronous Generator

Fig. 1.12 Circuit diagram for


the test for the determination
of q-axis parameters

i
A
ib
+

+
vb

ia

va +

vc
+ c
ic

1.5.2 Determination of q-Axis Parameters


The q-axis parameters are determined by conducting a test, the circuit diagram for
which is shown in Fig. 1.12 [1, 6]. The rotor position should be such that = 90 .
Since the rotor is at standstill, = 0. From the circuit diagram in Fig. 1.12,


2
v
3

3
iq =
i
2
Vq (s)
2 V (s)
=
Iq (s)
3 I (s)

(1.182)

dq
= Ra i q vq
dt

(1.184)

vq =

(1.181)

(1.183)

Since = 0, from (1.126),

Laplace transform of this equation gives


q (s)
Vq (s)
Ra
=

Iq (s)
s Iq (s)
s

(1.185)

Iq (s)
1
=
q (s)
[2/(3s)]V (s)/I (s) Ra /s

(1.186)

From (1.183) and (1.185),

From the values of the magnitude and the phase angle of this transfer function at
 , T  , T  , and T 
different frequencies, the values of the q-axis parameters L q , Tqo
qo q
q

1.5 Tests for the Determination of Parameters

33

 , and T  are real and


are estimated. It can be verified from the test that Tq , Tq , Tqo
qo




positive. The notations are such that Tq > Tq and Tqo > Tqo .

1.6 Time Domain Model with Standard Parameters


From (1.141) to (1.145), (1.152), (1.159), and (1.166),








1 + sTdo
1 + sTdo
Md f 1 + sTdc


 d (s)


 V f (s)
Id (s) =
L d 1 + sTd 1 + sTd
R f L d 1 + sTd 1 + sTd
(1.187)





1 + sTqo
1 + sTqo


 q (s)
Iq (s) =
(1.188)
L q 1 + sTq 1 + sTq
By partial fraction expansion,

Id (s) = r1 +

Iq (s) = r6 +

r2
r3
+
1 + sTd
1 + sTd
r7
r8
+

1 + sTq
1 + sTq

Md f
d (s) +
Rf

r4
r5
+
1 + sTd
1 + sTd


V f (s)
(1.189)

!
q (s)

(1.190)

where
r1 

 T 
 )(T  T  )
 )(T  T  )
Tdo
(T  Tdo
(T  Tdo
do
d
do
d
do
, r2  d
(1.191)
, r3  d






L d Td Td
L d Td (Td Td )
L d Td (Td Td )

r4 


 T 
Td Tdc
Tdc
d
, r5 


L d (Td Td )
L d (Td Td )

r6 

 T 
Tqo
qo

L q Tq Tq

, r7 

 )(T  T  )
(Tq Tqo
q
qo

L q Tq (Tq

Tq )

(1.192)
, r8 

 )(T  T  )
(Tq Tqo
q
qo

L q Tq (Tq Tq )

(1.193)

Equations (1.189) and (1.190) can be put in the block diagram form shown in
Figs. 1.13 and 1.14 where
Ef 

Md f
vf
Rf

(1.194)

The time domain model is obtained with F , 1D , 1Q , and 2Q as the state


variables associated with the rotor windings. The time domain model will be obtained
in terms of the inductances L d and L q and other inductances defined below, instead
of the residues r1 to r8 .

34

1 Synchronous Generator
r1

1
1 + sTd

1
1 + sTd

1D

+
+

r2

id

r4
r2
Ef
r5
r3
+
+

r3

Fig. 1.13 Block diagram form of (1.189)


r6

1Q

r7

1 + sTq

iq

2Q

r8

1 + sTq

Fig. 1.14 Block diagram form of (1.190)

L d and L q are related to the residues by


Ld =

1
,
r1 + r2 + r3

Lq =

1
r6 + r7 + r8

(1.195)

An inductance L d called d-axis subtransient inductance and an inductance L q called
q-axis subtransient inductance are defined [4] as
L d


d (s) 
 lim
,
s Id (s) 
v f =0

L q  lim

q (s)
Iq (s)

(1.196)

1.6 Time Domain Model with Standard Parameters

35

From (1.189), (1.190), and (1.196),


L d =

1
,
r1

L q =

1
r6

(1.197)

An inductance L d called d-axis transient inductance and an inductance L q called


q-axis transient inductance are defined as
L d 

1
,
r1 + r3

L q 

1
r6 + r8

(1.198)

These inductances satisfy the inequality relations: L d > L d > L d > 0 and L q > L q
> L q > 0.
 , T  , T  , and T 
Td , Td , Tq , and Tq are called short-circuit time constants. Tdo
qo
do qo
are called open-circuit time constants. These time constants and the inductances
L d , L d , L d , L q , L q , and L q are called standard parameters [2, 6]. The synchronous
generator model with standard parameters is given by the following equations:
dd
dt
dq
dt
d0
dt
d F
dt

= q Ra i d vd

(1.199)

= d Ra i q vq

(1.200)

= Ra i 0 v0


L  (T  T  )
1
=  F + d +  d dc  d  E f
Td
(L d L d )(Td Td )


 T  )
L d L d (Tdc
1
d
=  1D + d +
Ef
Td
L d (L d L d )(Td Td )

1 
=  1Q + q
Tq

1 
=  2Q + q
Tq

(1.201)

d1D
dt
d1Q
dt
d2Q
dt
d
= o
dt

1 
d
=  Tm d i q + q i d
dt
J




1
1
1
1
1
i d =  d +
 F +
 1D
Ld
Ld
Ld
L d
Ld
!
!
1
1
1
1
1
i q =  q +
 1Q +
 2Q
Lq
Lq
Lq
L q
Lq
i0 =

1
0
L0

(1.202)
(1.203)
(1.204)
(1.205)
(1.206)
(1.207)
(1.208)
(1.209)
(1.210)

36

1 Synchronous Generator

b
a

vf

Fig. 1.15 Three-phase short circuit at the synchronous generator terminals

1.7 Time Constants


1.7.1 Short-Circuit Time Constants
Td , Td , Tq , and Tq are called short-circuit time constants since these are the time
constants during short circuit at the synchronous generator terminals. Suppose the
synchronous generator is run at = o , and then, its terminals are shorted by closing
the switch shown in Fig. 1.15. During this condition, vd = vq = v0 = i 0 = 0. If Ra is
neglected, the equations governing the synchronous generator during this condition
are
dd
= o q
dt
dq
= o d
dt


L  (T  T  )
1
d F
=  F + d +  d dc  d  E f
dt
Td
(L d L d )(Td Td )


 T  )
L d L d (Tdc
d1D
1
d
=  1D + d +
Ef
dt
Td
L d (L d L d )(Td Td )

d1Q
1 
=  1Q + q
dt
Tq

1 
d2Q
=  2Q + q
dt
Tq

(1.211)
(1.212)
(1.213)
(1.214)
(1.215)
(1.216)

The eigenvalues of the system during this condition are 1/Td , 1/Td , 1/Tq ,
1/Tq , jo , jo . Hence, Td , Td , Tq , and Tq are called short-circuit time constants.

1.7 Time Constants

37

b
a

+
f

vf

Fig. 1.16 Field excited under open-circuit condition

1.7.2 Open-Circuit Time Constants


 , T  , T  , and T  are called open-circuit time constants since these are the time
Tdo
qo
do qo
constants under open-circuit condition. Suppose the synchronous generator is run at
= o , and then, the switch shown in Fig. 1.16 is closed. During this condition,
i d = i q = i 0 = 0. The equations governing the synchronous generator during this
condition are



 T  )
L d (Tdc
1
d F
d
=  F + d + 
Ef
dt
Td
(L d L d )(Td Td )


 T  )
L d L d (Tdc
d1D
1
d
=  1D + d +
E
f
dt
Td
L d (L d L d )(Td Td )

d1Q
1 
=  1Q + q
dt
Tq

1 
d2Q
=  2Q + q
dt
Tq




1
1
1
1
1
0 =  d +
 F +
 1D
Ld
Ld
Ld
L d
Ld
!
!
1
1
1
1
1
0 =  q +
 1Q +
 2Q
Lq
Lq
Lq
L q
Lq

(1.217)
(1.218)
(1.219)
(1.220)
(1.221)
(1.222)

The last two equations imply that d and q are not independent and can be
expressed in terms of state variables. Using (1.191), (1.193), (1.197), and (1.198),
it can be shown that the eigenvalues of the system during this condition are
 , 1/T  , 1/T  , 1/T  . Hence, T  , T  , T  , and T  are called open1/Tdo
qo
qo
qo
do
do do qo
circuit time constants.

38

1 Synchronous Generator

1.8 Model in Per Unit Quantities


It is convenient to analyze the equations governing a power system, if the values of
the quantities (variables and parameters) are normalized by dividing them by their
respective base values. The normalized value is said to be in per unit of the base
value. The base values of some quantities can be chosen independently, and from
these values, the base values of other quantities are obtained. If the base values of
angular frequency, power, and voltage are chosen, the base values of other quantities
can be obtained. Base angular frequency and base power are same for the entire
power system, whereas base voltage is different on the two sides of a transformer.
The obvious choice for base angular frequency B is the rated or nominal angular
frequency. A choice for the base power S B is the rated voltamperes of the largest
synchronous generator in the system. Base voltage VB is taken as the rated rms value
of line-to-line voltage. The base values of other quantities are obtained as follows:
Base current,

IB 

SB
VB

VB
IB
VB
Base flux linkage, B 
B
B
Base inductance, L B 
IB
SB
Base torque, TB 
B
Base impedance,

ZB 

(1.223)
(1.224)
(1.225)
(1.226)
(1.227)

The quantities in per unit are denoted by a bar over the notation; for example,
d  d / B . Dividing (1.199)(1.205) by B gives
d d
=
dt
d q
=
dt
d 0
=
dt
d F
=
dt
d 1D
dt

q B R a id B v d

(1.228)

d B R a iq B v q

(1.229)

B R a i0 B v 0


 T  )
X d (Tdc
1
d
F + d +
E f
Td
( X d X d )(Td Td )


 T  )
X d X d (Tdc
1
d
=  1D + d +
E f
Td
Xd ( X d X d )(Td Td )


1 
d 1Q
=  1Q + q
dt
Tq

(1.230)
(1.231)
(1.232)
(1.233)

1.8 Model in Per Unit Quantities

39


d 2Q
1 
=  2Q + q
dt
Tq

(1.234)

where X d  B L d , X d  B L d , and X d  B L d . It is to be noted that X d =


L d , X d = L d , and X d = L d .
Equation (1.207) can be written as
d
SB
= 
dt
J B

q i d
Tm
d i q

+
TB
B IB
B IB


(1.235)

A parameter called inertia constant denoted by H is defined [13] as


H

J  2B
2S B

(1.236)

Equation (1.235) can be written as


d
B 
=
(T d iq + q id )
dt
2H m

(1.237)

Dividing (1.208)(1.210) by I B gives


1
id =  d +

Xd

1
1


Xd
Xd

1
iq =  q +
X q

1
1

X q
X q

1
0
i0 =

X0

!
F +
!
1Q +

1
1



Xd
Xd

1
1

X q
X q

1D

(1.238)

!
2Q

(1.239)
(1.240)

where X q  B L q , X q  B L q , X q  B L q , and X 0 = B L 0 . It is to be


noted that X q = L q , X q = L q , X q = L q , and X 0 = L 0 . Equations (1.206),
(1.228)(1.234), and (1.237)(1.240) form the synchronous generator model in per
unit quantities.

1.9 Other Models


The equations obtained in the last section give a detailed model of the synchronous
generator. The choice of the model depends on the application. Some applications do
not need a detailed model. The simpler models in the order of decreasing complexity
are given below.

40

1 Synchronous Generator

1.9.1 Model in the Absence of Zero Sequence Variables and with


Tdc = Td
 = T  ,
If the zero sequence variables are equal to zero and if it is assumed that Tdc
d
the equations governing the synchronous generator are

d d
= q B R a id B v d
dt
d q
= d B R a iq B v q
dt
!
X d
d F
1
E f
=  F + d +
dt
Td
X d X d

1 
d 1D
=  1D + d
dt
Td

d 1Q
1 
=  1Q + q
dt
Tq

1 
d 2Q
=  2Q + q
dt
Tq
d
= o
dt
B 
d
=
(T d iq + q id )
dt
2H m
!
!
1
1
1
1
1
 F +
 1D
id =  d +
X d
X d
X d
X d
X d
!
!
i q = 1 q + 1 1 1Q + 1 1 2Q
X q
X q
X q
X q
X q

(1.241)
(1.242)
(1.243)
(1.244)
(1.245)
(1.246)
(1.247)
(1.248)
(1.249)
(1.250)

1.9.2 Model with Stator Transients Neglected


If high-frequency transients are not of interest, stator transients are neglected by
neglecting the terms d d /dt and d q /dt in (1.241) and (1.242), respectively. Then,
these equations become algebraic equations.
0 = q B R a id B v d
0 = d B R a iq B v q

(1.251)
(1.252)

1.9 Other Models

41

Another assumption made is that = B in (1.251) and (1.252). The synchronous


generator model is given by the following equations:
1
d F
= 
dt
Td

F + d +

X d

X d X d

E f

(1.253)


d 1D
1 
=  1D + d
dt
Td

d 1Q
1 
=  1Q + q
dt
Tq


1 
d2Q
=  2Q + q
dt
Tq
d
= o
dt
B 
d
=
(T d iq + q id )
dt
2H m
0 = q R a id v d
0 = d R a iq v q
!
!
i d = 1 d + 1 1 F + 1 1 1D
X d
X d
X d
X d
X d
!
!
1
1
1
1
1
iq =  q +
 1Q +
 2Q
X q
X q
X q
X q
X q

(1.254)
(1.255)
(1.256)
(1.257)
(1.258)
(1.259)
(1.260)
(1.261)
(1.262)

1.9.3 Two Axis Model


The time constants Td and Tq are set to zero, and hence,
1D = d
2Q = q

(1.263)
(1.264)

The synchronous generator model is given by the following equations:


1
d F
= 
dt
Td

F + d +


1 
d 1Q
=  1Q + q
dt
Tq

X d

X d X d

!
E f

(1.265)
(1.266)

42

1 Synchronous Generator

d
= o
dt
d
B 
=
(T d iq + q id )
dt
2H m
0 = q R a id v d
0 = d R a iq v q
!
1
1
1
 F
id =  d +
X d
X d
X d
!
1
1
1
iq =  q +
 1Q
X q
X q
X q

(1.267)
(1.268)
(1.269)
(1.270)
(1.271)
(1.272)

1.9.4 One Axis (Flux Decay) Model


The time constant Tq is set to zero, and hence,
1Q = q

(1.273)

The synchronous generator model is given by the following equations:


1
d F
= 
dt
Td

F + d +

X d

X d X d

d
= o
dt
B 
d
=
(T d iq + q id )
dt
2H m
0 = q R a id v d
0 = d R a iq v q
!
i d = 1 d + 1 1 F
X d
X d
X d
1
q
iq =
X q

!
E f

(1.274)
(1.275)
(1.276)
(1.277)
(1.278)
(1.279)
(1.280)

1.9 Other Models

43

1.9.5 Classical Model


Td is assumed to be infinite. This implies d F /dt = 0, i.e., F is constant. It is also
assumed that R a = 0 and X q = X d . Since R a = 0, q = vd and d = v q . The
synchronous generator model is given by the following equations:
d
= o
dt
B 
d
=
(T v q iq v d id )
dt
2H m
E
1
id =  v q 
X
X
d

1
iq =  v d

Xd

(1.281)
(1.282)
(1.283)

(1.284)

where E  ( X d X d ) F / X d is a constant.

References
1. K.R. Padiyar, Power System Dynamics: Stability and Control, 2nd edn. (BS Publications,
Hyderabad, 2002)
2. P. Kundur, Power System Stability and Control (Tata McGraw-Hill, Noida, 1994)
3. P.W. Sauer, M.A. Pai, Power System Dynamics and Stability (Pearson Education, Singapore,
1998)
4. P.C. Krause, Analysis of Electric Machinery (McGraw-Hill, New York, 1987)
5. P.M. Anderson, B.L. Agrawal, J.E. Van Ness, Subsynchronous Resonance in Power Systems
(IEEE Press, New York, 1990)
6. A.M. Kulkarni, Power system dynamics and control (2012), http://www.nptel.iitm.ac.in/courses/
108101004/

Chapter 2

Transformer, Transmission Line, and Load

2.1 Transformer
2.1.1 Single-Phase Transformer
The single-phase transformer consists of a core and two or more windings. Figure 2.1
shows a transformer with two windings.
Let the transformer be ideal: the windings have zero resistance and the core has
infinite permeability [1]. Infinite permeability means that there is no flux outside the
core.
N1 and N2 are the number of turns in the windings. If the flux in the core is ,
the induced emfs in the windings are
d
dt
d
e2 = N2
dt
e1 = N1

(2.1)
(2.2)

From (2.1) and (2.2),


e1
N1
=
e2
N2

(2.3)

The relation between i 1 and i 2 is obtained from Amperes law. Due to infinite
permeability, the magnetic field intensity in the core is zero. Application of Amperes
law to the closed path in the core, shown in Fig. 2.1, gives
N2
i1
=
i2
N1

S Krishna, An Introduction to Modelling of Power System Components,


SpringerBriefs in Electrical and Computer Engineering,
DOI: 10.1007/978-81-322-1847-0_2, The Author(s) 2014

(2.4)

45

46

2 Transformer, Transmission Line, and Load

Fig. 2.1 Single-phase


transformer with two windings

Closed path
i2

i1

+
e1

e2

N2

N1

Fig. 2.2 Single-phase


transformer with three
windings

i2

i1
+
e1

N1

Fig. 2.3 Representation of


ideal transformer

+
e2

i3
+
e3

N2

N3

i1

i2
+
e1

+
e2

N1 N2

For the three-winding transformer shown in Fig. 2.2, where the number of turns
in the windings are N1 , N2 , and N3 ,
e1
N1 e1
N1
=
,
=
e2
N2 e3
N3
N1 i 1 = N2 i 2 + N3 i 3

(2.5)
(2.6)

The two-winding ideal transformer can be represented by the equivalent circuit


shown in Fig. 2.3. The dots shown at a terminal of each winding indicate the winding
terminals which simultaneously have the same polarity due to the emfs induced.
There are applications where the ideal transformer cannot be used. Then the
equivalent circuit of the transformer is given by Fig. 2.4. R1 and R2 are the resistances
of the two windings. Though the permeability of the core is high, it is not infinite,
and hence there is flux outside the core which links some or all turns of only one
winding and induces an emf. This flux is called leakage flux and its effect is modelled
by leakage inductances L 1 and L 2 . e1 and e2 are related by (2.3). Due to finite
permeability of the core, (2.4) is not exact but is used as an approximation.

2.1 Transformer

47
R1

L1

i2

i1

R2

L2

+
+
e1

v1

+
e2

v2

N1 N2

Fig. 2.4 Equivalent circuit of transformer

2.1.2 Three-Phase Transformer


A three-phase transformer can be obtained from three identical single-phase transformers. Figure 2.5 shows the equivalent circuit of the wye-wye-connected transformer. The two windings of a single-phase transformer are shown parallel to each
other.
Let v1a , v1b , and v1c be the potentials of the terminals 1a, 1b, and 1c, respectively,
with respect to the neutral. Let v2a , v2b , and v2c be the potentials of the terminals
2a, 2b, and 2c, respectively, with respect to the neutral. The equations for the wyewye-connected transformer are
di 1a
+ e1a
dt
di 1b
+ e1b
= i 1b R1 + L 1
dt
di 1c
+ e1c
= i 1c R1 + L 1
dt
N2
N1
N1 di 1a
=
e1a
i 1a R2
L2
N1
N2
N2
dt
N2
N1
N1 di 1b
=
e1b
i 1b R2
L2
N1
N2
N2
dt
N2
N1
N1 di 1c
=
e1c
i 1c R2
L2
N1
N2
N2
dt

v1a = i 1a R1 + L 1

(2.7)

v1b

(2.8)

v1c
v2a
v2b
v2c

(2.9)
(2.10)
(2.11)
(2.12)

Elimination of induced emfs from (2.7) to (2.12) gives

N2
N2
N2
N1
v2a =
v1a
R1 +
R2 i 1a
L1 +
N1
N1
N2
N1

N2
N2
N2
N1
v2b =
v1b
R1 +
R2 i 1b
L1 +
N1
N1
N2
N1

N2
N2
N2
N1
v2c =
v1c
R1 +
R2 i 1c
L1 +
N1
N1
N2
N1

di 1a
N1
L2
N2
dt

di 1b
N1
L2
N2
dt

di 1c
N1
L2
N2
dt

(2.13)
(2.14)
(2.15)

48

2 Transformer, Transmission Line, and Load


2c

1c
R1

L2

i1c
L1

i2 c
R2

+
N1
e 1c

e 1a +
N1

e 1b

N1

+
N2
e2c

R1

L1

e 2a +

1a
i1a

L2
2a
i2 a

+
R2

i2 b

L1

i1b

e2b

N2

N2

R2

L2

R1
2b

1b

Fig. 2.5 Wye-wye-connected transformer

In order to obtain the equations in per unit quantities, the equations are divided by
the base voltage. If V1B is the base voltage on the transformer side with N1 turns,
the base voltage on the other side of the transformer is
V2B 

N2
V1B
N1

(2.16)

The base values of other quantities are obtained as follows.


I1B =

SB
SB
V1B
V2B
, I2B =
, Z 1B =
, Z 2B =
V1B
V2B
I1B
I2B

(2.17)

Dividing (2.13)(2.15) by V2B gives

1
v 2a = v 1a R 1 + R 2 i a
B

1
v 2b = v 1b R 1 + R 2 i b
B

1
v 2c = v 1c R 1 + R 2 i c
B

X1 + X2

di a

dt

di b
X1 + X2
dt

di c
X1 + X2
dt

(2.18)
(2.19)
(2.20)

where X 1  B L 1 and X 2  B L 2 . The subscripts 1 and 2 are not necessary in the


notation for currents since i 1a = i 2a , i 1b = i 2b , and i 1c = i 2c .

2.1 Transformer

R1

49

i1c
L1
+
N1
e 1c

N1

i2 c

i1a

R2

R2

i2 b

+
N2
e2c

L2

e 1b
+

R2

e 2a +
N2

L1

i1b

L2

R1

L1

e 1a +

N1

e2b

N2

L2
i2 a

R1
Fig. 2.6 Wye-delta-connected transformer

e 1b

N1

R1

e2b

N2

i1c

L2

L1
L1

i1b

+
N1
e 1c

R1

R1
e 1a +
N1

i2 c

L1

R2

R2

i2 b

+
N2
e2c

L2
e 2a +
N2

i1a

R2

L2
i2 a

Fig. 2.7 Delta-delta-connected transformer

The equivalent circuit of the wye-delta-connected transformer is shown in Fig. 2.6.


For the wye-delta-connected transformer, if V1B is the base voltage on the transformer
side with N1 turns, the base voltage on the other side of the transformer is
N2
V1B
V2B 
3N1

(2.21)

For balanced sinusoidal operation, if resistance and leakage inductance are neglected,
the phase shift between terminal voltages on the two sides of the transformer is 30 .
The equivalent circuit of the delta-delta-connected transformer is shown in
Fig. 2.7. For the delta-delta-connected transformer, if V1B is the base voltage on the

50

2 Transformer, Transmission Line, and Load

transformer side with N1 turns, the base voltage on the other side of the transformer
is
V2B 

N2
V1B
N1

(2.22)

2.2 Transmission Line


A transmission line has four parameters: series resistance, series inductance, shunt
conductance, and shunt capacitance. These parameters are distributed uniformly
throughout the length of the transmission line. The series resistance in each phase is
denoted by R. For an overhead transmission line, the shunt conductance represents
the effects of leakage current over the surface of the insulator and corona. The shunt
conductance in each phase is denoted by G.
The expression for inductance and capacitance are derived for overhead transmission lines. The derivations assume that the conductors are straight.

2.2.1 Inductance
2.2.1.1 Transmission Line with Three Conductors
Let the transmission line consist of three conductors, one for each phase, of radius r
as shown in Fig. 2.8. Let the current in these conductors be i a , i b , and i c with uniform
current density. The expression for inductance is derived assuming that
ia + ib + ic = 0

(2.23)

It is assumed that the three conductors are transposed if not spaced symmetrically,
in order to have a symmetrical system; the transmission line is divided into three
sections of equal lengths and each conductor occupies each of the three positions 1,
2, and 3 for one third of the transmission line length. Let the conductors a, b, c occupy
positions 1, 2, 3, respectively, in the first section, positions 2, 3, 1, respectively, in
the second section, and positions 3, 1, 2, respectively, in the third section.
Consider a tube of radius x < r and thickness dx in phase a conductor in section 1
as shown in Fig. 2.8; the tube is coaxial with the conductor. Consider a filament in
this tube with cross-sectional area xdxd ; d is the angle subtended at the axis of the
conductor by the filament [2]. Consider the closed path consisting of this filament
and an arbitrarily located (at P) straight line parallel to the conductors. Let f a ,
f b , and f c be the flux linkage of this closed path in section 1, due to i a , i b , and
i c , respectively. The power delivered to this closed path, due to i a , is equal to

2.2 Transmission Line

51

3
r
D3

P
D23

b
D2

D31

D1

Filament

D12

d
x
a

dx
1

Fig. 2.8 Cross section of transmission line conductors

d f a xdxd
di a
1
ia
xdxd
=
fa
2
2
dt
r
r
dt

(2.24)

The power delivered to phase a, due to i a , is obtained by integrating this expression


over the cross-sectional area of the conductor as follows.
1 di a
p=
r 2 dt

2
=0

f a xdxd

(2.25)

x=0

Let aa , ab , and ac be the flux linkages of phase a in section 1 due to i a , i b , and


i c , respectively. The expression for p can also be written in terms of flux linkage of
phase a due to i a , as
p=

di a
daa
i a = aa
dt
dt

(2.26)

From (2.25) and (2.26),


aa =

1
r 2

2
=0

r
x=0

f a xdxd

(2.27)

52

2 Transformer, Transmission Line, and Load

Similarly, it can be shown that


2 r
1
=
f b xdxd
r 2 =0 x=0
2 r
1
=
f c xdxd
r 2 =0 x=0

ab
ac

(2.28)
(2.29)

Hence, the flux linkage of a phase is the average of the flux linkages of all the
filamentary closed paths in that phase.
The flux linkage of a phase can be determined from the flux densities due to
currents in the three conductors. The flux density Ba due to i a can be obtained from
Amperes law.

0 i a x 

2r 2
Ba =

0 i a
2 x 

if x  r
(2.30)
if x  r

where x  is the distance from the axis of conductor a and 0 is permeability of free
space; permeability of air and conductor are almost equal to that of free space. Then,
fa

l
=
3

D1
x

0 i a l
Ba dx =
6


1
x2
D1
2 + ln
2 2r
r

(2.31)

where l is the length of the transmission line. From (2.27) and (2.31),
aa

1
=
r 2

=0

r
x=0

0 i a l
6

1
x2
D1
2 + ln
2 2r
r

xdxd =

D1
0 i a l
ln  (2.32)
6
r

where r   e1/4 r .
The flux density due to i b is
Bb =

0 i b
2 D 

if D  r

(2.33)

where D  is the distance from the axis of conductor b. Then,


fb =

l
3

D2
D

Bb dD  =

D2
0 i b l
ln
6
D

(2.34)

2
1/2
where D = D12
+ x 2 2D12 x cos
. From (2.28) and (2.34),
ab

1
=
r 2

=0

r
x=0

D2
0 i b l
D2
0 i b l
ln
xdxd =
ln
6
D
6
D12

(2.35)

2.2 Transmission Line

53

Similarly,
ac =

D3
0 i c l
ln
6
D31

(2.36)

The flux linkage of phase a in section 1 is


a1 = aa + ab + ac

0 l
=
6

D1
D2
D3
(2.37)
i a ln  + i b ln
+ i c ln
r
D12
D31

Similarly, the flux linkage of phase a in sections 2 and 3, a2 and a3 , respectively,


are given by

0 l
D2
D3
D1
i a ln  + i b ln
+ i c ln
6
r
D23
D12

0 l
D3
D1
D2
i a ln  + i b ln
=
+ i c ln
6
r
D31
D23

a2 =

(2.38)

a3

(2.39)

The flux linkage of phase a is


a = a1 + a2 + a3

(2.40)

From (2.37) to (2.40),

0 l
(D1 D2 D3 )1/3
D1 D2 D3 1/3
D1 D2 D3 1/3
a =
+ i b ln
+ i c ln
i a ln
2
r
D12 D23 D31
D12 D23 D31

(2.41)
The coefficient of i a is self inductance and the coefficients of i b and i c are mutual
inductances. Using (2.23), the self and mutual inductances can be replaced by an
equivalent self inductance L.
L=

0 l (D12 D23 D31 )1/3


ln
2
r

(2.42)

2.2.1.2 Composite Conductors


A composite conductor consists of two or more individual conductors. Examples
of composite conductor are bundled conductor, stranded conductor, and conductor
of a multi-circuit transmission line. Figures 2.9, 2.10, and 2.11 show a double circuit transmission line, a transmission line with bundled conductors, and a stranded
conductor, respectively.

54

2 Transformer, Transmission Line, and Load

Fig. 2.9 Double circuit


transmission line

d4

a
d1

d1

d3

d6

d5

d1
d4

c
Fig. 2.10 Transmission line
with bundled conductors

c
d2

a
d

d3

D
Fig. 2.11 Stranded conductor

Fig. 2.12 Transmission line


with composite conductors

12
11

13

1n

23
2n
22

31
3n
32

33

21

Consider the transmission line consisting of a composite conductor in each phase,


as shown in Fig. 2.12 [2]. Let each phase consist of n individual conductors of
radius r .
It is assumed that the three phases are transposed if not placed symmetrically;
the transmission line is divided into three sections of equal lengths and each phase
occupies each of the three positions 1, 2, and 3 for one third of the transmission line
length. Let phases a, b, c occupy positions 1, 2, 3, respectively, in the first section,
positions 2, 3, 1, respectively, in the second section, and positions 3, 1, 2, respectively,
in the third section. The position of each individual conductor is identified by two
numbers as in Fig. 2.12; the first number is that of the position of the phase and the
second number is that of the position of the individual conductor. Let the current in
the individual conductors of phases a, b, c be i a /n, i b /n, i c /n, respectively. This is
true if the individual conductors in each phase are transposed so that each individual

2.2 Transmission Line

55

conductor occupies each of the n positions for equal lengths along a section. i a , i b ,
and i c satisfy (2.23). Consider the closed path consisting of the individual conductor
of phase a at position 1k for length l/(3n), and the straight line (at P) parallel to the
conductors. Similar to (2.37), the flux linkage of this closed path is
a1k =

n
0 l
D1m
D2m
D3m
ln
+
i
ln
+
i
ln
i
a
b
c
6 n 2
D1k1m
D1k2m
D1k3m

(2.43)

m=1

where D pkqm ( p and q are 1, 2, or 3, and pk = qm) is the distance between the axes
of conductors at positions pk and qm, D pkpk = r  , and D pk is the distance between
point P and the axis of conductor at position pk. It is evident from (2.27) to (2.29)
that the flux linkage of a phase is the average of the flux linkages of the closed paths
formed by individual conductors in that phase. Therefore, the flux linkage of phase
a in section 1 is
a1

n
n
0 l
D1m
D2m
D3m
i a ln
=
+ i b ln
+ i c ln
6 n 2
D1k1m
D1k2m
D1k3m

(2.44)

k=1 m=1

Similarly, the flux linkage of phase a in sections 2 and 3, a2 and a3 , respectively,


are given by

n
n
0 l
D2m
D3m
D1m
i
ln
+
i
ln
+
i
ln
a
b
c
6 n 2
D2k2m
D2k3m
D2k1m
k=1 m=1

n
n
0 l
D3m
D1m
D2m
i a ln
=
+ i b ln
+ i c ln
6 n 2
D3k3m
D3k1m
D3k2m

a2 =
a3

(2.45)

(2.46)

k=1 m=1

The flux linkage of phase a is


a = a1 + a2 + a3

(2.47)

From (2.23) and (2.44) to (2.47), the equivalent self inductance of each phase is
L=

Dm
0 l
ln
2
Ds

(2.48)

where

Dm 

n 
n


1/(3n 2 )
D1k2m D2k3m D3k1m

k=1 m=1

Ds 

n 
n


k=1 m=1

(2.49)
1/(3n 2 )

D1k1m D2k2m D3k3m

(2.50)

56

2 Transformer, Transmission Line, and Load

Dm is known as mutual geometric mean distance (GMD) and Ds is known as self


GMD.
For the double circuit transmission line shown in Fig. 2.9,
1/6

Dm = 2d12 d22 d3 d4
1/6

Ds = e3/4 r 3 d52 d6

(2.51)
(2.52)

where r is the radius of the individual conductors. For hexagonal spacing (d4 = d1
and d6 = d5 ), transposition is not necessary.
For the transmission line with bundled conductors shown in Fig. 2.10, where each
bundle (composite conductor) consists of two individual conductors,


2 
1/12
6
2
2
2
2
4D d
Dm = 4D D d

(2.53)


1/2
Ds = e1/4 r d

(2.54)

where r is the radius of the individual conductors.


For the stranded conductor shown in Fig. 2.11,
Ds = 2 (364.5)1/49 e1/28r

(2.55)

where r is the radius of the strands.


It is to be noted that the transposition of the individual conductors in the composite
conductor is not necessary for the three phases to be symmetrical; but the assumption
of transposition helps in easily obtaining the expression for inductance.

2.2.2 Capacitance
2.2.2.1 Transmission Line with Three Conductors
Consider the transmission line with three conductors shown in Fig. 2.8. Let the charge
per unit length on the conductors of phases a, b, and c be qa , qb , and qc , respectively,
such that
qa + q b + q c = 0

(2.56)

The radius of the conductors is assumed to be very small compared to the distance
between any two conductors. Therefore, the potential of conductor a with respect to
the point P is

2.2 Transmission Line

57

va P =

1
2 0

D1
D2
D3
qa ln
+ qb ln
+ qc ln
r
D12
D31

(2.57)

0 is the permittivity of free space; permittivity of air is almost equal to that of free
space. The potential of the conductor is obtained by allowing P to recede to infinity.
As P recedes to infinity, using (2.56), the potential of conductor a for symmetrical
spacing of conductors (D12 = D23 = D31 = D) is
1
va =
2 0

1
1
1
qa ln + qb ln + qc ln
r
D
D

(2.58)

From (2.56) and (2.58), the capacitance in each phase is


C=

2 0 l
ln(D/r )

(2.59)

If the conductors are not spaced symmetrically, it is assumed that transposition


is done in order to have a symmetrical system. Let the conductors a, b, c occupy
positions 1, 2, 3, respectively, in the first section, positions 2, 3, 1, respectively, in the
second section, and positions 3, 1, 2, respectively, in the third section. The charge
per unit length is not same in all three sections for any phase whereas the potential
is same. It is assumed that the charge per unit length is same in all the three sections
[3]; let the charge per unit length on the conductors of phases a, b, and c be qa ,
qb , and qc , respectively, which satisfy (2.56). With this assumption, the potential of
conductor a in sections 1, 2, and 3, va1 , va2 , and va3 , respectively, are given by
va1
va2
va3

1
1
1
1
qa ln + qb ln
=
+ qc ln
2 0
r
D12
D31

1
1
1
1
qa ln + qb ln
=
+ qc ln
2 0
r
D23
D12

1
1
1
1
qa ln + qb ln
=
+ qc ln
2 0
r
D31
D23

(2.60)
(2.61)
(2.62)

The potential of conductor a is assumed to be given by the following equation [3].


va =

1
(va1 + va2 + va3 )
3

(2.63)

From (2.56) and (2.60) to (2.63), the capacitance in each phase is


C=

2 0 l


ln (D12 D23 D31 )1/3 /r

(2.64)

58

2 Transformer, Transmission Line, and Load

2.2.2.2 Composite Conductors


Consider the transmission line with composite conductors shown in Fig. 2.12. It is
assumed that the three phases are transposed if not placed symmetrically. Let the
composite conductors of phases a, b, c occupy positions 1, 2, 3, respectively, in
the first section, positions 2, 3, 1, respectively, in the second section, and positions
3, 1, 2, respectively, in the third section. Let each individual conductor of a composite
conductor occupy each of the n positions for equal lengths along a section. Let the
charge per unit length on the individual conductors of phases a, b, and c be qa /n,
qb /n, and qc /n, respectively, which satisfy (2.56). It is assumed that the charge
per unit length on all individual conductors is same along the entire length of the
transmission line. The potential of phase a individual conductor at position 1k is
va1k =

n
1
1
1
1
qa ln 
+ qb ln 
+ qc ln 
2 0 n
D1k1m
D1k2m
D1k3m

(2.65)

m=1

where D pkqm ( p and q are 1, 2, or 3, and pk = qm) is the distance between the axes of
conductors at positions pk and qm; D pkpk is the radius of the individual conductors.
The potential of phase a composite conductor in section 1, va1 , is assumed to be
equal to the average of the potentials of the individual conductors.
va1 =

n
n

1
1
1
1
q
(2.66)
ln
+
q
ln
+
q
ln
a
b
c



2 0 n 2
D1k1m
D1k2m
D1k3m
k=1 m=1

The potential of phase a composite conductor in sections 2 and 3, va2 and va3 ,
respectively, are given by

n
n

1
1
1
1
q
(2.67)
ln
+
q
ln
+
q
ln
a
b
c



2 0 n 2
D2k2m
D2k3m
D2k1m
k=1 m=1

n
n

1
1
1
1
q
(2.68)
=
ln
+
q
ln
+
q
ln
a
b
c



2 0 n 2
D3k3m
D3k1m
D3k2m

va2 =
va3

k=1 m=1

The potential of phase a composite conductor is assumed to be given by


va =

1
(va1 + va2 + va3 )
3

(2.69)

From (2.56) and (2.66) to (2.69), the capacitance in each phase is


C=

2 0 l

ln Dm /Ds

(2.70)

2.2 Transmission Line

59

Fig. 2.13 Conductors and


their images

2
1

D12

D23

D 31
h31
h2
h1

h12

h 31

h 23

h 23

h12

h3
Earth

D31
1

D 23

D12

where

Dm 

n 
n


1/(3n 2 )



D1k2m
D2k3m
D3k1m

k=1 m=1

Ds

n 
n


(2.71)
1/(3n 2 )




D1k1m
D2k2m
D3k3m

(2.72)

k=1 m=1

2.2.2.3 Effect of Earth


The earth affects the distribution of the electric field due to charge on a conductor.
The earth is at zero potential. The effect of earth is same as that of the image of the
conductor [3]. The image of the conductor with charge qa per unit length is a conductor with charge qa per unit length located at the same distance from the earths
surface below it as shown in Fig. 2.13. The images of conductors at positions 1, 2,
and 3 are at positions 1 , 2 , and 3 , respectively.
Transposition of conductors is assumed. Let the conductors a, b, c occupy positions 1, 2, 3, respectively, in the first section, positions 2, 3, 1, respectively, in the
second section, and positions 3, 1, 2, respectively, in the third section. Let the charge
per unit length on the conductors a, b, and c be qa , qb , and qc , respectively, which

60

2 Transformer, Transmission Line, and Load

satisfy (2.56). It is assumed that the charge per unit length is same in all the three
sections. The potential of conductor a in sections 1, 2, and 3, va1 , va2 , and va3 ,
respectively, are given by
va1 =

1
1
1
1
1
1
1
qa ln ln
+ qb ln
+ qc ln
ln
ln
2 0
r
h1
D12
h 12
D31
h 31

(2.73)

1
1
1
1
1
1
1
qa ln ln
+ qb ln
+ qc ln
va2 =
ln
ln
2 0
r
h2
D23
h 23
D12
h 12

(2.74)

1
1
1
1
1
1
1
qa ln ln
+ qb ln
+ qc ln
va3 =
ln
ln
2 0
r
h3
D31
h 31
D23
h 23

(2.75)
where r is the radius of the conductors. The potential of conductor a is assumed to
be given by
va =

1
(va1 + va2 + va3 )
3

(2.76)

From (2.56) and (2.73) to (2.76), the capacitance in each phase is


C=

2 0 l
(D12 D23 D31 )
ln
r

1/3

h 12 h 23 h 31
ln
h1h2h3

1/3

(2.77)

Since h 12 h 23 h 31 > h 1 h 2 h 3 , the effect of earth is to increase the capacitance.

2.2.3 Transmission Line Model


Let va , vb , and vc be the voltages with respect to the neutral, i a , i b , and i c be the
currents, at the point which is at distance x from the receiving end, as shown in
Fig. 2.14. The currents satisfy (2.23). Then, the voltages and currents are related by
the following equations.
R
L i a
va
= ia +
x
l
l t
vb
R
L i b
= ib +
x
l
l t
vc
R
L i c
= ic +
x
l
l t

(2.78)
(2.79)
(2.80)

2.2 Transmission Line

61

ia

Sending
end

ib

Receiving
end

ic
x
Fig. 2.14 Transmission line

G
i a
C va
= va +
x
l
l t
i b
G
C vb
= vb +
x
l
l t
i c
G
C vc
= vc +
x
l
l t

(2.81)
(2.82)
(2.83)

The equations for the three phases are decoupled.


Dividing (2.78)(2.80) by VB , and (2.81)(2.83) by I B gives the equations in per
unit quantities.
v a
R
X i a
= ia +
x
l
l B t
v b
R
X i b
= ib +
x
l
l B t
v c
R
X i c
= ic +
x
l
l B t
i a
G
B v a
= va +
x
l
l B t
i b
G
B v b
= vb +
x
l
l B t
i c
G
B v c
= vc +
x
l
l B t

(2.84)
(2.85)
(2.86)
(2.87)
(2.88)
(2.89)

where X  B L, B  B C, base admittance Y B  1/Z B , and base capacitance


C B  Y B / B . Two special cases: lossless transmission line and sinusoidal operation,
are considered.

62

2 Transformer, Transmission Line, and Load

2.2.3.1 Lossless Transmission Line


If R = G = 0, then (2.78)(2.83) can be written as follows.
va
L i a
=
x
l t
i a
C va
=
x
l t
L i b
vb
=
x
l t
i b
C vb
=
x
l t
vc
L i c
=
x
l t
C vc
i c
=
x
l t

(2.90)
(2.91)
(2.92)
(2.93)
(2.94)
(2.95)

If the subscripts a, b, and c are not shown, the solution for any phase is
i(x, t) = f 1 (x v p t) f 2 (x + v p t)

(2.96)

v(x, t) = Z c f 1 (x v p t) Z c f 2 (x + v p t)

(2.97)

where v p  l/ LC and Z c  L/C [4]. v p is called phase velocity and Z c is


called characteristic impedance. f 1 and f 2 are functions of x and t. Let subscripts
S and R denote sending end quantities and receiving end quantities, respectively.
If only terminal response is of interest, the following method known as Bergerons
method is used. From (2.96) and (2.97),
i R (t) = i(0, t) = f 1 (v p t) f 2 (v p t)
v R (t) = v(0, t) = Z c f 1 (v p t) Z c f 2 (v p t)

l
l
iS t
= i l, t
= f 1 2l v p t f 2 v p t
vp
vp


l
l
= v l, t
= Z c f 1 2l v p t Z c f 2 v p t
vS t
vp
vp

(2.98)
(2.99)
(2.100)
(2.101)

Elimination of f 1 v p t , f 1 2l v p t , and f 2 (v p t) from (2.98) to (2.101) gives

l
1
l
1
+

vS t
v R (t)
i R (t) = i S t
vp
Zc
vp
Zc

(2.102)

2.2 Transmission Line

63

This equation relates the receiving end current and voltage. Similarly, one can obtain
the following equation which relates the sending end current and voltage.

l
1
l
1

+
vR t
v S (t)
i S (t) = i R t
vp
Zc
vp
Zc

(2.103)

Dividing (2.102) and (2.103) by I B gives the equations in per unit quantities.

l
+
i R (t) = i S t
vp

i S (t) = i R t
vp

l
1

vS t
v R (t)
vp
Zc
Zc

1
l
1
+
vR t
v S (t)
v
Zc
Zc
p
1

(2.104)
(2.105)

2.2.3.2 Sinusoidal Operation


Let the voltages and currents be sinusoidal with angular frequency o . Then voltages
and currents can be represented by phasors. Let V and I be the notations for phasor
representation of v and i, respectively. If subscripts a, b, and c are not shown, (2.78)
(2.83) can be written in the following form for each phase.
R + jo L
dV
=
I
dx
l
dI
G + jo C
=
V
dx
l

(2.106)
(2.107)

If V R and I R are the receiving end voltage and current, respectively, the solution of
(2.106) and (2.107) is
V = cosh( x)V R + Z c sinh( x)I R
1
sinh( x)V R + cosh( x)I R
I=
Zc

(2.108)
(2.109)

where Z c is called characteristic impedance and is called propagation constant.



Zc 

R + jo L
, 
G + jo C

(R + jo L)(G + jo C)
l

(2.110)

If V S and I S are the sending end voltage and current, respectively, then from (2.108)
and (2.109),
V S = cosh( l)V R + Z c sinh( l)I R
1
sinh( l)V R + cosh( l)I R
IS =
Zc

(2.111)
(2.112)

64

2 Transformer, Transmission Line, and Load

Fig. 2.15 Equivalent circuit


of transmission line

IS

IR

+
VS

Y /2

Y /2

Fig. 2.16 Nominal circuit


of transmission line

IS

VR

R+j o L

IR
+

+
(G+j oC) /2
VS

(G+j oC) /2

VR

Each phase of the transmission line can be represented by the equivalent circuit
shown in Fig. 2.15, where Z is impedance and Y is admittance. For this circuit, the
following equations can be written.

YZ
VR + Z IR
VS = 1 +
2

YZ
YZ
VR + 1 +
IR
IS = Y 1 +
4
2

(2.113)
(2.114)

Equating the coefficients of V R and I R in (2.111) and (2.113) gives


Z = (R + jo L)

sinh( l)
tanh( l/2)
, Y = (G + jo C)
l
l/2

(2.115)

It is to be noted that
lim Z = R + jo L , lim Y = G + jo C

l0

l0

(2.116)

If Z and Y in Fig. 2.15 are replaced by the values of their respective limits as l 0,
the circuit shown in Fig. 2.16 is obtained. This circuit is called nominal circuit.
For transmission lines of length less than 240 km, the nominal circuit shown in
Fig. 2.16 is a good approximation [3].

2.3 Krons Transformation

65

2.3 Krons Transformation


Krons transformation does a transformation of the three-phase voltages and currents
as follows [5].



iD
va
ia
vD
v Q  TK vb , i Q  TK i b
v0
vc
i0
ic

(2.117)

where

2 cos (o t) 2 cos (o t 2/3) 2 cos (o t + 2/3)


1
TK 
2 sin (o t) 2 sin (o t 2/3) 2 sin (o t + 2/3) (2.118)
3
1
1
1
where o is the operating frequency. It can be verified that TK1 = TKT .
If v0 = i 0 = 0, then (2.117) can be written as

vD
vQ



va
ia
i
D
= TK vb ,
= TK i b
iQ
vc
ic

(2.119)

where
1
TK 
3

2 cos (o t) 2 cos (o t 2/3) 2 cos (o t + 2/3)


(2.120)
2 sin (o t) 2 sin (o t 2/3) 2 sin (o t + 2/3)

In certain studies, high-frequency transients in the transformer and the transmission line are neglected. Then, Krons transformation results in simplification of
equations. Krons transformation also enables generalization of the definitions of
certain electrical quantities.

2.3.1 Definitions
There are quantities such as voltage magnitude, phase angle, frequency, reactive
power etc. which are well defined in steady state when voltages and currents are
sinusoidally varying and balanced. The definition of these quantities will be generalized so that they can be used even in the presence of harmonics and during a transient
when the voltage and current are not sinusoidal; however, these definitions are made
with the assumption that v0 = i 0 = 0.
Consider the shunt-connected equipment shown in Fig. 2.17. Let va , vb , and vc
be the voltages of terminals a, b, and c, respectively, with respect to the neutral.

66

2 Transformer, Transmission Line, and Load

Fig. 2.17 Shunt-connected


equipment

ia
a
ib
b
ic
c

The magnitude V and phase angle of the voltage of the three-phase bus, at
which the equipment in Fig. 2.17 is connected, are defined as
V 

v 2D + v 2Q
vD
 tan1
vQ

(2.121)
(2.122)

In other words, V = v Q + jv D . v Q and v D can be expressed in terms of V and


as follows.
v Q = V cos , v D = V sin

(2.123)

If va , vb , and vc are obtained from these expressions for v Q and v D using (2.119),
then

2
V sin(o t + )
va =
(2.124)
3


2
2
V sin o t +
(2.125)
vb =
3
3


2
2
vc =
V sin o t + +
(2.126)
3
3
Therefore, if va , vb , and vc are sinusoidal with angular frequency o and balanced,
V is the rms value of the line-to-line voltage and is the phase angle of va .
The frequency at the three-phase bus f is defined as
f  fo +
where f o  o /(2 ).

1 d
2 dt

(2.127)

2.3 Krons Transformation

67

Similar to voltage magnitude and phase angle definitions, the magnitude I and
phase angle of the current drawn by the equipment in Fig. 2.17 are defined as
I 


2 + i2
iD
Q

 tan1

iD
iQ

(2.128)
(2.129)

If i a , i b , and i c are sinusoidal with angular frequency o and balanced, then I is


times the rms value of i a , i b , or i c , and is the phase angle of i a .
The power drawn by the equipment in Fig. 2.17 is

P = va i a + vb i b + vc i c

(2.130)

P = vD i D + vQ i Q

(2.131)

P = Re[V I ()]

(2.132)

From (2.119),

It can be seen that

P is also known as active power. The reactive power Q drawn by the equipment in
Fig. 2.17 is defined as
Q  Im[V I ()] = v D i Q v Q i D

(2.133)

The active current i A and the reactive current i R drawn by the equipment in
Fig. 2.17 are defined as
i A  I cos( ) = i Q cos + i D sin

(2.134)

i R  I sin( ) = i Q sin i D cos

(2.135)

It is to be noted that i A > 0 P > 0, and i R > 0 Q > 0. The reactive current
is said to be inductive if it is positive, and is said to be capacitive if it is negative.
Consider the series-connected equipment shown in Fig. 2.18. The magnitude V
and phase angle of the voltage across the equipment in Fig. 2.18 are given by
(2.119)(2.122) using va , vb , and vc of Fig. 2.18. Similarly, the magnitude I and
phase angle of the current through the equipment in Fig. 2.18 are given by (2.119),
(2.120), (2.128), and (2.129) using i a , i b , and i c of Fig. 2.18. The active voltage v A
and the reactive voltage v R across the equipment in Fig. 2.18 are defined as
v A  V cos( )
v R  V sin( )

(2.136)
(2.137)

68

2 Transformer, Transmission Line, and Load

Fig. 2.18 Series-connected


equipment

ia

va

vb

vc

ib

ic

If v A is positive, active power is supplied by the equipment, otherwise, active power


is drawn by the equipment. The reactive voltage is said to be capacitive if it is positive
and inductive if it is negative.

2.3.2 Application to Transformer


Equations (2.18)(2.20) of the wye-wye-connected transformer can be written as

v 2a
v 1a
ia

di a /dt

1
v 2b = v 1b R 1 + R 2 i b
X 1 + X 2 di b /dt (2.138)
B
v 2c
v 1c
di c /dt
ic

By Krons transformation,
TKT

v 2D
v 2Q

= TKT

v 1D
v 1Q

R 1 + R 2 TKT

iD
iQ

d
1
X1 + X2
B
dt

TKT

iD
iQ

(2.139)


ia
v 1a
v 2a
v
v
i
where 1D  TK v 1b , 2D  TK v 2b and D  TK i b .
v 1Q
v 2Q
iQ
v 1c
v 2c
ic
Pre-multiplying (2.139) by TK gives

v 2D = v 1D (R 1 + R 2 )i D

di D
o
1
(2.140)
X1 + X2 i Q
X1 + X2
B
B
dt

2.3 Krons Transformation

69

Fig. 2.19 Equivalent circuit


of phase a of transmission line

iSa

ia

Re

iRa

Le

+
vSa

Ge
2

Ce
2

v 2Q = v 1Q (R 1 + R 2 )i Q +

Ge
2

Ce
2

vRa

di Q
o
1
(2.141)
X1 + X2 i D
X1 + X2
B
B
dt

If the high-frequency transients are to be neglected, then the last term on the righthand side of (2.140) and (2.141) are set to zero. For balanced sinusoidal operation at
angular frequency o , all transformed variables are constant and hence the last term
on the right-hand side of (2.140) and (2.141) is equal to zero. The factor o / B in
one of the terms of (2.140) and (2.141) is usually approximated to 1. Therefore,

v 2D = v 1D (R 1 + R 2 )i D X 1 + X 2 i Q

v 2Q = v 1Q (R 1 + R 2 )i Q + X 1 + X 2 i D

(2.142)
(2.143)

2.3.3 Application to Transmission Line


For sinusoidal operation, the equivalent circuit of the transmission line shown in
Fig. 2.15 is applicable. Let Z = Re + jo L e and Y = G e + jo Ce . The circuit of
Fig. 2.15 can be redrawn as shown in Fig. 2.19 for phase a.
From the circuit diagram in Fig. 2.19,
di a
dt
Ge
Ce dv Sa
i Sa i a =
v Sa +
2
2 dt
Ge
Ce dv Ra
i a i Ra =
v Ra +
2
2 dt

v Sa v Ra = Re i a + L e

(2.144)
(2.145)
(2.146)

Similarly, for phases b and c,


di b
dt
di c
= Re i c + L e
dt

v Sb v Rb = Re i b + L e

(2.147)

v Sc v Rc

(2.148)

70

2 Transformer, Transmission Line, and Load

Ge
Ce dv Sb
v Sb +
2
2 dt
Ge
Ce dv Sc
i Sc i c =
v Sc +
2
2 dt
Ge
Ce dv Rb
i b i Rb =
v Rb +
2
2 dt
Ge
Ce dv Rc
v Rc +
i c i Rc =
2
2 dt
i Sb i b =

(2.149)
(2.150)
(2.151)
(2.152)

By Krons transformation, (2.144)(2.152) can be written as


di D
(2.153)
dt
di Q
v S Q v R Q = Re i Q o L e i D + L e
(2.154)
dt
Ce
Ge
Ce dv S D
iSD iD =
v S D + o v S Q +
(2.155)
2
2
2 dt
Ce
Ge
Ce dv S Q
v S Q o v S D +
(2.156)
iSQ iQ =
2
2
2 dt
Ce
Ge
Ce dv R D
iD iRD =
v R D + o v R Q +
(2.157)
2
2
2 dt
Ce
Ge
Ce dv R Q
iQ iRQ =
v R Q o v R D +
(2.158)
2
2
2 dt

v Sa
v Ra
i Sa
vS D
v
i
RD
where
 TK v Sb ,
 TK v Rb , S D  TK i Sb ,
vS Q
vR Q
iSQ
v Sc
v Rc
i Sc


i Ra
ia
iRD
i
 TK i Rb , and D  TK i b . For balanced sinusoidal operation
iRQ
iQ
i Rc
ic
at frequency o , all transformed variables are constant. Therefore, (2.153)(2.158)
can be written as
v S D v R D = Re i D + o L e i Q + L e

v S D v R D = Re i D + o L e i Q

(2.159)

v S Q v R Q = Re i Q o L e i D
Ce
Ge
v S D + o v S Q
iSD iD =
2
2
Ce
Ge
v S Q o v S D
iSQ iQ =
2
2
Ce
Ge
v R D + o v R Q
iD iRD =
2
2
Ce
Ge
v R Q o v R D
iQ iRQ =
2
2

(2.160)
(2.161)
(2.162)
(2.163)
(2.164)

2.3 Krons Transformation

71

Dividing (2.159) and (2.160) by VB , and (2.161)(2.164) by I B gives the following


equations in per unit quantities.
o
X ei Q
B
o
Rei Q
X ei D
B
Ge
o B e
vSD +
vSQ
2
B 2
Ge
o B e
vSQ
vSD
2
B 2
Ge
o B e
vRD +
vRQ
2
B 2
Ge
o B e
vRQ
vRD
2
B 2

v S D v R D = Rei D +

(2.165)

vSQ vRQ =

(2.166)

i SD i D =
i SQ i Q =
i D i RD =
i Q i RQ =

(2.167)
(2.168)
(2.169)
(2.170)

where X e  B L e and Be  B Ce . The factor o / B in one of the terms in all


equations is usually approximated to 1. As an approximation, (2.165)(2.170) are
used even during transients.

2.4 Load
In many system studies, the effects of the subtransmission and the distribution networks along with the connected load devices are represented by an aggregated load
at a transmission substation. The load model is given by the expressions for active
power P and reactive power Q drawn, in terms of voltage magnitude and/or frequency
[5, 6]. Two commonly used models are:


V a
1 + k p f ( f fo )
Vo
b


V
1 + kq f ( f f o )
Q = Qo
Vo
P = Po

(2.171)
(2.172)



V 2
V
+ p2
+ p3 1 + k p f ( f f o )
P = Po p1
Vo
Vo



V 2
V
Q = Q o q1
+ q2
+ q3 1 + k q f ( f f o )
Vo
Vo

(2.173)
(2.174)

72

2 Transformer, Transmission Line, and Load

Subscript o identifies the values of the respective variables at the operating point.
a, b, p1 , p2 , p3 , q1 , q2 , q3 , k p f , and kq f are constants; p1 + p2 + p3 = 1 and
q1 + q2 + q3 = 1. If frequency dependence is not to be considered, k p f and kq f are
set to zero.
Equations (2.171)(2.174) in per unit quantities are
P=


Po VBa a 
V 1 + k p f ( f fo )
S B Voa


Q o VBb b 
V 1 + kq f ( f f o )
b
S B Vo



VB2 2
VB
Po
P=
V + p3 1 + k p f ( f f o )
p1 2 V + p2
SB
Vo
Vo



VB2 2
VB
Qo
Q=
V + q3 1 + k q f ( f f o )
q1 2 V + q2
SB
Vo
Vo
Q=

(2.175)
(2.176)
(2.177)
(2.178)

References
1. C.A. Gross, Power System Analysis, 2nd edn. (Wiley, New York, 1986)
2. S. Krishna, Teaching calculation of inductance of power transmission lines. Int. J. Electr. Eng.
Educ. 48(4), 434443 (2011)
3. J.J. Grainger, W.D. Stevenson Jr, Power System Analysis (Tata McGraw-Hill, Noida, 1994)
4. P.W. Sauer, M.A. Pai, Power System Dynamics and Stability (Pearson Education, Upper Saddle
River, 1998)
5. K.R. Padiyar, Power System Dynamics: Stability and Control, 2nd edn. (BS Publications,
Hyderabad, 2002)
6. P. Kundur, Power System Stability and Control (Tata McGraw-Hill, Noida, 1994)

Chapter 3

DC and Flexible AC Transmission Systems

Under certain circumstances, it is advantageous to transmit power over a DC


transmission line instead of an AC transmission line. DC transmission requires conversion from AC to DC at one end of the transmission line, and conversion from
DC to AC at the other end of the transmission line. These conversions are done by a
circuit called converter which consists of power semiconductor devices.
Flexible AC transmission system (FACTS) is an AC transmission system incorporating equipment, made up of power semiconductor devices, in order to enhance
controllability and increase power transfer capability [1]. Such equipment is called
a FACTS controller.

3.1 Power Semiconductor Devices


The power semiconductor devices are used as switches. A mechanical switch is said
to be in the on state if it is conducting current and is said to be in the off state if it
is not conducting current. Similarly, power semiconductor devices can be in the on
state or the off state. The transition from off state to on state is called turn on, and
the transition from on state to off state is called turn off. It will be assumed that these
devices are ideal. In other words, the following assumptions are made:
1. The voltage across a device is zero when it is on.
2. The current through a device is zero when it is off.
3. The time taken to turn on and turn off is zero.
The power semiconductor devices can be classified into diode, thyristor, and
controllable switches [2].
The diode has two terminalsanode (A) and cathode (K)and its circuit symbol
is shown in Fig. 3.1a. The diode conducts current from anode to cathode if the
potential of anode is higher than that of cathode, and it does not conduct current if
the anode potential is less than cathode potential. The state of the diode depends on
the circuit conditions. Hence, the diode is said to be an uncontrolled device.
S Krishna, An Introduction to Modelling of Power System Components,
SpringerBriefs in Electrical and Computer Engineering,
DOI: 10.1007/978-81-322-1847-0_3, The Author(s) 2014

73

74
Fig. 3.1 Power
semiconductor devices:
a diode, b thyristor, and
c controllable switch

3 DC and Flexible AC Transmission Systems


A

(a)

(b)

(c)

The thyristor has three terminalsanode (A), cathode (K), and gate (G)and its
circuit symbol is shown in Fig. 3.1b. The thyristor starts conducting current from
anode to cathode if the anode potential is higher than the cathode potential, and a
pulse of positive current flows from gate to cathode. The thyristor stops conducting
if the anode to cathode current goes to zero, and the cathode potential is higher than
anode potential for a certain minimum duration. The gate is said to be the control
terminal. Since the thyristor can be turned on by a control signal but cannot be turned
off by a control signal, it is said to be a semi-controllable device.
The circuit symbol of a controllable switch is shown in Fig. 3.1c. The circuit
symbol has two terminals1 and 2and the control terminal is not shown. The
switch can be turned on by a control signal if the potential of terminal 1 is higher
than that of terminal 2, and the current flows from terminal 1 to terminal 2. The
switch can also be turned off by a control signal. Gate turn off (GTO) thyristor and
insulated gate bipolar transistor (IGBT) are examples of controllable switch.
A device is said to be forward-biased if the potential of anode (or terminal 1) is
higher than that of cathode (or terminal 2). On the other hand, the device is said to be
reverse-biased if the potential of cathode (or terminal 2) is higher than that of anode
(or terminal 1).
When used at the transmission level, many devices are connected in series and/or
parallel to obtain the required voltage and/or current rating [1, 3]. This combination
of devices is called a valve. It is assumed that a valve behaves as a single device; all
the devices in a valve turn on and turn off simultaneously. The circuit symbol of a
valve is the same as that of the device.

3.2 DC Transmission System


A DC transmission line requires two converters one at each end where one converter
operates as a rectifier and the other operates as an inverter. A converter is said to
operate as a rectifier if the average power flow is from the AC side to the DC side.
If the average power flow is from the DC side to the AC side, the converter is said
to operate as an inverter. The type of converter used for long-distance bulk power
transmission is line-commutated converter [4].

3.2 DC Transmission System

75

Fig. 3.2 Line-commutated


converter

1
ea

L
ia

eb

L
vd

ib
ec

Id

L
ic
4

3.2.1 Line-Commutated Converter


The circuit diagram of a six-pulse line-commutated converter is shown in Fig. 3.2.
The converter has six thyristor valves. The circuit on the AC side is represented by
a three-phase balanced voltage source with an inductance L in each phase. A large
smoothing reactor is used on the DC side [4, 5]. Therefore, the circuit on the DC
side is represented by a constant current source.

3.2.1.1 Two-Valve Conduction Mode


For the sake of starting with a simplified analysis, let L = 0. If the gate currents are
continuously applied, the valves behave as diodes. At any instant, one valve among
1, 3, and 5 conducts and one valve among 2, 4, and 6 conducts. Among the valves 1,
3, and 5, the valve whose anode is at the highest potential conducts and the other two
valves are reverse-biased. Among the valves 2, 4, and 6, the valve whose cathode is
at the lowest potential conducts and the other two valves are reverse-biased. Let ea ,
eb , and ec be given by

2
V sin(o t + 150 )
ea =
(3.1)
3

2
V sin(o t + 30 )
eb =
(3.2)
3

2
V sin(o t 90 )
ec =
(3.3)
3
The plots of ea , eb , and ec are shown in Fig. 3.3. Each valve conducts for 120 . The
valves turn on in the sequence 1, 2, 3, 4, 5, and 6. The duration between any two
consecutive instants of turn on is 60 . One cycle of the AC voltage can be divided
into six equal intervals where each interval corresponds to the conduction of a pair

76

3 DC and Flexible AC Transmission Systems

Fig. 3.3 Voltages on the AC


side

ea

ec

ea, eb, ec

6,1

1,2

2,3

3,4

4,5

5,6

6,1

ot

Table 3.1 DC-side voltage and voltage across a valve for two-valve conduction mode
Interval
+ 60

< o t <
+ 60 < o t < + 120
+ 120 < o t < + 180
+ 180 < o t < + 240
+ 240 < o t < + 300
+ 300 < o t < + 360

Valves that conduct

vd

Voltage across valve 1

2, 3
3, 4
4, 5
5, 6
6, 1
1, 2

eb ec
eb ea
ec ea
ec eb
ea eb
ea ec

ea
ea
ea
ea
0
0

eb
eb
ec
ec

of valves. The six intervals and the conducting valves during each interval are shown
in Fig. 3.3.
The instant of natural conduction of a valve is the instant at which the valve starts
conducting if the gate current is continuously applied. For example, the instant of
natural conduction of valve 3 is when o t is equal to zero. Instead of continuous
gate currents, gate current pulses are applied which are delayed by an angle with
respect to the instants of natural conduction; is called delay angle. For example,
valve 3 is turned on at o t = . Valve 1 stops conducting when valve 3 is turned on.
The transfer of current from one valve to another
is known as commutation. Voltage
across valve 1 when valve 3 is turned on is 2V sin(o t), and this voltage appears
across valve 1 for < o t < + 120 . Immediately after valve 1 stops conducting,
it takes some time for it to withstand positive voltage [2, 4, 6] and this time is given
in terms of an angle denoted by o . Therefore, can take a value between 0 and
180 o .
For each interval, the valves that conduct, DC voltage vd , and voltage across valve
1 are given in Table 3.1.
The average DC voltage is

3 2
V cos
Vd =

(3.4)

3.2 DC Transmission System

77

Fig. 3.4 AC-side current


Id

ib

120o

60o

120o

60o

Id

o t

The converter operates as a rectifier if 0 < 90 and as an inverter if 90 < <


180 o .
The harmonic components in the DC voltage are of order h = 6k, k = 1, 2, 3, . . .
The rms value of the hth-order harmonic component is
Vh =

1/2
6V
2
2
1
+
(h

1)
sin

(h 2 1)

(3.5)

The plot of an AC-side current is shown in Fig. 3.4.


The rms value of the fundamental component of the AC-side currents is

6
Id
I1 =

(3.6)

The harmonic components in the AC-side currents are of order h = 6k 1, k =


1, 2, 3, . . . The rms value of the hth-order harmonic component is
Ih =

I1
h

(3.7)

An angle known as angle of advance, denoted by , is defined [4] as


 180

(3.8)

If > 60 , gives the duration for which the voltage across a valve is negative after
it stops conducting; should be greater than o . An angle known as commutation
margin angle is defined as the duration for which the voltage across a valve is negative
after it stops conducting [4]. This angle is relevant for inverter operation for which
=

(3.9)

78

3 DC and Flexible AC Transmission Systems

Fig. 3.5 Effective circuit


in the first subinterval for
two- and three-valve
conduction mode

ea
+

i1
eb
+

vd

i3
ec

Id

3.2.1.2 Two- and Three-Valve Conduction Mode


Due to the inductance L on the AC side, the current in a valve can vary only at
a finite rate, and hence, commutation from one valve to another takes some time.
For example, when valve 3 is turned on, the current transfer from valve 1 to valve
3 requires a certain time period during which both valves conduct. This duration is
measured in terms of an angle known as overlap angle or commutation angle denoted
by u. For normal operation, 0 < u < 60 .
One cycle of the AC voltage can be divided into six equal intervals. Each interval
can be divided into two subintervals. Three valves conduct in the first subinterval, and
two valves conduct in the second subinterval. For steady-state analysis, it is sufficient
to consider only one interval. The interval < o t < + 60 is considered.
The effective circuit to be analyzed in the first subinterval ( < o t < + u) is
shown in Fig. 3.5.
From the circuit diagram in Fig. 3.5,
L

d(Id i 3 )
di 3
L
= eb ea
dt
dt

(3.10)

Since i 3 () = 0, the solution of this equation is


i3 =

V
2o L

[cos cos(o t)]

(3.11)

Substituting i 3 ( + u) = Id gives
V
Id =
[cos cos( + u)]
2o L

(3.12)

From the circuit diagram in Fig. 3.5, using the relation ea + eb + ec = 0,


3
vd = ec
2

(3.13)

3.2 DC Transmission System

79

Table 3.2 DC-side voltage and voltage across a valve for two- and three-valve conduction mode
Subinterval

Valves that conduct

vd

Voltage across valve 1

< o t < + u
+ u < o t < + 60
+ 60 < o t < + u + 60
+ u + 60 < o t < + 120
+ 120 < o t < + u + 120
+ u + 120 < o t < + 180
+ 180 < o t < + u + 180
+ u + 180 < o t < + 240
+ 240 < o t < + u + 240
+ u + 240 < o t < + 300
+ 300 < o t < + u + 300
+ u + 300 < o t < + 360

1, 2, 3
2, 3
2, 3, 4
3, 4
3, 4, 5
4, 5
4, 5, 6
5, 6
5, 6, 1
6, 1
6, 1, 2
1, 2

3ec /2
eb ec
3eb /2
eb ea
3ea /2
ec ea
3ec /2
ec eb
3eb /2
ea eb
3ea /2
ea ec

0
ea eb
3eb /2
ea eb
3ea /2
ea ec
3ec /2
ea ec
0
0
0
0

In the second subinterval ( + u < o t < + 60 ),


vd = eb ec

(3.14)

3
Vd = V [cos + cos( + u)]
2

(3.15)

The average value of vd is

The harmonic components in the DC voltage are of order h = 6k, k = 1, 2, 3, . . .


The rms value of the hth-order harmonic component is

u
u
u
u
cos2 (h1) 2 + cos2 (h+1) 2 2 cos (h+1) 2 cos (h1) 2 cos(2+u)

3 2
(h1)2
(h+1)2
h 2 1
V
Vh =

(3.16)

For each subinterval, the valves that conduct, DC voltage vd , and voltage across
valve 1 are given in Table 3.2.
The AC-side currents possess half-wave symmetry. The expression for one of
these currents for half cycle is given by

[cos cos(o t)] if o t + u

2
oL

Id if + u o t + 120
ib =
I V [cos + cos( t + 60 )] if + 120 t + u + 120

o
o
d

2o L

0 if + u + 120 o t + 180

(3.17)

80

3 DC and Flexible AC Transmission Systems

The rms value of the fundamental component of the AC-side currents is

6Id
I1 =
2

[cos + cos(

+ u)]2

2u + sin(2) sin(2 + 2u)


+
2 cos 2 cos( + u)

2
(3.18)

The harmonic components in the AC-side currents are of order h = 6k 1, k =


1, 2, 3, . . . The rms value of the hth-order harmonic component is

6Id
Ih =
h

sin2 (h+1) u2
sin2 (h1) u2
2 sin (h+1) u2 sin (h1) u2 cos(2+u)
+

(h+1)2
(h1)2
h 2 1

cos cos( + u)

(3.19)

An angle known as extinction angle and denoted by is defined [4] as


 u

(3.20)

For normal inverter operation, 120 < < 180 , and the commutation margin
angle is
=

(3.21)

Equations (3.12) and (3.15) can be written in terms of instead of as follows:

3
Vd = V cos( + u) cos
2

V
cos( + u) + cos
Id =
2o L

(3.22)
(3.23)

3.2.1.3 Three- and Four-Valve Conduction Mode


Under certain abnormal conditions, u > 60 . If u 120 , vd = 0. For 60 < u <
120 , one cycle of the AC voltage can be divided into six equal intervals and each
interval can be divided into two subintervals. Four valves conduct in the first subinterval, and three valves conduct in the second subinterval. For steady-state analysis,
it is sufficient to consider only one interval. The interval < o t < + 60 is
considered.
When valve 3 is turned on at o t = , valves 6, 1, and 2 are still conducting. The
effective circuit to be analyzed in the first subinterval ( < o t < + u 60 ) is
shown in Fig. 3.6.
From the circuit diagram in Fig. 3.6,

3.2 DC Transmission System

81

Fig. 3.6 Effective circuit


in the first subinterval for
three- and four-valve
conduction mode

ea

i1
3
eb

L
6

vd

Id

i6
ec

di 1
d(Id i 1 i 6 )
L
= eb ea
dt
dt
d(Id i 6 )
d(Id i 1 i 6 )
+L
= eb ec
L
dt
dt
L

(3.24)
(3.25)

Multiplying (3.25) by 2 and subtracting from (3.24) gives


3L

di 6
= 6V cos(o t)
dt

(3.26)

The solution of this equation is

i6 =

6V
[sin sin(o t)] + i 6 ()
3o L

(3.27)

Substituting i 6 ( + u 60 ) = 0 gives

i 6 () =

6V
[sin( + u 60 ) sin ]
3o L

(3.28)

Substituting this in (3.27) gives

i6 =

6V
[sin( + u 60 ) sin(o t)]
3o L

(3.29)

Multiplying (3.24) by 2 and subtracting from (3.25) give


3L

di 1
= 6V sin(o t + 150 )
dt

(3.30)

The solution of this equation is

6V
[cos( + 150 ) cos(o t + 150 )] + i 1 ()
i1 =
3o L

(3.31)

82

3 DC and Flexible AC Transmission Systems

Substituting i 1 () = Id gives

i1 =

6V
[cos( + 150 ) cos(o t + 150 )] + Id
3o L

(3.32)

The effective circuit in the second subinterval ( + u 60 < o t < + 60 )


is same as that shown in Fig. 3.5. From the circuit diagram in Fig. 3.5,
L

di 1
d(Id i 1 )
L
= eb ea
dt
dt

(3.33)

The solution of this equation is


i1 =

V
2o L

[cos(o t) cos( + u 60 )] + i 1 ( + u 60 )

(3.34)

From (3.32),

i 1 ( + u 60 ) =

6V
[cos( + 150 ) + sin( + u)] + Id
3o L

(3.35)

Substituting this in (3.34) gives


V
i1 =
[cos(o t) cos( + u 60 )]
2o L

6V
+
[cos( + 150 ) + sin( + u)] + Id
3o L

(3.36)

It is to be noted that
i 1 ( + 60 ) = i 6 ()

(3.37)

Substituting for i 1 ( + 60 ) from (3.36) and for i 6 () from (3.28) in (3.37) gives
V
Id =
[cos( 30 ) cos( + u + 30 )]
6o L

(3.38)

For each subinterval, the valves that conduct, DC voltage vd , and voltage across
valve 1 are given in Table 3.3.
The average value of vd is

3 6
V [cos( 30 ) + cos( + u + 30 )]
Vd =
2

(3.39)

3.2 DC Transmission System

83

Table 3.3 DC-side voltage and voltage across a valve for three- and four-valve conduction mode
Subinterval

Valves that conduct

vd

Voltage across valve 1

< o t < + u 60
+ u 60 < o t < + 60
+ 60 < o t < + u
+ u < o t < + 120
+ 120 < o t < + u + 60
+ u + 60 < o t < + 180
+ 180 < o t < + u + 120
+ u + 120 < o t < + 240
+ 240 < o t < + u + 180
+ u + 180 < o t < + 300
+ 300 < o t < + u + 240
+ u + 240 < o t < + 360

6, 1, 2, 3
1, 2, 3
1, 2, 3, 4
2, 3, 4
2, 3, 4, 5
3, 4, 5
3, 4, 5, 6
4, 5, 6
4, 5, 6, 1
5, 6, 1
5, 6, 1, 2
6, 1, 2

0
3ec /2
0
3eb /2
0
3ea /2
0
3ec /2
0
3eb /2
0
3ea /2

0
0
0
3eb /2
0
3ea /2
0
3ec /2
0
0
0
0

Equations (3.38) and (3.39) can be written in terms of instead of .

3 6
V cos( + u + 30 ) cos( 30 )
Vd =
2

V
Id =
cos( + u + 30 ) + cos( 30 )
6o L

(3.40)
(3.41)

3.2.2 12-Pulse Line-Commutated Converter


The circuit diagram of the 12-pulse converter is shown in Fig. 3.7 [4, 7]. The two
six-pulse line-commutated converters are connected in series on the DC side to obtain
a higher DC-side voltage. One transformer is wyewye-connected, and the other is
wyedelta-connected. The magnitude of the line-to-line voltages on the AC side of
the two six-pulse converters should be same. Therefore, the number of turns in the
transformer windings are as shown in Fig. 3.7.
If the transformers are assumed to be ideal, the three-phase voltage supplied by
the delta-connected winding to one of the six-pulse converters lags the three-phase
voltage supplied to the other six-pulse converter, by 30 . The instants of turn on of the
thyristor valves of the six-pulse converter supplied by the delta-connected winding
are delayed by 30 with respect to the instants of turn on of the thyristor valves of
the other six-pulse converter. If h is allowed to take a value 1 (for the fundamental
component) or 6k 1, k = 1, 2, 3, . . . (for the harmonic components), the current
phasors shown in Fig. 3.7 are given by
Iah = Ih 0, Ibh = Ih (120 h), Ich = Ih (120 h)

Iah
= Ih (30 h), Ibh
= Ih (150 h), Ich
= Ih (90 h)

(3.42)
(3.43)

84

3 DC and Flexible AC Transmission Systems

N1

N2

Iah

N1
IAh

N2

Ibh

N2

N1

Ich
IBh

ICh
N1

Iah

I
ah

Ich
3 N2

N1

Ibh

N1

Ibh

I
ch

Fig. 3.7 12-pulse line-commutated converter

, I , and I satisfy the following equations:


The currents Iah
bh
ch

Ich
= Iah
Iah

Ibh
Iah
= Ibh

Iah
+ Ibh
+ Ich
=0

(3.44)
(3.45)
(3.46)

Equation (3.46) is obtained by applying Kirchhoffs current law to the neutral of the
+ I +
wye-connected winding. Solving (3.44)(3.46) and using the relation Iah
bh

Ich = 0 gives
1

(I Ibh
)
3 ah
1

= (Ibh
Ich
)
3
1

= (Ich
Iah
)
3

Iah
=

(3.47)

Ibh

(3.48)

Ich

(3.49)

3.2 DC Transmission System

85

Fig. 3.8 Schematic diagram


of a typical converter station

The currents in the AC system are

N2

2
I1 0 if h = 1

N1

N2
N2
I Ah =
Iah + 3 Iah
= 2 N2 I 0 if h = 6k 1, k = 2, 4, 6, . . .
(3.50)

h
N1
N1

0 if h = 6k 1, k = 1, 3, 5, . . .

N2

2
I1 (120 ) if h = 1

N1

N2
N2
I Bh =
Ibh + 3 Ibh
= 2 N2 I (120 ) if h = 6k 1, k = 2, 4, 6, . . .(3.51)

h
N1
N1

N1

0 if h = 6k 1, k = 1, 3, 5, . . .

2 2 I1 120 if h = 1

N1

N2
N2
N
ICh =
Ich + 3 Ich = 2 2 I (120 ) if h = 6k 1, k = 2, 4, 6, . . .(3.52)

h
N1
N1

N1

0 if h = 6k 1, k = 1, 3, 5, . . .

Therefore, the harmonic components in the AC system currents are of order


h = 12k 1, k = 1, 2, 3, . . . The harmonic components in the DC-side voltage
are of order h = 12k, k = 1, 2, 3, . . .
Typically, a converter station consists of two 12-pulse converters connected as
shown by the schematic diagram in Fig. 3.8 [4, 7]. One terminal on the DC side is at
a positive potential with respect to earth, and the other is at a negative potential. The
operation is such that the DC-side currents in the two 12-pulse converters are equal.

86

3 DC and Flexible AC Transmission Systems

3.3 FACTS
FACTS controllers can be classified into the following types [3]:
1.
2.
3.
4.

Shunt controller.
Series controller.
Combined shuntseries controller.
Combined seriesseries controller.

Depending on the power semiconductor device used, the FACTS controllers can
be classified into variable impedance-type controller and voltage source converter
(VSC)-based controller. The device used in a variable impedance-type controller is
thyristor, whereas a VSC-based controller uses a controllable switch.
The prominent FACTS controllers are as follows:
1. Static var compensator (SVC): variable impedance-type shunt controller.
2. Thyristor-controlled series capacitor (TCSC): variable impedance-type series
controller.
3. Static synchronous compensator (STATCOM): VSC-based shunt controller.
4. Static synchronous series compensator (SSSC): VSC-based series controller.
5. Unified power flow controller (UPFC): VSC-based combined shuntseries
controller.
6. Interline power flow controller (IPFC): VSC-based combined seriesseries
controller.
The primary function of shunt FACTS controller is regulation of voltage and that
of series FACTS controller is regulation of power flow.

3.3.1 SVC
The SVC consists of a three-phase thyristor-controlled reactor (TCR) in parallel with
three capacitors connected in wye or delta.

3.3.1.1 TCR
The TCR consists of a reactor and two thyristor valves connected in antiparallel. The
effective reactance of TCR is varied by varying the instant of gate current pulses
to the thyristor
valves. Figure 3.9 shows a voltage source connected across a TCR
where v = 2V cos(o t) and L is inductance.
If the gate current pulses are at the instants of voltage peaks, the effective reactance
is minimum and is equal to that of the reactor. In order to increase the reactance,
the gate current pulses to the thyristor valves are delayed. The gate current pulses
to thyristor valves 1 and 2 are at o t = 2 k + and o t = 2 k + + ,

3.3 FACTS

87

Fig. 3.9 TCR connected to a


voltage source

L
+
v

k = 0, 1, 2, . . ., respectively, where 0 90 . The circuit of Fig. 3.9 is governed


by the following equation:

2V

cos(o t) if < o t <

L
di
0 if < o t < +
=
2V
dt

cos(o t) if + < o t < 2

L
0 if 2 < o t < 2 +

(3.53)

The solution of this equation is

2V

[sin(o t) sin ] if o t

o L
0 if o t +
i=

2V

[sin(o t) + sin ] if + o t 2

0 if 2 o t 2 +

(3.54)

Figure 3.10 shows the plots of TCR voltage and current.


The rms value of the fundamental component of current is
I1 =

V
[ 2 sin(2)]
o L

(3.55)

The current contains odd harmonic components only. The rms value of the odd
harmonic component of current, of order h, is
Ih =

4V |sin cos(h) h cos sin(h)|


o Lh(h 2 1)

(3.56)

88

3 DC and Flexible AC Transmission Systems

Fig. 3.10 Plots of TCR


voltage and current

v and i

50

100

150
200
ot (degree)

250

300

350

Fig. 3.11 Six-pulse TCR

If the harmonic components are neglected, the susceptance of the TCR is


I1
2 sin(2)
=
V
o L

(3.57)

Connection of three TCRs in delta as shown in Fig. 3.11 results in the three-phase
six-pulse TCR; the triplen harmonic components are eliminated in the line currents.

3.3.1.2 12-Pulse TCR


Half the number of harmonic components in the line currents of the six-pulse TCR
can be eliminated using the 12-pulse TCR [3, 7]. Figure 3.12 shows the circuit

3.3 FACTS

89

I ah
N2

N2

I bh

N2

IAh

N1

I ch
I Bh
N1

I
ah

N1

Iah

ICh
I
ch
3 N2

I
bh

I
bh

I
ch

Fig. 3.12 12-pulse TCR

diagram of the 12-pulse TCR which consists of two six-pulse TCRs and a threephase three-winding transformer.
If transformers are assumed to be ideal, based on the explanation given in
Sect. 3.2.2,
Iah = Ih 0, Ibh = Ih (120 h), Ich = Ih (120 h)

Iah
= Ih (30 h), Ibh
= Ih (150 h), Ich
= Ih (90 h)

(3.58)
(3.59)

From (2.6),
N2
N2
Iah + 3 Iah
N1
N1
N2
N2
=
Ibh + 3 Ibh
N1
N1
N2
N2
=
Ich + 3 Ich
N1
N1

I Ah =

(3.60)

I Bh

(3.61)

ICh

(3.62)

90

3 DC and Flexible AC Transmission Systems

Application of Kirchhoffs current law to the neutral of the wye-connected windings


gives Iah + Ibh + Ich = 0 and I Ah + I Bh + ICh = 0; therefore, adding (3.60)(3.62)
gives

+ Ibh
+ Ich
=0
Iah

(3.63)

, I , and I satisfy the following equations:


The currents Iah
bh
ch

Ich
= Iah
Iah

Ibh
Iah
= Ibh

(3.64)
(3.65)

+ I + I = 0 give
Solving (3.63)(3.65) and using the relation Iah
bh
ch

(I Ibh
)
3 ah
1

= (Ibh
Ich
)
3
1

= (Ich
Iah
)
3

Iah
=

(3.66)

Ibh

(3.67)

Ich

(3.68)

The currents drawn by the 12-pulse TCR are

I Ah

I Bh

ICh

N2

2
I 0 if h = 1

N1 1
N2
=
2
Ih 0 if h = 6k 1, k = 2, 4, 6, . . .

N1
0 if h = 6k 1, k = 1, 3, 5, . . .

2 N2 I1 (120 ) if h = 1

N1
N2
=
2
Ih (120 ) if h = 6k 1, k = 2, 4, 6, . . .

0 if h = 6k 1, k = 1, 3, 5, . . .

N2

2
I 120 if h = 1

N1 1
N2
=
2
Ih (120 ) if h = 6k 1, k = 2, 4, 6, . . .

0 if h = 6k 1, k = 1, 3, 5, . . .

(3.69)

(3.70)

(3.71)

Therefore, the harmonic components in the currents drawn by a 12-pulse TCR are
of order h = 12k 1, k = 1, 2, 3, . . .

3.3 FACTS

91

Fig. 3.13 SVC and the


Thevenin equivalent circuit of
the network

ia

Lt

iTCRa

Lt

vtc

vta

vtb

vb
+

LTCR C

iTCRc

LTCR

+
Lt

LTCR
ib
ic

+
v
a

iTCRb

vc +

3.3.1.3 Controller for SVC


In steady state, if the harmonic components in the TCR currents are neglected, the
TCR can be represented by three wye-connected inductors of inductance L T C R . Let
the capacitors of capacitance C be wye-connected. SVC is connected at a bus in a
network. The network is represented by the Thevenin equivalent circuit shown in
Fig. 3.13. The resistance in the Thevenin equivalent circuit is neglected, and L t is
inductance.
The equations governing the circuit in Fig. 3.13 are
di a
dt
di b
Lt
dt
di c
Lt
dt
di T C Ra
LTC R
dt
di T C Rb
LTC R
dt
di T C Rc
LTC R
dt
dva
C
dt
dvb
C
dt
dvc
C
dt
Lt

= vta va

(3.72)

= vtb vb

(3.73)

= vtc vc

(3.74)

= va

(3.75)

= vb

(3.76)

= vc

(3.77)

= i a i T C Ra

(3.78)

= i b i T C Rb

(3.79)

= i c i T C Rc

(3.80)

92

3 DC and Flexible AC Transmission Systems

By Krons transformation,
i Q + ji D = j(BT C R BC )(v Q + jv D )
v Q + jv D = vt Q + jvt D jX t (i Q + ji D )



vta
va
ia
v
i
 TK vtb , D  TK vb , D  TK i b , BC  o C,
vQ
iQ
vtc
vc
ic
 1/(o L T C R ), and X t  o L t . Equations (3.81) and (3.82) can be written

v
where t D
vt Q
BT C R
as

(3.81)
(3.82)

I ( 90 ) = jB SV C V

(3.83)

V = Vt jI ( 90 )X t

(3.84)



2 + i 2 ,  tan1 (v /v ) = tan1 (v /v ),
where V  v2D + v2Q , I  i D
D Q
tD tQ
Q

Vt  vt2D + vt2Q , and B SV C  BT C R BC . Equations (3.83) and (3.84) can be
written as
i R = B SV C V

(3.85)

V = Vt i R X t

(3.86)

where

iR =

I if current is inductive
I if current is capacitive

(3.87)

i R is reactive current as defined by (2.135). For a given value of B SV C , (3.85) is


the equation of a straight line in the i R V plane. A few representative straight lines
corresponding to different values of B SV C are shown in Fig. 3.14. Let 0 BT C R
B L ; the minimum value of B SV C is BC , and the maximum value of B SV C is
B L BC . Equation (3.86) is that of a straight line shown in Fig. 3.15.
For a given steady-state operating condition, the values of V and i R are obtained
by solving (3.85) and (3.86). In order to regulate the voltage, a regulator is used to
vary the value of B SV C such that in steady state, the voltage magnitude V is equal
to the desired value Vr e f . The block diagram of the regulator is shown in Fig. 3.16
where the regulator is typically a proportionalintegral controller. The proportional
and integral gains are negative; the sign of the proportional and integral gains should
be such that the change in the regulator output results in reduction in the magnitude
of the error (regulator input). Br e f is the desired value of B SV C . In steady state, the
V versus i R characteristic is shown in Fig. 3.17. It consists of three straight lines:
One straight line corresponds to the inductive limit at which B SV C = B L BC ,
one straight line corresponds to the capacitive limit at which B SV C = BC , and the
straight line with zero slope corresponds to the control range.

3.3 FACTS

93

Fig. 3.14 Plot of voltage


magnitude versus reactive
current for different values of
SVC susceptance

BSVC =BC

BSVC =BLBC

iR
Fig. 3.15 Characteristic of
the network

Slope = Xt
V = Vt

iR
Fig. 3.16 SVC controller

BL B C
Vref

Regulator

Bref
BC

As the power system operating condition changes, the values of Vt and X t change,
and this may result in operation at the inductive or capacitive limits. To avoid frequent hitting of the limits, the steady-state characteristic is altered using the controller
shown in Fig. 3.18 [3]; X s is positive. Due to this controller, the steady-state characteristic is as shown in Fig. 3.19.

94

3 DC and Flexible AC Transmission Systems

Fig. 3.17 Steady-state


characteristic of SVC

BSVC =BLBC
V=Vref
BSVC = BC

iR

Fig. 3.18 Modified SVC


controller

BL BC
Vref

Regulator

B ref

BC

Xs

iR

+
V

Fig. 3.19 Steady-state


characteristic of SVC with
positive slope in the control
range

Slope = Xs

BSVC =BLBC

V = Vref
BSVC = BC

iR

3.3.2 TCSC
A TCSC is connected in series with a transmission line. The TCSC consists of a
single-phase TCR and a capacitor in parallel, in all the three phases.

3.3 FACTS

95

Fig. 3.20 TCSC connected in


series with a current source

L
iTCR
C
+

3.3.2.1 Analysis of TCSC


The effective reactance of TCSC is varied by varying the instant of gate current pulses
to the thyristors. Figure 3.20 shows one phase of a TCSC where L is inductance and
C is capacitance; the transmission line current is assumed to be independent and is
modelled by a current source i = Im cos(o t) [3, 8].
If the thyristors are turned on at the instants of peaks of current i, the current
through the TCR is zero. In order to vary the effective reactance of the TCSC, the
gate current pulses to the thyristors are advanced by an angle with respect to the
instants of peaks of current i; 0 /2. The equations governing the circuit,
when the TCR conducts, are
iT C R
Im cos o t
dv
=
+
dt
C
C
di T C R
v
=
dt
L

(3.88)
(3.89)

In steady state, i T C R and v possess quarter-wave symmetry; the instant at which i T C R


attains the peak value and the instant of zero crossing of v are the same as the instant
at which i attains the peak value. The steady-state solution of (3.88) and (3.89), in
the interval o t , is given by the following two equations:
v=

Im X C
2

sin(o t + ) sin(o t)
sin(o t + + ) + sin(o t)
+
+1
1

+v() cos(o t + )

Im cos(o t) cos(o t + + )
iT C R =
2
+1
+


C
cos(o t) cos(o t + )
v() sin(o t + )
+
1
L

(3.90)

(3.91)

where  1/(o LC) and X C  1/(o C). Substituting v(0) = 0 in (3.90) gives
v() =

Im X C
[sin cos tan()]
2 1

(3.92)

96

3 DC and Flexible AC Transmission Systems

Substituting this in (3.90) gives

cos sin(o t)
Im X C
sin(o t) +
v= 2
1
cos()

(3.93)

In the interval o t , the equation governing the circuit is


dv
Im cos(o t)
=
dt
C

(3.94)

v() = v(). Therefore, the solution of (3.94) is


v=

Im X C
[ cos tan() sin ] + Im X C [sin(o t) sin ]
2 1

(3.95)

The peak value of the fundamental component of v is

V1m

4
=

/4
v sin(o t)d(o t)

(3.96)

If the harmonic components are neglected, the reactance of TCSC is




V1m
2 2 X C
1 2 + 1
2 cos2 tan()
X T C SC =
= XC +
2
sin(2) +
Im
2 1
2 1
2 1

(3.97)
X T C SC is positive when it is capacitive and negative when it is inductive. Parallel
resonance occurs at certain values of , denoted by r es . X T C SC as
r es = (2k + 1)/(2), k = 0, 1, 2, . . . Since 0 /2, there is only one value
of r es if < 3. Typically, < 3 [3, 8], and hence, r es = /(2). Figure 3.21
shows the variation of X T C SC with .

3.3.2.2 Controller for TCSC


The primary purpose of TCSC is to increase the power flow in a transmission line.
Since the reactance of TCSC can be varied, it can be controlled in order to regulate
the power flow in the transmission line. Normally, X T C SC is capacitive. Operation
near resonance results in a large voltage across the TCSC. To limit the voltage across
TCSC, 0 max where max < r es . At = 0, X T C SC = X C , and at
= max , X T C SC = X max .
The most basic type of control is open-loop control where is decided based on the
desired TCSC reactance X r e f . The other type of control is closed-loop control. One
type of closed-loop control is the constant current (CC) control [3, 8]. In CC control,

3.3 FACTS

97

TCSC

Fig. 3.21 Plot of TCSC


reactance as a function of
angle

Fig. 3.22 CC controller for


TCSC

10

20

30

40
50
(degree)

60

70

80

90

X max
Iref

Regulator

Xref
XC

Fig. 3.23 Steady-state


characteristic of TCSC with
CC controller

TCSC

=X

I ref

X TCSC =X max

the magnitude of the current in the transmission line is regulated. The block diagram
of CC controller is shown in Fig. 3.22. I is the magnitude of the current through
TCSC as defined in Sect. 2.3.1, and Ir e f is the desired value of I . The regulator is
typically a proportionalintegral controller; the proportional and integral gains are

98

3 DC and Flexible AC Transmission Systems

P
S1

ea

eb

S3

D5

vd

ib
ec

D3 S 5

ia
n

D1

Rp

ic
S4

D4

S6

D6 S 2

D2
N

Fig. 3.24 Two-level VSC

positive. The steady-state characteristic with CC controller is shown in Fig. 3.23. V


is the magnitude of voltage across TCSC as defined in Sect. 2.3.1. In steady state, if
the harmonic components are neglected, V and I are related by V = X T C SC I .

3.3.3 VSC
A VSC employs controllable switches.

3.3.3.1 Two-Level VSC


The circuit of a two-level VSC is shown in Fig. 3.24. The VSC consists of three
legs and a capacitor. The resistance R p in parallel with the capacitance C represents
the loss in the capacitor. Each leg has two bidirectional switches. A bidirectional
switch consists of a controllable switch and a diode connected in antiparallel. The
bidirectional switch consisting of S1 and D1 is called switch 1. Similarly, the other
five bidirectional switches are called switches 2 to 6. A bidirectional switch is said to
be on if the controllable switch or the diode conducts. At any instant, in each leg, one
bidirectional switch is on and one bidirectional switch is off. The DC-side voltage
vd is always positive. The possible states of one of the legs are shown in Table 3.4.
va N is the voltage of terminal a with respect to the DC-side negative terminal N . The
circuit on the AC side is represented by a three-phase balanced voltage source with
a resistance R and an inductance L in each phase. The currents i a , i b , and i c flow

3.3 FACTS

99

Table 3.4 Possible states of a leg of the two-level VSC


Controllable switch with control Controllable switch with control va N
signal to turn on
signal to turn off

Device that conducts


ia > 0 ia < 0

S1
S4

D1
S4

S4
S1

vd
0

S1
D4

through inductances and hence can vary only at a finite rate. If D1 is conducting and
a control signal is given to turn on S4 , then S4 starts conducting and the capacitor
voltage reverse-biases D1 ; if S1 is conducting and it is turned off, then D4 starts
conducting. Similarly, if D4 is conducting and a control signal is given to turn on
S1 , then S1 starts conducting and the capacitor voltage reverse-biases D4 ; if S4 is
conducting and it is turned off, then D1 starts conducting.
The adjective two-level means that for a given DC-side voltage, the number
of possible values of the voltage of an AC-side terminal with respect to a DC-side
terminal is two. For example, va N is equal to either vd or 0.
Let van , vbn , and vcn be the voltages of the terminals a, b, and c, respectively,
with respect to the neutral n of the AC side. Let vbN and vcN be the voltages of the
terminals b and c, respectively, with respect to the terminal N . Let v N n be the voltage
of N with respect to n.
van = va N + v N n
vbn = vbN + v N n

(3.98)
(3.99)

vcn = vcN + v N n

(3.100)

The switches are turned on and off such that the fundamental components of
AC-side currents i a , i b , and i c are equal in magnitude and displaced by 120 . The
triplen harmonic components of these currents are zero. For any other harmonic
order, the harmonic components of the three currents are equal in magnitude and
displaced by 120 . Hence, the drops across the series combination of R and L in the
three phases add to zero. Therefore, since the AC-side source voltages, ea , eb , and
ec are balanced,
van + vbn + vcn = 0

(3.101)

Adding (3.98)(3.100) and using (3.101) give


1
v N n = (va N + vbN + vcN )
3

(3.102)

Hence, from (3.98) to (3.100) and (3.102),


2/3 1/3 1/3


va N
van
vbn = 1/3 2/3 1/3 vbN
1/3 1/3 2/3
vcn
vcN

(3.103)

100

3 DC and Flexible AC Transmission Systems

The voltages va N , vbN , and vcN can be written in terms of switching functions u a ,
u b , and u c .


va N
ua
vbN = u b vd
(3.104)
vcN
uc
where u a = 1 if switch 1 is on and u a = 0 if switch 4 in on, u b = 1 if switch 3 is on
and u b = 0 if switch 6 in on, and u c = 1 if switch 5 is on and u c = 0 if switch 2 in
on. From (3.103) and (3.104),

sa
van
vbn = sb vd
vcn
sc

(3.105)



sa
2/3 1/3 1/3
ua
sb  1/3 2/3 1/3 u b
1/3 1/3 2/3
sc
uc

(3.106)

where

sa , sb , and sc are also called switching functions.


The switching of the GTO thyristor is slow. If GTO thyristors are used as
controllable switches, in order to minimize the switching losses, each GTO thyristor
is turned on and off only once in a cycle [3]. The plots of u a , u b , and u c are shown in
Fig. 3.25; the plots of sa , sb , and sc are shown in Fig. 3.26. With this switching, for
a given constant vd , the magnitude of the fundamental component of van , vbn , and
vcn is fixed. If vd is constant, the harmonic components present in van , vbn , and vcn
are of order h = 6k 1, k = 1, 2, 3, . . .
If IGBT is used as the controllable switch, the switching frequency can be
increased. Then, the magnitude of the fundamental component of the voltages van ,
vbn , and vcn can be varied for a given constant vd using the switching function u a
shown in Fig. 3.27 instead of u a . This switching function has a notch in both positive and negative half cycles and possesses quarter-wave symmetry. In addition to
controlling the magnitude of the fundamental component, some harmonic components can be eliminated by introducing more notches in the switching function. For
example, using the switching function u a shown in Fig. 3.28 which has two notches
in both positive and negative half cycles, a harmonic component can be eliminated
in addition to controlling the magnitude of the fundamental component; there are
two degrees of freedom, namely and . u a should possess quarter-wave symmetry.
This type of switching is called selective harmonic elimination.

3.3 FACTS

101

ua

+180

o t (degree)

+360

ub

0
+120

+480

+300
o t (degree)

uc

0
+60

+240
o t (degree)

+420

Fig. 3.25 Switching functions u a , u b , and u c

3.3.3.2 Multi-Level VSC


Diode-clamped converter is a type of multi-level VSC. Figure 3.29 shows the circuit
diagram of a three-level diode-clamped converter. In addition to the bidirectional
switches, there are six diodes Dc1 to Dc6 which are called clamping diodes. The
resistances R p in parallel with the capacitances C represent the loss in the capacitors.

102

3 DC and Flexible AC Transmission Systems

2/3

sa

1/3
0

1/3
2/3

+180

+360
ot (degree)

2/3

1/3
0

1/3
2/3
+120

+300

+480

ot (degree)
2/3

1/3
0

1/3
2/3
+60

+240

+420

ot (degree)

Fig. 3.26 Switching functions sa , sb , and sc

It is assumed that vd1 vd2 vd /2. vd is always positive. For each leg, there are
three possible states. The possible states of one of the legs are shown in Table 3.5.
vam is the voltage of the terminal a with respect to the midpoint m of the DC side.
Let van , vbn , and vcn be the voltages of the terminals a, b, and c, respectively,
with respect to the neutral n of the AC side. Let vbm and vcm be the voltages of the
terminals b and c, respectively, with respect to the midpoint m of the DC side. vam ,
vbm , and vcm can be written in terms of switching functions u 3a , u 3b , and u 3c .

3.3 FACTS

103

Fig. 3.27 Switching


function u a

ua

2+180

Fig. 3.28 Switching


function u a

+180

2+360

+180
o t (degree)

+360

+180
o t (degree)

+360

u"

vam = u 3a vd

(3.107)

vbm = u 3b vd
vcm = u 3c vd

(3.108)
(3.109)

where u 3a , u 3b , and u 3c are equal to 1/2, 1/2, or 0, depending on the status of the
switches. For a given value of vd , the number of possible values of vam , vbm , and
vcm is three; hence, this VSC is called three-level VSC. Similar to the derivation of
(3.103), it can be shown that

2/3 1/3 1/3


vam
van
vbn = 1/3 2/3 1/3 vbm
1/3 1/3 2/3
vcn
vcm

(3.110)

104

3 DC and Flexible AC Transmission Systems

S1

D1

S1

D1

Dc1 S3

D3

D5

Dc3 S 5

Dc5
vd 1

ea

eb

S5

D5

Rp

ib
ec

D3

ia
n

S3

+
C

ic
S4

D4

S6

D6

S2

D2
vd 2

S4

D4

Dc4 S6

D6

Dc6 S2

D2

Dc2

+
C

Rp

Fig. 3.29 Three-level diode-clamped converter


Table 3.5 Possible states of a leg of the three-level diode-clamped VSC
Controllable switches with control Controllable switches with vam
signal to turn on
control signal to turn off
S1 , S1
S4 , S4
S1 , S4

S4 , S4
S1 , S1
S1 , S4

Devices that conduct


ia > 0 ia < 0

vd /2
D1 , D1
vd /2 S4 , S4
0
S4 , Dc4

S1 , S1
D4 , D4
Dc1 , S1

If GTO thyristors are used as controllable switches, each GTO thyristor is turned
on and off only once in a cycle and u 3a is as shown in Fig. 3.30. u 3a possesses
quarter-wave symmetry. The magnitude of the fundamental component of van , vbn ,
and vcn is controlled by varying . If vd is constant, the harmonic components present
in van , vbn , and vcn are of order h = 6k 1, k = 1, 2, 3, . . . If IGBT is used as the
controllable switch, in addition to controlling the magnitude of the fundamental component, some harmonic components can be eliminated by increasing the switching
frequency.

3.3.3.3 Multi-Pulse VSC


The VSCs described in Sects. 3.3.3.1 and 3.3.3.2 are called six-pulse VSCs. A multipulse VSC is used to reduce the harmonic content in the AC-side voltages. The circuit
diagram of the 12-pulse VSC, which is a multi-pulse VSC, is shown in Fig. 3.31 [3].

3.3 FACTS

105

Fig. 3.30 Switching


functions u 3a

u3a

0.5

0.5

+180
o t (degree)

+360

The subscript h takes a value of 1 for the fundamental component and 6k 1,


k = 1, 2, 3, . . . for the harmonic components. Let Vah , Vbh , and Vch be the voltages
, V , and
of terminals a, b, and c, respectively, with respect to the neutral. Let Vah
bh

Vch be the voltages of terminals a , b , and c , respectively, with respect to the neutral.
The control signals to the controllable switches of the two six-pulse VSCs are given
such that
Va1 = V1 0, Vb1 = V1 (120 ), Vc1 = V1 120

(3.111)

Va1
= V1 (30 ), Vb1
= V1 (150 ), Vc1
= V1 90

(3.112)

The general expressions for the voltage phasors applicable to both the fundamental
component and the harmonic components are
Vah = Vh 0, Vbh = Vh (120 h), Vch = Vh (120 h)

Vah

= Vh (30 h),

Vbh

= Vh (150 h),

Vch

(3.113)

= Vh (90 h)

(3.114)

If the transformer is assumed to be ideal, from Fig. 3.31, the emfs in the transformer
windings are given by

Vbh
, V Bh = Vbh
Vch
, VCh = Vch
Vah
V Ah = Vah
1
1
1

= VCh
VAh = V Ah , VBh = V Bh , VCh
3
3
3

(3.115)
(3.116)

The voltages at the terminals a , b , and c with respect to the neutral are

Vah
= VAh + Vah

2V1 0 if h = 1
= 2Vh 0 if h = 6k 1, k = 2, 4, 6, . . .

0 if h = 6k 1, k = 1, 3, 5, . . .

(3.117)

106

3 DC and Flexible AC Transmission Systems

a
Six-pulse
VSC 1

b
c

a
Six-pulse
VSC 2

b
c

VBh
+

3N

VAh +

VCh

+ N

VCh

VAh +

c
a
b

VBh
+

Fig. 3.31 12-pulse VSC

Vbh
= VBh + Vbh

= VCh
+ Vch
Vch

2V1 (120 ) if h = 1
= 2Vh (120 ) if h = 6k 1, k = 2, 4, 6, . . . (3.118)

0 if h = 6k 1, k = 1, 3, 5, . . .

2V1 120 if h = 1
= 2Vh (120 ) if h = 6k 1, k = 2, 4, 6, . . . (3.119)

0 if h = 6k 1, k = 1, 3, 5, . . .

Therefore, the order of the harmonic components in the AC-side voltages of the
12-pulse VSC is h = 12k 1, k = 1, 2, 3, . . .
The concept of harmonic cancellation can be generalized. The schematic diagram
of a general multi-pulse VSC is shown in Fig. 3.32. It consists of n identical sixpulse VSCs which are connected in parallel on the DC side and in series on the AC
side through transformers. Each transformer has three windings in each phase. One
winding is wye-connected, and the other two windings are connected in zigzag. The
wye-connected winding is connected to the VSC. The circuit diagram of transformer
i is shown in Fig. 3.33. The subscript h is equal to 1 for the fundamental component
and is equal to 6k 1, k = 1, 2, 3, . . . for the harmonic components.
The control signals to the controllable switches of the six-pulse VSCs are given
such that the fundamental components of the voltages, with respect to the neutral, at
the AC-side terminals of (i + 1)th VSC are V1 (60 i/n), V1 (60 i/n 120 ),
and V1 (60 i/n + 120 ), i = 0, 1, 2, . . . n 1 [9]. For i = 1, 2, 3, . . . n 1, the
voltages at the terminals of the (i + 1)th VSC are nothing but the voltages across the
wye-connected winding of transformer i shown in Fig. 3.33. Therefore,

3.3 FACTS

107

Six-pulse
VSC 1

Six-pulse
VSC 2

Transformer 1

Six-pulse
VSC n

Transformer n1

Fig. 3.32 Schematic diagram of multi-pulse VSC

Viah +

Ni 2
+
N
Vich

+
Ni1

Vich

+
Viah

Viah +

Ni1

N
N

Vibh

Ni2

Ni1

Vibh

Vibh

+
+
N i2

Vich

Fig. 3.33 Transformer i of multi-pulse VSC

Via1 = V1 (60 i/n)


Vib1 = V1 (60 i/n 120 )

(3.120)
(3.121)

Vic1 = V1 (60 i/n + 120 )

(3.122)

If it is assumed that the transformers are ideal, the other voltages shown in Fig. 3.33
are given by

108

3 DC and Flexible AC Transmission Systems

Ni1
Ni1
Ni1

Viah , Vibh
Vibh , Vich
Vich
=
=
N
N
N
Ni2
Ni2
Ni2

Viah , Vibh
Vibh , Vich
Vich
=
=
=
N
N
N

Viah
=

(3.123)

Viah

(3.124)

The transformers provide a phase shift such that the fundamental component of the
voltage in zigzag winding is equal to the VSC 1 voltage both in magnitude and in
phase. This condition is given by the following equations:

Via1
Vib1
= V1 0

Vib1 Vic1 = V1 (120 )

(3.125)
(3.126)

Vic1
Via1
= V1 120

(3.127)

From (3.120) to (3.127),


Ni1
Ni2
V1 (60 i/n)
V1 (60 i/n 120 ) = V1 0,
N
N

i = 1, 2, . . . n 1
(3.128)

Equating the real and imaginary parts gives


60 i
1
Ni1
60 i
= cos
sin
N
n
n
3
Ni2
2
60 i
= sin
N
n
3

(3.129)
(3.130)

The zigzag windings are connected in series; the series-connected windings are connected to VSC 1 such
component of the resultant
n1that the hth-order harmonic
voltage in phase a is i=0
[(Ni1 /N )Vh (60 i h/n) (Ni2 /N )Vh (60 i h/n
120 h)], where N01  N and N02  0. Substituting h = 6k 1, k =
1, 2, 3, . . . gives
n1

Ni1
i=0

Ni2
Vh (60 i h/n)
Vh (60 i h/n 120 h)
N
N

nVh 0 if h = 6k 1 and k is a multiple of n


0 if h = 6k 1 and k is not a multiple of n

(3.131)

The VSC shown in Fig. 3.32 is called 6n-pulse VSC; for example, n = 3 for the
18-pulse VSC. The diagram of 18-pulse VSC is shown in Fig. 3.34.
3.3.3.4 Quasi Multi-Pulse VSC
The transformer requirement of a multi-pulse VSC is complicated. An alternative
to multi-pulse VSC is a quasi multi-pulse VSC which consists of only wye- or

3.3 FACTS

109

Six-pulse
VSC 1

Six-pulse
VSC 2

Six-pulse
VSC 3

Fig. 3.34 18-pulse VSC

delta-connected transformers [3]. The building block of a quasi multi-pulse VSC is


a 12-pulse VSC shown in Fig. 3.35. This circuit is different from the one shown in
Fig. 3.31.
Equations (3.113)(3.116) are applicable to the circuit shown in Fig. 3.35. The
voltage of terminal A with respect to terminal A is
VAh


+ Vah =

2Vh 0 if h = 1 or h = 6k 1, k = 2, 4, 6, . . .
(3.132)
0 if h = 6k 1, k = 1, 3, 5, . . .

Similarly, expressions can be written for voltages of terminals B and C with respect
to terminals B and C , respectively. The order of the harmonic components in the
AC-side voltages is h = 12k 1, k = 1, 2, 3, . . .
A quasi multi-pulse VSC is obtained by connecting 12-pulse VSCs in parallel on
the DC side and in series on the AC side. The schematic diagram of a general quasi
multi-pulse VSC is shown in Fig. 3.36. This VSC is called quasi 12n-pulse VSC.
For example, n = 2 for the quasi 24-pulse VSC.
The control signals are given such that the phase angle of the fundamental component of the AC-side voltage of the ith 12-pulse VSC is 30 (i 1)/n. Let Vh(12)

110

3 DC and Flexible AC Transmission Systems

B
A
C

a
Six-pulse
VSC 1

+
N
Vch

c
N
+

+
N
Vch

b
c

VBh
+

Vbh

Vbh

N
+

a
Six-pulse
VSC 2

Vah +

Vah +

+
VCh

3N

+
N

VCh

C
A
B

VBh

N
VAh +

VAh +

Fig. 3.35 Building block of quasi multi-pulse VSC


Fig. 3.36 Quasi multi-pulse
VSC

AC

DC
12-pulse VSC 1

12-pulse VSC 2

12-pulse VSC n

be the rms value of the voltage in each phase of a 12-pulse VSC, where h is equal
to 1 for the fundamental component and 12k 1, k = 1, 2, 3, . . . for the harmonic
components. The rms value of the voltage in each phase of the quasi multi-pulse
VSC is

3.3 FACTS

111



(12)
for different values of n and h
Table 3.6 Value of Vh / nVh
n=2
n=3
n=4

h=1

h = 11

h = 13

h = 23

h = 25

h = 35

h = 37

h = 47

h = 49

0.9914
0.9899
0.9893

0.1305
0.1053
0.0981

0.1305
0.0952
0.0861

0.9914
0.0952
0.0648

0.9914
0.1053
0.0648

0.1305
0.9899
0.0861

0.1305
0.9899
0.0981

0.9914
0.1053
0.9893

0.9914
0.0952
0.9893

Vh =

(12)
Vh

2 3
2 2 cos(30 h/n)

(3.133)

(12)

The value of Vh /(nVh ) for different values of n and h are given in Table 3.6. The
AC-side voltage of a quasi multi-pulse VSC has harmonic components of all order
which are present in a 12-pulse VSC. Though there is no elimination of harmonic
components, the magnitudes of the harmonic components are reduced.

3.3.3.5 Classification of VSC


There are two types of VSC depending on whether, for a given constant DC-side
voltage, the magnitude of the fundamental component of the AC-side voltages can be
controlled or not [3]: type 1 VSC and type 2 VSC. The magnitude of the fundamental
component of the AC-side voltages can be controlled in type 1 VSC for any given DC
voltage. In a type 2 VSC, the magnitude of the fundamental component of AC-side
voltages can be controlled only by changing the DC voltage. A two-level VSC is a
type 2 VSC if all controllable switches are turned on and turned off only once in a
cycle. A three-level VSC is an example of type 1 VSC.

3.3.4 STATCOM
A STATCOM is a shunt FACTS controller connected at a bus. The schematic diagram
of STATCOM is shown in Fig. 3.37.
STATCOM is mainly used to regulate voltage by generating or absorbing reactive
power. The reactive power is varied by varying the magnitude of the converter voltage.
If losses in the converter are neglected, and the transformer does not introduce a
phase shift, the fundamental component of the converter voltage and the fundamental
component of STATCOM bus voltage are in phase. In a STATCOM with type 2 VSC,
the converter voltage magnitude is altered by varying the DC voltage; the DC voltage
is varied by drawing/supplying active power from/to the network at the STATCOM
bus. Figure 3.38 shows the controller for STATCOM with type 2 VSC [3]. It consists
of two control loops: the outer voltage control loop and the inner reactive current
control loop. The voltage control loop is similar to the one shown in Fig. 3.18. V is

112

3 DC and Flexible AC Transmission Systems

VSC

Fig. 3.37 Schematic diagram of STATCOM


iRmax
Vref

i Rref +

Voltage
regulator

iRmin

Xs

iR

Reactive
current
controller

iR

+
V

Fig. 3.38 Controller for STATCOM with type 2 VSC

the magnitude of the STATCOM bus voltage, and i R is the reactive current drawn by
STATCOM. Vr e f and i Rr e f are the desired values of V and i R , respectively. X s is
positive. In steady state, is the angle by which the fundamental component of the
converter voltage leads the fundamental component of the STATCOM bus voltage;
if losses are neglected, = 0. In general (in steady state and during a transient),
decides the instant of switching on/off of the controllable switches. Change in
changes the amount of active power drawn/supplied by STATCOM, thereby changing
the DC voltage. The voltage regulator and reactive current controller are typically
proportionalintegral controllers. The proportional and integral gains of the voltage
regulator are negative, and the proportional and integral gains of the reactive current
controller are positive.

3.3.5 SSSC
A SSSC is connected in series with a transmission line. The schematic diagram of
SSSC is shown in Fig. 3.39. The three-phase winding of the transformer connected
to the VSC is wye- or delta-connected; the other three windings are connected in
series with the transmission line.

3.3 FACTS

113

VSC

Fig. 3.39 Schematic diagram of SSSC


vd

vRref

1/ k

Absolute

vdref

Sign

DC voltage
regulator

+
90

Fig. 3.40 Controller for SSSC with type 2 VSC

SSSC generates or absorbs reactive power. The reactive power is varied by varying
the magnitude of the converter voltage. If the losses in the VSC and the transformer
are neglected, the fundamental component of the voltage injected by SSSC lags or
leads the transmission line current by 90 . Figure 3.40 shows a part of the controller
for SSSC with type 2 VSC [3]. The output of the Absolute block is the absolute
value of its input. The output of the Sign block is 1 if its input is positive and 1 if
its input is negative. The controller consists of two control loops. The outer control
loop consists of a controller, for example CC controller shown in Fig. 3.22, which
generates v Rr e f (instead of X r e f ). v R and vd are the reactive voltage injected by
SSSC and the DC-side voltage, respectively, v Rr e f and vdr e f are the desired values
of v R and vd , respectively. In steady state, if vd is assumed to be constant, the reactive
voltage injected by SSSC is v R kvd , where k is a constant which includes the effect
of transformer also. In steady state, is the phase angle by which the fundamental
component of the voltage injected by SSSC leads the fundamental component of
the transmission line current; the polarity of the voltage and the direction of current
are as shown in Fig. 3.40. In the inner control loop, the reactive voltage is regulated
indirectly; change in changes the amount of real power drawn by SSSC, thereby
changing the DC voltage. The DC voltage regulator is typically a proportional
integral controller; the proportional and integral gains are positive. In steady state, if
losses are neglected, = 90 . In general (in steady state and during a transient),
decides the instant of switching on/off of the controllable switches.

114

3 DC and Flexible AC Transmission Systems

Fig. 3.41 Schematic diagram


of UPFC

VSC

VSC

Fig. 3.42 Schematic diagram


of IPFC

VSC

VSC

3.3.6 Multi-Converter FACTS Controllers


STATCOM and SSSC generate or absorb reactive power which can be controlled;
in other words, these FACTS controllers have one degree of freedom. If two such
VSC-based FACTS controllers are at the same location, then there can be three
degrees of freedom if the two FACTS controllers are connected in parallel on the
DC side, resulting in a multi-converter FACTS controller [1, 3]. The third degree of
freedom is amount of power flow via the DC link. If the losses are neglected, the net
active power drawn by the multi-converter FACTS controller is zero. Figures 3.41
and 3.42 show the schematic diagrams of multi-converter FACTS controllers: UPFC
and IPFC. UPFC has a VSC connected in shunt through a transformer and a VSC
connected in series through a transformer. IPFC has both VSCs connected in series
through transformers.

References
1. N.G. Hingorani, L. Gyugyi, Understanding FACTS: Concepts and Technology of Flexible AC
Transmission Systems (IEEE Press, New York, 2000)
2. N. Mohan, T.M. Undeland, W.P. Robbins, Power Electronics: Converters, Applications, and
Design, 2nd edn. (Wiley, New York, 1995)
3. K.R. Padiyar, FACTS Controllers in Power Transmission and Distribution (New Age International, New Delhi, 2007)
4. K.R. Padiyar, HVDC Power Transmission Systems, 2nd edn. (New Age International, New Delhi,
2010)

References
5.
6.
7.
8.

115

E.W. Kimbark, Direct Current Transmission (Wiley, New Delhi, 1971)


J. Vithayathil, Power Electronics: Principles and Applications (McGraw-Hill, New York, 1995)
P. Kundur, Power System Stability and Control (Tata McGraw-Hill, Noida, 1994)
R.M. Mathur, R.K. Varma, Thyristor-Based FACTS Controllers for Electrical Transmission
Systems (IEEE Press, New York, 2002)
9. G. Joos, Power Electronics: Fundamentals, ed. by Y.H. Song, A.T. Johns Flexible AC Transmission Systems (FACTS), (IEE, London, 1999)

Chapter 4

Prime Movers and Excitation System

4.1 Prime Movers


4.1.1 Steam Turbine
A steam turbine has two or more turbine sections. Figure 4.1 gives the model of
one type of steam turbine known as tandem compound single reheat turbine [1, 2]
which has three turbine sections: high pressure (HP), intermediate pressure (IP), and
low pressure (LP) turbine sections. Y is valve position; 0 Y 1. TCH is the time
constant of the steam chest and the inlet piping. TRH is the reheater time constant and
TCO is the time constant of the crossover piping. FHP , FIP , and FLP are the fractions
of the total power generated by HP, IP, and LP turbine sections, respectively. THP , TIP ,
and TLP are the torques generated by HP, IP, and LP turbine sections, respectively.
Tmax is the maximum total mechanical torque.
The per unit mechanical torque generated by the turbine driving the equivalent
synchronous generator with two field poles is
Tm 2Tmax
T m =
Tmax pf TB

(4.1)

4.1.2 Hydraulic Turbine


Figure 4.2 gives the model of the hydraulic turbine [1, 2]. Pm is the mechanical
power, G is the gate position, and Tw is the water starting time. Pmo and Go are the
initial values of Pm and G, respectively. G = Go + G and 0 G 1.
The per unit mechanical torque generated by the turbine driving the equivalent
synchronous generator with two field poles is

S Krishna, An Introduction to Modelling of Power System Components,


SpringerBriefs in Electrical and Computer Engineering,
DOI: 10.1007/978-81-322-1847-0_4, The Author(s) 2014

117

118

4 Prime Movers and Excitation System


+
THP
Tmax

TIP
Tmax

1
1+ sTCH

Tm
Tmax

FIP

FHP

TLP
Tmax

FLP
1
1+ sTCO

1
1+ sTRH

Fig. 4.1 Model of tandem compound single reheat turbine

Pmo 1 Tw s
Go 1+0.5 Tw s

Pm

Fig. 4.2 Model of hydraulic turbine

HP turbine
section

IP turbine
section

LP turbine
section

Synchronous
generator
rotor
1

Fig. 4.3 Structure of tandem compound single reheat turbine

Pmo + Pm
T m =
TB

(4.2)

4.2 Torsional Dynamics


A steam turbine is made up of turbine sections connected by shafts of finite stiffness. Figure 4.3 shows the structure of a synchronous generator driven by a tandem
compound single reheat turbine. Each turbine section and the synchronous generator
rotor can be represented by a mass. Masses 2, 3, and 4 in Fig. 4.3 represent LP, IP,
and HP turbine sections, respectively, mass 1 represents the synchronous generator
rotor.
Let the torques generated by LP, IP, and HP turbine sections, driving the equivalent
, T , and T , respectively. The
synchronous generator with two field poles, be TLP
IP
HP
equations governing the torsional dynamics are [1, 2]
d1
= 1 o
dt
d2
= 2 o
dt
d3
= 3 o
dt

(4.3)
(4.4)
(4.5)

4.2 Torsional Dynamics

119

d4
= 4 o
dt

1
d1
= K12 (2 1 ) Te
dt
J1

d2
1
= TLP + K23 (3 2 ) K12 (2 1 )
dt
J2

1
d3
= TIP
+ K34 (3 4 ) K23 (3 2 )
dt
J3

d4
1
= THP
K34 (3 4 )
dt
J4

(4.6)
(4.7)
(4.8)
(4.9)
(4.10)

where i is the angular position of mass i in electrical radian with respect to a reference
rotating at speed o , i is the speed of mass i in electrical radian per second, Ji is
the moment of inertia of mass i, and Kij is the stiffness of the shaft between masses
i and j. In some studies, the stiffness of shafts are assumed to be large and hence
1 = 2 = 3 = 4 = and 1 = 2 = 3 = 4 = . Under such an assumption,
(4.3)(4.6) are equivalent to (1.132); similarly, (4.7)(4.10) are equivalent to (1.133),
where
J  J1 + J2 + J3 + J4

(4.11)

Tm  TLP
+ TIP
+ THP

(4.12)

If equations in per unit quantities are required, (4.7)(4.10) are replaced by the
following equations.
d1
dt
B
d2
=
dt
2H2
B
d3
=
dt
2H3
d4
dt

B
K12 (2 1 ) T e
2H1

(4.13)

T LP + K 23 (3 2 ) K 12 (2 1 )

+ K 34 (3 4 ) K 23 (3 2 )
T IP

B
THP K 34 (3 4 )
2H4

(4.14)
(4.15)
(4.16)

where Hi  Ji B2 /(2SB ).

4.3 Speed Governor


A governor is used to regulate the frequency/speed. The governor adjusts the turbine
valve/gate to change the mechanical power/torque. Figure 4.4 shows a governor for
the steam turbine, which is an integral controller; K is positive. Y is the change in

120

4 Prime Movers and Excitation System

Fig. 4.4 Isochronous


governor

K
s

Fig. 4.5 Governor with droop

K
s

B
R

Fig. 4.6 Governor with droop


and load reference setpoint

K
s

Y
+

B
R

Load
reference
setpoint

valve position. This governor is called an isochronous governor since it tries to bring
the frequency back to the nominal value. The steady-state or operating speed o is
equal to B .
Isochronous governor works satisfactorily when there is a single synchronous
generator in the system or when only one synchronous generator is required to
respond to changes in frequency. The isochronous governor cannot be used when
two or more synchronous generators in a system are required to respond to changes
in frequency. This is because, due to inevitable errors in the speed measurement,
which are different for different governors, a steady-state frequency would not be
reached. Therefore, a feedback loop with positive gain R is added as shown in Fig.
4.5 [2]; R is referred to as droop. In steady state,
1
Y = ( B )
R

(4.17)

If the value of R of different governors are nearly equal, then for a given change in
frequency, change in power output of the generators is nearly in proportion to the
rating. The relation between speed and valve position can be adjusted by changing
an input known as load reference setpoint which is added in the feedback loop of the
governor as shown in Fig. 4.6 [2]. Therefore, in steady state,
1
Y = ( B ) + Load reference setpoint
R
With this governor, the operating speed o is in general not equal to B .

(4.18)

4.3 Speed Governor

121

RT

sT R
1+ sT R

K
s

G
+

Load
reference
setpoint

Fig. 4.7 Governor for hydraulic turbine

Exciter
transformer

Controlled
rectifier

Slip
ring

Synchronous generator
Field Armature

Potential
transformer
Voltage
regulator

Fig. 4.8 Schematic diagram of static excitation system

The hydraulic turbine has a peculiar response: if G is a step function, from


Fig. 4.2,
Pm =

Pmo
(1 3e2t/Tw )G
Go

(4.19)

The initial change in turbine power is opposite to that sought. Therefore, the governor
for a hydraulic turbine is provided with a large temporary droop (RT ) with a long
reset time (TR ), as shown in Fig. 4.7 [2].

4.4 Excitation System


The excitation system provides voltage to the synchronous generator field winding. This voltage is varied in order to regulate the synchronous generator terminal
voltage. One type of excitation system which is prevalent nowadays is the static excitation system the schematic diagram of which is shown in Fig. 4.8 [2]. The exciter
transformer steps down the voltage.

122

4 Prime Movers and Excitation System

Fig. 4.9 Model of static


excitation system

Efmax
Vref

Ef
Efmin

The delay angle of the controlled rectifier is set by the voltage regulator. Let
the magnitude of the voltage input to the controlled rectifier be V . Let it be assumed
that the field current is constant and that the input voltage to the controlled rectifier
is sinusoidal and balanced. If the instantaneous value of Ef is approximated by its
average value, then from (1.194) and (3.4),

3 2Mdf
V cos
Ef =
Rf

(4.20)

The voltage regulator is designed such that the static excitation system is represented
by the model shown in Fig. 4.9 where K is positive and
Ef max

3 2Mdf
3 2Mdf
=
V , Ef min =
V cos max
Rf
Rf

(4.21)

where max is the maximum value of . V is the magnitude of the voltage at the
generator terminals and Vref is the desired value of V .
The cosinusoidal dependence of Ef on is negated by the appropriate choice of
[3] given by

K(Vref V )

1 Rf

cos
3 2Mdf V
=

max
0

if Ef min K(Vref V ) Ef max


if K(Vref V ) < Ef min
if K(Vref V ) > Ef max

(4.22)

In per unit quantities, the equation governing the excitation system is

V ref V )
K(
Ef = E f min

Ef max
where K  KVB /B .

if
if
if

V ref V ) E f max
E f min K(

K(Vref V ) < E f min


V ref V ) > E f max
K(

(4.23)

References

123

References
1. K.R. Padiyar, Power System Dynamics: Stability and Control, 2nd edn. (BS Publications,
Hyderabad, 2002)
2. P. Kundur, Power System Stability and Control (Tata McGraw-Hill, Noida, 1994)
3. J. Machowski, J.W. Bialek, J.R. Bumby, Power System Dynamics: Stability and Control, 2nd
edn. (Wiley, New York, 2008)

Appendix A

Solution of Linear Ordinary Differential


Equations with Constant Coefficients

A set of n first-order linear ordinary differential equations with constant coefficients


can be written as follows:
dy
= Ay + u
(A.1)
dx
where y is a n 1 vector, A is a n n matrix, x is the independent variable, and u
is a n 1 vector dependent on x.
i is said to be an eigenvalue of A if there exists a non zero n 1 vector vi such
that
(A.2)
Avi = i vi
vi is called a right eigenvector of A corresponding to the eigenvalue i . The n n
matrix A has n eigenvalues 1 , 2 , . . . n . For a real matrix A, if i is a complex eigenvalue of A and vi is a corresponding right eigenvector, then i is also an eigenvalue
of A and vi is a right eigenvector corresponding to i .
If the eigenvalues are distinct, then the solution of (A.1) is
y = Te

(xx0 )

y(x0 ) + T e

e T 1 u( )d

(A.3)

x0

where x0 is a value of x for which y is known, and


T  [ v1 v2
x
e 1
0

.
x
e 
.

.
0

. . . vn ]

0 ... 0
e2 x . . . 0

.
.

.
.

.
.
0 . . . en x

S Krishna, An Introduction to Modelling of Power System Components,


SpringerBriefs in Electrical and Computer Engineering,
DOI: 10.1007/978-81-322-1847-0, The Author(s) 2014

(A.4)

(A.5)

125

Appendix B

Fourier Series

The Fourier series of a periodic function f (o t), with period 2 , is given by

f (o t) =

a0
+
[ah cos(ho t) + bh sin(ho t)]
2

(B.1)

h=1

where
1
ah =

bh =

c+2

f (o t) cos(ho t)d(o t)

(B.2)

f (o t) sin(ho t)d(o t)

(B.3)

c+2
c

c can be chosen arbitrarily. a0 /2 is the average value


of f (o t). The rms value of
the hth-order harmonic component of f (o t) is (ah2 + bh2 )/2. The Fourier series
of f (o t) is also given by
f (o t) =

ch ejho t

(B.4)

f (o t)ejho t d(o t)

(B.5)

h=

where
ch =

1
2

c+2

From (B.2), (B.3), and (B.5),


a0
2
ah jbh
, h = 1, 2, 3, ...
ch =
2
c0 =

S Krishna, An Introduction to Modelling of Power System Components,


SpringerBriefs in Electrical and Computer Engineering,
DOI: 10.1007/978-81-322-1847-0, The Author(s) 2014

(B.6)
(B.7)

127

128

Appendix B: Fourier Series

ch =

ah + jbh
, h = 1, 2, 3, ...
2

The rms value of the hth-order harmonic component of f (o t) is


If f (o t) is an odd function, i.e., f (o t) = f (o t),

(B.8)

2|ch |.

ah = 0

2
f (o t) sin(ho t)d(o t)
bh =
0

(B.9)
(B.10)

If f (o t) is an even function, i.e., f (o t) = f (o t),


2

bh = 0

ah =

f (o t) cos(ho t)d(o t)

(B.11)

(B.12)

If f (o t) possesses half-wave symmetry, i.e., f (o t + ) = f (o t),

0
ah = 2

bh = 2

if h = 0, 2, 4, ...
c+
f (o t) cos(ho t)d(o t) if h = 1, 3, 5, ...

(B.13)

if h = 2, 4, 6, ...
c+
f (o t) sin(ho t)d(o t) if h = 1, 3, 5, ...

(B.14)

f (o t) is said to possess quarter-wave symmetry if f (o t) possesses half-wave


symmetry and there exists a such that f (o t + ) = f (o t + ). If f (o t)
possesses quarter-wave symmetry, then for the function f (o t + ),

0 if h = 0, 2, 4, ...
ah = 4 /2

f (o t + ) cos(ho t)d(o t) if h = 1, 3, 5, ...


0
bh = 0

(B.15)
(B.16)

It is to be noted that the rms values of a harmonic component of f (o t) and f (o t+)


are equal.
The harmonic components of order 3 and its multiples are called triplen harmonic
components. The triplen harmonic components of f (o t), f (o t 2/3), and
f (o t + 2/3) are equal. Hence, the triplen harmonic component of f (o t) is
equal to one third of the triplen harmonic component of f (o t) + f (o t 2/3) +
f (o t + 2/3). Therefore, the triplen harmonic components are equal to zero if
f (o t) + f (o t 2/3) + f (o t + 2/3) = 0.

About the Author

S Krishna is Assistant Professor of Electrical Engineering at Indian Institute of Technology Madras, Chennai, India. He received B.E. degree from Bangalore University
in 1995 and M.E. and Ph.D. degrees from Indian Institute of Science, Bangalore, in
1999 and 2003, respectively. He worked with Kirloskar Electric, Bangalore, from
1995 to 1997 and M.S. Ramaiah Institute of Technology, Bangalore, from 2003 to
2008. His areas of research interest are power system dynamics and control. He is
an Associate Editor of SADHANA, a Journal of the Indian Academy of Sciences.

S Krishna, An Introduction to Modelling of Power System Components,


SpringerBriefs in Electrical and Computer Engineering,
DOI: 10.1007/978-81-322-1847-0, The Author(s) 2014

129

About the Book

The brief provides a quick introduction to the dynamic modelling of power system
components. It gives a rigorous derivation of the model of different components
of the power system such as synchronous generator, transformer, transmission line,
FACTS, DC transmission system, excitation system, and speed governor. Models of
load and prime movers are also discussed. The brief can be used as a reference for
researchers working in the areas of power system dynamics, stability analysis, and
design of stability controllers. It can also serve as a text for a short course on power
system modelling or as a supplement for a senior undergraduate/graduate course on
power system stability.

S Krishna, An Introduction to Modelling of Power System Components,


SpringerBriefs in Electrical and Computer Engineering,
DOI: 10.1007/978-81-322-1847-0, The Author(s) 2014

131

Index

A
Active current, 67
Active power, 67, 68, 71
Active voltage, 67
Amperes law, 45, 52
Angle of advance, 77

B
Bergerons method, 62
Bidirectional switch, 98, 101
Bundled conductor, 53, 56

C
Characteristic impedance, 62, 63
Commutation angle, 78
Commutation margin angle, 77, 80
Composite conductor, 53, 58
Current magnitude, 67, 97
Current phase angle, 67

D
DC transmission system, 74
Delay angle, 76, 122
Diode, 73
Diode-clamped converter, 101
Distribution network, 71
Double circuit transmission line, 53, 56

E
Eigenvalue, 125
Eigenvector, 125
Equivalent circuit of transformer, 46, 47, 49

Equivalent circuit of transmission line, 64, 69


Even function, 128
Excitation system, 121
Extinction angle, 80

F
FACTS, 86
Fourier series, 127
Frequency, 66, 71

G
Gate turn off thyristor, 74, 100, 104

H
Half-wave symmetry, 79, 128
Hydraulic turbine, 117, 121

I
Ideal transformer, 45, 83, 89, 105, 107
Insulated gate bipolar transistor, 74, 100, 104
Interline power flow controller, 86, 114

K
Krons transformation, 65, 92

L
Line-commutated converter, 75
Linear differential equation, 125
Load, 71

S Krishna, An Introduction to Modelling of Power System Components,


SpringerBriefs in Electrical and Computer Engineering,
DOI: 10.1007/978-81-322-1847-0, The Author(s) 2014

133

134
M
Multi-converter FACTS, 114

Index
steady-state characteristic, 94
Steam turbine, 117119
Stranded conductor, 53, 56
Subtransmission network, 71

N
Nominal circuit of transmission line, 64
O
Odd function, 128
Overlap angle, 78

P
Periodic function, 127
Phase velocity, 62
Power semiconductor device, 73, 86
Prime mover, 117
Propagation constant, 63

Q
Quarter-wave symmetry, 95, 100, 104, 128

R
Reactive current, 67, 92, 112
Reactive power, 67, 71
Reactive voltage, 67, 113

S
Selective harmonic elimination, 100
Series FACTS controller, 86, 94, 112
Shaft stiffness, 118
Shunt FACTS controller, 86, 111
Speed governor, 119
droop, 120
isochronous governor, 120
load reference setpoint, 120
Static excitation system, 121
Static synchronous compensator, 86, 111
Static synchronous series compensator, 86, 112
Static var compensator, 86
controller, 91

T
Tandem compound single reheat turbine, 117,
118
Three-phase transformer, 47
Three-winding transformer, 46, 89, 106
Thyristor, 73, 75, 86, 95
Thyristor-controlled reactor, 86, 94
Thyristor-controlled series capacitor, 86, 94
constant current control, 96
reactance, 96
resonance, 96
steady-state characteristic, 98
Torsional dynamics, 118
Transmission line, 50
capacitance, 56
effect of earth on capacitance, 59
inductance, 50
lossless transmission line, 62
model, 60
Triplen harmonic component, 88, 99, 128

U
Unified power flow controller, 86, 114

V
Variable impedance-type FACTS, 86
Voltage magnitude, 66, 71, 93, 98, 112, 122
Voltage phase angle, 66
Voltage source converter, 86, 98
multi-pulse converter, 104
quasi multi-pulse converter, 108
three-level converter, 101, 111
two-level converter, 98, 111
type 1 converter, 111
type 2 converter, 111, 113
Voltage source converter (VSC)-based FACTS,
86, 114