View
217
Download
0
Category
Preview:
Citation preview
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 1/120
Detailed analysis of a multi-pulse STATCOM
Ricardo Dávalos Marín
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 2/120
2
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 3/120
Abstract
The rapid development of power electronics technology provides opportunities
to develop new power equipment to improve the performance of the actual power
systems. During the last decade, a number of control devices called ”Flexible
AC Transmission Systems” (FACTS) technology have been proposed and imple-
mented. FACTS devices can be used for power flow control, loop-flow control,
voltage regulation, enhancement of transient stability and damping of power os-
cillations. FACTS devices can be used as a series controller, shunt controllers or
by a combination of both.
Since the early 1970s high power, line-commutated thyristors in conjunction
with capacitors and reactors have been employed in various circuit configurations
to produce variable output such as the shunt connected static VAR compensators(SVC) and the series connected thyristor controlled series capacitor (TCSC), based
on the traditional thyristor-switched capacitors (TSC) and thyristor-controlled re-
actors (TCR), which have been widely used for the AC voltage regulation in power
systems by controlling the injection of reactive power. With the advent of high
power gate turn-off thyristor (GTO) and other power semiconductors with an in-
ternal turn-off capability as the insulated gate bipolar transistor (IGBT), a new
generation of power electronic equipment has been implemented in switching con-
verter circuits, the voltage source inverters (VSI), to generate and absorb reactive
power without the use of AC capacitor or reactor banks.
The new generation and most dominant converters needed in FACTS con-
trollers such as the static synchronous compensator (STATCOM), the static syn-
chronous series compensator (SSSC) and by the combination of both the Unified
Power Flow Controller (UPFC) are based on the voltage-source inverters (VSI).
This work presents the detailed analysis to deduce the expressions for the AC
(phase currents and output voltage) and DC signals (capacitor current and voltage)
of 6-, 12-, 24-, and 48-pulse VSI-STATCOM. The expressions obtained allow
to estimate an appropriate DC capacitor value. Based on a switching model, a
state-space representation in the dq-reference frame is deduced. The accuracy of
the results are validated by comparison the analytical results along with digital
simulations using the EMTDC/PSCAD.
The relevance of such studies is to exhibit the detailed STATCOM operation,
not found commonly in the literature, emphasising assumptions and limitations
when the dq0 model is used to carry out dynamic studies in large power systems.
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 4/120
ContentsContents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I
List of figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Multi-pulse converter configuration . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Multi-level converter configuration . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Pulse Width Modulation (PWM) . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.4 Objetives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9
2 Analysis of a STATCOM based on 6- and 12-pulses VSI . . . . . . . . . . . 11
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2 Voltage source-inverter (VSI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.1 Harmonic analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3 Static synchronous compensator (STATCOM) based on
six-pulse VSI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3.1 Reactive power exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Analysis of the AC current signals . . . . . . . . . . . . . . . . . . . 23
Conduction period transistors and diodes . . . . . . . . . . . . . 27
Capacitor current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28DC capacitor voltage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.3.2 Reactive and active power exchange . . . . . . . . . . . . . . . . . . . . . . 32
Capacitor current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.4 12-pulse converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.4.1 AC current signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Six-pulse AC current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.4.2 Capacitor current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.4.3 DC capacitor voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.4.4 PSCAD/EMTDC Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3 24- and 48-pulse STATCOM operation . . . . . . . . . . . . . . . . . . . . . . . . 59
I
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 5/120
3.1 24-pulse operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.1.1 24-pulse voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.1.2 Magnetic interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.1.3 AC current signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.1.4 Capacitor current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.1.5 DC capacitor voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.2 48-pulse operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.2.1 48-pulse voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.2.2 AC current signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.2.3 Capacitor current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.2.4 DC capacitor voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.3 Digital simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4 STATCOM Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.1 Switching functions model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.1.1 12-pulse converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.2 STATCOM model at fundamental frequency . . . . . . . . . . . . . . . . 92
4.2.1 12-pulse converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
4.2.2 24-pulse converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
4.2.3 48-pulse converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
4.3 dq0 Reference frame model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1014.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
II
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 6/120
List of figures
1.1 One-line diagram of a STATCOM . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Three phase converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Multi-pulse staircase voltage waveform . . . . . . . . . . . . . . . . . . . . 5
1.4 Five-level voltage source inverter . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.5 Single-phase three level Chain circuit . . . . . . . . . . . . . . . . . . . . . 6
2.1 Simple voltage sourced-inverter (two level-pole) . . . . . . . . . . . . . 11
2.2 Six-pulse VSI with a resistive load . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3 Firing control signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4 Line-to-line voltage waveform . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.5 Equivalent circuit for the sequence 1-5-6 . . . . . . . . . . . . . . . . . . . 14
2.6 Equivalent circuit for the sequence 1-2-6 . . . . . . . . . . . . . . . . . . . 15
2.7 Equivalent circuit for the sequence 1-2-3 . . . . . . . . . . . . . . . . . . . 16
2.8 Line-to-neutral voltage waveform . . . . . . . . . . . . . . . . . . . . . . . . 16
2.9 vab(t) voltage Fourier spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.10 0.5 inductive power factor load conduction period . . . . . . . . . . . 20
2.11 0.8660 inductive power factor load conduction period . . . . . . . . . 21
2.12 Inductive load conduction period . . . . . . . . . . . . . . . . . . . . . . . . 21
2.13 Six-pulse VSI-STATCOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.14 Phase relationship between the inductor voltage V L and the
fundamental curren I a1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.15 Relationship between DC voltage, fundamental and
harmonics current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.16 AC system voltage and voltage compensator van(t) . . . . . . . . . . . 25
III
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 7/120
2.17 AC current waveform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.18 Q1 and D1 conduction period . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.19 Capacitor current; a)generating reactive power; b)absorbing
reactive power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.20 DC capacitor voltage; a)generating reactive power;
b)absorbing reactive power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.21 Voltage waveforms with a finite DC capacitor . . . . . . . . . . . . . . . 33
2.22 Capacitor current and DC capacitor voltage waveform . . . . . . . . 33
2.23 AC system voltage and compensator voltage; van(t) . . . . . . . . . . 34
2.24 AC current waveform with φ = 15o . . . . . . . . . . . . . . . . . . . . . . . 36
2.25 Instantaneous capacitor current . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.26 Test circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.27 Instantaneous capacitor voltage . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.28 AC instantaneous current with: a)φ = 0o;b)φ = −0.5o . . . . . . . . 39
2.29 a)vab(t) and vabY (t)2; b)12-pulse voltage . . . . . . . . . . . . . . . . . . . 42
2.30 12-pulse VSI-STATCOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.31 12-pulse STATCOM line-to-neutral voltages . . . . . . . . . . . . . . . . 43
2.32 vab(t)12 voltage Fourier spectrum . . . . . . . . . . . . . . . . . . . . . . . . 45
2.33 Phasorial diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.34 Relationship between the DC voltage and the fundamental
and harmonics current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.35 AC system voltage and fundamental voltage compensator
van(t) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.36 AC current waveform, ia(t) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.37 AC current of each six-pulse VSI . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.38 Q1 and D1 currents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
IV
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 8/120
2.39 First converter capacitor current; a)generating reactive
power; b)absorbing reactive power . . . . . . . . . . . . . . . . . . . . . . . 53
2.40 Second converter capacitor current; a)generating reactive
power; b)absorbing reactive power . . . . . . . . . . . . . . . . . . . . . . . 53
2.41 Capacitor current; a)generating; b)absorbing reactive
power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.42 DC capacitor voltage; a)generating; b)absorbing reactive
power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.43 a)12-pulse AC current ia(t); b)12-pulse line-to-line voltage
vab(t) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.44 Capacitor current and DC capacitor voltage waveform . . . . . . . . 56
3.1 24-pulse voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.2 24-pulse voltage Fourier spectrum . . . . . . . . . . . . . . . . . . . . . . . . 62
3.3 24-pulse STATCOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.4 12-pulse grounded transformer with PST on leading and
lagging configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.5 Phasor diagram of the phase-shifting mechanism . . . . . . . . . . . . 65
3.6 AC current waveform, ia(t) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.7 Phasorial diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.8 Capacitor current; a)generating reactive power; b)absorbing
reactive power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.9 DC capacitor voltage; a)generating reactive power;
b)absorbing reactive power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.10 48-pulse STATCOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.11 48-pulse voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.12 48-pulse voltage Fourier spectrum . . . . . . . . . . . . . . . . . . . . . . . . 76
3.13 AC current waveform, ia(t) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.14 Capacitor current; a)generating reactive power; b)absorbing
V
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 9/120
reactive power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.15 Capacitor voltage; a)generating reactive power; b)absorbing
reactive power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3.16 24-pulse behaviour with a fixed V DC value . . . . . . . . . . . . . . . . . 81
3.17 24-pulse behaviour with a finite DC capacitor . . . . . . . . . . . . . . . 82
3.18 48-pulse behaviour with fixed V DC value . . . . . . . . . . . . . . . . . . . 83
3.19 48-pulse behaviour with a finite DC capacitor . . . . . . . . . . . . . . . 84
4.1 Six-pulse VSI-STATCOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.2 phase ’a’ arm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.3 Six-pulse behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.4 12-pulse behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.5 24-pulse behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
4.6 48-pulse behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
4.7 Phase current; a)Switching model; b)model at fundamental
frequency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
4.8 id and iq current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4.9 Capacitor voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
VI
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 10/120
Chapter 1
Introduction
In recent years, voltage stability and control are increasingly becoming a limiting factor in the planning
and operation of some power systems, mainly in longitudinal ones. However, a variety of considerations
constrains the construction of new transmission lines. This has been reflected in the necessity to maximise
the use of existing transmission facilities. On steady state, bus voltages must be controlled on a specified
range. A suitable voltage and reactive power control allows to obtain important benefits in the power systems
operation such as the reduction of voltage gradients, the ef ficient transmission capacities utilisation and the
increase of stability margins. By different control means and operating techniques, the voltage control task
in transmission levels can be got; some solution technologies can involve a series voltage injection, or a
shunt reactive current injection in strategic sites of the power system. When a disturbance occurs, changes
in the voltage system are presented and the restoration to the reference values depends on the dynamic
response of the excitation systems and the control devices employed.
In the last decade commercial availability of Gate Turn-Off thyristor (GTO) devices with high power
handling capability, and the advancement of other types of power-semiconductor devices such as IGBT’s
have led to the development of controllable reactive power sources utilising electronic switching converter
technology [1]. These technologies additionally offer considerable advantages over the existing ones in
terms of space reductions and performance. The GTO thyristors enable the design of solid-state shunt
reactive compensation equipment based upon switching converter technology. This concept was used to
create a flexible shunt reactive compensation device named Static Synchronous Compensator (STATCOM)
due to similar operating characteristics to that of a synchronous compensator but without the mechanical
inertia.
The advent of Flexible AC Transmission Systems (FACTS) is giving rise to a new family of power
electronic equipment emerging for controlling and optimising the performance of power system, e.g. STAT-
COM, SSSC and UPFC. The use of voltage-source inverter (VSI) has been widely accepted as the next
generation of reactive power controllers of power system to replace the conventional VAR compensation,
such as the thyristor-switched capacitor (TSC) and thyristor controlled reactors (TCR).
1
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 11/120
DC-AC
Switching
converter
System bus VAC
Coupling
Transformer
Transformer leakage
inductor
Vs
VDC
C
I
Fig. 1.1 One-line diagram of a STATCOM
Some researchers are aiming their efforts to apply FACTS in different ways to enhance the power sys-
tems operation. The major applications are: voltage stability enhancement, damping torsional oscillations,
power system voltage control, and power system stability improvement. These applications can be imple-
mented with a suitable control (voltage magnitude and phase angle control) [2-4].
The Static Synchronous Compensator (STATCOM) is a shunt connected reactive compensation equip-
ment which is capable of generating and/or absorbing reactive power whose output can be varied so as to
maintain control of specific parameters of the electric power system. The STATCOM provides operating
characteristics similar to a rotating synchronous compensator without the mechanical inertia, due to the
STATCOM employ solid state power switching devices it provides rapid controllability of the three phase
voltages, both in magnitude and phase angle.
The STATCOM basically consists of a step-down transformer with a leakage reactance, a three-phase
GTO or IGBT voltage source inverter (VSI), and a DC capacitor. The AC voltage difference across the
leakage reactance produces reactive power exchange between the STATCOM and the power system, such
2
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 12/120
that the AC voltage at the bus bar can be regulated to improve the voltage profile of the power system, which
is the primary duty of the STATCOM. However, for instance, a secondary damping function can be added
into the STATCOM for enhancing power system oscillation stability [4]. The basic voltage-source inverter representation for reactive power generation is shown schematically in Fig.1.1.
The principle of STATCOM operation is as follows. The VSI generates a controllable AC voltage
source behind the leakage reactance. This voltage is compared with the AC bus voltage system; when the
AC bus voltage magnitude is above that of the VSI voltage magnitude, the AC system sees the STATCOM
as an inductance connected to its terminals. Otherwise, if the VSI voltage magnitude is above that of the
AC bus voltage magnitude, the AC system sees the STATCOM as a capacitance connected to its terminals.
If the voltage magnitudes are equal, the reactive power exchange is zero. If the STATCOM has a DC source
or energy storage device on its DC side, it can supply real power to the power system. This can be achieved
adjusting the phase angle of the STATCOM terminals and the phase angle of the AC power system. When
the phase angle of the AC power system leads the VSI phase angle, the STATCOM absorbs real power from
the AC system; if the phase angle of the AC power system lags the VSI phase angle, the STATCOM supplies
real power to AC system [5-7].
Typical applications of STATCOM are:
• effective voltage regulation and control.
• reduction of temporary overvoltages.
• improvement of steady-state power transfer capacity.
• improvement of transient stability margin.
• damping of power system oscillations.
• damping of subsynchronous power system oscillations.
• flicker control.
• Power quality improvement.
• distribution system applications.
The voltage source-converter or inverter (VSC or VSI) is the building block of a STATCOM and other
FACTS devices. A very simple inverter produces a square voltage waveform as it switches the direct voltage
source on and off. The basic objective of a VSI is to produce a sinusoidal AC voltage with minimal harmonic
distortion from a DC voltage. Three basic techniques are used for reducing harmonics in the converter output
3
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 13/120
Q1
Q4
Q3 Q5
Q6 Q2
D1 D3 D5
D1 D6 D2
a b c
g1 g3 g5
g4 g6 g2
Fig. 1.2 Three phase converter
voltage. Harmonic neutralisation using magnetic coupling (multi-pulse converter configurations), harmonic
reduction using multi-level converter configurations and the pulse-width modulation (PWM).
1.1 Multi-pulse converter configuration
Multi-pulse operation is achieved, by connecting identical three-phase bridges, Fig. 1.2, to transformers
which have outputs that are phase-displaced with respect to one another. Star and delta-connected windings
have a relative 30o phase shift and a 6-pulse converter bridge connected to each transformer will give an
overall 12-pulse operation eliminating 5th and 7th harmonics. This principle can be extended to 24- and
48-pulse operation summing at the primary windings the transformed outputs of several 6-pulse converters
(4 for 24-pulse and 8 for 48-pulse operation). The harmonic cancellation is carried out into the transformer
secondary windings.
The basic issue in structuring a high-power, multi-pulse converter is the complexity of the magnetic
structure that is needed.
The converter operation is carried out applying low frequency (usually line frequency) firing pulse to
the power switches. Due to the low switching frequency, only about one third of the converter losses are due
to the switching losses, the remaining two thirds are due to the magnetic interface (conduction losses) [1].
A typical multi-pulse waveform is depicted in Fig.1.3.
4
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 14/120
0 0 . 0 0 5 0 . 0 1 0 . 0 1 5 0 . 0 2 0 . 0 2 5 0 . 0 3 0 . 0 3 5-4
-3
-2
-1
0
1
2
3
4
Tim e ( s )
V o l t s
Fig. 1.3 Multi-pulse staircase voltage waveform
1.2 Multi-level converter configuration
The multi-level inverters synthesize a staircase voltage wave, Fig.1.3, from several levels of DC voltage
sources, obtained from capacitor voltage source. As the number of levels increases, the synthesized staircase
wave approaches the sinusoidal wave resulting in reduced harmonic distortion. Fig.1.4 shows a single-phase
five level voltage source-inverter, this converter is more complex and requires the DC voltage source to be
split or centre-tapped in order to provide a zero voltage reference [8,9].
The fundamental magnitude and the harmonic spectrum are controlled varying the switching angles, α.
The fundamental voltage component can also be changed by keeping α constant and changing V DC .
Alternative forms of multi-level converters is the chain circuit [ 10,11], in which several converter
bridges, each with its own source capacitor, are connected in series as illustrated in Fig.1.5. This topology
is simpler than that presented in Fig.1.4 due to it does not include diodes-clamp and flying capacitors.
The main advantage of the multi-level inverter circuits is their ability to produce quasi-harmonic neu-
tralised output voltage waveforms without magnetic waveform summation circuits. This advantage is offset
by the complexity and size of the DC capacitor, and/or the need for additional power circuit components,
e.g. power diodes, and control functions, e.g. DC voltage equalisation.
5
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 15/120
2
VDC
2
VDC
2
VDC
2
VDC
2
VDC
2
VDC
2
VDC
2
VDC
Neutral
Vout
2
VDC
2
VDC
DCVDCV
α1
α2
Fig. 1.4 Five-level voltage source inverter
Vout
Neutral
Fig. 1.5 Single-phase three level Chain circuit
6
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 16/120
1.3 Pulse Width Modulation (PWM)
In multi-pulse and multi-level converters, there is only one turn-on, turn-off per device per cycle. An-
other approach is to have multiple pulses per half-cycle, and then vary the width of the pulses to vary the
amplitude of the AC voltage. The pulse width modulation (PWM) technique is commonly employed to
generate high quality output waveforms by relatively low power converter used in variable frequency AC
motor drives and distribution applications [12-14]. With this technique, the output of each converter pole is
switched several times during a fundamental cycle between the positive and negative terminals of the DC
source.
PWM requires a considerable increase in the number switch operations (high switching frequency);
thereby it generally increases the switching losses of the converter [1,5,6]. However, the always increasing
switching frequency of modern solid-state power switches could made possible the use of PWM in high
power applications [15, 16].
Among thevarious VSI topologies themulti-pulseconfigurations and the multi-level configurationhave
become popular for advanced Static VAR compensation applications. In these configurations the switching
frequency can be kept low in order to minimise device stresses switching losses and electromagnetic inter-
ference [8]. Some commercials STATCOM installed are: The STATCOM installed in Japan in 1991 uses
eight six-pulse VSI, each of 10 MVA rating, connected to a main transformer resulting in 48-pulse STAT-
COM. A 100 MVA 48-pulse STATCOM was installed in 1995 for the Tennessee Valley Authority (TVA) at
the Sullivan Substation in North-Eastern Tennessee [17]. Another application in high power system of the
multi-pulse VSI is in the 160 MVA UPFC installed at the Inez substation of the American Electric Power
(AEP) in Kentucky USA, this is based on two identical 48-pulse VSI.
1.4 Objetives
The STATCOM and other FACTS devices have been widely studied by analytical models but the phys-
ical functionality is unknown for a lot of power researches. The objective of this doctoral work is to present,
analyse and experimentally verify the operation of a multi-pulse-base on three-phase STATCOM.
To accomplish the objective, the following research tasks are performed:
• Present the detailed analysis of 6-, 12-, 24- and 48-pulse-VSI STATCOM
This part presents the relevant details of the voltage source-inverter (VSI), the building block of a STAT-
7
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 17/120
COM and other FACTS devices. A detailed analysis is carried out to deduce the expressions for the AC
(phase currents and output voltage) and DC signals (capacitor current and voltage). The analysis allows
to understand how the power exchanged is achieved using a DC capacitor as a DC source, the operating
modes, and the control voltage.• STATCOM modelling.
Based on a switching model, a state-space representation in the dq-reference frame is deduced. The
model is used to calculate the control parameters.
• STATCOM design
A 12-pulse-VSI STATCOM is designed and simulated using the EMTDC/PSCAD.
• STATCOM laboratory prototype.
Using the theoretical analysis and simulations an IGBT-12-pulse-VSI STATCOM will be developed to
experimentally verify the operation of a three-phase STATCOM.
8
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 18/120
References
[1] CIGRE, ”Static Synchronous Compensator”, working group 14.19, September 1998.
[2] Z. Yang, C. Shen, L. Zhang, M. L. Crow, ”Integration of a STATCOM and Battery Energy Storage”,
IEEE Trans. on Power System, Vol. 16, no. 2, May 2001, pp. 254-260.
[3] L. Chun, J. Qirong, X. Jianxin, ”Investigation of Voltage Regulation Stability of Static Synchronous
Compensator in power system”, IEEE Power Engineering Society, Proceedings of the Winter Meet-
ing 2000, IEEE Vol. 4, pp. 2642-2647.
[4] H. F. Wang, ”Applications of damping torque analysis to StatCom control”, Electrical Power and
Energy Systems, Vol. 22, 2000, pp. 197-204.
[5] Narain G. Hingorani, Laszlo Gyugyi, ”Understanding FACTS”, IEEE Press 2000.
[6] Yong Hua Song, Allan T. Johns, ”Flexible AC transmission systems FACTS”, IEE Power and Energy
Series 30, 1999.
[7] Zhiping Yang, ”Integration of battery energy storage with flexible AC transmission system devices”,
Ph. D. Thesis, University fo Missouri-Rolla, 2000.
[8] Ekanayake, J. B., Jenkins, N., ”A three-level advanced Static VAR Compensator”, IEEE Trans. on
Power Delivery, Vol. 11, no. 1, pp. 540-545.
[9] C. J. Hatziadoniu, F. E: Chalkiadakis, ”A 12-pulse Static Synchronous Compensator for the distribu-
tion system employing the 3-Level GTO-Inverter”, IEEE Trans. on Power Delivery, Vol. 12, no. 4,
October 1997, pp. 1830-1835.
[10] Krshnat V. Patil, ”Dynamic Compensationof Electrical Power SystemsUsing a New BVSI-STATCOM”,
Ph. D. Thesis, University of Wester Ontario, London Ontario, Canada, March 1999.
[11] B. Han, S. Back, H. Kim, G. Karady, ”Dynamic Characteristic Analysis of SSSC bases on multibrige
inverter”, IEEE Trans. on Power Delivery, Vol. 17, no. 2, April 2002, pp. 623-629.
[12] J. G. Kassakian, M. F. Schlecht, G. C. Verghese, ”Principles of power electronics”, Addison-Wesley,
1992.
[13] N. Mohan, T. M. Undeland, W. P. Robbins, ”Power Electronics: Converters, Applications, and
Desing”, John Wiley and Sons, 1995.
9
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 19/120
[14] Olimpo Anaya-Lara, E. Acha, ”Modeling and Analysisof CustomPowerSystems by PSCAD/EMTDC”,
IEEE Trans. on Power Delivery, Vol. 17, no.1, January 2002, pp. 266-272.
[15] Pablo García Gonzalez, Aurerio García Cerrada, ”Control System for a PWM-Based STATCOM”,
IEEE Trans. on Power Delivery, Vol. 15, no. 4, October 2002, pp. 1252-1257.
[16] G. Venkataramanan, B. K. Johnson, ”Pulse Width Modulated series compensator”, IEE Proc.- Gener.
Transm. Distrib., Vol 149, no. 1, January 2002, pp. 71-75.
[17] C. Schuder, M. Gernhardt, E. Stacey, T. Lemark, L. Gyugyi, T. W. Cese, A. Edris, ”Operation of
±100 MVAR TVA-STATCON”, IEEE Trans. on Power Delivery, Vol. 12, no. 4, October 1997.
10
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 20/120
Chapter 2
Analysis of a STATCOM based on 6- and 12-pulses
VSI
2.1 Introduction
This chapter presents the detailed analysis of a six and twelve-pulse VSI-STATCOM. The analysis
allows to understand how the power exchanged is achieved using a capacitor as a DC source, the operating
modes, the control voltage, and to give us an important insight into the comprehension of higher pulse
arrangements. To validate the accuracy of the equations encountered digital simulations will be presented.
The VSI is analysed as a linear network with a topology that changes depending on the state of the six
(ideal) switching devices. The analysis exploits the fact that the system is piecewise linear; consequently,
over each interval during which the switches do not change their state, the circuit equations may be solved
using standard linear techniques.
Fig. 2.1 Simple voltage sourced-inverter (two level-pole)
The voltage-sourced converter or inverter (VSC or VSI) is presently considered the best candidate for
the implementation of a high power STATCOM and other FACTS devices based on VSI such as the UPFC
and the SSSC, and all existing and planned new installations known are based on this approach [1]. A two-
11
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 21/120
Q1
Q4
Q3 Q5
Q6 Q2
D1 D3 D5
D1 D6 D2
R
R R
n
a b c
g1 g3 g5
g4 g6 g2
ia
i b
ic
VDC
Fig. 2.2 Six-pulse VSI with a resistive load
level pole is the simplest switching arrangement capable of producing AC output from a DC source in the
form of a simple square wave as it switches the direct voltage source on and off, as illustrated in Fig.2.1.
2.2 Voltage source-inverter (VSI)
The major aim of a VSI is to generate an AC voltage from a DC voltage, so that it is often referred
to as a DC-AC converter or inverter. It has to be able to generate a symmetric AC voltage with a desired
magnitude and frequency. The magnitude and frequency can be fixed or varied (according to the application,
e.g. for traction applications).
The three phase basic configuration is called six-pulse inverter consisting of six asymmetric turn off
devices such as a GTO or IGBT with reverse-parallel diodes connected as a 6-pulse Graetz bridge [2]. The
six-pulse VSI is shown in Fig.2.2.
The firing control signals gi applied to the transistors are controlled such that each one conducts over
180owhen the inverter is connected to a resistive load. Fig.2.3 shows the firing control signals for each
12
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 22/120
0
0
g1
0
0
g2
0
0
g3
0
0
g4
0
0
g5
0
0
g6
ωt
ωt
ωt
ωt
ωt
ωt
3
π
3
2π π π2 π3 π4 π5
Fig. 2.3 Firing control signals
transistor. The inverter can be seen as the combination of three single-phase inverters where each phase
leg produces an output phase shifted by ±120owith respect to the outputs of the other two legs. The firing
control signals are shifted by 60ofrom each other.
The switching sequence shown in Fig.2.3 generates the line-to-line voltages vab(t), vbc(t) and vca(t)
illustrated in Fig.2.4. In this figure it is noticed that these line-to-line voltages have 120o pulse width with
peak voltage magnitude of V DC . For each 60ointerval three different operating modes exists [3]; in the first
half cycle the operating sequence is 1-5-6 (Fig.2.5), 1-2-6 (Fig.2.6) and 1-2-3 (Fig.2.7). In order to obtain
the line-to-neutral voltages the corresponding different operating modes will be analysed.
Sequence 1-5-6; 0 ≤ ωt ≤ π/3
van = vcn =1
3V DC
vbn
=−
2
3V
DC
Sequence 1-2-6; π/3 ≤ ωt ≤ 2π/3
van =2
3V DC
13
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 23/120
0
0
0
0
0
0
ωt
ωt
ωt
vab
vbc
vca
vDC
-vDC
vDC
-vDC
-vDC
vDC
3
2π π π2 π3 π4 π5
Fig. 2.4 Line-to-line voltage waveform
R
R R
a
bc
n
R b
R
c
R
a
n
+
VDC
VDC
+
+
VDC
+V b n
-
+Vcn
-
+Van-
Fig. 2.5 Equivalent circuit for the sequence 1-5-6
14
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 24/120
R
R R
a
bc
n
R
R
c
R
n
b
a
+
VDC
+
VDC
+Van -
+Vcn
-
+V bn -
VDC
+
Fig. 2.6 Equivalent circuit for the sequence 1-2-6
vbn = vcn = −1
3V DC
Sequence 1-2-3; 2π/3 ≤ ωt ≤ π
van = vbn =1
3V DC
vcn = −2
3V DC
The following three operating modes are obtained in a similar way, and they are the negative of those
shown previously. Fig.2.8 depicts the line-to-neutral voltages van(t), vbn(t) and vcn(t). The synchronisation
and frequency of the generated voltages depend directly on the frequency and synchronisation of the firing
control signals and not on the load type. The peak magnitude voltage depends on the DC voltage.
2.2.1 Harmonic analysis
The harmonic content of the line-to-line and line-to-neutral voltages can be obtained applying a Fourier
analysis to the waveforms shown in Fig.2.4 and Fig.2.8. Half-odd wave symmetry is obtained if vab(t)
voltage is lagged 30o, therefore the instantaneous value of vab(t) based on Fourier analysis is given by
vab(t) =∞X
n=1
V abn sin³
nωt +π
6n´
(2.1)
15
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 25/120
R
R R
a
bc
n
a
R
R
R
n b
c
+
VDC
+
VDC
+
VDC
+Van -
+V b n
-
+Vcn -
Fig. 2.7 Equivalent circuit for the sequence 1-2-3
0
0
0
0
0
0
ωt
ωt
ωt
3
π
3
2π π π2 π3 π4 π5
van
vbn
vcn
DCV3
2
DCV3
2
DCV32
DCV31
DCV31
DCV31
DCV3
1−
DCV3
1−
DCV31−
DCV32−
DCV3
2−
DCV32−
0
0
0
0
0
0
ωt
ωt
ωt
3
π
3
2π π π2 π3 π4 π5
van
vbn
vcn
DCV3
2
DCV3
2
DCV32
DCV31
DCV31
DCV31
DCV3
1−
DCV3
1−
DCV31−
DCV32−
DCV3
2−
DCV32−
Fig. 2.8 Line-to-neutral voltage waveform
16
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 26/120
where:
V abn =4
T
5π/6
Z π/6
V DC sin(nωt) dωt (2.2)
T = 2π
so that
V abn =2
nπV DC
µcos
³π6
n´− cos
µ5π
6n
¶¶(2.3)
From eq. (2.3) it is worth to note that only the 6r ± 1 (n = 6r ± 1) terms exist, where r is any positive
integer, that is, n = 1th, 5th, 7th, 11th, 13th, ..., therefore eq. (2.3) can be reduced to
V abn =4
nπV DC cos³π
6n´∀n = 6r ± 1, r = 0, 1, 2,... (2.4)
Eqns. (2.5) and (2.6) give rise to the fundamental and harmonic components of the line-to-line voltages,
V ab1 = 1.1026V DC peak; V ab1 = 0.7797V DC RMS (2.5)
V abn =1.1026
nV DC peak; V abn =
0.7797
nV DC RMS (2.6)
Finally, voltage vab(t) is expressed by eq. (2.7). Voltages vbc(t) and vca(t) have a similar pattern except
phase shifted by 120oand 240o, respectively, from vab(t).
vab(t) = 4π
V DC
∞Xn
1n
cos³π
6n´
sin³
nωt + π
6n´
(2.7)
vbc(t) =4
πV DC
∞Xn
1
ncos
³π6
n´
sin³
nωt − π
2n´
(2.8)
vca(t) =4
πV DC
∞Xn
1
ncos
³π6
n´
sin
µnωt − 7π
6n
¶(2.9)
∀ n = 6r ± 1; r = 0, 1, 2,...
In a similar way the harmonic content of the line-to-neutral voltages van(t), vbn(t) and vcn(t) is ob-
tained; analysing van(t) a half-odd wave symmetry is found, thus
van(t) =∞X
n=1
V ann sin(nωt) (2.10)
17
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 27/120
where
V ann =4
3T V DC
π/3
Z 0
sin(nωt) dωt + 2
2π/3
Z π/3
sin(nωt) dωt +
π
Z 2π/3
sin(nωt) dωt
(2.11)
T = 2π
so that
V ann =2
3nπV DC
µcos
³π3
n´− cos
µ2π
3n
¶+ 1 − (−1)n
¶(2.12)
Although the waveforms of the line-to-line voltages and the line-to-neutral voltages are different, they
present a similar harmonic content, the line-to-neutral voltages also have 6r ± 1 (n = 6r ± 1) order terms,
where r is any positive integer, that is, n = 1th, 5th, 7th, 11th, 13th, ..., thereby eq. (2.12) is expressed by,
V ann =4
3nπV DC
³cos
³π3
n´
+ 1´∀ n = 6r ± 1, r = 0, 1, 2,... (2.13)
the peak and the RMS fundamental and harmonic components result:
V an1= 0.6366V DC peak; V an1
= 0.4502V DC RMS
V ann=
0.6366
nV DC peak; V ann =
0.4502
nV DC RMS
The following equations describe the line-to-neutral voltages based on Fourier analysis. Fig.2.9 depicts
the Fourier spectrum for voltage van(t) .
van(t) =4
3πV DC
∞Xn
1
n
³cos
³π3
n´
+ 1´
sin(nωt) (2.14)
vbn(t) =4
3πV DC
∞Xn
1
n
³cos
³π3
n´
+ 1´
sin
µnωt − 2π
3
¶(2.15)
van(t) =4
3πV DC
∞Xn
1
n
³cos
³π3
n´
+ 1´
sin
µnωt− 4π
3
¶(2.16)
∀ n = 6r ± 1; r = 0, 1, 2,...
It is worth noting that the fundamental component and the harmonic components of the line-to-line
voltages and the line-to-neutral voltages are phase shifted by 30ofrom each other. The amplitude of the
18
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 28/120
0 5 10 15 20 25 30 35 40 45 500
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
nth. harmonic
M a g n i t u d e ( p e r u n i t f u n d a m e n t a l )
Fig. 2.9 vab(t) voltage Fourier spectrum
line-to-line voltages is√
3 times the line-to-neutral voltage amplitude, and the harmonics components not
included in the set n = 12r ± 1 are in phase opposition. This is illustrated by the following expressions:
V ab1 =√
3V an1
V ab
5=−√
3V an
5
V ab7 = −√ 3V an7
V ab11 =√
3V an11
V ab13 =√
3V an13
that may be reduced to the following one:
V abn = (−1)r√
3V ann (2.17)
where: n = 6r ± 1 and r = 0, 1, 2,...
If the VSI of Fig.2.2 has a resistive load (unitary power factor) the diodes do not conduct at any
time, but if the load is inductive regardless of the power factor the conduction period of the transistors
would be between 90oand 180oand that of the diodes between 0oand 90o. The following figures illustrate
19
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 29/120
0 0 .01 0.02 0 .03 0 .04 0 .05 0.06 0 .07-80
-60
-40
-20
0
20
40
60
80
A m p
0 0 .01 0.02 0 .03 0 .04 0 .05 0.06 0 .07-10
0
10
20
30
40
50
60
70
Time (s)
A m p
iQ 1
iD 1
Ph ase a curr e nt , ia(t )
Transistor Q 1 and d iode D1 curren t; i Q1 ; iD1
Fig. 2.10 0.5 inductive power factor load conduction period
the conduction period for the first transistor and diode. A wye connected load with a 0.5 inductive power
factor was considered in Fig.2.10 (R = 1 Ω , L = 4.5944 mH); in this case the transistors conduct during
120oand the diodes for 60o. In Fig.2.11 a wye connected load with a 0.8660 inductive power factor was
considered (R = 1 Ω, L = 1.5315 mH); in this case the transistors conduct during 150oand the diodes for 30o.
Finally, Fig.2.12 depicts a pure inductive load (zero power factor); it shows that for a pure reactive load the
conduction period for the transistors and diodes is equal, that is, 90o.
2.3 Static synchronous compensator (STATCOM) based on six-pulse VSI
The static compensators are devices with the ability to both generate and absorb reactive and active
power, but the most common applications are in reactive power exchange between the AC system and the
compensator.
The static synchronous compensator (STATCOM) based on six-pulse VSI, Fig.2.13, is the basic build-
ing block of high power static VAR compensator. In high power applications the six-pulse configuration
does not have an appropriate performance, because of it exhibits a high harmonic rate. Consequently, the
20
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 30/120
0 0 .01 0.02 0 .03 0 .04 0 .05 0 .06 0 .07-150
-100
-5 0
0
50
100
150
A m p
0 0 .01 0.02 0 .03 0 .04 0 .05 0 .06 0 .07-2 0
0
20
40
60
80
100
120
Ti me (s)
A m p
iQ 1
iD 1
Phase a current; ia(t)
Transistor Q 1 and diode D1 current; iQ1, iD1
Fig. 2.11 0.8660 inductive power factor load conduction period
0 0.01 0.02 0 .03 0.04 0.05 0.06-1
0
1
2
3
4
Time (s)
A m p
iQ1
i D1
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07-4
-3
-2
-1
0
1
2
3
4
A m p
Phase a current; i a(t)
Transistor Q1 and diode D1 current; iQ1, iD1
Fig. 2.12 Inductive load conduction period
21
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 31/120
Q1
Q4
Q3 Q5
Q6 Q2
D1 D3 D5
D1 D6 D2
a b c
g1 g3 g5
g4 g6 g2
L L L
C
ICD
vAN
(t) vBN
(t) vCN
(t)
VCD
+
ia(t) i b(t) ic(t)
Fig. 2.13 Six-pulse VSI-STATCOM
current analysis is aimed to give us an important insight into the analysis of more complicated higher pulse
arrangements.
2.3.1 Reactive power exchange
The reactive power exchange between the AC system and the compensator is controlled by varying the
magnitude of the fundamental component of the inverter voltage above and below that of the AC system.
The compensator control is achieved by small variations in the switching angle of the semiconductor
devices, so that the fundamental component of the voltage produced by the inverter is forced to lag or
lead the AC system voltage by a few degrees. This causes active power to flow into or out of the inverter
22
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 32/120
modifying the value of the DC capacitor voltage, and consequently the magnitude of the inverter terminal
voltage and the resultant reactive power. If the compensator supplies only reactive power, the active power
provided by the DC capacitor is zero. Therefore, the capacitor does not change its voltage. One could saythen that the capacitor plays not any role in the reactive power generation [1].
2.3.1.1 Analysis of the AC current signals
The current flowing between the compensator and the AC system is determined by the voltage across
the tie inductor. Let the AC voltage be a pure sinusoidal function ean(t) = V m sin(ωt), then the magnitude
of the fundamental and harmonic components are given by eqns. (2.18) and (2.19). From eq. (2.13)
van(t) = 0.6366V DC sin(ωt), then,
ia(t)1 = −V m − 0.6366V DC ωL
cos(ωt) ; (fundamental component)
ia(t)n =0.6366V DC
n2ωLcos(nωt) ; (nth harmonic component)
the fundamental current is also called fundamental reactive current I q.
I a1 = I q =V m − 0.6366V DC
ωL(2.18)
I an = I q n =0.6366V DC
n2ωL, n > 1 (2.19)
The fundamental current will be leading when V m < 0.6366V DC ; that is, if the amplitude of the inverter voltage is increased above that of the AC system (the current flows from the converter to the AC system); in
this case the compensator is seen as a capacitor by the AC system. The fundamental current will be lagging
when V m > 0.6366V DC , that is, if the amplitude of the inverter voltage is decreased below that of the AC
system (the current flows from the AC system to the compensator); in this case the compensator is seen as a
inductor by the AC system, Fig.2.14.
From eq. (2.19) can be deduced that the harmonic currents only flow from the compensator to the
AC system. Fig.2.15 depicts the relationship between the DC voltage (V DC ) and the fundamental reactivecurrent. The lower-order harmonic currents as a function of the fundamental reactive current are shown too.
Taking into account the voltage across the tie inductor over each 60oconduction interval, Fig.2.16, the
equation that describes the AC current ia(t), is derived. The other two-phase currents, ib(t) and ic(t), will
23
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 33/120
Lagging
Vm > 0.6366VDC
Leading
Vm < 0.6366VDC
VL
Ia1
Ia1
Ian
Lagging
Vm > 0.6366VDC
Leading
Vm < 0.6366VDC
VL
Ia1
Ia1
Ian
Fig. 2.14 Phase relationship between the inductor voltage V L and the fundamental curren I a1
-1 .5 -1 - 0.5 0 0 .5 1 1 .50
0.5
1
1.5
2
Iq
D C
v o
l t a g e
( p . u
)
-1 .5 -1 - 0.5 0 0 .5 1 1 .50
0.0 2
0.0 4
0.0 6
0.0 8
0.1
0.1 2
0.1 4
Iq
I q n
( n
t h h a r m o n
i c m a g n
i t u
d e
)
Iq 5
Iq 7
Iq 1 1
Leading (capaciti ve region)Lagging (induc tive region )
Lag ging (ind uctive regi on ) Lea ding (cap acitive region )
Fig. 2.15 Relationship between DC voltage, fundamental and harmonics current
24
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 34/120
0
0
ω t
3
π
3
2 π π
3
4 π
3
5 π π2
mV
mV
D CV3
2
D CV
3
2−
D CV3
1
D CV3
1−
Fig. 2.16 AC system voltage and voltage compensator van(t)
be identical except phase shifted by 120oand 240o, respectively, from ia(t).
At any time vL(t) = Ld
dtia(t), where vL(t) is the instantaneous inductor voltage, which is the instan-
taneous difference across the AC system voltage, ean(t), and the compensator voltage van(t). The following
differential equations can be written for each 60oconduction interval.
• Interval: 0 ≤ ωt ≤ π/3
vL(t) = V m sin(ωt) − 13
V DC = L ddt
ia(t)
ia(t) = −V mωL
(cos(ωt) − 1)− 1
3LV DC t + I 0 (2.20)
where I 0 is the initial condition at t = 0; ia(0) = I 0.
• Interval: π/3 ≤ ωt ≤ 2π/3
vL(t) = V m sin(ωt) − 2
3V DC = L
d
dtia(t)
ia(t) = −V mωL
(cos(ωt) − 0.5)− µ 23L
t − 2π9ωL
¶V DC + I 1 (2.21)
where I 1 = ia
³ π
3ω
´• Interval: 2π/3 ≤ ωt ≤ π
25
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 35/120
vL(t) = V m sin(ωt) − 1
3V DC = L
d
dtia(t)
ia(t) = −V m
ωL (cos(ωt) − 0.5)− µ1
3Lt −2π
9ωL¶
V DC + I 2 (2.22)
where I 2 = ia
µ2π
3ω
¶It is known that in steady state the AC current waveform is symmetric, therefore,
ia
³ π
2ω
´= 0
taking into account the above property, the steady state initial condition I 0 can be calculated as follows:
ia ³ π
2ω´ =
−V m
ωL³cos³π
2´− 0.5´−µ
2
3L·π
2ω −2π
9ωL¶V DC + I 1 = 0
I 1 = −V mωL
0.5 +π
9ωLV DC (2.23)
I 1 = ia
³ π
3ω
´= −V m
ωL
³cos
³π3
´− 1
´− 1
3LV DC ·
π
3ω+ I 0
I 1 =V mωL
0.5 − π
9ωLV DC + I 0 (2.24)
equating eqns. (2.23) and (2.24), the steady state condition is estimated:
I 0 = −V mωL +
2π
9ωL V DC (2.25)
substituting eqn. (2.25) into eqn. (2.20) the steady state equations are derived,
ia(t) = −V mωL
cos(ωt) −µ
1
3Lt − 2π
9ωL
¶V DC (2.26)
0 ≤ ωt ≤ π/3
ia(t) = −V mωL
cos(ωt) −µ
2
3Lt − π
3ωL
¶V DC (2.27)
π/3 ≤ ωt ≤ 2π/3
ia(t) = −V mωL
cos(ωt) −µ
1
3Lt − π
9ωL
¶V DC (2.28)
2π/3 ≤ ωt ≤ π
26
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 36/120
0 0 .0 1 0 .0 2 0 .03 0 .0 4 0 .0 5 0 .0 6-2
-1
0
1
2Ph ase a: i
a(t)
A m p
0 0 .0 1 0 .0 2 0 .03 0 .0 4 0 .0 5 0 .0 6-2
-1
0
1
2 Ph ase b: ib(t )
A m p
0 0 .0 1 0 .0 2 0 .03 0 .0 4 0 .0 5 0 .0 6-2
-1
0
1
2 Ph ase c : ic(t )
T i m e ( s )
A m p
L eading Lagging
Fig. 2.17 AC current waveform
Over the interval π ≤ ωt < 2π the AC current waveform is the negative respect to that described in the
above equations. Fig.2.17 depicts the AC current waveform. For the leading case V DC = 6 V, V m = 2.5 V,
and L = 3 mH were used; and V DC = 6 V, V m = 4.5 V, and L = 3 mH for the lagging case.
2.3.1.2 Conduction period transistors and diodes
With the precedent AC current equations it is possible to predict the conduction period of each transistor
and diode. When only reactive power is generated (zero power factor) the transistors and diodes of the
inverter circuit conduct for 90o.
If the compensator is absorbing reactive power, the transistors naturally turn off at zero current. When
it is generating reactive power, the transistors turn off at the peak of the AC current waveform. The following
algorithm proposes a way to determining the conduction period of each device at any power factor based on
the AC current waveforms and the firing control signals.
• Leg 1
Pulse g1 on:
– If ia(t) is positive, D1 conducts.
27
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 37/120
– If ia(t) is negative, Q1 conducts.
Pulse g4 on:
– If ia(t) is positive, Q4 conducts.
– If ia(t) is negative, D4 conducts.
• Leg 2
Pulse g3 on:
– If ib(t) is positive, D3 conducts.
– If ib(t) is negative, Q3 conducts.
Pulse g6 on:
– If ib(t) is positive, Q6 conducts.
– If ib(t) is negative, D6 conducts.
• Leg 3
Pulse g5 on:
– If ic(t) is positive, D5 conducts.
– If ic(t) is negative, Q5 conducts.
Pulse g2 on:
– If ic(t) is positive, Q2 conducts.
– If ic(t) is negative, D2 conducts.
The Q1 and D1 conduction period for leading and lagging current are presented in Fig.2.18.
2.3.1.3 Capacitor current
The capacitor current is made up of segments of the three AC phase currents and is dependent on which
semiconductor devices are conducting over each 60ointerval. The capacitor current can be deduced in the
following way:
iD1(t) + iD3
(t) + iD5(t) + iDC (t) = iQ1
(t) + iQ3(t) + iQ5
(t)
so that
iDC (t) = iQ1(t) + iQ3
(t) + iQ5(t) − (iD1
(t) + iD3(t) + iD5
(t)) (2.29)
28
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 38/120
0 0 .0 1 0 .0 2 0 .0 3 0 .0 4 0 .0 5 0 .0 60
0.5
1
1.5
Q1
cu rrent : iQ1
(t)
A m p
Leading
Lagging
0 0 .0 1 0 .0 2 0 .0 3 0 .0 4 0 .0 5 0 .0 60
0.5
1
1.5
T ime (s)
A m p
D1
current: iD1
(t)
Fig. 2.18 Q1 and D1 conduction period
• Interval: 0 ≤ ωt ≤ π/3
iDC (t) = ia(t) + ic(t)
where
ic(t) =
−V m
ωL
cosµωt +2π
3¶−µ
1
3L
t +π
9ωL¶V DC
so that
iDC (t) =V mωL
sin³ωt− π
6
´−µ
2
3Lt − π
9ωL
¶V DC (2.30)
• Interval: π/3 ≤ ωt ≤ 2π/3
iDC (t) = ia(t)
iDC (t) =V mωL
sin³ωt− π
2
´−µ
2
3Lt − π
3ωL
¶V DC (2.31)
• Interval: 2π/3
≤ωt
≤π
iDC (t) = ia(t) + ib(t)
where
ib(t) = −V mωL
cos
µωt − 2π
3
¶−µ
1
3Lt − 4π
9ωL
¶V DC
29
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 39/120
so that
iDC (t) =V mωL
sinµωt − 5π
6 ¶−µ2
3Lt− 5π
9ωL¶V DC (2.32)
These expressions are equivalent except phase shifted by 60oamong them; therefore the capacitor cur-
rent waveform over each of the remaining three conduction periods is identical to that described by eq.
(2.30) and yields, in a repetitive waveform, six times the corresponding AC power system frequency.
2.3.1.4 DC capacitor voltage
The previous analysis assumes a constant DC voltage; it is equivalent to consider an infinite capacitor
and consequently zero DC ripple voltage. If a finite capacitor is considerated a DC ripple voltage exist which
is dependent on the capacitor value and the capacitor current.
Assuming that the capacitor current remains significantly unchanged from that given by eqn. (2.30),
the capacitor voltage can be estimated; under this condition a minimum DC ripple voltage is obtained [ 4].
The capacitor voltage over the first 60o period is given by:
vcap(t) =1
C
tZ 0
iDC (t)dt + V 0 (2.33)
where V 0 is the initial condition at t = 0; V 0 = vcap(0). Substituting eqn. (2.30) into eqn. (2.33)
vcap(t) = − V mω2LC
cos³ωt − π
6
´− 13LC
V DC t2 + π9ωLC
V DC t + √ 3V m2ω2LC
+ V 0 (2.34)
The value of V 0 is calculated from the average component of eq. (2.34) with a period T = π/3ω . At
the same time the DC voltage level V DC is determined.
V DC =1
T
π/3ωZ 0
vcap(t)dt (2.35)
V DC =3ω
πÃ−
V m
ω3
LC −π3
243ω3
LC V DC +
π3
162ω3
LC +
√ 3π
6ω3
LC V m +
π
3ωV 0!
Simplifying this expression results:
V 0 = 0.0889V m
ω2LC − 0.0609
1
ω2LC V DC + V DC (2.36)
30
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 40/120
0
-1
-0.5
0
0.5
1
(a)
A m p ( p u )
0
-1
-0.5
0
0.5
1
(b)
A m p ( p u )
ωt
ωt
0
-1
-0.5
0
0.5
1
(a)
A m p ( p u )
0
-1
-0.5
0
0.5
1
(b)
A m p ( p u )
ωt
ωt
Fig. 2.19 Capacitor current; a)generating reactive power; b)absorbing reactive power
The current and voltage waveforms of the capacitor obtained using the expressions (2.30) and (2.34)
are shown in Fig.2.19 and Fig.2.20. Fig.2.20 illustrates that the peak capacitor voltage V pk, occurs when the
compensator is generating reactive power (leading current), ωt = 30o.
V pk = vcap³ π
6ω´
= −V m
ω2LC −1
3LC V DC ³ π
6ω´2
+
π
9ωLC V DC ³ π
6ω´
+
√ 3
2ω2LC V m
+0.0889V m
ω2LC − 0.0609
1
ω2LC V DC + V DC
Simplifying, gives rises to:
V pk = −0.0451V m
ω2LC + 0.0305
1
ω2LC V DC + V DC (2.37)
The above peak voltage is important because of the capacitor voltage is applied directly to the semi-
conductor devices, therefore those ones must be able to support that voltage.
Fig.2.21 shows the line-to-neutral voltage, van(t), and the line-to-line voltage, vab(t), when the com- pensator is operated with a finite capacitor as a DC source. This figure illustrates the capacitor ripple voltage
effect. The capacitor voltage and current waveforms are presented in Fig.2.22. These results were obtained
using the PSCAD\EMTDC software with the following parameters: C = 500 µF, L = 3 mH, V m = 2.5 V,
31
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 41/120
00.85
0.9
0.95
1
1.05
1.1
(a)
V o l t s ( p u )
00.98
0.99
1
1.01
1.02
1.03
1.04
(b)
V o l t s ( p u )
ωt
ωt
00.85
0.9
0.95
1
1.05
1.1
(a)
V o l t s ( p u )
00.98
0.99
1
1.01
1.02
1.03
1.04
(b)
V o l t s ( p u )
ωt
ωt
Fig. 2.20 DC capacitor voltage; a)generating reactive power; b)absorbing reactive power
and V DC = 6 V.
If the compensator only exchange reactive power the voltage capacitor does not vary, thus the capacitor
current is the one shown in Fig.2.19. Based on Fourier analysis it is given by:
iDC (t) =∞X
n=1
I DC n sin(nωt) (2.38)
where:
I DC n =2
T
T R 0
iDC (t)sin(nωt) dt
T =π
3ωThereby the capacitor voltage is made up only by sinusoidal functions with fixed amplitude and a DC
offset, Fig.2.20.
2.3.2 Reactive and active power exchange
Considering a phase shift φ across the AC power system and the fundamental compensator voltage,
Fig.2.23, both reactive and active power is exchanged. To obtain the AC current equations a similar proce-
32
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 42/120
0 0 .00 5 0 .0 1 0 .01 5 0 .0 2 0 .0 25 0 .03 0 .0 35 0 .0 4 0 .0 4 5 0 .0 5-5
0
5
V o
l t s
0 0 .00 5 0 .0 1 0 .01 5 0 .0 2 0 .0 25 0 .03 0 .0 35 0 .0 4 0 .0 4 5 0 .0 5-8
-6
-4
-2
0
2
4
6
8
Ti me (s)
V o
l t s
P hase to neu tra l vo l tage; v an(t )
Ph ase to p hase vo l tage; v ab(t )
Fig. 2.21 Voltage waveforms with a finite DC capacitor
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
A m
p
current waveform
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.050
1
2
3
4
5
6
7
Time (s)
V o l t s
DC c apacitor voltage
Fig. 2.22 Capacitor current and DC capacitor voltage waveform
33
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 43/120
0
0
D Cv3
2
mv
D Cv
3
1
D Cv
3
1−
DCv3
2−
mv
3
π
3
2π π
3
4 π
3
5 π π2t
Fig. 2.23 AC system voltage and compensator voltage; van(t)
dure to that presented in the section 2.3.1.1 is carried out.
• Interval: 0 ≤ ωt ≤ π/3
vL(t) = V m sin(ωt − φ) − 1
3V DC = L
d
dtia(t)
ia(t) = −V mωL
(cos(ωt − φ) − cos(φ)) − 1
3LV DC t + I 0 (2.39)
where I 0 is the initial condition at t = 0; ia(0) = I 0.
• Interval: π/3 ≤ ωt ≤ 2π/3
vL(t) = V m sin(ωt − φ) − 2
3V DC = L
d
dtia(t)
ia(t) = −V mωL
³cos(ωt − φ) − cos
³π3− φ
´´−µ
2
3Lt− 2π
9ωL
¶V DC + I 1 (2.40)
where: I 1 = ia
³ π
3ω
´• Interval: 2π/3
≤ωt≤π
vL(t) = V m sin(ωt − φ) − 1
3V DC = L
d
dtia(t)
ia(t) = −V mωL
µcos(ωt − φ) − cos
µ2π
3− φ
¶¶−µ
1
3Lt − 2π
9ωL
¶V DC + I 2 (2.41)
34
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 44/120
where: I 2 = ia
µ2π
3ω
¶
Taking into account that ia(0) = −ia³πω´
, the steady state initial condition I 0 can be calculated,
I 0 = −ia
³πω
´=
V mωL
µcos(π − φ) − cos
µ2π
3− φ
¶¶+
µ1
3L·π
ω− 2π
9ωL
¶V DC − I 2
so that
I 0 =V mωL
µcos(π − φ) − cos
µ2π
3− φ
¶¶+
π
9ωLV DC − I 2 (2.42)
where:
I 2 = −V mωL
µcos
µ2π
3− φ
¶− cos
³π3− φ
´¶− 2π
9ωLV DC + I 1 (2.43)
I 1 = ia³ π
3ω´
= −V
mωL
³cos
³π3 − φ
´− cos(φ)´− π
9ωL V DC + I 0 (2.44)
Substituting eqns. (2.43) and (2.44) into eqn. (2.42) yields
I 0 = −V mωL
cos(φ) +2π
9ωLV DC (2.45)
using eqn. (2.45) the steady state equations are:
ia(t) = −V mωL
cos(ωt − φ) −µ
1
3Lt − 2π
9ωL
¶V DC (2.46)
0 ≤ ωt ≤ π/3
ia(t) = −V mωL
cos(ωt− φ) −µ
2
3Lt − π
ωL
¶V DC (2.47)
π/3 ≤ ωt ≤ 2π/3
ia(t) = −V mωL
cos(ωt − φ) −µ
1
3Lt − π
9ωL
¶V DC (2.48)
2π/3 ≤ ωt ≤ π
Over the interval π ≤ ωt < 2π the AC current waveform is the negative of that described in the aboveequations. As the phase shift φ increases the AC current changes regarding to the one shown in Fig.2.17.
Fig.2.24 illustrates the AC current with φ = 15o.
35
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 45/120
0 0.005 0.01 0.015 0.02 0.0 25 0.03 0 .035 0.04 0 .045 0.05-2
-1
0
1
2
A m p
0 0.005 0.01 0.015 0.02 0.0 25 0.03 0 .035 0.04 0 .045 0.05-2
-1
0
1
2
A m p
0 0.005 0.01 0.015 0.02 0.0 25 0.03 0 .035 0.04 0 .045 0.05-2
-1
0
1
2
Time (s)
A m p
Phase a: ia(t)
Phase b: ib(t)
Phase c: ic(t)
Fig. 2.24 AC current waveform with φ = 15o
2.3.2.1 Capacitor current
The capacitor current is obtained in a similar manner to that presented previously and has a similar
behaviour, a waveform made up by six segments where each segment can be represented by eq. (2.49) with
a phase shifted by 60ofrom each other.
iDC (t) =V mωL
sin³ωt − φ− π
6
´−µ
2
3Lt − π
9ωL
¶V DC (2.49)
0 ≤ ωt ≤ π/3
It is noteworthy that if the angle φ increases the capacitor current would have a greater DC level, this
effect is illustrated in Fig.2.25 where iDC (t) is presented taking into account φ = 2o and φ = 15o.
If an angle φ exists between the AC power system voltage and the fundamental compensator voltage
the instantaneous capacitor voltage and current, based on Fourier analysis are given by:
iDC (t) = I DC 0 +∞P
n=1I DC n sin(nωt)
vcap(t) = V DC 0 +∞P
n=1V DC n cos(nωt)
36
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 46/120
0 0.002 0.004 0.006 0.008 0.01 0 .012 0 .014 0.0 16 0.018-0.5
0
0.5
1
1.5
Time (s)
A m p
Instantaneous DC current with φ = 15º
0 0.002 0.004 0.006 0.008 0.01 0 .012 0 .014 0.0 16 0.018-1
-0.5
0
0.5
1
A m p
Instantaneous DC curren t with φ = 2º
Fig. 2.25 Instantaneous capacitor current
The above expressions show that a DC power or active power flows through the DC side, thus the
capacitor voltage will be increased or decreased, depending if the I DC 0 is positive or negative. To reveal this
effect consider the circuit shown in Fig.2.26.
The circuit is initially at steady state with φ = 0o, at time t = 0.0333 s the angle φ was stepped from
0oto -0.5o. Fig.2.27 shows the capacitor voltage. This figure illustrates that the phase shift φ affect the
capacitor voltage, due to the DC component into iDC (t), that is, the active power exchange, and that the
mechanism of phase angle adjustment is used to control the VAR generation or absorption by increasing or
decreasing the capacitor voltage, and thereby the amplitude of the output voltage produced by the converter.
The steady state DC component of the capacitor current is given by:
I DC 0 =1
T
T Z 0
iDC (t)dt
where iDC (t) is given by eq. (2.49) and T = π/3ω, so that
I DC 0 = − 3V mπωL
sin(φ) (2.50)
37
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 47/120
IC(t) 500µF
Fig. 2.26 Test circuit
If the converter only supplies reactive power the AC power system voltage and the fundamental con-
verter voltage are in phase, if the converter is ideal (without losses). In practical converters, however the
semiconductor switches are not lossless, and therefore the energy stored in the DC capacitor would be used
up by the internal losses. However, these losses can be supplied from the AC system by making the output
voltages of the converter lag the AC system voltages by a small angle; this allows an active power flow from
the AC system to the converter, which compensates the converter losses and keeps the capacitor voltage at
the desired level.
The above concept is illustrated in Fig.2.28. The system is operating in steady state with a fixed DC
source at t = 0.1 s, the DC source is changed by a DC capacitor. In Fig.2.28(a) both the AC system voltage
and the compensator voltage are in phase, in this case the energy stored in the DC capacitor is used up bythe internal losses. In Fig.2.28(b) with a small angle between the AC system voltage and the compensator
voltage the losses are supplied from the AC system and the capacitor voltage is kept at the desired level.
Fig.2.28 was obtained using PSCAD/EMTDC.
2.4 12-pulse converter
The six-pulse STATCOM is the simplest arrangement used in this kind of devices; in high power ap-
plications it does not offer a good performance, due to the high harmonic content. Combining two six-pulse
converters, a better performance is obtained. This new configuration is called twelve-pulse STATCOM. The
twelve-pulse circuit is the lowest practical pulse-numbered circuit for power system application to achieve
a satisfactory harmonic behaviour [4].
38
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 48/120
0 0.02 0.04 0.06 0.08 0.1 0.125
5.5
6
6.5
7
7.5
8
8.5
9
9.5
Time (s)
V o l t s
Fig. 2.27 Instantaneous capacitor voltage
0 0.05 0.1 0.15 0.2 0.25 0.3
-1.5
-1
-0.5
0
0.5
1
1.5
(a)
A m p .
0 0.05 0.1 0.15 0.2 0.25 0.3
-1.5
-1
-0.5
0
0.5
1
1.5
(b)
Time (s)
A m p .
Fig. 2.28 AC instantaneous current with: a)φ = 0o; b)φ = −0.5o
39
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 49/120
The six-pulse converter output are the line-to-line voltages, vab(t), vbc(t), and vca(t). vab(t) is given by
eq. (2.1), expressing it by its Fourier series results:
vab(t) = V ab1 sin(ωt + 30o) + V ab5 sin(5ωt + 150o) + V ab7 sin(7ωt + 210o)
+V ab11 sin (11ωt + 330o) + V ab13 sin (13ωt + 30o) + V ab17 sin (17ωt + 150o)
+V ab19 sin (19ωt + 210o) + V ab23 sin (23ωt + 330o) + ... (2.51)
If this compensator is connected to a Y −Y transformer with a 1:1 turn ratio, the line-to-neutral voltage,
van(t) is given by,
van(t) =1√
3(V ab1 sin(ωt) − V ab5 sin(5ωt) − V ab7 sin (7ωt) + V ab11 sin (11ωt)
+V ab13 sin (13ωt) − V ab17 sin (17ωt) − V ab19 sin(19ωt) + V ab23 sin (23ωt) + ...) (2.52)
van(t) =1
3
∞Xn=1
V abn
(−1)r sin(nωt) (2.53)
∀ n = 6r ± 1, r = 0, 1, 2,...
From eqns. (2.51) and (2.52) could be noted that the line-to-line voltage amplitudes are√
3 times
the line-to-neutral voltage amplitudes and the harmonic components not included in the set n = 12r ± 1,
where r = 0, 1, 2,..., are in phase opposition. This feature is useful to cancel the harmonic components not
included in the set n = 12r ± 1.
Suppose that a second six-pulse converter produces line-to-line voltages lagged 30owith respect to the
other converter and with the same magnitude. That is,
vab(t)2 = V ab1 sin(ωt) + V ab5 sin (5ωt) + V ab7 sin(7ωt) + V ab11 sin (11ωt)
+V ab13 sin (13ωt) + V ab17 sin (17ωt) + V ab19 sin (19ωt) + V ab23 sin (23ωt) + ... (2.54)
vab(t)2 =∞
Xn=1V abn sin(nωt) (2.55)
If the second converter is connected to a ∆ − Y transformer with a 1 : 1/√
3 turn ratio, the line-to-
neutral voltage in the Y -connected secondary would be:
vanY (t)2 =1√
3(V ab1 sin(ωt) + V ab5 sin(5ωt) + V ab7 sin(7ωt) + V ab11 sin (11ωt)
40
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 50/120
+V ab13 sin (13ωt) + V ab17 sin (17ωt) + V ab19 sin (19ωt) + V ab23 sin (23ωt) + ...)(2.56)
vanY (t)2 =
1
√ 3∞X
n=1V abn sin(nωt) (2.57)
∀ n = 6r ± 1, r = 0, 1, 2,...
then the line-to-line voltage in the wye-side is,
vabY (t)2 = V ab1 sin(ωt + 30o) − V ab5 sin(5ωt + 150o) − V ab7 sin(7ωt + 210o)
+V ab11 sin (11ωt + 330o) + V ab13 sin (13ωt + 30o) − V ab17 sin(17ωt + 150o)
−V ab19 sin (19ωt + 210o) + V ab23 sin (23ωt + 330o) + ... (2.58)
vabY (t)2 =∞X
n=1
V abn
(−1)r sin³
nωt +π
6n´
(2.59)
∀ n = 6r ± 1, r = 0, 1, 2,...
that can be expressed in other form,
vabY (t)2 =√
3∞X
n=1
V ann sin³
nωt +π
6n´
(2.60)
∀n = 6r ± 1, r = 0, 1, 2,...
The two waveforms given by eqns. (2.1) and (2.59) are added using a summing transformer to give a
third waveform vab(t)12 closer to being a sine wave; this voltage is named twelve-pulse voltage,
vab(t)12 = vab(t) + vabY (t)2 (2.61)
vab(t)12 = 2(V ab1 sin(ωt + 30o) + V ab11 sin (11ωt + 330o) + V ab13 sin (13ωt + 30o)
+V ab23 sin (23ωt + 330o) + ...) (2.62)
thus, vab(t)12 is the line-to-line voltage of a twelve-pulse converter. These waveforms are shown in Fig.2.29.
In the arrangement of Fig.2.30, the two six-pulse converters are connected in parallel on the same DC bus,
working together as a twelve-pulse VSI-STATCOM.
41
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 51/120
(a)
(b)Time (s)
DCV5774.0
DCV1547.1
DCV
DCV5774.0−
DCV1547.1−DCV−
6
π
2
π
3
2π π π2 π3 π4 π5tω
6
π
2
π
3
2π π π2 π3 π4 π5tω
0
DCV5774.0
DCV5774.0−
DCV5774.1
DCV5774.1−
DCV1547.2
DCV1547.2−
0
)t(vab
2abY )t(v
(a)
(b)Time (s)
DCV5774.0
DCV1547.1
DCV
DCV5774.0−
DCV1547.1−DCV−
6
π
2
π
3
2π π π2 π3 π4 π5tω
6
π
2
π
3
2π π π2 π3 π4 π5tω
0
DCV5774.0
DCV5774.0−
DCV5774.1
DCV5774.1−
DCV1547.2
DCV1547.2−
0
)t(vab
2abY )t(v
Fig. 2.29 a)vab(t) and vabY (t)2; b)12-pulse voltage
six-pulse
converter
six-pulse
converter
vAN
(t ) vBN
(t ) vCN
(t)
ia(t)
i b
(t)
ic(t)
iC D12(t)
C VDC
+
Y-Y
Y- ∆
ia1 (t)
ia2
(t)
Fig. 2.30 12-pulse VSI-STATCOM
42
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 52/120
Van
(t)
Vbn
(t)
Vcn
(t)
DCV3
1
DCV3
1
DCV3
1
DCV3
1
−
DCV3
1−
DCV31−
DCV9107.0
DCV9107.0
DCV9107.0
DCV9107.0−
DCV9107.0−
DCV9107.0−
DCV2440.1
DCV2440.1
DCV2440.1
DCV2440.1−
DCV2440.1−
DCV2440.1−
0
0
0
2
π π π2 π3 π4 π5tω
Van
(t)
Vbn
(t)
Vcn
(t)
DCV3
1
DCV3
1
DCV3
1
DCV3
1
−
DCV3
1−
DCV31−
DCV9107.0
DCV9107.0
DCV9107.0
DCV9107.0−
DCV9107.0−
DCV9107.0−
DCV2440.1
DCV2440.1
DCV2440.1
DCV2440.1−
DCV2440.1−
DCV2440.1−
0
0
0
2
π π π2 π3 π4 π5tω
Fig. 2.31 12-pulse STATCOM line-to-neutral voltages
The twelve-pulse voltage given by eq. (2.61) expressed as a Fourier series is given by:
vab(t)12 =∞X
n=1
V ab12n sin³
nωt +π
6n´
(2.63)
∀ n = 12r ± 1, r = 0, 1, 2,...
where:
V ab12n = V abn +√
3V ann
V ab12n =√
34
nπV DC , ∀ n = 12r ± 1, r = 0, 1, 2,... (2.64)
The line-to-neutral voltages are shown in Fig.2.31.
From Fig.2.32 can be observed that the twelve-pulse voltage vab(t)12, holds only harmonics of order
n = 12r ± 1, where r is any positive integer, that is, n = 1th, 11th, 13th, 23th, 25th..., with amplitudes
1/11th, 1/13th, 1/23th, 1/25th..., respectively.
43
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 53/120
2.4.1 AC current signals
Applying a similar procedure to that presented in the previous section, the AC current analysis is
carried out. Let the AC voltage be a pure sinusoidal function, ean(t) = V m sin(ωt) , then the magnitude
of the fundamental and harmonics components are given by eq. (2.65) and eq. (2.66), where van1(t) =
1.2732V DC sin(ωt).
ia(t)1 = −V m − 1.2732V DC
ωLcos(ωt) ; (fundamental component)
ia(t)n =1.2732V DC
n2ωLcos(nωt) , ∀ n > 1; (nth harmonic component)
so that,
I a1 = I q =V m − 1.2732V DC
ωL(2.65)
I an = I q n =1.2732V DC
n2ωL, ∀ n > 1 (2.66)
From now on, voltages vab(t), vbc(t) and vca(t) will be referred as the twelve-pulse STATCOM line-
to-line voltages and the voltages van(t), vbn(t) and vcn(t) as the twelve-pulse STATCOM line-to-neutral
voltages.
The fundamental current will be leading when V m < 1.2732V DC ; thus, the compensator is seen as a
capacitor by the AC system and the current flows from the compensator to the AC system; the fundamental
current will be lagging when V m > 1.2732V DC , thus the compensator behaves as an inductor by the AC
system and thecurrent flows from theAC systemto the compensator. Fig.2.33 exhibits the phasorial diagram
across the tie inductor and the AC current.
Fig.2.34 depicts the relationship between the DC voltage (V DC ) and the fundamental reactive current.
The lower-order harmonic currents as a function of the fundamental reactive current are displayed too.
To obtain the AC current equations, the procedure is carried out in a similar way as that of the six-pulse
circuit, taking into account that the width over each conduction period is 30o, Fig.2.35.
• Interval 0 ≤ ωt < π/6
44
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 54/120
0 5 10 15 20 25 30 35 40 45 500
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
nth. harmonic
M a g n i t u d e ( p e r u n i t f u n d a m e n t a l )
Fig. 2.32 vab(t)12 voltage Fourier spectrum
Ia1IanIa1
VL
Lagging
Vm > 1.2732VDC
Leading
Vm < 1.2732VDC
Ia1IanIa1
VL
Lagging
Vm > 1.2732VDC
Leading
Vm < 1.2732VDC
Fig. 2.33 Phasorial diagram
45
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 55/120
-2 -1.5 -1 -0.5 0 0.5 1 1.5 20
0.5
1
1.5
2
Lagging (inductive region) Iq
Leading (capacitive region)
D C v o l t a g e ( p
. u )
-2 -1.5 -1 -0.5 0 0.5 1 1.5 20
0.005
0.01
0.015
0.02
0.025
0.03
0.035
Lagging (inductive region) Iq
Leading (capacitive region)
I q n (
n t h h a r m o n i c m a g n i t u d e )
Iq11
Iq13
Iq23
Fig. 2.34 Relationship between the DC voltage and the fundamental and harmonics current
mV
DCV3
1
DCV3
1−
DCV9107.0
DCV9107.0−
DCV2440.1
DCV2440.1−
mV−
6
π
3
π
2
π
3
2π
6
5π π
6
7π
3
4π
2
3π
3
5π
6
11π π2tω
Fig. 2.35 AC system voltage and fundamental voltage compensator van(t)
46
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 56/120
vL(t) = V m sin(ωt) − 1
3V DC = L
d
dtia(t)
ia(t) = −V m
ωL (cos(ωt) − 1)−1
3L V DC t + I 0
(2.67)
where I 0 is the initial condition at t = 0; ia(0) = I 0.
• Interval π/6 ≤ ωt < π/3
vL(t) = V m sin(ωt) − 0.9107V DC = Ld
dtia(t)
ia(t) = −V mωL
Ãcos(ωt) −
√ 3
2
!−µ
0.9107
Lt − 0.9107π
6ωL
¶V DC + I 1 (2.68)
where: I 1 = ia ³ π
6ω´
• Interval π/3 ≤ ωt < π/2
vL(t) = V m sin(ωt) − 1.2440V DC = Ld
dtia(t)
ia(t) = −V mωL
(cos(ωt) − 0.5) −µ
1.2440
Lt − 1.2440π
3ωL
¶V DC + I 2 (2.69)
where: I 2 = ia
³ π
3ω
´
In steady state ia ³π
2ω´ = 0, then the steady state initial condition I 0 is computed as:
ia
³ π
2ω
´= −V m
ωL(−0.5)−
µ1.2440
L·π
2ω− 1.2440π
3ωL
¶V DC + I 2 = 0
I 2 = − V m2ωL
+1.2440π
6ωLV DC (2.70)
I 2 = ia
³ π
3ω
´= −V m
ωL
Ã1
2−√
3
2
!−µ
0.9107
L·π
3ω− 0.9107π
6ωL
¶V DC + I 1
I 1 =
−
√ 3
V m2ωL
+2.1547π
6ωLV DC (2.71)
I 1 = ia
³ π
6ω
´= −V m
ωL
Ã√ 3
2− 1
!− 1
3L·π
6ωV DC + I 0
I 0 = −V mωL
+7.4641π
18ωLV DC (2.72)
47
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 57/120
Substituting eq. (2.72) into eqn. (2.67) the ia(t) steady state equations are,
ia(t) =
−
V m
ωL
cos(ωt)
−µ1
3L
t
−
7.4641π
18ωL¶V DC (2.73)
0 ≤ ωt < π/6
ia(t) = −V mωL
cos(ωt) −µ
0.9107
Lt − 3.0654π
6ωL
¶V DC (2.74)
π/6 ≤ ωt < π/3
ia(t) = −V mωL
cos(ωt) −µ
1.2440
Lt − 3.7320π
6ωL
¶V DC (2.75)
π/3 ≤ ωt < 2π/3
ia(t) = −V mωL
cos(ωt) −µ
0.9107
Lt − 2.3988π
6ωL
¶V DC (2.76)
2π/3 ≤ ωt < 5π/6
ia(t) = −V mωL
cos(ωt) −µ
1
3Lt +
1.4641π
18ωL
¶V DC (2.77)
5π/6 ≤ ωt < π
Over the interval π≤ω
t < 2π the AC current waveform is the negative of that described for the above
expressions. Fig.2.36 shows the ia(t) AC current waveform. For the leading case V DC = 3.0 V; V m = 2.5
V and L = 3 mH were used; V DC = 3.0 V; V m = 4.5 V and L = 3 mH for the lagging case.
2.4.1.1 Six-pulse AC current
The current flowing into the wye-connected secondary of both transformers (Y − Y and ∆ − Y ) is
equal to the AC line current, this is due to both secondary windings are connected in series. Using the
instantaneous power P (t) the AC current of each six-pulse VSI can be obtained.
Y −
Y connection
• Interval 0 ≤ ωt < π/6
P a(t)s = van(t)sia(t) (2.78)
where: P a(t)s is the phase ”a” instantaneous power in the secondary winding.
48
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 58/120
van(t)s =1
3V DC is the line-to-neutral voltage in the secondary winding.
so that,
P a(t)s = 13
V DC µ−V m
ωLcos(ωt) − µ 1
3Lt − 7.4641π
18ωL
¶V DC
¶(2.79)
At the primary side the instantaneous power is:
P a(t) p = van(t) pia1(t) (2.80)
where: P a(t) p is the phase ”a” instantaneous power in the primary winding.
van(t) p =1
3V DC is the line-to-neutral voltage in the primary winding.
ia1(t) is the AC current in the primary winding
then,
P a(t) p = 13
V DC ia1(t) (2.81)
Neglecting losses, the instantaneous power in both windings is equal, then the AC current in the primary
winding is:
ia1(t) = −V mωL
cos(ωt) −µ
1
3Lt − 7.4641π
18ωL
¶V DC (2.82)
therefore the wye-connected primary current is equal to the AC line current.
∆− Y connection
The instantaneous phase ’a’ power of the ∆− Y transformer is given by:
P a(t)s = van(t)sia(t) = P ab(t) = vab(t) piba(t) (2.83)
where: vab(t) p is the line-to-line voltage in the primary winding (∆).
iba(t) is the current in the primary winding.
van(t)s is the line-to-neutral voltage in the secondary winding (Y ).
Due to
vab(t) p =√
3van(t)s (2.84)
then
iba(t) =1√
3ia(t) (2.85)
49
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 59/120
0 0.005 0 .01 0.015 0.02 0.025 0 .03 0.035 0.04 0 .045 0 .05-1.5
-1
-0.5
0
0.5
1
1.5
Time (s)
A m p
LeadingLagging
0 0.005 0 .01 0.015 0.02 0.025 0 .03 0.035 0.04 0 .045 0 .05-1.5
-1
-0.5
0
0.5
1
1.5
Time (s)
A m p
LeadingLagging
Fig. 2.36 AC current waveform, ia(t)
icb(t) =1√
3ib(t) (2.86)
iac(t) =1√
3ic(t) (2.87)
The AC current of the second VSI ia(t) p is given by:
ia2(t) = iba(t) − iac(t) (2.88)
over the interval 0 ≤ ωt < π/6 the phase current of the second six-pulse VSI is given by eq. (2.89).
ia2(t) = −V mωL
cos³ωt− π
6
´+
µ1
3Lt +
6.4641π
18ωL
¶V DC (2.89)
From the above analysis can be concluded that the current flowing into the delta-connected primary is
phase-shifted by 30owith respect to the wye-connected primary. Fig.2.37 depicts the AC current waveform
of each six-pulse VSI and the current across the diode and transistor one, D1, Q1 of the first six-pulse VSI
is shown in Fig.2.38.
50
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 60/120
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-2
-1
0
1
2
Phase a: ia(t)
A m p
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-2
-1
0
1
2
Phaseb: ib(t)
A m p
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-2
-1
0
1
2
Phasec: ic(t)
Time (s)
Y-Y VSI ∆-Y VSI
A m p
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-2
-1
0
1
2
Phase a: ia(t)
A m p
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-2
-1
0
1
2
Phaseb: ib(t)
A m p
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-2
-1
0
1
2
Phasec: ic(t)
Time (s)
Y-Y VSI ∆-Y VSI
A m p
Fig. 2.37 AC current of each six-pulse VSI
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.050
0.2
0.4
0.6
0.8
1
1.2
1.4
Q1
current: iQ1
(t)
A m p
LeadingLagging
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.050
0.2
0.4
0.6
0.8
1
1.2
1.4
D1
current: iD1
(t)
A m p
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.050
0.2
0.4
0.6
0.8
1
1.2
1.4
Q1
current: iQ1
(t)
A m p
LeadingLagging
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.050
0.2
0.4
0.6
0.8
1
1.2
1.4
D1
current: iD1
(t)
A m p
Fig. 2.38 Q1 and D1 currents
51
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 61/120
2.4.2 Capacitor current
The capacitor current is made up of the DC currents contributed by each six-pulse converter; the ca-
pacitor current is given by,
iDC (t)12 = iDC (t)1 + iDC (t)2 (2.90)
where: iDC (t)12 is the twelve-pulse capacitor current.
iDC (t)1 is the first compensator capacitor current.
iDC (t)2 is the second compensator capacitor current.
To obtain the current of the DC side, the procedure is carried out in a similar way as that of the six-
pulse circuit, taking into account that the width over each conduction period is 30o. The current of the first
six-pulse compensator at the DC side iDC (t)1, during the first 60ois calculated as follows. In all time t,
iDC (t)1 is given by,
iDC (t)1 = ia1(t) + ic1(t)
Taking into account that the conduction interval is 30oyields,
iDC (t)1 = −V mωL
cos³ωt +
π
3
´− 1.2440
L
³t − π
6ω
´V DC (2.91)
0 ≤ ωt < π/3
Therefore, eq. (2.91) gives the first six-pulse compensator current at the DC side of during the first
60ointerval. The current of the second compensator iDC (t)2 presents a similar behaviour lagged 30o. Fig-
ures Fig.2.39 and Fig.2.40 exhibit these waveforms for both cases lagged and leading.
iDC (t)2 = −V mωL
cos³ωt +
π
6
´− 1.2440
L
³t − π
3ω
´V DC (2.92)
π/6 ≤ ωt < π/2
Substituting eqns. (2.91) and (2.92) into eqn. (2.90) results,
iDC (t)12 = −1.9319V mωL
cos
µωt +
5π
12
¶− 2.4880
L
³t − π
12ω
´V DC (2.93)
0 ≤ ωt < π/6
52
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 62/120
-1.5
-1
-0.5
0
0.5
1
1.5
(a)
A m p ( p u
)
-1.5
-1
-0.5
0
0.5
1
1.5
(b)
A m p ( p u )
3
π
3
2π π
3
4π
3
5π π2tω
tω-1.5
-1
-0.5
0
0.5
1
1.5
(a)
A m p ( p u
)
-1.5
-1
-0.5
0
0.5
1
1.5
(b)
A m p ( p u )
3
π
3
2π π
3
4π
3
5π π2tω
tω
Fig. 2.39 First converter capacitor current; a)generating reactive power; b)absorbing reactive power
-1.5
-1
-0.5
0
0.5
1
1.5
(b)
A m p
( p u
)
-1.5
-1
-0.5
0
0.5
1
1.5
(a)
A m p
( p u
)
3
π
3
2π π
3
4π
3
5π π2tω
tω
-1.5
-1
-0.5
0
0.5
1
1.5
(b)
A m p
( p u
)
-1.5
-1
-0.5
0
0.5
1
1.5
(a)
A m p
( p u
)
3
π
3
2π π
3
4π
3
5π π2tω
tω
Fig. 2.40 Second converter capacitor current; a)generating reactive power; b)absorbing reactive power
53
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 63/120
-1.5
-1
-0.5
0
0.5
1
1.5
(a)
A m p ( p
u )
-1.5
-1
-0.5
0
0.5
1
1.5
(b)
A m p ( p u )
3
π
3
2π π
3
4π
3
5π π2tω
tω-1.5
-1
-0.5
0
0.5
1
1.5
(a)
A m p ( p
u )
-1.5
-1
-0.5
0
0.5
1
1.5
(b)
A m p ( p u )
3
π
3
2π π
3
4π
3
5π π2tω
tω
Fig. 2.41 Capacitor current; a)generating; b)absorbing reactive power
The capacitor current waveform over the remaining 11 conduction periods is identical to that described
by eq. (2.93) and results in a repetitive waveform at twelve times the AC power system frequency. Fig.2.41
presents the twelve-pulse capacitor current.
2.4.3 DC capacitor voltage
The capacitor voltage over the first 30oconduction interval is:
vcap(t)12 = −1.9312V m
ω2LC sin
µωt +
5
12π
¶+ 1.8681
V mω2LC
− 1.2440
LC V DC t
2
+1.2440π
6ωLC V DC t + V 0 (2.94)
where V 0 is the initial condition at t = 0; V 0 = vcap(0). The initial condition is calculated using the average
component V DC of eq. (2.94) with T = π/6ω; therefore,
V DC = − 6πω2LC
V m + 1.8681ω2LC
V m + 0.4147π2
72ω2LC V DC + V 0 (2.95)
54
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 64/120
0.95
1
1.05
1.1
(b)
V o l t s ( p u )
0.75
0.8
0.85
0.9
0.95
1
1.05
1.1
(a)
V o l t s ( p u )
3
π
3
2π π
3
4π
3
5π π2tω
tω
0.95
1
1.05
1.1
(b)
V o l t s ( p u )
0.75
0.8
0.85
0.9
0.95
1
1.05
1.1
(a)
V o l t s ( p u )
3
π
3
2π π
3
4π
3
5π π2tω
tω
Fig. 2.42 DC capacitor voltage; a)generating; b)absorbing reactive power
simplifiying
V 0 = 0.04181
ω2LC V m − 0.0568
1
ω2LC V DC + V DC (2.96)
Fig.2.42 depicts the twelve-pulse DC-capacitor voltage. This figure also shows that the peak capacitor
voltage occurs at ωt = 15o when the compensator is generating reactive power. Eq. (2.97) gives the peak
voltage,
V pk = −0.02201
ω2LC V m + 0.0284
1
ω2LC V DC + V DC (2.97)
2.4.4 PSCAD/EMTDC Simulations
To validate the accuracy of the equations developed, digital simulations using the PSCAD/EMTDC
simulator are carried out. Fig.2.43 exhibits the line-to-line, vab(t) and the AC current ia(t), for a twelve-
pulse STATCOM when the compensator is operated with a finite capacitor as a DC source. This figure
illustrates the capacitor ripple voltage effect. The capacitor voltage and the current waveforms are presented
in Fig.2.44. The parameters employed are: C = 220 µF, L = 3 mH, V m = 2.5 V and V DC = 3 V.
55
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 65/120
0 0. 005 0 .01 0.015 0. 02 0. 025 0 .03 0.035 0. 04 0.045 0 .05-1
-0.5
0
0.5
1
(a)
A m p
0 0. 005 0 .01 0.015 0. 02 0. 025 0 .03 0.035 0. 04 0.045 0 .05-8
-6
-4
-2
0
2
4
6
8
(b)
Time (s )
V o
l t s
Fig. 2.43 a)12-pulse AC current ia(t); b)12-pulse line-to-line voltage vab(t)
0 0 .002 0.004 0.006 0.008 0.01 0.012 0.0 14 0. 016-0.5
0
0.5
(a)
A m p
0 0 .002 0.004 0.006 0.008 0.01 0.012 0.0 14 0. 0160
0.5
1
1.5
2
2.5
3
3.5
4
(b)
Time (s)
V o l t s
Fig. 2.44 Capacitor current and DC capacitor voltage waveform
56
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 66/120
2.5 Conclusions
This chapter presents the relevant details of the voltage source-inverter (VSI), the building block of
a STATCOM and other FACTS devices. A detailed analysis is carried out to deduce the expressions for
the AC (phase currents and output voltage) and DC signals (capacitor current and voltage) of a 6- and 12-
pulse STATCOM. The resultant expressions allow to approximate an adequate DC capacitor value. A VSI
generates an AC voltage from a DC voltage, when a VSI is used as a typical STATCOM a fixed DC capacitor
is used as a DC source. An important questions about that are answered: how the energy conversion is
achieved using a DC capacitor?, and if a DC source is not used, how the capacitor voltage is kept fixed?.
Such questions are answered establishing that, when the compensator (neglecting losses) is exchanging
reactive power only, the average power across the DC capacitor is zero, therefore the capacitor does not
change its voltage. The 12-pulse converter is the building block of higher converter (24- and 48-pulse) but
due to the high harmonic content it does not offer a good performance in high power applications, although
it could be applied in industrial applications where the harmonic distortion is not so detrimental.
57
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 67/120
References
[1] Cigre, ”Static Synchronous Compensator”, Working Group 14.19, September 1998.
[2] Yong Hua Song, Allan T. Johns, ”Flexible AC transmission systems FACTS”, IEE Power and Energy
Series 30, 1999.
[3] Muhammad H. Rashid, ”Power Electronics Circuits, Devices, and Applications”, Prentice Hall Inc.,
1995.
[4] D. R. Trainer, S. B. Tennakoon, R. E. Morrison, ”Analysis of GTO-based static VAR compensators,”,
IEE Proc.-Elect. Power Appl., Vol. 141, No. 6, pp. 293-302, November 1994.
58
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 68/120
Chapter 3
24- and 48-pulse STATCOM operation
The 24- and 48-pulse converters are obtained combining two and four 12-pulse VSI respectively, with
an adequate phase shifted between them. For high power applications the best option is the 48-pulse con-
verter, although using filters tuned to the 23th - 25th harmonics a 24-pulse converter could be adequate. In
this chapter the general features of each converter (24- and 48-pulse) will be examined. The analysis is sim-
ilar to that presented in the previous chapter. The transformer arrangement is also treated in detail. Finally
to verify the analytical results digital simulations are carried out using the EMTDC/PSCAD.
3.1 24-pulse operation
The 24-pulse converter is obtained combining two 12-pulse VSI with itsfiring pulse shifted by 15 from
each other, and providing 15 phase-shift windings on the two transformers of one of both 12-pulse VSI. It
is preferred to use transformers of the same characteristics in the two 12-pulse VSI getting a symmetrical
system phase-shift of +7.5 on both transformers of one 12-pulse VSI, and -7.5 on both transformers (see
Fig.3.3) of the other one [1]. A phase-shift of -7.5 and +7.5 in the firing pulse is needed.
3.1.1 24-pulse voltage
This section explains how to get a 24-pulse voltage. To produce the 24-pulse voltage two 12-pulse
voltages must be combined with a phase-shifted between them. Both the 24-pulse voltage and 24-pulse
current exhibit low harmonic rate on AC and DC side. Its AC output voltage presents an harmonic order
content of 24r±1, ∀ r positive integer; that is, n = 1th, 23th, 25th, 47th, 49th, ..., with magnitudes of 1/23th,
1/25th, 1/47th respectively, respect to the fundamental AC voltage. Eq. (3.1) shows the AC output of a basic
12-pulse converter (see chapter 2),
vab12(t) = 2[V ab1 sin(ωt + 30) + V ab11 sin (11ωt + 330) + V ab13 sin(13ωt + 30)
+V ab23 sin (23ωt + 330) + V ab25 sin (25ωt + 30) + ...] (3.1)
59
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 69/120
the compact expression is,
vab12(t) = 2∞
Xn=1
V abn sin(nωt + 30n) (3.2)
∀ n = 12r ± 1, r = 0, 1, 2,...
Assume 12-pulse converters, the AC output voltage of each converter is given by eq. (3.1), the phase
shifts needed to each 12-pulse converter are as follows.
• Lagging 7.5 the firing pulse of the first converter results in,
vab12(t)1 = 2[V ab1 sin(ωt + 22.5) + V ab11 sin (11ωt + 247.5) + V ab13 sin(13ωt + 292.5)
+V ab23 sin (23ωt + 157.5) + V ab25 sin (25ωt + 202.5) + V ab35 sin (35ωt + 67.5) (3.3)
+V ab37 sin (37ωt + 112.5
) + V ab47 sin (47ωt + 337.5
) + V ab49 sin (49ωt + 22.5
) + ...]
• Leading 7.5 the AC output of the first converter, eq. (3.3), using a phase-shifter transformer (PST),
vab12(t)1 = 2[V ab1 sin(ωt + 30) + V ab11 sin (11ωt + 240) + V ab13 sin (13ωt + 300)
+V ab23 sin (23ωt + 150) + V ab25 sin (25ωt + 210) + V ab35 sin (35ωt + 60) (3.4)
+V ab37 sin (37ωt + 120) + V ab47 sin (47ωt + 330) + V ab49 sin (49ωt + 30) + ...]
• Leading 7.5 the firing pulse of the second converter results in,
vab12(t)2 = 2[V ab1 sin(ωt + 37.5) + V ab11 sin (11ωt + 52.5) + V ab13 sin (13ωt + 127.5)
+V ab
23sin (23ωt + 142.5) + V
ab25
sin (25ωt + 217.5) + V ab
35sin (35ωt + 232.5) (3.5)
+V ab37 sin (37ωt + 307.5) + V ab47 sin (47ωt + 322.5) + V ab49 sin (49ωt + 37.5) + ...]
• Lagging 7.5 the AC output of the second compensator, eq. (3.5), using a phase-shifter transformer
(PST),
vab12(t)2 = 2[V ab1 sin(ωt + 30) + V ab11 sin (11ωt + 60) + V ab13 sin (13ωt + 120)
+V ab23 sin (23ωt + 150) + V ab25 sin (25ωt + 210) + V ab35 sin (35ωt + 240) (3.6)
+V ab37 sin (37ωt + 300) + V ab47 sin (47ωt + 330) + V ab49 sin (49ωt + 30) + ...]
From eqs. (3.4) and (3.6) can be noticed that the harmonic components not included in the set n =
12r ± 1 are in phase opposition. Therefore, if the output voltage of both converters (Eqs. (3.4) and (3.6))
are added, it results in a voltage closer to be a sine wave free of 12-pulse harmonic components; that is, the
60
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 70/120
24-pulse voltage vab24(t). Thus, vab24(t) is the line-to-line voltage of a 24-pulse compensator,
vab24(t) = vab12(t)1 + vab12(t)2 (3.7)
vab24(t) = 4[V ab1 sin(ωt + 30) + V ab23 sin (23ωt + 150) + V ab25 sin(25ωt + 210)
+V ab47 sin (47ωt + 330) + V ab49 sin (25ωt + 30) + ...] (3.8)
The general expression is given by,
vab24(t) = 4∞X
n=1
V abn sin(nωt + 22.5n + 7.5i) (3.9)
∀ n = 24r ± 1, r = 0, 1, 2,...
where: i = 1 for positive sequence harmonics, abc sequence.
i = -1 for negative sequence harmonics, cba sequence.
The line-to-neutral voltage, phase ‘a’, is given by,
van24(t) =
4√ 3
[V ab1 sin(ωt)−V ab23 sin(23ωt)−V ab25 sin (25ωt) +V ab47 sin (47ωt)+V ab49 sin(25ωt) + ...]
(3.10)
That is,
van24(t) = 4√ 3
∞Xn=1
V abn sin(nωt + 22.5n − 22.5i) (3.11)
∀ n = 24r ± 1, r = 0, 1, 2,...
Voltages vbn24(t) and vcn24
(t) are phase shifted by -120 and -240, respectively, from van24(t). Fig.3.1
depicts the line-to-line voltage vab24(t) and the line-to-neutral voltage van24(t).
From Fig.3.2 can be observed that the 24-pulse voltage vab24(t), hold only harmonics of order n =
24r ± 1, r = 0, 1, 2, ..., i.e. n = 1th, 23th, 25th, 47th, 49th, ..., with amplitudes 1/23th, 1/25th, 1/47th, 1/49th,
..., respectively.
61
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 71/120
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035-5
-4
-3
-2
-1
0
1
2
3
4
5
Time(s)
V o
l t s (× V
D C
)
vab
(t)
van
(t)
Fig. 3.1 24-pulse voltage
0 10 20 30 40 50 60 70 80 90 1000
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
nth. harmonic
M a g n
i t u d e
( p e r u n
i t f u n
d a m e n
t a l )
Fig. 3.2 24-pulse voltage Fourier spectrum
62
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 72/120
3.1.2 Magnetic interface
To obtain the 24-pulse compensator two 12-pulse converters phase-shifted from each other are needed;
this relation can be obtained through shifting the firing pulse and the PST. Fig.3.3 presents the layout of a
24-pulse STATCOM. It is comprised by four 6-pulse VSI and two 12-pulse transformers with PSW (phase-
shifter windings). Fig.3.4 depicts the 12-pulse transformer with the PSW on lagging and leading configura-
tion.
The phase shifter transformer (PST) [2,3] is a power transformer that provides regulation on voltage
magnitude and/or phase-angle. The variation on the phase-angle is achieved injecting a quadrature voltage
with respect to the sending voltage; this is ilustrated in Fig.3.5, where V x is the sending voltage (input
voltage), V y is the quadrature voltage and the V 0x is the receiving voltage (output voltage).
From Fig.3.5 is clear that the phase-shift θ, is determined by the magnitude of the quadrature voltage
V y. For the 24-pulse STATCOM the phase-shifter winding is fixed to provide a phase-shift of ±7.5 (leading
or lagging), this can be got with a 1:tan(7.5) turn ratio.
3.1.3 AC current signals
The current flowing between the compensator and the AC system is obtained solving eq. (3.12) for
each phase,
ein(t) − vin24(t) = L
d
dtii(t) + Rii(t) (3.12)
where i = a, b, c, ein(t) is the AC system voltage, vin24(t) is the 24-pulse STATCOM output voltage, R
represents the sum of the transformer winding resistance losses and the inverter conduction losses; and L
is the leakage inductance. Fig.3.6 shows the typical 24-pulse-current waveform, it can be noticed that the
behaviour of the current is close to a sinusoidal waveform, therefore it will be considered as a pure sinusoidal
function, that is, just the fundamental component is taken into account.
Let be the AC system voltage a pure sinusoidal function ean(t) = V m sin(ωt) , and the STATCOM AC
voltage van(t) = 2.5463V DC sin(ωt) (from eq. 2.13 and eq. 3.11 ), then the AC current for the phase ’a’
(neglecting losses, R = 0 Ω) is given by:
ia(t) = −I max cos(ωt) (3.13)
63
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 73/120
C
V C C
V C C
V C C
V C C
a a
a a a a b b
b b
b b
c c
c c c c
G N D
G N D
G N D
G N D
V S I
V S I
V S I
V S I
- 7 . 5
°
- 3 7 . 5
°
+ 7 . 5
°
- 2 2 . 5
°
+ 7 . 5
°
- 7 . 5
°
a ’
a ’
b ’
b ’
c ’ c ’
a Y
a Y b
Y b
Y c Y c Y a∆
a∆
b b c∆ c∆
Y - Y
Y - Y
Y -∆
Y -∆
L
R
i a ( t )
i b ( t )
i c ( t )
i a ( t ) 1
i a ( t ) 2
i a ( t ) 3
i a ( t ) 4
i D C
( t ) 1
i D C
( t ) 2
i D C
( t ) 3
i D C
( t ) 4
i D C (
t )
Fig. 3.3 24-pulse STATCOM
64
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 74/120
Vcb
(leading)
Vbc(lagging)
Vac(leading)
Vca(lagging)
Vba(leading)
Vab(lagging)
Vcb(leading)
Vbc(lagging)
Vac(leading)
Vca(lagging)
Vba(leading)
Vab(lagging)
ain
bin
cin
Signals from
a 6-pulse VSI
a’in
b’in
c’in
aout
bout
cout
Signals from
a 6-pulse VSI
Y-Y Transformer
1:1 Turn ratio
Fig. 3.4 12-pulse grounded transformer with PST on leading and lagging configuration
Vx
Vx’
Vy
θ
Fig. 3.5 Phasor diagram of the phase-shifting mechanism
65
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 75/120
where:
I max =V m − 2.5463V DC
ωL
In general, applying phasor analysis and considering the effect of losses and a phase-shift between the
AC system voltage and the STATCOM, yields
I a =ean − V anq R2 + (ωL)2
= I max]θ (3.14)
The current will be leading when V m < 2.5463V DC ; thus, the compensator is seen as a capacitor by
the AC system and the current flows from the compensator to the AC system. The current will be lagging
when V m > 2.5463V DC ; thus the compensator behaves as an inductor and the current flows from the AC
system to the compensator, this is illustrated in Fig.3.7.
3.1.4 Capacitor current
The capacitor current is made up by the contribution of each one of the four 6-pulse VSI that comprised
the 24-pulse STATCOM,
iDC (t) = iDC (t)1 + iDC (t)2 + iDC (t)3 + iDC (t)4 (3.15)
where: iDC (t) is the 24-pulse capacitor currentiDC (t)i is the ith 6-pulse VSI capacitor current, i = 1, 2, 3, 4
To obtain the contribution of each VSI the procedure presented in Chapter 2 (see 6- and 12-pulse
analysis) will be carry out. iDC (t)1 over the interval 7.5 ≤ ωt < 67.5 is given by,
iDC (t)1 = ia(t) + ic(t)
iDC (t)1 = I max cos(ωt + θ + 60) (3.16)
7.5 ≤ ωt < 67.5
The current of the second VSI, iDC (t)2 is lagged by 30 with respect to the first one, thus
iDC (t)2 = I max cos(ωt + θ + 90) (3.17)
7.5 ≤ ωt < 37.5
66
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 76/120
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
Time (s)
A m p
LeadingLagging
Fig. 3.6 AC current waveform, ia(t)
I a 1I a nI a 1
V L
L ag g in g
V m > 2 .5 4 6 3 V D C
L ea d in g
V m < 2 .5 4 6 3 V D C
I a 1I a nI a 1
V L
L ag g in g
V m > 2 .5 4 6 3 V D C
L ea d in g
V m < 2 .5 4 6 3 V D C
Fig. 3.7 Phasorial diagram
67
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 77/120
Combining iDC (t)1 and iDC (t)2 a 12-pulse capacitor current is obtained,
iDC 12(t)1 = 1.9319I max cos(ωt + θ + 75) (3.18)
7.5 ≤ ωt < 37.5
For the remaing VSI’s,
iDC (t)3 = iDC (t)1; 0 ≤ ωt < 52.5
iDC (t)4 = iDC (t)2; 0 ≤ ωt < 22.5
iDC 12(t)2 = 1.9319I max cos(ωt + θ + 75) (3.19)
7.5
≤ ωt < 22.5
Therefore the 24-pulse capacitor current is:
iDC (t) = 3.8638I max cos(ωt + θ + 75) (3.20)
7.5 ≤ ωt < 22.5
The capacitor current over each of the remaining twenty-three conduction periods is identical to that
described by eq. (3.20), and yields a waveform where its frequency is 24 times the frequency of the AC
system. Fig.3.8 depicts the 24-pulse capacitor current.
3.1.5 DC capacitor voltage
The capacitor voltage over a 15 conduction interval is:
vcap(t) =1
C
Z t
tx
iDC (t)dt + V x (3.21)
where: V x = vcap(tx)
tx =7.5π
180ωso that,
vcap(t) = −3.8638
ωC I max [sin(ωt + θ + 75) − sin (82.5 + θ)] + V x (3.22)
68
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 78/120
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016-1.5
-1
-0.5
0
0.5
1
1.5
(a)
A m p
( p
u )
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016-1.5
-1
-0.5
0
0.5
1
1.5
Time (s )
A m p
( p u
)
Fig. 3.8 Capacitor current; a)generating reactive power; b)absorbing reactive power
7.5 ≤ ωt < 22.5
To calculate the initial condition V 0 at t = 0 s, first V x will be calculated using the average component of eq.
(3.22).
V DC =1
T
tyZ tx
vcap(t)dt (3.23)
where: T =π
12ω; ty =
22.5π
180ω
V DC = −46.3656
πωC I max [cos (97.5 + θ) − cos (82.5 + θ)] − 3.8638
ωC I max sin(82.5 + θ) + V x (3.24)
Eq. (3.24) yields,
V x = V DC +46.3656
πωC
I max [cos (97.5 + θ)−
cos (82.5 + θ)] +3.8638
ωC
I max sin (82.5 + θ) (3.25)
if R = 0 then θ = 0,therefore,
V x = V DC −12.1039
πωC I max +
3.8307
ωC I max (3.26)
69
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 79/120
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.0160.985
0.99
0.995
1
1.005
1.01
(a)
V o l t s ( p u )
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.0160.996
0.998
1
1.002
1.004
1.006
1.008
(b)
Time (s)
V o l t s ( p u )
Fig. 3.9 DC capacitor voltage; a)generating reactive power; b)absorbing reactive power
calculating V 0; V 0 = vcap(tz), where tz =15π
180ω
V 0 =3.8638
ωC I max [sin (90 + θ) − sin (82.5 + θ)] + V x (3.27)
Fig.3.9 shows the 24-pulse DC-capacitor voltage.
3.2 48-pulse operation
Two 24-pulse converters, phase-shifted by 7.5 from each other, can provide a 48-pulse converter.
Using a symmetrical criterion the 7.5 are provided in the following way: phase-shift windings with -3.75
on the two transformers of one 24-pulse converter, and +3.75 on the two transformers of the other one. The
firing pulse needs a phase-shift of +3.75 and -3.75, respectively.
The 48-pulse converter is comprised by four 12-pulse converter linked by four 12-pulse transformers
with phase-shift windings (see Fig.3.4). Fig.3.10 depicts the schematic diagram of a 48-pulse STATCOM.
The transformer connexion and the necessary firing pulse logic to get the 48-pulse operation will be treated
in the next section.
70
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 80/120
3.2.1 48-pulse voltage
The 48-pulse converter can be used in high power applications without AC filters due to its high per-
formance and low harmonic rate on the AC side. The output voltage have harmonics n = 48r ± 1, where
r = 0, 1, 2,...; i.e., 47th, 49th, 95th, 97th,..., with magnitudes of 1/47th, 1/49th , 1/95th, 1/97th, ..., respec-
tively, respect to the fundamental; on the DC side the lower circulating harmonic current will be the 48th.
The phase-shift pattern on each 12-pulse converter is the following:
1th 12-pulse converter
PST: +7.5 Necessary to eliminate the 24-pulse harmonics
+3.75 Necessary to eliminate the 48-pulse harmonics
Total +11.25 Winding turn rate 1:tan(11.25)
Driver: -7.5 Necessary to eliminate the 24-pulse harmonics
-3.75 Necessary to eliminate the 48-pulse harmonics
Total -11.25
The resultant output voltage generated by this 12-pulse converter is,
vab12(t)1 = 2[V ab1 sin(ωt + 30) + V ab11 sin (11ωt + 195) + V ab13 sin (13ωt + 255)
+V ab23 sin (23ωt + 60) + V ab25 sin (25ωt + 120) + ...] (3.28)
2th 12-pulse converter PST: -7.5 Necessary to eliminate the 24-pulse harmonics
+3.75 Necessary to eliminate the 48-pulse harmonics
Total -3.75 Winding turn rate 1:tan(3.75)
Driver: +7.5 Necessary to eliminate the 24-pulse harmonics
-3.75 Necessary to eliminate the 48-pulse harmonics
Total +3.75
The resultant output voltage generated by this 12-pulse converter is,
vab12(t)2 = 2[V ab1 sin(ωt + 30) + V ab11 sin (11ωt + 15) + V ab13 sin (13ωt + 75)
+V ab23 sin (23ωt + 60) + V ab25 sin (25ωt + 120) + ...] (3.29)
3th 12-pulse converter
71
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 81/120
Fig. 3.10 48-pulse STATCOM
72
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 82/120
PST: +7.5 Necessary to eliminate the 24-pulse harmonics
-3.75 Necessary to eliminate the 48-pulse harmonics
Total +3.75
Winding turn rate 1:tan(3.75
)Driver: -7.5 Necessary to eliminate the 24-pulse harmonics
+3.75 Necessary to eliminate the 48-pulse harmonics
Total -3.75
The resultant output voltage generated by this 12-pulse converter is,
vab12(t)3 = 2[V ab1 sin(ωt + 30) + V ab11 sin (11ωt + 285) + V ab13 sin (13ωt + 345)
+V ab23 sin (23ωt + 240) + V ab25 sin (25ωt + 300) + ...] (3.30)
4th 12-pulse converter
PST: -7.5 Necessary to eliminate the 24-pulse harmonics
-3.75 Necessary to eliminate the 48-pulse harmonics
Total -11.25 Winding turn rate 1:tan(11.25)
Driver: +7.5 Necessary to eliminate the 24-pulse harmonics
+3.75 Necessary to eliminate the 48-pulse harmonics
Total +11.25
The resultant output voltage generated by this 12-pulse converter is,
vab12(t)4 = 2[V ab1 sin(ωt + 30) + V ab11 sin (11ωt + 105) + V ab13 sin (13ωt + 165)
+V ab23 sin (23ωt + 240) + V ab25 sin (25ωt + 300) + ...] (3.31)
These four 12-pulse AC output voltages, given by eqs. (3.29-3.31), are added connecting in series the
secondary windings of the transformers. The 48-pulse AC output voltage is given by:
vab48(t) = vab12(t)1 + vab12(t)2 + vab12(t)3 + vab12(t)4 (3.32)
vab48(t) = 8[V ab1 sin(ωt + 30) + V ab47 sin (47ωt + 150) + V ab49 sin(49ωt + 210)
+V ab95 sin (95ωt + 330) + V ab97 sin (97ωt + 30) + ...] (3.33)
73
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 83/120
The general expression for the 12-pulse AC output voltage is given by:
vab12(t) = 2∞
Xn=1
V abn sin(nωt + 30n) (3.34)
∀ n = 12r ± 1, r = 0, 1, 2,...
Applying a 11.25 phase-shift provided by a PST to eq. (3.34),
vab12(t) = 2∞X
n=1
V abn sin(nωt + 30n + 11.25i) (3.35)
∀ n = 12r ± 1, r = 0, 1, 2,...
where: i = 1 for positive sequence harmonics, abc sequence.i = -1 for negative sequence harmonics, cba sequence.
Now lagging the firing pulse 11.25, results,
vab12(t) = 2∞X
n=1
V abn sin(nωt + 18.75n + 11.25i) (3.36)
∀ n = 12r ± 1, r = 0, 1, 2,...
Therefore the general expression of the line-to-line voltage associated to a 48-pulse converter is given
by eq. (3.37),
vab48(t) = 8∞X
n=1
V abn sin(nωt + 18.75n + 11.25i) (3.37)
∀ n = 48r ± 1, r = 0, 1, 2,...
The line-to-neutral voltage is,
van48(t) =
8√ 3
[V ab1 sin(ωt)−V ab47 sin(47ωt)−V ab49 sin (49ωt) +V ab95 sin (95ωt)+V ab97 sin(97ωt) + ...]
(3.38)
that can be expressed as,
van48(t) =
8√ 3
∞Xn=1
V abn sin(nωt + 18.75n − 18.75i) (3.39)
74
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 84/120
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035-10
-8
-6
-4
-2
0
2
4
6
8
10
Time (s)
V o
l t s (× V
D C
)
vab
(t)
van
(t)
Fig. 3.11 48-pulse voltage
∀ n = 48r ± 1, r = 0, 1, 2,...
Voltages vbn48(t) and vcn48
(t) have a similar pattern except phase shifted by 120 and 240, respec-
tively, from van48(t). Fig.3.11 depicts the 48-pulse line-to-line and the line-to-neutral voltage; the harmonic
content is exhibited in Fig.3.12.
3.2.2 AC current signals
The low harmonic rate of a 48-pulse STATCOM allows to approach its AC current just by the funda-
mental signal, the typical AC current waveform is shown in Fig.3.13.
The AC current for the phase ’a’ is given by eq. (3.14), although for a 48-pulse converter I max is given
by eq. (3.40),
I max = V m − 5.0927V DC
ωL(3.40)
The current will be leading when V m < 5.0927V DC ; thus, the compensator is seen as a capacitor by
the AC system and the current flows from the compensator to the AC system; the current will be lagging
75
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 85/120
0 10 20 30 40 50 60 70 80 90 1000
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
nth. harmonic
M a g n
i t u d e
( p e r u n
i t f u n
d a m e n
t a l )
Fig. 3.12 48-pulse voltage Fourier spectrum
when V m > 5.0927V DC ; thus the compensator behaves as an inductor and the current flows from the AC
system to the compensator.
3.2.3 Capacitor current
The capacitor current is made up by the contribution of each one of the four 12-pulse VSI that com-
prising the 48-pulse STATCOM. The capacitor current will be calculated in similar way to that presented
for the 24-pulse. Summing the DC current of each 6-pulse VSI,
iDC (t) = iDC (t)1 + iDC (t)2 + iDC (t)3 + iDC (t)4 + iDC (t)5 + iDC (t)6 + iDC (t)7 + iDC (t)8 (3.41)
The current of each 6-pulse VSI is the following one,
iDC (t)1 = I max cos(ωt + θ + 60) ; 11.25 ≤ ωt < 71.25
iDC (t)2 = I max cos(ωt + θ + 90) ; 11.25
≤ωt < 41.25
iDC (t)3 = I max cos(ωt + θ + 60) ; 0 ≤ ωt < 56.25
iDC (t)4 = I max cos(ωt + θ + 90) ; 0 ≤ ωt < 26.25
iDC (t)5 = I max cos(ωt + θ + 60) ; 3.75 ≤ ωt < 63.75
76
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 86/120
iDC (t)6 = I max cos(ωt + θ + 90) ; 0 ≤ ωt < 33.75
iDC (t)7 = I max cos(ωt + θ + 60) ; 0 ≤ ωt < 48.75
iDC (t)8 = I max cos(ωt + θ + 90
) ; 0
≤ ωt < 18.75
The 48-pulse capacitor current shown in Fig.3.14 is given by the expression eq (3.42). Its frequency is
48 times the frequency of the AC system. The remaining 47 segments are expressed symilarly.
iDC (t) = 7.7276I max cos(ωt + θ + 75) (3.42)
11.25 ≤ ωt < 18.75
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
Time (s)
A m p
Leading
Lagging
Fig. 3.13 AC current waveform, ia(t)
3.2.4 DC capacitor voltage
The capacitor voltage over a 7.5 conduction interval is given by eq. (3.43):
vcap(t) =7.7276
ωC I max [sin(ωt + θ + 75) − sin (86.25 + θ)] + V x (3.43)
11.25 ≤ 5ωt < 18.7
77
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 87/120
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016-1.5
-1
-0.5
0
0.5
1
1.5
(a)
A m p ( p u )
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016-1.5
-1
-0.5
0
0.5
1
1.5
(b)
Time (s)
A m p ( p u )
Fig. 3.14 Capacitor current; a)generating reactive power; b)absorbing reactive power
where: V x = vcap(tx); tx =11.25π
180ω.
The initial condition V 0 at t = 0 s given by the eq. (3.46) is calculated by the similar procedure used
previously,
V DC = 1T Z ty
txvcap(t)dt (3.44)
where: T =π
24ω; tx =
11.25π
180ω; ty =
18.75π
180ω.
V x = V DC +185.4624
πωC I max [cos (93.75 + θ) − cos (86.25 + θ)] +
7.7276
ωC I max sin (86.25 + θ) (3.45)
V 0 = vcap(tz); tz =15π
180ω
V 0 =7.7276
ωC I max [sin (90 + θ) − sin (86.25 + θ)] + V x (3.46)
Fig.3.15 shows the capacitor voltage; the peak capacitor voltage occurs at ωt = 7.5 when the com-
pensator is generating reactive power.
78
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 88/120
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016
0.994
0.996
0.998
1
1.002
1.004
(a)
V o l t s ( p u )
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.0160.998
0.999
1
1.001
1.002
1.003
1.004
(b)
Time (s)
V o l t s ( p u )
Fig. 3.15 Capacitor voltage; a)generating reactive power; b)absorbing reactive power
3.3 Digital simulations
In this section digital simulations using the EMTDC/PSCAD are carried out in order to verify the
analytical expressions developed in previous sections. Two cases will be presented: (1) using a fixed DC
value, that is, considering a infinite capacitor, and (2) considering a finite capacitor.
The first one is shown in Fig.3.16 and Fig.3.17. (24- and 48-pulse respectively). In the second case
the capacitor ripple voltage effect is illustrated, Fig.3.18 and Fig.3.19 (24- and 48-pulse respectively). It is
worth noting that to get a good harmonic performance and small DC voltage ripple an appropriate size of
DC capacitor must be chosen. The parameters employed are:
C = 470 µF (24-pulse), 1000 µF (48-pulse).
L = 3 mH
R = 0.2 Ω
V m = 2.5 V
V DC = 1.5 V (24-pulse); 0.75 V (48-pulse).
Note that a resistor R is used as internal losses; thereby, a small phase-shift between the AC system
voltage and the STATCOM voltage should be used to supply the internal losses.
79
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 89/120
3.4 Conclusions
The simplest three-phase converter is the six-pulse converter however even at low rating a simple six-
pulse converter is unlikely in high power applications due to its harmonic distortion. To reduce the harmonic
distortion to an acceptable level the pulse order must be increased to 24- or 48-pulse.
To get the 24- and 48-pulse converter complex phase-shifting transformers are needed but an approx-
imate behaviour can be obtained using simple 12-pulse transformers; two 12-pulse transformers having
relative phase displacements of 15, quasi 24-pulse operation, and four 12-pulse transformers having 7.5,
quasi 48-pulse operation. The quasi 48-pulse operation allows a small circulation of 12- and 24- pulse order
harmonic due to the 12-pulse harmonics which are characteristic of each 12-pulse converter are not perfectly
cancelled, but the residual magnitude is acceptably small [4].
80
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 90/120
Fig. 3.16 24-pulse behaviour with a fixed V DC value
81
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 91/120
Fig. 3.17 24-pulse behaviour with a finite DC capacitor
82
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 92/120
Fig. 3.18 48-pulse behaviour with fixed V DC value
83
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 93/120
Fig. 3.19 48-pulse behaviour with a finite DC capacitor
84
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 94/120
References
[1] Naraing G. Hingorani, Laszlo Gyugyi, Understanding FACTS, IEEE Press 2000.
[2] M. R. Iravani, P. L. Dandeo, K. H. Nguyen, D. Zhu, d. Maratukulam, ”Applications of static phase
shifters in power systems,” IEEE Trans on Power Delivery , Vol 9, No. 3, July 1994, pp. 1600-1608.
[3] A. Krämer, J. Ruff, ”Transformers for phase angle regulation considering the selection of On-load
tap-changers,” IEEE Trans on Power Delivery , Vol. 13, No.2, April 1998, pp. 518-523.
[4] Yong Hua Song, Allan T. Johns, Flexible AC transmition systems FACTS, IEE Power and Energy
Series 30, 1999.
85
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 95/120
Chapter 4
STATCOM Modelling
This chapter establishes a detailed procedure to model the STATCOM; a generalised model based on
switching functions [1] is developed. Based on such switching functions model, a state-space representation
in the dq-reference frame is deduced. The accuracy of the developed models are validated comparing the
analytical results and digital simulations using the EMTDC/PSCAD.
4.1 Switching functions model
The general model or switching functions model is based on the on/off transistor switching strategy,the nonlinear characteristics of the transistors such as the turn off time and the turn on time and others are
neglected, the semiconductor devices are treated as ideal switches.
Fig.4.1 shows a three-phase bridge or six-pulse converter, this figure illustrates that each converter leg
is composed of a transistor (thyristor) and a diode to permit bi-directional current flow due to both devices
can be represented by a switch (neglecting losses). To describe the modelling procedures, the phase ‘a’
arm of the six-pulse converter is considered, where the semiconductor devices are represented by switches
Sw1 and Sw4; the resistance R in series with the inverter represents the sum of the transformer winding
resistance losses and the inverter conduction losses; the inductance L represents the leakage inductance of
the transformer, Fig.4.2.
The loop equation that described the behaviour of the STATCOM circuit can be expressed as
ean(t) − van(t) = Ld
dtia(t) + Ria(t) (4.1)
where van(t) is the converter output voltage determined by the gating signals and the DC voltage.
Regardless of the PWM technique the switches in one arm are complementary, that is, if the upper
switch is on the lower is off and vice-versa. f s1 is defi
ned as the switching function of switch Sw1and f s4 of Sw4. f s1 and f s4 are either 1 or 0 corresponding to on and off switch states respectively. Due to the switches
are complementary,
f s1 + f s4 = 1 (4.2)
86
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 96/120
Q1
Q4
Q3 Q5
Q6 Q2
D1 D3 D5
D1 D6 D2
a b c
g1 g3 g5
g4 g6 g2
L L L
C
ICD
vAN
(t) vBN(t) vCN(t)
VCD
+
ia(t) i b(t) ic(t)
Fig. 4.1 Six-pulse VSI-STATCOM
87
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 97/120
Q1
Q4
D1
D1
a
g1
g4
C
LR
ean
ia
VDC
iDC
Fig. 4.2 phase ’a’ arm
voltage van(t) is given by
van(t) = vF H (t) + vHn (4.3)
If Sw1 is on, f s1 = 1 and f s4 = 0, therefore,
vF H (t) = V DC f s1 (4.4)
on the other hand, if Sw4 is off, f s1 = 0 and f s4 = 1,then,
vF H (t) = 0 (4.5)
Therefore expressing eq. (4.3) as a function in terms of the switching function
van(t) = V DC f s1 + vHn (t) (4.6)
Substituting eq (4.6) into (4.1)
Ld
dtia(t) + Ria(t) = ean(t) − V DC f s1 − vHn (t) (4.7)
88
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 98/120
Similarly, for phases ‘b’ and ‘c’,
Ld
dt
ib(t) + Rib(t) = ebn(t)
−V DC f s3
−vHn(t) (4.8)
Ld
dtic(t) + Ric(t) = ecn(t) − V DC f s5 − vHn (t) (4.9)
The vHn (t) voltage is measured from the negative VDC terminal to the neutral AC point. It is obtained
adding the equations of the three phases assuming a balanced system (ia + ib + ic = 0), so that
vHn(t) = −V DC
3
Xi=1,3,5
f si (4.10)
Substituting (4.10) into (4.7) the final expression for phase ‘a’ is deduced.
L ddt
ia(t) = −Ria(t) −f s1 − 1
3
Xi=1,3,5
f si
V DC + ean(t) (4.11)
similarly for phases ’b’ and ’c’,
Ld
dtib(t) = −Rib(t) −
f s3 −
1
3
Xi=1,3,5
f si
V DC + ebn(t) (4.12)
Ld
dtic(t) = −Ric(t) −
f s5 −
1
3
Xi=1,3,5
f si
V DC + ecn(t) (4.13)
The above equations show three of the four equations, three currents and the capacitor voltage-the DC
voltage-. Deducing the expression for the DC capacitor voltage, we have
iDC (t) = C d
dtvDC (t) (4.14)
From chapter 2 we know that
iDC (t) = ia(t)f s1 + ib(t)f s3 + ic(t)f s5 (4.15)
iDC (t) =X
i=1,3,5
X j=a,b,c
i j(t)f si (4.16)
therefore
d
dtvDC (t) =
1
C (ia(t)f s1 + ib(t)f s3 + ic(t)f s5) (4.17)
89
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 99/120
The generalised model or switching function model for a six-pulse converter is given by equations
(4.11)-(4.17), where the adopted pulse-width modulation (PWM) scheme is taking into account on the
switching functions f si .The multi-pulse converters 12-, 24- and 48-pulse are comprised by the union of 6-pulse converters, the
lowest practical and building block of multi-pulse converters is the 12-pulse converter; it is comprised by
two 6-pulse converters linked by a 12-pulse transformer (see chapter 2) with a common DC capacitor. In the
next section the model developed above will be extended to a 12-pulse converter, the interaction between
the two six-pulse converter will be studied.
4.1.1 12-pulse converter
For a 12-pulse converter the contribution of the second six-pulse converter linked by the ∆-Y trans-
former must be taken into account. The line-to-neutral voltage van(t) due to a 12-pulse converter is given
by eq. (4.18).
van(t) = van(t)Y Y + van(t)∆Y (4.18)
where van(t)Y Y is the line-to-neutralvoltage in the secondary winding of theY-Y transformer, and van(t)∆Y
is the line-to-neutral voltage in the secondary winding of the ∆-Y transformer. van(t)Y Y was calculated
above, eq. (4.6).
The line-to-neutral voltage in the secondary winding of the ∆-Y transformer is calculated using the
line-to-line voltage of the three phase bridge. The line-to-line voltage of the six-pulse converter linked by
the Y-Y transformer vab(t)Y is given by
vab(t)Y = van(t)Y Y − vbn(t)Y Y (4.19)
where: van(t)Y Y = vDC (t)f s1 + vHn (t)
vbn(t)Y Y = vDC (t)f s3 + vHn (t)
so that
vab(t)Y = vDC (t) (f s1 + f s3) (4.20)
in the∆-Y six-pulse converter the line-to-line voltage on the delta side is
vab(t)∆ = vDC (t) (f s12 + f s32) (4.21)
90
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 100/120
where f si2 are the switching functions in the second six-pulse converter. The line-to-neutral voltage in the
Y side of the ∆-Y taking into account the turn ratio results in.
van(t)∆Y = 13
vDC (t) (f s12 + f s32) (4.22)
Similarly for voltages vbn(t)∆Y and vcn(t)∆Y
vbn(t)∆Y =1
3vDC (t) (f s32 + f s52) (4.23)
vcn(t)∆Y =1
3vDC (t) (f s52 + f s12) (4.24)
Eq. (4.25) expresses the current capacitor due to a 12-pulse converter where iDC (t)1 is the current of
the first six-pulse converter (Y-Y connection) and iDC (t)2 is the current in the second six-pulse converter
(∆-Y connection).
iDC (t) = iDC (t)1 + iDC (t)2 (4.25)
The current due to the second converter is calculated in a similar way to that presented for the first
converter (for the eqs. (4.27) to (4.29) see 12-pulse analysis in chapter 2). Thus, the current of the second
converter is
iDC (t)2 = ia(t)2f s12 + ib(t)2f s32 + ic(t)2f s52 (4.26)
where:
ia(t)2 = iba(t) − iac(t) =1
3(ia(t) − ic(t)) (4.27)
ib(t)2 = icb(t) − iba(t) =1
3(ib(t) − ia(t)) (4.28)
ic(t)2 = iac(t) − icb(t) =1
3(ic(t) − ib(t)) (4.29)
Substituting the above expressions into eq. (4.26) yields
iDC (t)2 =1
√ 3[(f s12
−f s32) ia(t) + (f s32
−f s52) ib(t) + (f s52
−f s12) ic(t)] (4.30)
Such equations represent the generalised switching model for a 12-pulse converter regardless of the
adopted pulse-width modulation (PWM) scheme. The 24-, and 48-pulse converters are comprised by the
combination of two and four 12-pulse converters respectively. To model a system greater than 12-pulse, the
91
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 101/120
contribution of each 12-pulse unit must be taken into account with the appropriate phase shift in the gating
signals.
From Fig.4.3 to Fig.4.6 the behaviour of the 6-, 12-, 24-, and 48-pulse converter obtained with thedeveloped model is shown. The parameters used are as follows.
• Six-pulse
V m = 2.5 V, V DC = 6 V, R = 0.2 Ω, L = 3 mH, C = 1000 µF, α = −6o
• 12-pulse
V m = 2.5 V, V DC = 3 V, R = 0.4 Ω, L = 6 mH, C = 1000 µF, α = −5.5o
• 24-pulse
V m = 2.5 V, V DC = 1.5 V, R = 0.8 Ω, L = 12 mH, C = 1000 µF, α = −6o
• 48-pulse
V m = 2.5 V, V DC = 0.75 V, R = 1.6 Ω, L = 24 mH, C = 1000 µF, α = −5.85o
The results obtained in this section are comparable with the ones shown in previous sections.
4.2 STATCOM model at fundamental frequency
The main advantage of the switching model is that it is applicable to various forms of pulse-width
modulation (PWM) or to other switching strategies [1], although it is a model that is hard to work applying
classical control theory, there are a few methods and tools in the theory of nonlinear dynamics that have
proved be very useful in studying the behaviour of power electronic converters; sliding mode schemes have
also been proposed on the basis of switching models [2-4].
Due to the complexity of the switching models, the modelling is typically performed using an approx-
imate continuous time representation (average models) of the converter in the synchronous reference frame
(dq). The conventional averaging technique gives a useful representation of the system and allows simple
design procedures for operation in certain regimes; however, there are several limitations to this modelling
approach. These include the inability to [5]:
• Represent the inherent discrete time nature of the VSI switching, which limits the closed loop perfor-
mance of the VSI.
• Account for the effect harmonics on the steady-state fundamental frequecy behaviour of the VSI.
• Model resonances, occurring between the AC and DC sides of the VSI, as well as those between the AC
92
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 102/120
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-2
-1
0
1
2
AC current ia(t)
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-10
-5
0
5
10Line-to-line and Line-to-neutral voltage
V o
l t s
vab
(t)
van
(t)
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-1
-0.5
0
0.5
1Capacitor current
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.355.6
5.8
6
6.2
6.4
Capacitor voltage
Time (s)
V o
l t s
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-2
-1
0
1
2
AC current ia(t)
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-10
-5
0
5
10Line-to-line and Line-to-neutral voltage
V o
l t s
vab
(t)
van
(t)
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-1
-0.5
0
0.5
1Capacitor current
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.355.6
5.8
6
6.2
6.4
Capacitor voltage
Time (s)
V o
l t s
Fig. 4.3 Six-pulse behaviour
93
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 103/120
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-1
-0.5
0
0.5
1
AC current ia(t)
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-10
-5
0
5
10Line-to-line and Line-to-neutral voltage
V o l t s
vab
(t)
van
(t)
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-0.4
-0.2
0
0.2
0.4Capacitor current
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35
2.95
3
3.05
Capacitor voltage
Time (s)
V o l t s
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-1
-0.5
0
0.5
1
AC current ia(t)
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-10
-5
0
5
10Line-to-line and Line-to-neutral voltage
V o l t s
vab
(t)
van
(t)
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-0.4
-0.2
0
0.2
0.4Capacitor current
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35
2.95
3
3.05
Capacitor voltage
Time (s)
V o l t s
Fig. 4.4 12-pulse behaviour
94
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 104/120
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-0.4
-0.2
0
0.2
0.4
AC current ia(t)
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-10
-5
0
5
10Line-to-line and Line-to-neutral voltage
V o l t s
vab
(t)
van
(t)
0.3 0.302 0.304 0.306 0.308 0.31 0.312 0.314 0.316-0.2
-0.1
0
0.1
0.2Capacitor current
A m p
0.3 0.302 0.304 0.306 0.308 0.31 0.312 0.314 0.316 0.318 0.32
1.46
1.48
1.5
1.52
1.54
1.56Capacitor voltage
Time (s)
V o
l t s
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-0.4
-0.2
0
0.2
0.4
AC current ia(t)
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-10
-5
0
5
10Line-to-line and Line-to-neutral voltage
V o l t s
vab
(t)
van
(t)
0.3 0.302 0.304 0.306 0.308 0.31 0.312 0.314 0.316-0.2
-0.1
0
0.1
0.2Capacitor current
A m p
0.3 0.302 0.304 0.306 0.308 0.31 0.312 0.314 0.316 0.318 0.32
1.46
1.48
1.5
1.52
1.54
1.56Capacitor voltage
Time (s)
V o
l t s
Fig. 4.5 24-pulse behaviour
95
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 105/120
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-0.2
-0.1
0
0.1
0.2
AC current ia(t)
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-10
-5
0
5
10Line-to-line and Line-to-neutral voltage
V o l
t s
vab
(t)
van
(t)
0.3 0.302 0.304 0.306 0.308 0.31 0.312 0.314 0.316-0.1
-0.05
0
0.05
0.1Capacitor current
A m p
0.3 0.302 0.304 0.306 0.308 0.31 0.312 0.314 0.316
0.745
0.75
0.755
Capacitor voltage
Time (s)
V o l t s
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-0.2
-0.1
0
0.1
0.2
AC current ia(t)
A m p
0.3 0.305 0.31 0.315 0.32 0.325 0.33 0.335 0.34 0.345 0.35-10
-5
0
5
10Line-to-line and Line-to-neutral voltage
V o l
t s
vab
(t)
van
(t)
0.3 0.302 0.304 0.306 0.308 0.31 0.312 0.314 0.316-0.1
-0.05
0
0.05
0.1Capacitor current
A m p
0.3 0.302 0.304 0.306 0.308 0.31 0.312 0.314 0.316
0.745
0.75
0.755
Capacitor voltage
Time (s)
V o l t s
Fig. 4.6 48-pulse behaviour
96
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 106/120
system and the VSI controls.
• Calculate AC and DC side harmonic injections generated by the converter switching.
Models based on time averaging theory have been proposed [6, 7]. This section presents the continuous
STATCOM time model at fundamental frequency. The output voltages (line-to-line and line-to-neutral) and
the phase currents are approximated just by their fundamental components; for high pulse-numbered (24-
and 48-pulse) converters it is a good approximation. The capacitor current and voltage are approximated by
its average components.
To develop the model the only assumption is that the switching pattern is symmetrical. The functions
f 0si are the average functions (fundamental frequency) of the switching functions defined previously. The
function f s1 expressed by its Fourier series is given by eq (4.31),
f s1 = a0 +∞X
n=1
(an cos(nωt) + bn sin(nωt)) (4.31)
where
a0 =1
2; an = 0,
bn =2
nπ∀ n odd
so that the fundamental component of f s1 is
f 0s1 =1
2+
2
π
sin(ωt) (4.32)
The expressions for f 0s3 and f 0s5 aresimilar to eq. (4.32) just phase-shift by -120oand-240o, respectively,
that is,
f 0s3 =1
2+
2
πsin(ωt − 120o) (4.33)
f 0s5 =1
2+
2
πsin(ωt − 240o) (4.34)
substituting eqs. (4.32), (4.33), and (4.34) into eq. (4.6) and (4.10) the voltages van(t) and vHn(t) at
fundamental frequency, v0an(t), v0
Hn
(t) are obtained given by eqs. (4.35) and (4.36).
v0Hn(t) = −1
2vDC (t) (4.35)
v0an(t) =2
πvDC (t)sin(ωt) (4.36)
97
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 107/120
Substituting eq. (4.36) into eq. (4.1) yields
d
dt
i0a(t) =
−
R
L
i0a(t)
−
2
πL
v0DC (t)sin(ωt) +1
L
ean(t) (4.37)
similarly for phases ’b’ and ’c’
d
dti0b (t) = −R
Li0b (t) − 2
πLv0DC (t)sin(ωt − 120o) +
1
Lebn(t) (4.38)
d
dti0c (t) = −R
Li0c (t) − 2
πLv0DC (t)sin(ωt − 240o) +
1
Lecn(t) (4.39)
For the capacitor we have that,
i0DC (t) = i0a(t)f 0s1 + i0b (t)f 0s3 + i0c(t)f 0s5 (4.40)
so that
i0DC (t) =2
π
¡i0a(t)sin(ωt) + i0b (t)sin(ωt − 120o) + i0c (t)sin(ωt − 240o)
¢(4.41)
Therefore the last state that complete the model is the voltage capacitor.
d
dtv0DC (t) =
1
C i0DC (t) (4.42)
The set of equations above represent the average behaviour of a six-pulse converter operating at line
frequency switching.
4.2.1 12-pulse converter
A 12-pulse converter is comprised by the combination of two six-pulse converter where the switching
pattern between the two six-pulse converters is phase-shifted by 30o, that is, the switching pattern of the
second converter is lagged by 30orespect to the first one, therefore the switching functions of the second
converter are given by,
f 0s12 =1
2+
2
πsin(ωt − 30o) (4.43)
f 0
s32 =
1
2 +
2
π sin(ωt − 150o
) (4.44)
f 0s52 =1
2+
2
πsin(ωt − 270o) (4.45)
98
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 108/120
The fundamental component of the voltage in the secondary winding (Y side) of the ∆− Y transformer is
v0an(t)∆Y =2
π
v0DC (t)sin(ωt)
therefore the output voltages of a 12-pulse converter at fundamental frequency are
v0an(t)12 =4
πv0DC (t)sin(ωt) (4.46)
v0bn(t)12 =4
πv0DC (t)sin(ωt − 120o) (4.47)
v0cn(t)12 =4
πv0DC (t)sin(ωt − 240o) (4.48)
To calculate the contribution of the second converter in the capacitor current, the fundamental switching
functions are substituted into eq. (4.30), yielding,
i0DC (t)2 =1√
3
¡¡f 0s12 − f 0s32
¢i0a(t) +
¡f 0s32 − f 0s52
¢i0b (t) +
¡f 0s52 − f 0s12
¢i0c(t)
¢(4.49)
so that,
i0DC (t)2 =2
π
¡i0a(t)sin(ωt) + i0b (t)sin(ωt − 120o) + i0c (t)sin(ωt − 240o)
¢(4.50)
Therefore the average capacitor current due to a 12-pulse converter is given by eq. (4.51)
i0
DC (t) =4
π¡
i0
a(t)sin(ωt) + i0
b (t)sin(ωt − 120o
) + i0
c (t)sin(ωt − 240o
)¢
(4.51)
4.2.2 24-pulse converter
The 24-pulse converter is comprised by two 12-pulse converters with a phase-shift of 15o between them
(one of them lagged 7.5oand the other leaded 7.5o). From chapter 3 we know that the 24-pulses line-to-
neutral voltage is given by
v0an(t)24 =4√
3V ab1 sin(ωt) (4.52)
where
V ab1 =2√
3
πv0DC (t),
therefore the 24-pulse line-to-neutral voltages at fundamental frequency are given by the following expres-
99
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 109/120
sions:
v0an(t)24 =8
πv0DC (t)sin(ωt) (4.53)
v0bn(t)24 =8
πv0DC (t)sin(ωt − 120o) (4.54)
v0cn(t)24 =8
πv0DC (t)sin(ωt − 240o) (4.55)
To obtain the capacitor current we have to take into account the contribution of each 12-pulse converter.
When the firing angle of one 12-pulse converter is zero the capacitor current is given by eq. (4.51); in the
24-pulse converter one 12-pulse converter is lagged 7.5oand the other one leaded 7.5o, the DC current of
each 12-pulse converter is given by the expressions (4.56) for the lagging case and (4.57) for the leading
case.
i0DC 12(t)1 =4
π
¡i0a(t)sin(ωt − 7.5o) + i0b (t)sin(ωt − 127.5o) + i0c (t)sin(ωt − 247.5o)
¢(4.56)
i0DC 12(t)2 =4
π
¡i0a(t)sin(ωt + 7.5o) + i0b (t)sin(ωt − 112.5o) + i0c (t)sin(ωt − 232.5o)
¢(4.57)
Thus, the capacitor current in the 24-pulse converter is
i0DC (t)24 = i0DC 12(t)1 + i0DC 12(t)2
so that,
i0DC (t)24 =8
π
¡i0a(t)sin(ωt) + i0b (t)sin(ωt − 120o) + i0c (t)sin(ωt − 240o)
¢(4.58)
4.2.3 48-pulse converter
A similar procedure to the above presented is carried out to obtain the signals of a 48-pulse converter.
For a 48-pulse converter we have that,
v0an(t)48 = 2v0an(t)24 (4.59)
i0DC (t)48 = 2i0DC (t)24 (4.60)
According to the results obtained previously and including, the output voltage angle α or firing angle
as a control variable, that is, the switching pattern applied to transistors, the STATCOM state-space model
100
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 110/120
at fundamental frequency is given by,
·
x (t) = Asx(t) + Bsu(t) (4.61)
where
x(t) =£
i0a(t), i0b (t), i0c (t), v0DC (t)¤T
u(t) = [ean(t), ebn(t), ecn(t)]T
As =
−R
L0 0 −k1 sin(ωt + α)
0 −R
L0 −k1 sin(ωt + α− 120o)
0 0 −R
L−k1 sin(ωt + α− 240o)
k2 sin(ωt + α) k2 sin(ωt + α− 120o) k2 sin(ωt + α− 240o) 0
Bs =
1L 0 0
01
L0
0 01
L
k1 =2
πL; k2 =
2
πC For a six-pulse converter
k1 =4
πL; k2 =
4
πC For a 12-pulse converter
k1 =8
πL; k2 =
8
πC For a 24-pulse converter
k1 = 16πL
; k2 = 16πC
For a 48-pulse converter
The model given by eq. (4.61) although appears to be linear, it is nonlinear. The nonlinearity of the
STATCOM is manifested by the inclusion of the control angle α on the state equation. Changes in the
control angle α will results in nonlinear responses in the STATCOM states [8].
4.3 dq0 Reference frame model
The Park’s transformation is used to develop the time-invariant form of eq. (4.61) in the rotating
reference frames (dq0). The Park’s transformation expressions used are given by the follows relationships
[9].
X dq 0 = T X abc (4.62)
101
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 111/120
X abc = T −1X dq 0 (4.63)
T =
cos(ωt) cos (ωt − 120o) cos (ωt− 240o)
− sin(ωt) − sin(ωt − 120o
) − sin(ωt − 240o
)1
2
1
2
1
2
(4.64)
The application of the Park’s transformation to an ’abc’ system eq. (4.65) yields to its equivalent in the
dq0 reference frame eq. (4.66).
·
x (t)abc = Ax(t)abc + Bu(t)abc (4.65)
·
xdq 0= T AT −1xdq 0 + T BT −1udq 0 + ωxqd0 (4.66)
Applying the Park’s transformation to the fundamental frequency model eq. (4.61) of the STATCOM
yields "·
idq 0·
vDC
#=
·T 00 1
¸" ·
iabc·
vDC
#+ ω
·iqd0
0
¸(4.67)
where: iqd0 = [iq ,−id, 0]T , so that"·
idq 0·
vDC
#=
·T As11T −1 T As12
As21T −1 As22
¸ ·idq 0
vDC
¸+
·T BsT −1 0
0 0
¸ ·edq 0
0
¸+ ω
·iqd0
0
¸(4.68)
where:
T As11T −1
= diagonal½−R
L¾
; T As12 = [−k1 sin(α), k1 cos(α) , 0]T
;
As21T −1 =
·3
2k2 sin(α) ,−3
2k2 cos(α) , 0
¸
As22 = [0] ; T BsT −1 = diagonal
½1
L
¾Finally the STATCOM model in the dq0 reference frame can be represented by
·
xdq 0= Adq 0xdq 0 + Bdq 0udq 0 (4.69)
where:
xdq 0 = [id, iq , i0, vDC ]T
udq 0 = [ed, eq , e0, 0]T
102
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 112/120
Adq 0 =
−R
Lω 0 −k1 sin(α)
−ω −R
L0 k1 cos(α)
0 0 −RL
0
3
2k2 sin(α) −3
2k2 cos(α) 0 0
Bdq 0 =
1
L0 0 0
01
L0 0
0 01
L0
0 0 01
C
Expressing in other form
ddt
id = −RL
id − k1 sin(α) vDC + 1L
ed + ωiq
d
dtiq = −R
Liq + k1 cos(α) vDC +
1
Leq − ωid (4.70)
d
dtvDC =
3
2k2 [id sin(α) − iq cos(α)]
The states of the model are the dq0 components of the phase current and the voltage capacitor; the
inputs are the dq0 components of AC voltage where the STATCOM is connected.
Figures Fig.4.7-Fig.4.9show the response of each model, the fundamental frequency model and the dq0
model. Figures depict the response of a 48-pulse STATCOM using the parameters previously mentioned.
The converter is initially at steady state with α = −6o, at time t = 0.3 s the angle α was stepped from -6oto
-9o. It should be noted that the model at fundamental frequency and the dq0 model do not take into account
the switching effects; this can be seen in the capacitor voltage, Fig.4.9.
4.4 Conclusions
Independent of the application, modelling for power system analysis is typically performed using an
approximate continuous time representation of the converter in the synchronous reference frame (dq0).
The continuous time model of the VSI yields a set of three differential equations which represent the VSI
operation with reasonable accuracy under most conditions, but these models present some limitations such
103
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 113/120
0.1 0.15 0. 2 0 .25 0.3 0. 35 0. 4 0.45 0.5 0.55 0.6
-0.2
-0.1
0
0. 1
0. 2
0. 3
(a)
A m p
Switch ing mode l
dq0 model
0.1 0.15 0. 2 0 .25 0.3 0. 35 0. 4 0.45 0.5 0.55 0.6
-0.2
-0.1
0
0. 1
0. 2
0. 3
(b)Time (s)
A m p
Fig. 4.7 Phase current; a)Switching model; b)model at fundamental frequency.
104
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 114/120
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.60.15
0.2
0.25
id
component
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6-0.04
-0.035
-0.03
-0.025
-0.02
-0.015
Time (s)
iq component
Fig. 4.8 id and iq current
105
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 115/120
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.60.7
0.75
0.8
0.85
0.9
0.95
Time (s)
V o l t s
dq0 model
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.60.7
0.75
0.8
0.85
0.9
0.95
V o l t s
Switching model
Fig. 4.9 Capacitor voltage
106
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 116/120
as the inability to represent the switching effects.
Switching models describe very well the dynamic of the VSI taking into account the switching effects
but these kinds of models are very complicated and hard to work with linear theory, although there are afew methods and tools in the theory of nonlinear dynamics that have proved very useful in studying the
behaviour of power electronic converters.
107
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 117/120
References
[1] Seyed Ali Nabavi-Niaki, ”Modelling and applications of unified power flow controller (UPFC) for
power systems”, Ph. D. Thesis, University of Toronto, 1996.
[2] Banerjee, Verghese, ”Nonlinear Phenomena in power electronics”, IEEE Press, 2001.
[3] J. Kassakian, M. Schlecht, and G. Verghese, ”Principles of power electronics”, Addison-Wesley,
1991.
[4] IEEE, ”Special Issue on hybrid systems”, IEEE Trans. on Automatic Control, May 1998
[5] P. W. Lehn, ”Exact modeling of the Voltage Source Converter”, IEEE Trans. on Power Delivery, Vol.
17, No. 1, pp. 217-222, January 2002
[6] P. W. Lehn, M. R. Iravani, ”Experimental Evaluation of STATCOM Closed Loop”, IEEE Trans. on
Power Delivery, Vol. 13, No. 4, pp. 1378-1384, October 1998.
[7] K. R. Padiyar, A. M. Kulkarni, ”Design of reactive current andvoltage controller of static condenser”,
Electrical Power and Energy Systems, Vol.19, No. 6, pp. 397-410, 1997.
[8] Pranesh Rao, M. L. Crow, Zhiping Yang, ”STATCOM Control for power system voltage control
applications”, IEEE Trans. on Power Delivery, Vol 15, No. 4, pp. 1311-1317, October 2000.
[9] P. Kundur, ”Power System Stability and Control”, McGraw Hill, 1994.
108
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 118/120
Conclusions
The STATCOM and other FACTS devices have been widely studied by analytical models but the physi-
cal functionality is unknown for a lot of power researches. Understand the physical functionality and present
clearly the design and development of a real three-phase STATCOM prototype laboratory is the biggest mo-
tivation of this work.
Chapter 2 explains the relevant details of the voltage source-inverter (VSI), the building block of a
STATCOM and other FACTS devices. A detailed analysis is carried out to deduce the expressions for
the AC (phase currents and output voltage) and DC signals (capacitor current and voltage) of a 6- and
12-pulse STATCOM. The resulting expressions allow to approximate an adequate DC capacitor value. A
VSI generates an AC voltage from a DC voltage, when a VSI is used as a typical STATCOM a fixed
DC capacitor is used as a DC source. An important questions about that are answered: how the energy
conversion is achieved using a DC capacitor?, and if a DC source is not used, how the capacitor voltage
is kept fixed?. Such questions are answered establishing that, when the compensator (neglecting losses) is
exchanging reactive power only, the average power across the DC capacitor is zero, therefore the capacitor
does not change its voltage.
The control mechanism to regulate the capacitor voltage and therefore, the control of the active and
reactive power is demonstrated using the active power exchange to decrease/increase the capacitor voltage,
controlling the switching angle pattern of the semiconductor devices.The simplest three-phase converter is the six-pulse converter however even at low rating a simple six-
pulse converter is unlikely in high power applications due to its harmonic distortion. To reduce the harmonic
distortion to an acceptable level the pulse order must be increased to 24- or 48-pulse. To get the 24- and
48-pulse converter complex phase-shifting transformers are needed but an approximate behaviour can be
obtained using simple 12-pulse transformers; two 12-pulse transformers having relative phase displacements
of 15, quasi 24-pulse operation, and four 12-pulse transformers having 7.5, quasi 48-pulse operation. The
quasi 48-pulse operation allows a small circulation of 12- and 24- pulse order harmonic due to the 12-pulse
harmonics which are characteristic of each 12-pulse converter are not perfectly cancelled, but the residual
magnitude is acceptably small.
In power system analysis, modelling is typically performed using an approximate continuous time rep-
resentation of the converter in the synchronous reference frame (dq0). The continuous time model of the VSI
109
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 119/120
yields a set of three differential equations which represent the VSI operation with reasonable accuracy under
most conditions, but these models present some limitations such as the inability to represent the switching
effects. Switching models describe very well the dynamic of the VSI taking into account the switching ef-fects but these kinds of models are very complicated and hard to work with linear theory, although there are
a few methods and tools in the theory of nonlinear dynamics that have proved very useful in studying the
behaviour of power electronic converters.
110
7/29/2019 Pre Doctoral
http://slidepdf.com/reader/full/pre-doctoral 120/120
Future work
Using this theoretical analysis an IGBT-12-pulse-VSI STATCOM will be developed first in digital sim-
ulations using the EMTDC/PSCAD software and then as a laboratory prototype. The prototype is comprised
by the power part (IGBT-12-pulse-VSI) and the control part (internal control and external control).
The internal control is an integral part of the converter. Its main function is to operate the converter
power switches so as to generate a fundamental output voltage waveform with the needed magnitude and
phase angle in synchronism with the AC system that forces the reactive power exchange required for com-
pensation.
The internal control is comprised by the IGBT-gate driver and by a microcontroller device that compute
the magnitude and phase angle of the required output voltage from the reactive and/or real power reference
signals, generating a set of coordinated gating pattern which determines the ”on” and ”off” periods of each
powerdevices. The reference signalsare provided by the external control. The external control is determined
by the functional operation of the STATCOM, that is, the final application: power flow control, loop-flow
control, voltage regulation, enhancement of transient stability or damping of power oscillations.
Recommended