Voltage Fluctuation in Industrial Network and Compensation Measures

UNIVERSITY OF LJ UBLJ ANA
Faculty of Electrical Engineering

Ljubia Spasojevi

VOLTAGE FLUCTUATION IN INDUSTRIAL
NETWORK AND COMPENSATION
MEASURES

DOCTORAL DISSERTATION

Ljubljana, 2014

UNIVERSITY OF LJ UBLJ ANA

Faculty of Electrical Engineering

Ljubia Spasojevi

VOLTAGE FLUCTUATION IN INDUSTRIAL
NETWORK AND COMPENSATION
MEASURES

DOCTORAL DISSERTATION

Mentor: prof. dr. Igor Papi

Ljubljana 2014

UNIVERZA V LJ UBLJ ANI

Fakulteta za elektrotehniko

Ljubia Spasojevi

KOLEBANJE NAPETOSTI V
INDUSTRIJSKEM OMREJU IN
KOMPENZACIJSKI UKREPI

DOKTORSKA DISERTACIJ A

Mentor: prof. dr. Igor Papi

Ljubljana, 2014

To my father Milan and my mother Zora

Acknowledgements

It would not have been possible to finish this thesis without contributions and support
from many people.

I would like to sincerely thank prof. dr. Igor Papi for his generous support, new ideas and
expert advice, which were very useful for my research and the development of this doctoral
dissertation.

Thank you to all my colleagues (Leopold Herman, Milo Maksi, Mitja Prelovek, Marko
Kolenc, Matej Kobav) from the Laboratory of Electricity Networks and Devices at the
Faculty of Electrical Engineering, who gave me a pleasant working environment, support
and assistance every time that I needed it, while I would especially like to thank dr. Botjan
Blai for numerous occasions of guidance and assistance when writing articles.

I would like to thank my father Milan, my mother Zora and my girlfriend Aleksandra for
endless support during the study.

Finally, special thanks go to Mr Bojan Seni, the company Store Steel d.o.o. and the agency
SPIRIT, without whose financial support I would not be able to finish my doctoral
dissertation.

Operation partly financed by the European Union, European Social Fund. Operation
implemented in the framework of the Operational Programme for Human Resources
Development for the Period 2007-2013, Priority axis 1: Promoting entrepreneurship and
adaptability, Main type of activity 1.1.: Experts and researchers for competitive enterprises.

Ljubljana, July 2014. Spasojevi Ljubia

Ljubia Spasojevi Doctoral dissertation
Content

LIST OF ABBREVIATIONS AND SYMBOLS............................................................................................................... 11
ABSTRACT ............................................................................................................................................................ 17
RAZIRJENI POVZETEK .......................................................................................................................................... 21
1. INTRODUCTION ........................................................................................................................................... 29
1.1 SUBJECT OF THE DOCTORAL DISSERTATION .............................................................................................................. 30
1.2 CONTRIBUTION TO SCIENCE .................................................................................................................................. 31
2 POWER QUALITY ......................................................................................................................................... 33
2.1 STANDARD EN 50160 ....................................................................................................................................... 33
Frequency of the supply voltage ........................................................................................................................ 34
Declared supply voltage .................................................................................................................................... 34
Voltage deviation .............................................................................................................................................. 34
Voltage dip ........................................................................................................................................................ 35
Interruption of voltage supply ........................................................................................................................... 36
Temporary overvoltage between live conductors and earth ............................................................................. 36
Transient overvoltage between live conductors and earth ............................................................................... 37
Voltage unbalance ............................................................................................................................................. 38
Rapid voltage fluctuations ................................................................................................................................. 39
3 FLICKER ....................................................................................................................................................... 41
3.1 OCCURRENCE OF FLICKER AND FLICKER SOURCES ...................................................................................................... 42
Occurrence of flicker .......................................................................................................................................... 43
Arc furnace ........................................................................................................................................................ 45
Welding machine ............................................................................................................................................... 46
Wind turbines .................................................................................................................................................... 46
3.2 ESTIMATE OF THE FLICKER LEVEL ........................................................................................................................... 46
3.3 FLICKERMETER .................................................................................................................................................. 48
Block 1 input voltage adapter ......................................................................................................................... 49
Block 2 - squaring multiplier .............................................................................................................................. 49
Block 3 filters .................................................................................................................................................. 50
Block 4 - non-linear variance estimator ............................................................................................................. 51
Block 5 - statistical calculation block ................................................................................................................. 52
4 ELECTRIC ARC FURNACE .............................................................................................................................. 55
4.1 FURNACE WITH A DIRECT ELECTRICAL ARC ............................................................................................................... 55
4.2 INDIRECT ARC FURNACES ..................................................................................................................................... 57
4.3 FURNACE WITH A SUBMERGED ARC ........................................................................................................................ 57
4.4 ELECTRIC ARC .................................................................................................................................................... 58
Electric arc of direct current .............................................................................................................................. 59
Electric arc of alternating current ...................................................................................................................... 62
5 DYNAMIC REACTIVE-POWER COMPENSATION ............................................................................................ 65
5.1 FLEXIBLE ALTERNATING-CURRENT TRANSMISSION SYSTEM .......................................................................................... 65
5.2 STATIC VAR COMPENSATOR ................................................................................................................................ 67
FC-TCR structure ................................................................................................................................................ 67
6 THE MODELLING OF ELECTRIC ARC FURNACES ............................................................................................ 73
6.1 REPRESENTATIVE SAMPLES OF VOLTAGE AND CURRENT ............................................................................................. 74
The Selection of Representative Samples .......................................................................................................... 75
6.2 MODEL OF THE FLICKERMETER .............................................................................................................................. 80
Testing of S(t) ..................................................................................................................................................... 80

9
Testing of the P
st
................................................................................................................................................ 82
6.3 BASIC CONCEPT OF THE MODEL OF THE ELECTRIC ARC FURNACE ................................................................................ 82
System Configuration ........................................................................................................................................ 84
Modeling the Electrical Arc Length.................................................................................................................... 86
Calculation of the Voltage Envelope ................................................................................................................. 87
6.4 THE RESULTS OF THE SIMULATION ........................................................................................................................ 88
7 DEVELOPING THE CONTROL ALGORITHM ................................................................................................... 95
7.1 VECTOR ILLUSTRATION OF THREE-PHASE QUANTITIES ................................................................................................ 95
Instantaneous active and reactive power ......................................................................................................... 98
7.2 COMPENSATION PRINCIPLE AND MATHEMATICAL MODEL OF TCR .............................................................................. 100
7.3 CONTROLLER DESIGN ........................................................................................................................................ 103
Transfer function of the P controller ............................................................................................................... 105
7.4 MODEL TESTING .............................................................................................................................................. 107
The static analysis .......................................................................................................................................... 107
The dynamic analysis ...................................................................................................................................... 109
7.5 THE SIMULATION RESULTS ................................................................................................................................. 111
The system configuration ................................................................................................................................ 112
7.6 MODELLING OF THE ELECTRIC-ARC LENGTH USING SINUSOIDAL FUNCTION ................................................................... 114
Simulations without a connected SVC ............................................................................................................. 114
Simulations with connected SVC ..................................................................................................................... 120
Reactive power and flicker-regulation mode .................................................................................................. 120
Voltage-regulation mode ................................................................................................................................ 127
7.7 MODELLING OF THE ELECTRIC-ARC LENGTH USING A SIGNAL THAT INVOLVES THE ALL-IMPORTANT FREQUENCIES FOR FLICKER 130
Simulations without the connected SVC .......................................................................................................... 130
Reactive power and flicker regulation mode ................................................................................................... 135
Voltage-regulation mode ................................................................................................................................ 139
7.8 SIMULATION WITH ANOTHER FREQUENCY SPECTRUM .............................................................................................. 141
7.9 ASSESSMENT OF THE CONTROLLER CHARACTERISTICS .............................................................................................. 142
8 CONCLUSION ............................................................................................................................................ 145
9 REFERENCES: ............................................................................................................................................ 147

10
List of abbreviations and symbols

List of used abbreviations

AC Alternating current
BESS Battery energy storage system
DC Direct current
DFT Discrete Fourier transformation
EAF Electric arc furnace
EMC Electromagnetic compatibility
FACTS Flexible Alternating Current Transmission System
FC Fixed-capacitor
FC-TCR Fixed-capacitor Thyristor Controlled Reactor
FFT Fast Fourier transformation
IPFC Interline Power Flow Controller
MV Medium-voltage
PCC Point of common coupling
SRCS Synchronous rotation coordinate system
SSG Static Synchronous Generator
SSSC Static Synchronous Series Capacitor
STATCOM Static Synchronous Compensator
SVC Static VAr compensator
TCPST Thyristor-controlled phase-shifting transformers
TCR Thyristor Controlled Reactor
TCR Thyristor Controlled Reactor
TCSC Thyristor Controlled Series Capacitor
TCSR Thyristor Controlled Series Reactor
TSSC Thyristor Switched Series Capacitor
TSSR Thyristor Switched Series Reactor
UPFC Unified Power Flow Controller

Symbols used in section 3

fb Basic frequency of the voltage
fm Frequency of modulation
m Factor of modulation
Pinst Instantaneous flicker sensation
Pinst, max Maximum instantaneous flicker sensation
Plt Long-term flicker index
Pst Short-term flicker index
t Time
U(t) Basic amplitude of the voltage

11
Angular frequency


B Correction factor due to the Schottky effect
A, B, C, D Constants
Earc Electric field in the arc
f Frequency
Iarc RMS value of arc current
iarc Waveforms of the arc current
J Emission density
l Arc length
M Frist characteristic constant that depend on type of material
N Second characteristic constant that depend on type of material
T Temperature of cathode
Uarc RMS value of arc voltage
uarc Waveforms of the arc voltage
U-I Voltage-Current characteristic


BTCR Reactive admittance (susceptance) of the TCR
cos Power factor
i(t) Instantaneous current
IFC Root-mean-square values of constant capacitive reactive current
Irms Root-mean-square values of i
ISVC Root-mean-square values of total reactive power current
ITCR Variable inductive reactive current
L Inductance of the thyristor-controlled reactor
U Peak value of the applied voltage
u(t) Instantaneous voltage
Urms Root-mean-square values of u
uS (t) Instantaneous source voltage
Thyristor firing angle
Angular frequency of the supply voltage


u Voltage drop
A Sum of the anode and cathode voltage drops
Am Amplitude
B Voltage drop per unit of the arc length

12
C, D Constants
F(f, A, ) Function for control of the variable voltage source
fs Sampling frequency
G Gain
iA Instantaneous value of arc current
iRS Representative sample of the current
k Ratio between the arc voltages threshold value UAT(l) at length l, and
UAT(l0) at the reference length l0
l Length of the arc
l0 Reference length of the arc
LC Cable inductance
Lf Furnace inductance
LLSC Short-circuit inductance
p(t)
Characteristic signal with frequencies from 1 Hz to 35 Hz of electric
arc furnace
p'(t)
Characteristic signal of electric arc furnace, fingerprint of electric arc
furnace
P1 Active power of the selected representative sample
Pm Mean values of the active power of the ten-minute measured signal
Ps Active power of the one-second sample
Psim Active power obtained by simulation
Pst Flicker level of the ten-minute measured signal
Pst, sim Flicker level obtained by simulation
Pst, s Flicker levels of one-second sample
Pst,1 Flicker levels of the selected representative sample
Q1 Reactive power of the selected representative sample
Qm Mean values of the reactive power of the ten-minute measured signal
Qs Reactive power of the one-second sample
Qsim Reactive power obtained by simulation
r(t) Law of the arc-length variation
RC Cable resistance
Req Equivalent resistance
Rf Furnace resistance
RLSC Short-circuit resistance
S(t) Instantaneous value of flicker
Ssc Short-circuit power at the 110-kv bus
T Total simulation time
uA Instantaneous value of arc voltage
uA Voltage at the point of the arc furnaces connection to the network
uA0 Arc voltage corresponding to the reference length of the arc
UAT The threshold value to which the voltage tends for the reference
length of the arc

13
uRS Representative sample of the voltage
Xeq Equivalent reactance
Phase displacement of interharmonic


Firing angle
b Variable factor
BTCR Susceptance of the TCR
cos Power factor
dq0, dq Direct quadrature zero transformations
F Control function of the controllable voltage sources
F(s)tr Transfer functions of the system
i
components of the current

i
components of the current

ia_TCR, ib_TCR,
ic_TCR
Currents through the reactors in the star equivalent scheme
iab, ibc, ica Currents through the reactors in the delta equivalent scheme
iB Base values of the current
id d components of the current
id d components of the current that flows into the TCR in p.u.
iFC Currents of FC
il Load current
IL RMS value of the current through the reactor
iq q components of the current
iq q components of the current that flows into the TCR in p.u.
iq
*
Reference current
iqFC Reactive components of the FC current
iql Reactive component of the load current
iqs Reactive component of the system current
iqs
*

Measured value of the reactive component of the total system current
iqs
iqSVC Reactive components of the current of SVC device
iqTCR Reactive component of the TCR current
is System current
isa, isb, isc System currents, currents from the power system
iSVC Current of the SVC device
iTCR Currents of TCR
Kp Gain of P controller
L Inductance of the TCR
La, Lb, Lc Inductance of inductor in star configuration
Lab, Lbc, Lca Inductance of inductor in delta configuration

14
Lf inductance of the cable line between the furnace transformer and the
furnace electrodes
LLSC Short-circuit inductance
p Instantaneous active power in dq rotation coordinate system
P Proportional controller
PI Proportionalintegral controller
Pst Short-term flicker index
q Instantaneous reactive power in dq rotation coordinate system
qins Instantaneous values of the reactive power
Ra, Rb, Rc Resistance of resistor in star configuration
Rab, Rbc, Rca Resistance of resistor in delta configuration
Rf Resistance of the cable line between the furnace transformer and the
furnace electrodes
RLSC Short-circuit resistance
Ssc Short-circuit power at the 110-kV bus
T Total simulation time
Tdq Transformation matrix for dq transformations
TTCR Response time of the TCR
T Transformation matrix for transformations
U RMS value of the connected voltage
u
components of the voltage

u
components of the voltage

ua, ub, uc, Phase-to-neutral voltages at the point of common coupling
uB Base values of the voltage
ud d components of the voltage
ud d components of the voltage in p.u.
ud d components of the measured voltage at the SVC connection point
uMV Voltage at the MV level
uq q components of the voltage
uq q components of the voltage in p.u.
uq q components of the measured voltage at the SVC connection point
URMS RMS voltage at the MV level
XL Maximum reactance of the TCR
ZB Base values of the impedance
0, transformations
Angle between the voltage vector and the d-axes of the dq SRCS
Angular speed of the rotation coordinate system
b Synchronous angular speed of the fundamental network component

15

Abstract

The main theme of this doctoral dissertation is the compensation of the negative impacts of
electric arc furnaces (EAFs) in a power system. Is generally known that an EAF represents
a large consumer of electrical energy, which because of its nonlinear characteristics has a
strong feedback influence on the power quality in the electric power system, and as such
requires the installation of compensation devices. The development and design of
compensation devices for the EAF is a major problem because of the non-linear electrical
characteristics of the EAF. In the development and design process of compensating devices
an important role is played the simulation model of the EAF. In order to be able to develop
a cost-effective and at the same time very efficient compensating device we need to have
the most accurate simulation model of the EAF. The key contributions to science of this
doctoral dissertation are from the fields of the development and design of a real model of
an EAF and a control algorithm for the compensation device in order to eliminate the
negative impacts of the EAF on the power system.

This dissertation can be divided into two parts: first, is the theoretical part and, second, is
the research-development part of the work. The theoretical part includes the first 5
chapters and in it is a theoretical presentation of the existing situation and issues, while the
research-development part of this work is included in the last three chapters, and in it is a
presentation of the results of the research work and the solutions for some existing issues.

In Chapter 2 is the basic structure of the power-quality concept (what the term 'power
quality' means and why it is necessary to define it). On the basis of the standard EN 50160
thirteen basic parameters are defined that are essential to the power quality. These
parameters will be respected by consumers and by suppliers. For all 13 defined parameters
the minimum and maximum allowable values that must be followed are presented.

Chapter 3 focuses on flicker. In this chapter, one of 13 pre-defined parameters of power
quality is explained in detail. In this chapter we present the causes of flicker, how flicker
appears and how it is transferred from its source to the human eye. To measure the
intensity of flicker in a power system a Flickermeter is used. Using a flickermeter it is
possible to measure the instantaneous intensity of the flicker in the power network. From
further statistical processing of the instantaneous value of the flicker the flicker severity
can be obtained. Depending on how long the period of time is taken to calculate, there are
two different flicker severities, i.e., the index for long-term flicker and the index for short-
term of flicker.

Chapter 4 introduces the theory of electric arc furnaces. These EAFs transform electrical
energy into heat by means of an electric arc. Their use in the past few decades is increasing
rapidly. Depending on the way the steel is melted, furnaces can be divided into three main
groups: furnace with a direct electrical arc, furnace with an indirect and electrical arc

17
furnace with a submerged arc. Direct arc furnaces melt the content of a furnace in such a
way that the electrical arc is established between the graphite electrodes and the batches
that gradually heat up and melt. In the indirect arc furnaces the melting is effected by the
arcing between two horizontally opposed carbon electrodes or graphite electrodes. In
furnaces with a submerged arc the electrodes are directly put into the molten alloy. The
electric arcs are established in the gaseous area within the slag. An electric arc is the
electrical breakdown of a gas that produces an on-going plasma discharge, resulting from a
current through normally non-conductive media. It is characterized by non-linear voltage-
current characteristics and a high current density.

Chapter 5 deals with reactive power. This reactive power is necessary for the proper work
of certain consumers, but its transfer to larger distances increases the stress on the
transmission capacity of the power system. Furthermore, the transfer of reactive power at
longer distance in a power system increases the losses and reduces the power factor. For
this reason there is a need for reactive power compensation in the place of consumption.
Electrical loads depending on their electrical characteristics can be consumers or
producers reactive energy. The development of a new technology has enabled the
development of compensation devices that at the same time can be both consumers and
producers of reactive power. The flexible alternating current transmission system (FACTS)
are compensating devices based upon power electronics that are capable of absorbing or
generating reactive power. This chapter gives an overview of FACTS devices with special
emphasis on the Static VAr compensator (SVC).

Chapter 6 deals with the modelling of realistic models of the EAF. There are several
different methods for modelling the arc of the EAF. Usually, the arc of an EAF is modelled as
a controlled voltage source. For modelling the time-varying characteristics of an electric
arc two approaches are used: the stochastic and the deterministic approaches. The realistic
model of the EAF proposed in this thesis is based on representative samples of voltage and
current, which are one-second or one-minute intervals extracted from longer voltage (or
current) intervals. On the basis of the voltage envelope of selected representative samples
the time-varying characteristics of the arc are defined. The new model of EAF was
implemented in the real model of the power system and on that occasion a number of
simulations and testings was made. In addition to the new model of the EAF in this section
a complete analysis of the data obtained from simulations of the new model EAF is
presented.

Chapter 7 shows the complete mathematical derivation and development of a new
controller for SVC. This mathematical derivation is based on a simplified mathematical
model of the SVC in the dq coordinate system. Based on the parameters of a simplified
mathematical model of SVC controller the parameters are determined. Also, during the
mathematical development and the testing of the reliability the maximum frequencies of
the thyristors was taken into account. A model of the SVC with the new controller was

18
implemented on a realistic simulation model of the steel factory and a large number of
simulations were made.

Keywords: arc furnace, flicker, interharmonics, load modelling, non-linear system,
representative voltage samples control algorithm, SVC, TCR, voltage
fluctuation,

19

Razirjeni povzetek

Kolebanje napetosti v industrijskem omreju in kompenzacijski ukrepi

Elektrina energija je najpogosteje uporabljena in najbolj razirjena oblika energije, ker jo
je mogoe enostavno pretvoriti v druge oblike energije. Znailnosti elektrine energije
(razen amplitude in frekvence napetosti) so bile standardizirane zelo pozno. Tradicionalno
se je mislilo, da je kakovost elektrine energije preprosto njena zanesljivost, z drugimi
besedami, neobstoj stalne prekinitve v oskrbi z elektrino energijo. Vendar pa to ni ve
tako. V sodobnem elektrinem omreju je sprejemanje kakovosti elektrine energije
odvisno od pomembnih fizikalnih lastnosti dobavljene napetosti. V zvezi s tem, so problemi
s kontinuiteto dobave veinoma reeni, v fazi nartovanja in gradnje elektrinega omreja,
medtem ko je fizini problem kakovosti napetosti tesno povezan z njeno uporabo.

Obstoj velikega tevila nelinearnih bremen na distribucijskih omrejih vodi do tevilnih
negativnih uinkov. Ti negativni uinki preko omreja vplivajo na vse druge uporabnike v
omreju. Skupni interes proizvajalcev in porabnikov elektrine energije je zanesljiva, varna
in visoko kakovostna oskrba z elektrino energijo. Prevladujoi vplivi na kakovost
elektrine energije prihajajo iz tako imenovanih nelinearnih potronikov, kot so naprave
energetske elektronike, elektrini stroji, elektrooblone pei, itd. Popaenje kakovosti
elektrine energije pomeni kritev osnovnih parametrov napetosti v ustaljenem oziroma
prehodnom stanju in deformacijo valovnih oblik.

Glede na trno usmerjenost, poveanje porabe in vse veje povpraevanje po prenosu
elektrine energije, so obstojei energetski sistemi vedno bolj obremenjeni in zanejo delati
na meji lastne stabilnosti. Zato obstaja poveana nevarnost delnega ali popolnega zloma
sistema, kar se kae v niji kakovosti dobavljene elektrine energije za potronike. Problem
pomanjkanja prenosnih zmogljivosti in nizkega pretoka energije v elektroenergetski sistem
je mogoe reiti z dodajanjem novih daljnovodov ali z zamenjavo obstojeega daljnovoda z
daljnovodom, ki ima vejo kapaciteto (kar pomeni vejo zmogljivost prenosa) . Te reitve
so zanesljive, vendar je njihova izvedba drag in dolgotrajen proces.

En alternativni pristop k reevanju teh teav je uporaba FACTS naprav. FACTS tehnologija
ni le ena naprava, ki rei obstojee teave v energetskem sistemu, temve skup naprav, ki
se lahko uporabljajo samostojno ali v sodelovanju z drugimi napravami za upravljanje
enega ali ve parametrov sistema.

Z uporabo FACTS naprav je mogoe dosei bolji nadzor pretoka energije skozi prenosni
vod brez ogroanja toplotne meje daljnovoda, z minimalnimi izgubami in vejo stabilnostjo
meja. Delovni princip FACTS naprav temelji na elementih energetske elektronike, ki
omogoajo bolje in enostavneje rokovanje, kot tudi hitrejo in bolj zanesljivo delovanje
naprav FACTS. Zaradi teh lastnosti, so naprave FACTS zelo prirono sredstvo za raziritev

21
stabilnosti sistema v primerih motnje. V bistvu, naprave FACTS igrajo kljuno vlogo pri
uinkoviti in ekonomini proizvodnji in prenosu elektrine energije v prihodnosti.

Z liberalizacijo energetskih trgov, je elektrina energija postala trno blago kot katerikoli
drug proizvod, in takna mora izpolnjevati ustrezne standarde kakovosti. Prav tako je treba
ohraniti doloeno raven kakovosti napetosti do omreja za zagotovitev pravilnega
delovanja prikljuene opreme. V svetu energetskih omreij, morajo oblikovalci in
proizvajalci elektrinih naprav, izpolnjevati celo vrsto standardov in priporoil na podroju
kakovosti elektrine energije. Vsak uporabnik omreja (potronik ali proizvajalec
elektrine energije) mora omejiti negativen vpliv lastne opreme na kakovost napetosti
omreja glede na vnaprej dogovorjeni nivo.

Osrednja tema te doktorske disertacije je kompenzacija negativnih vplivov elektroblonih
pei v elektroenergetskem sistemu. Splono je znano, da elektroblone pei predstavljajo
velike porabnike elektrine energije, ki imajo zaradi svojih nelinearnih znailnosti mone
povratne vplive na kakovost elektrine energije v elektroenergetskem sistemu in kot
takne zahtevajo vgradnjo kompenzacijskih naprav. Razvoj in nartovanje kompenzacijskih
naprav za elektroblone pei je velik problem, zaradi njenih nelinearnih elektrinih
karakteristik. Pri procesu razvoja in nartovanja kompenzacijskih naprav ima pomembno
vlogo simulacijski model elektroblone pei. Da bi sploh lahko razvili in nartovali
strokovno, uinkovito in hkrati zelo uinkovito kompenzacijsko napravo, moramo imeti
karseda natanen simulacijski model elektroblone pei.

To doktorsko disertacijo lahko razdelimo na dva dela: prvi del teoretini in drugi
raziskovalno-razvojni del. Teoretini del zajema prvih 5 poglavij in v njem so teoretino
predstavljene obstojee razmere in problematika, medtem ko raziskovalno-razvojni del
tega dela zajema zadnja tri poglavja, in v njem so predstavljeni rezultati raziskovalnega
dela in reitve za nekatera obstojea vpraanja.

V drugem poglavju lahko vidimo osnovno strukturo koncepta kakovosti elektrine energije
(kar pomeni termin "kakovost elektrine energije" in zakaj ga je potrebno opredeliti). V
veini evropskih drav imajo inkorporirani standard EN 50160 "Voltage characteristics of
electricity supplied by public distribution systems" ki jih je leta 1994 sprejel Evropski
odbor za standardizacijo v elektrotehniki. Standard EN 50160 daje kvantitativne
znailnosti kakovosti napetosti pri normalnem stanju delovanja. Obdobje merjenja, ki je
doloena s standardom EN 50160, je sedem dni brez prekinitve. Merilni posnetki, ki jih
spremljamo za vsak parameter so deset-minutni intervali, razen za frekvenco, ki jo je bilo
opaziti v asovni razini deset sekund. Slovenija je kot lanica EU sprejela evropsko
Direktivo o elektromagnetni zdruljivosti in jo preslikala v pravilnik o elektromagnetni
zdruljivosti (EMC). Na podlagi standarda EN 50160 je doloeno in predstavljeno 13
osnovnih parametrov, ki imajo bistven pomen elektrine energije. Te parametre morajo
spotovati potroniki in dobavitelji. Za vseh 13 doloenih parametrov so podane njihove

22
minimalne in maksimalne dovoljene vrednosti, ki jih je treba upotevati. V preglednici A so
podane vrednosti za naslednje parametre.

omrena frekvenca,
velikost napetosti,
odkloni napetosti,
hitre napetostne spremembe,
upadi napetosti,
porasti napetosti,
kratkotrajne prekinitve napetosti,
dolgotrajne prekinitve napetosti,
prehodne prenapetosti,
neravnoteje,
harmonska napetost,
medharmonska napetost,
signalna napetost.

V tretjem poglavju je natanno razloen pojav nastanka flikerja, od povzroiteljev
napetostnega kolebanja v omreju do odziva svetlobnega vira na napetostno kolebanje in
konno do odziva loveka na nihanje svetlobnega toka svetil.

Fliker je definiran kot vtis nestalnosti vidnega zaznavanja zaradi svetlobnega draljaja,
katerega svetlost ali spektralna porazdelitev asovno niha. Ta nihanja osvetljenosti se
pojavijo kot posledica nihanja napetosti, tako da je mogoe vzpostaviti jasno razmerje med
flikerjem in nihanjem napetosti. Na splono, utripanja lahko bistveno zmanjajo nao vizijo
in povzroijo splono nelagodje in utrujenost. V nekaterih primerih lahko celo povzroi
nesree na delovnem mestu. Vtis flikerja oziroma njegova jakost je odvisna od amplitude in
frekvence nihanja svetlobnega toka svetil, ta pa je odvisna od nihanja amplitude napetosti v
omreju. Tipino frekvenno obmoje za nihanja, ki jih lahko opazi loveko oko je od 0,5
Hz do 35 Hz z amplitudami, ki se e zanejo z 0,2% od amplitude pri 50 Hz. Takno
odvisnost opisujejo U-f krivulje flikerja.V nadaljevanju je prikazano fizikalno ozadje nihanja
svetlobnega toka razlinih svetil pri napetostnem kolebanju pri razlinih frekvencah, ter
tudi loveko zaznavanje svetlobnega nihanja.

Za merjenje jakosti flikerja v elektroenergetskem sistemu se uporablja flikermeter, s
katerim se merijo trenutne jakosti flikerja v elektrinem omreju. Za ocenjevanje kakovosti
napetosti v omreju je potrebno statistino obdelati dobljene vrednosti trenutnega flikerja.
Odvisno od tega, koliko asa je potrebnega za izraun, obstajata dva razlina indeksa
flikerja:

Pst indeks za jakost kratkotrajnega flikerja,
Plt indeks za jakost dolgotrajnega flikerja.

V etrtem poglavju bodo teoretino predstavljene elektrooblone pei. Elektrooblone pei
preoblikujejo elektrino energijo v toploto s pomojo elektrinega obloka. Elektrini oblok

23
je elektrina razdelitev plina, ki proizvaja stalno izpraznitev skozi plazmo, zaradi toka skozi
obiajno neprevodni medij. Zanj je znailna nelinearna napetostno-tokovna karakteristika
in visoka gostota toka. Pretvorba elektrine energije v toplotno energijo z oblokom
omogoa veliko koncentracijo gostote moi na relativno majhni prostornini, kar izkoriajo
elektrooblone pei za taljenje jekla. Odvisno od naina kako talijo jeklo, lahko pei
razdelimo v tri glavne skupine:

pei z direktnim elektrinim oblokom,
pei s indirektnim elektrinim oblokom,
pei s potopljenim elektrinim oblokom.

Pei z direktnim elektrinim oblokom talijo elezo v pei tako, da se elektrini oblok
vzpostavi med elektrodo in elezom, ki se postopoma segreva in topi. Pri pei z indirektnim
elektrinim oblokom, se taljenje eleza v pei opravi na tak nain, da je elektrini oblok
vzpostavljen med dvema nasprotno leeima elektrodoma. Pri peeh s potopljenim
elektrinim oblokom so elektrode neposredno potopljene v taljeno zlitino. Elektrini loki so
ustanovljeni v plinastem obmoju znotraj lindre.

Poglavje pet obravnava dinamino kompenzacijo jalove moi. Jalova mo je potrebna za
dobro delo nekaterih potronikov, vendar njen prenos na veje razdalje povea
obremenitev na prenosne zmogljivosti elektroenergetskega sistema. Poleg tega prenos
jalove moi na dalje razdalje v elektroenergetskem sistemu povea izgube in zmanja
faktor moi. Iz tega razloga obstaja potreba po kompenzaciji jalove moi v kraju potronje.
Elektrine naprave so lahko glede na njihove elektrine lastnosti potroniki ali proizvajalci
jalove energije. Razvoj nove tehnologije je omogoil razvoj kompenzacijskih naprav, ki so
lahko istoasno potroniki in proizvajalci jalove moi. The Flexible alternating current
transmission system (FACTS) so pridobljene naprave, ki temeljijo na energetski
elektroniki, ki so sposobne absorbirati ali ustvarjati jalovo mo. Uporaba te naprave je
posebej utemeljena v aplikacijah, ki zahtevajo eno ali ve od naslednjih znailnosti:

hiter dinamien odziv,
monost za pogoste spremembe izhodnih vrednot,
fino nastavljivimi izhodnimi vrednostmi, hitro izvajanje, da bi dosegli znatno
poveanje zmogljivosti,
zmanjanje strokov prenosa.

Na splono, FACTS naprave, glede na priklopitev na prenosni sistem, lahko razdelimo na
naslednji nain:

serijske naprave,
vzporedne naprave,
kombinirani serijski-serijska naprave,

24
kombinirana serijsko-paralelna naprave.

Serijske naprave FACTS so zelo uinkovite pri upravljanju pretoka energije, kakor tudi pri
poveanju stabilnosti sistema. Z uporabo serijskega kompenzatorja se celotna serijska
impedanca med dvema tokama prenosnega voda lahko zmanja, kar naprej vpliva na
pretok delovne moi. Vrste serijskih naprav FACTS:

statini sinhroni serijski kondenzator,
tiristorsko krmiljen serijski kondenzator,
tiristorsko krmiljena serijska duilka,
tiristorsko vklopljena serijska duilka,
tiristorsko vklopljen serijski kondenzator.

Vzporedna kompenzacija se uporablja za vpliv na elektrinih lastnostih daljnovoda, da bi
poveali zaetno vrednost moi, ki se lahko prenaa preko prenosnega voda in kontrolne
napetosti vrednosti vzdol daljnovoda. Vzporedne naprave FACTS:

Statini sinhronski generator,
Statini sinhroni kompenzator,
Statini VAr kompenzator,
Sistem za shranjevanje energije.

Kombinirane FACTS naprave so sestavljene iz vzajemnih kombinacij serijskih in paralelnih
naprav. Statini VAr kompenzator je skupno ime za ve vrst FACTS naprav. Statini VAr
kompenzator je vzporedno prikljuen v sistem in ima monost generiranja ali absorpcije
reaktivne energije da bi dosegli doloene parametre v EES.

Z uporabo statinega VAr kompenzatorja se lahko povea zmogljivost prenosnega
elektrinega omreja, napetost sistema se lahko stabilizira, nizke frekvence nihanja sistema
se zaduijo. Te vrste naprav so veinoma krmiljene s tiristorji in najpomembneje med
njimi so tiristorsko krmiljena duilka in tiristorsko preklopljen kondenzator. Kombinacija
teh dveh vrst zagotavlja vejo fleksibilnost pri uresnievanju kontrolnega dela in
zmanjanje harmoninih injekcij toka. Tiristorsko krmiljena duilka zagotavlja
neprekinjeno nadzorovanje jalove moi toda samo v induktivnem obmoju. Da bi dosegli
dinamino krmiljenje jalove moi tudi v kapacitivnem obmoju, je struktura s fiksnim
kondenzatorjem vzporedno povezana.

Poglavje est se ukvarja z modeliranjem realistinih modelov elektrooblonih pei. Eden od
glavnih virov kolebanja napetosti v elektroenergetskem sistemu so elektrooblone pei.
Zato je pomembno razviti imbolj natanne modele elektrooblone pei, da bi bilo mogoe
simulirati pogoje, ki povzroajo fliker. Obstaja ve razlinih metod za modeliranje oblok
elektrooblonih pei, ampak vse metode so lahko razvrene v tiri naslednje skupine:

25
Analiza v vremenskem domenu,
Analiza v frekventnem domenu,
Kaotina metoda variacije,
Modeli elektrooblonih pei, ki temeljijo na periodini porabi delovne in jalove moi.

Obiajno se oblok elektrooblone pei modelira kot kontrolirano napetostni vir. Amplitude
frekvennih komponent napetostnega vira so asovno modelirane v cilju opisa variacija
doline elektrinega obloka, ki povzroa nihanje omrene napetosti. Doloanje frekvenno
karakteristine doline elektrinega obloka je zelo zahtevna naloga. Za simulacijo variacij
doline elektrinega obloka, se lahko uporabljata stohastini ali deterministini pristop.
Stohastini pristop zagotavlja nakljune variacije doline obloka, katere so blije realnosti,
vendar je matematina analiza zelo zahtevna in zahteva dalji as simulacije, zaradi
potrebe statistinega vrednotenja.

Ob deterministinem pristopu doline obloka se spreminja kot sinusna funkcija z izbrano
oziroma nastavljeno frekvenco. Vendar ta pristop ne opisuje popolnega obratovanja
elektrooblonih pei, omogoa pa laje simulacije s krajim simulacijskim asom in brez
dodatnega statistinega ovrednotenja rezultatov. Deterministini model obiajno
zagotavlja vijo stopnjo flikerja v doloeni toki obratovanja elektrooblone pei v
primerjavi z stohastinim modelom. Vendar deterministino modeliranje ne zadovoljuje
popolnoma, ker se nelinearna pojava modelira kot linearna. Nekateri avtorji dolino obloka
elektrooblone pei modelirajo kot kontrolirani napetostni vir, ki temelji na nelinearni u-i
karakteristiki. asovna modelacija napetostnega vira se dosee z dvema razlinima
pristopoma, ki temeljita na sinusoidni in nakljuni funkciji.

V prvem primeru (sinusni pristop) je izbrana frekvenca od 8,8 Hz za spremembo doline
obloka, ker je to frekvenca, ki lei v srediu lovekovega opaanja flikerja. Ta pristop
modeliranja, ko se upoteva samo najhuja frekvenca flikerja, ne upoteva celotnega
spektra karakteristinih frekvenc flikerja, ki pomembno vplivajo na nivo flikerja. Bolj
realno modeliranje doline obloka elektrooblone pei se lahko dosee s asovno
modelacijo doline loka z nakljuno funkcijo. Ampak problem je, kako doloiti nakljuno
funkcijo, ki bi bila sposobna natanno opisati karakteristini spekter frekvenc flikerja.

Alternativni pristop k modeliranju doline obloka je s pomojo teorije kaosa. Slabost
stohastinega pristopa k modeliranju s teorijo kaosa je mogoe reiti in tak model lahko
opie pojavo v elektrini pei bolj natanno. Pomanjkljivost te vrste modeliranja je, da
podatki pridobljeni s simulacijo ne ustrezajo podatkom izmerjenim v omreju in zato
morajo biti vsi dobljeni rezultati previdno sprejeti. Da bi premostili opisane
pomanjkljivosti, v tej doktorski disertaciji, je predlagana nova metoda za modeliranje
doline obloka. S predlagano metodo je mogoe dosegati zadovoljivo natannost pri
modeliranju asovne modelacije doline elektrinega obloka. Metoda modeliranja doline
loka, predlagane v tej doktorski disertaciji, temelji na nakljuni funkciji. Nakljuna funkcija
je pridobljena iz reprezentativnih vzorcev napetosti in toka.

26
Reprezentativni vzorci napetosti in toka so izloeni iz daljih intervalov, pridobljeni iz
velikega tevila tokovnih in napetostnih meritev v energetskem sistemu. Reprezentativni
vzorci napetosti (ali toka), so 1-sekundni ali 1-minutni intervali kateri so izloeni iz daljih
intervalov napetosti (ali toka). Vsak vzorec predstavlja karakteristino in edinstveno
kombinacijo harmonika in interharmonika napetosti katere so karakteristine za vsak
delovni ciklus pei. Tako reprezentativni vzorci napetosti predstavljajo "prstni-odtis"
daljega intervala napetosti. Reprezentativne vzorce izloimo na mestu vklopa oblone pei
v omreje. V tem primeru so potrebne meritve trenutnih napetosti na tem mestu.

Potem se iz desetminutnih izmerjenih valovnih oblik napetosti in toka, na 110-kV nivoju
lahko izraunajo reprezentativni vzorci napetosti in toka. Najprej desetminutne izmerjene
napetostne in tokovne signale razdelimo na 600 eno-sekundnih intervalov. Pri izdelavi
novega vzorca je pomembno zagotoviti, da se vsak napetostni vzorec zane in kona v
trenutku, ko ima vrednost 0, pri tem pa morajo biti tokovni vzorci sinhronizirani z vzorci
napetosti. Zdaj za vsak novi 1-sekundni napetostni vzorec formiramo novi dalji napetostni
deset-minutni signal. Nov dalji deset-minutni signal se formira s ponavljanjem 1-
sekundnih napetostnih vzorcev za 600-krat. Nadalje, za novoustanovljene deset-minutne
napetostne signale, pridobljene na ta nain bomo izraunali kratkoroni nivo flikerja.
Potem je izbrano est reprezentativnih vzorcev s priblino enakimi karakteristikami,
kakrne ima dolgi deset-minutni izmerjeni signal. Izbor reprezentativnih vzorcev je bil
izveden na tak nain, da imajo vsi izbrani vzorci priblino enak nivo flikerja, tudi delovne in
jalove moi, kot deset-minutni izmerjeni signal. Zdaj je, od estih izbranih reprezentativnih
vzorcev, izbran reprezentativni vzorec z najbolj natanno vrednostjo flikerja.

Nadalje, z uporabo metode kvadratine-demodulacije izraunamo napetostno ovojnico
izbranega vzorca. V tem primeru so uporabljeni visokopasovni filterji prvega reda z mejno
frekvenco 3 dB pri 0,05 Hz, in Butterworth nizkopasovni filter 6. reda z mejno frekvenco 3
dB pri 35 Hz. Od izraunane napetostne ovojnice izbranega vzorca se izraunava niz
interharmonike, ki se lahko teje kot edinstvena za vsako intalirano pe v omreju in zato
predstavlja svoj prstni odtis v sistemu.

Dobljeni signal zajema vse frekvence signala od 1 Hz do 35 Hz, katere elektrooblona pe
generira med njenim delovanjem. Na ta nain so bile izraunane interharmonike, ki
temeljijo na podatkih iz dejanskih meritev. Navsezadnje, ta signal se uporablja kot
kontrolna funkcija za krmiljenje kontroliranega vira napetosti v simulacijskem modelu.
Rezultati simulacije kaejo, da tisti model pei generira signale, ki imajo priblino enak
spekter kot signali dobljeni z realnimi meritvami.

V zadnjem poglavju je prikazana celotna matematina izpeljava in razvoj novega
regulatorja za SVC napravo. Matematina izpeljava temelji na poenostavljenem
matematinem modelu SVC naprave, zastopane v dq koordinatnem sistemu. Na podlagi
parametrov poenostavljenega matematinega modela SVC-ja so doloeni parametri

27
regulatorja. Prav tako je, v matematinem razvoju in preizkuanju zanesljivosti najvije
frekvence obratovanja tranzistorja, le-ta upotevana.

Na zaetku je najprej razvit matematini model SVC kontrolorja v dq sinhrono vrteem
koordinatnem sistemu. Popolna analiza stabilnosti sistema (analiza v ustaljenem oziroma
prehodnem stanju) je narejena.

Na samem koncu je uinkovitost predstavljenega regulatorja dokazana s pomojo
raunalnikih simulacij dejanskega modela tovarne jekla. Rezultati simulacije so pokazali,
da se predlagani kontroler lahko uspeno uporablja za kompenzacijo jalove moi, zmanja
stopnjo flikerja ali regulira napetost.

28
1 Introduction

Because it can be easily converted into other forms of energy, electricity is the most
commonly used and most widespread form of energy. The characteristics of electricity
(apart from the amplitude and the frequency of the voltage) were standardized very late.
Traditionally, it was thought that the quality of electricity was simply its reliability, in other
words, the lack of any permanent interruption to the electricity supply. However, this is no
longer so. In a modern power network the acceptance of power quality depends on
important physical characteristics of the delivered voltage. In this respect, the problems of
the continuity of supply are mainly solved during the planning and construction phases of
the power network, while the physical problem of the voltage quality is closely related to
its exploitation.

With energy markets becoming liberalized, electricity is become a commodity like any
other product and so it must satisfy the relevant quality standards. It is also necessary to
maintain a certain level of voltage quality in a network to ensure the proper operation of
the connected equipment. In the world of power networks, the designers and
manufacturers of electric devices are required to meet a large number of standards and
recommendations in the field of power quality. Each network user (a consumer or
producer of electricity) must limit the negative impact of its own equipment on networks
voltage quality with respect to a pre-agreed level.

The existence of a large number of nonlinear loads in distribution networks leads to a
number of negative effects. These negative effects have an influence throughout the
network on all the other consumers in the network. The common interest of producers and
consumers of electricity is a reliable, safe and high-quality supply of electrical energy. The
dominant influences on the quality of the electrical power come from so-called non-linear
consumers, with such as devices power electronics, electrical machines, EAFs, etc.
Distortion of the power quality implies a violation of the basic parameters of voltage in the
steady-state or transient condition and the deformation of waveforms.

Because of the markets orientation, the increase in consumption and the growing need for
electric-power transfer, the existing power systems are becoming increasingly burdened
and are beginning to operate on the limit of stability. Consequently, there is an increased
risk of a partial or a total collapse of the system, which is reflected in a lower quality of the
electricity supplied to consumers. The problem of a lack of transmission capacity and a low
power flow in the power system can be solved with the addition of new transmission lines
or by replacing the existing transmission line with a transmission line that has a higher
capacity (meaning a higher transmission capacity). These solutions are reliable, but their
implementation is an expensive and time-consuming process.

An alternative approach to solving these problems is the use of FACTS devices. FACTS
technology is not just a one device that solves an existing problem in power system, but

29
rather a set of devices, which can be applied individually or in coordination with other
devices, for the management of one or more parameters of the system. Using FACTS
devices it is possible to achieve better control of the power flow through the transmission
line without endangering the thermal limits of the transmission line, with minimal losses
and increased border stability. The working principle of FACTS devices is based on
elements of power electronics that allow better and simpler handling, as well as a faster
and more reliable operation of the FACTS devices. Because of these qualities, FACTS
devices are a very convenient means of extending the systems stability in the disorder
conditions, such as outage and overload power lines and generators. Essentially, FACTS
devices play a key role in the efficient and economical production and transmission of
electricity in the future.

1.1 Subject of the doctoral dissertation

This doctoral dissertation is focused on the problem of working with electric arc furnaces
and eliminating their negative impact on the power quality in a power system. EAFs
represent a nonlinear group of consumers that have the ability to convert electrical energy
into heat. During this process their negative impact on the power quality in the power
system cannot be neglected. To be able to develop an appropriate compensation device,
which will reduce the negative impacts of an EAF to an acceptable level, it is necessary to
develop an accurate and realistic simulation model of the EAF.

In the first part of the doctoral dissertations a new method for modelling an EAF is
presented which is based on representative samples of voltage and current. A detailed
procedure in which actual data obtained from measurements can be used to model the EAF
are shown. Also, a review of existing EAF models and all the advantages of the new method
compared to existing methods are presented. According to the proposed method, a
detailed, three-phase model of an industrial network in a steel factory was designed with
the PSCAD software. A large number of simulations was made and based on the obtained
results it was concluded that the developed arc-furnace model provides voltage and
current waveforms (harmonics, interharmonics, flicker level) that are almost equal to the
measured waveforms of real steelworks. This fact is confirmed by a comparison of the data
obtained with the simulation and data obtained from actual measurements.

The second part of the of research work relates to the development of the control
algorithm for SVC devices. A detailed development process for the controller, from the
mathematical equations to the implementation, to a realistic simulation model of the steel
plant is shown. The developed controller has the ability to compensate the reactive power,
the power factor correction, reduce the level of flicker and regulate the voltage at the point
of connection. Also, in this case a large number of simulations was made and the obtained
results show the set of advantages of the developed controller as compared to existing
controllers.

30
1.2 Contribution to science

The major contributions to science of this doctoral dissertation are listed below:

Development and description of a new method for modelling an electric arc furnace
Development of a new controller for the SVC device
New analyse method of simulations with model of EAF

List of the authors publications:

1) Lj. Spasojevi, B. Blai, I. Papi: Uporaba tiristorsko krmiljene serijske duilke za
zmanjanje flikerja elektrooblone pei, Journal of electrical engineering and
computer science, Vol. 78, No 3, pp. 112-117.

2) Lj. Spasojevi, I. Papi, B. Blai: A New Approach to the Modelling of Electric Arc
Furnaces with Representative Voltage Samples, International Transactions on
Electrical Energy Systems, 14 March 2014, DOI: 10.1002/etep.1900.

3) Lj. Spasojevi, I. Papi, B. Blai: Development of the Control Algorithm for
Industrial Use of Static VAr Compensator (SVC), processing at Journal of Power
Electronics.

4) Lj. Spasojevi, Z. Ivanovi: Realizovanje pozicionog elektromotornog pogona
pomou savremenog industriskog energetskog pretvaraa, Proceedings of the
52nd conference for Electronics, Telecommunication, Computers, Automation and
Nuclear Technique, ETRAN 2008 : 8-12. June 2008, Pali.

5) Lj. Spasojevi, B. Blai, I. Papi: Reduction of arc furnace flicker by thyristor-
controlled series reactor, 6th International Workshop on Deregulated Electricity
Market Issues in South-Eastern Europe, DEMSEE, 20.-21. September 2011, Bled,
Slovenia.

6) Lj. Spasojevi, I. Papi, B. Blai: Implementation of a controller for a static VAr
compensator in large industrial networks, International Conference on
Renewable Energies and Power Quality ICREPQ13, 20-22 March, 2013 Bilbao,
Spain.

31

2 Power quality

As a member of the EU, Slovenia adopted a European directive on electromagnetic
compatibility (EMC) and based on that created a rulebook for electromagnetic
compatibility. This ordinance defines the criteria required to ensure the electromagnetic
compatibility of devices and thus the proper operation of electrical devices connected to
the public network. For this directive to enter into force it has been necessary to adopt a
set of standards that will support it. These standards relate to power quality and the
definition of the allowable interference in the network. Simultaneously with the change in
the Energy Act is the definition of the requirements in terms of voltage quality in the
network. With this the network operator is bound to provide quality power to the place of
takeover in accordance with the requirements on the quality of the legislation in the
Energy Act. A violation of power quality involves a disturbance of the basic parameters of
voltage in the steady-state or the transient condition and the distortion of the waveform.

2.1 Standard EN 50160

Most European countries have incorporated the standard EN 50160 "Voltage
characteristics of electricity supplied by public distribution systems" that in 1994 were
adopted by the European Committee for Electrotechnical Standardization. Standard
EN 50160 gives the quantitative characteristics of voltage quality in the case of normal
operating status. Its purpose is to describe and identify the characteristics of the
distribution voltage, while not describing the average values of the observed parameters
for defining the largest deviations of certain parameters, which can be expected in the
electricity distribution network. The period of the measurement that is determined by
standard EN 50160 is seven days without interruption. The measuring clips that are
observed for each parameter are ten-minute intervals, except for the frequency, which is
observed in time-slices of ten seconds. Standard EN 50160 describes the limits or values
within which the voltage characteristics can be expected to remain at any supply terminal
in public European electricity networks and does not describe the average situation usually
experienced by an individual network user [1], [2], [3]. In Slovenia the quality of the
voltage is assessed using Slovenia's standard SIST EN 50160, based on the Europeans
standards EN 50160. Among other things, this standard covers the following thirteen
characteristics of the voltage:

frequency of the supply voltage,
declared supply voltage,
voltage deviation,
rapid fluctuations of voltage (flicker),
voltage dip,
short voltage interruption,

33
long voltage interruption,
temporary overvoltage between live conductors and earth,
transient overvoltage between live conductors and earth,
voltage unbalance,
harmonic voltage,
interharmonic voltage,
mains signalling voltage.

Frequency of the supply voltage

The nominal frequency is the number of occurrences of the supply voltage per one second,
and this should be 50 Hz. Under normal operating conditions the mean value of the
fundamental frequency measured over 10 s shall be within the range of:

for systems with synchronous connection to an interconnected system:
50 Hz 1 % (i.e., 49.5 Hz... 50. 5 Hz) during 99.5 % of a year;
50 Hz +4 % / - 6 % (i.e., 47 Hz... 52 Hz) during 100 % of the time [1], [2],

for systems with no synchronous connection to an interconnected system
(e.g., supply systems on certain islands):
50 Hz 2 % (i.e., 49 Hz... 51 Hz) during 95 % of a week;
50 Hz 15 % (i.e., 42.5 Hz... 57.5 Hz) during 100 % of the time[1], [2].

Declared supply voltage

The declared supply voltage is agreed by the network operator and the network user. The
voltage level is defined as the RMS value of the voltage at the point of delivery of the
electric energy at a particular time, measured within a certain period. Under normal
operating conditions, not taking into account the disruption of supply, 95% of the value of
the 10-minutes average value of the RMS value of the voltage during an interval of one
week must be in the range 10% Un, while the remaining 5% Un the observed time period
must be within the limits of -15 / +10% [1], [2].

Voltage deviation

Voltage changes are the deviation voltage values from the nominal values. They are mainly
caused by changes in the loads of the energy consumer, by switching in the system, or by
faults. Under normal operating conditions the change generally does not exceed 5% Un [1],
[2]. Rapid voltage changes can cause changes in the luminance of lamps, which may create

34
the visual phenomenon called flicker. Figure 2.1 shows an example of voltage changes
caused by changing loads.

Figure 2.1 Voltage changes caused by changing loads

Voltage dip

Figure 2.2 shows an example of unpredictable voltage dip. A drop in voltage is defined as a
temporary, sharp (unpredictable) reduction in the value of the voltage below a
predetermined limit.

Figure 2.2 Example of voltage dip

35
These limits are usually in the range of 90% to 1% of the nominal voltage value. Voltage
dips are usually caused by a malfunction in the system or a sudden increase in the load.
They are classified by their depth and duration. Most of the dips have a length of less than
1 s and a depth of less than 60% Un [1], [2].
Interruption of voltage supply

Figure 2.3 shows one example of voltage interruption. The interruption of the supply
voltage refers to a condition in which the supply voltage at the point of delivery is less than
1% of the nominal voltage.

Figure 2.3 Example of voltage interruption

There are planned and unplanned supply interruptions as well as short-term and long-
term supply interruptions. Short interruptions to the voltage supply are interruptions that
last less than 3 minutes. The number of such occurrences is from tens to hundreds per
year. The duration of approximately 70% of short interruptions should be less than 1
second [1], [2]. For long-term interruptions (longer than 3 minutes) it is not possible to
determine the limits for the number occurrences and duration. They are usually caused by
external events and the distributor of the electricity is not able to prevent them.

Temporary overvoltage between live conductors and earth

An overvoltage is a condition in which the supply voltage rises above a predetermined
limit. Short-term overvoltages in the network appear because of malfunctions or are
caused by switching manipulations, lightning strikes, etc. An overvoltage is registered as

36
soon as the values of the voltage supply exceeding the limit of +10% Un. Figure 2.4 shows
an example of an overvoltage [1], [2].

Figure 2.4 Example of an overvoltage

Transient overvoltage between live conductors and earth

Figure 2.5 shows a couple of examples of a transient overvoltage.

Figure 2.5 Example of a transient overvoltage

37
Transient overvoltages are high-frequency voltage changes between two or more
conductors or earth that last from microseconds to milliseconds, considerably shorter than
one period. They can be caused by switching operations in the network, lightning strikes,
the switching of capacitive or inductive loads, etc. Transients in the network may appear as
random impulses or repetitive impulses [1], [2]. Random impulses occur at unpredictable
times, at remote locations and necessitate the installation of special equipment for their
detection. Repetitive impulses (commutation voltage spikes, loads switching, etc.) are
more easily noticed, defined and confined.

Voltage unbalance

Figure 2.6 shows one example of a voltage unbalance in a three-phase system. A voltage
unbalance in a three-phase system is a condition when the magnitude of at least one phase
of voltage is not within the acceptable limits of the nominal voltage, or if the angle between
two phases is not equal to 120 degrees.

Figure 2.6 Voltage unbalance in a three-phase system.

There are several causes of voltage unbalance:

1) Variations in single-phase loading cause the current in the three-phase conductors to be
different, producing different voltage drops
2) Blown fuses or overcurrent protective devices
3) Unbalanced single-phase loadings
4) Fault on one of the phases.

38
In the standards SIST EN 50160 a voltage unbalance is defined as a period of one week
where under normal operating conditions 95% of the 10 min mean RMS values of the
negative phase sequence component of the supply voltage shall be within the range 0% to
2% of the positive phase sequence component [1], [2].

Rapid voltage fluctuations

Voltage fluctuations can be described as repetitive or random variations of the voltage
envelope due to sudden changes in the real and reactive power drawn by a load. The
fluctuating loads in the electrical power system, e.g., welding machines and arc furnaces,
are the main sources of these voltage fluctuations. The characteristics of voltage
fluctuations depend on the load type and size, and the power systems capacity. The voltage
waveform exhibits variations in magnitude due to the fluctuating nature or intermittent
operation of the connected loads. The frequency of the voltage envelope is often referred to
as the flicker frequency. Thus, there are two important parameters with respect to the
voltage fluctuations, the frequency of the fluctuation and the magnitude of the fluctuation
[1], [2]. Both of these components are significant in the adverse effects of voltage
fluctuations. These two parameters can be constants, and in this case there are periodic
voltage oscillations in the network, Figure 2.7a. Also, these parameters can be variable, and
in this case there are non-periodic voltage oscillations in the network, Figure 2.7b.

Most electrical and electronic equipment is designed to operate properly and within the
specifications if the voltage supply varies within 10% of the nominal value [1], [2].
However, some sensitive devices require a stable incoming voltage for them to perform
accurately, such as computers, medical equipment, telecommunications and test
equipment. Voltage fluctuations are usually evident in nuisance variations of the light
output from incandescent and discharge lighting sources. This is commonly known as
flicker, which is a subjective visual impression of the unsteadiness of a lights flux, whose
luminance fluctuates with time [1], [2].

39

Figure 2.7 Example of voltage oscillations: a) non-periodic; b) periodic

40
3 Flicker

Voltage fluctuations in power systems can cause a number of harmful technical effects,
resulting in a disruption to production processes and substantial costs. But flicker, with its
negative physiological results, can affect worker safety as well as productivity. Voltage
flicker is a problem that has existed in the power industry for many years. Flicker is defined
as the impression of unsteadiness of the visual sensation induced by a light stimulus whose
luminance or spectral distribution fluctuates with time [3]. In other words, flicker is
defined as the unpleasant sensation experienced by a person when are subjected to
changes that occur in the intensity of the illumination of light sources. These illumination
variations appear as a consequence of voltage fluctuations, so that a clear relationship can
be established between these disturbances (flicker and voltage fluctuations). This is reason
why flicker directly correlates with an evaluation of voltage quality.

As already mentioned, humans can be sensitive to the light flicker caused by voltage
fluctuations. Generally speaking, flicker can significantly impair our vision and cause
general discomfort and fatigue. Also, flicker affects our vision process and brain reaction,
almost always producing discomfort and deterioration in work quality. In some situations,
it can even result in workplace accidents because it affects the ergonomics of the
production environment by causing operator fatigue and reduced concentration levels [4],
[5].

Since the beginning of the twentieth century there have been a lot of measurements on the
effects of electric flicker on human vision [4], [6]. These measurements studied the
relationship between the light intensity threshold and the luminance variation frequency
as observed by human beings. The aim was to find the condition in which the observer
cannot see flickering light while the fluctuation with the same amplitude can cause a
flickering sensation with a lower variation frequency. Tests were performed on people who
were exposed to different variations of waveform voltage, levels of illumination and types
of lighting. The typical frequency range for flicker that can be noticed by the human eye is
from 0.5 Hz to 35 Hz with amplitudes that begin from 0.2 % of the amplitude at 50 Hz [5],
[6], [7].

Figure 3.1 shows the flicker detection threshold, i.e., the minimum required amplitude of
the voltage oscillation at a particular frequency, in which case 50% of the population is able
to detect the flicker. Figure 3.1 shows the border case of the flicker curve (U-f curve) for
sinusoidal and rectangular modulations of the voltage according to [6], [7]. From the figure
it is clear that the human eye is most sensitive to a flicker frequency of approximately 9 Hz.
A relative modulation voltage of 0.25% of the basic voltage amplitude at a frequency of
8.8 Hz can cause flicker [3], [6]. The flicker of electric lighting causes discomfort in humans
and is especially dangerous for people with epilepsy [8], [9].

41

Figure 3.1 Flicker detection threshold for signals with sinusoidal and rectangular waveforms

Flicker can be measured with a UIE/IEC flickermeter, which is an instrument designed to
measure any quantity representative of flicker [5]. There are two major indices used in the
evaluation of flicker in power systems, the short-term flicker index, (Pst), and the long-term
flicker index, (Plt). The short-term and long-term flicker indicators, which are the output of
the UIE/IEC flickermeter, are used to describe the characteristics of flicker. The short-term
flicker indicator Pst is the flicker severity evaluated over a short period (10 minutes is used
in practice). Pst=1 is the conventional threshold of irritability [5]. The long-term flicker
indicator Plt is the flicker severity evaluated over a long period (two hours is used in
practice) using successive Pst values.

3.1 Occurrence of flicker and flicker sources

The basic principle of creating flicker can be simply described with Figure 3.2.

Figure 3.2 Creating flicker from voltage oscillations in a power system

Voltage fluctuations at the high-voltage level of the power system, from its source (source
flicker) spread to the rest of the power system. Over the transmission lines and
transformers, these oscillations are transmitted to the low-voltage level of the power

42
system. The low-voltage level supplies the light sources (light bulbs, lamps, etc.) which
respond differently to oscillations in the voltage supply system. In the worst case for the
determined frequency spectrum, the range oscillation of the amplitude is large enough to
cause discomfort in people.

Occurrence of flicker

The voltage signal of a power system with the basic frequency fb, can be written as:

() = () sin( 2
),
(3.1)

where U(t) is the basic amplitude of the voltage and t is the time. If the basic amplitude of
the voltage is constant (U(t)=const), there is no flicker in the network. As a result of
nonlinear dynamic loads, the quick connection and disconnection of some consumers, the
basic amplitude U(t) of the voltage can fluctuate with time. However, each switching on
and off of consumers in the network, changes in their power, active and/or reactive, causes
voltage fluctuations across the entire network. When the value of the basic amplitude U(t)
is not constant (U(t)const), it can be any variable with time. For example, let us take the
function defined by (3.2) with some random frequency fm (frequency of modulation). The
amplitude that describes the fluctuation of U(t) is equal to 20 % of the amplitude of the
basic signal U (factor of modulation m=0.2).

() = +sin( 2
).
(3.2)

The modulation factor m is defined by equation (3.3), Figure 3.3.

=
. (3.3)

Based on everything mentioned above, equation (3.1) can now be written as

() = [1 +sin(2
)] sin(2
),
(3.4)

or in the general form:

() = [1 +
,
=1
sin(2
,
+
,
)] sin(2
),
(3.5)

43
where m
m,i
, f
m,i
and
m
,
i
are the amplitude, frequency and phase shift i component of the
modulation signal, respectively, and N is the number of all the modulation components. In the
case when there is only one frequency of modulation (i=1, f
m
=10 Hz, U=1 p.u. and m=0.2),
equation (3.5) becomes

() = [1 +0.2 sin( 2 10 )]sin(2 50 ).
(3.6)

Figure 3.3 shows an example where the voltage signals basic frequency (fb) is modulated
with a sinusoidal signal of a voltage frequency of 10 Hz (fm), equation (3.6). It is clear that
the basic frequency of the voltage is still 50 Hz, but now the peak values of the signal
fluctuate with a frequency of 10 Hz (see the blue line in Figure 3.3). These oscillations of
the voltage-peak values represent the oscillations of the envelope voltage. If this voltage
envelope contains a frequency from 0.5 Hz to 35 Hz as a consequence, flicker will occur in
light sources.

Figure 3.3 Voltage frequency of 50 Hz and an amplitude of 1 p.u. modulated by f
m
=10 Hz
and m=0.2 p.u.

Figure 3.4 shows the spectrum of the modulated voltage signal shown in Figure 3.3. It is
clear that there are three frequencies, the fundamental frequency (50 Hz) and on each side
one additional frequency (40 Hz and 60 Hz). These two frequencies are obtained by
subtracting (adding) the frequency of modulation from the basic frequency. In this way it is
possible to obtain a spectrum of the amplitude for any voltage envelope. Now, based on the
U-f curve shown in Figure 3.1 it is possible to assess whether the level of flicker will cause
discomfort in people.

44

Figure 3.4 Amplitude spectrum of the signal from Figure 3.3

The voltage envelope can be calculated in many ways. The first and simplest way is to use
the algorithm of a discrete Fourier transformation (DFT). Using DFT we obtain the
spectrum components of the signal, as shown in Figure 3.4, in the case of the modulated
voltage signal shown in Figure 3.3. The speed of the calculation can be improved with a fast
Fourier transform (FFT), while the accuracy can be improved by using a variable window.
Another approach that can be used to calculate the voltage envelope is based on a structure
with a squaring demodulation, as shown in Figure 3.5. Also, this structure is used in the
flickermeter and will be described later.

Figure 3.5 Square modulation structure

As previously mentioned, the main sources of severe voltage fluctuations are industrial
loads with fluctuating power demands. Also, wind turbines and wave power, etc., can
generate flicker [9], [10].

Arc furnace

An arc furnace is probably the load that produces the most flicker in a power system [9],
[10]. When the arc furnace operates, an unstable arc is established between the electrodes
and the scrap causing fluctuations of the power consumption and thus a potential flicker
problem. As a rule of thumb, the ratio between the short-circuit capacity at the point of
common coupling (PCC) to the maximum demand of the arc furnace should be greater than
80 in order to limit the risk of severe flicker caused by the arc furnace. The best way to

45
investigate the actual flicker situation is to perform on-site measurements using a
flickermeter based on [6]. If the arc furnace is connected to a network with changing loads
over time, a good idea is to measure the flicker permanently and thereby observe the trend
of the flicker. A common method to reduce flicker originating from an arc furnace is to
increase the short-circuit level by installing a new main transformer with a higher capacity,
installing active mitigation equipment like a SVC, etc., or to improve the control strategies
of the arc furnace [11].

Welding machine

Welding machines can also introduce major power-quality problems. A welding machine
normally has a much shorter on-time than off-time, resulting in both current harmonics
and sudden current spikes, leading to a voltage drop at the point of common coupling. If a
welding machine is a major load on the power network, the current spikes can cause
significant flicker levels if the repetitive frequency is within the flicker frequency window.

Wind turbines

The flicker emission from a wind turbine is mainly caused by a fluctuating power
production due to wind-speed changes, since the generated power from a wind turbine is
proportional to the cube of the wind-speed. This means a small change in the wind-speed
will result in a much greater change in the produced output power. If the wind turbine is
installed in a weak network, the change in output power due to a fluctuating wind speed
can cause considerable voltage changes at the point of common coupling and thereby
flicker. Other reasons for flicker emission originating from a wind turbine are the tower
shadow effect and, for pitch-controlled wind turbines, the limited bandwidth of the pitch
mechanism [12]-[15].

3.2 Estimate of the flicker level

The simplest way to estimate the level of flicker is the use of flicker curves, which are given
in the IEEE standard [6], [7], Figure 3.6. The term changes per minute is used as the unit
for the modulating voltage frequency in Figure 3.6. For a modulating voltage, one cycle is
two changes. Thus, 120 changes per minute are equal to 1 Hz. The curves 230 V, 120 V and
100 V are obtained from 230 V, 120 V and 100 V, 60 W incandescent-bulb measurements
with a rectangular modulated voltage, respectively. These curves indicate the relative
voltage fluctuation (U/U) with respect to the modulating voltage frequency. The flicker
curves shown in Figure 3.6 give the borderline values of the relative voltage fluctuation for
a corresponding modulating voltage frequency when the short-term flicker indicator (Pst)
is equal to one. A relative voltage fluctuation below the flicker curve will cause Pst to be less

46
than one. A relative voltage fluctuation above the flicker curve will cause Pst to be higher
than one.

Figure 3.6 Flicker curve for rectangular modulation frequencies presented in IEEE 141-1993,
P
st
=1

As the flicker curve is the simplest way to estimate the flicker level, the relative voltage
fluctuations for different lamp types to obtain the same flicker level can be compared using
the flicker curves of different lamp types. Figure 3.7 shows the instantaneous flicker curves
of the four lamp types. Figure 3.7 shows that the incandescent lamp is the most sensitive to
flicker, since its modulating voltage amplitudes are the lowest when Pinst, max equals one.

Figure 3.7 Instantaneous flicker curves for different lamp types [8]

47
The energy-saving lamp is the most insensitive to flicker because of its highest modulating
voltage amplitudes [16]. To obtain the flicker curves of different lamp types, the reference
flicker level needs to be defined. Therefore, the reference values of the relative voltage
fluctuations and the corresponding modulating voltage frequencies, which can cause the
reference flicker level, should be defined first. The Standard IEC 61000-4-15 gives the
values of the sinusoidal modulated voltage for a 230 V incandescent lamp when the
maximum instantaneous flicker sensation Pinst, max equals one.

3.3 Flickermeter

To evaluate the flicker level in power systems, it is necessary to develop a measurement
tool that can represent the relationship between the voltage fluctuations and the human
vision system. Therefore, a device called a flickermeter has been developed to measure any
quantity representative of flicker. There were four important national flickermeters that
already existed before the flickermeter was defined [6]. The major reason why an
international flickermeter had to be defined was that the national flickermeters did not
give universal results that can indicate the flicker level. All the specifications and all the
necessary provisions of the flickermeter were defined in [6]. In it there is a fundamental
mathematical formulation of the device and the implementation process for testing and
measurement. The specifications of the flickermeter are applicable only to a 60 W,
230 V/120 V incandescent lamp. A universal flickermeter model that would cover all types
of lamps does not exist because there is no universal curve of flicker for all lights.

It is important to bear in mind that the primary objective of the flickermeter is not to
provide an evaluation of voltage fluctuations, but of the flicker perception caused by these
fluctuations. To reach to this, the equipment has to be designed so that it is able to
transform the input voltage fluctuations into an output parameter that is proportionally
related to the flicker perception. This is possible by simulating the process of physiological
visual perception, which is the so-called lamp-eye-brain chain, [17]. The architecture of the
flickermeter is described in Figure 3.8., which shows the flickermeter can be divided into
two parts, each performing one of the following tasks:

Simulation of the response of the lamp-eye-brain chain.
Online statistical analysis of the flicker signal and the presentation of results.

The first task is performed by blocks 2, 3 and 4 of Figure 3.8, while the second task is
accomplished by block 5. The response of each block will be analysed in the following
subsections.

48

Figure 3.8 Block scheme of flickermeter
Block 1 input voltage adapter

The supply voltage is the input for this block. In this block there is a voltage-adapting
circuit that can scale the RMS value of the input voltage down to an internal reference level
and does not modify the modulated voltage waveform. This means that all the flicker
measurements can be made independently of the input carrier voltage level. The output of
this block is the normalized RMS value of the input voltage [8].

Block 2 - squaring multiplier

For the referenced incandescent lamp, the light produced by the lamp depends on the
energy consumed by the lamp. The consumed energy by the lamp is proportional to the
square of the input lamp voltage. This block is used to simulate this squaring behaviour of
the lamp. Since the incandescent lamp has a thermal time constant because of the filament
of the lamp, a low-pass filter function is used to present this time-constant function and it is
part of the weighting filter in block 3 [8]. Let

() = (sin)(1 +sin
),
(3.7)

where U(t) is the supply voltage with amplitude A and angular frequency , which is
modulated by a sinusoidal waveform with the amplitude m and angular frequency m. The
signal at the output of the squaring multiplier has the following form:

()
2
=
2
2
1 +
2
2
+
2
2
1 +
2
2
sin2 +
2
8
sin2(+
) +
2
8
sin2(
) +
2
2
sin(2+
) +
2
2
sin(2+
) +
2
sin
2
4
sin 2

(3.8)

49
Block 3 filters

This block consists of series circuits of two filters, i.e., the demodulator filter and the
weighting filter [8]:

Demodulator filter (high-pass filter +low-pass filter)
The demodulator filter includes a first-order high-pass filter (3 dB cut-off frequency at
0.05 Hz) to filter the DC-component caused by the squaring function in block 2 and a low-
pass filter to filter all the components that are equal to, or greater than, the fundamental
frequency of the carrier voltage [8]. A sixth
-
order low-pass Butterworth filter (gives a 3 dB
cut-off frequency at 35 Hz for a 50 Hz system and a 3 dB cut-off frequency at 40 Hz for a 60
Hz system) is recommended in this block. By filtering the DC component and all the
frequency components higher than the fundamental angular frequency , only the
following terms remain:

=
2
sin
2
4
sin2
.
(3.9)

Weighting filter
The weighting filter simulates the frequency response of a coiled-filament gas-filled lamp
and the human visual system. As mentioned in block 2, the filament lamp has a thermal
time constant function; it is considered as a low-pass filter with a cut-off frequency of about
6 Hz. The transfer function of the weighting filter is described as [8]:

() =
2
+2 +
2
x
1 +
2
1 +
3
1 +
.
(3.10)

For a 60 W 230 V incandescent lamp:

k = 1.74802
= 2 4.05981
1
= 2 9.15494
2
= 2 2.27979
3
= 2 1.22535
4
= 2 21.9.

For a 60 W 120 V incandescent lamp:

= 2 4.167375
1
= 2 9.077169
2
= 2 2.939902

50
3
= 2 1.394468
4
= 2 17.31512.

Block 4 - non-linear variance estimator

This block consists of a squaring block and a first-order low-pass filter with a time constant
of 300 ms. The block is used to simulate the delay effect of the human brain. The output of
this block is called output 5 of the flickermeter [6], [8], and it is the instantaneous flicker
level. A flickermeter is calibrated to produce a maximum instantaneous flicker value of 1, in
case when the amplitude and frequency of the modulated voltage are equal to values
shown in Table 1, where m is the modulation factor defined by (3.3) and fm is the
modulation frequency.

Table 1 Values of the parameters f
m
and m for instantaneous values of the flicker equal to 1 [6]
f
m
[Hz] m [%] f
m
[Hz] m [%]
0.5 2.340 10.5 0.270
1.0 1.432 11.0 0.282
1.5 1.080 11.5 0.296
2.0 0.882 12 0.312
2.5 0.754 13 0.348
3.0 0.654 14 0.388
3.5 0.568 15 0.432
4.0 0.500 16 0.480
4.5 0.446 17 0.530
5.0 0.398 18 0.584
5.5 0.360 19 0.640
6.0 0.328 20 0.700
6.5 0.300 21 0.760
7.0 0.280 22 0.824
7.5 0.266 23 0.890
8.0 0.256 24 0.962
8.8 0.250 25 1.042
9.5 0.254 33.33 2.130
10.0 0.260 - -

A value of one unit of flicker levels is based on the human perceptibility threshold for 50%
of observers looking a 60 W 230 V incandescent lamp. The output of this block is the
instantaneous flicker level Pinst. Since the input signal (voltage) of the flickermeter is a
periodic signal, Pinst is also a periodic signal. The maximum value of Pinst, which is also
called the maximum perceptibility at output 5 in [6], is an important value in the flicker
measurement. It is used in the standard IEC 61000415 to define the voltage modulation
amplitude for each specific voltage modulation frequency. The information about the
modulated voltage can be used also to evaluate the performance of the flickermeter. There
is a ratio of 0.68 between the maximum value of Pinst and the short-term flicker-level
indicator Pst for the voltage modulation frequency range 0.525 Hz [8].

51
Block 5 - statistical calculation block

Block 5 is used for the statistical calculation of the short-term flicker indicator Pst and the
long-term flicker indicator Plt using the time-at-level method [6].

Short-term flicker indicator (P
st
)
The Pst value represents the perceived flicker severity averaged over a ten-minute
measurement interval. A value of 1.0 or higher represents a flicker level that will bring
annoyance to an observer. The Pst value can be calculated using the equation:

= 0.0314
0.1
+0.0525
1
+0.0657
3
+0.28
10
+0.08
50
,
(3.11)

where the percentiles P0.1, P1, P3, P10 and P50 are the instantaneous flicker levels exceeding
0.1 %, 1 %, 3 %, 10 % and 50 % of the time during the observation period. The suffix s in
the equation indicates that the smoothed values should be used, which are shown in the
equation below. P1s, P3s, P10s and P50s can be determined as:

P
1s
=(P
0.7
+P
1
+P
1.3
) / 3,
(3.12)

P
3s
=(P
2.2
+P
3
+P
4
) / 3,
(3.13)

P
10s
=(P
6
+P
8
+P
10
+P
13
+P
17
) / 5,
(3.14)

P
50s
=(P
30
+P
50
+P
80
) / 3.
(3.15)

The IEC standard gives four observation intervals (1 min, 5 min, 10 min and 15 min).
However, only the 10-min interval will be used for the Pst assessment during field
measurements. The other three intervals can be used for the laboratory measurements [8].
The 10-min interval was chosen because it is sufficiently short to characterize in detail a
load whose operating cycle is long (e.g., an arc furnace) and also sufficiently long so that it
is not affected by very isolated short-term variations.

Long-term flicker indicator (P
lt
)
Since the short-term flicker indicator is only suitable to assess the flicker caused by an
individual source with a short duty-cycle, another flicker indicator, which can assess the

52
combined effects of several flicker sources (e.g., welders, motors) operating randomly, or
flicker sources having a long and variable duty-cycle, should be defined. The long-term
flicker indicator Plt has been defined for this purpose. It can be derived from the short-term
flicker indictor Pst values over a certain measuring period that relates to the duty-cycle of
the flicker sources. The mathematical equation used to calculate Plt is shown below:

3
=1
3
. (3.16)

Here, N is the number of the Pst measurements during the Plt observation period [8].
Normally, in practice, 2 hours is selected as the observation period for the long-term flicker
indicator Plt. To calculate Plt values, 2 hours must be a discrete time interval. Two-hour
sliding windows are not allowed. Since the Pst observation interval is 10 minutes and the Plt
interval is 2 hours, the number of measurements N will be 12 [8].

53

4 Electric arc furnace

The use of electric arc furnaces (EAFs) has grown rapidly over the past 30 or so years. The
main reason for this is their productivity, precision, flexibility and some advanced
applications for steelmaking. Of all the steel made today, 36% is produced by the electric-
arc-furnace route and this share will increase to 50% by 2030 [16]-[18]. On the other hand,
their time-varying nature and non-linear voltage-current characteristics have a large,
negative influence on the power quality. They are major sources of voltage oscillations in
power system that can produce the effect known as flicker. Also, EAFs need a significant
quantity of reactive power, which may give rise to the voltage instability in a power system
without proper compensation [16]-[18].

An electric arc furnace converts electrical energy to thermal energy in the form of an
electric arc in order to melt the raw materials held by the furnace. The arc is established
between the electrode and the melting bath and is characterized by a low voltage and a
high current. Electric arc furnaces are used for smelting and refining metals, mainly the
iron in the steel production. Using an electric arc we can achieve an extremely high
temperature and a high power density. Due to the high concentration of power in a small
area, an equal allocation of heat is not possible [16]-[18]. Therefore, electric arc are used
for melting metals and alloys, while for heat treatments they are inappropriate.

Electric arc furnaces based on alternating current (AC) and direct current (DC) represent
one of the most intensive, disturbing loads in sub-transmission or transmission electric
power systems. They are characterized by rapid changes in the absorbed powers,
especially during the initial stage of melting. Nowadays, arc furnaces are designed for very
large power-input ratings and due to the nature of both the electrical arc and the melt-
down process, these devices can cause large power-quality problems on the electricity net,
mainly harmonics, inter-harmonics, flicker and voltage imbalances. The Voltage-Current
(U-I) characteristic of the arc is non-linear, which can cause harmonic currents [18], [19].
In general, electric arc furnaces are classified according to their method of heating, as
follows:

furnace with a direct electrical arc
furnace with an indirect electrical arc
furnace with a submerged arc [19] - [21].

4.1 Furnace with a direct electrical arc

For these furnaces there is a specific low voltage, a high current and the electric arc is
established between graphite electrodes and batches that gradually heat up and melt.
Direct arc electric furnaces are very popular for the melting of alloy steels and range in size

55
from a few kilograms, for laboratory units, to in excess of 100 tonnes per batch. Typically,
the units found in foundries are in the range of 1 to 10 tonnes [19] - [21].

The scheme of an EAF with a direct electrical arc is presented in Figure 4.1. The furnace
consists of a spherical hearth (bottom), a cylindrical shell and a swinging water-cooled
dome-shaped roof. The roof has three holes for consumable graphite electrodes held by a
clamping mechanism. The mechanism provides for the independent lifting and lowering of
each electrode. The water-cooled electrode holders also serve as contacts for transmitting
the electric current supplied by water-cooled cables (tubes). The electrode and the scrap
form the star connection of three-phase current, in which the scrap is the common junction.
The furnace is mounted on a tilting mechanism for tapping the molten steel through a tap
hole with a pour spout located on the rear side of the shell. The charge door, through which
the slag components and alloying additives are charged, is located on the front side of the
furnace shell. The charge door is also used for removing the slag (de-slagging). The scrap is
commonly charged from the furnace top. The roof with the electrodes is swung aside before the
scrap charging. The scrap arranged in the charge basket is transferred to the furnace by a crane
and then dropped into the shell [19], [20].

Figure 4.1 Furnace with a direct electrical arc

Advantage of a direct arc furnace:

Quicker readiness for use
Longer hearth life
Ease of repair
Great independence of the quality of the charge
Analysis of the metal can be kept to an accurate limit.

The major disadvantage of this type of furnace is the high cost of the electricity and the cost
of the equipment. The heating costs are higher than for other furnaces [19] - [21]. This,
Slag
Arc
Spout
Charging
door
Molten steel
Electrodes

56
however, can be adjusted, to some extent, by using low-cost scrap turning or boring as the
metal charge.
4.2 Indirect arc furnaces

Generally, indirect arc furnaces consist of a horizontal barrel shape steel shell lined with
refractories. The melting is effected by the arcing between two horizontally opposed
carbon electrodes or graphite electrodes. The heating is via radiation from the arc to the
charge. The barrel-shaped shell is designed to rotate and reverse through approximately
180 C in order to avoid excessive heating of the refractories above the melt level and to
increase the melting efficiency of the unit. The furnace is mounted on the rollers and the
rollers can be driven to rock the furnace. The metal melts because of the heat radiated from
the arc and the heat from the hot refractory lining. Indirect arc furnaces are suitable for
melting a wide range of alloys, but are particularly popular for the production of copper
base alloys. The units operate on a single-phase power supply and hence the size is usually
limited to relatively small units. The scheme of an EAF with an indirect electrical arc is
presented in Figure 4.2 [19], [20].

Figure 4.2 Furnace with an indirect electrical arc

4.3 Furnace with a submerged arc

A submerged arc furnace is mainly used for ferroalloy production, e.g., ferromanganese,
ferronickel, ferrosilicon, ferrochrome, silicon manganese. In furnaces with a submerged arc
the electrodes are directly put into the molten alloy. The electric arcs are established in the
gaseous area within the slag. The charge is heated by the heat emitted in the arcs and also
by the Joule heat developed as the current passes through the charge. The energy
consumption of these furnaces is relatively high, and for high-carbon ferrochrome the

57
electrical energy consumption varies between 2000 kWh/T alloy, with pre-reduction to
4000 kWh/T alloy, without pre-reduction and feed preheating [20]. These furnaces
represent the most powerful consumers of electricity. Figure 4.3 shows the scheme of the
submerged arc furnace.

Figure 4.3 Furnace with a submerged arc

4.4 Electric arc

An electric arc is the electrical breakdown of a gas that produces an on-going plasma
discharge, resulting from a current through normally non-conductive media such as air.
The electric arc was first described in 1802 as a "special fluid with electrical properties", by
Vasily V. Petrov, a Russian scientist [21].

The arc occurs in the gas-filled space between at least two conductive electrodes that are
made of tungsten or carbon, and it results in a very high temperature. The electric arc is a
continuous discharge and may occur either in direct-current circuits or in alternating
circuits, while a similar electric spark discharge is momentary. Under the influence of an
electric field the particles between the electrodes are electrified and their movement is
directed. Positively charged particles are directed towards the cathode, while negatively
charged particles move towards the anode. The electric arc has a non-linear relationship
between the current and the voltage characterized by a small cathode voltage drop (from
10 to 20 V) and high current density (from 100 A/cm
2
to over 1000 A/cm
2
). The current
level between the electrodes is directly proportional to the gradient of the electric
potential. Figure 4.4 shows the voltage-current (U-I) characteristics of the electric arc.
Normally, gas is an electrical insulator [19], [20].

58
With a small voltage difference between the electrodes a small current (negligible) that is
proportional to the voltage will flow. This can be explained by the higher speed of the free
electrons and the positive ions in the stronger electric field, Figure 4.4 point 1. Increasing
the voltage at point 2 does not cause an increased current. This can be explained by the
limited number of charge carriers. If the voltage level rises above point 2, the electrons
accelerated at that level cause the ionization of other molecules. In this way the number of
charge carriers is increased and the current begins to increase with increasing voltage [19],
[20], [21].

Figure 4.4 Voltage-current characteristic of an electric arc

The voltage at point 3 is characteristic for a given gas environment and the type of
electrodes. Under the influence of this voltage, the secondary ionization of space occurs,
which is manifested in the form of sparks. The quantity of the charge carrier has increased
to such an extent that the environment can be treated as being conductive. The voltage
between the electrodes is decreased and this generates the negative U-I characteristic of
the arc. At point 4, the arc is established; electrons with a high speed collide with the
neutral atoms (molecules) and ionize them [20], [21].

Electric arc of direct current

An electrical arc of a direct current is obtained with the short circuit of two graphite
electrodes connected to a source of direct-current voltage. Following a brief short circuit
the electrodes move apart and between them the arc is established, Figure 4.5. Positively
charged particles move to the cathode, become neutral and accumulate on it. That is the
reason why the top of the cathode is a conical surface, while at the anode crater is formed.
Electric arc that burns between the electrodes has a surface temperature of over 4000C
and is characterized by a different electrode voltage drop. The anodes voltage drop is

59
about 10 V, while the cathode's voltage drop is larger and amounts to 2540 V. The voltage
drop along the length of the arc is practically negligible [19], [21].

Figure 4.5 Electrical arc of direct current

There are three distinct zones of the arc: the anode area, the body of the electric arc and the
cathode area. The length of the anode area is about 10
-3
cm and it is filled with electrons
that move towards the anode because the anode does not emit positive ions. Because of
these electrons, there is voltage drop on the anode. The length of the cathode area is less
than the anode area and its length is about 10
-5
cm. The voltage on the body of the electric
arc is increased in proportion with the distance from the cathode. The current density
emitted from the cathode is given by the Richardson-Dushman equation [21]:

=
2
2
,
(4.1)

where:

J is the emission density in [amperes/cm
2
]
M and N are constants (characteristic for each type of material) in [amperes/cm
2
(
K)
2
]
T is the temperature of cathode in [
K]
B is a correction factor due to the Schottky effect

The U-I characteristic of the direct current arc is negative and its shape is parabolic. Figure
4.6 shows the U-I characteristic of the direct current arc.
l
Arc
U
arc
A
K
U
k
U
k
U
a

60

Figure 4.6 Static characteristic of a DC electric arc

Changing the arc length (l), i.e., the distance between the electrodes, means the shape of the
curve is changed. Also, the voltage level that is necessary for the electric arc is modified.
Increasing the current of the arc as result, there is an increase of the temperature into the
electric arc. The temperature of the electric arc directly influences the ionisation, i.e., a
higher temperature means better ionisation (higher conductance of arc). Based on the
results of much experimental research, the equations that describe the characteristic
values and lengths of the electric arc have been formulated [19]. Equation (4.2) describes
the characteristic values of the electric arc.

= + +
+
arc
,
(4.2)

where:

U
arc
is the voltage of the arc [V]
I
arc
is the current of the arc [A]
l is the length of the arc [m]
A, B, C, D are constants that depend on the type electrodes and the gas environment.

Based on the results of many experimental measurements, a simplified equation that
describes the length of the arc can be written as:

arc
=
arc
l
+40,
(4.3)

where:

U
arc
is the voltage of the arc [V]
E
arc
is the electric field in the arc (approximately constant 1700-1900 V/m)
40 is a constant that represents the drop voltage on the electrodes.

61
Electric arc of alternating current

In case when the voltage supply is alternating, the conditions of the arc burning and its
characteristics are very different. The forms of the U-I characteristic of the electrical arc
depend on the conditions of burning of the arc, essentially from its cooling conditions. The
dynamic characteristics of a low-power electric arc for the conditions of intensive cooling
are shown in Figure 4.7.

Figure 4.7 Dynamic characteristics of the arc

During each pass of the current through zero, the area is cooled and deionized.
Furthermore, the resistance of the space between the electrodes is increased and there is
demand for a higher voltage level for the establish arc (voltage peak). This voltage peak
represents the voltage level of the arc ignition. With the advent current of the arc, the
voltage of arc is dropped to minimum for maximum value of current. The decreasing
current causes the voltage to increase to the point when the arc disappears (the voltage
level of shutdown). In the case when the power of the arc is higher and the conditions of
the cooling are lower, the voltage peaks and the voltage level of the shutdown are lower.
The dynamic characteristics of a high-power electric arc for low conditions of cooling are
shown in Figure 4.8.

62

Figure 4.8 Dynamic characteristics of a high-power electric arc

This form of the U-I features is characteristic for high-power electric arc furnaces for
melting steel, were the arc burns almost constantly. For the values of the arc current equal
to 10-15 kA, the measured current density in the AC electric arc is approximately 10
7
A/m
2
,
while the diameter of the arc is about 2 cm. For larger values of the arc current, the current
range is 100 kA, the current density is 108 A/m
2
and the diameter of the arc is about 3 cm.
The major useful characteristic of the arc is its high density heat, which can reach a value of
2 GW/m
2
. The temperatures on the surface of the arc are in the range 40005000
C, while
the temperature in its interior can reach 15000
C [19], [20].

63

5 Dynamic reactive-power compensation

The lack of static compensation means an inability to control the power factor in conditions
of variable load. In the dynamic type of reactive-power compensation the reactive power is
produced in a way that there is the possibility to quickly change the production level,
regardless of the voltage level at the connection point. Dynamic reactive-power
compensation can be stepped or continuous, depending on the configuration of the static
compensators. Therefore, dynamic sources can increase (or decrease) the level of reactive
power in the conditions of the voltage drop (overvoltage) and thus avoid a voltage collapse.
Some of the major advantages of the use of dynamic reactive-power compensation are:

compensation of the reactive power, which requires consumers with rapid changes in the
load
fast switching on and off of the capacitor, without a time delay for the discharge of a
capacitor
no transient phenomena
extending the lifetime of the system to compensate

Static compensators based on power-electronic, synchronous generators and
compensators represent the dynamic sources of reactive power.

5.1 Flexible alternating-current transmission system

The technological advancement of power-electronics has enabled the production of more
modern compensating devices and devices for the management of more complex power
systems. The Flexible Alternating Current Transmission System (FACTS) is a concept based
on power-electronics that enhance the value of transmission networks by increasing the
use of their capacity [22]. FACTS technology has created new opportunities for power
management and achieving the full capacity of transmission networks. The idea for the
development of these devices was born in the 1980s in order to solve problems around the
construction of transmission lines and facilitate the control of transmission power [23],
[24]. However, as already mentioned, the main reasons were the increase of the maximum
transmission capacity of the existing transmission system and the management of power
flows on certain routes.

FACTS devices work by acting on certain parameters, such as serial or parallel impedance,
current, voltage, phase angle, etc [24]. Using these devices is particularly justified in
applications that demand one or more of the following characteristics:

fast dynamic response
ability for frequent changes of output values

65
finely adjustable output values
rapid implementation in order to achieve a significant increase in capacity
reduction of transmission costs.

FACTS technology is not just a device that solves a certain problem, it is more of a set of
devices that can be applied individually or in coordination with other devices in order to
manage one or more parameters of the system [24]. In general, FACTS devices, depending
on the connection to the transmission system, could be divided as follows:

serial devices
parallel devices
combined serial-serial devices
combined serial-parallel devices.

Serial FACTS devices are very effective in the management of power flow as well as
increasing the stability of the system. Using a serial compensator the total effective series
impedance between two points of the transmission line can be reduced by influencing the
flow of active power. The possibility of controlling the power flow can be used to increase
the limit of the transient stability and the damping of the power oscillation [22], [24].
FACTS devices for series compensation are composed of capacitors, reactors and thyristor
converters. Types of serial FACTS devices are [24]:

Static Synchronous Series Capacitor SSSC
Thyristor Controlled Series Capacitor TCSC
Thyristor Controlled Series Reactor TCSR
Thyristor Switched Series Reactor TSSR
Thyristor Switched Series Capacitor TSSC.

Parallel compensation is used to influence the electrical characteristics of the transmission
lines in order to increase the initial value of the power that can be transmitted through a
transmission line and control the voltage values along the transmission line [24]. Parallel
FACTS devices are:

Static Synchronous Generator SSG
Static Synchronous Compensator STATCOM
Static Var Compensator SVC
Battery energy storage system BESS.

Combined FACTS devices are composed of mutual combination serial and parallel devices.
Someone of most important FACTS devices of this type are:

66
Unified Power Flow Controller UPFC
Thyristor-controlled phase-shifting transformers TCPST
Interline Power Flow Controller (IPFC).

5.2 Static VAr Compensator

Static VAr compensator (SVC) is the common name for several types of FACTS devices. SVC
is a parallel connected static VAr generator or absorber whose output is adjusted to
exchange the capacitive or inductive current in order to achieve the control of specific
parameters of the power system (typically the voltage on the BUS). With the application of
SVC, the power-transmission capability of the electric network can be increased, the
system voltage can be stabilized, the low-frequency oscillation of system can be damped,
and sub-synchronous oscillation can be suppressed, while the application of SVC in power
distribution network, voltage fluctuation, flicker, negative sequence and harmonic
interference caused by nonlinear loads and impact loads can be reduced, the power quality
can be improved, the productivity and power factor can be increased, the reactive tide can
be decreased, and the net loss can be reduced. These types of devices are mainly controlled
by a thyristor and the most important of these are thyristor-controlled reactors (TCRs) and
thyristor-Switched capacitors (TSCs). The structure of the SVC is the obtained combination
of TCR and TSC, such as the Fixed Capacitor Thyristor Controlled Reactor (FC-TCR)
structure and the Thyristor Switched Capacitor Thyristor Controlled Reactor (SC-TCR)
structure. This combination is ensures greater flexibility in the realization of control work
and reducing the injection harmonics of the current.

FC-TCR structure

The TCR structure can ensure continuously controllable reactive power only in the lagging
power-factor range. In order to achieve dynamic, controllable, reactive power also on the
leading power-factor domain, a fixed-capacitor (FC) structure is connected in parallel. The
FC-TCR compensator consists of a fixed capacitor bank utilized as high-order current
harmonics filters, and parallel reactors whose fundamental current harmonic is controlled
by means of thyristor AC switches [24]. The FC-TCR structure can be seen in
Figure 5.1. This type of SVC can provide an adjustment of the reactive power in both
quadrants, i.e., capacitive and inductive. The capacitors ensure the constant capacitive
reactive power (current, IFC) for the devices and usually inductance (L) is added in series,
Figure 5.1, which forms a resonant circuit that filters harmonic and/or interharmonic from
the network [24]. The variable inductive reactive power (current, ITCR) is ensured by the
thyristor-controlled reactors. The sum of the capacitive and inductive reactive powers
represents the total reactive power (current, ISVC), equation (5.1), which the device
generates or absorbs over the range from IFC to ITCR+IFC.

67

(5.1)

Figure 5.1 FC-TCR structure of SVC

The voltage-current characteristic for the FC-TCR is shown in
Figure 5.2. The compensator voltage-current characteristic encompasses the area of
inductive and capacitive loadings within the boundaries determined by the FC and the
thyristor-controlled reactor powers. It should be emphasized that the compensator is a
source of odd harmonics and, if the control angles of antiparallel-connected thyristors are
unequal, even harmonics can occur too. The third harmonics in the compensator current
can be cancelled by the delta connection of the reactor branches.

68

Figure 5.2 Voltage-current characteristic for FC-TCR type of SVC

The current in the reactor can be controlled in a continuous manner from maximum
(thyristor valve closed) to zero (thyristor valve open) by the controllable range of the
thyristor firing angle . The firing angle is defined as the delay angle from the point at
which the voltage becomes positive to the point at which the thyristor valve is turned on
and the current starts to flow. Figure 5.3 shows a simplified electric scheme of the TCR,
while Figure 5.4 shows the waveforms of the voltage and the current through the TCR
structure [24].

Figure 5.3 Structure of TCR

69

Figure 5.4 Waveforms of voltage and current through the TCR structure

Let us define the source voltage as:

() =
sin,
(5.2)

where US is the peak value of the applied voltage and is the angular frequency of the
supply voltage [24].

The current through the inductance L is obtained with:

()
() = 0,
(5.3)

() =
1
() +.
(5.4)

For the boundary condition i(t=)=0,

() =
(cos +cos ).
(5.5)

Using Fourier analyses we can derive an expression for the fundamental reactor current as
a function of angle , equation (5.6) [24].

() =
(1
2
sin2),
(5.6)

where U is the amplitude of the applied ac voltage, L is the inductance of the thyristor-
controlled reactor, and is the angular frequency of the applied voltage. Equation (5.6)
can be rewritten as:

70

() =
(),
(5.7)

where BTCR represents the reactive admittance (susceptance) of the TCR defined as:

() =
(1
2
sin2),
(5.8)

=
1
.
(5.9)

The variation of the perunit value of BTCR with the firing angle is shown in
Figure 5.5. The perunit value of BTCR is obtained with respect to its maximum value Bmax as
the base quantity [24].

Figure 5.5 Control characteristics of the TCR susceptance, B
TCR
()

71

6 The Modelling of Electric Arc Furnaces

One of the main sources of voltage fluctuations in power systems are electric arc furnaces.
It is therefore important to develop accurate models of the electric arc furnace for network
power-quality analyses that make it possible to simulate the conditions that cause flicker.
There are various methods for modelling electric arc furnaces. All these models can be
classified into four groups:

Time-domain analysis methods [25], [26], [27],
Frequency-domain analysis methods [28], [29],
Chaotic variation methods [30], [31],
Models based on the electric arc furnaces periodic draw of active and reactive power
[32].

Usually, the electric arc furnace (the electric arc) is modelled as a voltage source. The
amplitudes of the frequency components of the voltage source are time-modulated to
describe the arc-length variations that cause network-voltage fluctuations. Determining the
frequency characteristic of the arc length can be a difficult task. To simulate the arc-length
variation, a stochastic or a deterministic approach can be used. The stochastic approach
provides random variations of the length, which is closer to reality, but makes the
mathematical analysis complex and requires longer simulation times due to the need for a
statistical evaluation. With the deterministic approach the arc length varies as a sine
function with a selected frequency. However, this approach does not fully describe the
operation of the electric arc furnace, but allows a much easier simulation with shorter
simulation times and without any additional statistical assessment of the results. The
deterministic model usually provides higher flicker levels at a certain operating point of the
electric arc furnace compared to the stochastic model. However, deterministic modelling is
not entirely satisfactory, because a nonlinear phenomenon is modelled as a linear one.

In [26] the arc length is modelled as a controlled voltage source based on a nonlinear u-i
characteristic. The time modulation of the voltage source is achieved by two different time-
variation laws based on sinusoidal and white-noise functions. In the first case (sinusoidal
law) a frequency of 8.8 Hz for the changing arc length is chosen, as this is the frequency
that lies at the centre of perceptivity of the flicker. This modelling approach, when only the
worst frequency is considered, ignores the whole spectrum of characteristic flicker
frequencies that can have a significant effect on the flicker level. A more realistic modelling
of the furnace operation can be achieved with the addition of a white-noise, time
modulation of the arc length, as in [26] and also in [33]. However, neither author explains
or mentions how to specify the white-noise function that would be able to accurately
describe the spectrum of characteristic flicker frequencies.

An alternative approach to modelling the arc length is described in [34], where chaos
theory is used. The irregular and aperiodic behaviour of the arc length is based on the

73
Lyapunov exponent. With chaos theory the disadvantages of the stochastic approach to
modelling can be resolved, and such a model can describe the phenomena in an electric arc
furnace more accurately. Also, Chuas chaotic circuit can be used for modelling the
variation of the arc length. A disadvantage of this type of modelling is that the data
obtained with the simulation does not correspond to the measured data in the network and
therefore all the obtained results must be accepted with caution.

In [27] the arc is modelled as a controlled voltage source that is based on a piece-wise
linear approximation of the u-i characteristic of the load. The approximation and
simplification of the data for the modulation characteristics of the arc could make the
accuracy and reliability of this method questionable.

In order to overcome the described shortcomings, a new method of modelling the arc
length is proposed. With this method it is possible to achieve a satisfactory accuracy for
modelling the time-modulation of the electric arcs length. The advantages of this model
with respect to other models are presented.

6.1 Representative Samples of Voltage and Current

Representative voltage and current samples are obtained from a large number of current
and voltage measurements from electric-arc furnaces in the power system. Representative
samples of the voltage (or current) are one-second or one-minute intervals extracted from
longer (e.g., 100-second, 10-minute) voltage (or current) intervals [35]. Each sample
represents a characteristic and unique combination of the harmonic and interharmonic
voltage components for every operating cycle of the arc furnace. The representative
samples represent a unique fingerprint of each individual arc furnace, as shown in [35] and
[36]. The flicker levels of short, representative samples Pst,r are also almost equal to the
flicker level of the longer interval they substitute Pst, as shown in Figure 6.1.

74

Figure 6.1 Value of P
st
for a long signal and P
st,r
value of a representative sample

The Selection of Representative Samples

The ten-minute measured waveforms of voltage and current obtained at the 110-kV level
are presented in Figure 6.2 and 6.3. From measurements we know all the characteristic
values for these two signals (active and reactive power, flicker level).

Figure 6.2 Waveforms of voltage measured at the 110-kV level

75

Figure 6.3 Waveforms of current measured at the 110-kV level

In the next step it is necessary to conduct mathematical signal processing, which consists of
the several steps described below:

Creating one-second samples In this step 10-min measured voltage and current
signals are divided into 600 one-second intervals (one-second samples). When creating the
sample it is important to ensure that each voltage sample begins and ends at the point of passing
through 0. Current samples must be synchronized with voltage samples. In Figure 6.4 and Figure
6.5 the voltage and current samples are shown.

Figure 6.4 One-second sample of voltage at the 110-kV level

76

Figure 6.5 One-second sample of current at the 110-kV level

Calculation P, Q and Pst In this step is necessary to calculate the active and reactive power
and the flicker level for each generated one-second sample, i.e., for each one-second sample
we calculate the values of the flicker level (Pst, s) as well as the active (Ps) and reactive
power (Qs). The values obtained in this step are shown in Figure 6.6 and will be used as a
reference for selecting representative samples.

Figure 6.6 a) Active power of one-second samples and the mean value; b) reactive power of one-
second samples and the mean value

77
The red line in Figure 6.6a represents the active power of each individual, one-second
sample, while the green line represents the mean value of the active power of all the one-
second samples. Figure 6.6b is the same, where the blue line represents the reactive power
of each one-second sample and the yellow line represents the mean value of the reactive
power of all the one-second samples.

The value of the flicker level for each individual one-second sample is calculated as follows.
Firstly, from the one-second voltage sample we form a longer ten-minute voltage signal.
The longer ten-minute voltage signal is formed by repeating the one-second voltage sample
600 times. Next, for the newly created ten-minute voltage signal obtained in this way we
calculate the short-term of flicker level (Pst). The easiest way to do this is to use the
flickermeter. The calculated Pst of the newly created ten-minute signal represents the Pst of
the one-second voltage sample. The calculated values of Pst for each individual one-second
voltage sample are presented in Figure 6.7.

Figure 6.7 Flicker level (P
st
) of all the one-second samples

Selection of a representative sample In this step we select the six representative
samples with approximately the same characteristic as the longer ten-minute measured signal.
The selection of representative samples was made in such a way that all the selected samples
have approximately the same level of flicker (the most important condition), as well as active
and reactive power, as the ten-minute measured signal. The reference value of the flicker level
for selecting the representative samples is the measured value of the flicker at the 110-kV level
(P
st
). The values P
m
and Q
m
are used as a reference for the active and reactive powers.

The Figure 6.8 shows the values of the flicker level Pst, s and the active Ps and the reactive Qs
powers for the six one-second selected representative samples, s=1, 2, 3, 4, 5, 6. All the
parameters are related to one phase. Also, in the same figure we can see the parameters of
the measured ten-minute signal. The Pst represents the ten-minute value of the flicker level
measured at the 110-kV level, while Pm and Qm represent the mean values of the active and

78
reactive powers of the same signal at the same voltage level. In Table 2 the values of these
parameters are presented numerically.

Now, from the six presented representative samples the representative sample with the
most accurate values of Pst should to be selected. The representative sample numbered one
with values of parameters P1 and Q1 was selected.

Figure 6.8 Parameters of the measured ten-minute signal (P
st
, P
m
, Q
m
) and the representative
samples (P
s
, Q
s
active and reactive power of representative sample s=1, 2, 3, 4, 5, 6) for one
phase

Table 2 Numerical values of the parameters P
st
, P and Q of the measured ten-minute signal and
of the representative samples (phase values)
P
m
[MW] Q
m
[MVAr] P
st

Measured
signal
13.49 10.09 1.450
Representative
sample
P [MW] Q [MVAr] P
st

1 12.62 10.44 1.453
2 12.88 11.41 1.447
3 14.20 10.90 1.454
4 12.72 10.50 1.455
5 12.47 9.91 1.443
6 12.94 9.09 1.442

79
6.2 Model of the flickermeter

A schematic model of the flickermeter used in this work is shown Figure 3.8. The model
flickermeter was designed with the PSCAD software. It is based on [6] and already
described in Section 3.3. The input signal for this model (u(t)) is the voltage signal from
which should be calculated the flicker level, while the output signal (S(t)) represents the
instantaneous value of the flicker. Statistical processing of the S(t) signal provides us with
the short term value of the flicker level (Pst). Figure 6.9 presents the model of flickermeter
designed in PSCAD software, while in Table 3 the parameters of designed flickermeter
model are presented.

Figure 6.9 Scheme of the flickermeter model

Table 3 Parameters of the model shown in Figure 6.9
Parameter Value
N
1
(s)

N
2
(s)

N
3
(s)
100.5499s
0.06981s+1
1
D
1
(s)
D
2
(s)
D
3
(s)
s
2
+51.017s+3308.8017
0.129885s+1
0.007267s+1
G
T
1280000
0.3 s

Testing of S(t)

In order to validate the accuracy of the designed model it is necessary to test the model.
The Flickermeter model was tested according to the standard IEC 61000-4-15 [6] with sine
and rectangular signal modulation. The signal used for the testing is given with the
equations

() = sin(250) [1 +sin(250
)], (6.1)

where m is the modulation factor defined by (3.3) and fm is the modulation frequency. In
Table 1 are the standard prescribed values of these two parameters.

The prescribed accuracy is achieved if the values of the output signal S(t) (for sine
modulations with different amplitudes and frequencies) are equal to 1, with a tolerance of

80
5% [6]. In Table 4 the values obtained from the test were shown, while in Figure 6.10
they are graphically represented. In Figure 6.10 it is clear that the deviations of S(t) are
within the prescribed limits, except for the first two and the last value. If we take a look at
Figure 3.1 it is clear that these frequencies do not significantly influence the flicker level.

Table 4 The values of S(t) obtained by tests
f
m
(Hz) S(t) f
m
(Hz) S(t)
0.5 0.750 10.5 1.012
1.0 0.875 11.0 1.010
1.5 0.925 11.5 1.007
2.0 0.925 12.0 1.005
2.5 0.975 13.0 1.000
3.0 0.991 14.0 0.995
3.5 0.991 15.0 0.993
4.0 0.993 16.0 0.995
4.5 1.011 17.0 0.993
5.0 1.005 18.0 0.998
5.5 1.015 19.0 1.000
6.0 1.015 20.0 1.009
6.5 1.007 21.0 1.005
7.0 1.008 22.0 1.008
7.5 1.015 23.0 1.010
8.0 1.015 24.0 1.018
8.8 1.015 25.0 1.038
9.5
10.0
1.015
1.010
33.3 1.544
- -

Figure 6.10 Graphical overview of the values from Table 4

81
Testing of the P
st

In the second test, the flickermeter model was tested with a rectangular modulation signal
according to IEC 61000-4-15 with a rectangular signal modulation. The Pst for different
modulation frequencies was calculated. In each case, the value of Pst has to be 1.000.05
[6]. In Table 5 the characteristics of the rectangular signal and the calculated Pst values are
presented. In Figure 6.11 the table values are graphically displayed. On the basis of figures
Figure 6.10 and Figure 6.11 we can conclude that the flickermeter model designed in
PSCAD software has an acceptable accuracy. Borderline case in which deviation is larger
than allowed can be ignored, because it does not have a major impact on accuracy.

Table 5 Characteristics of the rectangular signal and P
st
values
Rectangular
changes per minute
Voltage changes U/U [%]
230 V lamp 50
Hz system
Calculated
P
st

1 2.724 1.0228
2 2.211 1.0288
7 1.459 1.0259
39 0.906 1.0325
110 0.725 1.0228
1620 0.402 1.0035
4000 2.4 1.2710

Figure 6.11 Values of P
st
for rectangular signal

6.3 Basic Concept of the Model of the Electric Arc Furnace

The model has to be developed so that it can satisfactorily describe every operating point
of the furnace. Moreover, the model should be general, i.e., it should make it possible to

82
model different arc furnaces. To achieve this, a correlation between the voltage and the
current of the electric arc has to be established. That correlation is described [26] by a
non-linear voltage-current u-i characteristic, given with equations (6.2) and (6.3):

), (6.2)

, (6.3)

where uA and iA are the arc voltage and current, respectively, UAT is the threshold value to
which the voltage tends when the current increases and C and D are constants, whose
values determine the difference between the increasing and decreasing current parts of the
u-i characteristic. The correlation between the threshold value of the voltage and the arc
length is explained by (6.4):

= + , (6.4)

where l represents the length of the arc in centimetres, A is the constant that represents the
sum of the anode and cathode voltage drops and B represents the voltage drop per unit of
the arc length. By inserting equation (6.4) in (6.3) we obtain:

= + +

(6.5)

The values for all the above-mentioned constants are given in Table 2.

Table 6 Values of the constants
Parameter Value
A
B
C
D
40 V
10 V/cm
30,000 W
5,000 A

In order to simulate the time dependence of the arc model, a time-variable arc length is
used:

=
0
(
) ,
(6.6)

0
=
0
+
= +
0
+
.
(6.7)

In the above equation u
A0
is the arc voltage corresponding to the reference length of the arc l
0
.
The parameter k is given by (6.8) and represents the ratio between the arc voltages threshold

83
value U
AT
(l) at length l, and U
AT
(l
0
) at the reference length l
0
[26].

=
()
(
0
)
=
A +B
A +B
0
.
(6.8)

The time-variation of the arc-length can be expressed by:

() =
0
(), (6.9)

where l0 represents the reference arc length when the furnace does not generate flicker
and r(t) is the law of the arc-length variation [26]. In the case when the length of the arc l
does not change over time (the length of the arc is equal to the reference length of the arc,
l=l0) the arc voltage-current characteristic would be time-invariant. The furnace would not
produce flicker, but only voltage and current harmonics would be generated and injected
into the power system, which is a consequence of the non-linear characteristics of the arc.
The arc length for this case can be calculated as:

0
= +
0

0
=
.
(6.10)

System Configuration

Figure 6.12 shows the part of the steel plants network under investigation.

Figure 6.12 Single-line scheme of the plant network

The network model involves a high-voltage network equivalent uHVNE with its short-circuit
inductance LLSC and resistance RLSC, a 110/35-kV transformer substation, the resistance RC
and inductance LC of the cable line between the substation and the furnace transformer,
and the 35/0.555-kV furnace transformer. The inductance and the resistance of the cable
line between the furnace transformer and the furnaces electrodes and the inductance and
the resistance of the furnaces electrodes are represented by the furnaces inductance Lf
and resistance Rf. The values of these parameters and the elements of the network are
shown Table 7. The value for the inductance Lf will be specified later in such a way that the
reactive power corresponds to the measured value. The other values correspond to the

84
actual parameters of the steel plant. The short-circuit power Ssc at the 110-kV bus equals
3750 MVA. Respecting the turn ratio of the transformers, all the measured values were
calculated at a low-voltage level.

Table 7 Values of the parameters of the model shown in Figure 6.12
Component Parameter Value
High-voltage
network equivalent
S
sc

u
HVNE
R/X
3750 MVA
110 kV
0.1
Tr 110/35
S
tr1

u
ntr1

u
ktr1
R/X
40 MVA
110/ 35 kV
10.49 %
0.1
Tr 35/0.555
S
tr2

u
ntr2

u
ktr2
R/X
35 MVA
35/0.555 kV
5.96 %
0.1
Values of the parameters at 0.555 kV voltage level
cable resistance
cable inductance
furnace resistance
RC
LC
Rf
2.7610
-6

1.3110
-8
H
1.8010
-3

The value of l0 is calculated so that the furnaces active power is the largest for every tap
position of the furnace transformer. This is achieved when the value of UAT0 is equal to 30%
of the voltage value (for that tap position) on the low-voltage side of the furnace
transformer. This correlation was determined experimentally based on real measured data.
Since the furnaces inductance Lf is not known from the real factory network, its value was
also determined experimentally based on the data from the selected representative sample
(P1, Q1). The values of these parameters were set such that the simulation model of the
electric arc furnace has approximately the same values of the active and reactive power as
the representative sample. Table 8 presents the experimentally determined parameters
based on real measured data.

Table 8 Values of experimentally determined parameters of the model used in the simulation
Parameter of
Representative Sample
Value
Determined
Parameter
Value
P
1
12.62 MW
U
AT0

l
o

167 V
12.84 cm
Q
1
10.44 MVAr L
f
8.0210
-6
H

The law of the arc-length variation r(t) in the case when the length of the arc l is not equal
to the reference length of the arc l0 but changes with time will be presented in the following
sections.

85
Modeling the Electrical Arc Length

The basic idea of the proposed approach to modelling the electric arc furnace is that the
electrical arc is modelled as a controlled voltage source [26]. This voltage source has to be
controlled so that it generates waveforms at the 110-kV level that correspond to the
waveforms of the representative (measured) samples. To achieve this, the amplitudes of all
the harmonic and interharmonic components of the voltage source have to be controlled.
Therefore, the function F(f, A, ) (f frequency, A amplitude, phase displacement of
interharmonic) for control of the voltage source must be determined. This function, in fact,
describes the arc-length variation.

All the resistances and reactances of the network in Figure 6.12 (of the 110/35-kV and
35/0.555-kV transformers, of the cable and of the furnaces resistances and reactances) can
be replaced with one equivalent resistance Req and one equivalent reactance Xeq, while the
electric arc furnace can be represented as a controlled voltage source. In this case, the
simplified single-line scheme shown in Figure 6.13 is obtained.

Figure 6.13 Simplified single-line scheme of the plant network

The waveforms of the voltage and current (i.e., the amplitudes and phase angles of all the
harmonics and interharmonic components) at node B (measured at the 110-kV bus and
recalculated to the 0.555-kV level) are known from the measurements and are
representative samples of the voltage uRS and the current iRS. The voltage drop u can be
calculated with (6.11) and, consequently, the waveforms of the voltage at point A can be
obtained with (6.12). The voltage uA is the voltage at the point of the arc furnaces
connection to the network.

=
),
(6.11)

)
(6.12)

86
Using the representative samples of voltage and current and the equivalent impedance, the
voltage waveforms at point A can be calculated. Equation (6.12) is an ordinary differential
equation that has to be solved numerically, because the data are given in discrete form. In
this way the obtained voltage waveforms at the point A correspond to the actual
waveforms at that point. The voltage drops in phases b and c are calculated in the same
way.

Next, it is necessary to define the law of the arc-length variation r(t) in the case when the
length of the arc l is not equal to the reference length of the arc l0 but changes with time.
This can be calculated from the voltage envelope of the representative sample spectrum.
The arc voltage value uA in (6.3) corresponds to the calculated voltage (from
measurements) at point A in Figure 6.13. The arc current can be easily calculated
(respecting the turn ratio of the transformers) from measured representative samples of
the current. The obtained voltage fluctuations at point A correspond to the actual voltage
fluctuations at the 0.555-kV level. Since the voltage waveform at point A is known, the
voltage envelope at 0.555-kV can be calculated.

Calculation of the Voltage Envelope

To calculate the voltage envelope of uA the squaring-demodulation method is used [37].
The structure of this method is shown in Figure 6.14, while Figure 6.15 and Figure 6.16
show the signal waveforms before and after passing through the squaring-demodulation
structure. The filter used in this case is a first-order high-pass filter with a cut-off frequency
of 3 dB at 0.05 Hz, while a 6
th
-order Butterworth low-pass filter has a cut-off frequency of
3 dB at 35 Hz. The high-pass filter is used to eliminate the DC component and the
Butterworth filter to reject the double main frequency components caused by the squaring
demodulator, Figure 6.16. By applying a fast Fourier transform (FFT) to the signal p'(t) a
set of interharmonics (amplitudes A and phase angles ) for this signal are calculated. A
sampling frequency for the FFT equal to 1 Hz was chosen.

Figure 6.14 Squaring-demodulation structure

The calculated set of interharmonics can be considered as being characteristic for each
individual installed furnace and therefore represents its fingerprint in the system. The
signal p(t) is computed by applying an inverse fast Fourier transform (IFFT), covering the
frequency range from 1 Hz to 35 Hz, to the obtained set of interharmonics. Since this range
of interharmonics involves all the important frequencies for flicker [38], [39] it will be used

87
for modelling the time variation of the arc length r(t) in (6.9). The obtained signal involves
all the signals with frequencies from 1 Hz to 35 Hz generated by the electric arc furnace
during its operation. In this way the interharmonics were calculated based on the data
from actual measurements.

Figure 6.15 Waveform input signal of the squaring-demodulation structure

Figure 6.16 Waveform output signal of the squaring-demodulation structure

6.4 The Results of the Simulation

A detailed three-phase model of an industrial network in a steel factory, shown in Figure
6.12, was designed with the PSCAD software. The network equivalent is represented with a
voltage source respecting the short-circuit power Ssc at the 110-kV bus. The non-linear

88
arc-furnace model was implemented in the simulation program with three single-phase
controllable voltage sources. The control function F(f, A, ) for the control of the voltage
source at the arc-furnace model is based on equations (6.6) and can be written as:

=
= ()
0
(
). (6.13)

Substituting the calculated signal p(t) in equation (6.9) the law of the arc-length variation
is derived. In order to achieve identical levels of flicker Pst in the simulation and the
representative sample, the signal p(t) was multiplied by the constant G (gain), which was
determined experimentally (based on measured data) and was set to 6.25, equation (6.14).

() = ().
(6.14)

Applying the above-mentioned equation (6.9) for the time-variation of the arc-length
becomes:

() =
0
(). (6.15)

Furthermore, based on the time-variation of the parameter k expressed by (6.8) and
(6.15), the parameter k can be written as:

() =
+ ()
+
0
= 1
()
+
0
. (6.16)

Finally, the equation that describes the control function F of the controllable voltage sources
representing the electric arc furnace model is given by:

= 1
()
+
0
+
0
+
.
(6.17)

A total simulation time (T) of 50 seconds was chosen with a sampling frequency of 6.4 kHz.
In the simulation the waveforms of the voltage at the 110-kV and 0.555-kV levels were
observed along with the flicker level at the 110-kV level. The interharmonics spectrum of
the voltage at 110 kV, which is the cause of the flicker, is shown in Figure 6.17. Two sets of
interharmonic spectra are shown: the simulated and the measured. Although the
differences in the amplitude between some interharmonics can be seen, both voltages
cause approximately the same level of flicker at 110 kV. The amplitudes of the 50-Hz
component of the measured and obtained signals are identical and equal to 89.90 kV. In
Figure 6.17 they are given with different lines because of the ordinate scaling.

89

Figure 6.17 The spectrum of interharmonics at the 110-kV level; a) measured, b) simulated

The waveforms of the voltage and current simulated at the 0.555-kV level are shown in
Figure 6.18 and Figure 6.19. These correspond to the waveforms of the voltage and the
current of the electrical arc.

Figure 6.18 Waveforms of the voltage simulated at the 0.555-kV level

90

Figure 6.19 Waveforms of the current simulated at the 0.555-kV level

If we put the waveforms of the voltage and current simulated at the 0.555-kV level in the
same figure, Figure 6.20, it is clear that there are no phase displacements between them.
Also, it is clear that irrespective of the polarity of the voltage, the arc remains permanently
conductive. In Figure 6.20 the voltage of the arc uA is given in volts, while the arc current is
given in kiloamperes.

Figure 6.20 Simulated waveform of the voltage u
A
and the waveform of the current i
A
of the
electrical arc

The absence of a phase shift between the arc voltage and the arc current, and the fact that
the arc contains only ohmic resistance was confirmed. Based on all these simulation
results, the obtained arc voltage-current u-i characteristic of this electric arc furnace model
is shown in Figure 6.21.

91

Figure 6.21 Dynamic voltage-current u-i characteristic obtained from the simulation results

The time variation of the arc length l obtained with this method and the reference length of
the arc l0 are shown in Figure 6.22.

Figure 6.22 Reference arc length l
0
and the arc length obtained in the simulation l

92
In the picture it is clear that the main characteristic of the electric arc is fast and large
changes of the length in a very short period of time. Also, the picture clearly shows that at
any point the arc length is not equal to 0. Based on that can be concluded that the electric
arc is uninterrupted.

Figure 6.23 shows a representative samples instantaneous values of the flicker S(t) at the
110-kV level together with the simulated values. It is clear that the waveforms of these two
signals are approximately the same. After the statistical processing of both signals,
approximately the same value of Pst was obtained, as shown in Table 9.

Figure 6.23 Representative samples S(t)
RS
and simulated S(t)
sim
instantaneous values of flicker
at the 110-kV level

The simulated flicker level Pst, sim, the flicker level of a representative sample Pst, r and the
flicker level measured at 110 kV Pst are shown in Table 9; the measured active and reactive
power (Pm, Qm), active and reactive power of representative sample (P1, Q1) and the
simulated values (Psim, Qsim) are shown in the same table.

Table 9 Values of active power, reactive power and flicker level
Parameter Value
P
st

P
st, r

P
st, sim

1.450
1.453
1.457
P
m

P
1

P
sim

13.49 MW
12.62 MW
12.62 MW
Q
m

Q
1

Q
sim

10.09 MVAr
10.44 MVAr
10.44 MVAr

93

7 Developing the control algorithm

It is widely known that the voltage fluctuations at the point of connection are mainly
caused by a rapid change in the reactive power of the electric arc furnace. The TCR can
quickly alter the inductive current in a continuous way as the system demands, via
changing the firing angle () of the thyristors connected in series with the reactor. Using a
suitable controller for the firing angle () the consumption of reactive power from the
power system and the voltage fluctuations can be decreased to an acceptable level. The
control algorithm has an important role in the proper and efficient operation of the TCR
[40], [41].

Static VAr compensators (SVCs), based on conventional thyristor-phase-controlled
technology, can generate or consume reactive power. SVCs represent the most common
and most widely used type of FACTS devices that are used to compensate for the negative
impact of EAFs [42], [43]. In order to achieve the maximum dynamic characteristics of the
SVC, which are limited by delays in the reactive power measurement and the ignition of the
thyristors, an appropriate control algorithm has to be developed. The control algorithm of
the SVC is an important part for SVC operation, because the dynamic characteristic of the
SVC mostly depends on it [44].

Over time, many different types of control algorithms have been researched and applied to
SVCs. In [44] and [45] the proposed control algorithm is based on a robust control strategy:
in [46] a self-tuning PID controller is proposed, while in [47] a conventional PID is
proposed. The authors proposed control schemes of the SVC based on fuzzy logic in [40],
[48] and [49]. Fuzzy logic is suitable for implementation because of the variety of
advantages that can be used over conventional computational systems.
In this work a procedure for developing and testing a P controller for an industrial SVC will
be shown and its operation will be demonstrated by means of a simulation for a real
industrial network case.

7.1 Vector illustration of three-phase quantities

A three-phase system can be described by three vectors in a complex plane, which with
respect to the coordinate system rotate with a synchronous speed. In order to simplify the
analysis, the conversion of the three-phase system to a rectangular coordinate system is
made. The conversion of the three-phase system to a rectangular coordinate system is
based on the following assumption. The set of current variables in the three-phase system,
whose sum is equal to zero, can be represented as unique vector in the complex plane,
Figure 7.1a. Then, that vector can be represented in the rectangular coordinate system.

95
For the conversion of the three-phase system onto a rectangular coordinate system we use
direct quadrature zero (dq0 or dq) and (0) transformations. The dq transformation is
a transformation of the quantities from the three-phase stationary coordinate system to the
quantities in the dq rectangular rotating coordinate system. The transformation is
conceptually similar to the dq transformation, and whereas the dq transformation is the
projection of the phase quantities a onto two-axis rectangular rotating coordinate system,
the transformation is the projection of the phase quantities onto a stationary
rectangular coordinate system, Figure 7.1b.

Figure 7.1 Vector illustration of three-phase quantities: a) vector in the complex plane b) in the
stationary rectangular coordinate system

In the case when the sum of the three-phase variables is not equal to zero, it is necessary to
determine a zero component. This case will be neglected, because the compensation device
that will be analysed in this doctoral dissertation is connected to the system with three
phase conductors and therefore the zero component of the current in each case is equal to
0. Moreover, there is a possibility of the occurrence of a zero-voltage component, as the
consequence of various external impacts, but in combination with the zero-current
component there is no impact on the instantaneous power of the compensation device.
This is the reason why we can leave out the zero components in further work.

The transformation from the three-phase stationary coordinate system to the two-phase,
so-called , rectangular stationary coordinate system is made using the transformation
matrix T, defined as:

96

=
2
3
1
1
2
1
2
0
3
2
3
2
1
2
1
2
1
2
.
(7.1)

The variables of the coordinate system are calculated from the variables of the three-
phase system as follows:

0
=
3
and
0
=
3
.
(7.2)

The calculated components and represent the vector projection to the perpendicular
axis, Figure 7.1b. From the obtained and components of the voltage and current (i
,
u
), the amplitudes can be calculated using the following equations:

=
2
+
2

(7.3)

and

=
2
+
2
.
(7.4)

The transformation from the three-phase stationary coordinate system to the two-phase,
so-called dq transformation, rectangular rotating coordinate system is made using the
transformation matrix Tdq, defined as

=
2
3
cos() cos
2
3
cos +
2
3
sin() sin
2
3
sin +
2
3
1
2
1
2
1
2
.
(7.5)

The variables of the dq coordinate system are calculated from the variables of the three-
phase system as follows

0
=
3
and
0
=
3
.
(7.6)

In the conversion to the dq rotating coordinate system it is necessary to indicate that the
transformation matrix is no longer constant. Now, the transformation matrix is time-

97
dependent and as a variable contains the angular speed . The angular speed represents
the synchronous speed of the rotating dq coordinate system. A vector illustration in the
two-phase synchronous rotating coordinate system is shown Figure 7.2b.

Figure 7.2 Vector illustration of three-phase quantities: a) vector in the complex plane b) in the
dq rotating coordinate system

The transformation to the dq rectangular rotating coordinate system gets real practical
value when the angular speed of the dq coordinate system is equal to the angular speed of
the three-phase system. In such a case, the AC quantities, voltage and current, in the three-
phase AC system are converted to the constant quantities of the dq rotating coordinate
system (only in steady-state conditions). This property is very suitable for the
development of a control algorithm for system devices.

Instantaneous active and reactive power

The instantaneous active power in the three-phase system quantities can be calculated
using the following equation:

() =
1
()
1
() +
2
()
2
()+
3
()
3
().
(7.7)

The instantaneous reactive power can be described as part of the power that exists in each
phase and whose sum of all three phases is equal to zero. Because of it, there is no influence
of the instantaneous reactive power on the instantaneous active power.

98
Using the transformation matrix (7.1) the instantaneous active and reactive power can be
written in the rectangular coordinate system as:

=
3
2
,
(7.8)

=
3
2
.
(7.9)

In the same way, using the transformation matrix (7.5) the instantaneous active and
reactive power can be written in the dq rectangular rotating coordinate system as:

=
3
2
,
(7.10)

=
3
2
.
(7.11)

If we make an appropriate synchronization of the dq rectangular rotating coordinate
system, the value of uq will be:

= 0.
(7.12)

This synchronization is achieved when the angle () between the voltage vector of phase
L1 and d-axe of the dq rotating coordinate system is equal to 0. According to all the
mentioned equations for the instantaneous active and reactive power in the dq coordinate
system become:

=
3
2
,
(7.13)

=
3
2
.
(7.14)

On the basis of (7.14), we can conclude that the q-component of the current is proportional
to the reactive power, while the d-component is proportional to the active power.

99
7.2 Compensation principle and mathematical model of TCR

For the design of the control algorithm for a TCR the simplified equivalent scheme shown
in Figure 7.3 is used.

Figure 7.3 The simplified equivalent scheme of a TCR a) delta configuration; b) star
configuration

The symbols used are as follows:
,

and
are the system currents (currents from the

power system),
and
are the phase-to-neutral voltages at the point of common

coupling (PCC),
,

and
are the currents through the reactors in the delta equivalent

scheme, while
_
,
_
and
_
are the currents through the reactors in the star
equivalent scheme. In the star circuit, the currents through the reactors and the line
currents of the TCR are the same. For simplicity, the star configuration will be used for the
controller development, as the results are not dependent on the star or delta
representation of the TCR circuit.

The proposed control algorithm is based on continuous measurements of the system
currents. When the SVC operates in dynamic conditions, the control algorithm calculates
the angle of the thyristors ignition that changes the reactance of the TCR. Such an
operation may be modelled by a variable factor b, as shown in Figure 7.3. Factor b is
expressed in per-unit (p.u.) and can have a value between 0 and 1. For a value equal to zero
the current of the TCR is zero, while for the value of one the current through the TCR is
equal to one per-unit. The reader should note that the factor b is equal for the star and for
the delta TCR equivalent scheme (Figure 7.3). The resistance in series with the reactor
represents the losses of the TCR. For a further derivation, for the parameters in Figure 7.3b,
the following per-unit system is used:

100

,
(7.15)

,
(7.16)

,
(7.17)

,
(7.18)

where
and
are the base values of the current, voltage and impedance,

respectively, while
is the synchronous angular speed of the fundamental network

component. When the per-unit system is adopted, the following set of equations can be
written, based on the equivalent scheme shown in Figure 7.3b:

(7.19)

or on p.u.:

.

(7.20)

Applying the dq transformation, a 3-phase system can be transformed into a dq
synchronous rotation coordinate system (SRCS). Assuming that the phase inductances and
phase resistances are equal (X and R respectively), equation (7.19) can be written as (7.21)
in d-q SRCS and using the Laplace transform.

101

X
=

(7.21)

where s=d/dt. In this case both components (
and
) of the current that flows into the

TCR are coupled by the angular speed . A common choice would also be to set
= .
In steady-state conditions the values of
and
are constant, which is convenient for the

derivation of the control algorithm. The currents can be expressed as:

X
=
,
(7.22)

,
(7.23)

=
1
+
,
(7.24)

=
1
+
.
(7.25)

Based on equations (7.24) and (7.25) a simplified mathematical model of the TCR device in
dq SRCS is obtained and is shown in Figure 7.4.

Figure 7.4 Simplified mathematical model of TCR

102
7.3 Controller design

The derivation of the control algorithm of the TCR device in the d-q SRCS starts from the
simplified mathematical model shown in Figure 7.4. Firstly, a proper synchronization of the
coordinate system must be made. With this proper synchronization the reactive
component (
) of the TCR current is proportional to the reactive power (q) of the TCR
device. The dq SRCS is synchronized to the medium-voltage (MV) level so that the angle
between the voltage vector of uMV (voltage at the MV level) and the d-axes of the dq SRCS
equals 0.

From Figure 7.4 it is clear that a change in the factor b influences the value of
. Therefore,
the desired values of
can be generated by suitable values of the factor b. In fact, the

susceptance of the TCR (BTCR) can be controlled by the parameter b.

Now, the connection between the reactive component of the TCR current (
) and the
reactive component of the system current (
) should be established. The system current

(
) flows from the power system to the EAF and its reactive component should be
compensated to zero. Figure 7.5 shows a simplified system configuration.

Figure 7.5 Simplified scheme of the system

For point A we can write:

,
(7.26)

where
and
are the system current, the SVC current and the load current,
respectively. Subsequently, the current of the SVC device (isvc) can be written as:

,
(7.27)

where
and
are the TCR and FC currents. Based on equations (7.26) and (7.27), the
reactive current components of the system current (
) and of the SVC device (
) in
the dq SRCS are obtained:

103

In a fully compensated system, the reactive component of the system current (
) should
be equal to zero. In other words, the reactive component of the TCR current should be
adjusted so that
equals zero:

In this way, when the TCR controller regulates the system current (
) to zero, the reactive

component of the load current (
) is fully compensated, as follows from equation (7.30).

Figure 7.6 shows the control scheme in the dq frame. Proportionalintegral (PI) controllers
are used in this control algorithm to set the reference for the P controller. Only one of the
two PI controllers is used for setting the reference. The choice depends on the aim of the
regulation (voltage, reactive power or flicker control). The power
and the voltage
represent the instantaneous values of the reactive power and the RMS voltage at the
MV level. The reference current
is compared to the q-component of the system

current
.

Figure 7.6 Basic idea for the design of the control algorithm,

,
(7.28)

,
(7.29)

= 0

= (
)
(7.30)

104
The control algorithm is based on the continuous monitoring and computing of the reactive
current from the power system. The frequency range of 0.5 to 35 Hz should be extracted
from the whole spectrum of the reactive current since this frequency spectrum (with
observable magnitudes starting at less than 1.0 %) represents the main source of the
flicker phenomena [50], [51]. Frequencies outside this range should not influence the
operation of the controller because they could reduce its efficiency to reduce the flicker.
This can be achieved by filtering the input signals of the P controller. In this study filters
based on a fast Fourier transform (FFT) were used.

Based on the actual values of the inputs, the control algorithm calculates the required value
of the susceptance (BTCR) in p.u. A correlation between the susceptance BTCR and the firing
angle is nonlinear and is described by the following equation [52]:

=
2 2 +2
,
(7.31)

where XL represents the maximum reactance of the TCR. By changing the susceptance of
the BTCR, the RMS value of the current through the reactor is changed and can be expressed
as a function of the firing angle as follows:

(2 2 +2),
(7.32)

where Urms is the RMS value of the connected voltage, L is the inductance of the TCR, and
is the angular frequency of the voltage at the connection point [53].

Transfer function of the P controller

The transfer function of the P controller from Figure 7.6 can be expressed as follows:

where
represents the reference value of the reactive current component on the MV level,
while
is the measured value of the reactive component of the total system current (
).
It should be noted that
, as already shown.

Subsequently, the response time of the TCR (TTCR) is limited by the thyristors ignition
delay [52]. Namely, the thyristor phase angle can be varied once during each half-cycle (in
each phase). The ignition delay is not fixed and depends on the instant of the reference
change (thyristor dead-time) and the delay between the sampling instant and the instant
when the thyristor starts to conduct (firing delay). The required frequency of changing the

,
(7.33)

105
value of the parameter b is six times the fundamental frequency for a three-phase 6-pulse
TCR. In order to model the TCR delay, the parameter b should be multiplied by a first-order
block with the time constant TTCR representing the average delay time, as shown in
equation (7.34).

=
1 +
.
(7.34)

Taking into account equation (7.33) we obtain:

=
(1 +
)
.
(7.35)

By using equation (7.35), a direct connection between the controller output and the
mathematical model of the TCR is established. Based on this definition of the parameter b',
and taking intoaccount equations (7.24), (7.25) and (7.35), the following set of equations is
obtained for the TCR current:

=
1
+
1 +
,
(7.36)

=
1
+
1 +
,
(7.37)

where
are the d- and q- components of the measured voltage at the SVC connection
point. This set of equations represents the mathematical model of a TCR controlled with a P
controller. The block diagram of the mathematical model can be seen in Figure 7.7.

106

Figure 7.7 The simplified mathematical model of the TCR with the P controller

7.4 Model testing
The static analysis

In order to analyse the static accuracy, the dynamic accuracy and the stability of this model,
its transfer function has to be determined. By inserting equation (7.36) into (7.37) the
transfer functions of the system shown in Figure 7.7 are calculated as follows (equations
(7.38) and (7.39)):

=
1
+
1 +
1
+
1 +
1
+
.
(7.38)

=
+
2
(1 +2
) +2 +
2
2
,

where:

=
1
, =
1
, =
1
.
(7.40)

x
R'
b
X'
s +
1
R'
b
X'
s +
1
i
d
i
q
K
p
i
q
b
b

X'
b

X'
x
1
1+sT
TCR
'
'
'*
b'
u
d
'
u
q
'
()
=
(7.39)

107
For the determination of the static accuracy of the system, the static error in the stationary
state of the system (s=0) is observed. Theoretically, this error is minimal in the ideal case
when the condition
is fulfilled. In this case, from equation (7.39), the gain Kp of the

P controller can be determined as follows:

()
2
1,
.
(7.41)

Based on equation (7.41) it can be concluded that this controller with a negative feedback
loop in the stationary state is not able to achieve its reference properly, i.e., for every value
of the gain Kp there will be a static error.

If we define the transfer function of the P controller as

equations (7.36) and (7.37) become

=
1
+
1 +
,
(7.43)

=
1
+
1 +
.
(7.44)

By inserting equation (7.43) into (7.44), the transfer function of the system becomes:

()
=
(7.45)

=
+
2
(1 +2
) +
2
+
2
2
.

As previously discussed, for s=0 we have:

()
2
= 1.
(7.46)

for

,
(7.42)

108

= 0.50.
(7.47)

Therefore, for the value of Kp = 0.50 there is no static error. The value of Kp was
determined from the transfer function for the case when the ratio between X1 and R1 is
constant and equal to 10/1.

The dynamic analysis

For the dynamic analysis of the system, the output values in transient conditions have to be
observed. It is very important that the mathematical model describes, as much as possible, in
detail a real TCR. Therefore, the value of the TCR delay time T
TCR
is very important and has to
be selected carefully.

In this work, the value of TTCR was determined based on the real simulation model of a TCR,
which was modelled in the PSCAD simulation software. The step change in the reference
current was initiated and the response of the TCR simulation model to this change was
observed. Based on simulations of the different instants of the reference change and the
different SVC operating points, an average response was estimated. Next, the step response
of the mathematical model described by equation (7.45) for different values of TTCR was
calculated.

Based on a mutual comparison of the average response obtained from the PSCAD
simulation model and the results obtained from the mathematical model, the value of TTCR
was chosen. The selected value gives a response that best fits the response of the
simulation model. In Figure 7.8, line 1 represents the response of the TCR in the PSCAD
simulation model, while lines 2, 3 and 4 represent the response of the mathematical model
for different values of TTCR.

The controller tracks its reference properly, reaches its value and the static error can be
neglected for all cases. Based on this it can be concluded that the P controller with a gain
Kp=0.50 and a value of TTCR=4 ms operates with minimal static error and with a delay time
that corresponds to the average delay time of an actual TCR. The SVC current in the
simulation model also contains harmonics.

109

Figure 7.8 The step response of the system; 1 response of the real TCR model; 2 response of
the mathematical model for the case T
TCR
=6 ms; 3 response of the mathematical model for the
case T
TCR
=4 ms; 4 response of the mathematical model for the case T
TCR
=2 ms

The stability was tested using the Bode stability criterion. According to this criterion, a
closed-loop system will be stable if the value of the gain is less than 1 at the critical
frequency. The critical frequency is equal to -180 degrees. The obtained results are shown
in Figure 7.9. Based on Figure 7.9 it can be concluded that the system will be stable for the
selected parameters Kp and TTCR.

110

Figure 7.9 Bode diagram of F(s)
tr
for values of T
TCR
=0.004 s and K
p
=0.50;

7.5 The simulation results

The SVC control system proposed in this work has been applied for the realistic model of a
steel factory. The electric arc furnace is modelled as described in Chapter 6. The control
function F of the controllable voltage sources is described by (6.17).

Firstly, for modelling the time variation of the electric arc length a sinusoidal function with
different frequencies was used. Frequencies of 5, 9 and 13 Hz were used. For this case,
equation (6.17), which defines the control function of the controllable voltage sources,
becomes:

= 1
(2
)
+
0
+
0
+
,
(7.48)

where fsin can have values of 5, 9 or 13 Hz. Since the frequency of 9 Hz is the worst for the
flicker production, waveform of simulated data only for this frequency will be graphical
presented. For other frequencies the simulated data will only be presented numerically.

After simulations with the sinusoidal function for modelling the time variation of the
electric arc length, simulations with a signal that involves all the important frequencies for

111
flicker were made. The calculation procedure for this signal is described in Chapter 6. The
equation (6.17) describes the control function of the controllable voltage sources used in
this case. In each simulated case the parameters of the electric arc furnace model were
adjusted so that the model generated the same level of flicker (Pst) at 110 kV. The model
was calibrated according to the field measurements.

The system configuration

In Figure 7.10 a simplified single-line diagram of the steel plants network under
investigation is shown. The network model involves a high-voltage network equivalent
with its short-circuit inductance (LLSC) and resistance (RLSC), a 110/35-kV transformer
substation, the resistance (RC) and inductance (LC) of the cable line between the substation
and the furnace transformer, the 35/0.74-kV furnace transformer and inductance (Lf) and
the resistance (Rf) of the cable line between the furnace transformer and the furnace
electrodes.

Figure 7.10 Single-line scheme of the plant network

The values of these parameters are shown in Table 10 and correspond to the actual
parameters of the steel plant. The short-circuit power (Ssc) at the 110-kV bus equals
3200 MVA.

112
Table 10 Values of the parameters of the model shown in Figure 7.10
High-voltage
network equivalent
S
sc
u
n
R/X
3200 MVA
110 kV
0.1
Tr 110/35
S
tr1

u
ntr1

u
ktr1
R/X
100 MVA
110/35 kV
11 %
0.1
Tr 35/0.747
S
tr2

u
ntr2

u
ktr2
R/X
80 MVA
35/0.7473 kV
6.135 %
0.1
cable resistance RC 21010
-3

cable inductance LC 2.810
-3
H
furnace cable
resistance
Rf 4310
-3

furnace cable
inductance
Lf
1.1110
-5
H

The scheme of the SVC device used in these simulations is shown in Figure 7.1. The SVC is
connected on the 35-kV voltage level and its parameters are presented in Table 11.

Figure 7.11 Scheme of the SVC device

Table 11 Parameters of the SVC device
reactor
resistor
filter 2. harmonic
filter 3. harmonic
filter 4. harmonic
L
SVC

R
SVC
FC
2H
FC
3H
FC
4H

0.102 H
3.204
35 MVAr
35 MVAr
35 MVAr

113
7.6 Modelling of the electric-arc length using sinusoidal
function
Simulations without a connected SVC

A simulation of the steel-factory model without the SVC was carried out. A total simulation
time (T) of 20 seconds was chosen, with a sampling time of 15 s. Waveforms of the
simulated signals (frequency of arc length modulation 9 Hz), active (P) and reactive (Q)
powers of the furnace, are shown in Figure 7.12. Pfurnace and Qfurnace represent the active and
reactive powers of arc furnace, respectively, measured on the low-voltage side of the
furnace transformer. The figure shows fluctuations of the active and reactive powers.
These oscillations have a frequency of 9 Hz, and through the furnace transformer are
transferred to the medium-voltage level. Furthermore, over the substation transformer
these oscillations are farther transferred to the 110-kV voltage level.

Figure 7.12 Simulated signals of the active and reactive powers without the connected SVC at
the low-voltage level of the furnace transformer (frequency of the arc-length modulation 9 Hz)

Figure 7.13 and Figure 7.14 show the simulated signals of the active and reactive powers
that are obtained at different voltage levels. Figure 7.13 shows the simulated signal of the
active and reactive powers at the medium-voltage level, while Figure 7.14 shows the same
simulated signal at the 110-kV level. If we compare Figure 7.14 and Figure 7.12 it is clear
that there are differences between the quantities of active and reactive powers
downloaded from the power system and consumed in the arc furnace. This difference
represents the quantity of energy that is required to cover the losses in the transformers
and the cables.

114

Figure 7.13 Simulated signals of the active and reactive powers of the furnace without the SVC
at the 35-kV level (frequency of arc length modulation 9 Hz)

Figure 7.14 Simulated signals of the active and reactive powers without the connected SVC at
the 110-kV voltage level (frequency of the arc-length modulation 9 Hz)

Furthermore, in the next figures (Figure 7.15-Figure 7.22) are the waveforms of other
quantities obtained by simulations. Figure 7.15 and Figure 7.17 show the waveforms of the
voltage and current, respectively, at the medium-voltage level. If we look carefully at Figure
7.15 we can see fluctuations of the peak value of the voltage. These fluctuations can be
easily seen if we look at the waveform of RMS value of this voltage. The waveform of the
RMS voltage value at the medium-voltage level is shown in Figure 7.16.

115

Figure 7.15 Waveform of the voltage at the medium-voltage level without the SVC (frequency of
the arc-length modulation 9 Hz)

Figure 7.16 Fluctuations of the RMS voltage at the medium-voltage level (frequency of the arc-
length modulation 9 Hz)

The current waveforms at the medium-voltage level are presented in Figure 7.17.

116

Figure 7.17 Waveform of the current at the medium-voltage level without the SVC (frequency of

The phase displacement between the voltage and the current at the medium voltage level
without the SVC can be seen in Figure 7.18.

Figure 7.18 Phase shift between the voltage and current at the medium-voltage level without the
SVC

The waveforms of the voltage and the current at the 110-kV level are shown in Figure 7.19
and Figure 7.20. In Figure 7.19 and Figure 7.20 we can see the fluctuations of the peak
values of the voltage and current, respectively, while Figure 7.21 shows the phase shift
between them.

117

Figure 7.19 Waveform of the voltage at the 110-kV level without the SVC (frequency of the arc-
length modulation 9 Hz)

Figure 7.20 Waveforms of the current at the 110-kV level without the SVC (frequency of the
arc-length modulation 9 Hz)

118

Figure 7.21 Phase displacement between the voltage and the current at the 110-kV level without
the SVC

Based on the above, the negative impacts of the electric arc furnace on the power system
can be clearly seen in the figures. The voltage and current fluctuations at the 110-kV
voltage level, the oscillations of the active and reactive power as well as the quantity of
reactive power that withdraws from the power system are at unacceptably high levels. As a
consequence of this, the voltage fluctuations appear as flicker. Figure 7.22 shows the
instantaneous value of the flicker measured at the 110-kV level. A statistical analysis of the
presented signal results in the value of Pst, the value of which is listed in Table 12. Also, in
Table 12 are all the numerical values of P, Q and Pst obtained with these simulations on a
different voltage level and with a different frequency of the arc-length modulation.

Figure 7.22 Instantaneous value of the flicker at the 110-kV level (frequency of the arc-length
modulation 9 Hz)

119
Table 12 Simulated values of the active (P) and reactive (Q) powers and the flicker level (P
st
)
without the connected SVC, for different arc-length modulation
Without SVC
Arc length
modulation
U
35
= 1 p.u. P [MW] Q [MVAr] cos P
st

5 Hz
Network 110 kV 59.77 80.91 0.59
1.55 Network 35 kV 58.17 69.97 0.64
Furnace 0.74 kV 56.70 56.25 0.70
9 Hz
Network 110 kV 59.65 81.43 0.59
1.55 Network 35 kV 57.74 70.62 0.63
Furnace 0.74 kV 56.57 56.75 0.70
11 Hz
Network 110 kV 59.72 81.09 0.59
1.55 Network 35 kV 58.39 70.29 0.63
Furnace 0.74 kV 56.11 56.32 0.70
13 Hz
Network 110 kV 59.63 81.08 0.59
1.55 Network 35 kV 58.29 70.27 0.63
Furnace 0.74 kV 56.03 56.30 0.70

Simulations with connected SVC

Afterwards, the simulation of the realistic model steel factory with the SVC connected at
the 35-kV voltage level was carried out. The parameters of the P and PI controllers used in
the simulations are given in Table 13. The parameters of the PI controller were chosen
such that a first-order response of the system was achieved.

Table 13 Parameters of controllers
Controller Parameter Value
P Kp 0.5
PI (voltage control)
Kp
Ti
10
2.5 ms
PI (reactive power
control)
Kp
Ti
0.001
100 ms

Reactive power and flicker-regulation mode

Firstly, the case when the SVC operates in flicker-compensation mode with different
frequencies of the arc-length modulation was simulated. In Table 14 the numerical values
of P, Q and Pst obtained by simulations for the full load with the connected SVC (for
different frequencies of the arc length) are listed.

120
Table 14 Simulated values with connected SVC for the period t<6 s at the 35-kV level
With SVC - P controller
Arc length
modulation
P [MW] Q [MVAr] cos P
st

5 Hz
Network 110 kV 61.43 5.25 0.99
0.56 Network 35 kV 60.85 0 1
Furnace 0.74 kV 56.47 56.43 0.70
9 Hz
Network 110 kV 61.30 5.22 0.99
0.65 Network 35 kV 60.72 0 1
Furnace 0.74 kV 56.49 56.15 0.70
11 Hz
Network 110 kV 60.57 7.57 0.99
0.76 Network 35 kV 60.75 2.25 0.99
Furnace 0.74 kV 56.24 56.26 0.70
13 Hz
Network 110 kV 61.09 12.58 0.98
0.96 Network 35 kV 60.50 6.693 0.99
Furnace 0.74 kV 56.09 56.16 0.70

Waveforms of the all simulated signals (frequency of arc-length modulation 9 Hz) are
presented in the figures from 7.23 to 7.28. Figure 7.23 shows the waveforms of the active
and reactive powers measured at the low-voltage level of the furnace transformer. For the
period t<6 s the furnace operates with full load. If we compare Figure 7.23 (t<6 s) and
Figure 7.12 we can see that the fluctuations of the active and reactive powers are
approximately the same. Based on this we can conclude that the furnace works at the same
operating point, i.e., the industrial process in the furnace is not changed, irrespective
whether the SVC is connected or not. At the moment t=6 s, the furnace load is changed.

In Figure 7.24 we can see the positive effects of the operation of the SVC controller. We can
see that the fluctuations of the active and reactive powers at the low-voltage side of the
substation transformer (meddle-level) are significantly reduced compared to the
fluctuations of the active and reactive powers that can be seen in Figure 7.13. Also, it is
clear that the quantity of the downloaded reactive power from the power system is
practically negligible. At the moment t=6 s, the furnace load is changed. As can be seen in
Figure 7.24, this change of the furnace load does not have any impact on the proper
operation of the controller, i.e., the controller continues to minimize the fluctuations and
the quantity of the downloaded reactive power from the power system also for the period
t>6.

121

Figure 7.23 Simulated signals of active and reactive powers with connected SVC at the low-
voltage level of the furnace transformer (frequency of the arc-length modulation 9 Hz)

Figure 7.24 Simulated signals of the active and reactive powers of the furnace with the SVC at
the low-voltage level of the substation transformer before and after the change of furnace load
(frequency of the arc-length modulation 9 Hz)

The waveforms of the active and reactive powers at the 110-kV level are presented in
Figure 7.25. It should be noted that there is still downloading of the reactive energies from
the power system. In fact, it is the reactive energy which is spent for magnetizing the core
of the substation transformer. Although there is a consumption of reactive power the
power factor (cos) is within the prescribed limits, as can be seen in Table 14. In most
cases the minimum allowed value of cos is 0.95. If necessary, this quantity of reactive
energy can also be fully compensated using the proposed controller.

122

Figure 7.25 Simulated signals of the active and reactive powers at the 110-kV level downloaded
from the power system with the SVC (frequency of the arc-length modulation 9 Hz)

In Figure 7.26 we can see a waveform of the RMS value of the voltage at the medium-
voltage level. We can see that the fluctuations of the RMS value of the voltage at Figure 7.26
are lower than the fluctuations that can be seen in Figure 7.16 (the case when the
controller does not operate). As a direct consequence of the lower voltage fluctuations,
there is lower level of flicker at 110 kV.

Figure 7.26 Fluctuations of RMS voltage at the medium-voltage level with the connected SVC

At the moment t=6 s when the furnaces load is changed, as a consequence of a decreased
active power there is a lower value of the current. This lower value of the current means a

123
lower voltage drop on the transformers and consequently there is an increase in the RMS
value of the voltage, Figure 7.26.

The current waveforms at the medium-voltage level with the connected SVC can be seen in
Figure 7.27. The transitional process is done quickly and without introducing an
overcurrent. Although the arc furnace operates at the same operating point as in the case
without the connected SVC, (same level of active and reactive powers for t=6 s), now the
peak values of the current are significantly smaller than the peak values that can be seen in
Figure 7.17. This means the arc furnace now operates more efficiently, and cos is higher.

Figure 7.27 Waveform of current at the medium-voltage level during the transient process
reactive power and flicker mode (frequency of the arc-length modulation 9 Hz)

The phase shift between the voltage and the current at the 110-kV level before and after
the change of the furnace's load is shown in Figure 7.28. We can see that the phase shift is
constant, i.e., independent of the load of furnace it remains unchanged. The amplitude of
the voltage is given in kV, while the amplitude of the current is given in kA.

124

Figure 7.28 Phase displacement between the voltage and the current at the 110-kV level

The waveforms of the thyristors ignition angle can be seen in Figure 7.29.

Figure 7.29 Values of the tyristor's ignition angle, , the reactive power and the flicker-
regulation mode (frequency of the arc-length modulation 9 Hz)

Furthermore, the figures from Figure 7.30 to Figure 7.32 present the waveforms of the
active and the reactive powers of the SVC device. Here, it should be mentioned that the
direction of the power is defined as being positive when the power flows from the SVC
device to the arc-furnace transformer (shown in Figure 7.10). As a consequence of that, we
have negative values of the active power of the SVC and TCR. The negative value of the
active power means that the SVC and TCR consume active power from the system and it
actually represents the active losses in the SVC device. In Figure 7.30 we can see the total

125
power losses of the SVC device, and the power losses that occur in the TCR and filters. The
following figure, Figure 7.31, shows the losses incurred in each filter.

Figure 7.30 Active power (losses) of SVC and TCR (frequency of the arc-length modulation
9 Hz)

Figure 7.31 Active power (losses) of the filters (frequency of the arc-length modulation 9 Hz)

The waveforms of the total reactive power that the SVC device exchanges with system can be
seen in Figure 7.32. Also, in the figure we can see the waveforms of the reactive power of the
filters and the reactive power of the TCR. The capacitive reactive power has a positive sign,
while the inductive reactive power has a negative sign.

126

Figure 7.32 Simulated signal of the reactive power of the SVC, TCR and FC (frequency of the
arc-length modulation 9 Hz)

Voltage-regulation mode

Finally, the case when the SVC operates in voltage-regulation mode was simulated. Like in
the mode of reactive power and flicker regulation, for the period t<6 s the furnace operates
with a full nominal load and at the moment t=6 s the load of the furnace is changed. Figure
7.33 shows the waveforms of active and reactive powers at the medium-voltage level
downloaded from the power system in this case.

Figure 7.33 Simulated signals of the active and reactive powers at the medium-voltage level
downloaded from the power system (frequency of the arc-length modulation 9 Hz)

127
At the moment t=6 s the load of the furnace is changed and as a consequence of this
reduced load, there is an increase in the voltage, Figure 7.34. In Figure 7.34 we can see that
the system reaches the desired value (in this case 1 p.u.) of the voltage after a change of the
load for a short period. If we compare Figure 7.34 and Figure 7.26, it is clear that the
fluctuations of the RMS voltage value in both figures are almost the same, i.e., the flicker
level is almost the same. Based on this it can be concluded that the SVC with this controller
when it operates in voltage-regulation mode also has the ability to reduce the level of
flicker.

Figure 7.34 Fluctuations of RMS voltage at the 35-kV level with the connected SVC, voltage
regulation (frequency of the arc-length modulation 9 Hz)

The current waveforms which system draws from power system at the medium-voltage
level with the connected SVC can be seen in Figure 7.35. Also in this case, like in the
reactive-power and flicker-regulation mode, the transitional process is done quickly and
without an overcurrent occurring.

128

Figure 7.35 Waveform of the current which system draws from power system at the medium-
voltage level during the transient process voltage-regulation mode (frequency of the arc-length
modulation 9 Hz)

The waveforms of the thyristors ignition angle can be seen in Figure 7.36.

Figure 7.36 Values of the tyristor's ignition angle , the voltage-regulation mode (frequency of

129
7.7 Modelling of the electric-arc length using a signal that
involves the all-important frequencies for flicker

As in the previous section, firstly the simulation of the realistic model of the steel factory without
the SVC was carried out. All the simulation parameters are unchanged, except the function
which is used for modelling the arc length. Also, the procedure for the simulation is the same.
Frist, simulations without the connected SVC were made. After these simulations, simulations
with the connected SVC were made. During the simulating process with the connected SVC at
the moment t=6 s the furnaces load is changed.

Simulations without the connected SVC

In Table 15 the numerical values of P, Q and P
st
obtained by simulations for the full load without
the connected SVC are listed. If we compare the data from Table 15 with the data in Table 12 we
can see that the active and reactive powers and the power factor have approximately the same
values. This means that the furnace works at the same operating point.

Table 15 Simulated values of the active (P) and reactive (Q) powers and
the flicker level (P
st
) without the connected SVC
Without SVC - P controller

st

Network 110 kV 58.89 79.28 0.60
1.55 Network 35 kV 57.68 69.34 0.64
Furnace 0.74 kV 56.07 56.77 0.70

In the next figures are all the signals obtained with these simulations. In Figure 7.37 we can
see the simulated signals of the active and reactive powers at the low-voltage level of the
furnace transformer. Now, we can see that the fluctuations are higher and include more
than one frequency. For this case it can be said that it is more realistic than the case that
includes only one frequency.

130

Figure 7.37 Simulated signals of the active and reactive powers without the SVC at the low-
voltage level of the furnace transformer

The simulated signals of the active and reactive powers that the system withdraws from
the power system can be seen in Figure 7.38.

Figure 7.38 Simulated signals of the active and reactive powers without the SVC at the 110-kV
level

The voltage fluctuations at the medium-voltage level can be seen in Figure 7.39, while the
fluctuations of the current which system draws from power system at the same voltage
level are presented in Figure 7.40. The waveform of the RMS voltage value at the medium-
voltage level is shown in Figure 7.41.

131

Figure 7.39 Waveform of the voltage at the medium-voltage level without the SVC

voltage level without the SVC

132

Figure 7.41 Fluctuations of the RMS voltage at the medium-voltage level

The waveforms of the voltage and the current at the 110-kV level are presented in Figure
7.42 and Figure 7.43.

Figure 7.42 Waveform of the voltage at the 110-kV level without the SVC

133

Figure 7.43 Waveforms of the current at the 110-kV level without the SVC

Figure 7.44 shows the waveforms of the instantaneous value of the flicker obtained at the
110-kV level.

Figure 7.44 Instantaneous value of the flicker at the 110-kV level

The waveforms of the total active power (losses) of SVC can be seen in In Figure 7.45, while
Figure 7.46 shows the reactive power that the SVC device exchanges with system for this
case.

134

Figure 7.45 Active power of the TCR and the SVC

Figure 7.46 Reactive power of the FC, the TCR and the SVC

Reactive power and flicker regulation mode

In this section are the results of simulations with the connected SVC. In Table 16 the
numerical values of P, Q, cos and Pst obtained by simulations for the periods t<6 s and
t>6 s with the connected SVC are listed. At the moment t=6 s, the furnaces load is changed.

135
Table 16 Simulated values with the connected SVC for the period at MV level reactive power
and flicker-regulation mode
With SVC - P controller
t < 6 s t > 6 s

st
st

Network 110 kV 61.78 5.00 0.99
0.80
49.11 3.35 0.99
0.68 Network 35 kV 61.24 0 1 48.77 0 0.99
Furnace 0.74 kV 56.03 57.07 0.70 41.60 46.18 0.67

Figure 7.47 shows the simulated signals of the active and reactive powers with the
connected SVC at the low-voltage level of the furnace transformer before and after the
change of the furnaces load. Like in all the simulations before, the furnace operates at the
same operating point.

Figure 7.47 Simulated signals of the active and reactive powers with the connected SVC at the
low-voltage level of the furnace transformer

In Figure 7.48 we can see the simulated signals of the active and reactive powers with the
SVC at the low-voltage level of the substation transformer before and after the change of
the furnaces load. From the figure we can see that the change of the furnaces load does not
have an influence on the work of the controller. The controller continues to minimize the
fluctuations and the quantity of the downloaded reactive power from the power system,
independent of the level of the load.

136

Figure 7.48 Simulated signals of the active and reactive powers of the furnace with the SVC at
the low-voltage level of the substation transformer before and after the change of the furnaces
load

Simulated signals of the active and reactive powers at the 110-kV level downloaded from
the power system with the SVC are shown in Figure 7.49. In this case there is also the
downloading of reactive power from the power system.

Figure 7.49 Simulated signals of the active and reactive powers at the 110-kV level downloaded
from the power system with the SVC

In Figure 7.50 we can see the waveform of the RMS value of the voltage at the medium-
voltage level. We can see that the fluctuations of the RMS value of the voltage in Figure 7.50
are lower than the fluctuations that can be seen in Figure 7.41 (the case when the
controller does not operate).

137

Figure 7.50 Fluctuations of the RMS voltage at the medium-voltage level with the connected
SVC

The waveforms of the currents at the medium-voltage level during the transient process
are presented in Figure 7.51.

Figure 7.51 Waveforms of the currents at the medium-voltage level during the transient process
reactive power and flicker mode

Figure 7.52 shows the instantaneous value of the flicker at the 110-kV level.

138

Figure 7.52 Instantaneous value of the flicker at the 110-kV level

Voltage-regulation mode

In this section are the figures with the waveforms of the signal obtained by simulations in
the voltage-regulation mode. In Figure 7.53 we can see the RMS voltage fluctuations at the
MV level. At the moment t=6 s the load of furnace is changed. From Figure 7.53 it is clear
that the controller operates correctly before and after the furnaces load is changed. In
Figure 7.54 we can see the waveforms of the currents which system draws from power
system at MV level and it is clear that the transitional process is done quickly and without
any overcurrent occurring.

Figure 7.53 Fluctuations of the RMS voltage at the 35-kV level with the connected SVC, voltage
regulation

139

voltage level during the transient process voltage-regulation mode

Figure 7.55 presents the simulated signals of the active and reactive power (at the low-
voltage level of the substation transformer) downloaded from the power system before
and after the change of the furnaces load.

Figure 7.55 Simulated signals of the active and reactive powers (medium-voltage level)
downloaded from the power system

In Figure 7.56 we can see the waveform of the tyristor's ignition angle , before and after
the furnaces load.

140

Figure 7.56 Values of the thyristor's ignition angle , voltage-regulation mode

7.8 Simulation with another frequency spectrum

In order to verify the accuracy of the designed controller model, in this section the
numerical results of additional simulations are presented. All the parameters of the
simulation model are unchanged, except for the law of the arc-length variation. The
calculate procedure of the law of the arc-length variation is identical to the procedure
described in Chapter 6. The only difference is that the starting point for the calculation of
the law of the arc-length is changed, i.e., a different measured signal (signal obtained by
real measurement at other arc furnace location) is taken. According to the previously
described procedure, a representative sample from the measured signal was calculated and
selected. Furthermore, from that representative sample the law of the arc-length was
calculated. Figure 2 shows the waveform of the electric arc length obtained using this
simulation. If we compare Figure 6.22 with Figure 7.57, it is clear that these are two
different arcs with a different frequency spectrum. Also, we can see that the length of arc l1
is longer than the length of arc l shown in Figure 7.57.

In Table 17 the numerical values of P, Q, cos and Pst obtained by simulations for this case
with and without the connected SVC are listed. We can see that the active and reactive
powers and the power factor of the furnace have approximately the same values. This
means that the furnace works at the same operating point. The active power that the
system draws from the power system now is larger. That is a consequence of the existing
losses in the SVC device. In addition, we can see the difference between the power factor
and the level of flicker. With the connected SVC the power factor is almost equal to 1, while
the flicker level is lower than 1. Based on this we can conclude that in this case the
proposed controller also works correctly.

141

Figure 7.57 The waveform of the electric arc length

Table 17 Simulated values with the second law of the arc-length
Without connected SVC

st

Network 110 kV 59.32 79.48 0.59
1.55
Furnace 0.74 kV 56.23 56.77 0.70
With connected SVC
Network 110 kV 61.96 5.21 0.99
0.80
Furnace 0.74 kV 56.25 57.00 0.70

7.9 Assessment of the controller characteristics

In the first part of the experimental work we made a simulation with only one frequency of
the arc length. Observing the data in Table 14 it can be concluded that the proposed
controller has the capabilities to fully compensate for the reactive power, the power factor
(cos) and reduce the level of flicker to an acceptable level for each modulation of the arc
length. In Figure 7.58 the flicker levels for different modulations of the arc length can be
seen. Based on this it can be concluded that increasing the frequency modulation, the
capabilities of the controller to reduce the flicker level are reduced, i.e., the controller has
more difficulty in following the fluctuations with higher frequencies. It is widely known
that when using the SVC the level of the flicker can be reduced by a factor of up to 2. On the
basis of this and the data shown in Figure 7.58 it can be concluded that a frequency of 11
Hz for the modulation length of the arc is more useful in studies than a frequency of 9 Hz.
Although the human eye is more sensitive to the frequency of 9 Hz, the frequency of 11 Hz
gives more realistic results. With a frequency of 9 Hz the reduction factor of the flicker is
approximately 2.5, which is almost impossible to achieve in practice.

142

Figure 7.58 Simulated values of the flicker level with the connected SVC for different
modulation frequencies (for P
furnace
56.21 MW)

In the second part of the experimental work, the length of the arc is modulated by a signal
that included the full spectrum of characteristic frequencies for the occurrence of the
flicker. Based on the data obtained with these simulations, whose numerical values are
presented in Table 16, we can see that in this case, the controller successfully minimizes
the negative influences of the EAF on the power system. If we compare the waveforms
shown in the figures obtained by the simulations when we have only one frequency for the
modulation of the arc length with the figures when we have the full spectrum of
characteristic frequencies we can see that there is the same trend of behaviour for the
controller. The ability to compensate for the reactive power and voltage control in both
cases can be realized. Only that the compensation of the flicker level for some individual
frequencies is better than in the case if we have the whole spectrum of characteristic
frequencies. Also, we can see that in both cases the transition process in the EAF (a sudden
change of power) does not affect the stable operation of the controller.

In Figure 7.58 we can see the level of flicker with and without the connected SVC device
with the first law of the arc-length. The level of flicker with and without the connected SVC
device with the second law of the arc-length is same as shown in Figure 7.58.

We can see that the correction factor of the level of flicker in both cases is equal to 1.92. As
already mentioned, using the SVC the level of the flicker can be reduced by a factor of 2.
Based on all this, we can conclude that by using the proposed controller we can achieve
practically the maximum efficiency of the SVC, regardless which spectrum produces
electric arc furnace at the connection point. If we compare the pictures 7.58 and 7.59 we
can see that modelling the electric arc length with a frequency of 11 Hz produces
acceptable results for the compensation, i.e., the results obtained in this way are

143
approximately the same as the results obtained using simulations that included the whole
spectrum of characteristic frequencies. It is important to note that the active power of the
electric arc furnace in all the simulations remains the same.

Figure 7.59 Simulated values of the flicker level with and without the connected SVC; the arc
length modulated with first law of the arc-length to occur

144
8 Conclusion

This doctoral dissertation deals with the negative impacts of electric arc furnaces in the
power system. Generally, we know that the EAFs represent large consumers of electrical
energy, which because of their nonlinear characteristics have a strong feedback influence
on the power quality at the electric power system, and as such require the installation of
compensation devices. In the development and design process of compensating devices an
important role is played by a simulation model of the EAF. The process of developing and
designing the compensation devices for the EAF represents a major problem. In particular,
this problem is expressed (because of the non-linear electrical characteristics) when we
need to develop and design the simulation model of the EAF.

This dissertation is divided into two parts: first is the theoretical part and second is the
research-development part of the work. The theoretical part includes the first 5 chapters
and presents the existing situation and issues, while the research-development part of this
work is contained in the last three chapters, and in it are the results of the research work
and solutions to some existing problems. Also, the research-development part of the
dissertation includes the contributions to science. The key contributions to science of this
doctoral dissertation are from the fields of the development and the design of the real
model of the EAF and the control algorithm for the compensation device in order to
eliminate the negative impacts of the EAF on the power system.

In this doctoral dissertation a new method for modelling electric arc furnaces based on
representative samples of voltages and currents is presented. The proposed arc-furnace
model provides the waveforms of the voltage and current as well as the characteristics
(spectrum of interharmonics, flicker level) that are almost equal to the waveforms in the
real plant.

The proposed method is based on a deterministic approach for the generation of the arc
length. The arc-length variation is derived from the measured interharmonic spectrum of
an arc furnace. An advantage of the approach proposed in this thesis is that it can generate
a full set of frequencies that cause the flicker (i.e., interharmonics) which are generated by
the furnace.

The new model can be used for a comprehensive analysis of the influence of the electric arc
furnace on the network, as it produces similar conditions in the network to the actual
furnace. The proposed model is general and can be used for all arc furnaces. However,
furnaces differ substantially, not only due to their different sizes, but also due to the
different melting processes, different electrode regulators and different connections to the
network. Therefore, for an accurate representation of a particular arc-furnace,
measurements are needed for the calibration of the arc-furnace model. Generally, it would

145
be possible to group the furnaces based on their effect on the network. This would require
an analysis of a larger number of furnaces and is a goal of our future work.

The simulation results showed that the modelling of an electric arc length with a frequency
of 11 Hz produces results that are comparable with the results obtained by simulations in
the case when the length of electric arc is modelled with the whole spectrum of
characteristic frequencies.
The research-development part of this doctoral dissertation relates to the development
and implementation of an SVC controller. In this part of the doctoral dissertation we
presented a method for developing and implementing an SVC controller in a feedback-loop
strategy. The presented controller is based on the transfer function of the TCR
mathematical model, and was designed in the d-q coordinate system. The proposed
controller was designed with the objective being to eliminate the negative impact caused by
electric arc furnaces (EAFs) on a power system.

A mathematical model of the proposed SVC controller in the d-q synchronous rotating
coordinate system (SRCS) was developed first. Then, a completely stability analysis of the
system (analysis in dynamic and steady-state conditions) was carried out. In the end, the
efficiency of the presented controller is demonstrated by means of computer simulations of an
actual steel-factory network model.

The major advantages of the proposed controller are the better flicker compensation, the
ability to regulate the voltage and the requirement for only single-point network
measurements. Using an actual industrial network model, the proposed controller was
validated by means of simulations. The simulation results showed that the proposed model
can be used successfully to compensate for the reactive power, reduce the flicker level or to
regulate the voltage. The main advantages of the proposed controller are good flicker
compensation, the ability to regulate the voltage and that its implementation requires only
a single point for the network measurement. A single measurement point also means a
lower price for the installation of the system and fewer possibilities for failure.

The control algorithm proposed in this thesis can also be applied to other types of FACTS
devices, such as a STATCOM or thyristor-controlled series compensator. The control
algorithm was developed for positive system currents. The arc-furnace unbalance will be
investigated in future work. Also, future work on the subject will include an evaluation of
the proposed controller in a real-time digital simulator (hardware in the loop).

146
9 References

[1] Standard SIST EN 50160:2011: Voltage characteristics of electricity supplied by public
electricity networks.

[2] Standard SIST EN 61000430:2003: EMC Part 4, Section 30 Testing and
measurement techniques Power quality measurement methods.

[3] M. Maksi, I. Papi, Calculating flicker propagation in a meshed high voltage network
with interharmonics and representative voltage samples, Electrical Power and Energy
Systems 42 (2012) 179187.

[4] Flicker Measurement and Evaluation, UIE publication, 1992

[5] IEC 6100033, Edition 1.1: Electromagnetic compatibility (EMC) Part 33: Limits
Limitation of voltage changes, voltage fluctuations and flicker in References 168 public
lowvoltage supply systems, for equipment with rated current <=16 A per phase and not
subject to conditional connection.

[6] IEC/TR 61000415 Standard, EMV Testing and measuring techniques, Flickermeter
Functional and design specifications.

[7] IEEE standard 141 1993: IEEE recommended practice for electric power distribution
for industrial plants, 1993.

[8] Cai R.: Flicker Interaction Studies and Flickermeter Improvement, doctoral dissertation,
Technische Universiteit Eindhoven, 2009.

[9] S. Mendis, M. Bishop, and J . Witte. Investigations of Voltage Flicker in Electric Arc
Furnace Power Systems. IEEE Industry Applications Magazine, J an./ Febr. 1996.

[10] P. Ashmole. Quality of Supply-voltage fluctuations - part 2. Power Engineering
J ournal, Apr. 2001.

[11] Z. Zhang, N. Fahmi and W. Norris. Flicker Analysis and Methods for Electric Arc
Furnace Flicker Mitigation (A Survey). 2001 IEEE Porto Power Tech Conference,
Porto, Portugal, Sept. 2001

[12] T. Thiringer, T. Petru and S. Lundberg. Flicker contribution from wind turbine
installations, IEEE Trans. Energy Conv. vol. 17, pp. 114118, Mar. 2002.

[13] J . Tande and K. Uhlen. Wind Turbines in Weak Grids Constraints and Solutions.
16th Int. Conf. On Electricity Distribution, CIRED 2001, J une 2001.

[14] . Larsson. Flicker Emission of Wind Turbines during Continuous Operation. IEEE
Trans. Energy Conv. vol. 17, pp. 114-118, Mar. 2002

147
[15] . Larsson. Flicker Emission of Wind Turbines during Continuous Operation. IEEE
Trans. Energy Conv. vol. 17, pp. 114-118, Mar. 2002.

[16] M. Maksi, Calculating flicker propagation in a meshed high voltage network with
interharmonics and representative voltage samples, doctoral dissertation, Faculty of
Electrical Engineering, University Ljubljana 2011

[17] A. Baggini, Handbook of Power Quality, Wiley, J une 16, 2008

[18] H. Andrei1, C. Cepisca, S. Grigorescu, Power Quality and Electrical Arc Furnaces,
Politehnica University of Bucharest, Romania

[19] J. Vori, J. Bratina, Electrotermija, Univerza v Mariboru, Maribor 2000

[20] Y. Yang, Y. Xiao and M.A. Reuter, Analysis of transport phenomena in submerged arc
furnace for ferrochrome production, Delft University of Technology, Dept. of Applied
Earth Sciences Mijnbouwstraat 120, 2628 RX Delft, The Netherlands.

[21] A. H. Howatson, An Introduction to Gas Discharges, Pergamon Press, Oxford

[22] R. M. Mathur, R. K. Varma, Thyristor-based FACTS controllers for electrical
transmission systems, ISBN 0-471-20643-1, 2002

[23] D. Lagator: Kontrola napona i jalove snage, seminarski rad, sveuilite Osjek 2011

[24] N. G. Hingorani, L. Gyugyi, Understanding FACTS concepts and technology of flexible
AC transmission systems, ISBN 0-7803-3455-8, 1999

[25] Tang, L., Kolluri S., McGrangghan, M. F.: Voltage flicker prediction for two
simultaneously operated ac arc furnaces, IEEE Trans. Power Delivery, Vol. 12, No. 2,
1997.

[26] Montanari, G. C., Loggini, M., Cavallini, A., Pitti, L., Zaninelli, D.: Arc-furnace model
for the study of flicker compensation in electrical networks, IEEE Trans. Power
Delivery, Vol. 9, No. 4, 1994.

[27] Varadan, S., Makram, E. B., Girgis, A. A.: A new time domain voltage source model for
an arc furnace EMPT, IEEE Trans. Power Delivery. Vol. 11, No. 3, 1996

[28] Mayordomo, J . G., Asensi R.: A Frequency Domain Arc Furnace Model For Harmonic
Power Flows Under Balanced Conditions, Proceedings of the IEEE 7th International
Conference on Harmonics and Quality of Power, Las Vegas, pp. 419-27, 16-18 Oct.
1996.

[29] Mayordomo, J . G., Beites, L. F., Asensi, R., Izzeddine, M., Zabala, L., Amantegui, J .: A
new frequency domain arc furnace model for iterative harmonic analysis, IEEE Trans.
Power Delivery, Vol. 12, No. 4, 1997.

148
[30] Ozgun, O., Abur, A.: Development of an arc furnace model for power quality studies,
Power Engineering Society Summer Meeting, IEEE Vol. 1, pp. 507 511, 18-22 J uly
1999.

[31] Ozgun O., Abur A.: Flicker study using a novel arc furnace model, IEEE Trans. Power

[32] Zhang L., Liu Y., Ingram M. R., Bradshaw D. T., Eckroad S., Crow M. L.: EAF Voltage
Flicker Mitigation By FACTS/ ESS, Power System Conference & Exposition, Vol. 1,
pp. 372-378, Oct. 2004.

[33] ONeill-Carrillo, E., Heydt, G.T., Kostelich, E.J ., Venkata, S.S., Sundaram, A.:
Nonlinear deterministic modelling of highly varying loads, IEEE Trans. Power

[34] Ozgun, O., Abur, A.: Flicker study using a novel arc furnace model, IEEE Trans.

[35] Maksi, M., Papi, I.: Analysis of flicker propagation with representative samples of
network voltage, IEEE Trans. Power Delivery, Vol. 26, No. 3, 2011.

[36] Maksi, M., Papi, I.: Calculating flicker propagation in a meshed high voltage network
with interharmonics and representative voltage samples, Electrical Power and Energy
Systems, Vol. 42, 2011.

[37] Chang, G. W., Chen, C. I. A.: A Comparative Study of Voltage Flicker Envelope
Estimation Methods, Conference Publications IEEE, pp. 1-6, 20-24 J uly 2008.

[38] IEEE 1453-2004: Recommended Practice for Measurement and Limits of Voltage
Fluctuations and Associated Light Flicker on AC Power Systems, 2005.

[39] H. Samet, M.E. Hamedani Golshan, A wide nonlinear analysis of reactive power time
series related to electric arc furnaces, Electrical Power and Energy Systems 36, 127-134,
2012.

[40] J . Wang, C. Fu, Y. Zhang, SVC Control System Based on Instantaneous Reactive Power
Theory and Fuzzy PID, IEEE Trans. Industrial Electronics, Vol. 55, No. 4, April 2008.

[41] Q. Zhao, J . J iang, Robust SVC controller design for improving power system damping,
IEEE Trans. Power Systems, Vol. 10. No. 4, 1995.

[42] M. C. Lima, H. Tyll, An overview of static Var compensators technology evolution on a
Brazilian generation and transmission utility, Proc. IEEE 10th Int. Conf. Harmonics
Quality Power, Vol. 2, pp. 601606. 69 Oct. 2002.

149
[43] A. F. Zobaa, M. A. Jovanovi, A comprehensive overview on reactive power
compensation technologies for wind power applications, Proc. IEEE 12th Int. Power
Electron. Motion Control Conf., pp. 18481852. Aug. 2006.

[44] H. Samet, M. Parniani, Predictive method for improving SVC speed in electric arc
furnace compensation, IEEE Trans. Power Del., Vol. 22, No. 1, pp. 732734, J an. 2007.

[45] M. Noroozian, M. Ghandhari, G. Andersson, A robust control strategy for shunt and
series reactive compensators to damp electromechanical oscillations, IEEE Trans.

[46] C. Cheng, Y. Y. Hsu. Damping of generator oscillations using an adaptive static VAr
compensator, IEEE Trans.Power Systems, Vol. 7, No. 2, 1992.

[47] X. Du, L. Peng, L. Zhou. A hybrid reactive power compensation system consisting of
multi capacitor banks and a SVG, Power System Technology, Vol. 32, No. 14, 2008.

[48] E. H. Mamdani, Application of fuzzy algorithm for control of simple dynamic plant,
Procedings of IEEE, 121(12): 1585 1588. 1974

[49] T. Abdelazim, O. P. Malik, Intelligent SVC control for transient stability enhancement,
IEEE Power Engineering Society General Meeting. New York: IEEE, 1701 1707.
2005.
[50] C.J . Wu, T.H. Fu, Effective voltage flicker calculation algorithm using indirect
demodulation method, Power, Energy, & Industry Applications, Vol: 150, July 2003.

[51] J . W. Smith, Voltage Flicker Primer, Submitted to the IEEE SCC 21 P1547 Working
Group, 17. J une 1999.

[52] R. M. Mathur, R. K. Varma, Thyristor-based FACTS controllers for electrical
Transmission systems, IEEE Press, pp. 45-50

[53] A. Garca-Cerrada, P. Garca-Gonzlez, R. Collantes, T. Gmez, J . Anzola, Comparison
of Thyristor-Controlled Reactors and Voltage-Source Inverters for Compensation of
Flicker Caused by Arc Furnaces, IEEE Trans. Power Delivery, Vol. 15, No. 4, Oct
2000.

150

Voltage Fluctuation in Industrial Network and Compensation Measures

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Voltage Fluctuation in Industrial Network and Compensation Measures

Uploaded by

Copyright:

Available Formats

UNIVERSITY OF LJ UBLJ ANA

Faculty of Electrical Engineering

components of the current

components of the current

components of the voltage

components of the voltage

), the amplitudes can be calculated using the following equations:

are the system currents (currents from the

are the phase-to-neutral voltages at the point of common

are the currents through the reactors in the delta equivalent

are the base values of the current, voltage and impedance,

is the synchronous angular speed of the fundamental network

) of the current that flows into the

are constant, which is convenient for the

can be generated by suitable values of the factor b. In fact, the

) should be established. The system current

) and of the SVC device (

) to zero, the reactive

) is fully compensated, as follows from equation (7.30).

and the voltage

is compared to the q-component of the system

is fulfilled. In this case, from equation (7.39), the gain Kp of the

You might also like