Propagation of Transmission Level Switching Overvoltages PDF

Propagation of transmission level switching overvoltages,
related to the distribution level
a product of :
Projectgroup - EPSH-810
Institute of Energy Technology
Aalborg University, AAU
31st may 2007

Pontoppidanstrde 101
9220 Aalborg st
Telephone +45 96 35 92 40
http://www.iet.aau.dk
Title:
Theme:
Control and Surveillance
of Electrical Power Systems
Project period:
1st feb 2007 - 31st may 2007
Group:
EPSH-810
Members:
Vir Mr Atlason
Jakob Kessel
Supervisor:
Claus Leth Bak
Publications: 5
Number of pages: 193
Finished: May 31st 2007
SYNOPSIS:
This project focuses on the problem with overvoltages, generated by reconnection of a grid segment in a transmission system. In order to evaluate this problem, the 150 kV grid segment between Ferslev and Tinghj is examined which is
a combination of two underground cable sections,
one overhead line section and one shunt reactor.
Measurements are performed on the grid segment between Ferslev and Tinghj, manifesting
the presence of switching overvoltages. An analysis of these overvoltages is performed. First, the
frequencies experienced after switch-on are evaluated theoretically. Thereafter, a simplified model
of the 150 kV grid is analyzed mathematically,
before a model of the system is implemented in
the simulation software PSCAD.
Both the mathematical analysis and the simulation in PSCAD shows lack of damping compared
with the measurements on the system. This difference is considered to originate from an insufficient simulation model for the reactor component
and frequency independence of the short circuit
resistance.
Two proposals for minimizing the switching overvoltages on the 150 kV level are given: the zero
crossing closure principle and the pre-insertion
resistor principle. Simulation of both proposals
shows limitation of the experienced overvoltages
to a maximum of 6 kV, compared with a maximum of 64 kV without transient limitation.
From an evaluation of the switching overvoltages
in the system between Ferslev and Tinghj, it was
concluded that such solution is not necessary.
Institut for Energiteknik

Aalborg Universitet, AAU
Institut for Energiteknik
Pontoppidanstrde 101
9220 Aalborg st
Telefon 96 35 92 40
http://www.iet.aau.dk
TITEL:
Udbredelse af koblingsoverspndinger p transmissionsniveau
relateret til distributionsniveau
TEMA:
Regulering og overvgning
af elektriske anlg
Projekt periode:
1. feb 2007 - 31. maj 2007
Gruppe:
EPSH-810
Medlemmer:
Vir Mr Atlason
Jakob Kessel
Vejleder:
Claus Leth Bak
Oplag: 5
Antal sider: 193
Afsluttet: 31. maj 2007
SYNOPSIS:
Dette projekt fokuserer p problemet omkring
overspndinger forrsaget af genindkobling af
en transmissionslinje. Til behandling af problemet, undersges 150 kV linjen mellem Ferslev
og Tinghj, bestende af to kabelstykker, et
linjestykke samt en reaktor.
Mlinger p linjen mellem Ferslev og Tinghj,
viser tilstedevrelsen af koblingsoverspndinger.
Der foretages derfor en analyse af disse overspndinger. Frst foretages en teoretisk evaluering af de frekvenser der optrder efter indkoblingen. Herefter udfres en matematisk analyse af en
simplificeret model for 150 kV linjen, fr en model
for systemet implementeres i simuleringsprogrammet PSCAD.
Bde den matematiske analyse samt simuleringen
i PSCAD viser mangel p dmpning, sammenlignet med mlinger p systemet. Denne forskel
antages at stamme fra en utilstrkkelig model for
reaktoren, samt manglende frekvensafhngighed
i systemets kortslutningsmostand.
Der opstilles to forslag til minimering af indkoblingsoverspndingerne p 150 kV linjen:
et omhandlende nulgennemgangen for spndingskurverne samt et omhandlende en opladningsmodstand. Simuleringer af de to forslag
viser en begrnsning af overspndingerne til
maksimalt 6 kV, sammenlignet med 64 kV uden
transient begrnsning.
Ud fra en evaluering af overspndingerne p system mellem Ferslev og Tinghj konkluderes det,
at en begrnsning af overspndingerne ikke er
ndvendig her.
Preface
This report is prepared as an 8th semester study by project group EPSH-810 on Institute of
Energy Technology, Aalborg University. The report is produced in the period from the 1st of
February to the 31th of May 2007, under the project theme Control and Surveillance of Electrical
Power Systems.
The report contains eight chapters, and eight appendices, supporting the report. Each chapter
is structured by the IMRaD model. Thereby, each chapter starts with a Introduction where
the objectives are presented, then Methods and Results are stated, and Discussion of the most
significant findings given, leading over to next chapter.
Citations are given in square brackets, first with a unique number followed by name of the author
and the used page number(s), e.g. [1, S. Vrts, p. 33].
The attached CD, contains files used for calculations and simulations, articles referred to in the
report, data regarding the shunt reactor and transformers, and the report self in .pdf and .ps
form. Details of the CD content, are given on page H-65.
The authors wish to express appreciation to the following:
Kim Sgaard and Lars Rasmussen from Energinet.dk and NV Net respectively, for informations
regarding different aspects of the reactor, the line sections and the substations, David Eklund
from ABB for quickly responding to request for informations regarding the shunt reactor, and
Stefan Srensen from HEF for good response regarding data for transformers.
Vir Mr Atlason
Jakob Kessel
Contents
1
Introduction
Problem analysis
2.1
Problem statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2
Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Description and modelling of the 150 kV grid
3.1
Main system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2
150 kV source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3
Circuit breaker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12
3.4
Underground cable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12
3.5
Overhead transmission lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
3.6
Shunt reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
20
3.7
Transformers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
26
3.8
Three phase representation of the grid
. . . . . . . . . . . . . . . . . . . . . . .
29
3.9
Summary & discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
30
Analysis of switching overvoltages on the 150 kV grid
31
4.1
Estimation of the transient behaviour . . . . . . . . . . . . . . . . . . . . . . . .
31
4.2
Simplified model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
34
4.3
Mathematical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
37
4.4
PSCAD simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
44
4.5
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
46
4.6
54
Configuration and simulation in PSCAD
57
5.1
PSCAD model subparts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
57
5.2
Verification of reactor model . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
68
5.3
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
71
5.4
80
viii
CONTENTS
Evaluation of the PSCAD simulation
83
6.1
Estimation of cable and line models . . . . . . . . . . . . . . . . . . . . . . . . .
83
6.2
Estimation of models for shunt reactor and short circuit impedance . . . . . . .
87
6.3
Correction of cable capacitance . . . . . . . . . . . . . . . . . . . . . . . . . . .
90
6.4
Estimation of missing resistance . . . . . . . . . . . . . . . . . . . . . . . . . . .
93
6.5
95
Proposal for minimizing switching overvoltages
97
7.1
Evaluation of switching overvoltages at the 150 kV level . . . . . . . . . . . . . .
97
7.2
Evaluation of propagation of switching overvoltages . . . . . . . . . . . . . . . .
97
7.3
Proposal for minimization of switching overvoltages . . . . . . . . . . . . . . . .
98
7.4
Summary & discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Conclusion
107
8.1
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
8.2
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Bibliography
110
Appendix
A-1
A-1
Parameter determination for the overhead line system
A.1 Series impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-1

A.2 Shunt admittance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-14
A.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-24
B
Parameter determination in the shunt reactor
B-25
B.1 Theoretical background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-25

B.2 Inductance determination for the shunt reactor . . . . . . . . . . . . . . . . . . . B-27
B.3 Calculation of the inductances in the shunt reactor . . . . . . . . . . . . . . . . B-32
B.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-34
C
Measurements of transmission line switching
C-35
C.1 General test procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-35

C.2 Measurements on 150 kV and 20 kV level . . . . . . . . . . . . . . . . . . . . . . C-36
C.3 Measurements on 0,4 kV level . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-41
C.4 Conclusion & discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-44
CONTENTS
Transient recording using Omicron CMC 256-6
ix
D-47
D.1 Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-47

D.2 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-47
D.3 Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-48
D.4 Harmonics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-48
E
Single-line diagrams
E-49
E.1 Ferslev substation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-49

E.2 Tinghj substation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-50
E.3 Explanation of symbols in single-line diagrams . . . . . . . . . . . . . . . . . . . E-51
F
Parameter matrices for PSCAD
F-53
M-files used for calculations
G-57
G.1 M-files for determination of parameters in the overhead line system . . . . . . . G-57
G.2 M-files for determination of parameters in the shunt reactor . . . . . . . . . . . G-59
G.3 M-files for mathematical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . G-60
G.4 M-files for miscellaneous purposes . . . . . . . . . . . . . . . . . . . . . . . . . . G-63
H
CD content
H-65
Chapter 1
Introduction
Electricity is one of the cornerstones of modern society. Over the past decades, a steady rise
of electric power consumption has been experienced, which is also forecasted for the foreseeable
future. This is shown in figure 1.1.
TWh
60
50
40
30
20
10
0
1990
1992
Wind
1994
1996
1998
Decentral
2000
2002
Central
2004
2007
Hp Central
2009
2011
2013
2015
Consumption
Figure 1.1: Development in electricity production and consumption in Denmark the last 15 years,
including forecast for the next 10 years [2, www.miljorapport2006.dk]. Hp Central (High price central),
means that central production is activated when export prices are high.
In order to keep up with the increasing power consumption, the electric power generation must
respond. As shown in figure 1.1 the increment of power generation consist mainly of wind energy
and decentralized production. This was decided politically, in order to reduce the emission of
CO2 . This is in accordance with sharper focus on environmental issues, visual pollution amongst
other, where renewable power sources for electric generation, and grounding of transmission lines
are to be favoured.
In the early years of electric power systems, the power was consumed close to its generation.
Nowadays electric power systems are connected across national borders in large networks, to
ensure high reliability of electric power supply to the costumer. Generally, the electric power
system is divided into three subparts, namely: power generation, power transmission and power
distribution, shown in figure 1.2.
An important aspect in the design of electric power systems is the efficiency. Higher efficiency
indicates lower prices for the consumers. A key issue in order to obtain a high efficiency in the
transmission system, is to avoid transmission of reactive power. Beside reactive power caused
by the consumers consumption, the reactive power is caused by inductances and capacitances in
the transmission system. A transmission system, consisting of overhead lines, is predominantly
inductive in normal drift as the inductances are larger than the capacitances, and the system
consumes thereby reactive power. Underground cables are, on the other hand, mainly capacitive,
and are therefore seen as producing reactive power.
In order to minimize transmission of reactive power, compensation of the production or
consumption of reactive power is made.
Introduction
Generation
Transmission
Step-up transformer
Distribution
Step-down transformer
Consumer
Consumer level transformer
Figure 1.2: Graphical overview of the electrical power system
By law, the generation and transmission of electric energy, is to be separated. As the payload
from the power plant is active power, the production is according to this, and compensation is
therefore mostly left to the transmission company. This compensation is generally approached
by distributing reactive components along the power grid. These reactive components are chosen
to cancel out the reactive power, e.g. if the grid is inductive a capacitance is added, and vice
versa if the grid is capacitive an inductance is added. These reactive components can either be
stationary connected to the grid, or connected by circuit breakers to allow disconnections.
The power consumption is dynamic in its nature, and to compensate for different kind of loads,
the reactive components are connected/disconnected as the situation demands. A connection of
a single shunt reactor e.g. would therefore be performed in a case of a capacitive grid, and again
disconnected with the grid becoming inductive.
As the transmission lines at and below the 150 kV level, are to be substituted with underground
cables, an increment of the capacitance in the grid, and thereby reactive power production will
occur. This reactive power production has to be compensated for, and the predominant method
is implementation of shunt reactors. This power factor correction, is not achieved without a cost,
as problems due to connection of these reactive components to the grid arise. These are mainly
in the form of charging/discharging currents. Connection of an uncharged reactive component,
causes large inrush current resulting in overvoltages at the moment of the switching. These
overvoltages can propagate from the transmission level to the distribution level, and thereby
affecting the consumers. As the voltage quality affects the consumers electronic devices, the
propagation of these overvoltages is an actual problem, e.g. sensitive devices could be harmed
by overvoltages, resulting in malfunction. This leads to the general problem:
Inrush currents caused by switching of grid components on the transmission level,
affect the power quality at the distribution level.
In the year 2004 it was decided by the Danish Parliament, to build an offshore windmill park at
Horns Rev, in the North sea, called Horns Rev 2. The park is placed 23 km north-west of an
existing park Horns Rev 1. The geographical configuration of the two offshore parks, and the
energy transfer systems are shown in figure 1.3. The unified energy production of the Horns Rev
2, will be 215 MW.
Cable station
150 kV station
Future overhead lines
Future underground cables
Existing overhead lines
Existing underground cables
North sea
Ringkbing Fjord
Blbjerg
Land cable
Galtho
Karlsgrde
Sea cable
Overhead lines
Blvand
Horns Rev 2
Esbjerg
Endrup
Horns Rev 1
10 km
Figure 1.3: The geographical configuration of the two offshore parks, Horns Rev 1 and Horns Rev 2 [3,
hornsrev2.energinet.dk].
The electricity produced by the windmills is collected at a transformer platform, transforming

their voltages, from 33 kV to 150 kV. From the transformer platform the electricity of the windmill
park is transported to the mainland by a 42 km long 150 kV sea cable. In a cable station, placed
at Blbjerg Klitplantage, the sea cable is connected to a 30 km long 150 kV underground cable,
establishing connection to Galtho in the Tistrup region. In Galtho the cable will be connected
to a 26 km long overhead line. The overhead line connects to an existing transformer station
in Endrup, north east of Esbjerg, where the 150 kV is transformed to 400 kV, enabling the
electricity from the windmill park to enter the 400 kV system [3, hornsrev2.energinet.dk].
The resulting transmission system, with two long cable sections, will consequently require a
large reactor to compensate the reactive power production. As the windmill park only produces
electricity when the wind blows, the cable and hence the shunt reactor are disconnected from the
400 kV transmission system when the windmills are shut down. When the grid is reconnected,
over-voltages can occur. This leads to the initiating problem:
Connection of a grid segment, can cause overvoltages at transmission level, that
propagate to the distribution level.
In order to clarify and improve this problem, a model of the system is needed.
As Horns Rev 2 is not constructed, it is not possible to verify a model of the system through
measurements. Instead a model is constructed, on the basis of a similar system. Simulation
results can then be compared with measurements to verify the model.
In association with Energinet.dk, it is chosen to focus on the 150 kV grid between Ferslev and
Tinghj as shown in figure 1.4.
This system has similar characteristics as the system at Horns Rev 2, as it contains a shunt
reactor, underground cables and overhead lines.
Introduction
Cable station
150 kV station
Overhead lines
Underground cables
Nrresundby
Limfjorden
Aalborg
To dalen
To Mosbk
Ferslev
Aalborg Bugt
Stvring
Haverslev
Aars
Hadsund
Mariager Fjord
Tinghj
To Hornbk
10 km
Hobro
Figure 1.4: Location of the 150 kV grid from Ferslev to Tinghj.
The grid is subdivided into three different subsections, containing two underground cable sections
and one overhead line section. A simplified outline of this system is shown in figure 1.5.
Reactor
Section 3
Tinghj
Section 2
Section 1
Haverslev
21 km
Ferslev
18,93 km
0,43 km
Figure 1.5: Subsections of the system.
At the transformer station in Tinghj, a 40 Mvar shunt reactor is mounted, to compensate the
reactive power production of the system. The system can be disconnected from the rest of the
150 kV system through circuit breakers, in each of the two transformer stations.
In the next chapter, the problem of the experienced switching overvoltages is analysed, and the
solution method presented.
Chapter 2
Problem analysis
As stated in the initial problem, the switch-on of a grid segment at transmission level can
cause overvoltages that propagate to lower voltage levels. In order to improve on this problem,
the nature of this phenomenon must be analyzed. To investigate the magnitude of the problem,
measurements were made at the moment of switch-on, of the 150 kV grid section betweenTinghjFerslev 1 . The voltage waveforms shown in figure 2.1 were recorded in the test, where the TinghjFerslev section was isolated.
Voltage [kV]
100
0
100
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0.065
0.05
0.055
0.06
0.065
0.05
0.055
0.06
0.065
Time [s]
Voltage [kV]
20
10
0
10
20
0.03
0.035
0.04
0.045
Time [s]
Voltage [V]
500
500
0.03
0.035
0.04
0.045
Time [s]
Figure 2.1: Voltage waveforms recorded in Ferslev at the moment of switch-on. From top-to-bottom
150 kV, 20 kV, 0,4 kV.
In figure 2.1, it is apparent that transients arise after the switch-on of the section. These are most
significant at the 150 kV level, where a high frequency component is apparent for a brief period,
plus a lower frequency component. At the 20 kV and 0,4 kV levels, the high frequency component
is no longer apparent. The lower frequency component is still present on both the 20 kV and
the 0,4 kV level, though damped in comparison with the 150 kV level. After approximately one
period of the fundamental frequency, the distortion at the 0,4 kV level is no longer present.
1
The measuring reports are in appendix C on page C-35.
Problem analysis
The magnitudes of the recorded voltages are summarized in table 2.1.

Level [kV]
150
20
0,4
Peak voltage
before switch-on [kV]
134
16,8
0,316
Peak voltages at
switch-on instant [kV]
+159 -162
+17,3 -17,4
+0,381 -0,332
Percent
20,8%
3,5%
18,5%
Table 2.1: Recorded voltages from the test in Ferslev, sees in relation to figure 2.1.
From figure 2.1 and table 2.1, it is apparent that the switching overvoltages propagate to the
distribution level.
2.1
Problem statement
The switching overvoltages experienced when grid sections are switched-on, are unwanted as they
affect the consumer, and can in worst case result in damaged equipment. It is therefore desirable
to investigate this phenomenon, and provide a proposal for minimization of the switching
overvoltages.
Solution Method
In order to examine the problem, a dynamic model is constructed, and an analysis of the 150
kV grid between Ferslev and Tinghj is performed. First a simplified mathematical model is
arranged, where the most significant parameters are included, in order to give a theoretical
understanding of the system. Later, a detailed version is implemented in the simulation software
PSCAD. In relation to the switch-on for the section, a method that minimizes the switch-on
transients is to be implemented in the model.
The main objectives for this project are listed below:
Arrange a three phase representation of the 150 kV transmission system between FerslevTinghj, valid for the switch-on in Ferslev.
Explain the experienced transients from a theoretical analysis viewpoint.
Arrange a simplified dynamic model for the 150 kV system between Ferslev-Tinghj.
Arrange a mathematical solution for the simplified model in MatLab.
Verify the MatLab model through measurements and simulation in PSCAD.
Arrange a simulation of the model in PSCAD.
Verify the PSCAD model by comparing to measurements and calculated values.
Arrange a proposal for an optimization of the reconnection.
2.2 Limitations
2.2
Limitations
To limit the scope of the project the following limitations are made:
The 400 kV system present in the overhead line section of the 150 kV grid between Ferslev
and Tinghj is neglected.
For the simplified model, other voltage levels than 150 kV are neglected.
Further limitations are described in the respective sections.
Chapter 3
Description and modelling of the 150

kV grid
The objective of this chapter is to give a description of the system subparts, and their respective
electric equivalents. This is done in order to generate a three phase electric representation of
the 150 kV grid segment, valid for switch-on in Ferslev. A simplified line diagram of the system
is given, where all subparts can be seen, where after the system subparts are described. The
chapter concludes with a three phase electric representation of the grid.
3.1
Main system
In figure 3.1 is the single-line diagram1 for the connection between the Ferslev (FER3) and
Tinghj (TH3) transformer stations, and the rest of the transmission system shown.
TH3
FER3
Reactor
ABB
40 Mvar
3
ABB
40 Mvar
400 kV
1
KT51
ABB
400 MVA
MOS
HNB
ADL1
Siemens
80 MVA
60 kV
KT31
Strmberg
80 MVA
60 kV
S1
S1
S2
KT31
Strmberg
80 MVA
20 kV
16 MVA
0,630 MVA
Figure 3.1: Simplified single-line system of the two stations and their interconnection.
The single line diagram, shown in figure 3.1, shows the main components of the system, namely:
underground cables, overhead line, shunt reactor, and transformers. These components are
described in the following sections, along with circuit breaker and the voltage source.
3.2
150 kV source
The 150 kV grid section, is energized from Ferslev substation. When producing a model for
the 150 kV system, the remainder of the transmission system is represented by a voltage source
behind short circuit impedance, as shown in figure 3.2.
1
0,4 kV
It should be noted that the single-line diagram is simplified, the non-simplified single-line diagrams for the
two transformer stations are shown and explained in appendix E.
10
3.2.1
Electrical representation
Generally, the energization of grid sections is made through substations, whose characteristics can
vary considerably. The two extreme cases are: Energization from the busbar, that is solely fed
from generators, and energization from a busbar solely fed from long transmission lines/cables.
The former appears as a source behind an inductance, while the latter is represented by a source
behind a resistance [4, Bickford, p.97].
In this project, the energization of the Ferslev-Tinghj grid section, is through FER3 substation.
The source impedance, seen from the 150 kV grid in Ferslev, is somewhere in between the above
mentioned extremes, as a 400/150kV transformer placed at the station, feeds the busbar from the
Nordjytland power plant, but transmission lines on the 150 kV level from dalen and Mosbk,
supply short circuit power to Ferslev, as is shown in figure 3.1. The short circuit apparent power
in Ferslev substation, is not known at the moment of testing. Under normal conditions, the
short circuit impedance is mostly reactive, with angle of 6 80 to 6 85 assumed.
Zs
ua
Zs
ub
Zs
uc
Figure 3.2: The remainder of the transmission system can be equated with a voltage source and a
source impedance.
3.2.2
Electrical parameters
The voltage was measured at the 150 kV grid in Ferslev to a line voltage of 165 kV, which
will be used in simulations and calculations, as this will give more comparable results to the
measurements.
The maximum and minimum short-circuit apparent power is given by Energinet.dk, and is listed
in table 3.1.
S
U
Maximum
5908 MVA
165 kV
Minimum
2366 MVA
165 kV
Table 3.1: Short circuit information for the 150 kV grid in Ferslev, given by Energinet.dk.
The maximum apparent short-circuit power is given for the following conditions:
3.2 150 kV source
11
Intact transmission network, including AC-connection to Germany

All available central power plants are running
Synchronous compensation is activated for both machines
All decentralized power plants are running
All grid-connected wind mills are running
The minimum apparent short-circuit power is given for the following conditions:
Three central production units are on the grid
Synchronous compensation is not activated
One important grid component is out of drift
The short-circuit apparent power in the German 400 kV net is low
Two thirds of the total decentralized power production is out
All installed windpower is out
The short circuit impedance for each phase is given by:
Zs =
U2
S
(3.1)
The angle is unknown, and the short circuit impedance is therefore calculated for four different
angles, both for the maximum and minimum short circuit apparent power. The results are listed
in table 3.2.
Maximum power
Zs = 4, 61 6 70
Rs
Ls
1,58 13,78 mH
Minimum power
Zs = 11, 37 6 70
Rs
Ls
3,94 34,4 mH
Zs = 4, 61 6 75
Rs
Ls
1,19
14,2 mH
Zs = 11, 37 6 75
Rs
Ls
2,98 35,6 mH
Zs = 4, 61 6 80
Rs
Ls
0,800 14,5 mH
Zs = 11, 37 6 80
Rs
Ls
2,00 36,1 mH
Zs = 4, 61 6 85
Rs
Ls
0,402 14,6 mH
Zs = 11, 37 6 85
Rs
Ls
1,00 36,5 mH
Table 3.2: Values for short circuit impedance, calculated for different angles and different apparent
power.
For a first approach, the short circuit impedance will be used for the maximum short circuit
apparent power with an angle of 80 .
12
3.3
Circuit breaker
The circuit breaker connects the 150 kV grid segment to the rest of the 150 kV grid when switched
on.
CB
Figure 3.3: Model diagram for the circuit breaker, shown for one phase.
3.3.1
The electrical equivalent for the circuit breaker is shown in figure 3.4. Generally, a circuit breaker
will contain a leakage resistance when switched off, and a series resistance when switched on.
Ron
Rof f
Figure 3.4: Equivalent diagram for the circuit breaker, shown for one phase.
As focus is at the switch-on instant, electric arcs are considered to be of no influence, as the
breaker closes almost instantaneously, and the breaker can therefore be considered ideal in this
operation.
3.3.2
In table 3.3 are the equivalent series resistances for the circuit breaker given, both for the on-state
and the off-state. These values are proposed by Energinet.dk.
Ron
Rof f
100
1 M
Table 3.3: Resistances in the circuit breaker in its two states.
3.4
Underground cable
There are two underground cable sections in the system, a short section of 0,43 km (Section
1) connecting the overhead lines with the transformer station in Ferslev, and a 21 km section
(Section 3) connecting the cable station in Haverslev with the transformer station in Tinghj.
In these two sections, the cables are of the same type: 1200 mm2 aluminium conductors, with
each phase screened with 95 mm2 copper conductors. The maximum operating voltage for the
cable is 170 kV and the current capacity is 750 A.
3.4 Underground cable
3.4.1
13
Physical layout
The cable is from ABB, of the type AXLJ 1x1200/95 170kV, its physical layout is shown in figure
3.5.
Conductor: Aluminium, round, compact, A=1200 mm2 , D=41,5 mm
Conductor screen: Thickness 1,2 mm
Insulation: XLPE, thickness 16 mm, D=75,9 mm, = 2,3
Insulation screen: Thickness 1,0 mm
Bedding: Semi conductive tapes, thickness 0,1 mm
Metallic screen: Copper wires, A=95 mm2
Longitudinal water seal: Conductive swelling tapes, thickness 0,65 mm
Radial water seal: Aluminium laminate, thickness 0,2 mm
Outer sheath: PE, thickness 3,9 mm
Complete cable: D 92 mm
Figure 3.5: The physical layout for AXLJ 1x1200/95 mm2 170 kV cable.
As each phase is independently screened, mutual capacitances between phases can be excluded.
The capacitance consist therefore only of the capacitance between conductor and ground (C0 ),
as shown to the right in figure 3.6.
Grounding & Cross bonding

A triangular cable layout is used in the two cable sections. The cables are buried at a depth of
1200 mm as shown in figure 3.6.
Conductor
Isolation
Screen
1200 mm
C0
C0
C0
Figure 3.6: Cable duct and triangular cable layout.
In the cable section between Haverslev and Tinghj cross bonding is applied. In cross bonding,
the metallic screen for each phase is crossed with the other phases, and grounded along the
section. By doing this, the total induced voltage in the screen will be minimized, thereby
reducing losses. The cross bonding layout for Haverslev-Tinghj is shown in figure 3.7. At every
second joint, the cable is grounded. Joints with screen separation are crossed and grounded, but
straight-through joints are grounded without crossing.
14
Haverslev
Tinghj
Straight-through joint
Joint & screen-seperation
Linkbox for crossbonding
Linkbox for earth-connection

of straight-through joint
Figure 3.7: The cross bonding of the cable section between Haverslev and Tinghj.
In the cable section leading into the transformer station in Ferslev, one end of the screen is
grounded. No cross bonding is used in this section, due to its short distance.
3.4.2
In order to obtain a three phase representation for the two underground cable sections, a general
representation for one conductor is needed. A conductor of an infinitesimal length is shown in
figure 3.8.
rdx
ldx
gdx
cdx
x
Figure 3.8: An infinitesimal line section[1, Vrts, p. 158].
As shown in figure 3.8, the conductor contains four parameters, namely: the series resistance,
the series inductance, the shunt conductance and the shunt capacitance. These parameters
are distributed over the full length of the conductor. In order to make a less complicated
representation, the distributed parameters are lumped in one component for each parameter.
The conductor can now be represented in to ways, by the equivalent or by the T equivalent. In
the equivalent one half of the shunt admittance is placed at each end of the conductor, and the
full series impedance is placed in between as shown to the left in figure 3.9. In the T equivalent,
the full shunt admittance is placed in between one half of there series impedance, as shown to
the right in figure 3.9.
3.4 Underground cable

PSfrag
15
u1
Y
2
Z
2
Z
2
Z
Y
2
u2
u1
u2
Figure 3.9: Electric representation of one conductor. To the left equivalent. To the right T equivalent.
Where:
Z = R + jX is the series impedance
Y = G + jB is the shunt admittance
[]
[S]
In the three phase representation of the cable sections, the equivalent will be used.
Series impedance
The series impedance is determined by the series resistance and series reactance. These are
mainly dependent on conductor type, length, geometrical configuration and frequency.
Shunt admittance
The shunt admittance is determined by the conductance and the susceptance of the conductors.
The shunt conductance consists of the leakage conductance and losses due to partial discharges.
In a cable the partial discharges will consist of internal discharges in the isolation material.
The shunt susceptance is determined by the capacitance between the conductor and ground, and
the capacitance between the phase conductors. As mentioned previously, the screen in the cable
allows the capacitance between the conductors to be neglected.
Three phase representation
In order to represent a three phase system, the flux linkage between the phases the must be
taken into consideration. The flux linkage gives rise to mutual inductances between the phases,
which is included in the three phase representation for the cable section, as shown in figure 3.10.
16
Gan
2
Gbn
2
Ra
La
Can
2
Mab
Can
2
Rb
Lb
Mac
Cbn
2
Rc
Gcn
2
Mbc
Gan
2
Cbn
2
Gbn
2
Ccn
2
Gcn
2
Lc
Ccn
2
Figure 3.10: -model for a three phase cable section with mutual inductances.
Where:
Mab is the mutual inductance between conductor a and conductor b
Mac is the mutual inductance between conductor a and conductor c
Mbc is the mutual inductance between conductor b and conductor c
3.4.3
[H]
[H]
[H]
Electrical paramters
In table 3.4 are the most important electrical characteristics from the datasheet2 listed.
a.c. resistance at 57 C
d.c. resistance at 20 C
Capacitance per phase
Inductance between conductors,per phase
Dielectric loss per phase
Positive sequence impedance Z1
Negative sequence impedance Z2
Zero sequence impedance Z0
0,0330 /km
0,0247 /km
0,243 F/km
0,343 mH/km
0,1144 kW/km
0,0331 + j0,1077 /km
0,0331 + j0,1077 /km
0,1876 + j0,0706 /km
Table 3.4: Electrical characteristics for the AXLJ 1x1200/95 mm2 170 kV cable.
The conditions for which the zero sequence impedance is calculated by the producer, are
unknown. The mutual inductances for the cable are not provided.
The shunt conductance per phase can be found from the dielectric loss per phase and the phase
voltage, here calculated for section 3:
Pdiel = U 2 G
2
The datasheet can be found on the CD
(3.2)
3.5 Overhead transmission lines
17
G =
G =
Pdiel l
U2
0, 1144 21

3
16510
(3.3)
2 = 265 nS
(3.4)
The phase vales for the two cable sections are listed in table 3.5, due to the symmetric
configuration, the phase vales are equal for all phases.
Rx
Lx
Cxn
Gxn
Section 1 (0,43 km)

0,0142
0,147 mH
104,5 nF
5,42 nS
Section 3 (21 km)

0,693
7,20 mH
5,10 F
265 nS
Table 3.5: Electric parameters fro the two cable sections. x=a,b,c.
3.5
Overhead transmission lines
Section 2 is a 18,93 km long overhead line section. The conductor is of the type Martin simplex
772 mm2 , with a current rating of 835 A. The ground conductor is of the type, Dorking simplex
153 mm2 .
3.5.1
Physical layout
The section makes use of the Donau mast, which is shown in figure 3.11. As shown in figure
3.11, there can be two systems on the mast, which also is the case as a 400 kV system is present
at the opposite side of the tower.
20,8 m
10,4 m
41,6 m
33,5 m
24,7 m
5,8 m 7,2 m
Figure 3.11: Donau mast, type D11, with different systems each side of the tower.
18
The simplex configuration is used in the section, meaning that there is only one conductor per
phase. In this section the conductor material is aluminium over a steel core, as shown in figure
3.12.
r1
Aluminium
r2
Steel
Figure 3.12: Aluminium conductor with steel core.
3.5.2
The overhead line section can be represented with the equivalent as described in section 3.4.2
on page 14. Beside the inductance between the phases, capacitance between the phases will also
be present in an overhead line system. This gives the representation shown in figure 3.10.
Ra
La
Gan
2
Can
2
Mab
Can
2
Cab
2
Rb
Lb
Mac
Gbn
2
Cac
2
Cbc
2
Gcn
2
Cbn
2
Rc
Ccn
2
Mbc
Cbn
2
Gan
2
Cab
2
Gbn
2
Cac
2
Lc
Ccn
2
Cbc
2
Gcn
2
Figure 3.13: -model for a three phase line section with mutual inductances.
3.5.3
The provided information for the phase conductor and the ground wire used in the overhead line
section are listed in table 3.6.
3.5 Overhead transmission lines
Type
Self GMD
Resistance
Length
19
Phase conductor
simplex 772 mm2
0,0146 m
0,0423 /km
18,93 km
Ground wire
153 mm2
0,0070 m
0,2982 /km
18,93 km
Table 3.6: Parameters for the conductor and ground wire in the overhead line system.
The calculated DC resistances in table 3.6 will always have smaller value than the actual
resistance, due to the following:
Temperature
Skin effect
Proximity effect
Conductor sag
Spiraling
These terms are considered to give a correction factor of 1,153 , resulting in an AC resistance of
0,898 for section 2.
The shunt conductance of the system has not been calculated, however an estimation of its value
is performed in the following. Generally, the shunt conductance in an overhead line system is
given by the leakage losses and dielectric losses.
The leakage conductance is given by the surface conductance of the isolators, from the conductors
to earth. The leakage losses on a 132 kV line have been found to vary between 0,2 and 0,6 kW/km
[5, Weedy, p. 130]. This voltage level is close to the 150 kV level, and the worst case loss of 0,6
kW/km is therefore used. This gives the following conductance for the leakage part:
3
Gleakage =
0,610
Pleakage
9
3
S/km
=
= 22, 0 10

3 2
U2
16510
(3.5)
The partial discharges in the overhead line are in the form of corona, which occurs as the air
near the high voltage conductor is ionized. A method to evaluate the corona losses is presented
by A.-Salam & A.-Aziz [6]. Here the three phase power losses for a 210 kV overhead line (the
closest value to the 150 kV of the system), is found to vary 1-3 W/m (1-3 kW/km). By taking
the highest value of 3 kW/km, and knowing that the losses increase with higher voltage, this
value is used as a worst case scenario. This gives the conductance per phase, per/km for the
partial discharges part:
3
Gpd =
310
Ppd
9
3
=
S/km

= 110 10
3 2
U2
16510
The correction factor is found in appendix A.1.3 on page A-10.
(3.6)
20
The approximated worst case shunt conductance is found from the parallel configuration of the
leakage conductance and partial discharge conductance:
(3.7)
Gshunt = Gleakage + Gpd

9
Gshunt = 22, 0 10
+ 110 10
= 132 10
S/km
(3.8)
The electrical parameters of the overhead line section are calculated in appendix A. In table 3.7
are the phase values for the parameters in figure 3.13 listed.
Ra
Rb
Rc
0,898
0,898
0,898
Gan
Gbn
Gcn
2,5 S
2,5 S
2,5 S
La
Lb
Lc
36,2 mH
36,3 mH
34,4 mH
Can
Cbn
Ccn
0,149 F
0,149 F
0,165 F
Mab
Mac
Mbc
13,6 mH
10,9 mH
11,0 mH
Cab
Cac
Cbc
27,5 nF
11,4 nF
12,7 nF
Table 3.7: Electrical parameters for the overhead line section.
3.6
Shunt reactor
The shunt reactor is mounted at the transformer station in Tinghj, to compensate the reactive
power production in the cable. The shunt reactor is produced by ABB Transformers, and shown
in figure 3.14.
Figure 3.14: The shunt reactor placed at Tinghj transformer station.
3.6.1
Physical layout
The shunt reactor has a five-limbed core, with 8 gaps distributed across the three middle limbs,
which are for the phase windings. These gaps are inserted to avoid saturation and have the
reactor operating in the linear region of its B/H curve. Furthermore, it helps to decrease the
3.6 Shunt reactor
21
mutual inductances between the phases. The geometry of the reactor core is shown in figure
3.15.
d
c
p
m
i
j
g
n
o
k
c
h
f
b
Figure 3.15: Geometry of the reactor core from the front.
The distances from figure 3.15 are listed in table 3.8.

a
b
c
d
e
f
g
h
2213
4182
210
888
1203
600
1548
70
mm
mm
mm
mm
mm
mm
mm
mm
i
j
k
l
m
n
o
p
151,9
32,7
10
2,5
193
109
76
65
mm
mm
mm
mm
mm
mm
mm
mm
Table 3.8: Distances from figure 3.15.
The geometry of the reactor core from above is shown in figure 3.16.
f
c
a
b
Figure 3.16: Geometry of the reactor core from above.
The distances from figure 3.16 are listed in table 3.9.
22
a
b
c
794 mm
4182 mm
210 mm
d
e
f
1038 mm
976 mm
600 mm
Table 3.9: Distances from figure 3.16.
3.6.2
In this section, the electric equivalent for the shunt reactor described. In order to arrange such
model, the following terms must be included:
Self inductances
Mutual inductances
Core losses
I 2 R losses in the windings
These terms are described in the following, resulting in a three phase electric representation for
the shunt reactor.
Self inductances
The reactor core has five limbs, which give the flux paths for the flux generated by phase a,
shown in figure 3.17, where it is assumed that all the flux passes through the core.
a1
ab
ac
a2
Figure 3.17: The flux paths in the reactor core for phase a.
The solid line in figure 3.17 represent the primary flux path for the flux generated by the phase
a winding and is determined by the reluctances in the core and the gaps. The self inductance
for phase a, can be found if the reluctance for the primary flux path and the number of turns in
the winding are known4 .
4
The inductances in the shunt reactor are determined in appendix B.
3.6 Shunt reactor
23
Mutual inductances
The dashed lines in figure 3.17 represents the alternative flux path for the flux generated by the
phase a winding. Here the flux passes through the limb for phase b and phase c inducing voltages
in those two phases. In the same manner phase a will be affected with a contribution from phase
b and phase c. These contributions are represented by the mutual inductances, and determined
by the reluctances and number of turns in the same manner as the self inductance.
Core losses
As a consequence of the fluctuating magnetic field in the core, losses due to eddy currents and
hysteresis will occur. The magnetic flux in the core induces circulating currents, which causes
heat. In order to reduce the loss caused by these induced currents, the core is laminated, thereby
reducing the current pathway.
The hysteresis losses occur as the magnetic field changes polarity. Due to the magnetic
characteristics of the core material, a hysteresis loop will be present, where the hysteresis losses
is determined by the area in the hysteresis loop. Hence, is a slender hysteresis loop desirable.
Usually, a compromise between high saturation and slenderness of the hysteresis loop is taken,
due to the characteristics of the core material.
Copper losses in the windings
The conductor in each winding contains a certain resistance. Hence, losses will arise due to the
current passing through the resistance of the conductor, as given by:
Pcu = I 2 Rcu
(3.9)
Electric circuit model for the shunt reactor

In figure 3.18 is the electric circuit equivalent for phase a of the reactor shown.
ia (t)
+
Rcu,a
ua (t)
Rf e,a
La
Mab
Mac
Figure 3.18: Electric equivalent diagram for phase a in the shunt reactor.
Where:
Rcu,a represents the I 2 R losses in the windings
Rf e,a represents the losses in the core
La is the self inductance for phase a
Mab represents the mutual inductance between phase a and phase b
Mac represents the mutual inductance between phase a and phase c
[]
[]
[H]
[H]
[H]
24
The electric equivalent of the shunt reactor is arranged with the winding loss resistance in
series with the parallel connection of the core loss resistance and the self inductance. The
parallel connection represents that the current flowing through the core loss resistance, does not
contribute to the induction. The inductance is linking with the inductances in the two other
phases, represented by the mutual inductances Mab and Mac .
This model does not account for saturation of the reactor core, or the magnetic remanence, as
the system is assumed totally discharged at the moment of switch-on.
The representation for phase a of the shunt reactor, can be used for the two other phases as well.
In figure 3.19 is the electric circuit equivalent for the shunt reactor shown for all three phases.
ia (t)
+
Rcu,a
ua (t)
Rf e,a
La
ib (t)
+
Rcu,b
ub (t)
Mab
Rf e,b
Mac
Lb
Mbc
ic (t)
+
Rcu,c
uc (t)
Rf e,c
Lc
Figure 3.19: Equivalent diagram for the three phases in the shunt reactor, including mutual inductances.
3.6.3
Electrical characteristics
The rated values of the shunt reactor is listed in table 3.10.

Rated power
Rated voltage
Rated current
40 Mvar
170 kV
135,8 A
Table 3.10: Information of reactor [7, ABB Reactor].
3.6 Shunt reactor
25
The calculated parameters for the reactor are listed in table 3.11. Values for resistance are given
in the user manual, found on the CD. Mutual inductances are calculated from the geometry of
the reactor.
Rcu,a
Rcu,b
Rcu,c
La
Lb
Lc
Mab
Mac
Mbc
Calculated
1,745 H
1,744 H
1,745 H
-607 H
-434 H
-607 H
Designed
2,300 H
2,300 H
2,300 H
-32,0 mH
-23,0 mH
-32,0 mH
Measured
0,681
0,680
0,680
2,307 H
2,304 H
2,307 H
Table 3.11: Electrical parameters for the shunt reactor [7, ABB Reactor].
As the calculated values for the mutual inductances in the shunt reactor, are small compared
with the expected values, the expected values in the 1-2% region are used for simulations.
The losses in the reactor are given to Ploss =64,3 kW @ 170 kV. This includes both the copper
loss and the core loss for all three phases. In order to find the core loss, the copper losses must
be subtracted from the total loss:
Pcu,a = Rcu,a I 2
(3.10)
2
Pcu,a = 0, 6812 135, 8 = 12, 56 kW
(3.11)
The resistance representing core loss is calculated from the loss in the core, and the voltage across
the resistance Rf e,a shown in figure 3.19:
URf e
Uf e
(3.12)
= Rf e
Rf e,a =
(Ua I Rcu,a )2
Ploss,a Pcu,a
(3.13)
Rf e,a =
135, 8 0, 681)2
( 17010
3
64,3103
3
12, 56 103
Rf e,a
Rf e,b
Rf e,c
= 1, 08 M
(3.14)
1,08 M
1,08 M
1,08 M
Table 3.12: Calculated loss resistance for the core.
With all values for the components in the reactor, shown in figure3.19 known, it is possible to
simplify this model, to the one shown in figure 3.20.
26
ia (t)
+
Rreactor,a
ua (t)
Lreactor,a
ib (t)
+
Rreactor,b
ub (t)
Mab
Mac
Lreactor,b
Mbc
ic (t)
+
Rreactor,c
uc (t)
Lreactor,c
Figure 3.20: Simplified equivalent diagram for the three phases in the shunt reactor, including mutual
inductances.
The simplification is done by calculating a replacement impedance:

(3.15)
Zreactor,a = Rcu,a + (Rf e,a )||(jLa )

Zreactor,a = 0, 681 + 1, 08 106 ||(j 2 50 2, 307)
Zreactor,a = 1, 17 + j725
(3.16)
(3.17)
This is done for all three phases resulting in the values for the components in figure 3.20 listed
in table 3.13
Rreactor,a
Rreactor,b
Rreactor,c
1,17
1,17
1,17
Lreactor,a
Lreactor,b
Lreactor,c
2,307 H
2,304 H
2,307 H
Mab
Mac
Mbc
-32,0 mH
-23,0 mH
-32,0 mH
Table 3.13: Electrical parameters for the shunt reactor. For mutual inductances is estimate of 1% used.
3.7
Transformers
There are three transformers between the 150 kV level and the 0,4 kV level, namely: 150/60
kV, 60/20 kV and 20/0,4 kV. These transformers must be represented, in order to evaluate the
3.7 Transformers
27
voltages at the distributions level. The parameters needed to implement in PSCAD are presented
following.
These are:
Rated power
Winding voltages (line-to-line)
Rated frequency
Coupling type
Leakage reactance
No load losses
Copper losses
3.7.1
150/60 kV transformer
The values given in table 3.14 and 3.15, are taken from the 150/60 kV transformer test report,
found on the CD.
Rated power
Rated voltage
Rated current
Primary winding
80 MVA
165 kV
280 A
Secondary winding
80 MVA
67 kV
689 A
Table 3.14: Rated values for the 150/60 kV transformer.
Coupling
No-load losses
er
ex
YNd11
41 kW
0,35%
12,76%
Table 3.15: Electrical parameters for the 150/60 kV transformer.
In order to find the copper losses, Pcu , the short circuit impedance of the transformer is found,
and the rated current is used to evaluate Pcu . Impedances are calculated from the high voltage
side of the transformer, as this is seen from the source.
er + jex U 2
100
S
2
0, 35 + j12, 76 165
=
= 0, 9843 + j35, 88
100
80
= 3 Ir2 Rk
Zk = Rk + jXk =
(3.18)
Zk
(3.19)
Pcu
Pcu = 3 280 0, 9843 = 231, 5 kW

231, 5 kW
Pcu =
= 0, 00289 pu
80 MVA
(3.20)
(3.21)
(3.22)
28
3.7.2
The data in table 3.16 and 3.17, for the 60/20 kV transformer, is given by HEF net A/S. This is
to be found on the CD.
Rated power
Rated voltage
Rated current
Primary winding
16 MVA
60 kV
154 A
Secondary winding
16 MVA
20 kV
462 A
Table 3.16: Rated values for the 60/20 kV transformer.
Coupling
No load losses
er
ek
YNyn0
7,92 kW
0,44%
7,91%
Table 3.17: Electrical parameters for the 60/20 kV transformer.
The copper losses are calculated in the same manner as for the 150/60 kV transformer:
2
Pcu = 3 154
Pcu =
3.7.3
0, 44 602
100 16
= 70, 4 kW
70, 4 kW
= 0, 0044 pu
16 MVA
(3.23)
(3.24)
20/0,4 kV transformer
The data given for the 20/0,4 kV transformer, are limited as only the name plate was avalible.
This is to be found on the CD. Due to this, losses in the transformer are not given, but a short
circuit voltage drop of 4% is given. Hence, typical values for transformer of this size, given in [1,
Vrts, p. 207], are used. The data is listed in table 3.18 and 3.19, estimated values are annoted
with .
Rated power
Rated voltage
Rated current
Primary winding
630 kVA
24 kV
15,15 A
Secondary winding
630 kVA
0,420 kV
866 A
Table 3.18: Rated values for the 20/0,4 kV transformer.
Coupling
No load losses
ek
er
ex
ZNyn5
0,0025 pu
4%
1%
3,87%
Table 3.19: Electrical parameters for the 20/0,4 kV transformer.
3.8 Three phase representation of the grid
29
The copper losses are calculated in the same manner as above:

2
Pcu = 3 15, 15
Pcu =
3.8
242
1
100 0, 630
(3.25)
= 6, 3 kW
6, 3 kW
= 0, 01 pu
0, 630M V A
(3.26)
Three phase representation of the grid
The electric representations, obtained in the previous sections, are gathered as shown in figure
3.21.
Short
Circuit
circuit
breaker
impedance
Section 1
Section 2
Section 3
Reactor
ua
ub
uc
Figure 3.21: Cumulated model for the system. Arrows indicate mutual inductances between phases.
The diagram shown in figure 3.21, represents the three phase 150 kV grid section between FerslevTinghj. Transformers have been omitted from this representation. Parameter values, for each
part of the electric representation shown in figure 3.21, are given in table 3.20 on the next page.
30
Voltage Supply
ua
165sin((t)) kV ub
165sin((t) 2/3) kV
Short circuit impedance
Zs
4, 61 6 80 Rs
0,800
Circuit breaker
Ron
100 Rof f
1 M
Section 1 - 0,43 km long cable
Can /2
52,2 nF Ra
0,0142
Cbn /2
52,2 nF Rb
0,0142
Ccn /2
52,2 nF Rc
0,0142
Gan /2
2,71 nS Gbn /2
2,71 nS
Section 2 - 18,93 km long line
Can /2
74,5 nF Ra
0,898
Cbn /2
74,5 nF Rb
0,898
Ccn /2
82,5 nF Rc
0,898
Mab
13, 6 mH Mac
10,9 mH
Gan /2
1,25 S Gbn /2
1,25 S
Cab /2
13,75 nF Cac /2
5,7 nF
Section 3 - 21 km long cable
Can /2
2,55 F Ra
0,693
Cbn /2
2,55 F Rb
0,693
Ccn /2
2,55 F Rc
0,693
Gan /2
132,5 nS Gbn /2
132,5 nS
Reactor
Ra
1,17 La
2,307 H
Rb
1,17 Lb
2,304 H
Rc
1,17 Lc
2,307 H
uc
165sin((t) + 2/3) kV
Ls
14,5 mH
La
Lb
Lc
Gcn /2
0,147 mH
0,147 mH
0,147 mH
2,71 nS
La
Lb
Lc
Mbc
Gcn /2
Cbc /2
36,2 mH
36,3 mH
34,4 mH
11,0 mH
1,25 S
6,35 nF
La
Lb
Lc
Gcn /2
7,20 mH
7,20 mH
7,20 mH
132,5 nS
Mab
Mac
Mbc
-32,0 mH
-23,0 mH
-32,0 mH
Table 3.20: Parameters for the three phase representation, shown in figure 3.21.
3.9
Summary & discussion
In this chapter, the system subparts and their electric equivalents have been described. The last
section of the chapter summarizes the parameters and gives the three phase representation of
the grid segment that is to be analyzed.
The short circuit impedance was not known at the moment of measurements, but here a
value found from the maximum apparent power, and a typical angle was used. The reactor
representation does not account for magnetic saturation, as this is not expected to be a factor of
influence. With these limitations, the three phase model of the 150 kV grid section between
Ferslev and Tinghj is assumed to represent the system adequately for analysis, which is
conducted in the next chapter.
Chapter 4
Analysis of switching overvoltages on

the 150 kV grid
The objective of this chapter is to analyze the transients that are experienced at the switch-on
moment. In the first section of the chapter, the measured transients are analyzed, and estimated
from a theoretical point of view. In order to conduct a mathematical analysis of the system, a
simplified version is produced and two different analysis approaches are taken. The results from
the mathematical analysis are compared with a simulation in PSCAD of the simplified circuit.
The chapter ends with a discussion of the validity of the simplified model, and the factors that
have the greatest influence on the transient behaviour of the model. In order to give a clearer
analysis, focus is only on the 150 kV, and other voltage levels are therefore not considered here.
4.1
Estimation of the transient behaviour
Switching transients occur when the switching circuit changes from one state to another. In
the following section, the frequencies experienced after switch-on is analyzed from a theoretical
viewpoint.
In the transition between states, transient currents and voltages are superimposed on the power
frequency of the current and voltage. The frequency of these superimposed currents and voltages,
is determined by the natural frequencies of the power grid [8, Laubst, p.31]. For a series element,
the natural frequency is determined by the inductance and the capacitance:
f0 =
2 L C
(4.1)
Voltage [kV]
The voltage waveform, during switch-on of the grid section, is measured1 and shown in figure
4.1.
100
0
100
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0.065
Time [s]
Figure 4.1: The recorded voltage waveforms for 150 kV.
A Fourier analysis of the frequencies of the measured voltage waveform was made, the results of
this can be seen in figure 4.2. Results for a Fourier-analysis for all phases, and voltage levels can
be seen in appendix C.
1
The measurements are described in appendix C on page C-35.
32
10
Percentage of the fundamental [%]

150 kV level
5
0
0
1000
2000
3000
4000
5000
6000
1000
2000
3000
4000
5000
6000
1000
2000
3000
4000
Frequency [Hz]
5000
6000
10
5
0
0
10
5
0
0
Figure 4.2: Fourier analysis of the 150 kV waveform, of harmonic frequencies up to 6,5 kHz.
The Fourier analysis is performed over one period of the fundamental frequency, after the
switching instant. The Fourier analysis is conducted in MatLab, with the m-file shown in appendix
G.10. As shown in figure 4.2, two frequency band with large magnitude are present for all three
phases, one around 350 Hz and one around 3800 Hz. These are the frequencies that can be seen
in the time domain in figure 4.1.
The grid section and its equivalent model, described in the previous chapter, can be analyzed for
one phase to estimate the frequencies that will occur at switch-on. This is shown in figure 4.3.
Short
circuit
impedance
Circuit
breaker
Short circuit
impedance
Seen as an inductance
Section 1
Section 2
Section 1
430 m cable
Seen as a capacitance
Ls
ib
ia
C1
Section 2
18,93 km line
Seen as an inductance
L2
Section 3
Reactor
Section 3
21 km cable
Seen as a capacitance
C2
Reactor
Lr
Figure 4.3: Electric circuit for estimation of the transient behaviour. Top: non simplified representation.
Bottom: simplified representation for frequency analysis.
In order to obtain the circuit shown at the bottom of figure 4.3, the following simplifications
4.1 Estimation of the transient behaviour
33
are made. All resistances are stripped from the circuit, as these do not affect the frequencies,
only the damping. The short cable section is seen as purely capacitive, the overhead line is seen
as purely inductive, and the 21 km cable section is seen as purely capacitive. Furthermore, the
circuit breaker is eliminated, along with the mutual inductances and capacitances in the system.
From the current flowing into the circuit, two currents are interesting: Current a that charges
the capacitor in section 1, and current b that charges the capacitor in section 3. The parameters
for the circuit are:
Ls
14,4 mH
C1
0,105 F
L2
35,6 mH
C2
5,10 F
Lr
2,306 H
Table 4.1: The parameters used to evaluate the frequencies of the Ferslev-Tinghj section. These are
calculated average phase values.
For the current ia the natural frequency is found:

f0,ia =
f0,ia =
1
2 Ls C1
1
= 4, 1 kHz
2 0, 0144 0, 105 106
p
(4.2)
(4.3)
For the current ib the natural frequency is found:

f0,ib =
f0,ib =
1
2 (Ls + L2 ) C2
1
p
= 315 Hz
2 (0, 0144 + 0, 0356) 5, 1 106
p
(4.4)
(4.5)
The two calculated frequencies f0,ia and f0,ib , can be recognized in the measured frequencies
of ~3800 Hz and ~350 Hz. As the short circuit impedance at the time of the measurements is
unknown, the calculated frequencies deviate from the measurements. Furthermore, inductances
in the cable sections are neglected. These would also give slight changes in the calculated vales.
The natural frequency in the parallel coupling between reactor and the capacitor of section 3 is
found:
f0 =
1
= 46, 4 Hz
2 2, 306 5, 10 106
p
(4.6)
This frequency is expected, as the size of the reactor is chosen to compensate the capacitance of
the cable. If the reactance of the cable equals the reactance of the reactor, the frequency would
be 50 Hz.
The Fourier analysis of the measurement shown in figure 4.2, shows other frequencies than
the two frequencies caused by the capacitances in the cables. These frequencies arise due to
different capacitances in the system. E.g. the capacitances in the overhead line will cause
natural frequencies in the same way as the two cable sections did.
34
4.2
Simplified model
In this section, a simplified model is produced on the basis of the system, shown in figure 3.21 on
page 29. For this simplification, different parts of the system are compared, and parameters
deemed with little significance are neglected. The objective of this simplification is a less
complicated model, which can be analyzed mathematically with more ease.
A circuit breaker in a closing operation behaves near ideally, and as the project deals with the
switch-on of a grid segment, the circuit breaker is assumed to operate as an ideal switch. The
series resistance in the circuit breaker is assumed of no influence compared to the other resistances
in the circuit, and is therefore neglected.
As the length of section 1 is only 0,43 km, this section is negligible by comparison to the length
of the other sections2 . This leaves two sections, one 18,93 km long overhead line section and one
21 km long underground cable section. As the phase values for the parameters in the system
do not differ much between the three phases, average values are used in the simplified model,
resulting in equal parameters for the three phases. The simplification is therefore explained for
only one phase. A comparison of the parameters in the two sections is shown in table 4.2.
C
L
R
G
Section 2
Overhead line
154 nF
35,6 mH
0,898
132 nS
Section 3
Underground cable
5,10 F
7,20 mH
0,693
265 nS
Reactor
2,306 H
1,17
Table 4.2: Comparison of parameters between overhead lines and underground cable. Average phase
values are used.
From table 4.2 it can be seen, that the capacitance in the cable is approximately 30 times the
capacitance of the overhead lines. It is therefore reasonable for the overhead line section, to
neglect the capacitance, leaving only the series impedance. The shunt conductance in both
sections are considered to be of no influence, and are therefore neglected.
The mutual inductances in the shunt reactor, the overhead line section and the underground
cable section are compared in table 4.3.
Shunt reactor
Mab
Mac
Mbc
-32,0 mH
-23,0 mH
-32,0 mH
Section 2
Overhead line
13,6 mH
10,9 mH
11,0 mH
Table 4.3: Comparison of mutual inductances between the reactor, overhead line and underground
cable.
2
With this simplification, the high frequency transient caused by this section will not be present in the analysis.
The simplification is taken to obtain a simple circuit.
4.2 Simplified model
35
From the table it can be seen that the mutual inductances in the shunt reactor are greater than the
mutual inductances in the overhead line section with a factor of ca. 2. The mutual inductances
in the cable is not provided but is assumed to be smaller than the mutual inductances in the
overhead line, due to the small distance between the conductors. As the mutual inductances in
the reactor are the greatest, these are the only mutual coupling included in the simplification. The
issue concerning mutual couplings will therefore be included, without complicating the model too
much. All mutual inductances and capacitances in the transmission lines are therefore neglected.
With the above-mentioned considerations, phase a from figure 3.21 can be simplified as shown
in figure 4.4.
Section 2
Rs
Ls
Section 3
Rsection2 Lsection2
ua (t)
Rsection3
Lsection3
1
C
2 section3
1
C
2 section3
Reactor
Rreactor
Lreactor
Mab
Mac
Figure 4.4: Simplified representation for phase a.
The simplified circuit for phase a in figure 4.4 has two capacitances. The analysis is considered
to be less complicated if only one capacitance is present, the cable section is therefore expressed
with a T -equivalent as shown in figure 4.5.
Section 2
Rs
Ls
Rsection2 Lsection2
Reactor
Section 3
1
R
2 section3
1
L
2 section3
1
R
2 section3
1
L
2 section3
Rreactor
ua (t)
Csection3
Lreactor
Mab
Mac
Figure 4.5: Simplified model of phase a with cable represented by a T-equivalent.
The circuit in figure 4.5 is simplified further by unifying the series impedances resulting in the
circuit shown in figure 4.6
36
Rs
Ls
R1
L1
R2
ua (t)
C
L2
Mab
Mac
Figure 4.6: Simplified representation of phase a.
Where the following applies:

R1 = Rsection2 + 21 Rsection3
R2 = 21 Rsection3 + Rreactor
L1 = Lsection2 + 21 Lsection3
L2 = 21 Lsection3 + Lreactor
C = Csection3
With these simplifications, the three phase representation for the system, can be expressed as
shown in figure 4.7.
Rs
Ls
ia (t) R1
L1
i2,a (t)
+
i1,a (t)
ua (t)
u2,a (t)
u1,a (t)
R2
C
L2
Rs
Ls
ib (t) R1
L1
i2,b (t)
+
i1,b (t)
ub (t)
u2,b (t)
u1,b (t)
R2
C
Mab
Mac
L2
Rs
Ls
Mbc
ic (t) R1
+
L1
i2,c (t)
+
i1,c (t)
uc (t)
u1,c (t)
u2,c (t)
R2
C
L2
Figure 4.7: The simplified model of the system.
4.3 Mathematical analysis
37
The voltages u1,a (t), u1,b (t) and u1,c (t) and currents ia (t), ib (t) and ic (t) shown in figure 4.7
correspond to the voltages and currents measured in Ferslev.
The parameter values for the circuit model, shown in figure 4.7, are listed in table 4.4.
1,245
1,517
0,800
R1
R2
Rs
L1
L2
Ls
39,2 mH
2,310 H
14,4 mH
Mab
Mac
Mbc
-32,0 mH
-23,0 mH
-32,0 mH
5,10 F
Table 4.4: Parameters for the simplified circuit shown in figure 4.7.
The equations describing the circuit can be found by analyzing the circuit. Only one phase is
shown in figure 4.8, as this will give a clearer view of the analysis. The principle used for phase
a is used for the other two phases as well.
ia
Rs
R1
Ls
+
ua
u1,a
L1
i2,a
i1,a
R2
u2,a
C
L2
Mab
Mac
Figure 4.8: Phase model for mathematical analysis

l
Under the general notation of the voltage drop across an inductor, ul = L di
dt , and the current
dul
into a capacitor, ic = C dt the equations describing the system can be found. The circuit in
figure 4.8, can be expressed by four equations:
dia
dia
+ R1 ia + L1
+ u2,a
dt
dt
di2,b
di2,a
di2,c
= R2 i2,a + L2
Mab
Mac
dt
dt
dt
du2,a
= C
dt
= i2,a + i1,a
ua = Rs ia + Ls
u2,a
i1,a
ia
(4.7)
(4.8)
(4.9)
(4.10)
In the next section, a solution for the equations describing the system, is found.
4.3
Mathematical analysis
In the following section, the simplified model of the system is analyzed. The analysis is conducted
with two different approaches, namely an analytical approach and a numerical approach. An
analytical solution of the equations describing the system will have a general appearance. Such
solution can be obtained without knowing the values for the components, allowing for a direct
38
analysis of the significance of each component in the system. A numerical solution will only give
results for a specific system, and will thereby not provide any information about the significance
of the particular parameters.
An analytical solution of the circuit would be preferable, as this will give a general expression for
the voltages and currents in the system. An analytical approach is attempted in the following.
4.3.1
Analytical approach
The principle for the analytical solution is to find the current into the circuit. The desired voltage
u1,a after the short circuit impedance, shown in figure 4.8, is then found from the current. This
approach lets the short circuit resistance and inductance being unified with R1 and L1 from
figure 4.8. The mutual inductances are neglected in this approach, as this allows analysis of one
phase at a time. Furthermore, it is assumed that the mutual inductances will only have a small
influence on the voltages and currents in the circuit.
The simplified circuit is shown in figure 4.9, where the circuit is expressed in the Laplace domain.
R1
Ia (s)
sL1
I2,a (s)
I1,a (s)
Ua (s)
R2
1
sC
Zin
+
U2 (s)
sL2
Figure 4.9: Phase model for mathematical analysis
Where the components in the circuit shown in figure 4.9 are:

R1 = Rsection2 + 12 Rsection3 + Rs
L1 = Lsection2 + 21 Lsection3 + Ls
C = Csection3
The initial values for the currents and voltages in the circuit are assumed to be zero, as the
system is assumed totally discharged at the switch-on of the system. The current into the circuit
Ia (s) can be found from the voltage Ua (s) and the input impedance Zin . The input impedance
is found to be:
Zin = R1 + sL1 +
Zin =
s3 CL1 L2 +
1
sC (R2 + sL2 )
1
sC + (R2 + sL2 )
s2 C(L1 R2 + L2 R1 ) + s(CR1 R2
s2 CL2 + sCR2 + 1
(4.11)
+ L 1 + L 2 ) + R1 + R2
(4.12)
The current Ia (s) can now be found:

1
Ia (s) = Yin Ua (s) = Zin
Ua (s)
Ia (s) =
(4.13)
s2 CL
s3 CL
1 L2
s2 C(L
+ sCR2 + 1
Ua (s) (4.14)
1 R2 + L2 R1 ) + s(CR1 R2 + L1 + L2 ) + R1 + R2
2
39
The voltage Ua (s) is known, and can be expressed in the Laplace domain:
Ua (s) = L [Up sin(t )]
(sin()s + cos())
Ua (s) = Up
s2 + 2
(4.15)
(4.16)
Where:
Up is the peak voltage of the voltage source
is the power frequency of the voltage source
is the phase angle of the voltage source
The expression for the input current is then:
Ia (s) =
(s2 CL2 + sCR2 + 1) Up (sin()s + cos())

(4.17)
(s3 CL1 L2 + s2 C(L1 R2 + L2 R1 ) + s(CR1 R2 + L1 + L2 ) + (R1 + R2 ))(s2 + 2 )
With the current obtained, the voltage U1,a (s), represented with u1,a (t) in figure 4.8, can be
found by subtracting the voltage drop across short circuit impedance from the source voltage:
U1,a (s)
U1,a (s)
Ua (s) Ia (s) (Rs + sLs )

(4.18)
(sin()s + cos())
...
= Up
s2 + 2
(s2 CL2 + sCR2 + 1) Up (sin()s + cos()) (Rs + sLs )
...
(4.19)
(s3 CL1 L2 + s2 C(L1 R2 + L2 R1 ) + s(CR1 R2 + L1 + L2 ) + (R1 + R2 ))(s2 + 2 )
=
In order to obtain an expression in the time domain, the inverse Laplace transform must be
applied. To use this, an often used method is partial fraction. In partial fraction, the polynomial
is fractioned into an expression from which the inverse Laplace transform can easily be found.
In general, partial fraction is performed by evaluating the poles of a polynomial. The poles are
then used to expand the denominator into fractions, containing one pole each:
F (s) =
num(s)
(s p1 )(s p2 ) . . . (s pn )
(4.20)
By use of partial fraction the coefficients (r1 , r2 , . . . , r3 ) can be found from the following relations:
r1
r2
rn
num(s)
=
+
+ ... +
(s p1 )(s p2 ) . . . (s pn )
(s p1 ) (s p2 )
(s pn )
(4.21)
The coefficients are found by replacing s with each of the poles (p1 , p2 , . . . , pn ), one at a time:
r1 =
r2 =
..
.
rn =

num(s)

(s p2 )(s p3 ) . . . (s pn ) s=p1

num(s)

(s p1 )(s p3 ) . . . (s pn ) s=p2

num(s)

(s p1 )(s p2 ) . . . (s pn1 ) s=pn
(4.22)
(4.23)
(4.24)
40
When the coefficients are found, the inverse Laplace transform can be applied on the right hand
side of equation 4.21, which gives:
f (t) = r1 ep1 t + r2 ep2 t + . . . + rn epn t
(4.25)
As shown in equation 4.17 and 4.19, the denominator of the expressions for Ia (s) and U1,a (s)
will be of 5th order. The algebraic expression for each pole will then become large, and thereby
difficult to analyze and comprehend. It is therefore chosen to insert the values for each of the
components, resulting in an numerical solution, and not the desired analytical representation.
The expressions are therefore solved numerically in MatLab3 . The partial fraction and inverse
Laplace transform of the current in phase gives:
ia (t) = (0, 233 j652, 073)e(18,651+j1934,601)t + (0, 233 + j652, 073)e(18,651j1934,601)t . . .
+ (0, 202 + j15, 163)ej314,159t + (0, 202 j15, 163)ej314,159t 0, 870e1,507t
(4.26)
In the same manner, the partial fraction and inverse Laplace transform of the voltage is found:
u1,a (t) = (18166 + j340)e(18,651+j1934,601)t + (18166 j340)e(18,651j1934,601)t . . .
+ (67429 j13)ej314,159t + (67429 + j13)ej314,159t + 0, 677e1,507t
(4.27)
The solution of u1,a is explained in the following. From the Euler identities, cos() is expressed
as:
ej + ej
(4.28)
cos() =
2
By taking the absolute value of the complex magnitude (18166 + j340) and (18166 j340)
a magnitude of -18169 is found. The first term of equation 4.27 is be simplified:
18169 (e(18,651+j1934,601)t + e(18,651j1934,601)t )
18,651t j1934,601t
18169 (e
18,651t j1934,601t
+e
(4.30)
18,651t
(4.31)
18169 2 cos(1934, 601t) e18,651t
(4.32)
j1934,601t
18169 (e
j1934,601t
(4.29)
+e
)e
By use of equation 4.28, the term is further simplified:

18,651t
(4.33)
36338 cos(1934, 601t) e
With this procedure a new expression of u1,a (t) is found:

1,507t
u1,a (t) = 134858cos(314, 159t ) 36338cos(1934, 601t ) e| 18,651t
(4.34)
{z } + 0, 677e
{z
50Hz
{z
308Hz
=53,6 ms
{z
DC =664 ms
It can now be seen that the voltage consist of a 50 Hz cosine with 134,858 kV magnitude.
Superimposed is a 308 Hz cosine with a magnitude of -36,338 kV, further damped by a 53,6 ms
time constant. Furthermore, the voltage has a DC component of 0,677 V damped by a time
constant of 664 ms.
It is now possible to obtain expressions in the time domain for the currents and voltages in the
two other phases, by using the same method. The results for the three phases are plotted and
compared in the last section of this chapter.
As shown, the solution to a second order system can be difficult to conduct, with an analytical
method, in this case even with further simplification in the form of mutual inductances. In the
following section, the simplified circuit is analyzed through a numerical approach.
3
The m-file is shown in appendix G.8
4.3.2
41
Numerical approach
The numerical solution is conducted with the same principle as in the analytical approach, by
first finding the current into the circuit. The voltage u1 after the short circuit impedance, is then
found from this current. The circuit for this analysis is shown in figure 4.10.
ia
R1
L1
i2,a
i1,a
ua
R2
u2
C
L2
Mab
Mac
Figure 4.10: Phase model for mathematical analysis.
Where the components in the circuit shown in figure 4.10 are:

R1 = Rsection2 + 12 Rsection3 + Rs
L1 = Lsection2 + 12 Lsection3 + Ls
C = Csection3
These equations describing the voltages an currents in the simplified system are:
dia
+ u2,a
dt
di2,b
di2,a
di2,c
= R2 i2,a + L2
Mab
Mac
dt
dt
dt
du2,a
= Ca
dt
= i2,a + i1,a
ua = R1 ia + L1
u2,a
i1,a
ia
(4.35)
(4.36)
(4.37)
(4.38)
The wanted values from these equation are besides ia and u2,a , the current i2,a flowing through
inductance L2 , as this is needed for the mutual couplings. These three variables are isolated on
the left side:
dia
dt
di2,a
L2
dt
du2,a
Ca
dt
L1
(4.39)
= ua R1 ia u2,a
= u2,a i2,a R2,a Mab
= ia i2,a
di2,b
di2,c
Mac
dt
dt
(4.40)
(4.41)
By integrating on both sides of equation 4.39, 4.40 and 4.41, expressions for the currents and
voltages are found. As this analysis focuses on the reconnection of the grid, the system is assumed
42
totally discharged at the moment of switch-on of the system, and initial values of currents and
voltages therefore zero. The integration constants can therefore be neglected.
ia =
i2,a =
u2,a =
Z
ua dt R1
Z
ia dt
1
u2,a dt R2
L2
Z
1
(ia i2,a ) dt
C
u2,a dt
1
L1
(4.42)
i2,a dt Mab i2,b Mac i2,c
(4.43)
(4.44)
The expressions above are solved by constructing a numerical solver in Matlab-Simulink. The
numerical solver, consists of the blocksets shown in figure 4.11.
Signal block, for input voltage
[U_a]
1
s
1/L_1
[I_1_a]
From block, takes input from goto blocks
du/dt
Integral block
Gain block
Goto block, sends signals to from block
Differentiation block
Summation block
Subsystem, used to write data to workspace
Ramp, used to trigger switch
Switch block
Figure 4.11: The blocks used in the numerical solver.
The solver is constructed from equation 4.42, 4.43 and 4.44 and shown in figure 4.12. The values
for the parameters are given in table 4.4 on page 37, where R1 and Rs is unified along with
L1 and Ls . The values for the mutual inductances are inserted with opposite sign, as they are
defined to be negative in the numerical solver.
[U_a]
1
s
1/L_1
[I_a]
1
s
1/L_2
R_1
1/C
[U_2_a]
[I_2_a]
R_2
1
s
1
s
1
s
Mab
Mac
[I_a]
[U_2_a]
[I_2_a]
[I_2_b]
[I_2_c]
Figure 4.12: Numerical solver for ia , i2,a and u2,a , corresponding to equations 4.42, 4.43 and 4.44.
43
By use of the solver, shown in figure 4.12, numerical solutions for the signals ia , i2,a and u2,a are
achieved.
The next step is to find the voltage u1,a measured in Ferslev, shown in figure 4.8. This voltage
can be found by subtracting the voltage drop across the short circuit impedance from the grid
voltage ua . This operation is shown in figure 4.13 and is described by:
u1 = ua us
u1 = ua (Rs ia + Ls
(4.45)
dia
)
dt
(4.46)
The subsystem shown in figure 4.13, serves the purpose of writing results to the Matlab
workspace.
[U_a]
[I_a]
du/dt
[I_a]
Ls
Rs
[U_b]
[I_b]
du/dt
[I_b]
Ls
Rs
U_Ferslev
voltage
writeout
[U_c]
[I_c]
[I_c]
du/dt
Ls
Rs
Figure 4.13: Solver to find ua , by subtracting us from ua , corresponding to equations 4.45 and 4.46.
By using the above described method for all phases, a Simulink model is constructed that can
solve the simplified model, shown in figure 4.7. This solver is shown in figure 4.14.
44
[U_a]
1
s
[U_a]
1/L_1
[I_a]
1
s
[U_b]
1/C
1/L_2
R_1
[U_c]
[I_2_a]
R_2
1
s
1
s
1
s
Mab
Mac
[I_a]
[U_2_a]
[I_2_a]
[I_2_b]
[I_2_c]
[U_b]
[U_2_a]
[I_a]
[I_b]
1
s
[I_c]
1/L_1
[I_b]
1
s
1/C
1/L_2
R_1
Input
current
writeout
[U_2_b]
[I_2_b]
R_2
[U_2_a]
1
s
1
s
1
s
Mab
Mbc
[I_b]
[U_2_b]
[I_2_b]
[I_2_a]
[I_2_c]
[U_2_b]
[U_2_c]
[U_c]
U_2
voltage
writeout
1
s
1/L_1
[I_c]
1
s
1/L_2
R_1
[U_a]
1/C
[U_2_c]
[I_a]
du/dt
Ls
[I_2_c]
[I_a]
R_2
Rs
[U_b]
1
s
1
s
1
s
Mac
Mbc
[I_c]
[U_2_c]
[I_2_c]
[I_2_a]
[I_2_b]
[I_b]
du/dt
[I_b]
Ls
Rs
U_Ferslev
voltage
writeout
[U_c]
[I_c]
[I_c]
du/dt
Ls
Rs
Figure 4.14: The constructed simulink solver, which is used for analysis of the simple model.
The results from the numerical approach are compared with the analytical approach in the last
section of this chapter. In order to verify the results from the two approaches, a simulation of
the simplified circuit is performed in PSCAD, this is described in the next section.
4.4
PSCAD simulation
In order to verify the results obtained through the numerical solution in Simulink, a simulation
in PSCAD is performed. The setup for the PSCAD simulation is shown in figure 4.15.
4.4 PSCAD simulation
BRK
A
B
45
R=0
0.800 [ohm] 0.0144 [H] U_1a
I_a
0.0392 [H]
1.245 [ohm]
U_2a
0.0392 [H]
1.245 [ohm]
U_2b
0.0392 [H]
1.245 [ohm]
U_2c
I_b
0.800 [ohm] 0.0144 [H] U_1b
B
I_c
C
0.800 [ohm] 0.0144 [H] U_1c
ic
Rc
5.10 [uF]
5.10 [uF]
5.10 [uF]
BRK
Timed
Breaker
Logic
Open@t0
Lc
Rb
Lb
oc
ib
ia
Ra
La
ob
oa
Figure 4.15: Graphical configuration in PSCAD for simulation.
The component values given in table 4.4 are entered. As the source impedance is present in the
circuit model, an ideal source is used in PSCAD.
4.4.1
The reactor component in PSCAD
The reactor component, is constructed with background in the reactor described in section 3.6 on
page 20. PSCAD does not contain a component for the shunt reactor, a new component is therefore
created, in order to include the mutual inductances in the reactor. The graphical layout of the
created component is shown in figure 4.16.
ic
ib
Rc
Rb
Ra
Lc
Lb
La
oc
ob
ia
oa
Figure 4.16: Graphical representation of the shunt reactor in PSCAD.
In PSCAD, mutually coupled windings are constructed through the use of the # TRANSFORMERS
script directive. A basic Transformers Segment has the general format [9, PSCAD User manual]:
1
2
3
4
5
# T R A N S F O R M E R S < Number_of_ T r an s f or m e rs >

< Prefix > < Number_of_Wi n di n gs > /
< Node_1 > < Node_2 >
< R_11 > < L_11 > /
< Node_2 > < Node_3 >
< R_12 > < L_12 > < R_22 > < L_22 > /
...
Here <Node_1> <Node_2> represents the two nodes, which the following expression takes place
between. The expression <R_12> <L_12> <R_22> <L_22> can then be conceived as the equation
unode1 unode2 = R12 i1 + L12 didt1 + R22 i2 + L22 didt2 .
Using this, the computation script can be written from the inductance matrix for the shunt
reactor and the resistance in each winding:
1
2
3
4
5
# TRANSFORMERS 1
3 /
$ia $oa
$Ra $La
$ib $ob
0
$Mab
$ic $oc
0
$Mac
0
$Mab
$Rb $Lb
0
$Mbc
0
$Mac /
0
$Mbc /
$Rc $Lc /
46
Where:
$ia, $ib and $ic are the input nodes for each phase.
$oa, $ob and $oc are the output nodes for each phase.
$Ra, $Rb and $Rc are the resistances in the windings.
$La, $Lb and $Lc are the inductances in the windings.
$Mab, $Mac and $Mbc are the mutual inductances between the windings.
Note that the mutual resistances are set to zero, in order not to contribute with voltage between
the two nodes.
The values for the reactor is entered in the configuration as shown in figure 4.17.
Figure 4.17: Configuration of the shunt reactor in PSCAD.
The resistance R2 and inductance L2 from figure 4.7 are entered in the reactor interface, resulting
in: R2 = Ra = Rb = Rc and L2 = La = Lb = Lc . The values are found in table 4.4 on page 37.
With the parameters entered, the system is simulated. The results from this simulation can now
be compared with the two mathematical solutions and the actual measured values.
4.5
Results
In the following section, the waveforms from the analytical approach (section 4.3.1), the numerical
approach in Simulink (section 4.3.2) and the simulation in PSCAD (section 4.4) are compared with
the measured waveforms4 . The comparison will be performed for u1 and i, which represent
the voltage and current measured in Ferslev. The switch-on instant, for simulations and
measurement, is at t=0.
4
The measurements in Ferslev are described in appendix C.
4.5 Results
4.5.1
47
Comparison of u1
Measurements [kV]
PSCAD simulation [kV]
Numerical approach [kV]
Analytic approach [kV]
The voltage waveforms for u1 on phase a are shown in figure 4.18. The waveforms are taken
from the analytical approach, numerical solution in Simulink, the simulation in PSCAD and the
measurements in Ferslev.
200
100
0
100
200
0
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
200
100
0
100
200
200
100
0
100
200
200
100
0
100
200
Time [s]
Figure 4.18: Comparison of the voltage u1 on phase a.
The voltage u1 on phase b is shown, for the analytical approach, numerical solution in Simulink,
the simulation in PSCAD and the measurements in Ferslev, in figure 4.19.
Measurements [kV]
48
200
100
0
100
200
0
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
200
100
0
100
200
200
100
0
100
200
200
100
0
100
200
Time [s]
Figure 4.19: Comparison of the voltage u1 on phase b.
Measurements [kV]
The voltage u1 on phase c is shown for the analytical approach, numerical solution in Simulink,
the simulation in PSCAD and the measurements in Ferslev, in figure 4.20.
200
100
0
100
200
0
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
200
100
0
100
200
200
100
0
100
200
200
100
0
100
200
Time [s]
Figure 4.20: Comparison of the voltage u1 on phase c.
The three figures above, indicate that the voltage u1 is identical for the numerical solution in
Simulink and the simulation in PSCAD. The analytical approach gives a slightly different result,
which is assumed to stem from the neglected mutual inductance in the reactor. Compared to
4.5 Results
49
measurements, there is much lesser damping in the simulation results. The simulation results
also indicate, that switch-on at the peak gives the greatest oscillation.
4.5.2
Comparison of i for a duration of 0,15 s
Measurements [kA]
PSCAD simulation [kA]
Numerical approach [kA]
Analytic approach [kA]
The current i in phase a is shown for the analytical approach, numerical solution in Simulink,
the simulation in PSCAD and the measurements in Ferslev, in figure 4.21, for a duration of 0,15 s.
1
0
1
0
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
1
0
1
1
0
1
1
0
1
Time [s]
Figure 4.21: Comparison of the current i in phase a for a duration of 0,15 s.
The current i in phase b is shown for the analytical approach, numerical solution in Simulink,
Measurements [kA]
50
1
0
1
0
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
1
0
1
1
0
1
1
0
1
Time [s]
Figure 4.22: Comparison of the current i in phase b for a duration of 0,15 s.
Measurements [kA]
1
0
1
0
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
0.05
0.1
0.15
1
0
1
1
0
1
1
0
1
Time [s]
Figure 4.23: Comparison of the current i in phase c for a duration of 0,15 s.
The simulation current waveforms are all comparable, while the measured deviates from the
others. Lack of damping is clearly seen compared with measurements, as the damping time of
the current is larger than the damping time of the measurements.
4.5 Results
4.5.3
51
Comparison of i for a duration of 2,8 s
Measurements [kA]
The current i in phase a is shown for the analytical approach, numerical solution in Simulink,
0.5
0.5
0.5
1.5
2.5
0.5
1.5
2.5
0.5
1.5
2.5
0.5
1.5
2.5
0.5
0.5
0.5
0.5
0.5
0.5
Time [s]
Figure 4.24: Comparison of the current i in phase a for a duration of 2,8 s.
Measurements [kA]
0.5
0.5
0.5
1.5
2.5
0.5
1.5
2.5
0.5
1.5
2.5
0.5
1.5
2.5
0.5
0.5
0.5
0.5
0.5
0.5
Time [s]
Figure 4.25: Comparison of the current i in phase b for a duration of 2,8 s.
52
Measurements [kA]
The current i in phase c is shown for the analytical approach, numerical solution in Simulink,
0.5
0.5
0.5
1.5
2.5
0.5
1.5
2.5
0.5
1.5
2.5
0.5
1.5
2.5
0.5
0.5
0.5
0.5
0.5
0.5
Time [s]
Figure 4.26: Comparison of the current i in phase c for a duration of 2,8 s.
Measured
110 A
Simulated
60 A
Table 4.5: A comparison of steady state magnitude at 150 kV level, for the simplified system.
Once again, all the current waveforms from simulations are all comparable, while the measured
deviates from the others. The time constant for the DC component is though of the same size in
both simulations and measurements. As it can be seen from the figures and table 4.5.3, the steady
state current from the measurement has considerably higher magnitude than for the simulated
currents.
4.5.4
Harmonic analysis of u1
The harmonic analysis for u1 on phase a is shown in figure 4.27.
4.5 Results
53
20
Analytical
10

150 kV level
0
0
1000
2000
3000
4000
5000
20
6000
Simulink
10
0
0
1000
2000
3000
4000
5000
20
6000
Pscad
10
0
0
1000
2000
3000
4000
5000
20
6000
Measurement
10
0
0
1000
2000
3000
4000
Frequency [Hz]
5000
6000
Figure 4.27: Harmonic analysis of the voltage u1 for phase a.
The harmonic analysis for u1 on phase b is shown in figure 4.28.
20
Analytical
10

150 kV level
0
0
1000
2000
3000
4000
5000
20
6000
Simulink
10
0
0
1000
2000
3000
4000
5000
20
6000
Pscad
10
0
0
1000
2000
3000
4000
5000
20
6000
Measurement
10
0
0
1000
2000
3000
4000
Frequency [Hz]
5000
Figure 4.28: Harmonic analysis of the voltage u1 for phase b.
The harmonic analysis for u1 on phase c is shown in figure 4.29.
6000
54
20
Analytical
10

150 kV level
0
0
1000
2000
3000
4000
5000
20
6000
Simulink
10
0
0
1000
2000
3000
4000
5000
20
6000
Pscad
10
0
0
1000
2000
3000
4000
5000
20
6000
Measurement
10
0
0
1000
2000
3000
4000
Frequency [Hz]
5000
6000
Figure 4.29: Harmonic analysis of the voltage u1 for phase c.
The Fourier analysis shows the same frequencies for the analytical approach, the numerical
approach and the simulation in PSCAD. Here, the largest frequency component is the 6th harmonic.
For the measurements the largest frequency component is the 7th harmonic. As expected the
frequencies around 3,8 kHz experienced in the measurements is not present in the simulations,
due to the neglected short cable section.
4.6
The solutions from the numerical approach in Simulink and the simulation in PSCAD give
identical results, both for the voltages u1,a,b,c , and the input currents ia,b,c . Small difference
is apparent between the analytical approach conducted in MatLab and the results from Simulink
and PSCAD. This difference is assumed to stem from the neglected mutual inductance in the
analytical approach, as that is the only factor separating the solutions. The small magnitude
of this difference indicates, the mutual inductances have only small influence on the generated
frequencies.
As expected, the high frequency voltage component appears in none of the simulations. This is
due to the fact that the short cable section is neglected in the simulation.
Common for all simulated voltage waveforms, is the appearance of the 6th harmonic, which is the
dominant one, and in accordance to the frequency predicted in section 4.1. For the measurements,
the 7th harmonic has the highest magnitude. This difference can be explained by the fact, that the
source inductance is estimated, which will influence the frequencies. Furthermore, the difference
can originate from the fact, that simplifications are made using average values, and distributed
parameters are in the measured system, whereas the cable section cannot be seen purely as a
capacitance.
Similarities between voltage waveforms and the Fourier analysis of the first period after switching,
in all simulations, indicate correct solution of the simplified model.
4.6 Summary & discussion
55
Similarities between simulation and test results are present, however a noticeable deviation is in
the steady state current magnitude, and the damping rate. The steady state current magnitude,
has a value of ca. 110 A for measurements, whereas in simulation the magnitude is ca. 60 A.
This difference can be explained by the neglected shunt capacitances and conductances in the
overhead lines and the capacitances in the short cable section.
Damping in the measurements is much larger than in the simplified model. In an oscillating
system, the damping is given by resistances, which are energy consuming elements. When
simplifying the three phase representation of the system, shown in figure 3.21, care was taken
not to neglect elements adding to the damping ratio of the system. The AC-resistance was
therefore estimated for the overhead line. There are though a few factors that can influence the
transient behaviour of the system, these are discussed below:
The simplified model does not use distributed parameters
It is therefore difficult to make valid assumptions about frequency dependent components,
as the parameters are distributed and not lumped as assumed. The use of T equivalent,
for the long cable section, results in only one frequency component and thereby a high
magnitude in the Fourier analysis. The measured frequencies will, due to the distributed
parameters, contain several of frequency components, this can also be seen in the Fourier
analysis of the measurements. This is assumed to be the reason for difference between the
Fourier analysis for measurement and simulations.
Mutual inductances in overhead lines and underground cable are not included in the simplified model
As shown, only very small differences between the analytical (without mutual inductances)
and the numerical approach (with mutual inductances) were present. The mutual inductances are therefore considered to have only little effect.
Neglected capacitances in the overhead line section
These capacitances will give rise to high frequency components, though with small
magnitude. If the transmission line is considered as a -equivalent, the current running into
one-half of the line capacitance, will give rise to frequency around 6 kHz. This frequency
band can be seen in the measurements, where frequencies around 5700 Hz is present.
Shunt conductance neglected in simplified model
This is one of the simplifications that decrease the damping ratio. There is a reason to
believe that this will not add significantly to the damping, as simulations were briefly
conducted with this component present, without noticeable increase of damping ratio.
Neglected short cable section
The 430 m cable section in Ferslev, together with the source inductance, gives rise to a
frequency component of 4 kHz which is therefore not experienced in simulations. This is
one of the greater deviations between measurements and simulations.
Skin- and proximity effect for higher frequencies
The resistance in the transmission line is frequency dependent as a function of the skin
depth. This will lead to greater resistance for frequencies higher than 50 Hz, and thereby
greater damping. This, along with proximity effect in the cables, is assumed to be of
great significance for the validity of simulations, as the overvoltages are caused by higher
frequencies.
56
This chapter has dealt with analysis of the transient behaviour of the Ferslev-Tinghj grid
section, with focus solely on the 150 kV level. First, the system was simplified to find its
natural frequency response, in order to explain the measured frequencies. In order to ease
the mathematical analysis, the three phase representation of the system, shown in figure 3.21,
was simplified. When attempting to solve the differential equations describing the simplified
system, it appeared that the analytical solution was too complicated to comprehend. These
equations were therefore solved numerically, and a simulation of the circuit model conducted, in
order to evaluate the validity of the mathematical solutions. Simulations bear resemblance with
measurements, large difference in the damping is though present.
In order to perform a more precise analyze of the system, a more detailed system simulation is
conducted with the simulation software PSCAD. Here models for overhead lines and underground
cables are included, which eliminates the simplifications. This is done in the next chapter.
Chapter 5
Configuration and simulation in

PSCAD
The objective of this chapter is to obtain a more precise simulation of the system, in order to
obtain a more valid estimation of the switching overvoltages. The configuration of the subparts,
shown in figure 5.1, are first described, whereafter the simulation is performed and the results
are presented.
5.1
PSCAD model subparts
The simulation model for the Ferslev-Tinghj grid section, is shown in figure 5.1.
C
Section1
BRK1
A
B
R=0
0.8 [ohm]
14.4 [mH]
0.8 [ohm]
14.4 [mH]
I150k_a
I150k_b
S1
0.8 [ohm]
14.4 [mH]
C1 Section3
S1
S1
1
C2
C2
S2
Section2
Section2
1 [ohm]
1 [ohm]
Section3 C1
S1
U_a
S2
U_b
C3
C3
S3
3 [ohm]
3 [ohm]
S3
1 [ohm]
1 [ohm]
C2
S2
C3
S3
1 [ohm]
C2
S2
1 [ohm]
C3
C
Section3
Section1C1
1 [ohm]
I150k_c
T
Section2
C1 Section1
1 [ohm]
1 [ohm]
S3
U_c
ic
Rc
Lc
BRK1
Timed
Breaker
Logic
Open@t0
Rb
80.0 [MVA]
U60k_a
B
#1
#2
U150k_b
150 [kV]
C
60 [kV]
#1
#2
U04k_a
B
#1
60.0 [kV]
20.0 [kV]
C
U20k_c
oa
0.630 [MVA]
#2
24 [kV]
1 [ohm]
U20k_b
C
U60k_c
A
U20k_a
B
U60k_b
C
U150k_c
16.0 [MVA]
La
ob
U150k_a
ia
Ra
Lb
oc
ib
U04k_b
1[ohm]
U04k_c
1 [ohm]
C
0.42 [kV]
Figure 5.1: PSCAD model of the system.
The model is a combination of a voltage source, source impedances, circuit breakers, two cable
sections, one overhead line section, a shunt reactor, and three transformers. The configuration
of these subparts is described in the following.
5.1.1
Voltage source
The voltage source is represented by an ideal Three Phase Voltage Source, as shown in figure 5.2,
with resistance and inductance in series, to represent the short circuit impedance. The values
for the short circuit impedance is found in section 3.2 on page 9.
A
B
R=0
0.8 [ohm]
0.0144 [H]
0.8 [ohm]
0.0144 [H]
0.8 [ohm]
0.0144 [H]
Figure 5.2: Representation of the source in PSCAD.
58
The configuration of the source is shown in fig 5.3. The voltage level is configured to 165 kV, as
this was the voltage measured in Ferslev, and a phase shift of -90 , in order to match a switch-on
moment at peak for phase a.
Figure 5.3: Configuration of the source in PSCAD.
5.1.2
Circuit breaker
To simulate the reconnection of the grid, the Three Phase Breaker is used as shown in figure 5.4.
BRK1
A
B
C
BRK1
Timed
Breaker
Logic
Open@t0
Figure 5.4: Representation of the breaker in PSCAD.
The breaker is controlled by the Timed Breaker Logic, shown at the bottom in figure 5.4. The
breaker is configured to switch-on after 0,01 s, as shown in figure 5.5, in order to have phase a in
peak at the moment of switch-on. It has been checked, that the current is in steady state at this
moment, thereby will the switching transients not be influenced by transformer inrush current.
5.1 PSCAD model subparts
59
Figure 5.5: Configuration of the circuit breaker and timed breaker logic.
The values for the resistances in figure 5.5 is given in section 3.3, where the earth resistance is
assumed 1 as the cable is grounded in the station, and will hence have low resistance value.
5.1.3
Section 1
Section 1, which is a 430 m long underground cable section, is represented by the Cable
Component and Cable Interface as shown in figure 5.6.
C
Section1
C1
Section1
S1
1 [ohm]
Section1
S1
C1
C2
C2
S2
1 [ohm]
S2
C3
C3
S3
1 [ohm]
S3
Figure 5.6: Representation of section 1 in PSCAD.
The physical layout configuration of the cable component is shown in figure 5.7.
60
Ground Resistivity: 100.0 [ohm*m]

Relative Ground Permeability: 1.0
Earth Return Formula: Analytical Approximation
Cable # 1
0 [m]
-0.046 [m]
Cable # 2
1.134[m]
Cable # 3
1.21367 [m]
Conductor
Insulator 1
Sheath
Insulator 2
1.21367 [m]
Conductor
Insulator 1
Sheath
Insulator 2
6.97 [mm]
20.75 [mm]
39.05 [mm]
39.435 [mm]
46 [mm]
0.046 [m]
Conductor
Insulator 1
Sheath
Insulator 2
6.97 [mm]
20.75 [mm]
39.05 [mm]
39.435 [mm]
46 [mm]
6.97 [mm]
20.75 [mm]
39.05 [mm]
39.435 [mm]
46 [mm]
Figure 5.7: Configuration of section 1 in PSCAD.
The analytical approximation for solving the ground impedance integral is chosen in preference
to the numerical integration, as this is more time effective and recommended in the PSCAD help
file. The solution is normally accurate to within 5% of exact solution [9].
Note should be taken, that the conductor is configured to contain a hollow centre. This is done
in order to keep the outer diameter in the right dimension, without increasing the conductor
cross section. In the datasheet for the cable, the conductor cross section is defined as 1200 mm2 ,
and the diameter of the conductor is defined to be 41,5 mm, as shown in figure 3.5 on page 13.
The area difference1 can then be calculated:
41, 5
2
2
A = r2
(5.1)
= 1353 mm2
(5.2)
A = 1353 1200 = 153 mm2
(5.3)
The area difference of 153 mm2 c is considered to consist of the air between the threads in the
conductor. As the outer radius of the conductor is important in determination of the capacitance
this is kept as described in the datasheet. In order to obtain the correct resistance in the
1
The difference between the area calculated from the cross section of 1200 mm2 and the area calculated from
.
the given radius 41,5
2
61
conductor, the area difference is implemented, by defining the conductor as it is hollow. The
radius of the hollow part is found from the above calculated area:
r =
153
= 6, 97 mm
(5.4)
The configuration of conductor resistivity and isolation permittivity, of the cable component is
shown in figure 5.8 and 5.9.
Figure 5.8: Electrical configuration of the cable component, section 1 in PSCAD.
Figure 5.9: Electrical configuration of the cable component, section 1 in PSCAD.
62
The 1st Conducting Layer Data represents the phase conductor, which is aluminium for the
cable.
The Insulator 1 Data represents the isolation material between conductor and screen.
The 2nd Conducting Layer Data represents the metallic screen in the cable, which is made up
of copper wires. The resistivities and permeabilities for copper and aluminium is found in [10,
Serway, p. 959]. The insulation data is given in the user guide for the cable [11, ABB, cable user
guide].
In PSCAD, there are three different types of distributed transmission line models, namely the
Bergeron model, the Frequency dependent (mode) model and the Frequency dependent (phase)
model. These are described in the PSCAD help file [12, PSCAD, p.342], direct quote.
Bergeron model
The Bergeron model represents the L and C elements of a PI section in a
distributed manner (not using lumped parameters like PI sections). It is accurate
at the specified frequency and is suitable for studies where the specified frequency
load-flow is most important (e.g. relay studies).[12, PSCAD, p.342]
Frequency dependent (mode) model
The Frequency-Dependent (Mode) model represents the frequency dependence of
all parameters (not just at the specified frequency as in the Bergeron model). This
model uses modal techniques to solve the line constants and assumes a constant
transformation. It is therefore only accurate for systems of ideally transposed
conductors (or two conductor horizontal configurations) or single conductors.[12,
PSCAD, p.342]
Frequency dependent (phase) model
The Frequency-Dependent (Phase) model also represents the frequency dependence of all parameters as in the Mode model above. However, the Frequency
Dependent (Phase) model circumvents the constant transformation problem by direct formulation in the phase domain. It is therefore accurate for all transmission
configurations, including unbalanced line geometry. The Frequency-Dependent
(Phase) model should always be the model of choice, unless another model is chosen for a specific reason. This model is the most advanced and accurate time
domain line model in the world! [12, PSCAD, p.342]
As the Frequency dependent (phase) model is recommended in the user guide, and fulfills the
simulation demand it is the model of choice for all cable and line sections.
5.1.4
Section 2
Section 2 is represented with the Transmission Line Component and Transmission Line Interface
as show in figure 5.10. The earth resistance is chosen to 3 , as good grounding conditions are
assumed [13, Vrts, p. 50].
63
T
Section2
Section2
Section2
3 [ohm]
3 [ohm]
The line component is configured as shown in figure 5.11. Instead of using a graphical tower, x-y
coordinates are used. The graphical configuration of the Donau mast, is shown in figure 3.11 on
page 17.
Tower: Donau
Conductors: Martin
X (from
Cond. # Connection
Phasing # tower centre)
1
1
10.4 [m]
2
2
7.2 [m]
13 [m]
3
3
Tower Centre 0 [m]

Y
(at tower)
33.5 [m]
24.7 [m]
24.7 [m]
GW. #
1
Ground_Wires: Dorking
Connection
X (from
Y
Phasing # tower centre) (at tower)
4
8.35 [m]
41.6[m]
Ground Resistivity: 100.0 [ohm*m]

Relative Ground Permeability: 1.0
Earth Return Formula: Analytical Approximation
Figure 5.11: Configuration of section 2 in PSCAD.
The electrical configuration for the tower, the phase conductors and the ground conductors are
shown in figure 5.12 and 5.13.
Figure 5.12: Electrical configuration of the overhead line model, section 2 in PSCAD.
64
Figure 5.13: Electrical configuration of the overhead line model, section 2 in PSCAD.
The parameters entered in figure 5.12 and 5.13 are found in section 3.5 on page 17.
5.1.5
Section 3
Section 1 is represented with the Cable Component and Cable Interface as shown in figure 5.14.
C
Section3
C1 Section3
S1
1 [ohm]
1 [ohm]
Section3 C1
S1
C2
C2
S2
1 [ohm]
1 [ohm]
S2
C3
C3
S3
1 [ohm]
1 [ohm]
S3
Section 3 is configured in the same way as section 1. Only the length of the section differs, for
section 3 the length is 21 km. For this cable section both ends are grounded as described in
section 3.4 on page 12.
5.1.6
65
Shunt reactor
The shunt reactor model, is not included in the Master Library in PSCAD. A new component
was therefore constructed, as described in section 4.4.1 on page 45. The graphical layout of the
created reactor component is shown in figure 5.15.
ic
ib
Rc
Rb
Ra
Lc
Lb
La
oc
ob
ia
oa
Figure 5.15: The layout of the reactor component.
The configuration of the shunt reactor is shown in figure 5.16.
Figure 5.16: The electrical configuration of the reactor model.
The parameter values entered in figure 5.16 are found in section 3.6 on page 20.
5.1.7
The modelling of the transformers in the system, is made with 3 Phase 2 Winding Transformer
component. The 150/60 kV transformer is coupled in YNd11, meaning Y- transformer with
30 phase shift. The Y- transformer is configured, with the winding leading the Y winding,
to represent the phase shift. Magnetic saturation is not included in this component. The 150/60
kV transformer representation is shown in figure 5.17.
66
A
B
C
100.0 [MVA]
#1
#2
150 [kV]
B
C
60 [kV]
Figure 5.17: Representation of 150/60 kV transformer in PSCAD.
The parameters, shown in figure 5.18, for the 150/60 kV transformer model, and entered in the
transformer parameter interface, are calculated in section 3.7.1 on page 27.
Figure 5.18: Configuration of the 150/60 kV transformer in PSCAD.
5.1.8
The 60/20 kV transformer is coupled in YNyn0, resulting in no phase shift between the primary
and secondary side. The transformer is represented with the 3 Phase 2 Winding Transformer as
shown in figure 5.19: Magnetic saturation is not included in the model.
67
A
B
C
16.0 [MVA]
#1
#2
60.0 [kV]
20.0 [kV]
B
C
Figure 5.19: Representation of 60/20 kV transformer in PSCAD.
The configuration of the 60/20 kV transformer in PSCAD, is shown in figure 5.20, the values are
calculated in section 3.7.2 on page 28.
Figure 5.20: Configuration of the 60/20 kV transformer in PSCAD.
5.1.9
20/0,4 kV transformer
The 20/0,4 kV transformer is coupled in ZNyn5, meaning ZigZag-Y transformer with 210 phase
shift. A model for this exact transformer type is not given in the Master Library in PSCAD,
which is why the -Y transformer is used. This will though not give the same characteristics,
such as the earth reference of the Zig-Zag. To represent the phase shift, the -Y transformer is
configured, so that the winding leads the Y. The voltage will then obtain a wave form which
can be compared with the measurement. This procedure will give 0 displacement, and hence
a difference phase sequence after switch-on, compared with the measurements2 . It is though
2
The sequence in the measurement would be: phase b, phase c, phase a due to the 240 (30 + 210 ). For the
simulation the resulting phase shift from 150 kV to 0,4 kV will be 0 , hence the same sequence as for the 150 kV
level: phase a, phase b, phase c.
68
chosen not to proceed to obtain the correct model for this transformer. Magnetic saturation is
not included in this component. The 20/0,4 kV transformer is represented with the 3 Phase 2
Winding Transformer as shown in figure 5.21.
A
B
C
0.630 [MVA]
#1
#2
24 [kV]
B
C
0.42 [kV]
Figure 5.21: Representation of 20/0,4 kV transformer in PSCAD.
The configuration of the 20/0,4 kV transformer in PSCAD, is shown in figure 5.22, the entered
values were calculated in section 3.7.3 on page 28.
Figure 5.22: Configuration of the 20/0,4 kV transformer in PSCAD.
The components in the PSCAD simulation and their configuration has now been described. The
validity of the created reactor model, is examined in the next section, where the steady state
behaviour is controlled against calculated parameters.
5.2
Verification of reactor model
The reactor component is created in PSCAD in order to include the mutual inductances. The
reactor component is therefore tested for mutual inductances by applying voltage on phase a,
and measure the induced voltages on phases b and c. The equivalent circuit is shown in figure
5.23, where the resistances have been eliminated.
5.2 Verification of reactor model
69
ia (t)
ua (t)
La
Mab
ub (t)
Lb
+
uc (t)
Mac
Lc
Figure 5.23: Equivalent circuit for the reactor. The phases b and c are not connected to the voltage
source.
As phases b and c are not connected to the voltage source, the voltages in these two phases are:
dia (t)
dt
dia (t)
uc (t) = Mac
dt
ub (t) = Mab
Here
dia (t)
dt
(5.5)
(5.6)
can be found from the voltage equation for phase a:

dia (t)
dt
(5.7)
1
ua (t)
La
(5.8)
ua (t) = La
dia (t)
dt
The expressions for the voltages at the two non-connected phases can then be found:
ua (t)
La
ua (t)
uc (t) = Mac
La
ub (t) = Mab
(5.9)
(5.10)
The peak voltage for the two non-connected phases can then be found, by inserting the peak
voltage for phase a and the values for Mab , Mac and La :
165
2
3
ub,peak = 0, 032
2, 307
= 1, 34 kV
(5.11)
uc,peak = 0, 023
2, 307
= 1, 87 kV
(5.12)
165
2
3
The test setup in PSCAD is shown in figure 5.24 to the left, and the parameters for the reactor
component is shown to the right.
70
Ia
A
B
Ua
R=0
Ub
C
Uc
ic
ib
Rc
Rb
Ra
Lc
Lb
La
oc
ob
ia
oa
Figure 5.24: Steady state verification of the reactor component in PSCAD.
Phase a [kV]
The results from the simulation are shown in figure 5.25 for the three phases.
100
0
100
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.01
0.02
0.03
0.04
0.05
Time [s]
0.06
0.07
0.08
0.09
0.1
Phase b [kV]
4
2
0
2
4
Phase c [kV]
4
2
0
2
4
Figure 5.25: Voltages in the three phases. Note the different scale on the y-axis.
The values for phase b and c are measured at the peak voltage for phase a, to -1,34 kV and -1,87
kV respectively. These are the same as the calculated value, and the mutual inductances in the
reactor component are therefore verified.
A new simulation is performed with the resistance Ra , Rb and Rc set to 1,17 k as stated in
table 3.13 on page 26. The current into the reactor is simulated, as shown to the left in figure
5.24. The current in phase a is shown in figure 5.26.
5.3 Results
71
0.4
0.3
Current [kA]
0.2
0.1
0
0.1
0.2
0.3
0.4
0.02
0.04
0.06
0.08
0.1
Time [s]
0.12
0.14
0.16
0.18
0.2
Figure 5.26: Current flowing into phase a.
The peak current is 188 A, resulting in a rms current of 133 A. This is in accordance with the
rated current for the reactor of 135,8 A. The difference can be explained by the fact that the
rated current is provided for 170 kV, whereas the simulation is performed with 165 kV.
5.3
Results
As the validity of the reactor model has been confirmed against steady state, simulation of the
system is conducted. The general configuration of the simulation project is first described.
5.3.1
General simulation configuration
The general simulation settings are shown in figure 5.27.
Figure 5.27: General configuration of the simulation.
72
Here, the time step is set to 1 s. This is done in order to keep to time step below the travel
time for section 1 of 430 m. The travel time for section is:
=
0, 43
l
=
= 1, 43 s
c
300000
(5.13)
Where:
is the travel time for the cable
l is the length of the cable
c is the propagation time (assumed speed of light)
[s]
[km]
[km/s]
The PSCAD simulation is performed and the results are compared with measurements in the
following. Each figure shows all phases for each voltage level both for simulation and measurements. Comparison is made for the voltage at 150 kV, 20 kV and 0,4 kV. Current waveforms
are compared for the 150 kV level. No comparison is for the 60 kV level, as these measurements
were corrupt.
5.3.2
Comparison of the voltages at 150 kV level
Figure 5.28 shows the results for the voltages at 150 kV level.
Phase A
PSCAD [kV]
200
100
0
100
Measurements [kV]
200
0
0.01
0.02
0.03
0.04
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.05
0.06
0.07
0.08
0.09
0.1
0.05
0.06
0.07
0.08
0.09
0.1
0.05
0.06
0.07
0.08
0.09
0.1
200
100
0
100
200
PSCAD [kV]
Phase B
200
0
Measurements [kV]
200
0
0.01
0.02
0.03
0.04
0.01
0.02
0.03
0.04
200
100
0
100
200
Phase C
PSCAD [kV]
200
0
Measurements [kV]
200
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.01
0.02
0.03
0.04
0.05
Time [s]
0.06
0.07
0.08
0.09
0.1
200
100
0
100
200
Figure 5.28: The simulation and measurement results for the voltage at 150 kV level.
5.3 Results
73
The voltage behaviour on the 150 kV is shown in figure 5.29 for phase a two periods after
switch-on in t = 0.01 s, in order to show the appearance of the high frequencies.
Phase A
PSCAD [kV]
200
100
0
100
Measurements [kV]
200
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
0.005
0.01
0.015
0.02
0.025
Time [s]
0.03
0.035
0.04
0.045
0.05
200
100
0
100
200
Figure 5.29: Voltage behaviour for phase a, shown for two periods after switch-on in t = 0.01 s.
For the voltages at 150 kV level, a comparison of peak voltages are made:
PSCAD
Measurement
Phases [kV]
a
b
c
+199/-159 +141/-140 +146/-154
+159/-134 +154/-133 +146/-162
Table 5.1: A comparison for the peak voltages at 150 kV level.
From figure 5.28 and table 5.3.2, it can be seen that overvoltages reach a value of 199 kV,
which is greatly above the measured value of 159 kV. It is also apparent how the lower
frequency component is less damped in the simulation than in measurements. The high frequency
component seems to be damped at almost the same rate in simulation and measurement.
5.3.3
Comparison of the currents at 150 kV level
Figure 5.30 shows the results for the currents at 150 kV level.
74
Phase A
PSCAD [A]
1000
Measurements [A]
1000
0.1
0.2
0.3
0.4
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0.5
0.6
0.7
0.8
0.9
0.5
0.6
0.7
0.8
0.9
0.5
0.6
0.7
0.8
0.9
1000
1000
Phase B
PSCAD [A]
1000
Measurements [A]
1000
0.1
0.2
0.3
0.4
0.1
0.2
0.3
0.4
1000
1000
Phase C
PSCAD [A]
1000
Measurements [A]
1000
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0.1
0.2
0.3
0.4
0.5
Time [s]
0.6
0.7
0.8
0.9
1000
1000
Figure 5.30: The simulation and measurement results for the currents at 150 kV level.
For the current at 150 kV level, a comparison is made of the highest peaks, and steady state
magnitude.
PSCAD peaks
Steady state magnitude, peak-to-peak
Measurement peaks
Steady state magnitude, peak-to-peak
a
+1474/-1377
14
+1103/-1423
110
Phases [A]
b
+798/-589
16
+1134/-1773
110
c
+493/-896
14
+666/-1195
110
Table 5.2: A comparison of peak currents, and steady state magnitude at 150 kV level.
Comparison of the currents show that, apart from phase a, the initial inrush current reaches
higher peak values in measurements than in simulation, as shown in table 5.3.3. This inrush
period is also much shorter in the measurements. As shown in table 5.3.3, the steady state
magnitude obtain a much higher magnitude than in simulations.
5.3 Results
5.3.4
75
Comparison of voltage at 20 kV level
Figure 5.31 shows the results for voltage at 20 kV level.
Measurements [kV]
PSCAD [kV]
Phase A
20
0
20
0
0.01
0.02
0.03
0.04
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.05
0.06
0.07
0.08
0.09
0.1
0.05
0.06
0.07
0.08
0.09
0.1
0.05
0.06
0.07
0.08
0.09
0.1
20
0
20
Measurements [kV]
PSCAD [kV]
Phase B
20
0
20
0
0.01
0.02
0.03
0.04
0.01
0.02
0.03
0.04
20
0
20
Measurements [kV]
PSCAD [kV]
Phase C
20
0
20
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.01
0.02
0.03
0.04
0.05
Time [s]
0.06
0.07
0.08
0.09
0.1
20
0
20
Figure 5.31: The simulation and measurement results for the voltage at 20 kV level.
For the voltage at 20 kV level, a comparison of peak voltages is made:
PSCAD
Measurement
Phases [kV]
a
b
c
+18,4/-21,1 +18,3/-17,7 +19,8/-26,5
+16,6/-17,4 +17,3/-16,7 +17,3/-17,2
Table 5.3: A comparison for peak voltages at 20 kV level.
The noticeable deviation between simulation and measurements at the 20 kV level, is the damping
rate of the lower frequency component. That applies especially for phases a and c. Furthermore,
the overvoltages reach considerably higher values in the simulation than in measurement. The
higher frequency component experienced in the simulation is not present in the measurements
as these are damped/filtered.
76
5.3.5
Comparison of voltage at 0,4 kV level
Figure 5.32 shows the results for voltage at 0,4 kV level.

Phase A
PSCAD [V]
400
200
0
200
Measurements [V]
400
0.01
0.02
0.03
0.04
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.05
0.06
0.07
0.08
0.09
0.1
0.05
0.06
0.07
0.08
0.09
0.1
0.05
0.06
0.07
0.08
0.09
0.1
400
200
0
200
400
Phase B
PSCAD [V]
400
200
0
200
Measurements [V]
400
0.01
0.02
0.03
0.04
0.01
0.02
0.03
0.04
400
200
0
200
400
Phase C
PSCAD [V]
400
200
0
200
Measurements [V]
400
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.01
0.02
0.03
0.04
0.05
Time [s]
0.06
0.07
0.08
0.09
0.1
400
200
0
200
400
Figure 5.32: The simulation and measurement results for the voltage at 0,4 kV level.
For the voltage at 0,4 kV level, a comparison of peak voltages is made:
PSCAD
Measurement
Phases [V]
a
b
c
+440/-368 +328/-326 +339/-357
+369/-311 +339/-319 +318/-332
Table 5.4: A comparison for peak voltages at 0,4 kV level.
As the case for the 20 kV level, the damping at the 0,4 kV level is too low for the lower
frequency component. The peak voltages reach considerably higher values in the simulation
than in measurement. The higher frequency component is not damped through the 20/0,4 kV
transformer as experienced in measurements.
5.3 Results
5.3.6
77
Harmonic analysis at 150 kV level
A frequency analysis is made for the voltage at the 150 kV level, which represents the voltage
measured in Ferslev. This is compared with a Fourier analysis of the measured voltage.
The harmonic analysis for phase a is shown in figure 5.33.
Phase A
20
Pscad
10
0
0
1000
2000
3000
4000
5000

150 kV level
20
6000
Measurement
10
0
0
1000
2000
3000
Phase B
4000
5000
20
6000
Pscad
10
0
0
1000
2000
3000
4000
5000
20
6000
Measurement
10
0
0
1000
2000
3000
Phase C
4000
5000
20
6000
Pscad
10
0
0
1000
2000
3000
4000
5000
20
6000
Measurement
10
0
0
1000
2000
3000
4000
Frequency [Hz]
5000
6000
Figure 5.33: Results for the Fourier analysis of the voltage vaveform at 150 kV level, 20 ms after
switching.
The most significant results from the frequency analysis of the voltage at 150 kV are discussed.
The two frequency bands are apparent for both simulation and measurement, though with some
deviation. For the simulation, magnitudes of all frequencies above the fundamental are much
higher than for the measurement, indicating yet again lack of damping. In the simulation, the
appearance of the 8th harmonic is the most dominant for the lower frequency band, whereas the
7th harmonic is the dominant in the measurements. The simulation gives both higher magnitude,
and higher frequency for the high frequency band. The measurements give a center frequency of
ca. 3,8 kHz, whereas the PSCAD simulation gives a frequency of ca. 4,2 kHz.
78
A frequency analysis of the voltage at the other voltage levels is not conducted, as non significant
damping is experienced through the transformers, and hence will this yield the same result as
for the 150 kV level.
As seen from the results, both by comparison of voltage waveforms, and from Fourier analysis,
is a lack of damping apparent in the simulation. There is a suspicion that the skin effect may
not be accounted for in the simulation software. Before further conclusions can be made about
the validity of the simulations, an investigation for this is performed.
5.3.7
Test of frequency dependency of the transmission line model in PSCAD
In order to investigate the frequency dependency of the resistance in the transmission line model,
a test is conducted on section 2. An estimation of the expected PSCAD results is needed, in order
to evaluate the results from PSCAD. This is done by calculating the input impedance of the
transmission line for the case of a frequency dependent series resistance and for the case of a
frequency independent series resistance. The equivalent circuit for the transmission line is shown
in figure 5.34, when a equivalent is used. To simplify the analysis, the cable end is short
circuited.
1
2R
ua (t)
1
2R
1
2L
1
2L
Zin
1
2 Rg
1
2 Rg
Figure 5.34: A T equivalent circuit for the transmission line, where the resistance in the ground is
included. The end of the transmission line is short circuited.
The input impedance to the circuit shown in figure 5.34:

Zin
1
1
1
= ( R + Rg + j L) +
2
2
2
1
1
1
1
jC ( 2 R + 2 Rg + j 2 L)
1
1
1
1
jC + ( 2 R + 2 Rg + j 2 L)
(5.14)
This equation is used to calculate the input impedance for both the frequency dependent
resistances and the frequency independent resistances. The resistance in the ground Rg can
be found as a function for the frequency [1, Vrts, p. 116]:
Rg = 0, 00099 f
[/km]
(5.15)
The AC resistance is calculated by using a method presented by Hurley [14, Hurley, p. 6]. Here,
an approximated method for calculating the frequency dependent resistance for a round wire is
presented. First the skin depth is calculated for the given frequency:
=
(5.16)
5.3 Results
79
Where:
is the skin depth
is the resistivity of the conductor
is the permeability of the conductor
The frequency dependent resistance Rac can now be calculated:
Rac = Rdc ks
ks = 1 +
(5.17)
(r/)4
48 + 0, 8(r/)4

r
3
= 0, 25 + 0, 5
32
r
ks
. . . r/ < 1, 7
(5.18)
. . . r/ > 1, 7
(5.19)
Where:
ks is the correction factor between the DC and the frequency dependent
resistance
r is the radius of the conductor
In the case of a radius of 18,1 mm, equation 5.18 is used for frequencies below 65 Hz and equation
5.19 is used for frequencies above 65 Hz.
The input impedances are now calculated, both with and without the frequency dependent
resistance, for frequencies between 1 Hz and 2000 Hz, by using equation 5.143 . The results are
In order to evaluate the frequency dependency of the series resistance in the transmission line
model, simulations are performed as shown in figure 5.35. Here only one phase is used in order
to eliminate induced voltage. The end of the line is grounded through a small resistance, acting
as short circuit. The transmission line is configured as described in section 5.1.4 on page 62.
T
Section2
Ia
R=0 U_an
Section2
Section2
3 [ohm]
3 [ohm]
0.001 [ohm]
Figure 5.35: Setup for evaluation of frequency dependency of the resistance of the transmission line
model in PSCAD.
The simulation is conducted for 15 frequencies in the interval between 50 Hz and 2000 Hz. The
phase angle is measured along with the peak voltage and the peak current. The input impedance
can then be found:
Ua,peak
6
Zin =
(5.20)
Ia,peak
(5.21)
3
This procedure is performed in MatLab as shown in appendix G.11.
80
The real part of the input impedance is now plotted as a function of the frequency. This is done
in figure 5.36, where the simulation results are compared with the calculated results for both the
frequency independent and the frequency dependent series resistance.
50
Calculated with frequency independent series and ground resistance
Calculated with frequency dependent series and ground resistance
Calculated from PSCAD simulation
45
Real part of input impedance ()
40
35
30
25
20
15
10
200
400
600
800
1000
Frequency [Hz]
1200
1400
1600
1800
2000
Figure 5.36: Comparison of real part of the input impedance, between measurements in PSCAD
and calculated values with both frequency dependence and independence of the series resistance of the
transmission line.
In figure 5.36 the real part of the input impedance for the transmission line is found for different
frequencies.
The calculated real part of the input impedance increases with the frequency, as shown by the
dottet line in figure 5.36.
As shown by the slight rise of the solid line in figure 5.36, the real part of the input impedance,
calculated with frequency independence resistances, does rise with the frequency. This rise
is caused by the change in the reactances of the capacitance and inducatance in the parallel
connection.
As shown by the o line in figure 5.36, does the real part of the simulated input impedance
increase with frequency.
The real part of the input impedance measured in PSCAD varies from the expected values, as
shown in the figure. This difference can be caused by different calculation methods, and the fact
that the PSCAD model s more detailed than the simplified model which calculation is conducted
from. From the above-mentioned it can be concluded that the transmission line model do account
for frequency dependence of the resistances. This is also assumed to be the case for the cable
model.
5.4
As shown, there are some differences between the simulation and the measurements.
difference is reviewed in the following.
The
81
Lack of damping
As shown in section 5.3.7, the skin effect is accounted for in the frequency dependent (phase)
model4 . This rules out, that the lack of damping is originated from frequency independency
of the transmission line and cable models. Therefore, the only frequency independent,
resistive component in the PSCAD model, is the real part of the short circuit impedance.
Implementation of frequency dependency in the short circuit resistance would increase the
damping factor of the system. The short circuit power seen from Ferslev is given for
maximum and minimum apparent power, resulting in a maximum and minimum short
circuit impedance. The angle of this impedance is unknown, which can cause difference
between the measurements and the simulation. The angle is set to a typical value of 80
in the simulation, here a decrease in the angle would give a larger short circuit resistance,
and thereby more damping.
Frequencies
The frequency components from the Fourier analysis did indicate difference between the
simulation and the measurements. This difference could stem from the short circuit
impedance, as the short circuit inductance can obtain values varying between 14,4 mH
and 35,6 mH, depending on whether the maximum or minimum is used. Another factor
could be that the capacitance of the cable is too small.
Steady state current
The steady state current in the simulation did show a smaller magnitude than the measured.
This indicates that the simulated system has too high impedance. A possible reason for
this would be a too small capacitance in the cable sections.
Damping through transformers
In the measurements, the higher frequencies are filtered through the transformers. This is
not the case for the simulation, where the high frequencies from the 150 kV level propagate
to the lower voltage levels, only with a little damping.
Due to the difference between the simulation and measurements, an evaluation of the simulation
models is conducted in the next chapter.
This was also confirmed by mail from Dr. Dharshana Muthumuni from PSCAD.
Chapter 6

As shown in the previous chapter, large differences between the simulation and measurements
were present. In this chapter, the simulation is examined in order to minimize simulation errors.
6.1
Estimation of cable and line models
In order to investigate the correctness of the configuration, the parameters given in the cable
datasheet are compared with the parameters calculated by PSCAD. A line constant subprogram
is included in the PSCAD simulation software. This allows an output file to be written, containing
the parameters for the given system. These parameters are given in a matrix form, both in a
sequence and phase form. The output files are shown in appendix F.
6.1.1
Cable shunt admittance
The phase data for the shunt admittance, is a 66 matrix, as the cable has three phases each
containing one phase conductor and one screen conductor.
The shunt admittance for conductor 1 in section 3, is given for the full length of the cable (21
km), in the matrix shown in figure F.2 on page F-55:
Yshunt = 0, 656 106 + j1, 34 103 S
(6.1)
The conductance and capacitance can now be found per km, and compared with values given in
the datasheet. This is done in table 6.1.
Can
Gan
Datasheet [per km]

243 nF
12,6 nS
PSCAD [per km]

202 nF
31,2 nS
Table 6.1: A comparison of the shunt conductance and shunt capacitance of the cable, from values
calculated by PSCAD and calculated from datasheet.
6.1.2
Cable sequence impedance
The sequence impedance for the cable is compared with the values from the datasheet, given in
[/km]:
Z0
Z1
Z2
Datasheet [/km]
0,188+j0,0706
0,0331+j0,108
0,0331+j0,108
PSCAD [/km]
0,226+j1,87
0,0787 +j0,0711
0,0787 +j0,0711
Table 6.2: A comparison of the sequence impedance of the cable, from values calculated by PSCAD
and given in datasheet.
84
From the comparison of the shunt admittance in table 6.1 and sequence impedance in table 6.2
the following can be concluded:
The shunt conductance calculated by PSCAD is 2,47 times the value given in the datasheet.
The shunt capacitance in PSCAD is only 83% of the value given in the datasheet.
The zero sequence resistance in PSCAD is 1,2 times the datasheet value.
The zero sequence reactance in PSCAD is 26,5 times the datasheet value.
The positive sequence resistance in PSCAD is 2,37 times the calculated value.
The positive sequence inductance in the datasheet is 1,51 times the PSCAD value.
A comparison of the zero sequences is difficult to make, as the circumstances it is calculated
for are unknown.
The above-mentioned comparison, verifies that the cable section is not configured to give the
same results, as supplied in the datasheet.
The importance of this is significant, as the capacitance of the simulated cable section, will give
rise to higher frequency components than were measured. Furthermore, the amplitude of the
steady state current will obtain a too low value, as the input impedance will obtain too high
value. A re-evaluation of the cable configuration is needed. In figure 6.1 the configuration of the
cable is shown.
Conductor
Insul tor 1
She th
Insul tor 2
a
a
6.97 [mm]
20.75 [mm]
39.05 [mm]
39.435 [mm]
46 [mm]
Figure 6.1: The physical configuration of one of the cable conductors. The representation to the right,
is shown in connection to equation 6.2
From the distances shown in the figure above, the capacitance is calculated. The capacitance of
a cylindrical conductor is found from [10, Serway, p.808]:
C =
ke =
C =
l
2 ke ln

b
a
[F]
1
= 8.987551787 109
[N m2 /C2 ]
40
1000 2, 3

= 202, 3 nF/km
39,05
2ke ln 20,75
(6.2)
(6.3)
(6.4)
6.1 Estimation of cable and line models
85
Where:
Where ke is the Coulomb constant.
This equation
a is the radius of the first conductor
b is the radius to the second conductor
is given for isolation material that is air. In the case of other dielectric material, the relative
permittivity, in this case 2,3, is multiplied by the capacitance. The length is set to 1000 m, as a
capacitance per/km is desired. This gives identical result to the one found by PSCAD.
A new cable configuration is obviously needed, and this is calculated before inserting in PSCAD The
1,2 mm thick conductor screen, is defined as a part of the conductor, and the dielectric material as
only the isolation material, but not insulation screen and bedding for screen. Thereby is the new
radius a, defined to be 21,95 mm, and radius b, is found by the thickness of the isolation material,
defined in figure 3.5 on page 13 to be 37,95 mm. Thereby is a new capacitance calculated:
C =
1000 2, 3
2ke ln
37,95
21,95
(6.5)
= 233, 7 nF
As the conductor cross section has been increased, the hollow center is increased accordingly to
keep the conductor areal 1200 mm2 .
Aconductor = 21, 952 Acenter = 1200 mm2
2
Acenter = 21, 95 1200 = 313.6 mm

rcenter =
Acenter
= 9, 99 mm
(6.6)
(6.7)
(6.8)
The configuration of the corrected cable is shown in figure 6.2
Conductor
Insulator 1
Sheath
Insulator 2
9.99 [mm]
21.35 [mm]
37.95 [mm]
38.89 [mm]
46 [mm]
Figure 6.2: The configuration of the cable, with corrected distances.
After this configuration, the parameter values are solved for again in PSCAD, the values are listed
in table 6.3 and table 6.4.
Can
Gan
Datasheet [per km]

243 nF
12,6 nS
PSCAD [per km]

222 nF
21,4 nS
Table 6.3: A comparison of the shunt conductance and shunt capacitance for the corrected cable, from
values calculated by PSCAD and calculated from datasheet.
86
The new sequence impedance for the cable is compared with the values from the datasheet, given
in [/km]:
Datasheet [/km]
0,188+j0,0706
0,0331+j0,108
0,0331+j0,108
Z0
Z1
Z2
PSCAD [/km]
0,191 + j1,86
0,0434 + j0,0682
0,0434 + j0,0682
Table 6.4: A comparison of the newly found sequence impedance of the cable, from values calculated
by PSCAD and given in datasheet.
As the capacitance has yet not obtained completely the same value as given in the datasheet,
the permittivity of the isolation material is increased. By increasing the permittivity of the
isolation material, the correct value for the capacitance is obtained. In section 6.3, a simulation
is performed with this configuration. With this correction of the cable, the positive sequence
resistance has obtained a value more comparable to the datasheet value. Other parameters, given
in table 6.4 are accepted as adequately accurate. The matrices describing the parameters for the
corrected cable are given in appendix F.3 on page F-56.
6.1.3
Line parameters
The parameters for the overhead line section, calculated in PSCAD are compared with the
calculated parameters from appendix A.
Line shunt admittance
The phase data for the shunt admittance, is a 44 matrix, as the overhead line section consist
of three phase conductors and one ground conductor. The shunt admittance for conductor 1 is
given for the full length of section 2, 18,93 km, shown in the matrix in figure F.1 on page F-54:
Yshunt = 2, 5 + j46, 1 S
(6.9)
The conductance and capacitance can now be compared with the values calculated in appendix
A:
Can
Gan
Calculated [per/km]
7,87 nF
132 nS
PSCAD [per/km]
7,74 nF
132 nS
Table 6.5: A comparison of the shunt admittance of the overhead line, from values calculated by PSCAD
and values calculated in appendix A.
6.2 Estimation of models for shunt reactor and short circuit impedance
87
Line series impedance

The sequence impedance for the line in PSCAD is compared with the calculated values in table
6.1.3.
Z0
Z1
Z2
Calculated [/km]
0,175 + j0,983
0,0477 + j0,395
0,0477 + j0,395
PSCAD [/km]
0,186 + j1,32
0,0439 +j0,397
0,0439 +j0,397
Table 6.6: A comparison of the sequence impedance of the overhead lines, from values calculated by
PSCAD and values calculated in appendix A.
From the comparison of the shunt admittance and series impedance the following can be
concluded:
The shunt capacitance and shunt conductance are approximately the same for PSCAD and
calculations.
The positive sequence impedance is approximately the same for PSCAD and calculations.
An 8,6% difference in the resistive part is though present.
The zero sequence reactance calculated by PSCAD is 34% to high compared to calculated
value.
The values from PSCAD are in good accordance with the calculated values. The PSCAD model is
therefore considered to be correct configured.
In the next section, a discussion of the validity of the short circuit impedance and reactor model
is performed.
6.2
Estimation of models for shunt reactor and short circuit

impedance
An estimation of the factors which can affect the simulation, are described in the following, for
the short circuit impedance and the reactor model.
6.2.1
Short circuit impedance
As described in the section 3.2 on page 9, the short circuit impedance was estimated to 4, 66 80 .
As this is an estimate, the actual short circuit impedance can vary from the estimate due to the
following reasons:
88

The short circuit impedance is calculated for maximum and minimum apparent power
In section 3.2 on page 9, the short circuit impedance was calculated from both the maximum
and minimum short circuit power. At the time of the measurement the short circuit
impedance is unknown. As the data for the maximum short circuit power are used in the
simulation, this could give rise to results that are different to those measured. If the short
circuit impedance is calculated with a value closer to the minimum short circuit power,
the inductance will obtain a larger value. If this is the case, the frequencies obtained in
simulation will be affected.
The angle for the short circuit impedance is unknown
In continuation of the previous item, the angle of the short circuit impedance is unknown.
This will affect the short circuit impedance in such way that a larger angle will decrease
the resistance/increase the inductance, and vice versa, a smaller angle will increase the
resistance/decrease the inductance. The angle of the short circuit impedance is an
important factor, as the short circuit resistance will affect the damping for the whole
system, and the inductance has influences on the generated frequencies.
The real part of the short circuit impedance is frequency independent
The short circuit impedance used in the simulation consists of a series connection of a
resistor and an inductor. The resistance represents the replacement resistance of the current
path between Ferslev and the power plants. As this path will consist of a combination
of transmission lines and bus bars, the resistance will be frequency dependent. The
frequency dependency of the short circuit resistance can be difficult to determine. In a
single transmission line consisting of round wires, the skin dept at different frequencies can
be calculated, and thereby is it possible to find the frequency dependent resistance. This
is not the case for the short circuit impedance, as the path from the power plants consist
of multiple transmission lines and bus bars of different sizes.
6.2.2
Reactor model
The reactor model used in the simulation is a simple model which only includes the losses in the
core and windings, and the self and mutual inductances. The possible impacts caused by the
simplified model are described in the following.
Saturation of core is not included

When the grid is switched-on, large inrush currents will be experienced in the system.
Though these currents will mostly charge the cable capacitances, a side effect of these
transient is that the currents obtain a slowly decreasing DC-component. As this DC
component will cause a larger current in the reactor and hence high flux density, saturation
can occur. In order to check the inrush current, a simulation is performed, resulting in the
current, shown in figure 6.3. This simulation is performed precisely in the same manner
as the simulation described in chapter 5. The only difference is the measurement of the
current that flows into the reactor at the switch-on moment measured.
6.2 Estimation of models for shunt reactor and short circuit impedance
89
Current [kA]
0.5
0.5
0.05
0.1
0.15
Time [s]
Figure 6.3: The initial current sequence in the reactor.
The magnetic flux density can be calculated from F = N I = R. Knowing = BA and

2
R = NL , an expression for B is given:
B =
Li
N A
(6.10)
The reactor is described in appendix B, where the number of windings is 1194, the
inductance for phase a is 2,307 H and the cross section of the air gap in the limb is
0,2827 m2 . The maximum current in the reactor has a magnitude of 341 A, from which
the peak magnetic flux density is calculated:
Bpeak =
Bpeak =
L Ipeak
N A
2, 307 H 341 A
= 2, 330 T
1194 0, 2827m 2
(6.11)
(6.12)
The saturation limit for the reactor is defined to 2,03 T [15]. It is therefore apparent, that
the simple reactor model is not sufficient in the start sequence, where the reactor will enter
saturation. When the core goes into saturation, the permeability will approach the one of
air, resulting in a high reluctance, and thereby a small inductance, thereby causing larger
H
~ d~s = N I), can it be seen that when
currents to be present. From Amperes law ( H
the core enters saturation, non-linearity occurs due to hysteresis, and hence will harmonic
currents be present.
Remanence
Remanence occurs when the magnetization is removed, as some flux density will remain.
As the focus is on the switch-on, and the system assumed totally discharged, it is assumed
that there will be no remanence present. Furthermore, the air gaps in the phase limbs of
the core will decrease the appearance of remanence due to its large reluctances.
As the short circuit impedance can not be estimated correctly, the value used earlier is unchanged.
The estimations above for short circuit impedance and reactor are deemed to give minimal impact
on the frequencies in simulations, compared to the significance of the size of cable capacitance.
In the next section is the system therefore simulated with the cable sections configured to give
the capacitance given in the datasheet.
90
6.3
Correction of cable capacitance
In this section, the system is simulated with the corrected cable section. The configuration of the
cable has been corrected as to obtain the right capacitance, as described earlier, given the value
of 0,243 F/km in the datasheet. The simulation is conducted in precisely the same manner
as the simulation described in chapter 5. Only voltages and currents at 150 kV level are shown
here, as is the Fourier analysis.
The shunt admittance and sequence impedance of the corrected cable component, is found by
reading the line constant program output file, the matrices for the corrected cable are given in
appendix F.3:
Yshunt
Z1
Corrected Cable Parameters [per km]

21,4109 + j74, 9 106 S/km
43,4103 + j68, 3 103 /km
Table 6.7: The parameter values for the corrected cable component.
6.3.1
The voltage waveform
The simulated voltage waveforms, that represent the voltage measured in Ferslev, are shown in
figure 6.4, for the corrected cable component.
Voltage [kV]
200
100
0
100
200
0.005
0.01
0.015
0.02
0.025
0.03
0.035
Time [s]
0.04
0.045
0.05
0.055
Figure 6.4: The simulated voltage waveform, with corrected cable component
The peak values for the recorded voltages are given in table 6.8
Phases
Corrected cable
Earlier simulation
Measurement
Peak voltages [kV]

a
b
c
+193/-155 +144/-138 +141/-150
+199/-159 +141/-140 +146/-154
+159/-134 +154/-133 +146/-162
Table 6.8: A comparison for peak voltages at 150 kV level, simulated with the corrected cable component.
Results of this simulation show that the overvoltages for phase a are still out of proportion,
while the other phases are more like the measured overvoltages. Lack of damping of the lower
harmonic is still present.
6.3 Correction of cable capacitance
6.3.2
91
The current waveform
The initial current sequence, that represents the current that was measured in Ferslev at 150 kV
level is shown in figure 6.5, simulated with the corrected cable component.
1.5
Current [kA]
1
0.5
0
0.5
1
1.5
0.005
0.01
0.015
0.02
0.025
0.03
0.035
Time [s]
0.04
0.045
0.05
0.055
Figure 6.5: The initial current sequence, with the corrected cable component.
The current waveform is recorded over 3 seconds, to show how the change in capacitance
influences the steady state current magnitude. The peak and steady state currents are compared
in table 6.6, with the measurements and the earlier simulations where the cable capacitance was
deemed too small.
1.5
Current [kA]
1
0.5
0
0.5
1
1.5
0.5
1.5
Time [s]
2.5
Figure 6.6: The current waveform simulated over 3 s. The steady state magnitude is here 90 A, peak
to peak.
Corrected cable
Earlier simulation
Measurement
Corrected cable
Earlier simulation
Measurement
Peak currents [A]

Phase a
Phase b
Phase c
+1564/-1502
+893/-615
+558/-992
+1474/-1377
+798/-589
+493/-896
+1103/-1423 +1134/-1773
+666/-1195
Steady state magnitude, peak-to-peak [A]
90
90
90
14
16
14
110
110
110
Table 6.9: A comparison of peak currents, and steady state magnitude at 150 kV level.
Here is the steady state current obviously more comparable to the measured. Initial peaks have
risen considerably.
92
6.3.3
Fourier analysis
Percent of the fundamental [%]
A Fourier analysis of the frequency components was conducted for the simulation. This is shown
in figure 6.7.
30
20
10
0
0
1000
2000
3000
4000
Frequency [hz]
5000
6000
Figure 6.7: A Fourier analysis for one period after switching.
The Fourier analysis shows that the corrected cable component has lowered the center frequency
of the lower frequency band. The center frequency of the lower frequency band is now 350 Hz,
which is identical to the measured. The magnitude of the this harmonic component, of 26 % of
the fundamental frequency, is though still significantly higher than in measurements, where the
magnitude of the 7th harmonic was ca. 5% of the fundmental. For the high frequency component
is the magnitude ca. 8% of the fundamental, while in measurement it was ca. 3%.
6.3.4
Summary of simulation with corrected cable capacitance
The most significant results, obtained by correcting the capacitance of the cable, is summarized.
The initial peak currents have risen considerably in for all phases in simulation with the
newly configured cable.
The steady state currents are now comparable to the measured steady state current.
The Fourier analysis indicates the 7th harmonic as the dominant one, as in the measurements, though with larger magnitude. In measurements this component had ca. 5% of the
fundamental frequency.
Lack of damping is still apparent.
The damping of the lower harmonic component was not increased by correcting the cable. The
resistance needed to obtain the desired damping rate, is found in the next section.
6.4 Estimation of missing resistance
6.4
93
Estimation of missing resistance
This estimation is made with the same simulation configuration as before. By increasing the
source resistance, until the voltage waveforms for the simulation and measurements, are close to
identical, the missing damping is found.
By tuning the source resistance to 20 , the desired damping rate is obtained. This is shown in
figure 6.8.
Simulated
Voltage [kV]
200
100
0
100
200
0.005
0.01
0.015
0.02
0.025
0.03
0.02
0.025
0.03
Measured
Voltage [kV]
200
100
0
100
200
0.005
0.01
0.015
Time [s]
Figure 6.8: The voltage waveforms, with a 20 resistance as a source resistance.
Estimation of the voltage waveforms shows how the desired damping rate is obtained for
two phases, whereas phase b is to much damped. A slight difference is apparent between
measurements and simulation, as the phases appear to experience different damping.
94
Current [kV]
Simulated
1000
0
1000
0.5
1.5
2.5
2.5
Current [kV]
Measured
1000
0
1000
0.5
1.5
Current [kV]
Measured
100
0
100
2.72
2.73
2.74
2.75
Time [s]
2.76
2.77
2.78
Figure 6.9: The current waveforms over 100 ms, with the 20 source resistance
The high amplitude current spikes are present for approximately the same amount of time. There
is though a significant difference in that the simulated system is seemingly free of transients, 20
ms after the switch-on instant. An interesting phenomena is seen, when closer look is taken at
the measured steady state current, shown in the bottom diagram in figure 6.9. A significant
DC-component is still apparent, which might stem from error in measurement, and the current
waveforms are highly distorted, indicating non-linearities of some kind in the system.
30
Percent of the fundamental [%]
25
20
15
10
0
0
1000
2000
3000
4000
Frequency [hz]
5000
6000
Figure 6.10: A Fourier analysis of the frequency components when the system contains the corrected
cable, and a 20 source resistance.
The Fourier analysis of the harmonic currents is now comparable to the Fourier analysis of the
95
measurements, shown in appendix C.2.5 on page C-39. In the simulation has the 7th harmonic a
magnitude of 8%, whereas in the measurements this was ca. 5% of the fundamental frequency.
The 77th harmonic in simulation has a magnitude of 3,5% and the 76th in measurements has a
magnitude of approximately 3%.
6.5
The cable section was evaluated by comparing the calculated values from PSCAD and the provided
data for the cable. Here, differences in the values were found, both for the capacitance and the
series impedance. As the capacitance in the cable is an important component, this was changed
to its correct value, first by configuring the cable component differently, and then by giving
new value to the permittivity of the isolation material. A new simulation with the corrected
capacitance was conducted, indicating more comparable results in relation to steady state current
magnitude and harmonic frequencies.
Both the magnitude of the short circuit impedance and its angle are estimated values, as the
actual value at the time of measurement was not provided. Ideally, a simulation of the system
would incorporate the correct short circuit impedance, and a more advanced model of the shunt
reactor, where saturation is accounted for.
A new simulation where other parameters are varied has been decided against. It is though
acknowledged, that an increased resistance, could be achieved by changing the short circuit
impedance, and by implementing a frequency dependent short circuit resistance. The increased
damping, obtained with such simulation is assumed too low. By replacing the source resistance
with a 20 resistance, the sufficient damping is obtained, and waveforms are obtained that bear
resemblance to the measured waveforms. As this resistance is out of proportion for this type of
system, and that difference between phases is apparent, it is assumed that the damping is not
only to be found in the form of resistivity, but also in the simplified model for the reactor.
Closer look at the measured current waveform in assumed steady state, show how much the
current is distorted. This indicates non-linearity in the system, which is not accounted for in
the constructed model. As for this, a simulation that has identical waveforms compared with
measurements will not be possible to obtain.
In the next chapter, proposals for minimizing the switching overvoltages are made. These
proposals are supported by simulations of the system, where the above described cable component
is used.
Chapter 7
Proposal for minimizing switching

overvoltages
As shown both through measurements and in simulation, switching overvoltages do occur. In
the following chapter, the problem with switching overvoltages is reviewed, and a proposal for
minimization of such overvoltages is given.
7.1
Evaluation of switching overvoltages at the 150 kV level
Through measurements and simulations, a generation of overvoltages on the 150 kV level has
been shown. The voltages from simulations and measurements are compared with the insulation
level for 170 kV in table 7.1.
Phase
a
b
c
Measured [% of prior voltage]

159 kV [18,7%]
154 kV [14,9%]
-162 kV [20,9%]
Simulated [% of prior voltage]

199 kV [23,0 %]
141 kV [0,8 %]
-154 kV [4,2 %]
BIL
550-750 kV
550-750 kV
550-750 kV
Table 7.1: A comparison of simulated and measured magnitude of the overvoltages, with the insulation
level for 170 kV given in [16, Kuffel, p. 493].
As shown in table 7.1, the overvoltages stay below the insulation level for 170 kV by a large
margin. It can therefore be concluded that the switching overvoltages are not likely to cause
faults in the system.
Even though the switching overvoltages stay below the isolation level by a large margin, such
overvoltages are still unwanted, as the equipments are subjected to stress load. This stress load
can cause a decrease in the lifetime of the insulation material, and thereby lower the lifetime for
the equipments. Furthermore, the switching overvoltages will propagated to lower voltage levels.
This issue is evaluated in the following section.
7.2
Evaluation of propagation of switching overvoltages
The voltages, on the 0,4 kV level, from measurements and simulations are listed in table 7.2.
Phase
a
b
c
Measured [% of prior voltage]

382 V [18,4%]
339 V [7,2%]
-332 V [4,5%]
Simulated [% of prior voltage]

440 V [23,0 %]
328 V [0,8 %]
-357 V [8,8 %]
Table 7.2: Maximum overvoltages at the 0,4 kV level shown in measurements

and simulations. It should
be noted that the measured prior voltage of phase b and c, is not 230 2 V, for more details see appendix
C.3.
98
From the table it can be seen that the overvoltages from the 150 kV level do propagated to lower
voltage levels. It can though be concluded that the magnitude of these overvoltages are not very
high.
The variation in the supply voltage is defined in the standards [17, DS/EN 50160] and quoted
below:
2.3 Supply voltage variations

Under normal operating conditions, excluding situations arising from faults or voltage
interruptions,
during each period of one week 95 % of the 10 min mean rms values of the supply voltage
shall be within the range of Un 10%.
NOTE 1: Until the year 2008 the range of voltage may differ from these standard values, in accordance with HD 472 S1.
all 10 minutes mean rms values of the supply voltage shall be within the range of Un
+10%/-15%.
NOTE 2: In cases of electricity supplies in remote areas with long lines the voltage could be outside the range of Un
+10%/-15%. Costumers should be informed.
As this is defined for mean values for a period of 10 minutes, it is considered that the propogated
overvoltages will have minimal influence on the power quality, as they are present for only a few
hundreds of milliseconds or less.
From the above-mentioned, the propagation of switching overvoltages is deemed of little
significance. Though there are no problems with switching overvoltages in the system FER3TH3, problems can occur in other systems. The transmission system from Horns Rev 2 contain
a series connection of two cables with a total length of 72 km. In such system, the size and the
propagation of switching overvoltages can become an issue. A proposal for minimization of
switching overvoltages is therefore examined in the following section.
7.3
Proposal for minimization of switching overvoltages
Switching overvoltages can be reduced in different manners. Two different principles are described
in the following, and afterwards simulated in order to evaluate the effect of such implementations.
7.3.1
Zero crossing closure
In this principle, each phase of the circuit breaker closes at the zero crossing of the voltage, as
7.3 Proposal for minimization of switching overvoltages
99
Switch-on instant
400
300
Phase a [kV,kA]
200
100
0
100
200
300
400
0.005
0.01
0.015
0.02
Time [s]
0.025
0.03
0.035
0.04
Figure 7.1: Principle of zero crossing closure.
This principle will limit the transient voltages, as none of the phases experience instantaneous
change in voltage. Instead, the transition follow the 50 Hz waveform, and thereby a more slowly
charging of the capacitances in the system. The downside of this, is that a large asymmetry in
the current will occur, as two phases will obtain the same DC component, and the last phase
will obtain a opposite DC component. This will cause a considerable zero sequence current at
the moment of switch-on.
7.3.2
Pre-insertion resistors in circuit breakers
In this principle, a resistance is inserted in series with the circuit breaker in order to increase the
initial resistance of the system and thereby limit the inrush current. The resistor is only present
for a short duration, typically for 5-15 ms [18, Beanland], or until the system is fully charged,
determined by the system time constant. The size of the resistor is typically 300 - 500 , but can
vary for different cases [19, Ryan, p. 52]. At the time the resistor is short circuited in the setup,
transients will occur as the system is not totally charged at this time, due to the voltage drop
over the resistor. These transients are though small compared with those experienced without
pre-inserting resistors.
7.3.3
Simulation of zero crossing closure
The main simulation for the zero crossing closure is configured as described in chapter 5. As
described in section 6.3 on page 90 the capacitance is corrected to the value provided from the
datasheet. This correction is included in the following simulation. The setup for the simulation
is shown in figure 7.2.
100
C
Section1
R=0
0.8 [ohm] 14.4 [mH]

0.8 [ohm]14.4 [mH]
I150k_a
BRK_B
I150k_b
BRK_C
I150k_c
S1
1 [ohm]
C1 Section3
S1
1
C2
C2
S2
Section2
Section2
1 [ohm]
1 [ohm]
Section3C1
S1
U_a
S2
C3
U_b
C3
S3
3 [ohm]
3 [ohm]
S3
1 [ohm]
1 [ohm]
C2
S2
C3
S3
1 [ohm]
C2
S2
1 [ohm]
C3
0.8 [ohm]14.4 [mH]
C
Section3
Section1 C1
S1
C1 Section1
BRK_A
T
Section2
1 [ohm]
S3
____1 [ohm]
U_c
ic
Rc
Lc
Rb
80.0 [MVA]
16.0 [MVA]
U60k_a
B
#1
#2
U150k_b
150 [kV]
C
60 [kV]
#1
#2
0.630 [MVA]
U04k_a
B
#1
60.0 [kV]
20.0 [kV]
C
U20k_c
oa
#2
24 [kV]
1 [ohm]
U20k_b
C
U60k_c
A
U20k_a
B
U60k_b
C
U150k_c
La
ob
U150k_a
ia
Ra
Lb
oc
ib
U04k_b
1[ohm]
U04k_c
1 [ohm]
C
0.42 [kV]
Figure 7.2: Simulation setup for the zero crossing closure.
To represent the circuit breakers, three individual breakers are used (BRK_A, BRK_B and BRK_C
in figure 7.2), each with the same configuration of resistances as described in 5.1.2 on page 58.
Each breaker is configured to close at the zero crossing of each phase resulting in the following
times, when phase a crosses zero at t = 0, 015 s, as shown in figure 7.1:
ta,zero = 0, 015 s
(7.1)
Tperiod
0, 020
= 0, 015 + 2
= 0, 02167 s
6
6
Tperiod
0, 020
= ta,zero +
= 0, 015 +
= 0, 01833 s
6
6
tb,zero = ta,zero + 2
(7.2)
tc,zero
(7.3)
As shown, the sequence of the connection is: phase a, phase c, phase b. This sequence is chosen
to lower the asymmetry of the currents after switch on. As the chosen sequence will force the
DC component for phase c to be opposite of the other phases.
The three breakers are configured as shown in figure 7.3.
BRK_A
Timed
Breaker
Logic
Open@t0
BRK_B
Timed
Breaker
Logic
Open@t0
BRK_C
Figure 7.3: Configuration of the timer for the three breakers.
Timed
Breaker
Logic
Open@t0
101
Simulation results
The voltage behaviour is shown in figure 7.4 for zero crossing closure, where the switch-on time
for the three phases is given in equation 7.1, 7.2 and 7.3.
Phase a [kV]
200
100
0
100
200
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
0.005
0.01
0.015
0.02
0.025
Time [s]
0.03
0.035
0.04
0.045
0.05
Phase b [kV]
200
100
0
100
200
Phase c [kV]
200
100
0
100
200
Figure 7.4: Voltage on the three phases with the zero crossing closure principle.
The simulation of the zero crossing closure shows significant decrease of switching overvoltages
on the 150 kV level. The peak voltages for the simulation with and without the zero crossing
closure are shown in table 7.3.
Phase
a
b
c
Without zero crossing

199 kV
141 kV
-154 kV
With zero crossing

-140 kV
-137 kV
140 kV
Table 7.3: A comparison of simulated overvoltages, with and without zero crossing switching.
The current behaviour is shown in figure 7.5 for the zero crossing closure.
102
Phase a [kV]
0.5
0.5
0.5
1.5
0.5
1.5
0.5
1.5
Phase b [kV]
0.5
0.5
Phase c [kV]
0.5
0.5
Time [s]
Figure 7.5: Current in the three phases with the zero crossing closure principle.
All the currents in figure 7.5 have a DC component. Here two of the DC components are negative
and one is positive. This will result in a negative DC component, flowing as zero sequence current.
The peak currents from simulation with and without zero crossing closure are listed in table 7.4.
Phase
a
b
c

1474 A
798 A
-896 A
With zero crossing

-405 A
-405 A
402 A
Table 7.4: A comparison of simulated peak currents, with and without zero crossing switching.
As shown are the inrush currents greatly reduced by use of this principle.
7.3.4
Simulation of pre-insertion resistor
The main simulation is configured as described in chapter 5. Like in the previous simulation,
the corrected capacitance1 is used in the simulations. The setup of the simulation is shown in
figure 7.6.
1
As described in section 6.3 on page 90.
C
Section1
C1 Section1
BRK1
A
R=0
0.8 [ohm] 14.4 [mH]
0.8 [ohm]14.4 [mH]
I150k_a
S1
1 [ohm]
Section1
S1
BRK
C1
C1
Section2
C3
1 [ohm]
1 [ohm]
Section3 C1
S1
U_a
S2
C3
3 [ohm]
S3
1 [ohm]
Section2
3 [ohm]
S3
1 [ohm]
C2
S2
BRK
1 [ohm]
C2
S2
1 [ohm]
C3
400.0 [ohm]
Section3
S1
C2
S2
I150k_c
C
C
Section3
C2
I150k_b
0.8 [ohm]14.4 [mH]
T
Section2
400.0 [ohm]
103
U_b
C3
S3
1 [ohm]
1 [ohm]
S3
U_c
Rc
BRK1
Timed
Breaker
Logic
Open@t0
Lc
ic
Rb
ia
Ra
La
Lb
400.0 [ohm]
oc
BRK
ib
ob
oa
BRK
Timed
Breaker
Logic
Open@t0
80.0 [MVA]
U150k_a
16.0 [MVA]
U60k_a
B
#1
#2
U150k_b
#1
#2
U60k_b
C
150 [kV]
C
60 [kV]
U150k_c
0.630 [MVA]
U04k_a
B
#1
#2
60.0 [kV]
20.0 [kV]
24 [kV]
U04k_b
1[ohm]
U04k_c
1 [ohm]
C
0.42 [kV]
U20k_c
1 [ohm]
U20k_b
C
U60k_c
A
U20k_a
Figure 7.6: Simulation setup for the pre-insertion resistor.
The size of the pre-insertion resistor can be determined from demands for the time constant or
the maximum inrush current. The time constant for the system FER3-TH3 can be calculated
for the typically values of the pre-insertion resistor 300-500 , when neglecting other capacitances
than the one of section 3 with a vaule of 5,1 F:
300 = R C = 300 5, 1 106 = 1, 53 ms
6
500 = R C = 500 5, 1 10
= 2, 55 ms
(7.4)
(7.5)
As shown, both time constants are below the 10 ms which the resistor is configured to stay in the
circuit. The choice of the resistor is now a compromise between limitation of the inrush current
and the voltage drop across the pre-insertion resistor. Choosing the largest resistor will give a
better limitation of the inrush current. However, the voltage drop across the pre-insertion resistor
will increase, resulting in larger transients when the resistor is short circuited. This will occur as
the voltage across the capacitance will change when the pre-insertion resistor is short circuited.
From these two considerations a compromise has been made, selecting the pre-insertion resistor
to 400 , resulting in a time constant of 2,04 ms.
For each resistance one breaker is implemented in parallel, as shown in figure 7.7.
BRK1
A
B
400.0 [ohm]
BRK
C
400.0 [ohm]
BRK
400.0 [ohm]
BRK
Figure 7.7: Implementation of pre-insertion resistor.
104
Three single phase breakers are used to short circuit the charging resistance 10 ms after switchon. This is done with the same timer module for all three phases (BRK). The two timer modules
for the breakers (BRK1 and BRK) are configured as shown in figure 7.8.
Timed
Breaker
Logic
Open@t0
BRK1
BRK
Timed
Breaker
Logic
Open@t0
Figure 7.8: Configuration of the two different timer modules.
Simulation results
The voltage behaviour is shown in figure 7.9
Phase a [kV]
200
100
0
100
200
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
0.005
0.01
0.015
0.02
0.025
Time [s]
0.03
0.035
0.04
0.045
0.05
Phase b [kV]
200
100
0
100
200
Phase c [kV]
200
100
0
100
200
Figure 7.9: Voltage on the three phases with the pre-insertion resistor principle.
The transient overvoltages are limited significantly, compared to earlier simulations, as shown in
figure 7.9. As shown, transient voltages occur again when the resistor is short circuited a t = 0.2.
The peak voltages for simulation with and without the pre-insertion resistor are listed in table
7.5
Phase
a
b
c
105

199 kV
141 kV
-154 kV
With zero crossing

-141 kV
135 kV
135 kV
Table 7.5: A comparison of simulated peak currents, with and without pre insertion of resistors. Shows
no overvoltages.
The switching overvoltages are limited drastically by use of the pre-insertion resistor. For phases
b and c, no overvoltages are present.
The current behaviour is shown in figure 7.10
Phase a [kV]
0.5
0.5
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.02
0.04
0.06
0.08
0.1
Time [s]
0.12
0.14
0.16
0.18
0.2
Phase b [kV]
0.5
0.5
Phase c [kV]
0.5
0.5
Figure 7.10: Current in the three phases with the pre-insertion resistor principle.
The peak currents for simulation with and without the pre-insertion resistor are listed in table
7.6
Phase
a
b
c

1474 A
798 A
-896 A
With zero crossing

-336 A
146 A
-185 A
Table 7.6: A comparison of simulated overvoltages, with and without pre-insertion of resistors.
From table 7.6, it can be seen that resistor pre-insertion, limits both the transient peak currents,
and the DC component of the current.
7.4
As shown in the simulations, both the principle with zero crossing closure and the principle with
pre-insertion resistor, shows large improvements of the switching overvoltages. With the results
106
from the simulation it has been concluded that the pre-insertion principle is the most favorable.
In comparison with the zero crossing closure, only slight differences in overvoltages are present.
The currents experienced in the resistor pre-insertion principle, are symmetrical with only a
small DC component. Whereas, the zero crossing closure has larger DC components, resulting
in a large zero sequence current flowing in the initial state.
Though the pre-insertion resistor principle is the most favorable when focus is on the electrical
behaviour of the system, the zero crossing closure is generally the one of choice. This is based
on the implementation of the two principles. The zero crossing closure is considered to be
relatively easy to implement compared with the pre-insertion resistor. The zero crossing closure
can be implemented with a basic circuit breaker, and an implementation of a control algorithm
to operate the breakers correctly. This principle will therefore not add mechanical parts to the
system. The pre-insertion resistor requires new implementation of equipment, which will increase
the cost and complexity of this solution.
In the system FER3-TH3, it was concluded that improvements were not needed. The abovementioned proposals are therefore only to be implemented in a system where it is deemed
necessary.
Chapter 8
Conclusion
In the following chapter, a summary of the report is conducted, and the most essential conclusions
extracted. The chapter ends with a discussion of the experienced problems and proposals for
future works.
8.1
Conclusion
This project deals with the generated overvoltages in the 150 kV transmission system between
Ferslev-Tinghj, which is a combination of underground cable and overhead line sections. The
main objectives for this project are listed in the problem statement, section 2.1 on page 6, and
discussed in the following.
Arrange a three phase representation of the 150 kV transmission system between FerslevTinghj, valid for the switch-on in Ferslev.
In chapter three, each component in the system was described. From these descriptions,
electrical models and their respective parameter values were generated for the respective
subsystems, ending up with a three phase representation of the transmission system
between Ferslev-Tinghj.
Explain the experienced transients from a theoretical analysis viewpoint.
From the three phase representation, an examination of the switching transients was
conducted. Here the frequencies seen in the measurements were explained from a theoretical
viewpoint. The frequencies calculated from the theory were slightly different compared with
the measured frequencies. This difference was explained with simplifications made in the
theoretical calculations.
Arrange a simplified dynamic model for the 150 kV system between Ferslev-Tinghj
In order to ease the analysis of the system, a simplified version of the three phase model
was produced, by neglecting components deemed of little significance. This simplification
resulted in a simple electric circuit for analysis. As a part of the simplification, the short
cable section in the grid was neglected.
Arrange a mathematical solution for the simplified model in MatLab.
The differential equations describing the simplified model, were solved by two different
methods. First, an attempt was made to solve these analytically, where it appeared that
the solution was to extensive to comprehend, resulting in a numerical solution. Another
solution method was implemented by arranging a numeric solver in Simulink.
Verify the MatLab model through measurements and simulation in PSCAD.
The two solutions obtained, were verified with simulations of the simplified model in PSCAD.
This gave identical results to the numeric solver. A slight difference was seen, compared to
the analytical approach, explained by the fact that the mutual inductances in the reactor
model were neglected there. The two solution methods were therefore considered correct.
However, a comparison with the measurements showed lack of damping in all solutions.
108
Conclusion
Arrange a simulation of the model in PSCAD.
In order to evaluate the lack of damping in the simplified model, a simulation of the
transmission system between Ferslev-Tinghj, was made in PSCAD. Here each model, for
the overhead transmission lines and underground cables, was constructed from physical
parameters. A simple reactor component was created in PSCAD to include mutual inductances. This component does not account for magnetic saturation, which can occur at the
switch-on moment, as shown in section 6.2.2 on page 88.
Verify the PSCAD model by comparing to measurements and calculated values.
Simulations of the system indicated an error in the model, as significant deviation was
apparent in the damping of transients and the frequency components in the system. The
resistances in the line and cable models in PSCAD were therefore examined for frequency dependency. Frequency dependency of the models was verified. For further evaluation of the
simulation, the parameters calculated by PSCAD were controlled against calculated values
and values given in datasheet. From this comparison, it was appeared that the overhead
line parameters, calculated in appendix A did match the same parameters calculated by
the PSCAD line constant program. However, the parameters for the cable sections deviated
significantly from the datasheet values. Adjustment of the cable configuration was made,
which contributed significantly to a solution closer to measurements. A lack of damping
was though still present. In order to evaluate the lack of damping, the value of the source
resistance was varied until the voltage waveform was comparable to the measurement. This
evaluation gave a series resistance of 20 which is an excessive size in such system.
Arrange a proposal for an optimization of the reconnection.
From simulations and measurements, the problem concerning overvoltages was estimated
for the specific case of the transmission system between Ferslev-Tinghj. Two well known
methods, for minimizing the switching overvoltages were given, and simulation of these were
conducted. The first incorporated the idea of switch-on at zero crossing of the voltages.
This principle showed great improvement in reduction of switching overvoltages, though
a large asymmetry in the current was present in the initial states. The second proposal
was to charge the system through a resistor that is later short circuited. Simulation of
this proposal also showed great improvement in minimizing generated overvoltages. Both
simulations showed limitations of the overvoltages to a maximum of 6 kV, compared with
a maximum of 64 kV without transient limitation.
In the measurements on the 150 kV level in Ferslev two frequency bands was present. These
frequencies are concluded, to stem from the two cable sections, in series with the inductances in
the system.
The two different solutions of the differential equations, describing the simplified model, are close
to identical. Furthermore, a simulation of the simplified model showed good accordance with the
solutions. It is therefore concluded that the solutions are correctly solved.
The overhead line parameters calculated by the project group, are in accordance with the
parameters calculated by the line constant program in PSCAD. It is therefore concluded that
these are correctly calculated.
From simulation with different frequencies is concluded, that the skin effect is accounted for in
the Frequency dependent (phase) model, in PSCAD.
8.2 Discussion
109
From simulation and measurements, it is concluded that the propagated overvoltages to lower
voltage levels, are not of a magnitude considered harmful. The proposals for minimizing the
switching overvoltages focuses therefore solely on the 150 kV transmission level.
From simulation of the two different proposals, it is concluded that the switching overvoltages
are limited significantly. Of the two proposals, the zero crossing principle is preferable, as this
does not add to the number of mechanical parts.
The constructed PSCAD model, does not represent the transmission system between FerslevTinghj adequately, as the switching transients in the simulations, are not damped as the
measured transients. It is therefore concluded that a more detailed simulation model of the
reactor is needed, along with implementation of frequency dependency in the short circuit
impedance of the system.
8.2
Discussion
Throughout the simulation, a lack of damping was present. The size of the needed series
resistance, from which the desired damping was achieved, is unreasonable in a transmission
system like between Ferslev-Tinghj. It is therefore concluded, that a part of the reason is to
be found elsewhere. As the included models in PSCAD are very detailed, the created reactor
component is incomparable in precision, as nonlinearities like saturation, are not included. As
indicated in section 6.2.2, the reactor operates in saturation in the initial states. As saturation
will decrease the inductance in the reactor, the impedance of the reactor will consequently be
lowered, and thereby change the amplitude, and harmonic components of the currents in the
system.
In order to obtain a comparable simulation with the measurements on the system, a more detailed
reactor component in PSCAD is needed. This component should include saturation, in order to
obtain a sufficiently detailed model. The exact value of the short circuit impedance is also
imperative, as is a realistic frequency dependency of this, if more exact simulation results are to
be obtained.
Proposition for further work, is obviously a more detailed model of the reactor and transformers,
which do not include magnetic saturation. Also of great interest, is why the PSCAD model does
not give identical results to measurements. Is this only caused by a insufficient model of the
aforementioned components, or are there other factors adding to the described inaccuracy in
damping ratio?
Bibliography
[1] S. Vrts. Elektriske Fordelingsanlg. Polyteknisk forlag, 4th edition, 1990. ISBN: 87-502-0707-5.
[2] Energinet.dk. http://www.miljorapport2006.dk/composite-284.htm. Visited: 28/2 2007.
[3] Energinet.dk. http://hornsrev2.energinet.dk. Visited: 16/2 2007.
[4] J. P. Bickford, N. Mullineux, J. R. Reed. Computation of power system transients. Peter Peregrinus
Ltd., 1st edition, 1980. ISBN: 0-906048-35-4.
[5] B. M. Weedy, B. J. Cory. Electric Power Systems. John Wiley & Sons Ltd., 4nd edition, 2001. ISBN:
0-471-97677-6.
[6] M. Abdel-Salam, E. Z. Abdel-Aziz. A charge simulation based method for calculating corona loss on
AC power transmission lines. Journal of Physics D: Applied Physics, vol. 27 no. 12, page 2570-2579,
1994.
[7] ABB Power Technology Products AB, Transformers. Instruction Manual. ABB, 2003, Rev. October
2004. Version: 2ZSE 460 022-BUD.
[8] T. Laubst. Overspndinger og overspndingsbeskyttelse i strkstrmssystemer.
genirakademi, 1st edition, 1983.
Danmarks In-
[9] Manitoba HVDC Research Centre Inc. PSCAD Manual v4.2.0. 2005.
[10] R. A. Serway, R. J. Beichner. Physics for scientists and engineers with modern physics. Saunders
College Publishing, 5nd edition, 2000. ISBN: 0-03020972-2.
[11] ABBs high voltage cable unit in Sweden. XLPE Cable Systems Users Guide. ABB, 2003, rev. 2.
[12] Manitoba HVDC Research Centre Inc. PSCAD Users Guide. 2005.
[13] S. Vrts. Overspndinger og Overspndingsbeskyttelse i elektriske strkstrmsanlg. Akademisk
forlag, 2th edition, 1965.
[14] W. G. Hurley, W. H. Wlfle, J. G. Breslin. Optimized Transformer Design: Inclusive of HighFrequency Effects. IEEE Transactions on Power Electronics, vol. 13 no. 4 page 651-659, 1998.
[15] ABB Power Technology Products AB. David Eklund and Pr Carlsson. 2007. Information provided
by ABB are attached on the CD.
[16] E. Kuffel, W.S.Zaengl and J. Kuffel. High Voltage Engineering: Fundamentals. Newnes, 2nd edition,
2000. ISBN: 0-7506-3634-3.
[17] DS/EN. Voltage characteristics of electricity supplied by public distribution systems. CENELEC,
1999. Corrected: 2004-04-16 and 2004-10-01.
BIBLIOGRAPHY
111
[18] M. Beanland, T. Speas, R. Rostron. Pre-insertion Resistors in High Voltage Capacitor Bank
Switching. Western Protective Relay Conference, October 2004.
[19] H. M. Ryan. High voltage engineering and testing. Peter Peregrinus Ltd. on behalf of the IEE., 1st
edition, 1994. ISBN: 0-86341-293-9.
[20] P. M. Anderson. Analysis of Faulted Power Systems. IEEE Press, 1st edition, 1995. ISBN: 0-78031145-0.
[21] C. A. Gross. Power system analysis. John Wiley & Sons, Inc., 2nd edition, 1986. ISBN: 0-471-862061.
[22] L. M. Faulkenberry and W. Coffer. Electrical Power Distribution and Transmission. Prentice-Hall,
1st edition, 1996. ISBN: 0-13-249947-9.
[23] V. Del Toro. Electric Power Systems. Prentice Hall, 1992. ISBN: 0-13-678228-0.
[24] D. E. Johnson, J. R. Johnson, J. L. Hilburn, P.D. Scott. Electric Circuit Analysis. John Wiley &
Sons, Inc., 3th edition, 1999. ISBN: 0-471-36571-8.
[25] E. Kreyszig. Advanced Engineering Mathematics. John Wiley & Sons. Inc., 8nd edition, 1999. ISBN:
0-471-15496-2.
[26] Omicron electronics. CMC 256 Operations Manual. 1999. Version: C256.AE.1.
[27] Omicron electronics. Omicron test universe 1.3 EnerLyser. 1999. Version: ENLY.AE.1.
Appendix
a product of :
Projectgroup - EPSH-810
31st may 2007
Appendix A
Parameter determination for the

overhead line system
In this appendix, the calculations of the parameters for the overhead transmission lines are
explained. The appendix is divided into two, one section for the series impedance, and one
section for the shunt admittance. Each section starts by giving theoretical background for the
calculations, before calculations of the parameters are performed.
A.1
Series impedance
In this section the series impedance for the transmission lines is calculated. The MatLab file used
for these calculations is to be found in appendix G.3, and on the attached CD.
A.1.1
Theoretical background
The theoretical background for calculating the inductance in the overhead lines is subsequently
presented.
Magnetic field around a conductor

When a conductor carries current, there will, as a result of that current, be generated a magnetic
field around it. This situation is shown in figure A.1.
B
I
x
r
Figure A.1: The figure shows the variation of the magnetic flux density with distance from the conductor
[1, Vrts p.102].
The relation between the magnetic field strength and the current, is described by Amperes law:
I
~ d~s = I N
H
Where:
~ is the field strength
H
N is the number of current carrying conductors
(A.1)
[A/m]
[-]
A-2
~ o at a distance x (see figure A.1) from the center of the conductor is:
The field strength H
~o 2 x = I
H
~o =
H
I
2x
(A.2)
[A/m]
~ and the field strength H

~ is given by:
The relation between the magnetic flux density B
~ = H
~
B
(A.3)
[T]
Where:
is the permeability constant for the material
[H/m]
For the case shown in figure A.1 the permeability of air applies, which approaches permeability
of free space:
= r 0 4 107
[H/m]
(A.4)
By combining equation A.2, equation A.3 and equation A.4, the following is valid for the magnetic
flux density at a distance x outside the conductor:
4 107 I
2x
I
[T]
= 2 107
x
~o =
B
(A.5)
A magnetic flux density generated by the current, will also exist inside the conductor, as shown
in figure A.2.
r
~i
B
Figure A.2: The figure shows a cross section of a current carrying cylindrical conductor.
The field strength, at any point x inside the conductor, is found to be:
2
~i 2 x = I x
H
r2
I x
~i =
H
2 r2
(A.6)
[A/m]
(A.7)
If the current in a conductor, at 50 Hz is assumed equally spread, over its cross section. The
2
magnetic flux density at the point x, shown in figure A.2, is only generated by the fraction xr2 ,
and the current running in the conductor as equation A.6 states:
~i = H
~i
B
I x
=
2 r2
(A.8)
[T]
(A.9)
A.1 Series impedance
A-3
Inductance between two conductors

In figure A.3, two long parallel, identical conductors are shown.
B
F ield1
F ield2
x
r
d
Figure A.3: Two parallel conductors, where the current I generates flux linkage between them.
The current I flows forward in conductor 1, and backward in conductor 2. This system shows
some similarity with a loop of a coil. Between the two conductors, there will, as in a coil, exist
mutual inductance. This inductance can be found as [1, p. 102]:
L=
N
=
I
I
(A.10)
[H]
Where:
is the magnetic flux
N is the number of turns in the coil
is the magnetic flux linkage.
[Wb]
[-]
[Wb turn]
The flux linkage is given as:

=N =BAN
Wb turns
m
(A.11)
To calculate the flux linkage of the system, first the flux linkage 1 from conductor 1 is calculated.
This is done by adding the two components from the inner and outer flux densities as shown by:
1 = 1i + 1o
(A.12)
= Bi A + Bo A
~ i and for the outer field B
~ o are known, from
Equations for the flux densities of the inner field B
equation A.5 and equation A.8. The flux linkage for the inner field of conductor 1 is calculated
as:
2 107 I x x2
2 dx
r2
r
0
Z r
7
2 10 I
x3 dx
=
r4
0
I
= 2 107
4
1i =
(A.13)
For the outer field, the flux linkage is calculated in the same way, to be:
Z
2 107 I
dx
x
r
= 2 107 I (ln(d) ln(r))

d
= 2 107 I ln
r
1o =
(A.14)
A-4
The flux linkage in conductor 1, is found by combining equations A.12, A.13 and A.14:
1 = 1i + 1o

d
2 107 I
7
+ 2 10 I ln
=
4
r

d
1
+ ln
= 2 107 I
4
r
(A.15)
By combining equation A.10, with equation A.15 and assuming, because of symmetry of the
conductors, that 1 = 2 , the total inductance of the system is found as:
1 + 2
I 7

2 10 2 I + 107 I ln dr
L =
=
= 4 107
I
1
d
+ ln
4
r
(A.16)
[H/m]
The inductance of each conductor per meter is:

7
L = 2 10

1
d
+ ln
4
r
(A.17)
Flux linkage in overhead line transmission system

In order to calculate the parameters for the overhead lines, equations describing the flux linkage
for multiple conductors are needed.
In the following, a generalized equation will be derived, for a system consisting of n-number of
conductors, as shown in figure A.4.
y
2
4
d2x
d12
n
1
d1x
r1
3
Figure A.4: The figure shows a cross section of, n conductors.
The sum of the currents in the system is presumed to be zero. The flux linkage for conductor 1
is calculated as in equation A.15:
11 = 2 107
d1x
I1
+ I1 ln
4
r1

Wb turns
m
Where:
d1x is the distance between the center of conductor 1 to the point x
r1 is the radius of conductor 1
(A.18)
[mm]
[mm]
A-5
And the effect of the current in conductor 2, is found by:

12 = 2x 21
7
= 2 10
d2x
I2 ln
d12
(A.19)
The total flux linkage contribution in conductor 1, from the currents in all n conductors is:
1 = 11 + 12 + ... + 1n

I1
d1x
d2x
dnx
1 = 2 107
+ I1 ln
+ I2 ln
+ ... + In ln
4
r1
d12
d1n
(A.20)
As stated the sum of the currents in the system, is equal to zero:

(A.21)
In = I1 I2 I3 ... In1
And then equation A.20 is re written as:
7
1 = 2 10
I1
d1n d1x
+ I1 ln
4
r1 dnx
d1n d2x
+ I2 ln
d12 dnx
d1n d(n1)x
+ ... + In1 ln
d1(n1)dnx
!!
(A.22)
If the point x is moved to infinity along the x-axis (x ), then equation A.22 is simplified to:
7
1 = 2 10
I1
d1n
+ I1 ln
4
r1
d1n
+ I2 ln
d12
d1n
+ ... + In1 ln
d1(n1)
!!
(A.23)
This equation can be rewritten from the fact that

1 = 2 107
1
I1
+ I1 ln
4
r1
I = 0, simplifying equation A.23 to:
+ I2 ln
1
d12
+ ... + In ln
1
d1n

(A.24)
Simplified again gives:

7
1 = 2 10
1
I1 ln
r1
1
+ I2 ln
d12
1
+ ... + In ln
d1n

(A.25)
Here is r1 the self GMD of the conductor. This equation gives the flux linkage for conductor 1,
per meter. The same is derived for the other n-conductors.
A.1.2
Determination of series impedance for overhead line transmission

system
The objective of this section, is to produce equations for the transmission system, which can be
used directly to find the impedances of the overhead line section.
The overhead lines in the system, are arranged in a triangle, with a ground conductor above the
phase wire, which apart from giving return path for the currents, serves as a lightning protection
for the phases. The geometry of the overhead lines, is shown in figure A.5, where the distances
between phases are annotated.
A-6
g
dcg
hg
dag
dbg
dac dbc
dab
hc
ha,b
De
Figure A.5: The geometry of the conductors at the Donau mast.
The generalized equation A.25 can be used to describe the inductances, in the case shown in
figure A.5, a more convenient way, is to write the equation in its matrix form.
= LI
a
b
c
g
d
Wb turns
m
ln( r1 )
1
) ln( d1ad )
ln( d1ab ) ln( d1ac ) ln( dag
ln( d1da )
ln( 1 )
rb
ln( d1cb )
ln( d1gb )
ln( d1db )
ln( 1 )
d
ba
1
= 2 107
ln(
dca )
ln( 1 )
dga
ln( d1bc )
ln( r1 )
c
ln( d1gc )
ln( d1dc )
ln( d1bg )
ln( d1cg )
ln( r1 )
g
ln( d1dg )
ln( d1bd )
ln( d1cd )
ln( d1gd )
ln( 1 )
rd
Ia
Ib
Ic
Ig
Id
(A.26)
The resistance matrix for a system with 5 current paths, is given by the line resistance on
the diagonal, and zero on non-diagonal, where the earth resistance can be approximated to 50
m/km [1, Vrts p. 118].
A-7
Rabcgd
Ra 0
0
0
0
0 Rb 0
0
0
0
0 Rc 0
0
0
0
0 Rg 0
0
0
0
0 Rd
(A.27)
The series-impedance matrix for the system is then found by adding the resistance and inductive
reactance.
X = L/
(A.28)
Zabcgd = Rabcgd + jXabcgd
(A.29)
For the case shown in figure A.5, there are five current paths. One for each phase, one in the
ground conductor, and one in the earth. The schematic of this is shown in figure A.6.
a
Ia Zaa
Ib Zbb
Ic Zcc
g
Va
Vb
Vc
Ig
a
Zab
b
Zac
Zbc
Zgg
Zcg
Zag c
h
Zbg
Vg = 0
Zad
Zbd
Zcd
Zgd
Vd =0
Id
De
Zdd
d
Lenght
Figure A.6: The three phase oh-transmission lines with ground wire, and earth return
The notation in figure A.6 will subsequently be used, where a-b-c indicate values for each phase,
g indicates value for the ground conductor, and d indicates value for the earth return. For the
system shown in figure A.6, the voltage equation is written:
Vaa
Vbb
Vcc
Vgg
Vdd
Va Va
Vb Vb
Vc Vc
0 Vg
0 Vd
Zaa
Zba
Zca
Zga
Zda
Zab
Zbb
Zcb
Zgb
Zdb
Zac Zag Zad

Zbc Zbg Zbd
Zcc Zcg Zcd
Zgc Zgg Zgd
Zdc Zdg Zdd
Ia
Ib
Ic
Ig
Id
(A.30)
A reduction of the 55 impedance matrix to a 33 matrix is desired, in order to transform

the impedances to a sequence form. An inspiration for this reduction is found in [20, Anderson
p.113]. In the system shown in figure A.6, the sum of the currents, is zero, and the earth return
is shunted with the ground conductor, giving the following current relation:
A-8
Id Ig = Ia + Ib + Ic
(A.31)
Id = Ia Ib Ic Ig
The matrix A.30, can thereby be reduced by its 5th column, giving 5 equations and four
unknowns:
Vaa
Vbb
Vcc
Vgg
Vdd
Zaa Zad Zab Zad Zac Zad Zag Zad

Zba Zbd Zbb Zbd Zbc Zbd Zbg Zbd
Zca Zcd Zcb Zcd Zcc Zcd Zcg Zcd
Zga Zgd Zgb Zgd Zgc Zgd Zgg Zgd
Zda Zdd Zdb Zdd Zdc Zdd Zdg Zdd
Ia
Ib
I
c
(A.32)
Ig
Now are the impedances for the earth return (row 5) subtracted from the other impedances (rows
1 to 5):
Va
Vb
Vc
Vg = 0
=
=
=
=
Vaa Vdd
Vbb Vdd
Vcc Vdd
Vgg Vdd
(A.33)
Giving the following impedance matrix, where the 5th row has been removed, as all entries are
equal to zero:
ad ) (Zda Zdd )
(Zaa Z
(Zba Zbd ) (Zdb Zdd )
cd ) (Zdb Zdd )
(Zca Z
dd )
(Zga Zgd ) (Zdb Z
db Zdd )
(Zab Zad ) (Z
bd ) (Zdb Z
dd )
(Zbb Z
dd )
(Zcb Zcd ) (Zdb Z
gd ) (Zdb Zdd )
(Zgb Z
dd )
(Zac Zad ) (Zdc Z
bd ) (Zdc Zdd )
(Zbc Z
(Zcc Zcd ) (Zdc Zdd )
dd )
(Zgc Zgd ) (Zdc Z
ad ) (Zdg Z
dd )
(Zag Z
dd )
(Zbg Zbd ) (Zdg Z
cd ) (Zdg Zdd )
(Zcg Z
dd )
(Zgg Zgd ) (Zdg Z
(A.34)
Reduction of this matrix gives a 44 matrix Zabcg :
aa + Z
dd Zad Zda
Z
ba + Z
dd Z
bd Zda
Z
da
Zca + Zdd Zcd Z
ga + Zdd Z
gd Z
da
Z
ab + Z
dd Zad Z
db
Z
bb + Zdd Z
bd Zdb
Z
cb + Zdd Zcd Zcb
Z
gb + Zdd Zgd Z
cb
Z
dd Zad Z
dc
Zac + Z
bd Zdc
Zbc + Zdd Z
Zcc + Zdd Zcd Zdc
dc
Zgc + Zdd Zgd Z
ag + Z
dd Z
ad Z
dg
Z
bg + Zdd Z
bd Zdg
Z
bg + Zdd Z
cd Z
dg
Z
gg + Zdd Zgd Z
dg
Z
(A.35)
Each element in equation A.35 is in the following form:

Zij = Zij Zid Zdj + Zdd
(A.36)
Using this fact, the matrix can be simplified:
Va
Vb
Vc
Vg = 0
Zaa
Zba
Zca
Zga
Zab
Zbb
Zcb
Zgb
Zac
Zbc
Zcc
Zgc
Zag
Zbg
Zcg
Zgg
Ia
Ib
Ic
Ig
(A.37)
A-9
For the diagonal in this matrix, the impedance is given by:

7
Zii = Ri + Rd + j 2 10
De
ln
ri
(A.38)
And for the non diagonal terms:

7
Zij = Rd + j 2 10
De
ln
dij
(A.39)
Where:
[m]
De is the distance to the earth return
With the knowledge that De >> h, the distance between the earth return and conductors is
approximated to De .
r
[m]
(A.40)
De = 660
f
Where:
is the specific resistance of the earth
[ m]
f is the power frequency
[Hz]
In order to reduce the impedance matrix from a 4 4 matrix to a 3 3 matrix, the following
operation is performed [20, Anderson p.114]:
abc
Z
Zag h
Zaa Zab Zac
= Zba Zbb Zbc Zbg

Zcg
Zca Zcb Zcc
1
Zgg
This leaves a 3 3 matrix for the impedance:

abc
Z
ih
Zaa Zab Zac

= Zba Zbb Zbc

Zca Zcb Zcc
Zga Zgb Zgc
(A.41)
(A.42)
Where each element is given by the general expression:
Zig Zgj
Zij = Zij
Zgg
(A.43)
With a general expression for the series impedance for each phase known, the next step is to find
the sequence matrix. In order to do this, the transformation matrix T, is introduced [21, Gross,
p. 61]:
1 1 1
T = 1 a2 a
1 a a2
a = e
j2
3
(A.44)
(A.45)
The transformation matrix explains the relationship between the phase values and the sequence
values for both voltage and current. This relationship can now be used to transform the phase
values of the impedance into the sequence values. This is done with the following equation [21,
Gross, p. 62]:
Z012 = T1 Zabc T
(A.46)
The sequence matrix for the series impedance can now be calculated by use of the explained
method.
A-10
A.1.3
Calculation of the series impedance for the overhead line section
To calculate the impedances in the overhead line transmission system, physical values and
dimension for each section is needed. Those information is listed in table A.1 and A.2.
Conductor name
Length
Conductor resistivity
Cross section total
Cross section aluminum
Self GMD
Number of threads in conductor
Ground conductor
Cross section of ground conductor
Section 2
Martin
18,93 km
0,028264 m
772 mm2
685 mm2
0,0146 m
54
Dorking
153 mm2
Table A.1: Physical values for parameter calculation.
dab
dac
dag
De
5,8
9,364
16,939
660
m
m
m
m
9,176 m
17,528 m
8,355 m
dbc
dbg
dcg
Table A.2: Distances between conductors.
Series resistance
The DC resistance for conductor is given by:
Rdc =
l
A
(A.47)
Where:
=0,028264 is the specific resistivity for aluminum
[mm2 /m]
l is the length of the conductor
[m]
A is the cross section of the aluminum conductor
[mm2 ]
The series resistance is calculated for DC at 20 C. The actual resistance in the overhead line will
be affected by the terms listed below:
Temperature
Skin effect
Proximity effect
Conductor sag
Spiraling
In order to include the above-mentioned terms, a correction factor is estimated in the following.
A-11
Temperature The operation temperature will have an effect on the series resistance according
to the following general equation [10, NKT, Kabelteknisk hndbog p. C3]:
(A.48)
Rt = 1 + R20 (1 + (t 20)) = R20 (1 + ktemp )

Where:
R20 is the reference resistance at a temperature of 20 C
is the temperature coefficient for the conductor material
t is the actual temperature
[]
[ C1 ]
[ C]
For the overhead line system the conductor material is aluminium with a temperature coefficient
= 0,00403 C1 . The reference resistance R20 is the calculated DC resistance. The operational
temperature is unknown as the focus is at the switch-on of the system. The actual temperature
is therefore dependent on the time between the disconnection and reconnection of the grid. In
order to find a correction factor for the resistance, the operational temperature at the switch on
is assumed to 40 C, which gives the following temperature correction factor:
ktemp = 0, 00403(40 20) = 0, 0806
(A.49)
Skin effect The skin effect influences the resistivity of a conductor when an AC source is
used. Due to the alternating current, will the current density increase near the surface of the
conductor, and vice versa decrease in the middle of the conductor. This phenomena will cause
a smaller effective conducting area, hence a larger resistance. An approximation for calculating
the AC resistance is presented in [14, Hurley, p. 6]. First is the skin depth calculated:
r
f
= 11, 96 mm
=
(A.50)
(A.51)
Where:
is the skin depth
is the resistivity of the conductor
is the permeability of the conductor
For a non-magnetic conductor, as the case is with the aluminum conductor, the total permeability
is the permeability of free space[10, Serway, p.959]. A factor describing the ratio between RAC
and Rdc is given.
Rac
Rdc
(A.52)
= 1 + kskin
kskin =
kskin =
(r/)4
48 + 0, 8(r/)4

14,77 4
11,96
48 + 0, 8

14,77 4
11,96
for
= 0, 0465
ro
> 1, 7
(A.53)
(A.54)
Proximity effect When conductors are placed near each other, the current will be affected,
as the magnetic field presses the electrons, causing higher density. Due to this, the current
density will not be equally distributed in the conductor. This will result in an increased effective
resistance. However, the proximity effect of overhead lines is normally small, due to the large
distances between the conductors. It is therefore neglected here.
A-12
Conductor sag When the conductors are mounted between two towers, the length of the
conductor between the two towers will be larger than the actual length between the two towers.
The sag of the conductor will approximately add 0,2% of the total length. This is found from
the maximum sag of 7,6 m and the average distance between the towers of 292 m.
Spiraling The conductors in an overhead line system, consists of a bundle of small conductors.
These small conductors are twisted around the axis of the conductor, resulting in an increase of
the length of the current path in the conductor. The spiraling of the conductor usually adds 2%
to the length of the conductor, and there by an increase of the total resistance [22, Faulkenberry,
p. 22].
The total correction factor can be calculated from each of the correction factors:
(A.55)
ktot = (1 + ktemp ) (1 + kskin ) (1 + kprox ) (1 + ksag ) (1 + kspir )
ktot = (1 + 0, 0806) (1 + 0, 0465) (1 + 0) (1 + 0, 002) (1 + 0, 02) = 1, 15
(A.56)
The actual resistance can now be calculated, by using the correction factor:
(A.57)
Rtot = Rdc ktot
The resistances for the overhead line section and the ground conductor are calculated in MatLab1 .
The series resistances are shown in table A.3.
Total resistances for section 2
DC resistance AC resistance
Phase a-b-c (Rdc )
0,781
0,898
Ground wire (Rdc,g )
3,497
Earth return (Rd )
0,947
Table A.3: Series resistances for the overhead line.
Series inductance
The calculations for the series inductance are carried out in MatLab. The m-files2 used for this
are impedance.m, constants.m and geometry.m.
The calculated inductance matrix for section 2 is:
[H]
(A.58)
[]
(A.59)
36, 188 13, 570 10, 890

Laa Mab Mac
3
L = Mba Lbb Mbc = 10 13, 570 36, 267 11, 014
10, 890 11, 014 34, 384
Mca Mcb Lcc
The calculated sequence impedance matrix for section 2 is:
Z012
1
2
3, 311 + j18, 618 0, 399 + j0, 246 0, 429 + j0, 189
= 0, 429 + j0, 189 0, 903 + j7, 474 0, 319 j0, 160

0, 399 + j0, 246 0, 312 j0, 169
0, 903 + j7, 474
The m-file is shown in appendix G.3 on page G-58.

The m-files are shown in appendix G.1 on page G-57.
A-13
In table A.4 the calculated sequence inductances for the overhead line section are listed, including
values per km.
Inductance
L0
L1
L2
Section 2
59,3 mH
23,8 mH
23,8 mH
Per km
3,13 mH/km
1,26 mH/km
1,26 mH/km
Table A.4: Positive, negative and zero sequence inductance for the overhead line section.
For comparison, typical values for the inductances in an overhead line system is found in
Elektriske fordelingsanlg [1, Vrts, p. 133-134]. Here a typical value for the positive sequence
inductance is stated to be 1,27 mH/km, which is in accordance with the calculated value of
1,26 mH/km. For the zero sequence inductance, typical values is given in the range of 3,82
mH/km to 4,77 mH/km for overhead lines without earth conductor. For overhead lines with
earth conductor, the values are given to be 10-20% lower, which gives a range from 3,06 mH to
3,82 for 20% lower values. The calculated value for the zero sequence inductance is therefore
within this interval, and the calculated values for the inductances are assumed valid.
In table A.5 are the calculated sequence resistances for the overhead line section listed, including
values per km.
Resistance
R0
R1
R2
Section 2
3,311
0,903
0,903
Per km
0,175 /km
0,0477 /km
0,0477 /km
Table A.5: Positive, negative and zero sequence resistance for the overhead line section.
The typical value for the positive sequence resistance is very dependent on the cross sectional
area of the conductor. The largest cross section given in Elektriske fordelingsanlg [1, Vrts,
p. 133-134], for a typical value is a 402 mm2 + 52 mm2 Steel-Aluminium conductor. Here the
typical resistance is specified to 0,07 /km. The conductors used in the overhead line system
have a cross section of 772 mm2 , and the resistance for these conductors is therefore expected
to be lower than the typical value of 402 mm2 . This is also the case with a positive sequence
resistance of 0,0477 /km. For the zero sequence resistance, are the typical values usually slightly
above 0,15 plus the positive sequence resistance (0,15 + R1 ). This is in accordance with the
calculated value of 0,175 /km. These calculated values are therefore assumed valid.
A-14
A.2
Shunt admittance
The shunt admittance of a transmission line system is determined by the conductance and
capacitances between the conductors. Here is the conductance neglected, and focus only on
finding the capacitance of the overhead lines. The principle for determination of the shunt
capacitance will be described in this section.
A.2.1
In order to obtain equations for calculations of the shunt admittance, the theoretical background
for those equations is presented.
Electric field created by conductors
The cross section of a single conductor, is shown in figure A.7. The conductor has the radius r
and is carrying a charge q.
d2
d1
q
Figure A.7: Cross section of a single conductor.
If the conductor shown in figure A.7 is considered to be infinitely long, the electric flux out of the
conductor will spread radially. According to Gausss law, the total electric flux out of a closed
surface will be equal to the total charge in the conductor:
I
~ dA
~ = qtotal
D
(A.60)
[C]
Where:
~ is the electric flux density
D
~ is the area over which the surface integral is performed
dA
qtotal is the total charge the conductor is carrying
[C/m2 ]
[m2 ]
[C]
The electric flux density can now be found by differentiating both sides of equation A.60, resulting
in:
~ = dq
D
~
dA
[C/m2 ]
(A.61)
As the flux density is determined by the charge on a given surface area, the flux density can be
found by using the total charge and the total surface area of the conductor:
~ =
D
~ =
D
q
A
q
2rl
(A.62)
(A.63)
A.2 Shunt admittance
A-15
Where:
r is the radius of the conductor
l is the unit length of the conductor
[m]
[m]
Here the unit length is chosen to 1 m, in order to obtain the results per meter. The electric flux
density is then:
~ = q
D
2r
(A.64)
The relationship between the electric flux density and the electric field strength, is determined
by the permittivity of the material the electric field passes:
~ = E
~
D
(A.65)
Using this, the electric field around the conductor can be determined:
~ =
E
~ =
E
~
D
q
2r
(A.66)
(A.67)
[V/m]
With the electric field known, it is possible to find the potential difference between two points.
Potential difference
The potential difference U , between two points, a and b, is determined by the integral of the
electric field between these two points: [10, Serway, p. 768-774]:
Uab = U = Ua Ub =
~ d~l
E
Where:
Uab is the potential difference between a and b
Ua is the potential at point a
Ub is the potential at point b
(A.68)
[V]
[V]
[V]
[V]
The potential difference between the two points d1 and d2 in figure A.7, can then be calculated:
U12 =
U12 =
U12 =
d2
q
d~r
d1 2r
Z d2
q
1
d~r
2 d1 r
q
d2
ln
2
d1
(A.69)
(A.70)
(A.71)
With the ability determine the potential difference between two points, it is possible to determine
the capacitance between two conductors.
A-16
Capacitance between two conductors

Figure A.8 shows two parallel conductors, with opposite charge. The considered conductors are
located in air.
Conductor 1
Conductor 2
r1
r2
+q
d
Figure A.8: Two parallel conductors.
For a conductor located in air, the relative permittivity r is approximately unity, hence the
permittivity approaches the permittivity of free space [23, Del Toro, p. 192]:
1
= r 0
109
= 0 =
36
[F/m]
(A.72)
The electric field around the conductor can then be determined by replacing in equation A.67
with the permittivity in air:
q
109 r
~ = 18 109 q
E
r
~ =
E
1
2 36
(A.73)
(A.74)
The potential difference between two conductors can be found through superposition, by applying
a sample charge on each conductor, one at a time, and note the potential created by each
conductor. In a overhead transmission line system, the distance d will be much greater than the
radius of the conductor. This allows the radius rb to be neglected when integrating the electric
field between the conductors, hence simplifying the calculations. Likewise, the electric field can
be assumed uninfluenced by the other conductors and therefore radially shaped [1, Vrts, p. 136].
The potential difference between the two conductors, stems from the charges on each conductor.
is the contribution created by the charge +q, and can be expressed as:
The potential U12
U12
1
d~r
r1 r

d
9
= 18 10 q ln
r1
U12
= 18 109 q
(A.75)
(A.76)
is the contribution created by the charge q, and can be expressed as:

The potential U12
U12
U12
U12
r2
1
d~r
d r

r2
= 18 109 (q) ln
d

d
= 18 109 q ln
r2
9
= 18 10 (q)
(A.77)
(A.78)
(A.79)
A-17
The potential difference between the two conductors is now calculated by adding the two
contributions:
U12 = U12
+ U12
U12
U12
(A.80)

d
d
+ 18 109 q ln
= 18 109 q ln
r1
r2
!
2
d
= 18 109 q ln
r1 r2
(A.81)
(A.82)
If the radius r = r1 = r2 , the equation is simplified:

U12
U12
d2
= 18 10 q ln
r2

d
= 36 109 q ln
r
9
(A.83)
(A.84)
The relationship between the potential difference and the charge is [1, Vrts, p. 137]:
U
U
(A.85)
= pq
1
q
=
C
(A.86)
Where:
p is the potential coefficient
C is the capacitance
[m/F]
[F/m]
1
C12
Cp
Cp
Figure A.9: Schematic of the capacitance in a system of two parallel conductors.
Figure A.9 shows the three capacitances in a system with two conductors. The capacitance C12
is now found from the relation between the potential difference and the charge in equation A.84.
p12 =
C12 =

d
U12
= 36 109 ln
q
r
q
1

=
U12
36 109 ln dr
(A.87)
(A.88)
The capacitance per conductor Cp is representing the capacitance between the conductor and
neutral as shown in figure A.9. When r = r1 = r2 the capacitances per conductor will have the
same value, namely two-times the value of C12 according to the schematic on figure A.9:
pp = 18 109 ln
Cp =
1
18 109 ln

d
r

d
r
(A.89)
(A.90)
A-18
Capacitance to ground
A conductor placed above ground, will have a capacitance to ground. The electric field around
the conductor will, because of the neutral potential of the ground, have the same appearance
as if there, instead of ground, were a conductor carrying the opposite sign of the charge, in its
mirror image as shown in figure A.10 [1, Vrts, p. 137].
r
h
G
h
q
Figure A.10: A single conductor with its mirror image under ground.
The capacitance and potential coefficient to ground will then be the same as the capacitance and
potential coefficient per conductor in equation A.89 and equation A.90 according to figure A.9.
In figure A.10 the distance between the conductors is 2h, this gives the potential coefficient and
capacitance:
pa0
Ca0
2h
= 18 10 ln
r
1

=
18 109 ln 2h
r
9
(A.91)
(A.92)
This capacitance is annotated Caa and can be conceived as a self capacitance. The self potential
coefficient is then given as:
paa
2h
= 18 10 ln
r
9
(A.93)
If two conductors are placed above ground, the capacitance between the conductors will be
affected by the capacitances to ground.
In equation A.93 the potential coefficient to ground was found for a single conductor. If another
conductor b is present, the conductor a will be affected (shown in figure A.11). The potential
created between a and G is determined by contributions from conductor b and its mirror image:
Ua UG = Uab+ UG + Uab UG
(A.94)
The potential at ground UG is defined to be neutral, which gives:

Ua = Uab+ + Uab
(A.95)
A-19
b+
dab
h
G
Dab
h
b q
Figure A.11: Two conductors above ground.
The potential at conductor a can now be expressed with the charge on conductor b and its
potential coefficients to conductor a:
(A.96)
Ua = Uab+ + Uab
Ua
Ua
h
h
= 18 109 q ln
+ 18 109 (q) ln
d
Dab
ab
D
ab
= 18 109 q ln
dab
(A.97)
(A.98)
The mutual potential coefficient and capacitance between conductor a and conductor b are now
found:
pab
Cab
Dab
= 18 10 ln
dab
1

=
ab
18 109 ln D
dab
9
(A.99)
(A.100)
With the ability to determine both the mutual capacitance between two conductors (equation
A.100) and the capacitance between a conductor and ground (equation A.92), it is possible to
determine the capacitances in a system consisting of multiple conductors.
Capacitance between multiple conductors
In a system of multiple conductors, capacitances will exist between all of the conductors.
The potential on conductor a in such system, is determined by the contributions from all the
conductors, including contribution from its mirror image.
1
1
1
1
qa +
qb +
qc + ...
qn
Caa
Cab
Cac
Can
n
X
1
=
qj
C
j=a aj
Ua =
(A.101)
Ua
(A.102)
Ua =
n
X
j=a
paj qj
(A.103)
A-20
The potential in the other conductors is found by the same principle and gives:
Ub =
Uc =
n
X
j=a
n
X
j=a
..
.
Un =
n
X
j=a
pbj qj
(A.104)
pcj qj
(A.105)
pnj qj
(A.106)
These equations can be arranged in a general matrix equation:
Ua
Ub
Uc
..
.
Un
paa
pba
pca
..
.
pab
pbb
pcb
..
.
pna pnb
pac . . .
pbc . . .
pcc . . .
..
..
.
.
pnc . . .
pan
pbn
pcn
..
.
pnn
qa
qb
qc
..
.
qn
(A.107)
With a potential matrix produced for a system with multiple conductors, it is now possible to
determine the appropriate equations for the overhead transmission line system.
A.2.2
Determination of shunt admittance for overhead line transmission

system
To determine the admittances for the overhead line system shown in figure A.12, a potential
matrix will be constructed.
g
3
1
D11
d12
D12
Figure A.12: Distances between conductor 1, 2 and its mirror images.
A-21
The general matrix equation, for a system of multiple conductors is shown in equation A.107.
For the system shown in figure A.12 the matrix equation is found:
U1
U = Pq
U2
U
3
Ug
(A.108)
p11 p12
p22
p32
pg1 pg2
p13
p23
p33
pg3
p21
=
p
31
p1g
p2g
p3g
pgg
q1
q2
q
3
(A.109)
qg
The conductors self potential coefficient is placed on the diagonal of the potential matrix and
represent the potential coefficient to ground. This coefficient was found in equation A.93, and
can be written for a general case:

Dii
pii = 18 10 ln
r
9
(A.110)
[m/F]
Where:
Dii is the distance between the ith conductor and its own mirror image
[m]
The mutual potential coefficient represent the potential coefficient between two conductors. This
coefficient was found in equation A.99, and can be written for a general case:
Dij
pij = pji = 18 109 ln
dij
(A.111)
[m/F]
Where:
Dij is the distance between the ith conductor and the j th mirror image
[m]
Inserting the potential coefficients for the system in figure A.12 into equation A.109 gives:
U1
U2
U3
Ug
D13
1g
12
ln( Dr11 ) ln( D
d12 ) ln( d13 ) ln( d1g )
ln( D21 )
d21
9
18 10
ln( D31 )
d31
23
ln( Dr22 ) ln( D
d23 )
D33
32
ln( D
d32 ) ln( r )
D
q1
D2g
ln( d2g
) q2
D3g
q
ln( d3g
)
3
g1
g2
g3
ln( dg1
) ln( dg2
) ln( dg3
) ln(
Dgg
rg )
(A.112)
qg
In order to use the transformation matrix T to find the sequence matrix for the shunt admittance,
the constructed potential matrix needs to be reduced to a 3 3 matrix. To do this the potential
matrix is subdivided into four matrices:
Pabcg =
Where:
P1
P2
P3
P4
is
is
is
is
a
a
a
a
33
13
31
11
matrix
matrix
matrix
matrix
.
P1 .. P2
.
P3 .. P4
(A.113)
A-22
As the potential at the ground wire is neutral (Ug = 0) equation A.112 can be simplified:
U1
U2
U3
0
..
.
..
.
..
.
D12
D13
D11
ln( r ) ln( d12 ) ln( d13 )
ln( D21 )
d21
18 109
ln( D31 )
d31
Dg1
ln( dg1
)
ln( Dr22 )
23
ln( D
d23 )
D33
32
ln( D
d32 ) ln( r )
Dg2
ln( dg2
)
Dg3
ln( dg3
)
1g
)
ln( d1g
D2g
)
ln( d2g
D3g
)
ln( d3g
..
D
)
. ln( rgg
g
q1
q2
q3
qg
(A.114)
With the above-mentioned matrix equation defined, a 3 3 matrix can be obtained through
reduction [20, Anderson, p. 169]:
Pabc = P1 P2 P1
4 P3
(A.115)
The capacitance matrix is now found by taking the inverse of the potential matrix:
Cabc = P1
abc
(A.116)
The susceptance is found from the capacitance through their relationship:

1
1
Babc =
= 1
Xabc
Cabc
Babc = Cabc
(A.117)
(A.118)
As the conductance in overhead line systems is very small, it is neglected here. The admittance
matrix can therefore be seen as only imaginary, determined by the susceptance:
Yabc = 0 + jBabc
(A.119)
The sequence matrix can then be found using the transformation matrix:
Y012 = T1 Yabc T
A.2.3
(A.120)
Calculation of sequence matrix for the shunt admittance
In order to make the calculations described previously. The distances in equation A.114 must
be known. In figure A.13 those distances are shown.
g
d1g
3
d13
d12
1
2
D11
D12
D13
D1g
d21
1
D21
d2g
3
d23
2
D22
D23
D2g
d3g
3
dg3
3
d31
1
d32
2
dg1
1
dg2
2
D31
D32
Dg1
Dg2
D33
D3g
Figure A.13: Distances for the three phase conductors and the ground wire.
Dg3
Dgg
A-23
The distances in figure A.13 are calculated3 and listed in table A.6.
Distance
Length
Distance
Length
Distance
Length
Distance
Length
d12
5,80 m
d21
5,80 m
d31
9,36 m
dg1
16,94 m
d13
9,36 m
d23
9,18 m
d32
9,18 m
dg2
17,53 m
d1g
16,94 m
d2g
17,53 m
d3g
8,36 m
dg3
8,36 m
D11
49,40 m
D21
49,74 m
D31
58,29 m
Dg1
66,31 m
D12
49,74 m
D22
49,40 m
D32
58,26 m
Dg2
66,46 m
D13
58,29 m
D23
58,26 m
D33
67,00 m
Dg3
75,13 m
D1g
66,31 m
D2g
66,46 m
D3g
75,13 m
Dgg
83,20 m
Table A.6: Distances for the three phase conductors and the ground wire shown in figure A.13.
The capacitance matrix for section 2, can now be calculated with the method presented
previously4 .
Cabc
0, 149
0, 0275 0, 0114
Caa Cab Cac
6
= Cba Cbb Cbc = 10 0, 0275
0, 149
0, 0127
0, 0114 0, 0127
0, 165
Cca Ccb Ccc
[F]
(A.121)
(A.122)
The sequence admittance sequence matrix for section 2 is calculated to:
Y012
0, 000 + j37, 624 2, 935 j1, 509

2, 935 j1, 509
6
= 10 2, 935 j1, 509
0, 000 + j53, 829
1, 433 + j0, 618
2, 935 j1, 509 1, 433 + j0, 618 0, 000 + j53, 829
The sequence susceptances for the section are shown in table A.7.
Susceptance
B0
B1
B2
Section 2
37,62 S
53,83 S
53,83 S
Per km
1,988 S/km
2,844 S/km
2,844 S/km
Table A.7: Positive, negative and zero sequence susceptance for the overhead line section.
Table A.7 can be recalculated in order to find the capacitance using the relation in equation
A.118. The results for the capacitance are listed in table A.8.
Capacitance
C0
C1
C2
Section 2
0,1198 F
0,1713 F
0,1713 F
Per km
6,328 nF/km
9,049 nF/km
9,049 nF/km
Table A.8: Positive, negative and zero sequence capacitance for the overhead line section.
3
4
The distances are calculated in appendix G.2

The calculations are performed in MatLab as shown in appendix G.4.
A-24
To verify the calculated values of the line parameters, the per km values are compared with
normal values found in literature. In Elektriske Fordelingsanlg [1, S. Vrts, p. 157] the
admittance for overhead transmission lines is stated to have a typical magnitude of 3 S/km.
This is in accordance with the calculated values of 2,844 S, shown in table A.7. For the zero
sequence susceptance the typical value is given to 1,8 S, which is also in accordance with the
calculated values of 1,988 S. On the basis of these comparison, the calculated values are assumed
to be valid.
A.3
Summary
In this chapter has the determination of the series impedance and the shunt admittance for the
overhead line section been presented. Each section started with a theoretical review in order to
achieve general equations for the calculations. On the basis of these equations, parameter values
for the overhead transmission system were determined. The calculated phase values for the series
inductance, series resistance and the shunt capacitance are shown in table A.9.
Inductance
La
Lb
Lc
Section 2
36,2 mH
36,3 mH
34,4 mH
Resistance
Ra
Rb
Rc
Section 2
0,898
0,898
0,898
Capacitance
Can
Cbn
Ccn
Section 2
0,149 F
0,149 F
0,165 F
Table A.9: Phase values for self inductance and resistance for the overhead line section.
The calculated mutual inductance and mutual capacitances are shown in table A.10.
Mutual inductance
Mab
Mac
Mbc
Section 2
13,6 mH
10,9 mH
11,0 mH
Mutual capacitance
Cab
Cac
Cbc
Section 2
0,0275 F
0,0114 F
0,0127 F
Table A.10: Phase values for mutual inductance and mutual capacitance in the overhead line section.
The calculated phase values listed in table A.9 and A.10 will be used for the electric representation
of the overhead line.
Appendix B
Parameter determination in the shunt

reactor
In this appendix an analysis will be performed on the magnetic relations in the shunt reactor,
resulting in a determination of the inductances in the shunt reactor.
B.1
Before the inductances in the shunt reactor are determined, a brief theoretical review of magnetic
field theory will be performed to find the essential equation for determination of the inductances.
In figure B.1 is a simple element with the permeability shown.
H
F
l
Figure B.1: An element with a flux passing though it, generating a magneto-motive force.
If a flux is passing through the element, a magneto-motive force (mmf ) will occur between two
points of the element. The mmf is determined by the magnetic field strength and the distance
between the two points:
[A turns]
F = Hl
(B.1)
The relationship between the magnetic field strength and the magnetic flux density is determined
by the permeability the field passes through:
H=
[A/m]
(B.2)
The flux density can be expressed as the flux on a given surface, in this case the cross sectional
area of the element in figure B.1:
B=
[T]
(B.3)
By combining equation B.2 and B.3 and inserting in equation B.1, a new expression for the mmf
can be found:
F=
(B.4)
B-26
The fraction in equation B.4 is termed the reluctance and is expressed as:
R=
l
A
A turns
]
Wb
(B.5)
With this stated equation B.4 can be reduced, stating the mmf as:
(B.6)
F = R
This equation can be seen as an equivalent to Ohms law, allowing a comparison to a electric
circuit where F can be seen as a voltage source with magnitude N i according to Amperes law.
can be seen as an equivalent to the current, and the reluctance R can be seen as an equivalent
to a resistance. A core with N turns and a current i in the winding is shown in figure B.2.
(t)
i(t)
+
u(t)
-
Figure B.2: Core with N turns and permeability .
By using the previously mentioned relation with electric circuit, a magnetic equivalent of figure
B.2 is shown in figure B.3.
R2
R1
R3
(t)
F = N i(t)
+
-
R4
Figure B.3: Magnetic equivalent circuit of figure B.2.
In the magnetic circuit shown in figure B.3 each side of the core has its own reluctance. By
unifying the four reluctances in one (Rc = R1 + R2 + R3 + R4 ), the circuit is greatly simplified,
and the following relation can be expressed:
N i(t) = Rc
(B.7)
Faradays law states that the electro-motive force (emf ) is given as:
u(t) = N
d(t)
dt
(B.8)
B.2 Inductance determination for the shunt reactor
B-27
Solving equation B.7 for and inserting in equation B.8 gives:

N 2 di(t)
Rc dt
di(t)
u(t) = L
dt
(B.9)
u(t) =
(B.10)
The inductance can now be found from the number of turns and the reluctance:
L=
N2
Rc
(B.11)
By use of this general theory, the inductances in the shunt reactor can be found.
B.2
Inductance determination for the shunt reactor
As described, a magnetic flux will flow in the core as a consequence of the current in the windings.
The mean flux path in the five limbed reactor core is shown in figure B.41 .
a (t)
b (t)
ua (t)
ub (t)
uc (t)
c (t)
Figure B.4: Mean flux path in the shunt reactor.
In order to ease the determination of the inductances, the following limitations are taken:
The cross sectional area of the flux path in the core limb air gap, is assumed equal to the
cross sectional area of the core
In order to reduce the cross sectional area of the flux path in the gap, air gaps are distributed
over the length of the limb, as shown in figure B.5. Furthermore, the distance elements,
consisting of a type of porcelain, are used between the limb segments. A cross section of a
limb segment with distance elements is shown in figure B.5 [15, ABB].
1
The reactor core is briefly described in section 3.6 on page 20.
B-28
Distance element
Core limb
Limb segment
Figure B.5: Mean flux path in the shunt reactor.
It is assumed that this distribution of the air gap and the distance elements minimizes the
flux path area in the gap to the area of the limb. The cross sectional area of the air gap is
therefore assumed to be equal to the cross sectional area of the core limb.
The permeability of the gap equals air
Due to the use of distance elements between the core elements the permeability in the gap
will obtain a different value than the one of air. This value has not been provided by ABB,
but it is assumed that it stays in the region just above air, as a lower value would increase
the cross sectional area flux path, and much higher value would diminish the purpose of
the air gap, as the reluctance would then approach the reluctance of the core.
The flux passes only through the mean flux path
As shown in figure B.4 the flux is represented by a mean path. As a consequence of this
representation, it is assumed that the flux only passes through the core, and not through
different paths in the air surrounding the core. This is reasonable as the reluctance in the
air is much greater than the one of the core.
The geometry of the core is symmetrical, this allows the core being represented by five different
flux paths (s1 , s2 , s3 , s4 and sg ) as shown in figure B.6.
A3
A2
b
A4
s3
Ag
sg
A1
s1
s2
s4
s2
s3
Figure B.6: Flux path in the shunt reactor.
The lengths and areas in figure B.4 is calculated in MatLab2 from the geometry described in the
2
The m-file is shown in appendix G.5.
B-29
system description, section 3.6 on page 20. The calculated values are listed in table B.1.
s1
s2
s3
s4
sg
1,716
1,203
0,783
2,003
0,2866
m
m
m
m
m
A1
A2
A3
A4
Ag
0,2827
0,1667
0,1667
0,1667
0,2827
m2
m2
m2
m2
m2
Table B.1: Flux paths and areas.
The permeability in the core can be found by use of the DC magnetization for the core shown
in figure B.7.
Induction B (Tesla), (x 10; Kilogauss)
B-H
DC MAGNETIZATION &
DC PERMEABILITY
M=5 0,30mm
Tested by Epstein test apparatus. Samples were sheared
in the rolling direction and subjected to Stress-relief
annealing at 800C (1472 F). Assumed density 7,65 kg/m 3
Frequency 50 Hz.
Magnetizing Force H (A/m)
Figure B.7: DC magnetization curve for the core.
This is done by calculating the peak value of B and read off the corresponding value for H. The
permeability can then be found through the relationship between B and H.
The magnetic flux density can be calculated by using the expression for the mmf :
F = N i = R
Knowing = BA and R =
N2
L ,
(B.12)
an expression for B is given:

B =
Li
N A
(B.13)
B-30
The peak magnetic flux density can now be found by use of the rated peak current in the winding:
Bpeak =
Bpeak =
L Ipeak
N A
2, 3 135, 8 2
= 1, 3087 T
1194 0, 2827
(B.14)
(B.15)
With this value for Bpeak , the magnetic field strength H is read off the B-H curve to 20 A/m.
The relative permeability can now be found from the relation between B and H.
(B.16)
B = H = 0 r H
B
1, 3087
r =
=
= 52074
0 H
4 107 20
(B.17)
The permeability in the iron core and the gap are listed in table B.2.
Permeability
Iron core
Gap
r
52074
1
Table B.2: Relative permeability.
With flux paths, cross sectional areas and permeability known, it is possible to make an equivalent
diagram for the reactor core, shown in figure B.4. This is done by transforming the flux path
into reluctances by use of equation B.5. The equivalent diagram is shown in figure B.8.
R3
R4
+
-
R3
R2
R2
R3
Rg
Rg
Rg
R1
R1
R1
+
-
Fa
R2
+
-
Fb
R2
R4
Fc
R3
Figure B.8: Equivalent diagram for the reluctances in the reactor.
In order to simplify the analysis of the reactor core, the equivalent diagram is simplified by
unifying reluctances. The simplified equivalent diagram for shunt reactor is show in figure B.9.
R2
R2
a
R1g
R1g
R1g
R343
2
+
-
B-31
+
-
Fa
+
-
Fb
R2
R343
Fc
R2
Figure B.9: Simplified equivalent diagram for the reactor core.
Where:
(B.18)
R343 = R3 + R4 + R3
(B.19)
R1g = R1 + Rg
To analyze the equivalent circuit, general methods from electric circuit analysis can be used. In
this case the mesh analysis is the most convenient, as the equivalent contains both multiple
sources and meshes [24, Johnson, p. 138 and p. 372]. The four mesh equations from figure B.9
are stated below:
0 = Fa + 1 (R343 + R1g ) 2 R1g
(B.20)
0 = Fc Fb + 3 (R2 + R1g + R2 + R1g ) 4 R1g 2 R1g
(B.22)
0 = Fb Fa + 2 (R2 + R1g + R2 + R1g ) 3 R1g 1 R1g
(B.21)
0 = Fc + 4 (R343 + R1g ) 3 R1g
(B.23)
These equations can be simplified:

(B.24)
Fa = +1 (R343 + R1g ) 2 R1g
(B.25)
Fa Fb = 1 R1g + 2 (2 R2 + 2 R1g ) 3 R1g
(B.26)
Fb Fc = 2 R1g + 3 (2 R2 + 2 R1g ) 4 R1g
(B.27)
Fc = 3 R1g + 4 (R343 + R1g )
These four equations can be arranged in a matrix equation:
Fa
Fa Fb
Fb Fc
Fc
R343 + R1g
R1g
0
0
R1g
2R2 + 2R1g
R1g
0
0
R1g
2R2 + 2R1g
R1g
0
0
R1g
R343 + R1g
1
2
3
4
(B.28)
To determine the inductances in and between each phase, the flux in each limb is required.
According to the equivalent diagram in figure B.9, the flux in each limb can be expressed by the
four mesh fluxes:
a = 2 1
(B.29)
c = 4 3
(B.31)
b = 3 2
(B.30)
B-32
In order to use this fact, the equation for each mesh flux must be found from equation B.28.
This is done by taking the inverse of the reluctance matrix, creating a permeance matrix:
1234 = R1 F
(B.32)
(B.33)
1234 = PF
The inversion of the reluctance matrix shown in equation B.28, will give a long algebraic equation
and is therefore performed through a numerical Gauss-Jordan Elimination [25, Kreyszig, p. 350]
in MatLab:
AA1 = I
(B.34)
The inversion of the reluctance matrix will give a 4 4 matrix, and the flux equation shown
below:
1
2
3
4
P11
P21
P31
P41
P12
P22
P32
P42
P13
P23
P33
P43
P14
P24
P34
P44
Fa
Fa Fb
Fb Fc
Fc
(B.35)
This matrix can be reduced to a 4 3 matrix by multiplying the 4 1 F-matrix, and afterward
isolate a new F-matrix:
1
2
3
4
P21 P11
P22 P12
P23 P13
P24 P14
P31 P21
P32 P22
P33 P23
P34 P24
P41 P31
P42 P32
P43 P33
P44 P34
Fa
Fb
(B.36)
Fc
Equation B.29, B.30 and B.31 can now be used to set up an expression for each limb flux:
"
a
b
c
"
(P22 P12 ) (P21 P11 )

(P23 P13 ) (P22 P12 )
(P23 P13 ) (P22 P12 )
(P32 P22 ) (P31 P21 )

(P33 P23 ) (P32 P22 )
(P33 P23 ) (P32 P22 )
(P42 P32 ) (P41 P31 )

(P43 P33 ) (P42 P32 )
(P43 P33 ) (P42 P32 )
#"
Fa
Fb
Fc
(B.37)
The inductance matrix can now be found by use of equation B.11, as the permeance is the
reciprocal of the reluctance:
N2
R
L = N 2P
L =
B.3
(B.38)
(B.39)
Calculation of the inductances in the shunt reactor
Calculations of the inductance matrix are performed by the above-mentioned method in MatLab3 .
The results from the calculations are shown below:
1, 767
358 106 256 106
Laa Mab Mac
6
L = Mba Lbb Mbc = 358 10
1, 767
358 106
256 106 358 106
1, 767
Mca Mcb Lcc
The m-files are shown in appendix G.5, G.6 and G.7.
(B.40)
B.3 Calculation of the inductances in the shunt reactor
B-33
Due to the symmetry of the core: Mab = Mba , Mac = Mca and Mbc = Mcb . Furthermore,
Mba = Mbc as phase b is placed on the symmetrical axis.
The calculated self inductances are compared with the designed and the measured values provided
by ABB, in table B.3.
La
Lb
Lc
Calculated
1,767 H
1,767 H
1,767 H
Designed
2,300 H
2,300 H
2,300 H
Measured
2,307 H
2,304 H
2,307 H
Table B.3: Self inductances in the shunt reactor[7, ABB Reactor].
The calculated self inductance in table B.3 does vary from the values stated by ABB. This
variation is assumed to be inherited from the following assumptions:
The relative permeability in the gap equals air.
The cross sectional area of the gap equals the cross section of the limb.
As both the permeability and the cross sectional area of the gap will have larger values than
assumed, the reluctance in the gap will, according to equation B.5, obtain a lower value. In order
to find a suitable estimate of the mutual inductances, the reluctances in the gap are adjusted to
give a inductance per phase of 2,3 H. When this is the case, the reluctance in the gap can be
assumed to be correct. The reluctance can be adjusted, by changing both the cross section in
the gap and/or the relative permeability in the gap. The results will be the same whether the
cross sectional area or the permeability is changed, as long as the reluctance obtains the same
value. In this case the relative permeability in the gap is changed to a value of 1,3021, while the
cross sectional area of the gap is kept equal to the limb cross sectional area. The new inductance
matrix is:
2, 300
607 106 434 106
Laa Mab Mac
L = Mba Lbb Mbc = 607 106

2, 300
607 106
434 106 607 106
2, 300
Mca Mcb Lcc
(B.41)
The new results are listed in table B.4.
La
Lb
Lc
Calculated
2,300 H
2,300 H
2,300 H
Designed
2,300 H
2,300 H
2,300 H
Measured
2,307 H
2,304 H
2,307 H
Table B.4: Recalculated values of the self inductances in the shunt reactor[7, ABB Reactor].
Now the calculated self inductance matches with the values stated by ABB. The reason this
correction is made is to obtain a more credible estimate of the mutual inductances, as the
measured values for the self inductances will be used in simulations.
B-34
The mutual inductances in the reactor have not been calculated nor tested by ABB. However,
ABB states that the value usually stays in the region of 1-2 % of the phase values [15, ABB]. The
caluclated values for the mutual inductances are listed along with the 1 %-value of the designed
inductance in table B.5.
Mab
Mac
Mbc
Calculated [% of phase value]

-607 H [0,03 %]
-434 H [0,02 %]
-607 H [0,03 %]
Designed [% of phase value]

-32,0 mH [1,39 %]
-23,0 mH [1,00 %]
-32,0 mH [1,39 %]
Table B.5: Mutual inductances in the shunt reactor[15, ABB].
The calculated values of the mutual inductance do not stay in the expected region of 1-2 % as
ABB stated. Instead the percentage value stay in the 0,02-0,03 % region. The calculated value
have though the internally expected relationship, as Mab equals Mbc which are larger than Mac .
The factor affecting the mutual inductances the most is the permeability of the core, which is
found to have a very high value, and thereby causing low mutual inductances. Whether, the
information about the mutual inductances of 1-2% is wrong, or miscalculations are present is
unexplained. As this is the case, mutual inductances in the region of 1-2% will be used, for worst
case. For Mac values of 1% will be used, and for Mbc and Mab values of 1,39% will be used.
The difference in the values of the mutual inductances is due to the symmetrical structure of the
reactor.
B.4
Summary
In the previous appendix, the determination of the inductances in the shunt reactor was described.
The inductances were determined and are listed in table B.4 and B.5. For simulations the
inductances in table B.6 will be used.
La
Lb
Lc
Mab
Mac
Mbc
Inductances
2,307 H
2,304 H
2,307 H
-32,0 mH
-23,0 mH
-32,0 mH
Table B.6: Self and mutual inductances in the shunt reactor.
Appendix C
Measurements of transmission line

switching
In order to determine the scope of the problem concerning switching overvoltages, measurements
are made at the switching instant, on the 150 kV, 20 kV and 0,4 kV level. Measurements on the
150 kV and 20 kV level are performed by staff from Energinet.dk and measurements on the 0,4
kV level are performed by the project group.
C.1
General test procedure
A simplified single-line diagram of the TH3-FER3 section is shown in figure C.1.

TH3
FER3
60 kV
150 kV line
40 Mvar
Sec 3
Sec 2
400 kV
20 kV
0,4 kV
Sec 1
Figure C.1: A simplified single-line diagram for the Tinghj-Ferslev section and stations, showing the
most important components.
In order to record the switching transients, the following four steps are taken:
1. Disconnection of TH3
The line is switched-off in Tinghj.
2. Disconnection of FER3
The line is switched-off in Ferslev.
Measurements during switch-off are performed.
3. Reconnection of FER3
The line is switched-on in Ferslev.
Measurements during switch-on are performed.
4. Reconnection of TH3
The line is switched-on in Tinghj.
C-36
C.2
Measurements on 150 kV and 20 kV level
C.2.1
Purpose
This experiment is performed to investigate the switching transients of the 150 kV grid segment
FER3-TH3.
C.2.2
Instruments
The instruments used in the experiment are listed in table C.1.

Equipment
2 Transient recorder
2 PC
3 150 kV voltage transformer
Type
Omicron CMC 256-6
IBM
ASEA EMFC 145
Information
See appendix D
With EnerLyzer installed
0,1
/
kV
Transformation ratio: 160
3
3
3 150 kV current transformer

3 20 kV voltage transformer
ASEA IMBE 145 A3
Transformation ratio: 1000/1 A

20
/ 0,13 kV
Transformation ratio:
3
Table C.1: Instruments used for transient measurements.
C.2.3
Procedure
This test is performed by engineers from Energinet.dk. The principle of the setup is shown in
figure C.2.
TH3
FER3
150 kV section
60 kV
S
R
Sec 3
40 Mvar
Sec 2
20 kV
Sec 1
150 kV voltage
measurement
150 kV current
measurement
OMICRON
+
-
CMC 256-6
20 kV voltage
measurement
Trigger (400 Vdc)
ch9
ch8
ch7
ch6
ch5
ch4
0V
OMICRON
CMC 256-6
ch3
ch2
ch1
Figure C.2: Test setup for transient measurements.
T
S
R
C.2 Measurements on 150 kV and 20 kV level
C.2.4
C-37
Data
The voltage behaviour on the 150 kV line after a disconnection is shown in figure C.3.
Phase a [kV]
200
100
0
100
200
0.5
1.5
2.5
0.5
1.5
2.5
0.5
1.5
Time [s]
2.5
Phase b [kV]
200
100
0
100
200
Phase c [kV]
200
100
0
100
200
Figure C.3: Measured voltages on 150 kV level after disconnection.
The voltage behaviour on the 150 kV line after a reconnection (t = 0, 034 s) is shown in figure
C.4.
150
100
Voltage [kV]
50
50
100
150
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0.065
Time [s]
Figure C.4: Measured voltages on 150 kV level after reconnection.
The current behaviour on the 150 kV line after a reconnection (t = 0, 034 s) is shown in figure
C.5 for a short time period.
C-38
Phase a [A]
2000
1000
0
1000
2000
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.02
0.04
0.06
0.08
0.1
Time [s]
0.12
0.14
0.16
0.18
0.2
Phase b [A]
2000
1000
0
1000
2000
Phase c [A]
1000
0
1000
2000
Figure C.5: Measured current on 150 kV level after reconnection.
The current behaviour on the 150 kV line after a reconnection (t = 0, 034 s) is shown in figure
C.6 for a longer time period.
Phase a [A]
500
500
0.5
1.5
2.5
0.5
1.5
2.5
0.5
1.5
Time [s]
2.5
Phase b [A]
500
500
Phase c [A]
500
500
Figure C.6: Measured current on 150 kV level after reconnection.
The voltage behaviour on the 20 kV line after a reconnection (t = 0, 034 s) is shown in figure
C.7.
C.2 Measurements on 150 kV and 20 kV level
C-39
25
20
15
10
Voltage [kV]
10
15
20
25
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0.065
Time [s]
Figure C.7: Measured voltages on 20 kV level after reconnection.
C.2.5
Results
150 kV
The voltage behaviour on the 150 kV line after disconnection is shown in figure C.3, and shows
how the voltages slowly decreases.
The peak overvoltages are evaluated from the voltage waveforms in TransView1 :
Phase
R
S
T
Prior peak voltage

134 kV
134 kV
134 kV
Switching peak voltage

159 kV
154 kV
-162 kV
Percentage
18,7%
14,9%
20,9%
Table C.2: The peak voltages at the 150 kV level, before and after switch-on.
In figure C.8 are the results from the Fourier analysis2 of the first period after switch-on shown.
1
2
TransView is a part of the Enerlyzer software package.

The Fourier analysis is performed in MatLab with the m-file shown in appendix G.10.
C-40
10

150 kV level
5
0
0
1000
2000
3000
4000
5000
6000
1000
2000
3000
4000
5000
6000
1000
2000
3000
4000
Frequency [Hz]
5000
6000
10
5
0
0
10
5
0
0
Figure C.8: Fourier analysis of the 150 kV waveform, of harmonic frequencies up to 6,5 kHz.
Beside the fundamental, the Fourier analysis for the 150 kV level shows two large frequency
bands, one around 350 Hz and one around 3,8 kHz.
The peak inrush currents, during reconnection of the 150 kV line, are shown in figure C.5 and
the values are listed in table C.3.
Phase
R
S
T
Peak current
1423 A
1195 A
1773 A
Table C.3: Peak magnitude of the inrush current on the 150 kV level.
In figure C.6 are the currents in each phase, shown over a longer time period. At the reconnection
moment, can it be seen, that the currents obtain a slowly decreasing DC-offset.
20 kV
Evaluation of overvoltages in TransView yields:
Phase
R
S
T
Prior peak voltage

16,8 kV
16,8 kV
16,9 kV

-17,4 kV
17,3 kV
17,3 kV
Percentage
3,6%
3,0%
3,0%
Table C.4: The peak voltages at the 20 kV level, before and after switch-on.
In figure C.9 are the results from the Fourier analysis3 of the first period after switch-on shown.
3
C.3 Measurements on 0,4 kV level
C-41
10

20 kV level
5
0
0
1000
2000
3000
4000
5000
6000
1000
2000
3000
4000
5000
6000
1000
2000
3000
4000
Frequency [Hz]
5000
6000
10
5
0
0
10
5
0
0
Figure C.9: fft analysis of the 20 kV waveform, of harmonic frequencies up to 6,5 kHz
The Fourier analysis of the 20 kV level shows harmonics around 350 Hz. On the 20 kV level is
higher frequencies not present.
C.2.6
Summary
Voltage and current waveforms were measured, during switching on/off at the 150 kV level.
Measurements on the 20 kV level, show that the overvoltages caused by reconnection on the
150 kV grid, propagate to the 20 kV level, though considerably damped. Furthermore, the high
frequencies experienced at the 150 kV level did not propagate to the 20 kV level.
C.3
C.3.1
Measurements on 0,4 kV level

Purpose
This experiment is performed to investigate the propagation of overvoltages caused by switching

of 150 kV grid segment to 0,4 kV level.
C.3.2
Instruments
The instruments used in the experiment are listed in table C.5.

Equipment
Transient recorder
PC
Type
Omicron CMC 256-6
Dell Inspiron 8200
Information
See appendix D
AAU #: 55657
With EnerLyzer installed
Table C.5: Instruments used for transient measurements.
C-42
C.3.3
Procedure
The setup is connected as shown in figure C.10. As only three of the ten inputs are necessary,
input #1, #3 and #5 are used, as those are galvanic isolated from each other.
Three phase outlet
a
b
c
n
Parallel
OMICRON
+
-
CMC 256-6
PC
Figure C.10: Test setup for transient measurements.
The hardware is configured4 to a measurement range of 600 V with a sampling rate of 28,4 kHz.
The transient recording is started manually, which gives a maximum recording length of 11,72 s
with the chosen sampling rate and three channels used. The switching of the 150 kV grid is then
performed manually within this time.
C.3.4
Data
The voltage behaviour on the 0,4 kV level after a reconnection (t = 0, 0378 s) is shown in figure
C.11.
4
The configuration of CMC 256-6 is described in appendix D.
C.3 Measurements on 0,4 kV level
C-43
500
400
300
200
Voltage [V]
100
100
200
300
400
500
0.04
0.045
0.05
0.055
Time [s]
0.06
0.065
0.07
Figure C.11: Measured voltages on the 0,4 kV level after reconnection of the 150 kV section.
C.3.5
Results
As shown in figure C.11, do the switching overvoltages on the 150 kV propagate to the 400 V
level. In table C.6 are the peak values of the overvoltages on the 0,4 kV level listed.
Phase
a
b
c
Prior peak voltage

322,1 V
316,1 V
316,8 V

381,5 V
339,0 V
331,2 V
Percentage
18,4%
7,2%
4,5%
Table C.6: The peak voltages on the 0,4 kV level, after switch-on.
At the reconnection instant, switching harmonics are visible, shown and evaluated in figure C.12
with a Fourier analysis5 .
5
C-44
10

0,4 kV level
5
0
0
1000
2000
3000
4000
5000
6000
1000
2000
3000
4000
5000
6000
1000
2000
3000
4000
Frequency [Hz]
5000
6000
10
5
0
0
10
5
0
0
Figure C.12: fft analysis of the 0,4 kV waveform, of harmonic frequencies up to 6,5 kHz
C.3.6
Summary
Voltage and current waveforms were measured, during switching on at the 150 kV level. The test
indicates that switching transients on the 150 kV grid, propagate to the 0,4 kV level. Fourier
analysis shows that the high frequency experienced at the 150 kV, is filtered.
C.4
Conclusion & discussion
The measurements on the different voltage levels, show how switching transients on the 150 kV
level propagates to lower voltage levels. The magnitude of the propagated switching overvoltages
is though damped on the lower levels, from 150 kV level 20,9%, to a maximum of 3,6% for 20
kV and a maximum of 18,4% for 0,4 kV.
A collection of the voltage waveforms is given in figure C.13.
Voltage [kV]
C.4 Conclusion & discussion
C-45
100
0
100
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0.065
0.05
0.055
0.06
0.065
0.05
0.055
0.06
0.065
Time [s]
Voltage [kV]
20
10
0
10
20
0.03
0.035
0.04
0.045
Time [s]
Voltage [V]
500
500
0.03
0.035
0.04
0.045
Time [s]
Figure C.13: Voltage waveforms recorded in Ferslev at the moment of switch-on. From top-to-bottom
150 kV, 20 kV, 0,4 kV.
A high-frequency voltage component, is measured for a short period on the 150 kV level, to stays
around 3,8 kHz. This frequency is filtered through the transformers and is not present at the
lower voltage levels.
From the Fourier analysis voltage components around 350 Hz was also measured, which takes
slightly longer time to be damped than the high frequencies.
A single spike in measurements at 0,4 kV, gives 18,4% overshoot. A reason for this high overshoot
could be that the ground obtains a potential for a short period after the switch-on instant, caused
by a zero sequence current. The proposed current path of the zero sequence current is shown in
figure C.14.
Grid
150 kV
60 kV
20 kV
0,4 kV
Figure C.14: The proposed zero sequence current path.
The 150 kV level is grounded in the neutral-point, as is the 0,4 kV neutral-point. It is assumed
that the 20 kV level is grounded through a Peterson coil. This could explain why the spike is not
present at the 20 kV level, and why the overvoltages have a lower values than the overvoltages
at the 0,4 kV level.
Throughout the three voltage levels, is the 7th harmonic the dominant one, with 7% of the
fundamental at 150 kV, 4,4% at 20 kV and 5% at 0,4 kV.
Appendix D
Transient recording using Omicron

CMC 256-6
In this appendix is the device for transient recording described. This appendix is written by use
of [26, CMC 256 user manual] and [27, EnerLyser user manual].
D.1
Hardware
The front panel of the CMC 256-6 is shown in figure D.1.

Voltage ouput
aux dc
Analog dc input
Binary output
0..20mA
OMICRON
CMC 256-6
Current ouput A
0..10V
+
!
Binary/analog input
10
Current ouput B
Figure D.1: Front panel of Omicron CMC 256-6.
The Binary/analog input is galvanically isolated in pairs from each other.
D.2
Configuration
In order to perform transient voltage measurements, the Binary/analog input shown in figure
D.1 is used. The hardware configuration is set in the software EnerLyzer. For analog
measurements five different measurement ranges and three different sampling rates can be chosen
as listed in table D.1 and table D.2.
100 mV
Measurement ranges
1V
10 V
100 V
600 V
Table D.1: Measurement ranges for analog measurements using Omicron CMC 256.
3,16 kHz
Sampling rates
9,48 kHz
28,4 kHz
Table D.2: Sampling rates for analog measurements using Omicron CMC 256.
For the 0,4 kV measurements, the range is chosen to 600 V and the sampling rate is chosen to
28,4 kHz.
D-48
D.3
Transient recording using Omicron CMC 256-6
Accuracy
The accuracy of the sampled values depends on the chosen measurement range as shown in table
D.3. As the measurement is on the 0,4 kV level, an accuracy of 0,5% is guaranteed.
Accuracy of the sampling value
typical
0.2 %
guaranteed
0.5 %
Table D.3: Accuracy of the sampling value for measurement range of 600 V, 100 V, 10 V, and 1 V.
The accuracy is relative to the full scale measurement.
D.4
Harmonics
The EnerLyzer software, has an inbuilt harmonic analyzer, described in the user manual:
Harmonics are calculated by means of a full-cycle DFT (Discrete Fourier Transformation) and
are always r.m.s. values. The measuring window is placed to the left of the reference instant
(e.g. cursor position) and its length corresponds to one period of the nominal frequency TN .
Appendix E
The single line diagrams for the two substations are shown in this appendix. In the last section,
the symbols used in the single-line diagrams are breifly explained.
E.1
Ferslev substation
1600
2500 1000 1600
Disp.
11.8
11.8
1600
1600 1000 2500 1600
2500 1000 1600

11.8
1600 1000 2500 1600

11.8
TH
1 x 1200 AI
1 x 772
1 x 1200 AI
Disp. KT52
2500
3150 2000 2500
2500 1600 1600

11.8
ADL1
1 x 454
600
1600
11.8
11.8
1600
2500 2000 3150 2500
400 kV
1600 1600 2500 1600

11.8
KT51
ABB
1600
2500 1600 1600

11.8
400 MVA
1600 1600 2500 1600

11.8
MOS
1 x 772
1600
300
11.8
ZL1
1600
Siemens
300
2500 1000 1600
80 MVA
1600 1000 2500 1600

11.8
2500 1000 1600

11.8
1600 1000 2500 1600

11.8
60 kV
Disp. KT32
s
S1
S2
KT31
Strmberg
80 MVA
Figure E.1: Single-line diagram of transformer station in Ferslev (FER3).
E-50
E.2
Tinghj substation
S1
7.5
1250
2000
800
1250
ABB
40 Mvar
Disp.
FER
1 x 1200 AI
1 x 772
1 x 1200 IAI
Disp.
7.5
1250
2000
800
1250
ABB
40 Mvar
7.5
60 kV
300 2000
1250
HNB
1 x 1200 AI
Disp.
KT31
Strmberg
80 MVA
Figure E.2: Single-line diagram of transformer station in Tinghj (TH3).
1 x 772
E.3 Explanation of symbols in single-line diagrams
E.3
Explanation of symbols in single-line diagrams

Busbar.
Busbar connection with disconnector.
1250
Earthing switch of busbar.

1600
2500
Disconnector.
Circuit breaker.
11.8
Capacitive voltage transformer.

Inductive voltage transformer.
1000
1600
Current transformer.
Earthing switch.
Surge arrester.
Cable connection to station.
Overhead line connection to station.
Shunt reactor.
Autotransformer.
Transformer.
Figure E.3: Explanations of symbols used in single-line diagrams.
E-51
Appendix F

The matrices written by the PSCAD line constant program are shown in this appendix.
Figure F.1 shows the output file for the overhead line (section 2) configured as described
in section 5.1.4 on page 62.
Figure F.2 shows the output file for the underground cable (section 3) configured as
described in section 5.1.3 on page 59.
Figure F.2 shows the output file for the corrected underground cable ( section 3) corrected
as described in section 6.3 on page 90.
F-54
SECTION 2
Corresponds to PSCAD simulation in chapter 5
SERIES IMPEDANCE MATRIX (Z) [ohms/m]:

0.908318063E-04,0.704128248E-03
0.472961756E-04,0.296014235E-03
0.472961756E-04,0.296014235E-03
0.915579651E-04,0.703353087E-03
0.472962763E-04,0.297286539E-03
0.476608051E-04,0.325722308E-03
0.466072857E-04,0.303917061E-03
0.469641348E-04,0.259125595E-03
0.472962763E-04,0.297286539E-03
0.476608051E-04,0.325722308E-03
0.915579651E-04,0.703353087E-03
0.469635548E-04,0.256978115E-03
0.466072857E-04,0.303917061E-03
0.469641348E-04,0.259125595E-03
0.469635548E-04,0.256978115E-03
0.344722914E-03,0.756378016E-03
SHUNT ADMITTANCE MATRIX (Y) [mhos/m]:

0.132230000E-09,0.243401255E-08
0.000000000E+00,-.356286202E-09
0.000000000E+00,-.356286202E-09
0.132230000E-09,0.251071841E-08
0.000000000E+00,-.366967625E-09
0.000000000E+00,-.523915305E-09
0.000000000E+00,-.453693548E-09
0.000000000E+00,-.182825993E-09
0.000000000E+00,-.366967625E-09
0.000000000E+00,-.523915305E-09
0.132230000E-09,0.251126867E-08
0.000000000E+00,-.170187185E-09
0.000000000E+00,-.453693548E-09
0.000000000E+00,-.182825993E-09
0.000000000E+00,-.170187185E-09
0.000000000E+00,0.206434327E-08
LONG-LINE CORRECTED SERIES IMPEDANCE

0.171898642E+01,0.133278577E+02
0.894985334E+00,0.560274895E+01
0.894987022E+00,0.562683206E+01
0.881920983E+00,0.575239048E+01
MATRIX [ohms]:
0.894985334E+00,0.560274895E+01
0.173271281E+01,0.133131305E+02
0.901874291E+00,0.616505762E+01
0.888665793E+00,0.490452437E+01
0.894987022E+00,0.562683206E+01
0.901874291E+00,0.616505762E+01
0.173271268E+01,0.133131306E+02
0.888654969E+00,0.486387573E+01
0.881920983E+00,0.575239048E+01
0.888665793E+00,0.490452437E+01
0.888654969E+00,0.486387573E+01
0.652445415E+01,0.143171389E+02
LONG-LINE CORRECTED SHUNT ADMITTANCE

0.250354656E-05,0.460775500E-04
0.617432785E-10,-.674454238E-05
0.594313587E-10,-.694674825E-05
-.125759253E-09,-.858857234E-05
MATRIX [mhos]:
0.617432785E-10,-.674454238E-05
0.250355759E-05,0.475297408E-04
0.440640851E-10,-.991781750E-05
-.445111647E-11,-.346083512E-05
0.594313587E-10,-.694674825E-05
0.440640851E-10,-.991781750E-05
0.250355734E-05,0.475401582E-04
0.108056924E-11,-.322157456E-05
-.125759253E-09,-.858857234E-05
-.445111646E-11,-.346083512E-05
0.108056924E-11,-.322157456E-05
0.766497849E-09,0.390794778E-04
Phase A
Phase B
Phase C
----------------------------------------------------SEQUENCE COMPONENT DATA @

50.00 Hz:
----------------------------------------------------SEQUENCE TRANSFORM MATRIX:
0.577350269E+00,0.000000000E+00
0.577350269E+00,0.000000000E+00
0.577350269E+00,0.000000000E+00
0.577350269E+00,0.000000000E+00
-.288675135E+00,-.500000000E+00
-.288675135E+00,0.500000000E+00
0.577350269E+00,0.000000000E+00
-.288675135E+00,0.500000000E+00
-.288675135E+00,-.500000000E+00
SEQUENCE IMPEDANCE MATRIX (Zsq) [ohms/m]:

0.186151417E-03,0.131629353E-02
-.730861826E-06,-.943222438E-05
0.370319963E-08,-.943228249E-05
0.438981598E-04,0.397270446E-03
-.730861826E-06,-.943222438E-05
-.733565143E-06,0.196397257E-04
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.370319963E-08,-.943228249E-05
0.735564908E-06,0.196396095E-04
0.438981598E-04,0.397270446E-03
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
SEQUENCE ADMITTANCE MATRIX (Ysq) [mhos/m]:

0.132230000E-09,0.165388712E-08
0.292461573E-11,0.284357996E-10
-.292461573E-11,0.284357996E-10
0.132230000E-09,0.290105625E-08
0.292461573E-11,0.284357996E-10
0.632576761E-11,-.133852592E-09
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
-.292461573E-11,0.284357996E-10
-.632576761E-11,-.133852592E-09
0.132230000E-09,0.290105625E-08
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
Zero
Positive
Negative
Figure F.1: The figure shows the impedance and admittance matrix for the overhead line section,
simulated in chapter 5.
----------------------------------------------------PHASE DOMAIN DATA @

50.00 Hz:
-----------------------------------------------------
PHASE DOMAIN DATA @

50.00 Hz:
-----------------------------------------------------
SECTION 3
Corresponds to PSCAD simulation in Chapter 5

0.492649635E-04,0.633515432E-03
0.751203844E-04,0.686881613E-03
0.492649635E-04,0.633515432E-03
0.233613922E-03,0.633412694E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.751239380E-04,0.686892684E-03
0.492685171E-04,0.633526503E-03
0.492139199E-04,0.580160462E-03
0.492139199E-04,0.580160462E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492685171E-04,0.633526503E-03
0.233617476E-03,0.633423765E-03
0.492139199E-04,0.580160462E-03
0.492139199E-04,0.580160462E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492139199E-04,0.580160462E-03
0.492139199E-04,0.580160462E-03
0.751239380E-04,0.686892684E-03
0.492685171E-04,0.633526503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492139199E-04,0.580160462E-03
0.492139199E-04,0.580160462E-03
0.492685171E-04,0.633526503E-03
0.233617476E-03,0.633423765E-03

0.000000000E+00,0.635748641E-07
0.000000000E+00,-.635748641E-07
0.000000000E+00,-.635748641E-07
0.000000000E+00,0.528921349E-06
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.635748641E-07
0.000000000E+00,-.635748641E-07
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,-.635748641E-07
0.000000000E+00,0.528921349E-06
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.635748641E-07
0.000000000E+00,-.635748641E-07
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,-.635748641E-07
0.000000000E+00,0.528921349E-06

0.145526307E+01,0.136712923E+02
0.831085123E+00,0.125565016E+02
0.911457828E+00,0.114323377E+02
0.836067738E+00,0.114384935E+02
0.911457828E+00,0.114323377E+02
0.836067738E+00,0.114384935E+02
MATRIX [ohms]:
0.831085123E+00,0.125565016E+02
0.461895548E+01,0.125885485E+02
0.836067738E+00,0.114384935E+02
0.760665677E+00,0.114445052E+02
0.836067738E+00,0.114384935E+02
0.760665677E+00,0.114445052E+02
0.911457828E+00,0.114323377E+02
0.836067738E+00,0.114384935E+02
0.145533208E+01,0.136715129E+02
0.831152685E+00,0.125567226E+02
0.911493777E+00,0.114323941E+02
0.836103306E+00,0.114385501E+02
0.836067738E+00,0.114384935E+02
0.760665677E+00,0.114445052E+02
0.831152685E+00,0.125567226E+02
0.461902160E+01,0.125887699E+02
0.836103306E+00,0.114385501E+02
0.760700863E+00,0.114445620E+02
0.911457828E+00,0.114323377E+02
0.836067738E+00,0.114384935E+02
0.911493777E+00,0.114323941E+02
0.836103306E+00,0.114385501E+02
0.145533208E+01,0.136715129E+02
0.831152685E+00,0.125567226E+02
0.836067738E+00,0.114384935E+02
0.760665677E+00,0.114445052E+02
0.836103306E+00,0.114385501E+02
0.760700863E+00,0.114445620E+02
0.831152685E+00,0.125567226E+02
0.461902160E+01,0.125887699E+02

0.655855576E-06,0.133523572E-02
-.492205859E-05,-.133520987E-02
-.267528142E-11,-.272280742E-10
-.520540718E-07,0.508068918E-08
-.267527888E-11,-.272280733E-10
-.520540718E-07,0.508068916E-08
MATRIX [mhos]:
-.492205859E-05,-.133520987E-02
0.497152012E-04,0.112169078E-01
-.520540718E-07,0.508068918E-08
0.974687681E-05,0.100660699E-03
-.520540718E-07,0.508068916E-08
0.974687681E-05,0.100660699E-03
-.267528143E-11,-.272280740E-10
-.520540718E-07,0.508068918E-08
0.655855576E-06,0.133523572E-02
-.492205957E-05,-.133520987E-02
-.267537234E-11,-.272282147E-10
-.520543411E-07,0.508085876E-08
-.520540718E-07,0.508068918E-08
0.974687681E-05,0.100660699E-03
-.492205957E-05,-.133520987E-02
0.497158405E-04,0.112169097E-01
-.520543411E-07,0.508085876E-08
0.974720432E-05,0.100661220E-03
-.267527889E-11,-.272280731E-10
-.520540718E-07,0.508068916E-08
-.267537233E-11,-.272282148E-10
-.520543411E-07,0.508085876E-08
0.655855576E-06,0.133523572E-02
-.492205957E-05,-.133520987E-02
-.520540718E-07,0.508068916E-08
0.974687681E-05,0.100660699E-03
-.520543411E-07,0.508085876E-08
0.974720432E-05,0.100661220E-03
-.492205957E-05,-.133520987E-02
0.497158405E-04,0.112169097E-01
Phase A

50.00 Hz:
0.577350269E+00,0.000000000E+00
0.577350269E+00,0.000000000E+00
0.577350269E+00,0.000000000E+00
Phase B
Phase C
0.577350269E+00,0.000000000E+00
-.288675135E+00,-.500000000E+00
-.288675135E+00,0.500000000E+00
0.577350269E+00,0.000000000E+00
-.288675135E+00,0.500000000E+00
-.288675135E+00,-.500000000E+00

-.264426031E-04,-.279647560E-04
0.226412156E-03,0.186494929E-02
-.263721310E-04,0.635700150E-04
0.787230445E-04,0.711188508E-04
-.264426031E-04,-.279647560E-04
0.198107463E-04,0.368452469E-04
0.147656267E-03,0.175826415E-02
-.154180342E-04,-.887846026E-05
0.181990735E-07,0.177886803E-04
-.154145485E-04,-.887857927E-05
0.181990735E-07,0.177886803E-04
-.154145485E-04,-.887857927E-05
-.263721310E-04,0.635700150E-04
-.726783698E-04,-.545979167E-04
0.787230445E-04,0.711188508E-04
0.153979893E-04,-.891317920E-05
0.153963495E-04,-.891010099E-05
0.153963495E-04,-.891010099E-05
0.147656267E-03,0.175826415E-02
0.153979893E-04,-.891317920E-05
-.154180342E-04,-.887846026E-05
0.279250535E-03,0.184714502E-02
0.109622431E-04,-.724533102E-04
0.418325381E-04,0.190529766E-04
0.181990735E-07,0.177886803E-04
0.153963495E-04,-.891010099E-05
-.154145485E-04,-.887857927E-05
0.418325381E-04,0.190529766E-04
0.131554178E-03,0.532975952E-04
0.110168404E-04,-.190872695E-04
0.181990735E-07,0.177886803E-04
0.153963495E-04,-.891010099E-05
-.154145485E-04,-.887857927E-05
0.109622431E-04,-.724533102E-04
0.418871353E-04,0.724190174E-04
0.131554178E-03,0.532975952E-04

0.115981477E-06,-.881535582E-07
0.000000000E+00,0.176307116E-06
-.115981477E-06,-.881535582E-07
0.000000000E+00,0.239881980E-06
0.115981477E-06,-.881535582E-07
-.171038924E-06,-.563661261E-07
0.000000000E+00,-.211916214E-07
0.183524824E-07,0.105958107E-07
0.000000000E+00,-.211916214E-07
0.183524824E-07,0.105958107E-07
0.000000000E+00,-.211916214E-07
0.183524824E-07,0.105958107E-07
-.115981477E-06,-.881535582E-07
0.171038924E-06,-.563661261E-07
0.000000000E+00,0.239881980E-06
-.183524824E-07,0.105958107E-07
-.183524824E-07,0.105958107E-07
-.183524824E-07,0.105958107E-07
0.000000000E+00,-.211916214E-07
-.183524824E-07,0.105958107E-07
0.183524824E-07,0.105958107E-07
0.000000000E+00,0.331422611E-06
0.134333959E-06,0.987493688E-07
-.134333959E-06,0.987493688E-07
0.000000000E+00,-.211916214E-07
-.183524824E-07,0.105958107E-07
0.183524824E-07,0.105958107E-07
-.134333959E-06,0.987493688E-07
0.000000000E+00,0.394997475E-06
0.134333959E-06,0.351745048E-07
0.000000000E+00,-.211916214E-07
-.183524824E-07,0.105958107E-07
0.183524824E-07,0.105958107E-07
0.134333959E-06,0.987493688E-07
-.134333959E-06,0.351745048E-07
0.000000000E+00,0.394997475E-06
Zero
Positive
Negative
Zero
Positive
Negative
F-55
Figure F.2: The figure shows the impedance and admittance matrix for the underground cable section,
simulated in chapter 5.
-----------------------------------------------------
F-56
PHASE DOMAIN DATA @

50.00 Hz:
-----------------------------------------------------
SECTION 3
Corresponds to the corrected cable section simulated in chapter 6

0.492661464E-04,0.634845066E-03
0.758331451E-04,0.682683339E-03
0.492661464E-04,0.634845066E-03
0.126386606E-03,0.634588855E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.758366987E-04,0.682694410E-03
0.492697000E-04,0.634856137E-03
0.492139199E-04,0.580160462E-03
0.492139199E-04,0.580160462E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492697000E-04,0.634856137E-03
0.126390160E-03,0.634599926E-03
0.492139199E-04,0.580160462E-03
0.492139199E-04,0.580160462E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492139199E-04,0.580160462E-03
0.492139199E-04,0.580160462E-03
0.758366987E-04,0.682694410E-03
0.492697000E-04,0.634856137E-03
0.492120739E-04,0.580157503E-03
0.492120739E-04,0.580157503E-03
0.492139199E-04,0.580160462E-03
0.492139199E-04,0.580160462E-03
0.492697000E-04,0.634856137E-03
0.126390160E-03,0.634599926E-03

0.000000000E+00,-.749271569E-07
0.000000000E+00,0.749271569E-07
0.000000000E+00,-.749271569E-07
0.000000000E+00,0.501703800E-06
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.749271569E-07
0.000000000E+00,-.749271569E-07
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,-.749271569E-07
0.000000000E+00,0.501703800E-06
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.749271569E-07
0.000000000E+00,-.749271569E-07
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,0.000000000E+00
0.000000000E+00,-.749271569E-07
0.000000000E+00,0.501703800E-06

0.147917825E+01,0.136435749E+02
0.890255341E+00,0.126414123E+02
0.920430106E+00,0.114926792E+02
0.891470771E+00,0.114951657E+02
0.920430106E+00,0.114926792E+02
0.891470771E+00,0.114951657E+02
MATRIX [ohms]:
0.890255341E+00,0.126414123E+02
0.247764775E+01,0.126433483E+02
0.891470771E+00,0.114951657E+02
0.862509528E+00,0.114976311E+02
0.891470771E+00,0.114951657E+02
0.862509528E+00,0.114976311E+02
0.920430106E+00,0.114926792E+02
0.891470771E+00,0.114951657E+02
0.147924770E+01,0.136437964E+02
0.890324230E+00,0.126416340E+02
0.920466274E+00,0.114927360E+02
0.891506793E+00,0.114952226E+02
0.891470771E+00,0.114951657E+02
0.862509528E+00,0.114976311E+02
0.890324230E+00,0.126416340E+02
0.247771608E+01,0.126435701E+02
0.891506793E+00,0.114952226E+02
0.862545404E+00,0.114976881E+02
0.920430106E+00,0.114926792E+02
0.891470771E+00,0.114951657E+02
0.920466274E+00,0.114927360E+02
0.891506793E+00,0.114952226E+02
0.147924770E+01,0.136437964E+02
0.890324230E+00,0.126416340E+02
0.891470771E+00,0.114951657E+02
0.862509528E+00,0.114976311E+02
0.891506793E+00,0.114952226E+02
0.862545404E+00,0.114976881E+02
0.890324230E+00,0.126416340E+02
0.247771608E+01,0.126435701E+02

0.449379481E-06,0.157367584E-02
-.237634492E-05,-.157366405E-02
-.544875861E-12,-.554937552E-11
-.215233875E-07,0.203904846E-08
-.544871327E-12,-.554937765E-11
-.215233875E-07,0.203904846E-08
MATRIX [mhos]:
-.237634492E-05,-.157366405E-02
0.228333559E-04,0.106280469E-01
-.215233875E-07,0.203904846E-08
0.770283689E-05,0.844406496E-04
-.215233875E-07,0.203904846E-08
0.770283689E-05,0.844406496E-04
-.544875861E-12,-.554937560E-11
-.215233875E-07,0.203904846E-08
0.449379481E-06,0.157367584E-02
-.237634533E-05,-.157366405E-02
-.544893406E-12,-.554940584E-11
-.215234988E-07,0.203911822E-08
-.215233875E-07,0.203904846E-08
0.770283689E-05,0.844406496E-04
-.237634533E-05,-.157366405E-02
0.228338831E-04,0.106280485E-01
-.215234988E-07,0.203911822E-08
0.770310898E-05,0.844410870E-04
-.544871327E-12,-.554937773E-11
-.215233875E-07,0.203904846E-08
-.544893449E-12,-.554940585E-11
-.215234988E-07,0.203911822E-08
0.449379481E-06,0.157367584E-02
-.237634533E-05,-.157366405E-02
-.215233875E-07,0.203904846E-08
0.770283689E-05,0.844406496E-04
-.215234988E-07,0.203911822E-08
0.770310898E-05,0.844410870E-04
-.237634533E-05,-.157366405E-02
0.228338831E-04,0.106280485E-01
Phase A
Phase B
Phase C

50.00 Hz:
0.577350269E+00,0.000000000E+00
0.577350269E+00,0.000000000E+00
0.577350269E+00,0.000000000E+00
0.577350269E+00,0.000000000E+00
-.288675135E+00,-.500000000E+00
-.288675135E+00,0.500000000E+00
0.577350269E+00,0.000000000E+00
-.288675135E+00,0.500000000E+00
-.288675135E+00,-.500000000E+00

-.651709507E-05,0.252038515E-05
0.191145680E-03,0.186342892E-02
-.103172190E-04,0.317366066E-04
0.434553853E-04,0.682688441E-04
-.651709507E-05,0.252038515E-05
0.370165632E-04,0.434599724E-05
0.147656661E-03,0.175870736E-02
-.158020636E-04,-.909972440E-05
0.185933687E-07,0.182318915E-04
-.157985779E-04,-.909984340E-05
0.185933687E-07,0.182318915E-04
-.157985779E-04,-.909984340E-05
-.103172190E-04,0.317366066E-04
-.539049498E-04,-.247765680E-04
0.434553853E-04,0.682688441E-04
0.157816243E-04,-.913512628E-05
0.157799845E-04,-.913204806E-05
0.157799845E-04,-.913204806E-05
0.147656661E-03,0.175870736E-02
0.157816243E-04,-.913512628E-05
-.158020636E-04,-.909972440E-05
0.208004033E-03,0.184741613E-02
-.547669809E-05,-.408411660E-04
0.222906651E-04,-.116541116E-04
0.185933687E-07,0.182318915E-04
0.157799845E-04,-.913204806E-05
-.157985779E-04,-.909984340E-05
0.222906651E-04,-.116541116E-04
0.603064929E-04,0.522390670E-04
-.542091798E-05,0.138545084E-04
0.185933687E-07,0.182318915E-04
0.157799845E-04,-.913204806E-05
-.157985779E-04,-.909984340E-05
-.547669809E-05,-.408411660E-04
0.223464452E-04,0.430415628E-04
0.603064929E-04,0.522390670E-04

0.000000000E+00,0.167234600E-06
0.101570198E-06,-.836173000E-07
-.101570198E-06,-.836173000E-07
0.000000000E+00,0.242161757E-06
0.101570198E-06,-.836173000E-07
-.166459019E-06,-.461537215E-07
0.000000000E+00,-.249757190E-07
0.216296071E-07,0.124878595E-07
0.000000000E+00,-.249757190E-07
0.216296071E-07,0.124878595E-07
0.000000000E+00,-.249757190E-07
0.216296071E-07,0.124878595E-07
-.101570198E-06,-.836173000E-07
0.166459019E-06,-.461537215E-07
0.000000000E+00,0.242161757E-06
-.216296071E-07,0.124878595E-07
-.216296071E-07,0.124878595E-07
-.216296071E-07,0.124878595E-07
0.000000000E+00,-.249757190E-07
-.216296071E-07,0.124878595E-07
0.216296071E-07,0.124878595E-07
0.000000000E+00,0.309493481E-06
0.123199805E-06,0.961051595E-07
-.123199805E-06,0.961051595E-07
0.000000000E+00,-.249757190E-07
-.216296071E-07,0.124878595E-07
0.216296071E-07,0.124878595E-07
-.123199805E-06,0.961051595E-07
0.000000000E+00,0.384420638E-06
0.123199805E-06,0.211780026E-07
0.000000000E+00,-.249757190E-07
-.216296071E-07,0.124878595E-07
0.216296071E-07,0.124878595E-07
0.123199805E-06,0.961051595E-07
-.123199805E-06,0.211780026E-07
0.000000000E+00,0.384420638E-06
Zero
Positive
Negative
Zero
Positive
Negative
Figure F.3: The figure shows the impedance and admittance matrix for the corrected underground
cable section, simulated in chapter 6.
-----------------------------------------------------
Appendix G

G.1
M-files for determination of parameters in the overhead line

system
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
% Predefined constants :
p = 0.028264;
p_jord = 49.5;
f = 50;
w = 2* pi * f ;
%
%
%
%
Resistivit y of aluminium
Resistivit y of earth return
Frequency from power source
Angular frequency form power source
A_dorkin = 153;
A_martin = 685;
l = 18930;
Ds = 0.0146;
Ds_g = 0.007;
%
%
%
%
%
Cross section of dorkin

Cross section of martin
Distance for section 2
Self GMD for Martin conductor
Self GMD for Dorkin conductor
a = exp ( j *((2* pi ) /3) ) ;

T = [[1 1
1 ]
[1 a ^2 a ]
[1 a
a ^2]];
% D i s p l a c e m e n t between phases (120 degrees )

% T r a n s f o r m a t i o n matrix
Rdc = p / A_martin ;
Rdc_g = p / A_dorkin ;
Rd = 49.5 e -6;
k_ac = 1.15;
Rac = Rdc * k_ac ;
%
%
%
%
Resistivit y of martin conductor

Resistivit y of ground conductor
Resistivit y of earth return
Correction factor for resistance
r = sqrt ( A_martin / pi ) /1000; % Radius of conductor

rg = sqrt ( A_dorkin / pi ) ;
% Radius of ground conductor

MatLab script G.1: constants.m-M-file for constants for calculation of impedance and admittance.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
% Geometry
%
G (8.35 41.6)
%
|
%
|
%
- - - - - - - - T (10.4 33.5)
%
/ \
%
/
\
%
/
\
% (7.2 24.7) S - - - - - - - R (13.0 24.7)
% Coordinat e s [ x y ]
Sxy = [7.2 24.7];
Rxy = [13.0 24.7];
Txy = [10.4 33.5];
Gxy = [8.35 41.6];
% Mirror images coordinat es [ x y ]
Sxym = [7.2 -24.7];
Rxym = [13.0 -24.7];
Txym = [10.4 -33.5];
Gxym = [8.35 -41.6];
Dj
= 660* sqrt ( p_jord / f ) ;
% distance to earth - return
% Distances between conductors
d12 = Rxy (1) - Sxy (1) ;
d13 = sqrt (( Txy (1) - Sxy (1) ) ^2 +( Txy (2) - Sxy (2) ) ^2) ;
d1g = sqrt (( Gxy (1) - Sxy (1) ) ^2 +( Gxy (2) - Sxy (2) ) ^2) ;
% phase s - to - r / t / g
d21 = d12 ;
d23 = sqrt (( Rxy (1) - Txy (1) ) ^2 +( Rxy (2) - Txy (2) ) ^2) ;
d2g = sqrt (( Gxy (1) - Rxy (1) ) ^2 +( Gxy (2) - Rxy (2) ) ^2) ;
% phase r - to - s / t / g
d31 = d13 ;
d32 = d23 ;
d3g = sqrt (( Gxy (1) - Txy (1) ) ^2 +( Gxy (2) - Txy (2) ) ^2) ;
% phase t - to - s / r / g
dg1 = d1g ;
% ground - to - s / r / t
G-58
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
dg2 = d2g ;
dg3 = d3g ;
% Distances to mirror images
D11 = 2* Sxy (2) ;
% phase S - to - S / R / T / G
D12 = sqrt (( Sxym (1) - Rxy (1) ) ^2 + ( Sxym (2) - Rxy (2) ) ^2) ;
D13 = sqrt (( Txym (1) - Sxy (1) ) ^2 + ( Txym (2) - Sxy (2) ) ^2) ;
D1g = sqrt (( Gxym (1) - Sxy (1) ) ^2 + ( Gxym (2) - Sxy (2) ) ^2) ;
D21
D22
D23
D2g
=
=
=
=
D12 ;
% phase R - to - S / R / T / G
2* Rxy (2) ;
sqrt (( Rxym (1) - Txy (1) ) ^2 + ( Rxym (2) - Txy (2) ) ^2) ;
sqrt (( Gxym (1) - Rxy (1) ) ^2 + ( Gxym (2) - Rxy (2) ) ^2) ;
D31
D32
D33
D3g
=
=
=
=
D13 ;
% phase T - to - S / R / T / G
D23 ;
2* Txy (2) ;
sqrt (( Gxym (1) - Txy (1) ) ^2 +( Gxym (2) - Txy (2) ) ^2) ;
Dg1
Dg2
Dg3
Dgg
=
=
=
=
D1g ;
D2g ;
D3g ;
2* Gxy (2) ;
% ground - to - S / R / T / G
MatLab script G.2: geometry.m-M-file for geometry of the overhead line system.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
% overhead line series impedance c a l c u l a t i o n s

clc ; close all ; clear all ;
% Load constants & geometry
constants
geometry
% DC - resistance matrix
R_abc = [[ Rac + Rd
Rd
Rd
[
Rd
Rac + Rd
Rd
[
Rd
Rd
Rac + Rd
[
Rd
Rd
Rd
% Inductance matrix
L_abc = 2 e -7* [[ log ( Dj / Ds )
[ log ( Dj / d21 )
[ log ( Dj / d31 )
[ log ( Dj / dg1 )
]
Rd
Rd
Rd
Rdc_g + Rd
log ( Dj / d12 )
log ( Dj / Ds )
log ( Dj / d32 )
log ( Dj / dg2 )
]
]
]
]];
%[ ohm / m ]
log ( Dj / d13 )
log ( Dj / d23 )
log ( Dj / Ds )
log ( Dj / dg3 )
log ( Dj / d1g ) ]
log ( Dj / d2g ) ]
log ( Dj / d3g ) ]
log ( Dj / Ds_g ) ]]; % [ H / m
% Inductance phase values

Labc = ( L_abc (1:3 ,1:3) - L_abc (1:3 ,4) *(1/ L_abc (4 ,4) ) * L_abc (4 ,1:3) ) ;
% Impedance
Z_abc = R_abc + j * w * L_abc ;
% [ ohm / m ]
% Reduction of 4 x4 impedance matrix to 3 x3 matrix
Z_abc = Z_abc (1:3 ,1:3) - Z_abc (1:3 ,4) *(1/ Z_abc (4 ,4) ) * Z_abc (4 ,1:3) ;
% Sequence matrix
Z_012 = T ^ -1 * Z_abc * T ;
% Reactance
X_012 = imag ( Z_012 ) ;
% Resistance
R_012 = real ( Z_012 ) ;
% [ ohm / m ]
% [ ohm / m ]
% [ ohm / m ]
Z _ a b c _ s e c t i o n 2 = Z_abc * l
R _ a b c _ s e c t i o n 2 = real ( Z _ a b c _ s e c t i o n 2 )
L _ a b c _ s e c t i o n 2 = imag ( Z _ a b c _ s e c t i o n 2 ) / w
Z _ 0 1 2 _ s e c t i o n 2 = Z_012
L _ 0 1 2 _ s e c t i o n 2 = X_012 / w * l
% [ ohm ]
% [H]
R _ a c _ s e c t i o n 2 = Rac * l
R _ d c g _ s e c t i o n 2 = Rdc_g * l
% [ ohm ]
% [ ohm ]
MatLab script G.3: impedance.m-M-file for calculations of the impedance sequence matrix.

1
2
3
4
% overhead line admittance c a l c u l a t i o n s

G.2 M-files for determination of parameters in the shunt reactor
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
G-59
constants
geometry
% Potential matrix
P_abcg = 18 e9 *[[
[
[
[
log ( D11 / r )
log ( D21 / d21 )
log ( D31 / d31 )
log ( Dg1 / dg1 )
log ( D12 / d12 )

log ( D22 / r )
log ( D32 / d32 )
log ( Dg2 / dg2 )
log ( D13 / d13 )

log ( D23 / d23 )
log ( D33 / r )
log ( Dg3 / dg3 )
log ( D1g / d1g ) ]

log ( D2g / d2g ) ]
log ( D3g / d3g ) ]
log (( Dgg ) / rg ) ]];
% Reduction to 3 x3 matrix
P_abc = P_abcg (1:3 ,1:3) - P_abcg (1:3 ,4) *(1/ P_abcg (4 ,4) ) * P_abcg (4 ,1:3) ;
C_abc = P_abc ^ -1;
X_abc = 1./( w * C_abc ) ;
B_abc = 1./ X_abc ;
Y_abc = 0 + j * B_abc ;
%
%
%
%
Capacitan ce
Reactance
Susceptan ce
Admittance matrix
% Sequence matrix
Y_012 = T ^ -1 * Y_abc * T ;
% Writeout of parameters
C _ a b c _ s e c t i o n 2 = imag ( Y_abc ) / w * l * 1 e6
Y_012_section2 =
Y_012_section2km
C_012_section2 =
C_012_section2km
Y_012 * l * 1 e6 % [ muS ]
= Y_012 * 1000 * 1 e6 % [ muS / km ]
imag ( Y_012 ) / w * l * 1 e6 % [ muF ]
= imag ( Y_012 ) / w * 1000 * 1 e6 % [ muF / km ]

MatLab script G.4: admittance.m-M-file for calculations of the admittance sequence matrix.
G.2
M-files for determination of parameters in the shunt reactor

1
2
3
4
5
% Predefined constants :
mu_0 = 4* pi *10^ -7;
% Permeability
mu_r = 52073.5;
% Relative p e r m e a b i l i t y in the core
mu_r_gap = 1.302;
% Relative p e r m e a b i l i t y in the gap
N = 1194;
% Number of windings
MatLab script G.5: constants.m-M-file for constants for calculation of the shunt reactor.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
%
%
%
%
%
%
l3
l2
l2
l3
------------------------|
|
|
|
|
h |
|
|
|
|
|
|
|
|
|
-------------------------
l = 4182 e -3;
l2 = 1203 e -3;
l3 = 888 e -3;
h = 2213 e -3;
d_core = 210 e -3;
w_core = 794 e -3;
d_limb = 600 e -3;
%
%
%
%
%
%
%
Total lenght
Length l2 in figure above
Length l3 in figure above
Total height
Diameter of outer core
Width of outer core
Diameter of limb
h_seg = 151.9 e -3;

% Height of one segment
h_plate = 70 e -3;
% Cross flux plates
N_seg = 9;
% Number of segment in one limb
h_limb = h_seg * N_seg +2* h_plate ; % Total limb height
h_g =
N_g =
h_air
h_gap
32.7 e -3;
8;
= 2*10 e -3+2*2.5 e -3;
= h_g * N_g + h_air ;
%
%
%
%
Height of one gap

Number of gaps
Gap at flux plates
Total height of gaps
%
s3
s2
s2
s3
%
---------------------------------------%
|
|
|
|
|
%
|
| | sg
| | sg
| | sg
|
% s4 |
|
|
|
| s4
%
|
| s1
| s1
| s1
|
%
|
|
|
|
|
%
----------------------------------------
G-60
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
s3
s2
s2
s3
% Flux mean paths :

s2 = l2 ;
s3 = l3 - d_core /2;
s4 = h - d_core ;
s1 = s4 - h_gap ;
sg = h_gap ;
% Flux areas :
A1 = pi *( d_limb /2) ^2;
Ag = A1 ;
A2 = d_core * w_core ;
A3 = A2 ;
A4 = A2 ;
MatLab script G.6: geometry.m-M-file for geometry of the shunt reactor.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42

constants
geometry
% Reluctance calculatio n :
R1 = s1 /( mu_0 * mu_r * A1 ) ;
R2 = s2 /( mu_0 * mu_r * A1 ) ;
R3 = s3 /( mu_0 * mu_r * A2 ) ;
R4 = s4 /( mu_0 * mu_r * A3 ) ;
Rg = sg /( mu_0 * mu_r_gap * Ag ) ;
% Uniting reluctance s :
R1g = R1 + Rg ;
R343 = R3 + R4 + R3 ;
% Reluctance matrix :
R = [[ R343 + R1g
- R1g
[ - R1g
(2* R2 + 2* R1g )
[0
- R1g
[0
0
0
0
- R1g
0
(2* R2 + 2* R1g ) - R1g
- R1g
R343 + R1g
];
];
];
]];
% Inversion of reluctanc e matrix :

R_inv = R ^ -1
% Reduction premeance to 3 x3 :
P_123 = [(( R_inv (2 ,2) - R_inv (1 ,2) )
(( R_inv (3 ,2) - R_inv (2 ,2) )
(( R_inv (4 ,2) - R_inv (3 ,2) )
(( R_inv (2 ,3) - R_inv (1 ,3) )
(( R_inv (3 ,3) - R_inv (2 ,3) )
(( R_inv (4 ,3) - R_inv (3 ,3) )
(( R_inv (2 ,4) - R_inv (1 ,4) )
(( R_inv (3 ,4) - R_inv (2 ,4) )
(( R_inv (4 ,4) - R_inv (3 ,4) )
( R_inv (2 ,1) - R_inv (1 ,1) ) ) ...

( R_inv (3 ,1) - R_inv (2 ,1) ) ) ...
( R_inv (4 ,1) - R_inv (3 ,1) ) ) ;
( R_inv (2 ,2) - R_inv (1 ,2) ) ) ...
( R_inv (3 ,2) - R_inv (2 ,2) ) ) ...
( R_inv (4 ,2) - R_inv (3 ,2) ) ) ;
( R_inv (2 ,3) - R_inv (1 ,3) ) ) ...
( R_inv (3 ,3) - R_inv (2 ,3) ) ) ...
( R_inv (4 ,3) - R_inv (3 ,3) ) ) ;]
P_123 ^ -1
1./( P_123 ^ -1)
% Inductance matrix :
L = N ^2.* P_123
G.3
MatLab script G.7: reactor.m-M-file for calculations of inductances in the shunt reactor.
M-files for mathematical analysis

1
2
3
4
5
6
7
8
9

Up = 165 e3 * sqrt (2) / sqrt (3) ;
% Parameters
R1 = 1.245;
L1 = 39.2 e -3;
R2 = 1.517;
L2 = 2.310;
G.3 M-files for mathematical analysis
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
Rs = 0.800;
Ls = 14.4 e -3;
C = 5.10 e -6;
% Unify short circuit impedance
R1 = R1 + Rs ;
L1 = L1 + Ls ;
omega = 2* pi *50;
a = 2* pi /3;
t =0:0.00005:2.8;
Zs = tf ([ Ls Rs ] ,1) ;
Yin = tf ([ C * L2 C * R2 1] ,[ C * L1 * L2 ( R1 * C * L2 + R2 * C * L1 ) ( C * R2 * R1 + L1 + L2 ) ( R1 + R2 ) ]) ;
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Current
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Phase a :
alpha = pi /2;
A = sin ( alpha ) ;
B = cos ( alpha ) ;
Ua = tf ([ A * Up B * Up * omega ] ,[1 0 omega ^2]) ;
Ia = Yin * Ua ;
% Extract numerator and denumerato r
[ Ia_num , Ia_denum ] = tfdata ( Ia , v ) ;
% Partitial fraction
[r , p , k ] = residue ( Ia_num , Ia_denum ) ;
ia = 0;
for k = 2: length ( r ) +1
ia = ia + r (k -1) * exp ( p ( k -1) * t ) ;
end
% Phase b :
alpha = pi /2 - a ;
A = sin ( alpha )
B = cos ( alpha )
Ub = tf ([ A * Up B * Up * omega ] ,[1 0 omega ^2]) ;
Ib = Yin * Ub ;
[ Ib_num , Ib_denum ] = tfdata ( Ib , v ) ;
[r , p , k ] = residue ( Ib_num , Ib_denum ) ;
ib = 0;
ib = ib + r (k -1) * exp ( p ( k -1) * t ) ;
end
% Phase c :
alpha = pi /2 + a ;
A = sin ( alpha ) ;
B = cos ( alpha ) ;
Uc = tf ([ A * Up B * Up * omega ] ,[1 0 omega ^2]) ;
Ic = Yin * Uc ;
[ Ic_num , Ic_denum ] = tfdata ( Ic , v ) ;
[r , p , k ] = residue ( Ic_num , Ic_denum ) ;
ic = 0;
ic = ic + r (k -1) * exp ( p ( k -1) * t ) ;
end
figure
plot (t , ia *1 e -3 , t , ib *1 e -3 , t , ic *1 e -3) ; grid ;
title ( \ bfi ( t ) )
G-61
G-62
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
xlabel ( t [ s ] )
ylabel ( i [ kA ] )
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Voltage
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Phase a :
U1a = Ua - Ia * Zs ;
% Extract numerator and denumerat o r
[ U1a_num , U1a_denum ] = tfdata ( U1a , v ) ;
[r ,p , k ] = residue ( U1a_num , U1a_denum )
u1a = 0;
u1a = u1a + r (k -1) * exp ( p (k -1) * t ) ;
end
% Phase b :
U1b = Ub - Ib * Zs ;
[ U1b_num , U1b_denum ] = tfdata ( U1b , v ) ;
[r ,p , k ] = residue ( U1b_num , U1b_denum ) ;
u1b = 0;
u1b = u1b + r (k -1) * exp ( p (k -1) * t ) ;
end
% Phase c :
U1c = Uc - Ic * Zs ;
[ U1c_num , U1c_denum ] = tfdata ( U1c , v ) ;
[r ,p , k ] = residue ( U1c_num , U1c_denum ) ;
u1c = 0;
u1c = u1c + r (k -1) * exp ( p (k -1) * t ) ;
end
figure
plot (t , u1a *1 e -3 , t , u1b *1 e -3 , t , u1c *1 e -3) ; grid ;
title ( \ bfu1 ( t ) )
xlabel ( t [ s ] )
ylabel ( u [ kV ] )
MatLab script G.8: simpel.m-M-file for numerical approach in MatLab.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
clc , clear all , close all

t = 1:0.1:100;
Ua = 165 e3 * sqrt (2) / sqrt (3) ;
alpha = - pi /2; % phase delay
% Parameters
R1 = 1.245;
L1 = 39.2 e -3;
R2 = 1.969;
L2 = 2.310;
Rs = 0.800;
Ls = 14.4 e -3;
C = 5.10 e -6;
Mab = 32.0 e -3;
Mac = 23.0 e -3;
Mbc = 32.0 e -3;
% [H]
% [H]
% [H]
% Unify short circuit impedance

R_1 = R1 + Rs ;
L_1 = L1 + Ls ;
R_2 = R2 ;
G.4 M-files for miscellaneous purposes
23
24
25
G-63
L_2 = L2 ;
Fs = 20000;

MatLab script G.9: simplified_model_init.m-Initialization of calculated parameters for simulink
model.
G.4
M-files for miscellaneous purposes

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

fs = 1/0.035 e -3; % sampling frequency (0.035 e -3 s per sample )
% Load file and voltage waveform
ans = csvread ( m e a s u r e m e n t s _ f e r s l e v / Omicron /150 kV . txt ) ;
U = ans (3 ,:) /1000; % in kV
% Shorten data to one period after
N = 572; % period time = 0.02 s = > 572 samples / period
U = U ( 2 8 6 : 2 8 6 + 5 7 1 ) ; % switch on t =0.01 s = > 286 samples
X = abs ( fft ( U , N ) ) /( N /2) ;
X = fftshift ( X ) ;
F = fs * [ - N /2 : N /2 - 1]/ N ; % Mirrored round x - axis
figure
bar (F , X ) ; % plot fft as bar s
grid ;
axis ([0 6500 0 10]) ; % 0 - >6500 Hz
xlabel ( Frequency [ Hz ] ) ;
ylabel ({ Percentage of the fundamenta l [%] ; 150 kV level }) ;

MatLab script G.10: fft_for_report.m-M-file for FFT in MatLab. This file shows the principle for
FFT analysis of one phase.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
clc ; clear all ; close all

mu = 4* pi *10^ -7;
rho = 2.826*10^ -8;
sigma = 1/ rho ;
r = 18.1 e -3;
Rdc = 0 . 0 4 2 3 * 1 8 . 9 3 ;
L = 36.2 e -3;
C = .149 e -6;
Rg_dc = 0 . 0 0 0 9 9 * 5 0 * 1 8 . 9 3 ;
Rc = 0.693;
Lc = 7.20 e -3;
Cc = 5.10 e -6;
for f = 1:2000
d ( f ) = sqrt ( rho ./( f * pi * mu ) ) ;
omega ( f ) = 2* pi * f ;
if f < 65
ks ( f ) = 1 + ( r / d ( f ) ) ^4/(48 + 0.8*( r / d ( f ) ) ^4) ;
else
ks ( f ) = 0.25 + 0.5*( r ./ d ( f ) ) + 3/32*( d ( f ) / r ) ;
end
Rac ( f ) = Rdc * ks ( f ) ;
Rg ( f ) = 0.00099* f *18.93;
Z_ac_in ( f ) = (( Rac ( f ) + Rg ( f ) ) /2+ j * omega ( f ) * L /2) + (1./( j * omega ( f ) * C ) ) * (( Rac ( f ) +
Rg ( f ) ) /2+ j * omega ( f ) * L /2) /((1./( j * omega ( f ) * C ) ) + (( Rac ( f ) + Rg ( f ) ) /2+ j * omega ( f ) * L
/2) ) ;
R_ac_in ( f ) = real ( Z_ac_in ( f ) ) ;
30
31
32
33
34
35
Z_dc_in ( f ) = (( Rdc + Rg_dc ) /2+ j * omega ( f ) * L /2) + (1./( j * omega ( f ) * C ) ) * (( Rdc + Rg_dc )
/2+ j * omega ( f ) * L /2) /((1./( j * omega ( f ) * C ) ) + (( Rdc + Rg_dc ) /2+ j * omega ( f ) * L /2) ) ;
R_dc_in ( f ) = real ( Z_dc_in ( f ) ) ;
end
G-64
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
f_ps = [50 100 150 200 250 300 350 400 500 750 1000 1250 1500 1750 2000 2500 3000] ;
U = [165* sqrt (2) / sqrt (3) ];
I = [11.4477 5.93649 4.03 3.25 2.433 2.036 1.75 1.53 1.228 .8128 .5958 .4642 .37075
.30045 .243 .1572 .08842];
tu = [.180];
ti = [.184484 .18233593 .18157438 .18118904 .18095537 .18079782 .180684628 .180599897
.180480785 .18032086 7 8 .18024049 .180192066 .18015975 .180136694 .18011871 9
.180094297 5 . 1 8 0 0 7 5 9 0 9 0 ] ;
for k = 1: length ( f_ps )
theta ( k ) = ( ti ( k ) - tu (1) ) /(1/ f_ps ( k ) ) * 360;
Z ( k ) = U (1) / I ( k ) ;
R ( k ) = cos ( theta ( k ) *2* pi /360) * Z ( k ) ;
end
f = 1: f ( end ) ;
plot (f , R_dc_in ,f , R_ac_in , -. , f_ps ,R , o : ) ; grid ;
axis ([0 f ( end ) 0 50]) ;
ylabel ( Real part of input impedance (\ Omega ) ) ;
xlabel ( Frequency [ Hz ] )
legend ( Calculate d with frequency independe nt series and ground resistance ,
Calculated with frequency dependent series and ground resistance , Calculated
from PSCAD simulation ,2) ;
set ( gcf , PaperUnits , centimeters )
set ( gcf , PaperPosition ,[1 0 32 16])
% saveas ( gca , ../ report / billeder / pscad / v e r i f i c a t i o n _ f r e q u e n c y . eps ) ;

MatLab script G.11: pscad_frekvens_verification.m-M-file for calculating and plotting the frequency
dependence of the transmission line in PSCAD.
Appendix H
CD content
In this appendix, the content of the CD is briefly described. Besides the three main folders described in
the following, an electronic copy of the report is placed in the root of the CD.
Data & information

Articles
A_charge_simulation_based_method_for_calculating_corona_loss.pdf
This article concerns calculation of corona losses on AC transmission lines.
Optimized_Transformer_Design_Inclusive_of_high_frequency_effects.pdf
This article concerns the frequency aspect in transformer design. Among others frequency
dependence of a round wire.
Pre-insertion_resistor_in_HV_capacitor_bank_switching.pdf
This article concerns limitations of switching overvoltages, with different approaches.
Miscellaneous
Description_of_the_system_Horns_Rev.pdf
Contains information about the two windmill parks at Horns Rev.
Miljorapport2006.pdf
Contains information about energy consumption, production, transmission etc. in Denmark.
Overhead line
Donau.bmp
Contains information about the physical layout of the overhead line section.
PSCAD
PSCAD help file.chm
Contains the help file for PSCAD v4.2.
PSCAD Users Guide.pdf
Contains the users guide for PSCAD v4.2.
Shunt reactor
Core_and_winding_geometries.pdf
Contains information about the physical layout of the shunt reactor.
Core_design.pdf
Contains information about the principle structure of the shunt reactor.
DRAFT_IEC_60076-6_Reactors_Chapter_7_Shunt_reactors.pdf
Contain a draft of the standard IEC60076-6.
Instruction_Manual.pdf
Contains information about the shunt reactor, including test report and rating plate.
H-66
CD content
Magnetising_curve.pdf
Contains the DC magnetization curve for the reactor core (M-5 0.30 mm).
The_ABB_Shunt_Reactor.pdf
Contains information about the shunt reactor, including test report and rating plate.
Winding_and_core_data.pdf
Contains information about design of the shunt reactor.
Transformers
60-20kV_&_20-0,4kV.pdf
Contains information about the 60-20 kV and the 20/0,4 kV transformer.
150-60kV.pdf
Contains information about the 150-60 kV transformer.
Underground cable & stations
one-line_diagram_and_cable_information.PDF
Contains the one-line diagrams for FER3 and TH3. Furthermore, information about parameter,
layout etc. for the cable are included in this file.
XLPE_Cable_Systems_Users_Guide.pdf
Contains the users guide for the cable.
M-files
FFT analysis
principle_of_fft.m
M-file containing the principle of the Fourier analysis.
Mathematical analysis
simpel.m
M-file for solving the simplified model through the analytical approach.
Parameter determination of the overhead line system
admittance.m
M-file for calculations of the admittance sequence matrix.
constants.m
M-file for constants for calculation of impedance and admittance.
geometry.m
M-file for geometry of the overhead line system.
impedance.m
M-file for calculations of the impedance sequence matrix.
Parameter determination of the shunt reactor
constants.m
M-file for constants for calculation of the shunt reactor.
H-67
geometry.m
M-file for geometry of the shunt reactor.
reactor.m
M-file for calculations of inductances in the shunt reactor.
PSCAD frequency dependent resistance
pscad_frequency_dependent_resistance.m
M-file for calculating and plotting the frequency dependence of the transmission line in PSCAD.
PSCAD
Ferslev-Tinghoj_1.psc
Simulation of the system FER3-TH3, described in chapter 5 on page 57.
Ferslev-Tinghoj_2.psc
Simulation of the system FER3-TH3, with the corrected cable as described in chapter 6 on
page 83.
Pre-insertion_resistor.psc
Simulation of the pre-insertion resistor principle as described in chapter 7 on page 97.
Reactor_verification.psc
Simulation of the reactor verification as described in section 5.2 on page 68.
Simple_model.psc
Simulation of the simplified model as described in section 4.4 on page 44.
Zero_crossing_closure.psc
Simulation of the zero crossing closure principle as described in chapter 7 on page 97.

Propagation of Transmission Level Switching Overvoltages PDF

Uploaded by

Copyright:

Available Formats

Propagation of Transmission Level Switching Overvoltages PDF

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

Propagation of Transmission Level Switching Overvoltages PDF

Uploaded by

Copyright:

Available Formats

Propagation of transmission level switching overvoltages,

related to the distribution level

Institute of Energy Technology

Institut for Energiteknik

Description and modelling of the 150 kV grid

Overhead transmission lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Three phase representation of the grid

Summary & discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Analysis of switching overvoltages on the 150 kV grid

Estimation of the transient behaviour . . . . . . . . . . . . . . . . . . . . . . . .

Summary & discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Configuration and simulation in PSCAD

PSCAD model subparts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Verification of reactor model . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Summary & discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Evaluation of the PSCAD simulation

Estimation of cable and line models . . . . . . . . . . . . . . . . . . . . . . . . .

Estimation of models for shunt reactor and short circuit impedance . . . . . . .

Correction of cable capacitance . . . . . . . . . . . . . . . . . . . . . . . . . . .

Estimation of missing resistance . . . . . . . . . . . . . . . . . . . . . . . . . . .

Summary & discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Proposal for minimizing switching overvoltages

Evaluation of switching overvoltages at the 150 kV level . . . . . . . . . . . . . .

Evaluation of propagation of switching overvoltages . . . . . . . . . . . . . . . .

Proposal for minimization of switching overvoltages . . . . . . . . . . . . . . . .

Summary & discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

Parameter determination for the overhead line system

A.1 Series impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-1

Parameter determination in the shunt reactor

B.1 Theoretical background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-25

Measurements of transmission line switching

C.1 General test procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-35

Transient recording using Omicron CMC 256-6

D.1 Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-47

E.1 Ferslev substation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-49

Parameter matrices for PSCAD

M-files used for calculations

Consumer level transformer

Figure 1.2: Graphical overview of the electrical power system

The electricity produced by the windmills is collected at a transformer platform, transforming

Figure 1.4: Location of the 150 kV grid from Ferslev to Tinghj.

Figure 1.5: Subsections of the system.

The measuring reports are in appendix C on page C-35.

The magnitudes of the recorded voltages are summarized in table 2.1.

Description and modelling of the 150

Description and modelling of the 150 kV grid

3.2 150 kV source

Intact transmission network, including AC-connection to Germany

Description and modelling of the 150 kV grid

Table 3.3: Resistances in the circuit breaker in its two states.

3.4 Underground cable

Grounding & Cross bonding

Figure 3.6: Cable duct and triangular cable layout.

Description and modelling of the 150 kV grid

Joint & screen-seperation

Linkbox for crossbonding

Linkbox for earth-connection

3.4 Underground cable

Three phase representation