570

570
Switching Phenomena for

EHV and UHV Equipment
Working Group
A3.28
February 2014
SWITCHING
PHENOMENA FOR
EHV AND UHV
EQUIPMENT
WG A3.28
Members
J.Fan, U.Riechert, M.Kosakada, Y.Yamagata, P.C.Fernandez, A.Rocha, E.Kynast, S.Poirier,
Z.Xiang, X.Wang, W.Li, A.Giboulet, S.Kale, F.Lo Monaco, H.Kajino, S-W Bahng, C.van der
Merwe, A.Alfredsson, K.Anantavanich, A.Keri, B.Shperling, V.Rashkes, J.Amon, R.Smeets,
M.C.Bhatnagar, B. Khodabakhchian, D.Peelo
D. DUFOURNET, Convenor (FR), A. JANSSEN, Secretary (NL),
H.ITO, Former Convenor (JP)
Copyright © 2012
“Ownership of a CIGRE publication, whether in paper form or on electronic support only infers right
of use for personal purposes. Unless explicitly agreed by CIGRE in writing, total or partial
reproduction of the publication and/or transfer to a third party is prohibited other than for personal
use by CIGRE Individual Members or for use within CIGRE Collective Member organisations.
Circulation on any intranet or other company network is forbidden for all persons. As an exception,
CIGRE Collective Members only are allowed to reproduce the publication.
Disclaimer notice
“CIGRE gives no warranty or assurance about the contents of this publication, nor does it accept
any responsibility, as to the accuracy or exhaustiveness of the information. All implied warranties
and conditions are excluded to the maximum extent permitted by law”.
ISBN : 978-2-85873-265-4
Switching phenomena for EHV and UHV Equipment
ISBN : 978-2-85873-265-4
Page 1
Abbreviations
1LG Single-phase line fault to ground
2LG Two-phase line faults to ground
3LG Three-phase line faults to ground
AC Alternating Current
AF Amplitude Factor (= kaf)
AG Advisory Group
AIS Air Insulated Switchgear
CB Circuit Breaker
DC Direct Current
DS Disconnecting Switch
EHV Extra High Voltage
EM Electromagnetically
EMC Electromagnetic Compatibility
ES Earthing Switch
FACTS Flexible AC Transmission Systems
FFO Fast Front Overvoltage
FRA Frequency Response Analysis
GIS Gas Insulated Switchgear
HGIS Hybrid Gas Insulated Switchgear
HSES High Speed Earthing Switches (also called HSGS)
HVDC High Voltage Direct Current
IEC International Electrotechnical Commission
IEEE The Institute of Electrical and Electronics Engineers
ITRV Initial Transient Recovery Voltage
LIWL Lightning Impulse Withstand Level
LIWV Lightning Impulse Withstand Voltage
LLF Long Line Fault
MOSA Metal Oxide Surge Arrester
MOV Metal Oxide Varistor
MPAR Multi-phase rapid Auto-Reclosing
OHL Overhead line(s)
RRRV Rate of Rise of Recovery Voltage
SC Series Capacitor
SFO Slow-Front Overvoltage
Page 2
SPAR Single-Phase rapid Auto-Reclosing

SR Shunt Reactor
SVC Static Var Compensator
T30 Terminal test duty at 30% of rated short-circuit breaking current
T100s Terminal test duty at 100% of rated short-circuit breaking current with symmetrical current
T100a Terminal test duty at 100% of rated short-circuit breaking current with asymmetrical current
TCV Trapped Charge Voltage
TFF Transformer Fed Faults
TLA Transmission Line Arrester
TLF Transformer Limited Fault
TSF Transformer Secondary Fault
TOV Temporary Overvoltage
TPAR Three-Phase Rapid Auto-Reclosing
TRV Transient Recovery Voltage
UHV Ultra High Voltage (exceeding 800 kV)
VFTO Very Fast Transient Overvoltage
VT Voltage Transformer
WG Working Group
Page 3
Switching phenomena for EHV and

UHV Equipment
Table of Contents
EXECUTIVE SUMMARY ........................................................................................................ 6
1 Introduction ...............................................................................................................................11
2 Network configuration and technical solutions .................................................13

2.1 Introduction............................................................................................................................13
2.2 Design criteria ......................................................................................................................13
2.3 Secondary arc extinction .................................................................................................14
2.4 Series capacitor banks ......................................................................................................15
2.5 Insulation coordination ....................................................................................................17
2.6 Out-of-phase .........................................................................................................................18
3 TRV of circuit breakers for EHV and UHV networks ........................................22

3.1 Introduction............................................................................................................................22
3.2 Radial and meshed network models used for TRV study.................................22
3.3 System and Equipment parameters used for TRV study....................................23
3.4 Comparison of TRV evaluated in a radial network model with that in
Japan’s 1100 kV network..................................................................................................32
3.5 TRV for bus terminal faults and long-line faults in radial network model33
3.6 TRV for no-load line switching duty ..........................................................................36
3.7 Influence of system and equipment parameters on TRV ..................................44
3.8 Influence of system and operation parameters on TRV ....................................58
3.9 Conclusions ...........................................................................................................................71
4 Transformer limited faults ...............................................................................................74

4.1 Introduction............................................................................................................................74
4.2 Voltage drop..........................................................................................................................74
4.3 First-pole-to-clear factor ..................................................................................................76
4.4 Comparison with network simulations .....................................................................76
4.5 Single frequency approach .............................................................................................80
4.6 Multi-frequency model......................................................................................................81
Page 4
4.7 Advanced transformer models......................................................................................86

4.8 External capacitances........................................................................................................87
4.9 Resonance frequencies .....................................................................................................88
4.10 Simulation results on the influence of capacitance on TRV..........................88
4.11 Conclusions.........................................................................................................................95
5 High-speed earthing switch (HSES) ............................................................................97

5.1 Introduction............................................................................................................................97
5.2 Secondary arc Extinction by HSES ............................................................................ 102
5.3 Successive faults .............................................................................................................. 105
5.4 Parametric study for TRV requirements with basic model ............................ 108
5.5 HSES switching duties based on model network ............................................... 115
5.6 Recommendation for specifications......................................................................... 121
5.7 Conclusions ........................................................................................................................ 122
6 Disconnector and earthing switch............................................................................ 123

6.1 Introduction......................................................................................................................... 123
6.2 Bus transfer switching by disconnector switches.............................................. 123
6.3 Bus-charging current switching by disconnectors ............................................. 133
6.4 Earthing switch ................................................................................................................. 142
7 Switching experience during and after system commissioning ............ 160

7.1 Introduction......................................................................................................................... 160
7.2 Experience of UHV/EHV AC System in China...................................................... 160
7.3 Experience of UHV/EHV system in Japan ............................................................. 178
7.4 Experience of UHV/EHV system in Russia............................................................ 183
7.5 Experience of EHV system in Canada..................................................................... 186
7.6 Experience with Single Phase Switching in USA ............................................... 187
7.7 Conclusions on switching experience during and after system
commissioning ................................................................................................................ 189
8 Conclusions ............................................................................................................................ 190
Bibliography/References ........................................................................................................ 193
Annex ................................................................................................................................................. 200
Page 5
EXECUTIVE SUMMARY
Long distance bulk transmission of electrical energy in recent years, due to new contexts of power systems in some
countries, has forced (and keeps forcing) utilities in a number of countries to implement or plan electrical grids with
rated voltages equal or higher than 800 kV in AC and higher than 600 kV in DC. With the introduction of these
voltage levels, utilities felt the need to share knowledge and experience. At the request of IEC SC17A that had to
revise its standards to cover UHV switchgear, CIGRE SC A3 decided in 2006/2007 to establish group (WG) A3.22
“Technical Requirements for Substation Equipment exceeding 800 kV”. This WG collected information about UHV
pre commercial tests, pilot-UHV systems in some countries and commercial in-service experience as well as about
the particularities of UHV AC in comparison to EHV. In addition detailed studies of the background for technical
specifications, based on the information available, have been performed, and two CIGRE Technical Brochures
have been published: TB 362 (2008) and TB 456 (2011).
At the end of 2010 a new working group WG A3.28 “Switching Phenomena for EHV and UHV Equipment” started
investigations into the effects of possible network topologies and the system design criteria on the technical
requirements. Furthermore, a number of technical questions that were not solved yet have been addressed by WG
A3.28. Some questions had a wider scope than UHV, for instance a request to give recommendations for
Transformer Limited Fault clearing for rated voltages of 100 kV and above. The results of the studies of both
working groups have been used as input for the standardization work of IEC, especially within SC 17A.
The work of WG A3.28 was undertaken in three main areas: (1) by means of digital simulations with UHV/EHV
network models to verify the influence of system parameters on the TRV, concerning conditions such as terminal
faults, transformer limited faults, long line faults and unloaded line switching; (2) by means of the network models
and substation models, the switching and testing requirements for disconnectors, earthing switches and high speed
earthing switches have been determined and (3) field experience and switching behavior during and after
commissioning has been collected and analyzed.
After an introduction in Chapter 1, network topologies and technical solutions for bulk electric power transmission
are treated in Chapter 2.
A large variety of technical solutions has been considered to solve the problems with the redundance of OHL
circuits, the reduction of temporary and switching overvoltages, the voltage profile along the OHL, the unbalance
between phases, secondary arc extinction, and fast auto-reclosure after fault clearing. An overview of utilities’
policies is given in Chapter 2 and application aspects are discussed. A certain relationship between several
technical solutions can be seen. For instance, utilities with three parallel single circuit OHL can afford three-phase
auto-reclosure, that allows them to use simple shunt reactors, independent whether phase transposition has been
applied. With double circuit OHL (overhead lines) or two single circuit OHL in parallel, utilities prefer to apply single
phase auto-reclosure, meaning that the shunt reactors have to be adapted for that auto-reclosure condition. When
the phases of the OHLs are transposed, four-legged shunt reactors are usually applied to achieve a reasonable
probability that the secondary arc will be extinguished in due time. Otherwise, special shunt reactor switching
schemes have to be applied or even high speed earthing switches.
The application of shunt and series compensation is dealt with in Chapter 2, including the policies to switch shunt
reactors on and off (at line switching and auto-reclosing) and to by-pass series capacitor banks during fault
conditions in the power system (MOV, spark gap, fast triggered spark gaps, by-pass switch).
The reduction of temporary and switching overvoltages, by means of several technical solutions, is addressed
shortly and a general trend is that utilities try to reduce switching overvoltages to about 1.6 p.u. at the line ends.
Temporary overvoltages are most severe under conditions of severe system disturbances, including load and/or
generated power rejection. Under such conditions surge arresters are supposed to withstand the temporary
overvoltages. Not much knowledge is available about out-of-phase switching in real service, but simulations, as
presented in Chapter 2, show the important influence of travelling waves and the consequently high values of the
peak of the transient recovery voltages across the circuit beaker involved.
Page 6
Chapter 3 presents detailed investigations on the influence of several technical solutions, as described before, on
the requirements for circuit breakers, especially on the specification of the transient recovery voltage (TRV). The
investigations included a large number of digital simulations of transients. To that purpose two network topologies
have been selected to represent radial networks and meshed networks, so that it can be verified whether radial
networks show more onerous transient recovery voltage waveforms compared to those in meshed networks.
Utilities’ policies with respect to series and shunt compensation, OHL phase transposition, single or double circuit
OHLs, tower configurations and surge arrester characteristics have all been included in the network models. The
influence of topology, parameters and variables has been investigated, such as fault location, unloaded line
switching with/without earth fault, line height and line sag, earth resistance and earth wires, load currents, degree
of shunt and series compensation, (non, beta, gamma) transposition, surge arrester characteristics, single versus
three phase faults. The relevant parameters known from the participating utilities (500 kV up to and including
1200 kV) have been considered in the evaluation works and the results have been compared. The characteristics
like tower configuration, transposition of phase conductors, shunt and series compensation have been adapted in
the whole network model. This means that by adapting one parameter many variables may be influenced, so that
the impact of that parameter may become less salient than expected. For instance a higher surge impedance
coincides with a higher impedance per km of all OHL in the model. Consequently, the fault current level will
decrease and the combined effects on the RRRV of a higher surge impedance and a lower short-circuit current
cancel each other to a certain degree.
Partly because of this effect, many parameters have a negligible influence, but overall it can be stated that the
results of simulations of terminal fault and long line clearing, as well as unloaded line switching, confirm that TRV
requirements for EHV and UHV circuit breakers are properly covered in the IEC and/or IEEE standards.
Applications with series capacitors need special consideration as already explained in Chapter 2.
Transformer limited faults (TLF) are addressed in Chapter 4. TLF are short-circuit conditions predominantly
determined by a transformer impedance in a given circuit. Seen from the circuit-breaker, the fault to be cleared may
be at the terminals of the transformer at the other side (transformer secondary fault: TSF) or the fault may be at the
busbar side (transformer fed fault: TFF). The voltage drop across the transformer is an important component of the
recovery voltage. A short-circuit current equal to 10% of the rated short-circuit current of the circuit breaker will give
a voltage drop of 90% and a short-circuit current of 30% will give a voltage drop of 70%. The first pole-to-clear
factor will be rather low, since the transformer itself with its low X0/X1-ratio will reduce the overall X0/X1-ratio that
determines the fault current.
For a simplified TRV determination a network can be represented by the transformer short-circuit inductance and
the capacitance of the transformer and the equipment between transformer and circuit breaker. The capacitance is
to be determined for the frequency range of interest; i.e. some kHz to some tens of kHz. The most onerous
waveform comes from situations with a minimum of capacitance between transformer and circuit breaker. In those
cases, the transformer response dominates the transient recovery waveform. Such a response is more complicated
than represented by a single frequency circuit as can be learned from frequency response analysis (FRA), a
diagnostic test method used to monitor a transformer’s condition.
Multiple resonance frequencies can be detected in the FRA-characteristic and a rather irregular curve at the higher
frequencies is likely to occur due to the interactions among inductances and capacitances between coils, windings,
leads, bushings, magnetic circuit and transformer tank. A more accurate method to calculate the waveform of the
transient recovery voltage requires taking into account the relevant resonances. This can be done by a simulation
with several parallel-resonance circuits (L//C//R) in series. Other, more detailed transformer models have been
used as well, based on special measurements and based on design characteristics of the transformer. Calculations
and simulations, though, show that the amplitude factor and the steepness of the waveform as calculated by the
simple single frequency response are more stringent than those of the more accurate calculations.
The consequences for the first and last pole to clear have been investigated and an overview of the minimum
capacitances between transformers and circuit breakers is given: 0.5 to 1.0 nF. All information collected forms input
for IEC SC 17A to implement a special test duty to cover TLF for circuit breakers with rated voltages from 100 kV
up to and including 800 kV connected to a transformer with a connection of low capacitance. However more
Page 7
statistical information is needed on transformer natural frequencies to give recommendations for the
standardization of the time to TRV peak.
High speed earthing switches (HSES) form the subject of Chapter 5. Single phase or multi-phase (i.e. multiple
single phase in a double circuit OHL) auto-reclosing schemes are applied for high-voltage transmission systems to
enhance system reliability. When an earth fault occurs, circuit-breakers located at both ends of the line clear the
fault by single pole tripping. After interruption of the short-circuit current, a small current remains flowing at the fault
point. This current is called secondary arc current and is caused by the electrostatic or electromagnetic coupling
with the other phase conductors. Generally it is not sure whether the secondary arc current will self-extinguish in a
short time. From a system stability point of view it is preferable to apply an auto-reclosing scheme with a reclosing
time in the order of 1 s maximum. To achieve auto-reclosing in due time some precautions are necessary to
extinguish the secondary arc before circuit breakers reclosure.
Especially for short distance OHL without shunt reactors or for double circuit systems, where four-legged shunt
reactors are not suitable, a very useful method is to apply a special earthing switch that by-passes the secondary
arc and leads to its extinction. This earthing switch is designed for high-speed operation to ensure that the required
switching performance within the dead time of the reclosing cycle is met, and is called high-speed earthing switch
(HSES). While an earthing switch as well as a fast acting earthing switch require the capability to withstand the full
short circuit current, the function of a HSES is to short-circuit and thereafter to clear the induced current and to
withstand the related TRV.
The requirements for a HSES are described in Chapter 5, taking also into account the possibility of a second earth
fault occurring right at the moment of induced current switching by the HSES. The probability of such incidents can
be learnt from thunder storms and the detection of multiple lightning strokes. The study gave input for the standard
developed by IEC SC 17A.
Disconnectors (DS) are subjected to bus transfer currents and bus charging current switching, while earthing
switches (ES) are subjected to electromagnetically and electrostatically induced current switching. In Chapter 6
attention is given to these switching duties of DS and ES. A phenomenon in relation with disconnectors in GIS is
that of very fast transient overvoltages (VFTO), that are caused by the travelling waves in short bus sections inside
a GIS. VFTO is addressed in earlier publications of WG A3.22 and in WG C4.306 Technical Brochure 542
“Insulation Coordination for UHV AC Systems”, June 2013. The topic has been investigated in depth by a
cooperation between WG A3.22/28, WG B3.29, WG C4.306 and AG D1.03 and the results are published in the
Technical Brochure 519 “Very fast Transient Overvoltages (VFTO) in Gas-Insulated UHV Substations”, December
2012, under coordination AG D1.03. A three step procedure to assess the risk of dielectric problems due to VFTO
has been established: (1) calculate peak value and rise time of the VFTO, (2) calculate the required VFTO
withstand strength, of the equipment involved by applying a co-ordination factor, a safety factor and a test
conversion factor to the LIWL and (3) definition of measures to be taken. An extensive elaboration of each step is
described in the Technical Brochure.
Simulations of disconnector switching and field data showed that the limitation to 1600 A for the bus transfer
current should be deleted. The limited statistical data provided in the report show that the 90 percentile of the ratio
between the maximum load current and the rating of a bay is between 0.5 and 0.8. Further statistical analysis is
required to assess DS bus transfer current switching duties in EHV systems. The specified bus-transfer voltage for
UHV (750 V) is at the low side and should be re-considered. A bus charging current for GIS applications evaluation,
based on substation layout, showed the need of 1 A for UHV.
The electromagnetically and electrostatically induced currents to be switched by earthing switches strongly depend
on the tower configuration of the OHL, and the load current in the circuit in operation. Besides, the induced voltage
(i.e. the recovery voltage) depends also on the line length. The rate of rise of the recovery voltage is determined by
travelling waves and is proportional to the induced current. Transposition of the phase conductors reduces the
induced currents and voltages to a large extent.
Many different combinations of earthed phases, at one side or at both sides, and tower configurations have been
simulated for rated voltages of 800 kV and above. In general, the electromagnetically induced currents and
voltages are proportional to the load current in the parallel circuit and tend to be larger at higher system voltages.
Page 8
The induced currents and voltages are larger than those of the electrostatically induced situations. Results show
induced currents of hundreds of Ampere (more than 1 kA for heavy loaded lines) and induced voltages of hundreds
of Volt for OHL with a length of hundreds of km. These values are much higher than stated in the IEC Standard.
The electrostatically induced currents and voltages reach up to tens of A and tens of V.
The field experience on switching behaviour in 800 kV and above systems is very limited. Investigation results from
China, Japan, Russia, USA and Canada, that are very important to better understand the UHV equipment
performance, are summarized in Chapter 7. The investigation includes transient recovery voltage, fault clearing,
shunt capacitor making and breaking experience, load current switching values (to calculate bus transfer currents),
induced current breaking, secondary arc current extinction and experience with VFTO. The investigation on service
experience with switchgear showed no failures due to overstresses by a transient recovery voltage, a bus transfer
current switching or an induced current interruption.
Due to the residual voltage of series capacitors (SC), the transient recovery voltage at clearing line faults in series
compensated lines is significantly higher than the peak value without SC, and the TRV may exceed the value in the
IEC standard. Hydro-Quebec has evaluated the risk and replaced circuit breakers in series compensated lines by
new ones with a higher TRV withstand capability. In the near future, with the application of reliable fast protective
devices by-passing the series capacitors during faults, Hydro-Quebec expects to specify standard TRV values for
series-compensated line circuit-breakers.
During the system commissioning of UHV AC demonstration project in China, switching operations have been
performed successfully. The switching operations included switching of 1000 kV unloaded transformer, of 110 kV
reactors and capacitors at the tertiary side, and of 1100 kV unloaded lines. System overvoltages and over-currents
showed enough margins compared with the design values. In the UHV AC demonstration project the effectiveness
of four-legged shunt reactors to extinguish a secondary arc in due time has been proven. The secondary arc
extinction time was less than 0.12 s.
Due to power system needs for steady state voltage profile control and power transfer, at UHV transformer’s
tertiary side four shunt capacitor banks are connected. The capacitor banks are switched on and off very frequently
and consequently for this important application the capacitive current switching duty must be performed with a high
number of operations. It has been proved that a circuit breaker with controlled switching could pass the
5000 operations of capacitive current switching.
The bus transfer current switching performance of an 1100 kV disconnector was verified during commissioning of
the UHV AC demonstration project, based on a requirement of 1600 A under 400 V. Experts in Japan collected
statistical information about weekday maximum load current field data of 550 kV DS and 275 kV DS versus the
rated bay current.
Experts in Japan carried out several tests on 126 km 275 kV line to study the performance of HSES for secondary
arc extinction. Results showed that, for 1LG or 2LG on double circuit lines HSES operated correctly, and the
secondary arc was extinguished within 50 ms.
In conclusion, WG A3.28 has shown that, apart from TLF and series compensated line faults, the present IEC and
IEEE Standards for circuit breakers properly cover the system requirements. Several mitigation methods exist to
reduce or deal with the TRV requirements of series compensated lines. For TLF a special test duty should be
introduced for circuit breakers with rated voltages from 100 kV up to and including 800 kV connected to a
transformer with a connection of low capacitance. It is proposed to apply a first pole-to-clear factor of 1.3 (EHV),1.2
(UHV) or 1.5 (non-effectively earthed neutrals), a voltage drop factor which is rather high (e.g. 90%) for low values
of the fault current (e.g. 10% of the rated short-circuit current) and rather low (e.g. 70%) for higher values of the
fault current (e.g. 30%), and an amplitude factor of 1.7. In this way the specified TRV peak are at the safe side.
Special attention is requested for the collection of statistical information on transformer natural frequencies to give
recommendations for the standardization of the time to TRV peak for TLF.
The results of the investigations on HSES formed direct input for standardization by IEC SC 17A. The requirements
for bus transfer current switching have to be adapted in the next revision of the standard for DS. This applies also
Page 9
for electromagnetically and electrostatically induced current switching by ES. For UHV a bus charging current of
1 A seems to fulfill the near future needs.
In addition, CIGRE WG A3.28 would like to draw CIGRE’s attention to a few still open topics, such as new
developments, application aspects and field experience of series and shunt compensation; field data about out-of-
phase switching; and more field data on e.g. VFTO measurements, TRV for TLF, bus transfer current by DS,
induced currents for ES, especially for UHV.
Page 10
1 Introduction
Major factors affecting the recent, and ongoing, development of AC transmission systems up to 1200 kV are the
environmental and economic scrutiny to which utilities are subjected. This has created a need to optimize proposed
infrastructure whilst maintaining expected levels of system performance. In response to these demands, the design
of compact, cost effective systems for bulk, long distance power transmission at the highest system voltages has
progressed significantly in recent years.
From 2007 to 2009, CIGRE WG A3.22 has extensively studied technical requirements for UHV equipment. Results
are given in Technical Brochures 362 [1] and 456 [2] and presented in a number of other publications during
conferences [3] to [10].
After experience in Russia with 1200 kV, in the 1980’s, and UHV pilots in Italy, USA and Japan, a significant event
occurred in January 2009 when the 1000 kV UHV AC Demonstration Project in China was put into operation with a
transmission capacity of 2800 MW.
Following on from the work published by WG A3.22, WG A3.28 addressed specific aspects of EHV and UHV
switching phenomena, equipment & testing in greater depth. The theoretical aspects and simulation of specific
phenomena were explored and correlated with service experience from existing systems and from early UHV
projects where available. The WG undertook a benchmark study of switching requirements based on model UHV
networks reflecting the utilities practices and policies (China, India, Japan, etc.) for UHV and EHV system
configurations.
Work was undertaken in three main areas:
1- Switching requirements of circuit breakers based on simulations with model UHV/EHV networks
- Transient recovery voltages (TRVs) for terminal faults, transformer limited faults (TLF) and long line faults (LLF) in
different UHV/EHV systems;
- Influence of system and equipment parameters on TRV;
- Unloaded line switching including the effects of shunt reactor and series capacitors;
- Testing requirements for UHV/EHV circuit breakers.
2- Switching requirements of disconnector (DS), earthing switch (ES) and high-speed eathing switch (HSES) based
on simulations.
- Switching requirements of DS and ES based on model UHV/EHV substations;
- Switching requirements of HSES based on model UHV/EHV networks;
- Technical comparison between four legged shunt reactor and HSES;
- Testing requirements for DS, ES and HSES.
3- Field experience and switching behaviour during and after commission
- Interrupting current and TRV after clearing a fault or an artificial grounding fault
- Interrupting current and TRV of DS, ES and HSES;
- Primary and secondary arc extinction behaviours in single/double circuits;
- Out-of-phase switching behaviour.
Page 11
Results from simulations and service experience were analyzed and recommendations made to standardization
bodies for specification and testing.
The WG gathered information related to field experience and switching behaviour during and after commissioning
of EHV and UHV systems in China, Japan, Russia and Canada and compared them with existing information from
other sources and studies.
The work was done by the 30 experts of WG A3.28, representing 15 countries, with contributions from coordination
participants from other WGs and Study Committees.
In Chapter 2 an overview will be given of the variety of system design at voltage levels of 800 kV and above. The
consequences of the different technical solutions on transient phenomena, especially on the TRV, have been
investigated by simulations by means of benchmark network models. The investigation and the results are
described in Chapter 3. Two specific cases are the specification of transformer limited faults and of the duties for
high speed earthing switches (HSES) which are addressed in the Chapters 4 and 5, respectively. The requirements
for disconnectors and earthing switches are treated in Chapter 6. Chapter 7 deals with the experience with
switching tests and operation, during and after commissioning. The conclusions are given in Chapter 8.
Page 12
2 Network configuration and technical solutions

2.1 Introduction
Long distance bulk transmission of electrical energy has forced (and forces) utilities in a number of countries to
implement or plan electrical grids with rated voltages equal or higher than 800 kV in AC and higher than 600 kV in
DC. With the introduction of these voltage levels, utilities optimize the network topology and the system design
criteria. Therefore, many different network topologies and system design policies exist around the world.
As the work of WG A3.22 was based on the design criteria of specific utilities (Japan, Russia, China), the need for
further investigations of the effects of individual design criteria on the TRV waveform and amplitude was felt. To
that purpose a large number of computer simulations have been performed, the description and results of which
will be treated in the next Chapter. Here, in Chapter 2, a general description of the relevant design criteria and the
related phenomena will be given.
2.2 Design criteria

Power systems show a large variation in the application of single or double circuit OHL (overhead line), of
transposition of the lines, of shunt and series compensation, of location and operation of shunt reactors, of auto-
reclosing and secondary arc extinction methods (single-phase/three-phase/multi-phase rapid auto-reclosure or
SPAR/TPAR/MPAR, four-legged shunt reactors, modified switching schemes for shunt reactors, HSES: high speed
earthing switches), of insulation coordination, etc. As transient phenomena are dependent on the network topology
and the system design, a number of representative bulk power transmission systems is given in the Tables 2.2.1
and 2.3.1.
According to IEC terminology, the rated voltage is the maximum continuous system voltage that is specified in
discrete values. Therefore, the standard definition related with a rated voltage of 800 kV includes systems
designated as 735 kV, 765 kV and 787 kV (nominal voltage). These voltages belong to the EHV-class. The class
UHV is used for power systems with a rated voltage above 800 kV AC; i.e. 1100 and 1200 kV. All these voltages
are typical for long distance bulk power transmission. The transmission corridors consist of two or three parallel
single circuit OHL or of double circuit OHL. For UHV level, the transmission capacity is typically 6000 to 7000 MVA
per circuit. The power generation sites are 10,000 to 15,000 MW or even greater. Three-phase transformer groups
have, for example, a capacity of 3000 to 4500 MVA.
Table 2.2.1 gives an overview of typical maximum line lengths (per section; i.e. between substations) for rated
voltages of 800 kV and above, the number of circuits per OHL, and the application of transposition of the phase
conductors, of shunt compensation (connected to the OHL), and of series compensation.
Country Utility Voltage Line length Single/double Transposition Shunt Series

(kV) (km) reactors Compensation
RSA ESKOM 800 440 Single Yes Yes Yes
Canada Hydro Québec 800 400 Single Yes Yes Yes
S. Korea KEPCO 800 160 Double No No No
USA AEP 800 300 Single No Yes No
Brazil Furnas 800 340 Single Yes Yes Yes
Russia EES Rossii 800 500 Single Yes Yes No
India PowerGrid 800 400 Single/Double Yes Yes No
Japan TEPCO 1100 200 Double No No No
China SGCC (pilot) 1100 358 Single Yes Yes Yes
China SGCC 1100 336 Double Yes Yes No
Russia EES Rossii 1200 700 Single Yes Yes No
India PowerGrid 1200 400 Single No Yes No
Table 2.2.1: Voltage level, line length and compensation parameters in several countries
Page 13
2.3 Secondary arc extinction

The reclosing policy and the secondary arc extinction methods are summarized in Table 2.3.1. As most faults in
OHL are of single phase to ground type, SPAR is preferred, although some utilities apply MPAR and utilities with
triple parallel circuits apply TPAR¹ without impairing system stability. In South Korea and Japan MPAR is applied to
double circuit OHL (in total six bundle-conductors for the 2 x 3 phases) in which all faulted bundle-conductors are
tripped and re-closed independently provided that at least 2 out of 3 phases remain connected.
Note 1: SPAR schemes increase quite significantly the risk of commutation failures in nearby HVDC inverter stations.
During SPAR and MPAR, the current flowing through the healthy phase conductors induces currents and voltages
in the switched off phase(s), thus maintaining the hot gas in the original fault channel and continues to conduct the
secondary arc current. At lower rated voltages (and shorter line lengths) there is a fair chance that the secondary
arc will self-extinguish in a short time, but at the highest voltage levels special measures have to be taken in order
to extinguish the secondary arc. If SPAR or MPAR is not successful all three poles of the circuit breakers will be
tripped subsequently.
Secondary arc extinction is largely influenced by the existence of a parallel circuit on the same towers, the length of
the line and the application of four-legged shunt reactors. four-legged shunt reactors – i.e. with a neutral reactor
that reduces the secondary arc current – are less effective for double-circuit and for non-transposed OHL. In that
case utilities may consider to apply TPAR or high-speed earthing switches (HSES), that, at both line ends, connect
the faulty phase to earth in order to divert the induced current from the secondary arc. These methods for
secondary arc extinction are reported to be very effective and allow short reclosing times [16] [17].
A solution for single circuit non-transposed OHL with a horizontal conductor configuration is AEP’s special
switching scheme as illustrated in Figure 5.1.3.3 (see Chapter 5).
th
Since January 6 , 2009, in China a single circuit OHL of 650 km with 3 substations is successfully in operation. The
1100 kV single circuit line is fully transposed along each section, so that four-legged shunt reactors could be
applied. Special earth fault tests showed that the secondary arc extinction was done in 118 ms in one line section
and in 42 ms in the other, leading to reclosing times to be set for 0.7 to 1.0 s.
Country Utility Voltage Re- four-legged Special shunt reactor HSES

(kV) closing1 shunt reactor switching scheme
RSA ESKOM 800 SPAR Yes No No
Canada Hydro Québec 800 TPAR No No No
S. Korea KEPCO 800 MPAR¹ No No Yes
USA AEP 800 SPAR Yes ² No
Brazil Furnas 800 TPAR No No No
Russia EES Rossii 800 SPAR Yes ³ No
India PowerGrid 800 SPAR Yes No No
Japan TEPCO 1100 MPAR No No Yes
China SGCC (pilot) 1100 SPAR Yes No No
China SGCC 1100 SPAR Yes No No
Russia EES Rossii 1200 SPAR Yes No4 No4
India PowerGrid 1200 SPAR Yes No No
Table 2.3.1: Voltage, auto-reclosing and secondary arc extinction practice
¹ SPAR = Single-phase auto-reclosure, TPAR = Three-phase auto-reclosure, MPAR = Multi-phase auto-reclosure
² AEP’s special switching scheme
³ Air blast breakers with external gaps across half of the arcing chambers so that at high overvoltages the shunt reactors are automatically
connected to the OHL
4
In Russia the AEP’s special switching scheme and HSES were both applied successfully in a few UHV test occasions, but no secondary arc
extinction method is used for UHV-lines, as they are operated nowadays at 525 kV
In order to avoid re-ignitions and restrikes, that may damage the shunt reactor and/or the circuit breaker, controlled
switching is applied to the shunt reactor circuit breakers. In addition MOSA, pre-insertion resistors or MOV across
the arcing chambers may be applied to limit overvoltages. The performance of a circuit breaker to interrupt small
Page 14
inductive currents, like the currents through shunt reactors, is covered by IEC Standard (Std) 62271-110. Shunt
reactors are normally connected to the line ends, mostly switchable, but may also be connected to the substation
busbars or even to transformer tertiary windings (or at the lower system voltages). Under light load conditions shunt
reactors are switched on and under heavy load conditions switched off. Before energizing and de-energizing an
OHL most utilities switch on the shunt reactors belonging to that line. At SPAR or TPAR, some utilities will delay the
healthy phase switching until the shunt-reactors are connected to all phases, some will switch on the shunt reactors
by the tripping signal to the line breakers and other utilities will automatically switch on the shunt reactors at a
certain overvoltage level. For secondary arc extinction the four-legged shunt reactors have to be connected to the
OHL, as it has to be tuned to that circuit. (In Table 2.3.1, under switched SR, the regular connection of shunt
reactors to the OHL is not meant, but the more special switching cases; see the notes.)
The HSES are special devices with particular requirements to switch induced currents. The most severe operating
conditions occur during thunderstorms, with multiple lightning strokes hitting successively different phase
conductors, for instance at the moment of opening of the HSES. Within IEC SC 17A, a standard for such devices is
under development (see Chapter 5).
2.4 Series capacitor banks

Series compensation is applied on normally loaded long transmission lines to increase the line transfer capability
(steady state load flow conditions) or on heavily loaded lines to reduce the voltage angle difference between their
two ends (transient stability conditions). Series capacitor banks are applied at 420, 550 and 800 kV levels (although
some lower voltage applications exist also). Investigations of the series compensation on UHV lines have been
performed by experts in China where a single circuit 1100 kV pilot OHL has been equipped with series capacitor
banks since 2011.
The normal procedure to clear an internal fault on the series compensated line is as follows.
There are two scenarios:
1. If fault current is lower than high current protection setting of MOV.
a. MOV operates instantaneously and may conduct to limit the overvoltage across the series
capacitor
b. Distance protection of line operates
c. Line circuit breaker clears the fault current about 50-70 ms after fault initiation
d. The by-pass switch may close 150 ms after fault initiation
e. The line circuit breaker recloses about 1s after fault initiation
f. If fault persists the line breakers trip three-phase.
2. If fault current is higher than high current protection setting of MOV
a. MOV operates instantaneously
b. MOV high current protection operates (in less than 5ms), triggers the spark gap and send
command to close the by-pass switch
c. Distance protection of line operates
d. Spark gap is triggered within maximum 20 ms and by-pass circuit breaker closes within 50ms
e. Line circuit breaker clears the fault current about 50-70 ms after fault initiation
f. The line circuit breaker recloses about 1s after fault initiation
g. If the reclosing was successful, the by-pass switch will open some 500 ms after the reclosure
of the line circuit breakers
h. If fault is of permanent nature, the line breakers trip three-phase.
Apart from the complexity of a series capacitor bank, experts are studying the impact of the capacitor on the line
circuit breaker current interruption performance. At fault current interruption, in principle the trapped DC voltage
across the series capacitor bank adds up to the TRV waveform, thus leading to higher than standard TRV peak
Page 15
values. A MOV across the capacitor bank, protected by a spark gap and a by-pass switch, reduces the DC-voltage,
but still the TRV peak value may become large. Moreover, at the source side, other series capacitors may add as
well to the TRV peak. Possible mitigations to cope with the TRV-stresses are a delay in fault clearing (to allow by-
passing the capacitor bank) or circuit breakers equipped with opening resistors/MOV across the arcing chambers.
Especial thyristor controlled capacitor banks or Fast Protection Devices (that, triggered by the line protection relay,
short-circuit very fast and reliably the capacitor bank [25] [31]) are developments that will reduce to a large extent
the effect of series compensation on circuit breakers stresses when clearing faults. Figure 2.4.1 is representative
for one phase of a 3-phase series capacitor bank, with its MOV, spark gap, current limiting reactor and by-pass
switch (not visible at the left side of picture) [19]. Another example is given in Figure 2.4.2 with two modules for
reliability reasons.
Figure 2.4.1: Hydro Québec single phase of a 735 kV series capacitor bank
The spark gap is triggered to protect the MOV or fired to eliminate the DC trapped charge on the series capacitor
bank. To protect the spark gap and to short-circuit the capacitor bank, a by-pass switch is used, but it closes only
after several cycles of the power frequency. As the by-pass switch is an inevitable element of the series capacitor
bank its specifications have been standardized in IEC Std. 62271-109 [13]. The Standard includes a number of
examples of applications of a by-pass switch. Examples of UHV applications, as used in China, still have to be
added.
Without by-passing the series capacitor bank in due time, high peak values of the TRV will occur when clearing line
faults. In both [19] and [23] it has been recommended to develop standard specifications for circuit breakers that
are applied for series compensated lines. Like with LLF or OP, the RRRV will be relatively low, say 1 kV/μs, and
the fault current between 10% and 30% of the rated short-circuit current. But the peak value may reach values as
high as 3.0 to 3.5 p.u.; values that may be regarded as covering most cases. These results are in line with [24],
where, in addition, it has been shown that by applying a Special Protection Device, TRV parameters are equal to
those for LLF, as covered in the Standards by test duties T10 and T30.
However, a spark gap triggered by the line protection can only by-pass the series capacitor bank if the voltage
across the bank is large enough. So, for line faults far away the capacitor bank will not be by-passed and the TRV
will become higher than without series capacitor bank. On the other hand, nearby faults will cause fault currents
large enough for the MOV to clip the voltage across the series capacitor and a conventional spark gap to by-pass
the capacitor anyway; [26] [27]. The Special Protection Device will by-pass the series capacitor bank independent
from the voltage across the bank. A third remark is that by the Special Protection Device the effect of series
capacitor banks at the source side, for instance in a parallel circuit, on the TRV cannot be avoided.
There is a need for further studies on the impact of series capacitor banks on temporary and switching
overvoltages, on TRVs and secondary arc extinction, especially for UHV [28]. The impact of countermeasures and
the (adapted) specifications for HV equipment, like by-pass switches [27], MOV, capacitors, four-legged shunt
reactors, disconnectors, by-pass gap, Fast Protection Device [26] [27], FACTS (Thyristors, SVCs) have to be
Page 16
investigated, as well as phenomena such as low frequency (15-30 Hz) and DC-components in fault currents.
Service experience with special equipment and detailed system studies are necessary [24].
Figure 2.4.2: Furnas single phase of a 800 kV series capacitor bank
2.5 Insulation coordination

The substation lay-out for the bulk power transmission systems of 800 kV AC and above is typically a 1½-breaker
or a 2-breaker scheme. The applied technology is AIS, GIS and H-GIS (or compact switchgear assemblies) and the
dimensions are quite large, so that phenomena related to the voltage drop across bus sections (bus transfer
current), related to travelling waves (ITRV) and related to the bus capacitance (TLF) cannot be disregarded. Due to
the large dimensions MOSAs have to be installed at several locations [2]
Insulation co-ordination is very important at the highest system voltages. Precise calculations and simulations of
slow front, fast front and very fast front overvoltages are required in order to optimize the margins in the
specification of insulation requirements as much as possible. Switching overvoltages, including the slow front
overvoltages due to the occurrence of faults, are important for air insulation (OHL, air gaps in substations). Several
measures can be taken to reduce these overvoltages: shunt reactors (as they reduce TOV), closing and opening
resistors, MOSA, controlled switching and transmission line arresters (TLA). The more compact designs of 800 kV
and UHV-lines and the mechanical stiffness of the heavy bundle-conductors (less contraction by the
electrodynamic forces of fault currents) lead to a lower inductance per km, a higher capacitance and a lower surge
impedance, in comparison with designs that are extrapolated from the lower rated voltages. As long as the impact
of the reduced inductances and the increased capacitances have a negligible impact on the amplitude of the short-
circuit current, the lower surge impedance leads to a reducing effect on the steepness of the switching
overvoltages and transient recovery voltages. By a proper system design and the application of modern MOSA the
SFO design criteria may become as low as 1.6 p.u. (at the line’s end).
Lightning overvoltages are controlled by an optimal design of the surge impedance of the phase conductors, of the
shielding by the earth wires, and of the transient voltage rise of the towers by their surge impedances and earth
resistances. As fast front overvoltages (FFO) are especially detrimental for the insulation of equipment such as
GIS-installations and power transformers high performance surge arresters are applied at strategic locations within
a substation. The V/I-characteristic of the multi-column MOSAs in combination with the number of MOSA in parallel
give a protection level that is relatively low in comparison to the protection level of MOSAs applied at lower rated
voltages; [2] [22]. By a proper system design and the application of modern MOSA the FFO design criteria may
become as low as 1.8 p.u.
Power frequency overvoltages (TOV), such as occur by load rejection or load rejection in combination with phase
to earth faults, are to a large extent controllable by shunt compensation and tele-tripping. Also the relatively low
X0/X1-ratio helps to reduce the TOV at healthy phases. MOSAs have to be specified to withstand the short-duration
TOV, this may force them to dissipate a large amount of energy. Hydro-Québec applies switchable MOSA that are
Page 17
sacrificed under the very special circumstances of system separations under heavy overloading. For UHV systems
TOV values are given as: 1.3 p.u. for 3 seconds, 1.4 p.u. for 0.44 seconds and 1.5 p.u. for 0.17 seconds [2].
Ferroresonance may lead to severe TOVs, but known precautions can be applied to avert risk of its occurrence.
This applies also for resonance due to inrush or magnetizing currents. Note that for the highest system voltages the
natural damping is even lower [29]. Special attention must be paid to ferroresonance phenomena when applying
SPAR in shunt compensated lines. For instance, the shunt reactors need a more linear magnetization curve in
comparison to transformers as they are subjected to more severe TOV (at the line’s end). Resonance and
ferroresonance may also occur when transformers at the end of OHL are energized by only two phases, as may
occur during the dead time of SPAR or by an incomplete operation of a circuit breaker (stuck pole) [30]. More
information is given in the annex.
Capacitive current switching of long unloaded lines differs from clearing the capacitive currents on small or medium
length lines, as the capacitive currents will be higher (certainly during disturbances or during the restoration of an
UHV transmission system), the induced voltages will be higher, the Ferranti effect more dominant and switching
under high TOV-conditions more probable. On the other hand, usually shunt compensation will be applied and
possibly also series compensation, both leading to less severe recovery voltage stresses for the circuit breaker
when switching off the long OHL.
2.6 Out-of-phase
Network developments caused by larger distances between power generation sites and load centres, by multiple
power transfers between regions, and/or between nations, and by the limited number of available corridors, lead to
power systems that are operated near to their stability limits. The long distances in combination with high loadings
of the OHL give a higher probability of stability problems and cascade effects, when faults or disturbances occur,
resulting in system separation.
Shortly before the system separation an out-of-phase condition may be faced. The point of equilibrium can be at a
power source (power plant) or on the OHL. In the latter case the RRRV of the circuit breaker that separates the
system will be determined by the equivalent surge impedance seen by the breaker and the out-of-phase current:
dIOP
RRRV  Z eq  . At the line side Zeq is equal to the line first pole-to-clear surge impedance (i.e. about 300 Ω for
dt
UHV) and the TRV builds up until the reflected travelling wave returns from the substation at the other side of the
OHL. The reflection coefficient depends on the number of connected OHL circuits at the other substation, in a
similar way as Zeq at the busbar side of the circuit breaker that depends on the number of connected circuits to the
busbar. IOP is determined by the out-of-phase angle between line ends, the busbar voltage and the line reactance.
For better understanding two simulated cases from the 1100 kV network in China are given [20] [21]
Case 1, single circuit 1100 kV line, 3 substations, out-of-phase condition, 2.73 kA, Figures 2.6.1 and 2.6.2. Involved
line length 281.3 km, with a positive reflection at the end gives an increase in steepness after approximately 2 ms,
as can be seen in Figure 2.6.2. At the source side the line length is 358.5 km and an increase in steepness is to be
expected after 2.39 ms, but is hardly visible due to the dominant influence of the natural frequency phenomena.
The initial steepness is 1300 kV within 2000 µs: 0.65 kV/µs; after 1.88 ms it rises to 1.3 kV/µs, as may be expected.
The initial steepness of 0.65 kV/µs with 2.73 kA corresponds to an equivalent surge impedance of 540 Ω, so that
the line surge impedance (in each direction) is 270 Ω, quite reasonable.
Figure 2.6.1: Case 1
Page 18
Figure 2.6.2: TRV waveform of Nanyang CB during out-of-phase while the oscillation center is
in Jingmen line
Case 2, double circuit 1100 kV line, 4 substations (Huainan, Wannan, Zhebei, Huxi), out-of-phase condition on line
Wannan – Huainan, length 336 km, 3.18 kA:
Figure 2.6.3: Case 2

The out-of-phase angle is 60°. The positive reflection on Huainan can be seen after 2.2 ms, corresponding to about
twice 336 km. At the source side the total line length is 317 km, so that the positive reflection here comes together
with that of the line with the oscillation centre (equilibrium point).
The line Wannan – Huainan with the oscillation centre has 2 circuits, but one circuit has already been opened, in
order to get the out-of-phase condition on the other circuit. The travelling waves seen by the circuit breaker in
Wannan, face the line surge impedance in the direction of Huainan and half the surge impedance into the direction
Zhebei, as here both circuits are in operation. From Figure 2.6.4 an initial steepness of the TRV of 0.542 kV/µs can
be deduced. With 3.18 kA this corresponds to an equivalent surge impedance of 383 Ω and a line surge impedance
of about 256 Ω, quite realistic for an 1100 kV double circuit overhead line.
The system natural frequency at the source side can be estimated by the following scheme: 2 transformers in Huxi,
each 0.4 H, connected through a 165 km double circuit OHL, 0.08 H, to Zhebei gives 0.28 H. In parallel to the
single transformer in Zhebei, 0.4 H, it gives 0.16 H, which is connected to Wannan through a 152 km double circuit
OHL, 0.08 H, gives 0.24 H. In parallel to the single transformer in Wannan, 0.4 H, the equivalent inductance is
about 0.15 H. The equivalent capacitance comes mainly from the lines, 320 km double circuit: 6 mF. The natural
frequency is about 170 Hz and half a period is 3 ms, as can be seen in Figure 2.6.4 (assuming a 1-cos shape).
Page 19
Figure 2.6.4: TRV in case 2 [6]

The examples show that out-of-phase conditions on long lines are determined by travelling wave phenomena and
system natural frequencies. The discussion on out-of-phase conditions focuses on the probability of out-of-phase
conditions, and the possible out-of-phase angles. System engineers and protection experts take out-of-phase into
consideration as they implement synchro-check and synchronisation equipment, as well as out-of-phase blocking
protection in addition to system wide protection. If circuit breakers have to be specified for out-of-phase conditions
the fore-mentioned phenomena have to be considered in detail.
Two examples of simulations of real out-of-phase cases in Japan and in Thailand are shown in Figures 2.6.5 and
2.6.6, respectively. Within a second, system disturbances lead to large phasor angles, unless special protection
systems prevent escalation by generator and/or load shedding.
Figure 2.6.5: Example of power swing simulation for severe system disturbance in case with
and without SSC (System Stabilizing Controller) in 275 kV system in Japan
Page 20
Figure 2.6.6: Example of simulation of an unstable power swing in Thailand
In the next Chapter the effect of the design parameters and the related phenomena will be elaborated in a
quantitative way. The importance of travelling waves will be confirmed [32].
Page 21
3 TRV of circuit breakers for EHV and UHV networks

3.1 Introduction
CIGRÉ WG A3.22 “Technical Requirements for Substation Equipment exceeding 800 kV” has studied transient
phenomena occurring in UHV AC networks, which are considered to be different from those in EHV-networks, and
recommended technical specifications for UHV substation equipment, including phenomena such as Initial
Transient Recovery Voltage (ITRV), line surge impedances of multi-bundle conductors, DC component in short-
circuit currents, secondary arc extinction, and very fast transient overvoltages (VFTO).
TRV recommendations to IEC SC17A were based on inherent TRV analysis for both bus terminal fault (BTF) duties
and those for Long Line Fault (LLF) considering electromagnetically and electrostatically induced voltages under
three-phase fault to ground (3LG) conditions in the 1100 kV transmission systems without MOSA using the system
and equipment parameters in Japan. Therefore, a remaining question was whether TRV requirements based on
the 1100 kV radial transmission systems in Japan could apply to transmission systems in other countries with
different compensation schemes and if TRVs for meshed systems are covered by the standard TRV.
For the purpose of explanation for the query, CIGRÉ WG A3.28 conducted TRV analysis with radial and meshed
network models using different system and equipment parameters and different compensation schemes surveyed
in various national projects of rated voltages from 550 kV to 1200 kV, since transient phenomena depend on the
network topology along with the system and equipment designs. Many different network topologies and system
design policies are applied to UHV and EHV transmission systems that show a large variation in the combination of
single or double circuit OHL, transposition of the lines, shunt and series compensation, location and operation of
shunt reactors, secondary arc extinction schemes (single or three-phase rapid auto-re-closing, 4-legged shunt
reactors, modified switching schemes for shunt reactors, high speed grounding switches: HSES), and insulation
coordination. The influence of various system and equipment parameters on TRV was also investigated in detail.
3.2 Radial and meshed network models used for TRV study
UHV and 800 kV networks generally feature long distance and bulk power transmission with long radial distance
lines. On the other hand, 550 kV, 420 kV and lower voltage networks are comparatively meshed with relatively
short lines. In addition, large capacity UHV transmission lines employ multi sub-conductor bundles, leading to less
damping of travelling waves, in comparison to the lower voltages. Furthermore, the radial network topology can
provide smaller short-circuit currents and TRV generated after a fault clearing leads to simple reflection patterns of
the travelling waves and therefore less damping due to refraction and distortion. In the future, UHV and 800 kV
transmissions could develop with a more meshed network topology and provide larger short-circuit currents,
occasionally applied with series capacitor compensation.
Considering these situations, both radial network and meshed network models are used to evaluate TRV with
various system and equipment parameters. The benchmark radial network model has 4 power sources and three
transmission lines of 120, 240 and 360 km lengths as shown in Figure 3.2.1. Fault points are located every 120 km
along the 360 km transmission line between B s/s and D s/s (s/s: substation), which provides TRVs for BTF and
LLF interrupting conditions.
The benchmark meshed network model has 6 power sources and 7 transmission lines as shown in Figure 3.2.2.
Fault points are provided at 120 km and 360 km transmission lines in the vicinity of B s/s.
Page 22
Figure 3.2.1: Radial network model with four power sources
50 kA 50 kA 50 kA
Tr: 2 units Tr: 2 units Tr: 2 units
E-s/s D-4 D-s/s F-s/s

120 km 240 km
360 km 360 km 360 km

LLF breaking for D-4 breaker
F11 at 0 km from B-2 CB, BTF breaking for B-4 breaker
3LG condition
120 km 240 km
A-s/s B-4 B-s/s C-s/s

Tr: 3 units Tr: 3 units Tr: 3 units
50 kA 50 kA 50 kA
Figure 3.2.2: Meshed network model with six power sources
3.3 System and Equipment parameters used for TRV study

Table 3.3.1 summarizes voltage, line-length and compensation schemes for different transmission systems
surveyed in cooperation with the utilities in China, India, Japan, Russia, the United States, Canada, South Africa,
South Korea and Thailand. WG A3.28 investigated TRV requirements based on these system parameters and
utilities policies.
These UHV and EHV systems show a large variation in the combination of single and double circuit OHL, line
transposition, shunt and series compensation, location and operation of shunt reactors, and secondary arc
extinction measures. UHV and EHV systems feature long distance bulk transmission with two or three single circuit
or double circuit OHL with multi-conductor bundle. The transmission capacity is typically 6000 to 7000 MVA per
circuit. Three-phase transformer groups typically show 3000 to 4500 MVA. The typical results of the survey on
equipment parameters such as power transformer, tower and conductors, and Metal Oxide Surge Arrester (MOSA)
are shown as follows.
Page 23
Table 3.3.1: System and compensation parameters in different counties

Note: SPAR: Single-phase Rapid Auto-reclosing, TPAR: There-phase Rapid Auto-reclosing, MPAR: Multi-phase Rapid Auto-reclosing
3.3.1 Transformer specifications

Tables 3.3.2 to 3.3.4 show typical specifications of three phase large capacity power transformers with primary,
secondary and tertiary windings used in the TRV study.
Country Japan China India

Rated capacity Primary 3000/3 3000/3 3000/3
(MVA) Secondary 3000/3 3000/3 3000/3
Tertiary 1200/3 1000/3 1000/3
Rated voltage Primary Star 1050/ 3 1050/ 3 1150/ 3
(kV)
Secondary Star 525/ 3 525/ 3 400/ 3
Tertiary Delta 147 110 33
Short-circuit Primary and secondary %Xps 18.5 18 18
impedance Primary and tertiary %Xpt 61.1 62 120
Primary capacity base Secondary and tertiary %Xst 34.3 40 60
Winding resistance Primary 0.49 ---
(ohm) Secondary 0.24 ---
Tertiary 0.08 ---
Capacitance Primary to ground Cpe 9000 9000 6000
(pF) Secondary to ground Cse 8000 15000 3000
Tertiary to ground Cte 24000 18000 12000
Table 3.3.2: Typical specifications of 1100-1200 kV/3000MVA transformers
Page 24
Country Canada Canada Canada

(MVA) Secondary 510/3 2000/3 1200/3
Tertiary 57 60/3 150/3
Rated voltage Primary Star 700/ 3 765/ 3 512.5/ 3
(kV)
Secondary Star 300/ 3 345/ 3 242/ 3
Tertiary delta 11.9 23 12.6
Short-circuit Primary and secondary %Xps 17.5 18 17.0
impedance Primary and tertiary %Xpt 41.7 200 141.3
Primary capacity base Secondary and tertiary %Xst 23.9 200 121.8
Winding resistance Primary X/R=48.5 0.21
(ohm) Secondary X/R=38.9 0.097
Tertiary X/R=23.1 0.0233
Capacitance Primary to ground Cpe 7200 (7000) 4000
(pF) Secondary to ground Cse 9600 (10000) 2500
Tertiary to ground Cte 12400 (20000) 12000
Note to Table 3.3.3: values in brackets are assumed
Table 3.3.3: Typical specifications of 765 kV 1650MVA/550 kV 1200MVA transformers
Country Thailand Thailand Thailand

(MVA) Secondary 1000/3 1000/3 1000/3
Tertiary 50/3 50/3 50/3
Rated voltage Primary Star 525 525 525
(kV) Secondary Star 242 242 242
Tertiary delta 22 22 22
Short-circuit Primary and secondary %Xps 17.0 16.45 16.68
impedance Primary and tertiary %Xpt 246.4 250 235
Primary capacity base Secondary and tertiary %Xst 220.0 230 220
Winding resistance Primary 0.1707 0.182 0.197
(ohm) Secondary 0.0600 0.078 0.0594
Tertiary 0.0620 0.022 0.0426
Capacitance Primary to ground Cpe 4950 13798 6700
(pF) Secondary to ground Cse 5940 16557 8040
Tertiary to ground Cte 10460 4611 8600
Table 3.3.4: Typical specifications of 550 kV1000MVA transformers
Page 25
3.3.2 Lower voltage system conditions
Short-circuit conditions in the lower voltage system was basically defined by a breaking current of 50 kA (full short-
circuit current) with a DC time constant of τ = L/R = 120 ms and TRV of T100 conditions, where RRRV = 2 kV/μs,
kpp = 1.3, and kaf = 1.4 with a 2 parameter TRV waveform. For example, in the case of the lower system voltage of
525 kV, associated with a higher system voltage of 1100 kV, the TRV peak is 784.9 kV and RRRV is = 2.04 kV/μs.
The zero sequence impedance was R0 of 0.52 Ω, and L0 of 62.72 mH and positive sequence impedance was R1 of
0.16 Ω and L1 of 19.30 mH. Other parameters were Re of 80 Ω, Ce of 1.85 μF and Cp of 0.02 μF.
In order to check the influence of short-circuit, short-circuit conditions at the lower voltage system were also given
by the breaking current of 15 kA with a DC time constant of τ = L/R = 120 ms and TRV of T30 conditions, where
RRRV = 5 kV/μs, kpp = 1.3, and kaf = 1.54 with a 2 parameter TRV waveform. For example, in the case of the lower
system voltage of 525 kV associated with a higher system voltage of 1100 kV, the TRV peak is 861.2 kV and
RRRV is 5.05 kV/μs. The zero sequence impedance was R0 of 1.74 Ω, and L0 of 209.07 mH and positive sequence
impedance was R1 of 0.54 Ω and L1 of 64.33 mH. Other parameters were Re of 480 Ω, Ce of 0.076 μF and Cp of
0.02 μF.
The voltage distribution in the network model was first checked under a no-load condition with the maximum short-
circuit current (50 kA) at the bus terminal in the lower voltage networks. Even in the no-load condition, the voltages
at the line end are increased due to the Ferranti effect. Therefore, voltages at the bus terminal are adjusted below
the maximum voltage by reducing the system voltage by about 5 % in the lower voltage networks. However, the
voltage increase due to the Ferranti effect attains about 20 % of the system voltage under no-load conditions with a
lower short-circuit current of T30 short-circuit current (15 kA) in the radial network model with a 360 km
transmission line. The adjustment of the lower system voltage by more than 20 % can significantly affect the TRV
amplitudes. Therefore TRV studies were decided to evaluate with the rated breaking current at the lower voltage
system in most cases.
3.3.3 Tower and conductor configurations

The configuration of UHV and EHV transmission towers with multi sub-conductor bundles employ either a
horizontal single circuit tower arrangement, often two or three parallel single circuits, or a vertical double circuit
tower arrangement. The tower normally has a couple of grounding wires on the top. A parallel single circuit and a
double circuit consisting of 6 phases were simulated by the J. Marti model. The circuit was transposed at 1/3 and
2/3 length of a transmission line. The sag, when considered, was settled in the average height of the conductors.
The earth resistivity is normally 100 Ω-m, 500 Ω-m was also used to confirm the influence on TRV.
Figures 3.3.1 to 3.3.4 show typical tower and conductor designs applied to 550, 800, 1100 and 1200 kV single and
double circuit transmission in China, India, Japan, Canada, Korea and Thailand. Tables 3.3.5 to 3.3.8 show
detailed configurations of multi-bundle conductors.
Figure 3.3.1: 1100 kV double circuit tower
Page 26
Dimension Japan China

Conductor Diameter of conductor D (cm) 3.84 3.36
Thickness T (cm) 1.44 -
Ratio (T/D) 0.375 0.375
DC Resistance at 20 degree Celsius (Ù/km) 0.0356 0.04633
Number of conductors in the bundle 8 8
Bundle Spacing (cm) 40 40
Sag of the Line (m) 20 20
Grounding Diameter of grounding wire D (cm) 2.95 1.75
wire Thickness T (cm) 1.15 -
Ratio (T/D) 0.39 (0.5)
Number of conductors in the bundle 1 -
Bundle Spacing (cm) - -
Sag of the Line (m) 18 18
Table 3.3.5: Dimension of 1100 kV double circuit tower
Figure 3.3.2: 1200 kV/765 kV single circuit tower
Dimension India Canada

Thickness T (cm) - 1.26
Ratio (T/D) 0.391 0.36
Bundle Spacing (cm) 45 45.7
Sag of the Line (m) 0 7.6
Grounding Diameter of grounding wire D (cm) 1.9 1.27 / 2.29
wire Thickness T (cm) - -
Ratio (T/D) 0.329 -
DC Resistance at 20 degree Celsius (Ù/km) 0.221 3.1 / 0.137
Number of conductors in the bundle - 1/1
Sag of the Line (m) 0 11.7 / 11.7
Table 3.3.6: Dimension of 1200 kV/ 765 kV single circuit tower
Page 27
Figure 3.3.3: 550 kV and 800 kV double circuit tower
Dimension Thailand Korea

Thickness T (cm) 1.4495 1.014
Ratio (T/D) 0.427 0.333
Bundle Spacing (cm) 45.7 40
Sag of the Line (m) 18.06 13.41
Grounding Diameter of grounding wire D (cm) 1.05 1.85
wire Thickness T (cm) 0.262 -
Ratio (T/D) 0.250 0.284
Sag of the Line (m) 12.37 16.00
Table 3.3.7: Dimensions of 550 kV and 800 kV double circuit tower
Figure 3.3.4: 550 kV single circuit tower
Page 28
Dimension Canada Thailand

Thickness T (cm) - 1.4495
Ratio (T/D) 0.333 0.427
Bundle Spacing (cm) 45.72 45.7
Sag of the Line (m) 4 18.06
Grounding Diameter of grounding wire D (cm) --- 1.05
wire Thickness T (cm) --- 0.262
Ratio (T/D) --- 0.250
DC Resistance at 20 degree Celsius (Ù/km) --- 0.864
Number of conductors in the bundle --- 1
Bundle Spacing (cm) --- -
Sag of the Line (m) --- 12.37
Table 3.3.8: Dimension of 550 kV single circuit tower
3.3.4 Phase arrangement of the transmission lines

In the TRV analysis, 120 km, 240 km and 360 km transmission lines are gamma (half-) transposed twice to
minimize the influence of current and transient phenomena among the phases depending on the position and the
height of conductors. The influence of different transposition on TRV is described in 3.7.8.
Figure 3.3.5: Transposition for 120 km transmission line
Page 29
K-BUS
K-BUS
Z-BUS
Z-BUS
K-BUS
K-BUS
Z-BUS
Z-BUS
3.3.5 Modelling of bus terminal

Since TRV requirements for TLF conditions strongly depend on the capacitance between a power transformer and
a circuit-breaker, it is very important to have a precise estimation of capacitances between windings of power
transformers as well as those between windings and the ground.
WG A3.28 conducted a limited survey on the minimum capacitance between a power transformer and a circuit-
breaker. Several capacitances were measured with different large capacitor power transformers by Frequency
Response Analysis as well as the Daini-Kyodai method (see Chapter 4) as summarized in Table 3.3.9. Estimated
values of capacitances are given between brackets in Table 3.3.9.
Page 30
Country China Japan Canada Canada Thailand

Rated voltage (kV) 1100 1100 765 550 550
Primary winding to ground (pF) (9000) 9000 5680 (4000) 4950
Secondary winding to ground (pF) (15000) 30000 7570 (2500) 5940
Tertiary winding to ground (pF) (18000) 24000 18200 (12000) 10460
Between Trans. and CB at Trans. (pF) 5800 1750 700 (1600) 1600
(3000) (1620)
Bus terminal including CVT (pF) 8200 10000 7400 3000 22000
Between CB at line side and line (pF) 10000 --- --- --- 1500
Table 3.3.9: Bus terminal capacitance
3.3.6 MOSA parameters

The influence of MOSA on TRV for terminal fault duties was analyzed using different V-I characteristics of MOSAs.
The application of MOSAs at a rated voltage of 1100 kV and above could reduce the amplitude factor of TRV for
BTF, TLF and out-of-phase due to reduced restriction voltage of UHV class MOSAs. However, it may not lead to a
reduction large enough for LLF where the TRV is generated at both the source and line sides of the breaker
terminal. Depending on the V-I characteristic of the MOSA, the influence on TRV for BTF and TLF duties is also
investigated at a rated voltage of 800 kV and the results are summarized in sub-clause 3.4.8.
I (A) 1 10 100 1000 2000 5000 10000 20000 40000

V (kV) 1225 1260 1315 1415 1450 1505 1550 1620 1705
Table 3.3.10: V-I characteristic of 1100 kV MOSA in Japan
I (A) 1 10 100 500 1000 2000 5000 10000 20000
V (kV) 1270 1295 1310 1336 1406 1460 1505 1554 1620
Table 3.3.11: V-I characteristic of 1100 kV MOSA in China
I (A) 1 10 100 500 1000 2000 5000 10000 20000
V (kV) --- --- --- 1380 1440 1500 --- 1600 1700
Table 3.3.12: V-I characteristic of 1200 kV MOSA in India
I (A) 1 10 100 500 1000 2000 5000 10000 20000
V (kV) 878.5 921.8 977.7 1041.3 1076.9 1120.1 1202.8 1271.4 1410.0
Table 3.3.13: V-I characteristic of 765 kV MOSA in Canada
I (A) 1 100 500 1000 2000 5000 10000 20000 40000
V (kV) --- --- 1210 1230 --- 1280 1320 1400 1480
Table 3.3.14: V-I characteristic of 800 kV MOSA in Korea
I (A) 1 125 250 500 1000 2000 3000 5000 10000
V (kV) --- 735 740 772 793 816 838 859 898
Table 3.3.15: V-I characteristic of 550 kV MOSA in Canada
I (A) 1 100 1500 3000 5000 10000 15000 20000 40000
V (kV) (800) --- 936 963 1006 1070 1112 1177 1310
Table 3.3.16: V-I characteristic of 550 kV MOSA in Thailand
Page 31
3.4 Comparison of TRV evaluated in a radial network model with that in Japan’s 1100 kV
network
In order to confirm the effectiveness of TRV evaluation in the radial and meshed network models, the TRVs were
calculated in the radial network model which consists of double circuit lines with a line length of 120, 240 and
360 km. System and equipment parameters for the radial system are assumed to be the same as in the Japanese
system to allow a good comparison with the TRV analysis for the 1100 kV transmission system in Japan which
consists of double circuit lines with line lengths of 40, 50, 138 and 210 km as shown in the Figure 3.4.1.
Figure 3.4.2 shows that TRVs calculated for the radial network model. The TRV waveforms show rather simple
triangular shape. However, both TRV peak and RRRV values show good agreement with those analyzed in the
Japanese 1100 kV transmission system, despite the difference in transmission line lengths.
Therefore the TRV analysis in the network models can be considered as an effective method to evaluate TRV
requirements for different system and equipment parameters and can be applicable to investigate the influence of
system and equipment parameters on TRV for various switching duties.
In the next sub-clause, it will be confirmed that TRV requirements for UHV ratings recommended based on Japan’s
1100 kV transmission systems can also be applicable to the TRV calculated based in the 1100 kV and 1200 kV
radial and meshed network models with system and equipment parameters in India, China and Japan. Then the
method will be applied to evaluate whether TRVs calculated in 550 kV, 800 kV and 1200 kV radial and meshed
network models with system and equipment parameters can support the existing IEC standards.
Figure 3.4.1: Radial network model and 1100 kV transmission system in Japan
2500 2500
2000 2000
1500 1500
TRV (kV)
TRV (kV)
1000 1000
500 500
0 0
0 0.5 1 1.5 2 2.5 3 0 0.5 1 1.5 2 2.5 3
Time (ms) Time (ms)
Figure 3.4.2: TRVs comparison calculated in the radial network model and in the 1100 kV
transmission system in Japan
Page 32
3.5 TRV for bus terminal faults and long-line faults in radial network model
3.5.1 Introduction
Inherent TRV analysis for bus terminal fault (BTF) duties and those for long-line-fault (LLF) were investigated with
parallel single circuit or double circuit radial network models (without MOSAs) using different system and
equipment parameters for different national projects at rated voltages of 550, 765, 1100 and 1200 kV.
Since the radial network model is composed of three double circuit transmission lines with line lengths of 120 km,
240 km and 360 km, each inflection point observed on the TRV waveforms was analyzed to clarify the reflected
point of the propagated waveform.
Voltage across circuit-breaker

2.15ms Vs : Source voltage
1.29ms
Breaking point
(A)
(B) (C) (D) (E)
2.58ms VL : Line voltage
3.01ms
Figure 3.5.1: Radial network model and 1100 kV transmission system in Japan
Point A: Breaking point of D-s/s CB
Point B: Arrival of transient propagated from B s/s to D s/s with 360 km travel at 1.29 ms after breaking
Point C: Arrival of transient propagated from B s/s to A s/s and back to B s/s, then from B s/s to D s/s, total wave
travel is 120 km x 2 (0.43 ms x 2) + 360 km (1.29 ms) or 600 km, at 2.15 ms after breaking
Point D: Arrival of transient propagated from D s/s to B s/s and back to D s/s, total wave travel is 360 km x 2
(1.29 ms x 2) or 720 km, at 2.58 ms after breaking
Point E: Arrival of transient propagated from B s/s to C s/s and back to B s/s, then from B s/s to D s/s, total wave
travel is 240 km x 2 (0.86 ms x 2) + 360 km (1.29 ms) or 840 km, at 3.01 ms after breaking, assuming a
propagation velocity of 280 m/µs
The propagation velocity of the transients can generally be given by 300 m/µs through transmission line. The
propagation speed of zero-sequence travelling wave is slightly reduced to 280 m/µs, since it is influenced by the
earth resistivity. So, a velocity of 280 m/µs fits better for calculations of the transient phenomena.
In the following, for the purpose of comparing calculated TRVs with standard TRVs the shortest time t2 defined for
OP in IEC Std 62271-100 is used.
3.5.2 550 kV double circuit radial network model

Figure 3.5.2 shows the calculated TRV peak and RRRV for BTF and LLF conditions using the system and
equipment parameters in Thailand. Figure 3.5.3 shows typical TRV waveforms calculated in the 550 kV radial
network model. Most of the TRV peak and RRRV are covered by TRV requirements for test duties T10, T30, T60
and T100 standardized in IEC Std 62271-100. A few TRV peaks exceed by about 4% (uc = 1071 kV at maximum)
the value for T10 (uc = 1031 kV) and the breaking currents are smaller than 5 kA, corresponding to TRV in LLF
conditions generated 360 km from the circuit-breaker. However, these values can be covered by the TRV specified
for the out-of-phase duty (uc = 1123 kV).
Page 33
1200 8
Rate of rise of TRV (kV/μs)

1000 7
TRV peak (kV) 6
800 5
LLF LLF
600 4
TLF TLF
400 BTF 3 BTF
2
200 1
0 0
0 10 20 30 40 50 0 10 20 30 40 50
Breaking current (kA) Breaking current (kA)
Figure 3.5.2: TRV for BTF and LLF conditions in 550 kV radial network model with shunt reactor
using system and equipment parameters in Thailand
1200
1200
1000
1000
800
800
TRV (kV)
TRV (kV)
600
600
400
400
200
200
0
0
0 0.5 1 1.5 2 2.5 3 3.5
0 0.5 1 1.5 2 2.5 3 3.5
Time (ms) Time (ms)
Figure 3.5.3: TRV waveforms for LLF conditions in 550 kV radial network model at CB located in
B s/s (left) and in D s/s (right). The breaking currents are less than 5kA

equipment parameters in the Korean network. Figure 3.5.5 shows typical TRV waveforms calculated with the
800 kV radial network model. Most of the TRV peak and RRRV are covered by TRV requirements for test duties
T10, T30, T60 and T100 standardized in IEC Std 62271-100. A few RRRV values (5.4 kV/µs) slightly exceed the
standard value for T30 (5 kV/µs). They correspond to TLF conditions with breaking currents smaller than 10 kA.
TRV for TLF are introduced in Edition 2.1 of IEC Std 62271-100 for rated voltages 1100 and 1200 kV but not yet for
rated voltages 100 kV to 800 kV. TRV for TLF conditions in the IEC standard will be considered for 800 kV and
lower voltages based on the investigations by CIGRE WG A3.28 including TRV measurements. The TRV
requirements for TLF conditions will be described in detail in Chapter 4.
2000 8
7
1500 6
TRV peak (kV)
5
LLF
1000 LLF
4
TLF
TLF 3 BTF
500 BTF 2
1
0 0
0 10 20 30 40 50 0 10 20 30 40 50
using system and equipment parameters in Korea
Page 34
2000 2000
1500 1500
TRV (kV)
TRV (kV)
1000 1000
500 D-S/S 500
F24
0 0
0 0.2 0.4 0.6 0.8 1 0 0.5 1 1.5 2 2.5 3 3.5
Time (ms) Time (ms)
Figure 3.5.5: TRV waveforms for LLF conditions in 800 kV radial network model at CB located in
B s/s (left: less than 5 kA) and in D s/s (right: less than 10 kA)

equipment parameters in the Chinese network. Figure 3.5.7 shows typical TRV waveforms calculated in the
1100 kV radial network model. Most of the TRV peak and RRRV values are covered by TRV requirements for test
duties T10, T30, T60 and T100 standardized in IEC Std 62271-100. A few RRRV values (5.4 kV/µs) slightly exceed
the standard value for T30 (5 kV/µs), However, the values are covered by newly standardized RRRV of 17.2 kV/µs
at 12.5 kA for TLF conditions with a rated voltage of 1100 kV. TRV peak for LLF condition at 240 km distance from
the circuit-breaker shows a TRV value of 1660 kV that is 2 % higher than the standard value for T30, it is also
covered by the standard TRV (uc = 2245 kV) for the out-of-phase duty.
2500 8
7
2000
6
TRV peak (kV)
1500 5
LLF
LLF 4
TLF
1000 TLF
3 BTF
BTF
2
500
1
0 0
0 10 20 30 40 50 0 10 20 30 40 50
using system and equipment parameters in China
2500 2500
2000 2000
1500
T R V (kV)
1500
T R V (kV)
1000 1000
500 500
0 0
0 0.5 1 1.5 2 2.5 3 3.5 0 0.5 1 1.5 2 2.5 3 3.5
Time (ms) Time (ms)
Figure 3.5.7: TRV waveforms for LLF conditions in 1100 kV radial network model at CB located
in B s/s (left: less than 5 kA) and in D s/s (right: less than 15 kA)
Page 35

equipment parameters in the Indian network. Figure 3.5.9 shows typical TRV waveforms calculated in the 1200 kV
radial network model. Most of the TRV peak and RRRV values are covered by TRV requirements for test duties
T10, T30, T60 and T100 standardized in IEC Std 62271-100. An RRRV of 5.9 kV/µs is obtained when one circuit is
open. The values are covered by newly standardized RRRV of 18.0 kV/µs at 12.5 kA for TLF conditions with a
rated voltage of 1200 kV.
3000 8

2500 7
6
TRV peak (kV)
2000 5
LLF
1500 4
LLF TLF
1000 TLF 3 BTF
BTF 2
500 1
0 0
0 10 20 30 40 50 0 10 20 30 40 50
using system and equipment parameters in India
3000 3000
2500 2500
2000 2000
T R V (kV)
T R V (kV)
1500 1500
1000 1000
500 500
0 0
0 0.5 1 1.5 2 2.5 3 3.5 0 0.5 1 1.5 2 2.5 3 3.5
Time (ms) Time (ms)
Figure 3.5.9: TRV waveforms for LLF conditions in 1200 kV radial network model at CB located
in B s/s (left: less than 5 kA) and in D s/s (right: less than 15 kA)
It is shown that TRV for BTF and LLF conditions calculated in the radial network models can be covered by TRV
requirements specified in IEC for rated voltages from 800 kV to 1200 kV using the system and equipment
parameters for different national projects. Even though TRV peaks for 360 km LLF conditions may exceed the
standard values, e.g. for 360 km LLF conditions, they can be covered by the TRV requirements for the out of phase
condition.
3.6 TRV for no-load line switching duty

3.6.1 Introduction
TRV for no-load line breaking in the normal service condition without any fault on other line(s) was investigated in
the radial network model using different system and equipment parameters (considering a practical configuration of
a transmission tower and multi-sub conductor bundle). Figure 3.6.1 shows the radial network model used to
evaluate TRV requirements for no-load switching conditions. The requirements under normal service were
calculated with the other line side being open. The highest TRV peak appears when a fault occurs at the remote
end.
Page 36
Figure 3.6.1: Radial network model to study TRV for no-load switching condition
Case 1: No-load line breaking between Bs/s and As/s (CB of As/s in open position)
Case 2: No-load line breaking between Bs/s and Cs/s (CB of Cs/s in open position)
Case 3: No-load line breaking between Bs/s and Ds/s (CB of Ds/s in open position)

Figure 3.6.2 shows the calculated TRV peak and breaking current for no-load switching conditions using the
system and equipment parameters in Thailand. The TRV peak is covered by the standard value based on a voltage
factor of 1.2 in IEC Std 62271-100. The breaking current slightly exceeds the preferred value of 500 A given in IEC
Std 62271-100 when the line length is longer than 360 km. The interrupting current is 552 A under normal service
conditions and 559 A for the single-line-to-ground fault (1LG) condition.
(a) TRV peak and Breaking current for different transmission lines
Page 37
(b) TRV peak for different distances from CB to fault point at the line between B s/s and D s/s
Figure 3.6.2: TRV for no-load line breaking conditions in 550 kV radial network model without
shunt reactor using system and equipment parameters in Thailand
Figure 3.6.2 (b) shows the TRV peak for different distances from the CB to the fault location on the transmission
line between B s/s and D s/s. When the distance from CB to the fault point is longer than 120 km, TRV peak is only
slightly increased with an increase of the distance from the CB to the fault point. Therefore the longer distance
provides a more severer TRV peak for no-load line breaking under 1LG condition.
Figure 3.6.3 shows the typical TRV waveforms for no-load line breaking under normal service without faults on the
lines. The difference in TRV peak and breaking current with and without fault conditions seems to be negligible,
even though TRV waveforms for an faulted phase under 1LG condition shows an oscillation with triangular shape.
Voltage factor= 1.13 (at 1018kV ) Voltage factor= 1.13 (at 1014kV )
Voltage between CB’s
terminals (kV)
terminals (kV)
Capacitive breaking current Capacitive breaking current for

= 552A rms healthy-phase= 559A rms
current (kA)
current (kA)
Breaking
Breaking
voltage (kV)
voltage (kV)
Source-side
Source-side
voltage (kV)
voltage (kV)
Line-side
Line-side
Figure 3.6.3: TRV waveforms in 550 kV radial network model under normal service condition
without fault (left) and under 1LG condition (right)
Figure 3.6.4 shows the typical TRV waveforms for no-load line breaking under normal service with and without
shunt reactors. The first TRV peak and the breaking current are relatively lower but the subsequent TRV peak
becomes higher, when a shunt reactor is connected.
Page 38
Voltage factor= 1.13 (at 1018kV )

terminals (kV)
terminals (kV)
Capacitive breaking current Capacitive breaking current
= 552A rms = 290A rms
current (kA)
current (kA)
Breaking
Breaking
voltage (kV)
voltage (kV)
Source-side
Source-side
voltage (kV)
voltage (kV)
Line-side
Figure 3.6.4: TRV waveforms in 550 kV radial network model under normal service condition Line-side
without shunt reactor (left) and with shunt reactor (right)
Figure 3.6.5 shows the initial TRV for no-load line breaking under normal service and 1LG condition with line
lengths of 120, 240 and 360 km. The deformation of the initial TRV (ITRV) becomes prominent with an increase of
the line length and more prominent at 1LG condition. They exceed a TRV with a (1-cos) waveform and a voltage
factor of 1.2 due to the propagation of transients caused by a fault.
The ITRV continues for a few milliseconds after the interruption, which tends to increase the minimum arcing time
resulting in mitigating the switching duties for the circuit breaker. TRV peak for healthy lines (without a fault and
with 1LG conditions) can be covered by (1-cos) waveform with a capacitive voltage factor of 1.2.
TRV peak (1257kV) for voltage factor of 1.4 TRV peak (1257kV) for voltage factor of 1.4
TRV peak (1078kV) for TRV peak (1078kV) for

voltage factor of 1.2 voltage factor of 1.2
(a) Initial TRV with 120 km line length under normal service condition (b) Initial TRV with 240 km line length under normal service condition
ITRV by traveling- ITRV by traveling-

wave on the line wave on the line
(c) Initial TRV with 360 km line length under normal service condition (d) Initial TRV with 360 km line length under 1LG condition
and 1LG condition for different line length.
Page 39
3.6.3 800 kV parallel single circuit radial network model

Figure 3.6.6 shows the calculated TRV peak and breaking current for no-load switching conditions using the
system and equipment parameters in Canada. The TRV peak is covered by the standard value based on a voltage
factor of 1.2 in IEC Std 62271-100. The interrupting current slightly exceeds the preferred value of 900 A when the
line length is longer than 360 km. The breaking current is 926 A in the 1LG condition.
Figure 3.6.6: TRV for no-load line breaking conditions in 800 kV radial network model without
shunt reactor using system and equipment parameters in Canada
Figure 3.6.7 shows typical TRV waveforms for no-load line breaking under normal service with and without shunt
reactors. The first and subsequent TRV peaks that appear immediately after clearing become lower without a shunt
reactor and the breaking current is also smaller than that with a shunt reactor.

terminals (kV)
terminals (kV)

= 841A rms = 403A rms
current (kA)
current (kA)
Breaking
Breaking
voltage (kV)
voltage (kV)
Source-side
Source-side
voltage (kV)
voltage (kV)
Line-side
Line-side
without shunt reactor (left) and with shunt reactor (right)
Figure 3.6.8 shows the initial TRV for no-load line breaking under normal service and 1LG condition with line
lengths of 120, 240 and 360 km. The deformation of the initial TRV becomes prominent with an increase of the line
length and more prominent at the 1LG condition. They exceed a TRV with a (1-cos) waveform and a voltage factor
of 1.2 due to the propagation of transients caused by a fault. ITRV continues for a few milliseconds after the
interruption.
Page 40
TRV peak (1499kV) for
voltage factor of 1.2 TRV peak (1499kV) for
voltage factor of 1.2

and 1LG condition for different line length.

Figure 3.6.9 shows the calculated TRV peak and breaking current for no-load line breaking conditions using the
system and equipment parameters in Japan. The TRV peak for both normal service and 1LG conditions are
covered by the standard TRV with a voltage factor of 1.2. The breaking current is also covered by the preferred
value of 1200 A when the line is shorter than 360 km.
2400 1800
Uc=2155 kV for 1100 kV (voltage factor=1.2)
2000 1500
Breaking current of 1200 A for 1100 kV LC duty
1600 TRV for first-pole-to-clear 1200
1200 900
800 600 TRV for first-pole-to-clear
TRV at normal service condition (without fault)

400 300 TRV at normal service condition (without fault)
TRV at 1LG condition
TRV at 1LG condition
0 0
0 120 240 360 0 120 240 360
Line length (km) Line length (km)
Figure 3.6.9: TRV for no-load line breaking conditions in 1100 kV radial network model using
system and equipment parameters in Japan
Figure 3.6.10 shows the typical TRV waveforms for no-load line breaking under normal service without faults on the
lines. The difference in TRV peak and breaking current with and without fault conditions seems to be negligible,
even though TRV waveforms for a faulted phase under 1LG condition show an oscillation with a triangular shape.
Page 41

terminals (kV)
terminals (kV)
current (kA)
current (kA)
Breaking
Breaking
voltage (kV)
Source-side
voltage (kV)
Source-side
voltage (kV)
voltage (kV)
Line-side
Line-side
Figure 3.6.11 shows the initial TRV for no-load line breaking under normal service and 1LG condition with a line
length of 360 km. The deformation of the initial TRV becomes prominent with an increase of line length and more
prominent at 1LG condition. However, the TRV is covered by the standard TRV with a 1-cos waveform and a
voltage factor of 1.2. ITRV continues for a few milliseconds after the interruption.

Figure 3.6.11: TRV waveforms in 1100 kV radial network model under 1LG condition
3.6.5 1200 kV parallel single circuit radial network model

Figure 3.6.12 shows the calculated TRV peak and breaking current for no-load line breaking conditions using the
system and equipment parameters in India. The TRV peak for both normal service and 1LG conditions are covered
Page 42
by the standard TRV with a voltage factor of 1.2. The breaking current is also covered by the preferred value of
1300 A when the line is shorter than 360 km.
Figure 3.6.12: TRV for no-load line breaking conditions in 1200 kV radial network model
without shunt reactor using system and equipment parameters in India
Figure 3.6.13 shows the typical TRV waveforms for no-load line breaking under normal service without faults and
under 1LG condition on the lines. The difference in TRV peak and breaking current with and without fault conditions
seems to be negligible, even though the TRV waveforms for a faulted phase under 1LG condition shows an
oscillation with a triangular shape.
terminals (kV)
terminals (kV)

current (kA)
current (kA)
Breaking
Breaking
Source-side
voltage (kV)
Source-side
voltage (kV)
voltage (kV)
voltage (kV)
Line-side
Line-side
Figure 3.6.14 shows the initial TRV for no-load line breaking under normal service and 1LG condition with a line
length of 360 km. The deformation of the initial TRV becomes prominent with an increase of line length and more
prominent at 1LG condition. However, the TRV is covered by the standard TRV with a 1-cos waveform and a
voltage factor of 1.2. ITRV continues for a few milliseconds after the interruption.
Page 43


and 1LG condition for different line length
In the case of no-load line-charging current interruption for both normal service and single-phase line fault to
ground (1LG) conditions the peak value of TRV is covered by the standard TRV with a voltage factor of 1.2. But in
case of a single-phase line fault the waveform after say 5 ms requires a larger voltage factor in order to be covered
by a 1-cos waveform. The breaking current is covered by the standard value of 1200 A for 1100 kV and 1300 A for
1200 kV, when the line is shorter than 360 km.
ITRV is observed during a few milliseconds after the interruption, which tends to increase the minimum arcing time
resulting in a less severe switching duty for circuit breaker.
3.7 Influence of system and equipment parameters on TRV

3.7.1 Introduction
The TRV analysis in the network models can be considered as an effective method to evaluate TRV requirements
for different system and equipment parameters and can be used to investigate the influence of system and
equipment parameters on TRV for various switching duties.
Therefore, CIGRÉ WG A3.28 investigated the influence of various system and equipment parameters listed below
on TRV requirements for BTF, LLF, TLF, out-of-phase and no-load line-charging current breaking as well as
HSGS/HSES requirements in radial and meshed network model using the different system and equipment
parameters.
 Influence of system configuration, radial versus meshed network, single and double circuit
 Influence of line length and fault location, distance from the circuit-breaker to fault point
 Influence of line transposition
 Influence of earth resistance and grounding wire
 Influence of tower configuration, line height and line sag including extreme cases
Note: in case one of the parameters is changed in the model many other variables are automatically altered such as short-circuit current levels.
The influence of these altered variables is incorporated in the results presented.
Page 44
3.7.2 Influence of system configuration (radial vs meshed) on TRV

The influence of the system configuration difference between radial and meshed networks on TRV was
investigated. The maximum breaking current is 28 kA in the radial network model so the influence of breaking
currents on TRV is another concern. Therefore the number of power transformers per bus terminal at substations A,
B and C in the meshed model network is increased to three units as compared to two units per a bus for the radial
network model, resulting in a short circuit current that exceeds 40 kA for the BTF condition.
Figure 3.7.1: Radial and meshed network models
Figure 3.7.1 shows the fault locations for BTF (fault location: F21, B-4 CB located in B s/s, Breaking current:
46.1 kA at maximum) and LLF (fault location: F21 at 360 km distance from the D-4 CB located in D s/s, Breaking
current: 4.4 kA) conditions in the radial and meshed network models using the system and equipment parameters
in China. Figure 3.7.2 shows typical TRV parameters for BTF conditions and LLF conditions calculated in the
1100 kV radial and meshed network models.
Most of the TRV peak and RRRV are covered by TRV requirements for test duties T10, and T100 standardized in
IEC Std 62271-100. TRV for LLF conditions at 360 km distance from the circuit-breaker shows uc=1735.7 kV with
0.72 kV/µs at 4.4 kA in the meshed network. (uc = 1890 kV with 0.79 kV/µs at 4.5 kA in the radial network). TRVs
corresponding to T10 calculated in the meshed network model have a lower peak than those in the radial network.
It is found that TRVs in the radial network provide a slightly higher peak and rate of rise of TRV, especially for
breaking currents corresponding to T10-T30 duties, than those in the meshed network. TRVs corresponding to
T100 calculated in the meshed network model are also covered by TRV requirements in IEC 62271-100.
Figure 3.7.2: TRV for BTF and LLF conditions in 1100 kV radial and meshed network models with
shunt reactor using system and equipment parameters in China
Page 45
Figure 3.7.3 shows TRV waveforms in the 1100 kV meshed network model, TRV for BTF conditions (left: more
than 30 kA) and LLF conditions (right: less than 5 kA). As compared with TRV for LLF conditions in the radial
network model shown in Figure 3.5.7, the shape of TRV in the meshed network is more rounded due to an
increased number of propagation arrivals through different paths reflected from different discontinuities in the
network.
2500 2500
2000 2000
1500 1500
T R V (kV)
T R V (kV)
1000 1000
500 500
0 0
0 0.5 1 1.5 2 2.5 3 3.5 0 0.5 1 1.5 2 2.5 3 3.5
Time (ms) Time (ms)
Figure 3.7.3: TRV waveforms in 1100 kV meshed network model, TRV for BTF conditions (left:
more than 30 kA) and LLF conditions (right: less than 5 kA)
Figure 3.7.4 compares TRVs for 360 km LLF conditions at source-side and line-side CB terminals as well as those
across the CB calculated in radial and meshed network models. The TRV at the line side shows a triangular shape,
which propagates through the 360 km transmission line and comes back to the D2-CB after it is reflected at the
discontinuity of B-s/s. On the contrary, the TRV at the source side has a relatively round shape with a different time
of peak. There is one propagating path for the radial network model on a 360 km line, but in the meshed network
model there are three paths connected to B-s/s so several inflection points are observed on the waveform. The
TRV peaks for both source side and line side in the radial network model occur at the same instant resulting in a
higher TRV peak across the circuit-breaker.
1500
2500
1000 2000
BUS, Line voltage [kV]
500 1500
TRV [kV]
0 1000
12 13 14 15 16 17 18
500
-500
0
-1000
12 13 14 15 16 17 18
Time [ms]
Time [ms]
Figure 3.7.4: Comparison of TRVs at source-side and line-side CB terminals in radial and
meshed network models
Figure 3.7.5 shows TRV inflections arriving through different paths in the meshed network model. There are three
different paths with total line lengths of 360 km (from B-s/s to D s/s), 600 km (from B-s/s to A s/s and then from E
s/s to D s/s) and 840 km (from B-s/s to C s/s and then F s/s to D s/s). The resulting inflections on the source-side
TRV appear respectively 1.3 ms, 2.1 ms and 3.0 ms after fault generation. Three different and longer propagating
paths in the meshed network model delays the TRV peak instant and rounds the peak shape on the source side
resulting in a less severe TRV.
Page 46
TRV at source side

(Meshed network model)
3.0 ms
2.1 ms
1.3 ms
3.4 ms
2.6 ms
TRV at line side
LLF condition
Figure 3.7.5: TRV inflections arrived through different paths in meshed network model
In conclusion, TRV peaks across the circuit breaker calculated in radial networks tend to be slightly higher than
those in meshed networks even though both can be covered by IEC standards.
3.7.3 Influence of fault location on TRV

When clearing faults occurred on a long transmission line (Long line fault: LLF), the TRV across the circuit-breaker
terminals is of a similar physical nature as the saw-tooth shape TRV that appears when clearing short-line faults
(SLF). However, compared with the TRV for SLF, the RRRV is less steep and the TRV peak can reach much
higher values that increase with the distance between the circuit-breaker and fault location.
The RRRV of LLF is lower because the di/dt is lower than in case of SLF due to the fact that the large impedance
of a long line reduces the fault current. The LLF frequency is typically several hundred Hz on the line-side, while
that for SLF it is in the range of ten thousand to more than a hundred thousand Hz. The TRV peak value of the LLF
increases with the voltage drop on the line-side.
The TRV peak value for LLF conditions may be higher than the standard values for T10 or T30 as there are
contributions from both sides of the circuit breaker, instead a contribution from one side only in the case of terminal
fault. However, this TRV peak value does not exceed the TRV peak for out-of-phase.
For example, in the case of a LLF occurring in the 1100 kV model network at 360 km from the circuit-breaker, the
TRV peak is 1897 kV and the RRRV is 0.81 kV/µs (standard values are 1897 kV for T10 and 2245 kV for out of
phase). The fault current in this case is 4.4 kA.
In the case of a LLF occurring in the 550 kV model network at 360 km from the circuit-breaker, the TRV peak is
1071 kV and the RRRV is 0.46 kV/µs (standard values are 1031 kV for T10 and 1123 kV for out of phase). The
fault current in this case is 1.9 kA.
The severity of TRV peaks for LLF conditions can be estimated in the case of a double circuit transmission line.
Figure 3.7.6 shows schematic drawings of TRV of a sine shape waveform at the source side and a triangular shape
waveform at the line side terminals of the circuit-breaker. It is given for different distances from the circuit-breaker
to the fault points without considering the travelling wave propagated through another circuit from the remote
circuit-breaker.
Page 47
Shorter Distance to the fault point Longer

V Source side voltage V V
(a)
(a) (a)=(b)
US US (b) US=US’
(d)
(b) (d) US’
(d) US’ UL
UL 0 UL
t 0 t 0 t
(c)
(c)
Line side voltage (c)
t2 2L/c
tS  LSCS
t2 2L/c tS  LSCS t2 2L/c tS  LSCS t 2  t s 
(i) short distance (ii) middle distance (iii) long distance
Figure 3.7.6: TRV at both source side and line side terminals of CB without considering the
travelling wave through another circuit
The maximum TRV for LLF tends to appear when the first voltage peak on the line side (at 2 L/c) is almost identical
to the instant as the voltage peak on the source side (at pLsCs) without considering the travelling waves
propagated through another circuit of parallel single circuit and double circuit lines, as shown on the Figure 3.7.6.
Figure 3.7.7 shows some examples of TRV contributions at both terminals calculated for different distance to the
fault points. Table 3.7.1 shows the most severe distances from the circuit-breaker to the fault point, with the
corresponding maximum TRV peak and time to TRV peak, for rated voltages 550 kV, 800 kV and 1100 kV.
Breaking current =11.3 kA rms (di/dt=5.02A/s) Breaking current =7.1 kA rms (di/dt=3.15 A/s) Breaking current =5.1 kA rms (di/dt=2.26 A/s)
Source side voltage
Voltage across CB
1st TRV [kV]
1st TRV [kV]

1st TRV [kV]
Uo=602kV Uo=666kV
Line side voltage
Uo=458kV
Up=1084kV Up=1539kV
Up=1401kV
Tp=0.796ms
Tp=1.62ms Tp=2.41ms
Figure 3.7.7: TRV for LLF conditions for different line length in 1100 kV radial network model
using system and equipment parameters in Japan
Table 3.7.1: Expected maximum TRV for LLF conditions

In the case of double circuit (and parallel single circuits), the maximum TRV peak increases with the distance to the
fault point due to the superimposed transient propagated from the remote circuit-breaker. The effect of the transient
from another circuit is observed when the distance is longer than 180 km. The travelling wave propagated through
another circuit is added in the Figure 3.7.8, which arrives at the source side of circuit breaker terminal after a period
of travelling through the another circuit line from a remote circuit breaker.
Page 48
Shorter Distance to the fault point Longer

V Source side TRV V V
(a) Source side TRV Source side TRV
(a) Traveling (a)=(b)
wave
US US (b) US=US’ (e)
Traveling Wave (d) Traveling
from another line wave
(b) (d) US’
(d) US’
UL
UL 0 UL
(c) t0 t 0 t0 t 0 t0 t
(c)
Line side Line side
voltage voltage (c) Line side
t2 2L/c voltage
t2 2L/c tS  LSCS t2 2L/c tS  LSCS tS  LSCS
(i) Short distance (ii) Middle distance (iii) Long distance
Figure 3.7.8: TRV at both source side and line side terminals of CB considering the travelling
wave through another circuit
The influence of fault location, i.e. the distance from the fault to the circuit-breaker, on TRV for LLF conditions was
also investigated in the radial network model using the system and equipment parameters in India. Figure 3.7.9
shows the breaking current and TRV peak, Figure 3.7.10 shows the RRRV dependence on the fault location.
When the distance from the circuit-breaker to the fault point is increased, the breaking current and the RRRV are
decreased, but the TRV peak is increased. Relatively higher RRRVs for BTF conditions, similar phenomena to TLF,
were observed in the case of 3-phase to ground faults when one circuit is open (single circuit condition).
Breaking current (kA)
TRV peak (kV)
Figure 3.7.9: TRV and breaking current for BLF and LLF conditions in 1200 kV radial network
model using the system and equipment parameters in India
Figure 3.7.10: Rate of rise of TRV for BLF and LLF conditions when another line is open
Page 49
When one circuit of a double circuit line is open, this circuit is the only path for current carrying and TRV
propagation. In this case, the current carrying and TRV propagation path is completely disconnected when the
circuit breaker clears a line fault and the source side TRV is totally determined by the transformer.
Figure 3.7.11 compares the TRV across the circuit-breaker and its contributions on the source and line side circuit-
breaker terminals for LLF and BTF conditions in two cases:
- when one circuit of a double circuit is already open at one substation that is connected to only one line so that
there is no path for current carrying and TRV propagation when this line is faulted and the circuit breaker has
interrupted the current (then the double circuit is completely disconnected).
- when five lines are connected to a substation so that four remaining lines are still connected on the supply side of
a line circuit breaker after interruption of the current in a faulted line.
In the first case where the substation is connected to only one alternative transmission line (single circuit condition),
after current interruption a steep increase of TRV for BTF and LLF conditions on the source side (transformer side)
is observed due to the propagation of transient waves with a high frequency determined by the power transformer
parameters. For this reason a higher RRRV for LLF is observed when a circuit breaker is opened in a single circuit
condition.
B S/S 360 km LLF D S/S TLF (D S/S is

(B S/S is connected by only
2000 D S/S 360 km LLF 2000 B S/S BTF (B S/S is
connected by one line which is
(D S/S is connected connected by
some lines) faulted)
by only one line some lines)
which is faulted)
1500 1500
TRV [kV]
TRV [kV]
1000 1000
500 500
0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
time [ms] time [ms]
(a) 360 km LLF TRV (b) BTF, TLF TRV
D S/S B S/S
1500 BUS side voltage BUS side voltage D S/S 360 km LLF
1000
(D S/S is connected
by only one line
BUS, Line voltage [kV]
500 which is faulted)

0
12 13 14 15 16 17 18
-500
B S/S 360 km LLF
D S/S
B S/S (B S/S is
-1000 Line side voltage
Line side voltage
Time [ms] connected by
some lines)
Figure 3.7.11: Comparison of TRV for LLF and BTF conditions in case that one line of a double
circuit is open at one substation connected to one or five remaining lines
3.7.4 Influence of tower design (parallel single circuit or double circuit tower) on TRV
The influence of tower design on TRV was investigated using the 550 kV radial network model with parallel single
circuit and double circuit towers applied in the networks of Thailand.
Figure 3.7.12 shows a tower and line configuration of parallel single circuit and double circuit towers. Table 3.7.2
summarizes the comparison of the line surge impedance, first and third-pole-to-clear factors along with line
Page 50
parameters between the single and double circuit. The results show that the influence of the tower design on the
positive sequence impedance (Z1) and inductance (L1) is negligible, because the line characteristic does not
change due to the tower design. Accordingly TRV waveforms shown in Figure 3.7.13 do not show much difference
between the networks with parallel single circuit and double circuit towers even though line height and arrangement
are significantly different.
Figure 3.7.12: Double circuit OHL and two parallel circuits OHL for 550 kV networks in Thailand
Surge impedance Inductance Capacitance First-pole-to-clear Third-pole-to-clear

Type of OHL Z0 Z1 L0 L1 C0 C1 Ze Peak Ze Peak
(ohm) (ohm) (mH/km) (mH/km) (nF/km) (nF/km) (ohm) factor (ohm) factor
Double circuit 486 243 2.02 0.83 8.6 14.0 291 2.40 324 1.80
Two parallel
478 236 1.85 0.80 8.1 14.3 284 2.41 317 1.87
single circuits
Table 3.7.2: Influence of tower design on line parameters in 550 kV networks in Thailand
Figure 3.7.13: TRV waveforms for parallel single circuit and double circuit towers
Page 51
3.7.5 Influence of grounding wire on TRV
The influence of a grounding wire on line surge impedance and TRV was investigated using the 1100 kV radial
network model with the double circuit tower in Japan. Table 3.7.3 summarizes the line surge impedance of the
1100 kV double circuit transmission line with and without a grounding wire. Even though the impedance to ground
due to the existence of a grounding wire is larger, the difference of TRV between with and without the grounding
wire was confirmed to be negligible as shown in the Figure 3.7.14.
As can be learned from Table 3.7.3, as well as from the sections 3.7.6 and 3.7.7, the application of ground wires
has a large influence on the surge impedances, especially the zero sequence surge impedance, but its effect on
the TRV wave-shape is masked by the fact that the fault current is lower without ground wires in comparison to the
fault current with ground wires. This is due to the higher inductance per km (without ground wires) and, in the
simulations, all OHL are equipped with or without ground wires, so the total system inductance is influenced.
Surge impedance Inductance First-pole-to-clear Third-pole-to-clear

Tower height Sag of Ground
L0 L1 Peak Peak
(m) conductor (m) wire Z0 (ohm) Z1 (ohm) Ze (ohm) Ze (ohm)
(mH/km) (mH/km) factor factor
120 0 with 530 230 2.15 0.78 283 2.47 329 1.81
120 0 without 669 232 3.12 0.78 296 2.56 378 1.63
Table 3.7.3: Influence of grounding wire on line surge impedance (1100 kV double circuit tower)
Figure 3.7.14: TRV waveforms for 1LG/3LG conditions with and without grounding wire
3.7.6 Influence of line sag on TRV

The influence of line sag on TRV was investigated using the 1100 kV radial network model with system and
equipment parameters in Japan. The extent of the line sag depends on line weight, line tension and the distance
between the towers. Larger line sag leads to lower line height from the ground resulting in a larger coupling
capacitance, which in turn changes the value of zero-sequence impedance. The inductance (L), capacitance (C)
and surge impedance (Z) of a single conductor can be given by the following equations, where v is propagation
speed, h is the line height and r is the radius of the conductor. m0 is the magnetic permeability in vacuum, e0 is the
permittivity (h>>r).
 0 2h 2h 1 2h
L ln , C  2 0 / ln , v , Z  60 ln
2 r r LC r
Note: the formulae are applicable for perfect earth conditions; i.e. zero earth resistivity and therefore no penetration depth of electric and
magnetic fields.
The line parameters were calculated by the EMTP program built-in function for line constant calculation in order to
evaluate the influence of line sag on the line impedance and TRV for LLF conditions. In the EMTP program, the
influence of line sag is considered as an equivalent straight line without line sag whose height is decreased by 1/3
of the maximum length of line sag.
Page 52
3.7.5 Influence of grounding wire on TRV
The influence of a grounding wire on line surge impedance and TRV was investigated using the 1100 kV radial
network model with the double circuit tower in Japan. Table 3.7.3 summarizes the line surge impedance of the
1100 kV double circuit transmission line with and without a grounding wire. Even though the impedance to ground
due to the existence of a grounding wire is larger, the difference of TRV between with and without the grounding
wire was confirmed to be negligible as shown in the Figure 3.7.14.
As can be learned from Table 3.7.3, as well as from the sections 3.7.6 and 3.7.7, the application of ground wires
has a large influence on the surge impedances, especially the zero sequence surge impedance, but its effect on
the TRV wave-shape is masked by the fact that the fault current is lower without ground wires in comparison to the
fault current with ground wires. This is due to the higher inductance per km (without ground wires) and, in the
simulations, all OHL are equipped with or without ground wires, so the total system inductance is influenced.
Surge impedance Inductance First-pole-to-clear Third-pole-to-clear

Tower height Sag of Ground
L0 L1 Peak Peak
(m) conductor (m) wire Z0 (ohm) Z1 (ohm) Ze (ohm) Ze (ohm)
(mH/km) (mH/km) factor factor
120 0 with 530 230 2.15 0.78 283 2.47 329 1.81
120 0 without 669 232 3.12 0.78 296 2.56 378 1.63
Table 3.7.3: Influence of grounding wire on line surge impedance (1100 kV double circuit tower)
Figure 3.7.14: TRV waveforms for 1LG/3LG conditions with and without grounding wire
3.7.6 Influence of line sag on TRV

The influence of line sag on TRV was investigated using the 1100 kV radial network model with system and
equipment parameters in Japan. The extent of the line sag depends on line weight, line tension and the distance
between the towers. Larger line sag leads to lower line height from the ground resulting in a larger coupling
capacitance, which in turn changes the value of zero-sequence impedance. The inductance (L), capacitance (C)
and surge impedance (Z) of a single conductor can be given by the following equations, where v is propagation
speed, h is the line height and r is the radius of the conductor. m0 is the magnetic permeability in vacuum, e0 is the
permittivity (h>>r).
 0 2h 2h 1 2h
L ln , C  2 0 / ln , v , Z  60 ln
2 r r LC r
Note: the formulae are applicable for perfect earth conditions; i.e. zero earth resistivity and therefore no penetration depth of electric and
magnetic fields.
The line parameters were calculated by the EMTP program built-in function for line constant calculation in order to
evaluate the influence of line sag on the line impedance and TRV for LLF conditions. In the EMTP program, the
influence of line sag is considered as an equivalent straight line without line sag whose height is decreased by 1/3
of the maximum length of line sag.
Page 52
Table 3.7.4 gives the line surge impedance, first and third-pole-to-clear factors along with line parameters for
different line sag based on the 1100 kV transmission tower used in Japan. The result shows that the influence of
line sag on the positive sequence impedance (Z1) and inductance (L1) is negligible. This is due to the fact that the
distance between phases does not change much even though the length of the line sag is changed. They slightly
increase with an increase of the line sag but the influence is limited for relatively taller UHV transmission tower.
As may be expected (see also Table 3.7.5), the zero sequence surge impedance increases with a higher position
of the conductors. This applies for the first row in Table 3.7.4 (120 m, without sag) in comparison to the third row
(109 m, without sag) or for the second row (120 m, sag 40m/28m) to the fifth row (109 m, sag 40m/28m). But when
comparing a condition without sag to one with sag (thus a lower average height), the opposite can be seen.
Compare, for instance, row 1 to row 2 or row 3 to the rows 4 and 5. This phenomenon can be explained by the
large influence of the coupling between phase conductors (especially the upper conductors) with the ground wires.
The ground wires show less sag than the phase conductors, so that more sag leads to less coupling and thus a
relatively higher zero sequence surge impedance.
Sag of Sag of Surge impedance Inductance First-pole-to-clear Third-pole-to-clear

Tower height
conductor Ground wire Z0 Z1 L0 L1 Ze Peak Ze Peak
(m)
(m) (m) (ohm) (ohm) (mH/km) (mH/km) (ohm) factor (ohm) factor
120 0 0 530 230 2.15 0.78 283 2.47 329 1.81
120 40 28 537 230 2.25 0.78 284 2.47 332 1.78
109 0 0 526 229 2.14 0.78 282 2.46 328 1.80
109 27 20 528 230 2.21 0.78 283 2.47 329 1.78
109 40 28 529 230 2.25 0.78 283 2.46 329 1.76
Table 3.7.4: Influence of line sag on line surge impedance (1100 kV double circuit tower)
Figure 3.7.15 shows typical TRV waveforms for 360 km LLF condition in the 1100 kV radial network model with and
without line sag. The results show that the influence of the line sag on TRV waveform is negligible in the case of
the 1100 kV transmission tower.
(a) Tower 120 m high (b) Tower 109m high
Figure 3.7.15: Influence of line sag on TRV for 360 km LLF conditions in 1100 kV double circuit
network model using the system and equipment parameters in Japan
In conclusion, the influence of line sag on TRV for LLF (both 1LG and 3LG conditions) is negligible (less than 1 %)
in the case of 1100 kV networks with double circuit towers higher than 100 m with line sag up to 40 m.
Page 53
3.7.7 Influence of line height on TRV

The influence of line height on TRV was investigated in the 550 kV radial network model using the system and
equipment parameters in Canada. Figure 3.7.16 shows a schematic design of parallel single circuit towers and
conductor dimensions. The actual tower configuration is a height of 14.6 m with maximum 4 m sag and consists of
a four bundle conductor with ACSR 507 mm2. In analysis, the tower configuration is changed to heights of 24.6 m,
14.6 m and 4.6 m without sag and without the grounding wire to evaluate the influence for the extreme cases.
Conductor (ACSR) Case 1 Case 2 Case 3 Case 4

Diameter D (cm) 2.54
Ratio T/D (T: Thickness) 0.333
DC resistance at 20 °C, R (ohm/km) 0.084
Number of conductors in the bundle 4
Bundle Spacing, S (cm) 45.72
Height (m) 4.6 14.6 14.6 24.6
Sag (m) 0 4 0 0
Figure 3.7.16: Dimensions of 550 kV parallel single circuit tower including extreme cases
Table 3.7.5 shows the line surge impedance, first and third-pole-to-clear factors for different line heights in the
550 kV transmission tower used in Canada. The zero-sequence and positive sequence impedance (Z0, Z1)
decrease with a decrease of line height, because the capacitance to ground increases with a decrease of the line
height. The first-pole and third-pole-to-clear factors also slightly decrease with a decrease of line height. Line surge
impedance and peak factors of the line side TRV are shown in Figure 3.7.17 for the first and third pole to clear.
Surge impedance Capacitance First-pole-to-clear Third-pole-to-clear

Line height Sag of Ground
Z0 Z1 C0 C1 Ze Peak Ze Peak
(m) conductor (m) wire
(ohm) (ohm) (pF/km) (pF/km) (ohm) factor (ohm) factor
4.6 0 without 500 239 13400 15300 289 2.43 326 1.42
14.6 4 without 609 254 9020 13560 315 2.49 372 1.52
14.6 0 without 626 255 8530 13450 318 2.50 379 1.54
24.6 0 without 688 257 7100 13230 325 2.53 400 1.61
Table 3.7.5: Influence of line height and sag on line parameters

(550 kV single circuit tower)
Figure 3.7.18 shows typical TRV waveforms in the case of a 360 km LLF condition for different line height
calculated in the 550 kV parallel single circuit network model including extreme cases. Figure 3.7.19 shows the
values of TRV peak and RRRV dependence on the line height.
Figure 3.7.17: Influence of line height on line surge impedance and peak factor in 550 kV
parallel single circuit network model including extreme cases
Page 54
When TRV parameters are compared at the line heights between 24.6 m (higher than that for the practical case)
and 4.6 m (extreme lower case), the difference of TRV peak is 3-6 % and that of RRRV is 12-17 %. Therefore, the
influence of the line height (even the extreme lower height) on TRV is not generally significant.
Case 1 (4.6m)
1200 Case 1 (4.6m) 800 Case 2 (14.6m with sag)
Case 2 (14.6m with sag) Case 3 (14.6m)
1000 Case 3 (14.6m) Case 4 (24.6m)
Case 4 (24.6m) 600
800
1st TRV[kV]
1st TRV[kV]
600 400
400
200
200
0 0
0 0.5 1 1.5 2 2.5 3 3.5 0 0.5 1 1.5 2 2.5 3 3.5
time[ms] time[ms]
Figure 3.7.18: Influence of line height and line sag on TRV waveforms in 550 kV parallel single
circuit network model including extreme cases
Figure 3.7.19: Influence of line height on TRV peak and RRRV in 550 kV parallel single circuit
network model including extreme cases
From Table 3.7.5 it can be learned that the higher the conductors the lower the zero sequence capacitance and the
higher the zero sequence inductance and surge impedance (without ground wires). The height has hardly any
influence on the positive sequence parameters, apart from unrealistic low values of the height. A same trend can
be seen in Table 3.7.4 (with ground wires), when comparing cases with identical sag conditions. Cases with
different sag conditions, though, show that the coupling between phase conductors and ground wires is at least as
important as the coupling between phase conductors and earth. Therefore, with ground wires, the impact of height
and sag of phase conductors and ground wires is less straightforward and a greater average height does not
automatically lead to a higher inductance or surge impedance.
3.7.8 Influence of line transposition on TRV

Line transposition is employed to balance the normal current for each phase suspended from single and double
circuit towers. In case of short lines, a line is often transposed at the substation. The influence of the different type
transposed or non-transposed lines on TRV of first-pole-to-clear was investigated in the 550 kV radial network
model using system and equipment parameters in Thailand. Figure 3.7.20 shows non-transposed double circuit
lines used for the TRV study.
Page 55
Figure 3.7.20: Non-transposition for double circuit lines
Figures 3.7.21 and 22 show that -transposition (full-transposed lines) and -transposition (half-transposed lines)
for double circuit lines, respectively. The -transposition has two transpositions for one circuit and six transpositions
for another circuit and each circuit is transposed at different points of the tower. The -transposition has two
transpositions for both circuits at the same points of the tower.
Figure 3.7.21:  (beta)-transposition or full-transposition for double circuit lines
Figure 3.7.22:  (gamma)-transposition or half-transposition for double circuit lines
Page 56
Figure 3.7.23 shows TRV waveforms of first-pole-to-clear for 360 km LLF conditions with 3 different types of
transpositions (non-, - and -transpositions). In the case of the non-transposed line, the TRV peak for the phase
located at the middle position of the tower shows the highest value and that located at the upper position is the
lowest value. In cases of - and -transposition, TRV for different phases are almost identical so the line
transposition can also balance the TRV for each phase. The difference of TRV peaks between  or -transposition
and non-transposition is a few %.
Transposition First –pole-to clear Breaking current (kA) TRV for first-pole-to-clear
TRV peak (kV) RRRV (kV/s)
Non Lower case (C) 2.5 1047 0.46
Middle case (A) 2.5 1080 0.48
Upper case (B) 2.6 1019 0.47
 C phase 2.4 1028 0.45
beta A phase 2.4 1039 0.45
B phase 2.4 1038 0.44
 C phase 2.5 1044 0.47
gamma A phase 2.5 1044 0.46
B phase 2.5 1043 0.46
Figure 3.7.23: TRV comparison for different types of line transposition

1200 1200
1000 1000
800 800
600 600
400 beta-transposition (first clear at C-phase) 400 ganma1-transposition (first clear at C-phase)
beta-transposition (first clear at A-phase) ganma1-transposition (first clear at A-phase)
200 200
beta-transposition (first clear at B-phase) ganma1-transposition (first clear at B-phase)
0 0
0 0.5 1 1.5 2 2.5 3 0 0.5 1 1.5 2 2.5 3
Figure 3.7.24: TRV comparison for different phases in case of  and  transposition
In general, the influence of line transposition on TRV is not significant.
3.7.9 Influence of earth resistivity on TRV

The influence of earth resistivity on TRV of first-pole-to-clear was investigated using the 1100 kV radial network
model with system and equipment parameters from Japan. The earth resistivity has a negligible influence on the
TRV peak and the breaking current as shown in Table 3.7.6.
Page 57
Earth Resistivity Line sag Grounding wire sag Breaking current TRV peak
(ohm-m) (m) (m) (kA) (kV)
100 without sag without sag 2.78 1842
500 without sag without sag 2.78 1850
100 20 18 2.76 1831
Table 3.7.6: Influence of earth resistivity on TRV

in the case of 1100 kV double circuit tower
3.8 Influence of system and operation parameters on TRV

3.8.1 Introduction
The influence of various compensation schemes, utilities’ system and operation parameters on TRV requirements
for BTF, LLF, TLF and no-load line-charging current interruption conditions was also investigated in radial and
meshed network model using the different system and equipment parameters.
 Influence of 1LG versus 3LG conditions;
 Influence of series-capacitor, compensation degree;
 Influence of shunt reactor;
 Influence of short-circuit power or power flow;
 Influence of equipment specifications such as MOSA for different national projects.
3.8.2 Influence of MOSA V-I characteristic on TRV

Metal Oxide Surge Arresters (MOSA) may clip the TRV peak depending on their V-I characteristics (voltage
restriction level). Figure 3.8.1 shows typical TRV waveforms with a TRV representation by IEC Std 62271-100
calculated under 3LG conditions in the 1100 kV transmission system with and without the MOSA in Japan. UHV
class MOSA can reduce the TRV peak values at least in case of terminal fault (BTF) duties.
Figure 3.8.1: Influence of MOSA on TRV waveforms for 360 km LLF condition in 1100 kV
transmission system in Japan
Note that surge arresters are supposed to suppress switching overvoltages, these overvoltages are equal to TRV
only in the case of terminal faults. In the next examples terminal fault are simulated.
The TRV suppression levels due to MOSA can be determined by means of an equivalent surge impedance of T100
seen by the circuit-breaker’s clearing pole and the number of MOSAs connected at the source side of the circuit-
breaker.
Page 58
Figure 3.8.2: Simple network model to study different number of MOSAs
Figure 3.8.3 shows the equivalent surge impedance of a 1100 kV double circuit network model and three MOSA V-
I characteristics for different numbers of MOSAs at the bus terminal. The intersection points can provide the TRV
suppression level, which is slightly reduced with the number of MOSAs.
2000
1500
Voltage (kV) .
1000
500
0
0 2 4 6 8 10
MOSA Current (kA)
Figure 3.8.3: Intersection of system response line with different number of MOSAs
Figure 3.8.4 shows the intersection of system response line with different MOSA V-I characteristics in 550 kV,
800 kV and 1100 kV radial network models. In the case of the 1100 kV radial network model, the intersection with
the 1100 kV MOSA V-I characteristic gives 1452 kV corresponding to the restriction voltage around a few kA. The
inherent TRV peak without MOSA is 1568 kV.
On the contrary, the intersection with the 800 kV and 550 kV MOSA V-I characteristics are the same level of TRV
peaks so the TRV suppression by MOSA would not be observed in the EHV networks considering the
characteristics of the existing MOSA.
Furthermore in the future UHV meshed network models, the TRV suppression may not be observed because the
TRV can be reduced due to multiple-reflections of an increasing numbers of transmission lines.
Page 59
2000
1500
Voltage (kV
1000
500
0
0 10 20 30 40
C urrent (A)
Figure 3.8.4: Intersection of system response line with different MOSA V-I characteristics in
radial network model
3.8.3 Influence of fault conditions: 1LG vs. 3LG

In case of three-phase line faults to ground (3LG) condition, the TRV peak for the first-pole-to-clear becomes more
severe as compared to that of single-phase line faults to ground (1LG) because the voltage transients induced from
other fault lines are imposed on the TRV. Figures 3.8.5 and 3.8.6 show the TRV for 1LG and 3LG conditions in
765 kV and 1100 kV radial network models using, respectively, system and equipment parameters in Canada and
China. As expected, the results shows that higher TRV peaks are obtained for 3LG conditions, as compared with
1LG conditions, even though the RRRV does not show any significant difference.
Figure 3.8.5: TRV for 1LG and 3LG conditions in 765 kV radial network model with parallel
single circuits using system and equipment parameters in Canada
Note: single circuit means one circuit of the parallel single circuits is open, double circuit means both parallel single circuits are connected
Figure 3.8.6: TRV for 1LG and 3LG conditions in 1100 kV radial network model with double
circuits using system and equipment parameters in China
Page 60
Figures 3.8.7 and 3.8.8 gives the comparison of TRV across the CB, with contributions at the source side and line
side terminals, between 1LG conditions and 3LG conditions in the 1100 kV radial network model. There isn't much
difference between the source side contribution to TRV for the 3LG condition and the 1LG condition. However, the
line side contribution to TRV for the 3LG condition is higher than that of the 1LG condition although Ze of the 3LG
condition is lower than that of 1LG. This is caused by the higher short-circuit current: in this model plus 25% in this
model. As a result; the TRV across CB of 3LG condition is higher than that of 1LG condition.
TRV of 1st pole-to-clear for 3LG condition TRV of 1st pole-to-clear for 1LG condition
Breaking current =5.1kArms, di/dt=2.26A/s Breaking current =4.1kArms, di/dt=1.82A/s
TRV of 1st pole-to-clear (kV)

Uo=666kV Uo=697kV
Up=1539kV Up=
1320kV
Tp=2.40ms
Tp=2.41ms
K=Up/Uo=2.31, Ze=Up/Tp/di/dt=283 ohm K=Up/Uo=1.89, Ze=Up/Tp/di/dt=302 ohm
Figure 3.8.7: Comparison of TRV across CB, at source side and at line side terminals between
1LG conditions for single circuit and 3LG conditions for single circuit
TRV of 1st pole-to-clear for 3LG condition TRV of 1st pole-to-clear for 1LG condition
Breaking current =4.4 kA , di/dt=1.95 A/s Breaking current =2.75 kA , di/dt=1.22 A/s
Uo=550kV Uo=632kV
(ms) (ms)
Up=1456kV
Up=1148kV
Tp=2.41ms
Tp=2.42ms
K=Up/Uo=2.64, Ze=Up/Tp/di/dt=308 ohm K=Up/Uo=1.81, Ze=Up/Tp/di/dt=390 ohm
Figure 3.8.8: Comparison of TRV across CB, at source side and at line side terminals between
1LG conditions for double circuits and 3LG conditions for double circuits
Table 3.8.1 summarizes the results of the TRV peak for 1LG and 3LG conditions in single and double circuit
networks. TRV peaks for the 3LG condition are higher than those for the 1LG condition and more severe in the
case of a double circuit. This is due to the arrival of a travelling wave through the parallel circuit as described in
3.7.3.
TRV at line side TRV at source side TRV across CB

Line Fault
Ze=Up/Tp/di/dt Peak factor Usc Uc
arrangement condition
(ohm) K=Up/Uo (kV) (kV)
Single 3LG 283 2.31 837 1640
Single 1LG 302 1.89 813 1370
double 3LG 308 2.64 993 1890
double 1LG 390 1.81 939 1456
Table 3.8.1: TRV peak at source side and line side for 1LG and 3LG conditions
Page 61
3.8.4 Influence of power flow on TRV

In the TRV analysis made with EMTP, the circuit-breakers located at both line ends can interrupt the fault for BTF
and LLF conditions at the same instant without considering the influence of a power flow. If the effect of power flow
is considered, the interrupting instants of the circuit-breaker for BTF and LLF conditions differ and significantly
affect the TRV magnitude. For example, the CB can first interrupt the BTF (typically, the CB for BTF duty interrupts
1.3 ms earlier than the CB for the LLF duty) and generate a transient surge that propagates to the remote fault,
resulting in a significant reduction of the TRV peak for the LLF duty.
A time-consuming statistical approach is required to evaluate the TRV requirements in order to consider the
influence of power flow. However, several utilities have conducted TRV evaluation considering power flow in their
actual networks conditions.
In cooperation with a Canadian utility, the influence of power flow on TRV is evaluated in an actual 765 kV network
with series capacitor banks and compared with the results calculated with the radial network model using system
and equipment parameters in Canada. In the statistical approach, a total of 14400 simulations were carried out with
different fault locations.
In the statistical approach, different fault types (1LG, 3LG) and different compensation degrees (20, 40 and 60 %)
under light and heavy load conditions in Canada were modelled as shown in Figure 3.8.9. In the analysis, the
triggering gap of series capacitor is shorted (electrically-connected) either when the energy of the protective
varistor across the series capacitor is higher than 15 MJ or when the current peak through the varistor is higher
than 60 kA. For series capacitor banks see 3.8.6.
Figure 3.8.9: 765 kV transmission systems in Canada used for TRV analysis (left: light load
condition, right: heavy load condition)
Tables 3.8.2 summarizes the TRV peaks calculated at different fault locations with 1LG and 3LG conditions and
different compensation degrees (20, 40 and 60 %) under light and heavy load conditions. The brown shaded
values show TRV peaks when the triggering gap operated and bypassed the series capacitor. TRV peaks tend to
be severe for higher compensation degrees and in the case of 3LG conditions rather than 1LG. However, it does
not show a significant difference in TRV peaks between heavy load and light load conditions.
Figure 3.8.10 shows a typical example of the fault current, the current through the varistor, and the TRV waveform
at the compensation degree of 40 % under light load condition.
Page 62
Table 3.8.2: TRV peaks with triggering gap
Figure 3.8.10: Typical analytical example of fault current, current through the varistor and TRV
waveform in case of 40 % compensation degree and 3LG condition under light load flow in
765 kV transmission systems in Canada
Figure 3.8.11 shows the TRV waveform with and without MOSA located at both ends of transmission line at a
compensation degree of 40 % under light load condition. The suppression effect of MOSA between lines to the
ground is limited.
Figure 3.8.11: TRV waveforms with and without MOSA
Figure 3.8.12 gives the comparison of TRV peaks evaluated in 765 kV series compensated networks under heavy
and light power flow with those calculated without power flow in a radial network model using system and
equipment parameters in Canada.
Page 63
It is confirmed that TRV peaks calculated with and without power flow agree well. Therefore, TRV requirements can
be generally evaluated accurately using system and equipment parameters in the network even without a power
flow condition.
Series capacitor without bypassed measures Series capacitor with bypassed measures
without power flow

2500 2500 with Light power flow
TRV peak value for OP TRV peak value for OP
with Heavy power flow
2000 2000
TRV peak [kV]

TRV peak [kV]
1500 1500
1000 1000
TRV peak value for T10 without power flow
500 with Light power flow 500
with Heavy power flow TRV peak value for T10
0 0
0 100 200 300 400 0 100 200 300 400
distance from CB to faulting point [km] distance from CB to faulting point [km]
Figure 3.8.12: Comparison of TRV peaks between the values calculated by the statistical
approach considering power flow and those calculated without power flow in the 765 kV radial
network model with parallel single circuits using system and equipment parameters in Canada
3.8.5 Influence of shunt reactor on TRV

The influence of shunt reactor applications on TRV for LLF condition was investigated using the 550 kV double
circuit network model in Thailand. Figure 3.8.13 shows 110 MVA and 55 MVA shunt reactor arrangements located
at both ends of the transmission lines.
Figure 3.8.13: 550 kV radial network model with shunt reactors using system and equipment
parameters in Thailand
Figure 3.8.14 shows the TRV waveforms for the 360 km LLF condition with and without shunt reactors. The shunt
reactor increases the breaking current by 3 % and decreases the TRV peak by 2 % as compared with these values
without a shunt reactor.
Figure 3.8.15 shows the voltage distributions with and without shunt reactors when a fault has occurred at the bus
terminal of the D substation. The voltage is slightly reduced with the shunt reactor but not significantly.
Page 64
800
700
600
500
TRV (kV)
400
300
200
100
0
18 19 20 21 22 23 24
time (ms)
Figure 3.8.14: Influence of shunt reactor on TRV waveforms in 550 kV double circuit network
model using the system and equipment parameters in Thailand
Figure 3.8.15: Voltage distributions with and without shunt reactor along the 360 km
transmission line when a fault is generating at the bus terminal of D substation
In conclusion, the influence of shunt reactor application on TRV during a LLF is negligible, because the TRV
frequency is much higher than the oscillation generated by the shunt reactor and line.
The influence of shunt reactors on TRV for a 360 km no-load line breaking condition was also investigated using
the 550 kV double circuit network model in Thailand. Figure 3.8.16 shows typical TRV waveforms in the network
with and without shunt reactors. The TRV frequency depends on the inductance of the reactor and the capacitance
of the line, it has a frequency of several tens of Hz so the TRV peak can be considerably reduced in the network
when a large shunt reactor is connected. For example, the TRV peak for line charging condition in a 550 kV
network and a fault occurring at a remote distance of 360 km is 1018 kV without shunt reactor, while it is reduced to
less than 400 kV when a shunt reactor is connected.
Page 65

terminals (kV)
terminals (kV)
= 552A rms = 290A rms
current (kA)
current (kA)
Breaking
Breaking
voltage (kV)
voltage (kV)
Source-side
Source-side
voltage (kV)
voltage (kV)
Line-side
Line-side
Figure 3.8.16: Influence of shunt reactor on TRV for no-load line breaking in 550 kV double
circuit network model using system and equipment parameters in Thailand
Therefore, shunt reactors can significantly reduce the first TRV peak for no-load line charging breaking conditions.
3.8.6 Influence of series-capacitor on TRV

In the case of networks with series-capacitors, the TRV peak increases due to the increase of the TRV amplitude
factor at the line side and also the breaking current increases because an equivalent line inductance decreases
due to the capacitance of the series-capacitor. WG A3.13 investigated field experience on switching requirements
in several networks with series-capacitors along with some countermeasures and summarized its findings in
CIGRE Technical Brochure 336 [19].
As an example, the TRV peak for a 3-phase LLF in a 800 kV system with a fault occurring 240 km from the circuit-
breaker is 1181 kV without a series-capacitor (series-capacitor is by-passed) with an RRRV of 0.90 kV/µs
(standard value is 1499 kV for T10), while it is 1543 kV with an RRRV of 1.1 kV/µs with series-capacitor and a
compensation degree of 20 %. Fault current is respectively 6.3 kA and 7.7 kA without and with series-capacitor.
TRV peak can reach 1799 kV (standard value is 1633 kV for out-of-phase duty) if the compensation degree is 40 %.
Figure 3.8.17 shows a simplified circuit of 550 kV transmission lines with a series-capacitor in the middle of a
360 km line. Table 3.8.3 gives the line characteristics such as power factor and equivalent surge impedance of the
series compensated transmission line with series-capacitor (40 % compensation degree not bypassed) and without
series-capacitor (bypassed by a bypass switch). The series capacitor with 40 % compensation degree can increase
the breaking current and peak factor by a factor of 1.7 as compared to the case without series-capacitor due to the
decrease of the equivalent inductance of the line.
Source Us UL
E, f Transmission line Transmission line
Xs CB XL/2, Z1, Z0, L/2 SC XL/2, Z1, Z0, L/2
BPS
I Fault
E=550 kV/√3=318 kV, f=60 Hz, Xs=6.4 ohm at 50 kA, w=2πf

Transmission line : L=360 km (180 km + 180 km), LL=360 mH (1mH/m), XL=136 ohm =L1 w,
Z1=250 ohm, Z0=500 ohm, c=280 m/us
Series capacitor : SC=47.4 mF/phase, Xc=56 ohm
Figure 3.8.17: Simplified circuit of 360 km transmission line with series-capacitor
Page 66
BPS closed position

BPS opened position (with SC)
(without SC)
E
Breaking current (kA) I 3.7 2.2
(Xs  XL  Xc)
3Z1Z 0
Equivalent surge impedance (ohm) Z e1  300 300
( Z1  2Z 0 )
Equivalent inductance of the line side (mH) Le  ( X L  X c ) /  212 360
Up Z e1  2 L
Peak factor of line TRV P.F .   3.64 2.14
Uo Le  c
Table 3.8.3: Line characteristics with and without series-capacitor of 40% compensation
Figure 3.8.18 shows the TRV for a 360 km LLF condition with a series-capacitor and 40 % compensation degree
and without the series-capacitor (bypassed) in the 550 kV parallel single circuit network model using system and
equipment parameters from Canada. Since the peak factor on the line side is as high as 3.92 due to the series-
capacitor, a severe TRV peak of 1150 kV is obtained, which exceeds the standard value for out-of-phase (1123 kV).
Breaking current
Breaking current
Breaking current =1.9kArms Breaking current =3.0kArms
[kA]
Voltage across circuit-breaker [kA] Voltage across circuit-breaker

Uc=1150kV
Uc=720kV
Us
Uo=342kV UL
1st TRV [kV]
Uo=292kV
1st TRV [kV]
Up=776kV
Up=1144kV
Peak factor = Up/Uo = 2.27 Peak factor = Up/Uo = 3.92
Figure 3.8.18: TRV waveforms for 360 km LLF conditions with 40 % series-capacitor (right: not
bypassed) and without series-capacitor (left: bypassed by BPS)
Figure 3.8.19 shows the TRV peak for LLF at 1LG conditions in the 765 kV radial network model with a series-
capacitor (30 % compensation degrees) using the system and equipment parameters in Canada. The range of the
calculated TRVs show good agreement with the TRV evaluated based on the actual 765 kV networks in Canada. In
the case of 1LG conditions (90 % faults are covered), TRV peaks are less than 1250 kV with the series-capacitor
(30 % compensation), which can be covered by the standard TRV for test duty T30. The influence of MOSA on
TRV is negligible in the case of LLF because there are source and line side contributions to TRV and MOSA is
subjected to only a part of the TRV.
Figure 3.8.19: TRV for BTF and 120, 240, 360 km LLF at 1LG conditions in 765 kV radial network
model with series-capacitor
Page 67
Figure 3.8.20 shows the TRV peak and RRRV for BTF and 120, 240, 360 km LLF in the case of 1LG and 3LG
conditions in the 765 kV radial network model with series-capacitor (40 % compensation degree, not bypassed by a
bypass switch) using the system and equipment parameters from Canada. In the case of single-phase faults, most
of TRV peaks are covered by the standard values; those obtained in 3LG conditions (not bypassed) exceed the
standard TRV peaks for T10 and T30.
Figure 3.8.21 shows the TRV peak and RRRV for 3LG conditions in the 765 kV radial network model with series-
capacitor (bypassed or not bypassed) using the system and equipment parameters in Canada, assuming different
compensation degrees. The TRV peak can be covered by the IEC standard when the series-capacitor is bypassed
by either a spark gap or bypass switch. However, the TRV peaks exceed the standard requirements if the series-
capacitor is not bypassed and severe TRV peaks are expected with an increase of the compensation degrees.
2500 8

7 Ser ies-capacitor without bypassed
2000 Ser ies-capacitor with bypassed by spar k gap
6
5
TRV (kV)
1500
4
1000 3
2
500
Ser ies-capacitor without bypassed 1
Ser ies-capacitor with bypassed by spar k gap
0 0
0 10 20 30 40 50 0 10 20 30 40 50
2500 8
Ser ies-capacitor without bypassed 7 Ser ies-capacitor without bypassed

2000 Ser ies-capacitor with bypassed by spar k gap Ser ies-capacitor with bypassed by spar k gap
6
5
TRV (kV)
1500
4
1000 3
500 2
1
0 0
0 10 20 30 40 50 0 10 20 30 40 50
Figure 3.8.20: TRV peak and RRRV for 1LG and 3LG conditions in 765 kV radial network model
with series-capacitor
Page 68
2500 8

Ser ies-c apac itor without bypassed 7 Ser ies-c apac itor without bypassed
2000 Ser ies-c apac itor with bypassed by spar k gap Ser ies-c apac itor with bypassed by spar k gap
6
TRV (kV) 1500 5
4
1000 3
2
500
1
0 0
0 10 20 30 40 50 0 10 20 30 40 50
2500 8

7 Ser ies-c apac itor without bypassed
2000 Ser ies-c apac itor with bypassed by spar k gap
6
5
TRV (kV)
1500
4
1000 3
2
500
Ser ies-c apac itor without bypassed 1
Ser ies-c apac itor with bypassed by spar k gap
0 0
0 10 20 30 40 50 0 10 20 30 40 50
2500 8

7 Ser ies-c apac itor without bypassed
2000 Ser ies-c apac itor with bypassed by spar k gap
6
5
TRV (kV)
1500
4
1000 3
2
500
Ser ies-c apac itor without bypassed 1
Ser ies-c apac itor with bypassed by spar k gap
0 0
0 10 20 30 40 50 0 10 20 30 40 50
Figure 3.8.21: TRV peak and RRRV for 3LG conditions in 765 kV radial network model with
series-capacitor using the system and equipment parameters in Canada
Figure 3.8.22 shows the TRV peak and breaking currents for a 240 km LLF at 3LG conditions in the 765 kV radial
network model with a series-capacitor (not bypassed) for different compensation degrees. The TRV peak increases
due to an increase of the TRV amplitude factor on the line side. Also, the breaking current increases due to the
capacitance of the series-capacitor. TRV peaks at 3LG condition exceed the standard values if the compensation
degree exceeds 20 %. MOSA for line protection can reduce these severe TRV peaks especially in the cases of the
compensation degrees exceeding 40 %.
Figure 3.8.22: TRV for 240 km LLF at 3LG conditions in 765 kV radial network model with series-
capacitor for different compensation degrees, with and without MOSA
Figure 3.8.23 shows TRV waveforms for no-load line breaking conditions with a series-capacitor of 20%
compensation degree and without series-capacitor (bypassed) in the 765 kV parallel single circuit network model
using the system and equipment parameters from Canada. The difference in TRV peaks with and without the
series-capacitor is not significant and the interrupting current slightly decreases, because the voltage increase at
the circuit-breaker terminals before interruption is reduced due to the capacitance of the series-capacitor.
Page 69
Breaking Voltage between CB’s

Breaking Voltage between CB’s
terminals (kV)
terminals (kV)

= 841A rms
current (kA)
= 876A rms
current (kA)
Source-side
voltage (kV)
Source-side
voltage (kV)
voltage (kV)
voltage (kV)
Line-side
Line-side
Far end line-side

Far end line-side
voltage (kV)
voltage (kV)
Figure 3.8.23: TRV waveforms for no-load line breaking with 20 % series-capacitor (right: not
bypassed) and without series-capacitor (left: bypassed by BPS) in the 765 kV parallel single
circuit network model
3.8.7 Comparison of TRV for different compensation schemes (shunt reactor, series-capacitor,
HSES without shunt reactor nor series-capacitor)
TRV requirements in the network models with different compensation schemes such as shunt reactor, series-
capacitor and HSES applications without shunt reactor nor series-capacitor were investigated and it was confirmed
that the influence of different compensation schemes on TRV is not significant and severe TRVs are not observed
except in the case with a series-capacitor with the compensation degree higher than 40 % that is not bypassed by
a trigger gap or bypass switch.
Figure 3.8.24 show the TRV for BTF, and LLF conditions in the 1100 kV double circuit network model without shunt
reactor nor series-capacitor application, and Figure 3.8.25 shows the corresponding results in the case of a
1100 kV double circuit radial model with shunt reactors. Similar results are given in Figure 3.8.26 for an 800 kV
parallel single circuit network model with series-capacitor application. Both TRV peak and RRRV plots do not show
any significant differences due to different compensation schemes and are well covered by the standard values
even though some RRRV for TLF conditions slightly exceed the standard values for the terminal fault T30 testing
duty, when another circuit is open.
Figure 3.8.24: TRV for BTF and LLF conditions in 1100 kV double circuit radial network model
without shunt reactor nor series-capacitor using system and equipment parameters in Japan
Page 70
Figure 3.8.25: TRV for BTF and LLF conditions in 1100 kV double circuit radial network model
with shunt reactor using system and equipment parameters in China
Figure 3.8.26: TRV for BTF and LLF conditions in 800 kV parallel single circuit radial network
model with series-capacitor using system and equipment parameters in Canada
3.9 Conclusions
Simple extrapolation of assumptions from lower voltages is not appropriate in many cases for UHV specifications.
Several distinctive phenomena were observed that may have significant impact on UHV substation equipment due
to applications of multi-conductor bundles with large-diameter, large-capacity power transformers and higher
performance MOSA.
Extremely low voltage protective levels of higher performance MOSA lead to high specific stresses under
continuous operation conditions resulting in higher operating temperatures as well as excessive energy handling
requirements for MOSA itself, besides suppression of switching overvoltages.
The influence of system and equipment parameters on TRV were investigated in detailed using the system and
equipment parameters of different UHV and EHV projects. The results are summarized in the Tables 3.9.1 and
3.9.2.
The following conclusions can be drawn from network model analysis.
WG A3.28 investigated TRV requirements based on UHV and EHV radial and meshed network models using
system and equipment parameters for different national projects and confirmed that the TRV values specified in
IEC for rated voltages 550 kV to 1200 kV can adequately cover present system values except in particular cases of
TLF with negligible capacitance between the circuit breaker and the transformer, and in applications of
compensated lines with series capacitors not by-passed when the circuit breaker is interrupting a line fault. CIGRE
recommends that TRV values need to be introduced by IEC to cover TLF conditions for EHV. By the simulations, it
has also been shown that WG A3.22 recommendations for UHV applications are appropriate.
Page 71
In the case of terminal faults and long line faults (LLF), the influence of shunt reactors, line transposition, line height,
line sag, earth resistivity on TRV was confirmed to be small. Note that the line characteristics are applied to all lines
on the system, including the line at the load side, when applicable. LLF TRVs are covered by the standard TRV of
T10 and/or T30 and/or OP if the shortest time t2 specified for OP is used. In a number of cases considered for
550kV and 800kV, LLF is even covered if the longest time t2 specified for OP is used.
In applications with series capacitors, the TRV is covered by standard values during faults with high current as met
during terminal faults when the series-capacitors are bypassed. Severe TRV peaks are expected when series
capacitors are not bypassed; for example, in the case of low fault currents and a compensation degree higher than
40 %. Mitigating measures exist to maintain the TRV within limits defined by standards e.g. by using fast protecting
devices.
A more precise understanding of switching phenomena was obtained through simulation models, e.g. in case of
capacitive switching. The condition leading to the highest TRV peak in case of long line faults was defined, taking
into consideration the influence of a parallel line circuit.
It is relevant to note that technical characteristics like tower configuration, bundle geometry, transposition of phase
conductors, shunt and series compensation, when considered, have been adapted in the whole network model.
This means that by adapting one parameter many transient variables may be influenced, so that the impact of that
parameter on the values of such variables may become less salient than initially expected. For instance a higher
surge impedance coincides with a higher impedance per km of all OHL in the model. Consequently, the fault
current level will decrease, and the combined effects on the RRRV of the TRV regarding a higher surge impedance
and a lower short-circuit current cancel each other to a certain degree. Such a trend of reducing effects, when more
interactions among technical paramenters and variables have to be considered, can be recognised within many
topics addressed in this Technical Brochure: e.g., (a) multiple resonances in transformers’ admittance (in the
frequency domain) tend to decrease the amplitude factor for TLF; (b) travelling waves and phase interaction reduce
the amplitude factor with unloaded line switching (with/without earth fault); (c) meshed networks show a lower
amplitude factor than radial networks, etc
Page 72
System conditions Bus terminal fault (BTF) Long line fault (LLF)
Shunt reactor Influence on TRV is negligible Influence on TRV is negligible
Series-capacitor is bypassed for larger Severe TRV is expected when series
current such as BTF before the fault capacitor is not bypassed in case where
Series capacitor
clearing. In such case, severe TRV does the compensation degree is higher than
not appear. 40%
MOSA can suppress the severe TRV for
MOSA can suppress TRV peak at the
LLF at 3LG conditions in the 800 kV
rated voltage of 1100 kV and above.
MOSA networks with series-capacitor not
However, the influence is limited at
bypassed if the compensation degree is
550 kV and 800 kV networks
higher than 40%
Line height Influence on TRV is generally negligible. Influence on TRV is generally negligible.
Line sag Influence on TRV is generally negligible. Influence on TRV is generally negligible.
TRV peak is more severe with an
increase of line length, while the breaking
Line length NA
current is decreased with an increase of
line length.
Transposition Influence on TRV is generally negligible. Influence on TRV for 360 km is a few %
TRV at 3LG conditions is much more TRV at 3LG conditions is much more
3LG vs. 1LG
severe than those at 1LG. severe than those at 1LG.
Earth resistivity Influence on TRV is generally negligible. Influence on TRV is generally negligible.
Compensation scheme Influence on TRV is generally negligible. Influence on TRV is generally negligible.
Power flow Influence on TRV is not significant Influence on TRV is not significant
Table 3.9.1: Influence of various system and equipment parameters on TRV
Note: Line height and sag have a negligible influence on TRV because of presence of earth wires.
System conditions No-load line breaking (LC)

Shunt reactor First TRV peak is reduced because TRV at line side oscillates due to discharge of
trapped charges through the shunt reactor.
Series capacitor Influence on TRV is negligible
MOSA NA
Line height Influence on TRV is generally negligible.
Line sag Influence on TRV is generally negligible.
Line length Breaking current increases with an increase of line length, but not for TRV peak
Transposition Influence on TRV for 360 km is a few %
3LG vs. 1LG NA
Earth resistivity Influence on TRV is generally negligible.
Compensation scheme Influence on TRV is generally negligible.
Power flow NA
Table 3.9.2: Influence of various system and equipment parameters on TRV
Page 73
4 Transformer limited faults

4.1 Introduction
Where in CIGRE Technical Brochure 456 TLF has been addressed by a first order representation of the power
transformers, WG A3.28 investigated more in depth the effects of multiple resonance frequencies, as well as the
differences between TRVs at the primary and secondary or tertiary side and the effects of the first, second and last
pole-to-clear. Moreover, on request of IEC SC 17A, WG A3.28 paid attention to the TLF at rated voltages of 100 kV
to and including 800 kV. Apart from the TRV parameters, that include voltage drop across the transformer,
1st/2nd/3rd clearing pole factors, rate of rise of recovery voltage, fault current levels have been investigated,
especially in relationship to the voltage drop.
Transformer limited faults (TLF) are short-circuit conditions predominantly determined by a transformer in a given
circuit. In that case the transformer impedance is the dominant factor for the RMS-value of the short-circuit current.
Its X/R-ratio is determinant to the short-circuit current peak factor and its X0/X1-ratio is determinant to the earth-fault
versus multi-phase fault currents. At fault current clearing, the high-frequency characteristics of the transformer are
essential for the development of the TRV.
Four TLF conditions can be distinguished for a circuit breaker at a certain voltage level (in Figure 4.1.1: EHV). The
transformer under consideration may be one that at the other side is connected to a higher voltage level (UHV) or
to a lower voltage level (HV); the fault may be at the terminals of the transformer at that higher or lower voltage side
(transformer secondary fault: TSF) or the fault may be at the breaker terminals at the busbar side (transformer fed
fault: TFF).
UHV EHV
EHV
TSF
UHV TFF
EHV EHV
UHV TSF
UHV TFF
EHV HV HV
TSF TFF
EHV
Figure 4.1.1: Transformer secondary faults (TSF) and transformer fed faults (TFF)
4.2 Voltage drop

As the transformer reactance is the dominant impedance determining the symmetrical short-circuit current
amplitude, in the specification of the type test for transformer limited fault clearing the voltage drop along the
transformer is assumed to be close to 100%. To IEC Std 62271-100, the voltage drop is 90% in the voltage range
from 100 to 800 kV and covered by terminal fault test duty T10 at 10% of the rated short-circuit current (Isc). The
short-circuit current is relatively low: 10% to 30% of the rated short circuit current of the circuit breaker, as usually
the circuit breaker has to cope with a much larger fault current fed by other sources that are connected to the
busbar. Therefore, the peak value of the short-circuit current is not a relevant stress factor, although the X/R-ratio
will be very high. However, TRV conditions may become more severe than those specified for T10 and T30.
To show the relationship between voltage drop across the transformer and the TLF-fault current, the simple
scheme of Figure 4.2.1 is introduced. The next definitions are used:
Page 74
 Ip(net) is the short-circuit contribution by the network at the primary side, without the transformer’s
contribution, but including that of parallel transformers, when applicable.
 Is(net) is the short-circuit contribution by the network at the secondary side, without the transformer’s
contribution, but including that of parallel transformers, when applicable.
 Ip(TSF) is the TLF current with a fault at the secondary terminals, to be cleared by the CB at the primary
side.
Primary side
Secondary side
Figure 4.2.1: Simple scheme to show the definitions of fault currents

to determine the voltage drop
By Ip(TSF) and the transformer ratio the TFF short-circuit current at the secondary side, to be cleared by the
secondary CB, can be determined: Is(TFF). Moreover, by Ip(TSF) and Ip(net) the voltage-drop across the
transformer can be calculated. As a percentage the voltage drop ∆V is equal to 100% - Ip(TSF) as a percentage of
Ip(net):
Ip (TSF )
V  1 
Ip ( net )
The larger the ratio Ip(TSF)/Ip(net) the smaller the voltage drop (i.e. the smaller is the influence of the transformer
impedance).
Considerations made for the transformer secondary faults are also applicable to transformer fed faults, both for
circuit breakers at the primary side and at the secondary side.
Assuming a constant value for Ip(TSF), then the higher Ip(net) the larger is the voltage drop across the transformer
(i.e. the smaller is the net impedance). Let us express Ip(net) in relationship to the rated short-circuit current of the
breakers. Ip(net) will be less than the rated short-circuit-current, because of margins users apply, because of less
onerous system conditions and because of the contribution of the transformer itself to the total short-circuit current.
For various values of the TLF current, the following voltage drops can be deduced. In Figure 4.2.2, I(net) is either
Ip(net) or Is(net) respectively when the current is fed from the primary or the secondary side of the transformer. ITLF
refers to the corresponding transformer limited fault current shown on Figure 4.2.1. It is assumed that I(net) is 25%,
50%, 75% and 100% of the rated short-circuit current of the circuit breaker (Isc).
Page 75
Figure 4.2.2: Voltage drop across the transformer for various values of I(net) and I(TLF)
4.3 First-pole-to-clear factor

Depending on the neutral treatment of the network and of the involved transformer, the X0/X1 ratio will vary within
wide bands. Transformers with an earthed neutral on the primary side and a delta-winding on the
secondary/tertiary side will show an X0/X1-ratio smaller than 1.0. As the influence of the transformer on the short-
circuit current is rather dominant its X0/X1-ratio has a reducing effect on the first pole-to-clear factor kpp. Depending
on the network impedance and its X0/X1-ratio, the overall ratio will become close to 1.0 or even less. Consequently
kpp will be close to 1.0 as well. In other conditions, though, where transformer neutrals are not (always) connected
to earth kpp may rise up to 1.5.
At voltage levels of 100 kV and above transformer windings are Y-connected and the neutral solidly earthed (apart
from resonance earthed systems and other not-effectively earthed systems up to 170 kV). Maybe the neutral is
earthed by means of a reactor or a resistor is added to limit the ratio of the single to three-phase short-circuit
current but in that case this ratio will become close to 1.0. In conclusion, it can be stated that at the transformer
side, which is connected to networks of 100 kV and above, the first pole-to-clear factor for TLF will be close to 1.0
or even lower. The factors specified in IEC (i.e. 1.2 for UHV and 1.3 up to 800 kV) are certainly higher than those
commonly observed in real cases.
An exception on the above conclusion refers to tertiary windings (∆-windings) in UHV-transformers as applied in
China and Japan. These windings, with rated voltages of 123 or 145 kV in China and 154 kV in Japan may be used
for shunt compensation. When circuit breakers are applied in connection to such tertiary windings the duties are
very extreme and therefore cannot be considered as basis for standardization.
4.4 Comparison with network simulations

The fore-mentioned theoretical considerations can be compared with simulations of TLF-conditions in UHV and
EHV-networks. An example of the TEPCO UHV configuration will be examined.
The TLF short-circuit currents and the voltage drop across transformers in the UHV network in Japan are
investigated as follows. Figure 4.4.1 shows the distribution of three-phase to earth short-circuit currents calculated
for the network, at the stage immediately after upgrading the UHV network to four substations.
Page 76
Figure 4.4.1: Distribution of three-phase to ground short-circuit current (kA)

In Figure 4.4.2 the calculated TLF cases are presented. To simplify the calculations, it has been assumed that for
each substation shown in Figure 4.4.1 only one transformer is connected and that the circuit breaker to be
investigated interrupts after half a cycle.
Note 1: The impedance of the UHV transformer is 18 % (1050 / 525 kV, 3000 MVA base), which is the optimum taking into
consideration the suppression of short-circuit currents, voltage changes, economy, etc.
Note 2: The rated short-circuit current of UHV circuit breakers will be specified as 50 kA, and 63 kA for 500 kV circuit breakers,
considering the maximum short-circuit current contributed from lines and transformers, possible future development, etc.
3LG
UHV UHV UHV UHV
3LG
3LG
500kV 500kV 500kV 500kV

3LG
Case 1 (TFF) Case 2 (TSF) Case 3 (TFF) Case 4 (TSF)
Figure 4.4.2: Calculated TLF cases

Figures 4.4.3 and 4.4.4 give the relation between TLF short-circuit current and the ratio of the voltage drop across
the transformer to the supply voltage. The TLF current is given in kA and in percentage of rated short-circuit current
(Isc).
The voltage drop ratio of transformer tends to be larger as TLF current is larger. Maximum values are as follows.
- UHV side CB (50 kA rating): TLF current is 13 %, voltage drop ratio is 0.72
- 500 kV side CB (63 kA rating): TLF current is 23 %, voltage drop ratio is 0.72
Page 77
※100% current= Rated short-circuit current of 50 kA

0.80
Voltage drop ratio of transformer (p.u.)

Case 1 (TFF)
0.75 Case 2 (TSF)
0.70
0.65
0.60
0.55
0.50
0 2 4 6 8 10 12
(0%) (4%) (8%) (12%) (16%) (20%) (24%)
TLF short-circuit curremt (kA)
Figure 4.4.3: UHV side circuit breaker
※100% current= Rated short-circuit current of 63kA

0.80
Voltage drop ratio of transformer (p.u.)
Case 3 (TFF)
0.75 Case 4 (TSF)
0.70
0.65
0.60
0.55
0.50
0 5 10 15 20
(0%) (8%) (16%) (24%) (32%)
TLF short-circuit current (kA)
Figure 4.4.4: 550 kV side circuit breaker
For reference TRVs were also calculated. Figures 4.4.5 and 4.4.6 show the relation between TLF short-circuit
current and TRV peak voltage, for the UHV circuit breaker and the 550 kV circuit breaker.
Page 78
※100% current = Rated short-circuit current of 50 kA

2000
With MOSA
1800
0.9×1.7×1.2 =1649kV Without MOSA
1600
TRV peak voltage (kV)
1400
1200
1000
800 IEC 62271-100 am1 Ed. 2
600
(Approved in Aug. 2012)
400
200 - Current; 10 kA and 12.5 kA
0
0 2 4 6 8 10 12
(0%) (4%) (8%) (12%) (16%) (20%) (24%)
Figure 4.4.5: Relation between TLF short-circuit current and TRV peak voltage
for 1100 kV circuit breaker
※100% current = Rated short-circuit current of 63kA

1000
0.9×1.7×1.3 = 893kV With MOSA
900
Without MOSA
800
TRV peak voltage (kV)
0.9×1.7 = 687kV
700
600 IEC-1st draft-TLF
500
One of TEPCO’s specifications when
400 one-break 550 kV CB type tests were
300 done ten years or more before.
200 - Current; 15.8 kA
100
0
0 5 10 15 20
(0%) (8%) (16%) (24%) (32%)
Figure 4.4.6: Relation between TLF short-circuit current and TRV peak voltage
for 550 kV circuit breaker
Note 1: V-I characteristics of MOSA (V10kA) are 1550 kV for UHV side, 870 kV for 500 kV side.
Note 2: Transformer’s damping characteristics were assumed to give an amplitude factor of 1.7.
The examples of TEPCO give voltage drops in the same range as those that have based on theoretical
considerations. In the TEPCO-cases and in the theoretical calculations the TLF currents has been expressed as a
percentage of the rated short-circuit current of the circuit breaker. But under service conditions, it is rather the
transformer characteristics than the circuit breaker rating that determines the TLF current. It may be considered to
specify TLF currents not as a certain percentage of the circuit breaker rating, but as a fixed kA value (or range of
fixed kA-values) taken from the R10 series, as done already by IEC for UHV circuit breakers.
Page 79
4.5 Single frequency approach
Based on publications of R. H. Harner [33] a trial use test duty for transformer limited faults has been introduced in
the IEEE standards. To that purpose the transformer and the connections between transformer and circuit breaker
have been represented by a single frequency circuit, giving high RRRV-values. The circuit essentially consists of
the transformer’s short-circuit inductance, the transformer’s surge capacitance and the capacitance of the
connected equipment. The surge capacitance is not well-defined, at least for the purpose to calculate or simulate
TRV waveforms. Transformer specialists are more interested in the phenomena that occur at much faster transient
phenomena, such as lightning impulses or VFTO, and will tune their measurements for the MHz range. Moreover,
their main interest is in overvoltages between windings and within coils, rather than in overvoltages at the
transformer terminals. The bandwidth relevant to TRV phenomena is some 100 kHz, at maximum, with dominant
frequencies around ten to some tens of kHz.
In the IEEE Guide C37.011 indications are given for the capacitance values of transformers at several voltage
classes. These values are extrapolated from Harner’s measurements, which included only a limited number of
transformers for transmission voltages. In a recent publication considering a number of twenty one 230/115 kV
auto-transformers [34], it was claimed that the capacitance values for transmission transformers as indicated in
IEEE C37.011 may be too high. But the authors support the single frequency transformer model to specify TRV
requirements. These authors stress the importance of using or calculating/estimating capacitance values
representative for higher frequencies; i.e. the surge capacitance mentioned in the former paragraph.
As it is well known that a power transformer shows a much more complex response than the one related to a single
frequency model, Michael Steurer, Wolfgang Hribernik and John Brunke proposed to use FRA-measurements to
calculate TRV waveforms [35] [36]. FRA-measurements give the transformer impedance or admittance as a
function of the frequency in a wide bandwidth. By multiplying the impedance Z(ω) with the current in the frequency
domain I(ω), and performing a backwards Fourier transform, they calculated the TRV(t) in the time domain. FRA-
measurements are usually available from the manufacturer or utility, but care should be taken for proper and
adequate FRA-measurements, for instance representing the first-pole-to-clear. The effect on the frequency
response of an unloaded transformer versus a short-circuited transformer, registered by FRA-measurements, can
be seen in Figure 4.5.1.
107
Z (ohm)
106
105
LV terminal open
104
103
LV terminal shorted
102
Frequency (Hz)
104 102 103 104 105 106 107
Figure 4.5.1: FRA measurements on a 400 kV, 80 MVA transformer (red line for LV terminal open, blue
line for LV terminals short-circuited)
From Figure 4.5.2 it is clear that the lowest frequency range of the FRA-measurement shows the predominant
characteristic of an inductance with a linear increase of the impedance proportional to the frequency. The blue line
gives exactly the short-circuit inductance of the transformer. In the higher frequency range a pure capacitance will
give a characteristic that decreases linearly proportional to the frequency, as can be recognized in both curves, but
not as clearly illustrated as in the inductive part of the FRA-measurement.
Page 80
107
Z (ohm)
106
105
LV terminal open
104
103
LV terminal shorted
102
Frequency (Hz)
104 102 103 104 105 106 107
Figure 4.5.2: 420 kV 80 MVA transformer, L = 0.64 H, C = 400 pF

The inductance and capacitance deduced in this way from the FRA-measurement form the main elements for the
single frequency model. The damping, represented by a parallel resistance R, can be read off from the peak value
of the (blue) FRA-measurement around the resonance frequency due to the main L and main C: 200 kΩ. The
characteristic impedance Z  L / C  40 k . The ratio R/Z = 5 gives the amplitude factor for a single frequency
response: 1.73, as, for instance, can be learned from [46] (especially: Figure 4.7 and the formulas 4.2.26 to 4.2.28
in [46]). The relationship between R/Z and the amplitude factor AF is complicated; as can be seen in Figure 4.5.3
for R/Z between 3.0 and 10.0.
Figure 4.5.3: Relationship between R/Z and the amplitude factor AF
4.6 Multi-frequency model

To compare the response from a single frequency approach with the multi-frequency model, the FRA-plot of a
similar transformer is used; Figure 4.6.1. The deduced single frequency parameters are: 1.0 H, 350 pF and 500 kΩ
and the corresponding TRV waveform can be seen in Figure 4.6.2.
Figure 4.6.1: FRA and TRV waveforms of 80 MVA 400 kV transformer
Page 81
Horizontal axis: angle (in rad) instead of time
Figure 4.6.2: TRV waveform of single frequency model of 80 MVA 400 kV transformer
In comparison, the waveforms are different. The two-parameter envelopes show that the simple model gives more
severe TRV characteristics: an RRRV of 35 kV/μs instead of 28 kV/μs, an AF of 1.84 instead of 1.69.
The main reason for the lower amplitude factor in case of the multi-frequency model is the interference between the
three resonance frequencies around 10 kHz. The influence of resonance frequencies further away is weak, as the
Fourier transform of the current damps the higher frequencies. In a schematic way this is explained in Figure 4.6.3.
Note: The Fourier transform of the ramp-function I (t )  S  t with S  2 f 2  I rms is I ( )  S /  2
Figure 4.6.3: The effect of higher resonance frequencies
Therefore, above, say, 500 kHz, the FRA-measurement may be replaced by just a straight line representing the
capacitance, while, making no difference upon the results, at the lower frequency end of the FRA-plot, the graph
may be extrapolated with the straight line for the short-circuit inductance. It is also clear that the so-called surge
capacitance can be deduced from the straight line that can be drawn through the higher frequency part of the FRA-
plot.
For a number of transformers, measured in an identical way at the High Power Laboratory of KEMA the two-
parameter envelopes determined for the backward Fourier analysis, and determined for the simple approach
(single frequency model), have been compared: Table 4.6.1. Apart from case 4, all cases show higher RRRVs with
the simple model, and always AF is higher for the simple model. When results are close, the FRA-plot is quite
similar to a single frequency plot, as it is shown in Figure 4.6.4 for case 9, a rather small transformer.
Page 82
Case 4 Case 8 Case 9 Case 10 Case 11 Case 12

KEMA t3 (μs) 39.2 61.1 33.4 64.6 50.5 46.3
Simple t3 (μs) 43.5 53.0 35.0 49.0 56.5 33.5
KEMA AF 1.75 1.69 1.66 1.65 1.54 1.71
Simple AF 1.80 1.84 1.74 1.80 1.78 1.78
KEMA kV/μs 44.6 27.7 49.7 25.5 30.5 36.9
Simple kV/μs 41.4 34.7 49.7 36.7 31.6 53.1
Table 4.6.1: Two parameter envelopes for six transformer cases
Figure 4.6.4: Case 9 with a more or less single frequency FRA plot
Similar conclusions may be drawn from FRA-measurements on some large power transformers in Japan, in
particular a large 1500 MVA, 525/275/63 kV transformer [37]. A number of measurements, simulations and
calculations have been made for this transformer, showing a single frequency response at the 525 kV-side, a
double frequency response at the 275 kV-side and a triple frequency response at the 63 kV-side. By a current
injection method the TRV wave-shapes have been measured. By a step-voltage applied to each side of the
transformer the natural oscillation and damping has been measured (Daini-Kyodai method). And for each side an
FRA-measurement has been performed. From the FRA-measurements values, equivalent R//L//C circuits have
been deduced, whereas for the double and triple frequency simulation two and three such parallel circuits have
been put in series. In Figures 4.6.5, 4.6.6 and 4.6.7 the three models are shown together with the FRA-plots and
the comparison between measured FRA and the impedance Z(ω) of the models. Besides, the TRV-measurement
and the results from the Daini-Kyodai method are given.
Page 83
Figure 4.6.5: Measurements and modelling of 500 kV-1500 MVA transformer at 525 kV primary side
Figure 4.6.6: Measurements and modelling of 500 kV-1500 MVA transformer at 275 kV secondary side
Page 84
Figure 4.6.7: Measurements and modelling of 500 kV-1500 MVA transformer at 63 kV tertiary side
The Z(ω)-plots fit quite well with the FRA-plots, especially for the single frequency case. As visible in Figure 4.6.8,
the TRV waveforms reproduced by the Z(ω)-plots are reasonable comparable with the measured TRV-traces. The
similarity between the TRV waveform as measured by the current injection method and as deduced from the Daini-
Kyodai method is not so good (Figures 4.6.5 to 4.6.7). Also the comparison with the simple (single frequency)
method as applied to the KEMA FRA-plots shows that the dominant frequency (f) can be approached, but the
amplitude factors (AF) are too high (see Table 4.6.2).
f (meas.) f (simple) AF (meas.) AF (simple)

Primary side 8.0 kHz 7.9 kHz 1.62 1.73
Secondary side 8.5 kHz 10.0 kHz 1.40 1.69
Tertiary side 11.2 kHz 12.5 kHz 1.57 1.77
Table 4.6.2: Comparison between single frequency method and measurements
The discrepancy in the amplitude factors is mainly caused by the reducing effect of interference (interaction)
between different resonance frequencies. The lines in Figure 4.6.9 (this is Figure 8 (b) of [38]) shows the
dependence of the amplitude factor and RRRV on the frequency ratio of a higher resonance frequency to the main
resonance frequency (f2/f1) in the case of k = 0.25 and in the case of k = 0.125; k being the ratio of the amplitude
belonging to f2 to that to f1. It is shown that the amplitude factor is reduced considerably without any dissipating
elements in the circuit topology. As reference two measured TRV amplitude factors are plotted in Figure 4.6.9.
Compared with the values for k = 0.25, the measured amplitude factors (AF) are 20% lower for the primary side
and 10% for the secondary side. Some portion of the reductions of AF is caused by losses, however, at the same
time distributed circuit characteristics of the practical transformer have a large impact to the reductions of AF.
Page 85
Figure 4.6.8: Measured and calculated TRV waveforms
Figure 4.6.9: Dependence of AF and RRRV on frequency and amplitude ratio
The author of [43] draws attention to the fact that for three-winding transformers with two secondary windings the
FRA characteristics - and therefore the transient response at clearing a short-circuit current due to a three-phase
fault at any of the secondary terminals - will show different patterns, due to the slightly different geometric position
of the windings.
4.7 Advanced transformer models

The authors of [38] show that by modeling the power transformer in more detail, a much better approach of the
measured amplitude factor can be achieved, even without using dissipating elements. They applied a model, the
so-called L-C multi-mesh model, where each transformer coil of the 1500 MVA transformer has been represented
by its self- and mutual inductances and parasitic capacitances, in total 12 coils per phase (4 per winding) with 12
self-inductances and 66 mutual inductances. As they explain in [37], by adding resistance to present some losses
(< 1%), the measured TRV-waveform can be simulated with high accuracy.
To the authors of [44] and [45] the best accuracy to estimate the TRV-waveform is the direct measurement by a
current injection method. A next best choice is by an FRA-measurement in combination with or alternatively to the
voltage step-method (Daini-Kyodai). Another method would be to construct a meshed network model based on
construction details of the transformer windings. The most inaccurate method is by a single frequency model based
Page 86
on the short-circuit impedance and an estimate of the capacitance between windings, core and other parts of the
transformer.
Another aspect to be mentioned is the difference in TRV-waveform for the second and last pole to clear the fault
current in comparison to the first pole [37] [38]. From the X0/X1 ratio, which is lower than 1.0 for transformers with
earthed neutrals, it is known that the recovery voltage of the first pole is lower than for the second pole and also
lower than that for the last pole (with a recovery voltage of 1.0 p.u.). But, as the natural frequency of the zero
sequence equivalent network will be slightly higher than that of the positive sequence network, this difference in
frequencies may influence the TRV waveform. As the zero sequence network is less dominant for the first clearing
pole, its influence will be exacerbated for the second and last clearing pole. However, the time to peak shows to be
rather similar for all three poles, so that the two-parameter envelope varies only by the first, second and last pole-
to-clear factor. As, due to the low X0/X1 ratio, the last pole-to-clear factor is the highest, opposite to other fault
clearing conditions, the third pole may face the most onerous duty [47].
4.8 External capacitances

Surge capacitances of transformers range from several hundreds of pF to 1 nF per phase for power transformers
up to a few hundred MVA and from several nF to tens of nF for large transformers (> 500 MVA). Minimum values
for the external capacitances range from a few hundred pF to some nF with the higher values for the higher rated
voltages.
When the external Cext is incorporated in the Z(ω) or FRA-plot, it can be visualized by a shift of the surge
capacitance part (Ctot) to the left; Figure 4.8.1. The resonance frequencies will be reduced with a factor
Cr / (Cr  Cext ) ; Cr being the dominant capacitance at a specific resonance frequency. The resonance peak
value of Z(ω) is determined by the equivalent resistance and will not change, but the ratio R/Z will increase as
L / (Cr  Cext ) is smaller than L / Cr . A larger R/Z means a higher quality factor and thus a higher amplitude.
Figure 4.8.1: Schematic picture of shift of Z() by external capacitance Cext
In case the external capacitance is large enough, it will force the system to a more or less single frequency
response, dominated by the transformer short-circuit inductance and the transformer surge capacitance plus the
external capacitance [37]. The effect is a rather simple TRV waveform, but with a higher amplitude, more or less
the opposite as described with the multi-frequency case and the LC multi mesh model.
The connections between circuit breaker and transformer may be constructed by cables or gas-insulated busbars,
but mostly they are air insulated conductors. The distance between circuit breaker and transformer may be large,
but often distances are short as well: some tens up to hundred meters. The connected equipment is limited and
with limited capacitance to earth: current transformers, surge arresters, disconnectors, earthing switches. Some
typical examples of minimum capacitances:
 TEPCO 168 to 525 kV AIS or hybrid substations: 500 to 1100 pF (some tens of m)
 EGAT 115 to 500 kV AIS: 1 to 2 nF (30 to 150 m)
 ENEL 145 to 420 kV AIS: 1.3 to 2 nF (60 to 120 m)
 China 500 kV AIS: 2 nF (100 m)
Page 87
 HQ 230 to 735 kV AIS: 0.7 to 1.1 nF (23 to 65 m)

 NL 150 to 400 kV AIS: 300 to 1300 pF (20 to 80 m)
 Brazil 345 to 800 kV AIS: 1250 to 1650 pF (80 to 120 m).
In Figure 4.8.2 the capacitances as a function of the voltage classes are given, blue for transformers to a higher
voltage and red for transformers to a lower voltage class.
Figure 4.8.2: Minimum capacitance values per voltage class for AIS connection CB-transformer
Class 1 to 4: < 200 kV, < 400 kV, < 600 kV, ≤ 800 kV
Because of the bushings and the cable connections or the gas-insulated busbars, GIS-substations tend to show
larger capacitances. The capacitance to earth of current and voltage transformers may be very low, in the order of
150 pF, but much higher values are also possible. Capacitive voltage transformers, when applied, show higher
capacitances (nF).
From Figure 4.8.2, it may be concluded that a minimum additional capacitance of 0.5 to 1.0 nF may be assumed
for voltages between 100 and 800 kV. In many cases distances will be longer and the capacitances of involved
high voltage equipment larger, so that a more representative figure will be that of several nF. Such values have a
noticeable effect on the TRV frequency, certainly with regular power transformers (not the very large transformers
or those for the highest system voltages).
4.9 Resonance frequencies

Under special circumstances, depending on certain configurations and distances within substations, the
transformer resonance frequencies may coincide with substation resonance frequencies. Some typical situations
are studied by CIGRE WG A2/C4.39 [42] [43], but such situations with a danger of damage to the transformer are
outside the scope of WG A3.28.
The damage will occur at frequencies that correspond to internal resonances inside the windings of a transformer,
which usually are higher to much higher (tens of kHz up to MHz or more) than the external resonance frequencies
that play a role in the TRV wave-form (around 10 kHz up to some tens of kHz). Transformer manufacturers will be
able to provide a transient model that reproduces the transformer behavior up to some tens of kHz, so that its
interaction with the TRV waveform can be calculated.
4.10 Simulation results on the influence of capacitance on TRV

4.10.1 TRV TLF conditions at the primary side
TRV at the primary side of a power transformer for TLF conditions is calculated in the circuit including a power
transformer shown in Figure 4.10.1.1 using different system and transformer parameters. Since the influence of
capacitance on TRV is prominent, WG A3.28 conducted a limited survey on the minimum capacitance in the
transformer circuit to evaluate TRV requirements in case of first-pole-to-clear and second-pole-to-clear factors at
Page 88
severe side (see Figure 4.10.1.2). In the analysis, the first-pole and the second-pole are supposed to clear a three-
phase to earth fault. The short-circuit condition for the secondary side is given as 50 kA with the first-pole-to-clear
factor of 1.3 (Thevenin source).
The results show that the first-pole-to-clear factor at primary side is around 1.1 because the zero-sequence
impedance decreases due to existence of the tertiary delta-connection within large-capacity power transformers
used in 1100 / 800 / 550 kV networks.
Figure 4.10.1.3 shows the plots of TRV peak and RRRV calculated with different system and transformer
parameters. The RRRV at the primary side can be covered by the new recommendation for UHV ratings, but
exceed the existing specifications in the IEC standard for 800 kV and 550 kV ratings. The maximum RRRV at the
primary side are calculated as 10.1 kV/μs for 1100 kV, 11.2 kV/μs for 800 kV and 12.5 kV/μs for 550 kV
transformers.
B_BSZA
B-7
B_BSKA
B8-1A B9-1A
B S/S
B-9
Interruption
B-8
BTR1HA BTR2HA
BTR2LA
BTR1MA BTR1LA BTR2MA
B_MBSA
B_SOUA
Figure 4.10.1.1: Transformer circuit used to evaluate the TRV for

TLF conditions at the primary side
Line
Cl Cl
Cm Cm
Main-bus
Ctr Ctr
TR
TR
Source
Figure 4.10.1.2: Capacitance investigated in the transformer circuit
Page 89
China Japan Canada

Items
1100kV 1100kV 765kV 735kV *1
System Voltage [kV] (Source Voltage) 550 550 328 315
Rated frequency [Hz] 50 50 60 60
Short-circuit (3LG) current from source [kA] 50 50 50 45
DC time constant of short-circuit(3LG) current [ms] 120 120 120 75
S L1 [mH] 20.21 20.21 10.04 10.7
O R1 [ohm] 0.1685 0.1685 0.0836 0.14
U Impedance
L0 [mH] 65.68 65.68 32.63 11.2
R
C R0 [ohm] 0.5476 0.5476 0.2717 0.21
E Re [ohm] 81.5 81.5 68 76
TRV Circuit Ce [uF] 1.95 1.95 1.4 1
Cp [uF] 0.02 0.02 0.02 0.02
Uc [kV] 817.7 817.7 488.6 365.3
TRV
RRRV [kV/us] 2.05 2.05 2.04 2.01
Primary 3000/3 3000/3 1650/3
Rated Capacity
(MVA) Secondary 3000/3 3000/3 1650/3
Tertiary 1000/3 1200/3 36
T Primary Y 1050/√3 1050/√3 700/√3
Rated Voltage
R (kV) Secondary Y 525/√3 525/√3 300/√3
A Tertiary Delta 110 147 11.9
N
S Primary and Secondary %Xps 18 18.5 12
F Short-Circuit Impedance
Primary and Tertiary %Xpt 62 61.1 32
O (Base of primary capacity)
Secondary and Tertiary %Xst 40 34.3 18
R
M Primary - 0.49 X/R=74
E Winding resistance
Secondary - 0.24 X/R=50
R (ohm)
Tertiary - 0.08 X/R=50
Primary to ground Cpe (9000) 9000 5680
Capacitance
Secondary to ground Cse (15000) 30000 7570
(pF)
Tertiary to ground Cte (18000) 24000 18200
B Between Transformer-CB and Transformer, Ctr [pF] 5800 3000 (1750) 1620 (700)
U Main Bus (including CVTs'), Cm [pF] 8200 10000 (10000) 7400 (7400)
S Between Line-CB and Transmission line, Cl [pF] 10000 - -
1st-pole 1518.8 1534.8 (1535.3) 984.6 876.3 (882.5)
Uc [kV] 2nd-pole 1452.8 1467.8 (1469.1) 956 887.7 (894.1)
T 3rd-pole 1366.9 1370.4 (1372.8) 919.2 898.2 (904.6)
L 1st-pole 1393.1 1384.2 (1378.5) 860 814.4 (808.5)
F u1 [kV] 2nd-pole 1329.1 1320.0 (1315.5) 834.2 825.12 (819.1)
R 3rd-pole 1248.7 1232.1 (1228.4) 801.2 834.96 (828.9)
E 1st-pole 8.64 9.54 (10.07) 11.2 10.53 (11.24)
S RRRV [kV/us] 2nd-pole 8.24 9.11 (9.61) 10.94 10.73 (11.45)
U
L 3rd-pole 7.75 8.47 (8.95) 10.49 10.89 (11.63)
T TRV Frequency [kHz] 2.97 3.21 (3.41) 6.13 6.33 (6.54)
S kpp 1.1 1.1 (1.1) 1.06 0.97 (0.97)
ksp 1.06 1.06 (1.06) 1.04 0.99 (0.99)
Breaking Current [kA] 6.9 6.8 (6.8) 7.8 7.4 (7.4)
Table 4.10.1.1: TRV results for TLF conditions at the primary side
Page 90
Note: values in black are for actual values of the capacitance between transformer and circuit-breaker, values in red are for lowest possible
values of the capacitance.
Canada Thailand
Items
550kV 550kV
System Voltage [kV] (Source Voltage) 260 254
Rated frequency [Hz] 60 50
Short-circuit (3LG) current from source [kA] 50 50
DC time constant of short-circuit(3LG) current [ms] 120 120
S L1 [mH] 7.955 9.327
O R1 [ohm] 0.0663 0.0777
U Impedance
R L0 [mH] 25.853 30.313
C R0 [ohm] 0.2154 0.2525
E Re [ohm] 69 83
TRV Circuit Ce [uF] 1.1 0.92
Cp [uF] 0.02 0.02
Uc [kV] 387.8 378.5
TRV
3rV [kV/us] 2.05 2.02
Primary 1200/3 1000/3
Rated Capacity
(MVA) Secondary 1200/3 1000/3
Tertiary 150/3 50/3
T Rrimary Y 512.5/√3 525/√3
Rated Voltage Secondary
R Y 242/√3 242/√3
(kV)
A Tertiary Delta 12.6 22
N
S Primary and Secondary %Xps 17 17
Short-Circuit Impedance
F Primary and Tertiary %Xpt 141.3 246.4
(Base of primary capacity)
O Secondary and Tertiary %Xst 121.8 220
R
M Primary - 0.1707
E Winding resistance
Secondary - 0.06
R (ohm)
Tertiary - 0.062
Primary to ground Cpe (4000) 4950
Capacitance
Secondary to ground Cse (2500) 5940
(pF)
Tertiary to ground Cte (12000) 10460
B Between Transformer-CB and Transformer, Ctr [pF] - 1600 1600
U Main Bus (including CVTs'), Cm [pF] 3000 3000 22000
S Between Line-CB and Transmission line, Cl [pF] - - 1500
1st-pole 773.4 780.7 792.2
Uc [kV] 2nd-pole 732.9 737.3 749.7
T 3rd-pole 672.1 672.3 685.2
L 1st-pole 693.6 700.8 742.2
F u1 [kV] 2nd-pole 654.4 661.1 700.5
R 3rd-pole 595.2 601.6 638.9
E 1st-pole 12.47 10.54 8.3
S RRRV [kVus] 2nd-pole 11.78 9.95 7.85
U
L 3rd-pole 10.51 8.91 7.14
T TRV Frequency [kHz] 8.13 6.8 5.23
S
kpp 1.12 1.12 1.12
ksp 1.08 1.08 1.08
Breaking Current [kA] 6.3 6.3 5.2
Table 4.10.1.2: Additional TRV results for TLF condition at the primary side
Page 91
2000 20
1100 kV TRV peak for T10 duty
1100 kV RRRV for T10 duty
1500 15
TRV peak Uc (kV)
RRRV (kV/ s)
1000 10
500 5
1100 kV TRV for TLF condition
0 0
0 2 4 6 8 10 0 2 4 6 8 10
Breaking Current (kA) Breaking Current (kA)
2000 20

1500 15
TRV peak Uc (kV)
1000 RRRV (kV/ s) 10

500 5
0 0
0 2 4 6 8 10 0 2 4 6 8 10
2000 20
1500 15
TRV peak Uc (kV)
RRRV (kV/ s)

1000 10
500 5
550 kV TRV for TLF condition 550 kV RRRV for T10 duty
0 0
0 2 4 6 8 10 0 2 4 6 8 10
Figure 4.10.1.3: TRV parameters for TLF conditions at the primary side
4.10.2 TRV TLF conditions at the secondary side

TRV at the secondary side for TLF is calculated in conformance with the circuit including a power transformer
shown in the Figure 4.10.2.1 using different system and transformer parameters. The TRV requirements are
evaluated for the first-pole-to-clear and the second-pole-to-clear at the secondary side. The short-circuit condition
for the primary side is given as 50 kA with the first-pole-to-clear factor of 1.2 or 1.3 (Thevenin source).
The results of the Table 4.10.2.1 show that the first-pole-to-clear factor at secondary side is less than 1.03 because
the zero-sequence impedance decreases due to existence of a tertiary delta-connection within large-capacity
power transformers used in 1100 / 800 / 550 kV networks. Figure 4.10.1.2 shows the plots of TRV peak and RRRV
calculated with different system and transformer parameters. The RRRV at the secondary side exceeds the
existing specifications in the IEC standard for 525 kV (1100 kV for primary), 300 kV (800 kV for primary) and
Page 92
242 kV (550 kV for primary) ratings. The maximum RRRV at the secondary side is calculated as 11.0 kV/μs for
550 kV, 13.3 kV/μs for 300 kV and 17.0 kV/μs for 245 kV transformers for secondary side TFF.
B_BSKA
B8-1A B9-1A
B S/S
B-9
B-8
BTR1HA BTR2HA
BTR1LA BTR2LA
BTR1MA BTR2MA
B10_1A
Interruption B-1 0
B-1 1
B10_2A
B_MBSA
Figure 4.10.2.1: Transformer circuit used to evaluate the TRV

for TLF conditions at the secondary side
4.10.3 Summarizing
The simulation results confirm the theoretical considerations in the former sections.
The simulation results also conform that the present TRV specifications in the IEC Standard for TLF conditions in
the voltage range from 100 kV up to and including 800 kV are not covering the circumstances with a limited
capacitance between transformer and circuit breaker.
Page 93
China Tepco HQ BCH EGAT

Items
550/1100kV 550/1100kV 328/765 260/550kV 254/550kV
System Voltage [kV] (Source Voltage) 1100 1100 765 550 550
Rated frequency [Hz] 50 50 60 60 50
Short-circuit (3LG) current from source [kA] 50 50 50 50 50
DC time constant of short-circuit(3LG) current [ms] 120 120 120 120 120
S First-pole-to-clear factor (kpp) 1.2 1.2 1.3 1.3 1.3
O L1 [mH] 40.43 40.43 23.41 16.83 20.21
U R1 [ohm] 0.3369 0.3369 0.1951 0.1403 0.1685
Impedance L0 [mH]
R 80.86 80.86 76.08 54.7 65.68
R0 [ohm] 0.6738 0.6738 0.6341 0.456 0.5476
C
Re [ohm] 59.5 59.5 63.5 66.0 81.5
E
TRV Circuit Ce [uF] 3 3 3 2.34 1.95
Cp [uF] 0.02 0.02 0.02 0.02 0.02
Uc [kV] 1617.9 1617.9 1137.9 819.4 817.7
TRV 3rV [kV/us] 2.05 2.05 2.04 2.04 2.05
Primary 3000/3 3000/3 1650/3 1200/3 1000/3
Rated Capacity Secondary 3000/3 3000/3 1650/3 1200/3 1000/3
(MVA) Tertiary 1000/3 1200/3 36 150/3 50/3
Rrimary Y 1050/√3 1050/√3 700/√3 512.5/√3 525/√3
T Rated Voltage
Secondary Y 525/√3 525/√3 300/√3 242/√3 242/√3
R (kV) Tertiary Delta 110 147 11.9 12.6 22
A Primary and
%Xps 18 18.5 12 17.0 17.0
N Short-Circuit Secondary
S Primary and
Impedance (Baseof Tertiary
%Xpt 62 61.1 32 141.3 246.4
F primary capacity Secondary and
O %Xst 40 34.3 18 121.8 220.0
Tertiary
R Primary - 0.49 X/R=74 - 0.1707
M Winding resistance
Secondary - 0.24 X/R=50 - 0.06
E (ohm)
Tertiary - 0.08 X/R=50 - 0.062
R Primary to ground Cpe (9000) 9000 5680 (4000) 4950
Capacitance Secondary to
Cse (15000) 30000 7570 (2500) 5940
(pF) ground
Tertiary to ground Cte (18000) 24000 18200 (12000 10460
Between Transformer-CB and Transformer
1000 1000 750 500 500
Ctr [pF]
B
Main Bus (including CVTs')
U Cm [pF]
- - - - -
S Between Line-CB and Transmission line
Cl [pF]
- - - - -
1 689.5 642.8 377.3 376.5 375.0
Interruption
Uc [kV] 2 781.7 750.8 423.9 349.0 344.2
order
T 3 811.5 805.9 445.2 361.9 354.4
L 1 645 603.1 346.2 350.0 353.5
Interruption
F U1 [kV]
order
2 759 - 403.9 343.1 -
3 792.4 725.0 392.6 344.1 344.7
R 1 8.49 5.78 12.54 16.56 9.33
Interruption
3rV [kV/us] 2 10.46 6.7 12.82 16.38 9.16
E order
3 11.03 6.56 13.34 16.95 9.00
S TRV Frequency [kHz] 5.86 3.26 14.8 19.7 11.4
U kpp 0.86 0.85 0.83 1.02 1.03
L ksp 0.94 0.92 0.94 1.01 1.02
T Breaking Current [kA] 16.2 15.9 23.1 15.5 13.0
S Ratio of TLF current for rated breaking current(50kA) [%} 32.4 31.8 46.2 31.0 26.0
Voltage drop of transformer [%} 56.6 57.3 42.6 56.7 61.2
Wave forms Fig.2.1 Fig.2.2 Fig.2.3 Fig.2.4 Fig.2.5
Table 4.10.2.1: TRV for TLF condition at the secondary side
Page 94
2000 20
1500 15
TRV peak Uc (kV)
RRRV (kV/ s)
1000 10
500 5
0 0
0 2 4 6 8 10 0 2 4 6 8 10
800 20
600 15
TRV peak Uc (kV)
RRRV (kV/ s)
400 10
200 5
0 0
0 2 4 6 8 10 0 2 4 6 8 10
600 20
450 245 kV TRV peak for T10 duty 15

TRV peak Uc (kV)
RRRV (kV/ s)
300 10

150 5
0 0
0 2 4 6 8 10 0 2 4 6 8 10
Figure 4.10.2.2: TRV parameters for TLF at the secondary side
4.11 Conclusions
WG A3.28 investigated the transients related with transformer limited fault clearing by circuit breakers with a rated
voltage from 100 kV up to and including 800 kV. The results of the studies lead to the following conclusions.
 The dominant frequency to determine TRV parameters can be determined from the short-circuit impedance
and surge capacitance of the transformer. The surge capacitance can be deduced from FRA-plots in the
frequency range from some tens of kHz up to a few hundred kHz at most.
 The additional capacitance of the connection between transformer and circuit breaker is some nF and cannot
be neglected with respect to the TRV frequency (and amplitude factor). Minimum values range from 0.5 to 1 nF.
Page 95
 In several publications the frequency dependent behaviour of the short-circuit inductance has been addressed
[39] [40], but for (large) power transformers the relevance of this phenomenon still has to be investigated. It is,
however, beyond the scope of this Technical Brochure.
 When estimating the two-parameter characteristics of the TRV-envelope, a simple, single frequency model of
the transformer with its additional capacitance tends to show in general higher RRRV-values and always a
higher amplitude factor. Besides giving rather conservative values, the required parameters for the simple
method are quite easy to obtain.
 There is not enough statistical information collected on transformer natural frequencies to give
recommendations for the standardization of the time to TRV peak.
 The voltage drop across the transformer is a function of the ratio for the transformer impedance to the short-
circuit impedance of the substation, without the influence of the transformer itself. The higher the TLF current,
the larger the voltage drop will be; see Figure 4.2.2. A fault current of, say, 6.3 kA will give a voltage drop of
roughly 90% while a fault current of 12.5 kA will show a voltage drop more closer to 70%. Note that 12.5 kA in
practice approximates 30% of the rated short-circuit current of the circuit breaker [41].
 As the TLF current is mainly dependent on the transformer characteristics and not on the circuit breaker rating,
fixed values for the fault current instead of a percentage of the rated short-circuit current are recommended.
 Usually for EHV-transformers the X0/X1-ratio is less than 1.0, so that, depending on the network X0/X1-ratio, the
last-pole-to-clear factor may become the highest. In that case, depending on the amplitude factor belonging to
the first, second, or third pole-to-clear, the last clearing pole may face the most onerous condition. A first-pole-
to-clear factor of 1.3 (100 kV up to and including 800 kV) or 1.2 (UHV) is regarded as being at the safe side of
actual values, which are closer to 1.0. Though an exception are networks with non-effective earthed neutrals
(possible up to and including 170 kV).
 The amplitude factor is influenced by the mutual interaction of several natural frequencies of the transformer
coils. As power transformers are designed for low losses, dissipating elements play only a role at fine-tuning
the TRV waveforms. An amplitude factor of 1.7 is regarded as belonging to maximum values.
 Calculations and simulations show that the present TRV specifications in the IEC Standard for TLF conditions
in the voltage range from 100 kV up to and including 800 kV are not covering the circumstances with a limited
capacitance between transformer and circuit breaker.
Page 96
5 High-speed earthing switch (HSES)

5.1 Introduction
5.1.1 Background information on HSES
High-speed earthing switches (HSES) are special devices with particular requirements for switching induced
currents. In some references such as [53] to [57], this earthing switch is referred to as a HSGS (High Speed
Grounding Switches).
Several technical solutions are available to extinguish the secondary arc within a short time, as can be learned
from Table 2.3.1. At higher rated voltages four-legged shunt reactors and TPAR are the most frequently used
technological solutions. A very reliable alternative is the application of HSES. Before describing the specific duties
of HSES, attention will be given to four-legged SR in order to compare the benefits of several technologies.
Single-phase or multi-phase auto-reclosing schemes are generally applied for high-voltage transmission systems to
enhance system reliability. When a fault occurs on an overhead line, circuit-breakers located at both ends of the
line open to clear the fault. As more than 90% of the faults are single phase to earth faults, single phase auto-
reclosure (SPAR) is applied in most situations. In the case of high-voltage overhead lines (especially for system
voltages equal to or greater than 550 kV), where the conductors are located in the vicinity of each other and the
transmission systems are single phase operated, a low current may remain at the fault point after interruption of the
short-circuit current. This current, called a secondary arc current, is caused by electrostatic or electromagnetic
coupling with the other adjacent live conductors and may be difficult to self-extinguish in a short time. From a
system stability point of view it is preferable to apply an auto-reclosing scheme with a reclosing time on the order of
1 s maximum. To achieve auto-reclosing in this time some means are necessary to extinguish the secondary arc
before re-closing the circuit breakers.
Especially for short distance lines without shunt reactors or for double circuit systems with multi-phase auto-
reclosing schemes, and where four-legged reactors are not suitable, one of the useful and important means is to
apply a special earthing switch for the purpose of secondary arc extinction. This earthing switch is generally
designed for high-speed operation to ensure that the required switching performance is met, and is called a high-
speed earthing switch (HSES).
This HSES should be distinguished from the fast acting earthing switch. Table 5.1.1.1, taken from IEC [71], gives a
comparison of the different types of earthing switches.
The following main differences distinguish an HSES from the other types of earthing switches:
- The HSES needs to be operated in a well-defined operating cycle,
- It needs a clearing capability that is higher than that specified for some earthing switches with defined induced
currents together with a defined TRV withstand capability.
While an earthing switch, as well as a fast acting earthing switch, requires the capability to withstand the full short
circuit current, the function of a HSES is to ground a line and thereafter to clear the induced current and to
withstand the related TRV.
Page 97
Requirement Earthing switch class E0 Earthing switch with short- High speed earthing switch
circuit current making for secondary arc extinction
capability class E1 (and E2) (HSES)
Closing operation Low speed, hand or motor Fast (high-speed) closing Fast (high-speed) closing
operated operation operation, controlled
Opening operation Low speed, hand or motor Low speed, may be hand Fast opening, controlled
operated operated
Making capability None, shall carry the full Shall be able to make and to Shall be able to make and to
short-circuit current for a carry the full short-circuit carry the full short-circuit
specified time current current
Interrupting None If specified Shall be able to interrupt
capability induced current and to
withstand the associated TRV
Operating cycle None Close Close- open
Electrical endurance Withstand capability against 2 (or 5) closings against full 2 closings against full short
full short circuit current short circuit current circuit current
Table 5.1.1.1: Comparison of earthing switches

5.1.2 Secondary arc and its extinction
Secondary arcs, induced by sound lines at system voltage, are observed when a phase to earth fault is generated.
Single phase or multi-phase auto-reclosing can be completed successfully after the secondary arc extinction [1].
The secondary arc extinction performance depends on the primary current amplitude and duration, the degree of
shunt compensation, and will also be influenced by the recovery voltage and secondary arc current at the fault
location, both of which are influenced by the following:
- System voltage;
- Overhead line / tower configuration, e.g. single or double circuit lines, distance between phases and circuits,
height of lines above ground level, etc.
- Length of the transmission line;
- Transposition of the transmission lines (un-transposed or transposed);
- Wind speed, weather conditions;
- Occurrence of successive earth faults on the other line,
Therefore, the time duration between duty cycles is specified by the user.
The secondary arc generated in OHL is generally self-extinguished within one second. However, under certain
conditions, especially at system voltages of 500 kV and above, including UHV, secondary arcs may not extinguish
within one second. In such cases, in order to extinguish the secondary arc in the desired time, certain
countermeasures are required.
The principle of secondary arc generation is explained in Figure 5.1.2.1. After the occurrence of a single line fault,
circuit breakers at the both ends of the faulted phase open to clear the fault. After clearing the fault, the
electrostatic and/or electromagnetic induction from the healthy phases generate a certain voltage in the faulted
phase, in addition hot gases due to the fault current exists around the faulted point on the line. As a consequence,
a small arc current continues to flow to ground, this is the secondary arc.
If the dielectric strength does not recover at the fault location because the secondary arc continues to flow until the
time of re-closing of the CBs, re-closing may fail and system stability may be threatened. Therefore it is important to
extinguish the secondary arc within a desired time. As a countermeasure for the extinction of secondary arc in the
desired time, the application of HSES and four-legged reactors are possible.
Page 98
Transmission fault occurs
CB CB
Primary arc
CBs clear the faulted phase
Hot Gas
Electrostatic induction
Secondary arc induced from sound phases
Electromagnetic induction
from sound phases
Secondary arc
Figure 5.1.2.1: Secondary arc
5.1.3 Secondary arc extinction by four-legged shunt reactor

The use of four-legged shunt reactor schemes to enable single-pole auto-reclosing (SPAR) was proposed by
Knudsen [49] and Kimbark [50] in the early 1960s. The basic principle is as follows: When a single-phase line-to-
ground (1LG) fault occurs on a transmission line, the line circuit breakers are single-pole tripped to clear the fault,
i.e. they interrupt the so-called primary fault current. However, due to electrostatic and/or electromagnetic coupling
with the healthy phases, a secondary arc current may flow through the previous arc channel of the primary fault
current [66]. Higher operating voltages and long OHLs lead to a larger secondary arc current, that varies between
less than 10 A to more than 100 A. Electrostatic coupling is the primary contributor to generating the secondary arc
compared to the electromagnetic coupling with the current increasing in proportion to the line length. The strength
of electromagnetic induction varies depending on the location of the fault on the line and increases when the line
length exceeds 200 km.
Depending on the voltage and the tower configuration with line lengths up to about a hundred km, the secondary
arc current will usually extinguish within one second without special countermeasures. Line lengths of a few
hundreds of km or more require shunt reactors to control the reactive power generated by the line and the voltage
along the line. Shunt reactors compensate the capacitive coupling with the healthy phases and, thereby, facilitate
extinction of the secondary arcs. When proper values are applied for the inductances, arc extinction occurs within
desired re-closing times, which are generally less than one second.
Through the shunt reactors and the interphase capacitances, an electromagnetic induced voltage appears across
the secondary arc. The strength of electromagnetic induction depends on the load current through the healthy
phases and its direction. Further, it may vary depending on the fault location along the line. Electrostatic induction
by capacitive coupling is the primary contributor to sustaining the secondary arc, whereas the electromagnetic
induced voltage plays a relevant role with respect to the recovery voltage after an attempt to self-extinguish.
Figure 5.1.3.1 shows a three phase circuit with single line ground fault on phase C. The complicated configuration
of capacitances can be replaced by the Neptune-scheme, in a similar way as shown for four-legged shunt reactors
in Figure 5.1.3.2. This applies for equal as well as unequal values of the capacitances. If the circuit is fully
transposed, three capacitances between phases (Cij) and three capacitances between each phase to ground (Cii)
are respectively equal.
By adding three reactors (Lij) in order to cancel the coupling between phases, the electrostatically induced voltage
on phase C will also be cancelled, and, therefore, the secondary arc by the electrostatically induced current by
Page 99
capacitances (Cij), is theoretically not generated. For an un-transposed single circuit line, unbalanced reactors
effectively contribute to this cancellation.
In cases where reactive compensation of lines is needed (usually compensation by shunt reactors is in the order of
40-50%), reactors are installed at the both ends of the circuit and the value of each reactor is equal to the needed
capacity divided by two.
An additional neutral reactor contributes to compensation of the phase to phase coupling for secondary arc
extinction, which is lacking when normal three phase reactors are used. The combination of earthed (for
compensation) and unearthed reactors (for secondary arc extinction) is generally replaced by four-legged reactors.
Four-legged shunt reactors, as shown in Figure 5.1.3.2, offer the possibility to tune for both the compensation of
reactive power and the electrostatic coupling. The maximum secondary arc current decreases with an increase of
the compensation factor (< 100 %) and decreases with decreasing line length [63].
Figure 5.1.3.1: Three phase circuit (fully transposed)
Figure 5.1.3.2: Principle of four-legged reactor
In the case of a double circuit line, effects between the circuits have to be considered. Theoretically, 15 (number of
2-combinations with 6 elements or  ) reactors between phases are necessary. However, it is difficult to install
6
2
such reactors between phases. In reality, 6 （2 x 3 phases） reactors or two four-legged reactors are installed. In this
case it is impossible to compensate completely the capacitance between the six phases of a double circuit.
For non-transposed lines, capacitances to be compensated are different in each phase. Although it could be
possible to design shunt reactors with adapted inductances per phase, this is not practical. Moreover due to
weather conditions the capacitances may vary. For double circuit-lines, the theoretical optimum cannot be achieved
since the fault location and the healthy phase currents play a role, among others.
A modified four-legged shunt reactor for non-transposed transmission lines was proposed and installed on AEP
765 kV lines in the late 1970s [63] [69] [70] [72].
It has also been applied to 750 kV-lines in Russia [68]. This shunt reactor configuration, shown in 5.1.3.3,
effectively compensates unequal phase-to-phase line capacitances of the non-transposed lines. The non-
Page 100
transposed single circuit 765 kV OH- lines have a tower configuration with the three phases in one horizontal plane.
This gives an equal capacitance between the outer phases and the middle phase, but a much smaller capacitance
between the outer phases. In fact the middle phase bundle-conductor shields more or less the other outer phase.
That effect has been used by the special switching scheme of the four-legged shunt reactors. In case of an earth
fault in one of the outer phases, the shunt reactor pole of the other outer phase is switched-off.
Conventional 4 Legged Modified 4 Legged

Shunt Reactor Shunt Reactor
Figure 5.1.3.3: Schematic diagrams of a conventional four-legged shunt reactor bank and the modified
four-legged shunt reactor bank used on AEP non-transposed lines.
In combination with series compensated lines, four-legged shunt reactors may not be effective in limiting secondary
arc currents. By-passing the series capacitors during single pole auto-reclosure may solve this problem [65]. Also,
under certain circumstances, such as with a higher compensation factor, harmonic currents and overvoltages may
occur that, despite solutions as described above, will hamper the application of single phase auto-reclosure in
combination with shunt reactors [30]. When four-legged shunt reactors are useful to limit secondary arc currents, it
is obvious that they should be directly connected to the OHL and not to busbars in substations or to tertiary
windings of transformers.
Reactance values of the four-legged reactor have to be tuned and modified when transmission line configurations
and/or connections are modified.
Another solution to extinguish the secondary arc within a desired time is the application of high-speed earthing
switchs (HSES).
Utilities can choose the HSES solution when the reactive compensation by shunt reactors is not needed, or if they
want to extinguish the secondary arc in the required time with high confidence under various fault modes.
Especially for short length lines without shunt reactors or for double circuit systems with a multi-phase auto-
reclosing scheme, where four-legged reactors are not suitable, one of the most useful and important solution is to
apply a special earthing switch for the purpose of secondary arc extinction. Table 5.1.3.1 summarizes the
comparison between four-legged reactor and HSES [71].
Four-legged reactor HSES

- Effective for single-phase faults that hold the
majority of the faults.
Secondary arc
- Difficult to choose a reactance value of reactors that - Quick extinction for all fault modes.
extinction
effectively reduce the secondary arc current for all
fault modes.
Flexibility to the - In case a substation is constructed in the middle of
- No effect on the substation equipment
change of a line, it might be required to substitute a reactor that
that was already installed.
network has already installed.
- Automatic sequential control such as fault
Control - Special control is unnecessary for secondary arc detect  HSES close  HSES open  CB
/Protection extinction. close” is necessary in each phase, which
can be easily realized.
Page 101
- Four-legged shunt reactor is appropriate for transmission lines which require shunt reactors for
Economy
voltage control, while HSES would be economical for the lines without shunt reactors.
- Detailed analysis is necessary in order to not cause
- A highly reliable control system is
resonance between shunt reactor inductance and
Concern required since a mal-function leads to a
line capacitance not only at the power frequency of
ground fault.
50/60Hz but also in the high frequency band.
Table 5.1.3.1: Comparison between four-legged reactor and HSES
5.2 Secondary arc Extinction by HSES

5.2.1 Operation of HSES for secondary arc extinction
The operating sequence of HSES is determined by the time desired to maintain system stability, the high-speed
auto-reclosing sequence of the circuit-breaker, the dielectric recovery characteristics at the fault location on the
transmission line, and time coordination with protection relays including the time for confirming the condition of the
open/close condition of the circuit-breaker and HSES.
The basic operating sequences of HSES and CB are shown in Figure 5.2.1.1 with typical timing. In this example a
single-phase line-to-earth fault occurred on the overhead transmission line. Both CB at the ends of the affected line
are opened. After this, both HSESs at both line ends are closed and the secondary arc will be extinguished.
Afterwards, both HSESs at both line ends are opened. In case that there is a certain time difference in interrupting
instant between two HSESs, one HSES opens first and interrupts the electromagnetically (acronym EM) induced
current. Later the second HSES opens and interrupts the electrostatically (acronym ES) induced current. At the end
of the sequence both CBs at the line ends will close.
Figure 5.2.1.2 shows typical operation of CBs under the multi-phase auto-reclosing scheme applied to double
circuit overhead lines.
It is preferable to re-close the CBs as early as possible after an earth fault on the transmission line has occurred.
One second is generally applied for EHV/UHV systems in order to ensure dielectric recovery of the air at the faulted
point [2] and to avoid re-ignition at the faulted point on the line due to the hot gas and/or plasma.
Figure 5.2.1.1: Operation of HSES
Page 102
Figure 5.2.1.2: Multi-phase auto-reclosing scheme
Figure 5.2.1.3 shows the relation between insulation recovery voltage at the faulted point of the transmission line
and switching surges by the CB and HSESs [59].
A typical timing chart of the relationship between the transmission line circuit-breakers that interrupt the fault and
the HSESs is shown in Figure 5.2.1.4. This figure shows the first O - C operation of the circuit breakers and the first
C – O operation of the HSESs.
Figure 5.2.1.3: Recovery voltage at faulted point
Page 103
1 Energizing of the closing circuit of the HSES

2 Current start in HSES
3 Contact touch of HSES
4 Energizing of the opening release of the HSES
5 Contact separation of HSES
6 Arc extinction in HSES
Figure 5.2.1.4: Timing chart of the HSES in relation to the transmission line circuit-breakers
Figure 5.2.1.5 shows typical values of an operating sequence assuming a time interval of one second from the
initiation of a fault to the completion of reclosing of the circuit-breakers at both ends.
time
0 100 200 300 400 500 600 700 800 900 1000
(miliseconds)
First fault
⑭
①
Circuit breaker
⑮
② ⑪
③
Protection
relay ④ ⑫
⑤ ⑦ ⑬
HSES
⑯ ⑧ ⑩
successive
fault occurs in A
B
the adjacent
phases/lines C
Figure 5.2.1.5: Typical timing chart showing the time between fault initiation and a successful re-close of
the transmission line circuit-breakers
Page 104
Notes
A There may be successive faults. However these successive faults do not affect the HSESs interruption since the successive
faults on the other phases/ lines will have been cleared by CBs prior to the HSESs opening
B Successive fault may affect HSESs interruption. Common value of break time is up to 100 ms.
C Arcing time may be longer in case delayed current zero phenomena occurs
1 CB1, CB2 open 2 Confirmation of CB1 and CB2 in open position
3 Main relay function recovery 4 Confirmation of re-close condition
5 HSES1, HSES2 close command 6 HSES1, HSES2 close
7 HSES1, HSES2 open command 8 HSES1, HSES2 opening time
9 HSES1, HSES2 arcing time 10 HSES1, HSES2 open
11 Confirmation of HSES1, HSES2 in open position 12 Confirmation of CB1,CB2 re-close condition
13 CB1, CB2 close command 14 CB1, CB2 re-close at 1 s
15 CB1, CB2 remain open 16 HSES1, HSES2 remain closed
The HSESs need automatic sequential control for each phase such as, fault detection – circuit-breakers open -
HSESs close - HSESs open – circuit-breakers close. The HSESs need a high reliable control system since a
maloperation will lead to an earth fault. The HSES is also able to interrupt the induced current and to withstand a
TRV caused by electromagnetic and/or electrostatic coupling effects. Note that the primary fault current has been
cleared by the circuit-breakers at both ends of lines.
For HSES, a successive fault would be of concern if an additional earth fault occurs in an adjacent phase circuit(s)
during the time interval of a single-phase earth fault prior to reclosing by the circuit-breakers. If an additional fault
occurs during CB re-closure and HSES sequential operation, the breaking duty of the HSES will be more severe
(see Figure 5.2.1.5).
5.2.2 HSES requirements for different projects

The first HSES was applied in early 1980's in USA 500 kV systems [53][54] and since 2002 HSES have been
applied in Korea in their 765 kV system [55]. For 1100/1200 kV AC transmission systems HSES are envisaged for
application. [56][60].
Table 5.2.2.1 shows technical requirements for HSES applied in different projects.
Requirements USA Korea Japan

Highest voltage (kV) 550 800 1100
Interrupting current (A) 700 8000 7000
TRV peak (kV) 260 700 900
RRRV (kV/µs) 0.2 1.3 1.15
Table 5.2.2.1: HSES requirements for different projects
5.3 Successive faults

5.3.1 Possibility and Influence of successive faults on HSES interruption
During a period of eight years, lightning behaviour has been observed in TEPCO’s 1100 kV OHL, that are operated
at 500 kV [58][67]. The observation system, see Figure 5.3.1.1, consists of
 5 observation devices to detect lightning surges in substations
Page 105
 48 cameras to record lightning strokes
 60 devices to measure lightning currents through OHL towers
 a lightning localization system.
Major results are summarized in Table 5.3.1.1. It should be noted that the direct lightning strokes on phase lines
account for 88% of the observations. Figure 5.3.1.2 shows an example of direct lightning.
In Table 5.3.1.2 successive faults (i.e. direct lightning surges within one second) are listed, as measured in the
substations. Yellow cells indicate the case of a 1LG occurrence. These data are for cases of multiple lightning
strokes. It should be noted that in case (2) multiple lightning strokes on different phases occured.
Other field data obtained by the lightning location system are shown in Figure 5.3.1.3. Typical lightning location
observed in the summer has 11 strokes in area I and 9 strokes in area II, all observed within 1 second. There are
many successive lightning strokes in the close vicinity of or on the UHV transmission line within one second. From
these data it is shown that there is certain probability that multiple flashovers in different phases may occur.
East-West Route
240km [1999]
Minami-Iwaki
Kashiwazaki-Kariwa
Higashi-
Nishi- Gunma Shin-Imaichi
Gunma
North-South Route
190km [1993]
Tokyo
Higashi-
Yamanashi
Figure 5.3.1.1: Observational location map
Fault rate of UHV designed transmission lines *

(cases / 100 km-year)
Direct lightning Back- flashover Direct or Back- Total of

strokes 1LG 1LG flashover 2LG lightning fault
0.617 0.059 0.0294 0.705
*now operating at 550kV.
Table 5.3.1.1: Observational results (8 years)
Page 106
a) Direct lightning stroke to the upper phase line b) Direct lightning stroke to the middle phase line and flashover
Figure 5.3.1.2: Observation of direct lightning stroke to UHV transmission line
Transmission line date No. time Phase of lightning
1 02:14:06.895 1L-Middle
(1) Higashi-Gunma 2000/05/08
2 02:14:07.077 1L-Middle
1 12:41:17.850 1L-Upper
(2) Minami-Iwaki 2000/07/04 2 12:41:18.028 1L-Lower
3 12:41:18.missed 1L-Lower
1 11:29:14.957 2L-Upper
(3) Minami-Niigata 2004/08/07
2 11:29:15.449 2L-Upper
Table 5.3.1.2: Measurement data of the direct lightning surges at substations

Note: Yellow cells show 1LG occurrence
Figure 5.3.1.3: Multiple lightning observed by Lightning location system

Note: Observation time: 1 hour, 18:00-18:59, 2002/7/21, red underline shows the lightning strokes in 1 second.
Page 107
5.3.2 Impact of successive fault on network and user’s policy on UHV / EHV line
Multiple lightning strokes within one second as well as successive direct lightning strokes were definitely observed
along the UHV OHL. The possibility of having a successive fault due to a direct lightning stroke cannot be
disregarded.
UHV / EHV transmission lines carry huge bulk power and, therefore, a failure of HSES interruption due to a
successive fault could have a serious impact (ex. collapse of the stability) on the network.
As a fact, regarding the switching duties considering successive faults, a majority of HSES users adopt a duty that
covers successive faults. A detailed study of successive fault occurrence is needed to support the specification of
HSES on the market.
5.4 Parametric study for TRV requirements with basic model

5.4.1 Basic model
Parametric studies for interrupting duties were conducted by IEC SC17A WG48 experts in collaboration with
CIGRE WG A3.22/28 [51] [52]. The basic model for calculation consists of one transmission line with double
circuits, three-phases in each circuit. HSESs are located at both ends of the line (see Figure 5.4.1.1) Two power
sources are used at each end of the transmission line with different phase angles in order to generate the voltage
differences and the power flow on the line.
The conditions for parametric study are as follows,
- Line length : From 40 km to 300 km
- Power flow: From 8 GW to 10GW (Partly 13 GW for a delayed current zero phenomena)
- The configuration of the tower: UHV tower of Japanese and Chinese projects (see Figure 5.4.1.2)
- Fault conditions: Single-phase, two-phase and three-phase line-to-earth faults (acronym 1LG, 2LG, 3LG) are
considered and the effect of successive fault is studied.
- Short-circuit current: The lumped back impedances and UHV transformers on each side are based on a
short-circuit current of 50 kA - 1LG.
154kV 154kV
HSES 550kV
550kV 1100kV 1100kV
Power flow
Figure 5.4.1.1: Calculation model
Page 108
Japan China
D1 (m) 38.0 36.0
D2 (m) 11.5 16.0
D3 (m) 31.0 28.0
D4 (m) 17.5 20.5
D5 (m) 32.0 35.0
D6 (m) 17.5 20.5
D7 (m) 33.0 31.0
D8 (m) 62.5 30.0

Figure 5.4.1.2: Analysis model of tower (Japan/China)

Japan China
Item Phase conductor Ground conductor Ground conductor
OPGW 500 Phase conductor
ACSR 810 Type A Type B
Diameter D (cm) 3.84 2.94 3.36 2.00 2.11
Thickness T (cm) 1.44 1.15 - - -
Ratio (T/D) 0.375 0.39 - - -
DC resistance at 20deg-C R
0.0356 0.103 0.0463 0.3601 0.494
(ohm/km)
Number of conductor in the bundle 8 1 8 1 1
Bundle Spacing S (cm) 40 - 40 - -
Earth resistivity (ohm-m) 50 260
Table 5.4.1.1: Parameters for model of tower (Japan/China)
5.4.2 Categories of fault and operating sequence for parametric study
Five categories of fault cases are considered depending on the number of faults, with/without successive
faults/delayed current zeros, and single- or multi-phase faults. The five categories introduced in IEC standard
62271-112 are listed in Table 5.4.2.1.
Three major cases discussed during the standardization are as follows. For the simplest case, HSES located at
both ends of the transmission line have to interrupt both electromagnetic and electrostatic induced currents due to
a single-line grounding fault, and is classified as category 0. The first opening operations of the HSES will clear the
electromagnetically induced current and then during the following opening operation an HSES will clear a
electrostatically induced current. For the second case, HSES needs to interrupt the induced currents caused by
multiple phase grounding faults occurring on different lines, and is classified as category 1.
The last case is that of a successive single-phase earth fault occurring during the HSES opening operation on the
phase where the first single-phase earth fault occurred. The successive fault may occur in the same circuit or in the
other circuit located in the vicinity of the circuit with a faulted line.
Category Description
This is a basic case. Only one single-line earth fault occurs within the transmission circuits. For
Category 0 both electromagnetic and electrostatic duties, the currents to be interrupted and recovery voltages
are low. The values of category 0 are covered by those of Category 1.
One single-phase earth fault plus another single-phase earth fault on different circuit without
Category 1 successive fault. This is the case of up to one single-phase earth fault within each circuit in a
double-circuit system.
This is the case that a successive single-phase earth fault occurs during HSES opening operation
Category 2 at the phase where the first single-phase earth fault occurs. Successive fault may occur in the same
circuit or in the other circuit located in the vicinity of the circuit with a faulted line.
This is the case that a single-phase earth faults with delayed current zero phenomena occurs in the
Category 3 presence of a successive single-phase earth fault. During the delayed current zero period HSES
should withstand the stress caused by the arc generated between the contacts of HSES.
Category 4 This is the case that multi-phase faults occur within two or more phase circuits which are located in
the vicinity each other. At least two different phases should be remained without fault condition.
Table 5.4.2.1: Fault modes depending on the HSESs/ CBs operating sequence
Page 109
(a) Basic phenomena
The time difference between the two HSES (at both ends of the transmission line) opening times leads to
differences in current and voltage wave shapes. Major features of each case are as follows,
- When the induced current is interrupted simultaneously by the HSESs at both ends of the transmission line, the
interrupting current is electromagnetically induced and the recovery voltage is electrostatically induced (see Figure
5.4.2.1).
- In the case where the HSESs at both ends of the transmission line break the induced current with a small time
difference, the first HSES interrupts the electromagnetically induced current and a triangular travelling voltage and
current wave appears on the transmission line. Then the recovery voltage for first HSES will also have a triangular
shape. The last HSES interrupts with a recovery voltage that is electrostatically induced (see Figure 5.4.2.2). The
current for the last HSES changes from electromagnetically induced current to a triangular travelling current just
before current zero. Recovery voltage for last HSES to interrupt will be a small triangular wave superimposed on
the electrostatically induced recovery voltage.
Considering the actual operation of HSESs, the time difference between opening times of two HSESs is generally
small enough so that the triangular travelling voltage and current waves can be covered by simple
electromagnetically (EM) and electrostatically (ES) interruption duties.
Figure 5.4.2.3 shows HSES opening operations with interruption of both electromagnetically and electrostatically
induced currents that were calculated.
(a) Voltage wave form
(b) Breaking current on HSES
(c)TRV waveform on HSES
Figure 5.4.2.1: Example of calculated waveforms

(Induced current is interrupted simultaneously)
Page 110
Breaking current on 1st and 2nd breaking HSES
TRV waveform on HSES
Figure 5.4.2.2: Example of calculated waveforms

（induced current is interrupted with small time difference）
Analytical circuit for Category- 0 & 1 Analytical circuit for Category- 2
Electromagnetic (EM) Electrostatic (ES) Electromagnetic (EM) Electrostatic (ES)
coupling coupling coupling coupling
L2-1LG
L2-1LG
L2 L2
L2 L2
Power flow Power flow Power flow

Power flow
L1 L1 L1
L1 A S/S B S/S A S/S B S/S A S/S B S/S
A S/S B S/S
Case 7-9 Case 13-15 Case 19-21
Case 1-3
EM induced ES induced EM induced ES induced

current breaking current breaking current breaking current breaking
L2-1LG L2-1LG
L2 L2 L2
L2
Power flow Power flow Power flow
Power flow
L1 L1 L1
L1 A S/S B S/S A S/S B S/S A S/S B S/S
A S/S B S/S
Case 10-12 Case 16-18 Case 22-24
Case 4-6
EM induced ES induced EM induced ES induced

current breaking current breaking current breaking current breaking
Figure 5.4.2.3: HSES opening operations

(b) Effect of successive fault
Electromagnetically induced current (EM current) in HSES may increase due the existence of a successive fault on
a conductor of the neighbouring transmission line. This tendency is more pronounced if the successive fault occurs
at the beginning or at the end of a transmission line. In case of a successive fault at the centre of the transmission
line, two fault currents I1 and I 2 are generated from both line ends into the successive fault location with an
opposite direction, it follows that the EM current will be smaller (see left hand side of Figure 5.4.2.4).
Page 111
Figure 5.4.2.4: Successive fault location of the transmission line

(Left: at the centre of the line, Right: near the line end)
Figure 5.4.2.5 shows the calculated electromagnetically induced currents if a successive fault occurs at the end of
the neighbouring transmission line. The EM current decreases with increasing line length. This is due to the fact
that the single line earth fault current on the neighbouring transmission line (I2), which increases the EM induced
current, is limited by the impedance of the line itself. So a shorter transmission line generates a higher fault current.
20
EM induced current (kA peak)
18
16
Successive fault occurred on
） 14
1LG on V Phase
12
V phase circuit - 1
10
1LG on W Phase
8
W phase circuit -1
6
4 1LG on U Phase
U phase circuit -1
2
0
0 50 100 150 200 250 300 350
送電線長（k
m）
Line length (km)
Figure 5.4.2.5: A calculation result of interruption by HSES for 1LG

with successive fault on Line #1 at the terminal
5.4.3 HSES parametric studies
Parametric studies are carried out primarily for category 0/1, and 2 as a basic study, since categories 3 and 4 are
thought to be a special duty for HSES. Market need for category 0 is limited and duty for this is covered by that of
category 1. Two conditions are calculated by using EMTP-ATP.
The results are summarized in Tables 5.4.3.1 and 5.4.3.2.
These Tables contains dark square marks representing Class H1’ and Class H2 duties, which were shown in Table
B.3 and Table 1 (proposed duty) of [48] respectively.
a) Categories 0 and 1: duties for 1LG single or double fault without successive fault
These results are summarized in Table 5.4.3.1. The key points from categories 0 and 1 are as follows,
- EM duties increase in proportion to the power flow.
- The line length does not impact the EM breaking current, as the breaking currents are obtained with fixed power
flow. Similarly the rate of rise of the recovery voltage (acronym RRRV) of EM does not depend on the line length
because RRRV is proportional to the product of breaking current × line surge impedance.
- Rated power frequency recovery voltage of EM increases in proportion to the line length.
Page 112
The power flow does have an impact on HSES duties, because there is no power flow at the time of HSES
breaking.
- HSES breaking current, which is due to line capacitance, increases in proportion to the line length.
- Recovery voltage of HSES is independent of line length.
Breaking current Rate of the rise of the recovery voltage Rated power frequency recovery
(A rms) [RRRV] (kV peak/s) voltage (kV rms) or (kV peak)
Electro-magnetically (EM)
300
1200 0.22
波高値[kV](kVrms)
0.2 250
class H1 class H1
current (A rms)
1000
0.18
class H1
[rrrV] (kV peak/us)
0.16 200
上昇率[kV/μA]
800
Breaking 遮断電流[A]
Rated power frequency

0.14
0.12 150
600
0.1
0.08 100
400
0.06 　　　　　13GW
　　　　　13GW 　　　　　13GW
200 　　　　　10GW 0.04 50 　　　　　10GW
　　　　　10GW
　　　　　 8GW 0.02 　　　　　 8GW
　　　　　 8GW
0 0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Line length (km)
線路長[km] Line線路長[km]
length (km) 線路長[km]
Line length (km)
250
250
Electro-statically (ES)
class H1 class H1
Rated power frequency (kVp)

Breaking current (A rms)
200
200
遮断電流[A]
150
波高値[kV]
150
100 100
50 　　　　　13GW 　　　　　13GW
50
　　　　　10GW 　　　　　10GW
　　　　　 8GW 　　　　　 8GW
0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Line length (km)
線路長[km] Line線路長[km]
length (km)
Table 5.4.3.1: Calculation results of HSES parametric study [Categories 0 & 1]
b) Category 2: duties for 1LG with 1LG successive fault
These results are summarized in Table 5.4.3.2. The key points from category 2 are as follows,
- The power flow does not impact HSES duties since the successive 1LG fault current dominates the induced
currents from the first line fault.
- The EM breaking current and the RRRV decreases with line length as it is proportional to the 1LG fault current
which decreases when line impedance increases. In reverse, the EM recovery voltage increases when the
transmission line is longer.
- HSES breaking current, which is due to line capacitance, increases in proportion to the line length.
- Recovery voltage of ES is independent of the line length.
Page 113
Rate of the rise of the recovery voltage Rated power frequency recovery
[RRRV] (kV peak/s) voltage (kV rms) or (kV peak)
8000
Electromagnetically (EM)
1 700
class H2 class H2 class H2
0.9
(kVrms)
7000
600
0.8
6000
500
[rrrV] (kV peak/us)

0.7
上昇率[kV/μs]
遮断電流[A]
実効値[kV]
5000
0.6
Rated power frequency

400
4000 0.5
0.4 300
3000
0.3
2000 200
0.2
　　　　　13GW 　　　　　13GW 100 　　　　　13GW
1000 0.1
　　　　　10GW 　　　　　10GW 　　　　　10GW
0 0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Line length (km)
線路長[km] Line length (km)
線路長[km] Line length (km)
線路長[km]
300
250
class H2
Electrostatically (ES)
Rated power frequency (kVp)

250
class H2
)
(A rms)
200
遮断電流[A]
200
(kV
波高値[kV]
150
Breaking current
150
lt
100 100
50 50
　　　　　13GW
　　　　　13GW
　　　　　10GW
0 　　　　　10GW
0 50 100 150 200 250 300 350 0
0 50 100 150 200 250 300 350
Line length (km)
線路長[km]
Line線路長[km]
length (km)
Table 5.4.3.2: Calculation results of HSES parametric study [Category 2]
5.4.4 Results of parametric studies with the basic model
Based on the parametric study calculations, the more severe values for 1100-1200 kV HSES categories 1 and 2
are listed in Table 5.4.4.1. The proposed values from IEC committee draft 17A/969/CD [48] are also listed in Table
5.4.4.1 for reference.
Comparison Electromagnetic coupling Electrostatic coupling

between calculated Rated induced Rated power Time to first Rated induced Rated induced
First TRV peak
value and proposed current frequency recovery peak current voltage
(kV)
value (A rms) voltage (kV rms) (ms) (A rms) (kV rms)
Calculated values 744 1,92
76 189 228 196
for category 1 (1) (832) (3) (1.0) (4)
Values in IEC draft
standard for 830 80 200 1,0 230 200
category 1 [48]
Calculated values
6832 232 576 0,6 177 235
for category 2 (2)
Values in IEC draft
standard for 6800 240 580 0,6 230 235
category 2 [48]
Table 5.4.4.1: Calculated values and proposed values for 1100 – 1200 kV HSES
categories 1 and 2
Notes
(1) Up to 300km transmission line is considered for calculation in category 1
(2) Up to 200km transmission line is considered for calculation in category 2.
(3) Calculated value in parentheses is obtained by China tower model.
(4) Calculated value in parentheses is obtained by China tower model and also duty in BPA 800 kV HSES specification
(5) Values in IEC committee draft for international standard [48]
Page 114
5.5 HSES switching duties based on model network

Using a model network, further studies on HSES switching duties were carried out by CIGRE WG A3.28 to check
the following aspects [61] [62]:
1) Effect of system voltage, line length, power flow, tower configuration
2) HSES duties for different classifications proposed by IEC
The results were compared with results obtained with the basic model considering some differences in simulation
conditions.
5.5.1 Model network
TRV requirements for HSES were investigated with the parallel single circuit or double circuit radial network models
shown in Figure 5.5.1.1 (without MOSA) using different system and equipment parameters at rated voltages of
800 kV and 1100 kV.
50 kA
Tr x 2
D-s/s
360 km
120 km 240 km
A-s/s B-s/s C-s/s

50 kA 50 kA 50 kA
Tr x 2 Tr x 2 Tr x 2
Figure 5.5.1.1: Parallel single circuit or double circuit radial network models
Figures 5.5.1.2 and 5.5.1.3 show the tower configurations used for the simulations. Four kinds of tower
configurations were used: Japanese 1100 kV, Chinese 1100 kV, Canadian 800 kV, and Korean 800 kV. Canadian
800 kV tower consists of the two-single line towers of Hydro Quebec and others are double circuit towers of
TEPCO, SGCC, and KEPCO
Conditions for the study are as follows,
- Line length : A s/s to B s/s 120 km, B s/s to C s/s 240 km, B s/s to D s/s 360 km
- Power flow: 6 GW for TEPCO and SGCC, 4 GW for KEPCO, 3 GW for HQ
- Fault conditions: single-phase, two-phase line-to-earth faults are considered and effect of a successive fault
is studied
- Short-circuit current: 50 kA at the low voltage side of transformer, with two transformers connected at each
1100/800 kV substation
Page 115
18m 18m
19.0m 19.0m
14m 14m
15.5m 15.5m
14.8m 14.8m
16.0m 16.0m
109m
98 m
15.5m 15.5m
97.5m
16.5m 16.5m
81.5 m
80.0m
61.7 m
62.5m
42 m
Earth Resistivity = 50 ohm-m Earth Resistivity = 500 ohm-m
1100kV double circuit tower of Japan 1100kV double circuit tower of China
Figure 5.5.1.2: Tower configuration (for 1100 kV)
15.91 m 15.91 m
16.0 m
13.84 m 13.84 m
14.44 m 14.44m
99.12 m
20.1m 20.1m 20.1m 20.1m

76.79 m
14.94 m 14.94 m
12m 12m 12m 12m
58.29 m
13.41 m
75m
39.78 m
22.8m
35.0m
Earth Resistivity = 1000 Ohm-m
Earth Resistivity = 100 Ohm-m
800 kV double circuit tower of Korea 765 kV single circuit tower of Canada
Figure 5.5.1.3: Tower configuration (for 800 kV)
Figure 5.5.1.4 shows typical TRV waveforms obtained when HSES interrupts both electromagnetically and
electrostatically induced currents caused by a single-line to ground fault as well as multi-phase to ground faults with
an additional successive grounded fault on the faulted line using 1100 kV system parameters in China and Japan.
Electromagnetic (EM) Electrostatic (ES)

TRV
Uc=128kV, dv/dt=0.088kV/us
TRV (kV)
TRV (kV)
TRV Uc=559kV
Current (kA)
Current (kA)
Breaking current=610A rms Breaking current=178A rms
Figure 5.5.1.4: Typical waveform obtained for 1100 kV system
Page 116
5.5.2 Results for 1100 kV HSES
Figures 5.5.2.1 and 5.5.2.2 show the simulation results obtained with TEPCO (Japanese) and SGCC (Chinese)
models. Green line and dot in the figures indicate the IEC standardized values.
1) Categories 0 and 1
Line lengths of the model network calculations are 120 km and 240 km, while it was 300 km in the basic study
described in 5.4. The line length does not have a strong effect on EM breaking current, and the rated power
frequency recovery voltage of EM increases in proportion to line length. These tendencies are the same as in the
basic study.
For electromagnetically induced current breaking for categories 0 and 1, the study in model networks is conducted
with a 6 GW power flow, while it was 8 to 10 GW in the basic study. EM induced current is proportional to the
current flowing in healthy phases and therefore to power flow. The electromagnetically induced currents are less
than 830 A and the recovery voltage increases with line length (it is less than 200 kV). The IEC standardized
values are a little bit higher than those obtained by model network calculations.
Breaking current for 1100kV of Recovery voltage for 1100kV of IEC draft
IEC draft （8GW-10GW）（8GW-10GW）
1200 300
TEPCO 1100kV - 6GW TEPCO 1100kV - 6GW
Recovery voltage peak (kV)
1000 China 1100kV - 6GW 250 China 1100kV - 6GW

EM Induced current (A)
800 200
600 150
400 100
200 50
0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
(a) Electromagnetically (EM) induced current for categories 0 & 1
Breaking current for 1100kV

of IEC draft
250 250
Recovery voltage for
1100kV of IEC draft
Recovery voltage rms (kV)
200 200
ES Induced current (A)
150 150
100 100
50 TEPCO 1100kV - 6GW TEPCO 1100kV - 6GW

50
China 1100kV - 6GW China 1100kV - 6GW
0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Line length (km)
Line length (km)
(b) Electrostatically (ES) induced current for categories 0 & 1
Figure 5.5.2.1: Induced current for TEPCO and China 1100 kV model for category 0 & 1
Electrostatically induced current which is caused by line capacitance increases in proportion to line length (less
than 200 A) and the corresponding recovery voltage is less than 200 kV and independent of line length, as
obtained in the basic study. These values are covered by the IEC standardized values for 1100 kV.
Page 117
2) Category 2
The electromagnetically induced current and corresponding RRRV decrease with an increase of line length (less
than 4000 A) since the 1LG fault current depends on total impedances at the main bus bar connected to HSESs
which decreases when line impedance increases. Short circuit current in the model network calculations was 50 kA
at the 500 kV busbar side with two transformers connected between the 1100 kV and 500 kV busbars. This set up
yields around 16 kA to 20 kA for actual fault current. On the other hand, short circuit current in the basic study was
set as 50 kA at the 1100 kV busbar. Considering these differences in conditions, results with the model network are
well in agreement with that of the basic study. The electromagnetically induced currents decrease with an increase
of line length (less than 4000 A) and the recovery voltages increase with an increase of line length (less than
600 kV).
Electrostatically induced current which is caused by line capacitance increases in proportion to the line length.
Recovery voltage of HSES is independent from line length. The calculated values are also covered by the IEC
standard values for 1100 kV HSES.
Breaking current for 1100kV of IEC draft Recovery voltage for 1100kV of IEC draft
（Isc at Bus =50kA）（Isc at Bus =50kA）
8000 700
7000 600
Recovery voltage peak (kV)
EM Induced current (A)
6000 TEPCO 1100kV - 6GW 500

5000 China 1100kV - 6GW
400
4000
300
3000
200
TEPCO 1100kV - 6GW
2000 China 1100kV - 6GW
100
1000
0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Line length (km)
Line length (km)
(a) Electromagnetically (EM) induced current for category 2
Recovery voltage for 1100kV of IEC draft

Breaking current for 1100kV of IEC draft
300 300
250 250
Recovery voltage rms (kV)
ES Induced current (A)
200 200
150 150
100 100
TEPCO 1100kV - 6GW TEPCO 1100kV - 6GW
50 China 1100kV - 6GW 50 China 1100kV - 6GW
0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
(b) Electrostatically (ES) induced current for category 2
Figure 5.5.2.2: Induced current for TEPCO and China 1100 kV model for category 2
5.5.3 Results for 800 kV HSES
Figures 5.5.3.1 and 5.5.3.2 show the simulation results for KEPCO (Korean) and Hydro Quebec (Canada) models.
Again, the green line and dot indicate the values proposed by the IEC.
Page 118
1) Categories 0 & 1
Similar to the 1100 kV cases, line length in the model networks are 120 and 240 km and values obtained with a
300 km line length were used in the IEC SC17A WG48 study.
For electromagnetically (EM) induced current, study in model networks are conducted for 3 or 4 GW power flow
compared to 8 to 10 GW in the basic study. EM induced current is proportional to the flowing current in healthy
phases and then to power flow. The line length does not have a strong effect on EM induced current. The
electromagnetically induced currents are well less than 830 A and their recovery voltages increase with an increase
of line length (less than 200 kV).
Electrostatically induced current is proportional to the line length (less than 170 A) and the corresponding recovery
voltages show a constant value less than 140 kV. These requirements are covered by the 800 kV IEC standardized
values with margin.
Breaking current for 800kV Recovery voltage for 800kV of IEC

of IEC draft (8GW-10GW) draft (8GW-10GW)
1200 300
Korea 800kV - 4GW Korea 800kV - 4GW
(kV)
(A)(A)
250 Canada 800kV - 3GW

(kV)
peak
current
peak
200
current
800
voltage
voltage
induced
600 150
Induced
Recovery
400 100
Recovery
EM
50
EM
200
0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Line
Line length
length (km)
(km) Line
Line length
length (km)
(km)
a) Electromagnetically (EM) Induced current for categories 0 & 1
Breaking current for 800kV of IEC draft

Recovery voltage for 800kV of IEC draft
250 250
rms (kV)
(A) (A)
200 Canada 800kV - 3GW Canada 800kV - 3GW

200
peak (kV)
current
current
voltage
150 150
induced
voltage
100 100
Induced
Recovery
Recovery
ESES
50 50
0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Line
Line length
length (km)(km) Line
Line length (km)
length (km)
(b) Electrostatically (ES) Induced current for categories 0 & 1
Figure 5.5.3.1: Induced current for KEPCO and HQ 800 kV model for categories 0 & 1
2) Category 2
The EM induced current decreases with an increase of line length since the fault current depends on the total
impedances at the main bus bar connected to HSESs which consist of the transmission line impedance or short-
circuit impedance etc. Short circuit current in the model network calculations was set to 50 kA at the 345 kV busbar
Page 119
side and two transformers are connected between the 800 kV and 345 kV busabars. This set up yields around 11
to 20 kA in the model network. On the other hand, proposed duties in the study by IEC SC17A were based on a
50 kA short circuit current at the 800 kV busbar. Considering these differences in conditions, results of the model
network calculations are in agreement with that of the basic study.
The recovery voltages increase with an increase of line length.
Electrostatically induced current which is caused by line capacitance increases in proportion to the line length (less
than 170 A). Recovery voltage electrostatically induced is independent of the line length and less than 170 kV.
These values are also covered by the 800 kV IEC standard values.
Breaking current for 800kV Recovery voltage for 800kV of IEC
of IEC draft (8GW-10GW) draft (8GW-10GW)
1200 300
(kV)
(A)(A)
(kV)
peak
current
peak
200
current
800
voltage
voltage
induced
600 150
Induced
Recovery
400 100
Recovery
EM
50
EM
200
0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Line
Line length
length (km)
(km) Line
Line length
length (km)
(km)
(a) Electromagnetically (EM) induced current for category 2
Breaking current for 800kV of IEC draft Recovery voltage for 800kV of IEC draft
300 250
(kV)
peak(kV)
Canada 800kV - 3GW

(A)
250 Canada 800kV - 3GW 200

current(A)
voltage rms
inducedcurrent
200
150
Recoveryvoltage
150
ES Induced
100
Recovery
100
50
ES
50
0 0
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350
Line length (km)(km)
LineLine length
length (km) (km) Line length
(b) Electrostatically (ES) induced current for category 2
Figure 5.5.3.2: Induced current for KEPCO and HQ 800 kV model for category 2
Figure 5.5.3.3 gives the comparison of secondary fault currents in the case of single-phase grounding faults as well
as with successive fault for different tower configurations. The electromagnetically induced currents are produced
by multi-phase grounding faults with an additional successive grounding fault on the faulted line (lowest line) using
different tower designs. The larger fault current tends to generate a larger secondary induced current.
Page 120
1LG (successive fault) : Fault phase (Current based on the 1LG is flowing)
A s/s B s/s : Healthy phase (Current based on the power flow is flowing)
L2 (120km) : HSES closed phase (EM induced current is flowing)
Fault current Power flow : 6GW (China),

CB CB 4GW (Korea), Chinese 1100kV system
3GW (Canada)
L1 (120km) (double circuit tower)
Only one phase HSES HSES Only one phase
is opened is opened Korean 800kV system L1 L2
Closed position EM induced

(double circuit tower) 6.0kA
current breaking A C
L1 L2
4.0kA
Canadian 800kV system B B
A C
39.5m
(single circuit tower)
19.7m
5.2kA
37.0m
4.0kA B B
4.0kA C 31.0m A
18.5m
63m 3.6kA
L1 12m L2 C 29.9m A
42.0m
C 3.7kA
A B C B A
26.4m
22.8m 2.3kA 1LG (15.1kA)
1LG (20.1kA)
3.0kA 1LG (12.2kA) Successive fault
Successive fault Successive fault
EM breaking duties of HSES EM breaking duties of HSES EM breaking duties of HSES

I=3.0kA, Uc=387kV, 3rV=0.53kV/s I=2.3kA, Uc=307kV, 3rV=0.38kV/s I=3.7kA, Uc=394kV, 3rV=0.49kV/s
( Z= dv/dt / di/dt = 331ohm ) ( Z= dv/dt / di/dt = 310ohm ) ( Z= dv/dt / di/dt = 298ohm )
R-11 network with China 1100kV parameters Japan 1100kV system（Simplified model）
Power flow : 6GW , Line length : 120km Power flow : 10GW , Line length : 100km
L1 L2
L1 L2 8.7kA
5.6kA
6.0kA A C
3.7kA
A C
1.7kA 1.6kA
B B
1.2kA 1.2kA
35.0m
B B
39.5m
17.5m
7.6kA
19.7m
5.2kA C 33.0m A
C 31.0m A
62.5m
4.4kA
42.0m
3.7kA
1LG (20.1kA) 1LG：45.7kA
Successive fault Successive fault
EM breaking duties of HSES EM breaking duties of HSES

I=3.7kA, Uc=394kV, 3rV=0.49kV/s I=4.4kA, Uc=382kV, 3rV=0.57kV/s
(Z= dv/dt / di/dt = 298ohm) (Z= dv/dt / di/dt = 292ohm)
Figure 5.5.3.3: Induced currents when interrupting both electromagnetically induced currents caused by a
single-line grounding fault as well as multi-phase grounding faults with an additional successive
grounding fault on the faulted line using different tower designs
5.6 Recommendation for specifications

Based on the parametric study with the basic model and model network, switching duties for HSES are
investigated. The results are in line with the duties proposed by IEC SC 17A. Considering present market needs,
CIGRE WG A3.28 recommends a simplified single set of duties for categories 0, 1, and 2 as shown in Table 5.6.1.
As for the duties for categories 3 and 4, CIGRE WG A3.28 did not conduct further calculations than that of IEC SC
17A WG48. The market for these duties, which require the considerations of the occurrence of delayed current
zero and a multi-phase reclosing scheme, are thought to be limited. Specifications for categories 3 and 4 should be
informative only at this moment.
Page 121
Electromagnetic coupling Electrostatic coupling

Rated Rated Power
voltage Rated induced
Rated induced frequency First TRV peal Time to peak Rated induced
voltage
(kV) current (Arms) recovery voltage (kV) (ms) current (Arms)
(kVrms)
(kVrms)
550 6800 240 580 0.6 120 115
800 6800 240 580 0.6 170 170
1100-1200 6800 240 580 0.6 230 235
Table 5.6.1: Switching duties for HSES
5.7 Conclusions
HSES (High Speed Earthing Switch) is a kind of earthing switch used for secondary arc extinction of overhead line
at higher voltage systems such as UHV/EHV. In some references HSES is called HSGS (High Speed Grounding
Switches). Besides four-legged reactors, utilities can choose HSES as a countermeasure for fast secondary arc
extinction on overhead transmission lines in order to assure reclosing within desired times under various fault
conditions.
The possibility of having a successive fault and its effect on the interrupting duties of HSES were investigated.
Multiple lightning strokes within one second as well as successive direct lightning strikes were observed along a
UHV designed overhead line. The possibility of having a successive fault due to a direct lightning stroke cannot be
disregarded. UHV/EHV transmission lines carry huge bulk power and, therefore, an interruption failure of HSES by
a successive fault may cause serious impact on the network, such as collapse of its stability. Taking this factor into
account, a majority of HSES users specify a duty that covers successive faults; therefore this condition was
considered in the study.
Switching duties were studied with a basic model and a model network. Simulations were performed taking into
account several switching conditions. These investigation results by CIGRE WG A3.28 support the requirements
introduced by IEC SC17A in HSES standard IEC 62271-112.
Page 122
6 Disconnector and earthing switch

6.1 Introduction
In this chapter attention is given to switching of bus-transfer and bus-charging currents by disconnector switches
(DS), and to switching of electromagnetically and electrostatically induced currents by earthing switches (ES).
Previous work done by CIGRE WG A3.22 provided data to support the revision of IEC 62271-102 by IEC SC17A to
cover rated voltages higher than 800 kV [2] [12]. These new studies give additional input for UHV and provide data
on the switching duties of DS and ES in EHV applications that will support a coming revision of the IEC standard. A
particular attention is given to the bus-transfer current that is, in edition 1.2 of IEC 62271-102, equal to 80% of the
rated normal current with a limitation to 1600 A for rated voltages up to and including 800 kV and without limitation
for rated voltages higher than 800 kV.
A phenomenon in relation with disconnectors in GIS is that of very fast transient overvoltages (VFTO), that are
caused by the travelling waves in short bus sections inside a GIS. This topic has been addressed by several WGs,
but most extensively by WG C4.306 in a Technical Brochure on insulation coordination for UHV [22]. In Technical
Brochure 519 [75] a method to assess the insulation coordination with respect to VFTO has been published by WG
D1.03, a joint task force between WG A3.22/28, B3.29, C4.306 and D1.36. With reference to Technical Brochure
519, the method on how to evaluate certain countermeasures is shortly addressed in this chapter.
The electromagnetically and electrostatically induced currents to be switched by earthing switches strongly depend
on the tower configuration of the OHL, and the load current in the circuit in operation. Besides, the induced voltage
(i.e. the recovery voltage) depends also on the line length. The rate of rise of the recovery voltage is determined by
travelling waves and is proportional to the induced current. Many different combinations of earthed phases, at one
side or at both sides, and tower configurations have been simulated for rated voltages of 800 kV and above.
6.2 Bus transfer switching by disconnector switches

6.2.1 Introduction
In this chapter, calculations of bus-transfer voltage and current are presented for disconnectors capable of bus-
transfer switching for substation layouts with GIS or HGIS (hybrid switchgear or compact switchgear assemblies).
The calculations are based on loop lengths calculated for the worst case transfer scenario (as per replies to WG
questionnaire to participating utilities) for rated as well as IEC specified maximum bus-transfer currents.
With multiple parallel busbars in the switchgears, the need for a method to direct the current between different
feeding bays arises. Disconnector switches can be used for bus-transfer switching since it greatly reduces the size
and cost of GIS as compared to circuit-breaker use. The voltage drop over open disconnector switch is typically
very small following the breaking of bus-transfer currents as compared to isolation of switchgear sections [2].
While the resistance and inductance per length in GIS is quite low, the considerable differences in length which can
occur in large GIS or Hybrid IS (or Compact Switchgear Assemblies according to IEC) can lead to difficult switching
operations. IEC 62271-102 (Annex B) specifies the maximum bus-transfer current which a disconnector switch with
bus-transfer capability should be able to break: 80 % of its rated current up to a maximum of 1600 A [2]. The
highest bus-transfer voltage which is allowed to appear in steady-state at the contacts before making or after
breaking of bus-transfer currents is specified in Table B.1 of IEC 62271-102 (Annex B) and is called bus-transfer
voltage. The resistance R and the inductance L are determined by the geometry and layout of the relevant
conductors. The bus-transfer current I2 is calculated by the equation below:
Where I2 Bus-transfer current

Itot Total (rated) current
d1 Length of loop 1 (see Figure 6.1.2.1)
Page 123
d2 Length of loop 2 (see Figure 6.1.2.1)

Z = R  (L) 2
2
The power frequency voltage appearing across the terminals of the disconnector can be calculated by equation
Where,
U Bus transfer voltage
Loop length
R’ Resistance per unit length
L’ Inductance per unit length
Figure 6.2.1.1 shows an example of bus-transfer switching operation (close DS 2 open DS 1). Following the
closing of DS 2, the load current is distributed into two parallel paths in accordance with the impedance ratio of
both paths. The current from the transformer bay is branched into two directions: one part is going through DS 1 (I1-
1) and the other part is through DS 2 and the bus-coupler bay (I1-2).
Figure 6.2.1.2 shows an example with currents from two different bays (transformer bay and No.1 feeder bay) that
are flowing to No. 2 feeder bay through a loop circuit. In this case the bus-transfer current is the sum of currents
from the transformer bay and the No. 1 feeder bay.
In a double bus-bar system, the bus transfer current generally becomes the sum of distributed currents from
different bay as follows:
Bus transfer current = Σ (distributed current from each bay)
Therefore, it should be noted that in a double bus-bar system the bus transfer current can be higher than the load
current from one bay.
Figure 6.2.1.1: Bus-transfer current distribution (example 1)
In Figure 6.2.1.2 the disconnector under consideration is represented by a black dot, the other disconnectors are
represented either by a cross or an open circle for open respectively closed position. As can be learned from this
figure the maximum current through this disconnector under bus transfer conditions may become larger than the
current of an individual bay.
Page 124
Considering these phenomena, Japanese practice for 550 kV and UHV substations is that the rated current for the
bus disconnectors of the transformer bay is specified as 4000 A although the rated current through the transformer
bay itself (circuit breaker, transformer DS) is 2000 A (ex. In case of 550 kV transformer bank; 1500 MVA/√3*500 kV
=1730 A.).
Figure 6.2.1.2: Bus-transfer current distribution (example 2)
6.2.2 Bus-transfer switching by GIS disconnector switches

A typical switchgear layout is used in the following. The GIS ratings considered are as follows: Urated = 550 kV, Irated
= 6000 A and frated = 50 Hz.
The whole GIS is considered to be with only busbars and no other components, therefore the connecting lengths
are not considered. The corresponding loop lengths are as per Table 6.2.2.1.
A pictorial representation of the loop is given in Figures 6.2.2.1 and 6.2.2.2.
Figure 6.2.2.1: Bus transfer current loops in GIS – Cases 1 & 3
Page 125
Figure 6.2.2.2: Bus transfer current loops in GIS – Cases 2 & 4
Case 2 is for a loop with worst case lengths. A reactance of 0.2 µH/m and 0.3 µH/m are typical values respectively
for 550 kV and 1100 kV. As seen from the table, the bus-transfer currents highlighted in Italics exceed the
allowable current limit of 80% of the rated current or 1600 A maximum (except for UHV levels of 1100/1200 kV as
per the new IEC draft), even though the bus-transfer voltages are lower compared to the allowable IEC values for
550 kV GIS.
Scenarios Length Length IBT-loop1 IEC values

Imax (A) Zavg (µΩ/m) VBT-loop2 (V)
(Japan) Loop1 (m) Loop2 (m) (A) (Vrms)
Case 1 6000 64.13 11 33 4500 12.70 40
Case 2 6000 64.13 5.5 88.6 5649 34.09 40
Case 3 2133 64.13 11 33 1600 4.51 40
Case 4 1699 64.13 5.5 88.6 1600 9.66 40
Table 6.2.2.1: Calculation for the bus transfer voltages and currents in GIS
Note: in the second column 6000 A is the rated current of the disconnector
If the bus-transfer currents are to be limited to 1600 A as per IEC values, case 4 becomes a worst case and the
maximum current is to be limited to 1699 A, in this case the bus-transfer voltage is much lower than the IEC value.
In case the maximum current of 6000 A is considered for a bus-transfer switching, the bus-transfer voltage reaches
a value of 34.09 V (within IEC permissible limits). For case 2, if the IEC bus-transfer voltage levels are to be
reached, the maximum allowable loop length would be equal to 104 m. Similarly, for case 4, the maximum loop
lengths can be as long as 367 m.
6.2.3 Bus-transfer switching by hybrid switchgear

A typical hybrid switchgear layout is selected with the following ratings : Urated = 420 kV, Irated = 4000 A, Frated =
50 Hz. The corresponding loop lengths are as per Table 6.2.3.1. A pictorial representation of the loop is given by
Figures 6.2.3.1 and 6.2.3.2.
Page 126
OHL OHL
BB 1
BB 2
BB 3
Loop 1
Loop 2
CB
DS
OHL OHL
Figure 6.2.3.1: Bus transfer current loops in Hybrid Switchgear – Cases 1 & 3
OHL OHL
BB 1
BB 2
BB 3
Loop 1
Loop 2
CB
DS
OHL OHL
Figure 6.2.3.2: Bus transfer current loops in Hybrid Switchgear – Cases 2 & 4
Similar to GIS, case 2 here is for a loop with worst case transfer lengths. As seen from Table 6.2.3.1 the bus-
transfer currents highlighted in Italics exceed the allowable IEC values, even though the bus-transfer voltages are
lower compared to the allowable IEC values for 420 kV AIS.
Scenarios Zavg Length Length IBT-loop1 IEC values

Irated (A) VBT-loop2 (V)
(Germany) (µΩ/m) Loop1 (m) Loop2 (m) (A) (Vrms)
Case 1 4000 250.00 18.5 120.68 3468 120.68 300
Case 2 4000 250.00 18.5 194.68 3653 194.68 300
Case 3 1845 250.00 18.5 120.68 1600 55.67 300
Case 4 1752 250.00 18.5 194.68 1600 106.62 300
Table 6.2.3.1: Calculation for the bus transfer voltages and currents in HGIS
Note: in the second column 4000 A is the rated current of the disconnector
If the transfer currents are to be limited to 1600 A, case 4 becomes a worst case and the maximum current is to be
limited to 1752 A , the bus transfer voltages are much lower than the IEC value. In case the rated current of 4000 A
is considered for a bus-transfer switching, the bus-transfer voltage reaches a value of 213.18 V (within IEC
Page 127
permissible limits). For case 2, if the IEC bus-transfer voltage levels are to be reached, the maximum allowable
loop length would be equal to 300 m. Similarly, for case 4, the maximum loop lengths can be as long as 685 m.
6.2.4 Field data for 500 kV and 275 kV systems in Japan

Field data were collected for 500 kV and 275 kV systems in Japan (see also Chapter 7). The bay current data are
derived for eleven 550 kV substations (60 bays, 40 line bays and 20 transformer bays) and twenty-three 275 kV
substations (167 bays, 93 line bays and 74 transformer bays). Figures 6.2.4.1 and 6.2.4.2 show the cumulative
distribution of the ratio of the maximum load current to the rated current of the bay. A tabular representation is
given in Table 6.2.4.1. It should be noted that since the collected data were obtained on the third wednesday of
each month, only between 2008 and 2009; they may not represent the real maximum current ratio of the year.
550 kV(feeder bay） (bank bay）

Ratio Ratio
100.0% 100.0%
90.0% 90.0%
80.0% 80.0%
70.0% 70.0%
60.0% 60.0%
50.0% 50.0%
40.0% 40.0%
30.0% 30.0%
20.0% 20.0%
10.0% 10.0%
0.0% 0.0%
0 20 40 60 80 100 0.0 20.0 40.0 60.0 80.0 100.0
Cumulative distribution (%) Cumulative distribution (%)

Figure 6.2.4.1: Cumulative distribution of the ratio of the maximum load current to the rated
current of the bay (550 kV)
275 kV(feeder bay） (bank bay）
Ratio Ratio
100.0% 100.0%
90.0% 90.0%
80.0% 80.0%
70.0% 70.0%
60.0% 60.0%
50.0% 50.0%
40.0% 40.0%
30.0% 30.0%
20.0% 20.0%
10.0% 10.0%
0.0% 0.0%
0 20 40 60 80 100 0 20 40 60 80 100
Cumulative distribution (%) Cumulative distribution (%)

Figure 6.2.4.2: Cumulative distribution of the ratio of the maximum load current to the rated
current of the bay (275 kV)
Ratio of maximum load current to rated current
50% cumulative 90% cumulative 100% cumulative
Level kV Current (A) Source Path
distribution distribution distribution
550 8000, 6000, 4000 Field data feeder bay 20.7 36 55
2000 Field data bank bay 36.7 68 82.5
Field data total 25.5 52 82.5
275 8000, 4000, 2000 Field data feeder bay 28.8 38 52
4000, 2000 Field data bank bay 22.6 34 65.6
Field data total 26 48 65.6
Selected values (worst case) 26 52 82.5
Table 6.2.4.1: Tabular representation of probabilistic current distribution
Page 128
Based on the collected data a worst case assumption, using the 90 % probability value of occurrence, gives a load
current close to 50 % of the rated current.
As explained in 6.2.1, the bus-transfer current is the sum of distributed currents from different bays, then it is larger
than that of current from one bay. Some utilities require a bus-transfer current up to or near the rated current to
avoid any constraint in operation. For UHV and EHV class, normally 3 to 4 feeders and/or bank bays are
connected to the each divided bus-bar (such as 1/2 bus for double bus-bar - 2 bus sections system), and thus the
bus-transfer current consists of 2 to 3 bay currents. A bus-transfer current corresponding to 80% of the rated
current is suggested taking into account the 90% probability of occurrence of load current, the number of bays that
supply the bus-transfer current and the distribution ratio in the bus-transfer current.
6.2.5 Bus-transfer voltage

A survey was made on GIS and AIS layouts, the relevant replies for bus-transfer switching have been considered
for the following calculations.
GIS configurations:
Substation Zavg (µΩ/m) Length Loop1 (m) Length Loop2 (m) IBTL1 (A) VBTL2 (V)
Japan (550kV, 6000A) GIS 65 6 500 1600 53
Japan (550kV, 6000A) GIS 64 6 300 6000 118
EGAT (550kV, 4000A)GIS 70 15 500 1600 58
EGAT (550kV, 4000A) GIS 67 15 350 4000 98
Japan (1100kV, 8000A) GIS 104 11 150 8000 134
China (1100kV, 8000A) GIS 95 36 150 8000 141
Table 6.2.5.1: Calculation for the maximum loop lengths in GIS

As can be seen from Figure 6.2.5.1 loop length increases with the increase in allowable bus-transfer voltage but is
limited to a maximum value by the allowable IEC limits (Black color) for a particular voltage level. The allowable
loop length decreases with the increase in the bus-transfer current and with the increase in the circuit parameters
namely R and L.
Figure 6.2.5.1: Characteristic loop lengths comparison for different GIS ratings
Page 129
AIS configurations:
Substation Zavg (µΩ/m) Length Loop1 (m) Length Loop2 (m) IBTL1 (A) VBTL2 (V)
EGAT (550kV, 4000A) AIS 314 56 650 1600 355
Eskom (400kV, 3150A) AIS 315 50 650 1600 352
Eskom (765kV, 3150A) AIS 346 69 650 1600 398
Eletrobras (550kV, 3150A) AIS 377 69 650 1600 434
Eletrobras (800kV, 3150A) AIS 452 69 650 1600 520
Hydro Quebec (735kV, 3150A) AIS 452 69 500 1600 412
Japan (1100kV, 8000A) Hybrid IS 270 69 285 8000 765
China (1100kV, 8000A) Hybrid IS 377 70 200 8000 811
Table 6.2.5.2: Sample calculations for the maximum loop lengths for HGIS
As can be seen from Figure 6.2.5.2 loop length increases with the increase in allowable bus-transfer voltage but is
limited to a maximum by the allowable IEC limits for a particular voltage level. The allowable loop length decreases
with the increase in the bus-transfer current and with the increase in the circuit parameters namely R and L.
Figure 6.2.5.2: Characteristic loop lengths comparison for different AIS/HGIS ratings
6.2.6 UI characteristic based on the relation between maximum load current and rated bus-
transfer current
As seen in 6.2.4, a worst case assumption using the 90 % probability value of occurrence the load current is about
50 % of the rated current. The 52 % ratio of maximum load current to rated current (see Table 6.2.4.1) and 130 m
and 400 m as the maximum loop lengths for GIS and AIS respectively are used for IEC comparison and the result
is as indicated in the Figures 6.2.6.1 and 6.2.6.2. A sample calculation is as per Table 6.2.6.1 considering a case of
rated currents available per rated switchgear voltage designs.
Urated (kV) Irated (A) IBT (A) 52% Zavg (Ω/m) UBT-GIS (V) Zavg (Ω/m) UBT-AIS (V)
245/300 3150 1638 6.30E-05 13.42 3.14E-04 205.73
420 3150 1638 6.70E-05 14.27 3.46E-04 226.70
550 3150 1638 6.70E-05 14.27 3.61E-04 236.53
800 6300 5040 7.60E-05 32.37 3.55E-04 465.19
1100 6300 5040 8.00E-05 34.07 3.88E-04 508.44
Table 6.2.6.1: Sample calculations for U BT for GIS & AIS respectively
Page 130
GIS
(52% of rated busbar current)
IEC 3150 A 4000 A 5000 A 6300 A 8000 A
140
120
Bus transfer voltage (U)
100
80
60
40
20
0
245/3001kV 2
420kV 3
550kV 4
800kV 5
1100kV
Figure 6.2.6.1: UI characteristics for bus transfer in GIS

AIS
IEC 3150 A 4000 A 5000 A 6300 A 8000 A
800
700
Bus transfer voltage (U)
600
500
400
300
200
100
0
245/300
1 kV 420kV
2 3 550kV 4 800kV 51100kV
Figure 6.2.6.2: UI characteristics for bus transfer in AIS
The bus-transfer voltage could be defined in dependency on the rated voltage as well as on the rated current.
Because the dependency on rated current is lower, especially for GIS the rated bus-transfer voltage could be
defined using the maximum values indicated in Figure 6.2.6.1. For AIS and Hybrid IS the influence of the rated
current is higher and significant, two or more classes or the rated current as additional parameter could be
introduced.
6.2.7 Conclusion on bus transfer switching by disconnector switches

The maximum current though a bus disconnector may be larger than the load current of the corresponding bay
since the bus-transfer current is the sum of distributed current from each bay. The limited statistical data provided
in the report show that the 90 percentile of the ratio between the maximum load current and the rating of a bay is
between 0.5 and 0.8 (see Table 6.2.4.1). The bus transfer current to be switched could be relatively high e.g. up to
80 % of the rated current, depending on the loop length and the number of the bay that supply the bus-transfer
current, but more information is needed.
Page 131
The IEC bus-transfer voltages are too high compared to the actual values and thus the loop lengths can be
sufficiently long. This was verified from the actual substation layouts available. With the maximum allowable bus
transfer currents as per IEC for voltages up to 800 kV of 1600 A, the allowable loop lengths are longer than actual
GIS setup whereas these become near a real GIS case in case of transfer currents with rated capacity.
The values based on statistical evaluations are given in Figure 6.2.7.1 for GIS and in Figure 6.2.7.2 for AIS and
Hybrid IS.
GIS
300kV 420kV 550kV 800kV 1100kV

45
40
35
)
(V
e
g
a
tl 30
o
V
r
e
f
s
n 25
a
tr
s
u
B
20
15
10
1500 2000 2500 3000 3500 4000 4500
Bus transfer current (A)
Figure 6.2.7.1: UI characteristics for bus transfer in GIS

AIS
300kV 420kV 550kV 800kV 1100kV

650
600
) 550
V
(
e
g 500
a
tl
o
V
r 450
e
f
s
n
a
rt 400
s
u
B
350
300
250
200
150
1500 2000 2500 3000 3500 4000 4500
Bus transfer current (A)
Figure 6.2.7.2: UI characteristics for bus transfer in AIS
Thus, the IEC value of 130 V for 1100 kV is supported by this study. For the AIS/Hybrid case, with the maximum
allowable bus transfer currents as per IEC for voltages up to 800 kV of 1600 A, the allowable loop lengths are
longer than actual AIS setup whereas these become near a real AIS case in case of transfer currents with rated
capacity. The IEC value of 750 V for 1100 kV is on the low side since the allowable loop lengths are very low
(< 500 m).
Page 132
6.3 Bus-charging current switching by disconnectors

6.3.1 Maximum bus-charging current
The bus-charging current switching is the current interruption capability of disconnectors when de-energizing long
busbars or other energized parts, for example short length of cables, etc. Very fast transient overvoltages (VFTO)
arise within a GIS whenever there is an instantaneous change in voltage. More often this change occurs as a result
of the opening or closing of a disconnector switch (DS). It was decided to prove the correct design by special
switching tests. Annex F of IEC 62271-102 [12] describes the requirements for switching of bus-charging currents
by disconnectors for rated voltages of 72.5 kV and above.
DS normally have a capacitive current interrupting capability. For air-insulated DS, the interrupting current does not
exceed 2 A for the system operating voltages up to 800 kV. Regarding GIS DS, the interrupting current for test duty
3 ranges from 0.1 A at 72.5 kV and up to 0.8 A at 800 kV [12]. Extrapolation of specified bus-charging currents
suggest that the required bus-charging current switching capability of GIS DS at UHV is in the range from 1 A to
1.2 A.
The switching current (I), which corresponds to the bus-charging current, can be calculated by using equation:
I =ω C U , (where C is load capacitance and U is the rated voltage)
The load capacitance is the product of capacitance per meter of the bus multiplied by the busbar length. Typical
values for EHV and UHV are given in Table 6.3.1.1.
Based on this equation, the following table gives the comparative analysis for the data provided by the participants
of the CIGRE survey.
Voltage level UHV UHV UHV EHV EHV

Design GIS / Japan GIS / China HIS / China AIS / India GIS
Rated voltage (kV) 1100 1100 1100 800 550
Rated current (A) 8000 8000 8000 4000
Rated frequency (Hz) 50 50 50 50 50
Bus-charging current (A) 0.4 0.26 0.23 0.16 0.04
Maximum length of the gas-insulated busbar at load side (m) 32 11 7.6 22.6 7
Maximum load side capacitance (nF) 2 1.295 1.142 1.087 0.37
Average capacitance per meter (pF/m) 62.5 117.7 150.3 48.1 53
Table 6.3.1.1: Bus-charging current and related values
6.3.2 Very fast Transient Overvoltages (VFTO)

Very fast transient overvoltages (VFTO) arise within a GIS any time there is an instantaneous change in voltage.
Most often this change occurs as a result of the opening or closing of a disconnector switch (DS). Other events,
such as the operation of a circuit-breaker (CB), the occurrence of a line-to-ground fault or the closing of an earthing
switch can also cause VFTO. However, during a DS operation a high number of restrikes and pre-strikes occur due
to the low operating speed of DS compared to a circuit-breaker. Therefore, DS switching is the main source for
generating VFTO. The transients are characterized by their short duration and very high frequencies. The rise
times are in the range of some ns, with dominant frequency components up to 100 MHz. The generation and
propagation of VFTO from their original location throughout a GIS can produce internal and external transient
overvoltages. The main concerns are internal overvoltages between the conductor and the enclosure. Internal
VFTO cause high stress of the insulation system. It has been found that, particularly at 420 kV and higher system
voltage levels, disruptive discharges to earth might occur when switching small capacitive currents by gas-insulated
DS. The development of an earth fault by branching of the leader during DS switching depends on parameters
such as voltage, gap distance, electrode geometry, contact speed, gas pressure and magnitude and frequency of
VFTO. A proper design of the DS has shown, that in practice earth faults can be eliminated [2].
Page 133
However, external VFTO can be dangerous for secondary and adjacent equipment. These external transients
include transient voltages between the enclosure and ground at GIS-air interfaces, voltages across insulating
spacers in the vicinity of GIS current transformers, when they do not have a metallic screen on the outside surface,
voltages on the secondary terminals of instrument transformers, radiated electromagnetic fields (EMF) which can
be dangerous to adjacent control or relay equipment [74].
Different CIGRÉ working groups (WG D1.36, WG A3.22, WG A3.28, WG C4.306, WG B3.29) have studied the
phenomena occurring in UHV AC substations, which are considered to behave differently from HV-substations.
CIGRÉ Working Group C4.306 has finalized the insulation coordination studies and a CIGRÉ Technical Brochure
was published in June 2013 [76]. Within CIGRÉ Study Committee D1 (Materials and Emerging Test Techniques),
AG D1.03 and WG D1.36, in co-operation with WG A3.22 / WG A3.28 and WG C4.306 have also finalized their
studies and published a common CIGRÉ Technical Brochure on behalf of Study Committee D1 (Materials and
Emerging Test Techniques), A3 (High Voltage Equipment), B3 (Substations) and C4 (System Technical
Performance) [75].
VFTO in GIS are of greater concern at the highest rated voltages, for which the ratio of the lightning impulse
withstand voltage (LIWV) to the system voltage is lower [2]. As the rated voltage increases, the difference between
the rated lightning impulse withstand voltage (LIWV) and the VFTO decreases. The maximum calculated VFTO in
GIS system may reach the insulation level of LIWV. In case of Hybrid IS a maximum calculated VFTO of 2.2 pu is
reported, because of the lower length of busbar sections [2].
Summarizing the different experiences a procedure was proposed, following the general insulation co-ordination
approach [77]. The procedure consists of the following three steps:
Step 1 Calculation of VFTO (peak value and rise time)
 System analysis (travelling wave computer simulation program)

 Calculation of the maximum peak value and rise time for the GIS and the connected equipment
 Use of real trapped charge behaviour of the disconnector switch, if known
o 99 % probability value determined by simulation or
o maximum values measured during testing
 Otherwise the worst case assumption of a trapped charge voltage of -1 pu should be used for the
simulation.
 The accuracy of the simulation model must be verified.
Step 2 Calculation of required VFTO withstand voltage UCW_VFTO for the different equipment by using:
 Co-ordination factor Kc (statistical distribution, inaccuracy of simulation, frequency of occurrence).
In case of a proved simulation tool a coordination factor of 1.05 is recommended.
 Safety factor Ks (aging behaviour in service, quality of installation atmospheric correction). For
external insulation a safety factor of 1.05 is recommended, whereas for internal insulation a safety
factor of 1.15 is recommended.
 Test conversion factor Ktc (conversion of VFTO peak value to LIWV for a given equipment or
insulation configuration, the factor to be applied to the required withstand voltage, which describes
the different withstand behaviour under VFTO stress compared to the stress with standard LI
voltages). For SF6 insulated systems like GIS or GIL the recommended Ktc is 0.95. The
recommended test conversion factor Ktc for oil or oil/solid insulated systems is 1.0.
Comparison of calculated required VFTO withstand voltage values with LIWV level
Step 3 Definition of measures according to the insulation co-ordination
 No damping measure required

 Damping measure required (DS with low TCV, Damping resistor – definition of required resistance
value, other mitigation methods)
Detailed presentation of each step is given in the following.
Page 134
Step 1 – VFTO Calculation

The accuracy of a simulation depends on the quality of the model of each individual GIS component. In order to
achieve reasonable results even for time periods of some micro-seconds or for very complex GIS structures, highly
accurate models for each internal component and also for external components, connected to the GIS, are
necessary. An accurate modelling of each individual GIS component makes it possible to reproduce VFTO
waveforms with a relatively high precision (differences lower than 5 %), especially in short GIS structures or test
equipment. Nevertheless, variations of more than 10 % depending on the calculation methods are reported.
Therefore, it is important to verify the simulation results by measurements. On-site test of VFTO has been
systematically carried out and their measured data have been compared with calculated ones in China [78].
Step 2 – Required VFTO withstand level - Comparison with LIWV

Basis for the insulation co-ordination is the calculation of the required VFTO withstand voltage. The insulation
withstand strength is equal to LIWV divided by the safety factor [79]. The necessary safety depends on many
factors and could be different for the different kinds of equipment. Generally, the safety factor to be defined is
influenced by the breakdown behaviour of the insulating material, the frequency of occurrence and the probability
of trapped charge voltages as a basis for the simulation. Further aspects should not be disregarded. The absolute
number of occurrence of overvoltage processes caused by DS switching during the total lifetime of the equipment
is in the range of some thousands up to ten thousand whereas the number of overvoltage processes due to
lightning is in the range of some tens up to a few hundred.
An accurate modelling of each individual GIS component makes it possible to reproduce VFTO waveforms with a
relatively high precision (differences lower than 5 %), especially in short GIS structures or test equipment. In case
of a proved simulation tool a coordination factor of 1.05 is recommended. Exceptions are possible in case of
unsecure simulation.
Sometimes, variations of more than 10 % depending on the calculation methods are reported. Generally, it is
important to verify the simulation results by measurements. But in the described case a coordination factor of 1.1 is
recommended. Otherwise, if it is verified that the simulation results are higher in any case compared to the
measured VFTO peak values a lower a coordination factor of 1 could be used.
The maximum value of the VFTO depends on the voltage drop at the DS just before striking and on the location
considered. For the calculation of VFTO stresses, the trapped charges remaining on the load side of the DS must
be taken into consideration. Switching by a slow acting DS generates numerous re-strikes between the moving and
stationary contacts, but 99 % of the trapped charge voltages were limited to 0.4 pu [75]. For these cases, the
resulting VFTO is in the range of 1.7 pu and reach 2.0 pu for very specific cases. Fast operating DS on the other
hand can leave trapped charge levels corresponding to 1 pu in a non-negligible number of cases [2]. The trapped
charge voltage behaviour strongly depends on the contact speed, as shown in the simulation plot [75]. A lower
trapped charge voltage gives a higher safety margin compared to the calculation based on a trapped charge
voltage of 1 pu. For the insulation co-ordination this additional margin has to be considered.
More detailed investigations about the trapped charge voltages are reported in [80]. The trapped charge (TCV)
level depends on the contact speed and on the dielectric withstand of the disconnector gap as a function of the
stroke. The prediction of the multiple restrike scenarios and finally the trapped charge characteristics can be done
by the analysis of the inter-contact breakdown voltage, representing the difference between the load side and the
source side voltages necessary for a restrike (ΔU). This data can be plotted in relation with the sign of the inter-
contact breakdown voltage as presented in Figure 6.3.2.1.
Page 135
Figure 6.3.2.1: Positive and negative inter-contact breakdown voltages versus stroke - Test
results [80]
From these curves, it is possible to build a geometric pattern which represents the disconnector switch behavior
regarding trapped charge voltages. Four curves are necessary to plot the geometric pattern of Figure 6.3.2.2 and
their equations are the followings.
Curve1(t) = -1pu + |ΔU+|min(t)
Curve2(t) = +1pu - |ΔU-|min(t)
Curve3(t) = -1pu + |ΔU-|max(t)
Curve4(t) = +1pu - |ΔU+|max(t)
Figure 6.3.2.2: Positive and negative inter-contact breakdown voltages versus stroke - Test
results [80]
Curve 1 and curve 2 define the possible values of the load side voltage after a restrike. Curve 3 and curve 4 intend
to determine if a restrike can occur for a given load side voltage. From these four curves, four voltage-time areas
could be defined.
First one is the restrike area (grey area). If the load side voltage appears in this area, the phenomena of restrike
will continue till the load side voltage arrives into the stability area delimited by curves 3 and 4 (white cross-
hatching area). The phenomena of restrike stop when the load side voltage arrives inside the stability area [81] and
this final load side voltage is the trapped charge. The third area to consider is the exclusion area between curves 1
and 2 (blue area). It is the voltage-time domain where no load side voltage can be found. The exclusion area
appears later than the stability area and geometrically superimposed on the stability and restrike areas. Finally, the
Page 136
subtraction of the exclusion area (green area) from the stability area defines the existence area of the trapped
charges.
Each disconnector has its own geometric pattern defining the possible values of trapped charges. The
consideration of the voltage-time domain of trapped charge is however very instructive for design optimization
regarding trapped charge mitigation. As a consequence the trapped charge repartition follows a statistical
distribution around one or two main poles depending on the shape of the existence area as shown in Figure 6.3.2.3.
Figure 6.3.2.3: Example of trapped charge distributions around one or two poles (two different
disconnectors) [80]
To reduce the trapped charge voltages the disconnector design must have an existence area around low voltage
value as long as possible during the stroke of the disconnector. Several parameters can be tuned to reach this
target. First parameter is the speed of the disconnector contact. A high opening contact speed will provide a fast
enlargement of the existence area to high values while a slow opening contact speed will provide an existence area
concentrated around low voltages (Figure 6.3.2.4). The second parameter to control the stability area is the
asymmetry of the inter-contact breakdown voltages. The asymmetry is linked to the ratio between positive and
negative inter-contact breakdown voltages. For instance, a -20% asymmetry means that the negative inter-contact
breakdown voltage is higher, in absolute value, than the positive one by 20%.
The limitation of trapped charges has several interests. First one is the reduction of the highest VFTO magnitude
during an opening and a closing operation. The maximum of a VFTO can be calculated with the following formula:
VFTO = K × ΔU + Q
where K is a factor depending on the circuit, ΔU is the inter-contact breakdown voltage before the strike and Q is
the load side voltage before the strike. By limiting the maximal trapped charge, both Q and ΔU are decreased
leading to lower VFTOs.
Figure 6.3.2.4: Trapped charge distributions for the same disconnector opening at different
speeds. [80]
The safety factor Ks describes the aging behaviour in service, the quality of installation and the atmospheric
correction. The safety factor for the VFTO insulation coordination is identical to the safety factor used for LIWV. For
Page 137
external insulation a safety factor of 1.05 is recommended, whereas for internal insulation a safety factor of 1.15 is
recommended.
Breakdown caused by VFTO is improbable in a well-designed GIS insulation system during normal operations.
However, breakdown values can be reduced by insulation irregularities like protrusions. The breakdown probability
is very low for low VFTO amplitudes. It increases with the frequency of the oscillations and the degree of the field
homogeneity. The VFTO stress has been related to the LIWV, which is generally the base for GIS design. For
sound insulating system the VFTO stress is covered by the withstand capability for standard LI voltages. Caused
by the statistical and formative time lag for the breakdown channel, all VFTO breakdown or flashover voltages are
above the LIWV.
Special attention has to be paid when defects are present. Irregularities of the insulation system like needle shape
protrusions or triple junctions of insulators cause extremely inhomogeneous fields. The inhomogeneous fields due
to defects give considerably lower breakdown values compared to a sound system. But also in case of
inhomogeneous fields the minimum breakdown voltage occurs for lightning impulse waveform having a front time
of 5 μs. For times shorter than 1 μs, the breakdown voltage increases with decreasing time caused by the changing
leader inception conditions. Moreover, fixed protrusions on live parts are usually avoided by a proper design,
quality control and adequate testing in both factory and on-site. They can be detected by sensitive diagnostic
measurements under AC voltage stress.
The test conversion factor Ktc describes the comparison between the breakdown strength of the insulating system
between VFTO and LIWV stress for a given equipment or insulation configuration. The factor to be applied to the
required withstand voltage, which describes the different withstand behaviour under VFTO stress compared to the
stress with standard LI voltages. For SF6 insulated systems like GIS or GIL the recommended Ktc is 0.95. The
recommended test conversion factor Ktc for oil or oil/solid insulated systems is 1.0.
Step 3 – Measures according to the insulation co-ordination

If the calculated amplitude of the VFTO is higher compared to the insulation withstand strength of the equipment, it
is necessary to define measures reducing the risk of failures. There are two possibilities: an increase of the LIWV
or a mitigation of VFTO. The first choice is easy to realize, but cost-intensive. Nevertheless in some cases this
solution has advantages. The second choice aims for mitigation of amplitudes of VFTO and finally for a reduction of
the effect of VFTO on the equipment.
The damping of VFTO by integration of a damping resistor is a well proven technology. Service experiences exist
since more than 10 years [82]. Generally, the mitigation effect of the damping resistor depends on the value of the
resistance. To verify the mitigation effect, calculations and measurements were performed for the Japanese and
Chinese UHV projects.
Figure 6.3.2.5 shows a relation between resistance and VFTO peak for a typical GIS layout with 8 bays and a
double busbar scheme. The calculation is based on the assumption that re-striking occurs when the trapped
charge voltage at the load side and the voltage at the source side were -1 pu and +1 pu respectively. Without
damping resistor, the VFTO amplitude reaches a value of 2.49 pu. This exceeds the LIWV according to the
insulation co-ordination. In case of integration of a damping resistance higher than 200 Ω, the VFTO amplitude can
be suppressed below 1.5 pu. A damping resistance of more than 500 Ω mitigates the maximum VFTO amplitude to
a level lower than 1.13 pu [82]. Consequently the resistance of the damping resistor could be chosen and defined
according to the maximum calculated VFTO and the required mitigation effect. A 110 Ω damping resistor was used
in the Italian 1000 kV project [83]. For the Chinese, Japanese and Korean UHV and EHV projects, it was decided
to use a 500 Ω damping resistor [83]. An example of gas insulated DS structure with damping resistor is shown in
Figure 6.3.2.5.
Page 138
1.13 pu
500 
Figure 6.3.2.5: VFTO in relation to the resistance of the damping resistor (left) Structure of DS
with damping resistor (right) [83]
The damping resistor has to withstand the dielectric stress during striking. The highest voltage across the resistor
occurs shortly after the first pre-strike during close operation. Therefore, it is necessary to prove the voltage
withstanding characteristic and the energy absorption capability of the resistor in case of re-strikes and pre-strikes
between the moving contact and the arcing electrode of the resistor. A flashover across the resistor may lead to
high VFTO comparable to a DS without damping resistor and has to be avoided.
A higher resistance value leads to a higher voltage stress across the damping resistor and can reach values in the
range of 2 pu. Moreover the rate of rise of the voltage across the resistor could be very high and depends on the
set-up and the capacitances on the load and source side. The rate of rise of the voltage across the resistor has to
be considered especially during testing.
The absorption energy strongly depends on the load side capacitance and the voltage across the DS. The required
capacity of thermal energy absorption for the resistor could be calculated by summing up all close-open operations
containing a high number of strikes. For typical applications the required energy absorption for a 500 Ω resistor
ranges between 20 kJ and 35 kJ for one close-open operation [2]. Mostly the thermal absorption capability is
defined to withstand the thermal stress for one close-open operation. The possibility to operate more than one
close-open operation within some minutes which corresponds to the thermal time constant of the damping resistor
is estimated to be very short.
If the maximum VFTO is below the LIWV, no measures need to be taken. Otherwise it is necessary to design
considering the VFTO level as dimensioning criteria or to suppress VFTO by damping devices.
A maximum mitigation effect of 20 % to 25 % is required to supress the VFTO peak below the LIWV for UHV
applications considering the worst cases, as shown in Figure 6.3.2.6.
Figure 6.3.2.6: Structure of DS with damping resistor
Page 139
One way to overcome the drawback of an expensive DS design with damping resistor is to use other internal
damping measures. The new proposal concerning insulation co-ordination requires detailed knowledge about the
trapped charge voltage (TCV) of the DS. The new DS model for the entire VFT process analysis allows for trapped
charge voltage calculation. The assumption of a more realistic TCV value leads to more realistic insulation co-
ordination calculations. The application of a slow-acting DS provides significant reduction of the VFTO up to 25 %
[84]
A new approach for damping is to implement compact electromagnetic high-frequency resonators with low quality
factor specially designed to cover a wider frequency range. The novelty of this idea is not only in designing the
resonators but also in dissipating the received VFTO energy. The VFTO damping effect of the developed RF
resonator tuned to the dominant harmonic component was confirmed by full-size experiments [85]. HF resonators
for damping of VFTOs in GIS should have a special geometrical shape so as not to interfere with the GIS dielectric
design. A suitable shape of the resonator should be elongated parallel to the GIS inner conductor axis in order to
achieve a low enough resonant frequency, with a minimum size in the radial direction [73], [76], [85].
Rings of a nanocrystalline material placed around the GIS conductor leads also to a significant mitigation effect
(10 -20 %). Depending on number, material and size of the rings good results could be achieved. Even the
amplitude of the first peaks was significantly reduced by 20 % [86].
A summary of the different mitigation methods is given in Table 6.3.2.1. By combination of different methods an
increased mitigation effect could be achieved [86].
Page 140
Method Service Mitigation Example

experiences effect
Slow acting DS >20 Years 15-25 %
(low trapped [14]
charge voltage)
VFTO for different contact speed and trapped charge

voltage behaviour for 1100 kV DS
Damping resistor >10 Years Up to 90 %
[11]
Nanocrystalline Full scale tests Up to 25 %

material [14] 500
voltage / kV
400
300
200
100
-100
0 10-7 2·10-7 3·10-7
time / s
VFTOs without magnetic rings (continuous line) and with 20

magnetic rings (dashed line)
High frequency Full scale tests Up to 20 %

RF resonator for [14]
VFTO damping
The measured VFTOs in frequency domain of a 550 kV GIS

with and without resonator
Table 6.3.2.1: Comparison of different mitigation measures
6.3.3 Conclusion on bus-charging current switching by disconnectors
Page 141
For UHV, bus-charging current switching has been addressed in Edition 1.2 of IEC 62271-102 with special
attention to VFTO. There are three test duties specified in Annex F of IEC 62271-102 [12]:
1. Switching of a very short section of busbar duct, where VFTO plays a role;
2. Switching of parallel capacitors for circuit-breakers under 180° out-of-phase conditions;
3. Current switching capability.
Several solutions are available to limit the amplitude of VFTO, that in UHV applications may reach values as high
as the BIL or even higher. If the maximum VFTO is below the LIWV, no measures need to be taken. Otherwise it is
necessary to design considering the VFTO level as the dimensioning criteria or to suppress VFTO by damping
devices. A maximum mitigation effect of 20 % to 25 % is required to suppress the VFTO peak below the LIWV for
UHV applications considering the worst cases. New mitigation methods are available today.
The out-of-phase condition is relevant when the DS on the transformer side is required to be closed before
synchronizing the system such as in the case of a generator circuit directly connected to the UHV network. The
amplitude of the temporary power frequency overvoltage results from the combination of the trapped charge
voltage and the over-coupled voltage from the CB grading capacitors. The amplitude of the power frequency
overvoltage, greatly exceed those of the rated withstand voltages, especially for high speed DS. A test has to be
carried out according to the specification by agreement between manufacturer and user with deep consideration of
the insulation co-ordination. It has to be considered that the test arrangement reflects the arrangement in service.
Especially the distance between DS and CB should be as short as possible corresponding to the assembly
conditions. There is no need to change the requirements for test voltages and for the test circuit as given in [12].
The test duty for current switching capability is of interest as the value of the charging current, as specified by the
user, may become quite large. Within WG A3.28 such situations have been considered and the common opinion is
that for long lengths of busbars, utilities will prefer to apply circuit-breakers rather than disconnectors to switch the
charging current, especially with vital installations, such as for the highest voltage classes. Maximum values of the
charging current in GIS are in line with [12] where 0.8 A is given for 800 kV and an extrapolation to 1.0 or 1.2 A for
UHV seems to be reasonable.
6.4 Earthing switch

Switching conditions by earthing switches (i.e. current, recovery voltage and TRV) were studied in cases of
electromagnetically (EM) and electrostatically (ES) induced current interruption. A parameter study was done for
800 kV and 1100 kV earthing switches.
Line lengths from 120 km to 360 km were considered and the load current was varied from 2 kA to 8 kA.
6.4.1 Configurations of the transmission lines and towers

The configuration of towers and characteristics of conductors in transmission lines are as shown in Figure 6.4.1.1
and Table 6.4.1.1.
Page 142
Figure 6.4.1.1: Tower models for simulations
(a) Overhead grounding wire

Overhead grounding wire
Tower Tower A Tower B Tower C Tower D
Ground
Type and size of OPGW(500mm2), AW200, A.G.-1/2",
conductor,
lines 2wires 2wires 4wires
2wires
Resistance of
0.112 ohm/km 0.2000ohm/km 3.1ohm/km 0.489 ohm/km
conductor
Sag of line ― 16m 11.7m 18m
Rated voltage 1100kV 800kV 1100kV
Current for
2,000,4,000, 6000, 8,000A
energized line
Transmission
120km, 240km, 360km
line length
(b) Overhead power line
Page 143
Overhead power line

Tower Tower A Tower B Tower C Tower D
Phase conductor
ACSR Cardinal Bersimis Phase conductor
Type and size of ACSR(810mm2),
480, 689.5mm2 630 (500mm2),
lines 8wires
6wires 42/7ACSR, 8wires
4wires
Resistance of 0.0356 ohm/km 0.0599 ohm/km 0.0427ohm/km 0.04633ohm/km
conductor (per wire) (per wire) (per wire) (per wire)
Sag of line ― 13.41m 7.6m 20m
Rated voltage 1100kV 800kV 1100kV
Current for
2,000,4,000, 6000, 8,000A
energized line
Transmission
120km, 240km, 360km
line length
Table 6.4.1.1: Characteristics of conductors in transmission lines for calculation
6.4.2 Calculation of electromagnetically (EM) induced current switching
The calculation was carried out assuming 120 km / 240 km / 360 km for 800 kV and 1100 kV untransposed
transmission lines and a load current of 2000 A, 4000 A, 6000 A and 8000 A in one operated line of a double
transmission line (two single transmission lines with Tower C) configuration, while the other is de-energized. The
simulation model for EM induced current switching is as shown in Figure 6.4.2.1.
The de-energized line conditions for the maximum EM current, recovery voltage, and RRRV are shown in Figure
6.4.2.2. These conditions are chosen from 96 possible combinations of each phase and each side open-close
condition/operation (see Figure 6.4.2.3). Results of simulations are given in Table 6.4.2.1.
Figures 6.4.2.3 and 6.4.2.4 give the EM current, recovery voltage and RRRV as function of load current and line
length. Typical waveforms of transient recovery voltage are as shown in Figure 6.4.2.5.
Figure 6.4.2.1: De-energized line conditions (EM)
Figure 6.4.2.2: Calculation cases
Page 144
(a) EM induced current

Line Load EM induced current
Length current Tower A Tower B Tower C Tower D
2000 212.0 259.2 205.2 294.1
4000 424.1 518.7 410.7 588.4
120
6000 636.1 778.1 616.0 882.6
8000 848.2 1037.4 821.3 1176.7
2000 217.8 259.2 205.3 297.6
4000 435.7 519.1 410.7 595.4
240
6000 653.5 778.6 615.9 893.1
8000 871.3 1038.1 821.3 1190.8
2000 217.0 261.0 208.8 297.1
360 4000 434.3 522.2 417.6 594.4
6000 651.4 783.3 626.3 891.5
8000 868.5 1044.4 835.1 1188.7
(b) Recovery voltage

Line Load Recovery voltage
2000 9.7 16.0 20.1 15.1
120 4000 19.5 32.0 40.2 30.3
6000 29.3 48.1 60.3 45.4
8000 39.0 64.1 80.4 60.6
2000 20.5 33.1 42.7 31.3
240 4000 40.9 66.1 85.4 62.7
6000 61.4 99.2 128.0 94.0
8000 81.8 132.2 170.7 125.3
2000 31.5 52.8 71.1 48.7
360 4000 63.2 105.5 142.2 97.5
6000 94.7 158.3 213.3 146.2
8000 126.3 211.0 284.4 194.9
(c) RRRV
Line Load RRRV
2000 31.0 45.3 37.0 43.8
4000 60.4 88.7 76.8 87.4
120
6000 92.9 138.5 115.2 131.1
8000 123.9 179.4 153.6 174.9
2000 31.0 44.0 39.2 43.8
4000 62.0 90.7 76.6 87.7
240
6000 92.4 135.2 115.6 131.5
8000 124.8 181.4 156.0 175.4
2000 29.9 41.7 37.2 41.5
360 4000 58.9 86.2 76.3 83.7
6000 88.4 128.2 114.9 125.6
8000 116.9 172.4 152.6 167.5
Table 6.4.2.1: Results of electromagnetically induced current switching for 1100 kV
Page 145
EM induced [Arms] Recovery voltage [kV rms] RRRV [kV/ms]
1600 300 300
EM recovery voltage[kVrms]
120km 120km 120km
EM induced current[Arms]
1400 250 250

240km 240km 240km
1200
RRRV[kV/ms]
200 200
Tower A
360km 360km 360km

1000
150 150
800
100 100
600
400 50 50
200 0 0
2000 4000 6000 8000 2000 4000 6000 8000 2000 4000 6000 8000
Load Current[Arms] Load Current[Arms] Load Current[Arms]
1600 300 300

120km
120km 120km
1400 250 250

240km 240km
240km
1200 360km
RRRV[kV/ms]
360km 200 200
Tower B
360km
1000
150 150
800
100 100
600
400 50 50
200 0 0
2000 4000 6000 8000 2000 4000 6000 8000 2000 4000 6000 8000
1600 300 300

120km 120km
120km
1400 250 250

240km 240km 240km
1200
RRRV[kV/ms]
360km 200 360km 200 360km
Tower C
1000
150 150
800
100 100
600
50 50
400
200 0 0
2000 4000 6000 8000 2000 4000 6000 8000 2000 4000 6000 8000
1600 300 300

120km
120km 120km
1400
240km 250 250
240km 240km
1200 360km
RRRV[kV/ms]
200 200 360km

Tower D
360km
1000
150 150
800
100 100
600
50 50
400
0 0
200
2000 4000 6000 8000 2000 4000 6000 8000
2000 4000 6000 8000
Load Current[Arms] Load Current[Arms]
Load Current[Arms]
Figure 6.4.2.3: Result of electromagnetically induced current switching for 1100 kV or 800 kV ES as
function of load current
Page 146
320 320
2000A 2000A
280 280
4000A 4000A
240 6000A 240 6000A
200 8000A 200 8000A
160 160
120 120
80 80
40 40
0 0
0 100 200 300 400 0 100 200 300 400
Line Length[km] Line Length[km]
(a) Tower A (b) Tower B

320 320
280 2000A 2000A
280
4000A 4000A
240 240
6000A 6000A
200 8000A 200 8000A
160 160
120 120
80 80
40 40
0 0
0 100 200 300 400 0 100 200 300 400
Line Length[km] Line Length[km]
(c) Tower C (d) Tower D
Figure 6.4.2.4: Relationship between line length and EM recovery voltage
Page 147
80 20
[kV]
[kV]
60
0
40
20
-20
-20 -40
-40
-60
-60
-80 -80
0.00 0.02 0.04 0.06 0.08 [s] 0.10 16.0 16.5 17.0 17.5 18.0 18.5 [ms] 19.0
(file towerA.pl4; x-var t) v:P2AU (file towerA.pl4; x-var t) v:P2AU
(a) Tower A: 120 km - 4000 A

80 20
[kV] [kV]
60
0
40
20 -20
0
-40
-20
-40
-60
-60
-80 -80
14.0 14.5 15.0 15.5 16.0 16.5 [ms] 17.0
0.00 0.02 0.04 0.06 0.08 [s] 0.10 (file towerB.pl4; x-var t) v:P2AU
(file towerB.pl4; x-var t) v:P2AU
(b) Tower B: 120 km – 4000 A

80 20
[kV]
[kV]
60
0
40
20
-20
-40
-20
-40
-60
-60
-80 -80
0.00 0.02 0.04 0.06 0.08 [s] 0.10 14.0 14.5 15.0 15.5 16.0 [ms] 16.5
(file towerC.pl4; x-var t) v:P2AW (file towerC.pl4; x-var t) v:P2AW
(c) Tower C: 120 km – 4000 A
80 10
[kV] [kV]
60
-5
40
-20
20
0 -35
-20
-50
-40
-65
-60
-80 -80
0.00 0.02 0.04 0.06 0.08 [s] 0.10 16.5 17.0 17.5 18.0 18.5 19.0 [ms] 19.5
(file towerD.pl4; x-var t) v:P2AU (file towerD.pl4; x-var t) v:P2AU
(d) Tower D: 120 km – 4000 A
Figure 6.4.2.5: Typical wave form of recovery voltage of electromagnetically induced current switching
(left: overall, right: enlarged)
Page 148
6.4.3 Calculation of electrostatically (ES) induced current switching

The configuration and parameters of 6.4.2 are used for the electrostatically induced current analysis. The
simulation model for ES induced current switching is as shown in Figure 6.4.3.1.
The de-energized line conditions for the maximum recovery voltages are shown in Figure 6.4.3.2. Results of
simulations are given in Table 6.4.3.1 and Figure 6.4.3.3.
Figure 6.4.3.1: Simulation model of electrostatically induced current switching

for 1100 kV or 800 kV ES
Figure 6.4.3.2: De-energized line conditions (ES)

ES recovery voltage[kVrms]
Line Length[km]
Tower A Tower B Tower C Tower D
120 42.1 19.6 17.5 49.1
240 40.2 20.8 15.9 48.1
360 40.8 21.1 12.2 49.1
ES induced current[Arms]
Line Length[km]
Tower A Tower B Tower C Tower D
360 59.8 37.8 23.8 72.8
Table 6.4.3.1: Results of electrostatically induced current switching for 1100 kV and 800 kV ES
80
TowerA
70
ES recovery voltage[kVrms]
TowerB
60 TowerC
TowerD
50
40
30
20
10
0
0 100 200 300 400
Line Length[km]
Figure 6.4.3.3: Relationship between line length and ES recovery voltage

6.4.4 Influence of line transposition
Page 149
Line transposition is used to balance the normal current in each phase of single or double circuit lines. Figure
6.4.4.1 shows non-transposed double circuit lines used for the study. Figures 6.4.4.2 and 6.4.4.3 show β-
transposition (full-transposed lines) and γ-transposition (half-transposed lines) for double circuit lines respectively.
The β-transposition has two transpositions for one circuit and six transpositions for another circuit and each circuit
is transposed at different points on the line. The γ-transposition has two transpositions for both circuits at the same
points on the tower.
Parameter study of electromagnetically (EM) induced and electrostatically (ES) induced current switching duty by
earthing switches were carried out with the following conditions:
－line length: 360 km

－load current of energized line: 8 kA
－Tower: A
Figures 6.4.4.4 to 6.4.4.7 show the calculation results obtained for the cases shown on Figures 6.4.4.1 to 6.4.4.3.
Figure 6.4.4.1: non-transposition for double circuit lines
Figure 6.4.4.2: β (beta)-transposition or full-transposition for double circuit lines
Page 150
Figure 6.4.4.3: γ (gamma)-transposition or half-transposition for double circuit lines
Figure 6.4.4.4: Electromagnetically induced current and existing IEC standard value for 800 kV (influence
of transposition)
Figure 6.4.4.5: Recovery voltages for electromagnetically induced current interruption and existing IEC
standard value for 800 kV (influence of transposition)
Page 151
Figure 6.4.4.6: Electrostatically induced current and existing IEC standard value for 800 kV (influence of
transposition)
Figure 6.4.4.7: Recovery voltages for electrostatically induced current interruption and existing IEC
standard value for 800 kV (influence of transposition)
Figure 6.4.4.4 shows the comparison of electromagnetically induced current. Induced currents are greatly reduced
when ideal transposition is applied with fully or half-transposed lines. In case of γ-transposition (half-transposed
lines), the current is roughly divided by two and in case of β-transposition (fully-transposed lines) the inductive
effect is cancelled so that the induced current is divided by ten.
Figure 6.4.4.5 shows the comparison of recovery voltages for electromagnetically induced current breaking. The
tendency seen in Figure 6.4.4.4 is also observed in Figure 6.4.4.5.
Figure 6.4.4.6 shows that the electrostatically induced current calculated is close to the IEC value in case of
transposed lines.
Figure 6.4.4.7 shows the comparison of recovery voltages for electrostatically induced current breaking. Again
recovery voltages are greatly reduced with full- and half-transposition. Calculated results with β- and γ-transposition
are covered by the value defined in IEC standard for 800 kV systems.
Attention must be paid to the fact that these results are based on the condition of ideal transposition. In reality, it is
very difficult to construct β-transposed (full-transposed lines) transmission lines. Also, even in case of γ-
Page 152
transposition (half-transposed lines) the length of each segment between transposing points cannot be same.
Therefore the calculated results with transposition are minimum values.
Transposition is applied to long transmission lines, typically more than 200 km. In Japan UHV designed
transmission lines are not transposed up to 210 km.
6.4.5 Switching duty without transposition

Using data from previous reports by CIGRE WG A3.22 [1], [2], typical tower designs and network conditions are
summarized in Table 6.4.5.1.
Considering that the nominal current is carried on the line in each model, i.e. 8000 A for Tower A and B, 4000 A for
Tower D, and 3150 A for Tower C, calculated results are shown in Figures 6.4.5.1(a) and 6.4.5.2(a). Usually utility’s
design keeps some allowances on the current carrying capability. The examples of field data collected for 500 kV
and 275 kV systems in Japan (see also Chapter 7) show that the 90 % probability value of the maximum current for
a feeder bay is close to 50 % of the rated current. Considering that induced current switching is conducted with one
line of a double circuit de-energized and that the number of the field data is limited, a margin of 20% in the current
value is introduced in the calculation. Then 60% (50% times 1.2) of the rated current is used for evaluation. Figure
6.4.5.1(b) shows an evaluation of induced current based on these references and Figure 6.4.5.2(b) shows the
recovery voltage.
Table 6.4.5.1: Network conditions (CIGRE WG A3.22 survey)
6.4.5.1 Electromagnetically induced current

As shown in Fig. 6.4.5.1(a), calculation results exceed the value for 800 kV system (160 A) in existing IEC
Standard [12]. Especially in case of Tower A and B, induced currents are five to six times larger due to the high
current of 8000 A considered for the nominal current.
Page 153
(a) nominal current on the operating line
(b) 60% of nominal current on the operating line
Figure 6.4.5.1: Electromagnetically induced current and existing IEC standard value for 800 kV
Page 154
(a) nominal current on the operating line
(b) 60% of nominal current on the operating line
Figure 6.4.5.2: Recovery voltages for electromagnetically induced current interruption and existing IEC
standard value for 800 kV
Page 155
6.4.5.2: Recovery voltage for electromagnetically induced current interruption
Recovery voltages are in proportion to the line length as shown in Figure 6.4.2.4. Figure 6.4.5.2 (a) shows an
example of EM recovery voltage in case of a line length of 240 km. The recovery voltage with tower C tends to be
higher because the surge impedance of the overhead line is high compared to the other tower models. However,
due to the high nominal current considered, calculated recovery voltages for towers A and B are higher.
Calculation results for 360 km lines are more than five times higher than the value for 800 kV system (20 kV) in
existing IEC Standard [12].
6.4.5.3: Recovery voltage for electrostatically induced current interruption

Recovery voltages are theoretically independent from the line length, and approximately in proportion to the system
voltage for electrostatically induced current interruption. Because of this, as shown in Figure 6.4.5.3, existing IEC
Standard [12] covers 800 kV systems (towers B and C). Calculated values for 1100 kV (towers A and D), however,
are not covered by IEC values.
Figure 6.4.5.3: Recovery voltages for electrostatically induced current interruption and existing IEC
standard value for 800 kV
6.4.6 Switching duty with transposition

A line length of 360 km is considered in the following except when indicated differently.
6.4.6.1 Switching duty with rated current on healthy line (w/o transposition)
The maximum values obtained for each system voltage, with two tower types, with rated current on the energized
line are listed hereafter.
a) Electromagnetically induced current
1100 kV (Tower A) 869 A 1100 kV (Tower D) 595 A
800kV (Tower B) 1044 A 800 kV (Tower C) 329 A
b) Recovery voltage for electromagnetically induced current breaking
(1) line length 360 km
1100 kV (Tower A) 126 kV 1100 kV (Tower D) 98 kV
Page 156
800 kV (Tower B) 211 kV 800 kV (Tower C) 112 kV
c) RRRV for electromagnetically induced current breaking
1100 kV (Tower A) 117 kV/ms 1100 kV (Tower D) 83 kV/ms
800 kV (Tower B) 172 kV/ms 800 kV (Tower C) 60 kV/ms
d) Electrostatically induced current
800 kV (Tower B) 38 A 800 kV (Tower C) 24 A
e) Recovery voltage for electrostatically induced current breaking
This duty is independent on line length.
1100 kV (Tower A) 42 kV 1100kV (Tower D) 49 kV
6.4.6.2 Switching duty with reduced current on healthy line (w/o transposition)
The maximum values on each system voltage and tower with 60% of rated current on the energized line are listed
hereafter. The value of load current is taken from experience as given in Chapter 7.
800 kV (Tower B) 626 A 800 kV (Tower C) 197 A
Page 157
c) RRRV for electromagnetically induced current breaking
d) Electrostatically induced current: this duty is independent from current on the energized line and then the
same as in 6.4.6.1
e) Recovery voltage on electrostatically induced current breaking: this duty is independent from current on the
energized line and then same as in 6.4.6.1.
The results of simulations show that the electromagnetically induced currents, voltages and RRRV are proportional
to the current in the healthy circuit and therefore 60% of the values in section 6.4.6.1.
6.4.6.3 Switching duty with γ-transposition (half-transposed lines)

Calculations are conducted for the case with a 360 km line and Tower A.
1100 kV (Tower A) 348 A (40% of non-transposition condition)
1100 kV (Tower A) 41 kV (32.5% of non-transposition condition)
c) Electrostatically induced current
1100 kV (Tower A) 29 A (48% of non-transposition condition)
d) Recovery voltage for electrostatically induced current breaking
1100 kV (Tower A) 21 kV (51% of non-transposition condition)
6.4.6.4 Overall requirements for switching duty with transposition

When studying the effect of transposition, it must be taken into account that it is difficult to realize an ideal trisected
γ-transposition (half-transposed lines). Considering the variation in segment length in actual designs of
transmission lines, a margin of 20% is introduced for evaluation. On this basis, the requirement for the
electromagnetically induced current with γ-transposition (half-transposed lines) is thought to be 50% (40% times
1.2) the value without transposition, and the recovery voltage and RRRV are 40% (32.5% times 1.2) the value
without transposition. Electrostatically induced current and its recovery voltage should be 60% (48% times 1.2 or
51% times 1.2) the value without transposition.
Transposition may not be applied to line lengths from 120 km to 240 km, but is generally applied to 360 km
transmission line. Therefore, values without transposition up to 240 km and with transposition in case of 360 km
lines are considered.
Page 158
Regarding electromagnetically induced current, comparison of results shows that the duties with a transposed
360 km line are less severe than that of a non-transposed 240 km line. Then the duties of a non-transposed
240 km line give the more severe condition possible.
A similar tendency is seen for the recovery voltage of electromagnetically induced current switching. Even though
the duties become more severe with line length, non-transposed 240 km lines give the most severe condition
possible.
Line length does not affect the RRRV of electromagnetically induced current switching, the more severe condition
possible is obtained in case of short untransposed lines.
The recovery voltage of electrostatically induced current breaking is not affected by the line length. The highest
values are obtained in case of short untransposed lines.
The electrostatically induced current is proportional to line length. The most severe condition is obtained with a
non-transposed line of 240 km.
Based on the above, estimated switching duties for 800 kV and 1100 kV, including a 20% margin in case of γ-
transposition (half-transposed lines) are listed in Table 6.4.5.2. This table shows the comparison of the following
three combinations of parameters; 360 km transposed line with rated current on the healthy line, 240 km
untransposed line with 60% of rated current on the healthy line, and existing IEC standard [12]. Higher values for
Towers A and D are listed for 1100 kV in Table 6.4.5.2, and higher values of Tower B and C are listed for 800 kV.
Load current 60% 100%

Existing IEC
Transposition non-transposition γ-transposition
(class B)
Line length 240 km 360 km
induced current [Arms] 626 522 160
800 kV Recovery voltage [kV rms] 79 84 20
RRRV [kV/ms] 109 - -
EM
induced current[Arms] 521 434 440
1100 kV Recovery voltage [kV rms] 49 50 65
RRRV [kV/ms] 75 - -
induced current [A rms] 25 23 25
800 kV
Recovery voltage [kV rms] 21 13 32
ES
induced current [A rms] 49 44 50
1100 kV
Recovery voltage [kV rms] 49 29 40
Table 6.4.5.2: Comparison of duty between cases with combinations of parameters and existing
IEC standard
6.4.7 Conclusion on earthing switch interruption duties

Characteristics and tendency for electromagnetically and electrostatically induced current interruption duties for ES
in 800 kV and 1100 kV systems were studied taking into account double tower and 2 x single tower configurations.
Other than these parameters, requirements for earthing switches are greatly affected by the line length, current on
the healthy line, and by transposition. Effects of these parameters were studied using EMTP analysis. These
results will support the future revision of IEC 62271-102.
Page 159
7 Switching experience during and after system commissioning

7.1 Introduction
One of the main objectives of WG A3.28 is to investigate field experience and switching behaviors during and after
UHV/EHV system commissioning. A questionnaire was sent to experts through the WG members and the
information has been received from China, Japan, Russia, Canada, India, USA and Italy. The main results are
summarized in this chapter.
7.2 Experience of UHV/EHV AC System in China

7.2.1 Experience of UHV AC Demonstration Project in China
Rapid growth in electrical power demand in China creates urgency for the development of UHV AC transmission
and the construction of a demonstration project.
The Changzhi-Nanyang-Jingmen UHV AC demonstration project was officially put into operation by the State Grid
Corporation of China (SGCC) on January 6th, 2009. As shown in Figure 7.2.1.1, in this project Changzhi and
Jingmen are both UHV stations with one UHV transformer each. Nanyang is a UHV switching station. The lengths
of Changzhi-Nanyang and Nanyang-Jingmen UHV transmission line are 358.5 km and 281.3 km respectively. UHV
shunt reactors with neutral reactors are installed at each terminal of the lines.
Figure 7.2.1.1: System scheme of China UHV AC demonstration project

7.2.1.1 Switchgear used in UHV AC demonstration project
The requirements and associated test circuits for the type tests for UHV substation equipment has been
investigated based on the experience obtained through the developments and field demonstrations in China. The
main items of the UHV AC switchgear type tests are shown in Table 7.2.1.1.
Type tests are the most important demonstrations to verify the performance of UHV switchgear, which had been
confirmed with the testing conditions extrapolated or presumed from the 550 kV standards, before the CIGRE WG
A3.22” Technical Requirements for Substation Equipment exceeding 800 kV” started to review the UHV
specifications in 2006 and recommended them to IEC in 2008. National standards of China, including GB/Z 24836,
th
GB/Z 24837 and GB/Z 24838, were published on November 30 , 2009. Later, the amendment to cover UHV circuit
breakers was completed in IEC Std. 62271-100 based on the CIGRE recommendations and was published in
September 2012.
Modern 1100 kV circuit breakers are composed of 2 or 4 breaking units in series. Because of limitations in
laboratory test capability, it may be difficult to perform all test duties for 1100 kV circuit breakers on a full-pole.
Therefore, unit tests as well as multi-part tests can be applied to demonstrate the performance specified in IEC
62271-100, except for T100s and T100a. To verify interrupting capability under existing interactions between
multiple breaking units in hot-gas and electro-magnetic force, test duties of T100s and T100a were performed by
full-pole tests according to the test circuit described in CIGRE Technical Brochure 456 [2]. Other short circuit tests
were performed by a unit test, but other breaking units were applied as auxiliary circuit breaker to have equivalent
hot gas volume.
For unit tests, the out-of-phase test duty OP2 was performed with the triple-circuit method and other duties were
performed with the double-circuit method also described in the TB456 [2].
Page 160
GIS GCB DS & ES

Dielectric test Tested Tested Tested as GIS
Radio interference voltage (RIV) test Tested Tested Tested as GIS
Measurement of resistance of main circuit Tested Tested Tested as GIS
Temperature-rise test Tested Tested Tested as GIS
Short-time withstand current and peak withstand
Tested Tested Tested as GIS
current test
Verification of the degree of protection Tested Tested Tested as GIS
Tightness test Tested Tested Tested as GIS
Pressure test on partitions Tested Not applicable Not applicable
Electromagnetic compatibility (EMC) test Tested Tested Tested as GIS
Additional test on auxiliary and control circuits Tested Tested as GIS Tested as GIS
Mechanical operating and mechanical endurance test Not applicable Tested Tested
Test to verify the proper functioning of position-
Not applicable Not applicable Tested
indicating device
Low and high temperature test Not applicable Tested Tested
Operating test under severe icing conditions Not applicable Not applicable Tested
Static terminal load test Not applicable Tested Not applicalbe
Short-circuit current making and breaking test Not applicable Tested Tested for ES
Heat capacity test of opening and closing resistor Not applicable Tested Not applicable
Short-line fault tests Not applicable Tested Not applicable
Out-of-phase making and breaking tests Not applicable Tested Not applicable
Capacitive current switching tests Not applicable Tested (LC1, LC2) Not applicable
Bus transfer current switching tests Not applicable Not applicable Tested
Induced current switching tests Not applicable Not applicable Tested
Bus charging current switching tests Not applicable Not applicable Tested
Electrical endurance test Not applicable Tested Not applicable
Noise level test Not applicable Tested Not applicable
Seismic test Not applicable Tested Not applicable
Table 7.2.1.1: Main items of type tests for UHV AC switchgear
All type tests of circuit breakers have been performed for a short-circuit current rating of 50 kA, which is the
specification applied in the UHV demonstration project in China, while 63 kA is expected in future extension
projects. Characteristics and pictures of these products are shown in Tables 7.2.1.2 and 7.2.1.3, and Figs. 7.2.1.2
to Fig.7.2.1.4.
Test quantities Manufacturer A Manufacturer B Manufacturer C
Rated continuous current (CB/Main busbar) (A) 6300/8000 8000/8000 6300/8000

Rated short-circuit breaking current (kA) 63/50 63/50 63/50
Electrical endurance test number 16 16 16
Rated short-time/peak withstand current (kA) 63/171 63/171 63/171
Rated duration of short circuit 2s 2s 2s
Current for temperature rise test (A) 6930 6930 6930
Closing resistor (Ω) 560 580 600
Number of break 4 2 2
Rated SF6 pressure CB 0.58 0.6 0.65
(MPa) Other compartments 0.36 0.5 0.45
Spring and hydraulic
Operating device type Hydraulic charged Hydraulic charged
charged
Operating power (MPa) 57.3 31.5 31.5
Number of operations conducted for
5000 5000 5000
mechanical endurance test
Table 7.2.1.2: Main characteristics of circuit breakers from three manufacturers
Page 161
Test quantities Manufacturer A Manufacturer B Manufacturer C

Rated continuous current (A) 6300 8000 8000
Rated short-time/peak withstand current (kA) 63/171 63/171 63/171
Rated duration of short-time withstand current 2s 2s 2s
DS switching bus-transfer current test 1600/400V 1600/400V 1600/400V
Operating device type Motor Motor spring Motor spring
Number of operations conducted for
2000 5000 2000
mechanical endurance test
Capacitive switching current (A) 2 2 2
Table 7.2.1.3: Main characteristics of disconnectors from three manufacturers
Figure 7.2.1.2: Switchgear supplied by manufacturer A
Figure 7.2.1.3: Switchgear supplied by manufacturer B
Figure 7.2.1.4: Switchgear supplied by manufacturer C
Page 162
7.2.1.2 High electrical endurance 110 kV capacitor switch

Circuit breakers have to endure high recovery voltages for a number of switching operations when they are applied
for capacitor bank switching. Overvoltages generated by restrikes during small capacitive current interruptions may
cause serious consequences to the reactive power compensation devices and/or to system insulation. In order to
secure the required frequent capacitor switching capability, it is necessary to minimize the probability of restrikes.
On the other hand, capacitor bank energisation can be the most severe duty for circuit breakers. Back-to-back
capacitor bank energisation is the most onerous condition for inrush current since high amplitudes and high
frequencies of inrush current are likely if no countermeasures are taken. The inrush current amplitude can be tens
of times higher than the normal current for a capacitor switch, and the inrush current frequency may reach several
hundreds or thousands Hertz. Excessive inrush currents and associated overvoltages may cause CB contact and
nozzle erosion (or puncture) and other potential damage. In a back-to-back configuration, capacitor banks are
installed close to each other, so the inductance of the interconnections is quite small. Therefore, when a capacitor
bank is switched on, the other capacitor banks will charge it with an inrush current that may be of high amplitude
and high frequency. Adequate electrical endurance is necessary for the switching device as a high number of
switching operations is required with high inrush current. In the case of UHV networks, as capacitor banks are
installed on the tertiary of the transformer, a high number of making and breaking operations must be specified.
Therefore, it is very important to test and verify the ability of the circuit breaker for the making and breaking of back-
to-back capacitor bank currents for this particular application with capacitors banks on the tertiary of the
transformer.
Class C2 test parameters of capacitive current switching have been defined by IEC Std. 62271-100. However, the
test method, test procedure and test criteria for extended capacitive current switching tests have not been specified
by IEC. Also, test current of capacitor bank switching exceeded the preferred values of international standards.
Therefore, it is important to determine the test method, test procedure, test criteria and technical parameters.
As a survey of operating conditions of UHV AC equipment in China, switchgear at the tertiary side of a UHV AC
transformer will operate 2-4 times per day or 730-1460 times per year. The result shows that economically
reasonable lifetime of the switchgears applied to capacitive switching would be 5000 time operations. According to
the power system needs, the voltage at the tertiary side of the transformer varies with the system voltage and
position of the tap-changer, and may exceed 126 kV. Therefore, the rated voltage of the capacitor switchgear is
required to be 145 kV.
There are three options for the specification of the switching device at the tertiary side of the UHV transformer.
a) Circuit breaker
During capacitor bank de-energization, a restrike may occur with a certain probability of occurrence depending on
the design of circuit breakers. Decreasing the restrike probability is a complex problem of matching electrical and
mechanical characteristics. The ideal situation is obtained with a medium arcing time and an appropriate arc-
extinguishing capability. When the arc is extinguished, the dielectric strength rapidly recovers to a relatively high
level, and continuously increases with the separation of contacts and should always be larger than the recovery
voltage. Experimental studies show that methods to improve the capacitive current switching performance include
increasing the closing and opening speed appropriately, improving the ablation resistance ability of the contact
materials, improving the structure of the nozzle, increasing gas pressure, etc.
Dielectric recovery characteristic between circuit breaker contacts depends on contact roughness and opening
speed. Increasing closing speed leads to shorter pre-arcing time and smaller erosion of contacts and nozzle,
therefore, faster closing speed can improve capacitive current switching endurance and lower the probability of
restrike. Increasing opening speed may result in less mechanical reliability of circuit breakers because a higher
mechanical stress could be applied to moving parts due to a larger driving force. Therefore an optimization in
opening speed is important for the design of circuit-breakers.
b) Load switch
Capacitor bank nominal load switching current is much lower than short-circuit current and the requirements for
capacitor switching and short-circuit current switching are different. Therefore the two switching duties can be done
Page 163
by separate devices, a load switch can be used for capacitor switching and a circuit breaker can be used for short-
circuit switching.
Adopting a load switch is relatively easy to realize, but it will increase the amount of switchgears and capital
investment in the project.
c) Controlled switching
When energizing capacitor banks, controlled closing allows the electrical contact within the CB chambers to occur
at a time when the voltage is close to zero so that it will optimally decrease the pre-breakdown time, the inrush
current amplitudes, and, consequently, the CB contact ablation. When de-energizing shunt capacitors, controlled
opening ensures that current interruption occurs with medium arcing times instead of minimum arcing time.
Therefore, a larger contact gap will be obtained and the restrike probability will decrease.
A highly stable operating time is needed to apply controlled switching for both closing and opening operations, with
a deviation less than ± 1ms. In controlled capacitor banks energisation, independent pole operated circuit breaker
is required, but a three phase simultaneous (gang) operated circuit breaker is normally applied to 145 kV ratings. In
addition, accurate switching controller is also necessary for satisfactory effect of controlled switching. To evaluate
capacitor bank switching capability in the tertiary side of UHV transformer, capacitive current switching tests have
been performed following the sequence shown in Table 7.2.1.4, which is the general procedure of capacitive
current switching tests in China. Test voltage, breaking current and voltage factor kc are set as the values for
145 kV isolated neutral capacitor banks in accordance with IEC 62271-100.
1 T60 preparation test

2 120 making inrush current test
3 120 breaking capacitive current test (BC2)
4 48 breaking 10~40% of rated capacitive current test (BC1)
5 Tests 2 and 3 in cycle
Table 7.2.1.4: Test procedure for capacitor bank switch

According to IEC 62271-100, preferred values of rated breaking current for 145 kV capacitor switch are shown in
Table 7.2.1.5, and do not exceed 400 A.
Line Cables Single capacitor bank Back-to-back capacitor bank

Rated Rated line- Rated cable- Rated breaking Making inrush
Rated breaking current Frequency of inrush
voltage charging charging current of back-to- current of back-to-
of single capacitor current
Ur breaking current breaking current back capacitor back capacitor
Isb (A rms) fbi (Hz)
(kV rms) I l (A rms) I c (A rms) Ibb (A rms) Ibi (kA peak)
145 50 160 400 400 20 4250
Table 7.2.1.5: Preferred values of rated capacitive switching current in IEC 62271-100
According to 4.111.7 of IEC 62271-100, for direct single-phase laboratory tests, the test voltage measured at the
circuit-breaker location immediately before the opening shall be not less than the product of U r / 3 and the
following capacitive voltage factor kc :
a) 1.0 for tests corresponding to normal service in solidly earthed neutral system
b) 1.4 for tests corresponding to
- breaking during normal service conditions in systems other than earthed neutral systems;
- breaking of capacitor banks with an isolated neutral point.
In the UHV AC demonstration project, the 110 kV reactive power compensation device is in a non-effectively
earthed neutral system. Therefore, the single-phase test voltage (r.m.s.) should be 1.4 145 / 3  117 kV The
power frequency test voltage and the DC voltage resulting from the trapped charge should be maintained at least
0.3 s after breaking.
Page 164
Single capacitor bank breaking is tested with synthetic circuits. The circuit used for current making is shown in
Figure 7.2.1.5. Inrush current is produced by an LC oscillating circuit as shown in Figure 7.2.1.6. A photo of the test
circuit is shown in Figure.7.2.1.7.
BDc Lc
Lc HKc FK CD Cv Rv
R0
R0
P3 P2 P1 T2
U T1 C0
C0 SP
P3 P1
SP
BDa La HKa
La
Note: T1:Current source transformer T2: Voltage source step-up transformer

BDa, BDc: Protection circuit breaker FK: Auxiliary switch
La, Lc: Current circuit reactor HKa, HKc: Closing switch
P1, P2, P3: Arrester SP: Tested device
R0, C0: Resistor and capacitor for current source frequency modulation
CD: Operating switch Cv-load capacitor Rv-damping resistor
Figure 7.2.1.5: Principle of current making by synthetic test circuit
Figure 7.2.1.6: Principle of inrush current test circuit

Note: Ch ， Lh —Capacitor and reactor in LC oscillating circuit S t —Tested CB
In order to satisfy the requirement of capacitor switching, four manufacturers (hereafter, manufacturer A, B, C and
D respectively) have developed special switchgears and repeated the test for nearly four years. In summary, only
one type of circuit breaker with controlled switching device has satisfactory completed 5000 times making and
breaking tests with required duties above. Detailed test results are described below.
The special load switch of manufacturer A has completed 5000 switching tests, but the switching current is only
1250 A and the inrush making current is 7.2 kA, which is different from the system calculated value. Restricted by
the test conditions, the recovery voltage is applied one millisecond after current zero. The test results are not
adequate for the system requirements.
Page 165
Figure 7.2.1.7: Test circuit of capacitive current switching
Manufacturers B and C initially adopted the circuit-breaker scheme according to the following test parameters:
capacitive breaking current of 1600 A, and capacitor bank inrush making current of 9.3 kA (see Figures 7.2.1.8 and
7.2.1.9). Breaking failed when tests were performed at the 1200th and 2000th operation respectively. Therefore, the
circuit breaker scheme with controlled switching was adopted by manufacturer B. Due to the dispersion of
operating times and misoperation by the synchronous controller, the maximum value of actual inrush making
current was 1.8 kA. The 5000 operations specified have been performed. The special load switch (with SF6 live
tank structure and three-phase spring operating device) was developed by manufacturer C without a synchronous
control technique. The expected electrical durability was 5000 times.
Manufacturer D has recently started the development of a special switchgear for capacitor bank switching adopting
a circuit-breaker with high electrical durability, which takes into account the switching capability of both short-circuit
current and capacitor banks. The expected electrical durability is 3000 operations without controlled switching and
5000 times with this technique.
Figure 7.2.1.10 and 7.2.1.11 show the contacts and nozzles of the switchgear without controlled switching device
after capacitor bank switching tests as examples. Heavy erosion can be observed especially on the top of arcing
contacts.
Figure 7.2.1.8: Test circuit breaker Figure 7.2.1.9: Test circuit breaker
without controlled switching with controlled switching
Page 166
Figure 7.2.1.10: Erosion of fixed contact
Figure 7.2.1.11: Erosion of nozzle
7.2.1.3 Switching phenomena in commissioning test

The commissioning test of the China UHV AC demonstration project was implemented in December, 2008. Among
all the tests, there were 7 items about switching phenomenon including different voltage level CBs and 1100 kV
DSs. Details are summarized as follows:
(1) Switching on/off no-load UHV transformer at 500 kV side
During the system commissioning test, switching on/off a no-load UHV transformer at its 500 kV side was
conducted five times. The switching overvoltage and inrush current were measured. The maximum phase-to-
ground overvoltage measured at the 1000 kV and 500 kV sides of the UHV transformer were 1.54 p.u. (1 p.u. =
1100  2 / 3 kV ) and 1.48 p.u. (1 p.u. = 550  2 / 3 kV ) respectively, which are below the limits of 1.6 p.u.
and 2.0 p.u. required in Chinese standards. The maximum inrush current was 4.82 kA. Figure 7.2.1.12 shows the
measured typical overvoltage and inrush current waveforms.
Page 167
Figure 7.2.1.12: Typical waveforms of overvoltage and inrush current during switching on no-load UHV
transformer at 500 kV side in Changzhi station
(2) Switching on/off no-load UHV AC transmission line
Through a test switching on and off a UHV AC no-load transmission line, the operation function of a UHV circuit
breaker can be checked, and the transient voltage and current of a surge arrester can be measured, which can
help define the characteristics of switching UHV lines. For the two UHV AC transmission lines, switching on/off no-
load line tests were conducted three times from each terminal. The maximum phase-to-ground switching
overvoltage measured on the busbar of the substation was 1.25 p.u. and the value at both terminals of the UHV
line was 1.26 p.u. with 1 p.u .  1100 2 / 3 kV . The value of overvoltage met the design criterion of 1.6 p.u. Figure
7.2.1.13 shows typical overvoltage waveforms measured.
(a) Closing (b) Opening

Figure 7.2.1.13: Typical overvoltage waveforms during energizing Nanyang-Jingmen UHV no-load line at
Nanyang substation
(3) Switching on/off 110 kV 240 MVAr reactors at the tertiary side of UHV transformers
There are two 110 kV reactors with rated capacity of 240 MVAr each and four 110 kV capacitors with a rated
capacity of 210 MVAr each on the tertiary of the UHV transformer.
Page 168
The transient overvoltage at the tertiary winding side of a UHV transformer and the impact on the steady state
voltage of the UHV system caused by 110 kV reactor switching can be verified in this test. Each 110 kV reactor
was switched on and off three times. The maximum phase-to-ground switching overvoltage was 166 kV, which is
significantly lower than the insulation level. Figure 7.2.1.14 shows typical overvoltage test waveforms.

Figure 7.2.1.14: Typical overvoltage waveforms during switching of 110 kV reactors
(4) Switching on/off capacitors at the tertiary side of UHV transformers
The transient overvoltage at the tertiary winding side of a UHV transformer and the impact on the steady state
voltage of the UHV system caused by 110 kV capacitor switching can be verified in this test. Each 110 kV capacitor
was switched on and off three times. The maximum phase-to-ground switching overvoltage is 171 kV, which is
significantly lower than the insulation level. Figure 7.2.1.15 shows some typical overvoltage test waveforms.

Figure 7.2.1.15: Typical overvoltage waveforms during switching capacitors at the tertiary side of UHV
transformers
(5) Interconnecting / Splitting of UHV systems
The interconnecting / splitting of UHV systems test is conducted by closing and opening the UHV line circuit
breakers at each side. The impact on power system stability, the switching overvoltage and steady voltage
fluctuation phenomenon caused by the interconnecting / splitting operation can be observed in this test. The
frequency and voltage differences of the two systems can also be observed, which is necessary for the setting of
synchronization devices. The maximum switching and temporary overvoltages of interconnecting / splitting
operation measured in this test are much lower than the design criterion.
(6) UHV disconnector switching bus transfer current
The bus transfer current switching capability by the 1100 kV disconnector was verified during the commissioning
test. The test was completed successfully.
(7) Artificial single-phase grounding
Page 169
An artificial single-phase grounding test on UHV lines was performed with two goals. One is to check the
performance of the UHV line protection relay and single-phase automatic re-closing; the other is to check the
secondary arc current limitation by the shunt reactor and neutral point reactor. The artificial single-phase grounding
fault tests were conducted on the Changzhi-Nanyang and Nanyang-Jingmen UHV transmission lines close to the
Nanyang switching station respectively. The grounding positions are shown in Figure 7.2.1.16.
Figure 7.2.1.16: Artificial single-phase grounding fault tests on Changzi-Nanyang-Jingmen UHV lines
In the Chinese UHV power system, single phase auto re-closing (SPAR) is adopted and the re-closing time is set
as 0.7~1.0 s. Four-legged shunt reactors are applied in UHV single/double-circuit transmission lines to limit the
secondary arc.
The secondary arc current in Changzhi-Nanyang-Jingmen UHV AC demonstration project has been measured in
this test. Table 7.2.1.6 shows the system conditions for the two tests.
Line Capacity of shunt reactor Impedance of neutral point Shunt

Test line Transmission
length (MVAr) reactor (Ω) Compensation
(single circuit line) Power (MW)
(km) Sending side Receiving side Sending side Receiving side degree (%)
Changzhi - Nanyang 600 358.5 946.5 710.5 275 365 82
Nanyang - Jingmen 1000 281.3 710.2 602.6 450 530 87
Table 7.2.1.6: System condition for artificial single-phase grounding test

Table 7.2.1.7 shows the field test results of primary short circuit current, secondary arc current, and recovery
voltage during the artificial single-phase grounding test of the Changzhi-Nanyang and Nanyang-Jingmen UHV AC
lines. Figure 7.2.1.17 shows the short circuit current and secondary arc current waveform of the test on the
Nanyang-Jingmen line.
The field test was successful. Results show that, four-legged reactors can effectively limit UHV lines secondary arc
current, and that the secondary arc extinction occurs within 0.12 s, which meets the requirement for single phase
auto-reclosing.
During this test, the maximum overvoltage on the shunt reactor was 1.35 p.u. (1213 kV), and the maximum
overvoltage on the neutral reactor was 290 kV, which meets the design criterion and equipment technical
specification.
Test line Primary short circuit current Secondary current Max Recovery voltage Secondary arc
(single circuit line) (kA peak) (kA rms) (A peak) (kV peak) extinction time (ms)
beating wave
14.5
Changzhi-Nanyang 10.4 4.6 24.5 for trough 118
(after fault 105ms)
145 for crest
beating wave
11.6
Nanyang-Jingmen 9.7 4.7 11.8 for trough 42
(after fault 79.3ms)
44.6 for crest
Table 7.2.1.7: Field test results of secondary arc current and recovery voltage during artificial
single-phase grounding test
Page 170
10
I (kA)
0
-5
-10
0 50 100 150 200 250
t (ms)
Figure 7.2.1.17: Short-circuit current and secondary arc current waveform of Changzhi-Nanyang UHV line
during artificial single-phase grounding test
7.2.2 Experience of EHV system in China

China has made a survey on the switching behavior of 550 kV and 800 kV switchgears during and after system
commissioning. Field test results of several switching scenarios are mentioned below, such as interrupting short-
circuit current, opening a no-load line, earthing switch (ES) interrupting and secondary arc extinction in
transmission lines.
7.2.2.1 Circuit breaker

There are 101 cases recorded in the field of CB (550 kV-800 kV) short-circuit current breaking operations. Table
7.2.2.1 gives the number of fault currents in four ranges: less than 10% Isc, 10% Isc – 30% Isc, 30% Isc – 60% Isc,
and more than 60% Isc. The number of occurrences for 550 kV circuit breakers is respectively 35, 52, 11 and 0,
accounting for 35.7%, 57.1%, 9.2% and 0% of all faults.
Figure 7.2.2.1 shows the breaking fault current distribution in the form of a pie chart.
During this investigation, the maximum TRV peak value recorded is approximately 597 kV (due to limitations by the
VT and fault recorder, there may be a significant measurement error), which is much lower than the specified value
in the circuit breaker standard. Measured results are shown in Table 7.2.2.2.
Rated voltage Number of faults with short circuit current breaking

Number of records
(kV) <10% ISC 10%~30% ISC 30%~60% ISC 60%~100% ISC
550 97 35 52 10 0
800 4 0 3 1 0
Note: ISC is the rated short circuit breaking current of circuit breaker.
Table 7.2.2.1: Statistics on faults with short-circuit current breaking
Page 171
Figure 7.2.2.1: Distribution of CB short-circuit breaking currents

Rated voltage Fault current TRV peak value Whether SC is Interrupting result
ISC (kA)
(kV) (kA) (kV) installed (success or failure)
500 50 14.0（28%） 503 NO Success
500 50 20.9（41.8%） 597 NO Success
Notes:
1) ISC: the rated short circuit breaking current of circuit breaker.
2) SC: series capacitor in transmission line.
3) TRV is calculated from the measured voltages at bus side and line side of CB. Due to limitations by the VT and fault
recorder, there may be a significant measurement error.
Table 7.2.2.2: TRV during CB breaking fault current
a) Typical example 1
Figure 7.2.2.2 shows the main electrical wiring of the Linhe 500 kV substation, which includes two transformers and
four out-going lines. During the fault breaking of the Lin-Peng line CB in Linhe substation, the maximum TRV peak
value reached was 597.4 kV, measured by the fault recorder on site. The voltage waveform is shown in Figure
7.2.2.3.
Page 172
Figure 7.2.2.2: Single line diagram of Linhe 500 kV substation
Figure 7.2.2.3: Voltage waveform by site recorder during Lin-Peng line CB breaking
Page 173
b) Typical example 2
Figure 7.2.2.4 shows the main electrical wiring of the Shibei 500 kV substation, which includes two transformers
and ten out-going lines. During interrupting the fault on the Bei-Qing line II, the maximum TRV peak value reached
was 503 kV, with an RRRV of 1.3 kV/μs and a breaking fault current of 14 kA. According to IEC 62271-100, the
TRV peak value for the T30 test of a 550 kV CB is 899 kV and the RRRV is 5 kV/μs. The TRV peak applied to
Shibei 550 kV CB is well below the value specified in the IEC standard. The phase to ground voltage, TRV and
breaking fault current waveform are shown in Figures 7.2.2.5 to 7.2.2.7.
Figure 7.2.2.4: Single line diagram of Shibei 500 kV substation
Figure 7.2.2.5 shows the fault current breaking operation by the 550 kV CB in the Shibei substation. In the figure,
UBUS represents the bus side voltage of the Bei-Qing line II CB, ULINE represents the line side voltage of Bei-Qing
line II CB, UBUS-LINE represents the voltage difference between the two sides of Bei-Qing line II CB or the TRV, and
ICB represents the current passing through Bei-Qing line II CB.
Figure 7.2.2.5: Voltage and current of Shibei 550 kV CB during short-circuit interruption
Figure 7.2.2.6 shows in more detail the current and voltages near the interruption instant at current zero. Figure
7.2.2.7 shows a further extension of Figure 7.2.2.6 with the addition of the two-parameter TRV specified in
IEC 62271-100 for test duty T30. The TRV in the Shibei 500 kV substation is covered by the standard TRV even
Page 174
though the TRV record may include a significant measurement error due to limitations by the VT saturation and
fault recorder.
Figure 7.2.2.6: Voltages and current of Shibei 550 kV CB near current zero
Figure 7.2.2.7: TRV (in red) and voltages on both sides of Shibei 550 kV CB near current zero –
Comparison with IEC standard TRV for T30
During this investigation eighteen field test records were made of line-charging current switching by 550 kV circuit
breakers (the length of the lines is between 55 km and 170 km), seventeen were successful and one failed (due to
leakage of the closing resistor tank that lead to a decrease of the internal insulation strength and closing resistor
explosion).
No record of out-of-phase breaking was made during this investigation.
7.2.2.2 Earthing switches

Three field test records were made of induced current switching by 550 kV AIS earthing switches. As shown in
Table 7.2.2.3, all of them were successful. The maximum electrostatically induced voltage measured in the field
test was 109 kV, occurring on the Xin’an-Liaocheng 500 kV double circuit lines, with shunt reactors at one line side.
The electrostatically induced current simulation value is 26.8 A, and electromagnetically induced current simulation
value is 56 A. As the induced current and voltage are much higher than the standard values, they are specified for
the double circuit line project.
Page 175
Rated voltage Electrostatically induced Electrostatically induced Electromagnetically Electromagnetically

(kV) current (A) voltage (kV) induced current (A) induced voltage (kV)
4.7 56.0
500 109（measured） —
(simulation) (simulation)
26.8
500 50.7（measured） — —
(simulation)
500 — 54.5（measured） — —
Table 7.2.2.3: Induced current and voltage of ES during field tests

7.2.2.3 Secondary arc extinction field tests
Secondary arc extinction is necessary for single phase auto re-closing. In China, high voltage shunt reactors with
neutral reactor is the main countermeasure for secondary arc extinction in EHV/UHV systems with single/double-
circuit transmission lines. In case of lines with series capacitors, bypassing the series capacitor in the faulted phase
is required to eliminate its impact on the secondary arc current. Its effectiveness has been proven by field tests.
a) Typical example 1
Table 7.2.2.4 shows the field test results of secondary arc current extinction during the single-phase artificial
grounding fault of the 500 kV Yantan-Shatang line. The single-phase re-closing time is set as 1.10 s. The two field
tests were successful, which confirmed that the single-phase re-closing time is feasible.
Figure 7.2.2.8 shows the secondary arc current and recovery voltage test waveforms.
Primary short circuit current Secondary arc current of different time span Recovery
Peak value rms value Time Current (A) Time voltage
(A) (Arms) (s) b-c c-d d-e (s) (peak/kV)
Case 1 6872 3041 0.054 82 6.3 26.5 0.0667 18.6
Case 2 5920 2680 0.082 16 — — 0.0305 14.7
Table 7.2.2.4: Field test results of single-phase artificial grounding fault

on Yan-Sha 500 kV line
Figure 7.2.2.8: Secondary arc current and recovery voltage test results of single-phase artificial
grounding fault on Yan-Sha 500 kV line
Page 176
b) Typical example 2
Table 7.2.2.5 shows the field test results of secondary arc current extinction during the single-phase artificial
grounding fault test of the 750 kV Guanting-Lanzhoudong line. The single-phase re-closing time was set as 0.6 s.
The successful field test proved the feasibility of the single-phase re-closing time
Figures 7.2.2.9 and 7.2.2.10 show the secondary arc current and recovery voltage test waveforms.
Primary short circuit current

Secondary arc current
Recovery
Measurement Peak Rms Duration
Tested result of voltage
position value value time Tested value (A Tested secondary
the moment of (peak/kV)
(kA) (kA) (s) rms) arcing time(s)
Reignition (s)
beating wave
Guan Ting 5.66 3.80 10.5 0.038 0.151 33 for trough
192 for crest
0.039
beating wave
Lanzhou dong 3.38 2.29 — — — 42 for trough
200 for crest
Table 7.2.2.5: Field test results of single-phase artificial grounding fault on

Guan-Lan 750 kV line
Figure 7.2.2.9: Secondary arc current waveform of Guan-Lan 750 kV transmission line
Figure 7.2.2.10: Recovery voltage test result of Guan-Lan 750 kV transmission line
Page 177
7.3 Experience of UHV/EHV system in Japan

Some field test experience of circuit breakers (CB), earthing switches (ES), disconnectors (DS) and HSES in Japan
are provided in the following.
7.3.1 Circuit breakers

Several statistical analyses have shown actual short-circuit current in Japanese EHV systems. Table 7.3.1.1 shows
the latest data collected from 2006 to 2010 for 550 kV.
Based on the oscillograms mentioned above, fault clearing times of 27 records can be derived. In addition, the fault
clearing time below 40 ms accounts for 30% with a minimum value of 34 ms, the clearing time between
40 ms~50 ms accounts for 44%, and the clearing time above 50 ms accounts for 26% with a maximum value of
59 ms. The DC time constant of the fault currents recorded in the oscillograms is within the standard value of 45 ms.
Due to the special system wide protection method used, Japan has no experience with out-of-phase condition
where the phase angle difference reaches 180 degrees and the network is going to be separated [112] [113].
Probability of short-circuit fault current interruption

Rated voltage (kV) Number of records
<10% ISC 10%~30% ISC 30%~60% ISC 60%~100% ISC
51 37 % 45 % 18 % 0%
550
(21) (0 %) (29 %) (57 %) (14 %)
Note:
1) ISC: the rated short circuit breaking current of circuit breaker. (Isc= 63kA for 550 kV CB)
2) Upper low values are for short circuit breaking currents of CBs at both ends of line.
3) Lower low values in brackets show fault currents at the fault point calculated with those of both line ends.
Table 7.3.1.1: Probability of short-circuit fault current interruption
7.3.2 Earthing switches

The specified values for 550 kV and 1100 kV ES of GIS type are listed in Table 7.3.2.1.
Rated voltage Electrostatic induction Electrostatic induction Electromagnetic Electromagnetic

(kV) current (A) voltage (kV) induction current (A) induction voltage (kV)
550 40 50 1000 70
1100 40 50 1000 70
Table 7.3.2.1: Rated induced current and voltage switching parameters of ES
7.3.3 Disconnectors: bus charging and bus transfer current

For Japan 1100 kV GIS substation, since the load side capacitance is 2000 pF, the bus charging current passing
through the DS is about 0.5 A.
Though no actual data showing bus-transfer current is available, distribution in bay current may give reference to
analyze bus-transfer current. The bay current data were derived on the third Wednesdays of each month, from
April 2008 to March 2009 (one year), including eleven 550 kV substations (60 bays, 40 line bays and 20
transformer bays) and twenty-three 275 kV substations (167 bays, 93 line bays and 74 transformer bays). The
rated current is 8000 A, 6000 A and 4000 A for the different feeder bays and 2000 A for transformer bays.
Figures 7.3.3.1 and 7.3.3.2 show the cumulative distribution of the ratio of the maximum load current to the rated
current of the bay at a typical moment. The maximum ratio of the maximum load current to the rated current of the
550 kV bay was 82.5% and the 90 percentile cumulative probability value of the ratio was 52 %. As for the 275 kV
bay, the maximum ratio was 65.6% and the 90 % percentile cumulative probability value of the ratio was 48 %. It
should be noted that since the data was collected only on the third Wednesday of each month, it may not represent
the real maximum current ratio of the year.
Page 178
It should be noted that the load currents distributed along the different loops can contribute to the bus-transfer
current so the bus-transfer current is the summation of these distributed load currents, depending on the open and
closed conditions of the bus DS and their currents as described in section 6.1. Therefore, the bus-transfer current
will be higher than the load current of each bay.
Figure 7.3.3.1: Cumulative distribution of the ratio of the maximum load current to the rated current of
the bay on typical weekdays (550 kV)
Page 179
Figure 7.3.3.2: Cumulative distribution of the ratio of the maximum load current to the rated current of
the bay on typical weekdays (275 kV)
7.3.4 VFTO
Japan has carried out measurements of 550 kV and 1100 kV VFTO. Test results are shown in Table 7.3.4.1.
Rated voltage Amplitude of VFTO Type of disconnector Whether or not switching

Pre-charged DC voltage
(kV) （p.u.）（fast or slow） resistor installed
1100 kV 1.21 fast With resistor Approximately max. 1.0 pu
550 kV 2.72 fast Without resistor Approximately max. 1.0 pu
Note:
1) 1100 kV data is measured at UHV test site in Japan, which are reported in CIGRE TB 456.
2) 550 kV data is measured at a full GIS substation. 2.7p.u. is the corrected maximum value obtained from the measured
oscillation amplitude by assuming the restriking occurred at +1.0 p.u. on one side of the DS and –1.0 p.u. on the other side
Table 7.3.4.1: 550 kV and 1100 kV VFTO measurement results
Figure 7.3.4.1 shows the distribution of breakdown voltage across contacts obtained during 119 opening operations
of a 550 kV DS.
Page 180
Figure 7.3.4.1: Distribution of voltages across 550 kV DS contacts at opening operation

Note: The maximum value of 1.86 p.u. is the maximum breakdown voltage across DS measured during random operations. The value of 2.7 p.u.
shown in Table 7.3.4.1 is obtained by extrapolation of measured values with the assumption that the residual voltage is 1 p.u. on one side of the
DS and -1 p.u. on the other side.
7.3.5 Secondary arc extinction

Several field measurements on secondary arcs of 550 kV OHL have been carried out in Japan. Twenty-seven
faults which have records of the line side voltages were examined. With respect to the type of faults, twenty five are
1LG and two are 2LG (1LG + 1LG, i.e. 1LG in each circuit of double-circuit OHL). One of the 2LG faults resulted in
unsuccessful reclosing because the secondary arc remained over the reclosing time of 1.0s.
Figure 7.3.5.1 shows an example of an oscillogram analysed in case of a 1LG fault. From the line side phase
voltage during the dead-time for re-closing, secondary arc extinction time and recovery voltage on the faulted
phase were derived. The recovery voltage was evaluated by the induced voltage at a relatively stable state after
secondary arc extinction.
Figure 7.3.5.1: Oscillogram 1 of line side phase voltage during dead-time of reclosing
(1LG; Line length is 138 km; Primary fault current is 23 kA)
Figure 7.3.5.2 shows an oscillogram of an unsuccessful reclosing. In this case, line length is relatively long and
recovery voltage is high due to 2LG (1LG + 1LG). As for the secondary arc voltage of the unsuccessful reclosing
phase, a stable repeated waveform can be observed for more than 10 cycles just before reclosing. It is thought that
this phenomenon is caused by the short-circuiting of a stretched arc path.
Figure 7.3.5.2: Oscillogram 2 of line phase voltage during dead-time for reclosing
(1LG +1LG; Line length is 195 km; Primary fault current is 10 kA)
Page 181
There were two special cases with multiple reignition of the secondary arc due to multiple lightning surge intrusions
produced by direct lightning strokes to a phase conductor. Figure 7.3.5.3 shows one oscillogram of these cases.
Figure 7.3.5.2: Oscillogram 3 of line side phase voltage during dead-time for reclosing
(1LG; Line length is 59 km; Primary fault current is 37 kA)
Note: In this case, secondary arc extinction time was evaluated as 77 ms, which was the first extinguishing time.
Field data obtained at 550 kV have shown that secondary arc currents obtained by simulation are in line with field
data [114].
From the field data, it can be concluded that the secondary arc extinction time becomes longer as the secondary
arc current and the recovery voltage become larger. As a secondary factor, wind velocity influences the extinction
time to some extent. The extinction time tends to be shorter as the wind velocity becomes higher.
7.3.6 HSES
To verify the reliability of the reclosing system with HSES, field tests were carried out from 1993 to 2000 on a
275 kV OHL (126 km) using a circuit breaker as a HSES. Thirteen cases of line faults caused by lightning were
recorded, during which the sequence of the re-closing system and total system reliability were verified. Among the
thirteen recordings, there were nine 1LG faults and one 2LG fault during which the HSES operated correctly. The
other three cases were with 4LG faults on double circuit lines and the HSES did not operate correctly.
Figure 7.3.6.1 shows the time chart of a successful single-phase reclosing after 1LG on one circuit of a 126 km
th
275 kV double circuit line, which occurred on May 14 , 1997.
Figure 7.3.6.1: Time chart of a successful single-phase reclosing after 1LG on one circuit of a 275 kV
double circuit line (1LG; Line length is 126 km)
Page 182
7.4 Experience of UHV/EHV system in Russia

Russia has built 750 kV EHV and 1150 kV UHV AC transmission lines in 1967 and 1985 respectively. The
maximum operating voltage for 750 kV lines is 787 kV, and the maximum operating voltage for 1150 kV lines is
1200 kV.
7.4.1 Tests of 800 kV and 1200 kV circuit breakers

During the development of UHV air-blast circuit-breakers, Russian manufacturers were guided by Russian National
Standards that were essentially based on IEC recommendations. Particularly, these standards permitted tests to be
performed on a module of the circuit-breaker (800 kV air blast CB contains 3 modules; 1200 kV CB contains 5
modules of the same type with pre-inserted resistors being used only in the 1200 kV circuit breakers). The nominal
short-circuit current for circuit-breakers is 40 kA.
The modules of air blast circuit-breakers successfully withstood the following tests (data below show parameters
required by the standards for the whole pole, real module loads, being recalculated per pole, were somewhat
higher), and are summarized in Table 7.4.1.1.
Rated voltage Rated breaking First reference TRV peak RRRV

Test Mode
（kV） current（kA） voltage (kV) Value (kV) (kV/μs)
1512
40 T100 2.8
(1.54pu)
1512
40 50% dc component 2.8
(1.54pu)
SLF 531 1372
30
(L75) (8.1kV/μs) (1.40pu)
1890
1200 12 T30 11.2
(1.93pu)
2840
10 OP1-OP2 2.25
(2.90pu)
1
Capacitive current - - -
/50Hz
0.6
Shunt reactor current 1760
/0.715~0.8kHz
Table 7.4.1.1: Tests duties for 1200 kV circuit breaker
7.4.2 Field Tests on 750 kV experimental-commercial line

As EHV transmission plays an extremely important role in power supply, at the beginning of construction of the
750 kV grid it was decided that a 87 km long 750 kV experimental-commercial line with two substations was to be
erected between Moscow and Konakovo, in order to:
(1) Test various newly developed 750 kV equipment in full scale and under real conditions;
(2) Obtain experience in maintenance and operation, and improve equipment if necessary;
(3) Test the design and maintenance of lines in full scale and make improvement if necessary.
Test programs were developed and executed during several years by joint teams of researchers from leading
scientific research institutions of the Ministry of Electrotechnical Industry (VEI) and Ministry of Energy &
Electrification (VNIIE) with active participation of the Long Transmission Company (PODEP) who operated this
transmission line as well as some planned 750 kV transmission lines and all 1150 kV transmission lines.
Hundreds of commutations were performed, including switching-on and off 750 kV lines, transformers and shunt
reactors, short-circuits tests (including secondary arc measurements with proper oscillograph recordings and some
with speed filmings) as well as one and two phase commutations of the line terminated with a transformer (such
schemes led to extremely high temporary overvoltages). Some improvements resulting from these tests were made
in the design of 750 kV equipment and lines.
Page 183
It was decided that similar tests, but limited in numbers, would be conducted by the same joint team on every newly
constructed 750 kV transmission line before putting it into service.
The success of full-scale testing on the 750 kV experimental transmission line and commercial 750 kV lines led to
the decision that before constructing 1150 kV lines, test installations for full-scale 1150 kV equipment testing
should be developed and then the program of commissioning tests to newly constructed 1150 kV lines should be
widened. Two 1150 kV test installations were built, one at the 750 kV Bely Rast substation, and the other at
Togliatti City (this installation belonged to the manufacturers and was served by the VEI team).
It is important to underscore that field full-scale EHV/UHV equipment tests, being a very important source for
verifying newly developed equipment, cannot in some cases reproduce extreme parameters requested from
manufacturers, such as the rated short-circuit current or asymmetrical short-circuit current. Therefore, when
standardizing circuit-breaker TRV, field test results have to be used mainly to check calculated estimates, rather
than being directly used in standards.
7.4.3 Field Tests on VFTO

In the Russian operating experience, the following failures could be attributed to VFTO, which led to the necessity
to investigate VFTO.
(1) At least one case of 787 kV autotransformer failure at DS operation;
(2) Gapped arrester blowouts, especially at 126-252 kV substations;
(3) Flashovers to ground across 525 kV and 787 kV busbar external insulation;
(4) A flashover across the longitudinal insulation of a de-energized 525 kV air-blast circuit breaker;
(5) Some cases of distant current transformer failures.
7.4.4 Experience of secondary arc extinctions

The analysis of operational statistics [68] for Russian EHV transmission lines showed that single- pole faults
account for 60% (for 330 kV) to 98% (for 787 kV) of all faults. Therefore SPAR is the main tool to restore power
supply in Russian EHV transmissions and TPAR is used only as a second line of defense. Successful SPARs
amount to 60-70% of all SPAR cases which is somewhat less than in US EHV lines.
Based on many field SPAR tests and automated operational recordings in Russian EHV transmission lines of 363-
525-787-1200 kV, the following main conclusions were derived.
1 The secondary arc, even if its current is very small, cannot extinguish until the fault current arc column loses its
high initial conductivity. If shunt reactors are connected to the OH-line, the magnetic energy stored in the SR of the
failed phase will be discharged into the secondary arc, and the attenuating transient discharge current of several
hundred amperes may delay the extinction of the secondary arc. It might be noted that in the case of TPAR, SR in
healthy phases also create attenuating current through the single-pole arc, but with a resonance frequency close to
the power frequency, due to the capacitance coupling of the failed phase with healthy ones. Then, at a high degree
of line capacitance compensation with inductance of the SR, the voltage recovering on a failed phase after arc
extinction at SPAR can reach 1.3-1.5 of normal operational phase voltage, depending on the line design and being
limited by corona losses. In lines without SR, it is much lower, (0.1-0.2 p.u.), but rises much faster (see the Annex).
Secondary arc extinction time depends on many parameters, mainly the magnitude of the secondary arc current
but also on current in the primary arc and its duration, arc length and position, wind velocity, presence of shunt
reactors, recovery voltage, etc. It can be lower than 0.1 s in very short lines (with secondary current about 5-10 A
crest) without shunt reactors or greater than 2.5-3.0 s with secondary arc “steady-state” current exceeding 90-
100 A crest (see Figure 7.4.4.1). For the SPAR dead time it is necessary to add to a statistically determined time of
extinguishing of secondary arc current, (with a probability of about 90% preferred), and some time for de-ionization
of the arc channel and restoring its dielectric strength to withstand the full operating voltage and transient
overvoltage at re-closing. In Russia, estimated extinction times were calculated statistically from EHV field tests
Page 184
and operational recordings. As the dispersion of such times is large and executing numerous tests in operating
transmission lines is difficult, all available data (over 700 results) from lines 363 to 787 kV, with or without SR, were
combined, after proper statistical checking, as a function of the most significant influencing parameter- secondary
arc current magnitude. To the 90% extinguishing time determined above, an additional time for restoring dielectric
strength was added, taken cautiously from TPAR tests and records, of approximately 0.3 s [68]. Because 0.3 s was
estimated for arcs with large short-circuit currents, much larger than secondary-arc currents, such an estimate
seemed sufficiently conservative. It has also been noted that in 1000-1200 kV lines, with longer arcs a stronger
non-uniformity in the distribution of voltage across the arc channel is expected, which will lead to faster
extinguishing times compared with 363-787 kV lines. This was observed in a limited series of tests on a 1200 kV
line, which is also included in Figure 7.4.4.1
Figure 7.4.4.1: Secondary arc extinction time and reclosure time for SPAR as function of secondary steady-
state currents (peak value) in case of OHL 330-550 kV.
2. Secondary arc measurements, with oscillograph recordings and speed filming, have shown that the secondary
arc column is not uniform along its length. It contains less bright spots between very bright large zones, and in time,
these less bright spots increase in their relative length. Therefore, the extinction of an AC secondary arc does not
correspond to the mechanism of DC arc extinction which is achieved purely through elongation of the arc column
that leads to an increase in its resistance and current interruption. In a DC arc the extinguishing is reached once-
when the current power loss exceeds power supply. In AC arc we observed several, and sometimes numerous,
attempts of extinguishing. At such unsuccessful attempts, a non-uniform voltage distribution along the arc column
increases the voltage over less conductive parts of the column thus temporarily restoring current flow.
3. Of course, fast elongation of the secondary arc column, especially under strong wind conditions, reduces the
current through the arc and shortens the time to extinguishing. But short-circuiting of parts of the column, does not
play an essential part in the arc restoration and cannot explain numerous extinguishing attempts through such
short time intervals as 0.01-0.05 s.
4. The higher the line voltage the longer the secondary arc and the stronger its non-uniformity. The impression was
that, with approximately the same secondary arc current, the extinguishing of arc on 787 kV lines happened faster
than on 363 kV and 525 kV lines, and on a 1200 kV line even faster. Unfortunately, there are not enough statistical
data to confirm this.
Page 185
5. Russian experts tried initially to estimate statistics of extinguishing time versus the magnitude of the secondary
current separately for three groups of EHV lines: without shunt reactors (restoring voltage is small, but rises fast),
with reactors where restoring voltage is between 0.5 and 1.5 p.u and with 4-legged reactors where the restoring
voltage is both small (less than 0.5 p.u.) and rises relatively slowly. They found that the influence or restoring
voltage was statistically insufficient compared with the influence of the secondary arc magnitude. Therefore, they
combined all data presenting them versus one parameter only: current magnitude. In total 725 experimental points
are summarized in the CIGRE Session Report N°34-207 of 1990 [19]. It also contains data on successful 3- and 1-
phase re-closings in EHV lines. That approach allowed the estimation of dead times with a probability of
extinguishing approximately equal to 90%. Such an approach somewhat overestimated the dead-time necessary
for 1200 kV lines, and the limited tests on such line confirmed this.
7.5 Experience of EHV system in Canada

7.5.1 Hydro-Quebec EHV system
Hydro-Quebec EHV power system is a complex network requiring the use of most advanced technologies specific
to large-scale transmission systems. In the early 90's, Hydro-Quebec decided to install thirty-two series
compensation banks, for a total of 11 200 MVAr in its 735 kV transmission system [109]. This decision triggered a
major simulator study related to the TRV stresses of series compensated 735 kV line circuit-breakers. Results have
shown that, by using surge arresters at both line ends, standard TRV values specified by IEC 62271-100 for
800 kV circuit breakers were not sufficient to cover all calculated values of TRV peaks.
Countermeasures were studied to solve this problem. This included the use of a fast protective device by-passing
the series capacitors following a fault. It was judged at that time however that the device was not sufficiently
reliable to be used on a large scale on the 735 kV system. Since 148 line circuit-breakers were subjected to the
issue of TRV exceeding IEC specified values, a statistical approach based on the failure probability was adopted
[110]. This study was based on the combined probability of fault occurrences on the 735 kV network in the previous
40 years (see Table 7.5.1.1) and the probability of exceeding the IEC TRV values.
Type of fault Annual occurrences No. of faults per 100 km-line per year
1LG 38.3 0.33953
2LG 1.43 0.01267
2L 2.21 0.01959
3LG 0.48 0.00425
3L No occurrence
Table 7.5.1.1: Annual fault occurrences on Hydro Quebec 735 kV network [24]
By accepting a MTBF of one in 400 years for line circuit-breakers, eleven existing circuit-breakers were replaced by
those having a higher TRV capability in mid 1990's. For the new 40 kA series compensated line circuit-breakers,
TRV crest values of 1920 kV (3.2 p.u.) and 2040 kV (3.4 p.u.) were specified for breaking currents of 40 kA and
15 kA respectively. Very recently, since an upgraded fast protective device was qualified to have a high degree of
reliability, series compensated line circuit-breakers will be specified with standard TRV values.
7.5.2 BC Hydro EHV system

The BC Hydro 500 kV system came into being in the late 1960s with the development of hydroelectric generation
about 1000 km northeast of Vancouver. The transmission line lengths between intermediate substations were in
the range 200 to 300 km and were series and shunt compensated. The line insulation was designed for 2 p.u.
switching surge achieved through the use of closing resistors on air blast circuit breakers with six or eight
interrupters in series. The substation LIWL was 1800 kV and the gapped surge arresters in use were rated 396 or
420 kV.
The advent of SF6 circuit breakers and metal oxide surge arresters and controllers brought about many changes.
Lines built after 1990 were designed to 1.7 p.u. switching surge, this being by the use of 372 kV line terminal
Page 186
arresters and controllers to achieve high-speed auto-reclosing at minimal voltage across the circuit breaker.
Closing resistors are thereby eliminated and the circuit breakers now have two interrupters in series for the
unchanged fault current rating. The substation LIWL is now 1550 kV and 396 kV rated surge arresters are used
except at generating stations where the rating is 420 kV because of the potential higher TOVs. High-speed single
pole auto-reclosing is used on all lines and secondary arc extinction is achieved though the use of the common 4-
legged reactor scheme.
The 500 kV lines are all mid-line 50% series compensated. The earliest series capacitor banks had simple gapped
protection only and minimum oil type bypass breakers. The first transition was the use of air blast bypass breakers
and later to MOV protected banks with SF6 bypass breakers and non-linear damping circuits including retrofitting of
all the earlier banks with the latest technology.
It is BC Hydro practice to perform staged fault testing on all new installations and the most recent such test
provided some insights into secondary arcs and their extinction. The secondary arc current was measured using
optical CTs which showed that the current has two modes. The first is the quasi 'steady state' mode, almost
triangular shaped current half-cycle loops. The second mode is the 'restriking mode' leading to extinction. The
sooner the latter mode is initiated, the shorter the extinction time. The proposed theory for this behaviour is that the
steady state arc is maintained by the power input from the source which of course is wirelessly connected. As the
arc elongates, the source cannot provide the power input necessary to maintain the steady state mode and
restriking is initiated. The arc is now very complex, very non-linear and no doubt involving space charges and
ultimately extinction occurs. None of the measurements were during windy conditions but the effect of wind is to
elongate the arc hastening the steady state to restriking transition.
7.6 Experience with Single Phase Switching in USA

Application of single phase switching on EHV and future UHV transmission lines might be required to maintain
system stability following single phase to ground faults. Achievement of successful high speed single phase
reclosing on long EHV and UHV lines usually requires special means and methods to ensure timely extinction of
the secondary arc. Some of these methods, namely the symmetrical and asymmetrical utilization of neutral shunt
reactors were tested and applied on the AEP 765 kV system [63] [64] [69] [70].
Single phase switching tests were performed on two untransposed 765 kV lines: Kammer – Marysville, 243 km long,
equipped with two 300 MVAr shunt reactor banks and the Rockport – Jefferson line, 177 km long, with two
150 MVAr shunt reactor banks. All shunt reactors were equipped with neutral reactors: 800 Ω and 300 Ω
correspondingly at Kammer and Marysville and 2000 Ω at Rockport and Jefferson.
A total of 44 single phase to ground fault tests were executed on the mentioned above lines. The calculated 60 Hz
secondary arc currents for these tests varied from a very small value of 5 A rms to 86 A rms. These secondary arc
current values versus arc extinction time are represented in Figure 7.6.1.
The variation of the secondary arc current value during the tests was achieved by using different taps on the
neutral reactors as well as applying the fault on outer and middle phases.
As shown in Figure 7.6.1, the secondary arc current in the majority of tests was lower than 40 A rms. Single phase
reclosings in these tests were successful at 0.5 second while the secondary arc extinguished in less than 0.25
seconds. Four single phase reclosings with the 0.5 second reclosing time were unsuccessful in tests with the
secondary arc current between 44 A rms and 80 A rms. After the reclosing time was increased to about 2 seconds
three successful single phase reclosings were recorded for the secondary arc current from 51 A rms to 58 A rms.
Finally two unsuccessful reclosings took place at the reclosing time close to 2 seconds when the secondary arc
current was equal to 86 A rms.
Page 187
Figure 7.6.1: Secondary arc extinction time versus secondary arc current
Successful single phase reclosings
Failed single phase reclosings
In addition to the secondary arc current value, the secondary arc extinction time depends on a number of
parameters, such as the faulted phase recovery voltage (especially its rate of rise across the extinguished
secondary arc channel), the primary fault current, the location and configuration of the air gap, the meteorological
conditions, etc.
Secondary arc interruptions and restrikes took place in about a third of the tests and represented a unique
opportunity for determining the withstand voltage characteristic of the air gap which was ionized by the fault and
secondary arc currents. The air gap restrike voltages during these tests are presented in Figure 7.6.2 as a function
of time from the secondary arc current interruption to the moment of gap restrike. Peak values of the faulted phase
recovery voltages which did not reach the gap flashover levels are also presented in this figure. The importance of
the air gap withstand voltage capability as a function of time is evident. Indeed, since the rate of rise of the faulted
phase recovery voltage for the reactor compensated transmission lines is easily calculable the air gap withstand
voltage characteristic represents an important parameter in determining the success of the single phase reclosing.
Figure 7.6.2: Air gap (4.2 m) rate of rise withstand voltage characteristic from the moment of the
secondary arc interruption:
Withstand voltages
Flashover voltages
Page 188
7.7 Conclusions on switching experience during and after system commissioning

(1) The information on EHV and UHV field experience about switching behavior is very limited. Investigation results
from China, Japan, Russia, USA and Canada, that are very important to better understand the equipment
performance, are summarized in this chapter. It includes TRV, fault interruption, shunt capacitor making and
breaking, bus transfer current breaking, induced current breaking, secondary arc current, VFTO.
(2) Concerning EHV or UHV switchgear, there is no evidence in the investigations performed in these countries of
CB, DS and ES failures due to overstress by TRV during short-circuit, bus transfer current or induced current
interruption.
(3) During system commissioning of the China UHV AC demonstration project, switching operations have been
performed successfully, include switching of 1000 kV unloaded transformer, 110 kV reactors and capacitors at the
tertiary side, and 1100 kV unloaded lines. Switching characteristics were verified comprehensively in the field.
System overvoltages and over-currents have enough margin compared with the design values, which ensures the
safe operation of the project and shows the proper withstand requirements for the specified equipment.
(4) Switchgear is used at the UHV transformer’s tertiary side for frequent capacitor bank switching in Chinese UHV
system due to power system needs for steady state voltage profile control and power transfer. Such frequent
switching associated with high inrush making currents can lead to fast ablation of CB contacts and nozzle.
Therefore a high electrical endurance is required for capacitor bank switching devices. It has been proved that a
circuit breaker with controlled switching could pass the 5000 operations durability test. Field operating experience
still needs to be gained. Development of circuit breakers without controlled switching is difficult for this type of
application, and research is in progress.
(5) Performance of 1100 kV DS switching bus transfer current was verified during commissioning of the China UHV
AC demonstration project, based on a requirement of 1600 A under 400 V. Japan collected some field data of
550 kV and 275 kV bay current to assist studying the bus transfer current.
(6) The use of shunt reactors with neutral reactor as the countermeasure against secondary current and recovery
voltage is widely applied. Its effectiveness was proved in several countries covering voltage levels from 800 kV to
1200 kV. In the China UHV AC demonstration project, the secondary arc extinction time was less than 0.12 s.
(8) Japan carried out several tests on a 126 km 275 kV line to study the performance of HSES for secondary arc
extinction. Results showed that, for 1LG or 2LG on double circuit lines HSES operated correctly, and the secondary
arc was extinguished within 50 ms.
(9) Due to the residual voltage of series capacitors (SC), the TRV of series compensated lines during CB breaking
of line fault is significantly higher than that without SC, and may exceed the value in the IEC standard. Hydro-
Quebec has evaluated the risk of TRV during fault interruption and replaced some existing circuit breakers in series
compensated lines with some having a higher TRV withstand capability. In the near future, with the application of a
reliable fast protective device by-passing the series capacitors during faults, standard TRV values will be specified
by Hydro-Quebec for series-compensated line circuit-breakers.
Page 189
8 Conclusions
Following on from the work done by WG A3.22 that is published in Technical Brochures 362 and 456, WG A3.28
studied more in depth specific aspects of EHV and UHV switching phenomena.
A worldwide inventory of systems with rated voltages 800 kV and above showed a large variety of system design
criteria and technical solutions such as single and double circuit overhead lines, secondary arc extinction solutions,
series and shunt compensation, transposition, reclosing policies. The influence of such solutions on the TRV
waveshape and amplitude has been studied by means of benchmark models.
Switching conditions for circuit breakers, disconnectors, earthing switches and HSES (high-speed earthing
switches) were studied through simulations performed on a model network and compared with service experience
from existing EHV and UHV systems and from early UHV projects.
The main findings and recommendations from WG A3.28 are summarized hereafter for the different types of
switching devices.
1. Circuit breaker
The recommendations from WG A3.22 are shown to be appropriate for UHV applications and results of simulations
confirm that TRV requirements for EHV and UHV circuit breakers are properly covered in the revised IEC and IEEE
standards except in particular cases of TLF with negligible capacitance between the circuit breaker and the
transformer, and in applications of compensated lines with series capacitors not by-passed when the circuit breaker
is interrupting a line fault.
In the case of terminal faults and long line faults, it has been found that the following factors have a negligible
influence on TRV: presence of shunt reactors, line transposition, line height, line sag, earth resistivity. The
influence of power flow is complicated but the effect of different phenomena cancel out each other so that the
overall influence is usually small.
It is relevant to note that technical characteristics like tower configuration, bundle geometry, transposition of phase
conductors, shunt and series compensation, when considered, have been adapted in the whole network model.
This means that by adapting one parameter many transient variables may be influenced, so that the impact of that
parameter on the values of such variables may become less salient than initially expected. For instance a higher
surge impedance coincides with a higher impedance per km of all OHL in the model. Consequently, the fault
current level will decrease, and the combined effects on the RRRV of the TRV regarding a higher surge impedance
and a lower short-circuit current cancel each other to a certain degree. Such a trend of reducing effects, when more
interactions among technical parameters and variables have to be considered, can be recognized within many
topics addressed in this Technical Brochure: e.g., (a) multiple resonances in transformers’ admittance (in the
frequency domain) tend to decrease the amplitude factor for TLF; (b) travelling waves and phase interaction reduce
the amplitude factor with unloaded line switching (with/without earth fault); (c) meshed networks show a lower
amplitude factor than radial networks, etc.
In applications with series capacitors, TRV is covered by standard values during faults with high short-circuit
current as met when the series-capacitors are bypassed. Severe TRV peaks are expected when series capacitors
are not bypassed, for example in case of low fault currents and a compensation degree of 40% and higher.
Mitigating measures exist to maintain the TRV within limits defined by standards e.g. by using fast protecting
devices.
Concerning TLF (Transformer Limited Faults), recommendations made in Chapter 4 are summarized here briefly:
- A first pole-to-clear factor of 1.3 (rated voltages of 100 kV up to and including 800 kV) or 1.2 (UHV) is regarded as
being at the safe side of actual values, excepted for networks with non-effective earthed neutrals (possible up to
and including 170 kV) for which 1.5 is required.
- The voltage drop across the transformer is a function of the ratio of the transformer impedance to the short-circuit
impedance of the substation (without the influence of the transformer itself). A fault current of, say, 6.3 kA will give
a voltage drop of roughly 90% while a fault current of 12.5 kA will show a voltage drop closer to 70%. Note that a
fault current of 12.5 kA corresponds in practice to approximately 30% of the circuit breaker’s rated short-circuit
current.
Page 190
- An amplitude factor of 1.7 is sufficient for the determination of the TRV peak value.
- As the TLF current is mainly dependent on the transformer characteristics and not on the circuit breaker rating,
fixed values for the fault current instead of a percentage of the rated short-circuit current are recommended.
A more precise understanding of switching phenomena was obtained through simulation models e.g. in case of
out-of-phase and capacitive switching. Based on simulations the peak TRV is covered by using a voltage factor of
1.2 in both cases with or without an earth fault, but a higher value (e.g. 1.4) is needed to cover the waveform after a
quarter cycle in case of an earth fault on another phase.
The condition leading to the highest TRV peak in case of long line faults was defined, taking into consideration the
influence of a parallel line circuit.
2. Disconnector
Simulations and field data showed that the limitation to 1600 A for the bus transfer current should be deleted in IEC
standard for EHV as was done already for UHV. The bus transfer current is generally the summation of distributed
currents from different bays. The limited statistical data provided in the report show that the 90 percentile of the
ratio between the maximum load current and the rating of a bay is between 0.5 and 0.8 (see Table 6.2.4.1). Further
statistical analysis is required to assess DS bus transfer current switching duties in EHV systems.
A bus charging current for GIS applications evaluation, based on substation layout, showed the need of 1 A
maximum for UHV. Values in IEC Std. 62271-102 are confirmed to be correct for EHV applications, as indicated
already in TB 456 [2].
3. Earthing switch
The study gave characteristics and tendency for electromagnetically and electrostatically induced current
interruption by earthing switches in 800 kV and 1100 kV systems with either double tower or 2 x single tower
configurations. Other than these parameters, requirements for earthing switches are greatly affected by line length
and the nominal current on the healthy line. These results will support the future revision of IEC Std. 62271-102.
4. High-speed earthing switch

The simulations performed by WG A3.28 supported the standardization work done in IEC that lead to a new
standard (IEC Std. 62271-112) for HSES [13].
5. Experience in service
Field data obtained at 550 kV have shown that secondary arc currents obtained by simulation are in line with field
data. Very short extinction times (far less than 1 s) of secondary arc currents were obtained from field tests in the
Chinese UHV network in applications with four-legged reactors.
In the UHV AC demonstration project of China, capacitors are switched on the tertiary of transformers to control
reactive power. The rated voltage of reactive power compensation capacitor banks is 110 kV. It will be a severe
duty for circuit breakers that have to operate two to four times per day, leading to a possible requirement of 5000
cycles of operation. For such applications, synchronized switching should be applied to reduce the electrical stress
on interrupting chambers and achieve the required number of switching operations.
In addition, CIGRE WG A3.28 would like to forward the following comments and recommendations to CIGRE.
Some switching conditions that are only partially addressed by WG A3.28 need further investigations. The following
topics should be studied more in detail in the future by CIGRE SC A3:
- New developments and application aspects of series and shunt compensation: impact of series compensation on
temporary/switching overvoltages and related restraints to the application of series compensation, impact on
Page 191
secondary arc extinction and consequentially requirements for involved equipment like four-legged shunt reactors
and/or HSES, impact of the new developments on TRV requirements for the line circuit breakers and whether
special recommendations for standardization are necessary, field experience with former and new applications of
series/shunt compensation means and its consequences for utilities policy.
- In applications with series capacitors, TRV are covered by standards when mitigations means are used. A
remaining question is whether or not it can be considered that these mitigations means are sufficiently reliable to
consider that standard circuit breakers can be used for these applications.
- More field data should be collected e.g. on VFTO measurements in case of GIS and other switching phenomena
(TRV for TLF, bus transfer current for DS, induced currents for ES) especially for UHV.
- Special attention is requested for the collection of statistical information on transformer natural frequencies to give
recommendations for the standardization of the time to TRV peak for TLF.
- No recommendation could be made for out-of-phase switching due to insufficient field data.
Page 192
Bibliography/References
[1] CIGRÉ Technical Brochure 362, December 2008, “Technical Requirements for Substation Equipment
exceeding 800 kV AC”, by CIGRE WG A3.22: H. Ito (convenor); A.L.J. Janssen (secretary); J.Amon, S.-W
Bahng, M.C. Bhatnagar, J. Brunke, E. Colombo, R. Diaz, W. Due, D.Dufournet, P.C. Fernandez, Y. Filion, A.
Giboulet, J. Jäger, S. Kale, A. Keri, T. Kobayashi, M. Kosakada, E. Kynast, Z. Liu, A. Lokhanin, C. van der
Merwe, M. de Nigris, V. Rashkes, D. Peelo, U. Riechert, B. Shperling, R. Smeets, L. Stenström, G. Sun, S. Wang, Y.
Yamagata, Y. Sili, J. Yesuraj, R. Yeckley.
[2] CIGRÉ Technical Brochure 456, April 2011, “Background of Technical Specifications for Substation
Equipment exceeding 800 kV AC“, by CIGRE WG A3.22
[3] CIGRÉ SC A3 Session 2008, Report A3-211 “Technical Requirements for Substation Equipment exceeding
800kV”
[4] Électra 241, December. 2008, “Technical Requirements for Substation Equipment exceeding 800kV AC”
[5] Second International IEC/CIGRÉ Symposium on Standards for Ultra High Voltage Transmission, New Delhi,
January 2009, “System Impacts on UHV Substation Equipment”
[6] Second International IEC/CIGRÉ Symposium on Standards for Ultra High Voltage Transmission, New Delhi,
January 2009, “UHV Equipment Requirements: state of the art & prospects for equipment”
[7] CIGRÉ SC A2/A3/B3 Colloquium and Regional Conference 2009, South Africa “Comparison of UHV and
800 kV Specifications for Substation Equipment”
[8] 2009 International Conference on UHV Power Transmission, Beijing, May 2009, “Switching Transients in
UHV Networks”
[9] Technical paper A3-101 presented at 2011 SC A3 Colloquium in Vienna, September 2011, “Background
information & study results for specifications of UHV substation equipment“
[10] Technical paper 296 presented at 2011 CIGRE Symposium in Bologna, September 2011, “Consideration
& recommendation for specifications of UHV substation equipment“
[11] IEC 62271-100 High-voltage switchgear and controlgear – Part 100: Alternating-current circuit-breakers
(2012)
[12] IEC 62271-102 High-voltage switchgear and controlgear – Part 102: Alternating current disconnectors
and earthing switches
[13] IEC 62271-109 High-voltage switchgear and controlgear – Part 109: Alternating-current series capacitor
by-pass switches
[14] IEC 62271-112 Ed 1.0: High-voltage switchgear and controlgear - Part 112: Alternating current high-
speed earthing switches for secondary arc extinction on transmission lines
[15] A. Greenwood, Electrical Transients in Power Systems, 2nd edition, Wiley Interscience
[16] R.M. Hasibar, A.C. Legate, J. Brunke, W.G. Peterson, “The Application of High Speed Grounding Switches
for Single-Pole Reclosing on 500 kV Power Systems” (IEEE PAS, Vol. 100, No. 4, April 1981, pp. 1512-1515.
[17] H. Mizoguchi, I. Hioki, T. Yokota, Y. Yamagata, K. Tanaka, “Development of an interrupting chamber for
1000 kV High-Speed Grounding Switch” (IEEE PD, Vol.13, No.2 April 1998, pp. 495)
Page 193
[18] J. Berger, N. Bhatt, M. Chau, R. Gutman, A. Keri, “American Electric Power Experience with 765 kV
Transmission” (International Conference of UHV Transmission Technology, Beijing, China, 2006, Key technology
of UHVAC transmission system)
[19] CIGRÉ Technical Brochure 336, December 2007, “Changing System Conditions and System Requirements –
Part II: The impact of long distance transmission on HV equipment”
[20] Prof. Jiming Lin, e.a., “A Study on Transient Recovery Voltage of UHV Circuit Breakers” (International
Conference of UHV Transmission Technology, Beijing, China, 2006, Key technology of UHVAC transmission
system)
[21] Prof. Jiming Lin, e.a., “Study on Transient Recovery Voltage of circuit breakers in UHV AC double circuit
system” (IEEE International Conference on Power System Technology PowCon 2010, Hangzou, China, Oct.
2010)
[22] CIGRÉ Technical Brochure 542, June 2013, “Insulation Coordination for UHV AC Systems”, by WG C4.306
[23] A. Alfredsson, “Analysis and proposal for standardization of transient recovery voltages on series
compensated lines”, Cigré SC A3 Colloquium, Rio de Janeiro, 2007
[24] Q. Bui-Van, e.a., “Control of Switching Overvoltages and Transient Recovery Voltages for Hydro-Québec
735-kV Series Compensated Transmission System”, IPST’07, Lyon, June 4-7
[25] J. Redlund, e.a., “New Fast Protective Device for High Voltage Series Capacitors”, IEEE Power Engineering
Society general Meeting 2006, Montreal
[26] Zutao Xiang, Jiming Lin, Liangeng Ban, Bin Zheng, “Investigation of TRV across Circuit-breaker of Series
compensated Double-circuit UHV Transmission Lines”, IPST 2010 ??
[27] Bin Zheng, Zutao Xiang, Liangeng Ban, Jiming Lin, Nihong Gu, Gang Sun, Bin Han, “TRV Phenomenon in
Chinese 1100kV UHV Series Compensated System”, IPST 2011, paper 119
[28] Zhang Yuanyuan, Ban Liangeng, Xiang Zutao, Han bin, Zheng bin, Ma Wenyuan, “Transient Analysis on
Second Arc Current and Recovery Voltage of UHV Series Compensated Transmission Line”, IPST 2011, paper
96
[29] Lin Jiming, Wang Xiaogang, Ban Liangeng, Xiang Zutao, Zheng Bin, “Resonance overvoltage caused by
energizing no-load transformers in UHV system”, International Conference of UHV Power Transmission
Technology 2006, Beijing
[30] B. Khodabakhchian, “EHV Single-Pole-Switching: It’s Not Only a Matter of Secondary Arc Extinction”,
Submitted for presentation at IPST 2013, Vancouver, BC, July 2013
[31] R. Grünbaum, G. Ingeström, B. Ekehov and R. Marais, “765 kV Series Capacitors for Increasing Power
Transmission Capacity to the Cape Region”, IEEE PES PowerAfrica Conference and Exposition, Johannesburg,
South Africa, July 2012.
[32] A.L.J. Janssen, D. Dufournet, H. Ito, H. Kajino, Y. Yamagata, M. Kosakada, S. Poirier, J. Fan, Z. Xiang, P.
Fernandez, “Impact of UHV and EHV System Design on TRV Waveshapes”, UHV Symposium 2013, New Delhi
[33] R.H. Harner, J. Rodriguez, Transient Recovery Voltages associated with Power System, Three-Phase
Transformer Secondary Faults, IEEE PAS-91, no.5, Sep.1972, pp. 1887-1896.
[34] R. Horton, R.C. Dugan, K. Wallace, D. Hallmark, Improved Autotransformer Model for Transient Recovery
Voltage (TRV) Studies, IEEE PD, Vol.27, no.2, April 2012, pp. 895-901.
Page 194
[35] M. Steuer, W. Hribernik, J.H. Brunke, Calculating the Transient Recovery Voltage Associated With Clearing
Transformer Determined Faults by Means of Frequency Response Analysis, IEEE PD Vol. 19, No.1, Jan. 2004, pp.
168-173
[36] W. Hribernik, L. Graber, J.H. Brunke, Inherent Transient Recovery Voltage of Power Transformers – A Model-
Based Determination Procedure, IEEE PD Vol. 21, No.1, Jan. 2006, pp. 129-134
[37] H. Kagawa, T. Maekawa, Y. Yamagata, S. Nishiwaki, T. Chigiri, T. Saida, O. Hosokawa, Measurement and
Computation of Transient Recovery voltage of Transformer Limited Fault in 525 kV-1500 MVA Three-Phase
Transformer, IEEE T&D Conference, Orlando, May 2012
[38] Y. Yamagata, H. Kagawa, M. Kosakada, M. Toyoda, J. Kida, H. Ito, Considerations on Transformer Limited
Fault duty for GCB in UHV and EHV networks with large capacity power transformers, CIGRE SC A3 Session 2012,
Report A3-108
[39] H. Ikeda, E. Haginomori, H. Toda, M. Hikita, T. Koshizuka, Investigation of TRV after Interrupting
Transformer Limited Fault, CIGRE SC A3 Technical Colloquium 2011, Vienna, Report A3-102
[40] T. Koshizuka, T. Nakamoto, E. Haginomori, M. Thein, M. Hikita, H. Ikeda, TRV under Transformer Limited
fault Condition and Frequency-Dependent Transformer Model, IEEE PES General Meeting, San Diego, July 2011
[41] D.Dufournet, A.Janssen, J.Hu, Transformer Limited Fault Transient Recovery Voltage for EHV and UHV
circuit-breakers, ISH 2013, Seoul, August 2013
[42] A.C.O. Rocha, G.H.C. Oliviera, R.M. Azevedo, A.C.S. Lima, Assessment of Transformer Modeling Impact on
Transient Recovery Voltage of Transformer Limited Fault, CIGRE SC A3 Session 2012, Report A3-107
[43] A.C.O. Rocha, S.O. Moreira, L.H.S. Duarte, CEMIG experience in the analysis of transient recovery voltages
with Transformer-fed faults, IPST 2009, Zagreb, Report 0204
[44] A.C.O Rocha, contribution to Question 1.8 of the CIGRÉ SC A3 Session 2012 Special Report, Proceedings
of the CIGRÉ 2012 Sessions
[45] M. Kosakada, M. Toyoda, contribution to Question 1.8 of the CIGRÉ SC A3 Session 2012 Special Report,
Proceedings of the CIGRÉ 2012 Sessions
[46] M. Toyoda, M. Kosakada, contribution to Question 1.8 of the CIGRÉ SC A3 Session 2012 Special Report,
Proceedings of the CIGRÉ 2012 Sessions
[47] A.L.J. Janssen, D. Dufournet, H. Ito, H. Kajino, Y. Yamagata, M. Kosakada, “Transformer Limited Fault
Duties for UHV and EHV”, UHV Symposium 2013, New Delhi
[48] IEC SC 17A/969/CD, committee draft for HSES standard
[49] N. Knudsen, “Single Phase Switching of Transmission Lines Using Reactors for Extinction of the Secondary
Arc” CIGRE Paper No. 310, 1962
[50] E.W. Kimbark, “Suppression of Ground-Fault Arcs on Single-Pole-Switched EHV Lines by Shunt Reactors”
IEEE Transactions on Power Apparatus and Systems, March 1964
[51] M. Toyoda, et al., “Considerations for the standardization of high-speed earthing switches for secondary
arc extinction on transmission lines”, CIGRE 2011 Bologna 2011, 295
[52] M. Toyoda, et al., “Considerations for the Standardization of High-speed earthing switches for secondary
arc extinction on transmission lines (part 2)”, CIGRE 2011 SC A3 Technical Colloquium, A3-103 2011
Page 195
[53] R. M. Hasibar, et al, “The application of high-speed grounding switches for single-pole reclosing on 500
kV power systems”, IEEE Tr. on PAS, Vol. PAS-100, No. 4, April 1981
[54] G.E. Agafonovet al., “High Speed Grounding Switch for Extra-high Voltage Lines”CIGRE 2004, A3-308
[55] S. P. Ahn,et al.,”The Investigation for adaptation of High Speed Grounding Switches on the Korean 765kV
Single Transmission Line”,IPST 05-096 Montreal 2005
[56] Y. Yamagata et al., “Development of 1,100kV GIS - Gas Circuit Breakers, Disconnectors and High-speed
Grounding Switches –”, CIGRE 1996, 13-304
[57] LI Zhen-qiang et al., “Effect of UHV Ground Wire Disposition and HSGS on Second Arc Current”, UHV
transmission technology in 2009 an international conference CP0282
[58] J. Takami, S. Okabe, “Observation results of lightning current on transmission towers.”, IEEE Tr. on Power
Delivery, Vol..22, No.1 Jan. 2007
[59] Y.Goda et al, “Insulation Recovery Time after Fault Arc Interruption for Rapid Auto-reclosing on UHV
(1000kV class) Transmission lines” , IEEE Tr. on PWRD, Vol. 10, No.2, April 1995
[60] LI Zhen-qiang et al., “Effect of UHV Ground Wire Disposition and HSGS on Second Arc Current”, UHV
transmission technology in 2009 an international conference CP0282
[61] M.Kosakada, “High Speed Earthing Switch (HSES) /High Speed Grounding Switch (HSGS)”, CIGRE tutorial
2012 Rio de Janeiro
[62] M.Kosakada, “DS, ES, HSES/HSGS”, CIGRE WG A3.22/28 tutorial at IEEE Switchgear Committee 2012
Fall meeting San Diego
[63] B. R. Shperling and A.J. Fakheri (A. J. F. Keri), “Single Phase Switching Parameters for Untransposed EHV
Transmission Lines”, IEEE Transactions on Power Apparatus and Systems, Mar/Apr. 79
[64] H.N. Scherer, B.R. Shperling, J.W. Chadwick, N.N. Belyakov, V.S. Rashkes, K.V. Khoetsian, “Single Phase
Switching Tests on 765 kV and 750 kV Transmission Lines”, IEEE PAS, ne 1985, pp. 1537-1548
[65] Zhang Yuanyuan, Ban Linageng, Xiang Zutao, Han Bin, Zheng Bin, Ma Wenyuan, “Transient Analysis on
Second Arc Current and Recovery Voltage of UHV Series Compensated Transmission Line”, IPST 2011, Delft
(Netherlands), Report 96
[66] M. Roland, J. Jagar, Enhancement of the reliability of extra and Ultra high voltage Transmission Systems”,
IEEE PowerTech 2009
[67] J. Takami, S. Okabe, “Characteristics of direct lighting strokes to phase conductors of UHV transmission
lines”, IEEE Tr on Power Delivery, Vol..22, No.1 Jan. 2007
[68] N.N. Belyakov, V.S. Rashkes et al, “Application of Single-Phase Auto-reclosing in a Complex EHV Network
Containing 1200 kV Transmission Lines” CIGRE SC 34 Session 1990, Report 34-207
[69] A. J. Fakheri (A.J.F. Keri), T.C. Shuter, J.M. Schneider and C.H. Shih, "Single Phase Switching Tests on the
AEP 765 kV System - Extinction Time for Large Secondary Arc Currents", IEEE Transactions on Power Apparatus
and Systems, August 1983
[70] B. R. Shperling, A.J. Fakheri (A.J.F. Keri), C.H.Shih, B.J.Ware “Analysis of Single-Phase Switching Field Tests
on the AEP 765 kV System”, IEEE Transactions on Power Apparatus and Systems, April 1981
[71] IEC document 17A/1042/FDIS, Final Draft for International Standard for HSES
Page 196
[72] H. N. Scherer, Jr., B. R. Shperling, J. W. Chadwick, N. N. Beliakov, V. S. Rashkes, K. V. Khoetsian, “Single

Phase Switching tests on 765 kV and 750 kV Transmission Lines”, IEEE Transactions on Power Apparatus and
Systems, June 1985
[73] Smajic, J.; Holaus, W.; Kostovic, J.; Riechert, U.: “3D Full-Maxwell Simulations of Very Fast Transients in
GIS”, IEEE Transactions on Magnetics, Vol. 47, No. 5, pp. 1154-1517, May 2011
[74] CIGRÉ Working Group 15.03: “GIS Insulation Properties in Case of VFT and DC Stress”, 36th CIGRÉ
Session, 1996, Paris, France, CIGRÉ Report 15-201
[75] CIGRÉ Brochure No. 519, December 2012, “Very Fast Transient Overvoltages (VFTO) in Gas-Insulated
UHV Substations”, by CIGRÉ Advisory Group D1.03: Riechert, U.; Neumann, C.; Hama, H.; Okabe, S.; Schichler,
U.; Ito, H.; Zaima, E.
[76] Smajic, J.; Holaus, W.; Troeger, A.; Burow, S.; Brandl, R.; Tenbohlen, S.: “HF Resonators for Damping of
VFTs in GIS”, Proceedings of the 9th Int. Conference on Power System Transients (IPST2011), Paper ID:185,
Delft University of Technology, The Netherlands, June 2011
[77] Zaima, E.; Neumann, C.: “Insulation Coordination for UHV AC Systems based on Surge Arrester Application
(CIGRÉ C4.306)”, IEC – CIGRÉ International Symposium on International Standards for UHV Transmission, New
Delhi, India, 2009, pp. 108-118
[78] Y.Shu et al, Experimental Research on Very Fast transient Overvoltage in 1100kV Gas Insulated
Switchgear, IEEE PD vol.28 N°1, pp 458-466 (2013-01)
[79] Riechert, U.; Holaus, W. “Ultra High Voltage Gas-Insulated Switchgear – A Technology Milestone”,
European Transactions on Electrical Power ETEP, Special Issue: UHV-AC Transmission, Volume 22, Issue 1, pages
60–82, January 2012, Published online 18 May 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI:
10.1002/etep.582
[80] Th. Berteloot, A. Girodet, P. Vinson, M. Bernard, Development of a Gas Insulated Disconnector for UHV
Networks, CIGRE Colloquium in New Delhi 2013 (2013-04).
[81] M. Collod, E. Thuries, S.A. Boggs, N. Fujimoto, The modeling of statistical operating parameters and the
computation of operation-induced surge waveforms for GIS disconnectors (CIGRE, 1984)
[82] Ban Liangeng, Xiang Zutao, Wang Sen, Lin Jiming, Wang Xiaogang, Zheng Bin, Wang Xiaotong:
“Estimation of VFTO for GIS and HGIS of China 1000 kV UHV pilot project and its suppressing
countermeasures”, IEC - CIGRÉ UHV Symposium, Beijing, China, 2007, paper 2-3-4
[83] Hemmi, R.; Shinohara, R.; Kitazumi, Y.; Yatsuzuka, H.; Hirasawa, K.; Yamagiwa, T.: ”Evaluation of VFTO
(Very Fast Transient Overvoltage) and Its Reduction by Parallel Resistor during Switching Operation of
Disconnecting Switch in Future UHV Gas Insulated Substations”, Int. Conference of UHV Power Transmission
Technology, Beijing, China, 2006, pp. 460-466
[84] Riechert, U.; Krüsi, U.; Sologuren-Sanchez, D., Very Fast Transient Overvoltages during Switching of Bus-
Charging Currents by 1100 kV Disconnector, CIGRÉ Report A3-107, 43rd CIGRÉ Session, August 22-August
27, 2010, Palais des Congrès de Paris, Paris, France
[85] Riechert, U.; Bösch, M.; Smajic, J.; Shoory, A,; Szewczyk, M.; Piasecki, W.; Burow, S.; Tenbohlen, S.:
“Mitigation of Very Fast Transient Overvoltages in Gas Insulated UHV Substations”; CIGRÉ Report A3-110,
44th CIGRÉ Session, August 26-August 31, 2012, Palais des Congrès de Paris, Paris, France CIGRÉ Session
2012
Page 197
[86] Riechert, U., Mitigation of Very Fast Transient Overvoltages, CIGRÉ Contribution A3, PS1-Q1-05, 44th
CIGRÉ Session, August 26-August 31, 2012, Palais des Congrès de Paris, Paris, France
[87] Rama Rao, J. V. G., Amarnath, J.; Kamakshaiah, S.: ”Simulation and Measurement of Very Fast Transient
Overvoltages in a 245 KV GIS and Research on Suppressing Method using Ferrite Rings”; ARPN Journal of
Engineering and Applied Sciences, Vol. 5, No. 5, pp. 88-95, May 2010
[88] Wang Xiaogang, Yin Yonghua, Ban Liangeng, et al. “Summary of Commissioning Test on the 1000kV UHV
AC Demonstration Project”, Proceedings of the CSEE, vol.29, no.22, pp.12-18, 2009
[89] Zheng Bin, Ban Liangeng, Zhang Yuanyuan, et al. “Contrastive Analysis on Simulated and Measured
Overvoltages of 1000kV Pilot Project in China”, Power System Technology, vol.33, no.16, pp.24-28, 2009
[90] Han Bin, Lin Jiming, Ban Liangeng, et a1. “Study on Single-Phase Reclosing Issues for 1000kV UHV AC
Transmission Pilot Project”, Power System Technology, vol.33, no.16, pp.20-23, 2009
[91] Wang Qingling, “Analysis of the Grounding Switch in 500kV Dieling Substation Unable to be Normally
Operated under Certain Condition”, Southern Power System Technology, vol.5 no.3, 2011
[92] Fu Yiping, Li Meng, et.al. “Alteration and Switching of 500kV Double Circuit lines Earthing Switches”, High
Voltage Technology, vol.33 no.4, 2007
[93] Deng Yurong, Li Minggui. “Test and Research on Single-phase Instantaneous Artificial Earthing Fault of
500kV Yan-Sha Line”. Guangxi Electric Power Technology, Vol.3, 1996
[94] Jiang Wei-ping, Zhu Yi-ying, et.al., “Analysis on Measured and Calculated Results of Artificial Single
Phase Grounding Test in 750kV Transmission Pilot Project”. Power System Technology, vol.30 no.19, 2006
[95] S. Ogawa et al.. “Estimation of Restriking Transient Overvoltage on Disconnecting Switch for GIS”，1985
IEEE SM 367-8, 1985.7
[96] S.V. Biriykov, O.V. Volkova et al., “Switching Equipment for 1150kV Transmissions: Circuit-Breaker and
Shunt Reactor Breaker”, Collection of Papers “1150 kV Transmissions”, Vol. 2, Energoatomizdat Publishing
House, Moscow, 1992, pp. 46-72 (in Russian).
[97] A.A. Akopyan et al., “The 750kV Experimental-Commercial Transmission Line Konakovo-Moscow”, CIGRE
Session, 1964, #413.
[98] A.A. Akopyan et al., “Switching Overvoltages and the System of Protection against Them in 750 kV
Networks of the USSR”, CIGRE Session, 1972, #33-07.
[99] A.A. Akopyan et al., “Results of Tests on the Experimental-Commercial Transmission Line 750 kV Konakovo-
Moscow. Part 1-Tests in the Starting Scheme of the Transmission”’, Collection of Papers “Long-Distance 750 kV
Transmissions”, Vol. 2, Energia Publishing House, Moscow, 1975, pp. 42-61 (in Russian).
[100] E.M. Bazelian et al., “Switching Impulse Shape and Dielectric Strength of External EHV/UHV Insulation”,
CIGRE Session, 1976, #33-02.
[101] A.A. Akopyan et al., “Results of Tests on the Experimental-Commercial Transmission Line 750 kV
Konakovo-Moscow. Part 2- Tests in the Full Scheme of the Transmission”’, Collection of Papers “Long-Distance
750 kV Transmissions”, Vol. 2, Energia Publishing House, Moscow, 1975, pp. 61-86 (in Russian).
[102] A.N. Komarov, V.S. Rashkes, N.V. Shilin, “EHV Shunt Reactor Inductive Current Switching-Off”, CIGRE
Session, 1978, #13-06.
Page 198
[103] N.N. Beliakov, A.N. Komarov, V.S. Rashkes, “Results of Internal Overvoltages and Electrical Equipment
Characteristics Measurements in the Soviet 750kV Networks”, CIGRE Session, 1978, #33-08.
[104] N.N. Belyakov et al., “1150kV Experimental Installation at the Bely Rast Substation”, CIGRE Session,
1976, #23-03.
[105] I.M. Bortnik, N.N. Belyakov et al., “1200kV Transmission Line in the USSR- the First Results of Operation”,
CIGRE Session, 1988, #38-09.
[106] N.N. Belyakov et al., “Application of Single-Phase Autoreclosing in a Complex EHV Network Containing
1200kV Transmission Lines”, CIGRE Session, 1990, #34-207.
[107] V.Z. Annenkov et al., “Testing 750-1150kV Electrical Equipment on Commercial 750kV Transmissions and
at the 1150 kV Bely Rast Test Site”, Collection of Papers 1150 kV Transmissions, vol. 2, Energoatomizdat
Publishing House, Moscow, 1992, pp.185-203 (in Russian).
[108] N.N Belyakov et al., “Testing Electrical Equipment and Measuring Overvoltages on the First Commercial
1150kV Transmission Ekibastuz-Kokchetav”, Collection of Papers 1150kV Transmissions, vol. 2,
Energoatomizdat Publishing House, Moscow, 1992, pp.204--223 (in Russian)
[109] G. Trudel et al., “Designing A Reliable Power System: Hydro-Québec Integrated Approach”, Proceedings
of the IEEE, Vol. 93, No. 5, May 2005
[110] B. Khodabakhchian et al., “TRV and the Non-Zero Crossing Phenomenon in Hydro-Québec's Projected
735 kV System”, CIGRÉ paper 13-303, 1992.
[111] Q. Bui-Van et al., “Control of Switching Overvoltages and Transient Recovery Voltages for Hydro-
Québec 735-kV Series-Compensated Transmission System”, Int. Conference on Power Systems Transients (IPST),
2007.
[112] K.Taguchi, M.Matsuura, MAbe, H.Kameda, T.Matsushima, “Failure extension protection systems in Japan
(N°1) – Application Policy and Major Failure Experiences”, 2001 CIGRE SC 34 Colloquium, Sibia, report 104.
[113] T.Yamada, A.Takeuchi, F.Kumura, H.Kameda, K.Dekiguchi, “Coordination Experiences in Japanese
Transmission Transmission System”, 2007 CIGRE SC B5 Colloquium, Madrid, Report 308.
[114] T.Tsuboi, J.Takami, S.Okabe, K.Aoki, Y.Yamagata, “Investigation on Secondary Arc Extinction Time for
Large-sized Transmission Lines”, IEEJ 8th International Workshop on High Voltage Engineering 2012, pp.123-
128, 2012.
Page 199
Annex
Resonance due to coupling amongst lines compensated by shunt reactors
Considering double circuit lines, Figure A.1 shows the profile of the induced voltage on all phases of an open line
as a function of the amount of the open circuit shunt compensation, whereas the parallel line is transmitting power.
Both lines are non-transposed but resonance may also occur at transposed lines, with a shift of the resonance
points along the MVAr range. The horizontal dotted line in the figure corresponds to the rated phase-to-neutral
voltage. The neutral of the shunt reactors is directly earthed (no four-legged shunt reactor).
Figure A.1: Induced voltage on an open line due to a parallel line coupling, as a function of
shunt compensation level
In fact, such resonance phenomenon in lines using single-phase auto-reclosing (SPAR) is a particular case of what
occurs in resonant parallel lines due to coupling between them. For this phenomenon, the influence of the
transmitted power is also quite relevant. Figure A.2 shows three profiles for induced voltages in 500 kV parallel
lines.
 Profile (1) corresponds to the parallel line no-loaded;

 Profile (2) assumes the parallel line is loaded with approximately 1900 MW. Note the elevation of the profile (2)
around the first resonance peak;
 Profile (3) gives the same situation as profile (2), but only one phase floating, as applied with SPAR.
Although the voltage cannot achieve levels likely to cause damage to equipment, it can extend the extinction of the
secondary arc. For the results shown, half transposed (γ transposition) transmission lines and shunt reactors with
directly earthed neutral have been assumed.
Page 200
Figure A.2: Induced voltages in a 500 kV line with three conditions of a parallel line coupling
From Figures A.1 and A.2 it is clear that transposition will lead to a considerable reduction of the induced voltages.
Furthermore, the application of four-legged shunt reactors will generally lead to a reduction of the induced voltages.
In Figure A.3, the effect of neutral reactors (with and without power transmission: curve 2 resp. 1) and of neutral
resistors (with and without power transmission: curve 4 and 3 resp.) can be seen. However, there is a danger of
resonance by the combination of shunt-reactor and neutral reactor: Figure A.4.
Figure A.3: Effectiveness of neutral reactors and resistors
Page 201
Figure A.4: Evolution of induced voltage in case of possible resonance of the main and neutral
reactor.
Though, transposition (γ- or preferably β-transposition) and four-legged shunt-reactors will generally lead to more
acceptable levels of induced voltages derived from coupling between parallel OHL. [1][2][3]
References
[1] Amon J. F. et alii - Electrical Transients and Insulation Coordination - Application in HV Electric Power
Systems – Text book published (in portuguese) by FURNAS/UFF (Fluminense Federal University), 1987 –
Chapter 19 - Luiz Eduardo Nora Dias (English version: Jorge Amon Filho, Thiago Ferreira da Silva Costa and
Daniel Flores Silva – March, 2013 – Rio de Janeiro, RJ, Brazil)
[2] L. E. Nora Dias - Resonance by induction in lines compensated by shunt reactors – Article submitted to
COPPE/UFRJ (Federal University of Rio de Janeiro) to get the MsC. Degree in Electrical Engineering – 1977.
[3] Forest, J. J. La et alii - Resonant Voltages on Reactor Compensated Extra-High-Voltage Lines - IEEE
Transactions on Power Apparatus and Systems New York, 91 (6): 2528-2536, November/December 1972.
Page 202

570

Diunggah oleh

Informasi Dokumen

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

570

Diunggah oleh

Hak Cipta:

Format Tersedia

570

Switching Phenomena for

SPAR Single-Phase rapid Auto-Reclosing

Switching phenomena for EHV and

2 Network configuration and technical solutions .................................................13

3 TRV of circuit breakers for EHV and UHV networks ........................................22

4 Transformer limited faults ...............................................................................................74

4.7 Advanced transformer models......................................................................................86

5 High-speed earthing switch (HSES) ............................................................................97

6 Disconnector and earthing switch............................................................................ 123

7 Switching experience during and after system commissioning ............ 160

8 Conclusions ............................................................................................................................ 190

Bibliography/References ........................................................................................................ 193

Annex ................................................................................................................................................. 200

Work was undertaken in three main areas:

- Influence of system and equipment parameters on TRV;

- Testing requirements for UHV/EHV circuit breakers.

- Switching requirements of DS and ES based on model UHV/EHV substations;

- Switching requirements of HSES based on model UHV/EHV networks;

- Technical comparison between four legged shunt reactor and HSES;

- Testing requirements for DS, ES and HSES.

3- Field experience and switching behaviour during and after commission

- Interrupting current and TRV of DS, ES and HSES;

- Primary and secondary arc extinction behaviours in single/double circuits;

- Out-of-phase switching behaviour.

2 Network configuration and technical solutions

2.2 Design criteria

Country Utility Voltage Line length Single/double Transposition Shunt Series

2.3 Secondary arc extinction

Country Utility Voltage Re- four-legged Special shunt reactor HSES

2.4 Series capacitor banks

Figure 2.4.2: Furnas single phase of a 800 kV series capacitor bank

2.5 Insulation coordination

Figure 2.6.1: Case 1

Figure 2.6.3: Case 2

Figure 2.6.4: TRV in case 2 [6]

Figure 2.6.6: Example of simulation of an unstable power swing in Thailand

3 TRV of circuit breakers for EHV and UHV networks

Figure 3.2.1: Radial network model with four power sources

Tr: 2 units Tr: 2 units Tr: 2 units

E-s/s D-4 D-s/s F-s/s

360 km 360 km 360 km

A-s/s B-4 B-s/s C-s/s

Figure 3.2.2: Meshed network model with six power sources

3.3 System and Equipment parameters used for TRV study

Table 3.3.1: System and compensation parameters in different counties

3.3.1 Transformer specifications

Country Japan China India

Country Canada Canada Canada

Table 3.3.3: Typical specifications of 765 kV 1650MVA/550 kV 1200MVA transformers

Country Thailand Thailand Thailand

3.3.2 Lower voltage system conditions

3.3.3 Tower and conductor configurations

Figure 3.3.1: 1100 kV double circuit tower

Dimension Japan China

Figure 3.3.2: 1200 kV/765 kV single circuit tower

Dimension India Canada

Figure 3.3.3: 550 kV and 800 kV double circuit tower

Dimension Thailand Korea

Figure 3.3.4: 550 kV single circuit tower

Dimension Canada Thailand

3.3.4 Phase arrangement of the transmission lines

Figure 3.3.5: Transposition for 120 km transmission line

Figure 3.3.7: Transposition for 360 km transmission line