Insulation Coordination For UHV AC Systems PDF

INSULATION
COORDINATION
FOR
UHV AC
SYSTEMS
WG C4.306
TypethenameofTB
Members
E.Zaima,Convenor(JP),T.Koboyashi,Secretary(JP),J.Takami,AsistantSecrery(JP),P.C.
Fernandez(BR),D.Peelo(CA),Q.BuiVan(CA),W.Chen(CN),A.Sabot(FR),F.Gallon(FR),E.
Kynast(DE),A.Pal(IN),R.N.Nayak(IN),S.Malgarotti(IT),T.Yamagiwa(JP),E.Shim(KR),A.
Lokhanin(RU),P.Tlhatlhetji(ZA),A.Amod(ZA),C.vanderMerwe(ZA),U.Kruesi(CH),D.
Sologuren(CH),Y.Vachiratarapadorn(TH),A.J.F.Keri(US),A.Villa(VE),G.Carrasco(VE),H.Ito
(JP),T.Yokota(JP),Y.Shirasaka(JP),B.Richter(CH),U.Riechert(CH)
Coordinationwith
P.Zhou(CN),J.Lin(CN),Z.Li(CN),K.Uehara(JP),Y.Ishizaki(JP),S.Okabe(JP),M.Miyashita(JP),
H.Kajino(JP)
Copyright 2011
Ownership of a CIGRE publication, whether in paper form or on electronic support only infers right
of use for personal purposes. Are prohibited, except if explicitly agreed by CIGRE, total or partial
reproduction of the publication for use other than personal and transfer to a third party; hence
circulation on any intranet or other company network is forbidden.
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 : (To be completed by CIGRE)
Page 1
TypethenameofTB
ISBN : (To be completed by CIGRE)
Page 2
TypethenameofTB
INSULATION COORDINATION
FOR
UHV AC SYSTEMS
NoExtraCoverpageorblankpages
UseCIGREabbreviationsinsofaraspossible:StudyCommitteeSC,TechnicalBrochureTB,WorkingGroupWG
JointWorkingGroupJWG,TechnicalCommitteeTC
Photos:mustbeofreasonable definition (preferably 300 dpi);allfiguresandtablesmustbetitled,legibleandnumbered
withlegendsprovided.
NoCompanylogos
Table of Contents
1 Introduction .................................................................................................................6
2 Concept of recent practices on insulation coordination for the UHV and the 800kV
system ..........................................................................................................................8
2.1 Insulation coordination throughout substation and transmission line...................8
2.2 Reduction of insulation levels using overvoltage suppression measures .............16
References ..............................................................................................................19
3 Recent practice on insulation coordination for the UHV and the 800 kV system...........20
3.1 China UHV project .............................................................................................20
3.2 India UHV project ..............................................................................................29
3.3 TEPCO 1100 kV project .....................................................................................35
3.4 Brazil 800 kV project .........................................................................................46
3.5 China 750 kV AC Project....................................................................................60
3.6 India 800 kV project ..........................................................................................64
3.7 Korea 765 kV project.........................................................................................68
References ..............................................................................................................82
Page 3
TypethenameofTB
4 Overvoltage in UHV range...........................................................................................83

4.1 Determination of stresses (TOV, switching overvoltage, lightning overvoltage and
VFTO)......................................................................................................................83
4.2 TOV due to load rejection and ground fault .......................................................85
4.3 Switching overvoltages caused by closing and opening with ground fault overvoltage
............................................................................................................................... 93
4.4 Lightning overvoltages caused by back-flashover and direct lightning .............115
4.5 VFTO stress in GIS due to DS switching............................................................123
4.6 Influence of metal oxide surge arresters on circuit breaker TRVs .....................162
4.7 Conclusions ....................................................................................................168
5 Evaluation of overvoltage study results .....................................................................170
5.1 Overvoltage simulation tools and verification by measuring results..................170
References ............................................................................................................179
5.2 Main Characteristics of Metal-Oxide Surge Arresters (MOSAs) ..........................180
References ............................................................................................................188
5.3 Evaluation of waveform - Conversion in shape of field overvoltage to standard
impulse waveform in determining representative overvoltages - ............................189
References ............................................................................................................200
5.4 Conclusions ....................................................................................................200
6 Switching Overvoltage Mitigation Measures for Future UHV Systems .........................201
6.1 Introduction ....................................................................................................201
6.2 Fast Insertion of Shunt Reactors ......................................................................202
6.3 Closing Resistors.............................................................................................202
6.4 Staggered Pole Closing ....................................................................................203
6.5 Line Surge Arresters ........................................................................................203
6.6 Controlled Closing...........................................................................................205
6.7 Comparison and Relevance to Future UHV Systems ..........................................207
6.8 Conclusions ....................................................................................................209
References ............................................................................................................209
7 Some aspect of insulation coordination of air gaps in the UHV range (phase-to-earth
and phase-to-phase insulation).................................................................................211

7.1 Introduction ....................................................................................................211
7.2 Air Gap Clearances chosen for UHV Projects in different countries ...................211
7.3 IEC Minimum Air gap clearance according to IEC 60071-1.............................212
Page 4
TypethenameofTB
7.4 Air Gap Clearance Calculation for SIWV ............................................................214

7.5 Background data on flashover characteristics for UHV air clearance .................220
7.6 Recent investigations on air gap clearance in the UHV range ............................227
7.7 Non standard switching impulse waveforms in the insulation coordination ......245
7.8 Conclusion ......................................................................................................248
7.9 References ......................................................................................................249
8 Selection of insulation levels.....................................................................................251
8.1 Procedure for Insulation coordination ..............................................................251
8.2 Determination process for LIWV and SIWV........................................................261
8.3 Consideration of VFTO for insulation coordination...........................................266
8.4 Power frequency (AC) voltage tests for substation equipments.........................276
REFERENCES ..........................................................................................................284
8.5 Conclusion ......................................................................................................285
9 Conclusion and Recommendation .............................................................................286
9.1 Recent Practices on insulation coordination for UHV and 800 kV system ..........286
9.2 Overvoltages in UHV range ..............................................................................286
9.3. Evaluation of overvoltages ..............................................................................287
9.4 Switching overvoltage mitigation measures for future UHV systems .................287
9.5 Review on insulation coordination of air gaps in the UHV range .......................288
9.6 Selection of insulation levels............................................................................288
Page 5
TypethenameofTB
1 Introduction
Different countries in the world are planning and realizing UHV AC systems with operating voltages exceeding
800kV. When planning a new power system, in particular at a new voltage level, insulation coordination is one of
the most important subjects. The main task is the determination of stresses and the assessment of the strength of
the system and the equipment installed.
The general procedure of insulation coordination is described in IEC 60071-1 (2010). This standard does not give
precise advice regarding new voltage levels although it provides insulation levels for Um values of both 1100 kV
and 1200 kV. These insulation levels are based on both past experience available from former CIGR work that
also considered the 1000kV voltage level, and recent works in Japan, China and India. The research activities
within previous CIGR SC 33 in the topic of UHV transmission provided a good basis on overvoltages and air
insulation performance to make possible the design of air insulation for both 1100kV and 1200kV highest voltage of
equipment.
Since 1990s, metal oxide surge arresters have been applied to UHV substation design. Insulation coordination for
UHV has been changed based on these arresters throughout substation and transmission line. Also, gas insulated
switchgears (GIS, Hybrid-IS) have been generally applied to UHV substation design.
Considering the above issues, CIGR WG C4.306 has reviewed and discussed insulation coordination practice in
the UHV AC range taking into account the state-of-the-art technology, with special reference to surge arresters.
Such a review has been taken into account the accumulated knowledge of various CIGR working bodies, and
accomplished in collaboration with related CIGR SC A3 and B3 (WG A3.22, A3.28, B3.22 and B3.29).
Recommendation, for application guide IEC 60071-2 (1996) and IEC apparatus standards has been proposed.
The task of CIGR WG C4.306 is divided into four main sections dealing with (see Figure 1.1):
Recent practice on insulation coordination for UHV system:
- Insulation coordination throughout substation and transmission line
- Reduction of insulation levels by application of higher performance surge arresters and other overvoltage
suppression measures
Overvoltage in UHV range: (especially focused on peculiarity to UHV AC system)
- Determination of stresses (TOV, switching overvoltage, lightning overvoltage and VFTO) by simulation tools
and verification by measuring results
- TOV due to load rejection and ground fault
- Switching overvoltages caused by closing and opening (fault-clearing) with ground fault overvoltage
- Lightning overvoltage caused by back-flashover and direct lightning, VFTO stress in GIS due to disconnector
switching (ref to CIGRE brochure "Monograph on GIS Very Fast Transients 1989)
Review on insulation coordination of air gaps in the UHV range:
- Phase-to-phase insulation
Selection of insulation levels:
- Coordination withstand voltages and safety factors for equipment
- Selection of insulation levels for equipment and transmission lines
More than 250 written technical contributions have been prepared by 29 experts from 15 countries during the
investigations.
Page 6
TypethenameofTB
Recent practice of
UHV insulation
coordination
1.2
Collaboration with
A3.22&28 and
B3.22&29
IEC 60071-1 Ed 8.1

(2010)
(UHV LIWV & SIWV)
U r/ 3
1. Recent practice on insulation coordination for UHV system

2. Overvoltage in UHV range
(especially focused on peculiarity to UHV AC system)
3. Review on insulation coordination of air gaps in the UHV range
4. Selection of insulation levels
Proposal of recommendation for application guide IEC 60071-2

(1996) by the end of 2012
Figure 1.1 Task of CIGR WG C4.306
Page 7
TypethenameofTB
2 Concept of recent practices on insulation coordination for the

UHV and the 800kV system
The design of UHV and 800kV power system should achieve both economic efficiency and high reliability while
being capable of heavily loaded, long-distance transmission. UHV transmission lines and substation equipment are
inherently large, therefore they should be designed as compact as possible by applying effective insulation
coordination.
Metal oxide surge arrester, which has been applied since 1990s, is the key technology for UHV insulation, and
equipment can be designed optimally by applying them based on detailed overvoltage analysis. The insulation
concept examined in this report is expected to reflect in IEC 60071.
This chapter summarizes the concept of UHV recent practice and is related to the other chapters which mention
each topic in detail.
2.1 Insulation coordination throughout substation and transmission line

The general procedure of insulation coordination is described in IEC 60071-1 (Insulation co-ordination-Part1:
Definitions, principles and rules), and IEC 60071-2 (Insulation co-ordination - Part2: Application guide).
Insulation design of UHV system is required to achieve high reliability. UHV equipment sizes also tend to be large
compared to lower-voltage equipment. Therefore economical and highly reliable transmission lines and substations
with environmental considerations are essential in the UHV system.
From these system requirements for UHV systems, reasonable specifications should be determined by analyzing
overvoltage accurately, and applying sophisticated technologies, such as higher performance metal-oxide surge
arrester (MOSA). The main characteristics of the higher performance surge arrester are described in 5.2.2.
2.1.1 INSULATION COORDINATION RESEARCH IN UHV SYSTEM BY CIGRE AND IEC

CIGRE had researched insulation coordination for UHV within previous CIGRE SC33 since 1970s, and published
Technical brochure No.32 (Final report of the UHV Ad Hoc Group, 1972), and Technical Brochure No.85 (Ultra
High Voltage Technology, 1994).
Rated insulation levels for UHV system are standardized in Amendment 1 of IEC 60071-1 Ed.8.1 (March 2011).
The standard specifies rational insulation levels with the assumptions that several higher performance surge
arresters are installed at adequate locations, and utilities can choose the reasonable insulation level to meet their
own specifications. Table 2.1.1 shows the standard insulation levels in IEC 60071-1. LIWV for UHV system are
1950, 2100, 2250, 2400, 2550, 2700 kV and SIWV 1425, 1550, 1675, 1800, 1950 kV. But air insulation clearances
described in IEC 60071-1 Ed.8.1 are under consideration as shown in Table 2.1.2, therefore the proposal reported
in Chapter 7 of this report is very important. 1100 kV, 1200 kV were added as highest voltage of equipment when
IEC 60038 was revised in June 2009. Table 2.1.3 shows the standard voltages.
The UHV equipment and substation design were researched by CIGRE A3.22 and B3.22, and reported in
Technical brochure No.362 (Technical Requirement for Substation Equipment exceeding 800 kV AC, 2010) and
Technical brochure No.400 (Technical Requirement for Substation exceeding 800kV, 2010). In these technical
brochures, more adequate technical requirements are stipulated by analysing, with the latest tool, suppression of
overvoltages by higher performance arrester and the resistor insertion of disconnector. Detailed specifications of
UHV circuit breakers and disconnectors and field tests have been discussed previously in CIGRE WG A3.28 and
B3.29.
Page 8
TypethenameofTB
Table 2.1.1 Standard insulation levels for UHV (IEC 60071-1 Ed.8)
Table 2.1.2 Standard rated switching impulse withstand voltage and minimum
phase-to-phase clearance for UHV (IEC 60071-1 Ed.8.1)
Table 2.1.3 Standard voltage for UHV (IEC 60038)
Page 9
TypethenameofTB
2.1.2 RECENT PRACTICE OF INSULATION COORDINATION FOR UHV AC TRANSMISSION

SYSTEM
Economical and highly reliable transmission lines and substations equipment with environmental considerations
are essential in the UHV system. Therefore reducing the size of transmission lines and substation equipment are
practical countermeasures.
In Chinese, Indian, Japanese UHV projects, suppressing overvoltage by higher performance surge arresters is a
common countermeasure, and additional countermeasures, such as suppressing overvoltage by the circuit
breakers with closing and/or opening with pre-insertion resistors, are adopted in each project shown in Figure 2.1.1.
In these projects, overvoltages are simulated by the latest analyzing technology such as EMTP.
Practical application of higher performance

metal oxide surge arrester
Reliable circuit breaker with closing

and/or opening resistor
Rational Insulation Specification
LIWV
(Substation)
SIWV
(Substation)
Switching Overvoltage
Insulation Design Level

(Transmission line)
Figure 2.1.1 UHV insulation coordination concept

Figure 2.1.2 shows the flow chart of insulation coordination referred from IEC 60071-1. The basic concept has not
been changed, but the concept is desirable to be reviewed with the latest point of view, by taking account of the
analysis tool improvement, quality improvement, and safety factor which is included in analysis condition.
To design the substation equipment rationally, detailed analysis is more recommended than just applying the
insulation level based on LIPL (lightning impulse protective level of a surge arrester) and SIPL (switching impulse
protective level of a surge arrester), which are calculated by simplified method in IEC 60071-2 (Ref. Chapter 8.2),
because the insulation level has much influence on the construction cost in UHV design.
Page 10
TypethenameofTB
Origin and classification of stressing

voltages
Protective level of overvoltage limit ing
devices
Insulation characteristics
System analysis
Representative voltages and

overvoltage Urp
(1)
Performance criterion
Statistical distribution (+)
Inaccuracy of input deta (+)
(+) Effects combined in a co-ord ination
factor Kc
Selection of the insulation meet ing

the performance criterion
Co-ordination withstand voltages Ucw

(2)
Altitude correction factors Ka
(or at mospheric correction factors
Equipment test assembly *)

Dispersion in production *)
Quality of installation *)
Ageing in service *)
Other unknown factors *)
*) Effects combined in a safety
factor Ks
Application of factors to account for

the differences between type test
conditions and actual
service conditions
Required withstand voltage Urw

(3)
Test conditions
Test conversion factor Ktc
Standard withstand voltages
Selection of rated withstand voltages

or standard rated withstand voltages
Uw fro m the lists
Rages of Um
Rated or standard insulation level : set of Uw
NOTE In brackets the subclauses reporting the definition of the term or the description of the action.
Sided bo xes refer to required input
Sided bo xes refer to performed actions
Sided bo xes refer to obtained results
Figure 2.1.2 Flow chart for the determination of rated or standard insulation level
in IEC60071-1
NOTE: The definition of some terms in above figure as given in IEC60071-1 is summarized as follows;
Urp: Representative overvoltages: Overvoltages assumed to produce the same dielectric effect on the insulation as
overvoltages of a given class occurring in service due to various origins. They consist of voltages with the standard shape of the
class.
Ucw: Co-ordination withstand voltage: For each class of voltage, the value of the withstand voltage of the insulation
configuration in actual service conditions, that meets the performance criterion
Urw: Required withstand voltage: Test voltage that the insulation must withstand in a standard withstand voltage test to ensure
that the insulation will meet the performance criterion when subjected to a given class of overvoltages in actual service
conditions.
Uw: Standard rated withstand voltage: Standard value of the rated withstand voltage as specified in this standard. The rated
withstand voltage is value of the test voltage, applied in a standard withstand voltage test that proves that insulation complies
with one or more required withstand voltages. It is a rated value of the insulation of an equipment.
Kc: Co-ordination factor: Factor by which the value of the representative overvoltage must be multiplied in order to obtain the
value of the co-ordination withstand voltage.
Ks: Safety factor : Overall factor to be applied to the co-ordination withstand voltage, after the application of the atmospheric
correction factor (if required), to obtain the required withstand voltage, accounting for all other differences in dielectric strength
between the conditions in service during life time and those in the standard withstand voltage test
Page 11
TypethenameofTB
2.1.3 OVERVOLTAGES SPECIFIC TO UHV AC TRANSMISSION SYSTEM

Overvoltages which need to be considered in designing UHV transmission lines and substation equipment are
classified into four categories from the voltage characteristics as shown in Figure 2.1.3 (See Chapter 4). Each
shape of overvoltage is specified in IEC 60071 as shown Table 2.1.4.
Figure 2.1.3 Representative maxima of amplitude of overvoltages Um (pu)
Table 2.1.4 Classes and shapes of overvoltages, Standard voltage shapes and
Standard withstand voltage tests (IEC 60071)
2.1.3.1 T EMPORARY
OVERVOLTAGES
(TOV)
TOV includes healthy phase overvoltages due to transmission line ground faults and load rejections. In the case of
sudden load rejection on a heavily loaded, long line, such as a UHV system, the overvoltage is about 1.3 - 1.5 p.u.
This TOV is required not only to cover the peak voltage in the system, but also to cover the overvoltage generated
during their operation. Therefore power frequency withstand test was verified in both long time range and short
time range, because the voltage stress is different from both range as described in Chapter 8.4.
Page 12
TypethenameofTB
2.1.3.2 S LOW - FRONT
OVERVOLTAGES
(S WITCHING
OVERVOLTAGES )
The duration of wave front is about a few-hundred microseconds, such as the overvoltage in opening / closing
transmission lines and ground fault. This switching overvoltage has much influence on the insulation design of
towers, thus switching overvoltage is particular important for UHV systems because of the saturation effects of the
air insulation distance on the switching impulse strength.
As shown in Figure 2.1.4, for the 1100 kV voltage level, the flashover voltage of air-insulated gaps for switching
overvoltage has a tendency to saturate. Therefore, extremely high tower is required for air insulation. On the
contrary, to reduce the construction cost of UHV system, switching overvoltages can be reduced by adopting circuit
breakers with closing and/or opening resistors and higher performance arrester. Figure 2.1.4 shows the relation
between air insulation distance and switching overvoltage, and Figure 2.1.5 shows the comparison between the
double-circuit tower design based on 2.0 p.u. and 1.7 p.u.
Withstand Voltage (kV)
2000
(UHV-Switching OV: 2.0pu (same as 550kV))

(UHV-Switching OV: 1.7pu)
1500
(550kV-Switching OV:
2.0pu)
1000
500
3m
0
0
6m
9m
10
Air Insulation Distance (m)
Figure 2.1.4 Relation between air insulation and switching impulse withstand
voltages
ExpectinTra
g the
nsmission
ap plicaline
tion of 550k V
construtech
ctednology
this time
Tran sm ission lin e

constru cted this time
Figure 2.1.5 Size reduction of 1100kV tower
Page 13
TypethenameofTB
2.1.3.3 F AST - FRONT
OVERVOLTAGES
(L IGHTNING
OVERVOLTAGES )
Lightning strokes terminating on UHV transmission lines can generate overvoltages of several MV depending on
the front-steepness of the overvoltage and the height of the tower. Shielding failures as well as back-flashovers
have to be taken into account.
Lightning overvoltage is the predominant factor for substation equipment design. Therefore, lightning overvoltages
in the UHV substation are highly suppressed for size reduction within a rational level by installing several higher
performance surge arresters at adequate locations.
2.1.3.4 V ERY
FAST TRANSIENT OVERVOLTAGES
(VFTO)
The GIS disconnector, when switching a charging current, repeats restriking and generates VFTO, which can reach
up to approximately 3.0 p.u. At a UHV substation, lightning overvoltages are effectively suppressed by higher
performance surge arresters. Disconnector switching overvoltages are likely to exceed the lightning overvoltage if
no measures are taken to limit them. Therefore, the resistors can be a suppression measure for the VFTO.
2.1.4 SELECTION OF INSULATION LEVEL

Insulation coordination of substations and transmission lines can be achieved to set a reasonable insulation level
voltage without sacrificing supply reliability by installing higher performance surge arresters on specific locations in
substations, adopting resistor-fitted switching schemes of disconnectors and circuit breakers, and comprehensive
simulations and analysis of assumed overvoltage phenomenon. To select an appropriate insulation level and
insulation requirements for equipment, it is necessary to evaluate technical data of equipment and set reasonable
design margins to secure supply reliability while satisfying each projects design constraints, such as substation
type: open-air/Hybrid-IS/GIS type. Figure 2.1.6 shows substation designs and corresponding insulation levels
(LIWV and SIWV) of Chinese and Japanese projects.
Page 14
TypethenameofTB
TR
SW
LIWV SIWV
2250 kV 1800 kV
2400 kV 1800 kV
SW
LIWV SIWV
2400 kV 1800 kV
(a) Jindongnan substation (China)
(b) Nanyang switching station (China)
LIWV SIWV
TR
2250 kV 1800 kV
SW
2400 kV 1800 kV
(c) Jingmen substation (China)
LIWV SIWV
TR
1950 kV 1425 kV
SW
2250 kV 1550 kV
(d) Shin-Haruna testing site (Japan)
Figure 2.1.6 Substation designs and corresponding insulation levels (LIWV and
SIWV) of China and Japanese projects (TR: Transformer, SW: Switching equipment in
above captions)
Page 15
TypethenameofTB
2.2 Reduction of insulation levels using overvoltage suppression measures

The higher performance surge arresters, high voltage shunt reactors, resistor-fitted switching schemes of
disconnectors and circuit breakers have been utilized to suppress the overvoltages peculiar to UHV systems, and
to reduce insulation design level of each project.
2.2.1 OVERVOLTAGE SUPPRESSION WITH HIGHER PERFORMANCE SURGE ARRESTERS

The higher performance surge arrester, which has better protective performance (See Chapter 5.2), has been
utilized to suppress LIWV and SIWV. The reliability of higher performance surge arrester was confirmed throughout
its massive application in 550 kV systems, and it is recognized as an effective measure to suppress power system
overvoltages. Recent UHV projects in China and Japan employ higher performance surge arresters with highest
voltage of equipment of 1620 kV (1.80 p.u. at 20 kA) at 1100 kV system. On the other hand, a recent project in
India is developing an arrester with highest voltage of equipment of 1700 kV (1.74 p.u. at 20 kA) at 1200 kV system.
Typical locations of these higher performance arresters are transmission bays, busbars and transformer bays.
Table 2.2.1 shows the LIWVs and protective performance of arresters in recent projects. Although each project
adopts different insulation levels due to differences in location of arresters and substation types, all projects
succeeded in reducing insulation voltage level ranges: 1950 kV-2250 kV for transformers and 2250 kV-2400 kV for
switchgears.
Highest voltage (kV)

Type of substation
Residual voltage (@20 kA) (kV)
Transformer
LIWV (kV)
GIS and others
Italy
1050
GIS
1800
2250
2250
Japan
1100
GIS
1620
1950
2250
China
1100
GIS, Hybrid-IS
1620
2250
2400
India
1200
AIS
1700
2250
2400
Table 2.2.1 Protective performance of surge arresters in substation projects
Figure 2.2.1 Dead tank-type higher performance surge arrester (Japan, China)
Page 16
TypethenameofTB
Figure 2.2.2 Porcelain type
Figure 2.2.3 Porcelain type
higher performance surge
higher performance surge
arresters (China)
arresters (India)
2.2.2 RESISTOR-FITTED CIRCUIT BREAKERS

To suppress the switching overvoltage, pre-insertion resistor is employed for UHV circuit breakers. Chinese and
Indian UHV projects introduce resistor-closing technique, while Japanese project introduces resistor-closing /
opening technique. Both techniques suppress switching overvoltages of transmission lines to below 1.7 p.u. The
resistance of this switching scheme is usually between 500-700 depending on the size of UHV system and its
characteristics. Table 2.2.2 shows the insulation coordination of several UHV projects: (a) Closing overvoltage in
Indian project and, (b) Opening overvoltage in Japanese project
Figure 2.2.4 shows the study example of the relation between switching resistance and overvoltage suppression
effects in Indian and Japanese project, and Figure 2.2.5 is an example of GCB with pre-insertion resistors.

Suppression of switching
overvoltage
Switching overvoltage
insulation design level (p.u)
Italy
1050
MOSA
Closing &
opening R
(500 )
Japan
1100
MOSA
Closing &
opening R
(700 )
1.7
1.6 / 1.7
China
1100
India
1200
MOSA
Closing R
(600 )
MOSA
Closing R
(600 )
1.7
1.7
Table 2.2.2 Suppression methods and insulation designs of international projects

[5]
Page 17
TypethenameofTB
1.83
1.60
1.52
1.42
(a) Closing case in Indian project
(b) Opening case in Japanese project
Figure 0.4 Example of relation between resistance and overvoltage
Figure 2.2.5 GCB with pre-insertion resistors

2.2.3 THE DAMPING EFFECT OF THE RESISTOR-FITTED DISCONNECTORS EMPLOYED IN
GIS AGAINST THE SWITCHING OVERVOLTAGES.
In gas insulated substations, the resistor-fitted disconnectors are commonly utilized to suppress switching
overvoltages. Examples of applications of resistor-fitted disconnectors are shown in the Table 2.2.3 below. The GIS
system with fast-operating disconnectors can suppress the disconnectors overvoltage levels from 2.8 p.u. without
the resistors to less than 1.3 p.u. with pre-insertion resistors.

Type of substation
Pre-insertion resistor ()
Italy
1050
GIS
110
Japan
1100
GIS
500
China
1100
GIS
Hybrid-IS
500
None
India
1200
AIS
None
Table 2.2.3 Application of pre-insertion resistor in international projects
Page 18
TypethenameofTB
R esistor
M ovable
contact
Figure 2.2.6 Disconnector with pre-insertion resistor

2.2.4 OTHER MEASURES FOR THE REDUCTION OF INSULATION LEVEL
High voltage shunt reactors can be applied on long UHV transmission lines with adequate compensation degree
(generally in the range from 70% to90%) to maintain reactive power balance and suppress TOV below 1.4p.u.
Controlled switching and line arresters can be utilized as a mitigation measure for insulation level reduction,
although they have not been commercially applied to UHV systems.
References
[1] Eiichi Zaima, C.Neumann, Insulation Coordination for UHV AC Systems based on Surge Arrester Application
(CIGRE C4.306), IEC-CIGRE Second International Symposium on Standards for Ultra High Voltage Transmission
[2] Guagfan Li, Jianbin Fan, The experience of UHV substation facilities in China, International Conference on
Development of 1200kV National Test Station
[3] IEC 60071-1 Ed. 8.1, Insulation co-ordination - Part 1: Definitions, principles and rules, 2011
[4] IEC 60038, Standard voltages, 2009
[5] CIGRE brochure No.400, Technical Requirements for Substation exceeding 800kV, 2010
Page 19
TypethenameofTB
3 Recent practice on insulation coordination for the UHV and the

800 kV system
3.1 China UHV project
3.1.1 CHINA 1000 KV AC TRANSMISSION SYSTEM
The schematic diagrams of UHV transmission systems are shown in Figure 3.1.1 and Figure 3.1.2. The UHV AC
transmission test and pilot project (JindongnanNanyangJingmen) was put into operation in January 2009, as
shown in Figure 3.1.1. The planning construction Huainan-Shanghai UHV double-circuit tower arrangement
transmission project is shown in Figure 3.1.2.
Figure 3.1.1 Wiring schematic diagram for the China 1000 kV AC transmission test
and pilot project (single circuit) system
Figure 3.1.2 Schematic diagram for the China 1000 kV AC South Anhui-Shanghai
double-circuit tower arrangement transmission system
3.1.2 POWER FREQUENCY TEMPORARY OVERVOLTAGE (TOV) AND THE PARAMETER
SELECTION FOR METAL OXIDE ARRESTERS (MOA)
3.1.2.1 A MPLITUDE
AND DURATION OF
TOV
As for calculating the TOV, the following fault types shall be mainly taken into account.
Page 20
TypethenameofTB
As for the single-circuit transmission lines, two kinds of failures shall be usually taken into account, namely, the
load rejection under the normal operation and the load rejection in the case of the line single-phase grounding
failure.
As for the double-circuit tower arrangement transmission line, the double-circuit operation or one circuit out service
/the other circuit operation shall be considered and the failures causing double-circuit 6-phase load rejection shall
be taken into account.
Most of China 1000 kV lines are relatively long and the high voltage shunt reactors are generally installed in the
lines. The largest TOV may generally occur in the single-phase grounding fault case, in which the fault line circuit
breaker at one line side shall be three-phase tripped and the circuit breaker at the other line side shall not be
tripped. The above failure case may occur under the following two situations:
(1) During the normal operation process, the line single-phase is grounded and the single phase reclosing is not
successful the three-phase circuit breaker is tripped by relay protection.
(2) During the line live working process, the single-phase reclosing shall be required to withdrawal; at this time, the
single-phase grounding fault occurs and the line three-phase circuit breaker are also tripped.
The main measure to limit the power frequency overvoltage is to install the line high voltage shunt reactor.
The maximum TOV shall be allowed no more than 1.4 p.u. at the line side and the maximum TOV shall be allowed
no more than 1.3 p.u. at the bus side in China.
The TOV duration may play an important role in the choice of the arrester rated voltage and the equipment
insulation level. The relay protection mode is adopted in which the UHV line both side circuit breakers are tripped
synchronously, so as to shorten the duration of TOV and lower the energy absorbed by MOA. The maximum trip
delay for the both side circuit breakers shall generally be controlled within 0.2 seconds; the TOV duration shall be
no more than 0.5 seconds even if it is considered that the one side circuit breaker is failure to trip and the tripping
shall be carried out by the back standby circuit breaker.
3.1.2.2 MOA
PARAMETER SELECTION
In the past, the traditional MOA rated voltage selection principle was UnTOV; whereas, the traditional MOA rated
voltage selection principle has been broken through in China UHV project, namely, that Un is allowed to be less
than TOV. The MOA rated voltage Un of the UHV system was selected as 828 kV, which is equivalent to 1.3 p.u.
and is less than the maximum TOV (1.4 p.u.) of the UHV lines. Because the MOA is with the good short time power
frequency overvoltage withstand capacity, the Un being lower than TOV for short time shall be permitted. According
to the test data from the China MOA manufacturers, the TOV duration for MOA withstanding 1.4 p.u. is 10 seconds.
The main electrical parameters for MOA (Un=828 kV) are listed in Table 3.1.1.
The energy absorption allowable value is 40 MJ. The calculation results may show that the MOA actual maximum
absorption energy shall be less than 10 MJ while the maximum TOV duration is 0.5 seconds and under the
switching overvoltage cased by two times closing operation. Therefore, the Un is selected as 828 kV and there shall
be relatively great margin for MOA absorption energy.
The lower MOA rated voltage causes the MOA residual voltage to be lowered so as to lower the substation
overvoltage amplitude and the requirement for the equipment insulation level; as plays some certain role for
lowering the line overvoltage.
Page 21
TypethenameofTB
System
nominal
voltage
Installation
location
Rated
voltage
(RMS)
Continuous
operation
voltage
(RMS)
Switching impulse
residual voltage under
30/60 s and 2 kA
Lightning impulse
residual voltage under
8/20 s and 20 kA
1000
The line side, the

bus side and the
transformer side
828
638
1460
1620
Table 3.1.1 Switching overvoltage design level of transmission line

3.1.3 SWITCHING OVERVOLTAGE
The following 3 categories of switching overvoltage shall mainly be taken into account:
(1) The closing and reclosing no-load line overvoltage;
(2) The ground fault overvoltage;
(3) The clearing short-circuit fault overvoltage caused by circuit breakers tripping
3.1.3.1 C LOSING
AND RECLOSING NO - LOAD LINE OVERVOLTAGE
The closing and reclosing line overvoltage may play the control role to the insulation design of China 1000 kV lines.
The main measures to limit the closing and reclosing line overvoltage are that the closing resistor is installed on
line circuit breakers. The closing resistor is taken as 600 and the closing resistor pre-insert time is 9.51.5 ms.
The maximum phase-to-ground statistical switching overvoltage along the line shall be no more than 1.7 p.u. for
China 1000 kV lines; the substation maximum phase-to-ground statistical switching overvoltage shall be no more
than 1.6 p.u. and the maximum phase-to-phase statistical switching overvoltage shall be no more than 2.9 p.u. The
maximum statistical switching overvoltage at the substation bus side shall be no more than 1.55 p.u.
The front time of the 1000 kV line closing and reclosing line overvoltage shall generally be above 1000-3000 s,
which may greatly influence the air clearance selection of transmission line tower.
3.1.3.2 G ROUND
FAULT OVERVOLTAGE
The single-phase grounding fault type shall be taken into account for the calculation to the ground fault overvoltage
of the China UHV systems; the overvoltage amplitude is relatively low; the maximum 2 % overvoltage shall be
lower than 1.51 p.u.. As for the China UHV lines, the ground fault overvoltage shall not be the control factor to
determine the line insulation level.
3.1.3.3 C LEARING
FAULT OVERVOLTAGE
This clearing fault overvoltage means the overvoltage occurred in the adjacent non-fault lines while the short-circuit
fault in the fault line is cleared up.
The fault type may significantly influence the overvoltage amplitude. The clearing single-phase grounding fault
overvoltage amplitude shall be within the allowable range. The clearing 2-phase or 3-phase grounding fault
overvoltage may be relatively high and the overvoltages in some lines may exceed the allowable values. The
opening resistor may be installed in the line circuit breaker so as to lower the clearing fault overvoltage.
The following situations shall be taken into account:
(1) The opening resistor may absorb great energy and the operation may be complicated so that not only the cost
of the circuit breaker may be increased but also the probability of the circuit breaker failure may be increased.
(2) The probability of the 2-phase or 3-phase grounding fault shall be very low.
Page 22
TypethenameofTB
(3) The maximum overvoltage may occur on lines rather than in substations. It may cause the line insulation
flashover; however, the substation equipment may not generally be damaged by line overvoltage.
As for the JindongnanNanyangJingmen UHV transmission line, the opening resistor is not necessary to be
installed.
3.1.3.4 E NERGIZING
UNLOADED TRANSFORMER OVERVOLTAGE
As shown in the research and field tests in China, the 500 kV circuit breaker is not equipped with the closing
resistor for the energizing unloaded UHV transformer at the 500 kV side; there shall be no any relatively high
resonance overvoltage. The inrush current and the overvoltage are within the allowable range.
The possibility of the resonance overvoltage and the inrush current from the energizing unloaded UHV transformer
at the 1000 kV side may be greater than that at the 500 kV side. The closing resistor may be adopted so as to be
beneficial to lowering the resonance overvoltage and the inrush current; however, the closing resistor may not be
valid for all system construction and operation modes; moreover, the closing resistor may increase the equipment
cost and cause the switching operation mechanism to be complicated as well as lower the reliability.
Under the normal circumstances, the energizing unloaded UHV transformer at the 500 kV side shall be provided.
3.1.4 VERY FAST FRONT TRANSIENT OVERVOLTAGE (VFTO)

The GIS disconnector switching may generate the VFTO whose wave front is very steep and amplitude is very high
and which may damage three types of equipment insulations: (1) GIS body; (2) equipment with winding, such as a
transformer; (3) the secondary equipment.
As for the GIS UHV substation, this problem may be more remarkable. Because the higher the system rated
voltage is, the lower the ratio of the equipment lightning impulse withstand voltage LIWV and the system rated
voltage Un. In comparison the 1000 kV GIS substation with the 500 kV GIS substation, the rated voltage is
increased by 1 time, but the relative value of VFTO is basically the same; the absolute value of VFTO is
proportionally increased by 1 time with the rated voltage; however, the insulation level (LIWV) of the 1000 kV GIS
equipment is increased by 55 % in comparison with that of the 500 kV GIS equipment, which is not proportionally
increased. Therefore, VFTO may do more greatly harm to UHV GIS equipment than 500 kV GIS equipment.
Before the GIS rated withstand voltage is not determined under VFTO, we may temporarily adopt the GIS lightning
impulse withstand voltage LIWV as the GIS rated withstand voltage under VFTO;
The VFTO calculation research has been carried out by combining the UHV substation or switching station
characteristics in China; it is thought that not only the VFTO at the initial GIS layout of the substation or switching
station shall be calculated but also the VFTO at the long-term GIS layout; the substation or switching station GIS
layout (such as the bus length) may greatly influence the VFTO amplitude. Thus, the following viewpoints may be
put forward:
(1) The shunt resistor (whose resistance is 500 ) is required to use in the GIS substation so as to effectively limit
the VFTO.
(2) The maximum VFTO caused by the disconnector switching may be 2.15 p.u. in the Hybrid-IS substation or
switching station, which is not high and within the GIS insulation allowable range.
Therefore, the disconnector of the Hybrid-IS station may not be equipped with the shunt resistor; the differential
treatment is adopted so as to be safe and economical.
Because the transformer in the China UHV substation is not directly connected with GIS, which is connected
through the overhead lines and whose distance is far away; the VFTO is attenuated fast; the transformer VFTO
amplitude is relatively low (the maximum peak value is 925 kV) and the wave front time is increased. The
calculated wave front time is more than 1.5 s. Thus, there shall be no any danger to the transformer insulation.
Page 23
TypethenameofTB
3.1.5 SUBSTATION LIGHTNING INVADING OVERVOLTAGE CALCULATION AND THE

SUBSTATION ARRESTER LAYOUT
As for the insulation design of the UHV substation equipment, the lightning overvoltage may be predominant.
As for the calculation of the China UHV substation lightning invading overvoltage, the two relatively harsh
connection modes shall be taken into account, which may be listed as follows: (1). The single-line mode and the
line circuit breaker being tripped; (2). The single line + single bus + single transformer mode (as shown in Figure
3.1.3).
(a) single-line mode
(b) single line + bus + single transformer mode
Figure 3.1.3 Substation connection modes taken into account for the lightning
invading overvoltage calculation
Figure 3.1.4 Ground wires have been adopted in the entrance line section of the
UHV substation
The substation maximum lightning overvoltage may be caused by the lightning shielding failure invading wave in
the entrance line section of substation; two measures, namely, decreasing the maximum lightning shielding failure
current in the entrance line section [the entrance line ground wire protection angle is decreased to -4 and three
ground wires have been adopted in the entrance line section (as shown in Figure 3.1.4). and optimizing the arrester
layout, have been taken in China UHV AC transmission test and pilot project so as to decrease the lightning
shielding failure invading overvoltage. Finally the scheme with the small quantity of arresters has been adopted: 1
group of MOAs is installed in each circuit entrance location; 1 group of MOAs is installed in each bus section; 1
group of MOAs is installed beside the transformer. The overvoltage values may be different for various substations.
The typical values of the maximum lightning overvoltage of the equipment may be listed as follows: 2040 kV for
Page 24
TypethenameofTB
GIS, 1854 kV for the shunt reactance and 1796 kV for the transformers. The lightning impulse withstand rated
voltage of the transformer and the shunt reactance is 2250 kV and lightning impulse withstand rated voltage of
other devices is 2400 kV in China. The allowable values of the equipment lightning impulse insulation levels shall
be more than the maximum lightning invading overvoltage, which shall be meet the requirements of the internal
insulation margin (15 %) and the external insulation margin (5 %). As for the single-line mode, the internal
insulation margin may be lowered to 10 % because its occurrence probability is very small.
As for the calculation of the substation lightning invading overvoltage, the interval statistics method may be used
besides the deterministic method; moreover, the substation lightning MTBF (mean time between failures) shall be
required to be more than 1500 years.
3.1.6 LINE LIGHTNING PROTECTION

By taking the importance of the UHV lines as well as the characteristics of the UHV line high insulation level, the
expected lightning trip rate for the 1000 kV lines shall be lower than that for the 500 kV lines (according to the 500
kV operation experiences, statistic lightning trip rate value is 0.14 times/100kma), which may be 0.1 times/100kma
according to 70 % of the lightning trip rate for the 500 kV lines.
As shown in the operating experiences, the line insulation level shall be increased along with the transmission line
voltage level; the lightning back flash-over failure trip rate shall account for the less, of the total lightning trip rate,
which may account for less than 10 % of the total lightning trip rate in China 500 kV transmission lines. As shown in
the calculation results of the China 1000 kV line lightning protection, the lightning back flash-over of the line
insulation may basically not occurred ; the main cause to give rise to the lightning flashover shall be lightning
shielding failure. Therefore, the key for the UHV line lightning protection shall be against the lightning shielding
failure.
The main method for calculating the line trip rate is the improved electrical geometric model (EGM), in which the
influencing factors such as the terrain along the line, the correction coefficient of the lightning striking distance to
the earth as well as the probability distribution of the lightning leader incident angle shall be taken into account; at
the same time, the study on the line lightning shielding failure trip-out rate calculated by utilizing the leader
propagation model (LPM) has been carried out in China. Because some parameters and criteria used in the
calculation by various units may be different, the calculation results may be quite different; moreover, these
parameters and criteria have been lack of sufficient base for the time being. Therefore, the calculation result from
the improved electrical geometric model (EGM) shall be the primary base; on the other hand, the calculation result
from the leader propagation model (LPM) shall be a reference.
The main measure for lowering the lightning shielding failure trip-out rate shall be to lower the ground wire
protection angle ; and terrain along the line shall greatly influence the lightning shielding failure trip-out rate.
Based on the relevant researches, the following regulations have been applied to the ground wire protection angle
in China 1000 kV lines, which may be listed as:
(1) As for the single circuit transmission lines: <6 for the plain area and <-4 for the mountain area.
(2) As for the double-circuit tower arrangement transmission lines: <-3 for the plain area and <-5 for the
mountain area.
(3) As for the jumping-lines of the line strained angled towers: 6 for the single circuit transmission line in the
plain area and 0 for the single circuit transmission line or the double-circuit tower arrangement transmission line
in the mountain area.
The above described the protection angle between the ground wire and the conductor means the angle between
the perpendicular line of the horizontal surface from the ground wire at the tower and the connecting line from the
ground wire to the most outer sub-conductor.
Since China 1000 kV UHV AC transmission test and pilot project (JindongnanNanyangJingmen) was put into
operation in January 2009, there has yet been no any line lightning trip-out fault.
Page 25
TypethenameofTB
3.1.7 INSULATION COORDINATION
3.1.7.1 P RINCIPLES
OF INSULATION COORDINATION
By combining the research results on the China UHV overvoltage and internal and external insulation
characteristics and according to the safety and economy principals, the insulation level may be determined.
As for the China insulation coordination principle, some certain insulation safety margin has been required to
reserve; and the internal and external insulation safety margins are 15 % and 5 %, respectively.
As for the air clearance of UHV AC line and substation, discharge voltage test at power frequency voltage, lightning
and switching impulse voltage have been relatively systematically and comprehensively carried out with the 1:1
tested objects in China and a series of discharge voltage curves have been obtained; as may provide the
reasonable basis for the insulation coordination.
3.1.7.2 S UBSTATION
EQUIPMENT INSULATION LEVEL
The insulation levels of the main equipments in China UHV substation are shown in Table 3.1.2.
On the whole, the insulation levels of the main equipments in China UHV substation may be lower than those of the
Russia UHV equipment, but they may be higher than those of the Japan UHV equipment. It is determined by
combining the China UHV overvoltage level and the corresponding equipment manufacturing experiences; and it
shall be suitable for the Chinese situation.
(Unit: kV)
Power-frequency shortduration withstand voltage
Lightning impulse
withstand voltage
2250
(chopped impulse:
2400)
Switching impulse
withstand voltage
1800
1100 (5 min)
2400
1800
1100 (1 min)
2550
1800
1100 (1 min)
Voltage transformer (CVT)
2400
1800
1300 (5 min)
Bushing (transformer and reactor)
2400
(chopped impulse:
2760)
1950
1200 (5 min)
Bushing (GIS)
2400
1800
1100 (1 min)
Switching device longitudinal

insulation
2400+900
1675+900
1100+635 (1 min)
Equipment
Transformer and reactor
GIS (circuit breaker and
disconnecting switch)
Post insulator and disconnector
(open type)
Table 3.1.2 The rated insulation withstand voltage for 1000 kV equipment
Page 26
TypethenameofTB
3.1.7.2.1 T RANSFORMER
INSULATION LEVEL
The insulation level of China UHV transformers may be shown in Table 3.1.3.
(Unit: kV)
Country
China
Lightning impulse
withstand voltage
2250
Switching impulse
withstand voltage
1800
Power-frequency
withstand voltage
1100 (5 min)
Table 3.1.3 The insulation level of China UHV transformers

The lightning impulse withstand voltage and the power frequency withstand voltage may play the decisive role for
the UHV transformer insulation.
China UHV MOA is with good V-A curve saturation characteristics and the low lightning impulse protection level;
the measure for lowering the maximum shielding failure lightning current at the entrance line section; the lightning
overvoltage at the terminals of the transformer is not high. Therefore, the safety margin may be sufficient for
selecting 2250 kV as the lightning impulse withstand voltage.
As shown in the operating experiences, the majority of transformer failures occurred under operating voltage. The
power-frequency short-duration withstand voltage test shall be designed to inspect whether the partial discharge
(being considered to be the precursor phenomenon of the dielectric breakdown) exists and verify the insulation
strength as well as check the insulation aging characteristics to some extent. China experts advocate that the
power frequency withstand voltage test should be considered severely; the rated power frequency withstand
voltage should take as 1100 kV and the duration time for 5 minutes; it is more strict in comparison with the
recommendation value (1 min) for the UHV transformer from IEC standard
The switching impulse withstand voltage of China UHV transformers is 1800 kV.
3.1.7.2.2 L ONGITUDINAL
INSULATION TESTING VOLTAGE OF THE CIRCUIT BREAKER AND THE
DISCONNECTOR
The longitudinal insulation lightning impulse testing voltage of China UHV circuit breaker and disconnector should
be 2250 + 900 (kV), in which 900 kV is the peak value of opposite polarity working voltage,1100 2 / 3 =Un-m=
900 kV.
The amplitude of the opposite polarity power frequency component recommended in IEC60071-1 is 0.7Um 2 /
3 namely the amplitude of the working voltage should be multiplied with the coefficient 0.7.
In China national standard Insulation Coordination for High-voltage Transmission and Distribution Equipment
(GB311.1), the coefficient is stipulated among 0.7-1.0, which should also be determined together with the electric
power companies and the manufacturers.
If this coefficient is 0.7, in 1 cycle of the working voltage (the 360 phase), there is about 1/4 cycles not being
included. Therefore, the guarantee probability is 0.75.
As a result of the importance of the UHV equipments, China advocate that the amplitude of the opposite polarity
power frequency component is 11100 2 / 3 =900 kV while the longitudinal insulation lightning impulse tests for
the UHV circuit breaker and disconnector are carried out.
3.1.7.3 S UBSTATION
MINIMUM AIR CLEARANCE
The A values of the substation minimum air clearance shall be divided into three categories: A1' value, A1" value
and A2 values.
Page 27
TypethenameofTB
The A1' value is the minimum electrical distance of the substation conductor to the frame; the A1" value is the
minimum electrical distance of the substation equipment to the frame; and A2 value is the minimum electrical
distance between the phases in the substation.
As for the areas whose altitude being no more than 1000 m, the minimum air clearance for the 1000 kV substation
are listed in Table 3.1.4.
(Unit: kV)
The action voltage type
A1 value
A1
Power frequency
Switching impulse
Lightning impulse
A2 value
A1
4.2
6.8
7.5
5.0
6.8
10.1 (ring - ring)
9.2 (four sub-conductors - four subconductors)
11.3 (tubular bus - tubular bus)
5.5
Table 3.1.4 The minimum air clearance for the 1000 kV substation
3.1.7.4 L INE
INSULATION LEVEL
The contaminated insulator withstand voltage method or the specific creepage distance method may be used to
determine the insulator number.
With comprehensively considering the insulation margin and the simplified design and other factors, the insulator
configuration for the common lines in China UHV AC transmission test and pilot project may be listed as follows:
(1) he suspension string: 54 double-shed disk-type insulators (for 300 kN) are used in the II grade polluted area;
the composite insulator whose structural height is 9750 mm and creepage distance is 30300 mm are used in the III
and IV grade polluted areas.
(2) he tension string: the 44, 54 and 60 disk-type insulators (550 kN and creepage distance is 700 mm) are used in
II, III and IV grade polluted areas, respectively.
The minimum air clearance values for China 1000 kV lines are listed in Table 3.1.5.
As for the cup type-towers or cat-head type towers in the single-circuit line, the middle phase minimum air
clearance is controlled by the switching overvoltage; on the other hand, the out phase minimum air clearance is
controlled by the power frequency voltage under the weather condition of gale rather than the lightning overvoltage.
Therefore, there shall be no any regulation on the minimum air clearance under the lightning impulse.
However, the towers of the UHV double-circuit tower arrangement transmission line may be very high so as to
increase the attracting lightning area and weaken the conductor shielding effect from the earth; thus, the line may
easily stricken by the lightning. Because too large lightning current is not able to cause the conductor shielding
failure, so that the line maximum lightning shielding failure current amplitude is limited, the insulation level of the
UHV double-circuit tower arrangement transmission line may be enhanced appropriately so as to effectively lower
the lightning shielding failures. The main measure to enhance the line insulation level is to increase the distance
from the conductor to its lower cross arm or the tower body; especially the former may significantly enhance the
line insulation level. Therefore, the minimum clearance distance from the line lightning impulse voltage to the tower
has been regulated for 1000 kV double-circuit tower arrangement transmission line.
While the discharge voltage within the air clearance under the switching impulse voltage is determined, 1000 kV
line real type towers are tested in order to take fully account of the influence to the gap discharge voltage from the
tower width; The wave front time of the switching impulse test voltage is 1000 s.
Page 28
TypethenameofTB
The influence to the discharge voltage from the parallel connected number of gaps shall be taken into account in
the insulation coordination and the entire line insulation flashover rate under the switching overvoltage has been
calculated; the variation of the calculated switching overvoltage values at various locations along the line and the
actual distribution shall be taken into account. The entire line insulation flashover rate under the switching
overvoltage shall be no larger than 0.01 times/year.
(Unit: m)
The action voltage type
The type of lines
The minimum clearance/m

2.7 (altitude: 500 m)
Single-circuit and 1-tower
Power frequency voltage
double-circuit
Out phase: 5.9 middle phase: 6.7/7.9 (altitude: 500 m)
Single-circuit
Switching impulse
1-tower double-circuit
Single-circuit
No regulating
Plain area: 6.7 m; mountain area: 7.0 m (altitude: 500 m)
Lightning impulse
1-tower double-circuit
Table 3.1.5 The minimum air clearance in China 1000 kV lines

3.2 India UHV project
3.2.1 FEATURES OF POWERGRIDS 1200 KV TRANSMISSION SYSTEM
Major demand centres in Indian Power System are located towards Western and Northern parts of the country. On
the other hand, generation pockets are located mainly in the Eastern part. In order to facilitate power transfer from
Eastern to Western part of the country as well as to address right of way issue, high capacity East-West
transmission corridors comprising 765kV and 1200kV AC have been planned. Further, in the central part of India,
Wardha is a gateway for power transfer towards Northern and Western part. In view of the above, high capacity
765kV transmission corridor has been planned upto Wardha from Eastern part, whereas beyond Wardha, a hybrid
1200kV and 765kV transmission corridor i.e, Wardha Aurangabad Padghe has been planned towards Western
part. The 1200kV corridor includes Wardha Aurangabad 1200kV one circuit (about 380km) in parallel with 4 nos.
765kV lines. However, initially due to less power transfer requirement over this corridor, the 1200kV line is
proposed to be operated at a 400kV double circuit i.e, line insulation at 1200kV level with terminal equipment at
400kV level, which is likely to be commissioned by 2013. Subsequently, with the increased power transfer
requirement, this corridor shall be charged at rated voltage, sayby 2016-17. The power map showing proposed
1200kV transmission corridor is as under.
Page 29
TypethenameofTB
WR MAP SHOWING FUTURE

1200kV AND 800kV LINES
Figure 3.2.1 Power Map of Western Region showing proposed 1200kV corridor in
India
The connectivity diagram for the 1200kV corridor with 765kV system is shown in Figure 3.2.2.
Figure 3.2.2 Single Line Diagram showing 1200 kV corridor
Page 30
TypethenameofTB
3.2.2 INSULATION DESIGN FOR TRANSMISSION LINE

The nominal voltage of 1200kV system is considered as 1150kV. The line reactor size has been chosen as 660
MVAR (2 x 10 x 3) considering the maximum size of single phase reactor as 110 MVAR. This line reactor at each
end of the line amounts to about 50% compensation of the reactive power. The remaining compensations are
provided at the bus with 2 x 330 MVAR (2 x 3 x 110 MVAR, 1-ph) bus reactor.
The Switching Overvoltage studies are carried out under various source strength and the overvoltage value having
2% probability of being exceeded has been obtained as 1.7pu. To contain the overvoltage closing resistor of value
600 ohm has been considered which comes into the circuit initially for 10ms. The overvoltages are also contained
with the help of line surge arrestors while the temporary faults are taken care by the single phase auto reclosing
scheme.
The above information is tabulated below :
Design level (p.u.)
1.7 p.u.
Suppression measures for switching overvoltage
Closing resistor of 600 (10 ms insertion time)

and line surge arrester
Switching overvoltage (maximum and / or 2 %)

(p.u.) and occurrence case
1.7 p.u. for line reenergization
Types of overvoltage to be studied
line energization/reenergization/
de-energization, line fault clearing, Reactor
Switching, Transformer Energizing and tripping
on external fault as well as on no load
Shunt reactor: installed or not ?
Yes
Reclosing scheme
Single-phase reclosing

3.2.3 HIGHER PERFORMANCE SURGE ARRESTER
In order to determine the lightning impulse withstand voltage level (LIWV) and switching impulse withstand voltage
level (SIWV) for 1200 kV equipments, the Surge Arrester has been chosen with great care. Studies were
conducted for Voltage-Current (V-I) characteristics of ZnO blocks of the Surge arresters to achieve necessary
protective margins as per IEC-60071-1&2. The rated voltage of the Surge Arrester for 1200 kV is chosen as 850
kVrms, the corresponding V-I characteristics are given in Table 3.2.2 below:
Sr. No.
Surge Arrester Current (kA)
Residual Voltage (kVp)
1
2
3
4
5
0.5
1.0
2.0
10.0
20.0
1380
1440
1500
1600
1700
Table 3.2.2 V-I Characteristics of the Surge Arresters

The Surge Arrester Rated Voltage of 850 kVrms shall take care of power frequency voltage up to 1.15 p.u. for 10
seconds and 1.4 p.u. for about 1 second. From the V-I characteristics, the Switching Impulse Protective level
(SIPL) at 2.0 kA is 1500 kV, while the Lightning Impulse Protective Level (LIPL) at 20.0 kA is 1700 kV.
The following table shows the main characteristics for surge arresters.
Page 31
TypethenameofTB
Rated voltage (kV)

Residual voltage
(kV)
850 kV
Switching impulse
1500 kV at 2 kA
Lightning impulse
1700 kV at 20 kA
Table 3.2.3 Metal Oxide Surge Arresters (MOSA) main characteristics

3.2.4 INSULATION DESIGN FOR SUBSTATION
3.2.4.1 D ETERMINATION
PROCESS FOR
SIWV
AND
LIWV
The location of Surge Arresters is very critical in order to contain voltage rise due to Switching Surges as well as
Lightning Surges. For adequate protection of Equipments against Lightening surges, it was decided to place Surge
Arresters at Line Entrance and near Transformers/ Reactors. In addition, Surge Arresters in the Bus may also be
considered necessary.
The values of SIWV and LIWV are obtained considering adequate margin over the values of SIPL and LIPL as
obtained from V-I characteristics of the Surge Arrestor. The SIWV for 1200kV equipments has been considered as
1800kV with about 20% protective margin. The LIWV has about 35% protective margin for 1200kV equipments and
about 25% protective margin for 1200kV transformer and reactors. The values of LIWV for 1200kV equipments and
transformers/reactors have been considered as 2400kV and 2250kV respectively.
The following tables show us the LIWV and SIWV for 1200 kV equipments.
Substation type and MOSA layout
Substation Type (GIS, MTS, AIS)
MOSA layout
(unit)
AIS & MTS
Line entrance
Yes
Busbar
Yes (location under consideration)
Transformer
Yes
Determination process for LIWV

Representative
Value (kV)
Lightning overvoltage
1700 kV
Calculation method
Safety factor
Simplified method based on protection level V20kA

of surge arrester
Insulation Coordination being carried out keeping
IEC provisions in view
Transformer
2250 kV
Other equipment
2400 kV
LIWV (kV)
Table 3.2.4 Determination Process for LIWV
Page 32
TypethenameofTB
Switching
overvoltage
Representative Value
(Max and/or 2 % value)
(kV, p.u.)
Transformer
1500 kV (1.53 p.u.)
Other equipment
1500 kV (1.53 p.u.)

Simplified method based on
protection level V2kA of surge
arrester
Insulation Coordination being
carried out keeping IEC
provisions in view
Calculation method
Safety factor
Atmospheric correction factor, altitude (m)
SIWV (kV)
(phase-to-ground)
SIWV (kV)
(phase-to-phase)
NIL, Less than 1000 m
Transformer
1800 kV
Other equipment
1800 kV
Transformer
2970 kV
Other equipment
2970 kV
Phase-to-phase clearance of air gaps (m)
12.3 m (rod-cond.)
Table 3.2.5 Determination Process for SIWV

3.2.4.2 TOV
AND ENERGY ABSORPTION BY SURGE ARRESTER
For determination of temporary over voltages, preliminary studies have been carried out with single-line-to-ground
fault followed by three phase interruption at receiving end. The studies have been carried out with different source
strengths and reactive compensation as mentioned above. The study result showed a value of TOV as 1.33 p.u.
with 10000 MVA short circuit level at the sending end. For the purpose of insulation coordination study the value of
TOV is taken as 1.4 p.u..
To determine the discharge capability of the surge arrester, the most severe condition of operation has been
considered. The series of events start with single line to ground fault followed by opening of local and remote end
CB of the faulted phase. The single phase auto recloser becomes effective after a dead-time of 1000 ms. However
the reclosing becomes unsuccessful and the breaker could not be opened due to struck breaker condition. The
sequence of events is shown at Figure 3.2.3.
Steady
T=0ms
T=100ms T=120ms
T=1120ms T=1140 ms T=1260 ms T=1560 ms
Dead Time
1000 ms
SLG fault
Opening of
local end
CB of
faulted
Opening of
remote end
CB of
faulted
Stuck
breaker
Closing
of local
end CB
through
SPAR
Closing
of
remote
end CB
through
Opening of
CB due to
permanent
fault
LBB
Operates
and clears
the fault
Surge Arrestor Energy accumulation
Figure 3.2.3 Sequence of Events for calculation of Surge Arrester energy

accumulation
Page 33
TypethenameofTB
The discharge capability required for surge Arresters for above conditions is given below:
Total energy = 2Long discharge (IEC Class-5) + TOV + Margins
= 25 MJ + 35 MJ + 10 MJ = 55 MJ
The information for TOV and Energy absorption by surge arrester is tabulated below:
Overvoltage on healthy phase in case of ground fault
1.4
Maximum TOV (p.u.) and its duration (sec)
1.4 p.u., duration 1 sec
Case and condition for studies
(corresponding to the energy

absorption of 55 MJ)
Energy absorption of surge arrester (MJ)
1-Ph-Ground fault followed

by load rejection
Calculation method
EMTP/ATP
Power frequency test voltage for substation equipment

Assumed overvoltage condition for power
frequency test
Transformer
1200 kV
Other equipment
1200 kV
Analysis program
EMTP/ATP
Table 3.2.6 TOV and Energy absorption by surge arrester

3.2.5 1200 KV TEST STATION
A number of countries have gone for development of UHV AC technology but UHV AC technology is still in
evolving stage and not available commercially. Neither any International Standards for UHV AC system are
available. Thus POWERGRID is setting up 1200kV Test Station at Bina (M.P.) in association with CEA (Central
electricity Authority), CPRI (Central Power Research Institute) and Indian manufacturers to gain experience of field
tests/ trials so that the results and feedback can be used for developing field proven equipment of 1200KV system
in India as well as to gain initial operational experience.
The test station contains two nos. 1200kV bays comprising 1200kV class equipment like Instrument transformers,
Circuit Breakers and surge arresters and two nos. of 1000MVA transformer bank each comprising three single
phase 400kV/1200kV, 333 MVA auto transformers. In addition to the above substation equipment, two nos. 1200kV
AC test lines about a one km. long (one single ckt & one double ckt.) are being constructed and will be charged
through these two 1200kV Bays to study their performance by conducting measurements of various line
parameters. In the existing 400kV system through a Loop-in loop-out (LILO) arrangement, power flow through
1200kV test station shall be established.
Page 34
TypethenameofTB
400kV line
400kV line
Satna line
400kV Bina Bus
1200kV line
1200/400kV Transformer
1200/400kV Transformer
To 400kV Satna line
Figure 3.2.4 Power flow from Satna to Bina will be diverted via 1200kV test station
1200kV Equipment has been indigenously developed by Indian Manufacturers. 1200kV equipment will be field
tested under full operating voltage for fine tuning and optimization of system parameters. Further, 1200kV Test
station will help to learn safety, Quality & operational requirement for UHV AC Technology.
3.3 TEPCO 1100 kV project

3.3.1 FEATURES OF TEPCOS UHV TRANSMISSION SYSTEM [1], [2]
Tokyo Electric Power Company (TEPCO) has been working to expand their 550 kV network since the mid-1970s,
but it is difficult to obtain multiple power transmission routes in Japan. Countermeasures against the increase in
short-circuit current due to network expansion were also required. To cope with these problems, TEPCO decided to
introduce UHV (1100 kV) transmission system with a capacity of 3 to 4 times that of the 550 kV network.
By 1999, TEPCO had already constructed UHV-designed double-circuit transmission lines that ran 240 km from
east to west and l90 km from north to south, totally 430 km, as shown in Figure 3.3.1. These lines are now
operated at 550 kV and they are planed to be upgraded to UHV.
UHV systems are required to transmit very large power transmission (maximum 13 GW/route) and to ensure high
reliability. Additionally, various new technologies had been developed for substations. To establish carefully these
technologies towards UHV upgrading, field-testing of substation equipment has been carried out since 1996
connecting to the actual grid.
Page 35
TypethenameofTB
Sea of Japan
Tokyo
o
Pacific Ocean
Figure 3.3.1 1100 kV transmission routes of TEPCO

3.3.2 CONCEPT OF INSULATION COORDINATION AND SPECIFICATIONS [3], [4]
For economical insulation design of transmission lines and substations, overvoltages generated in the system must
be suppressed (limited) to a reasonable level. Fig. 3.4.2 shows the concept of insulation coordination on a TEPCO
UHV system.
Overvoltage on transmission lines and in substations can be effectively controlled by the higher performance
surge arresters (Refer to item 5.2 to see more details regarding higher performance MOSA). This metal oxide
surge arrester (MOSA) is a key technology for UHV insulation coordination. As shown in Fig. 3.4.3, it has excellent
protection characteristics with a residual voltage of 1620 kV (1.80 p.u.) at 20 kA (V20kA), flatter V-I characteristics
than conventional arresters, longer life under high-voltage stress, and higher discharging capability.
Switching overvoltages on transmission lines should be limited to a value as low as possible because the
predominant factor to determine the size (dimensions) of a tower is the switching overvoltage. Closing and opening
overvoltages are limited to the level of the ground-fault overvoltage, which can not be effectively controlled.
Flatter V-I
Characteristics
Practical application of high performance

metal oxide surge arrester
Voltage
550kVMOSA
Reliable circuit breaker with

closing/opening resistance
LIWV (Substation)
(Transmission line)
TOV
GIS:
2250kV
Transformer: 1950kV
1.61.7pu
1.5pu
1100kVMOSA
1620kV
(1.8pu)
Higher Voltage
Stress
mA 10kA20kA
Current
Figure 3.3.2 UHV insulation coordination
Figure 3.3.3 V-I characteristics of UHV

higher performance surge arrester
(1 p.u. =1100 kV 2 / 3 )
Page 36
TypethenameofTB
3.3.3 INSULATION DESIGN FOR TRANSMISSION LINE [3], [5]
3.3.3.1 S WITCHING
OVERVOLTAGE AND ITS SUPPRESSION
The predominant factors that determine the size of towers are the switching overvoltages generated in the system
that include closing, opening, and ground fault overvoltages. The level of these switching overvoltages must be
mitigated, limiting them to a value as low as possible. Among these overvoltages, ground fault overvoltages are
generated independently of circuit breaker switching, and generally, there are no effective means of controlling
them, except those near substations that can be reduced by surge arresters (Figure 3.3.4). In contrast, closing and
opening overvoltages can be controlled effectively by the insertion of closing and opening resistors. TEPCO
decided to control these overvoltages below the ground fault overvoltage level by employing the closing/opening
resistor (700 ).
Overvoltage [ pu]
1.8
1.5
Ground fault overvoltages near substations

can be suppressed by surge arresters
1.0 Nishi
UN
Gunma
49km
Higashi
Yamanashi
Nishi
Higashi
Gunma Gunma
138km
44km
Minami
Iwaki
194km
Figure 3.3.4 Calculated ground fault overvoltage profile

3.3.3.2 I NSULATION
DESIGN FOR TRANSMISSION LINE
[3], [5]
The evaluation of switching overvoltages is important for the insulated design of a transmission line. In an analysis
of ground fault overvoltages, the maximum overvoltage levels were obtained, using locations and phases of faults,
voltage phases at the time of faults, ground fault resistance, etc., as parameters. In an analysis of closing
overvoltages, considering that the overvoltage level depends on the closing timing lags of the main and resistance
contacts of 3-phase circuit breakers, maximum values and overvoltage distributions were obtained from a Monte
Carlo simulation using closing timings of 6 contacts as parameters. System conditions chosen are mainly the 1circuit trial closing from the 2-citcuit unenergized state of the transmission line, which is expected to cause high
overvoltages, or the reclosing of the transmission line. Opening overvoltage analysis mainly focused on
interruptions of ground fault currents, which are expected to cause high overvoltages. The values of ground fault
current were taken as currents at the maximum level in each phase of system growth, ultimately ranging up to 50
kA, which is the interrupting current of the circuit breakers.
From the results of overvoltage analysis, the insulation design levels for switching overvoltage (basic for air
clearance design) were determined as 1.6 p.u. (2.6 p.u. between phases) for the shorter north-south route of the
UHV system (about 140 km) and 1.7 p.u. (2.8 p.u. between phases) for the longer east-west route (about 240 km).
Page 37
TypethenameofTB
Expecting
Transmission
the application
line of 550kV
constructed
technology
this time
Transmission line
constructed this time
Figure 3.3.5 Size reduction of 1100 kV tower

Design level (p.u.)
1.6 - 1.7 p.u.
Suppression measures for switching overvoltage
Closing/opening resistors of 700 and line surge arrester
Switching overvoltage (maximum and / or 2 %) (p.u.)

and occurrence case
Ground fault, closing, reclosing and opening

None
Reclosing scheme
Fast multi-phase reclosing within 1 sec
1.6 - 1.7 (maximum) p.u. for ground fault

3.3.4 HIGHER PERFORMANCE SURGE ARRESTER [3], [6]
Overvoltages at substations can be controlled by higher performance surge arresters (Refer to item 5.2 to see
more details regarding higher performance MOSA). This metal oxide surge arrester is a key technology for
insulation coordination. They have excellent protection characteristics, a residual voltage of 1,620 kV (1.80 p.u.) at
20kA (V20kA), with flatter voltage-current characteristics than conventional arresters, long-term reliability under
high-voltage stress, and higher discharging capability. It effectively controls overvoltages from switching
overvoltage to lightning overvoltage with a wide range of flatness from low-current to high-current. Figure 3.3.6
shows a comparison of voltage-current characteristics between a higher performance metal oxide surge arrester
and a conventional arrester model.
Rated voltage (kV)

Residual voltage (kV)
826 kV
Switching impulse
Lightning impulse
1620 kV at 20 kA
Page 38
TypethenameofTB
Figure 3.3.6 V-I characteristics of UHV higher performance surge arrester

3.3.5 INSULATION DESIGN FOR SUBSTATION [3], [4], [7], [8]
3.3.5.1.1 L IGHTNING
PROCESS FOR
LIWV
OVERVOLTAGE STUDY
To analyze overvoltages accurately generated in the system, it is essential to determine a suitable system model.
Adding new knowledge to already established techniques, the study here prepared models that are as accurate as
possible. The main ones are as follows:
Lightning strike current:
200kA, 1/70 s ramp waves for tower back flashover
30kA, 1/70 s ramp waves for direct lightning to power line
Both lightings occur at the towers adjacent to substations.
Transmission line:
Eight (8)-phase frequency dependent distribution constant (Semlyen) model.
Tower:
Multistory transmission tower model shown in Figure 3.3.7. Constants, for example, Zt1=120 , were determined
by carrying out lightning impulse response characteristics tests on actual UHV towers and reflecting them [9].
Back flashovers between horns:
Leader model, time dependent inductance as shown in Figure 3.3.8, with leader strokes taken into account [10]
Substation circuits:
All buses are simulated by 3-phase separated GIS single-phase distribution constant circuits
Page 39
TypethenameofTB
Figure 3.3.7 Transmission tower model
Figure 3.3.8 Leader model
To analyze the estimated maximum overvoltage level, the following three types of severe operating conditions
involving high overvoltages generated at the substation equipment were determined, as shown in Figure 3.3.9,
Circuit condition I: High overvoltages are expected at the transmission line entrance.
Circuit condition II: High overvoltages are expected at buses.
Circuit condition III: High overvoltages are expected at the terminal of the transformer.
The validity of the insulation level determined from this maximum was proved by overvoltage distributions analyzed
as parameters in the unenergized equipment and lightning current during normal operation.
Figure 3.3.9 Service operating conditions of substation circuit for lightning surge
analysis
Page 40
TypethenameofTB
3.3.5.1.2 L IGHTNING
OVERVOLTAGE STUDY
On the basis of analyzed lightning overvoltage under severe circuit conditions, appropriate LIWV levels were
evaluated for different installation schemes of surge arresters. Table 3.3.3 shows that case 6 is economically most
favorable, which has two surge arresters per circuit at the line entrance, two per quarter bus, and one per bank of
transformers installed and employs the following LIWV levels.
- 1950 kV for transformers
- 2250 kV for GIS
Case 1 Case 2 Case 3 Case 4 Case 5 Case 6

No. of
Line entrance
1
surge
Bus
0
arresters
Transformer
1
TransRequired withstand
1950
former
voltage (kV)
GIS
LIWV(kV)
1950
Required withstand
2898
voltage (kV)
LIWV(kV)
Cost (100 for case 6)
2900
102
1
1
1
1
2
1
2
0
1
2
1
1
2
2
1
1943
1895
1943
1938
1896
1950
1950
1950
1950
1950
2854
2703
2628
2506
2208
2900
105
2900
109
2700
103
2550
103
2250
100
Table 3.3.3 Relationship between LIWV and Cost
Page 41
TypethenameofTB

GIS
Line entrance
2 units
MOSA layout
Busbar
2 units
(unit)
Transformer
1 unit
1896 kV for transformer
Representative
Lightning overvoltage
Value (kV)
2208 kV for GIS
Calculation method
EMTP analysis
Considered
Safety factor
(analyzed under severe circuit condition etc)
Transformer
1950 kV
LIWV (kV)
Other equipment
2250 kV
Case and condition for lightning overvoltage studies
Lightning current (kA) and its
200 kA with 1/70 s: ramp shape
waveshape
Lightning
No.1, 2 tower
stroke point
Back-flashover
- 1-line entrance, CB open (severest at line
entrance)
Substation circuit
- 1-line + 1-busbar (severest at busbar)
- 1-line + 1-busbar + 1-transformer
(severest at transformer)
Lightning current (kA) and its
30 kA with 1/70 s: ramp shape
waveshape
Lightning
No.1 tower
stroke point
Direct lightning
- 1-line entrance, CB open (severest at line
entrance)
Substation circuit
- 1-line + 1-busbar (severest at busbar)
- 1-line + 1-busbar + 1-transformer
(severest at transformer)
Table 3.3.4 Determination process for LIWV

3.3.5.2.1 S WITCHING
PROCESS FOR
SIWV
OVERVOLTAGE STUDY
In an analysis of ground fault overvoltages, the maximum overvoltage levels were obtained, using locations and
phases and phases of faults, voltage phases at the time of faults, ground fault resistance, etc., as parameters.
In an analysis of closing overvoltages, considering that the overvoltage level depends on the closing timing lags of
the main and resistance contacts of 3-phase circuit breakers, maximum values and overvoltage distributions were
obtained from a Monte Carlo simulation using closing timings of 6 contacts as parameters. System conditions
chosen are mainly the 1-circuit trial closing from the 2-circuit unenergized state of the transmission line, which is
expected to cause high overvoltages, or the reclosing of the transmission line.
Opening overvoltage analysis mainly focused on interruptions of ground fault currents, which are expected to
cause high overvoltages. The values of ground fault current were taken at the maximum level in each phase of
system growth, ultimately ranging up to 50 kA, which is the interrupting current of the circuit breakers. The
closing/opening resistors of 700 were employed to suppress (limit to a maximum value) switching overvoltage as
mentioned in section 3.4.3.
Page 42
TypethenameofTB
The switching overvoltages applied to the substation equipment, obtained under these conditions are shown in
Table 3.3.5. Here, the maximum overvoltage were generated under 1-circuit closing during 2-circuit unenergized
condition. For frequent overvoltages, the maximum value of ground fault and opening overvoltages at the end of
the substation was employed. Closing overvoltages in closing other than trial charges are lower.
Maximum overvoltage
1309 kV (1.46 p.u.)
1400 kV (1.56 p.u.)
Transformer
GIS
Frequent overvoltage
1250 kV (1.39 p.u.)
1250 kV (1.39 p.u.)
Table 3.3.5 Determination process for SIWV

3.3.5.2.2 D ETERMINATION
OF
SIWV
From the analysis results and equipment insulation characteristics, SIWV were determined as follows.
- 1425 kV for transformers
- 1550 kV for GIS
It is estimated that these withstand voltage levels are also economically acceptable, because insulation component
is basically determined by the withstand voltage values for lightning impulse or power frequency test voltages
described later.
(Max. and / or 2 %
Value)
(kV, p.u.)
Switching
overvoltage
Transformer
1309 kV (1.46 p.u.) (maximum)
Other
equipment
1400 kV (1.56 p.u.) (maximum)
Calculation method (2)
EMTP analysis
Safety factor
Considered (maximum overvoltage)

SIWV (kV)
Transformer
(phase-to-ground)
Other equipment
SIWV (kV)
Transformer
(phase-to-phase)
Other equipment
1.06, H=1000 m
1425 kV
1550 kV
10.5 - 11.5 m
Table 3.3.6 Determination process for SIWV

3.3.5.3 V ERY
FAST TRANSIENT VOLTAGE AND ITS SUPPRESSION
The GIS disconnector, when switching a charging current, repeats restriking and generates a very fast transient
overvoltage (VFTO). At a UHV substation, lightning overvoltages are effectively suppressed by higher performance
surge arresters. Disconnector overvoltages are likely to exceed the lightning overvoltage if no measures are taken
to control them. Therefore, the resistor insertion (500 ) is employed to suppress the VFTO.
Maximum
VFTO (p.u.)
Without
resistor
With resistor
()
Transformer
Other equipment
Transformer
Other equipment
2.8 p.u. (2514 kV)

1.3 p.u. (1167 kV) with 500
Table 3.3.7 Very fast transient voltage level
Page 43
TypethenameofTB
3.3.5.4 TOV
AND ENERGY ABSORPTION BY SURGE ARRESTER
[6]
Since a UHV system is looped with the existing EHV grid, load rejection of the UHV is exceedingly rare. However,
because it will generate high TOV with high energy, a scenario is taken into account, in which a UHV system with
several generators is isolated from the EHV grid due to double-circuit faults and successive malfunction of the
special protection scheme for tripping generators. A TOV analysis at load rejections, using the EMTP, as with
switching overvoltages analysis, simulated systems above 550 kV of all TEPCO systems, including UHV systems,
and also simulated in detail the AVR and governor characteristics of generators, which are expected to have a
large effect on voltage rises. The maximum value of TOV at load rejections in this case is about 1.5 p.u. and its
waveform is trapezoid as shown in Figure 3.3.10. The energy absorption by a surge arrester is so high that the
energy rating of an arrester is specified as 55 MJ and above.
The overvoltage protection relay system to open a no-load UHV transmission line concerned will be applied to
protect surge arresters if necessary, so that their energy absorption does not exceed the energy limit.
Overvoltage
The maximum TOV values for ground faults are about 1.15 p.u. on transmission lines and about 1.1 p.u. at
substations.
Figure 3.3.10 Example of analyzed temporary overvoltage

(p.u.)
1.15 p.u.
1.5 p.u. for 0.17 sec

(corresponding to the energy absorption of 55 MJ)
Load rejection followed by ground fault with selfexcitation of generators
55 MJ
Calculation method
EMTP analysis

3.3.5.5 P OWER
FREQUENCY TEST VOLTAGE
UHV power frequency test voltages were decided based on partial discharge tests (Long-duration tests) as shown
in Figure 3.3.11. This test is not only aimed to confirm dielectric strengths, but also to verify strengths by inspecting
whether or not partial discharges are generated as precursors of dielectric breakdowns throughout the test. The
test voltages are the combination of a short-duration section for verifying dielectric strengths to temporary
overvoltages at one-line ground faults and load rejections, and a long-duration section for assuring long-term
strengths to operating voltages.
Page 44
TypethenameofTB
Power frequency test voltages for UHV equipment were determined as follows.
[Transformer]
1.5 p.u. (1 h) +
[GIS]
1.5 p.u. (30 min) + 3 p.u. (1 min) + 1.5 p.u. (30 min)
3 p.u.(5 min) + 1.5 p.u. (1 h)
1.0 p.u. = 1100 kV /
V2
V1
t1
Longduration
section
V3
t2
t3
LongShortduration duration
section section
Figure 3.3.11 Pattern of 1100 kV power frequency test voltage
Transformer
Assumed overvoltage
condition for power
frequency test
Other
equipment
Transformer
Pattern
Other
equipment
Method of determination
1LG: 1.1 p.u. 0.1 seconds - 90times (period 30yr.)

Load rejection: 1.5 p.u. 0.83 second - 3times
(period 30yr.)
Normal voltage: 1.0 p.u. continuation time 30yr.
Load rejection: 1.5 p.u. 0.20 seconds* - 3times
(period 30yr.)
Normal voltage: 1.0 p.u. continuation time 30yr.
1.5 p.u. 1 h +
1.5 p.u. 30 min +
3 p.u. 1 min + 1.5 p.u. 1 h

3 p.u. 1 min + 1.5 p.u. 30 min
Combination of the short-duration section to confirm

dielectric strength against temporary overvoltage and
the long-duration section to confirm long-term dielectric
strength against operating voltage, based on the
systematically accumulated data
* Effect of trapezoid ACTOV waveforms was reflected in duration from insulation characteristics.
3.3.5.6 C OMPARISON
OF ANALYZED VALUE WITH MEASURED VALUE
TEPCO calculates various overvoltages by EMTP. The comparison between switching overvoltage analysis by
EMTP and actual measurement is shown in Figure 3.3.12. Computed values agree with measured ones in
TEPCOs UHV project very well. Therefore, the sophisticated design can be achieved by the precise overvoltage
analysis such as EMTP.
Page 45
TypethenameofTB
Calculation waveforms
Measurement waveforms
Figure 3.3.12 Comparison between switching overvoltage calculated by EMTP and

measurement in the case of closing circuit breaker with resistance
3.4 Brazil 800 kV project
3.4.1 INTRODUCTION
This presentation is basically divided in two parts: the first one shows the Brazilian knowledge and experience on
dealing with long distance EHV transmission systems (800 kV AC and 600 kV DC), describing the historical
reasons for that, faced problems/challenges and some aspects of the technical solutions to overcome those
problems/challenges, from planning and operating experience points of view. In the second part, it will be described
a summary of the Brazilian future trends and technical challenges, regarding the potential use/need of UHV
transmission system technologies, as for the transmission solution associated to the foreseen planned major hydro
powerplants in the Amazon region, that is, the transfer of high amounts of power (between 6000 and 10000 MW)
across long distances (in the range of 2000 km to 3000 km).
3.4.2 BRAZILIAN EXPERIENCE CONCERNING LONG DISTANCE TRANSMISSION SYSTEMS

To better understand the historical reasons for the countrys experience in dealing with long distance bulk
transmission technologies, it should be highlighted some particular features of the Brazilian Electric Power System:
the bulk electric power supply is typically provided by long EHV transmission lines belonging to the national
transmission grid, not only to connect far hydroelectric power plants of huge installed capacity to the main load
centers, but also to interconnect power systems of different regions, and from distinct geographical areas.
Regarding the geographic wideness of the country, several transmission lines and substations of 550 kV up to 800
kV had to be built in the country, soon after these voltage levels started to be used by utilities in other parts of the
world, due to the long distances between load centers and major generation powerplants.
Considering such a frame as above described, Itaipu hydroelectric power plant is located on Paran River, at the
border between Brazil and Paraguay and belongs to a company of same name: Itaipu Binacional. It is a power
generation utility, owned by both governments (of Brazil and Paraguay) in equal shares (50 % for each one). The
power plant has now-a-days 20 generating units of 700 MW each, resulting in a 14000 MW installed power
capacity enterprise, enable to produce a maximum of energy per year around 100 TWh. As in these countries the
nominal power frequencies are different from each other, the frequency of half of the generating units, belonging to
Brazil (10 units), is 60 Hz, while the other half belonging to Paraguay (another 10 units) has a frequency of 50 Hz.
Page 46
TypethenameofTB
This has been established as a consequence of an international treaty celebrated between both countries.
However, Brazil has agreed, by contract, to buy the whole of 50 Hz generated energy, except that consumed by
Paraguay, which means that 70 % to 80 % of the 50 Hz energy, in average, was consumed by the Brazilians in the
last years.
FURNAS Centrais Eltricas was responsible for developing the transmission solution in order to integrate Itaipu
powerplant generation into the Brazilian national transmission network, performing the planning studies, equipment
specification, factory and laboratory tests, erection on site, commissioning tests (factory and on-site acceptance
tests), operation and maintenance of the assets, amongst other tasks. Thus, FURNAS started the operation of the
800 kV AC and 600 kV DC systems in 1982 and 1984, respectively, integrating the Itaipu powerplant generation
into the Brazilian National Transmission Network, being, then, the detainer utility of the Brazilian know-how on
transmission issues of such voltage levels, completing now 28 years of successful operating experience.
The building up of the Brazilian knowledge and experience on dealing with 800 kV transmission issues have
started with these investigation studies, in order to define such hybrid transmission system of 900 km of length and
consisting of 3 800 kV AC parallel lines and 2 bi-pole 600 kV HVDC link. This hybrid AC/DC transmission
system planned, shown in Figure 3.4.2, is one of the most important in the occidental world, due to its nominal
voltage levels, rated power capacity and importance for the Brazilian electric industry. Its importance is not only
due to the usage of some of the highest voltage levels in commercial operation worldwide, but also because of the
high capacity of the transmitted power (rated to 12600 MW) for a long distance (about 900 km).
It is important to point out that there were no existing IEC Standards at that time for such a huge hybrid
transmission system, especially for the HVDC link. In this sense, some aspects and faced problems, from planning
and operating experience points of view, concerning the hybrid HVAC and HVDC features and equipment, had to
be deeply investigated as special requirements in the equipment specifications since the beginning of planning
works.
3.4.3 800 KV AC TRANSMISSION SYSTEM REMARKABLE ASPECTS

The HVAC transmission system interconnects Foz do Iguau substation (close to Itaipu powerplant) to Tijuco Preto
substation in So Paulo area (close to So Paulo metropolitan region), one of the main load centers of the Brazilian
electrical network.
Foz do Iguau substation contains both the UHV 800 kVAC switchyard and the 600 kVDC converter substation
(rectifier side). Together they form one of the biggest substations in operation, in terms of power and size.
HVDC +/-600kV
switchyard
800kVac
switchyard
Figure 3.4.1 Foz do Iguau substation, belonging to the Itaipu transmission system,
contains both the HVAC switchyard of the three 800 kV circuits and the HVDC
converter substation equipment/switchyard (HVDC link rectifier side)
Page 47
TypethenameofTB
The three circuits of 800 kV AC, each one of about 895 km long, are able to transmit to the Brazilian interconnected
network the rated generated power from the generating units in Itaipu powerplant operating with the nominal
frequency of 60 Hz. The planning criterion of such transmission system considered the possibility of dispatching
such amount of power even in the n-1 elements outage configuration. Also, this 800 kV AC system is responsible
for the energy interchange between the southern and southeastern geographic regions of the brazilian
interconnected national grid as well (refer to Figure 3.4.2).
Figure 3.4.2 More details about FURNAS transmission system for the integration of
Itaipu power plant within the Brazilian electrical network
The electrical requirements for the 800 kV AC equipment has been defined in the late 70s and early 80s. At that
time, the Brazilian network consisted of several sub-networks weakly connected to each other. Although each subnetwork could be considered as solidly grounded, the network as a whole could not. Besides, the 800 kV Brazilian
network would not be a meshed network but a radial one consisting of 3 parallel circuits. Thus, the planning
engineers at that time decided to adopt conservative insulation coordination requirements for the equipment of
such a new and unknown high voltage level in Brazil, derived from system simulations under severe operation and
emergency conditions.
The shunt reactors (fixed at line-ends and switchable at busbars and transformer tertiaries), series capacitors and
synchronous condensers provide, respectively, transient switching overvoltages mitigation (fixed SR at line-ends),
voltage regulation during steady state operating conditions (switchable SR at busbars and transformer tertiaries)
and system dynamic stability (series capacitors and synchronous condensers), therefore, contributing to the
insulation coordination matter from the switching transient overvoltage issue point of view.
Regarding the reliability of the overall transmission system, in terms of potential damages due to strong winds and
heavy thunderstorms, in order to keep the criterion of n-1 element outage as much as possible in real life, the rightof-way of the set of 5 transmission lines (3 AC lines and 2 HVDC bipoles) were defined with a security distance
among them as shown in the Figure 3.4.3 below.
Page 48
TypethenameofTB
Figure 3.4.3 Geographic arrangement of the set of 5 lines that belong to the
transmission system for the integration of Itaipu power plant within the Brazilian
electrical network (the distances are in kilometres)
3.4.4 800 KV SHUNT REACTOR APPLICATION PHILOSOPHY
In FURNAS, as well as in other utilities, the SR are used in power system transmission OH lines for 2 main
purposes: (a) switchable SR at busbars or tertiary-windings of power transformers for steady-state voltage control
issues; (b) fixed SR at the OH line ends to deal with switching transient overvoltage requirements, related to the
line closing and/or opening operation (including fast auto-reclosing issues). Therefore, the SR installed at this 800
kV transmission system follows the same philosophy, giving 70 % of shunt compensation for line charging. In the
next schematic diagram one can see the existing SR application philosophy installed in such transmission system.
For transient switching overvoltage mitigation purposes, the fixed SR have the aid of the arresters installed at the
line ends, having such arresters, along the last 30 years, two very distinct functions on such duty, depending on the
evolution of technology and applied materials by the arresters manufacturers: (a) arresters with active gaps and
non-linear resistances made of silicon carbonate (SiC); and (b) gapless metal oxide arresters (MOSA).
When it was used only arresters with active gaps and SiC non-linear resistances, the very first arrester that sparked
its active gaps was the only responsible for draining all the surge energy, as well as for controlling the switching
overvoltage in conjunction with the fixed SR at the line end.
After the introduction of devices such as gapless metal oxide arresters, the surge energy draining could be divided
amongst all the MOSA, depending on the VxI curve of the used non-linear material (Zinc Oxide - ZnO).
Page 49
TypethenameofTB
Equivalent Source 1
Equivalent Source 2
Zeq 2
Zeq 1
Line-length = 300 km
Figure 3.4.4 Fixed Shunt Reactor application philosophy for transient switching
overvoltages mitigation in the OH line
Regarding the gapless MOSA generation, those installed to protect the SR against lightning impulses could also be
used in conjugation with the MOSA installed at the OH line ends. Thus, using such set of surge arresters (i.e., not
only the ones installed at the OH line ends, but also those provided to protect the fixed SR against lightning surges),
the control of the switching transient overvoltages and the draining of switching surge energy as well, could be
done in a much more efficient/reliable way, helping with the mitigation/control of switching transient overvoltages as
mentioned, in comparison with the old SiC SA used before.
Figure 3.4.5 Shunt Reactor configuration in the 800 kV transmission system

3.4.5 SWITCHABLE SR: MITIGATION/CONTROL OF SWITCHING TRANSIENT
OVERVOLTAGES
Referring to the switchable SR, for mitigating/controlling the switching transient overvoltages generated when
switching off such SR, today, it is possible to use CB equipped with opening resistors or controlled switching. In
case of using CB equipped with opening resistors, for optimizing the transient switching overvoltage control, the
ohmic value in each phase should be of the same magnitude of the characteristic impedance of the SR (surge
impedance). In the 800 kV level, a normal range of surge impedance values for SR is 2000 to 4.000 .
The 800 kV switchable SR of this FURNAS transmission system, in the 1980, used to have CB equipped with
opening resistors for transient switching overvoltages mitigation, as it has been defined by the planning and design
works, developed in the late 70 and early 80. The design criterion for dimensioning the CB opening resistors was
to use the same ohmic-value-range of the SR surge impedance, meaning it was necessary to provide CB with
opening resistors in the range of 3.000 . In this case, there was no control over the arcing time of the SR current
in the CB chambers during the switching off procedures. Therefore, this could lead to reignitions within the CB
chambers when the arcing time was too small, causing new overvoltages with very high front steepness. During the
first decade of the switchable SR operation, this phenomenon caused some major damages to the SR and related
CB, even leading to the explosion of the switchable SR CB in Itaber substation. After this accident the company
decided to replace the CB equipped with opening resistors by controlled switching. The controlled switching
Page 50
TypethenameofTB
technology showed to have a very good performance for transient switching overvoltages mitigation, avoiding
completely the possibilities of resignations during the switching off procedures, since the convenient minimum
arcing time could be properly controlled for that purpose. The extensive field tests/measurements performed by the
company proved such expected performance during several switching off operations of the existing 800 kV busbar
SR. On the other hand, the larger arcing times, provided by the CB equipped with controlled switching device,
would lead to a higher and more frequent switching phenomenon defined as current chopping overvoltages.
The related current chopping switching transient can lead to a transient recovery voltage (TRV) across the circuitbreaker chambers that should be investigated, in principle, as a big concern, in order to avoid undesirable
damages to the SR, or even to the CB. This transient phenomenon is related to the fact that, during the SR
switching off, the nominal current is forced to zero suddenly and prior to the natural power-frequency-zero-crossing,
within the CB chambers, leading to such a phenomenon known as current chopping. Thus, the current chopping,
meaning a sudden change to zero in the magnitude of the SR nominal current, forces the magnetic field energy,
stored in the reactor coil, to be converted to electric field energy, stored within the stray capacitances, causing,
thus, sudden changes in the load-side transient voltage. This energy-transfer-phenomenon is known as chopping
overvoltage. The lower the stray capacitances are the higher the chopping overvoltages will be during the
occurrence of current chopping.
However, the field experience in FURNAS showed it was much more reliable and secure to have a complete
control over the possibility of no-reignition during the SR switching off, by means of using CB equipped with
controlled switching, compared to the increase in current chopping overvoltages. In the case of 800 kV switchable
SR, the increase of chopping overvoltages, derived from the increase of minimum arcing time (given by the CB with
controlled switching), is very small, due to the fact that the stray capacitances of the SR windings are quite
significant.
3.4.6 800 KV SERIES CAPACITOR (SC) APPLICATION PHILOSOPHY

The Series Capacitors provide system dynamic stability when dispatching huge amounts of power through the 800
kV transmission system, from Itaipu powerplant and/or high interchanges between the two interconnected regional
grids. The existing SC were designed to offer a compensation degree of 40 % to 50 % related to the OH lines
longitudinal inductive reactances, depending on the localization at the 800 kV transmission system (refer to Figure
3.4.6), being these equipment amongst the greatest in the world, in terms of rated power and voltage level. In this
dispatching condition, the voltage profile along the transmission system must be as much closer as possible of the
maximum continuous operating voltage level (at nominal frequency), or 800 kVrms between phases.
The SC were conceived to operate with two series segments, corresponding each segment to 20 % or 25 % of the
longitudinal reactance of the transmission lines. Each SC segment can operate together or separated from the
other segment, increasing, thus, the reliability of the SC themselves and of the 800 kV transmission system as a
whole.
Page 51
TypethenameofTB
Figure 3.4.6 Schematical one-line diagram of the 800 kV AC transmission corridor

- part (a) - and main power equipment data - part (b). The simplified CB diagram
showed in part (a) represents schematically the breaker-and-a-half busbar
configuration used in all 800 kV substations
The maximum operative voltage level (800 kVrms between phases) of this transmission system is measured right
after the SC (at the line side) in Ivaipor and Itaber substations, since the voltage profile suffers a sudden
increase in these points (proportional to the current flow) caused by the presence of the capacitive reactance in
series with the line inductive longitudinal reactance (refer to Figure 3.4.7)
During the planning studies the SC should be designed having their capacitive reactance, to compensate part of
the line inductance, in such a way that some main technical and economical concerns/requirements, regarding the
SC equipment itself, could be achieved, as follows: (a) SC self-overvoltage-protection devices (in the presence of
short-circuits); (b) protection coordination of parallel lines with series-capacitor compensation; (c) sub-synchronous
resonance.
In the presence of SC connected to the transmission lines, as they compensate the line inductive longitudinal
reactance, the short-circuit currents tend to increase (due to the lower values of short-circuit equivalent impedance
of the system amongst the source equivalents and the fault places).
Also, the transient overvoltages through the SC, caused by the short-circuit flowing, tend to be greater as well. In
order to control such short-circuit currents, as well as the transient overvoltages during the flowing of such shortcircuit currents in the compensated lines, the SC have a self-overvoltage-protection device, as shown with
schematic details in Figure 3.4.8.
Page 52
TypethenameofTB
Equivalent Source 1
Equivalent Source 2
Line flow
Zeq 2
Zeq 1
800 kV rms
Line-length
Figure 3.4.7 Voltage profile caused by Series-Capacitor placed in a OH line

The higher is the short-circuit current flowing through the SC, the greater will be the transient voltage raise at the
SC terminals. The definition of the dimensioning/design of the SC protection level is driven mostly by economic
reasons, regarding insulation withstand requirement costs. Thus, the non-linear varistor, the gap-protection-circuit
and the by-pass breaker operate whenever necessary, by-passing very quickly the SC, in order to keep the
transient overvoltage values within the designed withstand levels of the SC components defined by the insulation
coordination assessment.
Page 53
TypethenameofTB
SEGMENT 1
SEGMENT 2
Figure 3.4.8 Details of the 800 kV AC series-capacitor at Itaber substation

The varistors main role is to operate as an electronic switch, by-passing the SC during the transient flow of shortcircuit currents through it, at the same time the transient voltage across the SC increases. The purpose of the
parallel gap and the SC-segment-by-pass breaker is to protect the varistor (from overheating) when it drains higher
amounts of energy during more severe transient overvoltages in the presence of larger short-circuit currents.
Besides this presented self-overvoltage protection scheme, the SC main conventional protection features are:
Gap operated
Gap keep operating
Page 54
TypethenameofTB
Platform failure
Current Unbalance (only Alarm)
Current Unbalance (activating SC by-pass)
Optics fiber failure
By-pass breaker failure
SC bank overload
Air-Compressed Central failure (the gap operates with air-compressed-blast, provided by a dedicated central to
the SC)
Regarding the protection coordination of parallel lines with series-capacitor compensation, this issue is also of big
concern and has to do with the distribution of short-circuit currents amongst the parallel circuits, which may cause
undesirable protection operation of such parallel (and not faulted) lines. This may happen due to overvoltage
protection or SC by-pass misoperation.
There is still the issue, also of big concern, related to the sub-synchronous resonance, that may happen when
there is thermal turbines sited electrically near the series-compensated lines. Depending on the compensation
degree provided by the SC, they, in conjunction with the line inductances, may cause a natural oscillation mode
having typical values of resonance frequency below the rated power frequency of the system. Normally the thermal
turbines also have mechanical natural oscillation modes in that range of frequencies, leading, then, to the
possibility of under or sub-synchronous resonance occurrence.
3.4.7 OTHER 800 KV TRANSMISSION SYSTEM REMARKABLE ASPECTS

Regarding the 800 kV line-phases transposition designing criterion used, refer to the following schematic diagram
(Figure 3.4.9):
A
C
1/6
1/3
6/6
5/6
LENGTH
Figure 3.4.9 800 kV line phases transposition criterion

Regarding the fast auto-reclosing of the Itaipu 800 kV transmission system, the main related data is: there is only
3-phase fast auto-reclosing in operation (TPAR). The dead time is 600 milliseconds for all line segments (for all 3
circuits Foz do Iguau-Ivaipor; Ivaipor-Itaber and Itaber-Tijuco Preto). There is a big concern here related to
the use of Single Phase Auto-Reclosing (SPAR) in the 800 kV system, that is the potential risk of occurring severe
commutation failure in the inverter station thyristor valves of the HVDC link of Itaipu transmission system if SPAR
scheme were used, as can be seen in more details in item 3.4.8.
Page 55
TypethenameofTB
Auto-reclosing leader terminal
Auto-reclosing leader terminal
Figure 3.4.10 800 kV line 3-phase auto reclosing criterion

The leader terminal for the 3-phase auto-reclosing is Ivaipor, for all lines between Foz do Iguau-Ivaipor
substations, and Itaber, for all lines between Ivaipor-Itaber and Itaber-Tijuco Preto substations. This leader
terminal definition is related to the minimum power-accelerating impact in hydro generators in Itaipu powerplant
and other hydroelectric plants sited in the southern regional grid (refer to Figure 3.4.10).
Also, the 800 kV transmission system requirements should consider, for insulation coordination of equipment
designing purposes, the worst conditions regarding switching transients. Thus, it was decided at the beginning of
designing works to take into account, for instance, the criteria of switching transients derived from total 'oppositionof-phase' OP (i.e., 180 degrees instead of 120 degrees, foreseen in the international standards at that time). This
was reasonable to consider at that time for this transmission system, since the 800 kV AC system would be, in
practical terms, not only the connection system of the Itaipu powerplant into the Brazilian interconnected grid, but
also the main interconnection-tie between the Brazilian southern and southeastern regional transmission grids, the
two major sub-system of the country at that time. Before the existence of the Itaipu transmission system, these two
regional transmission grids were weakly interconnected, since they had only few tie-lines using voltage levels of 88
kV, 138 kV and 230 kV in the border between the state of So Paulo (south border of southeastern regional grid)
and the state of Paran (north border of southern regional grid).
Considering the same reason for the choice of OP premise, as above said, it was adopted, then, the value 2.0 for
out-of-phase factor for all TRV parameters. For the use of an amplitude-factor of 1.64 for Uc of TRV, instead of
1.25 defined by IEC at that time, the reason for that is the adopted 'grounding criteria', since, although each
regional grid could be considered as solidly grounded, the 800 kV system as a whole could not, as it was quite
radial and not a meshed network like the two regional grids interconnected by means of this AC transmission link.
Following this idea of adopting conservative criteria for defining the equipment withstand requirements (for such a
new and unknown high voltage level in Brazil at that time), derived from system simulations under severe operation
and emergency conditions, it was also found necessary to define other requirements based on rigid premises, such
as the maximum overspeed during total load rejection in the 800 kV AC system as 1.1 of nominal frequency (66
Hz), instead of adopting the standardized value of 1.05fn (63 Hz). This rigid overspeed criteria was used to verify
the possibilities of self-excitation regarding the Itaipu generators, and also to define the withstand requirements of
switching equipment due to overvoltage transients caused by full load rejection.
Finally, as other example, the value of 800 kV was defined as the 'maximum operating voltage' for the Itaipu
transmission system, which means only 1.045 p.u. of the nominal operating voltage (765 kV). In all other Brazilian
voltage levels the maximum operating voltage (non-continuous operating voltage limit, for emergency operating
Page 56
TypethenameofTB
conditions) is 1.10 p.u. of the rated or nominal voltage value (138 kV/152 kV; 230 kV/253 kV; 345 kV/380 kV; 440
kV/484 kV; 500 kV/550 kV).
3.4.8 600 KV DC TRANSMISSION SYSTEM REMARKABLE ASPECTS

The HVDC transmission system consists of two bi-poles of 600 kV AC, each one of about 900 km long, one
AC/DC rectifier side at Foz do Iguau substation and one DC/AC inverter side at Ibina substation, sited very close
to So Paulo city, Figure 3.4.2.
The HVDC transmission system is rated to transmit 6.300 MW from the generated power produced by the other 10
generating units of 50 Hz owned by the Paraguayan side of the bi-national Itaipu powerplant. In order to build up
the Brazilian know-how on HVDC transmission, being used for the first time in the country, the following constrains
have been established: services, equipment and products should be nationalized; Brazilian enterprises/engineers
should take part on the planning studies, designing, building and operation of the HVDC transmission system;
technology transfer on the job training should be practiced.
Figure 3.4.11 FURNAS 600 kV HVDC bipole tower structures (near Itaber
substation)
The 600 kVDC transmission system main characteristics are as follows:
Two substations (Foz do Iguau the rectifier side and Ibina, the Inverter side);
Nominal Power: 6300 MW;
Nominal Voltage: 600 kV(DC);
Nominal Current: 2625 A(DC);
Two Bipoles ( 600 kV);
Each Bipole composed by two poles;
Each Pole composed by two 300 kV series-connected convertors.
Both AC/DC rectifying and DC/AC inverting processes require great amounts of reactive power and produce
harmonic currents that cannot be injected into the related AC network. At Foz do Iguau substation AC filters
totalizing 1540 Mvar have been installed. Because of its vicinities to Itaipu power plant, the needed reactive power
is mostly supplied by the power plant itself, performing the filters the role of not allowing the harmonic currents flow
into the 50 Hz AC network. At Ibina substation, besides AC filters (totalizing 2480 Mvar) have been installed, also
Page 57
TypethenameofTB
shunt capacitors, totalizing 590 Mvar, and 4 synchronous compensators of 300 Mvar each were defined to supply
the needed reactive power and proper short-circuit power level (for proper inverter thyristor valves operation).
One big concern is related to the application of Single-Phase Auto-Reclosing schemes (SPAR), in the AC system
close to inverter stations of bulk HVDC transmission systems, since this auto-reclosing mode (single phase
operation) may cause several commutation failures (CF) in sequence in the inverter station thyristor valves after the
OH line short circuit application and the opening of only its faulted phase.
This CF process is caused by the big asymmetry among the 3 phase-to-phase AC voltages applied to the thyristor
valves of the inverter station during the dead time of the auto-reclosing scheme (normally from 0.8 to 1.5 seconds).
The operation of the OH line with only two sound phases and one open phase during the dead time of the autoreclosing scheme, at the same time the transient recovery process of the inverter station occurs, establishes an
uneven condition for the HVDC system operation, i.e., the three phase voltages at the AC side in the inverter
station are strongly unbalanced, causing the asymmetry of voltages applied to the thyristor valves, leading to
potential commutation failures in series. This possible behaviour is of big concern since its consequences could be
catastrophic to the Power System stability and reliability (loss of stability and/or voltage collapse could occur).
3.4.9 FUTURE TRENDS/CHALLENGES REGARDING POTENTIAL UHV TRANSMISSION IN

BRAZIL
Stepping into the future and trying to foresee possible needs, trends and challenges, regarding the use of UHV
transmission in Brazil, we have on one hand, 180 GW corresponding to the remaining Brazilian hydroelectric
potential still not used. A high percentage of it (around 70 %) is located in the Amazon region, according to the
official estimates. On the other hand, the consumption of energy is spread all over the country according to the
following geographical distribution:
Isolated Systems in Amazon Region (2 % of consumption)
Interconnected System in North/Northeast (19 % of consumption)
Interconnected System in South/Southeast/Central-West (79 % of consumption)
In terms of the Brazilian power system growth, it is reasonable to say that it will be necessary to transport huge
amounts of energy produced mainly in the Amazon region to the main load centers of the country, covering
distances of more than 2500 kilometers, as shown in Figure 3.4.12. Considering such a frame, the use of UHV
transmission systems (DC links and/or AC overhead lines) seems to be quite competitive and suitable alternatives,
in both economic and environmental views/aspects. In this case, it is not necessary to change the frequency from
the rectifier AC side to the inverter AC side, since it is just a matter of a set of issues such as: stability, loss
reduction, compactness, environmental friendliness and costs from the far amazonic powerplants to the main load
centers.
Figure 3.4.12 Geographical distribution of load centers in Brazil (left) and average
distance between the main load centers and future Madeira river complex (right).
Bulb-type turbines
Page 58
TypethenameofTB
For new long distance transmission systems, some important features and concerns related to the planning/design
stage, as well as to equipment withstand requirements, should also be deeply investigated, such as:
a- comparison between equivalent solutions of HVAC and HVDC systems and between equivalent GIS and AIS
equipment, in terms of engineering and economics issues (considering environmental and regulatory issues);
b- characteristics of HVDC system facilities, such as the so-called Dynamic Performance feature (concerning
impacts caused to the AC systems connected to the converter substations, both rectifier and inverter HVDC
system sides, in terms of mitigation/elimination of dynamic electromechanical transients/oscillations); HVDC
transmission system Forced Isolation protection scheme (implemented to give protection against electrical
transients derived from partial or total load shedding); High MVAr consumption operating mode; automatic fast
switching from one faulted DC line to another one under unrecoverable short-circuit conditions in the first line;
c- HVDC reliability features specification (ability of the system to transmit the rated power under contingency
conditions and outages), such as: operation modes (bipolar/monopolar) and related engineering/environmental
issues; temporary overloading/overcurrent capacity; AC System Faults/configurations that may create commutation
failures mainly in the inverter station;
d- importance/advantages of analogue/digital simulation facilities for HVDC/HVAC transmission systems
(necessity/usefulness of simulation tools for HVDC transmission systems in terms of
planning/specification/operational optimization and economics);
e- DC switchyard bushing isolation problems and the implemented mitigation solutions;
f- issues related to filtering requirements (short circuit level criteria, filter overloading, active filters usage since
they are not a sink for harmonics other than their own, which can be decisive in filtering performance, etc);
g- HVAC shunt and series compensation devices (and the related inherent aspects in terms of voltage profile
control, overvoltage transients and protection issues);
h- importance of equipment withstanding requirements definition in the planning phase, taking into account the
possible/foreseen evolution/expansion of the AC grid/network (up-grading in short-circuit level requirements);
i- enterprise staging definition, i.e., intermediate stages of transmitting power capacity, considering cost constraints,
needs of scaling the power transmission, time interval between stages, etc (series connection of converter groups
per pole easily allows the staging of 25 %, 50 %, 75 %, and 100 % of the total rated power);
j- oil chemistry issues of HVDC converter transformers;
k- possibilities of controlled switching usage.
Also, the following topics must be analyzed in detail in order to establish the reference UHV transmission
alternatives (DC, AC or hybrid transmission systems):
Line configuration /insulation/ clearances;
Corona and field effects performance of the line;
Level of power to be transmitted;
Weight and size of the equipment for shipping and transport (mainly the transformers for AC and DC
alternatives);
Power losses;
Spare parts;
Series connection of converter groups per pole (DC alternative);
Page 59
TypethenameofTB
Overload/ Stability requirements;

Need of reinforcing the receiving network;
Need of investigating procedures for limitation of short-circuit current levels exceeding existing equipment ratings.
Despite of the described Brazilian experience on hybrid parallel AC 800 kV and DC 600 kV systems, due to the
big challenges related to the new foreseen Amazon transmission system requirements, and depending on the
outcoming UHV transmission alternative references, the existing IEC Standards may not cover all the equipment
withstand requirements. This will lead to the necessity of developing new specification standards that might be of
planning and design concerns.
In order to face the challenges foreseen, several Brazilian scientific organizations, universities and utilities are
preparing themselves by means of studying new technologies and developing know-how.
3.5 China 750 kV AC Project

3.5.1 CHINA 750 KV AC TRANSMISSION SYSTEM
The 750 kV voltage level power grid has been built based on the 330 kV power grid in China Northwest area,
whose transmission system maximum operating voltage is 800 kV.
China first 750 kV transmission line was formally put into commercial operation in September 2005; thereafter,
more 750 kV transmission lines were gradually increased; for the time being, the total length of 750 kV
transmission lines are up to 3525 km.
3.5.2 POWER FREQUENCY TEMPORARY OVERVOLTAGE (TOV)

The following fault types shall mainly be taken into account for predicting the power frequency temporary
overvoltage.
As for the single-circuit transmission lines, two kinds of failures shall be usually taken into account, namely, the
load rejection under the normal state and the load rejection in the case of the line single-phase grounding failure.
As for the double-circuit tower arrangement transmission line, the double-circuit operation and one circuit out of
service/the other circuit operation shall be considered and the failures causing double-circuit 6-phase load rejection
shall be taken into account for the former case.
The main measure to limit the power frequency overvoltage is to install the line high voltage shunt reactor.
Most of China 750 kV lines are relatively long and the high voltage shunt reactors are generally installed in the lines,
whose largest TOV may generally occur in the single-phase grounding fault case; in which the circuit breaker at
line end shall be three-phase tripped and the circuit breaker at the line send side shall not be tripped.
The maximum TOV shall be no more than 1.4 p.u. at the line side and the maximum TOV shall be no more than
1.3 p.u. at the bus side.
The power frequency temporary overvoltage duration time shall be less than 0.5 seconds. The 750 kV voltage level
power grid has been built based on the 330 kV power grid in China Northwest area, whose transmission system
maximum operating voltage is 800 kV.
China first 750 kV transmission line was formally put into commercial operation in September 2005; thereafter,
more 750 kV transmission lines were gradually increased; for the time being, the total length of 750 kV
transmission lines are up to 3525 km.
Page 60
TypethenameofTB
3.5.3 MOA PARAMETER SELECTION

According to the different TOV values at the line sides and the bus sides, the different rated voltage Ur shall be
selected for the metal oxide arresters (MOA) at the corresponding positions: 600 kV for the bus side and 648 kV for
the line side.
The MOA main electrical parameters are listed in Table 3.5.1; and the MOA nominal discharge current is 20 kA.
System
nominal
voltage
750
750
Installation location
The bus side and
the transformer side
Line side
(Unit: kV)
The lightning
impulse residual
voltage under 8/20
s and 20 kA
Rated
voltage
(RMS)
Continuous
operating
voltage (RMS)
The switching
impulse residual
voltage under 30/60
s and 2 kA
600
462
1142
1380
648
498
1234
1491
Table 3.5.1 The main electrical parameters of 750 kV substation arrester

The following two categories of overvoltage shall mainly be taken into account:
The closing and reclosing no-load line overvoltage
The single-phase grounding fault and one side three-phase tripping overvoltage.
The main measures to limit the switching overvoltage is that the closing resistors being erected in the line circuit
breakers and the high voltage shunt reactor being erected in the line and MOA being erected in the both line ends.
The circuit breaker closing resistor is 575 . The switching overvoltage which may play the control role to the
insulation design of China 750 kV lines shall be the closing and reclosing line overvoltages; whereas, the singlephase grounding fault and one side three-phase tripping overvoltage shall mainly be used to determine the safety
distance from live working.
The maximum phase-to-ground statistical switching overvoltage along the line shall be no more than 1.8 p.u..
The circuit breaker used for the switching transformer shall not be equipped with the closing resistor.
3.5.5 LINE LIGHTNING PROTECTION

The lightning activity is relatively weak in China Northwest area; the calculated back flash-over trip rate is
4.16410-5 times/100kma, which is very low and much lower than the statistical value of the lightning trip rate for
China 500 kV lines. (0.14 times/100kma).
The lightning shielding failure trip-out rate is much larger than the back flash-over trip-out rate for 750 kV lines in
China Northwest area. The main measure to limit the lightning shielding failure trip-out rate shall be to lower the
ground wire protection angle. The ground wire protection angle shall be no more than 10 and 0 for 750 kV singlecircuit lines and the double-circuit tower arrangement transmission line, respectively.
The expected lightning trip rate of 750 kV lines shall be lower than the operating statistical value of the lightning trip
rate of 500 kV lines (0.14 times/100kma). Since the 750 kV line was put into operation, the actual lightning trips
have been 2 times. The following two categories of overvoltage shall mainly be taken into account:
Page 61
TypethenameofTB
3.5.6 SUBSTATION LIGHTNING INTRUDING OVERVOLTAGE

While the 750 kV substation lightning intruding overvoltage is calculated, the 1-line and 1-transformer connecting
mode with relatively high overvoltage has been taken into account; on the other hand, it has not been considered
that the single-line mode (the entrance line circuit breaker is in the trip state).
The arrester arrangement mode may be listed as follows: one group of arresters shall be erected in the bus and the
main transformer, respectively; one group of arresters shall be erected in each line entrance. However, the above
arrester arrangement mode is not used by all 750 kV substations, the different substations may have different
arrester arrangement.
According to the characteristics of the lightning activity being weak in China northwest area and the high lightning
insulation level of 750 kV lines, the probability for the lightning back flashover intruding overvoltage shall be very
small and the substation lightning intruding overvoltage shall be mainly derived from the lightning shielding failure;
decreasing the ground wire protection angle may bring about decreasing the maximum lightning shielding failure
current so as that it may be very beneficial to lower the shielding failure lightning intruding overvoltage.
While the 750 kV substation lightning intruding overvoltage is calculated, the lightning points shall be the various
towers in the 2 km entrance line and the conductors in the corresponding locations; and the lightning back
flashover current shall be taken as 225 kA. The maximum shielding failure lightning current shall be calculated and
determined by using the electrical geometric model and according to the specific tower sizes; and the calculated
lightning current shall be the double ramp shape waveform (2.6/50 s).
The various substation lightning intruding overvoltages are different, which shall be relevant to the tower sizes in
the entrance line, the substation connection modes and the electrical distance as well as the arrester layout and
other factors. As shown in the calculation results for several 750 kV substations in China Northwest area, the
representative maximum lightning intruding overvoltage may be 1591 kV, 1816 kV and 1695 kV for the
transformers, reactors and GIS, respectively.
As for the insulation coordination, the insulation margin of the equipment internal insulation shall be more than
15 % and the insulation margin of the equipment external insulation shall be more than 5 %. According to this
principle, the lightning impulse withstand voltage shall be 1950 kV and 2100 kV for the transformer and other
equipment, respectively. The lightning activity is relatively weak in China Northwest area; the calculated back flashover trip rate is 4.16410-5 times/100kma, which is very low and much lower than the statistical value of the
lightning trip rate for China 500 kV lines. (0.14 times/100kma).
3.5.7 INSULATION COORDINATION
3.5.7.1 L INE
INSULATION COORDINATION
The required values for the line tower minimum air gap are shown in Table 3.5.2.
The switching impulse air gap distance shall be selected according to the maximum switching statistical
overvoltage along the line, being 1.8 p.u..
The lightning impulse minimum air gap distance shall be selected according to 0.8 times of the lightning discharge
voltage of the insulator string.
The minimum air gap distance required for the safety distance of live working controls the line tower sizes, in the
Table 3.5.2 0.5m shall be considered for the human body action scope.
Page 62
TypethenameofTB
(Unit: m)
Power frequency
voltage
Side phase (I type insulator string): 1.9
Switching impulse
Side phase (I type insulator string): 4.0

and middle phase (V type insulator string): 4.8
Lightning impulse
4.2 (or the coordination shall be implemented according to the

0.8 times of the insulator string lightning discharge voltage.)
Safety distance
requirement for live
line working
Side phase (I type insulator string): 4.0+0.5

and middle phase (V type insulator string): 4.3+0.5
Table 3.5.2 Line minimum air gap (altitude: 1000m)

3.5.7.2 I NSULATION
COORDINATION OF THE SUBSTATION EQUIPMENT
The required values of the equipment insulation level are shown in Table 3.5.3.
Equipment
Rated lightning impulse withstand
voltage
Rated switching impulse withstand
voltage
Power frequency withstand voltage
for 1minute
(Unit: kV)
Circuit breaker
longitudinal insulation
Transformer
Reactor and other

equipment
1950
2100
2100+(650)
1550
1550
1300+(650)
900
Reactor 900
Other equipment 960
1270
Table 3.5.3 The rated insulation withstand voltage for 1000 kV equipment
No matter what the stress is the power frequency voltage, the switching impulse voltage or the lightning impulse
voltage, the insulation margin of the equipment internal insulation shall be more than 15 % and the insulation
margin of the equipment external insulation shall be more than 5 % for the insulation coordination.
The power frequency 1min withstand voltage selection shall be taking both TOV and the required power frequency
withstand voltage conversed from the lightning impulse withstand voltage into account.
The insulation coordination for the switching impulse withstand voltage shall be carried out based on the substation
switching overvoltage level (1.8 p.u.) and the arrester switching impulse protection level.
The insulation coordination for the lightning impulse withstand voltage shall be carried out based on the calculation
value of the substation lightning intruding overvoltage.
3.5.7.3 S UBSTATION
MINIMUM AIR GAP CLEARANCE
The A values of the substation minimum air gap distance shall be divided into three categories: A1' value, A1" value
and A2 values.
The A1' value is the minimum electrical distance of the substation wire to the frame; the A1" value is the minimum
electrical distance of the substation equipment to the frame; and A2 value is the minimum electrical distance
between the phases in the substation.
The recommended values of the minimum electrical distance for the substation are listed in Table 3.5.4.
Page 63
TypethenameofTB
(Unit: kV)
Phase to phase
A2
3.75
6.5
4.8
Phase to earth
Action voltage type
A1
2.3
5.1
4.3
Power frequency
Switching impulse
Lightning impulse
A1
2.3
5.5
4.53
Table 3.5.4 The recommended minimum electrical distance value for the substation
(altitude: 1000 m)
3.6 India 800 kV project
3.6.1 FEATURES OF POWERGRIDS 800 KV TRANSMISSION SYSTEM
The 800kV level in the EHVAC transmission system of Indian Power System has been planned as a next higher
level to 400kV with an aim to transfer bulk quantum of power from power surplus region/area to power deficit
region/area. A number of 800kV transmission lines are being planned for transfer of power from Eastern Region,
already a power surplus region, and Southern Region, a future power surplus region, to power deficit regions like
Western and Northern Region. In this process large 800kV transmission network is being established not only for
transfer of power across the various regions but also for dispersal of the same within the regions specially from
Eastern or generation dominated part to Western i.e. demand dominated part of Northern and Western Region.
Initially, some of the 800kV lines were planned to operate at 400kV level, however with requirement of enhanced
power flow the same are to be operated at its rated voltage. The first line came into operation at 800kV voltage
level is Sipat-Seoni S/c line in Nov. 2007.
The study results shown here is for the Sipat-Seoni 800kV line (350 km), the Sipat end is a generating station while
the Seoni end is connected to the grid. The installed capacity of generating station at Sipat is 2980 MW (2 units of
500MW and 3 units of 660 MW) and the same is evacuated through 2 nos. of 800kV line connected with Seoni.
The schematic diagram is shown below:
Bina
400 kV
3x1500 MVA
350 km
800 kV
Khandwa
Seoni
Sipat
(3x660+2x500 MW)
400kV
Satpura
Bhilai
800kV
Wardha
Figure 3.6.1 Schematic Diagram of Sipat Seoni 800 kV line
Page 64
TypethenameofTB
3.6.2 INSULATION DESIGN FOR TRANSMISSION LINE

The nominal voltage for 800 kV transmission system is 765 kV, 800 kV being the maximum voltage. The lengths of
transmission lines are in the range of 250-400 km. For shunt compensation generally 240 MVAR line reactor is
used, which provides for about 50-65 % compensation for each end of the line with length varying from 300-400 km.
At the substation end bus reactors of rating 240 as well as 330 MVAR are used.
The switching overvoltage limit is obtained as 1.9 p.u.. The switching overvoltage studies have been carried out
both for line energization/de-energization as well as for line fault clearing case. The Overvoltage value having 2 %
probability of being exceeded is obtained as 1.67 p.u. for line fault clearing case while the same has been obtained
as 1.48 p.u. in case of line energization/re-energization.
For controlling switching overvoltage, closing reactor as well as line surge arresters are used. The rating for closing
resistor is chosen as 450 and it would be in the circuit for 8ms before closing the main breaker contact. For
clearing of fault, both single and three phase auto reclosing scheme is used. The information regarding switching
overvoltage design level of transmission line is given below:
Design level (p.u.)
1.9
Suppression measures for switching

overvoltage
closing resistor (450 for 8 ms) and line

surge arrester
Switching overvoltage (maximum and / or 2 %)

(p.u.) and occurrence case
1.67, line fault clearing
line energization/Re-energization, line fault

clearing
Yes (long lines)
Reclosing scheme
Single phase & three phase auto Reclosing

3.6.3 HIGHER PERFORMANCE SURGE ARRESTER
The rated voltage and Maximum Continuous Operating Voltage of the metal oxide surge arrester for 800kV system
are specified as 624kVrms and 490kV rms respectively. The discharge capability of the surge arrestor has been
specified as 13kJ/kV which gives the total discharge capability as 8.1MJ. From the V-I characteristics of the surge
arrestor, the switching impulse residual voltage is 1180kV at 1kA and 1220kV at 2kA. The lightning impulse
residual voltage is 1480kV at 20kA. The information is tabulated below:
Rated voltage (kV)
Residual voltage
(kV)
624
1180 kV peak at 1 kA
Switching impulse
Lightning impulse

3.6.4 INSULATION DESIGN FOR SUBSTATION
The substations so far planned for 800kV system are AIS type. The surge arresters in the substation are used at
the line entrance and also at the transformer and reactor terminals.
Page 65
TypethenameofTB
PROCESS FOR
LIWV
The severe lightning overvoltages are caused by the lightning within the substation. The lightning overvoltages
taking place on the overhead line several kilometers away from the substation are less harmful to the substation
equipment due to loss of a large portion of the lightning energy along the overhead line.
Studies have been carried out for both types of lightning surges i.e. back flashover and direct lightning stroke. In
case of back flashover, the lightning stroke on the overhead shieldwire or tower may generate overvoltages
resulting sparkover one insulator string. The arc across the insulator string causes transfer of part of the surge
current to the substation. The tower footing grounding also has a major impact on the overvoltages within the
substation as the major part of the stroke current flows into the ground. In case of direct lightning stroke the
penetration into the substation takes place via the phase conductor.
The details of the study result are given at Table 3.6.3.
MOSA layout
(unit)
AIS
Line entrance
1 unit
Busbar
Transformer
1 unit

Lightning
overvoltage
Representative
Value (kV)
Transformers : 1300
Calculation method
EMTP Analysis
Other : 1837
Safety factor
LIWV (kV)
considered for the severe condition
Transformer
1950
Other equipment
2100
Case and condition for lightning overvoltages studies
Back-flashover
Direct lightning
Lightning current (kA)

and its waveshape
200 kA peak, 4.2/50 micro-sec, Ramp Shape
Lightning stroke point
Tower 1
Substation circuit
1 & CB, 3 dia, 3 transformers, 2 lines with

line reactors, 1 bus reactor
Lightning current (kA)

and its waveshape
10 kA peak, 1.0/50 micro-sec, Ramp Shape
Lightning stroke point
Tower 1
Substation circuit
1 & CB, 3 dia, 3 transformers, 2 lines with

line reactors, 1 bus reactor
Page 66
TypethenameofTB
3.6.4.2.1 S WITCHING
PROCESS FOR
SIWV
OVERVOLTAGE STUDY
The switching overvoltage studies are carried out both for line energization/ deenergization and for fault clearing
cases. of the 800kV Sipat Seoni line. The random nature of high switching overvoltages at Seoni substation due
to energization/re-energization of the 800kV Seoni-Sipat line is studied with the help of statistical tests consisting
100 energization per test.
In order to simulate the worst case, a trapped charge of 1.0 p.u. in the line is considered under line energization
condition. High switching overvoltages are also observed during line fault clearing. The magnitudes of overvoltage
vary depending upon source strength, line length, fault types and fault locations. Studies have been carried out for
phase to ground and phase to phase faults at various locations on Sipat Seoni 800kV line.
3.6.4.2.2 D ETERMINATION
OF
SIWV
The representative phase to ground and phase to phase overvoltages having 2% probability being exceeded are
observed as 1.48pu (968 kV peak) and 2.36pu (1543 kV peak) respectively. The same for fault clearing case has
been observed as 1.67pu (1093kV peak) and 2.67pu(1742 kV peak) respectively. Considering safety factor for the
severe condition, the SIWV has been considered as 1550kV peak for phase to ground and 2480kV peak for phase
to phase.
The details of the results are given in Table 3.6.4.
Switching
overvoltage
Transformer
1093 kV ph-earth; (1.67 p.u.)
Other equipment
1093 kV ph-earth; (1.67 p.u.)
(Max and/or 2 % value)

(kV, p.u.)
Calculation method
EMTP Analysis
Safety factor
considered for the severe

condition
NIL, Less than 1000 m
SIWV (kV)
Transformer
1550
(phase-to-ground)
Other equipment
1550
SIWV (kV)
Transformer
2480
(phase-to-phase)
Other equipment
2480
7.6m(cond-cond)
9.4m (rod-cond)
Table 3.6.4 Determination Process for SWV

3.6.4.3 TOV
AND
E NERGY
ABSORPTION BY SURGE ARRESTER
The power frequency overvoltage on the sound phases due to 3-phase load rejection at the receiving end in 1phase to ground fault condition was studied. It has been observed that in some operating condition the temporary
overvoltage can reach as high as 1.4pu. This happened particularly during energizing of Sipat-Seoni line section at
Sipat on a permanent 1-phase to ground fault at the Seoni open line end. Further, studies was carried out for line
charging overvoltage following a load shedding (at Seoni end) or generation rejection (at Sipat end) in the SeoniSipat 800kV transmission system. In certain conditions the overvoltage may go close to 1.4 p.u. The energy
stresses in the surge arrestor may exceed its limiting value i.e. 8MJ, if the overvoltage persists for more than 1 sec.
Page 67
TypethenameofTB
The power frequency test voltage for the substation equipments are 870kV (RMS) for breaker and 975kV (RMS) for
CVT/CT, both for 1 min under dry and wet condition. The data for TOV and energy absorption by surge arrestor are
tabulated below.
less than 1.2 p.u.
1.4 p.u. lasting 1 sec
Ground fault followed by load

rejection
8.1
Calculation method
EMTP Analysis
Power frequency test voltage for substation equipment

Assumed overvoltage condition for
power frequency test
Transformer
Other equipment
870 kV RMS (for breaker)

and 975 kV (for CVT/CT),
dry & wet for 1 min
Analysis program
EMTP-RV

3.7 Korea 765 kV project
3.7.1 OUTLINE OF INSULATION DESIGN
The basic principle of KEPCO 765 kV insulation coordination is to protect the substation from the external
abnormal voltage with proper location of the lightning arrester of higher performance, and to clear flashover of the
abnormal voltage caused in the system. The basic premise is that the ultimate voltage of the 765 kV system design
shall be 800 kV in accordance with IEC pub 60038 and ANSI C84.1. The type of substation is full GIS and the bus
arrangement is based on the 1 circuit breaker system. The insulation design and the rated insulation level are
basically based on the IEC specifications and the ANSI specification may be applied in case of inevitability. The
number of thunderstorm day (IKL) is based on IKL = 20.
3.7.1.1 T ARGET
OVERVOLTAGE
The switching overvoltage studies are carried out both for line energization/ deenergization and the kind of
overvoltage is classified as external and internal overvoltage. For the external overvoltage, direct stroke and back
flashover from transmission line is considered for the lightning overvoltage calculation. For the internal overvoltage,
the temporary overvoltage caused by the system operation or by fault is calculated. And the switching surge
overvoltage due to CB operation and the ground fault surge overvoltage of the transmission lines is also calculated.
As for the power frequency temporary overvoltage, overvoltage of sound phase due to phase-to-ground fault and
load rejection overvoltage has been analyzed. As for switching overvoltage, various kinds of fault condition have
been analyzed on the premise of application of the circuit breaker equipped with the closing resistors, and the
internal overvoltage for design of insulation level is shown in Table 3.7.1.
Page 68
TypethenameofTB
Classification
Temporary
overvoltage
Switching
overvoltage
Overvoltage p.u.
Phase conductor (Bus, line)
1.2 (555 kV)
Neutral point (floating)
0.3 (139 kV)

()
()
Phase to ground
1.8 (1176 kV),

1.9 (1241 kV)
Basis of p.u.
1 p.u. = 800 /
3 kV
1 p.u. = 800x 2 / 3 kV
3.5 (2287 kV)
Phase to phase
Note: The phase to ground switching overvoltage is 1.8 p.u. for the substation, and 1.9 p.u. for the transmission line design.
Table 3.7.1 Internal overvoltage p.u.

As a result of analysis of back flashover on the premise of the substation being completely shield from lightning and
on condition that the first steel tower of the outgoing the transmission line is hit by lightning stroke, the lightning
overvoltage value (external abnormal voltage) are listed in Table 3.7.2.
Classification
Transformer
Switchgear/Bus
HSGS
Calculated value
1710 kV peak
2015 kV peak
2025 kV peak
Table 3.7.2 External overvoltage p.u.

3.7.1.2 B ASIS
OF SYSTEM GROUNDING
The system is effectively grounded. The ratio of zero sequence impedance vs. positive sequence impedance shall
be maintained within the following range at any point under the condition of single phase to ground fault.
R 0 / X 1 1, X 0 / X 1 3
Neutral grounding shall be done solidly, not through any impedance. Grounding the whole system shall be done in
consideration of the following conditions in order to maintain the effective ground range. As for neutral grounding, a
bank of maximum capacity at every substation in order to the whole system to be within the effective ground range
(below 75 % of effective coefficient) shall be grounded.
If there is a limit to LIWV or the power station / substation are not sufficient to the fault current enough to the relay
operation, all the necessary transformers may be grounded. But, the capacity of the grounded transformers shall
be more than of the power capacity. The single phase to ground fault current shall not be less than 60 % of the
three phase short-circuit current at any point. The voltage rise of the neutral point that is not grounded shall be
maintained within 138 kV. The voltage rise due to single phase to ground fault are as follows;
Vb E a
( a 2 1)Z 0 ( a 2 a )Z 2
Z 0 Z1 Z 2
Vc E a
(a 1)Z 0 (a a 2 )Z 2
Z 0 Z1 Z 2
The voltage rise of sound phase due to single phase to ground fault shall not exceed 120 % of the maximum
system voltage (phase voltage). That is, Vb = Vc = 800 / 3 kV 1.2 = 555 kV. The standard lightning arrester can
be selected considering some safety factor to the above calculated voltage.
Page 69
TypethenameofTB
3.7.1.3 P ROTECTION
CHARACTERISTICS AND ARRANGEMENT OF LIGHTNING ARRESTER
The key protection characteristics of the lightning arrester in Table 3.7.3 are as.
Classification
For use of phase conductor (Line, Bus)
Rated Voltage (kV rms)
580 kV class (note)
Nominal discharge current (kA)
20
Line discharge class
Residual voltage for very fast-front wave (kV)
less than 1600
Residual voltage for fast-front wave (kV)
less than 1500
Residual voltage for switching impulse (kV)
less than 1400
Discharge energy (MJ)
more than 5
Note: The rated voltage has been determined as 580 kV class so that both 576 kV of IEC and 588 kV of ANSI can be adopted.
Table 3.7.3 Key protection characteristics of lightning arrester

The lightning arrester for the 765 kV substation shall be installed at the following position:
- For 765 kV
Connection point of outgoing bus from GIS and the transmission line
Connection point of branch-off bus of GIS and outgoing bus of the transmission line
Both ends of the main bus
Primary terminal of the main transformer
- For 345 kV
Incoming and outgoing terminal of the transmission and incoming and outgoing terminal of
Secondary terminal of the main transformer
- For 23 kV
Tertiary terminal of the main transformer
Page 70
TypethenameofTB
3.7.1.4 R ATED
INSULATION STRENGTH OF EQUIPMENTING ARRESTER
The key protection characteristics of the lightning arrester in Table 3.7.3 are as.
(170 kA, 170 s)
Rated insulation strength
of equipment
Transformer
(kV peak)
Switchgear/Bus
(kV peak)
HSGS
(kV peak)
Lightning impulse
insulation
2050 (550) note
2250
2250
Switching impulse
insulation
1500
1425
1425
Temporary overvoltage insulation
800 (230) note
830
830
Specification applied
ANSI
C57.12.00-1987
IEC 60694-1980
Note: Insulation of neutral point of transformer
Table 3.7.4 Key protection characteristics of lightning arrester

3.7.2 ANALYSIS OF OVERVOLTAGE
The Figure 3.7.1 represents the items to consider for the insulation design and the flow of outline, and the 765 kV
power system map used for analysis is shown in Figure 3.7.2.
Main Features of System
Determination of Equipment Specification
Transmission Line Characteristics
- Lightning
- Surge Propagation
- Back-Flashover
Power System Characteristics

- Operating Voltage
- Protective Relay
- System Parameters
- Transformer Operation
- Reactive Compensation
Substation Arrangement
- Scale of Substation
- Equip. Arrangement
- Surge Arrester Location
- Insulation type
- Overvoltage factor
- Arrester Energy
Air Clearance Design

- Phase-to-Ground Clearance
- Phase-to-Phase Clearance
- Selection of BSL
Power Freq. Design
- AC Withstand Voltage
- Leakage Distance of Bushings
Surge Arrester
- Ratings
- Energy
Lightning Overvoltage
- Overvoltage of Equip.
- Arrester Energy
T.O.V
- 1L-G Fault
- Load Rejection
Lightning Design
- Shielding of Substation
- Selection of BIL
Circuit Breaker Design

- TRV Rating
- Pre-Insertion Resister
HSGS Design
- TRV Rating
- Voltage, Current Rating
Figure 3.7.1 Flow chart of insulation design of substation
Page 71
TypethenameofTB
Figure 3.7.2 765 kV power system map used for simulation

3.7.2.1 P OWER
FREQUENCY TEMPORARY OVERVOLTAGE
The length of the line shall not be longer than 160 km in order to restrict the line end voltage rise within 800 kV
(1.05 p.u.) due to charging capacitance during the normal system operation.
3.7.2.2 A NALYSIS
OF UNBALANCE RATE OF NEGATIVE SEQUENCE CURRENT
The negative sequence current per power station due to un-transposed line of the 765 kV system during the heavy
load has shown 5.0 % in unbalance rate. This is because the voltage is stepped up from 22 kV to 765 kV on the
generator terminal.
3.7.2.3 A NALYSIS
OF LIGHTNING OVERVOLTAGE
For analysis of lightning surge, EMTP, commonly being used for analysis of the transient phenomena, has been
used and it has been assumed that the back flashover has penetrated into the substation by lightning stroke on to
the first steel tower located nearest to the substation. The insulation design of the substation against the lightning
stroke propagation into the substation is affected by the parameter such as magnitude of lightning, polarity, wave
front, transmission line configuration, place of lightning stroke, tower-footing resistance, etc., and especially by the
configuration of the substation and the operation condition. Table 3.7.5 shows the calculated results and margin to
LIWV.
Page 72
TypethenameofTB
Location of
overvoltage
Overvoltage occurred (kV)

Location of
overvoltage
Overvoltage
occurred (kV)
LIWV (kV)
Margin (%)
Remark
Incoming part of
substation
1937
2147
2250
4.8
10.8 % increased
Bus part of
substation
2120
2099
2250
7.2
1.0 % decreased
Transformer
1715
2050
19.5
Table 3.7.5 Analyzed result of LIWV of gas insulated bus

It can be proved that there will be about 5 % of margin in the overvoltage since the maximum overvoltage occurred
in the incoming part of the gas insulated bus is 2147 kV and LIWV is 2250 kV. The considered margin of about 5 %
is to be small, but it can be judged to have enough margin, because the simulating of the most severe operating
condition of single line and single bus and considering the improvement of performance of the lightning arrester to
be installed in the future.
3.7.2.4 A NALYSIS
OF SWITCHING OVERVOLTAGE
Switching overvoltage to be considered at insulation design for the transmission system can be divided into three
different types according to the cause of occurrence, the first one is, overvoltage due to phase to ground fault, and
the second one is overvoltage due to fault clearing and the last one is overvoltage due to line energizing. In case of
overvoltage due to line energizing, reclosing overvoltage may be divided separately considering residual electric
charge remains.
As for the 345 kV system and lower, switching overvoltage is not critical to consider since the magnitude of
overvoltage due to phase to ground fault and overvoltage due to fault clearing are relatively lower than the
overvoltage due to line energizing. However, as for the 765 kV system and above, the overvoltage due to phase to
ground fault and overvoltage due to fault clearing may show higher than the overvoltage due to line energizing
since the target value of the overvoltage due to line energizing is suppressed by pre-insertion resister.
The EMTP has been used for this analysis. The systems in 2005 and 2010 have been selected as the object and
the overvoltage of each case have been reviewed respectively. The calculation results of the switching overvoltage
are as follows:
Page 73
TypethenameofTB
(Unit: p.u.)
Classification
kind of surge
Phase to ground
Phase to phase
Remark
T/L
Ground fault, Energizing/

Re-energizing,
Fault clearing
1.9
3.5
Closing resistor
1000
Ground fault
1.6
2.3
Fault clearing
1.7
2.9
Energizing/
Re-energizing
1.8
3.5
Summary
1.8
3.5
0.3
S/S
Neutral point
of Tr.
Single phase ground fault
Supposing
failure of resistor
closing
Notes: 1 p.u.= 800 2 / 3 kV

Voltage rise at the neutral point during floating operation has been targeted 0.3 p.u. (r.m.s) since voltage rise at the neutral point
due to single phase ground fault has been set as object to restrict.
Table 3.7.6 Analyzed result of LIWV of gas insulated bus

3.7.3 INSULATION STRENGTH OF EQUIPMENT OF 765 KV SUBSTATION
3.7.3.1 F LOW
OF INSULATION COORDINATION
Insulation coordination in the power system shall mean a series of work in order to select each kind of equipment
comprising the system and insulation strength. The flow of insulation coordination according to the condition of IEC
60071-1 has been shown in Figure 3.7.3.
3.7.3.2 S ELECTION
OF REPRESENTATIVE OVERVOLTAGE
The representative overvoltages within the selective range are shown for the insulation design of the object
substation based on the analyzed results of lightning overvoltage and switching overvoltage shown in Table 3.7.7.
Table 3.7.8 shows the conditions and assumption of calculation conditions in representative overvoltage.
Page 74
TypethenameofTB

voltage
- Protective level of overvoltage
limiting device
- Insulation characteristics
Power system analysis
Determination of representative voltage

(Normal voltage and overvoltage)
- Insulation characteristics
- Performance criterion
- Coordination factor
(Statistical distribution,
Inaccuracy of input data)
Selection of insulation meeting the

performance criterion
Co-ordination withstand voltage
Application factors to account for the

differences between type test conditions
and actual service conditions
- Atmospheric correction factor

- Safety factor
(Equipment test assembly,
Dispersion in production,
Quality of installation,
Ageing in service, etc.)
Determination of required
withstand voltage
Selection of standard withstand voltage
- Test conditions
- Test conversion factor
- Standard withstand voltage
Selection of rated or standard

Insulation level
Figure 3.7.3 Flow chart for the determination of rated or standard insulation level
Kind
Value (kV)
Remark
Representative overvoltage of
power frequency
800 / 3 x 1.2 = 555
Voltage rise of sound phase due to

single phase to ground fault.
r.m.s value
slow front
fast front
Representative overvoltage
of slow front(phase to phase)
2 800/ 3 1.8 = 1176
1430-2020
2 800/ 3 3.5 = 2287
phase to ground
peak value
Lightning overvoltage for individual
equipment
peak value
(phase to phase)
peak value
Table 3.7.7 Representative overvoltage
Page 75
TypethenameofTB
Representative overvoltage
Conditions for assumption
Power frequency overvoltage
Voltage rise of sound phase during

single phase to ground fault
Reclosing overvoltage due to failure of pre-insertion

resistor closing
Lightning
overvoltage
Lightning current
170 kA, 1/70 s
Type of lightning stroke
Back flashover on to the first steel tower
Tower footing resistance
15
Configuration of substation
Full GIS of 1.5 circuit breaker system
Table 3.7.8 Conditions for assumption for selection of representative overvoltage

- Selection of coordination withstand voltage
The proper value of coordination factor shall be assumed in order to determine the coordination withstand voltage
as the representative overvoltage. It is general to calculate in consideration of the accuracy of the result of analysis
and the coordination factor which is assumed as 1.0.
- Selection of rated insulation level
Once the necessary withstand voltage is determined in consideration of the proper safety factor from the
coordination withstand voltage, the standard withstand voltage is selected based on the above. Lightning impulse
voltage is selected per each equipment, because lightning overvoltage occurred to each equipment in the
substation is different from it's magnitude.
Same switching overvoltage factor was selected, because the overvoltage due to energizing/re-energizing the
transmission line is the largest and this is applied to all equipment. Both the rated value and the selected value
which are available for selection as KEPCO's rated insulation strength are shown in Table 3.7.9.
Page 76
TypethenameofTB
Classification
Switching impulse
Transformer
Standard withstand
voltage (kV)
1425 I, 1500 A
Maximum
overvoltage (kV)
1176
1.32
1950 I
1.14
2050 A
1715
1.20
1.23
2100 I
1.04
2015
Bus and switchgear
1.12
2250 I
2100 I
1.11
1890
Sh.R/HSGS
ACWV
1.21
1550 I
2100 I
Lightning
impulse
Safety factor
2250 I
1.19
Transformer
800 A
1.44
Bus and switch
830 I
Sh.R/HSGS
830 I
555
1.50
1.50
Notes : IEC code, A : ANSI code

Thick Italic type letter underlined stands for the selected insulation strength
Table 3.7.9 Withstand voltage available for selection of rated insulation level
3.7.3.3 N EUTRAL
INSULATION OF THE
765 K V
TRANSFORMER
Insulation strength of the transformers which have the maximum voltage of 300 kV and above have been designed
as non-uniform insulation transformer. The neutral point of the non-uniform insulation transformer have been
directly / indirectly connected with the ground and in case of indirect connection with the ground, there shall be
installed proper protection device between the neutral point and the ground in order to restrict the overvoltage
(switching and lightning overvoltage, and etc.).
Table 3.7.10 below described the rated insulation strength of the transformer neutral point for the 765 kV system.
The rated insulation strength has been selected from the standard lightning impulse withstand voltage in
consideration of the safety factor of more than 1.25 at the protective level of lightning impulse of the lightning
arrester. The power frequency temporary overvoltage has been selected from the rated insulation strength referring
to the relation between the power frequency temporary overvoltage and the rated insulation strength which is
recommended by IEC.
Page 77
TypethenameofTB
Item
Value
Remark
Rated voltage of LA for

neutral point
144 kV
Protective level of LA for

neutral point
less than 450 kV
Protective level of lightning

impulse
Rated insulation strength
550 kV
Safety factor 1.25
Power frequency temporary

overvoltage
230 kV
Assumed from rated insulation

strength
Table 3.7.10 Rated insulation strength of the transformer neutral point

3.7.4 AIR CLEARANCE OF INCOMING PART OF 765 KV SUBSTATION
3.7.4.1 D EFINITION
OF
A IR
CLEARANCE
- Air clearance to ground

Minimum value: The minimum air clearance to ground means that the clearance between the metal and the ground
or the metal and the non-insulation materials of same voltage shall not be less than this value in any case.
Standard value: The standard air clearance to ground shall be such that the clearance between the center of the
conductor and the ground or the center of the conductor and the non-insulation materials of the same voltage shall
be designed, if no condition is specified otherwise.
- Phase to phase air clearance
Minimum value: The minimum phase to phase air clearance means the value that the clearance between metal to
metal shall not be less than that value in any case.
Standard value: The standard phase to phase air clearance mean such value that the clearance between the
center of the conductor shall be designed based on it, if no condition is specified otherwise.
Figure 3.7.4 Standard insulation distance of substation outgoing part
Page 78
TypethenameofTB
3.7.4.2 A IR
CLEARANCE
The air clearance between phase to ground and phase to phase of the 765 kV bus shall be in accordance with
Table 3.7.11.
Nominal
voltage (kV)
Air clearance (mm)
765
Standard
Minimum
Outdoor
Phase to phase
Phase to ground
11000
8500
7000
5000
Table 3.7.11 Design standard for air clearance of outdoor bus

3.7.5 LOCATION OF LIGHTNING ARRESTER AT INITIAL OPERATION OF 765 KV GIS
3.7.5.1 B ACKGROUND
OF ANALYSIS AND OUTLINE
It has been analyzed if installation of lightning arrester is necessary for the primary bus of the 765 kV GIS and for
the secondary main bus, incoming part and the shunt reactor terminal of the 345 kV GIS and if the overvoltage
exceeds the LIWV at the initial operation of the 765 kV substation.
The initial scheme of the 765 kV gas insulated substation is 4 circuits of transmission line and 2 banks of
transformer, half of the final size, and in this case the necessity of installation of the lightning arrester for the both
ends of the main bus has been analyzed using EMTP (Electro-Magnetic transient Program).
3.7.5.2 C ONCLUSION
OF ANALYSIS
- 765 kV side
The lightning arrester shall be installed at the both ends of the main bus even at the initial operation with 4 circuits
of transmission line and two banks of transformer and positioning lightning arrester shall be within six to eight
meters from the bus connection, so that the overvoltage can be reduced. The maximum overvoltage due to
lightning stroke in case of installation of lightning arrester at the both ends of the bus is 1765 kV and it has about
27 % margin for LIWV 2250 kV.
- 345 kV side
The maximum overvoltage will not exceed the standard LIWV allowing that the lightning arrester is installed at the
incoming part of the transmission line only and not installed for the bus and the shunt reactor additionally. The
maximum overvoltage at the incoming part and bus is 830 kV and has about 42 % margin for LIWV 1175 kV. The
maximum overvoltage of the shunt-reactor and the main transformers is 827 kV and has about 27 % margin for
LIWV 1050 kV.
3.7.6 LIGHTNING ARRESTER RATING 765 KV SYSTEM
3.7.6.1 R ATED
VOLTAGE OF LIGHTNING ARRESTER
The rated voltage of lightning arrester is the maximum permissible power frequency r.m.s value on both ends of
lightning arrester for operating duty test and it is normally decided based on power frequency temporary
overvoltage, that the voltage rise of sound phase during single line to ground faults.
The multiple factor of power frequency temporary overvoltage applied to the lightning arrester for the line is 1.2 p.u.
which was selected according to the guidelines for insulation design of the transmission line and its analysis.
Based on the above, the rated voltage of the lightning arrester for 765 kV system is shown in the Table 3.7.12.
Page 79
TypethenameofTB
Nominal voltage of Maximum voltage of

system (kV)
system (kV)
765
Power frequency
temporary overvoltage
(p.u.)
IEC code
(kV)
ANSI code
(kV)
1.2
(For phase)
2424 = 576
2425 = 600
588
612
0.3
(For neutral )
1212 = 144
1213 = 156
144
800
Table 3.7.12 Rated voltage of lightning arrester for 765 kV system

3.7.6.2 N OMINAL
DISCHARGE CURRENT AND DISCHARGE CLASS
The discharge characteristics are represented as the line discharge test of the lightning arrester and the energy
absorbed by the lightning arrester has been reviewed. Based on the above, the nominal discharge current and the
line discharge class to be applied to the 765 kV systems have been shown in Table 3.7.13.
Use
Nominal discharge current
Line discharge class
Phase
20 (kA)
Neutral
10 (kA)
Table 3.7.13 Nominal discharge current and discharge class of lightning arrester
for 765 kV system
3.7.6.3 P ROTECTIVE
CHARACTERISTICS OF LIGHTNING ARRESTER
Characteristics
580 kV class, 20 A
( phase )
144 kV, 10 kA
( neutral )
Discharge voltage of
fast-front wave
Protective level of
lightning impulse
Protective level of
switching impulse
less than 1600 kV
less than 1500 kV
less than 1400 kV
less than 500 kV
less than 450 kV
less than 400 kV
Notes: Discharge voltage of fast-front wave = Protective level of lightning impulse 1.1
Protective level of switching impulse = Protective level of lightning impulse (0.85-0.9)
Table 3.7.14 Protective characteristics of lightning arrester for 765 kV system

3.7.7 SHIELDING OF 765 KV GIS SUBSTATION
3.7.7.1 A NALYSIS
OF HEIGHT AND DISTANCE OF SHIELDING WIRE
Most of overvoltage in the substation is caused by lightning surge due to back flashover from the transmission line
and the direct lightning stroke. The required insulation strength of the substation equipment is determined mainly
by lightning surge due to back flashover. In case of a direct lightning stroke on to the equipment in the substation,
there may be a outage of power supply. In order to prevent outage it is a trend to protect the facilities from the
direct lightning stroke with installation of the overhead shielding wire. As for the 765 kV system facilities of the 765
kV substation, live parts are exposed up to the outgoing steel structure but the other facilities are enclosed in GIS.
Therefore, the height and distance and etc. of the overhead shielding wire to be installed on the dead-end steel
structure have been reviewed. At the same time, the voltage rise between the enclosure struck by lightning and the
relevant phase conductor, that is caused by lightning stroke and substation ground resistance, has been reviewed
by EMTP and the necessity of the shielding wire over the 765 kV switchyard has been reviewed.
Page 80
TypethenameofTB
3.7.7.2 E XAMPLE
OF ANALYSIS
The proper height and distance of the overhead ground wire in the 765 kV substation are in Table 3.7.15 below. It
is assumed that the 765 kV section of 765 kV substation is full GIS type, so the live part is the outgoing bushing
terminal of the transmission line and this is located under the dead-end steel structure. According to this calculation,
it is necessary to construct the gantry tower of 60 meter height for the overhead grounding wire every 2 Bay
(distance: about 130 m) in case that the height between the live part of the in/outgoing bushing of GIS and the
ground is 15 meters.
height of
shield wire
30m
35m
40m
45m
50m
55m
60m
10m
99.2
108.6
116.4
123.0
128.4
132.8
136.4
15m
87.6
99.2
108.6
116.4
123.0
128.4
132.8
20m
72.9
87.6
99.2
108.6
116.4
123.0
128.4
height
of live part
Table 3.7.15 Proper distance of the overhead ground wire

3.7.7.3 R EVIEW
OF PROTECTION OF
GIS
ENCLOSURE FROM DIRECT LIGHTNING STROKE
- Purpose of review
In general, the outdoor steel structure type substation has the overhead ground wire over the switchyard in order to
protect the equipment from lightning. However, in case of GIS substation, since the GIS enclosure is grounded, it
seems not to flashover on the conductor even if lightning strikes the enclosure. The in/outgoing bushing to be
installed near the gantry structure can be protected by the overhead grounding wire to be installed over the gantry
structure. Therefore, voltage rise and possibility of lightning stroke have been reviewed by using EMTP and the
possibility of lightning stroke in order to judge the necessity of overhead grounding wire over the switchyard.
- Result of analysis
Lightning overvoltage between the bus conductor and the enclosure caused by the direct lighting stroke on to the
enclosure is proved to be about 1311 kV in the worst case, that is, the case of the substation ground resistance of
2 and further lightning stroke of 200 kA. In case the LIWV of GIS is designed 2250 kV, this value proved to be
only about 58 %. It has been proved that the overhead grounding wire over the GIS switchyard would not be
necessary in consideration that the substation ground resistance could be kept below 1 and the lightning stroke
of 200 kA class might occur one time after 300 years.
Page 81
TypethenameofTB
References
Japan
[1] A. Nakamura, et al., 1100 kV AC Transmission Project in Japan, International Symposium on International
Standards for Ultra High Voltage, 2007
[2] H. Okamoto, Progress of, and prospects for, commercial applications of UHV 1100 kV Transmission System in
TEPCO; Issues un upgrading existing 1100 kV designed lines from 550 kV to 1100 kV, Second International
Symposium on Standards for Ultra High Voltage Transmission, 2009
[3] E. Zaima, et al., Insulation Coordination for UHV AC Systems based on Surge Arrester Application (CIGRE
C4.306), Second International Symposium on Standards for Ultra High Voltage Transmission, 2009
[4] T. Watanabe, et al., Insulation Coordination for UHV System, CIGRE paper 33-101, 1998
[5] T. Takebe et al., Insulation Design for UHV Transmission Line in TEPCO, ISH 2005
[6] Y. Ishizaki, et al., High Performance Surge Arresters for UHV Systems, Second International Symposium on
Standards for Ultra High Voltage Transmission, 2009
[7] T. Kobayashi, et al., Compactness and High-reliability Technology for UHV Circuit-breakers, disconnectors and
earthing switches, Second International Symposium on Standards for Ultra High Voltage Transmission, 2009
[8] Y. Yamagata, et al., Utilities experience on design and testing for UHV equipment in Japan, Second
International Symposium on Standards for Ultra High Voltage Transmission, 2009
[9] Y. Yamada, et al., Experimental Evaluation of a UHV Tower Model for Lightning Surge Analysis, IEEE Trans.
on Power Delivery, vol.10, no.1, pp.393-402, 1995
[10] T. Shindo, et al., A New Calculation Method of Breakdown Voltage-Time Characteristics of Long Air Gaps,
IEEE Trans. on Power Apparatus and Systems, vol.104, no.6, pp.1556-1563, 1985
Brazil
[1] Itaipu Transmission System Consolidation of FURNAS Experience - Book published only in portuguese
under the coordination of S. S. G. MOTTA and H. A. M. FADINI ISBN 85-85996-01-3 - Rio de Janeiro 1995
[2] Standards Library (International Electrotechnical Commission IEC)
[3] America: Achieving its Potential in Electrical, Electronic and Multimedia Technologies IEC Latin America
Regional Centre Seminar 2007 (IEC-LARC) - May 2nd and 3rd, So Paulo - Brazil
[4] Regulatory Acts Library of the Brazilian Regulatory Agency for electric energy (ANEEL)
[5] HVDC Converter transformer performance on Itaipu. system - Gilson Bastos, FURNAS (Brazilian Cigr
National Committee Seminar; published only in Portuguese)
[6] Real-Time Simulations for Long-Term Stability Analysis of FURNAS 750 kV HVAC Transmission System, S.
Esprito Santo et allis, XIII SNPTEE-1995, Cigr-Brazil, Florianpolis, Brazil, October 1995 (Brazilian Cigr National
Committee Seminar; published only in Portuguese)
[7] Technical Requirements for UHV Switching Equipment Hiroki Ito et allis (CIGR Working Group A3-22);
IEC/CIGRE UHV Symposium Beijing, July 2007.
Page 82
TypethenameofTB
4 Overvoltage in UHV range

4.1 Determination of stresses (TOV, switching overvoltage, lightning overvoltage
and VFTO)
Electrical equipment has to withstand different voltage stresses occurring in the network. Besides the continuous
power frequency voltage, temporary overvoltages and transient overvoltages have to be considered.
Representative maxima of amplitudes according to their time duration are generally described in Figure 4.1.1. The
values are referred from EHV systems. These stresses should be determined by simulation tools (EMTP etc) and
verified by measuring results. The specifications of each overvoltage are introduced briefly.
Figure 4.1.1 Representative maxima of amplitude of overvoltages Urp (per unit

values)
(1) TEMPORARY OVERVOLTAGE (TOV)
TOV includes healthy phase overvoltages due to the transmission line ground faults and TOV due to the load
rejections. In the case of sudden load rejection at the remote end of a heavily loaded long line such as a UHV
system, the voltage jump due to Ferranti effect under self-excitation of generators, resulting in a TOV, which will
lasts until the line is tripped at both ends. The amplitude of the TOV under such circumstances is system
dependent, but EMTP simulations show a typical maximum TOV of approximate 1.5 p.u. lasting the order of 1s.
While the peak values of TOVs caused by load rejections can be suppressed by surge arresters, their durations are
restricted by the energy absorption capability of surge arresters. The reason being that the ratio of protection level
and operating voltage is lower in the UHV class than in the lower voltage classes. The protective relay system is
planned to apply to avoid breakdowns of surge arresters due to lingering TOV in some network.
In Chinese UHV system, the TOV can be limited by high voltage shunt reactor for the long UHV ac transmission
line where the var supply from the line capacitance is greater than that the system can absorb. Since the reactive
power of the shunt reactors compensate partially the capacitive power of the line, the Ferranti effect is reduced,
therefore TOV is limited and the self-excitation of generators is prevented as well.
Page 83
TypethenameofTB
(2) SLOW-FRONT OVERVOLTAGE (SWITCHING OVERVOLTAGE)

Slow-front overvoltages (closing, opening, and ground fault overvoltages) are of particular importance for UHV
systems because of the saturation effect of the switching impulse strength of the air insulation due to the enhanced
leader propagation in air in the frequency and voltage steepness range in question. Among these overvoltages,
ground fault overvoltages are generated independently of circuit breaker switching, and there are generally no
effective means of controlling them, except those near substations that can be reduced by surge arresters. In
contrast, closing and opening overvoltages can be controlled effectively by the insertion of closing and opening
resistors. The target of reducing switching overvoltage is basically ground fault overvoltage level.
Overvoltage due to fault clearing occurs on the faulted phase, when the faulted line is switched off. A subsequent
insulation failure may occur on the part of the system still in operation. This type of overvoltage should be carefully
investigated for a given system.
In Chinese UHV system, shunt reactors of transmission lines can suppress the component of steady power
frequency voltage, on which the switching overvoltage superposed during energizing unload line. Since the TOV
can be reduced by shunt reactors, the line surge arrester can select a lower rated voltage, as well as lower SIPL to
reduce switching overvoltage.
(3) FAST-FRONT OVERVOLTAGE (LIGHTNING OVERVOLTAGE)

Fast-front overvoltages, such as lightning overvoltage, have the wave head of several micro-seconds, and are one
of important factors to determine the insulation designs of substation equipment. In UHV transmission system, UHV
equipment become larger in its dimension and this leads to difficulties in transportation, factory assembly, and onsite installation. Thus, test voltages for those equipment should be carefully lowered to a reasonable level by
employing higher performance surge arresters and analyzing the various overvoltage with the design parameters of
individual UHV substation. Since lightning is a natural phenomenon, it is impossible to forecast the lightning events
that UHV substations encounter, but possible to design the insulation level of UHV substation quite precisely based
on the observation data of recent lightning events and the technical improvements in overvoltage transient
simulations. To evaluate the lightning overvoltage, both back-flashover overvoltage and direct lightning overvoltage
should be studied. Although the simulation conditions such as circuit conditions, location of surge arresters, and
lightning currents for the overvoltage analysis are different in each project in each country, but every project
reduces LIVW to a reasonable level by installing higher performance lightning arresters with appropriate protective
characteristics on bus-bars near transmission line bays and/or transformer bays. CIGRE C4.302 proposes the
conversion method from the simulated waveforms to standardized lightning impulse waveforms, which promotes
the optimization of insulation designs.
(4) VERY FAST TRANSIENT OVERVOLTAGE (VFTO)

The disconnector restriking surge is an oscillation surge with very high frequency of several MHz. The disconnector
restriking surge can exceed the lightning impulse overvoltage in UHV GIS-type substations; these substations
insulation design should be coordinated accordingly. The disconnector restriking surge largely depends on the
design of disconnectors used, and the surge voltage reaches 2.8 p.u. in several cases of the disconnectors with
fast switching speed, 2.2 p.u. in the case of that with slow switching speed. The frequency of disconnector
restriking surge is much higher than that of lightning surges because every disconnector operation potentially
generates the overvoltage, and the surge could impose negative effects not only on main circuit insulation but also
on secondary system such as EMC. Thus, disconnectors with pre-insertion resistor can be installed to reduce the
overvoltages, and in some projects, the disconnector restriking surge level is reduced to less than 1.3 p.u.
Page 84
TypethenameofTB
4.2 TOV due to load rejection and ground fault

4.2.1 GENERAL
Temporary overvoltage (TOV) can cause representatively from faults, switching operations such as load rejection,
combination with fault and load rejection, and resonance phenomena occurred by large power transformer
energization or shunt reactor TOV may be slowly damped in general and its property is characterized by the
amplitude and the duration. TOV frequency band that is described in IEC 60071-1 is several times smaller and
higher than power frequency. [1]
A fault on single phase-to-ground in the three phase transmission system causes the overvoltage, which is
influenced on the two other healthy phases at any points in the system. The waveform of TOV in case of the
ground fault is close to sine wave in general case. In the mean time, a load rejection results in TOV of phase-toground depending on disconnected load, system configuration after rejection and control characteristics of source
including speed and voltage regulator of generator. Moreover, in some cases, Ferranti effect leads to an increase
in the voltage at a remote end of a long transmission line after load rejection. As for the combination load rejection
after ground fault, the overvoltage may lead to most sever level in some cases. In Chinese case, the shunt reactor
is adopted as one of the solution to suppress the TOV. After a fault occur on the line, open operation of circuitbreaker (CB) at the load side system is done and a load rejection overvoltage is caused by the disconnected load
until CB opens in the supply side system.
The amplitude and the duration of the overvoltage depend on configuration of transmission system, specifications
of equipments including arrester, protection system and so on. To consider the various conditions of power network
such as this, the computed calculations are useful for accurate estimation of waveform of TOV and its occurrence
probability. TOV affects decisions of surge arrester specification as well as insulation design of the equipments. As
low protection levels of surge arrester are selected, TOV duty of surge arrestor increases more increase. The
energy dissipation capability of surge arresters is the predominant factor for determination of their protection
characteristics. The necessary energy capability of the surge arrester under TOV depends on the overvoltage
amplitude and the duration. The amplitude of TOV level in UHV system is higher than in EHV systems and
therefore is required to be controlled lower.
The information of presumed TOV amplitude and duration in the system would help to design of equipments that
are considered with insulation coordination, energy absorption and magnetic core.
4.2.2 STUDIES FOR DIFFERENT COUNTRIES

Table 4.2.1 and Table 4.2.2 summarize condition of temporary overvoltage studies for UHV and 800 kV systems of
different countries. The load rejection with ground fault is considered by calculations in some countries and its
overvoltage values are comparably higher. The amplitude of TOV ranges from 1.4 to 1.5 p.u. in UHV systems and
from 1.09 to 1.4 p.u. in 800 kV systems are reported. The energy absorption of surge arrester against TOV of UHV
system are specified from 40 to 55 MJ and these values are higher than the different values specified for 800 kV
and EHV systems.
Page 85
TypethenameofTB
Country/Utility
China/SGCC
Highest Voltage (kV)

TOV (p.u.)
Duration (sec.)
1100
1.4
0.5
India/Power
Grid
1200
1.4
1
Case and condition for

studies
Single phase
ground fault,
Load rejection
Ground fault
followed by
load rejection
N/A
40
55
EMTP analysis
EMTP analysis
Energy absorption of
surge arrester (MJ)
Calculation method
Italy/CESI
Japan/TEPCO
Russia/
1050
N/A
N/A
1100
1.5
0.17
Load rejection
followed by
ground fault
with selfexcitation of
generators
1200
1.3
3
N/A
55
40 *1
N/A
EMTP analysis
N/A
N/A
*1 Estimated from a specified current impulse withstand capability of 2.8 kA with a shape of 3/8 ms and used MOSAs.
Table 4.2.1 Condition of temporary overvoltage studies for UHV system

Country/Utility
TOV (p.u.)
Duration (sec.)
studies
Brazil/Furnas
800
1.25
1 cycle of 60 Hz
load rejection, line
energization, transformer
no-load energization, or
auto reclosing schemes
operation
China/SGCC
800
1.4
N/A
India/Power Grid
800
1.4
1
Single phase ground fault,

Load rejection
Ground fault followed by

load rejection
surge arrester (MJ)
N/A
15.27kJ/kV
8.1
Calculation method
EMTP/TNA analysis
EMTP analysis
EMTP analysis
Country/Utility
South Korea/KEPCO
The United States/AEP

TOV (p.u.)
800
1.2
Duration (sec.)
10
800
1.09
Relays trip for 1.05 p.u. / 5
min.

studies
Load rejection, single line

to ground fault
Transformer Energization
caused sympathetic
saturation of other
transformers at that bus
Ground fault followed by

load rejection
surge arrester (MJ)
2.6 kJ/kV
7.76 (for station arresters)

4.7 (for line arresters)
7.7
Calculation method
EMTP analysis
EMTP/PSCAD analysis
ATP/EMTP analysis
Venezuela
/EDELCA
800
1.4
1
Table 4.2.2 Condition of temporary overvoltage studies for 800kV system
Page 86
TypethenameofTB
4.2.2.1 JAPANESE EXPERIENCE

As the single phase to ground fault leads the voltage of the faulted phase to zero, the voltage of healthy phases
increase depending on a ratio between zero phase sequence impedance X0 and positive phase sequence
impedance X1. Figure 4.2.1 shows an example of relationship between overvoltage on healthy phase and the ratio
X0 / X1, which was reported in CIGRE Technical Brochure 362. For UHV system, the X0 / X1 range become
relatively smaller than that in EHV systems and the ground fault factor tends to be smaller.
Figure 4.2.1 General tendency of ground fault factor in Japan

Figure 4.2.2 shows TOV in healthy phases during faults, which is simulated for TEPCOs future 1100kV system [2,
5]. The maximum TOV value at ground fault is nearly 1.1 p.u. at substation, and 1.15 p.u. at transmission line.
These results are reflected to the pollution design.
Page 87
TypethenameofTB
UN
UN
U2
U1
U2
U3
U3
UN
U2
U1
U5
U3
U1
U2 U5 U4
U3
p.u
1.2
p.u
1.2
Voltage on sound phase(pu)
UN
LG
LG
1.1
1.1
1.0
1.0
U2
UN U3
U2
U5
U4
U2
U1
UN U3
For future Pj,

1100
which is not
1 pu
kV
explained
3
here.
Resistance at fault point:0ohm
U2
U5
For future Pj,

which is not
explained
here.
U4
1 pu
U1
1100
kV
3
Resistance at fault point:0ohm
Figure 4.2.2 TOV in healthy phases during faults (Japan)
Overvoltage
The load rejection with the fault at long line that is connected through a transformer to a generator causes TOV
amplitude by Ferranti effect under self-excitation of generator. Figure 4.2.3 shows example of analyzed TOV in
1100kV system of Japan [3, 4, 6]. Such a case is quite rare in the system because the 1100kV system is looped
with the existing EHV grid. The simulation shows typical maximum TOV waveform whose amplitude and duration
are 1.5 p.u. and 0.17 second. Since the ratio of protection level and operating voltage is lower in the 1100kV class
than in the lower voltage classes, the energy absorption by a surge arrester is higher and the energy rating of an
arrester is specified as 55MJ and above in 1100kV system of Japan. To avoid an overload of the surge arrester,
the overvoltage protective relay system to open no-load 1100kV transmission line concerned will be applied.
Figure 4.2.3 Example of analyzed TOV in 1100kV system of Japan
Page 88
TypethenameofTB
4.2.2.2 CHINESE EXPERIENCE

In Chinese case, surge arresters are not normally sized to control temporary overvoltages except in the case of
without application of line shunt reactors, In that case TOV durations are restricted by the energy absorption
capability of surge arresters to avoid thermal runaway. Special temporary overvoltage protection relay can be used
to trip the three phases CB at both end of the UHV lines synchronously after load rejection or load rejection in
combination with earth fault, thus limit the duration of temporary overvoltage. The relay can be acted by checking
three phases CB tripped at any one end of line and measured TOV or line arresters energy.
In Chinese UHV AC system, the following three kinds of resonance or ferroresonance voltages are generally
considered.
- Resonance overvoltage in non-three-phase operation
- Resonance overvoltage caused by energizing no-load transformer
- Ferroresonance overvoltage
In a line with shunt reactors, resonance overvoltages may appear on opening phases during non-three-phase
operation. The resonance overvoltage is caused by the capacitive coupling between sound phases and opening
phase of the line and it can be limited by a grounding neutral reactor equipped on the neutral point of three phase
shunt reactors.
Energizing no-load transformer at long line end may conduct the transformer core magnetically saturated and
causes inrush current and periodic changes of the excitation inductance. When the frequency of some harmonic in
the inrush current matches the self-oscillation frequency of the system, high resonance overvoltage may be
generated.
Ferroresonance overvoltages developed from the interaction of capacitive and inductive elements possess
nonlinear magnetizing characteristics. In UHV system, the overvoltage is usually initiated by the resonance of GIS
bus potential transformer through circuit-breaker capacitance during isolation of a bus where potential transformer
is installed.
The shunt reactors compensate degree to the line reactive power is usually in the range of 60%90%, depending
on allowed TOV, system can absorb reactive power, and load transmission capacity. The higher compensate
degree is, the lower TOV is. In order to avoid power frequency resonant overvoltage in non-three-phase operation
near 100% compensate degree is not allowed.
IEC60071-2 points out that usually the selection of the rated voltage of the surge arrester is based upon the
envelope of the TOV expected, taking into account the energy dissipation capability of the surge arrester. The
energy capability of the MOA under TOV stress is expressed as an amplitude/duration characteristic furnished by
the manufacturer. Therefore the rated voltage of the arrester usually is chosen same as the amplitude of the TOV
where the arrester is installed. Although the TOV on line side(e.g. 1.4p.u) is higher than bus side(e.g. 1.3p.u) The
arresters at line side select with same rated voltage of the bus side, taking advantage of the excellent withstand
TOV amplitude/duration characteristic of the UHV higher performance arresters. The lower TOV is, the lower rated
voltage of the arrester is selected.
Page 89
TypethenameofTB
In the 1100kV system of China, since the transmission lines is long, the high voltage shunt reactors are installed.
The shunt reactors has the following function:
(1) To limit the TOV no more than 1.4 p.u. at the line side and no more than 1.3 p.u. at the bus side.
(2) To compensate the reactive power produced by line capacitance, smooth the voltage distribution along line,
prevent self-excitation of generators and favor synchronism parallel operation of the line CB.
(3) To suppress the second arc current and transient recover voltage during single phase ground fault and favor
the fast single phase reclosing, coordinating with neutral reactor of the shunt reactor.
(4) To help reduction of the rated voltage of arrestor and switching overvoltage.
The maximum TOV occur in the single-phase ground fault case, in which CBs at the fault line side operate with
three-phase tripped and the CB at the other line dont operate. This case occurs under follow situations.
(1) In case the single phase reclosing operation after the single line fault is not successful, the CBs at three phases
are tripped by relay protection.
(2) In case of Live-Line Working, the single phase reclosing operation is forbidden and the CBs at three phases are
also tripped when a single phase ground fault occurs.
The relationship between TOV and compensation degree of the shunt reactor is shown in Figure 4.2.4 for the
China UHV pilot project, which dose not consider the transformer saturation and the TOV limitation of the arresters
in the TOV calculation. The TOV can reach more than 2.0p.u at zero compensation degree. Since the
compensation degree of shunt reactor is 87.8% for the China UHV pilot project, the maximum TOV occur less than
1.4 p.u. at the line side and 1.3 p.u. at the bus side. The relay protection to trip three phase CBs of both end of a
line synchronous is adopted by communication to shorten the duration time of TOV for reason of suppression of the
arrester absorbed energy. The duration time of TOV is 0.2 second as maximum delay time of trip system and less
than 0.5 second in which the CB is failure and backup CB operate. The TOV duration for surge arrester
withstanding 1.4 p.u. is 10 seconds according to the test data. The energy absorption allowable value 40 MJ is
decided.
Applying controllable shunt reactor is the best solution to coordinate the TOV limitation and reactive balance of the
UHV line. The controllable shunt reactor can automatically adjust the compensation degree according to the load
change, e.g. controllable shunt reactor operates at lower compensation degree when heavy load is transmitted and
quickly adjusts to maximum compensation degree to limit the TOV when the heavy load is rejected.
There are two types of controllable shunt reactors in service in China. One is the magnetically controlled type. The
output inductive capacity of the reactor can be smoothly and continually be adjusted from 10% to 100% by ion-core
saturation, which is controlled by injecting dc current to the magnetizing winding of the reactor. Another is the
reactance stepping controlled type. The configuration is shown in Figure4.2.5. The output inductive capacity of the
reactor can be stepping adjusted at 25%, 50%, 75% and 100%. The thyristor is connected in parallel with circuit
breakers to provide a fast switching operation and reduce switching transient. Both types of controllable shunt
reactors have been operated for more than 8 yeas in 500kV system. The 750kV stepping controlled type shunt
reactors is put into operation in 2011year and it is going to be used for the UHV ac double circuit transmission line
in China.
Page 90
TOV
TypethenameofTB
Compensation degree
Figure 4.2.4 The relationship between TOV and compensation degree of the shunt
reactor for the China UHV pilot project (U0 is caused by load rejection, U1 is
caused by load rejection in combination with single phase ground)
Figure 4.2.5 the single phase configuration of the stepping controlled shunt
reactors
4.2.2.3 INDIAN EXPERIENCE
In the 1200 kV system of India, preliminary studies of TOV have been carried out with ground fault of single line
followed by three phase interruption at receiving end. The result showed a value of TOV as 1.33 p.u. with 10000
MVA short circuit level at the sending end and the representative value of TOV is decided as 1.4pu. The required
energy capability of the surge arrestor has been considered. Figure 4.2.6 shows sequence of events for calculation
of surge arrestor energy accumulation. The series of events start with single line-to-ground fault followed by
opening of local and remote end CB of the faulted phase. The single phase auto recloser becomes effective after a
dead-time of 1000 ms. However the reclosing becomes unsuccessful and the CB could not be opened due to stuck
CB condition. The capability of energy assumption for the surge arrester could be calculated as follows.
Total energy = 2 Long discharge (IEC Class-5) + TOV + Margins

= 2 5MJ +35MJ + 10 MJ = 55 MJ
Page 91
TypethenameofTB
Figure 4.2.6 Sequence of events for calculation of surge arrestor energy

accumulation
REFERENCES
[1] IEC 60071: Insulation co-ordination: Part 1: Definitions, Principles and Rules
[2] CIGR Technical Brochure 362 WG A3.22, Technical Requirements for Substation Equipment Exceeding 800
kV - Field experience and technical specifications of Substation equipment up to 1200 kV, December 2008, ISBN :
978-2 85873-049-0
[3] E. Zaima, T. Shindo and M. Ishii; System Aspects of 1100kV AC Transmission Technologies in Japan:
Solutions for Network Problems Specific to UHV AC Transmission System and Insulation Coordination, CIGRE
Paper, No.2-1-2, EC/CIGRE UHV Symposium, Beijing (2007)
[4] CIGR Technical Brochure 456 WG A3.22, Background of Technical Specifications for Substation Equipment
exceeding 800 kV AC, April 2011, ISBN: 978-2-85873-145-9
[5] H.Ito, A.L.J.Janssen, D.Dufournet, Y.Filion and D.Peelo, Technical Requirements for Substation Equipment
exceeding 800kV, CIGRE session, No.A3-211, Paris, 2008
[6] H.Ito, A.L.J.Janssen, D.Dufournet, C.V.Merwe, D.Peelo, U.Riechert, Y.Yamagata, Y.Filion, M.C.Bhatnagar,
P.Fernandez and S.Wang, System Impacts on UHV Substation Equipment, IEC-CIGRE UHV symposium, No.3-1,
New Delhi, 2009
Page 92
TypethenameofTB
4.3 Switching overvoltages caused by closing and opening with ground fault
overvoltage
4.3.1 GENERAL
UHV system demands more insulation clearance than current EHV systems, especially in case of clearance of
switching impulse voltage in air under rain condition. Since equipments sizes may be larger and total costs higher
according to the increasing insulation level, accurate estimation of overvoltage is needed. SIWVs are evaluated
based on amplitude of switching overvoltages (SOV) derived by calculation analysis, which are influenced on
decision of basic insulation design, fundamental allowance levels and selection of system components such as
overvoltage suppression equipments and protective devices. Electro-Magnetic Transient Program (EMTP) analysis
is generally used to evaluate amplitudes of overvoltages and shape of waveforms occurring in actual and planed
transmission circuits. (Table 4.3.1 and Table 4.3.2)
SOVs arise from closing, reclosing and opening operation in transmission line and at ground fault. Their circuits are
considered as representative circuits for transition analysis. The amplitudes of the overvoltages tend to depend on
the length of energized line, number of connections and length of other lines which are energized/reenergized, type
of circuit breaker (closing/opening resistor or not), arrangement and specification of arrestors and so on [1].
The occurrence probability of SOV is one of the factors to evaluate representative value. It is considered that the
probability can be approximated by Gaussian distribution. IEC 60071-2 express the representative value of
overvoltage that is characterized by its 2 % value of statistics distribution using lots of calculation results. Some
users evaluate maximum value as truncation voltage.
Representative switching overvoltage of UHV systems and 800 kV systems in each country would be suppressed
to become practical levels by application of CB with closing and /or opening resistor and effective specifications
and optimum placement of MOSAs. [4-5] Finally, SIWVs are decided from representative overvoltage and some
users added a safety margin.
4.3.2 CALCULATION CONDITIONS AND METHOD FOR SWITCHING OVERVOLTAGES
4.3.2.1 CLOSING OVERVOLTAGE

Figure 4.3.1 shows an example of line energization circuit with closing operation [2, 14]. The SOV with one line
energization in two lines re-energization tends to become higher than one in case of only one line re-energization.
Figure 4.3.2 shows switching overvoltage of a 1100 kV transmission line in Japan that is calculated using EMTP
with Figure 4.3.1 circuit. The double-circuit line from Higashi-Gunma to Minami Iwaki is 200 km long without shunt
reactors and the circuit breakers are installed with closing resistors of circuit breaker is 700 . Surge arrestors are
adopted in OH entrances of both substations. Maximum overvoltage values to earth are 1.7 p.u. and almost 2.8 p.u.
for phase-to-phase. Surge arresters are effective to reduce SOV for phase to earth. However they are not so
effective for phase to phase.
Figure 4.3.1 Line energization circuit with closing operation
Page 93
TypethenameofTB
(a) Phase to earth
(b) Phase to phase
Figure 4.3.2 Example of analysis results of switching overvoltage in 1100 kV

transmission line of Japan
Figure 4.3.3 shows the comparison of maximum values of SOV in two typical 800 kV transmission systems of
Canada and South Korea. [5] The simulations have been performed under following scenarios: a) without
mitigation, b) with a circuit breaker with closing resistor, c) with a MOSA at the line ends and, d) with MOSA. In
case of the Canadian transmission system, a closing resistor of 300 was applied and for the South Korean
system a closing resistor of 400 was used. According to the simulation results shown in the Figure 4.3.3, the
closing resistor can reduce the maximum value of the SOV by 36 %. MOSAs could reduce the SOVs peak by 14 %
to 28 % depending on its position and system components. A combination of the closing resistor and the MOSAs
would lead to a further decrease of the SOVs maximal amplitude.
Figure 4.3.4 shows the influence of the closing resistor on the overvoltage in the 1100 kV system of China pilot
project [5]. The values in the figure are 2 % probability values. Statistical SOV has minimum value from 200 to
500 of closing resistor and is gradually increased over 600 . The SOV is composed of power frequency steady
component and infinite high frequency harmonic components, which are quickly attenuated and superposed on
each others, due to the distribution parameters characteristics of the long line. Since the shunt reactors
compensate the capacitance of the long line, the power frequency steady component voltage is reduced; therefore
the SOV superposed on steady component is reduced too. Figure 4.3.5 shows the comparison between 2% SOV
distribution along Jindongnan-Nayan UHV ac line with shunt reactors and without reactors in the case of energizing
no load line at Jindongnan substation. The combination of application of the 600 closing resistor and the shunt
reactor can suppress the maximum 2% SOV to 1.49p.u. However, without shunt reactors, using the 600 closing
resistor alone, the maximum 2% SOV reaches 1.95p.u and the discharge current of the line arresters excess the
coordinate current 2kA related to its SIPL.
Figure 4.3.6 shows reclosing overvoltage with different resistance in 800 kV system of South Korea. The
overvoltage values calculated by Transient Network Analyzer (TNA) relatively correspond to the EMTP values.
SOVs at phase to earth and phase to phase are varied depending on the value of pre-insertion resistor in case of
reclosing condition.
Table 4.3.1 and Table 4.3.2 show the condition of switching overvoltage studies for 800 kV and UHV system. 800
kV systems adopt the closing resistor in the range from 325 to 1000 and UHV systems apply 600 or 700
resistor.
Page 94
TypethenameofTB
Highest Voltage
Country/Utility
800 kV
765 kV
South Korea / KEPCO Canada/Hydro Quebec
Suppression measures for Closing resistor 1000 Closing resistor 325

switching overvoltage
and line surge arrester and line surge arrester
800 kV
Brazil / Furnas
- Opening resistor
1000-4000
- Closing resistor
400-600
- MOSA
- CB with controlled
switching device
Installed
800 kV
The United States /
AEP
Closing resistor 350690 or line surge

arrester
Shunt reactor
Not installed
N/A
Reclosing scheme
Multi-phase reclosing
Fast singlephase
reclosing
TPAR scheme
Switching overvoltage(kV)
and occurrence case
(Transmission line)
1241 (1.9 p.u.)

(Maximum)
Ground fault Clearing
N/A
1372 (2.1 p.u.)

(Maximum)
1306 (2 p.u.)
(Maximum)
EMTP analysis
Simplified method
based on protection
level V2kA of surge
arrester
Gilman/Whitehead
model and EMTP
analysis
EMTP or PSCAD
Analysis
Tr.
1500
1520
1550
1700
Other
1425
1425
1550
1425
Tr.
N/A
N/A
N/A
N/A
Other
N/A
N/A
N/A
N/A
800 kV
The United States /
AEP
800 kV
800 kV
Venezuela/
EDELCA
800 kV
Calculation method
SIWV (kV)
(phase-to
ground)
SIWV (kV)
(phase-to
phase)
Highest Voltage
Country/Utility
India/Power Grid
Installed for longer lines

singlephase or three
phase reclosing in 0.5
sec.
South Africa
- Closing resistor
Closing resistor 450
- 37070 Ohm inserted Closing resistor 450
for 9 ms and line surge
for 8.3 ms and line
and line surge arrester
arrester
surge arrester
Suppression measures for

switching overvoltage
Closing resistor 350690 or line surge

arrester
Shunt reactor
Installed for longer lines
Reclosing scheme
Singlephase or three
phase reclosing in 0.5
sec.
Switching overvoltage(kV)
and occurrence case
(Transmission line)
1306 (2 p.u.)
(Maximum)
Calculation method
EMTP or PSCAD
Analysis
EMTP Analysis
ATP/EMTP analysis
EMTDC/PSCAD
Tr.
1700
1550
1550
1425
Other
1425
1550
1425+438
(Open CB)
1550/1425
Tr.
N/A
2480
N/A
N/A
Other
N/A
2480
N/A
N/A
SIWV (kV)
(phase-to
ground)
SIWV (kV)
(phase-to
phase)
Installed
(depends on line
Installed
length)
Not in service but has
singlephase or three
facilities to
Singlephase or threephase auto reclosing to single or three phase phase autoreclosing
auto reclosing
1090 (1.67 p.u.)
1306 (2 p.u.)
1241 (1.9 p.u., 2%
(2 % Value)
(2 % value)
value)
Ground fault clearing
Installed
Table 4.3.1 Condition of switching overvoltage studies for 800 kV systems
Page 95
TypethenameofTB
Highest Voltage
1200 kV
Country/Utility
Russia/
Suppression measures
for switching
overvoltage
Closing resistor
and line surge
arrester
Closing resistor
600 and line
surge arrester
Closing resistor
600 and line
surge arrester
Shunt reactor
Installed
Installed
Reclosing scheme
Fast single
phase
reclosing
Fast single
phase
reclosing
Installed
Fast single
phase
reclosing within
1 sec.
Closing/Openin
g resistor 700
and line
surge arrester
Not installed
Fast single
phase
reclosing within
1 sec.
1570(1.6 p.u.)
(Maximum)
1666 (1.7 p.u.)

(2 % Value)
Closing
1473 (1.64
p.u.)
(2 % value)
Closing
1437-1527
(1.6-1.7 p.u.)
(Maximum)
Ground fault
1715
(ph.- ground
2 % value)
2743(ph.- ph.
2 % value)
Special
programme
Simplified
method based
on protection
level V2kA of
surge arrester
EMTP analysis
TNA (Transient
Network
Analyzer)
Simulation
Tr.
1800
1800
EMTP or
Simplified
method based
on protection
level V2kA of
surge arrester
1800
1425
1800
Other
1800
1800
1800
1550
1675
Tr.
N/A
2970
N/A
N/A
N/A
Other
N/A
2970
N/A
N/A
N/A
Switching
overvoltage(kV) and
occurrence case
(Transmission line)
Calculation method
SIWV (kV)
(phase-to
ground)
SIWV (kV)
(phase-to
phase)
1200 kV
India/Power
Grid
1100 kV
1100 kV
1050 kV
China/SGCC
Japan/TEPCO
Italy/CESI
Closing resistor
500 and gap
surge arrester
Installed
Fast single
phase
reclosing *1
*1 Three phase reclosing was studied but not intended to be applied
Table 4.3.2 Condition of switching overvoltage studies for 1100 kV and above
Maximum value of SOV (p.u.)
systems
Canada 800kV
3
2.5
2.5
South Korea 800kV
2.2
1.8
1.6
1.8
2.04
1.9
1.4
1.5
1
0.5
0
No mitigation
With closing
resister
With MOSA
With MOSA
at the line ends at the middle of
the line
Figure 4.3.3 Reduction of SOV by closing resistor or MOSA in 800 kV systems of

Canada and South Korea
Page 96
TypethenameofTB
Figure 4.3.4 Typical closing overvoltage with different closing resistances in 1100
kV system of China
with reactors
without reactors
2% Overvoltage(p.u)
2.2
2
1.8
1.6
1.4
1.2
1
0
Jindongnan
20
40
60
% of line length
80
100
Nanyan
Figure 4.3.5 The comparison between 2% SOV distribution along Jindongnan-Nayan

UHV ac line with shunt reactors and without reactors in the case of energizing no
load line at Jindongnan substation
Figure 4.3.6 Reclosing overvoltage in 800 kV system of South Korea
Page 97
TypethenameofTB
4.3.2.2 OPENING OVERVOLTAGES

After fault current is interrupted, slow front overvoltages are generated at healthy circuits. Figure 4.3.7 shows
example of waveforms of overvoltage analysis in the 1100 kV system in Japan. The overvoltage depends on
several conditions, such as fault types, fault locations and faulty phases. Three three-phase line-to-ground faults is
higher, because the recovery voltage is high. Figure 4.3.8 shows opening overvoltage distribution. Maximum SOV
is generated in the middle of the line of healthy side for three-phase line-to-ground faults and three-phase line
breaker opening. The maximum value of the SOV depends on opening resistor value.
(a) Without resistor
(b) With resistor 500ohm
Figure 4.3.7 Waveforms of overvoltage analysis in 1100 kV system of Japan
Figure 4.3.8 Opening overvoltage distribution in 1100 kV system of Japan
Page 98
TypethenameofTB
4.3.2.3 G ROUND
FAULT OVERVOLTAGES
When a ground fault occurs, the phase voltage changes to zero. This phenomenon seems equivalent to inject the
surge waveform with reverse polarity voltage at the phase. The injected surge waveform induced to other healthy
phases of OH lines and multiple reflections generated in the lines are superimposed on AC waveform. Figure 4.3.9
shows analyzed waveform of maximum SFO at ground fault and Figure 4.3.10 shows ground fault overvoltage
distribution in a 1100 kV system of Japan [2, 4, 14]. Ground fault occurred at phase c and the maximum SOV of
1.62 p.u. has been induced in phase b in the middle of OH line. The overvoltage near substations can be reduced
by MOSA. In this case, the ground fault SOV gives the lower limit to determine the suppression level in the resistor
closing / opening of the circuit breaker.
Figure 4.3.9 Analyzed waveform of
Figure 4.3.10 Analyzed waveform
maximum SFO at the ground fault
of maximum SFO at the ground
in 1100 kV system of Japan
fault in 1100 kV system of Japan
Figure 4.3.11 shows example of analysis results of closing overvoltage in the 1100 kV system of China [3]
Jindongnen-Nanyang-Jingmen 1100 kV transmission line clearing ground faults leads to SOV. The maximum 2 %
voltage is 1.79 p.u. without resistor and 1.54 p.u. with 700 opening resistor for three-phase line-to-ground faults
and three-phase line breaker opening.
Page 99
TypethenameofTB
Jindongnan
Nanyang
Jingmen
(a) The circuit for overvoltage analysis when clearing line fault
(b) Example of analysis results
Figure 4.3.11 Example of analysis results of ground fault overvoltage in 1100 kV

system of China
Application of surge arrestor and study of its appropriate allocation are the basic idea in UHV system design.
Figure 4.3.10 and Figure 4.3.11 suggest that installation of surge arrestors have effect on the suppression of the
SOVs at the substations. Table 4.3.3 shows analysis result of maximum SOV at the ground fault clearing using
model of Figure 4.3.12. Results of the Table 4.3.3 suggests that the maximum SOVs at substations are lower than
on OH lines in the all cases and the protection level of surge arrestor is an important factor in reducing the SOV to
adequate level.
Page 100
TypethenameofTB
Protection level of
Surge arrestor
Maximum SOV at OHlines
Maximum SOV at three
substations under the
above condition
Without R
500
1280 kV
(1.42 p.u.)
1280 kV
(1.42 p.u.)
1210 kV
(1.35 p.u.)
1250 kV
(1.39 p.u.)
1640 kV
(1.83 p.u.)
1770 kV
(1.97 p.u.)
1280 kV
(1.42 p.u.)
1390 kV
(1.55 p.u.)
V20kA =1620 kV
V20kA =1800 kV
V20kA =1620 kV
V20kA =1800 kV
With Resistor
750
1360 kV
(1.52 p.u.)
1380 kV
(1.54 p.u.)
1230 kV
(1.37 p.u.)
1290 kV
(1.44 p.u.)
1000
1440 kV
(1.60 p.u.)
1470 kV
(1.64 p.u.)
1250 kV
(1.39 p.u.)
1310 kV
(1.46 p.u.)
Table 4.3.3 Maximum SOV due to the ground fault clearing
140km
223mH
U9
230km
140mH
U7
127mH
U6
Figure 4.3.12 Analyzed system model for the opening overvoltage for the
Jindongnan-Nanyang-Jingmen 1100 kV transmission line in China
Page 101
TypethenameofTB
4.3.3 REPRESENTATIVE OVERVOLTAGES

Table 4.3.4 and Table 4.3.5 show the representative overvoltages value and SIPL of surge arrestor protection
levels. The representative overvoltages, ranging from 1093 kV to 1318 kV in the 800 kV systems, are different from
thise of the UHV systems, which range from 1309 kV to 1570 kV. The ratios of the representative overvoltage and
SIPL are from 0.9 to 1.1 in 800 kV systems and less than 1 p.u. in UHV systems. These ratios are closed to 1.
Highest Voltage
Country/Utility
Representative
Value(kV)
Tr. /
reactor
765 kV
Canada/
Hydro
Quebec
800 kV
Brazil /
Furnas
800 kV
The United
States /
AEP
1306
(2.0 p.u.)
(Maximum)
1306
(2.0 p.u.)
(Maximum)
800 kV
800 kV
India/
Power Grid
Venezuela/
EDELCA
1093
(1.67 p.u.)
(2 % Value)
1093
(1.67 p.u.)
(2 % Value)
1318
(2.02 p.u.)
(Maximum)
1318+438
(Maximum)
(Open CB)
1140
(1.75 p.u.)
N/A
1140
(1.75p.u.)
N/A
1200
1140
1321
(at 3kA)
1200
1220
1197
Tr.
0.98
1.00
N/A
1.09
0.90
1.10
Other
0.98
1.00
N/A
1.09
0.90
1.10
Other
SIPL (kV)
at 2 kA
Representative
overvoltage /
SIPL
800 kV
South
Korea /
KEPCO
1176
(1.8 p.u.)
(Maximum)
1176
(1.8 p.u.)
(Maximum)
SIPL: Switching Impulse Protection Level of surge arrester
Table 4.3.4 Representative overvoltages and protection level of surge arresters in

800 kV system
Highest Voltage
1200 kV
Country/Utility
Russia/
1100 kV
1100 kV
1050 kV
China/SGCC
Japan/TEPCO
Italy/CESI
1450
(1.69p.u.) *2
(Maximum)
1450
(1.69 p.u.) *2
(Maximum)
1450 (at 3kA,
30/60us)
1570
(1.6 p.u.)
(Maximum)
1570
(1.6 p.u.)
(Maximum)
1570
(at 2.8kA)
1500
(1.53 p.u.)
(2 % Value)
1500
(1.53 p.u.)
(2 % Value)
1392
(1.55 p.u.)
(2 % value)
1437
(1.60p.u.)
(2 % value)
1309
(1.46 p.u.)
(Maximum)
1400 (1.56
p.u.)
(Maximum)
1500
1460
(1440) *1
Tr.
1.00
1.00
1.00
0.91
1.00
Other
1.00
1.00
1.00
0.97
1.00
Tr.
Representative
Value(kV)
Other
SIPL (kV)
at 2 kA
Representative
overvoltage /
SIPL
1200 kV
India/Power
Grid
SIPL: Switching Impulse Protection Level of surge arrester

*1 SIPL is not officially specified at present.
*2 with gap arresters.

1100 kV and above system
Page 102
TypethenameofTB
IEC 60071-2 proposes that the assumed maximum value of the representative switching overvoltage is equal to
the truncation value of the overvoltages or equal to the switching impulse protective level of the surge arrester,
whichever is the lower value. The truncation voltage at fault and fault clearing conditions can be estimated by
conservative method but the value of representative voltage is higher in comparison with maximum voltage that
occurred in actual power systems. In the meantime, since overvoltages calculated by computer simulations has
been recognized as accurate and credible, value, the calculated value is generally applied to representative value
without margin of error for simulation results. There are many results of simulations of actual and planned
transmission networks of above 800 kV systems, which are reported by in CIGRE A3.22WG Technical Brochure
362 and Technical Brochure 456 [4-5].
Some utilities have evaluated maximum value/2 % value calculated by simulations as representative overvoltage
and determined on SIPL which is in accordance with their policies. In India and some countries, the protection
levels of surge arrestors are selected on the basis of MCOV and energy capability due to TOV and SOV. As a
result, the representative value is closed to SIPL, which is consistent with the proposal of IEC 60071-2.
It may seem that some margins are not applied to the representative overvoltage value. However, in some cases,
the representative overvoltage is evaluated as the value with predominately low risk to failure. In case of South
Korea, since the most overvoltages will be much reduced with pre-insertion resistor of the 800 kV GCB and
maximum overvoltage is assumed for closing resistor insertion failure, which is quite rare, it is considered that a
higher margin is not necessary to the representative value.
As for other factors that include influence of small variations of the insulation strength and arrestor protective level,
some margin is considered as co-ordination factor to multiply the values of the representative overvoltages in some
systems.
4.3.4 SHAPES OF SWITCHING OVERVOLTAGES IN UHV SYSTEMS

Since the length of transmission lines in UHV systems is longer than that in EHV systems, the front time-to-peak of
switching surge in UHV systems tends to become long. It is important to figure out practical value of the time-topeak to consideration in the insulation design of transmission line. The air clearance strongly depends on the
insulation characteristic of the positive slow front overvoltage and its wave shape. In case of long air gap with rodplane, critical time-to-peak that cause minimum flashover voltage on the U curve is more than that of standard
switching impulse of 250 s.
Figure 4.3.13 shows the calculation result of switching overvoltage on the 1100 kV UHV GIS of Japan. The time-topeak of the switching surge waveform is 797 s. The occurrence situation is line energization with closing operation
shown in Figure 4.3.1. The closing resistor of CB is 700 and length of energized transmission line from is 200 km.
Page 103
TypethenameofTB
Figure 4.3.13 Calculation result of a switching overvoltage with a timeto-peak of

797s on the 1100kV UHV GIS of Japan
The time-to-peak of the waveform in the figure is calculated in accordance with the following theory.
The most important part of the front of a switching impulse is usually considered to be the upper 15 %. The rest of
the shape of the front would affect neither the 50 % breakdown voltage nor the standard deviation of the gap [6]. A
reasonable approximation for the upper 15 % by double exponential waveform is suggested as shown Figure
4.3.14.
The waveform V(t) can be represented as double exponential waveform by the following equation.
V (t ) k (e at e bt )
(1)
When b is equal to 50a, the ratio of the wave tail to the wave front is 10, and this wave form is often used for
switching impulse flashover test. If it is assumed that b= 50a, then
Vmax V (Tf ) k (e aT e 50aT )

f
(2)
0.85Vmax V (T0.85 ) k (e aT
0.85
e 50aT
0.85
(3)
dV (Tf ) / dt 0
(4)
From the equations given above, then the time-to-peak is determined by the following equation [7].
Tf 1.678(Tf T0.85 )
(5)
Page 104
TypethenameofTB
Figure 4.3.14 Reasonable approximation for upper 15 % by double exponential

waveform
The front part of the waveform shown in Figure 4.3.14 includes a negative spike-like-pulse that resembles a fastfront waveform such as a lightning impulse and has a lower peak voltage than SOV. The spike-like-pulse can be
neglected in the above calculation of the SOV front time, because the fast front waveform with a low voltage has no
influence on the dielectric performance of switching surge in accordance with the V-t characteristic of air insulation.
Page 105
TypethenameofTB
In China, some methods to define the front time Tf of slow front waveform are under consideration. The following 6
definitions of the front time are studied and tried to adopt analysis results of slow-front waveform that is occurring in
the Chinese UHV systems. Figure 4.3.15 shows the front time definitions.
Definitions of the front time

Definition 1; Tf = Tp Tp is the time interval between the zero-crossing point and the peak value U p of the
waveform. The definition is used in a time-to-peak evaluation of switching-impulse waveform for
withstand test [8-9].
Definition 2; Tf K (T90% T30% ) , where T90% is the time when the SOV reaches 90% voltage of U p and T30% is the
time when SOV reaches 30% voltage of U p . It can be derived by tracing a straight line through the
90% point of U p and 30% point of U p on the waveform. The straight line would intersect at points M
and N respectively with horizontal time-axis at zero voltage and horizontal line which is through the
peak value. Tf is the projection of line segment MN on the horizontal time-axis. K=1.67. The definition
is used in a front-time evaluation of lightning-impulse waveform for withstand test [9].
Definition 3; Tf K(Tp T85%) shown in equation (5) is defined by TEPCO. T85% is the time when the switching impulse
is equal to 85% of U p . K=1.678 [7].
Definition 4; Tf K(Tp T70%) , T70% is the rise time when the switching impulse is equal to 70% of U p and K=1.45 [10].
The definition is decided in accordance with a switching impulse flashover theory with having a model
where the inception and development of leader discharge occur between 60 - 70% value and 100%
value of the crest value corresponding to the 50% provability flashover voltage.
Definition 5; Tracing a tangent at the 85 % voltage of U p in the front wave shape, the line would intersect at points
M and N respectively with horizontal time-axis at zero voltage and horizontal line which is through the
peak value. Tf is the projection of line segment on the horizontal time-axis. Tf can be expressed as
follow equation. Tf 1.18 T 85 .
T 85 is the time interval between M and the instant when the voltage reaches its 85% of the peak
value [6].
Definition 6; Based on the actual waveform, the front time is determined by curve fitting with digital calculation. One
method is to adopt the double exponential fitting [9]. There are other methods such as modeling with
Steiglitz-Mcbride coefficients and low pass filtering [11].
Page 106
TypethenameofTB
1.5
1.6
*106
1.2
*106
Tf
1.0
Definition 1
Definition 2
0.8
0.5
0.4
T30%
T90%
0.0
0.0
-0.4
-0.5
-0.8
-1.0
-1.2
Tf
-1.6
-1.5
5
10
15
20
25
30
1.6
*106
1.2
35
[ms]
40
10
15
20
25
30
35
[ms]
40
[ms]
40
1.6
*106
1.2
Definition 3
Definition 4
0.8
0.8
T85%
0.4
Tp
0.0
0.0
-0.4
-0.4
-0.8
-1.2
-1.6
-1.6
10
15
20
25
30
Tp
-0.8
T f 1.678(Tp T85% )
-1.2
5
T70%
0.4
35
[ms]
40
T f 1.45(Tp T70% )
5
10
15
20
25
30
35
x 105
1.6
*106
1.2
U
vs. t
fit 1
14
Definition 5
Tangent
12
Definition 6
0.8
10
0.4
T85%
8
U(v) 6
0.0
-0.4
T85
-0.8
T f 1.18(T 85)
-1.2
-2
-4
-6
-1.6
5
10
15
20
25
30
35
[ms]
40
t(s)
8
x 10-3
Figure 4.3.15 Front time definitions

Figure 4.3.16 shows the single circuit of the Chinese UHV pilot project. Table 4.3.6 shows analysis results of the
SOVs on the fault line or its adjacent lines caused by single line to-ground fault and clearing by single-phase
tripping at both line ends in the circuit of Figure 4.3.16. Figure 4.3.17 shows the SOV of energizing NanyangJingmen unloaded line at Nangyang side and the front time was calculated by the method of definition 5.
Figure 4.3.17 shows the UHV double-circuit transmission system for the Huainan Shanghai lines. Table 4.3.6
shows analysis results of the SOVs of double-circuit transmission lines caused by line energization or single line-toground fault and fault clearing on the Huainan Shanghai transmission system.
500kV grid of North China
500kV grid of Central China

Jindongnan thermal power plant
358.7km
281.3km
Jinmen substation
Nanyang switchin station
Jindongnan substation
960MVAR
720MVAR
720MVAR
600MVAR
Figure 4.3.16 Single circuit of Chinese UHV pilot project
Page 107
TypethenameofTB
Operation
Line name
JingdongnanNanyang
Closing point
Jingdongnan
1.42
50%
Nanyang
1.58
40%
Nanyang
1.60
60%
Jingmen
1.42
90%
at Nanyang side of
Nanyang-Jingmen
line
1.65
60%
at Nanyang side of
JingdongnanNanyang line
1.55
40%
Line energizing
_
NanyangJingmen
Single-phase
ground fault and
clearing
Single phase
ground Fault point
Place where
maximum
2% SOV
overvoltage
along line
occurs
*1
(p.u.)
(% of total line
length)
JingdongnanNanyang
_
NanyangJingmen
Definition
front time /
time-to-half
ms
1
2
4
1
2
1
2
3
4
4*2
5
1
2
1
2
3
4
4*2
5
1
2.9 / 5.6
1.3 / 4.0
1.95 / 4.2
4.4 / 6.4
1.5 / 6.73
3.83 / 6.73
4.23 / 7.13
3.07 / 5.97
2.95 / 5.85
1.76 / 5.3
6.14/ / 9.04
3.56 / 5.76
1.11 / 3.31
2.46 / 5.46
2.35 / 5.35
1.0 / 4.0
1.46 / 4.46
1.14 / 4.5
3.7 / 6.7
2.88 / 5.18
2.51 / 4.81
*1: 1 p.u. = 1100kV x 2 / 3

*2: The additional definition of Definition 4. Tf K(Tp T70%) , K=1.86
Table 4.3.6 Front-times, time-to-half and amplitudes of SOVs of UHV singlecircuit transmission lines caused by line energizing or Single-ground fault and
clearing
Page 108
TypethenameofTB
1 .5
[M V ]
1 .0
0 .5
0 .0
- 0 .5
- 1 .0
- 1 .5
0 .0 0
0 .0 2
0 .0 4
0 .0 6
0 .0 8
[s ]
0 .1 0
0 .6 0
*10 6
0 .1 4
T85
- 0 .3 2
- 0 .7 8
Tangent
- 1 .2 4
- 1 .7 0
2 1 .0
2 2 .5
2 4 .0
2 5 .5
2 7 .0
2 8 .5
[m s ]
3 0 .0
Figure 4.3.17 Calculation result of switching overvoltage on the 1100kV UHV GIS of
China
Figure 4.3.18 The UHV double circuits transmission lines in Huainan Shanghai
system
Page 109
TypethenameofTB
Operation
Line
energizing
Single-phase
ground fault
and clearing
Line name
Huainan
Wannan
(with outage
of one circuit)
Closing point
Single phase
ground Fault
point
Huainan
Place where
maximum
2% SOV
front time /
overvoltage
Definition time-to-half
along line
occurs
(p.u.)*1
(ms)
(% of total
line length)
1.43
100%
1.26
100%
1.62
42% of
Huainan
Wannan line
_
Wannan
Zhebei
(with outage
of one circuit)
Huainan
Wannan
Wannan
at Wannan side
of Wannan
Zhebei line (with
outage of one
circuit)
1
2
3
4
4*2
1
2
3
4
4*2
1
2
3
4
4*2
4.59 / 5.79
3.33 / 5.79
1.84 / 5.79
2.44 / 5.79
2.03 / 5.79
4.01 / 7.25
3.61 / 7.25
0.81 / 4.25
0.76 / 7.25
0.90 / 7.25
4.08 / 5.93
3.68 / 5.93
1.71 / 5.93
2.35 / 5.93
1.90 / 5.93
*1: 1 p.u. = 1100kV x 2 / 3

*2: The additional definition of Definition 4. Tf K (Tp T70% ) , K=1.86
Table 4.3.7 Front-times, time-to-half and amplitudes of SOVs of UHV doublecircuit transmission lines caused by line energizing or Single-ground fault and
clearing
The maximum level of SOV is 1.65 p.u., which is less than the SOV insulation design level of 1.7 p.u. The results
correspond with those reported in CIGRE A3.22 for technical specifications of UHV systems.
The results of Table 4.3.6 and 4.3.7 suggest that the front times of switching overvoltages range from 0.76ms to
6.14ms. The front time of SOV on the 1100 kV systems of Japan and China tends to be longer than that of EHV
systems, because of long length of transmission lines. The definition of SOV front time with non-standard shape is
under consideration. It is necessary to clear the relationship between flashover voltage and front shape of nonstandard waveform regarding SOV. Further information on the definition of SOV front time will be provided by
CIGRE WG D1.
Page 110
TypethenameofTB
4.3.5 SWITCHING OVERVOLTAGES LEVELS OF TRANSMISSION LINES

As the system voltage of transmission line become higher, the necessary insulation distance in air tends to be
increased. Although the flashover voltages of various waveforms in long air gap are increased along the insulation
distance, the insulation strength of switching impulse, especially on positive polarity, has saturation characteristics
to the distance. Therefore, reducing the SOV is essential to achieve minimizing design and overall costs of UHV
system.
To realize the reduction of SOV, various mitigation methods are to be utilized. In UHV system of Japan, SIWV on
phase-to-earth is relatively lower. It is considered that circuit configuration, including length of transmission line,
application of CB with opening/closing resistor and installation of multiple surge arrestors at the substations, has
influence on reduction of the overvoltage. In particular, effect of a combination of mitigation technologies, such as
CB with opening/closing resistor and optimum configuration of surge arrestors, is reported [4-5, 12-13].
Figure 4.3.7 expresses suppressive effects of opening resistor and surge arrestor on the SOV in the transmission
lines. To investigate the fault type, the fault locations, CB with/without resistor and surge protection level the detail
analysis of SOV was carried out. The analyzed system model for the opening overvoltage is shown in Figure 4.3.12.
Figure 4.3.19 shows relationship between the locations of the fault and maximum SOV. The maximum SOV occurs
happened on the middle of the 230 km healthy line during three-phase opening operations after three phase line
faults to ground generated at location 3. In case of CB without resistor, the SOV is higher and the insulation design
of overhead (OH) line would become larger. Figure 4.3.20 shows maximum SOV for different fault types and
opening practice. [5] The figure suggests that SOV depends on CB with/without resistor and the fault type. The
maximum SOV for three-phase line faults and three-phase tripping is reduced from 1.83 to 1.42 by application of
the resistor. Therefore, the UHV system in Japan decided to employ the opening/closing resistor for line side CBs
to control the overvoltage level.
These technologies are effective to change the insulation design of transmission line. In case of conventional level
of overvoltage in systems such as the 550 kV system of Japan, the SVO is around 2 p.u. The SOV in UHV system
with above technology reduces to 1.6 - 1.7 p.u. Figure 4.3.21 illustrates a comparison between UHV transmission
tower based on existing 550 kV technologies and an actual UHV transmission tower in Japan. The height is
reduced from 143 m to 110 m [2, 4-5, 12-14].
Page 111
TypethenameofTB
Maximum overvoltage (pu)
2
1.8
1.6
1.76
1.69
1.83
1.62
1.42
1.4
1.22 1.22
1.2
1.66 1.69
Maximum value
The location of
maximum value generated
Without resistor
1.36
1.33
1.23 1.22
1.20
1.25
With resistor (500)

1
The location of the fault
3LG-3LO
1LG-3LO
Maximum overvoltage (p.u.)
Figure 4.3.19 Relationship between the locations of the fault and maximum SOV
Figure 4.3.20 Maximum SOV for different fault types and opening practice
Page 112
TypethenameofTB
1100 kV
Designed with
550 kV technology
Constructed tower
550 kV
S hin-tama line
(a) C omparison of transmiss ion tow er designs
(b) 1100 kV tow er
Figure 4.3.21 Tower design compaction in 1100 kV system of Japan

REFERENCES
[1] IEC 60071-2, Insulation Co-ordination Part 2: Application guide, 1996-12
[2] Eiichi Zaima, C.Neumann, Insulation Coordination for UHV AC Systems based on urge Arrester Application
(CIGRE C4.306), Second International Symposium on Standards for Ultra High Voltage Transmission, JICCG,
India symposium, 2009
[3] Gu Dingxie, Zhou Peihong, Xiu Muhong, Wang sen, Dai Min and Lou Ying, Study Overvoltage and Insulation
Coordination for 1000kV AC Transmission System, International conference of UHV Power Transmission
Technology, Beijing, 2006
978-2 85873-049-0
[5] CIGR Technical Brochure 456 WG A3.22, Background of Technical Specifications for Substation Equipment
exceeding 800 kV AC, April 2011, ISBN : 978-2-85873-145-9
[6] C.Menemenlis, K.Isaksson, The Front Shape of Switching Impulses and its Effect on Breakdown Parameters,
IEEE Transactions on Power Apparatus and Systems, volPAS-93, No.5, pp1380-1389, 1974
[7] T.Ono, H.Matsubara and T.Suzuki, Calculation of Risk of Failure of Switching Surge, Electrical Engineering
Laboratory, Report No. 182023, CRIEPI, 1982 (Japanese)
[8] IEC60071-1 2006 Insulation co-ordination-Part 1: Definitions Principles and ruls
[9] IEC60060-1 Ed.3.0: 2008 High-voltage test techniques Part 1: General definitions and test requirements
[10] K.J.Lloyd and L.E.Zaffanella, Insulation for switching surges in Transmission Line Reference Book 345kV
and above, J.J Laforrest, Ed. 2nd. Ed. Palo Alto CA: EPRI.1982.pp 503-544.
[11] Nunes, R.R. do Couto Boaventura, W. Dept. of Electr. Eng., Fed. Univ. of Minas Gerais (UFMG), Belo
Horizonte, Brazil, Insulation Coordination Considering the Switching Overvoltage WaveshapePart I:
Methodology, Power Delivery, IEEE Transactions on Oct. 2009
Page 113
TypethenameofTB
[12] H.Ito, A.L.J.Janssen, D.Dufournet, Y.Yamagata, U.Riechert, P. Fernandez, M. Kosakada and D.Peelo,
Background information and Study Results for the Specification of UHV Substation Equipment, SC A3 colloquium,
Vienna, 2011
[13] H.Ito, A.L.J.Janssen, D.Dufournet, Y.Yamagata, U.Riechert, P. Fernandez, M. Kosakada and D.Peelo,
Considerations and Recommendations for the Specification of UHV Substation Equipment, CIGRE symposium,
Bologna, 2011
[14] Eiichi Zaima and Hiroshi Okamoto, System Design and Insulation Coordination for UHV AC Transmission
System in TEPCO, International conference of UHV Power Transmission Technology, Beijing, 2006
Page 114
TypethenameofTB
4.4 Lightning overvoltages caused by back-flashover and direct lightning

4.4.1 GENERAL
For the insulation design of substation equipment, the predominant overvoltage is lightning overvoltage. LIWVs are
evaluated based on lightning overvoltages. As a calculation method of lightning overvoltages in the UHV and 800
kV system, most utilities have used analysis EMTP. The simplified method based on the protected level of a surge
arrester is at present applied in 1200 kV system in India. In this section, the calculation due to an analysis program
such as EMTP is mainly described.
The lightning overvoltages occurring in a substation are due to the back flashover at the transmission tower or the
direct lightning strokes to the power line, illustrated in Figure 4.4.1. When the back flashover at the transmission
tower or the direct lightning stroke to the power line occurs, a lightning surge will propagate on the transmission line
and get into substation. While the lightning overvoltages in a substation can be suppressed by arranging a surge
arrester effectively, they depend greatly on the assumed lightning stroke conditions and the configuration of a
substation circuit. In the case where the peak value of a lightning stroke current is large and the duration of
wavefront is short (the steepness of a wavefront being large), or when the branches of circuits are few and the
distance from the surge arrester becomes far, lightning overvoltages will be generally high.
It seems that LIWVs determined under severe conditions such as a large lightning current and/or other unusual
circuit conditions include a margin for the overvoltage generated under normal conditions. Overvoltage calculation
conditions are an important aspect of the utility's policy on insulation coordination.
Direct lightning stroke to phase line
Lightning stroke to tower

Back flashover
Lightning surge
Entrance of substation
Transmission lines
Transmission tower
Figure 4.4.1 Explanation of the invading lightning surge to substation

4.4.2 CALCULATION CONDITIONS AND METHOD FOR LIGHTNING OVERVOLTAGES
Tables 4.4.1 and 4.4.2 show the calculation method and conditions for lightning overvoltage studies of each
country.
Page 115
TypethenameofTB
Highest Voltage
Country/Utility
Calculation method
Lightning current
Amplitude (kA)
Lightning current
waveshape
Backflashover
Lightning
stroke point
Substation
circuit
Lightning current
amplitude (kA)
Lightning current
waveshape
Direct
lightning
Lightning
stroke point
Substation
circuit
Highest Voltage
Country/Utility
Calculation method
Lightning current
Amplitude (kA)
Lightning current
waveshape
Backflashover
Lightning
stroke point
Direct
lightning
Substation
circuit
Lightning current
amplitude (kA)
Lightning current
waveshape
Lightning
stroke point
Substation
circuit
800 kV
Brazil/Furnas
Gilman/Whitehead model and EMTP
800 kV
India/Power Grid
EMTP
No data available
200
No data available
Ramp shape (4.2 s/50 s)
Considered in the performed

simulations several stroke points
(near the line bays at the substations)
Considered in the performed
simulations several stroke points
1 & CB, 3 dia, 3 transformers, 2

lines with line reactors, 1 bus reactor
No data available
10
No data available
Ramp shape
(1.0 s/5 s)
Nule exposition for strokes above the

critical value considered in the
planning studies
Nule exposition for strokes above the
critical value considered in the
planning studies
1st tower
1st tower
1 & CB, 3 dia, 3 transformers, 2
lines with line reactors, 1 bus reactor
800 kV
China/
SGCC
EMTP
800 kV
Korea/
KEPCO
EMTP
800 kV
ATP/EMTP
225
170
20 to 300
Ramp shape (2.6 s/50

s)
Ramp shape
(1 s/70 s)
1st and 2nd tower
1st tower
1 line and 1 transformer
One and half CB system
Ramp shape
(0.5 to 3 s/50 s)
One tower of first five
line spans; usually the
second tower from S/S
line entrance
One and half CB or
Double CB
16
Not considered
Ramp shape (2.6 s/50

s)
N/A
1st span
N/A
1 line and 1 transformer
N/A
Venezuela
1 to 30
Ramp shape
(0.5 to 1 s/50 s)
One phase of 1st tower
from S/S line entrance
One and half CB or
Doble CB
Table 4.4.1 Condition of lightning overvoltage studies for 800 kV system
Page 116
TypethenameofTB
Highest Voltage
Country/Utility
Calculation method
Backflashover
Lightning current
Amplitude (kA)
Under evaluation
Lightning current
waveshape
Under evaluation
Lightning
stroke point
Under evaluation
Substation
circuit
Direct
lightning
1200 kV
India/Power Grid
Simplified method
based on
protection level
V20kA of surge
arrester
Lightning current
amplitude (kA)
Lightning current
waveshape
Lightning
stroke point
Substation
circuit
1200 kV
Russia/
1100 kV
China/SGCC
1100 kV
Japan/TEPCO
Special
programme
EMTP
EMTP
200-300
200
Ramp shape (2.6

s/50 s)
Ramp shape
(1s/70s)
Every tower in 2
km near S/S
1st and 2nd tower
-1 line and CB
open
-1 line and 2
busbar and 1
transformer
-1 line and CB
open
-1 line and 1
busbar
-1 line and 1
busbar and 1
transformer
Statistic method
taking into
account
Statistic method
taking into
account
Dependence
waveshape from
value of lightning
surge
Under evaluation
N/A
Under evaluation
N/A
Under evaluation
N/A
Under evaluation
N/A
Under evaluation
N/A
Max. value gained

by EGM
Ramp shape (2.6
s/50 s)
Every span in 2
km near S/S
Ramp shape
(1s/70s)
-1 line and CB
open
-1 line and 2
busbar and 1
transformer
-1 line and CB
open
-1 line and 1
busbar
-1 line and 1
busbar and 1
transformer
30
1st tower
Table 4.4.2 Condition of lightning overvoltage studies for 1100 kV and above
system
Page 117
TypethenameofTB
4.4.2.1 BACK FLASHOVER AT A TRANSMISSION TOWER

The ramp shape has been widely used as the assumed lightning current waveshape by China, India/800 kV, Korea,
Venezuela and Japan, as shown in Figure 4.4.2. This shape is simple and easy to use. A peak current and the
front duration of shape are set to be constant values in China/800 kV, India/800 kV, Korea and Japan, and those
are considered as the parameter of 20-300 kA and 0.5-3.0 s in Venezuela and that of 200-300 kA for lightning
current in China/1100 kV. While 300 kA in Venezuela is the largest of the peak current among these countries, 200
kA/s in Japan is the steepest wavefront.
I
If
If
tf
tf
t [s]
Figure 4.4.2 Ramp shape of current waveform: tf is the front time, tt is the time to
half value and If is the crest value of current [1]
According to the observational result of lightning stroke current to the transmission tower, the front duration tends
to be longer with larger amplitude of lightning stroke current, as shown in Figure 4.4.3. From the viewpoint of the
characteristics of the front duration in Figure 4.4.3, 1.0 s of the front duration in Korea and Japan seems to be
very short for higher lightning currents over 100 kA.
Page 118
TypethenameofTB
[kA]
0
PEAK : -120.0kA
Tf : 6.4s
-50
-100
0
20
40
60
80
100
First Peak amplitude

(s)
Maximum rate-of-rise
Maximum rate-of-rise(kA/s)
Example of observed current waveform
102
10 1
10 1
CIGRE (y=3.9x0.55)[5]
y=1.27 x0.81(r=0.846)
95% confidence
interval of data
95% confidence interval of data
10 -1
101
10 2
First Peak amplitude(kA)
101
10 2
First peak amplitude(kA)
Figure 4.4.3 Relationship between lightning stroke current amplitude and rate-ofrise [2, 5]
The lightning stroke positions are commonly the 1st or the 2nd tower near the substation. Meanwhile, China
considers every tower within 2km from the substation as the lightning location for the 1100 kV system.
With regard to the substation circuit, when calculating overvoltages, the normal operating configuration of one and
half CB system is assumed in Korea, India/800 kV and Venezuela. On the other hand, Japan has applied a circuit
configuration where the maximum overvoltage can occur at a line entrance, a busbar, or a transformer
independently. China provides the circuit conditions similar to Japan regarding to line entrance and transformer for
1100 kV system. As for the 800 kV system in China, the circuit of 1 line and 1 transformer is used.
The statistic method has been used in Russia. In USA/AEP, Brazil and Italy detailed data is not available. The
evaluation is proceeding in India/1200 kV.
4.4.2.2 DIRECT LIGHTNING STROKES TO A POWER LINE

Since the direct lightning strokes to a power line occur due to shielding failure of the ground wire such as Figure
4.4.4, a direct lightning stroke current is comparatively small as shown in Figure 4.4.5, and the front duration is
short [3]. The ramp shape has been applied to the lightning current waveform by China, India/800 kV, Venezuela
and Japan. The largest current value among these countries except China/1100 kV is 30 kA in Venezuela and
Japan. China has evaluated the largest direct lightning current for the 1100 kV system by an electric geometry
model (EGM) of lightning shielding. The lightning stroke position and the substation circuit configuration are the
same as those in the back-flashover study. On the other side, a direct lightning stroke has not been taken into
consideration in Korea, and detailed data is not available in Russia, Brazil USA/AEP and Italy. It is reported that the
resulting overvoltage exceeding LIWV could not be seen in Brazil.
Page 119
TypethenameofTB
Cumulative probability(%)
99.9
99
(1)
95
90
(1) Current of lightning strokes

on transmission towers
80
70
60
50
40
30
20
10
5
1
0.1
10
100
Current amplitude (kA)
Figure 4.4.4 Lightning stroke to
Figure 4.4.5 Distribution of
the upper phase line [3]
direct lightning stroke current

amplitude [3]
4.4.3 REPRESENTATIVE OVERVOLTAGES

The representative overvoltages are shown with the protection levels of surge arresters in Table 4.4.3 and Table
4.4.4.
Highest Voltage
800 kV
800 kV
800 kV
800 kV
Country/Utility
Brazil/
Furnas
China/
SGCC
India/
Power Grid
Korea/
KEPCO
1591
1300
1710
1716
1837
2025
1573
1380
(busbar or Tr.)
1491
(line entrance)
1480
1310
1420
1418
Tr.
N/A
1.15
0.88
1.31
N/A
1.03
Other
N/A
1.15
1.24
1.55
N/A
N/A
Representative
Value(kV)
Tr. / reactor
Other
LIPL (kV)
at 20 kA
Representative
overvoltage /
LIPL
No data
available
No data
available
800 kV
The United
States/
AEP
No data
available
No data
available
800 kV
Venezuela
1463 kV
No data
available
LIPL: Lightning Impulse Protection Level of surge arrester

800 kV system
Page 120
TypethenameofTB
Highest Voltage
1200 kV
1200 kV
1100 kV
1100 kV
1050 kV
Country/Utility
India/Power
Grid
Russia/
China/SGCC
Japan/
TEPCO
Italy/CESI
N/A
1796
1896
N/A
2040
2208
1760 (at
15kA)
1620
1620
1800
N/A
1.11
1.17
N/A
N/A
1.26
1.36
N/A
Representative
Value(kV)
Tr. / reactor
1700
Other
LIPL (kV)
at 20 kA
Representative
overvoltage /
LIPL
1700
Tr.
No data
available
No data
available
1.0
Other
LIPL: Lightning Impulse Protection Level of surge arrester

1100kV and above system
With regard to the representative overvoltages of the transformers for the 800 kV systems, the values range from
1710 kV to 1300 kV among the utilities, yielding a difference of as much as 400 kV. If the ratio of the representative
overvoltage to the protection level of a surge arrester is compared, the ratio in India is less than 1.0 in contrast with
that of Venezuela being almost equivalent to 1.03 and the others being more than 1.10. It means that India/800 kV
does assume less sever analysis condition such as the lightning stroke current and the circuit conditions than other
countries. Note that the discharge voltage at 20 kA is defined as the LIPL of surge arrester in the case of UHV
systems.
The representative overvoltage of other equipment is usually higher than those of transformers, and is over 1.15
times the protection level of surge arrester. As to both a transformer and other equipment, the representative
overvoltages of Korea are comparatively high and the ratio to the protection level of surge arrester is also large.
That could depend on the severest lightning current condition in Korea. The difference between each country is
less than about 300 kV.
The representative overvoltages of the transformers of the UHV systems, are 1700 kV in India, 1796 kV in China
and 1896 kV in Japan, yielding a difference of as much as 200 kV, which is less than the maximum of 400 kV for
the 800 kV systems. The representative overvoltage of India is in agreement with the protection level of a surge
arrester. The representative overvoltages of both transformer and GIS of Japan are the highest of all countries, and,
therefore, the ratio with the protection level of a surge arrester is also high. This is considered to be so because
lightning stroke current shape and a substation circuit configuration in Figure 4.4.6 are stricter. In general, the ratio
with a surge arrester protection level is small compared with that of an 800 kV system. The increase of the
protection level of a surge arrester is relatively higher than the increase of the lightning overvoltage due to the
severity of a lightning stroke current or substation circuit configuration.
Page 121
TypethenameofTB
Figure 4.4.6 Example of service operating conditions of substation in Japan [4]

In addition, representative overvoltages are defined as follows by IEC 60071-1;
overvoltages assumed to produce the same dielectric effect on the insulation as overvoltages of a given class
occurring in service due to various origins. They consist of voltages with the standard shape of the class , and may
be defined by one or a set of value or a frequency distribution of value that characterize the service condition.
Therefore, the representative overvoltages usually differ from the original wave shape of overvoltages which are
calculated with analysis programs. In order to convert analysis overvoltages to the standard impulse waveform
(1.2/50 s) in lightning surge time region, the evaluation of overvoltage waveshape explained in Section 5.3 can be
applicable.
REFERENCES
[1] IEC TR 60071-4, Insulation co-ordination - Part 4: Computational guide to insulation co-ordination and
modelling of electrical networks, First edition, 2004.
[2] J. Takami and S. Okabe, Observational Results of Lightning Current on Transmission Towers, IEEE Trans. on
Power Delivery, vol. 22, pp. 547-556, Jan. 2007.
[3] J. Takami and S. Okabe, Characteristics of Direct Lightning Strokes to Phase Conductors of UHV Transmission
Lines, IEEE Transactions on Power Delivery, Vol.202 No.1, pp.537-546, 2007.
[4] T. Watanabe, Y. Yamagata and E. Zaima, Insulation Coordination for UHV System, CIGRE Paper, No.33-101,
1998 Session.
978-2 85873-049-0
Page 122
TypethenameofTB
4.5 VFTO stress in GIS due to DS switching

4.5.1 GENERAL
In Gas Insulated Switchgear (GIS) or Hybrid-IS system, the switching of disconnecting switch (DS) will produce
very fast transient overvoltage (VFTO). VFTO has very high gradient and high frequency components. VFTO is
defined based on a lot of measurement and research. According to IEC 60071-1 (2006), VFTO is transient
overvoltage, usually unidirectional with time to peak Tf<0.1us, with or without superimposed oscillations at
frequency 30kHz<f<100MHz. The serious VFTO can damage the insulation of GIS and the interturn insulation of
transformer, and cause transient enclosure voltage (TEV) and electromagnetic interference (EMI) to low voltage
control systems. The amplitude and waveshape of VFTO will be decided by the practical condition of two sides of
DS, such as the length of busbar, the position of impedance changing, etc [1]. In the design of GIS and Hybrid-IS,
the influence of VFTO must be considered.
4.5.2 SIMULATION AND EXPERIMENTAL STUDY ON VFTO IN CHINA

State Grid Corporation of China (SGCC) has built the 1100kV UHV AC Demonstration Project. In this project, GIS
system was adopted in Changzhi substation, and Hybrid-IS systems were adopted in Nanyang switching station
and Jingmen substation.
In order to investigate the characteristics of VFTO generated in UHV GIS/Hybrid-IS, simulation study, experimental
study and field test on VFTO have been carried out by SGCC.
4.5.2.1 VFTO SIMULATION OF 1000KV UHV AC DEMONSTRATION PROJECT

4.5.2.1.1 SIMULATION CONDITION
The equivalent parameters of GIS and relative equipment for VFTO study are listed in Table 4.5.1. GIS busbar is
modelled as distributed parameters. Transformer, CVT, shunt reactor and GIS bushing are modelled as lumped
capacitors. Arrester is modelled as nonlinear resistance and its response characteristic to steep wave is considered.
Overhead line is modelled as transmission lines with distribution parameters. Closed circuit breaker and DS are
modelled as transmission lines. Open circuit-breaker is modelled as capacitors as shown in Figure 4.5.1. When DS
is open, there is no electric connection between two sides. Electrical arc model during restrikes has important effect
on VFTO simulation and a resistance and an inductance connected in series is used [2, 3, 4].
Equipment
GIS pipe
Restriking arc
Bushing
Transformer
Shunt Reactor
CVT
Overhead line
Calculation step time
Parameters
Z=92.68, v=273m/s
Ra=2, La=0.5H
C=126pF
C=5000pF
C=5000pF
C=5000pF
Z=310~363, v=300m/s
0.001s
Table 4.5.1 Equivalent parameters of GIS and connected equipment
Figure 4.5.1 Equivalent circuit and parameters of open circuit breaker
Page 123
TypethenameofTB
The initial voltage between DS contactors before restrike depends on the trapped charge voltage of load side and
the instantaneous voltage of source side. Considering the extreme unfavourable condition, it is supposed that
restrike occurs when the trapped charge voltage of load side is -1.0p.u. and the instantaneous voltage of source
side is 1.0p.u. (1.0p.u. = 1100kV 2 / 3 = 898kV, it's the peak value of maximum continuous operation voltage
of China 1100kV UHV AC system).
As there is no standard VFTO waveshape and withstand voltage of GIS, the required VFTO withstand voltage of
GIS is specified by the lightning impulse withstand voltage (LIWV) divided by safety factor. The safety factor for
China UHV AC Demonstration project was specified as 1.15, which takes into account that the VFTO withstand
voltage is higher than LIWV for sound GIS insulation. For the other types of equipment, the safety factor is the
same as GIS, excluding overhead busbar.
A deterministic approach is applied for VFTO insulation coordination. The calculated maximum value of VFTO at
each equipment is assumed as their representative VFTO and is compared with the required VFTO withstand
voltage shown in Table 4.5.2 [5]. The calculated maximum value of VFTO should be lower than the required VFTO
withstand voltage. Arrester is not able to protect equipment efficiently from VFTO due to the very fast front of VFTO,
the inductance and distance effect of arrester.
Standard LIWV (kV)

Required VFTO withstand voltage (kV)
Transformer
2250
1957
Shunt reactor
2250
1957
GIS
2400
2087
CVT
2400
2087
CT
2400
2087
Table 4.5.2 Standard LIWV and required VFTO withstand voltage of China 1100kV
equipment
4.5.2.1.2 SIMULATION RESULTS
1) VFTO in Changzhi GIS system
In Changzhi GIS substation, the double busbar and single busbar system are considered in design. Figure 4.5.2 is
the electric equipment arrangement 1 of single busbar system. It can be easily extended to one and a half busbar
system in the future. Figure 4.5.3 is the electric equipment arrangement 2 of double busbar system [6].
Figure 4.5.2 Electric equipment arrangement 1 in Changzhi GIS system
Page 124
TypethenameofTB
Figure 4.5.3 Electric equipment arrangement 2 in Changzhi GIS system

The overvoltage waveshapes at transformer, shunt reactor and CVT are not VFTO waveshape anymore since the
overhead line damps the VFTO strongly due to corona effect and conductor resistance of transmission line.
The maximum values of VFTO under various switching modes are listed in Table 4.5.3. The maximum value is
2249kV in arrangement 1 and 2260kV in arrangement 2. VFTO inside GIS exceeds the required VFTO withstand
voltage and some countermeasures are necessary. Damping resistor is a riper and most extensively used solution.
Arrangement
Arrangement 1
Arrangement 1
Arrangement 2
Damping resistor
Without
With
Without
GIS
2249
1250
2260
GIS bushing
2295
1141
1722
Transformer
954
905
942
Shunt reactor
1252
982
1012
CVT
1233
928
1015
Note: Resistor value of 500 was used in this calculation.
Table 4.5.3 Maximum values of VFTO in Changzhi GIS system (kV)

In order to verify the damping effect on VFTO, the voltage drop and energy consumption of damping resistor in DS,
damping resistor value varied from 100 to 700 is used to calculate VFTO. Damping resistor of 500 is finally
used in the Changzhi GIS, the maximum values of VFTO are limited to be much less than the required VFTO
withstand voltage and listed in Table 4.5.3.
2) VFTO in Nanyang and Jingmen Hybrid-IS systems
Figure 4.5.5 and 4.5.6 are electric equipment arrangements of Nanyang and Jingmen Hybrid-IS systems.
Page 125
TypethenameofTB
Figure 4.5.5 Electric equipment arrangement of Nanyang Hybrid-IS system
Figure 4.5.6 Electric equipment arrangement of Jingmen Hybrid-IS system

Switching modes of DS with and without damping resistor are calculated and the maximum values of VFTO are
listed in Table 4.5.4. The peak values of VFTO at GIS, bushing, shunt reactor and CVT don't exceed their required
VFTO withstand voltage. It is unnecessary to install damping resistor in DS.
The maximum overvoltage 2742kV (3.05p.u.) appears at the end of overhead busbar due to the full reflection of
overvoltage at the open end, the overvoltage should be lower if overhead busbar model with corona effect is used
in the simulation. The insulators and air clearance of overhead busbar have ability to withstand higher overvoltage
which has much shorter tail time than standard lightning waveform.
Page 126
TypethenameofTB
Substation
Nanyang
Jingmen
Damping resistor
Without
With
Without
With
GIS GIS bushing

1878
1850
1204
1199
1836
1948
1268
1266
Shunt reactor
1052
989
1018
943
CVT
1006
966
1041
993
Overhead busbar
2742
1306
2421
1314
Note: Resistor value of 500 was used in this calculation.
Table 4.5.4 Maximum values of VFTO in Nanyang and Jingmen Hybrid-IS systems
(kV)
3) Conclusions
(1) VFTO in Changzhi GIS system is severe. The maximum values of VFTO inside GIS in both arrangements reach
to 2249kV and 2260kV, exceed the required VFTO withstand voltage of GIS (2087kV), and it is necessary to use
damping resistor in DS.
(2) Damping resistor of 500 is used in Changzhi GIS system. The maximum value of VFTO can be reduced from
2249kV to 1250kV in arrangement 1, the amplitude and gradient of VFTO are decreased effectively and high
frequency components of VFTO are reduced obviously.
(3) VFTO values at all equipment in Nanyang and Jingmen Hybrid-IS systems dont exceed the required VFTO
withstand voltage, and it is unnecessary to install damping resistor in DS.
4.5.2.2 EXPERIMENT STUDY ON VFTO TEST CIRCUITS OF FULL SCALE UHV GIS
4.5.2.2.1 INTRODUCTION
In 2009, two full scale VFTO test circuits with practical 1100kV GIS were established by SGCC at UHV AC test
base in Wuhan, more than two thousand tests were carried out, and VFTO measurement systems were carefully
developed to capture the whole and accurate waveforms.
4.5.2.2.2 VFTO TEST CIRCUITS OF FULL SCALE UHV GIS

According to the practical situation of three substations in China UHV AC Demonstration Project, two test circuits
with full scale UHV GIS which can generate severe VFTO were built to investigate the characteristics of VFTO. The
circuit diagram of them is shown in Figure 4.5.7, equipment in the frame with dash lines refers to test GIS and
points marked with a number are VFTO measurement locations where porthole type sensors are installed. In order
to simulate the long busbar near DSs in Changzhi GIS system, a source side duct BS1 with changeable length
(three kinds of length) is connected to the source side of DS2.
The design of both test circuits is almost the same, the test GISs are the same as those used in China UHV AC
Demonstration Project. One test GIS is production of Xian XD Switchgear Electric Co., Ltd. (SHIKY) and the other
is one of Henan Pinggao Electric Co., Ltd. (PINGGAO). The layout diagram of both test GISs, which are
symmetrically placed at both sides of DC generator, is shown in Figure 4.5.8.
Page 127
TypethenameofTB
AS
BS1
BS2
BG2
:AC power source

:Source side duct
:Load side duct
:GIS bushing
BG1
DS2
DS1
DS
:GIS bushing
:Switching DS
:Auxiliary DS
:DC power source
Figure 4.5.7 Circuit Diagram of both VFTO test circuits
AC Tran.
SHIKY
DC gen.
PINGGAO
Figure 4.5.8 Layout diagram of both VFTO test circuits

In PINGGAO test GIS, DS1 is with 500 damping resistor and DS2 is without damping resistor in order to
investigate the mitigation effect of damping resistor on VFTO. In SHIKY test GIS, both DS1 and DS2 are without
damping resistor.
SHIKY DS is horizontal and operates slowly, while PINGGAO DS is vertical and operates quickly. The operation
speeds of both DSs are listed in Table 4.5.5.
PINGGAO DS
SHIKY DS
Opening speed (m/s)

1.7
0.54
Closing speed (m/s)

2.5
0.54
Table 4.5.5 Operation speeds of DSs

Each VFTO measurement system is comprised of a porthole type sensor, an oscilloscope, a power supply and a
shielding cage. For the porthole type sensor, which is shown in Figure 4.5.9, the high voltage capacitor C1 is
formed between the phase conductor of GIS and the plate electrode of sensor, the low voltage capacitor C2 is
formed by putting a thin film between the plate electrode of sensor and the blind flange of porthole. The low and
Page 128
TypethenameofTB
high frequency performance of measurement system were calibrated under voltage sources with different
waveshapes using a specially developed calibration setup in laboratory. The impedance transformation circuit was
designed to amplify the output impedance of capacitive divider and achieve the impedance matching with the
following circuit, thus the low cutoff frequency of the sensor can be extended greatly. The calibration results
showed that the bandwidth of measurement systems are about 0.01Hz-200MHz, there is no parasitic oscillation,
and the voltage ratio is stable, the performance is adequate for VFTO measurement.
Figure 4.5.9 Illustrative sketch of the porthole type sensor

In order to capture the whole VFTO waveforms with very high frequency components and repetitive restrikes
sustained for a long period, Tektronix DPO7054 and Yokogawa DLM2054 oscilloscopes with high bandwidth and
large memory were used in the measurement systems.
An optical fiber triggering system was developed to trigger all oscilloscopes at different measurement locations in
order to capture the VFTO waveforms synchronously.
To avoid the effects of EMI and TEV generated from switching operations of DS, each oscilloscope was placed in a
shielding cage directly connected to GIS enclosure and powered by batteries.
4.5.2.2.3 TEST RESULTS

Considering the effect of trapped charge voltage on VFTO, tests with and without precharging 1.0p.u. DC voltages
(1.0p.u.=1100kV/ 3* 2=898kV) were carried out using both SHIKY and PINGGAO GISs. In order to analyze the
effect of source side duct on VFTO, tests have been carried out using source side duct with different lengths. In
order to analyze the mitigation of VFTO by damping resistor, tests have been carried out using PINGGAO DS with
damping resistor. Test contents are shown in Table 4.5.6. Through total more than 2000 tests, waveform
characteristics of VFTO, distribution of VFTO magnitude and mitigation of VFTO by damping resistor are obtained.
Test
numbers
120
120
Precharging
1.0p.u DC
voltage
Yes
Source side duct

(BS1,m)
No
Damping
resistor
PINGGAO
With
Without
SHIKY
Without
PINGGAO
Without
SHIKY
Without
120
450
450
450
450
Test GIS
9
3
0
9
Table 4.5.6 VFTO Test contents
Page 129
TypethenameofTB
1) Waveform characteristics of VFTO

(1) Whole waveform
A large number of whole waveforms are captured under different test conditions, which differ obviously from the
source side to load side. As shown in Figure 4.5.10, the whole waveform of VFTO is with multiple voltage steps at
the load side, while it is without voltage step at the source side.
VFTO is a continuous process with multiple restrikes, the restrikes continue until the DS contactors close during
closing or the gap between DS contactors becomes large enough during opening. So the duration of VFTO
depends on the prestriking time during closing and arc extinction time during opening, the restrike numbers of
PINGGAO GIS during one operation are less than those of SHIKY GIS, no matter closing or opening, as shown in
Figure 4.5.11.
(a) Load side
(b) Source side
Figure 4.5.10 Whole VFTO waveform during closing
(a) PINGGAO GIS
(b) SHIKY GIS
Figure 4.5.11 Whole VFTO waveform during opening at the load side
The restrike numbers are statistically analyzed under condition that voltage across the DS contactors before
restrike is higher than 0.3p.u., as shown in Figure 4.5.12. The average restrike numbers are about 5 during
opening and 6 during closing of PINGGAO DS, which are about half of those of SHIKY DS.
Page 130
TypethenameofTB
SHIKY GIS
PINGGAO GIS
SHIKY GIS
PINGGAO GIS
30
Occurence probability / %
Occurence probability / %
35
30
25
20
15
10
5
0
25
20
15
10
5
0
10
12
14
16
Restrike number during operation
10
12
14
16
18
Restrike number during operation
(a) Opening operation
(b) Closing operation
Figure 4.5.12 Histograms of restrike numbers

The trapped charge voltage generated by the last strike during opening of DS is determined by the opening time of
DS and the phase angle of last strike which is affected by the strike characteristic of DS. The VFTO waveforms
captured at measurement location 1 are analyzed statistically to obtain the distributions of trapped charge voltage
in Figure 4.5.13. The moving contactor with near flat surface electrode is installed at source side, so the probability
of trapped charge voltage with negative polarity is higher than that with positive polarity due to the breakdown
voltage between DS contactors is lower at negative polarity because of polarity effect in GIS. If the moving
contactor of DS is connected to load side, the probability of trapped charge voltage with positive polarity will be
higher than that with negative polarity.
The induced AC voltage in the load side before closing or after opening of DS, which can be observed from Figure
4.5.10 and Figure 4.5.11, is generated by the capacitance divider formed with capacitor of opened DS contactors
and capacitor of the load side duct to earth. The induced AC voltage in the load side would be much lower when a
circuit-breaker is connected with DS.
Occurence Probability / %
14
SHIKY GIS
PINGGAO GIS
12
10
8
6
4
2
0
-1.0 -0.8 -0.6 -0.4 -0.2
0.0
0.2
0.4
0.6
0.8
1.0
Trapped charge voltage / p.u.
Figure 4.5.13 Distribution of trapped charge voltage during opening
Page 131
TypethenameofTB
(2) Single strike waveform

VFTO is a continuous process with multiple strikes, constituted with many single strike waveforms which are
typically illustrated in Figure 4.5.14. From 10 single strike waveforms of VFTO with maximum magnitudes
generated during closing, the magnitude and its rise time from the start base line of the first rising edge (indicated
as A1 in Figure 4.5.14) and the peak (indicated as A2 in Figure 4.5.14), the strike duration and maximum gradient
are calculated, these parameters differ between PINGGAO and SHIKY GIS, as shown in Table 4.5.7.
1000
Start Base Line
Amplitude / kV
500
A1
-500
A2
-1000
-1500
-300
-150
150
300
450
600
750
Time / ns
Figure 4.5.14 Typical single strike waveform of VFTO
PINGGAO
GIS
SHIKY GIS
Place
Magnitude
(p.u.)
Rise
time(ns)
A1
A2
A1
A2
0.77
2.38
0.99
2.64
8.8
366
6.4
298
Maximum
gradient
(kV/ns)
Strike
duration(s)
78.6
9.7
138.9
5.6
Table 4.5.7 Parameters of single strike waveforms of VFTO

2) Distribution of VFTO magnitude
VFTO magnitude is random affected by the switching time, the strike behaviour of contactors, and the electric arc
characteristic of DS etc. The trapped charge voltage has an important impact on the distribution of VFTO
magnitude. Two test cases, with and without precharging DC voltage, are considered to simulate the most severe
and natural distribution trapped charge voltage.
(1) Precharging DC voltage
The most severe VFTO is generally generated if a voltage collapse across DS contactors occurs when the
instantaneous voltage of power supply is at the peak value which polarity is opposite to the trapped charge voltage
of the load side during closing of DS. IEC62271-102 provides the test method of precharging maximum magnitude
DC voltage on the load side to consider the most severe trapped charge voltage. The maximum values of VFTO
during closing operation with precharging 1.0p.u. DC voltage are shown in Table 4.5.8, they are close to those
during opening operation.
Page 132
TypethenameofTB
Measurement location
PINGGAO GIS
SHIKY GIS
1
2.01
2.20
2
1.98
2.17
3
2.27
1.86
4
1.47
1.30
Table 4.5.8 Maximum VFTO during closing operation (p.u.)

The distributions of maximum VFTO are statistically analyzed and shown in Figure 4.5.15 with the waveforms
captured at measurement location 3 for PINGGAO GIS and measurement location 1 for SHIKY GIS. The maximum
magnitudes of VFTO present normal distribution during opening and the probability of high magnitudes of VFTO is
higher during closing for the effect of precharging DC voltage.
Closing operation
Opening operation
30
25
20
15
10
5
0
1.0
1.2
1.4
1.6
1.8
Closing operation
Opening operation
35
Occurrence Probability / %
35
2.0
2.2
2.4
30
25
20
15
10
5
0
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
Maximum VFTO/p.u.
Maximum VFTO/p.u.
(a) PINGGAO GIS
(b) SHIKY GIS
Figure 4.5.15 Distribution of maximum VFTO with precharging DC voltage

(2) Without precharging DC voltage
The maximum VFTO without precharging DC voltage are listed in Table 4.5.9, they differ slightly with those without
precharging DC voltage at the same measurement locations.
PINGGAO GIS
SHIKY GIS
1
2.01
2.03
2
2.00
1.92
3
2.23
1.86
4
1.33
1.27
Table 4.5.9 Maximum VFTO without precharging DC voltage (p.u.)

The distributions of maximum VFTO with and without precharging DC voltage are shown in Figure 4.5.16. The
distribution of maximum VFTO during closing varies widely and the probability of high magnitudes with precharging
DC voltage is higher. The distribution of trapped charge voltage without precharging DC voltage is natural, leading
to a lower probability of high magnitudes of VFTO.
Page 133
TypethenameofTB
with Pre-charging DC Voltage

without Pre-charging DC Voltage
with Pre-charging DC Voltage

without Pre-charging DC Voltage
35
30
35
25
20
15
10
5
0
30
25
20
15
10
5
0
1.0
1.0
1.2
1.4
1.6
1.8
2.0
2.2
1.2
2.4
1.4
1.6
1.8
2.0
2.2
2.4
M aximum VFTO/p.u.
Maximum VFTO/p.u.
(a) PINGGAO GIS
(b) SHIKY GIS
Figure 4.5.16 Distribution of maximum VFTO during closing

3) Effect of source side busbar on VFTO
All the above results are obtained with the source side duct (BS1) which length is 9m. For PINGGAO GIS, the
maximum magnitudes of VFTO without precharging DC voltage are compared with varied lengths of BS1 in Table
4.5.10. As the length of BS1 increases, the maximum magnitudes of VFTO tend to increase at measurement
location 1, 2 and 3, but decrease at measurement location 4. The maximum magnitude of VFTO is gotten at
measurement location 1 without BS1, as the length of BS1 is extended to be 9m, the maximum magnitude of VFTO
is generated at measurement location 3.
Length of BS1 (m)
0
3
9
1
1.79
1.90
2.01
2
1.65
2.00
1.88
2.23
4
1.53
1.33
Table 4.5.10 Maximum VFTO with different length of BS1 (p.u.)

4) Mitigation of VFTO by damping resistor
The maximum VFTO generated from PINGGAO DS with and without damping resistor are listed in Table 4.5.11,
they can be obviously suppressed by damping resistor from 2.27p.u. to 1.33p.u. As shown in Figure 4.5.17, high
frequency oscillation and gradient of VFTO have also been mitigated by damping resistor.
Without damping resistor
With damping resistor
1
2.01
1.33
2
1.98
1.31
3
2.27
1.25
4
1.47
1.22
Table 4.5.11 Maximum VFTO with/without damping resistor (p.u.)
Page 134
TypethenameofTB
(a) Without damping resistor
(b) With damping resistor
Figure 4.5.17 Waveforms of VFTO generated from PINGGAO GIS during closing
4.5.2.2.4 CONCLUSIONS
(1) The average restrike numbers are about 5 during opening and 6 during closing of PINGGAO GIS, which are
about half of those of SHIKY GIS, under condition that re-striking voltage across DS contactors is higher than 0.3
p.u.
(2) The probability of trapped charge voltage with negative polarity is higher than that with positive polarity when
moving contactor of DS is connected to source side. The probability of trapped charge voltage with high magnitude
of SHIKY GIS is obviously lower than that of PINGGAO GIS.
(3) The maximum VFTO is about 2.27p.u., which are close during opening and closing operation with or without
precharging 1.0 p.u. DC voltage.
(4) Damping resistor fixed in DS can mitigate the amplitude and gradient of VFTO significantly.
4.5.2.3 FIELD TEST OF 1100KV UHV AC DEMONSTRATION PROJECT

4.5.2.3.1 TEST CONDITIONS
In order to verify the safety of switching UHV GIS DS energized, field tests of switching all GIS DSs in Changzhi,
Nanyang and Jingmen substations have been carried out by SGCC from March to May, 2011 and VFTO was
monitored.
There were no installation conditions on-site to apply the porthole type sensor in UHV GIS/Hybrid-IS. However,
there were some disc spacers with metal rings embedded in Changzhi substation, which were used to form ring
embedded sensors to measure VFTO inside GIS. In addition, sensors installed at the last shielding of transformer
and reactor bushings were used to measure VFTO at transformer and reactor. Layout and measurement points of
each substation are shown in Figure 4.5.18, and Figure 4.5.19 shows the actual installation of measurement
devices on-site.
Page 135
TypethenameofTB
Chang-Nan
Line I
CVT
Disconnector
Br eak
Circuit Breaker
M1
VFTO-2
Bushing
T0111
VFTO-X
VFTO measuring point
T011
T0112
VFTO-3
T0121
T0131
VFTO-1
CVT
VFTO-4
500kV
system
T012
1Main power
tr ansfor mer
T0122
(a) Changzhi substation
(b) Nanyang switching station
(c) Jingmen substation
Figure 4.5.18 Equipment Layouts and measurement points of substations
Page 136
TypethenameofTB
(a) Ring embedded sensor
(b) Sensor at the last shielding of bushing
Figure 4.5.19 Actual installation of measurement devices on-site

4.5.2.3.2 TEST RESULTS
Typical measurement waveforms of VFTO from the ring embedded sensors are shown in Figure 4.5.20, Figure
4.5.20 (a) is a whole process waveform while Figure 4.5.20 (b) is a single strike one. Measurement results of VFTO
in Changzhi substation are shown in Table 4.5.12 and measurement results of overvoltage at Nanyang and
Jingmen substations are shown in Table 4.5.13.
(a) Whole waveform
(b) Single strike waveform
Figure 4.5.20 Measurement waveforms of VFTO from ring embedded sensors in

Changzhi substation
Switching DS
T0122
T0131
T0111
T0121
T0112
Maximum values of VFTO /p.u.

VFTO-1 VFTO-2 VFTO-3 VFTO-4
1.07
1.07
1.11
1.29
1.00
1.30
1.07
1.25
1.06
1.00
1.00
1.06
1.00
1.00
1.02
Table 4.5.12 Measurement results of VFTO in Changzhi substation
Page 137
TypethenameofTB
Substation and
switching DS
T0222
T0211
Nanyang
T0221
T0212
T0312
T0321
Jingmen
T0322
T0311
Maximum values of overvoltage /p.u.

VFTO-1
VFTO-2
1.00
1.00
1.05
1.02
1.22
1.13
1.12
1.11
1.05
1.07
Table 4.5.13 Measurement results of overvoltage at transformer and reactor in

Nanyang and Jingmen substations
It can be summarized as follows from the data in Table 4.5.12 and 4.5.13:
1)The whole level of VFTO inside GIS is relatively low in Changzhi substation, and the maximum VFTO is 1.29 p.u.
which is less than the simulation result 1.39 p.u. It can be inferred as the mitigation effect of the damping resistors
in UHV DSs.
2)The whole level of overvoltage at the bushings of transformer and reactoris relatively low in three substations.
The reason is that VFTO produced inside GIS travel along long overhead line to reach the transformer and reactor,
overhead line damps VFTO level.
3) The level of measured data might be relatively low because of their testing numbers and measuring methods.
4.5.3 COMPARISON OF MEASUREMENT AND SIMULATION
4.5.3.1 GENERAL OVERVIEW

VFTO simulation is a well-known instrument for the calculation of overvoltages needed for the insulation coordination process [17]. Because the accuracy of the simulation depends strongly on the quality of the model of
each individual component, it is important to verify the simulation results by measurements.
This chapter describes simulations of very fast transient overvoltages with the aim to compare the simulated values
with the measured ones. The impact of the GIS disconnector design on Trapped Charge Voltage (TCV)
distributions and Very Fast Transient Overvoltages (VFTO) was investigated. The study was performed for full
scale test set-up installed in State Grid Corporation China (SGCC) Wuhan GIS substation at 1100 kV voltage level
as described in chapter 4.5.2.2.
The maximum value of the VFTO depends on the voltage drop at the DS just before striking and the location
considered. Trapped charge remaining on the load side of a DS must be taken into consideration. A trapped
charge on the load side resulting in a voltage of -1 p.u. (2 p.u. across the DS) is normally taken into account as the
most unfavorable case for high speed DS or phase opposition conditions. This precondition is normally used for the
calculation of VFTO [9, 10]. For this case the maximum VFTO peak in GIS configuration has a typical value
between 1.5 p.u. and 2.8 p.u.
A comparison between simulation and measurement can verify the accuracy of the simulation [12]. Therefore, the
simulation method was also used during testing. According to IEC 62271-102, Annex F, VFTO amplitudes of at
least 1.4 p.u. are required without pre-charging of the busbar. Figure 4.5.21 shows a comparison of simulated and
measured VFTO without and with DC pre-charging of the busbar. The measured voltage progressions coincide
very well with the simulation results as regards VFTO amplitude and rise time[14].
Page 138
TypethenameofTB
1500
2000
1300
1500
Voltage [kV]
Voltage [kV]
1100
900
700
500
1000
500
0
-500
300
Measurement
Calculation
100
Measurement
Calculation
-1000
-1500
-100
1.8
1.9
2.1
2.2
2.3
2.4
1.8
2.5
1.9
2.1
2.2
2.3
2.4
2.5
Time [us]
Time [us]
Figure 4.5.21 VFTO calculation and measurement when switching busbars with a
GIS DS as per IEC 62271-102, without pre-charging (left), with pre-charging (right)
As stated above trapped charge remaining on the load side of a DS must be taken into consideration [13]. Totally
400, 520 records of measured trapped charge voltages during testing of DS have been evaluated for 550 kV, 1100
kV respectively. For the 550 kV DS, the maximum trapped charge reaches 0.6 p.u. during the tests with a source
voltage of 1.1 p.u. resulting in a most unfavorable voltage collapse of 1.7 p.u. In case of 1100 kV DS 90 % of the
trapped charge voltages were limited to 0.45 p.u. (see Figure 4.5.22).
8%
F(x) %
1100 kV
99,9
550 kV
99
6%
4%
Probability
95
90
70
50
30
10
5
1100 kV
2%
550 kV
0,10
0%
-1
-0.75
-0.5
-0.25
0.25
0.5
0.75
0.0
0.2
0.4
0.6
0.8
[pu]
1.0
Trapped charge voltage [pu]
Figure 4.5.22 Measured trapped charge voltage during DS switching under

laboratory conditions, probability density (left), approximation of absolute values
by normal distribution (right)
For all test duties the records have been evaluated separately and the trapped charge voltages do not show any
significant dependency of the different test duties. Figure 4.5.23 shows the cumulative probability of the trapped
charge voltages for all test duties (1100 kV). In order to compare the result of TD 3 with the others, its trapped
charge voltages had to be scaled for 1.1 p.u. source voltage.
The distribution of the trapped charge voltages have been simulated to investigate if the measured distributions can
be reproduced and to study the influence to the DS speed on these distributions.
For the simulations two parameters of the disconnector have to be known [15]. The first is the breakdown voltage
characteristic depending on the contact distance. The electric fields on the relevant parts of the moving and the
fixed contact side have been determined using dielectric calculations. The polarity effect of SF6, the surface
Page 139
TypethenameofTB
roughness and the gas pressure were taken into account. The second parameter is the contact speed. The
simulation model simulates the test circuit for the relevant test duty according IEC62271-102 Annex F. The
capacitance values of the source and load side have been determined according the length and typical value of
GIS bus duct capacitance. The disconnector is modeled as an ideal switch that closes if the command is >0 and
opens at the next current zero if the command is equal to 0. The command is set to 1 if the voltage across the
disconnector exceeds the breakdown voltage characteristic of the disconnector. The starting point of the contact
separation has been varied in 1 steps all around 360. It can be seen that the measured trapped charge
distributions can be reproduced with good agreement for the 1100 kV disconnector (see Figure 4.5.23 right).
20%
100%
Simulated
Measured
80%
40%
TD1
TD2
TD3
10%
Probability
all TD
Cumulative probability
60%
15%
5%
20%
TD3 (appr. for 1.1 p.u.)

0%
0.0
0.2
0.4
0.6
0.8
0%
1.0
0.0
0.2
Trapped charge voltage (abs.) [pu]
0.4
0.6
0.8
1.0
Figure 4.5.23 Measured trapped charge voltage during DS switching under

laboratory conditions, cumulative probability density for the different TD at 1100
kV (left), comparison between measured and calculated values at 1100 kV (right)
The trapped charge voltage distribution depends strongly on the contact speed. For contact speeds exceeding 1
m/s, the simulation results suggest that the 90 % trapped charge voltage increases significantly from around 0.5
p.u. to 0.8 p.u. One example for one specific design is shown in Figure 4.5.24. It has to be mentioned that beside
the contact speed also other parameter like gas pressure and local electrical fields have an influence on the TCV
distribution.
100%
40%
0.1 m/s
0.3 m/s
1.5 m/s
3 m/s
20%
Cumulative probability
60%
Trappedchargevoltage(abs.) [pu]
80%
calculated
90% probability
95% probability
99% probability
measured
0.8
0.6
0.4
0.2
0
0
0.5
1.5
2.5
3.5
Contact speed [m/s]
0%
0.0
0.2
0.4
0.6
0.8
1.0
Figure 4.5.24 Influence of contact speed on trapped charge voltage during DS

switching, cumulative probability (left), influence of contact speed, statistic values
(right)
However, depending on the design of the disconnector (especially contact speed, dielectric design of the contacts
and SF6 pressure) the assumption of trapped charge resulting in 1 p.u. voltage is a very conservative assumption
Page 140
TypethenameofTB
for VFTO calculations. The evaluation of type test results for the 1100 kV disconnector have revealed that the 90 %
voltage associated with trapped charges where 0.4 p.u. at a source voltage of 1.0 p.u. (see Figure 4.2.25). That
means for this 1100 kV disconnector having such a trapped charge behavior, a safety factor of around 1.43 (= 2
p.u./1.4 p.u.) is included when 1 p.u. trapped charge voltage is assumed for the VFTO calculation, assuming a
linear relationship between the voltage drop before closing and VFTO peak value. Moreover, it must be considered
that the VFTO peak value also depends on the location and the polarity of the trapped charge voltages. In case of
very low trapped charge voltages after opening, the VFTO peak values during opening could be higher compared
to a closing operation. A detailed study or the entire opening and closing operation is required in such cases.
4.5.3.2 TEST SET-UP MODELING

Simulations presented in this chapter relate to the two following set-ups, as presented in Figure 4.5.7 and Figure
4.5.8. The details are described in chapter in chapter 4.5.2.2.
Two options were considered for the layout modeling:
- Option 1: coaxial bus ducts are modeled by the surge impedance and speed of travelling wave given by
the dimensions of the test set-up layout
- Option 2: coaxial bus ducts are modeled by the surge impedance of 91 and speed of travelling wave of
290 m/us
Both options were used for modeling. Equivalent circuits of the GIS elements used in simulations are summarized
in Table 4.5.14.
Page 141
TypethenameofTB
Disconnector
under test
BDV characteristic:
- linear
- non-linear
Resistance at:
- Open state 1e16
- Closed state: 1e-16
Busbar model
Length: depends on the layout

Surge impedance depends on dimensions
Arc Model
Arc resistance = 2
Auxilay
disconenctor
Capacitance between contacts: 15pF

Capacitance to ground: 100pF
HV Bushing
Surge impedance:
- AIS side: 347
- GIS side: 81.73
Capacitance to ground: 50 pF
Table 4.5.14 Equivalent GIS models used in simulations

According to the description given above EMTP-ATP model of SHIKY set-up is presented in Figure 4.5.25.
Figure 4.5.25 EMTP-ATP model of SHIKY set-up
Page 142
TypethenameofTB
4.5.3.3 TRAPPED CHARGE VOLTAGE (TCV)

When the disconnector opening operation is completed, after the occurrence of the last re-strike, the Trapped
Charge Voltage (TCV) remains on the load side of the disconnector [11]. The trapped charge decay is a very slow
process, taking hours or days, resulting from the charge leakage across the spacers. During the subsequent
closing operation of the disconnector, the first pre-strike occurs when the source side voltage reaches its amplitude
with opposite polarity of the load side voltage (Trapped Charge Voltage) resulting from the preceding operation.
The VFT process depends on the voltage drop between the disconnector contacts just before a breakdown occurs,
hence the VFTO amplitude is directly affected by the TCV value.
Due to the statistical character of the VFT process, the trapped charge voltage has also a statistical distribution [14].
In state-of-the art approach used in co-ordination insulation calculations the single spark approach is used with the
very conservative assumption of -1 p.u. of trapped charge voltage [16]. This approach is used for verification
whether VFTO exceeds the Lightning Impulse Withstand Voltage (LIWV) of the GIS and/or adjacent equipment and
can be a limiting factor for the proper dielectric design of the GIS elements such as the disconnector itself.
Based on the TCV distribution, a real value of the TVC can be used for single spark calculations. This can lead to
an additional safety margin in the insulation co-ordination analyses and in consequence allow for reducing the total
cost of UHV substations.
The multi-spark disconnector model as a new approach has been employed as presented in [15]. In the model, not
only one spark, but the entire process of the disconnector operation is implemented, including many re- or prestrikes. The idea of the model is based on the concept which is commonly used when modeling of Vacuum Circuit
Breaker, as it is presented e.g. in [16]. The concept is to control the nonlinear resistance by using time dependent
resistance see Figure 4.5.26 [18].
ATP/EMTP
MODELS
i
r = r(t,uS,uL,i)
uS
uL
Figure 4.5.26 EMTP-ATP model of the disconnector [6]; r TACS controlled time
dependent resistance, uS disconnector source side voltage, uL disconnector
load side voltage
For the purpose of the model development the MODELS tool of the EMTP-ATP program has been used [16]. By
means of the control procedure, a decision is being made whether the spark is to be ignited or extinguished.
The decision is based on the values of potentials on the source and on the load side of the disconnector, which are
compared with the withstand voltage of the contact gap. The withstand voltage is calculated in each simulation step
on the basis of moving contact velocity obtained from mechanical characteristics of a real disconnector and the
withstand voltage characteristics of the disconnector contacts system. Trapped charge voltage is modeled with the
use of an additional capacitance, which is disconnected by the ideal switch after the ignition of the first spark in the
closing operation process.
Page 143
TypethenameofTB
Design parameters of the disconnector, such as contact speed, SF6 gas pressure and dielectric design of the
disconnector contact system, influence the withstand characteristics of the disconnector and hence influence the
Trapped Charge Voltage behavior of the disconnector.
The breakdown voltage (BDV) characteristic of the disconnector during the operation is one very important
parameter as input for simulation of the trapped charge behavior of the disconnector. Two types of breakdown
voltage characteristics were used for the simulations:
- Linear BDV (1)
- Nonlinear BDV (2)
The linear BDV characteristic was evaluated using the measurement of each breakdown voltage during one
switching operation. The obtained values are correlated to the position of the moving contact. Because the BDV
depends on the polarity of each contact a large scattering of measured values could be observed as shown in
Figure 4.5.27. The approximation used for the simulations is shown in Figure 4.5.27 (right). In general the linear
BDV does not include worst case of breakdown between contacts, thus electrical withstand rises faster, causing
higher TCV and lower sparking time.
Figure 4.5.27 Linear BDV (1) used as input, together with the approximation used
for simulations
The Nonlinear BDV characteristic was evaluated by lightning impulse and AC withstand voltage tests for different
contact distances and comparison with dielectric field calculations and is shown in Figure 4.5.28. For easier
simulation the BDV characteristic as shown in Figure 4.5.28 (right) was used. Generally the nonlinear BDV includes
worst case of the breakdown between contacts thus whole process of opening lasts longer and there is lower TCV
and longer sparking time.
Page 144
TypethenameofTB
Figure 4.5.28 Nonlinear BDV (2) used as input, together with the approximation
used for simulations
Trapped Charge Simulation for the SHIKY set-up
The simulation results for the slow acting disconnector in the SHIKY test set-up with use of linear BDV (1) are
presented in Figure 4.5.29. Two independent research institutes have performed the simulations of trapped charge
voltage using the same BDV characteristic. Both TCV distribution functions using the linear BDV characteristic are
in good agreement, but there are characterized by unsymmetrical distribution. The probability function (shape) is
the same for both simulations. Nevertheless it was proven that the differences between the simulations or model
parameter are very small.
Figure 4.5.29 Trapped Charge Voltage distribution, simulation results of two

different research institutes with use of linear BDV characteristics; opening
operation
But there are differences to the distribution function measured on-site. Especially the shape of the distribution
function is different. The probability of extreme values are much lower compared to the simulations. The measured
distribution function is shown in Figure 4.5.13.
The simulation results for the slow acting disconnector in the SHIKY test set-up with use of nonlinear BDV (2) are
presented in Figure 4.5.30. The only difference in the model from the previous one is that different nonlinear BDV
was now used.
Page 145
TypethenameofTB
Figure 4.5.30 Trapped Charge Voltage distribution, simulation results with use of
nonlinear BDV characteristics; opening operation
The simulation results have same probability function as for measurements. Trapped charge voltage obtained for
the model with nonlinear BDV has more realistic probability shape. It is close to the Gaussian function. But
simulation results have a lower TCV value than for measurements. There could be the following possible
explanations of the disagreement between simulation results and measurements: the nonlinear BDV characteristics
is likely to be more accurate characteristics and linear BDV is too much simplified.
To prove the accuracy of simulation results obtained with the use of nonlinear BDV, the following comparisons
were made: Trapped charge voltage distribution with use of nonlinear BDV vs. measurements during type tests of
1100 kV disconnector and sparking time simulations with use of nonlinear BDV vs. measurements from SHIKY test
set-up.
In Figure 4.5.31 the measurement results are presented which were used as a reference for comparison with
simulation results, see Figure 4.5.32. The simulation results were obtained with use of nonlinear BDV
characteristics. During the tests also the trapped charge voltages were measured using sensors on the load side.
Figure 4.5.31 Measured Trapped Charge Voltage Distribution (left); based on the
type Test measurements in Test Duty 1 set-up according to IEC (right)
Page 146
TypethenameofTB
Both simulations and measurements were made for the following data: 1100 kV design, opening operation at 0.4
m/s contact speed and test duty 1 set-up according to IEC. More details are given in [14, 15].
Figure 4.5.32 Trapped Charge Voltage distributions: comparison between

measurements and simulations
There is an excellent agreement in terms of mean values. The standard deviation is slightly different. The mean
value agreement is sufficient to prove of BDV characteristics used and for the application for TCV and VFTO
calculations.
The comparison of the simulation results for the sparking time is of importance since the measuring method for the
sparking time is easier and thus in some cases it can be much more reliable that the measuring method for
Trapped Charge Voltage. Sparking time was compared between the distributions obtained in simulations with use
of nonlinear BDV and measurements. In Table 4.5.14 the values from measurements are presented.
Table 4.5.14 Sparking time values from measurements

Simulation results are presented in Figure 4.5.33.
Page 147
TypethenameofTB
Figure 4.5.33 Sparking time simulated for nonlinear BDV

Simulation results are in agreement with the measurement results provided by measurements.
4.5.3.4 TRAPPED CHARGE VOLTAGE IMPACT ON VFTO IN UHV GIS SUBSTATION

To show the effect of the trapped charge voltage on the VFTO calculations were performed for a typical UHV 1100
kV GIS substation. The values of the Trapped Charge Voltage of a slow acting disconnector for probability of 99%
were used. The TCV values are presented in Table 4.5.15.
Contact speed [m/s]
TCV99% (abs) [p.u.]
0.1
0.24
0.3
0.36
1.0
0.55
1.5
0.68
3.0
1.00
Table 4.5.15 Trapped Charge Voltage for 99% probability

The value of the TCV99% for contact speed equals 3 m/s corresponds to the state-of-the-art conservative
approach of -1 p.u. typically used in insulation co-ordination analyses. The VFT process has been characterized by
the following parameters: VFT overvoltage (VFTO), number of sparks during the VFT process, and the sparking
time. The parameters were calculated during the disconnector closing operation for exemplary 1100 kV GIS
substation on the disconnector load side. For the purpose of VFT calculations for different levels of the TCV, the
model of an exemplary UHV 1100 kV GIS substation has been developed. The calculations were performed for the
closing operation of the disconnector.
The simulation results are presented in Figure 4.5.34. Each value represents VFT calculations for closing operation
of the disconnector in the UHV GIS substation for selected TVC values of 50%, 90%, 95%, and 99%. For the slow
acting disconnector the VFTO level significantly decreases with respect to that for the fast acting one.
Page 148
TypethenameofTB
Figure 4.5.34 VFTO peak values simulated for nonlinear BDV

Based on the results presented in Figures 4.5.34 the VFTO reduction factor can be expressed as the ratio of VFTO
for a given disconnector contact speed (hence given TCV level), and given TCV probability, to the VFTO at the
TCV of -1 p.u. The results presented in that way are depicted in Figures 4.5.35. The relation between VFTO level
and the voltage value when the spark occurs is linear, thus the relations depicted in Figures 4.5.35 for VFTO are of
the same character as the relations of TCV in the TD set-up depicted in Figure 8. In Figures 4.5.35 the reduction of
as much as 20% is visible for the disconnector contact speed reduced to the value of 0.1 m/s.
Figure 4.5.35 VFTO reduction as a function of disconnector contact speed

It has to be mentioned however, that such a speed reduction would lead to the increase in both sparking time and
the number of sparks during the disconnector operation process (see Figures 4.5.36). The optimum design of the
disconnector should be considered to achieve significant VFTO reduction with acceptable sparking time and
number of sparks.
Page 149
TypethenameofTB
Figure 4.5.36 TCV and contact speed impact on sparking time and number of
sparks
4.5.3.5 COMPARISON OF MEASURED AND SIMULATED VFTO WAVEFORM
The waveform was measured on the load side of the operated disconnector, thus the VFTO has a classical stepwise shape, see Figure 4.5.37 and Figure 4.5.38. The trapped charge voltage levels are indicated.
1.5
TCV = 1
1*724.5 kV = 0.81 p.u.
0.5
-0.5
TCV = 1.1
-1
1.1*724.5 kV = 0.89 p.u.
-1.5
-0.02
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
Figure 4.5.37 VFTO waveform measured by SGCC in SHIKY layout sensor 2; entire
closing operation
Page 150
TypethenameofTB
Max VFTO = 2.2535 * 724464 =

= 1.63 MV = 1.82 p.u.
2.5
2
1.5
1.5
1
0.5
1
0
-0.5
0.5
-1
0
-1.5
0.02
0.022 0.024 0.026 0.028
0.03
0.032 0.034 0.036 0.038
0.04
0.0307 0.0307 0.0307 0.0307 0.0307 0.0307 0.0307 0.0307 0.0307 0.0307
Figure 4.5.38 VFTO waveforms measured by SGCC in SHIKY layout sensor 2; closing
operation, max VFTO (second spark, TCV = -0.89)
The waveform was measured on the source side of the operated disconnector, thus the VFTO is superimposed on
the 50 Hz sine wave, see Figure 4.5.39.
Max VFTO = 2.1984 * 691254 =
= 1.5 MV = 1.69 p.u.
2
1.5
1
0.5
0
-0.5
-1
-1.5
-2
-0.02
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
Figure 4.5.39 VFTO waveforms measured by SGCC in SHIKY layout sensor 3; closing
operation, max VFTO (second spark, TCV = -0.89)
Comparison between simulation and measurement VFTO values for SHIKY set-up are given in Table 4.5.16. There
is a full agreement between measurements and simulations in sensor M2. Also simulation of VFTO value at
position M3 is in good agreement. This proves that the set-up was modeled correctly, in terms of surge impedance,
as based on the set-up elements dimensions, speed of electromagnetic wave, as based on the state-of-the-art
value and layout arrangement, as based on the layout drawing.
Page 151
TypethenameofTB
VFTO measured
VFTO calculated
with use of nonlinear BDV
Closing with TCV = 0.89 p.u.
Closing with TCV = 0.89 p.u.
M2
1.63 MV = 1.82 p.u.
1.63 MV = 1.82 p.u.
M3
1.5 MV = 1.69 p.u.
1.46 MV = 1.63 p.u.
Measuring point
Table 4.5.16 VFTO measured and calculated with use of nonlinear BDV
For verification the maximum VFTO peak values were calculated for the different measuring points of the SHIKY
test set-up. For the simulation the different BDV characteristics as well as the different trapped charge values are
used, summarized in Table 4.5.17.
nonlinear BDV
linear BDV
TCV distribution %
50%
98%
Worst case
50%
98%
Worst case
TCV p.u.
0.23
0.42
-1.0
0.50
0.92
-1.0
VFTO p.u. M1
1.82
1.94
2.34
2.00
2.28
2.34
VFTO p.u. M2
1.53
1.61
1.86
1.65
1.84
1.86
VFTO p.u. M3
1.83
1.95
2.35
2.02
2.29
2.35
VFTO p.u. M4
1.40
1.47
1.66
1.49
1.64
1.66
Table 4.5.17 Simulation results for max VFTO in SHIKY set-up

The VFTOs are lower for lower trapped charge voltages. VFTOs are lower for TCVs computed with nonlinear BDV
than linear BDV. There is a good agreement between VFTO simulations and measurements at two sensors
provided. Similar results are obtained using the PINGGAO set-up, using the same slow acting disconnector (see
Table 4.5.18).
nonlinear BDV
linear BDV
TCV distribution %
50%
98%
Worst case
50%
98%
Worst case
TCV p.u.
-0.23
-0.42
-1.0
-0.49
-0.92
-1.0
VFTO p.u. M1
1.87
2.00
2.42
2.06
2.37
2.42
VFTO p.u. M2
1.45
1.52
1.73
1.55
1.70
1.73
VFTO p.u. M3
1.94
2.08
2.53
2.14
2.47
2.53
VFTO p.u. M4
1.52
1.60
1.85
1.64
1.82
1.85
Table 4.5.18 Simulation results for max VFTO in PINGGAO set-up
Page 152
TypethenameofTB
The Layout of test set-up has negligible influence on TCV distributions. The highest impact parameter is the
disconnector design. This is by means of the disconnector Breakdown Voltage Characteristics: contact speed,
contact system design and SF6 pressure.
4.5.3.6 CONCLUSIONS
Normally the VFTO during closing is higher compared to the VFTO generated during opening operation. Therefore
the closing operation is more important, especially because of the trapped charge voltages as a result of the last
spark from the last opening operation.
The nonlinear BDV approach leads to the Trapped Charge Voltage (TCV) distributions with significantly lower
mean values than in the case of when linear BDV is applied. This in consequence leads to significantly lower VFTO.
By comparison of TCV simulations with TCV measurements, these results are in agreement in terms of the
distribution shape: both simulations and measurements have same Gaussian shape of probability function. The
prove of approach for BDV modeling is given by the comparison of sparking time distribution calculated with the
use of nonlinear BDV and the sparking time values measured. Here, good agreement was achieved for the
modified
The cumulative Trapped Charge Voltage distributions lead to the conclusion, that the value of -1 p.u. TCV is a
conservative assumption. The assumption of a more realistic value of the TCV, can lead to more realistic insulation
co-ordination calculations and in consequence allow for reducing the total cost of UHV substations. This can be
achieved by the reduction of the clearances and hence the size of the substations when the lower VFTO levels are
considered.
The application of slow operating disconnector doesnt produce high VFTO. It has to be mentioned however, that
such a speed reduction would lead to the increase of both sparking time and the number of sparks during the
disconnector operation process. The optimum design of the disconnector should be considered to achieve
significant VFTO reduction with acceptable sparking time and number of sparks.
4.5.3 INVESTIGATION FOR DS SURGE IN JAPAN

Concerning the disconnecting surge in UHV GIS, various investigations based on some analysis examples and
experimental data has been carried out in Japan.
4.5.3.1 ANALYSIS RESULT FOR ACTUAL SUBSTATION

Figure 4.5.40 shows the single line diagram of UHV GIS used for the VFTO calculation.
The substation has a double busbar scheme and consists of a GIS with four feeder bays and four transformer bays.
The disconnecting surges were calculated at various points in this GIS [7].
The maximum voltage across the contacts before restriking was set to be 2p.u. based on many observed data
among hundred operations in DS surge test of field verification by using 1000kV proto type facility [8]. Table 4.5.19
shows the conditions used in the calculation. Some waveforms in the calculation results, those are for no
resistance, 200 and 1k in DS (bus-tie DS) are shown in Figure 4.5.41 and Figure 4.5.42.
Page 153
TypethenameofTB
Figure 4.5.40 1100kV GIS circuitry used for the surge calculation
GIS
Transformer
ZnO surge arrester
Overhead transmission line
Bushing
Opened circuit breaker
DS resistance
Calculation step time
Z = 95 , = 270 m/s
4600 pF
V10kA = 1550kV, V1mA = 1080kV
Z = 230 , = 300 m/s
500 pF
400 pF
50 - 1 k
1ns
Table 4.5.19 Constants used in 1100kV GIS DS surge calculation
Page 154
TypethenameofTB
Figure 4.5.41 Results of calculation
Figure 4.5.42 Some waveforms from surge

calculation
(In each case of calculation, the

maximum values at various points
of GIS are plotted.)
The results of surge calculation for DSs at various points in Figure 4.5.40 are shown in Figure 4.5.41 (a), (b) and (c).
They show the followings:
With no resistor in the DSs, the crest value of overvoltages is 2.8p.u. (2510kV) and exceeds LIWV
2250kV (2.5p.u.).
f the resistance of the DS resistors is 200 or above, overvoltages can be suppressed below about
1.5p.u. When the resistance exceeds 500, they saturate approximately 1.2p.u.
The larger the resistance, the higher the voltage applied to the resistor. Dielectric capability of resistance
is required to be 1700kV when the resistance is 500.
Page 155
TypethenameofTB
The larger the resistance, the larger the energy consumed by the resistor, in the resistance range below
200, beyond which the energy eases off. Again, the larger the capacitance at the load side of the DS,
the larger the energy. The resistor energy shown in Figure 4.5.19 (c) was max when the longest bus
(35m) was used at the load side of the DS.
DS with resistor was adopted to reduce the disconnecting surge below LIWV level. To suppress the overvoltage as
low as possible and to reduce the dielectric and thermal capability of resistor, the value of resistor was determined
to be 500.
4.5.3.2 MEASUREMENT OF DS SURGE IN UHV FIELD TEST FACILITY

DS surge test was carried out using the UHV equipment test facility as shown in Figure 4.5.43 [8]. DS employs a
resistor of 500 to reduce the overvoltage of DS surge.
Figure 4.5.43 Switching operations for respective measurements

Figure 4.5.44 shows an observed waveform among hundreds operations. The maximum voltage across contacts
before restrike was 2p.u. There were neither voltage overshoots nor fast-rising oscillating waves, though these are
the general features of DS Restriking surge. The maximum surge observed during measurement was 1.20p.u.
Energy absorbed by the resistor was less than 20kJ (estimated value). The resistor switching is working effectively.
(a) Voltage waveform at closing operation of DS
(b) Enlarged waveform of part T
Figure 4.5.44 DS surge waveforms

EMI on secondary low voltage circuits was examined when main-circuit surges transferred through current
transformers (CT). Figure 4.5.45 shows the voltage waveform observed at the secondary circuits of CT at both
sides of the GCB. Figure 4.5.45 (a) shows main circuit surge, enlarged waveform of part A in Figure 4.5.44 (b).
Figure 4.5.45 (b) shows a voltage of secondary circuit at CT1 on the operating DS side of the GCB. It was
observed that only high-frequency component on main circuit affected the CT secondary circuit. Though the surge
amplitude of CT terminal exceeded 7kVp and the front shape was steep, the insulation performance of CT was
enough from the view of the V-t characteristics. Figure 4.5.45 (c) and (d) show voltage of secondary circuit of CT3
installed at the load side of the GCB. Figure 4.5.45 (c) is at the CT terminal; Figure 4.5.45 (d) at the interface panel
terminal. The highest level observed at interface panel and relay panels was at most 200Vp, and it was low enough.
Even over 7kVp were observed at CT terminals, this surge rapidly attenuated as travel in secondary cable.
Page 156
TypethenameofTB
(a) Main circuit (DS1 load side)
(c) Secondary circuit (CT3 terminal)
(b) Secondary circuit (CT1 terminal)
(d) Secondary circuit (CT3 Interface panel)
Figure 4.5.45 DS surge waveforms at CT secondary circuit

The resistor is not only working effectively to reduce the overvoltage in main circuits of GIS, but also secondary low
voltage circuits such as CT terminal and CT interface panel.
4.5.3.3 EVALUATION OF DIELECTRIC PERFORMANCE AGAINST DS SURGE FOR UHV

TRANSFORMER [8]
As transformer is directly connected to GIS, DS surge propagate the transformer terminal & windings. Depending
on the frequently components for intruding surge, local resonance may be generated inside the winding. Figure
4.5.46 shows DS surge measurement circuit estimate the voltage change for DS surge.
Figure 4.5.46 DS surge measuring circuit

Figure 4.5.47 shows voltage waveforms at the line terminal (X) with and without the resistor for the DS. The
waveforms almost match the analytical results by Electro Magnetic Transients Program (EMTP). The voltage
change at the wave-front was 1.35 p.u. without the resistor but 0.3 p.u. with the resistor, less than 1/4.
Page 157
TypethenameofTB
Without
resistor
1.0
0
-1.0
With
resistor
(Measured)
(Analyzed)
Figure 4.5.47 DS surge waveforms at transformer terminal (X)

Table 4.5.20 shows measured waveforms with and without the bushing under conditions without a resistor. When
DS surge invade into the bushing, a wave-front time of about 10 ns is extended to about 60 ns. This reduces the
overvoltage generated between turns or sections to 60 to 80% of that when a DS surge is applied directly without a
bushing.
Table 4.5.16 shows voltages generated in the winding (near the high-voltage line terminal). Compared with the
standard impulse voltage, a DS surge causes non-uniformity of potential sharing characteristics inside the winding
and generates higher voltages between turns or sections. This is because the DS surge wave-front time is shorter
than the capacitance charging time with the winding near the line terminal.
Page 158
TypethenameofTB
Without bushing
With busing
High-voltage line
terminal ground
Between first and

second sections
Table 4.5.20 DS surge waveforms in winding (without resistor)

DS surge (without resistor) (1)
Measurement point
Between turns of first
Section
Between first and second
sections
Between second and third
sections
Note:
Without bushing
With bushing
Lightning impulse
voltage(2)
(1.2/50 s)
23.1%
15.3%
2.3%
37.3%
23.7%
4.5%
16.8%
13.4%
4.5%
(1) DS surge is when the SW is turned on at 1 p.u. where 1 p.u. is 100%.

(2) Lightning impulse voltage is when the applied voltage is 100%.
Table 4.5.21 Voltage generated inside winding (near high-voltage line terminal)
Table 4.5.22 gives the calculated voltages based on Table 4.5.21 when a DS surge of (2/3)*1,100kV restrike and
lightning impulse voltage of 1,950kV are applied. These are rough estimates by conversion to the standard wave.
The values indicate that a voltage equivalent to or higher than LIWV may frequently be generated locally. Thus, it
would be preferable to adopt a DS with resistor for securing insulation reliability.
An intruding DS surge may also cause local increase of the voltage distribution in winding-type equipment such as
a voltage transformer (VT) or a bushing with internal shield for field control. Therefore, a DS without resistor would
be better to be used with careful attention given to DS surges, as with a transformer.
Page 159
TypethenameofTB
Measurement
point
Between turns of
first Section
Between first and
second sections
Between second
and third sections
DS surge (with resistor)

Without bushing (1)
With bushing (2)
Lightning impulse
voltage
(1.2/50 s)
2/ 3*1100kV*0.153/1.4 = 137kV
34kV
1950*0.023 = 45kV
2/ 3*1100kV*0.237/2.3 = 92kV
23kV
1950*0.045 = 88kV
2/ 3*1100kV*0.134/2.3 = 52kV
13kV
1950*0.045 = 88kV
Note: (1) Breakdown voltage of a steep wave having a pulse width of 70 ns was 1.4 times higher than that of the standard
wave between turns and 2.3 times higher at an oil gap [1] and their multiples were used for conversion to the
standard wave.
(2) DS surge with resistor was set to 1/4 of that without resistor.
Table 4.5.22 Evaluation of voltages generated in winding (near high-voltage line

terminal)
4.5.4 CONCLUSION
VFTO stress in GIS and Hybrid-IS due to DS switching could be concluded as follows:
1) VFTO is the abbreviation of very fast transient overvoltage defined in IEC 60071-1 (2006). VFTO is transient
overvoltage, usually unidirectional with time to peak Tf<0.1us, and with or without superimposed oscillations at
frequency 30kHz<f<100MHz.
2) As system voltage increases, it is a crucial issue to consider VFTO particularly in UHV application. The stress of
VFTO becomes severer compared with lower voltage classes.
3) The cumulative Trapped Charge Voltage distributions lead to the conclusion, that the value of -1 p.u. TCV is a
co-ordination calculations.
4) The application of slow operating disconnector doesnt produce high VFTO, but increase of both sparking time
and the number of sparks during the disconnector operation process. The optimum design of the disconnector
should be considered to achieve significant VFTO reduction with acceptable sparking time and number of sparks.
5) Damping resistor installed in DS can significantly suppress the amplitude and gradient of VFTO caused by
switching of DS. Application of DS with damping resistor or no use of damping resistor should be decided,
according to VFTO simulation considering the substation layout of GIS/Hybrid-IS. Regarding with neighbouring
equipment, the influence should be considered.
BIBLIOGRAPHY
[1] Ji Jiaqin. Analysis of very fast transient overvoltage in GIS. Journal of tsinghua univeristy, Vol.34 No.1 1994,
page 73~82.
[2] D. Povh, H. Schmitt, O. Vlcker and R. Witzmann. Modelling and analysis gridelines for very fast transients.
IEEE Transactions on power delivery, Vol.11, No.4, October 1996, page 2028-20
[3] Yanabu, S., Murase, H., et.al. Estimation of fast transient overvoltage in gas-insulated substation. IEEE
Transactions on Power Delivery. Vol.5, No.4, 1990.
[4] The Tokyo Electric Power Company, Inc. DS Overvoltage Analysis.[Technical Report] 2005.
Page 160
TypethenameofTB
[5] Du Shuchun, Zhang Cuixia, Ge Dong. Lightning protection for 1000kV UHV AC substation and switching station
[Technical Report]. China Electric Power Research Institute, Beijing. 2006.
[6] Ban Liangeng, Lin Jiming, Xiang Zutao, et.al. Very fast transient overvoltage analysis for GIS and HGIS of
China 1000kV UHV Pilot Project [Technical Report]. China Electric Power Research Institute, Beijing. 2006.
[7] Y. Yamagata, et al. Suppression of VFT in 1100kV GIS by adopting resistor fitted DS IEEE Transactions on
Power Delivery, Vol. 11, No. 2. April 1995
[8] YAMAGATA, Y.; OKABE, S.: Utility's experience on design and testing for UHV equipment in Japan, The
second IEC CIGR International Symposium on International Standards for UHV Transmission, 29-30 January
2009, New Delhi, India, proceedings pp. 142-154
[9] CIGR Working Group A3.22: Background of Technical Specifications for Substation Equipment Exceeding
800 kV AC, Brochure , to be published in 2010
[10] CIGR Working Group 15.03: GIS Insulation Properties in Case of VFT and DC Stress, Report 15-201, 36th
CIGR Session, Aug. 25 to 31, 1996, Paris, France
[11] CIGR Working Group 33/13-09: Monograph on GIS Very Fast Transients, Brochure 35, July 1989
[12] Riechert, U.; Holaus, W.; Krsi, U.; Sologuren, D. Gas-Insulated Switchgear for 1100 kV - Challenges in
Development and Testing, CIGR 6th Southern Africa Regional Conference, Colloquium of CIGR SC A2/A3/B3,
2009, Somerset West, 17 - 21 August 2009, South Africa
[13] Riechert, U.; Krsi, U.; Sologuren-Sanchez, D.: Very Fast Transient Overvoltages during Switching of BusCharging Currents by 1100 kV Disconnector, CIGR Report A3-107, 43rd CIGR Session, August 22-27, 2010,
Palais des Congrs, Paris, France
[14] Riechert, U.; Holaus, W.: Ultra High Voltage Gas-Insulated Switchgear A Technology Milestone, Euro.
Trans. Electr. Power (2011), (wileyonlinelibrary.com). DOI: 10.1002/etep.582
[15] Szewczyk, M.; Stosur, M.; Piasecki, W.; Florkowski, M.; Fulczyk, M.; Steiger, M.; Kostovic, J.: New
disconnector model for Very Fast Transient studies in High Voltage Gas Insulated Substations, in Proc. 2010
EEUG Conf., pp.154-163
[16] Szewczyk, M.; Piasecki, W.; Stosur, M.; Riechert, U.; Kostovic, J.: Impact of Disconnector Design on Very
Fast Transient Overvoltages in Gas-Insulated UHV Switchgear, Proceedings of 17th International Symposium on
High Voltage Engineering (ISH), August 22nd 26th, Hannover, Germany
[17] Riechert, U.; Neumann, C.; Hama, H.; Okabe, S.; Schichler, U., on behalf of CIGR WG D1.36 and AG D1.03:
Basic Information and Possible Counter Measures Concerning Very Fast Transients in Gas-Insulated UHV
Substations as Basis for the Insulation Co-ordination, CIGR SC A2 & D1 Joint Colloquium 2011, Kyoto, Japan,
PS3-O-5
[18] M. Szewczyk, M. Stosur, W. Piasecki, M. Kuniewski, P. Balcerek, M. Florkowski, Uwe Riechert:
"Measurements and Simulations of Very Fast Transients during Disconnector Type Testing in UHV Gas-Insulated
Switchgear, Electrotechnical Review" (accepted for publication)
Page 161
TypethenameofTB
4.6 Influence of metal oxide surge arresters on circuit breaker TRVs

4.6.1 GENERAL
Circuit breaker TRVs have evolved over the years and are standardized by IEC, generally based on work done
within Cigre environment. This standardization is a continuing process as more knowledge is acquired and power
systems grow in complexity.
Standard TRVs are inherent test circuit values to be achieved without regard to either possible circuit breaker and
test circuit interaction or network component influence. One such component is of course the metal oxide surge
arrester (MOSA). As AC system voltages have increased from HV through EHV (345 kV to 800 kV) to UHV (1100
kV or 1200 kV), the ratio between insulation levels and system voltages have decreased and, likewise, arrester
protective levels have decreased. The intent of this section is to examine the influence of MOSAs on TRVs and the
possible merits of consideration in UHV circuit breaker application for mitigating or limiting TRVs.
4.6.2 MOSA TRV LIMITATION

The notion of MOSA TRV limitation is not new. MOSAs are applied on EHV line terminals where the lines are
series compensated with MOV protected series capacitor banks for the purpose of limiting the TRV seen by the line
circuit breaker [1]. On the Turkish 400 kV, this limitation was achieved by connecting the MOSA across the line
circuit breakers [2].
For general purpose circuit breaker applications, the possible influence of MOSAs (and other network components)
is treated as having added value rather than fundamental value. Figures 4.6.1 to 4.6.4 show a comparison of IEC
TRV values to typical applied arrester switching impulse protective levels (SIPL) for circuit breakers rated at 500 kV
to 1200 kV.
Figure 4.6.1 500 kV circuit breaker TRVs and 396 kV rated surge arrester
Page 162
TypethenameofTB
Page 163
TypethenameofTB
Figure 4.6.4 1200 kV circuit breaker TRV and 850 kV rated surge arrester
Review of Figures 4.6.1 to 4.6.4 show that the MOSAs will limit the peak values for T60 and below at 800 kV and
for T100 and below at 1100 kV and 1200 kV. The results of simulation studies for the 1100 kV case are shown in
Figures 4.6.5 and 4.6.6[3-5]. The limiting effect of the MOSA is evident.
Figure 5.6.5 Simulation of MOSA influence on T100 and T60 TRV values [4]
Page 164
TypethenameofTB
2000
TLF
L L F (T 3 0 )
T60
T100
TRV peak in 1100 kV system with MOSA
1800
1600
1478kV
1411kV
1400
1200
1000
800
600
400
200
0
0
200
400
600
800
1000 1200 1400 1600 1800 2000
T R V p e a k in 1 1 0 0 k V s y s te m w ith o u t M O S A
Source: Cigre WG A3.22 Second Technical Brochure
Figure 4.6.6 Simulation of MOSA influence on T100, T60, T30 (LLF - longline fault)
and TLF (transformer limited fault)
Note: The TRV value is the voltage difference between the circuit breaker terminals in each phase after the current interruption,
while the MOSA SIPL is the arrester residual voltage phase to ground. MOSAs are applied in their usual role in power systems,
i.e., to ensure lightning and switching overvoltage protection of the apparatus in their vicinity. Therefore the MOSA are
considered as devices connected from each phase to ground, on one or both sides of circuit breaker.
4.6.3 INFLUENCE OF MOSAS ON TRVS AT UHV

The application of MOSAs, especially for UHV lines, can reduce the amplitude factors of TRVs for terminal faults,
TLF (transformer limited or secondary faults) and out-of-phase. However, it may not lead to a reduction large
enough for LLF (long line faults) where TRV is generated at both the source and line sides of the breaker terminals.
This phenomenon are not observed in transmission systems up to 800 kV. The purpose of these MOSA
applications is in quite normal cases in order to suppress the lightning and switching overvoltages to the specified
LIWV/SIWV levels and not to limit the TRV across the circuit breaker.
On the other hand, MOSA applications on lines with series capacitors are special cases intended to limit the TRV
across the circuit breaker, because the TRVs after clearing a fault may exceed the standard values if a series
capacitor in not bypassed by a triggering gap or bypass switch operation. In this case, the MOSAs are applied
Page 165
TypethenameofTB
across the circuit breaker rather than to ground. Note that the use of this possibility may be limited by system outof-phase switching considerations.
The MOSA applications for two examples have completely different purposes so it is recommend that they should
deal with different technologies as shown in Table 4.6.1.
Turkey 420 kV
series compensated line
All ratings
TRV limitation
Applications
Voltage
Purpose of MOSA
Installation of MOSA
Influence on TRV
Across GCB terminals

Special application
Limit TRV to the protection level
Compensation scheme
Series capacitor
1100/1200kV transmission lines

Only for UHV
Suppression of lightning overvoltage
and switching overvoltage (not
intended to suppress TRV)
Line terminal to the ground
General application
Limit switching overvoltage to the
ground if the value exceed the SIPL
All schemes
Table 4.6.1 Technical comparison

Figures 4.6.1 to 4.6.4 show the influence of MOSAs on TRVs applied to the lines with series capacitors or that on
TRVs for terminal faults. Since the MOSA applications in the UHV systems can suppress the overvoltage at line
and source sides to the protection level from a ground potential (not a voltage across the circuit breaker). The TRV
for a line fault with line of 40 km, which waveform is similar to that the case of the LLF duty, is provided below. In
this case, MOSA clips the source side TRV.
2000
Similar TRV for 40km line fault to LLF: Breaking current 9.6 kA (T30)
Standard TRV-Uc:1660kV
TRV across CB
Voltage(kV)
1500
Clipping level of MOSA
1000
Source side TRV without MOSA

Source side TRV with MOSA (B-character)
Source side TRV with MOSA (C-character)
Line side TRV with and without MOSA
500
Time(ms)
Figure 4.6.7 Influence of MOSA on TRV in the UHV system, where MOSA is applied
to suppress the lightning and switching overvoltages
Page 166
TypethenameofTB
4.6.4 CONCLUSIONS
The value of MOSAs in limiting TRVs, particularly for UHV systems, is evident and the following can be concluded:
The influence of MOSAs on mitigating or limiting circuit breaker TRV peak values has to date been viewed as
latent hidden value.
Consideration of this influence has considerable merit but generalization is difficult because MOSA ratings, and
hence protective levels, are system-specific being dependent on expected TOV magnitudes and durations. Also,
TOVs characteristics may vary over different parts of the system.
Users should recognize the technical advantage of TRV peak limitation and the associated economic value that
the approach offers by allowing the use of circuit breakers with fewer interrupters in series.
If the MOSA TRV limiting approach is pursued, then the user is advised to ensure that the actual application
TRVs are covered by available circuit breaker type test reports. The most recent amendment to the IEC circuit
breaker standard addresses the TRV requirements for 1100 and 1200 kV circuit and the calculation of the test
TRVs as influenced by opening resistors [6]. The additional influence of MOSAs can be calculated in a similar
manner.
REFERENCES
[1] B.L. Avent, D.F. Peelo, J.H. Sawada, Circuit Breaker TRV Requirements for a Series Compensated 500 kV
Line with MOV Protected Series Capacitors. Colloquium of Cigre Study Committee 13, Florianopolos, Brazil,
September 1995.
[2] F.M. Gatta, F. Illiceto, S. Lauria, B. Dilli, TRVs Across Circuit Breakers of Series Compensated Lines. Analysis,
Design and Operational Experience in the 420 kV Turkish Grid. Cigre Paper No. 13 109, Session 2002.
[3] Cigre Technical Brochure 362 Technical Requirements for Substation Equipment Exceeding 800 kV.
December 2008.
[4] Cigre Technical Brochure 456 WG A3.22, Background of Technical Specifications for Substation Equipment
Exceeding 800 kV AC, April 2011, ISBN: 978-85873-145-9.
[5] A.L.J. Janssen, Y. Yamagata, D. Dufournet, H. Ito, M.C. Bhatnagar, M. Kosakada, Y. Filion, J. Amon and P.
Fernandez, UHV Equipment Requirements: State of the Art and Prospects for Equipment, Paper No. 4-1, New
Delhi 2009.
[6] IEC 62271-100A1: Amendment 1 High voltage switchgear and controlgear Part 100: Alternating current
circuit-breakers (Addition of requirements and tests for 1100 and 1200 kV)
Page 167
TypethenameofTB
4.7 Conclusions
UHV clearance in the air may require longer distance than those applied to EHV system, because the dielectric
strength does not increase linearly with an increase of the air gap. Accordingly, optimal insulation co-ordination
becomes extremely important for UHV systems to achieve a compact, economic and environmental-friendly UHV
transmission. The precise calculation methods, advanced simulation and mitigation techniques for the overvoltages
in the system can achieve the reduction of the dielectric requirements. The analytical tools like EMTP are very
useful for the precise evaluation of the overvoltages. They can optimize the insulation co-ordination design for the
power network systems and rationalize the test voltage for the power equipment.
The amplitude and duration of temporary overvoltage occurring in the UHV systems are calculated. The amplitude
of temporary overvoltage levels and required energy absorption of surge arrestor in UHV systems are relatively
higher than those in EHV systems. The waveforms of switching overvoltage occurring in UHV transmission lines
are estimated and their front times are more than several hundred microseconds that is bigger than those in EHV
systems. As for lightning current, its waveform is investigated in actual UHV transmission lines and the ramped
shape waveform is used for the overvoltage analysis. The lightning currents more than 200 kA for back flashover
calculation are specified in the over 800 kV systems and the lightning overvoltages are estimated and lightning
currents for shield failure calculation are determined by EGM and other methods according to tower configuration.
In GIS or Hybrid-IS system, the switching of disconnecting switch will produce very fast transient overvoltage
(VFTO). VFTO has very high gradient and high frequency components. The serious VFTO can damage the
insulation of GIS and the interturn insulation of transformer, and cause transient enclosure voltage (TEV) and
electromagnetic interference (EMI) to low voltage control systems.Here, VFTO was investigated in detail based on
Chinese and Japanese data.The applications of the suppression technologies to reduce these overvoltages are
adopted in over 800kV systems. They contribute the downsizing of their power equipments and improve the
reliability of the system.
For 800 kV and the above systems, the representative values of overvoltages, their waveforms and occurrence
conditions have been practically investigated for the circuits of each country. The overvoltage characteristics
reported by each country are summarized as follows.
1) In UHV systems of various countries, representative levels of temporary overvoltages and switching
overvoltages are in the range from 1.3 to 1.5 p.u. and from 1.6 to 1.7 p.u., respectively. Here, the voltage of 1 p.u.
is defined as the peak value of phase to the earth of operating voltage.
2) The energy absorption of surge arrester for temporary overvoltages of UHV system is specified from 40 to 55 MJ,
which are bigger than those for 800 kV and EHV systems.
3) The methods such as CB with opening and/or closing resistor, shunt reactor and appropriate MOSA installation
are verified to be effective to mitigate the system overvoltages.
4) The front times of the switching overvoltages in the UHV systems, which are calculated by the simulations, tend
to be more than several hundred micro seconds. These are longer than those of the standard SI test waveform
defined in IEC 60060-1.
5) The front time duration of non-standardized waveform strongly depends on how to convert the waveform to the
standard test waveform. Accordingly, the clear understanding of the conversion method and the definition for the
front time is required.
6) The simulation results show that the CB with opening and/or closing resistor can reduce the maximum switching
overvoltage by 10% and more. MOSA installed at the line entrance is effective to reduce the switching overvoltage
levels of the substation and transmission line.
7) In Chinese system, the shunt reactors installed at the line side is special effectively to limit the TOV and
switching overvoltage for long transmission line. In addition, they have other functions such as balancing system
Page 168
TypethenameofTB
real power, suppressing the second arc current, preventing self-excitation of generators and reducing the MOSA
rated voltage (reducing SIPL and LIPL as well).
8) Regarding lightning overvoltage, the simulation method and the circuit conditions are studied by using EMTP
analysis. The results show that the lightning current waveform is much influenced by the lightning overvoltages.
The maximum lightning current more than 200 kA is generally used in the 800 kV and the above systems for back
flashover calculation and lightning currents for shield failure calculation are determined by EGM and other methods
according to tower configuration.
9) The recent investigation regarding lightning current waveform makes clear the relationship between the current
peak value and the front time duration. However, the lightning current waveform and its maximum value should be
defined based on the actual phenomena.
10) The ratios of representative overvoltages to LIPL for transformers in UHV systems are in the range from 1.0 to
1.17 p.u. Those for other equipment are from 1.0 to 1.36 p.u. In 800 kV systems, the ratios for transformers and
other equipment are in the range from 0.88 p.u. to 1.31 p.u. and from 1.15 p.u. to 1.55 p.u., respectively. The ratios
for UHV systems are likely to be smaller than those for 800 kV and the lower voltage systems, because the
overvoltage mitigation techniques including higher performance arrestors can drastically reduce the lightning
overvoltage levels.
11) The cumulative Trapped Charge Voltage distributions lead to the conclusion, that the value of -1 p.u. TCV is a
co-ordination calculations.
12) The application of slow operating disconnector doesnt produce high VFTO, but increase of both sparking time
and the number of sparks during the disconnector operation process. The optimum design of the disconnector
should be considered to achieve significant VFTO reduction with acceptable sparking time and number of sparks.
13) Damping resistor installed in DS can significantly suppress the amplitude and gradient of VFTO caused by
according to VFTO simulation considering the substation layout of GIS/Hybrid-IS. Regarding with neighbouring
14) MOSA is effective to suppress the overvoltage of phase to the earth. In UHV systems, MOSA can also improve
the peak voltages of TRV in case of terminal faults and TLF.
Furthermore, this chapter summarizes the detailed circuit conditions when the maximum overvoltages occur and
shows the statistical evaluation of overvoltage occurrence. The useful information in this chapter contributes to the
precise overvoltage evaluation of UHV systems. Also, these advanced analytical techniques and the knowledge
are recommended to be added to IEC60071 Insulation co-ordination.
Page 169
TypethenameofTB
5 Evaluation of overvoltage study results

5.1 Overvoltage simulation tools and verification by measuring results
5.1.1 INTRODUCTION
This chapter investigates the simulation method and the accuracy comparison between simulated and measured
overvoltages. In the simulation of switching overvoltages, the TNA was used in 1960s to 1970s, and digital
simulations such as EMTP and PSCAD/EMTDC have been used since 1980s. For the simulation of lightning
overvoltages or very fast transient, the digital simulation is used in general. The accuracy of simulation is mainly
dependent on the simulation conditions, which is not consistent with the measured one. Comparison between
simulated waveforms and measured waveforms of switching overvoltages are investigated for UHV transmission
lines, 765 kV transmission line, 500 kV line, 400 kV line, and 220 kV line. Comparison between simulated
waveforms and measured waveforms of lightning overvoltages are investigated only for one case in UHV line of
Japan. There are many simulation studies on the lightning overvoltage calculation, but measurement was very rare.
For the VFTO study in GIS, UHV, 765 kV and 500 kV case were reviewed.
5.1.2.1 TEPCO (T OKYO E LECTRIC P OWER C OMPANY ) [1-2]

Comparison between simulated waveforms and measured waveforms was carried out for switching overvoltages in
the UHV transmission line shown in Figure 5.1.1. The line, which is approximately 200km long, is designed for
1100kV and tentatively operating at 550kV. Simulated values by EMTP agree with measured ones by a lightning
observation device at a substation very well when switching at the same phase angle. Inaccuracy of analysis can
be neglected.
200km, closing R=1,000

Higashi-Gunma
M inami-Iwaki S/S
s/s
Actual measurement
1.61 [p.u.] 3%
(with measurement error)
EMTP Analysis
1.58 [p.u.]
Figure 5.1.1 Switching overvoltages measurement on actual UHV designed

transmission line
Page 170
TypethenameofTB
5.1.2.2 SGCC (S TATE G RID C ORPORATION
OF
C HINA ) [3]
Comparison between measured waveform and simulated waveform, shown in Figure 5.1.2 (a) and Figure 5.1.2 (b)
respectively, was carried out for switching overvoltage in the UHV transmission line shown in Figure 5.1.2. The line
is approximately 280km long. The simulated overvoltage waveforms are consistent with the measured ones. In the
calculation example, the maximum overvoltage is 959.9 kV, which is consistent with the measured result of
955.172 kV when switching at the same phase angle.
Figure 5.1.2 Measured and simulated waveforms at Nan Yang side during
energizing Nan-Jing TL from Jingmen (Maximum: (a) 955.2kV vs. (b) 959.9kV)
5.1.2.3 AEP (A MERICAN E LECTRIC P OWER ) [4-8]
Comparison between simulated waveform and measured waveform, shown in Figure 5.1.4(a), was carried out for
switching overvoltage in the 765kV transmission line shown in Figure 5.1.3. The line sections are 290km and
243km long. The simulated overvoltage waveforms by TNA and digital computer are 10 -15 % higher than
measured ones as shown Figure 5.1.4.
Page 171
TypethenameofTB
Figure 5.1.3 Switching diagram for 765kV field Tests
(a) Measured and simulated switching surges at Marysville (Energization from Dumont)
(b) Probability Distribution
(c) Measured waveform
Figure 5.1.4 Measured and simulated switching surges at 765kV line (290km,
closing)
Page 172
TypethenameofTB
5.1.2.4 SSPB (S WEDISH S TATE P OWER B OARD ) [9]

Comparison between simulated waveform and measured waveform, shown in Figure 5.1.5(b), was carried out for
switching overvoltage in the 400 kV transmission line shown in Figure 5.1.5 (a). The line is 349 km long. The
analyzed overvoltage waveforms by TNA and Digital computer are consistent with the measured ones.
(a) SSPB 400kV system.
(b) Comparison of field test and computer result
Figure 5.1.5 Comparison of field test and computer result [349 km, Reclosing A-B
in (a)]
5.1.2.5 EEA (T HE E GYPTIAN E LECTRICITY A UTHORITY ) [10]
respectively was carried out for switching overvoltage in the 500 kV and 220 kV transmission line which are 209 km
and 40 km long respectively and shown in Figure 5.1.6(c). The simulated overvoltage waveforms by EMTP are
consistent with the measured ones as shown in Figure 5.1.6 and Table 5.1.1.
Page 173
TypethenameofTB
(a) 220 kV and 500 kV systems
(b) Overvoltage waveforms
Figure 5.1.6 220kV line energization at no load (40 km)
Table 5.1.1 Measured and calculated overvoltage factors for 220kV line
5.1.3 LIGHTNING OVERVOLTAGE
5.1.3.1 TEPCO [11]

Comparison between measured waveform and simulated waveform, shown in Figure 5.1.7(b) and Figure 5.1.7(c)
respectively, was carried out for lightning stroke intrusion at the switching station in the UHV designed 550kV
transmission line shown in Figure 5.1.7(a). The front part of each waveform, which is important for the
characteristics of the wave, matches closely. There is a difference in the declining portion of the waveforms
because of an error due to the straight line of the wave tail of the lightning stroke current.
Page 174
TypethenameofTB
(a) UHV system
(b) Measured waveforms
(c) Simulated waveforms
Figure 5.1.7 Simulated and measured waveforms of lightning stroke intrusion

(MinamiIwaki substation)
5.1.4 VFTO
5.1.4.1 TEPCO [12]

respectively, was carried out for VFTO in UHV GIS. The initial part of the measured waveform agrees well with the
waveform obtained by the EMTP analysis. The disagreement of other parts between the experimental and
simulated results occurs because, in the EMTP analysis, only the first wave in the high-frequency circuit is used.
Page 175
TypethenameofTB
Figure 5.1.8 Comparison between measured and analyzed waveforms (VFTO

between k-terminal and E-terminal of CT in UHV-GIS under DS operation)
5.1.4.2 ESKOM [13]

respectively, was carried out for VFTO in 765 kV GIS Each waveform matches closely.
(a) Measured waveforms
(b) Calculated waveforms
Figure 5.1.9 Measured and calculated waveform of 765 kV substation Alpha
Page 176
TypethenameofTB
5.1.4.3 TEPCO [14]

Comparison between measured waveform and simulated waveform, shown in Figure 5.1.10(b) and Figure
5.1.10(c) respectively, was carried out for VFTO on 550kV GIS shown in Figure 5.1.10(a). Each waveform matches
closely.
a) Single-line diagram of 550kV GIS
b) Measured results
c) Simulated results
Figure 5.1.10 Measurement and simulation of Propagation Characteristics in 550kV

GIS
Page 177
TypethenameofTB
5.1.4.4 R ECENT S TUDY [15]

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 modeling 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 structure or test equipment. One such
model is proposed by Szewczyk, et al [16]. Figure 5.1.11 shows a comparison of simulated and measured VFTO.
The measured voltage progressions coincide very well with the simulation results as regards VFTO amplitude and
rise time. Nevertheless, variations of more than 10% depending on the calculation methods are reported. Therefore,
it is important to verify the simulation results by measurements.
Figure 5.1.11 VFTO calculation and measurement when switching busbars with a
GIS DS as per IEC 62271-102, without pre-charging (left), with pre-charging (right)
5.1.5 SUMMARY
(1) The accuracy comparison of switching overvoltage between simulations and measurement, the TNA and digital
simulation cases are 10 - 15 % higher than measured ones in early 1970s. In the case of UHV line in Japan and
China, the accuracy was in the range 1 - 5 %. So, the maximum error between measurements and simulation is
about 5 %, which is acceptable in the design of transmission lines.
(2) There are so many simulation studies, but both simulation and measurement case are rare. In the case of UHV
line in Japan, the measurement and simulation waveform are identical in magnitude but wave front time are slightly
different. The lightning waveform and peak amplitude has statistical distributions, so it is necessary to use the
cumulative distribution data than single waveform.
(3) An accurate modeling of each individual GIS component makes it possible to reproduce VFTO waveforms with
a relatively high precision (differences lower than 5%). Nevertheless, variations of more than 10% depending on
the calculation methods are reported. Therefore, it is important to verify the simulation results by measurements.
Page 178
TypethenameofTB
References
[1] E. Zaima, et al., Insulation Coordination for UHV AC Systems Based on Surge Arrester Application (CIGRE
C4.306), IEC-CIGRE Second International Symposium on Standards for Ultra High Voltage Transmission, New
Delhi, 2009
[2] N. Ryu, et al., Comparison between EMTP Waveform and Field Observed Waveform of Switching Surge in the
UHV Designed 500kV Transmission Line, Proc. of the Eleventh Annual Conference of Power & Energy Society,
IEE of Japan, No.380, Sapporo, Japan, Aug. 2000 (in Japanese)
[3] ZHENG Gin, et al., Comparative Analysis on Simulation and Test of Overvoltages in China UHV AC Pilot
Project, 2009 International Conference on UHV Transmission, Beijing, 2009
[4] Barnes, H. C. "Preliminary Analysis of Extensive Switching Surge Testing of American Electric Power's First 765
kV Line and Stations," IEEE Trans. Power Apparatus And Systems, vol. PAS-90, March 1971
[5] Scherer, H. N., Vassell, G. S., "Experience with the AEP 765 kV System, Part I: Overview, 1972 IEEE Winter
Power Meeting, New York, N.Y., Publication No. 72 CH0645-2 PWR.
[6] Haas, R. J., et al, "Experience with the AEP 765 kV System Part IV, Overvoltage and Staged Fault Tests:
Instrumentation and Control," IEEE Paper Submitted for Presentation at the 1972 Winter Power Meeting.
[7] Garrity T, et al, "Experience with the AEP 765 kV System Part IV, Overvoltage and Staged Fault Tests:
Analysis," IEEE Paper Submitted for Presentation at the 1972 Winter Power Meeting.
[8] Fakheri (Keri), A, J., Haahr, J, "Experience with the AEP 765 kV System-Dumont-Marys.-Kammer Field Tests,"
IEEE Trans. Power Apparatus And Systems, July/Aug, 77.
[9] G. Catenaccl The Calculation Of Switching Surges III: Transmission Line Representation For Energization And
Re-Energization Studies With Complex Feeding Networks, GIGRE Working Group 13.05, Electra, No. 62 (1079)
[10] M.M. Awad, et al The Egyptian Experience With New Technologies For Field Transient Overvoltage
Measurement, GIGRE 33.203, 1996
[11] J. Takami, et al., Characteristics of Direct Lightning Strokes to Phase Conductors of UHV Transmission Lines,
IEEE Trans. Power Del., vol. 22, no. 1, pp. 537-546, Jan. 2007
[12] Y. Yamagata, et al., Very Fast Transients in 1000 kV Gas Insulated Switchgear, IEEE Transmission and
Distribution Conference, pp. 501-508 vol.2, Apr. 1999.
[13] A. Eriksson, et al Studies of Very Fast Transient in a 765 kV substation, CIGRE Working Group 34.12., report
No. 35 Monograph On Gis Very Fast Transients pp. 43 ~ pp.48
[14] S. Ogawa, et al., Estimation Of Restriking Transient Overvoltage On Disconnecting Switch For GIS, IEEE
PES 1985 Summer Meeting, 7 Vancouver, 85 SM 367-8, 1985
[15] U. Riechert, et al., Basic Information and Possible Counter Measures Concerning Very Fast Transients in
Gas-Insulated UHV Substations as Basis for the Insulation Co-ordination, CIGRE SC A2 & D1 Joint Colloquium
2011, KYOTO JAPAN, PS3-O-5, 2011
[16] M. Szewczyk, et al., Impact Of Disconnector Design On Very Fast Transient Overvoltage In Gas-Insulated
UHV Switchgear, XVII International Symposium on High Voltage Engineering, No.B-013, 2011
Page 179
TypethenameofTB
5.2 Main Characteristics of Metal-Oxide Surge Arresters (MOSAs)

5.2.1 GENERAL
Regarding insulation coordination, it is important that overvoltage levels whose frequency ranges (hundreds of kHz)
from fast-front overvoltages to power frequency (50/60 Hz) voltages are reduced to rational levels to optimize the
size and costs of UHV system. The concept and specification values of insulation co-ordination for UHV systems
in Japan are shown in Figure 5.2.1. An insulation design of UHV systems can be optimised by incorporating GIS
type UHV higher performance MOSA, as developed in Japan.
The higher performance MOSAs have a higher protective performance due to their lower residual voltage level,
which is a decisive factor for UHV power transmission insulation design. The lightning impulse withstand voltage
(LIWV) level of UHV system could be remarkably reduced by installing higher performance MOSAs at adequate
points in the substation. In addition, by employing gas circuit breaker (GCB) using opening and closing resistors
and the MOSAs, switching surge level could be reduced from the range of 2.0 p.u. for 550 kV system to 1.6-1.7 p.u.
for UHV system voltage, where 1.0 p.u. equals peak value of maximum line-to-ground voltage).
Figure 5.2.1 Concept and specification values of insulation co-ordination for UHV
systems in Japan [1]
These results greatly contributed to optimize the reduction of size of UHV equipments and the economical design
of UHV power transmission insulation (reduction of steel tower size, ground area for substation, etc.).
Figure 5.2.2 shows an example of the insulation design optimization based on MOSAs.
In order to realize low LIWV level of UHV systems, the higher performance MOSAs with low residual voltage are
considered as essential, since their protective levels are remarkably lower compared with the conventional MOSAs
according to existing standard, and which have been developed by applying and comparing the two technologies,
as shown in Figure 5.2.3.
Page 180
TypethenameofTB
Figure 5.2.2 Insulation design optimization in Japan based on higher performance

MOSAs [1]
Note:
(*) MOSAs that have higher protective performance due to their lower residual voltage level, compared with conventional
MOSAs, are termed higher performance MOSAs.
(**) The validity of these values of LIWV based on higher performance MOSA has been confirmed by insulation co-ordination
analysis using EMTP program.
(***) The values of LIWV based on conventional MOSA are referred from the relevant existing standards.
Conventional
500kV MOSA
Higher performance 1000kV
Figure 5.2.3 Technologies to be applied to UHV higher performance MOSAs and

comparison of v i characteristics between conventional and higher performance
MOSA [1]
Page 181
TypethenameofTB
Main
performance
Protective performance
(Non-linearity)
Life
performance
Energy (TOV) durability
Requirements
1.55
90 %
300 J/cc
(55MJ as for MOSA)
Table 5.2.1 Improvement of MO element

The primary technology was improvement of Metal-Oxide (MO) element as indicated on Table 5.2.1 and the four
parallel connection of MO elements column which is another technology. However, the main issues upon
developing the MO elements are to secure the life performance, energy (TOV) durability and thermal stability of
them because of the remarkable reduction of its protective level (residual voltage) of the UHV higher performance
MOSAs.
5.2.2 MAIN CHARACTERISTICS OF MOSAS

Considering the following special requirements of MOSAs for the UHV system and issues of the existing IEC
standards for MOSAs, the main characteristics of MOSAs for the UHV pilot projects in the world have been
established as shown in Table 5.2.2, in which the main characteristics of MOSAs for the representative 800 kV
systems in the world have been covered for reference. The representative characteristics of GIS type MOSA,
porcelain-housed type MOSA and polymer-housed MOSA for UHV system have been investigated by the relevant
CIGRE Working Groups. Additionally, the studies for the UHV higher performance MOSAs have just been
commissioned by the IEC Technical Committee for surge arresters (TC37): maintenance team 4 (MT4).
The size of MOSAs for UHV systems, particularly AIS type UHV MOSAs, can become very huge, as shown in
Figure 5.2.4. Therefore, in addition to the general consideration concerning the protective level and energy
requirements of the MOSAs, the specific consideration concerning the insulation withstand requirements, and
mechanical requirements for the AIS type UHV MOSAs should be specified to enable proper and complete
evaluation of its performance. 5.2.2.1 Special requirements of UHV higher performance MOSAs
Protective level: It comes to the main issues for the UHV higher performance MOSAs with low protective level to
secure the long-life service performance, TOV durability and thermal stability.
Energy requirements: Energies for UHV higher performance MOSAs may significantly exceed the existing test
requirements as a result of realizing low protective levels of the MOSAs.
Insulation withstand requirements: For the MOSAs to be used for the systems exceeding 550 kV, the tests on the
housing of complete MOSAs are considered as essential under the realistic conditions as possible with the
MOSAs mounted on a pedestal.
Mechanical requirements: High reliability against earthquake for MOSAs mounted on pedestal is considered as
essential for AIS types UHV MOSAs because of their huge dimension.
Voltage distribution characteristic along the MO columns of MOSAs: This characteristic influences the life
performance and local heating of MOSAs, which is referred as the initial temperature in operating duty tests
Page 182
TypethenameofTB
UHV
China
Japan
(1) Max. system voltage (kVrms)
1100
1100
(2) Peak value of max. line to ground voltage (kVp)
898 (1.00)
898 (1.00)
for transformer
2250 (2.51)
1950 (2.17)
(3) LIWV of system (kVp) (*)
for other equipments
2400 (2.67)
2250 (2.51)
for transformer
1800 (2.00)
1425 (1.59)
(4) SIWV of system (kVp) (*)
1800 (2.00)
1550 (1.73)
GIS/AIS
GIS
(porcelain)
(5) Nominal discharge current (kAp)
20
20
20
(6) Rated voltage of surge arrester (kVrms)
850
828
826
(7) Continuous operating voltage (COV) (kVrms)
693, (MCOV: 723)
638
635
(8) LIPL at 20 kAp
1700 (1.73)
1620 (1.80)
1620 (1.80)
(9) DLR(*) (8)/(6)
2.00
1.96
1.96
(10) SIPL at 2 kAp
1500 (1.53)
1460 (1.64)
(11) TOV (MJ)
55
20
55
(12) 2 shots total energy at operating duty test (MJ)
10+10 (margin)
40
40
(13) Line discharge class as per IEC
5
5
(14) Insulation withstand of LIWV
2400 (2.45)
2400 (2.67)
2250 (2.51)
housing (kVp)
SIWV
1800 (1.84)
1800 (2.00)
1550 (1.73)
(15) Short circuit current rating (kArms)
40
50
50
IEC 60099-4
GB 50260-1996
JEAG 5003-1998
(16) Mechanical requirement
0.3 G
0.3 G resonant 5
0.3 G resonant 3
cycles beat wave
cycles wave
less than 1.05
< 1.10 (GIS type)
< 1.10
(17) Voltage distribution factor
< 1.15 (porcelain
type)
Note: (*) It is reported that LIPL and SIPL values are determined by the simplified method for the Indian UHV System. The
values of SIWV and LIWV are obtained considering adequate margin over the values of SIPL and LIPL as obtained from V-I
characteristics of the higher performance surge arrester.
Metal-Oxide Surge Arrester (MOSA)
System
India
1200
980 (1.00)
2250 (2.30)
2400 (2.45)
1800 (1.84)
1800 (1.84)
AIS (polymer)
800 kV
System
Canada
for transformer
(3) LIWV of system (kVp)
for transformer
(4) SIWV of system (kVp)
(8) LIPL at 20 kAp
(9) DLR(*) (8)/(6)
(10) SIPL at 2 kAp
(11) TOV (MJ)
housing (kVp)
SIWV
(a)
765
625 (1.00)
1950 (3.12)
2100 (3.36)
1425 (2.28)
1425 (2.28)
AIS (porcelain)
20
550
442
1340 (2.14)
2.44
1080 (1.73)
700 kV 1 sec.
8.0
2100 (3.36)
1425 (2.28)
50
sin wave 5
vibration
0.23 G
-
Page 183
(b)
765
625 (1.00)
1950 (3.12)
2100 (3.36)
1425 (2.28)
1425 (2.28)
AIS (porcelain)
20
612
442
1560 (2.50)
2.55
1200 (1.92)
700 kV 1 sec.
8.0
2100 (3.36)
1425 (2.28)
50
sin wave 5
vibration
0.23 G
-
Brazil
800
653 (1.00)
2100 (2.34)
1550 (2.37)
AIS (porcelain)
20
588
470
1440 (2.21)
2.44
1210 (1.85)
2.9 (5 kJ 588)
N.A.
5
2100 (3.22)
1550 (2.37)
50
IEC 6099-4
< 1.10
TypethenameofTB
800 kV
System
Korea
for transformer
(3) LIWV of system (kVp)
for transformer
(4) SIWV of system (kVp)
(8) LIPL at 20 kAp
(9) DLR(*) (8)/(6)
(10) SIPL at 2 kAp
(11) TOV (MJ)
housing (kVp)
SIWV

Note: (*) Discharge Level Ratio (DLR): (8)/(6)
800
653 (1.00)
2250 (3.45)
1500 (2.37)
1425 (2.18)
AIS (porcelain)
20
588
462
1400 (2.14)
2.38
1300 (1.99)
6.6
4
2250 (3.45)
1425 (2.18)
50
JEAG
0.3 G
< 1.18
China
for Bus
800
653 (1.00)
2100 (3.22)
1550 (2.37)
GIS/AIS
20
600
462
1380 (2.11)
2.30
1142 (1.75)
18
5
2100 (3.22)
1550 (2.37)
63
JEAG
0.25 G
< 1.15
for Line
800
653 (1.00)
1950 (2.99)
1550 (2.37)
AIS
20
648
498
1491 (2.28)
2.30
1234 (1.89)
18
5
2100 (3.22)
1550 (2.37)
63
JEAG
0.25 G
< 1.15
Table 5.2.2 Main characteristics of systems and MOSAs for UHV (1200, 1100 kV)
and 800 kV systems [2-12]
Figure 5.2.5 shows the UHV higher performance MOSAs in test stations of India, China and Japan. The long-term
reliability of UHV higher performance MOSAs for GIS has been confirmed since 1996. The long term tests of the
polymer-housed and porcelain type MOSAs have just started and their reliability and stable performances will be
conformed in the near feature.
Page 184
TypethenameofTB
Figure 5.2.4 Representative GIS, Polymer-housed & Porcelain type UHV higher
performance MOSAs [1-3]
Figure 5.2.5 UHV higher performance MOSAs in test station of India, China and
Japan [1, 9]
Page 185
TypethenameofTB
In Japan, 550 kV higher performance MOSA, whose characteristics are shown in Table 5.2.3, has been adopted
since 1988. Figure 5.2.6 shows operating experience of the 550 kV higher performances MOSA in Japan. About
1900 units of the MOSAs have already been operated in the fields, and their reliability has been confirmed to be
very high.
The characteristics of MO element of 550 kV higher performance MOSA are same to those of 1100 kV higher
performance MOSA. The LIWV level of 550 kV GIS is reduced to 1425 kV by adopting the higher performance
MOSA in Japan as shown in Figure 5.2.2.
In addition, many higher performance MOSAs for the systems of 550 kV and above have also been used around
the world.
(1) Rated voltage (kV)

(2) Residual voltage (kV)
(3) MO element arrangement
(4) Typical maximum residual voltage
according to IEC standard (DLR: (2) / (1))
550 kV system
420 (1.00 p.u.)
870 at 10 kA
(2.07 p.u.)
1 block
1100 kV system
826 (1 p.u.)
1620 at 20 kA
(1.96 p.u.)
4 parallel blocks
2.3 2.8
--
Table 5.2.3 Characteristic of higher performance MOSAs in Japan
Figure 5.2.6 Operational experience of 550 kV higher performances MOSA in Japan
Page 186
TypethenameofTB
5.2.3 SUMMARY
(1) For the insulation co-ordination studies, digital simulation packages, such as EMTP, have generally been used
since 1980s. The accuracy of simulations is mainly dependent on the simulation conditions, and may, in some
instances, not correlate with the measured results. Therefore, the simulation conditions are introduced in a
technical report on guidance for insulation co-ordination studies, IEC 60071-4 in detail [13]. It provides details and
examples of the subtleties associated with the representation of networks and equipment for various types of
analysis; lightning or switching transients and temporary overvoltages.
(2) In addition to the insulation coordination studies, the depth of energy analysis is considered as essential, in
order to achieve the desired accuracy for practical design of surge arrester of the complex and important network
system. The energy duties of switching surges and TOVs of UHV power transmission system should be analyzed
using EMTP program. Therefore, it is desirable that this analyzing measure, using EMTP program, is
recommended in the standard on application guide for insulation co-ordination, IEC 60071-2, as essential to UHV
power transmission system analysis. [14]
(3) The UHV higher performance MOSAs have a higher protective performance due to their lower residual voltage
level, which is a decisive factor for UHV power transmission insulation design. Therefore, the main characteristics
of UHV higher performance MOSAs, conforming to the special requirements for them, are recommended to be
introduced in the standard on application guide for insulation co-ordination, IEC 60071-2.
Page 187
TypethenameofTB
References
[1] T. Nakamura, T. Kobayashi, Y. Ishizaki, S. Watabiki, S. Ishibe and T. Kawamura: Insulation Co-ordination
based on 1100 kV High performance Surge Arresters, CIGRE SC C4 2009 Kushiro Colloquium
[2] T. Chiba, Y. Kayano, K. Hirai, et al.: Consideration Of Surge Arresters Used For The UHV Transmission
Systems In China, International Conference of UHV Power Transmission Technology, Oct., 2006, Beijing China,
pp989-995
[3] Y. Ishizaki, T. Kobayashi, M. Kan: The Consideration and Experiences of UHV Metal Oxide Surge Arresters
(MOSAs), CIGRE International Conference on Development of 1200 kV National Test Station, 29 &30 September
2010, New Delhi
[4] M. Reinhard, V. Hinrichsen, B. Richter and F. Greuter: Energy Handling Capability of High-Voltage Metal-Oxide
Surge Arresters Part 2: Results of Research Test Program CIGRE Aug. 2008, A3-309
[5] V. Hinrichsen, R. Gohler, M. Clemens, T. Steinmets and P. Riffon: External Grading System for UHV MetalOxide Surge Arresters A New Approach to Numerical Simulation and Dielectric Testing, CIGRE Aug. 2008, A3205
[6] IEC 60099-4: Ed.2.2, 2009-05, Surge Arresters Part4 Metal-Oxide Surge Arresters Without Gaps for a.c.
Systems
[7] Ning Qiu, Guangzhou Zhang, et al.: Environmental Considerations for UHV Substation, IEC/CIGRE UHV
Symposium Beijing 2007 Session 2-3-5, 2007
[8] Seo II, Especificao Tpica De Pra-Raios, DAQ.G/DEL.E Outubro 2007 Volume II Seo II 1/15,
FURNAS CENTRAIS ELTRICAS S.A
[9] LI Ming-gang, SONG Ji-jun, HAN Shu-mo, FANG Yu-ying: Development and Application of Tank Type Surge
Arrester for 1100kV System in China, UHV09_CP0410-EN, 2009 International Conference On UHV Transmission
(May 20-22 Beijing, China)
[10] WANG Bao-shan, XIONG Yi, TANG Lin, ZUO Zhong-qiu, WAN Ke, CHEN Li: The Summary And Conclusion
Of Chinese 1000 Kv Ultra High Voltage Porcelain Housed Metal Oxide Surge Arresters Development And Type
Test, UHV09_CP0616-EN, 2009 International Conference On UHV Transmission (May 20-22 BEIJING, CHINA)
[11] S. MaruyamaS. ShirakawaK. NakanoJ. Ozawa: Zinc Oxide Surge Arresters for the 735kV Power
Transmission LineCIGRE Open Conference on EHV Transmission Systems and Meeting Study Committees 22
and 23, 1987
[12] WANG Bao-shan, XIONG Yi, TANG Lin, ZUO Zhong-qui, WAN Ke and CHEN Li: The Summary And
Conclusion Of Chinese 1000 Kv Ultra High Voltage Porcelain Housed Metal Oxide Surge Arresters Development
And Type Test, 2009 International Conference On UHV Transmission (May 20-22 Beijing China)
[13] IEC TR 60071-4, Insulation Co-ordination Part 4: Computational Guide To Insulation Co-ordination And
Modeling Of Electrical Networks, 2004-06.
[14] IEC 60071-2, Insulation Co-ordination Part 2: Application Guide, 1996-12.
Page 188
TypethenameofTB
5.3 Evaluation of waveform - Conversion in shape of field overvoltage to standard

impulse waveform in determining representative overvoltages 5.3.1 INTRODUCTION
IEC 60071-1 Insulation co-ordination Part 1 [1] defines representative overvoltages Urp as standard shape
overvoltage assumed to produce the same dielectric effect on the insulation as overvoltages [of a given class]
occurring in service due to various origins (3.19), and there is also a description that this analysis shall determine
a representative voltage and overvoltage, taking into account the characteristics of the insulation with respect to the
different behavior at the voltage or overvoltage shapes in the system and at the standard voltage shapes applied in
a standard withstand voltage test (5.2). In the step of determining representative voltages based on system
analysis results in the flow chart for the determination of rated or standard insulation level in IEC 60071-1, it is
probable that real overvoltage waveforms have different effects on equipment dielectrics from the standard impulse
waveform (1.2/50 s).
As for this subject, it is described in IEC 60071-2 Insulation co-ordination Part 2 [2] that As a general
recommendation, the dependence on the overvoltage shape should also be considered in the determination of the
representative amplitudes, in particular, for external insulation and oil-paper insulation (2.3.4.5). Further, Annex H
illustrates with an example and states that if the assumed actual overvoltage has a shape different from test shape,
the representative overvoltage may have to be modified accordingly so that tests truly verify the insulation strength.
CIGRE WG.C4.302 reviewed relevant technologies and researches on this subject and found the following [3-4]:
Recent researches analyzed field surges, clarified insulating characteristics for field real surges in comparison with
for the standard impulse waveform in the lightning surge time region, and proposed the waveform evaluation
method for GIS and transformer. According to them, the decay of the field overvoltage is generally large, making
the insulation requirements not as severe as those of the standard lightning impulse waveform. Consequently, it
could be possible in some cases to use lower withstand test voltages. (3.1).
5.3.2 SIMPLE METHOD IN THE PAST

Before getting into the latest technology, the simple one in the past is described briefly [5]. Figure 5.3.1 shows the
typical V-t curves (short time region) of the various insulating media. The simple evaluation method of lightning
overvoltages is illustrated in Figure 5.3.2, which consists of the following two parts;
[a] Analysis of a wave front portion
1) Draw a straight line in parallel with a time-axis from the maximum peak value (Vm) of an analysis result.
2) Set the intersection of this straight line and a criteria V-t curve (Cc) to (Po).
3) The stress experienced by the apparatus can be withstood if the wave-peak envelope of the portion derived from
the analysis wave is below the LIWV characteristics at the time of Po.
[b] Analysis of a wave tail portion
1) It is based on actual breakdown voltage V-t curve (Ca).
2) Curve (Ca) is obtained by connecting two points, which are 2 s and 50 s (or 90 % value of LIWV at 10 s) of
the standard lightning impulse voltage (see Figure 5.3.2).
Page 189
TypethenameofTB
Figure 5.3.1 Criteria V-t curves for various apparatus
Figure 5.3.2 Evaluation method of lightning surge waveform with V-t curve
5.3.3 NON-STANDARD LIGHTNING IMPULSE WAVEFORM

From the results of lightning overvoltage and disconnector switching overvoltage analyses, and measurements at
UHV and 550 kV substations, representative non-standard impulse waveforms were extracted. The rise time of the
wave crest varies, as a whole, in the range of 0.1 - 1.0 s for GIS and 0.5 - 1.5 s for transformer. Figure 5.3.3
shows examples of waveforms A, B, C, D and E.
(a) Waveform A: Pulse-shaped waveform
(b) Waveform B: The wave crest has a steep pulse-shaped part, and the wave tail is flat (Ratio between peak and
flat part: 0.7 - 0.9 both for GIS and transformer)
(c) Waveform C: A damped oscillatory waveform whose first wave is a maximum
(Frequency: 0.5 - 5.0 MHz for GIS, 0.4 - 1.0 MHz for transformer)
(d) Waveform D: A damped oscillatory waveform whose second wave is the crest value
Page 190
TypethenameofTB
(Frequency: 0.5 - 5.0 MHz for GIS, 0.4 - 1.0 MHz for transformer)
(e) Waveform E: Double-frequency oscillatory waveform
(Lower frequency: Approx. 1.0 MHz, Upper frequency: Approx. 5.0 MHz for GIS)
Waveform B appears depending upon the positional relationship with protected points when the lightning arrester
operates, and waveforms C and D occur as a result of a negative and/or positive reflection inside and/or outside
the substation. Waveform A corresponds to the one when the decay of the DC component and/or oscillatory
component of waveforms B, C and D is large. Also, waveform E occurs during a re-arcing surge inside the
substation, as caused by a disconnector surge, when there exist two oscillation routes within the substation. Thus,
in general, it can be said that these waveforms are representative waveforms in view of the mechanism under
which they occur.
Figure 5.3.3 Examples of non-standard impulse waveforms
5.3.4 ADVANCED EVALUATION METHOD WITH GIS
Page 191
TypethenameofTB
5.3.4.1 I NSULATION
CHARACTERISTICS OF
GIS
ELEMENT FOR NON - STANDARD LIGHTNING IMPULSES
(1) Experimental results

The SF6 gas gap was used as the insulation element of GIS. The electrodes used in the test generated quasiequal electric fields whose utilization factors () were 0.60 and 0.45, and the gas has an absolute pressure of 0.50
MPa.
With changing waveform parameters such as the frequency and the decay time, more than 100 cases were
examined. Moreover, effects of other factors were investigated like the scale, the gas pressure, the voltage polarity,
superposition with DC, electrode material and roughness.
As an example of the results, Figure 5.3.4 shows the characteristics with respect to waveforms A and E. In the
case of waveform A, the minimum value of the insulation breakdown V-t characteristics is 264 kV, which is 1.20
times that of the case of a standard lightning impulse. In other words, it is possible to convert waveform A into a
standard lightning impulse waveform by dividing the crest value by 1.20. This interpretation applies to the other
experimental results as well.
(2) Duration time evaluation
The insulation characteristics are arranged according to the waveform duration time of the 80 % level of the peak
value. Figure 5.3.5 explains how to calculate the duration time td: the sum of td1+td2. All of the characteristics for
waveform A, the pulse parts of waveform B, and waveforms C and D, and the double-frequency oscillatory
waveform E, lie roughly along one characteristic line as shown in Figure 5.3.6. This characteristic corresponds to
equations in the Calculation of conversion factor K box in Figure 5.3.7, which is discussed later.
(a) Waveform A (Gas pressure: 0.50 MPa abs,

: 0.45, Gap length: 12 mm)
(b) Waveform E (Gas pressure: 0.50 MPa abs,

: 0.60, Gap length: 10 mm)
Figure 5.3.4 Examples of insulation characteristics with respect to the nonstandard waveform of the gap
Page 192
Voltage [pu]
TypethenameofTB
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
td1
Peak level
td2
80% level
0.5
1.5
2.5
Time [ s]
W aveform A
W aveform B
W aveform C
W aveform D
W aveform E
Standard lightning im pulse
(V min -0.8VLI,min)/VLI,min
Breakdown voltage ratio-0.8 [pu]
Figure 5.3.5 Calculation of duration time td
0.1
0.1
10
100
Duration
[ s]
Duration
timetime
[s]
Figure 5.3.6 (V min -0.8V LI, min )/V LI, min characteristics as a function of duration time
over 80 % voltage level. (V min : minimum breakdown voltage at waveform A, B, C and
D, VLI, min: minimum breakdown voltage at standard lightning impulse waveform).
This characteristic corresponds to equations in the Calculation of conversion
factor K box in Figure 5.3. 7.
Page 193
TypethenameofTB
5.3.4.2 E VALUATION
METHOD FOR
GIS
The waveform was divided into elements, including each oscillatory wave and the flat part. It was then evaluated
based on the characteristics of section 5.3.4.1, and finally the crest value that was converted into an equivalent
standard lightning impulse waveform as the overall waveform was calculated. Figure 5.3.7 shows the waveform
evaluation flow for GIS.
Figure 5.3.7 Waveform evaluation flow for a GIS
Page 194
TypethenameofTB
5.3.4.3 A PPLICATION
TO A LIGHTNING OVERVOLTAGE AT
GIS
IN THE
UHV
SYSTEM
Figure 5.3.8 shows an example of the lightning overvoltage analysis waveform applied to GIS in a UHV substation.
This waveform has the shape of waveform B whose tail level is approximately 60% of the peak, after the steep
oscillatory surge has decayed.
Figure 5.3.8 Application to a GIS
Table 5.3.1 shows the results of analyzing and evaluating this waveform using the flow of Figure 5.3.5. In the case
of this waveform and = 0.60, the wave crest is severer than the wave tail, and the overall waveform is equivalent
to a 2094 kV standard lightning impulse waveform. In other words, the crest value was converted into a value that
was about 25 % lower.
Waveform
element
80 % duration
time
Peak voltage
(kV)
Conversion ratio
Equivalent standard
LI Vol.(kV)
B (wave crest)
0.16s
2617
1.25 (1.30)
2094 [=0.60]
2016 [=0.45]
B (wave tail)
1517
0.90
1686
Table 5.3.1 Evaluation of the lightning surge waveform in the GIS of a UHV system
To think of the principle of construction, this method can be generally applicable to lower voltage classes as well as
550 kV and UHV. In fact, it has been put into practical use in 66 kV to UHV voltage classes in some electric power
companies.
Page 195
TypethenameofTB
5.3.5 ADVANCED EVALUATION METHOD WITH TRANSFORMER
5.3.5.1 I NSULATION
CHARACTERISTICS OF TRANSFORMER ELEMENT FOR NON - STANDARD LIGHTNING
IMPULSES
(1) Experimental results

Oil gap, section-to-section insulation and turn-to-turn insulation were used as the insulation elements of a
transformer. About 100 cases were examined with changing waveform parameters such as the frequency and the
decay time. As an example of the results, Figure 5.3.9 exhibits the characteristics with respect to waveform A of the
oil gap, and also the characteristics with respect to waveform C of the turn-to-turn insulation. In the case of
waveform A, the average value of the insulation breakdown V-t characteristics is 270 kV, which is 1.38 times that of
the case of a standard lightning impulse. In other words, this means that it is possible to convert waveform A into a
standard lightning impulse waveform with dividing the crest value by 1.38. This interpretation applies to the other
experimental results as well. In the case of waveform C, the minimum value of the insulation breakdown V-t
characteristics is 185 kV, which is 1.26 times that of the case of a standard lightning impulse.
250
-Oscillation frequency: 1
MHz
-Half-amplitude cycle
count: 1 cycle
400
Oil gap
A-waveform
(0.84/1.56 s)
300
Voltage
kV(kV)
Voltage
(kV)
kV
200
200
150
0.841.56
50
100
2
3
Time
(s)
: Breakdown
:1MHz,
:
Partial
:1
discharge
inception
100
10
Time
(s)
(a) Characteristics with respect to waveform A of

the oil gap
(b) Characteristics with respect to waveform C

of the turn-to-turn insulation
Figure 5.3.9 Examples of insulation characteristics with respect to non-standard

waveforms
(2) Duration time evaluation
Concerning the oil gap and section-to-section insulation, Figure 5.3.10 (a) shows that when the insulation
characteristics are arranged according to the waveform duration time of the 80 % level of the peak value, all of the
characteristics for waveform A, the pulse parts of waveform B, and waveforms C and D, lie roughly along one
characteristic line. Figure 5.3.11 explains how to calculate the duration time td: the sum of td1+td2.
Concerning the turn-to-turn insulation, Figure 5.3.10 (b) shows that when the insulation characteristics are
arranged according to the waveform duration time of the 90 % level of the peak value, all of the characteristics for
waveform A, the pulse parts of waveform B, and waveforms C and D, lie roughly along one characteristic line.
However, in contrast to the other models, the duration time for 90 % and higher in the case of an oscillatory
waveform is not a total value, but is calculated separately for each wave as waveform A.
Page 196
TypethenameofTB
(a) Oil gap (This also applies to the section-tosection insulation.)
(b) Turn-to-turn insulation
Voltage [pu]
Figure 5.3.10 Characteristics arranged according to duration time

1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
td1
Peak level
90% level
td2
80% level
td1
0
0.5
1.5
2.5
Time [ s]
Figure5.3.11 Calculation of duration time td

5.3.5.2 E VALUATION
METHOD FOR OIL - FILLED TRANSFORMERS
The waveform is divided into elements including each oscillatory wave and the flat part. It is then evaluated based
on the characteristics of Section 5.3.5.1, and finally the crest value that is converted into an equivalent standard
lightning impulse waveform as the overall waveform is calculated. Figure 5.3.12 presents the waveform evaluation
flow for an oil-filled transformer.
Page 197
TypethenameofTB
Overvoltage waveform
Resolving into
element waveforms
Pulse-like
(Waveform A and pulse
part of waveform B)
Oscillatory
Flat part
(Waveforms C and D)
(Wave tail of waveform B)
Calculation of duration time

*Oil gap / Section-to-section :
Sum total for 80% or higher
*Turn-to-turn :
90% or higher (Calculated
for each pulse)
Calculation of conversion factor K
(Ratio with respect to standard waveform)
K 1.37 t 0.11 (0.3 t 20s)
*Turn-to-turn :
K 1.06 t 0.15 (0.25 t 1.05s)
K 1.05 t 0.021 (1.05 t 12s)
Calculation of conversion factor K

(Ratio with respect to standard waveform)
K = 0.90
*Turn-to-turn :
K = 0.95
Calculation of crest value of equivalent standard waveform

(Calculation for all waveform elements i and insulation positions j)
Vsij = Vpi/Kij (for all i,j)
Vsij: Crest value of equivalent standard waveform
Vpi: Voltage of object element waveform
Overall evaluation
V = Max [Vsij] i,j
If the voltage exceeds the crest value as a result of conversion
of the flat part, the original crest value is used.
Figure 5.3.12 Waveform evaluation flow for an oil-filled transformer

5.3.5.3 A PPLICATION
TO A LIGHTNING SURGE AT OIL - FILLED TRANSFORMER IN THE
500 K V
SYSTEM
Figure 5.3.13 is an example of a lightning surge analysis waveform applied to a transformer in a 500 kV substation.
In this waveform, the wave tail maintains a level of about 70 % with respect to the peak after the steep oscillatory
surge has decayed.
Page 198
TypethenameofTB
Figure 5.3.13 Application to a transformer lightning surge
Table 5.3.2 summarizes the results of analyzing and evaluating this waveform using the flow of Figure 5.3.12. The
severest part of the waveform is the pulse-shaped one No.2 with respect to the turn-to-turn insulation. The overall
waveform is dominated by this part, and is equivalent to a 934 kV standard lightning impulse waveform. The crest
value was converted into a value that was about 24 % lower.
Insulation
element
Waveform
element
Oil gap /
Section-tosection
Oscillatory parts
(1~4): C or D
Flat part (5)
Max pulse (2)
Other pulses (1)
Flat part (5)
Turn-to-turn
Conversion
ratio
Equivalent
standard LI
Vol.(kV)
1158
1.40
827
784
1158
1084
784
0.90
1.24
1.24
0.95
871
934
874
825
Duration
time (s)
Peak
voltage (kV)
0.84
0.35
0.35
-
Table 5.3.2 Evaluation of re-striking surge waveform in the transformer of a 500kV

system
To think of the principle of construction, this method can be generally applicable to lower voltage classes as well as
500 kV and UHV ones. In fact, it has been put into practical use in 66 kV to UHV voltage classes at some electric
power company.
5.3.6 SUMMARY
Field real overvoltages of non-standard lightning impulse waveform are analyzed and the insulating characteristics
of SF6 gas and oil-filled transformer elements for these actual overvoltages are clarified to convert the waveform
into the standard lightning impulse waveform. It was found that the insulation requirements could not be as severe
as those of the standard lightning impulse waveform, since the decay of the field overvoltage is generally large.
Consequently, it could be possible in some cases to use lower withstand test voltages for GIS using SF6 and oilfilled transformers. The proposed conversion method for non-standard lightning impulse waveform is applicable up
to 1100 kV UHV system.
Page 199
TypethenameofTB
References
[1] IEC 60071-1, Insulation Co-Ordination Part 1: Definitions, Principles And Rules, 2006-01.
[2] IEC 60071-2, Insulation Co-Ordination Part 2: Application Guide, 1996-12.
[3] CIGRE WG.C4.302, Insulation Co-Ordination Related To Internal Insulation Of Gas Insulated System In Sf6
And N2/Sf6 Gas Mixtures Under Ac Conditions, CIGRE Technical Brochure No.360, 2008-10.
[4] S. Okabe, S. Yuasa, M. Koto and E. Zaima, Evaluation Of Lightning Surge Waveform For LIWV Reduction Of
Substation Equipment 13th ISH P.05.66, 2003.
[5] B. Wahlstrm, T. Yamagiwa, E. Colombo, F. Gallon, Possible Applications And Benefits Expected Through
Reduced Insulation Co-Ordination Voltages, CIGRE SC A3&B3 Joint Tutorial in Tokyo, 2005.
5.4 Conclusions
(1) In the analysis of power system, digital simulations such as EMTP have been used in many countries. In
addition, the comparison estimation is also performed the analysis results and the measured value in each country,
then, its validity has been confirmed up to UHV system. For example, in the case of UHV line in Japan and China,
the maximum error between measurements and simulation for switching overvoltages is about 5 %, which is
agreeable in the design of transmission lines.
(2) The main characteristics of MOSAs for the UHV projects in the world have been investigated in the WG. The
UHV higher performance MOSAs have a higher protective performance due to their lower residual voltage level,
which is a decisive factor for UHV power transmission insulation design. To illustrate the characteristics of the
higher performance MOSA is desirable to IEC60071-2.
In the UHV system is required to evaluate the capability (switching overvoltage and TOV) of the arrester by a
detailed analysis using EMTP, in IEC 60071-2, it is desirable to mention the need for tolerance evaluation.
(3) Field real overvoltages of non-standard lightning impulse waveform are analyzed and the insulating
characteristics of SF6 gas and oil-filled transformer elements for these actual surges are clarified to convert the
surge waveform into the standard lightning impulse waveform. Non-standard lightning impulse waveform evaluation
method shown in section 5.3 has already been put into practical use to UHV and could be possible in some cases
to use lower withstand voltages. The evaluation method can be reflected in the IEC60071-2 as specific method.
Page 200
TypethenameofTB
6 Switching Overvoltage Mitigation Measures for Future UHV

Systems
6.1 Introduction
Switching overvoltages are the primary dimensioning parameter for air-clearances on EHV and UHV systems. The
reason for this is that, while clearances for lightning impulse withstand increase linearly with gap distance, those for
switching impulse withstand tend to saturate with increasing gap distances (Figure 6.1.1). Clearly, therefore, there
is great merit to mitigate switching overvoltage levels.
U50 flashover value (kV peak)
8000
7000
6000
5000
Switching impulse
withstand
4000
Lightning impulse withstand
3000
2000
1000
0
0
10
15
20
25
Gap distance (m)
Figure 6.1.1 Lightning and switching impulse withstand versus gap distance for
rod-rod gap
The historical background to this subject is interesting. The first EHV system was the 400 kV system developed in
Sweden in the mid 1950s. No account was taken of switching overvoltages because their influence was unknown
and obviously no mitigation was considered. The development of the 400 kV system in Italy in the 1960s was the
first real recognition of switching overvoltages. A very large number of switching impulse tests were performed and
resulted in the now well-known U-curves and the notion of gap factors [1, 2]. The explanation for the U-curve was
provided by research work at EdF [3]. Switching impulse breakdown involves both streamer and leader
development (lightning impulse breakdown involves streamers only) with the latter being the main driver. The
minimum value of the U-curves thus represents optimal leader development and occurs at front-times in the range
of 100 to 400 s leading to a selection of a front-time of 250 s for standardization purposes. Detailed descriptions
of electrical discharges and breakdown in air can be found in [4, 5].
A number of switching overvoltage mitigation measures have been proposed or used starting in the 1960s:
Fast insertion of shunt reactors [6].
Closing resistors [6, 7, 8, 9].
Staggered pole closing [10, 17, 18, 20].
Line terminal arresters [10, 11, 12, 13, 14].
Page 201
TypethenameofTB
Controlled closing [15, 6, 8, 16, 17, 18, 19, 20, 21].

Each of these measures will be discussed in the next section and will then be evaluated as to relevance and value
for UHV systems.
Future UHV systems in China and Japan will operate at 1100 kV and at 1200 kV in India. In the following sections,
switching overvoltages will generally be discussed in terms of per-unit and for the above system voltages 1 p.u.
equals 898 kV and 980 kV, respectively.
6.2 Fast Insertion of Shunt Reactors

The primary purpose of shunt reactors is to provide compensation of transmission line capacitance and have
secondary value, if connected, of providing some reduction in switching overvoltage levels. No utility would
consider adding shunt reactors specifically for the latter purpose. The intent, therefore, with this proposal is to
switch in disconnected reactors preferably prior to energizing or re-energizing the line [6]. Compared to other more
measures discussed below, this measure is not of significant interest.
6.3 Closing Resistors

Once the significance of switching overvoltages was understood, closing resistors were applied on EHV air blast
circuit breakers starting in the 1960s [7, 8, 9]. The principle behind the use of closing resistors is as follows:
Assume a voltage UD across the open circuit breaker at the instant of closing. Closing effectively produces a
travelling wave of magnitude UD which divides between the resistor R and the line surge impedance Z. The incident
wave at the open end of the line is thus:
U oL
UD Z
RZ
and after voltage doubling, the line end voltage becomes

U eL
2U D Z
UTC
RZ
where UTC is the initial trapped charge on the line.

Taking, for example, UD 2 (1 p.u. positive source voltage and 1 p.u. negative trapped voltage)
U eL
4Z
1
RZ
If R 0, then UeL 3 p.u.

Clearly, the higher the value of R, the lower the value of UeL. However, the resistor is inserted only momentarily and
bypassing it generates a travelling wave given by:
UR
U RD R
RZ
where URD is the voltage across the resistor at the instant of bypassing.
At the line end with voltage doubling, the voltage is:
URe L
2U RD R
U pf
RZ
where Upf is the frequency voltage at the instant of travelling wave arrival.
Page 202
TypethenameofTB
An upper limit on the value of R is therefore required. The value of R is usually optimized through computer studies
and is typically in the range 250 to 600 dependent on the application.
Generally, the aim with closing resistors is to limit the voltage at the receiving end of the line to 2 p.u. However,
closing resistors do little for the voltage profile along the line and higher overvoltages than 2 p.u. at points where
incident and reflected waves cumulatively coincide.
6.4 Staggered Pole Closing

Staggered pole closing is more a complimentary switching overvoltage mitigation measure than a fundamental
measure [10, 17, 18, 20]. The principle is quite simply closing the individual poles one-half cycle apart in the
expectation that transients in the closed phase will have greatly attenuated before the next poles close. The effect
is to reduce the coupling contribution in any one phase from the other two phases. This measure is inexpensive,
easily implemented and reliable.
6.5 Line Surge Arresters

As an alterative to the use of closing resistors, metal oxide surge arresters were applied on line terminals starting in
the late 1980s [10, 11, 12, 13, 14]. This has since become common practice at EHV and even at HV but there for
lightning protection reasons.
The principle of surge arrester application is shown in Figure 6.5.1. In essence, there is an interaction between the
load line of the circuit and the U-I characteristic of the arrester. The arrester will conduct a current Iar to limit the
voltage to Var as shown in Figure 6.5.1.
The effect of the line surge arresters is shown in Figure 6.5.2 for a typical EHV application such as described in
[14]. With arresters at the line ends only, the overvoltages at those points only will be limited to the protective level
of the arresters. However, the voltage profile along the line shows much higher overvoltage levels and suggests a
need to even out the profile. This is achieved by adding arresters at intermediate points along the line as shown in
Figure 6.5.2.
The surge arresters applied for this purpose are special in the sense that their characteristics are derived from
EMTP system studies. A balance has to be achieved between the protective level and the energy absorbed which
work in opposite directions, i.e. decreasing the protective level increases the energy absorbed and vice versa. In
the ultimate, further reduction in the overvoltage level requires a more proactive mitigation measure.
Page 203
TypethenameofTB
Figure 6.5.1 Principle of surge arrester application
Page 204
TypethenameofTB
2% overvoltage value (pu)
4.5
4
3.5
No overvoltage control
measures
Line end arresters only
2.5
Line end and intermediate

arresters
2
1.5
1
0
20
40
60
80
100
Distance from sending end (% )
Figure 6.5.2 Effects of line surge arresters on switching overvoltage levels

6.6 Controlled Closing
Controlled closing of circuit breakers to limit switching overvoltages was suggested as early as 1966 [15]; was
viewed as less than practical 1969 [8]; was proposed as part of a combined closing resistor and controlled closing
solution in 1970 [6]; was actually tested on a power system in 1976 [16]; but finally did not become a reality until
the 1990s [17, 18, 19, 20, 21]. The reason for this delay is discussed below.
When an unloaded transmission line is switched out, the remaining voltage on the line will be a DC voltage if the
line is not compensated or an oscillatory voltage if the line is compensated, the frequency being dependent on the
degree of compensation. The intent of controlled switching then is to close the circuit breaker when the difference
between the source power frequency voltage and the line voltage is at a minimum. For the uncompensated line
case, the minimum will occur once every cycle when the polarities of the source and line voltages are the same.
For the compensated line case, the differential voltage minimums are frequency dependent as shown in Figure
6.6.1. The 45 Hz case corresponds to a high degree compensation (80%) with shunt reactors at both line ends and
the voltage minimums are clearly evident. At 33 Hz low degree of compensation (40%) with shunt reactor at one
end only the minimums are less evident by comparison.
Traces from actual field measurements by BC Hydro are shown in Figure 6.6.2 and Figure 6.6.3. Figure 6.6.3
shows that the line side oscillation is more complex than that shown in Figure 6.6.2. In reality, three oscillation
modes are involved in the 40% compensation case and the minimums become even less distinct.
Page 205
TypethenameofTB
2.5
60 Hz
Difference 60 vs 33 Hz
2
Difference 60 vs 45 Hz
1.5
Voltage (pu)
0.5
0
0
20
40
60
80
100
120
140
160
-0.5
-1
-1.5
-2
-2.5
Time (ms)
Figure 6.6.1 Differential voltage across circuit breaker
Figure 6.6.2 500 kV line 80% compensated
Figure 6.6.3 500 kV line 40% compensated
Page 206
180
TypethenameofTB
It is clear that this mitigation measure requires a dynamic controller which can analyze the differential voltage
across the circuit breaker, locate the minimums, predict the future minimums and close the breaker accordingly
when the close signal is applied, all within 0.5 s or perhaps less. The sophistication of the control device obviously
goes far beyond the zero crossing detectors and sequence timers available when this measure was first proposed.
As with controlled capacitor bank switching, the rate of decay of dielectric strength (RDDS) of the circuit breaker is
a factor in the application of this measure.
This measure combined with line end and mid-line 372 kV rated arresters has been successfully in-service at 500
kV on the BC Hydro system since 1995 [20, 21]. The intent was to limit switching overvoltages to 1.7 p.u.
6.7 Comparison and Relevance to Future UHV Systems

Figure 6.7.1 shows the cumulative frequency distribution for the overvoltages at the open end of a transposed 330
km, 500 kV line following high-speed reclosing for the various overvoltage limitation measures. The arresters
considered were rated at 372 kV and about an equivalent IEC Line Class 7. Clearly, staggered pole closing has
only incremental value and cannot be considered as stand-alone option. The value lies in the options of closing
resistors, line terminal (meaning arresters at both line terminals) and mid-line arresters and controlled closing
combined with line terminal arresters.
100.0
No control m easures
Cumulative frequency (%)
90.0
Staggered pole closing
80.0
70.0
Closing resistors
60.0
Line term inal arresters
50.0
40.0
Line term inal plus m id-line

arresters
30.0
Controlled closing plus line

term inal arresters
20.0
Closing resistor 400 ohm s

plus line end arresters
10.0
Closing resistor 800 ohm s

plus line end arresters
0.0
1.0
1.5
2.0
2.5
3.0
Overvoltage level (pu)
Figure 6.7.1 Cumulative frequency overvoltage distributions for various

overvoltage limitation options
The effect of the various options on the voltage profile along the line is of greater interest and this is shown in
Figure 6.7.2 for a 330 km long 500 kV line.
The two options that provide the most significant line overvoltage profile limitation are closing resistors or controlled
closing both combined with line terminal arresters. At 800 kV and below, closing resistors are now rarely used and
either an arrester only option or the controlled closing option is adopted. At the UHV levels of 1000 kV and above,
the situation is different as discussed below.
Page 207
TypethenameofTB
3.5
3
No control measures
Overvoltgae (pu)
Closing resistors (400 ohm)only

Closing resistors (400 ohm) plus
line terminal arresters
Line terminal arresters plus SPC
2.5
Line terminal and mid-line

arresters plus SPC
Controlled closing and line
terminal arresters
Closing resistors (800 ohm) only
Closing resistor (800 ohm) plus

line end arresters
1.5
1
0
10
11
12
Line Section No.
Figure 6.7.2 Overvoltage profile along 330 km long transposed 500 kV line for
various overvoltage limitation options
Circuit breakers applied at 1000 kV and 1200 kV will incorporate opening resistors in order to meet transient
recovery voltage (TRV) requirements during fault clearing. The resistors, though probably rated resistance-wise for
opening, can double as closing resistors. However, as shown in Figure 6.7.1 and Figure 6.7.2, the lower the
resistance value the greater the overvoltage limitation value. The choice between the closing resistor versus
closing controller option is a decision for the user and a comparison is given in Table 6.7.1.
Attribute
Proven technology
Complexity
Location
Maintainability
Closing resistor
Yes, in use since 1960
High; multiple moving mechanical
parts
At line potential
Low; requires circuit breaker outage
Provision of spares
Yes; parts only

complete module
Future improvement
potential
Limited if at all
Other
rather
than
Circuit breakers will incorporate

opening resistors which can readily
double as closing resistors
Closing controller
Yes, in use since early 1990s
Low; no moving parts, circuit
boards and associated software
only
At ground level in control room
High; does not require circuit
breaker outage
Yes; complete module or circuit
boards
Yes, component and software
advances
-
Table 6.7.1 Comparison of closing resistor and closing controller overvoltage

limitation options
Page 208
TypethenameofTB
6.8 Conclusions
The above discussion and study results for a transposed 500 kV transmission line illustrate the combined
contribution of line end arresters and closing resistors or a controller to switching overvoltage limitation. The results
can certainly be viewed as indicative of performance but, to determine absolute design parameters, the user needs
to conduct studies focused on the actual application details and considerations. Accurate representation of arrester
and controller characteristics, resistor insertion and line configuration is essential in order to achieve a valid design
basis for lines meeting dependability requirements.
References
[1] G. Carrara, "Investigation on Impulse Sparkover Characteristics of Long Rod/rod and Rod/plane Air Gaps".
Cigre Report No. 328, 1964.
[2] L. Paris and R. Cortina, "Switching and Lightning Impulse Characteristics of Large Air Gaps and Long Insulator
Strings". IEEE Transactions on Power Apparatus and Systems No. 87, 1968.
[3] J. N. Ross et al, "Positive Discharges in Long Air Gap at Les Renardieres." Electra No. 53. 1977.
[4] N. L. Allen, "Mechanism of Air Breakdown." Chapter 1, Advances in High Voltage Engineering (Book), IEE
Press 2004.
[5] V. Cooray, "Mechanism of Electrical Discharges." Chapter 3, The Lightning Flash (Book), IEE Press 2003.
[6] H. B. Thoren, "Reduction of Switching Overvoltages in EHV and UHV Systems." IEEE Trans. PAS-90, 1971.
[7] C. L. Wagner and J.W. Bankoske, "Evaluation of Surge Suppression Resistors in High-Voltage Circuit
Breakers." IEEE Trans. PAS-86, 1967.
[8] P. A. Baltensperger and P. Djurdjevic, "Damping of Switching Overvoltages in EHV Networks New Economic
Aspects and Solutions." IEEE Trans. PAS-88, 1969.
[9] R. G. Colclasser, C.L. Wagner and E.P. Donohue, "Multistep Resistor Control of Switching Surges." IEEE Trans.
PAS-88, 1969.
[10] A. C. Legate, J.H. Brunke, J.J. Ray and E.J. Yasuda, "Elimination of Closing Resistors on EHV Circuit
Breakers." IEEE Trans. PD, Vol. 3, 1988.
[11] J. R. Ribeiro and M.E. McCallum, "An Application of Metal Oxide Surge Arresters in the Elimination of Need for
Closing Resistors in EHV Circuit Breakers." IEEE Trans. PD, Vol. 4, 1989.
[12] J. K. Blakow and T.L. Weaver, "Switching Surge Control for the 500 kV California-Oregon Transmission
Project." Cigre Report No. 13-304, 1990.
[13] A. Eriksson, J. Grandl and O. Knudsen, "Optimized Line Switching Surge Control Using Circuit-Breakers
without Closing Resistors." Cigre Report No. 13-305.
[14] Y. I. Musa, A. J. F. Keri, J. A. Halladay, A. S. Jagtiani, J. D. Manderville, B. Johnnerfelt, L. Stenstrom, A. H.
Khan and W. B. Freeman, "Application of 800 kV SF6 Dead Tank Circuit Breaker with Transmission Line Surge
Arrester to Control Switching Transient Overvoltages." IEEE Trans. PD, Vol. 17, 2002.
[15] E. Maury, "Synchronous Closing of 500 and 765 kV Circuit Breakers: A Means of Reducing Switching Surges
on Unloaded Lines." Cigre Report No. 143, 1966.
[16] H. E. Konkel, A. C. Legate and H. C. Ramberg, "Limiting Switching Surge Overvoltages with Conventional
Power Circuit Breakers". IEEE Trans. PAS-96, 1977.
Page 209
TypethenameofTB
[17] B. Avent and J. Sawada, "BC Hydro's Experience with Controlled Circuit Breaker Closing on a 500 kV Line."
Canadian Electrical Association, Engineering and Operating Division Meeting, March 1995.
[18] K. Froehlich, C. Hoelzl, A. C. Carvalho and W. Hofbauer, "Transmission Line Controlled Switching." Canadian
Electrical Association, Engineering and Operating Division Meeting, March 1995.
[19] A. H. Khan, D. S. Johnson, J. H. Brunke and D. L. Goldsworthy, "Synchronous Closing Application in Utility
Transmission Systems." Cigre Report No. 13-306, 1996.
[20] K. Froehlich, C. Hoelzl, M. Stanek, A. C. Carvalho, W. Hofbauer, P. Hoegg, B. L. Avent, D. F. Peelo and J. H.
Sawada, "Controlled Closing on Shunt Reactor Compensated Transmission Lines: Part 1 Closing Control Device
Development and Part 2 Application of Closing Control Device for High-Speed Autoreclosing on BC Hydro 500 kV
Transmission Line." IEEE Trans. PD, Vol. 12, 1997.
[21] B. L. Avent, D. F. Peelo and J. Sawada, "Application of 500 kV Circuit Breakers on Transmission Line with
MOV Protected Series Capacitor Bank." Cigre Report No. 13-107, 2002.
Page 210
TypethenameofTB
7 Some aspect of insulation coordination of air gaps in the UHV

range (phase-to-earth and phase-to-phase insulation)
7.1 Introduction
CIGRE WG A3.22 summarized UHV insulation coordination and emphasized the importance of reducing
overvoltage to as low as possible. Optimal insulation coordination can be achieved based on higher performance
MOSAs [18]. CIGRE WG B3.22 further investigated the insulation coordination in order to assess the minimum
phase-to-earth and phase-to-phase clearances required at UHV levels which have been carried out [19]. The
clearances depend on the type of overvoltage like power frequency, lightning and switching overvoltage as well as
various other factors such as altitude, climatic condition, rain and pollution. The discussion in this chapter mainly
addresses to the question that for a given Switching Impulse Withstand Voltage (SIWV), whether the associated
minimum air gap clearance specified in IEC is reasonably safe and how to optimize the air gap clearance.
The studies [1-8] for evaluation of the air gap insulation in the EHV/UHV range have been carried out over the
years through actual experiments in order to find out expression of dielectric behaviour of various air gaps. These
studies show that the positive polarity switching impulse is the most important one to determine withstand and
flashover characteristic. The gap configurations are influencing the minimum gap lengths. The lowest breakdown
voltage occurs on a rod-plane gap which is also taken as the reference gap type. The breakdown voltage also
depends on the shape of the switching impulse. The U50 flashover voltage plotted against time-to-peak for impulses
can result in U-curves. The time to crest associated with the lowest of such U-curves point is called the critical timeto-peak (Tpcr) of the air gap configuration.
In this chapter at first the Clearances in air to assure a specified switching impulse withstand voltage installation
called minimum air-gap clearances according to IEC 60071-1 [9] are shown. The values for the UHV range have
been included in the normative Annex A of the latest version of IEC 60071-1 Edition 8.1 - 2011-03. As the
comparison of these values with the actual used gap clearances of UHV-projects show discrepancies, these
discrepancies are investigated here.
The ways to calculate air-gap clearances according to the Application Guide of Insulation Coordination IEC 600712 [10] are shown and discussed in some details. Studies from the past give the background of flashover
characteristics for UHV clearances. Further on recent experimental investigations and results from India, China and
Japan are shown.
The resume out of standard, calculation and recent experimental results will be given at the end of this chapter.
7.2 Air Gap Clearances chosen for UHV Projects in different countries
The air gap clearance values for phase-to-phase as well as phase-to-structure as reported by various countries for
highest voltage level 1000kV and above are given in [18]. For the sake of comparison the values of IEC 60071-1,
Annex A have also been enclosed.
Page 211
TypethenameofTB
Clearances as reported by different countries

Italy
Japan
China
Russia
India
Rated voltage (kV)
1050
1100
1100
1200
1200
SIWV(Phase-earth) (kV)
1675
1550
1800
1800
1800
Phase to structure (m)
9.5
7.5 - 8.5
7.5
7.5 - 9.7
6.3 -8.3
Phase to phase (m)
12
10.5-11.5
11.3
11.4 - 12.4
10.3 - 12.3
Minimum clearances for the chosen SIWV according to IEC 60071-1 Amd. 1
SIWV(Phase-earth) (kV)
Minimum phase-toearth (m)
(Conductor-structure /
Rod-structure )
Minimum phase-to-phase (m)
(Conductor-conductor parallel /
Rod-conductor )
1675
1550
1800
1800
1800
5.6 / 7.4
4.9 / 6.4
6.3 / 8.3
6.3 / 8.3
6.3 / 8.3
9.1 / 10.9
(R=1.65)*
8.4 / 10
(R=1.7)*
9.8 / 11.6
(R=1.6)*
10.3 / 12.3
(R=1.65)*
10.3 / 12.3
(R=1.65)*
* R = SIWV (Phase-to-phase value) / SIWV (Phase-to-earth value)
Table 7.2.1 Comparison of air gap clearance values for different countries
and IEC 60071-1 Ed. 8.1
The values are as per the enquiry sheets submitted by different countries to WG.A3.22 and given in TB 362, page
89 [18]. It may be observed that the air gap clearance data considered by India is same as that of IEC. The
Chinese and Russian data are also quite close to IEC data. Meanwhile, the Italian and Japanese data for both
phase to structure and phase to phase are much higher than the corresponding IEC data.
7.3 IEC Minimum Air gap clearance according to IEC 60071-1

IEC Minimum air gap clearances are guidance clearances that should guaranty that all air gap clearances of a
substation designed for a given SIWV will withstand that SIWV if they are equal or larger than the IEC Minimum
air gap clearances. The corresponding tables and formulas as per IEC 60071-1 and IEC 60071-2 are given below:
The aims and limits of the minimum air clearances defined in IEC 60071-1, Annex A with Tables A.1, A.2 and A.3
are explained in Annex A as follows:
In complete installations (e.g. substations) which cannot be tested as a whole, it is necessary to ensure that the
dielectric strength is adequate.
The impulse withstand voltages shall be equal to, or greater than the standard rated impulse withstand voltages.
So minimum clearances specified are determined with a conservative approach.
These clearances are intended solely to address insulation co-ordination requirements. Safety requirements may
result in substantially larger clearances.
Clearances may be lower if it has been proven by tests on actual or similar configurations that the standard rated
impulse withstand voltages are met.
These distances are not applicable to equipment which has a mandatory standard impulse type test included in
their specification.
Clearances may also be lower, where it has been confirmed by operating experience that the gap configuration is
more favourable than that assumed for the recommended clearances.
Page 212
TypethenameofTB
In IEC 60071-1 Annex A the minimum air gap clearances for rod-structure and conductor-structure are given. The
values of the UHV systems have been extracted and are shown in Table 7.3.1and Table 7.3.2 for phase-to-earth
insulation and phase-to-phase insulation respectively.
Standard rated switching impulse

withstand voltage
kV
1550
1675
1800
1950
Minimum phase-to-earth
mm
Rod-structure
6400
7400*
8300*
9500*
Conductor-structure
4900
5600*
6300*
7200*
* under consideration
Table 7.3.1 Part of Table A.2 of IEC 60071-1, Ed.8.1

Standard rated switching impulse
withstand voltage
Phase-toPhase-toPhase-to-phase value
earth
phase
Phase-to-earth value
kV
kV
1550
1.60
2480
1550
1.70
2635*
1675
1.65
2764*
1675
1.70
2848*
1800
1.60
2880*
1800
1.65
2970*
1950
1.60
3120*
Minimum phase-to-phase clearance

Mm
Conductor-conductor
parallel
Rod-conductor
7600
8400*
9100*
9600*
9800*
10300*
11200*
9400
10000*
10900*
11400*
11600*
12300*
13300*
* under consideration
Table 7.3.2 Part of Table A.3 of IEC 60071-1, Ed.8.1

Correlation between standard rated switching impulse withstand
voltages and minimum phase-to-phase air clearances
The minimum air gap clearance values for the UHV range are all covered with the remark under consideration.
They are derived out of the extrapolation from the lower voltage ranges and not yet proven in practical application.
The following explanations to the clearance values are given in IEC 60071-1 [9]:
for the phase-to-earth insulation (Table 7.3.1):
Table A.2 correlates the minimum air clearances for electrode configurations of the conductor-structure
type and the rod-structure type with the standard rated switching impulse withstand voltage phase-to-earth.
The conductor-structure configuration covers a large range for normally used configurations.
Page 213
TypethenameofTB
for the phase-to-phase insulation (Table 7.3.2):

Table A.3 correlates the minimum air clearances for electrode configurations of the conductor-conductor
type and the rod-conductor type with the standard rated switching impulse withstand voltage phase-tophase. The unsymmetrical rod-conductor configuration is the worst electrode configuration normally
encountered in service. The conductor-conductor configuration covers all symmetrical configurations with
similar electrode shapes on the two phases.
These air-gap clearances are normative in this standard. That means everybody can rely on them.
From optimum design point of view it is known that the necessary clearance is depending on many parameters.
The dependence on the geometry is shown in the Annex G of IEC 60071-2 [10] dealing with the calculation of the
clearances. Furthermore it has to be mentioned that most often insulators are present in the gaps. The influence of
the insulators under dry and wet condition is to be considered additionally to assess a minimum clearance.
7.4 Air Gap Clearance Calculation for SIWV

The Application Guide of Insulation Coordination IEC 60071-2 [10] gives in Annex G information concerning the
calculation of air gap breakdown strength from numerous switching impulse tests carried out in the past. According
the Annex G air gaps can be characterized by the minimum strength as a function of the geometrical
characteristics of the air gap which are mainly the gap spacing d and the electrode configuration described by the
gap factor K.
The given equations are applicable for altitudes at sea-level. For altitudes above sea level appropriate altitude
correction factors are to be used. (see atmospheric correction factor in the process of insulation coordination, IEC
60071-2).
The following gives the expressions for calculation of air-gap clearances for phase-to-earth and phase-to-phase
configurations for switching impulses under clean and dry conditions.
Phase-to-Earth Configuration
Under standard atmospheric condition the 50% positive (most critical) flashover voltage U50 of air gaps without
insulators is given by the following equations:
U50 = K 1080 ln(0.46 d + 1)
(1)
for critical time to peak switching impulse shape [6] and

U50 = K 500 d 0.6
(2)
for the standard switching impulse shape [2],

where U50 is 50% positive flashover voltage in kV, d is the gap spacing in meter and K is the gap factor.
Since in the following, reference to Standard Switching Impulse (SSI) is made, reference to equation 2 is
systematically made.
For the different types of gaps, the positively stressed rod-plane gap has usually the lowest withstand level and is
used as a reference. The gap factor K is the ratio between the flashover voltage of the investigated air gap to the
rod-plane gap when subjected to the same switching impulse waveform. The gap factor varies with different
configurations (see below).
The withstand voltage U10 with 10% flashover probability can be obtained as:
Page 214
TypethenameofTB
U10 = U50 (1 - 1.3 )
(3)
with the per unit value of the conventional deviation (p.u. of U50).
Unless otherwise specified by the relevant Technical Committee , the per unit value of the conventional deviation,
may be assumed conservatively as 0.06 for switching impulse tests on air insulation, without any other insulation
involved.
Phase-to-Phase Configuration
In this case also, the gap factor approach may be used however the gap factor is influenced not only by the gap
configuration but also by the following ratio which describes the balance between the two voltage values on the
two phases:
+
= U / (U + U )
Where, U+ is the maximum positive voltage applied to one electrode and U- is the corresponding value of negative
voltage applied at the opposite electrode.
7.4.1 DERIVATION OF IEC 60071-1 ANNEX A ED 8.1 UHV MINIMUM AIR GAP
CLEARANCES FOR SUBSTATIONS
To derive the UHV minimum-air-gap clearance-values of the IEC tables the following calculations and comparisons
were done by the IEC Maintenance Team in charge of.
With fixation of K and the gap distances can be calculated for standard switching impulse voltage with a 10%
flashover probability (according standard withstand voltage).
Following equation 2 and 3 will lead to equation 4 and the UHV minimum air clearances are then given by:
d = [SIWV / (500 (1-1.3 ) K)]1/0.6 (m, kV)
d = [(1.0846 SIWV) / (500 K)]1/0.6 1000 (mm, kV)
(4)
with an average gap factor K derived for each IEC structure from lower SIWV for which clearances are already
given in the IEC 60071-1:
- for phase to earth: Rod-structure 1.095; conductor.-structure 1.297
- for phase to phase: Conductor-conductor 1.5918; rod-conductor 1.432
- = 6%
With the Equation (4) and above average values of gap factor (K), the minimum phase-to-earth and phase-tophase air clearances are determined for the UHV switching impulse levels. The values thus obtained are compared
with the values given in the tables of IEC 60071-1 (Table 7.4.1 and Table 7.4.2).
Page 215
TypethenameofTB
Standard rated
switching impulse
withstand voltage
Minimum phase-to-earth (mm)

Calculated with per Eq. (4)
According IEC 60071-1 Amd.1
kV
Rod-structure
Conductor-structure
Rod-structure
Conductor-structure
1550
6487
4892
6400
4900
1675
7382
5567
7400
5600
1800
8323
6277
8300
6300
1950
9510
7172
9500
7200
Table 7.4.1 Comparison of minimum phase-to-earth air clearance calculation

to table values
Standard rated switching impulse withstand

voltage
Minimum phase-to-phase clearance (mm)

Calculated with per Eq.
According IEC 60071-1
(4)
Amd.1
ConductorConductorRodRodconductor
conductor
conductor
conductor
parallel
parallel
7617
9078
7600
9400
8427
10044
8400
10000
9126
10877
9100
10900
9593
11433
9600
11400
Phaseto-earth
kV
1550
1550
1675
1675
Ratio:
Phase-to-phase /
Phase-to-earth
1.6
1.7
1.65
1.7
Phase-tophase
kV
2480
2635
2764
2848
1800
1.6
2880
9773
11648
9800
11600
1800
1950
1.65
1.6
2970
3120
10287
11168
12261
13310
10300
11200
12300
13300
Table 7.4.2 Comparison of minimum phase-to-phase air clearance calculation

to table values
From the above tables it can be seen that the values obtained from equation (4) are close to the values considered
in IEC 60071-1 for the typical values of gap factor given above. It may be noted that the above average gap factor
values result purely of computation. So they can and furthermore must be given here with high accuracy to explain
in detail the IEC Maintenance Teams method used to derive the UHV standardized clearances. The final chosen
values in the IEC-tables were indeed rounded values, some above and some below the ones given by
computations, for convenient handling in practical use.
7.4.2 GAP FACTOR RANGES AND SENSITIVITY ON AIR GAP CLEARANCES

The gap factor depends on gap configuration and actual gap factor can only be determined accurately through
experiments. Typical values of gap factors are summarized in IEC 60071-2 for different phase-to-earth and phaseto-phase configurations which are shown in Table 7.4.3 and Table 7.4.4.
Page 216
TypethenameofTB
Gap Type
Conductor Crossarm
Conductor Window
Typical Range
1.36 1.58
1.22 1.32
Conductor Lower Structure
1.18 -1.35
Conductor Lateral Structure

Rod Lower Structure
1.28 1.63
1.03 1.66
Reference Value
1.45
1.25
Conductor-Plane : 1.15
Conductor-Rod : 1.47
1.45
1.35
Table 7.4.3 Typical gap factors ( K ) for switching impulse breakdown phase-to-earth
(Summary of Table G.1 of IEC 60071-2)
= 0.5
1.80
1.65
1.62
1.50
1.45
Configuration
Ring-ring or large smooth electrodes
Crossed conductors
Rod rod or conductor-conductor (along the span)
Supported busbars (fittings)
Asymmetrical geometrics
= 0.33
1.70
1.53
1.52
1.40
1.36
Table 7.4.4 Gap factors ( K ) for typical phase-to-phase geometries

(Table G.2 of IEC 60071-2)
The sensitivity of the minimum gap clearance calculation on variation of the gap factor is shown in Table 7.4.5 for
the critical impulse shape and the standard impulse shape [equation (1) and (2)]. In terms of absolute value for the
UHV range an increase of 0.05 of the gap factor value results in a reduction of 120 mm to 820 mm of minimum gap
clearance.
Increase in Gap Factor

(in steps of 0.05)
Reduction in minimum gap clearance (mm)

1550 kV
1675 kV
1800kV
1950 kV
From To
Eq-1
Eq-2
Eq-1
Eq-2
Eq-1
Eq-2
Eq-1
Eq-2
1.15 - 1.2
461
409
555
466
664
525
817
600
1.2 - 1.25
402
366
482
417
573
470
702
537
1.25 - 1.3
353
329
421
375
499
422
609
483
1.3 - 1.35
312
297
371
338
438
381
532
436
1.35 - 1.4
278
269
329
306
387
345
468
395
1.4 - 1.45
249
245
293
278
344
314
415
359
1.45 - 1.5
224
223
263
254
308
286
369
327
1.5 - 1.55
202
204
237
232
277
262
331
299
1.55 - 1.6
183
187
215
213
250
240
298
275
1.6 - 1.65
167
172
195
196
226
221
269
253
1.65 - 1.7
153
159
178
181
206
204
244
233
1.7 - 1.75
140
147
163
167
188
189
223
215
1.75 - 1.8
129
136
150
155
173
175
204
200
1.8 - 1.85
119
126
138
144
159
162
187
185
1.85 - 1.9
110
118
128
134
147
151
172
173
Table 7.4.5 Sensitivity of gap clearance against gap factor variation
Page 217
TypethenameofTB
14000
1950 kV - IEC equation 2
flashover distance [mm]
12000
10000
8000
6000
4000
2000
1,8
1,75
1,7
1,65
1,6
1,55
1,5
1,45
1,4
1,35
1,3
1,25
1,2
1,15
1,1
1,05
gap factor K
Figure 7.4.1 Dependency of gap clearance for gap factor variations in steps of 0.05
Representative for all SIWV values the sensitivity due to gap factor variations (0,05 steps of K) according Table
7.4.5 is plotted for 1800 kV in Figure 7.4.2. Here the sensitivity due to the 0,05 steps are related to the absolute
value of gap clearance shown in Figure 7.4.1. The related sensitivity due to the gap factor steps are in the range of
5% to 8% for the calculations with 1800 kV (Figure 7.4.2). The difference between low and high gap factor values is
only 4% for standard impulses and 6% for critical wave shapes.
percentage of flashover distance
12%
error [%] - IEC equation 1
error [%] - IEC equation 2
10%
8%
6%
4%
2%
1,85
1,8
1,75
1,7
1,65
1,6
1,55
1,5
1,45
1,4
1,35
1,3
1,25
1,2
1,15
1,1
1,05
0%
gap factor K
Figure 7.4.2 Sensitivity of gap clearance differences for gap factor variations
(0.05 steps) related to the absolute gap clearance
(Calculated for 1800 kV only)
Page 218
TypethenameofTB
7.4.3 TOLERATED DISCHARGE PROBABILITY FOR SELF-RESTORING INSULATION

The withstand voltage is the value of a test voltage which has been applied under specified conditions in a
withstand voltage test. During this test a specified number of disruptive discharges are tolerated. In the IEC 600711 a self-restoring insulation is specified with a statistical withstand voltages which tolerated approximately a
disruptive discharge probability of 10% and a specified withstand probability of 90%.
According to IEC 60060-1 [11] for self-restoring insulations the requirements of the withstand voltage test are
fulfilled if not more than two disruptive discharges occur out of fifteen impulses of the specified shape and polarity
at the withstand voltage value.
The statistical test procedures given in IEC 60060-1[11] permit either direct evaluation of U10 and U50 or indirect
evaluation of U10 using the relationship U10 = U50 (1 -1.3 ). The value of , the per unit value of the
conventional deviation, depends on the experimental result and is conservatively considered as 6% for switching
impulse voltages [10]. A lot of tests show a smaller per unit value of the conventional deviation, (between 3% and
5%), only a few mention values larger than 6%.
The difference in calculated results of 1800 kV for per unit values of the conventional deviation of 3%, 6% and 8%
is shown in Figure7.4.3. For a medium gap factor value of e.g. 1.4 the gap-distance difference will be about 400
mm for both equations between a calculation with 6% or 3% (for equation 2: 5526 5157 = 369 and for equation 1:
5733 5329 = 404).
In practical use, from a U50 test result, a calculation with the conservative 6% of the per unit value of the
conventional deviation will lead to a more reliable gap design.
r= =
8%8%
0,06
=
deviation
0,03
= 3% deviation
6% deviation
12000
IEC equation 2
IEC equation 1
flashover distance [mm]
10000
8000
5796
6000
6035
5733
5526
5329
5157
4000
2000
1,8
1,7
1,6
1,5
1,4
1,3
1,2
1,1
1,8
1,7
1,6
1,5
1,4
1,3
1,2
1,1
0
gap factor K
Figure 7.4.3 Dependency of gap-distance from per unit values of the conventional
deviation, for calculations with 1800 kV SIWV
Page 219
TypethenameofTB
7.5 Background data on flashover characteristics for UHV air clearance

In the past, various investigations have been carried out to understand the flashover characteristics for UHV
clearances. With the focus on different gap configurations and the dependence of the breakdown voltage on the
time to crest some characteristics with short descriptions are presented in the following.
Figure 7.5.1 to Figure 7.5.3 show the influence of front time on 50% flashover voltage (U50) curves for rod-plane
gaps of various lengths under positive polarity impulses. According to past experience as shown in Figure 7.5.1,
the critical time to peak of rod-plane gap less than 12 m has been measured lower than 400 s but there was a
large discrepancy of results.
Figure 7.5.1 50% flashover voltage curves for rod-plane gaps of various lengths d
under positive polarity impulses [4]
Figure 7.5.2 Influence of front-time on 50% positive polarity slow front impulse
flashover voltage of different rod-plane gaps [5]
Page 220
TypethenameofTB
flashover voltage of rod-plane gap [6]
For Rod-plane gap Watanabe results [6] show that critical time to peak is following Tpcr = 55 d (s, m) (Figure
7.5.3), values very similar to Cortina results Tpcr = 50 d (with the per unit value of the conventional deviation =
5%) but Cortina results for conductor plane (more actual substation and OVHL configurations) shows Tpcr = 35 d
(s, m) and only = 3% for conductor length of 50 m [7].
According to Figure 7.5.4, the phase to phase critical time to peak Tpcr of positive polarity slow front impulse
flashover voltage of parallel conductor gap is within 200 s and 300 s. Thus for phase to phase the critical time to
peak coincides with the time to peak of the standard switching impulse voltage: Tcr = 250 s. For such insulations
the distances proposed presently for UHV by IEC 60071-1 Annex A Ed 8.1 should be largely sufficient.
flashover voltage of parallel conductor gap [6]
Page 221
TypethenameofTB
Dependence of the flashover voltage on the impulse shape and gap distance has been investigated also with
reference to rod-rod and conductor-rod configurations. Examples of U - curves are reported in Figure 7.5.5 The
U50 versus gap length characteristics are given in Figure 7.5.6 for rod-rod gap with the comparison of critical to
standard impulse shape. In these figures the numbers within brackets indicate the percentage of discharges to the
earth plane with respect to the total number of discharges. For rod-rod gap the per unit value of the conventional
deviation is 7% - 8%, but for conductor-rod gap is 4%.
flashover voltage of different rod-plane and conductor-rod gaps (with in brackets
the percentage of discharges to the earth plane with respect to he total number of
discharges [7]
a: critical impulse shape
b: standard impulse shape
Figure 7.5.6 50% positive polarity slow front impulse flashover voltage of rod-rod
for critical time to peak and standard switching impulse with different ratios
"height of the earthed rod to the gap spacing".(With in brackets the percentage of
discharges to the earth plane with respect to the total number of discharges) [7]
Page 222
TypethenameofTB
For UHV tower windows with "V-insulator strings experimental data of the critical time to peak are given in Figure
7.5.7 [4]. The critical time to peak Tpcr is within 200 s and 400 s and the needed air gap spacing is within 8 m and
9 m for the UHV SIWV range.
Figure 7.5.7 50% flashover voltage Curves for tower windows with "V-insulator
strings" of various spacing d under positive polarity impulses [4]
Cortina results [7] show that for conductor-tower arrangements (conductor cross arm, conductor-window, conductor
tower leg in a circus arrangement, see Figure 7.5.8) up to 10 m the critical time to peak is following Tpcr = 35 d (s,
m) and that for positive polarity no influence of insulator strings on the flashover voltage was observed.
Page 223
TypethenameofTB
Figure 7.5.8 Investigated conductor-tower arrangements according to [7]

In substation the presence of post type insulators first completely flatten the U curve and second reduce the
flashover voltage as shown in Figure 7.5.9 [7].
Figure 7.5.9 50% flashover voltage curves versus the time to peak of the applied
impulse of conductor-structure configurations with post insulators
(six units of 1.33 m each) [7]
Page 224
TypethenameofTB
Regarding insulators Figure 7.5.10 shows additionally that the influence is quite negligible with positive polarity,
while it becomes significant in case of negative polarity, being the larger the shorter the length of the earthed rod
[7].
Figure 7.5.10 Influence of insulators on the 50% SI breakdown voltage

of a rod-rod gap [7]
The dependence of the critical time to peak with the minimum flashover voltage from the homogeneity of the
electric field is obvious and can be clearly shown by investigations conducted by the Les Renardieres Group [17].
Figure 7.5.11 shows the difference in the behaviour of a 10 mm tip in comparison to a large sphere with 0,75 m
diameter, both with a gap distance of 10 m. The tip shows the increase in the time-to peak for the minimum
flashover voltage with a front time of about 800 s, since the flashover voltage of the sphere is not influenced by
the time parameter in the interesting range.
Conclusion out of this is that for realistic gap designs the critical time-to-peak will primary tend to lower times,
because for very and ultra high voltage applications the aim to get a most possible homogeneous field can be
assumed, because of the higher withstand capability.
Page 225
TypethenameofTB
Gap:
10 m
Sphere: 0,75 m radius
Cone: 10 mm tip
Figure 7.5.11 Dependence of U50 from the time to peak for 10 m gap with the
electrode tips of a cone (10 mm Tipp) and a sphere (with 0.75 m diameter) [17]
Page 226
TypethenameofTB
7.6 Recent investigations on air gap clearance in the UHV range

Withstand and flashover characteristics are used for the design of external air gap insulation of lines in the
EHV/UHV range. The flashover voltage depends on the shape of impulse i.e. the time to peak and on the gap
configuration. A methodology to determine UHV air clearance is described in paragraph 7.6.1 according to Indian
air-gap studies. Further recent studies, carried out in China and Japan, are described in 7.6.2 and 7.6.3
respectively.
7.6.1 AIR GAP CLEARANCE STUDIES- INDIAN EXPERIENCE

The first stage of experiment deals with an approximate representation of the air gap, like clearance between a
tower and a conductor instead of V-insulator strings. Positive polarity switching impulse voltages with different time
to peak value (Tpcr) need to be applied under dry condition to obtain 50% flashover voltages. One would be the
standard switching impulse voltage of 250/2500 s. The 50% flashover voltages obtained need to be corrected for
reference atmospheric condition considering the atmospheric correction factor as per IEC-60060-1 [11]. From the
result, the minimum value of the 50% flashover voltages and corresponding value of time to peak of the waveform
i.e. critical time to peak (Tpcr) can be determined. Thus the optimum shape of the wave and clearance can be found
out.
The critical wave shapes obtained in the above tests can be used to obtain the optimum clearance corresponding
to the specified SIWV, with a more accurate representation of the air gap, like full window and suitable insulator etc.
The withstand voltages have been determined from the 50% flashover voltage with U10 = U50 (1 1.3 ) and
checked in experiments for different air-gaps with the specified SIWV. The gap will be considered as minimum, if it
meets the criterion of 2 or less flashovers out of 15 impulses. The gap clearance was than adjusted depending on
the above results.
Finally, with the insulator string in place, wet test simulating rain condition were carried out to ascertain the
withstand ability of the optimized tower window clearances. Withstand test with negative polarity switching were
also carried out under wet condition for the required test voltage.
At CPRI, Hyderabad, India, the field tests were carried out for optimization of conductortower air insulation
clearances for the 1200 kV system and corresponding to switching overvoltage of 1800 kV. The tests were carried
out on full-scale model UHV tower windows to obtain the critical time to peak for which 50% flashover voltages was
obtained for the estimated gap distances. Optimal window clearances of tower were determined at these gap
distances and adequacy of the optimized clearances was verified for the standard switching impulse withstand
voltage of 1800 kV required for the 1200 kV transmission lines.
The field study involved three types of tests:
(i)
Determination of 50% flashover voltages with Tower Top and Side
(ii)
Determination of 50% flashover voltages with Insulators and Tower Window
(iii) Withstand tests.

The test facility and the insulator string are shown in Figure 7.6.1. The test results are described in the following.
Page 227
TypethenameofTB
Figure 7.6.1 A view of mock up tower test facility (left) and a view of the insulator
strings (right)
7.6.1.1 TESTS WITH TOP AND SIDE TOWER
The conductor (eight BERSIMIS bundle with 35 m length) was strung between the end towers of the mock up tower.
The top and side tower was set up with MS and Aluminum channels, with a width of 2 m and a length of 18 m for
the top tower while a width of 5 m for the side tower. The minimum height of the bundle conductor above earth was
kept at about 20 m to avoid any disruptive discharges to earth or any other earthed object other than the top and
side tower. The arrangement for the setup is shown in Figure 7.6.2. The gap between the conductor-side tower and
conductor-top tower was maintained the same. The experiment was carried out with gap clearance 7 m and 10 m
under dry conditions and 50% flashover voltage were obtained with switching impulse voltage of positive polarity
and different times to peak in order to obtain critical time to peak value (Tpcr) for which the 50% flashover voltage is
minimum. up and down method as given in IEC-60060-1 with total number of impulses more than 20 and using one
level of test voltages in steps of less than 3% is used. Four wave shapes with different time to peak including
standard switching impulse voltage wave shape of 250/2500 s were used. Time to half value or Tail time during all
these tests was around 2500 to 2700 s.
18m
5m
2m
Y = X = 7 to 10 m
18m
Conductor
Bundle
20m
Ground
Figure 7.6.2 Sketch of the Tower Top Arrangement
Page 228
TypethenameofTB
The 50% flashover voltage obtained during the test was corrected for reference atmospheric conditions considering
the atmospheric correction factor as per IEC 60060-1. Table 7.6.1 gives the corrected value of 50% flashover
voltage corresponding to different peak time (Tp) of the wave shape applied for 7 m and 10 m gap clearance. The
corrected value of 50% flashover voltage plotted against time to peak (Tp) for the gap distance 10 m and 7 m is
shown at Figure 7.6.3. The result for the per unit value of the conventional deviation was = 0,03.
Gap
Distance
(m)
10
Time
to
peak
Tp
(s)
Temp.
Dry
( C )
Temp.
Wet
( C )
Atm.
Pressure
(mm Hg.)
50%
Flashover
Voltage
U50%
(kV)
Atm.
Corr.
Factor
( Kt )
315
250
220
150
315
250
220
150
29
30
31
30
29
28
31
32
22
22
23
22
21
20
22
21
716
715
716
715
717
715
719
717
2552
2489
2447
2550
2125
2079
2059
2131
0.992
0.991
0.993
0.994
0.984
0.982
0.985
0.927
50%
Flashover
Voltage after
atm.
correction
U50% (kV)
2572
2511
2463
2565
2159
2117
2090
2185
Table 7.6.1 50% flashover voltage with Tower top and one side representation for
10 m and 7 m gap clearance
Figure 7.6.3 50% flashover voltage plotted as a function of Tp for the gaps of 10 m
and 7 m for Top and Side Tower Configuration
From the test results, it is seen that the critical time-to-peak for which the minimal 50% flashover voltages occurred
is between 220 to 230 microseconds, which is slightly lower than the 50% flashover voltage at standard peak time
of 250 s. The lowest value of 50% flashover voltages obtained was 2090 kV for an air gap clearance of 7 m. The
Page 229
TypethenameofTB
withstand voltage is determined as U10 = U50 (1 1.3 ) = 2008 kV where the per unit value of the conventional
deviation, has been considered as 0.03. Even for = 0.06 (IEC 60071-2), the calculated value is 1927 kV. This
value is much higher than the required switching impulse withstand voltage of 1800 kV.
7.6.1.2 TESTS WITH INSULATOR AND TOWER WINDOW

In this test, full window of tower was set up along with introduction of suitable V-insulator strings using polymeric
insulators. From the top and side tower test, the withstand voltage corresponding to flashover voltage for 7 m gap
clearance was obtained as higher than the required withstand level of 1800 kV. However, in this case, due to tower
window and insulator strings, it is expected to have a considerable reduction in the 50% flashover voltage.
Accordingly, gap clearances of 8 m and 8.5 m were selected for the full tower window.
In this case, the tower top and all other side dimensions were kept the same as given above. The arrangement of
the total set up is shown in Figure 7.6.4. The conductor was strung at a height of 23 m above earth. Tests at the
bottom of the tower window were conducted at 11.5 m below the conductor and the distance between the bottom
window and the earth was 11.5 m.
Table 7.6.2 gives the results of 50% flashover voltage values obtained for 8 m and 8.5m gap distance and the
corrected 50% flashover voltage under reference atmospheric conditions.
Figure 7.6.4 Tower window for 1200 kV full-scale model

Gap
Distance
(m)
8
8.5
Time to
peak
(Tp)
Temp.
Dry
bulb
Temp.
Wet
bulb
(s)
160
250
270
350
115
250
270
350
( C )
22
27
27
24
29
25
25
29
( C )
16
17
17
17
19
19
19
20
Atm.
Pressure
(mm.
of Hg.)
718
713
713
718
716
716
716
717
50%
Flashover
Voltage
Atm.
Corr.
Factor
(kV)
1926
1879
1868
1854
2191
1950
1937
1948
(Kt)
0.987
0.94
0.94
0.982
0.98
0.992
0.992
0.988
50% Flashover
Voltage
(corrected)
U50
(kV)
1951
1915
1906
1910
2236
1963
1953
1961
Table 7.6.2 50% flashover voltages as against air gap clearance of 8 m and 8.5 m
Page 230
TypethenameofTB
Figure 7.6.5 shows the 50% Flashover voltage, corrected for atmospheric conditions, plotted as a function of Tp for
conductor-insulator string-tower window configuration and the gap distances 8 m and 8.5 m. The flashover voltages
were minimum for times to peak of 270 s and above.
Figure 7.6.5 50% flashover voltage plotted as a function of Tp for the gaps of 8 m
and 8.5 m conductor-insulator string-tower window configuration
From the experimental results for window clearance of 8 m the 50% flashover voltage obtained was 1906 kV for
which the withstand voltage determined as U10 = U50 (1 1.3 ), for = 0.03 will be 1832 kV. This level is just
about 1.75% higher than the 1800 kV rated standard Switching Impulse Withstand Voltage.
7.6.1.3 WITHSTAND TESTS

To ascertain the above, 15 positive polarity switching impulse voltages, corrected for ambient atmospheric
conditions corresponding to the rated standard Switching Impulse Withstand Voltage of 1800 kV, were applied for
the air gap clearance of 8 m. No flashover was observed. The test was repeated with reduced gap clearance to 7.5
m and four flashovers were observed out of 15 impulses applied giving very low withstand probability for 1800 kV
test voltage. The withstand voltage test was also carried out under wet condition with air gap clearance of 8 m. Test
was also carried out with negative polarity switching impulse for the same voltage and wave shape. No flashover
was observed in either case.
Thus, the gap distance of 8 m seems to be the optimized air clearances for tower window configuration in 1200 kV
transmission lines corresponding to the SIWV of 1800 kV. For substation, to be on the safe side, the optimum air
gap clearance was considered same as IEC value i.e. 8.3 m, which is about 3.75% higher than that for
transmission line i.e. 8 m.
Page 231
TypethenameofTB
7.6.2 AIR GAP CLEARANCE STUDIES - CHINESE EXPERIENCE

In order to optimize phase-to-earth and phase-to-phase air-Clearance for 1100 kV power transmission line, the
SGEPRI of China have carried out lots of flashover characteristic tests with full scale mock up tower and real tower.
In the past 3 years before foundation of the Chinese UHVAC pilot project, the research mainly focused on real
tower test. By a large number of air-gap clearance flashover tests, many important discharge curves were obtained,
which is greatly useful for the line tower design. The SGEPRI of China presents here representative results of tests
with UHV long clearance during recent years. Long air-gap 50% flashover characteristic curves were obtained with
real tower structure and fittings of Chinese pilot single-circuit UHV-AC transmission line, which includes lightning
impulse, power frequency voltage and switching impulse of different times to peak. The 50% phase-to-earth
flashover voltage curves of switching, lightning and power frequency voltage, and the 50% phase-to-phase
switching impulse flashover voltage curves were obtained with full scale external air-gap insulation of Chinese pilot
UHV substation apparatuses and bus bar.
During the tests, different times to peak switching impulse and the 1.2/50 s lightning impulse are generated by the
5400 kV / 527 kJ impulse voltage generator. The measuring system error of impulse crest value is lower than 3%.
The impulse voltage generator is controlled by computer through optical fiber cables with digitally charging,
triggering and recording function. This system has the high abilities to prevent interference and the test results are
repeated well. The power frequency voltage and switching impulse with 3000~5000 s times to peak are generated
by the 2250 kV power frequency test transformer. The capability of the test transformer is 9000 kVA. The power
frequency voltage and switching impulse measuring error for this test transformer measuring system is lower than
3%.
In the power frequency flashover tests, the number of flashovers for each one case is 5. The average of the 5
flashover voltages was taken as the flashover voltage of the conductor-tower air-gap clearance. In the impulse
voltage test, the up and down method specified by IEC-60060-1 was used. The number of impulses applied on the
air-gap was respectively 40 and 20 for each switching impulse and lightning impulse test. According to wave-shape
and air-gap distance, step voltage was 3% in lightning test and 4% in switching impulse test.
The 50% flashover voltage of air-gap clearance was corrected to the standard atmospheric conditions specified in
IEC-60060-1, and then they are plotted as flashover curves for the different test voltage types.
7.6.2.1 FLASHOVER CHARACTERISTICS TEST OF CAT-HEAD-TOWER SIDE PHASE

The arrangement of side phase air-gap test with 1100 kV UHV cat-head-tower is shown in Figure 7.6.6. The
suspension string (also called I string) consists of single composite insulator, conductor bundle and two grading
rings. Design length of the composite insulator Is 9.75 m and the air-gap between the two rings installed on the
insulator is 8.6 m. The conductor bundle is simulated by a cage of 8 steel bars, with length of 20 m.
According to insulation coordination standard and tower design procedure, the windage yaw angle of the
conductor is specified differently under the power frequency operating voltage, the switching impulse voltage and
the lightning impulse voltage due to consideration of correlation probability of one voltage type and the related wind
speed, e.g. the maximum wind speed of 30 m/s is considered to determine angle for power frequency operation
voltage.
In the test, the insulator and conductor bundle were pulled to the tower side and fixed with ropes to simulate the
windage yaw angle . There is little difference of the height h and the tower width w (width of tower body nearest to
the conductor bundle) for power frequency, switching impulse and lightning impulse test. Power frequency voltage
or impulse voltage was injected to the conductors. Figure 7.6.7 is the typical side phase air-gap flashover picture.
Page 232
TypethenameofTB
Figure 7.6.6 Arrangement of side phase air-gap test with I insulator string
Figure 7.6.7 Typical flashover picture for side phase air-gap test with
UHV cathead-tower
Page 233
TypethenameofTB
In the power frequency test, the gap distances between conductor bundle and tower were varied from 2.0 m to 4.5
m and the windage yaw angle of I-insulator string was fixed to 48. The height h of the conductor was 42 m. The
tower width w was 5.5 m. In the switching impulse tests, the gaps were varied from 4.5 m~8.2 m and the windage
yaw angle of I-insulator string was fixed to 18. The height of the conductor h was 38.5 m. The tower width w was
6.5 m. In the lightning impulse tests, the gap distances were varied from 4.0 m to 6.5 m and the windage yaw angle
of I-insulator string was fixed to 10. The h is 38 m and the w is 6.6 m.
The 50% flashover voltage curves as a function of different gap distances are shown from Figure 7.6.8 to Figure
7.6.10.
The per unit value of the conventional deviation, of the positive switching impulse flashover voltage varied within
the range of 4.0%~7.0%; this value range was obtained by analyses of the test results with the 4.5 m~8.2 m air-gap
clearances under standard switching impulse flashover test. Furthermore the value was irregularly rising with
increase of air-gap distance, as well as with the front time of switching impulse voltage. Since the values of were
obtained by up-and-down method rather than multistage method, they are not accurate. Investigations on the
statistics of the conventional deviation of switching impulse voltage with long times to peak will be the future
research work.
1800
U50( kV)
1600
1400
1200
1000
800
2
2. 7
3. 5
d ( m)
4. 5
Figure 7.6.8 50% Flashover voltage curve of side phase air-gap for
UHV cat-head-tower configuration as a function of gap distance
under power frequency voltage
Page 234
TypethenameofTB
U50 ( kV)
2400
2000
1600
1200
4. 5
5. 6
6. 5
7. 5
8. 2
d ( m)
Figure 7.6.9 50% flashover voltage curve of side phase air-gap for
under 250/2500 s positive switching impulse
3600
U50 ( kV)
3200
2800
2400
2000
3. 5
4. 5
5. 5
6. 5
d ( m)
Figure 7.6.10 50% flashover voltage curve of side phase air-gap for
under standard lighting impulse
7.6.2.2 FLASHOVER CHARACTERISTIC TESTS OF CAT-HEAD-TOWER MIDDLE PHASE
Since switching overvoltage on middle phase conductor suspended on V-insulator strings controls the cat-headtower windows size, the flashover voltage characteristic curve as a function of different gap lengths and of different
times to peak with a fixed gap length were obtained by tests, as well as lightning impulse flashover voltage for
different gap lengths. The test set up arrangement is shown in Figure 7.6.11. The Figure 7.6.12 shows the typical
flashover path for middle phase air-gap test with the UHV cat-head-tower.
The 50% switching impulse flashover voltage characteristic curve as a function of different gap lengths shown in
Figure 7.6.13 was obtained under the condition that the gap clearance d1 between V-insulator string grading rings
and the tower top were varied from 5.0 m to 8.3 m, d2 from grading ring to the tower side and d3 from conductor
bundle to the tower bottom were fixed to 7.8 m and 7.3 m respectively.
Page 235
TypethenameofTB
Figure 7.6.11 Arrangement of middle phase air-gap test with V-insulator strings
Figure 7.6.12 Picture of the typical flashover path for middle phase air-gap test
with UHV cat-head-tower
Page 236
TypethenameofTB
U 50 (kV)
1850
1700
1550
1400
4
d (m)
Figure 7.6.13 50% flashover voltage curve of middle phase V-insulator string air-gap
for UHV cat-head-tower window configuration as a function of gap distance
under 250/2500 s positive switching impulse voltage
The comparison of Figure 7.6.9 and Figure 7.6.13 shows that the 50% flashover voltage of the middle phase airgap between grading ring and tower top is 13%~15% lower than side phase air-gap between conductor and tower
side. This phenomenon can be explained as follows; the middle phase V-insulator strings and tower window form a
close electric field distribution around the high voltage electrode, so, the electric-field strength on V-insulator string
high voltage electrode is higher than that of the side phase I-insulator string high voltage electrode.
The 50% switching impulse flashover voltage characteristic curve as a function of different times to peak shown in
Figure 7.6.14 was obtained under the condition that the gap spacing between conductor and tower window
crossbeam was 7.9 m and the minimum gap between conductor and the side or the bottom of tower window was
6.7 m.
The gap spacing between high voltage ring and grounding ring of the insulator was 8.6 m. During the middle phase
air-gap test, the numbers of flashovers in paths from high voltage electrodes to different parts of the tower window
were counted. The table 7.6.3 presents the statistical results of flashover numbers in the different paths under
different times to peak of positive switching impulse voltage test. The flashover numbers in the path from (a) ring to
tower top is 38.4% of total flashover numbers, (b) ring to tower side and conductor to bottom are 29.3%
respectively, and (c) high voltage ring to grounding ring is 3.0%. Therefore, the flashover number in the path from
ring to top window was slightly more than that to the other parts of tower window, and the air-gap flashover number
between high voltage ring and grounding ring of the insulator occurred less. The flashover number in the path from
ring to top window significantly decreased when the time to peak of switching impulse voltage increased.
Page 237
TypethenameofTB
S.N.
Time to
peak (s)
Ring to
Tower top
Ring to
Tower side
1
2
3
4
5
120
250
500
1000
5000
11
9
9
8
1
38
38.4%
3
4
5
6
11
29
29.3%
Percentage
Conductor to
Tower
bottom
5
7
6
4
7
29
29.3%
Ring to
Grounding
ring
1
0
0
2
0
3
3.0%
Total
flashover
number
20
20
20
20
19
99
100%
Table 7.6.3 Flashover number statistical result in the paths of middle phase Vinsulator string air-gap for UHV cat-head-tower window under different times to
peak of positive switching impulse voltage test
U 50
kV
2600
2200
1800
1400
10
100
1000
Tp
10000
Figure 7.6.14 The 50% flashover voltage curve of middle phase V-insulator string
air-gap for UHV cat-head-tower configuration
as a function of time to peak of positive switching impulse
7.6.2.3 PHASE-TO-PHASE AIR-GAP FLASHOVER VOLTAGE TEST WITH UHV
TRANSMISSION LINE CONDUCTOR BUNDLES
The phase-to-phase air-gap flashover voltage test configuration is shown in Figure 7.6.15, the two 8-sub conductor
bundles representing two phase electrodes are horizontally suspended 18m high above earth. The two bundles
were injected separately with 250/2500 s positive and negative switching impulse voltage generated by 5400 kV
and 3000 kV impulse generators. The maximum negative switching impulse voltage is adjusted to maintain the
ratio = U- / (U+ + U-)=0.4 during test. The 50% standard phase-phase switching impulse flashover voltage
characteristics curve as a function of gap distance, which is varied from 5 m to 9 m, is shown in the Figure 7.6.16.
Page 238
TypethenameofTB
Figure 7.6.15 Arrangement of middle phase-to-phase test
3500
U50( kV)
3000
2500
2000
1500
1000
3
10
d ( m)
Figure 7.6.16 Phase-phase 50% standard switching impulse flashover voltage

characteristic curve as a function of gap distance
between two phases of UHV line conductor bundles
Page 239
TypethenameofTB
7.6.3 AIR GAP CLEARANCE STUDIES - JAPANESE EXPERIENCE

Fundamental properties for air gap clearance were investigated with a longer air gap and compare the values with
minimum phase-earth clearance shown in IEC 60071-1 Amd.1. The flashover tests were carried out using rodplane electrodes as the worst case to provide the flashover voltage. In some past studies, elderly calculation
method for time to peak of switching impulse was adopted and that was different from present method shown in
IEC 60060-1, 61083-1 and 61083-2. Moreover, the past time to peak data [1, 11-14] measured by using analogue
oscilloscope are supposed to include some measurement errors of reading by a measurer especially for the shorter
times to peak measured. For the accurate measurement of the parameters, the impulse waveforms in the tests
were measured by a digital recording system in accordance with IEC 61083-1 and 61083-2 [15-16].
Figure 7.6.17 shows the experimental setup. A brass rod with 15 mm square was employed as H.V. electrode. The
gap length was changed in the range from 4 to 10 meters. The tests under dry condition were carried out with up
and down method in accordance with IEC 60060-1. Step voltage was set below 2% to the flashover voltage and
total number of impulse more than 40 was applied in each case.
The photograph in Figure 7.6.17 expresses an example of photograph of flashover. Positive switching impulse
applied to H.V. electrode and the time to peak of the waveform was varied from 70 s to 800 s to find out the
critical time to peak. The time to half values of waveform tails was set to around 2500 s in all cases.
Positive switching impulse

H.V. electrode
(Rod)
Earth electrode
(Plane)
Gap length d
Figure 7.6.17 Experimental rod-plane setup and photograph of flashover

Figure 7.6.18 shows 50% flashover voltages and their deviations at each air rod-plane gap clearance and time to
peak. The data in the figure are corrected with considering atmospheric conditions. As the tests were performed,
conditions of absolute humidity were close to standard reference condition 11 g/m3 given in IEC 60060-1. The
results in the figure suggest as follows.
The characteristic of the 50% flashover voltages shows that each curve has a minimum value. These minimum
corresponding to the critical times to peak of the air rod-plane gap clearances becomes more pronounced with
longer air gaps.
The critical time to peak increases with gap length. They are 200 s for 4 m gap, 400 s for 8 m gap and 450 s
for 10 m.
The per unit values of the conventional deviation of flashover voltages is within around 5%, which is slightly lower
than the 6% recommended in IEC 60071-2 [10].
The per unit values of the conventional deviation of flashover voltages tends to be larger for longer air gap.
Page 240
TypethenameofTB
2200
d = 10m
Flashover voltage (kV)
2000
1800
d = 8m
1600
Critical time to peak
1400
d = 4m
1200
1000
10
30
100
300
1000
Time to peak Tf s
Figure 7.6.18 Curves of U50 flashover voltage variation for a rod-plane gap
of different clearances with time to peak of positive switching impulse
7.6.4 COMPARISON OF PAST AND NEW INVESTIGATIONS
Flashover characteristics of long air gap have been investigated under positive switching impulse with rod-plane
electrodes. Figure 7.6.19 shows time to peak Tp dependence of 50% probability flashover voltage for different air
gap lengths under dry conditions by using rod-plane electrodes. The data includes new experiments as shown in
Figure 7.6.18, as well as past experimental data of references [1, 11-14]. The flashover characteristic has the
critical time to peak for which the 50% flashover voltage is the minimum. The critical time to peak increases with the
gap length.
Figure 7.6.20 shows 50% flashover voltage for the critical time to peak and critical time to peak as a function of air
gap length. These experimental data of rod-plane electrodes are read off from Figure 7.6.19. The 50% flashover
voltages of the data agree well with formula (1), especially in the region of gap lengths larger than 8 meters. The
dotted curve is estimated by values of critical time to peak with rod-plane electrodes. Under around 5 or 6 m gap
length, EHV length range, the critical time to peak of rod-plane air gaps becomes the standard value 250 s. While
for the UHV length range (more than about 6 m length), the critical time to peak of rod-plane air gaps became
larger than 300 s.
In Figure 7.6.20 the data of India and Chinese actual UHV towers are also plotted. These experimental data are
shown in Figure 7.6.3 and Figure 7.6.5 for the Indian test results and in Figure 7.6.9, Figure 7.6.13 and Figure
7.6.16 for the Chinese test results.
Tpcr values of UHV tower measured recently in China and India are plotted at the little lower point of the fitting curve
in the figure. This tendency is similar to the past data of parallel conductor and V-insulator string for longer gap
lengths such as in Figure 7.5.4 and Figure 7.5.7. The geometry of air gap electrodes affects the critical time to peak
as well as the 50% flashover voltage. But for the UHV air gap lengths adopted in India and China the critical time to
peak of such actual Tower or substation configurations remain close to the 250 s associated to the Standard
Switching Impulse.
Page 241
TypethenameofTB
4000
U50 Flashover voltage (kV)
3500

3000
Past experimental
data *1
d=25m
2500
2000
d=10m
New experimental
data *2
1500
d=4m
1000
d=2m
500
0
10
100
1000
10000
Time to peak Tp (s)
Figure 7.6.19 Relationship between time to peak and 50% flashover voltage for
rod-plane gaps of lengths between 0.5 m to 25 m (*1 past and *2 new
experimental data)
*1
*1
*1
*2
*2
2400
Formula (2)
[Paris's formula]
2200
2500
2000
Formula (1)
[Critica sparkover voltage]
2000
1800
1600
U50 Flashover voltage (kV)
3000
Rod-plane, Past data

(Rod-plane, Past data)
(Rod-plane, New data)
(UHV tower)
(UHV tower)
1400
1500
1200
1000
1000
800
600
500
400
Critical time to peak Tcr (s)
50% Flashover voltage

Japanese experiments
Indian experiments
Chinese experiments
200
0
0
0
10
12
14
16
18
20
22
24
26
28
Gap length d (m)

Figure 7.6.20 Relationships between gap length, 50% flashover voltage and critical
time to peak (*1 new and past data from rod-plan gaps; *2 data from UHV towers)
Page 242
TypethenameofTB
7.6.5 COMPARISON OF MEASUREMENTS AND IEC GAP-CLEARANCES

For the comparison of the IEC minimal UHV gap clearance values with the measurements on actual configurations
in India and China the difficulty is that IEC gives withstand voltages and the testing results provide U50 values. To
permit a direct comparison of IEC and Indian and Chinese results the calculation of U10 values include the
uncertain knowledge of the deviations as described in chapter 7.4.3.
Assuming that the per unit value of the conventional deviation of 6% will lead to a reliable design some U50
measurements have been transferred to U10 values by formula (3) of chapter 7.4 and compared with the given
minimal gap clearances in IEC 60071-1 for UHV Rod-Structure and Conductor-Structure configurations. Only the
results of switching impulses with a time to peak in the range of 220 s to 270 s have been used, that means they
are within the tolerances of the standard wave shape.
For the comparison with the Conductor-Structure gap-clearances of IEC only the values were chosen where the
conductor against the tower was tested and no insulators were involved (Figure 7.6.21). The following results have
been used:
from the Indian measurements, the values of the Tower-Top and -Side investigations according to Table 7.6.1
(with Tp = 220 s and 250 s)
from Chinese measurements, the values of the Tower-Side Phase investigations according to Figure 7.6.9 (all
with Tp = 250 s).
Despite the uncertainties in this consideration and the low number of test series it can be concluded that the
Conductor-Structure clearances are rather close to the measurements and it supports the validity of the clearances
selected in IEC 60071-1 for the UHV levels.
2400
2200
2000
U10 [kV]
1800
1600
ph-e_Conductor-Structure (IEC)
1400
Tower -Top/-Side (India)

Tower-Side Phase (China)
1200
1000
4000
5000
6000
7000
8000
9000
10000
d [mm]
Figure 7.6.21 Calculated U10 values (from U50 measurements by means of s = 6 %)

of Conductor-Tower arrangements in comparison to the IEC withstand voltages
associated with minimal gap clearances for Conductor-Structure configurations
versus the gap clearance
Page 243
TypethenameofTB
In the same way as before (U50 to U10 and only standard wave shapes) the results from the configurations with
tower windows have been gathered in Figure 7.6.22. Here insulators were partly involved. The following results
have been used:
from Indian measurements, the Tower Window values according to Table 7.6.2 (with Tp = 250 s and 270 s)
from Chinese measurements, the Tower-Middle Phase values according to Figure 7.6.13 (all with Tp = 250 s)
Figure 7.6.22 shows much lower U10 values in comparison to the conductor-tower configurations of Figure 7.6.21.
The reason for that may be manifold. There is a more concentrated electrical field around the conductor due to the
tower window which leads to a higher electric-field strength on the high voltage electrode. Additionally there were
insulator strings in the windows; but it must be seen that the insulators only in some cases were directly involved in
the flashovers. Nevertheless they are a further increase in field-stress due to the armature and fixing points of the
insulators on the conductor.
The lower U10 values of the Tower Window configurations are compared in Figure 7.6.22 with the clearances for
the inhomogeneous Rod-Structure air gaps in IEC 60071-1. In the upper range the Chinese measurements are
slightly below the Rod-Structure values, but the Indian measurements fit very well. Therefore taking the RodStructure values for most unfavorable configurations should be handled with care. But this statement should not be
generalized due to the fact that there are not enough test results and due to the above mentioned uncertainty of
these considerations.
2400
2200
2000
U10 [kV]
1800
1600
ph-e_Rod-Structure (IEC)
1400
Tower Window (India)

Tower-Middle Phase (China)
1200
1000
4000
5000
6000
7000
8000
9000
10000
d [mm]
Figure 7.6.22 Calculated U 10 values (from U 50 measurements by means of s = 6 %)

of Tower Window arrangements versus the gap clearance. For comparison the IEC
UHV withstand voltages associated with minimal gap clearances for Rod-Structure
configurations are also shown
Page 244
TypethenameofTB
7.7 Non standard switching impulse waveforms in the insulation coordination

7.7.1 DISCUSSION ON CRITICAL TIME TO PEAK OR STANDARD SWITCHING IMPULSE
VOLTAGE
(PARIS VERSUS WATANABE)
In paragraph 2.3.8 page 60 of CIGRE Technical Brochure 400 [19], the following critic is mentioned about minimum
air gap clearances of IEC 60071-1 Amendment 1:
Since Paris equation (1) is made by relatively short gap (3~4m), this equation is not always enough for
very long gap in UHV system. This means present standard clearance (IEC 60071- Amd.1 Ed.8.0 CDV) is
not enough for long gap (Particularly the gap more than 10m).
The following tries to address, within the frame of Insulation Coordination, the difficulties if distances larger than 10
meters are really judged necessary. The conclusion of this chapter answers the adequacy or not of the present
standard clearance(s).
As already mentioned, in IEC 60071-1 amendment 1, the PARIS formula (equation 2; Chapter 7.2.2) was used to
express the UHV minimum air clearances for Rod-Plane gap valid up to 25m and standard rated SIWV with 250 s
time to peak.
The references [1,2] of Paris formula shows front time of impulse wave applied is only 120 s and Gap length is 2
to 7 m. Further references [5, 7] show additional experiments and comments that the Paris formula is applicable to
gaps of up to 17 m with 250 s front time impulse. More studies [4-7] and Chapter 4.3 show that the front duration
of switching surges in long length UHV systems tends to become longer than that in lower voltage systems. When
the length of a transmission line is about 100km and over, the switching surge with the long fronts of 300 s and
over may occur. This is discussed and shown in various figures at Chapter 4.3.
From various past data [Chapter 7.5] it may be concluded that for practical phase to earth insulations that are
involved for overhead lines or substations, the critical time to peak can be approximated by Tpcr = 35 d. As the
distances associated in IEC with the UHV standard switching impulse withstand voltages are lower than 9.5 m
[Table 7.3.1] their critical time to peak is of 332 s at maximum for the highest standard switching impulse
withstand voltage of 1950 kV.
The ratio of withstand voltage for Standard Switching Impulse (derived with Paris formula, Eq-1) and Critical time to
peak impulse (derived with Watanabe formula, Eq-2) versus the air gap spacing is shown in Figure 7.7.1.
According to Figure 7.7.1, for such distance range (9.5 m) the ratio between the withstand voltage for critical time
to peak switching impulse and standard switching impulse is at the maximum of 1.06. Beside this, in substation the
presence of post insulators flattens the U curves (Chapter 7.5).
Page 245
TypethenameofTB
Figure 7.7.1 Ratio of the withstand voltages for Standard Switching Impulse
(derived with Paris formula) and
Critical time to peak impulse (derived with Watanabe formula)
versus the air gap spacing
For phase to phase insulation involved in UHV overhead lines or substations, the critical time to peak is in the
range of 250 s whatever the air gap distance is. So for phase to phase distances of IEC there is no
underestimation of the distance needed.
In order to take into account the 1.06 difference for phase to earth air gap; instead of increasing the specified
substation or overhead line distances, post type insulators may be applied which flatten the U curve and reduces
the flashover voltage. Otherwise, a test could be carried out using rated standard switching impulse voltage.
7.7.2 HOW TO HANDLE NON STANDARD WAVEFORMS FOR INSULATION

COORDINATION PROCESS
The extract of the definitions and process of insulation coordination, as given in IEC 60071-1, is summarized in
Chapter 2 Concept of recent practice on insulation coordination and Chapter 8.1 Procedure for insulation
coordination. For convenience of reading the following some abbreviations are given here:
Urp :
Representative Overvoltages
Ucw :
Co-ordination Withstand Voltage
Urw :
Required Withstand Voltage
Ks :
Safety Factor
Kc :
Co-ordination Factor
Page 246
TypethenameofTB
The relations between the above parameters may be represented as below:

Ucw = Kc Urp
Urw = Ks Ucw = Ks Kc Urp
Based on the value of Urw, the rated or standard insulation level (Uw) is selected.
Selection of requested and standard rated SIWV for UHV insulation according to actual time to peak Tp of
overvoltages in the system. According to the distances that are really needed for actual air gap insulations the
insulation coordination process is described below for the example of a rod plane gap installed at sea level without
any surge arrester protection.
For actual distances above 10 meters:

U2 2% actual overvoltage in the system has
critical time to peak Tpcr for UHV clearances (400
to 500 s for 12-17 meter clearances) then:
Urp1/SI = Ktc (SSI/SITpcr) U2
with Ktc(SSI/SITpcr) "conversion factor" between
SI voltage with critical time to peak (SITpcr) and
the Standard SI voltage (SSI): 1.1 to 1.2 for 13
meters (Figure 7.7.1) :
Urp1/SI = (1.1 to 1.2) U2
For actual distances up to 10 meters:

U2 2% actual overvoltage in the system has time
to crest of 250 s (standard switching: critical
time to peak for clearances up to ~10 m) then:
Urp2/SI = Ktc(SI/250s) U2
with Ktc(SI/250s) = 1
For failure rate of 10-3, co-ordination factor Kcw =

1.1 so the needed coordination withstand
voltage Ucw1:
Ucw1 = Kcw Urp1/SI =
Ucw1 = 1.1 Ktc(SSI/SITpcr) U2
For failure rate of 10-3, Kcw = 1.1 so the needed

coordination withstand voltage Ucw2:
Ucw2 = Kcw Urp2/SI
Ucw2 = 1.1 U2
In air for a safety factor of Ks=1.05 and a needed

required withstand voltage Urw1:
Urw1 = 1.05 Ucw1
Urw1 = 1.05 1.1 Ktc(SSI/SITpcr) U2
Urw1 = 1.05 1.1 (1.1 or 1.2) U2
Urp2/SI = U2
In air for a safety factor of Ks=1.05 and a needed

required withstand voltage Urw2:
Urw2 = 1.05 Ucw
Urw2 = 1.05 1.1 U2
NOTE: Usually the Standard Switching Impulse is referred as SI but in this table to explicitly refers to the
Standard Switching Impulse an S has been added before: SSI. This is to clearly makes here (paragraph 7.7.2)
the difference with switching impulses of different time to peak (and even time to half value) called here SITpcr
as they refer to the critical time to peak of a given air gap.
In conclusion for the above example in case of a rod-plane gap configuration if distances above 10 meters are
actually needed for that gap type insulation then:
Urw (critical time to peak > 250 s) = (1.1 or 1.2 ) Urw (250 s time to peak)
dmin(critical time to peak > 250 s) = (1.17 or 1.35) dmin (250 s time to peak)
Of course in case of air gap configurations having Ktc(SSI / SITpcr) "conversion factor" between voltage
with critical time to peak (SITpcr) and the Standard SI voltage (SSI) lower than 1, the conclusion would be of
course the reverse in that cases.
Page 247
TypethenameofTB
7.8 Conclusion
In the following the conclusions out of Chapter 7 are gathered as hints and recommendations for IEC TC 28
Insulation Coordination concerning the special aspects of air gaps in the UHV range.
In this Chapter 7 the knowledge about air gap clearance and flashover characteristics of switching impulses is
refreshed, based on IEC Insulation Coordination and studies from the past. Recent investigations give some new
input. The only relevant voltage stress for air gaps in the UHV range is the positive switching impulse voltage.
Some recent investigations from Japan, India and China are presented.
Japanese investigations support the accuracy of Equation (1) for the calculation of the 50% flashover voltages for
rod-plane gaps with a decrease of the flashover voltage with increasing time to peak.
The systematic Indian investigations on a tower configuration show that the selected gap distance of 8 m is slightly
below the rod-structure value of the IEC-Table value of 8.3 m to meet the required switching impulse withstand
voltage of 1800 kV. The conductor-structure value of the IEC-Table with 6.2 m seems to be too small for general
dimensioning.
The Chinese results indicate that for a tower configuration with V-insulator-strings it is sufficient to consider the
standard impulse wave shape to find the lowest flashover voltage for the investigated air gap in the tower. Longer
times to peak lead to higher flashover voltages.
Transferring the U50 measurements from India and China to withstand voltages U10 by calculation and comparing
these results with the recommended UHV minimal air gap clearances of IEC 60071-1 show that the ConductorStructure clearances are rather close to the measurements and it supports the validity of the minimal air gap
clearances selected in IEC 60071-1 for the standard UHV switching withstand voltages. The much larger minimal
clearances for UHV Rod-Structure air gaps in IEC fit more or less to the measurements of tower window
configurations, but such a correlation should be handled with care due to various parameters which cannot be
figured out from these measurements. Special care has to be taken also for these comparisons because of the
uncertainties due to the calculation and the low number of test series.
The u-curve behavior of flashover voltages with non standard switching impulses is valid for rod-plane
configurations. From investigations in the past as well as from the recent measurements from India and China it
can be seen that configurations with a more homogeneous field distribution and configurations including insulators
do not show this behavior. For actual configurations in a tower with and without insulators the critical time to peak is
near to the peak of the standard wave shape. The necessity to use impulses with longer front times in tests seems
not to be compulsory so as for the calculations.
If longer front times are expected and unfavorable configurations have to be considered the relation between
standard and non-standard wave shape can be included in the insulation coordination process by means of the
coordination factor Kc.
The IEC-equations for the calculation of gap clearances include uncertainties in the gap factor K and the per unit
value of the conventional deviation . The gap factor can only be estimated with low accuracy based on the
examples given in IEC and the literature. The influence of the selected deviation cannot be neglected but the
recommended IEC value of 6% may be in most cases on the safe side. Therefore the calculations can only give
some support. For to achieve a reliable gap design tests are necessary.
Page 248
TypethenameofTB
7.9 References
1. L. Paris,Influence of Air Gap Characteristics on Line-to-Ground Switching Surge Strength, IEEE Trans. on
Power Apparatus and Systems, vol. 86, no.8, pp. 936947, 1967.
2. L. Paris and R. Cortina, Switching and lightning impulse discharge characteristics of large air gaps and long
insulator strings, IEEE Transactions on Power Apparatus and Systems, Vol. PAS-87, No.4, April 1968, pp. 947-957.
3. W.C. Pokony and R.W. Flugun, UHV tower insulation parameters determined by full scale testing, IEEE
Transactions on Power Apparatus and Systems, Vol. PAS-94, No.2, Mar-April 1975, pp. 518-525.
4. G. Gallet., G. Leroy, R. Lacey and I. Kromer, General expression for positive switching impulse strength valid up
to extra long air gaps, IEEE Transactions on Power Apparatus and Systems, Vol. PAS-94, No.6, Nov./Dec, 1975,
pp.1989-1993.
5. A. Pigini, G. Rizzi, R. Brambilla, and E. Garbagnati,Switching Impulse Strength of Very Large Air Gaps, 3rd
I.S.H. Milan 1979, article 526-15.
6. I. Kishizima, K. Matsumoto, Y. Watanabe, New facilities for Phase-to-Phase Switching Impulse Tests and Some
Tests Results, IEEE Transactions on Power Apparatus and Systems, Vol. PAS 103, No. 6, June 1984, pp. 12111216.
7. R. Cortina, E. Garbagnati, A. Pigini, G.Sartorio and L.Thione, Switching impulse strength of phase-to-earth UHV
external insulation - Research at the 1000kV project, IEEE Transactions on Power Apparatus and Systems, Vol.
PAS-104, No.11, November 1985, pp. 3161-3168.
8. Yoshibumi Yamagata, Akio Oe, Kunihiko Miyake, Yoshinori Aihara, and Takatoshi Shindo, Phase-to-ground and
phase-to-phase sparkover characteristics of external insulation at the entrance of a UHV substation, IEEE
Transactions on Power Delivery, Vol.17, No.1, January 2002. pp. 223-232.
9. IEC 60071 : Insulation co-ordination : Part 1: Definitions, Principles and Rules. Edition 8.1 - 2011-03
10. IEC 60071-2 Insulation co-ordination : Part 2: Application guide, Edition 3, 1996-12.
11. I.Kishizima, et al "New facilities for phase-to-phase switching impulse tests and some test results, IEEE,
Transactions on, Power Apparatus and Systems, Volume: PAS-103 Issue:6 pp.1211 1216, (1984)
12. R.Cortina et al.; Switching Impulse Strength of Phase-to-Earth UHV External Insulation -Research at the 1000
kV Project-, IEEE Transactions on, Power Apparatus and Systems, Volume: PAS-104, Issue:11, pp.3161-3168,
(1985)
13. R.Cortina et al.; Air Density Influence on the strength of external insulation under positive impulse:
Experimental investigation up to an altitude of 3000m a.s.l., IEEE Transactions on, Power Apparatus and Systems,
Volume:5, Issue:2, pp.730-737 (1990)
14. G.Galet et al.; General expression for positive switching impulse strength valid to extra long air gaps, IEEE
Transaction on Power Apparatus and Systems, Volume: 1. PAS-94, No.6, (1975)
15. IEC 61083-1 (2001) Instruments and software used for measurements in high-voltage impulse tests, Part 1:
Requirements for instruments
16. IEC 61083-2 (1996) Digital recorders for measurements in high-voltage impulse tests-, Part 2: Evaluation of
software used for the determination of the parameters of impulse waveforms
17. Les Renardieres Group: Research on long air gap discharges at Les Renardiers - 1973 Results; Electra No 35;
1974
Page 249
TypethenameofTB
18. CIGRE Technical Brochure 362;Technical Requirements for Substation Equipment Exceeding 800 kV: Field
experience and technical specifications of Substation equipment up to 1200 kV, Working Group A3.22; December
2008
19. CIGRE Technical Brochure 400; Technical requirements for substations exceeding 800 kV; Working Group
B3.22; December 2009
Page 250
TypethenameofTB
8 Selection of insulation levels

8.1 Procedure for Insulation coordination
Clause 8.1.1 outlines the procedure for the insulation coordination of IEC 60071-1 [1], and Clause 8.1.2 introduces
the computation of the lightning and switching impulse withstand voltages in UHV ac systems based on the
protection level of surge arresters in accordance with the insulation coordination procedure. A simplified method
described in IEC 60071-2 [2] is used here.
In the process of determining the insulation level, while various factors must be considered in the insulation
coordination procedure, Clause 8.1.3 focuses on the safety factor. Firstly, the inquiry results on the safety factor for
UHV systems in various countries are reported, followed by a brief overview of the process of how the safety factor
has been studied in IEC 60071-1 and 60071-2. Finally, among effects compensated by the safety factor, the data
on ageing is reported.
8.1.1 IEC 60071-1 FLOWCHART FOR DETERMINATION OF RATED OR STANDARD

INSULATION LEVELS
The flowchart in IEC 60071-1 Ed.8.0 2010 is shown in Figure 8.1.1

voltages
Protective level of overvoltage limit ing
devices
System analysis
Representative voltages and

overvoltage Urp
(1)
Performance c riterion
Statistical distribution (+)
Inaccuracy of input deta (+)
(+) Effects combined in a co-ord ination
factor Kc
Selection of the insulation meet ing

the performance criterion
Co-ordination withstand voltages Ucw

(2)
Altitude correction factors Ka
(or at mospheric correction factors
Equipment test assembly *)

Dispersion in production *)
Quality of installation *)
Ageing in service *)
Other unknown factors *)
*) Effects combined in a safety
factor Ks
Application of factors to account for

the differences between type test
conditions and actual
service conditions
Required withstand voltage Urw

(3)
Test conditions
Standard withstand voltages
Selection of rated withstand voltages

or standard rated withstand voltages
Uw fro m the lists
Rages of Um
Rated or standard insulation level : set of Uw
NOTE In brackets the subclauses reporting the definition of the term or the description of the action.
Sided bo xes refer to required input
Sided bo xes refer to performed actions
Sided bo xes refer to obtained results
Figure 8.1.1 IEC 60071-1 Flowchart for the determination of rated or standard
insulation levels
Page 251
TypethenameofTB
The following steps are followed in the determination of rated or standard insulation levels:
Step 1: Determination of the representative voltages and overvoltages (Urp)
These voltages and overvoltages are defined in IEC 60071-1as follows:
The voltage and the overvoltages that stress the insulation shall be determined in amplitude, shape and duration by
means of a system analysis which includes the selection and location of the overvoltage preventing and limiting
devices. For each class of voltages and overvoltages, this analysis shall then determine a representative voltage
and overvoltage, taking into account the characteristics of the insulation with respect to the different behaviour at
the voltage or overvoltage shapes in the system and at the standard voltage shapes applied in a standard
withstand voltage test.
The representative voltage and overvoltages may be characterized either by:
- an assumed maximum, or
- a set of peak values, or
- a complete statistical distribution of peak values.
Using the waveform evaluation described in Section 5.3, the voltage waveforms generated in the field can be
converted to the standard voltage shapes to determine these values.
Step 2: Determination of the co-ordination withstand voltages (Ucw) that satisfy the performance standard required
for the system
The co-ordination withstand voltages are defined in IEC 60071-1 as follows:
The determination of the co-ordination withstand voltage consists of determining the lowest values of the withstand
voltage of the insulation meeting the performance criterion when subjected to the representative overvoltages
under service conditions.
The co-ordination withstand voltages of the insulation have the shape of the representative overvoltages of the
relevant class and their values are obtained by multiplying the values of the representative overvoltages by a coordination factor. The value of the co-ordination factor depends on the accuracy of the evaluation of the
representative overvoltages and on an empirical, or on a statistical appraisal of the distribution of the overvoltages
and of the insulation characteristics. The co-ordination withstand voltages can be determined as either
conventional assumed withstand voltages or statistical withstand voltages. This affects the determination procedure
and the values of the co-ordination factor.
Simulations of overvoltage events combined with the simultaneous evaluation of the risk of failure, using the
relevant insulation characteristics, permit the direct determination of the statistical co-ordination withstand voltages
without the intermediate step of determining the representative overvoltages.
Step 3: Determination of the required withstand voltage (Urw)
The determination of the required withstand voltages of the insulation consists of converting the co-ordination
withstand voltages to appropriate standard test conditions. This is accomplished by multiplying the co-ordination
withstand voltages by factors which compensate the differences between the actual in-service conditions and those
in the standard withstand voltage tests.
The factors to be applied shall compensate atmospheric conditions by the atmospheric correction factor Kt and the
effects listed below by a safety factor Ks.
Effects combined in a safety factor Ks;
- the differences in the equipment assembly;
Page 252
TypethenameofTB
- the dispersion in the product quality;

- the quality of installation;
- the ageing of the insulation during the expected lifetime;
- other unknown influences.
Step 4: Selection of the rated insulation level
The selection of the rated insulation level consists of the selection of the most economical set of standard rated
withstand voltages (Uw) sufficient to prove that all the required withstand voltages are met.
The withstand voltages to prove that the required temporary, slow-front and fast-front withstand voltages are met,
for phase-to-earth, phase-to-phase and longitudinal insulation, may be selected with the same shape as the
required withstand voltage, or with a different shape, exploiting, for this last selection, the intrinsic characteristics of
the insulation.
8.1.2 IEC 60071-2: SIMPLIFIED METHOD (LIPL AND SIPL)

For simplified estimating methods on required withstand voltages from the lightning impulse protective level (LIPL)
and the switching impulse protective level (SIPL) of surge arrester, the rules and equations from statistical methods
are described in IEC 60071-2 (2) in detail. In this clause, the procedure is briefly introduced referring the existent
standards and the recent studies by the associated committees.
8.1.2.1 OVERVOLTAGES [1]

The following overvoltages are generally considered for insulation co-ordination studies:
a) Fast-front overvoltages
Fast-front overvoltages are often generated by lightning strikes to lines. The overvoltages have rise times typically
in the range 0.1 s - 20 s and tails as long as 300 s. Lightning current impulses (up to 300 kA) generate voltages
on circuits. The generated voltage is a function t of the surge impedance associated with the propagation of an
electromagnetic wave.
b) Slow-front overvoltages
Overvoltages occur whenever the initial voltage at the instant of switching is not equal to the final voltage. During
the transition from the initial state to the final state, the voltage overshoots by as much as 200 % if there are no
losses. The wave shape for the slow-front surges may vary over a wide range depending on the circuit involved.
Typically, the front times range from a 20 s up to 5 ms.
8.1.2.2 PROTECTIVE LEVEL OF SURGE ARRESTER
LIPL of surge arrester based on nominal discharge current (In)

According to IEC 60071-2, surge arresters with the following nominal discharge currents are generally applied for
protection against fast-front overvoltages:
- for systems with Um in range I: 5 kA or 10 kA;
- for systems with Um in range II: 10 KA or 20 kA
When currents flowing through the surge arrester are expected to be higher than its nominal discharge current, it
must be verified that the corresponding residual voltage is still within the acceptable protective limit.
The nominal discharge current (In) values are generally derived by following equation:
Page 253
TypethenameofTB
In = {(negative 50 % F.O.V. of the line) 2 V res.} / Z ---------- 8.1.1

where
F.O.V.: Flash-over voltage
V res.: Residual voltage of surge arrester
Z: Characteristic impedance of line
For the determination of the energy absorption due to lightning of surge arresters installed in a substation, it is
usually sufficient to assume that the representative amplitude of the prospective fast-front overvoltage reaching the
substation is equal to the negative 50 % fast-front withstand voltage of the line. However, for the determination of
the total energy absorption capability, one should take into account lightning flashes with multiple strokes.
The lightning impulse protective level (LIPL) of a surge arrester is the maximum residual voltage at nominal
discharge current. It is applicable to the protection of equipment from fast-front overvoltages.
The nominal discharge current of 20 kA is applied to surge arresters of system voltages of up to 800 kV. From
preliminary study by IEC TC37 MT4 Task Force 2, the nominal discharge current of 20 kA is considered suitable for
application on surge arresters for UHV systems.
The discharge current of surge arresters depends on the lightning current, the configuration of the transmission line
and the acceptable number of faults on the line. Therefore, when more practical results are desired, which include
travelling waves and statistical distributions of lightning, computer simulations, using EMTP program, should be
employed.
SIPL of arrester based on switching impulse discharge current (Ia)
According to the IEC 60071-2, the overvoltages originating from line energization and re-energization generates
currents with magnitudes ranging from 0.5 to 2 kA through the surge arresters. It is generally said that, within this
current range, the knowledge of the exact current amplitude is not so important owing to the extreme non-linearity
of the v-i characteristics of surge arresters. The current amplitude depends on the system conditions and the
residual voltage values in range of switching impulse discharge currents of the surge arresters. Therefore, the
current amplitude is not defined by the IEC 60099-4 [3]. For the determination of SIPL, the schematic procedure to
derive switching impulse discharge current amplitude of a surge arrester is introduced below. However, when more
accurate results are desired, which include travelling waves and statistical distributions of switch operations,
computer simulations should be employed.
The switching impulse protective level (SIPL) is the maximum residual voltage at the specified switching impulse
current. However, it is considered as better to use the exact current, which is expected by the supposing conditions
of the system, in order to determine the practical SIPL value. As an example, Figure 8.1.2 shows the schematic
procedure used to derive switching impulse discharge current amplitude of a surge arrester for UHV system in
Japan. From the maximum expected overvoltages (2E) at the surge arrester location without a surge arrester,
characteristic impedance of line (Z) and a surge arrester v-i characteristic, the residual voltage (V res.) and
discharge current (Ia) of a surge arrester are determined, because the equation 8.1.2 is written as the inclined
straight line. The values of Vres. and Ia are given as those at the intersection of the inclined straight line and nonlinear curve of v-i characteristics of surge arrester.
V res. = (2E Z Ia) -------- 8.1.2
Page 254
TypethenameofTB
Figure 8.1.2 Schematic procedure to derive switching impulse discharge current

amplitude of surge arrester for UHV system in Japan
8.1.2.3 PROCEDURE TO LEAD REQUIRED WITHSTAND VOLTAGE URW
Insulation coordination studies may be complex depending on the methods of analysis and modelling. The analysis
associated with system transients can be very subjective.
According to IEC 60071-2, the following are suggested:
The required withstand voltage to be verified on standard type test conditions and at standard reference
atmosphere is determined taking into account all factors which may decrease the insulation in service so that the
co-ordination withstand voltage is met at the equivalent location during the equipment life. To achieve this, two
main types of correction factors shall be considered.
- a correction factor associated with atmospheric conditions;
- correction factors (called safety factors) which take into account the differences between the actual in-service
conditions of the insulation and those in the standard withstand tests
There are some basic rules and equations which can be implemented for simple cases are introduced.
To obtain the required withstand voltage, Urw, the co-ordination withstand voltage (Ucw) shall be multiplied with an
altitude correction factor, Ka, considering 1000 m and a safety factor, Ks. as shown the equations 8.1.3.
The coordination withstand voltage Ucw is the lowest value of withstand voltage, for which the expected low failure
rate of the equipment is not exceeded over its full lifetime. Ucw is derived from the representative overvoltage (Urp)
and the coordination factor Kc.
The altitude correction factor, only applied for external insulation, is equal to em(H/8150) where H is the altitude (m
equals to 1 for co-ordination fast-front withstand voltage and according to the Figure 9 of IEC 60071-2 for coordination slow-front withstand voltage).
Required withstand voltage to fast-and slow-front overvoltage
Urw = Ka Ks Ucw = Ka Ks UrpKcd -------- 8.1.3
Page 255
TypethenameofTB
where
Urw: Required withstand voltage
Ka: Altitude correction factor
Ks: Safety factor
Ucw: Co-ordination withstand voltage
Urp: (LIPL+2 ST) for fast-front overvoltage, SIPL for slow-front overvoltage
Kcd: Deterministic co-ordination factor (1 for LIPL, 1.0 ~1.1 for SIPL as per IEC 60071-2 Figure 6)
S: Steepness of impinging surge
T: Travelling time of lightning surge from arrester point to the protected equipment terminal such as
transformer.
These basic rules and equations can be used for preliminary estimations. For more complex and/or practical
applications, results provided by these simplified formulas are considered to be less precise. For such cases, better
results are usually obtained from detailed computer simulations. It may incorporate such important features as
realistic line and station configurations with tower and ground transients, accurate models such as multi-phase
coupled lines with frequency-dependent parameters and corona, capacitances from major equipment such as
CVTs, non-linear behaviour of arresters plus many more which are described in IEC TR 60071-4.[4]
REFERENCES
[1] IEC 60071-1, Insulation co-ordination Part 1: Definitions, principles and rules, 2006-01.
[2] IEC 60071-2, Insulation co-ordination Part 2: Application guide, 1996-12.
[3] IEC 60099-4, Surge arresters Part 4: Metal-oxide surge arresters without gaps for a.c. systems, 2009-05.
[4] IEC TR 60071-4, Insulation co-ordination Part 4: Computational guide to insulation co-ordination and modeling
of electrical networks, 2004-06.
8.1.3 SAFETY FACTOR

The safety factor compensates for the differences in the equipment assembly and the dispersion in the product
quality as explained in Clause 8.1.1. The factors for these individual effects are not given. Instead, the following
safety factors are recommended as an overall factor in IEC 60071-2:
- for internal insulation Ks = 1.15;
- for external insulation Ks= 1.05.
The following research is reported:
(1) safety factors for rated insulation level of UHV ac system applied in various countries,
(2) background of derivation of safety factors in IEC, and
(3) evaluation of effects compensated by the safety factor (examples of the evaluation of actual ageing in GISs).
Page 256
TypethenameofTB
8.1.3.1 SAFETY FACTORS IN THE PRACTICE OF VARIOUS COUNTRIES

Table 8.1.1 and Table 8.1.2 summarize safety factors applied in the insulation design procedure of various
countries. In IEC 60071-2, the recommended safety factors are shown for internal insulation and external insulation,
respectively. However, since the inquiry about the safety factor in the WG was answered in the subject concerning
LIWV and SIWV, respectively, the safety factors are classified for LIWV and SIWV in Table 8.1.1 and Table 8.1.2.
Highest
Voltage
800kV
800kV
800kV
800kV
800kV
800kV
Country/Utility
Brazil/
Furnas
China/
SGCC
India/
Power Grid
Korea/
KEPCO
South
Africa
United
States/
AEP
According
to IEC and
ANSI
standard at
that time
According
to
IEC60071-2
(1996);
1.15, 1.05
Considering
for the
sever
condition*
1.12-1.2
N/A
N/A
1.25
N/A
1.32
N/A
N/A
1.15
N/A
LIWV
SIWV
800kV
765kV
Venezuela Canada
* The ratio of the insulation levels to representative voltages is above 1.14.
Table 8.1.1 Safety factor applied in 800kV system

Highest
Voltage
1200kV
1200kV
1100kV
1100kV
1050kV
Country/Utility
India/Power
Grid
Russia/
China/SGCC
Japan/
TEPCO
Italy/CESI
1.25
According to
IEC60071-2
(1996);
1.15, 1.05
N/A
N/A
LIWV
SIWV
According
to
IEC600712 (1996);
1.15, 1.05
1.15
Table 8.1.2 Safety factor applied in 1100kV and above system

It was revealed that safety factors actually used ranged from 1.00 to 1.32 besides those of 1.05 or 1.15
recommended in IEC 60071-2. In the case of Japan, the safety factor applied is 1. This is partly because there is
sufficient margin in the process to compute the representative overvoltage. Strictly speaking, this margin may
correspond to the co-ordination factor, but it could be playing a role of an implicit safety margin as a practical
matter.
8.1.3.2 BACKGROUND OF THE SAFETY FACTOR IN IEC STANDARD

The current safety factor was defined in IEC 60071-1 (1993) and IEC 60071-2 (1996). The details of these
standards had been discussed in IEC TC-28 WG01 and IEC TC-28 WG02 (both WGs active from 1986 to 1990)
and the definition of the safety factor was clarified in these two working groups. The information related to the
safety factor is extracted in the context of the process of how the safety factor has been studied. Since no formal
minutes were available, the research has been conducted based on the committee documents and the memoranda
of the committee members.
In IEC TC-28 WG01, the definition of terms was discussed. It seems that the safety factors then were 1.20 for
lightning overvoltages and 1.15 for switching overvoltages. The separation of altitude correction from these values
Page 257
TypethenameofTB
was proposed. In the meetings of the WG, no document on the numerical computation of the safety factor was
found.
In the draft IEC 71-1 submitted to National Committees in 1987, the safety factor was defined as follows:
The safety factor is a factor based on experience, that is introduced to account for all the elements which can
influence negatively the insulation performance, and which are not taken into consideration in the insulation coordination procedure, since they are either unknown, or difficult to be quantified, or related to insufficient knowledge
of the values introduced in the calculation.
IEC TC-28 WG02 studied the insulation coordination procedure and discussed overvoltages, withstand voltages,
and various factors. During this work, in order to clarify the details of safety factor, it was explained that the safety
factor would compensate for the following influences:
(1) dispersion in production
(2) contamination during the erection on site (GIS)
(3) aging of liquid and solid insulation
(4) parallel insulation
The draft report in 1989 indicated the following:
Safety factor
-Internal insulation: 1.15

- External insulation: 1.02
- Site assemble GIS: 1.15
The value of 1.02 for external insulation is different from 1.05 of IEC 60071-2(1996). However, no document
explaining the numerical value was found.
While the details of the safety factor were defined in these two working groups, it seems that not all the elements of
the safety factor were evaluated at that time. It is assumed that sufficient data was not yet available and individual
effects could not be quantified. Consequently, the safety factor was proposed as the overall value based on the
margin empirically used at that time, and that value was adopted.
8.1.3.3 EVALUATION OF THE EFFECTS COMPENSATED FOR BY THE SAFETY FACTOR

Based on recent studies and field data research, effects compensated by the safety factor on the insulation
performance have been gradually clarified. CIGRE WG C4.302 researched and evaluated the insulation
coordination related to internal insulation of gas insulated systems with SF6 and N2/SF6 gas mixtures under AC
condition and deals with the ageing of gas insulated equipment in service. Technical Report No.360 of WG C4.302
reports on ageing of an actual gas insulated equipment and switchgear after long-term service operation. Examples
cited in the report include performance of a 245kV current transformer (CT), insulating spacer made of epoxy resin
of 154kV and 72/84kV class and 123kV GIS.
(1) Performance of a 245 kV gas-insulated current transformer
Ageing and lifetime expectancy of a gas insulated instrument transformer for air insulated substations has been
investigated in Germany. The test object is a 245 kV CT with composite insulator manufactured in 1983, which was
continuously in service for more than 17 years. The insulation medium consists of a dielectric film, impregnated
with SF6 gas, which is wrapped around the conductor and supports the core housing.
A repeated non-destructive dielectric type-test was performed on the CT. To check the LI-withstand capability, five
LI voltages of each polarity were applied. The LI voltage was 80% of the specified 1050 kV. The results show that
according to the design philosophy to be free of PD up to the power frequency withstand voltage there is no
Page 258
TypethenameofTB
obvious ageing of the internal insulation. The results show that the dielectric performance of the current transformer
can be considered as similar to that of a new unit.
Other special tests were performed to characterise the performance of a 17 years old composite polymer insulator.
The composite insulator also showed excellent performance.
(2) Performance of epoxy insulating spacer
Ageing and residual performance of insulating spacers made of epoxy resin was studied, using the actual spacers
after long-term field operation in Japan.
The reference reports the test results on the ageing of 154kV insulating spacer after 17 years operation since 1971.
The verification tests by lightning impulse withstand voltage (LIWV), AC withstand voltage (ACWV) and PD
measurement have been conducted, which shows satisfactory insulating performance. The test results also
suggest that there was no deterioration after 17 years of operation. The interface between the high-voltage
embedded electrode and the epoxy resin spacer was observed, reporting that no sign of deterioration was found at
the interface.
Similar tests using 72/84kV epoxy insulating spacers after 23 years of in-service operation were conducted in terms
of the lifetime assessment. First, PD measurements at AC 128 kVrms and AC withstand tests at 160 kVrms for one
minute were conducted at 0.4 MPa for 12 spacers of 23 years old. No PD detection and no breakdown were
confirmed. Secondly, the ageing of the 23 years old spacers are accelerated by additional 1.5 times AC stress of
operating voltage during 5.5 months, which corresponds to 67 years of operating voltage. After the AC voltage
acceleration, PD measurements and breakdown tests at LI voltage are performed and no PD detection and no
reduction of LI breakdown voltages are confirmed again. Lastly, the interface between the high-voltage embedded
electrode and the epoxy resin is observed in detail. It is reported that there is no sign of ageing of epoxy resin
around the maximum electric field portion. Therefore, the series of the tests verified that the spacers have at least
90 years lifetime.
(3) Performance of a 123 kV GIS
The literature reports that overall the entire insulation system does not exhibit an increase in the failure rate with
time due to the failure rate of 123kV GIS in Germany since 1967 as exhibited in Figure 8.1.3
There is a general reduction of the failure rate with time, which proves that all further failures were caused by
teething faults during the first 6 years of operation. However, even after 25 up to 30 years of service no significant
increase of the failure frequency was noted. Therefore, the literature concludes that after 30 years of operation
there is no general ageing which recognizably affects the long-term performance.
Figure 8.1.3 Failure rate of 123kV GIS in Germany due to dielectric faults
Page 259
TypethenameofTB
(4) Summary
The investigation results above mentioned can be summarized as follows.
Equipment
(Insulated part)
Voltage
class
Age in the
field
CT
245 kV
17 years
Lightning impulse test

PD measurement
Epoxy spacer
154 kV
17 years
LIWV, ACWV, PD measurement, No deterioration,

Investigation of the boundary face No change in the boundary surface
Epoxy spacer
72/84 kV
GIS
123 kV
Test method
Results
No PD up to 460 kV
No PD, No breakdown,
PD test, AC voltage withstand test,
No PD even after the accelerated
Accelerated
deterioration
test
23 years
ageing test,
equivalent to 67 years, Investigation
No LI breakdown voltage decrease,
of the boundary face
No change in the boundary surface
Change in the failure rate over years
No ageing in 30 years
Table 8.1.3 Ageing of gas insulated equipment in service reported in Technical

Report No.360 of WG C4.302 reports
The CIGRE Brochure No.360 concludes;
"It is verified that the GIS components such as epoxy insulators and CTs after the operation from 17 to 23 years
had enough dielectric performance, which proves that the components have enough design margins and/or there is
not a significant ageing. The field failure data of 123 kV GIS suggest that after 30 years of operation there is no
general ageing which recognizably affects the long-term performance of the GIS.
Therefore, it is noted that a GIS insulation system of a proper design has a service life of 50 years even if
considering various reduction of insulation performance and no significant ageing is recognized in the actual GIS
after long-term operations."
8.1.3.4 REMARK ON THE SAFETY FACTOR

The current safety factor recommended in IEC 60071-2 has been adopted more than 20 years ago as a value
obtained from experience. During that period, data on insulation deterioration characteristics with ageing have been
accumulated and the quality control and assembly and installation technology of equipment components and
equipment manufacturing have also been significantly improved. Based on such a situation, it might be one option
to allow some range for the recommended value of the safety factor instead of making it a constant fixed value.
Practically, safety factors ranging from 1.00 to 1.32 are applied for rated insulation levels for UHV systems.
Meanwhile, CIGRE WG C4.304 is planning to investigate the deterioration characteristics of power transformer oil.
CIGRE WG A3.06 Reliability of high voltage equipment is currently also investigating equipment failures in
various countries. These investigations are very valuable in evaluating the validity of the safety factor. The findings
of the investigations may contribute to a review of the safety factor.
Page 260
TypethenameofTB
8.2 Determination process for LIWV and SIWV

8.2.1 DETERMINATION PROCESS FOR STANDARD RATED WITHSTAND VOLTAGES
The procedure for determining the rated or standard insulation levels, and major overvoltages and withstand
voltages are explained in 8.2.1.
The representative overvoltages are placed at the top of the flow chart of determining the standard insulation levels.
Overvoltages in UHV systems are explained in detail in Chapter 4, including reports on the process from analytical
approaches for the temporary overvoltage (TOV), switching overvoltage, lightning overvoltage, VFTO and TRV to
the determination of the representative overvoltage and their evaluation. Of these overvoltages, TOVs, switching
overvoltages and lightning overvoltages are estimated for determining the standard insulation level. A switching
overvoltage is classified as a slow-front overvoltage, and lightning overvoltage as a fast-front overvoltage in
IEC60071-1.
The next step in the flow chart is to determine the co-ordination withstand voltage. This process consists of
determining the lowest values of the withstand voltages of the insulation meeting the performance criterion when
subjected to the representative overvoltages under service conditions. This criterion is usually expressed in terms
of an economically and operationally acceptable failure rate of the insulation However, simulations of overvoltage
events combined with the simultaneous evaluation of the risk of failure, using relevant insulation characteristics,
permit the direct determination of the statistical co-ordination withstand voltages without the intermediate step of
determining the representative overvoltages.
The required withstand voltage is then determined by applying the atmospheric correction factor and safety factor.
Refer to the previous section for the roles of the safety factor.
The rated or standard insulation levels are determined based on the required withstand voltages. IEC60071-1
specifies that the rated insulation level of the equipment classified in range II (above 245kV) should be defined by
the combination of the standard switching impulse withstand voltage and standard lightning impulse withstand
voltage. This means, the rated insulation level for TOV is converted to a different standard waveform with the
conversion factor. IEC60071-2 1996 describes this process as follows:
"In range II, the standard switching impulse withstand voltage should cover the continuous power-frequency
voltage if no value is specified by the relevant apparatus committee, and the required short-duration powerfrequency withstand voltage."
Meanwhile, in Section 6.4 of IEC60071-1 2006, it is mentioned as follows;
"When it is too expensive or too difficult or even impossible, to perform the withstand voltage tests in standard test
situations, the apparatus committees, or technical committee 42, shall specify the best solution to prove the
relevant standard rated withstand voltages. One possibility is to perform the test in an alternative test situation."
The conversion factors for the conversion from the required short-duration power-frequency withstand voltage to
switching impulses indicated in IEC60071-2 1996 are shown in Table 8.2.1 below. It is also noted that the test
conversion factors should be used only in the specified direction.
Page 261
TypethenameofTB
Switching impulse
withstand voltage
Insulation
External insulation
air clearances and clean insulations, dry
clean insulators, wet
Internal insulation
GIS
liquid-immersed insulation
solid insulation
NOTE The test conversion factors include a factor of
1,4
1,7
1,6
2,3
2,0
2 to convert from r.m.s to peak value.
Table 8.2.1 Test conversion factors for range II to convert required short-duration
power-frequency withstand voltages to switching impulse withstand voltages
(According to Table 3, IEC60071-2 1996).
8.2.2 STANDARD RATED WITHSTAND VOLTAGES FOR UHV SYSTEMS
This WG conducted a research into the lightning and switching impulse withstand voltages. Table 8.2.2 and Table
8.2.3 summarize the research results for the insulation design of substations (only UHV systems were studied in
this draft).
8.2.2.1 LIGHTNING IMPULSE WITHSTAND VOLTAGES

Table 8.2.2 indicates that all countries set either 1950 kV or 2250 kV for the LIVW of transformers, and 2250 kV or
2400 kV for other substation equipment. Except for Italy/CESI, the LIWV of other equipment is higher than that of
transformers. The bottom two rows of the table indicate the ratios between LIWV and representative overvoltage. A
value of 1.02 used by TEPCO (Japan) for other equipment is the smallest, while 1.41 used by Power Grid (India) is
the largest. This may result from different analytical conditions assumed in these countries, i.e., Japan/TEPCO sets
higher representative overvoltages and smaller safety factors. India/Power Grid sets comparatively lower
representative overvoltages which correspond to a residual voltage of a surge arrester (V20kA), based on the
simplified method. When a calculation method such as EMTP is used for analyzing the representative overvoltage,
the effects of lightning strike and circuit conditions on the calculation results are large - refer to Section 4.4 for
details.
Page 262
TypethenameofTB
Highest voltage
1200kV
Country / Utility
Russia/
Substation type
AIS
Representative
overvoltage (kV)
1100kV
1100kV
1050kV
China/SGCC Japan/TEPCO Italy/CESI

GIS,
Hybrid-IS
GIS
GIS
N/A
N/A
1700
1700
1796
2040
1896
2208
N/A
N/A
Calculation method
Special
programme
Simplified
method
EMTP
EMTP
TNA,
EMTP
Safety factor
1.15
1.15/1.05*
1.15/1.05*
1.00
N/A
Tr.
Other
2250
2400
2250
2400
2250
2400
1950
2250
2250
2250
Tr.
N/A
1.32
1.25
1.03
N/A
Other
N/A
1.41
1.18
1.02
N/A
LIWV (kV)
LIVW/
Representative
overvoltage
Tr.
Other
1200kV
India/Power
Grid
AIS,
Hybrid-IS
* 1.15 for internal insulation, 1.05 for external insulation
Table 8.2.2 LIWV of substation equipment for UHV systems and determination
process
8.2.2.2 SWITCHING IMPULSE WITHSTAND VOLTAGES
Table 8.2.3 indicates that all countries set either 1425 kV or 1800 kV for the SIVW of transformers, and 1550 kV or
1675 kV or 1800kV for the SIWV of other equipment. Russia, India/Power Grid and China/SGCC set 1800 kV for
both transformers and other equipment. The bottom two rows of the table indicate the ratios between SIWV and
representative overvoltage. The smallest value of 1.09 is used by TEPCO (Japan) for transformers, and the largest
value of 1.29 is used by SGCC of China for transformers. Differences among countries are less significant than
those for the lightning impulse withstand voltage. In p.u., the representative overvoltages in Table 8.2.3 range
between 1.46 and 1.69, which are significantly lower than those in lower voltage systems in general. This is
because switching overvoltages are effectively suppressed in UHV systems by means of circuit breakers with
resistors, shunt reactors, higher performance surge arrestors, etc. Switching overvoltages are explained in detail in
Section 4.3.
Page 263
TypethenameofTB
Highest voltage
1200kV
Country / Utility
Russia/
Substation type
AIS
Representative
overvoltage (kV)
1050kV
Italy/CESI
GIS
Tr.
1570
(Maximum)
1500
(2% value)
1392
(2% value)
1309
(Maximum)
1450
(Maximum)
Other
1570
(Maximum)
1500
(2% value)
1437
(2% value)
1400
(Maximum)
1450
(Maximum)
Special
programme
Simplified
method
EMTP
EMTP
TNA,
EMTP
Tr.
Other
Tr.
1.15
1800
1800
1.15
1.15/1.05*
1800
1800
1.20
1.15/1.05*
1800
1800
1.29
1.0
1425
1550
1.09
N/A
1800
1675
1.24
Other
1.15
1.20
1.25
1.11
1.16
Calculation method
Safety factor
SIWV (kV)
(Phase-to-ground
SIWV/
Representative
overvoltage
1200kV
1100kV
1100kV
India/Power
China/SGCC Japan/TEPCO
Grid
AIS,
GIS,
GIS
Hybrid-IS
Hybrid-IS
* 1.15 for internal insulation, 1.05 for external insulation
Table 8.2.3 SIWV of substation equipment for UHV systems and determination
process
8.2.2.3 VERIFICATION OF SWITCHING IMPULSE WITHSTAND VOLTAGES FOR TOV
As explained in Section 8.2.1, it should be taken into account in determining the standard withstand voltage that the
standard switching impulse withstand voltage is required to cover short-duration power-frequency withstand
voltages in UHV systems. Hence, estimation was carried out to evaluate the possibility of using the switching
impulse withstand voltages for verifying the power-frequency withstand voltages (PFWV) in the standard procedure
of IEC60071-1 2006 and IEC60071-2 1996.
Table 8.2.4 lists the converted switching impulse withstand voltages equivalent to the power-frequency withstand
voltages.
Here, the maximum value of the temporary overvoltage was assumed to be the representative temporary
overvoltage. Co-ordination factor Kc = 1, that is the co-ordination withstand voltage is equal to the representative
temporary overvoltage, according to Section in 3.3.1 of IEC 60071-2 1996. Similarly, safety factor Ks is 1.15, and
conversion factor is 2.3 (Table 8.2.1) according to IEC 60071-2.
The bottom row of the table indicates the switching impulse withstand voltages estimated using a standard
conversion factor in IEC 60071-2 1996. These values, ranging from 2383 to 2566 kV, are considerably greater than
SIWVs in Table 8.2.3, and require the switching impulse withstand voltages of at least 2400 kV in UHV systems.
Since the SIWVs greater than the lightning impulse withstand voltages are economically and technically
disadvantageous, an alternative test is available to verify the withstand of internal insulation to TOV, as mentioned
in Section 6.4 of IEC60071-1 2006. Therefore, the power-frequency withstand voltage tests, which is explained in
Section 8.4, are provided for verifying withstand voltages by the relevant apparatus committees.
Page 264
TypethenameofTB
Highest voltage
1200kV
1200kV
1100kV
1100kV
1050kV
Country / Utility
Russia/
India/Power
Grid
China/SGCC
Japan/TEPCO
Italy/CESI
Substation type
AIS
AIS, Hybrid-IS
GIS, Hybrid-IS
GIS
GIS
1.3 p.u.
(901kV r.m.s)
1.4 p.u.
(970kV r.m.s)
1.4 p.u.
(889kV r.m.s)
1.5 p.u.
(953kV r.m.s)
N/A
EMTP
EMTP
TNA / EMTP
2521
N/A
Representative
overvoltage
(Maximum TOV)
Calculation method
Co-ordination factor
1.0*
Safety factor
1.15*
Conversion factor
2.3*
Equivalent SIWV
2383
2566
2351
* Reference: IEC60071-2 1996
Table 8.2.4 Required short-duration power-frequency withstand voltages and

conversion to switching impulse withstand voltages for internal insulation of UHV
systems (estimated)
Page 265
TypethenameofTB
8.3 Consideration of VFTO for insulation coordination

8.3.1 GENERAL
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 [1, 2]. As the rated voltage increases, the difference
between the rated lightning impulse withstand voltage (LIWV) and the VFTO peak value decreases, as shown in
Figure 8.3.1. Figure 8.3.1 also shows the calculation results for different GIS and Hybrid IS at different voltage
levels. The maximum calculated VFTO peak value in GIS systems may reach the insulation level of LIWV. In case
of Hybrid IS a lower maximum calculated VFTO peak value of 2.2 p.u. is reported, because of the lower length of
busbar sections [2].
Because the generated transients depend strongly on the specific configuration and on the superposition of
travelling waves, it is not possible to give generally admitted values, valid for each case. An accurate simulation for
each substation, especially in the UHV range, is necessary for the insulation coordination as basis for the decision
making about possible countermeasures [3, 4].
Concluding it could be necessary to design and maybe to test considering the VFTO level or to suppress severe
VFTO. Different approaches are known and used to determine the necessary measures according to the insulation
coordination.
5.0
Rated lightning impulse withstand voltage

Rated switching impulse withstand voltage
Rated power frequency withstand voltage (peak)
Voltage [pu]
4.0
3.0
GIS
GIS
GIS
2.0
MTS
MTS
1.0
Japan
South Africa
China
0.0
300
500
700
900
Rated voltage Ur [kV]
1100
Figure 8.3.1 Dependency of rated withstand voltages and calculated VFTO on rated
voltage as per IEC 62271-203 , calculated values from [2, 4, 5, 6]
Page 266
TypethenameofTB
8.3.2 INSULATION COORDINATION APPROACH
8.3.2.1 1 ST APPROACH MAINLY USED IN JAPAN

1100 kV substation would be located in the mountain areas, so it is necessary to apply full GIS to reduce the size
of substation area as much as possible to meet installation-space constrains, to harmonize with natural
environment and to minimize costs. The keys for these technologies are suppressing various incidents of
overvoltage and finding ways to reduce lightning and switching surges [7]. Substation equipment, such as higher
performance surge arrester and closing and opening resistors for CB and resistor fitted disconnector, were
developed and the insulation requirements were reduced. As a result, LIWV of GIS was determined as 2250 kV
and that of transformer was determined as 1950 kV. By adopting CB with opening and closing resistor, the
insulation specifications of all power network equipment including transmission lines are coordinated. Since a UHV
transformer is directly connected to GIS, overvoltages occurring due to switching of disconnectors or earthing
switches may directly enter the transformer terminal [8]. Moreover, digital protection and control equipment are
located close to the GIS. Therefore, it is very important to take care of the surges transferring into the secondary
equipment [9]. The levels of these switching overvoltages could also mitigated by applying a damping resistor [10].
To suppress VFTO below 1.3 p.u., disconnector with 500 resistor for both closing and opening was developed
and provided [11, 12, 13]. Another feature of this disconnector with resistor is that they prevent VFTO secondary
induced surges effectively. This is especially useful at UHV as the ratio between primary and secondary voltages is
significantly increased.
8.3.2.2 2 ND APPROACH MAINLY USED IN CHINA

The VFTO of each substation was calculated using the most unfavourable case corresponding to a remaining
trapped charge voltage of -1 p.u. The calculated values were compared to the LIWV for the different equipment.
For the comparison a safety factor was introduced. The safety factor was set to 1.15 for the equipment, except for
the air insulated busbar line in case of Hybrid IS. The required VFTO withstand voltage is equal to the LIWV
divided by the safety factor. If the calculated VFTO is higher compared to the required withstand voltage, special
measures to mitigate the VFTO are required [6]. Based on this study, SGCC applied disconnector with damping
resistor in case of GIS and disconnector without resistor in case of Hybrid IS in AC UHV Pilot Project in China.
Details are shown in section 4.5.2.1.
Page 267
TypethenameofTB
8.3.2.3 GENERAL INSULATION COORDINATION APPROACH

Summarizing the different approaches, a general insulation coordination procedure was proposed by CIGR [3, 4],
following the general insulation coordination approach [14]. The procedure consists of the following three steps
(see Figure 8.3.2):
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, if known
99 % probability value determined by simulation or
maximum values measured during testing
Otherwise the worst case assumption of a trapped charge voltage of -1 p.u. should be used for the
simulation.
The accuracy of the simulation model must be verified
Step 2 Calculation of the required VFTO withstand voltage UCW_VFTO for the different equipment by using:
Coordination 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 coordination
No damping measure required
Damping measure required (disconnector with low trapped charge voltage, Damping resistor definition
of required resistance value, other mitigation methods)
Page 268
TypethenameofTB
Trapped charge voltage

(TCV) behavior
known
NO
YES
System analysis
VFTO calculation
(verified calculation
method)
YES: UTCV = - 0.3 - 1.0 pu

NO: UTCV = - 1.0 pu
Evaluation of trapped
Maximum
calculated VFTO
Umax_VFTO
charge behavior
Insulation characteristic
Statistical distribution
Inaccuracy of simulation
KC
Co-ordination factor KC
Selection of the insulation

meeting the
performance criterion
Correction factors
Atmospheric correction
factor Kt
Aging in service
Quality of installation
Co-ordination withstand
voltage
Ucw_VFTO
KS
Application of factors to
account for the
differences between
type test and actual
service conditions
Safety factor KS
External
insulation
YES: Kt
Required withstand
voltage
Urw_VFTO
Test conditions
Ktc
Comparison with LIWV

LIWV Urw_VFTO
Standard lightning
impulse withstand voltage
(LIWV)
NO
Definition of required
damping measures
YES
Trapped charge behavior

Resistance value for
damping resistor
Other mitigation methods
No damping measures
necessary
Sided boxes refer to required input
Sided boxes refer to performed actions
Sided boxes refer to obtained results
Figure 8.3.2 Insulation coordination according to IEC 60071-1, proposed in [3, 4]
Page 269
TypethenameofTB
Step 1 VFTO calculation

The propagation of VFTO throughout GIS can be analysed by representing GIS sections as low-loss distributed
parameter transmission lines. Each section may be characterized by the surge impedance and the time of
travelling. Travelling waves are reflected and refracted at every point where they encounter a change in the surge
impedance. The generated transients depend on the GIS configuration and on the superposition of the surges
reflected and refracted on line discontinuities like circuit-breakers, T-junctions, cable connections or bushings. 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 [15].
An accurate modelling of each individual GIS component makes it possible to estimate VFTO level with a relatively
high precision (differences lower than 5 %), especially in short GIS structures or test equipment. Figure 8.3.3
shows a comparison of simulated and measured VFTO. The measured voltage progressions coincide very well
with the simulation results as regards VFTO amplitude and rise time [16]. Nevertheless, variations of more than
10 % depending on the calculation methods are reported. Therefore, it is important to verify the simulation results
by measurements [4].
1500
1300
Voltage [kV]
1100
900
700
500
300
Measurement
Calculation
100
-100
1.8
1.9
2.1
2.2
2.3
2.4
2.5
Time [us]
Figure 8.3.3 VFTO calculation and measurement when switching busbars with a GIS
disconnector as per IEC 62271-102 without pre-charging (left), test setup (right)
[16]
Step 2 Required VFTO withstand level - comparison with LIWV
Basis for the insulation coordination is calculation of the required VFTO withstand voltage. The insulation withstand
strength under VFTO stress must be compared with the LIWV. The necessary safety depends on many factors and
could be different for the different kinds of equipment. Generally, the VFTO withstand voltage 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 VFTO processes caused by disconnector 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
Page 270
TypethenameofTB
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 difference between the contacts of disconnector just
before striking and on the location considered. For the calculation of VFTO stresses, the trapped charge charges
remaining on the load side of the disconnector must be taken into consideration.
Switching by a slow acting disconnector generates numerous re-strikes between the moving and fixed contacts, but
99 % of the trapped charge voltages were limited to 0.4 p.u. [13]. For these cases, the resulting VFTO is in the
range of 1.7 p.u. and reach 2.0 p.u. for very specific cases. Fast operating disconnector on the other hand can
leave residual charge levels corresponding to 1 p.u. in a no negligible number of cases [16, 17]. The trapped
charge voltage behaviour strongly depends on the contact speed, as shown in the simulation plot (see Figure
8.3.4) [16]. A lower trapped charge voltage gives a higher safety margin compared to the calculation based on a
trapped charge voltage of 1 p.u. For the insulation coordination this additional margin has to be considered.
1
calculated
90% probability
95% probability
99% probability
measured
0.8
0.6
0.4
0.2
0
0
0.5
1.5
2.5
3.5
Contact speed [m/s]
Figure 8.3.4 Influence of contact speed
Figure 8.3.5 Principal voltage-time
on trapped charge voltage during
characteristic of SF6 [18]
disconnector switching [16]

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
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
in SF6-insulated systems 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 (see Figure 8.3.5) [18, 19]. 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
Page 271
TypethenameofTB
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 s. For times shorter than 1s, 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 coordination
If the required VFTO withstand voltage is higher than LIWV, 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 [2].
The damping of VFTO by integration of a damping resistor is a well proven technology. Service experiences exist
since more than 10 years [2]. 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 8.3.6 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 p.u. and +1 p.u. respectively. Without damping
resistor, the VFTO amplitude reaches a value of 2.49 p.u. This exceeds the LIWV according to the insulation
coordination. In case of integration of a damping resistance higher than 200 , the VFTO amplitude can be
suppressed below 1.5 p.u. A damping resistance more than 500 mitigates the maximum VFTO amplitude to a
level lower than 1.13 p.u. [20]. Consequently the resistance of damping resistor could be chosen and defined
according to the maximum calculated VFTO and the required mitigation effect. A 110 damping resistor was used
in the Italian 1000 kV project [2]. For the Chinese, Japanese and Korean UHV and EHV projects, it was decided to
use a 500 damping resistor [2]. An example of GIS disconnector structure with damping resistor is shown in
Figure 8.3.7.
1.13 pu
500
Figure 8.3.6 VFTO in relation to the
Figure 8.3.7 Structure of disconnector
resistance of the damping resistor [20]
with damping resistor [13]
Page 272
TypethenameofTB
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 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 disconnector 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 p.u. Moreover the rate of rise of the voltage across the resistor could be very high and depends on the
test setup 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 disconnector.
The required capacity of thermal energy absorption for the resistor could be calculated by summing up all closeopen operations containing a high number of strikes. For typical applications the required energy absorption for a
500 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.
8.3.2.4 INSULATION COORDINATION STATISTIC APPROACH

The above general insulation coordination approach is the deterministic method, which involves determining the
maximum VFTO and then choosing the LIWV of the GIS with a margin that will cover the uncertainties inherent in
the determination of these values. The statistical approach (corresponding risk of failure) recommended in the IEC
60071-2 could be used for the VFTO insulation coordination. The desired reliability caused by operation of
disconnector can be specified as MTBF, the acceptable criteria, of which bases on technical and economic
analysis and service experience, depended on the policy of each electric utility. The recommended MTBF is about
1500 years.
In the statistical approach the assumed VFTO withstand voltage is equal to LIWV and it has a withstand probability
of 100 %. The withstand probability at higher LIWV is assumed to be 0 %. This means that the co-ordination
withstand voltage is equal to the representative VFTO amplitude at an acceptable failure rate. The random
factories include:
- Statistical distribution of trapped charge voltage
- Statistical closing or opening time distribution of DS during a cycle of the power
- Statistical operation numbers of DS during a year
- Statistical magnitude, shape, duration and polarity of the VFTO
- The other influence factory, such as the possibility of closing on earth DS to discharge the trapped charge
voltage before switching on DS.
Page 273
TypethenameofTB
BIBLIOGRAPHY
[1] CIGR Working Group A3.22: Technical Requirements for Substation Equipment Exceeding 800 kV - Field
experience and technical specifications of Substation equipment up to 1200 kV, Brochure 362, December 2008
[2] CIGR Working Group A3.22: Background of Technical Specifications for Substation Equipment Exceeding
800 kV AC, CIGR Brochure No. 456, April 2011
[3] Riechert, U.; Neumann, C.; Hama, H.; Okabe, S.; Schichler, U., on behalf of CIGR WG D1.36 and AG D1.03:
Basic Information and Possible Counter Measures Concerning Very Fast Transients in Gas-Insulated UHV
Substations as Basis for the Insulation Co-ordination, CIGR SC A2 & D1 Joint Colloquium 2011, Kyoto, Japan,
PS3-O-5
[4] Riechert, U.; Neumann, C.; Hama, H.; Okabe, S.; Schichler, U., CIGR SC D1: Very Fast Transient
Overvoltages (VFTO) in Gas-Insulated UHV Substations, CIGR TB, accepted for publication in 2012
[5] 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
[6] 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 - CIGR
UHV Symposium, Beijing, China, 2007, paper 2-3-4
[7] KOBAYASHI, T.; SUN, G.; YOKOTA, T.; SASAMORI, K.; YAMAGIWA, T.: Compactness and High-reliability
Technology for UHV Circuit-breakers, disconnectors and earthing switches, The second IEC CIGR
International Symposium on International Standards for UHV Transmission, 29-30 January 2009, New Delhi, India,
proceedings pp. 169-177
[8] YAMAGATA, Y.; OKABE, S.: Utility's experience on design and testing for UHV equipment in Japan, The
second IEC CIGR International Symposium on International Standards for UHV Transmission, 29-30 January
2009, New Delhi, India, proceedings pp. 142-154
[9] Yamagata, Y.; Nakada, Y.; MuraYama, Y.; Hirasawa, K.; Yoshizumi, T.: Field Test of 1000 kV Gas Insulated
Switchgear, CIGR Report 13-209, 38th CIGR Session, Aug. 27 to Sept. 1, 2000, Palais des Congrs de Paris,
Paris, France
[10] Nakamura, A.; Taniguchi, H.; Yokoyama, A.: 1100kV AC Transmission Project in Japan, IEC/CIGR UHV
Symposium, Beijing, China, July 18-21, 2007, paper 1-2-1
[11] YAMAGATA, Y., KOBAYASHI, A.;YOSHIZUMI, T.; TSUBAKI, T.: Development of 1100kV GIS Gas Circuit
Breakers, Disconnectors and High Speed Grounding Switches, CIGR Report 13-304, 36th CIGR Session, Aug.
25 to 31, 1996, Palais des Congrs de Paris, Paris, France
[12] Kobayashi, T.; Ono, M.; Shimomura, T.; Yokota, T.: Basic design / specifications of GIS for UHV AC and its
verification test at site, IEC/CIGR UHV Symposium, Beijing, China, July 18-21, 2007, paper 2-3-3
[13] Kobayashi, A.; Murayama, Y.; Ohyama, S.; Kan, M.: Development and Verification Test of 1100 kV Gas
Insulated Switchgear, Int. Workshop of UHV Transmission Technology, Beijing, China, 2005, pp. 94-99
[14] Zaima, E.; Neumann, C.: Insulation Coordination for UHV AC Systems based on Surge Arrester Application
(CIGR C4.306), IEC CIGR International Symposium on International Standards for UHV Transmission, New
Delhi, India, 2009, pp. 108-118
[15] Riechert, U.; Holaus, W.: Ultra High Voltage Gas-Insulated Switchgear A Technology Milestone, Euro.
Trans. Electr. Power (2011), (wileyonlinelibrary.com). DOI: 10.1002/etep.582
Page 274
TypethenameofTB
[16] Riechert, U.; Krsi, U.; Sologuren, D.: Very Fast Transient Overvoltages during Switching of Bus-Charging
Currents by 1100 kV Disconnector; 43rd CIGR Session, 2010, Paris, France CIGR Report A3-107
[17] Szewczyk, M.; Piasecki, W.; Stosur, M.; Riechert, U.; Kostovic, J.: Impact of Disconnector Design on Very
Fast Transient Overvoltages in Gas-Insulated UHV Switchgear, Proceedings of 17th International Symposium on
High Voltage Engineering (ISH), August 22nd 26th, Hannover, Germany
[18] CIGR Working Group 15.03: GIS Insulation Properties in Case of VFT and DC Stress, 36th CIGR Session,
1996, Paris, France, CIGR Report 15-201
[19] CIGR Working Group 33/13-09: Monograph on GIS Very Fast Transients, Brochure No. 35, July 1989
[20] Ikebe, K.; Shimomura, T. ; Iyoda, I.: Simulation Technology for 1000 kV Power Systems, Mitsubishi Electric
Advance, Vol.77/Dec. (1996)
Page 275
TypethenameofTB
8.4 Power frequency (AC) voltage tests for substation equipments

8.4.1 UHV POWER TRANSFORMERS
As a result of the survey by this WG, it has been understood that the long-duration induced AC voltage test (ACLD)
apply to the Japanese and Chinese UHV projects. Indian 1200kV project is also applying the ACLD test. The actual
patterns for long-duration induced AC voltage test are as shown below.
- Japanese and Chinese long-duration induced AC voltage test time sequence:
1.5 p.u x 5min +
3 p.u. x 5min + 1.5 p.u. x 1h (1 p.u. = Um /
3)
- Indian long-duration induced AC voltage test time sequence:

1.5 p.u x 5min + 1.7 p.u. x 1min + 1.5 p.u. x 1h (before and after 1.1 p.u. x 5min )
This chapter shows the description in IEC standard, the theoretical ideas, evaluation of the test voltage and
example of a test voltage on the ACLD for the UHV transformer.
8.4.1.1 IEC STANDARD FOR POWER TRANSFORMERS

Power frequency (AC) voltage tests for the power transformers is standardised by the IEC 60076-3 Second edition
2000-03 Power transformers - Part 3: Insulation levels, dielectric tests and external clearances in air [1]. In this IEC
standard, the long-duration induced AC voltage test (ACLD) is specified as a routine test for Um 300 kV, but this
edition is standardising only up to Um = 800 kV.
The maintenance work of IEC 60076-3 has been started and the second CDV of the IEC 60076-3 Ed.3 was
published as IEC document 14/716/CD [2]. This second CDV of the IEC 60076-3 Ed.3 is including the descriptions
of the insulation levels and the dielectric tests for UHV transformers. Excerpts from the second CDV of IEC 600763 Ed.3 (14/716/CDV) are stated below:
Chapter 7 Dielectric tests,
Clause 7.1 Introduction
Induced voltage test with PD measurement (IVPD), see clause11.3
This test is intended to verify that the transformer will be free of harmful partial discharges under normal operating
conditions. The test voltage is applied in the same way as the voltage that the transformer will experience in
service. During the test, symmetrical voltages appear at all the line terminals and between turns, with no voltage at
the neutral. The test is performed three phase on three phase transformers.
Clause 7.2 Test requirement, 7.2.1 General Table 1
IVPD test is a routine test for the voltage level of Um > 170 kV.
Clause 7.2 Test requirements, 7.2.2 Test Voltage Levels Table 2.
The test voltage levels for UHV transformers are described in the Table 2. The excerpts of the test voltage levels
for the UHV transformers are shown below.
Part of the Table 2 from the CDV of the IEC 60076-3 (14/716/CDV)
Page 276
TypethenameofTB
Highest
Voltage for
equipment
winding
Um kV
Full Wave
Lightning
Impulse
(LI) kV
Chopped Wave
Lightning
Impulse
(LIC) kV
Switching
impulse
(SI) kV
1950
2250
2250
2145
2475
2475
1425
1800
1800
1100
1200
Applied
voltage or line
terminal AC
withstand
(AV) (LTAC)
kV
-
Clause 7.3.3 Tests on Transformers with Um > 170 kV, 7.3.3.1 Routine tests
4) Induced voltage with partial discharge measurement (IVPD)
A test shall be carried out using the method given in section 11.3, with and enhancement (phase to earth) voltage
level of 1.8 Ur / 3 and a one hour PD measurement voltage of 1.58 Ur / 3. Alternative higher voltage levels
may be used if specified by the purchaser, in particular an enhancement voltage of 3 Um / 3 and a one hour
PD measurement voltage of 1.5 Um / 3 may be used if higher.
Clause 11.3 Induced voltage with partial discharge measurement (IVPD), 11.3.2 Test duration and
frequency
The test time at the enhancement voltage shall be 60 s in case Um 800 kV and 300 s in case Um > 800 kV for any
test frequency up to and including twice the rated frequency, unless otherwise specified. When the test frequency
exceeds twice the rated frequency, the test time in seconds of the test shall be:
600
rated frequency
, but not less than 75 s for Um > 800 kV
test frequency
The duration of the test, except for the enhancement level, shall be independent of the test frequency
Clause 11.3.3 Test sequence: Figure 1 in the CDV (14/716/CDV)
Test time
Enhancement voltage
60 min.
1.2 Ur/ 3
One hour PD
1.2
0.4
U r/ 3
U r/ 3
Page 277
TypethenameofTB
8.4.1.2 THEORETICAL IDEA OF LONG-DURATION INDUCED AC VOLTAGE TEST
[3] [4] [5]
This clause describes the theory for the long-duration induced AC voltage test.
The long-duration induced AC voltage test can be used to verify the insulation performance of the equipment
across its lifespan and the insulation strength when subjected to temporary AC overvoltages with high accuracy.
The test is understood to be effective for very high voltage class of equipments. In IEC 60076-3, long-duration
induced AC voltage test are specified for equipment with Um > 170kV. The long-duration induced AC voltage test is
also specified for equipment with nominal system voltage of 187 kV and above in Japan.
The main features of the long-duration induced AC voltage test as a quality control test are listed below:
(1) The long-duration induced AC voltage test is totally low level of the test voltages and is applied over a long
period, so the equipment can be absolved of the damage due to an undue stress from the test. It is intended to
verify the insulation performance of equipment across its lifespan and its insulation strength against temporary AC
over voltages.
(2) The test voltages and duration should be calculated using temporary AC overvoltages and the insulation
characteristics of the equipment (V-t insulation characteristics) for evaluation of reliability of equipment using the
Weibull distribution function.
(3) The pattern of the test voltages, as shown Figure 8.4.1, is a combination of the lower voltages, V1 and V3, and
relatively higher voltage, V2. V2 is applied over a short period whereas V3 is applied over a longer period. V1 is set
slightly higher with a longer period in Japan, and a shorter period in China. The aim of V1 is to monitor the test
voltage and keep it constant. It is not intended to verify the insulation performance of the equipment.
(4) V2 is used to verify the insulation strength against temporary overvoltages and V3 is intended to verify the
insulation performance against the operating voltage across the entire lifespan of the equipment. Partial discharge
measurement should be done along the entire span of the test period.
Figure 8.4.1 Long-duration AC voltage test mode
Page 278
TypethenameofTB
8.4.1.3 EVALUATION OF THE TEST VOLTAGES

The distributions of probability P for the inception voltages of partial discharges should be shown in the Equation
(a), if the characteristics of partial discharges of the equipment would follow with the Weibull distribution function.
V m t a
P 1 exp --- (a)
V 1 t1
where, m=n x a
m: parameter of voltage form
n: gradient of V-t characteristic
a: parameter of time form
If the probability with the no partial discharge in the equipment in a test is Rt, the stress voltage in service is Vs, its
service period is ts and I iss applied N number of times, and Rs is the probability with no partial discharge in
operation, then the test voltage, Vt and its applied period, tt should be fixed. This relationship is be shown in
Equation (b) as
Vt m t ta N Vsm t sa
lnR t
--- (b)
lnR s
8.4.1.4 EXAMPLE OF EVALUATION FOR THE UHV TESTING VOLTAGES (JAPANESE CASE)
The calculated example for the UHV testing voltages in Japan is shown below.
(1) Characteristic of insulation for the equipment
Figure 8.4.2 and Table 8.4.1 show the necessary data of the insulation characteristics for the calculation of the
testvoltages [3] [6].
It should be applied to the insulation characteristics on the barrier-oil-duct models to evaluate the UHV testing
voltages.
Partial discharge inception voltage [%]
120
PD inception time
50% PD time
110
* A number in circle
represents a number of
models without PD
n=44.2
n=323
100
(70 days)
90
1m
80
10-3
1h
1d 4d
10-2 10-1
100
101
102
103
104
105
106
107
108
Figure 8.4.2 V-t characteristic of the barrier-oil-duct models of a core-type

transformer
Page 279
TypethenameofTB
Short time range
Long time range
Structure part of
transformer
Turn-to-turn
44.9
0.42
18.9
292
0.31
90.2
Coil-to-coil
39.9
0.40
16.0
200
0.35
70.0
Barrier-oil-duct
44.2
0.39
17.2
323
0.38
122.7
Table 8.4.1 Summary of V-t characteristics at partial discharge inception

(2) Condition of the stress under the test voltage and operating voltage
Table 8.4.2 shows the stresses in the operation within the 30 years.
Temporary over-voltage
Load rejection
Operating
voltage
1.5pu
1.0pu
Parameter
One-line ground fault
Over-voltage Vs
1.1pu
1.1pu
Duration Ts
0.1s
2s
0.83s
Frequency N
90
10
30years
-
* Considering the influence to the insulation characteristic by the voltage shape of trapezoid
[3]
Table 8.4.2 Power frequency voltage endured by UHV equipment

(3) Probability of zero partial discharge under the test and in the service
If the probability Rt of zero partial discharge in the test would be assumed to be 98.0% as a severe case, the test
voltage which has a value of probability with 99.8% of the no partial discharge in the service, is calculated.
Under the above condition, the reliability is evaluated by the Weibull distribution function, and it should be
considered to have sufficient tolerance to maintain high reliability in operation. The test voltages are calculated as
shown in Table 8.4.3 below.
Test voltages
V1
UHV transformers
1.5E
Applying periods
V2
3E
V3
t1
t2
t3
1.5E
5 min
5 min
60 min
Note: Vn and tn are shown in Figure 8.4.1.
Table 8.4.3 Long-duration AC voltage test mode for UHV Power Transformers
8.4.2 OTHER SUBSTATION EQUIPMENT
8.4.2.1 APPLIED VOLTAGE TEST FOR GIS, CB, ETC.

According to IEC 60071-1, no power frequency test voltage is required for voltage systems of range II.
Overvoltages are sufficiently covered by switching and lightning impulse withstand voltage tests.
For switchgear, IEC 62271-1 specifies a short-duration power-frequency withstand voltage for phase-to-earth and
between phases and for across open switching devices and/or isolating distance (see Table 8.4.4). The test across
Page 280
TypethenameofTB
the open switching device and/or isolating distance can be done with two power-frequency sources in out-of-phase
condition across the gap. In both cases the test voltage has to be applied for 1 minute. No flashovers are allowed.
The test procedure is described in IEC 60060-1.
Since the short-duration power-frequency voltage test is not necessary for insulation coordination purposes, it is
used more for quality assurance concerning line-to-earth voltages. The values serve as a basis for routine tests
and they are also the pre-stress voltage for the partial discharge measurement.
Concerning the AC test across the open switching device and/or isolating distance, the phase-opposition conditions
may be the background, but this is nowhere stated. For UHV systems (see Table 8.4.4), the values for lightning
impulse withstand voltages are considerably higher in comparison to the test voltages of the lower voltage systems.
Beside the pure power-frequency voltage tests, the AC voltage is also used for the combined voltage tests for the
longitudinal insulation - across the open gap of the switching devices. In this case the impulse voltage is
synchronized in the way that the peak-value of the impulse reach the peak-value of the power-frequency voltage in
the opposite polarity. The sum of both voltages is the test voltage across the gap.
In the combined voltage test, the test voltage is generally 0.7 p.u. (1 p.u. = peak of line-to-ground voltage) of the full
lightning impulse level and full 1 p.u. against a reduced switching impulse level.
Highest
voltage of
equipment
Ur
(kVr.m.s.
value)
Rated short duration

power frequency
withstand voltage
Ud
kV (r.m.s. value)
Phase-toearth and
between
phases
1100
1200
Rated switching impulse withstand

voltage
Us
kV (peak value)
Across
open
switching
device
and/or
isolating
distance
Phase-toearth and
across
open
switching
device
1100
Rated lightning impulse

withstand voltage
Up
kV (peak value)
Between
phases
Across
isolating
distance
Phase-toearth and
between
phases
Across open
switching
device and/or
isolating
distance
1550
2635
1550+(900)
2250
2250+(630)
1100+(635)
1800
2880
1675+(900)
2400
2400+(630)
1200
1800
2970
2400
2400+(685)
2550
2550+(685)
1100
1200
1675+(980)
1200+(695)
1950
3120
Table 8.4.4 Extension of highest voltages of equipment for UHV-Systems in the IEC
62271-1 A1 Ed.1: [7]
The situation in the standard for GIS (IEC 62271-203) is a little bit different concerning the power-frequency test
voltages. Generally the power-frequency test voltages are higher for line to ground and across the open switching
devices in comparison to IEC 62271-1. The reason for choosing higher test values was to achive higher reliability.
The values have been fixed with the factor of 0.45 in relation to the lightning impulse withstand voltage values.
Combined voltage tests in IEC 62271-203 are similar to those of IEC 62271-1.
Page 281
TypethenameofTB
The latest revision of IEC 62271-203 does not have any standardised values for UHV. The standard only lists
insulation withstand values as applied in some countries. Compared with systems in the lower voltage ranges (as
described before) the withstand values of the different countries are increased especially for the longitudinal
insulation.
Table 8.4.5 Insulation levels used for GIS in UHV systems (Annex G) [8]
8.4.2.2 LONG-DURATION OVER VOLTAGE TEST FOR GIS, CB, ETC. (JAPANESE CASE)
In Japan, gas insulated apparatus (GIS, GCB, etc.) were also examined with the same technique as abovementioned transformer [3]-[5]. The outline is described below.
(1) V-t characteristics of GIS
Tests were carried out as part of an investigation of the long-term V-t characteristics by using a 72kV single-phase
bus bar as shown in Figure 8.4.3[9]. Figure 8.4.3 also shows the results of the long-term V-t characteristics of 72kV
busbar. As the 50% breakdown V-t characteristics, they were expressed as inversely proportional to 82nd power of
the time. These results demonstrated that gas-gap insulation did not much deteriorate even after long-term voltage
application.
Table 8.4.6 show the necessary data of the insulation characteristics for the calculation of the testing voltages
obtained from the V-t characteristics. In addition, with GIS, there are no differences in short- and long-time
characteristics as the only insulation medium is gas.
Page 282
TypethenameofTB
CIGRE 1990 Session Paper:

No.15/33-03 Fig.3 & 4
n = 82
Figure 8.4.3 Long-term V-t characteristics of model GIS

Short time range
GIS
Long time range
82
0.232
19
82
0.232
19
Table 8.4.6 V-t characteristics in GIS [1]

(2) Condition of the stress under the testing voltage and operating voltage
Table 8.4.7 shows the stresses in the operation within the 30 years for GIS. This Table is similar to Table 8.4.2 for
Transformers, except for Duration Ts (0.2s) of Load rejection.
Temporary over-voltage
Load rejection
Operating
voltage
Parameter
One-line ground fault
Over-voltage Vs
1.1pu
1.1pu
1.5pu
1.0pu
Duration Ts
0.1s
2.0s
0.2s
30years
Frequency N
90
10
Table 8.4.7 Power frequency voltage endured by UHV GIS

(3) Short and long-term test voltages for GIS
According to the same manner of transformer (0 Evaluation of the test voltages), the following short and long-term
test voltages for GIS can be obtained from the conditional equation above (a) and (b).
Table 8.4.8 shows the test mode for GIS.
Page 283
TypethenameofTB
For GIS, because the insulation characteristics of the equipment do not differ between short- and long-term zones,
verification as long-term zones seem to be permitted with short-term (several minutes) voltage application.
However, long-term test are preferable since problems would occur under relatively long-term voltage application.
Test voltages
V1
GIS
1.5E
V2
3E
Applying periods
V3
t1
t2
t3
1.5E
30min
1min
30min
Note: Vn and tn are shown in Figure 8.4.1.
Table 8.4.8 Long-duration AC voltage test mode for GIS

REFERENCES
[1] IEC 60076-3 Second edition (2000-03) Power transformers - Part 3: Insulation levels, dielectric tests and
external clearances in air
[2] IEC document 14/716/CDV for IEC 60076-3 Ed. 3 Power transformers - Part 3: Insulation levels, dielectric tests
and external clearances in air
[3] T. Watanabe, K. Yamagata and E. Zaima, Insulation Coordination for UHV System, CIGRE Session 33-101,
1998
[4] E. Zaima, Insulation Design for 1100kV Substation in TEPCO, Proc. UHAVC Workshop in China, 2005.
[5] E. Zaima, T. Shindo and M. Ishii, System Aspects of 1100kV AC Transmission Technologies in Japan:
Solutions for Network Problems Specific to UHV AC Transmission System and Insulation Coordination,
IEC/CIGRE Symposium on International Standardization for UHV Transmission, Beijing, 2007.
[6] S. Okabe, T. Ohno and E. Zaima, V-t and V-N Characteristics of Insulation Elements with Large Scale OilImmersed Transformers, Proc. 14th ISH, I-13, 2005.
[7] IEC 62271-1 Amd.1 (2011-06) High-voltage switchgear and controlgear -- Part 1 - Common specifications UHV-addition
[8] IEC 62271-203 (2011-09) High-voltage switchgear and controlgear -- Part 203 - gas insulated metal enclosed
switchgear above 52 kV
[9] T. Kawamura, S. Kobayashi, Y. Mukaiyama, K. Saikawa, K. Sasaki, Y. Murakami, T. Nitta, Dielectric
Deterioration and Dielectric Diagnosis of GIS, CIGRE Session 15/33-03, 1990
Page 284
TypethenameofTB
8.5 Conclusion
Section 8.1 describes the procedure for insulation coordination, specifying the simplified method and the safety
factor. Section 8.2 shows concretely the process of determining the LIWV and SIWV, and verification of SIWV for
TOV. Section 8.3 covers the consideration of insulation coordination for VFTO and Section 8.4 explained the power
frequency voltage tests for substation equipment with reference to the relevant IEC standards.
The simplified method provides a preliminary estimation based on the residual voltage with the nominal discharge
current of a surge arrester. The residual voltage comes from V-I characteristics with the nominal discharge current.
The nominal discharge current for UHV surge arresters is studied in IEC TC37 MT4. Detailed computer simulations
are generally carried out for UHV systems.
quality. The current safety factor in IEC 60071-2 was adopted 20 years ago as an overall value obtained from field
experiences.
The safety factors used in practice range from 1.00 to 1.32 and those recommended in IEC 60071-2 are 1.05 or
1.15. Data on insulation deterioration characteristics due to ageing have been accumulated and also the quality
control, and assembly and installation technology of equipment components and equipment manufacturing have
been significantly improved. Based on such a situation, it might be one option to allow some range for the
recommended value of the safety factor instead of using it a constant fixed value.
According to IEC 60071-1 TOVs should be basically covered by SIWV. Following this way of IEC 60071-1 the
standard switching impulse voltages equivalent to TOV range from 2383 kV to 2566 kV in the UHV projects and
they require a standard switching impulse withstand voltages of at least 2400 kV in UHV systems. Since SIWVs
greater than LIWVs are economically and technically unrealistic and not feasible, the power-frequency voltage test
as an alternative test is available to verify the withstand capability concerning TOV. The power-frequency withstand
voltage tests are provided by the relevant apparatus committees.
In cases where the calculated VFTO caused by GIS-disconnector switching is higher than the withstand level,
special measures to mitigate the VFTO are required. Japanese and Chinese experiences are introduced. Generally
a three-step procedure is proposed as a general insulation coordination approach: 1) Calculation of VFTO (peak
and rise time), 2) Comparison of calculated VFTO values with LIWV level by using coordination factor, safety factor
and test conversion factor, and 3) Decision and definition concerning the necessity of measures to reduce the
VFTO level.
Japanese and Chinese UHV projects apply the long-duration induced AC voltage test (ACLD) for UHV power
transformers, whose time sequence is the same for both projects. The test voltages and time period are estimated
by the Weibull distribution function in consideration of reliability. In the meantime, the maintenance work of the IEC
60076-3 has been started and the description of the CDV of IEC 60076-3 Ed.3 (August, 2012) is introduced in the
chapter.
For other equipment the standard Common clauses of substation equipment IEC 62271-1 specifies a shortduration power-frequency withstand voltage for phase-to-earth and between phases and for across the open
switching devices. The test across the open switching device can be done with two power-frequency sources in
out-of-phase condition across the gap. In both cases the test voltage has to be applied for 1 minute. No flashover is
allowed. According to national requirements in Japan, gas insulated apparatus (GIS, GCB, etc.) were also
examined with the same technique as the above described procedure for transformers.
Page 285
TypethenameofTB
9 Conclusion and Recommendation

WG C4.306 has reviewed and discussed insulation coordination practice in the UHV AC range exceeding 800 kV
taking into account the state-of-the-art technology, with special reference to higher performance surge arresters.
Such a review has been taken into account the accumulated knowledge of various CIGR working bodies, recent
measured data of VFTO and air-gap dielectric characteristics, and accomplished in collaboration with related SC
A3 and B3 (WG A3.22, A3.28, B3.22 and B3.29).
Recommendations, such as the recent practice of insulation coordination based on the higher performance surge
arrester, estimation of overvoltages to peculiar to UHV (VFTO etc) and air clearance, will be proposed to the
application guide IEC 60071-2 (1996) and IEC apparatus standards.
9.1 Recent Practices on insulation coordination for UHV and 800 kV system
Economical and highly reliable transmission lines and substations equipment with environmental considerations
are essential in the UHV system. UHV clearance in the air may require much longer distance than those applied to
EHV systems, because the dielectric strength for switching overvoltage does not increase linearly with an increase
of the air gap. Therefore reducing the size of transmission lines and substation equipment are practical
countermeasures.
In Chinese, Indian, Japanese UHV projects, suppressing overvoltage by higher performance surge arresters is a
common countermeasure, and additional countermeasures, such as suppressing overvoltage by the circuit
breakers with closing and/or opening with pre-insertion resistors, are adopted in each project. In these projects,
overvoltages are simulated by the latest analyzing technology such as EMTP.
9.2 Overvoltages in UHV range

WG surveyed overvoltages and simulation method /conditions for overvoltages specific to UHV and 800 kV system
in nine countries.
The precise calculation methods, advanced simulation and mitigation techniques for the overvoltages in the system
can achieve the reduction of the dielectric requirements. The analytical tools like EMTP are very useful and
commonly applied for the precise evaluation of the overvoltages. These advanced analytical techniques and the
knowledge are recommended to be added to IEC 60071 Insulation Coordination.
(1) TOV
The amplitude of temporary overvoltage (TOV) levels and energy absorption of surge arrestor in UHV systems are
relatively higher than those in EHV systems.
In UHV systems of various countries, representative levels of temporary overvoltages range from 1.3 to 1.5 p.u.
respectively. The energy absorption of surge arrester for temporary overvoltages of UHV system is specified from
40 to 55 MJ, which are higher than those for 800 kV and EHV systems
(2) Switching overvoltage
The waveforms of switching overvoltage occurring in UHV transmission lines are estimated and their front times are
more than several hundred microseconds, higher than those in EHV systems. Representative levels of switching
overvoltages range from 1.6 to 1.7 p.u., Mitigation measures such as circuit breaker (CB) with opening and/or
closing resistor, shunt reactor and appropriate MOSA installation are verified to be effective against system
overvoltages.
Page 286
TypethenameofTB
(3) Lightning overvoltage

The ratios of representative overvoltages to LIPL of a surge arrester for transformers in UHV systems range from
1.0 to 1.17 p.u. for the ratios for other equipment range from 1.0 to 1.36p.u. The ratios for UHV systems are likely
to be smaller than those for 800 kV and the lower systems, because the overvoltage mitigation techniques such as
higher performance arrestors can drastically reduce the lightning overvoltage levels.
(4) VFTO
New on-site test data have been systematically measured in China. The cumulative Trapped Charge Voltage (TCV)
distributions lead to the conclusion, that the value of -1 p.u. TCV is a conservative assumption. The assumption of
a more realistic value of the TCV, can lead to more realistic insulation co-ordination calculations. The application of
slow operating disconnector doesnt produce high VFTO, but increase of both sparking time and the number of
sparks during the disconnector operation process.
Damping resistor installed in DS can significantly suppress the amplitude and gradient of VFTO caused by
according to VFTO simulation considering the substation layout of GIS/Hybrid-IS. Regarding with neighboring
9.3. Evaluation of overvoltages

(1) Overvoltage simulation tools and verification by measuring results
In the analysis of power system, digital simulations such as EMTP have been used in many countries. The
comparison estimation is also performed the analysis results and the measured value in each country, then, its
validity has been confirmed up to UHV system. For example, in the case of UHV line in Japan and China, the
maximum error between measurements and simulation for switching overvoltages is about 5 %, which is agreeable
in the design of transmission lines.
(2) Higher performance surge arrester
The main characteristics of MOSAs for the UHV projects in the world have been investigated in the WG. The UHV
higher performance MOSAs have a higher protective performance due to their lower residual voltage level, which is
a decisive factor for UHV power transmission insulation design. To illustrate the characteristics of the higher
performance MOSA should be desirable to IEC60071-2.
(3) Evaluation of overvoltage waveform
Actual field overvoltages of non-standard lightning impulse waveform are analyzed and the insulating
characteristics of SF6 gas and oil-filled transformer elements for these actual overvoltages are clarified to convert
the waveform into the standard lightning impulse waveform. The waveform evaluation method has already been
put into practical use to UHV and it could be possible in some cases to use lower withstand voltages. The
evaluation method should be reflected in the IEC60071-2 as specific method.
9.4 Switching overvoltage mitigation measures for future UHV systems

The study results for a transposed 500 kV transmission line illustrate the combined contribution of line end
arresters and closing resistors or a controller to switching overvoltage limitation and those results can certainly be
viewed as indicative of performance. Accurate representation of arrester and controller characteristics, resistor
insertion and line configuration is essential in order to achieve a valid design basis for lines meeting dependability
requirements.
Page 287
TypethenameofTB
9.5 Review on insulation coordination of air gaps in the UHV range

Insulation Coordination of air gaps in the UHV range are reviewed as hints and recommendations for IEC TC 28.
Recent investigations from Japan, India and China are presented.
Transferring the U50 measurements from India and China to a withstand voltage U10 by calculation and comparing
these results with the recommended air gap clearances of IEC 60071-1 show, that the Conductor-Structure
clearances are rather close to the measurements and it supports the validity of the clearances selected in IEC
60071-1 for the UHV levels. The much larger clearances for Rod-Structure air gaps in IEC fit more or less to the
measurements of tower window configurations, but a correlation should be handled with care due to various
parameters which cannot be figured out from these measurements.
The u-curve behavior of flashover voltages with non standard switching impulses is valid for rod-plane
configurations. From investigations in the past as well as from the recent measurements from India and China it
can be seen that configurations with a more homogeneous field distribution and configurations including insulators
do not show this behavior. For actual configurations in a tower with and without insulators the critical time to peak is
near to the peak of the standard wave shape.
9.6 Selection of insulation levels

(1) Safety factor
quality. The current safety factor in IEC 60071-2 was adopted 20 years ago as an overall value obtained from field
experiences.
The safety factors used in practice range from 1.00 to 1.32 and those recommended in IEC 60071-2 are 1.05 or
1.15. Data on insulation deterioration characteristics due to ageing have been accumulated and also the quality
control, and assembly and installation technology of equipment components and equipment manufacturing have
been significantly improved. Based on such a situation, it might be one option to allow some range for the
recommended value of the safety factor instead of using it a constant fixed value.
(2) LIWV
Either 1950 kV or 2250 kV for the LIVW of transformers, and 2250 kV or 2400 kV for other substation equipment
are adopted. The LIWV of other equipment is higher than that of transformers. As for the ratios between LIWV and
representative overvoltage, a value of 1.02 used in Japan for other equipment is the smallest, while 1.41 used in
India is the largest. This may result from different analytical conditions assumed in these countries, i.e., Japan sets
higher representative overvoltages and smaller safety factors. India sets comparatively lower representative
overvoltages which correspond to a residual voltage of a surge arrester (V20kA), based on the simplified method.
(3) SIWV
Each country sets either 1425 kV or 1800 kV for the SIVW of transformers, and 1550 kV or 1675 kV or 1800 kV for
the SIWV of other equipment. Russia, India and China set 1800 kV for both transformers and other equipment.
From the viewpoints of the ratios between SIWV and representative overvoltage, the smallest value of 1.09 is used
in Japan for transformers, and the largest value of 1.29 is used in China for transformers. Differences among
countries are less significant than those for the lightning impulse withstand voltage. In p.u., the representative
overvoltages range between 1.46 and 1.69, which are significantly lower than those in lower voltage systems in
general.
Page 288
TypethenameofTB
(4) VFTO
In cases where the calculated VFTO level due to GIS-disconnector switching is higher than the withstand level,
special measures to mitigate the VFTO are required. Japanese and Chinese experiences and approaches are
presented A three-step procedure is proposed as a general insulation coordination approach: 1) Calculation of
VFTO (peak and rise time), 2) Comparison of calculated VFTO values with LIWV level by using coordination factor,
safety factor and test conversion factor (the exact values of these factors cannot be given at the moment), and 3)
Decision and definition concerning the necessity of measures to reduce the VFTO level.
(5) Power frequency voltage test
Japanese, Chinese and Indian UHV projects apply the long-duration induced AC voltage test (ACLD) for UHV
power transformers, whose time sequence is nearly same for these projects. According to IEC 60071-1 TOVs
should be basically covered by SIWV. Following this way of IEC 60071-1 the standard switching impulse voltages
equivalent to TOV range from 2383 kV to 2566 kV in the UHV projects and they require a standard switching
impulse withstand voltages of at least 2400 kV in UHV systems. Since SIWVs greater than LIWVs are economically
and technically unrealistic and not feasible, the power-frequency voltage test as an alternative test is available to
verify the withstand capability concerning TOV. The power-frequency withstand voltage tests are provided by the
relevant apparatus committees. Japanese, Chinese and Indian UHV projects apply the long-duration induced AC
voltage test (ACLD) for UHV power transformers, whose time sequence is nearly same for these projects.
Page 289

Insulation Coordination For UHV AC Systems PDF

Enviado por

Dados do documento

Descrição original:

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Insulation Coordination For UHV AC Systems PDF

Enviado por

Direitos autorais:

Formatos disponíveis

INSULATION

ISBN : (To be completed by CIGRE)

ISBN : (To be completed by CIGRE)

4 Overvoltage in UHV range...........................................................................................83

and phase-to-phase insulation).................................................................................211

7.4 Air Gap Clearance Calculation for SIWV ............................................................214

IEC 60071-1 Ed 8.1

1. Recent practice on insulation coordination for UHV system

Proposal of recommendation for application guide IEC 60071-2

Figure 1.1 Task of CIGR WG C4.306

2 Concept of recent practices on insulation coordination for the

2.1 Insulation coordination throughout substation and transmission line

2.1.1 INSULATION COORDINATION RESEARCH IN UHV SYSTEM BY CIGRE AND IEC

Table 2.1.3 Standard voltage for UHV (IEC 60038)

2.1.2 RECENT PRACTICE OF INSULATION COORDINATION FOR UHV AC TRANSMISSION

Practical application of higher performance

Reliable circuit breaker with closing

Insulation Design Level

Figure 2.1.1 UHV insulation coordination concept

Origin and classification of stressing

Representative voltages and

Selection of the insulation meet ing

Co-ordination withstand voltages Ucw

Equipment test assembly *)

Application of factors to account for

Required withstand voltage Urw

Selection of rated withstand voltages

2.1.3 OVERVOLTAGES SPECIFIC TO UHV AC TRANSMISSION SYSTEM

Figure 2.1.3 Representative maxima of amplitude of overvoltages Um (pu)

2.1.3.2 S LOW - FRONT

Withstand Voltage (kV)

(UHV-Switching OV: 2.0pu (same as 550kV))

Air Insulation Distance (m)

Tran sm ission lin e

Figure 2.1.5 Size reduction of 1100kV tower

2.1.3.3 F AST - FRONT

FAST TRANSIENT OVERVOLTAGES

2.1.4 SELECTION OF INSULATION LEVEL

(a) Jindongnan substation (China)

(b) Nanyang switching station (China)

2.2 Reduction of insulation levels using overvoltage suppression measures

2.2.1 OVERVOLTAGE SUPPRESSION WITH HIGHER PERFORMANCE SURGE ARRESTERS

Highest voltage (kV)

Table 2.2.1 Protective performance of surge arresters in substation projects

Figure 2.2.2 Porcelain type

Figure 2.2.3 Porcelain type

higher performance surge

higher performance surge

2.2.2 RESISTOR-FITTED CIRCUIT BREAKERS

Highest voltage (kV)

Table 2.2.2 Suppression methods and insulation designs of international projects

(a) Closing case in Indian project

(b) Opening case in Japanese project

Figure 0.4 Example of relation between resistance and overvoltage

Figure 2.2.5 GCB with pre-insertion resistors

Highest voltage (kV)

Table 2.2.3 Application of pre-insertion resistor in international projects

Figure 2.2.6 Disconnector with pre-insertion resistor

3 Recent practice on insulation coordination for the UHV and the

The line side, the

Table 3.1.1 Switching overvoltage design level of transmission line

AND RECLOSING NO - LOAD LINE OVERVOLTAGE