Ep 21 00 00 01 SP

Engineering Standard
Electrical
Engineering Standard
EP 21 00 00 01 SP
INSULATION CO-ORDINATION AND

SURGE ARRESTER SELECTION
Version 3.1
Issued May 2013
Owner: Chief Engineer, Electrical
Approved Neal Hook Authorised Neal Hook

by: Chief Engineer by: Chief Engineer
Electrical Electrical
Disclaimer
This document was prepared for use on the RailCorp Network only.
RailCorp makes no warranties, express or implied, that compliance with the contents of this document shall be
sufficient to ensure safe systems or work or operation. It is the document user’s sole responsibility to ensure that the
copy of the document it is viewing is the current version of the document as in use by RailCorp.
RailCorp accepts no liability whatsoever in relation to the use of this document by any party, and RailCorp excludes
any liability which arises in any manner by the use of this document.
Copyright
The information in this document is protected by Copyright and no part of this document may be reproduced, altered,
stored or transmitted by any person without the prior consent of RailCorp.
UNCONTROLLED WHEN PRINTED Page 1 of 13

RailCorp Engineering Standard — Electrical
Insulation Co-ordination and Surge Arrester Selection EP 21 00 00 01 SP
Document control
Version Date Summary of change
3.0 May 2010 Application of TMA 400 format
3.1 May 2013 Update template
© RailCorp Page 2 of 13
Issued May 2013 UNCONTROLLED WHEN PRINTED Version 3.1
Contents
1 Introduction .............................................................................................................................4
2 References...............................................................................................................................4
3 Basic Insulation Level ............................................................................................................4
4 Lightning ..................................................................................................................................6
4.1 Typical Lightning Characteristics ..............................................................................................6
4.2 Lightning & Surge Arrester Current...........................................................................................7
4.3 Lightning Strike Rate.................................................................................................................8
4.3.1 Direct Strikes .............................................................................................................8
4.3.2 Nearby Strokes to Ground.........................................................................................9
4.4 Overhead Earth Wire Shielding ................................................................................................9
4.5 Outage Rates ............................................................................................................................9
4.5.1 Example of Outage Rate Calculation ........................................................................9
5 AC Surge Arrester Types .....................................................................................................10
6 Selection & Specification of Arresters................................................................................11
6.1 Parameters for Arrester selection ...........................................................................................11
6.2 Standard Arrester Ratings.......................................................................................................12
6.3 Procedure for Selecting Gapless MOV Arresters ...................................................................12
6.3.1 Example of Arrester Selection Procedure ...............................................................13
1 Introduction
This document explains the methods employed to ensure that equipment is not damaged
by lightning.
It provides data on lightning characteristics and frequency, and on the impulse withstand
levels of common insulating materials. It describes the various types of AC arresters in
use.
It details the steps necessary to calculate the type of arrester needed to protect a given
situation and provides an example calculation for this
2 References
AS 1824.1 Basic Principles, Standard Insulation Levels and Test Procedures.
AS 1824.2 Insulation Co-ordination-Application Guide.
AS 1307 Surge Arresters
AS 1768 Lightning Protection
AS 1429 Polymeric insulated cables up to 19/33 kV.
ESAA Annual Lightning Strike Counter Figures
M Darveniza: A Review of Modern Lightning Protection for Sub-transmission and
Distribution Systems
3 Basic Insulation Level

The process of ensuring that equipment is not damaged by overvoltages - usually due to
lightning on the RailCorp system - is called Insulation Co-ordination.
The Basic Insulation Level (BIL) of equipment such as transformers, is defined as the
peak voltage of a 1.2/50 μs wave which does not cause insulation to fail; see AS 1824.1.
This definition of a surge (1.2/50 μs) indicates a surge which reaches its peak value in
1.2 μs and decays to half the peak in 50 μs.
Lightning is a current wave. Testing of equipment BIL is performed using voltage waves
that have different rise times to what are now known to be typical lightning surges. These
test voltages have been ‘historically’ used for testing but it is now recognised that this not
a typical lightning representation. The waveform was originally derived from the response
time of gapped Silicon Carbide (SiC) arresters.
Transformers are also subjected to a chopped - wave test where a 1.5 * BIL wave is
chopped off after 3μs by a gap flashing over. This is to simulate the effect of a
transmission line flashing over. The equipment covered by these values includes
switchgear, transformers, cables, etc.
RailCorp specifies equipment BIL to AS 1824.1 as summarised in Table 1.
1 Minute, Dry, Power

BIL
Nominal System Voltage Frequency, Flashover
(kV peak)
voltage(kV RMS)
1500Vdc cable 60
Note 1
11kV 95 28
Note 2
33kV 200 70
66kV 325 140
132kV 550 230
Table 1 - BIL Levels
Note 1: If all 11kV feeders supplying the SS consist of at least 0.5 km of cable, and
all equipment is indoor, a BIL of 75 kV is used.
Note 2: If all 33 kV feeders supplying the SS consist of at least 0.5 km of cable, and
all equipment is indoors, a BIL of 170 kV is used.
Typical values of BIL for transmission lines can be estimated using the data in Table 2.
Material BIL
Wood/Plastics (wet) 350 kV/m
Air (wet) 700 kV/m
11 kV pin insulator 100 kV
33 kV pin insulator 200 kV
Disk type - 1 insulator 125 kV
Disk type - 2 insulators 255 kV
Table 2 - Typical BIL Values
The BIL of different types of materials when in series can be estimated as follows:
• The BIL of wood, plastic and fibreglass in series are added arithmetically.
• When combining series wood etc. and insulators, use ⌦ ∑ BIL2 as an
approximation.
Using values from Table 2, the BIL of some typical RailCorp lines is as follows:-
Timber poles
& crossarms Unpainted
Insulator Steel or Concrete (painted to timber poles &
BIL existing crossarms
level Poles or crossarms
standard) Wet
Wet
ϕ-ϕ kV ϕ-E kV ϕ-ϕ kV ϕ-ϕ kV
11 kV pins 200 100 330 415
11 kV suspension 250 125 275 320
11 kV suspension
700 350 - -
through air
33 kV post type
33 kV suspension 345 170 425 525
33 kV suspension
1000 500 - -
through air
66 kV suspension
760 380 800 900
4 insulators
66 kV suspension
990 490 1050 1100
5 insulators
66 kV suspension
1500 750 - -
through air
Table 3 - Typical Construction BIL
Since these transmission line values are above the equipment values, surge arresters are
required to protect the equipment.
4 Lightning
4.1 Typical Lightning Characteristics

Table 4 is derived from AS 1768 - Table A1, and shows the characteristics of lightning
strokes.
Minimum Typical Maximum

Peak current kA - 10-20 110
Front rise time μs 0.1 2 30
Time to decay to half peak μs 10 40 250
Total duration sec 0.01 0.2 2
strokes per flash 1 3-4 26
Table 4 - Typical Lightning Stroke Current
The stroke median peak current is 34 kA with a 2.5% chance of exceeding 100 kA.
peak current
10 kA
half peak
1 10 Tim e μs
Figure 1 - Typical Lightning Current Waveform
17% of surges currents have a rise time less than 1μs
57% of surges currents have a rise time less than 2.5μs
AS 1768, Appendix C suggests that the following represents an "idealised lightning stroke
current for design purposes" (Design of lightning protection systems - not surge
arresters).
First stroke: 150 kA 4.6/40 μs
Subsequent: 40 kA 0.2/20 μs
A more realistic worst case surge arrester current for distribution systems is 20 kA with a
1/10 μs shape with maybe 10 kA, 1/10 μs as typical. See Figure 1.
The standard AS 1307 calls for a 8/20 μs current wave for testing equipment. This has
been ‘historically’ used for testing but it is now recognised that this is not a typical
representation. It was originally derived from the response time of gapped Silicon Carbide
arresters.
4.2 Lightning & Surge Arrester Current

Table 4 shows the lightning stroke current. The current in an arrester is usually much
less than this. When the lightning hits a conductor, the stroke current divides in two, with
half the current going in each direction.
This current in the conductor causes a voltage to be generated. The approximate voltage
is calculated by multiplying the line current by the line characteristic impedance which is
around 500 - 600 ohms for RailCorp transmission lines. From Table 3 it can be seen that
the following stroke currents will almost certainly cause a flashover at the first crossarm:-
11 kV (say 400 kV ) 1.6 kA
33 kV (say 525 kV) 2.1 kA
66 kV (say 900 kV) 3.6 kA
These are very low values so it is almost certain that every lightning strike to an RailCorp
line will cause a flashover at the first pole.
When flashover occurs, the current divides into all the conductors. For a 3 phase line,
one sixth of the current travels down each line in each direction. ie One twelfth of the
stroke current continues to flow down each conductor to be ultimately discharged
to ground by an arrester.
A typical 40 kA strike only requires a 3.6 kA rated arrester and even a major 150 kA
strike means only 12.5 kA in an arrester.
Studies have shown that "Distribution arrester currents rarely exceed 40 kA because of
T.L. flashovers" and these high values are therefore chopped waves.
4.3 Lightning Strike Rate

It is possible to estimate how often a transmission line will be struck by lightning.
The number of Thunder days (TD) per year is documented (see AS1768). For the Sydney
area it can be taken as an average 30 with the Blue Mountains being 40. The annual
number of lightning strikes to ground can be approximated by Ng = 0.07 * TD
strokes/km2/year.
4.3.1 Direct Strikes

A tall structure, like a transmission line attracts lightning. A simple rule of thumb gives the
width of the attractive area either side of a transmission line as 6*h where h is the height
(of the highest conductor).
A line around 15m high attracts lightning for 90m on either side.
So a line of 1 km length and 15m high has a lightning ‘attractive area’ of around 0.18 km2
and can expect 0.18*30*0.07 = 0.4 strikes per year ( 0.5 in Blue Mountains and even
more because lines are higher over valleys).
This attraction only applies for flat ground and other high structures around the line will
reduce the value above by around a half.
Not all these strikes will cause a trip. The surge magnitude and wave shape affect
whether a flashover will occur and also whether a power follow occurs. Lines with wood
crossarms are able to extinguish the power follow arc without the protection operating. ie
no outage.
The percentage of strikes which do cause a trip is approximately as follows:-
ϕ-ϕ distance Wood crossarms steel crossarms

11 kV 1.5 m 35% 85%
11 kV 2.1 m 30% 85%
33 kV 2.1 m 60% 85%
33 kV 2.7 m 50% 85%
66 kV 3.25 m 60% 85%
With this data it is possible to calculate the theoretical outage rate of a transmission line.
See Section 4.5.1
4.3.2 Nearby Strokes to Ground

Nearby strikes to ground can induce currents into all 3 phase conductors. However the
induced voltages rarely exceed 200 kV and so do not cause flashover, even at 11 kV, as
long as wood crossarms are used.
4.4 Overhead Earth Wire Shielding

Overhead earth wires (OHEW) are installed on high voltage feeders to intercept lightning
strikes before they hit phase conductors. OHEW are installed on RailCorp lines as
follows:
• On all new 33kV lines.

• On existing 33 kV and above aerial lines, for 800 m adjacent to substations, to protect
substation equipment. Any direct lightning strikes to the phase conductors more than
800 m from the SS will have been attenuated to a level that the SS arresters can
handle by the time they reach the SS.
• On 33 kV and above aerial lines for 800 m from a cable to aerial termination.
• OHEW are not used on RailCorp 11 kV lines.
Conservative literature recommends a shielding angle as shown below depending on the
height of the OHEW:
Height (m) angle (deg)

20 35-40
30 25-30
40 15-20
50 5-10
4.5 Outage Rates

It is accepted industry practice that shielded aerial lines should give outage rates due to
lightning, of around 0.01 outages/km/year and unshielded lines about 0.02 to 0.08
outages/km/year.
4.5.1 Example of Outage Rate Calculation

Data:
803 feeder from Lawson to Katoomba.
Length 13.2 km.
Average height is 20m (to allow for high valley crossings.)
40 thunderdays pa.
0.5 shielding by trees.
wood crossarms.
Calculations:
attraction width = 6x20x2 = 240m total = 0.24 km.
attraction area of line = 0.24x13.2 = 3.168 km2
ground flash rate = 40x.07 = 2.8 /km2/year
line hit rate = 2.8x3.168x.5 = 4.4 per year
Therefore the outage rate = 4.4x.6 = 2.7 trips/year
5 AC Surge Arrester Types

There are 3 common types of surge arrester in use:
• Silicon Carbide (SiC).
These have a series of air gaps to withstand normal system voltage and the SiC
element has a negative resistance characteristic to clamp the voltage after the air
gaps flashover.
• Gapless Metal Oxide (MOV).
Gapless MOV types have no air gaps. The MOV element is a constant voltage
element with a set breakdown voltage. Because there are no gaps, some current
flows in the MOV at all times and this can cause thermal runaway in some
circumstances.
• Gapped Metal Oxide.
Gapped MOV types use the same MOV element (but less of them) in series with
air gaps. The air gaps and MOV element share the system voltage normally, but
once the gaps break down the MOV element limits the voltage (to a lower level
than the gapless type).
Table 5 shows the advantages and disadvantages of each type.
Type of Arrester Advantages Disadvantages

High initial sparkover voltage needed to
operate gaps. This is usually higher
No current flows than the following voltage 'clamped' by
SiC
normally. the SiC. There is also some time delay
before sparkover. Problems with
moisture ingress to gaps (breathing)
No time delay to
Because they are very voltage sensitive,
operate. Tightly
Gapless MOV overvoltages can cause deterioration of
controlled voltage during
the MOV with resultant thermal runaway.
operation.
Lower voltage needed
to operate gaps than
SiC so it operates faster
and at a lower voltage
Gapped MOV than the MOV clamping Possible moisture ingress
voltage. Clamp surges
to lower levels than
either SiC or gapless
MOV.
Table 5 - Types of Surge Arrester
6 Selection & Specification of Arresters
6.1 Parameters for Arrester selection

In order to correctly specify an arrester to protect given equipment, the following
parameters must be considered.
a) Arrester waveform.
For determining the arrester residual voltage, a waveform with 1 μs rise time &
10 kA peak is used. (This is the value given in AS 1307.2 Table 5.1 Cols 3 & 4
"Steep Current Impulse")
b) Leads Induced Voltage.
For determining the lead induced voltage, a waveform with 1 μs rise time, 20 kA
peak and a lead inductance of 1.25 μH/m is used. This corresponds to a lead
voltage drop of 25 kV/metre of lead. This wave has twice the rate of rise of the
wave in a) to allow for worst cases.
c) Margin of Protection.
This value should be greater than 0.2 for shielded locations.
This value should be greater than 0.5 for unshielded locations.
BIL − arrester Volts

M arg in of Pr otection ( MOP ) =
arrester Volts
In this expression, the term "arrester Volts" includes lead voltage drop and if
there is an open point in the system, voltage doubling can occur and so the
"arrester voltage" used in the above equation must be doubled.
d) System Earthing Considerations.
System earthing determines the voltages on unfaulted during a phase to earth

fault. Table 6 summarises the voltages for RailCorp lines.
Col 1 Col 2 Col 3

Maximum Continuous Temporary System
Nominal System Volts Volts applied to Overvoltage
Arrester kV rms
11 kV solidly earthed 6.9 kV 8.6
11 kV non-effectively
6.9 kV 11.9
earthed
20.6 kV 35.7
earthed
41.2 kV 71.3
earthed
Table 6 - RailCorp system continuous and overvoltage levels
As RailCorp 33 kV and 66 kV systems mainly originate from resistance earthed

supplies, these circuits must be considered to be "non-effectively" earthed.
Thus during faults, the unfaulted phase arresters are subjected to phase to
phase volts (for 2 seconds maximum.)
Most of the RailCorp 11 kV system has solidly earthed neutrals and so in parts
of the system that can only be fed from solidly earthed supplies, the unfaulted
phases are subjected to only phase to earth volts.
However, on solidly earthed lines, some allowance must be made for earth
resistance so the maximum arrester voltage is taken as 1.25 times phase-earth
volts.
Note
The actual earth resistance of the arresters (as opposed to the system earth )
has little effect on the efficiency of the arrester. However the earth resistance
usually needs to be kept low for safety reasons. Keeping the H.V. & L.V. earths
separate helps to stop L.V. surges damaging a transformer.
e) Temporary System Overvoltages.
As a worst case a maximum 8% overvoltage can be assumed to be provided by

supply authorities.
If assuming an 8% overvoltage will not allow a suitable arrester to be sourced,

then the actual overvoltage level that may be experienced may be obtained
from the relevant Distributor.
6.2 Standard Arrester Ratings

As a result of applying the above procedure, most RailCorp equipment is protected by
‘standard’ arresters as follows:
• 10 kA arresters are used:-
– at major substations with aerial connections

– on transformers 1 MVA and over
– on all aerial to cable terminations
• 5 kA arresters are used at all other locations.
If the ‘standard’ arresters do not provide sufficient protection, then specific arresters can
be chosen using the following procedure.
6.3 Procedure for Selecting Gapless MOV Arresters

To determine whether a given arrester will provide protection for equipment it is
necessary to check the following:-
a) That the normal Continuous overvoltage does not damage the arrester.
b) That the voltage on unfaulted phases, before the protection operates, does not
damage the arrester. The protection time can be taken as maximum 2 second
in the RailCorp system.
c) That the arresters have a MCOV (max continuous overvoltage) greater than Col
2 in Table 6.
Temporary overvoltage (during faults) data for arresters is usually available but
if not assume that typically arresters can handle a 25% overvoltage for 2
seconds.
eg. 33 kV system, non-effectively earthed.
i) max cont. ϕ-E = 21 kV
ii) temporary ϕ-E = 36 kV (33 x 1.08)
from i) could choose an arrester with MCOV > 21 kV
but because of ii) need MCOV > 36 kV/1.25 = 29 kV
d) Then consult the data table for 20 kA current wave to determine maximum
arrester voltage (around 130 kV for 33 kV).
e) Then add the lead voltage drop due to 20 kA/μs = 25 kV/m.
If the result exceeds 200 kV * (MOP+1), then that arrester does not provide suitable
protection.
6.3.1 Example of Arrester Selection Procedure

A 33 kV arrester is required for a 33 kV/240V transformer on an unshielded 33 kV line.
Data:
BIL = 200 kV. MOP = 0.5. Non-effectively earthed line
Lead Length = 2 m. Normal continuous overvolts = 8%
Available Arresters’ Characteristics are:-
5 kA type 10 kA type L.D. 10 kA type H.D.

MCOV 29 kV 29 kV 29 kV
Rated Voltage 36 kV 36 kV 36 kV
Residual voltage @ 10 kA 130 kV 130 kV 125 kV
Calculations:
Max allowable residual voltage = 200 / (1+MOP)) = 133 kV
33 x 1.08
Required max. cont. operating voltage MCOV > = 28.5kV
1.25
(The '1.25' is from the temporary overvolts characteristics, Section 6.1 d).
Therefore, choose an arrester with 29 kV.
Lead voltage for 2 m length = 2x25 = 50 kV
Conclusions:
Without the leads any of the arresters will protect the equipment since the residual
voltage of 130 kV (125 kV) exceeds the required 133 kV.
With the 2 m leads, none of the surge arresters will protect the transformer.

Ep 21 00 00 01 SP

Uploaded by

Copyright:

Available Formats

Ep 21 00 00 01 SP

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ep 21 00 00 01 SP

Uploaded by

Copyright:

Available Formats

Engineering Standard

INSULATION CO-ORDINATION AND

Issued May 2013

Owner: Chief Engineer, Electrical

Approved Neal Hook Authorised Neal Hook

UNCONTROLLED WHEN PRINTED Page 1 of 13

3 Basic Insulation Level

RailCorp specifies equipment BIL to AS 1824.1 as summarised in Table 1.

1 Minute, Dry, Power

Table 2 - Typical BIL Values

ϕ-ϕ kV ϕ-E kV ϕ-ϕ kV ϕ-ϕ kV

11 kV pins 200 100 330 415

11 kV suspension 250 125 275 320

33 kV suspension 345 170 425 525

Table 3 - Typical Construction BIL

4.1 Typical Lightning Characteristics

Minimum Typical Maximum

17% of surges currents have a rise time less than 1μs

40% of surges currents have a rise time less than 2μs

57% of surges currents have a rise time less than 2.5μs

80% of surges currents have a rise time less than 4μs

First stroke: 150 kA 4.6/40 μs

4.2 Lightning & Surge Arrester Current

11 kV (say 400 kV ) 1.6 kA

33 kV (say 525 kV) 2.1 kA

66 kV (say 900 kV) 3.6 kA

4.3 Lightning Strike Rate

4.3.1 Direct Strikes

The percentage of strikes which do cause a trip is approximately as follows:-

ϕ-ϕ distance Wood crossarms steel crossarms

4.3.2 Nearby Strokes to Ground

4.4 Overhead Earth Wire Shielding

• On all new 33kV lines.

Height (m) angle (deg)

4.5 Outage Rates

4.5.1 Example of Outage Rate Calculation

803 feeder from Lawson to Katoomba.

Length 13.2 km.

Average height is 20m (to allow for high valley crossings.)

0.5 shielding by trees.

attraction width = 6x20x2 = 240m total = 0.24 km.

attraction area of line = 0.24x13.2 = 3.168 km2

ground flash rate = 40x.07 = 2.8 /km2/year

line hit rate = 2.8x3.168x.5 = 4.4 per year

Therefore the outage rate = 4.4x.6 = 2.7 trips/year

5 AC Surge Arrester Types

• Silicon Carbide (SiC).

• Gapless Metal Oxide (MOV).

• Gapped Metal Oxide.

Table 5 shows the advantages and disadvantages of each type.

Type of Arrester Advantages Disadvantages

6 Selection & Specification of Arresters

6.1 Parameters for Arrester selection

b) Leads Induced Voltage.

This value should be greater than 0.2 for shielded locations.

This value should be greater than 0.5 for unshielded locations.

BIL − arrester Volts

d) System Earthing Considerations.

System earthing determines the voltages on unfaulted during a phase to earth