&,5(' QG,QWHUQDWLRQDO&RQIHUHQFHRQ(OHFWULFLW\'LVWULEXWLRQ 6WRFNKROP-XQH


3DSHU


CALCULATION OF LIGHTNING OVERVOLTAGE FAILURE RATES FOR A GAS

INSULATED SUBSTATION

Anthony Chabrol Neil McDonagh

ESB International – Ireland ESB International – Ireland
anthony.chabrol@esbi.ie neil.mcdonagh@esbi.ie

acceptable rate of failure for the substation due to lightning

ABSTRACT overvoltage. This paper details how to calculate a failure
rate for a given substation topology using ATP Draw
Lightning overvoltages can cause damage to the equipment software. The acceptable rate of failure can then be defined
in a substation, leading to poor utilisation of assets. In based on this calculation.
order to prevent this damage from occurring, an acceptable
level of lightning protection should be installed.
HIGH FREQUENCY SUBSTATION
Designing this level of protection requires that the MODELLING
acceptable rate of failure for the substation due to lightning
overvoltagest be defined. In order to model a substation in ATP Draw and to
sufficiently analyse resultant overvoltages resulting from a
An acceptable failure rate is normally calculated based on lightning strike, several pieces of equipment must be
experience and historic records of failure due to lightning modelled; busbar, surge arresters, overhead lines,
overvoltages. However there is no method freely available underground cables, towers and transformers. The source of
to enable designers to calculate acceptable substation lightning is another important component that must be
failure rates when this historical data is unavailable. modelled. The overhead lines were modelled as unshielded
and connected to the substation via underground cable,
The International Standard IEC-60071-2 “Insulation
which is common in Ireland.
Coordination” Third Edition 1996 states the following;
“For apparatus, acceptable failure rates Ra due to Standby feeding arrangements for events such as
overvoltages are in the range of 0.001/year up to maintenance or outages must be considered. Different
0.004/year depending on the repair times.” substation topologies such as multiple transformers or
multiple feeders connecting to the substation must also be
This paper deals with an approach to calculate a failure
considered. The ATP Draw model is shown in Figure 1.
rate for a substation based on a function of; lightning strike
magnitude, lightning strike location, surge arrester location
and substation topology. The substation, with 110kV
unshielded overhead line connected to the substation via
underground cable, is modelled using Alternative Transient
Program Draw (ATP Draw). c a
b f
ATP Draw is graphical user interface for Electromagnetic
Transient Program (EMTP), which is a software program
d e
used to simulate electrical transient phenomena.

Figure 1 – Simplified ATP Draw Illustration

INTRODUCTION
The aim of insulation coordination is to define the optimum
ATP Draw Modelling Components
network design at minimum cost. Sufficient insulation of the Busbar (a) – The busbar was modelled as a surge impedance
network, along with adequate protective equipment is [1].
required to help ensure the best performance of assets. Surge Arresters (SA) (b) – The surge arrester was modelled
Typically the overvoltages seen at a substation as a result of using the MOV component in ATP Draw.
switching or lightning are the most important factors to
consider when carrying out an insulation coordination study. Tower Footing Impedance (TFI) (c) – The tower footing
An important element in limiting the overvoltage at a impedance was modelled as an RLC circuit as shown in
substation is the employment of overvoltage protection section 7.6.11.1 of [2].
equipment i.e. surge arresters and adequate earthing. Overhead Lines (OHL) (d) – The OHL was modelled using
Designing this level of protection requires a definition of the a JMarti model in ATP Draw.

&,5('6HVVLRQ 3DSHU1R 
&,5(' QG,QWHUQDWLRQDO&RQIHUHQFHRQ(OHFWULFLW\'LVWULEXWLRQ 6WRFNKROP-XQH

3DSHU

Voltage at Transformer
Underground Cables (e) – The underground cables were 500

modelled as surge impedance [2]. [kV]

350
Transformers (f) – The transformers were represented as
capacitances to ground [1].
200

Lightning Source (g) – The lightning source was modelled

as a Heidler Source in ATP Draw [3]. 50

SIMULATIONS -100

A lightning strike is simulated at a point on the OHL

-250
connected to the substation. The energy from the strike will
result in an overvoltage on the OHL, which if large enough
will flash across the three phases. Part of the energy from -400
0 5 10 15 20 [us] 25
the strike will dissipate through the TFI of the angle masts (file case_1_TUES_1_RUNNING2.pl4; x-var t) v:TRAFOA-EARTH v:TRAFOB-EARTH v:TRAFOC-EARTH

and SA at the cable interface. The strike will also lead to an Figure 2 – Plot of Voltage at Monitored Equipment
overvoltage on the substation equipment. If the overvoltage
is large enough it will exceed the basic insulation level
(BIL) rating of the substation equipment affected. The highest voltage level for each separate simulation can
be recorded and displayed as shown in Figure 3. If any of
The voltage level at pieces of substation equipment deemed the voltage levels exceeds the BIL for a piece of substation
to be vulnerable, such as the busbar or transformer, should equipment the distance and magnitude of the strike should
be calculated. In the cases where the BIL is exceeded, the be noted for use in the failure rate calculation.
lightning strike magnitude and overvoltage at the piece of
equipment is recorded.
Lightning strikes of different magnitudes are simulated at Example Scenario Plot
varying distances from the substation. The furthest distance
away from the substation considered should be a point at
which the resultant overvoltage from any magnitude of
Voltage (kV)
lightning strike will not exceed the BIL of the substation
440
BIL
equipment. This distance can be estimated as shown in [4].
This will allow the results to show a definitive distance from
the substation at which lightning can cause dangerous
overvoltages, beyond which a strike should not cause the
BIL of the substation equipment to be exceeded. 100
500

1000
Table 1, shows the magnitude and distance from the 0
Distance (m) 200
substation at which lightning strikes were simulated. 10000
20

20000 5
Strike Magnitude (kA)
1

Figure 3 – Plot of Overvoltages

Figure 3 shows an example for a particular substation

topology and surge arrester deployment. In this case
Table 1 – Magnitude and Distance of Strikes lightning strikes that are above 80kA result in the BIL being
exceeded when the strike is 1km or closer to the cable line
interface. It can also be seen that strikes of 200kA result in
When a lightning strike is simulated, the voltage at a piece the BIL being exceed at 1km or closer to the cable line
of equipment can be plotted, as shown in the example in interface.
Figure 2.

&,5('6HVVLRQ 3DSHU1R 
&,5(' QG,QWHUQDWLRQDO&RQIHUHQFHRQ(OHFWULFLW\'LVWULEXWLRQ 6WRFNKROP-XQH

3DSHU


FAILURE RATE CALCULATION The lightning strike data provided for Region A was plotted
as shown in Figure 5. The probability of a lightning strike of
The failure rate was calculated considering the probability a given magnitude or higher occurring was calculated by
of a lightning strike on the OHL connected to a substation using this figure.
and the probability that this strike is of sufficient magnitude
to cause the BIL of substation equipment to be exceeded. For example, in Figure 5 if we have a lightning strike, the
probability that the magnitude of the strike will be 20kA or
The lightning flash collection rate is the number of lightning higher is approximately 9%. This probability is represented
flashes per 100km per year for a given region. This data
by S .
must be provided for the study.
The lightning flash collection rate for a given poleset or S is calculated for strike magnitudes at which the BIL of
substation equipment is exceeded.
tower per 100km/year, N L , is given by
28 H 0.6 + b
NL = Ng ⋅ Probability of Strike of a Given Magnitude at Region A
10 100

where 90

N g is the ground flash density (flashes/km 2 /year). 80

70
Probability in %

H is the pole / tower height in meters 60

b is the structure width in meters 50

40

The probability of a lightning flash N M , for a given line 30

20

length is given by 10

N
NM = L ⋅l
0
0 10 20 30 40 50 60 70 80

100 Magnitude of Strike

Where l is the length of line in km. [5]. Figure 5 – Lightning Strike Magnitude Distribution

N M is calculated for each of the distances at which a

lightning strike is simulated. N S is the probability of a lightning strike overvoltage
For example, if which exceeds the BIL of substation equipment occurring
over a given length of line.
N g = 0 .1
H = 10m NS = NM ⋅ S
b = 5m
The different values for N M would be as shown in Figure A value for N S is calculated at each distance from the
4; substation where a simulated lightning strike overvoltage
exceeds the BIL of substation equipment.

0.011647 The summation of the different values for N S will give an

overall rate of failure for that particular substation.
0.009318 N Stotal = N S 100 km + N S 200 km + N nkm ...
0.005824

0.002329

0.001165

Figure 4 – Probability of Strike per distance from

Cable Interface

&,5('6HVVLRQ 3DSHU1R 
&,5(' QG,QWHUQDWLRQDO&RQIHUHQFHRQ(OHFWULFLW\'LVWULEXWLRQ 6WRFNKROP-XQH

3DSHU


Example: REFERENCES
If the results from Figure 3 and the values for N M as
[1] IEEE Working Group on Modeling and Analysis of
shown in Figure 4 are used, the two lightning strike
System Transients, 1996, Modeling and Analysis
magnitudes that exceed the BIL are 80kA and 200kA, at a
Guidelines for Very Fast Transients in Gas Insulated
distance from the cable interface of 1km for both strikes.
Substations.
Since a strike of 80kA is sufficient to exceed the BIL at 1km
and the 200kA strike does not exceed the BIL at a distance
[2] IEC 60071-4, 2004, Insulation Coordination Part-4:
further than 1km, N S should only be calculated for 80kA. Computational Guide to Insulation Coordination and
Modelling of Electrical Networks.
S for an 80kA strike, using Figure 5 = 1.9% or 0.019
N M for 1km = 0.011647 [3] Working Group 01 (Lightning) of Cigre Study
∴ N S = 0.011647 * 0.019 = 0.000221293 / year Committee 3, 1991, Guide to Procedures for
Estimating the Lightning Performance of Transmission
Lines.
If we compare this to the IEC acceptable failure rate of
0.001/year we can see that in this example the given level of [4] IEC 60071-2, 1996, Insulation Coordination Part 2:
lightning overvoltage protection is sufficient. Application Guide.
Figure 3 show the results from 63 simulations, used to
calculate a failure rate, for one topology type. In order to [5] IEEE STD 1410, 2011, IEEE Guide for Improving the
calculate a failure rate for stations with different topologies Lightning Performance of Electric Power Overhead
and feeding arrangements, many more simulations would be Distribution Lines.
required.

CONCLUSIONS
This method for failure rate calculation is useful for
examining the effectiveness of overvoltage protection on
substations where there are no prior records of failures due
to lightning overvoltages.
It is also an efficient way of examining overvoltage
protection for new substation designs of different feeding
arrangements and topologies.

ACKNOWLEDGMENTS
The author would like to thank Danijela Klopotan, James
Dooley, Martin Campion and Damien Glennon, Power
Systems Studies Team, ESB International for their
contribution and suggestions with the compilation of this
paper.

&,5('6HVVLRQ 3DSHU1R