Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200

Journal of Mechatronics, Electrical Power,

and Vehicular Technology
e-ISSN: 2088-6985
p-ISSN: 2087-3379
mev.lipi.go.id

Effect of lightning mast placement on underground power

cable jacket stress within high voltage substations
Mostafa Nazih *
Future Energy, GHD Pty Ltd
999 Hay St., Perth, WA 6000, Australia

Received 13 July 2022; 1st revision 23 September 2022; 2nd revision 24 October 2022; 3rd revision 27 October 2022; Accepted 31 October 2022;
Published online 29 December 2022

Abstract
This study aims to investigate the impact of lightning masts placement on underground cables within high voltage
substations. While the subject of lightning discharges near to underground cables has been covered with open cable runs and
wind farms in many papers, this study focuses on lightning events within high voltage substations considering the associated
effective zones, which were not covered in the available literature. Substations built within areas prone to high lightning
activity experience frequent discharges that cause the potential rise of the earthing system into hundreds of kilovolts. The
potentials propagating within the soil and the earthing grid affect underground cables jackets terminated within the
substation. The numerical analysis of the problem is carried out using Current Distribution, Electromagnetic fields, Grounding
and Soil structure analysis (CDEGS) software engine for different configurations of lightning mast placements with varied
separation, electrode placement and length, soil resistivity, and lightning current. Study findings indicate that provision of
lightning masts/down conductors as far as possible or at least twice the effective zone radius from cable termination/route
electrodes ensures relatively lower stress voltages. Electrodes with effective zone radius length placed as close as possible to
lightning masts further reduce the attainable jacket stress voltages.
Copyright ©2022 National Research and Innovation Agency. This is an open access article under the CC BY-NC-SA license
(https://creativecommons.org/licenses/by-nc-sa/4.0/).
Keywords: substation earthing; lightning mast placement; high voltage; underground cable; effective zone radius.

I. Introduction system. A traditional sphere radius of 24 m is used,

which corresponds to a stroke design current of
Lightning protection air terminations represent a about 4.2 kA based on the electrogeometric method
vital part of high voltage substation switchyards [1]. (EGM) [2]. Some other utilities suggest a sphere
Many models are available to design the lightning radius of 43 m corresponding to 10 kA striking
protection systems based on various formulae for current, which is in line with the rated discharge
lightning stroke current and associated rolling current of most station type surge arrestors. The
sphere radius [2]. During a lightning stroke rationale behind that is to allow surges less than
discharged into the substation earthing system, the 10 kA to penetrate the shielding system, given these
soil will have a transient potential gradient within surges will be neutralised by surge arrestors
the substation and nearby [1]. The gradient causes typically installed near to important equipment
underground cable jacket stress that can exceed its within the substation. Surges higher than 10 kA will
insulation strength and cause a breakdown of the be intercepted by the lightning protection system
jacket compromising its integrity and the cable and diverted to earth. Typical locations for surge
lifetime [1]. arrestors within high voltage switchyards are line
The lightning protection systems intercepting entries, power transformers, and cable sealing ends.
lightning strokes exceeding the design current help The impact of lightning discharges on
rationalising the surge arrestors selection by underground cables has been studied extensively
diverting stroke currents to the substation earthing with respect to testing, open runs, and wind farms.
Gomes et al. [3] have used ATP modelling to study
the lightning discharge impact on cable sections
* Corresponding Author. Tel: +61-490181939 terminated to towers near substations with special
E-mail address: [email protected]

doi: https://dx.doi.org/10.14203/j.mev.2022.v13.189-200
2088-6985 / 2087-3379 ©2022 National Research and Innovation Agency
This is an open access article under the CC BY-NC-SA license (https://creativecommons.org/licenses/by-nc-sa/4.0/)
MEV is Scopus indexed Journal and accredited as Sinta 1 Journal (https://sinta.kemdikbud.go.id/journals/detail?id=814)
How to Cite: M. Nazih, “Effect of lightning mast placement on underground power cable jacket stress within high voltage substations,” Journal of
Mechatronics, Electrical Power, and Vehicular Technology, vol. 13, no. 2, pp. 189-200, Dec. 2022.
 190 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200

focus on the frequency dependency of soil and requirements on separate lightning electrodes.
towers earthing systems. It has used a lumped Goertz et al. [22] analysed the HVDC cables response
impedance representing the earthing grid instead of to lightning strokes and stress variation along cable
detailed modelling for the grid. Narajo-Villamil et al. length and the impact on insulation coordination
[4] studied lightning induced surges in underground with different station earthing resistances.
cables with focus on building reinforcements and Lennerhag et al. [23] developed a statistical method
internal earthing. Arshad et al. [5] listed a study on for estimating lightning overvoltages in HVDC cables
lightning effects on 132 kV underground cables and lines, considering shielding failure and
using ATP but without considering substation backflashover scenarios. Eriksson [24] researched
earthing systems. Cao et al. [6] studied the effect of the impact of lightning masts placement on low
lightning strikes on sheath voltage limiters in open voltage cable insulation within substations without
run applications. Tanaka et al. [7] document a considering the jacket breakdown or the effective
detailed study on the use of buried shield wires to zone concept and the effect of other design
improve the open run underground cable lightning parameters on attainable stresses.
performance considering the diversion of surges This paper studies cable jacket insulation
from the cable. Kangro et al. [8] detailed different impacted by discharges affecting high voltage
modelling approaches on cable sheath substation earthing systems, with a focus on
representation and the effect of lightning high lightning mast location and earthing system
frequency components on results. Wu et al. [9] modifications that were not covered in the available
presented a typical study on open run underground literature. The paper investigates the impact of the
cable metallic sheath induced voltages due to placement of lightning masts and down conductors
lightning discharges with focus on soil stratification within a substation on terminated underground
effects. Aniserowicz et al. [10] documented a cable jacket stress, taking into consideration the
comparative study between CDEGS modelling and effect of varied separation, electrode placement and
semi-analytical approaches regarding lightning length, soil resistivity, and lightning current.
voltage and current distributions within an
underground cable system. Chen et al. [11] II. Materials and Methods
demonstrated site measurements of transferred
surges through soil with rocket-triggered lightning A. Lightning discharge and modelling
strikes and associated response of surge arrestors. Lightning stokes show a probabilistic nature with
Jiao et al. [12] listed experimental results for coaxial current ranges from a few hundred amperes to about
cables lightning response and associated coupling 200 kA [2]. Several probability density functions had
using an impulsive generator setup. Güneri and been developed to match field measurements from
Alboyaci [13] analysed the accuracy and validity of various countries and territories with variable
different modelling techniques for high frequency median and exponent values [1][2]. The general
lightning response with underground cables return form of the density function is given by equation (1):
impedance. Shehab et al. [14] carried out a detailed
𝟏
lightning transient study of power transformers 𝐏(𝐈𝐩 ) = 𝐈𝐩 𝐱 (1)
within a thermal power plant where the effects of 𝟏+�
𝐈𝟓𝟓
�

lightning surges can be detrimental to equipment

where, P(lp) is probability that any peak return-
operation and cause excessive blackouts. Liu et al.
stroke in any given flash will exceed Ip (kA), I50 is
[15] proposed a new method to estimate the
50 % probability peak current, x is exponent, I is
lightning induced voltages on overhead lines and
single core underground cables open runs. Sekioka lightning peak current (kA). For this study, I50 is
[16] studied wind farm connected substation 24 kA and x is 2.6 are selected.
lightning overvoltage due to underground collector Hence, a typical stroke in any lightning flash is
cable sheath propagation of surges impending at expected to follow the pattern in Figure 1. The
wind turbines. Taha et al. [17] discussed the median indicates that 50 % of strokes crest current
performance of a lightning protection system for a will exceed this value at any time. The typical
110 kV substation within an urban area using CDEGS probability distribution shows that high stroke
software to estimate the touch and step voltages currents are less probable but not impossible. The
associated with lightning discharges into the probabilistic nature of the lightning phenomenon is
substation earthing grid. Sabiha et al. [18] studied captured in this distribution, where strokes can vary
the overvoltages and supply continuity associated along a very wide range of currents at the same
with backflow lightning surges propagating through location. Around any calculation period, typically a
lightning protection air terminations of photovoltaic few years, half of the strokes at a certain area will
installations. Aref and Anaraki [19] studied the exceed the median value where the other half will
propagation of lightning surges through a transition be equal to or less than the median value. Standard
between underground and overhead connections 1.2/50 μs current impulse is used to perform the
within sub-transmission substations. Alipio et al. studies.
[20] studied the lightning surge propagation in B. High voltage substation lightning protection
mixed overhead-cable lines taking soil frequency
dependency into consideration. Q. Liu et al. [21] Switchyards are typically protected against
investigated the lightning performance of weather lightning discharges by air terminations. This can be
stations and its effects on sheathed cables with in the form of shield wires and/or, preferably,
 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200 191

Figure 1. Probability of stroke current exceeding abscissa, 24 kA median value with 50 % exceedance probability is shown

lightning masts (LM), due to the risk associated with ground due to cost and ampacity requirements,
shield wires failure onto switchyard live conductors, jacket thickness is not perfectly uniform along the
as reported in [25]. The electrogeometric method cable due to possible abrasion during and after
(EGM) [2] utilises a rolling sphere with a striking installation. The compromised jacket thickness will
distance as its radius, depending on the striking result in a lower BIL due to a higher electric field
current. The sphere indicates the areas that a stroke within the jacket. Another detrimental factor that
is probable to hit and air terminations are provided affects the underground cable jacket is the presence
to intercept the strokes away from the protected of corrosive chemicals in the soil, which must be
equipment and conductors, as illustrated in Figure 2 verified during the installation planning/design stage
[26]. through soil chemical analysis. The latter is
especially true for installations within industrial
C. Cable jacket insulation
areas with contaminated soils.
Underground high and medium voltage cables
D. Earthning grid performance with lightning
are usually provided with a protective insulating
discharge
jacket to block moisture and protect the cable from
external damage. The jackets made of PVC or HD/LD The impact of lightning discharges on substation
PE are common. Jackets are not designed to earthing grids has been studied with several
withstand high voltage stresses since the cable parameters (grid dimensions, conductor spacing,
metallic sheaths are designed to have a limited impulse feed point location, soil resistivity and
potential with respect to soil outside cables [27]. ionization, wave front time and current amplitude)
Accordingly, the breakdown of the jacket can cause a [1]. The critical or effective length concept is
puncture pinhole, moisture migration and damage. introduced to identify the phenomenon of limited
Lightning damage to cables has been reported in [28]. lightning impulse propagation due to the dominant
Typical lightning impulse withstand voltages inductive effect at very high frequencies associated
(BIL) are provided in Table 2 of [27]. Given the cable with lightning discharges. The critical length concept
jacket insulation is non-self-restoring, peak recorded is extended into an effective zone with a radius re in
values of non-standard waveforms obtained at the metres given by equation (2) [1] for the centre feed
cable jacket shall be compared to jacket BIL [29]. point/lightning mast:
60 kV is a typical value to consider for jacket
𝑟𝑒 = 0.34𝜌0.42 𝑇 0.32 (2)
withstand voltage for HV cables. Although this value
is indicative, higher values may be verified through where ρ is soil resistivity in Ω.m and T is the front
testing. The jacket BIL is an important factor to avoid time in μs. The effective zone radius indicates the
breakdown and subsequent moisture damage. Since zone of conductors contributing to lightning
most HV underground cables are laid directly in the discharge dissipation. In other words, connected grid

Figure 2. Lightning masts around HV switchyard and rolling sphere concept [26]
 192 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200

Figure 3. Modelling flowchart

conductors and rods beyond this zone have (FFTSES) is then used to perform the forward
insignificant effect on grid lightning response, transform of the standard lightning impulse into
contrary to the power frequency behaviour where periodic components with different frequencies,
additional conductors are typically contributing to where the software calculates each frequency
reduce the grid resistance and attainable touch, step parameter separately. Inverse Fourier transform is
and transferred voltages. used to recompile the frequency-domain results into
time-domain results for attainable jacket stress
E. Methodology
variation with time. The process is repeated with the
The study uses state-of-the-art Current varied parameter as indicated in Table 1 for each
Distribution, Electromagnetic interference, case and the maximum jacket stress results
Grounding and Soil structure analysis (CDEGS) (magnitude and location along cable) are obtained
software engine. The process is summarised by the through the FFTSES and plotted against the variable
flowchart in Figure 3. The process starts with the parameter.
initial modelling of the earthing grid and cable using Three main cases are studied in this paper, with
CDEGS high frequency analysis module (HIFREQ) parameters tabulated in Table 1. Details of each case
software, as shown in Figure 4. The software engine are provided below.
is based on the field theory method and uses the
1) Case–I.1
method of moments (MoM) to solve Maxwell’s
equations in three dimensions. The software does The first case under study is for a typical HV
not consider soil ionization [30], a phenomenon substation earthing grid with overall dimensions of
encountered with lightning discharges exceeding 100 x 100 m and 10 m uniform spacing, as shown in
100 kA resulting in increases conductive area around Figure 4. The native soil resistivity is 100 Ω.m. One
affected conductors, especially with concentrated 15 m lightning mast of typical steel construction and
designs. CDEGS fast Fourier transform module a 1.5 m spike above it is considered for impulse feed.
 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200 193

a)

b)

Figure 4. 100 x 100 m, 10 m uniform spacing earthing grid, underground cable, and lightning mast CDEGS HIFREQ model: (a) 3D view; (b) Plan
view with LM possible locations shown in blue dots and separation distances in metres as marked. Cable chainage direction shown away from
grid

An underground cable terminated at the substation

3) Case–I.3
earthing grid and ran outside the substation grid for
1 km. Cable sheath is bonded to the substation The base design is modified considering a
earthing grid at the termination point (cable sealing practical spacing in outdoor high voltage substations
end). Lightning current is set at the median value of for the LM and cable sealing end (CSE) ≥ 4 m. The
24 kA. The mast location is moved away from the electrode depth is varied to identify the impact on
cable/sheath termination point at the earthing grid attainable jacket stress. The power frequency
and maximum jacket stress voltage is obtained earthing resistance is not affected by the increased
accordingly. The earthing grid power frequency electrode depth given the extended grid area.
resistance is calculated as 0.46 Ω.
4) Case–I.4
2) Case–I.2
The base design is modified considering a typical
The base design is modified with the addition of a 2.5 m deep electrode used to check the effect of
vertical rod next to LM. The power frequency moving the electrode away from the LM. The power
earthing resistance is not affected by the addition of frequency earthing resistance is not affected by the
a 2.5 m electrode at the lightning mast due to the increased electrode depth given the extended grid
extended grid area and uniform soil resistivity. area.

Table 1.
Case study parameters
Lightning Soil resistivity Mast separation Electrode depth Electrode Effective zone
Case
current (kA) (Ω.m) (m) (m) separation (m) radius (m)
I.1 24 100 1:78 0 0 2.5
I.2 24 100 1:78 2.5 0 2.5
I.3 24 100 4 0:30 0 2.5
I.4 24 100 4 2.5 0:15 2.5
II.1 24 500 1:78 0 0 4.9
II.2 24 500 1:78 2.5 0 4.9
III.1 120 100 1:78 0 0 2.5
III.2 120 100 1:78 2.5 0 2.5
 194 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200

120 80

Distance Along Cable (m)

Jacket Stress Voltage (kV) 70
100
60
80
50
60 40
Peak Jacket Stress (kV)
30
40
Max. Stress Chainage (m)
20
20
10
0 0
0 20 40 60 80
Mast Separation (m)

Figure 5. Peak jacket stress voltage and location variation vs. LM separation

5) Case–II.1
III. Results and Discussions
The soil resistivity is considered for this case as
500 Ω.m with a similar setup as Case – I. The earthing A. Results
grid power frequency resistance is calculated as 1) Case– I.1 results
2.31 Ω. Lightning current is set at the median value
of 24 kA. The attainable jacket stress voltage versus
variable separation distance from the lightning mast
6) Case–II.2 is shown in Figure 5. The attainable stress voltage
typically decreases with distance away from the LM
The design is modified with the addition of a
with little or no dependency on placement direction.
vertical rod next to LM. The power frequency
The location of the maximum stress varies with mast
earthing resistance is not affected by the addition of
a 2.5 m electrode at the lightning mast due to the separation and shows a tendency to move outside
extended grid area and uniform soil resistivity the substation grid with greater mast separation to
the cable termination point. The cable stress voltage
7) Case–III.1 suffers a sharp decline beyond 5 m (i.e., twice the
calculated effective zone radius). The behaviour
The soil resistivity is considered for this case as
follows the concept of effective length that beyond a
100 Ω.m with similar setup as Case – I but with the
certain distance along earthing conductor, the
injection lightning current of 120 kA peak
potential variation (and hence, resistance) is
(5 x 24 kA) representing a value of stroke current
insignificant regardless of the additional connected
with only 1.5 % exceedance probability (i.e., more conductors.
than 98.5 % of lightning strokes will be less than
120 kA). The earthing grid power frequency 2) Case– I.2 (modified design – addition of vertical
resistance remains as 0.46 Ω calculated in Case-I. rod next to LM) results

8) Case–III.2 The attainable maximum stress voltage and

location is shown in Figure 6. The attainable stress
The design is modified with the addition of a voltage typically decreases with distance away from
vertical rod next to LM. The power frequency the LM with little or no dependency on placement
earthing resistance is not affected by the addition of direction, similar to Figure 5. Electrode works to
a 2.5 m electrode at the lightning mast due to the dissipate the lightning current at the LM more
extended grid area and uniform soil resistivity. effectively away from the rest of the earthing grid.

100 80
Distance Along Cable (m)
Jacket Stress Voltage (kV)

70
80
60
60 50
40
40 30 Peak Jacket Stress (kV)
20 Max. Stress Chainage (m)
20
10
0 0
0 20 40 60 80
Mast Separation (m)

Figure 6. Peak jacket stress voltage and location variation vs. LM separation
 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200 195

25

20

Jacket Stress Voltage

Reduction (%)
15

10

0
0 10 20 30 40 50 60 70 80
-5
Mast Separation (m)

Figure 7. Peak jacket stress voltage reduction with/without electrode vs. LM separation (negative reduction % indicates stress increment with
electrode)

70 1.6

Distance Along Cable (m)

Jacket Stress Voltage (kV)

60 1.4
50 1.2
1
40
0.8
30 Peak Jacket Stree (kV)
0.6
20 0.4 Max. Stress Chainage (m)
10 0.2
0 0
0 5 10 15 20 25 30
Electrode Depth (m)

Figure 8. Peak jacket stress voltage and location variation with electrode depth

The effect is paramount with a separation distance of obtained at increased depth compared with an
less than 5 m (twice the calculated effective zone electrode at a depth equal to the critical length. The
radius) with a recorded maximum stress reduction behaviour is reversed after reaching a maximum
of about 20 % compared to the case with no value. No change to the locations of recorded
electrode. Spacings greater than 5 m showed slightly maximum stress is observed. The graph indicates
higher stress voltages compared with the no that a shorter electrode is more effective than a
electrode configuration (~2 %) then falls away deeper one for the same LM separation. While
(>15 m) to about 6 % average reduction. Figure 7 deeper electrodes are more useful in power
depicts the calculated reduction versus the mast frequency earthing design, it adds more inductance
separation. to the high frequency circuit impeding the lightning
discharge.
3) Case– I.3 (modified design – variable electrode
depth) results 4) Case– I.4 (modified design – variable electrode
location) results
The results are graphed in Figure 8. The jacket
stress increases with electrode depth The results are shown in Figure 9. The electrode
counterintuitively with more than 20 % rise separation from LM is critical with a sharp increase

60 1.6
Distance Along Cable (m)
Jacket Stress Voltage (kV)

58 1.4
1.2
56 1
54 0.8
Peak Jacket Stress (kV)
52 0.6
0.4 Max. Stress Chainage (m)
50 0.2
48 0
0 5 10 15
Electrode Spacing from LM (m)

Figure 9. Peak jacket stress voltage and location variation with electrode depth
 196 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200

350 200
180

Distance Along Cable (m)

Jacket Stress Voltage (kV)
300
160
250 140
200 120
100
150 80 Peak Jacket Stress (kV)

100 60 Max. Stress Chainage (m)

40
50
20
0 0
0 20 40 60 80
Mast Separation (m)

Figure 10. Peak jacket stress voltage and location variation vs. LM separation

in jacket stress voltages (18 %) with ~1 m separation. shown in Figure 11 with a pattern similar to Case-I.2.
The jacket stress is almost constant for any electrode Smoother maximum stress voltage chainage
separation of more than 1 m. No change to the variation is noticed compared with the 100 Ω.m case.
locations of recorded maximum stress is observed. Reduction of the attainable stress voltages nears
These results emphasize the importance of a close 25 % at LM separations about the effective zone
placement of electrode and LM for mitigating radius as shown in Figure 12. With the non-linear
underground jacket stress voltages. pattern, the reduction is negligible at about 50 m
(~10 times the effective zone radius) with masts far
5) Case– II.1 results
away from the cable sheath earthing point. The
The attainable peak jacket stress voltage with reduction again appears with a single peak value
variable separation distance from the lightning mast approaching 19 % with the mast at the corner
is shown in Figure 10. Like Case – I.1, the location of location. While most of the hazardous underground
the maximum stress varies with mast separation and cable jacket stresses occur with short mast
shows a tendency to move outside the substation separations (< twice the effective zone radius), the
grid with greater mast separation to the cable use of electrodes at lightning masts results in higher
termination point. The cable stress voltage suffers a reduction at these separations and is effective to
sharp decline beyond 10 m (i.e., twice the calculated control the attainable stresses.
effective distance). This indicates that soil resistivity Notwithstanding that the soil resistivity is five
affects the effective distance with a similar pattern times that used in Case-I.1 and I.2 (100 Ω.m) and
compared to the base case. Smoother maximum hence the power frequency resistance, the attainable
stress voltage chainage variation is noticed. While stress voltages are not elevated by the same ratio, as
the maximum cable stress shows an inverse pattern shown in Figure 13. The voltage rise is less than
with the mast separation distance, the chainage of 500 % and peaks at a separation slightly greater than
the maximum stress is much affected by the mast the effective zone radius. This indicates the non-
location rather than the separation. linear lightning response behaviour of the earthing
grid due to the pronounced induction effects of the
6) Case– II.2 (modified design – addition of vertical
high frequency component limiting the attainable
rod next to LM) results
voltages/stress. The maximum stress is about 82 %
It is evident that the addition of an electrode of the algebraic proportion, or in other words, a
reduces the attainable cable jacket stresses, as minimum of about 18 % reduction is obtained with

350 200
180
Distance Along Cable (m)
Jacket Stress Voltage (kV)

300
160
250 140
200 120
100
150 80 Peak Jacket Stress (kV)
100 60 Max. Stress Chainage (m)
40
50
20
0 0
0 20 40 60 80
Mast Separation (m)

Figure 11. Peak jacket stress voltage and location variation vs. LM separation
 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200 197

30
25

Jacket Stress Voltage

20

Reduction (%) 15
10
5
0
0 10 20 30 40 50 60 70 80
-5
Mast Separation (m)

Figure 12. Peak jacket stress voltage reduction with/without electrode vs. LM separation

normalised 500 Ω.m soil. At greater spacings, the a few metres away from the injection point
reduction is about half of the estimated algebraic regardless of the high stroke current conforming to
proportion value based on normalised soil resistivity. the effective zone concept due to the dominant
This pattern is also in line with the self-limiting inductive component with the high frequency
characteristics of lightning surge propagation within lightning discharge. Similar to Case-I and Case-II, the
earthing grids and the associated reduction in location of maximum stress varies with mast
underground cable jacket stress. separation, with a tendency to move outside the
substation the greater the mast separation to the
7) Case– III.1 results
cable termination point. The chainage of the
The attainable jacket stress voltage with variable maximum stress is much affected by mast location
separation distance from the lightning mast is rather than its separation. The self-limiting
shown in Figure 14. The majority of steep gradient behaviour is very useful in estimating the cable
appears near the LM (~effective zone radius) and the stress voltage with increased stroke currents where
rest of the separations return a very similar value to the criticality of mast placement is not affected by
the 24 kA injection case. The jacket stress “saturates” the high stroke current compared to separation.

450
400
Jacket Stress Voltage

350
Increment (%)

300
250
200
150
100
50
0
0 10 20 30 40 50 60 70 80
Mast Separation (m)

Figure 13. Peak jacket stress voltage increment with 500 Ω.m soil vs. LM separat ion

600 80
Distance Along Cable (m)
Jacket Stress Voltage (kV)

70
500
60
400
50
300 40
Peak Jacket Stress (kV)
30
200
20 Max. Stress Chainage (m)
100
10
0 0
0 20 40 60 80
Mast Separation (m)

Figure 14. Peak jacket stress voltage and location variation vs. LM separation
 198 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200

500 80

Distance Along Cable (m)

Jacket Stress Voltage (kV)
70
400
60
300 50
40
200 Peak Jacket Stress (kV)
30
20 Max. Stress Chainage (m)
100
10
0 0
0 20 40 60 80
Mast Separation (m)

Figure 15. Peak jacket stress voltage and location variation vs. LM separation

zone radius), as shown in Figure 17. Some locations

8) Case–III.2 (modified design – addition of vertical
(20 ~ 60 m separation) provide a reduction of the
rod next to LM) results attainable stress compared with the 24 kA injections.
The addition of a vertical electrode improves the This pattern demonstrates another non-linear
attainable jacket stress pattern similar to the one earthing grid lightning response behaviour limiting
observed in previous cases, as graphed in Figure 15 the attainable underground cable jacket
and Figure 16. As with other corresponding cases, voltages/stress. The maximum stress is about 62 %
with most of the hazardous jacket stresses taking of the algebraic proportion, or in other words, a
place with short mast separations, the use of an minimum of about 38 % reduction is obtained with
electrode at lightning masts is effective in mitigating a normalised 120 kA stroke. A 20 % less obtainable
underground cable jacket stresses at these jacket stress compared to the 24 kA case is
separations. A steep gradient of jacket stress voltage encountered around 1 x effective zone radius before
with mast separation is observed with the self- settling at a fractional increase (~10 % and less) at
limiting behaviour contributing to the improved longer spacings.
attainable voltages at an underground cable jacket
B. Significant Impact of LM Separation
with electrodes installed at LM. No change to the
locations of recorded maximum stress is observed The lightning response of earthing grids is a
with the electrode compared to the case without the complex phenomenon with soil and conductors’
electrode in Figure 14. Reduction of the attainable interactions [1]. The placement of lightning masts
stress voltages near 19 % at LM separations about within the substation affects the resultant voltage
the effective zone radius, as shown in Figure 16. stresses that appear on underground cables leaving
With the non-linear pattern, the reduction declines the substation. The provision of electrodes as close
sharply at about twice the effective zone radius as possible to the LM helps reduce the attainable
before dipping briefly and then sustaining a quasi- voltage at the cable jacket. The non-linear behaviour
steady pattern with masts far away from the cable of potential distributions with lightning discharge
sheath earthing point. current is evident with reduced jacket stress despite
Although the injection is five times the median higher injection currents. This is ascribed to the
current (24 kA), the attainable stress voltages are significant inductive effect with high frequency
almost identical to Case-I, excluding the placement components of lightning discharge and the
of LM near the cable termination point (~effective associated effective length phenomenon.

25

20
Jacket Stress Voltage

15
Reduction (%)

10

0
0 10 20 30 40 50 60 70 80
-5
Mast Separation (m)
Figure 16. Peak jacket stress voltage reduction with/without electrode vs. LM separation
 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200 199

400

Jacket Stress Voltage

300

Increment (%) 200

100

0
0 10 20 30 40 50 60 70 80
-100
Mast Separation (m)

Figure 17. Peak jacket stress voltage Increment with 120/24 kA vs. LM separation (negative increment % indicates stress reduction with 120 kA
lightning)

Since the effective length is much shorter than strength) and should be considered as a good
the installed earthing grid, soil resistivity is the practice for underground cables within substations,
single most important parameter in determining the especially in the vicinity of LMs.
lightning response of earthing grid. Along with
separation distance, the severity of the jacket stress Acknowledgements
on underground cables terminated at the grid can be
estimated. The author would like to express grateful
Cable jacket stresses exceeding the respective BIL gratitude to GHD Pty. Ltd for providing engineering
are likely to take place with almost 50 % of strokes software used in this study.
intercepted by LMs exceeding the median lightning
current distribution. The most critical zone is the Declarations
cable termination near the sealing end, where a
lightning mast placement is discouraged near a cable Author contribution
termination. The study finding suggests twice the M. Nazih is the main contributor of this paper. Author
effective zone radius as a buffer separation. read and approved the final paper.
The provision of non-metallic ducts/conduits
Funding statement
made of PVC/HDPE strengthens the jacket insulation.
For underground power cable routing within This research did not receive any specific grant from
substations, especially in the vicinity of LMs, it funding agencies in the public, commercial, or not-for-
should be considered as a good design practice by profit sectors.
typically adding about 60 kV or more to jacket Competing interest
dielectric strength depending on conduit
construction and wall thickness. The study did not The authors declare that they have no known
competing financial interests or personal relationships that
consider soil ionization; however, it is not expected
could have appeared to influence the work reported in this
to largely impact the results as the conductor paper.
diameter has a minimal effect on potential
distributions. Additional information
Reprints and permission: information is available at
IV. Conclusion https://mev.lipi.go.id/.
Publisher’s Note: National Research and Innovation
The introduction of lightning masts/air Agency (BRIN) remains neutral with regard to jurisdictional
terminations within substation earthing grids works claims in published maps and institutional affiliations.
as a current injection point. The location of the masts
affects the resultant underground cable jacket stress References
voltages with considerable impact taking place when
the cable sheath (or route) is terminated near the [1] J. He, R. Zeng, B. Zhang, Methodology and technology for
power system grounding. John Wiley & Sons, 2013.
lightning mast. The jacket stress is affected by soil [2] IEEE Guide for Direct Lightning Stroke Shielding of Substations
resistivity and mast separation. The greater the Sponsored by the Substations Committee IEEE Power and
separation of the lightning masts to the cable Energy Society, In IEEE Std 998-2012 (Revision of IEEE Std 998-
1996).
termination point or cable route, the lower the [3] T. V. Gomes, M. A. O. Schroeder, R. Alipio, A. C. S. de Lima, & A.
jacket stress voltages. Where this is unfeasible, the Piantini, “Investigation of overvoltages in HV underground
study finding suggests twice the effective zone sections caused by direct strokes considering the frequency-
radius should be used as a buffer separation between dependent characteristics of grounding,” IEEE Transactions on
Electromagnetic Compatibility, 60(6), 2002–2010, 2018.
the LM and cable (route/termination). The provision [4] S. Naranjo-Villamil, C. Guiffaut, J. Gazave, & A. Reineix, “On the
of relatively short (~effective zone radius) electrodes calculation of electrical surges in underground cables due to a
as close as possible to the LM effectively helps direct lightning strike,” Joint IEEE International Symposium on
reduce the attainable stress voltages. Non-metallic Electromagnetic Compatibility Signal and Power Integrity, and
EMC Europe, EMC/SI/PI/EMC Europe 2021, 682–687, 2021.
ducts/conduits strengthen the jacket insulation
(typically adding 60 kV or more to jacket dielectric
 200 M. Nazih / Journal of Mechatronics, Electrical Power, and Vehicular Technology 13 (2022) 189-200

[5] S. N. M. Arshad, A. Z. Abdullah, N. S. Tajudin, C. L Wooi, N. H. International Symposium on Lightning Protection, SIPDA 2021,
Halim, & N. Hussin, “Simulation on lightning effects to 132 kV 2021.
underground cable by using alternative transients program / [17] M. A. Taha, Li, L., & Wang, P., “Estimation performance of the
electromagnetic transients program (ATP/EMTP),” 2018 IEEE lightning protection system in an urban 110 kV grounding grid
7th International Conference on Power and Energy (PECon), pp. substation,” Results in Engineering, 6, 2020.
209-214, 2018. [18] N. A. Sabiha, M. Alsharef, M. K. Metwaly, E. E. Elattar, I. B. M.
[6] D. Cao, X. Liu, and X. Deng, “The suitability analyses of sheath Taha, & A. M. Abd-Elhady, ”Sustaining electrification service
voltage limiters for HV power cable transmission lines,” 2019 from photovoltaic power plants during backflow lightning
2nd International Conference on Electrical Materials and overvoltages,” Electric Power Systems Research, 186, 2020.
Power Equipment (ICEMPE), pp. 404-408, 2019. [19] S. Aref, & A. S. Anaraki, “An investigation of the factors
[7] H. Tanaka, Y. Baba, C. F. Barbosa, T. Tsuboi, & S. Okabe, affecting the transition of lightning surges through
“Protective effect of shield wires against direct lightning underground cables and sub-transmission substations: a case
flashes to buried cables,” IEEE Transactions on Power Delivery, study,” arXiv: Physics and Society, 2019.
33(4), 1628–1635, 2018. [20] R. Alipio, Xue, H., & A. Ametani, “An accurate analysis of
[8] T. Kangro, J. Kilter, & J. de Silva, “Comparative analysis of HVAC lightning overvoltages in mixed overhead-cable lines,” Electric
cable metallic sheath modelling approaches for network Power Systems Research, 194, 2021.
studies,” Proceedings of International Conference on [21] Liu, Q., Chang, M., & Wang, L., “Study on protection measures
Harmonics and Quality of Power, ICHQP, 2018-May, 1-6, 2018. of lightning electromagnetic pulse in automatic weather
[9] Wu, Y., Liu, Q., Yang, J. Y., & Liu, X, “Calculation of lightning station,” Journal of Physics: Conference Series, 2254(1), 2022.
induced voltage of buried cable metal sheath,” 13th [22] M. Goertz, S. Wenig, C. Hirsching, M. Kahl, M. Suriyah, & T.
International Symposium on Antennas, Propagation and EM Leibfried, “Analysis of extruded HVDC cable systems exposed
Theory, ISAPE 2021 – Proceedings, 2021. to lightning strokes,” IEEE Transactions on Power Delivery,
[10] K. Aniserowicz, & R. Markowska, “Analytical and numerical 33(6), 3009–3018, 2018.
calculations of overvoltages in underground cable of intrusion [23] O. Lennerhag, J. Lundquist, C. Engelbrecht, T. Karmokar, & M. H.
detection system directly hit by lightning,” 34th International J. Bollen, “An improved statistical method for calculating
Conference on Lightning Protection, ICLP 2018, 2018. lightning overvoltages in HVDC overhead line/cable systems,”
[11] Chen, S., Zhang, Y., Zhou, M., Yan, X., Lyu, W., & Zheng, D, Energies, 12(16), 2019.
“Transient response of surge protective devices during the [24] A. Eriksson, “Induced effects on cables laid in the vicinity of
potentials transferred between independent grounding grids,” substation lightning masts,” Master thesis, Department of
IEEE Transactions on Power Delivery, 35(2), 630-638, 2020. Electrical Engineering, Chalmers University of Technology,
[12] Jiao, X., Zhang, Q., Xing, H., & Zhou, J., “Coupling characteristics Gothenburg, 2020. [online].
of lightning impulsive current on the underground coaxial [25] Inclair Knight Merz, “Incident review – Otahuhu substation
cables,” IOP Conference Series: Earth and Environmental loss of supply 12th June 2006”, 23 June 2006.
Science, 440(3), 2020. [26] Primtech 3D, “Lightning Protection Calculation,” [online]
[13] M. Güneri, & B. Alboyaci, “Analysis of ground return [Accessed: 18-April-2022].
impedance calculation methods for modeling of underground [27] IEEE guide for bonding shields and sheaths of single-conductor
cables for lightning studies,” Turkish Journal of Electrical power cables rated 5 kV through 500 kV sponsored by the
Engineering and Computer Sciences, 26(1), 530–541, 2018. Insulated Conductors Committee IEEE Power and Energy
[14] W. F. A. Shehab, S. A. Ali, & M. I. Alsharari, “Lightning Society, in IEEE Std 575-2014 (Revision of IEEE Std 575-1988),
protection for power transformers of Aqaba thermal power 2014.
station,” Archives of Electrical Engineering, 69(3), 645–660, [28] S. Sekioka, H. Otoguro, & T. Funabashi, “A Study on
2020. overvoltages in windfarm caused by direct lightning stroke,”
[15] Liu, X., Zhang, M., Wang, T., & Ge, Y., “Fast evaluation of IEEE Transactions on Power Delivery, 34(2), 671–679, 2019.
lightning-induced voltages of overhead line and buried cable [29] A. R. Hileman, Insulation Coordination for Power Systems. 1st
considering the lossy ground,” IET Science, Measurement and ed., CRC Press, 1999.
Technology, 13(1), 67–73, 2019. [30] CDEGS knowledge base FAQ 329, “Effects of Soil Ionization,”
[16] S. Sekioka, “Simulation of lightning overvoltages in substation 2004. [online] [Accessed: 20-April-2022].
for lightning strike to wind turbine,” 2021 35th International
Conference on Lightning Protection, ICLP 2021 and 16th