2011 7th Asia-Pacific International Conference on Lightning, November 1-4, 2011, Chengdu, China

Application of a Partial Element Equivalent Circuit

Method to Lightning Surge Analyses
Peerawut Yutthagowith 1, Akihiro Ametani2, Naoto Nagaoka2, Yoshihiro Baba2
1
Electrical Engineering Department, Faculty of Engineering, King Mongkut’s Institute of Technology Ladkrabang
Chalongkrung Road, Ladkrabang, Bangkok, Thailand, 10520
2
Doshisha University, Kyo-tanabe, Kyoto 610-0321, Japan
1
[email protected]
2
[email protected]

Abstract— This paper presents the application of the partial lightning studies [15]-[17]. The calculated results by the full-
element equivalent circuit (PEEC) method in the time and wave PEEC method have shown good agreement in
frequency domains for calculating voltages across insulators on comparison with the results calculated by the MoM and the
an actual transmission tower and ground potential rise on FDTD methods, as well as with experimental data.
grounding electrodes. An effective way to increase the efficiency
For the formulation of the full-wave PEEC method in the
of the method in terms of computation time, which consists of the
appropriate combination of the PEEC method and the frequency domain time responses are obtained by using the
transmission line theory, is proposed. The transmission line inverse Fourier's transform. By employing the modified
model is adapted to combine with the method for reducing inverse fast Fourier transform (MIFFT) algorithm [19], the
elements in the PEEC method. Comparisons of the simulation results in the time domain are released from Gibb’s oscillation
results by the proposed methods in the time and frequency in their early time responses and from truncation errors in
domains with other numerical methods and with available their late time responses.
experimental data show satisfactory agreement. A full-wave time-domain PEEC formulation can therefore
be interesting, especially when dealing with nonlinearities.
I. INTRODUCTION
However, special care should be taken to avoid numerical
Predictive calculations of over-voltages due to lightning in instability problems. Voltages and currents along the structure
the transmission and distribution systems is most essential for should be memorized at every time step. To consider the
an economical insulation design of the transmission and presence of conducting ground, the image method can be
distribution systems. Traditionally, circuit-based models and employed but the solution matrix size is increased due to
transmission line approaches assuming transverse retardation effects. Therefore, the computational efficiency is
electromagnetic (TEM) propagation mode were employed for not always higher than that of a frequency domain PEEC
the calculation of over-voltages due to a lightning strike to a implementation.
transmission system [1]-[5]. The assumption of the TEM Neglecting retardation effects will increase the
propagation mode might not be correct for a tall tower in computational time efficiency of the PEEC method in the
EHV and UHV transmission systems, because of the non- time domain because the past time of voltages and currents
uniform behaviour of the tower and the grounding system, are memorized only at one previous time step. In addition, the
and reflections and scattering from the discontinuities and model of reduction (MOR) [20], [21] can be employed to
tower ends[6]-[8]. This behaviour cannot be expressed by a reduce the equation size. However, in a typical configuration
time-independent model unless the rise time of the current is involving a few thousands of elements and a few hundred
much longer than the round-trip time of the travelling wave. calculation points, over 80 percent of the computation time is
To obtain more accurate results, full-wave approaches such spent for the formulation of the impedance matrix and
as the method of moment (MoM) [6]-[11], the finite- inversion. Therefore, the MOR is not an effective way to
difference dime-domain (FDTD) technique [12]-[14], or a increase time efficiency of the method in this case.
hybrid electromagnetic-circuit method such as the PEEC As known [5], the transmission line approach assuming
method [15]-[17] are more appropriate than a circuit and TEM mode propagation can be successfully applied to
transmission line approach. systems of horizontal conductors. In the PEEC method, circuit
The PEEC method is derived from a mixed potential elements and also transmission lines can be implemented
integral equation (MPIE) and provides a full wave solution to readily. In the case of a lightning surge analysis in electrical
Maxwell’s equations. The method can be applied to both time power systems, most of elements (overhead or buried
and frequency domains. A detailed description of the conductors, grounding grid) can be represented as horizontal
approach can be found in [18]. The authors have developed a wires. Therefore, a combination of the transmission line
full-wave PEEC method based on the thin wire structure in theory and the PEEC method would be an effective way to
the frequency domain which has been successfully applied to increase the efficiency of the method in terms of computation

978-1-4577-1466-5/11/$26.00 ©2011 IEEE 809

 2011

5
time and memory requirement. In this paper, the full-wave 4
PEEC method and the PEEC method neglecting the

Current [A]
3
retardation effect in the time domain combined with the 2

transmission line theory is proposed for lightning surge 1

0
analysis.
-1
0 0.5 1.0 1.5 2.0 2.5 3 3.5 4.0

II. VERIFICATION OF THE PEEC METHOD AND EXAMPLE Time [Ps]

(b) Fast rise time current
Experimental results of transients on a transmission tower 5
and along grounding electrodes are presented in this Section 4
for comparison with simulated results by the PEEC method in

Voltage [V]
3

the time domain, by the full-wave PEEC method in the 2

1
frequency domain, and collected experimental data [1]. 0
(a) System configuration -1
A. Transient Performance of Transmission Towers 0 0.5 1.0 1.5 2.0 2.5 3 3.5 4.0
Time [Ps]
An experimental result of an actual transmission tower [1] (c) Slow rise time current
is used in this sub-Section for comparison with the simulated Fig. 1. System configuration and applied currents.
results by the PEEC methods.
An actual transmission tower was selected as a test case. 200
200
The PEEC methods are applied to determine the tower surge Voltage at cross arm
150
150
response of a 500-kV double-circuit transmission line and

[V]
voltages across insulators and are compared with
Voltage [V]
100
100 Insulator voltage
Voltage
experimental results reported by Ishii et al. [1]. The towers
composed of slant elements and cross arms including 50
overhead ground wires and phase wires are considered in the
0 Voltage on phase wire
simulation. The configuration of the experiment on an actual
system of a 500 kV double circuit transmission line is -50
-50
0 0.5 1.0 1.5 2.0 2.5 3 3.5 4.0
illustrated in Fig. 1(a). The simulated system is composed of
Time [Ps]
three towers with separation distances of 450 m and 560 m
(a) Upper phase
between towers, two overhead ground wires, and six phase
wires. A current pulse is applied to the top of the middle 200
200

tower. To measure the insulator voltages, two current pulses 150

150
Voltage at cross arm
characterized by different rise times were applied. The first
[V]
Voltage [V]

pulse (Fig. 1b) has about 0.2 Ps rise time (fast rise time 100
100
Insulator voltage
Voltage

current) and the second (Fig. 1c) has a time to crest of 3 Ps.
50
50
The peak amplitude of the two pulses is 3.4 A. To measure
voltages across insulator strings 10-k: resistive voltage 00
dividers were employed. Voltage on phase wire
The full-wave PEEC simulation in the frequency domain -50
-50
00 0.5
0.5 1.0
1.0 1.5
1.5 2.0
2.0 2.5 3 3.5 4.0
involves 256 frequencies upto 5 MHz and the element length Time [Ps]
is 5 m. The tower is composed of four main poles of which (b) Middle phase
elements have 0.2 m radius; slant and horizontal elements
200
200
have 0.1 m radius. The cross arms are composed of 0.2-m
radius elements. The tower-footing resistance is represented Voltage at cross arm
150
150
by a 17-: resistance by connecting four 68 : resistors at the Insulator voltage
Voltage [V]
[V]

bottom of four main poles of the tower. The same conditions 100
100
Voltage

are employed in the PEEC simulation in the time domain with

50
50
a 16 ns time step. The overhead conductors are modelled
using a lossless 8-conductor transmission line (6 phase wires 00
Voltage on phase wire
and 2 ground wires). The towers and parts of the transmission
lines close to the tower are modelled by the PEEC method for -50
-50
00 0.5
0.5 1.0
1.0 1.5
1.5 2.0
2.0 2.5
2.5 33 3.5
3.5 4.0
4.0
taking into account the retardation of electromagnetic fields Time [Ps]
around the tower by self and mutual impedances of the (c) Lower phase
elements. The total number of elements without the use of the
PEEC with a TL model PEEC without a TL model
transmission line model is 2488 while the total number of Full-wave PEEC Measurement from [1]
elements when using the transmission line model is reduced
to 604. Fig. 2. Comparison between simulated and measured waveforms, fast rise
time current injected.

978-1-4577-1466-5/11/$26.00 ©2011 IEEE 810

 2011

120
120 conductors is very effective to increase the computational
100
100
Voltage at cross arm efficiency of the method without losing accuracy.
80
80
Insulator voltage
Table 1
[V]
Voltage [V]

60
60 Comparison of Computation Time Relative to the
Voltage

40
40 Computation Time of the Full-wave PEEC Method
20
20
Formulation of Calculation of
Voltage on phase wire
00 PEEC in the an impedance result by Total
time domain matrix and multiplication (p.u.)
-20
-20
00 0.5
0.5 1.0 1.5
1.5 2.0 2.5 3 3.5 4.0 inversion (p.u.) (p.u.)
Time [Ps] without adopting
(a) Upper phase transmission line 5.56u10-3 5.00u10-4 6.06 u10-3
model
120 with adopting
100 transmission line 3.33 u10-4 6.66u10-5 4.00u10-4
Voltage at cross arm model
80
Insulator voltage
[V]
Voltage [V]

60
Voltage

40 B. Ground Potential Rise of Grounding Electrodes

20 In this sub-section experiments on simple arrangement of
0 horizontal and vertical electrodes carried out by Rochereau
Voltage on phase wire
[22] are selected for a comparison with the simulated results
-20
0 0.5 1.0 1.5
1.5 2.0
2.0 2.5
2.5 33 3.5
3.5 4.0
4.0 by the PEEC methods.
Time [Ps] The configuration and parameters of two experiments on
(b) Middle phase grounding electrodes carried out by H. Rochereau [22] are
120
120 illustrated in Fig. 4. In the first experiment, an impulse current
100
100
was applied to a vertical electrode. The voltages were
80
recorded at the injection point of the electrode. In the second
80
Voltage at cross arm experiment, an impulse current was applied to a horizontal
[V]
Voltage [V]

60
electrode. The voltages were recorded at the injection point of
Voltage

40 the electrode, and at distances of 3.5 m and 7 m from the

20 Insulator voltage injection point of the electrode, respectively. The injected
00
current amplitudes being relatively low, soil ionization effect
Voltage on phase wire was unlike to appear and was not taken into account in the
-20
-20
00 0.5
0.5 1.0
1.0 1.5
1.5 2.0
2.0 2.5
2.5 33 3.5
3.5 4.0
4.0 computation.
Time [Ps] The same conditions were simulated by the PEEC methods
(c) Lower phase in the time domain with a 10 ns time step and the full-wave
PEEC method in the frequency range from 19.53 kHz to 5
PEEC with a TL model PEEC without a TL model
Full-wave PEEC Measurement from [1] MHz with a frequency step 19.53 kHz and frequency range
from 39.06 kHz to 10 MHz with a frequency step 39.06 kHz
Fig. 3. Comparison between simulated and measured waveforms, slow rise for the first and the second cases, respectively.
time current injected.

Figs. 2 and 3 show a comparison between measured and

simulated waveforms. Experimental waveforms were taken
from [1].
Very good agreement of the results calculated by the full-
wave PEEC method and experimental results are observed,
and a satisfactory agreement of the PEEC method neglecting Fig. 4. Configuration of electrodes in the experiments [25].
retardation effects in the time domain with and without
Fig. 5 shows comparison results between experimental
adopting the transmission line model are also observed.
results and those calculated by the PEEC method in time
Table I shows computation times relative to the time
domain and the full-wave PEEC method in the frequency
corresponding to the full-wave PEEC method in the frequency
domain. Good agreements of the results calculated by the full-
domain. From the results in Table I, it can be seen that the
wave PEEC method including and neglecting electromagnetic
computation efficiency of the PEEC method in the time
retardation and experimental results are observed
domain is much better than that of the PEEC method in the
frequency domain. Furthermore, a combination with the
transmission line theory for the representation of overhead

978-1-4577-1466-5/11/$26.00 ©2011 IEEE 811

 2011

REFERENCES
[1] M. Ishii, T. Kawamura, T. Kouno, E. Ohasaki, K. Murotani and T.
600
Injected current Higushi, "Multistory transmission tower model for lishtning surge
500 analysis," IEEE Trans. Power Del., vol. 6, no. 3, pp. 1327-1335, July
20 A 1991.
Voltage [V]

400
[2] N. Nagaoka, "A development of frequency-dependent tower model,"
300 Trans IEE Japan, vol. 111-B, no. 1, pp.51-56, 1991 (in Japanese).
[3] A. Ametani, Y. Kasai, J. Sawada, A. Mochizuki and T. Yamada,
200 "Frequency-dependent impedance and a multiconductor tower model,"
100 IEE Proc.-GTD., vol. 141, no. 4, pp. 339-345, July 1994.
[4] M. T. Correia de Barros and M.E. Almeida, "Computation of
0 electromagnetic transients on nonuniform transmission lines," IEEE
0 0.5 1 1.5 2 2.5 3
Time [Ps] Trans. Power Del., vol. 11, no. 2, pp. 1082-1091, April 1996.
[5] A. Ametani and T. Kawamura, "A method of a lightning surge analysis
(a) Vertical electrode recommended in Japan using EMTP," IEEE Trans. Power Del., vol. 20,
no. 2, pp. 867-875, April 2005.
1000 [6] Y. Baba and M.Ishii, "Numerical Electromagnetic Field analysis of
Injected current tower surge response," IEEE Trans. Power Del., vol. 12, no. 1, pp.
800 483-488, Jan. 1997.
[7] Y. Baba and M.Ishii, "Numerical Electromagnetic Field analysis on
Voltage [V]

V (x= 0 m) 10 A
600 measuring methods of tower surge impedance," IEEE Trans. Power
V (x= 3.5 m) Del., vol. 12, no. 1, pp. 483-488, Oct. 1999.
400 [8] Y. Baba and M.Ishii, "Numerical Electromagnetic Field analysis on
lightning surge response of tower with shield wire," IEEE Trans.
200
Power Del., vol. 15, no. 3, pp. 1010-1015, July 2000.
V (x= 7 m) [9] S. Kato, A. Mochizuki, T. Narita, and E. Zaima, “A development of a
0
0 100 200 300 400 500 600 700 lightning surge simulation program employing TWTDA” (in Japanese),


Time [ns] in Proc. Conf. Switching and Protecting Engineering, IEE of Japan, pp.
(b) Horizontal electrode 149–158, Nov. 1994.
[10] R. K. Pokharel and M. Ishii, “Time-domain numerical electromagnetic
PEEC Full-wave PEEC Measurement from [22] simulation of tower surge response,” in Proc. Int. Workshop on High
Fig. 5. Experimental and simulated results of the transient voltages of the Voltage Engineering, vol. II, Tottori, Japan, pp. 77–82, Nov. 2000.
vertical electrode and the horizontal electrode. The injected current is also [11] T. Mozumi, Y. Baba, M.Ishii, N. Nagaoka, and A. Ametani,
shown. "Numerical Electromagnetic Field analysis of archorn voltages during
a back-flashover on 500 –kV twin-circuit line," IEEE Trans. Power
III. CONCLUSIONS Del., vol. 18, no. 1, pp. 207-213, Jan. 2003.
[12] T. Noda and S. Yokoyama, "Thin wire representation in Finite
This paper presents the application of the partial element DifferenceTime Domain surge simulation," IEEE Trans. Power Del.,
equivalent circuit (PEEC) method in the time and frequency vol. 17, no. 3, pp.840-847, July 2002.
domains for the analysis of lightning surge analysis. [13] T. Noda, "A tower model for lightning overvoltage studdies on the
The frequency- and time-domain PEEC methods were results of an FDTD simulation," Electrical Engineering in Japan., vol.
164, no. 1, pp.8-17, 2008.
applied to examples related to lightning effects on power [14] T. Noda, "A numerical simulation of transient electromagnetic fields
systems, namely, transient performance of transmission for obtaining the step response of a transmission tower using the
towers and ground potential rise on grounding electrodes. The FDTD method," IEEE Trans. Power Del., vol. 23, no. 2, pp.1262-1263,
obtained results have been compared with other numerical April 2008.
[15] P. Yutthagowith, A. Ametani, N. Nagaoka, and Y. Baba, "Influence of
methods and available experimental data and a very good a current lead wire and a voltage reference wire to a transient voltage
agreement has been found. on a vertical Conductor," IEEJ. TEEE., vol. 5, no. 1, pp.1-7, Jan. 2010.
We have shown that for applications involving long [16] P. Yutthagowith, A. Ametani, N. Nagaoka and Y. Baba., " Influence of
horizontal structures such as overhead power lines, a a measuring system to a transient voltage on a vertical Conductor,"
IEEJ. TEEE., vol. 5, no. 2, Mar. 2010.
combination of the PEEC method with the transmission line [17] P. Yutthagowith, A. Ametani, N. Nagaoka, and Y. Baba., "Lightning
theory is computationally very efficient while keeping the induced voltage over lossy ground by a hybrid electromagnetic-circuit
accuracy of a full PEEC approach. model method with Cooray-Rubinstein formula," IEEE Trans.
Electromagn. Compat., vol. 51, no. 4, pp.975-985, Nov. 2009.
It has also been shown that retardation effects can be
[18] A. E. Ruehli., "Equivalent circuit models for three-dimentional
neglected when the minimum wavelength of the applied multiconductor systems," IEEE Trans. Microw. Theory Tech., vol.
source is much larger than the size of the considered system. MTT-22, no. 3, pp.216-221, Mar. 1974.
Neglecting retardation give an overestimate result in [19] A. Ametani, "Application of the fast fourier transform to electrical
comparison with a results calculated by the full-wave PEEC transient phenomena," JEEE, vol. 4, no. 2, pp.277-287, 1972.
[20] A. Odabasioglu, M. Celik, and L. Pileggi, “Prima: Passive reduced-
method. Therefore, the results calculated by the proposed order interconnect macro modeling algorithm,” IEEE Trans. Comput.-
method are applicable for a conservative estimate a lightning Aided Des. Integr. Circuits Syst., vol.17, no.8, pp.645–654, Aug.1998.
surge. [21] S. Wang, J. He, B. Zhang, Rong Zeng and Z. Yu., "A time-domain
The analysis and the results presented in this paper show multiport of thin-wire system for lightning transient Simulation," IEEE
Trans. Electromagn. Compat., vol. 51, no. 4, pp.975-985, Nov. 2009.
that frequency- and time-domain PEEC method can be [22] H. Rochereau, “Response of earth electrodes when fast fronted
successfully applied to lightning overvoltage studies. currents are flowing out,” EDF Buletin de la Direction des Etudes et
Recherches., serie B, no. 2, pp. 13-22, 1998.

978-1-4577-1466-5/11/$26.00 ©2011 IEEE 812