하동혁
(Dong-Hyuk Ha)
1
최종기
(Jong-Kee Choi)
1
윤용범
(Yong-Beum Yoon)
2†
-
(Korea Electric Power Corporation (KEPCO))
-
(KEPCO International Nuclear Graduate School, Korea)
Copyright © The Korean Institute of Illuminating and Electrical Engineers(KIIEE)
Key words
Lightning Induced Voltage, ATP(Alternative Transient Program), Parametric Analysis
1. Introduction
Lightning strikes near a distribution line will induce voltages into the line from
the electric and magnetic fields produced by the lightning stroke. These induced voltages
are much less severe than direct strikes, but close strikes can induce enough voltage
to flash insulation and damage poorly protected equipment. The charge and current
flow through the lightning channel creates fields near the line. These fields induce
voltages on the line. The vertical electric field is the major component that couples
voltages into the line. As the highly charged leader approaches the ground, the electric
field increases greatly; and when the leader connects, the electric field collapses
very quickly. The rapidly changing vertical electric field induces a voltage on a
conductor, which is proportional to the height of the conductor above ground.
Most measurements of induced voltages have been less than 300kV, so the most common
guideline for eliminating problems with induced voltages is to make sure that the
line insulation capability (CFO) is higher than 300kV[1]. Therefore, lines with insulation capabilities less than 300kV have many more flashovers
due to induced voltages.
Fig. 1. Direct and indirect lightning strikes on distribution lines
There are methods to mitigate lightning overvoltage and its consequences, such as
back-flashover. In case of back flashover, reducing ground resistance of supporting
structure (i.e. tower, pole) is one way to limit excessive potential rise of supporting
structures,. which cause back-flashover, by lightning current discharge to ground.
In case of induced over-voltage, however, it is unclear that improving ground (ex.
reducing ground resistance) is really necessary to mitigate induced over-voltage.
For reader’s information, in EPRI’s handbook for lightning mitigation[1], improving ground is not a mitigation option for reducing induced over-voltage (refer
Table 1). But no further specific descriptions about why improving grounding is not included
in migitaion options are not given in the report.
Table 1. Typical power line lightning mitigation options[1]
In addition, as for overhead ground wires of distribution lines, a lot of KEPCO engineers
has believed that the purpose of overhead ground wires is to protect lines from direct
lightning. However, IEC 60071, which is an international standard of insulation coordination,
states in the related sections (i.e. 6. Special considerations for overhead lines,
6.4.1 distribution lines) that "Protection by shield wires is useless because tower
earthing and insulation strength cannot economically be improved to such a degree
that back flashovers are avoided."[2] Therefore, the lightning performance of distribution lines, therefore, is largely
determined by the ground flash density and the line height.
In this paper, a lot of numerical simulations using EMTP were conducted to understand
how much ground improvement, overhead ground wires and surge arresters contribute
to mitigation of lightning induced over-voltages. Firstly, compiled LIV-ATP process
briefly explained and the execution times are compared with the existing LIV-ATP.
And the paper also provide parametric analysis results on the induced voltage with
compiled LIV-ATP, for lighting arrester and ground resistance.
2. EMTP model for LIV calculation
ATP (Alternative Transient Program) version of EMTP (Electro-Magnetic Transient Program)
has been very widely used for lightning overvoltage analysis and insulation coordination
under lightning strikes. Since analysis of lightning induced overvoltage requires
rather rigorous electromagnetic field calculation comparing to circuit analysis, however,
original ATP code could not be used for lightning induced voltage analysis. To overcome
this limitation, LIV (Lightning Induced Voltage) computation model using MODELS, which
is an internal programming language for ATP, has been proposed recently.
2.1 LIV-ATP
The LIV-ATP program by Høidalen is proposed and described in [2][2]. It is based on the analytical solution of the lightning-electromagnetic field equations
for a lossless ground using the approach proposed in [3][3] by Rusck.
2.2 Compiled LIV-ATP
Software package MinGW32 includes appropriate tools to link of the object file generated
from c language version of LIV-ATP MODELS code with ATP object file (tpbig.a). By
gcc.exe, g77.exe and make.exe programs the fortran compiler will create the object
file from FORTRAN ATP written in FORTRAN language and C compiler will create the object
file from the relay source code written in C language.
These will be subsequently linked with other object files and libraries and compiled,
which will create a new simulation ATP program in file tpbig.exe containing a LIV-ATP.
The compile and link process is clearly described in Figure. 2.
Fig. 2. Linking and compiling the process of ATP in MinGW32[3]
Figure. 3(a) shows comparison of original and compiled version of LIV-ATP example, which has two
conductors and 500m of line length. The point of lightning strike to ground is 100m
horizontally away from the line. Figure. 3(b) shows that calculated induced voltages at both terminal is consistent.
Fig. 3. Comparison of calculation results of original and compiled LIV-ATP examples
Elapsed times for running ATPDraw models were compared and the results were shown
in Figure. 4. Since two conductos are possible to be modeled in the original LIV-ATP MODELS, case-1
results were selected and compared in regards to computational elapsed time. It can
be seen that computational speed is significantly improved by using compiled MODELS
version of LIV-ATP.
Fig. 4. Comparison of calculation time between original & compiled LIV-ATP
3. LIV calculation in distribution lines
In this chapter, EMTP models for lightning induced voltage on 22.9kV distribution
lines were described. Using the models, performances of various distribution line
protection schemes under indirect lightning strikes were evaluated.
3.1 ATPDraw Models
Figure. 5 shows examples of distribution line model with various lightning protection schemes
using ATPDraw, which is a public domain graphic pre-processor for ATP. The line length
is assumed to be 2km and lightning strikes at the center of the line with horizontal
distance of 100 meters.
Fig. 5. ATPDraw models for LIV study on 22.9kV distribution lines
The following 4 types of protection schemes were implemented to evaluate each model’s
lighting protection performance.
- Case-1 : Neither overhead groundwires nor Surge arresters. Neutral wires are grounded
at every 500 meters according to related KEPCO standard
- Case-2 : Overhead groundwires (No arrester)
- Case-3 : Surge arresters (No overhead groundwire)
- Case-4 : Both overhead groundwires and Surge arresters are installed
3.2 Lightning Performance Evaluation results
Figure. 6~Figure. 8 show the calculated voltage difference between phase and neutral conductor, which
is a voltage along line post insulator. If this voltage difference exceeds insulator’s
insulation strength, then flashover between conduct and neutral conductor will occur
causing line-to-ground fault. The magnitude of lightning current is assumed to be
100kA or 20kA.
1) 20kA of lightning current magnitude : Lightning falls on ground surface horizontally
distant from #20 pole, which is a center pole of 2km distribution line.
· Figure. 6(a) shows voltages along LP insulators at #20, #25, #30 and #35 poles (Case-1). The closest
#20 pole from lighting strike point shows the maximum voltage of about 50kV.
· Figure. 6(b) shows the case of overhead ground wires installed (no surge arresters, Case-2). By
shielding effect of overhead ground wires, the maximum insulator voltage is reduced
to about 40kV.
· Figure. 6(c) shows the case of surge arresters (no overhead ground wires, Case-3). From the voltage
waveforms, some surge arresters operated and insulator voltages were reduced consequently.
· Figure. 6(d) shows the case of Case-4 with overhead ground wires and surge arresters installed.
From the voltage waveforms, we can see no surge arrester was operated.
Fig. 6. Differential voltage between phase and neutral conductors under different
line protection conditions (20kA)
2) Figure. 7 show the cases of 40kA lightning strikes.
Fig. 7. Differential voltage between phase and neutral conductors under different
line protection conditions (40kA)
3) 100kA of lightning current magnitude : Lightning falls on ground surface horizontally
distant from #20 pole, which is a center pole of 2km distribution line.
· Figure. 8(a) shows voltages along LP insulators at #20, #25, #30 and #35 poles (Case-1). From
the voltage waveforms, we can see the insulator voltages exceeded their insulation
strength and flash overs are occurred.
· Figure. 8(b) shows the case of overhead ground wires installed (no surge arresters, Case-2). By
shielding effect of overhead ground wires, there was no flashover but the maximum
insulator voltage is close to 200kV, which is the maximum insulation strength under
impulse current.
· Figure. 8(c) shows the case of surge arresters (no overhead ground wires, Case-3). From the voltage
waveforms, some surge arresters operated and insulator voltages were greatly reduced
under about 50kV, which is the knee point voltage of surge arresters.
· Figure. 8(d) shows the case of Case-4 with overhead ground wires and surge arresters installed.
From the voltage waveforms, we can see surge arresters were operated and there is
no big difference from the case of Case-3.
Fig. 8. Differential voltage between phase and neutral conductors under different
line protection conditions (100kA)
3.3 Parametric Analysis of Lighting Induced Over voltages
In this section, overvoltages on phase and neutral conductor at the nearest pole from
lightning strike point were calculated varying the pole’s ground resistance from 50Ω
to 100Ω.
It can be seen from the Figure. 9 that as pole ground resistance increase, phase and neutral conductor voltage increases
also. However, the voltage difference, which is voltage applied along line-post insulator,
decreases as pole ground resistance increases.
Fig. 9. Pole ground resistances vs. conductor/ neutral voltages
These phenomena can be explained from the fact that neutral conductors are located
closer to pole ground than phase conductors so their voltages are more sensitive to
the change of resistance. This simulation results imply that reducing ground resistance
does not contribute to mitigation of overvoltage along line-post insulators and problems
of line post insulator flashover.
4. Conclusions
In this paper, we presented lightning induced voltage calculation technique using
compiled LIV-ATP. With the presented compiled MODELS version of original LIV-ATP,
its computational speed of LIV-ATP was significantly improved and parametric analysis
of LIV, which requires running a lot of simulation cases, was remarkably facilitated.
Based on a series of simulations using the code, evaluations of lightning protection
performances of various protection schemes were conducted. The calculation results
show the effectiveness of overhead ground wires and surge arresters as protection
methods from insulation failures at line-post insulators.
References
EPRI , 2004, Handbook for Improving Overhead Transmission Line Lightning Performance,
pp. 4-8
IEC 60071-2 , 1996-12, Insulation Coordination – Part 2 Applcation Guide, 3rd Edition,
pp. 107
Janicek Frantisek, Mucha Martin, September 22-24, 2006, Multifunctional Relay Developed
in ATP FOREIGN MODEL and C++, Proceedings of the 6th WSEAS International Conference
on Simulation, Modelling and Optimization, Lisbon, Portugal
2006, T & D System Design and Construction for Enhanced Reliability and Power Qualit,
EPRI Technical Report
Høidalen Hans Kr., 2003, Calculation of Lightning-induced Voltages in MODELS Including
Lossy Ground Effects, in IPST
Rusck S., 1958, Induced lightning over-voltages on power transmission lines with special
reference to the over-voltage protection of low-voltage networks, in Transactions
of the Royal Institute of Technology. No. 120, Stockholm, Sweden
Silva J. P., Araújo A. A., Paulino J. S., Dommel H. W., 2001, Calculation of Lightning-Induced
Voltages with RUSCK’s Method in EMTP, in IPST
Paolone M., Perez E., Borghetti A., Nucci C. A., Rachidi F., Torres H., 2005, Comparison
of Two Computational Programs for the Calculation of Lightning-Induced Voltages on
Distribution Systems, in IPST
Biography
He received B.S and M.S degree in electrical engineering from Kwang -Woon and ChonNam
university in 1979 and 2019 respectively.
He has joined Korea Electric Power Corporation since 1987.
He received B.S degree in electrical engineering and M.S/Ph.D in power system engineering
from Dong-guk and Hong-ik university in 1993, 1995 and 2007 respectively.
He joined Korea Electric Power Corporation in 1995 and is a principal researcher,
in which interested power system grounding analysis, at present.
He received B.S degree in electrical engineering and M.S/Ph.D in power system engineering
from Pusan National Univ. and Seoul National Univ. in 1984, 1986 and 1995 respectively.
He joined Korea Electric Power Corporation in 1986 and was a head of transmission
and substation research laboratory.
He is currently with the Department of Energy Policy and Engineering, KEPCO International
Nuclear Graduate School.
His current teaching and research interests include electric power economics and microgrid.