G.V.Nagesh
Kumar1,*,
Amarnath
Jinka2
J. Electrical Systems 10-3 (2014): 331-343
Regular paper
Assessing the Risk of Failure due to
Particle Contamination in GIS under
various Coefficient of Restitutions
JES
Journal of
Electrical
Systems
The development of compressed gas insulated switchgear (GIS) equipment has progressed
rapidly. Conducting particles in transmission and switching equipment insulated by compressed
sulphur hexafluoride (SF6) can result in loss of as much as 90% of the gas dielectric strength.
These particles may be free to move in the electric field or may be fixed on the conductors, thus
enhancing local surface fields. In a horizontal coaxial system with particles resting on the inside
surface of the enclosure, the motion of such particles is random but the randomness depends on
the coefficient of restitution and angle of incidence when approaching the coaxial conductors.
The power industry has utilized several methods to control and minimize the effect of particle
contamination in GIS. One such technique is to apply a dielectric (high resistivity) coating to
the inside surface of the outer GIS enclosure. The electric field necessary to lift a particle
resting on the inside surface of a GIS enclosure is much increased due to the coating. The
simulation of the particle movement was carried under different Coefficient of Restitutions for
coated busduct. The results have been presented and analyzed.
Keywords: Electric Field, Gas Insulated Substations, Metallic Particles, Particle Contamination
1. Introduction
Demand for electrical power has become one of the major challenges faced by the
developing countries. Considering the relatively low per capita power consumption, there is
a constant need for power capacity addition and technological upgradation whereas nonconventional energy systems have proved to be good alternative sources for energy. In
developing countries like India most of the additional power has been met by conventional
electric sources. Hence, the emphasis has shifted towards improving the reliability of
transmission and distribution systems and ensuring that the innovations are not harmful to
the environment. Air insulated power transmission and distribution substations suffer
variations in the dielectric capability of air to withstand varying ambient conditions and
deterioration of the exposed components due to oxidization and the corrosive nature of the
environment. The size of the sub-station is also substantial due to the poor dielectric
strength of air.
In order to enhance the life and reliability of a power transmission and distribution substation, it is desirable to protect the sub-station components from a corrosive and oxidising
environment. Metal encapsulation of the sub-station elements provides a simple and
effective solution to the problem of durability of the substations. The use of a bus duct, with
pressurized nitrogen gas, is a good example of devices with metal encapsulation used in
power sub-stations. The size of the container is a direct function of the dielectric strength of
*
Corresponding author: G.V.Nagesh Kumar, Associate Professor, Department of EEE, GITAM University,
Visakhapatnam, 530045, Andhra Pradesh, India E-mail: gundavarapu_kumar@yahoo.com
1
G.V.Nagesh Kumar, Associate Professor, Department of EEE, GITAM University
2
J. Amarnath, Professor, Department of EEE, Jawaharlal Nehru Technological University
Kukatpally, Hyderabad, 500085, Andhra Pradesh, INDIA
Copyright © JES 2014 on-line : journal/esrgroups.org/jes
G.V.Nagesh Kumar et al: Assessing the Risk of Failure in GIS...
the insulating medium. The container/enclosure sizes are thus large with a poor insulation
like air or nitrogen.
The electric power industry’s preferred gaseous dielectric sulfur hexafluoride (SF6), has
been shown to be a greenhouse gas. SF6 gas is mainly used for electrical insulation and,
especially, for arc quenching current interruption equipment used in the transmission and
distribution of electrical energy. Gas insulated sub-station systems offer a compact, costeffective, reliable and maintenance-free alternative to the conventional air insulated substation systems. Their compact size offers a practical solution to vertically upgrade the
existing sub-station and to meet the ever-increasing power demand in developing countries.
Compressed Gas Insulated Substations (CGIS) consist basically of a conductor supported
on insulators inside an enclosure, which is filled with sulfur hexafluoride gas (SF6). SF6 gas
provides phase to ground insulation so that space required is very much less. The voltage
withstands capability of SF6 bus duct is strongly dependent on field perturbations, such as
those caused by conductor surface imperfections and by conducting particle contaminants.
The contaminants can be produced by abrasion between components during assembly or
operations. As one is aware of the attractive features of a Gas Insulated Substation (GIS),
they also suffer from certain drawbacks. One of them is the outage due to seemingly
innocuous conducting particles, which accounts for nearly 50% of the GIS failures. Flash
over in a GIS is, in general, associated with longer outage times and greater costs than in a
conventional air insulated substation. A conducting particle can short-circuit a part of the
insulation distance, and thereby initiate a breakdown, especially if electrostatic forces cause
the particle to bounce into the high field region near the high voltage conductor.
Generally the allowable design levels are used because SF6 is very sensitive to field
perturbations such as those caused by conductor surface imperfections and by conducting
particle contaminants. A study of CIGRE group suggests that 20% of failure in GIS is due
to the existence of various metallic contaminations in the form of loose particles. These
particles may exist on the surface of support insulator, enclosure or high voltage conductor.
Under the influence of high voltage, they can acquire sufficient charge and randomly move
in the gap due to the variable electric field. Several authors [1-6] have reported the
movement of particles with reference to a few parameters.
The presence of contamination can therefore be a problem with gas-insulated
substations operating at high fields. If the effects of these particles could be eliminated,
then this would improve the reliability of compressed gas insulated substation. It would
also offer the possibility of operating at higher fields to affect a potential reduction in the
GIS size with subsequent savings in the cost of manufacture and installation.
Some of the methods of conducting particle control and de-activation are Electrostatic
trapping, Use of adhesive coatings to immobilize the particles, Discharging of conducting
particles through radiation and Coating conducting particles with insulating films. The
purpose of this work is to develop techniques, which will formulate the basic equations that
will govern the movement of metallic particles like aluminum, copper and silver and also
the maximum movement is to be reduced for increasing the dielectric strength of gaseous
Insulation. In This Paper dielectric performance is improved by coating the inner surface of
the enclosure with a light shade with micron thickness epoxy resin. The Coating thickness
is also varied and the results are presented and analyzed. The specific work reported deals
with the charge acquired by the particle due to macroscopic field at the tip of the particle,
332
J. Electrical Systems 10-3 (2014): 331-343
the force exerted by the field i.e., electric field on the particle, drag due to viscosity of the
gas and random behavior during the movement. Wire like particles of aluminum and copper
of 10 mm in length and 0.25 mm as radius on a 1-phase bus duct enclosure have been
considered. The movement pattern for higher voltages class has been also obtained.
2. Modelling Technique of Busduct
Figure 1 shows a typical horizontal busduct comprising of an inner conductor and an
outer enclosure, filled with SF6 gas is considered for the study. A particle is assumed to be
at rest at the enclosure surface, until a voltage sufficient enough to lift the particle and move
in the field is applied. After acquiring an appropriate charge in the field, the particle lifts
and begins to move in the direction of the field after overcoming the forces due to its own
weight and drag. The simulation considers several parameters like the macroscopic field at
the location of the particle, its weight, viscosity of the gas, Reynold’s number and
coefficient of restitution on its impact to the enclosure.
Fig 1: Typical Single Phase Gas Insulated Bus duct
2.1 Metallic Particles on Bare Electrodes in GIS
For particles on bare electrodes, several authors have suggested expressions for the
estimation of charge on both vertical/horizontal wires and spherical particles. The equations
are primarily based on the work of Felici [7]. The lift-off field for a particle on the surface
of an electrode can be estimated by solving the following equations. The gravitational force
acting on a particle of mass ‘m’ is given by
(1)
Fg = mg
The expression of the electrostatic force can be expressed as Fe = QE;
where Q is the particle charge and E is the ambient electric field.
The motion of the particle is simulated by using the equation:
m
d2y
dt
2
=
Fe - mg - Fd
(2)
(3)
333
G.V.Nagesh Kumar et al: Assessing the Risk of Failure in GIS...
The motion equation using all forces can therefore be expressed as [8-14]:
m &y& ( t )
=

 π ∈ l 2 E(t )
V Sin ω t
0
0

X
r
 ln( 2 l ) − 1
[r 0 - y(t)] ln ( 0

r
ri



)

-
mg
(4)
(
y& (t) π r 6 µ K
-
d
( y& ) + 2.656 [ µ ρ g l y& (t)]
0.5
)
So, two initial conditions are necessary for a solution:
.
.
m y(t = 0+ ) = −Rmy(t = 0)
and y(t = 0 + ) = 0
where R is the restitution coefficient given by the ratio of incoming-to-outgoing impulses.
The restitution coefficient for copper and aluminium particles seem to be in the range of 0.7
to 0.95: R=0.8 implies that 80% of the incoming impulse of the particle is preserved when
it leaves the enclosure. The motion equation is a second order non-linear differential
equation and in this paper, the equation is solved by using Runge-Kutta 4th Order Method.
2.2 Metallic Particle Movement on Coated Electrodes in GIS
The main reason for coating an electrode surface with a dielectric material is to increase the
breakdown voltage of the system. The dielectric coating decreases the high local fields
caused by conductor roughness. It impedes the development of pre-discharges in the gas
due to the resistance of the coating. The Charge mechanism of a metallic particle on a
dielectric coating is shown in Figure 2.
Fig 2. Charge mechanism of a metallic particle on a dielectric coating
Pseudo-resonance occurs when the particle impacts the outer electrode with a speed and
at a phase angle of electric field which is nearly the same as on previous bounce, it most
occurs with bounce of small amplitude. When Rc increases, the maximum height reached
by the particle and the amplitude of pseudo-resonance increase, which leads to higher
possibility of capturing the particle. Meanwhile, the time between the instant of particle liftoff, and the instant of onset of pseudo-resonance, would increase. The coefficient of
restitution, Rc, depends on the coating and the angle of impact of the wire particle with the
334
J. Electrical Systems 10-3 (2014): 331-343
coating. It is very difficult to get an accurate value for the coefficient of restitution, because
particle movement is random.
3. Results and Discussions
The influence of applied voltage on the maximum height reached by the particle for two
different values of the coefficient of restitution (neglecting drag force) is shown in Table I.
When the coefficient of restitution increases, the maximum height would increase, because
when impact with outer enclosure the particle gets higher inverse velocity with higher
coefficient of restitution which enables the particle to overcome more gravitational force.
Therefore, the particle may eventually cross the gap due to the large height of the bounce.
As the applied voltage increases the maximum radial movement also increases as given in
Table I. Further calculations may reveal the limiting voltage to enable the particle to reach
the high voltage conductor. The values given in the Table II are the maximum flight of
particles in radial and axial directions using Monte Carlo Technique. The main reason for
the lower movement in Coated Busduct is the lower charge acquired by the particles even at
higher fields.
The low magnitude of charge acquired by the particle ensures relatively feeble force on
the particles. Thus manufactures of GIS often resort to coating of the busduct to prevent
particle movement and consequential break down. For particles (Al, Ag and Cu) the nature
of movement is similar. However the maximum movement for aluminum particle is higher
than that of the copper particle. This is expected since aluminum particle is lighter. Figures
3 to 5 show the movement pattern for applied sinusoidal voltage of 100 kV. Figure 3 shows
the movement of the Al, Cu and Silver particles with the variation of voltage. Figures 4 to
27 shows the axial and radial movements of particle with and without Monte Carlo
Technique.
TABLE I
Maximum radial movement of aluminum, copper and silver particles
in a 1- phase coated Gas Insulated Busduct
Voltage
(kV)
75
100
132
145
Type
Al
Cu
Ag
Al
Cu
Ag
Al
Cu
Ag
Al
Cu
Ag
Maximum Radial Movement (mm)
RC = 0.8
0.11329
0.00494
0.00301
0.42659
0.01473
0.01144
0.91454
0.09136
0.04483
1.12572
0.16511
0.09917
RC = 0.95
0.34051
0.0125
0.00581
0.72828
0.09453
0.05773
1.28811
0.30877
0.23238
1.53034
0.41098
0.31804
335
G.V.Nagesh Kumar et al: Assessing the Risk of Failure in GIS...
TABLE II
Axial And Radial Movement Of Aluminum, Copper And Silver Particles
in A 1- Phase Coated GIB Using Monte Carlo Technique
Voltage (kV)
75
100
132
145
Type
Al
Cu
Ag
Al
Cu
Ag
Al
Cu
Ag
Al
Cu
Ag
Maximum Radial Movement (mm)
RC = 0.8
RC = 0.95
Axial
Radial
Axial
Radial
2.678
0.1132
7.3267
0.3405
0.1048
0.0049
0.1794
0.0125
0.0637
0.003
0.1053
0.0058
13.772
0.4266
17.481
0.7282
0.3696
0.0147
1.4128
0.0945
0.273
0.0114
0.7386
0.0577
34.346
0.9145
29.867
1.2881
2.0877
0.0913
6.482
0.3087
0.9852
0.0448
4.4998
0.2324
41.388
1.1257
34.45
1.5302
4.1865
0.1651
9.219
0.411
2.2955
0.0992
6.8015
0.318
Fig 3. Variation of Maximum Radial Movement to Sinusoidal Voltage
336
J. Electrical Systems 10-3 (2014): 331-343
337
G.V.Nagesh Kumar et al: Assessing the Risk of Failure in GIS...
338
J. Electrical Systems 10-3 (2014): 331-343
339
G.V.Nagesh Kumar et al: Assessing the Risk of Failure in GIS...
340
J. Electrical Systems 10-3 (2014): 331-343
341
G.V.Nagesh Kumar et al: Assessing the Risk of Failure in GIS...
342
J. Electrical Systems 10-3 (2014): 331-343
5. Conclusion
It has been observed that metallic particle contamination is often present in GIS and such
contamination adversely affects the insulation integrity. One of the techniques used to
reduce the particle movement and improve dielectric performance is to coat the inner
surface of the enclosure. Coating the inner side of the enclosure increases the electric field
necessary to lift a particle resting on the bottom of a GIS enclosure. If the coating thickness
is increased, the charge on the particle is reduced. For GIS of a given voltage class, the
busbar voltage and the clearance between busbar and enclosure are generally fixed. Hence,
the charge on the particle can be reduced by coating the enclosure inner surface by a
suitable dielectric material. In this paper epoxy coating of thickness 100 µm is taken. It is
also observed that as the coating thickness increases the movement of the particle
decreases. When the coefficient of restitution increases, the maximum height would
increase, because when impact with outer enclosure the particle gets higher inverse velocity
with higher coefficient of restitution which enables the particle to overcome more
gravitational force. The Randomness of the particle is also simulated by Monte Carlo
Technique to find the motion in Axial and Radial directions and it is observed that the
movement in radial direction is same without or with Monte Carlo Technique. It is also
inferred that the axial movement is strongly dependent on the random behavior of the
particle.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
L. G. Christophorou, J. K. Olthoff, R. J. Van Brunt, “SF6 and the Electric Power Industry”, IEEE Electrical
Insulation Magazine, DEIS, 1997, pp. 20-24.
“Hand Book of Switch Gears” BHEL, Tata McGraw Hill Publishing, 2005.
M. M. Morcos, S. Zhang, K. D. Srivastava and S. M. Gubanski, “Dynamics of Metallic Particle
Contaminants in GIS with Dielectric-Coated Electrodes”, IEEE Trans. on Power Delivery, Vol. 15, No. 2,
Apr 2000, pp. 455-460.
Arora, A.; Koch, H. ” Design features of GIS”, IEEE Power Engineering Society General Meeting, 2005.
Issue , 12-16 June 2005 Page(s): 927 - 929 Vol. 1
J .R. Lahari, A. H. Qureshi, “A review of particle contaminated gas breakdown”, IEEE Trans, on Electrical
Insulation, Vol, EI – 16, NO.5, pp 388-398, 1981.
G.V.Nagesh Kumar J.Amarnath, B.P.Singh, K.D.Srivatsava., “Influence of Power frequency and Impulse
voltages on particle movement in Gas Insulated Bus duct (GIB) with epoxy coatings”. at National Seminar
on Insulating materials for the Power Industry, organized by Central Power Research Institute, Bangalore
during 26-27 August, 2004
N.J.Felici; “Forces et charges de petits objects en contact avec une electrode affectee d’un champ
electrique”; Revue generale de I’ electricite, pp. 1145-1160, October 1966.
G.V.Nagesh Kumar, J.Amarnath, B.P.Singh, K.D.Srivatsava “Particle movement in a 245 KV, 300 KV and
400 KV Gas Insulated Substations with and without dielectric coating enclosure” National Conference APSC
2004 NIT, Rourkela
H. Anis, K.D. Srivastava; “Free Conducting particles in Compressed Gas Insulation”; “IEEE Trans. on
Electrical Insulation, Vol E1-16, pp.327- 338, August, 1981.
K.S. Prakash, K.D. Srivastava, M.M. Morcos, “Movement of Particles in Compressed SF6 GIS with
Dielectric Coated Enclosure”, IEEE Trans. DEI, Vol. 4 pp. 344-347, 1997.
A.Bargigia, W. Koltunowicz, A. Pigini; “Detection of Partial Discharges in Gas Insulated Substations”;
IEEE Transactions on Power Delivery, Vol 7, No.3, pp.1239-1249, July 1992.
G.V. Nagesh Kumar, J. Amarnath, B.P.Singh and K.D.Srivastava “Electric field effect on metallic particle
contamination in a common enclosure gas insulated busduct”, IEEE Transactions on DEI, April 2007,
Volume: 14, Issue: 2 Page(s): 334-340, ISSN: 1070-9878.
M.Venu Gopala Rao, G.V.Nagesh Kumar, J.Amarnath, S.Kamakshaiah, and K.D.Srivastava, “Movement
Patterns of Metallic Particles in a Single Phase Gas Insulated Busduct with Superimposed Lighting and
Switching Impulses” Journal of High Voltage Engineering, China, December 2008, Pages: 2642- 2647,
Volume: 34, Issue: 12
G.V. Nagesh Kumar, J. Amarnath, B.P.Singh and K.D.Srivastava “Particle Initiated Discharges in Gas
Insulated Substations by Random Movement of Particles in Electromagnetic Fields”, International Journal
of Applied Electromagnetics and Mechanics Volume 29, Number 2 / 2009, Pages 117-129.
343