Jour of Adv Research in Dynamical & Control Systems, Vol.

10, 12-Special Issue, 2018

Electric Field Effect on Metallic Particle

Movement for 145kV Gas Insulated Bus Duct
Using Different Numerical Methods
Ambati Giriprasad, Dept of EEE, VNR VJIET. E-mail:[email protected]
Poonam Upadhyay, Dept of EEE, VNR VJIET. E-mail:[email protected]
M. Suryakalavathy, Dept of EEE, JNTU, Hyderabad. E-mail:[email protected]
N. Karthik, Dept of EEE, VNR VJIET. E-mail:[email protected]
Abstract--- This paper presents the analysis of a metallic particle trajectory in 145kV common enclosure Gas
Insulated Bus duct (GIB) using various numerical methods which is compared with segregated GIB. A Gas insulated
Substation (GIS) is a modulated design equivalent of a conventional substation (Air Insulated substation). All the
equipment is housed inside an airtight enclosure filled with a dielectric medium. The dielectric strength of the
insulating medium is severely impacted by the presence of metal particles in it. CIGRE group in one of its study on
GIS systems has stated that of all the insulation failure cases of a GIS, metallic particle contamination has been the
cause of the failure in more than 20% of the cases. Therefore there is a need to study the trajectories of metallic
particles in GIB. In this paper, the movement of metallic particles in a common enclosure configuration of GIS has
been simulated for a 145 kV GIS up to withstand voltage of 275 kV. Trajectory of Al and Cu particles has been
simulated by using Analytical method, Charge Simulation method and Finite Element method. These results are
presented, compared and analysed.
Index Terms--- Maximum Radial Movement, Gas Insulated Bus Duct (GIB), Metallic Particle, Common Enclosure.

I. Introduction
Gas Insulated Substations offer an amicable solution to the problems faced by the modern power systems. There
is a need to upgrade the power system based on the growing needs of consumers. Conventional substation
technology fails to serve in many terrain conditions such as underground, Hilly areas, sea shores an urban territory.
Real estate is a very premium commodity in urban areas where the space required for establishing a new substation
or upgrading an existing one is very expensive and also scarce. GIS substations on an average require less than 10 %
of the space required for Air Insulated Substations.
A GIS is a closed system with a modular design in which all the components are housed in an enclosure filled
with a dielectric gas (SF 6 ). GIS systems come in different varieties such as horizontal or vertical configuration,
Segregated or common enclosure configuration. A horizontal common enclosure GIB [1] is modelled in this work.
The three phases are placed in a single duct. This is called as a common enclosure configuration. Metallic particle
contamination is prevalent problem experienced by GIS systems [2], [3]. During the many phases in the life of a GIS
system like manufacture, transport, erection and operation, metallic particle contamination is an inadvertent
occurrence [2], [3]. The dielectric capability of SF 6 gas falls drastically with metal particle contaminants [4], [5]. It
is a quite cumbersome task to keep this problem at bay as it is inevitable and very arbitrary. The primary and most
important task is to calculate the trajectories of metal particles which can then be used for diagnostic analysis [6].
The movement of metallic particles is calculated by finding out the electric stresses in the duct. The electric field
in a GIB depends on several variables such as conductor and enclosure dimensions, the operating voltage,
permittivity of the dielectric medium etc. All the variables such as gas pressure, restitution coefficient, particle
length are plugged in to corresponding equations to calculate the trajectories of metal particles [7], [8]. The
movement equation obtained after this is a second order differential equation. It is solved by using fourth order RK
method.
A GIS system of voltage rating 145 kV is considered in this work. A 132 kV GIS system is designed for an
operating range of 132 kV to 145 kV. In most of the literature study, it has been observed that a segregated
configuration is used for 145 kV. The demerits of doing so will be discussed in this work. The cost of the substation
increases if segregated configuration is used for this voltage. The literature also contains papers which have used up
to 400 kV which is not required because the withstand voltage is 275 kV [9]. This is not advisable since with or

ISSN 1943-023X 343

Received: 23 July 2018/Accepted: 21 Aug 2018
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, 12-Special Issue, 2018

without particle contamination, breakdown occurs. Common enclosure configuration for 145 kV rating is low cost
than segregated configuration for same MVA rating. This practice is also being followed by manufacturing
companies. The GIS system present in Mint compound in Hyderabad, Telangana, India is a 132/33 kV substation. In
this substation, both the 132 kV and the 33 kV sides are of common enclosure configuration. The analysis is being
done on three methods Analytical Method (AM), Finite Element Method (FEM) [10], [11] and Charge Simulation
Method (CSM) [12], [13] for two particles Al and Cu.

II. Mathematical Modelling

A horizontal common enclosure GIB configuration is considered in this work. The electric field calculations in a
GIB are carried out for various voltages using AM, FEM and CSM. The modelling of a Common enclosure GIB is
done as shown in Fig. 1

Fig. 1: Cross Sectional View of a Common Enclosure GIB

It can be seen in the figure that the three conductors are contained in a single enclosure. The contaminating metal
particle is resting on the enclosure and when a sufficiently strong electrostatic field comes into existence, the particle
lifts off. The electrostatic force can be considered as an actuating force and the opposing forces are the drag force
due to the viscosity of the dielectric medium and gravitational force. When the electrostatic force outweighs the
other two forces, the metal particle tends to move.
The movement of the metal particle [7] can be determined by using (1).
md2 y/dt2 = F e -F d -mg (1)
Where,
m = Mass of the contaminating metal particle
y = Displacement of the particle
F e = Electrostatic force
g = acceleration due to gravity
F d = Drag force due to the viscosity of SF 6
Electrostatic Force
The movement of the metal particle depends on the electrostatic force on it. The electrostatic force is
proportional to the electric field prevailing in the bus duct. The expression for finding the electrostatic force on a
contaminating metal particle can be written as,
Fe = Q X E (2)
E(t) is the electric field inside the bus duct.
The electric field can be calculated by using the following methods
1. Analytical method
2. Charge simulation method
3. Finite element method

ISSN 1943-023X 344

Received: 23 July 2018/Accepted: 21 Aug 2018
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, 12-Special Issue, 2018

Analytical Method
The electric field calculation is done by using Analytical method as in the equations

[Sinωti Cos (π ) ]
Vmax
E ay = (3)
h (h − x)
log( )
Rc

Cos (θ 2 )
[Sin(ωti − 2π )
Vmax
Eby = ]
h 3 ( Rbx )
log( )
Rc
(4)
V 4π Cos (θ 2 )
Ecy = max [ Sin (ωti − ) ]
h 3 ( Rcx )
log( )
Rc
(5)
E =E +E +E
y ay by cy
(6)
Charge Simulation Method
Charge simulation method is a powerful and reliable tool for calculating the electric fields in high voltage
engineering. Figure 2 shows the basic concepts involved in the calculation of electric field using charge simulation
method [9], [12].

Fig. 2: Model of Charge Simulation Concept

The electric field at a point P(x,y) is determined by (7) and (8)
3n λ x − xi
E x (t ) = ∑ i
i =1 2πε
[3 2 2
]sin ωt (7)
( x − xi ) + ( y − y i )

(8)
3n λi y − yi
E y (t ) = ∑ [ ]sin ωt
i = 1 2πε 3 ( x − xi ) + ( y − yi )
2 2

Where,
E x (t) = Component of Field along x-axis
E y (t) = Component of Field along y-axis
X i , yi are the ith fictitious charge coordinates
λ i is the line charge density
n is the number of fictitious charges

ISSN 1943-023X 345

Received: 23 July 2018/Accepted: 21 Aug 2018
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, 12-Special Issue, 2018

Finite Element Method

The calculation of electric field in a common enclosure configuration is shown in fig. 3

Fig. 3: Elements of a Duct

The equation used to calculate the total energy associated that is associated with the assemblage [10], [11] of all
the elements present in the GIB is
W = ∑𝑛𝑛𝑒𝑒=1 𝜀𝜀 [𝑉𝑉]𝑇𝑇 [𝐶𝐶][𝑉𝑉] (9)
Where, n=the no. of elements
V=node voltage matrix
C=Overall or global coefficient matrix
The solution region in Finite element method has a minimum total energy which satisfies Laplace’s and
Poisson’s equations. Hence, the partial derivatives of ‘W’ with respect to each potential at the nodes should be zero.
∂W ∂W ∂W
= ∂V2 = ⋯ ∂n = 0
∂V1
(10)
The Finite element mesh can be generalized as
∑nv=1 Vi Cik = 0 (11)
Where, i=no. of nodes
K=1,2,3,……n
It can be seen from the above that a set of ‘n’ equations is obtained. These equations can be solved by band
matrix method to find the unknown voltages at nodes (V F )
[V F ]=[C F ]-1(-[C FP ][V P ]) (12)
Where, V F = free node voltage matrix
V P = prescribed or fixed node voltage matrix
C FF = free node global coefficient matrix
C FP = free prescribed nodes global coefficient matrix.
The electric field intensity at any point in the bus duct is calculated by using (10)
E = -Grad V (13)
Drag Force
A contaminating metal particle moving in the gas filled enclosure experiences a force called as Drag Force. This
force is due to the viscous nature of the dielectric medium (SF 6 ). Any object moving in a fluid experiences a drag
force. The equation to calculate the drag force on a metal particle is
1 2
Fd = ρv Cd A (14)
2
Where, ρ = Density of the SF 6 ,
v = Speed of the metallic particle
A = Area of the cross section
C d = Drag Coefficient

ISSN 1943-023X 346

Received: 23 July 2018/Accepted: 21 Aug 2018
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, 12-Special Issue, 2018

The equation of the motion of the particle considering all the forces at once [6] can be written as
d2y πε l 2 E (t 0 ) dy dy (15)
m 2
=[ 0 X( E (t))] − mg − πr[6µK d ( )
dt 2l dt dt
ln( ) − 1
r
dy
+ 2.656[ µPg l ]]
dt
E(t) is the electric field intensity calculated by using AM, CSM and FEM which is given in (6), (7), (8) and (13)
respectively.
The movement of the particle can be calculated by solving the second order differential equation as seen in (15).
The solution of this can be obtained by using RK fourth order method.

III. Results and Discussion

The movements of Al and Cu particles [9] were calculated, compared and analysed up to the withstand voltage
(275kV) of a 145KV common enclosure GIB using three numeric methods AM, FEM [10,11] and CSM [9 ]. The
gas pressure of 4 bar, restitution coefficient 1 and a horizontal configuration of GIS are considered in this work.
Enclosure radius is 153 mm, the conductor radius is 27.5 mm and conductors centre radius is 76.5 mm for 145kV
common enclosure GIB. Hence, the gap between conductor and enclosure is 49 mm (Re -Rcc -Rc ) from fig. 4. A
metallic particle of length 12 mm and radius 0.25 mm is assumed to be resting on the inner surface of the enclosure.
The electric field is calculated using AM, FEM and CSM The particle movement has been simulated for123 kV, 132
kV, 145 kV, 170 kV, 200 kV, 245 kV, 275 kV and 300 kV voltages because withstand voltage of 145kV GIS is 275
kV and GIS design voltages are 123 kV, 132 kV, 145kV, 170 kV, 245 kV, 300 kV according to IEEE standards [5].
Maximum radial movements using AM, FEM and CSM methods are compared for Al and Cu particles
respectively in the Tables I and III. Maximum axial movements using AM, FEM and CSM methods are compared
for Al and Cu particles respectively in the Tables II and IV.
Table I: Al Particle’s Maximum Radial Movement in 145KV GIB Using Different Numerical Methods
S. No Voltages (kV) Max. Radial Movement ( mm)
Analytical Method Finite element Method Charge simulation Method
1 123 5.76 6.86 4.52
2 132 6.55 7.49 5.42
3 145 7.69 8.46 6.54
4 170 9.94 10.83 8.72
5 200 13.42 14.64 12.6
6 245 19.09 20.35 17.95
7 275 24.21 25.62 22.53
8 300 29.16 30.76 26.42

Table II: Al Particle’s Maximum Axial Movement in 145kV GIB Using Different Numerical Methods
S. No Voltages (kV) Max. Axial Movement ( mm)
Analytical Method Finite element Method Charge simulation Method
1 123 115.14 119.52 108.53
2 132 124.15 126.45 113.74
3 145 166.22 177.22 129.35
4 170 177.78 189.58 142.83
5 200 274.68 292.64 217.54
6 245 359.13 367.59 295.36
7 275 460.10 506.28 398.57
8 300 549.88 604.59 487.48

Table III: Cu Particle’s Maximum Radial Movement in 145 kV GIB Using Different Numerical Methods
S. No Voltages (kV) Max. Radial Movement ( mm)
Analytical Method Finite element Method Charge simulation Method
1 123 0 0 0
2 132 0 0 0
3 145 0 0 0
4 170 0 0 0
5 200 3.14 04.36 2.31
6 245 6.573 7.32 5.26
7 275 8.1 9.25 6.74
8 300 9.89 10.94 7.53

ISSN 1943-023X 347

Received: 23 July 2018/Accepted: 21 Aug 2018
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, 12-Special Issue, 2018

Table IV: Cu Particle’s Maximum Axial Movement in 145 kV GIB Using Different Numerical Methods
S. No Voltages (kV) Max. Axial Movement ( mm)
Analytical Method Finite element Method Charge simulation Method
1 123 0 0 0
2 132 0 0 0
3 145 0 0 0
4 170 0 0 0
5 200 64.77 69.47 46.28
6 245 88.35 92.59 75.83
7 275 126.51 132.38 117.51
8 300 152.06 158.75 144.57
Maximum axial movement of Al particle is greater than that of the Cu particle which is observed From Tables II
and IV in all the three methods. Fig.5 and Fig.6 show comparison among the maximum Radial movements which
are obtained by using AM, CSM and FEM for Al and Cu particles respectively. Maximum radial movements of
Aland Cu particles have been compared in the graphs shown in Fig.7, Fig.8 and Fig 9 for AM, FEM and CSM
methods respectively.

Fig. 5: Comparing an Al Particle’s Maximum Radial Movement for AM, CSM and FEM

Fig. 6: Comparing a Cu Particle’s Maximum Radial Movement for AM, CSM and FEM
From Tables I and II, Fig. 5 and Fig.6, it is observed that maximum radial movements for all the three methods
are almost same but radial movement obtained using CSM is slightly less than that of analytical method. The radial
movement obtained by FEM is slightly greater than analytical method.
The maximum Radial movements for Al and Cu have been compared using AM, FEM and CSM respectively in
Fig.7, Fig.8, and Fig.9.

Fig. 7: Comparison of Max. Radial Movement of Al, Cu and Ag Particles Using AM

ISSN 1943-023X 348

Received: 23 July 2018/Accepted: 21 Aug 2018
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, 12-Special Issue, 2018

Fig. 8: Comparison of Max. Radial Movement of Ag, Cu and Al Particles Using CSM

Fig. 9: Comparing Max. Radial Movement of Ag, Cu and Al Particles Using FEM
As the voltage increases, particle movement also increases. It can be therefore concluded that the max.
movement of the particle occurs at the peak value of the sinusoidal wave. Therefore only peak voltages and
maximum movements are studied. The testing of dielectric strength in high voltage engineering is also however
done for peak voltages.
Maximum radial movement of Al particle is greater than that of the Cu particle which is observed From Tables I
and II, Fig. 7, Fig. 8 and Fig. 9 in all the three methods.
The radial and axial movements of a contaminating metallic particle have been calculated in this work. Radial
movement of the particle causes a flashover between the conductor and the enclosure. This is a severe problem and
needs to be addressed promptly, which is why the radial movement is studied in tables and graphs. Axial movement
may cause a failure of the spacer insulation. But the gap in axial direction is more (length of a GIB is typically 4000
mm to 6000 mm). The failure rate of a GIB is less due to axial movement but they are however studied in tables.
The axial movement of particles in a 145 kV common enclosure GIS is always less than 4000 mm (max. 604.59
mm). There is no failure due to axial and radial movements. Hence, a common enclosure configuration is more
suitable for 145 kV rating.
Conductor radius of 27.5mm and enclosure radius of 76mm are used as dimensions of GIB for the 145kV
segregated GIS [8] and the remaining specifications are same as those of 145kV common enclosure GIS. .Hence gap
between conductor and enclosure is 48.5mm. The maximum radial movements of Al particle in common enclosure
and segregated GIS are compared in the Table 7. Gaps between particle and conductor in common enclosure and
segregated GIS are compared for Al particle in the Table 8.
Table V: Max. Radial Movement of Al Particle for Different Dimensions and Configurations Using AM for 145 kV
GIB
S. No Voltage (kV) Max. Radial Movement ( mm)
Common enclosure with optimum dimensions segregated GIS with withdrawn dimensions
1 123 5.76 26.38
2 132 6.55 30.74
3 145 7.69 35.23
4 170 9.94 CEG
5 200 13.42 CEG
6 245 19.09 CEG
7 275 24.21 CEG
8 300 29.16 CEG
CEG – Crossing Electrode Gap

ISSN 1943-023X 349

Received: 23 July 2018/Accepted: 21 Aug 2018
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, 12-Special Issue, 2018

Table VI: Gap between Particle and Conductor of Al Particle for Different Dimensions and Configurations Using
AM for 145 kV GIB
S. No Voltages (kV) Gap between particle and conductor ( mm)
Common enclosure GIB with optimum dimensions Segregated GIB with withdrawn dimensions
1 123 43.24 22.12
2 132 42.45 17.76
3 145 41.31 13.27
4 170 39.06 0
5 200 35.55 0
6 245 29.91 0
7 275 24.79 0
8 300 19.84 0
In fig. 10 and fig. 11, the instantaneous movement of Al particle for 145 kV for outdated and updated standards
are plotted respectively.

Fig. 10: Instantaneous Movement of Al Particle for 145 kV with Outdated Standards for Segregated Configuration

Fig. 11: Instantaneous Movement of Al Particle for 145 kV with Updated Standards for Common Enclosure
Configuration
It is observed that the radial movement of Al particle in common enclosure is less than radial movement of Al
particle in segregated from Table V and Fig. 11 and 12. Even though the electric field is the resultant of three
phases, this effect is dominated by the increase in the enclosure radius because the electric field intensity is inversely
proportional to product of enclosure radius and logarithm of the ratio of enclosure radius and conductor radius. The
gap between particle and conductor of common enclosure GIS is greater than that of segregated GIS which is
observed from Table VI. If the gap between the conductor and particle is increased, the probability of first ionization
(Particle to conductor) is decreased and the probability of second ionization (Particle to enclosure) is decreased
because First ionization leads to second ionization. Hence, the failure rate is less in a 145 kV common enclosure
GIS, in other words, the reliability is increased.

IV. Conclusion
The lifted off metal particle is neither crossing the electrode gap nor is it reaching near to the conductor. It is
lifted off of its position and falls back. Hence the reliability of a common enclosure GIB is more and the cost is less
than that of a segregated GIB. Hence, a common enclosure GIB is recommended for 145 kV rating.
The movement of metallic particle calculated by the three methods (AM, CSM, FEM) is almost same with very
slight changes. So, it is correct to use any of the three methods. The movement in Al particle is more when
compared to Cu particle due to its lesser density than copper.

V. Future Scope
Movement of the metallic particle using Finite difference method will be studied and presented for 145 kV
common enclosure GIS. The superimposition of Power frequency voltage with lightning and switching surges can
be studied for a 145 kV common enclosure GIB.

ISSN 1943-023X 350

Received: 23 July 2018/Accepted: 21 Aug 2018
Jour of Adv Research in Dynamical & Control Systems, Vol. 10, 12-Special Issue, 2018

Acknowledgment
The authors are thankful to CPRI, CEA boards for conducting valuable seminars and workshops on the subject.
The authors are also thankful to Dr. Mohan Rao and Dr. A. Raghuram for sharing their valuable experiences in the
field of GIS systems.

References
[1] Amarnath, J., Kamakshaiah, S., Radhakrishna, C., Singh, B.P. and Srivastava, K.D. Particle trajectory in a
common enclosure three phase SF6 bus duct. 12th International Symposium on High Voltage Engineering
during, 2001, 20-24.
[2] Anis, H. and Srivastava, K.D. Free conducting particles in compressed gas insulation. IEEE Transactions
on Electrical Insulation (1981) 327-338.
[3] Moreos, M.M., Srivastava, K.D. and Anis, H. Dynamics of metallic contaminants in compressed gases.
IEEE Int. Conf. on Industrial Applications, 1983.
[4] Cookson, A.H., Bolin, P.C., Doepken, H.C., Wootton, R.E., Cooke, C.M. and Trump, J.G. Recent research
in the United States on the effect of particle contamination reducing the breakdown voltage in compressed
gas insulated systems. CIGRE, 1976, 1-10.
[5] Laghari, J.R. and Qureshi, A.H. A review of particle-contaminated gas breakdown. IEEE Transactions on
Electrical Insulation (1981) 388-398.
[6] Doepken, H. Compressed-gas insulation in large coaxial systems. IEEE Transactions on Power Apparatus
and Systems (1969) 364-369.
[7] Amarnath, J., Singh, B.P., Radhakrishna, C. and Kamakshiah, S. Determination of particle trajectory in a
Gas Insulated Busduct. IEEE Conf. Electr. Insul.Dielectr. Phenomena (CEIDP), 1991, 399-402.
[8] IEC 62271-203 International Standards2
[9] Brunke, J.H. and Koch, H. Revision GIS standard rated above 52 kV–C37.122. IEEE Transaction,
Substation Technical Specifications, 2010.
[10] Matthew, S. Elements of electromagnetic. 4th edition, Oxford Publications, 2007, 694-711.
[11] Andersen, O.W. Laplacian Electrostatic field calculation by Finite elements with automatic grid
generation. IEEE PES winter meeting, New york, 1973.
[12] Singer, H., Steinbigler, H. and Weiss, P. A Charge Simulation Method for the Calculation of High Voltage
Fields. IEEE Power Engineering Society, 1974.
[13] Malik, N.H. A Review of Charge simulation method and its application. IEEE Trans. Electr. Insul. 24
(1989) 3-20.

ISSN 1943-023X 351

Received: 23 July 2018/Accepted: 21 Aug 2018