STUDY OF BREAKDOWN OF AIR-GAPS

AT HIGH VOLTAGES

Autumn Sem 2010 Report submitted in partial fulfillment of the requirements for
the degree of

Master of Technology
In
Power System Engineering

Submitted by:
Saswata Sekhar Roy
(09EE6314)

DEPARTMENT OF ELECTRICAL ENGINEERING

INDIAN INSTITUTE OF TECHNOLOGY
KHARAGPUR
Nov 2010
 Abstract

Corona inception and breakdown are investigated in air for point-to-Sphere,

sphere-to-sphere, rod(square)-to-sphere gaps under 50-Hz ac voltages. The
Breakdown Voltage increases if the Gap Distance increases. The equations of
Breakdown Voltage as a function of Gap Distance are obtained from the Curve
fitting method.

1
 Contents

Abstract..................................................................................…………………..1

1 Introduction 1
1.1 Types of Mechanism for Breakdown………………………………………………………………………………………………3
1.1.1 Townsend Theory………………………………………………………………………………………………………………………..3
1.1.1.1 Current Growth in the Presence of Secondary Processes…………………………………………………………4
1.1.1.2 Townsend’s Criterion for Breakdown……………………………………………………………………………………….5
1.1.2 Streamer Theory of Breakdown in Gases…………………………………………………………………………………….5
1.2 Paschen’s Law………………………………………………………………………………………………………………………………..7
1.3 Paschen’s curve………………………………………………………………………………………………………………………………7

2 Experimental Techniques 2
2.1 Circuit Diagram……………………………………………………………………………………………………………………………….8
2.2 Test procedures………………………………………………………………………………………………………………………………9
2.3 Air Density Correction factor…………………………………………………………………………………………………………..9
2.4 Humidity Correction Factors……………………………………………………………………………………………………………9
2.5 Standard Reference Atmosphere……………………………………………………………………………………………………9
2.6 According to IEC 60052 (2002) ………………………………………………………………………………………………………..9

3 Work done so far 3

3.1 Calculation……………………………………………………………………………………………………………………………………..12
3.2Results…………………………………………………………………………………………………………………………………………….13
3.2.1 Breakdown voltage in respect to gap distances of SPHERE-SPHERE…………………………………………….13
3.2.2 Breakdown voltage in respect to gap distances of SPHERE-POINT………………………………………………14
3.2.3 Breakdown voltage in respect to gap distances of POINT- SPHERE………………………………………………16
3.2.4 Breakdown voltage in respect to gap distances of ROD(SQUARE)- SPHERE…………………………………18
3.3 Conclusion………………………………………………………………………………………………………………………………………20

4 Future Work 4
4.1 Future Work…………………………………………………………………………………………………………………………………….21

2
 1
Introduction

Various phenomena occur in Air dielectric when a voltage is applied. When low
voltage is applied, small current flows between the electrodes and the insulation
retains its electrical properties. If the applied voltage is large, the current flowing
through the insulation increases very sharply and an electrical breakdown occur. A
strongly conducting spark formed during breakdown, practically produces a short
circuit between the electrodes. The maximum voltage applied to the insulation at
the moment of breakdown is called the breakdown voltage. For a uniform gap, the
effect of humidity on the breakdown voltage is negligible. For non-uniform gaps
such as rod-sphere, sphere-sphere like gaps the influence of humidity is found to
be of significant. A correction procedure recommended by IEC is commonly used
for converting the measured voltage or the test voltage from non-standard to
standard reference atmosphere. So the result in analysis is then corrected into the
standard reference conditions [in STP].

1.1 Types of Mechanism for Air Breakdown

There are two types of mechanism for Air breakdown i. Townsend Theory
ii. Streamer Theory
1.1.1 Townsend Theory

Townsend conducted experiments on the growth of these currents which led to

breakdown under d.c. voltage conditions, and he proposed a theory to explain the
phenomenon named as Townsend’s Current Growth Equation.

3
Assuming n0 electrons are emitted from the cathode and when one electron collides
with a neutral particle, a positive atom and electron formed. This is called as
ionization by collision.

If  be the average number of ionizing collisions made by an electron per

centimeter travel in the direction of the field where it depends on gas pressure p
and E/p, and is called the Townsend’s first ionization coefficient. At any distance x
from the cathode when the number of electrons, nx , travel a distance of dx they
give rise to ( n x dx ) electrons. Then, the number of electrons reaching the anode at
x=d, nd will be n0  n x x 0
………………………………....1

dn x
 n x or n x  n 0 e x ……………………….2
dx

And, at x=d. n d  n0 e x ………………………......3

Then, the number of electrons reaching the anode(x=d) will be-

nd  n0
e d  1  …………………………….4
n0

Therefore the average current in the gap, which is equal to the number of electrons
traveling per second is
I  I 0e d ………………………….5

Where I0 is the initial current at the cathode.

1.1.1.1 Current Growth in the Presence of Secondary Processes

When the initial set of electrons reach the anode the single avalanche process is
completed. Townsend’s secondary ionization coefficient  in the same way as  ,
as the net number of secondary electrons produced per incident positive ion,
photon, excited particle or metastable particle and the total value of  due to the
three different processes is    1   2   3 and is a function of gas pressure p and
E/p. Following Townsend’s procedure for current growth, it may be assumed that
n0'  Number of secondary electrons produced due to secondary  processes.
Let, n0''  total number of electrons leaving the cathode.

4
Then
n0''  n0  n0' …………………………………..6
The total number of electrons n reaching to the anode becomes,
n  n 0'' e d  (n0  n0' )e d ……………………….7
And n 0'   [ n  ( n0  n0' )] …………………………..8
n 0 e d I 0 e d
Eliminating n0' , n or I  …………..9
1   (ed  1) 1   (e d  1)

1.1.1.2 Townsend’s Criterion for Breakdown

Eqn. 9 give the total average current in a gap before the occurrence of breakdown.
As the distance between the electrode d is increased the denominator of the
equation tend to zero and at some critical distance d=ds
1   (ed  1)  0 ………………………………..10
For values of d<ds, I is approximately equal to I 0 and if the external source for
the supply of I 0 is removed, I becomes zero. If d=ds, I   and the current will
be limited only by the resistance of power supply and the external circuit. This
condition is called Townsend’s Breakdown Criterion and can be written as
 (ed  1)  1 Normally, ed is very large, and hence the above equation reduces to
ed  1 …………………………………….….11
For a given gap spacing and at a given pressure the value of voltage V which gives
the values of  and  satisfying the breakdown criterion is called the spark
breakdown voltage V, and the corresponding distance d is called the sparking
distance.
Townsend Mechanism explains the phenomena of breakdown only at low
pressures, corresponding to p  d values of 1000 torr-cm and below.
For the high pressure breakdown Streamer Theory is to be used.

1.1.2 Streamer Theory of Breakdown in Gases

The theory predicts the development of a spark discharge directly from a single
avalanche in which the space charge develop by the avalanche itself is said to
transform the avalanche into a plasma steamer. In fig 1, a single electron starting
at the cathode by ionization builds up an avalanche that crosses the gap. The
electrons in the avalanche move very fast compared with the positive ions. By the
time the electrons reach the anode the positive ions are in their original positions
and form a positive space charge at the anode. This enhances the field, and the
secondary avalanches are formed from a few electrons produced due to the photo-
5
ionization in the space charge region. This occurs first near the anode where the
space charge is maximum and a further increase in the space charge. This process
is very fast and the positive space charge extends to the cathode very rapidly
resulting in the formation of a streamer.

Fig. 1
Ionization proces

Comparatively narrow luminous tracks occurring at breakdown at pressures are

called streamers. As soon as the streamer tip approaches the cathode, a cathode
spot is formed and a stream of electrons rush from the cathode to neutralize the
positive space charge in the streamer; the result is a spark and the spark breakdown
occurs.
A simple quantitative criterion to estimate the electric field E r which is produced
by the space charge, at the radius r and that transforms an avalanche into streamer
 ex V
is given by Er  5.27  10 7 …………………………..12
x cm
p
where  is the Townsend’s first ionization coefficient, p is the gas pressure in torr
and x is the distance to which the streamer extends in the gap. When E r  E and
x  d the equation above simplifies into;
 d  ln  p  14.5  ln E p  0.5 ln d p . ……………..13
This equation is solved for  p and E p at a given p and d that satisfy the equation.
The breakdown voltage is given by the corresponding product Ed .
It is generally assumed that for pd values below 1000 torr-cm and gas pressures
varying from 0.01 to 300 torr, The Townsend mechanism operates, while at higher
pressures and pd values the streamer mechanism plays the dominant role in
explaining the breakdown phenomena.

6
1.2 Paschen’s Law
The breakdown criterion
1   (ed  1)  0 ……………………….14
where  and  are functions of E p , i.e.
  f 1  E 
p  p
and   f 2  E p  . Also E  V Substituting for E in the expressions
  d
 and  and rewriting equation 14 we have
f 2  V  e pd f (V pd )  1  1 …………………..15
1

 pd   
This equation shows a relationship between V and pd, and implies that the
breakdown voltage varies as the product pd varies. Knowing the nature of
functions f1 and f 2 we can write the equation V  f  pd  known as Paschen’s law
and has been experimentally established for many gases.

1.3 Paschen’s curve

Fig. 2 is shows the “Paschen Curve” for air, between two flat parallel copper
electrodes, separated by 1 inch, for pressures between 3x10-2 torr and 760 torr.

Fig. 2
Paschen’s curve

7
 2
Experimental Technique
2.1 Circuit Diagram
The test circuit is shown in Fig. 3. It consists of AC voltage source 220 V, Auto
Transformer (0-220 V), 220V/300 kV transformer, water resistor (RW)=560 kΩ.

Fig. 3
Circuit diagram

2.2 Test Procedures

In the Fig 3 adjust the gap distance to an initial value from 5mm to 80mm.220V
Power Frequency is applied to the Low Voltage side by a Auto Transformer.
Slowly raise the voltage till faint hissing audible sound is heard. This is the
beginning of corona. Hence the Corona Inception Voltage. Raise the Voltage
further till such time there is a faint visible glow at the high voltage electrode. This
is the Visible Corona Inception level. Then slowly reduce voltage further till such
time the hissing sound subside i.e., dies down or becomes extinct. This voltage is
called Corona Extinction Voltage. Once again raises the voltage till such time there
is a Break Down. This voltage is called Breakdown Voltage. Then the measured

8
voltage has been corrected by the Air density correction factor and Humidity
Correction factor. Reduce the voltage completely and open the circuit breaker.
Then further breakdown has been obtained for different gap distances.
2.3 Air Density Correction factor
The air density correction factor, kd is given by:

…………………16
Where,
p = atmospheric pressure under test
t = temperature (in 0C) under test conditions

2.4 Humidity Correction Factors:

The humidity correction factor, kh is given by:

…………………………………17
The constant k is given in Fig 3 as a function of absolute humidity, curve a or b
being applicable according to the type of voltage. The exponents m, n, and w
depend on the type and polarity of the voltage and on the flashover distance d as
given in Table 1 and Fig 5. Lacking more precise information, m and n are
assumed to be equal.

2.5 Standard Reference Atmosphere:

The standard reference atmosphere is :
Temperature to = 20°C
Pressure po = 101.3 kPa(760 mmHg)
Humidity ho.= 11 g water vapor per cubic meter

2.6 According to IEC 60052 (2002)

The breakdown voltage values v (kV) measured under actual conditions with the
temperature t(oC), the pressure p (mmHg) and the absolute humidity k (g/m3) are
reported to standard reference atmosphere as defined by Eq.
V0 =
×
………………….18
9
 Fig 4
Humidity Correction Factor k as a Function of Absolute Humidity

Fig 5
Values of the Exponents m and n for Air Density Correction and w for Humidity Corrections, as a
Function of Spark over Distance, in Meters.

10
 Table 1

Fig 6
Absolute Humidity of Air as a Function of Dry and Wet-Bulb Thermometer Readings

11
 3
WORKDONE SO FAR

3.1 Calculation
Wet condition temperature, Tw= 26 oC
Dry condition temperature, Td= 27 oC
Atmospheric Pressure, p = 76.45 Cm of Hg.
Standard Temperature, to = 20°C
Pressure , po = 101.3 kPa(760 mmHg)
Measured Voltage=V in kV
Humidity , ho= 11 g water vapor per cubic meter.
We are using m=n=w
Air density correction factor, kd = ( ) × ( )

.
kd = ( )×( )
kd=0.9824

Humidity Correction factor, kh =(k)w = k, where w=1

From the Fig 2 and Fig 4, we get Absolute Humidity, k= 0.865.
Therefore, kh = 0.865

Therefore, kd× kh=0.9824×0.865=0.8497

So the actual Breakdown voltage, Vo=

×

12
3.2RESULTS

3.2.1 Breakdown voltage in respect to gap distances of SPHERE-SPHERE

The gap configuration of Sphere to Sphere is shown in the fig 7. The electrodes are
used of two spheres. One sphere put into HV arm and another is grounded. Now
the gap distance vary from 5mm-80 mm and obtained the characteristics of corona
inception voltage, visible corona, corona extinction voltage and Breakdown
Voltage.

Fig 7

gap distances are vary from 5mm-80mm of SPHERE-SPHERE

In table 2 shows the measured values of Sphere to Sphere are obtained and
corrected by Air density correction factor as well as Humidity Correction factor
w.r.t standard temperature and pressure.

Table 2 of sphere-sphere
Measured and Corrected value with respect to gap distances
13
Curve of Sphere to Sphere is shown in the fig 8. The corrected values are put in
the graph w.r.t Gap distance.

Fig 8
Curve fitting of SHPERE-SPHERE

The curve equation has been obtained by Matlab. This curve of Sphere to Sphere is
fitted with 5th degree of polynomial equation. The solution of this equation is
followed by

f(x) = p1*x^5 + p2*x^4 + p3*x^3 + p4*x^2 + p5*x + p6

Coefficients (with 95% confidence bounds):
p1 = -6.222e-008 (-1.331e-007, 8.673e-009)
p2 = 1.413e-005 (-1.265e-007, 2.838e-005)
p3 = -0.001267 (-0.002289, -0.0002442)
p4 = 0.04431 (0.0131, 0.07553)
p5 = 1.382 (1.009, 1.754)
p6 = 0.1895 (-1.139, 1.518)
Goodness of fit: 4.492

3.2.2 Breakdown voltage in respect to gap distances of SPHERE-POINT

The gap configuration of Sphere to Point is shown in the fig 9. The electrodes are
used of one spheres and one point. The Sphere put into HV arm and the point is
grounded. Now the gap distance vary from 5mm-80 mm and obtained the
characteristics of corona inception voltage, visible corona, corona extinction
voltage and Breakdown Voltage.

14
 Fig 9
gap distances are vary from 5mm-80mm of SPHERE-POINT

In table 3 shows the measured values of Sphere to Point are obtained and
corrected by Air density correction factor as well as Humidity Correction factor
w.r.t standard temperature and pressure.

Table 3 of SPHERE- POINT

Measured and Corrected value with respect to gap distances

15
Curve of Sphere to Point is shown in the fig 10. The corrected values of Sphere
to Point are put in the graph w.r.t Gap distance.

Fig 10
Curve fitting of SHPERE-POINT

The curve equation of Sphere to Point has been obtained by Matlab. This curve is
fitted with 5th degree of polynomial equation. The solution of this equation is
followed by

f(x) = p1*x^5 + p2*x^4 + p3*x^3 + p4*x^2 + p5*x + p6

Coefficients (with 95% confidence bounds):
p1 = -1.147e-007 (-1.949e-007, -3.451e-008)
p2 = 2.628e-005 (1.015e-005, 4.241e-005)
p3 = -0.002038 (-0.003195, -0.0008812)
p4 = 0.05771 (0.02238, 0.09303)
p5 = 0.3095 (-0.1123, 0.7313)
p6 = 0.5417 (-0.9621, 2.045)
Goodness of fit: 5.752

3.2.3 Breakdown voltage in respect to gap distances of POINT- SPHERE

The gap configuration of Point to Sphere is shown in the fig 11. The electrodes are
used of one spheres and one point. The Poin is put into HV arm and the Sphere is
grounded. Now the gap distance vary from 5mm-80 mm and obtained the
16
characteristics of corona inception voltage, visible corona, corona extinction
voltage and Breakdown Voltage

Fig 11
gap distances are vary from 5mm-80mm of POINT-SPHERE

In table 4 shows the measured values of Point to Sphere are obtained and
corrected by Air density correction factor as well as Humidity Correction factor
w.r.t standard temperature and pressure.

Table 4 of POINT-SPHERE
Measured and Corrected value with respect to gap distances

17
Curve of Point to Sphere is shown in the fig 12. The corrected values are put in
the graph w.r.t Gap distance

Fig 12
Curve fitting of POINT-SPHERE

The curve equation of Point to Sphere has been obtained by Matlab. This curve is
fitted with 5th degree of polynomial equation. The solution of this equation is
followed by

f(x) = p1*x^5 + p2*x^4 + p3*x^3 + p4*x^2 + p5*x + p6

Coefficients (with 95% confidence bounds):
p1 = -1.607e-008 (-2.587e-007, 2.265e-007)
p2 = 1.59e-007 (-4.861e-005, 4.893e-005)
p3 = 0.0002603 (-0.003239, 0.003759)
p4 = -0.01329 (-0.1201, 0.09353)
p5 = 0.9657 (-0.31, 2.241)
p6 =-0.6048 (-5.153, 3.943)
Goodness of fit:52.61
3.2.4 Breakdown voltage in respect to gap distances of ROD(SQUARE)-
SPHERE

The gap configuration of Rod(square) to Sphere is shown in the fig 13. The
electrodes are used of one Rod of square cross section and another is Sphere. The
Rod is put into HV arm and the Sphere is grounded. Now the gap distance vary
from 5mm-80 mm and obtained the characteristics of corona inception voltage,
visible corona, corona extinction voltage and Breakdown Voltage

18
 Fig 13
gap distances are vary from 5mm-80mm of ROD(SQUARE)-SPHERE

In table 5 shows the measured values of Rod (square) to Sphere are obtained and
corrected by Air density correction factor as well as Humidity Correction factor
w.r.t standard temperature and pressure.

Table 5 of ROD(SQUARE)-SPHERE
Measured and Corrected value with respect to gap distances

Curve is shown in the fig 14. The corrected values of Rod(square) to Sphere are
put in the graph w.r.t Gap distance
3.2.4.2 Curve fitting of rod(square)- sphere

Fig 14
Curve fitting of ROD(SQUARE)-SPHERE

19
The curve equation of Rod(Square)-Sphere has been obtained by Matlab. This curve is
fitted with 5th degree of polynomial equation. The solution of this equation is
followed by

f(x) = p1*x^5 + p2*x^4 + p3*x^3 + p4*x^2 + p5*x + p6

Coefficients (with 95% confidence bounds):
p1 = 7.955e-008 (-2.334e-007, 3.925e-007)
p2 = -1.491e-005 (-7.43e-005, 4.447e-005)
p3 = 0.0009651 (-0.002919, 0.004849)
p4 = -0.02699 (-0.1293, 0.07535)
p5 = 1.021 (0.06245, 1.979)
p6 = 0.03645 (-2.406, 2.479)
Goodness of fit: 0.03829

3.3 Conclusion

In the experiment the Corona Extinction voltage is less than the Corona Inception
voltage. Among four set-up, point-to-sphere, sphere-to-point, sphere-to-sphere,
rod(square)-to-sphere the low breakdown voltage occurred in the set-up of sphere-
to-point. If the air gap is present between the high voltage electrodes is very
small, the Corona Inception, Visible Corona and Corona Extinction voltage
measuring is not possible in physically.

20
 4
Future Work

4.1FutureWork

• To find out the breakdown Voltages with respect to following types of

electrodes as Point-Point, Point-Rod(square), Rod(Square)-Rod(Square),
Rod(circular)-Rod(circular), Rod(circular)-Point.

• To find out the common Model Equation of Breakdown Voltages with respect to
various gap distances.

21
 References

1. IIEE Paper “High Frequency Breakdown Voltage”; by Thanh Duy Chu; March 1992.

2. IEEE paper “Influence of corona discharges on the breakdown voltage of airgaps”; K.

Feser, Dr.-lng; Reprinted from PROC. IEE, Vol. 118, No. 9, SEPT. 1971.

3. IEEE Paper “Evaluation of Humidity Correction Factor of Disruptive Discharge Voltage

of Standard Sphere Air Gaps”;by Osamu Fujii, TakahiroHayakawa, YukioMizuno,
KatsuhikoNaito; EEJTrans 2008; 3:100–105.

4. IEEE Paper “Tests on the breakdown of air at elevated temperatures in non-uniform

electric fields” by N.L.Allen, D.S.K.Lam and D.A.Greaves,in 2nd February 2000.

5. IEEE Paper “Effect Of Humidity On DC Breakdown Voltages In Ambient Air At High

Altitude”by P A Calva Chavarria and A Robledo-Martinez;in San Franicisco, October 20-
23, 1996.

6. IEEE Paper “Corona Inception and Breakdown in No uniform Field with Insulating
Support in Air”;by V. Navinchandra Maller and Krishan D. Srivastava; in January 16,
1987.

7. IEEE Paper “Dielectric Breakdown in Nonuniform Field Air Gaps”;by N. L. Allen and
M. Boutlendj, H. A. Lightfoot; in 1993.

8. IEEE Transactions on “Power Apparatus and Systems”, Vol. PAS-97, No. 6, Nov/Dec 1978.

9. “IEEE Standard Techniques for High-Voltage Testing”;6th edition; by Approved April

26,1982.

10. Book “High Voltage engineering”;M.S Naidu and V.Kamaraju,4th edition, Tata McGraw
Hill Education Private Limited.

11. Html link http://www.sayedsaad.com/High_voltge/insulating_gases/insulating_gases.htm

22