Breakdown in Gases

• Ionization and Generation of charged particles in a gaseous medium:

Ionization in gases may be produced by the following processes:

Ionization

Collision Emission from

Photo Thermal
Ionization Sloid Electrode
Ionozation Ionization
Surface

Collision Bombardment Bombardment

Direct Thermal Photo
Step-by-step Direct hit by Step-by-step due to Thermoionic of Electrode of Electrode Field
Collision of Radiation by Electric
Ionization Photon Ionization Thermal Emission Surface by Surface by +ve Emission
Atoms Photons Electrons Ions Emission
Motion

Collision of Collision between

Collision of Electron Electron with two Excited Atom
with Neutral Atom Excited Atom

1. COLLISION IONIZATION:

1.1 Ionization by Direct Collision:

Collisions are of two types: i) Elastic Collision and ii) In-elastic Collision

1.1.1 Elastic Collision:

It is a collision process between two particles (like atoms or molecules) where due to the collision there
is only change in the kinetic energy of the colliding particles but there is no change in the potential
energy. Since change in potential energy of the atom or molecule is an essential prerequisite for the
process of ionization, it is obvious that no ionization is possible due to elastic collision of the gas
particles.

1.1.2 In-elastic Collision:

In-elastic
elastic collision between two particles is the one where total change in potential energy of the system
occurs at the cost of kinetic energy of the particles. There may be some changes
change in the translational
kinetic energy of the system as well, which has relatively lower significance. Therefore,
Therefore in-elastic
collision is primarily responsible for ionization and production of charged particles (ions) in the gaseous
ectron, ion or neutral molecule) of mass m moving with a velocity v collides with
medium. If a particle (electron,

1 EE-805/1 HV Engg. - AD
another neutral atom or molecule, kinetic energy of the moving particle can be transferred to the
second particle. This may increase the potential energy of the second particle and may finally lead to
ionization if the criterion: ≥ holds good, where wi is the energy of ionization for the second
particle.

In-elastic collisions are of two types:

a) In the first type of in-elastic collision the colliding particles are: one neutral atom and an
electron, or two neutral atoms.The following diagrams give illustration of in-elastic collision of
the first type:

_ _

_ _
+ +
Electron Electron
with K.E.

Neutral
Excited
Atom Atom

_ _

Electron with K.E.

_ +
+
Positive Ion
Electron _
with K.E.
Electron with K.E.
Neutral
Atom

_ _ _
_

+ + +
+

Neutral Atom Neutral Neutral Excited

Atom Atom
with K.E. Atom

_ _
_
+
Positive Ion with K.E.
+ +
+
_

Neutral Atom Neutral Neutral Electron with K.E.

with K.E. Atom Atom

2 EE-805/1 HV Engg. - AD
b) In the second type of in-elastic collision, at-least one of the colliding particles are already in a
state of excitation.

_ _

Electron with K.E.

_ +
+
Positive Ion
Electron _
with K.E.
Electron with K.E.
Excited
Atom

_ _ _
+
Positive Ion with K.E.
+ + +
_

Neutral Atom Excited Neutral Electron with K.E.

with K.E. Atom Atom

_ _ _
+
Positive Ion with K.E.
+ + +

Excited Atom Excited Excited Electron with K.E.

with K.E. Atom Atom

1.1.3 Energy Transfer in an Elastic Collision:

Assume two particles 1 and 2 moving with velocities V1 and V2 respectively. Their velocities can be split
into vertical and horizontal components as shown in the diagram:

VY1 V2 VY2
V1

VX1 VX2

Particle - 1 Particle - 2

It is obvious that only horizontal components of their velocities: Vx1 and Vx2 take part in the process of
collision, where as the vertical components VY1 and VY2 have no role in the exchange of energy.
3 EE-805/1 HV Engg. - AD
Now, consider 2 particles:

For Particle-1:

Mass of the Particle = m1

Horizontal component of velocity before collision = v1
Horizontal component of velocity after collision = v2

For Particle-2:

Mass of the Particle = m2

Horizontal component of velocity before collision = 0 (the second particle is initially at rest)
Horizontal component of velocity after collision = V2

Then, from the principle of conservation of momentum:

= +

Again, from the principle of conservation of energy:

1 1 1
= + + −−−−−−−−− 2
2 2 2

Where, U is the final potential energy transformed from the initial kinetic energy.

From eqn. no.1:

−
= −−−−−−−−−−−−−−−− 3

Putting the value of V2 in eqn.2:

1 1 1 −
= + + −−−−−−−−− 4
2 2 2

To obtain the maximum (critical) value of potential energy (U), differentiate eqn.(4) with respect to v2:

1 1 −
0= ×2 −2× × + −−−−−−−−− 5
2 2

For maximum potential energy, dU/dv2 must be zero.

Hence,

−
= −−−−−−−−−−−−−−−− 6

4 EE-805/1 HV Engg. - AD
Or,

−
= −−−−−−−−−−−−−−−− 7

It can be concluded from eqn (7) that:

When m2 -> 0, v2 -> ∞

When, m1 = m2 ,
= = −−−−−−−−−−−−−−−− 8
+ 2

1.2 Step-by-step Ionization

When the density of electrons in a gaseous medium is high, the so called “step-by-step ionization”
can be produced even by slow moving electrons having energy less than wi . The step-by-step
ionization process has great significance in gaseous ionization and gas breakdown. The stepped
ionization can be brought about by the following processes:

i) An electron having energy less than wi brings an atom in excited state. An electron following
immediately after this after colliding with the excited atom imparts to it the difference of
energy to the excited atom necessary for ionization.
ii) During collision of an electron with an excited atom, the excited atom may return to ground
state, thereby imparting its energy to the electron. The electron thus gaining velocity and
energy then cause ionization during its next collision.
iii) When collision of two excited atoms take place, one of it returns to ground state and the
energy thus liberated is spent in ionizing the second atom.

Since all the three processes enumerated above require definite time (ranging from mille-second to
micro-seconds) for their completion, stepped ionization can take place only in those gases, molecules of
which posses meta stable molecules (excited state stability of the order of msec).

While electrons moving very slowly will not produce ionization, very fast moving electrons are also poor
ionizers. Initially with the increase in kinetic energy, the probability of ionization increases and at a
critical value, depending upon the nature of the gas, the maxim probability is achieved. Then any further
increase in kinetic energy reduces the probability of ionization. This is due to the fact that as kinetic
energy increases, velocity also increases and after the critical kinetic energy is exceeded, at such high
velocity the time of contact necessary between the colliding particles for the exchange of energy
becomes less, reducing the probability of ionization. Following fig. shows how the probability of photo
ionization varies with kinetic energy for Hg vapour.

5 EE-805/1 HV Engg. - AD
Fig. Variation of Collision Ionization with Kinetic Energy of Electrons

1.2 Photo Ionization:

Photo-ionization is an important phenomenon which involves formation of free electrons in a gas under
the action of short-wave radiation. The photons present in short-wave radiation are created by the
following processes:

a. Return of excited atoms to ground state and subsequent release of the excess energy in the
form of photons.
b. Photons generated during gas discharge process.

Photons can transfer energy to gas molecules in several ways:

a. It can directly ionize a neutral gas atom, provided it has got energy at least greater than
ionization energy of the gas atoms.
b. Photons may excite atoms of a neytral gas and new photons can bring the excited gas atoms
to ionized condition.
c. A photon can participate in step by step ionization by imparting its energy to another
photon; the later thereby acquiring adequate energy to produce ionization.

The energy of photon available from a source of radiation is given by:

=ℎ =ℎ × −−−−−−−−−−−−−−−−−− 9

Where, h is the Plank’s constant = 6.55 × 10 #

c = Velocity of light = 3×1010 m/sec

ν = Frequency of the radiation

λ =Wavelength of radiation

The essential criterion for photo-ionization is that the energy of the photon in the incident radiation
must be greater than the ‘energy of ionization (wi)’ of the gas. That is:

6 EE-805/1 HV Engg. - AD
=ℎ × ≥ − − − − − − − − − − − − − − − − − − − − − 10

It is known that the vapour of Caesium has the minimum ionization energy wi = 3.88 ev (1 ev = 1.6×10-12
Joule). For Caesium vapour, the ionizing radiation should have the maximum wave length:

×ℎ 3 × 10 $ × 6.55 × 10 #
= = × 10% &'
3.88 × 1.6 × 10

= 3180 &' − − − − − − − − − − − − − − − 11

It is, thus, imperative that radiations having wavelength greater that 3180 Ao shall not be capable of
ionizing a gas. Since λ ≤ 3180 Ao lies in the ultraviolet part of the spectrum, visible light practically can
not cause any ionization.

Photo ionization is enhanced in gases having suspended foreign particles like dust etc. In such cases,
electrons are not only liberated from the ionized gas atoms, but also from the surface of the suspended
solid particles, for which considerably less energy is required.

1.3 Thermal Ionization:

Thermal ionization is attributed by various phenomena which take place under high gas temperature as
enumerated below:

a. Ionization under collision of the molecules of a gas which move at high velocilities at high
temperature.
b. Photo-ionization on account of thermal radiation by heated gas molecules.
c. Ionization under collision of gas molecules with electrons liberated as a result of the first two
processes.

a. Ionization under collision of the gas molecules at high temperature:

According to the kinetic theory of gases, molecule of gases move with all possible velocities; therefore at
any temperature, there exists some probability of ionization during collision between molecules.

b. Photo-ionization on account of thermal radiation:

The probability of ionization on account of thermal radiation at low temperature is low, but with the
increase of temperature, the spectrum of thermal radiation moves in the region of shorter waves and
ionization becomes considerably more probable. The degree of ionization m of a gas is given by the
ration of the number of ionized molecules (Ni) to the total number of molecules (N) in an unit volume is
given by the Saha’s equation:
,-
(× = 2.4 × 10 )
×* .+
× ./ −−−−−−−−−−− 12
1−
Where, T = Temperature in °K

7 EE-805/1 HV Engg. - AD
Wi = energy of ionization in erg

K = Boltzmann’s constant = 1.37×10-16 erg/°K

P = pressure of the gas in mm of Hg

Simultaneously, with the ionization of molecules of a gas, the reverse process of recombination also
takes place. The intensity of ionization is determined only by the temperature of the gas, whereas,
recombination takes place with greater percentage as the number of ionized particles present in an unit
volume becomes more. Therefore, for each temperature, there exists a definite state of equilibrium at
which the number of particles produced and recombined in unit time becomes equal. The variation of
probability of ionization with gas temperature is illustrated in Fig.

Fig. Variation of probability of ionization with gas temperature

1.4 Emission from Solid Electrode Surfaces:

Electron can appear in a gas by way of emission from cathode surfaces. Liberation of an electron from
the metal surface requires a definite amount of energy called ‘energy of liberation’. This energy is
different for the different metals and depends upon the condition of the surface. There are several ways
in which the required energy may be supplied to release the electron:

1.4.1 Thermionic emission:

In metals, at high temperatures, some of the conduction electrons near the surface may posses
sufficient energy to overcome the natural potential barrier that exist at the surface and thus may be
liberated as free electron. The electrons receive their energy from the violent thermal lattice vibration in
the solids at high temperatures. The metal temperature required for thermionic emission may be in the
range 1500∼2500 °K. The emission current density (J) is related to the temperature (T) of the emitter by
the Richardson-Dushman relation:
12-
0 = &* × ./ − − − − − − − − − − − − − − − − − − 13

8 EE-805/1 HV Engg. - AD
Where, J = current density

T = temperature in °K

Øi = work function in Volts

e = electronic charge in Coulomb

K = Boltzmann’s constant = 1.37x10-16 erg/ 0K

The above equation shows that the current density increases with decreasing work function and
increasing temperature.

1.4.2 Electron Emission from Electrode Surface by Electron Bombardment

Electrons can be liberated from solid electrode surface provided and electron with sufficient kinetic
energy hits the solid surface and imparts its kinetic energy to the solid atom in such a way it exceeds the
‘work-function’ of the material:

1
≥ ×∅
2

The emission of electrons from solid surface will enhance with the kinetic energy of the bombarding
electron and is also dependent on the characteristic of the material. However, there is a critical value of
the kinetic energy, above which the emission decreases. This is due to the fact that at higher kinetic
energy, the penetration depth of the electron is more inside the solid, and electrons released from such
depth can’t escape out of the bondage of the material.

1.4.3 Electron Emission from Electrode Surface by Positive Ion Bombardment

Electrons can also be emitted from the metal surfaces by positive ion. Neutral excited (meta stable)
atoms or molecules incident upon the electrode are also capable of ejecting electrons from the surface.
The impinging ion, in such case should release two electrons, one of which neutralizes the ion’s charge.
The 2nd electron is then emitted out of the solid surface. The minimum energy to be acquired by the
positive ion should therefore be twice the work-function of the metal, i.e.:

1
≥ 2× ×∅
2

1.4.4 Photoelectric Emission

Photons, whose energy exceeds the work function (hγ ≥eφ) upon incidence on the cathode surface may
eject electrons from the surface. This can also be effected by irradiating the surface of a metal by short
wave radiations of sufficiently small wave length, so that, E= hγ = hc/λ ≥ eφ.

1.4.5 Field Emission

9 EE-805/1 HV Engg. - AD
Electrons may be drawn out of a metal surface by very high electrostatic field. The field intensity
required to produce emission current of a few microampere is generally of the order: 106 ~108 V/cm for
metals having work function around 4.5 electron-volts. Such high fields are observed at surfaces of very
fine wires, sharp points and other submicroscopic irregularities on the metal surfaces. In field emission
process, the electron emission takes place due to tunneling effect.

Table: Energy of liberation of electrons from different metals.

Metal Energy of liberation (ev)

Al 1.8

Cu 3.9

CuO 5.34

Fe 3.9

Ag 3.1

Pt 3.6

BaO 1.0

1.4.5.1 Enhancement of field emission process by oxide films

When the cathode surfaces is coated by a thin insulating oxide film (thickness ≤ 200A), the positive ions
produced by the initial breakdown, accumulate on the oxide film. The oxide film, accumulating positive
ion charges, thus behaves like a positive electrode (anode) created very close to the cathode. The
separation between the cathode surfaced and the oxide film (anode) being very thin, sets up an intense
field across the gap. Due to the influence of the high electric field, fresh electrons are now liberated
from the underlying metal surface by field emission process. Thus, oxide film coating of cathode
enhances the field emission process many fold.

2 Townsend’s Theory of Gaseous Breakdown

2.1 Growth of current in a gaseous gap under uniform field – Townsend’s experiment

In a gas discharge process, the initiatory electrons originate from cosmic radiations, radioactivity and
photoelectric irradiation of electrode surfaces. In the absence of electric field, there is an equilibrium in
which the rate of generation of electrons and positive ions is balanced by the rate of their decay by the

10 EE-805/1 HV Engg. - AD
process of recombination and straying away. On application of a sufficientlyy high field between the
electrodes, this equilibrium will be upset and there will be a resultant flow of current between the
anode and cathode. The variation of current in a gaseous gap between two parallel allel plate electrodes as a
function of the electric applied field was first studied by Townsend. Townsend used the setup shown in
the Fig 2.1 for his experiment.. The setup comprises of two metal electrodes enclosed in a glass tube
containing some gas. Stabilized
bilized d.c. voltage of high magnitude is applied across the two parallel plates to
setup a high electric field in the gap. The intensity of the field can be varied by varying the d.c. voltage.
The gap current is measured by a micro-ampere meter placed in the circuit. As shown in the Fig, guard
rings are placed around the anode to make the field uniform.

Fig.2.1 Townsend’s Experimental Setup

Observations:

When the applied d.c. voltage is gradually increased from the zero initial value, the gap current cu
characteristics is found to exhibit five distinct stages. The different current stages and the corresponding
ranges of the applied voltage are marked in the Fig. 2.2. Accordingly, the whole observation regarding
the growth of gap current shall be discussed
dis in five stages.

Fig. 2.2 Variation of Gap-current

Gap with Applied Voltage

11 EE-805/1 HV Engg. - AD
Stage I: voltage range 0 to V1

Initially when a small amount of voltage is applied, no current flow is observed in the gap. But, when
the cathode surface is irradiated with the help of an external UV source, current flow in the gap is
established. This current is setup due to the production of secondary electrons from the cathode surface
by UV irradiation. As the voltage is raised, current through the gap increases almost linearly. This is
because the electrons emitted from the cathode move through the gaseous gap with an average
velocity determined by their mobility corresponding to the applied field strength. In this stage, the gap
current I = q v/d.

Stage II: voltage range V1 to V2

After the initial linear increase, the current reaches a saturation value I0, and remains constant at this
value over the voltage range V1 to V2. The current I0 corresponds to the photo-electric emission current
produced at the cathode by external UV irradiation.

Stage III voltage range V2 to V3 (exponential rise of current and Townsend’s first ionization
coefficient - ‘α’ )

At still higher voltages, the current increased above the value I0 at a rate which increases rapidly with
increasing voltage. As the field increases, electrons leaving the cathode surface are accelerated more
and more between two successive collisions. Thus, the electrons gain enough energy to cause ionization
by collision. The additional electrons produced by such collisions also gain energy from the applied
electric field and involve themselves in the mechanism of further successful collision ionizations leading
to production of large no of free electron avalanches.

In order to explain the current growth in the gap, Townsend introduced a factor ‘α’ called the
Townsend’s first ionization coefficient. It is defined as the number of electrons produced by collision
ionization in the path of a single electron traveling a distance of 1 cm in the direction of the applied
field, considering each collision to be 100% effective.

In order to explain the rapid growth of gap current, let us consider that dn be the increase in the
number of electrons over a small distance dx in the gap.

It immediately follows from the definition of ‘α’: the Townsend’s first ionization coefficient, that:

dn = α . n . dx ------------------------------------ (2.1)

Let, n0 = Number of initial electrons leaving the cathode surface per sec.

n = Number of electrons reaching the anode surface after crossing the gap
12 EE-805/1 HV Engg. - AD
d = gap distance

Then, we can write:

6 56 5
46 = 4$ 8. 9
7 6
-------------------------------------- (2.2)

Integrating and putting the limits,

ln( n/n0) = α.d

or, n = n0 eα d -------------------------------------------- (2.3)

This explains that if I0 be the current leaving the cathode, the gap current will increase exponentially in
its path from cathode to anode according to the relation:

I= I0 eα d ------------------------------- (2.4)

But this still does not represent a self sustained discharge, because if I0 = 0, then I = 0; i.e. the gap
current is dependent upon the external UV irradiation. The current I disappears as soon as the U.V
irradiation I0 is withdrawn.

Stage IV: voltage range V3 to V4 (deviation from exponential current rise – current growth under
secondary process)

In order to explain the rapid increase of current beyond voltage V3 and to establish the criterion for self
sustained discharge, let us start with the equation (2.4), i.e. I = I0 eαd ,

or, ln I = ln I0 +αd ------------------------------- (2.5)

According to equation 2.5, the graph of ln I against gap distance d should yield a straight line with slope
α. But in experimental observation of ln I, a deviation from linearity was observed, particularly when
the gap distance d is large. The variation of gap current in the region V3 to V4 is elaborated in Fig. 2.3,
where ln I is plotted against gap distance d under different values of (E/p) ratio (E is the applied electric
field and p is the gas pressure). To explain this deviation from linearity, Townsend postulated that some
secondary mechanisms must be operative, which is affecting the current.

The following postulations were made by Townsend:

i) Liberation of electrons in the gap by collision of gas atoms with the positive ions (β-
mechanism), and
ii) Liberation of electrons from cathode surface by positive ion bombardment (γ- mechanism).

13 EE-805/1 HV Engg. - AD
Townsend introduced the factor ‘γ’ called the Townsend’s second ionization coefficient, which is defined
as the number of electrons released per positive ion bombardment on the cathode surface.

With the above two postulations, equations for the gap current derived in equation 2.5 should be
modified. However, out the two secondary mechanisms (β and γ mechanisms) postulated by Townsend,
the β- mechanism is too insignificant and can be ignored for practical calculations. The modified
equation for current derived in the following section will successfully explain the phenomena of self-
sustained discharge in a gas and shall establish the criterion for spark discharge.

Let,

n0 = Number of electrons emitted from cathode per second by UV irradiation

n = Number of electrons reaching anode per second

n’ = No of electrons emitted from cathode due to positive ion bombardment

γ = No of electrons released per positive ion bombardment on the cathode (Townsend’s 2nd coefficient)

d = gap distance between the two electrodes

Modifying the equation (2.3) to incorporate γ- mechanism , we have:

n= (n0 +n’ ) eαd -------------------------(2.6)

now,

n’ = γ . [number of positive ions bombarding the cathode]

= γ . [no of positive ions created by collision in the gaseous gap]

or, n' = γ . [ n- (n0 +n’)]

or, n’ = γn - γ n0 - γ n’

or, n’(1+γ) = γ (n- n0)

or, n’ = γ (n- n0) / (1+γ) ------------------------------ (2.7)

Substituting n’ in equation (2.6)

n= eαd [n0 + γ(n- n0)/(1+γ)]

or, n = eαd [ (n0 +γ n0+γn -γ n0)/ (1+γ) ]

or, n +nγ = (n0 + γn) eαd

14 EE-805/1 HV Engg. - AD
or, n +nγ - γn eαd = n0 eαd

or, n = n0 eαd /{1- γ ( eαd –1)}

or, in other words, the expression for current can be expressed as:

I= I0 eαd /{1- γ ( eαd – 1)} ------------------- (2.8)

This equation explains the deviation from linearity of the plot of ln I vs. d and also explains the deviation
of the current from its former path in the gap current vs. applied voltage curve of fig2.2 when voltage is
increased beyond V3.

The equation 2.8, however, does not reflect a self sustained discharge, since the gap current I is still
dependent on I0, indicating that the gap current will fall to zero as soon as the external UV irradiation is
withdrawn.

Stage V : voltage above V4 (Townsend’s criterion for self sustained or spark discharge)

In equation (2.8), eαd is very large for large values of gap distance d, and generally, eαd>> 1. Hence in
equation (2.8), ( eαd – 1) ≈ eαd . Thus, equation (2.8) can be modified as:

I= I0 eαd /(1- γ eαd ) ---------------------------- (2.9)

At voltages below V4, when the field E is lower, γ eαd is small compared to unity and, I ≈ I0 eαd . Hence I is
finite, indicating non self sustained discharge. However, in the stage: V since the applied voltage (and
hence the field E) increases, the value of γ eαd approaches 1 and in the limiting condition, when γ eαd =1,
from the equation (2.9), the denominator: (1 - γ eαd) → 0. Under this condition, the gap current,
I = I0/(1- γ eαd ) → ∝

The approach of gap current I towards infinity indicates initiation of a spark discharge. A spark discharge
actually refers to the breakdown of the medium. When a spark discharge begins, the gap current is only
limited by the resistance of the voltage source and the external circuit.

The condition for a spark discharge, that is: (1- γ eαd ) → 0, or in other words, γ eαd = 1 is called the
Townsend’s criterion for gas breakdown. The corresponding value of the supply voltage, which leads to
the breakdown is known as ‘sparking potential’. The critical value of the gap distance d, for which the
above criterion holds true, is called the ‘sparking distance’.

2.2 Physical Significance of Townsend’s Criteria

As per Townsend’s postulate, the secondary ionization coefficient γ is the number of secondary
electrons generated at the cathode per positive ion bombardment. An electron leaving the cathode and
moving a distance d towards anode is multiplied by the factor eαd (α = Townsend’s first ionization
coefficient), leaving behind (eαd – 1) number of positive ions. To continue the process, these positive

15 EE-805/1 HV Engg. - AD
ions on reaching the cathode must release γ (eαd – 1) number of fresh electrons. Now, γ (eαd – 1) = 1
would mean that at least one electron is released from the cathode due to positive ion bombardment to
continue the process of self sustained discharge by collision ionization. Thus, the criterion: γ (eαd – 1) = 1
(or, with some approximation, γ eαd = 1, as eαd >> 1) ensures that the ionization produced by successive
positive ion bombardments will increase the overall number of electrons cumulatively, creating large
number of free electrons (electron avalanches) in the gas, resulting favourable conditions for the spark
discharge to grow rapidly.

2.3 Dependence of Townsend’s First ionization Coefficient (α) on applied electric field and gas
pressure - Derivation of the Relation: α / p = f (E / p)

As discussed earlier, the Townsend’s first ionization coefficient α is defined as the number of ionizations
accomplished by one electron over a path of 1 cm while travelling in the direction of the applied electric
field. In order to determine the mathematical relation by which α depends upon the applied field
intensity and gas pressure, the following assumptions are made:

(a) An electron does not at all ionize a gas if Kinetic Energy of the electron is less than the energy of
ionization Wi of the gas atoms, that is: (K.E.)electron< Wi. On the other hand it ionizes in all cases when
(K.E.)electron ≥ Wi

(b) During each collision an electron gives away all its kinetic energy and then a new run starts from
zero velocity.

(c) The electrons always move along the lines of force of the applied electric field.

With the above assumptions, if an electron travels a path of length x in an electric field of intensity E
without any collisions, the energy stored by it at the end of the path equals to Eqx, where q is charge of
an electron. In order that collision of this electron with a gas molecule would end in ionization, the
following inequality should hold good:

E.q .x ≥ Wi ----------------------------- (1)

Hence, an electron should freely (without collision) cover a path

x ≥ Wi /E.q = Vi /E ----------------------------- (2)

Where, Vi is the ionization potential of the gas.

Let,

n0 = the number of electrons travelling along the direction of x axis from the point x = 0

λ = mean free path of the gas molecules

16 EE-805/1 HV Engg. - AD
n = the number of electrons which do not encounter even a single collision during its flight is from x = 0
to x = λ

Since, mean free path is λ, therefore there should be at least one collision over a length of travel λ.
Considering one collision over a path length λ , there should be dx/λ number of collisions per electron
over the path length dx. But, the number of collisions over the length ‘dx’ should also be equal to the
decrement in the number of particles ‘dn’ from the initial value of n, which do not collide.

Hence we can write:

- dn = n dx/ λ, [since, ndx/ λ = the number of collisions experienced by n electrons over the flight
length dx]

Integrating from over the path length o to x along x, we get:

6 56 < 5<
46 = 4$ − ; λ =
: 6
------------------------ (3)

or, n/n0 = e- x/λ -------------------------- (4)

The quantity e- x/λ denotes the probability that the electron completes a path ‘x’ (or greater) without a
single collision.

Now, total no of collisions experienced by an electron over unit path length is equal to 1/λ . Therefore,
the no of ionizing collisions over a unit path is determined by the product:
- x/λ
α =1/λ e ------------------------------------------ (5)

Though λ depends both on temperature and inversely on pressure of the gas, temperature can be
considered as constant for all practical calculations, where variation of temperature is insignificant.
Thus, considering λ to vary inversely with the gas pressure,

1/ λ = Ap -------------------------------------------------- (6)

Where, A is a constant whose value depends upon the type of gas, the temperature being assumed
constant.

Combining equations (5) and (6),

We get,
- x/λ
α = Ape --------------------------- (7)

substituting, x= Vi/E from equation (2) and, 1/ λ = Ap from equation (6), we get:
-ApV/E
α =Ape ------------------ (8)

Considering, AVi = B, another constant, we get:

17 EE-805/1 HV Engg. - AD
-Bp/E
α = Ape ------------------ (9)
-B/(E/p)
or, α/p = A e

i.e, α/p =f(E/p) ------------------ (10)

It can be concluded from equation (10) that Townsend’s first ionization coefficient, α is a function of the
ratio of the applied electric field to gas pressure (E/p).

2.4 Limitations of the Townsend’s Theory

Townsend’s Theory cannot explain all aspects of break down phenomenon in gases, such as:

1. It cannot explain the short formative-time of a long gap breakdown. According to Townsend’s
theory, for a long gap such as 1m, the formative-time required to arrive at the critical condition
for breakdown would be of the order of mSec, considering electron velocity of 3x105 m/sec. But
in actual practice, breakdown is found to occur within μSec.

2. The Townsend’s Theory cannot explain the forked nature of lightning discharge.

3. According to Townsend’s Theory, the Townsend’s 2nd coefficient γ depends on the cathode
material. But in long gap breakdown, no influence of the electrode material is observed in
practice. Moreover, in the case of lightning discharge the metal electrode cannot be involved at
all, as in lightning phenomena, the charged clouds and ground act as the electrodes.

2.5 Paschen’s Law - Minimum Breakdown Voltage of a Gas

The Paschen’s Law of gas breakdown can be stated as follows:

If the product of gas pressure and electrode separation distance is kept constant, the minimum
breakdown voltage of a gas will remain unaltered. For example, if the gas pressure is doubled, while the
electrode distance is halved, the minimum breakdown voltage of the gap will remain same as before.
The theorem gas extreme importance in the design of insulation of high voltage systems. It has many
useful applications such as optimization of the design of electrical appliances, optimization of the cable
insulation design, design of gas filled PTs and CTs etc, where increase in gas pressure permits reduction
in the gap between electrodes without changing the break down voltage.

2.6 Paschen’s Curve and Critical Breakdown Voltage of a Gas

18 EE-805/1 HV Engg. - AD
For a given gaseous gap, the product of gas pressure and electrode separation (pxd), can be varied by
changing the gas pressure p or varying the electrode distance d. If breakdown voltage of a gas is plotted
against variation of pxd value the plot as shown in Fig2.6 is obtained. In can be observed from the plot
that there exist a a point of minima on the breakdown voltage curve, indicating the m minimum
breakdown voltage (Vbmin ) for the gas.
gas Both on the left hand side as well as on the right hand side of the
minima, the breakdown voltage is found to increase.

Fig.2.6

Since Vbmin corresponds global minima on the plot, hence it can be concluded that there would be no
possibility of a flash over in the gas below the voltage Vbmin for any real value of gas pressure and gap
distance under uniform field. This minimum breakdown voltage is referred to a the ‘critical voltage’ and
the corresponding product of pxd is referred as the (pxd)critical. The typical nature of the breakdown
voltage vs pxd characteristic
ristic of any gas can be explained in the following way. On the right hand side of
(pxd)critical, as pressure is increased, the mean free path λ of the gas decreases.. Therefore the amount of
energy transferred in each collision (eEλ) also decreases, reducing the probability of ionization.
ionization As a
result, breakdown voltage of the gas increases. Conversely, on the left hand side of of (pxd)critical point, as
pressure is decreased, the gas is rarified and distance between moving particles (both gas molecules as
well as free electrons) greatly increases. Thus, the probability of ionization decreases, increasing the
breakdown voltage. For a given electrode distance, it i is therefore possible to increase the breakdown
voltage of a gap either by increasing or by decreasing the pressure from its critical value. For air, the
minimum breakdown voltage Vbmin is 384 volts. Which means, no breakdown is possible in air under any
circumstances, below the voltage of 384 volt.

2.7 Determination of critical (pxd)

(p value and sparking potential (Vbmin) for a gas

It has been derived in section 2.3 equation (9) that,

-Bp/E
α = Ape -------------------- (11)

19 EE-805/1 HV Engg. - AD
where,

p = Gas pressure

α = Townsend’s first ionization coefficient

E =Applied electric field

A, B = constants

Also, as per Townsend’s criterion for spark discharge:

[1 – γ (eαd – 1)]= 0 ----------------------------- (12)

or, γ eαd = 1+ γ

or, eαd = 1 + 1/ γ

or, αd = ln (1 + 1/ γ)

Substituting α, form equation (11), we have

– BP/E
Apd e = ln (1 + 1/ γ) -------------------- (13)

If, Vb be the breakdown voltage of the gap, then

substituting E = Vb/d in equation (13),

-BPd/ Vb
APd e = ln (1 + 1/ γ)

>?5
B@ ln 1 + 1BE
A =
&F

>?5 G?5
B@
A =
HI J BK
or,

>?5 G?5
= LM N P R
@A O6; JQ=
or,

>?5
or, S = ------------------------ (14)
UVW
O6T P [
XY;PZ =
Q

20 EE-805/1 HV Engg. - AD
From the Paschen’s curve illustrated in Fig. it is seen that the slope of the breakdown voltage curve
changes from negative to positive at (pxd)critical. Hence to find out the minimum breakdown voltage Vbmin
at (pxd) critical, dVb/ d(pxd) should be set to zero.

5@A
= 0,
5 ?×5
Considering,

5@A > >?5

= − ]
5 ?×5
=0
UVW
O6T P [ O6T
UVW
[ × ?5
XY;PZ = P
Q XY;PZ =
Q

> >
or, = ]
UVW
O6T P [ O6T
UVW
XY;PZ = P [
Q XY;PZ =
Q

G?5
or, LM N P R=1
O6; J =
Q

G?5
P =
O6; JQ=
or,

1
Hence, F × ^_ ` ^aO = LM ;1 + =
G K
--------------------------- (15)

Substituting the value of (pxd)critical in equation (14), the corresponding minimum breakdown voltage
can be obtained as:
cd
P
U.XY;PZ =
=
>1
LM ;1 + = --------------------- (16)
Q
Sb 6 & 1
LM;1+ =
=
G K
O6T P [
& E
XY;PZ =
Q

3. Use of gases as insulating material

The use of gases as insulating material in electrical equipments offers certain advantages over the
conventional liquid and solid dielectrics. In recent years, attempts have been made to utilize gaseous

21 EE-805/1 HV Engg. - AD
dielectric in circuit breakers, transformers and electric machines. This trend has led to the development
of gases with high dielectric strength. There has been a growing awareness that many compounds in
their gaseous state, particularly those which contain chlorine, fluorine, bromine, oxygen and other
electronegative atoms, have significantly higher dielectric strength. The dielectric superiority of these
materials can be attributed to the facts that:

i) They absorb electrons from the surroundings and become negative ions, which are slow moving.
Typical examples are: a) direct attachment such as: XY + e- → (XY)- or
b) dissociative attachment such as: XY + e- → X + Y-
Where X represents Carbon or Sulphur and Y represents a Halogen (Cl, F, I) eg.SF6, CCl2F2 etc.

ii) The free electrons get attached to the molecules and hence the chances of electron avalanches
leading to electric breakdown is much reduced.
iii) These electronegative molecules also tend to absorb much of the energy of the incident
electrons and thereby restricting chances of collision ionization.
iv) These compounds are usually of higher molecular weight and are too complex to produce
collision ionization.

3.1 Essential property of the gases to be used as dielectric

Gases to be used in electrical equipments as dielectric insulation must satisfy the following basic
requirements:

i) The gas must be chemically insert and should not react with other dielectrics in combination with
which it has been used, and also with other materials used in the construction of the apparatus.
ii) During ionization of the gap, which to a certain degree is always possible in high voltage
apparatus, chemically active substances should not evolve.
iii) The gas must possess a low temperature of liquification or it won’t be suitable neither for
increased pressure applications nor for outdoor applications where temperature may fall leading
to liquification of the gas.
iv) The gas should have high dielectric strength
v) The gas should have high thermal conductivity in cases where it is used simultaneously as insulant
and coolant
vi) The gas must have adequate thermal stability
vii) The gas should be non inflammable
viii) The gas should be physiologically non-toxic
ix) The availability should be good and cost must be sufficiently low so that its application can be
economically justified.

22 EE-805/1 HV Engg. - AD
x) In addition to the above requirements the gaseous insulation used for switchgears (C.B’s) should
also have arc-extinguishing ability, which includes:

a) low dissociation temperature

b) short thermal time constant to enable quick dissipation of heat
c) the gas should not produce conducting products (such as carbon) during arching

3.2 Advantages of gaseous dielectrics

The common advantages of using gaseous insulation in electrical equipments are:

i) their manufacturing processes are simpler and shorter

ii) there are no ‘voids’ and dielectric losses are less
iii) fire risks are eliminated (specially for indoor use)
iv) some saving in weight is achieved

Following are the comparative properties of some of the electro negative gases used for insulation
purpose:

Table: Comparative properties of some of the electro negative gases used as insulant:

Relative Dielectric
Temperature of
Gaseous dielectric Chemical Formula Strengths (taking air
Liquification in 0C
as the reference)

Air 1.0

Hydrogen H2 0.6

Nitrogen N2 1.0 -196

Sulphur Hexafloride SF6 2.5 -64

Freon (Dichloro
CCl2F2 2.5 -30
difloro methane)

Trichloro Fluoro
CCl3F 4.5 +49
Methane

Carbon Tetrachloride CCl4 6.3 +76

Hexafluro ethane C2F6 1.4 -78

23 EE-805/1 HV Engg. - AD
3.3 Suitability of different gases for use in circuit breakers, transformers, switchgears etc.

i) Air: Advantageously used in air blast circuit breakers and cables. Air is abundant in atmosphere.
It can be used at very high pressure. Only problem is the elaborate and costly arrangement
necessary to make the air dry by driving out the moisture. Also ionization of air is accompanied by
emission of ozone and NOx causing intensive corrosion of metallic parts and oxidation of organic
insulation.

ii) H2 : Used as insulator as well as coolant problem is that it is very difficult to seal the hydrogen in a
confined space at high pressure.

iii) N2: It has dielectric strength at good as air- it is cheap & also chemically an inert gas. However
application at high pressure requires complicated construction and also operation becomes more
difficult.

iv) CCl3F and CCl4: Both of them have very good dielectric strength. But due to their relatively high
liquification temperature- they can rarely be used for outdoor applications.

a. CCl4 is a gas at normal temperature and in the gaseous form, under the action of ionization it
decomposes emitting carbon which forms a conducting layer on the surface of solid dielectric
and also chlorine which corrodes metallic parts of the construction.
b. These are the reasons why in spite of having high dielectric strength, CCl3F and CCl4 rarely find
applications as dielectric in electrical switchgears.

v) CCl2F2: Freon has a dielectric strength that is two and a half times that of air. It is widely used as a
dielectric medium. Though being chemically inert, during ionization, it emits chemical substances
giving rise to corrosion. Owing to its low temperature of liquefaction, Freon can be used up to a
pressure of 5 atmospheric. However, Freon is quite costly.

vi) SF6 : SF6 is the most popular and most widely used gaseous insulation used in circuit breakers
owing to its high dielectric strength and excelled arc-quenching capabilities. Due to its low
liquification temperature, the SF6 can be com pressed to a pressure up to 12 – 15 atm. SF6 at a
pressure of 2.5 atm attains a dielectric strength almost equal to transformer oil. SF6 may however,
decompose in to SF4 and F2. The F2 on reacting with metals give metallic fluorides which are again
insulating in nature. Thus, dissociation of SF6 at high temperature and under pressure do not affect
the insulation system much. Moreover, the gaseous decomposition products, which may be toxic in
the presence of moisture, can easily be removed by using activated alumina as an absorber. That is
why, in spite of being costly, SF6 has the widest application as an efficient gaseous insulation.

4. Problems on Gas Breakdown

1. In a laboratory test on a gas a steady state current of 500 µA was established between two parallel
plate electrodes separated by a distance of 0.45 cm when 12kV potential difference is applied
between the electrodes. When the separation between electrodes is reduced to 0.25 cm, keeping

24 EE-805/1 HV Engg. - AD
the applied field constant, the gap current reduces to 50µA. determine Townsend’s first ionization
coefficient(α) from the above test data.

Solution: According to Townsend’s theory, steady state current in the gap between two electrodes is
expressed as:

I = e$ exp (αd) …………………………………………….. (1)

Where, I$ is the initial current at the cathode due to external UV irradiation.

α = Townsend’s first ionization coefficient

d = separation between electrodes

considering the first case I = 500 x 10 g A when d = 0.45 cm

Hence, 500 x 10 g
= I$ e$.)+i ……………………………………… (2)

Since the applied field E = v/d is kept constant,

α = f (E/d) should also remain constant. Thus considering the same value of α for the 2nd case

I = 50 x 10 g
= e$ $. +j
………………………………………………. (3)

Dividing equation (2) by equation (3)

+$$ k $lm $.)+ $. + j

+$ k $lm
=

$. j
Or, =10

Or, 0.28 = ln 10 = 2.3026

Or, 8 = 11.513 ionizing collisions/cm

2. In some laboratory experiment with certain gaseous dielectric, the following readings of gap current
as shown in Table – 1 was recorded between two parallel plate electrodes when electrode
separation was gradually increased from 0.5 mm to 5.0 mm in step of 0.5 mm. keeping applied
electric field constant at 20kV/cm. Determine Townsend’s first and second ionization coefficients (α
and γ respectively) from the given experimental data:

Table 1

Gap distance 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
in mm
Gap current in 2x 5x 7.5x 1.0x 1.5x 2x 5x 1.5x
Amp 10 10 10 10 10 10 10 n 10 %

25 EE-805/1 HV Engg. - AD
Solution: It is observed from the experimental data that there is a moderate increase in current in every
stage when gap distance is increased from 0.5 m upto 3.0 mm in steps of 0.5 mm. This suggests
Townsend’s current growth in absence of any secondary process upto a gap distance of 0.3 mm,
expressed by the equation:

I = e$ exp (αd) ………………………………………. (1)

Considering the current data corresponding to gap distance of 2.0 mm and 2.5 mm, and assuming
constant value for α (since applied field is constant at 20kV/cm)
We can write:
1.0 x 10 = e$ $. j ( d = 2.0 mm = 0.2 cm)
and,
1.5 x 10 = e$ $. +j ( d = 2.5 mm = 0.25 cm)

.+ k $lPP $. + $. j
.$ k $lPP
=

$.$+j
= 1.5

0.05α = ln 1.5 = 0.4055

8 = 8.11 ionizing collision/cm

Townsend’s first ionization coefficient α = 8.11 ionizing collision/cm

It is further observed that there are rapid growth of current in the last two reading, suggesting the
pressure of secondary ionization processes, leading to high growth of current, expressed by the
equation:

o7 1 pW
K 1 pW
I= …………………………………………………… (4)

From equation (3), putting the value of α = 8.11

e$ can be calculated as: 8.11 x 0.25

1.5 x 10 = e$ %. < $. +

1.5 x 10 = e$ x 7.595

e$ = 1.975 x 10 Amp

Now considering the last data for current, i.e I = 1.5 x 10 %

Amp corresponding to d = 0.4 cm,

From equation (4) we can write:

.n#+ k $lP] × 1 q.PP r 7.]s

1.5 x 10 %
K 1 q.PP r 7.t
=

26 EE-805/1 HV Engg. - AD
+.$gu < $lPP
1.5 x 10 %
).gugK
=

1 − 24.636E = 3.3754 x 10 u

E = 40.454 x 10 u

Hence, Townsend’s second ionization coefficient: γ = 40.454 x 10 u

3. In connection with the Paschen’s law, the constant A and B for a certain gas evaluated as: A = 15
and B = 360. Townsend’s experiment on the same gas reveals second ionization coefficient γ =
0.025. Obtain the minimum voltage that has to be applied to cause breakdown in the gas. Also find
the (p x d) critical value for the gas.

Solution: According to Paschen’s law, minimum breakdown voltage for a gas in given by the equation:
>1
Sb 6 =
G K
ln ( 1 + ) …………………………………. (1)

Substituting the value of A, B and γ and putting e = 2.7183

ug$ < .# %u
Sb 6 =
+
ln ( 1 +
$.$ +
)

= 242.27 volt

Hence minimum 242.27 volt must be applied to cause breakdown in the gas.

(p x d) critical value for the gas is obtained as:

1
F9 ^_ ` ^aO = G ln ( 1 + K
)

.# %u
×
+ $.$ +
= ln ( 1 + )

= 0.673 Torr-cm

4. The Paschen’s constant for air are A = 12, B = 365 and Townsend’s 2nd coefficient γ = 0.02.
(i) Determine minimum breakdown voltage for air.
(ii) What distance between the electrodes has to be maintained for breakdown to occur at this
voltage when pressure is maintained at 760 Torr?
(iii) If the pressure is maintained at 760 Torr, at what voltage breakdown will occur at 20$ C when
gap between the electrodes is changed to 1 cm?

27 EE-805/1 HV Engg. - AD
Solution: According to Paschen’s law:
>1
Sb 6 =
G
ln ( 1 +
K
)…………………(1)

(i) Putting the value of A, B and γ and e = 2.7183,

ug+ < .# %u
Sb 6 = ln ( 1 +
$.$
)
= 325 volts
So minimum breakdown voltage for air is 325 volt

(ii) According to Paschen’s law :

1 .# %u
F× ^_ ` ^aO G K $.$
= ln ( 1 + )= ln ( 1 + ) = 0.89 Torr-cm

For breakdown to occur at 325 volt, (p x d) must be 0.89 Torr-cm

If p = 760 Torr, then gap distance d = 0.89/760 cm = 0.00117 cm = 1.17 x 10 u

cm.

Thus gap between electrodes must be 1.17 x 10 u

cm for breakdown to occur at 325 volt at a pressure
of 760 Torr.

Based on experimental observations, breakdown voltage of air is expressed as:

nu ?5 nu ?5
S ] + 6.08 [ ] /
#g$/ #g$/
= 24.22 [

Where p = pressure in Torr

d = gap distance in cm

T = absolute temperature in $y

Putting p = 760 Torr, T = (273 + 20) $y and d = 1 cm

293 ×760 ×1 293 ×760 × 1 1/2
S = 24.22 [ 760 ×293
] + 6.08 [ 760 ×293
]

= 24.22 + 6.08 x 1

= 30 kV

Hence breakdown will occur at 30 kV when gap distance is 1 cm.

28 EE-805/1 HV Engg. - AD