Download Investigation into possible mechanisms of light pollution flashover of 275kV transmission lines as a PDF

TitleInvestigation into possible mechanisms of light pollution flashover of 275kV transmission lines as a
File Size1.6 MB
Total Pages87
Table of Contents
Table of contents
List of Abbreviations
List of Figures
List of Tables
A Review of Insulator Flashover Processes
Laboratory Investigations
The effect of a spark gap in series with a polluted insulator string
Appendix A: Type of glass insulator used in tests
Appendix B: 275kV tower
Appendix C: Pollution test methods
Document Text Contents
Page 1

Investigation into Possible Mechanisms of Light

Pollution Flashover of 275kV Transmission Lines as

a Cause of Unknown Outages

Kevin Kleinhans

Thesis presented in fulfilment of the requirements for the degree of

Master of Engineering at the University of Stellenbosch


Dr. J.P. Holtzhausen

April 2005

Page 2


I, the undersigned, hereby declare that the work contained in this thesis is my own

original work and that I have not previously in its entirety or in part submitted it at

any university for a degree.




Page 43

Chapter 3

much less than the breakdown of air (10kV/cm). This however confirms that at

higher voltages this could lead to the breakdown of air and resultant localised

flashover across these clean discs. Electric field inside fog chamber

It was decided to use the same setup for the electric field measurements in the

fog chamber of the University of Stellenbosch. The fog chamber has a rating

of 22kV rms, but it has a stronger source. Flashover was still very unlikely,

even under 100% relative humidity. The setup used is shown in Figure 3.11.

The spherical probe was protected from excessive exposure to humidity by

covering it with a Latex covering, which would keep it water tight, yet have no

adverse effect on the electric field measurements.

Figure 3.11: Schematic diagram of fog chamber setup




Remote reading AC fieldmeter

Fibre optic cable

Plane of movement

Spherical electric
field meter probe

Fog chamber


4 disc glass insulator string

The humidity inside the fog chamber caused excessive condensation on the

surface of the spherical probe. This distorted the electric field in the

immediate vicinity of the probe, and thus caused erratic value fluctuations.

Due to the sensitivity of the electric field probe, no accurate and coherent

readings could be taken of the electric field strength.


Page 44

Chapter 3

3.3 The effect of two clean discs in a lightly-polluted 16-disc I-string

The next logical step was to investigate the effect of a non-uniform distribution of a

conductive surface layer along the length of a 275 kV cap and pin insulator string

[Kleinhans 1999]. The worst case is when one or two clean discs are inserted in a

polluted string. The electric field and the actual voltage distribution are measured to

compare a polluted string with a clean string. The test environment used is the same

as in Figure 3.1.

3.3.1 Laboratory tests

For the following experiments, the test procedures were as follows:

A test transformer, rated at 350kV, was used. It had high source impedance,

and therefore did not comply with the source requirements for pollution tests.

The object of these tests, however was not to do pollution tests, but to

investigate the voltage and electric field distribution along the length of the

insulator string.

The type of insulator used in these experiments was the U120BS cap and pin

insulator string.

Two clean insulator discs were inserted in three different positions in the

insulator string: at the top, in the middle and at the bottom of the string. This

was done to see whether the positioning of the clean discs in the insulator

string held any significance with respect to the flashover voltage, as well as

the voltage and electric field distribution along the insulator string.

For these tests, two types of conducting layer were used as before to simulate

the effect of the conductivity of the layer on the process:

1. Aluminium foil (which effectively meant that the rest of the string was

shorted out).

2. A light pollution solution, which was sprayed onto the glass surface of the

insulators. The light pollution solution had an ESDD of 0.05 mg/cm2.


Page 86

Appendix C: Pollution test methods

materials with high viscosity and creates a thicker conducting layer that can remain in

a stable, wet condition for a long period of time.

In all tests done during this research study, the former mixture was implemented to

obtain the test results.

With both types of slurry, a relatively high flashover strength is found. The wet-

contaminant method has an advantage over other methods because of its simplicity,

ease of use, and low test cost.

The Clean-fog test method

This method may be separated into two types. The first type, also known as the pre-

deposit method [IEEE Standard 4-1978], insulators are contaminated, dried, and then

wetted by clean fog. Test voltage is applied to the insulators when leakage resistance

has reached its lowest value. Although the procedure is usually classified as a clean-

fog test, it should be considered as a variation of the wet-contamination test.

In the second type, voltage is applied to dry contaminated insulators, and then a

wetting condition is applied. This procedure is more commonly regarded as a

reasonable simulation for natural conditions; however, it is more complicated than the

other methods [Kawai 1968 and Kawai Milone 1969]. The wetting condition is

usually achieved using steam generated by evaporation of water from open containers.

The process of wetting the insulator surface is a key factor in this kind of test. In the

slow wetting condition of the steam-fog tests, the surface impedance gradually

changes according to the net rate of wetting, which are a function of the humidity in

the air and the drying effect of leakage current. This phenomenon would not appear if

the rate of wetting were fast enough to overwhelm the drying effect of leakage


Comparison of test methods

It is important to understand how contamination test methods differ from each other.

Each test method essentially simulates a different phenomenon. A factor that is

important for one method may not be significant for other methods. For example, the

wetting condition is the most important factor in the clean-fog tests, but not in the

wet-contamination tests.

There is no direct relationship between salt-fog, clean-fog and wet-contamination

methods, and one single method cannot simulate the breakdown phenomena created


Page 87

Appendix C: Pollution test methods

by the others. For practical designs, it is extremely important to choose the test

method that would simulate the particular natural condition found in service.

This was, however, quite difficult seeing that an unknown flashover method is being

studied, and only the symptoms of the flashover are known.

In most cases, tests were performed during this research using an array of the above

test methods, in order to simulate the symptoms shown by the unknown flashovers

experienced in the Eskom Transmission network.

An outline of each method is shown in Table C.1.

Test method Contamination

Time or
flashover or
test duration

Surface condition
during test


Salt-fog Contamination and
wetting are applied
simultaneously by
spraying salt water.
Degree of
contamination defined
by amount of salt in

Testing time is
1 hour.

Nearly pure resistive
impedance due to
the conducting layer
on the surface.

Flashover strength
linear with string

1. Light

Defined by amount of
dry contaminants.

Flashover after
20-30 seconds.

Wet changing to
dry, resistive, and
also capacitive
impedance in the
half-dry condition.

Small regular
discs perform
better than large

2. Heavy

Defined by surface

within 4-5

Wet also at instant
of flashover,
resistance close to
the initial value.

Flashover strength
proportional to
leakage distance.

1. Quick

Defined by amount of
dry contaminant.

Test time is 30-
60 minutes.
usually within
30 minutes

Dry changing to
wet, nearly resistive
impedance due to
the fast wetting

Linear; leakage
distance is the
main factor.

2. Slow

Defined by amount of
dry contaminant.

Test time is 2
usually after 1

Dry changing to
wet, capacitive and
resistive impedance
designated as
dynamic impedance.

Some non-
linearity; large
discs are better
than small disks.

Table C.1: Comparison of pollution test methods


Similer Documents