Download - MSc Thesis - University of Miskolc
University of Miskolc
Faculty of Material Science and Engineering
MSc Thesis
Benchabane Ahmed Elmehdi
2018
University of Miskolc
Faculty of Material Science and Engineering
Institute of Physical Metallurgy, Metalforming and Nanotechnology
Effect of surface treatments on the properties of bare
overhead conductors
Made by
Benchabane Ahmed Elmehdi
Consultant/s
Dr. Barkóczy Péter
2018
D E C L A R A T I O N
Benchabane Ahmed Elmehdi (Neptun code: ……, born: ……………, ………….) certifies,
and declares under penalty of perjury, that the submitted essay for a university MSc
degree is my own work effort.
…………., Miskolc
_________________________________
…………………
I have received the declaration:
………………………, Miskolc
__________________________________
Director of institute
ACKNOWLEDGEMENT
First and foremost, I thank God, for giving me the strength, the audacity to overcome the
difficulties, and the will and courage to accomplish this modest work.
I would like to express my gratitude to my supervisor Dr. Péter Barkóczy for the guidance, for his
help, and for the encouragement, he has given me during the thesis work and my internship in FUX
Zrt.
I wish to express my sincere gratitude to Prof. Dr. Valéria Mertinger, the head of the institute of
Physical Metallurgy, Metalforming and Nanotechnology for her constant help, guidance and care
during the whole period of studies at the institute.
To Dr. Katalin Voith for her endless assistance and patience with us during our studies at the
institute, for making it easier for us to do our best during our period of formation.
An enormous appreciation to my professors and lab workers in the institute, who helped and guided
me with their advices and positive critics: Prof. Dr. Kékesi, Dr. Veres, Dr. Gergely, Prof. Dr.
Kaptay, Prof. Krállics and Dr. Dr. Sándor Kovács.
I would like to express my gratitude and love to my country Algeria for providing me a scholarship
for my MSc studies.
I must express my very profound gratitude and love to my parents for their encouragement, love
and for pushing me to become the best person I could be. Also my thanks to my sisters for providing
me with unfailing support and continuous encouragement throughout my years of study.
My sincere thanks to all my friends and close ones especially Walid, Rachid, Jawad, Aleksandra,
Kassab, Aina and Zhanerke. This accomplishment would not have been possible without their help
and their support.
Thank you.
Benchabane A. Elmehdi
ABBREVIATIONS
LWC: Liquid Water Content.
HV: High Voltage line.
AC: Alternating Current.
EMI: Electromagnetic Interference.
AM: Amplitude Modulation.
FM: Frequency Modulation.
CR: Corrosion Rate.
ACCC: Aluminum Conductor Composite Core.
ACSR: Aluminum Conductor Steel Reinforced.
AAAC: All Aluminum Alloyed Conductors.
Content
Abstract p. 1
Introduction p. 2
Chapter 1: Icing of conductors
p. 4
1. Types of atmospheric ice accretion p. 4
1.1. Classifications of ice based on the appearance p. 4
1.2. Classifications of ice based on the formation mechanism p. 4
1.2.1. Precipitation icing p. 4
1.2.2. In-cloud icing p. 8
1.2.3. Sublimation icing p. 8
2. Main affecting factors of icing p. 9
2.1. Influencing factors of icing p. 9
2.2. Attribute classification p. 11
3. Relationship between characteristics of snow accretion and meteorological
parameters p. 11
3.1. Density of accreted snow p. 11
3.2. Snow accretion efficiency p. 12
3.3. Torsional rigidity of wires p. 13
3.4. Shape of sample wire p. 13
3.5. Shedding of accreted snow p. 13
4. Difficulty involved in the study of snow accretion on overhead wires p. 14
4.1. Physical properties of water p. 14
4.2. Rapid metamorphosis of snow p. 15
4.3. Diversity of snow accretion mechanisms p. 15
5. Major hazards of contact wires icing p. 17
5.1. Reduction in performance of contact wires p. 17
5.2. Divergence of the contact wires p. 17
5.3. Produced electric arcs p. 18
5.4. Galloping of the contact wire p. 18
6. Corona loss caused by icing p. 18
6.1. Electrical power system p. 18
6.2. Conductor corona onset gradient p. 19
6.3. The relation between audible noises caused and corona effect p. 20
6.3.1. Audible noise generated from conductors p. 21
6.3.2. Audible noise as a function of rain p. 22
Chapter 2: Corrosion and surface treatments of conductors
1. Surface treatments to increase the hydrophilicity p. 24
1.1. Wettability assessment p. 24
1.1.1. Surface tension and contact angle p. 25
1.1.2. Contact angle hysteresis p. 27
1.2. Increase of surface roughness p. 28
1.3. Coating of the conductors p. 29
2. Corrosion of overhead conductors p. 30
2.1. Corrosion of aluminum p. 30
2.2. Atmospheric corrosion p. 31
2.3. Kinetics of atmospheric corrosion p. 31
Chapter 3: Experimental work (Tests and results)
1. Surface treatments of conductors p. 33
1.1. Hydrophobic coatings of overhead conductors p. 33
1.2. Sandblasting of overhead conductors and contact wires p. 35
2. Wettability assessment on different conductors p. 35
2.1. ACCC conductors p. 35
2.2. ACSR conductors p. 36
2.3. AAAC conductors p. 37
2.4. Aged ACSR conductors p. 37
3. Microscopic investigation of surface treated conductors p. 38
3.1. ACSR conductors p. 38
3.2. AAAC conductors p. 43
4. Corona effect and audible noises of surface treated conductors p. 45
4.1. Corona test p. 45
4.2. Audible noise test p. 46
5. Corrosion of overhead conductors p. 49
6. Heat emissivity of treated conductors and contact wires p. 50
7. Effect of surface treatments on the mechanical properties of conductors and
contact wires p. 51
Summary/Conclusions p. 53
References p. 55
[1]
ABSTRACT
The icing of bare overhead conductors is causing serious problems to the transmission systems of
railway transportations due to several issues such as the sag of the conductors, corrosion of
conductors, and corona discharges, consequently, the emission of audible noises. The economical
need started several studies in order to find a solution for the icing and its effects on overhead
conductors and on contact wires.
This study is an evaluation of a collaboration product between FUX Zrt. and Henkel AG & Co.
KGaA using Bondrite EC2, and how the coating affects the mechanical, electrical, acoustical,
chemical and thermal properties of the bare overhead conductors.
Another side of the study examines the effect of a sandblasting, which is a technique that FUX uses
not only on different overhead conductors but also on contact wires. Where the aim was to
investigate the change in the properties of the products.
The literature shows some models of icing and the effect of emissivity to the icing. The effect of
the emissivity changes on the railway conductors is also analysed.
[2]
INTRODUCTION
The icing is an important matter in the railway catenary conductors because they are not under a
continuous electrical load. Therefore, the risk of icing is larger and it is no possibility of heating
the conductors with electrical load without train approaching. Articles show the effect of the
surface treated bare overhead conductors against icing, but all study analyse the glaze ice accretion.
This is the most dangerous case, because the glaze ice is an insulator layer on the conductor, and
in a case of catenary conductors, the pantograph cannot touch the contact wire.
According to the models from the literature, the glaze ice forms when a water film exists between
the conductor and the ice layer. This happens when the undercooling of the air at the conductor is
not high, the conductor does not consume large heat from the air nor from ice. If the emissivity of
the conductor becomes high, it can consume more heat from the environment. The undercooling
becomes larger, and it could be possible that the liquid film disappears from the surface, then rime
forms on the surface instead of the glaze ice. The analysis of the mentioned models are currently
in progress to calculate the extent of this effect, but the surface treatment can produce a safer
catenary system against icing. Because the pantograph can remove the rime from the conductor,
but the glaze ice can destroy the structure of the pantograph.
Cable producers regularly manufacture surface treated bare conductors. In case of high voltage
bare overhead conductors, this is a usual product. Generally there are two main reasons for the heat
treatment. First is to reduce the audible noise by coronary radiation. Second, is to terminate the
gleaming of the conductor. Both treatments have an importance next to the residential areas where
the two effects disturb people.
In the first case we change the nature of the surface. The as-manufactured conductors have a
hydrophobic surface. The water does not wet the surface but forms big drops on the surface in
rainy, moist or foggy weather. The large electric field can deform the drops, and initiate the
coronary effect. The corona radiation produces an audible noise, which can reach a high level in
larger moisture content of the air [1]. Besides the audible noise, the corona radiation means an
additional loss in the transmission system, which means a comparable amount with the Ohmic loss
in rainy periods. If a surface treatment can reduce the level of the audible noise, it will also reduce
[3]
the corona loss. Therefore, several solutions can be found for this in the literature. FUX develops
a technology which is based on a mechanical treatment without foreign material addition. With this
treatment, the surface of the conductor becomes hydrophilic so when the water nearly perfectly
wets the surface, it won’t form drops on the conductor. So, the audible noise level becomes lower
where it doesn’t disturb people.
In the second case, the aim of the treatment is to avoid the gleaming of the conductor in crossings.
The surface of the as-manufactured bare overhead conductor is shiny and smooth. To obtain the
desired effect, FUX uses a chemical treatment to change the colour of the conductor’s surface to
dark grey (RAL 9700).
It is well known that the same effect occurs due to pollution and corrosion after 15-20 years. The
surface of the conductor becomes grey (not as dark as RAL 9700) and the audible noise also
decreases, but within the mentioned period we have to prove the same properties of the bare
overhead conductors in the mentioned applications.
Both surface treatments change the emissivity of the conductors besides the primary effect. In both
cases the emissivity of the surface increases. The larger emissivity proves larger cooling of the
conductor, therefore, due to enhanced cooling, the same load means a lower operating temperature.
The resistivity of the conductor is smaller at lower operating temperatures, which means smaller
Ohmic loss while taking into consideration the maximal operating temperature because a larger
emissivity means larger current carrying capacity. This value can reach an 11% increment in high
voltage bare overhead ACSR conductors.
The surface treatment of the conductors in railway catenaries is not a usual technique. The main
reason is that there should be a robust electrical connection between the contact wire and the
pantograph. Several researchers have examined this contact where the treatment has been
performed without foreign material addition, and it can be applied in railway conductors, because
this does not have an effect on the electrical connection to the pantograph.
[4]
1. Types of atmospheric ice accretion
Atmospheric icing of structures, that is additionally referred to as icing or ice accretion, could be
a generic term for all sorts of accretion of frozen water substance [1-2]. Often, it happens in cold-
damp regions and it may be classified into three sections, i.e. precipitation icing, sublimation icing
and in-cloud icing [14].
Various processes in which water in numerous forms within the atmospherical freezes and adheres
to objects exposed to the air. The icing is influenced by factors, like temperature, micro- terrain
and micro-climate, wind speed and content of supercooled water in air, etc. Influenced by the
mentioned factors; the icing on overhead conductors is divided into several sections in keeping
with different classifications:
1.1. Classifications of ice based on the appearance
Glaze: transparent vitreum, unbreakable and strong in texture with density between 0.6 and
0.9g/cm3, can also be called ice slush or clear ice which covers wires with strong adhesive power
and never easy to shed.
Granular Rime: ivory opaque with density between 0.1 and 0.3g/cm3, loose and crisp in texture
with air bubble voids inside, sinuous surface and irregular shape.
Crystalline Rime: white crystal with many air bubbles inside, loose and soft in texture with density
between 0.01 and 0.08g/cm3, weak adhesive power on wires and easy to shed.
Wet Snow: ivory or off-white, usually soft in texture with density between 0.1 and 0.7g/cm3. Wet
snow on wires will turn into hard frozen body when the temperature continues to decrease.
Mixed Rime: ivory, large with many voids, is formed by the alternate freezing of glaze and rime
on wire surface, and the density ranges from 0.2 to 0.6g/cm3.
1.2. Classifications of ice based on the formation mechanism
1.2.1. Precipitation icing
Precipitation icing happens in many forms like freezing rain (rime or glaze) and it is caused by
freezing rain (supercooled water) or snowflake falling on the wire surface whose temperature is
near to 0℃ or below. The supercooled degree of water drops has an impact to the dimensions of
water drops; typically the lower degree is, the larger are the drops.
[5]
Dry snow and wet snow rely upon how the precipitation is influenced by variations in a temperature
close to the ground and up to some hundred meters higher than ground. Such icing is noticed
anywhere where freezing temperatures, together with precipitation take place.
Different types of precipitation icing are described below:
Glaze: Once touching the wires, the supercooled water droplets would freeze. A water film would
appear on the wire’s surface as a result of the slow speed of releasing latent heat throughout the
process of freezing, and therefore glaze is created. Glaze caused by precipitation icing is the most
dangerous to overhead conductors due the severe adhesive power and its high density. Freezing
rain usually happens in America, Canada, Russia China, etc., whereas snow covering is a lot
common in Japan and also the Alps.
Glaze includes a density above 900 kg/m3 [3]. Glaze grows in a very clear and sleek structure with
no air bubbles, as shown in Figure 1 It is typically formed from freezing precipitations, rain or
drizzle, or from clouds with big liquid water content and large drop size. The freezing rate of the
droplets is smaller than the impact rate, in which it causes a part of the drop to splash or flow on
the conductor before freezing. Whereas glaze contains no air bubbles per se, in strong wind
conditions, it grows in irregular shapes incorporating pockets of air.
Figure 1: Glaze on a conductor.
[6]
Figure 2: Soft rime on a conductor.
Rime: It includes a density of 300 - 900 kg/m3[4]. It is usually classified as soft rime or hard rime:
Soft rime: includes a density of less than 600 kg/m3. Figure 2 shows soft rime on a conductor in
experimental conditions where it was covered by soft rime which is a white ice deposition that
forms when the water droplets in mist freeze or light freezing fog to the outer surfaces of objects,
with light or calm wind. The fog freezes usually to the windward side of tree branches, wires, or
any other solid objects. Soft rime is similar in appearance to hoar frost; but whereas rime is formed
by vapor, first condensing to liquid droplets (of mist, fog or cloud) and then attaching to the surface,
hoar frost is formed by direct deposition from water vapor to solid ice. Soft rime grows in a pennant
or triangular shape pointed into the wind. The granular structure results from the rate of freezing
of individual drops, each drop freezing completely before another one impinges on the surface.
They are fragile and can be easily shaken off objects. Factors that favor soft rime are small drop
size, slow accretion of liquid water, high degree of super-cooling, and fast dissipation of latent heat
of fusion.
Hard rime: has a density starting from 600 to 900 kg/m3 and tends to grow in a layered structure
with a mixture of clear ice and ice containing air bubbles [5]. During this, the freezing rate of the
droplets is close to the impingement rate. Then, when the water droplets in fog freeze to the outer
surfaces, the white ice forms. It is often seen on high areas in winter, when low hanging clouds
cause freezing fog. This fog freezes to the wind-facing side of conductors, typically with high wind
[7]
velocities and air temperatures between -2 and -8 °C. Hard rime formations are difficult to shake
off, unlike soft rime, which looks feathery or spiky, or clear ice, which looks homogeneous and
transparent. Figure 3 shows accreted hard rime on a conductor.
Figure 3: Hard rime on a conductor.
Both rime types have less density than glaze and cling less doggedly, therefore the damage occurs
because rime is generally minor compared to glaze.
Wet snow: has a density of 300 - 800 kg/m3. It is usually defined as snow which falls at
temperatures equal to or above -5°C. Under these conditions, the snow is sticky enough to adhere
to surfaces easily and accumulate rapidly. Wet snow tends to build on tops and windward surfaces
of structures and in cylindrical layers around conductors [1]. At temperatures below about -2°C,
snow particles are usually too dry to adhere to surfaces in appreciable quantities. If the temperature
falls below 0°C after the accretion of wet snow, the accumulation freezes into a dense hard layer
with strong adhesion. Transmission line problems have occurred due to wet snow events.
Dry snow: Dry snow accretes at subfreezing temperatures [6-7]. This type of accretion appears
only when wind speed is very low i.e. below 2m/s. The density of dry snow is, in general, very
low, not exceeding 100 kg/m3. Hence, the accreted masses in most cases are much lower than the
loads the power lines are designed for.
[8]
1.2.2. In-cloud icing
In-cloud icing occurs only within clouds consisting of super-cooled droplets, which are droplets
that remain liquid at a temperature below 0°C [8]. Ice frozen by the supercooled cloud or fog in
the air on contact with the wires. Without rain or snow, icing mainly depends on humidity, air
velocity along with other factors. It can be formed as long as supercooled water drops exist in small
size, the fog droplets can release the latent heat quickly when freezing. Thus, it won’t form a water
film on wire surface. Therefore, in-cloud icing usually produces rime [14].
The super-cooled water droplets in a cloud or fog freeze immediately upon impact on objects in
the airflow, i.e. overhead lines in mountains above the cloud base. For the structures situated at
mountain summits, exposure to supercooled clouds or fog usually results in soft rime (300 - 600
kg/m3), however, precipitations resulting in hard rime (600 - 900 kg/m3) also occur and are most
frequent in early winter. Sometimes very large in-cloud icing occurs on overhead lines.
1.2.3. Sublimation icing
Hoar frost forms when water vapor in the air freezes directly on the surface of conductors. It is also
called crystalline rime. Formed through sublimation, but doesn’t have a big load because of its
weak adhesion and being easy to shed. Therefore, it won’t pose a big danger to the aerial conductor.
It has a density of less than 300 kg/m3. Figure 4 shows hoar frost that grows on pine needle tips
[9]. Hoar frost as mentioned previously, is generally a deposit of interlocking ice crystals formed
by direct sublimation of water vapor in the air onto objects. It forms when air with a dew point
below freezing temperature is brought to saturation by cooling. Hoar frost is featherlike in
appearance and builds occasionally to large diameters with very little weight. Normally, hoar frost
does not constitute a significant loading problem, however, it is a very good collector of super-
cooled fog or cloud droplets and at subfreezing temperatures with light winds, fog conditions
gradually become soft rime of significant volume and weight. Furthermore, its presence on the
overhead transmission lines can still cause very significant energy losses due to corona discharge
[10].
[9]
Figure 4: Surface hoar plates on pine needle tips.
In general, of all the above-mentioned types of ice accretion, rime and glaze are most dangerous to
the power system. Therefore, this study will deal only with rime and glaze ice accretion.
2. Main affecting factors of icing
Generally talking, icing of aerial conductors is a comprehensive physical phenomenon influenced
by factors such as micro-weather and small terrain changes in temperature, humidity, air
convection, circulation and wind [11]. The size rigidity and geometry of the overhead conductor
may also affect the icing, which, however, are not the premier influencing factors.
2.1. Influencing factors of icing
Micro-climate factors that influence icing mainly include the air temperature, wind speed, wind
direction, and liquid water content in the air, supercool water droplets diameter. These parameters
mainly control the flow and heat transfer in icing process.
Air temperature: Generally, the temperature that may cause icing is -20~0℃, while the most likely
temperature is -6~0℃. If the temperature is too low (below -10°C), supercooled water droplets
turn into snow, then there will be no chance of icing, however, if air humidity reaches the exact
[10]
value, still it is likely to ice despite the temperature is below 0°C, this also happens when the
temperature of supercooled water droplets is high (~1°C) and this will form dense glaze.
Air humidity: The relative humidity is usually more than 85%, not only it is easy to cause icing,
but also to form glaze. If air humidity reaches more than 90%, then the chances are massively
increased.
Wind speed: The possible wind speed for icing is 0~10m/s, at 0~6m/s it is very likely that icing
would occur when the supercooled water droplets touch the conductor’s surface. The wind sends
the supercooled water drops to the wire surface, and the wind speed also affects the heat transfer
in the process of icing. Therefore, the higher the speed of wind is, the more supercooled water
drops can touch the wire surface; the more easily the latent heat released in the process of icing can
be flew away, and the more icing on wire surface will be. In comparison with other factors, wind
speed especially affects the formation of granular rime.
Wind direction: also has certain influence on icing. The wind direction mainly affects the
effectiveness of supercooled water drop delivery. If the wind direction is perpendicular to the wires,
the delivery is the most effective; if the wind direction is parallel to the wires, the delivery is the
least effective, hence less icing.
Liquid water content: exerts an important influence on icing. It will not only influence the speed
of icing but also the types of icing. Generally speaking, if LWC is low and diameter of the droplets
is small, rime is formed and icing develops slowly, whereas if LWC is relatively high and diameter
of the droplets is comparatively big, glaze is formed and icing develops quickly. If the LWC is
very high and diameter of the droplets is very big, rain is formed.
Supercooled water droplets diameter: Supercooled water mainly influences the type of icing.
Droplet which is about 10~40μm will cause glaze; when it comes to rime ice, water droplet is
1~20μm; for mixed rime, the diameter is 5~ 35μm.
Clotting height: clotting height means the air freezing level based on the ground, it is the decisive
factor for high altitude mountain icing.
[11]
2.2. Attribute classification
Based on the information mentioned above; and taking weather, temperature, humidity, wind speed
into account. These four factors restrict each other, and they are strongly connected.
It is possible to categorize the factors into different attributes, the result is listed below:
Weather is divided into: sunny, foggy, rainy, snowy.
Temperature range, it is divided into: very cold (-40~-21℃), cold (-20~ 0℃) two interval
Humidity in 85% - 90% is normal, and large for over 90%.
Wind conditions: when the wind is blowing at the speed of 0~6m/s, we call it breeze, 6~10m/s as
strong breeze.
3. Relationship between characteristics of snow accretion and meteorological
parameters
Using the wind tunnel method, certain conditions and factors take place and affect the icing such
as the temperature of the air temperature, liquid water content, wind velocity along with the
conductor’s parameters such as the external diameter, electrical resistance, surface geometry
(external strands number and diameter).
3.1. Density of accreted snow
From the multiple regression analysis of the results obtained by the wind tunnel experiments, the
relationship between the density of accreted snow and the meteorological parameters is given by
the following equation [12]:
𝜌𝑆 = 0.0671𝑉 − 0.0102𝑉1.1 + 0.0574𝑇 − 0.0107𝑃𝑛 − 0.048 (1)
Where:
𝜌𝑆: Density of accreted snow (g/cm3).
𝑉: Wind speed (m/s)
𝑇: Air temperature (°C)
𝑃𝑛: Amount of snow passing around the wire per unit time (g/cm2.h) = precipitation intensity x
[1+ (wind speed/falling speed of snowflakes)2]1/2.
[12]
The equation implies that, since the temperature is between 0 and 3°C, wind speed affects the
density most significantly. This is quite understandable if we consider that greater snow
compaction takes place with the increased wind speed, and that the wind speed positively affects
the heat transfer from the air to the accreted snow via sensible heat transfer, resulting in melting
the snow. From the equation, it is also noticed that the density of accreted snow decreases with the
increase in precipitation intensity, which can be explained from the decrease in the melting factor
for snow.
3.2. Snow accretion efficiency
All snowflakes that pass around a wire do not accrete to the wire or the surface of the snow that
has already accreted on the wire. Particularly when the wind speed is above 4~5 m/s, the collision
factor can be observed as being close to 1. However, snowflakes collide with a wire and this leads
to breakage, which results some of them accreting on the wire and the rest are carried away by the
wind.
From some results of artificial accretion experiments in which accreted snow takes the shape of a
cylindrical sleeve, efficiency (α) was deduced using the density and the mass of snow accreted per
unit length that were measured soon after the experiments finished. Multiple regression techniques
were also used, with the accretion efficiency as the criterion variable and meteorological
parameters as functional variables, and the following empirical equation was obtained [12]:
𝛼 = 0.624 exp{−0.0865(𝑇 − 3.27)2} × exp(0.621𝑉 − 0.0744𝑃𝑛) (2)
This equation implies that the snow accretion efficiency has a dependency on air temperature
represented as a curve comprised of rising and falling phases with a peak in the middle, i.e., the
efficiency is small at low temperatures, and it increases to a maximum with rising temperatures,
then it turns to a decrease. In addition, the wind speed increases the accretion efficiency while the
amount of snow passing around the wire decrease the efficiency. However, since the amount of
snow that passes around the wire is a function of the wind speed, it is inferred, in a quantitative
sense, that the rate of snow load does not keep increasing with the increasing wind speed, but it
reaches the maximum at the middle wind speeds (about 7~10m/s).
[13]
3.3. Torsional rigidity of wires
The torsional rigidity of the wire has a large effect on the development of snow accretion growth.
As in the case of icing discussed by Poots, the shape of accreted snow/icing on a wire yields along
the span when the torsional rigidity of wire is large.
3.4. Shape of sample wire
The snow accretion growth develops in quite different ways between a multiple strand wire and a
smooth cylinder wire. In the case of the smooth cylinder, a cylindrical snow-sleeve is formed along
a sample wire that is rigidly fixed with no allowance for twisting. Meanwhile, in the case of a
multiple stranded wire, with no twisting, no cylindrical snow-sleeve is formed. Assumedly, this is
because the capillary force, which is the main adhesive force for snow accretion in a non-freezing
condition, is strong against the tensile stress but weak against the shearing stress, as Colbeck points
out. Therefore, the formation of the cylindrical sleeve on a cylinder wire occurs according to the
growing process that includes piling of snow on a wire (first stage), sliding of accreted snow on a
wire, which is horizontally projected, due to the force of gravitation (second stage), and repletion
of the first and second stages. If the cylinder has a slight uneven surface due to deformation,
somewhat like a stranded wire, accreted snow does not slide on wire, and follows the same growing
process as with the stranded wire.
3.5. Shedding of accreted snow
It is important to know the conditions where accreted snow drops off the wire when estimating
snow loads and unequal tensile stresses between conductor spans or phases. At the present time, it
is nearly impossible to quantitatively determine the conditions for accreted snow to drop off the
wire.
There were many cases where a part or all the snow accreted on a wire spontaneously dropped off
in the wind tunnel experiments [12]. However, shedding of snow occurred randomly; accreted
snow was formed into a cylindrical snow-sleeve in one case. In a second case, it was shed off the
wire, and in a third c, it was but partially shed. There was an attempt to find the difference in the
[14]
conditions between causing and not causing shedding, but it failed. Nonetheless, the following
tendencies were found in relation to the parameters:
Accreted snow of smaller density tends to drop off more easily than that of larger density.
When the wind speed is high, the external force, which exerts on accreted snow with the
eccentric weight, is greater. Although with the increase in wire twisting, the accreted snow does
not tend to drop off easily.
The probability that accreted snow drops off the wire does not significantly depend on the air
temperature.
A sample wire with greater torsional rigidity tends to shed the accreted snow more easily than
a sample wire with smaller torsional rigidity.
Once accreted snow completely wraps a wire, it is very unlikely to drop off.
It should be pointed out here that, when accreted snow begins to melt because of air temperature
rise, solar radiation, energizing the wire, etc., the liquid-water produced by melting flows down to
the bottom of the accreted snow due to gravity. This phenomenon is significantly recognizable
particularly with accreted snow of a density of less than 0.6 kg/m3.
These tendencies mean that, once snow accretion has been formed into a cylindrical sleeve, it
hardly drops off spontaneously until the melting substantially progresses. This suggests that
inhibiting the formation of a cylindrical snow-sleeve may be one way of reducing the load of
accreted snow. However, it needs to be noted that the shape of accreted snow, then, tends to be the
same over a span, and galloping is more likely to occur there.
4. Difficulty involved in the study of snow accretion on overhead wires
4.1. Physical properties of water
Water is a substance that is very familiar to human beings, but it has very distinctive properties.
Water can exist in three phases, as a gas, liquid and solid, in a temperature environment where
human beings normally live. The range of phase change between the three states of water is affected
by various causes and is broad. For example, a textbook on rudimentary physics normally discusses
water changing from a liquid state to a solid state at freezing, but it is common knowledge for
[15]
researchers of snow and ice that such a change is hardly a fact. It is also well known that increasing
the pressure on the solid ice can cause it to transform to a liquid and that water vapor exists at very
low temperatures [12].
It should always be taken into consideration that water normally co-exists in these three states and
the phase change occurs between them. The latent heat of water is generally larger than that of
other substances of the same kind. This fact makes snow accretion phenomena more complicated.
For example, snowflakes produced in an upper atmospheric layer with a temperature below
freezing do not melt when they go through a lower atmospheric layer with a temperature above
freezing, but are in a mixed state of solid and liquid, because of the large latent heat of melting.
Such peculiar properties of water make studies of snow accretion complicated.
4.2. Rapid metamorphosis of snow
As a result of the characteristics mentioned in the previous subsection, snow, whether it is a form
of snow accretion or snow cover, changes its physical properties rapidly due to the effects of
ambient temperature, solar radiation and rainfall. This is clearly seen from the fact that the
properties of snow differ (particularly in density, ice particle size, and structure) layer by layer
when digging into snow cover. Snow accretion is a mixture of water (in the solid and liquid states)
and air; and there are a lot of pores in accreted snow, particularly snow with a relatively low density
(< 0.5 kg/m3), where air and liquid-water can pass. This promotes the metamorphosis of snow,
which has a complex effect on snow observations.
4.3. Diversity of snow accretion mechanisms
In order to have snow accretion grow on overhead wires, there must be an adhesive force between
the surface of the wire and the snowflake and between the snowflakes themselves. In icing, the
adhesive forces mainly come from freezing. On the other hand, in snow accretion, the inferred
adhesive mechanisms are many as follows:
(1) Freezing (including pressure melting and re-freezing).
(2) Bonding through freezing of supercooled water droplets existing on the surface of snowflakes.
(3) Sintering.
(4) Condensation and freezing of vapor in the air.
[16]
(5) Mechanical intertwining of snowflakes.
(6) Capillary action due to liquid-water included.
(7) Coherent force between ice particles and water formed through the metamorphosis of
snowflakes.
Through those mechanisms, snow accretion on overhead wires can be produced over a wide range
of air temperatures from 3°𝐶 to – 7°𝐶. The authors once observed a cylindrical snow-sleeve with
a size (diameter) exceeding 50 cm where the air temperature was – 7°𝐶 and the density of accreted
snow was 0.1𝑘𝑔/𝑚3. They expected that, when the air temperature is relative low (< 0°𝐶),
adhesive mechanisms of (1), (2), (3) and (4) are expected to dominate. On the other hand, in the
case of “wet-snow accretion” where the ambient temperature is higher, mechanisms of (6) and (7)
are expected to dominate. One of the grounds for this prediction comes from the following
observations. Firstly, there is a thin layer of ice at the contact of the snow accretion with the wire
when removing the snow accretion on the wire produced under sub-freezing temperatures in natural
conditions. Secondly, the snow piled on a thin wire (having a small torsional rigidity) drops off
spontaneously due to the eccentric weight of accreted snow caused by wire twisting while the snow
piled on a thick wire does not easily drop off but begins to creep in a way to wrap the wire.
The cylindrical snow-sleeve under sub-freezing temperatures, mentioned above, has a larger size
(diameter) but its load is not so large because of its lower density, and therefore, it seldom causes
mechanical damage on modern technology power lines. Thus, this paper will only discuss “wet
snow accretion” at temperatures above freezing.
In the case of “dry snow accretion” at such low temperatures, all the accreted snow tends to fall off
from the wire easily at the same time by swings due to the wind because the adhesive force is small.
The wires then have sleet jump, occasionally causing short circuits between phases. The amount
of such a jump is a result of energy conversion, where the energy accumulated in the wire due to
the increased tension with the increasing snow load is converted into kinetic energy when the
accreted snow drops off. Once a part of the accreted snow starts to fall off, it develops into a total
snow drop-off over the span, often causing a large difference in snow load between spans or phases.
[17]
The difference in sag of each span often causes short circuits between phases of low-voltage power
lines.
5. Major hazards of contact wires icing
5.1. Reduction in performance of contact wires
The added weight to the wires resulting from icing can increase the tension of the wires, causes the
dropper clamp for messenger wire to break, deforms the mid-anchor clamp and damages the swivel
clip holder. Especially in glaze ice conditions, icicles may grow at the bottom of the clamps for
suspension, which may result in the increase of the variation of the elasticity, and lead to the
increase of the vertical contact force between collected strip and contact wire. This may reduce the
performance of the contact wire. This performance includes two meanings: one is the higher
smooth, and the other is the higher kilter. The “higher smooth” means the actual distance between
the contact wire and the orbit plane is smaller than the distance of the ideal parallel state; Second.
The “higher kilter” means that the contact wire itself is flat and straight all the way, so that the
collect strip can run smoothly with a high speed without any blocks. If there are icicles at the
bottom of the contact wires, a hard wear may occur when the collect strip runs. The contact wires
can not meet the ideal smoothness requirement, which is the main reason for being off-line of the
collect strip, volatility of collect current, and flash. The rising rate of electric locomotive collect
strip off-line not only damages the high-speed running locomotive, but also shortens its lifespan
[13].
5.2. Divergence of the contact wires
The changed shape of cross-section after icing makes aerodynamic characteristics of the contact
wire and lines change. Then the added loads under wind may make the contact wire diverge. In
order to make the contact wire wear in a uniform manner in the design, the installation of the
overhead contact wire should be in the shape of a "Z". The "Z" shape pull-value is generally
(480 ± 10) mm. If the contact wire diverges, it may break the smooth contact between the collect
strip and the contact wire. This not only affects the current collection quality, but also speeds up
the wear of collect strip and contact wire.
[18]
5.3. Produced electric arcs
Ice covered on the contact wire surface will reduce the conductivity when current flows from the
contact wire to the collect strip, and produce electric arc, which can burn out the contact wire and
the collect strip. If a heavy icing happened, and it is glaze or hard rime, the thickness of icing is so
big that the contact wire is apart from the collect strip, resulting in no current flow, then the
locomotive has to stop. There is a huge difference between icing of contact wire and icing of high
voltage line. Because as long as the tower won’t collapse and the line isn't broken, it still can
transmit power with ice on lines.
5.4. Galloping of the contact wire
Despite the fact that galloping of the overhead contact wire caused by icing is rare, its hazards are
massive. The damage caused to the parts of contact wire in the area influenced by the galloping
mainly includes the breakage of positive feeder cable, steady arm, and dropper clamp, deformation
of mid-anchor clamp, and the breakage of cantilever insulators, etc.
6. Corona loss caused by icing
6.1. Electrical power system
An important problem facing the early power systems was the efficiency of power distribution
using metallic wires, usually copper, over distances of even a few kilometers [15]. In fact, the
occurrence of heavy power losses in the resistance of the distribution wires at low voltages made
these systems uneconomical, as the need for larger quantities of power and longer distances
increased. For given quantity of power, the use of a higher transmission voltage results in a lower
current in the conductors and therefore in lower power losses and higher efficiency (Figure 5).
[19]
Figure 5: Corona discharge on a 500kV HV line [47].
Corona losses upon transmission lines may reach values worthy of serious consideration of the
designing and operating electrical engineers at or above potentials of 100𝑘𝑉 between wires [16],
depending upon the size and spacing of wires, the weather conditions and the elevation of the line
above sea level. Power loss due to corona atmospheric icing is very complicated and occurs in a
variety of forms as a result of the interplay of numerous physical processes.
6.2. Conductor corona onset gradient
Atmospheric air is probably the most important insulating material used on high voltage
transmission lines [17-18]. At sufficient high levels of conductor surface electric fields, complex
ionization processes take place in the air surrounding high voltage transmission line conductors,
resulting in discharge phenomena known as corona.
Onset voltage is defined as the initiation of a self-sustaining discharge near the conductor, and
occurs when the conductor surface electric field strength exceeds a critical value [19-20]. The
corona onset gradient is a function of the conductor diameter and its surface condition as well as
of the ambient temperature and pressure [17] [21-22]. The corona onset gradient of cylindrical
conductors has been studied experimental and empirical formulas have been developed for
alternating and direct applied voltages. In general, the corona onset gradient 𝐸𝐶 (𝑘𝑉𝑟𝑚𝑠
𝑐𝑚) of a
cylindrical conductor is given as:
𝐸𝐶 = 𝑚𝐸0𝛿(1 +𝐾
√𝛿𝑟𝐶) (3)
[20]
Where 𝐸0 and 𝐾 are empirical constants depending on the nature of the applied voltage, δ is the
relative air density factor (RAD), 𝑟𝐶 is the conductor radius (cm). These values for AC, according
to Peek, are 𝐸0 = 21.1𝑘𝑉𝑟𝑚𝑠/𝑐𝑚 and 𝐾 = 0.308 for the case of concentric cylindrical geometry,
and knowing that:
𝛿 =237+𝑇0
237+𝑇.𝑃
𝑃0 (4)
Where 𝑇 is the temperature (°C) and 𝑇0 is the reference temperature (°C), 𝑃 is the atmospheric
pressure (mmHg) and 𝑃0 is the reference atmospheric pressure (mmHg).
The empirical formula described above was derived from laboratory experiments on smooth
conductors with diameters much smaller than those used on practical transmission lines. The
diameter of the out cylinder was also small in these laboratory studies. However, it was found that
this formula can be extrapolated to practical conductor sizes.
Conductor surface irregularity factor 𝑚 is equal to 1 for ideally smooth and clean conductors [23].
Even microscopic imperfections on the conductor surface tend to reduce the value of 𝑚 below 1
[24].
Practical transmission line conductors are generally of stranded construction, comprised of several
layers of small diameter cylindrical strands. Experimental results have shown that the value of 𝑚
may vary between 0.75 and 0.85 for stranded conductors, depending on the ratio of strand to
conductor diameter. Under the rain and icing conditions, the value of 𝑚 may be reduced by the
water drops, icing treeing, and icicles.
6.3. The relation between audible noises caused and corona effect
The discharge activities associated with the ionisation of the air molecules cause the generation of
acoustic pressure waves. The random pressure waves of different pulses are observed
simultaneously under a conductor power line, which is perceived as frying noise. The noise is
mainly caused by positive streamer corona [25], [26], [27], [28], [29].
The corona effects that are considered in transmission line designs are power loss, electromagnetic
interference and audible noise [25], [26], [27], [29], [30], [31]. However, in modern designs, EMI
[21]
and corona power loss play lesser roles. Corona EMI is predominant at frequencies below 30𝑀𝐻𝑧
[25], [31], [32], [33], [34]. It therefore interferes with amplitude modulation (AM) radio receivers
(0.535– 1.605𝑀𝐻𝑧) [25], [35]. Frequency modulated (FM) radio (88– 108𝑀𝐻𝑧) virtually
replaced AM radio and therefore corona EMI does not affect the general public. EMI by power
lines is mainly caused by distribution lines, in particular, arcing between metal parts on wood poles
[32], [33], [34], [35], [36].
Corona losses are normally very small compared to 𝐼²𝑅 losses, under dry conditions, and are also
of no concern to the average person in the street. Audible noise, on the other hand, has become one
of the main design constants with the introduction of lines of 500𝑘𝑉 and above, including
compacted and double circuit 400 kV lines [25], [26], [27], [38], [39], [40],[41]. It is not easy to
mitigate excessive audible noise after a line has been built.
6.3.1. Audible noise generated from conductors
The discharge activities associated with the ionisation of the air molecules cause the generation of
acoustic pressure waves. The random pressure waves of different pulses are observed
simultaneously under a conductor power line, which is perceived as frying noise. The noise is
mainly caused by positive streamer corona [25], [26], [27], [28], [29].
At higher gradients or under foul weather conditions, negative streamers are also formed. This
means that the frying noise is generated twice, in one power frequency cycle (50𝐻𝑧 or 60𝐻𝑧), at
the positive and negative peak. A pure tone hum is therefore perceived by the human ear at a
frequency double that of the power frequency, i.e. 100𝐻𝑧 or 120𝐻𝑧. This noise component
sounds like transformer hum. (The human ear is not sensitive at 50𝐻𝑧 or 60𝐻𝑧, but can hear
100𝐻𝑧 and 120𝐻𝑧 – Figure 6) Conductor corona therefore produces a frying and a humming
noise.
[22]
Figure 6: Audible noise range according to the human ear sensitivity [47].
6.3.2. Audible noise as a function of rain
The corona activity, and therefore the audible noise produced by a conductor, is dependent on the
weather conditions, mainly rain and rain rate [25], [27], [28], [37], [42], [43], [44], [45]. For this
reason the corona performance is expressed for different rain conditions, namely fair (𝐿50 dry),
wet (𝐿50 Wet) and heavy rain (L5 Wet).
Fair weather: Foul weather refers to conditions where the conductor is subjected to forms of moist,
also referred to as dry condition. The term fair weather only refers to corona activity during
absolute dry conditions and not necessarily to pleasant conditions. Fair conditions exclude rain,
fog snow and ice. When corona cage measurements are performed and noise predictions are made,
the statistical term, 𝐿50 dry, is normally used for fair conditions. The term 𝐿50 dry means that the
noise level referred to, will only be exceeded 50% of the time, under dry conditions. A 𝐿5 dry
level will only be exceeded 5% of the time under fair weather conditions (the rest of the time
(95%) the noise will be below the 𝐿5 level.
[23]
Wet weather: The statistical term 𝐿50 wet is used to quantify noise levels that will be exceeded
50% of the time under measurable rain conditions. In corona cages, 𝐿50 Wet measurements are
made 1 minute after the conductor was sprayed with artificial rain in excess of 7.7𝑚𝑚/ℎ [39].
Heavy Rain: This term refers to rain rates of 7.7𝑚𝑚/ℎ and higher. The statistical term 𝐿5 Wet is
used to classify the noise levels that occur under this condition (the level that will only be exceeded
5% of the time under wet conditions). In corona cages, 𝐿5 Wet measurements are made while the
conductor is sprayed with artificial rain in excess of 7.7𝑚𝑚/ℎ.
[24]
1. Surface treatments to increase the hydrophilicity
1.1. Wettability assessment
An increase of the wettability of the cable is desirable, as it results in a better drip-off of the water
and the drops thus formed are fewer in number and have flatter shapes. This results in a smaller
enhancement of the electric field strength in the vicinity of these drops.
As a drop is sessile on a plane, a closed triple line is formed, where all three phases, air, water and
solid, meet. In the triple line the so-called contact angle is defined by the tangent perpendicular to
the triple line and the surface of the drop and its projection onto the plane (Figure 7). The contact
angle 𝜃 is a measure for wettability or the hydrophilicity of the plane, which originates, as in the
case of the surface tension, from the microscopic forces [48].
But only on perfectly smooth surfaces one single contact angle is given; for real, macroscopic
planes, this is not the case [49]. This is shown by the example of the drop sessile on an inclined
plane in Figure 7, where different contact angles are evident. The range of possible contact angles
is limited by the extremes of the receding and the advancing contact angles, 𝜃𝑟 and 𝜃𝑎 respectively.
Figure 7: Scheme of a water drop slipping on an inclined plane.
The more the contact angles 𝜃𝑎 and 𝜃𝑟 differ, the more the run-off is impeded: in the situation of
Figure 7 the plane could be inclined further until run-off occurs. If a better run-off is desired, the
homogeneity of the surface should be increased [49].
[25]
1.1.1. Surface tension and contact angle
Consider a liquid drop resting on a flat, horizontal solid surface (Figure 8). The contact angle is
defined as the angle formed by the intersection of the liquid-solid interface and the liquid-vapor
interface (geometrically acquired by applying a tangent line from the contact point along the liquid-
vapor interface in the droplet profile). The interface where solid, liquid, and vapor co-exist is
referred to as the “three-phase contact line”. Figure 8 shows that a small contact angle is observed
when the liquid spreads on the surface, while a large contact angle is observed when the liquid
beads on the surface. More specifically, a contact angle less than 90° indicates that wetting of the
surface is favorable, and the fluid will spread over a large area on the surface; while contact angles
greater than 90° generally means that wetting of the surface is unfavorable so the fluid will
minimize its contact with the surface and form a compact liquid droplet. For example, complete
wetting occurs when the contact angle is 0°, as the droplet turns into a flat puddle.
Figure 8: Contact angles formed by sessile liquid drops on a smooth homogeneous solid surface [50].
For superhydrophobic surfaces, water contact angles are usually greater than 150°, showing almost
no contact between the liquid drop and the surface, which can rationalize the “lotus effect” [50].
Furthermore, contact angles are not limited to the liquid-vapor interface on a solid; they are also
applicable to the liquid-liquid interface on a solid.
[26]
Figure 9: Full scale of a lotus leaf with a water drop on it (a), microstructure of the lotus leaf [53].
Ideally, the shape of a liquid droplet is determined by the surface tension of the liquid. In a pure
liquid, each molecule in the bulk is pulled equally in every direction by neighboring liquid
molecules, resulting in a net force of zero. However, the molecules exposed at the surface do not
have neighboring molecules in all directions to provide a balanced net force. Instead, they are
pulled inward by the neighboring molecules (Figure 10), creating an internal pressure. As a result,
the liquid voluntarily contracts its surface area to maintain the lowest surface free energy. From
everyday life, we know that small droplets and bubbles are spherical, which gives the minimum
surface area for a fixed volume. This intermolecular force to contract the surface is called the
surface tension, and it is responsible for the shape of liquid droplets. In practice, external forces
such as gravity deform the droplet; consequently, the contact angle is determined by a combination
of surface tension and external forces (usually gravity). Theoretically, the contact angle is expected
to be characteristic for a given solid-liquid system in a specific environment [50].
[27]
Figure 10: Surface tension caused by the unbalanced forces of liquid molecules at the surface [50].
According to the description of Young [50], the contact angle of a liquid drop on an ideal solid
surface is defined by the mechanical equilibrium of the drop under the action of three interfacial
tensions (Figure 8):
cos 𝜃𝑌 =𝛾𝑠𝑣−𝛾𝑠𝑙
𝛾𝑙𝑣 (5)
Where 𝛾𝑙𝑣 , 𝛾𝑠𝑣, and 𝛾𝑠𝑙 represent the liquid-vapor, solid-vapor, and solid-liquid interfacial tensions,
respectively, and 𝜃𝑌 is the contact angle. (Figure 8) is usually referred to as Young’s equation, and
𝜃𝑌 is Young’s contact angle.
1.1.2. Contact angle hysteresis
From Young’s equation applied to a specific liquid-solid system, three thermodynamic parameters
𝛾𝑙𝑣 , 𝛾𝑠𝑣, and 𝛾𝑠𝑙 determine a single and unique contact angle 𝜃𝑌 . In practice, however, there exist
many metastable states of a droplet on a solid, and the observed contact angles are usually not equal
to 𝜃𝑌. The phenomenon of wetting is more than just a static state. The liquid moves to expose its
fresh surface and to wet the fresh surface of the solid in turn. The measurement of a single static
contact angle to characterize wetting behavior is no longer adequate. If the three-phase contact line
is in actual motion, the contact angle produced is called a “dynamic” contact angle. In particular,
the contact angles formed by expanding and contracting the liquid are referred to as the advancing
contact angle 𝜃𝑎 and the receding contact angle 𝜃𝑟, respectively (Figure 11). These angles fall
within a range, with the advancing angles approaching a maximum value, and the receding angles
approaching a minimum value. Dynamic contact angles can be measured at various rates of speed.
[28]
At a low speed, it should be close or equal to a properly measured static contact angle. The
difference between the advancing angle and the receding angle is called the hysteresis(𝐻):
𝐻 = 𝜃𝑎 − 𝜃𝑟 (6)
Figure 11: Change between the advancing and receding contact angles [50].
The contact angle hysteresis arises from surface roughness and/or heterogeneity. For surfaces that
are not homogeneous, there exist domains that present barriers to the motion of the contact line.
For example, hydrophobic domains will pin the motion of the water front as it advances, causing
an increase in the observed contact angle; the same domains will hold back the contracting motion
of the water front when the water recedes, thus leading to a decrease in the observed contact angle.
1.2. Increase of surface roughness
Surface roughness has a substantial influence on the contact angle and can result in different
deviations of the contact angles reached on a plane consisting of the same material but with
perfectly smooth surface. According to Wenzel [51] the apparent contact angle 𝜃′ formed on rough
surfaces is connected to the contact angle of the ideally smooth surface 𝜃 and the roughness factor
𝑟 using the following equation:
𝑟 =cos𝜃′
cos𝜃 (7)
While the roughness factor 𝑟 is given by:
𝑟 =𝑎𝑐𝑡𝑢𝑎𝑙𝑠𝑢𝑟𝑓𝑎𝑐𝑒
𝑔𝑒𝑜𝑚𝑒𝑡𝑟𝑖𝑐𝑠𝑢𝑟𝑓𝑎𝑐𝑒 (8)
Therefore, 𝑟 of rough surfaces is greater than one. For contact angles 𝜃 > 90°, the apparent angle
is 𝜃′ is greater than 𝜃, in the case of 𝜃 < 90°, the apparent angle 𝜃′ is smaller than 𝜃. Roughness
can be used to increase the wettability of conductors, as for these the case 𝜃′ < 𝜃 < 90° applies.
[29]
1.3. Coating of conductors
Coatings may be either active or passive, depending on their need of external energy or not to start
their operation.
Active coatings require some electrical energy to be active. In general, these methods are applicable
to conductors. However, some semiconducting coatings applicable to insulator surfaces can also
be categorized as active coatings.
Passive coatings are ones which do not require energy to be activated. In the case of
superhydrophobic and icephobic coatings, the low surface energy and surface topography are
among the properties which reduce or prevent adherence of water and ice. Theoretically, icephobic
coatings prevent ice from sticking to the surface because of their anti-adherent properties, while
superhydrophobic coatings do not allow water to remain on the surface because of their repulsive
features. Passive coatings can be classified according to their application field:
- Coatings for insulators.
- Coatings for overhead conductors and ground wires.
- Coatings for other power network equipment.
Hydrophobic coatings
Hydrophobic surfaces have a low surface energy that causes water to bead into distinct water drops
that inhibit the formation of continuous conducting films. In another sense, upon contact with a
hydrophobic surface, water recedes to reduce contact with the surface. By receding, the amount of
water in contact with the surface is reduced to the minimum. The angle formed between the solid
surface and the tangent to the curve of the liquid droplet is referred to as the contact angle explained
previously in the wettability assessment section. The greater the difference between the surface
energy of the substrate and the surface tension of the liquid, the greater the contact angle and the
easier it is to repel liquid droplets.
[30]
2. Corrosion of overhead conductors
2.1. Corrosion of aluminum
Aluminum and its alloys generally exhibit excellent corrosion resistance because they are rapidly
covered by a thin, adherent oxide layer (usually thickness of 2 to 3𝑛𝑚) when exposed to a corrosive
environment at room temperature. The oxide layer is inert and continuous, thus protecting the
underlying metal from further corrosion. Alloying additions such as manganese, magnesium and
copper usually improve the strength of pure aluminum, but may affect its corrosion resistance.
Aluminum and its alloys form an oxide/hydroxide layer which protects them from further
corrosion. Localized corrosion occurs when there is a breakdown of this protective film at discrete
sites, leading to an accelerated attack of the passive aluminum alloy [54, 55]. The selective attack
of the metal surface leads to a higher corrosion rate at such regions than the general rate of the
metal. Differences in environmental and material composition are major causes of this form of
corrosion; the geometry of the structure may also influence the extent of localized corrosion.
Significant among localized corrosion processes are: pitting, crevice corrosion, galvanic attack,
intergranular corrosion and filiform corrosion.
General corrosion is an important type of degradation that affects aluminum alloys. This is a form
of corrosion in which there is a regular, general thinning of the metal. This regular depletion of the
metal surface is facilitated by fairly uniform access of the environment to the metal. For aluminum,
general dissolution spontaneously occurs in most acids or alkalis (except in solutions such as
concentrated nitric acid).
The corrosion behavior of aluminum and its alloys is dependent on the nature of the environment
in which they are employed. Various firms of corrosion ranging from uniform attack to localized
corrosion processes affect these alloys. Critical determinants of the corrosion mechanisms
encountered include the pH of the environment (solution), presence of aggressive ions in the
solution, alloy processing methods and composition of the alloy.
[31]
2.2. Atmospheric corrosion
“Atmospheric corrosion is an electrochemical process in a system that consists of a metallic
material, corrosion products and possibly other deposits, a surface layer of water (often more or
less polluted), and the atmosphere” [56].
The surface layer of water is the electrolyte and usually exists either as a thin, invisible film on the
surface of the metal (damp corrosion) or wet, visible films which exist as a result of rainfall, ocean
spray and other forms of water splashes such as dew (wet corrosion) [57].
These films are constantly affected by cycles of wetting and drying, which may also be influenced
by condensation and evaporation processes. Damp films are formed at a certain critical humidity
level (usually about 60% relative humidity); the composition of this electrolyte depends on the rate
of deposition of atmospheric pollutants, surface roughness and changes in relative humidity [58].
2.3. Kinetics of atmospheric corrosion
Corrosion rates encountered due to the interaction with the atmosphere are expressed either in
terms of the mass lost from samples exposed to the penetration rate. The assessment of corrosion
damage should take into consideration the entire exposed surface area of the sample which gives
the equation for computing the average corrosion rate as:
𝐶𝑅 = (𝐾 ×𝑊)(𝐴 × 𝑇 × 𝐷) (9)
Where 𝐶𝑅 is the corrosion rate.
𝐾 is a constant, whose value determines the units in which 𝐶𝑅 will be expressed in.
𝑊 is the lost weight (𝑔).
𝐴 is the area of the sample (𝑐𝑚2).
𝑇 is the exposure time (ℎ).
𝐷 is the density of the metal (𝑔/𝑐𝑚3).
The values of 𝐾 and corresponding 𝐶𝑅 units are shown in Table 1 below, where the values can be
used on the conversion from one 𝐶𝑅 unit to another.
[32]
Table 1: Corrosion rate units and values of K [59].
For aluminum and some other metals whose corrosion rates do not increase linearly with exposure
time, a power law below has been proposed to fit the data obtained from many studies.
∆𝑊 = 𝐴𝑡𝑛 (10)
Where: ∆𝑊 is the weight loss per unit area due to the corrosion process.
𝐴 is a constant which indicate the mass lost per unit area in one year.
𝑡 is the exposure time in years.
𝑛 is a coefficient which is usually less than 1 for indoor exposures, usually it equals 0.5 (parabolic
rate law).
Corrosion rate Constant K in corrosion rate equation
mm/year 8.76x104
µm/year 8.76x107
g/m2.h 1x104xD
µg/m2.s 2.78x106xD
[33]
TESTS AND RESULTS
The tested conductors and wires in the Laboratory of Wire and Conductor Diagnostics. The
laboratory supervised by jointly FUX and University of Miskolc and situated at the plant of the
company.
In this work we study the effect of coatings of overhead conductors, sandblasting of the conductors
and contact wires, also the effect of aging of conductors on their properties.
The effect of the surface treatment was tested by water spraying, microscopic observation of the
concerned coating on different surfaces, corona test, noise level measurements, corrosion tests,
heat emissivity and ampacity, and mechanical tests.
If the corona discharges, audible noises and corrosion are reduced/eliminated after using the
mentioned surface treatments, along with the increase of the heat emissivity; then, we can say that
the applied techniques are effective and the desired results are reached.
1. Surface treatments of conductors
1.1. Hydrophobic coatings of overhead conductors:
The used coating of bare overhead conductors in the experiment is based on titanium oxide particles
where the product is called Bondrite EC2 formerly called (Alodine EC2). According to Henkel, it
is a 2-15 micron ceramic coating which is a chromium free product specifically formulated for
coating aluminum, titanium and their alloys. The coating is applied by an immersion process, then,
the product is photocatalytically activated, i.e. the coated surface has to be exposed to ultraviolet
irradiation, such as sunlight, to attain the (super-) hydrophilic state. Bondrite EC2 may be used
unpainted in many applications, however, it also provides an excellent base for bonding of
adhesives, porcelain and organic finishes. However, the technique is usually applied on the
automobile industry but recently there was an interest towards applying the coating on overhead
conductors.
[34]
Figure 12: Coated ACCC conductor with titanium oxide layer (below), as-manufactured ACCC cable (above).
Figure 13: Coated ACSR conductor with titanium oxide layer (below), as-manufactured ACSR cable (above).
Figure 14: Coated AAAC conductor with titanium oxide layer (below), as-manufactured AAAC cable (above).
From the first observation of the coated conductors (Figures 12-13-14), it is appearing that the
treated conductors have a darker surface and don’t have the normal shiny substrate in which it
should increase the heat emissivity of the conductor. Also, the coated conductors have rougher
surfaces than the as-manufactured conductors.
[35]
1.2. Sandblasting of overhead conductors and contact wires:
Another solution to icing that has been used for several years by FUX is the sandblasting of both
overhead conductors and contact wires to increase the heat emissivity and ampacity at the same
operating temperature; also, to obtain a hydrophilic surface of the conductor/contact wire. The
surface treatment should also reduce corona discharges and audible noises.
The used sandblasting technique consists on blasting the conductors (Figure 15) and contact wires
with glass beads (silica), of course, taking in consideration the necessary parameters (pressure,
time, angle and distance of blasting). The obtained has a darker color than the as-manufactured
conductor which is supposed to increase the heat emissivity, and has a rough surface that could
decrease the segregation of water droplets, thus, the formation of water film is unlikely to happen
which leads to the prevention of glaze ice formation.
Figure 12: Sandblasted ACSR conductor.
2. Wettability assessment on different conductors
After a certain time with a continuous spraying of the conductors mentioned below, the drops were
visually analysed. After that, the main question was the size and the number of drops.
2.1. ACCC conductors:
Starting with the wettability test of the as-manufactured and the titanium oxide coated ACCC
(Aluminum conductor composite core) conductors. The test consists on a continuous water
spraying of both cables for 10 seconds which has been done in an ambient temperature (~17°C).
Then the observation and the comparison of both surfaces.
[36]
Figure 16: ACCC cables after the wettability testing (coated cable on the right, as-manufactured cable on the left).
After the wetting of both conductors, it can be seen that there were no droplets on the surface of
the coated conductor (Figure 16), contrary to the as-manufactured conductor that has been fully
covered with water droplets as expected. This means that the coating in this case is effective.
2.2. ACSR conductors:
The same process was applied to both coated and as-manufactured ACSR (Aluminum conductor
steel reinforced) conductors characterized by a galvanized steel core and stranded aluminium
wires; where they were continuously sprayed for 10 seconds in an ambient temperature (~17°C)
followed by the observation and the comparison of both surfaces.
Figure 17: ACSR cables after the wettability testing (coated cable on the right, as-manufactured cable on the left).
After the wetting of both conductors, it can be seen that there were extremely small droplets on the
surface of the coated conductor (Figure 17), contrary to the as-manufactured conductor that has
been fully covered with bigger water droplets. This means that the coating in this case is effective.
[37]
2.3. AAAC conductors:
The same process was applied to both titanium oxide coated and as-manufactured AAAC (All
aluminum alloyed conductors) characterized as AlMgSi0.5; where they were continuously sprayed
for 10 seconds in an ambient temperature (~17°C) followed by the observation and the comparison
of both surfaces.
Figure 18: AAAC conductors after the wettability testing (coated on the right, as-manufactured on the left).
After the wetting of both conductors, it can be seen that the surface of the as-manufactured
conductor was covered by small water droplets (Figure 18), on the other side, a part of the coated
conductor’s surface had water droplets on but it was drying fast. This means that the coating in this
case is effective.
2.4. Aged ACSR conductors:
To determine the effect of ageing on the wettability we have applied the same process on a 30 years
aged ACSR (galvanized steel core + stranded Aluminium wires) conductor that has not been treated
(Figure 19). The conductor was continuously sprayed in the same conditions as the previous cables
for the same period of time (10 seconds), then, followed by the observation of the surface.
[38]
Figure 19: a 30 years old ACSR conductor before the wetting.
After wetting the conductor, it can be seen that there were almost no formation of droplets on the
surface (Figure 20). This means that the aging of the conductors is effective.
Figure 20: a 30 years old ACSR conductor after the wetting.
Based on the results, we conclude that the both techniques (titanium oxide coating) and aging
decrease/prevent the wettability of conductors, resulting, the decrease of icing possibility and the
prevention of the occurrence of corona radiations.
3. Microscopic investigation of surface treated conductors
3.1. ACSR conductors:
After the wettability testing of the ACSR conductor, we investigated the nature of the coating by
separating all the wires and collection the powder (Figure 21) then we observe it using the SEM
(Figure 22).
[39]
Figure 21: The separation of the ACSR wires and the collection of the powder from the Aluminium wires and the steel core.
Investigation of the powder:
The collected powder could be AlF3, which is a soft material; meaning that it is unlikely to
face mechanical problem in the presence of this powder in the conductor. Also, the face of its
presence after the wettability proves that it does not dissolve in water.
From the SEM analysis and spectroscopy, we obtained the following results shown in Figure
22, we can see that the powder contains high quantities of Al, Ti, O, F, P.
Figure 22: SEM imagery of the ACSR coating powder (from ACSR coated cable).
[40]
Figure 23: Spectroscopy of the ACSR coating powder.
The spectroscopy of the ACSR powder (Figure 23) shows that the components were: OK (28.26
Wt%), FK (25.95 Wt%), TiK (23.69 Wt%), AlK (13.26 Wt%), PK (8.83 Wt%).
Investigation of surface of the wires:
From the SEM analysis and spectroscopy of the galvanized steel core wires of the ACSR conductor
(Figures 24-a and 24-b), it is clear that the surface of the core wires are deteriorated where the zinc
layer started to demolish because of the coating.
Figure 24-a: ACSR steel wire.
[41]
Figure 24-b: ACSR steel wire.
Figure 25: Spectroscopy of the ACSR wires.
The spectroscopy of the ACSR steel wires (Figure 25) shows that the components were: OK (25.48
Wt%), FK (24.87 Wt%), AlK (21.06 Wt%), TiK (17.57 Wt%), PK (5.58 Wt%).
[42]
From the SEM analysis and spectroscopy of the aluminum cladded steel wires of the ACSR
conductor (Figures 26-a and 26-b), it is clear that aluminum layer is demolished because of the
coating.
Figure 26-a: Aluminum cladded steel of ACSR conductor.
Figure 26-b: Aluminum cladded steel of ACSR conductor.
[43]
Figure 27: Spectroscopy of the aluminum cladded steel wire of the ACSR conductor.
The spectroscopy of the aluminum cladded steel wires of the ACSR conductor (Figure 27) shows
that the components were: OK (21.17 Wt%), TiK (17.48 Wt%), PK (5.5 Wt%), NaK (1.15 Wt%),
AlK (5.96 Wt%), ZnK (41.03 Wt%).
The presence of P and Na could be referred to the lubricant used in the cold drawing process
of the steel wire.
3.2. AAAC conductors:
After the wettability testing of the AAAC conductors, we investigated the nature of the coating by
separating all the wires and collection the powder from the surfaces of the wires (Figure 28) then
we observe it using the SEM (Figure 29).
[44]
Figure 28: The separation of the AAAC wires and the collection of the powder.
Investigation of the powder:
From the SEM analysis and spectroscopy, we obtained the following results shown in Figure 29,
we can see that the powder contains high quantities of Al, Ti, O, F, P.
Figure 29: SEM imagery of the AAAC coating powder.
[45]
Figure 30: Spectroscopy of the AAAC coating powder.
The spectroscopy of the AAAC powder Figure 30 shows that the components were: OK (30.95
Wt%), FK (16.77 Wt%), TiK (32.6 Wt%), AlK (9.73 Wt%), PK (9.95 Wt%) where the main
elements were: Ti, P, Al, O, F respectively.
4. Corona effect and audible noises of surface treated conductors
4.1. Corona test:
As it can be seen on (Figure 31), large drops were formed on the surface of the as-manufactured
conductor. Contrary to the first case, water drops were not formed on the surface treated conductor.
Figure 31: ACSR cables after the wettability testing (sandblasted cable on the right, as-manufactured cable on the left).
The corona radiation test was took place at VEIKI-VNL laboratory where both as-manufactured
and sandblasted ACSR conductors were put to test through the application of several line voltages
[46]
and loads; the measurements of the conductors gave us different values of Ohmic loss along with
corona loss respectively as shown in the Table 2.
Line voltage (kV) Load (MVA) I2R loss (kW/km) Corona loss (kW/km)
362 400 41 26
550 900 52 78
800 2000 93 208 Table 2: Obtained measurements of corona test of as-manufactured ACSR conductor.
From the observation of the corona effect during the test in Figure 32, it is clear that the corona
discharge is non-existent on the surface of the sandblasted conductor while it can be observed on
the surface of the as-manufactured conductor.
Figure 32: Corona radiation test, on the left side is the surface treated conductor.
After obtaining the previous results, we can say that the applied surface treatment (sandblasting)
prevents the occurrence of corona radiations on the surface of the conductors.
4.2. Audible noise test:
Noise level tests and measurements were performed at the Technical University of Graz on four
as-manufactured (Figure 33), and four surface treated (Figure 34) 339-AL1/30-ST1A ACSR
[47]
conductors in different conditions such as voltage, surface condition, temperature of the
conductors; where some conductors were tested in dry conditions and in different voltages while
other conductors were tested in wet conditions equal to rain rate of 6mm/h and for the same
different voltages. The noise levels were recorded and measured using a decibel meter for 2
minutes.
The as-manufactured ACSR conductors were tested as in Figure 33 and as follows:
Two dry conductors were tested in different voltages 220kV and 270kV; meanwhile, the other two
other as-manufactured conductors were tested in a wet condition and for the same voltages
mentioned previously.
Figure 33: Noise level measurements for as-manufactured ACSR conductors.
The surface treated ACSR conductors were tested as in Figure 33 and as follows:
Two dry conductors were tested in different voltages 220kV and 270kV; meanwhile, the other two
other surface treated conductors were tested in a wet condition and for the same voltages mentioned
previously.
[48]
Figure 34: Noise level measurements for sandblasted ACSR conductors.
The results of the noise level measurements after testing all the eight conductors were collected as
average values in Table 3:
Table 3: Noise level measurements for ACSR conductors.
From the obtained results it is proven as explained in the theoretical part that overhead conductors
have low noise levels in dry conditions, but they reach higher levels in larger moisture content and
in high voltages.
Concerning the hearing of noises, the as-manufactured conductors were silent in the dry state and
very loud in the wet state especially at a higher voltage. As for the sandblasted conductors, the
noises were audible in the wet state and at high voltage, however the noise level was much lower
than the case of as-manufactured conductor.
Due to the important reduction in the noise level of the conductors after sandblasting, we can say
that the surface treatment has fulfilled the noise level reduction requirement.
As-manufactured Surface treated
Dry Wet Dry Wet
Voltage (kV) dB(A) dB(A) dB(A) dB(A)
220 25 48 21 43
270 33 54 23 52
[49]
5. Corrosion of overhead conductors
Corrosion test was performed to examine the corrosion resistance of the applied coating. Using two
565-AL1/75-A20SA ACSR conductors; the first is coated using Bondrite EC2 product and the
second used conductor is an as-manufactured ACSR conductor.
The conductors were placed in a corrosion test chamber containing H2O and 5% w/w NaCl with
10 minutes of evaporation and 15 minutes of condensation for a duration of 7 weeks, and in a room
temperature.
From the Figure 35 taken after the corrosion test, it is clearly visible a change in color of the surface
of the as-manufactured conductor compared to the surface for the untested conductor below.
Meanwhile the surface of the coated ACSR conductor seem to be intact.
Figure 35: ACSR conductors: Coated and as manufactured after corrosion test and reference as-manufactured conductor (from
top to bottom).
From the Figure 36, it appears that the powder inside the conductor started to dissolve in a high
moisture content atmosphere, meaning that the steel core could be affected by the corrosion despite
that the outer surface of the conductor appears to be resistant to the corrosion.
[50]
Figure 36: Coated ACSR conductor after corrosion test.
6. Heat emissivity of treated conductors and contact wires
The emissivity of the conductors was measured by thermo-vision where the surface treatment was
interrupted after a certain length. The same surface treatment was applied on an AC-100 contact
wire as in the case of ACSR conductor. The conductors were heated up by current. (Figure 37)
shows the results of the observation, it can be seen that the treated surface gives more emissivity.
Figure 37: Measurement of the emissivity of sandblasted overhead conductor and catenary contact wire.
[51]
The measured results show that the emissivity of the treated surface in both cases is nearly four
times bigger than the as-manufactured surface. The ampacity calculations show that it means 11%
increment in the current carrying capacity. Therefore, the sandblasting of the conductors and
contact wires has fulfilled the requirement of increasing the heat emissivity to prevent the icing.
7. Effect of the applied surface treatments on the properties of overhead
conductors and contact wires
To investigate the efficiency of the coated layer against the sag of the conductors, a tensile test was
performed using an Instron tensile machine on wires from the ACCC coated conductor with the
same titanium oxide layer from Bondrite EC2.
The conductor had an initial length of 250 mm then it was subjected to different elongations 4%,
8%, 12%, 16% respectively as shown in Figure 38.
Figure 38: SEM imagery after the tensile test of coated ACCC wire with different elongations (4%, 8%, 12%, 16% respectively with
a,b,c,d) [University of Debrecen].
[52]
The obtained results have shown that the titanium oxide layer starts to crack starting from a 4%
elongation, along with the propagation of the cracks as the elongation increases before the breaking
at 22% to 25% elongation of the conductor. Therefore, we can say that the applied coating cannot
withstand the tensile stress, but probably because of previous initial deformation that the conductor
has passed through the manufacturing process.
Comparison of the mechanical properties of the wires:
However, the measurements of tensile stress gave close values for the as-manufactured aluminum
wires (198N/mm²) and the coated wires with the titanium oxide (193N/mm²) where the minimum
value was 152N/mm². Meaning, in general, the applied surface treatment does not change the
mechanical properties of the conductors. Knowing that the obtained results were as it follows:
Annealed as-manufactured conductors wires: 71.2N/mm² and 30% elongation. Surface treated
conductors wires: 68.8N/mm² and 28.4% elongation.
Elsewhere, it is proved by mechanical and electrical tests that the properties of the as-manufactured
and the surface treated (sandblasted) contact wires are nearly the same. (Figure 39) shows the
tensile strength of the wires, and the recrystallization behaviour. The strength and the starting
temperature of the softening is the same in both cases. The recrystallization starting temperature
was determined based on 1h isothermal heat treatments.
Figure 39: Comparison of the tensile strength and the recrystallization Ts (1h isothermal heat treatment).
[53]
CONCLUSIONS
In the case of low surface gradients, sandblasting, hydrophilic coatings or aging of overhead
conductors cause substantial reduction of wettability and icing and therefore reduction/prevention
of corona effect and audible noises. In these cases, the titanium oxide coating or sandblasting of
new conductors is recommended.
The accreted snow that has a smaller density tends to drop off more easily than in the case of
bigger density.
When the wind speed is high, the external forces exerting on the accreted snow with the
eccentric weight, are greater.
Audible noises that are caused by corona discharges depend directly on the moisture content
and on the rain rate.
The heat emissivity of the conductors and contact wires surfaces can be increased using
hydrophobic coatings, increasing the surface roughness or aging.
Ice accretion can be reduced by obtaining the proper contact angle to increase water drops run-
off using the wettability assessment.
Due to their increased heat emissivity, aged conductors have bigger possibility to prevent ice
formation on their surfaces.
The surface treatment changes the emissivity of the surface of the conductors. This has a great
effect to the current carrying capacity, and this also has some effect to the ice accretion. As the
results and the properties of the surface treatments of conductor were previously introduced, the
new product’s mechanical and electrical properties are the same as the as-manufactured one.
Due to the emissivity increment, the current carrying capacity is grater in the surface treated
conductors and contact wires. In addition, the emissivity increment has a good effect to the icing,
because the probability of the rime formation is lower than the case of as-manufactured conductors.
The applied coating was successful in the case of ACCC and AAAC conductors but failed in the
case of the both ACSR conductors (galvanized core and the aluminum cladded steel core). The
cause of the failure in the ACSR conductors case requires further investigation.
Concerning the cracking of the titanium oxide layer in the case of ACCC conductor; the
behavior of the layer and its stability in high voltage is still unknown. However we suppose that
it could be harmful, also it could reduce the corrosion resistance of the coating.
[54]
A possible technology was proposed by the FUX and Henkel that suggests treating the whole
conductor after stranding in order to avoid the cracking of the layer; because as in the previous
case, if the wires are treated before the stranding, then the cracking of the coating layer would
happen.
[55]
REFERENCES
[1]. M. Farzaneh, Atmospheric Icing of Power Networks, Springer, 2008.
[2]. P. Fu, Modelling and simulation of the ice accretion process on fixed or rotating cylindrical
objects by the Boundary Element Method, PhD thesis, University of Québec, 2004.
[3]. B. W. Smith, Communication structures, Thomas Telford Publishing, 2007.
[4]. ES ISO 12494:2012, Atmospheric icing of structures, Ethiopian Standards Agency, 2012.
[5]. COST 727, Atmospheric Icing on Structures: Measurements and data collection on icing, State
of the Art, Publication of Meteoswiss, 2006.
[6]. P. Admirat, Y. Sakamoto, “Calibration of a wet snow model on real cases in Japan and France”,
Proceedings of the 4th International Workshop on Atmospheric Icing of Structures (IWAIS 1988),
Paris, France, 1988.
[7]. S. M. Fikke, “Modern meteorology and atmospheric icing”, Proceedings of the 11th
International Workshop on Atmospheric Icing of Structures (IWAIS 2005), Montreal, Canada,
2005.
[8]. B. E. K. Nygaard, J. E. Kristjánsson, and L. Makkonen, “Prediction of In-Cloud Icing
Conditions at Ground Level Using the WRF Model”. Journal of Applied Meteorology and
Climatology, Vol. 50, NO. 12., pp. 2445-2459, 2011.
[9]. C. Shea and B. Jamieson, “The role of moisture in surface hoar growth”, The Journal of
Canada's Avalanche Community, Vol. 88, Issue Spring, pp. 61-64, 2009.
[10]. K. Lahti, M. Lahtinen, and K. Nousiainen, “Transmission Line Corona Losses under Hoar
Frost Conditions”, IEEE Trans. Power Del., Vol. 12, No. 2, pp. 928-933, 1997.
[11]. W. Jian, M. Weiqing, Z. Xiaolong, “A Predicting Method of Power Grid Transmission Line
Icing Based on Decision Tree’s modified model”, Proceedings of the 2nd International Conference
on Computer Science and Electronics Engineering (ICCSEE 2013), Hangzhou, China, 2013.
[56]
[12]. Y. Sakamoto, S. Tachizaki, N. Sudo, “Snow Accretion on Overhead Wires”, Proceedings of
the International Workshop on Atmospheric Icing of Structures (IWAIS XI), Montréal, Canada,
June 2005.
[13]. L. Heyun, G. Xiaosong, T. Wenbin, “Icing and Anti-Icing of Railway Contact Wires”,
Reliability and Safety in Railway, pp. 298-299, March 2012.
[14]. L. Heyun, G. Xiaosong, T. Wenbin, “Icing and Anti-Icing of Railway Contact Wires”,
Reliability and Safety in Railway, pp. 300-301, March 2012.
[15]. R. D. Begamudre, Extra High Voltage AC Transmission, Third edition, New Age
International Publishers, 2006.
[16]. C. F. Harding, “Corona Losses Between Wires at Extra High Voltages- II”, Trans. A.I.E.E.,
vol. XLIII, pp. 1182-1196, 1924.
[17]. P. S. Maruvada, Corona Performance of High-Voltage Transmission Lines. Research Studies
Press, England, pp. 91-96.
[18]. L. Hegy, and G. W. Dunlap, “Corona Loss Vs. Atmospheric Conditions”, Electr. Eng., Vol.
53, No. 2, pp. 272-273, 1934.
[19]. X. L. Jiang, J. Chen, L. C. Shu, J. L. Hu, Z.J. Zhang, and S. J. Wang, “Studying Corona Onset
Characteristics after Rime Ice Accumulation on Energized Stranded Conductors”, IEEE Trans.
Dielectr. Electr. Insul., Vol. 20, No. 5, pp. 1799-1807, 2013.
[20]. L. Chen, J.M.K. MacAlpine, X. M. Bian, L. M. Wang, and Z. C. Guan, “Comparison of
methods for determining corona inception voltages of transmission line conductors”, Journal of
Electrostatics, Vol. 71, No. 3, pp. 269-275, 2013.
[21]. F. W. Peek, “The law of corona and the dielectric strength of air - IV the mechanism of corona
formation and loss”, AIEE Trans., Vol. 46, No. 12, pp. 1009-1024, 1927.
[22]. M. M. Xu, Z. Y. Tan, and K. J. Li, “Modified Peek Formula for Calculating Positive DC
Corona Inception Electric Field under Variable Humidity”, IEEE Trans. Dielectr. Electr. Insul.,
Vol. 19, No. 4, pp. 1377-1382, 2012.
[57]
[23]. K. S. Iyer, and K. P. P. Pillai, “Analysis of irregularity factor of stranded conductors”, Proc.
IEE, Vol. 115, pp. 364-367, 1968.
[24]. M. M. El-Bahy, M. Abouelsaad, N. Abdel-Gawad, and M. Badawi, “Onset voltage of negative
corona on stranded conductors”, J. Phys. D: Appl. Phys., Vol. 40, 3094–3101, 2007.
[25]. EPRI, ”Transmission Line Reference Book: 200 kV and Above” (Third Edition) Palo Alto,
2005.
[26]. Eskom, ”The planning, design & construction of overhead power lines, 132Kv and above”,
Published by Crown Publications cc, Bedford Gardens, Johannesburg, February 2005.
[27]. EPRI, Transmission Line Reference Book: 345 kV and Above (Second Edition), Palo Alto,
1982.
[28]. Otto A. J., ”Direct Current Conductor Corona Modelling and Metrology”, Dissertation
presented for the degree of Doctor of Philosophy in Engineering at Stellenbosch University,
September 2009.
[29]. Maruvada P.S., ”Corona performance of high-voltage transmission lines”, Research Studies
Press LTD., Baldock Hertfordshire, England, 2000.
[30]. Juette, G.W. ; Zaffanella L.E., ”Radio Noise, Audible Noise, and Corona Loss of EHV and
UHV Transmission Lines Under Rain: Predetermination Based on Cage Tests”, IEEE Transactions
on Power Apparatus and Systems, July 1970.
[31]. University of Rome, ”IEEE/PES Special Course Corona and Field Effects of AC and DC
High Voltage Lines”, 1982.
[32]. Cigre Working Group 36.01, ”Guideline on interference produced by corona effect of electric
systems”, 1974.
[33]. Loftness M.O., ”A practical handbook for the correction of radio interference from overhead
power lines”, Bonneville Power Administration, May 1980.
[34]. Roets H.A. & Britten A.C., ”Guidelines for the identification, location and correction of radio
and television interference from high voltage lines”, National Energy Council, Pretoria, South
Africa, 1992.
[58]
[35]. Loftness M.O., ”A practical handbook for the location, prevention and correction of television
interference from overhead power lines”, Bonneville Power Administration, March 1977.
[36]. Roets H.A. & Britten A.C, ”Practical experience with power lines as a source of interference”,
Proceedings of Cigré Conference, Durban, May 1994.
[37]. Task Force, ”Measurement of audible noise from transmission lines”, IEEE Transactions on
Power Apparatus and Systems, Volume PAS-100, Issue 3, pages 1440-1452, March 1981.
[38]. Comber M.G., Cortina R, ”Audible noise generation on individual sub-conductors of
transmission line conductor bundles”, IEEE Transactions on Power Apparatus and Systems,
Volume 95, Issue 2, Pages 525-535, March 1976.
[39]. Britten A.C. and van der Westhuizen C., ”Eskom’s Corona Cage as a Tool for Research into
Corona Phenomena at High Altitudes” Power Industry Technology Trends Conference,
Rosherville, May 1989.
[40]. Jeong-Boo Kim, Dong-Il Lee, Koo-Yong Shin, Jermendy L., Forgarasi I, ”Comparison of
electrical environmental characteristics of different EHV transmission lines”, High Voltage
Engineering, Eleventh International Symposium, Volume1, Pages 262-265, 1999.
[41]. Juette, G.W. ; Zaffanella L.E., ”Radio Noise Currents and Audible Noise on Short Sections
of UHV Bundle Conductors”, IEEE Transactions on Power Apparatus and Systems, Volume PAS-
89, Pages 902-913, May 1970.
[42]. Kolcio N.; DiPlacido J.; Dietrich, F.M. ”Apple Grove 750 KV project - two year statistical
analysis of audible noise from conductors at 775 KV and ambient noise data”, IEEE Transactions
on Power Apparatus and Systems, Volume96, Issue 2, Pages 560-570, March 1977.
[43]. Roets H.A., Tlhatlhetji N.P. and Sibilant G.C., ”Corona Research”, Eskom Research Report
No. RES/RR/02/17511, November 2002.
[44]. Roets H.A., Tlhatlhetji N.P. and Sibilant G.C., ”Corona Research”, Eskom Research Report
No. RES/RR/02/17511, November 2002.
[59]
[45]. Sforzini M., Cortina R., Sacerdote G, Piazza R, ”Acoustic noise caused by AC corona on
conductors: results of an experimental investigation in the anechoic chamber”, IEEE Transactions
on Power Apparatus and Systems, Volume 94, Issue 2, Pages 591-601, March 1975.
[46]. “Corona effect on transmission lines”, www.electricalpowerenergy.com/2018/02/21/corona-
effect
[47]. Brüel & Kjær, “Measuring Sound” Brochure BR 0047-13.
[48]. E.M. Blokhuis, “Surface and Interfacial Tension, Measurement, Theory and Applications”,
Marcel Dekker, Inc., New York, Basel, vol. 119 of the series surfactant science series, pages 153–
155, 2004.
[49]. K. Katoh, “Surface and Interfacial Tension, Measurement, Theory and Applications”, Marcel
Dekker, Inc., New York, Basel, vol. 119 of the series surfactant science series, pages 375–379,
2004.
[50]. G. Bracco, B. Holst, “Surface Science Techniques”, Springer, Berlin, New York, 2013.
[51]. R.N. Wenzel, “Resistance of Solid Surfaces to Wetting by Water”, Industrial and Engineering
Chemistry, vol. 28, no. 8, pages 988–994, 1936.
[52]. P. Barkoczy, “Contact wires – conductivity is of prime importance”, Railway Pro, vol. IX,
No 1-2., 2014.
[53]. T. Wang, L. Chang, B. Hatton, J. Kong, G. Chen, Y. Jia, D. Xiong, C. Wong, “Preparation
and hydrophobicity of biomorphic ZnO/carbon based on a lotus-leaf template”, Materials Science
and Engineering, C 43, pages 310–316, 2014.
[54]. K. Sasaki, G.T. Burstein, “The Generation of Surface Roughness During Slurry Erosion-
Corrosion and its Effect on the Pitting Potential”, Corrosion Science, 38, pages 2111-2120, 1996.
[55]. K. Sasaki, G.T. Burstein, “The Birth of Corrosion Pits as Stimulated by Slurry Erosion”,
Corrosion Science, 42, pages 841-860, 2000.
[56]. E. Bardal, “Corrosion and Protection”, Springer, London, UK, 2004.
[57]. I.S. Cole, J.A.R. Tony, “Atmospheric Corrosion”, “Shreir’s Corrosion”, Elsavier, Oxford,
pages 1051-1093.
[60]
[58]. C. Leygraf, T.E. Graedel, “Atmospheric Corrosion”, Electrochemical Society Series, New
Jersey, 2000.
[59]. P. Enegela, “Ageing of overhead conductors”, University of Manchester, 2013.