…………………… ……………..Transformers ………………….………….

BSc. in Electrical Engineering

ELE3114 Electrical Machines and Drives I’

1 Properties of Insulating Materials & Liquid Breakdown Mechanisms

1.1 Breakdown of liquid insulating materials

In most applications insulating liquids, besides insulating voltage-carrying parts, have to satisfy further
requirements such as to cool the windings and cores in transformers, to extinguish the arc in circuit
breakers, or, as an impregnating medium in capacitors, to increase the dielectric constant of the paper
dielectric. The behaviour of technical insulating liquids in the electric field differs fundamentally from
that of gases and solids. It is critically governed by impurities, by the ageing condition as well as by space
charges. As a consequence of this there is no unified breakdown theory, even though certain mechanisms
are beyond doubt. Only these shall be discussed briefly here, whereas the effect of refining and ageing
on the breakdown performance, particularly critical in insulating oils, will be discussed in detail in Section
2.

1.1.1 Electric strength of technical configurations with insulating liquids

Mineral oils are predominantly used as liquid insulating materials, namely for open insulating paths in
conjunction with insulating supports, as well as for impregnation of laminated materials, especially
soft paper and pressboard. Very often they are mineral oils of low viscosity, designated transformer oils;
their viscosity, which is strongly temperature dependent, is matched to the application purpose
(cooling, impregnation) by mixing with suitable distillates.

As a result of their extraction from crude oil, insulating oils are a mixture of several hydrocarbons with
different properties. By contact with the atmosphere, insulating oils also contain impurities in the form
of dissolved gases, liquids (e.g. water, acids), as well as conducting and non-conducting particles (e.g.
fibre pieces, sludge). Experience shows that the presence of these impurities determines the
breakdown behaviour in practice much more than the properties of the ideally pure insulating liquid
itself. In fact, only liquefied gases represent really pure insulating liquids; for very low temperatures liquid
nitrogen (LN2) or liquid helium may become a practicable proposition for future applications in low
temperature technology.

The following table contains guiding values for some of the properties of insulating liquids:

Transformer oil Liquid nitrogen

Density in g/cm3 0.9 0.8

r 2.3 1.4

Besides depending upon the impurities, the electric strength also depends upon several other
parameters, particularly upon pressure and the stress duration. During impulse voltage stressing the
breakdown field strength of a configuration is many times the value for alternating voltages; in a
homogeneous field in practical insulations one may expect values up to Ed = 200 kV/cm. The impulse
voltage-time curve of an electrode configuration in transformer oil reproduced in Fig. 1-1 gives an idea
of the effect of stress duration.

Typical for breakdown measurements in insulating liquids with impurities is large dispersion and the
occurrence of irregular pre-discharges, even in a homogeneous field. Moreover, in pure liquid gaps
breakdowns can occur with subsequent self-healing. Fig. 1-2 shows the result of measurements of the
breakdown field strength Ed and the dissipation factor tan at 50 Hz as a function of water content v.

1
The reduction of Ed on exceeding v = 5010-6 (50ppm) can be attributed to the fact that a transition
from solution to emulsion takes place under room temperature. For a breakdown field strength of at
least 200kV/cm a residual water content of v < 10-5 (10ppm) must be ensured. Water/Moisture content in
oil is normally represented in ppm (parts per million) or mg/kg or g/g.

Fig 1 impulse voltage –time band of the rod-rod electrode configuration in oil for negative impulse voltage

In contrast to dissolved water vapour, dissolved gases have in general no effect upon the electric strength
of insulating liquids, apart from the ageing processes due to oxygen. However, the condition of

Fig 2 breakdown field strength and dissipation factor of transformer oil as a function of water contents
(mass fraction)

supersaturation could be critical if beyond the equilibrium condition dissolved gases appear in the
form of tiny bubbles released by mechanical vibration (forced circulation cooling) or by high electric
field strength.
Liquid insulation materials are used as impregnants in the dielectric of capacitors, in the soft paper
and pressboard insulation of transformers, as well as in oil-impregnated paper cables. Here very high
electric strengths are reached but at the cost of effective convection cooling. The following table gives
a few guide values for physical properties at 20C.

oil/paper oil/pressboard

Density in g/cm3 1.1 1.2

r 3.6 4.5

2
The electric strength of these mixed dielectrics is so high that continuous operating field strengths of
l00kV/cm and above can be permitted. For short-duration stressing values of Ed up to 1 MV/cm can be
measured. Especially at high ambient temperatures however, the possibility of a thermal breakdown must be
considered.

1.1.2 Breakdown mechanisms

The breakdown of liquids cannot be described by a single unified theory and there is also strong
dependence of the observed phenomena on the technical boundary conditions. Hence, only the two most
important types of breakdown will be discussed here.

a) Intrinsic breakdown

As in gases and solid insulating materials, an avalanche breakdown is also possible in very pure liquids. If
the ionization condition according to

E >Ui

is assumed, reduction of the mean free path  due to the liquefaction of the gas should be compensated by
a corresponding increase in the ionization field strength. An estimate for the example LN2 shows that
this theoretically desirable value of the field strength cannot ever be achieved in an experiment. Rather,
one should expect the kinetic energy of the electrons, even for the experimentally determined values of
the field strength, to be sufficient to effect partial vaporisation of the liquid by way of collision
processes with the liquid molecules. In the small gas bubbles so formed, on account of the larger mean
free path , the prevailing field strength can initiate collision ionization and avalanche formation. This model,
in conformity with the measured results for LN2, results in breakdown field strengths of about 300kV/cm.

In technical insulating oils more complicated mechanisms may be assumed as a consequence of the
presence of different components. In an electrode configuration with an insulating liquid in a dc field,
a current density S appears which only reaches a nearly constant value after a few minutes. The cause
of this phenomenon is the presence of charge carriers of different mobility. Finally, the steady current is
determined by the heavy electrolytic ions which are formed by dissociation. For low field strengths Ohm's
law is approximately valid, until a saturation current sets in, as in gases. If the field strength E is
increased further, the current increases disproportionately until breakdown occurs. According to the
expression for the current density

S = qnbE (q = ionic charge)

An increase in the charge carrier density n may be expected from this since there is no reason for the
mobility b to change. The presumption that, similar to a breakdown in liquefied gases, the charge
carrier multiplication may be attributed to collision ionization in the gaseous or vapour part of the liquid,
leads to the description of this breakdown as a "masked gaseous breakdown". This model also explains the
experimentally observed increase in the electric strength of insulating liquids with pressure. (refer to the gas
breakdown mechanisms).

b) Breakdown by fibre bridge formation

Technical insulating liquids always contain macroscopic impurities in the form of fibrous particles of
cellulose, cotton or other materials. Especially when these particles have absorbed moisture from the
insulating liquid, forces act upon them moving them into the zone of higher field strength and also
aligning them in the direction of the field. In contrast to the fibre particles, gas bubbles, because of
their lower dielectric constant, are removed from the region of highest field strength. In this manner, a
fibre bridge between the electrodes can result, and this represents a conducting canal. The resistance
loss can result in vaporization of the moisture contained in the particles with subsequent gas breakdown.
This phenomenon can also be interpreted as a local thermal breakdown originating from a weakly con-

3
ducting canal. The technically extremely important occurrence of a breakdown by the formation of
fibre bridges prohibits high electric stressing of free oil gaps. Fibre bridge breakdown can be
effectively prevented by insulating screens which should be arranged perpendicular to the electric field
if possible. A further effective measure is to embed the electrodes in a solid insulating material,
preferably using a paper bandage.

Fig 4 formation of a fibre bridge

Fig 3 current density in insulating oils in the dc

field
(a) as a function of the time after switching on
(b) as a function of the field strength

4
2 Natural Organic Insulating Materials
Organic materials are characterized by the element carbon which has the ability to form long chains or
ring-type structures. Mineral and vegetable oils belong to the group of natural organic liquid materials;
paraffin and bitumen as crude-oil products belong to the solid group, in addition so are wax, resins,
wood and fibrous materials like paper, silk, cotton and jute. Important to high-voltage technology are
mineral oil and paper.

2.1 Mineral oil

Mineral oil is obtained by fractional distillation of crude oil after degassing, dehydrating and desalination
of the raw product. Mainly saturated hydrocarbons with naphthene or alkane structure are employed
since they are chemically more stable than the unsaturated aromatic hydrocarbons. A few examples of the
structures are given below:

pentane (alkane): CH3-CH2-CH2-CH2-CH3

cyclohexane (naphthene): benzene (aromatic):

CH2 CH

CH2 CH2 CH CH

CH2 CH2 CH CH

CH2 CH

Removal of the undesirable aromatic hydrocarbons is achieved by a process, subsequent to the

distillation, referred to as refining.

The electrical properties of insulating oil deteriorate with increasing water and gas content. Insulating
oil must therefore be pre-treated before its use in high-voltage equipment. This is done in refining plants
for degassing and desiccation, as shown in Fig 2.6. Surface degassing is usually adopted where a thinly
flowing film of large area is produced. During this procedure the oil is exposed to a temperature of 50
to 60°C in a vacuum of about 10-2 mbar,

5
 Fig.2.6

Insulating oil should have a breakdown voltage of 50...60kV with the electrodes shown in Fig. 2-2, this
corresponds to a breakdown field strength of about 200 kV/cm.

Insulating oil is subject to ageing risk due to absorption of moisture, the solution of gas, impurities and, in
particular, oxidation. Under the combined effect of oxygen and heat, oxidation products are formed which
are soluble in oil, e.g. acids, and insoluble components which appear as sludge.

The oxidation of oil is accelerated by the catalytic action of copper, which is the reason why bare copper
conductors must be avoided in insulating oil. The neutralization number and the saponification number are
useful to characterize the ageing state, the former gives the quantity of potassim hydroxide (KOH)
necessary to neutralize the free acids contained in 1g of oil, while the latter describes the quantity of
KOH which neutralizes the free and bound acids, and so incorporates the neutralization number as well.

The dissipation factor of an aged oil is about ten times worse than that of new oil. Large oil-insulated
apparatus, e.g. transformers, must therefore be regularly checked for the ageing condition of the
insulating oil. For this purpose oil samples are withdrawn and examined for their breakdown field
strength, dissipation factor and impurities. If necessary, refinement must be undertaken on site or total
replacement of the oil carried out. Replacement of the oil is recommended when the neutralization number
exceeds 0.5 mg KOH/g oil, or when sludge soluble in chloroform is observed .

Fig.2.7

In addition to the air sealing method (sealed unit, N2 blanket transformers) which separate the contact
of air to oil during equipment operation, for the prevention or reduction of ageing of oil, "inhibition" of
mineral oil is also practiced. Here, the resistance to oxidation is improved by the addition of ageing
inhibitors. These react with the broken bonds in the oil molecules and interrupt the oxidation process
by forming stable, inert and dielectrically harmless compounds. They are used up in the process and must
therefore be replenished from time to time.

Low viscosity mineral oil in refined condition, such as it is used as transformer oil in devices or for oil-
filled cables, is a transparent and clear liquid has a solidifying point (pour point) at -40 °C. The
electrical properties depend upon the purity. The quoted breakdown field strength of 25kV/mm is
valid for spacings in the mm-range; for thin layers breakdown field strengths up to l00kV/mm have
been measured and for films in the m-range a value of 300kV/mm has been quoted. Combined with the

6
low dielectric constant, this is the reason for the excellent electric strength of oil-paper insulation systems.
Under the action of electric discharges oil decomposes with the formation of gas. In arcs thermal
degradation of the oil occurs with the formation of about 60% hydrogen, ca. 10% other gases and ca. 25%
saturated and unsaturated hydrocarbons. The high hydrogen content of the degradation products results
in concentrated cooling of the arc.

At sharp-edged electrodes, e.g. at the edges of metal foils in capacitors, continuous discharges can also
cause polymerization of the oil components with the formation of solid substances (X-wax formation);
these possess a lower breakdown voltage than the liquid components and therefore initiate the breakdown.

Of the thermal properties, the specific heat and the long-term thermal stability are of particular
significance. Since the oil, besides having insulating duties, frequently must also perform as convection
coolant, the relatively high specific heat is of advantage. The long-term thermal stability, however, lies at
around 90°C only and so in many devices limits the permissible rated power.

The application of oil as an insulating material occurs almost exclusively in combination with cellulose in
the form of paper or pressboard. Oil-impregnated paper, as an electrically extremely strong and under
continuous stressing a proven compound dielectric, is the most important high-voltage technical
insulating material, without which the present-day concept of many transformers, instrument transformers,
bushings, capacitors and cables would be inconceivable.

2.2 Paper

Paper for electrotechnical purposes is primarily manufactured from wood-pulp of the slowly growing
northern spruce or pine. Only capacitor paper down to about 10m thickness is made of rag-pulp. In
the first phase of the manufacture the cellulose pulp is separated from the other wood components
(e.g. lignin, resin) whereby either acidic or alkaline decomposition takes place in a pulp digester.
Cellulose to be used in electro-technical applications must be carefully washed in order to remove the
insulation reducing bleaching agents or acids. The raw cellulose thus obtained (mostly in the form of
unbleached lignosulphonates) is dissolved in water and the fibre suspension is then separated into
individual fibres and ground. The duration of the grinding and the method essentially determine the paper
quality. The watery suspension is then fed to the paper machine which produces machine-fine paper in
rolls. By pressure treatment in the glazing rollers, increased smoothness and shine of the paper, as well as
greater density, improved breakdown strength and an increased dielectric constant are also achieved.

Cellulose paper is manufactured into transformer paper with thicknesses of 0.05 mm to 0.08 mm, and
into cable paper of 0.08 mm to 0.2 mm. Pressboard is made by wet pressing several thin individual layers
without bonding material; per mm thickness it consists of 35 individual layers, each about 30m thick.
The theoretical density of 1.55 g/cm3 is not achieved in paper owing to the pore volume of 20...60%;
machine-fine paper has 0.65g/cm3, high-gloss paper 1.15 g/cm3 and press-board reaches 1.3 g/cm3.

The dielectric constant of cellulose at 20 °C is 5.6, of paper 1.5 to 3.5 and of pressboard 4.5. The
dissipation factor lies around (3 ...4)-10-3, the volume resistivity in the dry state between 1015 and 1017
cm. The volume resistivity decreases by about a factor of ten for every 1.5% of absorbed water.

7
 Fig.2.8

Paper is very hygroscopic and during storage in an atmosphere of average humidity absorbs 5...10%
water. Since ageing of paper is influenced primarily by water and heat, good drying is of particular
importance. The utilization of paper occurs in the form of hardboard, soft paper and pressboard.
Hardboard results on compression with epoxy or phenolic resins and is used for supports, insulating
barriers, etc. Soft paper or pressboard are used in the oil-impregnated form in transformers, instrument
transformers, bushings, capacitors and oil-filled cables.

2.3 Oil-impregnated paper

a) Properties, manufacture

Oil-impregnated paper is the most important compound dielectric for high-voltage insulation systems.
Since several layers of paper are usually used, and also on account of the fibrous nature of paper, for
the characteristic quantities one may assume a series connection of the insulating materials soft paper and
oil. If we consider a plate-type compound dielectric of thickness s and permittivity , this can be taken as
comprising a series connection of a pure oil dielectric (s l , l ) and a pure paper dielectric (s2, 2). For
the resultant dielectric constant we have:
   (s  s )
 1 2 1 2 with s1 + s2 = s
1  s 2   2  s1

The unknown thickness s1 and s2 are eliminated by introducing the fiducial pore volume v:
p
v  1
Z

Here, P is the density of the paper containing pores and Z is the density of pure cellulose.

For total impregnation v is equal to the oil volume v1 and for the volume of paper v2 = 1 – v1 holds.

Thus we have:
1   2

 1  v1 ( 2   1 )

Cellulose has a density Z = 1.55 g/cm3 and 2 =5.6. With l =2.2 for oil, the resultant dielectric constant is

for impregnated cable paper  = 3.1,

(P = 0.75 g/cm3, v1 = 0.516)

for impregnated capacitor paper  = 4.0,

(P = 1.15 g/cm3, v1 = 0.26)

8
for impregnated pressboard  = 4.5 .
(P = 1.3 g/cm3, v1 = 0.16)

With the usual values 1 =2.2 and 2 = 5.6, we have for the ratio of electric stress in the oil/paper
lamination:
E1  2
  2.55
E2  1

The oil is therefore electrically stressed more than the paper; using a finely graded lamination a large
number of thin oil films is obtained, the high electric strength of which is responsible for the excellent
breakdown field strength of an oil-paper dielectric. The paper promotes the formation of thin oil
layers, acts as a barrier to bridging impurities and ensures the mechanical stability of the insulating
system.

The breakdown field strengths of oil-paper dielectrics, even for large thickness, take
values up to 400kV/mm, and, as operating field strengths in direct voltage capacitors up
to l00kV/mm; during ac stressing, up to 20kV/mm are applied. In a high-quality oil-
paper dielectric the dissipation factor is tan  310 -3 , the volume resistivity   1015cm, and the
permissible temperature limit is ca. 100 °C.

The oil-paper insulation system of an apparatus must be carefully processed during manufacture in order to
prevent detrimental gas occlusions, which lead to partial discharges
and also reduce the breakdown field strength of oil by the solution of gas. Further,
moisture must be wholly removed since it causes a noticeable deterioration not only of
the electric strength of the oil, but also of the ageing stability of the paper.
Processing is carried out by subjecting the paper-insulated live parts in heated vacuum
chambers to a vacuum of 10-3...10-4mbar and temperatures up to 110°C; the drying
time increases quadratically with the thickness of the insulation and is of the order of
days or weeks. The drying procedure is controlled by permanent monitoring of the
dissipation factor. The relationship between residual moisture and degassing pressure
during drying is described by the adsorption isotherms of Fig. 2.9.
According to these, a pressure of about 0.1mbar at 110°C is necessary to achieve a residual moisture of
10-4.

Fig.2.9

The dried paper insulation is then impregnated, in vacuum if possible, with recently refined and warm
mineral oil. Its hygroscopic nature causes the dry paper to extract the moisture still contained in the oil;

9
in turn, the oil dissolves the residual gases in the paper and so contributes to an improvement of the partial
discharge performance.

b) Oil-impregnated paper as cable insulation

The oil-paper dielectric also plays a prominent role in cable insulation. In the voltage range up to about
60kV the so-called compound-filled cable was used, which has been practically supplanted by the PE-
cable; for 100kV and higher the oil-filled cable predominates.

The conductor is first provided with a paper tape insulation wound without overlap 20… 30mm wide
and 0.1 to 0.15 mm thick. The paper insulation is dried and impregnated. A low viscosity mineral oil is
used for oil-filled cables and a mineral oil thickened with resin additives for compound-filled cables.

In compound-filled cables, the permeating substance is of low viscosity at the impregnating temperature
and of high viscosity at ambient and operating temperatures so that bleeding of the cable does not occur
during transport and installation. The high viscosity at operating temperature prevents or hinders the
permeating substance from running down into the lower-lying cable sections when laid on inclined slopes.

The use of the compound-filled cable, which is of simple construction, is restricted to medium voltages
due to the risk of partial discharges. During thermal stress the compound expands more than the lead
sheath, which latter then experiences an irreversible expansion. After cooling, gas-filled cavities are
created which, with regard to the partial discharge inception, limit the permissible operating field strength
to 4kV/mm.

Compound-filled cables are in limited use for operating voltages over 60 kV in the form of internal and
external gas-pressure cables. In the former, cavity formation in the compound-paper insulation is
prevented from doing any harm by keeping the insulation at a gas pressure of 15bar nitrogen so
increasing the inception voltage in the cavities. Then, operating field strengths of 9kV/mm and, with the
addition of SF6, up to 12...13kV/mm can be attained. In the external gas-pressure cable, the normal
compound-filled cable is placed inside a steel pipe filled with nitrogen (15 bar); the lead sheath acts as
a pressure membrane and prevents the formation of cavities or ensures a high pressure in the cavity.

The low-viscosity, mostly inhibited, mineral oils used in oil-filled cables prevent the occurrence of
cavities. Expansion vessels are placed at regular intervals of a few km; these keep the cable at a certain
pressure. On warming Up the low-viscosity oil flows to the expansion vessel without expanding the lead
sheath, and from there, on cooling, it is pressed back into the cable insulation once more.

The value of the pressure influences the breakdown field strength as shown in Fig. 2.5-5, If the oil pressure
is a few bar, one speaks of low pressure oil-filled cables, and, for ca. 15 bar oil pressure, of high pressure
oil-filled cables. The operating field strength has values up to 14kV/mm. In Europe predominantly low
pressure oil-filled cables are used.

Fig.2.10

With regard to the necessary impulse voltage strength, thinner paper is often arranged in the region of
maximum stress, that is, at the inner conductor, whilst in the outer region normal cable paper is used.
Besides an increase in the a.c. strength, this kind of lamination results in homogenization of the field

10
distribution, since, due to the higher dielectric constant of the thinner paper layers, the voltage
distribution shifts towards the outer regions where the field is weaker. These measures, as well as the use
of special less porous -paper, make it possible to design a 400 kV oil-filled cable for an impulse level of
1640 kV with an insulation wall thickness of only 28mm; the maximum field strength at this stress is
93 kV/mm.

The dissipation factor of oil-filled cables lies around (2...4)10-3. It can be shown that dielectric losses
for increasing transmission voltage limit the transmitting power of a cable. Thus, as Fig. 2-6 proves, for tan
= 210-3 the transmitting power referred to the conductor diameter reaches a maximum at a transmission
voltage of 700kV. Higher transmission voltages are only practicable if dielectrics with a dissipation
factor lower than 210-3 are available. Insulation systems of this kind can consist of oil impregnated
plastic foils, synthetic paper or paper/plastic combinations.

Fig.2.11

11
3. Transformer Construction

The invention of the power transformer towards the end of the nineteenth century made possible the
development of the modern AC supply systems, with power stations often located many miles from
centres of electrical load.

The most commonly used power transformer is the oil-immersed transformer, which means the main
insulation material in transformers is mineral oil and cellulose paper. Dry-type transformer usually
uses cast-resin as the main insulation material. Looking from outside, a transformer is comprised of
tank, termination (bushing) and accessories (radiator, oil conservator, relay cables and etc). The
transformer tank contains windings, magnetic core and insulation systems.

3.1 Magnetic core

3.1.1 Core form and shell-form transformers

Although transformers are primarily classified according to their function in a power system, they also
have subsidiary classification according to how they are constructed. As an example of the latter type
of classification, perhaps the most important is the distinction between core-form and shell-form
transformers.

In a core-form design, shown in Fig.3-1, the coils are wrapped or stacked around the core. This leads
itself to cylindrical shaped coils. Generally high voltage and low voltage coil are wound
concentrically, with the low voltage coil inside the high voltage coil.

The most common magnetic circuit design for a three-phase transformer is the three-limb core-form:
three parallel, vertical limbs are connected at the top and bottom by horizontal "yokes". The five-limb
core-form magnetic circuit has three limbs carrying the windings, and two unwound limbs with a
lesser cross-section. The yokes connecting all five limbs also have a reduced cross-section in
comparison with the wound limbs.

(a) three-limb core-form (b) five-limb core-form

Fig.3-1 Core-form magnetic circuit

In the shell-form design the core is wrapped or stacked around the coils. This leads itself to flat oval
shaped coils called pancake coils, with the high and low voltage coil staked on to each other,
generally more than one layer each in an alternating fashion.

The conventional shell-form three phase design has a frame with the three wound limbs horizontal
and having a common centre line. The core steel limbs inside the windings have an essentially
rectangular cross-section, and the adjoining parts of the magnetic circuit surround the windings like a
shell.

12
A new common core for a three phase shell-form transformer is the seven-limb core which posses
advantages in both materials usage and losses.

(a) three phase conventional shell-form (b) seven-limb shell-form

Fig.3-2 Shell-form magnetic circuit

The principal difference between the designs, to be discussed here, lies in their behaviour when
subjected to an asymmetric three-phase set of voltages having non-zero sum, i.e. having a zero-
sequence component.

In the three-limb core-form transformer, positive and negative sequence flux components in the
wound limb cancel out via the yokes, but the residual zero-sequence flux must find a return path from
the upper yoke to the lower yoke outside the excited winding (without a delta winding). The zero-
sequence yoke leakage flux induced circulating and eddy currents in the clamping structure and the
tank, generating extra stray losses in these components. The zero sequence impedance of a star
winding is likely to be in the order of 75% to 200% of the positive sequence impedance between the
primary and secondary windings. In a five-limb core-form or a shell-form transformer, there are
return paths available for the zero-sequence flux through unwound parts of the magnetic circuit (side
limbs of five-limb core, outside parts of the shell form frame). Since this is a low reluctance return
path for the flux, the zero sequence impedance of a star winding will be the magnetising impedance of
the core configuration (10 to 50 times of the positive sequence impedance). In a single phase
transformer, the zero sequence impedance of a star winding will be also the magnetising impedance of
the core configuration (50 to 100 times of the positive sequence impedance).

3.1.2 Core materials

Cores are generally constructed of stacks of electrical steel laminations that are manufactured in a way
to enhance its magnetic properties. The steel lamination is made from iron and silicon alloy. Basic
cold-rolled flat alloy sheets known as 'cold-rolled grain-oriented silicon steel' have been used since the
1960s. It has a silicon content of about 3% which gives the alloy rather a high resistivity (~50*10 -8-
m) as compared with resistivity of low-carbon magnetic steel (~25*10-8-m). Although higher silicon
content can produce even higher resistivity, the brittleness increases with silicon content and this
makes the lamination difficult to handle.

Improved quality control producing better grain orientation and thinner sheets ('Hi-B steels') has
reduced no-load losses by some 15% compared to conventional cold-rolled grain-oriented silicon steel
types. Surface laser-etched steels (0.23mm thickness) are used by leading transformer manufacturers.
This results in a further 15% loss reduction and such treatment may be justified as a result of the
electrical supply utility's loss capitalisation formulae (see Appendix: cost of ownership of
transformers).

Thin sheets, 0.3 mm to 0.23 mm thick, each having a thin coating of insulation so that there is no
conduction path from sheet to sheet, minimise eddy currents in the core.

13
At the most onerous tap setting, typical flux density values for modern cold-rolled grain-oriented
silicon steel should not exceed 1.7 Tesla (Wb/m2) without manufacturers' advice.

Rapidly cooled, thin (typically 0.025mm thick) amorphous ribbon steel cores with lower magnetic
saturation limits of about 1.4 Tesla can reduce the core losses by up to 75% as compared to cold-
rolled grain-oriented silicon steel. Amorphous steel generally consists of iron, boron and other
additions. Distribution transformers using this material are widely used in North America.

A continuous magnetic circuit is desirable so that it can carry flux uninterrupted. It is obtained by
avoiding air gaps or non-magnetic components at joints. At joints sheets of laminations are
overlapped to ensure a continuous magnetic path. Core lamination clamping bolts are no longer used
in modern designs and the laminations are rather held together by the hoop stress of the windings, by
fibreglass banding or by pinching the yoke between external clamps.

3.2 Winding structure

3.2.1 Disc type and layer type of winding

Phase winding set is normally arranged concentrically, taking the centre line of the limb as the axis.
Lower voltage winding is wound near the limb and then higher voltage winding for the electrical
insulation purpose.

In disc-type windings, the adjacent turns are arranged radially as a disc and many discs are then
arranged axially, being separated by oil duct and spacers to make a winding, as shown in Fig.2.3.

In layer-type winding, the adjacent turns are no loner arranged radially but axially and the winding
consists of a certain number of concentric layers separately by oil paper insulation, as shown in Fig. 3-
3.

(a) disc-type winding (b) layer–type winding

Fig.3-3. Disc and layer type windings

3.2.2 Three phase winding connection

There are two principal three-phase connections of transformer windings: Star or Y-connection, and
Delta or D-connection.

Advantages of a Y-connected winding lie in the following aspects:

14
  More economical for a high-voltage winding,
 Neutral point available and permits direct earthing or earthing through an impedance.
 Permits reduced insulation level of the neutral,
 Permits winding taps and tap changer to be located at the neutral end of each phase - low
voltages to earth and phase to phase.

Advantages of a D-connected winding lie in the following aspects:

 More economical for a high-current, low-voltage winding.

 In combination with a star-connected winding, it reduces the zero-sequence impedance
current in that winding.

Two winding transformer (double wound) is the basic transformer type with two windings connecting
a higher voltage system to a lower voltage system. This type is the normal arrangement for step-down
transformers in distribution and sub-transmission systems, and for generator transformers.

Three winding transformer is used under the situations where, for design reasons, or because a third
voltage level is involved, a third winding, as called delta tertiary winding, is added to a double wound
transformer.

A star/star transformer is often supplied with a third (delta-connected) winding for one or more of the
following reasons:

 To reduce the transformer impedance to zero sequence currents and therefore permit the flow
of earth fault currents of sufficient magnitude to operate the protection.
 To suppress the third harmonics due to the no-load current in the earth connection when the
neutral is earthed. These harmonics have been known to induce disturbances in neighbouring
low voltage telecommunication cables.
 To stabilise the phase-to-phase voltages under unbalanced load conditions (e.g. a single phase
load between one phase and neutral). Without a tertiary winding the current in the
uncompensated phases is purely magnetising and, by saturation, causes deformation of the
phase voltages and displacement of the neutral point. The addition of a delta tertiary winding
balances the ampere turns in all three phases eliminating such phenomena.
 To enable overpotential testing of large high voltage transformers to be carried out by
excitation at a relatively low voltage. This requirement depends upon the transformer
manufacturers' test bay capabilities. However, for such test purposes the tertiary may only
need to be of a very low rating and connected only for the factory tests.
 To provide an intermediate voltage level for supply to an auxiliary load where a tertiary
winding offers a more economical solution than a separate transformer.

Because of these apparent advantages a general view that such a tertiary winding is essential has
flourished. However, this is not the case and the tertiary involves an increase in transformer cost of
approximately 6% to 8% with a corresponding increase in losses of some 5%. The cross-section of the
tertiary winding is usually determined by fault withstand considerations.

The basic transformer principle can be achieved using a single winding per phase. If a tap is made part
way down the winding, this can be the low voltage terminal just as though this were a separate
winding. These transformers are so called auto-transformers.

By eliminating the second winding, an auto-transformer is potentially cheaper than a two winding
counterpart. In practice, such cost savings only apply for voltage transformation ratios of up to about
3:1 if adequate power transfer is to be achieved. Thus, for example, a transformer with a voltage ratio
of 275/132kV will be a straightforward auto-transformer choice; a ratio of 275/66kV would, however,
probably favour a double-wound arrangement.

15
Both high voltage and low voltage systems have the same neutral (auto-transformers are usually star
connected) and this would often be undesirable except in transmission systems where solid earthing of
neutrals is common at all voltage levels.

x X

Tank

MV Winding

HV Winding
LV Winding
A
Core

a Am
Fig.3-4 Construction of phase-winding in a 220/110kV auto-transformer (tertiary 11kV)

3.2.3 Conductors

Transmission and distribution oil-immersed power transformer windings are usually made of copper
to reduce load losses. As ratings increase the winding conductors are connected in parallel to reduce
eddy current loss and transposed to avoid leakage flux circulating currents.

Unequal current distribution in the conductor, also called as skin effect, is an important source of loss
in the windings and it is current practice to sub-divide conductors. Paper insulated cable, supplied
directly to transformer manufacturers by manufacturers of insulated wire, consists of multiple wires
coated with epoxy resin. These wires are transposed within the cable to avoid circulating currents, as
shown in Fig.2.5.

Fig.2.5 Conductor with several insulated and transposed wires

Aluminium has a higher specific resistance than copper and therefore requires a larger cross-section
for a given current rating. It is not therefore generally used in power transformers. However,
aluminium has certain advantages over copper when used as foil windings in dry type cast resin
distribution transformers. Aluminium has a coefficient of expansion of approximately 24x10 -6K-1,

16
compared to 17x10-6K-1 for copper, and this is more similar to the resins used. The short circuit
thermal withstand time tends to be greater for aluminium compared to copper in an equivalent design
and aluminium foil eddy current losses are lower.

The impulse withstand tends to be lower for dry type transformers but this depends upon strip or foil
winding construction and the resins used. Some manufacturers offer 95 kV BIL designs whereas IEC
726 requires 75 kV BIL for 12kV systems.

3.3 Insulation Structure

Winding insulation in oil-immersed transformers is a cellulose paper material. Transformers are

vacuum impregnated with high quality, extremely clean, hot mineral-insulating oil with a water
content less than 2ppm (parts per million) and air content less than 0.2%.

Transformer winding and lead must operate at high voltage relative to the core, tank and structural
elements at earth potential. In addition, different windings and even parts of the same winding operate
at different voltages. This requires that some form of insulation between various parts be provided to
avoid voltage breakdown or corona discharges. Oil surrounding the windings provides some
insulation value. Further insulation is provided by paper covering over the conductor/cables. Other
insulating structures which are generally present in sheet form, often wrapped into a cylindrical shape,
are made of pressboard. This is a material made from cellulose fibres which are compacted together
into a fairly dense and rigid matrix.

Because of the different dielectric constants (relative permittivity) of oil (r=2.3) or air (r=1.0) and
paper (r=6.0 dry paper), the electric stresses are unequally divided between them. The electric
stresses are generally higher in oil or air then in the paper insulation, Unfortunately oil or air has a
lower breakdown strength than paper. In the case of oil, it has been found that subdividing the oil gap
by means of thin insulating barriers, usually made of pressboard, can raise the breakdown strength in
the oil. Thus large oil gaps between the windings are usually subdivided by multiple pressboard
barriers, as shown in Fig.2.6. The oil gap thickness is maintained by means of long vertical narrow
sticks around the circumference of the cylindrical pressboard barriers. This is referred to as the major
insulating structure as the oil duct +pressboard cylinder insulates the phase to earth and phase to phase
voltages. In transformers with disc type winding structure, the minor insulation structure consists of
the smaller oil gaps separating the discs and maintained by the key spacers which are narrow
insulators, usually made of pressboard, spaced radially around the disc’s circumference. In addition,
the turn to turn insulation, usually made of paper, can be considered as part of the minor insulation
structure.

The major and minor insulation designs, including overall winding to winding separation and number
of barriers as well as disc to disc separation and paper covering thickness, are often determined by
design rules based on extensive experience.

There is another mechanism for breakdown due to creep. It occurs along the surface of the insulation.
It requires sufficiently high electric stresses directed along the surfaces as well as sufficiently long
interrupted paths over which the high stresses are present. The sticks, barriers can be a source of
breakdown due to creep.

17
 Winding Section II

Winding Section I
Core

Oil duct Pressboard

Fig.3-6 Schematic arrangement of main insulation structure

3.4 Cooling

Transformer oil acts as not only the dielectric for electric insulation, also the cooling medium.

In the simplest cooling method, the heat conducted to the oil from the windings and core is
transmitted to the surrounding air at the tank surface. In practice, only the smallest distribution
transformers, for example 10kVA pole mounted, have enough tank surface area to dissipate the
internal heat effectively

As the transformer size increases the surface area for heat dissipation is deliberately increased by
attaching radiators to the tank. As the transformer rating increases further, the number of radiators
required becomes too large for all of them to be attached to the tank, and separate cooler banks are
used.

In the cooling method described above, no moving parts are used. As the oil is warmed inside the tank,
the warmer oil rises to the top of the tank and into the tops of the radiators. As the oil cools, it falls to
the bottom of the radiator then back into the bottom of the tank. This sequence then repeats itself,
giving a 'natural' circulation of cooling oil. Increased cooling efficiency is obtained by fitting fans to
the radiators to blow cooling air across the radiator surfaces.

A further increase in efficiency is achieved by pumping the oil around the cooling circuit, thereby
boosting the natural circulation into forced circulation. The oil can be forced into closer contact with
the winding conductors to improve the heat extraction rate through directing the oil flow paths.

The IEC cooling classification codes allow the desired type of cooling to be simply specified. The
codes indicate the primary cooling medium, i.e. the medium extracting the heat from the windings and
core, and the secondary cooling medium, i.e. the medium which removes the heat from the primary
cooling medium. The type of cooling method (how it is circulated) can also be specified. The
following codes are used:
Table 3.1 IEC cooling classification codes

Cooling medium Code Circulation Code

Mineral oil O Natural N

Water W Forced F
Air A Directed D

18
The coding method is to specify, in order, the primary cooling medium, how it is circulated; the
secondary cooling medium, and how it is circulated. For example, an oil-immersed transformer with
natural oil circulation to radiators dissipating heat naturally to surrounding air is coded as ONAN.
Adding fans to the radiators changes this to ONAF, and so on.

Oil cooling is normally achieved by heat exchange to the surrounding air. Sometimes a water jacket
acts as the secondary cooling medium. Fans may be mounted directly onto the radiators and it is
customary to use a number of separate fans rather than one or two larger fans.

Oil pumps for OFAF cooling are mounted in the return pipe at the bottom of the radiators. The motors
driving the pumps often use the transformer oil as their cooling system.

With ODAF cooling, oil is directed to flow through the windings to make sure that the heat is
extracted from all the conductors. Oil-to-air coolers tend to be compact and use relatively large fan
blowers. With this arrangement the cooling effectiveness is very dependent on proper operation of the
fans and oil pumps since the small amount of cooling surface area gives relatively poor cooling by
natural convection alone.

Water cooling (ODWF) has similar characteristics to the ODAF cooling described above and is
sometimes found in power station situations where ample and well-maintained supplies of cooling
water are available. Cooling effectiveness is dependent upon the flow of cooling water and therefore
on proper operation of the water pumps. Natural cooling with the out-of-service water pumps is very
limited. Operational experience has not always been good, with corrosion and leakage problems, and
the complexity of water pumps, pipes, valves and flow monitoring equipment.

The ODAF arrangement is probably favourable as a replacement for the ODWF designs. Double wall
cooler pipes give added protection against water leakage. The inner tube carries the water and any
leakage into the outer tube is detected and causes an alarm. This more secure arrangement is at the
expense of slightly reduced heat transfer for a given pipe size.

Normal practice with cooling plant is to duplicate systems so that a failure of one need not directly
affect operation of the transformer. Two separate radiators or radiator banks and duplicate oil pumps
may be specified. In larger ODAF cooling designs there may be four independent unit coolers giving
a degree of redundancy. The transformer may be rated for full output with three out of the four coolers
in service.

Dry type transformers will normally be naturally air cooled (AN) or incorporate fans (classification
AF).

3.5 Transformer accessories and operation

Some accessories are optional and will not necessarily be justified on every transformer; others are
important to the safety or operation of the transformer and will therefore be of a mandatory nature.

3.5.1 Buchholz relay

A Buchholz relay is connected in the oil feed pipe connecting the conservator to the main tank. The
relay is designed to:

 Detect free gas being slowly produced in the main tank, possibly as a result of partial
discharges. Under such conditions the relay may be set to give an alarm condition after a
certain amount of gas has evolved. Examples of incipient faults include broken down core
bolt insulation on older transformers, shorted laminations, bad contacts and overheating in
part of the windings

19
  Detect a sudden surge movement of oil due to an internal transformer fault. Under such
conditions the relay is normally set to trip the high and low voltage transformer circuit
breakers. Examples of such oil surge faults include earth faults, winding short circuits,
puncture of bushings and short circuits between phases
 Provide a chamber for collection and later analysis of evolved gas. Chemical analysis of the
gas and transformer oil can give maintenance staff an indication as to the cause of the fault.

Buchholz relays are considered mandatory for conservator type transformers since they are protective
devices. They should be installed in accordance with the manufacturers' instructions since a certain
length of straight oil piping is required either side of the relay to ensure correct operation.

From time to time maloperation of Buchholz relays is reported. This is often due to vibration effects
and a relay designed for seismic conditions may overcome the problem.

3.5.2 Sudden pressure relay and gas analyser relay

On non-conservator type transformers, the useful protective and gas analysis feature of the Buchholz
relay cannot be provided. In its place, a sudden pressure relay detects internal pressure rises due to
faults, and gas devices can be used to detect an accumulation of gases.

Sudden pressure relays are normal accessories for sealed transformers. Gas analyser devices tend only
to be used on large important transformers.

3.5.3 Pressure relief devices

A pressure relief device should be regarded as an essential accessory for any oil-immersed
transformer. Very large transformers may require two devices to adequately protect the tank. Violent
pressures built up in the transformer tank during an internal fault could split the tank and result in the
hazardous expulsion of hot oil. In order to avoid tank rupture resulting from the high pressures
involved in an internal transformer fault, a quick acting pressure relief device is specified and used to
give a controlled release of internal pressure.

Older transformers may have been fitted with a rupturing diaphragm type device where the excess
pressures breaks a fragile diaphragm and allows oil to be discharged. However, the overall operating
time may be too slow to protect the tank against splitting.

3.5.4 Temperature monitoring

A correctly specified and loaded transformer should not develop excessive temperatures in operation.
Oil and winding temperature is monitored in all but small (say, less than 200kVA) distribution
transformers.

Apart from the facility to monitor temperature (useful during controlled overloading), an important
feature of the winding temperature indicator is to initiate automatic switch-on and switch-off of
cooling fans and oil circulation pumps. In this way a dual rated transformer with a cooling
classification of, for example, ONAN/ONAF will automatically switch from ONAN to ONAF (and
back) according to the transformer loading conditions. The winding temperature monitor on an oil-
immersed transformer simulates the temperature by using an oil temperature sensor and injecting
additional heat into the sensor from a current transformer connected to one of the transformer
terminals. In this way the winding temperature monitor registers a temperature above that of the oil by
an amount that is dependent on the load current of the transformer. This arrangement is usually
calibrated on site and is used to indicate the hot spot winding temperature.

The oil temperature monitor is usually a capillary type thermometer with the sensor placed in the
vicinity of the hottest oil in the tank (i.e. at the top of the tank just prior to entering the radiators).

20
Both oil and winding temperature monitors are fitted with contacts which can be set to operate at a
desired temperature. Such contacts are used for alarm (and possibly trip) purposes and also to operate
auxiliaries as noted above.

Alarm and trip temperature settings are usually advised by the manufacturer. Note that it will usually
be necessary to modify the settings if the transformer is used for controlled overloading since winding
and oil temperature are allowed to reach higher temperatures during overloading than during normal
operation.

Dry type transformers incorporate thermistor probes through the resin, usually by the low voltage
winding hot spot area. Negative temperature coefficient thermistors are available with a resistance
range from some 2000 at 200C to 1M at ambient temperature with an accuracy of some 3%.
Settings for alarm and trip conditions may be made using an electronic control device.

3.5.5 Breathers

Breathers are placed in the vent pipes of conservators to dry the air entering the conservator as the
volume of oil contracts on transformer cooling.

Traditional breathers use the moisture absorbing properties of silica gel crystals. These crystals need
replacement when they become saturated with moisture. Replacement is indicated by a change in
colour of the crystals from blue to pink.

An alternative technique is to continuously extract moisture dissolved in the transformer oil by

freezing the moisture out of the air by passing it over refrigerating elements and then evaporating it
off to the atmosphere. The oil is kept particularly dry and researches have shown an improved life
span of the transformer. The 'Drycol' device is a standard fitting on 275 kV and 400kV transformers in
UK.

21
4 Basic Transformer Design
4.1 Volts Per Turn And Flux Density

Refer to Faraday’s law, the induced voltage (or the back e.m.f.) shown in Fig.4-1, is proportional to
the rate of change of flux linkage, and its polarity is such as to oppose that flux linkage change if
current were allowed to flow (as Lenz’s law states).

(V1: terminal voltage, E1: back e.m.f.. Im: magnetising current)

Fig.5-1 A no-load transformer

d
e  N (1.1)
dt

If the supply voltage is a pure sin wave, equation (1.3) can be further derived to give the relationship
between volts per turn and the flux density.

d
e  N
dt
d
2 E sin  t   N (1.2)
dt
2E
 cos  t
N

From equation (1.2) we can see that under a sinusoidal supply voltage, the flux also changes with the
time sinusoidally. The relationship between the maximum flux within the core and the volts per turn
for the windings is as
2E

N  2  f
E
 4.44 f  
N
(1.3)
E
 4.44 f  A  Bm
N

22
From (1.3), for a given supply frequency the relationship between the volts per turn and the total flux
within the core remains constant. Since for a given core the cross-sectional area of the limb is a
constant, this means that the relationship between volts per turn and the flux density also remains
constant at a given frequency. The designer of the transformer will wish to ensure that the flux density
is as high as possible consistent with avoiding saturation within the core. It is common in the UK for
the voltage of a system to be allowed to rise up to 10% above its nominal level, for example at times
of light system load. Under the nominal voltage, the designer can allow the nominal flux density to
approach closer to saturation and at the same time also ensure a safe margin exists below saturation.

Typical flux density values for modern cold-rolled grain-oriented silicon steel should not exceed 1.7
Tesla (Wb/m2) without manufacturers' advice.

4.2 Calculation example

The maximum flux density within the magnetic core of a 50Hz, 400kV/132kV transformer is
restricted to 1.55Tesla as the core has a circular cross section with a diameter of 1m. Calculate the
volts per turn for the winding and the number of turns for HV and LV windings.

E
The relationship between volt per turn and the maximum flux density is given as  4.44 fB m A , A
n
400kV/132kV three-phase transformer is connected as Yyn*, so the phase voltage for HV is
E1 400 E2 132
  230kV , and the phase voltage for LV winding is   76kV
3 3 3 3

Volt per turn is calculated as

E
 4.44 fBm A  4.44  50  1.55    0.5 2  270 (V )
n

The HV winding has n1=230000/270=851 turns

The LV winding has n2=76000/270=281.5 turns

4.3 Voltage regulation, tappings and tap changer

4.3.1 Voltage regulations

One of the main requirements of any electrical system is that it should provide a voltage to the user
which remains within closely limits regardless of the loading on the system, despite the regulation
occurring within the many supply transformers and cables, which will vary greatly from conditions of
light to full load.

Voltage regulation that occurs at the secondary terminals of a transformer when a load is supplied
consists of voltage drop due to the resistance of the windings and voltage drops due to the leakage
reactance between the windings. These two voltage drops are in quadrature with each other, the
resistance drop being in phase with the load current.

Long term voltage variation (duration exceeding ten seconds and up to several hours) are also due to
progressive start up or shut down of numerous pieces of equipment of low power and to changes in
operating conditions.

23
Transformers provide the option of compensating for system regulation, as well as the regulation
which they themselves, introduce, by the use of tappings which may be varied either on-load, in the
case of larger more important transformers, or off-circuit in the case of smaller distribution or
auxiliary transformers.

Almost all transformers incorporate some means of adjusting their voltage ratio by means of the
addition or removal of tapping turns. Principal tapping is the tapping to which the rated quantities are
related and, in particular, the rated voltage ratio.

In summary, transformer users require tappings for a number of reasons:

 To compensate for changes in the applied voltage on bulk supply and other system transformers.
 To compensate for regulation within the transformer and maintain the output voltage constant on
the bulk supply and other system transformers.
 On generator and interbus transformers to assist in the control of system VAR flows.
 To allow for compensation for factors not accurately known at the time of planning an electrical
system
 To allow for future changes in system conditions.

Note: In the Uganda system transformer is normally used to describe that class of transformer
which makes the transformation from 33kV or 66kV to 11kV, providing the bulk supply.

4.3.2 Transformer tappings and tapping range

Generation of VARs will be effected by tapping-up the generator transformer, that is, increasing the
number of HV turns for a given 400kV system voltage. Absorption of VARs will occur if the
transformer is tapped down. This mode of operation leads to variation in flux density which must be
taken into account when designing the transformer.

An on-load tap changer is required to allow for variation of the HV system volts and generator power
factor. LV volts will generally remain within 5% under the control of the generator automatic
voltage regulator (AVR).

Interbus transformers interconnecting 400,275 and 132kV systems are mostly likely to be
autoconnected. Tapping must be provided either at 400kV line end or at the 132kV common points as
shown in Fig.3.3. The former arrangement in Fig.3.3 (a) requires the tap changer to be insulated for
400kV but maintain flux density constant, the latter one in Fig.3.3.(b) allows the tap changer to
operate at a modest 132kV, therefore being utilised in most practical schemes, but still results in some
variation of flux density.

Alternatively these interbus auto-transformers are used without tapchangers thererby avoid the high
cost of tapchager itself as well as the many disadvantages associated. This simplicity results in
slightly reduced flexibility in the operation of 275kV and 132kV system but it can be compensated by
the tapping on the 275/33kV or 132/33kV transformers.

24
 (a) tappings at 400kV end (b) tappings at 132kV end

Fig.3.3 Alternative locations for tappings of 400/132kV auto-transformers

In the UK the 400kV system is normally maintained within 5% of its nominal value. The 275kV and
132kV system, just as here in Uganda, are normally maintain within 10% of nominal since the
interconnection transformers are not provided with taps and the voltage regulation will make the
system voltage variation greater. Hence for 275/33kV and 132/33kV transformers, it is common for
them to have tapping ranges of +10% to –20%. 33/11kV transformers need to be provided
with tapping range of only 10%. 11/0.433kV transformers, are thus normally provided with
off-circuit taps usually at 2.5% and 5%.

For Y and D connected three phase transformers, the tappings are normally positioned as

(a) tappings at Y neutral (b) tappings at D terminal

Fig.3.4 Tappings position at Y and D connected three phase transformers

25
4.3.3 On-load tap changer

On-load tap-changers enable the turns ratio of transformer to be changed under load. Two switching
principles for switching under load were employed, and they are the slow motion reactor principle and
the high speed resistor principle. High quality devices with a high degree of operating reliability have
been developed based on both principles, however the high speed resistor principle is used exclusively
practically world-wide as it enables more compact on-load tap-changers to be designed.

Only in the USA is the reactor principle still used on a large scale.

Principle of action

A change of the ratio is achieved by adding to or subtracting turns from wither the primary or
secondary winding.

The OLTC should switch the transformer load current, IL, from the tap in operation to the neighbour
tap without interruption. The voltage between the taps is the step voltage, it normally lies between
0.8% to 2.5% of the rated voltage of the transformer.

OLTC comprises a diverter switch and tap selector when being used in a transformer with larger
power and higher voltage, in the other cases the functions of the diverter switch and the tap selector
are combined in a selector switch.

With an OLTC comprising a diverter switch and tap changer, the tap change occurs in two steps, as
shown in Fig3.5. First the next tap is selected by the tap selector at no load (Fig.3.5.(a) – (c)), then the
diverter switch switches the load current from the tap in operation to the selected tap (Fig.3.5.(d)- (g)).
The OLTC is actuated by means of a motor drive. The tap selector is operated by gearing directly
from the motor drive. At the same time, a spring energy accumulator is tensioned; this operates the
diverter switch in a very short time independent of the motion of the motor drive. The gearing ensures
that this diverter switch operation always occurs after the tap selection operation has been finished.
The switching time of a diverter switch lies between 40-60ms. During the diverter switch operation,
transition resistor are inserted which are loaded for 20-30ms, i.e. the resistors can be designed for
short-term loading. The total operation time of an OLTC is between 30 –60ms depending on the
design.

(a) (b) (c) (d) (e) (f) (g)

Tap selector Diverter switch

Fig.3.5 Switching sequence of tap selector - diverter switch

A selector switch carries out the tap changer from the tap in service to the adjacent tap in one step.
The spring energy accumulator is wound up by the motor drive which actuates the selector switch
sharply after releasing.

26
 (a) (b) (c) (d) (e)

Fig.3.6 Selector switch – switching sequence

4.3.4 Calculation example of diverter switch resistance

A diverter switch is connected at the neutral end of a 400kV, start connected winding, on a 500MVA
3 phase transformer. Tapping range is 25% in 18 linear steps Circulating current is to be 60% of full
load current. Calculate the diverter switch resistor’s value.

400 25 1
Volts per tapping step, Vc     3.2 (kV )
3 100 18

500
Full load current on normal tap, I FL  3   721 ( A)
400

Circulating current, Ic  0.6  I FL  0.6  721  433 ( A)

Vc 3200
Current limiting resistance, r    7.4 ()
Ic 433

Total resistance 7.4 in 2 units of 3.7 (one each contact)

3.3.5. Basic connection of tapped winding

The following basic arrangements of the tapped winding are used.

Linear arrangement is generally used on power transformers with moderate regulating ranges up to a
maximum of 20%.

With a reversing change-over selector, the voltage of the tapped winding is added to or subtracted
from the main winding so that the regulating range can be doubled or the number of taps reduced.
This reversing operation is achieved with the help of a change-over selector which is part of the tap
selector or of the OLTC.

During coarse change-over selector operation, the tapped winding is connected to the plus or minus
tapping of the coarse tapped winding. The coarse change-over selector is also part of the OLTC.

27
(a) linear (b) single reverse change-over selector (c) single coarse change-over selector

Fig.3.7 Basic connections of the tapped winding

28
5. Transformer operation

5.1 Parallel operation

Two or more transformers operate in parallel when they are connected to the same system on both the
primary and the secondary sides. The need to make transformers operate in parallel arises from the
increase in demand for electrical energy resulting from the growth of the system supplied, or from
considerations of security of supply or from limitations of the short-circuit power. The objectives is to
obtain a total maximum power as close as possible to the sum of the rated power of the transformers.
This is achieved when the following operating conditions are fulfilled:

 No circulating current exists in the secondary at no-load

 The load is shared in proportion to the rated powers
 The currents supplied are all in phase

5.1.1 Phase relationship – Phasor groups

Satisfactory parallel operation requires that the secondary voltages are equal and in phase. Clearly the
rated HV and LV voltages of the transformers must be in the same.

A conventional notation indicating the connections of the high-voltage and low-voltage windings and
their relative phase displacement expressed as a combination of letters and clock-hour figure. Figure
3.1 shows some examples for three-phase transformer connections.

Fig.3.1 Examples of connections and corresponding time indices

Refer:
https://www.dropbox.com/sh/3qtcrm6mdnovk86/AACo0e31v5pRvoqzNfRVAo5Ea/Transformers?d
l=0&lst=&preview=Transformer+winding+and+configuration.PDF&subfolder_nav_tracking=1

29
Equality of phase implies that the transformers belong to the same connection group or to compatible
groups. The time index which characterises them corresponds to the positions on a dial of a vector
representing the voltage at the low voltage terminal ‘a’ when the corresponding vector for the voltage
at the high voltage terminal ‘A’ (left hand terminal when viewing the high voltage side) is at 12
o’clock. The time indices increase in unit steps from 0 to 11 for normal transformers.

There are four groups of connections:

Group I: time indices 0, 4, 8.

Group II: time indices 6, 10, 2.
Group III: time indices 1,5.
Group IV: time indices 7,11.

The primary and secondary winding connections and the time index are shown on the transformer
diagram plate. The most common are as follows:

Group I: Dd0, Yy0, Dz0

Group II: Dd6, Yy6, Dz6
Group III: Dy5, Yy5, Yz5
Group IV: Dy11, Yd11, Yz11

Fig.3.2 shows the possible parallel connections for two transformers of any time indices. It is evident
that transformers of group I and II can only be paralleled with transformers of their own group, while
transformers of group III and IV can be interconnected for parallel operation.

Fig.3.2 Connections of transformers in parallel

30
5.1.2 Parallel operation with no-load

If the transformation ratios of two transformers are identical, the no-load voltages in the secondaries
are equal and no significant current will circulate. This is the case with transformers of the same
power and same manufacture whose primary and secondary turns are respectively the same.

The effect of a difference in the transformation ratios is more marked. Because of a different
construction or different power rating it is generally impossible to obtain the number of primary and
secondary turns in precisely the same ratio. For doubled power, the number of primary and secondary
turns are reduced by 30% due to the increase in the cross section of the core. The no-load voltages of
the two transformers are slightly different and a circulation current arises which is limited by the sum
of the short-circuit impedances of the two transformers.

u
I  100 (3.1)
S1
Z1  Z2
S2
where u = voltage difference for no load (in percentage)
Z1, Z2 = transformer impedance (in percentage)
S1, S2 = rated apparent power capacity
I = circulating current as percentage of rated current of transformer 1.

5.1.3 parallel operation with load

Supposing two single phase transformers having different ratings, same voltage ratios and different
impedance, are connected in parallel, the division of power between these two transformer is given as

S  S ra S  S rb
Za Zb
Sa  , Sb  (3.2)
S ra S rb S ra S rb
 
Za Zb Za Zb
where S= total power flow
Sa, Sb = the power flowed through transformers a, b
Sra, Srb = rated power capacity of transformers a, b
Za, Zb = transformer impedance of transformers a, b

From (3.2), it can be seen that as the load increases, the transformer with the lower impedance is the
first to reach its rated current (rated power) and this fixes the total current that can be taken. Where
several transformers are in parallel, the calculation for maximum power is carried out using the
transformer with the lowest impedance as the reference unit. As this transformer reaches its nominal
power the remainder will not be fully loaded, and

Z1 Z
S max  S1  S 2  S 2 1  .... (3.3)
Z2 Z2

To improve parallel operation, there are two ways, by means of an external reactance or by adjusting
the transformation ratio. An external reactance is connected in series to the primary or secondary of
the transformer with the lowest impedance. This other possibility is to adjust the transformation ratio
of the new transformer to create a circulating current that will reduce the excessive load current, so
allowing an increased total maximum load. This is most effective when the consumption of reactive
energy by the consumer is high.

31
6. Extreme Conditions

6.1 Overloads

The heating of different part of a transformer must not exceed the limits set by the standard
specifications. These limits have been set after years of service experience to ensure that transformers
may be operated permanently at their rated power and in the predicted ambient conditions for a
normal service life of 20 years.

The insulating materials used in transformers, whether solid (paper) or liquid (oil), undergo a
chemical alternation with time under influence of heat and other agents such as oxygen and moisture.

Montsinger, author of the celebrated ‘eight degrees rule’, takes a 50% reduction in the tensile strength
of an insulating material as criterion for the end of its life. He concludes that each 8C rise in
temperature reduced the life of the insulating material by a half,



t  t0  2 8
(4.1)
where t = duration of life
t0 = constant (corresponding to the length of ‘normal’ life
 = the difference in C from the normal temperature

Samples taken from old transformers show that cellulose insulating materials retain their properties
well beyond the arbitrary limit set by Montsinger. Paper which has become brittle is still capable of
resisting normal electric stresses due to the impregnating oil and the barrier effect. Usually it is the
oscillatory electrodynamic force during a short circuit which causes disintegration of the insulating
material, followed by a breakdown between turns or coils.

The degree of polymerisation, which is about 1300 for new paper, drops to about 150 for completely
crumbly paper. Taking this as a more realistic value for the limit of the useful life of paper, CIGRE in
1961 took 6C as a more correct value for establishing a guide to loading of transformers.

Operation of the transformer windings at a temperature less than 80C, corresponding to a reduced
load (60%), or at a very low ambient temperature (0C) causes negligible ageing of the insulating
material. It is therefore possible, whenever necessary, and under well defined conditions, to apply
loads that exceeds the rated power at which accelerated ageing of the insulating material is caused.

Overloads are usually classed as

 Recurrent overloads (normal load cycle) corresponding to planned or programmed power
demands, repeating over a reasonable period of time
 Exceptional overloads (emergency conditions) due to unforeseen conditions such as transfer a
damaged transformer or accidental load peak

Equivalent ‘normal day’s loss of life needs to be calculated for the overload conditions.

6.2 Overcurrents

Overcurrents are distinguished from overloads by their short duration (from tenths of a second to tens
of seconds) and by their relative intensity ( 2 – 25 times the rated current).

Overcurrents are principally of the following nature:

 Short circuit at the secondary Icc = 6 to 25 In,
 Motor starting currents Id= 3 to 7 In,

32
 Switching-in transformers I0 = 32 to 82 In.

The effects of these surges fall into three categories:

 Electrodynamic forces involving high stress in the transformer windings
 High temperature rise of the winding
 Drop in secondary voltage

6.2.1 Electrodynamic forces

Laplace’s law states that if an element dl of a conductor in which a current i flows is in a field of flux
density B, the force dF on the element of the conductor is given by
  
dF  i dl  B (4.2)

The calculation of the forces in a winding requires a knowledge of the field B at all points in the
winding. Roughly speaking, the flux at mid-height of the windings is purely vertical. The forces
arising from this axial field of flux are purely radial and for the two windings they are mutually
repellent: the exterior winding is subject to tensile forces while the interior winding is subject to
compressive forces, as shown in Fig.4.1.

Fig.4.1 Radial forces experienced by inner and outer windings

For windings of equal height and uniformly distributed ampere-turns, the length of the actual winding
is obviously limited. At the ends the lines of force B curve and cause axial compression in the
windings, as shown in Fig.4.2.

Fig.4.2 Axial forces experienced by windings

33
6.2.2 calculation of temperature rise due to overcurrents

Suppose the windings of a transformer are at an initial temperature of i and that they are subjected to
an over-current for time t.

This time can be regarded as being sufficiently short for the heating effect in the windings not to be
dissipated. These losses will therefore cause an adiabatic temperature rise in the conductors.

The final temperature f is given by

 f   i  aJ 2 t10 3 (4.3)

where a  8 for copper windings

t = time in seconds
J = current density in Aeff/m2 during the overcurrent period.

6.3 Overvoltages

Although transformers experience temporary ac overvoltages during the operation, lightning impulse
overvoltage and switching overvoltages are the most concerning overvoltages for the transformer
operation.

Overvoltages cause nonlinear voltage distribution along the winding and produce excessive overstress
on the insulation near the striking point of impulse voltage, especially on the interturn insulation.

Using different winding structure can improve the series capacitance between the turns of winding
and thereby the voltage distribution.

6.3.1Disc winding configuration

Disc type windings are commonly used in power transformers manufactured at UK. The earliest disc
type winding construction has the discs in the form of continuous disc, as shown in Fig.4.3, in which
the conductor is wound radically turn after turn, turns in one disc are physically and electrically
adjacent to each other. The electric turns are denoted by the numbers.

10 9 8 7 6 5 4 3 2 1

11 12 13 14 15 16 17 18 19 20

Continuous double-disc
Fig.4.3 Continuous disc type structure

For the interleaved disc, the electric linked turns, denoted by the numbers, are interleaved physically
as shown in Fig.4.4.(a). This increases the series capacitance and thereby improves the performance
under impulse voltages.

For the intershielded disc, shield turns are put in between the “live” turns which carry the transformer
current, as shown in Fig.4.4(b). This increases the series capacitance and thereby improves the
performance under impulse voltages.

34
 8 7 6 5 4 3 2’ 2 1’ 1
15 5 14 4 13 3 12 2 11 1
9 10 11 12 13 14 2” 15 1” 16
6 16 7 17 8 18 9 19 10 20

(b) Intershielde d double-disc

(a) Interleaved double-disc [ ‘, ” denotes non-active, shield turns ]

Fig.4.4 Interleaved and intershielded disc winding construction

6.3.2 Layer type winding

The arrangement of the windings in concentric layers tends to decrease the shunt capacitance and the
layers forming coaxial capacitances being connected in series. Therefore layer type winding tends to
have better performance under lightning impulse, as compared with the conventional continuous disc
type winding. For very high voltage, the distribution of voltage is improved even more by the use of
electrostatic screens arranged concentrically to the layers, connected generally either to the line or to
the neutral, or to both, as shown in Fig.4.5.

Neutral shield

Line shield

Fig.4.5 Layer type winding with shield at both terminals

References

1. S. Austen Stigant and A. C. Franklin. J & P Transformer Book, 10th Edition, Newnes –
Butterworths, 1998 (ISBN 07506 1158 8)
2. Colin Bayliss. Transmission and Distribution Electrical Engineering, Newnes, 1999
3. B. Hochart. Power Transformer Handbook, Butterworths, 1989. (First English edition,
translated from the French by C.E. Davison)

35