/

(

(

(

( ( ( ( ( (

TRANSFonMERS

<.,

'\

I

{

.( I
I
( I
I
( I
j
(
(
Y. ) ;

. I \

TRANSFORMERS~

,

Bharat Heavy ElcctricaIs Limited' I Bhopal'

) '.

; ...

) (

\ (

) (

Tata McGraw-Ifill Publisbing Company Limited NEW DELli!

------

\

(

i

\

\

'_

© 1987, BHARAT HEAVY ELECTRICALS LTD .• PIPLANI. BHOPAL (M.P.)

Third Reprint i 991

RXDYCRY~lRXYCX .

No pan of this publication ('=:'1 be reproduced in any form or by any means without the prior written pe::-::ission of the publishers

This edition can be exportec from India only by the publishers. Tara McGraw-Hill Publishir.; Company Limited

--\

Published by Tara McGraw·Hill Publishing Company Limited. 4/12 A53:f Ali Road. New D~L'li-11 0 002 and printed at

Rekha Printers Pvt. Ltd .. ~ ;?\ ... Delhi-I 1 0 020.

"-- ( (

-,

_ i

(

\. . ~ \

~_ i

J.

-'-

\

Foreword

• I

I am happy to introduce this vbook on power transformers covering various aspects of design, manufacture and operation.

BHEL started' manufacturing' transformers in 1961 and has since made significant -strides. This book, written by nineteen experienced practising engineers, reflects the expertise gained by BHEL in the field and offers a wide and thorough coverage. I find the work informative as it follows a practical approach, lacking in most of the other books on the i subject.

The transformer industry is now more than a century old. Significant advances have been made and the growth in technology continues at a rapid pace. I sincerely hope that this book will serve the need of a useful practical reference book.

S. S. DASU Director (lS&P) BH EL, New Delhi

'<,

; . I

:

_'

Preface

The book covers' all the aspects of power transformers-basic fundamentals of the theory, design principles, manufacturing technology, materials specifications and standards. It also covers special transformers including reactors and inverter transformers for HVDe systems. It will serve as a reference book for practising engineers engaged in the design; manufacture, planning, testing, erection, operation and maintenance of transformers. It will also be useful to students and researchers.

This book is the collective effort of nineteen experts from Bhurut Heavy Elcctricals Ltd. (BHEL), a leading transformer manufacturer in India. An editorial committee went ·over the complete material to integrate it into a homogcnousbook and to ensure complete continuity between the chapters.

Bharat Heavy Elcctricals Ltd. (Bll EL), one of the largest engineering organizations in the country, is engaged in the design and supply of heavy equipment and total systems for generation, transportation and:.

utilization or energy. i

I

R. K .. S r \ll./"E

. Executive Director BH EL, Bhopal

Acknowledgement

', I

The editorial committee' acknowledges with thanks the permission given: .. by the Management of Bbarat Heavy Electricals Ltd. for publishing tbis book.

Thanks arc also due to all authors who contributed diflcrcnt :

chapters of this book. I

To integrate the entire material and present it at a standard lcv(;:, : at the same time, ensuring continuity between chapters was an uphill !. task. Our grateful thanks are due to the editorial committee members : who have put in exemplary efforts without which this book would not.: .

have taken this shape. i

i

T. K. 11uKHBRJEB \

Additional GC!1e:-z.: i Manager (Transformers :

Capacitors. Bushings)

.--~-.- ...

;

I

f.

. ,.

\

i·

Editorial Committee T.K. Mukherjee M.L. Mittal

I.C. :rayal

S.C. Verma

C. M. Sharma

V.K. Lakhiani

I •

Contents

1 I

\ \

I i

.1

I

I I !

I

i

vl'i '.' vf

J

IX

I ;

2'

I .

\5 .

I ,

! .

)

Foreword Preface Acknowledgement

1 Introduction V.X. Lakhiani

2 Principles of Transformers R. K. Singh

3 Materials Used in Transformers M. P. Singh

4 Magnetic Circuit K. N. Labb

5 Windings and Insulation

6 Voltage Regulation and Tapchanger B. L. Rawat ..

7 Electromagnetic Forces in Power Transformers 8 Cooling Arrangements

C .. M. Sharma

9./Design Procedure.

<:>: R. C. Agarwal '

10 Structural Design of Transformer Tank M. K. Shakya

11 Transformer Auxiliaries' and Oil Preservation Systems ,S. C. Verma

12 Assembly

V. B. Bhatia

'1 I

30

114

! .

)

I

128

;

146

\

I

165 I

212

;

229

13 Drying and Impregnation .M. P. Singh

14 Testing of Transformers and Reactors P. C. Moluijan Gild 1I1. L. Jnln

15 ..... Standards on Power Transformers

~O"~

':\.,/~ V. K. Lakhian!

233

245

.. ~

.------.~---

xlt Contents

16 Loading and Life of Transformers D. P. Gupta

17 Erection and Commissioning . C. M. Shrivastava

18 ...... Transformer Protection '~./ B. L. Rawat

19 Rectors

C. M~ Shrivastava

20 Traction Transformers J. M. Malik

21 Rectifier Transformers J. S. Sastry

22 Convertor Transformers for HVDe Systems I.C. Taya;, C. u. Sharma and S. C. Bhagerla

Index

312

370

384

393

411

420

445

"4)6

--_ .. _---.

--- - --_ -------- - -_-

"

i

(

1

Introduction

V.K. LAKHIANI

Senior Manager (Design), Bhopal I

)

i

;

A transformer is a device with two or more stationary electrical circuits

that are conductively disjointed but magnetically coupled by a common. ,. time-varying' magnetic field, Transiormers are basically passive devicd

for transforming voltage and current. One of the windings, generall~ termed as secondary .winding, transforms energy through the principle

of mutual induction and delivers power to the load. The voltage level4

at the primary and secondary windings· are usually different and any' increase or decrease of the secondary voltage is accompanied by corres-' ponding decrease or increase in current. .

Transformers are among the most efficient' machines 95% efficiency being common in lower capacity ranges, while an efficiency of the order of 99% is achievable in high capacity range. Theoretically, there is. no upper limit to the power handling capacity; transport constraints. handling facilities; etc. being the limiting factors. The lower limit is gO'/ern~ by the allowable no-load loss. :.

The physical basis of a transformer is mutual inductlon'between two! circuits linked by a common magnetic field. The primary circuit carrying a: current has associated with it, as a 'manifestation of the electrical phenomenon of current flow, a magnetic field in its immediate vicinity. When the; circuit is alternating, the magnetic field at any point in the surrounding medium will vary both in magnitude and direction in accordance; with the change of current with time. The secondary circuit being. in the! vicinity of the primary circuit will link some of the magnetic flux produced: by the primary. With an alternating primary current, and therefore flux,!

. • I

the changing linkages will produce in the secondary winding an e.m.f.,

The more closely the primary and secondary circuits are mutually linked,' the more direct becomes the exchange of energy between them. Chapter i deals with the theory of transformers and fundamental concepts of trans-i former operation. In introducing the theory, 'equations of real and ideal] transformers have been described in corresponding equivalent circuitsMajor electrical parameters of a transformer, viz., j ron and copper losses;'

l ,

~

._.:...;:;::; .' ~~

-_. -_._--- ---------------- - -------

1 Transformers

efficiency, regulation and parallel operation -have been described and three winding transformers arc. introduced.

Essentially the chief elements of construction of a transformer comprise materials for magnetic circuit, terminals, tapping switches and tank, oil as well as cooling devices. While manufacturing practices differ from manufacturer to manufacturer, the quality of the materials used, their guiding specifications and mode of usage, have profound influence on the active material content and long-term reliability of the equipment. In fact, the life span of a transformer depends, to a great extent, upon its insulation system; an amalgam of different insulating materials, processes and their interactions. The characteristics of an insulation system are influenced by electrode, geometry, composite 'dielectrics, temperature gradients and compatibility of different materials. Chapter 3 describes the materials used in transformers, their properties and applications.

The core forms the magnetic circuit of a transformer. Special steel of high resistance and low hysteresis loss is almost exclusively used in transformer cores. Over the years, continuous improvements have been made by steel 'manufacturers in terms of properties of core steel. It has, therefore, been considered appropriate tointroduce briefly reference. to such developmeats as wellas thecoustructionalfeatures and manufacturing aspects of cores: The emphasis has been primarily on 'core-type' construction which is more common these days rather than the 'Shell- Type'. Design principles of core are also adequately covered., Important topics such as permissible hot spot' temperature ,~nd overfluxing capabilities, which are still subjects of technical debate, have been covered briefly.

The windings, which form the electrical circuit are dealt with in Chapter 5. Any transformer winding must fulfil certain basic' requirements, particularly the dielectric, thermal and mechanical stresses imposed on it during testing and in service. This chapter is devoted to different types of winding, their behaviour with respect" to surge voltage' distribution and heat transfer. Since electrical circuit is .the costiiest part of a transformer, designer's approach is to achieve optimisation in terrnsof configuration, dimensions and electrical clearances. Different manufacturers, based on their experience and shop practices choose different types of windings. The chapter therefore covers a general-treatment of these aspects.

In order to control large high voltage networks and to maintain correct voltages, it is essential to provide some means of voltage variation in a power transformer. This variation is generally achieved by means of tappings on the windings. Tap changing equipment is thus an essential accessory to a transformer .and fall into two categories, (a) those by which turns

ratio .are changed when the transformer is electrically disconnected from the system on both sides and (b) those by which changes are carried- outwithout interrupting the load current. On load tap changers, particularly when associated with high voltage systems have to be designed to meet the surge voltages arising under impulse conditions. A tap changing equipment is thus an important accessory demanding high reliability and

.J

-I'

long term dependability. Chapter, 6, covers tho, principles of tap .Changing . as well as tho constructional and operational features of different ty~ of tap changing equipment.

In service the windings of a transformer are subjected to electro-magnetic forces under short circuit conditions. Tho windings must. therefore.: be designed to withstand such forces. This important aspect of design is covered in chapter 7. While it is feasible to verify the design through abo'> circuit testing of power transformers of lower ratings, such verificativE .for largo unit ratings is-precluded due to limitations of testing facilities. j

Chapter 8 presents an account of the principles included in design o'f cooling circuit as well,' as tho various cooling methods. Tho primary objective of the design is to ensure that hot spots in the windings or' mechanicai parts are avoided.

./ . I

) ,1//

)

"

Theories pertaining to design procedures, criteria of selection of core, winding, flux density, current densities, calculation of-number of turns," reactances, weights, and losses havo been discussed in chapter 9. Rapid

I

developments in computer science presently enables a designer use of

computer programmes to arrive ,at optimum designs and predict.performance 'under short circuit condition. It is also possible i.o find outth6

" surge voltage distribution, fieJd distribution pattern as well ~ stray 10s~ ,

, control parameters. . !

Similarly Chapter 10 provides a description of the methodology adopted for structural design of transformer tanks including the principles governing

, the classical and Finite Element methods of tank design. This chapter also' deals with the methods generally applied for measurement of stresses. Chapt ter II, on the other hand, is devoted to the description of various fittings and accessories with special reference to oil preservation systems in vogue. i

It is apparent that 'the inherent advantageous properties of improved varieties of cold rolled grain oriented sheet steel, insulation-materials and other items of construction can only be realised through accurate manufacturing methods, high degree of workmanship and controlled drying and processing techniques. Chapter 12 covers tbe aspects' relatedtd

I

assembly while Chapter 13 deals with the: drying methods including th~

latest technique of vapour phase drying and impregnation.

The establishment and achievement of a high standardof both materials and workmanship is possible through continuous inspection during thd manufacturing stages and subsequent testing of the components and thd finished product. Of all the operations-which attempts, achievement of:

, such a standard, no influence is greater than that of the system of tests to: which the product is subjected, and the final tests. Chapter J4 covers the testing codes and various methods employed in testing.

Chapter 15 provides details of the standards on transformers while, Chapter 16 deals with the loading and life with a view to serve as a: reference for application engineers.

)

) )

) )

)

)

)

)

)

;

·i

rt#fi!'P""",_,_,,_.

___ .. ._~ _ _c_ .,---...,..~ __

Transformers are the largest, heaviest and often costliest single piece of equipment in a power system. Obviously appropriate care is necessary in' dispatch, erection, commissioning, operation', and maintenance' of transformer's. It is .now generally recognised .hy the operating engineers that it pays to inspect th~ transfor'~e~s in: service regularly and maintenance is, as Important a factor as careful design and manufacture for onsuring trouble. free, reliable service. In Chapter 1 i .description of such aspects includmg"precautions to be taken at site have been' covered. Topics such as dissolved igas analysis and partial discharge testing at site have also been dealt with briefly. Chapter 18 provides guidelines for transformer protection as a ready reference.

The last three chapters of the monograph have been devoted to transformers of special category Chapter 19 covers in brief different types of reactors while chapters 20 and 21 cover special features of traction transformers and rectifier transformers, essentially low voltage DC transformers.

This treatise on power transformers is the.resukof contri.bution from a number of practising engineers of Bharat .. Heavy Electricals Ltd. an Organisation which, besides special transformers, has so far designed, manufactured and supplied more thatffifty five per cent (80,000 MV A) of tho power transformers in Indian grid. Tho publication is intended as a reference for all practising.electrical engineers and covers both theoretical as well as practical aspects of transformers. It.does not cover instrument

transformers. :_ ..

\ ,

) ,

2

I

! l.

,r',

Principles of Transformers

I •

R. K. SINGH

- )

Senior Manager (Production), Jhdnsi

I I

! .

,

2.1 WORKING PRINCIPLE OF A TRANSFORMER

I

I'

A transformer is a static piece .of apparatus used for transferring power

from one circuit.to another without change in frequency. It can raise !or lower tho voltage with a corresponding decrease or increase in curroht:

In its simplest form, a transformer consists of two conducting coils having a mutual inductance. The primary is the winding which receives electric power, and the secondary is the one which may deliver it. The coils art;

wound on a laminated core cf magnetic material. . !

I

The physical basis of a transformer is mutual inductance between two

circuits linked by a common magnetic flux through a path of low reluctance

as shown in Fig. 2.1. - I

LAMINATED CORE

PR

4~ I = ~
;1
(!
(.--------------¢------- -'.
·
• r
, I
I •
.- I
I
: i
!MARY j SECONDA
i i
: , ~
1 r
i
i I
• i
I
I
I I
• I
, I
• ;
"
-- - - - - - - - - -- - - -- - - - - - -- ---" , i RY

Fig.2.1 Diagram showing magnetic circuit and windings of a transformer.

6 Transformers

The two coils possess high mutual inductance. If one cuil is connected to a source of alternating voltage, an alternating flux is set _ up in the laminated core, most of which is linked up with the other coil in which it produces mutually induced emf (electromotive force) according to Faraday's laws of electromagnetic induction" i.e.

di e = M dt

e = induced emf

M = mutual inductance

If the second circuit is closed, a current flows in it and so electric energy

I

is transferred (entirely magnetically) from the first coil (primary winding).

to tho second coil "secondary winding).

where

2.2 ELEMENTARY THEORY OF AN IDEAL TRANSFORMER

An ideal transformer is one which has no losses, i.e. its. windings have . no ohmic resistance and there is no magnetic leakage. In other words, an ideal transformer consists, of two coils which arepurely inductive and wound

on a loss-free cO,re. .', . .

It may, ho,W.~ver, be noted that it is impossible to realize such a transformer i[lpracHce,~yet for convenience, we will. first analyse such a transformer and then an actual transformer.

V2 = E 2 t

Cal (b)

Fi~. 2.2 Vectorial representation of induced emf ofan ideal transformer under no-load.

. .

Consider an ideal transformer (Fig. 2.2 (a)] whose secondary is open

and whose primary is conn~ted to a sinusoidal alternating voltage VI. Under this condition, the primary draws curren~ from the source to build up a counter electromotive force equal and opposite to the applied voltage.

I

.\

\ \

I I

" '

-;

~

f ,.C .... -.

Principles of Transformers :7

Since the' primary coil is purely inductive and there is no output, t~e primary draws the magnctising current I,. 'only. The function of thjs . current is merely to magnetise the core, it is, small in magnitude and latis 1'1 by 90~. This alternating current I,.. produces an alternating flux cp which is proportional to the current and hence is in phase with it. This changing nux is linked with both the windings. Therefore, it produces self-induced emf in the primary. This self-induced emf e) is, at any instant, equ;1"i ~ and in opposition to VI. It is also known as counter emf of the primary.

Similarly in the secondary winding. an induced emf e2 is produce.' which is known as mutually induced emf: This emf is in phase opposition with VI' and its magnitude is proportional to the rate of change of flux and the I\UIl1[)l'r of sccondury IIII'IIS'. Figure 2.:!(b) SIIO\VS the vectorial rcprcscntnt ions

of tho above quantities. :

2.3_ EMF EQUATION OF A TRANSFORMER

N, = Number of turns in primary N2 == Number of turns in secondary

tPm . = Maximum flux .in the core in webers Em = Flux density in wcber/sq m (Tcsla)

A =. Net cross-sectional area of core in sq m f == Frequency of _J,C input in Hz

VI = Instantaneous value of applied voltage in pnmary winding in volis

VJm = Maximum value of applied voltage in volts.

The instantaneous value of counter electromotive force er is

ei = -Nl .~~ volt -, (2.1)

As discussed above in Sec. (2.2) the counter emf el is equal and opposite to applied voltage VI, i.o,

VI" =N. de/>

-, • dt

Let

--

lethe applied voltageis. sinusoidal, that is VI = VIm sin 21ift

Then Hence -

rfo = e/>m sin 27ift

el = - NI e/>m cas 27T If x 27T J

These equations are expressed as vectors as shown in Fig. ?2 (b), whcr~ VI and EI are thcrrns values of VI and, CI. To obtain the rms value of "counter emf et , divide its maximum value given above by V 2.

Then

(2.1.1) -

8 Transformers

The cosine term ha~ no significance except to derive the instantaneous ,va.lues.

1.0.

E: = 4A4 INl ,pm EI = 4.44JNIBmA

(2.2)

or.

Similarly nus value of emf .induced in secondary is, E2 , 4.44 J N'l SmA

(2.3)

r n an ideal transformer VI-EI

" ,

and V2 = E2

where V2 is the secondary terminal voltage [Fig. 2.2 (b)].

2.4 VOLTAGE TRANSFORMATION RATIO (K)

From EqSA (2.2) and (2.3), we get

~~=~ -K"

(2.4)

This constant is known as voltage transformation ratio.

(a) If N2 > Ni, i.e., K> 1, then the transformer is called as stop-up transformer.

(?) If N2 < l'iI,i.e., K < 1, then the transformer is called '~:; step-down transformer.

Again for an ideal transformer

Input = Output

VIlt == V2h (neglecting I,..)

or

(2.5)

where It and Iz are primary and secondary currents.

Hence the currents are iII the inverse' ratio of the transformation ratio.

2.5 IDEAL TRANSFORMER

We will consider two cases,

(a) 'when such a transformer is on no-load and

_ (b)_ wheuit jsIoaded. '

(a) Transformer on No-load

The primary-input current under no-load condition 'hal to 'supply (i) ironloss in the core 1.0'" hysteresis loss and eddy current loss and (ii) a very

..

. ;

. (

Prlnctples of Transformers ~

small amount of copper-loss in primary. 'Honce the no-load primary input current 10 is not at 90° behind VI but lags it by- a,n anglcBo which is Jes~ thar1.'90°. No-load primary input power Wo = VI lo'cos 00" No-load condition of an actual transformer is shown vectorially in Fig. 2.3. i

I.w

-._, 71

__.,..__J --

--- 1> If..

10

E,

Fig.2.3 No-load current of an ideal transformer (having core losses.only).

As seen from Fig. 2.3, primary current 10 has two components.

(i) One in phase with VI. 'This is known as active or working or ironloss component Iw, because it supplies the iron-los's plus a small

quantity of primary Cu-loss. - i

L; = 10 cos 80 (2.6)

(ii) Till: other component is in quadrature with VI and is known as magnetizing component because its function is to sustain the alternating flux in the core. It is wattless.

11'-= 10 sin 00

Obviously 10 is the vector sum of l ... and 11'-, lienee

10 = v' tj~2+1w2)

~2.7)

(2-8)' '

The no-load primary current 10 is very small as compared to full-load prim:l.ry current. A"" [0 is very small, hence no-19:1-<1 primary copper-loss is lle~ligibly sm.rll which 1111'::11\$ t.h.u no-loud primary input is practically equal to the iron-loss in ;L transformer.

~'-""'-.

10 Transformers

(b) Transformer on Load

.

When the secondary is loaded, secondary current h is set up. The

magnitude of h is determined by the characteristic of the load. The secondary currentsets up its own rnmf (= 112 Ii) and hence its own flux ,p2 which is in opposition to the main primary flux 4;, which is due to 10:

The opposing secondary flux ,pi weakens' the primary flux momentarily and primary back emf Ei tends to reduce. For a moment VI gains the upper hand ,over £1 and hence causes more current Ch') to flow in primary.

The current h' is known as load component of primary current. This current is in phase opposition to current /Z. The additional primary mmf NI hi . sets up a flux <P2' which opposes CPl (but is in the same direction as cp.} and is equal to it in magnitude. Thus; the magnetic effects of secondary

.. current Ii get neutralized immediately by additional primary current Iz', The whole process is illustrated in Fig. 2.4. Hence, whatever may be the

.....

1 __ ___; __ J

. I

,- - - -<I>- - - •• f t: -~----. I

I I I

.,.. .-~

.. - -<b- - -;

LOAD

-----(

'------!

\ I

Flr..1A Transformer on-load.

... _"'~-:"""'-":~'

i~ ·;~.i·

, . \

)

.i
) l'
)
!
)
)
j )

.. , .

J~.

-....-._

j

load conditions, ~h~ net flux passing through the core is approximately the same as at no-load. Duo to this reason the core-loss is also practicallv the., J

I

\

same- under all load conditions.

As

,pl = (P:zo' Nl12 = N1h' Nz '

12'= --xh=Kh.

NI

.r ..... ~I'. 111,

l'rinciple.t OJ ~ ransjormers

i

Hence, when, transformer is on load, the primary winding has two :' currents 10 and Iz' (which is antiphase with 1~ and K times its magnitude) .. :

The total' primary current. is thevector sum of io .and h'. In Fig. 2.5 arc

. "

£2 £2 E.2
K= 1 '. K~ 1 K= 1
( 10 NEGLECTED)
td) (b) ( c) . '"
" Fig. 2.5. Vector diagram of an ideal transformer. .

_"

shown the: vector diagrams for a loaded transformer. In Fig. 2.5 (a), .current h' is in phase with E2 (for non-inductive loads). In Fig. 2.S(b), it

is lagging behind E2 (for inductive loads). .

If we neglect 10 as compared to h' as shown in Fig. 2.5(c), then g,1 = ~2

and thus '

NJl2' = Nt/I = N212

II N2

--=.-. ==K

h NI

It shows that under load conditiQPs_" the ratio of primary and secondary

currents is constant. ----------- .. -. -,-----, ... ---- --.

12 Transformers

2.6 TRANSFORMER HAVING WINDING RESISTANCE sur NO MAGNlttlC LEAKAG'[--'_'---' ----.---.--.-- .. - .. --.-- ... -.~- -"-

, .~ ,~ -. ..._ .. - _ .. _._ ._ _ .. ---. -.,.~ , ...

An ideal transformer was supposed to possess no resistance but an actual transformer has primary and secondary windings with someresistances, Due to these resistances, there is some voltage drop in the two windings. The result is that:

(a) The secondary terminal voltage V2 is equal to the vector difference of the secondary induced emf £2 and h R2 where R2 is the resistance of . tho secondary winding.

(2.9)

(b) Similarly primary induced emf EI is equal to the vector difference of VI andfr Rs where Rl is tho resistance of tho primary winding.

(2.l0)

The vector diagrams -for-non-inductivc, .inductive and capacitive loads

are sM0w.n in Fig. 2.6 (3.). (brand (c) respectively. '

i:

'\

l V2 I2R2 t

£2 K: 1

(a)

£2 K~ t

(b)

(c)

1 NEGL sc -:-r: D

... 0

I

._- .. _\,

Fig.2.6 Vector diagram of an ideal transformer (having Core losses and

winding resistance). '

Prlnclples 0/ Transformers 13i

/0

, 2.7 MAGNETIC LEAKAGE

, . ~

=-~- ..

In an ideal case it, is assumed that all tho flux linked with tho primary: winding also links the secondary winding. But, in practice it is impossible:

I

to realize this condition as magnetic flux cannot be confined. The greater;

portion of the flux (i.e., the mutual flux) flows in the core while a small; proportion (Fig. 2.7) called the leakage flux links ono or the other winding, !

MAIN FLUX

" ,

I

LEAKAGE FLUX

',I .. __ - - _- __ -

: ::'.~:;~:~::~:::_::-_:::-_~~_-~~ ~ ~: :-~";':; i r :

, II ' I .: ,- - - _ .- _, - - - __ " .• I I I, .:' ,.-

I ,,' I r1~ _ ., , \ ' .r : -', I., ~ , . " '

,J I I • •. " ' '. ~ I I I I :

I' I I I I., •. I. I I II ,I I. I I

I I I 11'1' • ' I' t ,I J I

• ", I ' , . 'I . I I"

I I I, 1'1 I r I I

I. 1,1 I I I • I I I I I I I I

:: :'JI'~' : ;.:: '::::'" "

I I, ' ' I I'

I ' " I I I I I' I"" •

I t I I , I I "1"

................. ~-il II I I , . , I I, "1-'-~..x.o.::)Q

.' '." , II I I • "" ... - • - -. I I • I I'·

, './ I I I I " ' ... '_ .. ... ' ..... _ _ _ _ _ ,J ; I I I "';, ... '" .... _.

,. [': <r.r:': -_:.. -_: : :.._ -:: :: :. -":.: '.' I I '._.,

, I I ._ - - - - - - - - - - - - - - - - ... -": • ' . - - - - - - -- - - - - - - - - - - - _ ..

• _ - - _ - - - - - - - - - - - - - - - - _ •• 0

, '

" . ,

,_ ... - ...... ... ,'. , \ ,

', ,

, \

,,_ -

Fig. 2.7 Leakage flux.

but not both. On account of the leakage flux, both the primary and! secondary windings have leakage reactance, that is, each will become the; seat of an emf of self induction, of a magnitude equal to a small fraction; of the emf due to main flux. The terminal voltage VI' applied to the: primary must. therefore, have ; component 11 ,XI (where XI 'is leakage: reactance of primary) to balance the primary leakage emf. In the secondary, I similarly, an emf of self induction 12 X2' (where X2 is lcakago.rcactancc or; secondary) is developed. The primary and secondary coils in Fig. 2.7 are shown 'on separate limbs, an arrangement that would result in an exccptionally large leakage. Leakage between primary and secondary' could be. eliminated if the windings could be made to occupy the same space. This.. of course, is physically impossible, but an approximation to it is achieved: if the coils of primary and secondary are placed concentrically. Such an; arrangement leads to a marked reduction of the leakage reactance. If on the other hand. the primary and secondary arc kept separate and widely: spaced, there will be much room for leakage flux and the leakage reactance: will be greater.

2.8 TRANSFORMER \'lITH RESISTANCE AND LEAKAGE REACTANCE

Figure 2.8 shows the primary and secondary windings of a transformer

14 Transformers

Fig.2.8 Resistances and leakage reactances. 'I

with' resistance and leakage reactances taken out of the windings. The

primary impedance is; given by

ZI = V (R12+X12)

And the secondary impedance is given by'

Z2 = .y(Rl+X22)

Tr ... 1 .,/n; ''V) E_1J Z VI = .CIT.Ll\L\.I·,-j~l -. 1.- 1 1

E2 = V2+h(R2+jX2) = V2+hZ2

and

/.

(2.11) (2.12)

The vector diagram of such a transformer for different kinds of loads is shown in Fig. 2.9., In these diagrams vectors for resistive drops are drawn

----

I

I /

. ---(

Fig.2.9 Vector diagram of transformer.

..; 1',

Principles oj Transformers 15 I

parallel to current vectors, whereas rCB:~tive drops arc perpendicular to the . currentv,gctorL, The anglo O. between VI and 'I gives the 'power;factof -. , angle of the transformer.

"

"

2.9 EQUIVALENT CIRCUIT

) ,

I The transformer shown in Fig. 2.10 (a) can be represented by an cquiva]

lent circuit in which the resistance and leakage reactance of the transfr.rmer arc imagined to be external to the winding. The no-load current fu j~

. ,

simulated by a pure inductance Xo taking the magnetizing: component I,i.

and a n·oninductive resistance Ro taking the active component I", connected in parallel across the primary circu it as shown in Fig. 2.10 (b).

)i

j I

J f

) I

I
E2 E2
~ 1
IDEAL !
TRANSFORMER \
(0) (b) ·1 Fig.2.10 Parameters of transformer •

. To make the transformer calculations simpler, it is preferable to trans fer voltage, current and impedance to the primary side.

The primary equivalent of the secondary induced voltage is-,

E2' = E2/K = £1 (where K is the transformation ratio)

Similarly, primary equivalent of secondary terminal or output voltage is V2' = V2!K .

Primary equivalent of secondary current is 12' = 12K

and

R2' = R2IKl, X2' = X2!K2, Z2' = Z2}K2

The same relationship is used, for shifting an external load impedance to primary. The secondary circuit is shown separately in Fig. 2.11 (a) and its equivalent primary values in Fig. 2..11(b).

!

16 Transformers

(a)

, I 2

R:2=R2' K, ..

Fig.2.11 Secondary resistance and reactance referred to primary.

(bl

Vl

rl

The equivalent circuit with secondary parameters transferred on primary side is given in Fig. 2.12. Thi~is' known as exact equivalent circuit. A

Zl 72

E2=El

r

Fig. 2.12 Exact equivalent circuit.

simplification of the ,equivalent circuit can.be further made by transferring the eiditi:tig cltctiita:cross the input terminals as shown in Fig. 2.13 or in Fig. 2.14.

10

I (

v,

10

1,

Fig.2.13 Simplified equivalent circuit.

\
f
\
)
~o
( o-
J
"
!
) PI'inc/ples 0/ Transformers

L

XO V2

12 L

,-'----- __ --_ __ .. l

Fig. 2.14 Simplified equivalent circuit.

It shou ld be noted that in this case R" r- ':> VIII... und Xo = VI! r,.. Further simplification may be achieved by omitting fo altogether as shown III

Fig. 2.15 (a).

10

['

,./

ZOI

/

r

Fig. 2.15 (a) Simplified equivalent circuit.

,

",

Equivalent circuit for a three-winding transformer described in c 1. 2. 14!

is given in Fig. 2.15 (b) as a star network.

l o-----.JIW\!v\,---/Qi5Drj'>---

R, x\

Fig. 2.15 (b) Equivalent circuit for three-winding transformer.

18 Transformers

From Fig. 2.12 it is found thut total impedance between the inpu t torrninals is

(2.13)

where

ZI1I = impedance of the exciting circuit = RaJ jXo

VI = lx[Z I + --:::-_::0:-1 ]

-:: 1 1

Z::!'+ZL'+ Zm

(2.14)

2.10 REGULATION

When the transformer is loaded, with a constant primary-volrage, then tho secondary terminal voltage drops (assuming lagging power-factor): it will increase if power factor is leading, because of its internal resistance and leakage reactance.

Let

Vl = secondary terminal voltage. at no-load V2' = secondary terminal voltage at load

. ' .. _ r'V2~~ V2' ]

% Regulation =l. V2 _. x 100

.

Then, % regulation of a loaded transformer at any power factor is given

as

....... ,~,

-.~. '

( • 1." ) eX cos e-R sine)?

= R cos e -:'.1\ S.In 0 + 200 -~--.

R = percentage resistive drop X = percentage reactive drop cos e = lagging power-factor

It is to be noted' that in case of leading power-factor, fi will change to- e. Typical calculations for computation of percentage regulation at different load and power factor are given in Appendix 2.1.

(2.15)

where

2.11 LOSSES IN TRANSFOR\'lER-';~

LOS:;6S in transformers are

(a) No-load Losses

It includes both hysteresis loss and eddy current loss.

As the core flux

(

I /

I (

,

I (

's (

I ,

I I

I

I I

)

I

r

r i !

I

I

( I

::.L.

--

Principles (If Transformers 19

in a transformer remains practically constant at all loads, the core-loss is also constant at all loads.

Hysteresis loss WII = Kill Bm'·6 watts Eddy current loss W" = KcflKiBm2 watts

(2.16) (2.17)

where

Kit = the hysteresis constant

K" ;;.:" the eddy current constuut Kr = the form factor

These losses ar.£.minimik~d_llY_lJ:s.ing_ .. ~~l of high silic'on content - for the core and by using very th in. laminations _ The input: power of a transformer, when on no-load, measures the core-loss.

(b) Load Losses

This loss is mainly due to the ohmic resistance of the transformer windmg. Copper-loss also includes the stray loss occurring in the mcchanicul

t ucture and Wl·~A:~~ ~,.. rI .. ,.t" .. rill" t" tl-.". .,t,."" Anv.,.., Copper-loss

sr. .J\"J' b ·v _ ,, __ --- - _ _J " .

(J2R-Ioss plus stray loss) is measured by the short-circuit test.~,..

.F

2.12 EFFICIENCY

% Efficiency - Output X 100 Input

= (Input- Losses} X 100 Input

I. J..-{ ... ... - ., I"/! t:...\;\-,-:/:~ .•. ;~~/=( 1- i::~;S) X 100 !

\ --- .-.-.. - ._-- - '-- - ._-

2.12.1 Condition of Maximum Efficiency

Copper-Loss We = lJ2 ROI = h2 R02

. where ReI and Ro2 are equivalent resistance referred to primary and ary sides respectively.

Iron-Loss Wi = Wh+We

Primary input = Vili cos 01

E·ffi . [1 (h2Rol-- WI) J clency -I) =- .-

ViII cos 8J .

second-

_[ 1- llRol - Wi ]

-. V, cos 8, VtIJ cos 01 .

Differentiating both sides with respect to It we get

dT) _ 0-( Ro, )_'_( WI _)

dII - VI cos e 1 1 \ VxIJ2 cos 81

(2.18)

20 Transformers

for '1j to be maximum, dYj

dl, ,= 0,

hence the above equation becomes

Rot W,

VI cos e, = VIJ2t co's 0,

or W, = 112 Rot

or = Ii2 R02

or Iron-loss = Copper-loss

i.e., maximum efficiency occurs when iron-loss is equal to copper-loss, The load corresponding to maximum efficiency is given by .

/ ( Iron-loss . "')' - -

- tJ .rcu ... loadcopper_ldss;.>'<,F~ll-load

,Typical calculations fPI>CO])lPJltation of';percentage efficiency at different loads and power factorsare giyen in Appendix 2.1.

,-,';,-" ,'._" ,;" _,' ,'.: ,> " '. . <

I I r

( ! I

2.13 THE AUTO-TRANSFORMER

,.1'::

The auto-transformer has a single continuous winding which is used for the input and output voltages, as shown- in the Fig. 2.16. A portion Be' of the primary winding AB is used as secondary winding. _ It is used where

~ .

I,

...

V,
I
I
I
I
I
,
i
,
I
b -.::::
11 Fig 2.16 Auto transformer.

(

Principles 0/ Transformers 21

~(

,

transformation ratio differs a little from unity. Its theory and operation: principles arc similar to that of a two-winding transformer. If NJ is tho: primary winding turns and N']. (a portion of N,) is secondary winding I turns, the transformation ratio K can be represented in the same way as a ; two-winding transformer :

.V2, = N'!::_ = K V, n,

Tho current in tho secondary winding (may be called as common wind-I ing) is the vector difference of I: and I,. But as the 1 wo currents arc prac- : tically in phase opposition, the common winding current may be taken as i ~ritbmetical difference of hand 11 i.e. (h- I.) where Iz is greater than I,. i

j

2.13.1 Savings of Copper in Auto-transformers With reference to Fig. 2.16.

Weight of copper ill AC section (series winding) is proportional to (NI-N2)1J

Weight of copper in BC section (common winding) is proportional to R2(h-II)

So, total weight of copper in auto-transformer is proportional to (NI- N2) II + N2(h- II)

In a two-winding transformer,

total weight of copper is proportional to (NIh +N2h)

..

/

(

, "

)\.

Weight of copper in auto-transformer

- Weight' of copper in two-winding transformer

(NI- N2)lJ + N2(h- II) NllJ+}h12

,

i .

I'

t .

'\ i

(2K)

- 1- --2 = (I-K)

(where _!I_'! , II = K)

NI 12 ,

or weight of copper in auto-transformer (H/a)

== ([ _ K) x r weight of copper in ordinary tWO-J

L winding transformer (Wu)

_' .......... ..:0. -

21 Transformers

Hence, saving == Wo-Wa = Wo-(1-K)Wo = KWo

or Saving c::: Transformation ratio (K) X weight of copper in

, ordinary 'two-winding transformer

Thus, it can be visualized that saving will increase asK approaches unity.

2.14 THREE-WINDING TRANSFORMER

2.14.1 Purpose

A transformer may have~dditioml1 windings. apart from the two conventional main windings depending upon the particular application and type of connection (of the main-windings) .. In th!'ee-win<1ingJ~C\..nsformers, the third winding is normally called as tertiary winding and it is provided to meet one or more of the following requirements:

(0) For an additional load which f9r some reasonmustbe kept isolated from that of secondary ..

(b) To supply .phase- compensating devices, such as condensers, operated. at some-voltage net-equal to primary Or secondary or with some different eonnection (e.g. mesh).

(c) Instarjstar-cormected transformers, to allow sufficientearth fault current (zero-sequence component current) to flow for operation of protective gear, to suppress harmonicvoltages and to limit voltage unbalance when the main load is asymmetrical, the ~ary winding is delta-connected.:_,.

(el) As a voltage coil in a testing transformer. (e) To load charge split Winding generators.

(f) To inter-connect three supply systems operating at different voltages. Tertiary windings are mostly delta-connected.

Consequently, when faults occur On the primary or secondary sides (particularly between lines and earth), considerable unbalance of phase voltage, may be produced Which is compensated by large circulating currents.

In case of single line-to-ground fault, either on primary or secondary sides, the zero sequence current flowing through the delta-connected tertiary winding is given as:

/(Zo Tert.) =

t, X 100

(2.19)

2ZH-L +ZOL-T (or Zo H-T)"

(Z6L-T to by taken if the fault is on LY i.e, secoIlctaryside and 20 H-T, to be taken if fault is on HV i.e., primary side),

,

\

Ii

I,

\

\

.

(

•

(

Principles (If Transformers 2

t'

, I I

I I

I

I

I I !

\ )

I; = Rated current in tertiary winding at rated capacity:

I

of main (primary/secondary) winding. '

I

ZU-L c::: Positive sequence impedance (in percentage) between;

primary and secondary windings at rated capacity: of main windings.

20/_-'[ = Zero sequence impedance (in percentage) between I secondary and tertiary windings at rated capacity of: main windings. i

ZOIJ-T = Zero sequence impedance (in percentage) between: primary and tertiary windings at rated capacity of!

I, main windings.

Thus the reactance of the winding must be such as to limit the circulating i current to that value, which can be carried by the copper, otherwise the i tertiary windings may overheat and mechanically collapse under fault cou-l ditions.

,,'

Where

(

(

I' I

I~

2.14.2 Tertiary Windings in Star/Star Transformers

Star/star transformers comprising single-phase units, or three-phase five-limb \ core-type units, suffer from the following disadvantages:

(a) that they cannot readily supply unbalanced loads between line i

and neutral, and ;

(b).that their phase voltage may be distorted by third harmonic: emfs

By use of delta-connected tertiary windings, induced currents are caused to ; circulate in it, apportioning the load more evenly over the three phases. The: delta-connected tertiary provides a path for the third harmoniccurrents.

2.14.3 RuLing of Tertiary Winding

Rating or tertiary winding depends upon its usc. If it has to supply. additional Ioads, its winding cross-section and design philosophy is decided: as per load and three-phase dead short-circuit OIl its terminal with power now from both sides of HV and Mv.

In case it IS to' be provided for stabilization purposes only, its crosssection and design has to be decided Irorn the thermal and mechanical considerations for the short duration fault currents during various fault conditions single, line-to-grouudfault being the most onerous.

2 15 .PARALLE:L OPERA~llONS PF TRANSFORMERS

When operating two or more transformers in parallel, their satisfactory performance f~quirc:; that [b(~y have

24 Transformers

(a) The same voltage-ratio;

(b) The same per unit (or percentage) impedance; (c) The same polarity;

(d) The sa1mc phase sequence and zero relative phase displacement.

Of these, (c) and (d) are absolutely essential, and (a) must be satisfied to a close degree. There is more Iatitudc with (b), but the more nearly it is true, the better WIll be the load division between the several transformers.

(a) Voltage. Ratio '

If voltage readings on tbe open secondaries of various transformers, .to be tun in parallel do not show identical values, there will be circulating currents between the secondaries (and therefore between primaries also) when the secondary terminals are connected in para1lel.. The impedance of transformers is small, so thai a small percentage voltage difference may be sufficient to circulate a -considerable current and cause 'hddiiional [2 Rloss. When the secondaries are loadedrtheeircufating currentwill tend to produce unequal loading conditions and it may be impossible to take the combined full-load output from the parallel connected group without one of the transformers.becomingexcessively hot.

Casel: Eq~al Voltage Ratios

The equivalent circuit of two transformers running in parallel are shown in Fig. 2.17. (Magnetizing current of both the transformers ha ve been assumed more or less equal, resulting in voltages .coincident in phases also.)

V .....,.--+-ToR-,\ANSFORM EIl:Rf061r-D-'~~ t

Zl " \,~

. .' Z~'0Q60' j\

1------,-- v .. \ (c)

TRANSFORHER 2

PRIMARY (a)

Let

Fig.2.17 Transformers in parallel.

21, 22 = the impedances of the two tran-sformers

It, I: = their currents _

V = common terminal voltage I = totai combined current

v = voltage drop across the impedances in parallel

, ' ( • I

J -

~.

,

i'

._---:l( ~ \ i II

) (
) (
) f
\
) (
) (
) (
) (
) r,
,. ) (
) (
1.(
) (
) f
) , I;

\ '

l;

J,

Principles 0/ Transformers 25'

From Fig. 2.17, v = IIZI =12Z2 = IZ12, where Zl2 is tho impedance of:

Z, and Zs in parallel.

Therefore,

So

Similary,

and

h = [ (il!f_IZ2) ]

MUltiplying both sides by common terminal voltage V IZ2~< VI'

VII = (ZI + Z2)

n. = VI [ (ZI~~2) ] '

Writing VIx 10-3 == S, the combined load kV A, the kVA carried by i i

each transformer is \

SI = S [-(Z~+~2)J

and

[- ,ZlJ -

(ZI +Z,)

(2.20)

These expressions are complex, so' that SI and 52 are obtained in magni-.

tude and phase angle. - ,

Case II: Unequal Voltage Ratio

Let Ei, E2 = no-load secondary emf's

and Z = load impedance at secondary terminals

Other notations are taken as in case I above.

,

Then v = IZI2 = (II +12)212

Also the emf's of the transformers will be equal to the total drops in their, respective circuits.

EI = 1121+(h+h)Z

and

(2.21) ,

26 Transformers

Whence

£1-E2 _.-; (I.Z.-h22)

as might be expected from this expression

I (E.--E2)+hZ2

1=

. 2,

giving II in terms of Iz. Substituting this in Eq. (2.21) aboye for [2 gives

E2 = 1222+ [- (EI .- £2)+I1Z2 +h lX2

, 21 ~l

Whence

£221- (J~I-- £2) 2 ZIZl+Z(ZI+22)

"1 I

~

I J

(2.22)

and by symmetry

It is' clear from Eq. (t.2l) that, on rio-load, there will be a vector, circulating current between the two transformers, of amount

II = -h .- (EI~E2)/(ZI+22)

a result obtainable from Eq. (2.22) by writing;

Z = ex: (infinity).

On short circuit the expected result from Eq, (2.22) is given by

It = Et/Zt

and I: = E2/22

(b) Impedance

The currents carried by the" two transformers are proportional to their ratings if their numerical or ohmic impedances are inversely proportional to those ratings and their per-unit impedances are Identical.

A difference in the quality factor (i.e., ratio of reactance to resistance) of the per-unit impedance results in a -divergence of: the phase angle of the two currents, so that one transformer will be working with a higher, and the other with a lower power factor tban tbatof the combined output..

(c) Polarity'

The primary and secondary windings of any individual transformer may, under certain conditionsof coil winding, internal connections and connections to terminals, have tbe same or opposite polarity. When, respective inducedterminal voltages for primary and secondary windings are in the same d irection, the polarity of the two windings is the same. This polarity is generally sp ok en as subtractive, When, on the other band, the induced

}.

i

I P rlnclples of Transformers 2,7

I

,11,

1,

terminal voltages are in opposite direction, the windings. are of opposite

polarity which is usually referred to as additive.

030

.-

) (

(0) SUBTRACTIVE. POLARITY

°1 T' ' ;
Oz 1/'
SUPPLY AI J
- [:
)

) {b) ADDITIVE. POLARIT:Y. -----1---~ )f'

,..-_-J0z bz fez <a- . ~.

&2

,..---------_. __ .

---11'2.

r-----.- ... su .. p!..\"

I e.z ~C;--:- .•

~ ~

'---_.L J

(C) SUBTRACTIVE. POLARITY

Fj~. 2.18 Test connection Tor determining polarity.

-

Figure 2.18 shows the test connections at (a) and (b) respectively for,

single-phase transformers having subtractive and additive polarity. When i single-phase voltage is a.r>.R_lic_d to terminals Al and a2", the measured voltage between terminals A I and A2 shall be Tess than applied voTlagc-rn case of (a) and greater than applied' zoltage in case of (b) .. --.-.---.- .. r:

For three-phase transformer the testing procedure is similar, except that the windings must be excited from three-phase supply and more voltage measurements have to be TaKel-l--' for'-d~'t;-r~1inatlorl-' O('ex:1~t polarity and phase sequence, Figure 2_18 (c) shows the test connections and results for. a star/star-connected transformer with subtra:ctivcpo]arity:.

For parallel operation, the polarity should be the same. A wrong polarity results in a deud short-circuit.

~/ . Transformers

(d) Phase Sequence

The phase sequence, o~ the order in which the phases reach their maxi- ; I

mum positive voltages, must be identical for two paralleled transformers. Otherwise, during tho cycle, each pair of phases will be short circuited.

Any mixture of internal connections can be worked out if it is

. remembered that the primary and secondary coilson anyone limb have induced emf's that are in time-phase. The several connections produce various magnitudes and phases of secondary voltage; the magnitude can be adjusted forparallel operation by suitable choice of turn-ratio, but the pbase divergence cannot be compensated. Thus two sets of connections giving secondary voltages with a phase displacement cannot be used for transformers intended for parallel operation:

The following are .typical of the connections for which, fro m the view point of phase isequence and angular divergence, transformers can be operated in parallel:

Transformer 1 Transformer 2

Yy Dd

Yd Dy

Yd Yz

Internal connections of different windings can be done in a variety of ways, giving different vector groups. Various vector group arrangements and their connection symbols are given in Indian Standard IS: 2026 (Part IV)-1977.

APPENDIX

. CALCULATION OF % REGULATION AND EFFICIENCY

Transformer Specification 250 rvlV A, 15.75/240 kV, three-phase, 50-Hz generator transformer having no-load 10ss'120 kW, load-loss 500 kW at fullload at 75°C and reactance voltage drop lA%.at rated load.

(a)~~gulation at full-load, unity power-factor.

R == 0.2%, X = 14%

cos 6 = 1.0, sin C = 0

Substitute values in Eq. (2.15) Percentage regulation

" = Q,2,x 1.0+- (14 b~)2 1, 18 /~.

.... .._. .. _,2

(6) Regulation at 314-10ad~ D:8 lagging power-factor

R = 0.15 %, X = 10.5 %

cos 8 . 0.8, sin 8 = 0.6

) (

) (

) \

) (

'~'.'.

Principles 0/ Transformers .29 I

Substitute values in Eq. (2.15) Percentage regulation

[( 1 0.5 X 0.8)- (0.15x 0.6)F

= 0.15xO.8+10.5xO.6+ 200 '!

<=> 6.77 %

(c) Efficiency at full-load, unity power-factor Percentage efficiency

= (1- Total loss ) X 100 Input power

( 1- 500+120) 100

-= 250x 103 X

= 99.75 %

(d) Efficiency at 3/4 load, 0.8 lagging power-factor No-load loss = 120 kW

= 500X( ! r = "281.25 kW

Load loss

3 .

Input power= 250 x t 03 X 4 X 0.8 =·150 X 103 k W

)

Percentage efficiency

0::::0 ( 1- 281'25+ 120) 100

. . . 150 X 103 X

= 99.73 %

. ~ .

I

)

)

REFERENCES

"

...

1. Say, M.G. The Performance and Design of AC Machines (Book). The English Language Book Society and Sir Isaac Pitman & Sons. London.

-2. Puchstcin, A.F., T.e. Lloyd and A.G. Conrad, Alternating Current Machin~s (Book), Asia Publishing House.

3. Hellman, Charles 1. Elements 0/ Radio (Book). D. Van Nostrand Company. INC,

New York. .

4. Thcraja , n.L. A Text Book 0/ Electrical Technology ... Sultan Ch;I;IU & Co .• New Delhi.

3

Materials Used .in Transformers

M.P. SINGH Manager (Design), Bhopal

A part from active materials like copper and cold rolled grain oriented silicon steel, a number of ferrous. nonferrous and insulating materials are employed for building up a transformer. Optimum utilization of all materials in consonance with their electrical, mechanical, physical, chemical and thermal characteristics is necessary for obtaining a compact size transformer. Strict quality control measures like testing of raw materials,' therefore, become imperative. One basic requirement for all materials used in oil-filled transformer is that they should be compatible with insulating oil and should not react with or deteriorate oil.

Table 3.1 gives. application, applicable national and international standards of variousmaterials used in a transformer.

3.1 INSULATING OIL

Insulating oil forms a very significant part of the transformer insulation system and has the important functions of acting as an electrical insulation

as well as a coolant to dissipate heat losses. The basic raw material for the production of transformer oil is a low-viscosity lube termed as transformer oil base stock (TOBS), which' is normally-obta-ined-by-frattiQr.-al-. - distillation and subsequent treatment cf crude petroleum. Important characteristics of TOBS given in Table 3.2 must be kept within permissible limits in order to produce good insulating oils. TOBS is further refined by acid treatment process to yield transformer oil.

\

__ ,"7; _

'"

'"0

... CO '"0 Q

t"1 1,) -'"0 CI.l c<: u ....

.:cO

c<: t>

0. 0.

<

M. 'atcrlals Used ill Transformers :\1

....

1,)

c,

Vl

C;I

t-

o

0.

c-: 0.0 0;" ",0 til .o I=Q r:

.,
~.".
V! E 0 '0
E ,- -0
Q ;:: 0
C'J c-:: ,_
c vi
-0 .c: .D I
... ,~ E E E '0
E .c; ;:; :J o ...
'- c:l
'- .'- c: c-:: '- 0
0 0 .D
,_ ,_ E .c !;!JE- V)
'V 'v ::: c ::l Vl
0. 0
0. ?: 0. '-
..: ('j d '0
c, 0. t- :; '0 0.
.. 0 .c: ' c:;
~, 0
Cl. g' _::, OJ) 0. 0. 0 0 "0
(:1 c: .-:: .~ c-:: E ~ \I)
0. ...
c, :-. 0. 0
!2:E c-:: .:: ... ... - '0
~, 0 .:: "? ~
.!!! 0 "t:J J r: ::l c-:: 'V c, '- c:
0. c-::
.~ c.. C'J v- I;) en Li ('j ... c-:: ~
'-' ... c: C C .D ~ c-:: 0- ~ 0 <-
,'S .. tfJ ": c1 0. .D C c-::
c.,. E ... - C) 'IJ 0 "0 u
._ <V ._ 0. V) .:: ... (I)
N .= '- M C. E 'IJ V) c-:: 0. CO ." c:: .::
I~ r- .-: c-:: '.J IV- '\.)
c.::: .~ .- ... '- t ... ,_ ... U2 0 '- '-' 0
..... '" ~'r. ~ o 0... ~ . .D Po. ~ CI.l
.r.=. ~ 0. V) U;
ell U Vl ......_
c: ~ 'IJ .-- ---. ....... ,...... '-'
- [U >- >- ,_
r ;.:::. '-' > P.. '-'
I e: .~
I :J
'"
0 c: C-~ <'1
I~· ....
z
I~ co

"t ......

'" '"

~O

1;')\0 <'1 0\ MN

U U)l.LJ

'- 0\0

U~

o

0. .. .-. C/.) r-

32 Transformers

.... ... o 0. 0. ::J VI

.... o ~

.... o ~.

·c ....
Of)~_ 0
0. ..... ,_ \;')
·C~ c Ol
't;t:~ Ol
......... ~
...... 1StU-o
...... u
.... ....; 0 _~c 0
0 - - 0
("'.l - . - 0....... s'O U
.... ...... • ....
Q ~ (""fl C""l :----'£ cu:Gr-i o· "<t-
o .. M ~ 'g § 5·,:: ~ 0 "'0
C) ~_. \0 (1"\ UC') '0
M ~ ._J:;- _, ::l _, '0
'0 0\1; " C-'O\O CI.lOO 00 '"0 00 .....
0 ....
(>") 0- " ~C) C) ..14 •• < ....... 14 ..
rJl8C1.l~o(,') a
~~ f-<f;9 ~ fIl u·~ fIl
..... ~ ..... ~- H ...... .

"<t"

/

".'t!~ ... ~

til ......

-cl .

l

.... o·

8

Materials Used In Transformers 33

,,_

o

s

~lXl~

<t>N '0

~tIl Ot:q

z

v .... o u

.... v

E

.....

.E

'"

Q

("j

b

..... o ~

o o o

. \

\,.

-0

<t> ..c

'"

I

J4 Trallsformers

... ...... d .c

VI

.... o

o :-.

o

....

U

o

00

_-

..... .... 0-. U\O

'<t ••

AU)

.-

.... ·0

0 ...

._

......

o

0lci ~r.... 1'1 C/)r-

:. U) C)H

E"d Q3 til
til
~c: o<U
~d ~E 2
<U
"O"d d ...
.0 0 u~ ~
»::: ~ ___ . .3
-0 (J
0"d ~ til
E .... s ;:s
,d ro
S.!:l c;j '-"
8-0. vr
. 2 til ~
.... ::: ::s .~
f::o '- "d
0·- 8~ U
.... ~
..r:;U 0._ 0
uc:.> ... - Vl
_._
,Vl t:i.!:l
'6)\J C)
.5 (J
N ~ ::: '" -en
U d U
·1'2 ..... "2 :~ E ~
co:: .... ::: _ :J
~..a a.> ,o.;:! -.-:=
o "0 a. -en :::~.:::
.,., cu E::sS o~
~ t;2 < . z: ~C12
'<t" Or) \0
..... .2I
c;j
-......_ "0
a. .
'-' u (3
btl u
-v- I
<13_ <r.
;:s eo ~
0"1=: 0
"0 c;l U
.... _: cu
<13 U Vl
"d ~ .
~ ::: "d
ro Q I=:
~.c d
I u s:
.: .... .:::
"0 c;:
CUd .o ...
... ..0 c:.>
-
Vl ft CU eo:
~ .... CJ .,...
d ,c.
'-t::; Vl
;:s • '"
U t:1 ....... ....
...c> a
;:SO ."~~ ....
..... _ ...
..... - L.. 'J,..;. .<:.J
V. <!J ~"O ~
Vl
r...: 00 §
Z
0 ~

......

U U

~

o ~

... CU .c

-0

(I)

......

<o

00

00(1) r-Z

(I)":

He)

\

Materials Used ill Transformers ;35

I

'fable 3.2 Chaaracterlstlcs of Tons

S. No.

Characteristic

Requirement

1. Viscosity at 37.8°C

2. Pour point, Max.

3. Plash point (Pensky=-Martcn closed cup method) Min.

58-86 Sec. SUS -12 C

140cC

---

3.1.1 Chemical Composition

Transformer oil consists of four major generic class of organic compounds, namely, paraffins, naphthenes, aromatics and olefincs. All 'these are hydrocarbons and hence insulating oil is called a pure hydrocarbon-

J i

mineral oil. For good fresh insulating oil, it is desirable to have more of

saturated paraflins, less of aromatic und/or naphthenes and none of olcflncs. However, for better stability of properties, it is necessary to have optimum aromatic and/or naphthcnic hydrocarbons. Such an optimum balance' is struck by n carefully controlled rcflnin g process. Depending upon (1\0 predominance, oil is usually termed as of paraffinic base or naphthenic

, ,

3.1.2

Characterlstlcsof Oil and their Significance

,

!

,

i·

Table 3.3 gives characteristics of oil' as per IS : 335-1983 (third Table 3.3 Schedule of Charactcrlsttcs of Insulating 0 it

SI. No.

Characteristic

Requirement as per IS : 335-1983

1. De~~:~~' at 27°C, max.

2. Kinematic viscosity at 21°C, max.

3. Interracial tension at 21°C min.

4. Flash point, Pensky-=Marten (closed), min.

5. Pour point, max ..

6. Neutralization value (total acidity),

max. .

7. Corrosive sulphur

8. Electric strength (breakdown voltage) min.

(3.) As received

(b) After filtration

9. Dielectric dissipation factor (tan-delta) at 90cC, max.

10.' Specific resistance (resistivity), min. (a) At 90"C

(b) At27°C

II. Oxidation stability

(a). Neutralization value after oxidation, max.

(b) Total sludge after oxida tion, max.

12. Presence of oxidation inhibitor

0.89 g/cm3 27 cSt 0.04 N/m 140°C

-9°C

0.03 me KOH/g

NOll-corrosive

30 kV (rrns) 50 kV (rrns) 0.005

30XI012 n em 500XI012 n em

0.40 rng KOH{g

13. Water eontcnt (as received), max .

0.10 % by weight The oil shall .not contain antioxidant additives.

50 ppm by weight

. ~

~=~...........,_-- ------_. -_. _. _._--_. __ . -- -_._--_.-.- .

.. _-_._--_._---

36 Transformers

revision). Tho typical approximate values (all relating to 60<' C) of some other physical properties of oil considered 'in design' calculations arc given in Tabie 3.4. Figures it). parenthesis indicate the approximate temperature coefficient (per °C rise) for the property concerned. The significance of various tests conducted on oil to evaluate its quality is given below.

A. Physical properties

(0 Density, This test has special significance when transformer is operated in a very low temperature zone. The maximum value of . density . fix.dd'·· at 27~Censllres that water in the form of ice present in oil reniart1s! at the bottom and .does not tend to' float on the oil up to a temperature of about -10°C.

(H) Interfacial tension (IFT). This is a measure of the molecular attractive force between oil and' water molecules at their' interface. This test provides a means of detectingsoluble polar cont3Illinafitsand products of' deterioration, which decreasemolecular attractive force between oil and water. It is considered tnat It-T gives au indication of degree of sludging of en (Fig., 3.1).

~ ...... ~ ~I ..

o~ wo z"::>

:i ~_

« )0("

W'Z' '

_I z;

0<

.....

lJ...Z o ,0

w uO <z i ::l. we

u

c: 0 wz 0.'«

'"i.I_

s •

SLUOG E rORMt:.lION ,ZONE

.'

,,"

• r- 100r----~-:----___r~-"-r-,----,--.----_r__---_r_--__.

\ I
I
I
I
I
I
I I ~ .. '.

:i\ : I

I I \ I _ \

:~>r-:--_ -!-. _ __,\_~-~l \ \ \ I \

I I , I " \

'-:} ._-' _~L- ---l~-~-!,--- ~-1--~-1---- -k----ts

Fig. 3.1 Presence of sludge in oil" related to the in tcrf'acia 1 tension values at an ambient temperature Of 2TC.

,--(

,(iii) Moisture content, The amount 'of free and dissolved water present in the oil is its rnoi-ture content and is expressed in P;'l11 (carts per million by weight, i.e. mg'kg). Presence of moisture is :Iarmful since

Materials Used ill Transformers J7

it adversely affects tho electrical characteristics of oil and accelerates deterioration of insulating paper.

Table 3.4 Physlca! Constants of 011

51. No.

Property

Value

1. Permittivity

2. Specific heat

3. Thermal conductivity

4. Coefficient of expansion

5. Mean density correction factor over the normal range of operating temperature

2.2 ( - 0.001)

2.06 kJ/kg DC (0.0038) 0.l3 VI/moe O.OOO78te 0.OOO6sre

. (iv) Flash point. It is. the temperature at which oil gives so much vapour that this .vapOUr, ·Will-II. WiAC.U with air, forms an ignitable mixture and gives a momentary flash on application of flame under prescribed. conditions. A minimum flash point is specified in order to preven t i the risk of fire that might result by accidental ignition.

(v) Viscosity. It is a measure of oil resistance to continuous flow without the effect of external forces. The oil must be mobile, as heat transfer in transformers occurs mainly by convection currents. Since viscosity ~ increases with decrease in temperature, it is. necessary that viscosity ~ be as low as possible at low temperatures.

(vi) Pour point. The temperature at which oil will just flo; .... under the: prescribed conditions is known as the pour point. If the o~l becomes; too viscous or solidifies, it will hinder the formation of convection·, currents and thus cooiing of equipment will be severely affected.

B. Electrical properties

(i) Electric strength (breakdown voltage). BDV is the voltage at which breakdown occurs between two electrodes when oil is subjected to an electric iield under prescribed conditions. Electric strength is the basic parameter for insulation system design of a transformer. It serves to indicate the presence of contaminating agents like moisture, fibrous materials, carbon particles, prcci pitablc Sludge and sediment.

(ii) Resistivity (specific resistance). This is the mOSL sensitive properly of

oil ruquir ing utmost c.i rc [",_)1" its proper dctcrruinat ion. Rcxist ivi: y

in U~Cf;l is numcric.rlly equivalent t o the rcxistu ncc bot .. vccn opposite races Or:l centimetre .cubc oj" the liquid. Insulation resistance of

38 Transformers

windings of a transformer is also dependent upon the resistivity of oil: A low value indicates the presence of moisture and conductive

contarnina nts.

(

(iii)

Dielectric dissipation factor (DDF). DDF is numerically equal to sine of the loss angle (approximately equal to tangent of loss angle for dielectrics) and is a good tool to indicate the quality of an insulation. A high value of DDF is an indication of the presence of contaminants or deterioration products such as water, oxidation products, metal soaps, soluble varnishes and resins. ~

C. Chemical properties

(i) Neutralization value (total acidity), It is a measure of free organic and inorganic acids present in the oil and is expressed in terms of milligrams of KOH required to neutralize the total free acids in one gram of oil. Oxidation of oil in service is a consequence of . reaction between hydrocarbons present in the oil and oxygen. The oxygen may be atmospheric since ,. 'di.f . comes into contact With the atmospheric air during breathing of transformer,

or may have been dissolved in oil if oil is' hot degassed properly, or .niaybe liberated due to Ieffect of heat on cellulose insulation. Oxidation ()f oil, is a chain reaction by which organic acids and sludge are formed. Copper present in a large quantity in transformer, aC:3 as a strong catalyst in oxidation. Hence, as far as possible, no bare copper is allowed to be used in power transformers. Copper is covered with paper or enamel coated or electrotinned. The products of oxidation are injurious to the insulation system of transformer. Acids formed give rise to formation of sludge which precipitates out and deposits" on windings and other parts of transformer. This causes hindrance to proper oil circulation and heat dissipation. The acids also encourage deterioration of cellulose insulation: i.e. paper, pressboard arid wood. Water is produced during oxidation which reduces electric strength of oil and also accelerates corrosion of metals and deterioration of insulating materials. Hence the measurement of total acidity is the most convenient and direct method of assessing the capability of oil for nonformation of acids during service.

(ii) Oxidation stobility, This test is the measure of. neutralization value . 'and sludge after oil is aged by simulating the actual service conditions ofa transformer. The oxidation stability test is very, important for new oil but not for oil in service and shows the presence of "natural inhibitors which impart anti-oxidation characteristics to oiL

(iii) Sediment and precipitable sludge. These are oil deterioration products or contaminants which are insoluble after dilution of the

- {

, ,

MII/uf(//,r (/,1/(11 ln ·'hl".~J(If'I"I(.f J?

oil with II-heptane under prescribed conditions. However, prccipitable sludge is soluble in the solvent mixture of equal parts of toluene, acetone and alcohol but sediment is insoluble in this.', solvent mixture. These contaminants are determined for oils in, service. Oil is considered unsatisfactory for use if sediment 'or precipitable sludge is detected.

(iv) Corrosive sulphur. Crude petroleum usually contains sulpbur compounds, most of which are removed during the refining processes. This test is designed to detect any traces or' free corrosive sulphur

, J .

that may be present in oil. Presence of corrosive sulphur in oil

will result in pitting and black deposit on the surface of bare copper used in transformer, which will adversely affect the dissipation of beat and consequently performance of tbe equipment.

3.1.3 Stability of Oil

Paraffinic. base insulating oils available iudigenously have shown deterioration of electrical properties during, use, at a rate which is a little bit faster than of naphthenic base oils produced in the USA, the UK and other countries. Efforts are afoot to produce paraffinic base oil in the country having stability comparable to naphthenic base oil.

3.2 INSULATING PAPER

Paper is a fabric made from vegetable fibres which are felted to form a web or sheet. The fibrous raw materials are obtained, from plants including cotton, hemp, manila, straw and coniferous/deciduous trees. The cell or fibres of such plants consist mainly of cellulose. The molecular formula for cellulose is (C6HIOOS)n. The degree. of polymerization of the molecular unit indicated by the letter 11, varies widely in various' plants, The value of n can go up to 2500 or more for cotton fibres and up to 1200 or more for wood pulp. Various other materials, e.g. lignin, hemicellulose, mineral matter and resins are associated with the cellulose in the fibres. These contaminants arc removed by sulphate process treatment of wood ' pulp and careful water washing. The different -types of paper used andtheir application arc given in Table 3.1. Terminal-gear and high voltage ,lea~s arc norrnallv insulated using crepe kraft paper to obtain higher flexibility,

-I '

and smaller radii at bends. These papers have high stretchability varying ,

between 50 to 100%. '

3.2.1 Characteristics of Insulating Papers and their Significance Characteristics of different papers are given in Table 3.5 arid their significancc are described as Follows :

I (

Table 3 5 Schedule of Cb.nrncleristlcs of Electrical Grade Papers

Requirement

Sl. No.

Characteristic

Kraft paper of medium air permeability

Kraft paper of high air permeability

Press paper

1

2

4

3

5

6. Ai,t permeability 0.2-'-"0.5

'4gm:iPa'S ",'

II. IJecii~~ic~i pro~~rt~es

.-\" "~.:, :, :, ,'!.,_ '.. ,"

1. Tensile strength

(expressed. as tensile index), .': N mIg (min.),' '

_. Machine direction ';' ""

.~-.cross directio~, .._ ,

• • • ~ .... )0 • -

8. Elongation at break. %,(min.)

, .J Machine direction , .Cross direction

9.'Internattear- =. Substance "!,', ing resistance (r,t m2)-(expressed as ? 40-80 tear index). .>80-120

, rn'N rn2jg, ' >120

(min.)

10. Heat stability

, .' (decrease in bursting strength), % (max.)

III. Electrical Properties

11. Electric strength

, (BDV), kV/mm, (rnin.) (i) In air at 90°C

(ii) In oil at 90°C

12. Dissipation factor

of unimpregnated paper at 105"0 (max.)

] 3. Conducting paths

I. Physical Properties L Substance, g/mi

2). Density, g/cm3

3. Moisture content. %. (max.)

4. Oil absorption,

%, (min). '

"

5. Water absorption, , mro(min.)

lOO±5for 125 !Lm thk. 0.75 to 0.85

'8.0

lOO± 5 for 125'(.lm thk,

0.75 to 0.8'5 1.0 to 1.3

8.0 8.0

3'to 21% depending upon density

10

10

05--1.0

'83 78 69
30 ,28 24
,' .• OJ_
2 '2
4 " ,', 4
Tear resistance ,
MD XD MD XD
5 6 5 6
6 7 r 7
')
8 9
- 20 20 60 7.5 7.0 10
65 (
.-----,--\
0.003 0.003
should be free from any conducting ",'It},
r'-"~ I \
) i
"
i \
i Materials Used in Transformers 41

) !

3

4

5

2

) [

IV. Chemical Properties

14. Mineral ash, %, (max.)

15. Conductivity of 5 % aqueous extract mS/m, (max.)

16. pH value

11. Chloride content of aqueous extract, rngtkg, (max.)

18. Conductivity of organic extract, nS/m. (max.)

V. Methods of Test

6-8 6-8 5.0-8.5
50 50
.,
10 I
10
r
lEC 554-2 isc 554-2 IS: 8510 1.··.
!
IS : 9335 (Part II) IS : 9335 (Part II) 1.0 10

1.0 10

2.0 13.5

) \

) I

) '.

A. Physical properties ,

i

(i) Substance igrammage), It is the ratio of mass to the area and is a

fundamental parameter' which influences most of the mechanical: . and electrical properties.' This is a basic property for paper making" process.

(in Density. It is a function of thickness and grammage of paper., Density is also a basic property for setting process parameters for: the manufacture of paper. Papers in density range of 0.6 to' 1.3 g!cm3 are employed in transformer manufacture. .

(iii) Moisture c;ntent. Cellulosic fibr""esan~ hygroscopic. Water has the' effect of plasticizing the cellulose fibres and of relaxing and weaken-: ing the interfibre bonding. Hence moisture content has a significant: effect on many properties of paper. !

(iv) Oil and water absorption. .Paper attains a high value of electric: strength when impregnated in oil under vacuum, since place of; moisture liberated during vacuum drying is taken up by oil in the: molecular structure of the paper. Oil absorption is dependent upon: density and air permeability of paper. Sometimes as a substitute to;

\

oil absorption, water absorption is monitored. :

(v) Airpermeability, It is the measure of the rate at which paper allows: alfto penetrate through it. It is infi uenced by both the internal! structure and the surface finish of the paper. Electric strength of': the .paper is inversely proportional to its air permeability. '

)

) .

) '.

) ,

) ,

i \

B. Mechanical properties

(i) Tensile strength and elongaticn, Paper should be able to withstand, tension exerted during its wrapping over layer winding co·iIs. Paper tape is wound over-conductors and cables under specified tension at.

i

I

t- ,::;b:.."".

~;..~-

J.

42 Transformers

(ii)

a high speed by paper lapping machines. If specified tensile strength is met, breakages of paper will not occur in paper lapping operation. it has been proved theoretically that the displacement of windings during short circuit of a transformer depend very mi!ch, on, :ho Young's modulus of insulating materials used in windings whicb shows the 'irnp ortancc of tensile strength and elongation tests.

Internal tearing resistance. This gives. the load under which paper will tear off under specified conditions and shows' its capability against tearing.

Bursting strength. This is the pressure required to burst a disc of paper which is gripped firmly around its periphery, and to one side of which pressure is applied at a uniform rate, using liquid as a medium This test also gives an idea about the mechanical strength of paper.

(iv) Heat stability. It isth,.B., abilityo]' paper to witbstandtfierma! stresses

during service life 6fa:tran'sfbi:mer and is determined by measuring decrease in internal tearing resistance., bursting streflgth and degree . (}f.po,lymerizati()n after subjecting it to accelerated' temperature.

(iii)

C. ::t;Electri(a{pr,0perties

. .' ("t'~.J,; ...

. (i) .~le·c.ihc strength. This is a basic parameter in deciding the insula-

,:,,~tiOIi system design - of a transformer. Electric strength depends on ',,;- deIl~it);- and airpermeability of paper.

(in Dissipation factor. It is a good tool to indicate the quality of a dielectric. A high value of dissipation factor shows presence of con~al?-inants or depolymerization of paper.

(iii), Freedom from conducting paths. Electric strength of paper is adversely affected by the presence of conducting paths and hence . these are undesirable impurities.

D. Chemical properties

The. conductivity of aq ueous extract is a measure of· the electrolytes present in the 'paper which is extracted by gently boiling paper in distilled water. The electrolytic impurities are present as ionizable acids, bases, salts or a mixture of these. This property indicates chemical purity of paper and is a basic parameter for paper making process. Insulation resistance of p~per is dependent onthis property. -' Conductivity of organic

_ extract, pH value of aqueous, extract, .ash content.and chloride -Co-:'itent.oL aqueous extract are also measured to check purity. Mineral ash, solid residue remaining after complete combustion of paper are injurious to paper and should be within the permissible limits.

{

( I

\ ..

Materials Used {II Transformers 4:3

3.3 PRESSBOARD

I.

Pressboard is a widely used insulating material for making a variety of : components used in electrical. mechanical and thermal design of transfer- : mers, Types of pressboard used' and their applications arc given in:

.Table 3.1. Like paper, pressboard is also made entirely from vegetable' fibres, whose cells contain mainly cellulose, Pressboard is manufactured' using the following raw materials:

(a) Sulphate wood pulp

(b) Cotton

(c) Mixture of sulphate wood pulp and cotton or jute hemp (d) Mixture of cotton and jute hemp

3.3.1 Characteristics of Pressboard

The characteristics of different qualities of pressboard commonly used arc given in Table 3.6. Solid pressboard up to 6 mm thick calendered and up to 8' mm thick prccompresscd is made. Prccomprcsscd pressboard is manufac-: lured by pressing wet sheet at an elevatedtemperature. For higher thickness, laminated calendered and precompressed pressboard is : manufactured by building up laminae of solid pressboard using an adhesive (e.g, polyester: resin, polyvinyl alcohol, casien glue) for bonding under pressure and high temperature.

3.3.2 Moulding

Tho most difficult practical insulation problems in high voltage transformers occur at the ends of the windings and at the leadouts from the windings. Pressboard moulded components can be made to any required shape' to follow the contour of the equipotentials determined from the field plot. Angle rings and caps are the widely used mouldings (Fig. 3.2 and Fig. 3.3).

Angle rings are also made by stacking number of kraft paper sheets which are flanged and stitched together using pressboard sheet on two faces of the flange as support. These hand-made angle rings have the disadvantage of nonconforming radii and pave structural instability. Tightening of winding leads to deformation in electrically stressed. zones. It is difficult to maintai~ oil ducts due to buckling under compression, Hence, it is necessary to use moulded angle rings which are dimensionally stable. Other types of mouldings used in EHV transformers are snouts, . angle sectors and leadouts, etc.

Evaluation of moulding is another difficult task-due to intricate profiles ..

These cannot be tested without destroying them. Normally, bend/neck of the component is the weakest point. Hence, these are tested on regular, irregular and neck portions to ensure the overall quality .

. .. ,,,,-,k:'x.,

.....,,+c .... -.-a--~.

44 1'ra,..sj'ormt:rs

'C
..
c<I
0
.0
'"
'"
Cl
...
Po.
.... .....
0 C
'" II)
Cl E
Po
?; II)
....
.... .:;
t:I 0'
Co) II)
... ~
~
A
.....
0
.'"
.:!
...
'"
·c
Cl
s'~ :' ..... t)
Cl
...
S
0
....
0
4)
::
"0
Cl
-B
tr.l
\0
M
Cl
::;
c::
~ 0.
"3
0.
-0
''0 0
0 VI
'3 u ~ ~ --:
........
U «I §a:& ......
C! -g .. 8 ID ..... ~"1C'1
o U 8 ..... II)
CI)_«I 8.2, O\t-- ....
B- 0' 0. e v
"3 u
"0 0. 0 t:I
II) 2; til
'" "t:I <'"I ..
~"'d 0 II) ~
... 2 0 ...... .D
0. til VI ~ I VI t--'Ii) I I 0.
0.
E.S C'! ID '<t 00-'+ ~ til
0 a ~ .... ._ "t:I
u 8 II)
II) til
... - 0 bO
tl< ..... OIl
.9 til
~ ....
~ !:'
._ ~ "'d .0.
4) "3
ec
• <I.l 0.
1I)"t:I "'d
L.c .. ~
0.- 0
E ~ <'"I 0
0 ~
g ~
...
~ 8
...... . ~

~

CI)

.;;::

II)

"0

til .... ctI

.. ..c:I

U

o

i/.iZ

Q .9 ."t:

<n' o

C'I 0.

a

-c

U] II)

Vl"!"1~ 0000.

...

Q II) U til

:u

til

o

~

....

--:

.~ I eo ~

o

.,..,t--"1 co"<%,

(

(

.I

)

-L

t- ~"~"'_-'--

Materials Used in Transformers 45:

"<t-
V')
t"I
V')
.,.. 0\ V)
I I NO 0\ ...., VI ~N -- C! en
C"Ir<"I ..... 0 0 f.Q I I

---. <I)

C .~ c :t

.:: ~, E~

'J .c c _

0-0

V) o

o

......

o

00 ......

Ul en

o M

\0
t-
.,.. 101"\
0\ -< -c
- riM :i Z til
..... -til

til o C

~

.c

.....

o o

d

o

]

00 0 o 0.

... ~

C"I-

'""" C"~
.8 I-
6 '\.. -
'-" . $
»-> >
60000 0\
..§_~~ ~ ~ J. <: -< 0
>1000.,.. z z ~
_:.:r<"Ir<"I r<"I N 00 Vl ..... 1-

t-··_-

4(i Transformers

" ,

OPfN SCARF

. \~

- - - - - f-- -:- - - - - - - ~ tzz:z:z:z:nl

C7V-;-;/_L-;-;/......,/~L1S".·- - - - I. 'I~~~------------~--------------~

r ------l---------i

vZ22ZZZ~ .... 722722'1

f
~ 0- I t
V
/j .').,
c- Fig. 3.2 Typical pressboard angle rings.

3.3.3 Advantages of Pressboard Over Other Solid Insulating Materials

In EHV transformers care is taken to use such insulating materials in stress zones which will have the least partial discharge. Synthetic resin

bonded paper... based.laminates.and.cylinders and.laminated.wood.are.prone ._:_'_(

to give more P.D. in such zones compared to pressboard, since· air voids . are trapped due to extensive use of resin as the bonding material, Oil absorption in resin bonded laminates and' cylinders is negligible and hence

these are electrically inferior to pressboard. Efforts, are made to use

(

Materials U:eJ ill Transformers 4'1

,

cylinders, winding cleats and supports, ctc., made of pressboard as far as possible.

I

IOP£H SCARF

~.-.----

I

l--------------1---- --_ -------j

~,lllllll.ZZl : ""'Z"l'"

Fig. 3.3 Typical pressboard caps.

3.3.4 Significance of Pressboard Characteristics

Properties like density, tensile strength, elongation, ash content, pH of aqueous extract, conductivity of aqueous extract, oil absorption. moisture content, freedom from conducting particles and electric strength, described in Sec. 3.2 are also determined on pressboard. Apart from the above, the details of additional tests conducted fer the evaluation of pressboard are given below:

48 Transformers

(i) Compressibility. It is a very important property and is a measure of reduction in thickness of dried specimen when compressed under load, under prescribed conditions. Transformer windings are designed based on expected compressed thickness of dovetail blocks used for providing radial ducts between turns and sections. If compressibility of pressboard is more than the value taken in the design, the coil height will become less than the design value after vacuum drying and oil impregnation of transformer and' this J reduction can lead to serious problems of short circuit forces if not rectified.

(ii) Shrinkage. It is )a' measure of reduction in length, width and thi~kness of specimen after it is dried under specified conditions. Effect 'of shrinkage is also taken into consideration in the design.

(Iii) Cross-breaking strength. This is determined to check the proper adhesion of laminae in a laminated pressboard and ensures against

delamination. J •

(iv) Electric strength edgewise. It measures the creep strength across the glue line of laminated pressboard.

3.4' WOOD

Wood based laminates are manufactured from selected veneers (mostly 2 mm thick) obtained froin various timbers. The veneers are dried and partially or fully impregnatedt'with neutral phenol formaldehyde resin dep-: ending upon type of laminate required and then densified under heat and pressure. The placement of veneers one over the other is such as, to obtain the desired grain . orientation to achieve specific mechanical strength in the . required direction. The following types of wood based laminates are used in transformers whose applications are given in Table 3.1 and 'properties in Table 3.7.

Type

I~a) Unirnpregnated densified

lam inated wood-low density and high density .

. {b) Resin impregnated densified la mina tedwoodtypeV.

Direction of Veneers Alternate veneers with grain directions at right angles to each other.

All veneers have grain in the same direction.

(c) Resin impregnated .. densified.. ---Same-as .of{.a)-,--- _ Iarnintcd wood type VI.

The area, which require. higher mechanical and lower electric strength, densified laminated wood is used for making a variety of insulation components like coil clamping ring, cleat, support, core and yoke Clamp, wedge

- - ._--(

.';

lv/(/Icria/s Used ill Transformers 49

Table 3.1 Charactcr lstlcs of Densificd Laminated Wood

Sl. ~o.

, Unirnprcgnatcd laminated wood -r-Iow density

Unirnprcgnatcd -Iarnlnatcd weed

-high density

Characteristic

I. Physlcal Properties

1. Density, g/cmJ

2. Moisture content, % (rnax.)

3. Oil absorption, % (min.) '·4. Cohesion between laminae

0.90 to 1.09 7

1.10 to 1.20: 7

9 6

The specimen shall not readily split by delamination and. the exposed tern surfaces shall have ~ distinctly hairy or ragged appearance. I

II. Mechanical Properties

5. Tensile strength, MPa, (min.)

. 6. Cress breaking strength, MPa, (min.) (i) Along the grain

(ii) Across the grain

7. Compressive strength, MPa, (min.) Perpendicular to laminae

8. Shear strength, MPa, (min.) perpendicular to laminae

III. Electrical Properties

9. Electric strength (one minute proof') in oil at 90cC

(i) Flatwise, kV Imm (min.) (ii) Edgewise. kV (min.)

70

102

! .

90 67

I] 2 85

J so

1e5

35

50

4 60

4 60

IV. Chemical Properties
10. Effect of weed en oil
(i) Increase in acidity, 0.1 0.1
mg KOH/g, (rnax.)
(ii) Increase in sludge content, 0.05 0.05
% (rnax.) '<.
....
V. Methods of Test IS : 1998 IS : 1998. \ r

block, bolt and nut, terminal board and core step filler, etc. Haldu and teak wood seasoned planks are also used as a filler material between core limbs and enveloping coils and as yoke-step fillers.

3.4.1 Evaluation of Wood

Details and significance of .tests conducted on densified laminated wood as per Table 3.7 have been described in Sees. 3.2 and 3.3. The quality ~f laminated wood may vary widely from batch to batch and sometimes even sheet to sheet. The.rcasou has been attributed to many parameters, e.g. species of wood selected for veneers, their location and age, number of joints in veneers to make up the required wid til, gap in joints and defects in timber, d.". Hence, utmost care is taken in their use.

50 Trans/ormus
l:
u
s
en
:z: :z: :z: :z:
0
o III 0 0 IIJ
l!)Q.. l!) oJ l!) ,
:z:- l!) ~ en ~~
_0:: Z <I:
0:: .... 0:: 0:: tn
Will a: w u " W ~
0.. > W > > u
a: 0-' > 0 -' ". '0
... u<t 0 u <I: u 0
III a:3' u ::I w
w a: 0:: 0 oc:x:
a: w;; ~ IIJ > lIJu
0..- 0..- .·• .. r' Q; :z:
-< <to <to r s <1:::1
en o..~ Q..!; Q; en
. '. ;'
I :0 .~ ...
.-
I
.1
I ..........
;; ...0
~I I <:»
I
"
I '~ .. _':::'

, ....

.;.;

...: o

-

o

::s "C C o u

... <I.> 0- 0- o o

"C <I.> VJ o 0- VJ

C «I ....

-

I (

i ,

! (

,

I

I,

,

!

r ,

... «I

I

"--t ~

Material» Used in TraM/aniler.: 53

r

(.) l '
~ ,
d
~
.S
."
0.
b .'
'"
.....
0
r::
.2
-
'"
0
0.
'"
r::
...
~
- .
.....
0
"0
0
.t:l
0
~
\,:)
,
M,-
~ '. \

S4 Transformers

and transposed conductors as welt as on paper covered stranded and ficxi-

ble cables. .

High air permeability kraft paper with properties in accordance with

. Table 3.5 is used as covering. All layers except the outermost arc. butt wound, tho outermost layer being overlap wound. Maximum -thickness of paper for butt wound layers is normally 75 jLm. Increase in dimension due to paper covering is measured by difference method as well as by adding up the thickness of paper used in each layer to ensure reasonable tightness of covering.

3.5.1 Continuously Transposed Copper Conductor (CTC)

eTC comprises an odd number of high conductivity annealed copper strips insulated with a polyvinyl-acetal based enamel, arranged in two side-by-side stacks with the individual enamelled conductors continuously

transposedthroughout their length. Additional. insulation is provided by If

covering the transposed/stack by a number of paper layers, as for rect- I

angutersingie conductors. Figures 3.5 (a) and (bj show the general view

and constructional details and Fig. 3.6 depicts the transposition of individual

strips in a eTC.

Pitch of transposition is the distance between adjacent cross-overs measured on one side of the conductor and is normally 15, times the individual strip width. At least one full transposition should be made in a full turn of the winding. Hence .. the minimum winding diameter is decided

by the relationship. . ...

pitch of transposition X no. of conductors Dmln =

eTC offers the following technical advantages over single rectangular. conductor (economic aspects are also taken into account for optimum design):

(a) Winding time is reduced, since the need for hand transpositions is partially -or completely eliminated.

(b) The winding-space factor improves noticeably compared with single rectangular conductors.

(c) By employing a number of small conductors which are continuously transposed, eddy losses, due to' circulating currents between adjacent turns are reduced and hence total copper-losses of transformer due to windings are a.so reduced.

(d) Transposed conductor offers an improvement in the mechanical strength of the winding.

\

I

(

3.6 CREPE PAPER COVERED FLEXIBLE COPPER CABLE

I

(

I

..... _(

./

(

Tapping leads and high voltage leads are required to be given sharp bends up to a radius of four to. five times the overall diameter of cable. The

(

«!

v-

\

I

\

I

\

I •• j,';

Materials Used in Transformers 5S I

[

cables used have high flexibility due to usc of a fairly large number of thin,

wires. Clubbed with the benefits of crepe paper mentioned in Sec. 3.2)

, I

these cables provide extra-high flexibility and damage to paper insulation'

at sharp bends is eliminated. -j

3.7 SEALING MATERIALS

For all sealing purposes, to avid oil leakage at joints, gasket and mould-; cd components made from nitrile rubber or nitrile rubber bonded cork j arc used. Characteristics ofth~se materials are given in Table 3.9. PTFE i and polyamide washer, tape, etc. are used for plugs and valves. \ I

:\

Table 3.9 Schedule of Characteristics of Gasket Materials

s!.

Requirement

\

i I

1 ,i

---------------

No.

Characteristic

Nitrile rubber bonded cork

Nitrile rubber

1. Hardness, IRHD 2. _ Tensile strength.

MPa (min.)

3. Elongation at break % (min.)

4. Compressibility at

2.8 MPa. % (max.)

5. Recovery % (rnin.).

6. Compression set, % (rnax.)

7. Flexibility

70±lO 1.75

60±S 8.4

25-35

80 85

'No breakage through the granules of cork or separation of granules should occur.

30

I i I

i

. \ i

1 1,-

II. Effect of Ageing in Air

8. Increase in hardness, lRHD (max.)

III Compatibility with Insulating Oil 9. Increase in acidity,

" rng KOHlg (max.)

J.O. Increase in sludge

content, % (max.)

IV. Methods of Test

N.A.

10

0.1

. ~ . ,

0.05

IS: 4253 (Part II) ,

IS : 3400 '(Parts I, II and X)

3.7.1 Nitrile Rubber Bonded Cork

This is the most widely used general purpose gasket. The cork gasket is adaptable even to rough unrnach incd surface by virtue of its compressibility under li~h! seal ing loads. This gasket. is made by the cornpour ding of

56 Transformers

granular cork with medium nitrile butadiene rubber. The cork used is clean, soft-grade typo, uniformly granulated and freo from hard board, wood fiour,'dust and, other foreign materials. Nitrile rubber contributes to the 'scalability' factor that enables the bonded cork to effect a tight seal under very light seating loads and also permits considerable distortion of gasket without crack and damage. The presence of cork allows compression without excessive sI?rea~ing associated with solid rubbers.

3.7.2 Nitrile Rubber

This is used as gasket normally with a metal limiter to avoid excessive compression and as sealing rings and mouldings, etc. Nitrile rubber is a combination of butadiene and acrylonitrile rubber conforming to compound BA 60 of BS 2751. The resilience of nitrile rubber makes it possible to reuse the gasket after opening. a gasketted joint. However, this requires machined flange surfaces. For smaller flanged openings, a groove is cut into the flange for seating of nitrile rubber gasket. In such a case, limiter is normally net required.

3.8 COLD ROLLED GRAIN-ORIENTED ELECTRICAL STEEL SHEET (CRGOS)

eRGO electrical steel with an approximate silicon content 0[·3% is used for magnetic circuits of a transformer"· . The following features influence . selection of the type of steel sheet.

(a) Maximum magnetic induction to obtain a high induction. amplitude

in an alternating field.

(b) Minimum specific core-loss for low no-load loss.

(c) Low apparent power input for low no-load current. (d) Low maguetostrictiou for low noise level.

(e) High grade surface insulation.

(0 Good mechanicalprocessing properties.

,. ..: ... ,f ~ .•

eRGO made from a ferrous base presentmaximum magnetizability, i.e. permitting a high induction. Iron crystallizes into a body-centered cubic lattice, with the cube edges of the lattice pointing in the direction of easiest magnetizability and lowest, core-loss. Grain oriented electrical sheet consists of a silicon-iron alloy, with the crystallites being predominantly oriented by means of a specific manufacturing process, in such a way as to _have four cube edges pointingin the rolling direction. and. the diagonal plane being

parallel to the sheet surface. In tbis way the rolling direction betomes--!· the direction of maximum. magnetic properties and approaching the ideal properties of the individual crystallite. The more pronounced tho texture,

the nearer the properties become to those of an individual. crystallite.

Materials Used in Transformers 57

Table 3.10 Schedule of Characterlstlcs of eRGO Eli!ctrtcal Steel

2

S. No. Characteristic . Grade M4 Grade M-2H

I. Physical Checks

t. Thickness, mm

2. Edge camber, rnm (rnax.)

3. Edge burr, rnrn (rnax.j

4. Waviness, (rnax.)

5. Sl;lc~ing fuctor, r. (miu.) II. i\ tugnetlc Properties

6 Specific core-loss, W/kg (max.) (i) At 1.5 T

(ii) At 1.7 T

7. Ageing test (Increase in core-loss after ageing at 150°C for 14 days), % (max.)

8. Magnetic induction at magnetizing force of 800 A[m (Bs), T (min.)

9. S!,~cific <lpparent power at induction of 1.5 T, VA/kg (rnax.)

It). Saturation induction, T

IE. Electrical Properties

I!. Resistance of surface coating, .Q (min.)

IV. Mechanical Properties (Typical) 12. Ductility (Bend test)

13. Tensile strength, MPa (i) Longitudinal

(ii) Transverse

14. Yield point, MPa (i) Longitudinal (ii) Transverse

15. Elongation, % (i) Longitudinal (ii) Transverse

16. Hardness, HV (load SON)

V. Physical Properties (Typical)

17. Density, g/cm3

18. Specific electric resistance, !LQ. em

19. Thermal conductivity at 25"C, 11m sK vi. l'vI cthods of Test

VII. Applicable Standards

3 4
I
0.27 0.30 !
,. 2 for 2000 nun sample length

0.05 0.05

5 mm for 2000 rnm long sample, height of an.! wave not to exceed 1/80 of the length of a wave, or I mrn, whichever is the greater.

IJ6 %.5

0.89
1.40 1.17 ,-
3 3
1.75 1.·85
1.25 1.0
2.03 2.03 2 for'80% of the readings 5 for 50% of the readings

Min. one 1800 bend without fracture

350 " 330
420 ~~" 390
330 315
360 325
(J S
24 30
175 170 7.65 48 26

BS GOl-Part 5

7.65 45 28

BS 601-Part 2 DIN 46400

JJS C 2553

AISI M· 4

--- -----------_._-_ .. _._----------

58 . Transformers

3.8.1 Grades of eRGO

Convention,ai grain-oriented (CGO) steels are made in various gradesM4, MS, M6.and M7. while low loss HI-B steels are available in grades M-OH, M-IH, M-2H, M-3H and M-4H. The schedule of characteristics of some typical grades commonly used in transformers are given in Table 3.10.

Specific core-loss is made up of hysteresis and eddy current losses.

Hysteresis loss is usually reduced by making an improvement in grain orientation. As the (l00) pole figures in Fig. 3.7 show) the mean deviation

RO

rio

HI'S

C.G.O_

FIG 3~7 (100)POlE FiGURES SHOWING THE GRAiN ORIENTATION OF HI-8 AND C.G.O

,

Fig. 3..7 (l00) pole figures showing the grain orientation of HI-B eGo. angle of the (100) axis from the rolling direction is about 3~ for Hi-B, which is far lower than the. average 7~ for eGo. Thus in, .. terms of hysteresis loss, HI-B is one or two grades superior than conventional material : (Fig. 3.8). Eddy current loss depends to a large extent on the sheet thickness, the frequency of the alternating field and to a minor degree on the electrical conductivity. Eddy current loss is usually improved by applying

a surface coating which generally consists of a. glass film and a phosphate coating (Fig. 3.9). This surface insulation, uniformly coated on both sides withstands stress relief annealing, without deterioration in its adhesion or electricalinsulation value, at temperatures up to 840°C. The mean value of the insulation resistance of the coating is at least 10 Q em? on each side.

:~ ...

\ .. 1
f
0
..:.
r ....
~
0
r
.,'
'-J
VI ....
~
1-4
;z
0
c
n
.... Q,
a .VI
;Z
r2'
~
_.
cD
0 i McJ/~rl(JIJ Used In Transformers 59',

HYSTERESIS LOSS Wh is tso AND Wh 17/S0 (W/l(.9)

o

N

o y

o

?

.. , t,..-

,.

I

I,

o

o

°

-6Wh1S/S010'12W/K9"1 0 0

,-,tl'-

. 0 _\~

o 0 0% ° I <)'?I :

II

•• -- -- - - - - - - r .9.0. '0 ,- - - - o - - - -- .. - -- - - - - . - - - - - ... -.

° 00 0

0000

o

J

~~o_'-------------oL---------------~o----------------~o----------------~o---7--~

6 ..... .:... N .~

U1 0 VI 0 "f)

HYSTERESIS LOsS Wh 15 50AND Wh 17150 (W/lb.)

.. ,

o

./. ... /.

"". . . • •

.: .

•

•• • • • ... '

°

o

• •

o

- J)\.~ -<J .• .p.. ~ __

°0 0 °0

008 o

\---- 6 Wh 17/50 O·16W/K9 ____.; I

Fig. 3.8 Effects of Ba (induction at magnetizing force Of 800 AIm) value on hysteresis loss.

Fig. 3.9 Grain arrangement of GO--Hcavy arrows show the (100) direction.

I .

.\

· (

Figures 3.l0 and 3.1 i show the improvcmcats in core-loss of HI-B stc~l over eGo. when the induction is oyer 1.5 T. Therefore, HI-B permits

the manufacture of transformers with the same no-load loss as before, but ''t\T

',~ higher induction amplitude with the resulting economic advantages such

as smaller size, less weight and reduced amount of material as shown in

Fig 3.12.,

1·1

1f'I' If'I ' a 'l'2 _,

W

a: 0' u

1·0

M6S !-A 6

(GO :~H9S8

-"1"971

M1H

M2H

.'

1975

0'10

THICKNESS (m.m:)

o· s

Fig. 3.10 Improvement in core-loss.

(

(

(

".' ~

Yw"'~ .i.J'£.~ ;e;~-

1-1
;::
iii
z
2 ,.,
...
0
~
a
z
-
1-2 HI80·)Omm

UWL- __ ~~~ __ ~ __ -L __ -L __ ~ __ ~ __ ~ __ ~ __ ~ __ J~O~1_n_rJOI __ ._C~'G'_O'_O_'1~O_rn_m~ __ ~__j.,

0·5 1·0 i-s 2·\J i

CORE-LOSS (W/Kg 1

Fig.3.11 Comparisou of B.\V<curvcs:betwecn"jHI-n and eGo.

.,

Fig.3.12 Schematic comparison between a three-phase transformer with a mean

induction amplitude and one with a high induction amplitude. ;

62 Transformers

• -
\
\
I 0
\ w
'::£
I ::>
,
t ~
11 «
II
..
" -_ .. -
\ \
\ \
\
\ .'
\
I
\
\
.
\
\
\
\
\
\
\ E E
E E
0 0
,.. ,..
.; "
0
m 0
. ~
x- u
I I
I
I
I ,
I 8
0
0' 0
-
o
U
- 0 -e
0 t:l
0 '" c:l
U\ :z ~
-:t ......
" ::I: ;r:
t:
~ '-'
0
ct: ~
e 0
.0
tb '"
0
z >-
§ ~ ....
- ...., u
-
w ~
Z I
~, ~
- U
0 ::£
0 I
OJ' A
._
0
t:l
0
."
';:::
c:l
0.
E
0
o
0 M
0 "";
M
-I
- I;t)
;.. ti;
s \\,

'\

\ \

\

o N

."

o

oc o

Materla!s Used 111 TrlUUiormcrs ~

."lo

(

Alternatively, transformers can bo manufactured with 5 to 20% reduction in no-load loss depending upon tho grade and induction amplitude usin,g HI-B steel.

i

I

i

3.8.2 Core-loss in Cross-grain Direction I

If magnetization is applied in directions other than the rolling direction; tde core-loss increases substantially. Core-loss is more than three-folds at 9~· to the rolling direction and more than four-folds at 60°. This aspect is taken care of while-designing the magnetic circuit of transformers. Cor~ loss of HI-B steel in 90° direction is inferior to those of conventional steel, .... The B-H curves in Fig. 3.13 indicate that HI-D has a lligh permeability. However, saturation induction of both IH-B and eGO is comparable.

I

Due to lower specific apparent powor of HI-B steel. 10 to 50% rcductiop

in transformer exciting volt-amperage is obtained. I

(

I

i

3.8.3 ~~striction Ii .'-

For environmental protection, the noise level produced by the transformets

has recently been a significant factor for the user. Such transformer no~o

. I

is generated by the iron core of the transformer, as well as from that made

by the fans. In an ideal core, the laminations of which are pressed t<?gether in such a way as to prevent any fluttering or shifting, the noise [s generated by the magnetostriction of the laminations in the altern a ting' field, causing the core to vibrate and thus act. as an acoustic source; If tQe . above mechanical conditions of the core manufacture are not observed, other causes (such as magnetic vibrations) may a] so be the reason for nois~.

Magnetostriction, i.e. the changein configuration of a magnetizable body in a magnetic field, leads to periodic changes in the length of the body ih·an alternating magnetic field. The frequency of this magnetostriction- is . twice as large as that of the existing alternating field, i.e. when magnetizing an electrical sheet with 50 Hz the fundamental frequency of'Inaglletostric,tion is 100 Hz .. The precess of change in length as a function of induction

is not strictly a linear one, so that the magnetostriction of (in alterna ting: field gives both a fundamental and harmonic. frequencies. The fundamental frequency is predominant in an alternating magnetic field up to an induction .of 1.7 T, while asa result of the sophisticated process of alternating con~. tractions and expansions, the harmonic frequency prevails for induction

above 1.7 T. Magnetostriction is minimum in the rolling direction, whilst:-' . -

.....•. ·1- .

in the 90°-direction it is maximum. Except for the 90°-direction, all other .

directions -that present major deviations from the rolling direction have a' high percentage of harmonic frequencies in the spectrum, which are particularly undesirable because they generate harsh noise in frequency ranges) where the human ear is especially sensitive. Therefore, for sheets used in transformers, all transverse and oblique effects are kept to a minimum and very high induction amplitudes are prevented in the rolling direction .

. ~------.

Magnctostriction of Hl-B is lower than that of CGO.(Fig. ~.14) and consequently 2 to 7 dB reducti'on in the noise level of transformer is achieved .:

xi~r-------------------------------------------------~------~

<l

%

a

>: ~. ~

a:

I- 11\ a tw

:z: 2. I!>

<t

:f

0---0 HIS C?')Omm <>- __ '--.0 C.G.O. O-lOm m

60HZ'

I •• ' ,._:.. _.._ .:... .. " _.;A __ .../

»<" •

. ~_.o.,--<>---<>-_.(~-

1'0

1-1; . IE'

INDUCT ION .8 (T )

Fig. 3.14 Peak-to-peak magnetostricrion against induction for Hl-B and eGO.

3.8,4 Stress Sensitivity

Mechanical stresses exert a significant effect on the magnetic properties of eRGO electrical sheet. Such stresses are-introduced in a sheet. by way of external forces and' by way of' plastic deformation (internal stresses). External stresses are developed due to the following:

(a) When the sheets of a core are exposed to. tensile or compressive stresses exerted by the fixings with the longitudinal forces in the limbs and the yokes.

(b) Bending stresses, where the sheets are forced into waves (for instance, as a result of heavily' tightened clamp plates or end

frames). ·s~

(c) Bending stresses .due to deflection (for instance, in the case of nonuniform corner jointing).

(d) When. use is made -O("wavyor bent sheets, which are necessarily flattened in the core.

(e) Y\Tl¢en sheets cannot adopt their natural condition as a result of friction existing between them.

Internal stresses occur along the cut edges during each cutting operationand.as a .result of bending the sheet or subjecting it to tension beyond the yield point (in slitting lines, continuous furnaces , etc.). Bending stresses can be divided into' components of tension and compression.

T

, I

J 0

(

/ I I

Similarly in simple cases, internal stresses can 00 resolved into pairs of tensile and compressive components. A few examples are given in Fig.3.l5:

BENDING STRESS STRESSES

. FOOC(o FLATT_BETWEEN SHEETS

~~I~~~S ENIN(i OF A fCOMPRtSSION

CURVED _ ..•.

TENSION) _{jCOM- \STEEl .!..

'IPR£S

1510 N _,A .... H,....,::I

~ :

I g! 'h-'~~~~~-

I ;;

~MPRESSf- 1\ t~

I W I t-

TRUE SHAPE COMPRESS- FRICTIONAL LAYER ION

I (

, (

; ; ,

) (

. TENSILE AND COMPRESSIVE' COMPOt~EtHS ' IN CASE OF EXTERNAL J.(ECHWICAL STRES~

lONES CF INTERNAL ~TRESS

TENSILE AND COtz1PRESSIVE COMPONENT IN CASE OF INTERNAL STRESS IH A STRIP SeCTION

,.

Fig.3.15 Examples of mechanical stress in a sheet ••

d P2 RELATtVE CHANGE

..1 Ps

) ,

) I

) i

-to - 8 -6 -4 - 2 COM?RES'310N

RELATIVE CHANGE IN APPARENT POWER

ABSORPTION Pz. AND SPECIFtCCORE-LOSS Pr;; AS A RESULT OF MECHANICAL STRESS '0' -

MAGNEtiZATION AND STRESS IN ROLLING

DIRECTION. .

Pz ~P_O

-~"j 1'=+=""=; ~

o '2 j~ 6 8 10 MPo.

TENSION

Fig.3.16 Core-loss and apparent power absorption of CRGOS as a function of external rrr.chanical stress.

Figure 3.16 displays the specific core-loss P, and the apparent power 1\ for an induction of 1.7 T and various levels of mechanical stresses. Both P, and P.: arc n ighly dependent all t lic compressive stress but only slightly dependent on the tensile stress. Thus, the ar parent power is more sensitive than the core-loss.

---------- .

66 Transformers

:x:

., ...

~t.... m.;...

" "

(D'm

:z:
~
'" c
-0%,
w 0
I- 0
·c
CD ::l
._
C'!
':a
CI)
0
CJ
~
U
<...
0
C
0
u .n
._
...... ~
-
'" Lo
0
'0 '"
c ~
co
til J:l
Z E '"
:z: Q "0 .s::::
"0 u
r- 0
U t:::, E
<t bO -
a:: .S -;;;
... .._. J:l
Z "" t..
0 C 0
'-0 ~
U ~ 0
'< ".
0
e-
,....
:z: r.
<L> 0 ee
I ii\ G
\J\
W
et:
Q..
::£
0
V
cc
II
L "

a e,

::z:

."

/

"t"

> i
I
r- I
i
(
f
\ t . Materials Used in Transformers 67

Figure 3.17 shows the dependence of'magnetostriction upon the tensile and; compressive stresses for an induction of 1.7 T. Here too, effect of com-] presslve stress is much bigger than that of tensile stress. Hence compression beyond a certain level occurring in the plane of the sheet is harmful and" therefore, stresses must be removed by stress relief annealing. Sheets: affected by waviness or coil-set are maintained to a fiat shape since they; are subject to elastic stresses when being nested into the final shape of the' transformer cores due to pressure. Buckled sheets must be discarded. Stress: relief annealing of laminations is done in a continuous rolled hearth-type: furnace at 820±20°C for approximately 2 minutes in an air atmosphere and cooled quickly to ambient. temperature.

Internal residual stress inherent in HI-B strip in as-delivered condition is] lower than' that In eGo., thereby stress-relief annealing may be omitted in: many cases.

3.9 STRUCTURAL STEEL

Structural steel mainly in sheet and plate form is used in the fabrication \ ' of transformer tank, radiator, conservator, clamp plate, end, frame, mar-: shallingbox, cable box, roller, turret and inspection cover, etc. Normally;' standard quality mild steel is conventionally usedfor core damp plates. However, it is economical to use high tensile strength (HTS) steel plate.. since by reducing the thickness of damp plates, magnetic core area can be increased, resulting iii improvement in the utilization factor. In such 3. ' case, lifting pins required to connect Clamp plate to yoke end frameare: manufactured from HTS bar (1.5% nickel-chromium-molybdenum). having; a very high tensile strength in the range of 900 to 1050 MPa.

Stainless steel (austenitic chromium nickel) plates are used as cover for: turrets of high current bushings to neutralize the effect ofeddy current. • For medium currents, magnetic circuit perpendicular to a current carrying; conductor is broken by welding stainless steel inserts. These plates are' titanium stabilized for wc1dability with standard quality magnetic steels. Depending upon the requirements of application of the component.. particular quality of steel is used, details of which are given in Ch. 10.

Mechanical properties of a few qualities of structural steel are given III Table 3.11.

3.10 FUTURE TRENDS

In the recent years, intensive efforts have been undertaken to develop new transformer insulation systems which permit a more compact and rhus more economical design. However, these have been, more or less; limited to small transformers up to a few MVA. Dry type transformer,

1

\

i

i

,

, 1-

\

\

!

I

, !

I

\ ,

I \

6a Transformers

Table 3.11 Pr opertles of Structura I Sled

SI. No.

Material

Tensile strength MPa

Yield strength MPa (Min.)

Elongation %, (Min.)

1. Cold rolled carbon steel sheet

2. Hot rolled carbon steel sheet

275 330-410

205

18 (on80mm gauge length) 25 for thickness

.' above 3 mm.

3. Structural steel Std. 'quality Nominal thickness/diameter, mm (i) 6 to 20

(it) 20 to 40

(iii) Over 40

4. Bright steel bar and section-cold drawn

5. High tensile strength structural steel plate

Nominal thickness/diameter (mm) (i) 6 to 16

(ii) . 16 to 32

(iii) 32 to 63 (ivj Above 63

6. HTS Steel Bar(L.5 % nickelcbromium-mdlybdenum)hardened and tempered

7. Austenitic chromium-nickel-steeltitanium stabilized plate (stainless steel)

410-530 410-530 410--530 430 (min.)

250 2AO 230 335

23 23 23 14

,
540 (min.) 1')0 20
540 (min.) 340 20
510 (min.) 330 20
490 (min.) 280 20
900-1050 700 15
520 (min.) 40
_, silicone oil fined transformer, the vapour cooled and Sf's-insulated transformer are already available today. However these special transformers are used for a specific purpose, e.g.- for weight reasons in locomotives and are normally costlier than, conventional oil-cellulose transformer. The current level of knowledge and higher cost put limitations 011 these new insulation systems: Studies have shown that conventional oil barrier insulation system will cause noproblems up to operating voltages of 2000 kV. Hence, it is very unlikely that oil impregnated cellulose insulation system will be replaced by a new insulation system for large size power transformer, at least in the near future.

To increase the short-circuit strength of windings, annealed copper conductor is already being replaced by controlled proof stress copper. Copper-silver alloy will replace high conductivity copper, specially in higher rating transformers. Under short-circuit conditions in the windings, copper-silver alloy conductor is less susceptible to ariI1ea1ingand is thus easily able to retain its strength. Mechanical strength of transposed conductor is further, increased by' bonding together allthecouductors in a stack, by use of an epoxy bonding enamel over PV A enamelled strips. These changes are likely to take place faster, to cover almost.all transformers.

(~ Materials Used in Transformers 69

M

( However, copper is expected to be used unless it is replaced' by a new

~ generation development' of a superconducting material, since anodized

~. aluminium strip conductors have their limitations for HV windings

( I

~

r

j r

Another possible substitute for grain-oriented electrical steel for magnetic core of transformer is amorphous steel, also referred as metallic glass,

. which is a non-crystalline solid created by rapid quenching of metalmetalloid alloys. Amorphous steel (typical composition FCIlI Bn.s Six.s C2).of thickness 2S to 50 p.m and widths up to 100 mm are now commercially available It is expected that 175 mm wide strips would be commer-

. .

cially available shortly. However, there is apprehension about tho

technological capability of producing metallic glass in widths of 800-1000 mrn to match with CRGO width. There are numerous practical problems to be sorted out, e.g. complete transformer will have to be annealed for stress relieving, therefore pre-wound itransformer coils are to be made suitable to withstand annealing temperature. Further, saturation induction of metallic glass is lower than that of CRGO and hence transformer has to be designed at lower induction. Experimental transformers of maximum

. rating of 100. KVA have been made so far, using amorphous steel.

Nevertheless, the material holds some promise for the future, as it is expected that the problem due to its present prohibitive cost and other limitations can be overcome.

REFERENCES

i

,.

I. Viswanathan, P.N. et al.,

"Paper for Cable Insulation/rom Indigenous Wood Pulps". National Seminar on EleCtrical Insulation held in Nov. 1974 sponsored by the Institution of Engineers (India), Electrical Engg. Division, lEMA and lSI.

2.
I
i 3.
.i\
4.
,
:\
·i· 5.
:( Philip, P.K . ."Insulation Pressboard for Transformers" Electrical India, Vol. XIX. No. 24, pp. 5-13, Jan. 1980.

BICC Connolly's U.K. Publication No. 798 'Continuously Transposed Conductor'.

Thyssen Grillo Funke, VI.G. Brochure, "Grain Oriented Electrical Sheet", 1980 edition.

"Technical Data on Orient-Core, Hf-D"

Nippon Steel Corporation, Japan, Cat. No. EXE 367 Dec. 1976.

:(

,

.(

"

(

i \

::/

~\.

...... ----.- --.--

Magnetic Circuit

K.N.LABH

Senior Manager (Design)

In a transformer, energy is transferred from one electrical circuit to another through the magnetic field: Transformer core made 6 Ha:lnirlated steel sheets provides the magnetic-circuit. for theflowofm~grtetic flux mutually linking the electric~lcitttiits~ -As'against tbe ~~~bfe, -rrori_"core ' provides, a comparatively low reluctance path to the magnetic flux witb consequentbenefit of (a) smaller magnetizing current, (b) increase in the total fl'!J.x<linkHge and (c) a high ratio of mutual to leakage flux resulting in . reduction of stray losses. Its design, type and manufacturing methods have significant bearing on quality, transportability, operational limitations and guaranteed tecbn~cal performance of the .transformer.

4.1 MATERIAL

Some of the very early transformer cores were made of inferior grades of , ,,--. _j.9,minated steels which had, inherently higher core losses and-showed pro.' nouncedageiug effects, further aggravating the hysteresis component of iron'::lbsses 'rri 'ih{--equ·ipment. It was subsequently found that-very small quantities of silicon alloyed with low carbon content steel produced 'a

• . 6

material with low hysteresis losses and high permeability. These steel sheets

, alloyed with silicon mitigated the problem 'of . ageing and improved thepermeability and consequently 'reducing the magnetizing current and core losses. In the ever increasing pursuit of increasing the power ratings and reduction ofcore-losses, another innovative technique from steel manufacturers came in the form of cold rolling with orientation of the grain in the direction of rolling This core steel known as CRGOS (cold rolled grain oriented silicon steel) has the minimum epstein losses to the flow of magnetic flux along the directions of grain orientation and this material

is universally .. used for tbe_manufacture_oLtransformeLcru::.es...__ ._ ...

The adoption of CRGOS has brought about considerable reduction in the specific iron-losses (W /kg) over the earlier grades of core steel. However, CRGOS is susceptible to increased losses due to flux flow in directions

i

i

Magnetic Circuli 71

I

other than that of grain orientation, effect of mechanical strain duo to i

clamping pressure, bolt holes, jointing of 11mb with yokes,' etc, Apart i from this sensitivity to the direction of rolling, CRGOS sheets are also very susceptible to impaired performance due to impact of bending, blanking and cutting. Both surfaces of (he core s.tccl alicots uro provided with an insulation of oxide coating (commercially known as Carlite). (The, .stacking factor of lamination improves by using thicker laminations/out! ~dcfy·'currerit loss goes up in proportion to" square of the thickness of the! lamination. For reducing the eddy current losses, thinner laminations:

"

arc preferable even though the stacking factor goes down. Dcburrjng of :

, .-

the laminations improves, the stacking factor and minimizes the cddy :

losses. After machining, the material has to' be annealed at 80O-900°C in a neutral gas environment. Tile material properties of core stool are :

I

further discussed in Ch. 3.

4.2 DESIGN OF MAGNE'i'IC CIRCUIT

( )

For a transformer design, tho basic governing factors are (a) Rating of transformer and its performance,

(b) Operational conditions,

(c) Transport limitations (i.c, height, length, width and weight, etc.),

The design of the magnetic circuit, i.e. transformer core is also based; on the above considerations and it has significant bearing on the overall' economy of the transformer. For CRGOS, satura tion may occur at tho: ,,, magnetic flux densities exc~eding 1.9 tesla. Based on the input voltage'

......,.._.~-~:z.c...~ ---____ .

and frequency variations, a sllitabl"e"'"vaIue of flux density can be adop=r

ted to av"oid-al1ychance-6f core saturation under operating conditions. By i increasing the operating value of magnetic flux density (magnetic loading); the net weight of core can be reduced, but this leads to corresponding! increase in the core losses. One has to compute an optimum value of the: magnetic flux density, keeping in view all the above considerations,

4.2.1 Constructional Features

The trpc of transformer cora construction depends 011 the technical parti-! cular: of the transformer and transport' considerations. In general it is: prefer able to accommodate the windings of all the three phases in a single. core frame. Three-phase transformers are economical over a bank of three i single-phase transformers. /Another important advantage of three-phase;' transformer cores is that the component of the thjrd and its IIHlfilple : harmonics of mmf cancel' each other, consequently the secondary voltage ! wave shape are free fwmdistorti'o'ns due to the thirdharmonics in mrnf. i! However, if the three-phase ratings are large enough and difficult to i transport, one bas no choice but to go for single-phase transformer units. !

For single-phase and three-phase transformers, the cores can be broadly classed as

(a) Single-phase three-limbed core (b) Single-phase two-limbed core (c) Three-phase three-limbed core (d) Three-phase five-limbed core.

(a) Sing le-phase-three-limbedcore Fig. 4,1 (a)

The windings ate placed around' the central limb,' also known as main limb, The main magnetic flux generated in the central limb gets divided

1-MAIN' liMB

2 TOP YOKE'

3. BOTTOM YOKE

2

Fig. 4.1(!!!) Single-phase three-limbed core.

."'. :.

1. MAIN LlM8 2· TOP YOKE

3· BOTTOM YOKE

.\ L.~~__.J

Fig. 4.1(b) Single-phase two-limbed core.

(

I

Magnetic Circuit 73

into two parallel return paths provided by the yokes and auxiliary limbs. For tho same magnetic flux density as that in the main limb, tho auxiliary limbs, and the yokes need to have tho cross section only half of the main limb.; This type of transformer core is generally preferred for, single-phase trans-] former, as:this is more economical than two limbed construction discussed. below.

(b) Single-phase two-limbed core (Fig. 4.1 (b».

Sometimes, the single-phase power ratings of transformers are so large thatli if the windings of full power ratings were to be placed Ott the central limb, I its width would become too large to be transported. To mitigate such diffi-': culties the windings are split into two parts and placed around two separatcl limbs. Here the cross-sectional area of the legs (limbs) and the yokes are! identical. Consequently these cores are bulkier than the single-phase thrce-l limbed arrangements. Also the percentage leakage reactance for tbis type!1 of core construction is comparatively higher due to distributed nature of; the windings in the two limbs. separately.

(c) "1 'hree-phase three-limbed cores (Fig. 4.2). This type of ' core is generally used for three-phase power transformer of small and medium power ratings. Each phase of the winding is placed around one leg. For each phase of magnetic flux appearing in a limb, the yokes and the remaining two limbs provide the return path. If the phase! fluxes are denoted as t}A, ,pB, ,pc, their phaser summation at any instant of

" .

1\

)

,--------_._----------_._---_. __ ._----

--------------~~2.

, . I

___ --------~~3

l__ ~----------~

..

Fig.o12 Three-phase three-limbed Core: (1) main limbs, (2) top yokes, (3) bottom yoke.:

74 Transformers

time .is identically zero, which can be mathematically stated as ~A +91l+tf.C ='0. In this type of construction, all the legs and the yokes have identical cro~:; section. " .

I i

I

~

Fig.4:3 Three-phase.five-limbed.core: (J)Olaln limbs, (2) top main yokes, (3) bottom main yokes, (4) auxiliary top yokes, (5) auxiliary bottom y-okes, (6) auxiliary l irnbs.

(d) Three-phase, five-limbed cores(Fig. 4.3) .

For large rating' power transformers, cores have to be built in large diameters. In case of three-phase three-limbed cores, the yokes have the same

3
r
-,
f'\
;__'i
~
~ '~
I,
r·
2 .\
;_-c;·":- I
i
\ I! I

~------~-.. 0 s.: ~-,--------J~

Fig. 4.4 Circular core cross section: D-diamctcr of core, H-total lamniation stuck height, (1) iaro:natv.;·,,·: '~'l oil duct, (3) steel Clamp plates.

(

Magnetic Circuit !:J

diameter as the li~bs. In case of large diameter cores, the overall core height will go up leading to transport problem. For such cases the yo~e cross sections (and consequently yoke heights) are reduced by approximately 40% or more and auxiliary paths f~r the magnetic flux arc providcp through auxiliary yokes and limbs. The cross section and the height of the auxiliary yokes and limbs are lower than that of the main yokes.

I

4.3 OPTIMUM DESIGN 0:£1' CORE I

'I

I For the optimum design of magnetic core, the following aspects have to bp

decided.

'(a) Constructional features ' (b) Core cross-sectional area

(c) Number of oil ducts and location

4.3.1 Core Cross Section

i

, I

The ideal shape for tho section of the cores is a circle, since this would

waste no space beyond that taken up by the insulation between laminations .' A perfectly circular CON section, however, involves making ~ variation in dimensions for each successive lamination, which is possible but uneconomical. As a compromiso solution, the 'core cross section is made by laminations of varying widths ,and packet heights in such a wa'ythat the overall section approximates a circle. Such a typical core cross section is shown in Fig. 4.4. Oil ducts are needed for cooling the core, leh the hot-spot temperature rises dangerously and their number depends 6~ , the core diameter and the specific core-losses, which is a function of operating flux density. Additionally, clamp plates made of steel are needed on either' side of laminations for effectively clamping the laminations, These clamp plates should' be mechanically strong enough to prevent

I

buckling/bending of laminations and be able to withstand the lifting load

of core and windings and axial short-circuit forces. The steel sheet laminations, oil ducts and the clamp plates should all lie within the core circle. '

The net sectional area is calculated from the dimensions of various packets and an allowance is made for the space lost between laminations (known as stacking factor) which for sheet steer of 0.28 mm thickness with carlite insulation coatings is approximately 0.96. Area is also deducted for the oil ducts. The ratio of the net cross-sectional area and the gross area of the core circle is known as utilization factor (UF). By increasing the number of core steps UF improves. This, however, also increases the manufacturing cost Typical cost effective values for the number of steps (i.e, one-half of cross section from centro line) lie in the range of 6 (for smaller diameters) to 15 (for large diameters), For any particular core diameter based on other design considerations, this gives out not only tile

,

"

76 Transformers '

optimum area: and thereby reduction in theflux density and consequently iron-loss, but also helps tire designer to revert to a lesservalue of core diameter, wherever the computational margins allow this latitude. Seen from another angle, improvement in the core utilization factor increases the core area and hence the value of volrs/tur n for 'any particular core diameter and specified flux density. This, in (urn, results in the reduction in winding 'turns and thus reduction in copper. Therefore, core area optimization results in better economy of transformer designs. 0'

In the following, the core optimization is discussed under (a) optimum selection of laminations (b) optimum oil ducts.

4.3.2 Optimum Selection of Laminations

For any particular core diameter, the first and foremost point to be decided is the maximu 111 allowable height of the lamination packets;' 1 his, in turn, is determined by the design consideration of clamp plate (see Fig, 4.5) and the pertinent constraint in accommodating them inside the core circle.

h'o-------- F ('Y)

'\ "",(v - l v .

- ,,~\. "i

V \

J

t------~---D------..._._..---;

H

Fig.4.5 Circular cross section of core-dimensional relationship.

For a circle of diameter D, the length of the cord at distance Y from the centre is given by

(4.1)

Magnctlc Circuit ,77

If H is the maximum allowable packet height, the minimum allowable lamination width is given by

."

,L",/n. = 2

The maximum allowable width of the central packet is usually cqre

diameter less G mm (to accommodate wooden packing) :

I

(413)

I

If the core is to be built up in N~ number of steps, where widths oflaminations of individual steps are LI, i = 1, 2, ... Ns, these values must satisfy the

following relations .

I.e.,

Lmax. =D-O

i = 1· 2

. "

.. " NI

For building tho circular cross section of the core in say NI steps, wo have , to decide NI different widths of laminations to be stacked one over the

. I

other as shown in Fig. 4.4. The packet height of each lamination 1 is

computed by the difference of heights of cords equal to the width lor laminations and the adjacent next width below it. The gross cross-sectiorial area contributed by individual lamination packets is obtained by lamination widths times packet height. This is taken 2.S the return function for selection of a particular width. The optimum selection of .laminations is formulated on the method of dynamic programming. Recursive relationships are formulated, for return function and state transition function for each step 'of

<,. decision taking. This method gives-optimum selection of laminations-for

building in N, number of steps. ' ;

Out of the optimum area available by the above method w~ must dedJct areas for the oil ducts. The individual packet heights should be reduced slightly to allow for the overall manufacturing tolerance in xrhe core bu'ilt up. A computer program has been developed for the optimum selection 'of lamination packets based on the above algorithm. The computer program also computes (a) optimuni number of oil ducts and their locations and (b) core lamination dimensions and other details required for manufacturing activities. This has automated the complete transformer core design and manufacturing informa tions ... /

4.3.3 Computation of Optimum Number of Oil Ducts

Core-losses (iron-losses) take place due to magnetic flux flow in the 'Iaminations. For CRGOS these losses are minimum for flux How along the direction of grains orientation and maximum for flux flow along cross grain directions. Hot temperature spots arc developed inside the core as a rcsul t of core losses. and designers must ensure that the hot-spot temperatures are well below the permissible values .

. ~~ .

.. _--- _-_.-- ---~--------

78 Transformers

rHEAT flOW A~ROSS LAMINATIONS

I _ HOT SPOT .,

..... [l;.A_T _n.O~W_ ~!:P_N_GlL}~I~A_TI_Ol'l.? _~ x· OrR

'I

,

2Y

\------

2x

·1

Fig. 4.u - Rectangular stack of core laminations heat flow conditions.

Segments of the core cross section are rectangular as shown in Fig. 4.6.

Heat dissipation takes place along the laminations (x-direction) and across the laminations.Iy-direction). The temperature gradients along the .two directions are gh:,en by

... : .....

r, = W.x.(. 2~1 +h) r, = W.y.( 2"i2 +h)

(4.5)

(4.6)

Where

w - specific cross gain loss, W ikg,

Ks = thermal conductivity along the laminations,

WrC{mm"

K2 = thermal conductivity across the laminations,

~",Tlor"l ............. "I '-",U.Ull,

-li = surface heat transfer coefficient,OC{mm2/W,

Tx = temperature differential from hot spot to oil, assuming that all heat flows along the laminations.

T)" = temperature differential from hot spot to oil across lamination, assuming all heat flows across la.mination.

The hot-spot temperature gradient

Th = Tx·Ty (4.7)

,.T x±Ty, , __ ~ __

T/; (max, permissible) c::: maximum permissible hot-spot temperature+-Icil rise-l-ambient temperature)

I

\

I i

I

."

!

Magnetic Circuit 79:

Equation (4.7) can be rewritten as r.. r,

T» <= r:-=- Ti,

W.y ( 2~2 +") =

(4.8.1)!

I J !

i.o.

The value of W depends on the type of core construction and remains constant for a particular type of assembly. The value of T x increases: parabolically with the value of x. An interesting condition arises when;

i

x 0::::> x* such that

T... ~ r, {4.9}

The equality condition of (4.9) implies that y =~ 00, i.e, for x c.:> x·, Ts will! never achieve its specified maximum value for finite value of y, and, along: the lamination, dissipation of heat-will be sufficient to maintain the value' or Tit below the specified maximum permissible value. The second corr-' clition that T', <Tit leads to negative value of Tv, which is not physically' possible. This also implies that along the lamination, dissipation is: sufficient to contain the value of Ts below the specified limit. For such' conditions as obtainable in relation (4.9), no oil duct is required for cooling.

We now examine the case when Ts z-T; .. Equation (4.8.2) can be solved!

, I

for y as -, i

'. i

1 . '

y* = 2CI (-h+ V h2+4Cl. C2) (4.10)!

,

,

I

~

J

where

. (4.,1 1):

and

Here, we introduce the concept of critical height y~ y* obtainable fromi Eq. (4.10). For an r stack-height 2y such that y~y* for a specified value of x, the hot-spot t anperature will never exceed the specified value of r.; For actual stack h .ight H>2y*, cooling surface (in the form of oil duct), is required if the h rt-spot temperature is not to exceed Ts. :

For any core di .meter D, the minimum number of oil ducts is calculated: by iteratively US! 19 the Eqs. (4.10) and (4.11)·for computation of cl:iticali· heights y* startin] from the central packet. Since the above formulation' 'stipulates that the maximum hot-spot temperature in the packet will be less:

than or equal to : ne specified permissible value, the above calculation gives, the optimum nUD: ber of oil ducts. '

4.3.4 Location -fOll Ducts

For more than 0 ie oil duct, a suitable criterion is required regarding their location in the core section. For single duct, the obvious choice is its

8Q Transformers

location at the centre. Minimum difference of hot-spot temperatures (ideally zero) obtaining in different regions divided by the oil ducts is adopted here as the criterion of location of oil ducts. Figure 4.7 represents the case of two oil ducts in the core cross section.

ZONE-2

DUCTS

2Yl

....

. .

'"" .

Fig. 4.7 'Two oil ducts in core.

For Zone 1

1

. ~

1 ~

_" _ _L_"""_

Txt 1" Tyl

(4.12)

.. , ~'.,.' -

---

Where

T,;l' . w.x{ 2~1 + h) Tyl = W.yt ( ;~2 + h)

and ,. Ti« = Hot-spot temperature gradient of zone 1.

(4.13)

For Zone 2

wnere

1 "1 1

_" - ==~""_ +-"- ..

Th2 Tx2Ty2

W (X2. _L 1\ Tx2 = .X2\ 2Kl 1 .1)

Ty2 = w.y:{ ~2 + h )

(4.14)

.J

(4.15)

("

Magnetic Circuli 81

and Till c:=: Hot-spot temperature gradient of zone 2.

By equating the hot-spot temperature gradients of zones 1 and 2, we obtain the following:

2KJ 2K2

WXI (Xl +2Klh) + Wyl (yl + 2K1h)

2K2 2K2

c= WX2(X2+2Klh) + WY2 (y:z+2K2h)

(4.16),

By cancelling the common terms and expressing tho difference of expres-' sions of l.h.s. and r.h.s. along with a normalizing constant we compute, a; I

difference term as i

i

Co ( 1 1
" =:: P XI (Xl + 2K{ h) X2 (x2+2K2 h)+
I,
s: K:z ) ,
, . "y, , '\ Y:z (y:z+2K2 h) (4.17): .
y~ VI T .t..o..l fl)
) where K:z == K:z/Kl } (4.i8)~
and p = normalizing constant
Yl+2Y2 = H/2 1 'I
X:z == J ( T Y - yi (4.19)~ . !

Since Xl is a constant and y:z and X2 are related to X and Y'by Eq. (4.19),:

we have .

i

I

.Co = f (yl) (4.20):

I ;

The. desired value of Y is the one whichsatisfies Co c:=. O. This is obtained!

·1-..v nn iterative T'>"I"\,.",£1" ... ", 0'" computer Th ... distance I"\f' oil ducts +_,-"""i

UJ 0.1 1.,,'" "',YV l'Jr.""""'''''''''W-'&'V La. v .f,J...A. ~"'_.L. .£.L&.'" 'U.J. ""'- """"'" VJ. ,1. U v\. J.J.'V.LJ...l.

centre of core circle are similarly computed for nu.mber of oil ducts more'

than two on similar principle. .

I In the foregoing, the broad concepts of optimum design of transformer'

cores c f circular cross section has been discussed. Based on the optimum' 'core cross-sectional area and operating flux density, the windings for: individual phases can be designed and from .this .the core window size can:

be worked out. This decides the overall transformer core frame dimensions' for a particular type ofcore construction as illustrated in Figs. 4.1, 4.2' and 4.3.

4.4 MANUFACTURlNG

I ,

By adopting suitable technology for transformer cores and exercising care!

. I

in handling of laminations 'and core building, it is possible to achieve !

82 Transformers

(a) higher reliability

(b) reduction in iron-losses and magnetizing current (c) lowering material and labour cost

(d) abatement of noise levels

It is necessary to apply quality checks at different stages of manufacturing. to ensure quality and reliability. The core steel samples should be tested regularly for guaranteed epstein iron-loss values. The laminations should also be visually inspected and the rusted lot should be rejected. CRGOS sheet steels are susceptible to impaired losses due to cutting, punching) piercing, bending, etc. During cutting and piercing the edges develop burrs which may. cut, as knife-edge, the insulation coatings on the adjacent laminations, in addition to lowering the stacking factor. It is imperative that laminations are deburrcd and anncaled(for stress relieving) so that iron losses do not increase. For reducing the transformer noises) the laminations should be tightly clamped together ~nd punch holes should be avoided as far as possible. The air gap at the joints can becontrolled by working on tight tolerances so.that value of magnetizing currents are kept toa rninimum. These CImin.ufacluring, aspects are discussed inmore detail' in the following: .:

\ I I

4.4.1 Corner Jointing of Limbs with Yokes

Broadly speaking, the core losses can be split into (a) loss. due to magnetic flux flow along the direction of grain orientation (with grain iron-losses) ~~d (b) flux flow in cross-grain direction (cross-grain losses) occurring in the zones of jointing of limbs with yokes. The cross-grain.losses depend to , a large extent on the type of joints.- The two most commonly .used types, of corner joints are (a) interleaved, (b) mitred.

Fig.4.8 Three-phase three-limbed core-interleaved joints

(a) Interleaved joints (Fig. 4.8)

Interleaved joints are the simplest, from the point of view of manufacturing. However, in the cross-grain zones the magnetic fluxes leave/enter the laminations in perpendicular direction to the grains and these losses are

I' I

I

I !

.. .-;;;._==*"..,..·1'-""_.,..... ..... __ . _ .

I

I

I

I

1

i I

I

I

I

i "'{(Ilin~tlc Circuit IU

comparatively higher. Such type of joint ings are usually preferred only: for small rating transformers, where the total core-loss itself is very i

. .

small.

J

(b) Mitred joints (Figs. 4.9, 4.10 and 4.11)

When the corners of the laminations are cut at 45°, the jointing is known: as mitred joint. The cross-grain losses for this type of jointing is minimum. : as the magnetic flux leaving/entering at tho joint finds a smooth path i of its flow. This, however, entails extra manufacturing cost for preparing I the corner edges of the individual laminations. :

-. .. /

~
~ -, Fig. 4.9 Three-phase three-limbed core mitred joints.

Figure 4.9 shows one arrangement of laying. the laminations for three- . phase three-limbed cores with mitred joints. Figure 4.10 shows the lami- : nations laying arrnngement for the three-phase five-limbed cores with mitred; joints.

Fig. 4.10 Three-phase five-limbed core mitred joints.

84 Transformers

( \

Sometimes there may be constraints in preparation of laminations beyond certain. widths. If lamination widths larger than these arc: required. .as in case 01 large diameter corcs.uhcse are split into two halves, so that these can be handled easily in manufacturing.A typical core built with such split laminations is shown in Fig. 4.11.

- 7

I ' I

.. 1\

~ .. \

Fig .. 1. 11 Three-phase five-limbed core with mitred joints with split laminations for large core diameters.

4.4.2 Preparation of Lamination Sheets (a) Slitting

For building the transformer cores, lamination sheets of different widths and packet heights are needed. The manufacturing schedule may include cores of different diameters and differenttypes of constructions necessitating slitting laminations in many widths and lengths. CRGOS rolls cannot be orderedin so-many different widths and quantities. These.rolls are available instandard widths (If say 760~ 790, 840) 91'S,lOOOmm) etc, For slitting operation, some w ic ths can be combined together by suitably adjusting the cutter distances ie the slitting machine. Figure 4.12 (Plate 2) shows slitting operation of laminations.

It is evident that full width of roll cannot be utilized' at any time of slitting operation and the leftover material will vary from stage to stage and depending onthe widths selected in combination during the process of slitting. The meticulous care in planning is imperative to minimize wastage of core steeL

The slitting operation has been formulated based on the principle of dynamic programming to compute the optimum schedule. This provides the optimum combination of different widths to be taken together and the length for slitting. such that the scrap during slitting operation is minimum:

Figure 4.13 (Plate 2) shows the slitted rolls of different widths, which are collected at the other end of the slitting operation over a collapsible mandrel.

I-

I,,' _-

I

I '

, I

} ,

\ ' I

(b) Cropping of Laminations

Magnet(c Clr(/I/I 8~

\

i

1 I

The different shapes and sizes of laminations needed for core building are

illustrated in Figs. 4.8-4.11 for different types of transformer cores. Figur~ 4.14 (Plate 2) shows a cropping machine for cutting the laminations front slitted rolls. :

I

In case of mitred core laminations, these are first cut in a trapezoidal

shape. Finished shape is given by cutting off the corners (wherever needed) at the second stage of cutting by employing simple hydraulically operated guillotine machines.

(c) Piercing operation I

, i

The yoke punchings (laminations) usually need holes for bolting the yoke

laminations. These holes are punched after tho cropping/guillotine operation' by suitably adjusting tho- hole piercing positions in the piercing machine and selecting tho right piercing tool for specified hole punch size. Figure 4.15 (Plate 2) shows a piercing machine. However, some cores are also' constructed without yoke bolts.

(d) Deburring

During the process of slitting, cutting and piercing of laminations, the cut edges' get some burrs. These burrs are removed by passing the laminations through deburring operations. Presence of bun's impairs the stacking factor. Also burrs cut into the insulation coatings and bridge adjacent laminations and thereby increase the eddy losses.

(e) Annealing/Varnishing

If the insulation coatings at the 'edges are" scratched during deburring process or extra varnish coating is desired, the laminations are processed in a varnishing plant' which provides a thin coating of varnish and

. .""'!

quickly dried up at elevated temperature. However, varnish coating is

not considered as necessary if the carlite insulation. coating on the lamination is consistent. During the process of slitd'ng, croppingj piercing, deburr.ng, etc., mechanic-al strains are deveioped inside the

- - I

laminations, which disturb the original grain orientation and thereby

increase the iron-losses. This problem is mitigated by annealing the larninations in an annealing plant (Fig. 4.16 (Plate 2~). :

Annealingis done at an elevated temperature of ~OO--9DO°C, preferably

in a neutrai atmospheric zone and' subsequently 'cooled by a blast of air. .

The finished laminations are then taken to the core assembly area.

4.4.3 Core Assembly

Core building from the finished lamination sheets is done in horizontal position on specially raised platforms. The lamination sheets arc susceptiblc to mechanical stresses of bending, twisting, impact) etc. A lot of care is exercised while. handling and normally two persons arc needed to hold the two ends of the laminations at the time of laying. Figure 4.17

86 Transformers

(Plate 2) shows tho core building operations typically for a three-phase five-limbed core.

At first the clamp plates and end frame structure of one side of the core assembly are laic(out. Guide pins arc used at . suitable positions for maintaining the prop~'r alignments during core building process. Oil ducts are formed by' s\icking strips on lamination and put in position as required.

For each packet, the laminations are manufactured in two different lengths and these sets are laid out alternately, keeping at a time t'1,"O to four laminations tog~llicL_ Tj1e two_ill_tcrnate arrangements provide overlapping at the corner joints and when the lamination packets are clamped together, these overlapping edges providesufficient mechanical strength in holding the edges in tight grip. After laying out the completeIarninations, the clamp plates, and end frame structure of the other:stcie are laid out and the entire core-end frame structure. is properly secured through bolts and steel bands at a number of positions.

Theplatform on which the core building takes place is of special design and the core-end frame assembly can be raised to the vertical position alongwith the platform which serves as a cradle. Subsequently the platform is disengaged. In this process, the core assembly is spared from the mechanical strain of lifting and raising in the vertical position. Small-size corescan however be built up without these special platforms.

Steel bands used for tightening the laminations is only a temporary arrangement and are. later removed, otherwise these will form short circuited turns. Two commonly used methods of holding the leg laminations together is their clamping by either (a) resiglass tape or (b) using skin stressed bakelite cylinders. In case of resiglass tapes, these are tightly wound around the legs at specified pitch and cured by heating. The tape shrinks after heating and provides a firm grip. The 'tensile strength of resiglass tapes is even. higher than that of steel tapes. In the case of core legs tightened by skin stressed cylinders (base cylinder of innermost coil), these are.lowered from the top and the steel bands are cut progressively. Wooden. wedges are inserted along the packet corners' and hammered down, so that the enveloping bakelite. cylinder and the leg laminations are fitting tightly against each _ other. Figures 4.18. (Plate 3) and 4.19. show the complete core-end frame assemblies for single-phase three-limbed and threephase five-limbed cores.

Conventionally, the core is assembled along with all the yokes, and after assembly the top yokes are uolaced after removing the top-end frames for the purpose of lowering the. windings. This takes alot.ofIabour andmanu .. facturing time. The latest development is to assemble the core without top yokes and insert the top yokes after lowering all the windings in the core leg.

• f

,

~ .

I

I

\i

\

i

!

i

" "

. --- ·_---t t

i

I

(

!

!

I

J.

4.4.4 Fitting of Core 1" the Tank

~.

i The most commonly used method of putting the core assembly is to rest

the core frame on its f~et, which in turn is firmly fixed on the tank basel The extreme .cnd feet are enclosed in a steel bracket welded on the tan~ base. The top portion r)f core assembly is also suitably locked with the tank cover, so that any possible magnification of the vibration during transit is fully arrested. An alternative to this is construction of the tank base in the shape of a channel. The laminations along with tho bottomend frames are jacked from either side against the channel walls of the tank. As such, no boj~::; .are needed in the bottom yokes and iron-Iosse~ are comparatively lower because of the absence of these punch holesl Figure 4.20 illustrates this channel type of tank and core assembly fixind arrangement. This also has the advantage that the core is rigidly fixed td, the tank bottom and no possibility of core assembly shifting exists. As! a result of raising tho tank base (Item I, Fig.· 4 2.)) the inner oil volume inside the tank is reduced. Also, the windings and the insulation rings and: blocks are directly resting on the flat tank base. \

I I

2

6

6

r

I I

I :

...... =-_--

Fig. 4.20 Transformer core fitting in channel-shaped tank base

1. Channel-shaped tank base

2. End view of transformer core

3. Bottom end frame

4. Jacking of core end frame against channel walls

5. Tank base stiffener

6. Windings

7. Insula {ion blocks and rims

r;;; •.

88 Transformer«

REFERENCES

1. Brechha, H. "Some Asp~cts 0/ Modern Transformer Core Design" Bulletin Ocrlikon, No. 324. pp., 70-78.

2. Brcchna, H. "New Design Trends it Construction 0/ Transformer Cores" Bulletin Oerlikon, No. 326, pp. 6-14.

3. Austen Stignant and Franklin, J&P Transformer Book, Ncwness=-Butterworths, 10th Edition, 1973.

4. Bhusan Prabhakar, "Optimum Design 0/ Distribution Transformers". M. Tech.

Thesis. 1971. Deptt. of Electrical Engg., U.T. Kanpur.

=t.