T H I S P U B L I C AT I O N I S C O P YRI G H T P RO T E C T E D

C o p yri g h t © 2 0 1 8 I E C , G e n e v a , S wi tz e rl a n d

All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form
or by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from
either IEC or IEC's member National Committee in the country of the requester. If you have any questions about I EC
copyright or have an enquiry about obtaining additional rights to this publication, please contact the address below or
your local I EC member National Committee for further information.
IEC Central Office Tel.: +41 22 91 9 02 1 1
3, rue de Varembé info@iec.ch
CH-1 21 1 Geneva 20 www. iec.ch
Switzerland

Ab ou t th e I E C

The I nternational Electrotechnical Commission (I EC) is the leading global organization that prepares and publishes
I nternational Standards for all electrical, electronic and related technologies.
Ab o u t I E C p u b l i ca ti o n s

The technical content of IEC publications is kept under constant review by the IEC. Please make sure that you have the
latest edition, a corrigenda or an amendment might have been published.
I E C Catal og u e - webs tore. i ec. ch /catal og u e E l ectroped i a - www. el ectroped i a. org

The stand-alone application for consulting the entire
bibliographical information on IEC International Standards,
Technical Specifications, Technical Reports and other
documents. Available for PC, Mac OS, Android Tablets and
iPad.
I E C pu bl i cati on s s earch - webs tore. i ec. ch /ad vs earch form
The world's leading online dictionary of electronic and
electrical terms containing 21 000 terms and definitions in
English and French, with equivalent terms in 1 6 additional
languages. Also known as the International Electrotechnical
Vocabulary (IEV) online.
I E C G l os s ary - s td . i ec. ch /g l os s ary

The advanced search enables to find IEC publications by a
variety of criteria (reference number, text, technical
committee,… ). It also gives information on projects, replaced
and withdrawn publications.
I E C J u s t Pu bl i s h ed - webs tore. i ec. ch /j u s tpu bl i s h ed
67 000 electrotechnical terminology entries in English and
French extracted from the Terms and Definitions clause of
IEC publications issued since 2002. Some entries have been
collected from earlier publications of IEC TC 37, 77, 86 and
CISPR.

Stay up to date on all new IEC publications. Just Published I E C C u s tom er S ervi ce C en tre - webs tore. i ec. ch /cs c

details all new publications released. Available online and If you wish to give us your feedback on this publication or
also once a month by email. need further assistance, please contact the Customer Service
Centre: sales@iec.ch.
 –2– I EC 60076-7: 201 7 © I EC 201 7

CONTENTS

FOREWORD . .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... . .. .. ... ... ... ... ... ... ... ... .. 6
I NTRODUCTI ON .. ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. .. . ... ... ... ... ... ... ... ... ... 8
1 Scope .. ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... . 9
2 Norm ative references . ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... . 9
3 Terms and definitions . ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... . 9
4 Symbols and abbreviations . ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... 1 1
5 Effect of loading beyond nam eplate rating . ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... 1 3
5. 1 General . ... ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... ... ... ... ... ... 1 3
5. 2 General consequences ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... . .. ... ... ... ... .. ... ... ... ... ... ... ... . 1 3
5. 3 Effects and hazards of short-time em ergency loading . ... ... ... ... ... ... ... .. ... ... ... ... ... ... . 1 4
5. 4 Effects of long-time emergency loading ... ... ... .. ... ... ... ... ... ... ... ... ... ... .. . ... ... ... .. ... ... ... . 1 5
5. 5 Transformer size ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... .. . ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... .. 1 5
6 Relative ageing rate and transform er insulation life ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... . .. .. 1 5
61 General 15
6. 2 I nsulation life .. ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... .. . ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. 1 6
6. 3 Relative ageing rate ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... .. . ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... 20
6. 4 Loss-of-life calculation ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. .. 21
7 Lim itations ... ... ... ... ... ... .. ... ... ... ... ... ... . .. ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... .. . ... ... ... ... ... ... .. ... ... 21
7. 1 Temperature limitations .. ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... .. . ... ... ... ... .. ... ... ... ... ... ... ... . 21
7. 2 Current limitations . ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... . .. ... .. 22
7. 3 Specific lim itations for small transformers . .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . .. ... ... .. 23
7. 3. 1 Current and tem perature limitations . ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . 23
7. 3. 2 Accessory and other considerations ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... . .. ... ... .. . 23
7. 3. 3 I ndoor transformers .. .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... 23
7. 3. 4 Outdoor ambient conditions ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... . .. .. ... ... ... ... ... ... .. 23
7. 4 Specific lim itations for medium power transform ers ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... 23
7. 4. 1 Current and tem perature limitations . ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . 23
7. 4. 2 Accessory, associated equipment and other considerations . ... ... ... ... ... ... ... ... .. 23
7. 4. 3 Short-circuit withstand requirements ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... 24
7. 4. 4 Voltage limitations . ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . .. ... ... ... ... ... ... ... ... ... ... ... ... 24
7. 5 Specific lim itations for large power transform ers . ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... 24
7. 5. 1 General .. ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... . .. . 24
7. 5. 2 Current and tem perature limitations . ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . 24
7. 5. 3 Accessory, equipment and other considerations ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... 24
7. 5. 4 Short-circuit withstand requirements ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... 25
7. 5. 5 Voltage limitations . ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . .. ... ... ... ... ... ... ... ... ... ... ... ... 25
8 Determination of temperatures . ... ... ... ... ... ... .. ... ... ... ... ... .. . ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... . 2 5
8. 1 Hot-spot tem perature rise in stead y state .. ... ... ... ... ... ... ... ... .. ... ... ... ... ... .. . ... ... ... ... ... . 25
8. 1 . 1 General .. ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... . .. . 25
8. 1 . 2 Calculation of hot-spot tem perature rise from normal heat-run test data ... ... .. 25
8. 1 . 3 Direct measurem ent of hot-spot tem perature rise . ... ... ... ... ... ... ... ... ... ... .. ... ... ... 26
8. 1 . 4 Hot-spot factor .. ... ... ... ... .. ... ... ... ... ... ... ... ... ... . .. ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... .. 29
8. 2 Top-oil and hot-spot temperatures at varying am bient temperature and load
conditions ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... . . ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... . 31
8. 2. 1 General .. ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... . .. . 31
I EC 60076-7: 201 7 © I EC 201 7 –3–

8. 2. 2 Exponential equations solution ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... . .. ... ... ... ... ... .. 33
8. 2. 3 Difference equations solution . ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... .. 37
8. 3 Ambient tem perature.. .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... . .. 39
8. 3. 1 Outdoor air-cooled transformers ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... 39
8. 3. 2 Correction of am bient temperature for transform er enclosure .. ... .. ... ... ... ... ... ... 39
8. 3. 3 Water-cooled transformers . ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... . 40
9 I nfluence of tap-changers ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... . .. ... ... ... ... ... ... ... .. ... ... ... ... ... ... 40
9. 1 General . ... ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... ... ... ... ... ... 40
9. 2 Load loss ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... .. . ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... . 41
9. 3 Ratio of losses ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... . .. ... .. 41
9. 4 Load factor ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... . .. ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... .. . ... . 41
Annex A (informative) I nsulation life expectancy and relative ageing rate considering
oxygen and water effect ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... . .. 42
A. 1 I nsu lation life expectancy ... ... ... ... ... .. ... ... ... ... ... ... ... .. . ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... . 42
A. 2 Relative ageing rate considering oxygen and water effect ... ... ... ... .. ... ... ... ... ... ... ... .. 44
Annex B (informative) Core tem perature .. ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... .. . ... .. ... ... ... ... ... ... ... 47
B. 1 General . ... ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... ... ... ... ... ... 47
B. 2 Core hot-spot locations ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... . 47
Annex C (inform ative) Specification of loading beyond rated power .. ... ... ... ... ... ... ... ... ... ... ... .. 48
Annex D (inform ative) Description of Q, S and H factors .. ... ... ... ... ... ... ... ... ... ... ... ... ... .. . .. ... ... ... 50
Annex E (informative) Calculation of wind ing and oil time constant ... ... ... ... ... ... ... ... ... ... ... ... .. 53
Annex F (inform ative) Thermal m odel parameters . ... ... ... ... .. ... ... ... ... ... ... ... ... . .. ... ... ... ... ... .. ... .. 55
F.1 General . ... ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... ... ... ... ... ... 55
F.2 Thermal constant estim ation: experimental approach .. ... ... ... ... ... ... ... .. ... ... ... ... ... .. . . 55
F.3 Dynamic therm al modelling: further development ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... 57
Annex G (inform ative) Oil and winding exponents ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... . .. ... ... ... .. 58
G.1 General . ... ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... ... ... ... ... ... 58
G.2 Historical background .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... .. . ... ... ... ... ... ... ... ... ... ... ... ... ... 58
G. 3 Theoretical approach . .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... . .. ... ... ... ... ... ... ... ... ... ... ... ... ... 60
G.4 Extend ed tem perature rise test approach .. ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... ... ... ... ... ... . 62
Annex H (inform ative) Practical exam ple of the exponential eq uations m ethod . ... ... ... ... ... ... . 64
H. 1 General . ... ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... ... ... ... ... ... 64
H. 2 Time period 0 min to 1 90 m in .. ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... . .. ... .. ... ... ... ... ... ... ... ... .. 65
H. 3 Time period 1 90 min to 365 min ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... 65
H. 4 Time period 365 min to 500 min ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... 66
H. 5 Time period 500 min to 705 min ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... 66
H. 6 Time period 705 min to 730 min ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... 67
H. 7 Time period 730 min to 745 min ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... 67
H. 8 Comparison with measured values . ... ... ... ... ... ... ... ... ... ... ... ... ... .. .. . ... ... ... ... ... ... ... ... ... 68
Annex I (inform ative) Application of the d ifference equation solution method . ... ... ... ... ... ... ... . 70
I.1 General . ... ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. . ... ... ... ... ... ... 70
I.2 Example... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... . .. .. ... ... ... ... . 70
I.3 Use of m easured top-oil tem perature ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... . .. ... .. ... ... ... . 75
Annex J (inform ative) Flowchart, based on the exam ple in Annex H .. ... ... ... ... ... ... ... ... ... ... ... .. 76
Annex K (informative) Exam ple of calculating an d presenting overload data .. ... ... ... ... ... ... ... . 78
Annex L (informative) Geomagnetic ind uced currents ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. .. . ... ... 82
L. 1 Background ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... . .. ... ... ... ... . 82
 –4– I EC 60076-7: 201 7 © I EC 201 7

L. 2 GI C capability of power transform ers [54], [55] .. ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... . .. ... .. 82
Annex M (inform ative) Alternative oils . ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... . .. ... ... ... ... .. ... ... ... ... ... ... .. 84
Bibliograph y .. .. ... ... ... ... ... ... ... ... ... ... ... . .. ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... .. . ... ... .. ... ... ... ... ... ... ... ... 85

Figure 1 – Structural formula of cellulose .. ... ... ... ... ... ... ... ... ... ... .. ... ... . .. ... ... ... ... ... ... ... ... ... ... ... .. 1 6
Figure 2 – Correlation between tensile strength and DP value . ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... . 1 7
Figure 3 – Accelerated ageing in m ineral oil at 1 40 °C, oxygen and moisture contents
maintained at < 6  000 ppm and 0, 5 %, respecti vel y ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... . .. ... .. ... .. 1 8
Figure 4 – Expected life for non-therm ally upgraded paper and its dependence upon
moisture, oxygen and tem perature . ... ... ... ... ... ... ... ... ... ... ... ... .. . ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... . 1 9
Figure 5 – Expected life for thermall y upgraded paper and its dependence upon
moisture, oxygen and tem perature .. ... ... ... ... ... ... ... ... .. ... ... ... .. . ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... . 2 0
Figure 6 – Thermal diagram . ... ... ... ... .. ... ... ... ... ... ... ... ... ... . .. ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... .. . ... 26
Figure 7 – Tem perature rises above top-oil tem perature (in tank) 65, 8 °C of the zig -zag
cooled H V-winding of a 400 MVA ON AF cooled 3-phase transform er, load current 1 , 0
p. u. , tap position (-) ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... . .. ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... .. 27
Figure 8 – Coil edges, where the sensors should be located in the edge with the higher
calculated temperature rise ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... .. . ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... .. 28
Figure 9 – Tem perature rises above top-oil temperature at the end of an 8 h therm al
no- l oad test at 1 1 0 % su ppl y vol tag e ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... . .. ... ... .. ... ... ... ... ... ... ... ... 29
Figure 1 0 – Zigzag-cooled winding where the distance between all sections is the
sam e and the flow-directing washer is installed in the space between sections .. ... ... ... ... ... ... . 30
Figure 1 1 – Top view section of a rectangular winding with “collapsed cooling duct
arrangem ent” under the yokes .. ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... .. . ... ... ... ... ... ... ... ... .. ... ... ... ... ... . 31
Figure 1 2 – Block diagram representation of the differential equations .. ... ... ... ... .. ... ... ... ... ... ... 32
Figure 1 3 – Temperature responses to step changes in the load current . ... ... ... ... ... ... ... ... ... ... 34
Figure 1 4 – The function ∆ θh ( t)/ ∆ θ hr generated by the values given in Table 4 .. ... ... ... ... ... ... . 37
Figure 1 5 – Principle of losses as a function of the tap position . ... ... ... ... ... ... ... ... ... .. ... ... ... ... .. 41
Figure A. 1 – Arrhenius plot for an ageing process .. ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... . 43
Figure F. 1 – H ot-spot and top-oil overall m odel ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... . .. ... ... ... ... ... .. ... .. 57
Figure G. 1 – Extended tem perature rise test .. ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . 62
Figure G. 2 – Transformer exponent estim ation plots ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... . .. ... .. ... ... ... 63
Figure H .1 – Hot-spot tem perature response to step changes in the load current . ... ... ... ... ... .. 68
Figure H .2 – Top-oil temperature response to step changes in the load current . ... ... ... ... ... ... . 68
Figure I . 1 – Plotted input data for the exam ple .. ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. . 72
Figure I . 2 – Plotted output data for the example .. ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... . 75
Figure K. 1 – OF large power transform ers: perm issible duties for norm al loss of life .. ... ... ... .. 81
Figure L. 1 – GI C flow into a power transformer ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... . 82

Table 1 – Relative ageing rates due to hot-spot tem perature ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... . 21
Table 2 – Maximum permissible tem perature lim its applicable to loading beyond
nameplate rating ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... . .. ... ... ... ... .. ... . 22
Table 3 – Recommended current limits applicable to loading beyond nam eplate rating .. ... ... . 23
Table 4 – Recommended thermal characteristics for exponential equations .. .. ... ... ... ... ... ... ... . 36
Table 5 – Correction for increase in ambient tem perature due to enclosure .. .. ... ... ... ... ... ... ... . 40
Table A. 1 – Activation energy ( E A ) and environment factor ( A ) for oxidation, hydrol ysis . ... ... . 43
I EC 60076-7: 201 7 © I EC 201 7 –5–

Table A. 2 – Expected life of paper u nder various cond itions .. ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... 4 4
Table A. 3 – Relative ageing rates d ue to hot-spot tem perature, oxygen and m oisture
for non-upgraded paper insulation . ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... . 45
Table A. 4 – Relative ageing rates d ue to hot-spot tem perature, oxygen and m oisture
for upgraded paper insulation . ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... 46
Table H . 1 – Load steps of the 250 MVA transformer ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... . . ... ... ... 64
Table H . 2 – Tem peratures at the end of each load step . .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... . .. ... 69
Table I .1 – I npu t data for example .. ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... .. 71
Table I .2 – Output data for the example . ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... 74
Table K. 1 – Example characteristics related to the loadability of transformers ... ... ... ... ... ... ... . 78
Table K. 2 – An example table with the permissible duties and corresponding daily loss
of life (in “norm al” d ays), and m axim um hot-spot temperature rise during the load
cycle ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. .. . ... ... ... ... ... ... ... ... ... ... ... ... 80
 –6– I EC 60076-7: 201 7 © I EC 201 7

INTERNATI ONAL ELECTROTECHNI CAL COMMISSI ON

____________

POWER TRANSFORMERS –
Part 7: Loading guide for mineral-oil-immersed
power transformers
FOREWORD
1 ) The I nternati on al Electrotechni cal Comm ission (I EC) i s a worl d wid e organization for stan dardization com prisin g
all n ation al el ectrotechnical comm ittees (I EC N ational Comm ittees). The object of I EC is to prom ote
internati onal co-operation on all q uestions concerni ng stand ardi zati on in the el ectrical an d electronic fi elds. To
this en d and in addi ti on to other acti vities, I EC pu blish es I nternational Stan dard s, Techn ical Specificati ons,
Technical Reports, Publicl y Avail abl e Specifi cati ons (PAS) an d Gu ides (h ereafter referred to as “I EC
Publication(s)”). Th ei r preparation is entrusted to tech nical comm ittees; any I EC N ational Comm ittee interested
in the subj ect dealt with m ay partici pate in this preparatory work. I nternational, governm ental an d n on -
governm ental organ izations l iaising with th e I EC also participate i n this preparation. I E C collaborates closel y
with the I ntern ational Organi zation for Stand ard ization (I SO) in accordance with con ditions determ ined by
agreem ent between th e two organi zati ons.
2) The form al decisions or ag reem ents of I EC on tech nical m atters express, as n early as possible, an i nternati onal
consensus of opi nion on the rel evant subjects since each technical com m ittee has representati on from all
interested I EC N ational Com m ittees.
3) I EC Publications have the form of recom m endations for intern ational use an d are accepted by I EC N ational
Com m ittees in that sense. While all reasonable efforts are m ade to ensure that th e tech nical content of I EC
Publications is accu rate, I EC cann ot be h eld responsi ble for th e way in which th ey are used or for an y
m i sinterpretation by an y en d u ser.
4) I n ord er to prom ote intern ational u ni form ity, I EC N ational Com m ittees und ertake to apply I EC Publications
transparentl y to the m axim um extent possible i n their national an d regi on al publicati ons. Any d ivergence
between an y I EC Publication and the correspondi ng national or regi on al publicati on sh all be clearl y in dicated in
the latter.
5) I EC itself d oes n ot provi de an y attestation of conform ity. I n depend ent certificati on bodies provi de conform ity
assessm ent services and, in som e areas, access to I EC m arks of conform i ty. I EC is not responsi ble for an y
services carri ed out by ind ependent certification bodi es.
6) All users shou ld ensure that th ey h ave the l atest editi on of thi s publicati on.
7) N o liability shall attach to I EC or its directors, em ployees, servants or ag ents inclu din g in divi du al experts an d
m em bers of its tech nical com m ittees and I EC Nati on al Com m ittees for any person al i nju ry, property d am age or
other dam age of any n ature whatsoever, wheth er di rect or indi rect, or for costs (includ i ng leg al fees) and
expenses arisi ng out of the publ ication, use of, or relian ce upon, this I EC Publicati on or any other I EC
Publications.
8) Attention is drawn to th e N orm ative references cited in th is publ ication. U se of the referenced publ ications is
indispensable for the correct applicati on of this publi cation.
9) Attention is drawn to the possibility that som e of the elem ents of this I EC Publication m ay be the su bject of
patent rig hts. I EC shall not be held responsibl e for identifyi ng any or all such patent ri ghts.

I nternational Standard I EC 60076-7 has been prepared by I EC technical committee 1 4: Power

transformers.

This second edition cancels and replaces the first edition published in 2005. I t constitutes a
technical revision. This edition includes the following significant technical changes with
respect to the previous edition:
a) title has been updated from "oil-imm ersed power transformers" to "m ineral-oil-imm ersed
power transformers";
b) insulation life is updated by considering latest research findings;
c) tem perature limits have been reviewed and m aximum core temperature is recomm ended ;
d) number of fibre optic sensors is recom mended for temperature rise test;
e) Q, S and H factors are considered ;
f) therm al models are revised and rewritten in generally applicable m athematical form ;
I EC 60076-7: 201 7 © I EC 201 7 –7–

g) geom agnetic induced currents are briefl y d iscussed and corresponding tem perature lim its
are suggested;
h) extensive literature review has been perform ed and a num ber of references added to
bibliograph y.
The text of this standard is based on the following documents:
FDI S Report on votin g
1 4/933/FDI S 1 4/942/RVD

Full information on the voting for the approval of this standard can be fou nd in the report on
voting ind icated in the above table.

This publication has been drafted in accordance wi th the I SO/I EC Directives, Part 2.

A list of all parts of the I EC 60076 series, u nder the general title Power transformers , can be
found on the I EC website.

The comm ittee has decided that the contents of this publication will remain unchanged u ntil
the stability date ind icated on the I EC website und er "http: //webstore.iec.ch" in the d ata
related to the specific pu blication. At this date, the publication will be
• reconfirmed,
• withdrawn,
• replaced by a revised ed ition, or
• am ended.

A bilingual version of this pu blication m ay be issued at a later d ate.

I M P O RT AN T – Th e ' col ou r i n si d e' l og o on th e co ve r p ag e o f th i s p u b l i cati o n i n d i c ate s

th at it con tai n s c o l o u rs wh i c h a re c o n s i d e re d to be u s e fu l fo r th e c o rre c t

u n d e rs t a n d i n g of i ts c o n te n ts . U s e rs sh ou l d t h e re fo re p ri n t th i s d o cu m en t u sin g a

c o l o u r p ri n t e r.
 –8– I EC 60076-7: 201 7 © I EC 201 7

INTRODUCTION
This part of I EC 60076 provides guidance for the specification and loading of power
transformers from the point of view of operating temperatures and therm al ageing. I t provides
recom mendations for loading above the nameplate rating and guidance for the planner to
choose appropriate rated quantities and loading conditions for new installations.

I EC 60076-2 is the basis for contractual agreements and it contains the requirem ents and
tests relating to tem perature-rise figures for oil-imm ersed transform ers during continuous
rated loading.

This part of I EC 60076 gives mathem atical models for judging the consequence of different
loadings, with different tem peratures of the cooling m edium , and with transient or cyclical
variation with tim e. The models provide for the calculation of operating temperatures in the
transformer, particularl y the temperature of the hottest part of the winding. This hot-spot
temperature is, in turn, used for evaluation of a relative value for the rate of therm al ageing
and the percentage of life consum ed in a particular tim e period. The modelling refers to sm all
transform ers, here called distribution transformers, and to power transformers.

A m ajor change from the previous edition is the extensive work on the paper degradation that
has been carried out indicating that the ageing may be described by combination of the
oxidation, hydrol ysis and pyrol ysis. Also, providing possibility to estim ate the expected
insulation life considering different ageing factors, i. e. moisture, oxygen and tem perature, and
more realistic service scenarios. The title has been updated from "oil-imm ersed power
transform ers" to "m ineral-oil-imm ersed power transform ers". The tem perature and current
lim its are reviewed and the maxim um core temperature is recom mended. The use of fibre
optic temperature sensors has becom e a standard practice, however, the number of installed
sensors per transform er highl y varies. This issue and the description of Q, S and H factors
are now considered as well. The therm al m odels are revised and rewritten in generall y
applicable m athematical form. The geomagnetic induced currents are briefly discussed and
corresponding temperature limits are suggested.

This part of I EC 60076 further presents recommendations for lim itations of perm issible
loading according to the results of temperature calculations or measurements. These
recomm endations refer to different types of loading duty – continuous loading, norm al cyclic
undisturbed loading or tem porary em ergency loading. The recommendations refer to
distribution transform ers, to medium power transform ers and to large power transformers.
Clauses 1 to 7 contain definitions, comm on background inform ation and specific limitations for
the operation of different categories of transformers.

Clause 8 contains the determination of temperatures, presents the mathem atical models used
to estimate the hot-spot temperature in stead y state and transient conditions.

Clause 9 contains a short description of the influence of the tap position.

Application exam ples are given in Annexes A, B, C, D, E, F, G, H, I and K.

I EC 60076-7: 201 7 © I EC 201 7 –9–

POWER TRANSFORMERS –
Part 7: Loading guide for mineral-oil-immersed
power transformers

1 Scope
This part of I EC 60076 is applicable to m ineral-oil-im mersed transform ers. I t describes the
effect of operation under various am bient temperatures and load conditions on transformer
life.

NOTE For fu rn ace transform ers, the m anufacturer i s consulted in view of the pecul iar l oadi ng profil e.

2 Normative references
The following docum ents are referred to in the text in such a way that som e or all of their
content constitutes requ irements of this docum ent. For d ated references, onl y the edition
cited applies. For und ated references, the latest ed ition of the referenced d ocum ent (includ ing
an y am endm ents) applies.

I EC 60076-2, Power transformers – Part 2: Temperature rise for liquid-immersed transformers

I EC 60076-1 4, Power transformers – Part 14: Liquid-immersed power transformers using

high-temperature insulation materials

3 Terms and definitions

For the purposes of this docum ent, the following terms and d efinitions apply.

3.1
small power transformer
power transformer with out attached rad iators, coolers or tu bes includ ing corrugated tank
irrespective of rating

3.2
medium power transformer
power transformer with a m aximum rating of 1 00 MVA three-phase or 33, 3 M VA single-phase

3.3
large power transformer
power transform er with a m axim um rating of greater than 1 00 MVA three-phase or greater
than 33, 3 MVA single-phase

3.4
cyclic loading
loading with cyclic variations (the duration of the cycle usuall y being 24 h) which is regarded
in terms of the accum ulated amount of ageing that occurs d uring the cycle

Note 1 to entry: The cycli c loadin g m ay eith er be a n orm al l oadi ng or a l ong-tim e em ergen cy loadin g.
 – 10 – I EC 60076-7: 201 7 © I EC 201 7

3. 5
normal cycli c load i n g
loading in which a higher am bient tem perature or a higher-than-rated load current is applied
during part of the cycle, but which, from the point of view of relative thermal ageing rate
(according to the m athem atical model), is equivalent to the rated load at norm al ambient
temperature

Note 1 to entry: This is achieved by takin g ad vantag e of low am bient tem peratures or low l oad currents duri ng th e
rest of the load cycle. For plannin g pu rposes, this pri ncipl e can be exten ded to provid e for long peri ods of tim e
whereby cycles with relati ve therm al ageing rates g reater th an unity are com pensated for by cycles with th erm al
agei ng rates less than unity.

3. 6
lon g-tim e emerg en cy loadi n g
loading resulting from the prolonged outage of som e system elements that will not be
reconnected before the transform er reaches a new and higher stead y-state tem perature

3. 7
sh ort-tim e em erg en cy l oad in g
unusuall y heavy loading of a transient nature (less than 30 m in) due to the occurrence of one
or m ore unlikel y events which seriousl y disturb normal system loading

3. 8
hot-spot
if not specially defined, hottest spot of the windings

3. 9
rel ative th erm al ag ei n g rate
for a given hot-spot tem perature, rate at which transform er insulation ageing is reduced or
accelerated compared with the ageing rate at a reference hot-spot temperature

3. 1 0
tran sform er i n su l ati on l i fe
total time between the initial state for which the insulation is considered new and the final
state for which the insulation is considered deteriorated due to thermal ageing, dielectric
stress, short-circuit stress, or mechanical movement (which could occur in normal service),
and at which a high risk of electrical failure exists

3. 1 1
per cen t l oss of l i fe
equivalent ageing in hours over a time period (usuall y 24 h) times 1 00 divided by the
expected transform er insulation life

Note 1 to entry: The eq ui val ent agein g i n hours is obtained by m ultiplyi ng the relati ve ag ei ng rate with th e num ber
of hou rs.

3. 1 2
n on -th erm al l y u pgrad ed paper
kraft paper produced from unbleached softwood pulp under the sulphate process without
addition of stabilizers

3. 1 3
th erm al l y u pg rad ed paper
cellulose-based paper which has been chemically modified to reduce the rate at which the
paper decomposes

Note 1 to entry: Agein g effects are red uced either by parti al el im ination of water form ing agents (as in
cyan oeth yl ati on ) or by i nhi bitin g the form ation of water throu gh th e use of stabili zi ng agents (as in am ine ad dition,
I EC 60076-7: 201 7 © I EC 201 7 – 11 –

dicyandi am ide). A paper is consid ered as th erm all y u pgrad ed if it m eets th e life criteri a d efin ed in
AN SI /I EEE C57. 1 00 [1 ] 1 ; 50 % reten ti on i n ten si l e stren g th after 65 000 h in a sealed tube at 1 1 0 °C or an y oth er
tim e/tem peratu re com binati on given by th e eq uation:

 1 5 000   1 5 000 1 5 000 

 − 28,082   − 

Time (h) = e  ( θh + 273 ) 

≈ 65 000 × e  ( θh + 273 ) (110 + 273 ) 
(1 )

Because the th erm al upgradi ng chem icals used today con tain nitrog en, which is not present in kraft pulp, th e
deg ree of chem ical m odification is determ ined by testing for the am ount of nitrogen presen t in the treated paper.
Typi ca l val u es for n i trog e n con ten t of th erm al l y u pg rad ed papers are between 1 % an d 4 % wh en m easu red i n
accord ance with ASTM D-982 [2], but after the sealed tube test.

3. 1 4
n on -d i rected oil fl ow
OF
flow indicating that the pum ped oil from heat exchangers or radiators flows freel y inside the
tank, and is not forced to flow through the windings

Note 1 to entry: The oil fl ow i nside th e win din gs can be eith er axi al in vertical cooli ng d ucts or radial i n hori zontal
coolin g ducts with or without zi gzag flow.

3. 1 5
n on -d i rected oil fl ow
ON
flow indicating that the oil from the heat exchangers or radiators flows freel y inside the tank
and is not forced to flow through the windings

Note 1 to entry: The oil fl ow i nside th e win din gs can be eith er axi al in vertical cooli ng d ucts or radial i n hori zontal
coolin g ducts with or without zi gzag flow.

3. 1 6
directed oi l fl ow
OD
flow indicating that the principal part of the pumped oil from heat exchangers or radiators is
forced to flow through the windings

Note 1 to entry: Th e oil flow i nside the win din gs can be eith er axi al i n vertical coolin g d ucts or zig zag i n h ori zontal
coolin g ducts.

3. 1 7
desig n ambi en t temperatu re
temperature at which the perm issible average winding and top-oil and hot-spot tem perature
over am bient temperature are defined

4 Sym bol s and abbrevi ati ons

S ym bol M ean i n g U n i ts
C Therm al capacity Ws/K
c Specific heat Ws/(kg· K)
DP Deg ree of polym eri zati on
D Difference operator, in differen ce equ ations
gr Averag e-wi n din g-to-average-oi l (in tank) tem perature g rad ient at rated current K
H H ot-spot factor
k1 1 Therm al m odel constant
k21 Therm al m odel constant

______________
1 N um bers in squ are brackets refer to th e bi bliography.
 – 12 – I EC 60076-7: 201 7 © I EC 201 7

Sym bol M eani ng Uni ts

k22 Therm al m odel constant
K Load factor (load current/rated current)
L Total agei ng over the tim e peri od consid ered h
mA Mass of core an d coil assem bly kg
mT Mass of the tank an d fittings kg
mO Mass of oil kg
mW Mass of wi ndi ng kg
n Num ber of each tim e interval
N Total n um ber of interval s d uri n g the tim e peri od consi dered
OD Either ODAN , ODAF or ODWF coolin g
OF Either OFAN , OFAF or OFWF coolin g
ON Either ONAN or ONAF coolin g
P Suppli ed l osses W
Pe Relati ve win din g ed d y loss p. u.
PW Winding losses W
R Ratio of load losses at rated cu rrent to n o-l oad l osses at rated voltage
Rr Ratio of load losses to no-load loss at princi pal tappi ng
R r+ 1 Ratio of load losses to no-load loss at tappi ng r+1
R mi n Ratio of l oad losses to no-load l oss at m inim um tappin g
R ma x Ratio of load losses to no-load loss at m axim um tapping
RTD Resistance Tem peratu re Detector
RH Oil rel ati ve h um i dity %
s Lapl ace operator
t Tim e vari able m in
tap r Principal tappin g positi on
tap r+1 Tappi ng posi tion r+1
tap mi n Min im um tappin g positi on
tap ma x Maxim um tapping position
V Relati ve ag ein g rate
Vn Relati ve ag ei n g rate du rin g interval n
WOP Water content of oil ppm
WCP Water content of paper insul ati on %
x Exponential power of total losses versus top-oil (in tank) tem perature rise (oil expon ent)
y Exponential power of cu rrent versus wi ndi ng tem peratu re rise (win din g exponent)
θa Am bient tem perature °C
θE Yearl y weig hted am bient tem perature °C
θh Winding hot-spot tem perature °C
θ ma Monthl y averag e tem perature °C
θ ma - ma x Monthl y averag e tem perature of the hottest m onth, accord ing to I EC 60076-2 °C
θo Top-oil tem peratu re (in th e tan k) at the load consi dered °C
θ ya Yearl y average tem peratu re, accordin g to I EC 60076-2 °C
τo Oil tim e constant m in
τW Winding tim e constant m in
∆ θ br Bottom oil (in tank) tem peratu re rise at rated load (no-load losses + load losses) K
I EC 60076-7: 201 7 © I EC 201 7 – 13 –

Symbol M eani ng Uni ts

∆θ h Hot-spot-to-top-oil (i n tank) gradient at the load consi dered K
∆θ h i H ot-spot-to-top-oil (i n tank) gradient at start K
∆θ h r H ot-spot-to-top-oil (i n tank) gradient at rated current K
∆θ o Top-oil (in tank) tem perature ri se at the load consi dered K
∆ θ oi Top-oil (in tank) tem perature ri se at start K
∆θ o m Average oil (in tank) tem peratu re rise at the l oad considered K
∆ θ o mr Averag e oil (in tank) tem peratu re rise at rated l oad (n o-l oad losses + load losses) K
∆ θ or Top-oil (in tank) tem perature ri se in stead y state at rated l osses (no-load l osses + load K
losses)
∆θ or
′ Corrected top-oi l tem perature rise (in tank) d ue to enclosu re K

∆ ( ∆ θ or ) Extra top-oil tem peratu re rise (in tank) due to enclosure K

5 Effect of loading beyond nameplate rating

5.1 General
The norm al life expectancy is a conventional reference basis for continuou s duty under design
ambient temperature and rated operating cond itions. The application of a load in excess of
nameplate rating and /or an am bient tem perature higher than design ambient tem perature
involves a degree of risk and accelerated ageing. I t is the purpose of this part of I EC 60076 to
identify such risks and to indicate how, within lim itations, transform ers m ay be loaded in
excess of the nameplate rating. These risks can be red uced by the purchaser clearl y
specifying the maxim um loading cond itions and the supplier taking these into account in the
transformer design.

5.2 General consequen ces

The conseq uences of loading a transform er beyond its nam eplate rating are as follows.
a) The tem peratures of windings, cleats, leads, insulation and oil will increase and can reach
unacceptable levels.
b) The leakage flux density outside the core increases, cau sing additional ed d y-current
heating in m etallic parts linked by the leakage flux.
c) As the tem perature changes, the moisture and gas content in the insulation and in the oil
will change.
d) Bushings, tap-changers, cable-end connections and current transformers will also be
exposed to higher stresses which encroach u pon their d esign and application margins.
The combination of the m ain flux and increased leakage flux imposes restrictions on possible
core overexcitation [6], [7], [8].

NOTE For l oaded core-type transform ers havin g an en ergy flow from the outer wi ndi ng (u sually H V) to the i nner
wind ing (usuall y LV), the m axi m um m agnetic flu x densi ty i n the core, which is th e result of the com bination of the
m ain flu x an d the l eakage flu x, appears i n the yokes.
As tests have in dicated, this flu x is less than or eq ual to th e flu x g en erated by the sam e appl ied voltage on the
term inal s of the outer wind ing at no-l oad of the transform er. The m agnetic flu x in th e core l egs of the l oad ed
transform er is determ ined by the voltage on th e term inals of the in ner wi ndi ng and alm ost equals the fl u x generated
by th e sam e vol tage at n o-load .
For core-type transform ers with an energ y flow from the in ner win din g, the m axim um flux d ensity is present i n the
core-legs. I ts valu e is only sli g htly hig her than th at at th e sam e applied voltage un der no-l oad. The fl u x density i n
the yokes is then d eterm ined by the voltage on th e outer wind ing.
 – 14 – I EC 60076-7: 201 7 © I EC 201 7

Voltages on both sides of the load ed transform er, therefore, are observed du rin g loadin g beyond th e nam epl ate
ratin g. As lon g as voltages at the energ ized sid e of a l oad ed transform er rem ai n bel ow th e lim its stated in
I EC 60076-1 : 201 1 [5], Clause 4, no excitati on restrictions are needed du ri ng the l oadin g beyond nam eplate rati ng.
When higher excitati ons occu r to keep the loaded voltage in em ergency conditi ons in an area wh ere th e n etwork
can still be kept uprig ht, then the m agnetic flu x d ensities i n core parts never exceed val ues wh ere straying of th e
core flu x outside th e core can occur (for col d-roll ed g rai n-ori ented steel these satu rati on effects start rapi dl y above
1 , 9 T). Stray flu xes m ay cause unpredictabl y hi gh tem peratures at th e core su rface and i n nearby m etallic parts
such as wind ing clam ps or even in th e wi ndi ngs, du e to the presence of hi gh -frequ ency com ponents in th e stray
flu x. Th ey m ay j eopard i ze th e transform er. I n general, in all cases, the short overload tim es dictated by wi ndi ngs
are sufficientl y short n ot to overheat the core at overexcitati on. This is prevented by the l on g therm al tim e constant
of the core.

As a consequence, there will be a risk of premature failure associated with the increased
currents and temperatures. This risk m ay be of an immediate short-term character or come
from the cum ulative effect of therm al ageing of the insulation in the transform er over m an y
years.

5.3 Effects and hazards of short-time emergency loading

Short-tim e increased loading will result in a service condition having an increased risk of
failure. Short-time em ergency overloading causes the conductor hot-spot to reach a level
likel y to result in a temporary reduction in the dielectric strength. However, acceptance of this
condition for a short tim e m ay be preferable to loss of suppl y. This type of loading is expected
to occur rarel y, and it should be rapidl y reduced or the transformer disconnected within a
short tim e in order to avoid its failure. The permissible duration of this load is shorter than the
thermal tim e constant of the whole transformer and depends on the operating tem perature
before the increase in loading; typicall y, it would be less than half-an-hour.

The m ain risk for short-time failures is the reduction in dielectric strength due to the possible
presence of gas bubbles in a region of high electrical stress, that is the windings and leads.
These bubbles are likely to occur when the hot-spot temperature exceeds 1 40 °C for a
transform er with a wi nd i n g i nsu l ati on m oistu re con ten t of abou t 2 % . This critical tem peratu re
will decrease as the m oisture concentration increases.

NOTE Concernin g the bubble gen erati on, see also I EC 60076-1 4.
a) Gas bubbles can also develop (either in oil or in solid insulation) at the surfaces of heavy
metallic parts heated by the leakage flux or be produced by super-saturation of the oil.
However, such bubbles usuall y develop in regions of low electric stress and have to
circulate in regions where the stress is higher before an y significant reduction in the
dielectric strength occurs.
Bare metallic parts, except windings, which are not in direct therm al contact with cellulosic
insulation but are in contact with non -cellulosic insulation (for example, aramid paper,
glass fibre) and the oil in the transform er, may rapidl y rise to high tem peratures. A
tem perature of 1 80 °C should not be exceeded.
b) Tem porary deterioration of the m echanical properties at higher temperatures could reduce
the short-circuit strength.
c) Pressure build-up in the bushings m ay result in a failure due to oil leakage. Gassing in
condenser type bushings m ay also occur if the temperature of the insulation exceeds
about 1 40 °C.
d) The expansion of the oil could cause overflow of the oil in the conservator.
e) Breaking of excessivel y high currents in the tap-changer could be hazardous.
The limitations on the m axim um hot-spot temperatures in windings, core and structural parts
are based on considerations of short-term risks (see Clause 7).

The short-term risks norm all y disappear after the load is reduced to normal level, but they
need to be clearl y identified and accepted by all parties involved , e.g. planners, asset owners
and operators.
I EC 60076-7: 201 7 © I EC 201 7 – 15 –

5.4 Effects of lon g-time emergency loading

This is not a normal operating cond ition and its occurrence is expected to be rare but it m ay
persist for weeks or even months and can lead to consid erable ageing.
a) Deterioration of the mechanical properties of the cond uctor insulation will accelerate at
higher tem peratures. I f this deterioration proceeds far enough, it may redu ce the effective
life of the transform er, particu larl y if the latter is subj ected to system short circu its or
transportation events.
b) Other insulation parts, especiall y parts sustaining the axial pressure of the winding block,
cou ld also suffer increased ageing rates at higher tem perature.
c) The contact resistance of the tap-changers cou ld increase at elevated currents and
tem peratures and, in severe cases, therm al runaway could take place.
d) The gasket materials in the transformer may become m ore brittle as a resu lt of elevated
tem peratures.
The calcu lation rules for the relative ageing rate and per cent loss of life are based on
considerations of long-term risks.

5.5 Transformer size

The sensitivity of transformers to loading beyond nameplate rating usually d epends on their
size. As the size increases, the tendency is that:
• the leakage flux density increases;
• the short-circuit forces increase;
• the mass of insulation, which is subj ected to a high electric stress, is increased;
• the hot-spot tem peratures are m ore d ifficu lt to d etermine.
Thus, a large transformer could be m ore vulnerable to loading beyond nam eplate rating than
a smaller one. I n addition, the consequences of a transform er failure are more severe for
larger sizes than for sm aller units.

Therefore, in order to appl y a reasonable d egree of risk for the expected duties, this part of
IEC 60076 considers three categories:
a) small transform ers, for which onl y the hot-spot temperatures in the windings and therm al
deterioration should be considered ;
b) m edium power transformers where the variations in the cooling m odes should be
considered ;
c) large power transformers, where also the effects of stray leakage flux are significant and
the consequences of failure are severe.
For herm eticall y sealed transform ers without pressure relief devices the over pressure should
be considered to avoid permanent tank deformation during load ing beyond nam eplate rating.

6 Relative ageing rate and transformer insulation life

6.1 General
For the manufacture of paper and pressboard for electrical insu lation, m ainl y unbleached
softwood kraft pulp is used . The cellulose is refined from the tree by the so-called “sulphate”
or “kraft” process. After processing, the typical composition of unbleached kraft pulp is 78 %
to 80 % cel l u l ose, 1 0 % to 20 % h em i cel l u l ose an d 2 % to 6 % l ig n i n .

Cellulose is a linear condensation pol ym er consisting of anh ydroglucose joined together by

gl ycosid ic bonds, Figure 1 .
I EC 60076-7: 201 7 © I EC 201 7 – 17 –

1 20

Tensile i ndex (Nm/g )

80

40

0
1 250 1 000 750 500 250 0
DP-valu e
IEC

Figure 2 – Correlation between tensile strength and DP value

The degree of pol ym erization (DP) is the average num ber ( n ) of gl ycosidic rings in a cellulose
macromolecule, which ranges between 1 1 00 and 1 400 for unbleached soft wood kraft before
   

processing. Depending on the transformer drying process, the DP valu e may be red uced
further to a lesser or higher degree. During ageing, the lengths of these pol ym eric cellulose
molecules are reduced du e to breakage of the covalent bonds between the anh ydrous- - β

glucose m onom ers. The change of DP over time of non-thermall y and thermall y upgraded
paper exposed to a temperature of 1 40 °C, oxygen of < 6 000 ppm and water of 0, 5 is
shown in Figure 3 (see [1 5]). The nitrogen content of the thermall y u pgraded paper used in
  %

this experim ent was 1 , 8 % .

When the DP is reduced to 200 or ned tensile strength, the q uality of the paper
(i. e. the m echanical strength) is normall y considered so poor that this defines the “end of life”
% 3 5 % r e t a i

for such insu lating m aterial (see [1 1 ]), althou gh the insulating material dielectric strength may
be still at an acceptable level.

Annex A gives further elaboration of the paper ageing theory providing a m athematical
methodolog y for estim ation of the expected insulation life considering different ageing factors
such as moisture, oxygen and temperature. The corresponding results for the non-thermall y
and thermall y upgraded paper are presented in Figure 4 and Figure 5, respectivel y.

The illustrated d ifference in thermal ageing behaviour has been taken into accou nt in
industrial stand ards as follows:
• The relative ageing rate V = 1 , 0 corresponds to a tem perature of 98 °C for non-therm all y
upgraded paper and 1 1 0 °C for therm ally upgraded paper.

NOTE 1 Disagreem ent between laboratory tests could com e from testing procedu res. I t i s difficult to reprod uce
the sam e agein g process with accelerated ag ein g often at quite el evated tem peratu res com pared to service
conditions. The valu es gi ven in Table A. 2, Figu re 4 and Figu re 5 are considered as un confirm ed and can be
disputable. H owever, th e num bers gi ve a user th e possibility to sim ulate different agei ng scenarios.
 – 18 – I EC 60076-7: 201 7 © I EC 201 7

1 200

1 000

800
DP

600

400

200

0
0 500 1 000 1 500 2 000 2 500 3 000 3 500 4 000 4 500

IEC

Key
DP deg ree of polym eri zati on t tim e (h)
Δ val ues for therm all y upgrad ed paper ● val ues for non -th erm ally upgraded paper
Figure 3 – Accelerated ageing in mineral oil at 1 40 °C, oxygen and moisture contents
maintained at 6 000 ppm and 0,5 %, respectively
 
I EC 60076-7: 201 7 © I EC 201 7 – 19 –

1 000

1 00

10
Expected life

0, 1

0, 01
50 60 70 80 90 1 00 110 1 20 1 30 1 40
θh
IEC

Key
expected life (years) θh hot-spot tem perature (°C) □ 0, 5 % moisture, low oxygen
∆ 1 ,5 % moisture, low oxygen ◊ 3, 5 % moisture, low oxygen × 0, 5 % m oi sture, high oxyg en

Figure 4 – Expected life for n on-thermally upgraded paper an d its dependen ce

upon moisture, oxygen and temperature
 – 20 – I EC 60076-7: 201 7 © I EC 201 7

1 000

1 00
Expected life

10

0, 1
50 60 70 80 90 1 00 110 1 20 1 30 1 40
θh
IEC

Key
expected life (years) θh hot-spot tem perature (°C) □ 0, 5 % moisture, low oxygen
∆ 1 ,5 % moisture, low oxygen ◊ 3, 5 % moisture, low oxygen × 0, 5 % moisture, high oxygen

Figu re 5 – Expected life for thermally u pgraded paper and its dependence
upon moisture, oxygen and temperature

NOTE 2 Figu re 4 and Fig ure 5 indicate expected life values that are based on resid ual DP valu e of 200, and th at
are d eri ved u nder the l aboratory controlled conditi on as gi ven i n text above, (e. g. constant m oisture content,
constant an d hom ogenous tem perature, etc. ). H owever, to eval uate th e expected life of a transform er th e real
service conditi ons are considered (e. g. loadin g history an d prediction, am bient tem peratu re, insulation m aterial and
insulati on m oisture contam ination ). Th e m oisture contam ination estim ate is usuall y based on the correspondi ng
equi librium curves for m oisture partiti on between oil and paper, (e. g. WCO vs WCP or RH vs WCP).

6.3 Relative ageing rate

Although ageing or deterioration of insulation is a tim e function of temperature, moisture
content, oxygen content and acid content, the model presented in this document is based onl y
on the insulation tem perature as the controlling parameter.

An example of how all ageing param eters can be taken into account is given in Annex A.

Since the temperature distribution is not uniform, the part that is operating at the highest
temperature will norm ally undergo the greatest deterioration. Therefore, the rate of ageing is
referred to the winding hot-spot tem perature. I n this case, the relative ageing rate V is defined
according to Equation (2) for non-therm ally upgraded paper and to Equation (3) for therm ally
upgraded paper (see [27]).

V = 2 (θh − 98 ) / 6 (2)

 1 5 000 1 5 000 
 − 
1 1 0 + 273 θh + 273 
V = e (3)
I EC 60076-7: 201 7 © I EC 201 7 – 21 –

where
θh is the hot-spot temperature in °C.
Equations (2) and (3) imply that V is very sensitive to the hot-spot temperature as can be
seen in Table 1 .

Table 1 – Rel ative ag ei n g rates d u e to h ot-spot tem peratu re

θh N on -u pg rad ed paper i n su l ati on U pg rad ed paper i n su l ati on

°C V V

80 0, 1 25 0, 036
86 0, 25 0, 073
92 0, 5 0, 1 45
98 1 ,0 0, 282
1 04 2, 0 0, 536
110 4, 0 1 ,0
116 8, 0 1 , 83
1 22 1 6, 0 3, 29
1 28 32, 0 5, 8
1 34 64, 0 1 0, 1
1 40 1 28, 0 1 7, 2
The indicated relati ve agei ng rate V = 1 , 0 corresponds to a tem perature of 98 ° C for non-therm ally upgraded
paper and 1 1 0 ° C for th erm all y upg rad ed paper.

6. 4 Loss-of-li fe cal cu l ati on

The loss of life L over a certain period of time is equal to

t 2 N

L = ∫ V dt or L ≈ ∑ Vn × tn (4)
t 1 n =1

where
V is the
n
relative ageing rate d uring interval n , according to Eq uation (2) or (3);
t
n
is the n th tim e interval;

n is the number of each tim e interval;

N is the total num ber of intervals d uring the period consid ered .

The maximum tim e interval should be less than half the smallest time constant, τ w , in
Equation (4) for an accurate solution .

7 Li mi tati ons

7. 1 Tem peratu re l i m itation s

With loading valu es beyond the nam eplate rating, none of the ind ividual limits stated in
Table 2 should be exceeded and account should be taken of the specific lim itations given in
7. 3 to 7. 5.
 – 22 – I EC 60076-7: 201 7 © I EC 201 7

The limits given in Table 2 are applicable to transform ers specified to have temperature rise
requirem ents according to I EC 60076-2. For transform ers specified according to
IEC 60076-1 4, with a higher therm al class insulation materials, the lim its given in
IEC 60076-1 4 appl y.

Table 2 – Maximum permissible temperature limits applicable to loading

beyond nameplate rating
Types of loading Small Large and
tran sformers medium power
tran sformers
Normal cyclic loading
Winding hot-spot tem perature and m etallic parts in contact wi th 1 20 1 20
cellulosic insul ation m aterial (° C)
Other m etallic h ot-spot tem perature (i n contact with oil, aram id 1 40 1 40
paper, gl ass fibre m aterials) (° C)
I nner core hot-spot tem peratu re (°C) 1 30 1 30
Top-oil tem peratu re, in tank (° C) 1 05 1 05
Long-time emergen cy loadin g
Winding hot-spot tem perature and m etallic parts in contact wi th 1 40 1 40
cellulosic insul ation m aterial (° C)
Other m etallic h ot-spot tem perature (i n contact with oil, aram id 1 60 1 60
paper, gl ass-fibre m ateri als) (° C)
I nner core hot-spot tem peratu re (°C) 1 40 1 40
Top-oil tem peratu re, in tank (° C) 115 115
Short-time emerg ency loadin g
Winding hot-spot tem perature and m etallic parts in contact wi th See 7. 3. 1 1 60
cellulosic insul ation m aterial (° C)
Other m etallic h ot-spot tem perature (i n contact with oil, aram id See 7. 3. 1 1 80
paper, gl ass fibre m aterials) (° C)
I nner core hot-spot tem peratu re (°C) See 7. 3. 1 1 60
Top-oil tem peratu re, in tank (° C) See 7. 3. 1 115
N OTE For m ore i nform ation on the core tem perature, see An nex B.

7.2 Current limitations

There are limitations on current carrying capability of transform er other than temperature
limits given in Table 2, and these are described in 7. 3 to 7. 5. Therefore, it is recommended
that the current lim its given in Table 3 are not exceeded even if the circumstances of the
overload m ean that the temperatures in Table 2 are not exceeded. Specific examples would
be in cases of low am bient tem perature, low levels of preload or high therm al capacity of the
winding. The purchaser can specify higher current limits if required, but it should be
recognized that this could lead to a special transform er design. The recomm ended current
limits given in Table 3 should not appl y to very short duration overloads, i. e. less than 1 0 s.

NOTE 1 The breakin g capaci ty of tap-chang ers is li m ited to twi ce the rated current accord in g to I EC 6021 4-1 [3].
I EC 60076-7: 201 7 © I EC 201 7 – 23 –

Tabl e 3 – Re c o m m e n d e d c u rre n t l i m i ts ap p l i ca b l e to l o ad i n g

b e yo n d n am ep l ate ra t i n g

T yp e s o f l oad i n g S m al l M ed i u m p owe r L a rg e p o wer

t ra n s fo rm e rs t ra n s fo rm e rs t ra n s fo rm e rs

N o rm a l c yc l i c l oad i n g

Current (p. u. ) 1 ,5 1 ,5 1 ,3
L o n g -ti m e e m e rg e n c y l oad i n g

Current (p. u. ) 1 ,8 1 ,5 1 ,3
S h o rt - t i m e e m e rg e n c y l oad i n g

Current (p. u. ) 2, 0 1 ,8 1 ,5

NOTE 2 For specification beyond rated power, see Annex C.

7. 3 S p e c i fi c l i m i ta ti o n s fo r s m a l l t ra n s fo rm e rs

7. 3. 1 C u rre n t and t e m p e ra t u re l i m i tati o n s

The limits on load current, hot-spot tem perature, top-oil tem perature and temperature of
metallic parts other than windings and leads stated in Table 2 and Table 3 should not be
exceeded. N o limit is set for the top-oil, core and winding hot-spot tem perature under
short-tim e emergency load ing for d istribution transform ers because it is u suall y im practicable
to control the duration of emergency load ing in this case. I t should be noted that when the
hot-spot temperature exceeds 1 40 °C, gas bubbles m ay develop which cou ld j eopardize the
dielectric strength of the transform er (see 5. 3).

7. 3. 2 Ac c e s s o ry an d o t h e r c o n s i d e ra t i o n s

Apart from the wind ings, other parts of the transformer, such as bushings, cable-end
connections, tap-changing devices and leads may restrict the operation when loaded above
1 , 5 tim es the rated current. Oil expansion and oil pressure cou ld also impose restrictions.

7. 3. 3 I n d o o r t ra n s fo rm e rs

When transform ers are used ind oors, a correction should be mad e to the rated top-oil
tem perature rise to take account of the enclosure. Preferably, this extra tem perature rise will
be determined by a test (see 8. 3. 2).

7. 3. 4 O u td o o r a m b i e n t con d i ti on s

Wind, sunshine and rain may affect the loading capacity of d istribution transform ers, bu t their
unpred ictable nature makes it im practicable to take these factors into account.

7. 4 S p e c i fi c l i m i ta ti o n s fo r m e d i u m p o w e r t ra n s fo rm e rs

7. 4. 1 C u rre n t an d t e m p e ra t u re l i m i ta ti o n s

The load current, hot-spot temperature, top-oil temperature and temperature of m etallic parts
other than wind ings and leads should not exceed the limits stated in Table 2 and Table 3.
Moreover, it should be noted that, when the hot-spot tem perature exceeds 1 40 °C, gas
bubbles may develop which could jeopardize the dielectric strength of the transformer
(see 5. 3).

7. 4. 2 Ac c e s s o ry , a s s o ci ated eq u i pm en t an d o t h e r c o n s i d e ra t i o n s

Apart from the wind ings, other parts of the transformer, such as bushings, cable-end
connections, tap-changing d evices and leads, may restrict the operation when loaded above
 – 24 – I EC 60076-7: 201 7 © I EC 201 7

1 , 5 tim es the rated current. Oil expansion and oil pressure could also impose restrictions.
Consideration may also have to be given to associated equipment such as cables, circuit
breakers, current transformers, etc.

7.4.3 Short-circuit withstand requirements

During or directl y after operation at load beyon d nameplate rating, transformers can not
conform to the therm al short-circuit requirem ents, as specified in I EC 60076-5 [67], which are
based on a short-circuit duration of 2 s. However, the duration of short-circuit currents in
service is shorter than 2 s in most cases.

7.4.4 Voltage limitations

Unless other limitations for variable flux voltage variations are known (see I EC 60076-1 ), the
applied voltage should not exceed 1 , 05 tim es either the rated voltage (principal tapping) or
the tapping voltage (other tappings) on an y winding of the transformer.

7.5 Specific limitations for large power transformers

7.5.1 General
For large power transform ers, additional lim itations, m ainl y associated with the leakage flux,
should be taken into consideration. I t is therefore advisable in this case to specify, at the time
of enquiry or order, the amount of loading capability needed in specific applications.

As far as thermal deterioration of insulation is concerned, the same calculation method

applies to all transform ers.

According to present knowledge, the importance of the high reliability of large units in view of
the consequences of failure, together with the following considerations, m ake it advisable to
adopt a m ore conservative, more individual approach here than for smaller units.
• The combination of leakage flux and main flux in the limbs or yokes of the magnetic circuit
(see 5. 2) m akes large transform ers more vulnerable to overexcitation than sm aller
transform ers, especiall y when loaded above nam eplate rating. I ncreased leakage flux m ay
also cause additional edd y-current heating of other m etallic parts.
• The consequences of degradation of the m echanical properties of insulation as a function
of temperature and time, including wear due to thermal expansion, may be m ore severe
for large transform ers than for smaller ones.
• Hot-spot tem peratures outside the windings cannot be obtained from a normal
temperature-rise test. Even if such a test at a rated current indicates no abnormalities, it is
not possible to draw an y conclusions for higher currents since this extrapolation m ay not
have been taken into account at the design stage.
• Calculation of the winding hot-spot temperature rise at higher than rated currents, based
on the results of a temperature-rise test at rated current, m ay be less reliable for large
units than for sm aller ones.

7.5.2 Current and temperature limitations

The load current, hot-spot temperature, top-oil tem perature and temperature of m etallic parts
other than windings and leads but nevertheless in contact with solid insulating material should
not exceed the lim its stated in Table 2 and Table 3. Moreover, it should be noted that, when
the hot-spot temperature exceeds 1 40 °C, gas bubbles m ay develop which could jeopardize
the dielectric strength of the transformer (see 5.3).

7.5.3 Accessory, equipment and other considerations

Refer to 7. 4. 2.
I EC 60076-7: 201 7 © I EC 201 7 – 25 –

7.5.4 Short-circuit withstand requirements

Refer to 7.4. 3.

7.5.5 Voltage limitations

Refer to 7.4. 4.

8 Determination of temperatures
8.1 Hot-spot temperature rise in steady state
8.1 .1 General
To be strictl y accurate, the hot-spot temperature should be referred to the adjacent oil
temperature. This is assum ed to be the top-oil tem perature inside the winding. Measurem ents
have shown that the top-oil temperature inside a winding m ight be, d ependent on the cooling,
up to 1 5 K higher than the mixed top-oil tem perature insid e the tank.

For m ost transformers in service, the top-oil temperature inside a winding is not precisel y
known. On the other hand, for most of these u nits, the top-oil tem peratu re at the top of the
tank is well known, either by measu rem ent or by calcu lation.

The calculation rules in this docum ent are based on the following:
• ∆ θ or , the top-oil temperature rise in the tank above ambient tem perature at rated losses [ K ] ;
• ∆ θ hr , the hot-spot tem perature rise above top-oil tem perature in the tank at rated current [ K ] .
The param eter ∆ θ hr can be d efined either by d irect measurement d uring a heat-run test or by
a calculation m ethod validated by direct measurem ents.

NOTE The m ethods, princi pl es and calculati on proced ures g iven in 8. 1 . 2, 8. 1 . 3, 8. 1 . 4 and Annex D are u ltim ately
vali d for the con verter transform ers for H VDC appl ication, however, with the necessary con sideration of th e effect
of harm onics on th e transform er therm al perform ance with a reference to a specific converter operatin g poi nt an d
specific system conditions.

8.1 .2 Calculation of hot-spot temperature rise from normal heat-run test data
A therm al diagram is assum ed, as shown in Figure 6, on the und erstanding that such a
diagram is the sim plification of a m ore complex distribution. The assum ptions mad e in this
simplification are as follows.
a) The oil tem perature inside the tank increases linearl y from bottom to top, whatever the
cooling m ode.
b) As a first approxim ation, the temperature rise of the conductor at an y position up the
wind ing is assum ed to increase linearl y, parallel to the oil tem perature rise, with a
constant d ifference gr between the two straight lines ( gr being the difference between the
wind ing average temperature rise by resistance and the average oil temperature rise in
the tank).
c) The hot-spot tem perature rise is higher than the temperature rise of the conductor at the
top of the wind ing as d escribed in 8. 1 . 2 b), because allowance has to be made for the
 

increase in stray losses, for differences in local oil flows and for possible additional paper
on the conductor. To take into account these non -linearities, the difference in tem perature
between the hot-spot and the top-oil in tank is m ade equal to H × g r, that is, ∆ θ h r = H × gr.
NOTE I n m any cases, it has been observed th at the tem peratu re of the tank outlet oil is high er than that of
the oil i n the oil pocket. I n such cases, the tem perature of th e tank outl et oil is used for l oading.
 – 30 – I EC 60076-7: 201 7 © I EC 201 7

addition to the DC resistance loss. Connections that are subj ect to leakage flux heating,
such as coil-to-coil connections and som e tap-to-winding brazes, should also be
considered.
c) Conduction heat transfer effects within the winding caused by the various insulation
thicknesses used throughout the winding.
d) Local design features or local fluid flow restrictions.
• Layer insulation m ay have a different thickness throughout a layer winding, and
insulation next to the cooling duct affects the heat transfer.
• Flow-directing washers reduce the heat transfer into the fluid in the case of a
zigzag-cooled winding (Figure 1 0).

IEC

Fi g u re 1 0 – Zi g zag -cool ed wi n d i n g wh ere th e d i stan ce between al l secti on s i s th e

sam e an d th e fl ow-di recti n g wash er i s i n stal l ed i n th e space between secti on s

• Possible extra insulation on end turns and on winding conductors exiting through the
end insulation.
• Not all cooling ducts extend com pletel y around the winding in distribution transform ers
and small power transform ers. Some cooling ducts are located onl y in the portion of
the winding outside the core (see Figure 1 1 ). Such a “collapsed duct arrangem ent”
causes a circumferential tem perature gradient from the centre of the winding with no
ducts under the yoke to the centre of the winding outside core where cooling ducts are
located.
I EC 60076-7: 201 7 © I EC 201 7 – 33 –

The two equations are

d ∆θ h1
k 21 × K × ( ∆θ hr ) = k22 × τ w ×
y
+ ∆θ h1 (7)
dt

and

d ∆θ h2
( 21 − 1 ) ×
k K
y
× ( ∆θ hr ) = (τ o / k22 ) ×
dt
+ ∆θ h2 (8)

the solutions of which are combined in accordance with Equ ation (6).

The final eq uation for the hot-spot temperature is

θ h = θ o + ∆θ h (9)

Regard ing Equ ations (5) to (8), the com plexity is in order to accou nt for the fact that the oil-
cooling m edium has m echanical inertia in ad dition to thermal inertia. The effect is greatest for
natural cooling (ON ), somewhat less for non-directed-flow pum ped-oil cooling (OF), and
negligible for directed-flow pumped-oil cooling (OD), as regards power transformers. I t is also
negligible for small transform ers (see 8. 2. 2).

NOTE For ON an d OF cooli n g, the oil viscosity chang e cou nteracts the effect of th e ohm ic resistance vari ation of
the con ductors. I n fact, the cooli ng effect of the oi l viscosity chan ge is strong er than th e heatin g effect of th e
resistance change. This has been taken into accou nt im plicitly by th e win din g exponent of 1 , 3 in Tabl e 5. For OD
coolin g, the i nfluence of th e oil viscosity on tem peratu re ri ses is slight, and th e effect of the ohm ic resistance
variati on is considered. An approxim ate correcti on term (with its sign) for the h ot-spot tem perature rise at OD is
0, 1 5 × ( ∆ θ h – ∆ θ h r ).

8.2.2 Exponential equation s solution

Subclause 8.2. 2 describes the exponential eq uation solution to the heat transfer d ifferential
Equations (5) to (8).

An example of a load variation according to a step fu nction is shown in Figure 1 3 (the details
of the exam ple are given in Annex H).
I EC 60076-7: 201 7 © I EC 201 7 – 35 –

where two grad ients are

{ }(
∆θ h1 (t ) = ∆θ h1 i + k21 Hg r K y − ∆θ h1 i × 1 − e (−t ) /( k22 × τw ) ) (1 3)

and

∆θ h2 (t ) = ∆θ h2 i + {(k21 − 1 )Hgr K y − ∆θ h2 }× (1 − e ( t ) /(
i
− τ0 / k22 ) ) (1 4)

Correspondingl y, the hot-spot to top-oil temperature grad ient decrease to a level

corresponding to a load factor of K is given by:

{ }
∆θ h1 (t ) = k21 Hg r K y + ∆θ h1 i − k21 Hg r K y × e (−t ) /( k22 × τ w ) (1 5)

and

{ }
∆θ h2 (t ) = (k21 − 1 )Hg r K y + ∆θ h2 i − (k21 − 1 )Hgr K y × e (−t ) /( τ0 / k22 ) (1 6)

The final equation for the hot-spot tem perature is:

θ h (t ) = θ o (t ) + ∆θ h (t ) (1 7)

where
τw is the winding time constant (m in);
τ0 is the oil-time constant (m in).
The top-oil exponent x and the wind ing exponent y are given in Table 4 [23], [24].
The constants k1 1 , k21 , k22 and the time constants τ w and τ 0 are transformer specific. They
can be determ ined in a prolonged heat-run test during the “no-load loss + load loss” period, if
the supplied losses and correspond ing cooling conditions, for exam ple AN or AF, are kept
unchanged from the start until the stead y state has been obtained (see Annex F). I n this case,
it is necessary to ensure that the heat-ru n test is started when the transformer is
approximatel y at the am bient tem perature. I t is obvious that k21 , k22 and τ w can be defined
onl y if the transform er is equipped with fibre optic sensors. I f τ 0 and τ w are not defined in a
prolonged heat-run test they can be defined by calculation (see Annex E). I n the absence of
transformer-specific values, the valu es in Table 4 are recomm end ed. The corresponding
graphs are shown in Figure 1 4.

NOTE 1 U nless th e current and coolin g con ditions rem ain u nchan ged d uri ng the heati ng process long enoug h to
project the tang ent to the in iti al heati ng curve, the tim e constants cannot be d eterm ined from the heat-run test
perform ed accordi ng to I EC practice.

NOTE 2 The ∆ θ h ( t)/ ∆ θ h r graphs observed for sm all transform ers are sim ilar to graph 7 i n Figu re 1 4, i. e. sm all
transform ers do n ot sh ow such a hot-spot “oversh oot” at step increase i n th e l oad cu rrent as ON - and OF-cool ed
power transform ers do.
N OTE 3 The backgrou nd of the oil, x , an d wind ing, y , exponents and correspon din g d eterm ining proced ure are
given in Annex G.
 – 36 – I EC 60076-7: 201 7 © I EC 201 7

Table 4 – Recommended thermal characteristics for exponential equations

transformers
Small
M edium and large power tran sformers

Characteri stic

restricted a

restricted a

restricted a
ONAN
ONAN

ONAN

ONAF

ONAF

OD
OF

OF
Oil exponent x 0, 8 0, 8 0, 8 0, 8 0, 8 1 ,0 1 ,0 1 ,0
Winding expon ent y 1 ,6 1 ,3 1 ,3 1 ,3 1 ,3 1 ,3 1 ,3 2, 0
Constant k
11 1 ,0 0, 5 0, 5 0, 5 0, 5 1 ,0 1 ,0 1 ,0
Constant k
21 1 ,0 3, 0 2, 0 3, 0 2, 0 1 , 45 1 ,3 1 ,0
Constant k
22 2, 0 2, 0 2, 0 2, 0 2, 0 1 ,0 1 ,0 1 ,0
Tim e constant τ 0 , m in 1 80 21 0 21 0 1 50 1 50 90 90 90
Tim e constant τ w , 4 10 10 7 7 7 7 7
m in
a I f a wi ndi ng of an ON - or OF-cool ed transform er is zi g zag -cooled, a radial spacer thickn ess of less than
3 mm m ight cause a restricted oil circul ation, i. e. a hi gh er m axim um value of the fu nction ∆ θ h ( t)/ ∆ θ h r th an
obtai ned by spacers ≥ 3 m m .
 – 38 – I EC 60076-7: 201 7 © I EC 201 7

Equations (7) and (8) become

D ∆θ h1 =
k
Dt

22τ w
[
× k21 × ∆θ hr K − ∆θ h1 y
] (20)

and

D ∆θ h2
Dt

1( k22 )τ o
× [( k21 1 ) × ∆θ hr K y
∆θ h2 ] (21 )

The n th values of each of ∆ θ h1 and ∆ θ h2 are calculated in a way similar to Equation (1 9).

The total hot-spot tem perature rise at the n th time step is given by:

∆θ h( n ) = ∆θ h1 ( n ) − ∆θ h2( n ) (22)

Finall y, the hot-spot temperature at the n th time step is given by:

θ h( n ) = θ o( ) + ∆θ h( )
n n
(23)

For an accurate solution, the time step D t should be as small as is practicable, certainl y no
greater than one-half of the sm allest time constant in the therm al m odel. For example, if the
time constant for the winding considered is 4 min, the time step should be no larger than
2 m in. τw and τo should not be set to zero .

Also, there are theoreticall y more accurate num erical anal ysis solution methods than the
simple one used in Equations (1 8) to (21 ), for exam ple trapezoidal or Runge-Kutta methods.
However, the increased com plexity is not warranted here considering the imprecision of the
input data.

The loss of life of cellulose insulation differential equations of 6. 4 can also be converted to
difference equations. The fundam ental differential equation is

dL (24)
=V
dt

impl ying

DL ( n) = V( ) × Dt
n
(25)

and

L ( n) = L( n −1 ) + DL( )
n
(26)

An application example of the difference equations solution is given in Annex I .

I EC 60076-7: 201 7 © I EC 201 7 – 39 –

8.3 Ambient temperature

8.3.1 Outdoor air-cooled transformers
For d ynam ic considerations, such as m onitoring or short-tim e em ergency loading, the actual
temperature profile should be used d irectl y.

For d esign and test considerations, the following eq uivalent temperatures are taken as
ambient temperature:
a) the yearly weighted am bient temperature is used for thermal ageing calculation ;
b) the m onthl y average temperature of the hottest month is used for the m axim um hot-spot
temperature calcu lation.
NOTE Concernin g the am bient tem peratu re, see also I EC 60076-2.

I f the ambient tem perature varies appreciabl y during the load cycle, then the weighted
ambient tem perature is a constant, fictitious am bient temperature which causes the sam e
ageing as the variable tem perature acting during that tim e. For a case where a tem perature
increase of 6 K d oubles the ageing rate and the ambient tem perature can be assum ed to vary
sinusoidall y, the yearl y weighted ambient tem perature, θ E , is equal to

[ (
θ E = θ ya + 0,01 × 2 θ ma - max − θ ya ) ] 1 ,85 (27)

where
θ m a-m ax is the m onthl y average temperature of the hottest month (which is equ al to the sum
of the average daily maxim a and the average daily m inima, measured in °C, during
that m onth, over 1 0 or m ore years, d ivid ed by 2);
θ ya is the yearl y average tem perature (which is equal to the sum of the monthly
average temperatures, measured in °C, d ivid ed by 1 2).
EXAMPLE U si ng m onthl y averag e valu es (m ore accu ratel y u sing m onthly wei ghted valu es) for θ a :

θ ma- ma x = 30 ° C for 2 m onths 


θ ma = 20 °C for 4 m onths  Averag e θ ya = 1 5, 0 °C

θ ma = 1 0 °C for 4 m onths  Weighted average θ E = 20, 4 °C

θ ma = 0 °C for 2 m onths 

The am bient tem peratu re used in the calcul ation exam ples in Annex J is 20 °C.

8.3.2 Correction of ambient temperature for transformer enclosure

A transformer operating in an enclosu re experiences an extra temperature rise which i s about
half the temperature rise of the air in that enclosure.

For transformers installed in a m etal or concrete enclosure, ∆ θ or in Equations (1 0) and (1 1 )

should be replaced by ∆θ or
′ as follows:

′ = ∆θ or + ∆ (∆θ or )
∆θ or (28)
where
∆ ( ∆ θ or ) is the extra top-oil temperature rise und er rated load .
I t is strongly recom m ended that this extra tem perature rise be determined by tests, bu t when
such test results are not available, the values given in Table 5 for different types of enclosure
 – 40 – I EC 60076-7: 201 7 © I EC 201 7

may be used. These values should be divided by two to obtain the approximate extra top-oil
temperature rise.

NOTE When the enclosu re d oes not affect the coolers, no correcti on is necessary accordi ng to Equation (28).

Tabl e 5 – C o rre c t i o n fo r i n c re a s e in am b i en t t e m p e ra t u re due t o e n c l o s u re

C o rr e c t i o n to be ad d ed to we i g h te d

am bi en t t e m p e ra tu re

N u m ber of K
T yp e o f e n c l o s u re t ra n s fo rm e rs

i n s ta l l e d T r a n s fo rm e r si ze

kVA
2 50 50 0 750 1 000

1 11 12 13 14
U nd ergroun d vaults with natural ventilati on 2 12 13 14 16
3 14 17 19 22
1 7 8 9 10
Basem ents and bui ldi ngs with poor natural 2 8 9 10 12
venti lation
3 10 13 15 17
1 3 4 5 6
Buildin gs with good natural ventilati on an d
und ergroun d vaults an d basem ents with 2 4 5 6 7
forced ventilati on
3 6 9 10 13
Ki osks a 1 10 15 20 –
N OTE The above tem perature correction fig ures have been estim ated for typical substati on load ing conditions
using representati ve valu es of transform er losses. Th ey are based on the resu lts of a series of natu ral an d
forced coolin g tests in un derg rou nd vaults and su bstations and on ran dom m easurem ents in substations and
kiosks.
a This correction for kiosk enclosures is not n ecessary when the tem perature rise test has been carried out on
the transform er in the enclosu re as one com plete unit.

8. 3. 3 W a t e r- c o o l e d t ra n s fo rm e rs

For water-cooled transform ers, the am bient temperature is the temperature of the incoming
water, which shows less variation in time than air.

9 I n fl u e n c e o f t a p - c h a n g e rs

9. 1 G e n e ra l

All quantities used in Equations (1 0), (1 1 ), (1 3), (1 4), (1 5), (1 6) and (1 7) have to be
appropriate for the tap at which the transformer is operating.

For exam ple, consider the case where the H V voltage is constant, and it is required to
maintain a constant LV voltage for a given load. I f this requires the transform er to be on a
+ 1 5 % tap on th e LV si d e, th e rated oi l tem peratu re rise, l osses an d wi n d i n g gradients have to
be m easured or calculated for that tap.

Consider also the case of an auto transform er with a line-end tap-changer – the series
winding will have maximum current at one end of the tapping range whilst the comm on
winding will have maximum current at the other end of the tapping range.
 – 42 – I EC 60076-7: 201 7 © I EC 201 7

An n e x A

(informative)
I n s u l a ti o n l i fe e xp e c ta n c y an d re l a ti ve ag ei n g ra te

c o n s i d e ri n g o x yg e n an d w a te r e ffe c t

A. 1 I n s u l ati on l i fe e x p e c t a n c y

Ageing or change of polymerization of paper insulation is often described as a first order

process that can be described by the following Arrhenius equation:

EA
−
1 1 × (θ h + 273 ) (A. 1 )
− = A ×t×e R

DP end DPstart

where
DP
end is the insulation DP value at the mom ent of the sampling or the end of life criterion ;
DP
start is the initial insulation DP value;
A is the pre-exponential factor in 1 /h;
E
A is the activation energ y i n kJ/m ol;
t is the life time of a transformer in h;
R is the gas constant in J/(K· m ol);
θh is the hot-spot temperature, °C.
Arrhenius extrapolations assum e that a chem ical degradation process is controlled by a
reaction rate k proportional to exp( − EA / R T), where EA is the Arrhenius activation energ y, R the
gas constant (8, 31 4 J /(K· m ol), T the absolute tem perature and A the pre-exponential factor.
The pre-exponent value A is a constant depending on the chemical environment. I n an
Arrhenius plot, the natural logarithm of the ageing rate (ln k) is plotted against the inverse
absolute tem perature (1 / T) and a straight line is obtained as shown in Figure A. 1 – illustrating
how the ageing rate depends on temperature. The condition for achieving a straight line is
that it is the same ageing process over the whole temperature range. The activation energ y
describes how m uch the reaction rates depend on tem perature; if the ageing process is
independent of tem perature the activation energy is zero and the line becomes parallel with
the x -axis, while if it increases fast with increasing tem perature the line falls quickl y. One
should bear in m ind that A and EA values come in pairs. I n principle, EA is the slope of the line
in Figure A. 1 , and the A -value is the value of interception with a virtual y -axis from 0, the
higher the value the higher the location of the curve above the abscissa and the ageing is
faster. The point is that a sm all change in slope will influence the A -value significantl y.

NOTE The tem peratu re an d m oisture values used i n th e transform er life estim ation al ways refer to th e sam e
location.
 – 44 – I EC 60076-7: 201 7 © I EC 201 7

Table A.2 – Expected life of paper u nder various conditions

Expected li fe
years
Paper type/ageing temperatu re
Free from ai r Free from air Free from ai r Wi th ai r
and 0, 5 % and 1 ,5 % and 3, 5 % and 0,5 %
moisture m oisture moi sture m oisture
Non -th erm ally 80 °C 97, 3 26, 6 8, 9 1 4, 7
upgraded paper at
90 °C 29, 3 8 2, 7 6, 4
98 °C 1 1 ,7 3, 2 1 ,1 3, 4
1 1 0 °C 3, 2 0, 9 0, 3 1 ,4
Therm all y upgrad ed 80 °C 1 51 , 9 81 39, 9 1 9, 4
paper at
90 °C 67, 8 36, 1 1 7, 8 9
98 °C 36, 7 1 9, 6 9, 6 5
1 1 0 °C 1 5, 3 9, 6 4 2, 2

A.2 Relative ageing rate considering oxygen and water effect

I n 6. 3, the rate of ageing of the interturn insulation of transform ers under the effect of time
and temperature is referred to a hot-spot temperature of 98 °C and 1 1 0 °C for the non-
thermall y and therm all y upgraded paper, respectivel y. Further, the relative ageing rate V is
defined according to Equations (2) and (3). These equations are based on the life expectancy
form ula of Montsinger [25], and Dakin’s ageing rate formula [26], which are a sim plification of
the more general Arrhenius relation given in Equation (A. 1 ), and valid onl y in a lim ited
temperature range. On the other hand, the I EEE Loading Guide [27], recom mends
Equation (3), which is an equivalent to the acceleration ageing factor, FAA , for a wide range of
temperatures. Further, Equations (2) and (3) im ply that the ageing rate is onl y dependent on
the hot-spot temperature and do not consider different insulation conditions, which are
defined in references [9] to [1 5].

Therefore, if the ageing rate of the paper insulation is given as follows:

E
−

k = A×e R × (θ h + 273 )
(A. 3)

and if an ageing rate at a certain temperature and at an insulation condition is chosen to be

the rated one, kr , then the ageing rate, k, determ ined for an y temperature and insulation
condition can be related to this rated rate, kr , by a new relative ageing rate, V, given as their
ratio [28]:

1  Er E 
× 
k A R
−
 θ
h,+ 273 θh +273 
V= = e  
(A. 4)
kr Ar
where
the subscript r stands for the rated condition.
The chosen rated insulation condition for both the non-thermall y and thermall y upgraded
paper is “free from air and 0, 5 % moisture” taken from Table A. 1 . Also, similar to the approach
given in Clause 6, the rated relative ageing rate V = 1 , 0 at this condition corresponds to a
temperature of 98 °C for non-therm all y upgraded paper and to 1 1 0 °C for therm ally upgraded
paper.
I EC 60076-7: 201 7 © I EC 201 7 – 45 –

The calculated relative ageing rates for different temperatures and insulation contamination
cond itions are com pared with valu es given in Table 1 . The resu lts are summ arized in
Table A. 3 and Table A.4.

The same cou ld be applied to improve the I EEE equ ations for the ageing acceleration factor,
FAA , and the eq uivalent ageing factor, FEQA .
I t is obvious from the tables that the dominant ageing factor at higher tem peratures for the
kraft paper is the m oistu re. However, at lower temperatures the oxygen influence will prevail,
Table A. 3. On the other hand, the m ain factor responsible for the thermall y upgraded paper
ageing over the range of the tem peratures is oxygen, Table A. 4. This is in line with the
conclusions in [1 3].

Tabl e A. 3 – Rel ative ag ein g rates d u e to h ot-spot tem peratu re, oxyg en
an d m oistu re for n on-u pg rad ed paper i n su lation

Rel ati ve ag ei n g rate, V

Tem peratu re
Tabl e 1 Free from ai r Free from ai r Free from ai r Wi th ai r
°C an d 0, 5 % an d 1 , 5 % an d 3, 5 % an d 0, 5 %
m oi stu re m oi stu re m oi stu re m oi stu re

80 0, 1 25 0, 1 2 0, 44 1 , 323 0, 80
86 0, 25 0, 25 0, 91 2, 742 1 , 32
92 0, 5 0, 50 1 , 84 5, 548 2, 1 6
98 1 ,0 1 , 00 3, 66 1 0, 9 76 3, 47
1 04 2 1 , 94 7, 08 21 , 245 5, 50
110 4 3, 67 1 3, 43 40, 281 8, 58
116 8 6, 82 24, 96 74, 880 1 3, 21
1 22 16 1 2, 44 45, 53 1 36, 601 20, 07
1 28 32 22, 30 81 , 58 244, 755 30, 1 0
1 34 64 39, 27 1 43, 69 431 , 061 44, 62
1 40 1 28 68, 04 248, 93 746, 802 65, 39
The relati ve agei ng rates, V, for different agei ng factors at tem perature of 98 °C are in dicated to be com pared to
the rated i nsulati on condition, i . e. where th e relati ve agei ng rate is 1 .
 – 46 – I EC 60076-7: 201 7 © I EC 201 7

Tabl e A. 4 – Rel ati ve ag ei n g rates d u e to h ot-spot tem peratu re, oxyg en

an d m oi stu re for u pg rad ed paper i n su l ati on

Rel ati ve ag ei n g rate, V

Tem p eratu re
Tabl e 1 F ree from ai r Free from ai r Free from ai r Wi th ai r
°C an d 0, 5 % an d 1 , 5 % an d 3, 5 % an d 0, 5 %
m oi stu re m oi stu re m oi stu re m oi stu re

80 0, 036 0, 1 0 0, 1 9 0, 38 0, 79
86 0, 073 0, 1 6 0, 31 0, 63 1 , 25
92 0, 1 45 0, 26 0, 5 1 , 00 1 , 97
98 0, 282 0, 42 0, 78 1 , 59 3, 05
1 04 0, 536 0, 65 1 , 22 2, 48 4, 66
110 1 , 00 1 , 00 1 , 88 3, 81 7, 02
116 1 , 83 1 , 52 2, 84 5, 78 1 0, 45
1 22 3, 29 2, 27 4, 26 8, 66 1 5, 36
1 28 5, 8 3, 36 6, 30 1 2, 82 22, 32
1 34 1 0, 07 4, 91 9, 22 1 8, 74 32, 07
1 40 1 7, 2 7, 1 1 1 3, 33 27, 1 2 45, 60
The rel ative agei ng rates, V, for different ag ein g factors at tem peratu re of 1 1 0 °C are i ndi cated to be com pared
to the rated i nsulati on condition, i. e. wh ere the relati ve ag ein g rate is 1 .

The loss of life, L , over certain period of time is calculated as given in 6. 4.

I EC 60076-7: 201 7 © I EC 201 7 – 47 –

Annex B
(informative)
Core temperature

B.1 General
I n transform er cores, there are two different core hot-spots which, if not controlled, can cause
insulation m aterial degradation and subseq uent gassing.
a) Core hot-spot inside the core, shou ld be limited to 1 30 °C under conditions of highest core
excitation, rated load and maximum am bient tem perature. This is in order to prevent the
core heating which resu lts in the break-up of the thin oil film between core laminations,
the consequence of which is the generation of m ainly H 2 and CH 4 , in addition to sm all
quantities of other h ydrocarbons [29], [30]. I t is important to understand that an y possible
oil gas saturation should be prevented.
b) Core surface hot-spot, which is in contact with oil and solid insu lation m aterials, should be
lim ited according to Table 2.

B.2 Core hot-spot locations

The location of the internal core hot-spot depends largel y on the core type and whether it is a
shell form or a core form transformer. I n the m ost comm on core type (three-phase, three-limb
cores), this hot-spot is located in the m idd le of the top yoke between cooling ducts. I n other
core types, the location of the core hot-spot is typi call y at the top of the m id dle core lim b [30].

The total core surface temperature rise is the sum of the following three components:
a) tem perature rise due to the leakage flux impinging on the surface of the lam inations of the
outerm ost core step(s) – this value can vary from a few kelvins to several tens of kelvins
over the adj acent oil d epending on the transformer wind ing, core, and tank shield ing
design;
b) tem perature rise d ue to the main flux in the core – this value can again vary from a few
kelvins to several tens of kelvins over the adj acent oil depend ing on the transformer core
design (diameter and num ber of cooling d ucts), core induction, and core m aterial;
c) tem perature rise of the oil around the area of the surface hot-spot.
Therefore, in almost all cores, this core surface hot-spot is not located in the yoke but is
located at the top of the middle core-limb, where the leakage flux enters the surface of the
core laminations. Also, the relative magnitud es of the temperature rise due to the leakage flux
versus the rise due to the core main flux depends wholl y on the design of the transformer.

Consequ ently, the correct way to determ ine
tem perature increase at rated load (includ ing
the temperature rise d ue to the highest core
determined for the appropriate location of
evaluated .
the core surface hot-spot is to determ ine the
temperature rise of adj acent oil) and add to it
over-excitation. These temperatures are to be
the core surface hot-spot in the core type
 – 48 – I EC 60076-7: 201 7 © I EC 201 7

Annex C
(informative)
Specification of loading beyond rated power
This docum ent gives advice on the calculation of the capability of an existing transform er to
be loaded beyond rated power. All transform ers will have some overload capability. However
since no specific loading requirem ents beyond rated power are specified in I EC 60076-1 [5] or
I EC 60076-2 [52], it is up to the purchaser to specify an y particular loading requirements
(load, duration and ambient temperature).

Specification of loading beyond rated power can be done in the following ways.
a) Long tim e emergency loading
Since the hot-spot temperature limit in I EC 60076-2 is less than that given in Table 2, it is
possible to have an increased loading available for em ergency situations provided that the
loss of life is accepted. The extent of this loading capability will depend on am bient
tem perature, while preload and the duration of loading are onl y relevant to loss of life.
The following need to be specified at the enquiry stage:
1 ) the am bient temperature at which the loading is required;
2) the load as per unit (p. u. ) of rated current;
3) the winding(s) to which the loading is to be applied;
4) the tap position;
5) the cooling stage(s) in service.
I f loading according to this docum ent is specified, then the transformer should n ot exceed
the tem peratures and currents given in Table 2 and Table 3, respectivel y, under the
following conditions:
• a yearl y average am bient tem perature (20 °C unless otherwise specified);
• current flowing in the highest rated winding is considered;
• the tap position that gives the rated voltage on the lower voltage side with rated
voltage on the higher voltage side taking account of the voltage drop caused by th e
load for a unity power factor load;
• all norm al cooling in service but with no standby cooling capacity.
b) Short time em ergency loading
I f the transformer is used at a load less than rated current, then there will be an additional
short time loading capability caused by the therm al time constants of the oil and windings.
I f a specific short time loading capability is required then the following need to be
specified:
1 ) the ambient tem perature at which the loading is required;
2) the short time current that is requ ired in p. u. of rated current;
3) the winding(s) to which the load will be applied;
4) the tap position;
5) the preload current applied before the short tim e emergency loading in p.u. of rated
current;
6) the duration of the loading;
7) the cooling stage(s) in service.
I f loading according to this document is specified, then the transform er should not exceed
the tem peratures and currents given in Table 2 and Table 3, respectively, and under the
following conditions:
• a yearl y average am bient temperature (20 °C unless otherwise specified);
I EC 60076-7: 201 7 © I EC 201 7 – 49 –

• current flowing in the highest rated wind ing is considered ;

• the tap position that gives the rated voltage on the lower voltage sid e with rated
voltage on the higher voltage side taking account of the voltage drop caused by the
preload for a unity power factor load;
• a preload of 0, 75 p. u. ;
• a duration of 1 5 m in;
• all norm al cooling in service appropriate to the preload cond ition but with no stand by
cooling capacity
c) Loading according to a specific cycle
Specify in detail the load and am bient temperature cycles.
 – 50 – I EC 60076-7: 201 7 © I EC 201 7

Annex D
(informative)
Description of Q, S and H factors
I EC 60076-2 notes also that the hot-spot factor H is obtained by the product of the Q and S
factors:

H = QS (D. 1 )

The Q and S factors are dimensionless factors described in I EC 60076-2 as:

• Q is “a dim ensionless factor to estimate the increase of the average winding gradient due
to the local increase of additional loss.”
• S is “a dimensionless factor to estim ate the local increase of the average winding gradient
due to the variation in the oil flow stream .”
According these definitions, Q should be calculated by modelling the winding with the correct
heat loss distribution, but with uniform oil velocity. The Q factor is then the ratio of the
maxim al winding to local oil gradient over the average winding to average oil gradient.

On the other hand, S should be calculated by modelling the winding with uniform heat losses
and with the correct oil velocity inside the winding. The S factor is the ratio of the m axim al
winding to local oil gradient over the average winding to average oil gradient.

However after calculating Q and S in line with the I EC 60076-2 definition, the hot-spot factor H
cannot be calculated directl y as the product of Q and S factors, as m entioned in I EC 60076-2,
because:
– discs with maximal Q factor and discs with maximal S factor can (and probabl y will) be
different discs;
– when m odelling the winding with correct heat loss distribution and oil flows, Q and S
factors will not be independent from each other as explained later.
For the above reasons, CI GRE WG A2. 38 [1 8] proposes m ore practical definitions of Q and S
and H factors, as given below.

The H factor can be derived out of Figure 6 and is:

H= P−B (D. 2)
I− B+D
2

Equation (D. 2) is different from the current I EC 60076-2 definition, because the hot-spot
tem perature is referenced to the m ixed top-oil, while increase in local winding to oil gradient
refers to local winding oil.

This formula for H has the following advantages:

– the H factor can be calculated directl y from the calculation results (calculated with correct
loss and oil flow distribution);
– this is the correct hot-spot factor to predict the hot-spot tem perature out of the
tem perature rise results, obtained in the standard tem perature rise test;
– this hot-spot factor can also be checked in case fibre optic measurem ents are m ade (the
hot-spot tem perature is known);
I EC 60076-7: 201 7 © I EC 201 7 – 51 –

– the hot-spot is not always located at the top of the winding. H owever, this form ulation is a
practical solution to overcome this issue.
The Q factor is a dim ensionless factor as a ratio of two losses, and in cyli ndrical coordinates
can be defined as:

Q = Q (r, z, ϕ , T ) Qave (D. 3)

where
Q ( r, z , φ, T) is the local loss density at a location (W/m 3 );
r is the radial position;
φ is the angle in circum ferential position;
z is the axial position;
T is the local temperature at ( r, z, φ ), and
Q ave is the average loss of the winding at average temperature.
For calculation purposes, one can redefine the Q factor for a disc winding in which each disc
has several conductors in radial direction and consists of num erous discs in axial direction,
as:

Q = Q (conductor number in disc, disk number, ϕ , T ) Qave (D. 4)

The Q factor is a scalar function and is based on the stead y state condition of a defined
loading at a defined tap position (if applicable).

I t is im portant to note that the Q factor in this definition is not a ratio of tem peratures but a
ratio of losses.

Finall y the S factor used in this document is defined as:

S= H (D. 5)
Q
which can be easil y calculated as soon as H and Q are known.

This S factor is an indication of the local cooling inefficiency. H igher S factor means higher
local tem perature gradient thus worse cooling efficiency.

According to the current I EC 60076-2 calculation this S factor should be calculated as the
ratio of local hot-spot gradient over winding gradient with constant heat losses. With this
definition the S factor is proportional to the ratio of two therm al resistances, resulting in:

S ∝ S (r, z, ϕ , T ) Save (D. 6)

where
S( r, z, φ , T) is the local cooling resistance (K/W), and
Save is the average cooling resistance (K/W).
We should note that heat transfer can be in different directions. The (overall) local heat
transfer consists of series and parallel parts, such as:
– the insulation between the neighbouring conductors, that are in direct contact with each
other, in the radial direction.
 – 52 – I EC 60076-7: 201 7 © I EC 201 7

– the insulation paper and oil boundary layer between conductor and the oil flow in axial
direction. Note that heat transfer functions for the oil boundary layer are often function
of q ″ .
– the copper (which alm ost can be neglected) in the tangential direction.
This implies that the Q and S factors are not full y independent, because they are linked by
temperature, heat flux, etc. For exam ple, if the local losses are higher, the local tem perature
will also increase and will influence the local flow stream and the local convection heat
transfer coefficient from conductor to oil.

Rem ark 1 We should note that the size of one (paper insulated) cond uctor is the sm allest
elem ent in which one calculates the losses. I nside each elem ent there exists the same
temperature, so therm al resistances inside the elem ent are neglected. I n the case one
calculates the Q factor based on a number of conductors in one (or sometim es even 4) top
discs, one increases in essence the elem ent size to a large extent and one neglects the
temperature distribution between conductors in the disc (and even between discs), which
results in a too low estimate of the hot-spot. The approach of using one or more discs as
sm allest element results in a too low estimate of the hot-spot temperature and should be
rejected.

Rem ark 2 I n the case of a high Q factor in a transformer, one is able to lim it the hot-spot
factor by creating locally m ore cooling surface and so design for a low S factor at that
location. This principle is easy to do by adding an axial cooling channel inside a radial spacer
disc or by adding a radial spacer inside a winding with axial cooling channels. The location of
the hot-spot does not necessaril y have to correspond with the location of the m axim um
losses.
I EC 60076-7: 201 7 © I EC 201 7 – 53 –

Annex E
(informative)
Calculation of winding and oil time constant
The winding tim e constant is as follows:

τw =
mw × cw × g (E. 1 )
60 × Pw

where
τw is the winding time constant in min at the load considered;
g is the winding-to-oil gradient in K at the load considered;
mw is the bare mass of the winding in kg;
cw is the specific heat of the conductor material in Ws/(kg· K) ( 390 for Cu and 890 for Al);
Pw is the winding loss in W at the load considered.
Another form of Equation (E. 1 ) is

τw = 2,75 ×
g for Cu (E. 2)
(1 + Pe ) × s
2

τw = 1 ,1 5 ×
g for Al (E 3)
(1 + Pe ) × s 2
where
Pe is the relative winding edd y loss in p. u.;
s is the current density in A/mm 2 at the load considered.
The top-oil tim e constant is calculated according to the principles in references [27] and [31 ].
I t means that the therm al capacity C for the ON AN , ON AF, OF and OD cooling m odes is:

C = c W × m W + cFE × mFE + cT × mT + kO × cO × mO (E. 4)

where
m W is the mass of coil assembl y in kg;
m FE is the mass of core in kg;
m T is the mass of the tank and fittings in kg (onl y those portions that are in contact with
heated oil should be used, i. e. 2/3 of tank weight should be considered);
mO is the mass of oil in kg;
cw is the specific heat capacity of the winding material (390 for Cu and 890 for Al) in
Ws/kgK;
c FE is the specific heat capacity of the core (= 468) in Ws/kgK;
cT is the specific heat capacity of the tank and fittings (= 468) in Ws/kgK;
cO is the specific heat capacity of the oil (= 1 800) in Ws/kgK;
kO is the correction factor for the oil in the ON AF, ON AN , OF and OD cooling modes.
The correction factor for the oil, kO , is the ratio of average to m axim um top-oil tem perature
rise.
 – 54 – I EC 60076-7: 201 7 © I EC 201 7

NOTE I t is apparent that all the oil in a transform er is not h eated to the sam e tem peratu re as the top-oil. Th e ratio
of averag e to m axim um top-oi l tem peratu re rise ran ges from 65 % to 9 5 % d epen d i n g on d esi g n . I f th e
correspon din g tem perature rises are n ot known then an average val ue of these fi gu res, i. e. 80 % , is used as the oil
correction factor for ON AN and ON AF coolin g m odes. For the forced-oil cooli ng m odes, eith er OF or OD, and
distribution transform er without extern al radiators the correcti on factor is 1 00 % .

The top-oil tim e constant at the load considered is given by the following:

τo =
Co ∆×θo
(E. 5)
60 × P
where
τo is the top-oil time constant in min;
∆θ o is the top oil temperature rise above am bient temperature in K at the load considered;
P is the supplied losses in W at the load considered.
I EC 60076-7: 201 7 © I EC 201 7 – 55 –

Annex F
(informative)
Thermal model parameters

F.1 General

As d efined in 8.2. 3, the first grad ient ∆ θh1 (Equations (7) and (1 3) with thermal constants k21
and k22 ) represents the fundamental hot-spot tem perature rise, before the effect of changing
oil flow past the hot-spot is taken into account. The second grad ient ∆ θh2 (Equ ations (8) and
(1 4) with parameters k21 − 1 and k22 ) represents the varying rate of oil flow past the hot-spot,
a phenomenon which changes m uch m ore slowl y. The combined effect of these two terms is
to accou nt for the fact that a sud den rise in load current m ay cause an otherwise
unexpectedl y high peak in the hot-spot temperature rise, very soon after the sud den load
change. Sim ilarly, therm al constant k1 1 in Eq uations (5) and (1 1 ) acts as a correction factor
for the top-oil time constant, taking into account the fact that the time constant is being
affected in a length of the time by the viscosity change.

The procedure to estim ate these therm al constants from a prolonged heat-run test during the
“no-load loss + load loss” period is defined in Clause F. 2. The thermal constant estim ation
procedure from service d ata is defined in [1 8].

The model in this document represents the traditional way of mod elling based on man y years
of experience [21 ], [31 ], [33]–[39], and it has been verified u nder d ifferent operational
cond itions [1 8], [21 ], providing satisfactory hot-spot temperature estim ates. Being based on
constant parameters the m od el provides exponential response for predefined therm al
processes, and as such its further d evelopm ent in this form could be limited. However, due to
extreme simplicity in application in the dail y life of an engineer, it is still considered
appropriate.

On the other hand , it is possible to provide complete anal ytical solution for these ph ysical
processes consid ering all system variables (oil viscosity, loss change with temperature, etc.),
as is partly illustrated in Clause F.3. This approach would dem and a certain change in
m odelling practice, for exam ple d eviation from traditional x and y exponents.

F.2 Thermal constant esti mation: experimental approach

The thermal constant k1 1 should be estimated for the transient top-oil rise temperature curve
obtained during the test period with total losses as follows (see [32]).
1 ) Define fu nction f1 ( t), which describes the relative increase of the top-oil temperature rise
according to the unit of the stead y-state value:

1
f ( ) = (1 −
t e
(− )/(
t 1 1 × τ0 )
k
) (F. 1 )

2) Obtain m easured relative increase of the top-oil tem perature rise as per u nit of the stead y
state value, Mf1 , for the complete test period :
j

M θ o − Mθ a
Mf j1 = j j
(F.2)
∆θ or
3) Perform nonlinear regression by using the guess-error approach (alternativel y curve fitting
or optim ization software also could be used) to find the constant to minimize the sum of
squ ares of differences between f1 and Mf1 :
j j
 – 56 – I EC 60076-7: 201 7 © I EC 201 7

∑ =1 [ 1 ( ) − ]2
minimize :
N

j
f j
z 1
Mf j (F. 3)

where
M is a m easured variable;
Z is a vector whose element is onl y k1 1 thermal time constant;
f
1j is the relative increase of the top-oil tem perature rise as per unit of the stead y state
value as calculated from Equation (F. 1 );
Mf
1j is the measured relative increase of the top-oil temperature rise as per unit of the
stead y state value;
j is the index for each tim e step over the test period with the total losses.
The initial value for the k1 1 constant is taken from Table 4.

The oil time constant value to be set in the function f1 ( t) at the load considered is given in
Annex E.

Sim ilar to the top-oil thermal constant, k1 1 , estimation procedure, the therm al constants k21
and k22 for the hot-spot to top-oil thermal gradient are also obtained from the part of the
tem perature rise test with supplied total losses [32]. I n general, this approach is acceptable
due to fact that the shape of the thermal curve, ∆ θh ( t), Figure 1 4, is not affected by the total
supplied loss level [22], compared to when the unit is supplied with onl y rated load loss.
Alternativel y, a technicall y correct procedure would dem and additional tem perature tests,
which would be run from the cold start until the corresponding stead y state conditions would
be observed at the rated current. The procedure specifies that the transform er winding is to
be equipped with fibre optic sensors.
1 ) Define function f2 ( t), which describes the relative increase of the hot-spot-to-top-oil
gradient according to the unit of the stead y-state value:

2 ( )=
f
t k 21 (
× 1 − e (− ) /(
t 22 × τ w )
k
)− ( k 21 (
− 1 ) × 1 − e (− ) /( τ0 /
t k22 ) ) (F. 4)

2) obtain m easured relative increase of the hot-spot to top-oil gradient, Mf

2 j, as per unit of
the stead y state value for the complete test period:

M θ h − Mθ o
Mf j 2 = j j
(F. 5)
∆θ hr

3) perform nonlinear regression by using guess-error approach (alternativel y curve fitting or

optimization software could be used) to find the constant to minimize the sum of squares
of differences between f2 and Mf2 j j

minimize : ∑ =1 [ 2 ( ) −
N

j
f j
z Mf 2j ]2 (F.6)

where
M is a measured variable;
z is a vector whose elem ents are k21 and k22 constants;
f
2j is the relative increase of the hot-spot to top-oil gradient as per unit of the stead y state
value as calculated from Equation (F. 4);
Mf
2j is the m easured relative increase of the hot-spot to top-oil gradient as per unit of the
stead y state value;
 – 58 – I EC 60076-7: 201 7 © I EC 201 7

Annex G
(informative)
Oil and winding exponents
G.1 General
The traditional oil exponent, x = 0, 8, and winding exponent, y = 1 , 6, have been used since
1 91 6 [33] in the transform er loading calculus. Consequently, the sam e values were adopted
subsequentl y by the I EC 60076-2 [52] and I EC 60354 [4]. The background is further
elaborated in Clause G.2. H owever, these exponents are transform er specific and an y
suggested values should be used onl y in absence of transform er-specific values. I n addition,
it has been found that the recom mended winding exponent of 1 , 6 is quite conservative (see
[23], [24] and [45]), and this has been accounted for in this document. This change is still
included in this edition as given in Table 4.

Clauses G. 2 and G. 3 provide methods used to define the winding and oil exponents for
specific transform er design.

G.2 Historical background

I t was reported by Dulong and Petit in 1 81 7 that the velocity of the cooling due solel y to the
contact of a gas is proportional to the excess of tem perature in degrees centigrade raised to
the power 1 , 233 [33]. Based on this, Swift [22] concluded: “I t is fascinating that the reciprocal
of 1 .223 is approximately 0. 8, a value of x comm onl y used to this day, for natural cooling
conditions! ”.

In 1 881 , Lorenz derived for convection of heat from vertical plane surface the following
relation [33]:

3
= 0,548 × �푐휈××푔퐿××푘휃푡ℎ푚 × 휚 0 ,5 × ∆휃 1 ,25
4
푃 (G. 1 )

where
P is the loss dissipated per cm 2 ,
c is the specific heat of gas at constant pressure,
kth is the gas thermal conductivity,
ν is the gas viscosity,
θm is the gas average tem perature,
ρ is the gas average density,
g is the gravitational constant,
L is the height of plane, and
Δθ is the difference in tem perature of plane surface and of the gas at a great distance from
the plane.
For the given room tem perature, known plate design and for standard atm ospheric pressure,
the equation is further reduced to:

푃 = 퐶 푡푡 × ∆ 휃 1 ,25 (G.2)

The result of the number of tests conducted by Montsinger [33], where most of the heat is
dissipated by convection, was alm ost the sam e:
I EC 60076-7: 201 7 © I EC 201 7 – 59 –

푃 = 퐶 푡푡 × ∆ 휃 1 , 2 4 5 (G.3)

where the constant Cte was d etermined em pirically.

I t was then conclud ed that the convection for vertical surface and within operational range of
the tem peratures can be expressed by Equation (G. 3).

A sim ple eq uation rearrangem ent gives tem peratu re rise variation with loss as follows:

∆휃 = 퐶1 푡 × 푃 1 ⁄1 ,2 5 = 퐶1푡푡 × 푃 0 ,8
푡
(G. 4)

I t was found that Eq uation (G. 4) holds for top-oil rise and for horizontal d isc coils, bu t for the
vertical/layer winding design the correspond ing temperature rise over top-oil varies between
0, 9 and the first power of the loss. N ow the correspond ing temperature rises at an y load can
be found from the following mathem atical expression:
1 ) Top-oil temperature rise:
Based on Eq uation (G. 4), the tem perature rise at rated load is expressed as follows:

∆ 휃표 표 = 퐶1푡,푡표 × 푃표0 ,8 (G.5)

I f then Equation (G. 4) is divided by Equation (G. 5) the following ratio is obtained:
푡푡 0 ,8
∆휃표
∆ 휃표 표
= 퐶퐶1푡 × 푃푃 0 ,8
푡 (G.6)
1 ,표 표

Also, if for given design and operating tem perature range the following is defined:

Cte 1 , r= Cte 1 (G.7)

then Equ ation (G. 6) becom es:
0 ,8
∆ 휃표 = ∆ 휃표 표 × � 푃푃표� (G.8)

I t is also known that P represents the total losses when consid ering the top-oil
tem perature rise, which are equ al to the sum of the load loss, P l , and no-load loss, P fe :

푃 = 푃푙 + 푃푓푡 (G.9)
Further, if the loss ratio is d efined as:

푅 = 푃푃푓푙푡표 표
(G.1 0)

and the following correlations are valid:

푃푓푡표 = 푃푓푡 (G. 1 1 )

퐾2 = 푃푃 푙표 푙
(G.1 2)

the final equation for tem perature rise is:

2 0 ,8
∆ 휃표 = � 1+1+푅 ×푅퐾 � × ∆ 휃표 표 (G. 1 3)
 – 60 – I EC 60076-7: 201 7 © I EC 201 7

This expression is equivalent to the equation suggested by loading guide where “0, 8”
refers to the top-oil exponent “ x ” giving the final equation for the final top-oil temperature
rise at an y load as follows:
푥
= � 1+1+푅 ×푅퐾
2
∆ 휃표 � × ∆ 휃표 표 (G.1 4)

2) The hot-spot to top-oil tem perature rise

By following the same procedure as above for the top-oil temperature rise and knowing
that in this case the loss, P , represents onl y the load loss, P l , the following form ula is
derived:

∆ 휃ℎ푠− 표 = ∆ 휃ℎ푠 − 표 표 × 퐾 1 ,6 (G. 1 5)

Again, this is the same as the general form suggested in the loading guide where “1 , 6”
refers to the winding exponent “ y ”, giving the final equation for the final hot-spot
temperature rise over top-oil tem perature at an y load as follows:

∆ 휃 ℎ푠 − 표 = ∆ 휃 ℎ푠 − 표 표 × 퐾 푦 (G.1 6)

G.3 Theoretical approach

Based on heat transfer theory, the natural convection oil flow around vertical, inclined and
horizontal plates and cyl inders the tem perature rise and the average winding to average oil
gradient can be obtained from the following empirical correlation (see references [46] to [50]):

푁푢 = 퐶 × [퐺표 × 푃표 ] 푛 (G.1 7)

where

푁푢 = ℎ휆× 푙 (G. 1 8)

푃표 = 휌 ×휆휈 × 푐 (G. 1 9)

= 훽 × ∆휃휈× 푔 ×
3
퐺표 2
푙
(G.20)

where
푁푢 , 퐺표 , 푃표 are N usselt, Grashof and Prandtl numbers, respectivel y,

C and n are em pirical constants affected by oil flow,

h is the heat transfer coefficient,
l is the dim ension of the heated surface in the direction of flow,
휆 is the thermal conductivity,
휌 is the oil density,
휈 is the oil kinem atic viscosity,
훽 is the coefficient of thermal cubic expansion,
g is the gravitational constant, and
∆휃 is the corresponding surface drop, i. e. temperature gradient.
By substituting Equations (G. 1 8), (G.1 9) and (G.20) in Equation (G. 1 7), the following
expression is obtained:
I EC 60076-7: 201 7 © I EC 201 7 – 61 –

ℎ× 푙 3 푛
휆
= 퐶 × �휌 ×휆휈 × 푐 × 훽 × ∆휃휈×2 푔 × 푙
� (G.21 )

I f the tem perature grad ient is given as the heat loss to heat coefficient ratio:

∆ 휃 = 푃ℎ (G. 22)

and if Equation (G.21 ) is solved for heat transfer coefficient and substitu ted in Equation
(G.22), the following is obtained:

푃
∆ 휃 = 퐶× 휆 3 푛 (G.23)
푙
× �휌 × × 푐× 훽 × ∆ 휃 ×2푔 × 푙 �
휈

휆 휈

Also, as for the range of the operating tem peratures the ph ysical properties of the oil in
Equ ation (G. 23) can be consid ered constant the following is valid:

∆ 휃 = 퐶1 × 푃 1 /(1+ 푛) (G.24)

where

1
퐶1 = 훽 × 푔 × 푙3 − 푛
푛 (G. 25)
� 퐶 × 휆 푛 −1 × 휌 × 푐 × 휈
�

Equation (G. 24) is identical to Equation (G. 4) and by following the same procedure as given
above, it is possible to arrive at Equ ations (G.1 4) and (G. 1 6) by d efining correlation:

1
1+ 푛
(=) 푥 (G.26)

1
1+ 푛
(=) 푦 (G.27)

Apart of the oil ph ysical param eters the accuracy of the equation d epends also on accuratel y
determ ined C and n em pirical constants. This can be done according to the following
procedure:
1 ) Obtain the hot-spot/average winding and oil tem peratures from the extended tem perature
rise test.
2) Calculate average winding to average oil gradient.
3) Calculate m ean bou ndary layer temperature between the heating surface and coolant:

휃푚 = 휃 ℎ +2 휃표 (G.28)

4) Calculate average value of the heat transfer coefficient:

ℎ = ∆휃
푃
(G. 29)

5) For calculated mean bound ary layer tem perature obtain the ph ysical properties of the oil
and calculate N ussel, Prandtl and Grashof numbers
Plot the following graph in log-log system :

푁푢 = 푓 (퐺표 × 푃표 ) (G. 30)

 – 64 – I EC 60076-7: 201 7 © I EC 201 7

An nex H
(informative)
Practi cal exam pl e of the exponen tial eq uati ons meth od

H.1 G en eral

The curves in Figure 1 4 are taken from an exam ple in real life, and details of the case will be
given in this annex. A 250 MVA, ON AF-cooled transformer was tested as follows. During each
tim e period, the load current was kept constant, that is, the losses changed due to resistance
change during each load step. The corresponding flowchart is in Annex J.

Tabl e H . 1 – Load steps of th e 250 M VA tran sformer

Ti m e peri od Load facto r

m in
0 to 1 90 1 ,0
1 90 to 365 0, 6
365 to 500 1 ,5
500 to 705 0, 3
705 to 730 2, 1
730 to 745 0, 0

The two m ain windings were equipped with eight fibre optic sensors each. The hottest spot
was found in the innerm ost main winding (1 1 8 kV). I n this exam ple the variation of the hottest
spot tem perature during tim e period 0 m in to 745 min will be defined according to the
calculation method described in 8.2. 2. A comparison with the m easured curve will be made.

The characteristic data of the transform er, necessary for the calculation, are:

θ a = 25, 6 °C
∆ θ or = 38, 3 K
R = 1 000 (because the test was m ade by the “short-circuit m ethod”)
H = 1 ,4 (defined by measurem ent, see 8. 1 . 3)
gr = 1 4, 5 K
τ w = 4, 6 m in to 8, 7 min (depending on the loading case. The value in Table 4, that is,
7 m in, will be used in the calculation)
τo = 1 62 min to1 70 min (depending on the loading case. The value in Table 4, that is,
1 50, will be used in the calculation)
The winding is zigzag-cooled with a spacer separation ≥ 3 mm.
I EC 60076-7: 201 7 © I EC 201 7 – 65 –

H.2 Ti me peri od 0 mi n to 1 90 m i n

The initial data for the tim e period 0 m in to 1 90 min are as follows:

∆ θ oi = 1 2, 7 K (This test was started at 08: 20 in the morning. The preceding evening
an overloading test at 1 , 49 p. u. had been finished at 22: 00)
K = 1 ,0
∆ θ h i = 0, 0 K
Equ ations (1 0), (1 2), (1 3), (1 4) and (1 7) yield the hot-spot variation as a function of tim e,
hence from Eq uation (1 0):

 0, 8 

θ o (t ) = 25,6 + 1 2,7 +  38, 3 ×  1 + 1 000 × 1, 0
1 + 1 000
2 
 (
− 1 2, 7  × 1 − e (− )/(0,5 ×1 50 )
t
)
   

From Equation (1 3):

{ }(
∆θ h1 (t ) = 0, 0 + 2, 0 × 1, 4 × 1 4, 5 × 1, 01, 3 − 0, 0 × 1 − e (− ) /( 2, 0 × 7 ) t
)
From Eq uation (1 4):

{ }(
∆θ h 2 (t ) = 0, 0 + (2, 0 − 1, 0 ) × 1, 4 × 1 4, 5 × 1, 01, 3 − 0, 0 × 1 − e (− ) /(1 50 / 2, 0 )
t
)
From Equation (1 2):

∆θ h (t ) = ∆θ h1 (t ) − ∆θ h 2 (t )

From Eq uation (1 7):

θ h (t ) = θ o (t ) + ∆θ h (t )

H.3 Ti me peri od 1 90 m i n to 365 mi n

The initial data for the time period 1 90 m in to 365 m in are as follows:

∆ θ oi = 36, 2 K (calculated in H. 2)
K = 0, 6

∆ θ hi1 = 40, 6 K (calculated in H . 2)

∆ θ hi2 = 1 8, 7 K (calculated in H . 2)
Equations (1 1 ), (1 5), (1 6) and (1 7) yield the hot-spot variation as a function of tim e, hence
from Eq uation (1 1 ):

 1 + 1 000 × 0, 6 2 
0, 8   1 + 1 000 × 0, 6 2 
0, 8 
θ o (t ) = 25, 6 + 38, 3 ×  +  36, 2 − 38, 3 ×   × e (− ) /( 0, 5 ×1 50 )
  
t

 1 + 1 000    1 + 1 000  
 – 66 – I EC 60076-7: 201 7 © I EC 201 7

From Equation (1 5):

{ }
∆θ h1 (t ) = 2, 0 × 1, 4 × 1 4, 5 × 0, 61, 3 + 40, 6 − 2, 0 × 1, 4 × 1 4, 5 × 0, 61, 3 × e (− ) /( 2, 0 × 7 )
t

From Equation (1 6):

{ }
∆θ 2h(t ) = (2, 0 − 1, 0 ) × 1, 4 × 1 4, 5 × 0, 6 31, + 1 8, 6 − (2, 0 − 1, 0 ) × 1, 4 × 1 4, 5 × 0, 6 31, × e (− )
t 50/(×1 50 )

From Equation (1 7):

θ h (t ) = θ o (t ) + ∆θ h (t )

H.4 Ti me peri od 365 m i n to 500 mi n

The initial data for the time period 365 m in to 500 min are as follows:
∆ θ oi = 1 8, 8 K (calculated in H. 3)
K = 1 ,5

∆ θ h i1 = 20, 9 K (calculated in H. 3)
∆ θ h i2 = 1 1 , 25 K (calculated in H. 3)
The calculation is identical to the one in H . 2, when the following replacements are made in
Equation (1 0):

1 2, 7 replaced by 1 8, 8
1 , 0 replaced by 1 , 5

I n Equation (1 3):
0, 0 replaced by 20, 9
1 , 0 replaced by 1 , 5

I n Equation (1 4):
0, 0 replaced by 1 1 , 25
1 , 0 replaced by 1 , 5

H.5 Ti me peri od 500 m i n to 705 mi n

The initial data for the time period 500 m in to 705 m in are as follows:

∆ θ oi 63, 6 K (calculated in H. 4)
K = 0, 3

∆ θ h i1 = 68, 2 K (calculated in H. 4)
∆ θ h i2 = 30, 3 K (calculated in H. 4)
I EC 60076-7: 201 7 © I EC 201 7 – 67 –

The calcu lation is identical to the one in H . 3, when the following replacements are made in
Equation (1 1 ):

36, 2 replaced by 63, 6

0, 6 replaced by 0, 3

I n Equation (1 5):
40, 6 replaced by 68, 2
1 , 0 replaced by 1 , 5

I n Equation (1 6):
1 8, 6 replaced by 30, 3
1 , 0 replaced by 1 , 5

H.6 Ti me peri od 705 m i n to 730 mi n

The initial data for the time period 705 m in to 730 min are as follows:

∆ θ oi = 9, 4 K (calculated in H. 5)
K = 2, 1
∆ θ h i1 = 8, 5 K (calcu lated in H. 5)
∆ θ h i2 = 6, 0 K (calcu lated in H. 5)
The calcu lation is identical to the one in H . 4, when the following replacements are mad e in
Equ ation (1 0):

1 8, 8 replaced by 9, 4
1 , 5 replaced by 2, 1

I n Equation (1 3):
20, 9 replaced by 8, 5
1 , 5 replaced by 2, 1

I n Equation (1 4):
1 1 , 25 replaced by 6, 0
1 , 5 replaced by 2, 1

H.7 Ti me peri od 730 m i n to 745 mi n

The initial data for the time period 730 m in to 745 m in are as follows:

∆ θ oi = 42, 3 K (calculated in H. 6)
K = 0, 0
∆ θ h i1 = 90, 0 K (calcu lated in H. 6)
∆ θ h i2 = 1 9, 4 K (calcu lated in H. 6)
The calculation is m ade in the same way as in H. 3 and H . 5.
I EC 60076-7: 201 7 © I EC 201 7 – 69 –

Table H.2 – Temperatures at the end of each load step

Time (min) / load factor Top-oil temperature Hot-spot temperature
°C °C
Calculated Measured Calculated Measured
1 90 / 1 ,0 61 ,9 58,8 83,8 82,2
365 / 0, 6 44, 4 47, 8 54, 0 58, 6
500 / 1 ,5 89,2 80,8 1 27, 0 1 1 9, 2
705 / 0, 3 35, 0 46, 8 37, 54 49, 8
730 / 2,1 67,9 65,8 1 38, 6 1 40, 7
745 / 0, 0 60, 3 68, 2 75, 3 82, 4
N OTE Bold valu es ind icate load increase.

The calcu lation m ethod in this d ocument is intended to yield relevant values, especiall y at
load increase (noted by bold entries in Table H. 2).
 – 70 – I EC 60076-7: 201 7 © I EC 201 7

An n e x I

(informative)
Ap p l i c a ti o n o f th e d i ffe re n c e e q u a ti o n s o l u ti o n m e th o d

I.1 G e n e ra l

Annex I provides an example of the application of difference equation method described

in 8. 2. 3.

I.2 E xa m p l e

Suppose the obj ective is that an on-line m onitoring device is to generate hot-spot temperature
and loss-of-life inform ation. The steps in the solution are as follows:
1) establish the transformer param eters;
2) establish the input data;
3) calculate the initial conditions;
4) solve the differential equations;
5) tabulate the output data;
6) plot the output data.
The details are as follows.

1 – E s tabl i s h t h e t ra n s fo rm e r p a ra m e t e rs

The parameters used are chosen in such a way that the rated hot-spot temperature is 1 1 0 °C
at an ambient tem perature of 30 °C. Other param eters are typical.

∆ θor = 45 K τ o = 1 50 min R =8 y = 1 ,3 k21 =2

∆ θh r = 35 K τ w = 7 m in x = 0, 8 k1 1 = 0, 5 k22 =2

2 – E s tabl i sh th e i n pu t d ata

The input data for this example are listed in Table I . 1 and plotted in Figure I . 1 .
I EC 60076-7: 201 7 © I EC 201 7 – 71 –

Tabl e I.1 – I n p u t d a t a fo r e x a m p l e

S te p Ti m e Ti m e of d ay Am b i e n t t e m p e ra tu re Load fa c t o r

t θa K

m in h: m in °C
0 0 1 4: 00 30, 3 0, 81
1 3 1 4: 03 29, 9 0, 87
2 6 1 4: 06 29, 8 0, 88
3 9 1 4: 09 29, 5 0, 86
4 12 1 4: 1 2 29, 6 0, 90
5 15 1 4: 1 5 29, 5 0, 92
6 18 1 4: 1 8 29, 5 0, 95
7 21 1 4: 21 28, 9 0, 96
8 24 1 4: 24 29, 0 0, 97
9 27 1 4: 27 28, 6 1 , 00
10 30 1 4: 30 28, 0 1 , 70
11 33 1 4: 33 28, 7 1 , 70
12 36 1 4: 36 27, 8 1 , 73
13 39 1 4: 39 28, 1 1 , 72
14 42 1 4: 42 27, 9 1 , 69
15 45 1 4: 45 27, 1 1 , 68
16 48 1 4: 48 26, 9 1 , 71
17 51 1 4: 51 26, 7 1 , 69
18 54 1 4: 54 27, 2 1 , 67
19 57 1 4: 57 26, 7 1 , 68
20 60 1 5: 00 26, 9 1 , 63
21 63 1 5: 03 26, 5 1 , 59
22 66 1 5: 06 26, 2 1 , 53
23 69 1 5: 09 26, 3 1 , 49
24 72 1 5: 1 2 25, 4 1 , 41
25 75 1 5: 1 5 25, 6 1 , 38
26 78 1 5: 1 8 25, 3 1 , 32
27 81 1 5: 21 24, 8 1 , 28
28 84 1 5: 24 24, 5 1 , 21
29 87 1 5: 27 24, 3 1 ,1 9
30 90 1 5: 30 24, 1 0, 87
31 93 1 5: 33 24, 3 0, 88
32 96 1 5: 36 24, 1 0, 87
33 99 1 5: 39 23, 4 0, 86
34 1 02 1 5: 42 23, 6 0, 85
35 1 05 1 5: 45 23, 8 0, 87
36 1 08 1 5: 48 23, 1 0, 83
37 111 1 5: 51 23, 3 0, 86
38 114 1 5: 54 23, 1 0, 85
39 117 1 5: 57 22, 3 0, 82
40 1 20 1 6: 00 22, 2 0, 86
I EC 60076-7: 201 7 © I EC 201 7 – 73 –

4 – Solve the difference equations

At n = 0, t = 0, θ o(0) = 63, 9 (units are omitted; traditionall y °C for temperatures and K for
temperature differences)
∆ θh1 (0) = 53, 2
∆ θh2(0) = 26, 6
L
(0) = 0
At n = 1 , t = 3 m in, from Equations (1 8) and (1 9), the top-oil temperature changes as follows:

 2 0,8 
D θ o(1 ) = 3   1 + 0,87 × 8  × 45 − [63,9 − 29,9 ] = 0, 1 21 and
0,5 × 1 50  + 81 
   

θ o(1 ) = θ o(0) + Dθ o(1 ) = 63, 9 + 0, 1 21 = 64, 0

Similarl y, from Equation (20), the hot-spot temperature rise first term changes as follows:

D ∆θ h1 (1 ) =
3
2, 0 × 7
( )
2, 0 × 35 × 0, 87 1, 3 − 53, 2 = 1 , 1 2 and

∆θ h1 (1 ) = ∆θ h1 (0) + D∆θ h1 (1 ) = 53, 2 + 1 , 1 2 = 54, 3

Similarl y, from Equation (21 ), the hot-spot temperature rise second term changes as follows:

D ∆θ h2(1 )
3
(1 / 2, 0 ) × 1 50
( )
( 2, 0 1 ) × 35 × 0, 87 1, 3 26, 6 = 0, 1 04 and

∆θ h2(1 ) = ∆θ h2(0) + D ∆θ h2(1 ) = 26, 6 + 0, 1 04 = 26, 7

Then the total hot-spot temperature rise, from Equation (22) is

∆θ h(1 ) = ∆θ h1 (1 ) − ∆θ h2(1 ) = 54, 3 – 26, 7 = 27, 6

and, finall y, the hot-spot temperature is, from Equation (23)

θ h(1 ) = θ 0(1 ) + ∆θ h(1 ) = 64, 0 + 27, 6 = 91 , 6

The loss of life L over this tim e step is given by Equation (25):

 1 5 000 − 1 5 000 
 1 1 0 + 273 θ h(1 ) + 273 
DL (1 ) = V(1 ) × Dt =  e  × 3 = 0, 42 min
 
 

(Loss of life under rated conditions would have been 3 m in. )

The total loss of life to this point is:

L( ) 1 = L( 0 ) + DL (1 ) = 0 + 0, 42 min, or 0, 000 29 days.

 
 – 74 – I EC 60076-7: 201 7 © I EC 201 7

At n = 2, t = 6 m in, the entire calculation is repeated, with all subscripts increm ented by 1 , that
is, each variable X(1 ) becom es X(2) . At n = 3, t = 9 min, each variable X(2) becom es X(3) and
so on. Continue until n = 40, t = 1 20 min.

5 – Tabu l ate th e o u tp u t d ata

The results of the calculation are shown in Table I . 2 and Figure I . 2.

Tabl e I.2 – O u tp u t d ata fo r t h e e x a m p l e

S te p Ti m e Ti m e of d a y H o t -s p o t Loss o f l i fe Loss o f l i fe

t t e m p e r a t u re θh L L

m in h: m in °C m in days
0 0 1 4: 00 90, 5 0 0
1 3 1 4: 03 91 , 6 0 0, 00
2 6 1 4: 06 92, 7 1 0, 00
3 9 1 4: 09 93, 2 1 0, 00
4 12 1 4: 1 2 94, 3 2 0, 00
5 15 1 4: 1 5 95, 6 3 0, 00
6 18 1 4: 1 8 97, 2 3 0, 00
7 21 1 4: 21 98, 6 4 0, 00
8 24 1 4: 24 1 00, 0 5 0, 00
9 27 1 4: 27 1 01 , 6 7 0, 00
10 30 1 4: 30 1 1 8, 6 14 0, 01
11 33 1 4: 33 1 32, 1 39 0, 03
12 36 1 4: 36 1 43, 5 1 09 0, 08
13 39 1 4: 39 1 52, 4 258 0, 1 8
14 42 1 4: 42 1 58, 8 508 0, 35
15 45 1 4: 45 1 63, 6 875 0, 61
16 48 1 4: 48 1 68, 2 1 402 0, 97
17 51 1 4: 51 1 71 , 5 2 076 1 , 44
18 54 1 4: 54 1 73, 6 2 871 1 , 99
19 57 1 4: 57 1 75, 7 3 796 2, 64
20 60 1 5: 00 1 76, 1 4 754 3, 30
21 63 1 5: 03 1 75, 6 5 675 3, 94
22 66 1 5: 06 1 73, 8 6 480 4, 50
23 69 1 5: 09 1 71 , 5 7 1 56 4, 97
24 72 1 5: 1 2 1 67, 8 7 667 5, 32
25 75 1 5: 1 5 1 64, 3 8 055 5, 59
26 78 1 5: 1 8 1 60, 1 8 335 5, 79
27 81 1 5: 21 1 56, 0 8 534 5, 93
28 84 1 5: 24 1 51 , 1 8 668 6, 02
29 87 1 5: 27 1 46, 8 8 761 6, 08
30 90 1 5: 30 1 36, 9 8 800 6, 1 1
31 93 1 5: 33 1 29, 1 8 81 9 6, 1 2
32 96 1 5: 36 1 22, 8 8 830 6, 1 3
33 99 1 5: 39 1 1 7, 5 8 836 6, 1 4
 – 78 – I EC 60076-7: 201 7 © I EC 201 7

Annex K
(informative)
Example of calculating and presenting overload data
Annex K contains an example of how to calculate and present the overload data with the
equations presented in this document.

Table K. 1 gives, as an example, som e characteristics that might be used.

Table K.1 – Example characteristics related to the loadability of transformers

Small Large and m edium power transformers
Characteri sti c tran sform ers
ON AN ON AN ON AF OF OD
Oil exponent x 0, 8 0, 8 0, 8 1 ,0 1 ,0
Winding expon ent y 1 ,6 1 ,3 1 ,3 1 ,3 2, 0
Loss ratio R 5 6 6 6 6
H ot-spot factor H 1 ,1 1 ,3 1 ,3 1 ,3 1 ,3
Oil tim e constant τo 1 80 21 0 1 50 90 90
Wi nding tim e constant τw 4 10 7 7 7
Am bi ent tem perature θa 20 20 20 20 20
H ot-spot tem perature θh 98 98 98 98 98
H ot-spot to top-oi l (in tank) gradient 23 26 26 22 29
at rated current ∆ θh r

Averag e oil tem peratu re rise a ∆ θ omr 44 43 43 46 46

Top-oil (in tank) tem perature ri se ∆ θor 55 52 52 56 49
Bottom oil tem perature ri se a ∆ θ br 33 34 34 36 43
k1 1 1 ,0 0, 5 0, 5 1 ,0 1 ,0
k21 1 ,0 2, 0 2, 0 1 ,3 1 ,0
k22 2, 0 2, 0 2, 0 1 ,0 1 ,0
a Averag e oil tem peratu re rise and bottom oil tem peratu re rise are given for inform ation onl y.

With a spreadsheet programme, a 24 h period is created, with the time-scale in m inutes.

Equations (1 0) to (1 7) are used to calculate for each m inute the hot-spot tem perature as a
function of the load. The initial conditions for ∆ θoi and ∆ θh i1 , ∆ θhi 2 can be determined with
t ∞.
→

When the hot-spot tem perature is known, the relative ageing can be calculated with
Equation (2). With Equation (4) the loss of life, expressed in “norm al” days, can be calculated
by dividing the sum of the relative ageing of each m inute by 1 440.  

For exam ple, consider a case with a pre-load ( K1 ) of 0, 8, then an overload K2 = 1 , 4 during
30 m in and return to K1 = 0, 8 for the tim e remaining (1 41 0 min). The transform er is OF
 

cooled; therefore, the exam ple characteristics of Table H .1 (OF) are used.
I EC 60076-7: 201 7 © I EC 201 7 – 79 –

The initial valu es, after a stead y state pre-load are:

K
1 = 0, 8
∆ θoi = 38, 7 K
∆ θhi 1 = 21 , 4 K
∆ θhi 1 = 4, 95 K
The values after t = 30 min from start:
K
2 = 1 ,4
θo ( t = 30) = 76, 7 °C
θh ( t = 30) = 1 1 4, 2 °C

The valu es after t = 31 min from start:

K
1 = 0, 8
θo ( t = 1 ) = 76, 5 °C
θh ( t = 1 ) = 1 1 1 °C

The valu es after t = 1 440 min from start:

 

K
1 = 0, 8
θo ( t = 1 41 0) = 58, 7 °C
   

θh ( t = 1 41 0) = 75, 2 °C
 

This results in a loss of life of 0, 1 4 d ays and a maxim um hot-spot tem perature rise of 94 K.

The param eters used in the described method can be varied to obtain a table with the loss of
life as a function of K1 and K2 . When the overload tim e is changed, a com plete set of tables
can be obtained.

As an example, one table with an overload tim e of 30 m in is presented in Table K. 2.

I EC 60076-7: 201 7 © I EC 201 7 – 83 –

vend or should declare the behaviour of the transform er design under DC, if the unit is
exposed to GI C in the power grid. To evaluate the thermal risk of the transformer design, the
temperature lim its of short and long tim e em ergency loading in Table 2 can be used or if
specified by customer d ifferent temperature lim its can be applicable.
 – 84 – I EC 60076-7: 201 7 © I EC 201 7

Annex M
(informative)
Alternative oils
Synthetic esters and natural esters have been used for some years now in power
transform ers [56] to [59]. There has been an attempt to use these oils at increasing voltages
and power ratings, and there are some exam ples of these oils being used in high voltage and
high power transformers. CI GRE Brochure 436 includes an overview of how these liquids are
currentl y used, their properties, considerations that impinge on transformer designing and
manufacturing, testing regimes, handling precautions and reliability [60]; however, this
brochure was written in 201 0 j ust as the m arket for alternative liquids was devel oping. Since
the CI GRE Brochure 436 was published, I EC 62770:201 3 [61 ] – a new standard for unused
natural ester liquids for transform ers – has been published. I EC 60076-1 4 further discusses
the application of these oils in transform er design and consideri ng among others their thermal
characteristics and permissible temperatures, including suggested thermal limits when
operated above nam eplate rating. Moreover, a 50 M VA com mercial transformer designed for
mineral oil was comparativel y tested by filling first with natural ester and then with mineral
oil [62]. Some other ageing studies have been reported in references [63] to [65] and the
corresponding bubbling phenom enon was presented in [66]. While inconclusive, these tests
demonstrate the difference in the therm al perform ance that can be expected with these oils
com pared to m ineral oil. Consequentl y, introduction of these oils in this docum ent requires
further continuous research activities to give reliable and prompt answers to their effect on the
transform er design and service life.
I EC 60076-7: 201 7 © I EC 201 7 – 85 –

Bibliography
[1 ] C57. 1 00-201 1 – IEEE Standard Test Procedure for Thermal Evaluation of Insulation
Systems for Liquid-Immersed Distribution and Power Transformers
[2] ASTM D982 – 16 Standard Test Method for Organic Nitrogen in Paper and Paperboard

[3] I EC 6021 4-1 : 201 4, Tap-changers – Part 1: Performance requirements and test
methods
[4] I EC 60354: 1 991 , Loading guide for oil-immersed power transformers 2

[5] I EC 60076-1 : 201 1 , Power transformers – Part 1: General

[6] A. K. Bose and G. J H ulsink: Performance of the magnetic circuit of a transformer in

service – Elektrotechniek 56 (1 978)(7)
[7] E. Binder und W. Felber: Rueckarbeitsverfahren mit zwei 500 MVA Transformatoren –
ELI N-Zeitschrift 1 978 Heft 1

[8] A. K. Bose, C. Kroon, J. Wildeboer: The loading of a magnetic circuit – CI GRE 1 978
session, August 30 – Septem ber 7

[9] A. M . Emsley, X. Xiao, R. J . H eywood, and M. Ali: “Degradation of cellulosic insu lation
in power transformers. Part 3: Effects of oxygen and water on ageing in oil”,
I EE Proceedings – Science, M easurement and Technolog y, vol. 1 47, pp. 1 1 5-1 1 9,
2000

[1 0] L. E. Lu ndgaard, W. Hansen, S. I ngebrigtsen, D. Linhjell, and M. Dahlund: “Aging of

kraft paper by acid catal yzed h yd rol ysis”, in I EEE I nternational Conference on
Dielectric Liquids, 2005. I CDL 2005. pp. 381 -384

[1 1 ] CI GRE TF D1 . 01 .1 0 (L. Lundgaard): “Ageing of cellulose in mineral oil insulated

transformers”, CI GRE Brochure N o. 323, 2007

[1 2] L. E. Lu ndgaard , W. Hansen, and S. I ngebrigtsen: “Ageing of m ineral oil I m pregnated

cellulose by acid catal ysis”, I EEE Transactions on Dielectrics and Electrical I nsu lation,
vol. 1 5, pp. 540-546, 2008

[1 3] CI GRE WG A2. 24 (J . Declerec): Therm al perform ance of power transform ers, CI GRE
Brochure N o. 393, 201 0

[1 4] N. Lelekakis, D. Martin, and J . Wijaya, “Ageing rate of paper insu lation used in power
transform ers Part 1 : Oil/paper system with low oxygen concentration”, I EEE
Transactions on Dielectrics and Electrical I nsulation, vol. 1 9, pp. 1 999-2008, 201 2

[1 5] N. Lelekakis, D. Martin, and J . Wijaya, “Ageing rate of paper insu lation used in power
transform ers Part 2: Oil/paper system with medium and high oxygen concentration”,
I EEE Transactions on Dielectrics and Electrical I nsulation, vol. 1 9, pp. 2009-201 8,
201 2

______________
2 Withdrawn. I EC 60354: 1 991 was cancelled an d replaced by I EC 60076-7: 2005.
 – 86 – I EC 60076-7: 201 7 © I EC 201 7

[1 6] Lam pe, W., Petterson, L., Ovren, C. , Wahlstrom , B . , Hot-spot measurements in power
transformers – CI GRE, Rep. 1 2-02, 1 984

[1 7] Nordman, H. and Lahtinen, M., Thermal overload tests on a 400 MVA power
transformer with a special 2, 5 p. u. short time loading capability – I EEE Transactions on
Power Delivery, vol. 1 8, no. 1 , pp. 1 07–1 1 2, 2003

[1 8] CI GRE WG A2. 38 (J. Lapworth): Transformer thermal modelling, CI GRE Brochure

No. 659, 201 6

[1 9] Experimental determination of power transformer hot-spot factor – CI GRE WG 1 2-09,

Electra no. 1 61 , August 1 995

[20] I EEE 1 538: 2000, IEEE Guide for determination of maximum winding temperature rise
in liquid-filled transformers

[21 ] Nordman, H. , Rafsback, N. , and Susa, D.: Temperature responses to step changes in
the load current of power transformers – I EEE Transactions on Power Delivery, vol. 1 8,
no. 4, pp. 1 1 1 0-1 1 1 7, 2003

[22] Swift, G.W., Molinski, T. S. and Lehn, W. : A Fundamental Approach to Transformer

Thermal Modelling – I EEE transactions on power delivery, vol. 1 6, N o. 2, pp. 1 71 –1 77,
2001 ,

[23] Heat-run test procedure for power transformers – CI GRE WG 1 2-09, Electra no.1 29,
March 1 990, pp. 37–45

[24] Nordman H ., Takala O. : Transform er loadability based on directl y measured hot-spot

tem perature and loss and load current correction exponents, A2_307_201 0, CI GRE,
Paris, 201 0

[25] Montsinger V. M. : The loading transform er with temperature, Winter Convention of the
A. I . E. E. , New York, USA, Jan. 27-31 , 1 930, p. 783

[26] T.W. Dakin: Electrical I nsulation Deterioration treated as a chem ical rate
phenomenon”, AI EE Trans., Vol. 67, pp. 1 1 3-1 22, 1 948

[27] I EEE Std C57. 91 : 201 1 , I EEE Guide for Loading Mineral-Oil-I m mersed Transform ers
and Step-Voltage Regulators

[28] Susa D. , Liland K. B. , Lundgaard L. , and Vårdal G. : Generator step-up transform er post
mortem assessment, European Transactions on Electrical Power, Vol. 21 , I ssue 5,
pp. 1 802–1 822, 201 1

[29] Oomm en T. V. , Ronnau R. A. , and Girgis R. S.: N ew m echanism of m oderate hydrogen

gas generation in oil-filled transform ers, CI GRE Conference Paper 1 2-206, Paris
meeting, Aug-Sep 1 998

[30] teN yenhuis E.G. , Girgis R. S. , Mechler G.F. , Zhou G.: Calculation of core hot-spot
tem perature in power and distribution transformers, I EEE Transactions on Power
Delivery, Vol. 1 7, I ssue:4, pp. 991 –995, 2002

[31 ] Cooney, W.H.: Predeterm ination of self-cooled oil-immersed transformer temperatures

before conditions are constant, Presented at the regional meeting of dist. N o. 1 ,
Swam pscott, M A, U SA, May 7-9, 1 925
I EC 60076-7: 201 7 © I EC 201 7 – 87 –

[32] Susa, D. and N ordman H. : I EC 60076-7 loading guide thermal m odel constants
estimation, I nternational Transactions Electrical Energ y Systems, Vol. 23, I ssue 7,
pp. 946–960, 201 3

[33] Montsinger V. M.: “Effect of barometric pressure on temperature rise of self-cooled

stationary ind uction apparatus”, A. I . E. E. Trans., Vol. 35, p. 599, 1 91 6

[34] Montsinger V.M.: “Cooling of transform er windings after shut-down”, A. I . E. E. Trans.,

Vol. 36, p. 71 1 , 1 91 7

[35] Montsinger V. M ., and W.H. Cooney: “Tem perature rise of stationary electrical
apparatus as influenced by rad iation, convection, and altitud e”, A.I . E. E. Trans. ,
Vol. 43, p. 803, 1 924

[36] Montsinger V. M. , and Wetherill L. : “Effect of color of tank on the temperature of

self-cooled transform ers under service conditions”, A.I . E. E. Trans. , Vol. 49, p. 41 ,
1 930

[37] Montsinger V. M. : Loading transformer by tem perature, A. I . E. E. Trans. , Vol. 49, p. 776,
1 930

[38] Montsinger V. M., and W.M. Dann: “Overloading of power transformers”, Electrical
Engineering, Vol. 53, I ssue 1 0, p. 1 353, 1 934

[39] Montsinger V. M. , and Ketchum M. P. : “Emergency overloading of air-cooled

transform ers by hot-spot tem perature”, Electrical Engineering, Vol. 61 , I ssue 1 2,
p. 906, 1 942

[40] Susa D. , Lehtonen M. , and N ordm an H .: “Dynamic thermal modelling of power

transformers”, I EEE Transactions on Power Delivery, Vol. 20, I ssu e 1 , pp. 1 97–204,
2005

[41 ] Susa D., Lehtonen M. : “Dynam ic therm al mod elling of power transformers – further
development: Part I ”, I EEE Transactions on Power Delivery, Vol. 21 , I ssue 4,
pp. 1 961 -1 970, 2006

[42] Susa D., Lehtonen M. : “Dynam ic thermal m odelling of power transformers – further
developm ent: Part I I ”, I EEE Transactions on Power Delivery, Vol. 21 , I ssu e 4,
pp. 1 971 -1 980, 2006

[43] Susa D., Lehtonen M. , and N ordman H. : Dynamic therm al mod elling of distribu tion
transform ers, I EEE Transactions on Power Delivery, Vol. 20, I ssue 4, pp. 1 91 9–1 929,
2005

[44] Susa D., Palola J ., Lehtonen M ., and H yvarinen M.: Temperature Rises in an OFAF
Transform er at OFAN Cooling Mode in Service, I EEE Transactions on Power Delivery,
Vol. 1 8, I ssue 4, pp. 1 1 1 0–1 1 1 7, 2003

[45] CI GRE SC 1 2 (T. H urter, F. Viale): Therm al aspects of large transformers, test
procedures, hot-spot id entification, permissible lim its, their assessm ent in factory tests
and service, overload lim itation, effects of cooling system , 1 984 CI GRE Session 1 2-1 3

[46] I ncropera F. P., and DeWitt D. P. : “Fundamentals of Heat and Mass Transfer”, 4th ed. ,
J ohn Wiley & Sons, 1 996, 886 pp.

[47] Karsai K. , Keren yi D. , and Kiss L. : “Large Power Transformers”, Elsevier Science
Publishers, 1 987, 607 pp.
 – 88 – I EC 60076-7: 201 7 © I EC 201 7

[48] King W. J. , “The basic Laws and Data of Heat Transmission”, Mechanical Engineering,
March-August 1 932, p. 1 91

[49] Rice W. Chester, “Free and Forced Convection of Heat in Gases and Liquids”, Physical
Review, Vol. 21 , April 1 923 and A. I . E. E. Trans. , 1 923

[50] Rice W. Chester, “Free Convection of Heat in Gases and Liquids-I I ”, A.I . E. E. Trans. ,
Vol. XLI I I , p. 1 31 , 1 924

[51 ] I EEE Std C57. 1 1 9-2001 , IEEE recommended practice for performing temperature rise
tests on oil-immersed power transformers at loads beyond nameplate ratings
[52] I EC 60076-2: 1 993, Power transformers – Part 2: Temperature rise 3

[53] R. Pirj ola: Geomagneticall y induced currents in electric transmission networks at

different latitudes, EMC symposium , China 201 0

[54] J. Raith: Risk evaluation for power transformers during solar storms, A2-PS1 CI GRE
Zürich 201 3

[55] J. Raith: GI C strength verification of power transform ers in a high voltage laboratory,
GI C Workshop Cape Town, April 201 4

[56] K.J . Rapp, J . Luksich, A. Sbravati: “Application of natural ester insulating liquids in
power transformers”, Proc. M y Transfo conference, 1 8-1 9 Novem ber 201 4, Turin, I tal y

[57] D. Bingenheim er, K. J . Rapp, J. Luksich, A. Sbravati: “Application of Natural Ester

I nsulating Liquids in Power Transformers”, Proc. TechCon® N orth America,
23-25 February 201 6, Albuquerque, N M, U SA

[58] G.J. Pukel, G. Fleck, H . Pregartner, R. Fritsche, “Safe and environm entally friendl y
large power transform ers with ester, successful introduction of ester liquids at the
420 kV transmission level”, I EEE Electrical I nsulation Conference (EI C), 1 9-22 J une
201 6, M ontréal, Canada, p. 1 34–1 37, I SBN 978-1 -4673-8706-4

[59] Ronn y Fritsche, U we Rimm ele, Frank Trautm ann*, and Michael Schäfer: Prototype
420 kV Power Transformer using natural ester dielectric fluid, Proc. TechCon® N orth
America, 3-6 February 201 4, Phoenix, AZ, U SA

[60] CI GRE WG A2.35 (R. Martin): Thermal perform ance of power transform ers, CI GRE
Brochure N o. 436, 201 0

[61 ] I EC 62770: 201 3, Fluids for electrotechnical applications – Unused natural esters for
transformers and similar electrical equipment
[62] R. Girgis, M. Bernesjö, G. K. Frim pong: Detailed perform ance of a 50 M VA transformer
filled with a natural ester fluid versus mineral oil, A2. 1 07, CI GRE 201 0

[63] C. P. McShane, J. L. Corkran, K.J. Rapp, J. Luksich: N atural ester dielectric fluid
development update, I EEE Power & Energ y Society General Meeting, 2009. PES '09

[64] Gasser, H ans Peter et al: Aging of Pressboard in Different I nsulating Liquids. I EEE
I nternational Conference on Dielectric Liquids, Trondheim , N orway, June 201 1
______________
3 Withdrawn.
I EC 60076-7: 201 7 © I EC 201 7 – 89 –

[65] C. P. McShane, J. L. Corkran, K. J . Rapp, J . Luksich: Aging of paper insulation retrofilled

with natural ester dielectric fluid, Annual Report Conference on Electrical I nsulation
and Dielectric Phenomena, 1 9-22 October 2003, Albu querque, N M, U SA

[66] C. Y. Perkasa, N . Lelekakis, T. Czaszejko, J. Wijaya, and D. Martin: A comparison of

the form ation of bubbles and water droplets in veg etable and m ineral oil im pregnated
transform er paper, I EEE Transactions on Dielectrics and Electrical I nsulation,
Volume 21 , I ssue 5, p. 21 1 1 , 201 4

[67] I EC 60076-5, Power transformers – Part 5: Ability to withstand short circuit

___________
INTERNATIONAL
ELECTROTECHNICAL
COMMISSI ON

3, rue de Varembé
PO Box 1 31
CH-1 21 1 Geneva 20
Switzerland

Tel: + 41 22 91 9 02 1 1
Fax: + 41 22 91 9 03 00
info@iec.ch
www.iec.ch