ieee-deis (jun2006) electrical breakdown in hv winding insulations
TRANSCRIPT
7/23/2019 IEEE-DeIS (Jun2006) Electrical Breakdown in HV Winding Insulations
http://slidepdf.com/reader/full/ieee-deis-jun2006-electrical-breakdown-in-hv-winding-insulations 1/9
„Electrical Breakdown in High-Voltage Winding Insulations of DifferentManufacturing Qualities “
R. Vogelsang,andT. Weiers,andK. Fröhlich,
andR. Brütsch.
Important noticeCopyright and all rights in this work are retained by the authors.
This material may not be reposted without the explicit permission of the authors.
©2005 IEEE. Personal use of this material is permitted. However, permission toreprint/republish this material for advertising or promotional purposes or for creatingnew collective works for resale or redistribution to servers or lists, or to reuse anycopyrighted component of this work in other works must be obtained from the IEEE.
7/23/2019 IEEE-DeIS (Jun2006) Electrical Breakdown in HV Winding Insulations
http://slidepdf.com/reader/full/ieee-deis-jun2006-electrical-breakdown-in-hv-winding-insulations 2/9
May/June 2006 — Vol. 22, No. 3
F E A T U R E A R T I C L E F E A T U R E A R T I C L E F E A T U R E A R T I C L E F E A T U R E A R T I C L E F E A T U R E A R T I C L E
D
The results show that electrical tree
propagation is the main electrical
degradation mechanism that leads to
breakdown of the main wallinsulation, and that poor
impregnation of the insulation and
type of taping can reduce time to
breakdown significantly.
Introductionue to an increasing demand for new, or the refurbish-
ment of, power stations, high-voltage rotating machines
continue to play a significant role in generating electrical en-
ergy. An important component of these machines is their electri-
cal insulation. Although much research has been done to im-
prove the material, about a quarter of all failures still are related
to insulation problems [1] – [3]. The state of the art of winding
insulation is a composite of mica, binder resin, and a support
material, the so called mica tapes. A peculiarity of the insulation
is that the tapes are intermediate products, which usually are
produced by a different company than the manufacturer of the
final insulation or the entire machine. Such a discontinuity in
the production process results in many different types of insula-
tion systems and a variation in manufacturing quality [4] – [6].
Research to improve winding insulation is very complex be-
cause a large number of factors influence insulation life, among
them the raw materials, size of the mica tapes, geometrical char-
acteristics of the bar and taping- or manufacturing techniques
[6]. Additionally, the different service stress factors—electrical,
thermal, mechanical, and ambient load—also influence the life-
time of the insulations [6] – [10]. Although electrical degrada-
tion and breakdown in polymers have been studied extensively,the process that leads to failure in the composite structure of
industrially made winding insulations has not yet been conclu-
sively described. Especially, the influence of manufacturing qual-
ity is not fully understood. This is mainly because reliable re-
search on industrially made winding insulation is complex due
to the already-mentioned large number of influencing factors.
Additionally, test samples are expensive and their number often
is very limited, particularly when they come directly from the
production line of different companies.
Electrical Breakdown in High-Voltage
Winding Insulations of Different
Manufacturing Qualities
Key Words: high-voltage rotating machine insulation, ground (main) wall insulation, VPI,resin-rich, machine-taped insulation, hand-taped insulation, electrical treeing, time to
break down
R. Vogelsang, T. Weiers, K. Fröhlich,
and R. Brütsch
In order to provide a basis for material improvements, the first
aim of this work is to describe the influence of manufacturingquality on time to breakdown of high-voltage winding insula-
tion. Because tests accredited under IEEE Standard 1043 re-
quire much effort and are very expensive, a second aim of the
work is to use the knowledge about the main electrical degrada-
tion mechanism to develop a method that allows testing of such
insulation systems with less effort. The new test method should
be developed to allow investigation of industrially manufac-
tured insulation systems from the production line and should be
easily usable by industrial companies.
7/23/2019 IEEE-DeIS (Jun2006) Electrical Breakdown in HV Winding Insulations
http://slidepdf.com/reader/full/ieee-deis-jun2006-electrical-breakdown-in-hv-winding-insulations 3/9
6 IEEE Electrical Insulation Magazine
Structure and Manufacture of High-Voltage
Winding InsulationsThe main wall insulation for high-voltage (hv) rotating ma-
chines consists of three basic components—a barrier that is resis-
tant to partial discharges and tree growth, a support material that
gives mechanical strength, and a binder resin that fills any voids
between the mica and the support material. These materials are
produced as tapes and are wrapped around the surface of the
conductor to form the insulation [4], [5]. The wrapping process is
made either by hand or by machine as shown in Figures 1 and 2,respectively.
When the bar is fully covered by the tape, any spaces between
the mica tapes are filled with binder resin in order to reduce air
inclusions. For this purpose, two main technologies, resin-rich
and vacuum-pressure impregnation (VPI) are used [4], [5]. When
the insulation has been formed, the final bar is finished with a
corona protection for the slot- and end-winding coatings [4], [7].
Resin-rich bars are pressed under high temperatures and high
pressure. Therefore, the bars have an almost rectangular shape, a
tight arrangement of the mica tapes, and the whole cross section
of the insulation is impregnated with similar quality [4], [5]. A
disadvantage of the resin-rich process is the machine effort, which
is high especially for bent end-winding sections [4].
For VPI insulations, the whole bar is first set under vacuum,
followed by impregnation with resin under high pressure. An
advantage of VPI insulations is that the mica tapes are not sharply
bent during the pressure process. In addition, cavities at the bars
also are filled, especially when applying global VPI, in which
the whole machine is impregnated and cavities between the bars
and the slots of the machine also are filled with resin. Therefore,
the VPI-process is very suitable to impregnate bars with a com-
plex form. However, the resin may not impregnate all voids in
the insulation, and resin bubbles or voids are likely to form cracks
as the resin shrinks during cure. A disadvantage of the global VPI
process is its requirement for large and expensive equipment [4],[5]. After forming the insulation, the final bar is finished with a
corona protection for the slot- and end-winding coatings [4], [7].
The internal structure of the winding insulation with half-lap
tape wrapping and conductive coating is illustrated in Figure 3.
The combination of mica tapes and impregnation determine
the properties of the insulation [4] – [7], [11]. As a result of the
many combinations of mica, support material, resin and impreg-
nation, there is a great variety of mica tapes available on the
market [4] – [7].
Determining Tree Propagation by a TestArrangement with an Embedded Electrode
A. BasicsWhen voltage is applied, the breakdown process can be di-
vided into “tree inception” and “tree propagation” [6], [8], [12] –
[16]. As shown in Eqn. (1), the time to breakdown, Tbd
, depends
on the time interval for tree inception, Tti, and tree propagation,
Ttp:
T bd
= T ti + T
tp(1)
Electrical treeing is a process by which the insulation mate-
rial is eroded by discharges taking place over a long period [13].
However, the start of the erosion begins with inception of dis-
charges in the bulk, at the inner conductor or at the surface of theinsulation [7] – [10],[17]. Generally, it can be said that the result-
ing time to breakdown of the insulation is long when tree incep-
tion and tree propagation time intervals are long, and it is short
when both intervals are short. In a hv winding insulation, tree
inception is mainly influenced by the coating or the conductor,
and tree propagation is mainly dependant on the insulation ma-
terial.
B. Measurement of Tree Propagation by a Test
Arrangement with an Embedded ElectrodeAs outlined theoretically, the tree propagation time repre-
sents a measure for the material’s insulation resistance againsttreeing. Therefore, the insulation properties can be determined
by merely measuring the time interval a tree needs to propagate
through the material. In order to determine that value, an ar-
rangement with an embedded electrode was developed, which
causes tree inception immediately after voltage application. The
electrode is a copper sheet of 20 x 20 mm2, with a thickness of 0.2
mm, which was embedded in the insulation by the insulation
manufacturer. In contrast to former arrangements [18], the em-
bedded eledctrode was set directly in the insulation in order to
Figure 3. Structure of a winding insulation.
Figure 1. Mica tape
wrapping by hand [5].
Figure 2. Mica tape wrapping
by machine [5].
7/23/2019 IEEE-DeIS (Jun2006) Electrical Breakdown in HV Winding Insulations
http://slidepdf.com/reader/full/ieee-deis-jun2006-electrical-breakdown-in-hv-winding-insulations 4/9
May/June 2006 — Vol. 22, No. 3
avoid surface discharges. The principle design of the arrange-
ment is illustrated in Figure 4, and an actual cross-section is
shown Figure 5. The advantages of the arrangement with an em-
bedded electrode are that no surface discharges occur and that
the insulation properties are tested directly [6].
The edges of the embedded copper sheet have a radius of <10
µm, which ensures that the tree incepts immediately (Tti < 30 s)
after voltage application. Because the time to tree inception is
practically zero, the total measured time to breakdown is the
time for the tree to propagate through the material. Therefore,
such tests are referred to as “Electrical treeing tests” [6].
For all tests, insulation bars with a length of 1 m and a con-
ductor cross section of 8 x 39 mm2 were used. In each bar, four
electrodes were uniformly embedded into the insulation. In or-
der not to propagate in neighboring tree channels, the electrodes
were placed >100 mm from the end of the bar and > 200 mm
separation from each other [6].
C. Reference to Conventional Tests and Degradation Mechanism in the Insulation
To give a reference to current test standards, the results of the
electrical treeing tests were compared to results with the same
type of bars prepared according to IEEE Standard 1043 [19],
[20]. Tests according to the standards usually are made with three
times the rated voltage, and they are referred to as “3Un-tests”.
Samples for the 3Un-tests were prepared with outer conductive-
and field-grading coating [6], [11]. In the tests, three different
insulation systems were chosen. For better comparability, the
size of the insulation bar, insulation thickness of 2 mm, tempera-
ture of 20 ± 5°C, and voltage of 32 kV rms were kept the same for
each insulation system. There were 30 test specimens for the
3Un-tests, and 12 for the electrical treeing tests. Detailed infor-
mation about the test setup is given in [6], [8].
Figure 6 compares the Weibull-plots for the times to break-
down of the electrical treeing tests and the 3Un-tests for the
various materials. The results show that, for all three materials,
the times to breakdown for electrical treeing and 3Un-tests are
not significantly different as the 95% confidence intervals of the
63% quantiles overlap. However, there is significant variationbetween the three materials. The longest times to breakdown
were measured with material 3, followed by material 2 and mate-
rial 1. The higher scatter in the treeing tests is explained by the
lower number of samples. The significant differences between
the times to breakdown of the three materials show that the tree
propagation is different for each insulation. Because they vary
in their material properties, it can be concluded that the treeing
test indicates significant differences in the insulation properties.
Because the tree inception time in the treeing tests is zero,
there is strong evidence that electrical tree propagation is a main
determining degradation process in the high voltage winding
insulation. Therefore, it must be considered that, from the begin-
ning of voltage application, tree propagation takes place in test
specimens, even in the bars that were prepared with an outer
coating. When treeing takes place, many channels propagate
through the material [6], [11]. Because the tree cannot easily
penetrate an intact mica tape, it mainly propagates along them
and to the next layer at regions of tape overlaps, which is shown
in Figure 7.
For application of the test, it must be mentioned that the elec-
trodes usually cannot be embedded during wrapping as described
previously. For already manufactured bars, the electrodes must
Figure 4. Principle of sample for tests with the embedded,
high-voltage electrode.
Figure 5. Photograph of specimen with embedded, high-
voltage electrode.
Figure 6. Weibull-plots for the times to breakdown for 3Un,
and treeing tests for three different winding insulations.
7/23/2019 IEEE-DeIS (Jun2006) Electrical Breakdown in HV Winding Insulations
http://slidepdf.com/reader/full/ieee-deis-jun2006-electrical-breakdown-in-hv-winding-insulations 5/9
8 IEEE Electrical Insulation Magazine
be embedded afterward. Therefore, a method was developed to
place the copper sheet in the insulation on a manufactured bar.
Because this is very relevant, tests with subsequently-introduced
electrodes at the High Voltage Laboratory of the ETH Zürich
were carried out. The times to breakdown show no significant
difference to those shown in Figure 6 [6]. Therefore, it is also
possible to implement the electrodes in already manufactured
bars.
Furthermore, treeing tests give a detailed answer for material
properties of the main wall insulation. They are not suited for
tests considering the entire bar with slot and endwinding relief
coatings as described in the IEEE Standard 1043 (such as cus-
tomer acceptance tests).
The advantages of treeing tests compared to conventional
tests are that they require much less material because fewer bars
are needed, they represent a material test because tree propaga-
tion through the insulation is measured, and that they allow
investigation of particular regions of the bar, such as end-wind-
ings, bends, or straight sections separately. Electrical treeing tests,
therefore, are recommended to complement conventional tests
in which only a low number of samples is accessible, such as for
random tests in the production line; determination of residual
dielectric strength of bars after service; investigating the ten-dency of some influencing factors in research laboratories; fin-
gerprint-tests for newly developed material compositions; and
to investigate the insulation properties at particular regions of
the bar in order to find weak spots in the insulation.
Influence of Manufacturing Quality on Time
to Breakdown of Winding Insulations
A. VPI InsulationThe VPI insulations tested were the same material with differ-
ent types of manufacturing, such as:
· reference manufacturing—taped by machine and supervisedby a skilled worker at the application laboratory of the mica
tape producer (12 specimens),
· industrially taped by machine in company no. 2 (4 speci-
mens),
· industrially taped by hand in company no. 1 (10 specimens).
Figure 8 shows the Weibull-plots for the times to breakdown
for the VPI insulations. The results in Figure 8 show that there is
no significant difference in the times to breakdown for the refer-
ence- and the machine-taped insulation. This proves that indus-
trial manufacturing with machine taping (company no. 2) is of ahigh quality. In contrast, significantly lower times to breakdown
were measured for the hand-taped material. The difference for
the 63% quantiles is more than two orders of magnitude. To find
the reason for such behavior, the microscopic structure of the
insulations was analyzed. Figure 9 shows a micrograph of the
reference insulation. Figure 10 is a micrograph of the machine-
taped insulation, and Figure 11 is a micrograph of the hand-
taped insulation.
The micrographs clearly show that the reference material is
without any voids or imperfections, similar to the industrially -
manufactured machine-taped material from company no. 2. In
contrast, the hand-taped material as manufactured by company
no. 1 shows delaminations in nearly every layer of the insula-
tion. The reason for such a difference is assumed to be the varia-
tion of the wrapping tension of the tapes. With hand taping, it is
not possible to have such a constant taping tension as with ma-
chine taping. This can lead to loose attachment of the tapes, and
the binder resin cannot fill the regions between the tapes suffi-
ciently. In that area, an electrical tree can propagate quickly, thus
causing very short times to breakdown. The break-down values
and the micrographs confirm a significant influence of the tap-
ing on the potential lifetime of winding insulations.
B. Resin-Rich InsulationsTo consider the influence of manufacturing on the times to
breakdown of resin-rich insulations, two different materials were
chosen, each with a different type of tapings:
· reference manufacturing—taped by machine and supervised
by a skilled worker at the application laboratory of the mica
tape producer (8 specimens),
· same material as the first, but industrially taped by machine
in company no. 3 (8 specimens),
· material no. 2 of Figure 6 as reference manufacturing—taped
by machine and supervised by a skilled worker at the applica-
Figure 7. Tree propagation through winding insulation.
Figure 8. Weibull-plots for times to break down with a VPI.
7/23/2019 IEEE-DeIS (Jun2006) Electrical Breakdown in HV Winding Insulations
http://slidepdf.com/reader/full/ieee-deis-jun2006-electrical-breakdown-in-hv-winding-insulations 6/9
May/June 2006 — Vol. 22, No. 3
tion laboratory of the producer of the mica tapes (12 speci-
mens),
· same material as the previous, but industrially taped by hand
in company no. 4 (10 specimens).
Figure 12 shows the Weibull-plots for the times to breakdown
of resin-rich insulations with the type of taping as a parameter.
The results show that there is no significant difference in the
times to breakdown for the machine-taped insulations. In con-
trast, the times to breakdown for the hand-taped material are
much lower than those of the reference specimens. The differ-
ence in the 63% quantiles is more than four orders of magnitude.
To identify reasons for this difference, the microscopic structure
of the insulations also was analysed. Figures 13–16 show micro-
graphs of the tested resin-rich insulations. The micrographs show
that the structure of both insulations with reference manufactur-
ing and the machine-taped material of company no. 3 are similar.
All materials have good adhesion of the tapes, and no voids or
delaminations are present within the insulation. In contrast, the
hand-taped insulation of material no. 2, company no. 4 shows
many voids in the insulation (Figure 16). Such a poor structure is
very likely responsible for the dramatic drop in the times to
breakdown shown in Figure 12. The reason is assumed to be the
poor pressing cycle. As shown for VPI insulations, hand-tapingcan cause poor quality with voids and/or delaminations. An elec-
trical tree propagates very quickly along such regions and leads
to short times to breakdown.
The results for resin-rich materials show that the type of manu-
facturing can have a significant influence on insulation life. It
can be seen that the sensitivity to poor preparation is much more
severe for resin-rich than for VPI insulations. For resin-rich mate-
rials, the binder resin must fill the voids that arise during wrap-
ping. If the volume of the voids is larger than the resin content in
the tapes, all the voids cannot be filled, thus leaving voids and
delaminations in the material. It is therefore concluded that tol-
Figure 9. Structure of VPI insulation: reference material.
Figure 10. Structure of VPI insulation: industrially
manufactured with machine taping (company no. 2).
Figure 11. Structure of VPI insulation: industrially
manufactured with hand taping (company no. 1).
Figure 12. Weibull-plots for the times to breakdown of resin-
rich materials: reference and industrially taped.
7/23/2019 IEEE-DeIS (Jun2006) Electrical Breakdown in HV Winding Insulations
http://slidepdf.com/reader/full/ieee-deis-jun2006-electrical-breakdown-in-hv-winding-insulations 7/9
10 IEEE Electrical Insulation Magazine
Figure 13. Structure of resin-rich insulation: reference
material.
Figure 14. Resin-rich insulation: industrially taped by
machine (company no. 3).
Figure 15. Structure of resin-rich insulation—material no. 2:
reference taping.
Figure 16. Resin-rich insulation of material no. 2: industrially
taped by hand (company no. 4) and poorly pressed.erances in taping are more critical for resin-rich than for VPI
insulations.
It also should be mentioned that 4 of the 10 samples in Figure
12 would have passed the 1-minute proof tests according to IEC
Standard 60243-1, [21], and IEC Standard 1212.2 [22]. Although
an estimation of the insulation life based on the power law may
not always lead to valid results [23], these insulations cannot be
expected to be reliable. Even if there would be no rapid electric
breakdown, additional thermal stresses would hasten the delami-
nation of loosely attached tape layers [6], and/or mechanical
vibrations contribute to crack formation near resin-filled cavi-
ties [6], [24], [25], thus causing a further reduction in the time to
breakdown.
ConclusionsFor the breakdown of high-voltage, winding insulation and
the influence of manufacturing quality on the time to break-
down, the following conclusions are made:
· Electrical tree propagation is the main electrical degradation
mechanism leading to breakdown of high-voltage, winding
insulation. To increase insulation life, tree inception should
be delayed, and tree propagation must be slowed down. A
delay of tree inception can be achieved by preventing voids
at the conductors. To slow down tree propagation, the mica
content of the insulation and the number of tape layers must
be as high as possible.
· The newly developed method with an embedded electrode
allows testing of insulation with far less effort than the stan-
dard test according to IEEE Standard 1043. It is recommended
that the new test be used for tests of insulation life in which
only small numbers of samples are available, which might bedue to time, financial, and/or material reasons.
· As impregnation of the insulation has a very significant effect
on the times to breakdown and hand taping can lead to weak
points inside the main-wall insulation, machine taping is rec-
ommended. If hand taping is used, VPI tapes should be ap-
plied as they are found to be less sensitive to lower taping
quality.
· Because manufacturing by different companies may reduce
the times to breakdown significantly, insulation quality
7/23/2019 IEEE-DeIS (Jun2006) Electrical Breakdown in HV Winding Insulations
http://slidepdf.com/reader/full/ieee-deis-jun2006-electrical-breakdown-in-hv-winding-insulations 8/9
May/June 2006 — Vol. 22, No. 3 1
should be monitored by random tests in the production line.
As it requires much less material, tests with an embedded
electrode are recommended for this purpose, and the resulting
times to breakdown should be compared to reference values
given by the manufacturer of the mica tapes.
· At three times the rated voltage, breakdown in insulations
with poor quality occurred within a few minutes; but, in insu-
lations with good quality, it took up to 5000 hours. The dura-
tion of the standard 1-minute proof test therefore should be
extended to about 10 minutes. Insulations with poor quality,
thereby, could be identified, but high-quality insulations
would experience no significant degradation.
AcknowledgementsDr. Vogelsang gratefully acknowledges the support of the Swiss
Commission for Technology and Innovation (CTI), Von Roll Isola
AG and PD Tech Power Engineering AG. In particular, the inten-
sive cooperation and stimulating discussions with Dr. B. Fruth
(PD Tech Power Engineering AG, Stetten), Dr. T.H. Teich (High
Voltage Laboratory of the Swiss Federal Institute of Technology
Zurich – ETH) Prof. H. Sauvain (University of Applied Science
of Fribourg), and Dr. C.-E. Stephan (Alstom Power AG Birr, all
from Switzerland) are greatly appreciated.
References[1] O. V. Thorsen and M. Dalva, “A survey of faults on induction motors
in offshore oil industry, petrochemical industry, gas terminals and oil
refineries,” IEEE Trans. Ind. Appl., vol. 31, no. 5, pp. 1186–1196,
Sep./Oct. 1995.
[2] IEEE Motor Reliability Working Group, “Report of large motor
reliability survey of industrial and commercial installations: Part I
and Part II,” IEEE Trans. Ind. Appl., vol. IA-21, no.4, pp. 853–872,
Jul./Aug. 1985.
[3] N. Srb, “Erfahrungen mit Stossspannungsprüfungen an elektrischen
Maschinen,” Allianz Report 70, Heft 2, pp. 58–62, Apr. 1997.
[4] H. Schaumburg, ²Werkstoffe und Bauelemente der Elektrotechnik:Polymere,“ (K. Brandenburger, R. Brütsch Chapter 18, pp. 615–
642), Verlag B.G. Teubner, Stuttgart, Germany, 1997
[5] Von Roll Isola, “Electrical Insulating Materials,” Technical Data
Sheets, Von Roll Isola Breitenbach, Switzerland, 2002.
[6] R. Vogelsang, “Time to breakdown of high voltage winding insula-
tions with respect to microscopic properties and manufacturing quali-
ties,” Doctoral Thesis, Swiss Federal Institute of Technology Zurich ,
Hartung-Gorre Verlag Konstanz, Germany. [Au: What year?]
[7] G. C. Stone, E. A. Boulter, I. Culbert, and H. Dhirani, Electrical
Insulation for Rotating Machines, New York, IEEE Press, Wiley,
2004.
[8] K. Kimura and Y. Kaneda, “The role of microscopic defects in
multistress aging of micaceous insulation,” IEEE Trans. DEI, vol. 2,
no. 3, pp. 426–432, Jun. 1995.[9] K. Hee Dong, J. Young Ho, and R. Hong Woo, “Effect of aging on
the microstructure evolution, thermal and mechanical properties of
mica/epoxy composite,” in Proc. IEEE Ann. Report Conf. EI Dielect.
Phenomena (CEIDP), Oct. 1999, pp. 537–541.
[10] M. R. Naghashan, “Untersuchungen zur Teilentladungs-aktivität von
Maschinentypischen Hochspannungs-isolierungen”, Doctoral The-
sis, University of Dortmund, Germany, Shaker Verlag Aachen, 1996.
[11] R. Vogelsang, R. Brütsch, K. Fröhlich, “Effect of electrical tree propa-
gation on breakdown in mica insulations”, in Proc. 13th Int. Symp.
High-Voltage Eng., ISH, Delft, The Netherlands, pp. 1–4, Aug. 2003.
[12] K. Engel, “Bewertung von Teilentladungen in spaltförmigen
Isolierstoffdefekten”, Doctoral Thesis, University of Dortmund,
Germany, Shaker Verlag Aachen, 1998.
[13] L. A. Dissado and G. C. Fothergill, Electrical Degradation and
Breakdown in Polymers, London, Peter Peregrinus Ltd., 1992.
[14] J. H. Mason, “Assessing the resistance of polymers to electrical
treeing,” IEE Proc., vol. 128, pt. A, no. 3, pp. 193–201, Apr. 1981.
[15] W. McDermid, “Damage resulting from long term slot discharge
activity in a hydrogen environment,” IEEE Int. Symp. EI, Conf. Proc.,
Toronto, Canada, pp. 361–362.
[16] Y. Shibuya, S. Zoledziowski, and J. H. Calderwood, “Void forma-
tion and electrical breakdown in epoxy resin,” IEEE Trans. Power
Apparatures Syst., vol. PAS-96, no. 1, pp. 198–206, Jan./Feb. 1977.
[17] M. Kaufhold, K. Schäfer, K. Bauer, A. Bethge, and J. Risse, “Inter-
face phenomena in stator winding insulation – challenges in design,
diagnosis and service experience”, IEEE EI Mag., vol. 18, no. 2, pp.
27–36, Mar. 2002.
[18] H. Mitsui and Y. Inoue, “Statistical analysis on the electrical failure
properties of the form-wound insulations systems for rotating ma-
chines,” IEEE Trans. EI, vol. EI-8, no.3, pp. 109–113, 1977.
[19] IEEE Standard 1043™-1996, IEEE Recommended Practice for Volt-
age-Endurance Testing of Form-Wound Bars and Coils.
[20] IEEE Standard 1553™-2002, IEEE Trial-Use Standard for Voltage-
Endurance Test ing of Form-Wound Coils and Bars for
Hydrogenerators.
[21] IEC 60243-1, Electrical Strength of Insulating Materials – Test meth-
ods – Part 1: Tests at Power Frequencies, 2nd Ed. 1998-01.
[22] IEC 1212-2, Industrial Rigid Round Laminated Tubes and Rods
Based on Thermosetting Resins for Electrical Purposes – Part 2:
Methods of Test, 1995-07.
[23] G. Stone, “The statistics of aging models and practical reality,” IEEE
Trans. EI, vol. EI-28, no. 5, pp. 716-728, 1993
[24] T. Weiers, D. Keller, and R. Vogelsang, “The impact of low ampli-
tude 100 Hz vibrations on the winding insulation of rotating high
voltage machines”, 14th Int. Symp. High Voltage Eng., ISH, Beijing,
China, Aug. 2005.
[25] R. Vogelsang, R. Brütsch, and K. Fröhlich, “Time to breakdown of
high voltage winding insulations at mechanical vibrations,” 14th Int.
Symp. High Voltage Engineering, ISH, Beijing, China, Aug. 2005.
Ruben Vogelsang (M’01) was born in
1972 at Annaberg-Buchholz, Germany. He
studied electrical engineering at Dresden
University of Technology (TUD), Germany
and the University of Sheffield, England.
From 1999 to 2000, he worked at the TUD
on the development of polymeric cables for
voltage transmission. In 2000, Ruben
Vogelsang joined the Swiss Federal Institute
of Technology (ETH) Zurich, Switzerland,as a research and teaching assistant and completed his Ph.D. in
2004. In parallel to his work for the Ph.D. degree, he studied
business at the ETH Zurich and gained in 2004 the degree of
Master of Business Administration (MBA). Since 2005, Dr.
Vogelsang is working as R&D project manager in the generator
circuit breaker division at ABB Switzerland Ltd. in Zurich, Swit-
zerland. He is a member of IEEE and can be reached at
7/23/2019 IEEE-DeIS (Jun2006) Electrical Breakdown in HV Winding Insulations
http://slidepdf.com/reader/full/ieee-deis-jun2006-electrical-breakdown-in-hv-winding-insulations 9/9
12 IEEE Electrical Insulation Magazine
Tilman Weiers was born in 1978 in
Freiburg, Germany. From 1998, he studied
Physics at the University of Stuttgart and
graduated with a diploma (Dipl.-Phys.) in
2003. Hr then joined the Swiss Federal In-
stitute of Technology (ETH) as a research
and teaching assistant. He can be reached
Prof. Dr. Klaus Fröhlich (F’02) was
born in 1945 in Salzburg, Austria. He re-
ceived the M.Eng. and Ph.D. degrees in tech-
nical science from the Vienna University of
Technology, Austria. After 11 years in
Switchgear and High Voltage Technology
with BBC (later ABB) in Switzerland, he
became a full professor at the Vienna Uni-
versity of Technology in 1990. Since 1997
he has been a full professor of High-Voltage Technology at the
Swiss Federal Institute of Technology Zurich, Switzerland. Klaus
Fröhlich is a Fellow of IEEE and chairman of CIGRE Study Com-
mittee A3 (High Voltage Equipment). He can be reached at
Rudolf Brütsch was born in 1943. He
studied Chemistry at the University of Bern
and received his Ph.D degree in 1972 from
the same university. Between 1972 and
1986, Dr. Brütsch worked in various R&D
positions on metal and ceramic materials.
In 1986 he joined Von Roll Isola AG in
Breitenbach, Switzerland and is today man-
ager for technical marketing of the busi-
ness unit—electrical of that company. Dr. Brütsch can be reached