Open Research Online

The Open Universitys repository of research publications

and other research outputs

Todays problems with the evaluation methods of full

lightning impulse parameters as described in IEC
60060-1
Conference Item
How to cite:
Berlijn, S.; Kvarngren, M.; Garnacho, F.; Simon, P.; Gockenbach, E.; Werle, P.; Hackemack, K.; Wong,
K. C. P. and Watts, M. (1999). Todays problems with the evaluation methods of full lightning impulse
parameters as described in IEC 60060-1. In: Eleventh International Symposium on High Voltage Engineering
: Vol.1, Topic A: Measurements and test techniques; Topic E: EMC sources, tests, protection., Conference
Publication, Institution of Electrical Engineers, London, UK, 1.49-1.52.

For guidance on citations see FAQs.

c [not recorded]
Version: [not recorded]
Link(s) to article on publishers website:
http://ieeexplore.ieee.org/servlet/opac?punumber=6616

Copyright and Moral Rights for the articles on this site are retained by the individual authors and/or other
copyright owners. For more information on Open Research Onlines data policy on reuse of materials please
consult the policies page.

oro.open.ac.uk
 TODAYS PROBLEMS WITH THE EVALUATION
METHODS OF FULL LIGHTNING IMPULSE
PARAMETERS AS DESCRIBED IN IEC 60060-1
S. Berlijn, F. Garnacho, E. Gockenbach, P. Werle, M. Watts,
M. Kvarngren P. Simon K. Hackemack K.C.P. Wong

KEMA Nederland B.V., Fundacion para el Schering-Institute of High NGC Engineering &
High-Voltage Laboratory, Fomento de la Voltage Technique and Technology,
The Netherlands Innovacion Engineering University of University of
Kungl Tekniska Hgskolan, Industrial-LCOE, Hannover, Germany Sunderland, School
Dept. of Electric Power Spain of Engineering,
Engineering, Sweden United Kingdom

ABSTRACT In the framework of this project, todays problems

In this paper the present problems with the evalua- with present definitions were studied . A question-
tion methods for lightning impulse parameters, as naire was sent to several laboratories world-wide, to
defined in IEC 60060-1, are described. Also the obtain information about the present evaluation
current practice of evaluation in many laboratories problems and the evaluation methods used by dif-
world-wide, that is obtained by a questionnaire, is ferent laboratories.
presented. Some of the work performed up the pres- II. DEFINITIONS IN IEC 60060-1
ent time and the initial conclusions are reported,
then some recommendations are made for future In order to point out where the problems with the
work. interpretation are, hereafter a brief revision of the
definitions of IEC 60060-1 relating to full lightning
I. INTRODUCTION impulses is given.
Tests with impulses are designed to demonstrate the
response of high voltage equipment to transients
over a wide frequency range. Lightning impulse
voltages represent transients occurring in high vol-
tage systems under operating conditions. Therefor
tests with lightning impulse voltages are part of rou-
tine and type tests. For the evaluation of the test
results it is imperative to determine the characteristic
parameters of these lightning impulse voltages as
defined in international standards. At present the
standard lightning impulse voltage is defined in
IEC 60060-1 [1] by its peak value (Up) and its time Figure 1: Evaluation rules for a Full lightning
parameters; the front time (T1) and the time to half impulse according to IEC 60060-1
value (T2).
Reproducibility of the calculation of the parameters A. Value of the test voltage
with various algorithms is possible for smooth im- During testing of actual specimens, impulses with
pulses, but the definitions in IEC 60060-1 and IEEE different wave shapes can occur. Besides the stan-
Std. 4 are insufficient for the evaluation of wave dard full lightning impulse (Figure 1), also impulses
shapes with oscillations and/or overshoot [2]. It is with wave shapes as shown in Figure 2, Figure 3
well known that even bigger problems with the and Figure 4 can occur.
evaluation of parameters arise when the impulse is a For a lighting impulse without oscillations or over-
non-standard lightning impulse voltage, since no shoots, the value of the test voltage is its peak value.
guidelines are given in IEC 60060-1. For lightning impulse with oscillations or overshoots
KEMA, FFII-LCOE, Schering-Institute (University the evaluation method for test voltage is depending
of Hannover) and NGC have collaborated in a pro- on the amplitude and duration of the overshoot or
ject funded by the European Community through it's frequency of the oscillation. In any case, a mean
SMT program. The aim of this international project curve, that has neither overshoot nor oscillations has
is to define one or more sets of parameters to to be drawn to decide if there are oscillations or
characterise full lightning impulses, prove the rele- overshoot in the recorded impulse.
vancy of these parameters and establish unambi- Overshoot or oscillations in the neighbourhood of
guous algorithms to evaluate them and write a pro- the peak, measured by a system according to
posal for the relevant parts of IEC 60060-1 and IEC IEC 60060-2 [3], are tolerated provided their single
61083-2. The project was started at January 1 1997 peak amplitude is not larger than 5% of the peak
and it will end at July 1 1999.
value. If the amplitude of the oscillation or over- III. HOW TO INTERPRET IEC 60060-1?
shoot is larger than 5% the impulse does not repre-
Independent if manual or digital evaluation methods
sent a standard impulse. For these and other impulse
are used, there are some problems with the interpre-
shapes (see for example Figure 3) the relevant
tation of IEC 60060-1. The following list is of some
Technical Committee shall define the value of the
of the problems associated with the definitions and
test voltage taking into account the type of the test
interpretation of the standard. They are certainly not
and test object.
all of the questions to be resolved, but most of them
If the amplitude of the oscillation or overshoot are are probably included.
not larger than 5%, the duration of the overshoot or
the frequency of the oscillation has to be deter- What is the definition of the mean curve, espe-
mined. If the frequency of such oscillations is not cially for impulses with both overshoot and os-
less than 0,5 MHz or the duration of overshoot not cillations?
more than 1 s, as in Figure 2 a) and b) a mean How to determine the T30% and T90% when
curve that has neither overshoots nor oscillations there are oscillations close to these times (see
should be drawn. For the purpose of measurement, case no. 11 in IEC 61083-2 [4]]?
the maximum amplitude of this curve is chosen as For the calculation of the front time, T1, the
the peak value defining the value of the test voltage. peak value should be used as reference for the
In the cases of Figure 2 c) and d) the peak value of 100% value of the impulse. When you have
the original wave is considered to be the test overshoots or oscillations, which peak value
voltage. shall be used, the peak value of the original im-
pulse or the peak value of the mean curve?
B. Front Time T1
How should the amplitude and duration or fre-
The front time T1 of a lighting impulse is a virtual quency of overshoots and oscillations be calcu-
parameter defined as 1,67 times the interval T be- lated?
tween the instants when the impulse is 30% and
90% of the peak value Up, (points A and B, Figure How should a non standard wave shape, for
1). instance those that occur during real testing, see
Figure 3 or Figure 4 , be evaluated?
How to know if they are inside the tolerances of
IEC 60060-1?
What is the physical background that supports
the magical 0,5 MHz value for the frequency
of oscillations and the 1 s value for the dura-
tion of the overshoot?
Are the parameters described in IEC 60060-1
really the parameters that characterise the
breakdown process? Could other parameters de-
scribe this process better?
Should the same parameters be used for all ma-
terials (XLPE, SF6, oil, vacuum, etc.)?
Figure 2: Evaluation rules for lightning impulses
with overshoot or oscillation

Figure 3: Examples of non standard lightning

impulses
C. Time to half-value T2.
The time to half-value T2 of a lighting impulse is a Figure 4: Typical lightning impulses in trans-
virtual parameter defined as the time interval be- former testing
tween the virtual origin O1 and the instant when the
voltage has decreased to half the peak value.
 IV. CURRENT PRACTISE IN LABORATORIES tric materials, when overshoots and/or high fre-
quency oscillations are superimposed on the
In order to know how the high-voltage laboratories
lightning impulse. There was no agreement be-
around the world are facing the evaluation problems
tween the received answers.
and to note which evaluation methods are used, two
different questionnaires (one for manufacturers and C. Testing laboratories
testing laboratories and another one for calibration
laboratories) were prepared. For full impulses:
In the questionnaire questions were asked, amongst All laboratories consider the peak value of the
others, about the kind of measuring instrumentation mean curve as the test voltage, which is in ac-
used to measure voltage and time parameters of cordance to the standard IEC 60-1.
lighting impulses. Also questions were asked about The majority of the laboratories uses a mean
details of the test circuit that they think are relevant curve that removes the oscillations to determine
when oscillations and overshoot are present. Besides the time parameters.
this, questions were asked how these laboratories For chopped impulses:
evaluate voltage and time parameters in the cases of Almost all laboratories consider the maximum
full and chopped impulses with oscillations super- value at the test voltage.
imposed (f 0,5 MHz) and/or overshoot (d 1 s).
The majority of the laboratories considers the
Only manufacturers and testing laboratories were time to the point when the voltage decreases
asked about the possible influence on the breakdown suddenly as the time to chopping.
behavior of dielectric materials of these overshoots
and/or high frequency oscillations with small ampli- The answers from testing laboratories are in a better
tude superimposed to the lighting impulse. Below agreement than the ones of manufacturer laborato-
the result of the questionnaire is presented. ries, and are also in better agreement with
IEC 60060-1.
A. Results of the questionnaires
D. Calibration laboratories
The questionnaires were answered by 23 manufac-
turer laboratories, by 8 testing laboratories, and by For full impulses:
12 calibration laboratories, all over the world. Half of the laboratories considers the peak value
The manufacturers laboratories were classified in of the mean curve as the test voltage and the
power transformers, cables and other high voltage other half considers the maximum value the test
equipment. voltage.
Some of the questions were focused on the evalua- The majority of the laboratories uses a mean
tion of lightning impulse waveforms when high fre- curve that removes oscillations to determine the
quency oscillations (f 0,5 MHz) or short duration time parameters.
overshoots (d 1 s) are superimposed on the im- Calibration laboratories propose some alternative
pulse. The main conclusions are summarized below. methods for evaluating lightning impulses, which
are not described in this paper.
B. Manufacturers laboratories
For chopped impulses:
For full impulses: The majority of the laboratories considers
The majority of manufacturers of power trans- maximum value the test voltage.
formers and cables use the maximum value as There is no general agreement on how to deter-
the test voltage (not in accordance with mine the time to chopping, although in the ma-
IEC 60060-1), while the majority of manufac- jority of the answers the time to chopping is
turers of other high voltage equipment considers considered to be the time when the voltage de-
the peak value of the mean curve. creases suddenly.
The majority of the laboratories determines T1
from the original curve, using the peak value of V. STATE OF THE ART AND FUTURE WORK
the mean curve as the 100% value. The accuracy in the determination of impulse pa-
Only some manufacturers of transformers de- rameters is limited when using analogue measuring
termines T1 from the original curve, using the instruments and manual evaluation. The digital
peak value of the original curve as the 100% measuring techniques removed this limitation, being
value. the reason for the replacement of analogue measu-
For chopped impulses: ring equipment by digital measuring equipment [5].
For the two methods to determine the test vol- The use of digital measuring equipment have made
tage (maximum value, or the peak value of the it possible to improve the accuracy, but also intro-
mean curve) the majority of laboratories uses duced some more evaluation problems.
the maximum value. For many years people have attempted to find a so-
For the questions to the possible influence of oscil- lution for the evaluation problems, which have
lations and overshoot: mainly been focused on the definition of the mean
curve.
About 30% of the manufacturer laboratories
supplied information of the possible influence
on the breakdown behaviour of different dielec-
Because computers are used for the evaluation, manufacturers. The differences in the obtained pa-
modern methods for calculating the mean curve can rameters when evaluating the same impulse are
be used [6, 7]. large, because of these dissimilar evaluation
Besides that the present rules for the evaluation of methods.
parameters are ambiguous, it can be questioned Besides this, there is not a well-established physical
whether the present parameters are the most suitable background about the relevancy of the parameters,
ones for modern insulating materials. Also because and the information supplied by the laboratories on
the change from analogue to digital measuring this subject is contradictory. Because of this it can
equipment removed limitations in the number and be questioned whether the present parameters and
type of parameters used to characterise lightning evaluation methods used are the most suitable ones.
impulses, it was appropriate to start investigate the The conclusions mentioned above, were the reason
relevancy of present or new parameters. As men- the research project funded by the European Com-
tioned in the introduction KEMA, FFII-LCOE, munity was started. The results of tests performed or
Schering-Institute (University of Hannover) and to be performed are not available at the time of
NGC are collaborated in a project funded by the printing and will be published at a later time.
European Community to investigate the relevancy of
parameters for different insulating materials (SF6,
XLPE, oil, air and vacuum).
To test the relevancy of for instance Up, T1, and
dU/dt, tests mentioned in Table 1 will be performed
for homogeneous and non-homogeneous fields as
well as for both positive and negative polarities:
Figure 6: Example of a generated wave shape
Table 1 Tests to be performed VII. ACKNOWLEDGMENT
Parameters to vary Parameters to keep con- The authors thank the European Community for
stant financial support of the project "Digital Measure-
0,5s<T1<2,0s T2=50s ment of Parameters used for Lightning Impulse
200 kHz<f<5 MHz T1=1,2s, T2=50s Tests for High Voltage Equipment" that is within the
5%<A<20% scope of the SMT-programme contract no. PL-
95120-SMT4-CT96-2132 and the institutions in-
200 kHz<f<5 MHz T1=1,2s, T2=50s, volved in this project: KEMA (Netherlands), FI-
5%<A<20% Up=U10% LCOE (Spain), NGC (UK), Schering-Institute
(Germany). The authors also thank all laboratories
For this purpose, a special generating circuit, as who filled in the questionnaire.
shown in Figure 5, has been designed and built. The
special generating circuit comprises of two parts that [1] IEC 60060-1: High Voltage test techniques,
are very nearly independent. One part generates a Part 1: General definitions and test requirements :
standard lightning impulse and the other part gener- 1989.
ates the oscillations or overshoots superimposed. In [2] F. Garnacho et al.: Evaluation procedures for
Figure 6 an example of a generated impulse is lightning impulse parameters in case of waveforms
shown. It can be seen that this looks similar to the with oscillations and/or an overshoot. 96 SM 522-
impulses in Figure 4. 3 PWRD, IEEE Power Engineering Societe.
Generator 1. Lightning Impulse Divider 1
Summer Meeting. Denver 1996.
[3] IEC 60060-2: High Voltage test techniques,
R2
C1 R1 C2
Rad1 Part 2: Measuring systems : Second Edition 1994.
Rbd1
[4] IEC 1083-2: Digital recorders for measurement
in high-voltage impulse test-part 2: Evaluation of
Test
Cell
software used for the determination of the
Digitizer
parameters of impulse waveforms International
standard, First Edition, 1996.
C L Rad2
[5] Gockenbach, E.:A basic concept for the use of
R
Rbd2
digital recorders in high voltage impulse measuring
technique, 8th ISH, Yokohama Japan, 1993.
Generator 2. Oscillation and Overshoot Divider 2
[6] A. Brede, P. Werle, E. Gockenbach, H. Borsi:
A new method of determining the mean curve of
Figure 5: Special generating circuit lightning impulses according to IEC 60060-1 11th
ISH, London, 1999.
VI. CONCLUSIONS [7] K.C.P. Wong, J. Tindle, H.M. Ryan, M. Watts:
The main conclusion of the questionnaire is that Digital measurement of lightning impulse
different laboratories are using different algorithms parameters using curving fitting algorithms 11th
and evaluations methods. Some of them are quite far ISH, London, 1999.
of the IEC 60060-1 rules, as for power transformer