2013 IEEE International Conference on Solid Dielectrics, Bologna, Italy, June 30 – July 4, 2013

Effects of MV PILC Cables Ageing Process on the

Diagnostic Parameters: p-factor and tanδ

I. Mladenovic, Ch. Weindl, Th. Scharrer

University of Erlangen-Nuremberg
The Chair of Electrical Energy Systems
Erlangen, Germany
[email protected]

Abstract— During field operation, power equipment is that has been shown as very appropriate particularly in the case
exposed to numerous impacts (especially thermal and electrical of cables with impregnated paper insulation.
stress), which result in a continuous deterioration of its insulation
material. The grade of the deterioration can be weighted by In this article, results of two measurement series, before
different diagnostic measurements. Within an accelerated ageing (2009) and after the artificial ageing (2012), are presented.
experiment on paper insulated lead covered (PILC) cables, Beside influence of the ageing (for cables in different age
several diagnostic parameters have been measured regularly groups), selected measurement series enable analyzing of the
during the ageing process of ca. 2 years. The cable samples are temperature and test-voltage influence on the return voltage
PILC cables (20kV, 150mm2) of 13m length, which are differently curves and also on p-factor – a diagnostic parameter derived
pre-aged (from brand new, up to cables that have been in from the characteristic values of the return voltage curve that
operation for 60 years). In this article aging effects of this should be less sensitive to temperature and polarization
insulation system on dissipation factor and parameters of return voltage.
voltage curve as well as on p-factor will be shown and discussed.
Furthermore, the development of the tanδ and p-factor
Keywords—medium voltage cable; PILC cable; diagnostics; during the ageing process will be directly compared, analyzed
ageing; dissipation factor; return voltage and discussed. In addition time constant τ as a supplementary
way for the evaluation of return voltage measurements will be
I. INTRODUCTION suggested.
The applicable way to follow the irreversible changes that
an ageing process causes on the electrical and also other II. PHYSICAL BACKGROUND OF THE DIAGNOSTIC
insulation properties is to reproduce the ageing with PARAMETERS: TANδ AND p-FACTOR
accelerated and artificial but realistic conditions and to monitor A. Dissipation Factor - tanδ
the features of interests. The main goals of such experiments
are to form a sufficient knowledge databank for the purpose of The dissipation losses in the dielectric are the outcome of
statistical categorization of the measured parameters and an the moving and arranging of polarized molecules, ions and in
improvement of practical diagnostic interpretations. principle charge carriers within the insulation material under
the influence of the electric field. Measuring the dissipation
In order to point out the limits and characteristic values of factor - , the total losses (due to the polarization processes
the different diagnostic variables, an accelerated aging - , conductivity - , and partial discharge
experiment on PILC (paper insulated lead covered) cables has activity - ) the insulation condition can be weighted. On
been started. Therefore, a novel system for the accelerated the other hand, the charges mobility and therefore also
ageing of MV (medium voltage) cables called ICAAS appropriate dissipation losses are very dependent on the
(Integrated Cable Accelerated Ageing System) was developed frequency ( ) and strength of the applied electric field ( ) and
and realized, [1] [2]. The aim of this project is to find suitable on the temperature of the dielectric ( ). Therefore, it can be
parameters describing the aging process of PILC cables by a written:
constant monitoring of the relevant aging parameters and
mostly used diagnostic parameters: dissipation factor -tanδ and , , , , , ,
partial discharges. During almost two years of the ageing , , (1)
experiment, numerous studies have been carried out, where the
influence of temperature, voltage, cable condition etc. on the Measurement of dissipation factor is a well-known method
diagnostic parameters have been investigated. Actually, the so for the evaluation of insulating materials.
formed knowledge databank consists of over 270.000
measurements and more than 900GB of data, [3] [4]. B. Return Voltage Measurement - p-factor
The principle of return voltage measurement is based on the
In addition, also other diagnostic possibilities have been specific behavior of the insulation system of PILC cables that
tested. One of them is a return voltage measurement (RVM) consists of two different materials (paper and mass), resulting

978-1-4673-4461-6/13/$31.00 ©2013 IEEE

92
 2013 IEEE International Conference on Solid Dielectrics, Bologna, Italy, June 30 – July 4, 2013

in the occurrence of boundary polarization when

w a DC-voltage Moreover, measurements perfoormed before artificial ageing at
is applied to the cable for a longer polarizaation time tp. After a temperature of 23°C are show
wn in red.
short-circuiting for usually td of 2s, the retuurn voltage can be
measured. The return voltage curve is a dooubled exponential A. Influence of Polarisation Voltage
Vo
function caused by the differing time connstants of the two Fig. 2 shows the values off the p-factors of several cables
insulation materials. Because of several knoown dependencies that have been calculated out of the return-voltage
of return voltage measurements, p-factor haas been introduced measurements for different voltage levels. It can be seen that
in order to provide a reliable parameter forr the evaluation of there is no uniform behavior of the values with varying the
those measurements. p-Factor is given by b the following polarization voltage. Anywaay, in comparison with the
equitation. parameters of the return vooltage curve, especially return
⁄ ⁄
voltage maximum, as exempplarily shown in Fig. 3, the
. (2) variation of the p-factor values is rather small, so that it can be
·
concluded that p-factor is – ass expected from equitation (2) –
almost independent of the polarrization voltages value.
It depends on the three parameters maxiimum voltage Um,
initial incline s and time of maximum voltaage tm, Using the
equitations of these parameters, p-factor can be described only
by λ -the ratio of the insulation materials timee constants.
p-Factor is supposed to be independent of
o the polarization
voltage, geometry and nearly of temperature [5].
s
Um

tm t

Fig. 1 Return Voltage Curve and its characterristic parameters

With the time constant τ an addiitional diagnostic

parameter is regarded. τ is the time constantt of an exponential
function that fits the falling part of the returnn voltage curve that
nearly equates the time constant of the insulattion paper. Fig. 2 p-factors at voltages: 2kV (ssquare mark) and 4kV (round mark)
before (in red, 23°C) and after (in
( blue 11°C and in green 23°C)
III. TEST CONDITIONS AND CABLEE SAMPLES artificial aging
The cable samples of interest have beenn manufactured in
1987 and 1988 (service-aged cables), in 20000 (unused, stored
cables) and in 2008 (new cables). A first series of return
voltage measurements has been taken in 2009. After the
process of artificial aging, the measurem ments have been
repeated with identical parameters (tp=15min, td=2s;
polarization voltages 1, 2 and 4kV) in 20012. Generally the
three conductors of each cable have been connnected in parallel
for the measurements. Due to the failures occcurred during the
artificial aging, this was no longer possiblee for all cables in
2012.
IV. RESULTS
For the purpose of analyzing the influencce of different test Fig. 3 p-factors at voltages of 2kV and 4kV before and
parameters on p-factor and return voltage curve, exemplary after artifficial aging
results have been selected out of two seriess of RVM (before
and after accelerated and artificial ageing) and
a appropriate i.e. me constant τ (Fig. 6) shows a
Differing from p-factor, tim
comparable tanδ values. In the following
f figures clear voltage dependency. Vaalues are decreasing for higher
measurements performed with 2kV polarizzation voltage are polarization voltages.
symbolized by square signs and those taken ata 4kV polarization
voltage by round signs. Different temperaturres are symbolized
by different colors: blue for 11°C and green for 23°C.

978-1-4673-4461-6/13/$31.00 ©2013 IEEE

93
 2013 IEEE International Conference on Solid Dielectrics, Bologna, Italy, June 30 – July 4, 2013

B. Influence of Temperature D. Comparison between p-facttor and dissipation factor tanδ

To analyze only the influence of the tempperature, the RVM In Fig. 5 the values of dissipatiion factor are shown in addition
after the process of artificial aging have beenn taken at different to p-factor. It is recognizablee, that the same ageing effects
temperatures - 11°C and 23°C. It can be seen in Fig. 4 that caused different relative changes of p-factor and dissipation
temperature strongly influences the parameeters of the return factor.
voltage curve, but also (as shown in Fig. 1)) the calculated p-
factors (blue and red signs). For both p-factorr and time constant
of the paper τ, no clear correlation to tempperature is visible.
Anyway, it is to be recommended to ennsure comparable
environmental test conditions when perform ming return voltage
measurements.

Fig. 5 Direct comparison of thee development of two diagnostic

parameters: p-factor and tannδ during the ageing process
Fig. 4 Return voltage curves for two differennt temperatures
One of the single-conductoors of cable 1988a, which failed
Although the theoretical assumption abbout temperature- shortly after the repetition meeasurement, shows the expected
independency of p-factor does not always appply, it seems to be significant increase of the disssipation factor while, p-factor is
given for the stronger aged cable samples. As A especially these approaching the characteristic p-factor
p level of 0.3 for strongly
cables are of interest for diagnosis, p-factor therefore can be a aged cables [5] with not so remarkable
r increase due to the
useful diagnostic parameter. accelerated ageing.
C. Development of p-factor due to the ageinng process Comparing further the cablles from 1987, they both show a
In Fig. 5 the influence of the cable ageinng on the p-factors strong increase of their dissipaation factor and smaller changes
under comparable test conditions can be seenn. of p-factor. The p-factor of caable 1987a has even decreased.
Using only the p-factor as diaggnostic parameter, the effects of
With only one exception (1987a), the values of the p- artificial ageing could not be b recognized for these cable
factors after artificial aging in 2012 are higgher than in 2009. samples.
Differing to the others, cable sample 1988c shows less
development of the p-factor. This cable sam mple is aged only Cable 1988c has been put under electrical stress only. As
electrically, and comparing its values to thoose from the same mentioned above, regarding p--factor it would have aged very
pre-ageing-group (1988a, 1988b – cable samples s from the less. On the other hand, a strrong increase of the dissipation
same field-line that have been aged thermo-eelectrically), it can mparable to the tanδ increase of
factor can be stated. That is com
be stated that the influence of electrical agingg is rather small. the thermo-electrically aged cabbles from 2000.
Regarding the development of the p-facctor of the cables E. Development of time-coonstant τ due to the ageing-
which have been aged both thermally and electrically
process
(separately for each year of manufactture), they show
comparable increases of p-factor, with an exception of the In addition to p-factor the development
d of the time-constant
cables that have been unused and stored on the t cable drum for τ has been investigated. Therebby, it has to be regarded that the
ten years – cable samples 2000a and 2000b. Although these measurements before artificial ageing haven been taken using
cables have not experienced any service-aginng before, their p- the three conductors of each cable
c under test in parallel. As
factors show the strongest increase compaaring to all other failures occurred during the artiificial ageing, this was no longer
cables under the test. The values after arrtificial aging are possible for each of the cables at the repetitive measurements.
comparable to those of the pre-aged cables froom 1987 and 1988 Because of the short cable lengths the influence of the
or even higher. It is assumed that the manneer of storage could measuring resistor is therefoore no longer negligible and
have led to a local loss of mass in the upper part
p of the drum. measurement values obtained from f the affected cables (2008a,
1988a and 1987b) have to be corrected mathematically. Both

978-1-4673-4461-6/13/$31.00 ©2013 IEEE

94
 2013 IEEE International Conference on Solid Dielectrics, Bologna, Italy, June 30 – July 4, 2013

the original and the corrected values are shoown in Fig. 6. The test conditions have been consiidered. It is shown, that, in front
capacity of those cables with three conducctors in parallel is of all, temperature has a reemarkable but not correlative
about 50% higher than the capacity of thoose with only two influence on the parameters off the return voltage curve, much
conductors in parallel. As the measuring resistor
r is still the less intensive dependency of p-factor – but still not negligible.
same, the time-constants are different and itss values have to be The return voltage is dependency on the polarization voltage
corrected by a factor of 1.5. level is well compressed by the derivation of p-factor.
Moreover, an additional diagnostic parameter – time constant
of the paper is suggested.
Further, the direct comparisson of the ageing effects on two
different diagnostic parameterrs showed that strongly aged
cables leave the mark on both: tanδ and p-factor; in other cases
beside absolute values of theese parameters also additional
information or rather criteria, like
l parameter dependencies on
test voltage, must be applied.
It can be concluded that byb return voltage measurements
specific properties of the insuulation system can be analyzed.
Intelligent parameters like p-factor and time constants in
combination with the correlattion of the results from other
diagnosis methods can improvve the knowledge about aging
processes and hence changinng of the dielectrics electrical
parameters.
ACKNOWLLEDGMENT
The authors would like to thank the following companies
for their support: N-ERGIE E Service GmbH, Nürnberg,
Germany and N-ERGIE Netz GmbH,
G Nürnberg, Germany.
REFER
RENCES
Fig. 6 p-factors at voltages of 2 andd 4kV
before and after artificial agingg [1] Mladenovic I. and Weindl Chh., "Artificial Aging and Diagnostic
Measurements on Medium-Volltage, Paper-Insulated, Lead-Covered
For almost all of the cables under test an
a increasing time- Cables," IEEE Electrical insulatioon Magazine, vol. 28, no. 1, pp. 20-26,
constant can be noticed. Only time-constaants of the cables January/February 2012.
2000a and b show a different developmentt. The p-factors of [2] Mladenovic I. and Weindl Ch., "IICAAS – Integrated System for lasting
Accelerated Aging of MV Cables, Data Monitoring and Acquisition," in
these two cables have increased stronglyy due to artificial IEEE Conference on Electrical Insulation and Dielectric Phenomena
ageing. 2009, CEIDP, Virginia Beach, USAA, 2009.
As mentioned above, one of the single coonductors of cable [3] Ch. Weindl, "Verfahren zur Bestim
mmung des Alterungsverhaltens und zur
1988a failed after the repetition meaasurement. While Diagnose von Betriebsmitteln der elektrischen Energieversorgung,"
University of Erlangen-Nurembergg, Erlangen, Habilitation Thesis 2012.
dissipation factor therefore shows a signifiicant increase, the
[4] I. Mladenovic, "Determination ofo the Remaining Lifetime of Paper
change of time constant τ is rather smalll and its value is Insulated Lead Covered Cables based on the Diagnostics of Partial
comparable to the other cables from 1987 andd 1988. Discharges and the Dissipation Faactor," Friedrich-Alexander Universität
Erlangen-Nürnberg, Erlangen, Germmany, Dissertation 2012.
V. CONCLUSION
[5] Patsch R., Kouzmine O., and Meenzel J., "Zur Kurvenformanalyse von
The global condition of an insulatingg system can be Rückkehrspannungskurven von Öl-Papier-Kabeln,"
Ö in ETG-Fachtagung
weighted by applying different diagnostic methods as ‚Diagnostik elektrischer Betriebsm
mittel’, Kassel, Deutschland, 2006, pp.
measurements of dissipation factor or return voltage
v curve. 87-92.

In this paper results of these measuremeents on MV PILC

cables, during a long-lasting ageing experriment, have been
exemplarily presented. Besides the influencee of the insulation
condition on the diagnostic parameters, alsoo influences of the

978-1-4673-4461-6/13/$31.00 ©2013 IEEE

95