Medium-Voltage Shielded Busbar Long-Term Ageing Te
Medium-Voltage Shielded Busbar Long-Term Ageing Te
Medium-Voltage Shielded Busbar Long-Term Ageing Te
net/publication/321205321
CITATIONS READS
2 1,511
4 authors, including:
François Gentils
Schneider Electric
50 PUBLICATIONS 263 CITATIONS
SEE PROFILE
All content following this page was uploaded by Karim Helal on 02 February 2018.
Abstract: Classical air insulated switchgear busbar consist in a simple metal component using the ambient air as dielectric
insulating medium: its sensibility to the external conditions is very high and dielectric failure is likely to occur. An
innovative modular MV busbar, similar in principle to the MV cable shielded accessories, has been designed to reduce
the internal arc probability and increase the reliability of medium-voltage (MV) switchgear. The live part is overmolded
by a first insulating rubber layer which is also overmolded by a second conductive rubber, connected to the earth. The
study proposes a validation method to assess the lifetime of this technology, by providing a complete review of the
ageing behaviour. The use of several ageing models (parametric and non-parametric) is discussed. Physical and
chemical ageing is followed up at various temperatures during more than 12,000 h both on material samples and on
real parts. The choice of an end-of-life criteria based on a dielectric test is discussed and compared with a more
classical criteria. In the end, this study review all the steps to perform a complete ageing study of a real industrialized
component.
CIRED, Open Access Proc. J., 2017, Vol. 2017, Iss. 1, pp. 325–328
This is an open access article published by the IET under the Creative Commons 325
Attribution License (http://creativecommons.org/licenses/by/3.0/)
material, which will lead to an underrated lifetime. A good
approach is to dissociate each ageing mechanism.
This will allow us to sort the ageing mechanisms by their effects
on the ageing and provide long-term behaviour based on the
dominant ones.
There are many ageing mechanisms according to the material and
the mission profile that will be encountered. In the study presented in
this article, main degradation mechanisms were oxidation (chemical
ageing) and stress relaxation (physical ageing). These degradation
mechanisms, highly dependent of the temperature, lead step by
step to the deterioration of the insulating material by reducing its
Fig. 1 Modular architecture: single compact three-position device with mechanical and dielectric properties.
shielded solid insulation switchgear (2SIS) Main important information when performing accelerated ageing
test is to know the temperature for which the product or
component is designed. For example, accelerated ageing test will
not be the same if you are testing a component that will work at
25 or 80°C.
ALT provide information regarding whether or not each test unit has
survived to a particular stress. Only the failure times for units that fail
and the running times for units that have not failed are observed. In
contrast, accelerated degradation tests (ADT) provide information on
the evolution of a measured variable that can be related to the loss in
the performance of the product. ADT are in general more informative
Fig. 3 2SIS flat interface design than ALT. However, in many cases it is complex to model the
(a) Shielding is assured thanks to the contact between the deformable conductive relationship between the measured variable and the performance of
material and the conductive coating of the plug, (b) Insulation is assured by the product. Also, in ADT the same unit must be measured several
compression between the deformable insulating material and the epoxy plug, times during the test, so special care is needed to ensure that the
(c) Electrical conductor can be composed of aluminum or copper
measurement does not interfere with the degradation process [4].
The failure of a product can be related to a deterioration of the It is common practice to assume that the time-to-failure follows a
properties of its constituent materials beyond the minimum lognormal or Weibull distribution in which the scale parameter is
performance required by its design and application. Therefore, the a function of the stress level, x, and the shape factor does not
first step in understanding the durability of a given design consists depend on x. If xN denotes the normal stress, the time-to-failure at
in the analysis of its application, product use profil or mission profile. the accelerated conditions is scaled by a factor AF(x):
This last consist of the following:
time-to-failure xN = time-to-failure( x) · AF( x)
† operational and environmental stresses, their magnitude and
sequence;
The Arrhenius’ relationship is a widely used model to describe the
† the duration and number of sequence segments.
effect that temperature has on degradation mechanisms that depend
on a single chemical reaction, diffusion or migration process. The
The analysis of the mission profile should highlight the nature of Arrhenius’ relationship is given by
the stress factors, which can be external, like humidity, solar
radiation, temperature etc. or internal, like mechanical loading,
electrical current etc. It is obtained by choosing from one of the time-to-failure(T ) = Ce(Ea /kT )
following evaluation conditions: average use profile, aggressive
use profile, and a spectrum of use profile conditions. where T is the temperature (in K), k is either Boltzmann’s constant or
Accelerated tests should reproduce the same stresses as in the the universal gas constant, Ea is the activation energy and C is a
mission profile and not to add additional constraints to the model parameter to be determined (C > 0). In many cases, the
CIRED, Open Access Proc. J., 2017, Vol. 2017, Iss. 1, pp. 325–328
326 This is an open access article published by the IET under the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/3.0/)
degradation process consists of multiple steps, with each step having
its own activation energy.
Given T and TU the accelerated and normal use temperatures,
respectively, the Arrhenius acceleration factor can be written as
−C1 × (T − TR )
log aT =
C2 × (T − TR ) 5.2 Chemical ageing and Arrhenius model
CIRED, Open Access Proc. J., 2017, Vol. 2017, Iss. 1, pp. 325–328
This is an open access article published by the IET under the Creative Commons 327
Attribution License (http://creativecommons.org/licenses/by/3.0/)
parameters (shape of the parts, the temperature variations, the
electrical fields, the moulding process influence etc.) and have to
be realised to assess a lifetime of the component.
The ageing process of the material is obtained during the high
temperature phase but most of the dielectric failures are revealed
during the cold phase of the cycle, or the transition, because of a
loss of elasticity and compression of the elastomer (Fig. 6). The
ageing voltage is defined by 2.5U0, and was determined in order
to get a strong partial discharge activity in defective parts, and
therefore, be able to quickly detect them.
In this study, the well-known inverse power law (Vn × t = C st) was
not used, because it was demonstrated that the root cause of busbar
failures was thermomechanical effects near the interfaces yielding to
a strong partial discharge activity, and not pure electrical field effects
in the solid insulation material (busbar is designed with low and
non-critical field value in the solid material).
Ageing results under voltage on real parts allow us to provide a
Fig. 6 Example of busbar ageing conditions comparison between the ageing behaviour determined on material
samples and on real industrial moulded parts. It is a strong double
validation for the lifetime prediction. This method allows us to
6 Ageing tests on complete busbar certify the lifetime of the busbar according to the terms of use
specified to customers.
The last stage of an ageing study to validate the reliability of a busbar
is to carry out ageing tests on actual parts. Two kinds of tests have
been realised:
Thermomechanical ageing test of the busbar and monitoring of 7 Conclusion
the compression set: Ageing parameters have been selected by
referring to the TTS model and enabled to realise the tests on The validation of new technologies is a long and difficult process. In
small parts and at less ageing temperatures. Compression set addition to qualification tests, long-term ageing tests, taking into
measurements on actual busbar was at least 1.8 times better than account all different ageing phenomena, must be done. This step is
compression measured on samples. These results highlight that mandatory to check the long-term behaviour of the product or
lifetime determined on sample is critical comparing to real busbar component, and then to verify that the expected life time suits
behaviour. It also allows us to redefine the end-of-life criteria used with customers’ needs.
for the concerned ageing model.
Moreover, dielectric performance of the busbar was also checked
with regular partial discharges, power frequency and lightning
impulse tests at room temperature and at −25°C, critical 8 References
temperature for which the rubber shrinkage increases. 1 Hugoo, B.: ‘Eric Bonneville’, Green Premium, 2016, Schneider Electric eco-label
Ageing test of the busbar under voltage: During the study, it was that goes beyond compliance, p. 1
demonstrated that the dielectric failure mode of the real parts were 2 Reliability HotWire: ‘Determining reliability for complex systems, Part 2 –
not due to the electrical field presence in the material (it is not the simulation’, 2001, p. 1
3 Bonfils, J.-M.: ‘Connecting bar with flat interface’. US 7594822 B2, 2009, p. 1
root cause of the ageing process), but by a thermomechanical 4 Elsayed, E.A.: ‘Reliability engineering’ (Wiley, 1996), vol. 368–370, p. 2
ageing of the material, leading to the loss of compression in the 5 Toop, D.J.: ‘Theory of life testing and use of thermogravimetric analysis to predict
critical interfaces. However, these tests take into account all the the thermal life of wire enamels’, IEEE Trans. Electr. Insul., 1971, EI-6, (1), p. 3
CIRED, Open Access Proc. J., 2017, Vol. 2017, Iss. 1, pp. 325–328
328 This is an open access article published by the IET under the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/3.0/)
View publication stats