1 February 2014
This paper presents partial discharge (PD) inves- and security as well as creating less visual and envi-
tigation of dierent cable termination defects. The ronmental impact. The previous study revealed that
medium voltage power cables as rated of 3.6/6(7.2) although underground distribution system is presum-
kV are applied. Finite Element Method Magnetic ably reliable comparing to overhead counterpart, but
(FEMM) program is used as a simulation tool for failures of underground cable system cause consid-
electric eld stress investigation. The partial dis- erably longer repair time. Proper cable termination
charges patterns are detected by using a commer- and joint are vital for reliable delivery of electricity.
cial High Frequency Current Transformer (HFCT). Therefore, in this paper, the partial discharge (PD)
The simple cases for internal, surface and corona dis- measurement using commercial High Frequency Cur-
charge are rstly observed in order to investigate the rent Transformer (HFCT) is performed to investigate
performance of the HFCT. Then eight dierent case discharge pattern and PD inception voltage of vari-
studies of cable termination defects are further in- ous cases of cable defects due to human error. The
vestigated, which includes non-terminator, voids be- FEMM program is used as a simulation tool in order
tween XLPE and stress control, 20 mm. overlaps be- to investigate electric eld stress in cable termination.
tween semiconductor and stress control, particles on The trend toward breakdown of dierent defects is ad-
XLPE, non-smooth XLPE, needle tip on insulation ditionally analysed. Consequently, the human error
screen, impropriate cable bending, and proper termi- during cable terminator installation leading to termi-
nation. The results are then compared with the re- nation defects result in faster degradation of termi-
sults from a conventional PD diagnosis tool according nator due to high electric stress in the detected area.
Fig.1: Basic IEC 60270 Discharge Detection Cir- (a) Internal Discharge in Dielectric Voids
cuit.
3. EXPERIMENT SETUP
1. Non-stress Control
3. Bending Radius
4. Sandpaper Insulation
5. Particles on XLPE
7. Insulation Roughness
Case2 : Improper stress control position : Im- Using the test circuit in Fig. 4, the voltage from
proper installed position by sliding 20 mm. out from test transformer is raised until the PD occurs. The
installation standard due to viciousness of worker. acquisition period is 10 seconds for any test. The test
voltage is applied in sinusoidal waveform. Then, the
Case3 : Small bending radius : In some place, the
discharge is detected by ICMsystem and displayed in
sharp bending of power cable is inevitably required
form of dots as well as charge intensity whereas the
that causes air gap inside in terminator.
discharge detected by HFCT is displayed in form of
Case4 : Ef f ect of sandpaper : Sandpaper for
voltage spite with respect to sinusoidal waveform.
scrubbing insulation screen causes small insulation
damage and voids.
Case5 : P article on XLP E : Unclean XLPE 4. SIMULATION RESULTS BY FEMM
skin causes dust and semiconductor particles contam-
The electric eld stresses of eight simulated defects
inated on XLPE.
of cable terminators were observed by using Finite El-
Case6 : N eedle tip on screen : Improper remov- ement Method Magnetic (FEMM) program in order
ing of outer-semiconductor screen causes needle tip. to investigate electric eld stress, voltage distribu-
Case7 : Insulation roughness : Using ma- tion as well as the critical point in cable termination,
chine for removing outer-semiconductor produces which is expected to be the initial cause of partial dis-
non-smooth XLPE skin. charge and breakdown. Fig. 7 on the left shows the
Case8 : P roper termination : According to in- recommended dimension of each layers of the power
Investigation on Partial Discharge of Power Cable Termination Defects using High Frequency Current Transformer 19
(a) Detected by ICM system (b) Detected by HFCT accurate inception voltage and electric charge of all
cases can be detected only by using the conventional
Fig.13: Earth Needle Corona. tool. The results are given in Table 2. After that the
measurement using HFCT has been performed and
its result is presented in Table 3.
face discharge, HV needle, and earth needle cases, are However the comparison of electric charge mea-
given in Table 1. The results show that the HFCT can sured by both tools is not possible due to the dier-
detect the PD patterns and electric charge that are ence in detection principle. This shows that the con-
conrmed by the results from the conventional diag- ventional tool has better sensitivity than the HFCT.
nostic tool as shown in Fig. 10 to Fig. 13. However, Moreover, the phase resolved measurement results
the HFCT can clearly detect only when the number of from both tools for 8 cases are presented from Fig.
electric charge in pC. is high enough while the conven- 14 to Fig. 21 in order to compare the discharge pat-
tional tool can sensitively detect the inception voltage tern. It is clearly seen that the detected PD patterns
and electric charge. from both tools are similar and imply the combina-
B. P D Detection of Cable T ermination Def ects tion of internal and surface discharge phenomena.
For eight cases of cable termination defects, the
Investigation on Partial Discharge of Power Cable Termination Defects using High Frequency Current Transformer 21
(a) Detected by ICM system (b) Detected by HFCT Fig. 22 by plotting the relationship between the dis-
charge voltages and electric charge of all cases. The
Fig.15: Eects of Stress Control Position. trends are dierentiated into three groups. The most
severe group is case of non-terminator. The partial
discharges occurred around the range of rated voltage
and increased signicantly with the slight increasing
of the supply voltage. The second group consists of
needle tip on insulation screen and insulation rough-
ness. In this group, the partial discharges occurred
(a) Detected by ICM system (b) Detected by HFCT when the supply voltage was increased between two
to three times of the rated voltage. The last group in-
Fig.16: Eect of Bending Radius. cludes the cases of 20 mm. overlaps between semicon-
ductor and stress control, particles on XLPE, impro-
priate cable bending, non-smooth XLPE, and voids
between XLPE and stress control. The good cable
termination can withstand the highest voltage up to
16.02 kV. Thereafter, the partial discharge occurred.
[5] L.-J. Chen, T.-P. Tsao, and Y.-H. Lin, New di-
agnosis approach to epoxy resin transformer par-
tial discharge using acoustic technology, IEEE
Fig.22: Trend of Breakdown Voltage of Cable Ter- Trans. Power Del., vol. 20, no. 4, pp. 2501-2508,
mination Defects. Oct. 2005.
In the conclusion, human error during cable termi- tial discharge measurements, IEEE Trans. Di-
nator installation leading to cable termination defects electrics Elect. Insulation, vol. 18, no. 5, pp.
can cause the fast degradation of cable terminator be- 1798-1802, Jun. 2011.