2014 International Conference on Power System Technology (POWERCON 2014)
2014
Statistical Characteristics of Partial Discharge
Caused by typical defects in Cable Joint under
Oscillating Voltage
Guangmao.Li, Zuwei. Luo, Jun.Xiong, Zucheng.Huang, Yajun.Wang
Abstract—The oscillating voltage test is a nondestructive
detection method for partial discharge of XLPE cable and has
been applied recently. This paper made three kinds of varying
severity artificial defect model of cable joints in 10 kV XLPE cable.
Oscillating voltage is applied to the model, by use of pulse current
method to detect partial discharge signals. In order to study the
statistical characteristics of partial discharge of cable joint under
the oscillating voltage, three-dimensional statistical map has been
made. The results show that for the same kind of defects, with the
increases of the defect severity, the discharge interval extended,
the magnitude and the number of partial discharge increase; for
different kinds of defects, obvious difference exist among the
maps, this may established a foundation for the further study of
the partial discharge pattern recognition of XLPE cable under
oscillating voltage.
Index Terms—XLPE cable, cable joint defect model, partial
discharge, oscillating voltage, three-dimensional statistical map
I. INTRODUCTION
W ITH the advantage of simple process, reasonable
structure and excellent electrical performance, cross
linked polyethylene (XLPE) cable has gradually replace
the traditional paper oil insulation cable and widely applied
around the world. Bad construction technology, artificial
destruction and severe environment may cause insulation
defect after cable leave factories, especially for the cable
accessory defect, which will affect its insulating property [1].
Manuscript received July 29, 2014. This work was supported in part by the
Science and Technology Project of CSG K-GD2012-014 .
Guangmao.Li is with Guangzhou Power Supply Bureau, Guangzhou
510410, China. (e-mail: 819078565@qq.com).
Zuwei. Luo is with Guangzhou Power Supply Bureau, Guangzhou 510410,
China.
Jun.Xiong is with Guangzhou Power Supply Bureau, Guangzhou 510410,
China.
Zucheng.Huang is with Guangzhou Power Supply Bureau, Guangzhou
510410, China.
Yajun.Wang is with state key laboratory of power transmission equipment
& system security and new technology, Chongqing University, Chongqing
400030, China
POWERCON 2014 Paper No CP1746
Chengdu, 20-22 Oct.
Researches have shown that the severity of partial discharge in
XLPE cable can indicates the change of its insulation state.
Partial discharge test is recommended as the main diagnostic
method of the XLPE cable insulation by international authority
such as IEEE and CIGRE [2]-[3].
It has been found difficult to do the partial discharge test for
the power source is usually too big because of the long length
and high equivalent electric capacity of XLPE cable [4]. The
current XLPE cable handover and preventive test such as
ultra-low frequency voltage (0.1 Hz) method and DC voltage
method cannot work effectively in the off-line partial discharge
detection test, because they all need long action time and may
cause other defects [5]-[6]. However, oscillating wave testing
system (OWTS) is now widely used in partial discharge
detection in the off-line state of XLPE cable [7]-[8]. This
method has advantages such as short action time, harmless to
the insulation and its voltage waveform meet the demand of
IEC60270 [9].testing voltage characteristic. Due to the
favorable equivalence to power frequency voltage, oscillating
voltage has been widely applied as it could accomplish the
cable insulation defect test under the circumstance similar to
power frequency. Studies have shown that the application of
OWTS mainly concentrate on the detection and locate of cable
insulation defect, but the research on the statistical property of
partial discharge signal under the oscillatory wave are
insufficient [10]-[14].
This paper made three kinds of varying severity artificial
defect model of cable joints in real 10 kV XLPE cable,
oscillating wave testing system is used to carry out partial
discharge test. In order to study the effectiveness of partial
discharge test in detecting cable defect and the characteristic of
partial discharge signal under oscillating voltage,
three-dimensional statistical map of the partial discharge are
made to analyze the data.
II. TESTING CIRCUIT AND DEFECT MAKING
A. Oscillating Voltage Testing Circuit
In order to study the characteristic of partial discharge signal
under oscillating voltage, the testing adopt oscillating wave
Page 1/6 1368 Session 2
 2014 International Conference on Power System Technology (POWERCON 2014)
testing system to generate oscillating voltage, three kinds of
varying severity artificial defect on the cable joint which is of
high fault rate are made at the same time.
As shown in Fig.1, the OWTS constitutes of high voltage DC
source, resistance R, no halo reactor L, high voltage switch S
and test cable etc. The basic principle of oscillating voltage test
method is to build a LRC underdamped oscillator circuit with
inductance L, resistance R and the equivalent capacitance C of
the test cable [15].According to the state of switch S, the
generation of oscillating voltage can be divided into two stages:
First, off state of high voltage switch S. High voltage DC
source, no halo reactor L and test cable C will act as a DC
circuit till the cable reach the preset voltage. In the second stage,
the high voltage switch closed. The high voltage switch S
immediately closed when the preset voltage is reached. The
equivalent capacitance C of the cable, inductor L and the
resistance R formed a LRC underdamping oscillation circuit
and generate oscillating voltage. The decay rate of oscillating
voltage is rather high which only last for about 20-30 ms,
however, it is enough to stimulate the partial discharge signal in
the cable insulation defects without destroy the cable at the
same time. A sensor is used to collect the partial discharge
signal, the cable insulation detection could be accomplished
[16].
Fig. 1. OWTS testing system schematic diagram
B. Defect Making
The testing cable is copper core with cross-linked
polyethylene insulation tape armored PVC sheathed
flame-retardant power cable, the type is ZC-YJV22, with rated
voltage 8.7/15 kV and specifications is 3 * 70 mm2. The testing
cables all are new cable have not been put into operation, before
made detect. In order to guarantee the unicity of test variable,
we made artificial defect model in cable joint after the testing
cable has been confirmed that there is no partial discharge with
OWTS partial discharge test. Three kinds of typical artificial
defects are made separately at the middle of three 8m cables
with cold shrink joint. The armor layers and copper shielding
layers of XLPE cable joint were well-connected. The specific
defects are show in Fig. 2a~2c.
1) Air gap in outer semi-conductive layer defect: the poor
POWERCON 2014 Paper No CP1746
Chengdu, 20-22 Oct. 2014
quality of cable or improper over stripping of outer
semi-conductive layer may cause air gap. As shown in Fig 2a,
this defect can be simulated by scraping 5mm, 10mm, and
17mm gap on the outer semi-conductive layer.
2) Metal cusp at high potential defect: to simulate metal
particle mixed on the core between insulation and compression
coupling by accident, twin copper wire on the surface of core to
form 2mm, 3mm and 5mm tips. It is showed in Fig 2b.
3) Too short or long lap of stress cone defect: simulate the
improper installment that cause too long or short lap of stress
cone and outer semi-conductive layer. As shown in Fig 2c,
phase A represent too short lap of stress cone by 40mm, phase
B, C represent too long lap of stress cone by 12mm, 17mm.
Fig. 2a. Air gap in outer semi-conductive layer defect
Fig. 2b. Metal cusp at high potential defect
Fig. 2c. Too short or long lap of stress cone defect
Fig. 2. The picture of three kinds of defect in cable joint
III. OSCILLATING VOLTAGE TEST
If you are using Word, use either the Microsoft Equation
Editor or the MathType add-on (http://www.mathtype.com) for
equations in your paper (Insert | Object | Create New | Microsoft
Equation or MathType Equation). “Float over text” should not
be selected.
The test cable was connected to OWTS by high voltage
connection cable, the cupper shielding layer joined OWTS
grounding by one-point grounding. Calibration of discharge
quantity is required for the test of partial discharge in XLPE
cable. To get a more accurate apparent discharge magnitude,
the test cable was injected by different quantity of calibration
pulse signal by partial discharge calibrator after the testing
circuit are connected. According to electromagnetic wave
theory, the incident calibration pulse will refracted and
reflected when it reached the end of cable. The calibration of
apparent discharge magnitude could be realized by figuring out
its calibration factor extracting peak value of incident and
reflected impulse.
Page 2/6 1369 Session 2
 2014 International Conference on Power System Technology (POWERCON 2014) Chengdu, 20-22 Oct. 2014

Partial discharge test starts when the calibration is done. IV. TEST RESULTS ANALYSIS
Applied oscillating voltage to main insulation of XLPE cable,
To study the statistical property of partial discharge signal
increase voltage by step and the voltage should be no more than
under oscillating voltage, three-dimensional statistical maps
2.0 U0. Selecting two voltage classes above the partial
which base on statistical method for oscillating voltage phase
discharge inception voltage to collect partial discharge signals.
feature are built for the three defects [17]. It is shown in Fig 4
100 partial discharge tests had been done for each voltage class
-7.
and the data has been saved for further analysis.
Phi-Q-N
Phi-Q-N
Phi-Q-N

20
20
80
15
15 60
10
10 40
5
5 20
0 0
0 0
0 0
1000 360 360 360
270 5000 270 270
2000 180 180 5000
180
90 90 90
Q(pC) 3000 0 Q(pC) 10000 0 Q(pC) 10000 0
Phi(°) Phi(°) Phi(°)

0.9U0 1.3U0 1.5U0

Fig. 3. Phase A of air gap in outer semi-conductive layer defect

Phi-Q-N
Phi-Q-N Phi-Q-N

15 15 100

10 10
50
5 5

0 0 0
0 0 0
360 360 360
2000 270 2000 270 5000 270
180 180 180
90 90 90
Q(pC) 4000 0 Q(pC) 4000 0 Q(pC) 10000 0
Phi(°) Phi(°) Phi(°)

0.7U0 0.9U0 1.3U0

Fig. 4. Phase B of air gap in outer semi-conductive layer defect

Phi-Q-N Phi-Q-N Phi-Q-N

15 60 80

60
10 40
40
5 20
20

0 0 0
0 0 0
2000 360 360 360
270 5000 270 4 1 270
4000 180 180 x 10 180
90 90 90
Q(pC) 6000 0 Q(pC) 10000 0 Q(pC) 2 0
Phi(°) Phi(°) Phi(°)

0.7U0 0.9U0 1.5U0

Fig. 5. Phase C of air gap in outer semi-conductive layer defect

POWERCON 2014 Paper No CP1746 Page 3/6 1370 Session 2

 2014 International Conference on Power System Technology (POWERCON 2014) Chengdu, 20-22 Oct. 2014

Phi-Q-N
Phi-Q-N Phi-Q-N

8
10 20
6
15
4 5 10

2 5

0 0 0
0 0 0
500 360 1000 360 360
270 270 2000
270
1000 180 2000 180 4000 180
90 90 90
Q(pC) 1500 0 Phi(°) Q(pC) 3000 0 Q(pC) 6000 0
Phi(°) Phi(°)

0.9U0 1.3U0 1.5U0

Fig. 6. Phase A of metal cusp at high potential defect

Phi-Q-N
Phi-Q-N
Phi-Q-N
15
15
60
10
10
40
5
5
20

0 0
0 0 0
360 0
500 1000 360 360
270 270
1000 180 5000 270
2000 180 180
90 90 90
Q(pC) 1500 0 Phi(°) Q(pC) 3000 0 Phi(°) Q(pC) 10000 0 Phi(°)

0.9U0 1.3U0 1.5U0

Fig. 7. Phase B of metal cusp at high potential defect

Phi-Q-N
Phi-Q-N Phi-Q-N

15 20 80

10 15 60

10 40
5
5 20

0 0 0
0 0 0
500 360 360 360
270 2000 270 5000 270
1000 180 4000 180 10000 180
90 90 90
Q(pC) 1500 0 Q(pC) 6000 0 Q(pC) 15000 0 Phi(°)
Phi(°) Phi(°)

0.9U0 1.3U0 1.5U0

Fig. 8. Phase C of metal cusp at high potential defect

POWERCON 2014 Paper No CP1746 Page 4/6 1371 Session 2

 2014 International Conference on Power System Technology (POWERCON 2014) Chengdu, 20-22 Oct. 2014

Phi-Q-N
Phi-Q-N Phi-Q-N

20
8 8
15
6 6

4 10
4

2 2 5

0 0 0
0 0 0
500 360 360 2000 360
270 500 270 270
1000 180 1000 180 4000 180
90 90 90
Q(pC) 1500 0 Q(pC) 1500 0 Q(pC) 6000 0
Phi(°) Phi(°) Phi(°)

0.9U0 1.3U0 1.5U0

Fig. 9. Phase A of too short or long lap of stress cone defect

Phi-Q-N Phi-Q-N
Phi-Q-N

2 3
4
1.5
2 3
1 2
1
0.5 1

0 0 0
0 0 0
360 360 200 360
50 270 200 270 270
180 180 400 180
90 90 90
Q(pC) 100 0 Phi(°) Q(pC) 400 0 Q(pC) 600 0
Phi(°) Phi(°)

0.9U0 1.3U0 1.5U0

Fig. 10. Phase B of too short or long lap of stress cone defect

Phi-Q-N
Phi-Q-N Phi-Q-N

20
6 10
15
4
10
5
2 5

0 0 0
0 0 0
500 360 360 2000 360
270 2000 270 270
1000 180 180 4000 180
90 90 90
Q(pC) 1500 0 Q(pC) 4000 0 Q(pC) 6000 0
Phi(°) Phi(°) Phi(°)

0.9U0 1.3U0 1.5U0

Fig. 11. Phase C of too short or long lap of stress cone defect
By compare the three-dimensional statistical maps of same
kind defect, for the defect of air gap in outer semiconductive
layer, the severity positive increase with the length of air gap.
The partial discharge inception voltage drop from 0.9 U0 to 0.7
U0 together with larger magnitude and number. For the defect
of high potential metal tip, its discharge magnitude is little less
than the former one while the defect of stress cone is the lowest
one which has positive increase with the malposition.
By compare the three-dimensional statistical maps of three
different kinds of defects, the developing process as well as the
shape is vary from each other. The higher voltage classes, the
larger magnitude and number of apparent discharge magnitude
in partial discharge. Air gap in outer semiconductive layer and
high potential metal tip, these two defects has rather acute
partial discharge under high voltage classes. The bimodal
characteristics showed up in corresponding maps, which based

POWERCON 2014 Paper No CP1746 Page 5/6 1372 Session 2

 2014 International Conference on Power System Technology (POWERCON 2014) Chengdu, 20-22 Oct. 2014

on the statistical method for oscillating voltage phase feature. is [13] Xia Rong Zhao Jiankang Ouyang Benhong. Dielectric property
a result of the amplitude in different period changed its phase detection of 110 kV XLPE power cables using damped AC
voltages[J] High Voltage Engineering 2010 36(7) 1753-1760
when the oscillating voltage attenuation. For the defect of stress
[14] Feng Yi Liu Peng Xu Mingtao. Application of oscillating wave testing
cone, it hasn’t show bimodal characteristics because the system in the cable partial discharge detection[J] Distribution&
apparent discharge magnitude and nuber of partial discharge is Utilization 2009 26(3) 57-59
quite small. [15] Gulski E, Wester F. J, Smit J. J. Advanced partial discharge diagnostic of
MV power cable system using oscillating wave test system [J]. Electrical
Insulation Magazine, IEEE, 2000,16(2):17-25.
V. CONCLUSION [16] Gulski E, Smit J. J., Seitz P. N. PD measurements on-site using oscillating
This paper make three artificial defects in the joint of 10KV wave test system[C]. Electrical Insulation, 1998. Conference Record of
the 1998 IEEE International Symposium on, Year,2(420-423.
XLPE, oscillating voltage is applied to the test cable and partial [17] CHANG Wenzhi, GE Zhendong, SHI Xiang. Statistical Characteristic and
discharge signal which caused by defect has been detected. The Location of Partial Discharge Caused by Typical Defects in Power Cable
three-dimensional statistical maps for partial discharge signal Under Damped Oscillation[J] Voltage Power System Technology,
2013,37(3):746-752.
are made, statistical property of partial discharge under the
oscillating voltage has been studied, the conclusion are as
below:
1) Oscillating voltage can stimulate the partial discharge of
defect in the cable joint, and can detect it effectively.
2) Partial discharge will become severe when the oscillating
voltage amplitude become larger, obvious bimodal
characteristics can be seen in the three-dimensional statistical
maps at the same time.
3) The distinguish differences lie in three-dimensional maps
of three defects may lay a foundation for the further research in
pattern recognition.

REFERENCES
[1] SUN Guanghui, ZHUANG Xiaoliang, NIU Haiqing. Statistical analysis of
10 kV distribution cable fault in the pearl river delta region
Electrotechnical Application, 2013,11:023.
[2] Working Group 21.09. After laying tests on high voltage extruded
insulation cable system [J]. Electra 1977 (173) 33-41
[3] S. Boggs,J. Densley. Fundamentals of partial discharge in the context of
field cable testing [J]. Electrical Insulation Magazine, IEEE,
2000,16(5):13-18.
[4] LUO Junhua, CHEN Shouzhi, STUDY ON TESTMETHODSOF XLPE
POWER CABLE RATED UP TO 35 kV[J] Power System Technology,
2000,24(12):58-61.
[5] Luo Junhua,Yang Liming,Shi Jikang. Review of power cable and test
technology[J] High Voltage Engineering 2004 30(S1) 81-82.
[6] Xing Deqiang Detection technology of XLPE power
cables[J] Transactions of China Electro technical Society 2006
21(11) 7-8.
[7] Petzold F, Beigert M. Experiences of PD Diagnosis on MV Cables using
Oscillating Voltages (OWTS) [C]. Transmission and Distribution
Conference and Exhibition: Asia and Pacific, 2005 IEEE/PES, 1-7.
[8] B. Oyegoke, P. Hyvonen, M. Aro. Selectivity of damped AC (DAC) and
VLF voltages in after-laying tests of extruded MV cable systems [J].
Dielectrics and Electrical Insulation, IEEE Transactions on, 2003,
10(5):874-882.
[9] International Electrotechnical Commission IEC 60720 Third edition
[S] Geneva 1998 .
[10] CHEN Chengwei Study of 10kVXLPE cable detection using oscillating
voltage [D]. South China University of Technology, 2010.
[11] Zhang Hao Tang Jiating Zhang Lizhi. Typical case analysis of
oscillating wave test system in cable partial discharge test and
localization in power cable [J] Electrical Apparatus 2008 9(12)
31-34
[12] Tao Shiyang PD detection technique in XLPE power cables based on
oscillating wave test system[J] Electric Power 2009 42(1) 98-101

POWERCON 2014 Paper No CP1746 Page 6/6 1373 Session 2