Research Journal of Applied Sciences, Engineering and Technology 10(8): 903-913, 2015
ISSN: 2040-7459; e-ISSN: 2040-7467
© Maxwell Scientific Organization, 2015
Submitted: January 28, 2015
Accepted: February 24, 2015
Published: July 20, 2015
High Frequency Tan Delta Measurement Method for 132kV Transmission
Underground Cables
A.R. Avinash, Chandan Kumar Chakrabarty and Navinesshani Permal
Universiti Tenaga Nasional, Selangor, Malaysia
Abstract: Tangent Delta is a measurement technique to investigate cables insulation strength. Current techniques
utilize Very Low Frequency (VLF) at 0.1 Hz and power frequency at 50 Hz. However, high voltages are required,
thus requiring larger space and cost. Proposed method of tangent delta testing utilizes High frequency Low voltage
diagnoses. The phase between the current and the voltage is utilized to determine the tangent delta (tan δ). The aim
of this study is to develop a low voltage high frequency tangent delta measurement method and test if it can
discriminate manufactured 132 kV good conditioned cable sample from defect induced cables with void, scotched
and contamination in its insulation. Impurities are clearly discriminated using this method. Comparison of Tangent
Delta of cables manufactured simultaneously in good condition and defect induced is performed using High
Frequency Tangent Delta method and in 50 Hz conventional method to validate the effectiveness of the
measurement technique. The High Frequency AC setup utilizes a small testing environment which can sample small
lengths with minimum 1 m length of cable. The small lengths will result in the reduction of total capacitance of the
cable but using High Frequency induces high electric stress on XLPE layer thus resulting in measureable dielectric
current.
Keywords: 132 kV Cables, dissipation factor, HFAC, high frequency, tangent delta, XLPE
INTRODUCTION
Current Tangent Delta measurement techniques
utilize Very Low Frequency (VLF) at 0.1 Hz and power
frequency at 50 Hz. On the contrary, high voltages are
required, thus requiring larger space and cost. This
method of tangent delta testing utilizes High frequency
Low voltage diagnoses and discriminates good cables
from faulty ones (Gudmundsdottir et al., 2010). Hence,
this technique is more economical to assess dielectric
insulation condition for any lengths of transmission
cables in the laboratory.
Electromagnetic studio simulation (EM simulation)
was used to study the contribution of void on the
electric field distribution. Comparison between a good
insulation and a void induced insulation is made. Cable
Suite simulation was used to simulate the sufficiency of
dielectric current obtained to confirm the validity of the
experimental setup.
Following stage was the development and setup of
the high frequency tan delta measurement system. This
consists of the acquiring and setting up of the current
monitor, high voltage probe, high-speed high-voltage
power
amplifier,
function/arbitrary
waveform
generators and a screen cage in the lab.
Parallel to the setup is the preparation of the good
and defective cables. Contamination and void
embedded in the insulation of the cable were
manufactured to simulate a intermittent controlled
Fig. 1: Equivalent circuits of (a) an ideal capacitor and (b)
capacitor with dielectric loss
defect. Scotched cables were manufactured to represent
a severe defect. Factory acceptance test was conducted
on the good and defective cables to discriminate the
cables. The 132 kV 400 mm2 XLPE cable samples were
then subjected to High Frequency Tangent Delta
testing. Discrimination of intermittent and gross
defective cables from good cables is clearly visible with
good sensitivity.
Figure 1a shows the phase shift for a perfect
capacitor. Ideally Current leads Voltage by 90°
resembling that only pure capacitive current is present.
Figure 1b shows that current and voltage are no longer
shifted 90°. This is due to the presence of the resistive
current through the insulation because the resistivity of
the insulation has degraded (Fothergill et al., 2011;
Ponniran and Kamarudin, 2008; Hernandez-Mejia
et al., 2009). This may be the result of impurities like
voids, contamination or scotching effects on the
insulation.
Corresponding Author: A.R. Avinash, Universiti Tenaga Nasional, Selangor, Malaysia
903
Res. J. App
App. Sci. Eng. Technol., 10(8): 903-913, 2015
Fig. 2: Equivalent circuit of tan delta measurement
easurement method of a cable
Fig. 3: Equivalent modal of RS
The degree of change of phase shift, θ represents
the tan delta that depicts the level of aging in the
insulation. The equivalent circuit of tan delta
measurement method of a cable is shown in Fig. 2.
RP represents the resistance of the cable insulation
and C is the capacitance. Impurities in the insulation
reduce the resistance of the insulation thus increasing
resistive current through the insulation (Ponniran and
Kamarudin, 2008).
A series resistance RS with the cable insulation is
formed due to a potential drop across the cable. This is
influenced by the increase of frequency as displacement
current increase as well. RS is the combination of RSC
(series resistance of semiconductor)
emiconductor) and RCOPPER (series
resistance of copper i.e., metallic sheath and core) as
shown in Fig. 3 above. Since RSC is much larger than
RCOPPER therefore RS≈RSC (Hernandez
(Hernandez-Mejia et al.,
2009; Phung et al., 2008; Martí, 1988).
Hence, this study is to develop new technique for
the measurement of tan delta at Low Voltage High
Frequency (LV-HF)
HF) for 132 kV XLPE cable. The study
will provide an alternative means to practically and
effectively measure the tan-delta
delta of high voltage power
cable. This method off tangent delta testing utilizes High
frequency Low voltage will be able to and discriminates
good cables from faulty ones. Hence, this technique is
more economical to assess dielectric insulation
condition for any lengths of transmission cables in the
laboratory.
Fig. 4: Methodology
Tangent delta equation: From Fig. 2, equivalent
impedance of cable is:
METHODOLOGY
Z In this project, the research methodologies are as
Fig. 4 below:
Z 904
∗
(1)
(2)
Res. J. App
App. Sci. Eng. Technol., 10(8): 903-913, 2015
3 2 Z ²²²
²
²²²
(3)
The resistance and reactance components are separated:
!" #'" ²²²
(
( ( (
#$%&"
)*%+"
Since tan " =real " / imag "
²
,%'" ( (
(
,%'" Computer Simulation Technology Studio (CST EM)
was used to model to perform electric field simulation.
The cable 3D model is shown in Fig. 5 above. The
structure in the model was created based on all the
characteristics of 132 kV cable. The properties of the
cable and the boundary conditions were pre-set
pre
and
then a voltage source inserted between the core and
sheath as the excitation source. The arrow indicates the
oscillating voltage source. The applied source voltage is
1kV voltage and 10 kHz frequency. The outer sheath is
grounded.
Two cases for the electric field vector and scalar
plots were simulated which are cable without air void
and cable with air void (as discussed below). This
simulation was performed to show the effect of the
electric field and localized perturbation
ation by the presence
of void. Figure 6a and b shows the electric field vector
and scalar plots for the cable with no air void.
(5)
(7)
-
²
Since ,%'. Fig. 5: The cable model for the electric field simulation
(6)
- ,%'" (4)
(8)
3
/012
4567
/012
Since Rp>>Rs
,%'" (9)
/012
where,
Z = Impedance
Rs = Potential drop which forms a series resistance
Rp = Resistance of the cable insulation
8 = Angular frequency (2πf)
C = Capacitance
" = Phase between current and voltage
V = Voltage
I = Current
. = Dissipation angle
Cable with no air void: The results show clear
azimuthal field uniformity. Figure 7 below shows that
the field exhibits inverse relationship in which the field
gradually decreases towards the sheath.
Cable with air void: Figure 8a and b shows the electric
field vector and scalar plots for the cable with air void.
The void has 1 mm in diameter and located about 20
mm from the center of the core.
The results clearly show that the electric field
distribution is perturbed due to the presence of the air
void. A localized hot spot is also clearly visible in the
scalar plot as shown in Fig. 8b at the location of the
void. In order to depict this, the electric field vs radial
RESULTS AND DISCUSSION
Field and dissipation current simulation (CST EM
and CABLE): For the simulation, the Electromagnetic
(a): Electric Field 3D Vector Plot without air voids
905
Res. J. App. Sci. Eng. Technol., 10(8): 903-913, 2015
(b): Electric Field 3D Scalar Plot without air voids
Fig. 6: Electric field vector and scalar plots for the cable with no air void
Electric field (V/m)
5.00E+05
4.00E+05
3.00E+05
2.00E+05
1.00E+05
Electric field
-02
1.6
1E
-02
1.9
1E
-02
2.2
7E
-02
2.4
1E
-02
2.5
8E
-02
2.8
0E
-02
2.9
7E
-02
2.2
2E
-02
1.3
8E
1.1
9E
-02
0.00E+00
Radial distance (m)
Fig. 7: The plot of electric field vs. radial distance for the cable with no air void
distance in Fig. 9 explicitly shows the electric field
magnitude actually peaked at about 2 cm from the core
thus justifying effect of the void on electric field
distribution.
cable manufacturer. The cable samples consist of the
following:
Good cable preparation: The head is where 3
extruded layers are combined. Nitrogen is introduced as
inert gases to reduce the chances or corrosion at high
temperatures during curing.
Cable suite simulation for charging current: Cable
suite setup is shown in Fig. 10 below.
The comparison of current results at power
frequency is shown in Fig. 11a and High Frequency is
shown Figure 11b to d. The results for 132 kV from 10
to 100 Hz represents power frequency meanwhile
results at 2, 6 and 10 kV, represents High Frequency
simulation.
Figure 11, it is shown that the charging current for
2 kV, 6 kV, 10 kV at 1 kHz is 1.53, 4.36 and 7.20 mA
and is comparable with 132 kV 50 Hz which is 10.5
mA. From the simulation results, it seems that 1kV 10
kHz seem to be in good agreement with 132 kV 50 Hz.
Nevertheless, this simulated result is verified with
laboratory experiment.
Scotched cable preparation: Temperature of Zone 4
to 6 of Extruder 3 for outer semiconductor extrusion is
ramped to 150°C to create a scotching effect as shown
in Fig. 12.
Void cable preparation: The pressure of the XLPE
extrusion curing oven has been reduced from 10 to 4
bars to introduce void in the XLPE region.
Contaminated cable preparation: Parameters are the
same as good cable but filter is changed to 20 holes per
cm² type to create to allow wood ash contamination to
be positioned into XLPE extruder as shown in Fig. 13.
The
Tangent
Delta
measurement
using
conventional 50 Hz was conducted as Factory
Acceptance Test. Figure 14 shows that the order of
132 kV Sample preparation: Four types of 132 kV
400 mm2 Cu XLPE cable samples were prepared by
906
Res. J. App
App. Sci. Eng. Technol., 10(8): 903-913, 2015
(a): Electric Field 3D Vector Plot with air void
(b): Electric Field 3D Scalar Plot with air void
Fig. 8: Electric field vector and scalar plots for the cable with air void
Electric field (V/m)
5.00E+05
4.00E+05
3.00E+05
2.00E+05
1.00E+05
Electric field
-02
1.6
4E
-02
1.9
1E
-02
2.1
1E
-0 2
2.3
5E
-02
2.6
2E
-02
2.8
7E
-02
3.0
9E
-02
3.3
1E
-02
1.3
8E
1.1
3E
-02
0.00E+00
Radial distance (m)
Fig. 9: The plot of electric field vs radial distance for the cable with air void
Tangent Delta with Scotched being the worst followed
by Contamination Void and Good.
•
•
•
•
HF EXPERIMENT SETUP
Figure 17a to c shows the current vs. frequency for
2 kV, 6 kV and 10 kV.
Based on the figures above, at 2kV, scotched cable
is clearly differentiated from the rest. The dissipated
current measured for 2 kV, 6 kV and 10 kV for the 132
kV 400 mm² cable shows good<contamination<void
<scotching. At 6 and 10 kV, defect cables are clearly
discriminated from
rom good cable. However, current
reaches amplifier limit of 40 mA for 6 and 10 kV when
The laboratory test was carried out using the newly
developed tan delta measurement method. The setup is
shown in Fig. 15 and 16.
The equipment for HF Tan Delta Test consists of:
•
•
•
Oscilloscope
Waveform generator
Screen cage
132 kV 400 mm² single core cable (1 m length)
High voltage power amplifier
Current monitor
High voltage probe
907
Res. J. App. Sci. Eng. Technol., 10(8): 903-913, 2015
Fig. 10: Circuit diagram for simulation using CST cable
Fig. 11a: Current vs. frequency results for 132kV from 10 to 100 Hz
Simulated current (mA)
12.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
8.00
6.00
4.00
2.00
2.0
0
1.5
0
0
0.0
Frequency (Hz)
1. 0
0
0.00
2. 5
0
2.0
0
1.5
0
1.0
0
0. 5
0
0.00
0.0
0
Simulated current (mA)
10.00
Current (mA)
Current (mA)
Fig. 11b: Current vs. frequency results for 2kV from 1 to
3 kHz
0.5
0
8.00
3.5
0
Frequency (kHz)
Frequency (kHz)
9.00
3.0
0
0.0
0
0 00
00
00
00
0 .0
10
90 .
00
.0 0
80
00
.0 0
70
60
.0 0
50
.0 0
00
00
0
0 00
.0 0
40
00
3 0.
. 00
20
1 0.
0 00
0
0.00E+00
2.5
0
5.00E+05
2.50
2.00
1.50
1.00
0.50
0.00
2.0
0
1.00E+05
1.5
0
1.50E+05
Simulated current (mA)
1.0
0
2.00E+05
5.00
4.50
4.00
3.50
3.00
0.5
0
l
Current (mA)
Current magnitude (A)
2.50E+05
Frequency (Hz)
Fig. 11c: Current vs. frequency results for 6kV from 1 to 2 kHz
908
Fig. 11d: Current vs. frequency results for 10kV from 1 to
1.6 kHz
Res. J. App
App. Sci. Eng. Technol., 10(8): 903-913, 2015
(a)
(b)
(c)
Fig. 12: (a): Good semicon layer; (b): Scotched semicon layer
layer; (c): Scotched XLPE layer
3.000E-02
4.100E-03
2.500E-02
3.800E-03
2.000E-02
3.500E-02
Tangent delta
Tangent delta
Fig. 13: Wood dust particle inserted into machine to create contaminated cable
Good
Conversation
Void
Scotching
1.500E-02
1.000E-02
3.200E-02
2.900E-02
Good
Conversation
Void
2.600E-02
5.000E-03
2.300E-03
2.000E+03
60
55
50
45
40
35
30
25
20
15
10
5
0.000E+00
5
Voltage (kV)
10 15 20 25 30 35 40 45 50 55 60
Voltage (kV)
Fig. 14: (a and b) Tangent delta 50 Hz results (FAT)
Fig. 15: Laboratory setup for HF tan delta measurement system
using scotched cable (signal clipping), therefore the test
could not be continued for scotched cable at 6 kV and
10 kV.
Figure 18 shows the Current vs. Frequency
simulated vs. measured for (a) 2kV (b) 6kV (c) 10kV
good cables.
909
Res. J. App
App. Sci. Eng. Technol., 10(8): 903-913, 2015
Fig. 16: Equipment setup
Current mA
40.00
14.00
Good
Void
Conversation
Scotched
Good
Void
Conversation
12.00
Current (mA)
50.00
30.00
20.00
10.00
8.00
6.00
4.00
10.00
0.00
0.80
2.00
0.00
1.30
1.80
Freq (kHz)
20
2.80
2.30
0.80
1.00 1.20
Good
Current (mA)
Void
15
Concentration
10
5
Frequency (Hz)
Fig. 17: Current vs. Frequency for; (a): 2kV
2kV; (b): 6kV; (c): 10 kV
910
1. 7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0. 9
0.8
0
1.40 1.60 1.80
Freq (kHz)
2.00
2.20
Res. J. App. Sci. Eng. Technol., 10(8): 903-913, 2015
The best resolution of tan delta between all the
cables is clearly seen at 6kV at 1 kHz as shown in
Fig. 20.
50 Hz tan delta test was performed at 5 kV and 10
kV. Whereas, HF tan delta (1 kHz) was performed at 2,
6, 10 kV, respectively. Based on Fig. 21, at 50 Hz tan
delta values are poorly discriminated between all
the cables at 5 kV and 10 kV. At 1 kHz tan
delta values are clearly discriminated between all
the cables at 6 kV and10 kV. Discrimination is much
better using HF tan delta measurement method
developed.
5.00
Measured current (mA)
4.50
Simulation current (mA)
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50
Frequency (Hz)
12.00
Measured current (mA)
Simulation current (mA)
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
0.00
0.50
1.00
1.50
Frequency (Hz)
2.00
2.50
Measured current (mA)
Stimulation (mA)
10.00
Current (mA)
9.00
8.00
Current (mA)
Current (mA)
The measured current and simulated current are
very close to each other for good cable comparison.
Figure 19a to c shows the tan delta vs. frequency
for 2, 6 and 10 kV, respectively.
At 6 kV and 10 kV, defect cables are clearly
discriminated from good cable. The scatter plot for tan
delta 1 kHz at 6 kV and 10 kV able to discriminate
tangent delta value of contamination>void>good cable
samples.
The newly developed HF tan delta measurement
method able to discriminate tan delta value of
contamination>void>good cables at 6 kV and 10 kV.
8.00
6.00
4.00
2.00
0.00
0.00
0.50
1.00
1.50
2.00
Frequency (Hz)
Fig. 18: Current vs. Frequency simulated and measured for; (a): 2kV; (b): 6kV; (c): 10kV good cables
Good
Void
Contamination
Scotched
Good
1.60E-01
0.140
Void
Contamination
Scotched
1.40E-01
0.120
Tangent delta
Tangent delta
1.20E-01
0.100
0.080
0.060
0.040
1.00E-01
8.00E-02
6.00E-02
4.00E-02
0.020
2.00E-02
0.000
0.80
0.00E-00
0.80
1.30
2.30
1.80
Frequency (kHz)
2.80
3.30
911
1.00
1.20
1.40
1.60 1.80
Frequency (kHz)
2.00
2.20
Res. J. App. Sci. Eng. Technol., 10(8): 903-913, 2015
Good
Contamination
Void
0.08
Tangent delta
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
Frequency (kHz)
Fig. 19: Tan Delta vs. Frequency for (a): 2kV; (b): 6kV; (c): 10kV
Good
Void
Contamination
Scotched
2.00E-01
Tan delta
1.50E-01
1.00E-01
5.00E-02
5.00E-01
5
0
15
10
Voltage (kV)
Fig. 20: Tan delta vs voltage at 1kHz
Tan delta
Good 50 Hz
Contamination 50Hz
3.700E-03
3.600E-03
3.500E-03
3.400E-03
3.300E-03
3.200E-03
3.100E-03
3.000E-03
2.900E-03
2.800E-03
2.700E-03
2.600E-03
2.500E-03
2.400E-03
2.300E-03
2.200E-03
2.100E-03
2.000E-03
50Hz at 5kV
0
2
6
4
50Hz at 10kV
8
10
12
Voltage (kV)
Fig. 21: Tan delta vs. voltage at 50 Hz
These cable samples were tested with the newly
developed low voltage high frequency tangent delta
measurement method. The new tangent delta
measurement proves that phase shift angle by charging
a small capacitance cable of 1 m length can be used to
obtain tangent delta.
CONCLUSION
The 132 kV good conditioned cable sample and
several types of defect were manufactured and tested
using the new developed high frequency tangent delta
system.
912
Res. J. App. Sci. Eng. Technol., 10(8): 903-913, 2015
Measurements for 132 kV cable are attainable
using frequency range of 1-5 kHz and applied voltage
ranging from 1 kV-10 kV. The optimum was found to
be at 1 kHz. At 6 kV and 10 kV, defect cables are
clearly discriminated from good cable. The scatter plot
for tan delta 1 kHz at 6 kV able to discriminate tangent
delta value of contamination>void>good cable samples.
Thus this newly developed HF tan delta
measurement system can be used as an indicative test to
pre-determine the condition of 1 m cable in the
laboratory. By developing a correlation method, the tan
delta results in HF can be correlated to 50 Hz tan delta
values in the future.
Hernandez-Mejia, J., J. Perkel, R. Harley, N. Hampton
and R. Hartlein, 2009. Correlation between tan δ
diagnostic
measurements
and
breakdown
performance at VLF for MV XLPE cables. IEEE
T. Dielect. El. In., 16(1): 162-170.
Martí, L., 1988. Simulation of transients in
underground cables with frequency-dependent
modal transformation matrices. IEEE T. Power
Deliver., 3(3): 1099-1110.
Phung, B.T., Z. Liu, T.R. Blackburn, P. McMullan and
G. Burgess, 2008. Practical experience in on-line
partial discharge monitoring of power cables.
Proceeding of the Australasian Universities Power
Engineering Conference (AUPEC '08), pp: 1-5.
Ponniran, A. and M.S. Kamarudin, 2008. Study on the
performance of underground XLPE cables in
service based on tan delta and capacitance
measurements. Proceeding of IEEE 2nd
International Power and Energy Conference
(PECon, 2008), pp: 39-43.
REFERENCES
Fothergill, J.C., S.J. Dodd, L.A. Dissado, T. Liu and
U.H. Nilsson, 2011. The measurement of very low
conductivity and dielectric loss in XLPE cables: A
possible method to detect degradation due to
thermal aging. IEEE T. Dielect. El. In., 18(5):
1544-1553.
Gudmundsdottir, U.S., C.L. Bak and W.T. Wiechowski,
2010. Modeling of long high voltage ac
underground cables. Ph.D. Thesis, Seminar on
Detailed Modelling and Validation of Electrical
Components and Systems 2010 in Fredericia,
Denmark.
913