Practical aspects of PD localization for long length

Power Cables
M. Wild, S. Tenbohlen E. Gulski, R. Jongen F. de Vries
University of Stuttgart onsite hv technology AG Liandon member of Alliander N.V.
Stuttgart, Germany Luzern, Switzerland Duiven, The Netherlands
[email protected]

Abstract—The single sided PD measurement on power cable activity with the conventional standardized PD detection
circuits is a common used technique for onsite PD detection and synchronized with the damped AC test voltage and
localization. Time domain reflectometry (TDR) of the measured synchronized with the PD measurement at the near end of the
pulses is generally implemented for locating the PD pulse origin. cable. This two sided measurement provides higher detection
To get higher sensitivity in PD measurements and increased sensitivity, especially in the case of long cable lengths. This
localization achievements a double sided measurement can be technique provides more precise PD measurements and
performed. Crucial for localization is the attenuation of the PD enhances the possibility of localization of the PD origin in
pulse amplitudes as a result of travelling along cable length. For power cables. This method will be further discussed in this
this purpose some examples for classical one-sided TDR and
paper.
double sided measurements are discussed.

Keywords—Partial discharges; PD localization; On-site cable II. SINGLE SIDED MEASUREMENT AND PD ORIGIN
testing; Long power cables LOCALIZATION
The classical single sided measurement technique uses one
I. INTRODUCTION PD detection circuit. A schematic setup is shown in Fig. 1,
After-laying tests of new installed and diagnostic testing of which contains the PD detector, a coupling capacitor and the
service aged distribution power cables are an important issue to cable [2]. On the left side of the cable, the inner conductor of
obtain knowledge about the actual condition of the complete the cable is connected to the coupling capacitor and the DAC
cable system and to prevent breakdowns during service. The system. The far end of the cable has an open end. Fig. 1 shows
application of damped AC (DAC) voltages including a typical setup for energizing and measurement of PD in a
standardized conventional PD detection and analysis is world- power cable.
wide accepted for on-site testing and diagnosis of MV power
cables [3]. This technology is based on the off-line energizing
of a cable section with the possibility of testing with elevated
voltages. The damped AC technology makes it possible to
energize long lengths of power cable with a high capacitance
with a low input power demand [2]. Onsite testing with
damped AC voltages makes is possible to include a
standardized PD detection method, see Fig. 1.
There are different parameters which can influence the
quality of the partial discharge measurement. Especially in the
case of long length medium and high voltage cable systems,
the PD detection sensitivity is a known issue that can be Fig. 1. Measurement setup for single sided PD measurement with DAC
challenging. Also the PD origin localization in long cable voltage excitation and PD detection circuit.
lengths can become more complicated compared to shorter
cable lengths. Due to limited sensitivity, which is affected by The DAC excitation unit exists of an energizing source,
the noise level at on-site situations, the detectable PD which energizes the cable circuit over the resonance inductance
magnitude is an important factor for PD tests. Lind. After reaching the desired test voltage, the resonant
circuit will be closed over the HV switch. The resistive HV
To localize the origin of PD, a common used method is the divider measures the damped sinusoidal voltage.
time domain reflectometry (TDR). Using this technique, a Simultaneously the PD detector circuit records the partial
detectable reflection of the PD pulse from the far end of the discharge activity in the cable.
cable has to be detected by the measurement device.
At the fault location, every partial discharge displaces a
The PD measurement sensitivity and localization chance small amount of electric charge. This process emits two
can be extended by using an additional PD measuring system impulses in the cable, one into each direction away from the
at the second (far) end of the cable. This unit measures PD fault origin. These pulses are propagating through the cable. At
the moment tA the detector recognizes the first pulse, which is only has to travel to the far end to be detected there, so only
travelled directly from the PD origin to near end of the cable. A one time the cable length. For this double sided measurement
second pulse tB can be measured with a time difference ∆t. system a damped AC system for energizing the cable system is
This pulse travelled from the PD location to the far end, is used, see Fig. 3. This system uses a coupling capacitor with PD
reflected, travel’s in the other direction to the near end and can detector A on the left side and a PD detector B at the right side.
be measured a certain time later. This time difference Δt is
crucial for localizing the PD origin.
To calculate the distance XPD between the measurement
device and the PD origin, equation (1) is used, where Lcable is
the overall length of the cable and v the propagation speed
which is dependent on the cable characteristics.

X PD = Lcable − Δt ⋅ v 2 (1)

Especially in noisy environments or with long lengths of

Fig. 3. Setup for two sided PD measurement with localization feasibility.
cable this technique has limitations in locating the PD origin. DAC excitation circuit and synchronized PD detector units.
In particular the reflected pulse at the far end has a longer
distance to travel before reaching the measurement device.
This results in a higher attenuation and consequently A. Basic setup
decreasing pulse amplitude. Depending on the specific cable On both sides of the cable system, PD detectors are
characteristics there is a limitation for the localization of the installed. The detectors have to be synchronized to correlate the
PD origin with increasing cable length. The sensitivity of the measurement data of both sides. Fig. 4 shows the principle
PD localization decreases also with the environmental noise. setup and the distances used for the calculation of PD origin.
Disturbances can occur even if the cable system is not The overall length Lcable of the cable system is divided into two
energized. This sensitivity also depends on the used PD parts, the distance from the left side to the PD origin XA, and
detector measuring range. the distance XB from PD origin to the right side cable end.

Fig. 2. Power cable with PD source emits two travelling waves A and B with Fig. 4. Setup for the double sided PD measurement with fault localization
speed v. The PD detector on the left side can calculate the PD origin with the functionality. tA and tB are the times were the direct travelled partial discharge
aid of TDR analysis. waves are measured. Both PD detectors have to be synchronized for PD origin
localization.
This TDR technique is based on the reflection of the PD
pulse at the cable end. It can be possible, that the reflected
B. Localization functionality
pulse is not detectable anymore at the left side, where the PD
detector is placed. To overcome this problem, the double sided Both measurement units record data during the DAC
measurement can be performed. This technique increases the voltage excitation. Therefore the measurement settings and the
detection distance in which the PD origin localization is measurement data are communicated between the two
possible. measurement units. Furthermore the units are time
synchronized to obtain phase resolved PD patterns at both sides
III. DOUBLE SIDED MEASUREMENT as well a synchronized localization analysis.
Detection and localization of PD in cable system with long
length can be improved by performing PD measurements at
X PD = Lcable 2 − Δt ⋅ v 2 (2)
both sides of the cable circuit [2]. This will for the worst case
situation (PD at the near end) reduce the travelling distance for
PD pulses by a factor 2. In single side measurement setup, a To calculate the PD origin, equation (2) can be used. As in
near end partial discharge has to travel through the whole cable the chapter before, the distance XPD, is the distance from the
length to the far end and the whole cable length back to the left detection unit to the PD origin and Δt the time difference of
near end. The overall travelling distance is therefore two times t2 and t1.
the cable length. In double side measurement the near end PD
 In this particular configuration the PD pulses are directly In the case of a fault location between both ends, an
measured and there is no need to take reflections as is the case example is taken of a PD origin is at 2 km seen from the left
with the single sided TDR evaluation. As both units are side, shown in Fig. 6. Due to the attenuation, the decreasing PD
synchronized, the difference in the arrival times of the pulses at amplitude has about 70 percent of the original amplitude. In a
both sides together with the pulse velocity obtained from the single side measurement setup the reflected pulse over the far
calibration provides the location of the discharging defect. end decreases to 5 percent. A double sided measurement
increases the remaining PD amplitude to about 26 percent.
C. Advantages of the double sided PD measurement
The main advantage of the double sided measurement is the
better sensitivity compared to the single sided measurement.
The PD pulses have to propagate only to the ends of the cable.
No reflection at the far end is necessary. As a result there is no
additional attenuation due to imperfect reflection or attenuation
for travelling through the whole cable once again.
The attenuation decreases the amplitude of measureable
partial discharge pulses. This attenuation can be described with
an exponential function, with two constants c0 and c1 in the
exponent. Equation (3) describes this amplitude decreasing,
with PD0 as initial PD amplitude and x the variable of the
travelled length.

⎛ ⎛1 1 ⎞ ⎞⎟
a PD = PD0 ⋅ exp⎜ − x ⋅ ⎜⎜ + ⎟⎟ (3)
⎜ ⎟
⎝ ⎝ c 0 1 + c1 ⎠ ⎠

As can be seen from Fig. 5 and Fig. 6 the double sided

measurement results in a higher sensitivity of the PD detection
than the single sided measurement. As an example a 10 km Fig. 6. Normalized PD amplitude attenuation with a fault location at 2 km
long cable is simulated with single and double sided seen from the left side. The reflection on the right side is colored in grey.
measurements. For the single sided measurement, normalized
PD amplitude of 100 percent for the near end pulse and a Fig. 5 and Fig. 6 show the improvement of the sensitivity
reflected pulse of about 4 percent could be detected. With a with the double sided measurement. The worst case for
double sided measurement setup, the remaining PD amplitude localization is a PD fault location at the cable end. In such a
on the right side is about 19 percent. situation, the PD pulse attenuation is the highest, due to a
travelling distance of the whole cable length.
The double sided measurement brings advantages in
detecting sensitivity and localization possibilities of PD pulses
on long cable lengths. Due to the use of two PD measurement
units, the hardware is more complex, than compared to the
single side measurement. In particular the synchronization and
data transmission over long distances is a challenge for further
investigations.

IV. EXAMPLES

A. Typical measurement data used for time domain

reflectometry
For partial discharges occurring at the near cable end,
where in case of a single sided measurement the PD detector is
located, the first measurable pulse appears with a fast rise time
and high amplitude compared to its reflection. The second
detectable pulse is time delayed with two times the cable
travelling time with the specific wave propagation velocity.

Fig. 5 Normalized PD amplitude attenuation on both sides of a 10 km long

cable for double sided measurement (black). The reflection for the single
sided measurement is the grey curve.
Fig. 7. Near end PD fault location with reflection from far end on a 500 m Fig. 9. PD fault location at the far end from PD detector and reflection over
cable. near and far end.

In Fig. 7 the PD is located at the near end. After the In Fig. 9 a measurement example for the far end PD is
detection of the first pulse the reflection occurs in the graph shown. In this case both PD pulse amplitudes are already
after circa 3.5 µs. The amplitude decreased due to attenuation decreased due to the attenuation. The rise time of the first pulse
while travelling through the cable. Also the rise time is slower is not much faster than the rise time of the reflected pulse.
than compared to the first direct measured pulse.
B. Double sided PD measurement
This example shows a synchronized and combined double
sided measurement. The measured data from each PD detector
from the left and right side are plotted in one graph.

Fig. 8. PD fault location 200 m from the PD detector with reflection from far
end on a 500 m cable.

Fig. 8 shows the TDR measurement for a PD pulse origin

located 200 m away from the PD detector. The first pulse
detected travelled the 200 meter between the PD location and Fig. 10. Double sided PD measurement data. Detector A is located at the near
end side of the cable and Detector B on the opposite far end side. The PD
the PD detector. The reflection of the PD pulse arrives after pulses are injected at the near end close to Detector A.
about 2.5 µs. With the information of this time delay, the PD
origin can be calculated. Measured data from detector A is colored in black and the
The first pulse in Fig. 8 has a higher amplitude than its data from detector B in grey. The measurement in Fig. 10
reflection, which has the same behavior as in the near end case. shows a PD origin on the near end, located to detector A. After
Different in this case is that the rise time of both pulses is one cable travelling time the pulse reaches detector B. About 3
reduced, due to travelling along the cable for a not negligible µs later the reflection over the far end is detectable at detector
distance. A. This pulse is attenuated due to a travelling length of two
times the cable length.
Fig. 11. Double sided PD measurement data with PD origin at 220 m from
Detector A and 424 m from Detector B.
Fig. 13. Double sided measurement on a 644 m long medium voltage cable.
The data in Fig. 11 is measured in the same way as in
Fig. 10, with detector A on the one side, and detector B on the As it can be seen the PD pulse is reaching the PD detector
opposite side. The PD pulses are generated 220 m from A before detector B. That means the PD location is nearer to
detector A and 424 m from detector B. The first pulses of both the left side than to the right cable end. The time delay between
measurements are the direct travelled pulses. After the double the two pulses is about 0.25 µs. This results in a distance of
cable travelling time the reflections could be observed. circa 45 m of travelling length. It can be concluded, that the PD
It can be concluded, that the double sided PD measurement origin must be 45 m from the middle of the cable to the left
brings a benefit in higher sensitivity and better localization direction.
possibility. The higher sensitivity enables to test longer cables
with the same sensitivity compared to single sided V. CONCLUSION
measurement. Furthermore the decision between near end and
far end PD could be separated easily. In single sided The presented method of a synchronized PD measurement
measurements, the difference between near end and far end PD at both sides of the cable has shown that the sensitivity of on-
origin is not always clear. site measurements improves. In particular it provides a more
precise PD localization on long cable lengths and therefore a
better condition assessment.
C. Measurement on medium voltage cable
This example show a measurement on a 644 m long The single sided measurement in combination with a
medium voltage cable with PD detectors at both ends of the damped AC power source showed good results in the past
cable (Fig. 12). The PD detector A is on the left side of the years with onsite measurements. To improve the setup a
cable and is plotted with black color in Fig. 13. PD detector B second PD measurement unit was introduced and connected to
is placed at the right side and is plotted with grey color. the other cable end.
The double sided measurement is evaluated for improving
on the PD localization, especially in relation to longer cable
lengths. The attenuation of PD amplitudes was considered as a
function of cable length for single and double sided PD
detection. Moreover examples for double sided measurements
with different cable length were discussed. It can be concluded
that based on simulations and measurement the double sided
measurement technique, provides a solution for testing long
MV power cables and providing an increased detection and
localization ability of PD sources within the cable circuit.
Fig. 12. Measurement setup for double sided measurement with PD detector
A on the left side and PD detector B on the right side.
REFERENCES
The PD detectors are synchronized to a time resolution of [1] M. Wild, S. Tenbohlen, E. Gulski, R. Jongen, “Power Cable Modeling
10 ns. This results in an accuracy of about 2 m for locating the for PD Pulse Propagation and Sensitivity”, Electrical Insulation
PD origin. Conference (EIC) 2013, Ottawa.
[2] R. Jongen, B. Quak, E. Gulski, P. Cickecki, F. de Vries, “On-site testing
and diagnosis of long medium voltage cables”, Condition Monitoring
and Diagnosis (CMD) conference, 2012, p. 659-662.
[3] R. Jongen, B. Quak, E. Gulski, S. Tenbohlen, “New developments in on-
site testing of long lengths of (E)HV power cable”, Condition
Monitoring and Diagnosis (CMD) conference, 2012, p. 149-152.