Part III CIGRE TF D1.11/33.03.

02
POWER CABLES
39
PART III POWER CABLES
Contributors:
F.J. Wester, E.Gulski, E. Lemke, T. Okamoto, W. Boone, J.Pellis
3.1 GENERAL
In the last ten years, to determine the insulation condition of the power cable grid, several on-
site diagnostic measuring methods have been
developed [1-16]. As a result, on the basis of on-
site testing and measurements of partial
discharges, important information is generated
about discharging insulation defects. In
particular, applying off-line test the PD has
shown that PD knowledge rules can be defined
to support condition-based maintenance [17-19].
To maintain the assets e.g. power cable in
dependence on their availability important
detailed information on their actual insulation
condition of a cable section is necessary. In
particular, the service life of single cable section
elements has to be evaluated, see figure 3.1.
The history, the operational and the laying
conditions together with the importance of the
cable circuit in the network are necessary
conditions to decide to perform diagnostic
testing. In its turn the main goal of PD diagnosis
is to recognise high voltage insulation problems
and to identify the insulation defect causing the
discharge: e.g. internal or surface discharges,
corona, treeing, etc, see figure 3.2. This
information is vital for estimating the
harmfulness of the discharge.
3.2 CABLE INSULATION DEFECTS
By registration of the disturbances in the power supply by network owners, statistical analysis
can be made to indicate the major failure causes in the medium voltage power network. Figure
3.3 shows an example of statistical overview of the most important failure causes in a
distribution cable network of a utility having almost 95% of the distribution power cable
network exists of paper insulated lead cable (PILC). Approximately 43% of these breakdowns
are caused by non-electrical external causes. This can be damages by digging activities for other
types of underground cables (e.g. telephone, gas) or caused by movement of the soft wet
ground in some parts utility area. But still, more than half of the breakdowns in the cable
Cable Section:
Cable parts
Joints
Terminations
Different Types
Different Ages
Different History
Figure 3.1 Important service life elements of a cables
section.
Figure 3.2 Examples of power cable insulation
defects.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
40
network is caused by internal fault in the
insulation systems of the cable network.
The majority of the electrical faults occur in
the cable joints: 52%. The remaining
breakdowns occur in the cable
terminations (4%) and the insulation of the
cable parts itself (1%). For many years now,
failing joints and terminations are analysed
to find the repetitive fault causes.
Visual inspection of the disturbed components gives insight in the different types of fault causes
resulting in breakdowns. Figure 3.4 shows an example of the analysis of a disturbed mass filled
cable termination, which showed on all three phases sharp edges on the conductor connectors
to the switch gear. On the fault place, two phases were only insulated by the plastic cover of the
termination, resulting in a breakdown after several years, caused by PD activity from the sharp
edges. As a result of these visual inspections, a list of defects in the different elements of cable
network can be made, as reflected in table 3.1.
Table 3.1: Typical insulation defects for different types of power cable components.
Cable type Accessories Insulation
Low oil level Damage outer sheet
Sharp edges on connectors Tracking
Moisture penetration
Air/gas bubbles
PILC
Bad hardened resin
Internal damage (as a result of bending)
Sharp edges on connectors
Moisture penetration
Air/gas bubbles
Field grading movement
Bad hardened resin
Interface problems
XLPE
Remaining semicone
Damage outer sheet
Figure 3.3 Example of statistical distribution of cable
failure causes.
Figure 3.4 Fault cause analysis of disturbed cable
termination. Several sharp edges on the connector
were found on all three phases.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
41
3.3 SELECTION OF ADVANCED DIAGNOSTICS
For many years, DC withstand testing was the only testing method applied in the MV power
cables network, but nowadays also PD testing is applied. However, for some joints (resin filled),
the DC voltage test combined with PD test is the best option. DC test for moisture penetration
and PD test for sharp edges. Moisture penetration in the resin insulation, from the outside to the
inside, does not result always in PD. Because the moistured part of the insulation material is
conducting, the rest (dry part) of the resin is stressed with almost 99% of the applied DC voltage
and breaks down. For the other types of joints and termination, PD testing at AC voltage
stressing is the best diagnostic option, and is applied for almost two years now. As shown in
table 3.2, partial discharges (PD) are sensitive symptoms of possible discharging weak spots
(insulation defects, degradation products) in the HV insulation [19-22].
Table 3.2: Typical insulation degradation processes of the power cable insulation.
Accessories
interface problems PD tracking;
bad hardening cracking PD;
conductors problems overheating cracking PD;
local field concentrations PD
Extruded Insulation water trees electrical trees PD
insulation voids delamination electrical trees PD
local field concentrations PD
Paper/Oil Insulation
oil leaks dry regions overheating PD
water ingress load effects overheating PD
local field concentrations PD
For water-tree aged cables PD detection is not an effective diagnostics tool. Other methods
based on dielectric response e.g. return voltage, dissipation factor, dielectric spectroscopy or
isothermal relaxation current are better analysis methods to be recommended [23].
Already 30 years ago, it has been shown that using analogue PD detectors, a combination of PD
measuring quantities can be used to recognise discharging defects. In particular it was stated
that any discharge larger than a certain q0 measured below test voltage Uo leads to rejection of
the object under test. In addition using phase-resolved PD patterns the recognition of defect is
possible [ref. Cigre Electra No. 11, 1969]. At that time the major goal of using PD recognition
was to find the manufacturing faults during acceptance tests. Nowadays, the situation has been
slightly changed. In the last ten years, due to the fact that:
a) advanced digital PD analysers are available,
b) more knowledge about the materials and ageing processes is available;
c) due to liberalisation of the energy market and the introduction of condition based
maintenance (CBM) there is more need to have diagnostics giving information about
actual insulation condition of components in service.
the use of digital PD measuring techniques for quality assurance in the works, during on-site
testing as well as for monitoring purposes during service life of HV components like cables
have got increasing attention.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
42
Table 3.3: Standard methods for diagnostic testing of distribution power cables.
Voltage source PD detection PD information
50Hz ac energising:
on-line diagnosis
VHF PD detection at power line
frequency voltages
PD activity in the whole circuit at U0 level
20Hz 300Hz 50Hz ac
energising: AC
off-line diagnosis
Standardised (IEC) PD detection
at sinusoidal power line
frequency voltages
PD occurrence and location in the selected
cable section at different voltage levels
0.1Hz ac energising: VLF
off-line diagnosis
Standardised (IEC) PD detection
at sinusoidal 0.1Hz voltage
frequency
PD occurrence and location in the selected
cable section at different voltage levels
50Hz..500Hz Damped ac
voltages energising: DAC
off-line diagnosis
Standardised (IEC) PD detection
at damped sinusoidal voltages at
50Hz..500Hz frequencies
PD occurrence and location in the selected
cable section at different voltage levels
Due to the fact that PD diagnosis of medium voltage cables is used as a non-destructive means
of identifying critical weak points in a length of cable, the following requirements can be
defined for such a diagnostic tool:
a) Non-destructive for the cable insulation and using AC voltage stress conditions;
b) Using standard and derived quantities;
c) Providing distinction between different types of insulation defects;
d) Providing PD site location: PD mapping;
As a result, PD diagnosis has two ultimate purposes
1. to conclude on the condition of the component measured,
2. to recommend
a) about exchange of cable or accessory,
b) to perform a more accurate on-site visual inspection,
c) to retest within a certain time span.
From a physical point of view PD diagnosis tests of power cables should be performed under
electrical stresses most similar to service condition. In order to decrease the capacitive power
demand for energising the cable and hence to reduce the weight and costs of the test
equipment, specific voltage shapes have been introduced and employed since 1985. As a result
the most common applied excitation voltages are classified in table 3.3.
This report discusses systematic field experiences as obtained with all these methods whereas
with regard to knowledge rules generation three methods will get special attention.
3.4 KEY INFLUENCES FACTORS
Due to the fact that different elements can be used during construction of a cable section, each
cable section may represent a separate combination of elements like, see figure 3.1:
a) several different cable joints,
b) two or more different cable terminations,
c) different cable parts between particular joints/terminations.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
43
These elements from the point of view of construction, insulation materials applied as well as
due to different service conditions may age in different ways. In order to achieve a good end
result it is necessary to have a number of sources of information, concerning the cable and
accessories tested, available to support the operator in making a correct interpretation of the
measured PD signals. The information has to be collected from the network owner:
a) Type of cable and accessories
b) Accurate cable maps
c) Year of installation of cable and
accessories
d) Current and past loading of cables
e) Operating history including failure
behaviour
f) Installation conditions
g) Importance of the network
As a result, prior interpreting the diagnostics
results collecting relevant cable section specific
technical information is of great importance,
see figure 3.5. In particular, an example of
similar discharge activity (PD ~ 8.000 pC) as observed in two PILC cable sections is shown. In
the 1
st
case (cable A) the concentrated PD activity indicates in the cable a discharging defect. In
the 2
nd
case (cable C) no concentrations of PD are detected, but the PD activity is scattered over
the full lengths of the two cable parts between 180 and 770 meters. This is temporarily PD
activity (several hours) is caused by temperature/pressure changes, which are typical for
synthetic oil impregnated cable insulation and occurs after disconnecting the cable section, and
is not related to degradation of the insulation material.
Different to PILC cables XLPE and EPR insulated cables show mostly only sharp concentrations
of single PD events, see figure 3.6 [26]. The reason for that is, that the cable insulation is usually
Cable A
Discharges upto 2*U0
0
500
1000
1500
2000
2500
3000
3500
4000
0 100 200 300 400 500 600 700 800
cable length [m]
P
D

m
a
g
n
i
t
u
d
e

[
p
C
]
Cable B
Discharges upto 2*U0
0
100
200
300
400
500
600
700
0 400 800 1200 1600 2000 2400
cable length [m]
P
D

m
a
g
n
i
t
u
d
e

[
p
C
]
Cable C
Discharges upto 2*U0
0
2000
4000
6000
8000
10000
12000
0 150 300 450 600 750 900
cable length [m]
P
D

m
a
g
n
i
t
u
d
e

[
p
C
]
Figure 3.5 Examples of different types of PD location in a distribution power cable
(A) PD concentration in a cable part; in dependence on the PD magnitude as well as the concentration are
conclusions regarding local insulation degradation could be made;
(B) PD concentrations in cable joints; ; in dependence on the PD magnitude as well as the concentration are
conclusions regarding local insulation degradation could be made;
(C) PD concentration in the cable insulation only; due to temperature/pressure changes temporary PD activity
may occur in the cable insulation, no indication on insulation degradation.
Figure 3.6 PD fault position map of a 743 m long PE
cable (15/25 kV) subjected to a damped AC voltage of
29 kV. The local concentration of strong PD-events at
216 m appear in a not correct assembled joint.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
44
PD-free, because it is tested
carefully after
manufacturing in
laboratory. Hence, PD may
only appear in the
accessories due to not
correct assembling work
after laying or longer
service time.
Selecting testing criteria
have to be adapted to cable
specific configuration. It is
known that the XLPE cable insulation has to be PD free. Moreover, due to temporary over-
voltages occurring during network operation the cable at U0 service voltage has to stay PD free.
As a result for the XLPE cable insulation, the PD inception and extinction voltages have to occur
at higher levels than the service voltage U0. Therefore it is of importance for XLPE cables to be
able to detect this situation where due to over-voltage stress a permanent PD activity can be
ignited.
Furthermore from practical point of few due to large diversity in physical accessibility to cable
terminations and the possibility for a suitable sensor location it is of great importance to take
into account the limitations in applying any sensitive PD diagnostics, see figure 3.7.
3.5 RELEVANT PD QUANTITES; Cable
fingerprint
Although the measurements can be carried out
at preferred voltage levels, the following
aspects should be considered:
1. A PD measurement should be performed
at both U0 as well as at 2 x U0. Partial
discharges measured at phase voltage
appear constantly and therefore represent
a great threat.
2. Partial discharges measured with coupled
voltages only appear when there is an
earth fault and then it is a question of how
long the fault will last. A few minutes is
not a problem but several hours duration
is a threat.
3. When examining test results in cases
where the PD inception voltage is < U0, the
growth in PD amplitude and the increase
in PD intensity are of great importance.
4. For PD measurements with inception
(a)
(b)
(c)
Figure 3.7: Due to large construction diversity the accessibility for
diagnostic systems can be different: (a) due to very open construction almost
each of diagnostics systems can be connected (b, c) due to the cable
termination is inside the switchgear the accessibility is limited.
Figure 3.8 Example of a PD Finger print of a MV
power cable as obtained using off-line diagnosis.
Figure 3.9 Example of a PD Finger print of a MV
power cable as obtained using off-line diagnosis.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
45
voltages > U0, the increase in PD intensity
and PD level should be considered in
relation to the permitted values and used as
feedback to maintain non-destructive
voltage levels.
All relevant information from a PD cable measurement should be collected. To describe the PD
process in the cable section under investigation diagnostics method generates several PD
quantities, which can be divided into two groups:
1. basic quantities: e.g. PD level in [pC], [nC], PDIV in [kV], PDEV in [kV], PD spectrum;
2. derived quantities: e.g. q-V curve, phase-resolved PD pattern, PD magnitude/intensity
mappings.
This PD information which can be determined at different voltage levels e.g. up to 2U0 is
collected in a so-called PD fingerprint of a power cable section. Examples of PD fingerprint as
possible to obtain using different diagnostics are shown in figures 3.8-3.11. The characteristic of
these quantities measured on different cable sections may vary in dependence on factors like
type, age, service history, and location of the elements used.
Without any doubt, all diagnostics actions are to be considered as the indispensable final step in
the complete PD field testing process, because those actions should be able to avoid service
failures and consequently to contribute to improved reliability.
3.6 INTERPRETATION CRITERIA
For optimal interpretation of the results obtained from field measurements, it is important to
study the results from several points of view, collected in the cable fingerprint. In particular,
- the distinction between PD activity in the cable section under investigation and the
discharges originating from outside the cable section e.g. switchgear, corona,
- characterisation of different discharging defects by typical PD patterns
are important issues. Combining these interpretation rules the ageing stage of a fault can be
recollected.
PD i ncepti on vol tage;
PD exti ncti on vol tage;
PD magni t udes;
PD i ntensi ty;
PD pat t ern (2D & 3D);
PD mappi ng
Tg o o.
PD mappi ng
PD pat t ern
Figure 3.11 Example of a PD Fingerprint of a MV
power cable section as obtained using off-line
diagnosis.
(a)
(b)
Figure 3.10 Example of a PD Finger print of a MV
power cable as obtained using on-line diagnosis; (a)
PD spectrum (b) PD pattern detected at a selected
frequency e.g. 3.8MHz.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
46
a)
b)
Figure 3-13 Identification and location of a critical PD
source in a pre-fabricated 110 kV XLPE cable joint during
an after-laying test under power frequency AC voltage,
using the directional coupler technology; (a) Visualisation
of the phase-resolved PD pattern of the identified joint; (b)
Output signals of the four ports (A, B, C, D) of the both
integrated directional couplers [26]. Evaluating the time
delay of 1.4 ns the PD source inside the joint could be
pinpointed at a deviation from the real site less than 4 cm.
Comparison of the measured fingerprint with a database of other measured cable sections,
gives the opportunity to identify the criticality of a detected PD source and in this way optimise
the planning of the possible replacements of cable parts or accessories.
In tables 3.4, 3.5 schematic example of interpretation rules for off-line PD measurements on
distribution power cables are shown. In Table 3.7 schematic example is shown for off-line
diagnosis of transmission power cables. These are rules of thumbs supporting the analysis of
the measurement results from a cable section. Different aspects of the fingerprint are used in
these interpretation rules.
Regarding the interpretation criteria of
PD results the knowledge on the
position of critical PD sources is very
important. If PD diagnosis tests of
distribution cables are executed off-line,
the PD-site can be located by means of
the time domain reflectometry (TDR).
A practical example for a 480 m long
XLPE cable is shown in figure 3.12.
As already reported in chapter 3. for
diagnostics of transmission cables (off-
line and on-line) PD sensors can be
integrated in the accessories. Because
under this condition the maximum
reading of the PD instrument is
obtained, if the PD source is close to the
sensor, the location of critical PD events in cable accessories bases on these criteria. A practical
example, which refers to the identification of a PD failure in a prefabricated HV cable joint (110
kV) by using the directional sensor technology [26], is shown in figure 3-13.
Regarding the interpretation rule for the PD location, it is very important to keep in mind how a
cable joint or cable termination is constructed, because the harmfulness of a defect is depending
on the structure of the accessory.
The on-line based diagnosis method is at this moment under investigation and the
interpretation rules can be only made to make a distinction between general PD activity in a
cable circuit and the disturbances, see table 3.6.
a) b) c)
Figure 3.12 Practical example for the location of PD sites in a 480 m long XLPE cable (12/20 kV) by means of
a damped AC test voltage of 12 kVPeak using the time domain reflectometry [27] a) Visualisation of the PD
pattern appearing during the first half-cycle; b) Zoom of a single PD pulse selected from the record a), c)
Created PD fault location map; PD failures in joints at 7m and 452 m
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
47
Table 3.4: Example No. 1 of interpretation rules for an off-line PD diagnostics on power cables.
PD magnitude (depending
on component)
high low
PD value (depending on
component)
high low
PD density (depending on
component)
high low
PD pattern dangerous type of discharges * less dangerous type of discharges *
PD location cable accessory
Network condition insulated neutral solid earth
Installation condition duct direct buried
PD mapping scattered PD concentrated PD
* combining PD patterns to the dangerousness of discharge type requests special experiences
Table 3.5: Example No. 2 of interpretation rules for an off-line PD diagnostics on power cables.
Inception & Extinction
Voltage
> cable operating voltage < cable operating voltage
PD magnitude < typical values (e.g. < 4000 pC) > typical values (e.g. < 4000 pC)
PD intensity Low (e.g. < 5 pulses/period) High (e.g. < 5 pulses/period)
PD pattern harmful fault type * less harmful fault type *
PD location cable insulation cable accessories
PD mapping PD concentration scattered PD location
* identifying the harmfulness of discharge type requests special experiences
Table 3.6: Example of interpretation rules for on-line PD diagnostics on power cables.
VHF PD Spectrum typical for PD activity typical for disturbances
PD pattern typical for PD activity
typical for disturbances and cross-talking
effects *
* analysing cross-talking effects requests special experiences
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
48
Table 3.7: Example of interpretation rules for combination of tan and PD diagnosis to assess and
classify the condition of PILC cable diagnostics on transmission power cables.
Cable category PD at U0 [pC] tan o o at U0 [x10
-4
] remarks
A PD s 250 and s 30 -
B PD s 1000 and s 50 -
C
PD s 1000 or
Pd s 5000 + no
localisation
and
tan o > 50
tan o s 80
Follow trend
D
Pd > 1000 +
localisation
and tan o s 80
Inspect/replace PD
location
E
PD > 5000 + no
localisation
and/or tan o > 80
Think about cable
replacement
3.7 PRACTICAL EXPERENCES;
Distribution Power Cables
3.7.1 Interpretation at Off-line Conditions
Relevant aspects, regarding the
interpretation of the measurement data are
discussed based on systematic examples
from field measurements [21].
Especially for XLPE power cables it is of
importance to know in the relation to U0 of
the cable the voltages PDIV, PDEV. If the
voltage PDEV is lower than the operation
voltage of the XLPE cable section, PD can
occur continuously during service of the
cable section, see figure 3.14. The detected
PD source at PDIV (10kV = 1.1*U0) has been
located in phase L2 in the cable
terminations.
Due to decreasing voltage at oscillating the
PDEV can be determined. In this case, the
PDEV is 0.9*U0, lower than the nominal
voltage. This means that in a case of just a
small over-voltage, caused by for example
switching, the PD source will be ignited and
will be discharging continuously during
Westerh-317994
PD mapping up to 2*U0
0
1000
2000
3000
4000
5000
0 286 572 858 1144 1430 1716 2002 2288 2574 2860
[m]
[p
C
] L3 joint at 1268m:
epoxy resin type
L1 joint at 1301m:
oil filled type
L2 joint at 1301m:
oil filled type
Figure 3.15 PD diagnosis of a 22860m long 10 PILC
cable showing discharge activity in different types of
splice L3 located in the joint at 1268m. and splices
L1 and L2 located in the joint at 1300m.
PDIV = 1.1*U
0
PDEV = 0.9*U
0
Figure 3.14 PD measuring result where PDIV = 1.1
U
0
is just above U
0
and PDEV = 0.9U
0
is just below
U
0
.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
49
operation.
Figure 3.15 shows the measuring results
on a 3-phase PIL cable section with two
aged joints. The PD pattern of phase L3
differs from L1 and L2; in L3 PD occurs in
the negative as well as the positive side of
the oscillating voltage. As known this
refers to different PD source types, which
is true in this particular case.
The PD mapping of the cable section in
figure 3.15, shows that the detected PD is
originating from two different types of
joints at 1260 and 1300 m from the
measuring side.
Due to the fact that for PILC insulation a
certain PD level/intensity is allowed it is
of importance to determine the criticality
limits for PD activity. After discharges
cross such a limit maintenance has to be
performed resulting in improvement of
insulation condition. In figure 3.16 an
example of PD mappings made before and
after maintenance indicated by bad
insulation condition (too high PD activity
in a number of splices) shows that evaluation PD levels in combination with PD intensity
can be used to support the maintenance activities.
To detect and to locate in power cables insulation discharging weak-spots, the principles of
time dominie reflectometry (TDR) are generally used. In particular, using 1 sensor the actual
discharge source location can be derived from the differences in arrival times of the first and
reflected second pulse, see figure 3.17a. Due to attenuation and dispersion of the PD pulses
travelling through the cable system the
maximum length of cable circuits where the PD
Testlocation
Saratoga St.
MH 1399/2
Testlocation
Chestnut St.
MH 330/16
Testlocation
Power Plant
Branched
Network
Figure 3.18 Multi-terminal diagnosis of a branched
distribution power cable network.
reflection
based
multi-
sensor
based
defect
sensor
1
2
3
defect
1
2
3
time
(a)
(b)
(c)
Figure 3.17 Principals of signal processing of
discharging defects in branched distribution power
cable network. (a) single sensor PD detection on no-
branched cable section; (b) synchronized multiple
sensor PD detection on a branched cable network;
(c) PD pulse shapes as detected in (a) and (b).
PD activity before the maintenance
PD activity after the maintenance
Figure 3.16 PD mappings as made for a 3200m
long 10kV PILC cable before and after
maintenance: refilling all oil splices and
replacement of one of the splices.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
50
No PD present
Phaseu
(a) (a)
(b) (b)
(c) (c)
Figure 3.19: Typical three types of PD pattern to obtain using on-line diagnosis: (a) No PD present at
U
0
; (b) Disturbances at U
0
; (c) PD present at U
0.
activity can be located is about 4 km. With such a length of a cable the smallest measurable
discharge s are in the range of 100pC to 1000pC depending on noise conditions. By applying
sensors on both sides of the cable circuit the measurable cable length can be extended to 8
km. This increase is possible because the reflected partial discharge pulses do not have to
travel up and down the cable anymore before they will be measured.
In a branched cable circuit sensors can be placed at several cable branch ends. Each of the
sensors will receive partial discharge pulses that can be used to locate the discharge sources.
The principles are similar to that of a long cable discussed above. Figure 3.17b shows the
principle set-up in the case of working with three sensors.
A special case is a circuit with many branches, see figure 3.8. Here the diagnostic test can be
done in two or more rounds. In a first round of testing, the weakest branches can be
identified with only a few sensors (three or four), installed at the end of the longest
branches. Weak branches without sensors have discharges that are seen on the T-splices
only. In case of a dangerous discharge activity level at such a T-splice, this branch should be
evaluated more closely by having a sensor at the end of this branch too in a next round of
testing [24].
3.7.2 Interpretation at On-line Conditions
The diagnostics presented in section 3.7.1 work off-line, which means that during the test
the cable section has to be disconnected from the network and energized by an external
voltage source, e.g. 50 Hz AC, VLF or damped AC voltages. In that way, applying to the
cable sample AC test voltages between PD inception voltage (PDIV) and 2U0 the discharge
behavior of cable insulation can be analyzed to obtain most complete information about the
insulation condition. This is in contrast to an on-line diagnosis where PD detection can only
be performed at the line voltage U0. According to experiences of a utility applying on
regular base an off-line PD diagnosis about 10% of distribution power cables (PILC) as
tested in the last two years up to voltage level of U0 was showing discharges of negligible
levels. As result applying an off-line diagnosis was not necessary at that time.
Therefore, to obtain more efficiency in applying these advanced off-line diagnostics, a
certain pre-selection is preferred. This pre-selection can be done in different ways, e.g.:
a) Using failure history.
b) Identifying risky positions in the network.
c) On basis of an on-line diagnosis.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
51
In particular, to support the off-line diagnosis by prioritising particular cable sections, on-
line PD detection diagnostics can be a more and more interesting option [25,26]. Due to the
fact that
a) distribution power cable-systems are complex systems consisting of different cable
accessories and cable parts;
b) on-line energising may be accompanied by external disturbances;
c) the electrical (only U0) and thermal (actual) load condition may influence the
discharging defects;
before on-line methods will be applied, systematically investigation is still necessary.
In figure 3.19 three examples are show of PD patterns typical for on-line PD detection. It
follows from this figure that especially two cases could be of benefit in prioritising the cable
circuits for off-line in-depth diagnosis:
a) No PD present at U0 to the lower the priority of an off-lien diagnosis;
b) PD present at U0 to consider an off-line diagnosis;
Unfortunately in the other case Disturbances at U0 (20% of all on-line measurements) it is
not possible till now to make any statements about the PD activity.
As already reported (see figure 3.3), the majority of electrical faults occur in the joints of MV
power cables. Hence, for well accessible joints the UHF/VHF PD probe measuring
technology can advantageously be used in order to improve significantly the signal-to-
noise-ratio. A practical example for the PD probing is shown in figure 3.20 [28].
From the point of view of applicability of on-line PD detection to diagnosis of distribution
power experiences at Nuon West have shown that the use of on-line and off-line PD
diagnostics can be complementary and a combination can be of benefit for CBM.
Nevertheless, by using on-line PD diagnostics the following aspects have to be taken into
account [26]:
1. Physical accessibility to cable termination and the possibility for a suitable sensor
location.
2. The influence of disturbances in the network on PD signals originating from the cable
section.
Figure 3.20 Practical example for on-line PD probing of well accessible 20 kV cable joints; a) scanning the
critical region of the cable joint by means of an UHF-sensor; b) visualisation of measured PD events of a
defect joint, c) view of the identified PD degradation in the joint insulation.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
52
3. Interaction
between PD
activity and the
thermal (load)
conditions of the
cable during PD
test. It is know
that changes in
the operation
conditions of a
cable influence
both the PD
inception voltage
and the PD
amplitudes.
4. Distinction
among phase-
own PD sources.
Presence of multiple PD sources in different phases and the cross-talking processes.
5. Importance of PD analysis at different voltage stresses e.g. PDIV, U0, and up to 2U0.
6. The usual time domain reflectometry (TDR) can not be used for the location in the cable
section of PD sites.
The longer the cable sections the higher the minimal detectable PD activity. It followed that
above cable lengths of 0.5 km the cable must be tested at both ends.
3.8 PRACTICAL EXPERENCES; Transmission Power Cables
Approaching the CBM programs for HV cable networks (> 50kV) the following differences to
MV cable systems (< 50kV) have to be considered [17]:
1. there is not one dominant failure process in HV cable networks, PD in the cable insulation
cannot be seen as the only messenger for failures;
2. the complexity of cable systems in HV networks is often higher (e.g. oil-filled, gas-
pressurised cable systems).
Not detection of PD, but the prevention and detection of ageing mechanisms that cause
thermal instability, PD activity, leakages and other degradation of the cable, is the main
target for CBM on HV cable networks. This has resulted in the development of diagnostic
packages to assess the condition of a certain cable type. Figure 3.21 summarises how CBM is
applied to oil-filled cables. After a survey of all relevant information about the cables in the
network, diagnostics are carried out to assess the condition of each cable. Interpretation is
done based on criteria for each diagnostic. All results together are used for a classification of
the cable into five possible categories, see table 3.8. A maintenance program and/or
insulation
outer sheath
accessories
earthing system
hydraulic system
- oil analysis
- tano/PD meas.
- DC test
- visual inspection
- visual inspection
- resistance meas.
- impregnation coefficient
- test of pressure gauges
criteria
criteria
criteria
criteria
criteria
criteria
A
B
C
D
E
cable information diagnostics
interpretation/classification
more measurements
necessary?
Figure 3.21 Condition assessment of oil-filled cables.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
53
guideline for further operation of the cable is linked to each category. Depending on the
category, the condition assessment is repeated once every 3 or 5 years.
Table 3.8: Example of cable classification.
Cable
cat.
Description Availability in
service
Expected time for CBM (h/y) Expected life time
(years)
A new cable in good
condition
>99.9% <24 >30
B good condition but
extra attention is
required
>99% <48 >25
C stable situation, but
moderate condition
>95% <72 >15
D In-stable situation <50% - -
E end of life <95% - <3-5
The criteria for the interpretation of diagnostics are based on knowledge rules, which are
under continuous development using among others:
a) available knowledge (literature) on cable ageing and diagnostics;
b) experience obtained by carrying out diagnostics;
c) research in the laboratory with (accelerated ageing) tests on field aged cable.
d) Table 3.9 shows an example of criteria for interpretation of oil analysis.
Table 3.9: Reference values and criteria for oil analysis.
AC electric
strength
Good
> 50 kV/2,5 mm
Decreased
< 30 kV/2,5mm
tan o o Good Increased
40 C < 7710-4 > 12510
-4
60 C < 16810-4 > 27210
-4
80 C < 34810-4 > 56310
-4
100 C < 45910-4 > 74910
-4
DGA Good Increased
H2 < 658 l/l > 1093 l/l
CO < 50 l/l > 78 l/l
CH4 < 19 l/l > 32 l/l
C2H6 < 10 l/l > 17 l/l
C2H4 < 5 l/l > 8 l/l
C2H2 < 5 l/l > 8 l/l
C3H8 < 12 l/l > 20 l/l
C3H6 < 3 l/l > 5 l/l
iso-C4H10 < 2 l/l > 3 l/l
n-C4H10 < 34 l/l > 63 l/l
CO2 < 416 l/l > 577 l/l
3.9 INSULATION CONDITION KNOWLEDGE RULES
3.9.1 Distribution Power Cables
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
54
After measuring and analysing PD activity in a cable section, the second step is to make a
decision on the insulation condition of the tested cable sample. Using the measured PD
quantities and their interpretation rules (tables 3.4-3.5) three condition classes can be
derived from the analysis:
1. cable section NOT OK: weak spot in the cable section should be replaced;
2. cable section NOT OK?: trending on the cable component is required (e.g. 1 year 3
years);
3. cable section OK: no weak spots in the cable section, cable section is OK.
Figure 3.22 shows the decision diagram for power cables. Analysing the derived
measurement data through the diagram, the cable systems insulation condition can be
determined in one of the three classes. The decision diagram is based on all PD quantities
for measurements on power cables. Not concentrated PD in the cable insulation is not an
indication of ageing of the cable insulation materials, see figure 3.5.
For PILC related insulation, certain levels of PD can be accepted in the cable insulation, joint
and termination, depending on the design of the component. Also, XLPE related accessories
could stand certain PD levels (generally lower pC levels compared to PILC). However,
XLPE cable insulation is required to be PD free (PDIV>1.3U0, PDEV>1.1U0), where the
background noise during the measurement is not allowed to exceed 20pC. The typical PD
values for the different components can be derived from the statistical analysis of all
required measurement data in a database, see figure 3.23-24. As a result for particular
components threshold values can be derived to support the CBM, see table 3.10. In
PDIV/PDEV <
Oper at i on
Vol t age?
PD l ocat ed i n
accessor i es?
Cr i t i cal PD
pat t er n/
i nt ensi t y?
NOT OK
NOT OK ?
(Tr endi ng)
OK
Yes
No
PD l evel >
t ypi cal val ues
f or bot h cabl e
i nsul at i on and
accessor i es?
Concent r at ed
PD Mappi ng?
PD l ocat ed i n
accessor i es?
Cr i t i cal PD
pat t er n/
i nt ensi t y?
START
PD l evel <
t ypi cal val ues
f or bot h cabl e
i nsul at i on and
accessor i es?
Concent r at ed
PD Mappi ng?
No
Yes Yes
No
No
Yes Yes
No
No
No
Yes
No
Yes
No
Yes
Yes
PDIV/PDEV <
Oper at i on
Vol t age?
PD l ocat ed i n
accessor i es?
PD l ocat ed i n
accessor i es?
Cr i t i cal PD
pat t er n/
i nt ensi t y?
NOT OK NOT OK
NOT OK ?
(Tr endi ng)
NOT OK ?
(Tr endi ng)
OK OK
Yes
No
PD l evel >
t ypi cal val ues
f or bot h cabl e
i nsul at i on and
accessor i es?
Concent r at ed
PD Mappi ng?
PD l ocat ed i n
accessor i es?
Cr i t i cal PD
pat t er n/
i nt ensi t y?
START
PD l evel <
t ypi cal val ues
f or bot h cabl e
i nsul at i on and
accessor i es?
Concent r at ed
PD Mappi ng?
No
Yes Yes
No
No
Yes Yes
No
No
No
Yes
No
Yes
No
Yes
Yes
Figure 3.22 Decision diagram for diagnostic data analysis to categories the insulation condition in one of
the three condition classes.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
55
particular the insulation condition assessment of a power cable based on the PD quantities
the evaluation can be performed in two ways:
Table 3.10: Indication of some typical PD levels for different types of cable insulation and
accessories based on experiences from the Netherlands.
Cable element Type Trend values
PILC 10.000pC insulation
XLPE <20 pC
Oil-joint 5.000pC
Type 1(resin insulation 500pC (asymmetric)
(oil insulation) >10.000 pC
(oil/resin insulation) 5.000 pC
splices
Type 2 (resin insulation) 4.000 pC
Oil-termination 6.000 pC
Dry termination 3.500 pC
terminations
Type 3 termination 250 pC
a) comparing actual values to those obtained for the particular cable section in the past and
analysing/trending the differences in the terms of critical changes increases/decreases,
see figure 3.25.
b) comparing actual values to those obtained once for similar cable sections or particular
elements and evaluating the similarities/differences in insulation conditions, see table
3.10.
Defects generating PDs over a long period may result in a failure. It is very important to collect
all data received from PD diagnostic testing. Such data have to be analysed in a statistical way
and per discharging location parameters such as discharge number, intensity, density, pattern
and
phases involved have to be recorded. Typical defects often, but not always, show typical
discharging parameters. Using the systematic application of PD diagnostics can result in
Figure 3.24 Trending two diagnostics quantities
PDIV and PD level; Green area means the
degradation process is not critical for the
insulation condition; reed are means increased risk
of failure.
0
5
10
15
20
25
30
35
N
0
1
2
0
0
2
4
0
0
3
6
0
0
4
8
0
0
6
0
0
0
7
2
0
0
8
6
0
0
0
300
600
900
1200
1500
1800
2100
PD level (pC)
In-Service Time (Days)
Figure 3.23 PD frequency distribution versus service
life calculated from the measurement values in a
certain time interval. As a result aging effects could
become visible e.g. after 1550 days of service
discharging defects become visible.
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
56
0
10
20
30
40
50
60
70
80
10 20 30 40 50 60 70 80 90
T c o n d u c t o r [
o
C ]
T
a
n
o o

[
1
0
-4
]
On- s i te tan d
New c able (syntheti c oi l )
13 years (syntheti c oi l )
25 years ( mineral oi l )
Figure 3.27 Tan o o measurements, the maximum
safe conductor temperature for the cable in service
is 60C.
database and in knowledge rules which are of
considerable value [18], see figure 3.26. But it is extremely important, that network owners co-
operate.
Cable parts and accessories that were replaced on the basis of recommendations or due to
unforeseen failures have to be dissected with the purpose to verify the recommendation and to
understand ageing of the network. Such visual inspections are not required after each
replacement. Especially in the beginning of replacement based on diagnostic testing results,
such feedback is extremely important to obtain optimal knowledge rules and confidence in the
results. These knowledge rules describe the effect of discharge activity on the behaviour of
typical components, like splices, terminations and cable.
The knowledge rule is the key of the transfer of measured pC into the degree of severity of the
measured defect, resulting in recommendations to repair or to exchange components.
3.9.2 Transmission Power Cables
As stated in section 9, knowledge rules are necessary to define criteria. The criteria are used to
Test#1
LocationB
Test#2
LocationB
Test
Cap
Test#1
LocationB
Substation
Test#2
LocationB
644975
108498
078627
191638
122472
055384
055238
04886
080570
125765
6964741
540504
124605
138843
1210186
139202
114843
648505
828507
868510
988507
037505
826503
740503
505501
425500
245504
702252
8kV 4,300pC
12kV 7,700pC
Retestorreplacethe
sectionin1year
8kV 2,100pC
12kV 3,500pC
Inspectthesplice
8kV 2,100pC
12kV 3,300pC
Inspectthesplice.
Retestthesectionin
1year
8kV 2,100pC
12kV 3,300pC
Inspectthesplice
8kV 3,300pC
12kV 6,000pC
Inspectthesplice
8kV noPD
12kV 3,000pC
Retestthesection
in1year
8kV 5,200pC
12kV didnotincrease
Inspectthesplice
8kV 2,500pC
12kV didnotincrease
Retestthesectionin1year
Results areshownonthemapping
diagramfortest#2. All distances
areshownfromLocationAto
LocationB.
Results areshownonthemapping
diagramfortest#1. All distances
areshownfromLocationAto
LocationB.
PDZone
Figure 3.26 Example of insulation condition
assessment of a branched cable circuit. Based on the
database experiences critical parts of the cable
network are indicated.
PD mapping (05-1999)
0
900
1800
2700
3600
4500
0 52 104 156 208 260
cable length [m]
P
D
m
a
g
n
it
u
d
e
[
p
C
]
PD mapping (08-1999)
0
2,000
4,000
6,000
8,000
0 52 104 156 208 260
cable length [m]
P
D

m
a
g
n
i
t
u
d
e

[
p
C
]
PDIV = 10 kV
PD = 4500 pC
PDIV = 8 kV
PD = 8000 pC
Figure 3.25 Example of insulation ageing assessment
in a PILC distribution power cable using PD
diagnosis applied in a time inteval of 3 months. Due
to the decrease in the PD inception voltage and the
increase in the PD magnitude between the
measurements with a time interval of three months,
it was clear that the cable insulation at 90 m was
degrading.
Figure 3.28 Development of PD during accelerated
ageing at 2.5U0 of external gas-pressurized cable
(failure after 240 minutes).
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
57
Figure 3.29 Changing of the |-q-n pattern in the
investigated 275 kV XLPE power cable joint specimen with
k if
Figure 3.30 The -q-n patterns of 275kV XLPE power cable
specimen due to a needle-shaped void
a)
b)
Figure 3.31 The |-q-n patterns of a 66 kV XLPE power
cable specimen with a 6 mm cotton fiber at 70 kV (7
kV/mm); a) after 40 hours and b) after 315 hours voltage
application.
make an interpretation of measurement
results, to define necessary actions, and
to classify cables into a certain
category. Two examples of
development or improvement of
knowledge rules will be given in the
following.
Combination of laboratory research
and field measurements to define the
critical operating temperature of an 150
kV external gas-pressurised cable
system:
1. literature research and laboratory
measurements on new and field
aged cable to characterise the
thermal behaviour of external gas-
pressurised cable;
2. development of diagnostic program
to apply measurements in the field;
3. on-site tan o measurements
together with heating of the cable
conductor and temperature
measurements;
4. interpretation of results to define
guidelines for operation of the cable
(e.g. maximum load capacity), see
figure 3.27.
Accelerated ageing tests in parallel with
tan o and PD measurements on field
aged cable to develop knowledge rules
for critical tan o and PD levels, and to
estimate time to failure:
1. field aged cable is tested in the
laboratory under high voltage (>
2U0);
2. during the accelerated ageing tan o
and PD are measured continuously;
3. after breakdown, the measured data
are used to determine the fault
cause (thermal or electrical?) and to
derive several relations like PD/tan
o level remaining time to
breakdown, PD pattern fault
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
58
Figure 3.32 Tree growth in the 275 kV XLPE
power cable joint specimen with a 10 mm cotton
fiber defect
cause (see figure 3.28).
According to long experience of XLPE power
cable utilization, most probable defects in
XLPE power cables are selected to be knife
scar in the outer semiconducting layer,
needle shaped void in the insulation layer,
and cotton fiber in the insulation layer.
In the following on the base of typical
examples the PD behavior of those
discharging defects in a cable joint insulation
will be presented [29-33].
In figure 3.29 the |-q-n patterns during the first 10 hours are shown. The PD pulse pattern
changed rapidly and became almost steady state after 2-3 hours voltage application with
keeping polarity asymmetry.
Typical |-q-n patterns of a needle-shaped void is shown in figure 3.30. After 2 min of voltage
application q-max reached 2 pC and then about 100 pC just before its breakdown after 81 min.
The PD occurrence of a cotton fiber in the insulation of a XLPE power cable joint specimen is
shown in figure 3.31. Patterns after 40 hours (a) and 315 hours (b) show almost the same
polarity asymmetry and the pulse occurrence characteristics at the voltage phase angle among
90-180 degree.
The presence of cotton fibers may result in treeing activity.
In figure 3.32 an example of a tree growth of about 200 m is shown which was found in the
outer semiconducting layer side of the cotton fiber.
3.10 CONCLUDING REMARKS
The discussion in the previous chapters has shown that
a) Studying fault statistic and providing visual inspection of disturbed components results
in the identification of dominant faults. As most of these dominant faults will be PD
related, PD measurements on power cables are an useful diagnostic tool;
b) Using PD diagnostics, important information can be collected in the PD fingerprint of a
cable section, containing PD inception/extinction voltage, PD levels, PD patterns and
PD locations;
c) Standard interpretation rules which are used in a decision diagram, can be used for
optimal decision-making on the insulation condition of a cable section;
d) Trend-analysis and databases are important tools for condition assessment
Despite the availability of different advanced PD diagnostics several interpretation pitfalls need
to be mentioned also:
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
59
a) The main objective of a diagnostic measurement is to reach conclusions about the
severity and location of possible defects or potential failures, based on the measured
information. The degree of severity always relates to the question: Is this component
going to fail soon and when will that be?
b) So far it was not possible to translate accurately the diagnostic information into
remaining life estimation. Interpretation of the measured insulation condition is
dependent upon the situation i.e. what is valid for one particular cable network situation
is not necessarily applicable for another.
c) The network-owner has to accept a so-called risk of failure, based on the measured
information and collected experience laid down in knowledge rules.
d) These knowledge rules prescribe the typical behaviour of materials or combination of
different materials under a particular service regime. They are the basis of the
recommendation criteria and they should avoid that all decision power about
interpretation of measuring results has to be in the mind of the operator only.
3.11 REFERENCES
[1] E. Dorison, C. Aucourt: After laying test of HV and EHV cables. JiCable Versailles
(1984) paper BV-5.
[2] P. Grnefeld, R. v. Olshausen, F. Selle: Fehlererkennung und Isolationsgefhrdung bei
der Prfung water tree-haltiger VPE-Kabel mit Spannungen unterschiedlicher Form.
Elektrizittswirtschaft 84 (1985) 13, 501-506.
[3] E. Lemke, R. Rding, W. Weienberg: On-site testing of extruded power cables by PD
measurements at SI voltages. CIGRE Symp. Vienna (1987) paper 1020-02.
[4] F. Farneti, F. Ombello, E. Bertani, W. Mosca: Generation of oscillating waves for after-
laying test of HV extruded cable links. CIGRE Session Paris (1990) paper 21-10.
[5] G. Friese, E. Lemke: PD phenomena in polyethylene under AC and impulse stresses
with respect to on-site diagnosis tests of extruded cables. 7
th
ISH Dresden (1991) paper
75.07.
[6] R. Bach, W. Kalkner: Comparative study on alternative test voltages for layed
medium voltage cables. 7
th
ISH Dresden (1991).
[7] R. Plath: Oscillating voltages als Prfspannung zur Vor-Ort Prfung und TE-Messung
kunstoffisolierter Kabel. Doctoral Thesis, TU Berlin (1994).
[8] E. Lemke, P. Schmiegel: Complex Discharge Analyzing (CDA) - an alternative
procedure for diagnosis tests of HV power apparatus of extremely high capacity. 9
th
ISH Graz (1995) paper 5617.
[9] E. Lemke, P. Schmiegel, H. Elze, D. Ruwurm: Procedure for evaluation of dielectric
properties based on Complex Discharge Analyzing (CDA). IEEE Intern. Symp. on El.
Ins. Montreal (1996) 385-388.
[10] E. Pultrum, E. Hetzel: VLF discharge detection as a diagnostic tool for MV cables.
IEEE PES Summer Meeting Berlin (1997).
[11] J.T. Holboll, H. Edin: PD detection vs. loss measurements at high voltages with
variable frequencies. 10
th
ISH Montreal (1997).
[12] Lefvre, W. Legros, W. Salvador: Dielectric test with oscillating discharge on synthetic
insulation cables. CIRED Paris (1989) pp.270-273.
[13] V. Colloca, A. Fara, M. de Nigris, G. Rizzi, Comparison among different diagnostic
systems for medium voltage cable lines, paper 1-53, Cired 2001, Amsterdam
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
60
[14] C. Walton, Regulatory and business drivers for reducing underground cable failures,
paper 3.6 at ERA HV plant life extension conference, Brussels, 23-24 November 2000.
[15] N. van Schaik, E.F. Steennis, A. van Dam, B.J. Groetenhuis, M.J. van Riet, C.J.
Verhoeven, CBM on MV cable circuits as a part of asset management, CIRED 2001.
paper No. 1.28
[16] E.R.S. Groot, Ph. Wester, E. Gulski, F.J. Wester, C.G.N. de Jong, M. van Riet, D.M.
Harmsmen, J. Pellis, H. Geene, Tools, for quality assessment of distribution cable
networks, CIRED 2001. paper No. 3.17
[17] J. Popma, J. Pellis, Diagnostics For HV cable systems, HV Plant Life Extension, ERA
Technology Conference 2000, paper 2.6
[18] F.J. Wester, E. Gulski, J.J. Smit, CBM of MV power cable systems on the base of
advanced PD diagnosis, CIRED 2001. paper No. 1.29
[19] E.F. Steennis, E. Hetzel and C.W.J. Verhoeven, Diagnostic medium voltage cable test
at 0.1 Hz, Jicable 1991: 408-414.
[20] E. Gulski, J.J. Smit, P.N. Seitz, J.C. Smit, PD Measurements On-site using Oscillating
Wave Test System, 1998 IEEE International Symposium on EI, Washington DC, USA,
[21] E. Gulski, F.J. Wester, J.J. Smit, P.N. Seitz, M. Turner, Advanced PD diagnosis of MV
power cables using OWTS, IEEE EI Magazine March/April 2000, Vol. 16 No.2
[22] N. van Schaik, E.F. Steennis, W. Boone, E. Hetzel and D.M. van Aartrijk, MV cable
diagnostics, condition based maintenance on power cables, 17th Nordic Symposium
on EI, Stockholm, Sweden, 2001.
[23] S. Hvidsten, H. Faremo, J.T. Benjaminsen, E. Ildstad, Condition assessment of water
treed service aged XLPE cables by dielectric response measurements, Cigre Session
2000, paper 21-201
[24] N. van Schaik, E.F. Steennis, W. Boone and D.M. van Aartrijk, PD diagnostics on
very long and branched cable networks, 17th Nordic Symposium on EI NORD-IS,
Stockholm, Sweden, 2001.
[25] J.J. Smit Insulation condition estimation of distribution power cables by on-line and
off-line diagnostics, 2001 International Symposium on EI Materials, Himeij, Japan,
paper K-1
[26] E. Lemke Advanced diagnostic tool for on-site PD measurements of MV and HV
power cables, 1
a
Jornada Tcnica Sobre Coordinacin de Aislamiento, Madrid (1999)]
[27] E. Lemke, T. Strehl, M. Boltze: Advanced diagnostic tool for PD fault location in
power cables using the CDA-technology. 12
th
ISH Banaglore (2001) 6-46
[28] E. Lemke: Partial discharge detection (PDD) for cable accessories. ETG-PES Summer
Meeting, IEEE-Panel; Berlin (1997)
[29] A. Toya, T. Goto, T. Endoh, H. Suzuki, K. Takahashi, "Development of partial
discharge automatic-monitoring system for extra-high voltage XLPE cable lines",
JICABLE'95 No.D3.2, PP.655~660 (1995)
[30] G. Katsuta, H. Suzuki, H. Eshima, T. Endoh, "Discrimination of partial discharge from
noise in XLPE cable lines using a neural network", IEEE ANNPS'93 No.93
PP.193~198(1993)
[31] H. Suzuki, T. Endoh, "Pattern recognition of partial discharge in XLPE cable using a
neural network", Proc. of 3rd ICPADM, PP.43~46(1991)
Part III CIGRE TF D1.11/33.03.02
POWER CABLES
61
[32] H. Suzuki, T. Endoh, Y.Sekii, "Experimental study on the causes of electrical
breakdown of EHV XLPE insulation using pre-breakdown partial discharge detection
method", CEIDP. PP.255~260(1990)
[33] H. Eshima, H. Suzuki, T. Endoh, M. Fukasawa, "Partial-Discharge Automatic-
Monitoring System for Extra-High Voltage XLPE Cable Lines", Hitachi Cable Review
No.16 PP.9~18 (1997)