21, rue d’Artois, F-75008 PARIS B1_304_2010 CIGRE 2010

http : //www.cigre.org

ON-SITE COMMISSIONING TEST AND DIAGNOSTICS OF

220 kV XLPE CABLE SYSTEM

M.M.AWAD F.TAHOUN A.ELFARASKOURY* O.E.GOUDA

Egyptian Electricity Holding Company Faculty of Engineering
Cairo University
Egypt

SUMMARY

Most observed defects in today's HV and EHV cable systems cause partial discharges (PD) under
AC voltage stress. Partial discharge diagnostic testing of cable systems in the field is intended to
measure and locate those singular or multiple clusters of PD producing defect sites, in a non-
destructive way , that would result in a near term cable failure. Cable systems contain the cable
itself and connected accessories such as joints and terminations have to be tested to check the
performance of the transmission. The results of PD diagnostic tests are used to assess the
condition of the cable insulation. High voltage testing is that the applied test voltage simulates the
stresses which occur during the operation of HV apparatus. High voltage should provide the
information for the decision whether a defect in the insulation is dangerous. That means the
failure mechanism during the HV test and the later operation should follow the same physical
process. To accelerate this process, the test voltage is usually higher than the corresponding stress
during operation.On the other hand, experience also shows that the On-site PD measurements by
using frequency tuned resonant test systems gives high sensitivity and calibration is possible
using partial discharge calibrator on the cable terminations. This paper contains tests carried out
on-site for old and new 220 kV cables having different lengths to determine the faulty joints and
cable defects. This paper is an over view covering best practices for cable commissioning after
installations and applying predictive diagnostic programs to aging cable systems. It contains also
the results of many tests carried out on cable system. Withstand voltage test and the partial
discharge tests are carried out according to IEC 62067 and the obtained results are discussed.

KEYWORDS

High Voltage Cables- Variable Frequency- Partial Discharges- Cross Bonding Links- High
Frequency Current Transformers- Commissioning Test - On Site

1. INTRODUCTION

The Egyptian Electricity Holding Company (EEHC) Obtained a variable frequency test system
(20 Hz to 300 Hz) with 4.28 µF maximum test capacitance which corresponds to cable length of

1
16.4 km at a cable capacitance of o.26 µF/km for after installation tests carried out on 220 kV,
XLPE cables .The variable frequency test system is used to improve cables performance by using
up to date technology for On-site withstand tests of power cable systems. The tests are mainly
performed to check the quality of the accessories and their assembling. Damages of the cable
during lying can be detected too by using the frequency tuned resonant test systems. The
withstand test may be combined with PD measurements check the performance of cable systems
according to IEC. A general principle of the insulation co-ordination and HV testing is that the
applied test voltage simulates the stresses which occur during the operation of the HV apparatus.
High voltage tests should provide the information for decision whether a defect in the insulation
is dangerous or not for the later operation. That means the failure mechanism (caused by the kind
of defect and the kind of the voltage stress) during the HV test and the later operation should
follow the same physical process. To accelerate this process, the test voltage is usually higher
than the corresponding stress during operation. The most important stress of a XLPE cable in
service is the stress with the operational alternating voltage. If an on-site test is completed with a
partial discharge (PD) measurement, all the experience from the various tests can be transferred
to the various factory tests can be transferred to the on-site test [1].

The on-site testing of cables has to check the insulation condition after-laying and assembly of
cable system, as well as ageing of cables and accessories, since the performance of the cables and
accessories was tested during the type and routine tests in the factory. The after laying test of new
cables fills the quality assurance gap between the type and routine tests of the cable at the
manufactures site and the commissioning of the complete cable system on-site. During the
assembly or repair of a cable system, defects of the cable sheath and misassembled of joints and
terminations can occur. Frequency tuned resonant test system are meanwhile, state of the art for
on-site testing and diagnostics on high voltage (HV) and extra high voltage(EHV) XLPE
insulated cables. The components of the HV test system and load (power cable to be tested) from
an oscillating circuit of a certain natural frequency, this system operates in resonance. It
establishes a powerful high voltage at minimum power input. The control of test system searches
for the resonant frequency automatically and the HV test is carried out at this frequency. The test
system can also be designed in advance for use as two in-dependently running systems with
single reactor which also can be operated in series or parallel as shown in Figures (1, 2).Such this
system is available in Extra High Voltage Research Center (EHVRC) Laboratory.

Figure 1 Frequency tuned resonant test Figure 2 Frequency tuned resonant test
System (single reactor) system (parallel reactor)

2
2. TEST ARRANGEMENT

2.1 AC Cable Systems Testing after Installation

After installation testing of EHV cable system is actually carried out by alternating voltage of
variable frequency tests according to IEC 62067. These standards offer alternative for AC test
procedure, besides testing with sinusoidal waveform and the frequency between 20 Hz and 300
Hz. The voltage applied for 1 h, either with a voltage according √2 U0 or with 1.7 U0 depending
on practical operational conditions, it is also possible to test with U0 for 24 h [2]. While the
testing with operating voltage U0 for 24 h is not feasible for incidence of failure after the test
could be occur. High voltage tests should provide the information for the decision whether a
defect in the insulation is dangerous or not for the later operation. That means the failure
mechanism during the HV test and the later operation should follow the same physical process.
To accelerate this process, the test voltage is usually higher than the corresponding stress during
operation.

2.2 PD Measurements on long EHV cable systems

Partial discharges are a sensitive measure of local electrical stress and therefore the measurement
is very often used a quality check of the insulation [3]. Therefore the insulating material can be
tested with high stress but without damaging or reducing the performance of the insulation.

2.3. Measurement system

Measuring of partial discharges is carried out using the MPD 540 measuring instruments with
frequency tuned resonant test system as a source at the terminations and the clamp-on type high
frequency current transformer sensors (HVCT) at cross bonding box of 220 kV for XLPE cable
systems showed high sensitivity and calibration is possible using PD calibrator on the cable
terminations, and another type of measuring by use of a coupling capacitor according IEC 60270
is physically limited to a maximum detectable cable length of proximally 2 km becomes too low
sensitivity, depending on cable parameters and PD background noise.

2.3.1 PD Detection at cross-bonding links of 220 kV XLPE cables

Partial discharges measuring at Cross-Bonding (CB) links by using inductive sensors are
especially designed as an inductive sensor, such installation is even possible under on-line
conditions, as the sensor is a clamp-on HFCT that can be opened and clamped around a cross-
bonding link cable. Figure (3) shows the installed HFCT inside a cross-bonding links, also used
PD gating unit for gating purposes in order to provide the possibility to filter certain external
background noise. The PD sensitivity using HFCT the central measuring frequency is
recommended to lay between 2 MHz and 10 MHz in a flat zone of the frequency spectrum. The
spectrum is obtained from FFT of calibration PD pulses. Furthermore, the measuring frequency
must be set in order to obtain the greatest possible PD signal/noise ratio. In addition, an ''on-site
performance check" must be carried for the selected measuring frequency before the PD
measurement starts [4].

Detecting PD on long cables is a good start to the diagnosis of the system condition. However, it is
only a start, and the obvious next stage after detection is to locate where the PD originated from. Once
located, the final stage is to understand the problem which has generated the PD activity, and to make
some sort of remediation plan which may include replacement of the offending cable or plant item.
Long HV and EHV cable systems are usually designed as cross-bonding systems to minimize
screen losses and limit voltage rise. Using cross-bonding links for on-site PD measurements is

3
suited for direct buried systems, has no impact on the cable systems, needs no PD sensors
integrated in the cable joints, offers a low cost solution and so opens a wider range of
applications, in cable testing after installation as well as on service-aged or repaired cable systems
[5], [6] and [7].

The measurements carried out between two substations, one substation was disconnected from
the cable systems under test and the second system connected with frequency tuned resonant test
system with corona shield were mounted on the top of the outdoor terminations to avoid corona
discharges on the bushing bolts as shows in figures (1, 2). Each phase of three-phase system of 10
km cable installation included 14 joints, three outdoor terminations and three GIS sealing ends.
Inject a calibration pulse with known quantity of charge, at the outdoor system termination (i.e.,
between HV and LV terminals), the sensitivity of the PD detector has to be modified until the
detector shows the calibration charge.

Figure 3: CB Link with mounted three

HFCT (PD Sensor)

The PD Guideline Level given below for the on-line condition assessment of polymeric-insulated
XLPE cables, cable accessories and cable sealing ends of operating voltages from 66 kV up to
400 kV [8]:

From 0 to 250 pC : Discharge within tolerable limits

From 250 to 1000 pC : Some concern, regular monitoring recommended
> 1000 pC : Major concern, repair or replace

2.3.2 PD Fault Locations

The localized field stress at site of defect must be sufficient to exceed partial discharge inception
voltage (PDIV). Once this inception voltage has been met, repetitive partial discharge activity
will normally continue until the applied voltage is decreased to a level known as partial discharge
extinction voltage (PDEV). Therefore defects that have PDIV close to the operation voltage of
the cable system, are likely to be initiated by an over voltage transient. When a partial discharge
occurs in a cable, a traveling wave is created at the source of the PD site by the electromagnetic
pulse. One half of the traveling wave sets off towards the one end of the cable "near end" and the
other end "far end". The PD measuring software includes an internal PD location for long power
cables; the measuring principle is based on Tine Domain Reflectrometry (TDR).

4
3. TEST RESULTS AND EXPERIENCE

The tests have been carried out on-site according to IEC for 220 kV cables at 1.7 U0 having
different lengths to determine the faulty joints and cable defects as shown in Figure (4, 5).
Therefore, the XLPE cable insulation was subjected to AC tests after assembling and at the same
time partial discharge measurements were done on all accessories simultaneously for three-phase
PD measurement on the relevant joint box would be possible, also the cross-bonding can be
changed to straight-through connection, to minimize cross-talk between the three phases and to
clearly distinguish between the three joint of one group.

Figure: 4 Breakdown during test Figure 5: 220 kV cable termination failures

The results of the on-site withstand testing of 220kV XLPE cables by the use frequency tuned
resonant test system (20- 300 Hz) is in use since early 2007. To date more than 10 circuits
and about 72 km long, the withstand voltage of 1.7 U0 with duration of 1 hour reported six
faults during test , three joints , one termination and one cable insulation failed. Sources of
partial discharges in extruded cables include breakdowns in voids, cavities, along an interface
between an energized electrode and floating conductor, in an electrical tree etc. the pictures
given in Figure (6) shows a pending cable failure that occurred when PD activity between the
semi conducting and cone of cable termination resulted manufacturing process, the
breakdown occurred at 8 sec. from starting the test as given in Figure(4) and Figure (7) shows
failure that occurred in joint during commissioning test, resulted defective workmanship.

Figure 6: pending for 220 kV cable failures Figure 7: Joint failure

5
It is known, that HV power cables failure can occur as a result of the normally applied operational
voltage or during a transient voltage lightning or switching surges. The failure can occur if
localized electrical stresses are greater than the dielectric materials in the area of localized stress
or the bulk dielectric material degrades to the point where it cannot withstand the applied voltage.
Therefore performing non-destructive diagnostics on-site could be an important issue to
determine the actual condition of the cable system and to determine the future performances [9].
The results of the on-site PD measurements have been performed in conjunction with a
commissioning test at cross-bonding links using high frequency current transformer sensors
reported the discharge activity ranged from a 7 pC to 235 pC. The variation of noise level which
is experienced during all measurements resulting in higher external interference from ends of the
cables. Also corona effect caused by floating parts close to high voltage at the cables termination
as shown in Figure (8), the cable termination with and without corona shield. In this figure,
typical examples of PD pattern are shown.

a) PD measuring from CB link and cable termination without corona shield

b) PD measuring from CB link and cable termination with corona shield

Figure 8: PD measuring from CB link and cable termination

The results of on-site partial discharge measurements in PC for the three phases R, S and T of
high voltage power cables, cable accessories and cable sealing ends show similar behavior on
circuit I and circuit II as shown in Figure (9) for all joints at cross bond link box, solid link box
and straight through joint with screen earthed. In this figure, typical examples of PD level are
shown. Due to test procedure for joint group at different times and PD behavior had changed, the
PD pattern and damping show clear evidence for corona interference, which was introduced in
the cable systems by capacitive coupling.

6
 a) PD measurements for circuit I b) PD measurements for circuit II

Figure 9: Examples of on-site partial discharges results for 220 kV cable systems

4. CONCLUSIONS

The experiences from commissioning test of 220 kV – XLPE cables with frequency tuned
resonant test system have been gained in on-site testing, with all routine tests on cables before
laying, also using the HFCT sensor inside cross- bonding links given good results, and on-site
withstand voltage testing combined with sensitive partial discharge measurements reduces the
risk from the service, also after repair the joint reassembling was done exactly in the same place.
The experiences also show that the test voltage with U0 for 24 h is not feasible for incidence of
failure after the test could be occurred.

5. REFERENCE

[1] S.Schierig, D.Russwurm: HV on-site testing on cable by alternating voltage of variable

frequency, IEEE Insulated Conductors Committee (ICC)-2000
[2] IEC 62067 “Power cables with extruded insulation and their accessories for rated voltages
above 150 kV (Um = 170 kV) up to 500 kV (Um =550 kV) - Test methods and requirements”,
2006-1.
[3] IEC 60270 “High voltage test techniques- partial discharge measurements” – 2000-12.
[4] F.Garnacho, I.Trasmonte et al., “On-site measurements experiences in insulation condition for
medium and high voltage cables”, CIGRE, D1-201, 2008.
[5] W.Weissenberg, F.Farid et al., “On-site PD detection at cross-bonding links of HV cables”,
CIGRE, B1-106, 2004.
[6] C.Min, K. Urano et al., “Investigations of partial discharge measuring method via coaxial
bonding wire for direct-buried power cable”. T.IEE Japan, 121-B, No. 1,2001.
[7] C .Min, K. Urano et al., “Verification of partial discharge measuring method via coaxial
bonding wire for direct-buried power cable”. T.IEE Japan, 122 B, No.4,2002.
[8] HVPD Technical Guide for PD Levels in MV and HV Cables and Joints- Manchester, UK-
May 2009.
[9] E.Gulski, F.de Vries, et al., “Condition Assessment of Service HV Power Cables”, CIGRE,
D1-206, 2008.