Close and Return

C.4.1

8th International Conference on Insulated Power Cables

C.4.1

ON-SITE TESTING WITH COMPACT AC TEST-SYSTEM AT THE FIRST 500 KV XLPE

CABLE PROJECT IN SOUTH AMERICA
Andreas WEINLEIN, Gero SCHRDER, Sebastian EBERT, Sdkabel GmbH, Mannheim (Germany),
andreas.weinlein@suedkabel.com, gero.schroeder@suedkabel.com, sebastian.ebert@suedkabel.com,
Harald GEYER, agea kull ag, Derendingen (Switzerland), h.geyer@agea-kull.ch

ABSTRACT
The first 500 kV XLPE cable project in South America was
delivered, installed and commissioned in Colombia as
turnkey project. Due to the given local logistical and
technical requirements a new compact on-site AC high
voltage resonant test system was designed.
The test-set consists of 4 pieces cylinder reactors all
placed transport locked in a 20 ft container suitable for a
fast and reliable erection of the test system on-site. The
compact test-set is rated for 540 kV.
The on-site test was carried out successfully in
September 2010 including both, high voltage and PD
tests.

KEYWORDS
500 kV XLPE cable system, compact on-site AC HV
resonant test system, PD measurement on-site, compact
plug-in sealing end, mobile spare phase

A full scale commissioning test of the test-set was carried

out in the cable manufacturers factory under full load
conditions with 530 kV for the duration of 2 hours at a
cable drum of a capacitance of 170 nF. Additionally, a PD
measurement at both plug-in test cable sealing ends of
test set-up was carried out. The test frequency for this setup was 20 Hz, representing the minimum allowable
frequency for AC voltage tests after installation acc. to
IEC 62067 [11].
The on-site after installation test with 493 kV / 1 h was
carried out successfully at all cables including the spare
phase in September 2010 comprising both, AC HV tests
and PD measurements.
The testing of the spare phase was carried out by phaseplugging of the regular cable phase inside the GIB cable
enclosure which was performed in less than one day. This
was possible because of both, the modular set-up of the
test-set and the advantages of the compact plug-in
sealing end system.

INTRODUCTION

PROJECT DETAILS / REQUIREMENTS

The first 500 kV XLPE cable project in South America was

delivered, installed and commissioned in Colombia in
September 2010.

A turnkey project with 5.4 km 500 kV XLPE cable of type

A2XS(FL)2Y 1x800 RM/150 290/500 kV was planned,
manufactured, delivered, installed and commissioned for
the hydropower project Porce III near Medelln / Colombia
(Empresas Pblicas de Medelln E.S.P.).

The entire job was handled as turnkey project. Due to the

given local logistical and technical requirements a
compact on-site AC HV resonant test system was
designed to carry out the AC HV commissioning tests.
The XLPE cable connection consists of two systems
(additionally one spare phase) rated 500 kV with
individual cable length between 750 855 m. The
conductor cross section is 800 mm2 aluminium. Both
circuits are laid in flat formation in a vertical saddle
arrangement inside the cavern of the Porce III Hydro
Power plant. Total 7 pieces of factory pre-tested compact
plug-in SF6 sealing ends and 7 pieces of factory pretested and pre-assembled compact plug-in outdoor
sealing ends were installed.
The test-set consists of 4 pieces cylinder resonant
reactors (3 t each) all locked for transport in a 20 ft
container with removable roof suitable for a fast and
reliable erection of the test-system on-site. The step-up
transformer and frequency converter including control unit
for voltage and PD measurement are placed in a 10 ft
container.
The entire transport weight of the test-system is less than
24 t. Extension up to 6 resonant reactors is considered
and easy to implement. Because of the module-like set-up
the handling procedure on-site is optimized.
The test-set is rated for 540 kV, 11.6 A, max.170 nF cable
capacitance (set-up resonant reactors: 2 serial, 2 parallel)
or for 280 kV, 23.2 A, max. 680 nF cable capacitance
(set-up resonant reactors: 4 parallel).

The road distance to the next Atlantic sea port is more

than 600 km with very frequent unpaved road quality in
very high mountain regions.

Fig. 1: map of region Porce III in Colombia

The project consists of 2 systems with an approximate
phase length of 760 m. Additionally one spare phase of
855 m length was supplied and installed. In total 7 pieces
of factory pre-tested 500 kV compact plug-in SF6 sealing
ends and 7 pieces of factory pre-tested and preassembled 500 kV compact plug-in outdoor sealing ends
were installed:

Jicable11 19 23 June 2011, Versailles - France

Close and Return

C.4.1

8th International Conference on Insulated Power Cables

C.4.1

Although each single cable and accessory is subject to

routine tests at the manufacturers lab, incidents during
transport, cable laying and installation can lead to
unnoticed defects. In consequence, after-installation tests
of the insulation focus on defects in cable accessories,
e.g. interfacial problems, improper positioning, cuts or
scratches, contaminations etc. Such defects do not
necessarily lead to a breakdown during the HV test,
bearing the risk of breakdown later in service. Sensitive
on-site PD measurement significantly reduces this risk [4,
6, 8].
Fig. 2 (left): 500 kV compact plug-in SF6 sealing end
Fig. 3 (right): 500 kV compact plug-in OSE

design AC test voltage to ground:

design impulse voltage:
design switching impulse voltage:

580 kV
1550 kV
1175 kV

conductor

aluminium, round stranded

conductor
screen

conductive XLPE-compound

insulation

XLPE

insulation
screen

conductive XLPE-compound

bedding

swelling tape, semi-conducting

wire
screen

copper wires

bedding

swelling tape, semi-conducting

bedding

fabric tape, semi-conducting

9
10

metallic
sheath
outer
sheath

copolymer-laminated aluminium
HDPE, outer conductive and
flame-retardant layer

Fig. 4: cable design

The two circuits connecting the main transformers with the
overhead lines are fixed in the cavern by means of a
special short-circuit approved cable saddle system where
the cables are laid in flat formation in vertical arrangement
at 400 mm distance.

Fig. 5 (left): 500 kV compact outdoor sealing ends

Fig. 6 (right): 500 kV cables fixed with saddle system
The concept of using the spare phase leads to a
heightened final position of the outdoor sealing end
(middle phase) and a mobile fixation of the concerning
compact plug-in SF6 sealing end which can be connected
with any main transformer inside the transformer cavern
via the associated GIB enclosure. This concept allows to
replace a phase with the spare phase within a few hours.
During HV commissioning test this plug-in procedure has
to be demonstrated.
According to customers specification an AC HV test of
the insulation after installation acc. to IEC 62067 with a
voltage of 1.7U0 = 493 kV for the duration of 1 h had to
be carried out at each phase including spare phase/s. For
after-installation tests, PD measurements are actually not
required by IEC standards [11].

At each installed EHV accessory (Um 362 kV) a PD

measurement is carried out during AC voltage test of the
insulation as activity of quality assurance by Sdkabel
GmbH [6, 7, 10].
Because of the required high test voltage and the
secluded region in the high mountains of South America it
was nearly unfeasible to organise AC HV test without final
confirmed completion date. As there is no appropriate
test-set available in South America and in order to be
independent from test institutes it was decided to specify
an own compact AC HV test system with the main focus
on testing cable systems used as power plant links to
overhead lines, switchyard or GIS in regions difficult to
reach. The typical cable lengths of the power plant leads
in EHV systems are 200 - 1500 m, the conductor cross
2
sections are typically 1200 mm . For lower test voltages
even longer cables can be tested.
Acc. to IEC 60840 for cables with Um 170 kV and IEC
62067 for cables with Um > 170 kV requirements for the
shape of suitable AC test voltages and for the time of its
application are defined:
substantially sinusoidal waveform
frequency between 20 and 300 Hz
time of voltage application equal to 1 hour
Both standards define the AC test voltage level for on-site
tests of newly installed cable systems which depends on
the cable rated voltage with 1.1 - 2.0U0 [11, 12, 13]. Test
voltages up to 1.2U0 can be reached with insulated
operation and voltage supply by a generator which can
increase the voltage step by step from approx. 20% up to
120% for short time duration (increasing of excitation
voltage). Acc. to both standards alternatively, a voltage of
U0 may be applied for 24 h. For EHV systems this method
should be accompanied by a PD measurement [4].
To cover all international and local requirements a test
system for a maximum capacitance of 170 nF at AC test
voltage 540 kV was specified. Because of limited
availability of cranes on-site and limited access routes the
weight and design of one single resonant reactor was
specified to be as minor as possible also considering the
minimal allowed frequency of 20 Hz. To carry out a PD
measurement at each accessory of a cable system on-site
this compact mobile test system has to be free of internal
PD and free of corona up to the maximum test voltage of
540 kV (PD 5 pC).

Jicable11 19 23 June 2011, Versailles - France

Close and Return

8th International Conference on Insulated Power Cables

The result of this design is a compact reactor with low

weight compared to its testing power. Each resonant
reactor is designed for 280 kV and 5.8 A (inductivity:
360 H). The entire height of the 540 kV set-up is approx.
5.1 m. The HV divider is of three parts type.
8

HS- V$rbind$r 950

F

E
5195

3 x 400 V
max. 150 kVA
0 540 kV
max. 11.6 A
6265 kVA
20 144 Hz
3.3 nF 170 nF
2 h ON / 10 h OFF
> 100
approx. 3000 kg
4 pieces
approx. 1400 kg
approx. 330 kg
< 95%
approx. 18 t + 6 t
<4h

300

power supply (50/60 Hz):

power:
output voltage:
output current:
nominal power:
frequency range:
capacitance range:
duty cycle (at 20C / 100%):
quality factor:
weight reactor:
nos. reactors:
weight exciter transformer:
weight frequency converter:
humidity (no condensation):
system weight incl. containers:
erection time of test-set on-site:

905

1895

A system was designed with following characteristics:

The design of the reactor is based on the experience with

smaller units which were developed for on-site testing of
GIS installations since 1979 [01]. The reactors consist of
two high voltage windings in series. Both windings
enclose a common bar core at half potential. In difference
to most other reactors, the DSH6W has no closed iron
core with divided air gaps. The magnetic field generated
in the iron core closes back via the air. This has to be
taken into consideration during erection of the test circuit.

5069

DESIGN OF TEST-SET

C.4.1

2269

C.4.1

1140

440
7500
C

2390

3130

90
Anschluss A

2000

1980

Anschluss A

Prf syst $m - Hauptachs$

14
00

St ck G$g$nstand
II I nd$rung$n:

W$rkstoff

Pos. Mod$ll B$m$rkung$n

Ers. durch:
Ersat z f r:
Ausgab$dat .

A- 1136 Mass- St . G$z.

Aufbau: 2 R$akt or$n S$ri$

2 R$akt or$n Parall$l

1:30

El$ktrisch$ Apparat$
CH - 4552 D$r$nding$n

02.09.2009 Ko

G$pr.
Allg$m$intol$ranz$n
nach DIN ISO 2768 mk

3 HT 16 504
1

Fig. 9: 540 kV set-up of 4 reactors

Fig. 7: 540 kV layout of test system

High test power can only be efficiently generated by
mobile resonant test systems, where the weight to-power
ratio and feeding power demand is relatively low and the
transport volume is manageable [2]. When operated at
resonance, the feeding power is reduced to the real power
loss in the test circuit. Only the actual power loss has to
be supplied in order to maintain the test voltage. The
quality factor Q is the ratio of reactive power and real
power. The losses in the resonant reactors are by far
higher than inside the XLPE cable system. Frequency
tuned resonant circuits (ACRF) consist of constant
inductance(s), the capacitive load and a control module
with frequency converter unit [5].

The concept of the design considers a 20 ft transport

container with a removable roof suitable for sea transports
and fast and reliable erection of the test-system on-site.
All resonant reactors are fixed inside the container with
special holding devices locked for transport. All toroidal
corona rings are dismantled and stored volume optimized.

Fig. 10: 20 ft transport container with removable roof

All corona rings are made of conductive hard plastic. This
technique avoids perturbing buckles as noticed very often
on metallic hollow surfaces. To prevent from corona at
test voltages of 300-540 kV the diameter of the plastic
pipes is 300 mm with an outer ring diameter of 1.5 m.
Between resonant reactor(s) and HV divider a HV filter is
integrated to protect the set-up in case of breakdown of
the test object and to improve the quality of the PD
measurement at the test object.
Fig. 8: design of resonant reactor size DSH6W

The step-up transformer and the frequency converter

including the control unit for voltage and PD measurement
are placed in a 10 ft container.

Jicable11 19 23 June 2011, Versailles - France

Close and Return

C.4.1

8th International Conference on Insulated Power Cables

C.4.1

duty cycle (at 20C / 100%):

2 h ON / 10 h OFF
quality factor:
> 100
nos. of reactors:
4 pieces
erection time of test-set on-site: < 3 h
The entire height of the 280 kV set-up is approx. 3.8 m.
The HV divider consists of two parts. With the upgrade by
two further reactors cables with a maximum capacitance
of 1020 nF can be tested up to 280 kV.
8

300

0
$r 95
rbind
HS- V$

890

Fig. 11: conductive hard plastic corona rings

270

2574

3770

D
160

160

160

Anschluss A

Anschluss A

Anschluss A

Anschluss A

1150
Montag$wink$l b$r 10 Bohrung mont i$rt
ansonst$n mit t ig b$r Langloch

1980

90

C
2000

14
00

Further accessories of the test-set are a full size safety

circuit, a 50 m long connection cable of 200 kVA (5 x
400 V), a 12 m robust telescope HV connection (diameter
250 mm), a 20 m long medium voltage cable (1 x 25 kV)
and for each resonant reactor an oil-sump including post
insulators for isolated installation.

Prfsyst $m - Haupt achs$

2120

2120

2120

3150
B

9510

St ck G$g$nstand
II I nd$rung$n:

W$rkstoff

Pos. Mod$ll B$m$rkung$n

Ers. durch:
Ersat z f r:
Ausgab$dat .

A- 1136 Mass- St . G$z.

1:30 G$pr.

Aufbau: 4 R$akt or$n Parall$l

03.09.2009 Ko

Allg$m$intol$ranz$n
nach DIN ISO 2768 mk

El$ktrisch$

Apparat$

CH - 4552 D$r$nding$n

3 HT 16 503
1

Fig. 14: 280 kV set-up of 4 reactors

REAL HEATING TEST

The factory acceptance test of the AC HV test-set without
load capacitance at manufacturers test bay was carried
out for each single resonant reactor with 336 kV / 1 min.
and 280 kV / 30 min. at fTest = 112 Hz (including PD
measurement). Additionally the 540 kV set-up and the
280 kV set-up with 4 reactors each was tested at 540 kV /
1 h (at fTest = 130 Hz) resp. 280 kV / 1 h (at fTest = 125 Hz)
without load.
Fig. 12: 10 ft control container
The entire test-set is upgradeable with two further
resonant reactors type DSH6W to be stored in a second
10 ft container. With this upgrade cables with a maximum
capacitance of 255 nF can be tested up to 540 kV.
Alternatively the entire test-set can be built up for 280 kV
use. Test voltage 280 kV covers most of the tests (approx.
90%). For this application all 4 pieces of resonant reactors
have to be connected in parallel:

Fig. 15: set-up 4 x 280 kV / 1h

Fig.13: 280 kV layout of test system

output voltage:
output current:
frequency range:
capacitance range:

0 280 kV
max. 23.2 A
20 206 Hz
5 nF 680 nF

As final acceptance test the entire set-up was tested

under real heating conditions connected with a HV XLPE
cable. A cable length of approx. 1000 m was prepared
with 2 pieces of 500 kV SF6-plug-in compact sealing ends.
One end was connected via a gas filled bushing with the
test-set. The other side of the cable was closed with a gas
filled dead-end pipe. The entire capacitance of cable
including sealing ends, bushing and HV divider was
approx. 170 nF.

Jicable11 19 23 June 2011, Versailles - France

Close and Return

C.4.1

8th International Conference on Insulated Power Cables

C.4.1

test voltage:
530 kV
test duration:
2h
test capacitance:
170 nF
test frequency:
20.2 Hz
test current:
11.5 A
At both cable sealing ends a synchronous PD
measurement was carried out with inductive sensors.
Fig. 18 (left): sensor and PD acquisition unit at OSE
Fig. 19 (right): sensor and PD acquisition unit at GIB

Fig. 16: heating test with 530 kV / 2 h / 170 nF

Fig. 20: voltage time characteristics of all tests

ON-SITE TESTS
Besides the transport logistic of shipping the test-set to
site a concept was created to test all 7 phases with
minimal alteration works of resonant reactors considering
11 m phase distance at outdoor sealing ends. Only one
alteration of the resonant reactors had to be performed.

test voltage:
test duration:
test capacitance:
test frequency:
test current:

493 kV
1h
102 - 117 nF
24.5 - 26.2 Hz
8.2 - 8.8 A

(1)
(5)
(2)
(6)
(3)
(7)
(4)

Fig. 17: sequence of testing 7 phases

rd
After the 3 test (spare phase) the spare phase was unplugged from GIB at transformer no. 4 and the primary
compact plug-in SF6 sealing end was plugged into GIB
th
no. 4 (4 test). With all 6 pieces of 500 kV transformers
approx. 6 m GIB pipe was HV tested together with the
cable. All links to transformer were removed during the
tests.

The test set-up for on-site PD measurements should be

corona-free. Therefore, corona toroids and connections of
suited diameter have to be used. Sufficient clearance from
HV connections to any part of the construction should
prevent PD from earthed or potential free components. At
both sealing ends an inductive sensor (HFCT) was used
connected with a multi-channel PD detection system
whose acquisition units were connected with optical fibres
to controller and laptop in the control container. The
voltage signal for synchronization of PD measurement
was directly taken from the frequency converter. The
voltage was increased step by step PD controlled.

Fig. 21: set-up test phase no. 1

Fig. 22 (left): PD measurement at 493 kV at OSE

Fig. 23 (right): PD measurement at 493 kV at GIB
Synchronized two-side PD measurements enable time-ofarrival PD localization. In contrast, single side PD
detection is restricted to time domain reflectometry for PD
localization, which results in lower sensitivity and
accuracy [9, 4].

Jicable11 19 23 June 2011, Versailles - France

Close and Return

C.4.1

8th International Conference on Insulated Power Cables

C.4.1

reactors for extensive projects allows for more flexibility

regarding capacitance, set-up and above that it provides
redundancy. The on-site test was carried out successfully
in September 2010 including both, HV and PD tests.
REFERENCES

Fig. 24: set-up test phase no. 7

ON-SITE RE-PLUGGING PROCEDURE

For this project a special spare phase design and concept
was developed. The spare phase is mobile fixed at a
special carriage system under the ceiling of the
transformer cavern.

Fig. 25: spare phase carriage fixation in cavern

This design allows to reach each of the 6 pieces of
horizontal GIB cable enclosures to plug-in the 550 kV
compact plug-in SF6 sealing end of the spare phase. As
part of the commissioning tests this procedure shall be
demonstrated. First the spare phase was tested,
subsequently the regular phase in the same GIB cable
enclosure (transformer) was tested with 493 kV. The replugging procedure was performed in less than one day.
This was possible because of both, the modular set-up of
the test-set and the advantages of the plug-in sealing end
system.

Fig. 26 (left): carriage system for spare phase

Fig. 27 (right): ready installed spare sealing end

[1] F. Bernasconi, W.S. Zaengl, K. Vonwiller, 1979, A

rd
new HV-series resonant circuit for dielectric tests, 3
ISH Milan, paper 43.02
[2] W. Hauschild, W. Schufft, W. Spiegelberg, 1997,
Alternating voltage on-site testing of XLPE cables:
The parameter selection of frequency-tuned resonant
th
test systems, 10 ISH Montreal, Vol. 4, pp 75-78
[3] W. Schufft, W. Hauschild, 2000, Resonant test
systems with variable frequency for on-site testing
and diagnostics of cables, HV Testing Monitoring and
Diagnostics Workshop, Alexandria, Virgina
[4] U. Hermann, A. Kluge, R. Plath, 2007, Afterinstallation testing of HV/EHV extruded cable
systems procedures and experiences, Jicable07
[5] W. Hauschild, W. Schufft, R. Plath, U. Herrmann, K.
Polster, 2002, The technique of AC on-side testing of
HV cables by frequency-tuned resonant test systems,
CIGRE Session, paper 33-304
[6] R. Plath, U. Herrmann, K. Polster, R. Heinrich, W.
Kalkner, J. Spiegelberg, P. Coors, 1998, On-site PD
measurement on an extra-high voltage XLPE cable
line, Jicable98
[7] S. Sutton, R. Plath, G. Schrder, 2007, The St.
Johns Wood - Elstree experience testing a 20 km
long 400kV XLPE-insulated cable system after
installation, Jicable07
[8] CIGRE WG 21.16, 2001, Partial discharge detection
in installed HV extruded cable systems, CIGRE
technical report no. 182
th
[9] R. Plath, 2005, Multi-channel PD measurements, 14
ISH, Beijing, China
[10] J. Kaumanns, E. Plieth, R. Plath, 2003, On-site AC
testing and PD measurement of 345 kV / 2500 mm2
XLPE cable systems for bulk power transmission,
Jicable03, paper A8.4
[11] IEC 62067 Ed.1.1 2006-03, 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
[12] IEC 60840 Ed.3 2004-4 Power cables with extruded
insulation and their accessories for rated voltages
above 30 kV (Um = 36 kV) up to 150 kV (Um = 170
kV) - Test methods and requirements
[13] IEC 60060-3 Ed.1.0, 2006-02, Definitions and
requirements of on-site tests
GLOSSARY

CONCLUSIONS
For EHV projects in secluded regions a compact AC test
system has a lot of benefits. The typical cable lengths in
such projects are less than 1500 m. A mobile test system
as the described one meets the specification in all
matters. AC resonant testing in combination with
distributed multi-channel PD measurements using PD
sensors at each accessory ensures best efficiency for
EHV XLPE extruded cable systems. The modular
upgrading of the test system with two further resonant

AC
Alternating Current
ACRF Frequency Tuned Resonant Circuits
(E)HV (Extra) High Voltage
GIB/GIS Gas Insulated Bushing / Switchgear
HDPE High Density Polyethylene
HFCT High Frequency Current Transformer
OSE
Outdoor Sealing End
PD
Partial Discharge(s)
Q
Quality Factor
XLPE Cross-Linked Polyethylene

Jicable11 19 23 June 2011, Versailles - France