Close and Return

C3-6 10th International Conference on Insulated Power Cables C3-6

Integral sensing of HV cable joints – monitor operation and predict failures early
Ruben GRUND, Jens HOHLOCH; Pfisterer, (Germany/Switzerland), [email protected],
[email protected]
Rosalie ROGERS, Anja KAMMLER, Clemens POHL; AP Sensing (United Kingdom/Germany),
[email protected], [email protected], [email protected]
Henrik ROLAND; Energinet, (Denmark), [email protected]

ABSTRACT permissible temperature of a cable system, usually limited

at 90°C for XLPE-insulated cables [1], [2]. In normal
Due to space restriction and visual appearance, cable operation, operators prefer to have some capacity left in
systems are becoming an alternative to overhead lines, order to handle critical situations. Thermal overload due to
even in transmission networks. Furthermore, generally
speaking, utilities and network operators tend to increase excessive currents may lead to critical changes in material
the power transmitted which could be required in regard to properties and therefore shorten the service life.
overload behaviour. This additional load increases the During the project planning phase, cable transmission lines
requirements on cables, as well as the relevant
accessories. Nevertheless, safe operation is required and are designed in accordance with standards and technical
mandatory. This paper discusses the real life pilot guidelines with regard to a maximum limit current resulting
installation of three joints equipped with the world’s first- from the maximum permissible temperature. Influencing
time use of complementary sensing technologies: factors such as cable structure, if necessary, accumulation
Distributed Acoustic Sensing (DAS) and Distributed of several cables laid in parallel (often three-phase circuits)
Temperature Sensing (DTS). as well as soil and ambient conditions are taken into
account.
KEYWORDS This method of system design is based on relatively
roughly categorised calculations and empirical values,
FOC, DAS, DTS, smart joint, hipot test, underground which in practice do not always apply completely or are
power cable, monitoring, temperature, acoustic deliberately assumed to be very conservative.
During operation, the network operator intends to operate
INTRODUCTION AND MOTIVATION the cable system within the limits. But future cable systems
Energy distribution and transmission networks are are designed closer to the limits, so there will be times
constantly changing. This change is initiated by several when the operator will overload the cable for a limited
factors, such as changes in energy generation from large period, e.g. due to high feed-in power from renewables.
power plants to several medium sized power plants. Monitoring of decisive parameters (e.g. real thermal
Another additional aspect is power fluctuation which has conditions) during normal network operation can therefore
become a widely discussed aspect. If there is sun and wind make a contribution to optimized capacity utilization. In
there is a lot of additional power but a few days of cloudy addition, monitoring offers the possibility to detect
weather without a breeze requires additional backup characteristic changes at an early stage and to react to
generation. In this regard the expectation and our common emerging problems, thus contributes to achieving a high
approach in technology is another factor heavily influencing reliability of the entire grid. If a fault occurs in the system, it
our network setup. There is no acceptance of energy is also an important tool for fault location and makes it
disruption, a 100% continuous coverage of electricity at all possible to initiate repair measures rapidly and targeted.
time is expected. If a cable system is in service information about
These expectations in conjunction with the fluctuating temperatures is only available if an additional
sources, changes the requirement set for energy networks. measurement system is applied. Often these sensors are
Utilities need to plan for maximum single source load as optical fibres implemented in the cable’s screen or
well as 100% renewables. Therefore, there is a need to armouring as seen in Figure 1. Additionally, optical fibres
interconnect the distribution and transmission level, as well can also be used for measuring acoustic strain which can
as to increase the capacity for energy transmission. pinpoint cable faults to an exact location.

But how can this increase be managed? On one hand, one

could add additional infrastructure. Interconnecting the
mesh. At the same time bigger cross sections could be
used to increase capacity.
Nonetheless these are huge investments that need time.
Considering the planning and structuring of existing
networks, there are several potential methods to increase
energy transmission. Particularly cable systems have been
calculated and designed for a specific cause but usually are
not run at full capacity. But why aren’t these networks not Figure 1: Setup of FOC with power cable
run at full capacity?
These fibres usually are not implemented in cable joints but
A crucial factor limiting transmission power is the maximum in cables only. But cable joints are a crucial point of

Jicable'19 - Paris - Versailles 23-27 June, 2019 1/6

 Close and Return

C3-6 10th International Conference on Insulated Power Cables C3-6

information when it comes to temperatures and therefore insulated cables.

ageing and load indication. AP Sensing and Pfisterer
implemented a pilot installation in the network of Denmarks The outer housing offers protection against moisture and
utility Energinet at which distributed temperature and external mechanical forces. The space between the joint
acoustic sensing has been implemented within the joint as body and the outer housing is filled with filling compound
well as along the cable. This paper discusses the additionally leading to an improvement of mechanical
applicability and first on-site measurements of such a stability. The bonding cables (often coaxial bonding cables
system. are used) are connected to the screen of both sides of the
HV cable. Thereby the shields of different phases of a
GENERAL SETUP AND FUNCTION OF circuit can be cross-bonded externally in order to
CABLE JOINTS compensate shield voltages and currents. Alternatively, the
screens can also be grounded. Generally, treatment of
Due to transport restrictions, high voltage cables can only shielding ends is dependent on the earthing concept of the
be manufactured in limited lengths. For longer cable circuit and is determined in its design.
lengths, this means that cable joints (often installed
underground) are necessary for connecting cables. As with The design of a cable system is intensively proven in type
and prequalification tests according to IEC standards
the high-voltage cable itself, the cable joints must have a
[1], [2]. Beside electrical tests such as lightning impulse
high current carrying capacity on the one hand, and on the
other hand the electrical insulation between high-voltage voltage test, AC voltage test and partial discharge (PD)
and earth potential must be ensured. It is an essential measurement, heat cycles with temperatures slightly above
the maximum permissible conductor temperature as
requirement that cable joints and all other cable
occurring in real field operation are mandatory to ensure
accessories need to offer the same performance in regard
of transmission performance and service life as the cable functionality also under/after thermal stress.
itself. Nevertheless, cable joints are crucial components in the
network, e.g. improper installation may cause partial
6 discharges, accelerated ageing of electrical insulation and
2
1
4 finally may lead to a decreased reliability of the whole grid.
3
5 PILOT INSTALLATION – SENSOR
1 HV cable 4 Insulating joint body
EQUIPMENT INTEGRATED IN JOINTS
2 Bonding cable 5 Filling compound
3 Connector 6 Outer housing The pilot installation includes three cable joints (installed at
different phases of a three-phase circuit), the associated
Figure 2: Cable joint for Um = 145 kV linkbox and a separate box for electronic measurement
devices. In this setup some special sensor components
The basic structure of a complete state of the art cable joint
were implemented, which are evaluated as basis of a smart
is shown in Figure 2. The aluminium or copper cable
system consisting among others of:
conductor is contacted by a current-carrying connector. For
this purpose, usually screw or press connectors are used. • An optical sensor cable that has been routed
Conductor cross-sections commonly range up to 2500 through all three joints and allows the detection
mm2. According to the current state of the art, the electrical and localization of acoustic signals (DAS)
conductor connections are designed in such a way that no • Via further optical fibres, the temperature
higher temperature occurs at the connector than at the distribution within the cable joints can also be
current-carrying cable conductor [3]. measured (DTS)
The insulating joint body surrounding the connector • Two electrical temperature sensors each inside
consists of elastomeric material (usually silicone or EPDM) the cable joints for reference measurements
which has to withstand the electrical field stress not only The opposite side of the measuring cables, leading to the
under normal operating condition but also in the event of sensors inside the joints, ends in a manhole where the
circuit faults as well as transient events (e.g. switching electronic devices are placed and the linkbox is installed.
operations, lightning strikes). The insulating material During the commissioning test (hipot test) of the circuit, a
contains special conductive contours needed for keeping temporary setup was realized in order to check measuring
field stress at non-critical values. This technical principle is functions and to carry out parameterizations.
often referred to as geometric field control. During
installation the elastomeric joint body is temporary placed 2 1
at one cable end side and pushed back to the final position
after conductor connection is assembled.
The advantage of this slip-on technology lies in the fact that
the prefabricated joint body can be manufactured under
3
clean production conditions in constant high quality. 4

Electrical pre-tests are performed on each insulating part

at the end of production. At installation site, the insulation 5

part of the cable joint only has to be placed onto the

prepared and cleaned cable insulation. This procedure 1 Cable joint
2 Bonding cable
4 Fibre optic cabel (FOC)
5 DAS/DTS measuring system
minimizes the risk of assembly errors and helps saving 3 Linkbox
installation time compared to wrapping of insulation layers
as it was made in former times. The life span of such cable Figure 3: Principle setup of pilot installation
accessories corresponds to service life of XLPE- or EPR-

Jicable'19 - Paris - Versailles 23-27 June, 2019 2/6

 Close and Return

C3-6 10th International Conference on Insulated Power Cables C3-6

Optical Fibre Sensing Overview PROJECT DESCRIPTION

Two types of distributed optical fibre sensing systems were Three separate hipot tests took place on the 26th February
used to measure both the temperature and 2019 between 11:20 UTC and 16:32 UTC. The tests were
acoustic/dynamic strain. These two systems are described carried out on three separate circuit phases (L1, L2, and
below. L3) at a substation in Copenhagen, Denmark. The three
phases are shown in Figure 5.
DAS
Distributed Acoustic Sensing (DAS) systems use a fibre
optic cable (FOC), typically buried, to capture acoustic and
dynamic strain along the entire fibre’s length (Figure 4).
The system measures the strain on the FOC using
coherent optical time domain reflectometry (c-OTDR) [4].
AP Sensing have developed DAS systems able to use this
sensitive technique to acquire measurements up to a range
of 70 km, simultaneously recording acoustic signals for
every 1.28 meters continuously along the length of the
cable. This makes DAS systems very suitable for
measurement of large-scale linear assets.

Figure 5: Three circuit phases for power cables

Figure 4: DAS Setup (Copenhagen, Denmark)
DAS has been used for many applications, from intrusion Data was collected from both a DAS system and a DTS
detection [5] to cable faults. Recent research has shown system, each connected to two separate FOCs. Each FOC
that changes in temperature can also be measured using contained a single mode fibre connected to the DAS
the cOTDR technique, such as the AP Sensing Distributed system, and a multimode fibre connected to the DTS
Acoustic Sensor [6]. system. Both FOCs were mounted parallel to the circuits
however one FOC travelled outside of the joints as per
standard practice whilst the second FOC travelled inside
DTS the joints. Using a FOC inside the joints is a novel
technique of setting up FOCs with power cables that shows
Distributed Temperature Sensing (DTS), based on Raman
many benefits, as demonstrated in the data analysis
spectroscopy, is a method of measuring absolute
temperature along the length of a FOC. This technique is section.
well established and widely used to monitor the condition Each hipot test reached a full AC voltage of 134kV with a
of HV and EHV power cables. Like with DAS systems, DTS frequency of 22 Hz. In theory this should create a frequency
systems give the potential to simultaneously record the of around 44 Hz in the FOC’s. The first test took place on
temperature in 1 meter resolution. These measurements the first phase (L1) between 12:20 UTC and 12:57 UTC.
can be recorded along the full length of the FOC (up to The second test took place on the second phase (L2)
70km) without additional wiring or data transmission, starting at 13:38 UTC, reaching half voltage at 13:43 UTC,
providing an accurate picture of the temperature full voltage at 13:45 UTC, and ending at 14:49 UTC. The
distribution of the power circuit. final test took place on the third phase (L3), starting at
15:22 UTC, reaching full voltage at 15:32 UTC, finishing at
Temperature is the most critical physical parameter limiting
16:29 UTC. Data was collected for all three tests from both
the operation of XLPE/oil-filled power cables.
FOCs as well as both the DAS system and the DTS system.
Consequently, DTS systems are widely used for
temperature monitoring of HV and EHV underground
power cables, improving the safe operation of power DATA ANALYIS
circuits. Abnormal hot spots can be detected, showing
serious problems within joints, or indicating critical DAS
obstructions caused by environmental or infrastructural The results from the experiment highlight three main
conditions (e.g. backfill dry-out, district heat pipes etc.). benefits of installing FOCs inside joints: (i) the internal FOC
has lower levels of background noise compared to the
DAS and DTS combinations external FOC; (ii) it is possible to identify the time and
Combination of DTS and DAS systems provides a novel location of the hipot test by examining signal amplitude;
approach towards integrated monitoring of power cable and (iii) each hipot test demonstrates similar, expected
status. This includes continuous monitoring of the cable spectral characteristics for the internal FOC, showing that
conditions with regards to anomalous temperature and it is possible to detect hipot tests from frequencies alone.
acoustic characteristics that are indicative of a cable fault.
To obtain the energy from a DAS system, it is standard
The use of optical technology also makes DAS and DTS
practice to perform spectral analysis using a Fast Fourier
systems robust against electromagnetic interference from
Transform (FFT). The acoustic energy is then calculated
power cables.
by summing the magnitude of the FFT between two
frequencies. For this experiment, the energy was taken

Jicable'19 - Paris - Versailles 23-27 June, 2019 3/6

 Close and Return

C3-6 10th International Conference on Insulated Power Cables C3-6

between 4 and 250 Hz. Usually, when measuring the

energy at a joint location there is a lot of background noise.
This is due to the fibre traveling away from the joint
containing the power cable, resulting in relatively higher
levels of various environmental signals that the rest of the
FOC does not experience, since it is buried.
Below are distance/time plots of the acoustic data
measured by the DAS systems for the FOCs located both
internally and externally from the joint. Both plots were
taken prior to the hipot test at L2; the two yellow dashed
lines represent the joint location in either figure. The
external FOC shows a high level of background noise at
the location of the joint (Figure ).

Figure 8: Second hipot test on phase L2 (FOC inside

joint).
Figure , shows the same hipot test for the external FOC.
From this, the start and end times for the hipot test cannot
be distinguished from the high level of background noise,
whereas this is very clear for FOC running inside the joints
(Figure ).

Figure 6: FOC outside of joint at phase L2.

This is a usual result when looking at frequency energies

at joint locations. Comparing this to the FOC running
through the joint (Figure ) the level of background noise is
significantly lower. This enables analysis of the acoustic
signals emanating from the joint that would otherwise be
dominated by background noise.

Figure 9: Second hipot test on phase L2 (FOC outside

joint).
Another plot that supports this finding is a strain-frequency
plot for both FOCs before and during the same hipot test,
shown in Figure 10.

Figure 7: FOC inside joint at phase L2.

The second key result from the data analysis is being able
to identify correctly when and where the hipot tests are
taking place. Figure shows the start and end times of the
hipot test represented by two white horizontal lines.

Jicable'19 - Paris - Versailles 23-27 June, 2019 4/6

 Close and Return

C3-6 10th International Conference on Insulated Power Cables C3-6

Figure 10: Strain-frequency for Test 2.

Figure 11: Strain-frequency for FOC inside joints.

From this figure, the FOC outside the joint shows higher
levels of noise in frequencies of 80 Hz and above compared
to the internal FOC. Additionally, the external FOC shows
no clear difference between the strain-frequency response
before and during the hipot test.
Comparatively, the FOC inside the joint demonstrates
notably less noise compared to the external FOC at both
low and high frequencies. Secondly, there is a clear
spectral difference between the strain-frequency response
before and during the hipot test, as during the test the
signal was approximately 5 dB higher at low frequencies
and showed a sharp increase of approximately 30–40 dB
at 40–45 Hz.
The final finding from the DAS data analysis is that all
successful hipot tests produced the same frequency levels
which are only present during the hipot test. Figure 1 below
illustrates this by showing strain-frequency data for the
FOC running inside the second joint (L2) and the third joint Figure 12: Strain-frequency for FOC inside joints.
(L3). Data for both before and during the hipot tests for both
circuit phases were analysed to enable a direct comparison
between the different frequencies produced. Both hipot DTS
tests on either circuit phase showed an increase to 70 dB The temperature measurements by the DTS didn’t show
around 40-45 Hz. This frequency is not present when there any temperature variations during the hipot tests, neither
is no hipot test happening for either joint, concluding that on the fibre running in parallel to the power cable nor on the
this is due to the hipot tests. Additionally, this is very similar fibre installed directly in the three Smart Joints. This was
to the expected frequency to be produced which is 44 Hz. more or less expected as a hipot test applies only voltage
Figure 12 shows the same graph but zoomed in to the area on a power cable, no load. Below, Figure 3 shows the
of interest, showing that the DAS detects frequencies of 44 measurement results from the fibre directly installed inside
Hz which was expected. the Smart Joints. The variations at the beginning and at the
end are from the feeding fibre lengths and coming from
changing ambient temperatures. There is no temperature
change due to no load. In real operation temperature
increase are to be expected in relation to the load condition.
At the point of the hipot tests, the cable was not in
operation.

Jicable'19 - Paris - Versailles 23-27 June, 2019 5/6

 Close and Return

C3-6 10th International Conference on Insulated Power Cables C3-6

REFERENCES
[1] IEC 60840, Ed. 4.0, 2011, “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”
[2] IEC 62067, Ed. 2.0, 2011, “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”
[3] IEC 61238, Ed. 1.0, 2018, “Compression and
mechanical connectors for power cables”
[4] Zhengqing Pan, Kezhen Liang, Qing Ye, Haiwen Cai,
Figure 13: DTS measurements from FOC inside joints. Ronghui Qu, Zujie Fang, "Phase-sensitive OTDR
system based on digital coherent detection," Proc.
SPIE 8311, Optical Sensors and Biophotonics III,
83110S (29 November 2011)
CONCLUSIONS
[5] Cao, C., Fan, X. Y., Liu, Q. W., & He, Z. Y. (2015).
Usually FOCs for acoustic and thermal measurements are “Practical Pattern Recognition System for Distributed
laid parallel to cables, but cable accessories are not Optical Fiber Intrusion Monitoring System Based on
included. In a pilot project now some cable joints were Phase-Sensitive Coherent OTDR” Asia
included. During the commissioning test several Communications and Photonics Conference 2015,
measurements and comparisons were carried out, which ASu2A.145.
show promising results. Since the main purpose of the https://doi.org/10.1364/ACPC.2015.ASu2A.145
commissioning test is to prove a certain withstand level of [6] G. Cedilnik, G. Lees, P. E. Schmidt, S. Herstrøm and T.
the electrical insulation, high AC voltage is applied but no Geisler, "Pushing the Reach of Fiber Distributed
power is transmitted. Therefore no significant current flows Acoustic Sensing to 125 km Without the Use of
and no significant thermal effects are to be expected. Thus, Amplification," in IEEE Sensors Letters, vol. 3, no. 3,
further investigations of the DAS and DTS measuring pp. 1-4 (March 2019)
system must be carried out under different (including
extreme) load conditions.
The value of injoint monitoring have been proven by
comparing external FOC and internal FOC both during
hipot testing and without test voltages. There is a clear
advantage in having the FOC inside the joint from a data
analysis point of view, as temperatures and cable faults can
be more accurately located. Additionally, it is possible to
detect hipot tests from frequencies alone, as they are
expected to produce a frequency of around 44 Hz which is
evident from the data analysis. We have sensitive detection
means as well as potential to create an early warning
system for degrading joints, this is to be proven and will be
completed in future projects.

Jicable'19 - Paris - Versailles 23-27 June, 2019 6/6