A3-2 10th International Conference on Insulated Power Cables A3-2

Projects with Remote Installation (“Tube Post”) of Energy Cables in Ducts

Willem GRIFFIOEN, Christophe GUTBERLET, Alexandre UHL, Gregory LAURENT, Selim GROBETY;
Pumettaz SA, (Switzerland), [email protected], [email protected],
[email protected], [email protected], [email protected]

ABSTRACT safely been laid into the seabed. The position of the pipe
(i.e. also with respect to the seabed) can accurately be
A method was developed to install energy cables into ducts monitored using intelligent pigging, which has also been
and then further transport them through coupled ducts, like done in the Nissum Bredning project.
“tube post”, to any desired location, without the need to go
there with material, equipment and labour. There is almost And last but not least all cables can be installed from a
no limit to the FreeFloating distance over which cables can convenient launch location when using the FreeFloating
be transported. Many advantageous applications exist, like technique, enabling to reach crowded city centres, tunnels,
installing cables into crowded city centres from suburbs
and offshore cables from shore. A land project with and national parks, without the need to go there with
FreeFloating in Copenhagen is described, and a project in equipment, and even installing offshore cables from shore,
Thyborøn (Denmark), where offshore wind turbines were also array cables between the offshore wind turbines. The
connected by array cables installed from shore, even at latter enlarges considerably the offshore working window,
Beaufort wind force 8. allowing to keep on installing at bad weather conditions.
INSTALLATION CABLE IN DUCT (PIPE)
KEYWORDS
Energy cable; HV cable; Installation; Duct; Pipe;
FreeFloating; Water; Pressure; Pig; City centres; Offshore;
Wind Energy; Remote.

INTRODUCTION Pulling (winch)

Energy cables can be installed aerial or underground. In
the latter case the cables can be direct buried or installed
in ducts. Special methods have been developed to install
cables into ducts using water under pressure [1,2,3,4]. A
typical advantage is that operation is economical, direct
installation with all material, equipment and labour at one
Floating (WATUCAB)
(entry) side of the duct. Moreover, long lengths can be
installed and the methods are friendly to the cable. Maybe
the most appealing variant is FreeFloating where the cable,
once installed in the duct, is flown further by the sole action
of water, like “Tube Post”, from any convenient launch
location to any desired destination location, almost without
any limit to the distance over which the cable can be
WaterPushPulling (WATUCAB)
transported, the limits discussed in this paper. This
technique already proved to work in an installation trial at a
test site in Saint-Étienne-du-Grès (France) [3]. In this paper
two projects are described with FreeFloating, a land project
in Ballerup, a suburb of Copenhagen (Denmark), and the
Nissum Bredning offshore wind farm project in Thyborøn
(Denmark). FreeFloating (WATUCAB)
BENEFITS CABLE IN DUCT
Fig. 1: Cable-In-Pipe installation techniques
Several advantages can be recognized for cable in duct
solutions. In general cables can be removed or replaced Pulling (winch)
without digging up. Protection in the pipe is even better The traditional way to install cables into ducts is pulling
than for direct buried cables [4], because of the free space them with a winch, see Fig. 1. For this first a pulling rope
in the duct, a well-known fact in Telecommunications. has to be installed. Also installation equipment and labour
There are specific benefits for both land and offshore are required at both ends of the duct. Furthermore the
applications. On land ducts can be laid in short sections capstan effect (friction of the cable under tensile load in
and then simply be coupled together. No need to keep bends) limits the cable lengths which can be installed in one
trenches open for long lengths and long time, reducing pull. Synchronization between winch and drum pay-off is
neighbourhood disturbances. For offshore applications no often troublesome. Three “WATUCAB” techniques, using
cable armouring is needed, which allows to use standard water under pressure, have been developed to install
“land” cables and save a lot on costs. Also AC losses are energy cables into ducts, see Fig. 1. The typical drawbacks
minimized. Additionally, the risk of cable damage is smaller for winch pulling are taken away, and also the forces on the
because they are installed after the pipes (ducts) have cable (and wear) are reduced.

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

A3-2 10th International Conference on Insulated Power Cables A3-2

the duct trajectory, even when the latter is extremely curved

and with small bend radii (a 82 mm 3x36 kV cable could be
installed over 646 m into a HDPE duct with internal
diameter of 102 mm which was wound in 46 coils with a
continuous bend radius of 2 m [5]).
FreeFloating
The third technique, called FreeFloating, is maybe the most
appealing one. It starts after a cable has been entirely
installed by WaterPushPulling, with a special pig used.
Then the duct is extended at the entry side such that the
cable is entirely inside, and with some space to insert a rear
pig. The rear pig can either be attached to the cable, like in
Fig. 3, or placed loosely as a “bumper pig”. Next the duct is
closed and water under pressure admitted. The rear pig is
“communicating” with the front pig such that they share the
water pressure. In this way the cable is effectively
PushPulled by the sole action of water, and travels further
like “tube post”. The cable can be placed at any desired
location. There is in fact no limit how far the cable can be
transported, as the water pressure difference is mainly
Fig. 2: Cable installed by using water effective at the pigs. There might be some viscous pressure
loss over the feed length of duct, but this can be reduced
Floating at wish by reducing the cable (and water) speed. In fact,
In the first technique, called Floating, water under pressure higher cable speeds are reached with FreeFloating (in the
is injected into the duct with cable, creating a high speed Copenhagen and Nissum Bredning projects, see further,
(higher than the cable speed) water flow, while at the same 25 m/min and 28 m/min were reached, respectively) than
time the cable is pushed into the duct (and pulled from the with WaterPushPulling, for ducts not too narrow and long.
drum), see Fig. 2. The high-speed water flow creates a
distributed drag force propelling the cable. This distributed
force locally compensates the friction between cable and
duct, avoiding axial force build up in the cable, hence
eliminating the capstan effect. The same trick as with cable
blowing [2], a technique used worldwide today to install
optical cables into ducts. Extra beneficial effect with
Floating is the buoyancy of the water, reducing the friction
between cable and duct. With this technique extremely long
installation lengths can be reached (with Low Voltage
cables already 10 km has been reached), also in
trajectories with many bends. Moreover, there is the benefit
of single point entry (installation equipment, cable drums
and labour), reducing costs considerably. The technique is
user and cable friendly (low forces, no cable wear) with
compact equipment and does not suffer from
synchronization problems with the cable drum. With the
present equipment a comfortable cable speed of 15 m/min
can be used.
WaterPushPulling
The second technique, called WaterPushPulling is mainly Fig. 3: Rear end of cable prepared for FreeFloating
the same as Floating, except that a pig is mounted at the
foremost end of the cable. Now all forces exerted by the FIRST FREEFLOATING TRIAL
water under pressure are concentrated at the cable front
head and the water flows with the same speed as the cable. FreeFloating was demonstrated to work in a trial at the
The latter makes it possible to still use relatively small EHTP / Plumettaz facilities in St. Etienne du Grès (France),
pumps for larger diameter ducts (e.g. larger than 100 mm in a test circuit of a 976 m PVC duct of 160/152 mm, buried
internal diameter). The relatively high pulling force at the in the ground [3]. This circuit has been built in 4 loops and
cable front head also enhances passing sharp bends. But, contains 14 bends of 90° and 3 siphons. This circuit was
the capstan effect is back again. Fortunately, duct used earlier to demonstrate WaterPushPulling of a 82 mm
trajectories for energy cables are rather straight and 90 kV cable, with 1000 mm² aluminium core and weight of
buoyancy has not vanished. With balanced pushing and 68 N/m. To demonstrate FreeFloating, where the cable has
pulling forces (still lower than with winch pulling) installation to “travel” some length “loose from the machine”, a shorter
lengths can also be very long (3.3 km reached with cables (700 m) and slightly larger (89 mm, 84 N/m) cable was
with aluminium core), usually much longer than with winch used. After installing this cable by WaterPushPulling (see
pulling. When using a “sonic head” (pig with valve that Fig. 4) the cable could be FreeFloated further to the end,
opens at adjustable pressure) the advantages of Floating with a pressure of only 2 bar (maximum pushing and pulling
and WaterPushPulling can be combined and optimized to force on the cable < 200 daN).

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

A3-2 10th International Conference on Insulated Power Cables A3-2

recoupling of the ducts the cables for the 1st section could
be installed behind the installed cables by
WaterPushPulling again. Advantage is that all the work can
be done from one launch location, enabling deployment in
city centres from surrounding entry points. After the initial
learning curve of the first time FreeFloating in a real project
it was possible to install 2 cables in one day (1st cable
installed by WaterPushPulling, further installed by
FreeFloating and 2nd cable installed behind the 1st by
WaterPushPulling). In the project the water was recycled
and the remaining water brought to a waste water station.

Fig. 4: FreeFloating trial in St. Etienne du Gres

COPENHAGEN (LAND) PROJECT

The first real project where FreeFloating was used was in
Ballerup, a suburb of Copenhagen. Here 6 cables were
installed for Energinet in 2 sections of the connection
Veljeå-Ejbygård, see Fig. 5. Cable lengths were 1695 m
and 1574 m in Section 1 and 2, respectively. The cables
were 132 kV cables with 1400 mm2 solid Aluminium
conductor, with outer diameter 94 mm and mass 9.2 kg/m.
Their semi-conductive PE jackets contain graphite, Fig. 6: Installation of the cable in Copenhagen
resulting in a relatively high coefficient of friction. The
cables were installed into 160/140 mm PE ducts. Cable NISSUM BREDNING OFFSHORE WINDFARM
installation was carried out by NCC.
Siemens Gamesa Renewable Energy A/S has designed
and installed an offshore wind farm with 4 wind turbines of
7 MW in the Nissum Bredning (Limfjord) at Thyborøn,
Denmark, see Fig. 7. In this project they implemented 8 of
their innovations, one of them the Cable In Pipe solution.
Here the cables were not directly laid into the seabed, but
installed into HDPE pipes, jetted previously into the seabed
as a bundle of 4. Cable and pipe installation was
subcontracted to JD Contractor A/S. All three techniques
where water under pressure is used to install the cables in
the ducts were used in this project, Floating,
WaterPushPulling and FreeFloating, the latter allowing to
install from shore (see Fig. 8), also at Beaufort wind force
8, showing how much the working window can be enlarged
for offshore in the “bad season” (and bad it was!).
All connections were made with three 72 kV (1x630 mm2
Alu stranded core) cables with outer diameter of 68.1 mm
and with a mass of 4.6 kg/m, installed into 110/90 mm
HDPE ducts in a bundle of 4 configuration (one spare duct)
and a 40/29 mm duct in the centre for an optical cable.
Cable lengths were between 950 m and 1250 m.
In Fig. 7 two details are enlarged, showing the duct return
loops at the foot “bunkers” of the turbines. To FreeFloat a
cable to be placed between the two turbines in the front
(with the enlarged views) this cable is first installed by
WaterPushPulling from land (see Fig. 8) until it is entirely in
Fig. 5: Trajectory Copenhagen FreeFloating project the duct. Then it is FreeFloated further, passing the slack
The cables for the 2nd section were installed by duct loops at the foot of the turbine in the front at right, the
FreeFloating, after initial inserting the cables into the duct J-tube up, through the bunker to the other side of it, through
(1st section) by WaterPushPulling (see Fig. 6). After the duct return loop at right, back through the bunker, J-

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

A3-2 10th International Conference on Insulated Power Cables A3-2

tube and duct loops, and then further to the turbine in the now the cable also passing through the return loop in the
front at left, again through the duct loops and the J-tubes. turbine in the front at left, the cable finally reaching the
The duct return loop in the turbine at right has a bend turbine in the back at left. The second duct return loop has
diameter of 5.25 m and is reinforced with a steel cross to a diameter of only 3.25 m (radius 1.625 m!), without the
keep its shape. Also double loop FreeFloating was done, need for reinforcement, and also here the cable could pass.

Fig. 7: Nissum Bredning Offshore Wind Farm, with details of FreeFloating return loops
Although all cables could be installed from shore, 2 cables
were installed from the vessel at a calmer day, to show that LIMITS OF FREEFLOATING
this is also possible (solution for wind farms far away from In this paper the limits of the FreeFloating technique are
shore, but in the next section also FreeFloating from shore discussed. What is the maximum length over which the
to remote offshore wind farms is discussed). In total 12 cables can be FreeFloated, and at which speed (high
Medium Voltage cables were installed. For monitoring and enough for the installation to still be economical)? And what
communication optical cables were installed into HDPE about hydrostatic pressure differences at large elevation
pipes (central in the bundle) using the Floating technique, differences over long length? In this paper it is argued that
this time all the way from the substation. A MiniJet with FreeFloating is still possible over long distances (40 km!)
sonic head was used, longest distance reached of 3.8 km. while at the same time a high speed (40 m/min!) can be
The first kWh power was produced on February 18th 2018. reached. This means that in 24 hours a cable can be
installed, including preparation. And with a second or third
feeder duct the daily production can be enlarged (still with
one launch unit, with simple water feeding units for the
cables underway with FreeFloating), making installation
economical also for such long distances. Finally it will be
treated how to optimize pressures and how to handle ducts
with differing diameters.

Elevation differences
When a cable is FreeFloated over a very long length (many
cable lengths), there might be a large difference in
elevation between the launch point and the end point (or a
point halfway). As every 10 m elevation difference is
equivalent to a hydrostatic pressure difference of 1 bar, this
might be a limiting factor for the distance over which
FreeFloating can be used. For downhill installations the
maximum duct pressure can be limited using a pig with
safety valve [5]. Uphill there is no problem with water
pressure getting too high, but there will be reduced
pressure available for the pig, reducing the installation
Fig. 8: Installation of the cable at Nissum Bredning

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

A3-2 10th International Conference on Insulated Power Cables A3-2

performance. In some cases intermediate water pumps can Cable speed

be used to boost the pressure. This is all for land projects.
For offshore (submarine) projects the duct ends at When the cable end hits an obstacle and comes to a
destination are usually not much higher (or lower) in sudden stop, it will experience a compressive axial force
elevation than at the launching point, so here there is no under which it will buckle in the duct (note that with buckling
hydrostatic problem. And when there are deep dips in the is meant that the cable shows undulations, it does not mean
duct route, the water pressure outside the duct increases that the cable is out of specification). As the buckling
with the inside, so the duct will not see too high pressure. “absorbs” effective cable length, not the whole cable is
stopped at once. The portion of stopped and buckled cable
How far and fast can cables be FreeFloated? will increase, like a wave traveling backwards (same as for
FreeFloating is normally done at very low pressure losses water hammer). First it is calculated how much relative
(pig pressure close to applied pressure). However, when length εs of cable can be stored as a function of axial
FreeFloating (and water) speed are high and the total duct compressive force Fc. The worst case situation is
length is long, a viscous pressure drop limits the available considered that the duct is fixed in its position, not moving
pressure difference to move the cable. Fortunately, it is still sideward or elongating. The total “absorbed” relative length
possible to reach long distances (e.g. 40 km at a speed of εs of the stopped cable length Ls is given by [7]:
40 m/min!). And higher speeds are possible at shorter  2
lengths. As long as the cable speed remains smaller than  1 cb ( Dd − Dc ) 
εs
= + F [2]
60 m/min, sudden stops are still okay for cable and duct  kc 4π 2B  c
(water hammer, cable inertia), as will be explained.  

Water speed Here kc is the effective spring constant of the cable, B the
stiffness of the cable, Dc the diameter of the cable, Dd the
As for FreeFloating the water speed must be at least as internal diameter of the duct and cb a geometric factor
high as the cable speed (and for Floating even higher), it is which is equal to 2.23 for 2-dimensional (sinusoidal)
relevant to consider water hammer effects in this paper. buckling and equal to 4.93 (= ½π2) for 3-dimensional
Moreover, the theory which is developed in this paper for a (helical) buckling. The left term is the relative axial
sudden cable stop finds a lot of common ground with the compression of the “straight cable” and the right term the
existing theory of water hammer, which is therefore treated “buckling relative storage length”. When the cable with
first. Water hammer (or, more generally, fluid hammer, also initial speed vc suddenly stops, not the whole cable stops
called hydraulic shock) is a pressure surge or wave caused instantaneously. First the front end stops and then the
by a fluid (usually a liquid but sometimes also a gas) in amount of cable coming to a standstill grows backwards,
motion when it is forced to stop or change direction like a sound wave, with a speed vs given by:
suddenly (momentum change). A water hammer e.g.
occurs when a valve suddenly closes (also other causes of vc
vs = [3]
sudden blocking possible, e.g. when a cable or pig passes εs
or hits a duct narrowing) somewhere downstream in a duct
system, and an upstream pressure wave propagates The mass Ms stopped in a time Δt is given by:
through the duct. This pressure wave can cause major
problems, like duct bursting. When a valve in a duct is =
Ms mcv s ∆t [4]
suddenly closed, the moving column of water will stop. But,
this is not occurring instantaneously for the entire column Here mc is the mass of the cable per unit of length. The
of water (which would result in infinite pressure when the change of momentum Msvc of the stopped cable is equal to
valve is closed instantaneously). First the water at the valve FcΔt, so it follows:
stops and a pressure wave travels backwards, the amount
of water which has stopped growing with the speed of
Fc = mcv sv c [5]
sound c (about 1500 m/s in water). From this the
This equation looks similar to the Joukowsky formula for
Joukowsky formula follows for the water hammer pressure
water hammer (dividing by a surface in m2 changes force
p when a fluid with speed v is suddenly blocked [6]:
into pressure and mass per unit of length into density).
p = ρ cv [1] Writing out further, with equations [2] and [3], it is found:

Here ρ is the density of the fluid (1000 kg/m3 for water). mc

=Fc ⋅ vc [6]
Example: For a water speed of 1 m/s (60 m/min) this would 2
1 cb ( Dd − Dc )
result in a pressure of 15 bar. In pipes the speed of sound +
in water is lower because of expansion of the duct. For kc 4π 2B
even relatively thick-walled HDPE ducts with SDR 11 (duct
OD divided by wall thickness) the speed of sound would From equations [5] and [6] also the speed vs of the “buckled
already decrease to 23% of the speed in bulk water, and cable wave” follows:
the water hammer pressure decreases proportionally. As
1
such pipes are rated 16 bar for their lifetime, the short term vs = [7]
 2
water hammer will by far not be a problem for water speeds 1 cb ( Dd − Dc ) 
of 60 m/min. Moreover, when no valve is suddenly closed mc  +
 kc 4π 2B 
and the water hammer is caused by a sudden stop of cable  
and pig, the safety valve in the pig [5] will limit the pressure
to a safe value (of course the opening in the safety valve Example (cable and duct used in the Nissum Bredning
shall be large enough to release the full water flow, which project): Cable 72 kV (1x630 mm2 Alu stranded core),
is also dependent on the duct and cable diameter). diameter Dc of 68.1 mm, mass mc of 4.6 kg/m, effective

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

A3-2 10th International Conference on Insulated Power Cables A3-2

spring constant kc of 60 MN (educated guess, 0.1% strain pressure drops in duct 1 and duct 2 (using Blasius). But,
at 60 kN) and stiffness B of 2500 Nm2 (rule of thumb the best results are obtained when optimizing the pressure
estimation) in SDR 11 duct 110/90 mm. For 2-dimensional in a feedback loop, maximizing the flow that comes out at
buckling (worst case) a force of 12.9 kN is found for a cable the duct connection.
speed of 1 m/s. For this force the cable is compressed
axially by about 0.022% and buckling takes about 0.014%,
a total of 0.036%. The backwards “wave” travels with more
than 2800 m/s (faster than the water wave). Max pulling
force on the cable is 18.9 kN, so with 1 m/s cable speed the
maximum force at crash is still well below this maximum.
The force on the cable at sudden stop will probably be less,
because the duct will also expand during cable buckling.
The maximum sidewall forces and the minimum bending
radius of the buckled cable can also be calculated. Their
values are far away from the critical values.
Fig. 9: Schematic view of connection to smaller duct
The above equations were derived for a sudden cable stop
due to blocking at the front end. It is also possible that
CONCLUSIONS
blocking occurs at the cable inlet, e.g. when a lump is
present in the cable jacket. Now the cable stops from the The remarkable technique of installing a cable into a duct
cable inlet, and the elongation is tensile. In this case a by FreeFloating (from any suitable location to any desired
forward “wave of cable under strain” is travelling, not under location, avoiding difficult to access places) has proved to
compressive but under tensile stress. Buckling storage will work in a test trial and in two pilot projects, one on land and
not occur, so equations [6] and [7] will be with a cb value of one from land to offshore. Advantages are huge, like the
zero. For a cable speed of 1 m/s the forces on the cable at possibility to install offshore cables in extreme weather
sudden stop will be 16.6 kN (still okay) and the forward conditions. It is argued that this can even be done over long
“wave” travels with speed of more than 3600 m/min. distances, e.g. 40 km away from shore, with economical
speed of 40 m/min.
Viscous pressure drop along duct
The fluid speed v for a pressure drop p over a length of duct Acknowledgments
L is given by Blasius’ equation [6]: We wish to acknowledge the following persons: Niels-
Jørgen Borch-Jensen, Johnny Sørensen, Jesper Møller,
4/7
Dd5/7 p (Siemens Gamesa), Jesper Lykke Nielsen (JD Contractor),
v = 2.9   [8] Jean Fehlbaum (Nexans), Preben Søgaard-Jørgensen,
µ1/7 ρ 3/7 L
Lars Højsgaard (NKT), Anders Thanning Lemming
Here μ is the dynamic viscosity (10-3 Pas for 20 °C water) Anderssen, Daniel Johan Brøndum, Kenneth Pfeiffer
of the fluid (note that the majority of the duct is without Andersen, Rasmus Aabye Olsen (Energinet), Thomsen
cable, so Dd can be taken as the hydraulic diameter). In Kim Høgh, René Kühle Pedersen (NCC), Morten Kosiara
order to reach a speed of 40 m/min over a 160/130 mm (Entrepenør Morten Kosiara), Olivier Perignon (EHTP),
duct with length of 40 km a water pressure of 12.4 bar is Jean-Pierre Pousaz, Victor Chaves, Feti Kont and Yvan
sufficient, leaving some pressure to FreeFloat the cable. Chappuis (Plumettaz).
REFERENCES
Different duct diameters
Finally it will be treated how to optimize pressures and how [1] G. Plumettaz, J. Heinonen, 2009, "High voltage energy
to handle ducts with differing diameters, e.g. from export to cables go underground – how to improve installation
array (at platform) or at outer edges of a wind farm. The efficiency", Proc. 58th IWCS, 169-174.
cable then has to pass places with changes in inner duct [2] W. Griffioen, P. Blanchet, G. Plumettaz, 2011, “New
diameter, usually to a smaller one, see Fig. 9. In this case techniques to install power cables into ducts”,
the pig is replaced by a smaller one when the cable arrives, Proceedings of Jicable Conference, vol. 1, 72-77.
and reinserted into the smaller duct, closing again the duct [3] W. Griffioen, C. Gutberlet, G. Plumettaz, 2012, “New
system. FreeFloating can then be restarted until also the technique to install power cables into ducts”, Journal
rear pig comes out and is replaced, after which the cable is of Energy and Power Engineering 6, 1263-1275.
FreeFloated further into the smaller duct. No need to store
heavy cable drums at platforms in offshore wind farms! [4] W. Griffioen, C. Gutberlet, J. Mulder, L. Højsgaard, W.
Grathwohl, H. Bringsell, J. Sørensen, N-J. Borch
During the time that the cable is partly in the larger and in Jensen, 2015, “New approach to installation of
the smaller duct, pressure regulation is needed at the offshore wind energy cables”, Proceedings of Jicable
connection point. FreeFloating goes best when the force F2 Conference, paper B4.1.
on the front pig is about the same as the force F1 on the [5] W. Griffioen, 2018 “”Sonic head”, a pig enhancing
rear pig. For this a pressure pm follows in between the pigs: several techniques to install cables into ducts in many
ways”. Proc. 67th IWCS, paper 16.5.
D12 p1 + D22 p2
pm = [8] [6] V.L. Streeter, 3rd edition 1962 “Fluid mechanics”,
D12 + D22 McGraw-Hill Book Company, inc. New York.
The optimal pressure pm might differ a bit from the one of [7] W. Griffioen, W. Greven, 2008, “Protected Microducts,
equation [8]. Usually it is better to have a slightly higher F2 Overview and Current Trends”, Proc 57th IWCS, 152-
than F1. Also corrections have to be made for the viscous 159.

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