Close and Return
A.7.5
8th International Conference on Insulated Power Cables
A.7.5
DEVELOPMENT AND QUALIFICATION OF DEEPEST WATER POWER
UMBILICAL
Arild FIGENSCHOU, Jan Ole DUNSERUD, Aker Solutions, (Norway), [email protected],
[email protected]
Daniel ISUS, Grupo General Cable Sistemas S.A., (Spain), [email protected]
Emmanuel BIC, Silec Cable, (France), [email protected]
ABSTRACT
This paper describes the world’s first umbilical system to
deliver substantial power at 2700m water depth. The
system is developed to deliver power to subsea pumps
through a dynamic power cable hanging from a moored
turret buoy in a catenary installation down to the seabed.
Attention is focused on the development, design and
qualification of the umbilical and its main components, the
power cables, performed in close cooperation between
the different companies involved.
KEYWORDS
Power umbilical, dynamic, submarine, deep water,
boosting, pumps
INTRODUCTION
As the oil and gas industry looks for ways to keep pace
with the growing worldwide demand for oil, subsea
developments are moving into ever increasing water
depths and operators are looking towards subsea
boosting technology as a means of getting the most out of
their reservoirs. The power distribution umbilicals, which
are needed to supply electrical power to the subsea
boosting pumps, will be challenged by more extreme
tension loads due to the deeper waters.
This paper presents a technical description of a power
umbilical to be installed in the area of the central Gulf of
Mexico at 2700m water depth; the first of its kind to
transmit power over this depth sufficient to power subsea
boosting pumps. The requirements of the power umbilical
system required a transmission of 1.7MVA including two
three-phase supplies and a spare three-phase circuit;
therefore, a total of nine cables, rated 12/20 kV with a
conductor cross-sectional area of 150mm2 each. In
addition, due to the long length of the umbilical cable,
including dynamic (that part of the umbilical suspended in
water) and static (that part of the umbilical laying on the
seabed) sections, 2 sets of splices were included in static
section.
Attention is focused on the development and design
principles of the umbilical and its main components,
presenting the challenges encountered, which to date are
unique. Also, a thorough qualification program is
described that has verified the umbilical’s ability to survive
the rigours of installation and operational life at the most
extreme conditions.
ULTRA DEEPWATER CHALLENGES
The traditional way to limit the strain in umbilicals is to
increase the axial stiffness by adding steel armouring. But
for very deep water applications, steel as armouring
material does not work efficiently to reduce the strain of
an umbilical. If a structural member suspended from a top
point is hanging vertically and exposed to its gravity load
only, the strain close to the hang-off point is proportional
to the free hanging length and density, but the opposite
applies in relation to its stiffness. Therefore materials used
for stiffness enhancement lose effectiveness as the water
depth grows. At a certain water depth the added weight
from steel armouring increases the strain rather than
reducing it.
This has made the industry look for light but still stiff
material for umbilical stiffness enhancement. As a result
of many years of development and testing, a carbon fibre
rod reinforced umbilical system has been developed and
patented. The carbon fibre rods, which are bundled into
the umbilical alongside the other elements in the
umbilical, have about the same stiffness but only one fifth
of the density compared to steel. This change has
eliminated all practical water depth limitations in terms of
strain and stress induced by the gravity loads, and has
been used in a number of steel tube umbilicals over the
last several years.
The control of strain is particularly important for high and
medium voltage electrical power umbilicals. A typical safe
long-term strain limitation of power cables is in the range
of 0.15%; this can easily be achieved for dynamic power
umbilicals using the carbon fibre rod system.
Furthermore, by use of a special hang-off system
described below (Umbilical mechanical properties), the
cables are “free floating” in the dynamic bending zone,
thus avoiding any uncontrolled loads upon the cables in
the zone and limiting the strain to that caused by bending
only.
Comparison of carbon fibre rods vs. steel
as armouring material
In shallow waters in combination with a harsh
environment a high weight to diameter ratio is required for
the dynamic umbilicals in order to perform well. This is not
the case for power umbilicals in deep waters. Instead, the
primary concern is to reduce the global strain of the
umbilical system. The control of global strain of an
umbilical in deep waters may be achieved by two
methods: reduction of the underwater weight, and/or by
increasing the axial stiffness of the umbilical.
The following example provides a comparison of steel
versus carbon fibre rods as a stiffness enhancement
system for an umbilical:
The density of steel is 7850kg/m3 in air compared to
1600kg/m3 for carbon fibre reinforced vinylester rods. The
stiffness modulus of steel is 200000 MPa compared to
150000 MPa of a composite carbon fibre reinforced rod.
As the strain is proportional to the weight and the inverse
applies to the stiffness, reducing strain by the use of
Jicable’11 – 19 – 23 June 2011, Versailles - France
Close and Return
A.7.5
8th International Conference on Insulated Power Cables
carbon fibre rods is much more efficient than using steel.
By comparing the Young’s modulus to the density for steel
and for carbon fibre rods, the factor is 8.9, (density of sea
3
water is set to 1025kg/m ):
Steel:
Carbon fibre rods:
Ratio of stiffness enhancement efficiency:
Working example:
The following is a comparison of a steel reinforced power
umbilical to a carbon fibre rod reinforced power umbilical.
In this example the calculation is based on a power
umbilical with six medium voltage power cables, six ½”
and one ¾” steel tube service lines and two standard
signal cables. The stiffness contribution from the power
cables is disregarded due to thermal expansion and
creep, thus the power cables should not be considered as
strength members in operation. The diagram gives the
global strain of the umbilical systems as a function of the
amount of armouring in mm2, in the cross-section of the
umbilicals.
A.7.5
deepwater umbilicals.
POWER UMBILICAL DESIGN
The umbilical design is based upon the bundling of the
internal elements in a gradual spiral along the length of
the umbilical. The angles of the spiral relative to the
centre line of the umbilical range between one and three
degrees only. This gradual spiral provides several
advantages, such as low rotational forces, and a high
capacity to withstand axial compression forces of the
umbilical.
The internals of the umbilical are held in place and
separated by stiff plastic spacers that run along the length
of the umbilical. The spacers are shaped such that when
bundled together they form internal longitudinal voids or
conduits through which the umbilical elements pass. Each
conduit is exactly dimensioned to suit the internal
elements such as a tube or cable that it will contain, with
an all-round gap of approximately 1mm. This provides the
free movement within the umbilical during handling, and
allows the umbilical to be spooled and reeled to the
respective bending limits of flexibility of the internal
elements.
The use of plastic spacers also provides benefits in the
distribution of forces throughout the cross-section of the
umbilical when it is subjected to high squeeze pressures.
Such a benefit is especially advantageous for deep water
installation, as it helps prevent the caterpillar grip
pressures from causing damage to the umbilical internals.
Another important feature of the umbilical with bundled
elements in gradual spiral along the length of the umbilical
is the improved fatigue life due to low friction between the
components.
In Figure 2 the cross-section of the present umbilical is
shown, including a triad formation of three-phase circuits
with a total of nine cables rated 12/20 kV with a conductor
cross-sectional area of 150 mm² each.
Fig. 1: Umbilical strain versus steel or carbon rod
armouring in case of 3,000m water depth
The design limit of copper conductors is normally in the
range of 0.15% strain. As shown in the graph above, this
can easily be achieved by reinforcing with carbon fibre
rods to achieve global strain of 0.15% by including 2750
mm2 of rods in the example cross-section. This
corresponds to about 83 carbon fibre rods of 6.5mm
diameter. The rods can easily be integrated in the crosssection without changing the global diameter of the
umbilical and without any serious weight impact.
To achieve the same global strain limit by use of steel
reinforcement it takes in excess of 15000mm2 of steel to
reach a global strain of 0.15%. This result is due to the
added weight from the steel. As a further illustration, if a
single steel member is hung free from the surface 3000
meters vertical into sea water the strain is 0.1% by only
suspending its own weight. It is therefore clear that steel
is not a suitable material for reinforcement of ultra
Fig. 2: Power umbilical cross-section (dynamic
section – some rods are replaced with plastic fillers in
the static section)
Jicable’11 – 19 – 23 June 2011, Versailles - France
Close and Return
A.7.5
8th International Conference on Insulated Power Cables
A.7.5
combination of material creep and temperature effects
(expansion and contraction) in the copper cores makes
the evaluation of the fatigue life very uncertain.
Table 1: Table of power umbilical components in Fig.
2
Umbilical electrical properties
There is no transformation of the voltage, so with a motor
voltage of 4.4kV, it is necessary to have an output voltage
from the variable speed drive of 6.1kV to obtain the rated
voltage for the motor. This corresponds to a voltage drop
of 28% and reduces the air gap torque. This and the
voltage increase due to load reduction were found to be
acceptable for the operation.
The magnetic field set up by the load current will induce
currents and voltages in nearby conductive elements. The
induced current in the grounded elements such as
screens, steel weight elements and carbon fibre rods
gives additional losses that can be handled by thermal
analysis to check if any elements are overheated. The
voltage induced into the cores within the same circuit and
into the neighbouring circuit is more complicated because
it affects the operation of the pump motor. This negative
effect can be cancelled by twisting the circuits within the
cross-section. That method would increase the required
number of splices and the cost dramatically and was
rejected. Another solution is to have the 3 circuits in
different layers with opposite twisting direction. This would
increase the outer diameter significantly and a large
number of weight elements would be required to maintain
the weight to diameter ratio needed. It was therefore
decided to keep the most efficient design and study more
closely what impact the induced voltage would have on
the operation of the motor. By use of a Finite Element
method it was found that the negative sequence voltage
was less than 1%. This is acceptable according to IEC
60034-26 without de-rating the motor.
To minimize the fatigue of the cables in the dynamic
bending zone, a concept was developed with the electrical
power cables “free floating” in the dynamic bending zone,
thus avoiding any uncontrolled loads upon the cables in
the zone. The carbon fibre rods are anchored at the top
hang-off, but the electrical cables run through the hang-off
without any constraint and are therefore free to expand
and contract. Instead of a topside cable termination, the
cables are hung-off inside the umbilical with a “soft clamp”
system. This system is basically a vulcanized rubber
sheath that is applied to the outside of the cable over a
length of ten metres, increasing the outer diameter and
providing friction against the inside of the cable conduit in
the umbilical, without creating a “hard point” between the
surfaces. This is located some 15 to 20 metres below the
bend stiffener, and becomes the hang-off point of the
power cables. From this point and up through the dynamic
bending zone, the cables are free to expand thermally,
and are exposed to bending only, and therefore the forces
upon the electrical cables become highly predictable.
Descending below the hang-off point, the power cables
are secured by frictional points in a special pattern over a
length of 2 metres at intervals of 50 metres to counteract
any long-term creep/deformation of the copper cores. This
system protects the electrical cables from strain hardening
and fatigue failure in the dynamic section.
Power cable design
The maximum current required by the pumps was in the
order of 225 A. Due to the length of 22.7 km of one of the
umbilicals and the consequent voltage drop, the supply at
the host was at 6 kV approximately to achieve a voltage
required by the motor terminals of 4.4 kV. These
requirements were obtained by using a cable with a
conductor cross-sectional area of 150 mm² and, whereas
an insulation rating of 6/10 kV would have been sufficient,
12/20 kV rating was used to achieve a much lower
electrical stress and consequently a greater safety factor.
A drawing of the power cable is shown in Figure 3.
The induced voltage from the neighbouring circuit also
causes a variation in the air gap torque when the power
frequencies are different. The impact of this was also
studied and it was concluded that the amplitude of the air
gap torques was maximum 2.8% and was acceptable
because it was within a frequency range that was not
critical for the operation.
Thermal analysis was also performed to visualize the
temperature distribution in the cross-section. One of the 3
circuits was spare and it was interesting to know the
temperature difference between the elements as an input
to the mechanical analysis.
Umbilical mechanical properties
One of the main challenges of designing a safe and
reliable high power dynamic umbilical for deep water
application is to control the forces and strain in the copper
cores of the cables. The material properties of copper
make it difficult to predict the forces over time as the
Fig. 3: Power cable drawing
Jicable’11 – 19 – 23 June 2011, Versailles - France
Close and Return
A.7.5
8th International Conference on Insulated Power Cables
Item
Description
1
Conductor
2
3
Conductor screen
XLPE Insulation
4
Insulation screen (non metallic)
5
Copper wires screen
6
Outer sheath
A.7.5
strength members were playing an essential role, limiting
mechanical working conditions of power cable, the
umbilical design was adapted according to the power
cable characteristics established through testing.
Table 2: Table of power cable components in Fig. 3
The conductor was a multi-stranded Class 2 type design
according to IEC 60228. While following the requirements
of the mentioned standard, several design parameters
can be adapted in order to achieve the required final
conductor mechanical properties. Hence, after a
prototyping and trial testing process, it was of great
importance to well define and control the material type
and the drawing, annealing, stranding and compacting
processes. Electrolytic Tough Pitch (ETP) was used and
every layer of wires was semi-compacted; in addition, no
single wire weld, due to possible factory logistic limitations
or breaks during production, was allowed during the whole
conductor lengths manufacturing process. The final
objective was to obtain proper conductor behaviour under
all expected loads taken during the installation process
and operational life, while also assuring the constant
properties of it in the whole manufactured length.
The insulation system was constituted of a
semiconductive
conductor
screen,
Cross-Linked
Polyethylene (XLPE) and a semiconductive insulation
screen. The three layers were extruded at the same time
by means of a triple extrusion head and the curing and
cooling processes were performed in a dry nitrogen
catenary line. Both semiconductive layer thicknesses
were increased compared with a standard MV cable
design to avoid any possible insulation damage caused by
a conductor or screen copper wire embedding into those
layers during the dynamic related movements.
The metallic screen was constituted of 60 copper wires
laid up onto the outer semiconductive layer with a total
cross-sectional area of 16 mm². This screen design was
chosen after several trials, mainly taking into
consideration the proper behaviour under bending fatigue
and the shear load capacity between the insulation
system and the outer jacket; this last property being
crucial in the overall transfer of the copper conductor load
through all cable layers to the external umbilical strength
members (carbon fibre rods). Other types of screens,
including a copper tape helicoidally applied, were tested
with unsuccessful results. A High Density Polyethylene
(HDPE) outer sheath was applied onto the metallic screen
to protect it mechanically as well as to minimize the water
diffusion to the insulation.
Subsequently, when power cable working condition
limitations were determined and imposed by the final
umbilical strength members design, the second important
goal was to verify the characteristic cable operation at
those expected critical working conditions (with an extra
safety margin) by means of mechanical testing followed
by electrical testing.
Testing sequences
To verify the electrical suitability of the power cable during
and after the application of an external pressure of 1.5
times the equivalent to the maximum installation water
depth, the following test sequence was made on one
cable sample:
•
Hyperbaric Test at 410 bars for 24 hours; applying
voltage on the sample while pressurized.
•
Partial Discharge Test according to IEC 60502-2.
•
Impulse Voltage Test according to IEC 60502-2.
•
Voltage Test according to IEC 60502-2.
•
Visual Inspection (dissection).
The power cable stress-strain curve under axial tensile
load was obtained for information by means of tensile
testing. Eventually, multiple samples from multiple
manufacturing batches were tested to find any variation in
batches, using the most conservative results to fine-tune
the umbilical calculations.
To verify the electrical suitability of the power cable after
the application of the maximum working tensile load (with
an extra safety margin), the following test sequence was
made on another cable sample:
•
Tensile Test up to 0.3% elongation.
•
Partial Discharge Test according to IEC 60502-2.
•
Impulse Voltage Test according to IEC 60502-2.
•
Voltage Test according to IEC 60502-2.
The power cable properties under axial compression load
were also measured.
QUALIFICATION TESTING
Power cable qualification testing
Due to the specific particularities of the project (deepness,
dynamic behaviour, sea water environment…), a specific
qualification testing plan was designed to simulate the
mechanical and electrical conditions of the application
during both operation and installation of the cable.
The first important goal of the defined testing sequences
was to collect the mechanical properties of the cable both
under axial tensile and compression loads. As umbilical
Fig. 3: Axial compression testing rig at TECNALIALabein
Jicable’11 – 19 – 23 June 2011, Versailles - France
Close and Return
A.7.5
8th International Conference on Insulated Power Cables
To verify the electrical suitability of the power cable after
the application of the maximum working compression
value (with an extra safety margin) experienced in some
cases due to temperature differences inside the umbilical,
the following test sequence was made:
A.7.5
•
Bend testing
•
Low cycle fatigue testing under tension to
simulate worst case installation loads (performed
on a similar cable splice for another power
umbilical project)
•
Axial compression test up to 0.2% compression
strain.
•
Electrical testing according to IEC 60502-2
•
Partial Discharge Test according to IEC 60502-2.
•
Dissection
•
Impulse Voltage Test according to IEC 60502-2.
•
Voltage Test according to IEC 60502-2.
To determine the power cable creep behaviour during the
operation of the dynamic umbilical length at both
maximum working axial tensile and compression loads
and at maximum expected working temperature (50ºC),
two creep tests were done with a total duration of 720
hours each. See the constructed testing rig in Figure 4.
Umbilical qualification testing
In addition to standard umbilical testing as defined in ISO
13628-5, a sample of the completed umbilical was
subjected to rigorous mechanical testing. The following
tests were performed:
Flex fatigue test
The purpose of the test was to verify that the umbilical
and its bend stiffener would provide sufficient flex fatigue
life for use in a production environment where the
umbilicals are suspended from a floating platform
An umbilical sample complete with topside termination
and bend stiffener and subject to operational loads
distributed to the various strength members (see figure 5)
survived the flex fatigue test program without mechanical
or electrical faults. It fully complied with the requirements
of Section 10.2.5 of ISO/FDIS Standard 13628-5: 2005.
No tension load was applied to the power cables, as the
hang-off system described above under “Umbilical
mechanical properties” avoids any tension in the cables in
the dynamic bending zone.
Fig. 4: Axial tensile and compression creep testing
rigs at TECNALIA-Labein
In addition, the power cable was subjected to a complete
Type Testing according to IEC 60502-2.
Splice qualification testing
Ensuring that the cable splices in the static section of the
umbilical would survive the installation forces at 2700m
water depth without unduly influencing the umbilical
design or function, and without imposing special
installation requirements was a particular challenge.
A proprietary splicing method was specially developed
and qualified for this application, meeting the goals above.
Qualification testing included:
•
Tensile testing at the load corresponding to the
force induced in the unspliced cable by the
global strain in the static section of the umbilical
during installation, plus a 50% safety factor
•
Hyperbaric testing under nominal voltage
Fig. 5: Flex fatigue test setup at TMT Laboratories
Pull test
The purpose of this proprietary test was to determine the
friction capacity of the “soft clamp” design of the umbilical.
Displacement of individual umbilical components including
the power cables was measured during pulling to verify
that the soft clamp system performed as expected.
Jicable’11 – 19 – 23 June 2011, Versailles - France
Close and Return
A.7.5
8th International Conference on Insulated Power Cables
A.7.5
SUMMARY
The unique power umbilical design described here
combines field proven technologies with new solutions
specifically developed in close cooperation with the
involved companies for ultra deepwater power cable
applications. This has resulted in the successful
qualification and manufacturing of the world’s first
umbilical system to deliver substantial power at 2700m
water depths. The key technologies enabling this
achievement as described here are:
•
Fig. 6: Pull test setup at TMT Laboratories showing
displacement transducers attached to umbilical
components
Additional testing was performed on a one meter sample
of umbilical, where the soft clamp was applied to all
elements and weight was applied to each element, pulling
against the plastic profiles (see Figure 7). This provided a
more thorough verification of the creep behavior in the
soft clamp system, showing that the extrapolated creep
over time was well within the acceptable range for the
system.
•
•
Carbon Fibre Rods
o
Limit global strain of the umbilical,
including the power cables
o
Overcome the limitations of steel
armouring in deep water, allowing for
potential dynamic applications down to
much greater water depths than
otherwise practical
o
Eliminate the need for armouring in the
individual power cables
“Soft Clamp” system
o
Provides a hang-off point for the power
cables below the dynamic bending
zone, allowing for free thermal
expansion of the cables in the hottest
area of the umbilical system, and
minimizing fatigue in the cables by
eliminating the axial load in the dynamic
bending zone
o
Counteracts the effects of creep on the
cables in the full dynamic section down
to the seabed by ensuring sufficient
frictional contact with the conduits
formed by the plastic profiles, effectively
transferring forces to the strength
members
Cable design
o
Careful design of the individual power
cables ensures sufficient shear load
capacity throughout the layers and into
the umbilical
o
Careful design of the conductor,
semiconductive layers and metallic
screen ensures proper behaviour under
bending fatigue
Figure 7: Creep test rig for soft clamp system
Electrical final acceptance testing (FAT)
The completed power umbilicals, one of 22,7km and the
other of 5,6km, each spooled to an outdoor carousel,
were subjected to DC resistance, TDR and HV testing as
part of the umbilical final acceptance testing.
•
Splicing
o
The specially developed static cable
splice allows for safe installation without
impacting the umbilical design or
placing limitations on the installation
parameters
Jicable’11 – 19 – 23 June 2011, Versailles - France