ISOPE Conference, Toulon, France, 23-28 May 2004.

Floating Windfarms for Shallow Offshore Sites

A.R. Henderson
Ceasa, Stroud, United Kingdom

M.B. Zaaijer
Section Wind Energy, Delft University of Technology, Delft, The Netherlands

B. Bulder, J. Pierik
ECN, Petten, The Neterlands

R. Huijsmans, M. van Hees

MARIN, Wageningen, The Netherlands

E. Snijders
Marine Structure Consultants, Schiedam, The Netherlands

G. H. Wijnants, M. J. Wolf
TNO Bouw (Building and Construction Research), Delft, The Netherlands,

support structure allows.

ABSTRACT The research project being reported on within this paper has focused on
the shallow seas of around 50 m water depth in the Dutch sector of the
Offshore wind energy appears to be on the verge of a phase of North Sea. For floating support structures, at these depths, additional
enormous expansion to becoming a significant source of electricity for challenges beyond those listed above include:
a number of countries in northern Europe. Two major offshore • achieving stability,
windfarms at Horns Rev and Nysted are now in operation and dozens • designing appropriate moorings (paradoxically it being more
further projects are in various stages of preparation with a significant difficult to moor a vessel in the shallowest waters).
number now having planning permission and firm construction dates. This paper provides an overview of a feasibility study, DrijfWind or
The projects built to date have been in shallow seas, of up to around FloatWind performed in the Netherlands, describing the concept
20m in depth, in the North and Baltic seas, with the planned projects generation, evaluation and selection process, reporting on ancillary
extending the range to new seas and greater depths. However, issues such as grid connection and O & M and providing a report of the
numerous challenges remain in the greater depths, in particular relating in-depth analysis of the concept selected to be most suitable for the
to the necessary size of the support structures, the resulting wave loads, conditions considered, this being a triple floater shown below, Fig. 1.
handling equipment and natural frequencies. The conclusions are that although, in this case, this technology may not
At one point, the inherent advantages associated with a floating support yet be ready for commercial application, the margin to economic
structure (of compliance due to the flexible attachment to the ground) viability is closing.
will match the additional costs due to complexity and novelty.
Questions of course remain under what conditions this will be (water
depth, sea climate, distance to shore) and whether offshore wind energy
can be proven to be economic at all under such greater challenges.
Significant technical and cost challenges will remain, in particular
regarding:
• minimising wave induced motion
• the additional complexity for the windturbine design process
• understanding the coupling between the support structure and
the windturbine
• the construction, installation and O & M procedures.
On the positive side, these challenges are accompanied by new
opportunities for systems and procedures that the use of a floating

Paper No. 2004-ARH-03 Henderson Page number

 ISOPE Conference, Toulon, France, 23-28 May 2004.
PREVIOUS FLOATING OFFSHORE WINDENERGY RESEARCH
& DEVELOPMENT PROJECTS
To date, a limited amount of effort has gone into developing and
evaluating various floating windfarm concepts, which is briefly
summarised below. Several very different concepts were developed
since the early 1990s, including:
Fig. 1: Selected Concept, the Triple-Floater
INTRODUCTION
Currently numerous plans for major offshore wind energy projects are
being pursued. Near the Dutch coast at Egmond, two farms are being
developed and this year is almost certain to see major windfarms being
built off the coastlines of several north European countries with
developments further a field, USA and Japan, likely to follow shortly
afterwards.
To date, most of the focus has been on fixed offshore wind turbines in
shallow water since it is expected that this technology will be the first
to become viable economically in the coming years.
The main reasons for developing fixed bottom-mounted windturbines
are:
• proven technology for fixing the structure to the ground
• easy connection of power cables from the turbine to the shore
• low induced motion due to waves, wind and current
Of course there are also some negative aspects:
• restricted to shallow waters
• removal is difficult/costly
• Expensive installation
Only a limited amount of research effort has been expended on floating
windenergy, which have generally concluded that, due to the
economics, such support structures are not feasible as yet. Hence the
focus of this study has been to investigate a range of floating concepts
for technical and economical viability.
The advantages of floating turbines are that they can operate in deeper
water and removal at end of life is more straight forward.
Undetermined aspects include:
• affect of motion of the unit due to current, waves and wind
• installation procedures
• general design issues (such as stability)
In this project, a framework for developing a floating offshore
windfarm was established and the technical and economical feasibility
assessed. Existing and proposed fixed wind parks were taken as a
reference. Aspects assessed include the floaters, electrical system,
installation and operation and maintenance.
Paper No. 2004-ARH-03 Henderson
• In the United States, William E. Heronemus led research
work at the University of Massachusetts at Amherst from the
early 70s on various floating concepts he termed WindShips;
his earliest concepts would typically support an array of
dozens of small wind-turbines; he continued this work though
into his retirement and his last concepts used a variety of
technologies to support a single or at most three turbines
Stoddard, W. (2003)
• In the United Kingdom, Garrad Hassan and Technomare co-
operated in the evaluation of a single turbine concept, located
on a spar-buoy and kept in position using eight-point catenary
moorings Tong, K. (1994). This was a fairly detailed study
and aspects such as type of wind-turbine (downwind, free-
yawing with very-high tip-speed), multiple vs. single turbine
structures, sharing of anchoring systems and tower design
(lattice type to reduce wind loads and overturning moment)
were investigated. The costs were estimated to be
inhibitatively expensive at around twice that of bottom-
mounted alternatives.
• Also in the United Kingdom, a group at University College
London investigated the possibilities of locating several
turbines on a single structure with the potential advantages of
reduced motion response and shared moorings (hence
reduced anchoring costs). This concept was developed in a
PhD Halfpenny, A., (1998) and an EPSRC research project,
in which the first author was responsible for the wind-turbine
and floating structure aspects Henderson, A. R., (2000), to
develop research tools and evaluate the idea in greater detail.
The main conclusions were that it would be excessively
expensive as well as difficult to construct structures capable
of withstanding the wave loads in regions with an attractive
wind resource.
• In Italy, a group in Milan investigated placing a single turbine
on to a toroidal-shaped float, positioned with tensioned
moorings. The complex shape was chosen to minimise wave
motion response but had the disadvantage of being difficult
and expensive to build Bertacchi, P. (1994 ).
• More recently, also in Italy, a proposal has been made to
locate electrical generating and desalination plant on a
floating pontoon to provide temporary supplies to island
communities during the holiday season Cesari, F. (1999).
This could possibly develop into a niche market for floating
wind energy.
• In the Netherlands, Lagerwey have investigated concepts for
floating wind parks, generally with multiple rotors placed in
double rows; Lagerwey did also contribute to this project.
• In Ireland/United Kingdom, co-operation between a
windfarm developer, Sure Engineering, and an offshore
engineering consultant, Ocean Resource, has resulted in a
couple of innovative support structure for offshore
windturbines, including a tensioned quadruple-floater with a
telescopic tower Hannevig, D., (2003)
• In Japan, the JOIA (Japanese Ocean Industries Association)
is co-ordinating a group of interested parties to develop
promising concepts for floating wind energy in that country;
the first phases were completed in 2002 and further work
2
 ISOPE Conference, Toulon, France, 23-28 May 2004.

continues. This work is described within a separate paper at

Met-Ocean conditions Typical for Dutch North Sea
this conference Ushiyama, I. (2004) and the overall project
objective is to develop a prototype by the end of the decade. Soil conditions sand
• In Germany, several groups are proposing novel floating
concepts for the relatively shallow waters of the German Real Interest rate = 5%
sectors of the North and Baltic Seas; in these shallow depths, Economic parameters (inflation rate = 0%)
a cheap and workable mooring system will be critical and economic lifetime = 20 years
solutions proposed include using novel applications of the
proven tensioned moorings as well as mooring the floater to a 5-10 m
tower piled into the seabed. 10-15 m
Patents are another source of information, describing potentially viable 15-20 m
concepts, inspiration for concept generation and indication of the 20-25 m
general level of activity. Patent activity for floating offshore 25-30 m
30-35 m
windenergy concepts has increased recently to a level of several patents 35-40 m
each year. It should be noted that unfortunately the majority of 40-45 m
patented ideas are impractical and indicate a lack of knowledge of the 45-50 m
fields of either windenergy or offshore engineering or sometimes both.
So far, the furthest that any concept has been developed ahs been to
scaled tests in a wave-tank, of which there have probably been more
than half a dozen. The next stage of testing a full-size prototype will be
exceedingly expensive but necessary since windturbine aerodynamics
do not scale well.

TERMS OF REFERENCE

The reference design will be a large, approximately 500 MW, offshore

wind power plant as described in Table I.
Table I: Terms of Reference

Location North Sea

Water depth
Distance to shore
Weibull wind speed
parameters at 10 m height
Wind shear profile
Turbulence (IEC)
Wind farm turbine spacing
Wind farm array efficiency
Turbine description
Floater/Submersible
yawing
1
Minimum height determined by rotor radius, maximum wave height
and splash
around 50 m, see Fig. 2; this refers
to the relatively small deepest part
of the Dutch sector of the North Sea
but is chosen because for floating
concepts, shallow waters are
technically (and hence Fig. 2: Bathymetry in Dutch Sector of North Sea
economically) more challenging. The concepts were assessed on the basis both of estimate cost and on
potential of reducing this cost. The cost were determined according to
between 50 km and 200 km
the Levelised Production Cost method defined in Hunter R., (1994).
Levelised means that no variations in cost or energy yield are assumed
Vave = 9 m/s / k = 1.8 during the lifetime of the project.
The simplified method will be used, which means that the following
from roughness height of 0.1 mm equation needs to be evaluated
LPC = I /( a ⋅ AUE ) + TOM / AUE
I15 = 0.12 / a = 3
In which:
Approx. 8 Diameters apart. I Initial investment;
a annuity factor, depending on discount rate and economic
95% lifetime ;
AUE Annual utilised energy;
Rated Power = 5 MW TOM Total Levelised annual “downline cost”, i.e. Operations and
Diameter = 115 m maintenance, insurance, retrofit cost, and salvage cost.
Hub Height = >80 m1 This results in a yearly capital cost and operating and maintenance cost
3 blades divided by the net energy production minus electrical and aerodynamic
Direct Drive generator losses within the wind farm. To determine the cost of energy, it is
necessary to determine the following quantities:
single wind turbine
3-5 wind turbines • Energy yield, determined on the basis of the power curve,
wind conditions, wind turbine availability, wind farm losses,
nacelle, not the entire windturbine electricity losses in the wind farm and between wind farm
and grid connection;
• Total investment cost, i.e. cost of the wind turbines, floaters,
installation, electrical infrastructure in the wind farm and
between wind farm and grid;

Paper No. 2004-ARH-03 Henderson 3

 ISOPE Conference, Toulon, France, 23-28 May 2004.

• Operating and maintenance cost, including insurance;
• Economic parameters, such as interest and depreciation
period.
CONCEPTS GENERATION AND INITIAL EVALUATION
equally massive stability moments (which are already large with the
single turbine). Therefore, if more than one turbine is to be installed, it
is probably confined to two machines in a T-shaped arrangement, as
outlined in Fig. 4.

The concept generation process started by identifying critical choices,

which are described immediately below. From this, a large number of
candidate windturbine support-structure concepts were proposed and an
initial evaluation was undertaken, including the development of
simplified models within MARIN’s knowledge-based computer
program QUAESTOR, Van Hees, M.Th (1997).

Weather vaning or “fixed” floater ?

Should the floater be free to yaw so that the whole structure aligns itself
into the wind (“weather-vaning”) or should the floater not be allowed to
yaw e.g. by a spread mooring, Fig. 3? The latter implies the use of a
yaw mechanism at the nacelle, which is the usual solution for onland
and bottom-mounted offshore windturbines of course.
Fig. 3: Type of Yaw Mechanism
A weather-vaning floater would require no yaw-mechanism at the
nacelle however the cost savings for the windturbine would only be
marginal and it would be necessary to consider how the floater would
respond when the wind, waves and current were not in the same
direction and whether it would be possible to align the rotors in to the
wind. This might be possible by slightly yawing on off the turbines.
Another problem with the weather-vaning concept is the delivery of
generated power to the grid; a rotating- and most probably water-tight
connector is required. These connectors exist in the offshore industry
but are complex and expensive.
A vessel positioned using a spread mooring will experience a mean
yaw force due to the wind, waves and the current, which will be
compensated by the yaw restoring of the mooring system. Therefore the
position quality with respect to the wind will be better enforced using a
spread mooring system. Based on these considerations, it was decided
not to adopt the weather-vaning concept in this study.
Fig. 4: Multi-turbines Floater
The tower top mass of the twin turbine will be about three times that of
the single turbine due to the presence of the traverse. This requires a
larger floater, simply to deal with the increased wind moment and
vertical centre of gravity. The yaw mechanism should either be capable
to deal with one machine shut down or both machines should be shut
down in the event of a malfunction of one of the turbines, being the
most probable solution, hence reducing the availability. Tentative
calculations on the single floater concept with one or two turbines
indicate that a floater with two turbines contains about 170% of the
steel of that of a floater with a single turbine which makes this concept
unpromising.
For the purpose of the DRIJFWIND study, the single turbine solution
will be taken as the starting point. The twin turbine (“T”) arrangement
can be designed and studied as a separate system and can in principle
be fitted on each of the following floater concepts.
Floater concepts
The section identifies and describes briefly the parametric floater
concepts investigated within this project.
The single cylindrical floater (“pill-box”) or buoy.
The floater is a simple vertical cylinder, held in position by a spread
mooring, Fig. 5. This concept was the starting point in the discussions.
Tentative calculations were performed to establish the basic dimensions
of the “pill-box”. The stability range requires ballast water to achieve
sufficient draft. Initial stability requires a diameter of at least 37 m. In
the above results, about 3100 t of water-ballast is used to achieve a
draft of 4.27 m. This can either be stored in the pill box but this will
require a lot of additional structure to prevent free surface stability loss.

One or more turbine per floater?

In view of overall reliability and from a maintenance perspective, it is

attractive to apply the largest turbine which can be constructed on the
basis of currently or soon-to-be available technology. In the Terms of
Reference, a 115 m turbine with a rated power of 5 MW was
considered to be a feasible size and a 500 MW wind farm would consist
of 100 such units. A major conceptual decision is related to the number
of turbines to be installed on a single floater. Taking the diameter of the
turbine into account (115 m) positioning turbines above each other
would imply a tower height of about 200 m, with a massive weight and

Paper No. 2004-ARH-03 Henderson 4

 ISOPE Conference, Toulon, France, 23-28 May 2004.
Fig. 5: Cylindrical Floater
Single circular floater with skirts
This circular skirt will confine about 3000 tonnes of sea water and can
be considered as a ballast tank without bottom, Fig. 6. From a stability
perspective, a completely filled ballast tank can be regarded as a
flooded space. Although feasible from a stability perspective, this
concept is not feasible from a motion perspective; in particular the
heave period (Tz) of about 9 seconds is right within the high energy
range of the wave spectrum and the roll period at about 13 seconds will
be strongly affected as well. Both roll and heave period should be
longer than about 16 seconds and it is not possible to achieve that with
the single circular floater, i.e. it is not feasible to fulfill stability and
motion requirements at the same time. Therefore, the “pill-box”
concept was concluded to be technically unfeasible.
Fig. 6: Floater with Skirt
with a tension leg instead of a spread mooring
In order to fulfill stability requirements using a smaller diameter
floater, pretension can be introduce by means of tension leg (or
mooring), Fig. 7. The tension leg increases the vertical stiffness of the
floating system, which reduces the heave period. In this way, the heave
period can be moved out of the high energy region of the spectrum.
From a static stability point of view, this pretension can be considered
as a point mass located at the connection point of the tension leg. In
addition to the resulting downward shift of the virtual centre of gravity,
the centre of buoyancy is also moved downward in absolute sense since
Paper No. 2004-ARH-03
additional buoyancy is required to compensate for the pretension.
The use of pretension in shallow water offers reduced benefits
compared with the deeper waters where it is more commonly used
since only limited improvements in the stability can be achieved. As
noted above, a diameter of approximately 37 m was required to fulfill
the basic stability requirements when catenary moorings are used.
Increasing the ballast does not being significant benefits in such
shallow waters, hence smaller diameters are only possible with the use
of tension moorings. The pretension should be sufficient to avoid the
tension leg becoming slack in extreme wave conditions; this can only
be determined with some real accuracy by means of thorough motion
analyses. Initial calculations show that a large pretension is required, in
the order of 3,000 tonnes; the 3,000 tonnes water ballast and about
2,200 tonnes of steel results in a total displacement about 8,500 tonnes.
The diameter of the larger (upper) cylinder is in the range of 26-30 m
while the small lower cylinder has a diameter of about 16 m (around 60
per cent of the floater diameter). The total floater height is about 30 m
with the draft being in the order 25 m. These values can hardly be
considered as a feasible solution in terms of investment cost and
complexity for supporting a single 115 m turbine in waters up to 50 m
deep.
The question of whether tensioned moorings can inherently be
developed cheaply enough was also raised. Tensioned moorings have
only been used in a limited number of cases to date and even the
cheapest of these, the SeaStar concept, is a couple of orders of
magnitude more expensive than target costs here, albeit with
significantly larger payload requirements (orders of 6,000 tonnes versus
400 tonnes though at a lower height).
Fig. 7: Pre-tensioned Moorings
Similar to previous but with a box-shaped floater, i.e. a square or
rectangular barge.
This was summarily evaluated since it is expected to be similar to the
circular single floater described immediately above.
‘Catamaran’ type of floater with truces connecting the floaters
The floaters are prismatic and the truces are cylindrical, a spread
mooring is applied. Again not evaluated in detail, since it was felt it
was unlikely to be economical.
Henderson 5
 ISOPE Conference, Toulon, France, 23-28 May 2004.

‘Spar’ floater
This floater is a so-called ‘spar’ buoy with a large lower vertical
cylinder referred as the ‘disc’ and a smaller upper cylinder protruding
the water surface on which a single tower is located. The buoy is held
in position by a spread mooring, Fig. 8. In terms of initial stability, the
Spar as outlined in the above sketch is not feasible in water-depths
around 50 m due to its enormous size necessary to achieve sufficient
static stability.

Fig. 9: Triple Floater

As 0but with a single turbine located on one of the floaters

This concept was proposed and presented by MSC on the basis of the
initial calculations performed under 0. Re-assessment of this concept
showed that a lighter construction could be achieved by returning to
option 0since it allows lighter truces connecting the three floaters.

Triple floater with 5 turbines of 71 m.

This concept, Fig. 10, was developed by Lagerwey and Heerema to the
point of a preliminary construction design, which could be compared
with the independent analysis undertaken here. The weight of the
superstructure is 800 tonnes and of the five turbines together is
500 tonnes. This concept is moored by means of a single steel pile in
the centre of the triangle formed by the three floaters. The floater
should be weather-vaning, i.e. the floater should keep the turbines in
the wind due to the resulting turning moment of the wind force. An
Fig. 8: Spar obvious disadvantage of this concept is its inherent vulnerability; if one
out of five turbines needs to be shut down for maintenance or due to a
Triple floater concept with truces connecting the floaters and a single malfunction, the weather-vaning capability is lost which implies that
turbine located in the centre between the floaters. the other four have to be shut down too.
In order to improve the vertical motion response and to reduce overall
construction volume, the triple floater concept was proposed, see Fig. 9
for base and the figure in the Abstract above for the full structure. The
floater consists basically of a centre column carrying the wind turbine
which is connected with cylindrical floaters by means of tubes or
truces. Tentative relations were derived for the hydrostatic properties,
stability and weight on the basis of a limited number of describing
parameters. A concept variation was performed for a range of floater
distances. Floater dimensions are established on the basis of stability
requirements. Stability requires particular values of GM which can
only be fulfilled by a particular minimum distance between the floater
bodies.
The results indicate that the triple floater concept requires less steel
than the single floater/spar floater concepts. However, the vertical
motion response is still within a critical region and should be shifted
either to higher frequencies (only possible by introducing pretension) or
to lower frequencies in the order of 15-16 seconds. This can be done
e.g. by fitting large circular plates or cylinders underneath the floaters,
Fig. 10: Multi-Turbine Floater
increasing the (hydrodynamic) mass of the floater as indicated in Fig. 9.
This concept was selected as showing the greatest promise and
therefore was investigated in greater detail in the following sections. Quadruple floater concept with truces connecting the floaters
This concept, Fig. 11, is very similar to the triple floater. With equal
floater dimensions, the distance between the floaters can be somewhat
smaller. The steel weight of the quadruple floater is expected to be
higher due to the larger amount of connecting structure between the
four floaters.

Paper No. 2004-ARH-03 Henderson 6

 ISOPE Conference, Toulon, France, 23-28 May 2004.

INTEGRAL MAINTENANCE COST ESTIMATE FOR REMOTE

OFFSHORE WINDFARMS

An issue of interest in this study, is to what extent it is profitable to

perform “on site” maintenance in comparison with “on shore”
maintenance for which the floating platform needs to be shipped. Both
strategies are dependent on weather windows but it was calculated that
the case of the towing of the platform is more susceptible and hence
that “on site” maintenance is preferable for practically all failure
mechanisms.
Specific “on shore” activities such as recovering of the platform or
clustered activities within a “substantial overhaul” have been assumed
to be unnecessary due to a maintenance free platform and the use of
reliable components. The cost calculations assume the availability of
exchange parts, the costs of which are managed by using renewed cost-
intensive components that have failed.
Efficiency measures such as opportunity based maintenance or
implementation of clustered corrective maintenance actions, have not
been incorporated in the model since the failure rates are limited. This
factor therefore determines the maintenance costs only to a limited
portion of the accuracy of estimation.
Uncertainties with respect to the maintenance demand, resulting from
Fig. 11: Quadruple Floater the fact that no detailed design is present, are to be controlled by
incorporating a RAM (Reliability, Availability and Maintenance)
specification and assessment within the design phase of the final
The jackup platform on three or four legs construction. In a RAM assessment the final design is evaluated with
The jack-up concept was proposed as an option to allow simple
respect to its maintainability (with function loss during a specific time)
installation and convenient transportation to and from the wind farm. A
and the resulting availability (capability to produce), by using the
jack-up concept eliminates wave induced motions of the turbine and
reliability performance data of the specific components.
forms a stable foundation of a single or multiple turbine system.
The reliability data that are applicable for supposedly “maintenance
However, the jack-up concept has a major drawback: its cost.
free” components in order to safeguard the assumptions made within
According to data provided by MSC, a jackup suitable to carry a single
this study, are determined by a failure rate of ultimately 4×10-4 (yr-1).
115 m turbine will cost about € 12,000,000, an economically unviable
This guideline in combination with availability criteria is applicable
figure.
during the actual design phase.
The maintenance costs for a platform are estimated to be 2.2 % of the
ELECTRICAL SYSTEM, CONCEPTUAL DESIGN AND COSTS
investment costs (offshore position: 100 km). This is equivalent to
An important aspect in the determination of the feasibility of an approximately 35 % of the levelised production cost .
offshore wind farm is the choice of the electrical system, necessary to
LEVELISED PRODUCTION COST TRI-FLOATER WIND FARM
collect the power in the farm and transport it to shore. Based on the
results of ECN’s ERAO project, the two most promising system types The simplified method for the levelised production cost was applied to
for DrijfWind have been chosen: individual variable speed and park
calculate the economics, with variation in the distance to shore, the
variable speed. For these options, two park layouts based on platforms electrical system and the area of production of the floater, Table II.
with 1 and 5 turbines were investigated. These layouts correspond to
different cable layouts inside the park: string and star. The second Table II: Economics
parameter investigated is the distance between the wind farm and the
Distance to Coast 200 km 100 km 200 km 100 km
shore. The EEFARM computer program is used to calculate the
electrical and economic performance of the electrical architectures and Grid Option Pv Iv Pv iv
layouts. A single EEFARM run gives the load flow (voltages, currents, Construction Site Europe Asia
active and reactive powers) in all system nodes as well as the electrical Costs floater & install. € 4,500k € 4,500k € 3,500k € 3,500k
losses for all wind speed bins. EEFARM also estimates the contribution Mooring costs € 2,500k € 2,500k € 2,500k € 2,500k
of the electrical system to the kWh price, averaged over the life time of Turbine costs € 2,875k € 2,875k € 2,875k € 2,875k
the wind farm. The economic evaluation is based on budget prices for Electr. Infrastr. costs € 3,710k € 2,710k € 3,710k € 2,710k
the electrical components, received from manufacturers, and Capital Investment € 13,585k € 12,585k € 12,585k € 11,585k
aerodynamic performance of the wind farm calculated by
maintenance / year € 299k € 277k € 299k € 277k
FYNDFARM.
Insurance (1% of investment) € 135.85k € 125.85k € 125.85k € 115.85k
Based on economics only, the best choice for the DrijfWind 500 MW
wind farm will be the Individual Variable (iv) speed system for “downline cost” € 434.85k € 402.85k € 424.85k € 392.85k
distances below 140 km and the Park Variable (pv) speed system for Gain Wh gross 24.6 G 24.6 G 24.6 G 24.6 G
distance above 140 km. Differences in controllability and stability of transmission efficiency 88.5% 88.3% 88.5% 88.3%
the two options may influence the choice, but has not been investigated. Net Yield in Wh 20.7 G 20.6 G 20.7 G 20.6 G
annuity factor 12.462 12.462 12.462 12.462
Levelized Production
€ 0.074 € 0.068 € 0.069 € 0.064
Cost (kWh-1)

Paper No. 2004-ARH-03 Henderson 7

 ISOPE Conference, Toulon, France, 23-28 May 2004.
CONCLUSIONS
With the aid of the literature study the design criteria, boundary
conditions, references etc. for floating offshore wind turbines were
formulated. These parameters and the associated values were added or
adjusted as necessary through the duration of the project.
By means of brainstorm sessions with all partners, a number of
concepts for floating offshore wind turbines have been proposed. For
some of the concepts, primary dimensions were determined using the
knowledge-based system Quaestor. This computer program allows
relationships between weight, costs, dimensions, stability etc. to be
entered and uses these to search for an optimum solution.
Of the concepts evaluated in the early part of the project, the tri-floater
appears to be statically and dynamically stable and offers the greatest
promise for the types of sites considered and hence was analysed
further. Motion response calculations were undertaken followed by a
more thorough analysis on the strength and costs of production and
installation. A preliminary design of the mooring system was also
completed.
An electrical system analysis was made for a 500 MW wind farm
consisting of 100 turbines and several permutations of transmission
technologies and layouts were looked in to. Up to a distance of 140 km
from the coast, the individual variable speed system using 150 V AC
cabling seems to be the cheapest option. For distances to shore greater
than 140 km, a park variable speed system using 141 kV DC cable
connection appears to be the cheapest option.
An analysis has been made of the integral maintenance cost of an
offshore wind farm. By means of component failure rates, repair time
‘weather windows’, choice of transport equipment etc. the optimum
maintenance strategy has been defined. When using lightning
protection, the maintenance costs per year for one turbine are €277,000
for a site 100 km from the coast. Due to component failure and
maintenance, the wind turbine non-availability is expected to be around
33 days per year, which results in an availability of 91%.
The levelised production costs for a windfarm 200 km of the coast and
built in Asia is estimated to be 0.069 €/kWh, when built in Europe
0.074 €/kWh
The levelised production costs for a windfarm 100 km of the coast built
in Asia is estimated to be 0.064 €/kWh, when built in Europe 0.068
€/kWh
RECOMMENDATIONS
The tri-floater has been designed for water depths of 50 m and more.
However, it could also be used in water depths of 40-45 m. This
increases the area of the Netherlands continental shelf, which can be
used for floating offshore wind energy, to at least 14%, Fig. 2.
There remain uncertainties and hence improvements that could be made
to the analysis and evaluation methods utilized here.
Paper No. 2004-ARH-03
For the choice of the electrical system, a second major aspect is the
controllability and behaviour with respect to the (high voltage) grid.
This should be done for a final decision.
It is recommended to use a RAM-spec (Reliability, Availability,
Maintainability) during the design phase to reduces the maintenance
costs and hence the overall costs of ownership. Reducing the
maintenance costs can be achieved in the fastest way by reducing the
failure rate of those processes that appear to contribute heavily due to
the characteristics of the repair scenario (repair time, delay due to
weather window and repair time needed).
Acknowledgements
The work reported here forms part of the project Feasibility Study for
Floating Offshore Windenergy (DrijfWind), which was partly funded
by NOVEM within the TWIN-2 program under contract 224.721-0003.
The project was undertaken by five partners based in the Netherlands,
one industrial partner, Lagerwey and four research partners. TNO,
MARIN, ECN and Delft Technical University. In addition Marine
Structure Consultants (MSC), an offshore engineering consultancy was
requested to perform a preliminary design.
References
Bertacchi, P. (1994 ), Di Monaco, A., de Gerloni. M., Ferranti, G., ().
Eolomar - a moored platform for wind turbines; Wind Engineering
Vol. 18, No. 4, p189
Cesari, F. (1999) and Gaudiosi, G., Wind turbine on floating platforms
for desalination plants, Proceedings of the European Wind Energy
Conference, Nice,
Halfpenny, A., (1998), Dynamic analysis of both on- and off-shore wind-
turbines in the frequency domain; PhD Thesis, University College
London,
Hannevig, D., (2003) & Bone, D., A Semi-submersible Tension Leg
Platform, DOE/NREL Workshop on Deep Water Offshore Wind
Energy Systems, Washington, USA, 15-16 October
Henderson, A. R., (2000), Analysis Tools for Large Floating Offshore
Wind Farms, PhD Thesis, University College London.
Hunter R., (1994), J.O. Tande: “Recommended practices for wind turbine
testing and evaluation” # 2. Estimation of cost of energy from wind
energy conversion systems. International Energy Agency, 2 edition,.
Stoddard, W. (2003) & Heronemus-Pate, M., Windship Concepts,
DOE/NREL Workshop on Deep Water Offshore Wind Energy
Systems, Washington, USA, 15-16 October
Tong, K. (1994), Technical and economical aspects of a floating offshore
windfarm, Proceedings of the Offshore Wind Energy in the
Mediterranean and other European Seas {OWEMES} Seminar, Rome.
Ushiyama, I. (2004) Seki, K., & Miura, H.,, Feasibility Study on Floating
Offshore Wind Farms in Japanese Waters, ISOPE Conference 2004,
(submitted).
Van Hees, M.Th (1997): “QUAESTOR: Expert Governed Parametric
Model Assembling”, Doctors thesis, Technical University Delft,
ISBN: 90-75757-04-2
Henderson 8