Renewable Energy 36 (2011) 2831e2837

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Multi-criteria assessment of offshore wind turbine support structures

E. Lozano-Minguez a, *, A.J. Kolios a, F.P. Brennan b
a
Department of Offshore, Process & Energy Engineering, Room 137, Whittle Building (B52), Cranfield University, Bedfordshire MK43 0AL, United Kingdom
b
Department of Offshore, Process & Energy Engineering, Room 124, Whittle Building (B52), Cranfield University, Bedfordshire MK43 0AL, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history: Wind power, especially offshore, is considered one of the most promising sources of ‘clean’ energy
Received 10 March 2011 towards meeting the EU and UK targets for 2020 and 2050. Deployment of wind turbines in constantly
Accepted 17 April 2011 increasing water depths has raised the issue of the appropriate selection of the most suitable support
Available online 6 May 2011
structures’ options. Based on experience and technology from the offshore oil and gas industry, several
different configurations have been proposed for different operational conditions. This paper presents
Keywords:
a methodology for the systematic assessment of the selection of the most preferable, among the different
TOPSIS method
configurations, support structures for offshore wind turbines, taking into consideration several attributes
Multi-criteria decision making
Jacket
through the widely used multi-criteria decision making method TOPSIS (Technique for Order Preference
Monopile by Similarity to Ideal Solution) for the benchmarking of those candidate options. An application
Tripod offshore support structure comparing a monopile, a tripod and a jacket, for a reference 5.5 MW wind turbine and a reference depth
Offshore wind turbines of 40 m, considering multiple engineering, economical and environmental attributes, will illustrate the
effectiveness of the proposed methodology.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction constraints of high construction cost, especially foundation and

electrical connection with the shore, and limitations in operation
Wind energy has been used from ancient times, initially trans- and maintenance caused by limited access [1].
forming wind energy into mechanical work for operating windmills The depletion of fossil fuel reserves, the various crises of fossil
and wind pumps, and later employed to generate electricity. From fuel prices and its great contribution to climate change have forced
the first application of wind turbines at the end of the 19th century several countries in the EU to turn towards alternative forms of
[1], the world’s first onshore wind turbine was installed in 1980 [2], sustainable energy. In 2007, European leaders agreed that the EU
standing as the vanguard for the significant development which will reach a 20% share of energy from renewable sources by 2020,
followed in the field. The UK’s first commercial wind farm was built by setting individual targets for all Member States [10]. The UK aims
in 1991 at Delabole [3], with a total capacity of 4 MW, establishing to obtain 15% of its final energy consumption from renewable
wind energy as a safe, clean source of energy on the UK’s energy map. sources by 2020 and to cut CO2 emissions by a minimum of 26% by
Further developments in research throughout Europe, has 2020 and 60% by 2050 [11], having the best geographically varied
moved interest in wind energy offshore [4], taking advantage of the wind resources in Europe [12]. In order for those targets to be
unrestricted space, lower social impact and higher wind resource achieved, it is an essential requirement to make this industry more
conditions [5]. In Ref. [6], it is estimated that an additional 50% of economically efficient through the optimization of components
electricity can be generated by the same turbine in an offshore such as offshore wind turbine support structures. This paper aims
wind environment, compared to onshore. Although, the UK became to provide an analytical methodology for the selection of the most
involved later than other European countries in the research and preferable from the three most commonly used support structure
testing of offshore wind turbines, starting with the Blyth wind farm configurations e monopile, tripod and jacket e for a typical 5.5 MW
[7], it is currently the leader in this sector with 1 GW of offshore wind turbine in 40 m water depth. In this analysis; engineering,
wind farms [8], overtaking Denmark in 2008 [9] and representing economics and environmental assessment will be considered in
44% of Europe’s total installed capacity. Although transferring wind order to balance the socio-economic activities of the sustainable
energy technology offshore presents several benefits, it faces the energy sector [13]. The widely used multi-criteria decision making
method TOPSIS (Technique for Order Preference by Similarity to
Ideal Solution) will be applied, allowing consideration of both
* Corresponding author. Tel.: þ44 1234756162, þ44 7907774965 (mobile).
E-mail addresses: e.lozanominguez@cranfield.ac.uk, estivaliz@hotmail.com quantitative and qualitative criteria, in order to incorporate
(E. Lozano-Minguez). multiple attributes into the decision making process.

0960-1481/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.renene.2011.04.020
2832 E. Lozano-Minguez et al. / Renewable Energy 36 (2011) 2831e2837

Table 2
Soil characteristics.

Submerged Dense Sand

Young’s modulus (E) 75 MPa
Poisson’s coefficient (n) 0.3
Horizontal subgrade reaction (nh) 10 MN/m3

3. Design attributes

In the previous section, the design of the support structures was

discussed and the maximum displacement and von Mises stresses
were derived. This section will consider multiple attributes
regarding environmental and economical assessments of the
different configurations.

3.1. Environmental impact assessment

Fig. 1. Monopile, tripod and jacket support structures.

The assessment of the possible impact, negative or positive, that

2. Design and optimization of WIND turbine support
a project or a policy might have on the environment is known as
structures
environmental impact assessment (EIA). The European directive on
EIA, which was transformed into a legal obligation among Euro-
Within the scope of this study, the three most common
pean countries, requires an EIA study for offshore structures as part
configurations of offshore support structures were studied [14]:
of the licensing process [18]. Offshore wind farms have both
harmful and beneficial environmental consequences for the marine
 Monopile is a simple design, the foundation of which consists
environment. The main negative effects are: impact on birds,
of a tubular structure that extends into the seabed; it is used for
underwater noise and vibration, water turbidity, and electromag-
installations at water depths of up to 25 m.
netic interference. The main positive effect is that the foundations
 Tripod is a three-legged structure made of cylindrical steel tubes,
may act as artificial reefs. The degree of impact on biodiversity and
and is used for installations at water depths between 25 and 50 m.
ecosystem functions will depend on the chosen location [19]. The
 Jacket is a (usually) four-legged structure made of cylindrical
aforementioned are local impacts, whereas the global impact can
steel tubes and, as with the tripod, is used for installations at
be identified by distinguishing the fabrication process of wind
water depths between 25 and 50 m.
turbines and support structures as negative impacts, whereas the
low carbon emissions over the turbine’s life cycle, as well as
Fig. 1 presents a conceptual illustration of the different
negligible emissions of nitrous oxides, mercury and sulphur oxides,
configurations.
relative to conventional sources of energy are regarded as positive
The design of offshore structures is based on a combination of
global impacts [20]. In this section some of the most important
the Finite Element Method (FEM) and the provisions of design
impacts will be presented.
standards such as [15] and [16]. The methodology that is followed
involves the construction of an initial finite element model of the
3.1.1. Carbon footprint
structure, which takes into account the soil structure interaction
There seems to be some confusion about the definition of carbon
and the loads acting on the structure in two different case studies e
footprint [21]; the spectrum of definitions ranges from direct CO2
maximum operation and survival conditions. The basis for the
emissions to full life-cycle greenhouse gas emissions, with different
design is selected as Eurocode 3 for quasi-static analysis, generating
considerations of units of measurement. For this analysis, only CO2
a global safety factor based on its provisions for load combinations
emissions related to the production of steel from extraction,
and material properties. Design optimization took place through an
transportation, processing, manufacturing and construction will be
iterative process in order to efficiently utilize material properties.
considered. The amount of CO2 equivalent (CO2e) emissions per kg
For the design of all three cases, the same tower was considered
of steel produced will be calculated by the emissions of N2O, CH4
based on [17]; 90 m length, 6 m base diameter and 0.02 m thick-
and CO by the following empirical formula [22]:
ness, and 3.87 m top diameter and 0.02 m thickness. Tables 1e3
present the material, soil and environmental properties that have CO2 e ¼ 270  N2 O þ 24:5  CH4 þ 1:4  CO
been considered for a hypothetical site of deployment.
The structural models have been built with the Abaqus/CAE, For steel members, the unit emissions per each kg of production
which is a powerful engineering software tool based on the FEM. for N2O, CH4 and CO are correspondingly 0.07, 0.04 and 0.93 g. Based
After several iterations, dimensions of structural members were
Table 3
optimized; the results for each case are presented in Table 4e6.
Environmental conditions.
Table 7 presents the maximum displacement and maximum von
Mises stress to which the different support structures are subjected. Case 1: Extreme Case 2: Operating
wind conditions conditions

Table 1 Wave height (H) 10.6 m 10.6 m

Material properties. Wave period (T) 9.6 s 9.6 s
Wind velocity at 10 m 8 m/s 8 m/s
Young’s modulus (E) 210 GPa above the sea level [W10]
Shear modulus (G) 80.8 GPa Gust wind speed at 10 m 35 m/s 12 m/s
Density (r) 8,500 kg/m3 above Standing Water Level
Yield stress (sy) 355 MPa [UG(10)]
Poisson’s coefficient (n) 0.3 Aerodynamic Loads No Yes
 E. Lozano-Minguez et al. / Renewable Energy 36 (2011) 2831e2837 2833

Table 4 Table 6
Optimized monopile design. Optimized jacket design.

Length (m) Diameter (m) Thickness (m) Length (m) Diameter (m) Thickness (m)
Tower Top 87.6 4.1 0.013 Tower 87.6 5.1e7.02 0.02e0.028
Tower Base 87.6 6 0.021 Braces Hor. 1 16.4e24.8 2.1 0.016
Transition Piece 10 7 0.04 Brace Diagonal 1 22.69e27.30 2.1 0.016
Pile 35 þ 40 7 0.04 Braces Hor. Long 1 23.19e42.42 2.1 0.016
Piles 33.3 5 0.04

on the results of the analysis given in Section 2, and taking into

consideration the derived masses, Table 8 summarizes the emis- 3.1.5. Electromagnetic fields
sions of each different configuration, illustrating that the jacket Many fish species within UK waters are sensitive to magnetic
structure configuration has the greatest amount of CO2e emissions. and electric fields which can be caused by buried underwater
cables. Fish uses their perception of magnetic and electric fields for
3.1.2. Noise and vibration orientation and prey detection [20]. However, it is not yet known
Wind turbines can produce infrasonic vibrations which derive whether the fish will suffer any consequences caused by this
from the complex interaction of mechanical factors associated with interaction [26]. The choice of foundation will not, therefore, be
machinery, rotation bearings, effects of wind, etc. In several cases it considered as affecting the impact.
may be found impossible to predict them or their effect on the
natural environment [23]. There are plenty of sources of under- 3.1.6. Impact on birds
water noise including wind farm related geophysical surveys, pile Several studies exist which prove an increase in bird mortality
driving, foundation installation, drilling, cable trenching, rock due to the risk of colliding with wind turbines. However, the rate of
laying, wind turbine operation, vessels and machinery, turbine mortality is variable and relatively low, from 0.01 to 23 mortalities
structure installation, decommissioning, etc [24]. There is no per turbine per year [27]. Other studies, such as the flight of eider
identified potential for an adverse noise impact during the opera- duck at Utgrunden off Sweden [1], prove that some birds under-
tional phase of the offshore wind farm [25]. stand the presence of the wind turbines, even during night time, and
Measurements of underwater noise, such as the studies of the accordingly change their route. Therefore, the effects on bird pop-
sea mammals of the Bockstigen and Tuno Knob wind farms, off ulations can be summarised as: avian mortality through collision,
Sweden and Denmark respectively, have shown that offshore wind physical change of habitat, and changes in migratory flight paths.
farms do not have a major impact on marine mammals [1]. During The birds with a greater risk of collision are seabirds and migrating
pile driving operations, the number of sea mammals decreases, but passerines. Hence, before installing a wind farm it will be necessary
following the construction period they return although in lower to collect data on bird numbers, distribution and movement in order
numbers than before [20]. to predict impacts, because an inappropriate wind farm location can
As the machinery used is the same and the duration of the work adversely affect wild bird populations [27]. Obviously, the choice of
will not vary significantly, it can be assumed that the choice of foundation will not affect the impact on birds.
foundation will not affect the impact.
3.2. Economic assessment
3.1.3. Water turbidity
Installation and decommissioning of the foundations and
The aim of an economic assessment is to study the economic
cabling will result in considerable disturbance of the seabed,
efficiency of building a new offshore wind farm using specific
resulting in the removal or physical disruption of benthic
assumptions. For this study a hypothetical wind farm of 30 wind
communities and suspension of sediment. However, it is expected
turbines at a distance of 11 km is assumed. The life cycle period of
that the amount of deposited material is not enough to affect the
consideration depends both on the local climatic conditions and
morphology of the seabed [37]. For this study, it will be assumed
the quality of the components of the wind turbine. Although it can
that the environmental impact is proportional to the soil volume
be analytically derived, a normal service period of 20 years is
affected by the piles, as presented in Table 9, illustrating that the
usually assumed [28].
jacket type of configuration would produce more water turbidity.
3.2.1. Impacts identification
3.1.4. Artificial reefs
The positive impacts, which will be referred to as benefits, will
Wind farm structures and any scour protection provide addi-
account either for an improvement of goods that generate positive
tional substrata for colonization by epifaunal communities,
utility or a reduction in total price. The negative impacts, which will
potentially increasing the biodiversity of the locality [25], although
be referred to as costs, will either be any reduction in the quantity
the benthic community may be modified as a result. For this study,
or quality of goods or an increase in total price [29].
it will be assumed that this environmental benefit is proportional
The construction of wind farms involves several benefits,
to the total surface of each structural option, as presented in
including increasing employment and at the same time decreasing
Table 10, according to which the monopile structure would produce
fossil fuel consumption. In 2007, the EU wind energy sector directly
the lowest increase in biodiversity.
employed approximately 108,600 people and if indirect employ-
ment is included, a total of approximately 154,000 people [30].
Table 5
Optimized tripod design.
Table 7
Length (m) Diameter (m) Thickness (m) Maximum displacement and von Mises stresses.
Transition piece 10 8 0.4
Central Column 35 8 0.04e0.08 Maximum displacement Maximum von Mises stress
Braces 36.06 4.1 0.026 Monopile 2.37 m 177.6 MPa
Base 20e34.64 2.1 0.016 Tripod 3.30 m 184 MPa
Piles 35 3.6 0.04 Jacket 2.69 m 223.8 MPa
2834 E. Lozano-Minguez et al. / Renewable Energy 36 (2011) 2831e2837

Table 8 Table 10
CO2 emissions for different support structures. Total surface of each structural option.

Mass of steel (kg) N2O (kg) CH4 (kg) CO (kg) CO2e (kg) Monopile Tripod Jacket
Monopile 811565 57 32 755 17191 Total surface (m2) 1100 3857 8787
Tripod 1689552 118 68 1571 35788
Jacket 2447038 171 98 2276 51833
than private sector projects [38], due to the fact that they can pool
For the derivation of the amount of energy produced per year, it risks [39] and can borrow at far lower rates [38].
is important to introduce the two important concepts of availability Within the scope of this study, a public/private partnership will
and load factor. Wind turbine availability is its capability to operate be considered and a 6% discount rate will be applied. Once all PVs
when the wind is blowing and, for modern European machines, is have been calculated, the sum of discounted gains will be checked
in the range of 98% [31] or 95% as a more conservative approach. to see whether it exceeds the sum of discounted losses, accepting
The load factor is the ratio of the net amount of electricity gener- projects with an NPV greater than 1. The wind turbines’ value for
ated to the net amount which it could have generated if it were each different support structure configuration and the corre-
operating at its net output capacity e usually around 30% [32]. sponding NPVs are presented in Table 11.
For the theoretical wind farm under consideration, a total
output of 411,939 MWh will be assumed based on typical 5.5 MW 3.2.3. Sensitivity analysis
wind turbines that are adopted by [33], having an approximate cost This section will investigate the contribution of different
of 4 M£. The cost of electricity generation is selected as 9 p£/kWh parameters in the derived value of NPV. The first analysis will
based on [34] and [35], while the costs of the support structures is consider the cost of electricity for each offshore wind turbine type.
assumed to be £1500 per tonne [33], including steel costs, fabri- This is an important step because adding wind into the power mix
cation costs and protection against corrosion. The operational will have a significant influence on the resulting price of electricity.
expenditure (OPEX) cost of maintaining an offshore wind farm is in The cost of generating electricity for the monopile wind farm is
the range of 23% of the capital expenditure (CAPEX) cost [33], 5.03 p/kWh, for the tripod wind farm is 5.96 p/kWh, and for the
spread over the life cycle. Obviously initial costs will be low because jacket wind farm is 6.77 p/kWh; which are in accordance with the
of warranty and final costs will be higher as the farm comes to the average cost of generating electricity mentioned elsewhere.
end of its life, therefore it will be assumed that costs are spread Therefore, these projects would be profitable even with a fall of
following a linear increase from year 1. The costs of the electrical 2.2 p/kWh of selling price, which according to the Department of
infrastructure correspond to the costs of copper offshore and Energy and Climate Change (DECC) is unlikely [40].
onshore cables, cable laying, meteorological masts, a Supervisory As mentioned earlier, the discount rate is a crucial parameter in
Control And Data Acquisition (SCADA) system, and the onshore determining whether the PV is positive or negative, so the second
electrical works are based on [33] and [36]. Costs for transportation analysis was to calculate the Internal Rate of Return (IRR) for each of
and installation of the foundations of a wind turbine are based on the three proposed projects. IRR is the interest rate at which the
[33], appropriately incorporating potential downtime due to NPV of a project is zero [41]. The results of this analysis are pre-
extreme weather conditions. Installation of scour protection and sented in Fig. 3, proving that an increase in the discount rate
a cost for decommissioning has been included in the study and are implies a decrease in the NPV. Further it is derived that all projects
based on [33] and [37]. For the total derived costs for each wind are likely to receive private investment, since all the IRRs are
turbine, a cost breakdown chart can be formulated and this is greater than the cost of capital.
presented in Fig. 2, illustrating that the support structures cost
represents 15% of the total costs for the monopile wind farm, while 4. TOPSIS Method
for the jacket wind farm it represents 33% of the total costs.
4.1. Multi-criteria decision analysis
3.2.2. Net present value (NPV)
This parameter will convert the total cost of the service life of As mentioned in the introduction and noted throughout this study,
the structure to present value. The present value (PV) of a benefit or the selection of the optimum choice should be based on several
cost (X) is calculated as follows for t years and a discount rate of i: attributes; for instance, monopile is economically the best option but
it is the worst option for increasing biodiversity. Hence, a Multi-
h i
PV ðXt Þ ¼ Xt ð1 þ iÞt criteria decision analysis (MCDA) method can be employed, capable
of providing the best sustainable alternative in this decision making
The further in time a benefit or cost happens, the lower the (DM) process which has conflicting attributes. MCDA methods have
discount factor [(1 þ i)t] and the present value is. This is justified as been widely applied in DM for sustainable energy because of the
there is a level of uncertainty associated with the costs and benefits complexity of socio-economic and biophysical systems and the multi-
in the future, hence there is an expressed preference to obtain goods dimensionality of the sustainability target [13].
and services now, rather than later [29]. Therefore, the choice of
discount rate is of critical importance in determining whether the 4.2. Application of the TOPSIS method
PV is positive or negative. There is considerable controversy about
whether public sector projects should be discounted at a lower rate The TOPSIS method is a powerful MCDA method used
commonly in optimization problems, and has been chosen from the
Table 9
Affected soil volume. Table 11
Values of different support structures.
Number of piles Length Radius Area Volume
Length (m) (m) (m) (m2) (m3) NPV (£)
Monopile 1 40 3.5 38.5 1539.4 Monopile 187,682,531
Tripod 3 35 1.8 10.2 1068.8 Tripod 143,628,070
Jacket 4 30 2.5 19.6 2356.2 Jacket 105,570,459
 E. Lozano-Minguez et al. / Renewable Energy 36 (2011) 2831e2837 2835

j: 1, ., n
rij: value of the component that resides in row i and in column j
of the normalized matrix
Next, criteria weights (wj) have to be determined to indicate
their relative importance and to calculate the weighted normalized
values (vij) through:
vij ¼ wj  rij

Having obtained the Weighted Normalized Decision Matrix, the

PIS (Aþ) and the NIS (A), the ideal solutions are determined as:
    
Aþ ¼ vþ þ þ
1 ; .; vj ; .; vn ¼ maxj vij  j ¼ 1; .; n  i

¼ 1; .; m
n o   
A ¼ v  
1 ; .; vj ; .; vn ¼ minj vij  j ¼ 1; .; n  i

¼ 1; .; m

Finally, the ranking of the alternatives will be realized by

calculating the relative distance of each solution from the PIS (Sþ
i )
and to the NIS (Si ), as:
0 1
X  2
Sþ
i ¼ sqrt@ vþ
j
 vij A; S 
i
j ¼ 1 to n
Fig. 2. Cost breakdown of different configurations.
0 1
X  2
different types of weighting methods because its basic concept is ¼ sqrt @ v
j  vij
A
perfectly suited to this analysis, as [14] suggests. Its basic concept is j ¼ 1 to n
that the best alternative should have the farthest distance from the
Negative Ideal Solution (NIS) and the shortest distance from the The relative closeness of each solution to the ideal (Ci) will be esti-
Positive Ideal Solution (PIS) [42]. This method ensures the realiza- mated as follows, and the most favourable will be the one closest to 1.
 
tion of objective benchmarking among the options, taking into Ci ¼ S
i = Sþ þ S
i i
account both quantitative and qualitative attributes, and follows
a sequence of simple steps as presented in Fig. 4.
After the formulation of a design matrix with scoring for every
attribute of each option, normalization follows as: 4.3. Analysis of attributes
!
X Since this study aims to identify the best out of the three most
rij ¼ xij =sqrt x2ij commonly used support structure configurations, there are three
i ¼ 1 to m alternatives (m ¼ 3), and nine criteria have been selected (n ¼ 9) for
Where: this problem. The selection has been based on the conclusions
xij: value of the component that resides in row i and in column j obtained through the different steps of the study and the recom-
of the decision matrix mendations given in [14] of some attributes such as certification,
i: 1, ., m durability and depth compatibility. All these attributes are defined as
follows:

 Artificial reefs; this positive criterion, i.e. the higher the better,
aims to represent the increment of biodiversity provided by
additional substrata for colonization.
 Certification is a positive parameter; it shows if the support
structure has already been certified for wind turbines or not.
Considering a value of 1 if it has, 0.5 if not but it has already
been certified for the same structure used in the offshore oil
and gas industry, and 0 if it has not.
 CO2e emission is a negative attribute, i.e. the higher emissions
reflect the amount of CO2e emissions produced for the fabri-
cation of the different support structures.
 Depth compatibility is a negative criterion, representing the
compatibility of each support structure with 40 m water depth. It
is scored as 1 if 40 m depth is within the range of reference depth,
2 if it is within the extended range of reference depth: [minimum
reference value*0,75 e maximum reference value*1.25], and 3 if
it is not within the extended range of reference depth.
 Maximum displacement is a negative parameter representing
the resistance of the structure to lateral displacement as
Fig. 3. IRRs for proposed farms. a result of the forces acting on it.
2836 E. Lozano-Minguez et al. / Renewable Energy 36 (2011) 2831e2837

Fig. 4. TOPSIS Method flowchart.

 Durability is a positive criterion, related to the resistance to the present study, the weight vector will be based on the experi-
age-related deterioration. It is marked with values between 1 ence of the experts in this field within the Cranfield Offshore
and 5, obtained from [14], depending on the exposure to Renewable Energy Group. The relative importance of each criterion
corrosion and consequences of fatigue. was determined by a questionnaire based on a Likert scale. With
 NPV is a positive attribute that shows the economic benefit of this technique, the responder specifies a level of agreement or
selling energy according to the type of offshore wind farm. disagreement to the concept under study, using one of a number of
 Maximum von Mises Stress is a negative criterion chosen to positions on a five-point Likert scale [44]. In this particular case, the
account for the suitability of the support structures, as it relates meaning of Likert scale levels was modified, running from 1-Not
to the ductileebrittle transition [43]. important to 5-Very Important.
 Water turbidity, is a negative attribute aiming to represent
disturbance to the seabed caused by the support structure. 4.4. Application

Based on the above analysis, the initial Decision Matrix is defined as:
1 2 3 4 5 6 7 8 9
Monopile 1100 1 17191 3 2.37 4 187,682,531 177.6 1539
Tripod 3857 1 35788 1 3.30 5 143,628,070 184 1069
Jacket 8787 1 51833 1 2.69 5 105,570,459 223.8 2356

The normalized Decision Matrix is derived as:

1 2 3 4 5 6 7 8 9
Monopile 0.11 0.58 0.26 0.90 0.49 0.49 0.68 0.52 0.51
Tripod 0.40 0.58 0.55 0.30 0.68 0.62 0.57 0.54 0.36
Jacket 0.91 0.58 0.79 0.30 0.55 0.62 0.47 0.66 0.78

Results from the averaging of the questionnaires construct the weights vector.
1 2 3 4 5 6 7 8 9
0.65 0.65 0.91 0.91 0.87 1.00 1.00 0.83 0.74

And the Weighted Normalized Decision Matrix:

1 2 3 4 5 6 7 8 9
Monopile 0.07 0.38 0.24 0.83 0.42 0.49 0.68 0.43 0.38
Tripod 0.26 0.38 0.50 0.28 0.59 0.62 0.57 0.45 0.26
Jacket 0.59 0.38 0.72 0.28 0.48 0.62 0.47 0.54 0.58

The derived Positive and Negative Ideal Solutions are derived as:
A+ 0.59 0.38 0.24 0.28 0.42 0.62 0.68 0.43 0.26
A- 0.07 0.38 0.72 0.83 0.59 0.49 0.47 0.54 0.58

Weights influence directly the DM result and are based on the From the calculation of the relative closeness of each
practical engineering expertise of the decision makers; conse- support structure to the ideal solution, the tripod was found to
quently, the more experienced the decision makers are, the more provide the best option (0.61) against the monopile (0.44) and
objective the result. Although most of the attributes can be the jacket structure (0.55). Further to this analysis, a sensitivity
expressed in quantitative terms, this is every demanding task. For analysis was executed, considering equal weights (1); this study
 E. Lozano-Minguez et al. / Renewable Energy 36 (2011) 2831e2837 2837

will show the importance of the experienced allocation of weight [13] Wang J, Jing Y, Zhang C, Zhao J. Review on multi-criteria decision analysis aid
in sustainable energy decision-making. Renewable and Sustainable Energy
factors to the derived results. Calculation of the new values of
Reviews 2009;13:2263e78.
relative closeness will change the total performance of the [14] Kolios A, Collu M, Chahardehi A, Brennan FP, Patel MH. A multi-criteria
different options, ranking equally the Tripod and the Jacket decision making method to Compare support structures for offshore wind
structures (0.58) and still keeping the monopile in the third place turbines. Warsaw: EWEC; 2010.
[15] DNV. Design of offshore wind turbines structures. Offshore Standard DNV-OS-
(0.40). J101; 2010.
[16] CEN. Eurocode 3-Design of steel structures. BS EN 1993-1-1:2005.
5. Conclusions [17] Jonkman J, Butterfield S, Musial W, Scott G. Definition of a 5-MW reference
wind turbine for offshore system development. Colorado: National Renewable
Energy Laboratory; 2009.
The outcome of this comprehensive study, taking into account [18] Bruns E, Andersson A, Thor S-E. Environmental issues of offshore wind farms.
additional attributes more than absolute cost, illustrates that for Available from: http://www.ieawind.org/Task_11/TopicalExpert/Summary_
40_Offshore.pdf; 2002.
the assumptions considered, the tripod is the best option overall. [19] Punt MJ, Groeneveld RA, van Ierland EC, Stel JH. Spatial planning of offshore
This seems reasonable because, although the monopile is the most wind farms: a windfall to marine environmental protection? Ecological
economical option and less harmful to the environment, the tripod Economics 2009;69:93e103.
[20] Snyder B, Kaise MJ. Ecological and economic cost-benefit analysis of offshore
suffers less from wave-resonance than the monopile. Since the wind energy. Renewable Energy 2008;34:1567e78.
obtained results in the different sections provide a consistent end [21] Pertsova CC. Ecological economics research trends. New York: Nova Science
result, it can be concluded that the methodology that has been Publishers; 2007.
[22] Schleisner L. Life cycle assessment of a wind farm and related externalities.
followed and is proposed in this paper is appropriate. This
Renewable Energy 2000;20:279e88.
considers not only the methodology used for soilestructure inter- [23] Dean WD. Wind turbine mechanical vibrations: potential environmental
action, environmental and operational loads, but also imple- threat. Available from: http://www.sosmoray.org.uk/Dixie_dean.pdf; 2007.
[24] Sceirde offshore wind farm e environ impact statement. Aqua-Fact Interna-
mentation of the TOPSIS method in order to provide an objective
tional Services Ltd. Available from: http://www.fsteo.com/Non-Technical%
methodology for benchmarking the different support structure 2520Summary.pdf; 2008.
options, taking into account engineering, economic and environ- [25] Hiscock K, Tyler-Walters H, Jones H. High level environmental screening study
mental criteria. for offshore wind farm developments e marine habitats and species project.
Available from: http://citeseerx.ist.psu.edu/viewdoc/download?doi¼10.1.1.
A sensitivity analysis, with a more analytical consideration of 113.1621&rep¼rep1&type¼pdf; 2002.
the weight factors employed using the TOPSIS method, has [26] Gill AB, Gloyne-Phillips I, Neal KJ, Kimber JA. The potential effects of elec-
demonstrated the effect of each of the different attributes on the tromagnetic fields generated by sub-sea power cables associated with
offshore wind farm developments on electrically and magnetically sensitive
total scoring, increasing the level of objectiveness of the classifi- marine organisms - a review. Available from: http://www.offshorewindfarms.
cation of the different options. co.uk/Assets/1351_emf_phase_one_half_report.pdf; 2005.
Future work, following this study, should consider examination [27] Drewitt AL, Langston RHW. Assessing the impacts of wind farms on birds. Ibis
2006;148(s1):29e42.
of the effect of the water depth and the environmental conditions’ [28] DNV. Design of offshore wind turbine structures. Offshore Standard DNV-OS-
consideration to the decision making process towards the selection J101; 2007.
of the most suitable configuration as well as the performance of [29] Hanley N, Spash CL. Cost-Benefit analysis and the environment. Cheltenham:
Edward Elgar Publishing Inc; 2003. pp. 53e73, 261-273.
other attributes. Lastly, the quantification of qualitative attributes
[30] Blanco I, Kjaer C. Wind at work. Available from: http://www.ewea.org/fileadmin/
on the basis of a reference unit would contribute towards a more ewea_documents/documents/publications/Wind_at_work_FINAL.pdf; 2009.
informed decision making process. [31] Bwea. Wind turbine technology. Available from: http://www.bwea.com/pdf/
briefings/technology05_small.pdf; 2005.
[32] Renewable UK. Can we rely on the wind?, http://www.bwea.com/energy/rely.
References html; 2010. Available from:.
[33] ODE. Study of the costs of offshore wind generation. Department of trade and
[1] Nikolaos N. Deep water offshore wind technologies. Glasgow: University of industry. Available from: http://www.ode-ltd.co.uk/renewables/dti%
Strathclyde; 2004. 2520Costs%2520of%2520Offshore%2520Wind%2520Generation%2520by%
[2] U.S. Department of Energy. Historic wind development in New England: the 2520ode.pdf; 2007.
age of PURPA spawns the “wind farm”. Available from: http://www. [34] Nottingham Energy Partnership. Energy costs comparison. Available from:
windpoweringamerica.gov/ne_history_windfarms.asp; 2010. http://www.nottenergy.com/energy-costs-comparison3; 2011.
[3] Renewable UK. Onshore wind, http://www.bwea.com/onshore/index.html; [35] The Royal Academy of Engineering. Costs of generating electricity. Available from:
2010. Available from:. http://www.raeng.org.uk/news/publications/list/reports/Cost_Generation_
[4] World of wind energy. What is wind turbine sitting?. Available from: http://www. Commentary.pdf; 2004.
worldofwindenergy.com/vbnews.php?do¼viewarticle&artid¼31&title¼wind- [36] Morgan CA, Snodin HM, Scott NC. Offshore wind economies of scale, engi-
energy; 2010. neering resource and load factors. Available from: http://www.decc.gov.uk/
[5] Wind Energy Update. Offshore wind piques investor interest. Available from: assets/decc/what%20we%20do/uk%20energy%20supply/energy%20mix/
http://social.windenergyupdate.com/qa/offshore-wind-piques-investor-interest; renewable%20energy/policy/file22067.pdf; 2003.
2010. [37] Iberdrola Renewables. Buffalo ridge II wind farm decommissioning report.
[6] BERR, Department for Business Enterprise & Regulatory Reform. Review of Available from: http://puc.sd.gov/commission/dockets/electric/2008/el08e031/
reef effects of offshore wind farm structures and potential for enhancement Appendix%20h.pdf; 2008.
and mitigation. Available from: http://webarchive.nationalarchives.gov.uk/ [38] Grout PA. Public and private sector discount rates in public-private partnerships.
þ/http://www.berr.gov.uk/files/file43528.pdf; 2008. Available from: http://www.bristol.ac.uk/cmpo/publications/papers/2003/
[7] Pepper L. Monitoring & evaluation of Blyth offshore wind farm. Available wp59.pdf; 2002.
from: http://webarchive.nationalarchives.gov.uk/þ/http://www.berr.gov.uk/ [39] Arrow K, Lind RC. Uncertainty and the evaluation of public investment deci-
files/file20364.pdf; 2001. sions. The American Economic Review 1970;60:364e78.
[8] RenewableUK. Offshore wind energy reaches 1 GW. Available from: http:// [40] DECC. Cost financial support for offshore wind. Available from: http://
www.bwea.com/media/news/articles/pr20100423.html; 2010. webarchive.nationalarchives.gov.uk/þ/http://www.berr.gov.uk/files/
[9] Energy Efficiency news. UK becomes world leader in offshore wind. Available file51142.pdf; 2009.
from: http://www.energyefficiencynews.com/i/1456/; 2008. [41] Black J, Hashimzade N, Myles G. A dictionary of economics. Oxford: Oxford
[10] European Commission. Renewable energy e targets. Available from: http://ec. University Press; 2009.
europa.eu/energy/renewables/targets_en.htm; 2011. [42] Hwang C-L, Yoon K. Multiple attribute decision-makingmethods and appli-
[11] European Union Committee. The EU’s target for renewable energy: 20% by cation: a state-of-the-art survey. New York: Springer-Verlag; 1981.
2020. Available from: http://www.publications.parliament.uk/pa/ld200708/ [43] n/a ANTEC’96: plaastics-racing into the future. Society of Plastics Engineers
ldselect/ldeucom/175/175.pdf; 2008. 1996;vol. 3. Special areas.
[12] Sustainable Development Commission. Wind Power in the UK. Available from: [44] McDaniel CD, Gates RH. Marketing research essentials. USA: South-Western
http://www.sd-commission.org.uk/pages/wind-power-in-the-uk.html; 2005. College Publishing; 1998. pp. 247e248.