See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/360264842

Review of scaling laws applied to floating offshore wind turbines

Article in Renewable and Sustainable Energy Reviews · July 2022

DOI: 10.1016/j.rser.2022.112477

CITATIONS READS

46 6,050

6 authors, including:

Nataliia Y. Sergiienko Leandro Souza Pinheiro Da Silva

The University of Adelaide The University of Adelaide
107 PUBLICATIONS 1,225 CITATIONS 45 PUBLICATIONS 437 CITATIONS

SEE PROFILE SEE PROFILE

Erin E. Bachynski-Polić Ben Cazzolato

Norwegian University of Science and Technology The University of Adelaide
137 PUBLICATIONS 3,125 CITATIONS 401 PUBLICATIONS 5,679 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Nataliia Y. Sergiienko on 06 September 2022.

The user has requested enhancement of the downloaded file.

Review of scaling laws applied to floating offshore wind turbines
Sergiienko N.Y.a,∗ , da Silva, L.S.P.a , Bachynski-Polić, E.E.b , Cazzolato, B.S.a , Arjomandi, M.a
and Ding, B.a
a The University of Adelaide, School of Mechanical Engineering, Adelaide, 5005, SA, Australia
b Norwegian University of Science and Technology, Department of Marine Technology, Trondheim, 7491, Norway

ARTICLE INFO Abstract

Keywords: The wind energy industry is moving to offshore installations allowing for larger wind turbines
wind energy to be deployed in deep-water regions with higher and steadier wind speeds. Floating offshore
floating offshore wind wind turbines consist of two main subsystems: a wind turbine itself and a floating substructure
scaling factor that supports it and provides stability. While the wind turbine technology is mature, the floating
dynamics support structures for offshore wind turbines are still evolving and have not been deployed at
a commercial scale. Due to a significant increase in the size of wind turbines over the last
decade, it is important to understand how to design the floating platform to support larger wind
turbines, and how the dynamics of the entire system change with increasing scale. Firstly, this
article provides an overview of the trends in wind energy systems for offshore applications.
Secondly, a review of existing semi-submersible platforms designed to support 5 − 15 MW wind
turbines is provided. In addition, this article provides a comparative analysis of the techniques
proposed to upscale floating support structures for larger wind energy systems with a particular
focus on the system dynamics. The results demonstrate that the wind turbine mass, rated power
and rotor thrust force scale with close to square rotor diameter. Towers designed for floating
wind applications are usually significantly stiffer and heavier as compared to their fixed-bottom
counterparts to place the tower’s natural frequencies outside the wave excitation region. The
analysis of semi-submersible platforms revealed a strong correlation between the wind turbine
rotor diameter and the product of the distance to the offset columns and their diameter. Also,
it has been found that design practices adapted by the platform developers roughly follow
the theoretical square-cube (or ‘mass’) scaling law when designing platforms for larger wind
turbines.

1. Introduction
Wind energy is one of the mature renewable energy technologies that contributed to the estimated 743 GW of
electrical power worldwide in 2020 [1], accounting for 26% of the renewable energy market. Great potential for
expanding the wind energy industry to offshore locations is expected, where stronger and more consistent winds favor
the installation of larger wind turbines [2, 3]. Although offshore wind provided just 0.3% of global electricity supply
in 2020 [4], it is a rapidly growing global industry aiming to become the European Union’s largest source of electricity
in the 2050s [5].
Offshore wind energy is converted into electricity using wind turbines installed in coastal waters on either fixed-
bottom [6] or floating substructures anchored to the seabed [7]. About 80% of offshore wind resources are concentrated
in waters deeper than 60 m [8] where the installation of conventional fixed-bottom support becomes uneconomical
favoring the deployment of floating (buoyant) foundations. At present, floating support structures are more costly
compared to fixed-bottom systems but have many potential advantages. These include higher and steadier wind speeds
in deep-water regions, less visual and environmental impact, and easier assembly and installation [3, 9].
The technology for capturing wind energy has converged to a three-blade horizontal axis wind turbine for both
onshore and offshore installations. In contrast to wind turbines, floating platform technology for offshore wind turbines
is still evolving, and more cost-effective design solutions are emerging. The first pre-commercial floating wind turbines
were designed and manufactured based on the technology developed over decades for land-based or fixed-bottom
applications and adapted floating structure concepts from the offshore oil and gas industry [10, 11]. These first pilot
projects successfully demonstrated survivability of the platform and its power production capability, but many technical
∗ Corresponding author
[email protected] (S. N.Y.)
ORCID (s): 0000-0000-0000-0000 (S. N.Y.)

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 1 of 24
challenges remain unaddressed before the economic viability of this technology is assured. In addition, it should be
noted that design practices in the offshore oil and gas industry have very high safety factors in order to reduce the risks
of failures (and potential loss of life) leading to an order of magnitude increase in costs [12]. On the contrary, offshore
wind turbines have much lower consequences of failure, and are more susceptible to wind loads. Therefore, practices
adapted from offshore oil and gas applications might not be applicable for floating offshore wind turbines [13].
Due to the power capture benefits, wind turbines have grown significantly in size over the last decade. Currently,
offshore wind turbines are twice as big as their land-based counterparts, and their power ratings range from 3 to
15 MW and continue to grow as technology advances [9]. This increase in size leads to a significant reduction in
the cost of construction, installation, operation and maintenance, and higher capacity factors [2]. To support such
large turbines and provide the required static and dynamic stability to the system, floating platforms will also need
to be upscaled. However, due to the large number of possible design solutions and configurations of floating support
structures, there are no clear guidelines on how to increase the platform dimensions. Several different approaches have
been proposed in the literature, from simply applying geometric scale factors to an existing platform design [14–18]
to a full optimisation of an “integrated” wind energy system (including turbine and platform) following a systems-
engineering design approach [9]. In light of wind turbine rotor diameter approaching 300 m, the question arises, if
there are any limitations on the size of the offshore wind turbines, and what implications this might have on the design
of the associated floating platforms.
A number of studies have attempted to investigate scaling laws for wind turbines with an objective to answer a
question: how large a wind turbine can be built? Theoretical and practical scaling trends for onshore and offshore wind
turbines with a rated power of up to 7 MW have been investigated in [19–22]. Their analysis is focused primarily on
the technical and economic feasibility of multi-megawatt wind turbines with a rated power of more than 10 MW. It
has been found that the weight of large-size wind turbines increases with the cube of their rotor diameter, and if this
trend continues, it would be impossible to achieve a lower cost of energy by upscaling wind turbines without additional
technological improvements. Regarding the design of floating platforms for offshore wind turbines, to the best of the
authors’ knowledge, there is no systematic assessment of their scaling and its impact on dynamic response.
This paper has three main objectives:
1) to investigate the turbine and tower designs adapted by floating platform developers in order to understand how these
designs deviate from theoretical scaling;
2) to explore practical scaling trends for semi-submersible platforms depending on installed turbine sizes;
3) to study how the theoretical laws of platform scaling proposed in the literature affect the dynamic behaviour of
floating platforms and relate these theoretical laws to established industry practices.
This paper is primarily designed for the developers of floating support structures with a major focus on the system
dynamics. As dynamic behaviour is a substantial part of any techno-economic study, the application of this paper is
primarily for feasibility assessment and early stages of the development of floating platforms. Therefore, it is expected
that the results demonstrated in this research will be used as a starting point for designing new floating platforms, or
upscaling existing design solutions.
The remainder of this paper is organised as follows (see Figure 1). First, a review of reference offshore wind turbines
with rated power up to 20 MW is presented in Section 2 with a particular focus on the scaling laws and implications
in their design for floating offshore applications. Next, design practices adapted for floating offshore wind turbines are
discussed in Section 3 followed by a review of proposed floating semi-submersible platforms. Finally, an overview of
different scaling procedures proposed in the literature is provided in Section 4.
It is important to note that the analysis in this paper is based on the best currently available data for a range of
wind turbines and platforms. However, the reviewed technologies might be at different stages of commercialisation
and technological readiness, which introduces some uncertainty in the presented conclusions.

2. Reference offshore wind turbines

The detailed specifications of commercial wind turbines are scarce in the publicly available literature. Therefore,
a number of research institutions have published reference wind turbine models to accelerate the research and
development in the wind energy sector. These models include blade geometries, inertial properties of the rotor-nacelle-
assembly (RNA), drivetrain characteristics, baseline settings of the control system, and tower design. For offshore
applications, such information is available for wind turbines with a rated power between 1.5 MW and 20 MW. For this

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 2 of 24
 D
Reference
offshore wind turbines

Section 2

Existing Scaling laws

floating for floating
platforms platforms
L Section 3 Section 4

Figure 1: Visualisation of the paper structure.

work, the data have been collected from nine baseline (not commercial) wind turbines and are presented in Table 1,
including:
• WindPACT 1.5 MW (WP) turbine developed by National Renewable Energy Laboratories (NREL), US [23];
• 2 MW turbine proposed by the University of Tokyo and Hitachi, Ltd [24];
• NREL offshore 5-MW wind turbine [25];
• LEANWIND 8 MW reference turbine (LW) based on the Vestas V164 – 8 MW turbine [26];
• DTU 10 MW reference wind turbine developed by Technical University of Denmark [27];
• IEA 10-MW reference wind turbine developed within the second work package of IEA Wind Task 37 on
Wind Energy Systems Engineering (collaboration between National Renewable Energy Laboratories, USA,
Wind Energy Institute, Technische Universität München, Germany, SINTEF Energy Research AS, Trondheim,
Norway, and Wind Energy Department, Technical University of Denmark, Roskilde, Denmark) [28];
• INO WINDMOOR 12-MW wind turbine developed by a consortium led by SINTEF Ocean AS, Trondheim,
Norway [29];
• IEA 15-MW offshore reference wind turbine developed by National Renewable Energy Laboratories, USA, and
the Technical University of Denmark, Denmark [30]; and
• a 20 MW wind turbine proposed by University of Texas, University of Michigan and Delft University of
Technology [31].
The design of any wind turbine directly depends on its rated power, which in turn is closely linked to the rotor
diameter. Therefore, the latter is usually considered as the main independent variable that defines all other design and
performance parameters of the rotor-nacelle-assembly, support tower, and floating platform (for offshore applications).
To fully understand how mass, loads, and structural properties scale with turbine size, scaling laws have been identified
for each of the parameters in Table 1 depending on the rotor diameter 𝐷 using two different approaches:
• Best power fit as 𝑦 = 𝑎𝐷𝑏 , where 𝐷 is the rotor diameter, and coefficients 𝑎 and 𝑏 are unknown;
• Heuristic engineering approach as 𝑦 = 𝑐𝐷𝑑 + 𝑓 , where 𝑐 and 𝑓 are unknown (𝑓 ≠ 0 for the tower linear
dimensions), and the exponent 𝑑 is maintained constant according to the expected geometric upscaling.
The fitted coefficients have been obtained using the method of least squares with a goodness-of-fit represented
by the coefficient of determination R-squared, or R2 (refer to the MATLAB in-built function fit for more details).
In Table 1, values marked with * demonstrate design parameters for floating offshore applications, in particular tower
mass, thickness of tower walls, and tower natural frequencies. Showcasing only several examples from Table 1, Figure 2
demonstrates how turbine rated power, thrust force, nacelle mass and hub height scale with changes in rotor diameter.
In general, the proper fit is obtained using both the best power fit and heuristic engineering approaches that can be used

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 3 of 24
to estimate the properties of larger wind turbines. The following subsections describe the identified trends in more
detail, separately for the rotor-nacelle-assembly and the tower design.

(a) (b)

(c) (d)
Figure 2: Dependence of the (a) turbine rated power, (b) thrust force, (c) nacelle mass, and (d) hub height on the turbine
rotor diameter. In (c), gray circles correspond to the turbines with direct-drive, while black circles correspond to geared
wind turbines.

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 4 of 24
Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier

Table 1
Parameters for the reference (not commercial) offshore wind turbines and associated tower designs for both fixed-bottom and floating support structures. All the
values have been taken from the corresponding references provided in the table. The scaling laws have been identified by the authors based on the data in the table.
The scaling laws include: (i) best power fit as 𝑦 = 𝑎𝐷𝑏 , where 𝐷 is the rotor diameter, and coefficients 𝑎 and 𝑏 are unknown, and (ii) heuristic engineering approach
as 𝑦 = 𝑐𝐷𝑑 + 𝑓 , where 𝑐 and 𝑓 are unknown (𝑓 ≠ 0 for the tower linear dimensions), and the exponent 𝑑 is maintained constant according to the expected geometric
upscaling.
Scaling law 1.5 MW 2 MW 5 MW 8 MW 10 MW 12 MW 15 MW 20 MW
Wind turbine
Best Heuristic WP Hitachi NREL LW DTU IEA WINDMOOR IEA Texas
power fit (R2 ) engineering fit (R2 ) [23] [24] [25, 32] [26] [27, 33, 34] [28] [29] [30, 35] [31]
Rated power, MW 821𝐷1.79 (99.1%) 267𝐷2 (98.5%) 1.5 2 5 8 10 10 12 15 20
Rotor diameter, m 𝐷 𝐷 70 80 126 164 178.3 198 216.9 240 276
Rated wind speed, m/s 11.4 12 11.4 12.5 11.4 11 10.6 10.6 10.7
Tip speed, m/s 75 73.3 80 90 90 88.6 95 103.3
Rotor speed min, rpm 245𝐷−0.71 (96.2%) 982𝐷−1 (78.5%) 11.1 6.9 6.3 6 6 5.5 5 4.5
Rotor speed max, rpm 731𝐷−0.84 (99.1%) 1561𝐷−1 (95.5%) 20.5 19.6 12.1 10.5 9.6 8.7 7.8 7.6 7.1
Drive/gearbox ratio 88:1 98:1 97:1 N/A 50:1 Direct Direct Direct 164:1
Peak thrust, kN 84𝐷1.86 (97.3%) 40.3𝐷2 (97.1%) 250 300 700 1050 1500 1350 1950 2300
Blade mass, t 3.07𝐷1.83 (98.7%) 1.19𝐷2 (98.2%) 4.3 6.9 17.7 35 41.7 47.7 63 65 86.3
Hub mass, t 0.23𝐷2.46 (83.0%) 2.89𝐷2 (81.6%) 19.2 20.2 56.8 90 105.5 81.7 60 190 252.8
Rotor mass, t 5𝐷2.05 (97.2%) 6.46𝐷2 (97.2%) 32.2 41.2 110 195 230 225 250 385 511.8
Nacelle mass, t 24.3𝐷1.88 (97.6%) 12.3𝐷2 (97.4%) 52.8 75.2 240 285 446 543 600 630 945
Tower mass, t
- fixed-bottom 12.65𝐷2.07 (93.5%) 18.63𝐷2 (93.5%) 125.4 196.5 347.5 558 605 628 N/A 860 1588.3
- floating∗ 249.7∗ 1,257∗ 1162∗ 1,263∗
Tower height, m 3.41𝐷0.67 (86.8%) 0.43𝐷1 + 30 (86.9%) 82.4 75.8 87.6 106.3 115.6 115.6 110.2 129.6 155
Hub height, m 2.41𝐷0.75 (97.5%) 0.47𝐷1 + 31 (97.6%) 84 80 90 110 119 119 132 150 160
Tower diameter (bottom), m 0.23𝐷0.68 (94.5%) 0.032𝐷1 + 2.23 (92.7%) 5.66 4.3 6 7.7 8.3 8.3 9.9 10 10
Tower diameter (top), m 0.15𝐷0.68 (92.4%) 0.020𝐷1 + 1.39 (89.5%) 2.57 2.5 3.87 5 5.5 5.5 6.0 6.5 6.2
Tower thickness (bottom), mm
- fixed-bottom 0.16𝐷1.06 (97.3%) 0.22𝐷1 (96.8%) 17.39 32 27 36 38 38 N/A 55 63
- floating∗ 75∗ 90∗ 83∗
Tower thickness (top), mm
- fixed-bottom 0.18𝐷0.93 (79.1%) 0.12𝐷1 (78.8%) 10.26 18 20 22 20 20 N/A 25 39
- floating∗ 29∗ 30∗ 21∗
Page 5 of 24

Tower natural frequency, Hz

- fixed-bottom 26.4𝐷−0.91 (95.7%) 42.94𝐷−1 (94.7%) 0.3328 0.324 0.25 0.24 N/A 0.17 0.1561
- floating∗ 0.42∗ 0.55∗ 0.641∗ 0.496∗
2.1. Rotor nacelle assembly
2.1.1. Rated power
The power produced by the wind turbine depends on several parameters, including the wind speed, the
rotor diameter, and power generation efficiency. From the wind turbine design perspective, the amount of
wind energy that passes through the turbine is directly related to its swept area. Therefore, theoretically, the
wind turbine rated power should have a quadratic dependence on the rotor diameter, which can be clearly seen
from Equation (1):

1 1 𝜋𝐷2 3
𝑃 = 𝐶𝑃 (𝜆, 𝛽)𝜌𝑎 𝐴𝑡 𝑈𝑟3 = 𝐶𝑃 (𝜆, 𝛽)𝜌𝑎 𝑈 , (1)
2 2 4 𝑟
𝑃 ∼ 𝐷2 ,

where 𝐶𝑃 (𝜆, 𝛽) is the power coefficient that depends on the tip speed ratio 𝜆, and the blade pitch angle 𝛽, 𝜌𝑎 is
the air density, 𝐴𝑡 = 𝜋𝐷2 ∕4 is the wind turbine swept area, and 𝑈𝑟 is the wind speed at the nacelle. However,
the best power fit approach demonstrates a lower scaling exponent of 1.79 (refer to Figure 2a and Table 1).
A similar analysis for smaller commercial wind turbines (< 7 MW) [22] also demonstrated that the rated
power grows slower (𝐷1.8 ) than theoretically predicted (𝐷2 ). There are multiple reasons that might cause
this difference: nonlinear structural effects relevant to large multi-MW turbines [36], wake-induced effects
that have a strong variation with the azimuth [37], design choices for different specific power levels, design
constraints associated with allowable noise levels [38], etc. However, for floating offshore wind applications,
the power to weight ratio is a more important parameter to consider as the weight of the rotor, nacelle and
tower is highly correlated with the cost of energy from wind turbines.

2.1.2. Rated wind speed and tip speed trends

The rated wind speed is usually considered as a site-dependent design parameter that directly affects the
annual energy production of a wind turbine [38]. Therefore, it can be seen from Table 1 that the rated wind
speed does not depend on the turbine size and just a marginal decrease from 11.4 to 10.6 m/s (< 10%) can
be noticed for larger wind turbines. However, more powerful wind turbines exploit higher tip speeds, and an
increase of 30% can be noticed between 1.5 MW and 20 MW wind turbines. Increased tip speeds result in
a lower torque, less mass, and reduced costs of tower top systems [39]. As a result, the rotational speed of
the rotor decreases with the turbine rated power (inversely proportional to the turbine diameter) in order to
maintain the same level of the tip speed ratio of about 6 − 8 [40, 41] without affecting the turbine structural
integrity and its noise emission [38]. So the expected scaling of the rotor speed with turbine size can be seen
from the definition of the tip speed ratio:

Ω𝐷
𝜆= , 𝜆(𝐷) ≈ const, 𝑈𝑟 (𝐷) ≈ const, (2)
2𝑈𝑟
Ω ∼ 𝐷−1 ,

where Ω is the rotor speed, 𝜆 is the tip speed ratio, and 𝑈𝑟 is the wind speed at the nacelle.

2.1.3. Thrust force

The rotor thrust force is one of the key parameters that determine the tower bending loads [21, 22] for the
fixed-bottom support structures. In case of the floating foundations, the thrust force also affects the mean surge
position of the platform with respect to the equilibrium, and the peak rotor thrust dictates the requirements
for platform stability (for more details see Section 3.1.1).
The thrust force reaches its maximum value around the rated wind speed regardless the wind turbine size.
Similarly to the power, it is expected that the thrust force has a quadratic dependence on the rotor diameter
according to Equation (3):

1 1 𝜋𝐷2 2
𝐹𝑡 = 𝐶𝑇 (𝜆, 𝛽)𝜌𝑎 𝐴𝑡 𝑈𝑟2 = 𝐶𝑇 (𝜆, 𝛽)𝜌𝑎 𝑈 (3)
2 2 4 𝑟
Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 6 of 24
 𝐹𝑡 ∼ 𝐷2 ,

where 𝐶𝑇 denotes the thrust coefficient. However, according to the best-fit approach shown in Figure 2b, the
thrust force has a scaling exponent of 1.86.

2.1.4. Rotor and nacelle mass

For wind energy applications, weight (mass of the blades, nacelle, and tower) is commonly used as a
proxy for the cost of energy [21, 22, 42]. According to the scaling trends identified in Table 1, all mass
characteristics of the RNA and tower have an almost quadratic dependence on the rotor diameter (𝑚 ∼ 𝐷2 )
despite the expected cubic scaling factor which, in general, applies to mass scaling. Initial scaling models of
the blade designs indicated that blade mass can be scaled with the near-cube of rotor diameter [43]. However,
commercial blade designs have always maintained a scaling exponent lower than 2.4 [43–45]. This became
possible due to the use of advanced high-performance materials (i.e. carbon-fibre), improved aerodynamic
performance, and structural design (change of the aerofoil) [46]. Moreover, for wind turbines with rated power
higher than 2 MW, the bending loads on the blade are dominated by the gravity and inertia loads rather than
aerodynamic loads [43]. As a result, the analysis in Table 1 indicates that for larger wind turbines, the blade
mass scales as 𝐷1.83 .
Similarly to the blade mass, the nacelle mass has a near-quadratic dependence on the rotor diameter as
shown in Figure 2c. The mass of the nacelle is mainly composed of the mass of generator, power electronics,
and drivetrain. Thus, the nacelle mass is directly dependent on the choice of the drivetrain design, with or
without a gearbox. Traditionally, a gearbox is used between the turbine rotor and the electrical generator to
increase the rotational speed from approximately 5–20 rpm to 1,000–1,800 rpm. As a result, geared wind
turbines allow the use of high-speed, low-torque generators that are lightweight and inexpensive [47]. As
demonstrated in Table 1, wind turbines with a power rating higher than 10 MW use low-speed directly driven
electric generator systems that do not require a gearbox [48]. The use of gearless configurations prevents
gearbox failure, which is critical for offshore applications where maintenance costs are much higher than for
onshore wind turbines [47]. However, the weight and cost of direct-drive turbines are generally considered to
be higher than that of the gearbox-based turbines while also requiring the use of rare-earth elements [49].
Nevertheless, recent advances in the low-speed direct-drive synchronous generators have demonstrated
that both technological solutions can have a similar mass, which is clearly indicated in Figure 2c where the
scaling law does not seem to be affected by the drivetrain design: gearbox or direct drive (black and gray
circles on Figure 2c respectively).
There are a number of studies that have identified dependencies, similar to those reported in Table 1, but for
smaller wind turbines of up to 7 MW [19, 21, 22, 50]. Based on the reference turbines studied here, the basic
scaling laws for power, thrust force, and masses (except the nacelle mass) remain roughly the same for larger
wind energy systems up to 20 MW. The analysis of scaling trends identified within the Upwind project (as of
2012) [19, 21] found that the nacelle mass of wind turbines with a rotor diameter larger than 80 m (2–5 MW)
follows a cubic scaling rather than a quadratic one. Moreover, further economic analysis [21, 22] show that
with this trend, it would be uneconomical to increase the size of wind turbines above 5 MW. However, the
scaling exponent identified within this study for reference wind turbines of up to 20 MW is close to 1.88. This
shows that the wind turbine and platform developers have already taken into account potential technology
advancements when designing wind turbines > 10 MW.

2.2. Tower design

In contrast to the RNA, the properties of the tower usually depend on the installation site, and the type of
the support structure used to carry the rotor-nacelle assembly (fixed-bottom monopiles, gravity based, space-
frames, or floating structures). Moreover, the tower should be designed such that its first natural frequency
does not coincide with the main sources of the wind system excitation: the turbine rotational speed (commonly
referred to as 1P) and the rotor blade passing frequency (three times the rotational speed = 3P), noting that the
relevant excitation can be extended to 6P, 9P, etc. Therefore, three solutions have been proposed for the tower
design [51] as shown in Figure 3: (i) soft-soft frequency interval where the tower is designed to be ‘soft’ such
that its natural frequency is less than 1P; (ii) soft-stiff – a natural frequency is between 1P and 3P; and (iii)
stiff-stiff – the fundamental natural frequency is above 3P. It should be noted that the turbine rotational speed

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 7 of 24
is not constant and varies depending on the wind speed leaving a narrower interval for the soft-stiff design. In
addition, once the wind turbines are installed offshore, another significant source of excitation starts playing
an important role – ocean waves, which are stochastic in nature, with wave frequency components ranging
from 0.05 Hz to 0.2 Hz, equivalent to wave periods from 5 to 20 s.

Figure 3: Frequency intervals for the tower design from soft to stiff for the three-bladed variable speed wind
turbine, where the thickness represent the range of frequencies that may be excited by the wind turbine. 1P
refers to the turbine rotational speed; 3P is three times the rotor frequency, also called the rotor blade passing
frequency, or the frequency at which the blades pass by the tower; 6P is the twice the blade passing frequency.

Figure 3 demonstrates the possible tower design solutions for only one wind turbine. By extending Figure 3
to a range of wind turbines rated between 2 and 20 MW, and displaying tower natural frequencies from Table 1,
it is possible to clearly identify the trends and differences between tower designs for bottom-fixed and floating
offshore installations. As a result, Figure 4 demonstrates (i) the frequency intervals that should be avoided by
the tower designers: 1P, 3P, 6P and the wave excitation ranging from 0.05 Hz (wave period 20 s) to 0.2 Hz
(wave period 5 s); (ii) first natural frequencies (fore-aft) of the towers designed for the fixed-bottom offshore
wind turbines (marked by crosses); and (iii) first natural frequencies of the towers designed for floating offshore
wind turbines (marked by asterisk).
It is clear that fixed-bottom designs follow ‘soft-stiff’ approach where the tower natural frequency is
placed in the area between 1P (red) and 3P (dark gray). However, the soft-stiff frequency interval starts
overlapping with the wave excitation frequencies for larger turbines (> 15 MW), and this design approach is
not applicable anymore for floating wind applications. As an example, designers of the UMaine VolturnUS-S
semi-submersible platform that supports an IEA 15 MW wind turbine decided to stiffen the tower and place
its natural frequency above 3P, leading to 0.496 Hz [35]. In this way the mass of the tower increased by 47%
as compared to the monopile configuration. Note that this mass increase may be accompanied by a slight
reduction of the platform restoring stiffness in pitch and roll, and have effects on the stability and response
of the platform, which are discussed in the following section. Another reason for not choosing a ‘soft-stiff’
tower design for floating applications could be due to its unsatisfactory long term fatigue performance [52],
requiring the ‘stiff-stiff’ solution. According to Table 1, the increase in tower rigidity is mainly achieved by
thickening its walls while maintaining its diameter and height.
In Table 1, the tower natural frequencies for fixed-bottom applications have been calculated assuming a
‘fixed-free’, or clamped, cantilever system, while for towers installed on a floating foundation, a ‘free-free’
cantilever system has been used. It is important to note, that even for the same tower design, the natural
frequency of the first tower fore/aft bending mode will be higher if calculated using the ‘free-free’ condition
as compared with the ‘fixed-free’ approach [53–55].
The structural design of the tower, in particular its wall thickness and diameter, are determined by the
target natural frequency, and current manufacturing and installation technology [56]. It is expected that all
linear dimensions of the tower would scale almost linearly with increasing rotor diameter, which is supported
by the scaling laws identified in Table 1. Regarding the tower height, the towers for offshore wind turbines are

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 8 of 24
 OO-Star
INO WINDMOOR
VolturnUS-S
OC4-DeepCwind

NAUTILUS-10

Figure 4: Range of wind turbine excitation frequencies (1P, 3P, 6P and wave) and tower natural frequencies
depending on the rotor diameter. The dashed line corresponds to the best power fit for the natural frequencies
of the fixed-bottom towers identified in Table 1 𝑓 = 26.4𝐷−0.91 .

generally designed as short as possible [9], with a requirement to maintain at least 25 m blade tip clearance
with water surface to allow ship passage [57]. Therefore, data in Table 1 and Figure 2d demonstrate that the
lowest point of the passing blades is always approximately 30 − 31 m above the sea water level (Figure 5).

3. Floating platforms
Great emphasis in the development of floating support structures was given to the platforms designed
for the 5-MW wind turbine. These include a spar-buoy concept [58], a semi-submersible platform [59, 60],
a barge [61], and a tension leg platform [62]. With the increase in the turbine size, it became necessary to
develop new support structures that provide stability to the wind energy systems with rated power of higher
than 5 MW. Since the semi-submersible structure has been favored by many platform developers due to its
relatively low draught resulting in ease of assembly, towing, and commissioning [9], this paper focuses on
this platform topology.

3.1. Design practices of floating wind turbines

The design of floating support structures is directly related to the parameters of the supported wind turbine
and tower (i.e. linear dimensions, mass and moment of inertia, maximum overturning moment caused by the
rotor thrust force, etc.). Thus, DNV GL standards [63, 64] and recommended practices (i.e. [55, 65]) have
been developed to guide the design, construction and in-service inspection of floating wind turbines. Collu
and Borg [66] summarised the main principles and technical requirements as follows:
• Floatability. The buoyancy force of the platform should be sufficient to counteract the total weight of the
entire system (platform, tower, wind turbine) plus other downward forces including mooring forces.

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 9 of 24
 Figure 5: The vertical position of the turbines of different sizes in relation to the wind profile.

• Maximum inclination angle. It is desired that the wind turbine operates close to the upright position, and
the total angle of the FOWT inclination, caused by static (wind) and dynamic (wave) loads, should not
exceed a particular value. Even though all the literature refers to the value of 10 degrees as the maximum
allowed inclination angle (in roll and/or pitch), such a number cannot be found in the recent standards [64].
• Freeboard height and minimum draught. The minimum vertical distance from the sea water level to the
top of the support structure should be > 10 m, and the minimum draught should be > 15 m.
• Optimum dynamic response to wind and wave forces. Any oscillatory motion of the FOWT should be
minimised.

3.1.1. Stability
One of the key design criteria to maintain the platform stability during the operation of the wind turbine is
related to the restoring moments against pitch and roll motions. In order not to exceed the maximum allowable
static angle of inclination caused by the largest rotor thrust, the platform’s hydrostatic stiffness in pitch and
roll should satisfy the following criteria [66]:

𝐹𝑡max 𝑧ℎ𝑢𝑏
𝐾55 ≥ , (4)
𝜃 max
where 𝐾55 is the total restoring stiffness of a FOWT, 𝐹𝑡max is the maximum thrust force, 𝑧ℎ𝑢𝑏 is the height of
the wind turbine hub above the still water level, 𝜃max is the maximum allowable pitch angle.
Thus, the stability of the floating structure is directly related to the wind turbine’s performance, in
particular, the largest rotor thrust [64]. Usually, as shown in Figure 6a, the thrust force reaches its maximum
value around the rated wind speed regardless the wind turbine size. In order to reduce the maximum steady
state thrust force, the ‘peak shaving’ or ‘thrust clipping’ technique [67–69] is used to modify the blade pitch
controller. This technique allows the maximum thrust force to be reduced by up to 20%, thus easing the
requirements for the platform stability.
The total restoring stiffness of the floating platform in pitch and roll has three main components: the water
plane area, buoyancy, and mooring system [63, 66]:

𝐾55 = 𝜌𝑤 𝑔𝐼𝑦 + (𝐹𝑏 𝑧𝐶𝑜𝐵 − 𝑚𝑔𝑧𝐶𝑜𝐺 ) + 𝐾𝑚𝑜𝑜𝑟,55 , (5)

⏟⏟⏟ ⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏞⏟ ⏟⏟⏟
waterplane B-G relative position mooring

where 𝜌𝑤 is the water density, 𝑔 denotes the gravitational acceleration, 𝐼𝑦 is the waterplane second moment
of area (or area moment of inertia), 𝐹𝑏 is the buoyancy force, 𝑧𝐶𝑜𝐵 is the vertical coordinate of the centre of

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 10 of 24
 (a) (b)
Figure 6: (a) thrust force vs. wind speed for different turbine sizes, (b) thrust force with and without peak shaving.

buoyancy, 𝑧𝐶𝑜𝐺 is the vertical coordinate of the centre of gravity, 𝑚 is the total mass of the system, 𝐾𝑚𝑜𝑜𝑟,55
is the restoring stiffness in pitch caused by the mooring system.
The waterplane area is the main source of stability for the barge and semi-submersible platforms, the
buoyancy caused by the relative position of the centre of gravity and buoyancy forces is dominant for spar
platforms, and the mooring system is utilised to maintain stability of the tension leg platforms. As this paper
focuses on the semi-submersible floating structures, the restoring stiffness in pitch is dominated by the first
term in Equation (5):

𝐾55 ≈ 𝜌𝑤 𝑔𝐼𝑦 . (6)

In order to understand which geometric parameters play the key role in the design of a semi-submersible
platform, its second moment of area can be approximated as (assuming three offset columns):

( )
3 𝑑4 3
2 2
𝐼𝑦 ≈ 𝜋 𝑅 𝑑 + ≈ 𝜋𝑅2 𝑑 2 , (7)
8 8 8

where 𝑅 is the radius from the geometrical centre of the platform to the centre of the offset column, and 𝑑 is
the diameter of the upper part of the offset column.
By substituting Equation (7) into Equations (6) and (5), it is possible to identify the requirements to the
linear dimensions of the FOWT as:

√
𝐹𝑡max 𝑧ℎ𝑢𝑏 3
𝑅𝑑 ≥ . (8)
𝜃 max 8𝜋𝜌𝑤 𝑔

As a result, the main stability criterion that governs the design of a semi-submersible platform is the
product of the distance to the offset columns and their diameter (𝑅𝑑) that should be controlled to maintain
the required restoring stiffness of the FOWT in pitch and roll.

3.1.2. Dynamic response and natural frequencies

Floating offshore wind turbines have six additional degrees-of-freedom (DOF) as compared to their fixed-
bottom counterparts due to a soft coupling with a seabed, making the dynamics of the floating foundation
an important design criterion. The characteristic excitation frequencies of the wind and wave loads on the
FOWT are demonstrated in Figure 7. Turbulent wind is represented by the Kaimal spectrum [70] with a mean
wind speed of 11.4 m/s at a reference hub height (100 m) and turbulence intensity of 0.12. Ocean waves are

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 11 of 24
represented by the JONSWAP spectrum [71] with a significant wave height of 6 m, peak wave period of 10 s,
and peak enhancement factor of 2.87 [71]. Figure 7 clearly demonstrates that the turbulent wind excites the
low-frequency response of the system with a turbulent peak occurring around 0.01–0.03 Hz (30–100 s), while
the wave excitation usually occurs in the frequency range of 0.05 and 0.25 Hz (4–20 s).
Therefore, in order to minimise the oscillatory motion of the FOWT under the wind and wave loads, the
natural frequencies of the floating foundation should be selected with caution [63]. First of all, the platform
should be designed such that its natural frequencies in all six DOFs lie outside the frequency range of the first-
order wave excitation. In addition, if the natural period of heave approaches half the natural period of pitch or
roll, the platform may experience Mathieu instability [72, 73] resulting in large heave motions. As the natural
frequencies of a semi-submersible platform in heave and pitch are mostly related to its hydrostatic stiffness and
inertial properties, they can be controlled through altering the platform dimensions. The natural frequency in
surge depends on the mooring system. Usually, a semi-submersible platform has the surge natural frequency
in the low-frequency range (≈ 100 s), so the platform experiences slow motion due to the low-frequency loads
from the wind turbulence [74], difference-frequency loads from waves [75], and the mean drift force [76]. As
a result, the mooring system has to be stiff enough to to keep the system’s peak surge-sway offset under a
certain level [35], but soft enough to allow the inertial loads to take up the wave loads.

(a) (b)
Figure 7: Wind (Kaimal) and wave (JONSWAP) power spectral densities. Two different representations:
(a) common in naval engineering; (b) common in aerodynamics and wind engineering applications (logarithmic
scale).

To illustrate how these design practices are applied in the development of semi-submersible platforms, an
overview of existing prototypes is presented below.

3.2. Existing prototypes of semi-submersible platforms

The key concept of any semi-submersible design is to achieve static stability using large waterplane area
and distributed surface buoyancy. Different semi-submersible structures developed to support 5, 10, 12 and
15-MW wind turbines proposed in the literature are summarised in Table 2, including:
• platforms for a 5-MW wind turbine:
– WindFloat [77–80] developed by Princliple Power, Inc., US;
– OC4-DeepCwind [59] developed for the US-based DeepCwind project (collaboration between
National Renewable Energy Laboratories, US, University of Maine, US, and Norwegian Univer-
sity of Science, Norway);
– CSC-semisub [81, 82] prototype designed by the Norwegian University of Science and Techno-
logy;
– GustoMSC Trifloater [83, 84] developed by GustoMSC, Netherlands;
• platforms for a 10-MW wind turbine:
– OO-Star [33, 34, 85] and

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 12 of 24
 – NAUTILUS-10 [33] are the two platforms developed by the LIFES50+ consortium within EU-
funded project Horizon2020;
• the INO WINDMOOR 12 MW floating wind turbine [29] developed by a consortium led by SINTEF
Ocean AS, Trondheim, Norway;
• the UMaine VolturnUS-S reference platform developed for the IEA Wind 15-MW offshore reference
wind turbine [35] by the University of Maine and National Renewable Energy Laboratories, US.

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 13 of 24
Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 14 of 24

Table 2
Proposed designs of semi-submersible platforms for different turbine sizes.
5-MW 10-MW 12 MW 15-MW
Platform OC4- INO
WindFloat CSC GustoMSC OO-Star NAUTILUS-10 VolturnUS-S
DeepCwind WINDMOOR
[77–80] [59] [81, 82] [83, 84] [33] [33] [29] [35]
Hull displacement, m3 6,931 13,917 10,297 3,541 23,509 9,280 13,830 20,206
Platform mass (incl. ballast), t 6,251 13,473 9,730 2,920 21,709 7,781 11,974 17,854
Type of ballast active passive passive passive – active passive passive
Platform material steel steel steel steel concrete steel steel steel
Platform draught, m 23 20 30 13.2 22 18.3 15.5 20
Offset column diameter, m 10.7 12 6.5 8 13.4 10.5 15 12.5
Waterplane area, m2 270 372 133 192 537 346 530 447
Radius to column centre, m 33 28.9 41 36 37 38.7 35.2 51.75
Platform mass/FOWT mass 88% 96% 92% 80% 90% 82% 84% 86%
Rigid-body natural period, s
- surge 99.8 113 79.5 74 181.8 125 97.3 142
- heave 19.9 18 26 16 20.5 19 16.3 20.4
- pitch 43.3 27 31.3 29 31.3 30.3 31.4 28
- yaw 80 58 63 116 100 88 91
Platform image
 In order to identify trends and relate the platform designs to the characteristics of supported wind turbines,
several parameters from Table 2 are displayed in Figure 8 vs. the turbine rated power. Solid circles in Figure 8
correspond to the original designs listed in Table 2, while the hollow circles correspond to those studies where
rational up-scaling has been applied to develop platforms for larger wind turbines based on existing 5-MW
semi-submersible designs ([14, 16, 17, 86–89]). The reader is referred to Section 4 for more information on
rational up-scaling procedures suggested in the literature.

(a) (b) (c)

(d) (e) (f)

Figure 8: Properties of various semi-submersible platforms designed to support wind turbines with different power ratings:
(a) platform displaced volume, (b) waterplane area, and (c) product of the radius to the offset column and offset column
diameter, natural periods of the FOWT in (d) surge, (e) heave and (f) pitch. Solid circles represent the original designs
developed for a particular wind turbine, while the hollow circles correspond to those studies where rational up-scaling have
been applied to design platforms for larger wind turbines based on existing 5-MW semi-submersible prototypes. The shaded
area in (e) indicates a range of periods excited by ocean waves. Best-fit in (a), (b) and (c) has been identified for the
original designs only (solid circles).

The data in Table 2 demonstrate that even for the same wind turbine rated power (i.e. 5-MW), there is a
large variety of semi-submersible designs, depending on the number of columns, presence of heave plates,
placement of the wind turbine, and presence of pontoons/bracings. According to Figures 8a and 8b, this
has resulted in more than twofold difference in displacement and waterplane area between different design
solutions for the 5-MW wind turbine. In order to demonstrate the validity of findings presented in Section 3.1,
the best-fit curves were added to Figures 8a and 8b, where both the platform displacement and waterplane
area were related to the wind turbine rotor diameter. Low values of the goodness-of-fit (R2 ) show that neither
the hull displacement, nor waterplane area are fundamental criteria that drive the platform design. Instead,
Figure 8c shows a strong correlation between the rotor diameter 𝐷 and the product of 𝑅𝑑 for all floating
platforms. Keeping in mind that the unit of 𝑅𝑑 is [m2 ] and the unit of 𝐷 is [m], the scaling exponent of
1.09 means that the platform linear dimensions should not increase proportionally to the rotor diameter to
maintain the same level of system stability. Referring
√ back to the derived Equation (8), it should be expected
that 𝑅𝑑 scales as 𝐷1.5 (𝐹𝑡 ∼ 𝐷2 , 𝑧ℎ𝑢𝑏 ∼ 𝐷, so 𝐹𝑡 𝑧ℎ𝑢𝑏 ∼ 𝐷1.5 ). However, the identified scaling exponent
of 1.09 is lower than the anticipated value of 1.5. This might be explained by the fact, that semi-submersible
platforms designed for 12 and 15-MW wind turbines do not rely only on the restoring stiffness associated with
waterplane area, but also utilise the gravity-buoyancy relative position (see Equation (5)). In addition, unlike

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 15 of 24
the wind turbine technology, the floating platform is still under development, hence, the differences between
the platforms designs are more pronounced compared with the wind turbine designs.
Regarding the platform draught, all platform developers try to keep the draught close to 20 m regardless
the size of the supported turbine. Interestingly, despite the large variations in the platform mass and
waterplane areas, the platforms’ natural periods in both heave and pitch remain similar for all turbine sizes, at
approximately 20 s for heave (Figure 8e) and 30 s for pitch (Figure 8f). Also, from Table 2, the contribution of
the platform to the total mass of the system tends to decrease from 90-95% for the 5-MW turbines to 80-90%
for higher rated turbines of 10 and 15 MW. This is partially caused by the stiffening of the tower for the floating
support structures, as explained in Section 2, reducing the contribution of the platform in the total mass of
the system. Moreover, it is important to note that some platform developers try to decrease the platform mass
by using an active ballast system as implemented in NAUTILUS-10 design [33].
Based on such a sparse data set, it is difficult to draw a conclusion on how the floating platform design
changes for larger wind energy systems. Also, it clearly demonstrates that developers require more time
to converge to a specific design. While the economically feasible platform design remains uncertain, the
upscaling procedure to support larger turbines is still important for predicting the system response and
performance, which can be used as a first estimate.

4. Review of platform scaling laws suggested in the literature

There are a number of studies that have attempted to rationally upscale 5-MW platforms to support larger
turbines (7.5 and 10 MW). The choice of the geometric scaling factor depends on which key properties of
the turbine/platform system are preserved. As the floating offshore wind turbine experiences loads from both
wind and wave environments, it is important to understand how the scaling of the platform is related to the
scaling of the rotor. More importantly, how the ratio between these two parameters affects the loads, mass and
stability of the entire system.
The length scale of the wind turbine is defined by the rotor diameter 𝐷, while the length scale of the
platform can be defined by some characteristic length 𝐿 (i.e. column radius, distance to offset columns from
the platform center, draught, etc.), as illustrated in Figure 1. With increasing size, the wind energy system
scales with 𝑠, while the floating support structure scales with 𝑝, where the scaling factors 𝑠 and 𝑝 can be
defined according to:

𝐷𝑠𝑐𝑎𝑙𝑒𝑑 𝐿𝑠𝑐𝑎𝑙𝑒𝑑
𝑠= , 𝑝= . (9)
𝐷𝑏𝑎𝑠𝑒 𝐿𝑏𝑎𝑠𝑒

As explained in Section 2, power, mass, and maximum thrust force of the wind turbine scale with 𝑠2 ,
while the overturning moment (force × height) scales with 𝑠3 . Regarding the platform scaling factor, 𝑝, the
displacement (mass) is scaled with 𝑝3 , wave excitation forces in heave and surge scale with 𝑝2 , and the pitching
moment increases proportionally to 𝑝3 [90]. However, the ratio between 𝑠 and 𝑝 defines the dynamics of the
upscaled floating offshore wind energy system. Therefore, in order to upscale an existing FOWT (base) to
support a larger wind turbine (scaled), three different scaling laws for the platform design have been proposed
in the literature and summarised in Table 3.
Power [14, 87]. According to this scaling law, the linear dimensions of the rotor and the platform increase
proportionally for larger turbines. Therefore, the scaling factor for the platform can be calculated as:

√
𝑃𝑠𝑐𝑎𝑙𝑒𝑑
𝑝=𝑠≈ . (10)
𝑃𝑏𝑎𝑠𝑒

Thus, forces and moments acting on the floating platform from both wind and waves also increase proportion-
ally as 𝑠2 and 𝑠3 , respectively. However, the mass of the platform, including ballast, increases faster (𝑠3 ) than
the mass of the supported RNA and tower (𝑠2 ) for larger turbines. As a result, the platform’s contribution to
the total mass of the floating wind system will grow, affecting the cost.

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 16 of 24
 Mass [17, 86, 89]. This scaling factor has been proposed to preserve the ratio between the turbine and
platform masses using the following square-cube law:

√
3 𝑚𝑠𝑐𝑎𝑙𝑒𝑑
𝑝= = 𝑠2∕3 . (11)
𝑚𝑏𝑎𝑠𝑒
According to this scaling law, the wave excitation will have less impact on the platform dynamics than the
aerodynamic loads from the rotor (refer to Table 3).
Fixed draught [15–17, 91]. This scaling law is also designed to preserve the mass properties of the
turbine/platform system but without changing the draught of the floating support due to the constraints in the
port facilities. Therefore, the horizontal (𝑥𝑦) and vertical (𝑧) linear dimensions of the platform scale differently
as:

⎧ √
𝑚𝑠𝑐𝑎𝑙𝑒𝑑
⎪𝑝𝑥𝑦 ≈ = 𝑠,
⎨ 𝑚𝑏𝑎𝑠𝑒 (12)
⎪𝑝𝑧 = 1.
⎩
It is not straight forward to evaluate how wave loads scale with 𝑠, but some indicative numbers are presented
in Table 3 and discussed below.
In Table 3, the approximate scaling factors for the floating platform’s displacement, excitation force, added
mass and hydrostatic stiffness have been populated based on the definition of each of these physical quantities
[90]. Moreover, these scaling factors have been used to provide a rough estimate of expected changes (increase
or decrease) in the natural frequencies and response amplitude operators of the upscaled FOWTs compared
with a baseline design. Thus, the resultant scaling factors for the platform natural frequencies in surge, heave
and pitch have been evaluated using the simplified equation:
√
𝐾
𝜔𝑛 = , (13)
𝑚 + 𝐴(𝜔𝑛 )
and the response amplitude operator (RAO) for each scaling law for each degree-of-freedom has been
calculated (neglecting coupling between different DOFs) based on [92]:
𝐹𝑒𝑥𝑐 (𝜔)
RAO(𝜔) = , (14)
−𝜔2 (𝑚 + 𝐴(𝜔)) + 𝑖𝜔𝐵(𝜔) + 𝐾
where 𝐹𝑒𝑥𝑐 is the wave excitation force, 𝑚 is the total mass of the FOWT, 𝐴(𝜔) and 𝐵(𝜔) are the frequency-
dependent added mass and radiation damping coefficients respectively, and 𝐾 = 𝐾ℎ𝑠 + 𝐾𝑚𝑜𝑜𝑟 is the total
restoring stiffness of the floating platform including the hydrostatic stiffness 𝐾ℎ𝑠 and the stiffness provided
by the mooring system 𝐾𝑚𝑜𝑜𝑟 .
Using heave RAOs as an example, the following analysis demonstrates how the resulting scaling factors
from Table 3 for the natural frequencies and RAOs have been obtained for different scaling laws:
- power platform scaling
Using data from Table 3, the wave excitation force scales as 𝑝2 , mass and added mass scale as 𝑝3 , the
radiation damping coefficient scales as 𝑝2.5 [90, 93] and the hydrostatic stiffness scales as 𝑝2 . Then,
following Equation (14) and substituting the corresponding scaling factors for each physical quantity,
the RAO of the scaled platform can be approximated as:

| |
|RAO𝑠𝑐𝑎𝑙𝑒𝑑 (𝑝−0.5 𝜔)| =
| |
| 𝑝2 𝐹𝑒𝑥𝑐 |
| |
=| |
| −(𝑝−0.5 𝜔)2 (𝑝3 𝑚 + 𝑝3 𝐴) + 𝑖𝑝−0.5 𝜔𝑝2.5 𝐵 + 𝑝2 𝐾 |
| |
Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 17 of 24
Table 3
Review of proposed scaling laws for floating platforms.
Rotor diameter 𝑠
Wind turbine Rated power 𝑠2
Rotor mass 𝑠2
Tower mass 𝑠2
Wind thrust force 𝑠2
Overturning moment 𝑠3
Platform scaling Power Mass Fixed draught
Reference [14, 87] [17, 86, 89] [15–17, 91]
√ √
√ 𝑚𝑠𝑐𝑎𝑙𝑒𝑑
𝑃𝑠𝑐𝑎𝑙𝑒𝑑 𝑚 𝑝𝑥𝑦 ≈
Scaling factor 𝑝= 𝑝 = 3 𝑠𝑐𝑎𝑙𝑒𝑑 𝑚𝑏𝑎𝑠𝑒
𝑃𝑏𝑎𝑠𝑒 𝑚𝑏𝑎𝑠𝑒
𝑝𝑧 = 1
𝑝𝑥𝑦 = 𝑠
Platform length scale 𝑝=𝑠 𝑝 = 𝑠0.67
𝑝𝑧 = 1
Platform displacement 𝑝 3 = 𝑠3 𝑝3 = 𝑠2 ≈ 𝑝2𝑥𝑦 𝑝𝑧 = 𝑠2
Waterplane area and 𝑅𝑑 𝑝2 = 𝑠2 𝑝2 = 𝑠1.33 ≈ 𝑝2𝑥𝑦 = 𝑠2
Wave excitation force
Floating platform

surge 𝑝 2 = 𝑠2 𝑝2 = 𝑠1.33 ≈ 𝑝𝑥𝑦 𝑝𝑧 = 𝑠

heave 𝑝 2 = 𝑠2 𝑝2 = 𝑠1.33 ≈ 𝑝2𝑥𝑦 = 𝑠2
pitch 𝑝 3 = 𝑠3 𝑝3 = 𝑠2 ≈ 𝑝2𝑥𝑦 𝑝𝑧 = 𝑠2
Added mass (diagonal)
surge 𝑝 3 = 𝑠3 𝑝3 = 𝑠2 ≈ 𝑝𝑥𝑦 𝑝𝑧 = 𝑠
heave 𝑝 3 = 𝑠3 𝑝3 = 𝑠2 ≈ 𝑝2𝑥𝑦 = 𝑠2
pitch 𝑝 5 = 𝑠5 𝑝5 = 𝑠3.33 ≈ 𝑝2𝑥𝑦 𝑝𝑧 = 𝑠2
Hydrostatic stiffness
heave 𝑝 2 = 𝑠2 𝑝2 = 𝑠1.33 ≈ 𝑝2𝑥𝑦 = 𝑠2
pitch 𝑝4 = 𝑠4 𝑝4 = 𝑠2.67 ≈ 𝑝3𝑥𝑦 𝑝𝑧 = 𝑠3
Natural frequency
surge mooring system dependent
heave 𝑝−0.5 = 𝑠−0.5 𝑝−0.5 = 𝑠−0.33 ≈1
pitch 𝑝−0.5 = 𝑠−0.5 𝑝−0.5 = 𝑠−0.33 ≈ 𝑠0.5
RAO (noting that the peak response will shift to the lower frequency range following 𝑝−0.5 )
surge 1 1 ≈ 𝑠0.5 − 𝑠1
heave 1 1 ≈ 𝑠0.5 − 𝑠1
pitch −1
𝑝 =𝑠 −1 −1
𝑝 =𝑠 −0.67 ≈ 𝑠0.5

| 𝑠2 𝐹𝑒𝑥𝑐 |
| |
=| |
| −(𝑠−0.5 𝜔)2 (𝑠3 𝑚 + 𝑠3 𝐴) + 𝑖𝑠−0.5 𝜔𝑠2.5 𝐵 + 𝑠2 𝐾 |
| |
| 𝑠2 𝐹 |
= || 𝑒𝑥𝑐 | ≈ |RAO𝑏𝑎𝑠𝑒 (𝜔)| .
| | |
| 𝑠2 −𝜔2 (𝑚 + 𝐴) + 𝑖𝜔𝐵 + 𝐾 |
| |
|RAO𝑠𝑐𝑎𝑙𝑒𝑑 (𝑠−0.5 𝜔)| ≈ ||RAO𝑏𝑎𝑠𝑒 (𝜔)|| .
| |
It should be noted that for this simplified analysis, the wave frequency has also been scaled as 𝑝−0.5 .
The wave frequency at a particular test site does not change with increasing platform size. However,
this approach allows us to predict whether the peak response of the scaled platform will increase or
decrease compared with a baseline design, and how exactly it is shifted to the lower frequency range.
- mass platform scaling
| |
|RAO𝑠𝑐𝑎𝑙𝑒𝑑 (𝑠−1∕3 𝜔)| =
| |
Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 18 of 24
Table 4
Scaling factors used to increase a 5-MW OC4-DeepCwind semisubmersible platform to support a 10-MW DTU
wind turbine.
Scaling 10 MW FOWT
factor
Power Mass Fixed draft
Rotor, 𝑠 1.41
𝑝𝑥𝑦 = 1.41
Platform, 𝑝 1.41 1.26
𝑝𝑧 = 1

| 𝑠4∕3 𝐹𝑒𝑥𝑐 |
| |
=| |
| −(𝑠−1∕3 𝜔)2 (𝑠2 𝑚 + 𝑠2 𝐴) + 𝑖𝑠−1∕3 𝜔𝑠5∕3 𝐵 + 𝑠4∕3 𝐾 |
| |
| 𝑠4∕3 𝐹 |
= || 𝑒𝑥𝑐 | ≈ |RAO (𝜔)|
| | 𝑏𝑎𝑠𝑒 |
| 𝑠4∕3 −𝜔2 (𝑚 + 𝐴) + 𝑖𝜔𝐵 + 𝐾 |

- fixed draught platform scaling

| |
|RAO𝑠𝑐𝑎𝑙𝑒𝑑 (𝑠−0.5 𝜔)| =
| |
| 𝑠2 𝐹𝑒𝑥𝑐 |
| |
=| |
| −(𝑠−0.5 𝜔)2 (𝑠2 𝑚 + 𝑠2 𝐴) + 𝑖𝑠−0.5 𝜔𝑠2.5 𝐵 + 𝑠2 𝐾 |
| |
( )
≈  𝑠0.5 − 𝑠1 ||RAO𝑏𝑎𝑠𝑒 (𝜔)||

This analysis demonstrates that if the platform’s size increases proportionally in all directions (𝑥𝑦 and 𝑧),
following power or mass scaling laws, it is expected that the platform’s heave RAO will shift to longer wave
periods, without significant change in maximum amplitude. However, if only water-plane area of the platform
increases (fixed-draught scaling law), the platform will experience larger motions in heave.
Similar conclusions can be drawn from the results presented in Table 3 for other degrees-of-freedom. Thus,
the platform upscaled following the power and mass scaling laws will experience similar motion amplitudes
in surge, and will be less responsive to pitch excitation compared with the FOWT baseline design. However,
the platform designed using the fixed-draught scaling law will be more susceptible to wave excitation in
all degrees-of-freedom (scaling factors of 𝑠0.5 − 𝑠1 ). Also, it is expected that the heave and pitch natural
frequencies of the power and mass scaled platforms will decrease with increasing platform size. However, the
opposite trends are expected for the fixed-draught scaling law, as the platform’s pitch natural frequency will
increase with increasing dimensions of the platform.
To verify this analysis, numerical calculations have been performed by comparing three different scalings
(power, mass and fixed-draught). Thus, a 5-MW OC4-DeepCwind semisubmersible platform is upscaled
to support a 10-MW DTU wind turbine by increasing its linear dimensions in accordance with Table 3.
Increasing the wind turbine rated power from 5 to 10 MW, its diameter changes from 126 m to 178 m leading
to the turbine scaling factor of 𝑠 = 178∕126 = 1.41. Based on this value, the corresponding scaling factors for
the platform are outlined in Table 4 and calculated as 𝑝 = 𝑠 = 1.41 for the power scaling law, 𝑝 = 𝑠2∕3 = 1.26
for the mass scaling law, and 𝑝𝑥𝑦 = 𝑠 = 1.41, 𝑝𝑧 = 1 for the fixed-draught scaling law.
The numerical analysis of all FOWTs is performed in the frequency domain using the statistical
linearisation approach [74, 76]. The response of the entire FOWT (which includes the floating platform, tower,
and RNA) is estimated using white-noise excitation in a range of typical wave frequencies 0.05 − 0.25 Hz,
similar to [60].
Figure 9 demonstrates the difference in RAOs for surge, heave and pitch modes of semi-submersible
platforms scaled using three different laws. The peak responses in all modes (surge, heave and pitch) shift to
the lower frequency range. Also, as specified in Table 3, the platform that followed the fixed-draught design
experiences higher motion amplitudes in all degrees-of-freedom. In contrast, the scaling of the platform using

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 19 of 24
 (a) (b)

(c)
Figure 9: RAOs in (a) surge, (b) heave, and (c) pitch of the 5-MW OC4-DeepCwind semisubmersible platform
and its upscaled versions designed to support a 10-MW DTU wind turbine using three different scaling laws:
power (p), mass (m), and fixed draught (fd).

power or mass laws makes the system less responsive in pitch to wave excitation. Also, it should be noted
that the heave natural frequencies of the original and scaled platforms are located within or close to the
wave excitation region, which may present issues in large sea states. This numerical example verifies the
comparative study of three scaling laws and shows that data in Table 3 can be utilised to quickly predict the
changes in the response of the upscaled FOWT depending on the scaling law used.
However, comparing scaling laws proposed in the literature (Table 3) and trends identified in Section 3.2, it
can be concluded that none of the proposed scaling procedures is followed in practice. As already mentioned,
the power scaling law leads to the largest mass of the platform significantly affecting the cost of the entire
system. Moreover, it has been shown in Section 3.2 that the linear dimensions of the platform should not
increase proportionally to the rotor diameter to maintain the same level of platform stability. In terms of the
platform draught, the fixed-draught scaling law might be the closest one to existing trends as the platform
developers try to maintain a similar level of draught around 20 m. While both mass and fixed-draught scaling
procedures provide satisfactory increase in hydrostatic stiffness close to 𝐷3 (refer to Equation (7)), the value of
𝑅𝑑 scales closer to the identified trends (1.09) with the mass scaling procedure (1.33). Taking into account the
dynamic response of the platform, it seems that the platforms designed following a mass scaling law will not
experience any increase in the motion amplitudes with increasing size of the platform, while the fixed-draught
approach might lead to higher motion amplitudes in all degrees of freedom. As a result, it can be concluded
that the mass scaling procedure might be a better starting point in comparison with power and fixed draught
scaling laws for the platform design and further optimisation to support larger wind turbines.

5. Conclusion
This study combines data from 9 reference wind turbines with power ratings of up to 20 MW, and from 8
semi-submersible platforms developed for the 5-15 MW wind turbines in order to identify trends and scaling
laws related to their physical properties, which can be applied at the conceptual design of offshore semi-
submersible wind turbines. The analysis revealed that the wind turbine mass, power and thrust forces scale
quadractically with the rotor diameter and these characteristics are not really dependent on the drivetrain

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 20 of 24
design (with or without a gearbox). The main challenges associated with larger wind turbines for offshore
applications is related to the tower design. The ‘soft-stiff’ tower design, common for fixed-bottom installations,
are not applicable for floating offshore wind energy systems with power ratings higher than 10-12 MW. This is
mainly due to the fact that the frequency range between 1P and 3P starts overlapping with the wave excitation
frequencies so the tower should be stiffened to have natural frequency above the 3P region. This practice has
been adapted by all FOWT developers with wind turbines > 10 MW.
The review of existing semi-submersible platforms reveal a large variety of design solutions. Regardless
the wind turbine size, all platforms have a draught less than 20 m, and the natural frequencies of the FOWT
in heave and pitch are kept around 20 s and 30 s respectively. The contribution of the platform into the total
mass of the FOWT slightly decreases with the turbine size meaning that novel designs tend to be optimised
to bring the technology to a more mature level and become cost-competitive with fixed-bottom installations.
It has been found that the key geometric parameter that drives the design of any semi-submersible platform
is not related to its waterplane area, but to the product of the offset column diameter and the distance to the
offset column.
In addition, different scaling procedures developed in the literature to upscale platform dimensions for
larger wind turbines have been reviewed and compared with existing practices. The analysis demonstrates
that the mass-scaling procedure is the best starting point when the larger platform is required to support a
turbine with a higher rated power.
The analysis in this paper is based on the currently available data for a range of wind turbines and platforms.
However, it should be noted that the number of data points is relatively small, data are taken from publicly
available reference wind turbines and platforms; and reviewed technologies have different levels of design
maturity. Despite these limitations, the review undertaken in this paper clearly demonstrates the direction in
which industry and researchers are moving, while technology development is still ongoing.

Acknowledgment
The authors acknowledge the funding support of Australia-China Science and Research fund, Australian
Department of Industry, Innovation and Science.

CRediT authorship contribution statement

Sergiienko N.Y.: Conceptualisation, Data curation, Methodology, Software, Formal analysis, Invest-
igation, Visualisation, Validation, Writing - original draft, Writing - review & editing.. da Silva, L.S.P.:
Conceptualisation, Formal analysis, Methodology, Visualisation, Writing - original draft, Writing - review
& editing.. Bachynski-Polić, E.E.: Data curation, Methodology, Formal analysis, Writing - review &
editing.. Cazzolato, B.S.: Conceptualisation, Supervision, Writing - review & editing.. Arjomandi, M.:
Conceptualisation, Supervision, Writing - review & editing.. Ding, B.: Supervision, Writing - review &
editing..

References
[1] Global Wind Energy Council, GWEC| global wind report 2021, Technical Report, Global Wind Energy Council: Brussels, Belgium,
2021.
[2] W. Musial, P. Spitsen, P. Beiter, P. Duffy, M. Marquis, A. Cooperman, R. Hammond, M. Shields, Offshore wind market report:
2021 Edition, Technical Report, National Renewable Energy Lab.(NREL), Golden, CO (United States), 2021.
[3] M. Karimirad, Offshore energy structures: for wind power, wave energy and hybrid marine platforms, Springer, 2014.
[4] IRENA, Renewable capacity statistics 2021, International Renewable Energy Agency (IRENA), Report, 2021. Abu Dhabi.
[5] IEA, World Energy Outlook 2021, Report, International Energy Agency, 2021.
[6] M. S. Genç, F. Karipoğlu, K. Koca, Ş. T. Azgın, Suitable site selection for offshore wind farms in turkey’s seas: Gis-mcdm based
approach, Earth Science Informatics 14 (2021) 1213–1225.
[7] L. Eatough, Floating Offshore Wind Technology and Operations Review, Report, Offshore renewable energy Catapult, 2021.
[8] The Global Wind Atlas, 2020. Accessed on 26/10/2020.
[9] G. E. Barter, A. Robertson, W. Musial, A systems engineering vision for floating offshore wind cost optimization, Renewable
Energy Focus 34 (2020) 1–16.
[10] W. Musial, B. Ram, Large-scale offshore wind power in the United States: Assessment of opportunities and barriers, Technical
Report, National Renewable Energy Lab. (NREL), Golden, CO (United States), 2010.

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 21 of 24
[11] W. Musial, P. Beiter, S. Tegen, A. Smith, Potential offshore wind energy areas in California: An assessment of locations, technology,
and costs, Technical Report, National Renewable Energy Lab. (NREL), Golden, CO (United States), 2016.
[12] J. A. Schneider, M. Senders, Foundation design: A comparison of oil and gas platforms with offshore wind turbines, Marine
Technology Society Journal 44 (2010) 32–51.
[13] M. Atcheson, A. Garrad, L. Cradden, A. Henderson, D. Matha, J. Nichols, D. Roddier, J. Sandberg, Floating Offshore Wind Energy,
Springer, 2016.
[14] M. Leimeister, E. E. Bachynski, M. Muskulus, P. Thomas, Rational upscaling of a semi-submersible floating platform supporting
a wind turbine, Energy Procedia 94 (2016) 434–442.
[15] J. M. Hegseth, E. E. Bachynski, A semi-analytical frequency domain model for efficient design evaluation of spar floating wind
turbines, Marine Structures 64 (2019) 186–210.
[16] Y. Kikuchi, T. Ishihara, Upscaling and levelized cost of energy for offshore wind turbines supported by semi-submersible floating
platforms, in: Journal of Physics: Conference Series, volume 1356, IOP Publishing, p. 012033.
[17] J. M. George, WindFloat design for different turbine sizes, Master thesis, Instituto Superior Tecnico, Universidade de Lisboa, 2014.
[18] J. Wu, M.-H. Kim, Generic upscaling methodology of a floating offshore wind turbine, Energies 14 (2021).
[19] P. Jamieson, Innovation in wind turbine design, John Wiley & Sons, 2018.
[20] G. Sieros, P. Chaviaropoulos, Upscaling beyond classical similarity: case study W/T tower, Technical Report, UPWIND Project
Report, 2009.
[21] G. Sieros, P. Chaviaropoulos, J. D. Sørensen, B. H. Bulder, P. Jamieson, Upscaling wind turbines: theoretical and practical aspects
and their impact on the cost of energy, Wind Energy 15 (2012) 3–17.
[22] T. Ashuri, M. B. Zaaijer, J. R. Martins, J. Zhang, Multidisciplinary design optimization of large wind turbines—technical, economic,
and design challenges, Energy Conversion and Management 123 (2016) 56–70.
[23] K. L. Dykes, J. Rinker, Windpact reference wind turbines, Technical Report, National Renewable Energy Lab. (NREL), Golden,
CO (United States), 2018.
[24] M. Saeki, I. Tobinaga, J. Sugino, T. Shiraishi, Development of 5-MW offshore wind turbine and 2-MW floating offshore wind
turbine technology, Hitachi Review 63 (2014) 414.
[25] J. Jonkman, S. Butterfield, W. Musial, G. Scott, Definition of a 5-MW reference wind turbine for offshore system development,
Technical Report, National Renewable Energy Lab. (NREL), Golden, CO (United States), 2009.
[26] C. Desmond, J. Murphy, L. Blonk, W. Haans, Description of an 8 MW reference wind turbine, in: Journal of Physics: Conference
Series, volume 753, IOP Publishing, p. 092013.
[27] C. Bak, F. Zahle, R. Bitsche, T. Kim, A. Yde, L. C. Henriksen, M. H. Hansen, J. P. A. A. Blasques, M. Gaunaa, A. Natarajan, The
DTU 10-MW reference wind turbine, in: Danish Wind Power Research 2013.
[28] P. Bortolotti, H. Canet Tarrés, K. Dykes, K. Merz, L. Sethuraman, D. Verelst, F. Zahle, Systems engineering in wind energy-WP2.
1 reference wind turbines, Technical Report, Technical Report No. NREL/TP-5000-73492, National Renewable Energy, 2019.
[29] C. E. Silva de Souza, P. A. Berthelsen, L. Eliassen, E. E. Bachynski, E. Engebretsen, H. Haslum, Definition of the INO WINDMOOR
12 MW base case floating wind turbine, Technical Report, SINTEF Ocean, 2021.
[30] E. Gaertner, J. Rinker, L. Sethuraman, F. Zahle, B. Anderson, G. Barter, N. Abbas, F. Meng, P. Bortolotti, W. Skrzypinski, et al.,
Definition of the IEA 15-Megawatt offshore reference wind turbine, Technical Report, National Renewable Energy Lab. (NREL),
Golden, CO (United States), 2020.
[31] T. Ashuri, J. R. Martins, M. B. Zaaijer, G. A. van Kuik, G. J. van Bussel, Aeroservoelastic design definition of a 20 MW common
research wind turbine model, Wind Energy 19 (2016) 2071–2087.
[32] E. E. Bachynski, M. I. Kvittem, C. Luan, T. Moan, Wind-wave misalignment effects on floating wind turbines: motions and tower
load effects, Journal of Offshore Mechanics and Arctic Engineering 136 (2014).
[33] W. Yu, K. Müller, F. Lemmer, H. Bredmose, M. Borg, G. Sanchez, T. Landbo, Public definition of the two LIFES50+ 10MW
floater concepts.”, LIFES50+ Deliverable 4 (2017).
[34] A. Pegalajar-Jurado, F. Madsen, M. Borg, H. Bredmose, Qualification of innovative floating substructures for 10MW wind turbines
and water depths greater than 50m. Deliverable D4.5 State-of-the-art models for the two LIFES50+ 10MW floater concepts,
Technical report, LIFES50+ consortium, 2018.
[35] C. Allen, A. Viscelli, H. Dagher, A. Goupee, E. Gaertner, N. Abbas, M. Hall, G. Barter, Definition of the UMaine VolturnUS-S
Reference Platform Developed for the IEA Wind 15-Megawatt Offshore Reference Wind Turbine, Technical Report, National
Renewable Energy Lab. (NREL), Golden, CO (United States), 2020.
[36] V. Riziotis, S. Voutsinas, E. Politis, P. Chaviaropoulos, A. M. Hansen, H. Madsen, F. Rasmussen, Identification of structural non-
linearities due to large deflections on a 5mw wind turbine blade, in: Proceedings of the 2008 European Wind Energy Conference
and Exhibition, EWEA: Brussels, pp. 9–14.
[37] A. Madsen, F. Rasmussen, D. Risø, Influence of rotational sampling of turbulence in combination with up-scaling of wind turbine
rotors, Note in WP2 of UpWind Project (2010).
[38] A. Sedaghat, A. Hassanzadeh, J. Jamali, A. Mostafaeipour, W.-H. Chen, Determination of rated wind speed for maximum annual
energy production of variable speed wind turbines, Applied Energy 205 (2017) 781–789.
[39] A. R. Henderson, C. Morgan, B. Smith, H. C. Sorensen, R. Barthelmie, B. Boesmans, Offshore windpower: a major new source
of energy for europe, International Journal of Environment and Sustainable Development 1 (2002) 356–369.
[40] N. Cetin, M. Yurdusev, R. Ata, A. Özdamar, Assessment of optimum tip speed ratio of wind turbines, Mathematical and
Computational Applications 10 (2005) 147–154.
[41] M. Yurdusev, R. Ata, N. Çetin, Assessment of optimum tip speed ratio in wind turbines using artificial neural networks, Energy
31 (2006) 2153–2161.

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 22 of 24
[42] D. Malcolm, A. Hansen, WindPACT Turbine Rotor Design Study: June 2000–June 2002 (Revised), Technical Report, National
Renewable Energy Lab.(NREL), Golden, CO (United States), 2006.
[43] D. A. Griffin, Windpact turbine design scaling studies technical area 1-composite blades for 80-to 120-meter rotor, Technical
Report, National Renewable Energy Lab., Golden, CO (US), 2001.
[44] L. Fingersh, M. Hand, A. Laxson, Wind turbine design cost and scaling model, Technical Report, National Renewable Energy
Lab.(NREL), Golden, CO (United States), 2006.
[45] C. A. Crawford, The path from functional to detailed design of a coning rotor wind turbine concept, Proceedings of the Canadian
Engineering Education Association (CEEA) (2007).
[46] F. M. Jensen, K. Branner, Introduction to wind turbine blade design, in: Advances in wind turbine blade design and materials,
Elsevier, 2013, pp. 3–28.
[47] A. Hayes, L. Sethuraman, K. Dykes, L. J. Fingersh, Structural optimization of a direct-drive wind turbine generator inspired by
additive manufacturing, Procedia Manufacturing 26 (2018) 740–752.
[48] G. van de Kaa, M. van Ek, L. M. Kamp, J. Rezaei, Wind turbine technology battles: Gearbox versus direct drive-opening up the
black box of technology characteristics, Technological Forecasting and Social Change 153 (2020) 119933.
[49] J. Li, K. Peng, P. Wang, N. Zhang, K. Feng, D. Guan, J. Meng, W. Wei, Q. Yang, Critical rare-earth elements mismatch global
wind-power ambitions, One Earth 3 (2020) 116–125.
[50] The European Wind Energy Association (EWEA), 2021. Accessed on 22/06/2021.
[51] J. Van der Tempel, D. Molenaar, Soft-soft, not hard enough?, in: Proceedings of the World Wind Energy Conference, Berlin.
[52] J. M. Hegseth, E. E. Bachynski, J. R. R. A. Martins, Integrated design optimization of spar floating wind turbines, Marine Structures
72 (2020) 102771.
[53] G. Bir, J. Jonkman, Modal dynamics of large wind turbines with different support structures, in: ASME 27-th International
Conference on Offshore Mechanics and Arctic Engineering (OMAE2008), volume 48234, pp. 669–679.
[54] A. N. Robertson, J. M. Jonkman, A. J. Goupee, A. J. Coulling, I. Prowell, J. Browning, M. D. Masciola, P. Molta, Summary of
conclusions and recommendations drawn from the deepcwind scaled floating offshore wind system test campaign, in: International
Conference on Offshore Mechanics and Arctic Engineering, volume 55423, American Society of Mechanical Engineers, p.
V008T09A053.
[55] G. DNV, Coupled analysis of floating wind turbines, Recommended Practice DNVGL-RP-0286 (2019).
[56] L. Arany, S. Bhattacharya, J. Macdonald, S. Hogan, Design of monopiles for offshore wind turbines in 10 steps, Soil Dynamics
and Earthquake Engineering 92 (2017) 126–152.
[57] N. Baldock, F. Sevilla, R. Redfern, A. Storey, A. Kempenaar, C. Elkinton, Optimization of installation, operation and maintenance
at offshore wind projects in the US: Review and modeling of existing and emerging approaches, Technical Report 701216-UKBR-
R-01-F, Garrad Hassan America, Inc., San Diego, CA (United States), 2014.
[58] J. Jonkman, Definition of the Floating System for Phase IV of OC3, Technical Report, National Renewable Energy Lab. (NREL),
Golden, CO (United States), 2010.
[59] A. Robertson, J. Jonkman, M. Masciola, H. Song, A. Goupee, A. Coulling, C. Luan, Definition of the semisubmersible floating
system for phase II of OC4, Technical Report, National Renewable Energy Lab. (NREL), Golden, CO (United States), 2014.
[60] A. Robertson, J. Jonkman, F. Vorpahl, W. Popko, J. Qvist, L. Frøyd, X. Chen, J. Azcona, E. Uzunoglu, C. Guedes Soares, et al.,
Offshore code comparison collaboration continuation within IEA wind task 30: phase II results regarding a floating semisubmersible
wind system, in: International Conference on Offshore Mechanics and Arctic Engineering, volume 45547, American Society of
Mechanical Engineers, p. V09BT09A012.
[61] J. M. Jonkman, Dynamics modeling and loads analysis of an offshore floating wind turbine, Technical Report, National Renewable
Energy Lab. (NREL), Golden, CO (United States), 2007.
[62] E. E. Bachynski, Design and dynamic analysis of tension leg platform wind turbines, Doctoral thesis, Norwegian University of
Science and Technology, 2014.
[63] DNVGL-ST-0119, Floating wind turbine structures, Standard, Det Norske Veritas, 2018.
[64] DNVGL-ST-0119, Floating wind turbine structures, Standard, Det Norske Veritas, 2021.
[65] A. B. of Shipping, Guide for building and classing floating offshore wind turbines, 2021.
[66] M. Collu, M. Borg, Design of floating offshore wind turbines, in: Offshore wind farms, Elsevier, 2016, pp. 359–385.
[67] H. Kooijman, C. Lindenburg, D. Winkelaar, E. Van der Hooft, DOWEC 6 MW Pre-Design: Aero-elastic modeling of the DOWEC 6
MW pre-design in PHATAS, Technical Report DOWEC-F1W2-HJK-01-046/9, DOWEC Dutch Offshore Wind Energy Converter
1997–2003 Public Reports, 2003.
[68] B. Fischer, M. Shan, A survey on control methods for the mitigation of tower loads, Technical Report 01/104256, Fraunhofer-
Institute for Wind Energy and Energy Systems Technology, IWES, 2013.
[69] V. Petrović, C. L. Bottasso, Wind turbine envelope protection control over the full wind speed range, Renewable Energy 111 (2017)
836–848.
[70] T. Burton, D. Sharpe, N. Jenkins, E. Bossanyi, Wind energy handbook, volume 2, Wiley Online Library, 2001.
[71] The Specialist Committee on Waves, Final report and recommendations to the 23rd ITTC, in: Proceedings of the 23rd International
Towing Tank Conference, volume II, pp. 505–736.
[72] J. Orszaghova, H. Wolgamot, S. Draper, R. Eatock Taylor, P. Taylor, A. Rafiee, Transverse motion instability of a submerged
moored buoy, Proceedings of the Royal Society A 475 (2019) 20180459.
[73] J. Orszaghova, H. Wolgamot, S. Draper, P. H. Taylor, A. Rafiee, Onset and limiting amplitude of yaw instability of a submerged
three-tethered buoy, Proceedings of the Royal Society A 476 (2020) 20190762.

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 23 of 24
[74] L. S. P. da Silva, B. S. Cazzolato, N. Y. Sergiienko, B. Ding, Efficient estimation of the nonlinear aerodynamic loads of floating
offshore wind turbines under random waves and wind in frequency domain, Journal of Ocean Engineering and Marine Energy
(2021) 1–17.
[75] J. Orszaghova, P. Taylor, H. Wolgamot, F. Madsen, A. Pegalajar-Jurado, H. Bredmose, Second and third order sub-harmonic wave
responses of a floating wind turbine, in: 35th International Workshop on Water Waves and Floating Bodies (IWWWFB 2020).
[76] L. S. P. da Silva, B. Cazzolato, N. Sergiienko, B. Ding, Nonlinear dynamics of a floating offshore wind turbine platform via
statistical quadratization—mooring, wave and current interaction, Ocean Engineering 236 (2021) 109471.
[77] D. Roddier, C. Cermelli, A. Weinstein, Windfloat: A floating foundation for offshore wind turbines—part i: design basis and
qualification process, in: International Conference on Offshore Mechanics and Arctic Engineering, volume 43444, pp. 845–853.
[78] D. Roddier, C. Cermelli, A. Aubault, A. Weinstein, WindFloat: A floating foundation for offshore wind turbines, Journal of
Renewable and Sustainable Energy 2 (2010) 033104.
[79] C. Cermelli, D. Roddier, A. Aubault, Windfloat: a floating foundation for offshore wind turbines—part ii: hydrodynamics analysis,
in: International Conference on Offshore Mechanics and Arctic Engineering, volume 43444, pp. 135–143.
[80] H. Kim, M. Kim, J. Lee, E. Kim, Z. Zhang, et al., Global performance analysis of 5MW WindFloat and OC4 semi-submersible
floating offshore wind turbines (FOWT) by numerical simulations, in: The 27th International Ocean and Polar Engineering
Conference, International Society of Offshore and Polar Engineers.
[81] C. Luan, Design and analysis for a steel braceless semi-submersible hull for supporting a 5-MW horizontal axis wind turbine,
Doctoral thesis, Norwegian University of Science and Technology, 2018.
[82] C. Luan, C. Michailides, Z. Gao, T. Moan, Modeling and analysis of a 5 MW semi-submersible wind turbine combined with three
flap-type wave energy converters, in: International Conference on Offshore Mechanics and Arctic Engineering, volume 45547,
American Society of Mechanical Engineers, p. V09BT09A028.
[83] F. Huijs, J. Mikx, F. Savenije, E.-J. de Ridder, Integrated design of floater, mooring and control system for a semi-submersible
floating wind turbine, Proceedings of the EWEA Offshore, Frankfurt, Germany 19 (2013) 21.
[84] F. Huijs, R. de Bruijn, F. Savenije, Concept design verification of a semi-submersible floating wind turbine using coupled
simulations, Energy Procedia 53 (2014) 2–12.
[85] A. Pegalajar-Jurado, H. Bredmose, M. Borg, J. G. Straume, T. Landbø, H. S. Andersen, W. Yu, K. Müller, F. Lemmer, State-of-the-
art model for the LIFES50+ OO-Star wind floater semi 10MW floating wind turbine, in: Journal of Physics: Conference Series,
volume 1104, IOP Publishing, p. 012024.
[86] M. Leimeister, E. E. Bachynski, M. Muskulus, P. Thomas, Design optimization and upscaling of a semi-submersible floating
platform, in: WindEurope Summit 2016, Hamburg, Germany.
[87] M. T. Islam, Design, Numerical Modelling and Analysis of a Semi-submersible Floater Supporting the DTU 10MW Wind Turbine,
Master’s thesis, Norwegian University of Science and Technology, 2016.
[88] Q. Wang, Design of a steel pontoon-type semi-submersible floater supporting the DTU 10MW reference turbine, Master’s thesis,
Norwegian University of Science and Technology, 2014.
[89] Z. Zhao, X. Li, W. Wang, W. Shi, Analysis of dynamic characteristics of an ultra-large semi-submersible floating wind turbine,
Journal of Marine Science and Engineering 7 (2019) 169.
[90] J. N. Newman, Marine hydrodynamics, The MIT press, 2018.
[91] W. Xue, Design, numerical modelling and analysis of a spar floater supporting the DTU 10MW wind turbine, Master’s thesis,
Norwegian University of Science and Technology, 2016.
[92] J. Falnes, A. Kurniawan, Ocean waves and oscillating systems: linear interactions including wave-energy extraction, volume 8,
Cambridge university press, 2020.
[93] G. Payne, Guidance for the experimental tank testing of wave energy converters, SuperGen Marine 254 (2008).

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 24 of 24
 Nomenclature
Acronyms
1P turbine rotational speed, the same as Ω
3P rotor blade passing frequency, or the frequency at which the blades pass by the tower
6P twice the blade passing frequency
DOF Degree-of-freedom
DTU Denmark Technical University
FOWT Floating offshore wind turbine
IEA International Energy Agency
NREL National Renewable Energy Laboratories, US
RAO Response amplitude operator
RNA Rotor-nacelle assembly
WP WindPACT project
Symbols
𝛽 blade pitch angle
𝜆 tip speed ratio
Ω rotor angular speed
𝜔𝑛 natural frequency of the FOWT
𝜌𝑎 air density
𝜌𝑤 water density
𝐴 hydrodynamic added mass
𝐴𝑡 wind turbine swept area
𝐵 hydrodynamic damping coefficient
𝐶𝑃 power coefficient
𝐶𝑇 thrust coefficient
𝐷 rotor diameter
𝑑 diameter of the upper part of the FOWT offset column
𝐹𝑏 buoyancy force
𝐹𝑡 rotor thrust force
𝐹𝑒𝑥𝑐 wave excitation force acting on a FOWT
𝐼𝑦 waterplane second moment of area (or area moment of inertia)
𝐾 total restoring stiffness of the platform
𝐾ℎ𝑠 hydrostatic stiffness of a FOWT
𝐾𝑚𝑜𝑜𝑟 restoring stiffness from the mooring system
𝑚 total mass of the FOWT including wind turbine and platform
𝑃 rated power
𝑝, 𝑝𝑥𝑦 , 𝑝𝑧 floating platform scaling factor related to the characteristic length of the platform (in all directions,
in horizontal plane, and in vertical direction)
𝑅 radius from the geometric centre of a FOWT to the centre of the offset column
𝑠 wind turbine scaling factor related to the rotor diameter
𝑈𝑟 relative wind speed at the nacelle
𝑧ℎ𝑢𝑏 , 𝑧𝐶𝑜𝐵 , 𝑧𝐶𝑜𝐺 vertical coordinate of the wind turbine height, centre of buoyancy, centre of gravity w.r.t.
to the still water level

Sergiienko, da Silva, Bachynski-Polić, Cazzolato, Arjomandi, Ding: Preprint submitted to Elsevier Page 25 of 24

View publication stats