Wind Energy

WIND ENERGY
Wind Energ. 2016; 19:1161–1177

Published online 30 September 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/we.1886
PAPER PRESENTED AT AWEA WINDPOWER
Design and model confirmation of the intermediate

scale VolturnUS floating wind turbine subjected to its
extreme design conditions offshore Maine
Anthony M. Viselli, Andrew J. Goupee, Habib J. Dagher and Christopher K. Allen
University of Maine, Advanced Structures and Composites Center, 35 Flaggstaff Road, 04469-5793 Orono, Maine, United States
ABSTRACT
Floating offshore wind turbines are gaining considerable interest in the renewable energy sector. Design standards for
floating offshore wind turbines such as the American Bureau of Shipping (ABS) Guide for Building and Classing
Floating Offshore Wind Turbine Installations are relatively new and few if any floating wind turbines have yet
experienced the prescribed design extreme environmental conditions. Only a few pilot floating turbines have been
deployed in Europe and Japan. These turbines have been designed for long return period storm events and are not likely
to see their extreme design conditions during early deployment periods because of the low probability of occurrence.
This paper presents data collected for an intermediate scale floating semi-submersible turbine intentionally placed
offshore Maine in a carefully selected site that subjects the prototype to scale extreme conditions on a frequent basis.
This prototype, called VolturnUS 1:8, was the first grid-connected offshore wind turbine in the Americas, and is a 1:8
scale model of a 6 MW prototype. The test site produces with a high probability 1:8 scale wave environments, and a
commercial turbine has been selected so that the wind environment/rotor combination produces 1:8-scale aerodynamic
loads appropriate for the site wave environment. In the winter of 2013–2014, this prototype has seen the equivalent of
50 year to 500 year return period storms exercising it to the limits prescribed by design standards, offering a unique
look at the behavior of a floating turbine subjected to extreme design conditions. Performance data are provided and
compared to full-scale predicted values from numerical models. There are two objectives in presenting this data and
associated analysis: (i) validate numerical aeroelastic hydrodynamic coupled models and (ii) investigate the performance
of a near full-scale floating wind turbine in a real offshore environment that closely matches the prescribed design con-
ditions from the ABS Guide. Copyright © 2015 John Wiley & Sons, Ltd.
KEYWORDS
offshore wind; VolturnUS; floating; model; validation; Maine
Correspondence
A. M. Viselli, University of Maine, Advanced Structures and Composites Center, 35 Flaggstaff Road, Orono, Maine 04469-5793,
United States.
E-mail: [email protected]
Received 21 May 2014; Revised 22 May 2015; Accepted 29 June 2015
1. INTRODUCTION
This paper presents sample data collected for a 1:8 scale prototype floating semi-submersible wind turbine placed in a
carefully selected test site that produces extreme environmental conditions relative to the 1:8 scale unit on a frequent basis
with a high probability. This prototype, called VolturnUS 1:8, was the first grid-connected offshore turbine in the Americas
and is a 1:8 scale model of a full-scale 6 MW prototype situated farther offshore in the Gulf of Maine. The test site reliably
produces 1:8 scale wave environments, and a commercial turbine has been selected so that the wind environment/rotor
combination produces 1:8 scale aerodynamic loads appropriate for a given wave environment. In the winter of
2013–2014, this prototype has seen the equivalent of 50 year to 500 year return period storms exercising it to the limits
prescribed by design standards for floating offshore wind turbines. This data offers a unique look at behavior of a floating
Copyright © 2015 John Wiley & Sons, Ltd. 1161

10991824, 2016, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/we.1886 by Lehigh University, Wiley Online Library on [26/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Design and model validation of VolturnUS floating wind turbine A. M. Viselli et al.
turbine subjected to extreme design conditions. Performance data are provided and compared to the full-scale predicted
values obtained through numerical modeling. There are two objectives in presenting this data: (i) validate that the numerical
aeroelastic hydrodynamic coupled models can adequately predict the response and (ii) investigate the performance of a
intermediate-scale floating wind turbine in a real offshore environment that closely matches the prescribed design
conditions. The usefulness and applicability of the testing results are also discussed.
VolturnUS is a new floating wind turbine platform design developed by the University of Maine. The design uses inno-
vations in materials, construction and deployment such as a concrete semi-submersible hull to reduce the costs of offshore
wind. An offshore model test at an intermediate scale serves to de-risk the development of a commercial scale VolturnUS
system by: (i) generating properly scaled global motion data for validation of numerical design models, (ii) confirmation of
performance estimates, (iii) allowing for implementation of the new structural materials and (iv) demonstrating construction
and deployment methods.
Model tests of offshore structures at small scales (e.g. 1:50) are routinely conducted in the offshore oil and gas sectors in
wave basin facilities as part of the design process.1 A basin model test offers a low risk and cost venue to accurately eval-
uate the floating system’s dynamic characteristics. Some example test programs for floating offshore wind turbines include
Principal Power Inc.,2 Hydro Oil & Energy,3 WindSea AS,4 the University of Maine,5 Manjock et al.6 and Ren et al.7 These
experiments are aimed at validating dynamic global performance of various floating wind turbine platform types. These
past efforts provide guidance for conducting scale model tests of floating wind turbines but do not include the use or dem-
onstration of full-scale materials, construction methods, a commercial turbine or deployment methods as part of the test.
There are currently six known intermediate-scale testing efforts of floating wind turbines with less than 1MW capacity
by commercial and public entities. BlueH* installed a 100 kW tension-leg platform off the Italian coast. Floating Power
Plant A/S installed the Poseidon wind/wave prototype with three 11 kW turbines.8 Kyoto University completed fabrication
and installation of a 1:2 scale 100 kW spar.† A 1:4 scale testing program is planned by NORCOE in Norway,‡ and a 1 kW
vertical axis turbine atop a spar type platform is planned by the Deepwind Consortium in Europe.§ These testing efforts are
an important stepping stone towards commercial scale projects.
Full-scale floating offshore wind turbines pilot projects are gaining considerable interest in the renewable energy sector
and several full-scale floating turbines have been deployed globally to date. Statoil has deployed a 2.3 MW spar design off
the coast of Norway.8 Principle Power deployed a 2.0 MW semi-submersible off Portugal.9 In Japan, a 2 MW semi-
submersible. off the coast of Fukushima¶, and a 2 MW Spar were deployed off the coast of Kabashima Island∥ funded
by the Ministry of Environment.
Design standards for floating offshore wind turbines such as DNVGL-OS-J103 and the American Bureau of Shipping
(ABS) Guide for Building and Classing Floating Offshore Wind Turbine Installations dictate required design environmental
conditions.10,11 Given that floating wind turbines were only deployed in the past few years, few if any are likely to have yet
experienced the DNVGL or ABS prescribed extreme environmental conditions because of their low-probability of occur-
rence. These efforts have been commercial in nature, and as a result, significant experimental results are not publicly available
yet on the performance of these systems in extreme design environments although some numerical predictions have been
published.
The VolturnUS 1:8 intermediate scale test program was carefully designed to produce useful results for design valida-
tion, model validation and technology demonstration at the intermediate-scale. The testing took place in a carefully selected
site that applies the correctly scaled extreme wind and wave loads with high probability. The test utilized a properly scaled
commercial turbine to apply the primary wind forcing, thrust, included an instrumentation system capable of measuring key
metocean data and hull performance characteristics. The prototype included design and construction methods representa-
tive of full scale. The testing effort was developed to meet several objectives:
1. Validate the VolturnUS design behavior at intermediate scale by conducting a Froude scale test representative of a
6 MW floating turbine deployed far offshore in the Gulf of Maine.
2. Design, test and demonstrate advanced material systems, construction techniques and deployment methods for the
VolturnUS concept.
3. Collect data for validation of coupled aeroelastic/hydrodynamic numerical models for floating offshore wind
turbines.
4. Develop deep water offshore wind testing capabilities, procedures and methods.
*
http://www.ewea.org/fileadmin/files/library/publications/reports/Deep_Water.pdf.
†
http://www.nawindpower.com/e107_plugins/content/content.php?content.12215.
‡
http://www.norcowe.no/.
§
http://www.offshorewind.biz/2010/11/11/future-wind-turbines-go-offshore-%E2%80%93-deep-and-floating-denmark/.
¶
http://www.fukushima-forward.jp/pdf/pamphlet4en.pdf 4/25/2014.
∥
http://www.japanfs.org/en/news/archives/news_id034598.html.
1162 Wind Energ. 2016; 19:1161–1177 © 2015 John Wiley & Sons, Ltd.
DOI: 10.1002/we
A. M. Viselli et al. Design and model validation of VolturnUS floating wind turbine
The testing was performed near Castine Harbor, Maine as shown in Figure 1. The VolturnUS 1:8 prototype is shown in
Figure 2. The concrete hull and composite materials tower were designed following the ABS Guide for Building and
Classing Floating Offshore Wind Turbine Installations. The turbine and platform were constructed onshore and towed to
site fully outfitted to demonstrate full-scale construction and deployment methods. A comprehensive instrumentation pack-
age monitors key performance characteristics of the system to verify the full-scale design and coupled modeling software.
Data collected during the program are discussed in this paper and compared to numerical models, as well as ABS design
load cases. Although not the focus of this paper, the test program also allowed for a permitting/ecological monitoring pro-
cedure to be developed in the State of Maine for floating offshore wind turbines which is being applied in an ongoing par-
allel full-scale design effort.
2. PROTOTYPE SCALING METHODOLOGY AND SITE SELECTION

Details on the site selection, scaling laws and system geometry are discussed in detail elsewhere by the authors and sum-
marized herein.12 Froude scaling was maintained for the model geometry and weight distribution as is customary in off-
shore model tests. The VolturnUS 1:8 scale model was also designed to be constructed, assembled and deployed using
similar materials and techniques as the full-scale 6 MW design, which can be fully constructed and assembled dockside.
The concrete hull design, composite tower design and construction process replicate the full-scale system. The concrete hull
is constructed of 15 concrete members and representative of the full-scale design and construction process. The connec-
tions, thicknesses and reinforcements of the concrete hull are scaled. The 1:8 scale model can be fully dismantled for ship-
ping over the road on a truck. The 1:8 composite tower is constructed of the same reinforcements and resins as a full-scale
tower. Figure 3 shows the completed unit. Table I lists the gross properties of the model and the full-scale system. The
properties compare well with the target scale values derived from the full-scale properties. Some dissimilitude exists be-
cause of the competing goals of the test program. A commercial pitch regulated 20 kW turbine modified to produce
12 kW was used for the test. At this reduced power and given the wind speed regime at the site, the turbine applies the de-
sired peak scaled thrust force and torque at the rated wind speed. The power and rotor diameter are not perfectly Froude
scaled because the focus was on properly scaling the thrust load, the primary wind turbine forcing on the floating plat-
form.13 The platform is moored using three catenary chains anchored to the sea bed with drag embedment anchors. The
average water depth at the site matches the target scaled depth. The model was successfully deployed on June 2nd of
2014 near Castine Harbor. The site was studied with an instrumented buoy the year before which confirmed the suitability
of the site to produce the desired 1:8 scale wave, current and wind environments for the prototype with a high probability of
occurrence.
3. INSTRUMENTATION
An overview of the onboard instrumentation systems is now provided. Figure 4 depicts the types and locations of sensors
onboard the model, as well as the buoy sensors. Data collected include wind speed and direction at four locations on the
platform, real time and statistical wave height, wind turbine power and rotor speed from the turbine instrumentation,
Figure 1. Project test site.
Wind Energ. 2016; 19:1161–1177 © 2015 John Wiley & Sons, Ltd. 1163
DOI: 10.1002/we
Figure 2. Grid-connected VolturnUS 1:8 scale prototype. Note: The rotor diameter has been reduced to apply the correct scaled thrust
and toque when subjected to the wind conditions found at the test site.
Figure 3. 1:8 model outfitted and ready for deployment.
platform angular position, platform translational and rotational accelerations, tower loads, platform loads and mooring line
loads. Sample rates vary depending on the instrument and range from 10 to over 60 Hz. For example accelerometers are
sampled at 60 Hz while turbine power output is 10 Hz. In total, over 60 channels of data are being collected from the float-
ing wind turbine. Also shown are the labels for the three legs of the platform as well as the coordinate system utilized for
expressing the acceleration results. Further, an instrumented buoy located in the test site approximately 200 m away pro-
vides a separate wave height measurement, wind speed, atmospheric and current profiles from the surface to the seafloor.
All data was collected on shore via wireless communications. Accuracy of the sensors varies but in general is targeted to be
less than 5% of the smallest expected values of interest. This comprehensive set of data permits careful comparisons
between predicted performance and measured performance, and as noted earlier, can also be used to accurately predict per-
formance for a full-scale 6 MW VolturnUS floating wind turbine located farther offshore using Froude scaling techniques.
DOI: 10.1002/we
Table I. VolturnUS system properties.
Parameter VolturnUS 1:8 VolturnUS full scale
Draft 2.9 m 23.2 m

Hub height 12.2 m 97.6 m
Rotor diameter 9.6 m 152 m
Rated power 12 kW 6 MW
Peak thrust load 1.37 kN 700 kN
Hull material Concrete
Tower material Composite
Average water depth 21 m 168 m
Number of mooring lines 3 × catenary chain
Anchors Drag anchor
Figure 4. Instrumentation systems.
The VolturnUS 1:8 model has been operational for 10 months at the time of this writing. During this time numerous
weather events have occurred and subjected the VolturnUS platform to a wide array of environmental conditions. The on-
board instrumentations system has successfully collected data describing the system’s performance, motions and loads.
These data are now examined through two perspectives in the rest of this paper: (i) comparison with numerical model sim-
ulations and (ii) evaluation of the VolturnUS platform in operational, extreme and survival design conditions prescribed
by ABS.
4. SAMPLE NUMERICAL MODEL VALIDATION CASE STUDY

A summary of numerical model correlation with the testing data is now discussed. The comparisons discussed are based
upon numerical models of the 1:8 scale prototype using direct input of environmental parameters and as-built geometry. This
allows for direct comparison of the simulation to experiment in order to establish confidence in the models ability to predict
the behavior. The primary numerical software considered in the validation effort is FAST Version 7 developed by the
National Renewable Energy Laboratory (NREL).14 In addition to performing FAST simulations, ANSYS AQWA Release
14.5 was used to evaluate mooring tensile loads because of AQWA’s high-fidelity mooring load simulation capabilities.15
DOI: 10.1002/we
4.1. System ID tests: hydrostatic stiffness and free-decay
To confirm the static stability characteristics of the 1:8 scale model, a fully assembled unmoored heave and heel hydrostatic
stiffness test was completed dockside prior to deployment. Known weights were placed atop the platform columns to in-
duce heave and heel responses. The results are shown in Figure 5 and are compared to the estimated response. The testing
confirmed the expected properties of the model. The heave measurement variability is because of local waves created by
wind and small craft nearby. This issue was not experienced in the heel incline test.
As a confirmation of the dynamic characteristics of the floating wind turbine, a combined heave/heel decay test was com-
pleted dockside using a crane pulling on and then releasing one of the outside columns. Surge, sway and yaw decay tests
were not possible dockside as the mooring system was not installed. Onboard instrumentation measured heel and heave and
corrected for angular displacement. Figure 6 shows a power spectral density plot created using the free decay data. The heel
and heave natural frequencies matched closely to the estimated scaled values of 0.12 Hz and 0.16 Hz, respectively. Hydro-
dynamic damping values for the system were tuned based on this data for use in numerical model validation efforts
discussed later in this paper.
Figure 5. Hydrostatic heave (left) and heel incline (right) testing.
Figure 6. VolturnUS 1:8 measured power spectral density plots from heave (top) and heel angle (bottom) free-decay test.
DOI: 10.1002/we
4.2. Summary of model validation against experimental performance data
A summary of comparisons between results of numerical modeling software FAST and experimental performance data
collected is now presented. Given the broad scope of this paper, a full treatment of the validation of numerical codes is
not possible. Detailed model validation efforts are presented by Allen et al.16 and additional publications are planned on
this subject. FAST is a coupled aeroelastic and hydrodynamic model developed to simulate floating wind turbine global
performance and has been validated through comparisons with tank testing data.17–19 FAST is freely available and
possesses comprehensive capabilities for modeling floating wind turbines. A numerical model of the 1:8 scale prototype
was developed in FAST for use in the validation effort.
The date of the event used herein for model correlation was November 27th, 2013, 12:51:54 PM to 1:51:58 PM and the
unit is shown in Figure 7. The mean and maximum wind speeds were 15.4 and 23.6 m s 1, respectively. The significant
wave height was 1.6 m. The peak spectral period was 5.2 s. The maximum wave height was 2.6 m which, when scaled using
a scale factor of 8, is 20.8 m which is close to a scaled 500 year wave height of 22 m for the 6 MW VolturnUS. The wind
wave directions were aligned such that they induced primarily sway and roll response. The turbine was operating at this
time. The simulated time was 1 h. To facilitate comparison between the data and model, a custom version of FAST was
utilized that computes 1st order wave excitation forces from a measured wave elevation time series record collected during
this event. The wind speed time history was also applied and allows for time history comparisons between the numerical
model and VolturnUS 1:8. The standard version of FAST can only simulate regular waves or random waves based on a
user defined spectra. Wind speed and direction were also input directly. This approach allowed for a reasonable modeling
approximation of the measured environment and followed the work of Coulling et al.20
To begin, sample data pertaining to wind turbine performance are presented. A representation of the turbine controller
was included in the FAST model. The data provided are in the form of time series comparisons for the blade pitch angle,
shown in Figure 8. The correlation between the measured and predicted blade pitch depicted in Figure 8 is quite good.
This figure provides confidence that the wind turbine blade aerodynamics and blade pitch actuation logic of the commer-
cial 20 kW turbine are reasonably well captured with the FAST model. This observation is further substantiated by the
time-series comparison of the simulated and measured tower top shear force in the sway direction, shown in Figure 9.
While this measurement also captures the inertial component of the shear load due to the motion of the wind turbine
mounted atop the tower, the measurement also captures the bulk of the aerodynamic load produced by the wind turbine
for the wind environment orientation considered. As can be seen in the figure, the mean force is fairly similar between the
two quantities and FAST does a fair job of capturing the dynamic trends in the tower top shear force. The differences seen
the figure could largely be because of measurement uncertainties including wave height and wind speed. For example,
wave heights are measured close to the structure and are actually affected by the structure itself. Similarly the wind speeds
are also made on the platform and are affected by the structure itself. Despite these uncertainties, the model compares
favorably with the measured response. A frequency domain comparison of the measured and FAST-predicted tower
Figure 7. VolturnUS 1:8 in 500 year scaled extreme environment.
DOI: 10.1002/we
Figure 8. Comparison of measured and FAST-predicted wind turbine blade pitch angle time series.
Figure 9. Comparison of measured and FAST-predicted yaw bearing sway-direction shear force time series.
Figure 10. Comparison of measured and FAST-predicted yaw bearing sway-direction shear force frequency domain response.
DOI: 10.1002/we
top sway shear force is also provided in Figure 10. Both the simulation and data show similar energy in the important
wave (~0.1 Hz), the wind turbine 1P (~1.6 Hz) and 3P and fundamental tower bending (~3.3 Hz) frequency ranges.
Figure 11 shows comparison between the data and FAST for the tower base sway acceleration. The model results com-
pare well with the measured performance capturing the maximum amplitude and frequency of the acceleration. Similarly,
platform roll acceleration is shown in Figure 12 again confirming the ability of the numerical code to accurately predict the
peak performance of the VolturnUS system.
Table II lists maximum recorded values for select variables. The FAST model compares very well with measured
performance.
Figure 13 provides a comparison between the measured mooring line tension and two simulations. The measured
mooring line tension is shown as a light solid line. Shown as a dark solid line is the FAST-predicted fairlead mooring
tension for one mooring line. As can be seen in the figure, the FAST mean load under predicts the measured quantity. That
said, the FAST prediction hardly changes from nominal whereas the measured data exhibits a great deal of moderate
Figure 11. Comparison of tower base sway acceleration predicted by numerical modeling in FAST and model test data collected at
Castine on Nov. 27, 2013.
Figure 12. Comparison of platform roll acceleration predicted by numerical modeling in FAST and model test data collected at Castine
on Nov. 27, 2013.
DOI: 10.1002/we
Table II. Comparison of peak values calculated using FAST and those measured on Nov. 27, 2013.
Variable Data FAST % Diff.
Blade pitch angle (degrees) 31.7 29.7 6%

2
Tower base sway acceleration (m s ) 1.19 1.14 4%
2
Tower base heave acceleration (m s ) 1.00 0.89 10%
2
Platform roll acceleration (rad s ) 0.119 0.108 9%
Figure 13. Comparison of measured, FAST-predicted and AQWA-predicted leg C fairlead tension time series.
amplitude motion because of drag load on the mooring lines induced by the platform motion from second order wave
kinematics, which is a known deficiency in FAST.20 To overcome this shortcoming, which is essential for estimating
the peak mooring loads required, one must either augment FAST with an improved multi-body or finite element mooring
model (e.g. see Garret21 or Paulling and Webster22), or use an alternative software package that possesses these capabilities.
For this model correlation study the fluid–structure interaction software ANSYS AQWA (which possesses dynamic
mooring line simulation capabilities) is employed to determine if a better correlation can be obtained between mooring
load simulation predictions and experimental data. In this model, the wind loading from the turbine is modeled as
quasi-static horizontal load acting perpendicular to the rotor plane. The mass of the turbine and its inertia are included
in the rigid body mass matrix. The predicted response of the mooring line tension by AQWA is plotted on this figure
as a dotted line. As can be seen in Figure 13, the inclusions of 2nd order wave drift forces acting on the platform,
mooring dynamics, mooring line added-mass and viscous drag loads on the mooring line in AQWA lead to superior
predictions of the mooring tension as compared to FAST’s quasi-static mooring line approach. Table III provides
the recorded leg C mooring load statistics alongside the statistics derived from the FAST and AQWA results. As
the table clearly shows, AQWA not only predicts the mean tension well, but also adequately predicts the maximum
mooring line tension.
5. VOLTURNUS PERFORMANCE IN ABS DESIGN CONDITIONS

Measured test data from three weather events at the Castine site that closely simulated full-scale ABS design load
conditions for a full-scale 6 MW turbine placed far offshore Maine’s coast are now presented. Because: (i) the test site
Table III. Comparison of measured, FAST-predicted and AQWA-predicted mooring tensions.
Mean mooring line tension Extreme mooring line tension
Variable Data FAST AQWA Data FAST AQWA
C leg mooring tension (kN) 2.11 1.63 1.95 3.22 1.78 3.28
DOI: 10.1002/we
was carefully selected to produce Froude scaled environmental conditions with a high probability, (ii) a commercial
turbine was chosen that applies the correct turbine thrust and torque force for the wind/wave environments of interest
and (iii) the VolturnUS 1:8 scale prototype was properly sized, the performance data collected can be scaled to predict
the behavior of the full-scale unit similar to the use of model test results obtained in a wave basin facility. Additional
information on the development of this testing program is discussed by Viselli et al..12 Full-scale 6 MW turbine
behavior was predicted using FAST and ANSYS AQWA and compared to directly the scaled test data in this section.
The full-scale platform and mooring numerical models are scaled from the as-built 1:8 prototype following Froude scal-
ing laws. The turbine model used for the full-scale simulations is based on a commercial 6 MW machine with features
such as active blade pitch, variable speed generator control and is not a direct scale-up of the commercial 20 kW ma-
chine. The 20 kW turbine exhibits the correct mean Froude scaled thrust load which is the primary wind loading
imparted by the wind turbine on the floating platform for the wind and wave environments considered in this paper. For other
design loading not discussed in this paper such as full-scale turbine controller fault conditions or unique controller behavior,
the difference in behavior between the 20 kW and full-scale 6 MW turbines may not produce as representative test data be-
tween full scale and model scale.
Three environmental events are considered in this section. These events closely approximate three particular
design load cases (DLCs) prescribed by the ABS.9 These DLCs are shown in Table IV. The values are obtained
from a metocean analysis utilizing over ten years of wind and wave buoy data at a proposed 6 MW full-scale
design deployment site off Monhegan Island, Maine.23,24 Data collected includes significant wave height and peak
period, wave energy spectral parameters and 10 min average wind speeds at 4 m above sea level. Using this data,
full-scale DLCs were developed for the 6 MW design. The DLCs are shown along with the desired 1:8 Froude
scale values in Table IV. DLC 1.2 is an operational type loading condition used for fatigue analysis. DLC 1.6
is an extreme turbine operating condition with an associated 50 year return period significant wave height
concurrent with the turbine operating wind speeds. The table specifically shows a DLC 1.6 case where the turbine
is operating near rated wind speed and peak thrust with a 50 year return period significant wave height and wave
period associated with this wind speed. The last case shown is a survival load case (SLC) with a 500 year return
period with the turbine parked and is used to determine wave-structure air gap as well as evaluate the mooring and
anchoring systems.
For each event, key performance data are presented and compared to a 6 MW full-scale system simulated response. The
6 MW full-scale system global performance was simulated using FAST augmented with a separate analysis using ANSYS
AQWA for the prediction of mooring line tensions following ABS guidelines.25 The number of random environments and
length of simulation time are shown in Table V which follows ABS guidance. The full-scale system simulations use the
design values prescribed by ABS. The test conditions shown in the following section closely match the ABS values and
are presented to provide an indication that the VolturnUS 1:8 did in fact experience conditions highly representative of
the ABS design conditions.
Table IV. Design load cases for the 6 MW full-scale unit and 1:8 Froude scale values.
ABS design load cases Required Metocean design parameter Full scale 1:8 scale
DLC 1.2 Operational load case significant wave (m) 0–6.0 0–0.75
Associated wave period (s) 6.1–11.5 2.2–4.1
DLC 1.6 50 year significant wave height associated with turbine peak thrust (m) 8.0 1.0
Associated wave period (s) 12.7 4.5
Survival load case (SLC) 500 year significant wave height (m) 12 1.5
Associated wave period (s) 15.3 5.4
1
500 year 10 min wind speed at hub height (m s ) 45 15.9
Table V. Simulations plan for the 6 MW full-scale system.
Number of simulations with Length of individual

DLC unique random seed values simulation (s) Total simulation time (s) Design value derivation
1.2 18 600 10,800 Mean value of maximum responses

1.6 6 3600 21,600 Mean value of maximum responses
6.1 6 10,800 64,800 Mean value of maximum responses
SLC 6 3600 21,600 Mean value of maximum responses
DOI: 10.1002/we
Figure 14. VolturnUS in DLC 1.2 environment.
5.2. DLC 1.2 operational load case (operating turbine with mild waves)
DLC 1.2 prescribed by ABS is a fatigue driving operational load case for many of the components on the floating wind
turbine. As seen in Figure 14, on November 9th, 2013, 6:53:04 7:53:04 AM, the VolturnUS unit experienced wind,
wave and current conditions representative of this load case at 1:8 scale. The significant wave height, mean wind speed
and mean current speed were 0.34 m, 8 m s 1 and 0.08 m s 1 respectively. The turbine was operating at near peak thrust
and the wind and waves were co-aligned. At full scale this equates to a significant wave height, mean wind speed and
mean current speed of 2.7 m, 10.2 m s 1 and 0.23 m s 1 respectively. A summary of the measured data is provided in
Table VI. Full-scale values are determined by applying Froude scaling laws to the measured data and compared with
numerical modeling results for the 6 MW full-scale design subjected to the prescribed ABS environmental conditions.
The scaled test data compares very well with the full-scale response indicating the unit experienced a very similar con-
dition during the test. The largest differences in response are less than 20%. These differences are in part because of
the combination of higher concentration of wave energy at the peak period for the 1:8 scale site as compared to the
full-scale system. The measured and full-scale JONSWAP gamma parameter is 6 as compared to the desired ABS
full-scale value of 2.5. This table confirms that the model was subjected to a representative DLC 1.2 scale
conditions and the performance of the prototype closely matches the expected full-scale response.
5.3. Extreme operational load case ABS DLC 1.6 (50 year waves with turbine operating)
DLC 1.6 prescribed by ABS is an extreme operational load case governing the strength of many components of float-
ing wind turbines and requires a 50 year return period wave environment associated with a given turbine operating
speed. On November 1st, 2013, 5:14:49–6:14:49 PM the VolturnUS unit experienced wind, wave and current conditions
representative of this load case at 1:8 scale with the turbine operating at peak thrust and rated wind speed. The
prototype on this day is shown in Figure 15. The measured significant wave height, 1 h-mean wind speed and
30 min-mean current speed were 1.07 m, 8.2 m s 1 and 0.11 m s 1 respectively. The turbine was operating at near peak
thrust and the wind and waves were co-aligned. At full-scale this equates to a significant wave height, mean wind
speed and mean current speed of 8.5 m, 10.2 m s 1 and 0.31 m s 1 respectively. A summary of the measured data is
provided in Table VII. These values are then scaled up following Froude scaling laws and compared with numerical
modeling results for the 6 MW full-scale design subjected to the prescribed ABS environmental conditions which
DOI: 10.1002/we
Table VI. Comparison of measurements on 11/9/2013 with predicted DLC 1.2 response at full scale.
Test data 11/9/13 6:53:04 AM 7:53:04 AM Full scale per ABS DL 1.2
Parameter 1:8 scale Full scale value %diff
Significant wave height (m) 0.34 2.7 2.7 0.3%

Max wave height (m) 0.82 6.6 4.6 29.8%
Peak period (s) 2.5 7.0 7.0 0.6%
JONSWAP gamma factor 6 6 2.5 58.3%
1
Mean current speed (m s ) 0.080 0.226 0.225 0.6%
Mean turbine thrust (kN) 1.12 573 521 9.1%
Max turbine thrust (kN) 1.53 782 661 15.5%
Mean turbine torque (kN-m) 0.62 2554 3008 17.8%
Max turbine torque (kN-m) 1.15 4721 4716 10.7%
Turbulence intensity (%) 13% 13% 9% 27.6%
2
Max nacelle acceleration (m s ) 0.77 0.77 0.89 15.8%
Mean heel angle (degrees) 2.29 2.29 2.19 4.2%
Max heel angle (degrees) 3.51 3.51 3.01 14.4%
Mean mooring line tension (kN) 3.80 1943 1879 3.3%
Max mooring line tension (kN) 4.82 2468 2239 9.2%
Figure 15. VolturnUS in DLC 1.6 environment.
closely match the witnessed environment. The scaled test data compares with the full-scale response very well. The
largest differences are only about 18% different indicating that the prototype was subjected to very similar condition.
The significant wave heights was slightly greater than the prescribed full scale by 6% and the measured JONSWAP
parameter of 1.9 compares very well with the full-scale value of 2.0. Measured maximum nacelle accelerations reached
1.42 m s 2 which were within 2% of the full-scale modeled condition. These closely matching data and full-scale
modeling results confirm that the model was subjected to a representative DLC 1.6 scale conditions and the
performance of the prototype closely matches the expected full-scale response.
DOI: 10.1002/we
Table VII. Comparison of measurements on 11/1/2013 with predicted DLC 1.6 response at full scale.
Test data on 11/1/13 5:14:49–6:14:49 PM Full scale per ABS DLC 1.6
Parameter 1:8 scale Full scale Value %diff
Significant wave height (m) 1.07 8.5 8 6.3%

Max wave height (m) 1.79 14.3 13.1 8.4%
Peak period (s) 4.6 13.0 12.7 1.9%
JONSWAP gamma factor 1.9 1.9 2 3.6%
1
Mean turbine thrust (kN) 1.11 567 525 7.5%
Max turbine thrust (kN) 1.67 856 727 15.1%
Mean turbine torque (kN-m) 0.643 2632 3018 14.7%
Max turbine torque (kN-m) 1.114 4564 4847 6.2%
2
Figure 16. VolturnUS in ABS survival environment.
5.4. Sample survival load case (500 year waves with turbine parked)
To conclude this look at the performance of the VolturnUS 1:8 prototype subjected to ABS design conditions, we now look
at a survival case. ABS prescribes a survival load case for evaluation of mooring and anchor systems as well as determi-
nation of the required structure-wave air gap clearance. The turbine is parked for this load case, and the unit is subjected
to 500 year return period events (e.g. 500 year significant wave height with mean associated wind and currents). On
November 27th, 2013, from 12:51:54 to 1:01:54 PM the VolturnUS unit experienced wind, wave and current conditions
DOI: 10.1002/we
representative of this load case at a 1:8 scale. The measured significant wave height, 10 min-mean wind speed and
30 min-mean current speed were 1.61 m, 15.4 m s 1 and 0.21 m s 1 respectively. The turbine was parked during this time.
Figure 16 shows the prototype in this environment. At full scale this equates to a significant wave height, mean wind speed
and mean current speed of 12.9 m, 29 m s 1 and 0.60 m s 1 respectively. These full-scale conditions are within about 10%
the estimated full-scale 500 year return period conditions prescribed by ABS for the 6 MW full-scale unit located far offshore
Maine. A summary of the measured data is provided in Table VIII along with the full-scale values and numerical modeling
results for the 6 MW full-scale design subjected to the prescribed ABS environmental conditions.
The scaled test data compares well with the expected full-scale response. The significant wave height, peak period
and JONSWAP gamma parameter measured in the test were very close to the 500 year sea state prescribed by ABS.
The significant wave height was within 6.8%, the peak period was within 3.2% and the measured JONSWAP param-
eter of 2.5 matches the required full-scale value. Figure 17 shows the wave time record as well as the wave spectrum
with a fitted JONSWAP spectrum indicating the quality of the wave spectrum produced at the test site. The current
also matched very well and was within 13% of the full-scale value. The predicted performance of the structure agrees
well with the measured data. The differences seen in the table are because of known modeling limitations for nonlinear
wave loading in extreme sea states. Several breaking waves were experienced by the VolturnUS unit which would also
explain the slightly higher acceleration values seen during the test as compared to the model. FAST also does not
currently include wind drag forces on hull, tower or nacelle which is why the mean heel angle differs. Another source
for differences could be that only 10 min of data was available that matched the environmental criteria of the survival
case. This duration is equivalent to 30 min at full scale. ABS requires six 1 h simulations. As discussed in the model
validation section of this paper, better agreement is obtained when using the true wave time history, wind speeds and
current profiles. Nevertheless, the data compares well with the full-scale modeling results providing confirmation that
Table VIII. Comparison of test data from 11/27/13 with predictions from ABS survival case at full scale.
Test data 11/27/13 12:51:54–1:01:54 PM Full scale per ABS survival case
Parameter 1:8 scale Full scale Value %diff
Significant wave height (m) 1.61 12.9 12 6.8%

Max wave height (m) 2.32 18.6 20.9 12.5%
Peak period (s) 5.2 14.8 15.3 3.2%
JONSWAP gamma factor 2.5 2.5 2.5 0.0%
1
2
Figure 17. Wave time history and spectrum with JONSWAP model.
DOI: 10.1002/we
the model was subjected to a representative survival conditions that exceed the ABS specification. The prototype
performed without issue in this environment.
6. SUMMARY AND CONCLUSIONS

A new floating wind turbine hull and tower design called VolturnUS was successfully deployed using an intermediate scale
prototype off the coast of Castine, Maine. This was the first grid-connected offshore wind turbine in the Americas. The pro-
totype was a 1:8 Froude scale version of a 6 MW full-scale design to be deployed far offshore Maine. Unlike a traditional
model test completed in a wave basin, the intermediate 1:8 scale experiment allowed for the prototype to be designed, con-
structed, deployed and operated in a similar way as the full-scale 6 MW version. The hull is constructed of concrete, the
tower is made of composite materials and the moorings are made of steel chain connected to drag anchors. In addition
to properly scaling the practical features of the VolturnUS design, the prototype was deployed in a site that was determined
through a record of buoy data to produce with high probability 1:8 Froude-scaled metocean conditions representative of the
far offshore site. The performance data in representative design events therefore can be scaled up to predict the full-scale
response of the 6 MW floating turbine.
During November, the VolturnUS 1:8 unit experienced many environmental conditions which closely represent full-
scale design conditions prescribed by the ABS floating offshore wind guide.11 Instrumentation on board the platform
collected environmental and global performance data allowing for two testing objectives to be met. Global response
numerical models (e.g. FAST and ANSYS AQWA) of the 1:8 scale prototype using measured environmental data
and as-built geometry show excellent agreement with measured response of the platform. This model validation for
an intermediate scale floating turbine operating in extreme environments offers another milestone in the development
of numerical models for floating wind turbines. It augments previous published validations which have been completed
with tank test data at 1:50 scale. The experiment has also shown that the offshore test program has been able to collect
data in an operational ocean environment of sufficient quality to verify numerical tools.
While this model correlation study has been able to draw positive conclusions regarding the performance of the nu-
merical software tools FAST and AQWA, there are several pieces of additional work that have not been covered in the
limited space provided in this paper. These include additional environments included in the correlation study, methods
of refining environmental input conditions used in the analysis and expanding the capabilities of FAST in an attempt to
refine the agreement between the measured data and predictions (e.g. refined aerodynamics, addition of slow-drift
forces, permitting the use of realistic short-crested seas in the hydrodynamic load calculations, etc.).
The second objective of this testing program, which is to exercise the VolturnUS 1:8 near its extreme design con-
ditions in a short period of time, has also been met. The test data shows that the VolturnUS platform was successfully
exercised in service under operational, extreme and survival load cases similar to those put forth by the ABS Guide.11
Measured responses of VolturnUS 1:8 confirm the design’s ability to successfully withstand these ABS conditions. The
experiment further provides unique operational data on the behavior of floating wind turbine platforms exercised to
these ABS design limits. Operational, 50 year to 500 year scale conditions were witnessed, recorded and compared with
expected full-scale response. Because the VolturnUS 1:8 model was properly scaled, the test site carefully selected to
produce scaled conditions, and the turbine properly matched to provide the correct thrust, the performance could be
compared to a 6 MW full-scale unit subjected to comparable ABS conditions far offshore Maine.
Satisfying these objectives has also confirmed the usefulness of the intermediate scale deployment approach prior to
full-scale deployment. In addition to validating design software and the design itself at a representative scale, the inter-
mediate test program provided several advantages: First, the test program de-risked the novel hull and tower materials,
construction methods and deployment techniques at a reduced scale. Second, an instrumentation system was developed
for a floating wind turbine that successfully provided the data necessary for model validation and which can be used
for a future full-scale deployment. Finally, the test program also allowed for ecological monitoring/permitting proce-
dures to be developed in the State of Maine for floating offshore wind turbines. These accomplishments confirm the
usefulness of an intermediate scale testing approach in the early stages of a new technology development program
for novel floating onshore wind foundation technologies.
ACKNOWLEDGEMENT
The authors would like to acknowledge the support of the US Department of Energy Office of Energy Efficiency and
Renewable Energy Grant Number DE-EE0002981 and DE-EE0003278, the National Science Foundation (PFI) Grant
Number IIP-0917974, the State of Maine 2010 State Bond, the Maine Technology Institute grant numbers CIP 111 and
CIP 120, the Rockerfeller Brothers, the Davis Family Foundation, the University of Maine and the support of the members
of the DeepCwind Consortium including Cianbro and Ershigs.
DOI: 10.1002/we
REFERENCES
1. Chakrabarti SK. Offshore Structure Modeling. World Scientific Publishing Co. Pte. Ltd: Singapore, 2004.
2. Roddier D, Carmelli C, Weinstin A. A floating foundation for offshore wind turbines part I: design basis and qualifi-
cation process. OMAE 2009; 4: 845–853.
3. Skaare B, Hanson TD, Nielsen FG, Yttervik R, Hansen AM, Thomesn K, Larsen TJ. Integrated dynamic analysis of floating
offshore wind turbines. Paper presented at the European Wind Energy Conference and Exhibition, Milan, 7–10, May, 2007.
4. Windsea AS. Next generation floating wind farm. Scandinavian Oil-Gas Magazine 2010; 7/8: 250–253.
5. Goupee AJ, Koo BJ, Lambrakos KF, Kimball RW. Model test of three floating wind turbine concepts. OTC 23470,
Offshore Technology Conference, 2012.
6. Manjock A, Bredmose H, Azcona J, Lemmer F, Amann F, Campagnolo F, Montinari P. Rotor aerodynamics for tank
tests of scaled floating wind turbines. In Proceedings of the EWEA Offshore. Copenhagen, 2015.
7. Ren NX, Li YG, Ou JP. The effect of additional mooring chains on the motion performance of a floating wind turbine
with a tension leg platform. Energy 2012; 5: 1135–1149.
8. Athanasia A, Genachte AB. Deep offshore and new floating concepts, DeepWind’2013, 24–25 January, Trondheim,
Norway. Energy Procedia 2013; 35: 198–209. Elsevier
9. Carmelli CA, Roddier, Weinstein A. Implementation of a 2MW floating wind turbine prototype offshore Portugal.
OTC 23492, Offshore Technology Conference, Houston Texas, USA, 30 April–3 May 2012.
10. Det Norske Veritas AS. DNV-OS-J103, design of floating wind turbine structures. Available: http://www.dnv.com/.
(Accessed June 2013).
11. American Bureau of Shipping. 2013. Guide for Building and Classing Floating Offshore Wind Turbine Installations.
Houston: TX.
12. Viselli AM, Goupee AJ, Dagher HJ. Model test of a 1:8 scale floating wind turbine offshore in the Gulf of Maine. Pro-
ceedings from the ASME 2014 33rd International Conference on Ocean, Offshore and Artic Engineering. OMAE2014.
San Francisco, California, USA, June 8–13, 2014.
13. Goupee AJ, Koo BJ, Lambrakos KF, Kimball RW. Model test of three floating wind turbine concepts. OTC 23470,
Offshore Technology Conference. 2012.
14. Jonkman JM, Buhl Jr. ML. FAST user’s guide. National Renewable Energy Laboratory. Technical Report NREL/EL-
500-38230, 2005.
15. ANSYS, Inc. ANSYS AQWA reference manual, release 14.5. October 2012.
16. Allen CK, Goupee AJ, Viselli AM, Dagher HJ. Validation of global performance numerical design tools used for
design of floating offshore wind turbines. ASME paper No. OMAE2015-41437, 2015.
17. Coulling AJ, Goupee AJ, Robertson AN, Jonkman JM, Dagher HJ. Validation of a FAST semi-submersible floating
wind turbine model with DeepCwind test data. Journal of Renewable and Sustainable Energy 2013; 5: 023116.
18. Prowell I, Robertson A, Jonkman J, Stewart GM, Goupee AJ. Numerical prediction of experimentally observed
behavior of a scale model of an offshore wind turbine supported by a tension-leg platform. Proceedings of the Offshore
Technology Conference (OTC), Houston, Texas, USA, May 6–9, 2013.
19. Browning JR, Jonkman J, Robertson A, Goupee AJ. Calibration and validation of a spar-type floating offshore wind
turbine model using the FAST dynamic simulation tool. Proceedings of the Science of Making Torque from Wind
Conference. Oldenburg, Germany, October 9–11, 2012.
20. Coulling AJ, Goupee AJ, Robertson AN, Jonkman JM. Importance of second-order difference-frequency wave-diffraction
forces in the validation of a FAST semi-submersible floating wind turbine model. Proceedings of OMAE 2013, ASME
32ndInternational Conference on Offshore Mechanics and Arctic Engineering. Nantes, France, June 9–14, 2013.
21. Garrett DL. Dynamic analysis of slender rods. Journal of Energy Resources Technology 1982; 104: 302–306.
22. Paulling JR, Webster WC. 9186 A consistent, large-amplitude analysis of the coupled response of a TLP and tendon
system. Proceedings of the 5th International Offshore Mechanics and Arctic Engineering Conference, Tokyo, 1986.
23. Pettigrew NR, Xue H, Irish JD, Perrie W, Roesler CS, Thomas AC, Townsend DW. The Gulf of Maine Ocean
Observing System: generic lessons learned in the first seven years of operation (2001–2008). Marine Technology
Society Journal 2008; 42: 91–102.
24. University of Maine. 2013. Mooring E0130: Central Maine Shelf, Gulf of Maine Moored Buoy Program, Physical
Oceanography Group, School of Marine Sciences. http://gyre.umeoce.maine.edu/buoyhome.php.
25. American Bureau of Shipping. Guidance notes on global performance analysis of floating offshore wind turbine
installations. 2014.
DOI: 10.1002/we

Wind Energy - 2015 - Viselli - Design and Model Confirmation of The Intermediate Scale VolturnUS Floating Wind Turbine

Uploaded by

Copyright:

Available Formats

Wind Energy - 2015 - Viselli - Design and Model Confirmation of The Intermediate Scale VolturnUS Floating Wind Turbine

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Wind Energy - 2015 - Viselli - Design and Model Confirmation of The Intermediate Scale VolturnUS Floating Wind Turbine

Uploaded by

Copyright:

Available Formats

Wind Energ. 2016; 19:1161–1177

PAPER PRESENTED AT AWEA WINDPOWER

Design and model conﬁrmation of the intermediate

Received 21 May 2014; Revised 22 May 2015; Accepted 29 June 2015

Copyright © 2015 John Wiley & Sons, Ltd. 1161

2. PROTOTYPE SCALING METHODOLOGY AND SITE SELECTION

Figure 1. Project test site.

Figure 3. 1:8 model outﬁtted and ready for deployment.

Table I. VolturnUS system properties.

Parameter VolturnUS 1:8 VolturnUS full scale

Draft 2.9 m 23.2 m

Figure 4. Instrumentation systems.

4. SAMPLE NUMERICAL MODEL VALIDATION CASE STUDY

4.1. System ID tests: hydrostatic stiffness and free-decay

Figure 5. Hydrostatic heave (left) and heel incline (right) testing.

4.2. Summary of model validation against experimental performance data

Figure 7. VolturnUS 1:8 in 500 year scaled extreme environment.

Variable Data FAST % Diff.

Blade pitch angle (degrees) 31.7 29.7 6%

5. VOLTURNUS PERFORMANCE IN ABS DESIGN CONDITIONS

Table III. Comparison of measured, FAST-predicted and AQWA-predicted mooring tensions.

Mean mooring line tension Extreme mooring line tension

Variable Data FAST AQWA Data FAST AQWA

Table V. Simulations plan for the 6 MW full-scale system.

Number of simulations with Length of individual

1.2 18 600 10,800 Mean value of maximum responses

Figure 14. VolturnUS in DLC 1.2 environment.

Parameter 1:8 scale Full scale value %diff

Signiﬁcant wave height (m) 0.34 2.7 2.7 0.3%

Figure 15. VolturnUS in DLC 1.6 environment.

Parameter 1:8 scale Full scale Value %diff

Signiﬁcant wave height (m) 1.07 8.5 8 6.3%

Figure 16. VolturnUS in ABS survival environment.

Parameter 1:8 scale Full scale Value %diff

Signiﬁcant wave height (m) 1.61 12.9 12 6.8%

6. SUMMARY AND CONCLUSIONS

You might also like