Comparison of Loads From HAWC2 and OpenFAST For The IEA Wind 15 MW Reference Wind TurbineJournal of Physics Conference Series
Comparison of Loads From HAWC2 and OpenFAST For The IEA Wind 15 MW Reference Wind TurbineJournal of Physics Conference Series
Comparison of Loads From HAWC2 and OpenFAST For The IEA Wind 15 MW Reference Wind TurbineJournal of Physics Conference Series
Comparison of loads from HAWC2 and OpenFAST for the IEA Wind 15
MW Reference Wind Turbine
To cite this article: Jennifer Rinker et al 2020 J. Phys.: Conf. Ser. 1618 052052
Abstract. Reference wind turbines (RWTs) that reflect the state-of-the-art of current wind
energy technology are necessary in order to properly evaluate innovative methods in wind turbine
design and evaluation. The International Energy Agency (IEA) Wind Technology Collaboration
Platform (TCP) Task 37 has recently developed a new RWT geared towards offshore floating-
foundation applications: the IEA Wind 15 MW. The model has been implemented in two
aeroelastic codes, OpenFAST and HAWC2, based on an underlying common ontology. However,
these toolchains result in slightly different structural parameters, and the two codes utilise
different structural models. Thus, to increase the utility of the model, it is necessary to
compare the aeroelastic responses. This paper compares aeroelastic loads calculated using
different fidelities of the blade model in OpenFAST (ElastoDyn and BeamDyn) and HAWC2
(prismatic Timoshenko without torsion and Timoshenko with fully populated stiffness matrix),
where both codes use the DTU Basic controller and the same turbulence boxes to reduce
discrepancies. The aeroelastic responses to steady wind, step wind and turbulent wind (per
IEC 61400-1 wind class IB) are considered. The results indicate a generally good agreement
between the loads dominated by aerodynamic thrust and force, especially for the no-torsion
blade models. Discrepancies were observed in other load channels, partially due to differences
in the asymmetric loading of the rotor and partially due to differing closed-loop dynamics, and
they will be the subject of future investigations.
1. Introduction
To successfully investigate the impacts of innovation on the efficacy and cost of wind energy
systems, it is essential that well-tested, state-of-the-art reference wind turbines (RWTs) are
shared publicly. On the one hand, commercial wind turbine models (even those out of production
for many years) embody sensitive intellectual property information related to many system
design aspects that cannot be shared publicly. On the other hand, sub-suppliers, developers,
consultancies, and research institutes need models for wind turbines that closely resemble
the commercial state-of-the-art in order to investigate scientific phenomena and innovative
technologies that will help advance performance and reduce costs for next generation wind
energy systems. One of the most well-known examples in wind energy is the National Renewable
Energy Laboratory (NREL) 5 megawatt (MW) [1], but a large number of other RWTs have been
produced to capture evolution in commercial technology over the years, including the Technical
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
University of Denmark (DTU) 10 MW [2] and the Wind Partnership for Advanced Component
Technology (WindPACT) RWTs [3].
As the wind energy industry has developed, the need for new RWTs that reflect current
and future technologies has grown. To this end, the International Energy Agency (IEA) Wind
Technology Collaboration Platform (TCP) Task 37 on systems engineering has developed a
series of RWTs [4]: the IEA Wind 3.4 MW RWT, the IEA Wind 10 MW RWT and recently
the IEA Wind 15 MW, targeting offshore applications. The model was originally designed
by researchers from the National Renewable Energy Laboratory (NREL) before being further
developed collaboratively with researchers at the Technical University of Denmark (DTU). To
maximize its utility, a variety of design documents and aeroelastic input files are provided on
GitHub [5]. More information on the model specifications can be found in the design report [6].
The underlying definitions of the RWT are defined according to the IEA Wind Task 37
wind turbine ontology, windIO [7]—a formal, common definition of a wind turbine model—and
this common ontology was converted to OpenFAST and HAWC2 implementations using different
toolchains. DTU’s toolchain used BECAS [8], HAWC2 [9] and HAWCStab2 [10], whereas NREL
used Sonata/Anba4 [11], WISDEM [12], and OpenFAST [13, 14]. The differences in toolchains
result in slightly different aeroelastic parameters, which will affect aeroelastic responses. To
complicate matters, the implementations in both OpenFAST and HAWC2 have multiple blade
models with different fidelities: OpenFAST has ElastoDyn and BeamDyn, and HAWC2 has
prismatic Timoshenko without torsion, prismatic Timoshenko with torsion and a Timoshenko
model with a fully populated stiffness matrix. For the IEA Wind 15 MW RWT to be useful
to both industry and researchers, it is essential that its inter-code aeroelastic behavior is well
understood. The most detailed source of information on differences between aeroelastic codes is
the set of reports from the Offshore Code Comparison projects (OC3 [15], OC4 [16] and OC5
[17]). However, these reports compare too many different aeroelastic codes and use an aeroelastic
model of a much smaller wind turbine, so we cannot draw relevant conclusions about differences
in the aeroelastic response of the IEA Wind 15 MW. Thus, this paper seeks to provide a baseline
inter-code comparison for the RWT, to yield key insights into the aeroelastic performance of
ElastoDyn, BeamDyn and HAWC2 and to provide a case study in using windIO for collaborative
research.
The remainder of the paper is organised as follows. Section 2 describes the considerations
taken to reduce discrepancies in the blade structure, turbulence and controller. Comparisons of
the blade natural frequencies and steady-state, step-wind and turbulent responses are provided
in Sec. 3. Finally, the conclusions are drawn in Sec. 4.
2. Model comparisons
This section provides details on the modelling methodology, including particular efforts that were
made to reduce discrepancies between the two models. The version of the IEA Wind 15 MW
used in both OpenFAST and HAWC2 was the monopile model, but both hydrodynamics and
the monopile flexibility were disabled to mimic an onshore turbine. The remaining subsections
discuss specific aspects of the aerodynamics, blade structural modelling, turbulence modelling
and controller that could cause discrepancies and what measures were taken to reduce those
discrepancies.
2.1. Aerodynamics
The two aeroelastic codes use similar methodologies to calculate aerodynamic loading. The
OpenFAST model uses AeroDyn to calculate the aerodynamic forces. Tower shadow is modeled
using the baseline potential flow model and the unsteady airfoil aerodynamics are calculated
with the Beddoes-Leishman model. The HAWC2 model uses its standard potential tower shadow
model and a modified Beddoes-Leishman model for unsteady airfoil aerodynamics [18]. Both
2
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
models use Blade Element Momementum Theory, including empirical corrections like the Prandtl
tip loss model.
2.3. Turbulence
The default turbulence simulators for OpenFAST and HAWC2 use different modeling
methodologies: OpenFAST uses the Kaimal spectrum with exponential coherence (KSEC),
whereas HAWC2 uses the Mann model. The Mann model uses a three-dimensional spectral
tensor that includes spatial correlation, whereas the KSEC model uses three one-dimensional
spectra and a spatial coherence model. These differences in inflow generation could have
significant impact on the aeroelastic results. Therefore, the turbulence boxes used in the HAWC2
simulations were converted to an OpenFAST-friendly format using PyConTurb [19] and then
used as inflow for the ElastoDyn and BeamDyn simulations.
2.4. Controller
A significant potential source of discrepancy in the dynamical behaviour of the models is different
controller logics. The default controller for the OpenFAST implementation of the RWT is
3
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
NREL’s ROSCO (Reference Open Source Controller) [20, 21], whereas the HAWC2 model uses
the DTU Basic Controller [22]. For a code-to-code comparison, it is essential that as many
possible sources of discrepancy are removed from the comparison. Thus, for this paper, the
DTU Basic controller was compiled and coupled with OpenFAST, and the same controller
parameters were used for the corresponding OpenFAST and HAWC2 models. There are two
sets of controller parameters, one for the models with torsion and one for the models without
torsion. Thus, ElastoDyn and H2-PTNT used the same controller parameters, as did BeamDyn
and H2-FPM. Both sets of controller parameters were tuned using HAWCStab2 v2.16a [?] with
assumed partial-load poles (torque controller) and full-load poles (pitch controller) of 0.05 Hz,
70% critical, and 0.03 Hz, 70% critical, respectively. The controller parameters can be found in
the input files on the RWT GitHub [5].
3. Results
This section presents the comparisons of the aeroelastic results from HAWC2/HAWCStab2 and
OpenFAST. An analysis of the blade structural models is first presented in Sec. 3.1, which
includes the natural frequencies and a selection of distributed structural properties. Then,
Secs. 3.2, 3.3 and 3.4 compare the steady-state, step-wind and turbulent responses, respectively.
All OpenFAST results were simulated using version 2.2, the HAWC2 results were simulated
using version 12.8 1900 and the HAWCStab2 natural frequencies were generated using version
2.16a.
The structural natural frequencies of the blade calculated using OpenFAST’s linearization
option and HAWCStab2 [?] are compared in Table 1. The mode shapes were identified using
HAWCStab2’s mode-shape-visualization capabilities. The percent difference was calculated
4
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
according to pdif f = 2(fHAW C2 − fOpenF AST )/(fHAW C2 + fOpenF AST ) ∗ 100%. The agreement
between the two pairs of models (i.e., ElastoDyn vs. H2-PTNT and BeamDyn vs. H2-FPM) is
generally good. This small differences in natural frequencies are a natural result of the difference
in structural parameters, as discussed above. There is a larger difference in the 3rd flapwise
mode in the BeamDyn/H2-FPM models, but it may be ignored in further analyses because it
is sufficiently small and on a higher mode shape, which has a low modal participation factor in
the overall response.
5
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
A selection of steady-state loads and deflections are plotted in Fig. 3. From the plots, the
following observations may be made:
• Thrust- and torque-dominated loads for ElastoDyn and H2-PTNT match quite
well. The tower-base fore-aft moment, flapwise blade root moment and edgewise blade
root moment are all loads that are primarily dominated by the aerodynamic thrust or
aerodynamic torque. For the ElastoDyn and H2-PTNT models, the aerodynamic thrust
and torque match quite well (not shown), and therefore these load channels also feature
very good agreement.
• Small differences in the thrust- and torque-dominated loads for the BeamDyn
and H2-FPM models. The differences in the BeamDyn and H2-FPM blade structural
models and resulting difference in controller set-points result in thrust and torque curves for
the BeamDyn and FPM models (not shown) that do not match as well as their no-torsion
counterparts. These small differences in the thrust and torque curves for the BeamDyn and
H2-FPM models then manifest as differences in the tower-base fore-aft moment, flapwise
blade root moment and edgewise blade root moment. For example, the maximum flapwise
blade root moment for BeamDyn is approximately 7.6% lower than H2-FPM.
• Differences in the asymmetry-sensitive loads and deflections for all models.
The tower-base side-side moment and yaw-bearing pitch moment are both sensitive to
asymmetric rotor loading, which is caused by tilt, coning, etc. These load channels
demonstrate distinct differences not only between BeamDyn and H2-FPM but also
ElastoDyn and H2-PTNT. Thus, there are differences in the net lateral and vertical force
exerted on the rotor by the aerodynamic loading, which result in different tower-base side-
side and yaw-bearing pitch moment values. This difference in the net lateral and vertical
rotor loading could be caused by the difference in blade deflections (Fig. 3 bottom right
subplot) or by fundamental differences in the aerodynamic calculations.
6
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
behaviour from the start of the simulation has died out. The Region 1 wind speeds are from
220 to 300 s, Region 2 from 300 s to 540 and Region 3 from 540 s onwards.
The agreement between the two models is generally quite good, especially for the
ElastoDyn/H2-PTNT models, with more discernible differences between the BeamDyn and H2-
FPM responses. With the same controller and controller parameters between the two codes,
the differences can be attributed to the structural modeling, both in terms of the structural
property preprocessing and differences in the theoretical basis of the two codes. The resulting
blade deflections and torsion also change the respective aerodynamic responses of the blades.
Therefore, using the exact same controller parameters does not maximize agreement between the
models. Minor re-tuning that accounts for the differences in the aeroelastic response is expected
to further reduce discrepancies and will be the subject of future work. Generally though, the
7
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
Figure 5. Mean (circles) and standard deviation (error bars) of aeroelastic response to class
IB turbulence with H2-PTNT and ElastoDyn.
The statistics of a selection of load channels versus mean wind speed for ElastoDyn vs. H2-
PTNT and BeamDyn vs. H2-FPM are presented in Figs. 5 and 6, respectively. The following
observations may be made from the figures:
• Good agreement in the operational data and thrust- and torque dominated
channels for ElastDyn and H2-PTNT. Because the rotor thrust and torque for the two
models match quite well, the flapwise root moment and tower-base fore-aft have generally
good agreement.
• Higher thrust and related loads with H2-FPM. Due to differences in the structural
parameters, the H2-FPM model has a generally higher thrust than BeamDyn, which also
8
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
Figure 6. Mean (circles) and standard deviation (error bars) of aeroelastic response to class
IB turbulence with H2-FPM and BeamDyn.
results in generally higher flapwise root moment, tip deflections and tower-base fore-aft
moment.
• Differences due to asymmetric rotor loading in all models. Due to the addition of
wind shear, there are larger differences between OpenFAST and HAWC2 due to asymmetric
rotor loading. For example, for ElastoDyn vs. H2-PTNT, the mean yaw-bearing pitch
moment and the mean tower-base moments at higher wind speeds show poor agreement.
Similar trends are also seen in the BeamDyn vs. H2-FPM models.
• Larger variation in OpenFAST tower-base moments at low wind speeds. This
is likely due to the controller parameters being ill-suited to the slightly different open-loop
dynamics of the OpenFAST models. Future work could tune the controller parameters such
that the OpenFAST and HAWC2 models demonstrated identical closed-loop behaviour.
The extreme values for ElastoDyn vs. H2-PTNT and BeamDyn vs. H2-FPM are presented
in Figs. 7 and 8, respectively. From the plots, the following observations can be made:
• Larger max generator speed and power in OpenFAST above rated. This indicates
that the identical controller parameters used in the two models do not produce the same
closed-loop dynamics, resulting in larger OpenFAST overshoots. As noted above, future
work could include a controller study that seeks to match the closed-loop behaviour of the
two models.
• Generally good agreement for ElastoDyn vs. H2-PTNT loads and deflections.
Except for the yaw-bearing pitch moment, the maximum values for the flapwise root
moment, flapwise tip deflection and tower-base moments show generally good agreement.
9
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
Figure 7. Extreme values of aeroelastic response to class IB turbulence with H2-PTNT and
ElastoDyn.
As noted above, differences in the maximum yaw-bearing moment occur due to differences
in the asymmetric rotor loading.
• Larger maximum blade responses for H2-FPM than BeamDyn. Because H2-FPM
features larger mean thrust and torque than BeamDyn, the maximum values for the loads
dominated by thrust and torque in H2-FPM are larger than their BeamDyn counterparts.
• Relatively good agreement in tower-base loads for BeamDyn vs. H2-FPM.
• Similar maximum yaw-bearing pitch moment for BeamDyn vs. H2-FPM.
Considering that the mean value of the BeamDyn yaw-bearing pitch moment is more
negative than the H2-FPM moment but their variations have similar amplitude, this might
be indicative of suboptimal closed-loop behaviour. A future in-depth control study would
be needed to examine this issue further.
4. Conclusions
This paper presents a detailed aeroelastic comparison of the IEA Wind 15 MW RWT as
implemented in OpenFAST and HAWC2/HAWCStab2. The model was implemented in the
two aeroelastic codes using separate toolchains, resulting in differences in structural properties,
and it is important to quantify the effect these differences will have on the respective aeroelastic
responses. To reduce the complexity of the analysis, particular efforts were made to reduce as
many discrepancies between the two models as possible. First, two different HAWC2 models
were used (Timoshenko without torsion and Timoshenko with a fully populated stiffness matrix)
to better match the structural models used in ElastoDyn and BeamDyn, respectively. Second,
10
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
Figure 8. Extreme values of aeroelastic response to class IB turbulence with H2-FPM and
BeamDyn.
the Mann turbulence boxes from HAWC2 were reused in OpenFAST to eliminate differences
caused by inflow modelling. Lastly, the DTU Basic Controller with identical parameters was the
controller in both models. The aeroelastic responses of the two pairs of models were evaluated
by comparing their steady-state responses, step-wind responses and responses to turbulent wind
prescribed according to IEC 61400-1 design load case 1.1.
The responses of all four models showed generally good agreement, although differences did
arise that merit discussion. First, the load channels dominated by aerodynamic thrust and
torque (e.g., tower-base bending moment) showed very good agreement between ElastoDyn
and H2-PTNT. The differences in the structural values for BeamDyn and H2-FPM models
resulted in higher aerodynamic thrust, so the loads dominated by thrust were slightly higher
in the H2-FPM model. Interestingly, the load channels sensitive to asymmetric rotor loading
(e.g., tower-base side-side moment and yaw-bearing moment) showed differences, even for the
steady-state comparisons of ElastoDyn and H2-PTNT. Future work is needed to investigate
whether this difference is caused by differences in the blade deflected shape of in the fundamental
aerodynamics calculations. Lastly, although the same controller parameters were used in
both aeroelastic codes, the differences in structural parameters resulted in different closed-
loop behaviours. To remove this discrepancy, a future investigation must be made into tuning
the controller parameters for a better agreement of the OpenFAST and HAWC2 closed-loop
dynamics.
Despite the differences in certain responses, the results presented herein are of significant
value to other scientists who are interested in the aeroelastic response of the IEA Wind 15-MW
RWT. In particular, the use of the same turbulence boxes and the same controller sheds light
11
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
onto exactly how close the aeroelastic responses from the two codes can be. Thus, the results
serve as an essential basis of comparison for future work that may utilize different controllers,
turbulence models or other modifications to the models.
Acknowledgement The DTU authors were supported by the European Union’s Horizon 2020
research and innovation programme under the COREWIND project, grant agreement No 826042.
This work was authored in part by the National Renewable Energy Laboratory, operated
by Alliance for Sustainable Energy, LLC, for the U.S. Department of Energy (DOE) under
Contract No. DE-AC36-08GO28308. Funding provided by the U.S. Department of Energy
Office of Energy Efficiency and Renewable Energy Wind Energy Technologies Office. The
views expressed in this article do not necessarily represent the views of the DOE or the U.S.
Government. The U.S. Government retains and the publisher, by accepting the article for
publication, acknowledges that the U.S. Government retains a nonexclusive, paid-up, irrevocable,
worldwide license to publish or reproduce the published form of this work, or allow others to do
so, for U.S. Government purposes.
References
[1] Jonkman J M, Butterfield S, Musial W and Scott G 2009 Definition of a 5-MW reference wind turbine for
offshore system development Tech. Rep. NREL/TP-500-38060 National Renewable Energy Laboratory
Golden, CO
[2] Bak C, Zahle F, Bitsche R, Kim T, Yde A, Henriksen L C, Natarajan A and Hansen M H 2013 Description
of the DTU 10 MW reference wind turbine Tech. Rep. I-0092 DTU Wind Energy
[3] Rinker J and Dykes K 2018 WindPACT reference wind turbines Tech. Rep. NREL/TP-5000-67667 National
Renewable Energy Laboratory Golden, CO
[4] Bortolotti P, Tarrés H C, Dykes K, Merz K, Suthuraman L, Verelst D and Zahle F 2019 IEA wind TCP
task 37: Systems engineering in wind energy - WP2.1 reference wind turbines Tech. rep. IEA Wind URL
https://www.nrel.gov/docs/fy19osti/73492.pdf
[5] IEA Task 37 2020 IEA GitHub repository URL https://github.com/IEAWindTask37/IEA-15-240-RWT
[6] Gaertner E et al 2020 Definition of the IEA wind 15-megawatt offshore reference wind turbine Tech. Rep.
NREL/TP-5000-75698 National Renewable Energy Laboratory Golden, CO
[7] IEA Wind Task 37 2020 windio documentation URL https://windio.readthedocs.io/en/latest/
[8] Blasques J 2012 User’s manual for BECAS: A cross section analysis tool for anisotropic and inhomogeneous
beam sections of arbitrary geometry Tech. Rep. Risø-R-1785 DTU Wind Energy Roskilde, Denmark
[9] Larsen T J and Hansen A M 2007 How 2 HAWC2, the user’s manual Tech. Rep. Risø-R-1597 DTU Wind
Energy Roskilde, Denmark
[10] Hansen M, Henriksen L, Tibaldi C, Bergami L, Verelst D, Pirrung G and Riva R 2018 HAWCStab2 User
Manual Tech. rep. DTU Wind Energy Roskilde, Denmark
[11] Feil R, Pflumm T, Bortolotti P and Morandini M Journal of Composite Structures In review
[12] Dykes K, Graf P, Scott G, Ning A, King R, Guo Y, Parsons T, Damiani R, Felker F and Veers P
2015 Introducing WISDEM: An Integrated System Model of Wind Turbines and Plants URL https:
//www.nrel.gov/docs/fy15osti/63564.pdf
[13] Jonkman J and Buhl Jr M L 2005 FAST User’s Guide Tech. Rep. NREL/EL-500-38230 National Renewable
Energy Laboratory Golden, Colorado
[14] National Renewable Energy Laboratory 2019 OpenFAST Documentation URL https://openfast.
readthedocs.io/en/master/
[15] Jonkman J and Musial W 2010 Offshore code comparison collaboration (OC3) for IEA task 23 offshore wind
technology and deployment Tech. Rep. NREL/TP-5000-48191 National Renewable Energy Laboratory
Golden, CO
[16] Popko W, Vorpahl F, Zuga A, Kohlmeier M, Jonkman J, Robertson A, Larsen T J, Yde A, Saetertro K
and Okstad K M et al 2012 22nd International Society of Offshore and Polar Engineers Conference URL
https://www.nrel.gov/docs/fy12osti/54124.pdf
[17] Robertson A N, Wendt F, Jonkman J M, Popko W, Dagher H, Gueydon S, Qvist J, Vittori F, Azcona J,
Uzunoglu E et al. 2017 Energy Procedia 137 38–57
[18] Hansen M H, Gaunaa M and Madsen H A 2004 A Beddoes-Leishman type dynamic stall model in state-space
and indicial formulations Tech. Rep. Risø-R-1354 DTU Wind Energy
[19] Rinker J M 2018 Journal of Physics: Conference Series 1037 062032
12
The Science of Making Torque from Wind (TORQUE 2020) IOP Publishing
Journal of Physics: Conference Series 1618 (2020) 052052 doi:10.1088/1742-6596/1618/5/052052
[20] National Renewable Energy Laboratory 2019 ROSCO Version 0.1.0 GitHub repository URL https://
github.com/NREL/rosco
[21] Abbas N J, Wright A and Pao L 2020 vol 1452 (IOP Publishing) p 012002 URL https://doi.org/10.1088%
2F1742-6596%2F1452%2F1%2F012002
[22] Hansen M H and Henriksen L C 2013 Basic DTU wind energy controller Tech. Rep. DTU Wind Energy
E-0018 DTU Wind Energy Roskilde, Denmark
[23] International Electrotechnical Commission 2005 Wind turbines - part 1: design requirements 3rd ed.
13