Significant attention has been drawn towards the holistic test and analysis of Floating Wind Turbines (FWT) with an aim of derisking the technology. However, as the turbines grow in size, the down scaling process becomes more challenging. Being the main issue the down scaling of two fluid-structure interactions at different regimes, leading to two different scaling processes, the Froude and Reynolds law. There are different techniques to overcome this problem and most of them related to a software-in-the-loop approach. However, in this campaign, the turbine is not geometrically scaled but instead designed under constraints to generate similar effects than the full-scale turbine, at Froude scaled wind conditions, maintaining the physics in the origin of loads.
This work carried out in the deep-water basin at DHI Denmark (DHI) builds on a series of FWT experimental campaigns that are a result of collaboration between DTU, DHI and Stiesdal Offshore in 2017, 2021 and 2023. This iteration is highlighted by fully representing a FTW system on realistic scenarios. Featuring 5 x 5 independent fans that conform a 25 m2 open flow wind generator capable of generating realistic down-scaled turbulent wind flows. A 4-meter diameter rotor (1:70) designed to generate the thrust and aerodynamic damping of a down-scaled IEA-22MW turbine. A taut moored TetraSub design variant platform from Stiesdal Offshore as substructure. And a turbine active pitch control that all together allows to fully represent the technology. This campaign involved wave only, wind only and combined tests for irregular waves and constant to sheared turbulent wind profiles.
This campaign is also used as calibration and validation for a HAWC2 setup developed to include all the mentioned features, simulating the experimental tests. HAWC2 considers first and second order wave forcing (WAMIT) and Blade Element Momentum theory for the aerodynamic forcing, being considered a mid-fidelity approach, based on aerodynamic and hydrodynamic coefficients that are sea state or wind dependent. Therefore, tests such as wind and wave only are used to calibrate the model, in both the aerodynamic and hydrodynamic parts. As part of this work, a comparison between the experimental and simulated results is presented, considering wave only, wind only and wind plus wave cases, using uniformly (1D) turbulent wind forcing, highlighting the remaining challenges and advantages.
As part of the analysis, the environmental conditions, wind and waves are here described and used in a comparison of pitch performance by exceedance probability in a range of operational and extreme conditions. The exceedance probabilities are normalized in terms of its characteristic forcing condition, showing the linear or nonlinear dependency. An averaging exercise is carried out by taking the higher events of pitch response in both active control and fixed blade pitch conditions, and averaging them together with time conditioned signals. As conditioned signals both the wind and wave signals are used. These signals are averaged in the same time windows as the pitch response, allowing to identify the conditions that developed such response and their contribution, while comparing cases with or without blade pitch control.

To meet the ever-increasing renewable energy demands, floating offshore wind turbines and floating solar energy farms keep increasing in size. As a result, the complex demands on controlled laboratory tests move beyond the current state-of-the art of existing wave basin facilities. To support this trend, MARIN is developing a new high-quality test laboratory, which is an upgrade of its existing Offshore Basin.
Within this test facility large-scale, light-weight floating solar energy farms and the next generation floating wind turbines of 20MW and larger can be tested in combined wind, waves and current environmental conditions. For this purpose a new modular wind generation system is being developed alongside a wireless measurement system, an optical motion tracking system, and a dedicated wind turbine stock model for these 20+MW wind turbines.
The wind generation system consists of modular wind tunnel units including a fan, contraction nozzle, honeycombs and flow conditioners to generate a wind field of high homogeneity and low turbulence. Due to its modular units, the system can be build up to generate a 18mx3.4m wind field to test large scale floating solar energy farms floating offshore infrastructure. Alternatively, the wind field generation system can be transformed to a 9mx6.8m set-up to be able to test 20+MW floating wind turbines at scale 40-50. The fans that generate the wind can be individually controlled such that time-dependent wind spectra can be generated as well as a variation of the wind velocity over height to mimic the atmospheric boundary layer observed offshore.
Along with the new wind generation system, MARIN is developing a new stock wind turbine model to mimic a 22MW turbine at scale 1:50. Additionally, large steps are being taken to have a wireless measurement system. As renewable energy systems are relatively light-weight, the use of electrical measurement cables should be avoided as much as possible to avoid that these cables influence the platform motions. Therefore, the models can be equipped with small transducer units consisting of a battery and wireless data transfer to avoid the use of these measurement cables. Furthermore, to be able to measure the motions of large-scale floating solar farms and floating infrastructure consisting of multiple objects spread out over a large space, a new optical measurements system is being developed.
Besides providing support to the offshore industry in design verification and innovation, the new test facility is foreseen to play an important role in the validation of mulfi-fidelity hydro-aero-mooring-servo-elastic simulation methods that are being used in the innovation, design and optimization of renewable energy systems. With the larger wind generation system, the future renewable projects, i.e. large scale FOWT systems and floating solar farms, can be tested in combined environmental conditions, i.e. in multi-directional waves, physical current and high-quality wind As such these developments are an important next step in laboratory testing to provide reliable advice and high-quality data to the offshore renewable industry.

Floating Lidar Systems (FLS) enable offshore wind resource assessment where fixed masts are impractical [1]. Whereas windspeed measurements are of good quality, turbulence intensity (TI) readings are contaminated due to the motion of the FLS platform. FLS can typically be categorized into three moored, station-keeping platform types: buoy-based, ship-based and spar-buoy based systems, each with individual responses to the dynamic sea state [1], [3].
We investigated how a deterministic motion-compensation algorithm described in [3], performs across two different FLS platform designs: a buoy-based system with long-period motions and a ship-based system with shorter, higher-frequency motions.
Instead of only accounting for all DoF at once, we applied sequential DoF configurations of the motion compensation algorithm to each 10-minute interval, covering roll, pitch, yaw, heave, surge, sway, pitch/roll angular-rate terms, and shear corrections, resulting in parallel time series for the same intervals. Under the assumption that platform motion inflates wind speed fluctuations and thus TI, we identified which DoF configuration yielded the lowest standard deviation (σ). On the buoy platform, the full compensation sets consistently dominated, jointly resulting in the lowest σ in ~99.9% of intervals. This indicates that the buoy’s longer motion periods (relative to lidar sampling) result in measurable fluctuations that are effectively removed when all relevant DoFs, their angular-rates and shear effects are considered. The longer periods also simplify time synchronization, as linear interpolation of inertial measurement unit (IMU) data between the lidars line-of-sight (LoS) timestamps is adequate and less prone to timing-mismatch noise.
In contrast, the ship-based system showed a more diffuse outcome, consistent with a smaller motion imprint on TI measurements and stricter demands on sensor timing precision. Across 89.1% of intervals, motion-compensated data yielded the smallest σ (dominated by simpler configurations). In the remaining 10.9%, the raw, uncompensated data produced the lowest σ, suggesting that compensation introduced noise, likely due to sensor inaccuracies, imperfect time synchronization, or motions with periods short enough that interpolation and timestamp errors are penalized.
Overall, motion-compensated relative TI error (RRMSE, RMBE) was similarly low for both systems, although the buoy-based FLS exhibited highly overestimated raw TI. For the ship-based FLS, raw TI was already low, and compensation remained beneficial, but the motion impact was less pronounced.
This paper will show that the same algorithm behaves differently across platforms because their motion spectra differ. Increasing dominant motion frequency increases the required precision of time synchronisation and sensor/measurement accuracy. While accounting for more DoF reduces σ in low-frequency regimes, the benefits are less consistent at high frequencies and can even reverse, when synchronisation or measurement accuracy is insufficient. Beyond a point, additional DoF may therefore introduce noise and underperform simpler motion compensation strategies. Accordingly, compensation should reflect the platform’s dynamic response to the sea state, which sets the necessary sensor precision and time-synchronization tolerances. If those requirements are not met, simpler compensation strategies may perform better.

Hydrogen is a promising pathway for the integration of offshore wind energy into a renewable energy mix (Shi et al., 2023). Hydrogen produced from far offshore wind farms (OWF) could be an economically viable option. Depending on the location of the OWF, the electrolysis capacities can be deployed on bottom fixed substructures (Rudolph et al., 2023), on floating offshore wind turbines (“Dolphyn Hydrogen - phase 1 final report,” 2019) or on floating substations (Lee et al., 2023; Zhang et al., 2024). This study focuses on the last two options and presents a novel approach experimentally investigating one of the still pending questions: what is the influence of floaters motion on a Proton Exchange Membrane (PEM) electrolyser production rate?
Previous steps of the work proposed pre-designs of a 15 MW floating wind turbine and of a GW scale substation with electrolysis capacities. Hydrodynamical simulations have been carried to estimate their motion responses in different sea states. Building upon those numerical results, this presentation aims at describing the method and the results of the experimental tests conducted on a 1 kW PEM stack in order to quantify its performance under Froude scaled imposed motion.
The first part presents the experimental setup where a PEM stack is installed on top of a hexapod. This hexapod imposes sinusoidal motions representative of floater motion with applied scaling law. This approach allows for the decomposition and isolation of the different parameters susceptible of impacting the electrolyser performance. Hence offset positions and motions with different amplitude, frequencies and accelerations have been tested. The stack response is closely monitored.
The second part of the presentation covers the analyses performed on recordings of stack monitored channels. A focus is applied on quantifying the impact of motion on the PEM stack and quantifying uncertainties through a correlation analysis. Metrics wells described in literature such as polarisation curves and Electrochemical Impedance Spectroscopy (EIS) (Thomas Malkow et al., 2018; T. Malkow et al., 2018; Tsotridis and Pilenga, 2021), are combined with innovative approaches such as frequency domain analysis of high frequency resistance of the stack. An analysis quantifying the influence of Degrees of Freedom (DoF), motion parameters (i.e. frequency, amplitude…) with regards to the PEM stack performance will be presented.
Initial results indicate that an influence of the motion on the PEM stack can be observed. An approach to link the observable changes in the stack's performance metrics with the physics at the heart of the electrolyser will be presented as a conclusion.

Wind turbine rotor blades are rapidly increasing in size and a fast succession of new blade models is entering the market. Furthermore, there is a push towards life extension of existing assets and extended operating life (30-40 years) for new assets, pushing the limits of wind turbine blade design and raising questions on the lifetime assessment of rotor blades.

For this purpose, the Reliablade 2 project in the Netherlands aims to develop a digital twin framework combining structural health monitoring, modelling, and probabilistic analysis to reduce the risk of blade failure, allowing for timely inspections and preventive condition-based maintenance before catastrophic failure occurs. A set of detailed information on loading history and blade response, including damages and degradation, is necessary for the development and validation of this framework. However, this level of data is difficult to come by for blades in operational conditions.

To assess which info is essential for the purpose of the framework and to provide a first validation set for the framework, a test series has been conducted on a 3-meter long demonstrator beam designed and manufactured by Suzlon . This component represents the spar cap/shear web assembly of the wind turbine blade and was designed to investigate failure in thick adhesive bond lines. To monitor the beam during the test, it was equipped with several monitoring techniques, such as strain gauges, DIC, fibre optics, and Acoustic Emission sensors both on the inside and outside of the beam. All measurements combined form a high-density sensor network to enable an assessment of the digital twin with different levels of available data, by selecting and comparing different sensors sets for model updating and damage detection.

This presentation will present the experimental and numerical assessment of this beam and how it supports the development of a digital twin for remaining useful lifetime assessment of wind turbine blades.