Global investment in floating wind energy is accelerating to meet 2050 decarbonization goals. Even if no more a young technology, floating wind turbines still demand tailored design methodologies to guide their development. To address this need, a systems engineering approach is crucial—one that optimizes the entire system rather than its individual components in isolation.

This research presents a study on the holistic design of floating wind turbines with the objective of minimizing the Levelized Cost of Energy (LCOE). The approach, therefore, is to design all the main components together from the very beginning, seeking the right balance among the different parts in terms of performance and cost, while at the same time using analysis and simulation tools required for the turbine’s certification. The methodology builds upon a well-established, in-house wind turbine design suite originally developed for land-based turbines and validated across multiple research and industrial applications.

The design process is organized as a sequence of nested optimization loops. It begins with the co-design of the rotor’s aerodynamic shape and the substructure, which includes the floating platform and catenary mooring lines. Subsequent stages involve model-based controller design and detailed structural sizing of the blade and tower, using load cases defined by certification standards. During this phase, the floating substructure is iteratively updated to account for mass variations within the system while ensuring compliance with ultimate limit constraints. Depending on the desired level of detail, the methodology employs models of varying fidelity—ranging from reduced-order static aero-hydro-structural models to fully coupled multi-body simulations for time-domain load assessments.

To illustrate the capabilities of this multidisciplinary, multi-level optimization approach, a reference multi-megawatt turbine originally designed for fixed-bottom installations has been adapted for a floating spar platform. The results indicate substantial redesign requirements across multiple subsystems. Notably, the rotor’s aerodynamic characteristics and the structural configurations of the blade and tower show major deviations from the baseline model.

These results highlight the strong interdependencies among the components of floating wind turbines, underscoring the importance of a holistic design philosophy. Optimizing such systems requires simultaneous consideration of aerodynamics, hydrodynamics, structural behavior, and control strategies to achieve cost-effective and reliable performance.

The final presentation will detail the proposed integrated design methodology, discuss the optimized turbine configurations, and compare them against conventional fixed-bottom reference models, demonstrating the benefits and challenges of a system-level approach to floating wind turbine design.

This study provides recommendations to reduce embodied carbon in floating offshore wind developments, based on a comprehensive Life Cycle Assessment conducted by COWI for a representative 900 MW wind farm under the Floating Offshore Wind Centre of Excellence programme, commissioned by ORE Catapult. The analysis establishes whole-life carbon (WLC) baselines for five foundation types—steel and concrete semi-submersibles, monopile, jacket and gravity-based foundations—following PAS 2080 principles across the full lifecycle (A1–C4).

Using ORE Catapult datasets and emission factors from the Carbon Trust’s Offshore Wind Footprinting Tool, the study quantifies embodied and operational emissions over a 25-year lifecycle, including uncertainty in operation and maintenance (O&M). Results show that Production (A1–A3) and O&M (B1–B8) dominate total WLC, together exceeding 85% of emissions. Floating foundations exhibit higher embodied carbon than fixed-bottom designs due to material intensity and structural complexity. Steel floating foundations are the most carbon-intensive (≈2.2–2.6 Mt CO₂e).

The study has highlighted key recommendations including:
• Low-carbon materials, such as electric arc furnace steel and geopolymer concrete, can achieve significant upfront emission reductions, alongside circular design and improved carbon governance.
• Shared anchors and synthetic moorings can deliver up to respective 6% and 3.7% whole lifecycle carbon savings.
• Optimised O&M strategies can significantly reduce embodied carbon across all foundation types. For floating foundations, marine operations improvements (in-situ repair and component replacement) can reduce reliance on fuel-intensive tow-to-port procedures, cutting up to 7.3% of operational emissions.
• Circular design should be considered where possible, including decommissioning strategies. Current decommissioning assumptions typically involve partial or full asset removal; most components are rarely reused but may be repurposed. Integrating circularity at the design stage helps minimise whole-life carbon.

Beyond technical measures, the study emphasises robust carbon management aligned with PAS 2080—setting early baselines, integrating carbon targets into project governance and using transparent, scenario-based reporting that moves beyond sole reliance on carbon payback metrics. Policy mechanisms led by government bodies—such as carbon budgets, taxation, and PAS 2080-certified supply chains—are recommended to drive industry-wide decarbonisation.

In summary, this work establishes transparent WLC baselines and actionable strategies—technical, managerial and policy-driven—to support the floating wind sector’s contribution to the UK’s net-zero ambitions.

Offshore wind offers steadier winds, larger rotors and higher altitudes, boosting AEP while reducing onshore visual impact. Yet LCOE remains high. This study tests whether load estimates based on the Davenport turbulence coherence model used in IEC 61400 are conservative, potentially increasing material intensity. This may be especially true for platform dynamics in floating configurations, which display eigenfrequencies below 0.05 Hz.
The work quantifies how measurement height affects the coherence of turbulence and how coherence in turn influences load responses of large offshore Horizontal-Axis Wind Turbines (HAWTs). Three empirical coherence models are compared: the Davenport model (IEC 61400), the Bowen model (Bowen et al., 1983), and the modified Bowen model (Cheynet, 2018/2019), which introduces an additional decay parameter to better account for limited gust size. Offshore datasets (e.g., FINO1) show that the Davenport decay coefficient decreases with measurement height; the Bowen model incorporates this height dependence, while the modified Bowen model further adds a third decay coefficient to prevent coherence approaching unity at zero frequency for non-zero separation. Stochastic wind fields are generated in MATLAB and used in OpenFAST simulations to predict the response of a 22 MW offshore wind turbine under each coherence model at different wind velocities. The analysis explores how differences in coherence formulation affect response estimates, considering variations in substructure type (fixed-bottom versus floating). The dynamic response’s quantities of interest are blade root bending moment, and mooring line stress, as well as their corresponding damage-equivalent loads. The work addresses whether reliance on the Davenport model in current standards leads to conservative design assumptions for large offshore turbines, as suggested in Cheynet (2019). Showing the impact of alternative coherence models could reveal potential for reducing material requirements without compromising safety, thereby lowering the levelized cost of energy (LCOE) and strengthening the competitiveness and sustainability of offshore wind.

Reducing platform motion and structural loads using advanced control is crucial for the scale-up and industrialization of floating offshore wind turbines (FOWTs). The Nacelle Acceleration Feedback (NAF) control gain proposed by Abbas (2021) was derived analytically, but the simulation-based optimization is required as suggested for fixed-bottom offshore wind turbine by Yamaguchi et al. (2020). NAF reduced the floater motion, but the cumulative moving distance of blade pitch angle increased. Lidar based Feed-Forward Collective Pitch Control (FFCPC) proposed by Kakuya et al. (2022) reduced both of the floater motion and the cumulative moving distance of blade pitch angle, but the optimal gain was not evaluated with considering look-ahead time. Furthermore, Feed-Forward Independent Pitch Control (FFIPC) was proposed and its effect was investigated for a semi-submersible platform by Russell (2024). Its application to a spar-type floater needs to be evaluated.
In this study, the optimal control gain for NAF is systematically investigated considering the reduction effect of floater motion and tower loading and compared with the analytical solution. Next, the optimal control parameters for FFCPC are evaluated considering the look-ahead time. Finally, the effect of FFIPC combined with NAF and FFCPC on floater motion and load reduction is evaluated for spar-type floater.
Dynamic analysis is conducted using a coupled OpenFAST and Simulink model of the NREL 5MW turbine on the OC3 Hywind spar at various wind speeds. NAF is adopted on wind turbine control and the effect of gain on power production, floater motion and structural loading is systematically investigated. An optimized gain for NAF is found to be 0.3 to minimize the standard deviation of the tower-base fore-aft moment, which differs from the analytically derived value of 0.07. Compared to baseline PI-controlled case, platform pitch motion is reduced by 36%, while the cumulative moving distance of blade pitch angle increases by 137%.
Then, the Lidar based FFCPC with NAF algorithm is developed using wind speeds measured by a simulated lidar system, scanning eight points on a circle upstream of the rotor, with considering the six-degree of floater motion. The optimal FF gain is evaluated to be 0.3 to minimize the cumulative moving distance of blade pitch angle. The optimum look-ahead time is found to be 0.28s. By using FFCPC, it is possible to reduce the cumulative moving distance of blade pitch angle by 57% comparing to using only NAF, without changing the motion reduction effect.
Finally, a Feedback IPC (FBIPC) utilizing blade root out-of-plane bending moments with d-q transformations and a Feed-Forward IPC (FFIPC) using Lidar-previewed wind data are developed. It is shown that Feedback IPC with NAFCPC and FFCPC does not suppress fluctuations in the tower base moment, but increases the cumulative moving distance of blade pitch angle compared to using NAF and FFCPC. In contrast, FFIPC with NAFCPC and FFCPC further reduces tower-base moments at low-frequency and also blade root moments at low-frequency, wave, and 1P ranges comparing to using NAF and FFCPC.