Multi-fidelity actuator line modelling of tandem floating offshore wind turbines

Abstract. The motion of floating offshore wind turbine platforms strongly affects wake development, influencing energy production, farm layout, and turbine loads. Unlike fixed-bottom turbines, floating turbine wake interactions are less understood and require high-fidelity modeling. This study employs an Actuator Line Model approach to investigate two turbines, with the upstream platform undergoing surge and pitch motions. Multi-fidelity simulations (URANS, LES with laminar and turbulent inflow) are performed to separate the effects of platform motion and inflow turbulence on wake dynamics and to assess URANS capability in capturing floating turbine wakes. Simulations of the single floating turbine are validated against experimental load and wake data. Wake validation shows that LES with turbulent inflow best captures turbulence intensity distribution, while URANS reproduces mean velocity profiles accurately especially in the near-wake. Platform motion enhances wake recovery under laminar inflow, whereas under turbulent inflow the wake recovers faster and the effect of motion is reduced; URANS shows slower far-wake recovery compared to turbulent LES. Analysis of platform-motion-induced wake oscillations indicates that turbulent LES accurately reproduces amplitudes, while URANS underestimates them and laminar inflow LES is inadequate. Furthermore, no significant differences is found between surge and pitch cases. Wake meandering is found to be primarily driven by turbulent inflow rather than platform motion: turbulent LES captures wake displacement at a characteristic frequency, whereas URANS fails to reproduce it. The impact of the wake on a downstream turbine 5D away is finally assessed and reveals that URANS underestimates both mean values and amplitudes of the downstream turbine loads. Blade distributed load analysis shows that URANS captures platform-motion-induced variability upstream but misses turbulence-driven effects downstream. In conclusion, this study provides a detailed characterization of floating turbine wake dynamics, highlights the different accuracy of LES and URANS, and demonstrates that LES yields more reliable predictions of downstream turbine loads, essential for their structural assessment within wind farms.