Disparity analysis for three floating wind turbine aerodynamic codes in comparison

Abstract

This paper compares the predictions from three independent aerodynamic simulation tools modelling the time varying rotor thrust and shaft power of floating offshore wind turbines (FOWTs) under different sea wave conditions. These include a Blade-Element-Momentum (BEM) model, a Free-Wake Vortex model (FWM) and a Navier-Stokes based Actuator Disc (AD) model. The study is based on the NREL 5 MW baseline FOWT installed on the OC4 DeepCWind semi-submersible platform. The rotor speed is maintained constant throughout the analysis, though different rotor tip speeds and sea wave heights and periods are considered. While the three aerodynamic models apply different approaches for modelling the wake, they are all based on a blade element theory (BET) approach for simulating the blade loads. A common set of static aerofoil data is used and corrections to the data for unsteady effects such as dynamic stall are ignored. Thus disparity between the predictions for the surging rotor is primarily due to the different numerical approaches used for modelling the FOWT wake. The time-averaged rotor thrust and power coefficients predicted by the three models were found to be in close agreement with one another at low tip speed ratios and the sea state was found to have marginal effect on these results. However, the disparity in such predictions between the three models was found to increase at high tip speed ratios, with the FWM and the AD models yielding the largest and smallest rotor thrust and power coefficients, respectively. Furthermore, the AD model was observed to exhibit the highest sensitivity to sea state, with a significant increase in the time averaged power coefficient being predicted at the most extreme wave condition.

The amplitudes in the thrust and power expressed as a percentage of the corresponding time-averaged values estimated by the three aerodynamic models were found to be in close agreement with one another for the optimal and high tip speed ratios. However, at low tip speed ratios, the BEM predictions were significantly smaller than those estimated by the FWM and AD models.