Floating vertical axis wind turbines (FVAWT) are an interesting solution to harness wind power but are not yet as widely developed as horizontal axis wind turbines (HAWT). Therefore, for detailed design of the system, there are some challenging points to represent numerically because all the specificities of the VAWTs must be included in the global model.

This paper introduces the modelling of SeaTwirl 1MW S2 VAWT in Deeplines Wind. The turbine is a three-bladed VAWT mounted on a spar. The turbine and spar are both rotating together, connected with a bearing to the generator housing that is held in place by mooring lines. The first issue was the representation of the aerodynamic loading on the blades. Three methods were tested: multiple streamtubes, actuator cylinder and 2D (horizontal slices) vortex method. The actuator cylinder method was selected as a good compromise between load representation and computational time. The second issue was to incorporate a control system, based on the applied torque on the shaft as the blade pitch is fixed. A model of the turbine alone was used to optimize the rotational speed and controller's parameters before testing on the full model. Start-up (with grid energy), exclusion zone and a complete safety system based on DNVGL's guidelines for shutdown procedures were also included in the controller. The floater itself was represented by beam elements with hydrodynamic loading provided by Morison's formulation. The structural properties of the blade were obtained with NuMAD from Sandia National Laboratories. As the spar is rotating in the fluid, a Magnus effect was added when current is present, according to results from tank tests performed by SSPA. The resulting lineic drag and lift forces are a function of the speed ratio (velocity at spar outer radius divided by current velocity). Furthermore, a hydrodynamic friction torque was applied as it can influence electrical production.

The global model was tested on various environments (wind, wave, current) and shutdown procedures. The objective of the simulations is to provide data for assessing the integrity of the whole system as well as to check the performance in terms of electrical production. Therefore, global data were processed (floater's motions and accelerations, turbine performance), but also loads, stress or strains at all stations in the blades, struts, tower, shaft, and mooring lines. Those post-treatments generate approximately one gigabyte of data per case.