Novel methodology for performance characterization of vertical axis wind turbines (VAWT) prototypes through active driving mode

Vertical axis wind turbines (VAWT) are called to have an important role in the definitive penetration of
renewable energies in the near future. Although still under development, they are considered good candidates for
urban environments and offshore generation in deep waters. For its experimental characterization, wind tunnel
testing of small-scaled VAWTs is usually employed. However, self-starting issues, reduced aerodynamic torques
-for reliable measurement- and high rotational speeds -imposed by similarity- emerge as inherent operative
problems. To overcome these issues, Active Driving Mode (ADM) is commonly used to drive the turbine under
equivalent kinematic conditions introducing an external motor. Up to now, an accurate characterization of the
turbine performance, in terms of retrievable aerodynamic power, is not available for this methodology in the
open literature.
This work presents the development and application of an innovative methodology for testing VAWT prototypes
through ADM. Interesting advantages over the conventional determination of the turbine performance
using Passive Driving Mode (PDM) are shown. It is demonstrated that a deep level of characterization and hightest
control is achieved with a simple experimental set-up. In addition, the key aspects for the application of this
methodology have been identified and studied in detail, like the isolation of the mechanical losses from the blade
drag or the quantification of the parasitic drag of the turbine struts. Relevant conclusions have been revealed
concerning the marginal effect of the induction factor on the parasitic drag, or the importance of a correct blade
drag estimation for the computation of the mechanical losses. Finally, the methodology has been applied to a real
wind tunnel set-up, for the characterization of a 3-bladed VAWT with a H-rotor based on the DU 06-W-200
profile. The performance results obtained experimentally have been accurately compared to CFD simulations.