Is the free wind speed sufficient to determine aerodynamic turbine performance in complex terrain?

The performance of a wind turbine is usually characterized by its power curve, which relates the wind speed at hub height with its energy production. This relation does not take streamwise inhomogeneities of the inflow – as they might be present in complex terrain – into account. In this work, the consequences of this simplification for performance predictions are analyzed. Simulations using Reynolds-averaged Navier Stokes equations (RANS) with the k-ϵ-fp model as closure are performed. An actuator disc (AD) on the ridge of a quasi two-dimensional Gaussian hill subject to a neutral atmospheric inflow is investigated. Roughness length, hill width and thrust coefficient are varied and the respective induction evaluated. Results indicate that the induction at a given thrust coefficient depends on the terrain configuration; it can be higher or lower than predictions by momentum theory, translating to a decrease or increase in the power coefficient. In this work, a power decrease by up to 15.3 % compared to flat terrain is observed. The maximum power increase is approximately 1.6 %. It is concluded that hub height wind speed or rotor equivalent wind speed are no sufficient measures to universally characterize aerodynamic turbine performance in complex terrain.