Posters

Abstract

Wake effects can reduce drastically the energy production of large wind farms. In specific conditions losses led by wake can represent more than 50% of the whole wind farm yield. Offshore, wind farm sizes increase more and more reaching several GW of installed capacity. Wake is a complex phenomenon depending on atmospheric conditions (wind speed, wind direction, turbulence...), layout configuration (number of WTGs, inter-turbine spacing...) and wind turbine design (power curve, thrust curve...). Wake estimation remains a real difficulty for yield assessment engineer. Different models exist for wake with, for each, a different level of complexity and fidelity in term of simulated physic processes. Nevertheless the uncertainty of such approaches remains important. Wind turbine manufacturers and operators asses several methods to decrease wake impacts on wind farms from layout design optimization to WTG controller improvements. One of them, the wake steering, consists in changing yaw of upwind turbines to deviate the wake and reduce wind speed decrease for downwind WTGs. PYWAKE/TOPFARM and FLORIS, two python open-source libraries, respectively developed by DTU and NREL, include optimization algorithms permitting to modify yaw of each WTG of a wind farm in order to find the best yaw configuration minimizing wake effect. More, simulation of wake deflection due to yaw need specific wake model like Jimenez, Curl or Hybrid Gaussian/Curl models. EDF Renouvelables will present different results using FLORIS and TOPFARM, first on simple cases assessing impacts of wake models and layout configuration on wake deflection, then, secondly, on two realistic wind farms. AEP improvements and final yaw configurations will be clearly highlighted for each scenarii depending on selected inputs.