Posters

Abstract

This study is the continuation of previous works aimed at improving the accuracy of CFD-RANS simulations for Wind Resource Assessment. It is analyzed a complex site with a wide canopy of forests in the area selected for the development of a wind farm. Five met masts are deployed on ridges covering part of the designed wind farm. In the previous studies [1,2] the author has focused on a methodology to classify stability conditions and compute a value of Monin-Obukhov Length (MOL) to be assigned at the inlet of each simulated wind directional sector. In the software in use the MOL affects both the wind speed and the solved turbulence variables, k ande (being k the turbulent kinetic energy ande its dissipation rate), at the inlet of the computational domain. ERA5 reanalysis data, downscaled with the mesoscale model WRF, were used to generate a set of sector-wise MOLs. A dependence on the Height of the Atmospheric Boundary Layer (HABL) was also observed; therefore, the choice to improve the methodology by accounting for the HABL in addition to the thermal stability. The present work is introducing a computational method of the Height of the HABL based on ERA5 data, coupled with the MOL classification already proposed in [1,2]. A set of sector-wise couples (HABL, MOL) are now used for the same complex and forested site analyzed in the previous studies [1,2]. The accuracy of the new settings is evaluated by means of cross-predicted mean wind speeds between all five masts and reported as RMSE and MAE. The CFD simulations were run with a commercial tool, which makes use of a general-purpose finite-volume solver to integrate the RANS equations. The pressure-velocity coupling is the General Collocated Velocity (GCV) method, similar to the SIMPLEC algorithm. Turbulence is closed with the RNG k-e model, Coriolis force due to the Earth rotation is not accounted for, and buoyancy is modelled with a Boussinesq approximation. The domain was discretized with a structured mesh of hexahedral cells. The results of the CFD simulations are compared to a potential flow solver, stand-alone, and coupled with a WRF model. [1] Crasto Giorgio. Use of sector-wise MOL settings in a WRA study. WindSim User Meeting 2023 [2] Crasto Giorgio. Site specific CFD fine tuning of forest and thermal effects & comparison with other flow models. WindEurope Annual Event 2024