Posters

Abstract

One of the most effective strategies for pre-feasibility projects in the wind industry is the numerical weather prediction (NWP) method. Using wind mesoscale numerical weather prediction, early wind resource assessment studies can be performed quickly and at a minimal cost without the need for long measurement campaigns.  However, the NWP method has the limitation that it does not consider accurately the influence of local orography and forests on wind flow leading to significant errors in complex and forested terrains. Recently, many studies have highlighted the potential of combining NWP method with Computational fluid dynamics approach (CFD). The NWP-CFD approach uses the NWP wind speed data as a meteorological data input condition for computational fluid dynamics (CFD), which takes topographic features and surface roughness into account.  The objective of this work was to investigate if wind turbine production could be estimated for pre-feasibility wind studies using NWP-CFD approach in a short amount of time with acceptable errors. For this, we performed a large study in Italy over 4 different wind farms and more than 40 wind turbines using the New Wind Atlas (NEWA) based on 3 km-Weather research forecasting (WRF) simulations. This work's primary contribution is a detailed methodology following specific conditions for production assessment based on NWP-CFD. The second contribution is the global method uncertainty for the NEWA-based NWP-CFD using more than 40 validation cases in 4 Italian wind farms. We will show that combining CFD and WRF approaches can reduce the WRF simulation errors by 30-50% depending on the wind turbine. In this work absolute errors on wind farm production estimates vary from 7% to 16% that are still quite accurate for early studies without mast data. The advantages and disadvantages of this approach for wind resource assessment will also be discussed.