Posters

Abstract

Wind farm energy assessment ideally requires more than 10 years of data to account for seasonality, interannual variability and identify potential consistency changes; however, on-site wind measurements are generally available in the range of 1 to 2 years only, and thus, to improve the quality of the prediction, the long-term wind climate is then defined through correlations between measured data from the site and long-term reference data from a representative location, hence extending the segment of consistent records and reducing the risks related to the natural variability of wind. Italian territory has in general a complex topography and is often observed that the long-term sources do not show enough quality of correlations. This motivates the presented analysis, where R2 from different sources are investigated. . First, measured data from various masts, located mostly in Southern Italy, were investigated against the different long-term data sources by comparing the coefficients of determination (R2) used to define the quality of correlations between on-site and reference data. Reanalysis data (ERA5 and MERRA-2 from 2000 to 2020) were used as reference to determine the long-term wind speed at the position of 100 masts with at least 1-year of measurements to capture the intra-annual variability of the resource. The met-mast locations were then divided into subgroups according to the site characteristics and prevailing wind directions to investigate areas with similar features. The sites were identified as complex evaluating the land morphology and the wind variability over a buffer area around the site of interest based on the GWA maps. Additionally, virtual meteorological data (Vortex SERIES) were investigated for years between 2011 and 2020 in 19 locations: here, all three long-term data sources were compared and analyzed. The correlation values in both examined cases were computed through a DNV script which allows calculating the R2 coefficient based on a series of predefined inputs. The benchmark of R2 values defined for Italy in this study shows that ERA5 has better correlations with the measurement towers than MERRA-2, and Vortex furtherly improves the correlation quality increasing the final R2 values by 4% on average. Moreover, Vortex data introduce beneficial effects on the long-term wind speed evaluation by providing either the highest R2 value in the 58% of cases or a value close to the one calculated by ERA5, thus confirming its better results and constructively contributing to the determination of the ultimate long-term wind speed. In numbers, the three investigated long-term sources match the measured data with average index values of 0.86, 0.80 and 0.89 for ERA5, MERRA-2 and Vortex respectively. Finally, the ability of Vortex to better correlate with the site measurements in complex terrains was highlighted through the analysis of the subgroup of measurement locations with similar orography and wind regime. The uncertainties inevitably connected with the wind speed estimates are therefore reduced as well as the risks correlated to the wind farm project development.