Presentations

Abstract

Heavy precipitation and strong winds can cause leading edge erosion on wind turbine blades during operation, impacting on their aerodynamic performance resulting in lower energy production over time.  The AIRE HE project (Advanced study of the atmospheric flow Integrating REal climate conditions) aims to develop and improve models and tools to reduce the impact of erosion on wind turbines on onshore and offshore locations.  Inside the AIRE project context, this work evaluated the accuracy of three numerical models to predict rain and windspeed. The models are (1) the second Modern-Era Retrospective analysis for Research and Applications MERRA2, (2) the fifth generation ECMWF reanalysis for the global climate and weather ERA5 and (3) Weather Research & Forecasting Model (WRF), with four different configurations: the original NEWA project configuration, and derived setups testing different microphysics settings.  The evaluation was made comparing the models with in-situ records from 28 weather stations located in Navarra, Spain, across sites with a wide range of terrain complexities, surface characteristics and wind climates. These masts are equipped with wind anemometers and rain gauges. The evaluation of these models involves comparing wind speed and rainfall time series. The comparison was performed between February and December 2023 (11 months).   With this analysis we aim to see the benefit of using outputs from numerical models as inputs in an erosion onset prediction model that allows us to have an estimate of blade lifetimes in wind farm sites. Blade lifetime prediction models utilize a time series of rainfall intensity, wind speeds, and a turbine-specific tip speed curve. We assessed various statistical metrics, such as correlation coefficient, bias and root mean square error. Additionally, spatial and temporal characteristics of wind and precipitation patterns were analysed to identify the strengths and weaknesses of each model.  Previous studies [1] that have used ERA5 in a blade lifetime prediction model had shown certain deficiencies in the characterization of high-intensity rainfall phenomena and underestimation of wind speed in sites with complex topography, which is expected to improve with the higher spatial-temporal resolution WRF model. The hourly values obtained confirm that, indeed, for the estimation of both the mean wind speed and accumulated precipitation the results with WRF improve those of ERA5 especially in complex topographies.   This work is supported by the AIRE HE project, which has received funding from the European Union under the grant agreement 101083716.