Posters

Abstract

General summary To date, unavailability rates of wind turbines due to the presence of icing and assumed in the development phase are theoretical values mostly based on regional meteorological data. However, the uncertainty on icing losses is linked not only to the complexity of the meteorological phenomenon itself but also to the actual behaviour of the turbines subject to this phenomenon (i.e. strategies implemented by the manufacturers). The objective of this study is to overcome these uncertainties by conducting a feedback study on actual unavailability rates due to icing and recorded on more than 300 wind farms in operation mostly located in France. This large sample has enabled to get an average loss and a dispersion rate within each area. An analysis of the dependence of the actual losses on geographical and/or climatological and/or technological parameters was carried out in details in order to get a better understanding of the estimation of unavailability rates due to icing for any kind of project. Method In the framework of this study, unavailability rates due to frost for 300 wind farms operating mostly in France over the last 3 years were analysed. This first step has enabled to get a statistical analysis on numerous French areas with different meteorological conditions based on a significant sample of data. As a second step, weather data from meteorological utility but also from meteorological masts analysed by Eoltech, were considered to establish a correlation with the actual unavailability rates observed within each region. Another analysis was carried out taking into account the turbines characteristics, in particular the presence or not of an anti-icing system. Results At the moment, the analysis of the sample enabled to set up an atlas of actual availability rates due to icing observed for French regions (average and range of possible values). The second step, currently in process, should allow to provide some guidelines to estimate unavailability rates due to frost depending on meteorological conditions and turbines characteristics. Learning objectives Although downtime events due to icing are rare or negligible for many wind farms, they can constitute for some projects a significant part of production losses. This study allows to get orders of magnitude of unavailability rates based on actual operational data for wind farms located in France. The observations made depending on meteorological conditions and turbine characteristics also enabled to get some guidelines for projects located outside France.