Production of a Numerical Icing Atlas for Finland

Ice on wind turbine blades reduces efficiency and causes financial loss to energy companies. Thus, it is important to know the possible risk of icing already in the planning phase of a wind park. This paper presents a new Finnish Icing Atlas and the methodology behind it and is prepared by applying the mesoscale numerical weather prediction model AROME with 2.5km horizontal resolution and an ice growth model based on ISO 12494. The same meteorological dataset is used as was used in the Finnish Wind Atlas (published in 2009), and thus is fully compatible with and comparable with existing climatological wind resource estimations. Representation of the selected time period is evaluated from an icing point of view. Comparing reanalysed temperature and humidity datasets for both the past 20 years and the wind atlas period, we conclude that the used time period represents large-scale atmospheric conditions favourable for icing. We perform a series of sensitivity tests to evaluate how sensitive this ice model is to input from the weather model. The new atlas presents climatological distributions of active and passive icing periods and wind power production loss in map form for three different heights (50, 100 and 200m) over all of Finland. The results show that the risk for active icing is much greater in coastal areas, while the risk of passive icing is larger inland. © 2016 The Authors. Wind Energy Published by John Wiley & Sons Ltd.