Ice throw from wind turbine

Positions of found ice fragments on the observation site from one selected icing event. Circles indicate the tower positions. X indicate ice fragments. Different colors signify different turbines. Dotted lines indicate the most probable rotor position.

Ice formation on wind turbines is a complex phenomenon, depending on multiple influence factors such as rotational speed of the turbine, air temperature, rotor blade temperature, liquid water content of the air and droplet size distribution. This leads to different forms of ice on turbines, which are roughly classified as clear ice and rime ice. Clear ice has a much higher density. While rime ice is more frequently found on wind turbines under icing conditions typical for European sites, clear ice formation is typical for freezing rain, which causes large ice amounts in short time frames and occurs for example more frequently at Canadian sites.

Related to rotating systems a distinction must be made between ice throw, where ice ablates from the rotating blades of the turbine and is thrown away, and ice shed, where ice falls off the stationary blades and is transported by the wind Large wind turbines (in the megawatt range) make use of appropriate sensors to detect ice and shut down or curtail automatically during icing conditions to prevent or reduce ice throw A qualitative comparison of the observed ice formation on SWT (small wind turbines) with the formation on large wind turbines (LWT), shows that the aggregated ice mass relative to the blade surface was significantly larger on small wind turbines.

Observation of ice accumulation and fragments from a study 2016-2018

Observations of the formation of ice were made at the SWT (small wind turbines) test site in Lichtenegg (Austria) 2016-2018. On three days, icing intensity could be measured on a total of 11 turbines and the ice mass per meter rotor blade could be estimated. Linear ice mass density was found to be in the range of 0,6-6,5 g per cm blade, with median of 1,6 g/cm. This results in a total mass of 290 g for a three-bladed turbine with 60 cm blades. It was found that icing does not occur uniformly along the blade but increases linearly to the tip for large wind turbines. This can be explained by partial throw of accumulated ice and subsequent continuation of ice formation. Due to the varying environmental conditions at the different events, resulting in different icing strength and density of ice formation, quantitative analysis was limited to one event on 08.01.2016, with a high number of thrown, risk-relevant fragments. Positions of found ice fragments on the observation site from the selected icing event can be seen in the figure. Technical data about the wind turbines in figure below.

Technical characterization of the three turbines in the study where multiple ice fragments where found.

Homola et al. (2009) show possible variations of icing shapes. Risk relevant ice fragments need to reach a relevant mass and their maximum size is limited by the maximum ratio of length and thickness which again is determined by the brittleness of the relevant ice type.

Using a limit of 40 J for potentially deadly ice fragments as suggested in van den Bosch et al. (1992) , this limits risk relevance for fatalities to fragments with a weight above 200 g. To include minor injuries, the impact energy limit for risk relevance should be reduced to 20 J, which results in 100 g for fragment mass.