Icing in North
The term “cold climate site” refers to a site that either shows conditions favorable for icing to occur, or which consistently experience temperatures that are lower than the operational limits of standard wind turbines (Laakso et al., 2010a). It is difficult to describe a typical cold climate site, as site conditions under the definition may vary to a great extent3. Each wind project requires a specifically chosen set of measurements to evaluate the condition of the wind site. For example, some sites may experience low temperatures, but no atmospheric icing. Another site can be mild in temperature, but show periods of heavy icing (Laakso et al., 2010a). Figure 1 shows a map of icing regions in Europe. The map does not take under consideration the local topography, which is indeed significant for the local icing climates.
Figure 1 Icing map of Europe (Laakso et al., 2010; 11).
The installed cold climate capacity in Scandinavian countries is presented in table 1.
Table 1 Existing cold climate capacity in Scandinavia in 2010 (Laakso et al., 2010).
Figure 2 Mapping the different regions
In Finland there have been reports of turbine down time due to ice and low temperature, as reported by National Wind Energy of Finland (Laakso et al., 2010b). According to the statistics the low air temperature has lowered turbine availability annually between 0.2 and 2.8% between 1997 and 2010. Depending on the year, 1 to 27 turbines have been forced to shut down due to low air temperature each year (Laakso et al., 2010b).
Icing has lowered turbine availability by approximately 1.3 % of normalized annual operational hours on average for those turbines that have been reported for icing between 1996 and 2010 (Laakso et al., 2010b). The decrease in availability due to icing has been between 0.3 % and 4.1 % / year per turbine (Laakso et al., 2010b).
However, in Norway there is no centralized system for collection of operational experience from wind farms. Therefore there are no available central data on downtime and production loss due to icing or low temperatures. There is one test turbine at Sandhaugen, close to the city of Tromsø, which has reported 20-25 icing days a year, but there are no detailed statics on failure or energy loss reported publicly (Laakso et al., 2010a).
There is an empirical relation among icing, location and seasonal icing profiles by effective hub height for each region, see figure 2. Region 3 shows the area between Sweden and Norway and the area has a mean annual loss of 7% to 13% at 700-800m hub height. Figure 4 shows how the mean annual losses are 1 % to 3% in 300-500 hub height. In Figure 5 the region 1 is presented and the mean annual losses are 0% to 0.5% in 100-300m-hub height. This is a result from 250 turbines in 10 different sites (Lars Tallhaug, 2015).
Figure 3 Region 3 – Empirical relation among icing, location and seasonal icing profile (Tallhaug, 2015).
Figure 4 Region 2 – Empirical relation among icing, location and seasonal icing profile (Tallhaug, 2015).
Figure 5 Region 1– Empirical relation among icing and location and seasonal icing profile (Tallhaug, 2015).