|RER Renewable Energy Research||français|
This technical report was authored by the principal of RER Renewable Energy Research when he was an employee of GPCo Inc.
Ross, Michael. Impacts of Turbine Icing at Rumble Ridge. Report to BC Hydro. Varennes, Québec: GPCo Inc., 2002. (Technical Report)
The operation of wind turbines at sites prone to ice accumulation creates particular challenges that must be addressed through appropriate equipment selection and operating procedures. This report provides background information on icing, how it accumulates on turbine blades and the problems it causes for wind turbines. It also examines the icing conditions that will be encountered at Rumble Ridge, the proposed site for a 10 to 40 MW wind farm, and provides guidance on the selection of turbines for this site.
Glaze and rime are two forms of ice that accrete on turbine blades and reduce their aerodynamic performance significantly. Glaze is smoother than rime, tends to follow the contour of the blade, sheds more easily from the blade, and is therefore less problematic than rime, which is rough and can form large, long-lived accumulations.
Past experiences with turbines operating in icing environments have revealed three problems: increased structural loads, reduced power production, and safety concerns about falling and flung ice. Operation under heavy icing loads or during periods when asymmetric icing of the rotor causes vibrations is not recommended. At sites with severe icing, power production losses of 10 to 20% of annual production have been attributed to aerodynamic changes in the iced blade, turbine shut down due to iced instruments, and vibrations caused by asymmetric accumulation.
The principal forms of icing at Rumble Ridge will be:
Glaze and hard rime will accrete during a significant portion of the time that the site is below freezing. Wet snow and soft rime accretions will be both less frequent and occur at low wind speeds, and therefore have only a minor impact on energy production. Both glaze due to freezing rain and frost will be very rare. All types of accretions will tend to be short-lived due to frequent temperature excursions to levels above freezing.
At Rumble Ridge, electrical production losses due to icing would be about 20% of annual energy production if the turbines were shut down whenever some form of ice was present on the instruments; this is an upper limit. In reality, production losses are more likely to fall in the range of 5 to 10%. Turbines are likely to be operating with some ice on them 10 to 15% of the time, but accretions likely to cause heavy loads or serious imbalance would not occur more than 5% of the time. Falling or flung ice will be a safety hazard 10 to 15% of the time.
The following graph shows the fraction of each month during which measured data from the winter of 2001/2002 revealed that there was some ice present on one or more instruments at Rumble Ridge. Radiosonde data from Port Hardy was also used to estimate the fraction of each month during which ice would be present; for the months of September 2001 through January 2002, these estimates match the observed fraction very closely. Based on 12 years of radiosonde data, it appears that ice was more prevalent during the winter of 2001/2002 than is normal.
Since a blade heating system would increase the capital cost of a turbine by a fraction comparable to the fraction of potential production that would be lost to ice without blade heating, it is difficult to justify the installation of anti-icing technology at Rumble Ridge. Furthermore, no turbine manufacturer currently offers a tested blade heating technology.
Created 2005/07/18 Updated 2012/10/02 ©2005-2012 RER Energy Inc.