Canadian wind sector looks to de-icing systems to boost energy yield, revenues

Cold climate and de-icing technologies are improving the business case for wind power in Canada, but icing prediction and assessment remain problematic for the industry, experts told Wind Energy Update.

According to Bernard Moulins, commercial manager at international wind turbine design and sales company ENERCON Canada Inc., clients throughout Canada are increasingly recognizing the added value of having de-icing systems and cold climate packages.

Though most of the country is located in a cold climate environment, the incentive to install turbines with in-built de-icing technology is especially high in regions exposed to more severe icing, such as the hilly terrains in eastern Canada, not least because retrofit de-icing technology is so far limited and expensive.

Ice protection technology has been reported to increase wind turbine energy yield, and could help reduce icing-related risks and revenue losses, and improve project bankability for financiers and insurers.

Appropriate icing detection equipment, including icing sensors and/or heated ultrasonic anemometers, can also help better understand and predict the icing on a turbine.

Effects on revenues

The severity of icing is key to assessing its effect on production and revenues, and on estimating the business case for an ice protection technology.

While less severe icing (Class 1-2) would typically cut annual energy production (AEP) by up to 5%, severe icing (Class 4-5) can lead to AEP losses of more than 20%, according to the International Energy Agency (IEA) (see Figure 1).

Figure 1: IEA Task 19 Site Classification

The IEA Task 19 categorizes the severity of icing into five classes based on the duration of these types of icing and the resulting production loss. Source: René Cattin presentation at WinterWind 2016 - Validation of the IEA Task 19 Ice classification.

Wind RFP prices in Quebec currently range between 6.3 Canadian cents per kWh and 12.09 Canadian cents per kWh, for an average price of around 9 Canadian cents per kWh, said Charles Godreau, researcher at TechnoCentre Eolien (TCE).

Based on a PPA price of 9 cents per kWh, for example, the annual revenue from a 2 MW wind turbine with a typical capacity factor of 35% would be about CA$552,000 (about $418,000).

Such a wind turbine exposed to light icing, leading to 5% AEP losses, will suffer annual losses of about CA$28,000 ($21,000).

On the other hand, a turbine exposed to severe icing and 20% icing-related AEP losses could see its revenue drop by about CA$110,000 (about $83,000) per year.

An "icing double-whammy" can also affect investments – if the icing causes 10% of AEP losses, for example, then the added uncertainty will add a second penalty on the bankability of the project.

De-icing technologies

Developers requiring ice protection technologies can choose between hot-air systems, which Godreau said have been reported to reduce icing-related losses by up to 50% in some cases, and electro-thermal systems, which have been reported to cut icing-related losses by up to 80%.

However, these gain estimates require more extensive studies by the industry, he added.

Electro-thermal is generally more expensive and more efficient than hot air, according to Godreau, however lightning strikes risks must be taken into account either in the design or in the installation of such systems.

Hot air could be comparatively less expensive, but its performance when exposed to critical combinations of wind, icing and cold weather is limited, Godreau said.

Both of these active ice-protection systems also involve extra equipment that may require additional inspection and maintenance compared to a standard turbine.

However, there is currently no publicly available information on the exact cost of different de-icing systems and very little data on the availability gain of these technologies, Godreau said.

Passive ice protection systems, which do not require any activation and do not consume any power by themselves, are also on the horizon. Less expensive icephobic coating could become commercially available in the next five years, according to Godreau, but so far its performance and durability are still lacking. 

Finally, not all sites in Canada have icing that is severe enough to warrant using de-icing technology. Some owners use strategies such as wind farm curtailment or preventative stop – reduction of the ice accretion with the blade at rest instead of in motion – rather than a specific technology.

The Dokie General Partnership wind farm operated by Alterra Power in British Columbia, for example, has icing events that “don’t last for more than a few days,” said Gordon Galbraith, operations technician at Dokie General Partnership.

The wind farm project team is looking at ways to boost output by a few percent through analyzing collected data and optimizing turbine performance.

Ice prediction

Predicting icing severity and assessing the related losses early on also remain difficult because they depend a lot on a combination of factors, including the unreliable long-term historical database on atmospheric conditions, the project and the turbine location, the difficult correlation with specific turbine model behavior under icing conditions, and the lack of feedback on icing technologies, according to Sébastien Goupil-Dumont, asset manager at EDF EN Canada.

Icing is also affected by several variables, including annual weather variation and altitude. It can also be caused by freezing rain, melting snow or in-cloud icing, which occurs when super-cooled droplets of water freeze instantaneously at the contact of a surface.

This type of icing is strongly related to cloud-base height, or the distance between sea level and the base of cloud. Turbines at higher altitudes are more likely to be caught inside a freezing cloud.

Appropriate icing detection equipment, including icing sensors and/or heated ultrasonic anemometers, can help better understand the icing on a turbine.

Contrary to wind, which has a small inter-annual variability of about 6%, icing duration can vary by more than 100% from one year to the other. If the icing characterization has not been undertaken prior to facing an icing problem, it can be quite hard to tell if a wind park is facing a “typical” icing season.

Measuring icing technology efficiency

There will be a “tremendous improvement” in the next five years in assessing the two big unknowns about ice protection technologies: their efficiency – in terms of energy gained using such devices – and their life expectancy, Goupil-Dumont said. Most turbine manufacturers have only recently started commercializing these technologies.

“Only when we’ll have confidence in those numbers, will we be able to state if it is worth investing," he said.

By Zara Maung