Presentations

Abstract

The scaling in size of offshore wind turbines presents an interesting challenge for characterising the wind conditions at proposed offshore wind farm sites. Future offshore turbine technology is forecast to reach turbine rotor tip heights of over 300m, extending further into the atmospheric boundary layer. Such large rotor tip heights and rotor swept areas require an accurate understanding of the wind conditions, not just for the estimation of energy yield production and potential losses, but also as an input to the design basis of offshore foundations. With larger turbines there is also an increased likelihood of complex atmospheric phenomena, such as Low Level Jets (LLJs), impacting the rotor at these higher heights. There is, therefore, a pressing need to better understand the potential impacts of such atmospheric conditions on offshore turbine performance. Historically, there have been extensive investigations into onshore turbine performance under different atmospheric conditions (unstable/neutral/stable), however, there has been a lack of similar research presented for offshore turbines. Under stable atmospheric conditions onshore, turbulence levels are typically low (< 10%). Thus low turbulence is seen as a proxy for stable conditions. However, this proxy does not apply as easily for offshore sites, where the ambient turbulence levels are already very low (< 6%). Some industry evidence has previously been presented based on nacelle mounted Lidar campaigns to indicate that 'off design' low turbulence site conditions at offshore wind farms have a negative impact on performance. This study considers a real-world project in the US to investigate the impact of different atmospheric conditions on turbine power curves and further build on this limited evidence to date. A unique dataset from operational offshore turbines in the US has been analysed, which comprises meteorological measurements and concurrent turbine production data, covering a period of approximately 1 year. The meteorological measurements were used to filter the concurrent turbine production dataset based on different classifications of atmospheric stability. This then allowed for an investigation of turbine performance by deriving the mean power curve for different atmospheric classifications from the filtered dataset. The results revealed a difference of up to 2% in energy production between the power curves derived for stable and unstable atmospheric conditions. This work provides insights based on a real-world project which can be used to inform suitable turbine performance losses as part of pre-construction energy yield assessments. The authors shall also highlight opportunities for improvements to this study and recommendations for measurement campaigns that would allow for an improved understanding of the implications of atmospheric conditions on turbine performance, particularly as turbines continue to increase in size.