Presentations

Abstract

In recent years, ground-based lidars have gained popularity as an alternative to meteorological masts for wind resource evaluation and power curve testing due to their ease of installation and low maintenance costs. Despite their advantages, profiling lidars have not replaced traditional anemometry, and their use is only recommended in flat terrain. One of the main obstacles for fully adopting lidars for wind resource assessment is their limitation in producing turbulence intensity (TI) estimates comparable to those of cup or sonic anemometers. Due to the measurement principle of lidars, turbulence measurements are affected by noise, filtering of small scales of turbulence due to volume averaging in the probe length and variance contamination from the scanning trajectory. Generally, lidars tend to show an overestimation in TI bias of 0.65-2.40% (CFARS, 2021) compared to cup anemometers for the wind speed range 4 to 12 m/s. This overestimation is a concern for OEMs when performing site suitability studies of wind turbines since TI is one of the main drivers for fatigue loads. Consequently, there is a general opposition by the WTGs manufacturers/certification bodies to grant site suitability for wind farms designed with only lidar data. In the last years several working groups (CFARS, IEA Wind Task32/Task52, DNV JIP) are advancing the development of techniques to correct the lidar TI data and increase acceptance by the wind industry stakeholders. While a new Recommended Practice by DNV (DNV-RP-0661) provides acceptance criteria for the use of ground-based lidar TI data in simple terrain and offshore sites, there is still concern in using lidar turbulence data in complex terrain. The reason for this is the increased variance contamination from the scanning trajectory in complex terrain. To evaluate the impact of using lidar TI data in complex sites, we present here a study that compares lidar turbulence estimates from co-located lidars with met masts at 10 different sites with varying terrain complexity. Results on the comparison between lidars and cup anemometers for the wind speed range 4 to 12 m/s shows differences in TI MBE of up to 4.10%, with an average of 2.1%, much higher than the values reported by CFARS. The turbulence estimates are used as inputs for conducting WRA studies in the 10 wind farm projects. The focus of the study is on evaluating the impact on design turbulences, turbulence class, as well as wake losses and wind sector management if required. We will discuss the risks of using lidar-only data for site suitability and energy yield assessment in complex sites from the perspective of OEMs, wind farm developers, certification bodies and investors.