Hidden Uncertainties in Floating Lidar Buoy Time Series

Floating LiDAR buoys provide the lowest uncertainty measurements of the wind resource at hub height for offshore wind projects in the U.S. These measurements form the basis of bankable annual energy production (AEP) estimates, which dictate the cost of financing for offshore wind projects. AEP is often calculated using the 10-minute averaged time series that are provided by the floating LiDAR buoy provider. Buoy providers often incorporate additional processing and filtering of the raw LiDAR data when generating the 10-minute averaged time series. Thus, relying on the 10-minute time series data blindly, could introduce additional uncertainty into downstream calculations.

Raw floating LiDAR data require processing to properly align the directional data to true North that compensates for the yaw of the buoy caused by wind-, wave-, and ocean current-induced rotation. Additional processing may also be applied to account for tidal variations, motion compensation for pitch and roll, or additional filtering based on the rate-of-change of the measured data or other metrics. In this analysis, we investigate how much uncertainty is introduced by the raw data processing workflows of different floating LiDAR buoy providers. Since the precise details of the processing performed by buoy providers are often proprietary, we first establish a workflow to process raw the 1Hz Lidar data that results in minimally processed 10-minute time series. This workflow only accounts for the yaw of the buoy, which provides a straightforward baseline to determine if additional processing has been implemented by the buoy provider. We then apply the same baseline workflow to multiple floating LiDAR buoy campaigns deployed in the U.S. and Europe.

Next, we compare the 10-minute averaged time series data delivered by the floating LiDAR buoy providers against our own 10-minute time series of wind speed and wind direction using our baseline workflow that serves as a reference. Aggregating the differences between the reference and buoy-provided time series across multiple campaigns provides a means for assessing the uncertainties introduced by floating LiDAR providers.

Initial results demonstrate that the raw LiDAR data processing varies greatly across different buoy providers. While the processing applied by some aligns closely with our reference time series, others vary greatly leading to different estimates of wind speed and wind direction. Accounting for these processing differences is necessary to accurately quantify the wind resource uncertainty from floating LiDAR measurements.