Presentations

Abstract

Significant benefits of collocated (or lidar-only) measurement campaigns have been observed by the wind industry. Lidars are typically cheaper and measure at more and higher heights than masts, with no planning permission or working at heights needed. Wind farm developers are motivated to accelerate its adoption for use in energy yield assessments and, where possible, move to Lidar-only campaigns. However, standalone campaigns are still in their early stages and improvements to the measurement campaign design process are needed to support their uptake. One such improvement is optimising the design of the measurement campaign, to minimise the estimated overall wind speed uncertainty, through selecting the best Remote Sensing Device (RSD) type and location for a given project. A process has been developed, that has been used to review numerous measurement campaign designs, and, where suitable, suggest improvements to the Lidar locations. Where alternative locations have been suggested, the indicative uncertainties have been evaluated and reductions in the estimated overall wind speed uncertainty of 5-40% compared to the original designs have been demonstrated. This study demonstrates the key steps taken to achieve this uncertainty reduction, and the implications it will have on project financing through helping to produce more favourable P90/P50 ratios. This reduction in the overall estimated wind speed uncertainty is achieved through optimising for two variables: firstly, the estimated topographical uncertainty, and secondly, the wind speed measurement accuracy. Practical constraints such as terrain gradient, proximity to access tracks and forestry are other considerations. Topographical uncertainty is minimised by finding locations that minimise the weighted distance and elevation difference from the Lidars to the proposed turbine locations and hub heights, and the difference in mean wind speed at these locations. Several elements contribute to the uncertainty of a Lidar at the device level, including uncertainty components related to verification, classification, mounting and the complexity of flow. All other components being equal, a key consideration for the use of lidar is the extent to which its measurement is biased, due to flow inhomogeneity across the measuring volume. The process presented in this study relies on earlier work that has shown that high resolution CFD simulations can accurately quantify this bias. From this, directional flow complexity correction (FCC) factors can be calculated and used to convert Lidar measured wind speeds to point-like wind speeds comparable to cup anemometer data. Application of these FCCs has undergone careful validation to demonstrate a significant reduction (75% on average) in the deviation between Lidar and cup anemometer measurements. Areas within a site boundary with FCCs closest to 1 are then considered representative of lowest uncertainty. An iterative and multi-variate optimisation process has been developed for larger sites with the possibility of multiple device deployments to optimise uncertainties across both the spatial and measurement uncertainties. Finally, the exploration of FCCs across a variety of sites has yielded proxies which can be explored at low-cost in early stages of development in lieu of CFD modelling. This can form the basis of further work which the industry can benefit from.