Posters

Abstract

Lightning and thunderstorms represent a severe risk for the wind energy sector. In the USA last year, lightning damage cost the wind energy industry over $100 million, and accounted for 60% of blade losses. The damage - both visible and invisible - caused by lightning strikes following a thunderstorm is manifold. They can be material and economic: structural, mechanical, electrical - with the direct destruction or premature aging of blades, nacelles, masts, lightning protection components - as well as the loss or temporary stoppage of production. Conducting Wind Turbine (WT) inspections in the event of a thunderstorm therefore becomes essential for operators and is recommended by standards such as IEC 62305-3 or IEC 61400-24. Deciding on a check-up can hardly be done without a thunderstorm activity monitoring system. Relying on the occurrence of a thunderstorm within a zone, typically 20 km, corresponding roughly to the area around a wind farm where it can be visible, would be irrelevant and costly, as only one thunderstorm in 5 or even 10 generates lightning on a wind farm [1]. In this case, it makes more sense to install lightning counters on each WT of a wind farm: a costly choice, however, which probably explains why it has not been systematized, unlike Lightning Protection System (LPS) which are installed on each WT. The use of data from Lightning Locating System (LLS) is also an alternative for triggering targeted checks or avoiding them and requires no equipment installation on each WT. Some LLS can exceed cloud-to-ground (CG) detection efficiency higher than 97%, with a median localization accuracy of 100m (2), making the information usable for deciding whether to trigger a check. Nevertheless, and despite performance improvements in recent years, the information does not systematically enable to determine which WT has been affected and may necessitate the verification of several WTs into the park. Moreover, it is sometimes necessary to analyze and interpret the data, as the location of a CG in relation to a WT is not the only criterion to be taken into account. For example, the intensity of a CG may be a more relevant criterion than its closest distance to a WT when deciding whether to trigger a check. An analysis of 10 recent cases of lightning damage to wind turbines is a first step towards using combined criteria to determine the most relevant choices. By using both precision measurement criteria materialized by an ellipse of uncertainty, the distance and CG peak current, but also the occurrence of intra-clouds which can, in the specific field of WT, be the marker of upward lightning, the idea is to improve the diagnosis to conclude even more effectively when triggering verification. The results of this study will be presented at the Dublin workshop.