Preprints
https://doi.org/10.5194/wes-2026-187
https://doi.org/10.5194/wes-2026-187
16 Oct 2026
 | 16 Oct 2026
Status: this preprint is currently under review for the journal WES.

Rain-time operation, not rainfall alone, sets wind turbine blade rain-erosion exposure: synchronized disdrometer and SCADA measurements

Yisheng Liu, Xinchang Jia, Xiaoya Liu, Meng Jin, Fang Liu, and Yafei Li

Abstract. Rain erosion of wind-turbine blade leading edges accumulates while the blades turn in rain, yet exposure assessments prescribe a fixed rotor speed and rarely test how fast the turbines actually turn while it rains. We coupled drop-resolved disdrometer rainfall to the operating records of two whole wind farms over their common observation period: one wetter (Guangxi, 513 mm of reconstructed rainfall) and one drier (Jiangsu, 242 mm). Rain-time operation does not merely refine the exposure estimate; it sets its size. The observed tip speed replaces the fixed 90 m s−1 reference, the rounded value of the fleet's observed operating-speed ceiling. Coupling it in lowers the operation-coupled observed-period index (ESCADA) to 18 % of its fixed-speed value in Guangxi and 47 % in Jiangsu, so that reference overstates exposure by factors of 2.1–5.6. The microphysical input matters far less: replacing the parameterized fall-velocity relation with the measured spectra moves the same index by only 28–38 %. The operating state can therefore change which site appears worse. Under the common reference speed the wetter farm is roughly twice as exposed (E90 0.140 versus 0.066); with operation coupled in, the two point estimates change order (0.0251 versus 0.0311), although their bootstrap intervals overlap and the ranking is not statistically separated. Exposure is concentrated in time: the 10 % most exposed rain intervals carry 75–80 % of the total and the five most exposed rain events 78–90 %. Capping the rain-time tip speed at 80 m s−1 removes 42–58 % of the exposure while touching at most 48 % of rain time, within the model framework and before power-loss or control constraints. Both indices are normalized to a literature reference coating and are screening quantities rather than lifetime predictions, and the threshold relation is extrapolated below its 90–150 m s−1 calibration range. They supply the operational loading term that previous assessments had to assume.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.
Share
Yisheng Liu, Xinchang Jia, Xiaoya Liu, Meng Jin, Fang Liu, and Yafei Li

Status: open (until 13 Nov 2026)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
Yisheng Liu, Xinchang Jia, Xiaoya Liu, Meng Jin, Fang Liu, and Yafei Li
Yisheng Liu, Xinchang Jia, Xiaoya Liu, Meng Jin, Fang Liu, and Yafei Li
Metrics will be available soon.
Latest update: 16 Oct 2026
Download
Short summary
Rain erodes wind turbine blade leading edges, but risk estimates usually assume how turbines run during rain. We measured raindrop size and fall speed at two Chinese wind farms and aligned them with the whole fleet's operating data. The turbines run faster and produce more power in rain than out of it, and using the real operating state roughly halves the estimated exposure at one site. Because a few intense events dominate the total, briefly cutting tip speed then reduces exposure sharply.
Share
Altmetrics