Researchers at the University of Stavanger fabricated six versions of the S810 airfoil, a profile from the NREL S-series family designed specifically for horizontal-axis wind turbines. One preserved the original clean contour, while the other five carried controlled material removal centered on the stagnation region, with nominal erosion depths of 0.1, 0.2, 0.5, 0.8, and 1.2 millimeters. These staged geometries were defined in a CAD environment and mapped onto the airfoil profiles by surface projection, guided by the erosion pattern classification of Gaudern and conceptually aligned with the morphological categories of IEA Wind Task 46, which progress from pinholes and shallow pits to connected patches, coating loss, and eventually extensive material removal with laminate exposure.

All six sections were produced by high-resolution fused-filament fabrication with PLA filament at a layer height of 0.1 millimeters, giving each model a chord of 180 millimeters and a span of 410 millimeters. Because every specimen, including the clean baseline, shared the same printing process and layer texture, printer-induced surface features formed a common baseline, and all aerodynamic degradation was evaluated relative to this as-printed reference state. The authors stress that the geometries represent nominal CAD-defined erosion states rather than metrology-validated scans of field-eroded blades.

Testing took place in the low-speed open-circuit wind tunnel at the University of Stavanger, driven by a 22 kW fan and capable of freestream velocities up to 43 meters per second. With honeycomb flow conditioning and multiple mesh screens, turbulence intensity stayed below 0.5 percent. Each blade section was mounted vertically on a rotary table and measured with a six-axis force and torque transducer, with signals acquired at 1 kilohertz and low-pass filtered. The freestream velocity was fixed at 40 meters per second, corresponding to a Reynolds number of approximately 5.3 times ten to the fifth, and the angle of attack was swept from 0 to 10 degrees in steps of 2.5 degrees, covering the pre-stall regime where aerodynamic efficiency peaks for the S810 profile.

The measurements revealed a clear and systematic degradation with increasing erosion severity. The clean airfoil reached a maximum lift coefficient of about 0.92 near 10 degrees angle of attack, while the most severely eroded configuration achieved only roughly 76 percent of that peak. Lift saturation also shifted toward lower angles as erosion deepened, indicating earlier flow separation induced by the roughened leading edge. Drag rose monotonically across all eroded configurations, and the lift-to-drag ratio, the key measure of aerodynamic efficiency, declined steadily from the clean baseline to the extreme case. Comparisons with primary NREL benchmark data for the S810 showed very good agreement in lift-curve slope, confirming that the setup reliably reproduced the airfoil’s fundamental behavior.

To scale these airfoil-level findings to a full rotor, the team propagated the measured lift and drag polars through a Blade Element Momentum Theory model built on the open-source pyBEMT framework, using a representative multi-megawatt turbine consistent with the NREL 5-MW reference turbine, with a 126-meter rotor diameter and three blades discretized into 17 radial elements. For each erosion scenario, the inner 11 stations retained the clean polar while the outer six stations, spanning roughly 0.707 to 0.978 of the rotor radius, received the erosion-specific polar, reflecting the fact that erosion damage concentrates in the outer blade region where local velocities and torque contribution are highest. The maximum power coefficient fell from approximately 0.382 for the reference case to 0.377, 0.372, 0.364, 0.356, and 0.348 for the five eroded levels, representing reductions of 1.3, 2.7, 4.7, 6.8, and 9.0 percent respectively.

The physical mechanism behind these losses is coherent. Leading-edge roughness increases boundary-layer instability and accelerates the transition from smooth laminar flow to turbulence, raising skin friction drag. Geometric disruption of the contour alters the stagnation region and pressure recovery, while earlier separation reduces maximum lift and increases drag. Because these deficits concentrate in the outer blade sections, where local flow speeds are greatest, even shallow leading-edge deterioration produces noticeable rotor-scale penalties. Under below-rated operation, where aerodynamic efficiency governs energy capture, a first-order approximation links relative changes in annual energy production directly to relative changes in the power coefficient, meaning a 9 percent reduction in peak power coefficient implies roughly 9 percent less energy capture unless mitigated through control adaptation or repair.

The study then demonstrated how these aerodynamic penalties could be propagated through a realistic site wind distribution using an illustrative SCADA-based case study on the openly available Kelmarsh wind farm dataset from the UK. The researchers derived a clean reference power curve from a year of 10-minute operational records of a Senvion MM92 turbine, applying careful filtering to screen curtailment and non-representative operating points, and then imposed the modelled power-coefficient reductions as a first-order derating. The scenario-based estimates produced energy deficits of 63.5, 132.0, 229.7, 332.3, and 439.9 megawatt-hours over the analysed period for the mild through extreme erosion cases. The authors emphasize that no direct blade-inspection evidence of erosion was available for the analysed turbine, so these figures are hypothetical performance-propagation illustrations rather than measured field losses.

Beyond the immediate energy numbers, the findings carry significant implications for maintenance planning. Leading-edge erosion is a major driver of operational expenditure through increased inspection frequency, unplanned downtime, and recurrent blade repair or recoating campaigns, yet maintenance strategies still predominantly rely on fixed intervals rather than quantified relationships between erosion severity and energy-yield impact. The study introduces an aerodynamic health index, defined as one minus the relative power-coefficient reduction, as a compact ranking metric for fleet prioritisation, and maps erosion levels to indicative maintenance responses: mild and moderate erosion primarily motivate enhanced monitoring, while advanced, severe, and extreme erosion produce losses large enough to materially affect turbine-level energy capture, making repair prioritisation or corrective intervention increasingly justified.

The authors are careful to frame their contribution as a traceable integration of established methods rather than a new erosion model or validated maintenance decision model. Important limitations remain: the experiments covered a single airfoil profile at a single Reynolds number representative of laboratory-scale testing, no blockage or wall-interference corrections were applied to the reported coefficients, and the transfer of S810 erosion polars to the reference rotor geometry is a model-specific simplification. The power-coefficient penalties should therefore be viewed as relative sensitivity bounds for an outer-span erosion scenario rather than field-validated losses. Future work should validate the erosion-performance relationships with long-term SCADA data from operating turbines, multi-Reynolds-number experimental campaigns, and inspection-informed spanwise damage maps, potentially integrating drone- or lidar-based erosion inspection records with operational performance trends to build site-specific degradation trajectories for erosion-aware maintenance planning.

Subject of Research: Experimental evaluation of progressive leading-edge erosion as an aerodynamic failure mechanism affecting wind turbine performance and maintenance planning

Article Title: Experimental evaluation of progressive leading-edge erosion as an aerodynamic failure mechanism with implications for wind turbine maintenance planning

Article References: Ali, W., Teigen Giljarhus, K. E., Delimitis, A., & El-Thalji, I. (2026). Experimental evaluation of progressive leading-edge erosion as an aerodynamic failure mechanism with implications for wind turbine maintenance planning. Results in Engineering, 32, Article 113411. https://doi.org/10.1016/j.rineng.2026.113411

Image Credits: AI Generated

DOI: 10.1016/j.rineng.2026.113411

Keywords: wind energy, leading-edge erosion, wind turbine blades, aerodynamics, wind tunnel testing, blade element momentum theory, SCADA, maintenance planning, power coefficient, annual energy production, Experimental, evaluation