| 29 Jun 2026
On the effects of wind shear and veer on the power of a downstream turbine
Abstract. Understanding flow behavior within wind farms remains one of the central challenges in wind energy research, where wakes play a prominent role by coupling individual machines together. In fact, turbines are frequently operated under waked conditions, with wake effects on power production further modulated by atmospheric stability at the site. This paper investigates the influence of wind shear and veer on the power output of a waked wind turbine. The analysis is based on field measurements from two aligned wind turbines operating under inflow conditions that are often strongly sheared and veered. The results demonstrate that both shear and veer significantly affect wake characteristics and trajectory. Because isolating their individual contributions from field data alone is challenging, dedicated computational fluid dynamics (CFD) simulations were performed, confirming the experimental observations and enabling the effects of shear and veer to be disentangled.
The performance of several wake models was evaluated against experimental data, showing that prediction accuracy can be improved by explicitly accounting for shear and veer effects. This improved accuracy could be leveraged to support various applications. To explore the potential benefits in one exemplary use case, we consider shear- and veer-enhanced models in the context of wake-steering wind farm control. This application, however, requires real-time estimates of wind gradients, which are not available from standard onboard anemometry. To overcome this limitation, a wind sensing technique based on blade load measurements is employed to estimate shear and veer during operation. Furthermore, the strong correlation between these two quantities observed at the test site is exploited to simplify practical implementation. Wake-steering simulations indicate that incorporating shear and veer into the control strategy can lead to improved power capture.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Wind Energy Science.
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.
The manuscript addresses a relevant topic, but the central conclusions are not supported by the analysis presented. The study is described as a head-to-head comparison of UAV and LiDAR deployments, yet no side-by-side UAV and LiDAR measurements are used. Instead, the scenarios are constructed from one LiDAR dataset combined with literature-derived uncertainty values and assumed reductions in flow-model uncertainty.
These assumptions largely determine the outcome. Flow-model uncertainty is prescribed as 8\% for LiDAR_1, 5\% for LiDAR_2 and UAV_2, and zero for UAV_4. The resulting progression in P90, NPV, and LCOE therefore follows directly from the imposed scenario structure. In particular, assigning no flow-model uncertainty to UAV_4 produces an AEP standard deviation of only 0.02\%, which is not credible for a wind-resource assessment. The model excludes long-term correction, interannual variability, wake uncertainty, power-curve uncertainty, persistent calibration bias, and correlated spatial and temporal errors. Treating measurement errors as zero-mean perturbations of 10-minute data also allows much of the uncertainty to cancel over a full year.
The increase in P50 across the scenarios is also unexplained. Reducing zero-mean uncertainty should primarily narrow the distribution, not systematically raise its median. This suggests that the reported gains may be produced by the numerical formulation rather than by a physical difference between the measurement strategies.
The financial analysis introduces further problems. Equation 2 is described as an equity NPV calculation, but it discounts after-tax EBIT using WACC and subtracts total CAPEX. The debt-sizing method, debt-service schedule, and derivation of the reported debt shares are not shown. The use of P90 rather than expected production in the LCOE denominator also makes reduced uncertainty appear as a reduction in generation cost.
There are additional inconsistencies, including the use of 20 turbines rated at 6.2 MW while describing the project as 120 MW, and the extension of conclusions from a site classified as low-complexity terrain to moderate-to-high-complexity terrain.
These are not matters that can be resolved by limited revision. The uncertainty model, scenario construction, and financial analysis would need to be reformulated, and the claimed technology comparison would require independent comparative data or a substantially more cautious scope. Therefore, the paper should be declined for publication.