Wind-turbine wake effects on the rate of accumulation of fatigue damage in overhead conductors

TL;DR

This study addresses whether the UK guideline of keeping overhead conductors at least $3D$ from wind turbines is physically warranted under turbine wakes in forested atmospheric boundary layers. It employs high-fidelity wind-tunnel experiments with a scaled turbine upstream of an instrumented OHC, using Rayleigh backscattering distributed sensing to capture fine-scale, time-resolved strain data across multiple conductor heights and spacings. Fatigue is assessed via rainflow counting applied to a bending-stress history derived from the distributed strain, revealing that mean strain near the clamp is largely insensitive to wake effects, while fatigue damage depends strongly on conductor height and immersion in the wake, with sub-$3D$ spacings sometimes reducing damage. The results imply that, in forested conditions, the minimum safe distance could be reduced below $3D$ provided the conductor is not fully immersed in the wake, with significant implications for transmission planning and cost, and they offer a physically grounded basis for revising guidance under specific terrain and operating conditions.

Abstract

Guidance relating to the safe distance within which overhead conductors (OHCs) should not be built with respect to a wind turbine, and vice versa, varies from country to country. In the United Kingdom the recommendation is that OHCs are not installed within three rotor diameters (3D) of wind turbines due to concerns over wake-induced fatigue. To assess the physical basis for this recommendation, wind-tunnel experiments were conducted using a scaled wind turbine placed upstream of an instrumented conductor, in conditions representative of forested terrain. The conductor was equipped with distributed fibre-optic strain sensing based on Rayleigh backscattering, providing spatially resolved measurements at 2.6 mm resolution. Three conductor heights and four turbine-conductor spacings were tested, together with a no-turbine baseline, while maintaining a constant incident wind speed of 7 m/s. The results identify a critical region near the clamp where mean strain is maximised. Neither mean nor fluctuating strain at this location increased substantially due to the presence of the turbine wake. Strain fluctuations were dominated by aeolian vibration, with the wake increasing vibration amplitude whilst reducing its characteristic frequency. Rainflow counting fatigue analysis shows that damage accumulation depends strongly on conductor height. While fatigue damage rates increase when the conductor is fully immersed in the wake, reduced damage rates are observed for lower heights and closer spacings. These results suggest that, under forested atmospheric conditions, conductor-turbine separations smaller than the current 3D guidance may be feasible provided the conductor is not fully immersed in the turbine wake.

Figures (20)

• Figure 1: Front on schematic of the experimental setup in the wind tunnel.
• Figure 2: Side on schematic of the experimental setup in the wind tunnel giving an indication of the conductor heights $H$ tested. Note that the separation between the turbine and the conductors is not to scale.
• Figure 3: Top view schematic of the experimental setup in the wind tunnel giving an indication of the spreading of the wind-turbine wake, and the proportion of the conductor span that is exposed to the wake. Note that at the lowest conductor height it is expected that the wake will pass over the conductor for the two smallest turbine -- conductor spacings. Not to scale.
• Figure 4: Photograph of the surface roughness used in the wind tunnel to simulate an atmospheric flow over forested terrain. Visible far upstream, at the inlet to the test section, are the 1 m high Counihan-type spires and the fence. The cube roughness is visible covering the majority of the floor of the test section.
• Figure 5: (a) Non-dimensional velocity profile $U^\star$ as a function of non-dimensional distance from the ground $y^\star$. $y^\star = 1$ corresponds to the hub height of the turbine (i.e. $y = h$, simulating 122.5 m above the ground) with the velocity $U$ normalised by the mean wind speed at the hub height, i.e. $U^\star(y^\star=1) = 1$. (b) Turbulence intensity profile in the boundary layer as a function of $y^\star$.