No abstract

Efficient spatial and temporal resolution of simulated inflow wind fields is important in order to represent wind turbine dynamics and derive load statistics for design. Using Fourier-based stochastic simulation of inflow turbulence, we first investigate loads for a utility-scale turbine in the neutral atmospheric boundary layer. Load statistics, spectra, and wavelet analysis representations for different space and time resolutions are compared. Next, large-eddy simulation (LES) is employed with space-time resolutions, justified on the basis of the earlier stochastic simulations, to again derive turbine loads. Extreme and fatigue loads from the two approaches used in inflow field generation are compared. On the basis of simulation studies carried out for three different wind speeds in the turbine's operating range, it is shown that inflow turbulence described using 10-meter spatial resolution and 1 Hz temporal resolution is adequate for assessing turbine loads. Such studies on the investigation of adequate filtering or resolution of inflow wind fields help to establish efficient strategies for LES and other physical or stochastic simulation needed in turbine loads studies.

No abstract

Near-neutral atmospheric stability conditions form the basis for wind turbine design. This is surprising since such near-neutral conditions occur in so-called transition periods only twice each day (around sunrise and sunset). Unstable conditions occur during the day and stable conditions occur generally at night. During nighttime stable conditions, turbulence is typically generated by shear and destroyed by negative buoyancy. Wind shear (both magnitude and direction) under stable conditions is much larger in comparison to that during neutral conditions. Moreover, stable boundary layer (SBL) flows are often accompanied by low-level jets (LLJs); these LLJs can be low enough to impact today's large utility-scale turbines and thus influence loads. This study compares turbulence, turbine loads, and accumulated fatigue damage for a utility-scale wind turbine in stable versus neutral atmospheric conditions. Our focus is on the varying simulated atmospheric flows that result from (i) different surface cooling rates (which control buoyancy destruction); and (ii) different geostrophic winds (which control shear generation). Inflow turbulence time series are generated and applied over the rotor plane of a 90-meter hub-height 5MW wind turbine based on Large-Eddy Simulation (LES) with refined dynamic sub-grid scale modeling; similarly, neutral boundary layer flows are generated using conventional Fourier techniques for comparison. These simulated wind velocity fields are then fed into an aeroelastic model of the selected wind turbine and turbine fatigue loads are analyzed. Some differences are seen between fatigue loads resulting from neutral versus stable conditions but these are not significant especially when missing high-frequency content in the inflow turbulence from LES-generated flows is augmented by fractal interpolation.