Assessing Long-Term Hub Height Wind Specifications
Numerous wind energy yield assessments are based on measurements carried out with a met mast located at a lower height than the hub of the planned wind turbine. This article deals with the issue of vertical extrapolation of measured wind characteristics. It benchmarks four distinct empirical extrapolation methods based on a simple measurement set-up – two heights of anemometers, a wind vane, a short-term measurement of a remote sensing device (lidar, sodar) and different assumptions about the wind shear (α). These methods were tested on nine of Maïa Eolis’ 80-metre met masts in France and benchmarked to a WAsP calculation.
By Olivier Coupiac, Maïa Eolis, France
The first three methods, alpha-series, alpha1 (α1)-sampling and alpha-correlation assume the consistency of the wind shear along and above the met mast, which makes them quite sensitive to near-ground turbulence and to the height of the first anemometer. The last method, alpha2 (α2)-sampling, investigates the wind shear dependency on wind speed, direction and atmospheric stability.
A permanent challenge for industrial wind assessment is to select the most efficient way of assessing the long-term hub height wind specifications, though keeping the uncertainties at minimum. Performing long-term measurement campaigns at great heights can be incompatible to project timescale and budget. However, the vertical extrapolation can be a major source of uncertainty [refs 1, 2, 3].
Methods
The tested methods rely on a middle-term (typically one year) wind measurement campaign carried out with a met mast at a lower height than the planned wind turbine's hub (hh). We assume here a basic equipment set-up with one wind vane and anemometers at two different heights h1 and h2 (with h2>h1) in order to calculate the vertical wind shear – see Figure 1.
Alpha-series
We assume here that the vertical wind shear α2 between the mast and the hub equals at any time the wind shear along the mast α1 calculated from the two wind speeds u2 and u1 at heights h2 and h1. The hub height wind speed uhh time-series is then calculated from the highest wind speed u2 and α1.
Alpha1-sampling
Because of the near-ground turbulences, the calculation of α1 can rely on a noisy signal and result in non-realistic values for the hub height wind speed uhh. The α1 sampling method aims to bypass this drawback by averaging α1 in wind speed u2 and direction bins. Another version of this method uses atmospheric stability bins, which try to account for the strong time and seasonal dependency of the wind shear, as shown in Figure 2. The new assumption will then be that the two vertical wind shears α1 and α2 are on average equal for each direction, wind speed and stability bin.
Estimating atmospheric stability
As the measurement set-up of the met mast often does not allow a reliable estimation of atmospheric stability, this information is extracted from reanalysis data [ref. 4]. In this article we will only deal with the Monin-Obukhov length from MERRA [ref. 5] data. Monin-Obukhov length values (MOL) are sorted into three different stability classes, as detailed in Table 1.
