(a) Average wind speed, median wind speed, modal wind speed
If wind speed for a site is measured over a year, it can be easily observed from the gathered data that strong winds are often rare, while moderate and fresh winds are quite common. The wind variation for a typical site is usually described using the so-called Weibull distribution, as shown in the diagram below. The horizontal axis is the wind speed, and the vertical axis is the frequency of occurrence. The area under the whole curve is 1. This particular site has an average (mean) wind speed of about 7 m/s, a median wind speed of 6.6 m/s and the shape of the curve is determined by a so called shape parameter of 2.
The average wind speed is the average value of all the measured wind speeds. The blue line at the median wind speed divides the area under curve into two equal halves. Half of the times the wind speeds are below median wind speed, and half the times the wind speeds are above the median speed. Also note that both the average and the median values do not coincide with the modal wind speed of 5.5 m/s, which is the speed of most frequent occurrence.
(b) The Weibull shape parameter
The Weibull shape parameter determines the shape of the distribution. Two sites with wind speed distributions having the same average value but with different shape parameters will have different power densities (in W/m2 of rotor area). This is because the energy in the wind varies with the cube of the wind speed. Note that a wind turbine will be able to extract less than half of the power in the wind and convert it into electricity.
The following diagram shows a Weibull distribution with average (mean) wind speed of 5.94 m/s and shape parameter of 2.1:
If the average (mean) wind speed remains the same, but the shape parameter is changed to 1.5, then the distribution looks like this:
(c) Wind rose
In a wind rose, the horizontal plane is divided into 12 sectors of 30 degrees each, and the frequencies of occurrence of "bands" of wind speeds for wind blowing within the sector are shown in the form of contiguous bars. For example, the following diagram shows the wind rose for a typical island on the eastern side of Hong Kong.
(d) Wind measurement
Wind resource differs from location to location, particularly when the terrain of the area is hilly. The best way is to conduct actual wind measurement at or near to the proposed location.
Wind measurement is done using measurement masts mounted with anemometers, wind vanes, and temperature sensors. A data logger specially designed for wind measurement purpose is used to log the data. The logged data can be sent using GSM modem (if the location is covered by the GSM network) or other suitable wireless modems to a server or to an email account. Unless there are height restrictions, a measurement mast of 50m tall (or taller) can be used. The purpose of using tall masts is to obtain the wind data as close to turbine hub-height as possible.
Anemometers and wind vanes are mounted at different levels on the mast, so as to obtain the wind shear profile along the mast. There are guidelines governing the accuracy, calibration, and mounting arrangement of the anemometers. All these aspects are very important for getting an accurate set of wind data.
Due to the heavy lightning activities in Hong Kong during the summer months, care should be taken in the design of the lightning protection system for the mast, and surge arrestors fitted to the data logger can be used to minimize data acquisition problems.
The measurement period is usually one to two years, to gather the short-term year-round data for the site.
(e) Wind flow modelling
If the measurement site is some distance away from the wind turbine site, and the terrain is such that the wind characteristics of both sites could be somewhat different, then wind flowing modelling using special computer software (such as the Wind Atlas Analysis and Application Program (WAsP) of Riso National Laboratory, Denmark) would be necessary to infer the wind characteristics of the wind turbine site.
(f) Energy yield prediction
Having conducted the short-term wind measurement, the Weibull diagram of the site can be plotted, and the major parameters obtained (such as average wind speed) from the gathered data. Furthermore, the turbulence intensity data derived from the wind speed data will also be useful for selecting an appropriate wind turbine.
It is then necessary to carry out long-term prediction of the wind resource of the site. The long-term wind data of a reference site with similar wind exposure as the measurement site need to be obtained (in the case of Hong Kong, from the Hong Kong Observatory). The correlation parameters between the wind data of the measurement site and those of the reference site can be found using statistical techniques. These parameters are then used to obtain the long-term wind resource parameters of the site.
With the long-term wind resource parameters, the annual energy yield of a particular wind turbine model can be predicted using its power curve.
It should be noted that in the whole process, inaccuracies are inevitable (such as measurement errors) and the annual energy yield predicted is therefore subject to uncertainties.
If more than one wind turbine will be installed, it will be necessary to make use of computer software to study the annual energy yield of the whole wind farm, because of the wake effect arising from one turbine affecting the wind flow for another turbine downwind of it.