In order to measure the wind resource on a site, one or more meterological measurement masts are erected. Generally, the mast height is chosen considering the hub height of the planned wind turbine. Wind speed must be measured at at least two different heights with anemometers (usually cup anemometers are used) and wind direction is determined with wind vanes. It is important to notice that a good calibration of the measuring instruments is essential.

• Measure wind speed and direction, air temperature and pressure.
• Use good quality and calibrated instruments, especially for the anemometer.
• Preferably measure at the hub height of the planned wind turbine.
• Measurements to be taken at at least two different heights to calculate the wind sheer.
• One of the anemometers must be mounted on top of the mast as to avoid disturbances caused by the structure and all instruments most be well spaced to prevent interference.
• If the terrain is complex, more than one mast may be required.

Click here for a discription of meterological mast instruments

The method to record the measurements is generally the same: quantities are recorded and stored in a data logger as 10 minute mean values, max and min values and standard deviation. Of course, to take into account seasonal variations of wind speed and direction, the minimum period of measurement must be 12 months. Monthly overview can then be provided, including:

•  Calculation of wind shear (based on the measurement of wind speeds at different heights).
•  Determination of frequency distribution (Weibull distribution) and annual average wind speeds for different heights.
•  Determination of turbulence and gust (highest measured speed as 10 minute mean value or max value).
•  Map of wind direction distribution.
•  Presentation of diurnal variations in wind speed and direction.
•  Determination of annual or seasonal average air temperature and pressure.

These measurements over a period of at least a year can be considered as a reliable database for assessing the wind resource of the site. However, to take into account the annual and long term variations of the wind regime (over periods of ten or twenty years), these results must be correlated with long term data from a near meteorology station if available. Indeed, wind turbines having generally a lifetime of twenty years, these long term variations are important to assess the economic efficiency of the project. 

To achieve this correlation, it is essential to evaluate the comparability of the terrains of both sites. The quality of the data available must also be high. Ideally, the data available must cover a period of at least ten years and must come with the same format and resolution (10 minute mean values) as the actual measurements. Finally, it is important that the long term data covers the measurement period, this overlapping period giving a good idea of the quality and the accuracy of the data.

Once this correlation has been made, a function can been found covering the whole long term period, giving that the level of accuracy between the measurements and the long term data is satisfactory. The long term wind resource can then be assessed at the measurement site. 

As we saw, the combination of these two methods is usually needed to obtain reliable estimation of the wind resource. Modelling tools are used to transfer measured wind data from a known site to another. The presence of meteorological stations in the area is very useful to estimate the wind resource on a long term period. 

The following method is generally used:

•  Measure the wind conditions on the potential site for a period of at least one year.
•  Establish a correlation between the measurements and the long term data from the close meteorological station. This comparison can be used to correct the site data which has been monitored for a relatively short period, in order to validate this database for a long time period.
•  Use a modelling tool to analyse wind variations on the site in order to find the best location of the turbine.
•  In case of construction of a wind farm, use a modelling tool and measurement sensors to determine the wake effect of each turbine on the others to determine the best layout of the park.