Wind Measurement for Accurate Energy Predictions – An Overview

The main tasks of a professional wind measurement, used for wind power generation applications, are wind measurement for accuracy and reliability of measuring needed for profit prognosis (predict), and wind measurement for the monitoring of already installed wind farms (verify).

The following text summarises a number of important aspects, tips and instructions - based upon years of practical work - relating to wind measurement for wind power generation. This enables the user to meet the most important conditions for wind measurement. Ammonit highly recommends that inexperienced users contact professional wind consultants at the early stages of a project.

1. Prognosis (Predict)

1.1 General demand

Before the installation of a wind farm, it is recommended to analyse the site professionally. The collected meteorological data should accurately describe the wind potential of the site. This is why the measuring systems should meet the highest quality demands concerning the accuracy and the reliability.

For the evaluation of an energy prognosis based upon the meteorological data by a wind consultant, the following additional factors are important.

The correct choice and positioning of sensors is of vital importance. Mistakes here have the consequence that the whole data material cannot be evaluated properly. If the wind consultant is relying upon data generated by an insufficient or incorrectly-installed wind measurement system, this could lead to big miscalculations and builds up the risk of economically non-viable operation of the planned investment. Small deviations could be caused simply by the use of non-calibrated anemometers, for example.

The following example shows the risk an investor takes when relying on data from a poorly installed measuring system.

Assuming, for example, an average price of 0.08 €/kWh, one ends up losing Euro 50.000 in one year for a single 1500 kWh wind power generation unit. All due to the wrong prognosis. Over the total running time of a medium-sized wind farm, the loss adds up to several million Euro. In comparison with this, even the most highly developed measuring system is not expensive.

An example, what a wrong measurement
can cause for energy prognosis:
Correct Measurement
10 m  = 4.4 m/s
30 m  = 5.3 m/s
The result is a roughness length of 0.047 m or a wind speed in 85height of 6,15 m/s A WEC with 1500 kW power and hub height of 85 m will produce on this site per year:

2.892 MWh
Possible deviations
caused by
  • No calibration
  • Wrong fitting
  • Skew wind
10 m = 4.2 m/s
30 m  = 5.5 m/s
The result is a roughness length of 0.288 m or a wind speed in 85height of 6,73 m/s By this small measurement mistake the prognosis for the same WEC is now:

3.544 MWh

Measurement error at 10 m  = - 0,2 m/s ( - 4,5% )
Measurement error at 30 m = + 0,2 m/s (+ 3,8%)

Overestimation of the energy prognosis:

= 22,5%

Measurement results can be made more accurate through higher initial levels of investment (especially an advanced calibrated anemometer). However, many mistakes can be avoided without additional costs. What is needed for this is the specialist knowledge.

1.2 Measuring sizes and heights

Graph 1
Measurement heights

The measurement categories „wind speed“, „wind direction“, „atmospheric pressure”, temperature” and „relative humidity“ will be discussed in the following passages.

1.2.1 Wind speed

The ideal approach would be to measure the wind speed at the hub height of the wind power generation station that is to be installed.

Two arguments against this are that the exact hub height is most probably not yet known, since the final decision will be made on the basis of the measurement results, and secondly, that such as high measurement tower is very expensive and difficult to install.

The alternative is to use two anemometers to measure the wind speeds at two lower heights. The height profile at this location is determined ("roughness length Z0") to calculate the wind speed at other heights.

Since the calculation with a logarithmic formula represents only idealized wind circumstances, and the difference of the average speeds in different heights is small, this implies:

On a "simple" location (flat land, no obstacles), a measuring tower with calibrated anemometers at 10 and 30 m is sufficient. In more complex areas, the lower anemometer has to be fitted higher. In order to provide the minimum spacing, the measurement tower therefore must be higher. Here measuring at 20 and 40 m or even 30 and 50 m is needed.

Choice of sensors

Graph 2

Cup anemometers are the standard way of measuring wind speeds in wind energy measuring systems.

These sensors have some problems when recording wind streams (inertia of the cups, "overspeeding" effect), but these are only of minor importance. The key features are

  1. the linearity of the electronic signal and
  2. the insensitivity of the anemometer to turbulence and skew winds caused by tower or traverses.

Wind speed transmitters with large cups show much better attributes than anemometers with cups that are small in relation to the shaft.

Opto-electronic transformers and AC-generators have proved to be the most suitable transducers. One of the reasons is that they are robust. But most opto-electronic transformers supply a much higher pulse-rate (at least 10 Hz per m/s), which is needed for the recording in short measuring-intervals or for the evaluation of turbulence.

The consequences of the resulting wrong measurements have already been described. It is therefore highly recommended to use only anemometers of category 1 (according to IEC 61400), which are individually calibrated.

Anemometer calibration

Anemometer producers guarantee a certain accuracy for their products, for example ± 0.3 to 0.5 m/s (or 3 to 5 % for speeds above 15 m/s). The measurements usually remain well within this tolerance range and this is sufficient for all needs in weather forecast and in industrial processes. For a reliable prognosis in wind energy, this tolerance is not acceptable. When using a non-calibrated anemometer, it is necessary to reckon with a reduced accuracy for the predictions. This is a risk and we recommend using only anemometers which are individually calibrated.

The anemometer should be calibrated by a specialised institute in accordance with national and international standards and regulations, which should also provide an official certificate of compliance (ISO 3966 1977, IEA-guidelines, uniform measuring process of the MEASNET-Group). The members of the MEASNET Group ( are independent, international institutes which have specialised on applications in the field of wind energy. The development of standard measuring processes and the continuous flow of experience and information makes sure that the calibration of anemometers will be handled according to the strictest guidelines.

The results of each anemometer calibration are presented in a  calibration report, which  describes precisely  which aspects of the performance of the anemometer have been measured. In addition, the measurement equipment should be described in each report, including the reference tools and their last check-up.

In really well-designed measuring projects, the anemometers will be calibrated a second time after use: this makes sure that there have been no changes while measuring. The repeat calibration is part of the standard offer from many wind experts.

1.2.2 Wind direction

Measurement height

The monitoring of the wind direction is very important for the layout design of a wind farm.

The wind direction on a location needs to be monitored only at one height. The wind vane should be fitted about 1.5 m below the top of the tower in order not to influence the top anemometer.

Choice of sensors

For determination of the wind direction, potentiometric transmitters are increasing in use, because the resolution (1°) is excellent and they consume little power. It is important to keep in mind that the outgoing signal has to cover a full 360 degrees without gaps.

Because they have only a very simple potentiometer, cheap wind vanes often show a big north-gap. These “low cost" sensors can also have only a limited safe-life, because the electro-mechanical material used in their construction is not durable enough.

1.2.3 Athmospheric pressure, temperature and relative humidity

Measurement height

The influence of athmospheric pressure, temperature and relative humidity on the energy output is of secondary importance.

The atmospheric pressure can be measured at any convenient height. The measured pressure can simply be projected. Since the equipment (pressure sensors) will require additional weather protection, it is recommended that it should be installed in the shelter box of the data logger.

The temperature probe has to be supplied with a suitable shield against weather and solar radiation, and should be fitted at a height of at least 10 m to avoid the effects of heat radiation from the ground. Additionally, this height will protect the temperature probe.

The relative humidity has no influence on the energy output, but it is helpful to know this parameter to estimate the danger of icing.

Choice of sensors

As mentioned above, the quality of the sensors is important for the significance of data.

For the temperature, mainly PT100 sensors are used. Often temperature probes are used in combination with air humidity sensors, because the additional cost is low. One should not forget to ensure that there should be a good weather and radiation shield for the temperature probe. It stops the sensor heating up even if the sun shines directly on it, without causing an air-jam.

2. Verification (Verify)

Increasingly, measurement systems for recording meteorological data are used not only to investigate potential new sites, but also in the wind farm after it has subsequently been begun operations. Running in parallel to the ongoing operations, they provide reliable meteorological data which is needed for monitoring the performance and production of the wind power generators. It is important that meteorological data loggers offer the necessary functions as standard, so that can be used for all the applications without any other special equipment being needed.

It is important to meet two requirements:

Graph 3
Predict & verify

Frequently, two masts are installed at a location. One of these will later remain standing as a reference mast when the wind farm is in operation, and the other is erected directly on the proposed site of a wind power generator. This provides data for comparison purposes so that during the subsequent operation of the wind farm, the measurements on the reference mast can be converted to the wind potential at this turbine. This means that it is important to have measurement systems which can be precisely synchronised, for example by cascading via a data connection. Properly-installed professional equipment in the direct vicinity of the wind power generators is therefore also a necessary monitoring tool for the manufacturer of the plant.

3. Setting up the tower

The most important rules for the best possible tower build-up:

Graph 4
Setting up the tower

For the other components of a measuring station (shelter box with data logger, solar-power supply and installation of a remote data transmission), the general rule is: Fit them as high as possible, but still within reach for access and maintenance. Experience shows that the solar panel and the GSM-antenna are at special risk of theft and vandalism on stations with free access. Try, therefore, to make your measuring system look less attractive. A GSM-antenna for example still works as well if an old, grey plastic-pipe is put over it. And a station which works with a little, plain solar panel, is less interesting.

Make sure after installation that the tower is absolutely vertical. If you cannot climb up the tower, you must test the orientation at the lower part with a suitable measuring tool (for example an inclinometer) and then check the tower from all sides for any bends. The human eye is, according to experience, able to notice even small deviations. Electronic tilt sensors fitted to the top of the tower are a good help for installation. The inclination of the tower can be controlled at all times, and if remote data transmission is used, the operator of the station will also be warned of any impending danger.

3.1 Avoidable mistakes

Here are some of the most common mistakes that you should avoid when installing anemometers on the measurement tower:

Graph 5
Wrong traverses

Wrong traverses

Close to the tower and traverses there is always turbulence and shading, which can have a negative effect on the measurements. Wind transmitters installed on traverses should therefore not be fitted directly on the boom. In relation to the diameter of the tower, the boom must be as short as possible. At the same time, the traverse must remain stable, so that it does not oscillate.

Shaded by wind vane

Graph 6
Shaded by wind vane

The anemometer should be streamed upon from all directions without obstacles. The important top anemometer can be fitted in an ideal position, since it is situated over the tower. However, this advantage is often ruined by fitting a wind vane right next to it!

Marion Große
Ammonit Measurement GmbH