Anemometer and LIDAR

Daniel Weber, Wintersemester 2016/2017

Since wind has an influence on basically every outdoor structure or application, measurement of wind speed and direction has been an issue to humanity for a long time. This led to several different measurement methods and devices, suited for various measuring environments.

One up-to-date application for wind measurement is in wind energy, where wind speed measurement influences every stage of a project. On a possible wind energy site, wind expertise allows the assessment of profitability prior to construction. For wind turbines in operation, wind speed and direction are crucial inputs for control systems. The rotor has to be geared into wind direction when cut-in wind speed is available, while cut-out wind speed is essential for construction safety. The following article introduces two measurement systems, anemometry and LIDAR, with a focus on wind measurement for wind energy applications.

Anemometry

An anemometer is a common device for measurement of wind speed. Its name derives from the Greek word anemos, meaning wind, added with the suffix -meter, designing a measuring device. Anemometers are among the oldest devices for wind measurement. Nowadays there are basic mechanical, as well as high-tech devices. Wind velocity measurements are generally split into standard meteorological and micrometeorological observations. The former are measured on scales from several minutes and larger, while the latter happen on far smaller scales down to instantaneous measurement. Nowadays a multitude of different anemometer types are available. Mechanical, propeller or cup anemometers have their main use in standard measurements. They are characterized by high reliability, large dimensions and delayed action, due to its moment of inertia. For short-term measurements, e.g. atmospheric turbulence, acoustic anemometers are the most widespread precision devices in wind energy.^[1] Other methods are thermal anemometers, or Laser Doppler Anemometry, which are not further regarded in this article.

Rotating Anemometer

There are two types of rotating anemometers: a cup anemometer, which has three or four cup wheels, and the propeller anemometer with propeller blades, attached to a rotating axis. While the cup anemometer has a vertical axis, consequently being independent of wind direction, a propeller anemometer has a horizontal axis, thus has to be oriented in wind direction. This is generally done by a vane.

Measurement principle

Wind speed measurement is based on proportionality of wind speed to the revolution speed of the cup or propeller rotor.^[2] There are different types of rotating anemometers, in the following two types are briefly exemplified: generator-type and mechanical-type. The classic generator type has the cup or propeller axis directly coupled to the axis of a generator. The rotating axis generates voltage proportional to the revolution speed and by this proportional to the wind speed. Another configuration is the pulse generator, that comes in various setups. ^[3] A mechanical anemometer relies on counting the number of cup revolutions. Through gears connected to the rotating axis, a counter indicates the number of cup rotations. In its easiest form, the number of revolutions has to be read by hand, which allows powerless operation. However, this setup is unusual nowadays. ^[4]

Error consideration

An important parameter for wind speed measurement with rotating anemometers is response behaviour. For cup anemometers it is influenced by frictional force and the moment of inertia. Since both parameters cannot be reduced to zero, they cause a delayed response to changes in wind speed leading to errors in wind velocity measurement. As response is faster for increasing than for decreasing wind speed, the measured average wind speed is higher than the true average. ^[5] For propeller anemometers, the reaction behaviour of the vane causes delayed response in orientation leading to errors in measurement. ^[4] Another error aspect is the horizontal measuring nature of rotating anemometers. In practice, the effect of the vertical wind component on rotation velocity has to be considered. ^[5]

Application

Rotating anemometers, especially three cup anemometers, are widely used as a routine instrument for wind speed measurement, offering simplicity in installation and reliability in operation. It is in common use in airports, at weather stations and around wind turbines. Due to its limited time resolution, rotating anemometers are mainly used for mean wind velocity measurement. A common interval for would be 10 minutes, where rotating anemometers are perfectly suitable, given a proper calibration.

Advantages

• Comparably simple
• Reliable, with little downtime
• proven for mean values

Disadvantages

• delayed response behaviour
• Not precise enough for instantaneous measurement
• Small measuring volume
• Overspeeding
• Horizontal measuring, vertical wind component may influence the result

Sonic Anemometers

Sonic anemometers use ultrasonic sound waves to measure wind speed by evaluation of the runtime between two transducers. While ultrasonic waves propagate at a speed of about 330 m/s in wind-free condition, sonic speed changes slightly in wind. Wind travelling in the same direction than sound increases sound velocity, thus reducing runtime while the opposite is true for contrary wind. ^[5]

For a set pathlength L between two transducers, travel times of an ultrasonic signal are measured from transducer 1 to transducer 2 (travel time $//<![CDATA[ \begin{array}{l}t_1\end{array} //]]>$) and vice versa (travel time $//<![CDATA[ \begin{array}{l}t_2\end{array} //]]>$). For one way of the ultrasonic signal, the wind component parallel to the path acts as tailwind, while for the opposite way it acts as contrary wind, such that the resultant flow vector magnitude V and sonic speed C are expressed by:

$//<![CDATA[ \begin{array}{l}t_1 = L/(C+V)\end{array} //]]>$ and $//<![CDATA[ \begin{array}{l}t_2 = L/(C-V)\end{array} //]]>$

where,

$//<![CDATA[ \begin{array}{l}C = (L/2) (t_1^{-1} + t_2^{-1})\end{array} //]]>$

and,

$//<![CDATA[ \begin{array}{l}V = (L/2) (t_1^{-1} - t_2^{-1})\end{array} //]]>$

Based on the measurement of two or more opposed pairs of sensors, orientated in different directions, a resultant flow vector can be calculated by its components V. ^[6]

Error consideration

The main disadvantage is the influence of the transducer supporting structure on the wind flow. This causes aerodynamic turbulence, so that the device requires thorough calibration. Another disadvantage is weather sensitivity. Sonic anemometer measurement achieves lower accuracy in precipitation, when rain drops impair the speed of sound. ^[5]

Application

This method offers a very fine temporal resolution (20Hz or better), so that it is suitable for turbulence measurements. Another asset is the lack of moving parts, that makes them more applicable for long-term use in exposed locations, where salty air or large amounts of dust can impair the functioning of traditional rotating anemometers. Depending on the configuration, measurements can be performed 1-, 2- or 3-dimensional. For wind turbine installations a 2D-sonic anemometer is the most common configuration. For these one has to differentiate devices with two ultrasound paths, having four arms, and devices with three ultrasound paths having three arms. The former is apt to measure a horizontal wind vector, however measurement error amplifies for wind in direction of one ultrasound path, due to the influence of the supporting structure. A three-path-setup is technically more complex, but it offers one ultrasound path in redundancy, leading to higher accuracy and less aerodynamic turbulence. ^[7]

Methods using heat Radiation

Another anemometer type are hot-wire-anemometers, also called thermal anemometers. These rely on the principle of heat radiation, but are only briefly introduced, since they have no common application in wind energy. Wind velocity is calculated by electrically heating a very fine wire to a temperature warmer than ambient, and subsequently measuring convective heat loss to the surrounding fluid. The air flowing past the wire, has a cooling effect whose magnitude is related to the wind speed. For most metals, electrical resistance is a function of temperature (for thermal anemometers, tungsten is widely used for this reason), hence the cooling of the wire can be measured by the variance in current and voltage. ^[7] Thermal anemometry is the most common method for measurement of instantaneous wind velocity, because of its high-frequency resolution. However, thermal anemometry is extremely delicate, so that it is not appropriate for outdoor use on wind energy plants.

Literature Anemometry

• Kraus, H.: Die Atmosphäre der Erde. Springer Berlin Heidelberg (2004), p. 34 – 37.
• Kristensen, L. Boundary-Layer Meteorology (2002) 103: 163.
• Pedersen, B. M. et al.: Recommended Practices for Wind Turbine Testing and Evaluation. 11. Wind Speed measurement and use of Cup Anemometry, iea wind (2003), p.1 – 48.
• Solov'ev, Y.P., Korovushkin, A.I. & Toloknov, Y.N. Physical Oceanography (2004) 14: 173.
• Srivastava, G. P.: Surface Meteorological Instruments and Measurement Practices. Atlantic Publishers & Dist (2009), p.180 – 233.
• Walker, I. J.: Physical and logistical considerations of using ultrasonic anemometers in aeolian sediment transport Research. Geomorphology (2005) 68:1–2, p. 57-76.

Websources

• The University of Western Australia - CITS4419 Mobile and Wireless Computing: Wind Speed Sensor

http://teaching.csse.uwa.edu.au/units/CITS4419/lectures/wk2.submissions/leix03/Wind%20Speed%20Sensor.pdf

• Japan Meteorological Agency: Chapter 4 Measurement of Surface Wind

http://www.jma.go.jp/jma/jma-eng/jma-center/ric/material/1_Lecture_Notes/CP4-Wind.pdf

References Anemometry

1. Solov'ev, Y.P., Korovushkin, A.I. & Toloknov, Y.N. Physical Oceanography (2004) 14: 173.
2. Kraus, H.: Die Atmosphäre der Erde. Springer Berlin Heidelberg (2004), p. 34 – 37.
3. Srivastava, G. P.: Surface Meteorological Instruments and Measurement Practices. Atlantic Publishers & Dist (2009), p.180 – 233.
4. Japan Meteorological Agency: Chapter 4 Measurement of Surface Wind
5. Pedersen, B. M. et al.: Recommended Practices for Wind Turbine Testing and Evaluation. 11. Wind Speed measurement and use of Cup Anemometry, iea wind (2003), p.1 – 48.
6. Walker, I. J.: Physical and logistical considerations of using ultrasonic anemometers in aeolian sediment transport Research. Geomorphology (2005) 68:1–2, p. 57-76.
7. The University of Western Australia: CITS4419 Mobile and Wireless Computing: Wind Speed Sensor

LIDAR

LIDAR is a remote sensing, non-destructive measurement technique using the scattering and propagation properties of light by gases, liquids, and solids to deduce some of their physical characteristics.

LIDAR works similar to Radar, but the main difference is the usage of laser beams instead of radio waves. It´s applications are numerous, LIDAR can be based on ground, aircraft or in space and is used in several fields, e.g. oceanography, topography or measurement of atmospheric properties. ^[1] Two different etymologic origins of LIDAR are considered. While some sources treat it as an acronym for Light Detection and Ranging or alternatively Laser Detection and Ranging, another theory claims that it is a portmanteau of light and radar. This etymology is supported by The Oxford English Dictionary.

While the physical background explained in this article is generally true for LIDAR, the main focus in this article is measurement of atmospheric properties, such as: air density, wind speed and direction, air temperature etc.

Physical fundamentals ^[2]

Distance measurement is done by runtime determination. The distance is calculated by:

$//<![CDATA[ \begin{array}{l}l=(c*∆t)/2 \qquad \qquad \qquad \qquad \qquad \qquad \qquad (eq. 1)\end{array} //]]>$

With: distance emitter-target     $//<![CDATA[ \begin{array}{l}l [m]\end{array} //]]>$

speed of light in medium           $//<![CDATA[ \begin{array}{l}c ≈ 3*10^8 m/s\end{array} //]]>$ (in air)

runtime for the roundtrip            $//<![CDATA[ \begin{array}{l}∆t [s]\end{array} //]]>$

The factor 2 originates in the runtime being measured for a roundtrip from emitter to target and back to the emitter, equivalent to the double distance from emitter to target. For a given distance, the runtime can be calculated by:

$//<![CDATA[ \begin{array}{l}∆t=(2*l)/c \qquad \qquad \qquad \qquad \qquad \qquad \qquad (eq. 2)\end{array} //]]>$

The foundation of velocity measurement using LIDAR-technology is described by the Doppler effect. Established by Christian Doppler (1803-1853) for acoustic waves, LIDAR uses the principle of optical Doppler effect, describing the frequency shift due to the relative movement of a source with respect to a recipient of waves. Another application using Doppler effect in non-destructive measurement is Laser_Vibrometry. The perceived frequency $//<![CDATA[ \begin{array}{l}f_1\end{array} //]]>$ is expressed by:

$//<![CDATA[ \begin{array}{l}f_1=f_0*(1+v⁄c) \qquad \qquad \qquad \qquad \qquad \qquad (eq. 3)\end{array} //]]>$

With:

original wavelength:                            $//<![CDATA[ \begin{array}{l}λ_0 [m]\end{array} //]]>$

original frequency:                              $//<![CDATA[ \begin{array}{l}f_0 [Hz] = c/λ_0\end{array} //]]>$

speed of light in medium:                    $//<![CDATA[ \begin{array}{l}c ≈ 3*10^8 m/s\end{array} //]]>$ (in air)

relative speed along the line of sight:    $//<![CDATA[ \begin{array}{l}v [m/s]\end{array} //]]>$

Doppler shift:                                      $//<![CDATA[ \begin{array}{l}v/c [-]\end{array} //]]>$

A second Doppler shift of f1 leads to:

$//<![CDATA[ \begin{array}{l}f_2=f_1*(1+v⁄c) \qquad \qquad \qquad \qquad \qquad \qquad (eq. 4)\end{array} //]]>$

The total shift can be determined by combining both equations, while neglecting the quadratic term for v << c (generally true in LIDAR operation):

$//<![CDATA[ \begin{array}{l}∆f=f_2-f_0=2*f_0*v⁄c \qquad \qquad \qquad \qquad (eq. 5)\end{array} //]]>$

Accordingly, the total shift in wavelength $//<![CDATA[ \begin{array}{l}∆λ\end{array} //]]>$ from the emitted wavelength $//<![CDATA[ \begin{array}{l}λ_0\end{array} //]]>$ can be derived:

$//<![CDATA[ \begin{array}{l}∆λ=2*λ_0*v⁄c \qquad \qquad \qquad \qquad \qquad \qquad (eq. 6)\end{array} //]]>$

Technical fundamentals LIDAR ^[3]

The distance to the location of scattering is calculated by runtime measurement according to eq. 1. With this approach LIDAR can be aimed at a defined distance by evaluating the recorded signal of backscattered light after the corresponding runtime (eq. 2). LIDAR being used for determination of atmospheric parameters relies on interaction of the emitted light (laser) with atmospheric constituents (e.g. aerosols) and measures a Doppler shift by comparing the frequency of the original and the backscattered light. This allows the observation of parameters such as temperature, pressure, humidity and concentration of gases. At first LIDAR is the emitter of light which is backscattered by atmospheric constituents. The relative movement of the LIDAR system with respect to the reflecting particle leads to a first Doppler shift (eq. 3). In a second step LIDAR becomes the observer of the backscattered light, evoking a second Doppler shift, due to the same relative movement (eq. 4). Eventually, the total shift is described by eq. 5.

Because of their negligible fall velocity, aerosol-particles are particularly suitable to trace air movement. ^[4] In wind energy applications, LIDAR systems are mainly used to measure wind speeds. The scanning occurs in multiple points along 2 axes, scans are usually performed in horizontal, conical or vertical sections. The laser beam is directed by an interior rotating mirror, a simplified scanning along one axis is shown in figure 7. Since only axial wind speed is measured, a general wind speed vector has to be calculated from at least 3 point measurements.

Different types of LIDAR-measurement ^[3]

Coherent or direct detection

In coherent detection, the Doppler shift is measured by comparing the frequency of the original and the backscattered light. In direct detection systems, an optical frequency analyser, typically an interferometer, is used to gain information regarding the frequency of the return signal.

Infra-red (IR) or ultra-violet (UV) waveband

Whether IR or UV light is emitted, depends on the used laser source. IR-light is reflected by aerosols (small particles, dust, salt, water droplets etc.), whereas UV-light is reflected by air molecules.

Pulsed or continuous waveform ^[5]

Pulsed LIDAR systems allow simultaneous measurements at different distances by putting changes in the backscattered signal over time in reference to various distances. This is done by runtime measurement, while considering constant speed of light. (eq. 2). Continuous wave (CW) systems, in contrast, measure wind speed at a specific location by focusing the laser beam at that location. As refraction happens along the entire laser beam, all wind speed values are integrated according to a range weighting function W(R), where R designates the range that is focussed. By this, the parts of the signal backscattered in front of and behind the desired range are suppressed. Different focal distances can be adjusted, however, as the focal distance increases, spatial averaging occurs along a greater length of the beam. Typical focal distances for a CW LIDAR are less than 200m.

Accuracy and Limits

An example for the needed accuracy in emitter- and receiver-technology is given in table 1. Calculated are the Doppler shifts in wavelength (eq. 6) and frequency (eq. 5) for UV- and IR- laser, assuming a relative velocity of 1m/s. The ratio between the frequency shift and the original frequency is 6,67E-9 indicating the needed recipient accuracy in order to have precise results.

Table 1: Wavelength- and frequency- Doppler-shifts for UV- (355nm) and IR- (2µm) laser, assuming a relative velocity of 1m/s.

In practice, values for accuracy and limits are hard to give, since atmospheric composition, weather and most important manufacturer specific factors influence the result. ^[4] gives the following values for a 2 µm Doppler LIDAR (2004) by Lockheed Martin Coherent Technologies located at Karlsruher Institut für Technologie (KIT). A test measurement on 5th October 2004 showed a velocity precision smaller than 0,1 m/s for a distance as far as 2 km. A velocity precision of 0,2 m/s is achieved for a distance of up to 3,5 km. Only with further distance precision decreases significantly. The same paper gives the results of an ultrasonic anemometer measurement for comparison of the LIDAR results with a reference measurement. In general, both measurements show good accordance with a difference of generally smaller than ±1 m/s. This divergence can be explained by measurement precision of both devices. Only later, when wind direction changes the emerging discrepancies become greater. This is due to the different measurement principles of both devices, in particular the differences in measuring volume. While the anemometer has a measuring volume of 1dm³, LIDAR measures in a cylindrical volume of 362dm³. Thus, local influences like the mast on which the anemometer is mounted, have a greater influence on anemometer measurement, while LIDAR measurement is less sensitive.

Advantages

• Contactless, long-distance measurement
• On site analysis
• Based on ground, aircraft or in space (stationary or mobile usage)
• Insensitive to sound interference
• Multi-point measurement
• Fast measurement in a defined volume
• Application in several fields
• Depending on LIDAR-configuration, many target elements possible (gas, liquid, solid)

Disadvantages

• LIDAR transmitters and receivers with sub-picometer-accuracy needed
• Eye-safety of laser light has to be considered
• Backscattering depends on aerosol concentration
• Air yields very weak return signals, providing high requirements on laser sources and receivers
• Precision dependent on atmospheric composition/weather situation

Prospects for wind energy ^[3]

Today, the main use of LIDAR measurements in wind energy lies in site assessment before construction of wind turbine installations. Further potential in optimising energy production and reducing structural loads lies in nacelle-based LIDAR-systems. This allows improved operation of wind turbines due to preview information about wind and turbulence events being favourable compared to nowadays feedback control, only reacting after the event has occurred.

Applications

• Analysis of atmospheric properties
• Speed control in traffic
• Industrial production
• Meteorology
• Wind hazard detection, for instance on airports
• Site assessment for wind energy projects

Literature LIDAR

• Dunne, F., Simley, E., Pao L.Y.: LIDAR Wind Speed Measurement Analysis and Feed-Forward Blade Pitch Control for Load Mitigation in Wind Turbines. January 2010 – January 2011, National Renewable Energy Laboratory, USA (2011), p. 8 – 17.
• Henderson, S. W.: Review of Fundamental Characteristics of Coherent and Direct Detection Doppler Receivers and Implications to Wind Lidar System Design. Beyond Photonics, USA (2013), p. 1 – 5.
• Leblanc, T., Trickl, T., Vogelmann, H: Chapter 7. LIDAR, in: Kämpfer, N.: Monitoring Atmospheric Water Vapour. Ground-Based Remote Sensing and In-situ Mehtods. Springer New York (2013), p. 113 - 158.
• Marksteiner, U.: Airborne wind LIDAR observations for the validation of the ADM-Aeolus instrument. Dissertation Ingenieurfakultät Bau Geo Umwelt, Technische Universität München (2013), p. 5 – 14.
• Mikkelsen T.: On mean wind and turbulence profile measurements from ground-based wind LIDARs: limitations in time and space resolution with continuous wave and pulsed LIDAR systems – a review. Technical University of Denmark – DTU (2009), p. 1 – 10.
• Schlipf, D.: LIDAR-Assisted Control Concepts for Wind Turbines. Verlag Dr. Hut, München (2015), p. 23 – 25.
• Werner, C.: Doppler Wind LIDAR, in: Weitkamp, C.: LIDAR. Range-Resolved Optical Remote Sensing of the Atmosphere. Springer New York (2005), p. 325-354.
• Wieser, A., Träumner, K., Arnold, K.: Windmessungen mit Doppler-LIDAR in der atmosphärischen Grenzschicht. Doppler-LIDAR wind measurements within the atmospheric boundary layer. Fachtagung „Lasermethoden in der Strömungstechnik“, Karlsruhe (2008), p. 1 – 8.

References LIDAR

1. Leblanc, T., Trickl, T., Vogelmann, H: Chapter 7. LIDAR, in: Kämpfer, N.: Monitoring Atmospheric Water Vapour. Ground-Based Remote Sensing and In-situ Mehtods. Springer New York (2013), p. 113 - 158.
2. Marksteiner, U.: "Airborne wind LIDAR observations for the validation of the ADM-Aeolus instrument." Dissertation Ingenieurfakultät Bau Geo Umwelt, Technische Universität München (2013), p. 5 – 14.
3. Schlipf, D.: "LIDAR-Assisted Control Concepts for Wind Turbines." Verlag Dr. Hut, München (2015), p. 23 – 25.
4. Wieser, A., Träumner, K., Arnold, K.: "Windmessungen mit Doppler-LIDAR in der atmosphärischen Grenzschicht." Doppler-LIDAR wind measurements within the atmospheric boundary layer. Fachtagung „Lasermethoden in der Strömungstechnik“, Karlsruhe (2008), p. 1 – 8.
5. Henderson, S. W.: "Review of Fundamental Characteristics of Coherent and Direct Detection Doppler Receivers and Implications to Wind Lidar System Design." Beyond Photonics, USA (2013), p. 1 – 5. (http://www.tsc.upc.edu/clrc/wp-content/uploads/Manuscripts/clrc2013_submission_40.pdf).