Wind resource and site assessment - Wind Power Generation and Wind Turbine Design - page 58

Table 2: Beam half-length versus focal distance of a continuous-wave Lidar [ 11 ].

Altitude (m) Beam half-length, L (m)

40 2.5

60 6

100 16

200 65

spectrum. An extra second scan with a near-collimated output beam is inserted
into the height cycle. The spectra thus measured are used to remove the infl uence
of the clouds at the desired measuring heights [10].
The two systems have fundamental differences concerning the measurement
volume. The continuous-wave system adjusts the focus so winds may be sam-
pled at a range of heights above ground level. The backscattered signal comes
mainly from the region close to the beam focus, where the signal intensity is at
its maximum. While the width of the laser beam increases in proportion to focus
height, its probe length increases non-linearly (roughly the square of height;
Table 2). The vertical measuring depth of the pulsed system depends on the
pulse length and is constant with sensing range [10]. A continuous-wave system
can measure wind speed at heights from less than 10 m up to a maximum of
about 200 m. Pulsed Lidars are typically blinded during emission of the pulse
and this restricts their minimum range to about 40 m, with the maximum range
usually limited only by signal-to-noise considerations and hence dependent on
conditions.
Like the Sodar, both Lidar systems rely on the assumption of horizontally homo-
geneous fl ow. In complex terrain this assumption is violated, increasingly so as the
terrain complexity increases. There are indications that errors of 5-10% in the
mean speed are not uncommon [12-14]. Only a multiple Lidar system, in which
units are separated along a suitably long baseline, could eliminate this inherent
error as explained above.
In general, care has to be taken when performing short-term measurements with
remote sensing devices. The vertical profi le varies signifi cantly with different
atmospheric stabilities. Thus the measurement campaign using remote sensing
should, similarly to cup anemometry, be a minimum of 1 year.
Currently work is in progress for a Best Practice Guideline for the use of remote
sensing.

2.3 Calibration

As explained in Section 2.2.2, the turns of an anemometer are transformed into a
wind speed measurement by a linear function. The scale and offset of this trans-
fer function are determined by wind tunnel calibration of the anemometer. Strict
requirements concerning the wind tunnel test are specifi ed in [2]. Please note that