Environmental Engineering Reference
In-Depth Information
9
6
8
3
Wind farm
5
2
7
4
1
Figure 6.11
Wind farm located in complex terrain with surrounding NWP grid-
points
power and thrust curves and details of the surrounding orography, roughness and
obstacles. The terrain descriptions are required for the downscaling task, whereas
the wind farm details are required for the second key task of the wind power
forecasting model which is the conversion from a local wind speed and direction
forecast to a wind farm power output forecast. One simple method might involve
the wind turbine manufacturer's power curve, but an alternative approach is to use
a wind farm power curve based on the forecast wind speed and the measured wind
farm power output. The conversion can be carried out via a matrix of wind farm
power outputs for a range of wind speed and wind direction bins. An example of
such a power curve is shown in Figure 6.12.
The third major input to the wind power forecasting model is the measured
power from the wind farms. This can be dynamic if the power measurements from
the wind farms are available online or static if they are available offline only.
Measurements of other meteorological parameters from the wind farm might also
be included (e.g. wind speed, wind direction, atmospheric pressure and tempera-
ture) (see Section 5.3.4). The online measurements are usually available from the
SCADA system of the wind farm or TSO. Such measurements are not required for
all the wind farms connected to the TSO, but rather for a selected set of repre-
sentative wind farms. These online measurements then need to be up-scaled to
represent the total output of the wind farms in the TSO's area. This up-scaling is the
third key task of the wind power forecasting system. In addition, the online mea-
surements need to be pre-processed to account for consistency and for whether
wind farms or wind turbines are disconnected due to breakdown, or indeed for
failures in the SCADA system itself.
Search WWH ::
Custom Search