Environmental Engineering Reference
Five wind farms' variability - 24 hours
slightly different (or time shifted) wind regime, and so the combined wind farm
output is smoother and less variable than that from any individual wind turbine.
Figure 5.16 compares the normalised output of five equally sized wind farms
within the N. Ireland region with the averaged output of all the wind farms. The
distance between the wind farm locations ranges from 40 to 120 km, with an
average inter-site distance of approximately 65 km. It can be seen that the com-
bined output is much smoother that that of any individual wind farm. The standard
deviation of the 30 minute variations for the individual wind farms ranges from
5.5 to 8.4 per cent, expressed as a percentage of the wind farm capacity. For the
aggregated wind farm outputs, the standard deviation drops to 3.2 per cent, con-
firming the benefits of diversity. A similar study using measurements from Danish
onshore and coastal wind farms investigated the wind power gradients (15 minute
variation) for different topologies (Pantaleo et al. , 2003). For a 1,000 MW capacity,
distributed at three sites, the variability dropped by 50 per cent. Assuming perfect
geographical dispersion caused the ramping gradient to fall by a further 20 per cent.
Considering all the wind farms in the N. Ireland zone, a cumulative distribu-
tion plot can be constructed as Figure 5.17, similar to Figure 5.15. By increasing the
area of interest, short-term and local wind fluctuations will not be correlated and,
therefore, should largely balance out. Consequently, the maximum amplitude of
wind power fluctuations as seen by the power system should be reduced. For a time
delay of 30 minutes, the magnitude of the wind power fluctuations is most likely to
be less than 7 per cent of wind farm capacity, and unlikely to exceed 11 per cent.
Both figures are noticeably improved on the variability of a single wind farm.