UNCERTAINTY IN WIND RESOURCE ASSESSMENT - Wind Resource Assessment: A Practical Guide to Developing a Wind Project - page 220

MCP using data from three tall towers. While the results for two of the three towers
(diamonds and squares) are generally consistent with the theoretical curve based on
Equation 15.1, the errors for the third tower (open triangles) are considerably smaller
than predicted.
Historical wind records suggest that the standard deviation of annual mean wind
speeds typically falls between 3% and 6% depending on the location and data source.
In some regions, the variation may be either larger or smaller. Absent a thorough
analysis of historical wind data for a particular region (something that is not always
possible given problems with the long-term consistency of wind measurements), it
is reasonable to assume a value of 4%. For 1 year of overlapping data, a reference
record ranging from 7 to 15 years, and a correlation factor ranging from 0.6 to 0.9, the
result of Equation 15.1 is an uncertainty range of 1.6-2.8%. Where suitable reference
stations are lacking, the uncertainty is simply that of the period of on-site observation
(for example, for 1 year, 4%).

15.3 FUTURE WIND RESOURCE

The uncertainty in the future wind resource can be divided into two components:
that due to normal variability of the wind climate and that due to the risk of long-
term climate change. Assuming the two components are unrelated, the individual
uncertainties can be combined by taking the square root of the sum of the squares, 2
as follows:

2
normal

2
climate

σ future = σ + σ (15.2)

For the first component, σ normal , the same interannual variability used for estimating

the historical climate uncertainty can be assumed. Adapting Equation 15.1, we have,

σ R
N p σ normal = (15.3)

where N p is the number of years over which the average is to be calculated. This
could be the financial horizon of the wind project investment, which typically ranges
from 10 to 25 years. Assuming an interannual variation of 4%, the normal component
of the uncertainty is 1.3% over 10 years and 0.8% over 25 years.
Although the climate-change component of the uncertainty is more speculative,
it should not be ignored. Considering the studies conducted to date (reviewed in
Chapter 12), a plausible range of uncertainty due to climate change is 0.5-2%. The

2 The sum of the squares rule for independent sources of error can be derived from the equation for the
variance (standard deviation squared) of a sample of data. When two large samples of data are combined,
the variance of the combined distribution contains linear cross terms that tend to cancel out because they are
randomly distributed about the mean, leaving only the squared terms. The result is that the variance of the
combined distribution equals the sum of the variances of the individual distributions. For more information,
see any standard text on statistics.