Page 113 - Fluid Power Engineering

W i nd Measurement 91

unusual for a wind project to start with a small number of met-towers
to prospect and to subsequently add additional met-towers as fi-
nanciers demand higher accuracy in wind resources over the planned
wind farm area.
In 2009, the cost (in USD) of a wind measurement campaign is:

A fully instrumented 60-m met-tower costs about $18,000 to
$25,000.
Installation cost for a site with good access is $8,000–$10,000.
Annual data management with final measure-correlate-predict
(MCP) analysis and estimated energy computation can cost
about $16,000 to $20,000.
Decommissioning cost is $5,000 to $7,000.

Since wind is highly dependent on local conditions like contour,
roughness, and obstacles, in an ideal case, wind measurement should
be performed at the planned location of a turbine. This would yield
the highest accuracy in predicting energy output. There are two issues
with this:

1. One of the purposes of a wind measurement campaign is to
determine the best location of turbines, therefore, turbine lo-
cations are not known a priori. Unless wind measurements are
done, micrositing of turbines cannot be done.
2. When designing a wind farm consisting of multiple turbines,
it is very expensive to measure wind at each turbine location.

Therefore, a wind measurement campaign must start with what is
known and design a process that can economically generate reason-
ably accurate energy predictions. A process for designing a wind mea-
surement campaign is described below. It may be adapted to meet a
project’s specific needs.

1. Conduct a preliminary wind resource assessment of the area
under consideration. This applies to both single turbine and
wind-farminstallations.Theoutcomeofthepreliminarywind
assessment will be a wind resource map of suitable granular-
ity. For the purposes of this discussion, it is assumed that wind
resources are available for a 200 m × 200 m grid. The resource
map may be based on computer simulations using numer-
ical weather prediction models or wind resource-prediction
model like WAsP or others. The source of the wind data for
these models may be 10-m airport data and/or reanalysis data
from National Center for Atmospheric Research (NCAR). Pre-
liminary wind assessment is described in Chapter 7.
2. The preliminary location of Wind Turbine Generators (WTGs)
may be computed by running a layout optimization model