WIND FLOW MODELING - Wind Resource Assessment: A Practical Guide to Developing a Wind Project - page 184

Clearly, the success of CFD modeling is not assured, and there is a continuing need
to validate CFD results with high quality wind measurements. Problems have been
ascribed to various factors, including inaccuracies in initial and boundary conditions
(which are usually assumed to be homogeneous and follow a neutrally stratified,
logarithmic profile), limited grid resolution, and treatment of turbulence. The added
complexity of the models may be a problem, as some users may not be equipped to
run them properly. Another factor is that in general, CFD models are not designed to
take into account any circulations due to temperature gradients. The lack of a complete
prognostic equation for temperature in CFD models is, in turn, the result of another
assumption made in most CFD models, which is that the wind flow is steady state. In
a manner not unlike WAsP, most CFD models assume a constant incoming wind field.
Mesoscale Numerical Weather Prediction Models. The last class of wind
flow models covered in this chapter is the mesoscale NWP model. This type of model
has been developed primarily for weather forecasting. Like CFD models, mesoscale
models solve the Navier-Stokes equations. Unlike CFD models, however, they include
parameterization schemes for solar and infrared radiation, cloud microphysics and
convection (cumulus clouds), a soil model, and more. Thus, they incorporate the
dimensions of both energy and time and are capable of simulating such phenomena
as thermally driven mesoscale circulations (such as sea breezes) and atmospheric
stability, or buoyancy. In the world of mesoscale modeling, as in the real world, the
wind is never in equilibrium with the terrain because of the constant flow of energy
into and out of the region, through solar radiation, radiative cooling, evaporation, and
precipitation, a cascade of turbulent kinetic energy down to the smallest scales and
dissipation into heat.
Mesoscale models consequently offer considerable hope for simulating wind flows
accurately in complex terrain. They have, however, one big drawback: they require
enormous computing power to run at the scales required for the assessment of wind
projects. The typical model resolution for most mesoscale simulations is on the order
of kilometers, meaning a single grid cell is kilometers across. It is clearly impossible
to obtain a detailed picture of the wind resource within a project area at such a scale.
One way around this problem is to couple mesoscale models with a microscale
model of some kind. This could be a statistical model, if there is sufficient on-site
wind data to create reliable statistical relationships. More often, it is a simplified
wind flow model, usually either a mass-consistent model or a Jackson-Hunt model.
Examples include AWS Truepower's MesoMap and SiteWind systems (14), 3TIER's
FullView system, the Risø National Laboratory's KAMM-WAsP system (15), and
Environment Canada's AnemoScope system (16).
Research suggests, not surprisingly, that such methods can be more accurate than
simplified wind flow models where mesoscale effects play a significant role. One
example is the wind resource in a coastal mountain pass such as the Altamont Pass
in California. Here, a model such as WAsP predicts that the best winds should be at
the top of the pass, whereas a mesoscale-microscale modeling approach predicts the
acceleration of the relatively cool and dense marine air mass as it flows down the
slope. The result is a definite improvement in accuracy (Fig. 13-6).