Boundary-layer wind and turbulence-profile theories as described in most textbooks apply to flat prairies, not to the rugged terrain of British Columbia (BC). For the mountainous terrain of BC, different turbines at the same wind farm experience different winds and turbulence associated with their locations relative to small-scale (unresolved) terrain features. Convolution of the resulting wind-speed distribution with the wind-turbine power curve for an individual turbine yields a wind-farm power curve that differs from the theoretical power curve. For several wind farms in BC, the farm-average power curve does not achieve the cube of wind speed even between the cut-in and rated-power speeds.
To partially compensate for these wind variations, we run an ensemble of up to 51 NWP model runs each day, with fewer ensemble members covering the more distant wind farms. These runs are based on a variety of initial/boundary conditions (from gov’t centers in Canada, USA, France, Germany), a variety of model cores (WRF-ARW, WRF-NMM, MM5, MPAS), a variety of horizontal grid spacings, and a variety of physics parameterizations. Each forecast is individually bias corrected based on recent-past observations at any wind farm, and then the separate runs are combined to yield ensemble-average and calibrated-probabilistic forecasts.
UBC has been making operational limited-area NWP forecasts of hub-height winds for all the active wind farms in BC for the past decade. BC Hydro uses our ensemble forecasts to better manage the integration of wind power with their much-greater hydro-power generation. BC Hydro also uses our forecasts of Bonneville Power Administration (Columbia River region) wind-farm hub-height winds to optimize their energy-trading to the USA.
Based on our experience, we created a course ATSC 313 “Renewable Energy Meteorology”, which covers meteorology for hydro, wind, and solar power.
.