Exploring the Impacts of Wind Forecasting Error on the PJM Electrical Grid for Offshore Wind in the Mid-Atlantic

Efficient, economical operation of an offshore wind farm relies on frequent, accurate forecasts of the available wind resource. One of the challenges in wind forecasting is correctly predicting the timing and strength of sudden changes in the wind such as occurs in wind ramp events. We characterize the impact that forecasting error of wind ramps in the Delaware wind energy area (WEA) have on the regional electrical system, using the Regional Weather and Forecasting (WRF) model, in forecast mode. We explore the sensitivity of forecasting error to different input datasets including GFS3, GFS4, and NAM. Wind power production is estimated using 5 years of data from NDBC buoy 44009, extrapolated to hub height. The wind aloft is then used to calculate the power produced by a hypothetical wind farm at the Delaware WEA and compared to the changes in power demand on the PJM grid. Particular attention is paid to the midsprediction of wind ramps at times when the power demand is increasing.