Estimating wind forecast errors and quantifying its impact on system operations subject to optimal dispatch

Wind power is being added to the supply mix of numerous jurisdictions, and an increasing level of uncertainties will be the new reality for many system operators. Accurately estimating these uncertainties and properly analyzing their effects will be very important to the reliable operation of the grid. A method is proposed to use historical wind speed, power, and forecast data to estimate the potential future forecast errors. The method uses the weather conditions and ramp events to improve the accuracy of the estimation. A bilevel programming technique is proposed to quantify the effects of the estimated uncertainties. It improves upon existing methods by modeling the transmission network and the re-dispatch of the generators by operators. The technique is tested with multiple systems to illustrate the feasibility of using this technique to alert system operators to potential problems during operation.