A Mixed Integer Quadratic Program for Valuing the Impact of Price and Forecast Uncertainty for Wind Generators

Abstract

Owners of wind power plants are exposed to financial risk in wholesale electricity markets due to the uncertain nature of wind forecasts and price volatility. In the event of a wind shortfall, the plant may have to repurchase power at a higher price in the real-time market. However, reducing the power offered in the day-ahead market may also be interpreted by regulators as physical withholding. We formulate and solve a mixed-integer quadratic program (MIQP) that prices the uncertain portion of a wind generator's forecast to hedge against uncertainties and which addresses concerns around withholding. We exploit the structure of the MIQP inputs to introduce additional constraints to improve computation time. Additionally, we provide a qualitative approach for generators and regulators to interpret the results of the MIQP. Finally, we simulate a real-world application for a wind farm in New York using past wind forecasts and NYISO prices.