Two-Stage Unit Commitment Modeling for Virtual Power Plants

Abstract

The development of an increasingly decentralized, renewable power supply requires adequate planning approaches. Compared to unit commitment planning in regulated markets with a dominant share of dispatchable power generation, power systems with large shares of intermittent renewable power sources such as wind or photovoltaics are subject to uncertain supply as well as uncertain load forecasts and prices.

Virtual Power Plants have been developed to aggregate intermittent renewables with so-called flexibility options, which include dispatchable power plants, storage systems and flexible power consumers. Dispatchable power plants, such as biogas plants, include all that can actively be committed to supply power in a time interval. Storage systems, such as pumped-storage hydroelectricity, can store power in times of low prices and resell it when prices rise. Flexible power consumers, such as operators of electric vehicles, can attempt to use these time windows to load the batteries, lowering their power purchasing costs.

In the current German power market, power can be traded either in auctions on the day before physical delivery or in continuous intraday trading on the day itself. To determine optimal schedules for flexibility options in the context of day-ahead or intraday markets, a two-stage unit commitment model is presented to deal with the uncertainty of market prices resulting from the interplay of power generation in wind turbines and photovoltaic cells one the one hand with power demand on the other.