Comparison of Two Day-Ahead Offering Strategies for a Flexible CHP Plant in Germany

Abstract

Two different day-ahead bidding strategies for a flexible combined heat and power plant are discussed. The plant portfolio consists of multiple combined heat and power units, a heat storage, gas boilers and an electric boiler. The first strategy relies on two-stage stochastic optimization which uses scenarios for electricity prices and thermal loads as inputs. The second “flat-bidding” strategy uses deterministic forecasts for electricity prices and thermal load. The performance of the two strategies is evaluated for four months in 2016 by consecutively simulating the day-ahead auction participation and subsequent plant operation. Both strategies are benchmarked against a perfect information optimization. The stochastic optimization decreases the costs in the range of +4 to −1 percent compared to the flat bidding scheme yet requiring excessively more computation time.