Optimal Trading of Flexible Power Consumption on the Day-Ahead Market

As power generators shift from inert power plants to more volatile renewable sources of electricity, variable loads allow energy intensive businesses to monetize their flexibility in the electricity market. In the aluminum industry, engineering innovations have enabled such flexibility. A virtual battery can balance changing power inflows over a period of a few days. We have modeled the optimization problem of when and how to use this flexibility on the day-ahead energy market based on hourly price forecasts as a linear program. Besides the expected revenue, we also consider technical costs which occur in an unsteady energy schedule. In order to account for realistic trading settings, we embed the optimization problem in a daily rolling horizon planning framework. In this work, we generalize the specific planning problem to a broader range of applications and analyze the main influences on a profitable trading strategy. For a typical parametrization, we present a constructive heuristic that does not require an LP-Solver.