Mixed-Integer Dynamic Scheduling Optimization for Demand Side Management

Abstract With fluctuating electricity prices, demand side management (DSM) promises to reduce energy costs. DSM of processes and energy supply systems requires scheduling optimization to consider transient behavior and binary on/off-decisions resulting in challenging mixed-integer dynamic programs. In this work, we present an efficient scheduling optimization approach that captures scheduling-relevant dynamics in a linear scale-bridging model and relies on collocation for time discretization. The resulting mixed-integer linear program (MILP) can be solved with state-of-the-art solvers. We apply the approach to a case study on building DSM. A detailed simulation model represents an office building, which allows load shifting through dynamic concrete core activation and is heated by a heat pump with minimum part-load. The DSM scheduling optimization approach reduces energy cost significantly compared to a rule-based scheduler without DSM if electricity price volatility is high. At the same time, the optimization is sufficiently fast to perform online scheduling.