Master's Thesis Steffen Kaminski


Development and implementation of a cloud-based energy management system for optimized scheduling of building energy systems in city districts

Cloud-Energiemanagementsystems Copyright: EBC Illustration of Cloud-based energy management system approach. The clients communicate by a communication layer with the coordinator.

Modern residential heating systems can offer flexibilities to compensate imbalances of the electrical grid, which are introduced to the grid by the time-varying electricity generation of renewable energy systems and their uneven distribution across Germany. In order to utilize these flexibilities a cloud-based energy management system approach is developed and presented in this work.

The developed energy management system uses a hierarchical structure. Decisions are made by a central entity, the coordinator, and forwarded to the participants. A two-stage optimization approach is employed for an optimal control. On the client level, the participants objectives are considered. On the coordinator level, however, flexibilities are also taken into account. Therefor, the schedules determined on the client level are reevaluated in terms of grid stability. For the coordination of flexibilities, two different control strategies are introduced. The first strategy uses a set of client schedules to determine the most suitable combination of schedules. The second strategy uses a central optimization problem considering a modulation of each participants heating system. The determined operating schedules are applied to the residential heating systems by a controller. Within this work the developed cloud-based energy management approach was also implemented into a demonstrator platform.

In order to evaluate this platform, a use case simulation considering a city district of six buildings is evaluated using deterministic input data. It is shown that the developed energy management system is capable of saving operational costs or of reducing CO2 emissions while providing flexibilities for the electrical grid. Thereby, the second strategy shows a better performance addressing flexibilities.

Moreover, a profitability analysis is conducted validating that the presented approach is economically efficient.