Geothermal heat supply

Development of Concepts for the Optimal Control of District Heating Systems with Distributed Heat Supplies

Decentralized and regenerative energy sources will increasingly be used in district heating systems in the coming years, replacing today's mostly centralized heat supply relying on fossil fuels. The optimal distribution of heat with sufficient  pressure and temperature for all consumers will become more difficult. The goal of this project is to develop a concept using Model Predictive Control (MPC) to feed heat into the network with optimal distribution and supply temperatures. The project is carried out together with the energy supplier of Munich (Stadtwerke München - SWM).

Fast and accurate models are necessary for the use in a Model Predictive Controller. Therefore, machine learning and system identification models are researched in order to replace complex thermo-hydraulic models. For the optimization part of the MPC local and global optimization algorithms are implemented and tested, such as the optimization solver WORHP which is in part developed at the Center for Industrial Mathematics (ZeTeM) at the University of Bremen.