Control of bidirectional prosumer substations in smart thermal grids : A weighted proportional-integral control approach

Title data

Licklederer, Thomas ; Zinsmeister, Daniel ; Lukas, Lorenz ; Speer, Fabian ; Hamacher, Thomas ; Perić, Vedran S.:
Control of bidirectional prosumer substations in smart thermal grids : A weighted proportional-integral control approach.
In: Applied Energy. Vol. 354, Part A (2024) . - 122239.
ISSN 1872-9118
DOI: https://doi.org/10.1016/j.apenergy.2023.122239

Project information

Abstract in another language

In future smart thermal grids, decentralized prosumers shall supply each other over the network, optimized by an overarching management system. For this, the substation control must be able to directly process optimal power flow setpoints from the management, while considering the mutual influence of prosumers on the technical level. Addressing these requirements, this paper develops a control approach for bidirectional prosumer substations. Therefore, we identify the objectives and characteristics of the control problem, before relaxing it by reformulation. This enables to keep proportional–integral–derivative controllers (PIDs) for operating the pumps and valves as decentral actuators in the substations. We propose a control approach, exploring the combination of two key aspects: (1) The assignment of temperature objectives to actuators. (2) Weighted error functions that linearly combine temperature and power errors as input for the PIDs, allowing to handle multiple objectives with limited actuators. Through case studies, we validate the suitability of our proposed objective–actuator assignment and demonstrate the necessity and benefits of the weighted error functions. Like this, we conclude with a control approach for bidirectional prosumer substations that aims for the relevant temperature objectives, directly considers heat transfer setpoints, replicates and exploits the coupling within the network, effectively manages conflicting objectives, copes with prosumer interference, and is pragmatic for real-time operation. Our approach can be transferred to different network concepts, bridging the gap between decentralized field-level operation and the management level.