Research project

SuSeWK: system-level management for efficient heating and cooling

Chillers and heat pumps are usually controlled one component at a time. Each part does its job on its own. How the parts work together is left to chance, and that is where a large share of the savings sits.

In the SuSeWK project we are developing operational management that treats the plant as one system and runs directly on its programmable logic controller. It is built on physics-based models of typical plant components. We modularise those models far enough that they compute on standard control hardware, with no cloud connection required. Machine learning adds operational forecasts and predictive maintenance recommendations. We chose methods that work with the small data volumes real plants actually produce.

Project objectives

  • Represent chillers and heat pumps modularly on a PLC, accurate to within ±10 % against systems characterised by measurement
  • Build a hardware-in-the-loop test bench so that operating strategies can be validated before they reach the field
  • Trial the system manager in field tests with operators in data centres, hospitals and the energy sector
  • Reduce the instrumentation required, so that the approach also pays off on smaller plants
  • Automate the evaluation and reporting of plant operation

Expected results

These are the results the project is working towards. They are targets, not measured outcomes.

  • Operating strategies that reach around 40 % electrical savings in simulation against the state of the art
  • More than 30 % savings of electrical energy in real installations, measured continuously over twelve weeks
  • Evidence that machine learning contributes at least a further 10 % in efficiency, and that predictive maintenance recommendations are feasible
  • System-level management that transfers across typical plant configurations and can be deployed in existing installations without project-specific redevelopment

Project duration

1 January 2025 to 30 September 2026

Funding

Kofinanziert von der Europäischen Union

This project is supported under the Pro FIT programme of Investitionsbank Berlin, with funding from the European Regional Development Fund (ERDF) and the State of Berlin. It is assigned to policy objective 1 and specific objective 1.1 of the ERDF programme of the State of Berlin 2021 to 2027.