DataDrivenLM – Improving the system efficiency of district heating networks by intelligent and data driven load management

District heating networks are an essential part of the existing heat supply and a key technology of future decarbonised and cross-sector energy system solutions.

In addition to major challenges in the last 20 years (reduced heating demand, increased investment- and operating costs, …), many new challenges are upcoming as a result of continuous technical developments, the reduction of system temperatures and the integration of multiple, partly fluctuating renewable heat sources as well as central/decentralised storages.

Experience in heating network operation has shown that there is great potential in optimising plant operation and control and that intelligent, predictive control systems are becoming a basic requoirement for future-proof systems as a result of increasing system complexity. Modern data acquisition systems, which are also increasingly being used in the DH sector, provide detailed operating data. Based on systematic analysis methods these data can be used for an efficient, partially automated and time-saving system optimization. In addition, data-based approaches lead to the development of models able to predict consumption profiles and grid loads. Such models are the key to predictive control concepts.

Essential research work in the project comprises the potential analysis of measurement data (availability, usability, quality), data preprocessing (homogenisation and plausibility check), the coupling of different data sources (meteorological data, building data,…), the intelligent evaluation and processing of big data to generate usable information and the data-driven development of model-based controls focused on load management. The aim is to improve the control concept to a functional model able to carry out initial tests with real plant data.

Project manager

DI Philip Ohnewein, DI Harald Schrammel