Copenhagen Solutions Lab have just finished a partnership project about self-learning and data-driven solutions for district heating, that shows great potential for economic, energy-related, and CO2 related savings.
The project was initiated by DTU (Denmark’s Technical University) with the idea to test how methods from artificial intelligence can be utilized to optimize the flow temperature by predicting heat demand. Copenhagen Solutions Lab invited HOFOR to participate in the project as the supplier of district heating in Copenhagen. The Tingbjerg area in Copenhagen was appointed as test area, where 30 properties equipped with new digital meters have provided the project with data.
There is a lot of money to save by having a more precise control of the flow temperature instead of being on the safe side and operating with an unnecessary high temperature. DAMVAD Analytics have calculated the potential for a yearly cost saving to be 240-790 million kr. in Denmark if the flow temperature is lowered with 3-10 degrees Celsius.
The pilot project showed that the flow temperature in Tingbjerg could be lowered with around 5 degrees Celsius solely due to the higher precision that the data-driven methods give rise to. A further gain is the fewer temperature fluctuations and cost reductions in maintenance.
In addition to the economic savings, the pilot project has also shown a great potential to improve the energy consumption and carbon emissions from district heating through a higher degree of digitalization. The project therefore helps to support Copenhagen’s goal to become a carbon-neutral city where digitalization of the district heating can be an important part of reaching this goal.
The experiences from the project have been shared with Danish District Heating Association so the positive results and experiences from the pilot project hopefully can be spread out. The necessary data can be provided by the regular, modern meters that already are placed at a majority of the consumers. However, an improvement of the present data structure is required, as it is not geared to handle the big amounts of data. At the same time GDPR-rules and data safety pose a challenge as well.
About the project:
- The project is financed by The Capital Region of Denmark.
- The project is carried out in collaboration with Gate 21, HOFOR, DTU Compute and the DTU-spin-out company ENFOR.
- The title of the project, IDASC, stands for ‘Intelligent Data Use in Smart Cities’.
- The data-driven models used in the project are based on artificial intelligence that makes systems and models self-learning.
- Read the two technical reports here ("Energy data: mapping, barriers and value creation") and here ("Digitalization of District Heating").