Heat pump at AffaldVarme Aarhus. The photograph shows the first part of the heat pump, the steam-driven chiller, during installation.  Photograph: Finn Olesen, AffaldVarme Aarhus.

Digital twins to improve heat pump operation

Thursday 19 Nov 20


Wiebke Meesenburg
DTU Mechanical Engineering


Nanna Berthelsen, AffaldVarme Aarhus
mail: blna@affaldvarme.dk

Benjamin Zühlsdorf,Teknologisk Institut
mail: bez@teknologisk.dk

Sven Försterling, TLK-Thermo 
mail: s.foersterling@tlk-thermo.com
Advanced digital models are needed to optimize heat pump operation in district heating systems. It will happen by means of ‘self-learning’ digital twins, which will potentially play a major role for the green transition in the sector.

Large industrial heat pumps are increasingly gaining ground in Danish district heating plants, leading to a growing interest in the development of digital twins, which are in effect digital copies of the heat pumps. 

A digital twin can be used for several different purposes. First, it can monitor the operation of the heat pump and check that everything is working as it should. Using real-time measurements from the heat pump’s various components, the digital twin can also optimize these settings and feed them back to the physical heat pump, which adapts them and optimizes operation to the greatest possible extent. Real-time data is also used to continuously improve the model.

The digital twin can also be used to locate defects in the heat pump, and to predict when individual components need to be replaced or cleaned. In this way, maintenance can take place before the occurrence of a real breakdown that impacts operation and heat supply.

Universal model for all heat pumps
Digital models of heat pumps already exist, but they are very resource-intensive to develop and adapt to each individual unit. A new project, led by the Danish Technological Institute, will therefore go one step further and develop a more universal digital twin. It needs to be built up in a modular fashion, so that it can be adapted relatively easily to a new heat pump without having to start from scratch every time.

“We’re hoping that by using this data-driven modelling, statistical methods, and deterministic models we can make the digital twin as automatic as possible. If we can make it ‘self-learning’ or able to adapt itself, it will be able to establish itself to a large extent on a new heat pump and thereafter monitor operation and predict maintenance tasks,” says Wiebke Meesenburg, DTU Mechanical Engineering.

Together with the other project partners, Wiebke Meesenburg and DTU researchers will contribute the methodological work aimed at creating a more universal digital twin which can simply be adapted and applied to all large heat pumps.

More heat pumps welcome
The development is happening in close cooperation with a couple of district heating companies which use heat pumps for heating. One of them is the district heating and waste-processing company AffaldVarme Aarhus.

“We’re pleased to be contributing to the project, and thus help develop new technologies for the green transition. At the same time, with a digital twin, we can test new features and settings on our heat pumps without needing to take a screwdriver to the physical heat pump,” says Nanna Berthelsen, energy planner at AffaldVarme Aarhus.

The development of digital twins for heat pumps is supported by the Energy Technology Development and Demonstration Programme, and is being carried out in cooperation with DTU, the Danish Technological Institute, TLK-Thermo GmbH, and AffaldVarme Aarhus. The project would like to include data from more heat pumps, so anyone who is interested is very welcome to contact Wiebke Meesenburg.