By Peter Witten, AAU Communication and Public Affairs

AI can reduce both costs and CO2 emissions. This is demonstrated by an innovation project involving AAU researcher Per Printz Madsen. The innovation project RACE has been successfully tested at Brønderslev district heating supply.

"We have seen results that exceed our original hopes," says Associate Professor Per Printz Madsen from the Department of Computer Science, Aalborg University.

High precision

He has built and trained the digital twin, which predicts consumption patterns in district heating, weather conditions, and even faults before they occur.

"The digital twin can now predict changes in temperature and flow with unprecedented precision. This would have been unthinkable just a few years ago," reports Per Printz Madsen.

Using an advanced mathematical-physical model, the technology monitors data from district heating pipes.

Development continues, and the model will steadily become more precise and efficient.

At Brønderslev district heating supply, sensors currently measure water pressure and temperatures in three wells.

"We can get much better control and overview of what happens at the very edges of the district heating network. This helps ensure a stable and efficient operation for consumers," says Henrik Horsholt Christensen, team coordinator at Brønderslev Forsyning.

Saves up to 20 percent

By making continuous adjustments, instead of just once or twice a year, the Brønderslev district heating supply saves energy on pump operation, with an expected saving of 10-20 percent.

"This contributes to the green transition and reduces our and, therefore, consumers’ costs," adds Henrik Horsholt Christensen.

In Brønderslev, they had already optimized energy consumption for heating water. In other district heating systems that are not energy-optimized, even greater savings can be achieved with AI.