Photo: Mikal Schlosser

Software saves millions on urban heating

Friday 11 Mar 16
by Morten Andersen


Henrik Madsen
Professor, Head of section
DTU Compute
+45 45 25 34 08

Smart energy research

The research in control software for district heating takes place as a part of the programme CITIES (Center for IT-Intelligent Energy Systems) which is headed by Professor Henrik Madsen, DTU Compute.

The urban district heating systems can help stabilize the energy system, if they are controlled intelligently.

About Enfor

The company Enfor is a DTU spin-out from 2007, and provides software and consultancy supporting intelligent energy supply. In addition to district heating control software, another significant area of expertise is solutions for detailed weather forecasts that can be exploited for wind power control. Control of solar power installations is a growing area with the increasing number of installations in Denmark.

DTU still has a shareholding in the company, which regularly cooperates with the University on research projects. 

For the sake of safety, district heating water was previously heated more than adequately. Using software developed at DTU, the temperature in the network is lowered 3-10°C without leaving consumers freezing.

Danish district heating plants will realise huge savings if they integrate software that accurately controls the temperature in the district heating network. The public utility Roskilde Forsyning saved just over DKK 2 million on the company’s annual operating costs. And the heat loss in the network decreased by 18 per cent.

“The necessity of controller software in the industry is currently attracting a lot of attention. Throughout Denmark, district heating companies can save millions of kroner, and carbon emissions to the environment can be reduced by several tonnes when utilizing the energy more efficiently,” says Torben Nielsen, CEO of Enfor—a spin-out from DTU.

Some heat loss is unavoidable 

With 60 per cent, Denmark is the country in the world where district heating covers the largest part of the heat consumption of private households. However, other countries are following suit. District heating is smart, because you utilize waste heat from power production. And when you have the district heating network anyway, you can also exploit other energy sources—such as heat from waste incineration.

A disadvantage, however, is that there will always be a certain waste when the heat is supplied to the consumers. You therefore need to send the water at a temperature that is so high that it still has an acceptable temperature when it reaches the final consumer on the network.

Previously, the temperature in the system was regulated manually in a control centre. The precise flow temperature varies between the different areas and companies. But normally, it is in the range 70-80°C for most of the year and 80-90°C in the coldest months.

“With the right control software, you can typically do with 70°C for most of the year and 80-85°C in the coldest months. This temperature reduction is in itself worth a lot of money if the heat is coming from a CHP plant. In addition, the heat loss in the pipes will be less when the water leaving the plant is not so hot,” explains Torben Nielsen.

Many uncertainties 

In view of the large savings it may seem strange that control software has not been introduced in the industry a long time ago. Professor Henrik Madsen from DTU Compute explains:

“It has been a challenge to develop control software dedicated to this purpose. The traditional solutions could not handle the many uncertainties in the district heating system. The uncertainties is especially due to the fact that the water propagates at different speeds in different parts of the system. You must handle a large variation of temperatures and, at the same time, a great deal of variation in the uncertainties regarding the precise temperature."

Part of the solution are temperature sensors in various locations in the system.

“If you place the sensors appropriately, 5 to 6 sensors will be sufficient in a typical Danish city. But it’s clear that with the current sensor technology developments where you can, e.g., take hourly measurements, the precision will increase. In this way you can also achieve greater savings,” he concludes.

Taking weather forecast into account

In the latest generations of the system, the measurements are linked to consumption forecasts. The weather, in particular, is of course very important for how much consumers turn the heat up up and down. Therefore, the control depends on weather reports.

“Especially during transition periods in the spring and the autumn, there is a lot of uncertainty about consumer choices. In this context, it’s relevant to convert the local weather forecast to a forecast of the expected consumption 12 hours ahead of time,” says Torben Nielsen, CEO of Enfor.