Smarter wastewater treatment management can shave 10-20 per cent off treatment plants’ operating costs.
Wastewater treatment requires a lot of electricity 24-7 and accounts for 1-2 per cent of Denmark’s electricity consumption. A demonstration project shows that intelligent process management means that treatment plants can contribute to the green transition. The trial has been performed under the CITIES project, which comprises research into effective and intelligent energy solutions by means of artificial intelligence (AI).
Treatment plants remove nutrients, odours, and material from the wastewater before it is discharged into the aquatic environment. During the process, the so-called aeration—by which air is supplied to large tanks of wastewater—plays an essential role in the bacterial cultures being able to consume nitrogen, carbon, and phosphorus compounds. The aeration is typically started and stopped in suitable cycles, and the treatment time varies. The aeration constitutes 45-75 per cent of the electricity consumption, and great cost reductions can therefore be achieved by managing the process based on when the electricity price is low and on renewable energy.
This has precisely been shown in the demonstration project, in which Peter Stentoft—as part of his industrial PhD with Krüger A/S and Veolia Water Technologies and together with a CITIES team headed by Henrik Madsen, Professor at DTU Compute—has developed the price-based management tool Smart-Energy Operating-System (SE-OS). The system is based on an AI model which combines data gathered through multiple sensors in the treatment plant with statistical methods and process knowledge and automatically adapts itself to the actual conditions.
"The ‘CITIES’ demo project shows that the treatment plant can thus act as a green battery—a flexible device in the smart electricity system of tomorrow,"
Professor Henrik Madsen, DTU
AI makes it possible to predict what will happen in the plant in the future, but also takes into account the uncertainty of the predictions. By simulating different scenarios, the treatment plant system can choose precisely the solution that is most inexpensive, while also meeting the requirements for wastewater treatment.
“We've seen in simulations that—in relation to how you manage in the moment—you can save 10-20 per cent on operating costs by managing smarter,” says Peter Stentoft.
Intelligent management can also help the electricity market—for example in highly windy weather—where wind turbine owners today have to pay to get rid of their electricity or shut down completely if the wind turbines produce electricity in excess of the demand. Here—based on price signals—the automatic model can temporarily turn the processes up and down again when the electricity market can again take delivery of the electricity, and even further down during peak periods.
The ‘CITIES’ demo project shows that the treatment plant can thus act as a green battery—a flexible device in the smart electricity system of tomorrow,” says Professor Henrik Madsen.
The CITIES project is supported by Innovation Fund Denmark and will conclude after seven years of research by the end of 2020. Peter Stentoft’s industrial PhD has also been supported by Innovation Fund Denmark, and will likewise be completed this year.