At ARA Innichen-Sexten, a new path has been taken to reduce energy consumption. Artificial intelligence is to play a decisive role in this.
The aim of the project was to forecast the effluent values of the wastewater treatment plant based on models and thus to support the reliable compliance with the discharge conditions.
For this purpose, optimal operating settings for the activated sludge process were to be calculated and predicted on the basis of the model-based predictions with the help of an online optimiser.
This system has already been installed at ARA Meran and we visited it there. Together with the managing director of ARA Pustertal AG Dr. Konrad Engl, the process engineer of ARA Tobl Hannes Kirchler and the process engineer of Ara Innichen-Sexten Emanuel Kirchler, it was decided to implement artificial intelligence at ARA Innichen-Sexten.
Since this is a system that requires plant conditions that are as stable as possible, it made sense to realise this system in San Candido-Sesto, since there had been stable conditions there for a long time and there had been no major reconstructions of the plant.
The investment project I07_19 was elaborated by the managing director and started on 23.10.2020. The company AquaTune from Germany then implemented the project with the help of us operators, ARA Pustertal AG and the company ATS GmbH.
The optimisation system is to regulate the aeration and recirculation in the biology. Of course, the limit values must always be adhered to. It was hoped that this would lead to better effluent values and energy savings of 5 to 10%.
The test phase was started on 11.05.2021. After a few improvements to the software, the system ran very reliably in the end. Nevertheless, in the event of a failure of the optimisation system, safeguards were set up to automatically switch to the traditional control system in the event of a malfunction.
The project was finally completed on 15.10.2021. Our supervisor Dr. Andreas Nink from AquaTune thoroughly trained us in the system on site at the ARA Innichen-Sexten. On this occasion he was also able to get an overview of the plant, which was not possible before due to the pandemic.
The optimisation system now runs very reliably and is well protected against possible failures in order to avoid exceeding limit values. The system runs around the clock all year round and can be expanded as required.
The first results are very satisfactory, the data show in the comparison calculated by influent load:
Thus, the project goals have been clearly achieved and to our full satisfaction.