EU's Urban Wastewater Directive requires the efficient removal of organic matter from domestic wastewater. The required treatment efficiencies (COD>75%, BOD >70%) are not expected to be achieved without biological treatment. Norway has many treatment plants without biological processes and huge investments require to comply with the EU requirements. This article presents results documenting the possibility of achieving EU requirements at mechanical-chemical treatment plants, challenging the common perception of their inability to achieve such high removal rates. Data-driven models and hybrid soft sensors are utilized in achieving these results and possible mechanisms of removal are discussed.

This paper deals with the increasing of the operation efficiency of a wastewater treatment plant by reducing the level of pollutant concentration in the effluent. In essence, the main goal is the reduction of the concentration of organic substrate, Model Predictive Control algorithm being used for this, considering as a model of the wastewater treatment process a neural network model based on a simplified mathematical model of the 4th order. For the implementation of the control algorithm, two neural networks were trained. In the first version of the neural network, it was considered that all 4 state variables are measurable and in the second version, only the organic substrate was considered measurable. Regarding the training of the two neural networks, good results were obtained in the case of both versions. Instead, the control algorithm gave better results when the second neural network was used (lower values of the substrate, lower variations of the aeration rate, which means a lower cost of operation).

To increase the biogas production, the anaerobic digestion process requires advanced control tools. However, in order to achieve an advanced control of the anaerobic digestion process, two constraints must be taken into account: the need for available plants for experimentation and the need to reduce costs. In this paper, it is proposed to develop a test platform for the anaerobic digestion process using the Hardware in the Loop Simulations principle. The emulation of the anaerobic digestion process is ensured by a computer that communicates with a data acquisition board, and the process control is ensured by a Programable Logic Controller. Such a platform is effective because it allows testing in extreme situations, which in the case of environmental processes is very difficult. The platform also includes a cloud computing component for advanced data processing.