A Pragmatic Approach to Wastewater Treatment Modelling: The Källby Wastewater Treatment Plant as a Case Study

Abstract: A wastewater treatment plant prediction in your computer The Källby wastewater treatment plant was set-up in a mathematical model reflecting its actual behaviour. Are the commonly available measurements good enough to perform a simulation? What kind of outcomes can we expect? After you have read the heading and the introduction, I guess that two proper questions swirling around in your head might be: what is wastewater treatment and what is a model? First wastewater treatment is the process that allows the return of sewerage and industrial wastewater to water cycle with minimal environmental impact and in accordance with existing discharge limits. There are three main reasons why this is performed: protect people from waterborne diseases that may be originated by contact with untreated wastewater, safeguard water environment from pollution that may arise from discharge of untreated wastewater and support of relevant life and economic systems. Secondly I would answer that a model is the representation of a system of interest, that is something important to be studied. It often simplifies reality in order to describe only the relevant processes and makes use of software solutions to perform the required computations. Modelling applied to wastewater treatment allows to simulate the actual physical, biological and chemical processes taking place in a wastewater treatment plant. Now, why do we need to model a wastewater treatment plant? Nowadays modelling is used as a design tool, instead of traditional design procedures. It is the most feasible and maybe less costly way to attain a process optimisation and it can also help the plant operators to test some corrective actions without expensive and environmentally risky full scale tests...wow! Let's get to the point: The present work achieved to simulate the Källby wastewater treatment plant of Lund, Sweden. The work focused on the wastewater treatment lines, especially on carbon, nitrogen and phosphorus removal. The sludge treatment lines, necessary to dispose of sewage sludge produced during sewage treatment, were not included in the model. Only available real data from the database of the company managing the plant were used to feed the model, without performing any additional measurement campaign. A relevant part of the work regarded the data treatment concerning the influent concentration to generate the input file for the model. After that it was possible to perform the simulation. At this point comparison between real measurements and model outputs was allowed. The reasons causing the main mismatches between the model outputs and real data were investigated and a troubleshooting step was performed in order to try to fix them. Annual variations of nitrogen and phosphorus were shown by the model, meaning that, even with the limitation of poor data, modelling can still be a valuable tool to understand the behaviour of treatment plants and to predict the response of the plant to influent variations. The result of this thesis is a deeper understanding of the actual behaviour of the plant. It also allows evaluations of new scenarios and possible future changes of the operation. Knowledge achieved in this work could also be used by plant operator to troubleshoot or to find better control strategies. Enjoy the reading!