Biokinetic and Artificial Intelligence Models for the Simulation of Nitrous Oxide Emissions from Wastewater Treatment Plants
[EN] Nitrous oxide (N2O) is considered a potent and very harmful greenhouse gas (GHG), and wastewater treatment plants (WWTPs) are considered a potent source of it. Predicting N2O emissions is a first step in reducing these. One way of doing this is by using a process-based biokinetic model, based...
| Autores: | , , , |
|---|---|
| Tipo de recurso: | capítulo de libro |
| Fecha de publicación: | 2024 |
| País: | España |
| Institución: | Universitat Politècnica de València (UPV) |
| Repositorio: | RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia |
| Idioma: | inglés |
| OAI Identifier: | oai:riunet.upv.es:10251/205900 |
| Acceso en línea: | https://riunet.upv.es/handle/10251/205900 |
| Access Level: | acceso abierto |
| Palabra clave: | Nitrous oxide Wastewater treatment Biokinetic modelling Artificial intelligence Artificial neural network |
| Sumario: | [EN] Nitrous oxide (N2O) is considered a potent and very harmful greenhouse gas (GHG), and wastewater treatment plants (WWTPs) are considered a potent source of it. Predicting N2O emissions is a first step in reducing these. One way of doing this is by using a process-based biokinetic model, based on Activated Sludge Models (ASMs) that have been extended to include the N2O production pathways. Alternatively, data-driven Artificial Intelligence (AI) models can be used to predict N2O emissions. In this paper, a biokinetic model has been built and calibrated for the Amsterdam West WWTP (1.1 Million PE; 168 MLD), using the EnviroSim software, BioWin®. A comprehensive monitoring campaign was conducted to characterise the common quality parameters (COD, TKN, TP, TSS, etc.) into their fractions, which were then used as BioWin model inputs. The calibration was conducted in two stages to predict effluent quality followed by model calibration to predict N2O emissions. Additionally, an Artificial Neural Network (ANN) based model was developed using pertinent process parameters, such as the influent flowrate, and NH4 in the aerobic tank as inputs to predict the N2O concentration in the gas phase. Preliminary results demonstrate that the ANN model outperforms the BioWin model in terms of prediction accuracy. Still further work is required to better understand the pros and cons of the two modelling approaches. |
|---|