Improved prediction of nitrogen removal using updated asm1 default parameters verified at 13 full-scale plants
Amélioration des prédictions de traitement de l'azote par le modèle numérique ASM1 en utilisant un jeu de paramètres validés sur 13 stations d'épuration des eaux résiduaires urbaines
Résumé
The Activated Sludge Model n°1 (ASM1) published in 1987 is still one of the main models used in simulation projects for design and optimization of processes. It was published with a set of default parameter values at 20°C and at 10°C that reflected the best knowledge twenty years ago. Several recent lab-scale studies using experiments conducted under controlled conditions have pointed out that some of these default values may be inadequate. The endogenous decay rate of nitrifiers appears to be higher than originally assumed, while the heterotrophic yield was shown to have a lower value under anoxic conditions. As these results were issued by several different lab experiments, it is difficult to define the domain of validity of the measured parameter values, and to evaluate if a simple aggregation of the new information leads to a correct set of default parameters. Moreover most of the tests were conducted at 20°C, so the parameter values have not been confirmed at 10°C. To address these concerns, simulations were carried out with the updated set of parameters to evaluate its ability to provide improved predictions compared to observed data collected at 13 nitrifying-denitrifying municipal treatment plants. The measured maximum nitrification rates and the nitrogen concentration in the effluent were selected as the output variables to assess the model predictions. The work proves that simulations using the original ASM1 default parameters tend to overpredict the nitrification rates and underpredict the denitrification st½chiometry. The updated set of parameters allows more realistic predictions, over a wide range of operating conditions.