Shall we upgrade one-dimensional secondary settler models used in WWTP simulators? - An assessment of model structure uncertainty and its propagation
Benedek Gy Plósz,Jeriffa De Clercq,Ingmar Nopens,Lorenzo Benedetti,Peter A Vanrolleghem
DOI: https://doi.org/10.2166/wst.2011.412
Abstract:In WWTP models, the accurate assessment of solids inventory in bioreactors equipped with solid-liquid separators, mostly described using one-dimensional (1-D) secondary settling tank (SST) models, is the most fundamental requirement of any calibration procedure. Scientific knowledge on characterising particulate organics in wastewater and on bacteria growth is well-established, whereas 1-D SST models and their impact on biomass concentration predictions are still poorly understood. A rigorous assessment of two 1-DSST models is thus presented: one based on hyperbolic (the widely used Takács-model) and one based on parabolic (the more recently presented Plósz-model) partial differential equations. The former model, using numerical approximation to yield realistic behaviour, is currently the most widely used by wastewater treatment process modellers. The latter is a convection-dispersion model that is solved in a numerically sound way. First, the explicit dispersion in the convection-dispersion model and the numerical dispersion for both SST models are calculated. Second, simulation results of effluent suspended solids concentration (XTSS,Eff), sludge recirculation stream (XTSS,RAS) and sludge blanket height (SBH) are used to demonstrate the distinct behaviour of the models. A thorough scenario analysis is carried out using SST feed flow rate, solids concentration, and overflow rate as degrees of freedom, spanning a broad loading spectrum. A comparison between the measurements and the simulation results demonstrates a considerably improved 1-D model realism using the convection-dispersion model in terms of SBH, XTSS,RAS and XTSS,Eff. Third, to assess the propagation of uncertainty derived from settler model structure to the biokinetic model, the impact of the SST model as sub-model in a plant-wide model on the general model performance is evaluated. A long-term simulation of a bulking event is conducted that spans temperature evolution throughout a summer/winter sequence. The model prediction in terms of nitrogen removal, solids inventory in the bioreactors and solids retention time as a function of the solids settling behaviour is investigated. It is found that the settler behaviour, simulated by the hyperbolic model, can introduce significant errors into the approximation of the solids retention time and thus solids inventory of the system. We demonstrate that these impacts can potentially cause deterioration of the predictive power of the biokinetic model, evidenced by an evaluation of the system's nitrogen removal efficiency. The convection-dispersion model exhibits superior behaviour, and the use of this type of model thus is highly recommended, especially bearing in mind future challenges, e.g., the explicit representation of uncertainty in WWTP models.