Application of activated sludge models (ASMs) to full-scale wastewater treatment plants (WWTPs) is still hampered by the problem of model calibration of these over-parameterised models. This either requires expert knowledge or global methods that explore a large parameter space. However, a better balance in structure between the submodels (ASM, hydraulic, aeration, etc.) and improved quality of influent data result in much smaller calibration efforts. In this contribution, a methodology is proposed that links data frequency and model structure to calibration quality and output uncertainty. It is composed of defining the model structure, the input data, an automated calibration, confidence interval computation and uncertainty propagation to the model output. Apart from the last step, the methodology is applied to an existing WWTP using three models differing only in the aeration submodel. A sensitivity analysis was performed on all models, allowing the ranking of the most important parameters to select in the subsequent calibration step. The aeration submodel proved very important to get good NH(4) predictions. Finally, the impact of data frequency was explored. Lowering the frequency resulted in larger deviations of parameter estimates from their default values and larger confidence intervals. Autocorrelation due to high frequency calibration data has an opposite effect on the confidence intervals. The proposed methodology opens doors to facilitate and improve calibration efforts and to design measurement campaigns.

The paper investigates the impact of the way oxygen transfer is modelled and the frequency of influent data on the dynamic calibration of a full-scale WWTP. Oxygen transfer was modelled in 2 ways: by means of a fast "virtual controller" tracking dissolved oxygen and by means of a linear correlation between K L a and air flow rate. Influent data was retrieved from correlations derived from either off-line data or on-line data. The correlations in the latter case were found to be better. With regard to model performance, it was found that the oxygen transfer model based on the linear relation between K L a and air flow rate was able to sufficiently capture DO dynamics. Using the off-line data based correlation for influent data resulted in a decent DO profile prediction, but the nitrification prediction was not accurate implying that a calibration effort is required. However, when using on-line influent data correlations, NH 4 predictions were found accurate without any need for calibration. Using more influent data resulted in a small deterioration of NH 4 predictions but resulted in better NO 3 predictions.

The current model of the full-scale wastewater treatment plant model in Eindhoven uses a state-of-the-art model for the biological processes (ASM2d) and is calibrated for C-and N-removal in dry weather. However, for the "Kallisto" project, which is an innovation programme aiming at a smart improvement of the surface water quality of the river Dommel by applying cost effective integrated system measures, the WWTP model needs to be improved to predict the WWTP performance under all conditions foreseen in the scenarios (e.g. storm events). A project approach was developed with parallel improvements in the different submodels, based on the interaction between submodels and the availability of several on-line sensors in influent, in-process and effluent. This is in contrast to most WWTP modelling studies, where focus is only on one submodel. It should lead to a well-balanced dynamic model that is able to predict WWTP behaviour under various conditions and that will be included in the integrated model, which will serve as an important decision support tool. Keywords urban wastewater system, dynamic modelling, WWTP optimization INTRODUCTIONOverall improvement of efficiency, increasingly stringent effluent discharge limits, the aim for a better surface water quality, minimization of energy use and greenhouse gas emissions are current drivers for the optimization of urban wastewater systems and operational strategies of wastewater treatment plants (WWTPs). In this respect, the use of dynamic models has already proven to be of great value. Indeed, a model with high predictive power allows testing of different optimisation strategies without disrupting the actual operation of the plant i.e. without the risk of losing biomass or violating discharge permits. Waterboard De Dommel (WDD, The Netherlands) has shown long-term interest in the development and use of dynamic models, which resulted in a calibrated dynamic model (including COD, N and P removal) of WWTP Haaren (Sin et al.,2008). Building further on the gained knowledge, a full-scale model of Eindhoven was calibrated and validated for dry weather including C and N removal (Nopens et al., 2010). This model is used as basis in the wastewater treatment plant modelling part of the current "Kallisto" project, which studies the integrated urban water system in the water cluster Eindhoven.

No abstract

Aeration modeling is essential not only for insuring the correct nitrification and carbon oxidation, but also for reducing energy costs. Activated sludge models (ASMs) use the oxygen transfer coefficient (k L a) and the dissolved oxygen (DO) at saturation in the DO mass balance to predict oxygen transfer. The oxygen transfer efficiency (OTE) depends on: aeration system type, size and geometry; air flow rate; mixing rates and gradients; diffuser submergence; released bubble size and distribution; diffusers time in operation and consequent fouling and scaling. Moreover, due to the dynamic nature of plant influent load, and the therefore variable air flow rate, OTE is variable over the diurnal and seasonal cycles. Also, due to fouling and scaling, OTE declines with increasing time in operation. These effects are not included in current aeration models and although not crucial for modeling the biochemical processes in the activated sludge, they are necessary to quantify and minimize process energy footprint. Furthermore, it is also expected that this will reduce the need for re-calibration of ASM models. In this contribution, we propose an expanded aeration model by including the dependence of aeration efficiency to the aforementioned design and process parameters. The potential for energy footprint reduction during operation was calculated as the resulting energy savings from minimized excess DO.