Model extension, calibration and validation of partial nitritation–anammox process in moving bed biofilm reactor (MBBR) for reject and mainstream wastewater

Abstract

In the paper, the extension of mathematical model of partial nitritation-anammox process in a moving bed biofilm reactor (MBBR) is presented. The model was calibrated with a set of kinetic, stoichiometric and biofilm parameters, whose values were taken from the literature and batch tests. The model was validated with data obtained from: laboratory batch experiments, pilot-scale MBBR for a reject water deammonification operated at Himmerfjärden wastewater treatment and pilot-scale MBBR for mainstream wastewater deammonification at Hammarby Sjöstadsverk research facility, Sweden. Simulations were conducted in AQUASIM software. The proposed, extended model proved to be useful for simulating of partial nitritation/anammox process in biofilm reactor both for reject water and mainstream wastewater at variable substrate concentrations (influent total ammonium-nitrogen concentration of 530 ± 68; 45 ± 2.6 and 38 ± 3 gN/m - for reject water - and two cases of mainstream wastewater treatment, respectively), temperature (24 ± 2.8; 15 ± 1.1 and 18 ± 0.5°C), pH (7.8 ± 0.2; 7.3 ± 0.1 and 7.4 ± 0.1) and aeration patterns (continuous aeration and intermittent aeration with variable dissolved oxygen concentrations and length of aerated and anoxic phases). The model can be utilized for optimizing and testing different operational strategies of deammonification process in biofilm systems.

Combining numerical simulation with response surface modelling for optimization of reject water partial nitritation/anammox in moving bed biofilm reactor

“…Nitrite-and nitratenitrogen influent concentrations were assumed to be zero. Initial biomass concentrations and biofilm thickness were consistent with the settings for the reject water MBBR made by Trojanowicz et al [17].…”

“…Numerical studies of PN/A in the MBBR bioreactor were conducted with AQUASIM 2.1f computer program with the application of the Wanner and Reichert [16] biofilm model and the validated, mathematical model of PN/A process in biofilm reactors [17]. Detailed descriptions of the applied model structure with its parameter values and validation results have been presented by Trojanowicz et al [17].…”

“…The values were as follows: simulated value of TINrem and TINrem ± RMSE. The RMSE value was calculated by Trojanowicz et al [17] in the course of validation of the PN/A model for reject water treatment and its mean value was used.…”

“…Therefore, the first stage of the RSM procedurefactors screening was omitted in the current analysis. Factors selection was based on authors' experience from long-term, laboratory and pilot-scale studies regarding one-stage PN/A process in MBBR reactors, earlier modelling studies and literature data [11,15,17,19]. The proper selection of the value range of the chosen factors was based on the data regarding reject water quality, quantity and the variability of those values (Table 1).…”

“…The predicted values were consistent with those recorded in the efficient biofilm systems for reject water PN/A (Table 4) ranging from 7.2 to 7.8. It should be emphasized that the mathematical model of PN/A in MBBR [17] applied in the course of the presented studies considers only the indirect impact of pH on deammonification process performance. It is related to the changes in free ammonia (FA) and free nitrous acid (FNA) concentrations, which are in the model the real substrates for autothrophic growth and main process inhibitors (depending on their concentrations in the system).…”