Modelling aerobic granular sludge reactors through apparent half-saturation coefficients

Biological and physical selectors for mobile biofilms, aerobic granules, and densified-biological flocs in continuously flowing wastewater treatment processes: A state-of-the-art review

There have been significant advances in the use of biological and physical selectors for the intensification of continuously flowing biological wastewater treatment (WWT) processes. Biological selection allows for the development of large biological aggregates (e.g., mobile biofilm, aerobic granules, and densified biological flocs). Physical selection controls the solids residence times of large biological aggregates and ordinary biological flocs, and is usually accomplished using screens or hydrocyclones. Large biological aggregates can facilitate different biological transformations in a single reactor and enhance liquid and solids separation. Continuous-flow WWT processes incorporating biological and physical selectors offer benefits that can include reduced footprint, lower costs, and improved WWT process performance. Thus, it is expected that both interest in and application of these processes will increase significantly in the future. This review provides a comprehensive summary of biological and physical selectors and their design and operation.

High resolution functional analysis and community structure of photogranules

Photogranules are spherical aggregates formed of complex phototrophic ecosystems with potential for “aeration-free” wastewater treatment. Photogranules from a sequencing batch reactor were investigated by fluorescence microscopy, 16S/18S rRNA gene amplicon sequencing, microsensors, and stable- and radioisotope incubations to determine the granules’ composition, nutrient distribution, and light, carbon, and nitrogen budgets. The photogranules were biologically and chemically stratified, with filamentous cyanobacteria arranged in discrete layers and forming a scaffold to which other organisms were attached. Oxygen, nitrate, and light gradients were also detectable. Photosynthetic activity and nitrification were both predominantly restricted to the outer 500 µm, but while photosynthesis was relatively insensitive to the oxygen and nutrient (ammonium, phosphate, acetate) concentrations tested, nitrification was highly sensitive. Oxygen was cycled internally, with oxygen produced through photosynthesis rapidly consumed by aerobic respiration and nitrification. Oxygen production and consumption were well balanced. Similarly, nitrogen was cycled through paired nitrification and denitrification, and carbon was exchanged through photosynthesis and respiration. Our findings highlight that photogranules are complete, complex ecosystems with multiple linked nutrient cycles and will aid engineering decisions in photogranular wastewater treatment.

Start-up and optimization of a one-stage partial nitrification-anammox (PN-A) process treating low ammonium concentration wastewater: experimental results and modeling investigation

Although the partial nitrification-anammox process (PN-A) has achieved great success in nitrogen removal for the high ammonium concentration wastewater, its application is still limited in low ammonium concentration wastewater treatment due to its instability and low nitrogen removal efficiency. In this study, a sequencing batch reactor (SBR) with continuous aeration was employed to enrich ammonia oxidation bacteria (AOB) and suppress nitrite oxidation bacteria (NOB) first; then, the SBR was operated intermittently aerated SBR (IASBR), to which the anammox granular sludge (AMX) was added to achieve complete autotrophic nitrogen removal under low influent ammonium concentration of 100 mg/L. A mathematical model was used to optimize the IASBR aeration strategy to achieve sub-optimal nitrogen removal. The experimental results showed that high nitrite accumulation efficiency (above 80%) in the SBR and a fast start-up within 100 days and a stable TN (total nitrogen) removal efficiency of 70% were achieved in the IASBR. Meanwhile, the simulation results indicated that keeping aeration duration at 4 h, k_La (oxygen transfer coefficient) at 50 day^-1, or aeration duration at 2.5 h, k_La at 80 day^-1 could obtain a higher total nitrogen removal efficiency (TNR) (TNR > 80%), and the TN removal could also be improved by increasing hydraulic retention time (HRT) under the optimal oxygen supply rate.

N2 -fixation can sustain wastewater treatment performance of photogranules under nitrogen-limiting conditions

Wastewater characteristics can vary significantly, and in some municipal wastewaters the N:P ratio is as low as 5 resulting in nitrogen-limiting conditions. In this study, the microbial community, function, and morphology of photogranules under nitrogen-replete (N+) and limiting (N-) conditions was assessed in sequencing batch reactors. Photogranules under N- condition were nitrogen deprived 2/3 of a batch cycle duration. Surprisingly, this nitrogen limitation had no adverse effect on biomass productivity. Moreover, phosphorus and chemical oxygen demand removal were similar to their removal under N+ conditions. Although performance was similar, the difference in granule morphology was obvious. While N+ photogranules were dense and structurally confined, N- photogranules showed loose structures with occasional voids. Microbial community analysis revealed high abundance of cyanobacteria capable of N₂ -fixation. These were higher at N- (38%) than N+ (29%) treatments, showing that photogranules could adjust and maintain treatment performance and high biomass productivity by means of N₂ -fixation.

Successful startup of the single-stage PN-A (partial nitrification-anammox) process by controlling the oxygen supply

The single PN-A (partial nitrification-anammox) reactor offers a cost-effective solution for nitrogen removal. However, optimal control of the PN-A reactor is challenging due to the interactive mechanisms among the oxygen supply, bulk liquid DO (dissolved oxygen) concentration, and the balance of various functional bacterial species. In this study, a mathematical model was used to derive the optimal control variable for the maximum nitrogen removal, and an experimental PN-A reactor was operated to verify the model simulation results. The model simulation results indicate that the oxygen supply to the ammonium load ratio is the key factor to control the single-stage PN-A reactor for optimal TN removal. For optimal TN removal, the oxygen supply to the ammonium load ratio should be 1.9 mg O₂/mg N. The DO concentration is not the key control parameter to get the maximum TN removal as the optimal TN removal could be achieved under a wide range of DO concentration. The model simulation results were verified in the experimental PN-A reactor under oxygen transfer rate ([Formula: see text]) at 52 day^-1, HRT at 24 h, and ammonium load ratio of 0.55 kg N/(m³∙day).

Multidisciplinary characterization of nitrogen-removal granular sludge: A review of advances and technologies

Nitrogen-removal granular sludge (NRGS) is a promising technology in wastewater treatment, with advantages of efficient nitrogen removal, less footprint, lower sludge production and energy consumption, and is a way for wastewater treatment plants to achieve carbon-neutrality. Aerobic granular sludge (AGS) and anammox granular sludge (AnGS) are two typical NRGS technologies that have attracted extensive attention. Mounting evidence has shown strong associations between NRGS properties and the status of NRGS systems; however, a holistic view is still missing. The aim of this article is to provide an overview of NRGS with an emphasis on characterization. Specifically, the integrated nitrogen transformation pathways inside NRGS and the performance of NRGS treating various wastewaters are discussed. NRGS properties are categorized as physical-, chemical-, biological- and systematical ones, presenting current advances and corresponding characterization technologies. Finally, the future prospects for furthering the mechanistic understanding and engineering application of NRGS are proposed. Overall, the technological advancements in characterization have greatly contributed to understanding NRGS properties, which are potential factors for optimizing the performance and evaluating the working status of NRGS. This review will provide guidance in characterizing NRGS properties and boost the introduction of novel characterization technologies.

A Comprehensive View of the ASM1 Dynamic Model: Study on a Practical Case

The ASM1 model was elaborated by the IWA Task Group for Mathematical Modelling, with the aim of explaining and predicting the output values of organic matter concentration in activated sludge processes, especially for domestic wastewaters. In recent years, ASM1 has been completed with new components and extended to other biological processes, including biological membrane reactors, activated carbon filters, and microalgae bioreactors. In this article, the essentials of this model are studied by outlining the original topics that were formulated in the model, and by using a practical example of a wastewater treatment plant (WWTP), which can clarify the application of the ASM1. A protocol of approximation between the dynamic model and the experimental data for the COD effluent concentration is presented, based on three steps of tuning and fine tuning, and the corrected values of the kinetic parameters YH and μH,max are calculated in accordance with the minimum error. In the simulation procedure, the baseline and dynamism are controlled, comparing them to the experimental data line, and the values obtained for the kinetic parameters are YH = 0.60 and μH,max = 0.40 d−1. The kinetic parameters reflect the activity of the mixed community of microorganisms in the WWTP.

Machine learning modeling and analysis of biohydrogen production from wastewater by dark fermentation process

Dark fermentation process for simultaneous wastewater treatment and H₂ production is gaining attention. This study aimed to use machine learning (ML) procedures to model and analyze H₂ production from wastewater during dark fermentation. Different ML procedures were assessed based on the mean squared error (MSE) and determination coefficient (R²) to select the most robust models for modeling the process. The research showed that gradient boosting machine (GBM), support vector machine (SVM), random forest (RF) and AdaBoost were the most appropriate models, which were optimized by grid search and deeply analyzed by permutation variable importance (PVI) to identify the relative importance of process variables. All four models demonstrated promising performances in predicting H₂ production with high R² values (0.893, 0.885, 0.902 and 0.889) and small MSE values (0.015, 0.015, 0.016 and 0.015). Moreover, RF-PVI demonstrated that acetate, butyrate, acetate/butyrate, ethanol, Fe and Ni were of high importance in decreasing order.

Unaerated feeding alters the fate of dissolved methane during aerobic wastewater treatment

In municipal wastewater treatment plants, some dissolved methane can enter the aerobic bioreactors. This greenhouse gas originates from sewers and return flows from anaerobic sludge treatment. In well-mixed conventional activated sludge reactors, methane emissions are largely avoided because methane oxidizing bacteria consume a large fraction, even without optimizing for this purpose. In this work, the fate of dissolved methane is studied in aerobic granular sludge reactors, as they become increasingly popular. The influence of the characteristic design and operating conditions of these reactors are studied with a mathematical model with apparent conversion kinetics and stripping: the separation of feeding and aeration in time, a higher substrate transport resistance, a high retention time of granular biomass and a taller water column. Even for a best-case scenario combining an unrealistically low intragranule substrate transport resistance, a high retention time, a tall reactor, an extremely high influent methane concentration and no oxygen limitation, the methane conversion efficiency was only 12% when feeding and aeration were separated in time, which is lower than for continuous activated sludge reactors under typical conditions. A more rigorous model was used to confirm the limited conversion, considering the multi-species and multi-substrate biofilm kinetics, anoxic methane consumers and the high substrate concentration at the bottom during upward plug flow feeding. The observed limited methane conversion is mainly due to the high concentration that accumulates during unaerated feeding phases, which favours stripping more than conversion in the subsequent aeration phase. Based on these findings, strategies were proposed to mitigate methane emissions from wastewater treatment plants with sequentially operated reactors.

Potential of off-gas analyses for sequentially operated reactors demonstrated on full-scale aerobic granular sludge technology

This work shows how more variables can be monitored with a single off-gas sampler on sequentially operated than on continuously fed and aerated reactors and applies the methods to data from a full-scale aerobic granular sludge reactor as a demonstration and to obtain insight in this technology. First, liquid-gas transfer rates were calculated. Oxygen (O₂) absorption and carbon dioxide (CO₂) emission rates showed comparable cyclic trends due to the coupling of O₂ consumption and CO₂ production. Methane (CH₄) emissions showed a stripping profile and nitrous oxide (N₂O) emissions showed two peaks each cycle, which were attributed to different production pathways. Secondly, aeration characteristics were calculated, of which the gradual improvement within cycles was explained by surfactants degradation. Thirdly, liquid phase concentrations were estimated from off-gas measurements via a novel calculation procedure. As such, an average influent CH₄ concentration of 0.7 g·m^-3 was found. Fourthly, reaction rates could be estimated from off-gas data because no feeding or discharge occurred during reaction phases. The O₂ consumption rate increased with increasing dissolved oxygen and decreased once nitrification was complete. Fifthly, greenhouse gas emissions could be derived, indicating a 0.06% N₂O emission factor. Sixthly, off-gas gave an indication of influent characteristics. The CO₂ emitted per kg COD catabolized corresponded with the TOC/COD ratio of typical wastewater organics in cycles with balanced nitrification and denitrification. High nitrogen removal efficiencies were associated with high catabolized COD/N ratios as estimated from the O₂ absorption. Finally, mass balances could be closed using off-gas O₂ data. As such, an observed yield of 0.27 g COD/g COD was found. All these variables could be estimated with a single sampler because aeration without feeding creates a more homogeneous off-gas composition and simplifies liquid-phase mass balances. Therefore, off-gas analyzers may have a broader application potential for sequentially operated reactors than currently acknowledged.

Combination of 15N Tracer and Microbial Analyses Discloses N2O Sink Potential of the Anammox Community

Although nitrogen removal by partial nitritation and anammox is more cost-effective than conventional nitrification and denitrification, one downside is the production and accumulation of nitrous oxide (N₂O). The potential exploitation of N₂O-reducing bacteria, which are resident members of anammox microbial communities, for N₂O mitigation would require more knowledge of their ecophysiology. This study investigated the phylogeny of resident N₂O-reducing bacteria in an anammox microbial community and quantified individually the processes of N₂O production and N₂O consumption. An up-flow column-bed anammox reactor, fed with NH₄⁺ and NO₂^- and devoid of oxygen, emitted N₂O at an average conversion ratio (produced N₂O: influent nitrogen) of 0.284%. Transcriptionally active and highly abundant nosZ genes in the reactor biomass belonged to the Burkholderiaceae (clade I type) and Chloroflexus genera (clade II type). Meanwhile, less abundant but actively transcribing nosZ strains were detected in the genera Rhodoferax, Azospirillum, Lautropia, and Bdellovibrio and likely act as an N₂O sink. A novel ¹⁵N tracer method was adapted to individually quantify N₂O production and N₂O consumption rates. The estimated true N₂O production rate and true N₂O consumption rate were 3.98 ± 0.15 and 3.03 ± 0.18 mg_N·g_VSS^-1·day^-1, respectively. The N₂O consumption rate could be increased by 51% (4.57 ± 0.51 mg_N·g_VSS^-1·day^-1) with elevated N₂O concentrations but kept comparable irrespective of the presence or absence of NO₂^-. Collectively, the approach allowed the quantification of N₂O-reducing activity and the identification of transcriptionally active N₂O reducers that may constitute as an N₂O sink in anammox-based processes.

Explore the sludge stabilization process in sludge drying bed by modeling study from mesocosm experiments

Sludge drying bed (SDB) treatment is a valuable alternative to conventional sludge treatment methods. However, changes in sludge hydrotexture during dewatering present a barrier for direct modeling of the SDB process. This study proposes a modeling strategy to simulate the sludge stabilization process in SDB treatment by separating sludge dewatering and sludge solids stabilization into independent processes. Two cell decay theories widely used by activated sludge models (ASM), death-regeneration concept and endogenous respiration theory, are compared to describe the biokinetic processes of sludge digestion. Both cell decay theories are found to adequately describe effluent total COD, NH₄-N, NO₃-N, and sludge layer composition, but have pronounced differences in describing effluent COD compositions. Results show that natural aeration does not maintain adequate aerobic/anoxic sludge digestion within the sludge layer to fully nitrify NH₄-N released by cell decay. Results also indicate that the kinetics of sludge digestion are adaptable over time, indicating the need to adopt lumped values for biokinetic simulations. While lowered sludge dewatering rates (outflow) can increase biodegradable COD for cell metabolism, increased sludge loading rates (inflow) lead to higher effluent COD and NH₄-N concentrations. Contrary to conventional judgement, this study demonstrates the merit of sludge layer formation to reduce leaching loss of biodegradable COD. Overall, the proposed modeling strategy is proven capable of simulating deposited sludge digestion processes in an SDB.

Development of an ensemble of machine learning algorithms to model aerobic granular sludge reactors

Machine learning models provide an adaptive tool to predict the performance of treatment reactors under varying operational and influent conditions. Aerobic granular sludge (AGS) is still an emerging technology and does not have a long history of full-scale application. There is, therefore, a scarcity of long-term data in this field, which impacted the development of data-driven models. In this study, a machine learning model was developed for simulating the AGS process using 475 days of data collected from three lab-based reactors. Inputs were selected based on RReliefF ranking after multicollinearity reduction. A five-stage model structure was adopted in which each parameter was predicted using separate models for the preceding parameters as inputs. An ensemble of artificial neural networks, support vector regression and adaptive neuro-fuzzy inference systems was used to improve the models' performance. The developed model was able to predict the MLSS, MLVSS, SVI₅, SVI₃₀, granule size, and effluent COD, NH₄-N, and PO₄^3- with average R², nRMSE and sMAPE of 95.7%, 0.032 and 3.7% respectively.

Optimization of the Aeration Strategies in a Deammonification Sequencing Batch Reactor for Efficient Nitrogen Removal and Mitigation of N2O Production

In deammonification systems, nitrite-oxidizing bacteria (NOB) suppression and nitrous oxide (N₂O) mitigation are two important operational objectives. To carry out this multivariable analysis of response, a comprehensive model for the N cycle was developed and evaluated against experimental data from a laboratory-scale deammonification granular sludge sequencing batch reactor. Different aeration strategies were tested, and the manipulated variables comprised the dissolved oxygen (DO) set point in the aerated phase, aeration on/off frequency (F), and the ratio (R) between the non-aerated and aerated phase durations. Experimental results showed that a high ammonium utilization rate (AUR) in relation to the low nitrate production rate (NPR) (NPR/AUR = 0.07-0.08) and limited N₂O emissions (E_N2O < 2%) could be achieved at the DO set point = 0.7 mg O₂/L, R ratio = 2, and F frequency = 6-7 h^-1. Under specific operational conditions (biomass concentration, NH₄⁺-N loading rate, and temperature), simulation results confirmed the feasible aeration strategies for the trade-offs between the NOB suppression and N₂O emission. The intermittent aeration regimes led to frequent shifts in the predominating N₂O production pathways, that is, hydroxylamine (NH₂OH) oxidation (aerated phase) versus autotrophic denitrification (non-aerated phase). The inclusion of the extracellular polymeric substance mechanism in the model explained the observed activity of heterotrophs, especially Anaerolineae, and granule formation.

Heterogeneous diffusion in aerobic granular sludge

Aerobic granular sludge (AGS) technology allows simultaneous nitrogen, phosphorus, and carbon removal in compact wastewater treatment processes. To operate, design, and model AGS reactors, it is essential to properly understand the diffusive transport within the granules. In this study, diffusive mass transfer within full‐scale and lab‐scale AGS was characterized with nuclear magnetic resonance (NMR) methods. Self‐diffusion coefficients of water inside the granules were determined with pulsed‐field gradient NMR, while the granule structure was visualized with NMR imaging. A reaction‐diffusion granule‐scale model was set up to evaluate the impact of heterogeneous diffusion on granule performance. The self‐diffusion coefficient of water in AGS was ∼70% of the self‐diffusion coefficient of free water. There was no significant difference between self‐diffusion in AGS from full‐scale treatment plants and from lab‐scale reactors. The results of the model showed that diffusional heterogeneity did not lead to a major change of flux into the granule (<1%). This study shows that differences between granular sludges and heterogeneity within granules have little impact on the kinetic properties of AGS. Thus, a relatively simple approach is sufficient to describe mass transport by diffusion into the granules.

An improved ASM-GDA approach to evaluate the production kinetics of loosely bound and tightly bound extracellular polymeric substances in biological phosphorus removal process

This study established an extended activated sludge model no. 2 (ASM2) for providing a new recognition of the contributions of both loosely-bound EPS (LB-EPS) and tightly-bound EPS (TB-EPS) into phosphorus (P) removal by incorporating their formation and degradation processes during the anaerobic-aerobic cycle. For determining the best-fit values for the new model parameters (k _h,TB-EPS, k _h,LB-EPS, f _PP,TB-EPS, and f _PP,LB-EPS) in this extended ASM2, a novel and convenient gradient descent algorithm (GDA) based ASM (ASM-GDA) method was developed. Sensitivity analysis of f _PP,TB-EPS, f _PP,LB-EPS, k _h,TB-EPS, and k _h,LB-EPS on the model target outputs of S _PO4 , X _TB-EPS, X _LB-EPS, and X _PP proved the accuracy of the chosen parameters. Eight batch experiments conducted under different influential chemical oxygen demand (COD) and P conditions were quantitatively and qualitatively analyzed. Respectively, 9.37-9.64% and 4.17-4.29% of P removal by TB-EPS and LB-EPS were achieved. Self-Organizing Map (SOM) has shown its high performance for visualization and abstraction for exhibiting the high correlations of the influential COD/P concentrations and the P% removal by TB-EPS (and LB-EPS). Comprehensive analyses of the influences of influential COD and P concentration on the biological phosphorus removal process help us in successfully establishing the mechanism kinetics of production and degradation of P in a dynamic P biological-treatment model.

Deactivation mechanism of calcified anaerobic granule: Space occupation and pore blockage

Calcification and deactivation of high rate sludge bed reactors is a common and serious engineering problem in the application of anaerobic bioreactor. In this study, the characteristics and deactivation mechanism of calcified anaerobic granules were investigated. The results showed that the calcium content of calcified anaerobic granules was ten times higher than that of control anaerobic granules. A large part of the calcium accumulated in the center of anaerobic granules in the form of calcite, and a small part of the calcium distributed in the outer layer of anaerobic granules in the form of Ca-P deposit. The calcium core occupied a large space which was available for the functional microorganisms. The calcium salts deposited in the outer layer of granular sludge which led to the significant reduction of macropore volume. The porosity of calcified anaerobic granules decreased by 13% compared with that of control anaerobic granules, causing generally the decline of methanogenic activity (for example, by 13% at influent organic concentration of 6.6 g COD L^-1). The substrate gradient created by methanation of organic salts, including organic calcium salts, was deduced to be the driving force of anaerobic granule calcification, while the gradual accumulation of calcium salts in anaerobic granules was deduced to be the dominant factor for the decline of anaerobic granule activity.

Performance and microbial community structure of an aerobic granular sludge system at different phenolic acid concentrations

The present work aims to use aerobic granular sludge technology for the treatment of wastewater containing high organic matter loads and a mixture of phenolic compounds normally present in olive washing water. The physicochemical performance of five bioreactors treating different concentrations of mixture of phenolic acid was monitored to observe the response of the systems. The bioreactors that operated at 50, 100 and 300 mg L^-1 did not show relevant changes in terms of performance and granules properties, showing high ratio of phenolic compound removal ratio. However, the bioreactors operated with high phenolic compound concentrations showed low rates of organic matter, nitrogen and phenolic acid removal. In the same way, high concentrations of phenolic compounds determined the disintegration of the granular biomass. Next-generation sequencing studies showed a stable community structure in the bioreactors operating with 50, 100 and 300 mg L^-1 of phenolic acids, with the genera Lampropedia and Arenimonas, family Xanthobacteraceae and Fungi Pezizomycotina as the dominant phylotypes. Conversely, the reactors operated at 500 and 600 mg L^-1 of phenolic substances promoted the proliferation of Oligohymenophorea ciliates. Thus, this study suggests that aerobic granular sludge technology could be useful for the treatment of wastewaters such as olive washing water.

Artificial Neural Network (ANN) Approach to Modelling of Selected Nitrogen Forms Removal from Oily Wastewater in Anaerobic and Aerobic GSBR Process Phases

Paper presents artificial neural network models (ANN) approximating concentration of selected nitrogen forms in wastewater after sequence batch reactor operating with aerobic granular activated sludge (GSBR) in the anaerobic and aerobic phases. Aim of the study was to determine parameters conditioning effectiveness of selected nitrogen forms removal in GSBR reactor process phases. Models of artificial neural networks were developed separately for N-NH4, N-NO3 and total nitrogen concentration in particular process phases of GSBR reactor. In total, 6 ANN models were presented in this paper. ANN models were made as multilayer perceptron (MLP), which were learned using the Broyden-Fletcher-Goldfarb-Shanno algorithm. Developed ANN models indicated variables the most influencing of particular nitrogen forms in aerobic and anaerobic phase of GSBR reactor. Concentration of estimated nitrogen form at the beginning of anaerobic or aerobic phase, depending on ANN model, in all ANN models influenced approximated value. Obtained determination coefficients varied from 0.996 to 0.999 and were depending on estimated nitrogen form and GSBR process phase. Hence, developed ANN models can be used in further studies on modeling of nitrogen forms in anaerobic and aerobic phase of GSBR reactors.

Modelling anaerobic, aerobic and partial nitritation-anammox granular sludge reactors - A review

Wastewater treatment processes with granular sludge are compact and are becoming increasingly popular. Interest has been accompanied by the development of mathematical models. This contribution simultaneously reviews available models in the scientific literature for anaerobic, aerobic and partial nitritation-anammox granular sludge reactors because they comprise common phenomena (e.g. liquid, gas and granule transport) and thus pose similar challenges. Many of the publications were found to have no clearly defined goal. The importance of a goal is stressed because it determines the appropriate model complexity and helps other potential users to find a suitable model in the vast amount of literature. Secondly, a wide variety was found in the model features. This review explains the chosen modelling assumptions based on the different reactor types and goals wherever possible, but some assumptions appeared to be habitual within fields of research, without clear reason. We therefore suggest further research to more clearly define the range of operational conditions and goals for which certain simplifying assumptions can be made, e.g. when intragranule solute transport can be lumped in apparent kinetics and when biofilm models are needed, which explicitly calculate substrate concentration gradients inside granules. Furthermore, research is needed to better mechanistically understand detachment, removal of influent particulate matter and changes in the mixing behaviour inside anaerobic systems, before these phenomena can be adequately incorporated in models. Finally, it is suggested to perform full-scale model validation studies for aerobic and anammox reactors. A spreadsheet in the supplementary information provides an overview of the features in the 167 reviewed models.

The future of WRRF modelling - outlook and challenges

The wastewater industry is currently facing dramatic changes, shifting away from energy-intensive wastewater treatment towards low-energy, sustainable technologies capable of achieving energy positive operation and resource recovery. The latter will shift the focus of the wastewater industry to how one could manage and extract resources from the wastewater, as opposed to the conventional paradigm of treatment. Debatable questions arise: can the more complex models be calibrated, or will additional unknowns be introduced? After almost 30 years using well-known International Water Association (IWA) models, should the community move to other components, processes, or model structures like 'black box' models, computational fluid dynamics techniques, etc.? Can new data sources - e.g. on-line sensor data, chemical and molecular analyses, new analytical techniques, off-gas analysis - keep up with the increasing process complexity? Are different methods for data management, data reconciliation, and fault detection mature enough for coping with such a large amount of information? Are the available calibration techniques able to cope with such complex models? This paper describes the thoughts and opinions collected during the closing session of the 6th IWA/WEF Water Resource Recovery Modelling Seminar 2018. It presents a concerted and collective effort by individuals from many different sectors of the wastewater industry to offer past and present insights, as well as an outlook into the future of wastewater modelling.