The development of continuous flow reactors (CFRs) employing aerobic granular sludge (AGS) for the retrofit of existing wastewater treatment plants (WWTPs) using a continuous-flow activated sludge (CFAS) system has garnered increasing interest. This follows the worldwide adoption of AGS technology in sequencing batch reactors (SBRs). The better settleability of AGS compared to AS allows for process intensification of existing wastewater treatment plants without the difficult conversion of often relatively shallow CFRs to deeper AGS-SBRs. To retrofit existing CFAS systems with AGS, achieving both increased hydraulic capacity and enhanced biological nutrient removal necessitates the formation of granular sludge based on the same selective pressures applied in AGS-SBRs. Previous efforts have focussed mainly on the selective wasting of flocculent sludge and retaining granular sludge to drive aerobic granulation. In this study a pilot-scale CFR was developed to best mimic the implementation of the granulation mechanisms of full-scale AGS-SBRs. The pilot-scale reactor was fed with pre-settled municipal wastewater. We established metrics to assess the degree to which the proposed mechanisms were implemented in the pilot-scale CFR and compared them to data from full-scale AGS-SBRs, specifically with respect to the anaerobic distribution of granule forming substrates (GFS). The selective pressures for granular sludge formation were implemented through inclusion of anaerobic upflow selectors with a water depth of 2.5 meters, which yielded a sludge with properties similar to AGS from full-scale SBRs. In comparison to the CFAS system at Harnaschpolder WWTP treating the same pre-settled wastewater, a more than twofold increase in volumetric removal capacity for both phosphorus and nitrogen was achieved. The use of a completely mixed anaerobic selector, as opposed to an anaerobic upflow selector, caused a shift in EBPR activity from the largest towards the smallest size class, while nitrification was majorly unaffected. Anaerobic selective feeding via bottom-feeding is, therefore, favorable for the long-term stability of AGS, especially for less acidified wastewater. The research underlines the potential of AGS for enhancing the hydraulic and biological treatment capacity of existing CFAS systems.

The aerobic granular sludge (AGS) process treats wastewater with a significantly lower footprint and energy consumption compared to conventional activated sludge systems. Nevertheless, there is still potential for optimizing its performance, and mathematical models are most valuable tools to this end. Aeration energy consumption deserves particular attention, as it is the largest remaining operating cost for AGS systems. Batch-wisely operated reactors show an increasing oxygen transfer efficiency during aeration, which translates into a dynamic alpha factor. However, the dynamic nature of alpha is neglected in current models. The impact of this simplification on the operating performance was addressed for the first time in this study. Through the development of a novel 1-D biofilm reactor model, calibrated to a full-scale AGS plant, it was shown that the alpha dynamics affect both model structure and calibration, as well as the process performance. The description of the dynamic nature of alpha through the empirical relationship with the soluble biodegradable organic carbon required the addition of the state variable representing soluble slowly biodegradable organic carbon (S_CB) to the biokinetic ASM2d model. Simulation results showed that alpha dynamics significantly influences simultaneous nitrification and denitrification and therefore need to be included in mathematical models to optimize AGS process performance. Different process variables such as volume exchange ratio, aeration capacity and granule size can be manipulated to improve reactor design and performance. The practical application of these new insights were discussed regarding the optimization of AGS systems, as well as other batch-wisely operated aerobic wastewater treatment systems.

In the pursuit of reducing carbon footprint and in view of increasing energy prices, energy efficiency is more important than ever before. Batch-wise operated aerobic granular sludge reactors consume up to 50% less energy compared to conventional activated sludge systems because pumping energy is reduced and mixing equipment is not needed. Further energy reduction efforts should therefore target aeration energy requirements. The alpha factor is an important factor influencing the oxygen transfer efficiency, however the dynamic behaviour of alpha has hardly been investigated in general and never for an aerobic granular sludge reactor. This study showed that alpha increases during the aeration phase of a cycle due to the influence of different process parameters. Through a data analysis study of 175 batch cycles of the Prototype Nereda® installation in Utrecht over the summer and winter period of 2020–2021, the exchange ratio and temperature were identified as the main influencing factors on the rate of increase of alpha in a batch cycle. A higher exchange ratio was related to a slower increase in alpha over the aeration phase, while a higher temperature was related to a faster increase in alpha. Moreover, alpha was characterized by a same minimal value at the beginning of every aeration phase, which could be explained by the adsorption of soluble biodegradable organic carbon described by a Langmuir adsorption model. Two mathematical models, a decreasing exponential and a first order model, were set up to unravel the dynamic behaviour of alpha. Both models were discussed in view of their practical implications for the design and performance optimization of aerobic granular sludge reactors and other batch-wise operated aerobic wastewater treatment systems.

In this study a mathematical framework was developed to describe aerobic granulation based on 6 main mechanisms: microbial selection, selective wasting, maximizing transport of substrate into the biofilm, selective feeding, substrate type and breakage. A numerical model was developed using four main components; a 1D convection/dispersion model to describe the flow dynamics in a reactor, a reaction/diffusion model describing the essential conversions for granule growth, a setting model to track granules during settling and feeding, and a population model containing up to 100,000 clusters of granules to model the stochastic behaviour of the granulation process. With this approach the model can explain the dynamics of the granulation process observed in practice. This includes the presence of a lag phase and a granulation phase. Selective feeding was identified as an important mechanism that was not yet reported in literature. When aerobic granules are grown from activated sludge flocs, a lag phase occurs, in which not many granules are formed, followed by a granulation phase in which granules rapidly appear. The ratio of granule forming to non-granule forming substrate together with the feast/famine ratio determine if the transition from the lag phase to the granulation phase is successful. The efficiency of selective wasting and selective feeding both determine the rate of this transition. Brake-up of large granules into smaller well settling particles was shown to be an important source for new granules. The granulation process was found to be the combined result from all 6 mechanisms and if conditions for either one are not optimal, other mechanisms can, to some extent, compensate. This model provides a theoretical framework to analyse the different relevant mechanisms for aerobic granular sludge formation and can form the basis for a comprehensive model that includes detailed nutrient removal aspects.

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.

The nitrous oxides emission was measured over 7 months in the full-scale aerobic granular sludge plant in Dinxperlo, the Netherlands. Nitrous oxide concentrations were measured in the bulk liquid and the off-gas of the Nereda® reactor. Combined with the batch wise operation of the reactor, this gave a high information density and a better insight into N₂O emission in general. The average emission factor was 0.33% based on the total nitrogen concentration in the influent. The yearly average emission factor was estimated to be between 0.25% and 0.30%. The average emission factor is comparable to continuous activated sludge plants, using flocculent sludge, and it is low compared to other sequencing batch systems. The variability in the emission factor increased when the reactor temperature was below 14 °C, showing higher emission factors during the winter period. A change in the process control in the winter period reduced the variability, reducing the emission factors to a level comparable to the summer period. Different process control might be necessary at high and low temperatures to obtain a consistently low nitrous oxide emission. Rainy weather conditions lowered the emission factor, also in the dry weather flow batches following the rainy weather batches. This was attributed to the first flush from the sewer at the start of rainy weather conditions, resulting in a temporarily increased sludge loading.

The settling behavior of aerobic granular sludge (AGS) in full-scale reactors is different from the settling of normal activated sludge. Current activated sludge models lack the features to describe the segregation of granules based on size during the settling process. This segregation plays an important role in the granulation process and therefore a better understanding of the settling is essential. The goal of this study was to model and evaluate the segregation of different granule sizes during settling and feeding in full-scale aerobic granular sludge reactors. Hereto the Patwardhan and Tien model was used. This model is an implementation of the Richardson and Zaki model, allowing for multiple classes of particles. To create the granular settling model, the most relevant parameters were identified using aerobic granular sludge from different full-scale Nereda® reactors. The settling properties of individual granules were measured as was the bulk behavior of granular sludge beds with uniform granular sludge particles. The obtained parameters were combined in a model containing multiple granule classes, which then was validated for granular sludge settling in a full-scale Nereda® reactor. In practice a hydraulic selection pressure is used to select for granular sludge. Under the same hydraulic selection pressure the model predicted that different stable granular size distributions can occur. This indicates that granular size distribution control would need a different mechanism then the hydraulic selection pressure alone. This model can be used to better understand and optimize operational parameters of AGS reactors that depend on granular sludge size, like biological nutrient removal. Furthermore insights from this model can also be used in the development of continuously fed AGS systems.

The main processes contributing to elevated effluent suspended solids in the full-scale aerobic granular sludge process were studied. The two processes found to be most important were (1) rising of sludge due to degasification of nitrogen gas (produced by denitrification) and (2) wash-out of particles that intrinsically do not settle such as certain fats and foams. A mathematical model was made to describe the process of degasification of nitrogen gas during the feeding phase in an AGS reactor. The process of rising sludge due to degasification could be limited by stripping out the nitrogen gas before starting the settling phase in the process cycle. The wash-out of scum particles could be reduced by introducing a vertical scum baffle in front of the effluent weir, similar to weirs in traditional clarifiers. A full-scale Nereda^® reactor in the municipality of Utrecht, The Netherlands, was operated with a nitrogen stripping phase and scum baffles for 9 months at an average biomass concentration of 10 g L⁻¹ and an average granulation grade of 84%. In this period the influent suspended solids concentration was 230±118mgL⁻¹ and the concentration of effluent suspended solids was 7.8±3.8mgL⁻¹.

For a stable operation, the aerobic granular sludge process requires mechanically strong granules in balance with the shear forces in the reactor. Despite a wide general interest in granular stability, the mechanical strength of both anaerobic and aerobic granular sludge received very little attention. In this study, a high-shear method for strength characterization has been evaluated for full-scale aerobic granular sludge (AGS). Abrasion times up to 90 min showed a stable abrasion rate coefficient (K), while prolonged periods of abrasion up to 24 h resulted in a decrease in abrasion rate. Larger granules have higher abrasion rate than smaller granules. No abrasion was observed at low shear rates, indicating a threshold shear rate for abrasion. Lab-scale AGS showed a lower abrasion rate than full-scale AGS. Incubation of full-scale granules in NaCl led to a decrease in abrasion rate at 25 g L⁻¹ NaCl, but incubation in 50 g L⁻¹ NaCl led to a further decrease for only half of the tested granular sludge samples.