This study focuses on the metabolic impacts of simultaneous glucose and oxygen concentration gradients on penicillin production in an industrial-scale fermentor, using the computational fluid dynamics-cellular reaction dynamics approach. Inclusion of oxygen-coupling considerably impacts the glucose uptake and resulting penicillin productivity. This is characterised by six metabolic regimes; lifeline data reconstructed from experimental results, recorded from the cellular perspective, indicates rapid dynamics in glucose and dissolved oxygen uptake by the microorganisms. The results are highly sensitive to variations in the oxygen-related model parameters, requiring accurate insight into the multiphase hydrodynamics and metabolic processes. Hypothetical scenarios with stronger glucose-oxygen limitations than tested experimentally were further explored. A precision scale-down (SD) simulator was designed based on the lifeline data, requiring considerable operational dynamics, with increasing system complexity and implementation difficulty. These insights may inspire further research into alternative SD configurations better suited to mimic the rapid dynamics of large-scale fermentation processes.

Gas-mixing is commonly applied in anaerobic digesters, yet the resulting flow and hydraulic mixing are difficult to evaluate because of limited full-scale experimental data and uncertainties in integrating sludge rheological data. This study used computational fluid dynamics (CFD) to assess the impact of treated sludge rheology on flow and mixing characterisation in a full-scale biogas-mixed digester. The CFD model, which was firstly validated using a lab-scale setup, showed that flow and mixing predictions depended on the rheological properties, especially at low shear rates. The predicted dominant shear rate was out of the effective shear-rate range of the Ostwald model, leading to flow and mixing performance overestimation. The results indicated that there are limitations in applying the Ostwald model and the conventional approaches for determining dead-zone. The Herschel-Bulkley model was more appropriate for the prevailing low shear rates and predicted large viscosity gradients in the digester, indicating two distinct compartments with different flow and mixing behaviour based on the gas-sparging height: a plug-flow compartment with dominant vertical convection above, and a dead-zone compartment with considerable segregation below. The results showed that the applied gas-sparging induced insufficient flow and mixing, but contributed to the well-functioning of the digester. To correctly assess flow and mixing, the applied rheological data should be in agreement with the type of sludge that is treated in the digester. Our results indicate that the shear rate in the digester must be increased and various options for achieving this are proposed.

Accurate rheological characterisation of sewage waste activated sludge (WAS) is of high importance for downstream processing related to optimised sludge pumping and mixing, assessment of energy demands and overall process design. However, to elaborate rheological behaviour is often challenging under dynamic operational conditions in practice. In this study, two practical influencing factors were investigated: long-term shearing and temperature. Compared to anaerobic digestate (DGT), concentrated WAS had more complex and stronger thixotropic behaviour. Under the long-term shearing conditions, the sludge thixotropic behaviour was well characterised by two quantified limitation states. Temperature had a striking impact on the rheological properties, which was strongly correlated to solids content and digestion process. The impact discrepancy between the long-term shearing and temperature, implied different mechanisms to shift the equilibrium of hydrodynamic and non-hydrodynamic interactions for structure deformation and recovery. The distinct rheological properties between the two determined states were clearly reflected in pipe flow behaviour, revealing a concrete link between lab-measured sludge rheology and its practical flow performance. The pipe flows were well assessed using the developed Computational Fluid Dynamics model with effective rheological data integration, which is promising for practical design and optimisation of sewage sludge systems.

This study aimed to characterise the gas-liquid flow and mixing behaviour in a gas-mixed anaerobic digester by improving phase interaction modelling using Computational Fluid Dynamics (CFD). A 2D axisymmetric model validated with experimental data was set up using an Eulerian-Eulerian method. Uncertainty factors, including bubble size, phase interaction forces and liquid rheology were found to significantly influence the flow field. A more reliable and complete validation was obtained by critical comparison and assessment of the referred experimental data, compared to the models reported in other studies. Additionally, justifiable corrections and predictions in detail were obtained. Mixing was evaluated by trajectory tracking of a large number of particles based on an Euler-Lagrange method. The mixing performance approximated to a laminar-flow reactor (LFR) that distinctly deviated from expected continuous stirred tank reactor (CSTR) design, indicating limited enhancement from the applied gas-sparging strategy in the studied digester. The study shows the importance of a proper phase-interaction description for a reliable hydrodynamic characterisation and mixing evaluation in gas-mixed digesters. Validations, bend to experimental data without a critical assessment, may lead to an inaccurate model for further scaled-up applications.

For a proper operational performance assessment of excess sewage sludge digesters in practice, a better understanding of waste activated sludge (WAS) rheological behaviour is important, especially regarding the low-shear and poor mixing zones in anaerobic digesters. The potential rheological instability of WAS with different total solids (TS) concentrations was studied in this research. The obtained yield-pseudoplastic behaviour showed varying patterns in a wide shear rate range, of which characterisation depended on hybrid model fitting of defined shear rate segments. Although a new mathematical expression improved the fitting quality, limited applicability of the empiric models reflected the samples’ transient rheological behaviour rather than intrinsic properties, challenging the included parameters definition. Characterised by the distinct flow status and transitions, the observed rheological instability gives more insight in viscoelastic and thixotropic effects on sludge flow and mixing behaviour in full-scale WAS treatment systems. Recommendations for developing a rheological measurement protocol were also formulated.