Effect of feeding pattern and storage on the sludge settleability under aerobic conditions

The selection of filamentous bacteria is often assumed to be associated with specific microbial properties such as growth rate, substrate uptake rate, substrate affinity and potential for substrate storage. In this study we aimed to verify some of these factors. Sequencing batch reactor (SBR) systems were used to scale-down aerobic activated sludge systems with an aerobic selector. Adding acetate in different aerobic feeding periods allowed us to simulate a variable relative size of aerobic selector with different bulk liquid substrate concentrations. The experiments showed that as expected, the aerobic fill time ratio (FTRox) and the corresponding feast period, which can be assumed similar to contact time in an aerobic selector, had a strong effect on the sludge settleability. Promoting a strong substrate gradient in the SBR (FTRox<5.4%) resulted in good sludge settleability (SVI<120 mL g^-1). Whenever acetate was added in a limiting rate (FTRox>6.2%), a condition in which the acetate concentration in the reactor was always very low, the sludge settleability decreased (SVI>150 mL g^-1). Sludge settleability could be improved by changing the feeding strategy to a pulse feed. The maximum specific acetate uptake rate and poly (beta)-hydroxybutyrate (PHB) production rate of bad settling sludge, including bulking sludge, was similar to wellsettling sludge, which is not in accordance with the general assumptions that well settling sludge have a higher maximal substrate uptake rate and better storage capacities. An alternative hypothesis for the development of filamentous structures in biological flocs has been formulated. It is hypothesized that bulking sludge originates from the presence of substrate gradients in sludge aggregates. Whereas at low bulk liquid substrate concentration filamentous bacteria give easier access to the substrate at the outside of the flocs and thereby proliferate, at high bulk liquid substrate concentration there is no substrate advantage for filamentous organisms and smooth bacterial structures predominate. In this hypothesis there is no need for an intrinsic difference in kinetic parameters between floc and filamentous bacteria. Where presence of filamentous bacteria is related to process conditions, the presence of a specific filament is likely due to presence of a specific limiting substrate.