Limitations for transferring lab-scale microfiltration results to large-scale membrane bioreactor (MBR) processes

Abstract

Membrane bioreactor (MBR) processes for the treatment of municipal and industrial wastewater have been extensively investigated in the past years. For this purpose, synthetic model feeds containing dissolved macromolecules and/or suspended solids were used to allow for a stable and reproducible feed composition. Lab-scale setups equipped with single membrane filaments and pilot plants containing single-bundle configurations were applied. Despite the fact that numerous of the corresponding findings were successfully transferred to large-scale MBR processes, it is expected that the filterability of activated sludge differs significantly from that of common model substances. Moreover, phenomena linked to the highly complex multi-phase flow of liquid, suspended solids and dispersed air bubbles which physically interacts with the membrane filaments make downsizing difficult. Therefore, the overall objective of this study was to evaluate the transferability of lab-scale microfiltration results to large-scale MBR processes. Our investigations have shown that the rheology of a model feed containing silica particles is similar to that of activated sludge with MLSS⩽3.5g/L. A noticeable sub-critical fouling was observed, but was removable by an intensive chemical cleaning using sodium hypochlorite. Lower fouling rates when increasing the aeration frequency with the same net aeration rate were found for both single-fibre and single-bundle tests, but are more pronounced on the larger scale. A macroscopic circulation flow affects the filtration performance of full-scale MBRs, but is difficult to mimic on a smaller scale. Fibre motion enhances the filtration performance, but is insensitive to packing density variations for bundles of up to four filaments.