Biofilm development on new and cleaned membrane surfaces

This thesis presents a comprehensive research report on microbiological aspects of biofouling occurrence in full-scale reverse osmosis (RO) systems. Biofouling is a process in which microorganisms attach to membranes and develop into a thick film that can choke the entire RO system. Management of this problem requires basic understanding of the mechanism of this phenomenon. The basic questions of this PhD research project therefore addressed the origin, succession and spatiotemporal development of biofilms in full-scale RO systems, in particular in relation to operational aspects of RO systems. The multifaceted research strategy involving acquisitions of representative samples and use of many molecular and microscopic analysis techniques in parallel was employed. The investigation showed that biofilms are able to grow on any surface in a full-scale RO plant. This gives local niches for detachment of biomass, either as single cells or cell clumps, and results in a spreading of bacteria to the further stages of the plant. In the RO membrane modules, the enriched bacteria might more easily colonise the surfaces since they will be better adapted to growth in the system than bacteria present in the feed water. Initially, the single cell colonizers (sphingomonads) form a number of flat and abundantly EPS-embedded cell monolayers over the entire membrane surface. The clumps-associated pioneers (mainly Beta- and Gammaproteobacteria) appear to be trapped mainly in the first part of the module, most likely due to a filtering action of the spacer. In time, these bacteria develop in pillar-like structures and slowly spread throughout the whole membrane module on top of the established sphingomonads biofilm. The secondary colonisers (bacteria and eukaryotes) occur in the resulting biofilm formations. Although composition of the biofilm microbial community undergoes a succession in time, the architecture of an established biofilm appears to be rather stable. Conventional treatment of RO membrane modules with chemicals did not lead to cleaning: the sphingomonads cells can be detected under the collapsed but obviously not removed biofilm EPS matrix. After cleaning, the biofouling layer seemed to grow faster (within 6 days) than a fresh biofilm (16 days). To conclude, biofouling is a complex phenomenon with two appearances: a fouling layer on the membrane limiting the water flux and a fouling layer on the spacer limiting the water flow through the spacer channel and resulting in an increased pressure drop. It became clear that cleaning strategies should focus more on the removal of accumulated biomass and not only on the killing of cells. Moreover, the basal Sphingomonas layer requires further research to appropriately control biofouling in RO systems. It might also be possible to design the RO - membrane module in a different manner, leading to a different biofilm morphology which gives less rise to operational problems.