Formation mechanisms and control of biofouling in submerged MBRs

• Main Author: Luo, Jinxue
• Other Authors: Scott Rice
• Format: Theses and Dissertations
• Language: English
• Published: 2015
• Subjects: DRNTU::Science::Biological sciences::Microbiology::Microbial ecology; DRNTU::Science::Biological sciences::Microbiology::Microorganisms; DRNTU::Engineering::Environmental engineering::Water treatment; DRNTU::Engineering::Environmental engineering::Environmental pollution
• Online Access: https://hdl.handle.net/10356/62182
• Institution: Nanyang Technological University

Membrane bioreactors (MBRs) are increasingly being applied in modern wastewater treatment plants. Biofouling of the MBRs represents a significant challenge in the application of membrane based technologies for water purification. Biofouling is the build-up of organic and biological cake-layers on the membranes that may block the membrane pores. The biofouling layers will reduce membrane permeability and increase the hydraulic resistance of the membrane. Thus, membrane fouling results in an increase of the trans-membrane pressure (TMP) when operated at a constant flux or a decreased flux when operated at a constant pressure. Therefore, it is necessary to identify the relative contribution of microbes and macromolecules to the biofouling process in order to develop appropriate strategies to reduce fouling and hence increase operational efficiency. Two identical laboratory-scale MBRs were operated at a low, constant flux (13 - 15 LMH) to treat artificial synthetic wastewater (TOC of 200 mg/L). The TMP was maintained at a low pressure (3 - 15 kPa), steady state for the first 80 - 87 d of operation and then was observed to increase exponentially from 15 to 90 kPa over 30 d. The biofouling layers on the hollow fiber membrane surfaces were observed to contain significant amounts of α-polysaccharides, β-polysaccharides, proteins and microorganisms that were always present on the hollow fiber membranes, even during the early stages of MBR operation when the TMP was still in the low pressure phase (Chapter 2). Quantitative image analysis indicated that each of these components on membrane correlated positively with the TMP increase, Pearson’s correlation coefficients 0.7 - 0.95. Among the four components, the proteins increased fastest when the TMP was rapidly increasing and comprised the greatest proportion of the individual components when TMP increased. This indicated that the production of proteins was more important than the two types of polysaccharides or the cells during the transition of TMP from the low to high TMP stage. Additionally, co-localization analysis revealed that approximately 50% of the EPS co-localized with 80 - 90% of the cells. The co-localization data indicated that the majority of the EPS components were closely associated with the cells, suggesting that the EPS components may be the byproducts of microorganisms on membrane rather than being randomly distributed on the membrane from the aqueous phase. Therefore, the formation of microbial biofilms on the membranes is the key driver of the biofouling process in MBRs. Thus, it is vital to develop methods to prohibit or reduce biofilm formation on the membranes in situ or to disperse the mature biofilms. As part of the development of novel strategies to control biofilms on membranes, it is also essential to determine if the biofilm is formed by a selected subset of microorganisms in the sludge community or if biofilm formation is a stochastic process. Therefore, the microbial biofilm community, including bacteria (Chapter 3) and fungi (Chapter 4), were investigated through 16S and ITS tagged pyrosequencing at different stages of the fouling process. At low TMP, the biofilms were most highly similar to the sludge and as the biofilm developed and the TMP increased, the biofilm communities diverged from the sludge. Ultimately the biofilm community appeared to be distinct from the sludge and the greatest differences were seen for the bacterial community. In contrast, the fungal communities were overall less diverse and showed only minor differences in relation to the sludge. This was mostly seen in subtle shifts in the percentage composition of individuals, while the community was still dominated by the same groups. It was noted that the correation between the richness of biofilm bacterial community and TMP increase was not evident in this study. Compared to the sludge community, the bacteria including Burkholderiales, Pseudomonadales and Rhizobiales and fungi including the Archaeosporales and Hypocreales were enriched in the biofilm, indicating these microorganisms were more fit when growing as biofilm compositions relative to the growth and competition in the planktonic sludge. Additionally, during the process of TMP increase, the Alphaproteobacteria, represented by Rhodospirillales, Sphingomonadales and Rhizobiales in this project, and the fungi including Saccharomycetales and Hypocreales became more dominant in the late stage biofilms, indicating these organisms may contribute more to the construction of late rather the early biofilm and may play an important role to the TMP increase in the biofouling process. These results indicated that the change of microbial community that occurred before the TMP jump may be the most important in its effect on biofouling process. It may be possible to target those organisms to ultimately delay their incorporation into the biofilm and hence delay the TMP jump. Finally, strategies based on nitric oxide (NO) induced biofilm dispersal were tested to control biofilm formation and TMP increase in the MBR (Chapter 5). The potential for NO to control biofilms in MBR systems was tested using two distinct approaches. The first was to disperse pre-established, mature biofilms that had developed during MBR operation and the second was to prevent biofilm accumulation by applying the NO from the beginning of the MBR operation. The results showed that treatment using the NO donor PROLI NONOate resulted in a 50% (in pre-established biofilm dispersal experiment) and 28.2% (in the biofilm prevention experiment) reduction of fouling resistance. The CLSM analysis showed that, in the biofilm prevention experiment, the NO treatment also resulted in a reduction of biofilm biomass, for both cells (66.7% reduction) as well as macromolecules (e.g. 37.7% reduction for proteins). Analysis of the community composition indicated that the bacterial Orders of Thiotrichales, Gemmatimonadales and Xanthomonadales and fungal Orders of Hypocreales and Glomerales were reduced in abundance after the PROLI NONOate treatment. Furthermore, the development of the community associated with the late stage biofilm was delayed, but not entirely prevented, as a consequence of the PROLI treatment. These results demonstrated that the NO donor PROLI NONOate had the potential to control biofouling in MBRs.