Flux enhancers in biofilm membrane bioreactors : insight into fouling control

Conventional wastewater treatment technologies are not able to keep up with the population growth and new challenges and restrictions, induced by climate change. Membrane bioreactor (MBR) technology has a high potential to become a solution of choice when implementing the concepts of minimal/zero liquid discharge and 100% recycling. However, membrane fouling highly inhibits the expansion of MBR technology. Among the variety of methods to fouling mitigation, the integration of the biofilm membrane bioreactor technology (BF-MBR), which is the advancement of the MBR, with the coagulation approach into a concept of flux enhancement has been showing high potential. Due to the complexity of the fouling phenomenon and underlying mechanisms of its mitigation, process control in coagulant-assisted BF-MBR remains a major gap in its investigation. The development of the universal strategy for real-time fouling control and prediction in coagulant-assisted BF-MBR requires the comprehension of the membrane fouling patterns, determination of the underlying mechanisms of coagulant action, and system optimisation. According to the analysis of the latest studies, performed in Paper I, membrane fouling phenomenon is primarily caused by floc-bound extracellular polymeric substances (EPSs) and soluble microbial products (SMPs), whose accumulation causes pore blockage and cake and gel layer formation, resulting in reversible, irreversible and irrecoverable fouling. Polysaccharides of the SMPs were identified as the primary foulants in MBR. Among the most efficient approaches to flux enhancement, the use of coagulants was recognized as one of the most promising directions. The outlined mechanisms of fouling mitigation during chemical flux enhancement in MBR became the instrument of “navigation” in the development of an advanced fouling mitigation strategy with regard to BF-MBR at the further stages of the present research. Paper II was focused on the identification of membrane fouling patterns in the BF-MBR. During 114 days of system performance, a large representative, with regard to the application of the variety of operational conditions in BF-MBR, data array was analysed. One of the main contributions of this work was the development of the multivariate chemometric approach to fouling monitoring and control in BF-MBR based on the interrelation between the characteristics of the mixed liquor fouling propensity, membrane fouling indicators, and operational conditions. The introduction of the term “critical solids retention time” (SRT) allowed for determining SRT operational window with regard to the absence of severe fouling, which was less than 31 days for the investigated BF-MBR. The automated monitoring of the MLSS concentration in the moving bed biofilm (MBBR) and the separation chambers was implemented. The developed partial least squares (PLS) models were used to adjust the operational conditions (SRT, permeate flux, and sludge recirculation intensity) according to the characteristic of mixed liquor, which enabled the system to work below the critical fouling levels. In Paper III, the comparative analysis of the prepolymerized and non-prepolymerized coagulants with regard to flux enhancement in BF-MBR revealed that the extent of chemical flux enhancement via floc-bound EPSs removal highly correlated with the intrinsic charge concentration of the coagulants and their basicity. pH correction was shown to have a tremendous influence on fouling mitigating performance of every coagulant. The operational settings in the pilot system, which was the source of mixed liquor, were kept at the optimum levels, determined in Paper II. Paper III addresses one of the main concepts of the thesis, i.e., the development of the core model for the soft sensor for fouling control and prediction based on the electrostatic mechanism of the floc-bound EPSs removal during coagulation. The optimisation analysis applying the most effective coagulant revealed the significance of mixed liquor suspended solids (MLSS), coagulant dose, temperature, pH, and the interaction factors in the filtration processes. The derived optimisation models allow for minimising the adverse effects of low temperatures on the membrane filtration process by adjusting the operational conditions in BF-MBR. Paper III allowed for the determination of the most efficient coagulant with respect to flux enhancement and the optimum pH of its performance in BF-MBR. Based on these findings and according to the literature analysis, conducted in Paper IV, the active hydrolysis species of the optimum coagulant (i.e., Al13 2+ complex of the prepolymerized aluminium chloride with medium basicity) and the dominant carbohydrates in the structure of SMPs in MBR and BF-MBR systems were identified. By applying the quantum chemical and thermodynamic calculations, and multivariate chemometric analysis, the other main concept of the thesis was addressed, i.e., the deriving of the core model for the soft sensor for membrane fouling estimation and prediction based on the chemical coordination of the carbohydrates of SMPs to the Al13 2+ complex. In Paper IV, a two-stage supervised classifier was developed based on data mining and cluster analysis of the PLS Y-scores. The results of this study served as a basis for developing an unsupervised hierarchical classifier for online monitoring of the reduction of the mixed liquor fouling propensity during coagulation in BF-MBR based on the thermodynamic parameters of the system