Reduction of Membrane Fouling in Membrane Bioreactors - Development of Innovative and Sustainable Techniques

Over the past decade, membrane technology has become a worldwide implemented, acknowledged separation process. Although membrane bioreactors (MBRs) still require further research efforts re-garding operating and economic efficiency, they offer considerable advantages compared to conven-tional activated sludge systems, such as smaller foot print, modular construction, and superior effluent quality. They are especially preferred in cases where stringent water quality requirements have to be guaranteed. One of the main challenges in the operation of MBRs is membrane fouling, causing flux decrease and the increase of the transmembrane pressure. Its control and prevention result in high manpower re-quirements as well as in increased electrical power consumption and the need for cleaning chemicals. Therefore, most MBR researches aim to identify, investigate, and control membrane fouling. The research on fouling reduction and permeability loss in MBRs was carried out at three MBR pilot plants with synthetic and real wastewater. On the one hand, the effect of mechanical cleaning with an abrasive granular material on the performance of a submerged MBR process was tested. Additionally, scanning electron microscopy (SEM) measurements and integrity tests (rejection measurements and water microbiology testing) were conducted to check whether the membrane material was damaged by granulates. The results indicate that the fouling layer formation was significantly reduced by abrasion using the granular material. This technique allowed a long-term operation of more than 600 days at a flux up to 40 L/(m2h) without chemical cleaning of the membranes. However, at high flux rates (> 35 L/(m2h)) a significant decline in permeability was observed, presumably as a result of permeation drag forces that might draw the particulate and colloidal matter towards the membrane surface. It was demonstrated that the MBR with granulate could be operated with more than 27 % higher flux compared to conventional MBR operation. SEM images and integrity tests showed that in consequence of abrasive cleaning, the granular material left brush marks on the membrane surface; however, the membrane function was not affected. Rejection measurements indicated no damaging effects by granu-late, as the rates were comparable with those of new membranes (> 90 %). Water microbiology test-ing (removal of E. coli and total coliforms) showed that granulate did not have any negative impact on the membranes and still rejected almost all bacteria in the system. Additionally to the experiments with granulate, the impact of membrane backwash, using a salt brine solution, the impact of chemical and physical cleaning (relaxation breaks), and the effect of changed milieu conditions (aerobic, anaerobic) on membrane fouling was investigated. Among the chemical and physical methods relaxation breaks have proved to enhance membrane performance; however, this involves high specific costs due to operation interruptions, i.e. low membrane throughput. Research on permeability loss was carried out to investigate the influence of operating conditions (temperature, sludge retention time/sludge loading, and wastewater composition) and the influence of activated sludge characteristics (mixed liquor suspended solids, colloidal matter, truly soluble matter, viscosity) on fouling. It was shown that the concentration of the colloidal fraction obviously depends on the operating temperature and sludge retention time (SRT). A nearly linear correlation between the colloidal fraction (CODcolloidal) and the concentration of biopolymers (proteins and polysaccharides) was found. However, while a relationship between the retention of biopolymers and capillary suction time (CST) of activated sludge was found, a clearly connection with the permeability of the investi-gated membranes did not exist. It was observed that the addition of polyaluminum chloride coagulant resulted in the reduction of the colloidal fraction in the supernatant by up to 80 % of the initial value. The coagulant even bound soluble matter < 0.04 μm, as the chemical oxygen demand (COD) concen-tration of permeate was slightly higher. Fractionation tests showed that the colloidal fraction (0.04-0.1 μm) constituted the largest percen-tage of COD (71 %) and dissolved organic carbon (DOC) (81 %) of samples during high fouling in the MBR, whereas this fraction made up only 21 % (COD) and 20 % (DOC) of samples during low fouling. It was observed that the biological degradation of organic matter was much higher during periods of low fouling and only 41 % of organic compounds were retained by the membrane in the reactor com-pared to 84 % during periods of high fouling. This implies that most of the biopolymers (proteins, po-lysaccharides) were retained in the reactor by membrane separation. In a parallel experimental set up, the impact of the operationally fraction 0.04 μm rather than by the dissolved fraction of wastewater and activated sludge. The rheological measurements showed a structural viscous behavior of the investigated activated sludge samples. Moreover, it was found that with increasing mixed liquor suspended solids (MLSS) concentration, the viscosity also tends to increase. The viscosity amounted to 18 mPas at a “common” MLSS concentration of 12 g/L in MBRs and a shear rate of 40 s-1. A correlation between the apparent viscosity and the retention of proteins and polysaccharides in the mixed liquor filtrate of activated sludge was observed. The coefficients of both model approaches (Oswald de Waele and Herschel-Bulkley) could be described as a function of MLSS concentration using a non-linear regression analysis, however the Oswald de Waele approach showed a better plausibility with the measured values. It was demonstrated that the energy consumption of the coarse-bubble crossflow system for fouling control is about two times higher than the fine-bubble crossflow system for oxygen supply