ABSTRACT
Subcritical flux operation is widely practiced in membrane bioreactors (MBRs) to avoid severe membrane fouling and, thus, to maintain sustainable permeability. Filtration at a constant subcritical flux, however, usually leads to a two-stage increase in the transmembrane pressure (TMP): initially slowly, then abruptly. We have investigated the mechanism of this two-stage TMP increase through analyses of the structure and microbial characteristics of the bio-cake formed on the membrane. The MBR was operated under various subcritical and supercritical flux conditions. Under subcritical conditions, we observed the typical two-stage TMP increase. When a constant flux augmented and reached the supercritical conditions, however, the dual TMP change gradually transformed into a steeper, one-stage TMP increase. The second stage TMP increase under the subcritical flux was closely related to the sudden increase in the concentration of extra-cellular polymeric substances (EPSs) at the bottom layer of the bio-cake; we attribute the one-stage TMP increase under the supercritical conditions to the accumulation of microbial flocs and the reduced porosity of the bio-cake under compression. We explain the variation of the EPS concentration in the bio-cake in terms of the spatial and temporal changes of the live-to-dead ratio along the depth of the bio-cake.
Subject(s)
Bioreactors , Bacteria/metabolism , Biofilms , Biopolymers/metabolism , Polysaccharides/metabolism , PressureABSTRACT
The effects of a sequencing variation for dissolved oxygen (DO) concentrations on the membrane permeability in a submerged membrane bioreactor (MBR) were studied. An MBR was continuously operated under alternating DO conditions, e.g., 36 h of an aerobic phase, followed by 36 h of an anoxic phase. The rate of increase in transmembrane pressure (TMP) in the anoxic phase was always steeper than that in the aerobic phase, indicating that the fouling rate was higher in the anoxic than in the aerobic condition. Regardless of the phases, the rate of TMP increase became steeper as the cycles were repeated. However, this trend became less important as the cycle numbers increased. Even in identical microbial communities, the number of colloidal particles and soluble extracellular polymeric substances (EPS) in the bulk solution were increased during the anoxic condition, which caused a reduction in the porosity of the bio-cake. During analysis of the bio-cake profile along the cake depth, the temporal variation of the bio-cake structure was attributed to the temporal change in the number of colloidal particles as well as the change in compression forces acting on the bio-cake. The influence of the latter was found to be more important than that of the former, which was verified by comparing the various structures of bio-cake formed in differing DO environments.
Subject(s)
Bacteria/metabolism , Bioreactors , Membranes/chemistry , Oxygen/chemistry , Waste Disposal, Fluid/methods , Water Purification/methods , Aerobiosis , Bacteria/chemistry , Colloids/chemistry , Colloids/metabolism , Extracellular Matrix/chemistry , Extracellular Matrix/metabolism , Permeability , Porosity , Solubility , Time Factors , Waste Disposal, Fluid/instrumentation , Water Purification/instrumentationABSTRACT
The structures of biofilms deposited on the membrane surface under different dissolved oxygen (DO) conditions were characterized to identify its relation to membrane filterability in membrane bioreactors (MBR). The rate of membrane fouling for the low DO (LDO) reactor was 7.5 times faster than that for the high DO (HDO) reactor. Even though the biofilm deposited on the membrane surface in the HDO was thicker than in the LDO at the operating terminated (TMP reached 30 Kpa), biofilm resistance in both reactors were similar. Exactly, specific cake resistance of the HDO was lower than that of the LDO. Difference in biofilm characteristics as a result of different DO level was main factor affecting biofouling for both MBRs. The number of small particles ranging from 2-5mum in the biofilm as well as in the bulk solution for the LDO was greater than those for the HDO. The small particles in the bulk solution of the LDO more preferentially deposited on the membrane surface than those of HDO did. Hence, the biofilm porosity in the LDO (0.65) was smaller than that in HDO (0.85). The reduced porosity of LDO biofilm resulted in lower filterability than the HDO. The porosity data obtained from analysis of images of biofilm using confocal scanning laser microscopy (CLSM) was verified in terms of specific cake resistance (alpha) by comparing the experimentally measured values with the semi-theoretically computed values.
