Performance comparison of membrane bioreactor with dynamic membrane bioreactor

Membrane bioreactors (MBRs) used for wastewater treatment have begun to be used at an increasing rate in developed and developing countries in the last 10 years. It is anticipated that many conventional wastewater treatment plants will convert to the MBR technology in the coming years, and as a result, the final sedimentation basins will come to an end. MBRs are biological processes in suspension growth mode in which purified water and biomass are physically separated by membrane equipment. However, membrane costs, increased energy demand, clogging control and low flow bioreactor operating problems are among the major problems faced by membrane bioreactor systems. Dynamic membrane bioreactor systems (DMBR), due to its low operating cost, have provided a promising approach to solving these problems. The purpose of this study is to measure the pollution performance of two different membrane types, 0.45 μm and non-woven dynamic membranes, which are operated in the same aerobic membrane bioreactor (AeMBR) and fed with synthetic wastewater. During reactor operation, parameters of COD, turbidity, pH, temperature, electrical conductivity (EC), TMP, SADm and ORP were measured daily. The highest COD removal values were found to be 99.47% and 97.02% and the lowest turbidity values were found to be 0.12 NTU and 1.23 NTU for 0.45 µm real and non-woven dynamic membrane, respectively. The maximum and minimum permeate pH values of the AeMBR operated at room temperature, for 0.45 µm real and non-woven dynamic membrane, were found to be 7.98-7.94 and 7.37-7.43, respectively. The maximum and minimum values for electrical conductivity were measured as 574-577 μS and 470-471 μS, respectively, as 0.45 μm real and non-woven dynamic membrane. As a result of the TMP measurements, both membrane materials were subjected to chemical washing (1 hour 0.5% NaOCl and then 1 hour pH = 2 set in tap water) with the pressure rising 300 mbar or above. It is noted that in the measurements made after washing, the membranes acted as if they were used for the first time in operation. In the ORP measurements made in the reactor, the maximum and minimum values were measured as 177.4 and 114.2 for 0.45 µm real and non-woven dynamic membrane, respectively. Also, soluble microbial products (SMP) / extra polymeric substances (EPS), specific resistance to filterability (SRF) / supernatant filterability (SF), fourier transform infrared spectroscopy (FT-IR), capillary suction time (CST), scanning electron microscopy (SEM) and denaturing gradient gel electrophoresis (DGGE) analysis were included in this study. The study consisted of three experimental periods and on the first day of each experimental period, a new membrane was placed in the module. The first experimental period was successfully operated at 10 LMH flux. During this time, approximately 98% COD removal and 1 NTU turbidity values were measured for both membrane types. During operation, the membranes exhibiting high filtration performance with only physical washing were subjected to chemical washing only once, with the pressure reaching 300 mbar and above. During the first experimental study, once in a week SMP / EPS and only once SEM, SRF / SF, CST, FT-IR and DGGE analysis were performed. In the second period of the experimental run, the flux was increased to 15 LMH and the reactor was run successfully. The average COD removal efficiencies in percentage for 0.45 µm real and non-woven dynamic membrane were recorded as 97% and 90%, respectively. However, physical and chemical washing was applied to both types of membranes against the fouling problems originating from the high flux. The chemical washout was reduced by 97% of the membrane flocking for both membrane types. After chemical washing, the turbidity values of the 0.45 µm real membrane did not change much (around 0.2 NTU) but the turbidity values of the non-woven dynamic membrane gave high values until the cake layer was formed after washing. Minimum turbidity values were found to be between 1-2 NTU for the non-woven dynamic membrane. During the second experimental period, once in a week SMP/EPS and only once SRF/SF and CST were performed and the results compared with the previous experimental period. In the last experimental period, the flux was increased to 20 LMH, during which time the non-woven dynamic membrane fouled swiftly and thus did not perform a successful filtration process. While the turbidity values of the non-woven dynamic membrane layer reached approximately 8 NTU levels, the turbidity values of the 0.45 µm real membrane remained at about 0.2 NTU. Both membranes reached a TMP of approximately 10 mbar as a result of chemical washing which means 97% of the fouling reduction. During the last experimental period, only once SMP/EPS, SRF/SF, CST, FT-IR and DGGE analyses were performed and the results were compared with the other analyses. As a result of FT-IR analysis, organic pollutants on the sludge have been assigned on the graph and for the SEM analysis, organic substances on the membrane were identified, thickness of the cake layer formed on the membrane was measured by cross-sectional image, and the types of micro-contamination causing the fouling were observed and significant information was obtained. Within the DGGE study, the numbers of bacterial species in the sludge that cause fouling were determined. Findings in this study showed that the use of DMBR is very low in terms of cost compared to MBRs and is at least as successful as MBRs in terms of functionality. Based on this information it is expected to shed light on DMBR and MBR studies to be made in the future and also benefit on the use of DMBR in real scale installations.