AWWA MTC57674

The objective of this research project was to develop and optimize enhanced biological phosphorus removal (EBPR) technology in membrane bioreactors (MBR). Further investigations for nitrogen removal were conducted to optimize the overall performance of MBR processes. The aim was to establish cost effective strategies for upgrading small sewage treatment units (up to 10,000 PE), as needed in some decentralized areas of Berlin. One MBR bench-scale plant of 200 to 250L (BSP) and two MBR pilot plants of 1 to 3m3 each (PP1 and PP2) were operated in parallel to a conventional EBPR wastewater treatment plant (WWTP Berlin-Ruhleben, Germany). The membrane bioreactors were tested with EBPR configurations in predenitrification (PP1), and post-denitrification mode without additional carbon source (PP2). The MBR pilot and bench-scale plants were operated under sludge ages of respectively 26 and 15 days. Eight extended measurement programs (profile measurements, mass balance) were carried out during the considered period of 8 months (September 2001 - April 2002). The effluent concentration of total phosphorus was very low and stable between 0.05 and 0.16 mg/L in both configurations. A comparison to the Ruhleben WWTP showed, that the effluent concentrations for COD, TN and TP were slightly lower for the MBR due to the absense of particulate fraction. Measurements along the multi-stage reactors showed that P-uptake occurred in the anoxic and the aerobic zones. The P-content of the biomass was about 2.1 - 2.6%P . With a nitrogen load of 0.01 kgN/kgTS.d an average total nitrogen concentration of 2.5mg/L was monitored in the effluent of post-denitrification compared to 9.2mg/L for pre-denitrification. Under identical loading conditions, nitrogen removal was greater in PP2 with post-denitrification. The denitrification rate under operation conditions in the first anoxic reactor (AX1) was in average 1.4 mgN/gVSS.hr for PP1 and 0.43 mgN/gVSS.hr for PP2. In the following anoxic reactors (AX2 and AX3), the denitrification rates were much lower for both configurations. Multi-stage reactor design ensured that nitrate removal was complete in all profile measurements. In all measurements, denitrification rate was lower in PP with 26 days than in BSP and conventional WWTP with 15 days. But further investigations are necessary to evaluate the influence of parameter such as sludge age, oxygen transfer in anoxic reactor, etc. The assessment of membrane system performance was not the main focus of this first project phase, and the membrane units were operated below their optimum operation ranges. The objective was to minimize the use of chemicals due to the potential application in remote areas. The two PVDF modules of the company Memcor were operated with 15 - 17 L/m^2.hr. Further data are requested to conclude whether process configurations (pre- and post-denitrification) or process conditions have some influence on fouling rate. The project will be continued until April 2003. The second project phase will evaluate maximum nitrogen load different SRT and impact of various conditions on membrane performance. Further trials with the BSP will be undertaken with another type of raw water coming from a decentralized area with a short retention and separate sewer. Includes 17 references, tables, figures.