Low Energy Membrane Bioreactors for Decentralised Waste Water Treatment

Abstract

This research is part of PURATREAT project, an action taken by the European Commission (EC) to investigate waste water treatment (WWT) issues in the peri-urban areas of the Middle East and North Africa (MENA) countries. The main target of this work was to develop the application of, and to determine economic viability of, membrane bioreactor (MBR) technology as an alternative to conventional activated sludge (AS) processes for municipal waste water treatment (WWT). The research particularly focused on the long-term operation of three pilot submerged MBR systems - abbreviated as MBR1, MBR2 and MBR3 -, designed and constructed by three different membrane manufacturers. All trials were performed in Tunisia, in the city of Sfax, at the North Sfax “Office National de l' Assainissement” (ONAS) WWT site. ONAS is the country’s national sanitation utility. The MBR systems were tested under different combinations of operating conditions, namely solids residence times (SRTs) and hydraulic residence times (HRTs) and conclusions regarding treated permeate quality, membrane performance and energy consumption rates are drawn.

First, the capability of the MBR systems to produce treated permeate of the appropriate quality was tested. The treated permeate had to be suitable for reuse in unrestricted irrigation in Tunisia, therefore, chemical oxygen demand (COD) -1 concentration must be equal to or lower than 90 mg L and it has to be free from pathogens. Two different mixed-liquor suspended solids (MLSS) concentrations were applied. Initially, the MBR systems were operated under a low MLSS concentration -1 of about 4 - 5 g L . Under this low MLSS concentration, MBR1 failed to produce treated permeate of appropriate quality whereas MBR2 and MBR3 were successful. The COD concentration removal efficiency was only 71.4 % for MBR1, but it was 88 % for MBR2 and 87.7 % for MBR3. Then, all MBR systems were operated under a -1 higher MLSS concentration value of about 9 - 10 g L . Under this MLSS concentration, all MBR systems produced treated permeate of the appropriate quality. COD concentration removal efficiency was 89.4 % for MBR1, 89.7 % for MBR2 and 90.9 % for MBR3.

Then, the membrane performance was tested. This experiment was mainly conducted when the MBR systems were operated at the high MLSS concentration of about 9 - 10 g L . Real/Net membrane permeate fluxes (MPFs) were increased up to values that were no longer sustainable and membrane fouling phenomena appeared to be out of control. An average maximum sustainable net membrane permeate flux (MPF) for each MBR system was then estimated. Application of this net MPF could maximise the daily production of the treated permeate and, at the same time, a reliable long-term membrane performance could be achieved. The maximum sustainable net MPF was -2 -1 -2 -1 found to be equal to 13.77 L m h for MBR1 and 12.81 L m h for MBR2. With respect to MBR3, continual membrane fouling conditions during this experiment did not allow to predict an average maximum sustainable net MPF.

Finally, it was attempted to reduce the energy consumption rates below a specific -3 energy demand (SED) value of 3 kWh m , which is the current SED value of the full scale conventional AS plant and, therefore, the target for this research. Initially, short term power-analysis experiments were performed for MBR1 and MBR2. For MBR3 these experiments were not able to be conducted due to its three-phase power supply. The energy consumption rates per component were measured and SED values were calculated. At the same time, longer-term energy-analysis experiments were performed and SED values were re-calculated. By directly comparing the SED values taken by both sets of experiments, the short-term power-analysis experimental data was validated by the longer-term energy-analysis data. Finally, an Excel-based model, which was capable of predicting SED values for MBR1 and MBR2, was built. The model could predict SED values under different sets of operating conditions, e.g. varying SRTs/HRTs, but it was concluded that neither MBR1 nor MBR2 could produce treated permeate of the appropriate quality at SED values equal to or lower -3 than 3 kWh m , whilst simultaneously achieving stable membrane performance. However, a modified version of MBR1 was finally able to achieve the objectives of this research. Namely production of treated permeate of the appropriate quality, at a stable long-term membrane performance, and with SED values equal to or lower than 3 kWhm .

Date of Award	1 Apr 2010
Original language	English
Awarding Institution	University of Bath
Supervisor	Tom Arnot (Supervisor)