Could nonwoven filtration material be used as a viable, cheaper alternative to hollow fibre membranes in submerged membrane bioreactors, without a loss in performance? Filtration + Separation takes a look at one research report that thinks so.
Nonwoven for the MBR – why bother?
The Membrane Bioreactor (MBR) is an evolving process that combines membrane technology with biological technology and uses membrane separation to replace the conventional secondary sedimentation tank.
Compared with the conventional activated sludge process, MBR technology has a better performance in water treatment applications because of high sludge concentration, high pollutant removal efficiency, better effluent water quality, low production of excess sludge, simple operation and facility management.
In recent years, the MBR process has gained more and more interest from the water treatment sector. Because of the high efficiency isolation function of the membrane, the microbe can be withheld from the bioreactor completely, separating the hydraulic retention time and sludge retention time and eliminating the problem of sludge bulking that exists with the traditional activated sludge process. However, the MBR process isn't perfect. Issues arise such as the problem of membrane fouling, not to mention the high capital cost of the membrane itself. To date, these issues have restricted the commercial take up of MBR on a large scale.
To find a way of reducing the capital costs of the membrane bioreactor process, a new type of MBR with nonwovens as the filtration medium has been designed and tested by researchers. The performance of a domestic wastewater treatment system using this nonwoven bioreactor was then compared with the performance of a conventional hollow fibre membrane bioreactor – based on a polypropylene membrane. The experiment was designed to answer the question: If the filtration medium of an MBR was made from nonwovens (by a melt blowing process), could this potentially reduce the cost of the membrane, making the capital cost of the MBR cheaper and increasing its take up in wastewater applications?
Designing the experiment
The bioreactor was made of polymethyl methacrylate and had a working volume of 0.028m3. The air was supplied through an aeration pipe with an air compressor, and the airflow rate was regulated at 20L/min-1 with the flowmeter.
• The nonwovens used in the experiment were of pore size 3 and 5µm, and made of polypropylene.
• The hollow membrane had a pore size of 0.4µm.
The influent was synthetic domestic wastewater, prepared with sugar and nutrient salt.
Removal performance – how did the membrane elements compare?
There was little difference in the effluent water quality between the nonwoven bioreactor and the membrane bioreactor:
• The permeate BOD5 ranged from 3 to 7mg/L for both the nonwoven and hollow membrane.
• The average permeate TOC was very low at <3mg/L and the average COD was at <20mg/L.
• The variation in influent organic carbon had little effect on effluent removal efficiency.
• The average permeate NH4+-N was at <1.5mg/L and from day 7, the influent declined to pH 4 because of the acidification taking place in the influent storage tank. The removal of ammoniac nitrogen also dropped and after adjusting the influent pH figure to about 7.5, the permeate ammoniacal nitrogen recovered to 1 mg/L.
It was obvious that the pore size of nonwovens had little effect on the organic carbon reduction, probably due to the formation of a dynamic layer on the nonwoven membrane surface. The effluent water quality of the nonwoven bioreactor could be kept at a stable level for the effective interception of the nonwovens itself and the surface layer. The removal performance indicated that the nonwoven bioreactor was efficient in treating domestic wastewater.
Fouling performance – how did the membrane elements compare?
Figures 3.1 to 3.4 above show that a great deal of fibre constituted the surface of the nonwovens and that the interspaces of the nonwoven fibre were full of microbes after being polluted. Figures 3.5 and 3.6 showed that the hollow membrane had a smooth surface with pores distributed over the surface and that a cake layer was formed on the surface after being polluted.
In the experiment, 5µm nonwovens exhibited greater fouling than 3µm nonwovens and a hollow membrane, when the bioreactor ran in constant effluent flux (20L/m2·h) mode. (A previous study has pointed out the nonwovens exhibited greater fouling than the membrane).
Cleaning performance - how did the membrane elements compare?
Based on the specific characteristics of nonwovens, two different cleaning procedures were tested for the nonwovens used in this study.
• Flushing the nonwoven surface with tap water.
• Flushing the membrane surface with tap water after wiping the floc on the surface of the nonwovens, then immersing it in 0.3%(wt) NaClO solution for twelve hours.
The change of nonwoven surface flux before and after cleaning demonstrates that nonwoven should be cleaned through chemical and physical methods to eliminate the irreversible fouling after operating for a long time. In particular, cleaning with 0.3%(wt)NaClO solution was demonstrated to be an effective method for nonwoven surface flux recovery.
What does the experiment conclude?
• The effluent permeated from the nonwoven bioreactor (3µm and 5µm), and the hollow fibre membrane MBR, showed little difference at the same reaction condition. The average effluent contained COD<20 mg/L-1, TOC<3 mg/L-1, BOD5<6 mg/L-1, NH3-N<1.5mg/L-1. This data indicates that the nonwoven bioreactor was efficient in treating domestic wastewater.
• According to the difference in filtration resistance and the different conditions of the nonwoven surface fouling, it appeared that nonwoven fouling was mainly attributed to internal fouling. This means that the 3µm nonwoven bioreactor exhibited a greater advantage than the 5µm nonwoven bioreactor.
• The experiment also showed that nonwovens should be cleaned through chemical and physical methods, to eliminate the irreversible fouling that operates after a long period of operation. In particular, cleaning with a 0.3%(wt)NaClO solution was demonstrated to be an effectively method for nonwoven surface flux recovery.
• Both the nonwoven bioreactor and hollow MBR exhibited a high removal efficiency of COD, TOC, NH4+-N on domestic wastewater treatment, with no colour, smell or SS contained in the effluent water. Nonwovens could thus be considered suitable to be used in the MBR process for domestic wastewater treatment.