WaterSubject

Membrane bioreactors (MBRs) represent a newly developed wastewater treatment process in which secondary clarifier is replaced with membrane filtration devices. As the activated sludge is filtered by the physical barrier of a membrane, effluent does not contain suspended solids and the mixed liquor suspended solids (MLSS) level can be maintained very high (5000–30 000 mg/L). Consequently, the overall system becomes more compact and sludge retention time in the aeration basin can be increased to 20–100 days.

One of the biggest limitations of cross-flow-type MBR processes is membrane fouling. This results in higher operating and initial capital costs. Although this problem was alleviated by the invention of submerged-type membrane modules in the late 1980s, membrane fouling still remains a significant hurdle in the MBR process.

Various approaches have been tried to reduce membrane fouling such as (1) intermittent suction, (2) backflushing, (3) module design improvement, (4) optimization of aeration, etc. Combinations of these methods have reduced the costs of the MBR process significantly. However, the overall cost of the MBR process can be reduced further by decreasing membrane fouling.

It has been shown that soluble microbial products (SMPs) are major membrane foulants in MBR processes. Although some inorganic coagulants have been known to coagulate the SMPs and decrease membrane fouling (Lee 2001), excess demand of coagulant and increased mixed liquor production have limited their commercial success.

Recently, new membrane performance enhancer (MPE) products were developed to increase flux by reducing membrane fouling. These revolutionary products coagulate SMP and deactivate it. Lab- and pilot-scale experiments have shown that the membrane fouling rate was significantly reduced under high flux conditions and the process could be operated at an extremely high suspended solid level, e.g. 50 000 mg/L. In addition, permeate chemical oxygen demand (COD) was improved by approximately 30%, whereas no toxic effects on bioactivity were found.

In this study, MPE50 was tested in full-scale MBR plants to confirm the results obtained with the lab and pilot experiments. A method to determine the initial dosage will be demonstrated in addition to the pilot- and full-scale results.

Jar tests were performed to determine the initial dosage of MPE50. Different amounts of MPE50 were added to 500 mL of mixed liquor samples with vigorous mixing at 100–150 r.p.m. After 1–2 min of vigorous mixing, the mixture was slow mixed for 5–10 min. After 10 min settling time, a sample of supernatant was carefully obtained from the top layer and centrifuged at 3200 g for another 10 min. As most of the unreacted SMP exists in the supernatant of the centrifuged sample, the COD was measured in this layer using a spectrophotometric method (Hach Co., Loveland, Colorado, USA). The optimum dosage was determined by charting COD as a function of MPE50 dosage.

MPE50 was added directly to aeration basins or to the mixed liquor flow going to the aeration basin, depending on the turbulence at the injection point. The product was diluted one to three times in water. The extent of membrane fouling was monitored by measuring the transmembrane pressure (TMP). If necessary, additional MPE50 was added daily or intermittently to improve performance and to compensate for the loss of MPE owing to mixed liquor removal.