A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Environmental Engineering and Management

Thesis (M.Eng.) - Asian Institute of Technology, 2016

A novel attached growth membrane bioreactor process was investigated for treatment of the polluted surface water with CODMn approximately 10 mg/Las a raw water in drinking water production. Firstly, Aquaporousgel (APG) and Polyvinyl alcohol gel (PVA-Gel) were compared to select the suitable carrier. According to pollutants removal and Monod{u2019}s kinetic coefficients analysis, PVA-Gel was the most suitable carrier with CODMn removal of 79% and NH3-N removal of 67.13%. Growth yield, maximum specific growth rate, half velocity constant and microbial decay rate of the PVA-Gel were 0.4999 gVSS/gCOD, 0.048 d-1, 9.923 mgCODMn/L, 0.00145 d-1 respectively. Secondly, six different filling ratio of PVA-Gel(5, 10, 15, 20, 30, and 50%) were applied in batch moving bed reactors. 15% filling ratio showed the appropriate performance among others with 72.08% of CODMn removal and 47.65% of NH3-N removal. As a results of preliminary studies, two laboratory scale reactors; attached growth membrane bioreactor with 15% PVA-Gel and conventional membrane filtration reactor with a submerged polytetrafluoroethylene hollow fiber membrane module with the pore size of 0.1 om were conducted. As compared to membrane filtration process, attached growth membrane bioreactor achieved two times higher CODMn removal with 54.1±10.5%. Almost complete ammonia removal was occurred with 84.7% by membrane filtration and 95.6% by attached growth membrane bioreactor. Turbidity was completely removed. However, UV254 removal was still low around15-20% in both reactors. For fouling study, attached growth membrane bioreactor showed the greater ability to reduce membrane fouling rate which correlated to the slower increasing of transmembrane pressure and longer operation period. The fouling layer was still proved to be a biofouling from the membrane resistance analysis. Hydraulic retention time at 2 hours was selected to be the most suitable operational condition by providing the greatest performance and constant effluent concentration. Lastly, Modified Stover-Kincannon model was a preferable model to apply for attached growth membrane bioreactor system with R2>0.92 in order to predict the treatment plant performance both CODMn and NH3-N removal, control the operational condition and optimize the plant volume design.