• 상하수도학회지
• /
• 제23권1호
• /
• pp.23-30
• /
• 2009

Slow sand filtration (SSF) was the first engineered/mechanical filtration process used in drinking water treatment. In SSF, untreated water slowly percolate through a bed of porous sand. Biological activity within the sand bed have the strongest influence on removal efficiency of pollutants by slow sand filtration. In this study, the microbial population distributions in slow sand filters operated at the various operation conditions was evaluated. The concentrations of $10^4$ to $10^5$ CFU per g dry wt. were observed. No significant differences were seen between the number of filter-covered materials. The data indicate that the temperature has affect on population distribution. Also, the light exposure was influenced on microorganism in slow sand filtration according to the heterotropic plate counts. The role of microorganism within the sand media requires further study.

• 상하수도학회지
• /
• 제22권4호
• /
• pp.467-473
• /
• 2008

Natural organic matter (NOM) removal by physico-chemical adsorption and biological oxidation was investigated in five slow sand filters with different media depths. Non-purgeable dissolved organic carbon(NPDOC) and $UV_{254}$ absorbance were measured to evaluate the characteristics of NOM removal at different filter depths. Removal efficiency of NOM was in the range of 10-40% throughout the operation time. At start-up of the filters packed with clean sand media, NOM was probably removed by physico-chemical adsorption on the surface of sand through the overall layer of filter bed. However, when Schumutzdecke layer was built up after 30 days operation, the major portion of NPDOC was removed by biological oxidation and/or bio-sorption in lower depth above 50 mm. NOM removal rate in the upper 50 mm filter bed was $0.82hr^{-1}$. It was about 20 times of the rate($0.04hr^{-1}$) in the deeper filter bed. Small portion of NPDOC could be removed in the deeper filter bed by both bio-sorption and biodegradation. SEM analysis and VSS measurement clearly showed the growth of biofilm in the deeper filter bed below 500 mm, which possibly played an important role in the NOM removal by biological activity besides the physco-chemical adsorption mechanism

• 상하수도학회지
• /
• 제19권3호
• /
• pp.338-347
• /
• 2005

Horizontal-Flow Roughing Filtration (HRF) is one of altemative pretreatment methods e.g. prior to Slow Sand Filtration (SSF). However, some of its limitations are that the effluent quality drops drastically at higher turbidity (>200 NTU) and at higher filtration rate (>1 m/h). To overcome these drawbacks, we suggested Direct Horizontal-Flow Roughing Filtration (DHRF), which is a modified system of Horizontal-Flow Roughing (HRF) by addition of low dose of coagulant prior to filtration. In this study to optimize the DHRF configuration, a conceptual and mathematical model for the coarse compartment has been developed in analogy with multi-plate settlers. Data from simple column settling test can be used in the model to predict the filter performance. Furthermore, the model developed herein has been validated by successive experiments carried out. The conventional column settling test has been found to be an handy and useful to predict the performance of DHRF for different raw water characteristics (e.g. coagulated or uncoagulated water, different presence of organic matter, etc.) and different inital process conditions (e.g. coagulant dose, mixing time and intensity, etc.). An optimum filter design for the coarse compartment (grain size 20mm) has been found to be of 3 m/h filtration rate with filter length of 4-4.5 m.