• 상하수도학회지
• /
• 제29권3호
• /
• pp.337-345
• /
• 2015

Two step rapid filter system as a pre-treatment for the injected water into aquifer storage and recovery (ASR) in Korea was developed to reduce physical blockage and secure the volume of the injected water. First, single rapid sand filters with three different media sizes (0.4~0.7, 0.7~1.0 and 1.0~1.4 mm) were tested. Only two sizes (0.4~0.7 and 0.7~1.0 mm) satisfied target turbidity, below 1.0 NTU. However, they showed the fast head loss. To prevent the fast head loss and secure the volume of the injected water, a rapid anthracite filter with roughing media size (2.0~3.4 mm) were installed before a single rapid sand filter. As results, both the target turbidity and reduction of head loss were achieved. It was determined that the media size for a rapid sand filter in two step rapid filter system (i.e. a rapid anthracite filter before a rapid sand filter) was 0.7~1.0 mm. In addition, the effects of coagulant doses on the removal of natural organic matter (NOM), which might cause a biological clogging, were preliminarily evaluated, and the values of $UV_{254}$, dissolved organic carbon (DOC) and SUVA were interpreted.

• 상하수도학회지
• /
• 제20권6호
• /
• pp.813-822
• /
• 2006

Soluble manganese removal was analyzed as a function of filter media, filter depth, presence or absence of chlorination, and surface manganese oxide concentration in water treatment processes. Sand, manganese oxide coated sand (MOCS), sand+MOCS, and granular activated carbon(GAC) were used as filter media. Manganese removal, surface manganese oxide concentration, turbidity removal, and regeneration of MOCS in various filter media were investigated. Results indicated that soluble manganese removal in MOCS was rapid and efficient, and most of the removal happened at the top of the filter. When filter influent (residual chlorine 1.0mg/L) with an average manganese concentration of 0.204mg/L was fed through a filter column, the sand+MOCS and MOCS columns can remove 98.9% and 99.2% of manganese respectively on an annual basis. On the other hand, manganese removal in sand and the GAC column was minimal during the initial stage of filtration, but after 8 months of filter run they removed 99% and 35% of manganese, respectively. Sand turned into MOCS after a certain period of filtration, while GAC did not. In MOCS, the manganese adsorption rate on the filter media was inversely proportional to the filter depth, while the density of media was proportional to the filter depth.

• 상하수도학회지
• /
• 제24권3호
• /
• pp.341-349
• /
• 2010

It is well known that manganese is hard to oxidize under neutral pH condition in the atmosphere while iron can be easily oxidized to insoluble iron oxide. The purpose of this study is to identify removal mechanism of manganese in the D water treatment plant where is treating bank filtered water in aeration and rapid sand filtration. Average concentration of iron and manganese in bank filtered water were 5.9 mg/L and 3.6 mg/L in 2008, respectively. However, their concentration in rapid sand filtrate were only 0.11 mg/L and 0.03 mg/L, respectively. Most of the sand was coated with black colored manganese oxide except surface layer. According to EDX analysis of sand which was collected in different depth of sand filter, the content of i ron in the upper part sand was relatively higher than that in the lower part. while manganese content increased with a depth. The presence of iron and manganese oxidizing bacteria have been identified in sand of rapid sand filtration. It is supposed that these bacteria contributed some to remove iron and manganese in rapid sand filter. In conclusion, manganese has been simultaneously removed by physicochemical reaction and biological reaction. However, it is considered that the former reaction is dominant than the latter. That is, Mn(II) ion is rapidly adsorbed on ${\gamma}$-FeOOH which is intermediate iron oxidant and then adsorbed Mn(II) ion is oxidized to insoluble manganese oxide. In addition, manganese oxidation is accelerated by autocatalytic reaction of manganese oxide. The iron and manganese oxides deposited on the surface of the sand and then are aged with coating sand surface.

• 상하수도학회지
• /
• 제21권4호
• /
• pp.503-508
• /
• 2007

Small scale D-water treatment plant(WTP) where has slow sand filtration was using raw water containing high concentration of manganese (> 2mg/l). The raw water was pre-chlorinated for oxidation of manganese and resulted in difficulty for filtration. Thus, sometimes manganese concentration and turbidity were over the water quality standard. Two stage rapid manganese sand filtration pilot plant which can treat $200m^3/d$ was operated to solve manganese problem in D-WTP. The removal rate of manganese and turbidity were about 38% and 84%, respectively without pH control of raw water. However, when pH of raw water was controlled to average 7.9 with NaOH solution, the removal rate of manganese and turbidity increased to 95.0% and 95.5%, respectively and the water quality of filtrate satisfied the water quality standard. Manganese content in sand was over 0.3mg/g which is Japan Water Association Guideline. The content in upper filter was 5~10 times more than that of middle and lower during an early operation but the content in middle and lower filter was increased more and more with increase of operation time. This result means that the oxidized manganese was adsorbed well in sand. Rapid manganese sand filter was backwashed periodically. The water quality of backwash wastewater was improved by sedimentation. Thus, turbidity and manganese concentration decreased from 29.4NTU to 3.09NTU and from 1.7mg/L to 0.26mg/L, respectively for one day. In Jar test of backwash wastewater with PAC(Poly-aluminum chloride), optimum dosage was 30mg/L. Because the turbidity of filtrate was high as 0.76NTU for early 5 minute after backwash, filter-to-waste should be used after backwash to prevent poor quality water.

• 상하수도학회지
• /
• 제21권2호
• /
• pp.203-214
• /
• 2007

This study aimed at researching the process selection for two-stage and dual media filtration system, as a technology substituting the existing sand filter without expanding the site when retrofitting an old filter bed or designing a new one. In order to select the process for optimum complex filtration system, three running conditions have been tested. Test results demonstrated that Run 3 in which the 1st stage was filled with anthracite and coarse sand, and the 2nd stage was filled up with activated carbon and fine sand reduced the head loss and the load of turbidity substances. Also, Run 3 showed better performance in removing TOC, particle counts, THMFP and HAAFP, compared to other two conditions. 99 % of Cryptosporidium was removed. Bisphenol-A was rarely removed from the 1st stage of coarse sand and 2nd stage of fine sand, but 99 % of it was removed from the 2nd stage of activated carbon. In conclusion, when it is required to retrofit an old rapid filter bed or design a new one for the purpose of improving filtration performance, the following two-stage and dual media filtration system is suggested: the 1st stage of filter bed needs to be filled up with coarse sand to remove turbidity as the pretreatment for extending duration of filtering, the top part of 2nd stage needs to be filled up with granular activated caron to remove dissolved organic matters and others as the main process, and finally the bottom part of 2nd stage needs to be filled up with fine sand as the finishing process.

• 한국물환경학회지
• /
• 제18권3호
• /
• pp.253-260
• /
• 2002

The purpose of this research was to describe the mechanisms and prevention of initial degradation in turbidity of the sand filter effluent. The method used was by adding a coagulant (PAC) to the sand filter after backwashing as a means of reducing turbidity. It was found that adding 80 mg/L of PAC solution to the sand filter was very effective in improving the initial effluent turbidity. A turbidity removal efficiency of 99 % was observed in the initial term period as compared to a 70% efficiency without PAC addition. The PAC solution added to the sand filter resulted in high aluminum concentration at the upper layer as compared with the bottom layer of the sand filter column. A change in the zeta potential to a strong positive-ions at upper layer was observed at this time but only a small change was obtained at the bottom. This result showed that the zeta potential of the sand was changed to positive with PAC coating. The effect of pH on zeta potential with PAC addition was also investigated. Zeta potential was greatly changed to positive-ion at pH 4~6. A series of experiments was then conducted in this study to optimize the pH of the PAC solution to be added to the sand filter after backwashing. The removal efficiency of turbidity was found to be highest at pH 5. This result suggested that hydrolyzed aluminium species attached to the surface of the sand enhanced the removal of turbidity of the effluent.

• 상하수도학회지
• /
• 제21권3호
• /
• pp.359-366
• /
• 2007

Slow sand filtrations have been widely used for water treatment in small communities, however their capacity is often limited by high turbidity in the raw water. For this reason, several pre-treatment facilities were required for a slow sand filter. Turbidity removal from the highly turbid raw water was investigated in roughing filters as a pre-treatment process. The roughing filters followed by rapid mixing tank were operated in the form of a contact filtration. In several jar tests, the predetermined optimum aluminium sulfate (alum) doses for turbid water of 30 and 120NTU were 30 and 50mg/L, respectively. At the optimum alum dose, physically optimum parameters including G value of $220sec^{-1}$ and rapid mixing time of 3 minutes were applied to the contact filtration system. Without addition of alum, the filtrate turbidity from the roughing filters, packed respectively with different media such as sand, porous diatomite ball and gravel, was in the range of 5~30NTU at filtration velocities of 30 and 50m/day. However, the application of a contact filtration to roughing filters showed stably lower filtrate turbidity below 1.0NTU at filtration velocity of 30 m/day. Although the filtration velocity increased to 50m/day, filtrate turbidity was still below 1.0NTU in both single and double layer roughing filters. At influent turbidity of 120NTU, the filtrate turbidity was over 5 NTU in the triple layer roughing filter, which shortened the filter run time. The flocs larger than $10{\mu}m$, formed in the rapid mixing tank, were almost captured through the roughing filter bed, while the almost flocs smaller than $10{\mu}m$ remained in filtrate.

• 한국물환경학회지
• /
• 제20권6호
• /
• pp.652-656
• /
• 2004

Both surface wash and backwash are considered as one of the most important methods that can improve the filtration efficiency in the existing water treatment plant. This study has mainly focused on the improvement of filtering efficiency by controlling surface wash and backwash time, and water level before backwash (when drained up to the trough, when drained up to 10 cm above filter bed, and when drained below 10 cm filter bed). Filtration efficiency was shown a little bit of differences depending on the operating conditions like surface wash injection pressure, the distance control between filter bed and the facility, and the types of surface wash. When the water level before backwash was reached up to 10 cm below filter bed after draining, however, the filtration velocity and the turbidity removal efficiency in the filter bed was improved. When the surface wash followed by backwash is longer, it showed a similar result. Because the proper adjustment of surface washing time makes filtration efficiency higher, therefore, it is necessary to set up the backwash time moderately.

• Coupled systems mechanics
• /
• 제4권2호
• /
• pp.157-172
• /
• 2015

Control of potable water treatment plant (PWTP) is nowadays based on experience. The aim of this article is to show that model based control of treatment process is more efficient than process operation based on experience. Stimela environment is used for modeling of processes of potable water treatment. Application of the model was conducted on PWTP "Crkvice" in Zenica (BiH). This plant has used conventional rapid sand filters. By effective application of the model it is determined the optimal filter run time for different input turbidity of raw water. This results in the possibility of reducing the consumption of backwashing water, lower costs for its pumping and reducing the amount of coagulants. In the existing practice, based on experience, these benefits are not used.

• 한국환경보건학회:학술대회논문집
• /
• 한국환경보건학회 2005년도 국제학술대회
• /
• pp.381-385
• /
• 2005

Characterization of particle behavior is becoming more important in performance evaluation of water treatment system as well as in operation of the system because conventional parameter, turbidity has lack of explaining ability on small sized microorganisms such like Cryptosporidium etc. Accordingly, particle counter has been introduced in evaluation and operation of the treatment system. However researches on the relationship between turbidity, particle count and/or different sand/anthracite sizes have not been concurrent. Therefore in this study, the relationship was investigated to improve performance evaluation of sand filter so as to help choosing sand/anthracite effective size as a design parameter of water treatment facility. According to the results, too small or too large effective size media filter reached to turbidity limit(0.1 NTU)earlier. However, because shallow sand layer may cause early breakthrough, the depth of sand layer should be provided enough in order to compromise water quality and productivity.